PSYCHIATRY

THE STATE OF THE ART

Volume 2
Biological Psychiatry,
Higher Nervous Activity

Volume 1
PSYCHIATRY
THE STATE OF THE ART

Volume 1 CLINICAL PSYCHOPATHOLOGY

NOMENCLATURE AND CLASSIFICATION

Volume 2 BIOLOGICAL PSYCHIATRY, HIGHER NERVOUS ACTIVITY

Volume 3 PHARMACOPSYCHIATRY

Volume 4 PSYCHOTHERAPY AND PSYCHOSOMATIC MEDICINE

Volume 5 CHILD AND ADOLESCENT PSYCHIATRY,

MENTAL RETARDATION, AND GERIATRIC PSYCHIATRY

Volume 6 DRUG DEPENDENCE AND ALCOHOLISM, FORENSIC

PSYCHIATRY, MILITARY PSYCHIATRY

Volume 7 EPIDEMIOLOGY AND COMMUNITY PSYCHIATRY

Volume 8 HISTORY OF PSYCHIATRY, NATIONAL SCHOOLS, EDUCATION

AND NEW TRENDS IN PSYCHIATRY AND TRANSCULTURAL
PSYCHIATRY
PSYCHIATRY
THE STATE OF THE ART

Volume 2
Biological Psychiatry,
Higher Nervous Activity

Edited by
P. PICHOT
Academie de Paris
Universite Rene Descartes
Paris, France

and
P. BERNER, R. WOLF, and K. THAU
University of Vienna
Vienna, Austria

Volume 1

SPRINGER SCIENCE+BUSINESS MEDIA, LLC

 Library of Congress Cataloging in Publication Data
World Congress of Psychiatry (7th: 1983: Vieona, Austria)
Biological psychiatry, higher nervous activity.
(Psychiatry, the state of the ar!; v. 2)
"Proceedings of the VII World Congress of Psychiatry, held July 11-16, 1983, in Vien·
na, Austria"-T.p. verso.
Includes bibliographical references and indexes.
1. Biological psychiatry-Congresses. 2. Higher nervous activity-Congresses. 1.
Pichot, Pierre. II. Title. III. Series. [DNLM: Biological Psychiatry-congresses. 2. Higher
Nervous Activity-congresses. W3 W05385 7th 1983b / WM 100 W9238 1983b]
RC327.W63 1983 616.89'071 84-26481
ISBN 978-1-4684-8331-4 ISBN 978-1-4684-8329-1 (eBook)
DOI 10.1007/978-1-4684-8329-1

Proceedings of the VII World Congress of Psychiatry,

held July 11-16, 1983, in Vienna, Austria

© 1985 Springer Science+Business Media New York

Origina11y published by Plenum Press, New York in 1985
Softcover reprint of the hardcover 1st edition 1985

AII rights reserved

No par! of this book may be reproduced, stored in a retrieval system, or transmitted,
in any form or by any means, electronic, mechanical, photocopying, microfilming,
recording, or otherwise, without written permission from the Publisher
PREFACE

The purpose of the World Psychiatric Association is to coordinate

the activities of its Member Societies on a world-wide scale and to
advance enquiry into the etiology, pathology, and treatment of mental
illness. To further this purpose, the Association organizes mono- or
multithematic Regional Symposia in different parts of the world twice
a year, and World Congresses dealing with all individual fields of
psychiatry once every five or six years. Between these meetings the
continuation of the Association's scientific work is assured through
the activities of its specialty sections, each covering an important
field of psychiatry.

The programs of the World Congresses reflect on the one hand the
intention to present the coordinating functions of the Association and
on the other to open a broad platform for a free exchange of views.
Thus, the VII World Congress of Psychiatry, held in Vienna from July
11 to 16, 1983, was composed of two types of scientific events - those
structured by the Association and those left to the initiative of the
participants. The first type comprised Plenary Sessions, planned by
the Scientific Program Committee, and Section Symposia, organized by
the WPA sections; the second embraced Free Symposia, free papers,
video sessions, and poster presentations prepared by the participants.
Altogether, 10 Plenary Sessions, 52 Section Symposia, and 105 Free
Symposia took place, and 78 free papers and poster sessions and 10
video sessions were held.

The editors of the Proceedings of the VII World Congress of Psy-

chiatry were immediately faced with two major problems, namely how to
deal with such a great number of presentations and how to present them
to the reader. The only way to solve the first difficulty was to re-
strict the Proceedings to Plenary Sessions and Symposia. The second
obstacle was surmounted by grouping the Plenary Sessions and Symposia
according to their scientific content, which meant waiving the chrono-
logical order of the Congress. In order to achieve reasonable uni-
formity in the lengths of the volumes, it was not possible to devote
each of the eight books comprising the Proceedings to a single theme.
Nevertheless, we hope that the final arrangement will enable colleagues
interested in only certain subjects to restrict their purchases to the

v
particular volume or volumes of their choice. The Proceedings in
their entirety, however, represent a complete and comprehensive
spectrum of the current areas of concern in psychiatry - the state
of the art.

We are greatly indebted to our colleagues Rainer Wolf and

Kenneth Thau. Their untiring efforts made the publication of these
Proceedings possible.

Peter Berner

Secretary General, WPA

at the time of the VII
World Congress of Psychiatry
President, Organizing Committee
VII World Congress of
Psychiatry
Chief Editor, Congress Proceedings
ACKNOWLEDGMENTS

First and foremost, we should like to express our sincere

appreciation to all colleagues whose scientific contributions com-
prise the content of these Proceedings.

We should also like to thank the immediate administrators of

the VII World Congress of Psychiatry (Congress Team International),
as well as the staff of the Vienna Secretariat of the World Psy-
chiatric Association, for their collaboration in the compilation of
this publication.

We should finally like to explain that, for technical reasons

connected with the actual printing process, it has not been possible
in every instance to eliminate minor typing errors.

Various reasons also prevented the compilation of all chapters

in exact conformity with the presentations as contained in Plenary
and Symposium Sessions.

Despite these problems, we hope that our aim to structure the

content of the individual volumes as clearly as possible has met
with an adequate measure of success.

vii
INTRODUCTORY REMARKS

The World Psychiatric Association was born out of the Organizing

Committee of the Horld Congress of Psychiatry. The first World Con-
gress, held in Paris in 1950, was an event of the utmost importance.
For the first time, psychiatrists of the whole world met to exchange
their ideas and experiences and to promote the progress of our spe-
cialty. It later became obvious that such large congresses., convening
every five or six years, needed to be complemented by a more permanent
organization and by more frequent meetings smaller in scope and of a
more specialized nature. The national psychiatric societies decided
on the creation of a World Association which could assume all the re-
sponsibilities connected with such a complex task, I had the honor to
be elected President of this Association at the VI World Congress in
Honolulu and to hold this responsibility for six years until the advent
of the Vienna Congress.

Whatever the importance of the various functions of the WPA, the

organization of these World Congresses has remained its major task.
It has become fashionable to criticize World Congresses because they
attract too many participants, because the scientific presentations
are not always of the highest quality, and because the mult.iplicity of
the subjects discussed in simultaneous sessions obliges the partici-
pants to limit attendance to only part of the entire program. Some of
the criticisms may be justified, but the fact remains that such con-
gresses fulfill an important function. The majority of the psychia-
trists of the world are not highly specialized research workers but
practitioners. Many of them live in countries where they are rela-
tively isolated and where there is little opportunity for scientific
interchange. The World Congresses, by presenting not only the latest
technical discoveries but also general surveys through leading spe-
cialists in the different fields of psychiatry, allow every partici-
pant to keep abreast of the state of the art. There is no better
opportunity to become acquainted with developing trends, and personal
experience of this type cannot be replaced by the reading of scientific
journals. Of course, the value of such Congresses depends on the care
with which the program is prepared. The readers of these Proceedings
will have the opportunity to convince themselves that the Austrian
Organizing Committee, under the chairmanship of Prof. Peter Berner,

ix
Secretary General of the WPA at the time of the Congress, has attained
this goal, and that the scientific quality of the papers presented and
now printed is worthy of the tradition of our World Congresses of
Psychiatry.

Pierre Pichot

President, WPA
at the time of the VII
World Congress of Psychiatry
President, Scientific Committee

X
CONTENTS

BIOLOGICAL PSYCHIATRY
BIOLOGICAL ASPECTS - ORGANIC BRAIN SYNDROMES

Organic Brain Syndromes in DSM-III

Z. J. Lipowski

Neuropeptide s and Chemical Messengers in the Aging Brain 9

E. W. Busse

Cross Sectional and Longitudinal Studies of Organic

Brain Syndromes • • • • • • • • • • • • • • • 15
K. Kryspin-Exne r

BIOLOGICAL ASPECTS - FUNCTIONAL PSYCHOSES

Today's Problems and Results of Biological Research

in Schizophreni a • • • • • • • • • • • • • • 21
H. Hippius and M. Ackenheil
Evidence for Alterations of Brain Morphology and Function in
Subgroups of Schizophreni c Subjects • • • • • • • • 37
G. Sedvall, G. Bergstrand, L. Bjerkenstedt , H. Nyback,
G. Oxienstierna and F.-A. Wiesel
Psychobiolog ical Studies of Mania • 43
W. E. Bunney, Jr. and B. Garland

Biochemical and Morphologica l Abnormalitie s in Paranoid and

Nonparanoid Schizophreni a • • • • • • • • • • • • 51
R. Takahashi

xi
Peptides in Major Psychoses 57
0. J. Rafaelsen

Recent Trends in Biological Depression Research • • • • • • 61

H. M. van Praag

GENETIC ASPECTS OF PSYCHIATRY

Genetics of Affective Psychoses - Some Methodological

Considerations 67
J. Angst

Trait, State, HPA Activity in Depression 77

G. Winokur, B. Pfohl and B. Sherman

Heredity-Environment in Schizophrenia • 85
F. Schulsinger, J. Parnas, H. Schulsinger,
Th. W. Teasdale and S. A. Mednick

X-Linked Inheritance in Affective Disorders 95

J. Mendlewicz
Genetics of Psychogenic Diseases 101
H. Schepank

NEW PROSPECTS IN THE TREATMENT OF DEPRESSION

Preclinical Pharmacology of Antidepressants • • • • • • • • 105

w. Haefely
MAO-Inhibitors Revisited I: Andrenaline (A) and Noradrenaline
(NA) in Cerebrospinal Fluid (CSF) in Isocarboxazide
treated Rats • • • • • • • • • • • 111
A. Gjerris, D. I. Barry, N. J. Christensen
and 0. J. Rafaelsen

MAO-Inhibitors Revisited II: Clinical Implications 115

0. J. Rafaelsen and A. Gjerris

Serotonin in Depression and Suicide 121

H. M. van Praag

New Prospects in the Treatment of Depression 151

P. Turner

xii
 CLINICAL AND RESEARCH ASPECTS
OF AFFECTIVE DISORDERS

Biological Markers and Antidepressant Treatment

Response • • • • • • • • • • • • • • • • • • • • • • 159
D. G. Ostrow, A. Okonek, R. Gibbons, R. Cooper
and J. M. Davis

EEG Sleep in Borderline Personality Disorder • • • • 165

H. W. Lahmeyer, E. Val, F. M. Gaviria, and B. R. Prasad

Affective Disorders and Borderline Personality 171

E. Val, F. M. Gaviria, H. W. Lahmeyer, B. Prasad
and M. Weiler

Social Aspects of Depression 177

J. A. Flaherty and F. M. Gaviria

PATHOCHEMICAL MARKERS IN MAJOR PSYCHOSES

The Significance of Psychopathological Classification in

Interpreting Biochemical Findings • • • • • • • 183
E. Gabriel

Metabolic Disturbances in Psychiatric Disorder:

The Question of Diagnostic Speciticity 187
C. M. Banki, M. Arat6 and Z. Papp

Disturbances in Serine-Glycine Metabolism in Relation to

Acute Psychoses with Psychedelic Symptoms • • • 193
J. Bruinvels and L. Pepplinkhuizen

Peptides and Amines in Affective Disorders 197

A. Gjerris and L. J. Rafaelsen

Morphological Backgrounds of Pathochemical Studies in

Major Psychoses • • • • • • • • • • • • • • 203
K. Jellinger

Dopamine Function and Neuroleptics in Schizophrenia

Post-Mortem Studies of Human Brain Tissue • • • 209
G. P. Reynolds

Neuroendocrine and Receptor Binding Studies

in Schizophrenia • • • • • • • • • • 215
M. Ackenheil, M. Albus, B. Bondy, F. Muller-Spahn,
U. Munch and D. Naber

xiii
Decreased CSF Norepinephrine and Monoamine Metabolites in
Schizophrenic Patients with Brain Atrophy as
Shown with CT Scans • • • • • • • • • • • • • • • • • 221
D. P. van Kammen, L. S. Mann, M. Scheinin,
P. T. Ninan, W. B. van Kammen and M. Linnoila

STEROIDS IN PSYCHIATRY

Role of Testosterone in Male Sexual Impotence

Patients • • • • • • • • • • • • 223
0. Benkert and F. Holsboer

Psychotropic Effect of Combined Estrogen-Vit B6 Treatment

in Endogenously Depressed Females • • • • • • • • • • 229
F. Holsboer, L. Meier, A. Kreuz and 0. Benkert
Oestrogen Modulation of Dopamine Receptors: Autoreceptor
Implication for Antidepressant Activity • • • • • • • 233
M. P. Piccardi, M. Del Zompo, M. Meloni
and G. U. Corsini
Neuroleptic-Like Activity of Estrogens 237
T. Di Paolo, P. Bedard and F. Labrie
Interactions between Gonadal Steroids and
Neurotransmitters •• 243
G. Koster and H. Breuer

FRONTIERS IN PSYCHONEUROENDOCRINOLOGY
Neuroendocrine Involvement in Therapeutic Mechanisms of
Neuroleptic and Antidepressant Drugs: Studies of
Thyroid Axis • • • • • • • • • • • • • • • • • • 249
G. Langer, H. Aschauer, M. S. Keshavan, G. Koinig,
F. Resch and G. Schoenbeck
Pharmacoendocrinology and its Clinical Relevance 257
G. Laakmann
Chronobiology of Affective Disorders: Alterations in
Circadian Secretion of Pituitary and Pineal
Hormones in Bipolar and Unipolar Depression 263
J. Mendlewicz and D. P. Linkowski
Neuroendocrine Control of Pituitary Function in
Extrapiramidal Disorders • • • • • • • • • 273
T. Caraceni, P. Giovannini, E. Parati, G. Scigliano
and E. E. MUller
Neuroendocrinological Profile of Major Depression 279
C. N. Stefanis, B. Alevizos, C. Soldatos,
G. Papadimitriou, G. Tolis and M. Liparaki

POSITRON EMISSION TOMOGRAPHY

RO 15 1788-11c: A Specific Radioligand for an "In Vivo"

Study of Central Benzodiazepine Receptors, by
Positron Emission Tomography • • • • • • • • 289
M. Maziere, P. Hantraye, B. Guibert, M. Kaijima,
C. Prenant, J. Sastre, M. Crouzel, R. Naquet
and D. Comar
PET Studies During Hypnosis and Hypnotic Suggestion • • • • 293
L. Baer, R. H. Ackerman, 0. S. Surman, J. A. Correia,
J. L. Griffith, N. M. Alpert and T. P. Hackett
A Study of Epilepsy and Interictal Psychosis Using the
Oxygen-15 Inhalation Technique and Positron Emmission
Tomography • • • • • • • • • • • • • • • • • • 299
B. Gallhofer, M. R. Trimble, R. S. J. Frackowiak,
R. J. S. Wise and T. Jones

PET Studies on Brain Energy Metabolism and Dopamine Receptors

in Schizophrenic Patients and Monkeys • • • • 305
G. Sedvall, G. Blomqvist, T. DePaulis, E. Ehrin,
L. Eriksson, L. Farde, T. Greits, C. G. Hedstr5m,
D. H. Ingvar, J.-E. Litton, J. L. G. Nilsson,
S. Stone-Elander, L. Widen, F.-A. Wiesel and G. Wik

LATERALITY AND PSYCHOPATHOLOGY

Dynamic Hemispheric Imbalance in Schizophrenia 313

J. Gruzelier

Event-Related Potentials and Cognitive Deficits in

Schizophrenia • • • • • • • • • • • • 321
T. Kameyama, 0. Saitoh, K.-I. Hiramatsu, S.-I. Niwa,
K. Rymar and K. Itoh

XV
 SEROTONIN AND DISTURBANCES OF MOOD

Brain Serotonin, Mood-Regulation and Aggression-

Regulation . . . . . . . . _. . . . . . . . 327
H. M. van Praag

New Trends in the Pharmacology of Central Serotoninergic

Receptors . . . . . . . . . . . . . . . . . . •. 341
H. Gozlan, M. Hall, s. Bourgoin, S. El Mestikawy,
L. Pichat and M. Hamon
Indalpine, a New Serotonin Uptake Inhibitor ••• 347
A. Darragh, M. Kenny, R. Lambe, T. Lenehan, I. Brick
and C. Maulet
A Double Blind Out Patient Comparative Trial of
Indalpine Versus Mianserin . • • • • • • 353
B. Martin and G. J. Naylor
The Comparative Effects of Indalpine and Amitriptyline
on Saliva Production • • • • • • • • • • • . •• 361
M. Peet

PSYCHOBIOLOGY OF DEPRESSION -
RECENT FINDINGS AND THEORETICAL MODELS

Entrainment and Masking in the Chronobiology

of Depression • • • . • . • • • • • • 367
D. von Zerssen, G. Dirlich, P. Doerr, R. Lund,
K. M. Pirke and H. M. Emrich

The Origin of Early REM Sleep Episodes in Depression

and other Conditions •••••••••• 371
H. Schulz, R. Lund and S. Volk
Sleep Structure and Neuroendocrine Rhythms 377
D. J. Kupfer and D. B. Jarrett

Neuroendocrinological Findings in Depression Compared

with Findings in Malnutrition, Cushing's
Syndrome and Emotional Stress . 383
M. Berger, P. Doerr, J.-Ch. Krieg, K.-M. Pirke
and D. von Zerssen

ACTH and Cortisol Secretion in Cushing's Disease and in

Endogenous Depression: Indication of a Common
Pathway? • • • • • • . • • • • • • • • • • • . • 389
K. H. Voigt, s. Bossert, St. Bretschneider, A. Bliestle
and H. L. Fehm

xvi
Effect of Nitrous Oxide on Depressive Patients and
Volunteers • • • • • • • • • • • • 397
M. A. Gillman, N. Matussek and F. J. Lichtigfeld

THE OLD AMINE THEORY AND NEW ANTIDEPRESSANTS

Relevant Selectivity of Antidepressants • • • • • • • • • • 405
S. A. Montgomery

The Serotonin-Noradrenaline Link-Hypothesis of Affective

Disorders . . . . . . . . . . . . . . . . . . . . . . 411
F. Sulser

BIOLOGY OF MANIA
Biochemistry of Mania • • • • • • • • • • • 417
R. M. Post, D. R. Rubinow, P. W. Gold and M. Linnoila
The Psychopharmacology of Mania: Towards a Unifying
Hypothesis • • • • • • • • • • • • • • • • • 423
T. Silverstone
Prolactin, Growth Hormone and Cortisol in Manic
Illness . . . . . . . . . . . . . . . 431
J. C. Cookson
Amphetamine Induced Arousal in Human Subjects as a Model
for Mania . . • . . • . . . . • . . . . . . . . . 435
D. Jacobs

LITHIUM TRANSPORT RESEARCH: FROM CELLULAR

MEMBRANE TO CLINICAL PRACTICE
Ten Years of Lithium Transport Research • • • • • 443
D. G. Ostrow
Lithium Ion Transport Mechanisms in Human
Erythrocytes • • • • • • • • 449
J. Duhm

LI Transport and the Psychoses • • • • • • • • • 457

D. L. Garver, R. Hitzemann and J. Hirschowitz

Lithium Transport in Major Affective Disorders: A Model

System for Membrane-Plasma and Genetic-Physiologic
Interactions . . . . . . . • . • • . • . . • • . • • 463
D. G. Ostrow, A. Okonek, W. Dorus, W. Kitt
and J. M. Davis

xvii
Lithium Transport in Brain Cells and Human
Erythrocytes • • • • • • • • • • 471
I. Szentistvanyi and z. Janka

The Genetics of Lithium Metabolism in Major

Affective Disorders 479
J. Mendlewicz

CLINICAL APPLICATIONS OF PLASMA LEVELS

IN THE MANAGEMENT OF SCHIZOPHRENIA

Pharmacodynamics and Pharmacokinetics of Neuroleptics

and Neuroleptic Assay Methodology • • • • • • • 485
R. 0. Friedel

Plasma Levels and the Management of Chronic, Refractory

Schizophrenia • 493
L. E. Hollister

Plasma Levels as Predictors of Clinical Response and

Violent Behavior • • • • • • • • • • • • • • ~ • 499
J. A. Yesavage

Guidelines to Use of Plasma Levels from a Clinical

Perspective • • • • • • • • • • • • • • • • • 509
T. van Putten

DOSING NEUROLEPTIC MEDICATION WITH THE HELP

OF ELECTRIC REGISTRATION OF EXTRAPYRAMIDAL
FINE MOTORICITY

Optimum Dosage of Neuroleptic Agents by Electronic

Recording of Extrapyramidal Fine Motoricity
in Handwriting • • • • • • • • • • • • • • • 515
H.-J. Haase

Computerized Evaluation of Fine-Motor Symptoms 521

A. Dreher

Fine-Motor Control of Neuroleptic Dosage 525

I. Bitter

xviii
 MOVEMENT DISORDERS AND TARDIVE DYSKINESIA

Diagnosis of Tardive Dyskinesia • • • • • • • • • • • • 529

C. G. Goetz

Spontaneous Dyskinesias Amongst Psychotic Patients 537

C. D. Marsden

Animal Models of Tardive Dyskinesia • • • • • 543

H. L. Klawans, P. M. Carvey, C. M. Tanner
and C. G. Goetz

Experimental Tardive Dyskinesia 555

L. M. Gunne and J.-E. Haggstrom

Epidemiology of Tardive Dyskinesia 561

R. J. Baldessarini

Gaba-Ergic Treatment of Tardive Dyskinesia • • • • • 567

K. Heinrich, J. Tegeler, W. Woller, H. Quadbeck,
G. Arendt, E. Klieser, H. Lange and E. Lehmann

ECT: BACKGROUND AND CURRENT RESERACH

Historical Remarks on the Discovery of ECT 573

L. B. Kalinowsky

ECT, Acetylcholine Receptors, and Memory 577

B. Lerer, M. Stanley and H. Altman

Memory Disturbance after ECT in Low-Pressure Narcosis • • • 583

J.-0. Ottosson and K. Widepalm

The Influence of ECT on Regional Cerebral Blood

Flow (rCBF) • • • • • • • • • • 589
T. G. Bolwig, R. Hemmingsen and S. Vorstrup

PSYCHIC CHANGES IN PATIENTS WITH

CEREBROVASCULAR DISEASES

Influence of Rheological Therapy on Regional Cerebral

Blood Flow in Patients with Cerebrovascular
Disease • 593
A. Hartmann

xix
Organic Dementia: Clinical Picture Related to Regional
Cerebral Blood Flow and Neuropathological
Findings . . . . . . . . . . . . . . . . . . 605
L. Gustafson, A. Brun and J. Risberg

Reaction Time in Vascular (Multiinfarct) Dementia •• 613

G. Ladurner, M. Tschinkel, H. Klebl and H. Lechner

Computer as a Prosthesis in the Neuropsychological

Rehabilitation of Stroke Patients • • • • • • 619
F. I. Perez, G. A. Brown, M. Rusin, D. Koehler
and V. Rivera

Neuropsychological Aspects in Cortical Blindness due to

Cerebral Infarcts • • • • • • • • • • • • • 623
V. M. Rivera, G. R. Ravichandran, F. I. Perez,
W. Breitbach, E. Wulff and F. Grajeda

Clinical and Morphological Aspects in Diagnosis and

Prognosis of Vascular Dementia • • • • • 629
G. Ladurner and G. Bertha

Cognitive and Conative Function Changes Depending

on CVA and TIA • • • • • • • • • 635
A. Sever, A. Mesec and M. Cuk

Dementia and its Relation to Cerebrovascular Disease 641

s. Hoyer
Organic Brain Syndromes in Cerebrovascular Disease -
Early and Mild Cases without Focal Neurological
Symptoms and Signs • • • • • • • • • • • • • • 649
G. Kruger

PSYCHIC CHANGES IN PATIENTS WITH

CEREBROVASCULAR DISEASES

The Differentiation of Multi-Infarct and

Alzheimer Dementia 657
J. Marshall

Differential Dignosis between Degenerative and

Vascular Dementia • • • • • • • • • • • • 661
C. Loeb and C. Gandolfo

Long Term Clinical Observation of Multi

Infarct Dementia • • • • 667
H. Lechner, G. Bertha and E. Ott

XX
 SYSTEMS SCIENCE AND SYSTEMS THERAPY

General Living Systems Theory • • • • • • • • • • • 673

J. G. Miller

From Descartes to Pavlov to Anokhin: The Evolution of

General Systems Concepts in Biomedical Sciences
in Eastern Europe • • • • • • • • • • • • • • • 679
S. A. Corson and E. O'Leary Corson

Systeltls Therapy 685

G. Gunthern

PSYCHOBIOLOGY OF ANXIETY

Human Anxiety and Noradrenergic Function: Preliminary

Studies with Caffeine, Clonidine and Yohimbine 693
T. W. Uhde, J.-P. Boulenger, B. Vittone, L. J. Siever
and R. M. Post

Sympathetic-Adrenal and Pituitary-Adrenal Response to

Challenge • • • • • • • • • • • • • • 699
M. Frankenhauser and U. Lundberg

Psychophysiological Response Patterns in Anxiety 705

R. Hoehn-Saric and D. R. McLeod

Classification of Anxiety Disorders 711

D. V. Sheehan and K. E. Sheehan

PSYCHOBIOLOGY OF ANOREXIA NERVOSA

Anorexia Nervosa - An Addiction • • • • • • • • • • 717

D. Ploog

Depression, Anorexia Nervosa and Nutrition: Effects of

Starvation on Endocrine Function and Mood • • • • 721
M. M. Fichter, K. M. Pirke, F. Holsboer, W. Kempin
and W. Weiss

Perceptions of the Body in Anorexia Nervosa • • • • • • • • 729

P. E. Garfinkel and D. M. Garner

Naloxone in the Treatment of Anorexia Nervosa:

Metabolic and other Effects • • • • • 747
I. H. Mills and L. Medlicott

xxi
 PHYSIOLOGICAL BASIS OF ANXIETY

Benzodiazepine Receptor-Mediated "Anxiety"

in Patients • • . • . • . • . . • • • • . . . • 755
P. Skolnick, P. Ninan, T. Insel, J. Crawley and S. Paul
Benzodiazepine and Non-Benzodiazepine Anxiolitics • • 759
H. Goldberg
Mitral Valve Prolapse: A Mirror for Anxiety? 765
H. Boudoulas, B. D. King and C. F. Wooley

THE DEVELOPMENT OF HUMAN STRESS RESPONSE:

RESEARCH FINDINGS AND CLINICAL APPLICATION
Components of Human Stress Response: Coping with
Hospitalization in Adolescents • • • • • • 769
H. Steiner and S. Levine
Speech Correlates of Stress and Coping 775
H. Steiner and T. F. Anders
Social Environment and Stress: A Study of Psychiatric and
Psychosomatic Patients and their Families 783
H. Steiner, C. Mazer and J. Sobieski
The Long-Term Stress of Chronic Illness: An Illustration
from Pediatric Oncology • • • • • • • • • • • • • • • 789
G. K. Fritz and J. R. Williams
Anorexia Nervosa and Bulimia: Stress Syndromes? • • • • • • 795
H. Steiner and I. F. Litt
The Role of Psycho-Social Stresses in Bronchial
Asthma • • • • • • • • • • 801
C. M. B. Higgs and J. E. Harvey
Stress and the Adolescent's Reproductive System • • • • • • 807
I. F. Litt

STRESS AND THE HEART

Anti-Anxiety Drugs in Cardiac Disorders • • • 813

D. Wheatly

Psychogenic Hypertension: Experimental Review 819

R. Friedman and J. A. McCaughran, Jr.

xxii
The Relaxation Response and Reduced Norepinephrine
Reactivity • • • • • • • ••••• 825
R. Friedman and H. Benson
Reduction of Cardiac Risk Factors by Autogenic Training
and Physical Training • • • 833
M. Carruthers

ffiGHER NERVOUS ACTIVITY

PHYSIOLOGICAL INVESTIGATIONS OF PSYCHOLOGICAL

PROCESSES IN HEALTH AND PSYCHIATRIC DISEASES

Higher Nervous Activity in Psychiatric Patients • • • • • 841

C. Astrup

The Two Models of Conditioning of Neurotic Anxiety 849

J. Wolpe

Conditioning of Autonomic Functions, Schizokinesis, and

Psychosomatic Medicine • • • • • • • • • • • 855
s. A. Corson and E. O'Leary Corson
Biofeedback and Anxiety States 861
M. Agathon and J. B. Mazel

Biofeedback Treatment in Psychosomatic Illness 867

D. E. Sheer

Extinction Failure in Classical Conditionin~ as

a Mechanism of Psychosomatic Illness 871
A. Levey, I. Martin, R. Blizard and M. Cobb

Computer Assisted EEG-Analysis as an Approach to the

Exploration of Mental Processes • • • • • • 879
H. Pockberger, P. Rappelsberger and H. Petsche

Goal D.irected Behaviour and Movement Related Brain

Macropotentials • • • • • 885
C. L. Cazzullo and G. A. Chiarenza

ORIENTING REFLEXES IN PSYCHOPHYSIOLOGICAL

HEALTH AND DISEASE

Cardiorespiratory Orienting Responses as Predictors of

Individuals at Risk for Psychovisceral Disorders 895
S. A. Corson and E. O'Leary Corson

xxiii
The Orienting Reflex and Functional Stability of the
Nervous System • • • • • • • • • • • • • • • • 901
H. D. Kimmel

Anxiety and Attention • • • • • • • • • • • • • • • • 905

J. F. Orlebeke, R. J. M. Somsen and M. w. van der Molen
Some Characteristics of Orienting Reflexes 913
I. Maltzman

The Diagnostic Potential of the Orienting Reflex 917

J. Wolpe

Orienting Reflexes in Medical Practice 921

W. Knopp

Psychopharmacological Implications of the

Orienting Reflex • • • • • • • 927
C. H. Vranckx

BASIC RESEARCH AND CLINICAL INDICATIONS

OF PSYCHIATRIC SURGERY

Indications for Psychosurgery: Some Theoretical

Aspects • • • • • • • • • • • • • • • • • 931
P. K. Bridges and J. R. Bartlett

Follow-Up Study of Obsessive-Compulsive Patients

Treated by Psychiatric Surgery in Belgium and
The Netherlands from 1971 to 1981 935
P. Cosyns, W. P. Haaijman, H. A. M. Ceha
and N. Sijben

Author Index 941

Subject Index • 945

Summary Contents of Volume I - VIII • • • • • • • • • • • • 953

xxiv
ORGANIC BRAIN SYNDROMES IN DSM-III

Z. J. Lipowski

Clarke Institute of Psychiatry

University of Toronto
Toronto, Canada

INTRODUCTION

Psychiatric manifestations of brain disorders are fast becoming

a major area of research at the interface of psychiatry and the neuro-
sciences. Neglected for decades by investigators and clinicians
alike, especially in North America, the organic brain syndromes (OBS)
have attracted growing interest lately because of their high prev-
alence and incidence in the elderly, whose numbers are increasing
throughout the world. Moreover, advances in medicine, such as crit-
ical care and new drugs and surgical procedures, have helped to save
lives but at the same time allowed survival of many individuals
exhibiting evidence of transient or permanent brain dysfunction and
its psychiatric correlates. Thus, the frequency of organic psychi-
atric disorders has been rising and they pose a challenge to
psychiatrists to assist in their diagnosis, treatment and study.

As one of the first steps towards revitalization of this area

of psychiatry it was necessary to revise the existing concepts, ter-
minology and classification of the OBS in order to improve communica-
tion and case finding, and to stimulate clinical investigations. As
a consequence of its relative neglect, the field of organic psychi-
atry has been plagued by semantic confusion, which in turn probably
contributed to its being avoided by researchers. Not only were the
relevant terms poorly defined but also the very concept of organicity
was overly restricted and the classification of the OBS unduly sim-
plistic. Preparation of the new classification of mental disorders
of the American Psychiatric Association, or DSM-III 1 , provided a
timely opportunity to revise the definition and classification of
the organic disorders in the light of accrued clinical observations.
I will discuss first the core concepts and definitions of the
organic mental disorders as they are formulated in DSM-III, and will
then review briefly the individual organic brain syndromes.

THE CONTRAST BETWEEN DSM-III AND EARLIER CLASSIFICATIONS

The revised organic section of DSM-III represents a radical

departure from the two previous American classifications, i.e.
DSM-1 2 and DSM-II3. In preparing the revision, the DSM-III Task
Force tried to follow several guidelines: (1) to do justice to the
clinical observations made in recent years; (2) to provide a set of
definitions and explicit diagnostic criteria so as to facilitate case
finding and reporting; (3) to include all known types of psychopath-
ological manifestations associated with brain damage or dysfunction;
and (4) to bring in and redefine ancient terms, such as delirium and
dementia, which represent historical continuity of psychiatric
thought, and to introduce new terms only when necessary. As members
of the Task Force, we reviewed critically the existing concepts and
definitions in this area and discarded or revised them to reflect
current knowledge even if this meant challenging time-honored con-
ceptions and inviting controversy4.

The extent of the changes made in the organic section of DSM-III

may become clearer if one compares and contrasts it with the corre-
sponding sections of DSM-I and DSM-II. DSM-I, published in 1952,
featured a category called "Organic Brain Disorders," which were sub-
divided into acute or reversible and chronic or irreversible, re-
spectively. DSM-II, which took effect in 1968, substituted the term
"syndromes" for "disorders" and subdivided them into psychotic and
non-psychotic. The subdivision into acute and chronic syndromes
remained as an option and was thus no longer obligatory. In DSM~II,
the organic brain syndromes were defined as mental disorders caused
by or associated with impairment of brain function, and characterized
by a set of psychopathological features which included impaired
orientation, all intellectual functions, memory and judgment, as well
as lability or shallowness of affect. This set of defining features
was designated "the basic organic syndrome." Clearly, this diagnos-
tic definition of the organic brain syndromes was almost coterminous
with impairment of cognitive functions and hence precluded inclusion
among the syndromes of those manifestations of cerebral disorders
which do not feature cognitive impairment.

In DSM-III, the definition and classification of organic mental

disorders have been drastically changed. The concept of the basic
organic syndrome was rejected as one that was overly simplistic and
restrictive. The subdivisions into psychotic and non-psychotic as
well as acute and chronic syndromes have been discarded. The former
subdivision was judged to be too vague and of too little practical
value in this area of psychiatry. The subdivision into acute and
chronic syndromes, one based on unreliable prognostic judgment, was

2
found to be seriously flawed since it required clinicians to make
such judgment in the absence of valid prognostic criteria or of long-
term follow-up of the patient. As a result, a patient could be
erroneously diagnosed as suffering from a chronic, that is irrevers-
ible, brain syndrome and thus be viewed as a hopeless case, with all
the undesirable psychosocial consequences of such an ominous prognosis
for him or her. The DSM-III Task Force resolved that the prognostic
concept of reversibility should have no place in the new classifica-
tion.

THE CORE CONCEPT AND DEFINITIONS

Having discarded the previous concept, definition, and classifi-

cation of organic mental disorders, the Task Force has formulated a
set of appropriate substitutes. The traditional basic premise under-
lying the separation of these disorders from other psychiatric
syndromes has remained the same: They are differentiated on the as-
sumption that demonstrable brain damage or dysfunction constitutes
a necessary condition for their occurrence. By contrast, the so-
called functional psychiatric disorders fail to meet this criterion
at this time. Thus, the organic mental disorders are classified as
a separate category on the basis of an etiologic hypothesis. Such a
separation has certain practical consequences, as it implies that
the diagnosis and treatment of the underlying brain disorder should
have priority in the clinical approach to those syndromes that are
designated "organic."

In DSM-III, the term "Organic mental disorders" designates the

class of psychological or behavioral abnormalities which are asso-
ciated with transient or permanent dysfunction of the brain. It
should be noted that this definition does not specify the nature of
the psychological abnormality and thus underscores the assumption
that no single psychopathological syndrome is to be regarded as an
essential and diagnostic feature of the class of organic disorders as
a whole. Thus, in a departure from the traditional viewpoint, the
organic syndromes are now regarded as a clinically heterogeneous
group, one that does not invariably involve a fixed set of cognitive
abnormalities. One may say that the concept of organicity has been
liberalized. Furthermore, in DSM-III the traditionally sharp bound-
ary between the organic and the functional disorders has been delib-
erately blurred by the inclusion among the organic brain syndromes
of those schizophrenia-like, affective, and personality disorders
which are judged to be causally related to cerebral disorders.

The term "Organic mental disorders" designates the whole class

of those psychiatric syndromes which are associated with transient
or permanent brain dysfunction. A distinction has been made in
DSM-III between an organic brain syndrome and an organic mental
disorder. The former term refers to a cluster of descriptive

3
psychopathological symptoms which tend to occur together and are
either as a rule or more often than chance associated with cerebral
disorders. So defined, the OBS lack any etiological specificity,
that is to say, the presence of an OBS in a given patient does not
allow one to infer the specific cause of the underlying brain dys-
function. Once such a cause has been established or presumed, one
speaks of an organic mental disorder. In other words, the term
"organic mental disorder" designates an organic brain syndrome of
known or putative etiology. For example, if a clinician diagnoses de-
lirium on the basis of its clinical features, he has diagnosed an
organic brain syndrome only. If further inquiry indicates that the
delirium is due to pneumonia, for instance, the clinician will have
diagnosed an organic mental disorder, that is, delirium due to
pneumonia.

The explicitly acknowledged heterogeneity of the OBS is postu-

lated to reflect in part the characteristics of the underlying brain
disorder. Factors such as the degree of spread and the location of
the pathological process, and the rate of onset and of progression as
well as the duration of cerebral dysfunction are all believed to in-
fluence which OBS develops. For example, relatively widespread and
protracted cerebral disorder is likely to give rise to the syndrome of
dementia, while widespread but acute and transient derangement of
cerebral metabolism and neurotransmission is likely to result in de-
lirium. Either syndrome may be caused by a wide range of specific
etiologic factors, metabolic, toxic, infectious, vascular, and so
forth.

The OBS are defined in terms of their essential clinical features

which I will describe presently. In addition, the OBS include the
so-called associated features which may or may not be exhibited by a
given patient, and represent emotional, behavioral and motivational
disturbances that may be due directly to the cerebral disorder, or
represent the individual patient's psychological reactions to being
ill. Any patient may display a constellation of such associated
neurotic, psychotic, or behavioral symptoms in addition to those which
are considered to be essential for the diagnosis of a given OBS. This
implies that the evaluation and treatment of an organic patient should
involve not only the direct effects of cerebral dysfunction but also
the patient's reaction to it. The purpose of classifying some
psychiatric disorders as "organic" must not be misconstrued to mean
that one has substituted the brain for the whole person.

In the crucial realm of diagnosis, DSM-III requires that two

independent sets of criteria must be met in order to diagnose an or-
ganic mental disorder: (1) the essential clinical features of an
organic brain syndrome must be present, and (2) there must be evidence
from the history, physical examination, or laboratory tests of a

4
specific organic factor that is judged to be etiologically related to
the disturbance. What this implies in practice is that the diagnosis
has to involve two separate steps: first, obtaining a psychiatric
history and examination of the patient's mental status, and second,
performing a physical examination and carrying out of such laboratory
tests as blood chemistries, electroencephalography, or neuroradiology,
to ascertain the presence of brain dysfunction. It is no longer suf-
ficient to diagnose an organic brain syndrome on the basis of psychi-
atric examination alone. Furthermore, to diagnose an organic mental
disorder one also needs to identify the specific etiologic factor
judged to be responsible for both the brain dysfunction and the
associated organic brain syndrome. The etiologic factors fall into
four groups: (1) primary cerebral disorders; (2) cerebral disorders
secondary to systemic diseases; (3) exogenous toxic factors; and (4)
withdrawal from an abused substance such as alcohol or barbiturates.

THE SYNDROMES

By contrast to previous American classifications which included

only one or two organic brain syndromes, DSM-III features ten such
syndromes, but only seven of them are descriptively specific, while
the remaining three are residual categories. These ten syndromes are
grouped as follows:

1. Delirium and dementia, distinguished by global cognitive

impairment;
2. Amnestic syndrome and organic hallucinosis, which feature
relatively selective or circumscribed cognitive abnormality;
3. Organic delusional syndrome and organic affective syndrome,
whose features resemble those of the corresponding functional dis-
orders;
4. Organic personality syndrome, in which the patient's per-
sonality is altered as manifested by his or her behavior;
5. Intoxication and withdrawal, which are associated with in-
gestion or reduction in use of a substance and do not satisfy the
criteria for any of the preceding OBS;
6. Atypical or mixed organic brain syndrome, which constitutes
a residual category.

It should be noted that the OBS included in the first four cate-
gories are distinguished on the basis of their descriptive clinical
characteristics, while the remaining syndromes are diagnosed by ex-
clusion. Strictly speaking, only those seven descriptively specific
OBS constitute autonomous syndromes.

The seven syndromes are listed, with their essential features,

in Figures 1-5

5
 FIGURE 1

ORGANIC MENTAL DISORDERS

Organic Brain Syndromes
1. Delirium
2. Dementia
3. Amnestic Syndrome
4. Organic Delusional Syndrome
5. Organic Hallucinosis
6. Organic Affective Syndrome
7. Organic Personality Syndrome

FIGURE 2
DELIRIUM

Acute, fluctuating clinical picture

involving disturbances of attention,
memory, thinkin~, perception,
orientation, psychomotor behavior,
and sleep-wakefulness cycle.

FIGURE 3
DEMENTIA

Deterioration of previously acquired

intellectual abilities sufficiently
severe to interfere with social or
occupational functioning or both.
Memory, abstract thinking, and
judgment are impaired.
Sequentially or concurrently.
Defective control of impulses and
emotions.
Usual course: Progressive, static,
or remitting.

6
 FIGURE 4

AMNESTIC SYNDROME

Retrograde and anterograde memory

impairment is only or predominant
feature.

ORGANIC HALLUCINOSIS

Persistent or recurrent hallucina-

tions in awake and alert subject.
Attributed to known organic factor.

ORGANIC AFFECTIVE SYNDROME

Predominant symptom is mood disorder

attributed to known organic factor,
e.g. Cushing's syndrome.

FIGURE 5

ORGANIC DELUSIONAL SYNDROME

Predominance of delusions in waking,

alert state.
Other symptoms encountered in
schizophrenic disorders may be present.
Symptoms attributed to known organic
factor.

ORGANIC PERSONALITY SYNDROME

Essential feature is marked change in

personality involving expression and
control of emotions and impulses.
Social judgment impaired.

7
CONCLUSIONS

I have presented the salient features of the revised concept,

definitions, and classification of the organic brain syndromes con-
tained in DSM-III. The effected changes have been extensive and are
subject to criticism. Indeed, the new classification has been crit-
icized occasionally for being too inclusive and blurring the tradi-
tional distinction between the organic and the functional disorders,
and for lumping together the common syndromes, such as delirium and
dementia, with the relatively uncommon and inadequately investigated
ones such as the delusional and the affective syndromes. Such an in-
clusion, however, has been made deliberately so as to encourage
reporting of and research on the association of these psychiatric
disorders with various organic factors. The DSM-III Task Force hopes
that the new classification should have two salutary effects: (1)
improve patient care by emphasizing the importance of the search for
organic causative factors, and (2) stimulate research and formulation
of testable hypotheses. Future research will show if the proposed
changes have heuristic value and if the seven specific OBS are reli-
able and valid diagnostic categories. A psychiatric classification
must not be viewed as a collection of immutable platonic ideas but
rather as a set of working hypotheses and definitions that reflect
current knowledge, and have to be changed as data accrue.

REFERENCES

1. "Diagnostic and Statistical Manual of Mental Disorders

(Third Edition)," American Psychiatric Association,
Washington, DC (1980).
2. "Diagnostic and Statistical Manual, Mental Disorders,"
American Psychiatric Association, Washington, DC
(1952).
3. "Diagnostic and Statistical Manual of Mental Disorders,
ed. 2," American Psychiatric Association,
Washington, DC (1968).
4. Lipowski, Z.J., Organic mental disorders--an American
perspective, Br J Psychiatry, in press.

8
NEUROPEPTIDES AND CHEMICAL MESSENGERS IN THE AGING BRAIN

Ewald W. Busse

Department of Psychiatry
Duke University Medical Center
Durham, N. C.

PART 1: INTRODUCTION

This presentation is a review of the rapidly expanding scien-

tific knowledge regarding chemical messengers in the aging brain.
There are two major classes of chemic~! messengers--neurotransmitters
and neuropeptides. This paper focuses upon the peptide hormones--
brain and peripheral. In the past the two classes of chemical
messengers were assumed to be associated with either the nervous
system or the endocrine system and were believed to have
distinctly different modes of communication. Now it appears there
is a degree of overlap, a sharing of selected communication methods
and a combining of the activities of the two systems to affect the
functioning of groups of cells.

There are many chemical messengers that are abundant and

produced in the hypothalamus. The presence of these chemical
messengers supports the belief that the hypothalamus is the site of
major control of many body and brain processes and is the location
of the physiologic aging clock. Although some chemical messenger
cells within the hypothalamus appear to have their own internal
controls, many are responsive to environmental alterations via a
feedback mechanism. A common feature of aging is selective cell loss.
Consequently, the disappearance or functional decline of cells
producing chemical messengers is associated with other changes of the
aging brain. The cascading impact of the loss of cells beyond a
critical level is associated with the loss of ability to coordinate
various physiological processes including learning and memory and the
ability to respond to stress.

9
PART 2: WHAT IS A NEUROPEPTIDE? NEUROTRANSMITTER?

There are at least 100 different substances which are known to

be or suspected to be chemical messengers in the central nervous
system. Each of these chemical messengers has a characteristic
effect on neurons. The cells that generate neurotransmitters are not
randomly distributed throughout the brain. The clusters of similar
neurons have axonal projections to other highly specific brain
regions.

Generally, those messengers, which have become known as neuro-

transmitters, are of low molecular weight, often containing a single
amine. Sometimes the neurotransmitter is the amino acid itself. In
contrast, neuropeptides are much larger molecules and are made up of
chains of amino acids. The amino acids are linked together by a
peptide bond (see Figure 1). A peptide chain composed of 10 or more
amino acids is often called a polypeptide. ACTH is made of 39 amino
acids and is a polypeptide. Although monoamines usually behave as
neurotransmitters and neuropeptides as hormones, this distinction is
not consistent, as for example, norepinephrine can also behave as a
hormone. 1

r---------,
I I
I
I
I .

I
I I I
I H I
-1-~--]---N--~-T-
o
I
I I
L--------~

Fig. 1. Peptide Linkage--A Substituted Amino Bond

Neurohormones are present in the brain in a much lower concen-

tration than those of classical biogenic amine transmitters. The
lower concentration of neurohormones does not reliably indicate their
functional importance. Neuropeptides have a biological importance at
least equal to that of the biogenic amines.

Peptides are thought to be predominantly brain or peripheral.

Peptides are abundant in the pituitary gland. The posterior
pituitary is usually regarded as part of the brain and is referred
to as the neurohypophysis. The anterior lobe, referred to as the
adenohypophysis, produces hormones which may be regarded as
peripheral peptides.

10
PART 3: THE RECEPTOR SITES

Catecholamine and peptide hormone receptors appear to be differ-

ent from the receptors of steroid and thyroid hormones. The
receptors, or binding sites, of the catecholamine and peptide hormones
are located on the outside surfaces of the cell. The neurotransmitter
receptor site is usually confined to the synapsis, while the peptide
hormone receptors may be located at various places on the cell
membrane. Both steroid hormones and thyroid hormones penetrate the
cell membrane before reaching the receptor site.2

The role of the receptor is, when occupied, to influence a

regulatory protein which initiates a process within the cell.
However, the receptor must first distinguish a particular signal
from a jumble of other signals and accept the proper messenger.
There is a strange difference in the patterns of transmission. A
polypeptide hormone, along with its receptors, can be taken into a
cell by a process termed "internalization." Following internaliza-
tion, the hormones may be degradaded or they may be found
intracellularly and initiate a process. Although neuropeptides may
act as neurotransmitters, their major function appears to be that
of neuromodulator. It increases or decreases intracellular processes.

Table 1. Location Receptor Sites

Type Binding Site

Neurotransmitters Synapse
Neuropeptides Cell Membrane--Scattered
Steroid Hormone Within Cell--Involves Chromatin
Thyroid Hormone Within Cell--Involves Nuclei

PART 4: SYNTHESIS--LOCATION

Neurotransmitters are synthesized primarily in nerve terminals

and stored in proximity to the presynaptic terminus. When the nerve
is properly stimulated, the neurotransmitters are released into the
synaptic cleft and are picked up on the receptors of the cell next
in the neuronal chain. Some neurotransmitters are excitatory, others
are inhibitory.

The origin and distribution of peptides are somewhat confusing.

Some of these chemical messengers are apparently restricted to the
brain while others are found in organs and systems throughout the
body.3 For example, it has been known for a long time that somato-
statin occurs both in the gastrointestinal tract and in the brain.
Similar peptides may have different functions in various types of
cells, and cells exist that include both neurotransmitters and neuro-
peptides. Even more confusing is the recent recognition that not all
substances that act like hormones are produced in the body. Certain

11
substances that behave like hormones and resemble them chemically
are brought into the body by food. These are opiate-like substances,
and, when produced by the body, are called endorphins; when taken
into the body are called exorphins.4

There is another interesting phenomenon related to neuropeptides.

Polypeptides often include components that, when split, produce other
active, shorter-chained neuropeptides. These shorter peptides or
peptide fragments are usually neuronally active over short distances.
Of particular interest to the field of aging are fragments of the
ACTH as there are behavioral effects of the fragment labelled
ACTH 1-10 and ACTH 4-10.

PART 5: FUNCTION

Neurotransmitters seem to have specific functions in the brain,

while it is more likely that brain peptides have a broader function
that is related to the maintenance of the normal activity and
functions of the cell. Considerable attention has been given to the
neurotransmitters, and their changes in affective disorders, anxiety
reactions and panic, and the major degenerative disorders of the
brain.

In recent years there have been a number of peptides identified

and localized in the nervous system of vertebrates. However, only a
comparative few have been demonstrated in the human central nervous
system. This limitation is, in part, due to the fact that studies of
the human brain are based upon postmortem examinations. However, it
is assumed that man has similar, if not identical, peptides both in
the brain and the periphery.

The co-existence of peptides and classical transmitters in some

cells complicates the picture even further. It has been reported
that sympathetic neurons can contain both acetycholine and adrenalin
and that some endocrine-producing cells reveal a co-existence of
peptide and biogenic amines.S

PART 6: RELEVANCE TO AGING

The primary progressive deterioration disorders such as senile

dementia and Parkinsonism are characterized by pathological changes
and premature cell death in relatively discrete neuronal populations.
Often the cell, prior to death, is believed to show pathological
changes that lead to its malfunction and eventual death. Because of
the selectivity of cell death, it has also been demonstrated that
these diseases are characterized by a loss of associated neurotrans-
mitters. Parkinson's disease is said to be primarily due to the loss
of dopaminergic cells in the substantia nigra. But this loss of

12
dopaminergic cells does not necessarily take place in other parts
of the brain. For instance, such cells are preserved in the hypo-
thalamus. It is possible that these differences in the distribution
of surviving cells may be attributable to the effective presence of
peptides that properly maintain the health of the cell.

Alzheimer's disease is said to be due to a selective loss of

cholinergic neurons primarily in the nucleus basalis of Meynert.
But the cholinergic defect does not explain the presence of senile
plaques in the cerebral cortex. The cortex is devoid of cholinergic
cells although projections do exist. Hence if there is a common
etiology it is likely to be found elsewhere.

There are a limited number of studies of peptides in Alzheimer's

disease. One report indicates that vasoactive intestinal polypeptide
(VIP) and intrinsic cortical neuropeptide are unchanged in Alzheimer's
disease. However, somatostatin (SRIF) is especially reduced in the
temporal cortex of those with Alzheimer's disease. Another neuro-
peptide, cholecystokinin (CCK) did not differ between the group with
Alzheimer's disease and the control group. Alzheimer's patients did
show a significant fall in catecholamine transferace (ChAT) in all
areas of the cerebral cortex. In addition to the reductions in
somatostatin, a neurotransmitter, GABA, was also reduced.6

The distribution of quantity of neuropeptides and the monoamines

appeared to be associated with brain and bodily structure and function
that change with the passage of time. Aging in humans brings about
alteration in temperature regulation, reproduction, water balance,
sleep and a number of circadian rhythms that may be linked to neuro-
peptide changes.

With advancing age, the neurons that make neuropeptides in humans

undergo intracellular changes as well as a decline in cell numbers.7
A similar reduction of cells may not be found in some laboratory
animals. Such differences would be important to understand. Aging
brings about important changes in the hypothalamus and in the pitui-
tary gland. Although most of this information comes from studies in
rats and other animals, changes in man are consistent with such
alterations (for example, the decline in reproduction functions and
a rise in prolactin secretion). A number of authors have noted that
a decline in aging of both neurotransmitters and neuropeptides. The
interaction of these two groups, as related to aging, will be
particularly important. Epinephrine promotes release of gonaditropins
and growth hormones, while it inhibits the release of ACTH. Conse-
quently, a loss of epinepherine could have a profound effect on
important neuropeptides.

Somatotropin (growth hormone) in young adults and children is

released throughout a 24-hour period, but in relatively large amounts
during the first three hours of sleep. In old adults this peak of
growth hormone in sleep does not occur, but the total amount of
hormone extended over 24 hours is not significantly different from

13
the total of the young adults. 8 Somatomedins that represent a group
of at least six peripheral polypeptide hormones that seem to be
related to somatotropin are of particular interest in the field of
aging. The somatomedins are associated with cartilage-stimulating
activity, mitogenic activity, and insulin-like capacities. The site
of the production of somatomedins has not been established, but it
is possible they originate in the liver. Somatomedins are present
in the blood stream and appear to have widespread influence.9

Studies of pituitary growth hormone regulation have_, for several

years, suggested that some depressed patients secrete less growth
hormone than normals in response to a variety of stimuli.10 Although
the human pituitary content of growth hormone is relatively constant
with age, its pattern of excretion does alter with age. The disap-
pearance of a few cells in the hypothalamus may have far-reaching
consequences.

The behavioral effects of peptides are easily recognized only

when the relationship appears to be a direct one. Undoubtedly there
are indirect roles which are equally important. Current knowledge
of brain peptides is limited and enigmas exist, but there is no doubt
that peptides will play an important role in Alzheimer's disease.
The future is exciting but clinicians must await the day of the
peptides.

REFERENCES

1. F. E. Bloom, Neuropeptides, Sci Am 145(4):166 (1981).

2. J. D. Baxter and J. W. Funder, Hormone receptors, N Engl J Med
301(21):1149 (1979).
3. D. T. Krieger and J. B. Martin, Brain peptides, part 2, N· Engl_
J Med 304(16):944 (1981).
4. Food peptides--a new class of hormone, JAMA 247(17):2379 (1982).
5. T. Hokfelt, J. M. Lundberg, and 0. Johansson, Coexistence of
peptides and classical transmitters--morphological and
physiological aspects in "Chemical Neurotransmission: 75
years," L. Stjarne, P. Hedquist, H. Lagercrante, and
A. Wennmalm, eds., Academic Press, New York City (1981).
6. M. N. Rossor, P. C. Emson, and L. L. Iverson, Neuropeptides and
neurotransmitters in cerebral cortex in Alzheimer's disease,
Agi_gg_ 19:15 (1982).
7. J. R. ~iadek and B. C. Blanchard, Age-related declines in
perikaryal monoamine histoflourescence in the Fischer 344
rat, Aging 17:13 (1980).
8. P. Prinz, H. Halter, and M. Raskind, Diurnal variation of plasma
catecholamines and sleep-related hormones in man: relation to
age and sleep patterns. NIH Conference (1980).
9. L. S. Phillip and R. Vas.silopoulou-Sellin, Somatomedins, parts 1
and 2, N Engl J ~ed 302(7):371; 302(8):438 (1980).
10. N. Matussek, Neuroenaodrinologische untersuchunger bei
depressiven syndromen, Nergenarat 49:569 (1978).

14
CROSS SECTIONAL AND LONGITUDINAL STUDIES OF ORGANIC

BRAIN SYNDROMES

Kornelius Kryspin-Exner

Universitatsklinik fUr Psychiatrie

A-6020 Innsbruck, Austria

Tests, which are made to objectify and quantify

cognitive impairment and which can be used for longi-
tudinal studies are mostly of theoretical, but in some
groups of patients also of therapeutic interest. Metho-
dological difficulties of these studies are well-known.
Every paper in the international literature includes
the demand for a more thorough and exact and more popu-
lation oriented diagnosis. On the basis of studies, that
were carried out at the Universitatsklinik fUr Psychia-
trie Innsbruck, two examples will be presented.

Schubert investigated whether there is a correl-

ation between cognitive deficits and the observed be-
haviour disturbances within the various syndromes of
dementia. The study was designed as a longitudinal
program. 89 patients with the diagnosis "senile dementia
of the Alzheimer type (SDAT)" were examined five times
in a period of 35 days. In order to maintain the homo-
genity of the sample only 40 patients were used for
statistical evaluation. To measure the cognitive
deficits it seemed suitable to consider only the func-
tions "impaired memory" and "impaired performance"
(attention and concentration) . These two functions are
easy to measure and are said to represent the principle
symptoms of the intellectual deficit. It was necessary
to choose a patient-oriented test system, which means,
that instructions are easily understandable and that
the stress for the patients is minimal. In preliminary
experiments the test system "Syndromkurztest" (Erzig-
keit, 1977) and "Funktionpsychose-Skala-B" (Lehrl et

15
al., 1977) was chosen. Both procedures together seemed
suitable for demented patients.

The Geriatric Rating Scale (GRS, Plutchik et al.,

1970) for the nursing staff and the SCAG (Sandoz Clinical
Assessment Geriatrics, Shader et al., 1974), in which
the rating is done by psychiatrists and which also
assesses the depressive syndrome, were chosen following
the recommendations of Kochansky (1979) and Gurski(1981).

According to the authors these procedures are valid

to assess abnormalities in behaviour and psychopatholo-
gical disturbances in gero-psychiatric patients. They
are also very well suited to follow the course of ill-
ness.

Nine factors were gained from the Geriatric Rating

Scales by hierarchical cluster analysis (Anderberq 1973):
GRS factor 1 aggressiveness, GRS factor 2 sleep distur-
bance, GRS factor 3 incapacitation, GRS factor 4 social
dysfunction, GRS factor 5 severe social dysfunction,
SCAG factor 1 organic impairment, SCAG factor 2 loss of
appetite, SCAG factor 3 disturbed social adaption, SCAG
factor 4 depression.

After the clinical separation into 18 mild and 22

severe forms of dementia no significant correlation bet-
ween the degree of cognitive impairment and behavioural
disturbances was found during the whole course (fig. 1
and 2) .

This leads to the conclusion, that the measured

severity of cognitive deficits does not correlate with
the observed behavioural disturbances. One can also
deduce the fact, that the behaviour of the patients with
SDAT varies a great deal during the course of the ill-
ness. The changing abnormalities in behaviour do not
correlate with the significantly more stable cognitive
deficiencies. It can thus be concluded, that it is not
possible to deduce severity and therapeutical suscepti-
bility of cognitive impairment in SDAT from behavioural
parameters.

In a further study Demel et al. examined 117

patients, aged 54 to 60, who either suffered from a
well defined organic impairment through cerebrovascular
insufficiency or were suspected to have beginning
dementia or were alcoholics. All three groups reported
the subjective complaint of an "premature deficiency
syndrome". The patients complained about a lack of

16
 I
I I
I
I
I
I I
I I I
I
I I
I I I
I
I I I I
I
I I I I
I I I I I
I I I
I I
I I I
I
21 I
"t:
I I
I
I
I
I I I
I I I I
I I I I I
I I

n
I I I I I I

0
I

A)
I I
I I \4
b
I
I
I
I

I
I
I
I

I
I
0
I
I
10 "" I
I I
I
[) """' I
I

I
0 n.s.

I
I II •< o.os
1§\1 •< 0.01

•
I
I I I
I I I I I I
I
I
I
I
I
I
I I
' < 0 .001
+ KEiiDN.l -COA.FI:EL.AT ION
C'JEFF ICIE.PH '"'('

COrutf:LATI Ofii.S. Of= FP· S.TAriOARD SCOA:ES WITH GRS· FACroas 1· 5 AHO SC.V.4ACTORS 1·'~

Fig . 1

COIUIEI.ATII:*S FOR SEYIERE FOR"$ OIF D£~NT I A

q
GRS• J
~ I
SCAG- 1 )

I I ,I /I
I I I
I DA"' / 35
t'J ]j p
I I I

I I
0 I I
I
I I I I
I I I
I I I I I
I I I I I
I
I I I I I
I I I I
I I I I
I I I I
I I
I I I I
I
I
I
I I
I I I
I I I
I I I I I I

,I
I I
2\

/I
I I I I

.L) '?
10 ,U 0
I J

I
I
I
"""'
I
I
I I I I
I I
I
I
I

I
I I
I I

--r"r--~--.~~ 0 n.s.
- 11 • < 0.05

1§\1 • < 0.01

+ • P < O.OOI
KEHDALL•CII)RRE t,.AT I OH
CO£FFICIENT "(;.

COftULATIOHS OF FP·STAN~RO SCORES "ITH f;RS·FACfOIU 1-S >\NO SCAG·F.t.CTORS 1·~

Fig . 2

17
energy, difficulties to concentrate and fatigue. Among
the 117 persons tested there were 42 alcoholics. In
routine performance tests assessing memory, concentra-
tion and attention, the whole group of test persons did
not show any deficiency symptoms. The individual values
were at least above the lower normal range. In longer
stress tests deficits were found under different condi-
tions either in tests using reaction-controlled presen-
tation of signals ("Pauli apparatus", i.e. the appearance
of new signals depends on the reaction to the previous
one) or stimulus-controlled presentation of signals
("Determinationsger~t", i.e. the signal appears inde-
pendently and the patient can not influence the speed
of presentation).

Hierarchical cluster analysis was carried out and

the level was based on three clusters. For the above
mentioned reaction-controlled presentation of signals
performance patterns were parallel for all three
clusters and differed already at the starting point.
A more subtle assessment additionally reveals typical
performance profiles, in which the rates of increased
performance can be interpreted as a result of practice
and that of declined performance as a sign of fatigue.
The group of alcoholics can be found in all three
clusters. 21 % of the alcoholics show a performance
above average, a high percentage of these persons was
exceptionally intelligent and between 55 and 60 years
old (Fig. 3). If the test is stimulus-controlled, the
percentage of alcoholics in the cluster with the best
performance decreases to 14 %, although the distribu-
tion of the three groups in the clusters is similar to
that of the reaction-controlled tests (Fig. 4). This
could mean, that alcoholics deal worse with stress than
patients whose mild cognitive disturbances have different
causes. The study shows, that reactions in stress situ-
ations seem to be a more sensitive measurement for
possible impairment in alcoholics than other test
methods.

Summing up the results achieved in the field of

alcoholism research, one discovers different cognitive
deficit patterns after a certain duration of alcohol
addiction. The restitution of these deficits can take
place within a few weeks, a few months or within a
year, but sometimes - in a small percentage of cases -
it does not take place at all. The result of this last
study indicates, that evaluation of mean values of
groups are likely to imply great mistakes. Examinations

18
 REACT I ON-CONTROLLED PRESENT AT I ON OF SIGNALS

"p A U L I ~ P P fl. R fl. T U S"

TOTAL N = 117 "PREMATURE DEFICIENCY SYNDROMES"
NUMBER OF
ALCO~OLICS N = 42
CALCULATIONS
75

• •
50

25

10TH 15TH 20TH INTERVAL

(1 1/2 MIN, EACH)

---CLUSTE R TOTAL N=31, ALCOHOLICS N•9, I ,E, 21,4l OF ALL ALCOHOLICS

-·-·-·-CLUSTE R II N=33, N=14, I .E. 11,3%

----CLUSTE R III N=53 N=19,I.E. 45...3.%_
N=42 100,0%
e ~IORr1AL SCORES

Fig. 3

STIMULUS-CONTROLLED PRESENTATION OF SIGNALS

D E T E R M I N A T I 0 N S G E R A T

__.-4 I TOTAL N=ll7 "PRE~IATURE DEFICIENCY

CORRECT REACTIONS
SYNDROMES"
~----.------~·----
0
0
0 ALCOHOL! CS N=42
0
75

...... -·-·-·-·-·-·-·-·-·-·-·~I I
DELAYED CORRECT
REACTIONS
50 50

_________.._ .._ ______ ,:._ __ _.I I I

.... - - - ---....,.III
25 25

:·-·-·-·-·- ·-·-·-.... !I

~I
2 3 4 1 2 3 4
1
4 TESTS/100 SIGNALS, 1,2 SEC, EACH

--CLUSTER I TOTAL N=33, ALCOHOLICS N=6, I .E. 14,2% OF ALL ALCOHOLICS

-·-·-CLUSTER II N=64 N=27, I.E, 64,4%

----CLUSTER III N=20 ~=9, I , F.., __2l._fll_

N=42 100,0%
0 CONTROLS

Fig. 4

19
should be carried out intraindividually on the basis of
longitudinal designs, using exact methods that are
adapted to the symptoms. The course of brain performance
has to be considered with the same consequence as the
base line. The determination of the individual base line
of organic impairment will remain an unsolved problem,
but nevertheless, we think it is justified to carry on
experiments to objectify cognitive deficits.

REFERENCES
•
Anderberg, M., 1973, Cluster-analysis in application.
Academic Press, New York
Demel, I., Kryspin-Exner, K., Moser, C. und Schubert,H.,
1982, Klinik der Alkoholabh~ngigkeit, in "Krankheit
Alkoholismus", Wieck, H., Schrader, A.-,-Daun, H. 1
und W'itkowski, R., eds.
Perimed, Erlangen
Erzigkeit, H., 1977, Manual zum Syndrom-Kurz-Test.Formen
A-E. Vless, Vaterstetten.
Gurski, G.E., 1981, Evaluation of geriatric patients with
special reference to clinical trials of so called
nootropic drugs. Pharmacopsychiat. 14: 51
Kochansky, G.E., 1979, Psychiatric rating scales for
assessing psychopathology in the elderly: a
critical review, in: "Psychiatric symptoms and cog-
nitive loss in theelderly", Raskin, A. and Jarvik,
L.F., eds., Wiley & Sons, New York-Chichester-Bris-
bane-Toronto.
Lehrl, S., Fuchs, H.H., Lugauer, J., Schuhmacher, H. and
Nusko, G., 1977, Manual zur Funktions-Psychose-
Skala-B. Vless, Vaterstetten.
Plutchik, R., Conte, H., Lieberman, N., Bakur, M., Gross-
man, J. and Lehrman, N., 1970, Reliability and vali-
dity of a scale for assessing the functioning of
geriatric patients. J. Amer. Geriat. Soc. 18: 491
Schubert, H., 1983, Hirnleistung und Verhalten bei
senilen Demenzen vom Alzheimer-Typ. Wien.Klin.
Wschr., Suppl., in press.
Shader, R.I., Harmatz, J.S. and Salzman, C., 1974, A new
scale for clinical assessment in geriatric popula-
tions: Sandoz Clinical Assessment-Geriatric (SCAG).
J. Amer. Geriat. Soc. 22: 107

20
TODA Y'S PROBLEMS AND RESULTS OF BIOLOGICAL

RESEARCH IN SCHIZOPHRENIA

H. Hippius and M. Ackenheil

Psychiatric Clinic of the University of Munich, Germany

Since 1908, since exactly 75 years ago, the usage of the term "group of
schizophrenias" suggested by Eugen BLEULER for the endogenous psychoses
for whose central group Emil KRAEPELIN coined the term "dementia
praecox" in the 6th edition of his textbook in 1899, has become standard in
psychiatry.
In discussing the causes of dementia praecox, as early as 1899 KRAEPELIN
had no doubts as to the heredity of these clinical pictures. In view of the
goal of identifying a nosological entity, he concluded at that time, "These
considerations force us to assume that a palpable illness process in the brain
must be involved".
Kraepelin's conclusion, made at the close of the previous century for
dementia praecox, was readily extended to include the "group of
schizophrenias", whereby these were regarded as "nosological entities" from
the beginning by biological psychiatric cause researchers and continue to be
so regarded today.
Biological-psychiatric schizophrenia research thus began at the end of the
last century and since that time has been pursued almost exclusively from a
nosological perspective.
Since 1900 there has been an endless number of studies conducted with the
aim of discovering the "somatic cause" of schizophrenia. Despite the
intensity of these efforts, a period of more than 50 years elapsed without
the somatic basis of schizophrenia being uncoverd by either morphological,
neuro- and psychophysiological or endocrinological and biochemical
methods.
Only upon discovery, due to clinical therapeutic empiricism, of the
therapeutic efficacy of neuroleptics in the treatment of schizophrenic
psychoses was an important step forward taken. The explanation of the
activity of these neuroleptics using the methods of biochemical

21
 pharmacology delivered approaches for studies with the goal of deducing the
causes of the illness treated by medicaments by studying the activity of
these successfuly applied medicaments.
This is a completely legitimate and often successfully followed route in
medical research.
Thus, the most promising approaches for biochemical research on
schizophrenia within the past 30 years have resulted from the interplay
between psychopharmacological-pharmacotherapeutic and biochemical
pharmacological research in hospitals and laboratories.
The discovery by CARLSSON and LINDQVIST (1964-) that neuroleptics act
via a postsynaptic dopamine receptor blockade resulted in the Dopamine
Hypothesis of Schizophrenia (MA THYSSEE 1977).
This hypothesis, which assumed an increasing activity of dopaminergic
neurons to be the cause of schizophrenia, was supported above all by clinical
pharmacological findings in the final analysis. For one thing, dopamine
agonistic substances such as L-Dopa or amphetamine were able to
reinforece or trigger productive schizophrenic symptomes dependents on the
dose and length of medication, on the other hand, especially these
productive symptoms are sensitive to a blockade by neuroleptics (review:
ACKENHEIL et al. 1980). Prognostic studies show that patients with marked
florid symptoms generally respond better to treatment with neuoleptics and
that the prognosis is more favorable, whereas the so-called minus or
deficient symptoms are less responsive at the usual dosage (LECRUBIER and
DOUILLET 1983).
In this connection, it is understandable that if one is proceeding from the
classical categorization of schizophrenic illness, the prognosis for the group
of paranoid-hallucinatory schizophrenias is more favorable than that for
hebephrenia or schizophrenia simplex.
Research in recent years which attempted to identify a disturbance of
dopaminergic neuronal activities was generally accompanied by a lack of
success. Neither the assay of concentrations of dopamine nor its
catebolites, such as the homovanillic acid, in body fluids, were able to
establish unmistakable differences between schizophrenic patients, healthy
controls and other groups of psychiatric patients (review: ACKENHEIL et al.
1980).
The explanation for these failures are to be sought in the complexity of both
human behavior and its disturbances which manifests itself in the variety of
the schizophrenic clinical picture and the manner in which the brain
function, which even today has only slightly been explored.
During the period of the past 30 years of biological psychiatric research, one
suspected cause was the inadequate standardization and operationalization
of psychiatric diagnostics. This operatlonalization is without a doubt the
condl tio sine qua non for all biological research (KENDELL, 197 5), although
even here some limitations should be affected. Throughout the years,
instruments from the lCD to the modern DSM III, to mention some, were
developed to standardize nosological diagnoses and operationally compile
the symptoms. The development of the different rating scales, from the
English BPRS, (1), the German AMDP (2), which is also used most often in
France, to the PSE (3), represents enormous progress. Although these

22
 instruments are available in sufficient quantity today, we are still far from
the internationally valid standardized documentation of biological-
psychiatric research projects, whereby this is certainly not ascribable to the
possible inadequacies which are inherent to this system.
An operationalization of the diagnostics and the psychopathological finding
would necessarily have to proceed from a large number of patients. In so
doing, the disadvantage that the individuality of the patient would be
neglected would have to be accepted. But it is exactly the individuality of
the schizophrenic patient which we can intuitively comprehend (Ackenheil
et al., 1984), but which is difficult to operationalize that is of a great
importance in psychiatry. Often the patient does not display all the
symptoms listed in the rating but just a single one, which, however, is
especially conspicious. Other patients can demonstrate a change in
symptoms during the course of the illness. Only rarely is it possible to judge
the course and prognosis at the onset of the illness. The kaleidoscopic
picture of "schizophrenia" led to a search for basic disturbances, a search
which also occupied the classical psychiatrists GRIESINGER, KRAEPELIN
and BLEULER to a great extent.
The individuality of the schizophrenic patient also manifests itself in the
results of biological-psychiatric research. Even when considering the
opera tionalized clinical criteria, the results ob biological studies, whether
biochemical, neuroendocrine or psychophysiological paramters are used, are
characterized by their greater variance as compared to healthy controls.
Using neuroendocrine methods which allow to measure the receptor
sensitivity of dopaminergic and alpha-adrenergic receptors in the pituitary
gland in the hypothalamic-hypophyseal system, differences between
schizophrenic patients and healthy controls, respectively other psychiatric
patients could be compiled.
On the average, growth hormone secretion was increased after stimulation
by both clonidine (MA TUSSEK et al., 1980) and apomorphine (ACKENHEIL,
1981), as was also reported by other groups (ROTROSEN et al., 1977;
PANDEY et al., 1977). However, there were wide individual differences
ranging from a so-called blunted response to very high stimulation values.
These individual differences have not been able to be related to
psychopathological symptoms or subgroups of schizophrenic patients. There
are signs that the degree of growth hormone stimulation could be a
predictor for the response to neuroleptic therapy. The group in Liege
(ANSSEAU et al., 1983) recently found that patients with high stimulation of
growth hormone following apomorphine responded better to neuroleptic
therapy than did those with a blunted response.
Above all, however, social factors play along-term role in therapeutic
success. In psychophysiological experiments we found a greater activation
pattern in schizophrenic patients, reflected among other things in the heart
rate, EMG and serum noradreanaline secretion and also characterized by
very wide individual variability (ACKENHEIL and ENGEL, 1981). Patients
who lived in a family with high expressed emotions (HEE according to
TARRIER, 1979) showed a greater activation pattern and demonstrated a
greater relapse frequency than did those with a low activation pattern in a
family with low expressed emotion (EIBL-EIBESFELD et al., 1982). (Fig. 1
and 2).

23
 Herz-
frequenz

100

90 I

80
•
70

60

0
-..
low EE high EE
P"" O.D1

Fig. 1 Heart rate of schizophrenic patients living in families with low and high
expressive emotions (L and HEE)

-------------------------------low EE
100 % Pat.16
8 6 p"'-0.05

high EE
50%

Wochen nach
0%+- - -.-- - -..,.- - - T"-
15 0 Entlassung
0 50 100

Fig. 2 Relapse rate of schizophrenic patients with low EE and nigh EE

24
Social aspects are a further factor in addition to a possibly different initial
biological condition, which are important to the response of patients
receiving long-term treatment. The actual effect of the different
neuroleptics also plays an important role. The postsynaptic dopamine
receptor blockade is a substantial principle of the efficacy at the beginning
of the treatment, although it must also be considered that depending upon
the dose and pharmacokinetic aspects, there is an additional antagonistic
presynaptic dopamine and alpha recaptor effect which results in an
increased release of transmitters and thus increased neuronal activity
(GROSS et al., 1980). Such an effect was possibly put to use empirically on
the basis of clinical experience in the treatment of deficient symptoms,
which we shall discuss in detail later. Furthermore, the different neuroleptic
attack the nigrostriatal, the mesolimbic and tuberinfundibular dopamine
systems with varying potency. Comparison of clozapine, and eventually its
successor, fluponex, with neuroleptics are a prime example of this. Last but
not least, biochemical studies have revealed an adaptation of neurons and
synapses during the course of neuroleptic treatment. These changes can be
clinically compiled on the basis of the course of extrapyramidal motor
disturbances and other side effects. Following years of neuroleptic
treatment the synapse of catecholaminergic neurons has adapted to a
different level in the transmission of nerve impulses. These changes must be
considered when withdrawing therapy and can also be used for research
purposes.
After years of neuroleptic therapy the dopamine metabolism measured on
the basis of the homovanillic acid in the liquor, and the prolactin secretion
in the blood are normalized (ZANDER et al., 1981). The growth hormone
secretion stimulated by apomorphine and clonidine continues to be
suppressed, the noradrenaline secretion in the blood is clearly incresed at
this point (review, NABER and MUELLER, 1983). An increase in
noradrenaline secretion is found in acutely schizophrenic patients in blood
(ACKENHEIL et al., 1979), in liquor (LAKE et al., 1980) and also in the
nucleus accumbens (limbic system, autopsy material (FARLEY et al., 1978)
of schizophrenic patients. Acute, respectively subchronic treatment effects
via a feedback mechanism an increase in nonadrenaline secretion which does
not adapt but increses greatly during the course of years of treatment in the
awake of the alpha-adrenalytic effect of neuroleptics (Fig. 3). If the long-
term neuroleptic treatment is withdrawn, biochemical and neuroendocrine
methods measure changes which are most conspicious approximately 14 days
thereafter. The secretion of prolactin and noradrenaline in blood decreases,
beta endorphin and cortisol increases, the apomorphine-stimulated growth
hormone secretion is greater (Fig. 4). Clinically, an amelioration of above
all deficient symptoms, as well as a frequent increase in productive
symptoms which can finally lead to an exacerbation of the psychosis are
observed. On the whole, the sum score of the rating scales is improved,
although an opposite change in both these symptoms groups occured. Signs
of tardive dyskinesia increase, although no supersensitivity in the
hypothalamic-hypophyseal system can be measured with the apomorphine-
GH-situation-test compared to healthy controls and acute untreated
schizophrenic patients.

25
 ng/ml
Plasma- Norepinenephrine
2.5

X •
2.0
X •

•
•
1.5
X:

1.0
X
X•
..
•

••

·~
0.5

...
i •
(29)
•
•
(12) (58)
A B c
A: Controls
B: Acute Schizophrenics.untreated
C: Chronic Schizophrenics.
treated with Neuroleptics

Fig. 3 Plasma noradrenaline in schizophrenic patients

A Controls
B Acute schizophrenics, untreated
C Chronic schizophrenics, long-term treated with neuroleptics

+ p<O.l unpaired student t-test

++ p< 0.01 unpaired student t-test

26
 A NL- therapy
• 8 12 'days drug free
c 30 .....
•

=-
~

Q)
en
35
I
•

-E

5 :···:

A B C
acute chronic
schizophrenics

Fig. 4 Apomorphine-GH-stimulation test

acute schizophrenic patients, untreated
A chronic schizophrenic patients, long-term with neuroleptics (NL)
B 15 days withdrawal of chronic NL treatment
C 30 days withdrawal of chronic treatment
mean

27
 Attempts at withdrawal of chronic neuroleptic treatment and the
accompanying biochemical changes are well suited for studies of the
pathophysiological causes· of schizophrenic symptoms. It is obvious that no
direct relationship between biochemical and psychopathological changes
exists. The exacerebation of above all productive symptoms occurs
temporally independently of the biochemical changes: it appears to be more
a matter of vulnerability factors, in addition to which other, e.g. social and
stress factors appear. The partially opposite changes between productive
and dificient symptoms suggest that two independent symptom groups can
appear in schizophrenic illness. The appearance of minus symptoms has long
been described (CONRAD, 1958), but has only recently become a focal point
of interest. This was due primarily to studies on post mortem material of
schizophrenic patients. On the basis of binding studies using spiroperidol a
wide variability in binding sites was established. CROW described a Type I
and Type II of schizophrenia which substantially agree with the productive
and deficient symptom types (CROW, 1980).
Considering the aforementioned findings, above all the lack of a relationship
between changes in measurable neuronal dopaminergic activity and the
psychopathological symptomatology - this again sub-divided into productive
and negative smptoms - the attempt was made to include recent
biochemical discoveries concerning the interactions of other transmitters
and hormones, respectively peptides in reflections on and studies in
pathophysiology. Whether changes in noradrencergic neurons and alpha-
adrenergic receptors are important was also studied on the basis of
withdrawal studies after long-term neuroleptic therapy and also in therapy
experiments. In one study we found that a group with. especially high
noradrenaline secretion as compared to another group of schizophrenic
patients also reacted more strongly with psychopathological symptoms.
Perhaps these patients were more sensitive to stress since cortisol and beta
endorphin in serum also increased. Beta endorphin, which interacts with
cholinergic neurons, was found to be reduced in the liquor of schizophrenic
patients (NABER et al., 1983). The treatment of schizophrenic patients with
the endorphin antagonist naloxone was disappointing though (NABER et al.,
1983). A similar interaction with catecholaminergic neurons has been
described for cholecystokinin (CCK). Initial clinical trials with CCK-agonist,
ceruletide, were promising. An initial double-blind study of this substance
versus placebo in chronic schizophrenic patients who continued to receive
neuroleptics but who demonstrate residual psychotic symptoms
demonstrated an improvement in the symptoms which manifested itself
primarily in the subscores thought disturbances and hostility in the brief
psychiatric rating scale. Between the patients treated with ceruletide and
placebo there was no difference, so that one must assume that the
improvement occurred as a result of the increased attention paid, a social
factor (ALB US et al., 1984).
A further approach to the biochemical characterization of schizophrenic
patients is to be found in binding studies with radioactively marked ligands.
These methods date back to studies by CREESE and SNYDER (1975), who
located specific binding sites for pharmacological alpha-, beta- and

28
dopamine antagonists in the brain. Using a similar method, CROSS and co-
workers (1978) later showed increased spiroperidol binding sites in autopsy
material of chronic schizophrenic patients, whereby the problem of post
mortem studies has already been a subject of controversy. Proceeding from
these methods such binding sites were also found on human blood cells,
whereby their importance is not yet clarified. At our hospital we have
developed a method which allows us to measure the binding sites on
granulocytes, lymphocytes and thrombocytes from one blood sample.
Yohimbine was used as the ligand for alpha-2 receptors on thrombocytes,
dihydroalprenolol for beta-2receptors on granulocytes and spiroperidol for
dopamine receptors on lymphocytes (BONDY et al., 1984).
Markedly increased dopamine binding sites were found on the lymphocytes
of untreated acute schizophrenic patients; less conspicious was the
reduction in alpha- and beta-receptors (Fig. 5). The outcome of further
studies must be awaited in order to see whether this indeed a matter of
dopamine receptors or whether it is simply a matter of changes in the
lymphocytes indicative of a virus infection (CROW, 1980) and finally
whether these binding sites on the brains of acute schizophrenic patients
have undergone similar changes. Furthermore, it is important whether these
binding sites are specific for one group of schizophrenic illnes and whether
one is dealing with a trait or state variable.
Initial studies conducted in relatives reveal the same pattern of binding sites
in at least one parent (Fig. 6). However, the noradrenaline or dopamine
secretion in blood which at the same time exceeds the norm in schizophrenic
patients, remaining unchanged. When one assumes that the secretion of
these transmitters depends on stress factors, there exists a possible
connection between genetic prodisposition and ecological factors, i.e. life
events (KA TSCHNIG et al., 19 ). The specifi ty of these changes must still
be studied; theoretically it is plausible that changes in the receptors and
simultaneously incresed release of transmitters must coincide in order for
schizophrenic symptoms to originate. Hormones and neuropeptides are
important in that they modulate regulatory processes taking place on the
synapses. Further classification in view of the origin of schizophrenic
symptoms is offered by the study of symptomatic psychoses, e.g. alcohol
psychoses. Studies at our hospital showed reduced growth hormone secretion
following alcohol intake using the clonidine test (MA TUSSEK et al., 1984).
Following years of alcohol abuse an increase in noradrenaline is measured on
the basis of the MHPG when delirium occurs (ACKENHEIL and BECKMANN,
1978). The symptoms which appear differ from patient to patient and
throughout the course of alcohol abuse - still another example of the fact
that in order for psychopathological symptoms to originate, several must
coincide whose interaction remains unknown to us even today.

What do the previous examples illustrate:

1. The biochemical research on schizophrenia first made decisive progress

in combination with research on the activity of neuroleptics. The most
important result of this development to date is the dopamine hypothesis
of schizophrenia.

29
 ~r:-t·l
0
0
0

0 I 0 ~

--
0 ... 0

...
8 ,.. •
~... :1 ...• .. 3
0
-3
...
II ••
•••
8- u

il.
0
80
•II
•••
v
0

• •
8o
..
:I 0

•
• ::• ~~
E
.. 0
0

~
v

)( )(
• )(
Ill Ill Ill
E E E
ID ID ID

rec:trs NE
platelets serum
rec/An E
granulocytes serum
DA
receptors DA
lymphocytes serum

Fig. 5 binding studies on blood cells of acute schizophrenic patients and

serum norepinephrine (NE), epinephrine (E) and dopamine (DA)
WD controls (n=40)
o o o individual patients
tJ.. receptors measured with 3 H Yohimbin
13 receptirs measured with 3 H dihydroalprenolol
DA - receptors measured with 3 H spiperone
(Method: Bondy et al., 1984)

30
 SPECIFIC 3H- SPIPERONE BINDING TO
LYMPHOCYTES

-
-~
en • ~
0
G)~
(,)

1.0 0
0- • ....,
0

m
::::ao
0
Ec.!) ~
-N

~0
•
...... \ ~

-
~N ~-'l~·1
0

c
m ~

2 3 2 3

CATECHOLAMINES IN SERUM
NA A DA
~ •
0 .,c--
~
8
"'
E
:§ • .,
0
•
Cll

m
en
•• ., •
-
N
E~
•• •
Ol
Q. ~ I
,•
••
,
, 2 3 2 3 2 3
eHOSPITALIZE;J PATIENT
eMOTHER OFATHER

Fig. 6 3H-spiperone binding of lymphocytes of acute schizophrenic patients

and their families. Catcholamines in Serum.
~controls (n=40)

31
2. Despite this, in studies designed to uncover basic biochemical
disturbances in schizophrenia the perspective cannot be limited to a
single parameter, e.g. dopamine. This is illustrated by our studies of the
noradrenaline system in acute schizophrenics, in patients receiving long-
term (more than 10 years) treatment with neuroleptics prior to and after
withdrawal trials.

3. It must be recognized that the deviations from the biochemical norm

uncoverd in schizophrenics are not the causes but rather the
accompnying factors or possibly also unspecific consequential
phenomena. Such findings are not necessarily worthless for this reason,
as can be shown in terms of the understandable and elucidative findings
for patients from low and high expressed behavior families with regard
to stress concepts.

4. Also in interpreting differences which have been discoverd between

schizophrenics and normal controls it is always wise to exercise caution.
In the future it will be necessary to resort much more frequently to
control groups other than simply the so-called healthy controls.

5. The selection of the parameters to be studied should never be too

narrow. Even when, for example, the antipsychotic efficacy of
neuroleptics had drawn attention primarily to the dopamine system,
studies of the noradrenaline system in acute schizophrenics and above
all, prior to and after withdrawals in patients treated long-term with
neuroleptics have shown us that this transmitter system is also involved
in the biochemical processes in schizophrenics.

6. Changes in schizophrenics which have been detected and established

should never be simply etiologically interpreted without further ado.
Thus our findings on the noradrenaline system are meaningful possibly
only in terms of an interpretation based on the stress concept. Hereby,
however, criteria might be collected which would permit statements as
to the different vulnerability, finally as to the danger of relapse.

7. Biochemical and also all other findings acquired using biological methods
should not always be projected solely in terms of a relationship to
nosological perspectives. Every finding should be tested with regard to
its causal relationships, its specificity, for example, in terms of
syndromes. Recently there has been a positive trend towards at least
posing the question of whether trait or state causes are involved.
However, up to now the question of whether biological parameters do not
indeed correlate much more with characteristics of the course of the
illness rather than with symptomatological or even etiological aspects
has been completely neglected.

Further frames of reference for biochemical and neurophysiological findings

might be, in addition to course criteria, intensity and acuity parameters,
genetic aspects, therapeutic criteria, for example, the comparison or

32
"responders" and "non-responders" in different therapy procedures. Only if
these and other frames of references are considered in addition to the
nosological-diagnostic apsects when interpreting biological-psychiatric
findings, can the cognitive possibilities of biological-psychiatric research be
better utilized in the future. This also means that the future questions in
biological psychiatry can no longer be oriented solely toward certain
(nosological or syndromally defined) groups of psychopathological clinical
pictures.
The aim of our general remarks on the biological aspects of endogenous
psychosis at the beginning of the session was to guide biological-psychiatric
research away from far too strong an attachement to psychiatric nosology,
without, however, intending to cast a doubt as to its importance to the
research on the pathophysiological fundaments of certain psychiatric illness.

Literature

M. Ackenheil, D. Athen, H. Beckmann: Pathophysiology of delirious states. J

of Neural Transmission, Suppl. 14, 167-175 (1978)

M. Ackenheil, H. Hippius, N. Matussek: Ergebnisse der biochemischen

Forschung auf dem Schizophreniegebit. Der Nervenarzt 49, 634-649 (1978)

M. Ackenheil, R. Engel: Stressuntersuchungen bei psychiatrischen Patienten.

75-Jahrfeier der Psychiatrischen Klinik der Universitiite MUnchen.
MUnchener Med Wschr, 123, No 19,805-808 (1981)

M. Ackenheil: Biochemical effects of apomorphine: contribution to

schizophrenia research: apomorphine and other dopaminomimetics, Vl 2,
Clinical Pharmacol, 215-225, Raven Press N.W. Eds.: G.U. Corsini, G.L.
Gessa (1981)

M. Ackenheil, U. MUnch, H. Wissmann: Die Bedeutung des Mienenspiels in

der Psychiatrie. 99. Wanderversammlung sUdwestdeutscher Psychiater und
Neurologen Baden-Baden, 1983.

M. Albus, M. Ackenheil, U. MUnche, D. Naber: Ceruletide: A new drug for

the treatment of schizophrenic patients? Arch Gen Psychiatry, Vol 4, 528
(1984)

M. Ansseau, A. Doumont, J.J. Legros: Evidence for two subgroups of

schizophrenics with regard to dopaminergic sensitivity. 5th CA Symposium
Goteborg 1983. Abstracts. Suppl to Progress in Neuro-Psychopharmacology
and Biological Psychiatry, p 69

B. Bondy, M. Ackenheil, R. Elbers, M. Frohler: Binding of H-spiperone to

human lymphocytes: a biological marker in schizophrenia? Accepted by
Psychiatry Research 1984

33
A. Carlsson, w. Kehr, M. Linquist: Receptor mediated control of dopamine-
synthesis. In: Neuropsychopharmacology. Ed.: J.R. Boissier, H. Hippius, P.
Pichot. American Elsevier Publ Comp Inc New York, pp 468-471 (1975)

K. Conrad, W. Scheid, H.J. Weitbrecht: Die beginnende Schizophrenie.

Versuch einer Gestaltanalyse des Wahns. Sammlung psychiatrischer u,nd
neurologischer Einzeldarstellungen. Georg Thieme Verlag, Stuttart (1958).

I. Creese, D.R. Burt, S.H. Snyder: Dopamine receptor binding:

Differentiation of agonist and antagonist states with H-dopamine and
Halo per idol. Life Science 17, 99 3-1002 (197 5)

A.J. Cross, T.J. Crow, A. Longden, F. Owen, M. Poulter, G.J. Riley:

Evidence for increased dopamine receptor sensitivity in post mortem brains
from patients with schizophrenia. J Physiol 290:456 (1978)

T.J. Crow: Molecular pathology of schizophrenia: more than one disease

process. Br Med J 6207:66-68 (1980)

T. Eibl-Eiblsfeld: Elternhal tung, psychophysiologische Erregung und RUck-

fallquote bei schizophrenen Patienten. Diss. vor der med. Fakulti:it der LMU
MUnchen, 1983.

G. Grof3: Presynaptic alpha-adrenergic neurones. Nauyn-Schmiedeberg's

Arch Pharmacal Sppl 313:387 (1980)

H. Katschig: Sozialer Stress und psychische Krankheit. In: Fortschritte der

Sozialpsychiatrie, Vol 5. Urban und Schwarzenberg, MUnchen (1978)

R.E. Kendell: Waht are our criteria for a diagnosis of schizophrenia? In:
H.M. van Praag (ed.). On the origin of schizophrenic psychoses. E Evven
Bohn BV, Amsterdam (197 5)

CH. Lake, M. Ziegler, I. Kopin: Use of plasma norepinephrine for evaluation

of sympathetic neuronal function in man. Life Sciences 18, 1315-1326 (1976)

Y. Lecrubier and P. Douillet. In: N. Ma tussek et al., Noradrenalin plasma

level after sulpiride and other antidepressants during rest and ergomety. In:
N. Matussek and M. Ackenheil (eds.): Special aspects of
psychopharmacology. Possible clinical significance of recent biochemical
and pharmacological findings with ortho-methoxy-benzamides. Expansiom
Scientifique Francaise, Paris (1983).

S. Matthyse: Antipsychotic drug actions: a cue to the neuropathology of

schizophrenia. Fed Proc, 200-205 (1973)

N. Matussek, M. Ackenheil, M. Herz: The dependence of the Clonidine

growth hormone test on alcohol drinking habits and the menstrual cycle.
(Accepted for publication by Psychoneuroendocrinology, 1984)

34
D. Naber, F. MUller-Spahn: Effects of long-term neuroleptic treatment on serum
levels of prolactin, TSH, LH and norepinephine, alpha-adrenergic and
dopaminergic receptor sensitivity. Relation to tardive dyskinesia, 1983.

D. Naber, C. Finkbeiner, H. Fischer, K. Zander, M. Ackenheil: Effect of

long-term neuroleptic treatment on prolactin and norepinephrine levels in
serum of chronic schizophrenics: Relations to psychopathology and
extrapyramidal symptoms. Neuropsychobiology 6, 181-189 (1980)

D. Naber, M. Albus, F. MUller-Spahn, H. BUrke, D. Welter, M. Ackenheil, H.

Hippius: B-endorphin, cortisol and prolactin in serum of schizophrenic
patients during long-term neuroleptic treatment and after withdrawal.
Psychopharm Bull 18, 224-226 (1982)

D. Naber und D. Pickar: Endorphins in CSF and plasma of psychiatric

patients. Arzneim Forsch 32, 877-878 (1982)

D. Naber, U. MUnch, J. Wissmann, R. Grosse, R. Ritt, D. Welter: Naloxone

treatment for 5 days ineffective in schizophrenia. Neuroendocrine actions of
the opiate antagonist. Acta Psychiat Scand 67, 265-271 (1983)

G.N. Pandey, D.L. Garver, C. Tamminga, S. Ericksen, S.J. Ali, J.M. Davis:
Postsynaptic supersensitivity in schizophrenia. Am J Psychiat 134:518-522
( 1977).

J. Rotrosen, B. Angrist, S. Gershon, J. Paquin, L. Branchey, M. Oleshansky,

F. Halpern, E.J. Sachar: Neuroendocrine effects of apomorphine:
characterization of response pattern and application to schizophrenia
research. Brit J Psychiat 135, 444-456 (1979)

N. Tarrier, C.E. Vaughn, M.H. Lader, J.P. Leff: Bodily reactions to people
and events in schizophrenics. Archives of General Psychiatry, 36:311-315
(1979)

K.J. Zander, B. Fischer, R. Zimmer, M. Ackenheil: Long-term neuroleptic

treatment of chronic schizophrenic patients: Clinical and biochemical
effects of withdrawal. Psychopharmacology 73:43-47 (1981). (This
publication will be published more detailed in "Der Nervenarzt" (1984).

35
EVIDENCE FOR ALTERATIONS OF BRAIN MORPHOLOGY AND FUNCTION IN
SUBGROUPS OF SCHIZOPHRENIC SUBJECTS

Goran Sedvall, Gustaf Bergstrand, Lars Bjerkenstedt,

Henrik Nyback, Gabriella Oxenstierna and Frits-Axel Wiesel
Department of Psychiatry and Psychology
Karolinska institute
S-104 01 Stockholm
Sweden

INTRODUCTION
The concept of functional psychosis implicates the occurrence
of psychosis in patients where no structural brain damage can be
demonstrated. The development in neurobiology over the last few
years makes it reasonable to assume that the elaboration of new
sophisticated methods in the future will have great implications
for the diagnostic classification of functional psychoses. New
biochemical and neuroradiological methods have been developed du-
ring the past two decades. These methods are already today applied
to groups of psychiatric patients. A number of altered functions
in patients with functional psychosis have already been demonst-
rated with such methods.
Central Monoamine Metabolism
The profound effects of hallucinogenic and antipsychotic
drugs on mentation in healthy volunteers and schizophrenic sub-
jects implicate that these chemical compounds have profound ef-
fects on central nervous mechanisms related to higher mental func-
tion. The direct evidence from a number of neuropharmacological
studies indicate that minute concentrations of these compounds have
profound effects on receptor mechanisms related to synaptic trans-
mission mechanisms where serotonin, dopamine and noradrenalin are
involved.
By studying the concentrations of the major metabolites of
these transmitter amines in cerebrospinal fluid (CSF), we attemp-

37
ted at classifying cases of functional psychosis with regard to
such measurements. However the importance of a number of variance
factors for these indirect variables of monoaminergic function like
age, height, level of CSF sampling, position of subject during CSF
sampling, physical activity of subject, drug treatment and possibly
also season and day of sampling complicates the interpretation of
the results. In systematic studies in healthy volunteers we have
obtained evidence that familial factors, possibly of both genetic
and environmental type, play a role for the mono.mine.metabolites
concentrations (Sedvall et al 1982). Interestingly enough we found
in groups of both healthy volunteers and untreated schizophrenic
subjects that the variance of concentrations in cases with a family
history of psychiatric disorder was greater than in subjects with-
out family history for psychiatric disorders (Sedvall et al 1980,
Sedvall and Wode-Helgodt 1980). Cases of schizophrenic patients
with a family history for psychosis of schizophrenic type had not
only a wider range of values but there was also a subgroup with
high concentrations of both 5-hydroxy-indoleacetic acid (5-HIAA)
and homovanillic acid (HVA) who may have a familial disposition
for the disorder. On the basis of these transmitter metabolic stu-
dies, the results allowed the formulation of the hypothesis that
schizophrenic psychosis may be classified both with regard to fa-
milial disposition for the disorder and concentration of monoamin-
ergic transmitter metabolites in the CSF. Since the deviant mono-
amine metabolite concentrations were found also in healthy sub-
jects with a family history of psychosis, the high amine concentra-
tions per se do not seem to determine the development of psychosis
but rather reflect a disposition for the development of such a dis-
order, i.e. a trait and not state dependant variable.
New Neuroradiological Methods fer Research in Psychiatry
Deviant concentrations of monoamine metabolites in the CSF
may be related to changes in central monoamine metabolism or changes
in the circulation of cerebrospinal fluid (CSF). During recent
years, a number of new neuroradiological techniqu~s have been de-
veloped allowing the analysis of morphological and functional as-
pects of the central nervous system in patients and healthy volun-
teers. Using computerized axial tomography (CAT), several investi-
gators have presented evidence for the occurrence of degenerative
changes in the brain of some schizophrenic patients. In comparative
studies between young schizophrenic patients and healthy volunteers
we have found support for this view (Nyback et al 1982). We found
evidence for increased ventricular size in young schizophrenic
patients. Interestingly enough we also found a significant negative
correlation between concentrations of the monoamine metabolites
HVA and 5-HIAA and ventricular size in these patients (Nyback et
al 1983). These biochemical and morphological findings in schizo-
phrenic patients may be related to different causal mechanisms

38
which have to be explored. An alteration of CSF circulation may be
responsible for both these changes in some patients. The circula-
tion of the CSF within the brain has not previously been systema-
tically studied in schizophrenic patients. We have therefore ini-
tiated studies of CSF circulation in schizophrenic patients (Oxen-
stierna et al., 1983). Using isotope cisternography (ICG) we found
evidence that a substantial proportion (about 30%) of a group of
young schizophrenic patients shows qualitative evidence of an alte-
ration of CSF circulation. Thus, 15% of the patients have evidence
of a marked pathological change of CSF flux over the convexity of
the brain or evidence of retroflux of CSF within the ventricular
system. All the patients had normal pressure within the ventricu-
lar system and a normal Queckenstedt test. Two patients who had
been drug free for three weeks before the investigation also demon-
strated these changes. This finding and the fact that most of the
patients who had a normal CSF circulation were on high doses of
neuroleptics indicates that there is no relationship between the
pathological findings at ICG and drug treatment.
CAT studies of cerebral morphological changes in the patients
with disturbed CSF circulation indicated that brain atrophy could
occur in the presence or in the absence of a disturbed CSF circu-
lation. Several patients showed evidence of atrophy without any
disturbed circulation and vice versa.
In this study (Oxenstierna et al., 1983), we found no statis-
tical correlation between deviant concentrations of monoamine meta-
bolites and alteration of CSF circulation or atrophy. However, the
interpretation of these data is not clear since drug treatment
markedly influences the concentrations of monoamine metabolites in
the CSF. For ethical reasons, this series of schizophrenic patients
was studied with regard to CSF circulation without drug withdrawal.
We found no evidence for a relationship between disturbed CSF
circulation and familial occurrence of schizophrenia. On the other
hand, we found a significant relationship between the occurrence
of brain atrophy and familial form of schizophrenia in these sub-
jects (Table I).
Table I. Relationship between Brain Atrophy and Absence of Family
History for Psychosis in Schizophrenic Subjects
No family history Family history
for psychosis for psychosis
Atrophy 9 1 10
No atrophy 9 9 18
Fisher~s exact test p < 0.05
Oxenstierna et al 1983.

39
 The alterations of brain morphology and function reported in
this communication may or may not be related to the occurrence of
functional psychosis in these subjects. The high frequency of occu-
rrence of the deviations contradicts the view that they are unre-
lated. The fact that the different types of deviations were not
significantly related may indicate that they are due to different
ethiologies within the category of functional psychosis. Our
findings that high monoamine metabolite concentrations on one hand
and atrophy on the other were significantly related to familial
and non-familial forms of schizophrenia respectively may reflect
the occurrence of at least two different ethiologies. One category
of familial forms of schizophrenia may be related to defective
development or control of central monoaminergic mechanisms. A non-
familial form of schizophrenia with evidence of brain brain atrophy
may be the result of brain infection or trauma. Toxic environmental
influences on neuronal development, non-specific or specific, for
some of the monoaminergic systems may also be considered for the
schizophrenias who have evidence of brain atrophy.
The schizophrenic patients who showed evidence of a disturbed
CSF circulation were similarily distributed between the categories
of familial and non-familial forms of schizophrenia. Therefore, a
heterogenous ethiology may also be considered for these cases.
Some of these cases may be related to early brain trauma or infec-
tion. Other cases may have a genetic origin. Further studies with
more refined techniques for the measurement of CSF circulation in
families with schizophrenic cases are required to clarify these
questions.
CONCLUSIONS
A substantial proportion of patients with psychosis of func-
tional type have evidence of:
1. A local or general brain degeneration.
2. A disturbed circulation of cerebrospinal fluid.
3. De~iant concentrations of monoamine metabolites in the CSF.
These three different alterations of brain morphology or func-
tion do not seem to be correlated.
The majority of cases with cerebral atrophy have a non-fami-
lial type of schizophrenia.
ACKNOWLEDGMENT
The study was supported by grants from The Swedish Medical
Research Council (B84-21X-3560-13C), Bergvall~s Foundation, Karo-
linska Institute and F. Hoffman La-Roche & Co.

40
REFERENCES
Nyb~ck, H., Wiesel, F.-A., Berggren, B.-M. and Hindmarsh, T., 1982,
Computed tomography of the brain in patients with acute
psychosis and in healthy volunteers. Acta Psvchiat. Scand.,
65:403.
Nyb~ck, H., Berggren, B.-M., Hindmarsh, T., Sedvall, G. and Wiesel,
F.-A., 1983, Cerebroventricular size and cerebrospinal fluid
monoamine metabolites in schizophrenic patients and healthy
volunteers. Psvchiatrv Res., in press
Oxenstierna, G., Bergstrand, G., Sedvall, G. and Wik, G., 1983,
Evidence for a disturbed CSF circulation in cases of schizo-
phrenic psychosis. In preparation.
Sedvall, G., Fyro, B., Gullberg, B., Nyb~ck, H., Wiesel, F.-A. and
Wode-Helgodt, B., 1980, Relationships in healthy volunteers
between concentrations of monoamine metabolites in cerebro-
spinal fluid and family history of psychiatric morbidity.
Brit. J. Psvchiat., 136:366.
Sedvall, G. and Wode-Helgodt, B., 1980, Aberrant monoamine meta-
bolite levels in CSF and family history of schizophrenia.
Arch. Gen. Psychiat., 37:1113.
Sedvall, G., 1se11us, L., Nyb~ck, H., Oreland, L., Oxenstierna, G.,
Ross, S.B. and Wiesel, F.-A., 1982h Genetic studies. of CSF
monoamine metabolites. Paper presented at the Nobel conference
"Frontiers in Biochemical and Pharmacological Research in
Depression", at Skokloster, ~weaen, June ~~-~~

41
PSYCHOBIOLOGICAL STUDIES OF MANIA

William E. Bunney, Jr. and Blynn Garland

Department of Psychiatry, California College of
Medicine, University of California
Irvine, California 92717

INTRODUCTION

This paper will review some clinical and biological aspects of

mania including symptoms, diagnosis and treatments. Special empha-
sis will be placed on data which may be relevant to understanding
mania such as receptor studies, effects of lithium on behavioral su-
persensitivity, incidence of tardive dyskinesia in mania, and time-
dependent changes in vulnerability to the switch. Finally, some
newer treatments for mania will be briefly discussed.

CLINICAL OBSERVATIONS

A careful analysis of the onset, course and symptoms of mania

may provide clues to its etiology. The average age of onset of bipo-
lar illness which includes episodes of mania and depression has been
reported to be about 30 years with females having a slightly later
onset of about 35 years (Perris, 1969); a review of the literature
includes an increasing number of reports of mania in children and
adolescents (Hassanyeh and Davison, 1980). Briefly, the symptoms of
mania consist of an increase in motor and verbal activity, grandios-
ity, intrusiveness, euphoria, increased interest in sex, bizarre
manner, and loss of sleep. Using the Manic-State Rating Scale, Bie-
gel and Murphy (1971) describe two types of manic patients: one sho-
wing elation and grandiosity with little paranoia and -the other with
minimal euphoria but more paranoid and destructive behavior. It is
interesting to note that this distinction is consistent with reports
as early as 1674: "One variety may be crowded with horrible, violent
and truculent symptoms. Another variety is mild and is character-
ized by gentler symptoms like laughter, singing, etc."(Leyden, 1674)

43
Another possible distinction of mania may be made in view of the
rapidity of its onset. The switch into mania may occur in some
patients over a period of days to weeks while in others, manic
symptoms can occur suddenly, sometimes in a matter of minutes to
hours. Post and coworkers (1981) suggest that the two patterns of
onset may provide clues to underlying biological substrates, nota-
bly that patients with rapid onsets experience more affective epi-
sodes/year and have had a longer duration of the illness as compa-
red to slow onset patients, which may imply progressive increases
in behavioral sensitization as is observed in some drug-sensitiza-
tion paradigms.

THE SWITCH PROCESS

The switch into mania is a fascinating and striking phenomen-

on which occurs naturally or can be induced by most of the success-
ful treatments for depression including tricyclics, MAOis, ECT,
and sleep deprivation. In a review of 4000 patients in the world
literature, approximately 10% of the patients treated with antide-
pressants switched into mania (Bunney et al., 1978). Because a
number of switches occurred within weeks after drug administration
in patients with no history of hypomania/mania it appears highly
unlikely that these switches were spontaneous. For example, van
Scheyen and van Kammen (1979) randomly administered clomipramine,
a tricyclic antidepressant, to 25 unipolar patients and reported
that 6 patients switched into mania; the switch occurred abruptly,
within a couple of hours in two patients and gradually, over the
course of four days to two weeks in four patients. The mean age of
onset of the first episode of mania was 63 years, notably above
the age of risk, and occurred after an average treatment length of
57 days. Other types of antidepressants have also been associated
with switches into mania. More recently, the administration of a
potential new antidepressant compound, s-adenosylmethionine, was
reported to induce switches from depression into mania after only
six days of treatment in three of 12 depressed patients (Carney et
al., 1983). Another example is sleep deprivation. Wehr et al.,
(1978) demonstrated switches in 7/9 rapidly cycling manic-depres-
sive patients after one night of sleep loss. In another study,
Post et al., (1976) reported similar results in a patient whose
depression was interrupted with sleep deprivation which precipita-
ted mania. It is generally well accepted that bipolar patients
are more sensitive to treatment-induced manias but now there is
increasing evidence that it is possible to switch unipolar pati-
ents. Theoretical implications of this data suggest a possible
common biological defect which may be triggered to increase beha-
vioral sensitization in response to antidepressant treatment.
Further study is needed to identify markers of "high-risk" vulner-
ability to the switch process.

44
LITHIUM

It is well known that lithium is effective in the treatment

of mania and depression and is the treatment of choice for preven-
ting or attenuating recurrent episodes. Data has suggested that
Li may act by blocking the development of supersensitivity associ-
ated with the switch process, specifically the increase in behavi-
oral and receptor sensitivity to dopamine agonists (Bunney et al.,
1977). It has been proposed that changes in sensitivity may deve-
lop through the various cycles of manic-depressive illness; de-
pression associated with a decrease in catecholamines, resulting
in an increase in DA sensitivity and the initial stages of mania
characterized by increases in responsivity, possibly as a result
of increased levels of dopamine (Bunney et al., 1977).

There is a growing body of evidence which suggests that lith-

ium prevents the development of dopaminergic supersensitivity. In
animal studies, it was shown that treatment with haloperidol, aDA
antagonist, produced an increase in locomotor and stereotypy in
response to the DA agonist, apomorphine. Pretreatment with Li was
shown to be effective in blocking the development of DA behavioral
supersensitivity in 10/11 studies (Bunney and Garland, 1983). The
proliferation of DA receptors associated with behavioral supersen-
sitivity has been shown to be ameliorated with Li treatment in
some studies (Pert et al., 1978; Tanimoto et al., 1983) although
others were unable to replicate these results (Staunton et al.,
1982).

Other models which might be useful in studying the effect of

Li on the development of supersensitivity include those diseases
hypothesized to be dopaminergic supersensitivity illnesses. A
number of investigators have studied the effects of Li administra-
tion on tardive dyskinesia (TD), a possible DA supersensitivity
disease associated with the chronic administration of neurolep-
tics. In a recent review, Gardos and Cole (1983) reported that
lithium can strikingly decrease the incidence of tardive dyskine-
sia, but this effect certainly does not occur in all patients.
Similar beneficial results have been reported in the "on-off" syn-
drome observed in Parkinsonian patients following the administra-
tion of the DA agonist, L-dopa. This syndrome, characterized by
fluctuations in akinesia and dyskinesia, has been shown to be sig-
nificantly modified by lithium in some patients (Coffey et al.,
1982). Other hypothesized DA-related disorders include Hunting-
ton's Chorea and Gilles de la Tourette syndrome, and as reviewed
by Yassa and Ananth (1980), Li was reported to be efficacious in
some individuals. This is an area where there will be much re-
search in the future. Li treatment certainly does not work in
schizophrenia, an hypothesized DA illness, but one might suggest
the hypothesis that Li may block the development of fluctuations
associated with changes in receptors if it is administered prior
to the development of DA receptor supersensitivity.

45
 A possibly exciting clue to the understanding of mania cen-
ters around the recent reports in the literature describing
state-dependent dyskinesias associated with manic-depressive cy-
cles. This striking phenomenon was described by Cutler et al.,
(1981; 1982) who reported increased TD symptoms during periods of
depression and a remission of symptoms during mania in a rapidly
cycling manic-depressive patient, consistent throughout the
course of the patient's hospitalization. Other investigators
have reported similar patients (Weiner and Werner, 1982; Apple-
baum, 1982).

BIOLOGICAL RHYTHMS

There is evidence that rhythms might serve as biological

markers of "vulnerability" to the switch process. In a longitudi-
nal study of one patient with regularly occurring 48 hour manic-
depressive cycles, 173 switches into mania were recorded. A sig-
nificantly higher number of switches occurred at night between 12
and BAM with the peak incidence occurring between 4 and 6am (Sit-
aram et al., 1978a). Additional data from a study of 75 bipolar
I patients hospitalized at the NIMH showed that patients who swi-
tched into mania at night were rated as significantly more manic
during the 4 days following the switch and had significantly less
sleep during this period of time (Sitaram et al., 1978b).

One may raise the question of factors which may alter night-
time "high-risk" switch periods in rapidly cycling patients.
There is evidence that sleep does not block the switch process.
In 10 patients, 14 rapid switches into mania were recorded in the
early morning hours out of sleep (Sitaram et al., 1978b). Fur-
thermore, in an EEG study over a period of 58 nights in a rapidly
cycling patient, Gillin et al., (1977) showed that out of 8 swit-
ches into mania, 50% occurred during sleep. Sleep deprivation,
on the other hand, has been shown to precipitate mania in rapidly
cycling bipolar patients (Wehr et al., 1978).

It is interesting to note that most of the data from these

studies is specific to rapid cyclers (those who switched in 24
hours or less) which may be suggestive of a new bipolar subtype.
A study of 75 bipolar patients hospitalized at the NIMH from
1963-1975 reported that no patient showed both rapid and slow
switches, providing further evidence that this group may be dif-
ferent (Sitaram et al., 1978b).

Another variable may be that of phase rhythms associated

with the course of manic and depressive episodes. Some prelimin-
ary evidence suggests differences in early, middle and late

46
mania. Post et al., (1977) used telemetric measurements of ac-
t1v1ty in a rapidly cycling depressive/hypomanic (BPI) patient
and indicated that at the end of the depressive episode there was a
significant increase in activity indicating the imminent onset of a
switch. The highest levels during the manic cycle occurred in ear-
ly mania after which there was a gradual decrement. Other prelimi-
nary case study data supporting phase rhythms comes from studies of
plasma melatonin in manics as compared to controls. Highest total
melatonin secretion occurred in early mania followed by lower le-
vels in middle and late mania. Again, the greatest differences in
24-hr plasma melatonin levels between manics and controls occurred
at night (Lewy et al.,l979). Additional preliminary case study
findings concern state-dependent tardive dyskinesia which has been
demonstrated to be somewhat ameliorated during the afternoon as
compared to morning values in both mania and depression (Cutler et
al.,l981).

Another type of rhythm which might be relevant to understan-

ding the switch process is that of seasonal rhythms. In a recent
publication, Rosenthal et al., (1983) described a male bipolar pa-
tient with a 13-year history of depressive and hypomanic episodes
which are closely related to seasonal changes. In a follow- up
study (Rosenthal and Sack, in press) described 79 patients with
seasonal affective disorder (SAD) who developed hyperphagic, hyper-
somnic depressions regularly in the fall and winter and are euthym-
ic or hypomanic in the spring or summer. One of the factors which
is hypothesized to be associated with SAD is photoperiodicity.
This same research group obtained data to determine the average
photoperiod for each month of the year in Rockville, Maryland where
the study was conducted. They reported a significant negative cor-
relation between the percentage of patients depressed per month and
the mean photoperiod. Light was then proposed as a significant
factor in the onset of the switch. In a double-blind crossover
study (Rosenthal and Sack, in press) 9 depressed patients with sea-
sonal depressions were exposed either to bright fluorescent light
(approximately 2500 lux at three feet) or dim light (approximately
100 lux at three feet). Results showed a significant decrease in
depression ratings in patients exposed to bright light as compared
to the dim condition. In addition, 50% of manic-depressive pati-
ents had suppressed melatonin levels in response to 500 lux of
light; this intensity had little or no effect in normal controls or
1n euthymic patients, suggestin that manic-depressive patients may
be supersensitive to light (Lewy et al., 1982).

Future research might focus on identifying relevant biological

rhythms involved with the switch process and on new treatments
which may alter them. Recent work on lithium and the antidepres-
sants has been shown to slow or dissociate circadian rhythms
(Wirz-Justice and Campbell, 1982).

47
SUMMARY

A review of the literature reveals many details of the switch

process into mania and suggests several considerations including a.
possible refinement of diagnostic procedures. Although the switch
often appears to be uniform among bipolars, a closer analysis re-
veals several distinguishing features of possible subtypes. These
include rapid versus slow switchers, seasonal switchers, and night-
switchers. In addition, there are those manics who have a predomin-
ance of one set of symptoms over another. It may be possible to use
these clues to help in the identification of "high- risk" periods for
these individuals. Thus far, it appears that rapid cyclers have
increased vulnerability at night, seasonally affective disorder indi-
viduals appear to be more vulnerable during periods of change to
short daylight, and there seems to be behavioral and biochemical
variations in early and late manic phases. Research in the area of
biological rhythms is very prom1s1ng and may provide clues to the
etiology of affective illness.

REFERENCES

Applebaum, P.S., 1982, Dyskinesia and unipolar depression, Am~

Psvchiatrt 139: 140.
Biegel, A. and Murphy, D.L., 1971, Assessing clinical characteristics
of the manic state, Am J Psvchiatry 128: 688.
Carney, M.W.P., Martin, R., Bottiglieri, T., Reynolds, E.H., Nis-
senbaum, H., Toone, B.K., and Sheffield, B.F., 1983, Switch
mechanism in affective illness and s-adenosylmethionine,
Lancet 1: 820.
Bunney, W.E., Jr., 1978, Psychopharmacology of the switch process in
affective illness, in: "Psychopharmacology- A Generation of
Progress", M. Lipton, A. DiMascio, and K. Killam, eds., Raven
Press, New York.
Bunney, W.E., Jr., and Garland, B.L., 1983, Possible receptor effects
of chronic lithium administration, Neuropharmacology 22, 367.
Bunney, W.E., Jr., Post, R.M., Andersen, A.E. and Kopanda, R., 1977,
A neuronal receptor sensitivity mechanism in affective
illness: A review of evidence, Comm Psychopharm 1: 393.
Coffey, C.E., Ross, D.R., Ferren, E.L., Sullivan, J.L., and Olanow,
C.W., 1982, Treatment of the "on-off" phenomenon in Parkins-
onism with lithium carbonate, Ann Neurol. 12: 375.
Cutler, N.R., Post, R.M., Rey, A.C. and Bunney, W.E., Jr., 1981,
Depression-dependent dyskinesias in two cases of
manic-depressive illness, New En2 J Med 304: 1088.
Cutler, N. R. and Post, R.M., 1982, State-relate.d cyclical dyskinesias
in manic-depressive illness, J Clin Psvchopharm 2: 350.
Gardos, G. and Cole, J.O., 1983, Tardive dyskinesia and. anticholiner-
gic drugs, Am J Psvchiatrv 140: 200.

48
Gillin, J.C., Mazure, C., Post, R.M., Jimerson, D., and Bunney, W.E.,
Jr., 1977, An EEG sleep study of a bipolar (manic-depressive)
patient with a nocturnal switch process, Biol Psychiatry,l2:
711.
Hassanyeh, F., and Davison, K., 1980, Bipolar affective psychosis with
onset before age 16 years: Report of 10 cases, Brit J
Psvchiatry, 137: 530.
Lewy, A.J., Wehr, T.A., Gold, P.W., and Goodwin, F.K., 1979, Plasma
melatonin in manic-depressive illness, in "Catecholamines:
Basic and Clinical Frontiers", Usdin, E., Kopin, I.J. and
Barchas, J.D. (eds.), Pergamon Press, New York.
Lewy, A.J., Wehr, T.A., Rosenthal, N.E., Nurnberger, J.I., Siever,
L.J., Uhde, T.W, Newsome, D.A., Becker, L.E., Markey, S.P.,
Kopin, I.J., and Goodwin, F.K., 1982, Melatonin secretion as
a neurobiological "marker" and effects of light in humans,
Psvchooharmacology Bulletin, 18: 127.
Leyden, A.S., 1674, Dissertatwn on "De mania" by Lannoy, A. as
reviewed by Diethelm, 0., (1970), Mania: A clinical study of
dissertations before 1750, Confinia Psvchiatrica. 13: 26.
Perris, C., 1969, The separation of bipolar (manic-depressive) from
unipolar recurrent depressive psychoses, Behavioral
Neuropsychiatry, 1: 17.
Pert, A.·, Kosenolatt, J.E., Sivit, C., Pert, C.B., and Bunney, W.E.,
Jr., 1978, Long-term treatment with lithium prevents the
development of dopamine receptor supersensitivity, Science,
201: 171.
Post, R.M., Ballenger, J.C., Rey, A.C. and Bunney, W.E., Jr., 1981,
Slow and rapid onset of manic episodes: Implications for
underlying biology, Psvchiatrv RPs, 4: 229.
Post, R.M., Kotin, J. and Go.odwin, F.K., f976, Effects of sleep
deprivation on mood and central amine metabolism in depressed
patients, Arch Gen Psychiatry, 33: 627.
Post, R.M., Stoddard, F.J., Gillin, J.C., Buchsbaum, M.S., Runkle,
D.C., Black, K.E. and Bunney, W.E., Jr., 1977, Alterations in
motor activity, sleep, and biochemistry in a cycling
manic-depressive patient, Arch Gen Psvchiatrv, 34: 470.
Rosenthal, N.E., Lewy, A.J., Wehr, T.A., Kern, H.E. and Goodwin, F.K.,
1983, Seasonal cycling in a bipolar patient, Psychiatry Res,
8: 25
Rosenthal, N.E. and Sack, D.A. (in press), Affective disorders and the
seasons: The possible role of light, Biolo~ical Psvchiatrv.
Sitaram, N., Gillin, J.C. and Bunney, W.E., Jr., 1978a, Circadian
variation in the time of "switch" of a patient with 48-hour
manic-depressive cycles, Bioi Psychiatry, 13: 567.
Sitaram, N., Gillin, J.C. and Bunney, Jr., 1978b, The switch process
in manic-depressive illness: Circadian variation in time of
switch and sleep and manic ratings before and after switch,
Acta Psvchiat Scand 58: 267.

49
Staunton, D.A., Magistretti P.J., Shoemaker, W.J., Deyo, S. and
Bloom, F.E., 1982, Effects of chronic lithium treatment on
dopamine receptors· in the rat corpus striatum. II. No effect on
denervation of neuroleptic-induced supersensitivity, Brain Res,
232: 401.
Tanimoto, K., Maeda, K. and Terada, T., 1983, Inhibitory effect of
lithium on neuroleptic and serotonin receptors in rat
brain, Brain Res, 265: 148.
Van Scheyen, J.D. and van Kammen, D.P., 1979, Clomipramine-induced
mania in unipolar depression, Arch Gen Psychiatry, 36: 560.
Wehr, T., Goodwin, F.K., Wirz-Justice, A., Breitmaier; J., and
Craig, C., 1982, 48-hour sleep-wake cycles in manic-depressive
illness: Naturalistic observations and sleep deprivation
experiments, Arch Gen Psychiatry, 39: 559.
Weiner, W.J. and Werner, T.R., 1982, Mania-induced remission of
tardive dyskinesia in manic-depressive illness, Ann Neuro1,
12: 229. '
Wirz-Justice, A. and Campbell, I.e., 1982, Antidepressant drugs can
slow or dissociate circadian rhythms, Experentia, 38: 1301.
Yassa, R. and Ananth, J., 1980, Lithium carbonate 1n the treatment
of movement disorders, Internat Pharmacopsychiat, 15: 301.

50
BIOCHEMICAL AND MORPHOLOGICAL ABNORMALITIES

IN PARANOID AND NONPARANOID SCHIZOPHRENIA

Ryo Takahashi

Department of Neuropsychiatry
Tokyo Medical and Dental University
Yushima, Bunkyo-ku, Tokyo, Japan

The understanding of the biological mechanism of schizophrenia

has progressed during the past three decades since the advent of
antipsychotic drugs. However, the biochemical findings of the
dopamine system in the central nervous system of the patients are
still inconsistent. One reason for this gap is, of course, that
schizophrenia itself could be a heterogeneous condition. Therefore,
subjects for biological study should be chosen to represent as
homogeneous a group as possible. In this sense, patients with
paranoid schizophrenia form a representative group to study from the
biological point of view.

When we reclassified the 1st-contact schizophrenic patients ac-

cording to DSM-III and Wing's Catego program, paranoid schizophrenia
was relatively homogeneous. A higher percentage of paranoid schizo-
phrenics fell in Categq Class S.O.P. than hebephrenics. Most studies
on the biochemistry of schizophrenia in the literature are concerned
with chronic or acute patients not specified as to subtypes. Only a
few studies report different findings in the amine metabolism of
paranoid schizophrenics as compared to nonparanoid patients.

Dr. Carlsson's study 1 concerning potentiation by a-methyl-p-

tyrosine (a-MPT) of thioridazine in chronic schizophrenia provides
one of the important strategies for proving a dopaminergic
hyperactivity hypothesis in living sch~zophrenic patients. We
undertook a study 2 on the effects of a-MPT in acute schizophrenics
during haloperidol therapy using a double-blind technique. Five of
10 cases were given a-MPT and the other half a placebo. As a
result, four cases in the placebo group showed the predicted results
in that improvement assessed by BPRS did not appear during the
period of placebo administration and plasma prolactin (PRL) did not

51
 Result of studY on 0. -I'll effects__ d ·
p-·

·.a/
•.
...
'·o-··-o· .6

.•
.,; 40

Fig.l "

increase. On the co'ntrary, in -che a-r.1PT group, improvement was

observed and the PRL level increased in three cases. Fig. 1 shows
one of these cases.

It is interesting to note that the three improved cases were all

paranoid schizophrenics, and, of the two non-improved cases, one was
catatonic and the other hebephrenic. These findings are quite
consistent with the notion that a dopamine hyperactivity hypothesis
is applicable to acute paranoid schizophrenia.

Monoamine oxidase activity of platelets from schizophrenic

patients has been extensively studied in many laboratories. Low MAO
activity was often reported to be found in patients with
hallucination and delusion. This finding, however, is still
controversial 3 We have studied MAO activity in schizophrenia
classified by the DSM-III criteria. As a result, there were no
significant differences in MAO activity between the group with
positive symptoms and that with negative symptoms assessed by BPRS.
Then, when nonparanoid and paranoid schizophrenics were compared,
the mean value of MAO activity was not different between these two
groups either. However, as shown in Table 1 when the cut-off point
is placed at the lower limit of the range of MAO activity in normal
controls, low platelet MAO activity was found more frequently in
paranoid than in nonparanoid schizophrenic patients.

Table 1. Subclassification and Patients with MAO Activity

Under Lower Limit of Normal Value

> lower limit of < = lower limit of

normal MAO activity normal MAO activity

Nonparanoid 6 31

Paranoid 18 20

p< 0.01

52
This finding is consistent with Dr. Wyatt's report'+ and suggests
that low MAO activity is, in fact, observed in not all, but a
considerable number, of paranoid schizophrenics.

In this context, it is interesting to examine the urinary

excretion of B-phenylethylamine in paranoid schizophrenics. Since
phenylethylamine (PEA) is an endogenous amine with amphetamine-like
properties, the reports of high urinary PEA excretion in
schizophrenia are noteworthy. Dr. Potkin and his colleagues 5
Urinary Phenylethylamine Excretion reported increased 24-hour
36.3:+:25.8 urinary PEA excretion in
(14) paranoid, but not in nonpara-
PEA
ug/day
23 '3:+:24' 2 noid chronic schizophrenics,
(7)
by employing a mass frag-
mentography method for the
assay of urinary PEA. We
have also measured urinary
PEA excretion in paranoid and
nonparanoid schizophrenics
after wash-out of drugs by
normal schizophrenics
controls nonparonoid paranoid
gas chromatography/mass spec-
troMetry. The results are
Bar indicates S.E. P=0.052 (one way ANOVAl
• P<Q,05 ·•· P=O.l04 (Mann-Whitney u testl shown in Fig .2.
Fig.2
Paranoid schizophrenics excreted significantly higher amounts
of PEA than normal controls. The mean of the paranoid schizophrnics
was higher than that of the nonparanoid schizophrenics, although the
difference is only close to a significant level, with p = 0.1. The
one-way analysis of variance showed marginally significant overall
results. Of the 21 schizophrenics studied, 11 were high excreters
of PEA (more than 21.8 ~g in 24 hours, which is the mean of all
data), compared with two out of 21 normal subjects. The difference
w~s significant, at p = 0.03, by the Fisher exact probability test.
Of the 14 paranoid schizophrenics, nine were high excreters,
compared with two out of 21 normal subjects. This difference was
also significant, at p = 0.01.

Our previous controlled multi-clinical study6 on CT scan of 280

schizophrenics' brains indicated that 55 percent of relatively young
schizophrenics showed overall abnormalities of CT scans, and there
was significant association of overall CT scan abnormality as well
as frontal and temporal cortical atrophy with nonproductive
symptoms. Schizophrenics showed a significantly larger anterior
horn of the lateral ventricle and third ventricle than the matched
control group. When the symptoms assessed by BPRS were compared
with the VBR in discrete areas, nonproductive symptoms were
associated with enlargement of the lateral ventricle. However, the
VBR in each area was not different between the paranoid and
nonparanoid groups. We followed up the same patients for four years

53
 Table 2. Comparison of Mean VBR in Discrete Areas in
Schizophrenics Between 1978 and 1982

all cases matched cases

1978 1982 1978 1982

(n=75) (n=85) (n=64) (n=64)

VBR (L) 8.09 ± 1.83 7.84 ± 2.41 8.15±1.91 8.24 ± 2.48

(n=96) (n=89) (n=85) (n=85)

VBR (AH) 2.30 ± 0.94 2.19 ± 0.89 2.27 ± 0.97 2.21 ±0.90

(n=81) * (n=85) (n=68) ** (n=68)

VBR (III) 0. 70 ± 0. 24 > 0. 61 ± 0. 25 0. 70 ± 0. 23 > 0. 60 ± 0. 25

*p< 0.05 ** p < 0,02 Figure indicates Mean ± S,D,

with CT scanning, Abnormal findings of CT scan of the brain

assessed by visual inspection were significantly reduced in both the
ventricular system and cortical area four years later.

Table 2. indicates a comparison of the mean VBR in

schizophrenics between the index and follow up points. Almost all
VBR figures decreased, and a significant decline was found in the
third ventricle.

It is to be noted with interest that schizophrenics with a decrease

of abnormal brain CT findings in the third ventricle at the four-
year follow-up showed improvement of positive symptoms, particularly
hallucination and delusion, as shown in Table 3.

Table 3. Comparison of BPRS Item Score in Schizophrenics in

Whom CT-VBR Decreased during Follow-up Period

VBR (III) (n=46) 1978 1982

BPRS 46.6 ± 14.1 * 38.7 ± 12.2

Emotional withdrawal 3.3 ± 1.6 N.S. 2.9 ± 1.9
Conceptual disorganization 2.8 ± 1.4 N.S. 2.4 ± 1.6
Mannerism 2.6 ± 1.6 N.S. 2.9 ± 2.1
Hostility 2.3 ± 1.3 * 1.6 ± l.C
Suspiciousness 3.5 ± 1.8 *** 2.2 ± 1.6
Hallucinatory behavior 3.4 ± 1.8 *** 1.9 ± 1.5
Motor retardation 3.0 ± 1.6 N.S. 2.9 ± 1.7
Uncooperativeness 2.5 ± 1.7 N.S. 1.8 ± 1.3
Unusual thought content 3.7 ± 1.7 *** 2.4 ± 1.4
Blunted affect 3.7 ± 1.3 N.S. 3.7 ± 1.7

* p< 0.05 *** p< 0.01

54
 The same findings were observed in the case of the anterior
horn of the lateral ventricle. In other words, negative symptoms of
schizophrenia that were associated with central and cortical atrophy
of the brain did not correlate with the reversible change of brain
CT findings.

The hallucinatory paranoid symptoms are reversible and seem to

be associated with the reversible portion of the brain CT findings.
To further exaxmine the CT findings of paranoid schizophrenics as
compared with nonparanoid schizophrenics, 40 cases among the first-
contact schizophrenics were examined by CT scanning. The CT
findings were calculated as area ratio, compared with psychiatric
symptoms, clustered and scored in 38 syndromes from 140 PSE items by
an MRC 9SYN program. As a result, many significant correlations
between CT findings and syndrome check list items were found. As
shown in Fig. 3, when paranoid schizophrenics were compared with
nonparanoid schizophrenics, frontal cortex atrophy and enlargements
of the anterior interhemispheric fissure as well as the left Sylvian
fissure were significantly less pronounced in the paranoid subjects.

In this sense, it is interesting to note that the frontal

cortex atrophy was negatively correlated with positive symptoms and
the enlargement of the interhemispheric fissure was positively
correlated with negative symptoms. Frontal atrophy was also
negatively correlated with excitation. These findings coincide with
the well-known notion that cortical and central atrophy of the brain
of schizophrenics is correlated with negative symptoms and that the
paranoid schizophrenic preserves personality better than the
nonparanoid type.

Comparison of Mean Value of CT Area Ratio

r- -,

r----., n. s.
r-,

• p <0,05
100

Fig.3 80

0.5

n rj 11
I ~./a II
Frontal
i u I
VBR (Ill l
cortex ratio

~ normal controls •paranoid schizophrenics ~nonparanoid schizophrenics

55
 In conclusion, as compared with the nonparanoid type, paranoid
schizophrenia could be a relatively homogeneous and distinct
condition in terms of low platelet MAO activity and high PEA
excretion as well as less pronounced CT abnormality. In addition,
the paranoid type favors the dopamine overactivity hypothesis of
good prognosis schizophrenia or Crow's Type 1 syndrome?

Reference

1. A. Carlsson, B. E. Roos, J. Walinder, and A. Scott, Further

studies on the mechanism of antipsychotic action., Potentiation
by alpha-methyl-tyrosine of thioridazine effects in chronic
schizophrenia, J. Neuronal Transmission., 34: 125 (1973).
2. R. Takahashi, S. Yoshimoto, and H. Matsumoto. , Implication of
plasma level of neuroleptics in the treatment and
pathophysiological research of schizophrenia, in:"Cooperation
Franco-Japonaise, . Colloque Psychopharmacologie," P. Pichot, T.
Kobayashi eds., Paris (1979).
3. R. J. Wyatt, S. G. Potkin, T. P. Bridge, B. H. Phelps, and C. D.
Weise, Monoamine oxidase in schizophrenia, And overview,
Schizophrenia Bulletin. 6: 199 (1980).
4. S. G. Potkin, H. E. Cannon, D. L·. Murphy, and R. J. Wyatt, Are
paranoid schizophrenics biologically different from other
schizophrenics? The New Eng. J. Med. 298: 61 (1978).
5. S·. G. Potkin, K. Farouk, L. Chuang, H. E. Cannon-Spoor, L.
Phillips, and R. J. Wyatt, Phenylethylamine in paranoid chronic
schizophrenia. Science 206: 470 (1979).
6. R. Takahashi, Y. Inaba, K. Inanaga, N. Kato, H. Kumashiro, T.
Nishimura, T. Okuma, S. Otsuki, T. Sakai, T. Sato and Y.
Shimazono, CT scanning and the investigation of schizophrenia,
in:"Biological Psychiatry 198," C. Perris, G. Struve and B.
Jansson, eds., Elsevier, Amsterdam (1981).
7; T·. J. Crow, Molecular pathology of schizophrenia; more than one
disease process, Br. Med. J. , 280: 66 ( 1980 }·;

56
PEPTIDES IN MAJOR PSYCHOSES

Ole J. Rafaelsen

Psychochemistry Institute & Department of Psychiatry

Rigshospitalet
9, Blegdamsvej
DK - 2100 Copenhagen, Denmark

During the last decades, biological interest in depression

has focused on neuroendocrinological disturbances. Accordingly
the involvement of minor peptide hormones present in the brain
have been hypothesized and studied. Thus an impaired thyrotropin
(TSH) response to thyrotropin releasing hormone (TRH) has been
described (Prange et al., 1972; Kastin et al., 1972), and furthermore
Kirkegaard et al. (1975) and Kirkegaard and Smith (1978) showed that
changes in TSH response to TRH during electroconvulsive therapy (ECT)
seem to be of predictive value for the post-treatment course of the
disease. Moreover, Kirkegaard et al. (1979}, reported high levels of
TRH in cerebrospinal fluid (CSF) in endogenous depression.

Growth hormone (GH) response to various stimuli is reduced in

depression. This has been found in insulin-induced hypoglycemia
(Sachar et al., 1971), after administration of levodopa (Sachar et
al., 1975), after clonidin (Matussek et al., 1980; Checkley et al.
1981), after methylamphetamine (Checkley and Crammer, 1977), and
after the antidepressant drug, desipramine (Laakman, 1980).

Also vasopressin has been hypotesized to play a role in affective

disorders (Gold and Goodwin, 1978) and some studies seem to indicate
an influence of vasopressin on behaviour and memory (de Wied, 1972;
Weingartner et al., 1981; Koob and Bloom, 1982).

Recently, it has been discovered that peptides originally

classified as 'gut-peptides' are present in neurones in the brain
(Hokfelt et al., 1980), and several lines of evidence seem to indicate
that these 'gut-peptides' (Vasoactive Intestinal Polypeptide (VIP),
cholecystokinin (CCK) , and gastrin) have a function as neurotrans-
mitters or modulators (Fahrenkrug, 1979; Rehfeld, 1978).

57
 In the Department of Psychiatry and the Psychochemistry Institute,
Rigshospitalet, Copenhagen,we have studied the concentrations of TRH,
Vasopressin, VIP, CCK, and gastrin in CSF in patients with endogenous
depression, non-endogenous depression, mania, schizophrenia, and in
neurological controls. Concerning TRH no differences were observed
between the examined groups, and thus the findings of Kirkegaard et al.
(1979) was not confirmed. The median age of the depressed patients and
the control group was higher in the study by Kirkegaard, but as no
correlation to age could be demonstrated in either stud~ the age
difference cannot explain the differing results.

The hypothesis that vasopressin function is diminished in de-

pression and augmented in mania (Gold and Goodwin, 1978) was only
partly confirmed as no differences in CSF could be seen in the manic
group compared to controls, whereas in the depressed groups, compris-
ing a group of endogenous and a group of non-endogenous depressions
according to ICD-8, a slight. but significant decrease in CSF vaso-
pressin was demonstrated.

In a small number of patients CSF vasopressin was measured before

and after lithium treatment. No influence of lithium on CSF vasopressin
was demonstrated.

Concerning the 'gut-peptides' the only positive finding was

with VIP. According to Hokfelt et al. (198Gb) CCK coexists with
dopamine in neurones and has an inhibitory effect on dopamine release.
On this background it has been discussed whether the function of CCK
is impaired - or some kind of inbalance with dopamine exists - in
schizophrenic patients. A positive effect of CCK injections in
schizophrenic patients has been reported (Itoh et al., 1982; Maroji
et al., 1982), but from our study we must conclude, in agreement with
Gerner and Yamada (1982) , that the possible impaired CCK function is
at least not reflected in CSF-CCK, as CSF-CCK levels were equal in
all groups investigated.

The scatter of VIP-levels in depressive patients was quite

large, but when applying the ICD-8 and Newcastle rating scale (1965)
to distinguish between endogenous and non-endogenous depression, it
turned out to our surprise that the non-endogenous, but not
endogenous, depressions had a 50 per cent reduction in CSF-VIP.
Furthermore, there was no change of these low values when the patients
had recovered from their depression (Gjerris et al., 1981). We do not
know much about the function of VIP in the CNS (Johansson et al.,
1982), but it is noteworthy that VIP is present in CSF in concentra-
tions 5 to 10 times as high as in blood (Fahrenkrug, 1979).

Concerning the co-existence of classical neurotransmitters and

peptides in neurones, it has been reported that VIP is found in
acetylcholine containing neurones (Hokfelt et al., 1980a). As
acetylcholine function in some depressed patients seems to be

58
impaired (Sitaram et al., 1981) a possible correlation between
CSF-VIP and acetylcholine function should be investigated. The
phase-independency of CSF-VIP may indicate that we are dealing
with a trait marker allowing to diagnose some individuals as
having a reaction potential, a mode for depression: in addition
this marker seems to distinguish between endogenously and non-
endogenously depressed patients, and if this is true, it will
represent a new biological differentiation that will promote
research, diagnosis, and treatment in affective disorders (Rafaelsen,
1980) .

REFERENCES

Checkley, S.A., and Crammer, J.L., 1977, Hormone responses to

methylamphetamine in depression: A new approach to the
noradrenaline depletion hypothesis, Br J Psychiatry, 131:582.
Checkley, S.A., Slade, A.P., and Shur, E., 1981, Growth hormone
and other responses to clonidine in patients with endogenous
depression, Br J Psychiatry, 138:51.
De Wied, D., 1977, Peptides and behavior, Life Sci_, 20:195.
Fahrenkrug, J., 1979, Vasoactive intestinal polypeptide: measurement,
distribution and putative neurotransmitter function, Diqestio~,
19:149.
Gerner, R.H., and Yamada, T., 1982, Altered neuropeptide concentrations
in cerebrospinal fluid of psychiatric patients, Brain Res,
238:298.
Gjerris, A., Fahrenkrug, J., B¢jholm, S., and Rafaelsen, O.J., 1981,
Vasoactive Intestinal Polypeptide (VIP) in cerebrospinal fluid
in psychiatric disorders, in: "Biological Psychiatry", C. Perris,
G. Struwe, and B. Jansson, eds., Elsevier/Berth-Holland
Biomedical Press, Amsterdam , 565.
Gjerris, A., Rafaelsen, O.J., Vendsborg, P., and Rehfeld, J.,l983,
Gastrointestinal peptides in cerebrospinal fluid in psychiatric
disorders, .J Affective Dis, to be published.
Gold, p_w. and Goodwin, F.K., 1978, Vasopressin in affective illness,
Lancet, i:l233.
Hokfelt, T., Johansson, 0., Ljungdahl, A., Lundberg, J.M., and
Schultzberg, M., l980a, Peptidergic neurones, Natur~, 284:515.
Hokfelt, T., Rehfeld, J.F., Skirboll, L., Ivemark, B., Goldstein, M.,
and Markey, K., 1980b, Evidence for coexistence of dopamine and
CCK in meso-limbic neurones, Nature, 285:476.
Itch, H., Tanoue, S., Yagi, G., Tateyama, M., Kamisada, M., Fujii, Y.,
Takamiya, M., and Nakajima, S., 1982, Clinical study on the
psychotropic effects of caerulein - an open clinical trial in
chronic schizophrenic patients, Keio J Med, 31:71.
Johansson, B.B., Fahrenkrug, J., Wikkels¢, C., Andersen, 0., and
Blomstrand, C., 1982, Vasoactive intestinal polypeptide in
human cerebrospinal fluid, Front Horm Res, 9:189.

59
Kastin, A.J., Ehrensing, R.H., Schalch, D.S., and Anderson, M.S.,
1972, Improvement in mental depression with decreased thyro-
tropin response after administration of thyrotropinreleasing
hormone, Lancet, ii:740.
Kirkegaard, c., N¢rlem, N., Lauridsen, U.B., and Bj¢rum, N., 1975,
Prognostic value of thyrotropinreleasing hormone stimulation
test in endogenous depression, Acta Psvchiatr Scand, 52:170.
Kirkegaard, c. and Smith, E., 1978, Continuation therapy in endo-
genous depression controlled by changes in the TRH stimulation
test, Psychol Med, 8:501.
Kirkegaard,-c., Faber,· J., Hummer, L., and Rogowski, P., 1979,
Increased levels of TRH in cerebrospinal fluid from patients
with endogenous depression, Psychoneuroendocrinology, 4:227.
Koob, G.E. and Bloom, F.E., 1982, Behavioral effects of neuropeptides:
Endorphins and vasopressin, Annu Rev Phvsiol, 44:571.
Laakmann, G., 1980, Effect of antidepressants on the secretion of
pituitary hormones in healthy subjects, neurotic depressive
patients and endogenous depressive patients, Nervenarz~, 51:725.
Maroji, T., Watanabee, N., Aoki, N., and Itoh, S., 1982, Antipsychotic
effects of ceruletide (caerulein) on chronic schizophrenia,
Arch Gen Psychiatrv, 39:485.
Matussek, N., Ackenheil, M., Hippius, H., Muller, F., Schroder, H.-Th.,
Schultes, H., and Wasilewski, B., 1980, Effect of clonidine on
growth hormone release in psychiatric patients and controls.
Psvchiatrv Res., 2:25.
Prange, Jr., A.J., Lara, P.P., Wilson, I.C., Alltop, L.B., and
Breese, G.R., 1972, Effects of thyrotropinreleasing hormone
in depression, Lancet, ii:999.
Rafaelsen, O.J., 1980, Biology of manic-melancholic disorders,
Med J Aust, 1:637.
Rehfeld, J.F., 1978, Immunochemical studies on cholecystokinin. II.
Distribution and molecular heterogeneity in the central nervous
system and small intestine, J Biol Chem, 253:4022.
Sachar EJ, Altman, N./ Gruen, P.H., Glassman, A., Halpern, F.S.,
and Sassin, J., 1975, Human growth hormone response to
levodopa, Arch Gen Psvchiatrv, 32:502.
Sachar, E.J., Finkelstein, J., and Hellmann, L., 1971, Growth
hormone responses in depressive illness. I. Response to
insulin tolerance test, Arch Gen Psychiatry, 25:263.
Weingartner, H., Gold, P., Ballenger, J.C., Smallberg, S.A.,
Summers, R., Rubinow, D.R., Post, R.M., and Goodwin, F.K.,
1981, Effects of vasopressin on human memory functions,
Science, 211:601.

60
RECENT TRENDS IN BIOLOGICAL DEPRESSION RESEARCH

Herman M. van Praag

Department of Psychiatry
Albert Einstein College of Medicine/Montefiore Medical
Center, Bronx, NY 10461, U.S.A.

What happened in biological depression research since the last

World Congress of Psychiatry. Much, too much to cover in any compre-
hensive way in 15 minutes. Therefore, I confine myself to sketching
broadly a few trends. Without detailing; without any claim of
comprehensiveness.

First trend. There has been a shift in emphasis in the study

of the mode of action of antidepressants from interferences with
monamine (MA) processing to interferences with MA receptors (both
pre- and post- synaptic).

Until the mid-seventies, the focus in antidepressant research

was on MA-synthesis, -degradation, -turnover, -uptake and -release.
Then, thanks to the development of techniques to study isolated MA
receptors, interest shifted to the effects of antidepressants on
the effector-structures, the receptors.

The effects of antidepressants on MA metabolism and processing

are rather clear cut. Via different mechanisms - such as inhibition
of degradation or reuptake - they increase MA availability in the
synaptic cleft. The effects of antidepressants on receptors is
less clear cut and too complicated to be summarized in a few sen-
tences. Suffice it to say that certain receptor systems are up-
regulated, i.e., increase in functional activity, such as certain
(not all) serotonin (5-hydroxytryptamine; 5-HT) receptors and certain
a-adrenergic receptors. Others decrease in functional activity
(down regulation) such as S-adrenergic receptors.

What the net effect is of antidepressants on the functional

activity of MA ergic systems in the brain is much less clear than

61
it seemed in the seventies, when simply "increase" seemed to be the
appropriate answer.

Second trend. This is the attempt to measure MA receptor

activity in living men.

The effect of antidepressants on central MA led to the study of

MA metabolism in depression. When the way MA are processed was the
main focus of antidepressant research, affective disorder research
followed by studying metabolism-related variables, such as: MA
metabolite concentration in CSF, blood and urine; activity measure-
ment of enzymes involved in MA metabolism; 5-HT uptake in blood
platelets.

Preclinical receptor studies in animals, led to the search for

clinical counterparts. I mention 2 strategies used to achieve this
goal.

Direct strateqy. Search for centrally occuring receptors on

peripheral structures. I mention one example, the so-called
imipramine receptor, probably the uptake site of 5-HT or closely
related to that site. This receptor is located in presynaptic
nerve terminals of 5-HT ergic neurons but is equally found and
measurable on blood platelets.

Indirect strat~~· A substance whose release is regulated by

certain central receptor systems is peripherally measured. Example:
the pineal hormone melantonin released via S-adrenergic mechanisms
and measurable in plasma. A related strategy is: to challenge a
receptor system with either an agonist or an antagonist and measure
a product whose release is regulated. Example: measurement of
growth hormone in plasma after clonidine challenge.

Third trend. Increasing evidence that biological markers of

depression really exist, that there are biological variables that
can contribute to the firmness with which a diagnosis of depression
can be established.

Obvious examples are: deficient suppression of cortisol release

by dexamethasone, low CSF 5-hydroxyindoleacetic acid (5-HIAA) and
decreased REM latency. Unknown is to what degree these variables
interrelate, but there is more than circumstantial evidence that
they have indicative value for the diagnosis of certain types of
depression or for the occurence of certain symptoms within the
context of a depressive syndrome.

The predicitive validity of disturbed DST and increased REM

latency is yet known. There is evidence that low CSF 5-HIAA does
contain predictive information. First, with regard to the choice of
an antidepressant. The group with low CSF 5-HIAA seems to respond

62
best to 5-HT potentiating compounds. Second, w1th regard to
prognosis. The low 5-HIAA subgroup seems to have an increased
risk of relapse, as well as suicide.

The past 5 years have demonstrated that with regard to the

classification of depression the biological vantage point has
something important to contribute.

Fourth trend. What I would like to mention here is less a

trend than an actual accomplishment, i.e., the development of new
treatment modalities. Actually these developments can be divided
into 3 categories.

1. Drugs with a selective influence on one of the central MA.

Examples are the selective 5-HT reuptake inhibitors, recently intro-
duced. Selective noradrenaline (NA) reuptake inhibitors are on their
way. The MA precursors likewise belong to this category. There is
increasing evidence that both 5-HT precursors, tryptophan and 5-
hydroxytryptophan are effective. The first experiments with the
NA precursor tyrosine are encouraging.

2. Old drugs in new combinations. MAOI and tricyclics, once

considered to be incompatable have shown their value in therapy-
resistent depression. The latter statement is also true for tri-
cyclics and lithium. This combination too, can be effective in
patients in which other drugs have failed.

3. Old drugs for new indications. A perfect example is

carbamazepine used for years in temporal lobe epilepsy and known to
have psychotropic potential. Recently, it has been shown that it has
prophylactic potential in manic-depressive disorder and that it seems
particularly indicated in cases in which lithium has failed.

Fifth trend. Finally, I want to touch upon a fifth trend that

became more and mqre prominent in the past 5 years or so. A trend
which I define as a retreat from the specificity-concept or better
from the nosological concept.

Until recently the basic philosophy (or better; strategy) in

biological psychiatry has been to find biological variables indica-
tive of a psychiatric disease (e.g., schizophrenia) or at least of
a certain syndrome (e.g., vital depression). This, of course, is
a very classical attitude for psychiatrists and really for physicians
in general. Our thinking is very much imbued with the idea that
disease entities are the \ery building stones of classification.
However, biological psychiatry brought forward data in.compatable
with this philosophy. I mention some suggestive data:

- Low CSF 5-HIAA long considered as a likely marker of certain

types of affective disorders, was demonstrated to occur likewise in

63
suicidal and hostile individuals without affective disorders.
Tentative conclusion: low CSF 5-HIAA is not so much characteristic
for a certain depressive syndrome as it is for disturbances in
aggression regulation. These disturbances may occur in depressive
syndromes but are by no means specific to them.

- Low CSF homovanillic acid (HVA) is a typical finding in

Parkinson's disease and a sign of decreased dopamine (DA) metabolism.
Similar observations however, have been made in vital depression,
in particular if motor retardation is prominent.

Conclusion: Low HVA is not nosologically (Parkinson's Disease)

or syndromally (retarded vital depression)specific but linked to a
particular motor dysfunction: motor retardation, irrespective of
the disease entity in which it occurs.

Recent findings in psychopharmacology could be considered as

hints in a similar direction.

Tricyclic antidepressants are considered as rather specific

therapeutics for vital depression. Suggestive evidence indicates
that at least one member of this group - imipramine - is effective
in panic disorder. Another member of the group - clomipramine -
has been found effective in obsessive-compulsive disorder without
evident depression, although this Swedish finding still awaits
confirmation.

These findings indicate that in clinical psychopharmacology as

well one should have an open mind for deviations from the adage of
nosological or at least syndromal specificity of psychotropic drug
action.

Almost 10 years ago I formulated a plea for what I called a

functional psychopathology, i.e. dissection of psychiatric syndromes
into constituent psychological dysfunctions. I suggested that
relating biological variables to disturbed psychological functions
would be a better strategy, than to search for "the" biological
cause of a particular syndrome or nosological entity.

The developments over the past few years suggest the usefulness
of this conception.

I have discussed the main trends in biological depression

research. By and large they have been productive trends that have
led to a better understanding of the pathophysiology of depression
and towards more effective treatment. Moreover, the omens seem
favourable: I see in this domain no signs of stagnation.

64
REFERENCES

Beumont, P.J.V.; Burrows, J.D. (Eds.). Handbook of psychiatry and

endocrinology. Elsevier/North Holland Biomedical Press, Amsterdam,
1982.

Freedman, E.; Mann, J.; Gershon, J. (Eds.) Depression and anti-

depressants. Raven Press, New York, 1983.

Ho, B.T.; Schoolar, J.C.; Usdin, E. (Eds.). Serotonin in biological

psychiatry. Raven Press, New York, 1982.

Paykel, E.J. (Ed.) Handbook of affective disorders. Churchill

and Livingston, Edinburgh, London, Melbourne, New York, 1982.

Post, R.M.; Ballenger, J.C. (Eds.). The neurobiology of manic

depressive illness. Williams and Wilkins, Baltimore, 1983.

van Praag, H.M. Management of depression with serotonin precursors.

Biol.Psychiatry. 16:291-310, 1981.

van Praag, H.M. Neurotransmitters and CNS disease: depression.

Lancet. !:1259-1264, 1982.

van Praag, H.M.; Kerf, J.; Lakke, J.P.W.F.; Schut, T. Dopamine

metabolism in depressions, psychoses and parkinson's disease: the
problem of the specificity of biological variables in behavior
disorder. Psychol.Med. 5:138-146, 1975.

van Praag, H.M.; Mendlewicz, J. (Eds.). Management of depression

with monamine precursors. Karger, Basel, 1983.

65
GENETICS OF AFFECTIVE PSYCHOSES -

SOME METHODOLOGICAL CONSIDERATIONS

Jules Angst

Head of Research Department

Psychiatric University Hospital
8029 Zurich (Switzerland)

In the op~n~on of a number of experts, clinical-genetic research

has come to a dead end. Further studies with traditional methods on
adoptees, twins or entire families and kins are not likely to bring
about any important new findings. The application of modified mathe-
matical models has led to some controversial conclusions, and the
mode of inheritance of affective disorders is still unknown, even
if at present a polygenetic inheritance seems to be the most prob-
able. Important genetic researchers have turned away from clinical
research because of this critical balance, to look for a breakthrough
in the molecular biological area. Another new trend, observed in
recent years, consists in an ever increasing interest in the classi-
fication of psychiatric disorders. As validators for diagnostic
criteria, data on genetics, course and response to treatment are
used, in addition to biological variables. Within this context,
clinical genetics will continue to be relevant. For this reason,
I wish to focus my speech on associations between diagnostics and
genetics, and will deal with three general problems from a genetic
point of view: (1) Diagnostic concepts: homotypology, validation,
(2) case definition: continuum from normal to pathological, and
(3) sex differences in prevalence: artefact, or true?

DIAGNOSTIC CONCEPTS

Clinical genetic research is based on diagnostic concepts. If

the same disorder runs within families, this is termed 'breeding
true' or 'homotypology'. The intrafamilial similarity of psychiatric

67
disorders is also used as a tool for the validation of a diagnosis.
One of the best confirmations of a diagnostic concept occurs
when a subclassification of probands can be corroborated by a 'breed-
ing true' among relatives. Today, this is a frequently applied stra-
tegy of recognized value. The point I wish to make here is that this
must be done in an appropriate way. The history of genetic research
shows that many studies have been based on insufficiently discrimi-
nating, overinclusive diagnostic concepts, without a design able to
disprove them. Examples of overinclusive and probably heterogeneous
diagnoses are those of schizophrenia and manic-depressive psychoses,
created by Kraepelin, and of cycloid psychoses, created by Kleist
and Leonhard. I would like to show that the use of such unitarian
global concepts in genetic research is highly questionable and not
likely to foster advance. The history of genetic research in the
field of affective disorders shows clearly that substantial progress
can be achieved only if we try to break down current diagnostic ca-
tegories into subclasses, and to classify both probands and relatives
in the same way. Table 1 illustrates some of the points to be made.
Using an overinclusive concept of psychological disorders, Jung
found a 'breeding true' within families (1864). The same turned out
to be true for the concept of manic-depressive psychoses, based on
Kraepelin's definition and studied among relatives for the first
time by Hoffmann in 1921. Neither of these authors subclassified
psychological disorders or manic-depressive psychoses any further,
and they could therefore not disprove the underlying diagnostic
concept. In such a way, we can only confirm what we are looking for.

A great step forward in classification was made by Kleist, by

reintroducing the original distinction between mania, melancholia,
and bipolar psychosis. Unfortunately, the first family studies corning
from his school did not classify the relatives in an equal way, but
used again an overinclusive concept of endogenous psychosis. There-
fore, Neele (in 1949), later on Leonhard (1957) and Leonhard, Korff
and Schulz (1962) could not trace a true 'breeding true', but only
differences in overall morbidity for endogenous psychoses between
the various groups. Fortunately, Leonhard,Korff and Schulz also
distinguished personality disorders, and could thus provide some
evidence to justify a subclassification of manic-depressive psychoses.
As soon as the index probands and their relatives had both been
subclassified into different categories of depression, mania or bi-
polar psychoses, substantial evidence for the heterogeneity of manic-
depressive psychoses was found in 1966 and 1967 (Angst; Perris; Wino-
kur and Clayton). Future classificational research can only be ex-
pected to be fruitful if diagnostic groups are repeatedly subclassi-
fied, trying to disprove the homogeneity of current diagnostic con-
cepts. Progress in this direction was made by Dunner et al., and Ger-
shon et al., in 1980, by further subtyping the bipolar psychoses.

68
 Table 1. Classification of affective disorders

I Index-Cases, Probands I IRelatives I Authors

Psychol.Disorders Psychol.Disorders Jung 1864

Depression Depression Sioli 1885

Manic-Depr.Psychoses Manic-Depr.Psychoses Hoffmann 1921

Mania } Endogenous Psychoses

Depression,Melancholia Cyclothymic Persona- Neele 1949
Bipolar Psychoses lity Disorders

Depression,Melancholia }Endogenous Psychoses Leonhard 1957

Bipolar Psychoses Cyclothymic Persona- Leonhard,
lity Disorders Korff,Schulz 1962

Depression Depression, Mania Angst 1966

Bipolar Psychoses Bipolar Psychoses Perris 1966
Winokur+Clayton 1967

Bipolar I, and II Bipolar I, and II Dunner et al. 1980

Gershon et al. 1980

More recently, the diagnostic group of the schizoaffective or

cycloid psychoses has attracted full attention. A standard work in
that field is the monograph of Perris on cycloid psychoses. In his
genetic study, he classified probands and relatives using that global
diagnostic concept. This approach is questionable for the following
reasons: The concept of cycloid psychoses, evolved by Kleist and
further developed by Leonhard, includes at least three distinct
diagnostic classes: motility psychosis, confusion psychosis, and
anxiety-elation psychosis. The application of a generic term like
'cycloid psychosis' in family studies is useful only if it has pre-
viously been demonstrated that the three distinct subclasses do not
differ from a genetic point of view. A study on cycloid psychoses
should therefore be designed in such a way as to start with the
classification of the index probands or affected relatives into the
three subgroups.

With these examples, I merely wish to illustrate and stress

again how dangerous it is to use unitarian global diagnostic con-
cepts as a basis for genetic investigations. Their use is justified
only if it has previously been demonstrated that possible subgroups
do not differ genetically. Otherwise, we presume a homogeneity that

69
may just as well be absent. For future research, it could be produc-
tive not only to apply current diagnostic concepts, but to develop
new ones, based on new categories of syndromes, as can for instance
be generated by the application of multidimensional scaling and
cluster analyses of symptoms (Angst et al.,l983), or by the inclusion
of biological (for example neuroendocrine, immunological or electro-
physiological) criteria.

CASE DEFINITION

We know that case finding is not only dependent on the diagnos-

tic concept but, in the first instance, on a threshold value defined
by certain specific criteria. The validity of threshold values
applied in the genetics of affective or schizoaffective disorders has
not yet been examined. Apart from a few exceptions, research basi-
cally focuses on hospitalized probands, and on diagnosing their
affected relatives, so-called 'secondary cases'. But a recent study,
carried out by Weissman et al.(l982), shows that among the relatives
of outpatient depressives, an equally high morbidity risk prevailed
as among those of the hospitalized group. This is a finding of ge-
neral interest. Along with the results of epidemiological studies of
the general population, we have to presume the existence of a con-
tinuum of depressive mood changes from healthy to mildly depressed
and severely psychotic depressive disorders (Angst et al.l983).

Consistent with the hypothesis of a continuum is the fact that

most symptoms of normal depressive mood swings, minor and major
depression, do not differ in quality but in quantity only. In addi-
tion, life events do not seem to play a dominant role in minor de-
pression. Our results support the hypothesis that every human being
has an individual biological disposition for spontaneous depressive
mood changes which can be precipitated, deteriorated or prolonged
by life events.

If such a continuum from normal to pathological depression

existed, the definition of a threshold for the diagnosis of a case
would represent a highly difficult problem, and developing a quanti-
tative measure of the depressive continuum would be more adequate.
The application of such methods could provide genetic analysis with
completely new data, and lead to new developments. Traditional work
would not become worthless, however, because sophisticated mathema-
tical models, for instance as developed by Reich et al.(l979), are
applicable to soft qualitative psychiatric data too.

DIFFERENCES BETWEEN THE SEXES IN PREVALENCE

Traditional epidemiological studies as well as genetical family

70
studies of depressives commonly show a higher ~orbiditv risk at a
factor of 2 or 3 females compared to males. Several excellent review
articles on this fact exist (Weissman a.Klerman,l977; Clayton,l983),
as well as a rich discussion of sociological, psychological and bio-
logical hypotheses trying to explain these differences between females
and males. If we consider the hypothesis that the finding of an un-
equal morbidity of males and females for depression is a methodolo-
gical artefact, this may at first sight be frightening, but please
try and follow me nonetheless in some reflections. It is well known
that the excellent 'Amish Study', carried out be Egeland and Hostetter
(1983), based on a highly sophisticated methodology, revealed no
differences in male and female morbidity for unipolar depression.
In order to avoid a change in our views, our immediate natural re-
action to such unexpected findings is, of course, to look for an
explanation for the new data that would be compatible with the tra-
ditional concept of a true difference between the sexes. The surpris-
ing findings of the Amish Study may be due not only to cultural fac-
tors, but also to the case definition, which implied clear social
impairment. This assumption is supported to some extent by our own
field study (the 'Zurich Study') of a young general population of
males and females (table 2). If we define depression not based on
the presence of a certain number of symptoms, but on the presence
of social disablement, e.g. impairment in occupational functioning,
in consequence of a 2-week depressive episode, we do not find any
difference in point- or 3-month prevalence rates. We find a preva-
lence of 4.6% for males and 4.3% for females (MAJDEP 2+3). On the
other hand, we do obtain the well-known sex difference when applying
DSM-III criteria. In this way, we produce a ratio of 1:2 for 3-month
prevalences of male/female depression (table 2).

Table 3 lists some diagnostic criteria of potential influence

on a differing morbidity in males and females. There is some evidence
(Angst,l980) that the duration of depressive manifestations might
be the same in both sexes. The DSM-III requires a minimum duration
of two weeks for depression. In addition, a certain number of symp-
toms must be present. But it is well known that female depressive
patients report more symptoms than males, not only in self-ratings,
but also in interviews (Fig.l). An equal cut-off point for the
number of symptoms required for a diagnosis may therefore discrimi-
nate males. Thus, the application of the usual diagnostic criteria
may result in a preponderance of depressive females.

A second pitfall based on sex differences is the forgetting of

depressive manifestations. If we compare the prevalence figures over

I wish to express my thanks to Mrs. A. Dobler-Mikola for her coope-

ration in the epidemiological project (Zurich Study) mentioned above

71
 Table 2. Prevalence rates of depression (DSM-III and MAJDEP)
differences between 3 and 12 months

DSM-III MAJDEP (2+3)

m f ratio m f ratio

3 months 1.9 3.4 0.6 4.6 4.3 1.1

12 months 2.3 11.2 0.2 4.7 10.7 0.4

male morbidity increases less significantly with

length of observation period than female morbidity

Table 3. Factors favoring differences in morbidity

of males and females for depression

Sex difference
1 Definition of depression
Criteria: duration of episode ??
number of symptoms m< f
impairment at work m= f
2 Memory
Criteria: different time spans
1 month }
m = f
2- 3 months
4-12 months m< f

3 and 12 months, it can be seen that the less recently a depressive

episode occurred, the more males underreport depression. In this
way, the 12-month prevalence figures for males become underreported
and the sex difference returns. Among males there is no substantial
gain from 3- to 12-month prevalence figures. By contrast, females
gain by a factor of 2 to 3 (see table 2).

Both the Amish Study and our own findings favor the hypothesis
of an equal morbidity of males and females for depression. Should
this turn out to be true, epidemiologic and genetic research would
have to be revised to a certain extent. If we have to include minor
depression in genetic models, we shall perhaps meet obstacles which
will be extremely difficult to overcome. It is to be hoped that in-
telligent mathematicians will develop suitable new models which will
allow us to commit all the traditional errors and still reach valid
results. If that were to be the case, we could remain happy and

72
 % 0 20 40 60 80 100

Depression
Appetite
Sleep
Loss of energy
Retarded
Agitated
Loss of interest

Worthlessness
Thoughts of death
Loss of concentration

males -·-females
Fig. 1. Male and female symptom profiles (MAJDEP, sub-
jects with occupational impairment)

satisfied. But I am afraid I expect the growing interaction between

general epidemiology and genetics to lead us in the direction of
the hypothesis put forward above.

CONCLUSIONS

May I summarize once again some of the points this presentation

should have raised.

(1) The concept of homotypology of psychiatric disorders is basic

to genetic investigation.

(2) Homotypology is dependent on the diagnostic class~fication used.

(3) Systems of classification change because they are mainly

descriptive and not based on causal mechanisms.

(4) Symptoms of classifications currently used are poorly validated,

or the criteria used for validation are not very discriminating.

73
 Furthermore, by taking a short list of signs and symptoms as a
basis for a certain diagnosis, a serious loss of information has
to be taken into account. Newer methods for classification use
long lists of symptoms but are not suitable for retrospective
data, whereas most of the information collected for family
studies is retrospective.

(5) Mood changes and affective disorders may be differently per-

ceived by males and females or at least be differently reported.
For this reason, the application of identical diagnostic cri-
teria for males and females has to be seriously questioned.

(6) The binary distinction between 'healthy' and 'ill' is most cri-
tical in the field of affective disorders. Future research
should try to base itself on methods accounting for the conti-
nuum from normal to pathological and allowing the application
of quantitative genetic models for data analysis.

If some of these reflections turn out to be true, the carrying

out of further family studies in the traditional way may become
highly questionable. On the other hand, it would certainly also be
incorrect to accept a methodological nihilism, giving up clinical
research in genetics, and wait peacefully for a breakthrough in mo-
lecular biology. The history of interaction between diagnostic
classification and genetic research shows clearly that a fruitful
interrelationship is necessary for the further development of both.

REFERENCES

Angst, J., 1966, Zur Aetiologie und Nosologie endogener depressiver

Psychosen-Eine genetische,soziologische und klinische Studie.
Monogr.a.d.Ges.geb.d.Neurol.u.Psychiat.,vol.ll2, Springer,
Berlin.
English translation: The Etiology and Nosology of Endogenous
Depressive Psychoses. A Genetic,Sociological, and Clinical
Study. Foreign Psychiatry 1973; 2: 1-108.
Angst, J., 1980, Verlauf unipolar depressiver, bipolar manisch-
depressiver und schizoaffektiver Erkrankungen und Psychosen.
Ergebnisse einer prospektiven Studie, Fortschr.Neurol.Psych-
iat, 48: 3-30.
Angst, J., Scharfetter Ch., Stassen H.H., 1983, Classification of
schizoaffective patients by multidimensional scaling and
cluster analysis, Psychiat.clin. (in print).
Angst, J., Dobler-Mikola A., Binder J., 1983, Zur Diagnostik der
Depression - epidemiologische und genetische Aspekte, in:
"Der Depressive und seine Familie", W.Poldinger, ed.,
Huber, Bern (in print)

74
Clayton, P.J., 1983, Gender and Depression, in: "The Origins of De-
pression: Current Concepts and Approaches", J.Angst, ed.,
Springer, Berlin-Heidelberg (in print)
Dunner D.L., Go R.C.P., Fieve R.R., 1980, A Family Study of Patients
with Bipolar II Illness (Bipolar Depression with Hypomania),
in:Abstracts of Papers, Workshops, and Posters, Annual Meeting
of the American College of Neuropsychopharmacology,
December 16-18, 1980
Egeland, J.A., Hostetter, A.M., 1983, Amish Study, !:Affective Dis-
orders among the Amish, 1976-1980, Am J Psychiatry, 140:56-61
Gershon E.S., Goldin L.R., Jimerson D.C., Nurnberger J.I., Goodwin
F.K.Jr, Guroff J., 1980, Clinical,Biological, and Linkage
Data in Affective Disorders Pedigrees, in: Abstracts of
Papers,Workshops,and Posters, Annual Meeting of the American
College of Neuropsychopharmacology, December 16-18, 1980
Hoffmann, H., 1921, Die Nachkommenschaft bei endogenen Psychosen.
Genealogisch-charakterologische Untersuchungen, Springer,
Berlin
Jung, W., 1864, Untersuchungen tiber die Erblichkeit der Seelen-
storungen, All2.Z.Psvchiat., 21:534-653
Kleist, K., 1911, Die klinische Stellung der Motilitatspsychosen,
~eitschr.f.d.ges.Neur.u.Psvch., 3:914-917
Kleist, K., 1911, Die Streitfrage der akuten Paranoia (Ein Beitrag
zur Kritik des manisch-depressiven Irreseins), Zeitschr.f.d.
ges.Neur.u.Psvch., V/3
Kleist; K., 1921, Autochthone Degenerationspsychosen, Zeitschr.f.d.
ges.Neur.u.Psych., LXIX
Kleist, K., 1927, Diskussionsbemerkungen zum Vortrag Kolle. Ver-
sammlung der slidwestdeutschen Psychiater Freiburg 1926,
Zentralblatt f.d.ges.Neur.u.Psych •. , 45:830
Kleist, K., 1928, Ueber zykloide,paranoide und epileptoide Psychosen
und tiber die Frage der Degenerationspsychosen, Schweiz.Arch.
Neurol.Psychia~., 23:3-37
Kleist, K., 1953, Die Gliederung der neuropsychischen Erkrankungen,
• II
~:Jakob Klaes1• zum 70.Geburtstag.Festschr1
II 'f t, Psych.Univ.
Klinik Waldau-Bern, pp.526-538
Kraepelin, E., 1913, Ein Lehrbuch fur Studierende und Aerzte.
8.Aufl., vol.III, 2.Teil. Das manisch-depressive Irresein.
J.A.Barth, Leipzig, pp.ll83-1395
Leonhard, K., 1959, Aufteilung der endogenen Psychosen, Akademie-
Verlag, Berlin (2.Aufl.)
Leonhard K., Korff I., Schulz H., 1962, Die Temperamente in den
Familien der monopolaren und bipolaren phasischen Psychosen.
Psychiat.Neurol~, 143: 416-434
Neele, E., 1949, Die phasischen Psychosen nach ihrem Erscheinungs-
und Erbbild, J.A. Barth, Leipzig

75
Perris, C., 1966, A Study of Bipolar (Manic-Depressive) and Unipolar
Recurrent Depressive Psychoses, Acta Psychiat.scan~., Suppl.
194: 1-189
Perris, C., 1974, A Study of Cycloid Psychoses, Acta Psychiat.scand.,
Suppl.253
Reich T., Rice J., Cloninger C.R., WetteR., James J., 1979, The use
of multiple threshold and segregation analysis in analyzing
the phenotypic heterogeneity of multifactorial traits. Ann.
Hum.Genet., 42: 371-390
Weissman M.M., Kidd K.K. Prusoff B.A., 1982, Variability in Rates
of Affective Disorders in Relatives of Depressed and Normal
Probands. Arch Gen Psychiatry. 39:1397-1403
Weissman M.M., Klerman G.L., 1977, Sex Differences and the Epidemio-
logy of Depression, Arch Gen Psy~hiatry, 34:98-111
Winokur G., Clayton P., 1967, Family History ~tudies.I.Two Types of
Affective Disorders Separated According to Genetic and Cli-
nical Factors, in:"Recent Advances in Biological Psychiatry~
J. Wortis, ed., Plenum Press, New York, pp.35-50

76
TRAIT, STATE, HPA ACTIVITY IN DEPRESSION

George Winokirl, Bruce Pfohll, and Barry Sherman2

lnept. of Psychiatry and 2nept. of Internal Medicine

University of Iowa College of Medicine
Iowa City, Iowa 52242

Presently the only trait variables that we have in affective

disorder are the family history or the history of a mania. In
unipolar depression it is possible to have a family history of
depression or family history of alcoholism. Most patients are
sporadic because they would have no family history at all.

In a series of papers we have outlined three types of

unipolar depressive illnesses which are defined by their family
backgrounds (1). 1) Familial pure depression disease (FPDD) is a
depression seen in an individual who has a family history of
depression but nothing else. 2) Depression spectrum disease
(DSD) is an ordinary depression seen in a person who has a first
degree family history of alcoholism and possibly antisocial
personality, with or without depression. 3) Sporadic depressive
disease (SDD) is a depression seen in a person with a negative
family history for alcoholism, antisocial personality, depres-
sion, or mania. By definition, a family history is a trait.

The illnesses described above differ from each other in age

of onset, presence of a stormy life style and course of illness.
Studies have shown that these familial subtypes show different
frequencies of abnormal dexamethasone suppression tests. FPDD
patients are very likely to be nonsuppressors; DSD patients are
most likely to be suppressors; sporadic patients are in between
the previous two groups. Sporadic patients may be composed of
the above two groups in varying degrees of mixture. Alternately
some SDD patients may have a totally separate kind of depression.
Not all studies have been positive in separating out the familial
subtypes but most have (2,3,4). These findings add validity to
the concept of familial subtypes.

77
 A positive dexamethasone suppression test (DST) in depression
is a state variable. However, it is a test which is sometimes
abnormal during a depression but normal at other times even
though the patient is just as depressed (5). Thus, over several
depressive episodes the results may vary. A history of an
abnormal dexamethasone suppression test could be considered a
trait. Once positive at any point in an individual's life it
cannot become negative. Thus, in a sense a lifetime history of
an abnormal DST approximates a trait, i.e. an inherited charac-
teristic. This is true because it is related to another trait, a
specific family history of alcoholism or depression. As noted
above, a number of studies support the idea of abnormal suppres-
sor status in FPDD and normal suppressor status in DSD.

In this paper we will determine whether the dexamethasone

suppression test as a state variable (measured at time of index
admission) is as good a test to define subgroups as is history of
ever having had an abnormal DST as a trait variable.

Further, we will examine trait findings as defined by family

history in order to determine whether baseline cortisol secretion
and ACTH secretion show any abnormalities related to DST trait
abnormalities as defined above or whether any abnormalities are
simply related to current DST state. There is, in fact, some
question whether ACTH is related to cortisol secretion in depres-
sion (6,7). We will determine whether it is related to trait,
state or neither.

Methods

In this study a patient will be defined as having trait

abnormalities (i.e. trait positive) if any of the following are
present. Either a specific family history, (depression with
absence of alcoholism in a first-degree family) or a history of
bipolarity or an abnormal dexamethasone suppression test at
sometime during a person's life (lifetime history) was considered
a trait. Both familial pure depressive disease and bipolar
illness were included in this group because of the fact that in
both cases a high degree of nonsuppression has been reported in
the past.

Patients who met DSM-III criteria for depression were

admitted to the hospital antl the clinical research center. Of
these, there were four bipolars and six FPDD patients. In
addition, there were seven DSD patients. Evaluations of the
dexamethasone suppression test were done at time of entry into
the study (index admission) and also the records and history were
searched in order to determine the presence of dexamethasone
suppression tests which were done at other times besides that of
index admission.

78
 After admission to the study 25 patients (including the above
plus others) were evaluated over 48 hours with determinations of
plasma cortisol and ACTH accomplished every 20 minutes. Plasma
cortisol was measured by radio-immunoassay using a specific
cortisol in a serum obtained from Damon Diagnostics, Needham
Heights, MA. ACTH was measured by radio-immunoassay without
extraction using an antiserum to purified human ACTH obtained
from Immunonuclear Corporation, Stillwater, MN, and used at a
final dilution of 1:16,000 in a volume of 0.4 ml. The 125-I
labelled human ACTH was obtained from Damon Diagnostics, Needham
Heights, MA, and the ACTH standard from the National Pituitary
Agency.

Cortisol and ACTH were compared statistically in three groups

of patients, 1) nonsuppressors on the dexamethasone suppression
test at index (state+, trait+), 2) Bipolars and FPDD plus those
with a history of lifetime nonsuppression, all suppressors at
index (trait +, state -) and 3) DSD, SDD suppressors (trait -,
state-). Cortisol and ACTH plasma levels were compared between
the groups over three 8 hour time periods: 8 PM to 4 AM, 4 AM to
Noon, Noon to 8 PM. Mean serum levels were calculated for each
patient during each of the three time periods. These means were
then averaged and compared across three groups of patients using
Student T-tests.

Results

Of ten bipolar and FPDD patients (bipolar= 4, FPDD = 6), two

(10%) were abnormal suppressors at time of index admission. Of
seven DSD patients, one (14%) was an abnormal suppressor at time
of index admission. However, evaluation of tests done at other
times during the patient's life indicated that six (60%) of the
bipolars and FPDD patients had a lifetime prevalence of abnormal-
ity. No DSD patient was added with an abnormal suppressor status
to the one which has been noted before. Though the numbers are
small, there is a clear trend in favor of the fact that non-
suppressor status as a lifetime history figure separates the
bipolars and the familial pure depressives from the DSD patients.

If abnormalities in baseline ACTH were related only to DST

state, then only the nonsuppressors at index admission in this
study should be different from the suppressors. However, if ACTH
were related to trait, the bipolars, FPDD and those with a
lifetime history of nonsuppression should have a different ACTH
curve at time of entry into the study (even if they had a normal
suppressor status at index) when compared to those DSD and
sporadic patients who also had a normal suppressor status at
index.

79
 Figures 1 and 2 give the curves for the three groups. The
three groups are as follows: nonsuppressors independent of
family subtype (N=8). Bipolar and FPDD patients who were normal
suppressors at time of index admission plus those suppressors who
had shown nonsuppression in the past (N=8) and DSD and SDD
suppressors at time of index admission who had never shown
nonsuppression at other times (N=9). The second of these three
groups is thus state negative but trait positive according to the
definition given earlier. The third group is state negative and
trait negative. As regards the predexamethasone cortisols, a
comparison of the bipolar and FPDD suppressors, DSD and SDD
suppressors, and the nonsuppressors may be seen in Table 1. The
nonsuppressors significantly separate themselves from the two
other groups on the cortisol variable, p <001 when compared to
the trait +, state- group and p <01 when compared to the state -
trait - group. Looking at the ACTH over a period of 24 hours,
one may note that there are no significant differences between
any of the three groups for any of the sampling periods (see
Table 1).

c 15
0
R
T
I
s
0 I .'
L r:•:f· .. ···

u ,
. /i
9 :'I
: ....·.. \
/
/'<I ..,
d 5 ...\, ' ...... ·"·"'- _I ':;il
I ..
...... ____ .. .,- ___~.:.~,-·

-4 4 8 12 16

Predexamethasone cortisol over a period of 24 hours

Nonsuppressors, state + :_ __

Trait +, State -: ....... .

Trait _, State - : ---------

Figure 1

80
A
c
T
H

p
9
/
m
I

0~~~~~~~+-~-+~~+-~-+~~~-+~
-4 0 4 8 12 16
~ F'Ra1 MIDNIGHT

Predexamethasone over a period of 2Q hours

Nonsuppressors, State +:

Trait +, State -

Trait - State

Figure 2

81
00
1\J

TABLE 1
PREDEXAMETHASONE CORTISOL AND ACTH IN GROUPS OF DEPRESSIVES

II III
Bipolars + FPDD DSD + SDD Who
I + Lifetime Nonsuppresors Never Have Shown
Suppressors (Suppressors at Index) Nonsuppression

Mean (S.D.) Mean (S.D.) Mean (S.D.)

ACTH (pg/ml)
8PM-4AM 31.0 (16.8) 23.1 (9.3) 21.4 (16.0)
4 AM - 12 Noon 31.3 (14.5) 27.7 (9.3) 26.0 (14 .3)
Noon - 8 PM 29.3 (11.9) 23.3 (6.6) 24.3 (14.8)
Cortisol
8PM-4AM 7.9 ( 2.3) 4.2 (1.5) 4.8 ( 1.1)
4 AM - 12 Noon 14.1 ( 2.4) 9.1 (2.2) 9.6 (2.3)
12 Noon - 8 PM 11.5 ( 2.3) 6.2 (2.2) 7.7 (1. 9)

No significant differences across groups for ACTH.

Group I vs Group II significantly different at all time periods (p <001)
Group I vs Group III significantly different at all time periods {p <01)
Group II vs Group III no significant difference.
 These curves have been drawn for 24 hours prior to the
dexamethasone suppression test. What one can conclude is that
there is no evidence that being a bipolar or a FPDD or a lifetime
nonsuppressor patient is different from the DSD and SDD patients
in ACTH secretion, as long as both are nonsuppressors. The
suppressors clearly do separate as regards cortisol as might be
expected.

Because there are no significant differences in the predex

ACTH, that variable does not appear to have anything at all to
offer for a separation of state or trait. On perusal the non-
suppressors do look as if they have a higher ACTH than the other
two groups but this does not become significant at any poi.nt.

Discussion

A reasonable trait marker for one biologic subtype of

depression may well be a lifetime history of a positive dexa-
methasone suppression test. It seems clear that some patients
have abnormal tests when they are depressed; and at other times
when they are hospitalized and equally depressed, they have
normal tests (5). Thus, an individual test may not be as useful
as a lifetime history. Previous correlations between the DST and
family history suggests that a more inclusive definition of this
bi.ologic trait might include being e bipolar and/or a familial
pure depressive and/or having an abnormal DST test at some time
during one's lifetime. Such persons could be compared to
patients who are depression spectrum patients and/or sporadic
patients who have never had an abnormal suppressor test.

An evaluation of this definition of trait could be useful in

studying other tests of abnormalities in affective disorder. For
example, the blunting of thyroid stimulating hormone in response
to protirelin has been reported to be poorly correlated with DST
results among depressives. It may be that the protirelin test
would show better correlation with the depressive trait measure
defined above.

In this research, we looked at the predexamethasone cortisol

and ACTH and it was clear that abnormal cortisol went along with
an abnormal suppressor status but not with the trait markers such
as familial subtyping or being bipolar or having a lifetime
history of nonsuppression. Although the ACTH concentration was
higher in the suppressors, it was not significantly different
from compared to the two other nonsuppressor groups. As a
consequence the predex ACTH defines neither trait nor state in
this study.

83
Conclusions

The trait of dexamethasone nonsuppression as defined by

lifetime history of any abnormal response to this test may be
even more highly correlated with familial subtype of depression
than is suppressor status as defined by DST results at one point
in time.

In a 24-hour predexamethasone evaluation of cortisol,

abnormal suppressors showed significantly higher plasma cortisol
than those patients who had normal suppression. Among the latter
group, trait positive patients did not differ from trait negative
patients. This suggests that the baseline cortisol plasma level
is related to state rather than trait as defined by a positive
family history of depression or bipolarity or a history of
nonsuppression. Predexamethasone ACTH secretion was not found to
be significantly related to either state or trait in this study.

REFERENCES

1. Winokur, G•• 1982, The development and validity of familial

subtypes in primary unipolar depression. Pharmacoosvchia-
trica~ 15:142-146.
2. Schlesser, M., Winokur, G., Sherman, B., 1980,
Hypothalamic-pituitary-adrenal axis activity in depressive
illness. Arch Gen Psvchiat, 37:737-743.
3. Targum, S., Byrnes, S., Sullivan, A., 1982, Subtypes of
unipolar depression distinguished by the dexamethasone
suppression test. .T Affect Dis, 4:21-27.
4. Rudorfer, M., Hwu, H-G., Clayton, P., 1983, Dexamethasone
suppression test in primary depression: significance of
family history and psychosis. Rial Psvchiat, 17:41-48.
5. Coryell, W. , Schlesser, M. , 1983, Dexamethas-one suppression
test response in major depression: Stability across
hospitalizations. Psvchiat R~R, 8:179-189.
6. Fang, V.S., Tricou, B.J., Robertson, A., Meltzer, H.Y., 1981,
Plasma ACTH and cortisol levels in depressed patients:
Relation to dexamethasone suppression test. Life Sc~,
29:931-938.
7. Kalin, N.H., Weber, S.J., Shelton, S.E., 1982, Plasma ACTH
and cortisol concentrations before and after dexamethasone.
_?sychiat Res_, 7:87-92.

84
HEREDITY-ENVIRONMENT IN SCHIZOPHRENIA

F. Schulsinger 1 1 J. Parnas 1, H. Schulsinger 2,

Th. w. Teasdale , and s. A. Mednick 1
1) Psykologisk Institut, Dept. of Psychiatry
Kommunehospitalet, Copenhagen, Denmark
2} Institute of Clinical Psychology, University
of Copenhagen, Denmark

INTRODUCTION
Is there a real profit of learning about the gene-
tic liability for schizophrenia? The answer is, to some
extent, YES. The knowledge about the strength of the
liability, and of the mode of genetic transmission
justifies the search for the biological mechanisms being
active in this genetic transmission - be they metabolic
or neurophysiological. In addition, the knowledge of a
genetic liability makes it possible to compose a group
of subjects for research in which a higher outcome of
schizophrenia than the average one percent can be
expected - in other words: a high risk group.
Instead of reporting results from the wellknown,
and traditional studies of twins, discordant twins,
pedigrees, and adopted pedigrees, I shall - therefore -
present a brief, and almost sketchy, survey of some of
the more resent results from a prosoective study of
children at hioh risk for schizophrenia. These children
have severely schizophrenic mothers. The American
psychologist, Sarnoff A. Mednick and I began this study
in Copenhagen in 1962 when the children were between
10 and 20 years old, and not yet schizophrenic. The
study still goes on.
Fig. 1 shows the major assessment years in our
study. In addition we received information on the
subjects from various sources.
85
 High risk children n = 207
Low risk children n = 104
x age 15,1 years

1962 ~
1967 1972 1980
INITIAL 5 year 10 year subsample
ASSESSMENT follow-up diagnostic follow-up
follow-up
Fig. 1. Design of the project.

In the following we shall describe results from

four substudies on the same sample:

1) An analysis of pregnancy and birth complications

distributed over diagnostic outcome in 1972-74.

2) An analysis of institutional care during the

first five years of life - and the diagnostic outcome in
1972-74.

3) An analysis of premorbid behaviour until the

1962 assessment, and diagnostic outcome in 1972-74.

4) Data from a study carried out 1979-80 in order

to test a hypothesis on the nature of what is genetical-
ly transmitted, and what not.

OBSTETRICAL DATA

Detailed midwife reports were rated on a 0-4

weighted rating scale for specific complications. This
scale was developed with the assistance of professor
Frits Fuchs of the Department of Gynaecological Obste-
trics at Cornell University Medical School. The indivi-
dual scores could be utilized to create three different
global scores of pregnancy and birth complications.

1) Frequency score
2) Severity score
3) Total score

86
 The total score was derived by adding all the
individual weighted scores, and the total score correla-
ted highly with the other scores. Fig. 2 shows the total
scale scores for the four diagnostic groups: schizophre-
nia, borderline schizophrenia, other diagnoses, NO MENTAL
ILLNESS or well. The schizophrenic group had significant-
ly higher complication scores than the borderline group.
The mean values for the no mental illness group do not
differ significantly from either those of the schizo-
phrenics or those of the borderlines. In fact the mean
values for the no mental illness group lie close to the
midpoint between these two pathological groups. The
majority of the schizophrenics (67 %) experienced some or
other form of complication, which means that our results
do not come from very few schizophrenics suffering a
tremendous amount of severe complications. These results
which are reported in (1) will be discussed eventually.

PBC
total

Fig. 2. Pregnancy and birth complications in

high-r isk g roup .

87
BEHAVIORAL PRECURSORS OF SCHIZOPHRENIA SPECTRUM
During the first assessment of our subjects from
1962-64 we collected information on the: subjects' beha-
viour from infancy up to the time of assessment. We
interviewed their rearing parents or rearing agencies.
All the subjects were interviewed during the assessment,
and all the investigators involved completed an adjective
checklist on the subjects. Finally we got very detailed
information from the classmasters about the subjects'
achievements and behaviour in school. Fig. 3 demonstrates
the overall results from the study of premorbid behavioral
characteristics within the schizophrenia spectrum. Part of
the information is retrospective, as for example the early
childhood factors where we find that schizophrenia spec-
trum, i.e. schizophrenia and borderline schizophrenia,
exhibited passivity and poor attention according to the
parental descriptions. School behavior was retrospective
to some extent, but for the majority of the subjects the
information was actually present. We found, that schizo-
phrenia spectrum subjects in school were more isolated
and rejected by others, and they were more sensitive. In
addition, they were characterized by poor affect control,
they got more easily upset, and more liable to contain
their upsetness for a longer time than the comparison
groups. With regard to this item, the schizophrenics were
significantly worse than the borderlines which was the

Early childhood
passivity
poor concentration

School behavior
rejected by others
poor affect control*

Clinical assessment (15 years of age)

formal thought disorder
defective emotional rapport

* This item discriminated schizophrenics from borderline

schizophrenics.
Fig. 3. Behavioral precursors of schizophrenia spectrum.

88
only significant difference between these two parts of the
schizophrenia spectrum. The clinical assessment items
were non-retrospective, and they comprised cognitive as
well as emotional items as premorbid discriminators.

As a preliminary conclusion of the behavioral

precursor study which is described in detail in (2) we
can state that our results seem to point toward a basic
behavioral relationship between schizophrenia and border-
line schizophrenia. Both disorders were characterized by
premorbidly defective emotional rapport, and formally
disturbed cognition. Pre-schizophrenics, however, tended
to display more attention deficit, and exhibited poor
affective control. In their psychopathological picture
both disorders exhibit fundamental schizophrenic symptoms
but of course to a different degree. The distinction
between them, however, relies mainly on the intensity of
accessory psychotic symptoms like hallucinations and
delusions. This study of behavioral precursors substan-
tiate Eugen Bleuler•s concept of schizophrenia as a
disorder essentially characterized by formally disturbed
cognition and defective emotional rapport: 11 The disease
is characterized by a specific type of alteration of
thinking, feeling, and relation to the external world
which appears nowhere else in this particular fashion
.••. Hallucinations and delusions are partial phenomena
of the most varied diseases. Their presence is often
helpful in making the diagnosis of a psychosis, but not
in diagnosing the presence of schizophrenia ....

As the fundamental symptoms can be traced premorbid-

ly, we find that Bleuler•s concept of schizophrenia as
a development rather than a disease process hitting a
healthy person is substantiated also by our results.

INSTITUTIONALIZATION

The basic advantage of the prospective, high-risk

design is that it makes it possible to find out which
environmental factors might be correlated to the diffe-
rences in clinical outcome. Michael Rutter (3) concluded
his review of the consequences of deprivation in child-
hood by endorsing Caldwell's recommendation that future
should fulfil three criteria, namely: 1) taking genetic
factors into account, 2) employing stringently defined
measures of stress, and 3) focusing upon specific
psychopathological outcome variables. We have tried to
analyze the history of institutionalization in relation
to the clinical outcome within and outside the schizo-
phrenia spectrum in our sample of children of severely

89
schizophrenic mothers. We proposed, with respect to gene-
tic loading, that schizophrenics and borderline schizo-
phrenics share a similar genetic predisposition that is
more severe than that of those high-risk individuals who
remain mentally healthy. Operationally, this means that
we hypothesize that age of onset of maternal schizophre-
nia should be lower among schizophrenics and borderline
schizophrenics than among high-risk individuals who
remain healthy. With respect to the environmental factors
we have suggested that it is in particular the schizo-
phrenics themselves rather than borderlines who have
premorbidly experienced the most stressful conditions.
Operationally this means that schizophrenics should have
spent more time in institutions during the first five
years of life than did other diagnostic outcome groups.
Table 1 presents means and standard deviations for each
of the three independent variables as a function of diag-
noses in the high-risk group. In this study we have also
included a group of all the other diagnoses apart from
schizophrenia, borderline schizophrenia, and no mental
ilness. The three variables are: mother's age at first
hospitalization (MAH), the number of months spent in
institution during the first five years of life (INS),
and finally, the number of months spent with the schizo-
phrenic mother during the first five years of life (CM) .
Both the schizophrenics and borderlines have mothers
whose age at first hospitalization was significantly
younger than those of the no mental illness group. The
"other diagnoses" group occupied a middle position.

With respect to institutionalization (INS) the

schizophrenics have experienced significantly more
through the first five years of life than have the

Table l . Predictor variables as a function of diagnosis

(High risk group only).
Schizophrenics Borderline Other No Mental ANOVA p
Schizophrenics Diagnoses Illness
N = 13 N = 29 N =76 N =55
mean S.d. mean S.d. mean S.d. mean S.d. F

MAH (agel 27.7 5.4 29.5 8.3 32.1 8.2 34.5 8.3 3.86 .011
INS (months I 22.2 22.8 10.8 16.0 4.1 10.6 2.4 7.6 11.33 <.001*
CM (monthsl 32.5 23.5 41.6 21.0 50.1 17.2 50.3 18.1 4.66 .004*

* In view of the skewness of these two variables we have repeated the analyses using a non-parametric test
( Kruskai·Wallis).
For INS the corresponding probability is< .001 and for CM .003.

90
borde~lines, who in term lie significantly above the
"other diagnoses" as well as the "no mental illness"
group on this variable. In the case of contact with the
mother (CM) the schizophrenics had significantly less
contact with the mothers during the first five years of
life than had both the other diagnoses and no mental
illness groups which do not differ significantly from
each other. These three variables can be expected to be
heavily intercorrelated, and this is also what was found.

An analysis of co-variance of institutionalization

corrected for mother's age at first hospitalization and
for contact with the mother showed that institutionali-
zation significantly keeps its effect. If we entered
mother's age at hospitalization, and institutionaliza-
tion as co-variates for contact with the mother, the
four diagnostic groups did not differ significantly.
Thus, contact with the mother may only have significance
for later psychopathology in the child to the extent it
indirectly reflects the degree of genetic loading the
child has inherited and/or the amount of time spent in
institutions. A detailed report on institutionalization
is published as (4).

CEREBRAL VENTRICULAR SIZE WITHIN THE SCHIZOPHRENIA

SPECTRUM

A number of studies over the last three decades

indicate that chronic schizophrenia is associated with
cerebral ventricular enlargement. The NIHM studies of
twins discordant for schizophrenia showed that the
schizophrenic twins had more soft neurological signs
than had the healthy co-twins. Other studies, for
example by Max Pollock and co-workers, have shown that
siblings of schizophrenics have less neurological defi-
cit than had the schizophrenics themselves. As described
above, we, and many others, have shown that schizophre-
nics have suffered more pregnancy and birth complica-
tions than the other high-risk children. Taking into
consideration also that the American-Danish adoption
studies (5) have demonstrated that schizophrenia and
borderline schizophrenia appears among the biological
relatives of the same schizophrenic index probands, we
forwarded the hypothesis that the genetically transmitted
condition is not schizophrenia itself, but border-
line schizophrenia. Real schizophrenia then is a
complicated (frequently neurologically complicated)
form of borderline schizophrenia. Our concept of border-
line schizophrenia corresponds with the DSM-III schizo-
typal personality disorder.

91
 In order to test this bold hypothesis we studied
from 1979-80 a subgroup of the total high-risk sample
consisting of ten schizophrenics, ten borderlines, and
sixteen no mental illness cases, as diagnosed in our
1972-assessment during which diagnoses were made as a
consensus between two of three diagnostic instruments:
a clinical diagnosis, the Current And Past Psychopatho-
logy Scale (CAPPS), or the Present State Examination
(PSE), 9th edition. In 1979-80 these 36 subjects were
clinically reassessed by the same clinical interviewer,
and the subjects were also examined with a 1010 EMI-
scanner. We used the same method of measurement as
reported by Weinberger and his colleagues from the
United States.

There were no significant differences in the third

ventricle width or the ventricle-brain-ratio (VBR) as a
function of the 1972 consensus or even the 1972 clinical
diagnoses. However, the clinical diagnoses made in 1980
had changed to some extent from the clinical diagnoses
in 1972. Two of the schizophrenias had turned into
borderlines, one of the borderlines had turned into
schizophrenia, one into no mental illness, and one into
other psychiatric disorders. Comparisons based on the
1980 diagnoses showed that schizophrenics had the largest
third ventricles. Fig. 4 showed, that ventricular size as
measured by the VBR was the largest for the schizophrenics
and the smallest for the borderlines, with the no mental
illness group in between. No relationship was found
between ventricular size and the age of the subject,
length of psychiatric hospitalization, drug treatment
or electro-convulsive treatment.

We examined the relationship between ventricular

size in 1980 and the pregnancy-birth complication data
obtained as part of the initial 1962-assessment. We
found significant associations between enlarged ventric-
les in adulthood and signs of prematurity at birth.
There was also a modest correlation between the width
of third ventricle and a global rating of perinatal
complications. For further information see (6).

CONCLUSION

It is an almost insuperable task to report results

from more than 20 years pospective work with this group
of children of schizophrenic mothers and their controls.
In the references (1,2,4,6,7,8) detailed descriptions
and discussions about methodology, and about the concepts
of the various parts of the schizophrenia spectrum may

92
 ,.......30 --Mt:DIAN

~28
ii 26 "
>24
22
20
16
16
14
12 +
10 ~9.76

8 " ~ .. ~
~7.48
6 _:.._5.41 ""
4 " ";tli a,.
2
0~--~----~----.-----.------r--
SCHIZOPHR. BORDERUNE NO M. ILL
N=7 N=11 N=13
+ = 1972 BCl!DoRUNE +" 1972 SCHIZOPHRENIC +" 1972 EOROERUNE

Fig. 4. VBR by 1980 diagnosis.

be found. What we have tried to show is one development

within the genetics of schizophrenia: the utilization of
genetic knowledge to facilitate the study of possible
environmental effects. We have seen, that the clinical
outcome of the children of severely schizophrenic mothers
to some extent may depend on obstetrical complications,
and on institutional care. We have seen that premorbidly
there is a great overlap between the components of the
schizophrenia spectrum. We have also seen, that diffe-
rences in ventricular size to some extent might be a
reflection of differences in environment. We believe
that prospective studies of populations at a high risk
for schizophrenia is the most powerful way of studying
the interactions between heredity and environment.
First, it is not too burdened by the doubtful reliabi-
lity of retrospective information. Second, the value of
following the same specific individuals over time is a
strong advantage, as demonstrated with the diagnostic
changes over time in the cerebral ventricle study. The
high risk design provides us with the best possible
controls: those high-risk subjects who do not become
schizophrenics in spite of the genetic risk.

93
REFERENCES

1) Parnas, J., Schulsinger, F., Teasdale, T.W.,

Schulsinger, H., Feldman, P.M., and Mednick, S.A.:
Perinatal Complications and Clinical Outcome within
the Schizophrenia Spectrum. Brit.J.Psvchiat., 1982,
140: 416-420.

2) Parnas, J., Schulsinger, F., Schulsinger, H.,

Mednick, S.A., and Teasdale, T.W.: Behavioral
Precursors of Schizophrenia Spectrum. Arch.Gen.
Psvchiat~, 1982, 39: 658-664.

3) Rutter, M.: Maternal DePrivation Reassessed, 1981,

Pinquin Books Ltd., Hammondworths.

J) Parnas, J., Teasdale, T.W., Schulsinger, H.: Genetic

Risk and Early Childhood Experience in the Etiology
of Schizophrenia. Arch.Gen.Psvchiat. (Submitted for
publication) .

5) Kety, S.S., Rosenthal, D., Wender, P.H., Schulsinger,

F., Jacobsen, B.: Mental illness in the biological
and adoptive relatives of adopted individuals who
became schizophrenic. In Genetic Research in Psychia-
try, ed. R. Fieve, D. Rosenthal, H. Brill, l~:U!>,
Johns Hopkins University Press, pp. 147-65.

6) Schulsinger, F., Parnas, J., Petersen, E.T.,

Schulsinger, H., Teasdale, T.W., Mednick, S.A.,
and M¢ller, L.: Cerebral Ventricular Size in the
Offspring of Schizophrenic Mothers: A preliminary
study. Arch.Gen.Psychiat. (Submitted for publication).

7) Mednick, S.A., Schulsinger, F.: Studies of children

at high risk for schizophrenia. In Schizophrenia:
The First Ten Dean Award Lectures, 1973, New York:
MSS Inf. Serv., pp. 245-93. .

8) Schulsinger, H.: A ten-year follow-up of children of

schizophrenic mothers: Clinical assessment. Acta
Psychiatr .Scand., 1976, 53: 371.

94
X-LINKED INHERITANCE IN AFFECTIVE DISORDERS

Julien Mendlewicz

Deparbnent of Psychiatry, Erasme Hospital

University Clinics of Brussels, Free University of
Brussels. 1070 Brussels

ABSTRACT

After reviewing the evidence for genetic factors in affective

disorders, the question of X-linked inheritance is discussed
as a node of transmission in a subgroup of bipolar illness.
Recent linkage studies also enphasize the inportance of consi-
dering genetic heterogeneity in the interpretation of genetic
marker studies in affective illness. These genetic investigations
may have relevance to the biological classification and treat.rrent
of affective disorders.

Introduction and Results

Sane early and nore recent genetic research suggested the

presence of an X-linked dominant fonn of bipolar manic-depressive
illness (Winokur et al 1969 ; MerxUewicz et al 1979 ; Mendlewicz
et al 1 980) , although conflicting results have been published
concerning this node of transmission (Gersron et al 1979). '!his
apparent discrepency has been discussed in tenus of diverging
methods of ascertainment of infonnative pedigrees for linkage
analysis, diagnostic criteria and mathematical models used in
linkage analysis for the estimation of map distances on the
X-chronosane. X-chromosane inactivation should also be considered
although this last phenanenon has been given little attention in
linkage studies of psychiatric disorders.

95
TABLE I : SEX DISTRIBUI'ION IN FIRST DEGREE RElATIVES OF BIPOLAR
PROBANDS.

No ( %) OF .AFFEX:TIVELY ILL RELATIVES

------------------------------------------------
M %
------------------
F %
STUDY YEAR 'I'OI'AL
------------------------------------------------------------------
WINOKUR El' AL 1982 40 15 (38) 25 (63)
MENDLEWICZ 1974
& RAINER 229 93 (40) 136 (60)
WINOKUR El' AL 1969 76 20 (26) 56 (74)
KADRMAS El' AL 1979 102 54 (53) 48 (47}
STENSTEU.I' 1952 41 19 (47) 22 (53)
ANGST El' AL 1980 38 15 (39) 23 (61)
MENDLEWICZ & RAINER 1977 29 9 (31) 20 (69)
GERSHON El' AL 1978 79 38 (48) 41 (52)
TAYLOR & ABRAMS 1981 36 11 (31) 25 (69)
JAMES & CHAPMAN 1975 52 13 (25) 39 (75)
OOEI'ZL El' AL 1974 35 13 (37) 22 (63)
GERSHON El' AL 1975 36 20 (55) 16 (45}
ICMA COLIJ\BORATIVE
STUDY 54 22 (41) 32 (59)
------------------------------------------------------------------
TOI'AL 847 342 (39) 505 (61)

ADAPI'ED FRCM WINOKUR AND CReWE I 1983.

If a disorder is transnitted through a daninant gene situated on

the X-chrarosane, the rex ratio for this disorder should bring
about an excess of females affected men carpared to males.
Table I shows the sex distr:ihution of affective illnesses in first
degree relatives of probands suffering from bipolar illness in va-
rious recent ;family_ studies of bipolar illness.

out of a total of 847 affectively ill first degree relatives of

bipolar probands, there were 342 males with a diagnosis of affec-
tive illness (39 %) for 505 females (61 %) • 'Ihus a clear excess
of females over males with a diagnosis of affective illness (bipo-
lar or unipolar) was observed in first degree relatives of bipo-
lar manic -depressive patients. Accord:ing the X-linked daninant
hypothesis there srould be no male to male transmission of tha
trait, if assortative mating is not present, because the X-chrom:>-
sane of a male. can only. be transmit~ through the nnther'side of
the kindred. Table II summarizes these data from some recent stu-
dies on bipolar illness. N::> or very little male to male transmis-
sion of bipolar. disorders. has been observed in nnst studies

96
 TABLE I I : MALE 'IO MALE TRANSMISSION OF BIPOLAR IL.INESS

STUDY FAMILIES RESULTS

WINOKUR El' AL 1970 89 0

TAYLOR & ABRAMS 1973 55 0
MENDLEWICZ & RAINER 1977 30 0
GERSHON El' AL 1980 16*t 0
MENDLEWICZ & RAINER 1974 134 11%
PERRIS 1971 138 10%
DUNNER El' AL 1969 23* 18%
GREEN El' AL 1973 35 10%
HELZER & WINOKUR 1974 30* 5%
VON GREIFF El' AL 1973 16* 22%

* MALE BIPOLAR PROBANDS

SELECTIVE ASCERI'AINMENI'.
** LINKAGE STUDY

TABLE I I I : LINKAGE STUDIES IN BIPOLAR AFFECI'IVE DISORDERS

------------------------------------------------
STUDY YEAR GENETIC MARKER
----------------
LINKAGE RESULTS

WINOKUR & TANNA 1969 xrf + (-)

REICHEl' AL 1969 COLOR BLINDNESS +
MENDLEWICZ El' AL 1972 COLOR BLINDNESS +
MENDLEWICZ & FEISS 1974 COLO~JtLINDNESS +
+ (-)
BARON 1977 COLOR BLINDNESS +
JOHNSON & LEEMAN 1977 COLOR BLINDNESS - ?
GERSHON El' AL 1979 COLOR BLINDNESS - (+)
LEn<MAN El' AL 1979 X~ - ?
MENDLEWICZ El' AL 1979 COLOR BLINDNESS + (-)
GERSHON El' AL 1979 COLOR~SS + (-)
MENDLEWICZ El' AL 1980 G6PD +
VYAWAHARE &. DEO 1981 MUSCULl\R DYSTROPHY +
(BEKER)

+ LINKAGE POSITIVE CERI'AIN

(+) LINKAGE POSITIVE PROBABLE
- LINKAGE NEGATIVE CERI'AIN
(-) LINKAGE NEGATIVE PROBABLE
? LINKAGE UNKNCM

97
indicating, that male to male transmission of bir;olar illness,
i f present, seans to be a rather rare even in some families.

Table III provides a listing of all available linkage studies in

bipolar affective disorders in reference to the type of genetic
marker studied.

Among the genetic markers studied for the X-chranoscme are the
Xga blood group, color blindness (deuteranopia and _protanopia)
glucose -6- phosphate dehydrogenase deficiency and muscular
distrophy (Becker type). 'Ihe linkage results are generally in
favor of the X-linked dominant transmission in bipolar affective
disorders as illustrated in table IV.

TABLE IV : LINKJI.GE STUDIES ll! BIPOI.J\.R AFFECriVE DISORDERS.

IN FAVOR OF X-LINI<AGE AGAINST X-LINKAGE

NUMBER OF STUDIES 10 3
NUMBER OF PEDIGREES 47 11

Table IV shows that ten linkage studies are definitively in favor

of X-linkage on a total sample of 47 informative families while
3 studies on 11 families reach opposite conslusions.

Conclusion

Among nost relevant family studies of bipolar affective disorders,

a majority report a sex ratio distribution of affective illness
in first degree relatives (excess of fema.les over males) canpati-
ble with an X-linked dominant transmission. An investigation of
male to male transmisson of bipolar illness reveals that this
pattern of inheritance if not absent, is nevertheless a rare
event in the k.indreds of bipolar probands. Furthennore the great
majority of linkage studies are conclusive and daronstrate the
presence of linkage between bipolar manic depressive psychosis
and several X-linked genetic markers, in a subgroup of bipolar
illness. Clearly other forms of affective disorders are mt

98
transmitted through the X-chrorroSOire indicating that genetic hetero-
geneity is present in bipolar illness.

E. S. Gershon, s. D. Targums,s. Ma.tthysse, w. E. Bunney Jr.

Color blindness not closly linked to bipolar illness. Arch.
Gen. Psychiat. 36, 1423, 1979.

J. Mendlewicz, P. Linkoski, J. Wilnotte : Linkage Between

glucose - 6 - phosphate dehydrogenase deficiency and manic-
depressive psychosis. Brit. J. Psychiat. 137, 337, 1980.

J. Mendlewicz, P. Linkowski, J.J. Guroff,H. M. Van Praag

Color blindness linkage to bipolar manic-depressive illness.
Arch. Gen. Psychiat. 36, 142, 1979.

G. Winokur, R. J. Clayton , T. Reich : Manic-Depressive

Illness, St. Louis, C V t-t:>sby CD 1969.

G. Winokur, R.R. Crowe : Bipolar illness. 'rhe sex-polarity

effect in Affectively ill family rrernbers. Arch. Gen. Psychiat.
40, 57, 1980.

99
GENETICS OF PSYCHOGENIC DISEASES

Heinz Schepank
Psychosomatische Klinik am Zentralinstitut
fur Seelische Gesundheit, Postbox 59 70
6800 Mannheim 1, FRG

My topic "genetics of psychogenic diseases" seems to be

contradictory: Psvchoqenic I call such diseases in which
the psychosocial environmental influences play a consid-
erable aetiopathogenetic part. Such are more precisely:
the diseases according to ICD 300 to 307 (WHO, 8th rev.),
i.e. psychoneuroses, personality disorders, sexual de-
viations, alcohol and drug dependence as well as various
psychosomatic disorders.
Genetics in connection with the above diseases
means that hereditary factors have a share in the inter-
individual variance of the manifestation of such dis-
eases. I here do not refer to the simple knowledge that
a human being as a species actually has got in his genes
the possibility of producing psychogenic diseases.
The scientific question is: Can we detect genetic
determinants among psychogenic diseases or are they
caused exclusively by psychosocial influences?
We discern three methods of human genetic research
on psychiatric diseases:
Investigation of:
1. biochemical-molecular processes,
2. the morphological-chromosomal substratum, or
3. the phenomena of the psychopathologic manifestation
of a disease.
I shall mainly speak about results gained by the
third way. In this it is the strategy to use genealo-
101
gical and adoption studies and mainly twin methods. The
most important research instrument is still the classi-
cal comparison of the rates of concordance of the mono-
zygotic twin pairs with those of the dizygotic ones.
Scientifically valuable results gained by twin research
strictly demand methodically cleanliness although this
often is an alaborate task, f.i. in sampling techniques.
The influence of hereditary factors on the mani-
festation of psychoneurose~ in general has been shown
by several investigators: Bakwin and Davidson, Braconi,
Eysenck and Prell, Gottesman, Juel-Nielsen, Shields,
Sturnpfl. A clearly hereditary component in the mani-
festation of obsessive-comoulsiv~ neurotic symptoms and
character structures has been demonstrated by the twin
and genealogical studies of: Brown, Carey and Shields,
Gottesman, Ihda and Inouye, Marks, RUdin, Sakay, Slater
and Cowie. - Concerning anxiety and especially genera-
lized anxiety SYndrome~ as well as phobias there are done
twin investigations which prove a participation of a he-
reditary component by P.E. Becker, Slater and Shields,
Torgersen. - Results concerning neurotic deoression are
not as clear: Shapiro, Schepank. Some authors did not
find any hereditary influences here: Slater and Shields,
Stenstedt, Torgersen.- HYsteria. probably is not here-
ditarily determined.
Difficult to classify is the scope of character neu-
roses and of oersonalitv disorders and abnormal behavior.
In general, distribution patterns of concordance rates
among twins support a hereditary participation. Genea-
logical and recent adoption studies lead to the same re-
sults, especially in antisocial behav~/crirninality: v.
Baeyer, Christiansen, Dalgard, Hayashi, Hutchings and
Mednick, Kranz, Kringelen, Lange, Rosanoff and Handy and
Plesset, Sturnpfl, Yoshimasu. - According to the classi-
cal investigations of Kallmann as well as of Heston and
Shields we may suppose that the manifestation of an in-
clination-homosexualitv.in males is very much hereditari-
ly determined, although we do not know all the details
of the aetiopathogenetic mechanisms. - Since some years
we have got to know some certain facts about the mecha-
nisms and the participation of genetic components in the
occurrence of alcoholism. Here the co-operation of twin
research, transcultural epidemiology, neurophysiology,
biochemistry and adoption method has been especially
successful: Johnsson, Kaij, Nielson, Partanen, Propping
and Vogel.

102
The psychopathologically important suicidal behavior
has also been investigated. It is not genetically de-
termined as has been shown by many twin studies,-only
with one exception: If suicide occurred as the conse-
quence of an endogenic psychosis it is hereditarily de-
termined (Haberlandt, Harvald-Hauge, Juel-Nielsen and
Videbech, Kallmann).
For several psychosomatic disorders the twin me-
thod could prove a participation of genetic components
in the occurrence of the disease, as there are: some
functional disorders and here especially enuresis noc-
turna (Bakwin, Hallgreen, Gedda and Alfieri, Weitz and
Thiesen) stuttering (Godai and Tatarelli, Lenz, Nelson
et al., Seemann), the so called general veqetative syn-
drome of females (Curtius and Feyereis) and headache
(Harvald and Hauge, Spaich and Ostertag). -The same
applies to different psychosomatic disorders with patho-
logical-anatomic laesion, f.i. the pylorus neighboured
ulcus ventriculi, the ulcus duodeni (Eberhardt, Hauge
and Harvald, Gottlieb Jensen, Verschuer), asthma_pron-
chiale (Hauge and Verschuer) , urticcu:·J.a (Niermann, spaJ.ch
and os tertag) •
I am sorry I cannot mention the complete list of
authors as time is pressing.
In classifying these diseases there is a fundamental
difficulty: The border between healthy and ill is -
especially in psychogenic disorders - often fluent; the
classification of individual cases often leads to contro-
versial opinions. Psychoneurotically-psychosomatically
mixed patterns are frequent, change of symptoms is well
known to the clinician.
pur own investiqations of 100 psychogenically dis-
eased twin pairs - 36 monozygotic and 64 dizygotic
pairs - we began in 1963, together with A. Heigl-Evers.
SO index twins we took from a psychotherapeutic out-pa-
tient institution (Berlin) und SO pairs from a psycho-
therapeutic-psychosomatic hospital (Goettingen). There-
sult of the half-structured depth-psychological investi-
gation of each of the partners of our 100 twin pairs has
been: 1. The intrapair difference of the impairment by
psychogenic symptoms is signit1cantly less among monozy-
gotic twin pairs than among dizygotic ones. - 2. Compar-
ing the ~mptoms of our 100 twin pairs we found a signi-
ficantly n1gher rate of concordance among monozygotic
twins (31%) than among dizygotic ones(16%). Even after
subdividing the symptoms into psychoneurotic (ICD 300),

103
psychosomatic (ICD 305 and 306) and character neurotic
ones (ICD 301 to 304) the rates of concordance of mono-
zygotic pairs are significantly higher than thos of di-
zygotic ones. - Among our 18 investigated infant twin
pairs the rates of concordance of the monozygotic twins
were clearly higher than among adults.
In detail, we could clearly prove a hereditary par-
ticipation for the following symptoms: psychogenic socio-
communicative disorders, sleep disorders and neurotic de-
pressions; and - by means of an additional twin study -
for anorexia nervosa.
The following result seems to be very important:
we could not find any genetic connection between the
mentioned psychogenic diseases and the manifestation of
a __o_s_vchosi~;_. Among the twin partners of our neurotic in-
dex twins psychoses did not occur more frequently than
among our general population.
Summarv:
The twin investigations of the last decades show: In the
manifestation o the most psychogenic diseases hereditary
factors have a probable share in the intraindividual
variance. In some of these diseases hereditary factors
are even of great significance, f.i. in stuttering and
in inclination-homosexuality in males.-Finally, I want
to emphasize that in contrast to the psychoses the rates
of concordance in psychogenically diseased monozygotic
twins are not very high. That means: Environmental in-
fluences have a great share in the manifestation of psy-
chogenic disorders.

LITERATURE
A. Heigl-Evers and H. Schepank: " Ursprlinge seelische
bedingter Krankheiten". (1980/81) 2 Vol., Vandenheock
and Ruprecht, Gottingen, Zlirich

104
PRECLINICAL PHARMACOLOGY OF ANTIDEPRESSANTS

Willy Haefely
Pharmaceutical Research Department
F. Hoffmann-La Roche &Co., Ltd.
CH-4002 Basel, Switzerland

ANTIDEPRESSANTS AND BRAIN MONOAMINE$

Anti-depressant drugs have now been available for a quarter of
a century. Historically the two main categories of antidepressants,
the monoamine oxidase (MAO) inhibitors and the imipramine-like
drugs, were discovered by serendipity in the clinic.
The antidepressant action of the tuberculostatics, isoniazid
and iproniazid, was soon related to their inhibitory effect on MAO,
the enzyme that catabolizes the brain monoamines noradrenaline (NA),
dopamine (DA), and serotonin (5-hydroxytryptamine, 5-HT). Imipra-
mine had been originally considered a potential antipsychotic and
antihistamine based on its structural similarity to chlorpromazine.
Imipramine was found to enhance the contractile response of tissues
to exogenous NA and proposed to act as an antidepressant by poten-
tiating transmission in central noradrenergic synapses 1 • This was
supported by our finding that imipramine and similar compounds
potentiated the effect of endogenous NA released at peripheral nor-
adrenergic synapses by blocking its neuronal re-uptake 2 • The same
study also showed that the effect of imipramine on noradrenergic
transmission was a biphasic potentiation-depression with increas-
ing doses of the antidepressant. This complication, due to NA
receptor blockade, has been greatly neglected in later attempts
to understand the changes that might be induced by imipramine-
like drugs in central noradrenergic transmission. In spite of an
apparently obvious connection between monoamine metabolism and
antidepressant action, the convincing evidence that MAO inhibitors
increase the neuronal release or delay the removal of NA (or of any
105
other monoamine) released into the synaptic cleft, is still lacking.
It was tempting to assume an involvement of monoamines not only in
the main mechanism of action of antidepressants, but also in the
pathogenesis of affective disorders (monoamine hypothesis of de-
pression). The discovery of presynaptic control mechanisms regulat-
ing the release of monoamines by nerve impulses opened new ways for
pharmacologically manipulating the synaptic concentration of mono-
amines and, perhaps, for explaining the mode of action of mianserin.
Antidepressants were important probes in elucidating the role
of NA and 5-HT in the modulation of ponto-geniculo-occipital (PGO)
waves, a phasic phenomenon of REM sleep; vice versa, PGO waves, in-
duced in cats by a predominant elimination of either NA or 5-HT
mechanisms, became a very useful tool to study the effects of novel
compounds on the synaptic concentration of these two monoamines 3 •
Recent systematic investigations of sleep disturbances in depres-
sion and of the antidepressant effect of sleep deprivation~ as well
as the concept of phase advance of circadian rhythms all support
the notion that disturbances of monoaminergically modulated sleep
parameters are a cause or an epiphenomenon of depression and that
antidepressant drugs improve this disease by affecting sleep and
additional functions that are under the control of monoamines.
A great challenge to research has been the suspicion that the
effects of antidepressants observed in acute experiments, such as
e.g. inhibition of monoamine uptake and potentiation of monoamine
transmission, were unrelated to the therapeutic action because the
latter usually sets in with a latency of days or weeks 5 • This argu-
ment is not compelling; antibiotics, as an example, are bacteri-
cidal briefly after a first, appropriate dose, however, the general
condition of a patient with a severe infectious disease takes some
time to improve. Obviously, the time course of improvement of dis-
ease-disturbed functions depends not only on intrinsic properties
of a drug but also on the nature of the disease. There is no rea-
son why the psychobiologically complex state of depression, which
takes time to develop fully, and which contains a considerable
number of "learned" or conditioned elements, should be reversed
immediately after altering the dynamics of neuromodulators such as
the monoamines. Nevertheless, the hypothesis had the merit to em-
phasize possible drug-induced changes that may be slow to occur
with chronic treatment. It seems safe to conclude that the "down-
regulation" of monoamine receptor-mediated responses is the con-
sequence of a longer-lasting enhancement by antidepressants of
monoamine receptor stimulation and that this "down-regulation"
might reduce the effectiveness of antidepressant treatment (kind
of tolerance) rather than initiate the therapeutic process. Long-
term adaption processes have so far been studied only in apparently

106
normal animals and we may question whether the monoamine systems in
the brain of a depressed patient respond in the same way.
A most striking aspect of our circular reasoning about mono-
amines as targets of antidepressant drugs and as the biological
basis of depression is the old controversy about the effects of
monoamines on neuronal activity. Recent studies seem to indicate
that NA (and possibly the other 3 monoamines) are neither inhibi-
tory nor excitatory neurotransmitters in the classical sense of go
and no-go signals; they rather alter the integrative or discrimi-
natfve function of neurones by reducing their sensitivity to weak
excitatory and inhibitory inputs and by increasing their responses
tostrong (supraliminal) excitatory inputs (increaseofthe 11 Signal-
to noise ratio 11 ) 6 • Such a 11 biasing 11 action of monoamines would ap-
pear to be more appropriate for endogenous modulators thought to
be involved in alertness, attention, motivation and mood.
THE PHARMACOLOGIST'S APPROACHES TO ANTIDEPRESSANTS
The preclinical pharmacology of antidepressant drugs can be
studied by 3 principal approaches. A behavioural aooroach would a
priori appear to be the most appropriate one for drugs affecting
mood and psychomotor activit.v. In the absence of, at least de-
tectable, spontaneous animal forms of depression, several experi-
mental models have been proposed and used, the oldest being the
state induced by reserpine or other reserpine-like monoamine de-
pleters. The 11 depression 11 in this model concerns rather locomotor
activity and sympathetic nervous activity than mood. The model
does not detect agents interfering selectively with 5-HT. The other
models proposed (separation-induced behavioural changes, learned
helplessness, behavioural despair, rewarding brain self-stimula-
tion) are either too complicated to be useful in drug screening,
give too variable results with standard drugs, or give too many
false positive results. Of the electroph.vsioloQical methods avail-
able, only the study of REM sleep has given reliable results. In-
deed, all drugs generally accepted to have clinical antidepressant
activity were found to selectively depress REM sleep of cats in
appropriate doses 7 ' 8 • The biochemical methods are those frequently
and successfully used in the screening of drugs that inhibit mono-
amine catabolism or uptake. The value of studying the interaction
with specific imipramine binding sites remains to be demonstrated.
NEW PERSPECTIVES
For a quarter of a centur~ the monoamines have been, among
endogenous neuromodulators, the only strong candidates for primary
targets of antidepressants as well as for fundamental biochemical

107
disturbances in depressive states. Even though novel, endogenous
compounds (in particular neuropeptides) are being proposed as po-
tentially important biochemical markers of depression, none is
apparently able to invite serious medicinal chemists and pharma-
cologists to start major programmes on new antidepressants. After
twenty years of intense research, we have accumulated an impres-
sive amount of new information on monoamine mechanisms and their
response to antidepressants. And yet, there is still some doubt
about the mechnism of action of the present antidepressants and
the pathogenetic role of monoamines in depression; thus, with re-
spect to fundamental views on the biological basis of depression
and antidepressant drugs, no entirely new line has been opened.
Resignation might be one reaction of industrial drug research sci-
entists to the present state. Alternatively, the progress achieved
in the past might be used to maximally improve the monoamine ap-
proach of antidepressant drug therapy. Optimization of chemical
structures known to interfere with specific monoamine mechanisms
should either improve potency and/or specificity for one mechanism
or permit the design of a single molecule capable of interacting
with several mechanisms, as well as reduce the non-negliaible lia-
bility of presently available drugs for side effects. Adequate
dosing by the physician and improved patient compliance would en-
hance the chance of obtaining more favourable clinical results.
Three compounds, that were developed in our laboratories and that
are in phase II to III clinical studies, may serve as examples of
this strategy.
Cianooramine (Fig. 1) is the result of chemical optimization
of the imipramine structure with the aim of increasing the potency
for 5-HT uptake inhibitory acticity. In vitro as well as in vivo,
cianopramine is a most potent inhibitor of 5-HT reuptake. Inhibi-
tion of NA uptake is weak9 • Experience on about 700 patients con-
firms the very high antidepressant potency (effective daily dose
4 to 16 mg) and the predicted lower liability for anticholinergic
and cardiovascular side effects.
Diclofensine (Fig. 1) is an isoquinoline derivative with some
structural similarity to nomifensine, from which, however, it dif-
fers in several important aspects 10 • Diclofensine is unique in in-
hibiting the uptake of NA, DA and 5-HT to virtually the same extent
and it is a particularly potent inhibitor of DA uptake. In contrast
to nomifensine, diclofensine is very weak as a DA releaser and is,
consequently, less stimulant. The concerted action on the uptake
of the three amines seems to be the basis of the excellent toler-
ance and antidepressant activity which has been observed in over
800 patients studied so far.

108
 Cl Cl

"'
~

r c1--o-{~N~0
N I
H3CO 'cH3 H

Cianopramine Diclofensine Moclobemide

(Ro 11-2465) (Ro 8-4650) (Ro 11-1163)

Figure 1. Structural formulae of 3 novel antidepressants.

Moclobemide (Fig. 1) is a MAO inhibitor with a completely nov-

el structure 1 1 • It is neither a hydrazine nor an acetylenic com-
pound; MAO (preferentially type A) is inhibited in a reversible way
(duration 16 to 24 h). The compound has an excellent tolerance in
animals and man. The potentiation by moclobemide of the pressor
effect of ora 1 tyramine in anima 1s and man is negl i gi b1e in compa-
rison with previous MAO type A inhibitors. Moclobemide lacks the
central stimulant and the hypotensive effect of many previous MAO
inhibitors. Moclobemide, tested so far in about 600 patients, ap-
pears to be the promising newcomer in the field of improved MAO in-
hibitors, an area which meets a revived interest of clinicians.
These novel tailor-made antidepressants do not only promise
an improved therapeutic potential; the detailed analysis of their
preclinical pharmacology make them also extremely useful tools for
improving our understanding of the biological basis of depression
and/or of the mechanisms of action of antidepressants.
REFERENCES
1. E.B. Sigg, Pharmacological studies with tofranil, Canad.
psychiat. Ass. J. 4: 75 (1959),
2. ~. Haetely, A. HUrlimann, and H. Thoenen, Scheinbar paradoxe
Beeinflussung von peripheren Noradrenalin-Wirkungen durch
einige Thymoleptica, Helv. Physiol. Pharmacal. Acta 22: 15
(1964) .
3. W. Haefely, M.-A. Ruch-Monachon, M. Jalfre, and R. Schaffner,
Interaction of psychotropic agents with central neurotrans-
mitters as revealed by their effects on PGO waves in the cat,
Druq Research 26: 1036 (1976).
4. G.W. Vogel, Evidence for REM sleep deprivation as the mecha-
nism of action of antidepressant drugs, Proar. Neuroosvcho-
oharmacol. &Biol. Psychiat. 7: 343 (1983).

109
 5. F. Sulser, New perspectives on the mode of action of antide-
pressant drugs, Trends in Pharmacoloqical Sciences 1: 92 (1979).
6. D.J. Woodward, H.C. Moises, B.D. Waterhouse, B.J. Hoffer, and
R. Freedman, Modulatory actions of norepinephrine in the cen-
tral nervous system, Fed. Proc. 38: 2109 (1979).
7. P. Pole, J. Schneeberger, and W. Haefely, Effects of several
centrally active drugs on the sleep-wakefulness cycle of cats,
Neurooharmacoloqy 18: 259 (1979).
8. R. Scherschlicht, P. Pole, J. Schneeberger, M. Steiner, and
W. Haefely, Selective suppression of rapid eye movement sleep
(REMS) in cats by typical and atypical antidepressants, in:
"Typical and atypical antidepressants: molecular mechanisms,"
E. Costa and G. Racagni, eds., Raven Press, New York (1982).
9. M. Da Prada, H.H. Keller, W.P. Burkard, R. Schaffner, E.P.
Bonetti, J.M. Launay, and W. Haefely, Some neuropharmacolo-
gical effects of Ro 11-2465 - a novel tricyclic antidepress-
ant I:Jith potent inhibitory activity on the uptake of 5-HT,
in: "Typical and atypical antidepressants: molecular mecha-
nisms", E. Costa and G. Racagni, eds., Raven Press, New York
(1982).
10. H.H. Keller, R. Schaffner, M.D. Carruba, W.P. Burkard, M.
Pieri, E.P. Bonetti, R. Scherschlicht, M. Da Prada, and W.E.
Haefely, Diclofensine (Ro 8-4650) - a potent inhibitor of
monoamine uptake: biochemical and behavioural effects in com-
parison with nomifensine, in: "Typical and atypical antide-
pressants: molecular mechanisms'', E. Costa and G. Racagni,
eds., Raven Press, New York, (1982).
11. M. Da Prada, R. Kettler, H.H. Keller, and W.E. Haefely, Neuro-
chemical effects in vitro and in vivo of the antidepressant
Ro 11-1163, a specific and short-acting MAO-A inhibitor. Mod.
Probl. Pharmacoosvchiat. 19: 231 (1983).

110
MAO-INHIBITORS REVISITED I: ADRENALINE (A) AND NORADRENALINE (NA)

IN CEREBROSPINAL FLUID (CSF) IN ISOCARBOXAZIDE TREATED RATS

Annette Gjerris, David I. Barry,

Niels J. Christensen and Ole J. Rafaelsen

Psychochemistry Institute & Department of Psychiatry

Rigshospitalet
9, Blegdamsvej
DK - 2100 Copenhagen, Denmark

For many years the research on biology af depressions has focused

on noradrenaline and serotonin. However, it has been known for nearly
10 years that adrenaline is present in neurons in the brain (Hokfelt
et al., 1974), and experimental studies indicate that hypothalamic
adrenaline is of central origin (Mefford et al., 1981). Using a highly
specific and sensitive isotope derivative technique we are now able to
measure adrenaline in CSF and brain (Christensen et al., 1980; Gjerris
et al., 1981; Christensen et al., 1983). Before entering clinical stu-
dies of depressed patients to test a possible relation between changes
in central adrenaline and clinical outcome of treatment with different
antidepressant principles, we found it valuable first to investigate
the influence of antidepressant drugs on adrenaline and noradrenaline
in rat brains and to see, whether a change in brain concentrations
was reflected in the CSF.

MATERIAL AND METHODS

Druq administration

Isocarboxazide (Marplan®), Four groups of 8 Wistar rats

250-300 g received isocarboxazide 10 mg/kg. Isocarboxazide was
administered by gastric intubation.

Group 1 received a single dose and was studied 1 hour later.

Group 2 received two doses 12 hours apart and was studied 12 hours
later. Group 3 received one daily dose for seven days and was studied

111
24 hours later. Group 4 received one daily dose for 6 weeks and was
studied 24 hours later. The three control groups consisted of 8
untreated rats, 8 who received the vehicle by gastric intubation in
one daily dose for seven days, and 9 who received the vehicle in one
daily dose for 6 weeks.

Sample Collection

The sample collection pro~edure was the same for all animals.
Initial anaesthesia was introduces with 4% halothane in N2o:o2 70:30.

CSF. The CSF (100-150 ~1) was collected by puncture of cisterna

magna, using a glass capillary tube. It was transferred to a plastic
tube containing glutathione and EGTA1 and frozen immediately.

Brain. The brains from eleven 6 weeks isocarboxazide treated rats

were dissected using the procedure described by Glowinsky and Iversen
(1966) and frozen immediately. Adrenaline and noradrenaline were deter-
mined in hypothalamus, an area known for its high content of noradrena-
line, and in hippocampus, an area with a low noradrenaline content.
The results were compared with concentrations in 9 rats who had re-
ceived the vehicle in one daily dose for 6 weeks.

RESULTS

The medians of adrenaline and noradrenaline in CSF were equal

in the untreated control group and the groups treated with vehicle
for 1 and 6 weeks, respectively. After 24 hours of isocarboxazide
treatment a significant increase was seen for adrenaline reaching a
maximum increase of 250% after 6 weeks of treatment. on the other
hand, the increase in noradrenaline concentration did not become
significant until 1 week of treatment and the maximum increase after
6 weeks of treatment was only 40%.

The concentrations of adrenaline and noradrenaline in hypothala-

mus were in accordance with our findings in CSF. After 6 weeks of
treatment with isocarboxazide an increase in adrenaline of 135% was
seen compared with the increase in noradrenaline,which only reached
a maximum of 70%, both increases being statistically significant.

The concentrations of adrenaline in hippocampus were extremely

low in both treated and untreated rats, being measurable in only 4
out of 10 isocarboxazide treated rats and 5 out of 9 controls. No
tendency of increase in adrenaline- and noradrenaline concentrations
was demonstrated in the treated group. The sum of adrenaline from 19
rats including both groups was 24 pmol/g. The corresponding value for
noradrenaline was 52 nmol/g. Anticipating that adrenaline measured is
in fact noradrenaline, maximum cross-over from noradrenaline to
adrenaline is 0.05%.

112
DISCUSSION

The increase of adrenaline in rat brain and CSF is in accordance

with results by Fuller and Hemrick-Leucke (1981) and Da Prada (1983),
who reported an effect on brain adrenaline compared to the increase in
noradrenaline in rats treated with monoaminoxidase-inhibitors type A.
The present study indicates that the mixed MAO-I, isocarboxazide, has
a similar influence on brain adrenaline. On the other hand Scatton and
Bartholini (1981) reported an effect on adrenaline turnover both for
some antidepressants and for some antipsychotics; this underlines the
need for further studies of other drugs in order to elucidate whether
we are dealing with an unspecific reaction.

The very low concentrations of adrenaline in CSF and brain com-

pared to noradrenaline might lead to the supposition that the adrena-
line values were influenced by noradrenaline due to a methodological
failure. Against such an assumption speak the results from the hippo-
campus measurements showing a maximum cross-over of 0.05%.

In experimental studies in rats it has been suggested that hypo-

thalamic adrenaline concentration and turnover are particularly respon-
sive to acute and chronic stress (Roth et al., 1982) and other studies
have demonstrated an influence of central adrenaline on growth hormone
secretion and corticosterone release (Roth et al., 1981; Terry et al.,
1982).

The present and the above-mentioned studies indicate that central

adrenaline may play a role in the biology of depression, an assumption
supported by two previous investigations (Christensen et al., 1980;
Gjerris et al., 1981) showing significantly decreased CSF-adrenaline
in depression.

REFERENCES

Christensen, N.J., Vestergaard, P., S¢rensen, T., and Rafaelsen, O.J.,

1980, Cerebrospinal fluid adrenaline and noradrenaline in depres-
sed patients, Acta Psvchiatr Scand, 61:178.
Christensen, N.J., Galbo, H., Gjerris, A., Henriksen, J.H., Hilsted,J.,
Kj~r, M., and Ring-Larsen, H., 1983, to be published, Whole body
and regional clearances of noradrenaline and adrenaline in man,
Acta Phvciol Scand.
Da Prada, M., Kettler, R., Keller, H.H., and Haefely, W.E., 1983,
Neurochemical effects in vitro and in vivo of the antidepressant
Ro 11-1163, a specific and short-acting MAo-A inhibitor,
MQd Probl Pharmacopsychiatry, 19:231.
Fuller, R.W., and Hemrick-Leucke, S.K., 1981, Elevation of epinephrine
concentration in rat brain by LY51641, a seleotive inhibitor of
type A monoamine oxidase. RPs Commun Chern Pathol Pharmacol,
32:207.

113
Gjerris, A., Jensen, E., Christensen, N.J., and RafaelseP, O.J.,
1981, Adrenaline and noradrenaline in psychiatric disorders,
in: "Biological Psychiatry", c. Perris, G. Struwe, and B. Jans-
son, eds., Elsevier/North-Holland Biomedical Press, Amsterdam,
565.
Glowinski, J., and Iversen, L.L., 1966, Regional studies of catechol-
amines in the rat brain - I, J Neurochem, 13:655.
Hokfelt, T., Fuxe, K., Goldstein, M., and Johansson, 0., 1974,
Immunohistochemical evidence for the existence of adrenaline
neurons in the rat brain, Brain Res, 66:235.
Mefford, I.N., Roth, K.A., Paxinos, G., and Barchas, J.D., 1981,
Central and peripheral contributions to hypothalamic epinephrine,
Brain Res, 244:175.
Roth, K.A., Katz, R.J., Sibel, M., Mefford, I.N., Barchas, J.D., and
Carroll, B.J., 1981, Central epinergic inhibition of corticoste-
rone release in rat, Life Sci, 28:2389.
Roth, K.A., Mefford, I.M., and Barchas, J.D., 1982, Epinephrine, nor-
epinephrine, dopamine and serotonin: differential effects of
acute and chronic stress on regional brain amines. Brain Res,
239:417.
Scatton, B., and Bartholini, G., 1981, Drug-induced changes of
epinephrine turnover in the rat hypothalamus, Life Sci, 29:1161.
Terry, L.C., Crowley, W.R., and Johnson, M.D., 1982, Regulation of
episodic growth hormone secretion by the central epinephrine
system, .J Clin Invest, 69:104.

114
MAO-INHIBITORS REVISITED II: CLINICAL IMPLICATIONS

Ole J. Rafaelsen and Annette Gjerris

Psychochemistry Institute & Department of Psychiatry

Rigshospitalet
9, Blegdamsvej
DK - 2100 Copenhagen ¢, Denmark

For nearly forty years clinicians have had the feeling that
there were types of depressions responding better or only to
MAO-inhibitors. But whenever it was tried to prove this tenet by
comparisons in well-planned studies, the results were disappointing.
The most notorious case was the British Medical Research Council
from 1965 where the MAO-inhibitor phenelzine was inferior not only
to electroconvulsive therapy (ECT) and to imipramine, but also to
placebo, the latter difference was, fortunately enough, not
statistically significant (Shepherd, 1965). Ever since, many
clinicians has argued that it was indeed the 'atypical' depressive
patients who were candidates for MAO-inhibitor treatment and that
such patients did rarely enter psychiatric hospitals and depart-
ments and that it therefore was not surprising that the results
with hospitalized patients were rather disappointing (Quitkin
et al., 1981).

Giller and his co-workers (1982) compared Marplan(R)

(= isocarboxazide) with placebo and found trends. suggesting that
the markedly improved group showed less depression, anxiety,
sleep disturbance, weight loss and fewer gastrointestinal complaints,
but more helplessness and worthlessness. Others have also had to
conclude that there were only weak differences in the profile
between a tricyclic antidepressant like amitriptyline and a
MAo-inhibitor like phenelzine, amitriptyline producing greater
improvement in depressive impairment of work and interests while
phenelzine produces additional improvement in anxiety ratings
(Rowan et al., 1981). Nies and Robinson (1981) compared the same
two drugs and found no differences in measures of total symptom
severity or in patients with 'fears of recent onset'; they

115
suggested, however, that phenelzine might be particularly
efficacious in treating patients whose symptomatology included
'social fears'.

When treating neurotic patients with phenelzine or diazepam

Martin Roth and his group found no evidence of a difference in
pattern of response between the two treatment groups and it was
not possible to predict response to treatment from pretreatment
variables (Mountjoy et al., 1980) .

A new approach was undertaken by Davidson and his group when

making operational definitions of typical and atypical depression:
"we define typical depression as a state which is associated with
loss of any of the following: appetite, weight or libido.
Atypical depression is the converse, namely a gain in any of the
above" (Davidson et al., 1982). Their results applying this very
simple set of distinctions were remarkable. The 'typical'
depressions did not significantly better on isocarboxazide than
on placebo, although a uniform trend was noted which favoured
the MAO-inhibitor, whereas the 'atypical' depressions receiving
isocarboxazide showed significantly greater improvement than
those atypical cases receiving placebo on items of depression,
anxiety, phobia, sensibility, somatisation, and obsession
(Davidson et al., 1981).

The last-mentioned study by its simple operationalisations

of typical and atypical depression should invite for replication
studies to falsify or verify this interesting tenet.

The various reviews are immediately confronted with the

crucial problem: How can you discuss pfficacy as long as you have
not defined the patient population? (Quitkin, Rifkin & Klein, 1979).

Suffice it to say that a majority of phenelzine-placebo

studies showed positive effects in non-endogenous depression,
whereas the results were modest or outright not different from
placebo in some important studies in endogenous depression
(Greenblatt, Grosser & Wechsler, 1964; Shepherd, 1965; Raskin
et al., 1974).

It has been tried to explain some of the inherent problems

pharmacodynamically and pharmacokinetically. Pharmacodynamics
includes effective dosage and duration of treatment. Inhibition
of at least 80 per cent platelet-MAO led to clinical improvement
in 68 per cent of phenelzine treated patients, whereas only 44 per
cent of those with lower MAO-inhibition had a favourable response
(Robinson et al., l978a). The same researchers found that with
60 mg of phenelzine, maximum platelet MAO-inhibition took more
than two weeks to occur (Robinson et al., l978b). We here see

116
a possible explanation for the clinical observation that full
effect of MAO-inhibitor therapy takes four to six weeks.

It has been suggested that all MAO-inhibitor treated patients

should have a platelet MAO-analysis after 3 to 4 weeks to ascertain
whether sufficient MAO-inhibition had occurred knowing the wide
interindividual variation to drugs. This has at least not yet
gained widespread acceptance, but it is evident that no patient
has had an optimal chance to respond to this class of drugs if
he has not had phenelzine 60 mg per day, isocarboxazide 30 mg
per day or other MAO-inhibitors in similar dosage, for at least
four weeks and preferably six weeks.

From a pharmacokinetic viewpoint it is important that

MAO-inhibitors partly are metabolized by acetylation. Individuals
vary so considerably in their acetylation capacity that they can be
divided into fast-acetylators and slow-acetylators. It was con-
ceivable that this might influence the clinical response to
MAO-inhibitor treatment, but investigations have failed to confirm
this hypothesis.

We now kn9w that MAO enzyme (~ono~ine ~xidase) exists in two

forms: (l) MAO type (MAO-A) (its preferred substrates include
serotonin and noradrenaline) and (2) MAO type B (MAO-B) (its
preferred substrates include 'trace' amines as phenylethylamine).
Dopamine seems to be preferentially deaminated by MAO-B at
least in man, whereas tyramine is a good substrate for both forms
of the monoamine oxidase enzyme (Kline et al., 1982).

Most of the MAO-inhibitors are nonselective, i.e., they

inhibit both MAO-A and MAQ-B, but a few are selective (Pickar
et al., 1981).

Clorgyline is a selective inhibitor of MAO-A, whereas

pargyline and deprenaline (Deprenyl) are selective inhibitors of
MAO-B (Lipper et al., 1979; Potter et al., 1982). It was hoped
that one or the other type of selective MAO-inhibitors would have
a more favourable ratio between wanted and unwanted effects, but
these expectations have as yet not been substantiated (Pickar et al.
1982) .

Adrenaline has in .recent years been recognized as a brain

neurotransmitter, quantitatively much more sparsely present in
the brain and CSF than noradrenaline.

In collaboration with Dr. Niels Juel Christensen, also in

Copenhagen, we have performed two studies of CSF-adrenaline. Both
studies showed that CSF-adrenaline was reduced by some 75 per cent
in depression and normalized on clinical recovery (Christensen
et al., 1980); and that the reduction was equally pronounced in

117
endogenous and in non-endogenous depression (Gjerris et al., 198lb).
CSF-noradrenaline showed no differences between depressed and
recovered patients and controls.

We hope in CSF-adrenaline to have found a state-dependent

marker related to the presence of depression, but not to its clinical
subtype, a marker useful in the differential diagnosis of depression
and in assessing the effect of various treatments.

The combination of state- and trait-marker will be of paramount

importance for future psychochemical and psychopharmacological
research. It will perhaps also, at long last, give us the tools in
hand to tell us which patients will respond - preferentially or
only - to MAOI's and thus make us realize a differentiation in
therapy that will further a differentiation in our understanding
of the biology of depression.

MAO-inhibitors (MAOI) were among the substances that heralded

the psychopharmacological revolution in the early 1950'es. But after
early glory their role was taken over by other drugs due to unpredict-
ability of clinical response and troublesome unwanted effects in the
form of drug-drug and food-drug interactions (White & Simpson, 1981).

Many clinicians have recognized that MAO-inhibitor treatment

was of long duration. Those patients responding to MAO-inhibitor
seemed to need it for many months, even many years. Very few studies
have dealt with this problem which is one of the more difficult, but
also more crucial in psychopharmacology. In the Psychochemistry
Out-Patient-Clinic, Rigshospitalet, Copenhagen, we have accumulated
a number of patients, who have had a MAO-inhibitor up to twenty years
(Larsen & Rafaelsen, 1980). Our study suffers from many of the
classical problems being both open and retrospective; nevertheless
the individual case records are very impressive, and certainly urge
us to continue the efforts to find the right safe, specific,
reversible MAO-inhibitor for the subgroup of depressed patients,
who can only be helped by this type of drugs.

REFERENCES

Christensen, N.J·., Vestergaard, P., S¢rensen, T. and Rafaelsen, O,J.,

1980, Cerebrospinal fluid adrenaline and noradrenaline in
depressed patients, Acta Psvchiatr Scand, 61:178.
da Prada, M., Kettler, R., Keller, H.H. and Haefely, W.E., 1983,
Neurochemical effects in vitro and in vivo of the antidepressant
Ro 11-1163, a specific and short-acting MAO-A inhibitor,
Mod Probl Pharmacopsychiatry, 19:231.
Davidson, J., McLeod, M., Law-Yone, B. and Linnoila, M., 1978,
A comparison of electroconvulsive therapy and combined
phenelzine-amitriptyline in refractory depression,
Arch Gen Psychiatry, 35:639.

118
Davidson, J.R.T., Miller, R.D., Turnbull, C.D. and Sullivan, J.L.,
1982, Atypital depression, Arch Gen Psychiatry, 39:527.
Davidson, J., Weiss, J., Sullivan, J., Turnbull, C. and Linnoila, M.,
1981, A placebo controlled evaluation of isocarboxazid in
outpatients, in:"Monoamine oxidase inhibitors", M.B.H. Youdim,
E.S. Paykel, eds., John Wiley & Sons Ltd., Chichester, 115.
Gjerris, A., Berry, D.I., Christensen, N.J. and Rafaelsen, O.J.,
1983, MAO-inhibitor revisited I: Adrenaline (A) and noradrenaline
(NA) in cerebrospinal fluid (CSF) in isocarboxazide treated rats,
VII World Congress of Psychiatry, Vienna, to be published.
Gjerris, A., Fahrenkrug, J., B¢jholm, S. and Rafaelsen, O.J., 1981,
Vasoactive intestinal polypeptide (VIP) in cerebrospinal fluid
in psychiatric disorders, in: "Biological Psychiatry", C. Perris,
G. Struwe and B. Jansson, eds., Elsevier/North-Holland Bio-
medical Press, Amsterdam, 359.
Gjerris, A., Jensen, E., Christensen, N.J. and Rafaelsen, O.J., 1981,
Adrenaline and noradrenaline in psychiatric disorders. III World
Congress of Biological Psychiatry, Stockholm; Elsevier/North-
H?lland Biomedical Press, Amsterdam, 565.
Giller, E., Bialos, D., Riddle, M., Sholomskas, A. and Harkness, L.,
1982, Monoamine oxidase inhibitor-responsive depression.
Psychiatry Res, 6:41,
Greenblatt, M., Grosser, G.H. and Wechsler, H., 1964, Differential
response of hospitalized depressed patients in somatic therapy,
Am J Psychiatry, 120:935.
Kline, N.S., Cooper, T.B., Suckow, R.F., Hallstrom, C. and
Brebbia, D.R., 1981, Protection of patients on monoamine
oxidase inhibitors against hypertensive crises,
J Clin Psvchooharmacol. 1'410.
Larsen, J.K. and Rafaelsen, O.J., 1980, Long-term treatment of
depression with isocarboxazide, Acta Psychiatr Scand, 62:456.
Lipper, S., Murphy, D.L., Slater, S. and Buchsbaum, M.S., 1979,
Comparative behavioral effects of clorgyline and pargyline
in man: A preliminary report. Psychopharmacology, 62:123.
Mountjoy, C.Q., Marshall, E.F., Campbell, I.e., Garside, R.F. and
Roth, M., 1980, Prediction of response to treatment with
phenelzine in neurotic patients, Prog Neuropsychopharmacol,
~:303-308, (Pergamon Press Ltd., GBJ.
Nies, A. and Robinson, D.S., 1981, Comparison of clinical effects
of amitriptyline and phenelzine treatmep.t, in: "Monoamine
oxidase inhibitors", M.B.H. Youdim and E.S. Paykel, eds.,
John Wiley & Sons Lted., Chichester, 141.
Pare, C.M.B., Kline, N., Hallstrom, C. and Cooper, T.B., 1982,
Will amitriptyline prevent the "cheese" reaction of
monoamine-oxidase inhibitors?, Lancet, ii:l83.
Pickar, D., Cohen, R.M., Jimerson, D.C., Lake, C.R. and Murphy, D.L.,
1981, Tyramine infusions and selective MAO inhibitor treatment:
II. Interrelationships among pressor sensitivity changes,
platelet MAO inhibition and plasma MHPG reduction,
Ps~chopharmacology, 74:8.

11 9
Pickar, D., Murphy, D.L., Cohen, R.M., Campbell, I.e. and Lipper, s.,
1982, Selective and nonselective monoamine oxidase inhibitors.
Behavioral disturbances during their administration to depressed
patients, Arch Ge~ Psychiatry, 39:535.
Potter, w.z., Murphy, D.L., Wehr, T.A., Linnoila, M. and Goodwin, F.K.,
1982, Clorgyline: a new treatment for patients with refractory
rapid-cycling disorder, Arch Gen Psychiatry_, 39:505.
Quitkin, F.M., McGrath, P., Liebowitz, M.R., Stewart, J. and Howard, A.
1981, Monoamine oxidase inhibitors in bipolar endogenous
depressives, J Clin Psychopharmacology., vol 1, 2:70.
Quitkin, F.M., Rifkin, A. and Klein D.F., 1979, Monoamine oxidase
inhibitors. A review of antidepressant effectiveness,
Arch Gen Psychiatry, 36:749.
Raskin, A., Schulterbrandt, J.G., Reatig, N., Crook, T.H. and Odle, D.,
1974, Depression subtypes and response to phenelzine, diazepam,
and a placebo, Arch Gen Psychiatry,. 30:66.
Robinson, D.S., Nies, A., Ravaris, C.L., Ives, J.O. and Bartlett, D.,
1978, Clinical pharmacology of phenelzine: MAO activity and
clinical response, in: "Psychopharmacology: A generation of
progress", M. Lipton, A. DiMascio and K.F. Killam, Raven Press,
New York, 961.
Robinson, D.S., Nies, A., Ravaris, C.L., Ives, J.O. and Bartlett, D.,
1978, Clinical pharmacology of phenelzine, Arch Gen Psychiatr~,
35:629.
Rowan, P.R., Paykel, E.S. Parker, R.R., Gatehouse, J.M. and Rao, B.M.
1981, Tricyclic anti-depressant and MAD-inhibitor: Are there
differential effects?, in: "Monoamine oxidase inhibitors",
M.B.H. Youdim and E.S. Paykel, eds., John Wiley & Sons Ltd.,
Chichester, 125.
Shepherd, M., 1965, Clinical trial of the treatment of depressive
illness, Br Med J, 1:881.
White, K. and Simpson, G., 1981, Combined MAOI-tricyclic anti-
depressant treatment: A reevaluation, J Clin Psychopharmacol,
1:264.

120
SEROTONIN IN DEPRESSION AND SUICIDE

H.M. van Praag

Department of Psychiatry, Albert Einstein College of

Medicine/Montefiore Medical Center
1300 Morris Park Avenue, Bronx, NY 10461 USA

1. RELATION TO MOOD-REGULATION TO CENTRAL SEROTONERGIC SYSTEMS

Ever since the introduction of the antidepressants in the

late fifties and the discovery that central monoamine (MA) metab-
olism is influenced by them, central MA, for example serotonin
(5-hydroxytryptamine; 5-HT) has been studied in depression. What
is the state of the art of 25 years later? Is 5-HT related to
depression, or to certain features of depression, or is it not?

I will first formulate my answer, then present the argumenta-

tion. My answer is, cautiously, in the affirmative. Suggestive
evidence indicates a relationship between 5-HT disturbances in
the central nervous system (CNS) and dysregulation of mood.
Supportive evidence is derived from patients with a primary
diagnosis of depression as well as from patients suffering from
certain somatic disorders who, in addition, show features of
depression. Both groups of data are discussed consecutively.

Finally, I will discuss the issue to what extent 5-HT distur-

bances seem to be specific for (certain types of) depression.

2. DATA DERIVED FROM PATIENTS WITH DEPRESSION AS THE PRINCIPAL

DIAGNOSIS

It is not always clear in what types of depression the biolog-

ical abnormalities were found. My impression is that most find-
ings pertain to the syndrome of vital or endogenous depression
which is rather similar to the DSM III category of Major Depres-
sive Disorder, Melancholic type. If diagnostic categorization

121
is unclear, I will use the term "certain forms of depression".

2.1 CNS studies

2.1.1. CSF 5-hydroxyindoles

In the late sixties and early seventies three research

groups independently and almost simultaneously introduced the
probenecid technique in psychiatric research in order to study
the metabolism of 5-HT in the CSN (van Praag, et al., 1970;
Tamarkin et al., 1970; Roos and Sjostrom, 1969). Probenecid
blocks the efflux of 5-hydroxyindoleacetic acid (5-HIAA) from the
CNS, including the cerebrospinal fluid (CSF), to the bloodstream.
The resulting accumulation of 5-HIAA in CSF can be considered a
crude yardstick of 5-HT degradation in the brain. (For a
description of the probenecid technique see van Praag, 1977).
Several authors have reported decreased probenecid-induced CSF
5-HIAA accumulation in a subgroup of patients suffering from the
syndrome of vital (endogenous) depression (Figure 1). This
observation can be regarded as indicative of diminished 5-HT
metabolism in the CNS. The data on baseline CSF 5-HIAA in
depression are much more controversial (Review: van Praag, 1978,
1982; Baldessarini, 1983).

2.1.2. Postmortem studies

Regional postmortem studies of CNS 5-hydroxyindoles have

been performed in suicide victims and depressed patients who died
from natural causes (Lloyd et al., 1974; Birkmayer and Riederer,
1975). Lowered concentrations of 5-HT and 5-HIAA have been found
in certain raphe nuclei, these being the principal seat of cell
bodies of serotonergic nerve fibers. The number of regional post-
mortem brain studies is too small for definitive conclusions.

2.2. Peripheral data

2.2.1. Ratio tryptophan/competing amino acids in plasma

Danish investigators have repeatedly reported a lowered

plasma ratio of tryptophan vs. competing amino acids in certain
patients with endogenous depression (M~ller et al., 197&, 1980;
M~ller and Kirk, 1981). This would give rise to diminished
tryptophan influx in the CNS and in consequence to lowered
synthesis of 5-HT (Wurtman, 1982).

These experiments have not yet been repeated. The rela-

tionship to diminished post-probenecid CSF 5-HIAA, if any, has
not been established.

122
 NUMBER OF PATIENTS
CONTROL GROUP

::r
1J
J
J
DEPRESSION GROUP

25 50 75 100 125 150 175 200 225

INCREASE IN CSF 5-HIAA AFTER PROBENECID "to ABOVE
BASELINE VALUES

Fig. 1. Post-probenecid CSF 5-HIAA accumulation in untreated

patients suffering from vital (endogenous) depression and in a
control group. CSF 5-HIAA distribution is bi-modal in the
depression group (van Praag and Westenberg, 1982).

123
 2.2.2. "'
5-HT uptake in blood platelets

Several investigators found 5-HT uptake in blood plate-

lets to be diminished in certain types of depression (Review:
Rotman, 1983). This finding is relevant since the blood platelet
is regarded as a reasonable model of a serotonergic nerve ending.

The functional significance of this phenomenon is

unknown. Several explanations are conceivable. The serotonergic
neuron could be primarily hypoactive. The decreased 5-HT uptake,
leading to increased 5-HT availability in the synapse, is a
compensatory mechanism to restore function. Another explanation
is that decreased 5-HT uptake is secondary to hypersensitivity
of post-synaptic 5-HT receptors. Increased 5-HT availability
would lead to desensitization of the receptors and possibly to
normalization of function.

2.2.3. Direct receptor studies

Some MA-related receptors are measurable in peripheral

structures, in particular in white and red blood cells. In
recent years such peripheral receptor studies have been applied
in biological depression and schizophrenia research. Very
little is known however about the functional relationship between
peripheral and corresponding central MA receptors. Are disturb-
ances in central receptors reflected in the periphery and, vice
versa, is normality or abnormality of peripheral MA receptors
indicative of their similar state in the CNS? Considerable
animal work has to be done before the "marker-value" of peripher-
al MA receptors for their central counterparts can be properly
assessed. The following data should be viewed with these restri-
ctions in mind.

Imipramine recognition sites were discovered on blood

platelets (Langer et al., 1981). These so-called imipramine
receptors are probably modulators of the 5-HT uptake sites. The
natural ligand of the imipramine receptor has not yet been
identified but promising progress has been made (Barbaccia et al.,
1983). High affinity sites for imipramine, indistinguishable
from the ones on platelets have been found on CNS serotonergic
nerve endings (Rehavi et al., 1980).

At the present time the imipramine receptor is being

intensively studied in humans. Preliminary data indicate that
in depression their activity/density might be decreased (Briley
et al., 1980). Uncertainty exists as to whether this finding
constitutes a trait or a state factor. A factor which is limit-
ing progress in this area is the rather low interlaboratory

124
reliability of the imipramine receptor assay. Two postmortem
studies of suicide victims revealed decreased imipramine recep-
tor density in the brain (Stanley et al., 1982; Perry et al., 1983).
One study failed to confirm this (Meyerson et al., 1982).

Whether an association exists between imipramine recep-

tor and 5-HT uptake in blood platelets is controversial (Wood et
al., 1983). It is also unknown in which direction 5-HT uptake
is influenced by imipramine binding sites. Henceforth, it is
impossible even to speculate on the functional significance of
the alleged decrease in imipramine binding sites in depression.

According to McBride and Mann (1983), a 5-HT? recentor

exists on blood platelets. Nothing is known on its indicator
value for the central counterpart. No studies on peripheral
5-HT2 receptors in depression have as yet been published.
Stanley and Mann (1983) found the number of 5-HT2 receptors in
frontal cortex of suicide victims increased.

2.2.4. Indirect receptor studies

One indirect 5-HT receptor test has been described, i.e.

the prolactin response to compounds increasing 5-HT availability
in the brain. Challengers that have been used are the 5-HT
precursors: tryptophan and 5-hydroxytryptophan (5-HTP) (Kato et
al., 1974; Westenberg et al., 1982; Meltzer et al., 1983;
Heninger et al., 1984) and fenfluramine, an anorectic drug
increasing 5-HT release and decreasing 5-HT reuptake (Siever et
al., 1984). Although there is considerable evidence that
increased 5-HT "tone" is a prolactin releasing factor, the
effect of 5-HT precursors on plasma prolactin in man has been
controversial (Table 1, Figure 2). Differences in dose could
have been a factor (van Praag, et al., 1984). Perhaps even more
important in this respe~t is the fact that neither tryptophan
nor 5-HTP is "5-HT-specific". Both have measurable influences
on the metabolism of CNS noradrenaline (NA) and dopamine (DA).
5-HTP increases not only the metabolism of 5-HT but also that of
DA and NA (van Praag, 1983). Large quantities of tryptophan
will reduce the influx of tyrosine in the CNS (Wurtman, 1982),
and by doing so diminish the production Date of catecholamines.
The controversial data on prolactin response to 5-HT precursors
could be related to their dual effect on 5-HT and catecholamine
metabolism.

The prolactin response to fenfluramine is described as

being blunted in depression (Siever et al., 1984) (Table 2).
These findings suggest that in those patients the 5-HT system is
sub-responsive.

125
~
0)

Table 1. Maximal baseline, baseline, and maximal treatment values of prolactin

(ng/ml) during ~-5-hydroxytryptophan (5-HTP) loading in depressive
patients and controls (Westenberg et al., 1982),

Maximal Maximal
Treatment Baseline baseline treatment
and dose Group 1 Mean SD Mean SD Mean SD
Tryptophan Patients (n=9) 8.5 1 4.2 9.7 5.3 10.4 1 4.9
(5 g) Controls (n=5) 4.0 1.3 4.5 1.6 5.2 1.2
5-HTP Patients (n=9) 8.9 5.3 10.5 5.2 10.3 5.5
(200 mg) Controls (n=5) 5.0 1.1 5.9 1.3 5.2 0.5

For baseline and maximal baseline values plasma samples were collected at -40,
-20,. and 0 minutes before administration of the amino acids. For baseline
values plasma concentrations at these times were averaged. Thereafter
samples were collected at 20-minute intervals between +20 and +180 minutes.

1. Statistically significant compared to controls (Mann-Whitney U test,

p < 0.05).
 5-HTP 200 mg p.o.

40 1

-
.-
E
C)
c 30
y -- ...............

c
......
....,
u
res
.-
0
s...
a.
res 20
E
VI
res
.-

vi
a.

10

0
0 60 120 180
Fig. 2. Effect of the oral administration of 200 mg of 5-HTP
on the plasma prolactin levels in 21 normal subjects (Mean ± SEM
shown) (Kato et al., 1974).

127
Table 2. Prolactin response to fenfluramine in normal controls and
depressed patients (Siever et al., 1984).
Plasma Prolactin {n2/ml)

Percent
Increase Increase
Age After After
{Years) Baseline Fenfluramine Fenfluramine

Normal Controls 44.0 ± 4.0 8.5 ± 1.1 17.2 ± 3.3 248 ± 58

(n=10)

All Depressed 41.5 17.8 ± 2.5* 10.9 ± 2.7 63 ± 13**

Patients (n=18)

Depressed Patients 42.8 ± 3.4 12.6 ± 1.7 6.4 ± 2.3** 58 ± 16**

(excluding those
with elevated
[>25 ng/ml) base-
lines [n=13))

Depressed Patients, 43.5 ± 3.8 15.5 ± 2.4 6.8 ± 2.9** 52 ± 20**

Age-and-sex-matched
with controls (n=10)

Baseline plasma prolactin concentrations are the means ± SEM of the

-15 min and 0 min values; the plasma prolactin increases after
fenfluramine are the differences between the baseline values, and
the means of the 3, 4 and 5 hr. values following 60 mg fenfluramine.

*p <.OS; **p <.01, Student's t-test compared to controls; Means ± SEM.

128
 30

28

Q)

~
(/)
21
---.....-- PL
18
E
p + 01
="' 15
12
(/)
<lJ A
<...
0
'-'
(/)
9 A + P + 01
<:
"'
6
Q)

::E

0~~--~--~~---.--~--~--.--------
initial 0 3 6 9 12 15 18 21 days

Fig. 3. Hamilton scores in four groups of patients with

unipolar and bipolar vital depression, before and during a
3-week period of medication with serotonin-po tentiating
compounds. The groups comprised 10 patients each, and were
treated with clomipramine alone (225 mg/day; A) with 5-HTP alone
(P; 200 mg/day, in combination with 150 mg MK 486, a peripheral
decarboxylas e inhibitor; DI), with a combination of 5-HTP and
clomipramine , and with a placebo (PL), respectively . The design
was double-blind (van Praag, 1978, 1979).

129
 2.2.5. Pharmacological strategies

In double blind, controlled studies, we have repeatedly

demonstrated that £-5-HTP is therapeutically active in patients
with the syndrome of vital (endogenous) depression (van Praag
et al., 1972; van Praag, 1978). It proved to be more active
than placebo, equivalent to the tricyclic antidepressant clom-
ipramine (Anafranil), whereas the combination clomipramine/ 5-HTP
proved to be more efficacious than either compound alone (Figure
3). 5-HTP was always administered together with a peripheral
decarboxylase inhibitor. Several groups, though not all, have
confirmed our findings (Review: van Praag, 1981). We found the
low post-propenecid CSF 5-HIAA group of vital depression to be
preferentially responsive to 5-HTP treatment. This study has not
yet been repeated.

Several independent groups reported euphoric effects of

5-HTP in normal test subjects (Trimble et al., 1975; Puhringe
et al., 1976). In myoclonus patients treated with 5-HTP, (hypo)
mania has been reported to be a frequent side effect (van Woert
et al., 1977; Thal et al., 1980).

The data on £-trvutouhan in depression are much more

equivocal. The number of studies finding antidepressant effects
almost equals the number of negative studies. However it is
fair to add that the largest tryptophan study to date found
£-tryptophan to be an effective antidepressant in depressive
out-patients seen by general practitioners (Thomson et al., 1982).
Its efficacy was equivalent to that of amitriptyline (Tryptisol)
(Figure 4). In the only double blind comuarative trvutouhan/
5-HTP depression studv conducted so far (van Praag, 1984a), 5-HTP
turned out to be significantly better than either tryptophan or
placebo (Figure 5). Tryptophan tended to be more efficacious
than placebo but not significantly so. This study was exclusive-
ly concerned with depressive in-patients whose illness was con-
siderably more severe than the patients involved in the above
mentioned tryptophan study of Thomson et al. (1982).

The reason for the discrepant findings with the two 5-HT
precursorsmaybe in their disparate effects on central catechol-
amines (CA). As stated under 2.2.4. 5-HTP increases both the
metabolism of 5-HT and of DA and NA. This has been observed in
animals and humans (van Praag, 1983). In animals large doses of
tryptophan decreases central CA metabolism by decreasing the
transport of tyrosine into the CNS, via diminishing the ratio
tyrosine/competing amino acids (Wurtman, 1982). In humans this
issue has not yet been studied. According to the monoamine
hypothesis of depression, decreasing the CA availability in the
brain constitutes a depressogenic principle. This could counter-
act the therapeutic potential of increased 5-HT availability.

130
 100

90

80
• - --<::

70 I'
,-...

--
~
'-' 60
Vl

t:
-~
C':l 50
0..
...-o

J
40 D.........--

30

20
/I
10

2 3 4 6 8 10 12
Week
Fig. 4. Comparison of treatments showing the eumulative
% of patients in each group in remission, defined as reaching
a score of 0 (not depressed) on the global rating scale, • •,
tryptohan; 0----o, amitriptyline; • •, tryptophan-amitripty-
line combination; D - - - D , placebo (Thomson, et al., 1982).

131
 z
~
....J L·TRYPTOPHAN
~
<(
J:
z<( L·5·HTP

w
:::!!:

0 3 6 9 12 15 18 21 24 27 30
DAYS OF TREATMENT

Fig. 5. Hamilton ratings over 4 weeks in a double-blind

comparative study of i-5-HTP (200 mg/day) in combination with
carbidopa (150 mg/day), i-tryptophan {5g/day) and placebo. Each
group consisted of 15 patients suffering from the syndrome of
vital (endogenous) depression (van Praag, 1984a).

132
 There has been one study published in which an inhibitor
of 5-HT synthesis was used, i.e. para-chlorophenylalanine (PCPA)
(Shopsin et al., 1979). It was reported that the antidepressant
response of tricyclic antidepressants was blocked by PCPA.

Recently a number of selective inhibitors of 5-HT uptake

have been developed. All of them possess antidepressant proper-
ties. This finding also speaks in favor of involvement of
central serotonergic functions in mood regulation.

3. DATA DERIVED FROM PATIENTS WITH SOMATIC DISORDERS AND A

CONCOMITANT DEPRESSION

3.1 Parkinson's disease

Depression is diagnosed in approximately 50% of patients

with Parkinson's disease (Mindham, 1970; Mayeux et al., 1981).
Correlation between the severity of the Parkinson syndrome and
the severity of depression is lacking. Moreover depression not
infrequently antedates Parkinson's disease. Thus, the depressive
syndrome is not likely to be solely a psychological reaction to
a disabling disease.

Hornykiewicz (Review: 1981) demonstrated Parkinson's

disease to be, at least in part, a DA-deficiency syndrome. In
addition, he found CNS 5-HT to be lowered in this disease. In
accordance with postmortem findings, the post-probenecid CSF
concentrations of homovanillic acid (HVA) - the main degradation
product of DA - and of 5-HIAA were found to be lowered in Park-
inson's disease (Lakke et al., 1972).

Recently Mayeux et al., (1984) both confirmed our find-

ings and found a significant correlation in.Parkinson patients
between low CSF 5-HIAA and the occurrence of major depressive
disorder (Figure 6). This is suggestive evidence that depression
in Parkinson patients relates to involvement of serotonergic
systems. The concentration of HVA and 3-methoxy-4-hydroxypheny l-
glycol (MHPG) - the main degradation product of NA in the CNS -
did not relate to depression.

3.2 Coeliac disease

This is a malabsorption syndrome characterized by gluten

intolerance. Hallert et al. (Review: Hallert and Sedvall, 1983)
demonstrated an increased depression morbidity in this intestinal
disorder. Using the same reasoning as mentioned under Parkinson's
disease, this finding should not be exclusively considered as a
psychological reaction to chronic somatic suffering.

133
 •
ro

60 •

-•
~

u
•

•
{ 40
!
~

•
••

-..
c
c
i
30
• T
~
.••• •
" ••
•
20

- .. f•
-----
-----
•
10

0
MD D~ ND CON
N•9 N•7 N•25 N•15

Fig. 6. CSF content of 5-HIAA in Parkinson patients accord-

ing to their psychiatric diagnosis. MD refers to major depression,
DYS to dysthymic disorder, ND to no psychiatric disorder, and CON
to the age-matched controls. The horizontal lines reflect means
for each group (Mayeux et al., 1984).

134
 In patients with this disorder, Hallert and Sedvall (1983)
found CSF 5-HIAA to be lowered (Figure 7). The concentration of HVA
and MHPG was also decreased.

At this time, it is unclear to what extent decreased CSF

5-HIAA and mood lowering interrelate. The fact that depression in
coeliac patients responded to pyridoxine speaks in favor of such
a relationship. Pyridoxine, which comes from dietary sources, is
a co-factor required for 5-HT and CA synthesis, and is poorly
absorbed in coeliac disease.

4. ARE THE 5-HT DISORDERS RELATED TO DEPRESSION, TO SUICIDE OR

TO DISTURBED REGULATION OF AGGRESSION?

Assuming now that central 5-HT metabolism can be disturbed in

depression, the question arises to what extent those disturbances
are specific for depression. I confine myself to one variable:
CSF 5-HIAA.

Biological research in psychiatry so far has a definite

nosological/syndrornal orientation. The ultimate goal is to find
the biological causes, of for example, schizophrenia, depression,
etc. Previously we have questioned this approach and pleaded
another one (van Praag et al., 1975). We then summarized the
evidence that biological variables in behavioral disorders seem
to correlate better with particular psychological dysfunctions
than with a particular syndrome or nosological entity as a whole.
Particular psychological dysfunctions are seldomly syndrornally
or nosologically specific but tend to occur in many a behavioral
disorder. Biological dysfunctions are specifically related, we
hypothesised, to particular psychological dysfunctions; they are
syndromally and nosologically non-specific. The best basis for
biological psychiatry is, we suggested, a "functional psychopath-
ology", that is a psychopathology in which psychological dys-
functions rather than syndromes or nosological entities are the
units of classification. To this end, syndromes have to be
"dissected" into their components, often measurable psychological
dysfunctions, such as disturbances in motor activity, perception,
information processing, etc. In fact this consideration consti-
tutes application of Claude Bernard's principles to psychiatry.
Collaboration between psychiatrists and experimental clinical
psychologists is a prerequisite for this venture to be successful.

Recent data regarding CSF 5-HIAA and depression seem to

support this notion. Asberg's group (Asberg et al., 1976) report-
ed that low CSF 5-HIAA in depression occurs in particular, in
those patients who have committed (violent) suicide (Figure 8).

This finding has been confirmed by several groups (Agren, 1980,

135
 E
......
0
E
-
<(
<(
g,

J:
I
II')
1.&-
V)
u

Without With
Gluten-free diet

Fig. 7. 5-HIAA in CSF of 7 adult coeliac patients before and

during treatment with a gluten-free diet. Normal range 70-140
pmol/ml (shaded area) (Hallert and Sedvall, 1983).

136
 8 Endogenous depression • violent suicide attempt
0 drug overdose
6 t suicide

4
Ill

~ 2
iiiQ.
0

•
.1:1
E 2
:I
z
4

6
Reactive depression
8

Fig. 8. Baseline CSF 5-HIAA and suicidal behavior inpatients

with endogenous and reactive depression according to the
Newcastle Inventory. CSF 5-HIAA-distribution is bi-modal in the
endogenous, not in the reactive group. There is a preponderance
of (violent) suicide attempts in the low CSF 5-HIAA probands
(Asberg et al., 1976).

137
1983; Banki et al., 1981; Montgomery and Montgomery, 1982; van
Praag, 1982; Banki and Arato, 1983) (Table 3), but not all
(Roy-Byrne, 1983). A possible explanation for this discrepancy
is that in some studies CSF was examined proximate to the
suicide, while others selected cases on life time histories of
suicide.

Subsequently Aaberg's group contended that low CSF 5-HIAA is

also present in non-psychotic, non-depressed suicide attempters.
They concluded that low CSF 5-HIAA is related to suicidal
behavior rather than to depressive behavior (Traskman et al.,
1981). This conclusion cannot be justified since we have demon-
strated that a) suicide attempts in a majority of cases are
preceded by weeks or months of overt depression (van Praag, 1982;
van Praag and Plutchik, 1984) (Table 4) and b) the pre- and post-
suicidal mood condition are not highly correlated (van Praag,
unpublished). The diagnosis of non-depressed suicide attempter
is only justified after careful exploration of the state of mood
prior to the suicide attempt.

In order to answer the question whether low CSF 5-HIAA

relates to disturbances in aggression regulation rather than to
disturbances in mood-regulation, we studied this variable in
schizophrenic patients who had attempted suicide because they
were ordered to do so '(imperative hallucinations), but who were
not depressed according to DSM III and RDC criteria (van Praag,
1983). We found post-probenecid CSF 5-HIAA to be lower in
those patients in comparison to a matched group of schizophrenic
patients without suicidal antecedents and a control group with-
out psychiatric symptoms (Figure 9). This finding support the
view that low CSF 5-HIAA and suicidal behavior are indeed rela-
ted.

Some data suggest that low CSF 5-HIAA relates to outward

directed aggression as well. Low CSF 5-HIAA depressives did
surpass their normal CSF 5-HIAA counterparts not only in number
of suicides but likewise in all signs of outward directed
aggression that were studied (van Praag, 1984b). Moreover, in
four independent studies of non-depressed patients with severe
personality disorder, a negative correlation was established
between CSF 5-HIAA and outward directed aggression (Brown et al.,
1979; Bioulac et al., 1980; Brown et al., 1982; Linnoila et al.,
1983) (Table 5) (Figure 10). One study, that of Linnoila et al.
(1983), concerned criminals who had committed or attempted
murder. Low CSF 5-HIAA was found in those who had acted impul-
sively vis a vis those who had somehow premeditated the crime.

These data make plausible the assumption that disturbed

aggression regulation (and/or disturbed impulsive control) is a
behavioral correlate of low CSF 5-HIAA. These findings do not

138
 Table 3. Incidence of suicide attempts in 203 patients hospitalized
with depressions of varying symptomatology, related to
central 5-HT disorder (van Praag, 1962).
Total Admitted after Number of violent
suicide attempts suicide attempts

Low CSF 5-HIAA level 51 24 4

Normal CSF 5-HIAA level 152 30 3

Table 4. The incidence of depressions preceding the suicide attempt

in 100 consecutive patients hospitalized after suicide
attempts (van Praag, 1982).

Number of patients Number of violent suicide attempts

Vital depression 24 7

Non-vital depression 58

Not depressivea 18 0

aAnxiety neurosis, 3; alcoholism, 4; impulsive act after acute stress, 11.

139
 A Non-psychiatric control group
B Schizophrenic without suicidal
history
C Schizophrenic with suicidal history
0 Non-violent suicide
e Violent suicide

0
0
0
0
00 8
()
8
~
0 •
•
0
0 00
oe
0 0
0 +
•
oe

Fig. 9. Post-probenecid CSF 5-HIAA in non-depressed

schizophrenic patients with and without suicidal histories and
in non-psychiatrically disturbed controls (van Praag, 1983).

140
 Table 5. Studies in which CSF 5-HIAA and agression were negatively correlated

STUDY N" DIAGNOSES MEASUREMENT ASSAY POST-PROBENECID OTHER MA

Sex AGGRESSION CSF 5-HIAA OR BASELINE 5-HIAA METABOLITES IN CSF

Brown et al. 26d Personality -Checklist Fluorometrically Baseline HVA unchanged

(1979) disorder lifetime history MHPG increased
aggressive acts

Brown et al. l2d Borderline -Checklist Mass-fragmento- Baseline HVA unchanged

(1982) personality lifetime history graphy MHPG unchanged
aggressive acts
-Buss-Durkee
Inventory
-MMPI

Bioulac et al. 6d xyy Person- -Lifetime history Fluorometrically Post-probenecid HVA unchanged
(1980) ality disorder of aggressive MHPG not measured
behavior

Linnoila et al. 36d Severe perso- -21 had killed; Liquid chroma- Baseline HVA unchanged
(1983)* nality disor- iS had made tography MHPG unchanged
ders. attempts to kill.
All were alcohol
abusers.

* Low CSF 5-HIAA in impulsive violent offenders as opposed to those who had premeditated their acts.

~
 60

0 4 8 12 16 20
Mean aggression scare

Fig. 10. The relationship of aggression scores in young

military men to CSF 5-HIAA. The subjects had various personality
disorders and difficulties adjusting to military life. Aggres-
sion scores showed a significant negative correlation with 5-HIAA
(r = -0.78) (Brown et al., 1979).

142
per se contradict a relation between low CSF 5-HIAA and mood
disorder. It is conceivable that disturbances in serotonergic
functions could give rise to both mood and aggression-disregula-
tion. Two observations speak in favor of this assumption. First,
low CSF 5-HIAA has been found in depressed patients without
suicidal history (Asberg et al., 1976; van Praag, 1982). Second,
CSF 5-HIAA values have been found to be lowest in depressed
patients who had committed violent suicide attempts (Traskman et
al., 1982). A common factor in the central regulation of mood
and aggression would provide a biological explanation for the
clinical observation that disregulation of mood and disregula-
tion of aggression frequently go hand in hand (Freud, 1916;
Gershon et al., 1968; Weissman et al., 1973; Conte and Plutchik,
1974).

5. SUMMARY AND CONCLUSION

Ever since the discovery that the classical antidepressants -

tricyclics and monoamine oxidase inhibitors - exert an influence
on central 5-HT, this ~eurotransmitter has been studied in
depression, in particular, in those forms responsive to this type
of treatment. This paper reviews the evidence in favor of a
relationship between depression and central 5-HT dysfunctions.
Most of the findings are derived from pati~nts with depression
as the principal diagnosis. Some data originate from patients
suffering from a somatic illness and from depression as well.
Both peripheral and central data are discussed. The issue of
the specificity of the observed alterations in 5-HT metabolism
is also addressed.

The following conclusions are drawn. In depression, 5-HT

metabolism can be disturbed. The same seems to be true for
certain disturbances in aggression regulation. It is more
likely that the 5-HT disturbances are causally related to the
behavioral disorder, than that they are secondary to them or
mere coincidences. This conclusion is based on the following
considerations: 1) 5-HT precursors, in particular 5-HTP can be
effective in depression (Review: van Praag, 1981), 2) 5-HTP
exerts prophylactic effects in high frequency uni- and bipolar
depression (van Praag and de Haan, 1980), 3) Preliminary find-
ings suggest 5-HT precursors, in particular tryptophan, to
possess aggression decreasing potential in schizophrenic patients
(Morand, 1983), 4) In animals the relationship between certain
types of aggressive behavior and central 5-HT systems has been
well established (Review: Valzelli, 1981).

The hypothesis is launched that disturbances in serotonergic

function could give rise to both mood- and aggression-disregula-
tion. This would provide a biological explanation for the
clinical observation that mood disorders and aggression disorders
frequently go hand in hand.

143
REFERENCES

Agren, H. (1980) Symptom patterns in unipolar and bipolar depres-

sion correlating with monoamine metabolites in the cerebrospinal
fluid. Psychiat. Res.,3:225.

Agren, H. (1983) Life at risk:markers of suicidality in depres-

sion. Psychiat, Dev.,1:87.

Asberg, M.; Traskman, L. and Thoren, P. (1976) Serotonin

depression:a biochemical subgroup within the affective disorders?
Science,131:478-480 ••

Baldessarini, R.J. (1983) Biomedical aspects of depression.

American Academic Press.

Banki, C.M. and Arato, M. (1983) Amine metabolites and neuro-

endocrine responses related to depression and suicide.
J. Affect. Disord.,5:223.

Banki, C.M.; Molnar, G. and Vojnik, M. (1981) Cerebrospinal

fluid amine metabolites, tryptophan and clinical parameters in
depression:Psychopatholo gical symptoms. J. Affect. Disord.,3:91.

Barbaccia, M.L.; Gandolfi, 0.; Chaung, D. and Costa, E. (1983)

Modulation of neuronal serotonin uptake by a putative endogenous
ligand of imipramine recognition sites. Proc. Nat'l. Acad. Sci.
USA.,80:5134-5138.

Bioulac, B.; Benezich, M.; Renaud, B.; Noel, B. and Roche, D.

(1980) Serotonergic functions in the 47, XYZ syndrome. Biol.
Psychiat.,15:917.

Birkmayer, W. and Riederer, P. (1975} Biochemical post-mortem

findings in depressed patients. J. ·Neural. Transm.,37:95-109.

Briley, M.S.; Langer, S.Z.; Raisman, R.; Sechter, D. and

Zarifian, E. (1980) Tritiated imipramine binding sites and
decreased in platelets of untreated depressed patients. Science,
209:303-305.

Brown, G.L.; Ebert, M.E.; Goyer, P.F.; Jimerson, D.C.; Klein,

W.J.; Bunney, W.E. and Goodwin, F.K. (1982) Aggression, suicide
and serontonin:relationships to CSF amine metabolites. Am. J.
Psychiat.,139:741.

Brown, G.L.; Goodwin, F.K.; Ballenger, J.C.; Goyer, P.F. and

Major, L.F. (1979) Aggression in humans correlates with cerebro-
spinal fluid metabolites. Psychiat. Res.,1:131.

144
Conte, H.R. and Plutchik, R. (1974) Personality and background
characteristics of suicidal mental patients. J. Psychiat.
Res. ,10:181.

Freud, S. (1956) Mourning and melancholia. In: London, J.E., ed.

Collected Papers, Hogarth Press.

Gershon, E.S.; Cromer, M. and Klerman, G.L. (1968) Hostility and

depression. Psychiat.,31:224.

Hallert, C. and Sedvall, G. (1983) Improvement in central mono-

amine metabolism in adult coeliac patients starting a gluten-
free diet. Psychol. Med.,13:267-271.

Heninger, G.R.; Charney, D.S. and Sternberg, D.E. (1984) Seroton-

ergic function in depression. Arch. Gen. Psychiat.,41:398-402.

Hornykiewicz, 0. (1981) Biochemical determinants of Parkinson's

disease. In: van Praag, H.M.; Lader, M.H.; Rafaelsen, O.J. and
Sachar, E.J. eds. Handbook of Biological Psychiatry Part IV.
(Marcel Dekker Inc. New York, Basel).

Kato, Y.; Nakai, Y.; Imura, H.; Chihara, K.; and Ohgo, S. (1974)
Effect of 5-Hydroxytryptophan (5-HTP) on plasma prolactin levels
in man. J. Clin. Endoc. Metab.,38:695-697.

Lakke, J.P.W.F.; Korf, J.; van Praag, H.M. and Schut, T. (1972)
Predictive value of the probenecid test for the effect of ~-dopa
therapy in Parkinson's disease. Nature New Biology,236:208-209.

Langer, S.Z. and Briley, M. (1981) High-affinity 3H-imipramine

binding; a new biological tool for studies in depression.
Trends Neurosci.,4:28.

Linnoila, M.; Virkkunen, M.; Scheinin, M.; Nuutila, A.; Rimon, R..
and Goodwin, F.K. (1983) Low cerebrospinal fluid 5-Hydroxy-
indoleacetic acid concentration differentiates impulsive from
nonimpulsive violent behavior. Life Sci.,33:2609-2614.

Lloyd, K.J.; Farley, I.J.; Deck, J.H.N. and Hornykiewicz, 0.

(1974) Serotonin and 5-Hydroxyindoleacetic acid in discrete areas
of the brainstem of suicide victims and control patients. Adv.
Biochem. Psychopharmacol.,11:387-397.

Mayeux, R.; Stern, Y.; Cote, L. and Williams, J.B.W. (In press,
1984) Altered serotonin metabolism in depressed patients with
Parkinson's disease. Neurology.

Mayeux, R.; Stern, Y.; Rosen, J. and Leventhal, J. (1981) Depres-

sion, intellectual impairment and Parkinson's disease. Neurology 1
31:645-650.

145
McBride, P.A.; Mann, J.J.; McEwen, B. and Biegon, A. (1983) Char-
acterization of serotonin binding sites on human platelets. Life
Sci.,33:2033.

Meltzer, H.Y.; Tricou, B.J.; Robertson, A. and Lowry, M. (1983)

Hormone response to 5-HTP. Psychiat. Res.,10:151.

Meyerson, L.R.; Wennogle, L.P.; Abel, M.S.; Coupet, J.; Lippa,

A.S.; Rau, C.E. and Beer, B. (1982) Human brain receptor altera-
tions in suicide victims. Pharmacol. Biochem. Behav.,17:159-163.

Mindham, R.H.S. (1970) Psychiatric syndromes in parkinsonism.

J. Neurol. Neurosurg. Psychiat.,30:188-191.

M¢ller, S.E. and Kirk, L. (1981) Decreased tryptophan availability

in endogenous depression caused by disturbed plasma leucine
clearance. Prog. Neuro Psychopharmacol.,5:277-279.

M¢ller, S.E.; Kirk, L. and Fremming, K.H. (1976) Plasma amino

acids as an index for subgroup in manic depressive psychosis:
correlation to effect of tryptophan. Psychopharmacologia,49:205-
213.

M¢ller, S.E.; Kirk, L. and Honore, P. (1980) Relationship between

plasma ratio of tryptophan to competing amino acids and the
response to £-tryptophan treatment in endogenously depressed
patients. J. Affect. Disord.,2:47-50.

Montgomery, S.A. and Montgomery, D. (1982) Pharmacological pre-

vention of suicidal behavior. J. Affect. Disord.,4:291.

Morand, C.; Young, S.N. and Ervin, F.R. (1983) Clinical response
of aggressive schizophrenics to oral tryptophan. Biol. Psychiat.,
18:575-578.

Perry, E.K.; Marshall, E.F.; Blessed, G.; Tomlinson, B.E. and

Perry, R.H. (1983) Decreased imipramine binding in the brains of
patients with depressive illness. Brit. J. Psychiat.,142:188-
192.

Puhringe, W.; Wirz-Justice, A.; Graw, P.; Lacoste, V. and Gastpar,

M. (1976) Intravenous £-5-hydroxytryptophan in normal subjects:
an interdisciplinary precursor loading study. I. Implication of
reproducible mood elevation. Pharmakopsychiat.,9:260-268.

Rhavi, M.; Paul, S.M.; Skolnick, P. and Goodwin, F.K. (1980)

Demonstration of specific high affinity binding sites for
3H-imipramine in human brain. Life Sci.,26:2273-2279.

146
Roos, B.E. and Sjostrom, R. (1969) 5-Hydroxyindoleacetic acid and
homovanillic acid levels in the cerebrospinal fluid after proben-
ecid application in patients with manic-depressive psychosis.
J. Clin, Pharmacol.,1:153-155.

Rotman, A. (1983) Blood platelets in psychopharmacological

research. Prog. Neuropsychopharmacol. and Biol. Psychiat., 7:135-
151.

Roy-Byrne, P.; Post, R.M.; Rubinow, D.R.; Linnoila, M.; Savard, R.

and Davis, D. (1983) CSF 5-HIAA and personal and family history
of sui.cide in affectively ill patients: a negative study.
Psychiat. Res.,10:263-274.

Shopsin, B.; Gershon, S.; Goldstein, S.; Friedman, M. and Wilk, S.

(1974) Use of synthesis inhibition in defining a role for biogenic
amines during imipramine treatment in depressed patients.
Psychopharmacol. Bull.,10:52.

Siever, L.J.; Murphy, D.L.; Slater, S.; de la Vega, E. and

Lipper, S. (1984) Plasma prolactin changes following fenfluramine
in depressed patients compared to controls:an evaluation of
central serotonergic responsivity in depression. Life Sci.,34:
1029-1039.

Stanley, M. and Mann, J.J. (1983) Increased serotonin-2 binding

sites in frontal cortex of suicide victims. Lancet,I:214-216.

Stanley, M.; Virgilio, J. and Gershon, S. (1982) Tritiated

imipramine binding sites are decreased in the frontal cortex of
suicides. Science,216:1337-1339.

Tamarkin, N.R.; Goodwin, F.K. and Axelrod, J. (1970) Rapid

elevation of biogenic amine metabolites in human CSF following
probenecid. Life Sci.,9:1397-1408.

Thal, L.J.; Sharpless, N.S.; Wolfson, L. and Katzman, R. (1980)

Treatment of myoclonus with ~-5-hydroxytryptophan and carbidopa.
Clinical, electrophysiological and biochemical observation.
Ann. Neurol.,7:570-576.

Thomson, J.; Rankin, H.; Ashcroft, G.W.; Yates, C.M.; McQueen,

J.K. and Cummings, S.W. (1982) The treatment of depression in
general practice:a comparison of ~-tryptophan, amitriptyline,
and a combination of ~-trytophan and amitriptyline with placebo.
Psychol.Med.,12:741-751.

Traskman, L.; Asberg, M.; Berulsson, L. and Sjostrand, L. (1981)

Monoamine metabolites in CSF and suicidal behavior. Arch. Gen.
Psychiat.,38:631-636.

147
Trimble, M.; Chadwick, D.; Reynolds, E. and Marsden, C.D. (1975)
£-5-hydroxytryptophan and mood. Lancet,i:583.

Valzelli, L. (1981) Psychobiology of aggression and violence.

Raven Press, New York.

van Praag, H.M. (In press, 1984a) Studies in the mechanism of

action of serotonin precursors in depression. Psychopharm. Bull.

van Praag, H.M. (In press, 1984b) Brain serotonin and human
(autoaggression). In: Usdin, E., Neuroscience Research Symposium,
eds. Bunney, W.E. and Barchas, J.D.

van Praag, H.M. (1983) In search of the action mechanism of anti-

depressants. 5-HTP/tyrosine mixtures in depression. Neuropharm-
acol.,22:433-440.

van Praag, H.M. (1983) CSF 5-HIAA and suicide in non-depressed

schizophrenics. Lancet,II:977-978.

van Praag, H.M. (1982) Neurotransmitters and CSN disease:

Depression. Lancet,II:1259-1264.

van Praag, H.M. (1982) Depression, suicide and the metabolism of

serotonin in the brain. J. Affect. Disord.,4:275-290.

van Praag, H.M. (1981) Management of depression with serotonin

precursors. Biol. Psychiat.,16:291-310.

van Praag, H.M. (1978) Psychotropic drugs. A guide for the

pr·actitioner. Brunner/Mazel, New York.

van Praag, H.M. (1978) Amine hypotheses of affective disorders.

In: Iversen, L.L.; Iverson, S.D. and Snyder, S.H., eds. Handbook
of Psychopharmacology:Vol 13. Biology of mood and anti-anxiety
drugs. Plenum Press, New York, Amsterdam, London.

van Praag, H.M. (1977) Depression and schizophrenia. A contribu-

tion on their chemical pathologies. Spectrum Publication, New
York.

van Praag, H.M. and de Haan, S. (1980) Depression vulnerability

and 5-hydroxytryptophan prophylaxis. Psychiat. Res.,3:75-83.

van Praag, H.M.; Korf, J.; Dols, L.C.W. and Schut, T. (1972) A
pilot study of the predictive value of the probenecid test in
application of 5-hydroxytryptophan as antidepressant. Psycho-
pharmacologia,25:14-21.

148
van Praag, H.M.; Korf, J.; Lakke, J.P.W.F. and Schut, T. (1975)
Dopamine metabolism in depressions, psychoses and parkinson's
disease:The problem of the specificity of biological variables
in behavior disorders. Psychol. Med.,5:138-146.

van Praag, H.M.; Korf, J. and Puite, J. (1970) 5-Hydroxyindolea-

cetic acid levels in the cerebrospinal fluid of depressive
patients treated with probenecid. Nature,225:1259-1260.

van Praag, H.M. and Plutchik, R. (In Press) Depression-type and

depression-severity in relation to risk of violent suicide
attempt. Psychiat. Res.

van Praag, H.M. and Westenberg, H.G.M. (1983) The treatment of

depressions with ~-5-hydroxytryptophan. In: Mendlewicz, J. and
van Praag, H.M. eds. Management of Depression With Monoamine
Precursors. S. Karger/Basel.

van Praag, H.M.; Westenberg, H.G.M.; Thyssen, J.H.H. and de Jong,

J.T.V.M. (1984) Hormone response to 5-HTP:A reply. Psychiat.
Res.,11:79.

van Woert, M.H.; Rosenbaum, D.; Howieson, J. and Bowers, M.B.

(1977) Long-term therapy and myoclonus and other neurological
disorders with ~-5-hydroxytryptophan and carbidopa. N. Engl.
J. Med.,296:70-75.

Weissman, M.; Fox, K. and Klerman, J.L. (1973) Hostility and

depression associated with suicide attempts. Am. J. Psychiat.,
130:450.

Westenberg, H.G.M.; van Praag, H.M.; de Jong, J.T.V.M. and

Thyssen, J.H.H. (1982) Postsynaptic serotonergic activity in
depressive patients:evaluation of the neuroendocrine strategy.
Psychiat. Res.,7:361-371.

Wood, P.L.; Suranyi-Cadotte, B.E.; Nair, N.P.V.; LaFaille, F. and

Schwartz, G. (1983) Lack of association between [3H]imipramine
binding sites and uptake of serotonin in control, depressed and
schizophrenic patients. Neuropharmacol.,22:1211-1214.

Wurtman, R.J. (1982) Nutrients that modify brain function. Sci.

Am.,246:42.

149
 NEW PROSPECTS IN THE TREATMENT OF DEPRESSION

Paul Turner

Department of Clinical Pharmacology

St. Bartholomew's Hospital
London, EClA ?BE

Forecasts of developments in medicine have not been particularly

successful (Turner, 1981). This is because such forecasts can only
be based on what is known in the present, and rely on the philoso~hy
expressed by Bender et al (1969) - "The future must build on the
present, and the present is merely a cumulative summary of the
past". However, while this may be an acceptable retrospective
generalisation, it is not really a satisfying basis for a detailed
predictive exercise, for while it may be possible to trace present
events to their past causes, abrupt changes in the course of events
may be brought about by the exercise of human intuition and
ingenuity - "lateral thinking" perhaps - often associated with a
considerable element of serendipity. In other words, the forecast
of new prospects in anti-depressive treatment which is to follow
must be based on a personal view of the importance of recent
developments in psychopharmacological research. It cannot take
into account abrupt changes due to momentous discoveries which
may be made at any time in the future which may rapidly revolutionise
our understanding of the pathogenesis of depression or approaches
to its treatment.

Some recent developments in psychopharmacology and basic

pharmacology which may well have a profound significance on
future anti-depressive drug treatment will be discussed in the
following pages.

a). The role of dopamine as a neurotransmitter

The emphasis in monoamine research in depressive illness has

tended to be concentrated on 5 hydroxytryptamine (5-HT) and

151
noradrenaline (NA), but interest has recently increased in the role
of dopamine (DA). Some of the evidence has been reviewed by van
Praag (1982). In some patients with depression, particularly,
where there is marked motor retardation, accumulation in the CSF
of the main metabolite of dopamine after probenecid loading is
diminished indicating reduced dopamine metabolism. The blood
platelet is considered to be a model for the 5-HT presynaptic
neurone with 5-HT transport into human platelets and 5-HT storage
in platelet granules showing striking similarities to the
analogous processes in central serotonergic neurones (Sneddon, 1973).
Although the evidence for the human platelet as a model for dopa-
minergic neurones remains controversial, it can probably be
regarded as a model for the binding of DA by amine storage granules
(Stahl and Meltzer, 1978).

Tuomisto and Tukiainen (1976) found a decreased uptake of

5-HT into blood platelets from depressed patients, compared with
controls. This was confirmed by Ehsanullah and Mulgirigama (1979),
who also demonstrated a significantly decreased uptake of DA in
these patients, suggesting an involvement of DA in depression.

Further evidence for the involvement of DA in depression is

the therapeutic effectiveness of drugs such as nomifensine and
diclofensine which, unlike earlier monoamine reuptake inhibiting
antidepressive drugs, not only influence 5-HT and/or NA reuptake,
but also have a marked effect on DA uptake (Nicholson and Turner,
1977., Heinze and Orner, 1981., Keller et al. 1982).

The suggestion that a reduction in DA activity may be

associated with motor retardation is consistent with the absence
of sedative effects with these drugs. Furthermore, Heinze and
Orner (1981) found that diclofensine was most effective in those
patients in whom depression was accompanied by severe psychomotor
retardation.

Bromocriptine, an ergot polypeptide derivative, is a central

DA receptor agonist, and has been shown, in several controlled
trials, to be an effective antidepressant (Theokar et al. 1981.,
Bouras and Bridges, 1982). As expected from its pharmacological
action, it is not sedative, but produces a relatively high
incidence of nausea and vomiting. Nausea is also a recognised
side effect of nomifensine (Brogden et al. 1979), and probably
reflects the involvement of DA in emesis in man.

Enhancement of central DA activity would appear to offer

considerable promise of antidepressive. activity, without sedation
but rather with some stimulant effects. Research into sub-
populations of central DA receptors may lead to selective agonists
or reuptake inhibitors which will permit dissociation of a
psychotherapeutic action from nausea and emetic adverse effects.

152
b). The importance of presynaptic receptor function in
regulating transmitter release

There is considerable evidence that receptors exist on the

presynaptic cell membrane which modulate transmitter release
(Langer, 1977). Presynaptic alpha adrenoceptors appear to exert
an inhibitory effect on noroadrenaline release, while presynaptic
beta receptors may facilitate its release. This provides two
mechanisms, therefore, by which synaptic monoamine concentrations
may be increased, namely presynaptic alpha receptor blockade and
presynaptic beta receptor stimulation.

Presvnantic alnha adrenocentor blockade antagonises the

negative feedback effect of presynaptic receptor stimulation by
noradrenaline. Mianserin and trazodone have several pharmacological
actions, amongst which is presynaptic alpha receptor blockade
(Norman and Burrows, 1983) which may be important in their
therapeutic effect. It is of interest that presynaptic alpha
receptor stimulating drugs such as clonidine and methyldopa can
produce depression among their adverse effects.

Presvnautic beta adrenoceptor stimulation would be expected

to exert an antidepressive effect through 1ncrease of synaptic
monoamine concentrations, and conversely, blockade of these
receptors could produce depression.

In animal studies, isoprenaline and salbutamol were found to

antagonise the hypothermia induced in mice by reserpine, oxotremorine
or apomorphine, and to potentiate the toxicity of the presynaptic
alpha receptor antagonist yohimbine (Frances et al. 1978).
Preliminary open studies with intravenous salbutamol in depressed
patients suggested that it had a rapid antidepressant action
without significant adverse effects (Simonet al. 1978), and this
was confirmed in a controlled comparison with clomipramine in which
salbutamol was found to be therapeutically effective, and more
rapid in its onset of action than clomipramine (Lecrubier et al.
1980).
In other animal studies, the effects of salbutamol,
imipramine and dexamphetamine on reserpine-induced hypothermia
were antagonised by the beta-receptor antagonist propranolol
(Souto et al. 1979). It is tempting to associate this action
with the recognised although relatively uncommon incidence of
depression in patients treated with beta-receptor blocking drugs,
particularly propranolol (Patel and Turner 1981), although there
are other possible explanations for this (see below).

c). The relationship between beta-adrenoceptorand 5-HT activitJT

Several observations have indicated a close relationship

153
between beta-adrenoceptor and 5-HT activity. Most beta-adrenoceptor
antagonists inhibit the specific binding of 5-HT to brain membranes
(Middlemiss et al. 1977), and several of them block 5-HT induced
hyperactivity syndromes 'in the rat (Grahame-Smith and Orr, 1978).
Conversely, the lipophilic beta-receptor agonist clenbuterol has
been shown to enhance 5-HT mediated behavioural responses in
experimental animals, its action being blocked by the centrally-
acting beta- 1 adrenoceptor antagonist metoprolol (Cowen et al 1982).

A better understanding of this complex relationship may lead

to new approaches, not only to antidepressive therapy, but also to
anti-psychotic and anti-migraine treatment (Turner, 1983).

d). ~han~es in recentor activity associated with increased

synaptic monoamine concentrations

A major problem with the monoamine hypothesis of depressive

illness is the dissociation in time between the immediate action
of monoamine reuptake inhibiting drugs and the observed delay in
their therapeutic effect.

Within 1-2 hours of the first dose of desipramine, there is

a marked shift to the right in the tyramine pressor-dose response
curve (Ghose et al. 1976) indicating inhibition of the noradrenaline
uptake process, but its antidepressive effect takes at least 1-2
weeks to appear. It is necessary to enquire, therefore, what
pharmacological consequences to monoamine reuptake inhibition
might better be correlated with the time course of the therapeutic
action. The answer may lie in the changes in receptor number and
activity that follow changes in synaptic transmitter concentration.
An increase in synaptic transmitter concentration or receptor
stimulation is followed by "down regulation", or a decrease in
number, of receptors on the post-synaptic membrane. In association
with these changes in receptor density, longer term changes may
occur in the intracellular concentrations of cyclic AMP. Vetulani
and Sulser (1975) showed that various antidepressant treatments,
including monoamine reuptake inhibiting drugs, reduced the
reactivity of the noradrenaline-sensitive cyclic AMP generating
system in the rat limbic forebrain, and that the time course of
this change could be related to that of the antidepressive effects
of these drugs. If a reduction in postsynaptic monoamine receptor
density is indeed an essential part of the pharmacology of anti-
depressive drug action, this suggests that in untreated depression
there may be an excess of monoamine receptor activity. This
would be difficult to demonstrate within the brains of patients
with depression, but it may be relevant that the pressor response
to tyramine was found to be greater in a population of depressed
patients compared with matched control subjects (Ghose et al. 1975),
which could reflect increased receptor activity in the peripheral
noradrenergic system in these patients. It is uncertain to what

154
extent the peripheral system reflects changes in central
monoaminergic function. Nevertheless, it appears likely that
postsynaptic changes induced by increasing synaptic monoamine
concentrations may well deserve closer investigation in the
search for better antidepressive treatment.

e). Changes in responsiveness to other neurotransmitters

following changes in monoamine ~_eceptor activity

The concept of down-regulation of postsynaptic receptors in

response to increased synaptic transmitter concentration,
although important in its own right, is even more so if
considered from the possible stand-point of "heterologous"
regulation. In a paper which has not received the attention it
deserved, Amer (1977) suggested that the antihypertensive action
of beta-adrenoceptor blocking drugs could be due to an enhancement
of vasodilator responsiveness to endogenous dilator mediators
induced by up-regulation of beta-adrenoceptors, because specific
receptors activate adenylate cyclase through coupler subunits
which may be shared by more than one receptor. Support for Amer's
hypothesis has recently come from studies in which human lymphocyte
cyclic AMP responsiveness to prostacyclin (epoprostanol) was shown
to be increased in hypertensive patients treated with the beta-
adrenoceptor blocking drugs propranolol and timolol (Lima and
Turner, 1982).

It is possible, therefore, that a down-regulation of

monoamine receptors in central neurones reduces the intracellular
cyclic AMP responsiveness to one or more neurotransmitter substances
at present unrecognised, which have a more important and primary
role in the depressive illness. Bloom (1982) has recently reached
a similar conclusion; "The monoamines may not represent the primary
source of abnormality even in disorders in which the monoamine
directed drugs can bring about partial return of function". The
last decade has seen the discovery of many putative neuro-
transmitters, including amino-acids, peptides, and prostaglandins
as well as indoles and other monoamines. A disorder of one of
these, or other as yet unrecognised transmitters, may have a
greater role in depressive illness than do the monoamines. Whether
or not this is true, the possibility underlines the importance of
research into the basic neurochemistry of normal and abnormal
behaviour. A major breakthrough in this area could lead to rapid
and very important advances in drug treatment of psychiatric
illness.

Conclusion

The so-called "second generation" antidepressive drugs, with

their more selective reuptake inhibiting activity or presynaptic
effects represent important developments in the treatment of

155
depression. However, it is probable that advances of a greater
degree will depend upon major steps forward in our understanding
of psychopharmacology, some aspects of which have been discussed
in this paper.

More than three hundred years ago, the great British diarist
Samuel Pepys recorded in his diary that "Dr. Whistler told a
pretty story related by Muffet, a good author, of Dr. Caius that
built Caius College; that, being very old, and living only at
that time upon womans milk, he, while he fed upon the milk of an
angry, fretful woman, was so himself; and then, being advised to
take it of a good natured, patient woman, he did become so, beyond
the common temper of his age 11 • Perhaps we should search more
vigorously for endogenous lipid-soluble basic compounds, with low
protein binding, such as influenced Dr. Caius' behaviour and might
be involved in the pathogenesis or treatment of depression.
Who knows!
REFERENCES
Amer, M.S., 1977. Mechanism of action of beta-blockers in
hypertension.
Biochem. Ph arm ac • 26: 171.
Bender, A.D., Strack, A.E., Elright, G.W., Haunalter, G.V. 1969 •
Delphic study examines developments in medicine.
Futures 1: 289.
Bloom, F., 1982. Neurotransmitters and CNS disease.
Lancet 2: 1382 •
Bouras, N., Bridges, P.K. l982. Bromocriptine in depression.
Curr.med.Res.Opin. 8:150.
Brogden, R.N., Heel, R.C., Speight, T.M., Avery, G.S. 1979
Nomifensine: A review of its pharmacological
properties and therapeutic efficacy in
depressive illness.
Drugs 18:1.
Cowan, P.J., Grahame-Smit~ D.G., Green, A.R., Heal, D.J., 1982.
Beta-adrenoceptor agonists enhance 5-HT
mediated behavioural responses.
Br.J.Pharmac. 76:265.
Ehsanullah, R.S.B., Mulgirigama, D.L. 1979. Decreased uptake of
5-hydroxytryptamine and dopamine into blood
platelets from depressed patients.
Br.J.Clin.Pharmac. 7, 434P.
Frances, H., Puech, A.J., Simon, P. 1978. Psychopharmacological
profiles of isoproterenol and salbutamol.
J.Pharmacol. (Paris) 9:25.
Ghose, K., Turner, P., Coppen, A. 1975. -Intravenous tyramine
pressor response in depression.
Lancet 2, 1317.
Ghose, K., Gifford, L., Turner, P., Leighton, M. 1976. Studies of
the interactions of desmethylimipramine with tyramine in man and
its correlation with the plasma concentrations.
Br.J.Clin.Pharmac. L:335·
156
Grahame-Smith, D.G., Orr, M.W. 1978. Clinical psychopharmacology.
in Recent advances in clinical pharmacology.
ed. P. Turner, D.G. Shand.
Churchill-Livingstone, Edinburgh.
Heinze, G., Orner, L.M.O. 1981. Placebo-controlled trial of
diclofensine in the sympomatic treatment of
depressive illness.
Curr.Ther.Res. 29:567.
Keller, H.M., Schaffner, R., Carruba, M.O. et al. Diclofensine-
a potent inhibitor of monoamine uptake.
Advs.Biochem.Psychopharmac. 31:249.
Langer, S.Z., 1977. Presynaptic receptors and their role in the
regulation of transmitter release.
Br.J.Pharmac. 60:481.
Lecrubier, Y., Puech, A.J., Jouvent, R., Simon, P., Widlocker, D. 1980
A beta adrenergic stimulant (salbutamol) versus
clomipramine in depression: a controlled study.
Br.J.Psychiat. 1~6:354.
Lima, D., Turner, P. 1982. Beta-blocking drugs increase
responsiveness to prostacyclin in hypertensive
patients.
Lancet 2:444.
Middlemiss. D.N., Blakeborough, L., Leather, S.R. 1977.
Direct evidence for an interaction of beta-adrenergic
blockers with 5-HT receptors·
Nature 267:289.
Nicholson, P.A., Turner, P. 1977. Proceedings of a symposium on
nomifensine.
Br.J.Clin.Pharmac. 4. 53S.
Norman, T.R., Burrows, G.D. 1983. Clinical pharmacology of some
new antidepressant drugs.in Recent advances in
clinical pharmacology. Vol.3. ed. P. Turner,
D.G. Shand.
Churchill Livingstone. Edinburgh.
Patel, L., Turner, P. 1981. Central actions of beta-adrenoceptor
blocking drugs in man.
Medicin.Res.Rev. 1:387.
van Praag, H., 1982. Depression.
Lancet 2:1259
Simon, P., Lecrubier, Y., Jouvent, R., Puech, A.J., Allilaire, J.F.,
Widlocker, D. 1978. Experimental and clinical evidence of the
antidepressant effects of a beta-adrenergic stimulant.
Psychol.Med. 8:335.
Sneddon, J.M., 1973. Blood platelets as a model for monoamine
containing neurones.
Progr.Neurobiol. 1:151.
Souto, M., Frances, H., Lecrubier, Y., Puech, A.J., Simon, P. 1979.
Antagonism by propranolol of imipramine effects
in mice.
Eur.J.Pharmac. 60:105.

157
Stahl, S.M., Meltzer, H.Y. 1978. The human platelet as a model for
the dopaminergic neurone.
Exp.Neurol. 5q:1.
Theohar, C., Fischer-Cornelss en, K.,- Brosch, H., Fischer, E.K.,
Petrovic, D. 1982. A comparative multicenter trial between
bromocriptine and amitriptyline in the
treatment of endogenous depression.
Arzneim.Forsch. ?2:783
Tuomisto, J., Tukiainen, E. 1976. Decreased uptake of
5-hydroxytryptam ine in blood platelets
from depressed patients.
Nature 262:596.
Turner, P., 1981. 1969. Delphi study of medicine.
Futures 13:221·
Turner, P. 1983. Beta-adrenocepto r blocking drugs and the
central nervous system in man. in Recent
advances in clinical pharmacology. Vol.3.
ed. P. Turner, D.G. Shand.
Churchill-Liv.ing stone. Edinburgh.
Vetulani, J., Sulser, F. 1975. Action of various anti-depressant
treatments reduces reactivity of noradrenergic
cyclic AMP-generating systems in limbic
forebrain.
Nature 251:495.

158
BIOLOGICAL HARKERS Aim ARTIDEPRESSAIIT TREATMENT RESPONSE

David G. Ostrow, Anna Okonek, Robert Gibbons,

Richard Cooper and John M. Davis

Departments of Psychiatry and

Community Health & Preventive Medicine
Northwestern University Medical School
VA Lakside Medical Center
Department of Research
Illinois State Psychiatric Institute
Chicago, Illinois

The use of biological markers to predict antidepressant

treatment response may lead to more rational selection of specific
drugs for individual patients. Research must first determine
which biological measurements have the requisite properties to
serve as reliable guides to treatment selection before those
markers can be tested for their predictive abilities. Those pro-
perties include ease of performance, reliability and reproduci-
bility of the measurement, insensitivity or predictable response
to "background" physiological variables, ability to differentiate
meaningful subgroups of depressed patients, and cost effective-
ness. In this paper we review the status of several biological
markers being developed as potential predictors of antidepressant
treatment response. Although biological marker research is pri-
marily concerned wtih predicting response to drug treatment, it
may also prove to be useful in predicting which depressed patients
may respond best to psychological or combined biological and psy-
chological treatments.

Biological Harker Properties

Assay reliability and reproducibility refer to the ability of
a particular measurement to accurately reflect the biological
characteristic of interest and the stability of that measurement
when repeated over time. These properties have been very well
159
studied for the Dexamethasone Suppression Test (DST), as pioneered
by Carroll and co-workers (1). The reliability of the DST can be
looked at in terms of its sensitivity (i.e. ability to accurately
identify endogenously depressed patients) and its specificity
(i.e. ability to exclude "false positives"). Both of these per-
formance parameters affect the number of plasma cortisol measure-
ments that must be made to provide a reliable test result. As
Carroll and co-workers have shown, two serum cortisol
measurements , at 4 and 11 pm on the day after an 11 pm one mg oral
dexamethason e dose, are able to detect melancholic depressive
patients with a sensitivity of 50-60% and a specificity of greater
than 95%. Since most depressed patients are evaluated as
outpatients in settings where 11 pm cortisol measurements are not
feasible, we have evaluated the performance of a single 4 pm
cortisol measurement in outpatients seen in the Affective
Disorders Evaluation Unit of Northwestern Memorial Hospital (Table
1). As can be seen from these data, less than 20% of depressed
outpatients had abnormal DST results (defined as 4 pm cortisol
values greater than 5 micrograms/d l). This may be the result of
the relative rarity of endogenomorp hic depressions in outpatients
as compared to our inpatient sample. However, the specificity of
the outpatient DST procedure appears to be quite good; no non-
depressed patients showed nonsuppressi on. We conclude that a
simplified 4pm DST can be used to detect a subgroup of depressed
patients, but that this subgroup represents a minority of
depressed patients presenting to an outpatient affective disorders
evaluation unit. Furthermore, non-suppress ion cannot simply be
the result of severity of depression, as Hamilton Rating Scale
(HDRS) scores did not differentiate suppressors from non-
suppressors in our studies (Table 1). Discriminant function
analysis of individual HDRS items, as shown by Drs. Nasr and
Gaviria in the accompanying paper, yielded a "profile" of HDRS
items which were positively associated with non-suppress ion.
These items (middle and late insomnia, retardation, somatic
anxiety and obsessive-com pulsive symptoms) also support the
concept of dexamethason e non-suppress ion in a sub-populatio n of
endogenous depressives.

Background physiologica l variables which may affect any bio-

logical measurement include such factors as age, race, sex, blood
pressure, body mass, diet, medications, alcohol and drug intake,
Table 1
PERFORMANCE OF INPATIENT VS. OUTPATIENT DST

Inpatient Outpatient
Total Non-Suppress ors 18/40 (45%) 18/91 (20%)
Non-Suppress ion at
11 pm only 2/18 (11%)
HDRS of Suppressors 29.4 + 11.7 25.0 + 9.3
HDRS of Non-Suppress ors 32.9 .: 8.5 28.5.: 5.7

160
smoking, and other diseases, especially chronic conditions. It
would be ideal if we could find a marker ·unaffected by any of
these factors, but this is most unlikely as biological processes
reflect the physiological state of the individual organism. We
must therefore content ourselves with determining which variables
are important to the marker and whether the effect is predic-
table. If not predictable, the resulting "exclusion criteria"
can limit the applicability of the test to a relatively small pro-
portion of the total population of depressed patients. Because we
have a major interest in the utility of ion transport measurements
in the diagnosis of major affective disorders, we have spent
considerable effort exploring the effects of numerous
physiological variables on the red blood cell lithium efflux rate
(2,3).

Figure la shows the distribution of lithium efflux values in

a sample of 129 normal control subjects without family history of
major affective illness. I t is important to screen potential
control subjects for family history of psychiatric illness, as
that variable has itself been shown to affect various biological
markers (4). As previously shown, race, diastolic blood pressure,
sex, body mass, and clinical diagnosis of hypertension, determine
at least 14% of the inter-individual variation found in this bio-
logical marker. When the data shown in Figure la are corrected
for the predictable effects of these variables they assume the
unimodal normal distribution shown in Figure lb. This finding
tells us that we must control for at least these five variables in
any studies of lithium efflux in major affective disorder
patients. Furthermore, similar multimodal distributions of other
biological measurements in psyhciatric populations serve to
underscore the importance of analyzing for physiological effects
in marker research and clinical application.

Ability to differentiate aeaningful subgroups of depressed

patients is perhaps the most important property of a biological
marker that predicts antidepressant treatment response.
Depending on the types of depressed patients being studied, the
antidepressant treatments under investigation, and the theoretical
basis of the investigation, one or more of the following subtypes
may be of interest - unipolar vs. bipolar depressive; endogenous
or melancholic vs. non-endogenous or reactive; norepinephrine vs.
serotonergic deficient; psychotic vs. non-psychotic; agitated vs.
retarded; early age of onset vs. late onset; family history posi-
tive vs. negative.

When we examine lithiUIIl efflux in a large group of patients

with major depressive disorders, we again see a complex trimodal
distribution pattern (Figure 2a). However, in contrast to the
control group, when we correct the patient values for the effects
of race, sex, blood pressure, body mass, and clinical

161
 Table 2
CORRECTED LITHIUM KFFLOX LEVELS IN UNIPOLAB. Aim BIPOLAB. PATIENTS

Diagnostic Group Less Than Predicted More Than Predicted

Unipolar 18 32
Bipolar 26 9
44 41
Chi-square = 10.6, p <0.001

hypertension, we do not obtain a simple unimodal distribution

(Figure 2b). Analysis of the distribution of bipolar and unipolar
patients' corrected lithium efflux values shows that the majority
of the bipolar patients have lower than expected lithium efflux
rates and the majority of the unipolar patients have greater than
expected lithium efflux rates (Table 2). Thus, lithium efflux
measurements can be used in the aggregate to subdivide depressive
disorder patients according to their clinical polarity.

Two biological markers that have been shown to predict the

type or class of antidepressant therapy response are urinary MHPG
and amphetamine challenge test response. Schildkraut and
colleagues (5) have shown that there are three subgroups of de-
pressed patients on the basis of 24 hour urinary MHPG excretion.
Those with low MHPG excretion rates (less than 2 mg/24 hour)
apppear to respond preferentially to the noradrenergic tricyclics,
such as 1m1pramine, and the tetracyclic, maptroptiline.
Similarly, Fawcett and associates (6) have shown that those de-
pressed patients having a marked antidepressant response to low
dose d-amphetamine administration respond to noradrenergic
compounds, such as imipramine and maptroptiline, and also to the
serotonergic antidepressant, desipramine. The latter investigator
feels this indicates that the difference in tricyclic responsivity
seen in depressed patients may not be the result of differences in
their relative abilities to block NE vs. SHT reuptake systems. An
alternative explanation may be found in the recent work of
Richelson and colleagues. Their studies of the receptor binding
properties of antidepressant compounds reveal a predominantly
adrenergic profile for all of the currently available
antidepressant compounds (7). A true test of the adrenergic/-
serotonergic classification of depression would require the test-
ing of a relatively pure serotonergic compound, such as
fluoxetine, for specificity of treatment response.

Discussion
Each biological marker has a particular set of properties
that may be used to subgroup depressed patients according to
different diagnostic schemata. Since response to a particular
antidepressant treatment may vary among different subpopulations,
it may be more helpful to measure multiple biological parameters
to derive a multi-dimensional view of the depressed patient.

162
 24. ~ 1
(N z 12 7 )
18

-0.4 1.4 3.2 4 .9 &.7 8 .$ 10.3 12 .0 13.8

]~ -1 . 3 0 .2 1.8 3.3 4 .8 6.3
.ueq Ll .. /1. cell• 111 mln.
7 .9 8 .4 10. 8

Fig. lA. Control Sample - Lithium Fig. lB. Control Sample - Lithium
Efflux (Raw Va+ues) Efflux (Adjusted for
Race, Sex, BP, Hyper and
B:HI)

.
12 .
8. (N• 86)

...,.. ..
D.
.

.....
u u
< c
6. .
~
3. 2.

0.0 1.7 3.3 5 .0 • .• 8 .3 10.0 1 1.6 13.3

,Ueq LI+/ J, C:ellt X m in.

Fig. 2A. Affective Disorders Fig. 2B. Affective Disorders

Sample - Lithium Efflux Sample - Lithium Efflux
(Raw Values) (Adjusted for Race, Sex,
BP, Hyper and B:HI)

ez. o

10.0
&1. 5

F<;oE.Q

..
7.5
a:l
.1. 1.0

A I
..as
~
~ 5 .0
i/
•'
\
;~
""· 5 II :IE
'/ 2.5
I j}
..• •• 5 7. 7 10.0 0
2.5 5.0 7.5 10.0
14o\RKEA

Marker A
Fig. 3A. A Mixture of Two Normal Fig. 3B. A Mixture of Four Normal
Distributions Distributions

163
Coupled with clinical and genetic subdivisions, this approach may
provide a classification of depression for the systematic
examination of treatment response prediction. The cost of a
multivariable biological evaluation might range from $150 to $350
per patient. This would seem to be a modest expense if we can
shorten the period needed to treat a depressed patient by one or
more weeks.

Another way to visualize the subdivision of depressed

patients by using multiple biological markers is with
multidimensional graphics. If a heterogeneous population of de-
pressed patients composed of four clinically important subtypes,
is examined using only one biological marker, only two subgroups
might be apparent (Figur~ 3a). Examination of another biological
marker may likewise reveal only two subgroups. But the
simultaneous examination of both markers has the potential for
revealing the presence of all four subgroups (Figure 3b). The
division of patient populations into biologically homogeneous
subgroups may have treatment predictive capabilities far in excess
of that provided by one measurement alone. While such potential
uses of multiple biological markers are still only hypothetical,
studies currently underway may soon substantiate this approach to
the biological characterization of depressed patients.

UFERERCES

1. B. J. Carroll, M. Feinberg, J. F. Greden, et al., A specific

laboratory test for the diagnosis of melancholia, Arch.
Gen. Psychiatry 38:15-22 (1981).
2. D. G. Ostrow, G. N. Pandey, J. M. Davis, S. N. Hurt, and
D. c. Tosteson, A heritable disorder of lithium transport
in erythrocytes of a subpopulation of manic-depressive
patients, Am. J. Psvchiatrv 135:1070-1078 (1978).
3. D. G. Ostrow and J. M. Davis, Lithium transport studies in
affective disorders, in: "Biological Psychiatry Today,"
J. Obiols, E. G. Monclus, and J. Pujol, eds., Elsevier,
Amsterdam, (1979).
4. M. Baron, M. Levitt, and R. Perlman, Platelet monoamine
oxidase values and genetic heterogeneity in
schizophrenia research, J.A.M.A 246:1418-1421 (1981).
5. J. J. Schildkraut, P. J. Orsulak, A. F. Shatzberg, et al.,
Toward a biochemical classification of depressive
disorders, I: Differences in urinary MHPG and other cate-
cholamine metabolites in clinically defined subtypes of
depressions, Arch. Gen. Psvchiatrv 35:1427-1433 (1978).
6. H. c. Sabelli, J. Fawcett, J. L. Javaid, and S. Bagri, The
methylphenidate test for differentiating desipramine re-
sponsive from nortriptyline-responsive depression, Am. J.
~sychiatrv 140:212-214 (1983).
7. E. K1chelson, Are receptor studies useful for clinical
practice?, J. Clin. Psychiatrv 44 (1983, in press).

164
EEG SLEEP IN BORDERLINE PERSONALITY DISORDER

Henry w. Lahmeyer, Eduardo Val, F. Moises Gaviria

and Bhawani R. Prasad

University of Illinois at. Chicago

College of Medicine
Department of Psychiatry
P.O. Box 6998 Chicago, IL 60680

Sleep disturbances are well known concomitants of affective

disorders. Objective EEG studies have repeatedly confirmed that
many arousals occur and less slow-wave-sleep occurs. More specif-
ically, major depression is associated with rapid eye movement
(REM) sleep disturbances. Most studies have documented three REM
sleep abnormalities in depression. First, the number of minutes
from sleep onset to the first REM period is less than usual (short-
ened REM latency), the first REM period lasts longer than usual,
and the number of phasic rapid eye movements is much greater than
usual during REM especially during the first REM period of the
night.

These findings first described in adults with depression have

now been found in some patients with depression and concomitant
borderline personality(l),(2), in some patients with only border-
line personality(3),(4), and in depressed adolescents(5). The
REM abnormalities are, therefore, not specific to depressed adults
but may reflect a common underlying trait characteristic. The
degree to which the state of depression contributes in these condi-
tions is unknown but deserves study especially in borderline
patients because of the frequent occurrence of mood disorders in
these patients. Many of the studies of borderline patients, how-
ever, have not measured all the sleep parameters, have not measured
other biological markers, and have used incomplete psychiatric
evaluation to arrive at the diagnosis of either borderline person-
ality or major affective disorder.

In this study we compare the sleep architecture of three

165
independent groups. Group 1: pure borderline personality; Group
2: patients with coesixting borderline personality and major
affective disorder; Group 3: patients with major affective dis-
order only in the depressive phase.
Methods
Sixteen borderline patients, ages 29-46 (X= 29.3 7) , 6 male, 10
female, were identified after complete psychiatric evaluation --
including two-hour diagnostic interview, to make the presumptive
diagnosis of borderline personality disorder. This presumptive
diagnosis was further refined by the administration of the Diag-
nostic Interview for Borderlines (DIB) (6). Only those patients
with DIB ratings of 6 or greater were included in this study. Then
the Diagnostic Interview Schedule (DIS) (7) was administered to
determine DSM III diagnoses.

The Hamilton Rating Scale for Depression (8) was used to rate
severity of depression and the Carroll Depression Rating Scale (9)
was filled out by all subjects to obtain a subjective quantitative
confirmation of depression severity.

Sleep records were hand scored according to the methods of

Rechtschaffen and Kales (10) but sleep onset and REM latency, REM
density, and REM period length were scored according to the
criteria of Taska and Kupfer (11). In addition to these sixteen
patients, sixteen unipolar depressed patients were studied in an
identical fashion.

All subjects spent two or three nights in the sleep laboratory

after a medication-free period of at least 10 days. All parameters
were averaged for the two or three nights before further statistical
analysis. Insignificant "first night effect" is present in de-
pressed patients so this night was included in our data analysis.
Results
Of the original sixteen patients who met criteria for border-
line personality. twelve also had a history of major affective dis-
order and four did not. Seven had Hamilton Depression scores over
20. Of the twelve patients with major affective disorders, five
were given the diagnosis by the DIS of bipolar affective disorder
and seven the diagnosis of major depression.

REM latency for borderlines was 68.0 ± 21.2 and no~ signifi-
cantly different from the major depressive group: 63,4±14.7. REM
density for the borderline group was 1.54 ± ,65. not significantly
different from the major depressive group: x = 1. 40 ± • 36. However,
REM density for the first REM period was significantly higher in the
borderlines, x x
= 1. 21 ± • 86 than the major depressives, = , 726
± .45, ( t = 6.4, p < .01).

166
 Table I. Borderline Personality Disorder

REM Densitya REM Density REM

(total) 1st REM period latencyc

All borderline
cases (n = 16) 1.54 ± .65 1.21 ± .86 68.0 ± 21.2

Coexistence with
Major Affective
Disorder(n = 12) 1.46 ± .78 1.16 ± .90 75.4 ± 30.5

Pure Borderlines
(4 = 4) 1.52 ± .44 1.17 ± .78 63.4 ± 14.7

Major Affective
Disorder (Control
Group) (n = 16) 1.40 ± .36 .726 ± .45b 66.9 ± 26.4
a One way ANOVA N.S.
b First REM period density for the major affective disorder
group significantly less than for the borderline cases
(t= 6.4, p < .01)
c One way ANOVA for all groups N.S.

In comparing the borderlines who also had the diagnosis of

major affective disorder and those without major affective disorder
there was no significant difference in average REM latency, REM
density, OR REM density in the first REM period. However, if
the group was divided into those patients with moderately severe
depressions (Hamilton>25) and mildly depressed (Hamilton<l3) REM
density in the 1st REM period discriminated borderlines with and
without the state of depression. REM density in the first REM
period was defined as abnormal if greater than .6. Using this
definition REM density was elevated during the 1st REM period in
100% of cases where Hamilton's were greater than 25 and only 20%
of cases with Hamilton's less than 13.

Table II. Borderline Personality Disorder REM Latency

Short Using Rule of 90 (Age + REM - L < 90)
All Hamil tons Hamil ton 725
All borderline cases 9/16 56% 4/7 = 57%
Coexistance with major 6/12 50% 4/6 67%
affective disorder
Pure borderlines 3/4 75% 0/1 = 0%
x2N.S.

167
 When individual cases were examined (Table II) REM latency
was short by the rule of 90 (age + REM latency 90) in 55% of
borderlines with major affective disorder and in 67% of the same
subjects with Hamilton Depression ratings greater than 25. The
presence of significant depression appears to have less effect on
REM latency than REM density in these patients.

Discussion

Yerevanian et al.(4) also found reduced REM latency in 24

outpatients who met DSM III criteria for borderline disorder
(x=63 min.). Soloff et al(1) studied a group of ten inpatients
who met Gunderson's criteria for borderline. In this study the
mean Hamilton score (17 items) was 19.0. The mean REM latency was
54.3 ± 12.5 not significantly different from primary depressives
(x =48.8 ± 22.6). These findings are in essential agreement with
McNamara et al(2).

Although our REM latency findings are closer to the Akiskal

findings than Soloff data, several reasons for these differences
are apparent. The majority of our·patients were outpatients at the
time of the study. We averaged the REM latencies over two to three
nights. Only night two was used in the Soloff study. We also had
a young population; average age = 29.3. REM latencies are no longer
in young patients in general. Because of this, we prefer to use
the rule of 90 (age+ REM latency less than 90). So that age is
controlled for. Elevated REM density in the first REM in our Border-
line patients was related more to elevated Hamilton scores than to
the presence of a lifetime diagnosis of major affective disorder.
Previous studies have not investigated this association. In fact,
previous studies have emphasized REM density as a more stable trait
characteristic than REM latency often remaining elevated even during
clinical remission.(12) Preview studies have also emphasized REM
latency as a measure of depression severity(13,(14) while this
small group of patients demonstrated only a trend in that direction.

Elevated REM density in the first REM period has been shown to
be associated with poor responsiveness to tricyclic antidepressants.
(15) This may partly explain the fact that borderline patients are
often poorly responsive to mood altering agents.
Summarv
The EEG sleep architecture of 16 borderline patients was com-
pared to the architecture of 16 patients with major affective dis-
order (depressive phase) REM latency and overall REM density was
the same for both groups but REM density in the first REM period was
significantly greater in the first REM period for the borderline
group esp~cially for those borderline patients with moderately
severe depressions.

168
REFERENCES

1. Soloff, P.H., George, A.W.A., Nathan, R.S., Spiker, D.G.:

Pharmacotherapy of borderlines: An empirical trial; Paper
presented to the American Psychiatric Association, New York
May 1983.

2. McNamara, M.E., Reynolds, C.F. III, Soloff, P.H. Mathias, R.J.,

Rossi, A.J., Spiker, D.G., Coble, P.A., Kupfer, D.J.: Electro-
encephalographic (EEG) sleep findings in borderline personality
disorder; Paper presented to Association for the Psychophysia
ologic Study of Sleep, San Antonio, Texas, June 1982.

3. Akiskal, H.S.; Factors associated with incomplete recovery in

primary depression; J.Clin.Psychiatry 43:266-271, 1982.

4. Yerevanian, B.I., Akiskal, H.S., King, D., Davis, G.C.; REM

latency in borderline disorders; Paper presented American
Psychiatric Association, New York, May 1983.

5. Lahmeyer, H.W., Poznanski, E.O., Bellur, S.N.; EEG sleep in

depressed adolescents, Am.J.Psychiatry (to be published Sept. '83h

6. Gunderson, J.G., Kolb, J.E., Austin, V; The diagnostic inter-

view for borderline patients; Am.J.Psychiatry 138:896-903,1981.

7. Robins, L.N., Helzer, J.E. Croughan, J., Ratcliff, S; National

Institute of Mental Health: Diagnostic interview schedule;
Arch.Gen.Psychiatry 38:381-389, 1981.

8. Hamilton, M.; A rating scale for depression; J.Neurol.Neurosurg.•

Psychiatry 23:56-62, 1960.

9. Carroll, B.J., Feinberg, M., Smouse, P.E., Rawson, S.G. Greden,

J.F.; The Carroll rating scale for depression: I. Development,
Reliability and Evaluation, Br.J.Psychiatry 138:194-200, 1981.

10. Rechtschaffen, A., Kales, A.; A manual of standardized termi-

nology techniques and scoring system for sleep: Stages of
human subjects., Brain Information Services, UCLA 1968.

11. Taska, L.S., Kupfer, D.J.; Sleep scoring manual from Sle~
Center Western Psychiatric Institute and Clinic, University
of Pittsburgh School of Medicine, 1982.

12. Schulz, H., Trojan, B; A comparison of eye movement density in

normal subjects and in depressed patients before and after
remission, Sleep Research 8:49, 1973.

169
13. Kupfer, D.J., Foster, F.G.; EEG sleep and depression in
Williams, R.L., and Karacan (eds.) Sleep Disorders Diagnosis
and Treatment,, John Wiley & Sons, New York, 1978.

14. Cartwright, R.D., Rapid eye movement sleep characteristics

during and after mood disturbing events; Arch.Gen.Psychiatry
40: 197-201, 1983.

15. McPartland, R.J., Kupfer, D.J., Coble, P.A., Shaw, D.H.,

Spiker, D.G.; An automated analysis of REM sleep in primary
depression; Biol.P~ychiatry 14:767-776, 1982.

170
AFFECTIVE DISORDERS AND BORDERLINE PERSONALITY

Eduardo Val, F. Moises Gaviria, Henry W. Lahmeyer,

Bhawani Prasad and Martin Weiler

University of Illinois at Chicago

College of Medicine, Department of Psychiatry
P.O. Box 6998 Chicago, IL 60680

While the term borderline has been used in the literature with
certain degree of consistency to signify a particular type of dis-
order placed somewhere between psychosis and neurotic disturbances
the clinical characteristics and identifying criteria used have
varied significantly from author to author and with different
degrees of descriptive elements and levels of abstraction.

In summary, the conceptual usage of the term borderline encom-

passes a) a descriptive definition based on characteristic symptoms
and anamnestic data (Grinker 1 , Gunderson 2 , 3 , 4 , DSM 111 5 ; b)
a level of functioning using descriptive and psychoanalytic con-
cepts (Kernberg 6 , 7 ); c) a more abstract operationally defined entity
based on psychoanalytic concept of arrestation and individuation -
separation dynamics (Mahler 8 , }1asterson 9 ); d) a stable charactero-
logical organization resulting from lack of self-cohesion eventually
diagnosable after optimal therapeutic trial (Kohut 10 ). Also the
view that borderline disorders are atypical or subclinical forms of
classical psychiatric disorders with genetic-familial transmission
have been sponsored, eg; schizophrenic spectrum (Kety 11 ); schizo-
typal personality (Spitzer 12 ); subaffective disorders (Akiskal u)
or borderline syndromes (Stone 14 ).

The evidence for a possible relationship between affective

disorders and borderline disorders comes from both sides of the
interactional equation. From the perspective of affective disorders
Gaviria, Flaherty and Val 15 have demonstrated the coexistence of
affective disorder with underlying personality with borderline and
without borderline characteristicsr Our findings indicate that
bipolars with borderline personality behave differently in a number

171
of areas: bipolars with borderline personality had high frequency
of childhood pathology, poor school performances, had significantly
worse social functioning between episodes, were proned to an earlier
onset of their affective episode and tended to report more psychotic
symptomatology. From the side of the borderline personality, the
evidence for an affinity with affective disorder have been sug-
gested and shown in several areas by diverse workers; Stone found
a greater prevalence of affective disorders in relatives of border-
line patients diagnosed by Kernberg's criteria in comparison to
neurotic p~tients 14 ; also Stone el al found that borderline and
psychotic probands had significantly more first degree relatives
affected with psychiatric illness than control subjects, with more
representatons of manic-depressive disorders in borderline patients.
Akiskal studying 100 consecutive admitted patients diagnosed as
borderline found that history and follow up indicated a strong
association between borderline and affective disorders. 2

Carroll reported that 62% of patients referred by clinicians

as having borderline diagnosis were non suppressors in the dexa-
methasone suppression test, findings similar to patients with major
depression 16 • Akiskal 13 as well as Kupfer et al 17 found that
borderline patients had shortened REM latency similar to affective
disturbances.

However, the studies reviewed are limited by the lack of

utilization of rigorous diagnostic criteria and the use of reliable
instruments in a systematic way within the same index sample. For
instance in some studies the presence or absence of DSM III diag-
nosis were arrived at retrospectively from chart reviewing, raising
questions to the reliability of their findings. In other studies
the borderline diagnosis was left to the criteria of the referring
source. In some other instances the comparison between DST results
and EEG sleep studies were derived from different samples of
patients. Thus a need exists to study borderline patients using
reliable diagnostic instruments and external validation measures
centered around each individual case, and for taking into consider-
ation also the necessity for excluding from the study sample those
cases in which the diagnosis of affective disorders has been clearly
arrived at and coexisting with a possible borderline personality,
thus eliminating or minimizing the affective disorder side of the
interactive equation between affective disorders and borderline
personality.

The study design presented here aims at resolving some of the

previously mentioned shortcomings. Patients referred to the pro-
ject with a presumptive diagnosis of borderline are screened out
by the principal investigator, who makes certain that the referring
source had not diagnosed any major psychiatric disorder, in partic-
ular any affective disorder coexisting or grafted in a borderline
personality. The subjects accepted to enter the protocol are

172
further screened by the Diagnostic interview for Borderlines (DIB).
A DIB~ 6 allows the subject to become an index case. Subsequently
each subject is assessed by means of the Diagnostic Interview
Schedule (DIS) 18 , Hamilton questionnaire, the Luria-Nebraska
Neuropsychological test and a series of social support measuring
scales. In addition the subjects are evaluated through the DST,
Sleep EEG, MA0-1 and Lithium Transport Ratio. We will report the
findings on the DIS, Hamilton and DST on 18 completed cases.

Findings: To the present 28 subjects referred to the

project. Twenty-one were found to have a DIB~ 6; of the 21, 3
dropped out giving an end of 18 patients. According to the DIS
evaluation 12 had a diagnosis of major affective disorder (MAD):
2 of schizophrenia and 4 had no major primary affective disorder
(NMAD). That is 66.6% had a diagnosis of MAD and only 22.2% had
NMAD. If we exclude the 2 schizophrenia diagnoses, the proportion
increases to 75% with MAD versus 25% without MAD. Of the 12
borderlines with MAD, 5 had bipolar or atypical bipolar diagnoses
(41.6%), and 7 major depression diagnoses (58.3%). The mean
Hamilton scores for the borderline personality with MAD was 21.9
versus 20.7 for the borderlines without MAD- essentially the
same, but indicating that both groups are at least moderately
depressed.

According to the DIS each case carries multiple diagnoses with

a mean of 6.5 diagnoses per case.

Antisocial personality diagnoses appeared in 41% of the total

cases and dysthymic disorders were obtained in 61% of the total
cases. (66% of the borderline personalities with MAD and in 50%
of the borderline personalities without MAD) DST finding shows
that 19% of all 16 cases were nonsuppressor. Twenty-five percent
of the borderlines with MAD were nonsuppressor whereas the 4
borderlines without MAD had normal suppression.

Discussion: This very preliminary report fails to support,

inspite of the clear prevalence of affective symptomatology, any
majcr role for endogenous type of depressive phenomena. Our
findings are in clear contrast with those of Carroll (62% of NS)
and in agreement with those Soloff et al 17 which found a 16% of
NS in a sample of 19 borderlines diagnosed by the DIB. Despite
the possible artifact of the tendency towards overinclusivenes of
affective disorders by the DIS instrument, the finding of 66% of
our sample carrying a diagnosis of MAD is striking, especially in
view of our efforts to exclude patients with affective disorders
from the sample.

Phenomenologically our findings are in clear contrast with

Pope, Gunderson et al in which borderlines with MAD and without
MAD were almost evenly distributed 19 • Furthermore, they found only

173
one case with dysthymic disorder and one with agoraphobia. In their
sample the concomitant diagnosis of antisocial personality was
present in 9% of those cases. In contrast we found that 41% were
associated with diagnoses of antisocial personality; 38% qualified
for agoraphobia, and 61% for dysthymic disorder.

Keeping in mind the small nature of the sample the DIS findings
clearly indicate the presence of affective symptomatology across the
entire spectruP1, even when correcting for the tendency to overinclu-
siveness by the DIS. The lack of greater percentage of nonsupres-
sors may be explained on the basis of an absence of biological
affinity with primary affective disorders or on the basis that DST
is not as specific of a test as claimed. We are more inclined to
believe the latter. At any rate, the present state of the art
calls for the clinician to rely on careful and detailed longitudinal
anamnestic data to ascertain the presence of affective symptoms,
their length and duration since the affective disturbance appears
to be disguised and obscured by the characterological matrix when
more specific and highly structured clinical instruments are not
utilized.

From the research perspective future treatment outcome studies

will have to take into consideration the two subgroups with major
and without MAD for their double blind studies in order to increase
the treatment specificity. In particular the use of Lithium, which
so far has not been tested as much as antidepressants, or antipsy-
chotic agents 19 , since at least in our sample, a bipolar syndrome
appears more prevalent than reported by others 20 Also, long-
term followup and re-evaluation of the borderline pathology after
or during treatment interventions over long periods of time might
yield some evidence toward whether the borderline function is more
state than trait related.

The clear presence of a variety of multiple disorders, with

predominance of affective ones among them on Axis I suggests that
the clinician should take a rather cautious approach and avoid any
premature closing by arriving at the diagnosis of borderline
personality to the expense of overlooking possible or atypically
manifested Axis II disorders. From a clinical perspective, giving
the present status of knowledge, the concept of a borderline
behavioral disorder is perhaps a more functional and appropriate
one since it encompasses a variety of symptoms and behaviors
presumptive of, but not necessarily of, a borderline personality
condition. Indeed "pure" borderline personality seems to consti-
tute a small minority within the overall group displaying border-
line behavior.

174
 REFERENCES

1. Grinker, R.R., Werble, R., and Drye, R.C.: The Borderline

Syndrome, New York, Basic Books, 1968.

2. Gunderson, J.G., and Singer, M.T.: Defining borderline

patients: an overview. Amer J Psychiat, 132:1-10, 1975.

3. Gunderson, J.G., and Kolb, J.E.: Discriminating features of

borderline patients. Amer J Psychiat, 135:792-796, 1978.

4. Gunderson, J.G., Kolb, J.E., and Austin, V: The diagnostic

interview for borderline patients. Amer J Psychiat, 138:7:
896-903, 1981.

5. Diagnostic and Statistical Manual of Mental Disorders,

American Psychiatric Association, 1980.

6. Kernberg, O.F.: Borderline personality organization, JAm

Psychoanal Assoc, 15:641, 1967.

7. Kernberg, O.F.: The structural diagnosis of borderline person-

ality organization. (1977) in Borderline Personality Disorders,
Harticollis, P (Ed), International Universities Press, New York.

8. Mahler, M.S.: A study of the separation phenomena in the

psychoanalytic situa.tiol)., Psychoanalytic Study of the Child,
26403, 1971.

9. Masterson, J.F.: Treatment of the adolescent with borderline

syndrome: a problem in separation-individuation, Bull,
Menninger Clin, 355, 1971.

10. Kohut, H: The Restoration of the Self, International Univer-

sity Press, New York, 1977.

11. Kety, S., Rosenthal, D., Wender, P. and Schulsinger, F: Mental

illness in the biological and adoptive families of adopted
schtzophrenics. Am J Psychiat, 127:302, 1971.

12. Spitzer, R.L, Endicott, I. and Gibbon, M: Crossing the border

into borderline personality and borderline schizophrenia. Arch
Gen Psychiat, 36:17, 1979.

13. Akiskal, H.S., Rosenthal, T.L., Haykal, R.F., et al: Character-

ological depressions. Clinical and sleep EEG findings separa-
ting "subaffective dysthymias" from "character spectrum
disorders." Arch Gen Psychiat, 37:777-783, 1980.

175
14. Stone, M.H.: The Borderline Syndromes Constitution, Personality
and Adaptation, New York, 19~0.

15. Gaviria, M., Flaherty, J. and Val, E: A comparison of bipolar

patients with and without a borderline personality disorder.
The Psychiatric Journal of Univ of Ottawa. 7:191, 1982.

16. Carroll, B., Greden, J., Feinberg, M., Lohr, Mel, James, N.,
Steiner, M., Haskett, R., Abala, A., de Vigne, J. and Tapika,
J: Neuroendocrine evaluation of depression in borderline
patients. The Psychiatric Clin. North Am, 2:89, 1981.

17. Soloff, P., Anselm, G., Swami Nathan, Reynold, C., McNamara, E,,
and Kupfer, D.: Laboratory studies of borderline disorders,
Svllabus and Scientific proceedings. Am Psychiat Ass~, 155,
1983.

18. Robins, L.N., Helzer, J.E., Croughan, J. and Ratcliff, S.:

National Institute of Mental Health diagnostic interview
schedule. Arch Gen Psychiatry, 38:381- , (1981.)

19. Soloff, P: Pharmacotherapy of borderline disorders. Compre-

hensive Psychiatry_, 22:535, 1981.

20. Pope, H., Jonas, J., Hudson, J., Cohen, B. and Gunderson, J:
The validity of DSM III borderline personality disorder, Arch
Gen Psychiat, 40:23, 1982.

176
SOCIAL ASPECTS OF DEPRESSION

Joseph A. Flaherty and F. Moises Gaviria

University of Illinois at Chicago

College of Medicine, Department of Psychiatry
P.O. Box 6998 Chicago, IL 60680

Over the last ten years, American psychiatry has witnessed a

remarkable rise in interest in the social aspects of depression in
particular. A key factor in this rise has been better agreement on
research criteria for this group of disorders and new epidemiologi c
instruments such as the Diagnostic Interview Schedule (DIS), the
Present States Examination (PSE), and the Center for Epidemiologi c
Studies-Depr ession Scale (CED-S). In a related theme, Dohrenwend
has developed an instrument for measuring demoralizatio n as a non-
specific indicator for depression and other psychopatholo gy in
community surveys (1). Weissman has developed a reliable scale
for assessing social adjustnent and has shown how this outcome vari-
able is different than symptom ratings in depressive illness (2).
In addition, two well known concepts, social support and life events,
have been reexamined and applied to depressive illness. It is the
purpose of this report to examine the research on these two vardables
in particular. This emphasis is predicated on the assumption that
we are in danger of becoming complacent with our current understandin g
of and methodology for measuring these two crucial factors in the
etiology and clinical course of the affective disorders.

I. Life Events

The most basic hypothesis regarding life events and depression

is that stressful events can lead to the development of depression,
although more sophisticated versions including additional variables
have been proposed by the Dohrenwends (3) as well as Dean and Lin
(4). Holmes and Rahe advanced the research in this area by developing
a quantitative method for measuring life events. Using this, or
similar instruments, numerous studies have examined the relationship
of life events to psychosomati c illness and depression.

177
 There are three criticisms of this type of life events research.
First, certain events, such as a vacation or traffic tickets seem
relatively minor and have questionable predictive capacity over
negative outcomes. Second, while the numerical values for each
event may be accurate for the population surveyed, they have dubious
impact for the individual patient or for populations which are cul-
turally different. Unfortunately, this instrument has been used by
clinicians to determine whether an individual client is at risk for
depression. A third criticism is that while the scale may be pre-
dictive for the development of gastrointestinal or cardiovascular
illness, it is probably less predictable for depression.

Aware of the limits of the SRRS, several investigators have

developed new instruments and methods for measuring stress. George
Brown and colleagues, long known for their work on stress and
schizophrenia, began developing a more time-consuming method of
assessing life events in their work on depression (5). The threat
of each important event, the emotional impact, and the readjustment
in daily routine that resulted from it is solicited and quantita-
tively rated for each patient. A unique feature of Brown's method
is the solicitation of data to determine whether the event occurred
prior to and independent of any outcome symptoms or assessments.
Brown also has provided a measure of the support received from a
confidant and others surrounding each event. In addition to events,
this research group has provided similar assessments of "difficul-
ties" thus clarifying the previous confusion between two types of
stressors: discrete events and long-standing problems (e.g.,
financial, marital, occupational). While this method represents
the most optimal method of rating events, the two-hour interview
needed for its completion and the required training with Brown
and Harris make it impractical for most research projects.

Another venue in life events research has been the impact of

early life events on depression in later life. This is particularly
difficult because it is not possible to make an objective assessment
of the impact of an event in the past nor to estimate the level of
ongoing stress. such as marital conflict or single parent homes, on
the child's development. Perhaps most importantly, it is impossible
to measure the amount of support a child receives from others when
experiencing the stress and its aftermath. For this reason, most
research has concentrated on the death of a parent and the age the
death occurred. While the preponderance of evidence probably favors
the hypothesis that early parental loss renders one more vulnerable
to depression in later life, there are a number of studies which
fail to support this hypothesis. This variation is due in part to
the sampling problems as well as the lack of a reliable way to
retrospectively measure social support.

There are at least three new areas in which the life events
research can expand. The first is to attempt to integrate early

178
life events with current stresses in a single hypothesis. This might
be the following: there is an optimal amount of early life stress
that helps prepare the individual to successfully deal with subse-
quent events and stresses later in life. This positive outcome would,
of course, depend on several other factors including the developing
personality of the child and his coping skills and the amount of the
positive support received during the early loss.

A second area for future development is the preparation of

reliable and valid scales for measuring life events that would
include the following features: a) an assessment of whether the
event was both independent of and prior to the outcome variables,
and b) a measure of the adaptation stress caused by each event to
the specific individual to whom it occurred.

Finally, it should be acknowledged that different stressful life

events affect different outcome variables. While multiple life
events are probably additive in producing a nonspecific response to
stress, such as demoralization, specific events may lead to specific
outcome. Recently, Finlay-Jones, et al. have provided an initial
confirmation of this hypothesis by showing that events involving a
loss lead to depression while events showing a threat lead to anxiety.

II. Social Suooort

Cognizant of the failure of life events research to predict more

than 15% of the variance in depressive outcomes, in the mid-1970s
several theorists suggested that social support be included in
predictive studies. Social suppott has come to emphasize those
affective and nurturing components one receives from close friends
and family and has been differentiated from the social network con-
cept previously described by British social scientists. In rela-
tionship to depression, social support has been proposed as either a
buffer to stressful life events by Dean and Lin (4), and low social
support as an independent risk factor in Dohrenwend's additive bur-
den hypothesis. Specific aspects of social support have confirmed
i'ts utility in clinical studies. Brown, et al, found the presence
of a close confidant to be protective in depression development
among working-class women. Clayton provided evidence that widows
with children in their immediate geographic area are less susceptible
to depression in their first year of bereavement (6).

Over the last three years, instruments for quantitatively meas-

uring social support have been developed. Lin developed an instru-
ment for a Chinese-American population and showed how individuals
with low support are more likely to develop depression (7).
Surtees piloted a brief instrument in a well-designed study to show
the predictive power of support in the ongoing course of unipolar
patients (8). Over the last three years, Flaherty and Gaviria have
developed and psychometrically tested an instrument, the Social

179
Support Network Inventory (SSNI) (9,10). This instrument consists
of 11 questions which are asked regarding the five major individuals
in one's support network. The 11 questions have a high internal
consistency (alpha coefficient= .93), high test-retest reliability
and can differentiate populations predictive to be high in support
(a religious commune) from the general population.

As with life events, there are several methodologic and theo-

retical areas in which the social support research can expand.
Social support has been charged with being more subjective and
questions have been raised as to whether it should more properly be
considered the perception of support. As such, it may vary directly
with general well-being, i.e., when an.individual feels good he per-
ceives greater support. Such ideas should be tested by a longitudinal
design that includes measure of the perception of being supported as
well as objective assessments. It would then be particularly enlight-
ening to examine how perceived and objective support change in bipolar
patients during manic, depressive and euthymic periods.

One theoretical notion that we have proposed is that social sup-

port assessments such as ours are really measuring an ego function.
Certain individuals have a tremendous capacity for receiving support
regardless of their environment, while others are so deficient in
this regard that they form few close relationships even under optimal
circumstances. If one accepts this hypothesis, it would be logical
to conclude that the capacity to receive (and benefit from) support
is the most basic and predictive variable.

On a related theme, investigations need to be undertaken to

determine how one develops this capacity for social support. Primate
studies by Harlow and McKinney as well as Spitz's earlier observa-
tions of infants in fondling homes illustrate the failure to develop
attachments under extremely negative conditions. However, how it
may be different for boys and girls has not been elucidated. In fact
little consideration has yet to be given to the role of social sup-
port in the burgeoning area of childhood depression.

REFERENCES

1. Dohrenwend, B.P., Shrout, P.E., Esir, G., et al.: Nonspecific

psychological distress and other dimensions of psythopathology.
Arch.Gen.Psych. 37:1229-1236, 1980.

2. Weissman, M.M., Klerman, G., Paykel, E.S., et al.: Treatment

effects of social adjustment in depressed patients. Arch.Gen.
Psych. 30:771-778, 1980.

3. Dohrenwend, B.S., Dohrenwent, B.F. Overview and prospects for

research on stressful life events. In B.S.&B.F. Dohrenwend (eds)
Stressful Life Events: Their Nature artd Effects. New York,
Wiley & Co., 1974.

180
4. Dean, A., Lin, N.: The stress-buffering role of social support.
J.Nerv.Ment.Dis. 165:403-416, 1977.

5. Brown, G.W., Harris, T.: Social Origins of Depression. New

York, Free Press, 1978,

6. Clayton, P., Halikes, J., Mauvie, W.: The bereavement of the

widow. Dis.Nerv.Syst. 32:597-604, 1974.

7. Lin, N., Ensel, W., Simeone, R., et al.: Social support,

stressful life events and illness. J. Health Soc. Behavior,
20:109-119, 1979.

8. Surtees, P.B. Social support, residual adversity, and depressive

outcome. Social Psychiatry, 15:71-80, 1980.

9. Flaherty, J.A., Gaviria, F.M., Blak, E.M., et al. The role of

social support in the functioning of patients with unipolar
depression. Am. J. Psychiatry, 140(4):473-476, 1983.

10. Flaherty, J.A., Gaviria, F.M., Pathak, D. The measurement of

social support: the SSNI. Comprehensive Psy~hiatry (in press,
1983).

181
THE SIGNIFICANCE OF PSYCHOPATHOLOGICAL CLASSIFICATION
IN INTERPRETING BIOCHEMICAL FINDINGS

Eberhard Gabriel
Medical Direction
Psychiatric Hospital
A-Vienna 1140

1. INTRODUCTION
The empirical background of my contribution is a research pro-
gramme in schizophrenics executed by Jellinger as the morphologist,
Riederer and for some time Reynolds as the biochemists and myself
as the psychiatric clinician. It is a post mortem study on Spiro-
peridolbinding in brain tissue of patients who in most cases died
after a long stay in a psychiatric hospital. I stress these 3 dis-
ciplines we are representatives of and the special design of a
post mortem study because these facts both lead to the problem of
case definition and -identification and at the same time point
out preconditions of such kind of studies which perhaps influence
its results. I would like to mention only 2 points depending on the
design which are related closely to the topic of my communication.
a) The one is as I mentioned that our probands have been
patients of a mental hospital with in most cases very long stays
there, e.g. that they not only have been chronically ill but that
they belong at the same time to a risk population for institutio-
nalism. From a function oriented view point and with respect to
the cross sectional evaluation.of the patients behavior and sympto-
matology it may be difficult to distinguish between illness dependant
and situation dependant symptoms. A classical example for this dif-
ficulty is the so called residual state or defect in schizophre-
nics, but the same has to be considered for paranoid states too.
The solution of the problem only can be that case definition from
a psychopathological view point only can be a functional one which
does not represent etiological conditions of the described state.
I have to admit that this is true only in functional psychoses. In
contrast the notion of organic psychoses comprises an etiological
inclusion criterion.
183
 b) An other precondition of a study with our design I would like
to mention because of its possible influence on the results is con-
cerned with the stability of the psychopathological states most pa-
tients showed before death (80%). (We related the item to the last
3 month of life). So there are only few acute patients.

2. CLASSIFICATION
As I said the biochemists studied the spiroperidol-binding in
brain tissue of patients who had the diagnosis of schizophrenia and
died in a mental hospital. In case finding we started from the ICD
diagnosis but compared it with 4 other diagnostic systems for schizo-
phrenia, e.g. the criteria of Bleuler, K. Schneider, the Vienna
Research Criteria of Berner and Feighner's criteria (WPA Diagnostic
Criteria, 1983). The comparision showed as was expected that the
diagnoses using Bleuler's criteria have been practically the same
as using ICD. There is a great overlap with the diagnoses using
K. Schneider's 1st rank symptoms. With these criteria a minority
of ICD- and Bleuler-schizophrenics is not reached. That is not esto-
nishing. K. Schneider himself based his clinical diagnosis of schi-
zophrenia not exclusively on 1st rank symptoms. Nevertheless in our
material 80% of the reD-schizophrenics show 1st rank symptoms at
any time of the illness course. The Vienna criteria of Berner are
psychopathological criteria as are the ICD-, Bleuler- and K. Schei-
der-criteria. We used these Vienna criteria without considering af-
fective blunting as a criterion, thus basing the classification only
on formal thought disorders and neologisms. With these criteria about
one half (48%) of the reD-schizophrenics is considered positive. The
Feighner criteria are not only psychopathological criteria but com-
prise a time criterion towards chronicity. It is not estonishing
that these criteria are positive in the majority of cases (74%) who
have been long stay patients of a mental hospital. Of course the ICD
295.6 comprises some sort of time criterion as it stresses on resi-
dual states. Otherwise in our study the actual psychopathological
state is only considered by ICD. All other systems are considered
basing upon criteria observed at any time of oberservation. - Is
one using this time criterion in Feighner's system in addition to ICD
differences in the overlap with the other 3 systems emerge, namely
an increase of overlap with the group of Berner positive cases and
a decrease of overlap between Feighner negative ICD schizophrenics
and lst rank symptom-schizophrenics (50%) and Berner positive schi-
zophrenics (0%). Therefore 3 psychopathological differentiations
seem to be meaningful:
a) a syndromatological one including a large group of patients
(ICD 295.3/297 : 295.6)

184
 b) a psychopathological one which is much more restrictive
(Berner positive : Berner negative) and
c) a differentiation basing on a time criterion (ICD positive/
Feighner positive : ICD positive/Feighner negative).

3. RESULTS
In the following I do not concentrate on the differences bet-
ween patients treated with neuroleptics within 3 month before death
and patients who had no neuroleptic treatment in this space of time.
I only mention that under any classificatory condition the biochemi-
cal measures are higher in patients treated with neuroleptics than
in not treated patients. But this is not my topic. (Reynolds et al.,
1981)
Again I do not concentrate on the problem whether the size of
neuroleptic treatment in terms of dosage and/or time is influential.
(Reynolds et al., 1981) This is a problem not easily to solve with
because of methological reasons (problem of equivalence of doses of
different neuroleptics, problem of different distribution of the
medication in time). Our analyses on this topic are not finished yet.

Table 1. Bmx levels, psychopathological classification and neuro-

leptic treatment (NL)
13 14 15 16 17 18 19 20 ~ 22 23 24 25 26 27 28
c
X 295.3/297 0 X 295.3/297
NL- n NL+
t
X 295.6 r X 295.6
NL- 0 NL+
l
BER+ X s X BER+
NL- + NL+
1~95
X BER- X BER-
NL- NL+
X ICD/FEI+ X FEI+
NL- N~

X ICD/FEI- X ICD/FEI-
NL- NL+

185
 Thus I do not want to stress the high receptor number in the
different groups but to stress on the low receptor number in all the
groups without neuroleptic treatment, within the last 3 months be-
fore death. In any case (n=9) it relates to patients who have been
in a stable state at that time only few of them showing paranoid symp-
toms (ICD 295.3/297, n=2). It is well known to psychiatrists that
paranoid symptoms in chronic patients may not be based on the same
pathogenetic pathways than in acute psychoses. The Berner positive
probands too are few (n=3). Thus from a psychopathological view
point it seems to be possible that the different distribution of
biochemical measures is based on treatment effects on the one hand
and on pairs of psychopathological notions as are productive : non
productive, acute : chronic, stable : unstable characterizing morbid
states which probably are interrelated with the size of neuroleptic
treatment on the other hand. From this point of view the next steps
of research in field have to focuse on quantitative aspects of this
kind and to apply multivariate statistical methods. Both should be
possible.

REFERENCES

ICD 9th Rev., 1978, "Mental disorders: Glossary and Guide to their
Classification in accordance with the 9th Revision of the
International Classification of Diseases", WHO, Geneva.
Reynolds, G., Riederer, P., Jellinger, K., and Gabriel, E., 1981,
Dopamine Receptors and Schizophrenia: the neuroleptic
drug problem, Neuropharmacology, 20:1319-1320
WPA Diagnostic Criteria, 19~3, American Psychiatric Press,
Washington.

186
METABOLIC DISTURBANCES IN PSYCHIATRIC DISORDER: THE QUESTION

OF DIAGNOSTIC SPECIFICITY

Csaba M. Banki 8 , Mihaly Arat6b and Zsuza Pappc

~egional Neuropsychiatric Hospital, Nagykall6

bHamilton Psychiatric Hospital, Ontario, Canada LBN 3K7
cNational Institute for Nervous and Mental Diseases
Budapest, Hungary

INTRODUCTION

An increasing number of indirect neurochemical markers have

been proposed to reflect CNS biological dysfunctions underlying
some psychiatric disorders in man. CSF amine metabolites were
often reported to be decreased in depression, but several later
studies contradicted; there are some excellent reviews covering
this subject 1 • 2 • In contrast, more recent works seem to support
the view that CSF biochemical parameters /concentration of HVA,
5HIAA, MHPG, etc./ are more closely correlated with syndrome sub-
types, symptoms or other physical ~r 5 psychological variables than
with the clinical diagnosis itself - • Some "elementary"
symptoms were already found to be related to certain CSF meta-
bolites: motor activity to HVA6, anxiety and insomnia to 5HIAA7,
suicide and aggression to 5HIAA and MHPG2 • 8 • 9. However, the
study of isolated symptoms in relationship with biochemical
alterations has usually been. confined to a limited number of
clinical disorders4,lO.
Besides amine metabolites the precursor amino acid trypto-
phan /TRY/ was sometimes measured in the CSF~ it might reflect
brain tryptophan supply and was correlated to the 5HIAA concent-
ration, without showing diagnosis-dependent mean differences 11 •
Cortisol may be an index of ~~othalamic-pituitary dysregulation
observed mainly in depression , although CSF cortisol failed to
differentitje consistently the affective disorders from other
conditions •
This study aimed at investigating all the four of the above
substances in the CSF of 110 female psychiatric inpatients and
searching for all the correlations between clinical and biochem-

187
ical variables across four diagnostic categories.

PATIENTS AND METHODS

The inclusion criterion was a concordant diagnosis from 2

psychiatrists, based on DSM-III criteria, of an active episode
of either major depression /n=29/, or schizophrenic disorder
/n=32/, or adjustment disorder /"reactive" syndromes, n=2o/, or
alcohol dependence /n=29/. Significant physical, endocrinological
or other psychiatric disorders formed the exclusion criteria;
all patients had been free from major psychotropic medication
for at least 2 weeks before admission.
After obtaining informed consent lumbar punctures were per-
formed in a standard setting at 8.oo-9.oo a.m.; the samples were
analyzed within 14 days for 5HIAA and HVA /by fluorometric method/,
for cortisol /competitive protein-binding/ and for TRY /fluoro-
metry/. 5HIAA and HVA values were standardized to the mean age
and body height of the population5.
45 items of actual symptoms were rated during the interview,
using a shortened and modified version of the Comprehensive
Psychiatric Rating Scalel4, adapted by ourselves years ago. We
analyzed for correlation only those items which received a score
greater than l in at least 2o% of the whole population /to avoid
markedly skewed distribution and nearly constant variables/, and
obtained in this way 9 symptoms for further calculations. In
addition, the CGI /Clinical Global Impression/ scale was com-
pleted in each case. All ratings and diagnostic interviews pre-
ceded the LP's by one or two days.
In spite of symptom ratings being not really continuous
variables parametric statistics are acceptable4, therefore we
computed Pearson's correlation coefficients and related statistics.

RESULTS

Mean concentrations of four CSF biochemical parameters

proved to be homogeneous in four diagnoses except 5HIAA that
was almost significantly lower in schizophrenia as compared to
depression /table l./. Global clinical severity did not correlate
with any of these variables either in the total group or in any
of the diagnostic subgroups /table 2./.
The computation of correlation coefficients separately for
each possible symptom-metabolite pairs in all the four diagnosis
produced 144 values; to save space we only present the range
of these coefficients in table 2. Homogeneity tests /Fischer's
z-transformation and chi 2-test/ revealed only two significant
inter-diagnostic differences of these correlations: in depression
the coefficient for 5HIAA/agitation was -o.3l against positive
values in the other three groups; and the ~ for TRY/paranoia
was -o.42 in depression against positive values in the other
three groups.

188
Table 1. Means ± standard deviations of CSF 5HIAA, HVA, TRY
and cortisol in depression /DEP/, schizophrenia /SCH/,
alcohol dependence /ALC/ and adjustment disorder /ADJ/.

5HIAA HVA TRY Cortisol

DEP 144 + 41 162 ± 87 4.8 ± 2. 3 23 + lo

SCH 118 ± 36 166 + So 4.4 + 2.0 2o + 12
ALC 141 ± 59 158 + 78 5.2 ± 2.9 19 + 9
ADJ 135 ± 52 149 ±63 4.5 ± 1.8 24 +11

TRY is given as pmol es /L, the other metabolites as nmoles/L

Table 2. Correlatiot;J.S of CSF 5HIAA, HVA, TRY and cortisol with

CGI scores and 9 isolated symptoms across four dia-
gnostic groups /range of four r's per cell/

5HIAA HVA TRY Cortisol

C. G. I. from -o.l7 -o.l4 -o.2o -o.23

to o.l3 o.l2 -o.o6 o.28

Depressed mood from -o.l2 -o.l4 -o.o8 o.l8

to o.o8 o.l7 o.lo o.4o
Anxiety from -o.44 -o.o7 -o.l3 -o.l4
to -o.23 o.32 o.l8 o.l6
Somatization from o.o6 ~o.o7 -o.o8 -o.27
to o.34 o.l3 o.28 o.o5
Suicidal thought from -o.43 -o.o8 -o.l8 -o.32
to -o.l5 o.33 o.l3 -o.o3
Insomnia from -o.54 -o.o7 -o.33 -o.ol
to -o.33 o.45 -o.l8 o.42
Retardation from -o.o2 -o.52 -o.ll -o.o2
to o.28 -o.l7 o.o9 o.l7
Agitation from -o.31 o.34 -o.l7 -o.l7
to o.29 o.67 o.l2 o.l8
Hallucinations from -o.41 -o.26 -o. 35 -o.24
to o.o3 o.22 -o.o6 o.o3
Paranoia from -o.33 o.l8 -o.42 -o.5o
to o.o9 o.54 o.2o -o.l6

All correlations except 5HIAA/agitation and TRY/paranoia were

homogeneous across the four diagnoses.

189
 Table 3. Average correlations of CSF 5HIAA, HVA, TRY and
cortisol with CGI score and 9 isolated symptoms.

5HIAA HVA TRY Cortisol

C. G. I. -o.o2 -o.o7 -o.l2 o.l4

Depressed mood o.o4 o.o7 o.lo o.2o
Anxiety -o.28b o.o9 o.l2 o.o6
Somatization o.2oa o.oo o.2.3a -o.ll
Suicidal thoughts -o.2la o.27b o.o5 -o.o6
Insomnia -o.37c o.ol o.22a o.l6
Retardation -o.o2 -o.2oa -o.lo o.l2
Agitation o.l6 o.5oc o.l4 -o.l4
Halluc in at ions -o • .35c -o.o2 -o.l7 -o.l3
Paranoia -o.l8 o.22 o.l5 -o.27b
a = /p<._.o5/, b = /p<.ol/, c = /p .(,. ool/

Average correlations for the total group of llo patients

were calculated because of the previously demonstrated inter-
diagnostic homogeneity /table .3./. Five symptoms correlated sig-
nificantly with 5HIAA, three with HVA, two with TRY and one with
cortisol.

DISCUSSION

The four CSF biochemical variables investigated in this

study proved to be quite homogeneous in four diagnostic groups,
except a small difference of mean 5HIAA between depression and
schizophrenia in agreement with Gattaz et al. 1 5 Because of the
considerable overlap even between these two groups our data do 1
not support the hypothesis that any of these CSF variables could
help clinical psychiatric diagnosis.
On the other hand we found several significant correlations
between some symptoms and CSF metabolites. Many of them corres-
pond to earlier observations, as e.g. the inver~e relationship
between anxiety, suicide, insomnia and 5HitA,3' or the direct
correlation between motor activity and HVA • A more detailed
multivariate analysis yielded essentially similar results4. It
is important to be noted that all these correlations appeared
beside a practically complete lack of relationship of any of the
CSF variables with the global clinical severity. This is a con-
firmation of the view, first proposed by Van Praag,6 that CSF
biochemical investigation could be rather symptom- t:pan syndrome-
oriented. This conclusion is qu:i.te concordant with the everyday
experience that most of our current psychotropic drugs, thought
to act mainly by altering neurot:cansmitter function, are rather
symptom- than syndrome-specificlb.
The most important result of this study is that the

190
observed symptom/metabolite correlations were present not only
within depressed patients but in at least four diagnostic cate-
gories; in addition, these correlations - with a few exception -
proved to be quite homogeneous inter-diagnostically. This lends
further support to the above statement and underlines the import-
ance of a symptom-oriented line of biological research.
Such results are, of course, dependent on the definition
of the symptom variables; here we tried to reduce the number of
the items also by omitting symptoms which occurred rarely or
only within one syndrome; and contracting others into one vari-
able /e.g. "hallucinations"/. A more detailed analysis might
very likely discover other correlations or perhaps some inter-
diagnostic differences; this, however, do not alter the main
conclusion that these CSF parameters are principally related
to certain symptom patterns instead of any specific diagnosis.
This conclusion is, in addition, in line with several recent
reports which observed specific correlations of CSF metabolites
with e.g. suicide 5 projective test resultsl7 or electrophysiol-
ogical parametersi 8 , again specific biological or psychological
variables other than traditional clinical diagnosis. Future work
is needed to discover in detail the relationship between CSF
biochemical markers, behavioural symptoms and diagnoses - even-
tually as dependent variables - in human psychiatric research.

REFERENCES

L H.M. Van Praag, "Depression and Schizophrenia, a Contribution

on their Chemical Pathologies." Spectrum, New York /1977/
2. M.Asberg, and L.Bertilsson, Serotonin in depressive illness,
studies of CSF 5HIAA, in: "Neuro-Psychopharmacology,"
B.Saletu, P.Berner, L.E.Hollister, eds., Pergamon, Oxford /1979/
3. C.M.Banki, G.Molnar, I.Fekete, Correlation of individual
symptoms and other clinical variables with CSF amine meta-
bolites and tryptophan in depression. Arch.Psychiat.Nervenkr.
229:345 /1981/
4. H.Agren, Biological markers in major depressive disorders.
Thesis, Uppsala /1981/
5. M.Asberg, L.Bertilsson, E.Rydin, D.Schalling, P.Thoren, and
L.Traskman, Monoamine metabolites in CSF in relation to de-
pressive illness, suicidal behaviour and personality, in:
"Recent Advances in Neuro-Psychopharmacology," B. Angrist,
G.D.Burrows, M.Lader, O.Lingjoerde, G.Sedrau, D.Wheatley,
eds., Pergamon, Oxford /1981/
6. H.M.Van Praag, J.P.W.Lakke, and T.Schut, Dopamine metabolism
in depression, psychoses, Parkinson disease, the problem of
specificity of biological variables in behavioural disorders.
Psvchol.Med. 5:138 /1975/
7. C.M.Banki, Correlation of anxiety and related symptoms with
CSF 5HIAA in depressed women. J.Neur.r,ransmiss. 41:135 /1977/

191
8. G.L.Brown, J.C.Ballenger, M.D.Minichiello, and F.K.Goodwin,
Huma.n aggression and its relationship to cerebrospinal fluid
5HIAA, MHPG and HVA, in: "Psychopharmacology of Aggression, 1"
M. Sandler, ed , Raven, New York /1979/
9. C.M.Banki, and M. Arat6, Amine metabolites and neuroendocrine
responses related to depression and suicide J.Affect.Disord
/in press/
lo. C.M.Banki, Evidence of disturbance of monoamines in depression,
in: "Management of Depression with Monoamine Precursors,"
H.M.Van Praag, J.Mendlewitz, eds., Karger, Basel /in press/
11. G.Curzon, B.D.Kantamaneni, P.Van Boxel, and P.K.Gillman,
Substances related to serotonin in plasma and in lumbar and
ventricular fluids of psychiatric patients. Acta psychiat.
scand. suppl.28o: 3 /198o/
12. B.J.Carroll, The dexamethasone suppression test for melan-
cholia. Brit.J.Psychiat. 14o:292 /1982/
13. D.C.Jimerson, R.M.Bost, D.P.Van Kammen, J.S.Skyler, G.L.Brown,
and W.E.Bunney, CSF cortisol levels in depression and schizo-
phrenia. Amer.J.Psychiat. 137:979 /198o/
14. M.Asberg, S.A.Montgomery, C.Perris, D.Schalling, and G.Sedvall,
A comprehensive psychopathological rating scale. Acta nsychiat.
scand. suppl.271:5 /1978/
15. W.F.Gattaz, P.Waldmeier, and H.Beckmann, CSF monoamine meta-
bolites in schizophrenic patients. Acta psychiat . scand.
66:35o /1982/
16. H.Beckmann, Die medikametose Therapia der Depression,
Nervenarzt 52:135 /1981/
17. E.Rydin, D.Schalling, and M.Asberg, Rorschach ratings in
depressed and suicidal patients with low levels of 5HIAA
in cerebrospinal fluid. Psychiat. Res. 7:229 /1982/
18. K.L.Benson, V.P.Zarcone, K.F.Faull, J.D.Barchas, P.A.Berger,
REM sleep eye movement activity and CSF concentration of
5HIAA in pscchiatric patients. Psychiat.Re~. 8:73 /1983/

192
DISTURBANCES IN SERINE-GLYCINE METABOLISM IN RELATION TO ACU~E

PSYCHOSES WITH PSYCHEDELIC SYMPTOMS

Jacques Bruinvels and Lolke Pepplinkhuizen

Group Biological Psychiatry, Medical Faculty

Erasmus University Rotterdam, P.O.Box 1738
3000 DR Rotterdam, The Netherlands

Psychoses showing complete recovery mostly have an episodic course

and belong to the large group of acute psychoti~ syndromes which
cannot be classified either as schizophrenic, paranoid, organic
mental or affective disorders.
The first patient studied was suffering from acute psychotic epi-
sodes. After an operation she showed a psychosis identical with a
LSD- or mescaline-induced state. Although acute intermittent por-
phyria might be the cause of the psychosis no biochemical evidence
was obtained to support this diagnosis, although a high carbohydrate
diet resulted in complete recovery (Stein and Tschudy, 1970).
It was postulated that an increased porphyrin production is respons-
ible for the endogenous synthesis of hallucinogenic substances like
tetrahydro-S-carbolines (THSC).
The enhanced porphyrine synthesis increases the conversion of serine
into glycine since the latter is one of the precursor substances of
porphyrins. Methylene-tetrahydrofolate being disengaged in the pro-
cess of conversion of serine into glycine will accumulate, augment-
ing the chance of segregation of formaldehyde. A non~enzymatic re-
action of formaldehyde with monoamines will finally result in the
production of psychotogenic substances of the a-carboline and/or
isoquinoline type (Meller and Friedhoff, 1979; Pearson and Turner,
1979; Pepplinkhuizen et al, 1980).
An investigation into the literature on the incidence of such
psychotic syndromes in porphyria and also in metabolic diseases and
vitamin deficiencies, often supposed to create an accumulation of
active carbon moieties such as methylene- and methyl-tetrahydrofo-
late, turns out to be negative.
The biochemical literature appears also to offer few cues to pos-
tulate that vitamin B12 deficiency and homocystinurias a.o. will
cause a considerable accumulation of methylated folates. Only a

193
forced catabolism of serine and glycine, in porphyria or otherwise,
seems to make this possible as things stand now.
By means of a retrospective investigation of 71 case histories
(Pepplinkhuizen et al., submitted) it became clear that acute psy-
choses characterized by dysperceptions have been regularly observ-
ed (n=27). It appeared that these perceptual anomalies are mainly
mentioned at the onset of a psychosis. Besides this type of psy-
chosis generally shows a kaleidoscopic polymorphic picture. They
are, on the strength of their clinical description, so-called
'degeneration psychoses'. They are clearly distinct from a second
group of acute psychosis without dysperceptions - mostly diagnosed
as psychogenic psychoses - in which the clinical picture is deter-
mined by strong depersonalisatio n (n=35).
This approach may offer a completely new classification of the so-
called Unclassifiable Psychoses of the DSM III.
Based on the central thought that an accelerated conversion of
serine into glycine should be the cause in all patients suffering
from an episodic psychosis characterized by the occurrence of dysper-
ceptions, oral loadings of serine, glycine or glucose in restricted
quantities (2 mmoljkg bodyweight) were attempted. The investigation
was carried out double blind. About three hours after the adminis-
tration, patients reacted to serine- and a few to glycine- with
distinctive phenomena. In particular depersonalisatio n, changes of
mood and characteristic perceptual disturbances were induced. The
administration of glucose was without effect. It was therefore con-
cluded that a disturbance in serine and glycine metabolism may be
the cause of the psychoses (Pepplinkhuizen et al., 1980).
In a second double blind study loading tests with serine, glycine
or glucose and occasionally with alanine or methionine were per-
formed in recovered patients and healthy controls.
Out of a total number of 33 patients who were expected to react on
account of their clinical picture - which should be characterized
by dysperceptions - 24 patients turned out to show the expected
response to serine, glycine or both amino acids. Of 6 patients an
erroneous diagnosis (schizophrenia a.o.) and/or confusion regarding
the character of the dysperceptions appears to be the cause of the
wrong predictions. From the controlgroup (33 patients having suf-
fered from other psychoses and 15 healthy subjects), 1 patient also
turned out to react to serine and glycine. Administration of methio-
nine or alanine did not evoke psychotic symptoms.
Plasma amino acid analysis of a number of these patients and heal-
thy controls (n=8) indicated that serine-positive patients (n=7) -
after their recovery - had a significantly lower plasma level (62%)
of serine as compared to healthy controls and a significantly in-
creased plasma concentration (228%) of taurine, while in glycine
positive patients (n=3) and manic depressive patients (n=3) normal
plama levels of serine and taurine were found.
Determination of plasma concentrations of serine and glycine after
the administration of serine or glycine showed an increased con-
version of serine into glycine as compared to healthy controls
(0.128±0.013 and 0.086±0.008 mol glycine/mol serine respectively)

194
while the conversion of glycine into serine was significantly de-
creased (0.114±0.012 vs 0.208±0.011 mol serine/mol glycine in heal-
thy controls). Since the conversion of serine into glycine is an
aequilibrium reaction the increased conversion of serine into glyci-
ne is probably the result of an inhibition of the conversion of gly-
cine into serine. It was therefore, concluded that a diminished
availability of the co-substrate of the latter reaction, methylene-
tetrahydrofolic acid, was responsible for the impaired formation of
serine from glycine.
These data suggest that the increased plasma concentration of tau-
rine - which is formed by binding of serine to homocysteine - and
the impaired conversion of glycine into serine may be both respons-
ible for the lowered plasma level of serine found in serine-positive
patients. It was argued that the folate cycle and the methionine-
homocysteine-taurine cycle are operating at a maximal level in these
recovered patients and that administration of serine (as pure amino
acid or in food containing high amounts of this amino acid) will,
therefore, increase the formation of methylenetetrahydrofolic acid.
The latter will decompose into formaldehyde and tetrahydrofolic
acid and the formaldehyde may react with monoamines to form tetra-
hydro-S-carbolines and tetrahydroisoquinolines which may be res-
ponsible for the evoked psychotic symptoms as postulated previously
(Pepplinkhuizen et al.; Bruinvels et al., 1980).

REFERENCES

Bruinvels, J., Pepplinkhuizen, L., van Tuijl, H.R., Moleman, P.,

and W. Blom, 1980, Role of serine, glycine, and the tetrahydrofolic
acid cycle in schizoaffective psychosis. A hypothesis relating
porphyrin biosynthesis and transmethylation, in: Enzymes and
Neurotransmitters in Mental Disease, E. Usdin, T.L. Sourkes, and
M.B.H. Youdim, eds., John Wiley & Sons Ltd., Chapter 11.8.
Meller, E., and Friedhoff, A.J., 1979, 5-Methyltetrahydrofolate and
metabolism of biogenic amines, in: Folic Acid in Neurology, Psy-
chiatry and Internal Medicine, M.I. Botez and E.H. Reynold, eds.,
Raven Press, New York, p.157.
Pearson, A.G., and Turner, A.J., 1979, The formation of S-carboline
alkaloids mediated by serine hydroxymethyltransferase, FEBS Let-
ters 1: 96.
Pepplinkhuizen, L., Bruinvels, J., Blom, W., and Moleman, P, 1980,
Schizophrenia-like psychosis caused by a metabolic disorder, The
Lancet, March 1: 454.
Stein, J.A., and P. Tschudy, 1970, Acute intermittent porphyria a
clinical and biochemical study of 46 patients, Medicine, Balt~more
49: 1.

195
PEPTIDES AND AMINES IN AFFECTIVE DISORDERS

Annette Gjerris and Ole J. Rafaelsen

Psychochemistry Institute & Department of Psychiatry
Rigshospitalet
DK-2100 Copenhagen

Modern psychopharmacology gave rise to the amine hypotheses of

affective disorders and the search began for amine metabolites that
could fulfil our hopes for biological markers of depression and mania.
In spite of the intensive and extensive research over two
decades, by and large, it is my seasoned opinion that little of last-
ing value has come out of this dedicated struggle.
It is no disgrace that you study what you are able to study,
just as you should only move stones that are not too heavy. Concern-
ing the neurotransmitters this have for quite some years meant that
brain studies were concentrated on three amines: noradrenaline,
dopamine, and serotonin because these three amines could be visualized
by fluorescent microscopy, whereas many others could not, like GABA,
acetylcholine, glycine, and the Benjamin among amine brain neuro-
transmitters: adrenaline.
As a consequence it was the same neurotransmitters - or rather
their metabolites as only these could be detected with available
techniques - that were studied in various biological fluids: blood,
urine, cerebrospinal fluid (CSF). By and large, CSF-studies have
yielded the more interesting results. This fluid is after all nearest
to the brain, even if some contemptuously have nicknamed this precious
liquid the gutter of the central nervous system.

The three metabolites studied have been HVA (from dopamine), MHPG
(from noradrenaline - and incidentally also from adrenaline) and 5-
HIAA (from 5~HT (serotonin)) (van Kammen et al., 1982).

197
 But soon the problems and controversies started. Some found
only low values of HVA, but not of 5-HIAA, while others had opposite
findings (Banki et al., 1981). This led to a noradrenergic school
(Lingj~rde, 1983) and a serotoninergic school (Asberg, l976a). Some
found changes in totally drug-free patients, while others had to
apply preloading with probenecide (to block transport of amine
metabolites out of the CSF-compartment) in order to see a meaningful
pattern of changes (Maas et al., 1982; van Praag, 1982).
Some reported normalization in recovered patients, some found
unchanged values after recovery.
Some could make distinctions between unipolar and bipolar
patients (Ashcroft et al., 1973), while others could not. And some
were best in falsifying the results of others (Vestergaard et al.,
1978), but they were open to the criticism that the analytical
methods were not sufficiently specific.
A new avenue was opened when Asberg and her co-workers reported
a bimodal distribution of 5-HIAA in depression with the interesting
notion that patients who had tried to commit suicide applying violent
methods more often had low values than those who had used less
dramatic, aggressive methods (Asberg et al., 1976b). Reconfirmation
of these studies have turned the attention to the aggressive aspects
of personality and away from the depressive aspects (Asberg et al.,
1981).

Most of the classical cyclic antidepressants blocked the back-

pumping of both noradrenaline and serotoin, but to various degrees
in individual subjects, partly due to the added effects of their
active metabolites (Traskman et al., 1980).
With the advent of more specific noradrenergic or serotonergic
antidepressants the study of whether one or another type of anti-
depressant was more efficacious in individual patients, became
feasible. In a Swedish study it was found that patients with low 5-
HIAA responded better to zimeldine (a selective, but not specific 5-HT
transport blocker) but the dichotomy was not very strong (Agren,
1981). The near future will see studies with truly specific serotonin
blockers like citalopram, and we will then see - or not see - the
predictive value of CSF-metabolites for antidepressant therapy with
cyclic antidepressants (Cowdry et al,, 1981).

I will now come to our recent results that are characterized

by

l. Analysis of catecholamines proper and not their metabolites,

~d
2. Studies of a series of peptides that presumably function as
neuromodulators in the CNS - if not neurotransmitters in
their own right.

198
 Adrenaline has in recent years been recognized as a brain neuro-
transmitter, quantitatively much more sparsely present in the brain
and CSF than noradrenaline.
In collaboration with Dr. Niels Juel Christensen, also in Copen-
hagen, we have performed two studies of CSF-adrenaline. Both studies
showed that CSF-adrenaline was reduced by some 75 per cent in depres-
sion and normalized on clinical recovery (Christensen et al., 1980);
and that the reduction was equally pronounced in endogenous and in
non-endogenous depression (Gjerris et al., l98lb). CSF-noradrenaline
showed no differences between depressed and recovered patients and
controls.

We have next turned to animal experimentation, to study whether

antidepressant drugs influenced CSF-adrenaline and noradrenaline.
Isocarboxazide was administered intragastrically every day and CSF-
adrenaline doubled already after 24 hours and continued to increase,
reaching some 250 per cent after six weeks of continuous treatment.
In contrast, rat CSF-noradrenaline increased more slowly, and the
increase was percentage wise much smaller, reaching only some 40 per
cent after six weeks (Gjerris et al., 1983). I would like to mention
the study of Da Prada (1982) who using one of the new reversible MAO-
inhibitors found that the increase of adrenaline in rat hypothalamus
was faster, more pronounced, and longer lasting than that of nor-
adrenaline.

Among the brain and CSF peptides we have studied cholecystokinin

(CCK), thyreotropin stimulating hormone (TRH), vasopressin, vasoactive
intestinal popypeptide (VIP), and a few others. The only positive
finding was with VIP.

The scatter of results in depressive patients was quite large,

and when applying the Newcastle rating scales to distinguish between
endogenous and non-endogenous depression, it turned out - to our
surprise - that the non-endogenous, but not endogenous, depressions
had a 50 per cent reduction in CSF-VIP. Furthermore there was no
change of these low values when the patients had recovered from their
depression (Gjerris et al., l98la). We do not know much about the
function of VIP in the CNS (Johansson et al., 1982). It is noteworthy
that VIP is present in CSF in concentrations 5 to 10 times as high as
in blood. I do not remember any other substance with such a skewed
distribution between CSF and blood.

Concerning the co-existence of classical neurotransmitters and

peptides in neurons, it has been reported that VIP is found in
acetylcholine containing neurons (Hokfelt et al., 1980).

We are now hoping that others will start to replicate our

findings, which we at present interpret as two biological markers
for depression. One of these (CSF-adrenaline) is state-dependent

199
(disappearing when the depression disappears and undiscerning b~tween
endogenous and non-endogenous depression).
The other is state-independent and may thus be a trait marker
allowing to diagnose some individuals as having a reaction potential,
a mode for depression; in addition this marker seems to distinguish
between endogenously and non-endogenously depressed patients, and if
this is true, it will represent a new biological differentiation that
will promote research, diagnosis, and treatment in affective dis-
orders (Rafaelsen, 1980).
REFERENCES
Ashcroft, G.W., Crawford, T.B.B., Eccleston, D., Sharman, D.F.,
~1acDougall, E.J., Stanton, J.B., and Binns, J.K., 1966,
5-Hydroxyindole compounds in the cerebrospinal fluid of patients
with psychiatric or neurological diseases, Lancet, ii: 1049.
Ashcroft, G.W., Blackburn, LM., Eccleston, D., Glen, A.I.M.,
Hartley, W., Kinloch, N.E., Lonergan, M., Murray, L.G., and
Pullar, I.A., 1973, Changes on recovery in the concentrations
of tryptophan and the biogenic amine metabolites in the cerebro-
spinal fluid of patients with affective illness.
Psvchol Med, 3: 319.
Banki, C.M., Molnar, G., and Vojnik, M., 1981, Cerebrospinal fluid
amine metabolites, tryptophan and clinical parameters in depres-
sion. 2. Psychopathological symptoms. J Affect Dis, 3: 91.
Christensen, N.J., Vestergaard, P., Sgrensen, r:-, and Rafaelsen, O.J.,
1980, Cerebrospinal fluid adrenaline and noradrenaline in de-
pressed patients. Acta Psvchiatr Scand, 61: 178.
Cowdry, R.W., and Goodwin, ~.K., 1981, Biological and physiological
predictors of drug response, in: "Handbook of biological psychia-
try, part VI", H.M. van Praag, M.H. Lader, O.J. Rafaelsen, and
E.J. Sacher, eds., Marchel Dekker,Inc., New York, Basel.
Gjerris, A., Fahrenkrug, J., BG!Ijholm, S., and Rafaelsen, O.J.,
l98la, Vasoactive intestinal polypeptide (VIP) in cerebrospinal
fluid in psychiatric disorders, in: "Biological Psychiatry",
C. Perris, G. Struwe, and B. Jansson, eds., Elsevier/North-
Holland Biomedical Press, Amsterdam.
Gjerris, A., Jensen, E., Christensen, N.J., and Rafaelsen, O.J.,
l98lb, Adrenaline and noradrenaline in psychiatric disorders,
in: "III World Congress of Biological Psychiatry", Elsevier/
North-Holland Biomedical Press, Amsterdam.
Gjerris, A., Rafaelsen, O.J., and Christensen, N.J., 1981c,
Adrenaline in endogenous depression, in: "Adrenergic mechanisms
in human pathology", N.J. Christensen, 0. Henriksen, and N.A.
Lassen, eds., Clin Psvchiol, Suppl. 1: 98.
Gjerris, A., Barry, D.I., Christensen, N.J., and Rafaelsen, O.J.,
1983 to be published, MAO-inhibitors revisited I: Adrenaline (A)
and noradrenaline (NA) in cerebrospinal fluid (CSF) in
isocarboxazide treated rats •.

200
Hokfelt, T., Johansson, 0., Ljungdahl, A., Lundberg, J.M., and
Schultzberg, M., 1980, Peptidergic neurones, Natur~, 284:515.
Johansson, B.B., Fahrenkrug, J., Wikkels0, C., Andersen, 0., and
Blomstrand, C., 1982, Vasoactive intestinal polypeptide in
human cerebrospinal fluid, Front Horm Res, 9: 189.
Lingj~rde, 0., 1983, The biochemistry of depression. A survey of
monoaminergic, neuroendocrinolo gical, and biorhythmic distur-
bances in endogenous depression, Acta Psychiatr Scand, Suppl.
302: 36.
Maas, J.W., Koscis, J.H., Bowden, C.L., Davis, J.M., Redmond, D.E.,
Hanin, J., and Robins, E., 1982, Pretreatment neurotransmitter
metabolites and response to imipramine or amitriptyline treat-
ment, Psvchol Med, 12: 37.
Rafaelsen, O.J., 1980; Biology of manic-melancholi c disorders.
Med J Aust, 1: 637.
van Kammen, D.P., Sternberg, D.E., Lake, C.R., and Lerner, P., 1982,
Methodological issues in spinal fluid studies of schizophrenia:
The case of norepinephrine and dopamine-S-hydroxylase,
Front Horm RP.s, 9: 198.
van Praag, H.M., r982, Neurotransmitter s and CNS disease. Depression.
Lancet, ii: 1259.
Vestergaard, P., S0rensen, R., Hoppe, E., Rafaelsen, O.J., Yates,
C.M., and Nicolaou, N., 1978, Biogenic amine metabolites in
cerebrospinal fluid of patients with affective disorders,
Acta Psvchiatr Scand, 58: 88.
Traskman, L., Asberg, M.,· and Bertilsson, L., 1980, Serotonin and
noradrenaline uptake inhibitors in the treatment of depression -
relationship to 5-HIAA in spinal fluid, in: "Recent advances in
the treatment of depression", Proceedingsof an international
symposium, Corfu, Greece, Munksgaard, Copenhagen.
Agren, H., 1981, Biological markers in major depressive disorders.
A clinical and multivariate study.
Acta Universitatis Uosaliensis, Abstr. No. 405.
Asberg, M., Bertilsson, L., Rydin, E., Schalling, D., Thoren, P.,
and Traskman, L., 1981, Monoamine metabolites in cerebrospinal
fluid in relation to depressive illness, suicidal behaviour and
personality, in: "Recent advances in neuropsychopharmacology",
B. Amgrist, G:0. Burrows, M. Lader, 0. Lingj~rde, G. Sedvall,
and D. Wheatley, eds., Pergamon Press, Oxford.
Asberg, M., Thoren, P., Traskman, L., Bertilsson, L., and Ringberger,
V., 1976a), "Serotonin depression" - A biochemical subgroup
within the affective disorders? Science, 191: 478.
Asberg, M., Traskman, L., and Thoren, P., 1976b), 5-HIAA in the
cerebrospinal fluid. A biochemical suicide predictor?
Arch Gen Psychiatry, 33: 1193.

201
MORPHOLOGICAL BACKGROUNDS OF PATHOCHEMICAL
STUDIES IN MAJOR PSYCHOSES

Kurt Jellinger

L.Boltzmann-Institute of Clinical Neurobiology

Lainz-Hospital
A-1130 Vienna, Austria

INTRODUCTION

The morphological substrates of major psychoses are

controversial. While structural changes in the brain are
commonly present in dementias and organic psychoses, no
consistent deficits have been substantiated in schizo-
phrenia and other psychoses that are associated with a
variety of pathochemical changes. While a wide variety
of structural and cytological changes described in schi-
zophrenic brains1 are considered by most authors as non-
specific, coincidental or agonal changes unrelated to
the primary psych~sis2,3, both neuroradiological4-6 and
neuropathological studies suggest that structural defi-
cits including brain atrophy and neuronal loss and
gliosis in some brain areas may occur in some subsets
of schizophrenia, mainly in chronic cases and defect
states akin to the chronic type II syndrome of negative
symptoms6. A series of neuroradiological, CBF and histo-
logical data have been believed to indicate an "anatomic·
al locus of the pathology of schizophrenia" that has
been tentatively related to disorders of subcortico-
cortical, thalamic and mesolimbic systems1,6. It may
well be, however, that the postulated genetically de-
termined pathochemical processes do not necessarily lead
to structural deficits of the brain that can be substan-
tiated by current neuroradiology or pathologic methods4.
Two major problems will be briefly discussed:
(1) What is the current interpretation of hitherto re-
ported structural changes of the brain schizophrenia?
(2) Which morphological factors are to be considered in

203
pathochemical studies of psychoses in post-mortem brain?

NEUROPATHOLOGY OF CATATONIA AND CHRONIC SCHIZOPHRENIA

The neuropathological studies in schizophrenic

rain up to date have revealed no consistent gross or
microscopical evidence of CNS changes characteristic of
this type of disorder2,3,7. In acute lethal psychoses
including catatonia, frequent autopsy findings are brain
edema, signs of increased intracranial pressure,vascular
congestion, and occasional secondary anoxic lesions2,3.
In chronic schizophrenia either brain weight normal for
age or decreased brain weight and enlarged ventricles
have been observed1,2,6. Most of the reported cytologic-
al changes, e.g. neuronal loss and atrophy in cerebral
cortex, lipid degeneration, neuronal vacuolation and
dwarf cells in subcortical nuclei (mediodorsal thalamus,
nucleus basalis)1 have also been observed in non-psycho-
tic controls, and are being claimed either as non-speci-
fic or post-mortem changes, related to EC-therapy, extra
cranial or agonal disease2,3,7.

Neuronal loss in cerebral cortex up to 57% of con-

trols in chronic schizophrenia which is higher than in
non-demented patients but similar to senile dementia8
is considered as a morphological correlate of dementia
in defect states, while neuronal losses in the posterior
thalamus and striatum have been interpreted as hypopla-
sia or deficit of inhibitory Golgi-II microneurons9.
Thickened apical dendrites and axons in cerebral cortex
and hippocampus and hyperplasia of smooth endoplamic re-
ticulum in neurones, membrane-bound lysosomes in post-
synaptic areas, and abnormal synaptic membranes observed
in biopsy material suggest disorders of axonal and trans-
synaptic transport, and increased protein snythesis10.

Occasional glial nodules and fibrillary gliosis in

the brainstem considered as residuals of inflammatory
process or chronic viral infections1 are incidental
findings seen in non-psychiatric autopsies2,3,11. Fib-
'rillary gliosis in the diencephalon, periaqueductal
region of the brainstem, basal forebrain and mesolimbic
areas observed in up to three-quarters of schizophrenic
brains1 have not been confirmed by others2,7. In a
recent series of 90 autopsy cases of schizophrenia we
observed fibrillary gliosis in the basal forebrain and
brainstem only in 4 instances (4,2%), while another
three cases showed slight cerebellar cortical atrophy.
There is no evidence so far to support the hypothesis
of some viral agent inducing encephalitis-like disorders

204
leading to neuronal degeneration and structural changes
in the brain of schizophrenics6.
Neuronal loss in globus pallidus and substantia in-
nominata reported by several authors1,12 has not been
confirmed by recent morphometric studies in schizophre-
nic brains13. The cholinergic forebrain system represent-
ing the main cholinergic input to the cerebral neocortex
and limbic structures appears to play an important role
in cognitive and behavioural functions. In Alzheimer's
disease, senile dementia of Alzheimer type and in demen-
ted cases of Parkinson's disease there is a selective de-
generation of cholinergic neurones in the basal forebrain
with neuronal loss in the nucleus basalis of Meynert and
substantia innominata up to 70% of controls causing a
disorder of cortical cholinergic innervation which has
been related to deterioration of memory and cognition in
these disorders 13,14. Although there is some evidence
of a deficit of choline acetyl transferase in the septa-
hippocampal complex in schizophrenia15, from recent mor-
phometric studies by Arendt et al13 an personal (unpub-
lished) studies there is no evidence for a substantial
neuronal reduction in the nucl.basalis in schizophrenia.
Similar reservations are necessary with regard to often
postulated structural lesions in the nucleus accumbens
and substantia perforata anterior·l in which increased
levels of dopamine have been demonstrated in schizophre-
nia, but up to now no such changes have been confirmed.
On the other hand, a variety of organic disorders
may mimick the signs and symptoms of schizophrenia, and
there has been a considerable degree of unexpected
cerebral pathology in various samples of schizophrenic
patients examined by CT or at autopsy2,3,7. The inciden-
ce of organic brain lesions ranges from 9% in clinical
CT series to 31% in autopsy series?. They include acute
and chronic inflammatory lesions, e.g. herpes simplex
encephalitis, rheumatoid lesions, SSPE, Creutzfeldt-
Jakob disease, diffuse and multiple sclerosis, posttrau-
matic, postanoxic encephalopathies, intoxications, brain
infarction, senile and presenile atrophies, metabolic
disorders (neurolipidoses, leukodystrophies), Huntington
chorea, brain tumors and angiomas often affecting the
frontal and temporal lobes, basal ganglia and limbic
system2. In an unselected series of 101 autopsy cases
with the clinical diagnosis of schizophrenia in the pre-
CT era, we found only 24% normal brains, 32% had senile
atrophy, while 31% showed a variety of organic disorders.
The presence of such changes has to be excluded by care-
ful examination prior to pathochemical studies.

205
LIMITING FACTORS OF PATHOCHEMICAL STUDIES OF HUMAN BRAIN

When performing studies of human post-mortem brain

in search of biochemical markers in major psychoses, a
series of factors should be considered:
(1) Clinical informations, including diagnosis, classi-
fication, dementia, drug treatment, associated diseases.
(2) Circumstances and major causes of death, duration of
agony, general autopsy findings, post-mortem time,etc.
(3) Gross and histological findings of the CNS in order
to exclude organic brain lesions before forwarding the
tissue samples to the neurochemist.

The importance of neuropathology studies is demon-

strated by personal experiences in a consecutive 5-year
autopsy series of schizophrenia and other major psycho-
ses. Neuropathology studies in 95 brains using routine
methods revealed nothing abnormal in only 47,4%, while
7,3% showed either brainstem gliosis or mild cerebellar
cortical atrophy occasionally reported in chronic schi-
zophrenia1. Another 18% showed superimposed acute or
age-related change of the brain, while in 27,4% there
were other organic brain disorders, e.g. senile and pre-
senile atrophy, vascular, alcoholic or demyelinating
lesions, brain tumors (meningiomas) or Fahr's disease.
While the age at death between patients with paranoid
schizophrenia and defect states did not vary considerab-
ly in our series (mean age 62,3 and 67,1 years), the
schizophrenic patients with normal brains were older
(mean age 63,6 years) than in most other case series1,6
Patients with paranoid psychoses (mean age 76,5 years)
were considerably older than the average schizophrenics,
and often showed senile brain atrophy. Schizophrenic pa-
tients with primary or superimposed senile atrophy and
senile-vascular lesions (mean age 79,2 and 81,7 years)
were significantly older than those without pathology.
Although the relationship of both primary and superim-
posed organic brain deficits to the pathophysiology and
demonstration of pathochemical markers of schizophrenia
is unclear, all such cases were excluded from neuro-
chemical studies.

In this series of 95 cases of major psychoses the

clinical histories of 73,7% fulfilled Feighner's Re-
search Criteria16, while 26,3% did not. Among the FRC
positive cases, 57,1% had normap brain morphology, while
among those who did not fulfill FRC only 36% showed no
structural abnormalities of the brain. All the others

206
showed either superimposed and secondary CNS lesions or
other basic brain disorders, and were excluded from
pathochemical studies.
In addition to clinical, treatment, and neuropatho-
logical data, information on the premortem and agonal
states, and major causes of death as revealed by general
autopsy are essential factors for the validity of neuro-
chemical analyses in human post-mortem brain as demon-
strated by extensive correlative studies17. In our autop-
sy seris of major psychoses about 25% each died from
pulmonary embolism and pneumonia, and 10% from myocar-
dial infarction, aspiration or acute bolus death, all
preventing longstanding premortem cerebral hypoxia. In
patients dying from cardiac failure and pulmonary edema
(12%) the influence of terminal hypoxia on CNS neurotrans-
mitters and receptor kinetics cannot be definitely ex-
cluded. Patients dying from hepatic coma, uremia, severe
GI hemorrhage, acute leukemia and endocarditis (16%)
were automatically excluded from neurochemical analysis,
since the influence of these conditions on pathochemical
markers is well established. The causes of death in the
remaining cases were cancer, peritonitis, etc.
CONCLUSIONS
None of the hitherto reported structural changes in
schizophrenic brain appears to be consistent or typical
for this or other major psychoses, although it seems
likely that some morphological deficits including dila-
tation of the external and internal CSF spaces and neuro-
nal loss in some subcortical brain areas may occur in
chronic schizophrenia and defect states. The majority of
these lesions, however, are to be considered rather as
the correlates of intellectual impairment or advanced
age than of the basic disease process. In view of recent
morphometric and biochemical studies, the structure of
some subcortical areas in schizophrenic brain needs fur-
ther controlled investigation. On the other hand, the
fact that neuropathology demonstrated structural brain
deficits due to primary, secondary or superimposed orga-
nic disorders in more than 45% of an autopsy series of
major psychoses, mainly schizophrenics, precludes the
importance of exact neuromorphological examination of
the brain prior to neurochemical analysis. The frequent
occurrence of unexpected structural brain deficits, the
large variety of premortem conditions and other variab-
les possibly influencing pathochemical markers in the
brain may explain some discrepancies in recently pub-
lished data on human post-mortem brain pathochemistry.

207
REFERENCES

1. J.R.Stevens, Neuropathology of schizophrenia,

Arch.Psvchiat. (Chic.) 39,1131 (1982)
2. J.A.N.Corsellis, Psychoses of obscure pathology,
In: Greenfield's Neuropathology, W.Blackwood,
J.A.N.Corsellis,eds,E.Arnold, London (1976).
3. K.Jellinger, Zur Neuropathologie schizophrener
Psychosen, Curr.Top.Neuropath. 6,85(1980).
4. G.Gross, G.Huber, R.L.Schtitt.Ler, CT studies on
schizophrenic diseases. Arch.Psychiat.Nerven-
krankh. 231,519(1982)
5. D.R.Weinberger,L.E.DeLisi, G.P.Perman, CT in
schizophreniform disorder and other acute psy-
chiatric disorders. Arch.qen.Psvch.39,778(1982)
6. J.T.Crow, Two syndromes in schizophrenia. TINS
5,351 (1982).
7. P.Riederer, K.Jellinger, Biochemie und Morpholo-
gie der Schizophrenie. Schwerpunktmed.5,32(1982)
8. E.J.Colon, Quantitative cytoarchitectonics of
the human cortex in schizophrenic dementia.
Acta neuropath. (Berl.) 20,1 (1972)
9. R.Dom, J.De saecteleer, Quantitative cytometric
analysis in catatonic schizophrenia. Abstr.
3rd World Conqr.Biol.Psvchia~.Stockho.Lm 1982.
10. T.Miyakawa, S.Sumiyoshi, M.Deshimaru, Electron
microscopic study on schizophrenia. Acta neu-
ropath • ( Ber 1. ) 2 0 , 6 7 ( 1 9 72) .
11. L.D.Hankhoff, N.S.Peress, Neuropathology of the
brain stem in psychiatric disorders. Biol.
Psychiat. 16,945(1981) ·
12. K.Butt.Lar-Brenatno, Pathohistologische Feststel-
lungen am Basalkern Schizophrener. J.Nerv.Ment.
Dis. 116,646(1952) ·
13. T.Arendt, Y.Bigl, A.Arendt, A.Tennstedt, Loss of
neurons in the nucleus basalis of Meynert in
Alzheimer's disease, paralysis agitans and Kor-
sakoff disease. Acta neuropath.(Berl.) 61(1983)
14. J.T.Doyle, D.Pric~, M.R. DeLong, Alzheimer's
disease: a disorder of cortical cholinergic
innervation. Science 219,1184(1983).
15. E.D.Bird, E.G.S.Spokes, L.L.Iversen, Increased
dopamine concentration in limbic areas of pa-
tients with schizophrenia. Brain 102,347(1979).
16. J.P.Feighner, E.Robins, S.Guze, R.A.Woodruff,G.
Winokur, Diagnostic criteria for use in psych-
iatric reserach. Arch.qen.Psvchiat.26,57(1972)
17 D.M.Bowen, C.B.Smith,P.White, Chemical pathology
of organic dementias. Brain 100,397(1977).

208
DOPAMINE FUNCTION AND NEUROLEPTICS IN SCHIZOPHRENIA - POST-MORTEM

STUDIES OF HUMAN BRAIN TISSUE

Gavin P. Reynolds

MRC Neurochemical Pharmacology Unit, Brain Tissue Bank

Dept. of Neurological Surgery and Neurology
Addenbrooke's Hospital, Hills Road, Cambridge CB2 2QQ

INTRODUCTION

There is little doubt that biochemical investigation of human

post mortem brain tissue has made a substantial contribution to
the understanding and treatment of disease. It was as a direct
result of the observation of a dopamine deficit in the corpus
striatum of Parkinson's disease patients that L-dopa was introduced
and found to be so successful (Birkmayer and Hornykiewicz, 1962).
More recently, substantial advances in our understanding of
Huntington's chorea and senile dementia have been gained from post
mortem studies.

INCREASED DOPAMINE RECEPTORS IN SCHIZOPHRENIA

The most rigorously investigated hypothesis of the neuro-

chemistry of schizophrenia is that the disease derives from a
hyperactivity of dopamine neurotransmission. Concentrations of
the transmitter, its metabolites and associated enzymes have all
been investigated using post-mortem brain tissue from schizophrenics,
albeit with few conclusive and consistent results. Recently work
has focussed on a more consistently reported finding: an incr~ase
in the number of receptors to dopamine in post-mortem brain tissue.
Initial work from Crow and his colleagues (Owen et al., 1978)
suggested that the increase is a function of the disease process,
since they report an increase which appears independent of anti-
schizophrenic neuroleptic medication. Other groups, however, find
an increase only in drug treated patients (Mackay et al., 1982;
Reynolds et al., 1981). Certainly animal experiments also show an
increase in these receptors due to long-term neuroleptic medica-
tion (e.g. Clow et al., 1980).

209
 Table 1. (3H)spiperone binding in human putamen

B (pmol/g-1)
max

Controls (13) 22.7±1.2

Schizophrenics:
Chronic neuroleptic treatment (8) 32.2 ± 4.5*
Short term or no neuroleptic treatment (7) 17.7 ± 1.4

* p < 0.05 vs controls

In previous work we have attempted to answer the question of

whether these dopamine 02 receptors are increased due to the disease
process or to the drug treatment of schizophrenics.

Our findings are summarised in Table 1. Bmax values (which

correspond to receptor number) are significantly increased in
patients who received long-term (generally several years) neuro-
leptic treatment before death. On the other hand, we were able to
identify a group of patients, all with less than three months
neuroleptic treatment in the last year of life, who had no abnor-
mal increase in receptor number. In fact, the three patients
receiving no such treatment had a significantly lower number of
02 receptors (15.4 ± 1.7 pmol/g-1 tissue). The increase in long-
term treated patients was found to be incependent of any disease
sub-classification, i.e. whether they were paranoid (ICD 295.3) or
residual, defect state schizophrenics (ICD 295.6).

Thus the effect would appear to be independent of the disease

process and purely an effect of long-term neuroleptic medication.
Certainly there is other evidence that neuroleptics increase D2
receptor number in man. Samples we have studied from two neuro-
leptic-treated patients exhibiting the neuropathological change
of Alzheimer's disease also had a substantial and significant
increase in spiperone binding sites.

SITE SPECIFICITY OF NEUROLEPTIC ACTION

The observation that extrapyramidal side effects do not occur

to the same extent with all neuroleptic drugs is generally thought
to be due to inherent anticholinergic activity which is much
greater in some neuroleptics (e.g., thioridazine) than in others
(Snyder et al., 1974). However the theory that the mesolimbic
dopamine pathway may function abnormally in schizophrenia (Stevens,
1979) has led some workers (Borison et al., 1981) to propose a

210
 Table 2. Potencies of neuroleptics as inhibitors of specific
3H-spiperone binding in striatum (putamen) and limbic
system (nucleus accumbens)

IC 50 (nrnol/1)
Putamen N.accumbens
Thioridazine 170 200
Haloperidol 46 36
Chlorpromazine 146 140

'site specific blockade of dopamine receptors' and to claim that

thioridazine is several orders of magnitude more active on limbic
than on striatal dopa~mine receptors. We have investigated this
claim using post-mortem tissue from the accumbens and putamen of
human brain. The results, shown in Table 2, indicate no substantial
differences between the inhibition of (3H)spiperone binding to
receptors of either tissue (Reynolds et al., 1982). Thus the
evidence suggests that it is incorrect to invoke a 'site specific'
action for thioridazine on dopamine receptors in limbic areas of
human brain.

LIMBIC DOPAMINE IN SCHIZOPHRENIA

While some studies have been unable to identify a significant

increase in dopamine concentrations in post-mortem brain tissue
from schizophrenics (Crow et al., 1980), in two reports of studies
from a large series of cases (Bird et al., 1979; Mackay et al.,
1982) dopamine has been found to be increased, particularly in
limbic regions. Mackay et al. (1982) find this increase to be most
marked in younger patients and not obviously related to neuroleptic
medication.

As mentioned above, the limbic system is particularly impli-

cated in schizophrenia (Stevens, 1979). The temporal lobe is
possibly involved considering the psychotic behaviour often seen
in temporal lobe epilepsy (Flor-Henry, 1969). One limbic region
of the temporal lobe which receives a dopaminergic innervation is
the amygdala which has recently been studied in post-mortem brain
tissue: a substantial increase in dopamine concentrations has been
found. Table 3 shows this to be regionally and biochemically
specific since caudate dopamine and amygdala noradrenaline are
not significantly changed. It is, however, unclear whether
neuroleptic medication contributes to this effect. Finally, it
should be mentioned that preliminary results show the increase to
be restricted to the left temporal lobe (Reynolds, 1983), an

211
 Table 3. Catecholamines in post-mortem brain tissue

Controls Schizophrenics
(n=l9) (n=22)

Caudate
Dopamine 1777 2332

Amygdala
Dopamine 43.0* 72.9*
Noradrenaline 61.7 63.4

Values are geometric means in ng/g tissue.

* p < 0.01 by t-test of log-transformed data, other control-
schizophrenic comparisons are not significant (p > 0.05).
From Reynolds (1983).

observation which is the first biochemical support for the wide

range of psychological and neurophysiological evidence for temporal
lobe lateralisation in schizophrenia (reviewed by Gruzelier, 1981).
Thus it would appear that further study of mesolimbic dopaminergic
pathways, particularly those projecting to the amygdala, is
indicated and may eventually serve to combine the various bio-
chemical, neurophysiological and psychological interpretations of
the dysfunction in schizophrenia into a single comprehensive
hypothesis.

REFERENCES

Bird, E.D., Spokes, E.G.S. and Iversen, L.L., 1979, Increased

dopamine concentration in limbic areas of brain from patients
dying with schizophrenia, Brain, 102:347.
Birkmayer, w. and Hornykiewicz, 0., 1962, Der L-3,4-Dioxyphenyl-
alanin (L-Dopa)-Effekt beim Parkinson-Syndrom des Menschen,
Arch. Psychiatr. Nervenkr., 203:560.
Borison, R.L., Fields, J.Z. and Diamond, R.I., 1981, Site specific
blockade of dopamine receptors by neuroleptic agents in
human brain. Neurooharmacol., 20:1321.
Claw, A., Theodorou, A., Jenner, P. and Marsden, C.D., 1980,
Changes in rat striatal dopamine turnover and receptor
activity during one year's neuroleptic administration, Eur.
J. Pharmac., 63:135.

212
Crow, T.J., OWen, F., Cross, A.J., Ferrier, N., Johnstone, E.C.,
McCreadie, R.M., Owens, D.G.C. and Poulter, M., 1981,
Neurotransmitter enzymes and receptors in post-mortem brain
in schizophrenia: evidence that an increase in D2 dopamine
receptors is associated with the type 1 syndrome, in:
Transmitter biochemistry of human brain tissue, P. Riederer
and E. Usdin eds., Macmillan, London, P.85.
Flor-Henry, P., 1969, Psychosis and temporal lobe epilepsy: a
controlled investigation, Epilepsia, 10:363.
Gruzelier, J.H., 1981, Cerebral laterality and psychopathology:
fact and fiction, Psychological Med., 11:219.
Mackay, A.V.P., Iversen, L.L., Rossor, M., Spokes, E., Bird, E.,
Arregui, A., Creese, I. and Snyder, S.H., 1982, Increased
brain dopamine and dopamine receptors in schizophrenia,
Arch. Gen. Psychiatr., 39:991.
Owen, F., Crow, T.J., Poulter, M., Cross, A.J., Longden, A. and
Riley, G.J., 1978, Increased dopamine-receptor sensitivity
in schizophrenia, Lancet, ii:223.
Reynolds, G.P., 1983, Increased concentrations and lateral
asymmetry of amygdala dopamine in schizophrenia, Nature (in
press) .
Reynolds, G.P., Cowey, L., Rossor, M. and Iversen, L.L., 1982,
Thioridazine is not specific for limbic dopamine receptors,
Lancet, ii:499.
Reynolds, G.P., Riederer, P., Jellinger, K. and Gabriel, E., 1981,
Dopamine receptors and schizophrenia: the neuroleptic drug
problem, Neurooharmacol., 20:1319.
Snyder, S., Greenberg, D. and Yamamura, H.I., 1974, Anti-
schizophrenic drugs and brain cholinergic receptors, Arch.
Gen. Psychiatr., 31:58.
Stevens, J.R., 1979, Schizophrenia and dopamine regulation in the
mesolimbic system, Trends Neurosci._, 2:102.

213
NEUROENDOCRINE AND RECEPTOR BINDING STUDIES IN SCHIZOPHRENIA

M. Ackenheil, M. Albus, B. Bondy, F. MUller-Spahn,

U. MUnch and D. Naber

Psychiatric Hospital of the University of Munich

Nussbaumstr. 7, D-8000 Munich 2, F.R.G.

Dopamine agonists such as L-DOPA and amphetamine have been

found to exacerbate schizophrenia, while neuroleptics, dopamine (DA)
antagonists, have antipsychotic effects. These results of many clini-
cal studies sustain the DA hypothesis of schizophrenia.
The increased knowledge of neuroendocrinology and receptor bin-
ding mechanism extends research strategies. This article reports
several studies on the biochemical approach to investigate the patho-
physiology of schizophrenia.

NEUROENDOCRINE STUDIES

Disturbances of the dopaminergic system in the hypothalamic-

pituitary-axis can be evaluated by the apomorphine test, since the
stimulation of growth hormone (GH) secretion after apomorphine inject-
ion (0,5 mg s.c.) depends on the sensitivity of DA receptor. 1
In untreated acute schizophrenic patients GH response after
apomorphine was significantly elevated compared to controls (Fig. 1).
These results are in agreement with Rotrosen et al. 2 Stimul -
ated GH levels did not correlate with psychopathological scores before
or after neuroleptic treatment. Haloperidol treatment (20-40 mg/day)
significantly reduced GH response in all patients, indicating DA re-
ceptor blockade.
In 15 chronic schizophrenic patients, treated for 12 ± 6 years
with neuroleptic drugs, the apomorphine test was repeated during
treatment (day 0), after 12 (day 12) and 30 (day 30) days of with-
drawal. The mean response of GH to apomorphine on day 12 (15,9 ± 12,3
ng/ml) was significantly elevated in comparison to day 0 (7,8 ± 8,9
ng/ml), but lower than the GH response in acute schizophrenics and
controls (Fig. 1). Again, these findings agree with previous reports~• 4

215
On day 30, about 50% of the patients had a GH response ~5 ng/ml;
because of a high variation levels did not differ significantly from
those of day 0 and day 12.
Correlating psychopathologic al scores with maximal GH levels,
some significant relationships have been found. However, no obvious
pattern was revealed, the relevance of these correlations needs fur-
ther confirmations.
Stimulation of GH by clonidine, an alpha-2 agonist, has been

APOMORPHIN E- TEST

A NL- therapy
~ ~~:::days drug free

•
-, E
:I
Q)
Ul
35
E
.....
~25
Ul
~
tel
Q)
a.

I
(!)
5

A B C
controls acute chronic
schizophrenics

Fig. 1
Apomorphine stimulation (0,5 mg s.c.) of GH- secretion in schizo-
phrenic patients (GH maxima) (n=10)
Controls (age and sex matched) (n=14)
Chronic schizophrenics (n=15)
Day 0: during longterm neuroleptic treatment; day 12: after 12 days
neuroleptic withdrawal; day 30: after 30 days neuroleptic withdrawal.

216
demonstrated to be a useful tool measuring central alpha-adrenergic
receptor sensitivity. 5 This neuroendocrine test was carried out in
acute schizophrenic patients. The mean maximal GH stimulation was
slightly higher in schizophrenics than in controls; both groups had
significantly elevated levels compared to those of schizoaffective
patients. (Fig. 2).
Within the group of schizophrenics, there was a high variation
of GH serum levels ranging from 1,2 to 47 ng/ml. Schizophrenics and
schizoaffective patients did not differ in severity of psychopatho-
logical scores. Thus, it is intelligible that no correlations between
psychopathology and GH response were found.
Acute neuroleptic treatment did not change the maximal GH se-
cretion (before treatment 12,0 ± 14,6 ng/ml; after treatment 10,1 +
10,4 ng/ml).
The effect of long-term neuroleptic treatment on clonidine-
induced GH secretion was studied in 26 chronic schizophrenics,
treated for 13 ± 5 years. (Fig. 3). Compared to acute treated
patients, matched for sex and age, chronic patients showed a signi-
ficantly diminished GH response (9,7 ± 9,0 ng/ml vs. 4,8 ± 6,0 ng/ml).
Discontinuation of the neuroleptic treatment for 5 and 12 days did

CLONIDINE TEST WITH SCHIZOPHRENIC PATIENTS

30 GH
ng /ml

20 -·-.CONTROL
---SCHIZO.
-SCHIZOAFFECT.

/
/
/

10

I I ________..._--.J

-10 -30 0 30 10 10 120 1 (nun I

Fig. 2
GH stimulation with clonidine (0,15 mg i.v.)
Acute schizophrenics (13), schizoaffectives (9) compared to age and
sex matched controls (23)
p < 0,05; U-Test Mann-Whitney. Schizoaffectives vs. controls; schizo-
affectives vs. schizophrenics.

217
not induce significant changes of stimulated GH secretion, although
most of the patients showed distinct changes of psychopathological
scores.

RECEPTOR BINDING STUDIES

The normal function of aminergic neurons depends on the inter-

CLONIDINE- TEST

A NL- therapy
55 B 5 days drug free
C NL-therapy

..
0 30 days drug free

45
E
:::l
Ql
.
I
til 35
E

I
~ 25
til
..X

~ 15
a.

control acute
It£
A
chronic
B
chronic
c 0

SChiZOphreniCS

Fig. 3
GH- sti~ulation with clonidine (0,15 mg i.v.)
Acute schizophrenics (13) compared to chronic schizophrenics (30)
A and C during long-term NL treatment; B after 12, D after 30 days
withdrawal.
Age and sex matched controls (13).

218
action between transmitter-release, transmitter-turnover and recep-
tor sensitivity. A new method permitted the sim~ltaneous determin-
ation of alpha-2-adrenergic binding sites with H 3 yohimbine on
platelets, of beta-2-adrenergic binding sites with H - dihydroxy -
alprenolol on granulocytes and of DA-2-binding sites with 3 H -
spiroperidol on lymphocytes, thus offering a further possibility to
evaluate disturbances of rec~ptor sensitivity. Receptor binding and
serum levels of DA, norepinephrine (NE) and epinephrine (E) were
measured in a group of 21 acute untreated schizophrenic patients
and matched controls. (Fig. 4). The affinities of all 3 binding
sites did not differ between patients and normals, but there were
significant differences in the number of receptors: B max of patients
was reduced with regard to alpha-2-(0,41 ± 0,13 fmol/10 6 cells vs.
0,52 ± 0,14 fmol/10 6 cells; p < 0,05) as well as to beta-2-adrener-
gic receptor binding (1,6 ~ 0,7 fmol/10 6 cells vs. 2,8 ~ 0,6 fmol/
10 6 cells; p~ 0,01), while the capacity of the DA receptor was
5-fold elevated (14,4 ~ 9,3 fmol/10 6 cells vs. 2,5 ± 0,8 fmol/10 6
cells; p <: 0,0001).
Serum levels of NE and DA were both significantly elevated in
schizophrenic patients (NE 493 ~ 349 pg/ml vs. 210 ~ 40 pg/ml
p < 0, 0001; DA 50 ~ 47 pg/ml vs. 18 ~ 9 pg/ml, p ..::::. 0, 001) , but were
not significantly correlated to psychopathology or to receptor bind-
ing capacity.

z m

~l:J-i
.
0 ..,
0 ~-
0 0 CD ::I

I...
g CD
~- 0 "0 ~ • 3
CD ::::I
§ ..,::r -·
8~ -;·
••
0 ::I
N
::r ID
0 .., 0 I!! CD
••
0 ....
.• --
00 ,....
...
::I
CD "0

,..., §"' ,...,~

"0
0 "'§ • 0 0
03
0
"0
.....

-
0 -CD
0 Qj
Q)
.....
••
-3 Ql
• ......
•••
0 0
3
0
..... 8- .....0 8
(/)g
(UN ••
•
0
go
........ :s •I v (/)

..!
•
~

::I
•
8 -o
::::t~
• ::I 0

..
0
Ql 0
0
..! .!!§
o-.
...

~
~
,:) 0
0 0
Eo • ...8 .;!il •• Ul
0 E
0

~
...... ~ u
• 0
)(
Ill
E
0 )(
Ill
E
)(
Ill
E
~
m m m
a2
receptors NE rec/11ors E
DA
receptors DA
platelets serum granulocytes serum lymphocytes serum

Fig. 4
Alpha-2, beta-2-adrenergic, DA receptor binding to human blood cells
~controls, mean ± SD; • acute unmedicated schizophrenic patients.

219
CONCLUSIONS

The neuroendocrine studies show some marked differences between

acute or chronic schizophrenic patients and normal controls with re-
gard to the hypothalamic DA and the central alpha-adrenergic receptor.
Both catecholamine receptor systems retain a high responsiveness
even after long-term neuroleptic therapy and short-term drug with -
drawal. The high interindividual variation might indicate different
subtypes of schizophrenia.
The binding studies suggest a possible disturbance between
catecholamine binding sites and the circulating transmitters. Fur-
thermore, the marked elevation of spiperone binding to lymphocytes
of patients might be valuable as diagnostic criteria and marker in
schizophrenia.

REFERENCES

1. M. Ackenheil: Biochemical effects of apomorphine: contribution

to schizophrenia research. In: Apomorphine and other dopaminer-
gics. Vol. II Corsini GV, Gessa GL (eds) pp 215-225. Raven Press
New York (1981)
2. J. Rotrosen, B.M. Angrist, S. Gershon, E.J. Sachar, F.S. Halpern:
Neuroendocrine assessment of dopaminergic activity in schizo -
phrenia. In: Advances in Biochemical Psychopharmacol. Vol. 16
Gessa GL, Costa E. (eds) pp 112-118. Raven Press New York (1977)
3. P. Ettigi, N.P.V. Nair, S. Lal, P. Cervantes, H. Guyda: Effect
of apomorphine on growth hormone and prolactin secretion in
schizophrenic patients, with or without oral dyskinesia, with -
drawn from chronic neuroleptic therapy. J Neurol Neurosurg
Psychiat 39:870-876 (1976)
4. G. Pandey, D. Farver, C. Taminga, S. Ericksen, S. Ali, J.M. Davis:
Postsynaptic supersensitivity in schizophrenia; Am J Psychiat
134:518-522 (1977)
5. S. Lal, G. Tolis, J.B. Martin, G.M. Brown, H. Juyela: Effect of
clonidine on growth hormone, prolactin, luteinizing hormone,
follicle stimulating hormone, and thyroid-stimulating hormone
in the serum of normal men. J Clin Endocrinol Metab 41:827-836
(1975)

ACKNOWLEDGMENT

This study was supported by "Deutsche Forschungsgemeinschaft"

220
DECREASED CSF NOREPINEPHRINE AND MONOAMINE METABOLITES IN SCHIZO-

PHRENIC PATIENTS WITH BRAIN ATROPHY AS SHOWN WITH CT SCANS

*Daniel P. van Kammen, Lee S. Mann, Mika Scheinin,

Philip T. Ninan, Welmoet B. van Kammen, & Markku Linnoila

*Western Psychiatric Institute and Clinic

University of Pittsburgh, School of Medicine
3811 O'Hara Street
Pittsburgh, Pennsylvania 15213

Spinal fluid levels of homovanillic acid (HVA), 5-hydroxy-

indoleacetic acid (5-HIAA), 3-4 dihydroxyphenylacetic acid (DOPAC),
norepinephrine (NE) and 3-methoxy-4-hydroxyphenylglycol (MHPG)
were evaluated in off-medication chronic schizophrenic patients
with (N=B) and without (N=43) evidence of brain atrophy of
computerized tomography (CT) scans. Patients with brain atrophy
showed significantly lower levels of these spinal fluid consti-
tuents except MHPG. These CSF findings support the proposed sub-
division of schizophrenia with and without brain atrophy. Future
CSF studies need to take brain atrophy into account among
variables as diet, age, sex, sleep (hours), drug conditions,
clinical state, anxiety and activity.

221
ROLE OF TESTOSTERONE IN MALE SEXUAL IMPOTENT PATIENTS

O.Benkert and~ Holsboer

Department of Psychiatry
University of Mainz, Mainz, FRG

!.INTRODUCTION

The role of biological factors contributing to development of

sexual dysfunction (SD) has given rise to controversy over many
years. Whereas a broad consensus is achieved that schizophrenia
and affective psychoses have a biological etiology, such ·mechan-
isms receive less acceptance for SD. Conversely, the observation
that SD is frequently a concomitant factor in affective illness
is subject of current interest in research. For both diseases SD
and affective disorder impaired gonadal function is documented
(Rubin et al., 1981; Benkert and Holsboer, 1984), whereas both
disorders may well occur independantly from each other. In the
following section we outline briefly some of the presently
available data on biological indices of SD. Some of the forth-
coming research strategies are illustrated from which the emer-
gence of a timely and appropriate classification of SD is expec-
ted.

II. TESTOSTERONE AND SD

Hypogonadismus as evidenced by decreased testosterone secreting

capacity is followed by SD which can be restored by exogenous
testosterone-supplementation (Davidson et al., 1979; Spark et
al., 1980; Salmimies et al., 1982). Also, the mean testosterone
plasma contents were found to be significantly and negatively
correlated to maximum penile tumescence after visual erotic
stimuli (Lange et al., 1980). Besides a clinically manifest
hypogonadism there exists no clearcut relationship between

223
sexual activity and testosterone plasma concentrations (Brisson
et al., 1977). Nocturnal penile erections coincide with rapid-
eye-movement periods of sleep (Fisher et al., 1965; Karacan et
al., 1966). Penile nocturnal parameters as monitored with penis-
plethysmography, however, are not correlated to any changes of
testosterone secretion during night (Roffwarg et al., 1982).

In psychiatric patients a disturbed sleep-architecture, abnormal

secretion patterns of almost all hormonal axes and altered
responses to endocrine perturbation tests are well known. Within
such a population SD may etiologically be derived from psycho-
tropic drug~ (particularly through the anticholinergic or anti-
dopaminergic component) or from the psychiatric disease itself.

Study design

To further investigate the relationship between gonadal testo-

sterone secretion and sexual performance we investigated 117
consecutively admitted inpatients. Due to their psychiatric
disease 27 patients were unable to answer questions concerning
their sexual activity. Another 15 patients refused their consent
to participate. The remaining sample of 75 was attributed to the
group of patients with SD if they met the following criteria:

1. Difficulty in achieving or maintaining erection during inter-

course.
2. Persistent dysfunction for at least one year due to an un-
known etiology.
3. A prior history of normal penile erection
4. A regular female sexual partner
5. At least one attempt at performing coitus per week.
6. Normal findings in andrological investigation

Blood samplings were drawn between 7.00h and 8.00h and prolac-
tin, LH, FSH and testosterone were analysed by radioimmunoassay.
The total group was split into 31 patients with SD and compared
with an age matched comparison group of 25 patients without SD.

Results

The mean testosterone plasma concentration in patients with SD

was 506 +/- 157 ng/100ml and in patients without SD 709 +/-213
ng/100ml(p 0.001). If the unmedicated subsamples were extracted
also a significantly lower (p 0.05) mean testosterone plasma
concentration (529 +/- 179 ng/100ml) was seen in patients with
SD when compared to patients without SD (712 +/- 215 ng/100ml).
The prolactin concentration in patients with SD was not statis-
tically (t-test) different. The same was true for the gonadotro-
pins LH and FSH.

224
III. DISCUSSION

The finding of lowered testosterone plasma concentration in male

psychiatric patients with SD adds to previous observations which
indicate an association of decreased gonadal activity and im-
paired sexual function. It has to be recognized, however, that
the majority of this patient sample with SD showed testosterone
plasma concentration within the normal range: 300-1000 ng/100ml.
Only 4 patients with SD and one patient without SD had values
beyond this cutting point. A testosterone plasma concentration
at the lower end of the normal range may well contribute to the
development of SD. However, other precipitating factors, e.g. a
partner conflict or transient abuse of alcohol, may also be
present. The whole number of causative factors which result into
SD are not adequately recognized within the traditional diagnos-
tic dichotomy which differentiates between organic and psycho-
genic impotence. This bimodal conception depends strongly upon
the users theoretical and therapeutical preoccupations. An orga-
nically oriented investigator will possibly search more rigo-
rously than a psychoanalyst for subtle biological factors, which
can be considered to be of etiological importance. Such a con-
cept which allows for a neglect of etiologically important
biological indices prevents to derive a therapeutic strategy
from the diagnostic process.

Diagnostic classification

A multiaxial classification of SD seems to be more appropriate

since it allows to incorporate biological and psychological
mechanisms. We propose a classificatory system which divides SD
into the following subtypes:

Subtype Ia organic disorder*:

SD with the simultaneous presence of a defined organic disor-
der which may be of relevance for the occurence of the SD.
Subtype Ib physical condition*:
SD in the absence of subtype Ia and in the presence of un-
usual laboratory tests (e.g. low testosterone, abnormal NPT,
see figure 1) •
Subtype IIa psychiatric disorder:
SD in the presence of a defined (i.e. DSM III) psychiatric
disorder including personality disorder.

* The terms "organic disorder" and "physical condition" are used

in accordance with the DSM III-system.

225
 Simultaneous Recording of EEG-NPT-and Neuroendocrine Data

EEG AW~~~3------------lHJ~J~ n . r JlJU1 r~ nJ

~~~ ... a.r -.uy 11
22 23 24 2 3 4 5 6

NPT

'f~
PRL

am~~"''} l A . I
4
I
5 6
I •

1~L-~·jr'~fo:::::!"!:::::!•~~=..:.._-.:r=:::!:=f-..------r-~----r-1 -

1\
22 23 24 1 2 3 4 5 6
HGH
~~og/ml
~~ l~f
:--: ~ :.
T 1•

r~·::':'
LH

FSH 22 23 21. 1 2 3 4 5 6
1~ mU/ml

~ ;=-;=;
;:=; ·=;
I •
TESTO 1 2 3 4 5 6

Figure 1: Simultaneous recordings of neurophysiological and

endocrine data during sleep of a patients suffering
from SD.

226
Subtype IIb psychosocial condition:
SD in the presence of psychosocial stress situation (e.g.
partner conflict) and in the absence of subtype IIa.
Subtype III pharmacological condition:
SD in parallel with the intake of drugs (medical or alcohol
etc) insofar as these are most likely of relevance for the
occurence of SD.
Subtype IV cryptogenic disorder:
SD which can not be classified in any of the above mentioned
subtypes.

Patients with SD may on occasion display subtype Ia or Ib to-

gether with subtype IIa or IIb and subtype III.

Subtype Ib is of particular interest for biological research in

SD. It necessitates developement of subtle laboratory techniques
to detect minor biological aberrations of relevant functions. We
have tentatively elaborated a protocol which allows to collect
simultaneously series of relevant neurophysiological (sleep-EEG,
NPT) and endocrine data( prolactin, cortisol, growth hormone,
LH, FSH, testosterone) over time (see figure 1).

Such a strategy which incorporates an adequate biological data

base into an intregrative multiaxial classification system of SD
holds promise to be helpful for researchers and clinicians.

Acknowledgement
The study was supported in part by the Deutsche
Forschungsgemeinschaft (Ho 940/1-1).

REFERENCES

Benkert, 0., and Holsboer, F., 1984, Sexual dysfunction in male

patients: Biological indices and classification In: Psycho-
pharmacology of Sexual Disorders (M. Segal ed.), Libbey &
Company Ltd, London, in press.
Brisson, G.R., Volle, M.A., Desharnais, M., Dion, M., and Tana-
ka, M., 1977, Pituitary gonadal axis in exercising men.
Med. Sc. Sp. 9: 47.
Davidson, J.M., Camargo, C.A., and Smith, E.R., 1979, Effects of
androgen on sexual behavior in hypogonadal men. Journal of
Clinical Endocrinology and Metabolism 48: 955
Fisher, C., Gross, J., and Zuch, J., 1965, A cycle of penile
erection synchronous with dreaming (REM) sleep. Archives
of General Psychiatry 12: 29.

227
Karacan, I., Goodenough, D.R. Shapiro, A., and Starker, S.,
1966, Erection cycle during sleep in relation to dream
anxiety. Archives of General Psychiatry 15: 183.
Lange, J.D., Brown, W.A., Winzce,J.P, and Zwick, W., 1980, Serum
testosterone concentration and penile tumescence changes in
men. Hormones and Behavior 14: 267.
Roffwarg, H.P., Sachar, E.J., Halpern, F., and Hellman, L.,
1982, Plasma testosterone and sleep: relationship to sleep
stage variables. Psychosomatic Medicine 4: 73.
Rubin, R.T., Poland, R.E., Tower, B.B., Hart, P.A., Blodgett,
A.L.N., and Forster, B., 1981, Hypothalamo-pituitary-
gonadal function in primary endogenously depressed men:
Preliminary findings. In: Steroid Hormone regulation of
the Brain (K.Fuxe, ed.)Pergamon Press, Oxford,New York,
p387.
Salmimies, 0., Kockott, G., Pirke, K.M., Vogt, H.J., and Schill,
W.B., 1982, Effects of testosterone replacement on sexual
behavior in hypogondal men. Archives of Sexual Behavior 11:
345.
Spark, R.F., White, R.A., and Connolly, P.B., 1980, Impotence is
not always psychogenic. Newer insights into hypothalamic-
pituitary-gonadal dysfunction. Journal American Medical
Association 243: 750.

228
PSYCHOTROPIC EFFECT OF COMBINED ESTROGEN-VIT B6 TREATMENT IN

ENDOGENOUSLY DEPRESSED FEMALES

F. Holsboe~, 1. Meier, A. Kreuz and 0. Benkert

Department of Psychiatry, Research Unit

University of Mainz, Mainz, FRG

INTRODUCTION

Estrogens are among the most commonly prescribed substances

in females. Also endogenous estrogen levels change dramatically
throughout life and this biological variable has been associated
with several psychological signs like premenstrual tension syn-
drome and depression in older age. Along with clinical practice
there is increasing evidence from neuropharmacology suggesting a
psychotropic action of estrogens (review: Holsboer, 1982). The
most prominent findings are:
1) Reduction of monoamine-oxidase (MAO)-activity by estrogens
(McEwen et al., 1978);
2) Competitive inhibition of catechol-o-methyltransferase by 2-
hydroxyestrogens, which are major metabolites of estrogens in
the CNS (Breuer et al., 1978)
3) Estrogens seem to modulate the dynamic range of dopamine
autoreceptors. DiPaolo et al. (1984) have postulated that
estrogens may act as weak neuroleptics. By stimulation of
dopamine (DA)-autoreceptors they may enhance dopaminergic
transmission and down-regulation of presynaptic DA receptors
This mode of action is also reported for antidepressants by
some (Serra et al., 1981; Antelman and Chiodo, 1981) but not
all (Holcomb et al., 1982) researchers;
4) Some recent reports indicate that the cholinergic (Miller,
1983) and serotonergic (DiPaolo et al., 1983) systems are also
affected by estrogens.

229
These pharmacological data together with the widespread
acceptance of estrogen application in practice encouraged us to
further investigate the psychoactive profile of estrogens in
affective disorder.

METHODS
20 female patients suffering from a major depressive disorder,
endogenous subtype, were studied. If the menses has ceased for
more than five cycles the patient was attributed to the post-
menopausal (n=lO) group. Pretreatment with MAO-inhibitors or
neuroleptics were exclusion criteria, all other medications were
withdrawn at least four days and substituted by placebo prior to
treatment with 60 ug ethinylestradiol (EE2). To prevent a
functional Vit B6 deficiency (Adams et al., 1973) a daily dose of
120 mg pyridoxalphosphate was given. No other psychotropic drugs
including benzodiazepines were allowed. All patients except those
with a hysterectomy were cycled with 10 mg norethisterone during
the last five days of the four week treatment period. Hamilton
Rating Scores were collected weekly. A major goal of the present
study was to establish a set of biological indices which would
help to positively identify prospective hormone treatment respon-
ders. For that purpose estradiol, estrone, testosterone,
~4-androstendione and platelet MAO activity were assessed; in
addition, dexamethasone suppression tests (1 mg version) were
performed.

RESULTS
Similar to many pilot studies a significant (p < 0.05) decline of
total Hamilton Scores in comparison to baseline was observed. The
overall improvement in the postmenopausal group (22.5 +/- 3.1--.
8.6 +/- 7.6) was more marked than in the premenopausal group
(25.2 +/- 4.8 __.13.8 +/- 9.3). This is further documented by
individual data: 7 postmenopausal females responded well (HRS < 8)
to estrogens and only one case remained unchanged, whereas only 4
out of 10 premenopausal patients had a final score lesser than 8
points on the Hamilton-scale.
Neither testosterone nor ~4-androstendione-plasma-concentrations
differentiated betweeen responders and non-responders (testo:
39.4 +/- 31.8 ng/ml for responders and 28.5 +/- 9.5 ng/ml for
non-responders; androstendione: 88.3 +/- 70.7 ng/ml for respon-
ders and 117.8 +/- 65.3 ng/ml for non-responders). As expected,
no difference of estradiol plasma-concentrations between respon-
ders and non-responders is seen in the total group (82.7 +/- 87.9
pg/ml for responders and 59.3 +/- 39.5 for non-responders). In-
spection of the postmenopausal subsample revealed the tendency
toward lower estradiol-values in responders (41.6 +/- 13.8 pg/ml)

230
in comparison to non-responders (79.0 +!- 59.0 pg/ml). 8 out of
12 patients who were positive on the DST responded favourably
whereas 5 non-responders showed adequate DST suppression.
Platelet (substrate: p-tyramine and benzylamine) and plasma MAO-
activity were significantly attenuated by the EE2/B6-treatment
(MAO: p-tyramine: 40.7 +/- 12.7-18.0 +/- 10.5 nmol/mg pro-
tein/h, p/ 0.01; benzylamine: 27.5 +/- 8.9-11.5 +/- 8.2
nmol/mg protein/h, p<'O.Ol; plasma: 29.4 +/- 7.9-24.7 +/- 5.8
units/min; p< 0.05). The decrease of platelet MAO-activity
was significantly (p < 0.05, t-test) more marked in responders
(benzylamine: 21.2 +/- 8.9, p-tyramine: 29.0 +/- 15.1 nmol/mg
protein/h) than in non-responders (benzylamines: 10.1 +/- 8.9,
p-tyramine: 15.0 +/- 12.7 nmol/mg Protein/h).

DISCUSSION
The clinical result suggests an antidepressive effect of an
EE2/B6 treatment modality in female patients with lowered gonadal
function. However, no firm conclusion may be drawn from a clini-
cal pilot study even if a diagnostically homogenous sample is
studied.
From hormonal investigations prior to treatment no indication
which patient may respond can be obtained. However, this finding
is derived from a single blood specimen. A more thorough endo-
crine protocol including collections over a longer time period
may possibly reveal that a particular hormonal profile indicates
a beneficial outcome after a hormonal treatment. The mechanism by
which EE2/B6 exerts a psychotropic effect in depressed patients
remains unknown. Our finding of a pronounced decrease of MAO-
activity in responders suggests that the influence of estrogens
upon this enzyme system is at least partially involved.

REFERENCES

Adams, P.W., Wynn, V., Rose, D.P., Seed, M., Folkard, J., and
Strong, R., 1973, Effect of pyridoxine hydrochloride (vitamin
B6) upon depression associated with oral contraception. The
Lancet, 28: 897.
Antelman, S.M., and Chiodo, L.A., 1981, Dopamine autoreceptor
subsensitivity: a mechanism common to the treatment of de-
pression and the induction of amphetamine psychosis? Biologi~
cal Psychiatr~, 16: 717.
Breuer, H., Koster, G., Schneider, H.T., and Ladosky, W., 1978,
Interactions between estrogens and neurotransmitters: Effect
of estrogens on the enzymatic methylation of noradrenaline in
brain. In: Brain-Endocrine Interaction III. Neural Hormones
and Reproduction, D.E.Scott, G.Kozlowski, and A. Weindl,
eds., p274. Karger Verlag, Basel.

231
DiPaolo, T., Daigle, M., Picard, V., and Barden, N., 1983, Effect
of acute and chronic 17B-estradiol treatment on serotonin and
5-hydroxyindole acetic acid content of discrete brain nuclei
of ovariectomized rat. Experimental Brain Research 51: 73.
DiPaolo, T., Bedard, P., and Labrie, F., 1984, Neuroleptic-like
activity of estrogens. VII. World Congress of Psychiatry,
Plenum Press, New York- London-Washington D.C.-Boston, in
press.
Holcomb, H.H., Bannon, M.J., and Roth, R.H., 1982, Striatal
dopamine autoreceptors unifluenced by chronic administration
of antidepressants. European Journal of Pharmacology, 82: 173.
Holsboer, F., 1982, Hormones. In "Psychopha~'macology", Vol. 1,
Part 2: Clinical Psychopharmacology, H. Hippius, and G. Wino-
kur, eds., p. 144. Excerpta Medica, Amsterdam-Oxford-Prince-
ton.
McEwen, B.S., Krey, L.C., and Luine, V., 1978, Steroid hormone
action in the neuroendocrine system: When in the genome
involved? In: "The Hypothalamus", R.J. Baldessarini, J.G.
Martin, eds., p. 255, Raven Press, New York.
Miller, J.C., 1983, Sex differences in dopaminergic and choliner-
gic activity and function in the nigro-striatal system of the
rat. Psychoneuroendocrinology, 8: 225.
Serra, G., Argiolas, A., Fadda, F., Melis, M.R., and Gessa, G.,
1981, Psychopharmacology, 73: 194.

232
OESTROGEN MODULATION OF DOPAMINE RECEPTORS: AUTORECEPTOR
IMPLICATION FOR ANTIDEPRESSANT ACTIVITY

M.P. Piccardi, M. Del Zompo, M. Meloni, and G.U. Corsini

Dept. of Clinical Pharmacology
University of Cagliari
Via Parcell 4, 09100 Cagliari, Italy

Biochemical and pharmacological evidence in animals indicates

that oestrogens are capable of substantially modulating post-synap-
tic dopamine (DA) receptor sensitivity (1,2). Recent physiological
data, indicating that oestrogens significantly attenuate the abili-
ty of apomorphine to inhibit the firing activity of type B dopamine
neurons in the rat substantia nigra, suggest that these hormones
may also induce a subsensitivity of DA autoreceptors (3).
According to this hypothesis oestrogen-treated male rats showed
no remarkable decrease of motility counts when challenged with apo-
morphine (20 - 50 pg/kg) from 24 to 72 hours after the last hormone
administration. Similarly, the apomorphine (20 ~g/kg) induced de-
crease of dihydroxyphenylacetic acid (DOPAC) levels in the caudate
nucleus in oil-treated control rats was almost completely counter-
acted by oestrogen treatment. Various dopamine agonists such as
2-~-bromocriptine, Piribedil, 3 Hydroxyphenylpropylpiperidine and

n-propylnorapomorphine, failed to induce hypomotility when admini-

stered to oestrogen-treated rats unlike in oil-treated animals.
These behavioural and biochemical results, along with the electro-
physiological data, indicate that oestrogen treatment is able to in-
duce a hyposensitivity of DA autoreceptors in male rats (4,5).
The effect of oestrogen pretreatment (17 p Estradiol benzoate,
10 pg/kg s.c. daily for 3 days) on the hypomotility induced by a
low dose of apomorphine (20 pg/kg s.c.) was studied in female rats.
Apomorphine was administered to female Sprague-Dawley rats (Charles
River- ITALY) treated with sesame oil (0.3 ml s.c.) or oestrogens

233
24 hours after the last hormone administration. The animals were
counted for motility with respect to the animals treated with saline.
Although oestrogen pretreatment did not significantly modify the
spontaneous motor behaviour with respect to saline treated rats,
apomorphine induced a decrease (55%) in motility in oil treated ani-
mals. However, this effect was less marked than that observed in
male rats and a large variability of response is documented by a
high standard error.
Oestrogen pretreatment failed to significantly prevent the apo-
morphine-induced hypomotility in female rats (p = n.s.); in fact,
a very high degree of variability of response was observed in these
animals. Similar findings were obtained by studying DOPAC content
in striatum after apomorphine administration. In female rats, DOPAC
levels decreased by 27% after apomorphine but the standard error
was up to ~ 0.24 pg/g, indicating a marked variation of response.
Oestrogen pretreatment did not significantly affect this biochemi-
cal parameter, the difference with oil pretreated animals not being
statistically significant (p = n.s.). These data indicate that, un-
like in male rats, apomorphine is variably effective in decreasing
motility and striatal DA turnover in female rats, suggesting that
hormonal factors linked to oestrous phases in these animals may
account for the observed variations.
To ascertain the importance of gonadal influence on these actions
of dopaminomimetics in female rats, we studied the effects of apo-
morphine and oestrogen pretreatment after ovariectomy. Experiments
were carried out on ovariectomized animals (OVX) and in sham-opera-
ted female rats (SHO). Ovariectomy was performed 15 days before the
experiments on female Charles River rats. Apomorphine, at the same
dosage (20 pg/kg s.c.) was administered to oil treated OVX animals,
oestrogen treated OVX and oil and oestrogen treated SHO animals, 24
hours after the last hormone administration. The animals were coun-
ted for motility in respect to OVX rats treated with saline. Apomor-
phine markedly decreased spontaneous motor behaviour (75%) in OVX
but not SHO animals, as observed in intact female rats.
The analysis of variability, as observed by the low standard
error, revealed that OVX rats showed a rather homogenous behavioural
response to apomorphine administration, the responses being similar
to those observed in male rats. This difference between OVX and SHO
animals indicates that gonadal hormones are responsible for varia-
bility of the behavioural responses induced by a low dose of apo-
morphine in intact female rats.
Furthermore, oestrogen pretreatment in OVX animals significantly
antagonized the sedative effect, a response similar to that obser-

234
ved in male rats but not in intact female rats. Similarly, striatal
DOPAC content decreased by 33% in OVX rats after apomorphine
(100 yg/kg s.c.), but this effect was no longer present in OVX ani-
mals pretreated with oestrogens, indicating that female hormones
antagonize the decrease of DA turnover in striatum elicited by DA
agonists.
It is noteworthy that both OVX and male rats have the same beha-
vioural and biochemical responses after apomorphine with or without
oestrogen pretreatment. Female rats show a variable sensitivity of
DA autoreceptors (measured by behavioural and biochemical methods)
which is in close relation to oestrous cycle and in particular to
endogenous levels of 17 B-estradiol, indicating a physiological con-
trol of these hormones on dopaminergic activity. Accordingly, ova-
riectomy increases DA autoreceptor sensitivity (measured by beha-
vioural models and by biochemical methods in vivo). DA autoreceptors
control DA synthesis, release and firing activity of the DA neuron
itself and may therefore modulate global dopaminergic transmission.
These data may be relevant for mood variations during the menstrual
cycle, for puerperal or post-menopausal depression and for the an-
tidepressant activity of oestrogen.
REFERENCES
1. C.Euvrard, C.Oberlander and J.R.Boissier, "Antidopaminergic ef-
fect of estrogens at the striatal level", J.Pharmacol.Exp. 214:
179 (1980).
2. J.H.Gordon, R.A.Gorski, R.L.Borison, B.I.Diamond, "Postsynaptic
efficacy of dopamine: possible suppression by estrogen", Pharm~­
col.Biochem.Behav., .12: 515 (1980).
3. L.A.Chiodo and A.R.Caggiula, "Alterations in basal firing rate
and autoreceptor sensitivity of dopamine neurons in the substan-
tia nigra following acute and extended exposure to estrogen",
Eur.J.Pharmacol._, 67: 165 (1980).
4. P.Piccardi, F.Bernardi, Z.Rossetti, G.U.Corsini, "Effect of est-
rogens on dopamine autoreceptors in male rats", Eur.J.Pharmacol.
91 (In Press) (1983).
5. M.P.Piccardi, F.Bernardi, A.Bocchetta, G.U.Corsini, "Estrogen
modulation of dopaminomimetic actions: Implications for autore-
ceptor regulation", In: Lisuride and other dopamine agonists,
Eds. D.B.Calne et al., Raven Press, New York, 1983, pp.55-64.

235
NEUROLEPTIC-LIKE ACTIVITY OF ESTROGENS

Therese Di Paolo*, Paul Bedard** and Fernand Labrie*

*Department of Molecular Endocrinology, CHUL, Quebec G1V

4G2, Canada; and **Neurobiology Laboratory, Dept of
Anatomy, Laval University, Quebec G1J 1Z4, Canada

1. INTRODUCTION

Tardive dyskinesias are more frequent in postmenopausal women

than in men (Kane, 1982). Estrogens, when administered to parkin-
sonian patients can decrease Dopa-induced dyskinesia and in some
cases increase the signs of parkinsonism (Bedard et al., 1979).
The same antidyskinetic effect is observed in patients suffering
from tardive dyskinesia (Bedard et al., 1979; Villeneuve et al.,
1980). These clinical observations point to neuroleptic-like acti-
vity of estrogens at the level of the striatum. The following expe-
riments have been able to confirm these clinical results.

2. ESTRADIOL AND DOPAMINE TRANSMISSION

17~estradiol, at concentrations up to 10 mM, does not compete

for the striatal dopamine receptors labelled with [ ~]spiperone.
Thus, the neuroleptic-like activity of estradiol is not due to
binding to the receptor such as is the case for neuroleptics. By
contrast, a chronic treatment of two weeks with estradiol (10 ~g
b.i.d.) of adult ovariectomized (ovx) rats leads to an increase in
the density of striatal dopamine receptors with no change of their
affinity (Di Paolo et al., 1979; 1982).

3. COMBINED HALOPERIDOL AND ESTRADIOL TREATMENTS

Chronic neuroleptic treatment which is known to produce an

increase of dopamine receptors in the striatum was studied in com-
bination with estradiol treatment. The drug doses were 10 ~ b.i.d.

237
for estradiol for 4 weeks; haloperidol was given daily with a week-
ly increase from 1, 2.5 to 5 mg/kg for 3 weeks and stopped on the
fourth week. Another group received a combination of these two
treatments. After the administration of haloperidol, measurement of
catalepsy can be considered as reflecting the intensity of the
inhibition of dopamine receptors. After injections of 2.5 or 5.0
mg/kg of haloperidol, rats were cataleptic and estradiol potentia-
tes this effect (DiPaolo et al., 1981). The striatal dopamine
receptors density is increased in estradiol as well as in the halo-
peridol-treated animals and the combined treatment leads to even
higher increases (DiPaolo et al., 1981). Thus, with the drug sche-
dule used, the effect of estradiol and haloperidol on the number of
dopamine receptors is additive and this is in line with the beha-
vioral observation of an increased catalepsy with estradiol.

4. ROLE OF THE PITUITARY AND PROLACTIN

Estrogens stimulate the production of prolactin from anterior

pituitary mammotrophs and neuroleptics by blocking the inhibitory
dopaminergic control of prolactin secretion also elevate plasma
prolactin levels. Thus, the increase in striatal dopamine receptor
density produced by estrogens or haloperidol treatments may at
least partly share the common mechanism of increasing prolactin
levels. This was investigated in the following experiment.

Ovx and hypox rats treated with estradiol (10 ~. b.i.d.) or

bearing three anterior pituitary implants for two weeks were assay-
ed for their blood prolactin levels and striatal dopamine receptor
binding capacity. As expected the pituitary implants elevated blood
prolactin levels in hypox and ovx animals while estradiol elevated
prolactin levels only in ovx rats. Both estradiol or prolactin from
the implants elevated striatal spiperone binding in hypox and also
in ovx rats. However, hypox rats had about 20% less binding capaci-
ty than ovx rats (Di Paolo et al., 1982). Thus, estradiol can have
an effect on brain dopamine systems and the pituitary is not requi-
red for this. However, these results also indicate that the pitui-
tary and prolactin can modulate central dopamine systems.

5. EFFECT OF HALOPERIDOL AND ESTRADIOL IN HYPOPHYSECTOMIZED RATS

A similar protocol as previously shown in ovx animals was used

in hypox rats with a weekly increase of the haloperidol dose. After
the administration of 2.5 or 5 mg/kg of haloperidol, as previously
observed, the ovx rats were cataleptic and estradiol potentiates
the catalepsy duration in hypox or ovx rats (Fig. 1). However, the
hypox animals are much less cataleptic than the ovx rats.

Fig. 2 shows in hypox rats the effect of a chronic estradiol or

238
 Oovx + HYPOX
E:li:'l OVX + HYPOX + E,
Dovx
~ovx + E2
0
LJ.J
~
z
0

~
::l
0
>-
(/)
a.
LJ.J
...J

~
u

HALOPERIDOL 5 MG/KG
** p < O.D1

Fig. 1. Catalepsy measured in seconds on a 10 em horizontal bar

after the 3rd week of haloperidol treatment (5 mg/kg, s.c.) in ovx
and hypox rats. During the neuroleptic treatment, half of the rats
were also treated with estradiol (10 pg b.i .d.). Vertical bars
indicate S.E.H.

haloperidol treatment on blood prolactin levels and striatal dopa-

mine receptor binding. As expected in hypox rats, prolactin is low
and is not increased with the treatments. On the right panel we see
the effects of haloperidol or estradiol alone which both individual
ly increase the receptor levels. Thus, haloperidol can lead to dopa-
minergic supersensitivity (an increase of dopamine receptors) inde-
pendently of the pituitary. Estradiol treatment started during the
withdrawal period of haloperidol can significantly decrease the
haloperidol induced increase of dopamine receptors. This is an
interesting finding; estrogens given during the withdrawal period
from haloperidol can inhibit the development of supersensitivity.
In the last column of Fig. 2, we see that if estradiol is given
during all the haloperidol treatment it does not reverse the effect
of haloperidol. We have previously shown (Di Paolo et al., 1981)
that it potentiates the haloperidol increase in ovx rats.

Thus, as is observed with estradiol, haloperidol does not re-

quire the presence of the pituitary to produce an increase of stri-
atal dopamine receptors. However, the pituitary also plays a role
since the hypox rats striatal dopamine receptors are lower and they

239
 OVX + HYPOX RATS
DcoNTROL
C) ESTRADIOL (DAYS 1-28)
lllJ HALOPERIDOL (DAYS 1-21)
IIIII HALOPERIDOL (DAYS 1-21)
ESTRADIOL (DAYS 22-28)
[j HALOPERIDOL (DAYS 1-211
+ ESTRADIOL (DAYS 1-28)

*P < 0.05 **P < 0.01

Fig. 2. Effect of estradiol and/or haloperidol treatments on plasma

prolactin levels and striatal [3H]spiperone binding in hypox and
ovx rats. Means ± S.E.M. are shown and [ 3H]spiperone (0.1-0.5 nM)
binding data are given in percent of OVX + HYPOX controls.

are much less cataleptic after an haloperidol injection than the

ovx animals. Therefore, estradiol, prolactin and haloperidol may
have similar but independent activities. Thus, several factors can
interfere with dopaminergic transmission and this could partly
explain the contradictory observations showing that estrogens can
either decrease or increase the efficacy of dopaminergic drugs
(Gordon, 1980; Joyce et al., 1982).

6. DYSKINESIAS IN MONKEYS

Recent studies show a biphasic effet of one injection of estro-

gens, inhibitory at short term and stimulatory in the long term on
stereotypy behaviors in rats (Gordon, 1980; Joyce, 1982). We have
studied this effect of estradiol in a monkey model of dopamine
sensitive lingual dyskinesias. These monkeys (maccaca facicularis)
have a left midbrain electrolytic lesion and show bucco-lingual
movements similar to neuroleptic-induced dyskinesias in humans.
These movements occur spontaneously but are increased with dopamin-
ergic agonists and decreased with antagonists (Bedard et al.,1982).
A chronic estradiol treatment significantly decreases the apomor-
phine-induced dyskinesias in these monkeys (Bedard et coll., 1983).

The effect of only one injection of estradiol to these monkeys

is shown in Fig. 3. One day after the estradiol injection, the ef-

240
 HALOPERIDOL

% C.S.F. ESTRADIOL

f'
(/)
0 500 * 200 (/) % CS.F
z .. 0
z

j\
0 ! ... 200
u 400 150- 0 60
w
(/)
300
E
.......
100 OlQ
u
w
(/) 150 2/t/¥· · · ·. .2 40 :§
~I 200
(/) 50
c :
~:

.::t6
> ~I 100 Eb
0>0
.Si
20 <t 6
w
z QHA I (j)
w l >
I
:;;::
(/)
100 l 0 ~
:::.:::
50 0
>- (/)
0 >-
0
0 7 14 28 0 3 7 14 30
DAYS DAYS
'P<0.05 ''P<0.01 'P < 0.05
Fig. 3. Effect of haloperidol (1 mg/kg, i.m.) or estradiol benzoate
(0.5 mg, s.c.) on the apomorphine potentiation of lingual dyskine-
sia in monkeys. "Dyskinesia time" ± S.E.M. recorded in 4 monkeys
during 75 min and HVA levels obtained in the CSF immediately before
the apomorphine test are shown. The response to the first test was
arbitrarily set at 100%.

feet of apomorphine on the dyskinesia is reduced by 75%. There is

then a slow increase with an overshoot at 14 days and a return to
normal at 30 days. This biphasic effect of a single injection of
estradiol is similar but smaller to the effect of an injection of
haloperidol. CSF, HVA values were elevated at one day, after
either estradiol or haloperidol at the time of the reduced efficacy
of apomorphine and HVA decreased thereafter. This HVA increase may
be interpreted as an increased dopamine turnover. Prolactin itself
also biphasically influences dyskinesias with an inhibition follow-
ed by a facilitation of the effect of apomorphine (Bedard et al.,
1983).

7. CONCLUSION

1. The increase after estradiol of striatal dopamine receptors

which is also seen after haloperidol is compatible with the view
that estradiol may act as a weak neuroleptic. The biphasic dyskine-
sia response to apomorphine seen in monkeys after a single dose of
haloperidol or estradiol also shows the neuroleptic-like properties

241
of estrogens. Estradiol shares several properties with haloperidol.
Is estradiol an endogenous neuroleptic?
2. Both estradiol and prolactin may be independently capable
of modulating dopaminergic transmission. Indeed, chronically high
levels of estradiol or prolactin as well as only one increase of
prolactin levels or only one injection of estradiol can influence
dopaminergic transmission in the brains of rats and monkeys.

REFERENCES

Bedard, P.J., Boucher, R., and Larochelle, L., 1982, Experimental

tardive dyskinesia, Prog. Neuro-Psychopharmacol. & Biol.
Psvchiatr., 6: 551.

Bedard, P., Boucher, R., Di Paolo, T., and Labrie, F., 1983, Bipha-
sic effect of estradiol and domperidone on lingual dyskinesia in
monkeys, ~xp. Neurol._, 82: in press.

Bedard, P.J., Langelier, P., Dankova, J., Villeneuve, A., Di Paolo,

T., Barden, N., Labrie, F., Boissier, J.R., and Euvrard, C., 1979,
Estrogens, progesterone and the extrapyramidal system, Adv.
Neurol., 24:411.

Di Paolo, T., Carmichael, R., Labrie, F., and Raynaud, J.P., 1979,
Effects of estrogens on the characteristics of [3a)spiroperidol
and [ 3H)RU24213 binding in rat anterior pituitary gland and brain,
Mol. Cell. Endocrinol., 16:99.

DiPaolo, T., Poyet, P., and Labrie, F., 1981, Effect of chronic
estradiol and haloperidol treatment on striatal dopamine receptors,
Eur. J. Pharmaco~., 73:105.

Di Paolo, T., Poyet, P., and Labrie, F., 1982, Effect of prolactin
and estradiol on rat striatal dopamine receptors, Life Sci., 31,
2921.

Gordon, J., 1980, Modulation of apomorphine-induced stereotypy by

estrogen: time course and dose-response, Brain Res. Bull_., 5: 679.
1

Joyce, J.N., Smith, R.L., and Van Hartesveld, 1982, Estradiol sup-
presses, then enhances intracaudate dopamine-induced contralateral
deviation, Eur. J. Pharmacol., 81:117.

Kane, J.M. and Smith, J.M., 1982, Tardive dyskinesia, Arch. Gen.
Psychiatry, 39:473.

Villeneuve, A., Cazejust, T., and Cote, M., 1980, Estrogens in

tardive dyskinesia in male psychiatric patients, Neuropsvchobio-
_!_ogy. 6:145.

242
INTERACTIONS BETWEEN GONADAL STEROIDS AND NEUROTRANSMITTERS

Gertraud Koster and Heinz Breuer

Institut fUr Klinische Biochemie, and

Neurochirurgische Klinik
University Bonn, FRG.

The brain is a target for estrogens, as we know e.g. from the

fact that the site of steroid feedback on the secretion of peripheral
hormones is located in the hypothalamus (HT). For a long time, neuro-
endocrinologists have been investigating the mechanisms of central
hormone actions. Although their work is not concerned with behavioral
aspects, some of their findings m~ also be of value for psychiatry,
as similar mechanisms may be involved in other hormone actions in
the brain. Furthermore, under neuroendocrinological conditions they
may be studied more easily as the input into, and the output out of
the brain, are measurable parameters, i.e. the steroid levels in
the blood or the hormone secretion of the pituitary.

It is in this context tbat we have studied the metabolic inter-

action between noradrenaline (NA) and 2-hydroxyestradiol-178 (2-0H-E 2 ),
a catecholestrogen which in brain as well as in peripheral organs
1s formed from estradiol-178 (E 2 ) as a natural metabolite.

2-0H-E 2 acts as a hormone mainly at the central feedback system

and suppresses the elevated LH-levels in ovarectomized (ovex) rats.
It is easily methylated by the catechol-0-methyltransferase (COMT)
and is one of the most effective competitive inhibitors of COMT
in vitro.

COMT is almost equally important as MAO for the enzymatic de-

gradation of catecholamines, the neurotransmitters involved in the
central feedback processes. Thus the question arose, whether 2-0H-E 2
also might inhibit the COMT in the intact brain tissue with pre-
served metabolic compartments, and whether this might be one of the
mechanisms by which the hormone acts in vivo.

243
 In earlier experiments we studied the effect of E2 applied
in vivo on the metabolism of 3 H-NA when incubated with brain tissue
slices of ovex rats 2o h after the hormone treatment. We expected
the 2-0H-E 2 to be formed from the inj ected E2 at the physiological
sites and to be retained especially at the E2 -receptors in the HT.
A sufficient influence on the NA-metabolism then should result in
a shift away from COMT-deriving methylated metabolites to deaminated
catechol compounds formed by MAO (Koster and Breuer, 1981).

Figure 1 shows the amount of metabolites formed from 3 H-NA

during 3o min incubation periods with HT-slices (for detail, see
legend to Fig. 1). Compared to the controls, lower amounts of all
the methylated metabolites were found in the E -treated group. But
the reduced methylation was compensated for by an increased form-
ation of only one of the deaminated compounds, namely DOMA (for
abbreviation, see legend to Fig.1), which is of only minor importance
in vivo.

The results could have been caused by 2-0H-E 2 f ormed in vivo

from the injected E2 but not necessarily so. To test this
assumption, the direct effect of 2-0H-E 2 on the metabolism of NA was
studied in vivo. (Koster & Breuer, 1982 ). Again, the methylation was
reduced, and the deamination was increased, in ovex rats which had

Thalamus
Hypothalamus

4
~
)(

"'
0
E
c: ~

"'..
*
)(

Ol

E 0
E
2 c:
'--'
\II
!!l
]
E
~
..
Figure 1. Effect of estradiol on the metabolism of noradrenaline
in the brain. Brain tissue sli ces, taken from ovar-
ectomi zed rats 2o h after injection of solvent 11 or
estradiol~. were incubated with 3 H-noradrenaline for
3o min. Formation rates are given as nmol/g·h for the
catechol compounds DHPG = 3 ,4-dihydroxyphen ylglycol,
DOMA = 3 ,4-dihydroxymand elic acid, the methylation pro-
duct NM = normetanephrine, the ceaminat ed and methylated
metabolite s MHPG = 3-methoxy-4-hydro xyphenyl glycol,
VMA = vanillic mandelic acid .

244
been treated with 2-0H-E 2 • But the non-hydroxylated E2 did not alter
the metabolic pattern in vivo. That means, a specific effect of
2-0H-E 2 had been observed which supported the assumption that the
catecholestrogen might inhibit the COMT in the intact brain as well
as in vitro.

Then the question arose whether an inhibition of the COMT could

be of physiological relevance; that means, whether it could- some-
how - affect the aminergic neurotransmission, e.g. by affecting the
NA-turnover. It is often assumed that MAO may have such a modulatory
effect on the turnover of catecholamines, but similar considerations
concerning the COMT are rare in literature, as this enzyme is thought
to be localized in non-neuronal cells and to have almost no effect
on the neurotransmitter in the synaptic cleft.

These conclusions were derived from studies in the periphery.

Therefore it was tested whether also in the brain MAO acts only in
the presynaptic nerve ending, and COMT occurs only at extraneuronal
sites.

In order to distinguish between the various compartments in the

brain, the specific uptake systems were inhibited by drugs. Reserpine
is known to inhibit the vesicular storage and thus to stimulate the
intraneuronal NA-metabolism, whereas desmethylimipramine (DMI) blocks
the neuronal Uptake and thus stimulates the extraneuronal NA-metabol-
ism, but prevents the intraneuronal processes. After pretreatment
with these drugs the role of COMT in the various compartments could
be studied if the enzyme was inhibited by tropolone.

The metabolic patterns of intraventricularly injected 3 H-NA in

drug-treated rats revealed, that also in the brain, COMT acts on NA
exclusively outside the nerve endings, and that within the nerve
endings all the 3 H-NA will be deaminated by MAO. But in spite of
the fact that the methylation takes place only at extraneuronal
sites, the COMT-inhibitor tropolone had a protective effect on 3 H-NA
and increased its recovery.

When 3 H-NA was injected into drug-pretreated rats and its re-
covery determined either 1o min (exp. A) or 1 h (exp. B), the 3 H-NA-
values of tropolone-treated rats always exceeded those of the compar-
able tropolone-free group, irrespective of the pretreatment with
reserpine or DMI (see Tab. 1) The most pronounced effect was seen
in the DMI-group when more 3 H-NA than normal was methylated at extra-
neuronal sites

However, the interesting finding was that tropolone reduced

only the initial decline of 3 H-NA, but did not retard the metabolic
breakdown after the first 1o min. Instead, a greater fraction of
the injected 3 H-NA had been taken up into compartments with a slow

245
 Table 1. Drug effects un the recovery of 3 H-noradrenaline in
rat brain regions

Tissue Recovery of 3 H-noradrenaline (%) in treatment group

(exp,) C T R RT D DT

HT (A) 56 + 5 67 + 12 28 + 7 51 + 19 31 + 5 58 + 6
(B) 44 + 9 59+ 11 12 + 5 15 + 7 31 + 6 56 + 8
(C) 41 + 7 47 + 9 16 + 5 27 + 11 57 + 9 63 + 1o
PM (A) 39 + 11 51 + 1, 24 + 5 36 + 17 19 + 62 + 18
(B) 29 + 5 41 + 16 5 + 1 8 + 3 11 + 2 3o + 7
(C) 29 + 8 49 + 9 9 + 4 12 + 6 5o+ 11 57 + 13

SE (A) 53+ 11 58 + 18 25 + 4 53 + 18 37 + 6 66 + 16
(B) 19 + 4 41 + 1o 8 + 2 17 + 4 36 + 4 46 + 8
(c) 4o + 7 5o + 12 24 + 5 32 + 12 67 + 9 63 + 11

Rats were treated with several combinations of carrier (C), trope-

lone (T), reserpine (R), DMI (D). Exp. A, B: The drugs were injected
before the intraventricular injection of 3 H-NA; its recovery was
determined 1o min (A) or 1 h (B) later. Exp. C: The drugs were
given 2 h after 3 H-NA-injection; the recovery was determined 3 h
later. The recovery is given in %of the respective initial tissue
content of 3 H-NA. HT =hypothalamus, PM= pons +medulla, SE =
septum.

NA-decline. Thus, under the action of tropolone, some of the exogen-

ous NA must have been rejected from the COMT-compartment during the
intial processes of uptake and metabolism, and thus became available
to other spaces.

To explain this it was postulated that the COMT might be linked

to the extraneuronal Uptake and forms a saturable "extraneuronal
methylating system" similar to that described for peripheral organs
by Trendelenburg (198o). This linked system would be inhibited by
tropolone as a whole. It might be speculated that a reduced NA-up-
take into the COMT-space could retard the elimination of the amine,
not only from the intercellular space, but also from the synaptic
cleft. The increase of the NA-concentration in the synaptic cleft
would prolonge the NA-action on the postsynaptic B-receptors as well
as on the presynaptic a-receptors which mediate the feedback on the
NA-release during the nerve impulses - and in this way, an inhibition
of the extraneuronal COMT might retard the release of NA from the
distant vesicular pool.

246
Table 2. Ef'fect of estrogens on the initial vesicular uptake (U )
and on the half-time (t '2
1/ ) in the hypothalamus of ovex

rats

control E2 2-0R-E2
aRT pRT aRT pRT aRT pRT

tk (h) 4.2 3.9 3.5+ 3.4 7. 1+ 5.3+

uz
0
(. 10 cpm/g) 3. 1 1.9 2.2 1.8 1.8+ 1.4+

1o pg E 2 or 2-0R-E 2 or solvent were injected 1 h before the intra-

ventricular injection of R-NA into ovex rats. The half-life time
(t k ) and the initial uptake into the vesicles (U were calculated
0 )

frok the 3 R-NA-levels in hypothalamic tissue as determined bP.tween

. 5 and 6 h after amine injection. + P :( . o5 vs. controls.

This hypothesis gets some support from experiments in which the

drugs (reserpine, DMI, tropolone) were given, not before, but 2 h
after the injection of 3 R-NA, i.e. when all the 3 R-NA had been stored
in the vesicles. Then the elimination rate of 3 R-NA is no longer de-
termined by the uptake into the metabolic compartments, but by the
release from the nerve endings (exp. C in Tab. 1). Again tropolone
improved the recovery of NA, but in contrast to exp. A and B not in
DMI-treated rats. Since DMI, by inhibiting the neuronal re-uptake,
activated the a-receptors to the maximum extent, it alone caused an
increase of 3 R-NA compared to the controls, but tropolone could not
further improve the recovery of 3 R-NA. In contrast, tropolone now
increased the recovery of 3 R-NA in reserpine-treated rats though
here HAO is the more important enzyme but reserpine blocks the re-
storage of previously released 3R-NAously - and this release is
under the control of the a-receptors. his supported the hypothesis
that COMT might aff'ect the turnover of the NA-pool.

The question remained whether this mechanism might also be in-

volved in the hormone action of 2-0R-E 2 , which presumably inhibits
the methylation in the brain (see above). Thus, the NA-turnover was
studied in the anterior (aRT) and in the posterior part (pRT) of the
RT of hormone-treated ovex rats and correlated to the LR-levels in
the blood. (The pRT contains the median eminence where the influences
on the tonic LR-secretion are integrated, whereas in the aRT the cyc-
lic LR-surge is triggered.)

Between 1 and 7 h after application of either E2 or 2-0H-E 2

(1o pg/kg), both hormones decreased the LR-levels in the blood (see
Fig. 2), but they had different effects on the fate of 3 R-NA in the
RT (see Tab. 2). E2 slightly, but insignificantly accelerated the

247
 -
E
01
c:

~
~
400

-!-l-1--j ** * . *
. yf~
f -f{~f-t-1
0

tV
E
ell 200
0
iS.
I
:I: blood plasma
...J

0
3 5 7 3 5 7 3 5 7
Time after hormone injection (h]
Figure 2. Effect of estrogens on the LH-levels in blood

turnover of 3H-NA, whereas 2-0H-E 2 retarded it in aHT as well as in

pHT and caused significant longer half-life times t~. Both hormones
reduced the NA-storage in the neuronal pool (see U 0 -~alues in Tab.2).
Again, 2-0H-E 2 was more effective than E2 in both aHT and pHT.

The effects of the 2 hormones on the LH-secretion and on the

fate of NA are not similarly correlated. E2 suppressed the LH-levels
throughout the experimental period, but it did not affect the NA-
turnover. In contrast, 2-0H-E 2 had a significant effect on LH for
only a short time, but it markedly affected the turnover and the up-
take of 3 H-NA. It even seems that 2-0H-E 2 for only 2 h retarded the
NA-turnover much more than reflected by t , but there are not yet
enough data available to test this statistically. However, such a
short lasting effect would correspond very well to the short last-
ing decrease of LH, and would point towards a connection between the
2 effects of 2-0H-E 2 on eith.y LH or NA.

In summary, it is concluded that E2 and 2-0H-E 2 act by different

mechanisms. 2-0H-E 2 seems to have more short lasting direct effects;
one of them might be a metabolic influence on the NA-metabolism
brought about by an inhibition of COMT, and this mechanism possibly
is related to the modulatory effect on the LH-secretion. But there
are indications that other mechanisms occur as well.

REFERENCES

Koster, G., and Breuer, H., 1981, Acta Endocrinologica, 98:1

Koster, G., and Breuer, H., 1982, Acta Endocrinologica, 99:1
Trendelenburg, U., 198o, Rev. Physiol. Biochem. Pharmacal., 87:34.

248
NEUROENDOCRINE INVOLVEMENT IN THERAPEUTIC MECHANISMS OF NEUROLEPTIC

AND ANTIDEPRESSANT DRUGS: STUDIES OF THYROID AXIS

G. Langer, H. Aschauer, M.S. Keshaven, G. Koinig,

F. Resch and G. Schoenbeck

Department of Psychiatry
A-1097 Vienna, Austria

SUMMARY

Sixty-five depressed and 33 paranoid hallucinatory patients

were investigated longitudinally for one year to assess short- and
long term therapeutic outcome with antidepressant and neuroleptic
drugs, respectively. The patients' thyrotropin (TSH) response to
thyrotropin releasing hormone (TRH) was studied at admission,
during inpatient treatment and at discharge. Decreased TSH respon-
ses at outcome, and normalisation of these pathological responses
during treatment were associated with the highest chance for
recovery. TSH responses which persisted blunted at discharge were
associated with a higher relapse rate during the one year
following. It is hypothesized that the blunted TSH response may
reflect a nonspecific cerebral malactivation, which is disactivated
by the therapeutic effects of neuroleptic and antidepressant drugs.

INTRODUCTION

It is widely recognised that modern psychotherapeutic drugs

have revolutionised concepts and treatments in psychiatry.
However, little is known about the biochemical mechanisms
underlying the clinical response to antidepressant and neuroleptic
drugs.

The study of the hypothalamic-pituitary-thyroid axis has, over

the last decade, emerged as a major neuroendocrine strategy for the
investigation of the biological basis of various psychiatric
disorders and of the mechanisms of their therapy. In particular,
much attention has been directed towards the thyrotropin (TSH)
response to thyrotropin releasing hormone (TRH), first observed in

249
1972 to be diminished ("blunted") in a considerable portion of
depressed patients before treatment (1) and widely replicated
thereafter (2). Subsequent studies have revealed that this finding
may be seen in quite different illnesses such as schizophrenia,
alcoholism, chronic deseases and anorexia nervosa (2), suggesting
that the possible clinical value of this neuroendocrine parameter
may not be utilised to support the traditional diagnostic habits.
Only a few studies have raised questions of concern for therapy,
e.g. to examine the TRH-test for treatment outcome. It has been
observed that the "disblunting" of a blunted TSH-response during
therapy may predict successful outcome with antidepressants (3,4),
neuroleptics (5) and ECT (3). The biological "meaning" and the
mechanisms underlying this transient abnormality in the thyroid
axis remain unclear. The blunted TSH-response has been thought to
reflect a dysregulation on the hypothalamic level, posssibly based
on a serotonergic dysfunction (6). While this may be in line with
other findings in depressive illness, the fact of its occurance in
a variety of neuropsychiatric illness and its similiar alteration
by psychotherapeutic drugs of different chemical classes points to
a more general interpretation of the findings. In this paper, we
report the results of our longitudinal studies in depressed and
paranoid hallucinatory patients aimed at investigating the
following questions: 1. Can the blunted TSH-response to TRH at
admission predict outcome with antidepressant and/or neuroleptic
drugs? 2. Is the blunted TSH-response to TRH also a good predictor
of prophylaxis (long term outcome)? 3. Can various TSH-responses
during therapy, with different associations with outcome, be
identified? 4. How can the observation of the possible involvement
of the thyroid axis in both, the pathogenesis and therapy of
psychiatric syndromes, be formulated in a unified hypothesis?

MATERIAL AND METHODS

Patient population

The patient population consisted of 98 female patients who

were evaluated on the psychobiological research ward of the
department of psychiatry, University of Vienna. Sixty-five of them
presented with a depressive and 33 with a paranoid-hallucinatory
syndrome; by diagnostic criteria (ROC, 7) the former group was
classified as major depressive disorder, whereas the latter was a
mixed group of schizophrenic, schizoaffective and manic disorders.
Patients with manifest cerebral-organic and medical illnesses were
excluded from the study.

Study design

After admission, previous medication was withdrawn and oral

diazepam (20-40mg/d) was given until the first TRH-test was done
within one week. Thereafter depressed patients were treated with

250
clomipramine (AnafranilR; 50-250mg/d) and paranoid-hallucinatory
patients with haloperidol (HaldolR; 10-30mg/d), given over the
first week as intravenous infusions. The severity of the depressive
or paranoid-hallucinatory syndromes was assessed by applying the
Hamilton-rating scale for depression (HAM-D, first 16 items (8))
and the Brief Psychiatric Rating Scale (BPRS, (9)) respectively.
Recovery was either defined as a reduction of the total HAM-D score
to less than 9 or the BPRS score to less than 30 points at 3-9
weeks of drug treatment. After discharge, patients were followed up
and tested at regular intervals over one year. Relapse was defined
as reemergence or worsening of symptoms necessitating readmission.

The TRH-test (TRH-RelefactR; 0.4mg i.v. over 30 sec; at 8.30

AM) was carried out weekly during hospitalisation and in monthly
intervals during the follow up. TSH was analysed by radioimmuno-
assay. Intra- and interassay coefficients were below 15%. A hormo-
nal response to TRH was defined as blunted, if the maximal increase
of the serum concentration of TSH following TRH was less than
5uU/ml. For statistical analysis, the "risks" of the patients with
different TSH-response patterns for recovery or non recovery was
assessed by comparing the odds of all groups by cross-product
ratios. Equal risks is the null-hypothesis to be rejected.

RESULTS

Thirty-three of the 98 patients studied at admission showed a

blunted TSH-response to TRH with no apparent difference between
depressed (33%) and paranoid-hallucinatory patients (30%) in the
prevalence of this abnormality.

TSH-responses at admission in relation to therapeutic outcome at

3-9 weeks

TABLE 1. TSH response to TRH at admission and clinical outcome at

3-9 weeks of drug treatment in two groups of psychiatric
patients.

PSYCHOPATHOLOGICAL TSH-RESPONSE PSYCHOPATHOLOGICAL SIGN.

SYNDROME AT ADMISSION STATE AFTER 3-9 WEEKS
recovered non-recovered

DEPRESSIVE NORMAL 27 15 T=5.8

BLUNTED 21 2 p<0.05

PARANOID- NORMAL 17 6 T=7 .8

HALLUCIN. BLUNTED 10 0 p<0.05

TOTAL NORMAL 44 21 T=7 .4

BLUNTED 31 2 p<0.05

251
 Seventy-five of the 98 patients were considered recovered at
3-9 weeks of treatment; 44 depressed patients (74%) and 27
paranoid-hallucinatory patients (82%). The distribution of recovery
and non-recovery in relation to the psychopathological syndrome and
the TSH-response at admission is shown by table 1. The patients
with a blunted TSH-response at admission (regardless of syndromal
category) had a significantly higher chance to recover than
patients with a normal TSH-response.

TSH -response patterns Psychopathological state

during drug treatment of at 3-9 weeks of
3-9 weeks (schematic) drug treatment

"Recovered" "Non-recovered"

10 2
c 51JU/ml

8 3
c51JU/ml ;~;;
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::: 'T-6.38
pc0.05

34 2

'!'·11.82
pc0.05

;,51JU/ml

23 16
.:rrrrrttttttttttrrrrt
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .'9"1> . . . . . . . . . . . .

FIGURE 1: TSH-response patterns during drug 3-9 weeks

in relation to clinical outcome

252
TSH-resoonse patterns during treatment in relation to outcome

Serial observations of the TSH-response to TRH during the

inpatient phase of treatment revealed for response patterns: a) 12
patients showed persistently a blunted TSH-response ("blunted")
b).initially normal responses became blunted in 11 patients ("blun-
ting") c) 36 patients showed a normalisation of a blunted response
("disblunting") and d) 39 patients persistently had a "normal"
response. Those patients who showed a "disblunting" TSH-response
pattern had an almost twelve times higher chance of recovery than
those with persistently normal ISH-responses (log T=2.47;
z=3.39; p<O.OS) and an almost seven times higher chance of recovery
than patients showing the "blunting" response pattern (log T=1.85;
z=2.03; p<O.OS).

TSH-response at discharge and long-term outcome

Recovered patients who presented a normal TSH-response at

discharge were less likely to relapse during the one year follow up
than the patients with a blunted TSH-response (tab.2).

TABLE 2. TSH-response to TRH at discharge and long-term outcome.

TSH-RESPONSE AT DISCHARGE SIGN.

Normal Blunted

IMPROVEMENT MAINTAINED 31 8 T=3 .17

z=1.97
RELAPSE 11 9 p (0 .os

DISCUSSION

The presented findings allow for three major comments.

Firstly, depressed and paranoid-hallucinatory patients did not
differ significantly with respect to the frequency of the blunted
TSH-response at admission which suggests the absence of a
diagnostic specification along traditional categories. Secondly
neuroleptic and antidepressant drugs effected similarly the TSH-
response patterns and were similarly associated with successful
outcome. Thirdly, a blunted TSH-response prior to treatment and a
normalisation of the response during treatment predicted best the
recovery, irrespective of syndroma1 category or the type of drug
being used. These observations indicate that the common therapeutic
mechanism may be operative. We have suggested earlier that the
changes in the TSH-responses may reflect as yet unknown "autothera-
peutic mechanisms" of the brain, which, triggered by drugs, other
therapies or spontaneously, are ultimately responsible for recovery
into a normal state (10).

253
 In order to understand these observations in a coharent
unified model, we propose the following hypothesis. The blunted
TSH-response may be regarded as indicator of pathological state of
"malactivation", which appears to be associated with diverse
psychiatric and medical illnesses, and which seems to represent a
dysfunction amenable to antidepressant and neuroleptic drugs.
These drugs may reverse the psychopathological states by
"disac tivating" the activation. A poor drug response in patients
with a normal pretreatment TSH-response (i.e. no malactivation on
which the drugs could act) and a higher rate of relapse among
patients with persistently blunted responses (i.e. failure of
disactivation) may be interpreted by this model.

If the above hypotheses were true, drugs which produce a

blunted TSH-response ("malactivation") must be able to transiently
convert treatment refractory states into responsive states. Indeed,
this could apply to the administration of thyroid hormones, which
are well known to effect a blunted TSH-response (11). There is
accumulating evidence that thyroid hormones potentiate the clinical
response to antidepressants in euthyroid depressives resistent to
treatment (12). Dopaminergic agonists such as L-Dopa which result
in a blunted TSH-response (13) also have shown promise in therapy
of some schizophrenic patients (14).

It is to be emphasised that the blunted TSH-response may

represent only one neuroendocrine manifestation of a more
widespread cerebral malactivation. Recent suggestions that an
undifferentiated state of psychophysiological and emotional
overarousal as evidenced by abnormal skin conductance and evoked
potentials may lie at the basis of all major psychosis (15) are in
support of this view.

The "disactivation" hypothesis of the therapeutic effects of

psychotropic drugs may bear considerable heuristic value. We are
currently carrying out further investigations designed to the model
by applying it to further questions related to treatment.

This work was supported by "Fonds zur Foerderung der

wissenschaftlichen Forschung", No. 4416 and 4565, and by an IBRO
fellowship (M.S.K.).

REFERENCES

1. A.J. Prange, I.e. Wilson, P.P. Lara and 1. Alltop, Effects of

thyrotropin releasing hormone in depression, Lancet 2: 999-
1002 (1972).
2. P.T. Loosen and A.J. Prange, Serum thyrotropin response to
thyrotropin releasing hormone in psychiatric patients, A re-

254
 view, Am. J. Psychiatry 139, 4: 405-416 (1982).
3. C. Kirkegaard, N. Norlem, U.B. Lauridsen and N. Bjorum, Progno-
stic value of the thyrotropin stimulation test in endogenous
depression, Acta Psvchiatrica Scandinavia 52: 170-177 (1975).
4. G. Langer, G. Koinig, G. Schoenbeck, H. Aschauer and F. Resch,
The therapeuti~ effects of antidepressants appear to involve
neuroendocrine mechanisms, in "Biological psychiatry,"
c. Perris, G. Struwe and B.-yanssen, eds., Elsevier/North
Holland (1981).
S. G. Langer, H. Aschauer, G. Koinig, H. Reiter, G. Schoenbeck and
o. Lesch, Neuroendocrine factors in the treatment of schizo-
phrenic patients: relationship with haloperidol level and
treatment outcome, in: "Biological Psychiatry," c. Perris,
G. Struwe and B. Janssen, eds., Elsevier/North Holland (1981).
6. L. Kirkstein, M.S. Gold, D. Martin and A.L.C. Pottash, Clinical
correlates of the TRH infusion test in primary depression, J.
Clin. Psvchiatrv 43: 5, 191-194 (1982).
7. R.L. Spitzer, J. Endicott and E. Robins, "Research diagnostic
criteria for a selected group of psychiatric disorders,"
third edition, N.Y.S. psychiatric institute, New York 10032,
(1978).
8. M. Hamilton, A rating scale for depression, .T. neurol.
osvchiat 23: 56-62 (1960).
9. J.E. Overall and D.R. Gorhem, The brief psychiatric rating
scale, Psychol. Reo 10: 799-812 (1962).
10. J. Wilber, A. Jaffer, L. Jacobs and R.D. Utinger, Inhibition
of TRH stimulated TSH secretion in man by a single dose of
thyroid hormone, Horm. Metab. Res. 4: 508 (1972).
11. G. Langer, G. Schoenbeck, G. Koinig and H. Aschauer, Neuro-
endocrine mechanisms in the therapeutic effects of anti-
depressant drugs: The thyroid axis hypothesis, in: "Typical
and atypical antidepressants," E. Costa and G. Racagni, eds.,
Raven Press, New York (1982).
12. F.K. Goodwin, A.J. Prange, R.M. Post, G. Muscettola and
M.A. Lipton, Potentiation of antidepressant effects by
L-trijodthyronine in tricyclic nonresponders, Am. J.
Psvchiat 139: 1, 34-36 (1982).
13. J. Tuomisto, Neuropharmacological intervention on the
pituitary-hypothalamic relationship, ftnn. elin. res. 10:
120-132 (1978).
14. S.L. Horacz, The dopamine hypothesis, Schizophrenia Bulletin
8, 3: 438-470 (1982).
15. H. Heimann, Auf dem Wege zu einer einheitlichen psychophysio-
logischen Theorie depressiver Syndrome, Praxis Psychother.
Psychosom. 24: 281-299 (1979).

255
PHARMACOENDOCRINOLOGY AND ITS CLINICAL RELEVANCE

Gregor Laakmann

Psychiatrische Klinik der Universitat Mlinchen

Nussbaumstrasse 2
8000 Munich 2, F.R. Germany

Studies of effects of a variety of psychotropic

drugs on pituitary hormone secretion are based on the
theory that such drugs affect the functions of aminergic
neurons, which in turn can influence the effects of the
pituitary hormones. The three following sections include
primarily our own results on first, th~ effect of psycho-
tropic drugs on pituitary hormone secretion, second, the
effect of receptor blockers on the changes in pituitary
hormone secretion due to psychotropic drugs and third,
the GH stimulating properties of DMI in depressive
patients.

I. PHARMACOENDOCRINOLOGY - PSYCHOTROPIC DRUGS AND

PITUITARY HORMONE SECRETION

Growth hormone (GH): Secretion of GH in man can be

stimulated by dopaminergic (DA), noradrenergic (NA) and
serotonergic (5-HT) neurons or agonistic substances (Mar-
tinet al., 1977).
Desimipramine (DMI), an antidepressant inhibiting
primarily the reuptake of noradrenaline (NA) led to re-
liable GH increase. The GH stimulation following p.o.
administration occurs later than after i.m. administra-
tion. The DMI-induced GH stimulation is reproducible and
dose-dependent (Laakmann 1980).

This work was supported by: Deutsche Forschungsge-

meinschaft, Schwerpunktprogramm Neuroendokrinologie.

257
 Chlorimipramine (CI), an antidepressant primarily
inhibiting the reuptake of 5-HT also stimulates GH. This
effect is significantly lower compared to DMI in the same
dose (Laakmann et al., in press).
Nomifensine (NF), an NA-reuptake inhibiting and DA-
agonistic antidepressant, also stimulates GH secretion
(Laakmann et al., 1979).
Benzodiazepine derivatives such as diazepam and
metaclazepam stimulate GH secretion only in a part of the
healthy subjects (Laakmann et al., 1982a).
The DA-receptor blocking neuroleptics do not stimu-
late GH secretion (Martinet al., 1977).
Prolactin (PRL): Secretion of PRL is primarily
controlled by dopaminergic inhibition. The stimulating
of serotonergic neurons must also be assumed (Martin et
al., 1977). The literature reports partly contradictory
effects of antidepressants on PRL secretion.
DMI and CI given p.o. and i.m. have either a minor
effect on PRL secretion or none at all, whereas after
i.v. application of the substances a significant stimula-
tion was observed. This stimulation is dose-dependent
(Laakmann 1980). Comparison of equal doses of DMI and CI
demonstrate that CI, the primarily 5-HT reuptake inhibi-
tor, stimulates PRL secretion significantly higher than
DMI (Laakmann et al., in press). NF inhibits PRL secre-
tion (Laakmann 1980). These results indicate the clear
effect of antidepressants in high i.v. dosage. Thus,
the contradictory results in the literature are possibly
traceable to low doses and different administration.
Diazepam, a benzodiazepine derivative, had no effect
on PRL secretion in humans (Laakmann et al., 1982b).
Neuroleptics enhance PRL secretion, as has been
shown by many investigators (Langer 1979). This increase
occurs after acute and chronic administration and is
dose-dependent. It seems to correlate with the DA-re-
ceptor blocking potency of a substance.
Luteinizing Hormone (LH) and Thyrotropic Hormone
(TS~): It can be shown that LH and TSH, just like other
pituitary hormones, are controlled by hypothalamic re-
leasing hormones. Furthermore, a possible influence of
aminergic neurons on LH, respectively TSH secretion pos-
sibly with the aid of releasing hormones, has often been
discussed in the literature. No exact explanation of
these mechanisms has been successful in recent years. A
clear effect of antidepressant, anxiolytic or neuroleptic
drugs on LH or TSH release has not been stated to date.
In our study we found no change in LH or TSH release
after p.o. and i.m. administration of DMI, CI and NF
(Laakmann 1980).
Cortisol and ACTH: The statement that the effect

258
of aminergic neuron systems on ACTH secretion is contra-
dictorily reflected in the literature (Martinet al.,
1977) is still valid today. Without further discussing
the question of whether cortisol secretion can be influ-
enced only via ACTH secretion, we primarily studied the
effects of different psychotropic drugs on cortisol se-
cretion.
After administration of DMI i.m. no change in corti-
sol secretion could be measured, but after i.v. administ-
ration there is a clear dose-dependent effect. Stimula-
tion of ACTH was measured after DMI (Wittmann et al.,
unpublished). CI provides similar study results.
The benzodiazepine diazepam and neuroleptics had no
effect on cortisol secretion.

A survey on the mechanisms of action and effects of

various psychotropic drugs is given in Table 1. This
shows that different psychotropic drugs have different
endocrine profiles, which are probably de~endent on their
mechanisms of action. The possibility of a ~lassifica­
tion of other unknown substances ih groups like anti-
depressants, anxiolytics or neuroleptics according to
their endocrine profiles will have to be verified in
future research studies.

Table 1: Effects of antidepressant, anxiolytic and

neuroleptic substances on secretion of various pituitary
hormones and their mechanisms of action.

reuptake pituitary hormones

NA 5-HT DA GABA GH PRL LH TSH Cort

++ + DMI ++ + 0 0 ++
+ ++ CI + ++ 0 0 ++
++ ++ NF ++ 0 0
ago. Diaz. + 0 0
block. Sulp./ 0 ++ 0
Hal dol

II. EFFECTS OF RECEPTOR BLOCKERS ON THE DMI-INDUCED GH,

PRL AND CORTISOL SECRETION

In order to ascertain whether different receptor

blockers have different effects on DMI-induced secretion
of GH, PRL and cortisol, studies were conducted in which
in addition to DMI, subjects received also a receptor-

259
Table 2: Effects of receptor-b~ocking substances on DMI-
induced GH, PRL and cortisol secretion.

Phentol. Yohim. Prop. Methy.

a-1/a-2 a-2 b 5-HT

DMI-induced secretion of:

GH ++ 0
PRL 0 0 ++
Cort. 0 ++ 0 0

blocking drug (Laakmann et al., 1983). These studies

show that phentolamine, an alpha-1/-2-receptor blocker
significantly inhibits stimulation of GH and has no sig-
nificant effect on secretion of PRL and cortisol.
Yohimbine, a primarily alpha-2-receptor blocker, in-
hibits DMI-induced stimulation of GH, does not affect
that of PRL, but increases that of cortisol.
Propranolol, a beta-receptor blocker, increased the
DMI-induced stimulation of GH, increased PRL stimulation
and had no significant effect on cortisol secretion
(Table 2).
Methysergide, a 5-HT-receptor-blocking substance had
no significant effect on DMI-induced GH and cortisol
secretion, but inhibited that of PRL significantly.
The results of studies on receptor-blocking sub-
stances to date seem to allow the interpretation that
DMI-induced GH stimulation is increased by alpha-recept-
ors, decreased by beta-receptors and not affected by
serotonergic receptors. The stimulation of PRL secretion
with DMI seems to be transmitted by serotonergic recept-
ors. Whether increased PRL secretion following propranolol
is due to a stimulatory effect of noradrenergic or dopa-
minergic neurons cannot yet be answered.
DMI-induced cortisol secretion is probably influ-
enced by alpha-2-receptors, but not by serotonergic re-
ceptors. Particularly the influence on cortisol secre-
tion, but also the influence on GH and PRL seems to
require substantiation by further research with other
receptor-blocking substances.

III. GH STIMULATION FOLLOWING DMI IN DEPRESSIVE PATIENTS

The stimulation of GH in depressive patients was

studied in recent years by a varie~y of researchers with
a variety of results. In addition to the IHT test, the

260
 a b d
40 40

l:
e -'E
"!
l:

@)
" ~
30 30

20 ,zo

10 10

d
Probanden <30 lCD 301),4 >30 lCD 2811ft1,2,3
Probanden Petlenten Probanden Patient en

Figure 1: GH increases after DMI 75 mg i.m.: maximal

Increases in a. premenopausal and b. postmenopausal
healthy female subjects and endogenous depressive pa-
tients (ICD 296/0.2), c. healthy male subjects and neu-
rotic depressive male patients (ICD 304) under 30 years
of age and d. healthy subjects and endogenous depressive
male patients (ICD 296/0.2.3) beyond 30 years of age.
(The numbers in circles indicate the age.)

L-DOPA, apomorphine and clonidine tests were conducted.

In a small group of patients we found that among endogen-
ous depressive male patients only a slight stimulation of
GH was triggered by administration of DMI, as opposed to
neurotic depressive male patients whose GH stimulation
was significantly higher i~!~~-!l·
We found GH stimulation in healthy premenopausal
women, but not in endogenous depressive premenopausal
patients. Neither healthy postmenopausal women nor
endogenous depressive patients showed a GH stimulation
(Laakmann 1980). Sawa et al. were able to corroborate
these results in 1982. Mesters et al. (1983) also
studied the effects of 75 mg DMI i.m. in healthy
subjects and depressed patients (classified according to
RDC) and reported partially similar results.

Further DMI-tests which we conducted in the mean-

time are not completely evaluated.

261
REFERENCES

Laakmann, G., 1980, Beeinflussung der Hypophysenvorder-

lappen-Hormonsekretion durch Antidepressiva bei
gesunden Probanden, neurotisch und endogen de-
pressiven Patienten. Nervenarzt 51:725.
Laakmann, G., Treusch, J., Schmauss, M., Schmitt, E.,
Treusch, U., 1982a, Comparison of growth hormone
stimulation induced by desimipramine, diazepam and
metaclazepam in man. Psvchoneuroendocrinology 7:
141.
Laakmann, G., Treusch, J., Eichmeier, A., Schmauss, M.,
Treusch, U., Wahlster, U., 1982b, Inhibitory ef-
fect of phentolamine 'n diazepam-induced growth
hormone secretion and lack of effect of diazepam
on prolactin secretion in man. Psychoneuro-
endocrinol02Y 7:135.
Laakmann, G., Schoen, H.W., Wittman, M., and Zygan, K.,
1983, Effect of receptor blockers on the desimi-
pramine-induced stimulation of growth hormone and
prolactin in man. in: E. Endroeczi et al., eds.,
"Integrative Neurohumoral Mechanisms", Elsevier/
North Holland Biomedical Press, Amsterdam.
Laakmann, G., Gugath, M., Kuss, H.-J., and Zygan, K., in
press, Comparison of growth hormone and prolactin
stimulation induced by chlorimipramine and desimi-
pramine in man in connection with chlorimipramine
metabolism.
Langer, G., 1979, The "prolactin-model" for the study of
the pharmacodynamics of neuroleptic drugs in man.
in: E.E. MUller, A. Agnoli, eds., "Neuroendo-
crine Correlates in Neurology and Psychiatry,"
Elsevier/North-Holland Biomedical Press, Amsterdam
New York-Oxford.
Martin, J.B., Reichlin, s. and Brown, G.M., 1977, "Cli-
nical Neuroendocrinology," Davis, Philadelphia.
Mesters, P., Linkowski, P., Hoffmann, G., Charles, G.,
Kerkhofs, M., Decoster, C., Vanderelst, M., and
Mendlewicz, J., 1983, Growth hormone stimulation
by desipramine in de)ression. in: Abstracts of
the VII World Congress of Psychiatry, Vienna.
Sawa, Y., Odo, S., and Nakazawa, T., 1982, Growth hormone
secretion by tricyclic and non-tricyclic antide-
pressants in healthy volunteers and depressives.
in: S.Z. Langer et al., ejs., "New Vistas in De-
pression. Advances in the Biosciences," Vol. 40,
Pergamon Press, Oxford-New York-Toronto-Sydney-
Paris-Frankfurt.

(Further references to be obtained from the author)

262
CHRONOBIOLOGY OF AFFECTIVE DISORDERS: ALTERATIONS IN CIRCADIAN
SECRETION OF PITUITARY AND PINEAL HORMONES IN BIPOLAR AND UNIPOLAR
DEPRESSION

Julien Mendlewicz and D.P. Linkowski

Department of Psychiatry, Erasme Hospital
University Clinics of Brussels, Free University of
B~ussels
B-1070 Brussels, Belgium

ABSTRACT

Affective disorders are characterized by episodic

alterations in mood and physiological functions indicating that
these illnesses may be considered as biological clock disorders.
This paper reviews some of the data on biological rhythm
disturbances in affective illness with special reference to
circadian alterations in secretion of pituitary and pineal
hormones and disturbances in sleep EEG parameters in depressive
illness.

INTRODUCTION

Biological rhythms are an integral part of life. The day and

night 24 hour period is present in almost all biological systems,
remaining constant in various environment. This observation has
led to the concept of endogenous biological clocks. The exact
synchronisation of the 24 hour period also involves external
synchronizers ("Zeitgeber"). The exact nature of biological clocks
remains unclear, but the hypothalamo-pituitary axls seems to play
a role in the maintenance of circadian rhythms both in animal and
man (1). It is by now clear that the pituitary gland constitutes a
major link in the neuroendocrine axis in man. Central neuro-
transmitters regulate the secretion of the hypothalamic neuro-
hormones which in turn may affect brain monoamine metabolism.
These neuroendocrine parameters are subjected to circadian
variations both in animal and man and may be implicated in the
pathogenesis of periodic psychoses. A new approach called
chronobiology permits to study temporal changes in clinical and
physiological symptoms in diseases such as cancer, endocrine

263
disturbances, cardiovascular and psychiatric conditions.
Furthermore, treatment response may also show important temporal
variations probably due to circadian variations in drug metabolism
and receptor sensitivity (chronopharmacology). Some affective
disorders are characterized by an alternation of depressive and
manic episodes and by periodic as well as diurnal disturbances in
mood and biological functions such as sleep, energy, appetite and
sex (evening improvement and morning worsening). These diurnal
changes may be related to alternations in day-night variation of
the mood and drive system. According to this concept,
desynchronisation may be an important pathophysiological aspect of
despression with some biological rhythms following its non
endogenous, or free running, circadian period, deviated from the
normal 24 hour period (in general phase advanced)~ Furthermore,
experimental studies have shown jet day and night shift to modify
energy level and concentration abilities while sleep deprivation
has been reported to temporarily alleviate depression.

The study of circadian and ultradian rhythms of biological

functions are thus of great importance in psychopathology, in
particular, manic-depression. According to this desynchronisation
hypotesis, manic-depression may be conceptualized as a "biological
clock" disorder. Phase shifts in biochemical circadian rhythms in
manic depressive patients have also been suggested for steroids,
electrolyte rhythms, neurochemical metabolites and body
temperature. While these studies based on brief observations
suggested there are circadian disturbances in manic-depressives,
longitudinal observations are essential to demontrate consistent
circadian patterns in affectively ill patients. The existence of
circadian variations in the release of serveral pituitary
hormones, such as ACTH, TSH, PRL and adrenal hormones such as
cortisol in man has been well documented. Moreover, for several of
these hormones, such as ACTH, TSH, cortisol and melatonin, it has
been shown that higher frequency non-periodic oscillations,
corresponding to secretory episodes, were superimposed to the
basal circadian rhythm. As a consequence of this hormonal
variability, consistent studies of hormonal secretion imply
repeated blood sampling over long periods of time (i.e. 24 hour
periods). This approach has been used by Sachar and his colleagues
(2) who have found that the normal 24-bour pattern of cortisol
secretion was disrupted in some depressed patients. There was an
increase in the number of secretory episodes with active secretion
of all peaks of plasma cortisol throughout the 24 hours. The
patterns almost returned to normal after the patient recovered.
They postulated the existence of an "abnormal drive from limbic
areas", i.e. there was a central limbic dysfunction in depression.
Other studies suggest that these disturbances cannot be explained
entirely as a simple stress response, since these abnormalities
are present in unanxious patients, during sleep, and are not
corrected after the administration of large doses of sedative
medication (3).

264
 In light of the biogenic amine hypothesis of affective
illness, recent studies have shown alterations in circadian and
seasonal rhythms of various neurotansmitter substances in the
plasma of depressed patients (4).

The above arguments led us to examine ciradian variations of

pituitary-pineal hormone levels in manic-dpression.

METHOD

We have investigated pituitary activity in patients

suffering from primary affective disorders. In this paper, we are
now reporting results on the 24-hour secretion of prolactin, TSH,
GH and melatonin during the depressed phase of subjects diagnosed
as bipolar manic-depressive, i.e. patients experiencing both manic
and depressive episodes and unipolar depressives, suffering from
depression only.

Serum THS (,uU/ml), prolactin (.UU/ml) and GH \UU/ML) were

measured by a double antibody radioimmunoassay. Plasma
concentration of melatonine was also determined by
radioimmunoassay. Estimated amplitudes and phases (day and night)
of TSH, prolactin, GH and melatonin patterns observed in depressed
patients were compared to the estimations obtained for control
patterns recorded in healthy volunteers. All patients studied were
free of medications for at least one week prior to the
investigation and were hospizalized in an inpatient unit for
pirmary depressive episodes severe enough to warrant
hospitalization. Severity of the depressive illness was assessed
by the Hamilton Rating Scale. Blood samples were drawn for 24
hours through a plastic indwelling catheter. Blood samples were
collected every hour during day time and every thirty minutes
during the night. All patients were confined to bed had normal
breakfast, lunch, supper and their noctural sleep was not
interrupted. Day time sleep was prevented and sleep times were
recorded by trained nurses. All patients and controls were
investigated throughout the calendar year.

RESULTS

Prolactin (Table I)

The prolactin patterns of all depressive patients as a group

(n=18) showed no significant difference with patterns observed in
healthy subjects (n=6 males). Nevertheless, the mean prolactin
level over 24 hours was significantly lower in bipolar patients as
compared to unipolars. This was mainly due to the absence of sleep
related elevation of PRL in 6 out of 8 bipolar patients (75 %) in
whom maximum PRL secretion occurred during wakefulness (phase
advanced), wh~reas maximum PRL concentrations were observed during
sleep in all unipolar patients as in normal controls.

265
 TABLE I: 24-HOUR PROLACTINE PROFILES IN UNIPOLARS AND BIPOLARS

NORMALS UNIPOLARS BIPOLARS

N = 14 N = 10 N =7

24 HOUR MEAN (J...;U/ml) 283 .:t 144 351 -+ 156 165 .:t 62
WAKE MEAN (~U/ml) 235 .:t 136 321 -+ 158 160 .:t 59
DIFFERENCE BETWEEN
SLEEP MEAN AND WAKE MEAN + IN 14/14 + IN 8/10 + IN 2/8
RATIO SLEEP MEAN/WAKE
MEAN 1. 77 .:t 69 1. 37 .:t 46 1. 06 .:t 29

TSH (Table II)

The diurnal patterns of TSH levels studies in unmedicated

depressed female patients (n = 13) differed greatly from those
exhibited by normal subjects previously investigated (n = 6 males,
10 females). The mean 24 hour TSH level was lower in all depressed
patients when compared to normals. In these patients, the rhythm
appears to be desynchronized, no early morning peak being
evidenced. In some cases, a maximum occurred before midnight.
Higher frequency variation of plasma TSH could also be observed
between unipolar and bipolar patients. Furthermore, thyroid
function was found to be normal in all patients and controls.

TABLE II: NYCTOHEMERAL TSH IN DEPRESSION

UNIPOLAR DEPRESSION BIPOLAR DEPRESSION

NORMAL THYROID FUNCTION NORMAL THYROID FUNCTION

NORMAL BASAL TSH LEVELS NORMAL BASAL TSH LEVELS

ABNORMAL CIRCADIAN TSH RHYTHM NORMAL CIRCADIAN TSH RHYTHM

LOWER 24 HOUR TSH MEAN NORMAL 24 HOUR TSH MEAN
SLEEP-WAKE RATIO OF TSH IS LOWER NORMAL SLEEP-WAKE RATIO OF TSH
ABSENCE OF NOCTURAL RISE OF TSH NOCTURAL RISE OF THS

266
G. H.
The 24 hour profile of plasma GH was investigated in 12
depressive male patients and 10 healthy male volunteers.

Unipolar depressives secreted more GH than age-matched

controls, daytime peaks being larger and more frequent. GH
secretion was also slightly enhanced in bipolar patients but not
significantly so. Table III shows the association between night GH
secretion and sleep stages in our depressed subjects.

TABLE III: GH SECRETION AND SLEEP IN DEPRESSED SUBJECTS

AMOUNT OF GH WAKE RFH sw I + II

SECRETED IN (MIN) (MIN) (MIN) (MIN)
SLEEP (NG)

UNIPOLARS

L 28 275 4 0 123
R 241 194 14 20 146
GI 0 211 28 10 200
GO 0 387 15 1 193
MA 91 285 18 5 164
F 41 129 95 118 154
MEAN 662:91 2472:89 292:33 262:45 1632:29

BIPOLARS

p 0 232 33 26 192
J 229 193 52 112 148
D 130 70 105 5 302
Me 179 143 27 0 27
v 168 164 100 29 183
G 151 221 14 30 103
MEAN 1432:77 1712:60 552:39 34.!:40 1592:93

The amount of GH secreted during sleep is greater in

bipolars, but there is more wake time in the unipolars. There was
no difference in REM, slow wave sleep (SW) and stages I+II.

Table IV illustrates the association between slow wave sleep

(SW) and GH spikes in normal controls and unipolar and bipolar
patients.

267
 TABLE IV: SW SLEEP AND GI SPIKES

YOUNG NORMALS

TIME OF GH SPIKE % OF SW DURING ASCENDING

PORTION

B 02.15 96%
04.15 O%
c 01.00 91%
03.30 O%
D 00.45 89%
04.00 44%
E 00.45 50%
01.45 41%
F 01.15 86%

OLDER NORMALS

VH 01.15 O%
VG
VE 00.30 O%
05.30 1%
D 00.30 48%
Pi 00.15 60%
01.45 40%

268
 TABLE IV: SW SLEEP AND GH SPIKES (continuation)

UNIPOLARS

TIME OF GH SPIKE %OF SW DURING ASCENDING

PORTION

Ra 02.30 17%
05.00 O%

Ma 03.30 2%
F 01.30 O%
07.15 2%

Gi

Go

La

BIPOLARS

Po
J 02.00 79%

Du 02.00 1%
04.30 O%

Me 23.45 0%
01.15 0%
04.30 O%

v 00.45 0%
02.00 7%
04.15 28%

G 01.00 30%

All young controls have a GH spike shortly after sleep

onset. An average of 82 + 18 %of the ascending portion of these
spikes was sent in slow-wave sleep. In 3 of the 5 older controls,
an early GH spike was also observed. Only two of these spikes were
associated with SW sleep. In the unipolar depressives, no GH peak

269
occurred in the 3-hour period following sleep onset and only one
of the 5 noctural GH spikes was associated with concomitant SW
sleep. Three unipolars did not show any GH peak during the night.
In the bipolars, five of 6 patients had an early sleep GH spike
but only one of them was parrallel with SW sleep. Six out of a
total of 10 noctural GH spikes occured without any concomitant
stage III or IV. Thus the classical association between noctural
GH secretion and SW sleep was absent in older male controls as
well as in depressed patients. One interesting aspect of the
desynchronization hypotesis could be related to the increased
incidence of affective (manic and depressive) relapses in spring
and autumn when day light is either longer or shorter. It is thus
possible that in genetically predisposed subjects, some circadian
psychological clock parameters may be desynchronized during these
periods. In those susceptible individuals circadian
desynchronization may be triggered by ultradian seasonal
variations and then continue into a free running cycle. Melatonin
is a pineal hormone particularly sensitive to day-night changes in
exposure to light. It is also under central noradrenergic control.
Melatonin is thus of great interest in studying cyclic manic-
depressive patients. Preliminary data on 24 hour plasma melatonin
concentrations are available for 4 female depressed patients
before and after treatment and 5 normal controls (males).
Secretory episodes are observed during wakefulness in all subjects
(phase advanced(, but they are of higher magnitude in depressed
patients and seem to appear at abnormal times (late afternoon or
evening before onset of sleep). Furthermore the circadian rhythm
of melatonin is less apparent in depressed patients. The night/day
ratio for melatonin is 1.38 in depressed patients and 2.8 in
normals. Noctural rise of melatonin is almost absent in 3 of 4
depressed patients who show an elevation of the pituitary hormone
during day time. Finally, no significant changes in 24 hour
melatonin patterns can be seen in depressed patients after
antidepressant treatment and following remission. (Table V)

It is thus clear that the altered circadian secretion of

melatonin in depression is not state dependent.

Alterations in circadian rhythms for plasma pituitary and

pineal hormones such as prolactin, GH, TSH and melatonin are
present in some depressed patients and may be related to basic
disturbances in circadian oscilators of central catecholaminergic
and serotonergic activity in affective illness. Horeover, it is

270
 TABLE V

NYCTOHEMERAL MELATONIN NYCTOHERMERAL MELATONIN

IN DEPRESSION IN CONTROLS

NORMAL OR NO CIRCADIAN RHYTHM CIRCADIAN RHYTHM WITH NOCTURAL

RISE OF MELATONIN

REDUCTION OR ABSENCE OF NOCTURAL

RISE OF MELATONIN

AVERAGE NIGHT/DAY RATIO FOR NIGHT/DAY RATIO FOR MELATONIN

MELATONIN= 1,38 = 2,80
SECRETORY EPISODES THROUGHOUT SECRETORY EPISODES THROUGHOUT
THE 24-HOUR SPAN THE 24-HOUR SPAN

possible that cholinergic-adrenergic interactions are also of

significance. The rapid changes in mood and the abruptness of the
desynchronization observed in manic-depressive patients make it
unlikely that these phenomenon are related to primary oscillations
in central neurotransmitter synthesis or turnover rate. More
fruitful hypotheses may be formulated in terms of infradian,
ultradian or circadian variation in specific brain receptor
sensitivity or more complex behavioral modulation through
endogenous neuropeptide substances. The brain distribution of
other releasing factors and peptides has not yet been reporded and
when the brain effects of hypothalamic hormones and peptides will
be further elucidated, the clinical conditions in which their
actions are investigated may be better unterstood. Nevertheless,
the chronobiological studies described above combining sleep and
neuroendocrine evaluation over long periods of time in the study
of abnormal behavior are most promising and may enable us to
better understand cyclical alterations of hypothalamic and
pituitary functions in man, although it is still premature to draw
firm conclusions from the neuroendocrine abnormalities as to the
specific nature of the underlying neurotransmitter or neuropeptide
disturbances in psychopathology.

271
REFERENCES
1. Richter, C.P., Abnormal but Regular Cycles in Behaviour and
Metabolism in Rats and Catatonic-Schizophrenics. In:
Psychoneuroendocrinology, M. Reiss, Ed. Grune & Stratton New
York : 168-181, 19.58.
2. Sacher, E.J., Hellman, L. and Fukushima, D.K.: Cortisol
production in depressive illness. Arch. Gen. Psychiat. 23:
289-298, 1980.
3. Stockes, P.E., Studies on the Control of Adrenocortical
Function in Depression. In: Recent Advances in the
Psychobiology of Depressive Illnesses, Williams, Katz and
Shield, Washington D.C., U.S. Dehw Publ.: 199-220, 1972.

4. Riederer, P., Birkmayer, W., Neumayer, E., Ambrozi, L. and

Linauer, W., The Daily Rhythm of HVA, VMA, (VA) and 5HIAA in
Depression Syndrome. J. Neural Transm., 35:23-45, 1974.

272
NEUROENDOCRINE CONTROL OF PITUITARY FUNCTION IN EXTRAPIRAMIDAL

DISORDERS

T. Caraceni, P. Giovannini, E. Parati,

G. Scigliano and E. E. Muller

Istituto Neurologico C. Besta

Via Celoria 11
20133 Milano, Italy

INTRODUCTION

The purpose of this study was to reconsider the functional

status of the hypothalamus-pituitary axis in diseases of the
extrapyramidal system. The theme of the paper is behavior of the
secretion of growth hormone (GH), thyroid-stimulating hormone
(TSH), and prolactin (PRL), particularly in Huntington's disease
(HD) and Parkinson's disease (PD). The choice of these hormones
derives from the fact that their secretion is under a dopaminergic
control 1 and that this latter function is notedly involved in the
pathologies under examination.

HUNTINGTON'S DISEASE

HD is a hereditary degenerative pathology of the central ner-

vous system characterized by a dominant autosomal heredity, the
presence of pathological involuntary movements (choreic hyperkine-
sia), behavioral alterations and dementia. 2 There is evidence in
the literature to indicate an involvement of the hypothalamic
structures in this disease. 2 • 3 The most interesting neuroendo-
crine alterations reported in HD concern the secretion of GH. The
basal levels of the hormone have been reported to be normal or
slightly high. An increased response of GH secretion was demon-
~trated with the insulin tolerance test (ITT~ and by the arginine
test of Keogh et al., 4 Leopold and Podolsky, and Lavin et al. 6
A recent study conducted by Durso et al. 7 documented a marked
response in GH to various stimuli, which the authors attributed to
low levels of somatostatin in HD patients. These findings suggest
an altered hypothalamic neurotransmitter function as regards the
secretion of GH in patients affected by HD.
273
 Since most of the studies have been carried out on series of
patients who had been suspended, before the endocrinological
tests, from a preceding neuroleptic treatment for a relatively
brief time, we conducted a study on a selected group of 14 HD
patients who had not been treated with neuroleptic agents. Our
group of patients was subjected to dopaminergic stimuli (bromo-
criptine, 2.5 mg orally; lisuride, 200 ~g orally) and a stimulus
with a synthetic analog of met-enkephalin (FK 33-824, 0.5 mg
i.v.). Under these experimental conditions, GH levels were not
different than those of controls. The responses obtained with
these different stimuli were considered, on the whole, not differ-
ent from those obtained in normal subjects; however, analysis of
the individual responses allowed the identification of three dif-
ferent populations within the group of choreic patients: normore-
sponders, whose endocrine response was perfectly normal, hyper-
responders, whose response was more than twice the maximum
response obtained in controls, and hyporesponders. These data 8
together with recent evidence of Destee et al., 9 indicate a
variable involvement of the hypothalamic mechanisms of control of
GH secretion in HD patients.

As regards PRL, most studies in the literature have reported

normal basal levelsG,lO but the opposite has been reported. In
this context, the studies on PRL secretion during 24 h in such
patients by Polleri et al. 11 are interesting. In such conditions
of study, sinuisoidal rhythms of secretion not different from con-
trols have been found, with conservation of nocturnal peaks. The
responses induced by dopaminergic stimuli or by thyrotropin-
releasing hormone (TRH) showed discordant behaviors: normal 12 or
diminished. 13 Our previous results on 30 HD patients 1 D showed
normal basal PRL levels and normal responses to various stimuli
(TRH, apomorphine, L-dopa, bromocriptine, domperidone, sulpiride),
in agreement with other studies.

In our recent experience conducted on untreated HD patients, 8

tests of stimuli with lisuride, bromocriptine and FK 33-824 were
performed, and no variations in basal PRL levels or responses
induced by the stimuli were observed, with respect to controls.
In 10 cases of this series, the responses in PRL and TSH to TRH
were not different from those of controls.

Our data, in untreated patients, thus indicate that the

mechanisms of controls of PRL secretion are preserved in hunting-
tonian patients under these conditions and confirm that the varia-
tions are dependent in some way on the neuroleptic therapy.

PARKINSON'S DISEASE

PD is a degenerative disease of the central nervous system

that usually begins around the sixth decade of life and is

274
essentially characterized by alteration of motorial function,
which is expressed clinically by the classical symptomatological
triad - akinesia, rigidity and tremor. From the neurochemical
point of view, it is characterized by a reduction in dopamine at
the level of the nigrostriatal pathway. Involvement of the hypo-
thalamic structures 14 • 15 is apparent in PD and indicates that
there may be alterations of the mechanisms that regulate the func-
tion of the hypothalamus-pituitary axis. Moreover, the existence,
for some time, of drugs that can correct the dopaminergic hypo-
function, i.e., L-dopa and more recently the direct agonistic
dopamine drugs, suggests the possibility that such chronic treat-
ments modify neuroendocrine behavior.

Considering now the behavior of GH in parkinsonian patients

not previously treated with dopaminergic drugs, most reports in
the literature indicate the existence of normal basal levels. 16 • 17
Studies on the secretion of GH over 24 h have shown a flattening
of the curve for the disappearance of nocturnal peaks. 18 Such a
behavior is comparable to that observed in elderly subjectsl9 and
therefore ascribable to a phenomenon consequent to aging. The
data relative to the responses of GH to various stimuli are
discordant. In fact, a normal or reduced response was found with
the ITT and arginine, 20 as well as with two dopaminergic stimuli
(L-dopa and bromocriptine).20,21

In our recent experience conducted on 50 parkinsonian

patients, untreated and all in the initial phases of the disease,
we confirmed the normality of the basal GH levels and of the
responses, with respect to controls, to various stimuli such as
infusion to dopamine (5, 10, 20 mg/20 min i.v.), nomifensine (200
mg orally), lisuride (50 ~g i.v.), domperidone (10 mg i.v.), and
the ITT (data not reported and not published). Normal PRL and TSH
responses were also observed with these stimuli. In parkinsonian
patients chronically treated with L-dopa plus peripheral decarbox-
ylase inhibitors (PDI), the basal GH levels were not different
from those of controls, 16 whereas the response induced by dopamin-
ergic stimuli appeared reduced in such conditions.l7 Such
behavior can be explained as a result of a phenomenon of densiti-
zation of the DAergic receptors that control GH secretion, conse-
quent to chronic stimulation by 1-dopa. Such behavior was also
confirmed by us in a group of six patients chronically treated
with DA-agonistic drugs (bromocriptine and lisuride). In the same
subjects, an ITT was also effected, and in these conditions there
was a release of GH superimposable to that obtained in untreated
patients. This finding suggests that also in conditions of
chronic treatment with DA-agonists the pituitary reserve of GH is
maintained and that the secretory stimulus of the ITT acts through
a pathway not mediated by dopamine and perhaps by specific hypo-
thalamic receptors (glucoreceptors).

275
 As regards the behavior of PRL and TSH in untreated
parkinsonian patients, basal levels have been consistently
normal.l 7 , 22 • 23 Studies of the circadian secretion of PRL also
showed normal behavior of the curves, except for a reduced
nocturnal secretion. 24 The most interesting data obtained by us
in untreated patients and confirmed by other reports in the
literature 22 is that of a reduced release of TSH after
stimulation with TRH (Fig. 1). 23 This data cannot be considered
as the consequence of ageing, as suggested by Snyder and
Utiger 25 since the comparison was done on normal controls matched
for sex and age. A possible interpretation of this reduced TSH
release is that untreated parkinsonian patients develop a
hypersensitivity of DAergic receptors involved in control of
secretion of this hormone at the pituitary level that partially
blocks the response to TRH stimulus.

At this point, it is interesting to observe that in one of

our series of patients chronically treated with L-dopa plus PDI,
the TSH response to TRH was restored, being much closer to the
secretion curve of controls as if the chronic treatment reestab-
lished a more physiological DAergic receptor responsiveness. 23

In a group of six patients treated chronically with DA-

agonists (bromocriptine and lisuride) recently studied by us, we
found that TSH secretion after stimulus with TRH was superimpos-
able to that observed in patients treated chronically with DA
mimetics (Fig. 2). This appears to confirm the hypothesis of a
down regulation of the DAergic receptors involved in the control
of TSH secretion induced by chronic dopaminergic treatment.

0 CONTROL SUBJECTS
- - P O UNTREATED
hTSH
18 ~Ujml

-30 30 45 60
TIMI::Cmin>

Fig. 1. TSH secretion after TRH stimulation ( 200 11g i. v.) in

untreated PO patients with respect to controls. *, Significant.

276
 ._- - ePD under OA agonists
..,._.....PO untreated

t',
10 hTSH
pU/ml
I: '.,_ ___ _
I '

,, ! !',
f.

.
l 1'\
I
I \

TRH
200J!gi.¥.,
r
:
·r~
3o cs eo go
TIMEtmin)

Fig. 2. Secretion of TSH, after stimulation with TRH (200 ~g

i.v.) in PD patients chronically treated with DA agonists.

REFERENCES

1. E. E. Miiller, G. Nistico and U. Scapagnini, "Neurotransmit-

ters and Anterior Pituitary Function," Academic Press, New
York (1978).
2. G. W. Bruyn, Huntington's chorea, history, clinical and labo-
ratory synopsis, in: "Handbook of Clinical Neurology,"
vol. 6, P. J. Vinken and G. W. Bruyn, eds., North-Holland,
Amsterdam (1968).
3. E. D. Bird, Neuroendocrine changes in Huntington's disease -
An overview, Adv Neurol 23:291 (1979).
4. H. J. Keogh, R. H. Johnson, R. N. Nanda and W. R. Sulaiman,
Altered growth hormone release in Huntington's chorea,~
Neurol Neurosur2 Psvchiat 39:244 (1976).
5. N. A. Leopold and s. Podolsky, Exaggerated growth hormone
response to arginine infusion in Huntington's disease, J
Clin Endocrinol Metab 41:160 (1975).
6. P. J. M. Lavin, I. Bone and P. Sheridan, Studies of hypo-
thalamic function in Huntington's chorea, J Neurol
Neurosurg Psvchiat 44:414 (1981).
7. R. Durso, c. A. Tanninga, A. Denaro, S. Ruggeri and T. N.
Chase, Plasma growth hormone and prolactin response to
dopaminergic GABA mimetic and cholinergic stimulation in
Hunt5 n~ton' s disease, Neurology 33: 1229 (1983).
8. T. A. Caraceni, E. A. Parati, P. M. Giovannini et al.,
Neuroendocrinological findings in untreated huntingtonian
patients, Arch Neurol in press (1983).
9. A. Destee, H. Petit, P. Fossati and P. Warot, Choree de
Huntington et hormone somatotrope: explorations dynamiques
dans 27 cas, Rev Neurol, 137:21 (1981).
10. E. E. Miiller, E. A. Parat{, D. Cocchi, D. Zanardi and T. Cara-
ceni, Dopaminergic drugs on growth hormone and prolactin
secretion in Huntington's disease, Adv Neurol 23:319 (1979).

277
11. A. Polleri, F. Savoldi, A. Muratorio et al., Circadian
rhythmicity of prolactin secretion in Huntington's chorea,
Life Sci 26:1609 (1980).
12. R. J. Chalmers, R. H. Johnson, H. J. Keogh and R. N. Nanda,
Growth hormone and prolactin response to bromocriptine in
patients with Huntington's chorea, J Neural Neurosurg
Psvchiat 41:135 (1978).
13. M. R. Hayden, M. Paul, A. I. Vinik and P. Beighton, Impaired
prolactin release in Huntington's chorea, Lancet 2:423
(1977).
14. J. W. Langston and C. S. Forno, The hypothalamus in
Parkinson's disease, Ann Neural 3:129 (1978).
15. E. Ohama and F. Ikuta (1976) Parkinson's disease: distribu-
tion of Lewy bodies and monoamine neuron system, Acta
Neurooathol (Berlin) 34:311 (1976).
16. P. 0. Lundberg, Blood levels of FSH, LH, TSH and GH in
parkinsonian patients before and during L-dopa treatment,
Acta Neural Scand 48:427 (1972).
17. A. Martinez-Campos, P. Giovannini, E. Parati et al., Growth
hormone and prolactin stimulation by Madopar in Parkin-
son's disease, J Neural Neurosurg Psy~hiat 44:1116 (1981).
18. L. Murri, A. Iudice, A. Muratorio et al., ~pontaneous noctur-
nal plasma prolactin and growth hormone secretion in
patients with Parkinson's disease and Huntington's chorea,
.Eur Neural, 19:198 (1980).
19. H. E. Carlson, J. C. Gillin, P. Gorden and F. Snyder, Absence
of sleep-related growth hormone peaks in aged normal
subjects and acromegaly, J Clin Endocrinol Metab 34:1102
(1972).
20. F. Cavagnini, M. Peracchi, G. Scotti et al., Effect of L-dopa
administration on growth hormone secretion in normal
subjects and parkinsonian patients, J Endocrinol 54:425
(1972).
21. J. D. Parkes, A. G. Debono and C. D. Marsden, Growth hormone
response in Parkinson's disease, Lancet 1:483 (1976).
22. P. J. M. Lavin, M. J. Gadwell, P. K. Das et al., Effect of
levodopa on thyroid function and prolactin release, Arch
Neural 38:759 (1981).
23. A. Martinez-Campos, P. Giovannini, A. Novelli et al., Thyro-
tropin and prolactin responses to thyrotropin-releasing
hormone in patients with Parkinson's disease, Acta
.Endocrinol. 99:344 (1982).
24. A. Polleri, P. Masturzo, G. Murialdo et al., Chronobiology of
prolactin secretion: a marker in physiology and pathology,
in: "Central and Peripheral Regulation of Prolactin
Function," R. M. MacLeod and U. Scapagnini, eds., Raven
Press, New York (1980).
25. P. J. Snyder and R. D. Utiger, Response to thyrotropin-
releasing hormone (TRH) in normal men, J Clin Endocrinol
Metab 34:380 (1972). -

278
NEUROENDOCRINOLOGICAL PROFILE OF MAJOR DEPRESSION

C. N. Stefanis, B. Alevizos, C. Soldatos,

G. Papadimitriou, G. Tolis, and M. Liparaki

Dept. of Psychiatry, Athens National University

Eginition Hospital, Athens, Greece

INTRODUCTION

Postulates of endocrine involvement in the pathophysiology

of mental disorders date back to the beginning of the century
(Bleuler, 1954). However, it is only in the pasts few years, with
the advancement of biochemical techniques for accurately measuring
hormonal blood levels, that such an involvement could reliably be
tested. Hypothalamic-pituitary-adrenal (HPA) hyperactivity has
been proposed to the most prominent and most specific biological
marker of endogenous depression (Carroll, 1980). In fact, it has
been reported that about half of the patients suffering from major
depression fail to suppress cortisol secretion following
dexamethasone administration (Carroll, 1982). Other studies in
which the TSH response to TRH is used have indicated an
abnormality of the hypothalamic-pituitary-thyroid (HPT) axis in
patients with this type of affective disorders (Loosen and Prange,
1980; Gold et al, 1981). Recently additional provocative tests
like clinidine, apomorphine and insulin tolerance were used to
evaluate other hormonal systems in depression.

In the past three years neuroendocrinological

investigations, usually combined with studies on sleep and
biorhythms, dominated the biological research of the affective
disorders. But as new and very frequently conflicting data are
accumulating, extensive studies on control subjects and large,
clinically homogeneous patient populations, studied with multiple
neuroendocrinological tests, are urgently needed.
279
 In this presentation we will outline some of the results
obtained in the past four years from an ongoing study conducted in
our Department on a variety of neuroendocrinological parameters in
a carefully selected sample of 200 patients with major depressive
episodes (n.169) and schizoaffecttve or schizophreniform disorders
(n.31).

MATERIAL AND METHODS

In our ongoing study 200 patients are so far included. Of

these 169 wre diagnosed by RDC criteria, and classified according
to DSM III to the group of major affective disorders either of the
bipolar (n.58) or of the unipolar (n.111) type. The remaining
patients were diagnosed as suffering either of schizoaffective or
schizophreniform disorder (n.31). The age of the patients ranged
from 15 to 72 years, and females accounted for 62 %of the total
group. Every patient was diagnosed by two independent
psychiatrists, and in addition to his clinical assessment he was
rated for the severity of his depression by the Hamilton's
Depression Scale (HDS). Hormonal levels were measured with the
standard techniques. The laboratory staff was not aware of the
diagnosis and had no information on the clinical picture of the
patients.

RESULTS

A. Studies on the activity of the HPA axis

1. Dexamethasone suppression test (DST)

DST was performed in all 200 patients included in this
study. Cortisol plasmal levels were determined at 8.00, 16.00 and
23.00 hr of the day following the administration of 1 mg of
dexamethasone at 23.00 hr. According to established cirteria
(Carroll et al, 1981), an abnormal test response was defined as a
post-dexamethasone cortisol level of 7 50 ng/ml. A cortisol
radioimmunoassay kit for clinical assays was used for cortisol
measurements. All precautions were taken to exclude those factors
which might yield false positive DST's.

As shown in Table I, 38 %of unipolar depressed patients and

34 %of bipolar patients in the depressed phase failed to suppress
cortisol after dexamethasone. Two out of ten patients (20 %) with
schizoaffective psychosis also were found to be nonsuppressors.
There are 32 patients who are not included in the table because one
of the three Post-Dex cortisol measurements was missing.

280
 TABLE I

DEXAMETHASONE SUPPRESSION TEST

CORTISOL
PATIENTS N SUPPRESSORS NON-SUPPRESSORS

Unipolar
Depression 102 62% 38%
Bipolar
Depression 45 66% 34%
Schizoaffective
Psychosis 10 80% 20%

As shown in Table II suppressibility is not steady during

the 24 hs period following Dex administration. It is to be noted
that a very early excape of cortisol suppression (at 8.00 hr post
dex) occurred seven times more frequently in unipolar compared to
bipolar patients: Age, sex, severity of depressive mood rated by
the HDS, as well as family hisory of affective disorders, were
found not to be significantly associated either with the
supporessor or the non-suppressor group of patients.

TABLE II

PERCENTABE OF PATIENTS WITH DST POSITIVE

AT FIXED POST-DEX TIME INTERVALS (n.140)

Time UP BP SCHIZOAFF.

8 a.m 20 3 0
4 p.m 24 24 10
11 p. m 29 30 20
any time 38 34 20

2. N-POMC studies

N-terminal of the pro-opio-melanocortin (N-POMC) the

precusor of ACTH was studied in collaboration with Cr. Chretien's
Laboratory in Montreal. It was measured according to extablished
methods (Chan et al, 1983) at 8.00 and 23.00 hr prior to and at
8.00, 16.00 and 23.00 hr following the administration of 1 mg
dexamethasone, in the same plasma that immunoassayable cortisol was
determined. Thirty two patients, so far, all suffering from major
depression, took part in the study.

281
 TABLE III

N-POMC POST-DEX

Patients tested 32
Criterion A: N.-POMC ~ 100 pg/ml
Non suppressors = 86 %
Suppressors = 14 %
Criterion B: N-POMC ·~ 150 pg/ml
Non suppressors = 68 %
Suppressors = 32 %
Criterion C: N-POMC ~ 200 pg/ML
Non suppressors = 42 %
Suppressors = 58 %
Basal N-POMC values wre found to be igher than 100pg/ml in
80 %of patients. Report data in control subjects indicate normal
values below 100pg/ml.

Results concerning suppressibility of N-POMC following

dexamethasone administration, using three cirteria, are shown in
Table III. It is evident that dex suppression of N-POMC did not
occur but only in a small percentage of our patients when
criterion A was used. Even when criterion C (N-POMC~200pg/ml) was
used 42 %of our patients wre classified as non-suppressors.

Using again all three criteria for N-POMC post-dex

suppression and the standard one for cortisol suppression
( 50ng/ml) a pronounced disassociation between susceptibility to
suppression of cortisol and N-POMC by Dex was observed (Table IV).

TABLE IV

PATIENTS (32) TESTED WITH BOTH F AND N-POMC AFTER 1mg DEX
Suppressions in both 12.5 %
N-POMC 100 pg/ml Non Suppression in both 25 %
Disassociation 62.5 %

Suppressions in both 30 %
N-POMC 200pg/ml Non Suppressions in both 20 %
Disassociation 50 %

Suppressions in both 47.5%

N-POMC 200 pg/ml Non Suppressions in both 17.5%
Disassociation

282
3. Cortisol rhythm studies

Several hormones are known to ahve circadian rhythm closely

linked to the sleep-wake-cycle. Cortisol among them has a
prominent rhythm with the lower plasma levels in the early hours
of sleep and the highest about the time of awakening (Krieger,
1979b).

In 80 of our depressed patients (both unipolar and bipolar)

cortisol plasma levels were determined at 8.00 and 23.00 hrs. As
shown in table V the amplitude of the diurnal rhythm was found to
be considerably blunted in about half of our patients. Although
rhythm cannot be reliably studied only with two measurements in
time, due to the large number of patients, these results acquire
an indicative value and are in agreement with previous reports on
disruption of cortisol diurnal rhythm in depression (Sachar,
1975). This abnormality is found in pure endocrine disorders, such
as diencephalic Cushing's syndrome and led to the evaluation of
the suppressibility of the HPA axis in depression with the
dexamethasone suppression test.

TABLE V

DIURNAL RHYTHM OF PLASMA CORTISOL

RHYTHM
PATIENTS N ABSENT : <50 % <Ll> PRESENT

Unipolar
Depression 44 42% 58%

Bipolar
Depression 36 50% 50%

B. Studies on the activity of the HPT axis

The role of the hypothalamic-pituitary-thyroid (HPT) axis

has been extensively studied in recent years. Although conflicting
findings on thyroid activity in depression by measuring T4 and T3
were reported, the serum thyrotropin response to thyrotropin-
releasing hormone (TRH) in affective disorders yielded more
conistent and positive results. A blunted response to TRH in
patients with major depression has been reported by several
investigations. The sencitivity of the test however varies from one
study to the other and there are in the literature sporatic
reports of negative results (see P. Loosen et al, 1982).

283
 TABLE VI

TRH-TEST

TSH-RESPONS
PATIENTS N BLUNTED NORMAL
UNIPOLAR 84 40% 60%
BIPOLAR 52 32% 68%

From our material 136 patients were studied with the TRH
test using the standard procedure and techniques (M. Gold et al,
1980; L. Extein et al, 1982). A blunted TSH response to TRH was
defined by maximal TSH response of ~TSH 7 <piU/ml. the results
obtained in both unipolar and bipolar patients had a blunted
response but in both groups the percentage was not as high as
expected from the literature.

Sex, Age, severity of depressive symptomatology and family

history of depression were not significantly associated with the
TSH response to TRH. Blunted TRH response was not consistently
associated with DST positives. There was a considerable overlap
between responses to TRH and post-Dex cortisol suppressibility but
disassociation between the two was found in 28 % of the cases.

C. Prolactin (PRL) secretion studies

There are only few and conflicting results on the 24-hour

levels of PRL in depressed patients (J. Mendlewicz et al, 1980;
P. Tosca et al, 1982). In our study of the circadian rhythms of
PRL secretion we have used a continuous blood withdrawal pump and
we found that in 4 depressed patients the PRL secretion was
significantly lower than in 5 male control subjects. Furthermore
when we compared the PRL plasma values in a group of 41 depressed
patients in two consecutive days at 8.00 hr prior to and following
dexamethasone administration a substantial decrease of PRL
secretion in response to Dexamethasone was observed. However,
increase rather than decrease of PRL was seen in those patients of
the group with an early excape of cortisol suppression to
dexamethasone. Conversely, in 48 patients in which TRH test was
performed the mean PRL secretion for the total group was
increased. However, in patients with blunted TSH response to TRH
and PRL increase was less pronounced and in some of them did not
occur.

These resuls provide further evidence of PRL involvement in

the pathophysiology of depression.

284
DISCUSSION

It is beyond any doubt that a new era in psychiatric

research has been opened by recent neuroendocrine investigations.
Particularly in the fielf of affective disorders they yielded
promising results which provide, a emote but still valid link
between biological processes and psychopathology. However, at this
stage it is rather premature to draw conclusions regarding the
significance the observed hormonal changes may have on our
understanding of the biologic mechanism underlying the major
depressive disorders.

Our findings definitely provide no additional information

towards this direction. Since however they derive from a large and
carefully sleeted population of patients who were studied not with
one but with a battery of neuroendocrinological tests, they may
contribute to clarifying certain points which are currently raised
in the field.

We will first discuss the issue of sensitivity of the most

widely tests used, i.e. DST and TRH. Our findings confirm all
previous reports that by these two tests an abnormality in both
the HPA and HPT systems of patients with major (primary) depression
can be manifested. However the percentage of patients manifesting
this abnormality is not as high as appears in the literature: Both
tests fell below the sensitivity level of 50 %and in fact they
ranged from a high level of 40 % for blunted TSH response to TRH
in the unipolar depressives to a low of 34 % for DST pos in the
bipolar depressives. As our results indicated Post-Des cortisol
single measurements at 4 pm, a precedure prefered by increasing
number of investigators, lowers the sensitivity even further.
Compared to DST and TRH suppressibility of N-POMC to Dex was found
to be a strikingly more sensitive test. Even when the strickest
criterion C was applied almost half of the depressed pateints
failed to suppress N-POMC after Bexamethasone below the level of
250 pg/ml. It is though to be kept in mind that there are only few
studies on N-POMC and all of them in non-psychiatric conditions.
Hence, standardisation is still lacking and consequently our
findings on N-POMC have to be interpreted with caution.

In view of the low sensitivity of both DST and TRH and

regardless of issues of specificity raised by some authors. (Coppen
et al. 1980; Shulman and Diewold, 1977; Holsboer et al. 1980) it
is questioned what degree of diagnostic confidence these tests may
have in clinical practice. Overliance on these tests may lead to
their abuse by clinician who may overlook sound clinical evidence
and change their diagnosis and treatment schedule solely on the
basis of test-negatives from the laboratory.

285
 In this paper only results obtained from major primary
depression and a small number of schizoaffective and
schizophreniform disorders were reported. Since neither typical
schizophrenic and paranoid psychoses nor other types of depression
are reprinted in our present material, we will refain from
commenting at this time of the specificity issue. Neither shall we
deal with the treatment response (Langer et al, 1979) or the
state-trait (Brambilla et al, 1980) issues. It has though to be
noted that discrimination of the two types of depression (unipolar
and bipolar), was impossible to make by the tests we have used in
this investigation. Such a discriminatory value of DST was
previously suggested (Schlesser et al, 1980) but refuted by others
(Mendelewicz et al, 1982). In contrast with previous reports
(Schlesser et al, 1980; Coryell et al, 1982) and in agreement with
others (Hwu et al, 1981; Mendlewicz et al, 1982), no association
was found in our results between family history and abnormalities
in either the HPA or the HPT axis.

Suggestions by certain authors (Shulman and Diewold, 1977);

Holsboer et al, 1980), who found low sensitivity rates of DST in
major depression, that overdrive of the HPA system in depressives
may only relate to their symptomatology at the time of testing
reather than to their illness, by our present findings and are not
supported still unpublished data.

Our results deriving from a pool of patients subjected to

multiple neuroendocrinological testing indicate that in the same
patient, meeting all the clinical and research criteria of an
active primary depression, a variety of hormonal responses
associated with multiple endocrine systems may be observed. Even
within the same hormonal system, as the HPA axis, a differential
susceptibility of its components to provocative tests may be
manifested. The disassociation in cortisol and N-POMC
suppressibility to dexamethasone provides a substantial evidence.

In closing the discussion we may say that psychoneuroendo-

crinological research has grown so fast that has rapidly reached
the stage of maturation that calls for less enthusiasm, more
critical assessment of its own achievements and tighter
combination of imaginative design with factual interpetations.

REFERENCES

Bleuler, M., 1954, "Endokrinologische Psychiatrie", Georg Thieme,

Stuttgart.
Brambilla, F., Smeraldi, E., Bellodi, L., Sacchetti, E. and
Mueller, E., 1980, Neuroendocrine correlates and monoaminergic
hypothesis in primary affective disorder, in: "Progress in
Psychoneuroendocrinology", Brambilla, F., Racagni, G.,

286
 de Wied, D., ecs., Elsevier/North Holland Biomedical Press,
New York.
Carroll, B.J., 1982, The dexamethasone suppression test for
melancolia, Brit.J.Psychiat., 140:292.
Carroll, B.J., Schroeder, K., Mukhopadhyay, S., Greden, J.F.,
Fenberg, M., Ritche, J. and Tarika, J., 1980, Plasma
dexamethasone concentrations and cortisol suppression response
in patients with endogenous depression, j.CEM, 51:433
Carroll, B.J., Feinberg, M., Greden, J.F., et al., 1981,
A specific laboratory test for the diagnosis of melancholia.
Standardization validation and clinical utility,
Arch.Gen.Psychiat., 38:15
Chan, J.S.D., Seidah, N.G. and Chretien, M., 1983, Measurements
of N-POMC in human plasma, J.CEM., 56:791
Coppen, A., Rama Rao, V.A., Bishop, M., et al., 1980, Neuro-
endocrine studies in affective disorders, J. Aff. Dis., 2:311
Coryell, W., Gaffney, G. and Burkhart, P.E., 1982, The
dexamethasone suppression test and familial subtypes of
deppression- a naturalistic replication, 1982; Biol. Psych.,
17:33
Extein, I., Pottach, A.L.C., Gold, M.S., et al., 1982, Using the
protirelin test to distinguish mania from schizophrenia, Arch.
Gen, Psychiat., 39:77
Gold, M.S., Potash, A.L.C., Extein, I., et al., 1981, The TRH test
in the diagnosis of major and minor depression, Psychoneuro-
endocrinology, 6:159
Gold, M.S., Pottash, A.L.C., Ryan, N., Sweeney, D.R., Davis, R.K.
and Martin, D.M., 1980, TRH induced TSH response in unipolar,
bipolar and secondary depressions, Psychoneuroendocrinology,
5:147
Holsboer, F., Klein, H., Bender, W., et al., 1980, Hypothalamic-
pituitary-adrenal activity in a group of 100 heterogenic
patients: Diagnostic validity and biochemical aspects of the
cortisol response to dexamethasone suppression, in: "Progress
in Neuropsychopharmacology", A. Carlson, J. Engel, G.G.
Gottfries, et al., eds., Pergamon Press, New York.
Hwu, H., Rudorfer, M.V., and Clayton, P.J., 1981, Dexamethasone
suppression test and subtypes of depression, Arch. Gen.
Psychiat., 38:363.
Krieger, D.T., and Aschoff, J., 1979, Endocrine and other
biological rhythms, in: "Endocrinology", L. J. DeGroot, G.F.
Cahill, W.D. Odell, L. Martini, D. H. Nelson, J.T. Potts,
E. Steinberger and A. Winegrad, eds., Grune and Stratton, New
York
Langer, G., Schonbeck, G., Koinig, G., Lesch, 0., and Schussler,
M., 1979, Hyperactivity of hypothalamic-pituitary-adrenal axis
in endogenous depression, Lancet, ii: 524
Loosen, P.T., Prange, A.J. Jr., 1980, TRH: A useful tool for
psychoneuroendocrine investigation, Psychoneuroendocrinology,
5:63

287
Mendlewicz, J., Charles, G., and Franckson, J.M., 1982, The
dexamethansome suppression test in affective disorder:
Relationship to clinical and genetic subgroups,
Brit.J.Psvchiat., 141:464
Mendlewicz, J., Van Cauter, E., Linkowski, P., L'Herrnite, M.,
and Robyn, C., 1980, The 24-hour profile of prolactine in
depression, Life Sciences, 27:2015
Sachar, E.J., 1975, Neuroendocrine Abnormalities in depressive
illness, in: "Topics in Psychoendocrinology", E.J. Sachar,
ed., Grune and Stratton, New York
Schlesser, M.A., Winokur, G., and Sherman, B.M., 1980, Hypo-
thalamic-pituitary-adrenal axis activity in depressive
illness: its relationship to classification, Arch. Gen.
Psvchiat •• 37 737
Schulman, R., and Diewold, P., 1977, A two dose dexamethasone
suppression test in patients with psychiatric illness, Can.
Psvch. Ass. J., 22
Tosca, P., Fenoglio, L., Zerbi, F., Romani, A., Bezzi, G.,
Ferrari, E., Canepari, C., 1982, Neuroendocrinological spects
of depression and symptomatological picture, Psychiat. Clin.,
15:153

288
RO 15 1788- 11 C: A SPECIFIC RADIOLIGAND FOR AN "IN VIVO" STUDY
OF CENTRAL BENZODIAZEPINE RECEPTORS, BY POSITRON EMISSION
TOMOGRAPHY

M. Maziere, P. Hantraye*, B. Guibert*, M. Kaijima*, C.

Prenant, J. Sastre, M. Crouzel, R. Naquet*, D. Comar
Service Hospitalier F. Joliot, CEA Dpt. de Biologie, 91406
Orsay, France
*Lab. de Physiologie Nerveuse, CNRS, 91190 Gif-sur-Yvette,
France

Positron emission tomography (PET) is a visualization technique which

offers the feasibility to realize "in vivio" quantitative autoradiographic
measurements. This method yields, in a non invasive way, an image (which
can be quantified) of a previously administered radioligand in any transverse
section of the body.
So, for an "in vivo" investigation of BDZ receptors, the problem is to find a
suitable radioligand for an "in vivo" labelling of these finding sites i.e.: a
substance which reaches the binding sites very ~uickly after parenteral
administration and displays "in vivo" (hence at 37 C) a high affinity and
specificity for the receptors.
The radioligand must be labelled by a positron emitter, with a very high
specific activity in order that, at the level of the receptor sites, a
concentration of the order of magnitude of the dissociation constant
corresponds to a radioactivity high enough for PET quantitative
measurement.
RO 15 1788, an imidazobenzodiazepine derivative which exerts a selective
antagonist action on the central effects of BDZ by competitive, high
affinity intery<ftion with central type BDZ receptors (1 ), was isotopically
labelled with C (20.4 min half life) using a methylation process (2). An "in
vivo" study (3) of the binding of RO 15 1788- C, in the brain of the
immobilized baboon (10 kg weight) after I.V. administration of 10 mCi
(corresponding to 10 nanomoles) shows, in the cerebellar region known to be
ric~i:n BDZ receptors that: - the mjan brain uptake of the radioactivity is
5.2 % of the injected dose per em of brain tissue which according to the
specific ~tivity of the radioligand corresponds to a concentration of 5
pmo1e/cm • - In saturation experiments: when increasing doses of cold RO
15 1788 (0.05-5 mg) were coinjected with the labelled compound, the brain
uptake is reduced in a dose dependent mode, up to reach the blood
radioactivity level. - In displacement experiments: when increasing doses of
289
 _ SATURATION EXPERIHENTS-

i:?
5
,---- ...... ....
,,
11 C-Ro15-1788 CEREBRAL KINETICS • Change vith
coinjected incn~asing doses of Ro15-17881nHtkg l
:!. I ...........
"i
..,..
.....
~4 I
I
.... ....
0 I
.., I .... ....
.. .... ....
I
~ I

.
:E
~
3
... __
...•
!!
- __ Control
Q.
::0
2
"f"'
liD

Blood
oL--------.--------r--------.-------,r-- ------.-+
20 40 60 80 100 11in

Figure 1

290
cold RO 15 1788 were injected 20 minutes after the radloligand, a dose
related displacement of the cerebral radioactivity is observed. When using
for the displacement a BDZ agonist (Lorazepam 5 mg) or 2 stereolsomers (5
mg) R/ 11 6896 (active) or RO 11 6896 (inactive), a displacement of the
cerebral radioactivity is also observed but in the case of the inactive
isomer, this displacement is much lower.
In animals receiving increasing doses of cold RO 15 1788 for the
displacement an increasing amount of pentamethylene tetrazol is necessary
to induce generalized seizures. 11
Conclusion: Using RO 15 1788- C as an "in vivo" radlollgand, most of the
criteria necessary for the "in vivo" characterization of the BDZ receptors
(regional localisation - affinity - saturabllity - stereospecificity and
pharmacological response) were for the first time verified by PET in the
brain of a living baboon. This opens a new field in fundamental and clinical
research.
Acknowledgement: This study was partially supported by grants from CNRS
(ATP N~ 031927) and fondatlon pour la recherche medlcale.
The authors wish to thank Wyeth Byla (for lorazepam) and Professor W.
HAEFEL Y (Hoffmann - La Roche & Co) for kindly supplying the compounds
used in this study.

REFERENCES
1. Mohler, H. Burkard, W.P., Keller, H. H., Richards, .J.G., Haefely, W.
(1981): J.Neurochem.37:714-722

2. Mazlere, M., Godot, J.M., Berger, G., Prenant, C., Comar, D.

(1980): J.Radloanal.Chem.56:229-235

3. Mazlere, M., Prenant, C., Sastre, J., Crouzel, M., Co mar, D.,
Hantraye, P ., Kaljlma, M., Culbert, B., Naquet, R. (1983): C.R.
Acad.Sc. Paris, t.296 Serle III, 871-876

291
PET STUDIES DURING HYPNOSIS AND HYPNOTIC SUGGESTION

L. Baer, R.H. Ackerman, O.S. Surman, J.A. Correia,

J.L. Griffith, N.M. Alpert and T.P. Hackett
Massachusetts General Hospital and Harvard Medical
School, Boston, Mass., U.S.A.

Positron emission tomography (PET) is a sensitive measure of

brain function. It detects not only the effects of structured
stimuli, but also the effects of adventitious endogenous and exo-
genous events on regional blood flow and metabolism. As a result,
the "normal" PET study, both at rest and in response·to activation
procedures, may be more difficult to characterize than the abnormal
study (1).

The Cerebral Blood Flow Laboratory at the Massachusetts Gen-

eral Hospital (MGH) has embarked on a series of studies designed to
determine the optimum conditions for isolating stimuli of interest
in order to obtain reproducible baseline/activation studies. As-
suming the cerebral blood flow (CBF) and metabolism are coupled in
normals, blood flow is used as the index of brain function. CBF
is measured during continuous inhalation of c15 o2. The data col-
lection period for each baseline and activation trial is 24 minutes,
following an 8 minute equilibrium period.

A wide range of psychological events can intervene during this

30-40 minute period of equilibration and steady state imaging.
These effects include: anxiety, fatigue, inattention, habituation,
visual imagery, and silent speech. As a result, the PET image which
is obtained may be contaminated by these secondary psychological
phenomena, and may reflect activation of brain areas in response to
stimuli other than the test stimulus.

This report presents the results of pilot studies ongoing over

the past 2 years at the MGH by the Cerebral Bloodflow Laboratory
and Department of Psychiatry. These studies were designed to
evaluate whether with hypnosis one might be able to control the

293
patient's response to his PET study environment to selectively en-
hance or suppress his response to specific stimuli, and thereby
increase the "signal to noise ratio" of the PET image. Supporting
this hypothesis is current evidence that hypnosis is not a passive
activity, but rather involves heightened awareness, along with se-
lective attention and inattention to environmental events (2,3).

In addition, these studies attempted to determine the substrate

on which hypnosis and hypnotic suggestions act, because to date
there exist no good physiological markers of the hypnotic state
(4,5).

Subjects

Three white, male, right-handed subjects were examined. sub-

ject 1, age 26, was a normal volunteer with no neurological or psy-
chological problems. Subjects 2 and 3 were patients in a psychiatry
outpatient clinic who were included in the study because of their
excellent hypnotic ability, and their past clinical response to
hypnosis. Subject 2, age 37, suffered from diabetes mellitus and
chronic renal failure. He was taking 1 gram of Tegretol daily,
which eliminated anxiety episodes thought to be caused by partial
complex seizures. Subject 3, age 44, had a history of depression,
chronic back pain, irritable bowel syndrome, and recurrent plantar
warts. The former three problems were responsive to Triavil and
Imipramine. The warts were reportedly improved with hypnosis in
the past.

Procedure

Subjects were screened for high hypnotic susceptibility by

means of the Stanford Hypnotic Susceptibility Scale: Form C (6).
They then underwent training for at least 5 hours, with several
objectives: 1) to allow them to rapidly enter hypnosis during eye
fixation and counting by the hypnotist, so as to minimize visual
imagery during the induction; 2) to enable them to remain deeply
hypnotized with eyes open; 3) to permit them to rapidly experience
hypnotically-suggested blindness without associated anxiety; 4)
to familiarize them with the testing situation via "dry runs" of
the PET procedure.

In the experimental session, four consecutive PET trials were

performed for each subject during the following states: (1) awake
baseline; (2) "resting" hypnosis; (3) hypnotic suggestion for blind-
ness; (4) post-hypnotic waking state, except in Subject 3 (see
below).

Subjects were instructed to keep their eyes open throughout

the trials while looking at projected geometric shapes, which were
changed at 8 minute intervals to prevent habituation.

294
 Following the experimental session all subjects were debriefed
by a resident psychiatrist who was not present during training or
the PET studies. Subjects' self reports indicated that all had re-
mained deeply hypnotized throughout the hypnotic PET trials and
had experienced the suggested blindness.

Computer-generated images were produced at various cross-

sectional levels in a plane 10° to the orbital meatal line, and
these images were interpreted by Dr. Ackerman.

RESULTS

Subiect 1. At rest, Subject 1 showed no asymmetry in CBF

(Figure 1). Following hypnosis, there was a relative focus of
increased blood flow in the left fronto-temporal region as compar-
ed to both of his waking baseline scans (Figure 1). After the
blindness suggestion, no significant changes were noted in the
visual cortex areas.

Subiect 2. While at rest, Subject 2 showed a slight relative

left fronto-temporal focus of CBF activity. Following hypnosis,
CBF in this area was symmetric. No other changes in CBF were found
in this subject following hypnosis. Following the blindness sug-
gestion, no change in visual cortex activation was noted.

Subject 3. At rest, no asymmetries were noted in CBF for

Subject 3. During hypnosis there was bilateral activation of the
cerebellum and relative activation in the area of the left basal
ganglia. Following the suggestion for blindness, there was a re-
lative decrease in activity in both the left cerebellum and the
area of the left basal ganglia. No change was noted in Subject
3's visual cortex activity following the blindness suggestion.

Since this subject had, in the past, responded to hypnotic

suggestion for wart removal, it was decided to modify the fourth
experimental condition to investigate the CBF response to this
hypnotic suggestion. Thus, for the 4th trial, this subject re-
mained hypnotized and was asked to imagine the multiple warts on
his hands and feet tingling, in preparation for healing.

During this fourth PET trial, the count rate fell off due to
this subject's bodily movements, making the resulting image impos-
sible to interpret with confidence. There appeared, however, to be
some relative activation in the left fronto-temporal area during
this suggestion.

DISCUSSION

Although all of our subjects were excellent and experienced

295
Figure 1. Positron images of bloodflow in Subject 1. Increasing
brightness represents higher rates of physiologic activity. Trans-
verse-section images displayed as if looking from above, with left
hemisphere on observer's left and frontal pole located superiorly.
Left column contains lower level images. At rest no asymmetry was
present in CBF. Following hypnosis, there was a relative increase
in CBF in left fronto-temporal region (arrows).

296
hypnotic subjects, PET did not find among them a reliable CBF pat-
tern common to the hypnotic state. In addition, hypnosis did not
appear to reduce the inter-trial CBF variability of these subjects.
With regard to hypnotically-suggested blindness, although all 3
subjects reported experiencing the suggested blindness, no changes
in activation were found in the visual cortex. These findings raise
the possibility that hypnosis may not be a uniform state between
subjects, but rather that individuals may have unique patterns of
response to hypnosis.

The meaning of Subject l's relative increase in left fronto-

temporal activity following hypnosis is unclear. However, during
post-experimental inquiry, this subject stated that he had occa-
sionally engaged in silent counting to help him remain hypnotized.
This may have played a role in the left fronto-temporal activation.
This finding illustrates the importance of examining each subject's
ideosyncratic response to an experimental challenge. Other possible
contaminants also occurred in the other subjects: Subject 2 fre-
quently closed his eyes and drifted into.light sleep for short per-
iods during the experimental session, and Subject 3 experienced
severe back pain during the final PET trial.

The results from this series of three subjects must be con-

sidered preliminary. The subjects were not homogenous, and as noted
above, Subjects 2 and 3 had physical and emotional problems, and
periods of sleep and back pain were confounding variables for these
two subjects, respectively. Although it would have been desirable
to obtain arterial samples to permit quantification of the data,
arterial punctures cannot be routinely done in volunteer subjects.

In a related case we attempted to use hypnosis to minimize

the movements in a 63 year-old female patient with a history of
choreoathetosis of unknown etiology, Previously, a CT scan could
only be done after repeated Brevital injections, and it was desir-
able to avoid this method to obviate the effects of barbiturates
on CBF and metabolism.

During hypnotic suggestion to remain stationary this patient

was able to undergo two consecutive PET trials, which showed de-
creased function in the striatum. The cl5o 2 and 15o2 images, of
CBF and metabolism respectively, both showed bilateral activation
of the cerebellum; more in the cl5o 2 study when the patient's move-
ments were better controlled. One may wonder whether the relative
increase in CBF activity in this patient's cerebellum was related
to increased motor inhibitory activity.

297
CONCLUDING REMARKS

Our preliminary findings regarding PET and hypnosis illustrate

the variations in CBF response between subjects, even in those
highly trained with respect to the PET procedure. However, our
experience with hypnosis to control athetoid movements, as repor-
ed in our final case, suggests that more specific input by means of
hypnotic suggestion may help to reduce extraneous sources of vari-
ability. We plan to investigate this possibility in future studies.

+n summary, it appears that PET can detect changes in cerebral

activation during hypnosis and suggestion, but the significance of
these changes is unclear, and will require further study. Hypnosis
itself may be the manifestation of a heterogenous physiological
response that is subject-dependent, and this heterogeneity may be
present even in individuals who exhibit the same behavioral response
to hypnotic suggestions.

REFERENCES

1. Mazziota, J.C., Phelps, M.E., Carson, R.E., Kuhl, D.E. Tomo-

graphic mapping of human cerebral metabolism: sensory depri-
vation. Paper presented at the Annual meeting of the Ameri-
can Neurological Association, 1981.
2. Barabasz, A.F. and Lonsdale, C. Journal of Abnormal Psychology.
In press.
3. Hilgard, E.R. and Hilgard, J.R. "Hypnosis in the relief of pain"
Los Altos, Cali.£. William Kaufmann, 1975.
4. Tebecis, A.U., Provins, K.A., Farnback, R.W., and Pentony, P •
.Tnnrn::.l nf N~rvnns and Mental Disease, 1975, 161:1-17.
5. uiett, G.A., Ahpinar, S., !til, TM: Eiectroenceph. Clin.
Neurophysiolog~, 1972, 33:361-368.
6. Weitzenhoffer, A.M. and Hilgard, E.R. Stanford Hypnotic
Susceptibility Scale: Form C. Palo Alto. Consulting
Psychologists Press, 1962.

298
A STUDY OF EPILEPSY AND INTERicrAL PSYCHOSIS USING

THE OXYGEN-15 INHALATION TECHNIQUE AND POSITION EMISSION TOMJGRAPHY

B. Gallhofer, M. R. Trirrble, R. S. J. Frackowiak, R. J. S.

R.J.S. Wise, and T. Jones
National Hospital, Queen Square and the MRC cyclotron
unit, Hamuersmith Hospital, Du Cane Road, London UK

INTRODUcriON

There are now several studies of positron

emission tarography in epilepsy, and virtually all
have used fluoro-deoxyglucose as the preferred method
of investigation. They have mainly examined selected
patients undergoing evaluation for consideration of
temporal lobe surgery for intractable seizures.
Similarly, the published data using this technique in
sChizophrenia have also used glucose uptake to assess
metabolism. The Hanmersmi th group have developed the
technique using oxygen-15 and positron emission
tarography for the measurement of regional cerebral
blood flow (rCBF), and the regional oxygen extraction
fraction (rOER), whiCh permits the calculation of
regional oxygen metabolism ( rCMRO?.) , (Frackowiak,
Lenzi, Jones and Heather 1980). In ""this paper we
presentdata on a group of patients with epileptic
psychosis.
The literature on the relationship between
epilepsy and psyChosis has recently been reviewed in
detail (Trirrible 1982,), and it seems clear that a
growing number of authors accept that an increased
association between the two occurs. However, not all
patients with epilepsy suffer from psychosis, and the
issue has more recently been to clarify those patients
at most risk. There have been several authors who have
noted a link with temporal lobe epilepsy ( Hill 1953,

299
 Pond 1957, Slater and Beard 1963), confirmed for
certain forms of psychosis, especially those with a
sChizophreniform presentation, in more recent studies
using standardised and validated rating scale methods
for the evaluation of psyChopathology (Perez and
Trinble 1980.) The evidence in addition is that only a
small percentage of those with temporal lobe seizures
develop psyChosis, and again risk factors are becoming
defined (Stevens 1982). In particular patients with
carplex partial seizures, medial (linbic) temporal
lesions, and longstanding poorly controlled epilepsy
seem prone, although radiological data so far has not
clearly :inplicated structural pathology, identified
with the cr scan, as necessary for the psyChotic
evolution (Reider, Perez, and Trinble-in preparation).
We have therefore been interested to examine patients
with and without psyChosis using positron emission
tooography, noting differences between them.

MATERIAlS AND MIITHODS

The series consisted of two groups of psyChotic

patients, in addition to age rratChed non-psyChotic
epileptic patients (n=5), and age matChed non-
epileptic volunteer controls (n=5). The first group of
psyChotic patients were being treated with neuroleptic
rredication (n=6), and the second were free fran this
class of drug (n=6). The psyChosis was rated in all
the patients, except one, using the Present State
Exarninatia1 (PSE) of Wing (Wing et al 1974). All these
psyChotic patients had complex partial seizures,
secondary generalisation ocurring in sane of them.
Scans were taken inter-ictally. Three patients were
scaned on two occasions, once a1 a neuroleptic drug,
and a1ce drug free.
The scan proceedure has been fully outlined
elsewhere (Fracko.viak et al 1980), and will not be
presented here. In the rra jority of the patients and
controls, planes were scanned at CM+2, +4, +6, and +8
ems. All were corrected for attenuation by the
corresponding transmission scans, and tracer equations
WhiCh relate steady state measurements to tissue blood
flo.v,and oxygen extraction ratios were used to
calculate absolute quantitative values of rCBF, rOER,
and rCMR0 2 •
Follo.ving a camp~ter print out of the
quantitative data, 5. Ocm regions of interest were

300
chosen, corresponding to the frontal and occipital
regions en slices at CMt4 and CM+6 respectively. Three
areas 2were measured fran the tenporal cortex, each of
1.5a:n in a continuous strip en slice OM+4.
Additional areas examined included a "fronto-terrporal
bridge" (Firn), representing an island of cortical
tissue between the frontal and tenporal areas en slice
~· and the basal ganglia were assessed using a 2.6
an ROI area directly fran the visual display unit
en which scans were projected.
Statistical analysis of the data was py Student's
t test, one way ANJVA, and the post hoc means test.

RESULTS

In this psychosis study, the mean ages of the

groups were: control 37.8 years, epileptic non-
psychotic 34.0 years, epileptic psychotic untreated
38.2 years, and epileptic psychotic treated 42.3
years. The mean IQ of the three epileptic groups was
98. 0, 97.0, and 96. 5 respectively, and the mean
seizure frequency in the rronth prior to the scan 9. 2,
3. 9, and 1. 4. There were no statistical differences
between the groups for age, IQ, or seizure frequency.
The PSE ratings were NS (nuclear schizcphrenia) for 4
in the untreated group, and two in the treated group,
one of whan was scanned on two occasions. Other
psychotic classifications were mainly paranoid and
affective. The drug free psychotic patients were free
of neuroleptic medication for varying intervals. In
one patient this was nine days, in a second 24 days,
in two others over seven rronths, and two patients had
never received any such treatment.
When the cortical values for the frontal,
tenporal and occipital areas were examined, all three
epileptic sanples shONed consistently lONered values
canpared with non-epileptic volunteers, replicating
the findings of our first study. Again the maximal
differences were seen in the more posterior temporal
areas. When the psychotic non-neuroleptic treated and
the non-psychotic epileptic groups were canpared, the
former had lONer rCMR02 , higher rCBF, and lONer
rOER, the differences appearing in a constant pattern
on the majority of the areas • Significantly lONer
rOERs were recorded in the follONing regions: left
frontal, terrporal and Firn, and right frontal and
tenporal. When the psychotic non-treated, and the non-
pychotic epileptic patients were examined for

301
 laterality differences, lower values for the rCBF and
rCMR02 were seen en the left side accross the entire
tenporal cortex in the former, cpposite to the trend
seen for the ncn-psychotic patients, with the
exception of the anterior terrporal r:01R0 2 •
The treated psychotic patients generally showed
lower rCBF values ccmpared with the non-treated,
significant in frontal and anterior t.errporal areas.
This resulted in the rOER increasing in the treated
group.

DISCUSSION

The results of these studies of patients with

partial epilepsy using positron emission tooography
and the oxygen-15 inhalation technique demonstrate a
nunber of significant and interesting findings. With
regards to our first study of patients with catplex
partial seizures (Bernardi et a1 1983) the data nust
be viewed in the light of extensive results in
epilepsy gathered by others (eg Engel et al 1982
a, b, c) • In our studies patients showed bilateral
decrements of both the rCBF and the r:01R0 2 in
tenporal areas and on the side of the expected lesJ.cn
these Changes are of a wider distribution than would
be anticipated fran a relatively circumscibed lesion
in the anterior terrporal lobe. This is similar to the
local areas of hypanetebolism reported by others (eg
Engel et a1 1982a), suggesting that such changes are
seen in patients with carplex partial seizures
generally, and not cnly in those awaiting surgery, in
whan structural lesions may be :rrore prevalent. The
finding of law values in the contralateral
hemisphere, also in patients with a unilateral focus
raises the question of the extent of abnormalities in
the brains of these patients, and may have relevance
for the suggested association between epilepsy and
psychosis. Thus it is interesting that the Changes
recorded are mainly seen in the tenporal and frontal
areas, those regions of the brain that contain lirribic
system structures, suggesting more widespread
involvement of such structures in unselected patients
with partial seizures than may have been suspected.
The preliminary data fran the psychosis study
nust be interpreted with caution. The numbers of
patients examined is small, and the differences
recorded between the psychotic and non-psychotic
groups small. Nevertheless they do sugest a consistent

302
pattern of difference in values between the groups,
with greater changes for the CMR02 in the psychotic
patients. The rOER is lo.ver in selected regions in the
psychotic patients, perhaps reflecting real metabolic
differences.
The laterality differences noted, namely the
psychotic patients sho.ving lo.ver values on the left
side, especially in the temporal areas is of interest
in view of the gro.ving evidence of a link between left
temporal abnorm:tlities and sane forms of psychosis
(Trimble and Perez 1982), and the recent reporting of
positron emission tomography using fluoro-deoxyglucose
in schizophrenia in Which left temporal abnormalities
have been suggested (Di Lisi et al 1983). This effect
is unlikely to be due to an artefact of more patients
in one of the groups having a temporal lobe focus on
one side, as the groups were approximatly equal with
regard to the laterality of the focus Further
examination of these laterality differences are in
progress, especially noting links between type of
psychosis and the side of the lesion.
We feel that the results of our investigations, a
preliminary analysis of Which is given here, will be
of value not only for the further understanding of the
relationship between epilepsy and brain function, but
also, especially with further advances in scan
resolution, of the interesting but conplex
relationship between psychosis and epilepsy.

REFERENCES

Bernardi S, Trimble MR., Fracko.viak RJS, Wise RJS, and

Jones T, (1983), An interictal study of partial
epilepsy using positron emission torrography and the
oxygen-15 inhalation technique. J Neurol Neurosurg
Psychiat 46 473-477.

Di Lisi IE, Buchsbaum MS, Irving CA,IJo..lling S, Johnson

J, Holcomb HH, Kessler R, and Borono.v
J, (1983 ) , Clinical correlates of PEI' in schizophrenia.
New Research Abstracts. American Psychiatric
Association Annual Meeting May 1983 NR49.

Engel J,Kuhl DE, Phelps ME,and Mazziota JC, (1982),

Interictal Cerebral Glucose Metabolism in partial
epilepsy and its relation to Effi changes. Ann Neurol
12: 512-517.

Engel J, Brcwm WJ, Kuhl DE, Phelps ME, Mazziota JC,

303
 and Crandall PH,(l982), Pathological findings
underlying focal tenporal lobe hypanetabolism in
partial epilepsy. Ann Neural 12: 518-528

Engel J, Kuhl DE, Phelps ME, and Crandall PH, ( 1982),

Comparative localisation of epileptic foci in partial
epilepsy by Per and EEG. Ann Neural 12: 529-537.

FrackCMiak RS, Lenzi GL, Jones T, and Heather JD,

(1980), Quantitative measurement of regional cerebral
blood flCM and oxygen metabolism in man using 150 and
positron emission tarography. J Ccnput Assist Tooogr
4:727-736.

Hill D, (1953), Psychiatric disorders of epilepsy. Med

Press 229:473-475.

Perez MM, and Trirrible MR, ( 1980) , Epileptic Psychosis

- diagnostic camparison with process schizophrenia.
Brit J Psychiat 137:245-249.

Pond DA, (1957), Psychiatric aspects of epilepsy. J

Ind Med Prof 3:1441-1451.

Slater E, and Beard N!V, (1963), The schizophrenia-like

psychoses of epilepsy. Brit J psychiat 109:95-150.

Stevens JR, (1982), Risk factors for psychopathology

in individuals with epilepsy. In: Terrporal Lobe
Epilepsy, Mania, and Schizophrenia and the LiriDic
system. ed Koella WP and Trimble MR pp56-80. Basel
Karger.

Trirrible MR, (1982), The interictal psychoses of

epilepsy. In: Psychiatric Aspects of Neurologic
Disease. ed Benson DF and Blumer D pp 75-91 New York
Grune and Stratton.

Trirrible MR and Perez MM, (1982) The phenomenology of

the chronic psychoses of epilepsy. In: Terrporal lobe
epilepsy, Mania, and Schizophrenia and the Limbic
System. ed Koella WP and Trimble MR pp98-105 • Basel
Karger.

Wing JK, Cooper JE,and Sartorius N, (1974),The

description and classification of psychiatric
symptoms. London Cambridge University Press.

304
PET STUDIES ON BRAIN ENERGY METABOLISM AND DOPAMINE RECEPTORS IN

SCHIZOPHRENIC PATIENTS AND MONKEYS

G. Sedvall, G. Blomqvist, T. DePaulis, E. Ehrin, L.

Eriksson, L. Farde, T. Greitz, C.G. Hedstrom, D.H. Ingvar,
J.-E. Litton, J.L.G. Nilsson, S. Stone-Elander, L. Widen,
F.-A. Wiesel and G. Wik
Departments of Psychiatry and Psychology, Clinical Neuro-
Neurophysiology and Neuroradiology at the Karolinska
Institute , the Karolinska Pharmacy, Stockholm, and
Astra Lakemedel SBdertalje, Sweden

INTRODUCTION

Since about two years, the PET-scan group at the Karolinska in-
stitute in Stockholm has been engaged in analysing various aspects of
brain function in schizophrenic patients using the PET-scan technique.
Most studies performed so far have concerned the use of 11-C-labelled
glucose to study brain energy metabolism (Blomqvist et al., 1983).
The studies reported in a preliminary form in this communication had
the following objectives: (1) To compare regional brain net uptake of
11-C in schizophrenic patients and healthy volunteers after intrave-
nous injection of 11-C-glucose. (2) Examine variability of metabolic
data obtained from two measurements in healthy volunteers. (3) Exami-
ne the effect of a standardized neuroleptic treatment (sulpiride) on
regional glucose metabolism in schizophrenic patients. (4) Explore
the possibility to visualize brain dopamine receptors in monkeys
using 11-C-labelled substituted benzamides.
PET-scan Studies with 11-C-glucose in Healthy Volunteers and Schizo-
.Phrenic Patients
Demographic data for the subjects participating in the investiga-
tion are presented in Table 1. All the schizophrenic patients were
relatively young and all fulfilled the Research Diagnostic Criteria
(RDC) for schizophrenia (Spitzer et al., 1977). All patients had been
drug free for at least 3 weeks. Two were completely drug naive. A
standardized protocol was used (Fig. 1). The subjects were studied on
two occasions with an interval of about 5 weeks. This design was used
in order to evaluate (1) the reproducibility of the method in healthy

305
 Table 1. Demographic Data for Healthy and Schizophrenic
Subjects

Healthy Schizophrenic
Volunteers Patients
Men 10 8
Women 5
First episode 6
Second episode 3
Three or more episodes and chronic 4
Age (yea rs) 29 + 6.2 29 + 7.2
Age range 22 - 41 18 - 41

Mean + S.D.

volunteers and (2) the effects of a standardized neuroleptic treat-

ment in the schizophrenic patients.

Average net uptake of 11-C was determined during 25 min. after

an intravenous injection of 11-C-glucose. Regional uptake in defined
anatomical brain structures was determined according to the following
technique. Regions of interest were drawn on CAT-scan sections of the
brain from each subject (Fig. 2). These anatomical areas were fed
into the computer and the time averaged net uptake of 11-C from the
PET scan was determined and related to the surface of the total slice
area. Accordingly, metabolic data from a defined anatomical brain
region could be obtained (Bohm et al., 1983). We have so far calcu-
lated metabolic data from about 20 defined brain regions. The tech-

im.stigation I -Ngo6onJ:
PET Inc - Gluco>e l PET ( llC - ~o,.)

CT

~~y~k 0r'----------------------~
WEEKS

Schizophtenic
~ treotm.nt 800 mg / doy
0 WEEKS

Fig . 1. Protocol for the Fi g. 2. Regions of interest drawn

s t udy from CAT- s can s ections
sup er imposed on a PET- s can
picture obtained after IV-
injection of 11-C-glucose .

306
 -
1.2
n-9

I
15
l.o S2
r •0.96*** ....
• • n•17 ~
a..
z 0.1 • u 10

3
Ill ~
~ ~ 5
0.6 0
~ iii

~
0-6 0.1 l.o 1.2
0 2 s
11C IN BRAIN REGIONS, I WEEKS

Fig. 3. Correlation of uptake Fig. 4. Reduction of psychotic

data from two dif- morbidity during treat-
ferent measurements ment with sulpiride
(800 mg/day)

nique has the potential of allowing the estimation of metabolic 1n-

formation from about 100 regions in each human brain.

In order to study the reproducibility of the method used, time

averaged net uptake of 11-C from the two measurements were compared
in the healthy volunteers. As demonstrated in Fig. 3, there was a
very high correlation between uptake data obtained in a defined
region at the two measurements.

Metabolic data from the schizophrenic patients, were compared

with data from age matched healthy controls (Table 2). Thus, Brodmann
areas 32 sin. and 39+40 showed significantly lower average net uptake
of 11-C-glucose in the schizophrenic subjects. The nucleus lentifor-
mis on the left side also had a lower uptake. The data should indi-
cate that brain metabolism is lower in these regions in the schizo-
phrenic subjects. Brodmann areas 39+40 are considered to be involved
in integration of sensory information from several modalities. Our
data may indicate an alteration of such functions in young, untreated
schizophrenic subjects. We did not find any evidence of hypofrontali-
ty in these schizophrenic patients.

Treatment with the dopamine-2 receptor blocking antipsychotic

drug, sulpiride (800 mg/day), resulted in a significant reduction of
psychotic morbidity in the patients (Fig. 4). The metabolic data
demonstrated that drug treatment elevated 11-C-uptake in Brodmann
areas 39+40 and in the caudate nucleus. These findings indicate that
neuroleptic treatment counteracts the reduction of metabolic activity
in Brodmann areas 39+40 observed in the schizophrenic patients. More-
over glucose metabolism seems to be stimulated in the caudate nucleus,
an area known to contain a large number of dopamine receptors which
are potently blocked by neuroleptic drugs like sulpiride (Harnryd et
al., 1983a,b).

307
Table 2. Net Uptake of 11-C in 10 Healthy Subjects and 13 Drug Free
Schizophrenic Patients
Brain region Hol1th¥ subjects Pati.,ts

Brocl 32 dx 1.20 t 0.084 1.15t0.108

'sin 1.19 t 0.037 1.12 t 0.091.

Brocl 6+8 dx 1.14 t 0.044 1.14 t 0.094

sin 1.12 t 0.048 1.14 t 0.057

Brod 39+40 dx 1.14 t 0.043 1.08 t 0.067.

sin 1.13 t 0.041 1.08 t 0.063.

Brocl 6+8/39+40 dx 1.00 t 0.034 1.05 t 0.051•

sin 0.99 t 0.048 1.06 t 0.059**

Nuc1. coud. dx 0.98 t 0.067 0.98 t 0.086

sin 0.95 t 0.074 1.00 t 0.087

Nucl. 1.,t. dx 1.10 t 0.072 1.08 t 0.072

sin 1.14 t 0.043 1.07 t 0.052**

lleon tS.D.

* p < 0.05 •• p < 0.01

When data from the caudate nucleus were analysed with regard to
change between the two investigations, marked differences between
healthy volunteers and the schizophrenics were found. Thus, all the
volunteers but one had lower net uptake of 11-C in the caudate nuc-
leus at the second investigation. In the group of schizophrenics, 9
out of 10 had higher uptake at the second investigation. This finding
(Fig. 5) further demonstrates the significant effect of sulpiride
treatment on metabolic activity in the caudate nucleus, a brain
region which is in the centre of interest for research with regard to
schizophrenia and the mechanism of action of neuroleptic drugs.

PET-scan Imaging of Brain Dopamine Receptors Using 11-C-Labelled

Substituted Benzamides

Sulpiride is a neuroleptic drug belonging to the group of sub-

stituted benzamides (Justin-Besan~on et al., 1967). During recent
years, these compounds have attracted great interest since they in-
duce a number of effects on central dopaminergic mechanisms. A series
of substituted benzamides have been developed at the Astra Company,
Sodertalje, Sweden. Some of these have a very high affinity for
central dopamine receptors with a low dissociation constant (Hall,
Kohler & Ogren, personal communication). Recently it has been possib-
le to label one of these compounds with 11-C (Ehrin et al., unpub-
lished data). This compound was injected into nembutal anesthetized
cynomolgus monkeys and PET-scan images of the brain were produced at
various intervals after the injection. In Fig. 6 it can be seen (left
image) that shortly after the injection of the ligand, there is a
fairly strong non-specific uptake of the compound in several areas of

308
 1/0l~EIS SOt20I'ttiNCS
1.2 n -7 r .0.41 1.2 n - 10 • • o.a."
~ 1.1 ... 1.1
~ 1.
u"'
0
Oi 1.o
it
.::; 0 .9 ~0.9
~ 0 .8
0 , _...,._ _ _...,
Oa 0.9 1.o 1.1 1.2 0 0.1 0.9 lo 1.1 1.2
I NO OIUG
11C N CAUDA.T£

Fig. 5. 11-C-uptake in the Fig. 6. PET-scan picture of 11-C-

caudate nucleus in uptake in the brain from
healthy volunteers and a cynomolgus monkey after
schizophrenic patients. IV-injection of the sub-
Sulpiride-treated stituted benzamide FLB-
schizophrenic patients 542. Note specific uptake
demonstrate higher up- in the basal ganglia one
take during treatment hour after injection
(right projection). Left
projection was obtained
20 min. after injection

the brain. However, about 60 min. after the injection, the basal
ganglia represent the only part of the brain where uptake of 11-C
still persists. We have recently been able to displace this label
using systemic haloperidol treatment (Sedvall et al., unpublished
data). Interestingly enough, this ligand did not show any marked
binding to structures in the retina or the pituitary gland. Our
results indicate that this type of substituted benzamides may be
suitable for further studies aiming at labelling dopamine receptors
in the human brain. Such tools may be of value in neuropsychiatric
research for analysis of quantity and distribution of central dopamine
receptor populations in the brain of different patient categories.

Wagner et al. (1983) recently described the use of 11-C-labelled

methylspirone which binds to central dopamine and serotonin receptors
for PET imaging of dopamine receptors in the brain of baboons and a
human .subject.

CONCLUSIONS

Net uptake of 11-C-glucose in brain regions using PET scans in

healthy subjects has a good reproducibility.

Schizophrenic patients may have a reduced metabolic activity in

309
Table 2. Net Uptake of 11-C Healthy Subjects and 13 Drug Free
Schizophrenic Patients

Brain region Healthy subjects Patients

Brod 32 dx l. 20 + 0. 084 1.15 + 0.108

II II
sin 1.19 + 0.037 1.12 + 0.091*

Brod 6+8 dx 1 . 14 + 0. 044 l. 14 + 0.094

II II
sin 1.12 + 0. 048 1.14 + 0.057

Brod 39+40 dx 1.14 + 0.043 1.08 + 0.067*

-
II II
sin l. 13 + 0.041 1 . 08 + 0. 063*
-
Brod 6+8/39+40 dx 1.00 + 0.034 1 . 05 + 0 . 05 7*
II II
sin 0.99 + 0.048 1. 06 + 0.059**

Nuc1. caud. dx 0.98 + 0.067 0.98 + 0.086

II II
sin 0.95 + 0.074 1 .00 + 0. 087

Nuc1. 1ent. dx 1.10 + 0. 072 1.08 + 0.072

sin l. 14 + 0.043 l. 07 + 0.052**
-
II II

-
Mean + S.D.

* p < 0.05 ** p < 0. 01

310
Brodmann areas 32, 39+40, and nucleus lentiformis. Young, drug free
schizophrenics do not show any hypofrontality.

There is no consistent evidence of asymmetric metabolic activity

either in volunteers or schizophrenics.

Neuroleptic treatment (sulpiride) results in increased metabolic

acitivity in the caudate nucleus, Brodmann areas 39+40, and pons.

Brain dopamine receptors in monkeys can be selectively imaged

using PET-scan technique and substituted benzamides labelled with
11-C.

REFERENCES

Blomqvist, G., Bergstrom, K., Bergstrom, M., Ehrin, E., Eriksson, L.,
Gamelius, B., Lindberg, B, Lilja, A., Litton, J.-E., Lundqvist,
H., Malmborg, P., Mostrom, U., Nilsson, L., Stone-Elander, S.,
and Widen, L., in press 1983, Models for llC-glucose, in:
''Metabolism of the Human Brain Studied with PET," D. H. Ingvar,
T. Greitz, L. Widen, eds., Raven Press, New York.
Bohm, C., Berggren, B.-M., Greitz, T., Kingsley, D., and Olsson, L.,
in press 1983, A computerized individually varikble stereotaxic
brain atlas, in: ''Metabolism of the Human Brain Studied with
PET," D. H. Ingvar, T. Greitz, L. Widen, eds., Raven Press,
New York.
Harnryd, C., Bjerkenstedt, L., Bjork, K., Gullberg, B., Oxenstierna,
G., Sedvall, G., Wiesel, F.-A., Wik, G., and Aberg-Wistedt, A.,
in press 1983, Clinical evaluation of sulpiride in schizophrenic
patients - A double-blind comparison with chlorpromazine, Acta
Psychiatr. Scand., Suppl.
Harnryd, c., Bjerkenstedt, L., Gullberg, B., Oxenstierna, G., Sedvall,
G., and Wiesel, F.-A., in press 1983, Time course for effects of
sulpiride and chlorpromazine on monoamine metabolite and pro-
lactin levels in cerebrospinal fluid from schizophrenic patients,
Acta Psvchiatr. Scand., Suppl.
Justin-Besan~on, L., Thominet, M., Laville, C., and Margarit, J.,
1967, Constitution chimique et proprietetes biologiques du Sul-
piride, C. R. Acad. Sci. Paris, 265:1253.
Spitzer, R.L., Endicott, J., and Robins, E., 1977, Research Diag-
nostic Criteria (RDC) for a selected group of functional dis-
orders, 3rd ed, _Biometrics Research, New York.
Wagner, Jr., H.N., Burns, H.D., Dannals, R.F., Wong, D.F., Langstrom,
B., Duelfer, T., Frost, J.J., Ravert, H.T., Links, J.M., Rosen-
bloom, S., Dukas, S.E., Kramer, A.V., and Kuhar, M.J., in press
1983, Imaging of dopamine receptors in the human brain by posit-
ron tomography, in: "Metabolism of the Human Brain Studied with
PET," D. H. Ingvar, T. Greitz, L. Widen, eds., Raven Press, New
York.

311
DYNAMIC HEMISPHERIC IMBALANCE IN SCHIZOPHRENIA

John Gruzelier
Department of Psychiatry
Charing Cross Hospital Medical School

In a study of forty-eight unmedicated recent hospital

admissions diagnosed as schizophrenic with the help of the Present
State Examination (PSE) (Wing, Cooper and Sartorious, 1974) lateral
asymmetries in electrodermal orienting responses were found in either
direction (Gruzelier &Manchanda, 1982). Classification of patients
according to the nature of the asymmetry and comparison of their
symptoms and signs on the PSE and Brief Psychiatric Rating Scale
(BPRS) (Overall &Gorham, 1962) delineated two syndromes, one
representing positive symptoms and the other negative symptoms
(seeGruzelier 'Lateral imbalance and positive and negative symptoms•
in this collection). From consideration of the electrodermal lateral
asymmetry, reciprocal interactions between the hemispheres in the
control of arousal, and the character of the behaviour described in
the syndromes, e.g. pressure of speech in one and restricted quantity
of speech in the other, it was inferred that the positive syndrome
reflected hemispheric imbalance in the direction of overactivation of
the left hemisphere and the negative syndrome overactivation of the
right hemisphere.
The validity of the model was first examined through a review
of lateralised dysfunction in schizophrenia (Gruzelier, 1983a).
Evidence was sought for (1) asymmetries in both directions, (2)
opposite asymmetries coinciding with different clinical syndromes,
(3) those syndromes havinq an affinity with the ones delineated by
the electrodermal asymmetry and (4) consistence in the direction of
hemispheric imbalance across measures. Support was forthcoming from
a wide range of measures which are shown in Table I. These include
levels of electrocortical activity as seen in the electroencephalogram
(EEG); cortical evoked potentials to sensory stimuli; the direction
of lateral eye movements to questions desiqned to stimulate the
313
(,.)
_.
~
Table I: Measures on which schizophrenic subgroups differed in lateral asymmetries.
Measure Class ifi cation Hemispheric Authors
Activity
EEG Hallucinating Paranoid/Catatonia. Left/Right Stevens & Livermore (1982)
Dominant/Non-Dominant. Left/Right Coger & Serafetinides (1983)
Paranoid/Residual. Left/Right Etevenon et al (1983)
Cortical Retarded symptoms Right Rodin et al (1968)
Evoked Hallucinatory, florid, reactive Left Connolly et al (1983)
Potentials symptoms.
Lateral Paranoids more extreme (trend) Left Gur (1978)
eye movements Non-paranoids and Depressives. Right Sandel &Alcorn (1980)
Hoffman Good prognosis-reactive/Poor Left/Right Good et al (1981)
Reflex prognosis-withdraw, retarded.
Dichotic Hallucinating/Non-halluc inating Left/Right Alpert et al (1975)
Listening Paranoid/Non-paranoid Left/Right Lerner et al (1977)
Paranoid-electrodermally reactive/ Left/Right Gruzelier & Hammong (1980)
Non-paranoid electrodermally
unreactive.
Positive Symptoms. Left Wexler & Heninger (1980)
Auditory Hallucinating/Non-halluc inating Left/Right Bahzin et al (1975)
Thresholds
Somatosensory Young/Old Right?/Left? Scarone et al (1983)
Extinction
Handedness Positive/Negative Left/Rgiht Andreasen (1982)
hemispheres independently; the recovery curve of the Hoffman reflex
to somatosensory stimulation; measures of auditory processing
including the shadowing of dichotically presented textual passages,
the recall of digits presented dichotically and thresholds to very
brief stimuli; a test of somatosensory extinction whereby competing
textures are presented to the palmar surfaces simulataneously; finally
the incidence of left handedness.
It is noteworthy that all but one of the variables measured
processes that are dynamic and reversible. Some vary with the
activation levels of the hemispheres:- electrodermal responsiveness,
electrocortical activity, the Hoffman reflex and lateral eye move-
ments, others with the direction of attention in the lateral plane:-
eye movements, auditory processing and perhaps somatosensory extinction.
Although dynamic asymmetries may coincide with hemipsheric cognitive
specialisation, which tends to be of a fixed nature and is presumably
dependent on brain structure, by placing attention and activation
in competition with cognitive asymmetries dynamic process asymmetries
may be differentiated. One example concerns lateral asymmetries in
dichotic listening with schizophrenic patients which have been
variously interpreted as reflecting activational imbalances (Nachson,
1980) or structural problems in information transmission {Walker et al,
1981). By manipulating activation through varying stimulus intensity
and by instructing the direction of attention to either ear these
have been shown to be reversible and to belong to the dynamic class
(see Gruzelier, 1983a, b). In contrast when measures were employed
which are thought to depend on cognitive specialisation of a relat-
ively fixed nature, such as the recognition of verbal or spatial
stimuli presented in the visual half-field, there is no evidence of
two syndromes but more commonly a unitary deficit of the left hemis-
phere.
The one exception to the list of variables measuring dynamic
process asymmetries in Table I is handedness. While this clearly
represents a structural asymmetry it would not be unreasonable to
posit an activation asymmetry in favour of the contralateral hemis-
phere as one of its consequences (see also Kinsbourne, 1980).
Another issue in the literature review was whether the opposite
asymmetries coincided with different clinical syndromes and whether
these had an affinity with those delineated by the electrodermal
measure. The answer was affirmative on both counts and evidence
for this is shown in the Table. Furthermore an additional and
critical issue for the model was whether the direction of the
hemispheric imbalance and its relation to the syndromes was consistent
across measures and accorded with the electrodermal results. Here
again the answer was affirmative, see Table I, with only one
exception, somatosensory extinction, a discussion of which is given
elsewhere {Gruzelier, 1983a).

315
 The negative syndrome depicted by the electrodermal pattern
of larger right hand response~which was interpreted as representing
right hemispheric activation,had a correspondence with the following
categories:- the subtypes of catatonia, residual, nonparanoid; symp-
toms described as depressive, nonhallucinating, negative, poor-
prognosis-withdrawn-retarded; and 'non-dominant' hemisphere. These
described negative symptoms or subtypes largely characterised by
negative symptoms.
The positive syndrome depicted by the electrodermal pattern of
larger left hand responses, in which left hemisphere overactivation
was inferred, coincided with the following categories:- the paranoid
subtype; hallucinatory, florid, good prognosis-reactive, and positive
clinical features; and 'dominant' hemisphere. These described
positive symptoms or the paranoid subtype which in its acute reactive
form approximates the syndrome delineated by the electrodermal var-
iable, though a note of caution is sounded in identifying this
syndrome with a paranoid diagnosis per se (Gruzelier, 1981).
The nature of the asymmetry in terms of hemisphere function
was consistent with the model. Left hemisphere overactivation
coupled with right hemisphere underactivation characterised the
positive syndrome and right hemisphere overactivation and left
hemisphere underactivation the negative syndrome. In the case of
the 'dominant'/'non-dominant' hemisphere distinction this was made
by sorting the BPRS ratings into two categories, though without
independent validation (Coger &Sarefetinides, 1983). While there
was good correspondence with the electrodermal delineation of BPRS
ratings pertaining to the positive syndrome and the 'dominant'
category this was less satisfactory with the negative syndrome and
the 'non-dominant' category. It may be relevant that the EEG
asymmetry was clearer in the 'dominant' than 'non-dominant' subgroup.
The second step in establishing the validity of the model was
to use the electrodermal asymmetry and clinical syndrome as the
independent variable and measures of electrocortical activity as
the dependent variable. In a preliminary study (Gruzelier, Jutai,
Connolly and Hirsch, 1984) on eleven unmedicated patients diagnosed ,
with the help of the PSE with six belonging to the postive syndrome
and five to the negative syndrome, an EEG spectral analysis was
obtained of 360 epochs following the presentation of flashes of
six different intensities raging between 0.31 and 10 ft lamberts
which subjects viewed passsively. Patients were compared with 10
normal volunteer subjects in activiy recorded from bilateral occip-
ital (0 1 and o2) and temporal (T 3 and T4 ) placements as well as
vertex (Cz).
Of relevance to the question at issue here, the two syndromes
did differ in showing lateral asymmetries in opposite directions
in EEG activity such that activation was higher on average on the

316
 01-2 08-22 Hz)
-1.60 L>R R>L

-1.80

L
-2.00
2
log 10 j..IV /Hz
-2.20
L
-2.40
,R
/
-2.60
1 2 3 4 5 6 1 2 3 4 5 6
High Low High Low
INTENSITY
Fig. I. Lateral asymmetries in occipital beta activity in
positive (L > R) and negative (R > L) syndromes.
left hemisphere for the patients with the positive syndrome and
higher on average on the right hemisphere for those with the
negative syndrome. Specifically these results referred to fast
wave activity (18-22Hz) over primary sensory cortex, see Fig. 1,
and slow wave activity (2-6 Hz) over temporal cortex. A replication
study is currently in progress.
Thus while there is growing evidence for a lateralised deficit
in schizophrenia of the left hemisphere {Gruzelier &Flor-Henry,
1979; Flor-Henry &Gruzelier, 1983) the consequences of this for
patterns of hemispheric activation appear to depend upon the
syndrome.

REFERENCES
Alpert, M., Rubenstein, H. and Kesselman, M., 1976, Asymmetry of
information processing in hallucinators and non-hallucinators.
J. Nerv. Dis. 162: 258-265.
Andreasen, N.C. (1982) Negative symptoms in schizophrenia. Arch.
Gen. Psychiat. 39: 784-794.
Bazhin, E.F., Wasserman, L.I. and TonKonogii, M., 1975,
Auditory hallucinations and left temporal lobe pathology.
Neuropsycholoqia. 13: 481-487.

317
Coger, H.W. and Serafetinides, E.A. 1983, in : "Laterality and
Psychopathology", P. Flor-Henry and J.H. Grizelier, eds.,
Elsevier Science Publishers, Amsterdam, 225-248.
Connolly, J.F., Gruzelier, J.H., Manchanda, R. and Hirsch, S.R.,
1983, Visual evoked potentials in schizophrenia: Intensity
. effects and hemispheric asymmetry., Brit. J. Psych., 142:
152-155. . -
Etevenon, P., Peron-Magnon, P., Campistron, D., Vordeaux, G. and
Deniker, P., 1983, Differences in EEG asymmetry between
patients with schizophrenia, in: Laterality and Psychopathology,
P. Flor-Henry and J.H. Gruzel1er, eds., Elsevier Science Pub-
lishers, Amsterdam, 269-290.
Flor-Henry, P. and Gruzelier, J.H., eds., 1983, Laterality and
Psychopath9logy, Elsevier Science Publishers, Amsterdam.
Goode, D.J., Manning, A.A. and Middleton, J.F., 1981, Cortical
laterality and asymmetry of the Hoffman reflex in psychiatric
patients, Biol. Psvchiatrv, 16: 1137-1157.
Gruzelier, J.H., l983a, ~ critfcal assessment and integration of
lateral asymmetries in schizophrenia, in: Hemisyndromes:
Psychobiology, Neurology, Psychiatry, M. Myslobodsky, ed.,
Academic Press, London, 265-326.
Gruzelier, J.H., 1983b, Conundrums in methods and theory in
laterality research, in: Laterality and Psychopathology,
P. Flor-Henry and J. Gruzelier, eds., Elsevier Science Publishers,
Amsterdam, 19-28. ,
Gruzelier, J.H. and Flor-Henry, P., eds., 1979, Hemisphere
Asymmetries of Function and Psychopathology, Elsevier, Amsterdam.
Gruzelier, J.H., Jutai, J.W., Connolly, J.F. and Hirsch, S.R.,
1984, Cerebral asymmetries in unmedicated schizophrenic
patients in EEG spectra and their relation to clinical and
autonomic parameters, Advances in Biological Psychiatry,
S., Karger, Basel.
Gruzelier, J.H. and Manchanda, R., 1982, The syndrome of schizophrenia:
Relations between electrodermal response lateral asymmetries and
clinical ratings, Brit. J. Psychiat., 141: 488-495.
Gur, R.E., 1978, Left hemisphere dysfunction and left hemisphere
overactivation in schizophrenia, J. Abnorm. Psychol., 87:
226-238.
Kinsbourne, M., 1980, A model of the ontogeny of cerebral organisation
in non-right-handers, in: Neuropsychology of Left-Handedness,

318
 J. Herron, ed., Academic Press, New York.
Nachson, G., 1980, Hemispheric dysfunctions in schizophrenia.
J. Nerv. Ment. Dis., 168: 241-242.
Overan, J.L and Gorham, D.R., 1962, Psychol. Rep., 10: 799-812.
Rodin, E., Grisell, J. and Gottliev, J., 1968, Some electrographic
subjects, in: Recent Advances in Biological Psychiatry,
J. Wortis, ed., Plenum Press, New York.
Sandel, A. and Alcorn, J.D., 1980, Individual hemisphericity and
maladaptive behaviours, J. Abnorm. Psychol., 89: 514-517.
Walker, E., Hoppes, E. and Emory, T., T9En, A reinterpretation of
findings on hemispheric dysfunction in schizophrenia, J. Nerv.
Ment. Dis., 169: 378-380. · -
Wing, u.K., Looker, J.E. and Sartorius, N., 1974, in: The Measurement
and Classification of Psychiatric Symptoms, Cambridge Univ.
Press, London and New York.

319
EVENT-RELATED POTENTIALS AND COGNITIVE DEFICITS IN SCHIZOPHRENIA

Tomomichi Kameyama, Osamu Saitoh, Ken-Ichi Hiramatsu,

Shin-Ichi Niwa, Karen Rymar and Kenji Itoh
Department of Neuropsychiatry, Faculty of Medicine
University of Tokyo
7-3-1 Hongo, Bunkyo-ku, Tokyo, 113 Japan

INTRODUCTION

Measurements of Event-Related Potentials (ERPs) have been uti-

lized in studies on attentional functioning. Hink, Hillyard and
Benson (1978) recorded ERPs during syllable discrimination tasks,
and found that the N100 component in normal subjects was enhanced to
all stimuli in the attended ear, while the P300 component was en-
hanced only to the "target" stimuli in that ear. Based on these re-
sults, they hypothesized that the N100 component correlated with the
'stimulus set', while the P300 component correlated to the 'response
set' as defined by Broadbent (1971). Accordingly, ERP measurements
are considered to be a useful method for clarifying the pathophysio-
logical bases underlying those diseases that display disturbed at-
tentional functioning. A number of studies measuring ERPs in schiz-
ophrenics has already been reported (Levit et al., 1973; Shagass et
al., 1978; Roth et al., 1981; Saitoh et al., in press). These re-
ports are nearly consistent in that schizophrenics demonstrate a re-
duction in P300 amplitudes.

The purpose of this study is to investigate schizophrenic defi-

cits in attentional functioning in terms of ERPs during syllable
discrimination tasks that require subjects to activate two sets of
selective attention.

There have been several controversies concerning which of the

two, that is "state" or "trait", in schizophrenic disorders is re-
flected in a reduction of the P300 component. We investigated ERPs
in siblings of schizophrenic probands in order to gain more insight
into this issue.

321
 In this report, results for the siblings as well as schizophre-
nics concerning ERPs derived from the T3 and T4 regions during syl-
lable discrimination tssks, will be described, in comparison to
those of normal controls.

~ffiTHODS

Subjects
Schizophrenic subjects consisted of 12 medicated patients
(6 males and 6 females; mean age, 29.8) and 10 non-medicated pa-
tients (5 males and 5 females; mean age, 28.2). These are the same
subjects as reported in a previous study (Hiramatsu et al., in press).
All schizophrenics met the diagnostic criteria of DSM-III for schiz-
ophrenic disorders. All non-medicated patients had not received any
drugs for at least 4 weeks prior to the experiments. One out of the
12 medicated patients and 7 out of the 10 non-medicated patients
were in an active phase at the time of the experiments. Twenty sib-
lings of schizophrenic probands (10 males and 10 females; mean age,
28.5) were also subjected to this study. Twenty volunteers with no
family history of psychiatric disorders (10 males and 10 females,
mean age, 29.1) served as a normal control group. All the siblings
and normal controls had no history of psychiatric or neurological
diseases. All subjects were free from any hearing disability.
There was no difference among the four category groups concerning
male/female ratios as well as ages.

Auditory stimuli
In syllable discrimination tasks, four CV-syllables (/ba/,
/da/, /ga/, /za/) of a male voice were presented to one ear, with
the same four CV-syllables of a female voice being presented to the
other ear through headphones monaurally. Each of the eight stimuli
was presented randomly with an equal a priori probability of 0.125.
The duration of the stimuli was 150 msec, with an intensity of ap-
proximately 60 dBSL. Interstimulus intervals ranged between 800-
1000 msec.

Procedure
The subjects were seated in an anechoic room with eyes closed.
They were required to count silently the number of a particular syl-
lable in a given ear as a 'target' in each run. All subjects per-
formed 16 runs, that is, four target syllables X two voices (male or
female) X two attended channels (left ear or right ear). The order
of 16 runs was randomized for all subjects, with the number of the
target syllables for each run being set between 18-26. After the
conclusion of each run, subjects were asked to give the total number
of the target stimuli detected.

EEG recordings and data analysis

EEGs were recorded at the Cz, T3 and T4 regions monopolarly,
utilizing linked earlobe electrodes as references. After the ex-

322
periment, EEGs without artifacts were averaged. The averaging epoch
began 20 msec before stimulus onset and lasted for 620 msec there-
after. Each of the ERP components was labeled as follows: N100 as
the most negative peak in the 56-156 msec period after stimulus on-
set, P200 as the most positive peak in the 40-120 msec period after
the N100 peak, the late positive component (LPC: including the P300
component) as the averaged amplitude of ERP in the 50-330 msec period
after the P200 peak. Each amplitude of the ERP components was mea-
sured with respective to a 20 msec pre- and post-stimulus baseline
as zero level.

RESULTS

1) The amplitudes of "T3-N100" and "T4-N100"

Judging from the results of ANOVA, the main effect of the factor
"CHANNEL(Attend/Non-attend)" (abbreviated as channel effect) on the
amplitude of "T4-N100" was significant in siblings, as well as normal
controls (normals: F(1,143)=7.63, p<0.006; siblings: F(1,143)=11.13,
p<0.001). However, this channel effect was not significant in either
medicated or non-medicated schizophrenics. The channel effect on
the amplitude of "T3-N100" was not significant in normal controls,
however, the amplitude was larger when the stimuli were presented
to the attended ear, as compared to the non-attended ear (attend:
1.91~V; non-attend: 1.75~V). On the other hand, the amplitude of
"T3-N100" elicited by the stimuli presented to the attended ear was
almost equal to that of the non-attended ear in medicated patients
(attend: 1.61~v; non-attend: 1.59~V), and the amplitude of "T3-N100"
elicited by the stimuli presented to the attended ear was smaller
than that of the non-attended ear in non-medicated patients (attend:
1.38~V; non-attend: 1.51~V). The channel effect on the amplitude of
"T3-N100" was significant in siblings (F(1,143)=8.85, p<0.003)(see
Fig. 1).

IJV
2.0

1.0
o5
ATTEND NON-ATTEND ATTEND NON-ATTEND
Fig. 1. Mean amplitudes of the N100 components at the T3 and T4 re-
gions. "ATTEND" and "NON-ATTEND" means the conditions when
the stimuli were presented to the attended ear and non-at-
tended ear. o- - -o normal(N=20x4), ....___. sibling(N=20x4),
!).-~non-medicated ( N=10x4) , A-.---& medicated ( N=12x4) .

323
 I"'SEC
120 T3 T4

100
----:::::._-a---e
~--
-.4
--t:.

80 t:.- --- -0.

o3
LEFT RIGHT LEFT RIGHT
Fig. 2. Mean latencies of the NlOO components at the T3 and T4 re-
gions elicited by the stimuli presented to the attended ear.
"LEFT" and "RIGHT" means the conditions when the stimuli
were presented to the left ear and right ear.o-- -onormal
(N=20x2), e---e sibling(N=20x2), b - .......6. non-medicated(N=10x2),
~-~medicated(N=12x2).

2) The latencies of "T3-N100" and "T4-N100"

ANOVA revealed that the latency of "T3-N100" elicited by the
stimuli presented to the attended ear for the non-medicated patients
was significantly shorter as compared to that of normal controls
(F(1,103)=18.64, p<O.OOl). However, for the latency of "T4-N100",
there was no significant difference between normal controls and the
non-medicated group. There were no significant differences in the
latencies of "T3-N100" as well as "T4-N100" among the medicated
schizophrenics, siblings, and normal cotrols (see Fig. 2).

pV
1.0

0.5 '

-0.5
~I~~~~ No~!l~~~ET NO~A~~l~ND ~8~=~I~~~~
Fig. 3. Mean LPCs at the T3 region. ATTEND TARGET, Target stimuli
in the attended ear; ATTEND NON-TARGET, Non-target stimuli
in the attended ear; NON-ATTEND TARGET, Same syllables as
target stimuli in the non-attended ear; NON-ATTEND NON-
TARGET, Syllables different from target stimuli in the non-
attended ear. o- - -0 normal(N=20x2), ~ sibling(N=20x2),
6.- -6.non-medicated(N=10x2), .A-· -A medicated(N=12x2).

324
 3) The "T3-LPCs" and "T4-LPCs"
In normal controls, the main effect of the factor "SYLLABLE(Tar-
get/Non-target)" on the "T3-LPCs" elicited by the stimuli presented
to the attended ear was significant (F(1,71)=10.60, p<0.002). This
effect means that "T3-LPCs" elicited by the target syllables pre-
sented to the attended ear were significantly larger as compared to
those of the non-target syllables presented to the attended ear.
This main effect of SYLLABLE was not significant in either medicated
or non-medicated schizophrenics, nor in siblings of schizophrenic
probands (see Fig. 3). As for the "T4-LPCs", the main effect of
SYLLABLE was not observed in any of the four groups.

DISCUSSION

Based on the results of their experiment employing syllable dis-

crimination tasks, Hink et al. (1978) hypothesized that the 'stimulus
set', as defined by Broadbent, was reflected in the N100 component,
with the 'response set' being reflected in the P300 component. In
this experiment, normal controls demonstrated an increase in ampli-
tudes of the "T3-N100" and "T4-N100" components due to channel-se-
lective attention, although in the case of "T3-N100", this increase
did not reach statistical significance; moreover, normal controls
exhibited an augmentation of "T3-LPCs" upon detection of targets.
These results support the hypothesis of Hink et al •. Mesulam (1981)
reported that right-hemisphere mechanisms appear more effective in
the execution of attentional tasks. Results of this study concerning
the asymmetry in changes of the amplitudes of "T3-N100" and "T4-N100"
may be related to such a hemispheric differentiation in attentional
mechanisms. Results of this study concerning the asymmetry in
changes of the "T3-LPCs" and "T4-LPCs" suggest a possibility that
the left-hemisphere dominance in verbal tasks is observed in the
stage of 'response set'.

Both medicated and non-medicated schizophrenics failed to de-

monstrate an increase in amplitudes of the "T3-N100" and "T4-N100"
components by channel selection. These results indicate that schiz-
ophrenics are thought to have disturbances in channel-selective at-
tention, that is, schizophrenics are not able to attend to one chan-
nel effectively, nor sustain this attention. According to
Broadbent's model of information processing, schizophrenics demon-
strate disturbances in the 'stimulus set'.

Both medicated and non-medicated schizophrenics failed to dem-

onstrate an augmentation of "T3-LPCs" upon detection of target stim-
uli. Judging from these results, it is suggested that schizophren-
ics demonstrate disturbances in stimulus evaluation, that is, schiz-
ophrenics are thought to have disturbances in the 'response set' as
defined by Broadbent.

The non-medicated group of schizophrenics demonstrated shorter

325
latencies of the N100 component exclusively at the T3 region, as
compared to those of normal controls. While, the medicated group
of schizophrenics displayed nearly equal latencies of the N100 com-
ponent to that of normal controls even at the T3 region. Seven out
of 10 non-medicated schizophrenics were in an active phase at the
time of the experiment. Therefore, it is speculated that schizo-
phrenics demonstrate shorter latencies of ERPs at the T3 region dur-
ing their active phase. Hence, hemispheric dysfunction, specifi-
cally of the left hemisphere in schizophrenics, may be a phenomenon
which is closely related to positive psychotic symptoms.

The siblings in this study demonstrated an increase in ampli-

tudes of the "T3-N100" and "T4-N100" components due to channel-se-
lective attention. This result was similar to that of normal con-
trols, but different from those of both schizophrenic groups. On
the other hand, the siblings, similar to both schizophrenic groups,
failed to demonstrate an augmentation of "T3-LPCs" upon detection
of target stimuli. This abnormality of LPC in siblings may reflect
a genetic predisposition to schizophrenia.

REFERENCES

Broadbent, D. E., 1971, Decision and Stress, Academic Press, New

York.
Hink, R. F., Hillyard, S. A., and Benson, P. J., 1978, Event-related
brain potentials and selective attention to acoustic and pho-
netic cues, Biological Psychology, 6:1.
Hiramatsu, K., Kameyama, T., Niwa, S., Saitoh, 0., Rymar, K., and
Itoh, K., in press, Schizophrenic deficits in information proc-
essing as reflected in ERP abnormalities during syllable dis-
crimination tasks, in Advances in Biolo~ical Psvchiatrv,
Mendlewicz, J. and Van Praag, H. M. (eds), Karger, Basel.
Levit, A. L., Sutton, S., and Zubin, J., 1973, Evoked potential cor-
relates of information processing in psychiatric patients,
Psychological Medicine, 3:487.
Mesulam, M-M., 1981, A cortical network for directed attention and
unilateral neglect, Annals of Neurolo~, 10:309.
Roth, W. T., Pfefferbaum, A. F., Berger, P. A., and Kopel!, B.S.,
1981, Auditory event-related potentials in schizophrenia and
depression, ~svchiatrv Researeh, 4:199.
Saitoh, 0., Hiramatsu, K., Niwa, S., Kameyama, T., and Itoh, K.,
in press, Abnormal ERP findings in schizophrenics with special
regards to dichotic detection tasks, in T.ateralitv and Psycho-
pathology, Flor-Henry, P. and Gruzelier, J. (eds), Elsevier,
Amsterdam.
Shagass, C., Ornitz, E. M., Sutton, S., and Tueting, P., 1978,
Event-related potentials and psychopathology, in Event-related
Brain Potentials in Man, Callaway, E., Tueting, P., and Koslow,
S. H. (eds), Academic Press, New York.

326
BPJ\IN SEROTONIN, MOOD-REGULATION AND

AGGRESSION-REGULATION

H.M. van Praag

Department of Psychiatry, Albert Einstein College of

Medicine/Mo ntefiore Medical Center
1300 Morris Park Avenue, Bronx, NY 10461 USA

1. PROBLEM DEFINITION

Rather consistently , the CSF post-probene cid concentration of

5-hydroxyind oleacetic acid (5-HIAA) - the main degradation
product of 5-hydroxytryp tamine (serotonin, 5-HT) - has been found
to be lowered in a subgroup of patients suffering from the
syndrome of vital (endogenous) depression (Review: van Praag,
1982). The data on baseline 5-HIAA values in CSF are more con-
troversial (Asberg et al., 1976; Koslow et al., 1983). Several
groups have reported that the "subgroup" I referred to above was
characterized by strong suicidal tendencies (Banki et al., 1981;
Montgomery and Montgomery, 1982; van Praag, 1982; Banki and Arato,
1983). Low CSF 5-HIAA was in particular found in depressed pati-
ents who had attempted suicide with violent means (Asberg et al.,
1976; Traskman et al., 1982). Subsequently low CSF 5-HIAA was
likewise observed in what was called non-psychoti c and non-
depressed suicide attempters (Traskman et al., 1982). This
prompted the hypothesis that disorders in central 5-HT are not so
much related to depression as to suicidality irrespective of the
nosological context in which this phenomenon occurred.

In this paper, I will investigate two questions: 1) Are

depression and suicide sufficiently independent to justify the
conclusion that the alleged disturbances in central 5-HT relate
to (violent) autoaggressio n rather than to depression? 2) What
strategies could furnish additional information pertinent to the
question of mood- or aggression-r elatedness of low CSF 5-HIAA?

327
2. RELATEDNESS OF DEPRESSION, SUICIDE AND TYPE OF SUICIDE ATTEMPT

2.1 Relation depression and suicide

First, we explored the validity of the diagnosis "non-psycho-
tic, non-depressed suicide attempter", a key-concept in the
hypothesis of suicide-relatedness of low CSF 5-HIAA.

We studied the mood condition in a group of 100 consecutive

suicide attempters admitted to a psychiatric ward in the two
months prior to the attempt (van Praag, 1982). The diagnosis
of depression was considered to be likely if in that period the
patient evidenced mood-lowering continuously or so frequently
that work performance, lust for life and sleep quality suffered
considerably and he had felt in need of professional help.
According to those criteria, the diagnosis of depression prior
to suicide attempt was appropriate in 82% of the sample.

The pre-suicide state of mood correlated rather poorly with

the state of mood in the first week after admission. Pre-admis-
sion ratings tended to exceed post-admission ratings (van Praag,
unpublished).

From this study, we draw conclusions of three types: 1) In

hospitalized suicidal patients, the overlap between depressive
and suicidal behavior is considerable. This confirms previous
observations (e.g. Weissman et al., 1973); 2) The post-suicidal
mood-condition is not representative for the one prior to the
suicide; 3) The diagnosis "non-psychotic, non-depressed suicide
attempter" can therefore only be based on data pertaining to the
pre-suicidal mental condition. In the studies mentioned under 1
this diagnosis was based on post-suicidal data. Henceforth, the
conclusion that low CSF 5-HIAA relates to suicidality rather than
to depression is not warranted.

True, our sample was a select one. Only a subsample of non-

psychotic suicide attempters are hospitalized. Therefore, our
conclusion regarding the depression/suicide overlap should not be
generalized. For the sake of the present discussion, however,
this notion is irrelevant. Both the Scandinavian work and our
study pertained to hospitalized suicidal patients.

2.2 "Specificity" of suicide attempt with violent means

Low CSF 5-HIAA tends to accumulate in those depressed patients
who have attempted suicide by violent means. Does that mean that
this biological variable relates to disordered aggression-regula-
tion rather than to disordered mood-regulation? Not necessarily
so. If violent suicide tended to occur preferentially in a part-
icular type of depression, it would be impossible to decide
whether low CSF 5-HIAA relates to self-destructive tendencies
per se or rather to that particular type of depression. We

328
 A = Non-psychiatric control group
B = Schizophrenic without suicidal
history
C = Schizophrenic with suicidal history
0 = Non-violent suicide
• = Violent suicide

0
0
0
0
00 8
---o
8 --=
0 •
0
0 00
•
oe
_.o..__

••
0 0
0
oe

Fig. 1. Post-probenecid CSF 5-HIAA in non-depressed schizophrenic

patients with and without suicidal histories and in non-
psychiatrically disturbed controls. Mean CSF 5-HIAA in
group c is lower than in the other 2 groups (p <0.05,
Fisher's exact test). Violent/non-violent classifica-
tion showed no correlation with CSF 5-HIAA (van Praag,
1983).

329
studied the question whether violent suicide attempt is or is not
"depression-specific". We have focused ~n particular on two
aspects of depression: syndrome and severity (van Praag and
Plutchik, 1984).

We compared two groups of patients. One consisted of 31

patients hospitalized in connection with non-psychotic depression
and vi~lent suicide attempt. The other consisted of 31 depressed
patients admitted in the same period in connection with non-
psychotic depression and non-violent suicide attempt. These
patients were matched for age and sex with the former group.
Syndromally, patients were diagnosed as: vital depression,
personal depression and mixed forms. Syndromal diagnoses were
made using a structured standard interview (van Praag et al.,
1965). The closest DSM III diagnoses of vital and personal
depression are Major Depressive Episode with Melancholia and
Dysthymic Disorder respectively. Severity of the depression was
rated on a global 5-point scale. Syndromal diagnoses and sever-
ity ratings were made by a physician who was unaware of the
objectives of this study.

Suicide with violent means appeared to occur preferentially

in the group of patients with the syndrome of vital depression
(Table 1). Severity of the depression was similar in the group
of vital and non-vital depression. The severity of the depres-
sion in the violent group was not greater than in the non-violent
group.

We concluded that the risk of violent suicide attempts is

greater in vital depression than in other syndromal depression
types and that the severity factor is unrelated to the type of
suicide. Since violent suicide and vital depression are inter-
related, it is impossible to conclude whether low CSF 5-HIAA
found in violent suicide attempters relates to disturbed regula-
tion of aggression or to that particular depression type.

3. IS LOW CSF 5-HIAA RELATED TO DISTURBED MOOD-REGULATION OR TO

DISTURBED AGGRESSION-REGULATION

3.1 Research strategy

As discussed under 2.," the available data do not allow for a
decision whether low CSF 5-HIAA relates to disturbances in mood
or in aggression. To resolve this problem, we formulated a
research strategy, made up of the following questions. If low
CSF 5-HIAA were related to aggression rather than to depression,
one could, we reasoned, formulate the following assumptions:
1 In comparison with normal CSF 5-HIAA depressives, low CSF
5-HIAA depressives will manifest not only a preponderance of
suicide attempts but in addition, an excess of signs indicating

330
Table 1. Syndromal depression diagnosis in 31 depressed patients
who had committed violent suicide attempts and in 31
depressed patients who had committed non-violent suicide
attempts. Vital depressions were found significantly
more frequently in violent suicide attempters (chi
square= 11.68, p < 0.001). (van Praag and Plutchik,
1984).

DIAGNOSIS

Number Vital Other

Depression Depression
Types
Violent
suicide attempt 31 26 5

Non-violent
suicide attempt 31 13 18

331
disregulated outward directed aggression.
2 Low CSF 5-HIAA will be found in other forms of auto-aggres-
sive behavior than suicide in depression, in particular in
a. Non-depressed, non-psychotic individuals
b. Non-depressed, psychotic individuals
c. Self-mutilation
3 Low CSF 5-HIAA could be expected to occur in personality
disorders in which disturbed aggression regulation is manifested
in outward directed aggression.

3.2Disturbances in aggression regulation in low versus

normal CSF 5-HIAA depressives
We compared two groups, each consisting of 25 patients mani-
festing the full vital depressive syndrome (van Praag et al.,
1965; van Praag, 1978). The two groups were matched for age and
sex. In one group post-probenecid CSF 5-HIAA was low (two
standard deviations or more below the normal mean), in the other
group it was normal. The average number of previous admissions
for depression numbered 2.4 and 2.7. In both groups, the major-
ity of patients was unipolar (19 in the former, 17 in the latter
group).

By means of a semi-structured interview, emotion-laden events,

including hostility in interpersonal functioning which occurred
in the preceding 6 months, were recorded. In addition, the
occurrence of previous suicide attempts was explored. A suicide
attempt was defined as any intentionally self-inflicted injury
(including ingestion) with self-destructive intent. Only those
attempts were counted that had led to medical intervention.
Besides the patient, at least one "significant other" was inter-
viewed. Finally, signs of hostility in interpersonal functioning
during the interview were recorded.

We were able to confirm previous observations indicating pre-

ponderance of suicide attempts in the low CSF 5-HIAA depressives
compared to their normal 5-HIAA counterparts. In addition, the
former group surpassed the latter in all signs of outwarddirected
aggression that were studied (Table 2). Weissman et al. (1973)
reported comparable findings. Comparing a group of suicide
attempters with a group of depressed patients, they found incre-
ased hostility in the suicide group, to be the single most
important distinguishing characteristic between the two groups.
Rydin et al. (1982) reported that low CSF 5-HIAA depressives
showed on Rorschach testing more hostility than comparable patients
with normal 5-HIAA.

In conclusion, in depressives with low post-probenecid CSF

5-HIAA not only s~icide but likewise outward directed aggression
is more prominent than in normal 5-HIAA depressives.

332
Table 2. Low 5-HIAA depressives compared to normal 5-HIAA
depressives present:

1. More suicide attempts p <0.01

2. Greater number of contacts

with police p <0.05
3. Increased arguments with

- relatives p <0.05

- spouse p <0.01

- colleagues p <0.05

- friends p <0.05

4. More hostility
at interview p <0.05

5. Impaired employment
history (arguments) p <0.05

333
 3.3 CSF 5-HIAA in auto-aggressive behavior other than suicide
in depression

Suicide in non-depressed, non-psychotic individ-

3.3.1
uals
As discussed under 2~ this diagnosis should be based on
careful exploration of the pre-suicidal mood condition. To the
best of my knowledge, no studies have yet been published that
meet this condition.

3.3.2 Suicide in non-depressed, psychotic individuals

In schizophrenia, suicide attempts may occur on the
basis of imperative acoustic hallucinations, without signs of
affective disorder. We measured post-probenecid 5-HIAA in 10 of
those patients and compared them with 10 psychopathologically
similar patients without suicidal history and with a non-psych-
iatric control group. The probenecid test was done in the second
week after admission, when the patients had been medication-free
for at least 10 days.

Mean post-probenecid CSF 5-HIAA was lower in the suici-

dal than in the other two groups (Fig. 1), although in 7 patients
the value fell within the normal range. The violent/non-violent
classification showed no correlation with CSF 5-HIAA.

These data support the hypothesis that low CSF 5-HIAA

relates to suicidal behavior irrespective of diagnostic category.

3.3.3 CSF 5-HIAA and self-mutilation

This issue has only been studied in the context of the
Lesch-Nyhan syndrome, and at that only in few patients. The
Lesch-Nyhan syndrome is an inborn error of (purine) metabolism
due to a defect of the enzyme hypoxanthine guanine phosphori-
bosyl transferase, with the abnormal gene on the X chromosome.
The disorder is characterized by self-mutilation (biting after
eruption of teeth), mental and growth retardation, choreoathetosis
and hyperuricemia ~Lesch and Nyhan, 1964; Castells et al., 1979).

Castells et al. (1979) reported decreased post-proberecid

CSF 5-HIAA in a boy suffering from this syndrome. Administration
of 5-hydroxytryptophan (5-HTP) in combination with a peripheral
decarboxylase inhibitor increased CSF 5-HIAA and suppressed self-
mutilation, albeit only temporarily. Mizuno and Yugari (1975)
treated 4 patients with 5-HTP, successfully in terms of self-mut-
ilation. Anders et al. (1978) however, reported unresponsiveness
of one patient whom they had treated with 5-HTP and a peripheral
decarboxylase inhibitor.
The data on central 5-HT in the Lesch-Nyhan syndrome are
scanty and inconclusive. Further studies are obviously needed.

334
 Table 3. Studies in which- CSF 5-HIAA and agression were,negatively correlated.
STUDY No DIAGNOSES MEASUREMENT ASSAY POST-PROBENECID OTHER }!A
Sex AGGRESSION CSF 5-HIAA OR BASELINE 5-HIAA METABOLITES IN CSF
Brown et al. 26~ Personality -Checklist Fluorometrical1y Baseline HVA unchanged
( 1979) disorder lifetime history MHPG increased
aggressive acts

Brown et al. 12~ Borderline -Checklist Mass-fragrnento- Baseline HVA uncl\anged

( 1982) personality lifetime history graphy MHPG unchanged
aggressive acts
-Buss-Durkee
Inventory
-HHPI
Bioulac et al. 6~ xyy Person- -Lifetime history Fluorometrically Post-probenecid HVA unchanged
(1980) ality disorder of aggressive MHPG not measured
behavior

Linnoila et al. 36~ Severe perso- -21 had killed; Liquid chroma- Baseline HVA unchanged
(1983)* nality disor- 15 had made tography HHPG unchanged
ders. attempts to kill.
All were alcohol
abusers.

* Low CSF 5-HIAA in impulsive violent offenders as opposed to those who had premeditated their acts.

w
w
U1
 3.4CSF 5-HIAA and outward directed a22ression
If low CSF 5-HIAA were related to auto-aggression, it is
reasonable to assume that similar disturbances would occur in
aberrant outward directed aggression. In other words, it would
be reasonable to assume that disturbed central 5-HT would relate
to disturbed aggression regulation irrespective of the direction
the aggression takes.

So far four studies have been published relevant to

this assumption (Table 3). All of them pertain to non-depressed
individuals with severe personality disorder who tend to act out
in an aggressive and sometimes delinquent way. In three of them,
a negative correlation was found between CSF 5-HIAA and scores
for lifetime history of aggression (Brown et al., 1979, 1982;
Bioulac et al., 1980). In a group of criminals that had committed
or attempted murder, l~w CSF 5-HIAA was found in those who had
acted impulsively vis a vis those who had somehow premeditated
the crime (Linnoila et al., 1983). This suggests that low
impulse control rather than aggressivity could be the psychologi-
cal correlate of low CSF 5-HIAA. Only Bioulac et al. (1980)
tried 5-HT increasing compounds, i.e. 5-HTP, and reported decrea-
sed aggressivity to the same degree as had been observed with
traditional neuroleptics. Another 5-HT precursor, tryptophan,
has been found to decrease aggressivity in schizophrenics
(Morand, 1983).

Those data, though certainly not conclusive, suggest related-

ness of low CSF 5-HIAA and disordered aggression regulation.

4. FUTURE RESEARCH INTO THE BIOLOGY OF HUMAN AGGRESSION

The biology of outward directed human aggression is difficult

to study. It can be induced and influenced by numerous external
factors and lacks continuity. Aggressive acts are generally
short-lived in contrast to other deviant behaviors - e.g. depres-
sion, psychosis - that tend to last over time. Non-aggressive
control groups - sufficiently matched - are not easy to compose.
Future research should take into account the type and frequency
of aggressive acts and the prevailing circumstances under which
they occurred should be well defined and documented. The possi-
ble role of alcohol and drugs should be taken into account both
as aggression precipitants and in their possible role in inducing
biological changes.

336
 Biological suicide research is likewise in need of greater
methodological finesse. To define violent suicide attempt on the
basis of method alone, disregarding impulse strength and charac-
ter is of a crudity that does not behoove sophisticated psychia-
tric research. That statement is valid even if one recognizes
the complexities of retrospective impulse measurement. Moreover,
to date biological concomitants of suicide have been studied in-
discriminantly in patients who have recently attempted suicide
(time lapse between attempt and biological studies being seldomly
if ever stated), in those with a lifetime history of suicide
attempts and in patients with suicidal ideation. Those groups
should be differentiated. There is no prior reason to assume
them to be equivalent.

Stressing the need for methodological refinement detracts in

no way from the importance of biological aggression research as
an essential complement of psychological and sociological studies.

5. SUMMARY AND CONCLUSION

Low CSF 5-HIAA has been found in certain types of depression

and in violent suicide attempters. It has been suggested that
low CSF 5-HIAA is related to suicidality rather than to disordered
mood regulation. The data on which this tentative conclusion is
based, however, do not warrant that conclusion. This is the
theme of the first part of this paper. In the second part, a
strategy is outlined to collect relevant data and the results so
far are discussed. Those data support the tentative conclusion
that low CSF 5-HIAA does relate to suicidality; and in general
more possibly to disordered aggression regulation.

Is low CSF 5-HIAA in depression completely explained by dis-

turbed aggression regulation? I consider this to be unlikely
since 1) low CSF 5-HIAA does occur in depressed patients without
suicidal anticedents and 2) the lowest 5-HIAA values have been
found in patients who were both depressed and showed an urge to
violent suicide.

The most reasonable hypothesis seems to be that central 5-HT

systems relate to both mood and aggression disorder. From a
biological vantage point one could expect, as a consequence, that
disturbances in both emotional systems will often occur hand in
hand. Psychopathologically, this is precisely a relationship
that is so often observed in clinical practice.

337
REFERENCES

Agren, H. Symptom patterns in unipolar and bipolar depression

correlating with monoamine metabolites in the cerebrospinal
fluid. Psychiat Res. 3: 225-236, 1980

Agren, H. Life at risk:markers of suicidality in depression.

Psychiat Dev 1: 87-104, 1983

Anders, Th.F., Cann, H.M., Ciaranello, R.D., Barchas, J.D.,

Berger, Ph.A. Further observations on the use of 5-hydroxy-
tryptophan in a child with Lesch-Nyhan Syndrome. Neuropadi-
atrie 9:2, 157, 1978

Asberg, M., Traskman, L., Thoren, P. 5-HIAA in the cerebrospinal

fluid:a biochemical suicide predictor? Arch Gen Psychiatry
33: 1193-1197, 1976

Banki, C.M., Molnar, G., Vojnik, M. Cerebrospinal fluid amine

metabolites, tryptophan and clinical parameters in depression:
Psychopathological symptoms. J Affect Disord 3: 91-99, 1981

Banki, C.M., Arato, M. Amine metabolites and neuroendocrine

responses related to depression and suicide. J Affect Disord
5: 223-232, 1983

Bioulac, B., Benezich, M., Renaud, B., Noel, B., Roche, D.

Serotonergic functions in the 47, XYY syndrome. Biol Psychi-
atry 15: 917-923, 1980

Brown, G.L., Ebert, M.E., Goyer, P.F., Jimerson, D.C., Klein, W.,
Bunney, W.E., Goodwin, F.K. Aggression, suicide, and seroton-
in:relationships to CSF amine metabolites. Am J Psychiatry
139: 741-746, 1982

Brown, G.L., Goodwin, F.K., Ballenger, J.C., Goyer, P.F., Major,

L.F. Aggression in humans correlates with cerebrospinal
fluid metabolites. Psychiat Res 1: 131-139, 1979

Castells, S., Chakrabarti, C., Windsberg, B.G., Hurwic, M.,

Perel, J.M., Nyhan, W.L. Effects of £-5-hydroxytryptophan on
monoamine and amino acids turnover in the Lesch-Nyhan Syndrome.
J of Autism Dev Disord 9:1, 95, 1979

Koslow, S.H., Maas, J.W., Bowden, C.L., Davis, J.M., Hanin, I.,
Javaid, J. CSF and urinary biogenic amines and metabolites
in depression and mania. Arch Gen Psychiatry 40: 999-1010,
1983

Lesch, M., Nyhan, W.L. A familiar disorder of uric acid metabol-

ism and central nervous sytem function. Amer J Med 36: 561-
570, 1964

338
Linnoila, M., Karoum, F., Potter, W.Z. Effects of antidepress-
ant treatments on dopamine in depressed patients. Arch Gen
Psychiatry 40: 1015-1017, 1983

Linnoila, M. Virkunen, M., Scheinin, M., Nuutila, A., Rimon, R.,

Goodwin, F.K. Low cerebrospinal fluid 5-hydroxyindoleacetic
acid concentration differentiates impulsive from nonimpulsive
violent behavior. Life Sci 33: 2609-2614, 1983

Mizuno, T., Yugari, Y. Prophylactic effect of ~-5-hydroxytrypto­

phan on self-mutilation in the Lesch-Nyhan Syndrome. Neuro-
padiatrie 6:1, 13, 1975

Montgomery, S.A., Montgomery, D. Pharmacological prevention

of suicidal behavior. J Affect Disord 4: 291-298, 1982

Rydin, E., Schalling, D., Asberg, M. Rorschach ratings in de-

pressed and suicidal patients with low levels of 5-hydroxyind-
oleacetic acid in cerebrospinal fluid. Psychiat Res 7: 229-
243, 1982

Traskman, L., Asberg, M., Bertilsson, L., Sjostrand, L. Mono-

amine metabolites in CSF and suicidal behavior. Arch Gen
Psychiatry 38: 631-636, 1982

Weissman M., Fox, K., Klerman, J.L. Hostility and depression

associated with suicide attempts. Am J Psychiatry 130: 450-
454, 1973

van Praag, H.M. Psychotropic drugs. A guide for the practition-

er. Brunner/Mazel, New York, 1978

van Praag, H.M. Neurotransmitters and CNS disease:Depression.

Lancet ii: 1259-1264, 1982

van Praag, H.M. Depression, suicide, and the metabolites of

serotonin in the brain. J Affect Disord 4: 275-290, 1982

van Praag, H.M. CSF 5-HIAA and suicide in non-depressed schizo-

phrenics. Lancet II: 977-978, 1983

van Praag, H.M., Plutchik, R. Depression-type and depression-

severity in relation to risk of violent suicide attempt.
Psychiat Res (In press) 1984

van Praag, H.M., Uleman, A.M., Spitz, J.C. The vital syndrome
interview. A structured standard interview for the recogni-
tion and registration of the vital depressive symptom complex.
Psychiat Res Neurol Neurochir 68: 329-346, 1965

339
NEW TRENDS IN THE PHARMACOLOGY OF

CENTRAL SEROTONINERGIC RECEPTORS

H. Gozlan, M. Hall, S. Bourgoin, S. El Mestikawy,

L. Pichat*, and M. Hamon
INSERM Ull4, College de France, F-75005 Paris, and*
Service des Molecules Marquees, CEA, F-91191 Gif-sur-
Yvette, France

INTRODUCTION

Since the discovery of serotonin (5-HT) in brain about thirty

years ago, considerable progress has been made regarding the distri-
bution of serotoninergic neurones, their electrophysiolog ical and
biochemical characteristics and their functional role in the control
of sleep and wakefulness, thermoregulation , pituitary hormone release,
transmission of nociceptive messages (see I) etc. In addition, abnor-
mal serotoninergic neurotransmissio n seems to be associated with
some psychiatric disorders (notably depression) showing indirectly
that 5-HT may well be involved in the control of basic behavioural
reactions.

Thanks to the extraordinary development of neuropharmacolog y,

drugs are now available for affecting specifically any biochemical
step dealing with the synthesis, inactivation, release of 5-HT in the
CNS. Such drugs are of great interest for exploring not only the re-
gulatory mechanisms of central 5-HT neurones but also the possible
therapeutic treatment of disorders related to alterations in 5-HT
neurotransmissio n (sleep disturbances, aberrant pain sensitivity,
anxiety, affective disorders etc.). The present symposium illustrates
perfectly the two goals of drug design in the 5-HT field: indalpine
is a very potent and selective 5-HT uptake blocker which can be used
in fundamental research to facilitate 5-HT neurotransmissio n in ani-
mals; in man, this central effect of indalpine is probably responsi-
ble for its therapeutic value in the treatment of affective disorders
and chronic pain.

341
THE PRESENT STATUS OF CENTRAL 5-HT RECEPTORS

Thanks to simple in vitro binding tests, Peroutka et al.(2)

reached the conclusion that at least two class3s of 5-HT receptors
exist in the CNS: the 5-HT 1 class which binds H-5-HT and related
agonists with an affinity 1n the nanomolar range but is poorly reco-
gnized by 5-HT an§agonists, and the 5-HT~ class which §an be selecti-
~ely labelled by H3antagonists such as H-spiperone, H-mianserin,
H-metergoline and H-ketanserin, and exhibits much higher affinity
for antagonists than for agonists including 5-HT itself. In addition
to binding studies, other biochemical approaches have been used for
exploring central 5-HT receptors: Nelson et al.(3) measured the 5-HT-
sensitive adenylate cyclase, Quach et al.(4) investigated the charac-
teristics of 5-HT-dependent glycogenolysis in brain slices, and
Dolphin and Greengard (5) studied the 5-HT-dependent phosphorylation
of a specific neuronal protein. Since the pharmacological properties
of the receptors involved in the former two 5-HT-dependent responses
are clearly distinct from those of the 5-HT 1 and 5-HT 2 specific bin-
ding sites, it can be concluded that four separate classes of 5-HT
receptors likely exist in the rat brain (6).

Classically, 5-HT agonists by stimulating directly 5-HT recep-

tors should trigger or facilitate 5-HT neurotransmission in brain.
However, the administration to rats of minute doses of LSD, in the
same range as those required for stimulating 5-HT receptors, induces
behavioural responses normally associated with a reduction in central
5-HT neurotransmission (such as those evoked by a blockade of 5-HT
synthesis with p.chlorophenylalanine or by raphe lesion). This led
to the concept of presynaptic autoreceptors, the stimulation of which
(by LSD notably) triggering a decreased release and turnover of the
neurotransmitter (7~. Up to now, evidence for the existence of such
receptors was however indirect: first, Aghajanian et al. (see 8)
reported that intra-raphe injection of minute amounts of LSD, 5-HT
and other indoles, produces a dramatic decrease in nerve impulse flow
within 5-HT neurones; they could thus define the pharmacological
profile of a putative 5-HT autoreceptor located on cell bodies and/
or dendrites of 5-~l neurones. Second, several groups (7,9,10) men-
tioned that the Ca -dependent 5-HT release from isolated terminals
can be markedly reduced by 5-HT and related agonists and that some
antagonists such as methiothepin, metergoline and quipazine, counte-
ract this effect. Such data led to the proposal that 5-HT terminals
also possess specific 5-HT autoreceptors, their stimulation trigge-
ring a negative feed· back process reducing the efficacy of action
potentials to release 5-HT from nerve endings. However, neither the
autoreceptor located on cell bodies and/or dendrites nor that on
nerve terminals has yet been visualized directly so that their exis-
tence was the matter of debate even in the recent past. In the
present report, we describe the binding characteristics of a new
5-HT agonist, 8-0H-N,N-dipropyl-2-aminotetralin (PAT), which provide
in fact the first direct demonstration of the existence of presynap-

342
tic 5-HT autoreceptors at least on nerve terminals. The possible
therapeutic value of drugs acting selectively on these autoreceptors
is discussed briefly in the last section.

THE INTRODUCTION OF PAT IN NEUROPHARMACOLOGY

In 1981, Arvidsson et al.(11) described the synthesis and some

pharmacological properties of PAT and concluded that this compound
is a potent and selective 5-HT agonist in the CNS. Further investi-
gations by this group indicated that PAT stimulates male sexual be-
haviour (12) and facilitates the acoustic startle response (13) in
rats. Since both effects are observed following treatments which
block central 5-HT neurotransmission, they postulated that PAT inter-
acts likely with presynaptic autoreceptors. I2 our laborat~ry, we
explored the possible effect of PAT on theCa +-dependent H-5-HT
release from cortical and striatal slices and found that this mole-
cule reproduces the inhibitory effect of 5-HT (14). Furthermore, the
down regulation of 5-HT release by PAT could be prevented by a selec-
tive 5-HT antagonist, metergoline, but not by yohimbine and sulpiride
which block preferentially alpha 2-adrenergic and dopaminergic recep-
tors respectively. All these observations were compatible with a
preferential (if not exclusive) interaction of PAT with presynaptic
autoreceptors.

CHARACTERISTICS OF 3H-PAT BINDING TO RAT BRAIN MEMBRANES

Since PAT apparently acts as an agonist on presynaptic 5-HT

autoreceptors, we tried to label directly these receptor sites with
its tritiated derivative (3H-PAT) synthesized with a high specific
~adioactivity: 105 Ci/nnnoL The incubation of brain membranes with
H-PAT allows the identification of specific binding sites (Kd=
3-9 nM) from which the labelled ligand can be displaced by an excess
(10 pM) of "cold" 5-HT or PAT (15). Analyses of the subcellular dis-
tribution of 3H-PAT binding sites revealed a postsynaptic location
in the hippocampus but a presynaptic labelling in the striatum. In
agreement with this conclusion, we observed that the selective dege-
neration of 5-HT innervation in the whole forebrain of adult rats
treated with 5,7-dihydroxytryptamine (5,7-HT) was associated with a
significant reduction (-60%) in the number of 3H-PAT binding sites
in the striatum (fig.l) but not in the hippocampus (15). In contrast,
the degeneration of 5-HT-target cells induced by the local injection
of kainic acid resulted in the disappearance of half the 3H-PAT
binding sites in the hippocampus but did not affect significantly
3H-PAT binding to striatal membranes (15). Such findings are compa-
tible with a preferential location of 3H-PAT binding sites on affe-
rent 5-HT fibres and terminals in the striatum and on intrinsic
~ostsynaptic neurones in the hippocampus. Differential locations of
H-PAT binding sites were also found in other regions: in the spinal
cord and the brain stem, they seem located presynaptically on 5-HT
structures. In contrast, the cerebral cortex, septum etc., apparen-

343
 2
o Control
• 5,7-HT

N STRIATUM
0
......
0
~
LL

'
lXI
0

0~ 0.2
B(pmol./mg prot.)

Fig.!. Scatchard plot of 3H-PAT binding to membranes from

the striatum of control or 5,7-HT-treated rats.
5,7-HT (8 ~g) was administered intracerebrally one month
before sacrifice. The selective degeneration of 5-HT inner-
vation is associated with a marked reduction (-60%) 3 in the
number (Bmax, intercept with the abscissa axis) of H-PAT
binding sires in the striatum of 5,7-HT-treated rats. The
Kd (slope ) remains equal to 6-9 nM in both groups of ani-
mals.

tly contain mainly postsynaptic 3H-PAT binding sites (15).

3 Among neurotransmitters, 5-HT is far the most potent to displace
H-PAT bound to either hippocampal or striatal membranes showing that
both post- and pre-synaptic 3H-PAT binding sites must be considered
as being serotoninergic (15). Extensive investigations indicated
however that the pharmacological properties of hippocampal and stria-
tal 3H-PAT binding sites are different. In the hippocampus, the order
of relative efficacy of 5-HT agonists and antagonist~ to displace
bound 3H-PAT corres~onds exactly to that found with H-5-HT as a
ligand; therefore, H-PAT binds probably to the same site as 3H-5-HT,
i.e. the postsynaptic 5-HTJ receptor (15). In the striatum, the
pharmacological properties of 3H-PAT binding sites are very close to

344
those of postulated autoreceptors controlling 5-HT release: among
antagonists, those preventing the negative influence of 5-HT on its
own release, i.e. methiothepin, metergoline and quipazine, are also
the most potent against 3H-PAT binding to striatal membranes.

CONCLUSION - POSSIBLE THERAPEUTIC APPLICATIONS OF DRUGS ACTING ON

PRESYNAPTIC AUTORECEPTORS

In contrast to 5-HT agonists acting on postsynaptic receptors

and therefore promoting synaptic 5-HT neurotransmission, those sti-
mulating presynaptic autoreceptors trigger a negative feed back con-
trol of 5-HT release leading to a decreased 5-HT neurotransmission.
Such an affect may well explain why the sexual activity of male rats
is markedly enhanced following the administration of PAT (12). Another
behavioural field in which PAT might be of some therapeutic value is
anxiety. Indeed, convergent observations in animals demonstrate that
treatments which reduce central 5-HT neurotransmission are anxiolytic
(16). Although studies with PAT have yet to be done in this respect,
it must be emphasized already that buspirone, a non-benzodiazepine
with well proven anxiolytic properties (17), is very active on 3H-PAT
binding sites (15).

Since PAT and other agonists acting on presynaptic autoreceptors

reduce 5-HT neurotransmission (at least in some brain regions),
antagonists acting on these receptors should in contrast facilitate
5-HT release and improve 5-HT neurotransmission. The latter drugs
would be of potential therapeutic value for the treatment of some
affective disorders associated with a deficiency in 5-HT neurotrans-
mission.

In conclusion, the synthesis of PAT has allowed the direct de-

monstration of the existence of presynaptic 5-HT autoreceptors in
several brain regions (striatum, brain stem etc.) and spinal cord,
and opens a new era for the pharmacology of central 5-HT synapses.
We can bet already that drugs similar or antagonist to PAT will be
of considerable interest for exploring further the functioning of
central 5-HT neurones and developing possibly better therapies of
selected psychiatric diseases.

REFERENCES

I. A. Galas, "The serotoninergic neuron", Masson, Paris (1981).

2. S.J. Peroutka, R.M. Lebovitz, and S.H. Snyder, Two distinct
central serotonin receptors with different physiological func-
tions, Science, 212:827 (1981).
3. D.L. Nelson, A. Herbet, A. Enjalbert, J. Bockaert, and M. Hamon,
Serotonin-sensitive adenylate cyclase and (3H) serotonin binding
sites in the CNS of the rat. I. Kinetic parameters and pharma-
cological properties, Biochem. Pharmacol. 29:2445 (1980).

345
 4. T.T. Quach, C. Rose, A.M. Duchemin, and J.C. Schwartz, Glycogeno-
lysis induced by serotonin in brain: identification of a new
class of receptor, Nature (Lond.) 298:373 (1982).
5. A.C. Dolphin, and P. Greengard, Serotonin stimulates phosphoryla-
tion of protein I in the facial motor nucleus of rat brain,
Nature (Lond.) 289:76 (1981).
6. M. Hamon, ~. tlourgoin, S. El Mestikawy, and C. Goetz, Central
serotonin receptors, in:"Handbook of Neurochemistry", A. Lajtha,
ed., Plenum Publ. Corp., New York (in press).
7. M. Hamon, S. Bourgoin, J. Jagger, and J. Glowinski, Effects of LSD
on synthesis and release of 5-HT in rat brain slices, Brain Res.
69:265 (1974).
8. H.J. Haigler, and G.K. Aghajanian, Serotonin receptors in the
brain, Fed. Proc. 36:2159 (1977).
9. M. Gothert, Modulation of serotonin release in the brain via pre-
synaptic receptors, T.I.P.S. (Nov.):437 (1982).
10. F. Cerrito, and M. Raiteri, Serotonin release is modulated by
presynaptic autoreceptors, ~ur ._ J. Pharmacol. 57:427 ( 1979) .
Jl. L.E. Arvidsson, U. Hacksell, J.L.G. N~lsson, S. Hjorth, A. Carlsson,
P. Lindberg, D. Sanchez, and H. Wikstrom, 8-hydroxy-2-(-di-n-
propylamino)tetralin, a new centrally acting 5-hydroxytryptamine
receptor agonist, J. Med. Chern. 24:921 (1981).
12. S. Ahlenius, K. Larsson, L. Svensson, S. Hjorth, A. Carlsson, P.
Lindberg, H. Wikstrom, D. Sanchez, L.-E. Arvidsson, U. Hacksell,
and J.L.G. Nilsson, Effects of a new type of 5-HT receptor ago-
nist on male rat sexual behavior, Pharmacol. Biochem. Behav.
15:785 (1981).
13. L. Svensson, and S. Ahlenius, Enhancement by the putative 5-HT
receptor agonist 8-0H-2-(di-n-propylamino)tetralin of the acou-
stic startle response in the rat, Psvchopharmacol. 79:104 (1983).
14. H. Gozlan, S. El Mestikawy, S. Bourgoin, M. Hall, L. Pichat, J.
Glowinski, and M. Hamon, The specific labelling of pre- and post-
synaptic 5-HT receptors by 3H-PAT in the rat CNS, Proceedings
nf thP French-German Pharmacological Meeting, abst.73 (1983).
15. H. Gozlan, S. El Mestikawy, L. Pichat, J. Glowinski, and M. Hamon,
Identifi3ation of presynaptic 5-HT autoreceptors with a new
ligand: H-PAT, Nature (Lond.) 305:140 (1983).
16. M.H. Thiebot, M. Hamon, and P. Soubrie, Attenuation of induced-
anxiety in rats by chlordiazepoxide: role of raphe dorsalis
benzodiazepine binding sites and serotoninergic neurones,
Neuroscience, 7:2287 (1982).
17. H.L. Goldberg, and R.J. Finnerty, The comparative efficacy of
buspirone and diazepam in the treatment of anxiety, Am. J.
Psvchiat. 136:1184 (1979).

346
INDALPINE, A NEW SEROTONIN UPTAKE INHIBITOR

A. Darragh, M. Kenny, R. Lambe, T. Lenehan, I. Brick

and C. Maulet*

Institute of Clinical Pharmacology, Dublin 8, Ireland

*Pharmuka Laboratoires, Gennevilliers, France

INTRODUCTION

Many of the currently available antidepressant drugs produce

unwanted anticholinergic effects which range from minor effects
such as blurred vision and dryness of the mouth to urinary
retention and exacerbation of glaucoma. Anticholinergic effects
may be particularly severe in middle-aged and elderly patients. In
addition many of the peripheral symptoms associated with depression
may be exacerbated by treatment with those antidepressants which
have significant anticholinergic activity.

Indalpine, 4-[2-(3-indolyl) ethyl] piperidine, is a recently-

developed compound not structurally related to the tricyclic
antidepressants. It is selectively active on the central
serotonergic system and inhibits serotonin uptake in vivo and in
vitro without affecting that of the catecholamines. (Le Fur and
Uzan, 1977) It is practically devoid of anticholinergic or
monoamine oxidase inhibitory properties.

Results from animal studies suggest that indalpine is a

potential antidepressant with some anxiolytic properties(personal
communication, Groupe Pharmuka, France). The antidepressant
activity of indalpine has been confirmed in clinical trials in
depressed patients and the efficacy was similar to that of
clomipramine, while tolerance to indalpine was markedly superior
(personal communication, Groupe Pharmuka, France).

In the present study, platelet serotonin uptake was

investigated following single and repeated oral doses of indalpine
to healthy male volunteers. In addition, a series of assessments,

347
sensitive to anticholinergic effects was carried out and plasma
levels of indalpine were monitored.

MATERIALS AND METHODS

Design

The study was carried out as a double-blind 2-way crossover

study in two groups of 6 subjects. Oral doses of indalpine were
administered as outlined below. The order of treatment was
randomised so that 6 subjects were placed on the indalpine schedule
first, followed a week later by the placebo schedule, while this
was reversed in the other subjects. All subjects received 3
tablets each day for six days in each period and there was an
interval of 7 days between successive periods. The daily indalpine
dose increased from 50 mg on day 1 to 150 mg on days 3 and 4
(table 1).

Table 1. Treatment Schedule

A: Day 1: 50 mg Indalpine at 08.00 hours.

Placebo at 12.00 and 16.00 hours.
Day 2: 50 mg Indalpine at 08.00 and 12.00 hours.
Placebo at 16.00 hours.
Days 3-4: SO mg Indalpine at 08.00, 12.00 and 16.00 hours.
Days S-6: Placebo at 08.00, 12.00 and 16.00 hours.

B: Days 1-6: Placebo at 08.00' 12.00 and 16.00 hours.

Subiects

Twelve healthy male volunteers aged between 18 and 36 years

and within 10% of their ideal body weights were selected on the
basis of normal history and physical examination. The subjects
all had normal ECG and renal, hepatic and haematopoietic function.
Subjects were excluded if they had taken any medication within 14
days prior to the beginning of the study or if there was any
history of allergy, drug hypersensitivity or drug abuse. Before
the study began the protocol was approved by the Institutional
Review Board and written informed consent obtained from each
subject. The study was carried out in accordance with the
principles laid down in the Declaration of the World Medical
Assembly of 1975.

Assessments

A series of assessments was carried out at frequent intervals

after dosing each day in order to identify possible anticholinergic
side effects. These included:-
(i) Measurement of blood pressure and heart rate (both supine and
standing).

348
(ii) Ophthalmological examinations : pupil diameter and near and
far points of accommodation were measured under carefully
controlled light conditions.
(iii) Saliva output was measured essentially as described by
Dollery et al (1976).

Blood samples were taken pre-dose on day 1, and day 2 hours after
the first dose on each day in order to evaluate the effects of
indalpine on the uptake of 3H-serotonin by platelets. Tritiated
serotonin was added to platelet-rich plasma and uptake was assessed
by a standard method (Kenny et al, 1983). Plasma from treated
volunteers was incubated with platelets isolated from untreated
subjects to determine its effect on the uptake of tritiated
serotonin by normal platelets.

In order to measure plasma levels of indalpine, blood samples

were also taken before the first dose each day and 3 hours after
every dose. Indalpine levels were determined using a specific
HP1C assay (Jozefczak et al. 1982).

RESULTS

The data obtained from the anticholinergic tests were

statistically analysed by one and two way analyses of variance.
Comparisons were made between the two treatment groups (indalpine
versus placebo), and within each group the data obtained after the
first dose were compared with the pre-dose values. After both
single and repeated doses of indalpine no significant changes were
observed in blood pressure or heart rate (supine and standing), in
the ophthalmological examinations or in saliva output. The
evaluations were made at half-hourly intervals up to 2 hours and
at hourly intervals up to 9 hours after the first dose each day.

A marked inhibition of 3 H-serotonin uptake into platelets was

observed 2 hours after the first 50 mg dose of indalpine (figure 1).
The uptake was reduced to 10% of the pre-dose control value on day
4 of the study. Indalpine was replaced by placebo on days 5 and 6
and the uptake capacity of platelets increased, reaching 70% of
the pre-dose control value forty-two hours after the final dose of
indalpine.

When plasma from the treated subjects was incubated with platelets
isolated from untreated subjects the uptake of 3H-serotonin
followed a similar pattern or inhibition, reaching 11% of the
pre-drug control value on day 4 of the study. The correlation
between the latter results and plasma levels of indalpine is
illustrated in figure 2.

349
 100

80

Q)
,.!>G
<II
60
+
.j..)
p..
p

~I
Ll"l
40 -
~

20 f~"! /!
··i
------.
1 2 3 4 5 6

Time (days)

Fig. 1. 3 H-serotonin uptake by human platelets during oral dosing

with indalpine, expressed as a percentage of the pre-dose control
(mean.:!:_ S.E.M.).

DISCUSSION

Many of the classical antidepressants inhibit the uptake of

exogenous 3 H-serotonin into blood platelets, a procedure which is
generally accepted to reflect inhibition of serotonergic neuronal
re-uptake. The results of the present study indicate that indalpine
compares favourably with other recently developed compounds in its
ability to inhibit platelet serotonin uptake. In addition, no
significant anticholinergic side effects were encountered during
the study.

350
 100

80
Q)
~
13p. 60
::::>

~ 40
I
Lr)

~ 20

100-

•
B

80

,.-j 60 ~ :t ---------
-..s

.
gp 40 -------- :l .·
/:t
20 / I
·~
/
.
1 2 3 5 6

Time (days)

Fig. 2. a. The effect of plasma, obtained from the subjects treated

with indalpine on the uptake of 3 H-serotonin by normal platelets,
expressed as a percentage of the pre-dose control (mean+ S.E.M.)
b. Plasma levels of indalpine during the treatment period (mean +
S.E.M.) -

351
REFERENCES

Dollery, C.T., Davies, D.S., and Draffen, G.H. 1976, Clinical

pharmacology and pharmacokinetics of clonidine, Clin.
Pharmacal. Ther.,l9:11.
Jozefczak, C., Ktorza, N. and Uzan, A. 1982,High performance liquid
chromatographic determination of indalpine, a new
non-tricyclic antidepressant, in human plasma, J.
Chromatogr., 230:87.
Kenny, M., Lenehan, T.J., Lambe, R., Darragh, A., Brick, I. and
Omer, L.M.O. 1983, Effects of a single oral dose of 3-cyano-
imipramine on serotonin uptake and content of platelets in
healthy volunteers, Psychopharmacology, 79:304.
Le Fur, G. and Uzan, A. 1977, Effects of 4-(3~indolyl-alkyl)
piperidine derivatives on uptake and release of
noradrenaline, dopamine and 5-hydroxytryptamine in rat brain
synaptosomes, rat heart and human blood platelets, Biochem.,
Pharmacal., 26:497.

352
A DOUBLE BLIND OUT PATIENT COMPARATIVE TRIAL

OF INDALPINE VERSUS MIANSERIN

B. Martin and G.J. Naylor

Department of Psychiatry
University of Dundee
Dundee DD1 9SY, Scotland

Indalpine, inhibits the re-uptake of 5HT but it is obvious

that there is no single agreed hypothesis to explain depression.
Whatever the proposed mechanism of action of an antidepressant,
the main question for the clinician is "is the drug an effective
and safe antidepressant in clinical practice," and the following
trial was carried out to determine Indalpine's efficacy by
comparing it to an established antidepressant. This design was
chosen because of the ethical difficulties of placebo controlled
out-patient trials but it is scientifically the weaker design
because the null hypothesis is then that the two drugs are equally
effective and failure to reject this hypothesis may be the result
of a poorly designed trial, too small numbers, etc. Thus one
does not prove the two drugs to be similar, one simply fails to
prove that they differ. Despite such difficulties, however,
double blind control trials are essential tools for the assessment
of a new antidepressant and many such trials are necessary before
the effectiveness of a drug can be said to be established. With
placebo effect and observer bias, open trials are unsuitable for
assessing such drugs.

Patients

In this trial a total of 65 depressed out-patients were

randomly allocated into two groups to receive treatment either
with Indalpine or Mianserin. Six patients dropped out within the
first seven days and have, therefore, been excluded from the
analysis. Fifty two patients completed four weeks of treatment,
six completed seven days and one completed 14 days. For
statistical purposes, a total of 29 patients received Indalpine
and 30 patients received Mianserin. The patients aged from 17 to
353
68 years. All trial subjects were recruited from referrals to a
general psychiatric out-patient department. Patients were
considered for entry to the trial if in the opinion of the
investigator they were depressed and thought likely to benefit
from treatment with an antidepressant and were not suffering from
any other concomitant psychiatric condition apart from depression.
No formal diagnostic criteria were applied but patients were
divided into psychotic or neurotic groups using the Newcastle
Scale and given an I.C.D. diagnosis after entry to the trial: all
were suffering from either manic depressive psychosis depressed
type or neurotic depression.

Drugs

Patients were required to have been free from any

antidepressant treatment over the seven days prior to.the entry to
the trial and were excluded if in the previous six weeks they had
received Mianserin. The only other drug used during the trial
was Temazepam, as a night sedative.

During the trial, the patients received their medication in

identical capsule form, one capsule in the morning and one at
night. For the first three days, the patients received a smaller
increasing dose of the trial drug reaching a maximum of 150 mg
Indalpine or 60 mg Mianserin by the fourth day. This dosage then
remained stable for the remainder of the trial.

Methods

Patients were seen four times over a 28 day period: that is

at day 0, 7, 14 and 28. On each occasion the patients were rated
using the Hamilton Rating Scale for depression [1], the Beck Self
Rating Scale [2], a visual analogue self rating scale for anxiety,
a simple seven point global rating scale of depression and a
side-effect checklist. In addition, prior to the trial, the
Newcastle Scale [3] was completed and full psychiatric and
physical histories were obtained.

Blood samples were taken on days 0, 14 and 28 for

haematological and biochemical safety screens. Drug levels were
not estimated during the trial.

Results

At the start of the trial, the two groups were well matched.
The mean age of the Indalpine group was 45 years and of the
Mianserin 41 years. There were seven men in the Indalpine group
and five i~ the Mianserin group. The mean of the baseline Hamil-
ton Rating Scale score for de2ression was 17.04 for Indalpine.

354
 a HAMIL TON MEANS b BECK MEANS
18
MIANSERIN
INOALPINE

--,
''
'
''
''
' '·
'
'•

2
WEEK WEEK

Fig. 1a and b. Hamilton and Beck means.

and 17.46 for the Mianserin. The mean of the Beck Self Rating
Scales was 15.35 for the Indalpine group and 16.17 for the
Mianserin group and the mean of the visual analogue self rating
scale for anxiety was 5.48 for the Indalpine group and 5.46 for
the Mianserin group. Each of the two drug groups also contained
similar proportions of patients having psychotic or neurotic
depressive illness as categorised by the Newcastle Scale. Of
these scoring seven or more on the Newcastle Scale, five were in
the Indalpine group and six in the Mianserin treated group.

Eight of the drop-outs had been receiving Indalpine whereas

five had been on Mianserin. The main reason for the failure to
complete the trial was the development of side-effects (see
later). The age, sex distribution and severity of depression. was
roughly the same in the two groups of drop-outs as in those
remaining in the trial. Therefore, there was no suggestion that
older patients or patients at one end of the depressive spectrum
were more likely to drop-out and no suggestion that the drop-outs

355
 BECK> ~X. IMPROVEMENT HAMIL TON> SO'X. IMPROVEMENT

D MIANSERIN D MIANSERIN

100
(=:1 INOALPINE D INOALPINE

WEEK WEEK

Fig. 2. Beck and Hamilton >50% improvement histograms

biased the trial. Most of the drop-out patients required further

treatment in the form of another antidepressant.

The significance of the results were assessed in three ways

by non-parametric statistics:

1. comparison of the scores at each assessment point.

2. comparison of the changes in the scores from baseline.

3. calculation of the number of subjects who had improved by

more than 50% over the baseline score.

Both groups improved over the 28 days of the trial and

comparing the scores directly, there was no significant difference
between them, although on closer examination there was the
suggestion that on the Hamilton and Beck ratings, Indalpine
appeared to be slower or less effective in the first two weeks.

356
 VISUAL ANALOGUE MEANS

GLOBAL MEANS
• MIANSERIN
MIANSERIN INDALPINE
INDALPINE

4
..._

3 3
--- -·
2 2

WEEK WEI!K

Fig. 3a and b. Global and visual analogue means.

When changes from baseline were considered - in general,

there were significant differences between the groups on the Beck
at week two and the Hamilton at week one and two: the Mianserin
group showed significantly more change (p <0.02 for the Hamilton
scores and p <0.04 for the Beck scores), see Fig. 1a and 1b.

When the number of subjects improving by 50% was examined,

the Indalpine group did worse in the first two weeks and
significantly so on the Hamilton (p <0.02) but not significantly
so on the Beck (p <0.1). Thus at two weeks on the Hamilton Scale
only two patients on Indalpine had improved by 50% whereas ten had
done so on Mianserin. However, by 28 days, the response to the
two drugs was very similar. See Fig. 2a and 2b.

There was no statistically significant difference in measures

of anxiolytic activity over time. The global scores for severity
were very similar between the two groups throughout the trial.
See Fig. 3a and 3b.

357
 At the end of the four week trial period nine patients taking
Indalpine and nine Mianserin were considered sufficiently improved
to justify inclusion into a longer term trial.

Three patients, one receiving Indalpine and the other two

Mianserin required hospitalisation at the end of the trial period
because of persistant severe depressive symptoms. Four more
patients receiving Mianserin and two Indalpine were changed to
another antidepressant after the four weeks because of their poor
response during the trial.

Subjective Side Effects as Reported by the Patient

A full screen of side-effects was made. Statistical

analysis of the full side-effect checklist showed no significant
difference in the incidence of side-effects. For the purposes of
analysis, a side-effect was considered to have occurred if
reported during the trial at an intensity greater than that
reported at baseline.

Breaking down the list into the more usually accepted

categories of antidepressant side-effects, i.e. cardiovascular,
anticholinergic, central nervous system and a non-specific group
did not suggest that either drug had significant advantages in any
of these categories.

Adverse Effects Detected by Laboratory Screening

Fifteen patients recruited approximately equally from both

groups demonstrated persistently abnormal liver enzyme levels
during the trial and although these were not considered clinically
significant, they should perhaps be borne in mind.

One patient, a woman aged 32, in the Indalpine group,

developed a mild leucopenia as detected by a white cell count of
3.7 at day 14. Her baseline white cell count was 4.9. This
change affected mainly the neutrophils, their proportion falling
from 69.5% to 54.5% of the total couunt. This change was thought
to be due to drug toxicity and the patient was withdrawn from the
trial. Her white cell count returned to normal shortly after the
drug withdrawal. Scrutiny of the white cell count and blood
films for all patients over the trial period did not indicate any
generalised trend towards leucopenia. However, another patient
on Indalpine, who had been through this trial and had been
continued on the drug developed a more serious leucopenia after
three months on the drug, which again reversed on withdrawal of
the drug.

358
 Also a 62 year old lady in the Indalpine group suffered a
myocardial infarction ten days after commencing Indalpine and
shortly afterwards went into asystolic arrest and was not
successfully resuscitated and died. She had a past history of
myocardial infarction which was not known at the time of entry
into the trial despite a medical history being taken. There was
no obvious causal link between Indalpine and the myocardial
infarction.

Conclusion

This trial provides data on only 29 patients recelvlng

Indalpine for only 28 days. In the opinion of the investigators
a single brief study such as this does not provide adequate
information on which to base decisions about the usefulness or
safety of the drug. Further well designed and conducted studies
are required. In the opinion of the investigators, the
experience obtained in this trial suggests that Indalpine needs to
be used cautiously and only when patients can be carefully
monitored.

REFERENCES

1. M. Hamilton, A rating scale for depression. J. Neurol.

Neurosurg. Psychiat., 23, 56-65 (1969).

2. A. T. Beck, Archives of Gen. Psychiatry, 4, 561-571 (1961).

3. M. W. P. Carney, M. Roth, and R. F. Garside, The diagnosis of

depressive syndromes and the prediction of ECT response. Br.
J. Psychiat., 111, 659-674 (1965}.

359
THE COMPARATIVE EFFECTS OF INDALPINE AND AMITRIPTYLINE

ON SALIVA PRODUCTION

Malcolm Peet

Dept. of Psychiatry
Scarsdale Hospital
Chesterfield, Derbyshire
England

Indalpine is a potent inhibitor of 5-Hvdroxytryptamine uptake

by central neurones (Le Fur et al., 1978) and has been shown in
comparative clinical trials to be an effective antidepressant
agent (Guelfi et al., 1981; Gerard et al., 1981).

The more recently developed antidepressant drugs have been

found in general to have a similar spectrum of clinical efficacy
to the old~r tricyclic drugs, but to differ in side-effect profile.
Some of the most troublesome side-effects of tricyclic anti-
depressants stem from their anticholinergic properties. Amongst
these, dryness of the mouth due to dimished saliva production is
common, maybe troublesome enough to lead to discontinuation of drug
treatment, and in the longer term gives rise to an increased
incidence of dental caries and mouth infections. (Bertram et al.,
1979).
In a number of clinical studies the incidence of subjective
anticholinergic effects including dry mouth has been found to be
lower with indalpine than with tricyclic antidepressant agents.
However, subjective evaluation of side-effects is not always accu-
rate, and it is important to substantiate these clinical findings
with more objective methods of assessment. In animal studies,
indalpine was found to lack anticholinergic properties (Le Fur et
al., 1978). Single doses of indalpine given to healthy volunteer
subjects produced no significant change in saliva production
(Iliopoulou, A. et al., 1981). However, this finding cannot
necessarily be extrapolated to depressive patients. Such patients
have a significantly reduced saliva flow, even without drug treat-
ment, and salivation increases towards normal as the depression
remits (Bolwig and Rafaelsen, 1972).

361
 The present study was designed to objectiv ely measure saliva
productio n in depressiv e patients given indalpin e or amitript yline
over a 4 week period, and at the same time to gain an initial
impressio n of the clinical profile of activity of indalpin e.

Subjects were patients with primary depressiv e illness attending

a psychiat ric out-pati ent clinic. Patients were allocated at random
to be treated with either indalpine or amitript yline, with 2 patients
given indalpin e for every 1 on amitript yline. Drug doses were 50 mg
twice daily for the first week increasin g to 75 mg twice daily for the
remaining 3 weeks; drug administ ration was not double-b lind. Three
patients (2 unable to co-opera te with salivary volume measurem ent, 1
who dropped out during the first week) were replaced so that there
were 18 patients (12 on indalpin e, 6 on amitript yline) who had full
salivary volume data available for at least 1 week of treatmen t.
Following a base line drug free period which was usually 7 days,
salivary volume was measured using the method of Kingsley and Turner
(1974), and this was repeated after 1 week and 4 weeks of treatmen t.
In addition , clinical ratings were made using the Depressio n Inventor y
of Beck et al., (1961), and the Hamilton Rating Scale for depressio n
(Hamilton , 1967). Patients were asked to voluntee r side-eff ects and
subseque ntly a side-eff ect check list was complete d.

The age, sex, and weight distribu tion of the patients is shown
in Table 1. The groups do not differ significa ntly.

Table 1: Patient Characte ristics.

Sex
Weight Age
<ran gel (ran gel M F

Indalpine 68.8 46.2

1 11
=
(n 14) (48 -103) (17 -57)

Amitriptylinet 68.2 41.6

3 3
=
<n 6) (51-91) (24 -61)

tExclude s one patient who dropped out in first week.

362
Two patients on indalpine and one on amitriptyline did not complete
the full 4 weeks of the study. Data on salivary volume for the re-
maining patients are shown in figure 1. It can be seen that ami-
triptyline produced a marked reduction in salivary volume which
persisted throughout the 4 weeks of treatment, whereas indalpine
produced no significant fall in saliva production, the difference
between the two drugs being statistically significant.

q
e
a)
6
'
~\
n =10
.,; \
•
§
+I
'
\
\
p<0.002

r----------
Q)
E 'p<0.004

--
::I \ __ -o
~ \
....>. 4 n=5
~
fti
V'l

2
0 4
Weeks of treatment

Figure 1. Salivary Volume (mls in 2 mins) in Patients Taking

Indalpine • • or Amitriptyline o------o

The changes in clinical ratings and subjective side-effects are

less conclusive than the more objective measurement of salivary
volume, because the study was not double-blind, and also because of
the small numbers of subjects involved. However, there were no
obvious differences between the groups in clinical response. There
was a significant difference between the groups with regard to the
side-effect dry mouth, amitriptyline producing an increase in this
symptom from base-line values, but indalpine having no consistent
effect. It was noted that 6 patients taking indalpine complained of
gastric upset(stomach churning, nausea) but this did not reach
statistical significance as a number of patients complained of simi•
lar symptoms in the base-line side-effect record. Three patients on

363
indalpine and 1 patient on amitriptyline showed small elevations of
liver enzymes during treatment; such effects are quite common during
treatment with tricyclic antidepressants (Blackwell, 1981),

It is of interest to put these findings with indalpine into the

context of other studies that have been performed on saliva production
following administration of antidepressant drugs. These studies are
summarised in Table 2. It can be seen that tricyclic antidepressants
produce a fall in saliva ~roduction, either on single dose or on rep-
eated dose to volunteers or patients. The newer antidepressants, by
contrast, are much less prone to alter saliva production.

Table 2: Effect of Antidepressant Agents on Saliva Production.

Multiple dose

Single dose Depressive

!volunteers> Volunteers patients

Amitriptyline
Imipramine
Dothiepin '' ''
(J) '
Maproti I ine
Nortriptyline
Clomipramine
'
'' '
Lithium 0
Isocarboxazid 0
ECT t
Viloxazine 0
Nomifensine (J)
Zimelidine
Mianserin
(J) 0

' '
0 t
Indalpine 0 0

0 = no change; ~, f = fall or rise; (I) = non-significant fall

Sources: Bayliss et al, 1974; Bertram et al, 1979; Bolwig and
Rafaelsen, 1972; Ghose et al, 1976; Kopera, 1978; Sheth et al,
1979; Rafaelsen et al, 1980; Swift et al, 1981; von Knorring,
1981.
364
Indalpine fits into this table as producing no significant effect on
saliva production either on single dose to volunteers or, from the
present study, when given chronically to depressive patients. Whilst
this study did not include a placebo, and therefore no statement on
the absolute effect of indalpine can be made, it is certainly
encouraging that indalpine is markedly better than amitriptyline with
regard to effects on saliva production.

Nine patients have gone on to receive treatment with indalpine

over longer periods of 6 - 12 months. These patients have been
regularly monitered biochemically (liver function, haematology, blood
urea and creatinine, blood sugar) and there have been no untoward
changes in these parameters, apart from one man who became anaemic
because of a previously undiagnosed bleeding gastric ulcer thought to
be unconnected with his indalpine treatment. So far none of these
patients have shown an acute depressive relapse.

In conclusion, the main finding from this study is that indalpine

lacks anticholinergic properties relative to amitriptyline, as measured
by saliva production, and this may constitute a significant clinical
advantage over tricyclic antidepressant drugs.
REFERENCES

Bayliss, P.F.C., and Duncan, S.M., 1974. The clinical

pharmacology of viloxazine HCl - a new antidepressant of
novel chemical structure, Br.J.Clin. Pharmacol., 1:431.
Beck, A.T., Ward, C.H., Mendelson, M., Mock, J.", and
Erbaugh, J., 1961, An inventory for measuring depression,
Arch. G@n. Psvchiat., 4:561.
Bertram, u., Kragh-·Sorensen, P., Raf aelsen, o., and
Larsen, N.-E., Saliva secretion following long-term
antidepressant treatment with nortriptyline controlled
by plasma levels. Scand. J. Dent.Res., 87:58.
Blackwell, B., 1981, Adverse erreccs of antidepressant
drugs. Part 1: Monoamine oxidase inhibitors and tri-
cyclics, prugs, 21:201.
Bolwig, T.G., and Rafaelsen, O.J., 1972, Salivation in
affective disorders, Psvchol. Med., 2:232.
Gerard, A., Clerc, G., Guibert, M., Robert, G., and
Pagot, T., 1981, A double-blind investigation of a highly
selective 5-HT uptake inhibitor: Indalpine vs.
clomipramine in hospitalised depressive patients, presented
at Eighth International Congress of Pharmacology, Tokyo.
Ghose, K., Coppen, A., and Turner, P., 1976, Autonomic
actions and interactions of mianserin hydrochloride
(Org GB94) and amitriptyline in patients with depressive
illnes, Psvchonharmacol., 49:201.
Guelfi, J.D., Dreyfus, ·J.-F., Boyer, P., and Pichot, P.,
1981, A double-blind controlled multicenter trial com-
paring indalpine with imipramine, presented at III World
Congress of Biological Psvchiatrv, Stockholm.

365
 Hamilton, M. 1967, Development of a rating scale for
primary depressive illness, Br.J. Soc.Clin. Psvchol.,
6:278
Iliopoulou, A., Kaspi, T., Blackett, A. and Turner, P.,
1981, Clinical pharmacological studies with LM 5008, a
new antidepressant, Pharmatheraoeuti ca, 2:613.
Kingsley, P.J., and Turner, P., 1974, Class experiment
in clinical pharmacology with benzilonium bromide, an
anticholinergic drug, Euroo. J. Clin. Pharmacol., 7:141.
Kopera , H., 1978, Anticholinergic and blood pressure
effects of mianserin, amitriptyline and placebo,
Br.J.Clin. Pharm~~~l., 5 Suppl. 1:29S.
Le Fur, G., Kabouche, M., and Uzan, A., 1978, On the
regional and specific serotonin uptake inhibition by
LM 5008, Life Sci., 23:1959.
Rafaelsen, O.J., Clemmesen, L., Lund, H., Mikkelsen, P.L.,
and Bolwig, T.G., 1980, Comparison of peripheral anti-
cholinergic effects of antidepressants: dry mouth,
Act. Psvchiat. Scand., 63: Suppl. 290.
Sheth, U.K., Paul, T., Desai, N.K., and Pispati, P.K.,
1979, Comparative effects of imipramine and dothiepin
on salivary rate in normal volunteers, Br.J.Clin. Pharmacol.,
8:475.
Swift, C.G., Haythorne, J.M., Clarke, P., and Stevenson, I.H.,
1981, Cardiovascular, sedative and anticholinergic effects
of amitriptyline and zimelidine in young and elderly
volunteers, _A.ct. Psvchiat. Scan<!., 63: Suppl. 290.
von Knorring, L., 1981, Changes in saliva secretion and
accommodation width during short-term administration of
imipramine and zimelidine in healthy volunteers, Int.
Pharmacopsychia t., 16:69.

366
ENTRAINMENT AND MASKING IN THE CHRONOBIOLOGY OF DEPRESSION

Detlev von Zerssen, Gerhard Dirlich, Peter Doerr,

Reimer Lund, Karl Martin Pirke, and Hinderk M. Emrich

Max-Planck-Institut fur Psychiatrie

D-8000 Munich I FRG

In order to analyse the possible role of the main circadian

pacemaker (1) in endogenous depression (2), a large scale chrono-
biological investigation has been performed by an interdisciplinary
research group at our institute. 20 inpatients with the ROC-diag-
nosis of major depressive disorder, endogenous subtype, 10 of whom
could be reinvestigated after, at least almost, complete clinical
remission, and 10 healthy probands of similar sex- and age distri-
bution were included in the study. All subjects had been drug free
for usually at least one week prior to the onset of an observation
period of at least two weeks and were, with only one exception, not
medicated before the termination of it. During this time period
night sleep was recorded polygraphically (cf. H. Schulz et al. in
this volume), and a series of equidistant measurements was taken
six times every day from 7:00 a.m. to 10:00 p.m. and once every
night at 2:30 a.m. Several of the investigated variables were re-
garded as reflecting, more or less directly, the degree of severity
of depression (three mood scales, a self-assessment of daily activi-
ties, two measures of spontaneous motor activity, a motor speed test
and a calculation test). The other variables, although conceived as
possibly being related to the severity of depression, were primarily
selected as indicators of the function of the main circadian pace-
maker, i. e. salivation rate, urine volume and urinary excretion of
sodium, potassium and free cortisol, and, furthermore, core body
temperature which, at night, was recorded continuously by a rectal
probe. In addition, also room temperature and humidity were measured
at the times of the subjects's assessments.

With the exception of one severely depressed patient (first

episode of an agitated delusional depression), the individual data
sets were complete enough for an evaluation of the educed 24h wave-

367
form and the power spectrum of each variable in every subject. The
present analysis was primarily based on an intraindividual compari-
son of the sample of the 10 patients investigated in depression as
well as during remission (using the Wilcoxon-test for correlated
samples). Differences reaching at least the 10% level of significance
were accepted as hints to changes of chronobiological parameters
during depression. They were then used for predicting the results
of an interindividual comparison (by means of the u-test) of patients
in depression versus healthy controls. The one patient with an in-
complete data set and three others in whom, according to clinical
judgement (independent of the results of the chronobiological in-
vestigation), the diagnosis of an endogenous subtype of depression
seemed doubtful, were excluded from this analysis leaving data of
16 depressives for the comparison with those of the 10 controls.
A one-tailed test was applied whenever a difference had been pre-
dicted on the basis of the intraindividual comparison. In all other
instances a two-tailed test was employed. The 10% level of signi-
ficance was accepted as indicating a trend in the data. Most of the
differences were, however, either insignificant or reached at least
the 5% level.

The results of this analysis can be summarized as follows:

There was no indication of free-running rhythms in any of the
variables. The circadian pattern was well preserved during depression
in all indicators of the main circadian pacemaker, including urinary
free cortisol, although the range of oscillation was somewhat
attenuated in some of them (e.g •. urinary potassium and rectal
temperature). There was a slight tendency to phase advance in uri-
nary potassium, even in the remitted state, compared with the values
of the control sample. However, the salivation rate exhibited the
opposite tendency towards a phase delay.

The changes in the indicators of the main clock were minimal

in comparison with those of the mood scales and spontaneous motor
activity (recorded at the left wrist). Here, marked changes in level,
indicating a wors~ning of mood and a decrease in motor activity,
were elicited together with an attenuation of the range of the
circadian oscillation. In the mood data, but not in the measures of
motor activity, a pronounced phase delay was ascertained, the peak
in the scale values occurring around awakening and not in the middle
of the night as in a fully remitted state or i~ healthy subjects.

Room temperature was significantly higher during the investi-

gation of the patients in depression, obviously due to the fact
that, on average, they had been investigated more often during
summer. However, after controlling for the difference in seasonality
of the respective investigations most of the group differences in
the variables remained significant although, with respect to the
indicators of the main circadian pacemaker, only at a low or even

368
borderline {10%) level of significance.

In conclusion, no evidence of a specific dysfunction of the

main circadian pacemaker has as yet been found in depressives
living under a strict 24h routine. It is an open question whether
this is, at least in part, due to the time cues involved in the
design of the study. Diurnal variation in the symptomatology of
depression, most prominent in the mood data, may be interpreted as
reflecting an entrainment {1, 3) of the central pathogenetic
mechanism of endogenous depression to the internal clock, since the
maximum and the minimum of the average educed waveform of the mood
scale coincided with those of the respective waveform of urinary
free cortisol and correlated significantly with the acrophase of
cortisol-excretion. In addition, masking effects {1, 3) have to be
taken into account, particularly with respect to the temperature
data: A similar flattening of the diurnal pattern, mainly due to
an elevation of the circadian minimum during the rest period {as
observed in patients with endogenous depression), was reported to
occur in animal experiments as a response to repeated handling
procedures involved in taking rectal temperature {4; cf. 5: fig. 2).
These procedures seem to have induced emotional stress and interrup-
tions of sleep - phenomena that, independently of experimental
manipulations, belong to the classical symptom pattern of endogenous
depression. Therefore, the changes in body temperature of depressives
may be interpreted as a consequence of the symptomatology of their
disorder {i. e. emotional stress and sleep disturbances) rather than
a symptom of a dysfunction of the main circadian pacemaker.
REFERENCES

1. J. Aschoff, "Biological Rhythms," Handbook of Behavioral

Neurobiology, Vol. 4, Plenum Press, New York - London {1981).
2. R. A. Wehr and F. K. Goodwin, Biological rhythms and psychiatry,
in: "American Handbook of Psychiatry," 2nd ed., Vol. 7,
Advances and New Directions, s. Arieti, ed., Basic Books,
New York {1981).
3. R. A. Wever, "The Circadian System of Man," Springer, New York -
Heidelberg - Berlin {1979).
4. G. H. Miles, Telemetering techniques for periodicity studies,
~· N.Y. Aead. Sci. 98:858 {1962).
5. D. von Zerssen, Chronobiology of depression, in: "The Origins
of Depression," J. Angst, ed., Springer, Berlin- Heidel-
berg- New York - Tokyo {1983).

369
THE ORIGIN OF EARLY REM SLEEP EPISODES IN DEPRESSION

AND OTHER CONDITIONS

Hartmut Schulz, Reimer Lund, and Stephan Volk

Max Planck Institute for Psychiatry

Kraepelinstrasse 10
8000 Munchen 40, Fed.Rep. Germany

INTRODUCTION

Normal sleep onset in the human adult is characterized by a

transition from wakefulness to non-REM sleep. There is a pro-
gression of the sleep stages 1 2 3 4, which is ended
after about an hour by the emergence of the first REM sleep epi-
sode. Numerous studies have shown that this basic structure of the
first sleep cycle is significantly altered in many nights of de-
pressed patients. As Kupfer and coworkers have shown, REM latency
is abbreviated during depression (Kupfer, 1976). In addition, the
amount of slow wave sleep stages 3 and 4 may be reduced by the
illness (Gillin et al., 1979). While a reduction of slow wave
sleep has also been observed in other patient samples and in
healthy subjects of advanced age, short REM latency has been
claimed to be more specific and to represent a biological marker
of primary depression (Kupfer, 1976). However, specificity of
short REM latency as a marker for depression has been challenged
by the fact that short REM sleep latencies may also occur in nor-
mal aged subjects (Spiegel, 1981) and in patients with different
psychiatric diseases (Jus et al., 1973; Insel et al., 1982). In
addition, sleep onset REM episodes (SOREM, latency~ 20 minutes
after sleep onset) have been observed after a shift of the regu-
lar sleep time (Decoster and Foret, 1979), after a reduction of
sleep time (Mullaney et al., 1977), in nap studies (Karacan et
al., 1970), and in subjects living on an ultradian rest-activity
schedule (Weitzman et al., 1974).

Later studies have demonstrated that the distribution of REM

sleep latencies in depressed patients is bimodal with a first peak

371
in the latency class 0-20 minutes and a second peak at about 60-70
minutes (Schulz et al., 1979). The bimodal form of distribution
seems to be typical for all cases where a very short mean REM sleep
latency has been observed, as e.g. in young infants (Schulz et al.,
1983), in narcolepsy-cataplexy (Montplaisir et al., 1978), in free-
running sleep-wake cycles (Chouvet et al., 1974), and in sleep re-
corded in a disentrained environment (Campbell, in press).

Hypotheses on the origin of early REM sleep episodes

Different hypotheses have been proposed to explain the occur-

rence of early REM sleep episodes in depression and other condi-
tions. In the following we will discuss primarily two of them.

The phase advance hypothesis

Starting from the assumption that REM sleep propensity con-

stitutes a circadian rhythm, different authors have proposed that
early REM sleep episodes in the sleep of depressed patients may
result from a phase-advance of the REM sleep rhythm against the
rest-activity cycle (Wehr et al., 1983). This phase-advance hypo-
thesis also assumes that the circadian rhythms of body core tem-
perature and cortisol are phase advanced, since all these overt
rhythms have been proposed to be driven by the same oscillator
(Weitzman et al., in press). While phase advance of body core
temperature in depressed patients has been reported by some authors
(Avery et al., 1981) others did not find a phase advance in de-
pression (Lund et al., 1983). In addition, the assumed phase re-
lationship between REM sleep and body core temperature has been
seriously challenged by data of Lund et al. (in press) who found
a normal nightly distribution of REM sleep in cases where the cir-
cadian temperature minimum fell early. These latter results suggest
that the phase advance hypothesis in its present form is not ten-
able.

The REM sleep-slow wave sleep interaction hypothesis

The hypothesis of a reciprocal interaction between slow wave

sleep (SWS) and REM sleep has been proposed by Borbely (1982).
Borbely assumes that the occurrePce and maintenance of sleep is
regulated by a circadian process C and a sleep-dependent process S.
The hypothetical process S is best represented by the EEG delta
activity of SWS. It is postulated (a) that the process S is de-
ficient in the sleep of depressed patients, and (b) that a "de-
ficiency of process S leads to disinhibition of the REM sleep con-
trolling process in the first part of sleep. Consequently, the la-
tency to the first REM sleep periods is shortened and its duration
increased" (Borbely and Wirz-Justice, 1982, p. 207). This hypothe-
sis is in good agreement with many observational data. Since SWS

372
 dominates at the beginning of sleep, a deficient process S during
depression would result in a more or less local disturbance of the
REM sleep process and this is indeed the case in depression. REM
sleep, which is normally inhibited at sleep onset, seems to be dis-
inhibited in many depressed patients resulting in short REM latency,
increased amount of REM sleep in the first third of recording time
and a somewhat higher eye movement density in the first REM sleep
episode. A deficiency of process S could also explain the high rate
of sleep interruptions during depression as well as early morning
awakening.

If the disinhibition of REM sleep is really the result of a

deficit in process S, an intrapatient correlation between SWS
amount and REM sleep latency could be expected over a series of
nights. As Figure 1 shows, the respective correlations are essen-
tially distributed around zero and, in the mean, are not signifi-
cant. The nonsignificant correlations between amount of delta
sleep and REM sleep latency suggest that these two sleep states
are regulated independently •

• • • ••• • • • •••• •
I I I I I I I I I I I I I I
-.6 -.4 -.2 0 +.2 +.4 +.6

Figure 1: Intraindividual Spearman rank correlations between REM

latency and amount of delta sleep (%S3+4). N = 15 endogenous depres-
sives; N = 10 to 22 nights per patient. Each dot represents one
intraindividual correlation.

Further, the reciprocal interaction hypothesis is in conflict

with some data from the literature, since sleep onset REM episodes
occasionally have been observed after extended waking phases
(Mullaney et al., 1977; Meddis, 1968; Kales et al., 1970).

Since there seem to be experimental data which are in conflict

with the phase advance hypothesis and the reciprocal interaction
hypothesis, we have formulated an alternative hypothesis which may
be more suitable to explain the occurrence of very short REM la-
tencies in depression and other conditions.

The circadian amplitude hypothesis

The circadian amplitude hypothesis supposes that abbreviated

REM latency and the occurrence of sleep onset REM episodes is asso-
ciated with an alteration of the circadian arousal cycle. The hypo-
thesis states that a flattening of the circadian arousal cycle in-

373
creases the probability for the occurrence of SOREM episodes. Indi-
cators of this hypothetical arousal cycle are measures of subjective
sleepiness and measures of body core temperature. In accordance
with this assumption, we have recently shown that the difference
between the levels of the temperature values during daytime and
nighttime is significantly smaller in persons with SOREM episodes
than in persons without such critical events (Schulz and Lund, in
press) .

A review of all clinical and experimental conditions, where

SOREM episodes have been observed, suggests that these condition
have in common a flattening of the putative circadian arousal cycle.
Such conditions may be especially critical for susceptible persons,
e.g. those who already have a small circadian amplitude of body
core temperature as a trait variable. A poorly developed arousal
cycle, or a damped one, can be ass~ed in very young infants and in
old age, in narcoleptic patients and in depressed patients. Condi-
tions which may interfere with the putative circadian arousal cycle
are (a) displacements of the habitual sleep-wake cycle, either by
shift work, time zone shift or experimentally induced, (b) experi-
mentally scheduled non-24h rest-activity cycle (ultradian rest-ac-
tivity schedules; 48h day schedule (Meddis, 1968)), (c) nap sleep
schedules, (d) night sleep reduction (Mullaney et al., 1977), total
sleep deprivation (Kales et al., 1970), (e) freerunning sleep-wake
cycles, expecially spontaneous desynchronization, and (f) static,
basal behavioral conditions (Campbell, in press). A weak point of
the circadian amplitude hypothesis is that the proposed arousal
cycle cannot be measured directly, and that we do not know whether
the overt rhythm of body core temperature is a suitable indicator
for arousal. In future experiments, other measure of arousal should
be performed on a 24h basis. These measures could include objective
ones such as heart rate and catecholamine excretion, or subjective
ones such as sleepiness ratings.

References

Avery, D.H., Wildschiodtz, G., Rafaelsen, O.J.: Nocturnal tempera-

ture in affective disorder. J.Affective Dis., 1982, 4: 61-71.
Borbely, A.A.: A two process model of sleep regulation. Human Neuro-
biol~, 1982, 1: 195-204.
Borbely, A.A., Wirz-Justice, A.: Sleep, sleep deprivation and de-
pression. A hypothesis derived from a model of sleep regulation.
Human Neurobiol._, 1982, 1: 205-210.
Campbell, S.S.: Bimodal distribution of REM sleep latencies in nor-
mal young adults during prolonged bed rest. Sleep Research,
Vol. 12, 1983 (in press).

374
Chouvet, G., Mouret, J., Coindet, J., Siffre, M., Jouvet, M.: Peri-
odicite bicircadienne du cycle veille-sornmeil dans des conditions
hors du temps. Etude polygraphique. Electroencephal.Clin.Neuro-
physiol., 1974, 37: 367-380.
Decoster, F., Foret, J.: Sleep onset and first cycle of sleep in
human subjects: change with time of day. Electroencephal.Clin.
Neurophysiol., 1979, 46: 531-537.
Gillin, C., Duncan, W., Pettigrew, K.D., Frankel, B.L., Snyder, F.:
Successful separation of depressed, normal, and insomniac sub-
jects by EEG sleep data. Arch. Gen. Psychiat., 1979, 36: 85-90.
Insel, T.R., Gillin, J.C., Moore, A., Mendelson, W.B., Loewenstein,
R.J., Murphy, D.L.: The sleep of patients with obsessive-compul-
sive disorders. Arch. Gen. Psychiat., 1982, 39: 1372-1377.
Jus, K., Bouchard, M., Jus, A.K., Villeneuve, A., Lachance, R.:
Sleep EEG studies in untreated long-term schizophrenic patients.
Arch. Gen. Psvchiat., 1973, 29: 386-390.
Kales, A., Tan, T.L., Kollar, E.J., Naitoh, P., Preston, T.A.,
Malmstrom, E.J.: Sleep patterns following 205 hours sleep de-
privation. Psychosom. Med., 1970, 32: 189-200.
Kupfer, D.J.: REM latency: a psychobiologic marker primary depressi-
ve disease. Biol. Psvchiat., 1976, 11: 159-174.
Lund, R., Karnmerloher, A., Dirlich, G.: Body temperature in endo-
genously depressed patients during depression and remission,
in: Goodwin, F.K., Wehr, T.A. (eds.), Circadian Rhythms in
Psychiatry, Boxwood Press, Pacific Grove, Calif., 1983 (in press).
Lund, R., Schulz, H., Berger, M.: K6rpertemperatur und REM-Schlaf
bei affektiven St6rungen, in: D. Vaitl, R. Ferstl, E.R. Rey
(eds.), Klinische Psychologie. Psychophysiologische Merkmale
klinischer Symptome. Bd. 3, Depression und Schizophrenie, Beltz-
Verlag, Weinheim, 1983 (in press).
Meddis, R.: Human circadian rhythms and the 48 hour day. Nature,
1968, 218: 964-965.
Montplaisir, J., Billiard, M., 'Takahashi, s., Bell, I.R., Guillemi-
nault, C., Dement, W.C.: Twenty-four-hour recording in REM-nar-
coleptics with special reference to nocturnal sleep disruption.
~iol. Psvchiat., 1978, 13: 73-89.
Mullaney, D.J., Johnson, L.C., Naitoh, P., Friedmann, J.K., Globus,
G.C.: Sleep during and after gradual sleep reduction. Psvcho-
physiolog~, 1977, 14: 237-244.
Schulz, H., Lund, R., Cording, C., Dirlich, G.: Bimodal distribu-
tion of REM latencies in depression. Biol. Psychiat., 1979, 14:
595-600.
Schulz, H., Lund, R.: Sleep onset REM episodes are associated with
circadian parameters of body temperature. A study in depressed
patients and normal controls. Biol. Psychiat._, (in press).
Schulz, H., Salzarulo, P., Fagioli, I., Massetani, R.: REM latency:
development in the first year of life. Electroencephal. Clin.
Neurophysiol., (in press).

375
Spiegel, R.: Sleep and Sleepiness in Advanced Age, in: Weitzman,
E.D. (ed.), Advances in Sleep Research, Vol. 5, Spectrum Publ.,
Jamaica, N.Y., 1981.
Wehr, T.A., Gillin, J.C., Goodwin, F.K.: Sleep and circadian
rhythms in depression, in: Chase, M.H., Weitzman, E.D. (eds.),
Sleep Disorders: Basic and Clinical Research, Spectrum Publ.,
N.Y., 1983, pp. 195-225.
Weitzman, E.D., Czeisler, C.A., Zimmerman, J.C., Moore-Ede, M.C.:
Biological rhythms and sleep-wake relationships in man of cor-
tisol, growth hormone and temperature during temporal isolation,
in: Martin, J.B., Reichlin, S., Bick, K. (eds.), Advances in
Neurology, Raven, New York, (in press).
Weitzman, E.D., Nogeire, C., Perlow, M., Fukushima, D., Sassin, J.,
McGregor, P., Gallagher, T., Hellman, L.: Effects of a prolon-
ged 3-hours sleep-wake cycle on sleep stages, plasma cortisol,
growth hormone and body temperature. J.clin.Endocrinol., 1974,
38: 1018-1030.

376
SLEEP STRUCTURE AND NEUROENDOCRINE RHYTHMS

David J. Kupfer and David B. Jarrett

Department of Psychiatry, University of Pittsburgh

School of Medicine, Western Psychiatric Institute
and Clinic, Pittsburgh, Pennsylvania

INTRODUCTION:

Current knowledge of affective disorders leads to the con-

clusion that these disorders entail dysfunction in many systems
and do not reflect simply alterations in mood. Since many of the
core symptoms of affective disorder represent changes in
regulatory systems, investigators have frequently searched for
biological correlates, possible markers of vulnerability and ex-
plored the interdependencies among aspects of biological and
psychological functioning. Such an approach has yielded promising
findings with regard to electroencephalographic (EEG) sleep
abnormalities. Furthermore, the study of EEG sleep in affective
states is now affording an opportunity to examine (among other
biological variables) how aberrations in neuroendocrine pathways
and circadian rhythms may be responsible for the various
biological disturbances in depression.

The sleep of a depressed patient is characterized by the

presence of a characteristic cluster of abnormalities (Kupfer et
al., 1978; Kupfer et al., 1982): 1) usually sleep is shortened and
intermittent with a loss of continuity, but in 10-15% of cases (who
are often bipolar) sleep is prolonged and is interrupted less than
is usual; 2) REM latency (the number of minutes asleep until the
onset of the first REM period) is shortened; 3) the distribution of
REM sleep time and activity from the normal patterning is altered,
with a preponderance of REM during the first half of the night; and
finally, a reduction in slow-wave (delta) sleep time and delta wave
density is present (Figure 1). Our current studies demonstrate that
at least one, and usually more, of these abnormalities is found in
90% of all patients with major depressive disorder.

377
 f3AWAt
REM
C!l <E.~
~ 1
il; 2
~ 3+4

1 2 3 4 .s 6 7

HOURS OF SLEEP
EEG SLEEP FEATURES IN DEPRESSION

1• Shortened REM latency

2. Increased REM activity
(1st half of the night)
3. Sleep continuity decrease
4. Delta sleep decrease

Figure 1. Characteristic Sleep Features During Depression

The recent emphasis on slow-wave activity and NREM sleep in

depression also points to the need for integrating data on other
biologic rhythms with data on NREM sleep. In the past decade
increased attention has been devoted to efforts to identify neuro-
endocrine correlates in affective states. Research on the hypo-
thalamic-pituitary-adr enal (HPA) axis has been of particular
importance and reflective of the enthusiasm among biological
investigators to identify neuroendocrine abnormalities in depres-
sion. Such efforts have not been restricted to the HPA axis, but
have involved studies with "sleep-related" hormones, growth hormone
(GH) and prolactin, as well as numerous investigations on the
thyroid (TRH-TSH) axis. While the more systematic earlier neuro-
endocrine studies in depressed patients involved all night neuro-
endocrine measurements, possible relationships with sleep measures
received only perfunctory attention. Indeed, even the findings of
studies which involved 24-hour measurements were "linked" with the
24-hour clock rather than sleep-onset and offset "linkages." More
recently, however, investigators have begun to reassess the
possibilities inherent in examining potential inter-relationships
between various neuroendocrine and sleep measures. In such
analyses it is clear that the role of NREM sleep, especially slow-
wave sleep, may be very important in ascertaining the extent of
association between sleep and neuroendocrine measures both in
normal controls and patients. Therefore, the second half of this
current report will focus on preliminary findings on these
potential associations, first on cortisol, then on prolactin and
growth hormone.

378
 A rather precise 24-hour programming of the cortisol rhythm
relatively independent of the sleep-wake cycle is present in nor-
mals (Krieger, 1979). However, despite the sleep-independent cir-
cadian rhythm some ties to the sleep-wake cycle remain. For ex-
ample, sleep onset is usually associated with a significant de-
crease in cortisol levels. Secondly, it appears that the initial
cortisol rise during the night in normals is correlated positively
with the length of the second non-REM sleep period (Kupfer et al.
1983). Even though the greatest number of cortisol peaks and the
highest concentrations of cortisol are found in the second half of
the night, the initial cortisol rise occurs within four hours of
sleep onset (Jarrett et al., 1983).

Sleep disruption may be associated with increased cortisol

levels throughout the 24-hour cycle. Furthermore, prolonged
awakenings during the night in normal controls are associated with
increased cortisol levels. Perhaps alterations in the initial
cortisol rise may make an individual more susceptible to
dexamethasone suppression test (DST) non-suppression. This may be
especially true since a change in cortisol rise time appears to be
related significantly to sleep onset rather than to clock time
(Jarrett et al., 1983). Our data on depressed patients and
cortisol timing during the night seem to support such a view. For
example, it has been shown that the nocturnal plasma cortisol
concentrations are greater in depressed patients. The average
cortisol nadir for depressed patients demonstrates a significant
increase in cortisol levels as compared to matched normal
controls. The average cortisol rise time after sleep onset is
significantly shorter in these control patients than in the
depressed patients (Jarrett et al., 1983). This earlier cortisol
rise in depressed patients, which is probably related to increased
lengthening of NREM periods during the night (especially between
the first and second REM period) may represent a replicable
biological correlate to be used as a state-dependent correlate
with conjunction in other sleep neuroendocrine measures. While
cortisol levels are restored to normal in remitted depressed
patients, the time between the cortisol rise and sleep onset is
still somewhat shortened in these patients.

As discussed earlier, it is appropriate to investigate the two

"sleep-related" hormones, prolactin and growth hormone, in relation
to EEG sleep which have led us to the tentative conclusion that the
initial nocturnal increases in prolactin levels usually occur after
NREM periods both in normals and in depressed patients (Kupfer, et
al., 1983). More specifically, the initial prolactin rise after
sleep onset is associated with the transition of NREM sleep into
REM sleep at the early segments of the sleep period (either first
or second NREM period in normals and the first NREM period in
depressives). Thus, the shift in prolactin secretion reported in
depressed patients may be associated with concurrent sleep altera-

379
tions (such as the abbreviated first NREM sleep period). Further-
more, the earlier NREM periods are more likely to be associated
with changes in delta wave density.

In contrast, alterations in growth hormone in depressed

patients may be more clearly a function of delay of onset and/or
definite decreases in Stages 3 and 4 sleep, especially as reflected
in delta-wave intensity measures. In two preliminary series of 12
normal subjects, growth hormone peaks during sleep were consis-
tently associated with a marked increase in delta-wave activity
(Figure 2). However, it is important to emphasize that delta-wave
activity elevations can be present without a growth hormone peak.
In other words, delta-wave activity elevations appear to be a
necessary, but not a sufficient, condition in which growth hormone
elevations occur. In support of this proposition in normals is the
following data: ten of the 12 normals had an obvious set of delta
activity peaks in the first NREM period of the night. Of these ten
normals, eight also showed a clear growth hormone peak. One normal
subject did not demonstrate either a delta activity peak or a
growth hormone peak. Another normal subject showed poor delta
activity levels in the first NREM period but experienced a growth
hormone peak concurrent with high delta activity during the second
NREM period.

In contrast to the associations present between delta wave

sleep and growth hormone in the normals, growth hormone secretion
appears to have a more complicated pattern in depressed patients.
Of the first eight depressed patients studied, only three patients
showed considerable delta activity levels and two of these patients

~~0~18yrW. studyllatcAuo10
Dn19F,_

DELTA
counts

AWNC£1
STI«.I/REW

STAGE 2

STAGE 3/4

21:00-!-:--:22::-':00::--=23:00~·:--:2-:-o._,oo::---,:OO,-,-2::c,oo::----:J:-o::oo:--4:00::c•~--:s:oo-o::•-e::c:00~-::-!.,00
a.oocl'IIE

Figure 2. Relationship Between A Normal Control With Early Growth

Hormone Peaks During The First NREM Sleep Period

380
also had good growth hormone peaks. The other five depressed
patients showed reduced delta activity levels in the first NREM
period and no associated growth hormone peaks. Finally, one
patient did demonstrate marked delta activity levels in the second
NREM period associated with a growth hormone peak. Therefore, it
appears that a growth hormone elevation in the first NREM period is
associated with a marked delta activity level in the first NREM
period; one could therefore speculate that a good delta peak
appears to be a necessary but not sufficient condition for the ap-
pearance of growth hormone peaks in depressed patients as well as
for normal controls. Furthermore, it appears that the growth
hormone rise is not associated with the timing of the first REM
period.

It is appealing to conclude that abnormalities in growth

hormone in depressed patients may be a function of delay or
decreases in Stages 3 and 4 sleep, especially as reflected in delta
wave intensity measures. If this is true, then the two process
model of sleep regulation with its emphasis on process S as
reflected by delta wave density may be useful in understanding
sleep-neuroendocrine interrelationships in depression. Since it
would be reasonable to propose that growth hormone is positively
correlated with delta wave density and both prolactin and cortisol
levels are inversely correlated with delta wave density, an
impaired process S in depressive states could help explain the all-
night neuroendocrine changes in depression. Thus, our findings
suggest an interesting yet complicated set of relationships in the
slow-wave sleep cycle between both REM and NREM sleep and neuro-
endocrine activity. We have already suggested that the depressive
state is associated with a "flattening" or diminished amplitude in
hormone rhythms, and this flattening may account for the abnormali-
ties in cortisol, prolactin and growth hormone secretion noted in
such individuals (Kupfer & Jarrett, 1983). Although considerable
effort has been devoted to relating these neuroendocrine abnormal-
ities to disturbances of REM sleep, we have reservations about such
an overemphasis and propose an alternative set of explanations.
First, it is now well-accepted that peak secretion of growth
hormone and Stages 3 and 4 sleep are closely related. Indeed, as
our preliminary data suggest an even closer tie between growth
hormone changes and delta-wave activity may be present if quanti-
tative techniques are applied. Second, prolactin changes during
the night may be inversely related to the intensity of delta-wave
activity, especially in the early part of the sleep period. The
association of prolactin secretion with the onset of REM period may
represent a relationship with the rapid "falling-off" of delta wave
intensity, an occurrence which is present frequently. Third, with
respect to cortisol, data on follow-up studies in depressed
patients are demonstrating an inverse relationship between cortisol
and delta-wave intensity.

381
 Evidence is accumulating to support the concept that shifts
occur in the normal cycle of the sleep and neuroendocrine rhythms
and that these disturbances may persist following clinical recovery
from the illness. Furthermore, such abnormalities may serve as "trait"
or non-episodic characteristics for the illness and thereby signify
a level of vulnerability. Strategies of longitudinally following
these measures in remitted recurrent depressives may also allow the
application of pharmacologic probes during remission when one may
speculate that rather the "blunting" associated with episodic changes,
"hyperresponsivertess" may be associated with trait phenomena.

REFERENCES
Jarrett, D. B., Coble, P. A., and Kupfer, D. J., 1983, Reduced
cortisol latency in depressive illness. Arch. Gen. Psychiat.,
40:506-511.
Krieger, D. T., 1979, Rhythms in CRF, ACTH and corticosteroids in
endocrine rhythms, in: "Comprehensive Endocrinology Series,
Endocrine Rhythms," D-:-T. Krieger, ed., Raven Press, New York.
Kupfer, D. J., Bulik, C., and Jarrett, D. B., 1983, Nighttime
plasma cortisol secretion and EEG sleep- are they associated?
Psychiat. Res., 10:191-199.
Kupfer, D. J., Foster, F. G., Coble, P. A., McPartland, R. J., and
Ulrich, R. F., 1978, The application of EEG sleep for the
differential diagnosis of affective disorders. Am. J.
Psychiat., 135:387-396.
Kupfer, D. J., and Jarrett, D. B., 1983, Sleep-neuroendocrine
interrelationships in affective disorders. Psychopharma Bull.,
9:479-481.
Kupfer, D. J., Shaw, D. H., Ulrich, R. F., Coble, P. A., and
Spiker, D. G., 1982, Application of automated REM analysis in
depression. Arch. Gen. Psychiat., 39:569-573.
ACKNOWLEDGEMENT
This work was supported in part by NIMH Grants MH #24652 and MH
#30915 as well as a Grant from the John D. and Catherine T.
MacArthur Foundation Research Network on the Psychobiology of
Depression.

382
NEUROENDOCRINOLOGICAL FINDINGS IN DEPRESSION COMPARED WITH FINDINGS

IN MALNUTRITION, CUSHING'S SYNDROME AND EMOTIONAL STRESS

M. Berger, P. Doerr, J.-Ch. Krieg, K.-M. Pirke and

D. von Zerssen

Max-Planck-Institute of Psychiatry
Kraepelinstr. 10, D 8000 Munich 40, FRG

In the last 20 years there have been extensive publications on

neuroendocrine abnormalities in depressive disorders. In recent years
the main interest has focused on the hypothalamo-pituitary-adrenal
system and the dexamethasone suppression test (DST). Just since 1979
about 150 studies regarding the relevance of the DST for psychiatry
have been published.
Despite this amount of scientific work and despite published
recommandations to use, for example, an abnormal DST in clinical 1 2
routine, as a specific biological marker for endogenous depression '
the following questions have not been clarified:
First, is there really any single neuroendocrine abno:nnality which
is specific for depression in general or a subgroup of de-
pression?
Second, is there a typical pattern of neuroendocrine disturbances
in depression or a depressive subgroup?
Third, do the neuroendocrine abnormalities in depression reflect a
central transmitter defect also underlying the clinical symp-
toms ~f 4 the disorder, a hypothesis at present generally fa-
vored ' , or are they induced by the depressive psychopathol-
ogy and/or its behavioral consequences, as insomnia, agitation,
psychic pain, weight loss or stress caused by admission to
hospital 5?
Regarding the first question we will demonstrate our own data con-
cerning the DST, which is considered to be the single test for endo-
genous depression with the highest specificity. Regarding the second
question, studies about combined neuroendocrine challenge tests on
various hormonal axes will be discussed by means of results from the
literature. Finally the question will be readressed whether the
neuroendocrine findings may be the direct consequences of a specific
central transmitter disturbance or whether they may be better under-

383
stood within the more general framework of a multifactorial process.
Regarding the 1.5 or 1 mg DST with post-dexamethasone blood
samples drawn at least at 4 and 11 p.m. in 75 healthy subjects and
225 psychiatric inpatients, the following results were obtained: The
test was positive in approximately 13% of the healthy probands that
means at least one plasma value exceeded 5 ~g/dl. In 83 endogenous
depressives als well as in 64 neurotic or situative depressives, 36%
of the patients showed a positive DST. In 78 patients suffering from
non-depressive psychiatric disorders such as mania, schizophrenia or
non-depressive neuroses, 31% were non-suppressors. In essence, these
data contradict the results of several other authors about a high
specificity of the tert 2for depression in general and for endogenous
subtype in particular ' . However, the results arg ~n good agreement
with recent results from the literature (overview ' ). KLEIN has
just described in an extensive, up to date review of the published
DST data, that of 1713 depressed patients 35% revealed non-suppres-
sion, whereas out of 427 patients with other psychiatric disorders
about 30% were non-suppressed. Within depressive disorders KLEIN re-
ported, however, a higher percentage o7 positive DSTs in endogenous
compared to non-endogenous depressives .
Our studies demonstrated a decisive influence of a lot of inter-
vening variables on the DST regardless of the diagnostic classifi-
cation:
1. In 71 psychiatric patients with a variety of depressive or non-
depressive disorders the influence of hospital admission on the
1 mg DST was examined. Whereas the DST was positive in 45% of the
71 patients immediately after admission, the percentage of abnormal
DSTs decreased to 32% when the test was repeated 7 to 10 days later
(max. cortisol values, Wilcoxon-test for correlated samples: pc 0.05).
The ps:}Chopathological state did not improve sign~ficantly during this
time according to self-rating-scales (Bf-S/Bf-S' ).
2. In 92 patients immediately after admission the influence of with-
drawal of psvchoactive druas and/or alcohol on the 1 mg DST was l.n-
vestigated. In 34 subjects hospita~admission coincided with with-
drawal. In 62% of these patients the DST was abnormal. In those 58
subjects without withdrawal the test was abnormal only in 38%. This
difference was statistically significant (X.2. :p <~: 0. OS).
3. Furthermore we focused our attention on the relationship between
an abnormal 1 mg DST and preceeding suicidal attempts. Of 93 patients
11 had attempted suicide immediately prior to admission. Ten of
these 11 patients had an abnormal DST (x~:p< 0.01). Nine of these
had a situational depression without endogenous features. Whether
this high percentage is due to inner turmoil, drug withdrawal or
other factors is unclear.
4. In 45 depressed patients whose body weight were precisely docu-
mented in the week preceeding the DST, it was found that out of 13
patients who lost weight during that week 54% were non-suppressors
after 1.5 mg DST, whereas patients without weight loss were non-sup-
pressors only in about 7% (.x}:p<' 0.001) ~. Focusing on the question
of the influence of weight loss, we studied the effect of semifast-

384
ing on the DST in 24 healthy probands with normal body weights, who
had demonstrated two normal baseline DSTs. It could be shown that a
daily diet of 1000 to 1300 calories and a related weight loss of
1.5 kg per week provoked DST non-suppression in nearly 40% of the
subjects within a period of two weeks. In summary our studies suggest
that thg specificity of the DST for endogenous deprgssion is clearly
limited . Intervening variab~e~ 1~ch as weight loss or stress
caused by hospital admission ' ., seem to have a decisive influence
on the test results.
The presentation of our data on the DST is not complete without
mentioning a methodological difficulty at least of the 1 mg DST. In
160 DSTs with 1 mg we investigated the dexamethasone serum levels
at 9 a.m. to verify the compliance. While it was shown that all
patients had complied, the dexamethasone concentration was signifi-
cantly lower in the non-suppressors than in the suppressors~ This
result rises the question how far at least an abnormal 1 mg DST re-
flects hypercortisolism or in as much the result is also influenced
by the degree of resorption and clearance of dexamethasone. This
question requires further inquiry. w (u-test, p< 0.0001)
The second question of our presentation concerning a specific
pattern of neuroendocrine abnormalities in depression has been in
the focus of numerous investigations on various hormonal axes. Re-
viewing this research, it seems necessary to limit the discussion
to those studies with generally consistent findings. It should be
mentioned, however, that contradictory results have been published
on almost all findings. Regarding the neurofydocrine axes, the
following abnormalities have been described :
a) Hypothalamo-pituitary-adrenal (HPA-) system:
-increased amount of free cortisol and 17-0HCS in the 24h-urine,
- increased rate of cortisol production,
- a flattened 24 h cortisol profile in the serum,
lack of cortisol suppression in the DST
- blunted· response of co.rtisol to a hypoglycemic stimulus.
b) Growth hormone (hGH-) system:
- blunted response of hGH to a hypoglycemic stimulus, to (methyl-)
amphetamine, or clonidine.
c) Hypothalamo-pituitary-thyroid (HPT-) system:
- blunted response of TSH to TRH.
Investigations of the different neuroen?~Cf~nf 4 af5s, such as those
concerning the HPA- and thg rGT-systems ' ' ' , the HPA- and
the growth hormone systems ' , those of WINOf~is group concerning
the HPA-, HPT- and the growth hormone systems ' show an increased
number of abnormal, single test results, but no systematic relation-
ships among the tests. In other words, a consistent pattern of neuro-
endocrine disturbances has not been observed. In contrast to the
large number of unrelated and contradictory findings, two results
are essentially constant: - OrH~ is the activation of the HPA-system
in depression, - the other is that neuroendocrine abnormalities
thus far observed in ?§pression have, when studied, also been found
in Cushing's syndrome . Beside the alterations found in connection

385
with the hypercortisolism in Cushing's syndrome there is also a
blunted response of hGH and cortisol to a hypoglycemic stimulus and
of TSH to TRH. Corresponding changes have at§o been observed in the
therapy with synthetic glucocorticosteroids . This rises the ques-
tion whether the spectrum of neuroendocrine abnormalities in depres-
sion may be a consequence of hypercortisolism. This hypothesis would
allow a more comprehensive ~iew of the variety of results, which ap-
pear to be unrelated by now . The possible relevance of this hypo-
thesis is supported by investigations of KRIEGER in Cushing's syn-
drome, where she could demonstrate that the disturbances within
other hormonal axes caused by hypercortisolism may show a delayed
recovery after the elevated cortisol production had been corrected.
Regarding this observation, the different degree, durations and
fluctuations of hypercortisolism,in addition to the individual vary~
ing sensitivities of the other axes to an elevated cortisol level,
would explain the different and contradictory neuroendocrine results
in depression. Therefore, longitudinal studies of the HPA system
seem to be necessary to evaluate its influence on challenge tests
of the other hormonal systems.
Regarding our initial questions the results discussed so far
support the following conclusions:
I.) Up to now, neither a single neuroendocrine abnormality nor a
pattern of neuroendocrine disturbances has been proven to be
specific for depression or a depressive subgroup.
II.) Neuroendocrine studies have not yet convinciously demonstrated
a specific disturbance of central transmitter systems as the
pathogenetic basis of depressive disorders.
III.) There are many hints to a dominant role of hypercortisolism
within neuroendocrine disturbances in depression.
IV.) Our own data regarding the DST in psychiatric disorders indi-
cate that hypercortisolism is induced by the depressive psycho-
pathology and/or its behavioral consequences. Therefore, hyper-
cortisolism in depression has to be understood as being a
multifactorial phenomenon.
We will highlight only some of 6 these factors very briefly: (1) One
factor seems to be weiqht loss . This may explain the neuroendocrine
similarities between depression and anorexia nervosa, a topic which
will be discussed by Dr. Brambilla during this symposium. (2) A
second factor seems to be Siuuational stress. Our examination of
the DST indicates that stress due to hospital admission is chara-
terized by a marked activation of the HPA-system in these pati9nt5'
a result also confirmed by KENNETH DAVIS' and CARROLL's groups ' .
This suggesrethat even other kinds of situational stress due to,
e.g., partner conflicts, job or financial problems can also be part
of the reason for the activation of the HPA-system during depression
and other disorders. (3) A third factor seems to be stress induced
bv the depressive symptomatology itse}f- The findings of the DST in
patients with psychotic depression 20 • or after a suicidal attempt
suggest that it may be worthwhile to reconsider an earlier, now un-
popular concept which proposes that the depressive illness itself

386
represents a severe distress for the patients 22 • 23 Depressed
patients are suffering from helplessness, severe feelings of insuf-
ficiency, guilt feelings, unresolved conflicts ('silent sufferers'),
self-directed aggression and suicidal tendencies, which may corre-
spond to severe chronic distress. As a large number of daily stress
situatio~~ have been proven to stimulate the EPA-system in healthy
subjects , it would be surprising if this severe emotional stress
did not frequently lead to an activation of the EPA-system in de-
pressives.
Th~s presentation of factors probably~influencing the fWA activ-
ity is essentielly not exhaustive. Nonetheless, it may indicate the
necessity to take into consideration the complexity of neuroendo-
crine regulations when studying the neuroendocrinology of depression.

REFERENCES

1. B. J. Carroll, M. Feinberg, J. F. Greden, J. Tarika, A. A. Albala,

R. F. Haskett, N. Mci. James, z. Kronfol, N. Lohr, M. Steiner,
J.P. de Vigne, and E. Young, A specific laboratory test for
the diagnosis of melancholia, Arch. Gen. Psychiat., 38:15-20
(1981)
2. N. H. Kalin, S. c. Risch, D. S. Janowsky, and D. L. Murphy, Use
of the dexamethasone suppression test in clinical psychiatry,
J. Clin. Psychopharmacol., 1:64-69 (1981)
3. E. J. Sachar, Hormonal changes in stress and mental illness, in:
D. T. Krieger, and J. c. Hughes (Eds.) Neuroendocrinology,
Sinauer Assoc. Inc., Sunderland (1980)
4. B. J. Carroll, J. F. Greden, R. Haskett, M. Feinberg, A. A.
Albala, F. I. R. Martin, R. T. Rubin, B. Heath, P. T. Sharp.
W. L. McLeod, and M. F. McLeod, Neurotransmitter studies of
neuroendocrine pathology in depression, Acta Psychiat. Scand.
Suppl., 280:183-200 (1980)
5. D. von Zerssen, M. Berger, and P. Doerr, Neuroendocrine dys-
function in subtypes of depression, in: N. s. Shah, and A.
G. Donald (Eds.) Psychoneuroendocrine Dysfunction in Psychi-
atric and Neurological Illnesses, Plenum Publ. Co., New York,
in press
6. M. Berger, P. Doerr, R. Lund, T. Bronisch, and D. von Zerssen,
Neuroendocrinological and neurophysiological studies in major
depressive disorders: Are there biological markers for the
endogenous subtype?, Biol. Psychiat., 17:1217-1242 (1982)
7. H. E. Klein, The dexamethasone-suppression-test in psychiatry:
Myth and Realities, Advances in Psychopharmacol., in press
8. D. von Zerssen, with assistance of D. M. Koeller, Die Befind-
lichkeitsskala, Beltz, Weinheim (1976)
9. E. F. Coccora, J. Prudic, A. Rothpearl, H. G. Nurnberg, and K.
L. Davis, The effect of hospitalization on the DST. Abstract
82; presented at the Annual Meeting of the Society of Bio-
logical Psychiatry, New York (1983)

387
10. R. F. Haskett, P. z. Athanasios, A. A. Albala, and B. J. Carroll,
DST performance during first 48 hours of admission, Abstract
100; presented at the Annual Meeting of the Society of Bio-
logical Psychiatry, New York (1983)
11. G. F. s. Johnson, Endocrine dyfunction in depression, in: P. J.
v. Beumont, and G. D. Burrows (Eds.) Handbook of Psychiatry
and Endocrinology, Biomedical Press, Amsterdam, New York, Ox-
ford, p. 239-266 (1982)
12. A. A. Albala, L. Grunhaus, B. J. Carroll, and A. Arbor, Neuro-
endocrine identification of melancholia, Abstract 31; pre-
sented at the Annual Meeting of the American Psychiatry As-
sociation, New York, (1983)
13. S. D. Targum, A. C. Sullivan, and s. M. Byrnes, Neuroendocrine
interrelationships in major depressive disorder, Am. J.
Psychiat., 139:282-286 (1982)
14. c. Kirkegaard, H. Aggernaes, B. Kijne, I. Krog-Meyer, B.-M. Niel-
sen, c. B~rup, P. c. Eskildsen, and J. Kvist, The hypotha-
lamic-pituitary-thyroid axis in depression and mania, in: c.
Perris, G. Struwe, and B. Jansson (Eds.) Biological Psychiatry
1981, Biomedical Press, Amsterdam. New York , Oxford, p. 317-
320 (1981)
15. I. Exstein, A. L. c. Pottash, and M. s. Gold, Relationship of
thyrotropin releasing hormone test and dexamethsaone suppres-
sion test abnormalities in unipolar depression, Psychiat.
Res. 4:49-53 (1981)
16. M. Feinberg, and B. J. Carroll, Diagnostic tests series and
parallel, Abstract 34; presented at the Annual Meeting of the
Society of Biological Psychiatry, New York (1983)
17. A. Wjnokur, J. Amsterdam. S. Caroff, P. J. Snyder, and D. Bruns-
wick, Variability of hormonal response to a series of neuro-
endocrine challenges in depressed patient, Am. J. Psychiat.,
139: 39-44 (1982)
18. J. D. Amsterdam, A. Winokur, I. Lucki, S. Caroff, P. J. Snyder,
and K. Rickels, A neuroendocrine test battery in bipolar
patients and healthy subjects, Arch. Gen. Psychiat., 40:515-
521 (1983)
19. D. T. Krieger, Cushing's Syndrome, Monographs on Endocrinology,
Vol. . 22, Springer, Berlin, Heidelberg, New York (1982)
20. M. v. Ruhdorfer, H.-G. Hwu, and P. J. Clayton, Dexamethasone
suppression test in primary depression: Significance of family
history and psychosis, Biol. Psychiat., 17:41-48 (1982)
21. J. Mendlewicz, G. Charles, and J. M. Franckson, The dexametha-
sone suppression test in affective disorder: Relationship to
clinical and genetic subgroups, Brit. J. Psychiat., 141:464-
470 (1982)
22. E. J. Sachar, Corticosteroids in depressive illness, I. and II.
Arch. Gen. Psychiat., 17:544-533 and 554-567 (1967)
23. J. W. Mason, Clinical psychophysiology: Psychoendocrine mecha-
nisms, in: S. Arieti (Ed.) American Handbook of Psychiatry,
Basic Books, New York, pp 143-181 (1975)

388
ACTH AND CORTISOL SECRETION IN CUSHING'S DISEASE AND IN

ENDOGENOUS DEPRESSION: INDICATION OF A COMMON PATHWAY?

Karl H. Voigt, Sabine Bossert, Stephan

Bretschneider, Anette Bliestle and HorstL. Fehm

Lab. for Neuroendocrinology and Dept.of

Psychiatry, University of Ulm, D-7900 Ulm
Germany

INTRODUCTION

An impaired regulation of cortisol secretion can

be observed in patients suffering from one of two
otherwise unrelated diseases: The hypothalamo-pituitary-
dependent Cushing's Syndrome and a subpopulation of major
depressive illness. Among a number of different other
disturbances of neuroendocrine regulations (Table 1)
the cortisol-hypersecretion is the most important and
the typical diagnostic sign for Cushing's disease, and
on the other hand, the most consistently observed hor-
monal dysregulation in endogenously depressed patients.

The term Cushing's Syndrome is used for patients

who exert the physical signs of hypercortisolism regard-
less their different causes 1,2:
(i) Approximately 50 to 70% of cases are due to an in-
appropriate elevated ACTH/B-endorphin-secretion of the
pituitary gland (Cushing's disease, C.D.). The basic
disturbance is thought to be a diencephalic imbalance
of neurotransmitter/neuromodulator induced secretion of
the hypothalamic Corticotrophin-Releasing-Hormone (CRH),
which in turn leads to ACTH-hypersecretion often pro-
duced by pituitary microadenomas3. Because of the
CNS-origine of the cortisol-hypersecretion in C.D. and
their high incidence (40-60~of depression4 we examined
only these patients for the present study.
(ii) Hypersecretion of cortisol and other adrenocorti-
cal steroids is caused by adenomas o1 ~eoplasms of
adrenal tissue in 20-30% of the cases ' . Their pitui-

389
tary ACTH secretion is very low or undetectable.
(iii) The ectopic production (by carcinomas) of ACTH,
producing bilateral adrenal hyperplasia is seen in
approximately 10%.
(iv) A different, frequently occurring type of Cushing's
Syndrom is caused by chronic glucocorticoid administra-
tion.

Affective abnormalities are observed in 30 to 60%

of all'patients with Cushing's Syndrome including the
drug induced form (iv). However, only in C.D. patients
(i) the mood change is predominantly depressive, where-
as the patients belonging 4 t~ the other groups (ii-iv)
exert more manic symptoms ' .

The impaired glucocorticoid-feed-back-mechanism in

depressive patients, usually tested by the dexametha-
sone-suppression (DST)7,8,9 is obviously very similar
to the basic disturbance in C.D. Recently, we could
show, that the disturbance of cortisol-feed-back-mecha-
nism in C.D. is maintained even after surgical correc-
tion (e.g. adrenalectomy) of the hypercortisolism10.
In these patients the usually negative feed-back mecha-
nism is converted into a positive one, e.g. the
infusion of cortisol was followed paradoxically by a
marked elevation 9f ACTH during the first phase of the
steroid-feed-back . Furthermore, the feed-back-mecha-
nism is clearly influenced by brain monoamines, indica-
ted.by t~e T~f~ct of cyproheptadien14, reserpine12 and
deslpramlne ln man.

In regard to the above mentioned reports we scruti-

nized in the present study the question: Are there any
indications for a pathogenetic link between the
disturbed cortisol-feed-back-mechanism and dimensions
of depressive illness? In other words, do C.D. patients
exert depressive symptoms even if they do no more
suffer from hypercortisolism, but still from a CNS
mediated disturbance of cortisol-feed-back-mechanism?

SUBJECTS and METHODS

17 patients (14 females, 3 males) with Cushing's

disease, age range frbm 16 to 63 years, mean age ! S.D.
42.0 ± 13.5 were examined. The endocrinological diagno-
ses of all patients was assertained by the appropriate
procedures 1 and by the microscopical examination of the
removed tissue including specific immunocytochemical
reactions3. Their illness was treated by neurosurgery,
mostly microadenomectomy, of the pituitary (7 cases),

390
by bilateral adrenalectomy (7 cases) or by both methods
(3 cases). Most of the patients were already tested for
the occurrence of a paradoxical feed-back-mechanism 1 11
including the influence of neuropharmacologic drugs .
The surgery was performed about 5 years before this
study. All patients, except one, got a replacement
therapy of 25 to 35 mg cortisol/day.

We examined in parallel 13 indoor patients suffering

from psychiatric diseases, mostly depression, by both
a number of neuroendocrine tests and psychological pro-
cedures. Identical examinations, conducted by the same
psychiatrist and psychologist were used for C.D. pa-
tients: (1) Hamilton Depression-Rating-Scale containing
21 items. (2) Semi-standardized interview which con-
sisted in two parts: (a) item scales measuring social
interactions, depressive symptoms, family history for
psychiatric or endocrinologic disturbances, stress-
coping-skills including normal life events and the par-
ticular disease. (b) Non-verbal-behavior ratings.
(3) Self rating-scale for somatic complaints with 24
items according to v. Zerssen16. (4) Self f'ting-scale
for paranoid and depressive symptoms (PDS) 18
(5) Amnestic report. (6) Attention test (d2)

The C.D. patients were examined during a routi-

neously sheduled visit to the endocrinological ambu-
lance.

RESULTS and DISCUSSION

Among the vast number of tested items in the

different test systems, there was no indication for
depressive symptoms in the examined 17 patients with
C.D. (Hamilton Score: 4.2 ± 4.9; PDS-D-Scale: 54.7 ±
11.5; Somatic complaints: 58.7 ± 10.8). It is important
to remember, that all patients have been successfully
treated by surgical correction of their hypercortiso-
lism. This clear result was rather unexpected by us, in
view of the preexisting depression and severe life
events caused by the C.D. and its treatment. It does
not only mean, that the correction of hypercortisolism
is followed or in some cases paralleled by improvement
of depressive symptoms19,20,21, but that even the long
history (6.5 years) of the illness including high rise
surgical operations, the vital nec~ssity of replacement
therapy and a number of still existing somatic com-
plaints (osteoporosis, reduced sexuality) and a wor-
sened social situation were not able to induce and/or
maintain depressive behavior. Amnestic and hospital

391
Table 1: Clinical findings in patients with Cushing's disease
(hypothalamo-pituitary depending Cushing Syndrom with
bilateral hypertrophia of the adrenal gland) compared to a
subpopulation of endogenously depressed patients, charac-
terized by an impaired cortisol secretion.
CUSHING'S DEPRESSION
NEUROENDOCRINOLOGY DISEASE
Corti sol
Circadian Rhythm impaired impaired
Content (Plasma,Urin) increased increased
DST (1 mg) impaired impaired
Insulin induced hypo- blunted blunted
glycemia response response
ACTH/B-endorphin increased ?
Growth Hormone (GH) increase
after TRH
Gonadotropins (LH/FSH) blunted ?
Prolactin stimulation increased?
Tyroid Stimulating blunted TRH-
Hormone (TSH) response

CLINICAL SYMPTOMS Hypercorti- no symptoms of

solism hypercortisolism
Depression Endogenous
Depression

NEUROPHARMACA Serotonin-Anta- Noradrenaline-

gonists and Serotonin-
Reserpine uptake-Blockers
Dopamine-Ago-
nists

reports showed, that during the hypercortisolism (it is

before surgery), in eight patients depressive symptoms
were diagnosed (not by the criteria of this study) and
treated in six with antidepressive drugs. In two pa-
tients who have been adrenalectomized after an unsuc-
cessful neurosurgery, depression persisted for about
one year after the operations.
Our results do not confirm the speculation that the
CNS mediated disturbance of the cortisol-feed-back-me-

392
chanism11 is per sea pathogenetic link to depressive
mood changes. It is to our knowledge the first examina-
tion of C.D. patients after correction of their hyper-
cortisolism, thus allowing to test the mood changes due
to the still existing CNS-defect11, which is otherwise
hidden 19y high cortisol levels. According to Jeffcoate
et al. elevated cortisol during a rather long peroid
is essential for the development of depression in c 1 ~.
When the patients have been treated with metopirone to
reduce the cortisol secretion, the depressive symptoms
were diminished simultaneously. Similar results were
obtained by reserpine induced reduction of cortisol in
C.D.22. However, chronically elevated cortisol levels
do not usually induce depressive mood change4,6,21, and
the every day clinical experience in glucocorticoid
treated patients rather point to manic mood alterations.
Thus, in C.D. the diencephalic disturbance is able to
induce depressuve behavior in 30 to 60% of the patie~crs
only together with hypercortisolism. In some reports '
23, a premorbid psychopathology of C.D. patients is
suggested to be essential for their depression. In con-
trast, we were unable to observe any affective disorders
in the treated C.D. patients.

The coincidence of many factors including dis7urbed

cortisol feed-back-mechaoism 8 and hypercortisolism ,
duration of the illness 13 , 24 , scoping of stressful
situations9 and some unknown components (e.g. brain
monoamines, genetics, neuroendocrine dysregulation,
body weight loss9) may be responsible for the simulta-
neous occurrence of depression and hypercortisolism.

REFERENCES

1. H. L. Fehm and K. H. Voigt, Pathophysiology of

Cushing's Disease. in: "Pathobiology Annual",
H. L. Joachim, ed., Raven Press, New York (1979)
2. D. T. Krieger,"Cushing's Syndrom," Springer, Berlin
(1982)
3. R. Martin, Y. Cetin, H. L. Fehm, R. Fahlbusch, and
K. H. Voigt, Multiple cellular forms of cortico-
trophs in surgically removed pituitary adenomas
and periadenomatous tissue in Cushing's disease.
Am. J. Pathol. 106:332 (1982)
4. B. J. Carroll, Psychiatric disorders and steroids.
in: "Neuroregulators and Psychiatric Disorders,"
E. Usdin, D. A. Hamburger, J. D. Barchas, eds.,
Oxford Univ. Pr. (1977)

393
 5. K. H. Voigt, H. L. Fehm, and E. F. Pfeiffer, Disso-
ciation of in vivo and in vitro autonomy in a
human adrenocortical tumour. Acta endocr. 78:302,
( 1 97 5)
.6. P. C. Whybrow, and T. Hurwitz, Psychological distur-
bances associated with endocrine disease and hor-
mone therapy, in: "Hormones, Behavior, and Psy-
chopathology," E. J. Sachar, ed., Raven Press,
New York (1976
7. E. J. Sachar, Neuroendocrine abnormalities in depres-
sive illness. in: "Topics in Psychoendocrinology"
E. J. Sachar, ed., Grune & Stratton, New York
(1975)
8. B. J. Carroll, M. Feinberg, J. F. Greden, J. Tarika,
A. A. Albala, R. F. Haskett, N. Mci.James, Z.
Kronfold, N. Lohr, M. Steiner, J. P. de Vigneand
E. Young, A specific laboratory test for the dia-
nosis of melancholia, Arch. qen. Psvchiat. 38:
15 (1981)
9. M. Berger, P. Doerr, Ch. Krieg, K. M. Pirke, D. v.
Zerssen, Neuroendocrinological findings in de-
pression. This volume.
10. H. L. Fehm, K. H. Voigt, R. Lang, K. E. Beinert,
G. W. Kummer, and E. F. Pfeiffer, Paradoxical
ACTH response to glucocorticoids in Cushing's
disease, N. Enql. J. Med. 297:904 (1977)
11. H. L. Fehm, K. H. Voigt, G. N. Kummer, and E. F.
Pfeiffer, Positive rate-sensitive corticosteroid
feed-back mechanism of ACTH secretion in Cushing's
disease, ~- rlin. Invest. 64:102 (1979)
12. H. L. Fehm, R. Steck, J. Hohnloser, K. H. Voigt, and
E. F. Pfeiffer, Influence of neuroactive drugs
on corticosteroid feed-back regulation of ACTH
secretion in man, Horm. Metab. Res. 15:29 (1983)
13. K. H. Voigt, S. Bossert, St. Bretschneider, B. Rock-
stroh, and H. L. Fehm, Disturbance of cortisol
secretion in depressive patients: lack of corre-
lation with plasma ACTH, in: "Integrative Neuro-
humoral Mechanism," L. Angelucci, D. de Wied,
E. Endroczi, U. Scapagnini, eds., Elsevier,
North Holland, Amsterdam, in press
14. D. T. Krieger, L. Amorosa, and F. Linick, Cyprohep-
tadine-induced remission of Cushing's disease,
N. Enql. J. Med. 293:893 (1975)
15. M. Hamilton, Rating scale for depression, J. Neurol.
Neurosurq. Psychiat. 23:56 (1960) ·
16. D. von Zerssen,"Klinische Selbstbeurteilungs-Skalen
(KSb-S) aus dem Mlinchner Psychiatrischen Infor-
mations-System, (PSYCHIS Mlinchen). Die Beschwer-
denliste", Beltz, Weinheim (1976)

394
17. ibid, "Paranoid-Depressivittits-Skala und Depressivi-
ttits-Skala", Beltz, Weinheim (1976)
18. R. Brickenkamp,"Der Aufmerksamkeits-Belastungs-Test/'
Hogrefe, Gottingen (1972)
19. W. J. Jeffcoate, J. T. Silverstone, c. R. w. Edwards,
and G. M. Besser, Psychiatric manifestations of
Cushing's Syndrome: response to lowering of plas-
ma cortisol, Quarterly Journal of 1-iedicinP- 48:465
(1979) -
20. S. Gifford and J. G. Gunderson, Cushing's disease
as a psychosomatic disorder, Medicine 49:397
(1970)
21. M. N. Starkman, D. E. Schteingart, and M. A. Schork,
Depressed mood and other psychiatric manifesta-
tions of Cushing's Syndrome: relationship to hor-
mone levels, Psychosomatic 1-iedicine 43:3 (1981)
22. H. L. Fehm, K. H. Voigt, and E. F. Pfeiffer, Die Be-
deutung des Zentralnervensystems in der Xtiologie
des Morbus Cushing, in:"Pathologische Erregbar-
keit des Nervensystems und ihre Behandlung,"
H. G. Mertens, H. Przuntek, eds., Springer, Ber-
lin ( 1 980)
23. Q. R. Regestein, L. I. Rose, and G. H. Williams,
Psychopathology in Cushing's Syndrome, Arch.
In tern. Med . 1 3 0 : 11 4 ( 1 9 7 2 ) -
24. G. M. Asnis, E. J. Sachar, U. Halbreich, R. Swami
Nathan, H. Novacenko, and L. c. Ostrow, Cortisol
secretion in relation to age in major depression,
Psvchosomatic Medicine 43:235 (1981)

ACKNOWLEDGEJI.lENT

The study was supported by Deutsche Forschungsge-

meinschaft (SFB 87, Projekt B2 and B13; Heisenberg-Sti-
pendium for K.H.V.).

395
EFFECT OF NITROUS OXIDE ON DEPRESSIVE PATIENTS AND VOLUNTEERS

M.A. Gillman 1 - Univ. Wits. Med. School, Johannesburg

N. Matussek - Psychiatry Hospital, Univ. Munich F.R.G.
F.J. Lichtigfeldl - Psychiatrist, Port Shepstone

leo-Directors S.A. Brain Res. Inst.

P.O. Box 8078, Johannesburg

Nitrous Oxide (N20) at analgesic concentrations has been

shown to cause naloxone reversible analgesia both experi-
mentally 1 • 2 and clinically3,4.

In addition, in vitro studies have shown cone 1us i ve ly the

opioid effect of N20 at receptor level 5 '6. In view of the
opioid effect of N20, it occurred to us that this agent may be
antidepressant, since endogenous opioids have been claimed to
have an anti-depressant action7 ,8 We thus investigated the
possible anti-depressant action of N20.

Acute alcohol withdrawal states have been shown to respond

extremely positively to analgesic doses of N20, with particular
reference to the affective symptoms9.

Furthermore, we have postulated that the improvement in

this syndrome is mediated via inhibition of a supersensitive
adrenergic systemlO. Clonidine has also been shown to ameliorate
the alcohol withdrawal statell, also possibly as a result of an
immediate pre-synaptic effect, followed by a more delayed post-
synaptic agonistic effect. This would have the effect of down
regulating the supersensitive post-synaptic adrenergic receptors,
postulated to be present in the alcohol withdrawal statelO
Furthermore, clonidine can cause an increase in endogenous opioid
levels in animalsl2, an effect that may also potentiate its
effect in alleviating symptoms of the withdrawal statell, this
effect highlighting the possible connection between this disorder
and the opioid system9.

397
 Matussek et all3 have shown that the Growth Hormone (GH)
response to Clonidine is blunted in patients suffering from
endogenous depression but not in neurotic depressions. We have
therefore tested the effect of N20 at analgesic concentrations on
the GH response in alcoholic volunteers, since opioids have been
shown to increase the release of some pituitary hormones includ-
ing GH14' 15. We also measured LH responses to N20 in two alco-
holic patients, since opioids cause a decrease in levela of this
hormone, probably secondarily to inhibition of release of LHRH16,

In our previous work as mentioned above we have demon-

strated that N20 at analgesic concentrations can alleviate
depressive symptoms, which occur during alcohol withdrawal
states9. We now present further evidence of the antidepressive
property of N20.

METHOD

DEPRESSION STUDY

ALCOHOLIC SUBJECTS

Forty-four Male caucasiod alcoholic subjects aged between

28 and 38 (mean 32) were tested after obtaining informed written
consent. Of these, 25 subjects were tested during treatment with
N20 of their acute alcoholic withdrawal state; the other 19
subjects had been detoxified at least 6 weeks before the trial
and were resident in the alcohol rehabi lit at ion facility at the
Rand Aid Home, and therefore· were no longer drinking.

Depression was assessed using a visual analogue scale lOOmm

long, marked at one end worst (0) and the other best (100)17,18,
The subject marked with a vertical line his state of depression
before each experimental condition, i.e. prior to 100% oxygen,
prior to analgesic concentrations of N20, prior to 100% Oxygen
and at the end of treatment. Each phase lasted 20 minutes.

The gases were administered with a Quantiflex relative

dental analgesic machine (Medishield S.A. Ltd). The concen-
tration of the mixture of N20 and Oxygen (used in phase 2) were
titrated to produce relaxation till the point just prior to
euphoria.

NON ALCOHOLIC SUBJECTS

Twenty-five (16 males and 9 females) depressed non

a lcoho lie subjects, were referred to us by independent psych ia-
trists who had diagnosed them as suffering from various types of
clinical depression. Their ages ranged from 27-43 (mean 38.)

398
 These patients were then assessed by us using the visual
analogue scalel7 in the following manner. All patients were
given pure oxygen for 20 minutes, followed by N20 and Oxygen,
which was then followed by 100% Oxygen (for 20 minutes).

HORMONE STUDY

Eight fasting abstinent alcoholic male caucasoid patients

gave informed consent. Their ages ranged from 25-26 (mean 31).
These patients were tested betwewen 07h30 and 10h30, to exclude
alterations in secretion as a result of circadian rhythym, after
an overnight fast. An i.v. cannula was inserted in an antecubial
vein and after a preliminary 15 minute period the first blood
sample was taken. The N20 was then administered at a pre-
determined level as described above for 1 hour. After this
period 100% oxygen was given for 20 minutes, and then a final
blood sample was taken 1 hour after the N20 had been dis-
continued. Two other blood samples were taken at 30 minutes and
60 minutes during the N20 administration. The hormones were
assayed using a standard radio-immunoassay.

RESULTS

ALCOHOLIC SUBJECTS

All 25 alcoholic subjects showed distinct improvements 1n

their depressive symptoms or mood, immediately after analgesic
concentrations of N2o had been administered. This elevation of
mood was less marked after the final administration of oxygen.

NON ALCOHOLIC SUBJECTS

RESPONDERS

In most cases (N=21) these subjects responded in a similar

manner to the alcoholic subjects, i.e. relief of their depression
with N20. In some cases (N=ll) oxygen alone improved their
mood. An unexpected finding was that most of the responders
(N=l7) claimed that their depression was relieved for hours to
days.

NON RESPONDERS

Some patients did not respond either to N2o or 0 2 • Of

these (N=4), the history revealed chronic depression which had
not responded to anti-depressant therapy. One of these subjects
responded to a combination of benzodiazepine and beta-blockers,
indicating that this patient suffered from a chronic anxiety
state with a secondary depression. The other case that showed no

399
HORMONE DETERMINATION (See Tables 1 and 2)

Table 1 GROWTH HORMONE RESPONSE TO N20 (ng/ml)

Subject 1 2 3 4

1 0.6 0.6 0.5 1.2

2 0.7 0.8 0.7 14.6
3 0.6 0.6 0.7 1.0
*4 12.1 1.5 0.8 0.6
5 0.4 0.6 1.0 0.6
6 0.7 0.7 0.7 4.6
7 0.7 0.5 3.2 3.7

*This subject it was discovered had eaten, which might

account for the anomalous levels.

Table 2 LUTEINIZING HORMONE RESPONSE TO N20 (ng/ml)

Subject 1 2 3 4

1 6.2 5.3 13.1 8.6

2 7.5 16.4 11.5 8.6

response to the treatment with N20 responded to administration of

neuroleptics.

DISCUSSION

We have been able to confirm that the euphoric effect of

N20 is part of a more general rapid and therefore more transient
antidepressant effect in a number of normal and clinical situ-
ations. In view of the almost immediate effect of this agent,
when it is successful, a comparison with standard anti-depressant
therapy is irrelevant for the moment. We have in fact found that
weekly application of N20 has assisted a number of patients
(n=lO) in maintaining their confidence during the well-known
latent period of orthodox anti-depressant therapy (unpublished
observations Gillman and Lichtigfeld). However, there is one
practical method of anti-depressant therapy that seems to have a
similar time sequence in producing its effect viz. the use of
luteinizing hormone releasing hormone (LHRH). The therapeutic

400
effect of this agent is rapid but transient and must be repeated
regularly to maintain improvementl9. I t is possible that our
preliminary demonstration of increased LH levels following N20
administration may be rel ~ted to the anti-depressant effect of
N20. The action of LHRH >·Jas initially thought to only benefit
reactive disorders, but Dekeyser feels that this may not be
truel9, particularly if administration is continuous.
Amphetamine, methlyphenidate and cocaine are also rapidly acting
and transient anti-depressant agents particularly in reactive
states. The possibility thus exists that N2o may also exert its
effect via a similar mechanism, e.g. increased and immediate
post-synaptic nor-adrenergic stimulation through opioid
inhibition of the presynaptic alpha 2 auto-receptor. For this
reason N2o would, as we have found, be more effective in treating
reactive depression.

There 1.s evidence that N20 can exert a more long-lasting

and less evanescent therapeutic effect, when it was ostensibly
used as a placebo. In fact Brill et al, showed that the highest
percent of improvement in a mixed depressive group was obtained
using ECT(alone), ECT plus muscle relaxant and N20 (given to
anaesthetize), while ECT plus thiopental, and the group with
thiopental alone had the lowest rate of improvement20. From this
it follows that N20 used on a long term basis may have a similar
effect as more traditional therapy with anti-depressants.

Our finding related to the hormonal responses must be

viewed with caution since we were dealing with alcoholic
population. However, it appears that most of our subjects showed
a small increase 1.n GH concentration following N2o admini-
stration. In one case (subject 2) there was an enormous
increase. Mattusek, however, has already shown that men with a
diminished GH response imbibed 1-2 litres of beer regularly. It
is therefore possible that the alcoholic history of our subjects
contributed to our results. Furthermore, such a response may be
a useful marker for male chronic alcoholics, who would have to be
distinguished from endogenous depressions, who Mattusek found to
also have a decreased response of GH to clonidine21. Unfortu-
nately, in view of. the (our) small selected sample, it is not
possible to speculate further on the exact mode of action of N20
on the release of GH It is possible that N2o has a releasing
action on growth hormone, in common with other opioidsl4,15. The
effect of LH release by N20 is opposite to the effect of other
opioids in the 2 cases examined and further work is at present
being undertaken. In conclusion, it appears that N20 may well be
a most useful tool in investigating neurotransmitter interactions
between the mono-aml.ne and the opioid systems in the human
subjects. Furthermore, our preliminary work (Gillman &
Lichtigfeld presented elsewhere at this conference) indicates

401
that N20 may possibly be used to distinguish between dexa-
methazone suppressors and non-suppressors.

REFERENCES

1. Berkowitz, B. A., Ngai, S. H., and Finck, A. D., 1976,

Nitrous oxide analgesia : Resemblance to opiate action,
Science, 194 : 967
2. Berkowitz, B. A., Finck, A. D., and Ngai, S. H., Nitrous
oxide analgesia: Reversal by nalaxone and development of
tolerance J. Pharmacol Exo. Thez:, 203:539
3. Gillman, M.A., Kok, L., and Lichtigfeld, F. J., 1980,
Paradoxical effect of naloxone on nitrous oxide analgesia
in man, Eur. J. Pharmacol, 61:175
4. Gillman, M. A., and Lichtigfeld, F. J., 1981. A comparison
of the effects of morphine sulphate and nitrous oxide on
chronic pain states in man, J. Neurol. Sci. 49:41
5. Ahmed, S.M., and Byrne, W. L. Opiate receptor binding
studies influence of a reversible sulfhydryl reagent, in
"Endogenous and exogenous opiate agonists and
antagonists", E. L. Way ed., Pergamon, New York (1980)
6. Daras, C., Cantrill, R. C., and Gillman, M. A., 1983 ( 3H)
naloxone displacement: evidence for nitrous oxide as
opioid receptor agonist, Eur. J. Pharmacol, 89:177.
7. Bunney, W. E., 1978, Drug therapy and psychobiological
research advances in psychoses in the past decade, Am. J.
Psychiat. 135:8
8. Vereby, K., Volavka, J., and Clonet, D., 1978, Endorphins in
psychiatry, Arch. Gen. Psychiatry, 35:877.
9. Lichtigfeld, F. J., and Gillman, M.A., 1982. Treatment of
alcoholic withdrawal state with oxygen and nitrous oxide
S. A. Med. J. 61:349
10. Gillman, M. A., and Lichtigfeld, F. J. A receptor hypothesis
of the alcohol withdrawal state in "CNS receptors - From
molecular pharmacology to behaviour P Mandel and F. V. de
Feudis eds., Raven Press, New York (in press)
11. Bjorkqvist, S. E., 1975, Clonidine in alcohol withdrawal,
Acta Psychiat. Scand. 52:256
12. Pettibone, D. J., and Mueller, G. P., 1981. Clonidine
release inimunoreactive beta-endorphin from rats pars
distalis Brain Res. 221:409
13. Matussek, N., Ackenheil, M., Hippins, H., Muller, F.,
Schroder, Schultes, H., and Wasilewski, B., 1980. Effect
of clonidine on growth hormone release in psychiatric
patients and controls, Psychiat Res. 2:25
14. Lindberg, T., Larsen, V., Kehlet, H., ·and Jacobsen, E., 1981
Respiratory, analgesic and endorcrine responses to an
enkephalin analogue in normal man. Acta. Anaesth. Scand.
25:254

402
15. Morley, H. E., 1981 The endocrinology of the opiates and
opioid peptides Metabolism 30:195
16. Report, J., F., Quigley, M. E., and Yen, S. S.C., 1981
Endogenous opiates modulate pulsatile luteinizing
release in humans, J. ~lin! Endocrin Metab. 52:583
17. Luria, R. E. 1975. The va11d1ty and rehability of the
visual analogue mood state. J. Psvchiat Res. 12:51
18. Davis, K. L., Berger, P. A., Ho(lister, L. E.,-and Defraetes,
E., 1978. Physostigmine for mania Arch Gen. Psychiat.
35:119
19. Dekeyser, Y. F. L., 1980. Hypothalamic releasing factor
primary and situational depression, Int. J. Neurol.
14:219
20. Brill, N. Q., Crumpton, E., Eituson, S., Grayson, H. M.,
Hellman, L. I., and Richards, R. A., 1959. Arch.Neurol.
Psychiat. 81:627 ·
21. Matussek, N. Effect of amphetamine and clonidine on growth
hormone release in psychiatric patients and controls. jn
Depressive Disorders (symposie Medica Hoechst 13) S.
Gavattini e.d. F. K. Schattauer Verlag. Stuttgart. (1978)
22. Gillman, M. A., and Lichtigfeld, F. J., 1983. Therapeutic use
of N20 in Psychiatry. Free paper VII World Congress of
Pschiatry.

403
RELEVANT SELECTIVITY OF ANTIDEPRESSANTS

Stuart A. Montgomery
Academic Department of Psychiatry
St Mary•s Hospital
Praed Street, London W2

The amine deficit hypotheses of affective disorders

have dominated research on the biochemical correlates of
depression for the past twenty years. These hypotheses
derive from the observation that certain drugs which
deplete or block the action of monamines cause depressive
symptomatology. Conversely certain compounds which have
monoamine reuptake blocking properties, have marked
antidepressant efficacy. The indoleamine hypothesis
suggests that in depression there is a serotonin
deficiency, and a reduction of neurotransmission
associated with the disturbance in the availability of
5 hydroxytryptamine (SHT). The catecholamine hypothesis
concentrates on disturbances in the noradrenergic system
and suggests that reduced availability of noradrenaline
(NA) is the significant factor.
One of the tests of the hypotheses would be to
demonstrate that depressed patients have low levels of
either amine presynaptically compared with normal
controls. The direct measurement of neurotransmission
is of course not possible in man and unfortunately there
is no adequate substitute model of depression in animals.
The only te5ts possible have therefore been indirect
measures such as the level of amine metabolites in
cerebrospinal fluid (CSF) or in urine.
Several studies of MHPG, the metabolite of NA in
urine have reported a lower level of the metabolite in
depressed patients. Some investigators have even

405
suggested that this level could be used as a predictor
of treatment response with specific antidepressants. It
is however unclear what light is thrown on the amine
hypothesis by these findings since it appears that only
a small amount of the MHPG excreted in urine is derived
from central metabolism of NA. Findings from the
measurement of MHPG in CSF which should give a closer
measure of brain metabolism have been inconclusive.
Some investigators report a reduction of the metabolite
in depressed patients but others have found the levels
to be within normal limits.
Studies of SHIAA, the metabolite of serotonin,
have concentrated on levels in CSF. The results of
some investigators have been interpreted as indicative
of the existence of a subgroup of depression associated
with low levels of 5HIAA in CSF before treatment (1, 2).
It is particularly intriguing that these patients with
low 5HIAA levels are associated with a high risk of
suicidal behaviour and it seems that the biochemical
variable is associated with a behaviour across several
diagnostic categories (3, 4, 5).
Other experimental approaches for testing the
hypotheses are the loading with precursors eg tryptophan,
5HPT etc., which effect the turnover of amines, measure-
ment of neuroendocrine responses which may be affected
by altered amine function, measurement of transport to
detect whether there is a contributory functional impair-
ment, and post mortem studies of brain amines. The
methodological problems involvedin all these approaches
are a considerable constraint and the inconsistency of
results means that there is no very compelling evidence
for the amine hypotheses.
It has long been acknowledged that at best the
theories represent a gross oversimplication and they
cannot explain a number of well recognised phenomena.
There are a number of 1 atypical 1 antidepressants such
as mianserin and iprindole which do not have the usual
uptake blocking effects on monoamines. The acute effects
on the neur~transmitter systems are seen immediately
both in animals and man but the antidepressant effect is
delayed and is notseen reliably until four waeks have
passed. This delay needs to be taken into account.
Newer theories of antidepressant action which have
focussed attention on post synaptic events take some
account of these deficiencies. The reduction in
sensitivity of the NA sensitive adenylate cyclase system
which antidepressants have in common appears to happen

406
independently of which amine system an antidepressant
is supposed to be affecting. However this comprehensive
theory like all theories only stands provided there is
not a reasonable example of an effective antidepressant
which does not produce this reduction in sensitivity.
The logic of the argument in favour of the amine
deficit hypotheses, which appear simple and attractive,
is in part not soundly based. A compound with anti-
depressant properties which has amine uptake blocking
actions does not necessarily exert its antidepressant
effect via this particular mechanism. Nevertheless
the search for antidepressants has been very much
influenced by the theories so that most of the newer
antidepressants developed have been variants of
existing tricyclic antidepressants or have been developed
to have specific effects on monoamine metabolism.
One of the possible approaches to testing the
hypothesis is to examine differential clinical response
to treatment with antidepressants having quite different
pharmacological actions. Such an investigation could
not be carried out with the earlier tricyclic anti-
depressants which had mixed effects on NA and serotonin
as well as having effects on cholinergic, muscarinic
and histamine receptors. Earlier attempts were made
using clomipramine as a selective 5HT uptake inhibitor
but the results from these studies were confounded by the
fact that the active metabolite, desmethyl clomipramine
is a relatively potent blocker of the reuptake of NA.
It has also been shown that the level of metabolite at
steady state may be higher than that of the parent
drug (6). The introduction of newer antidepressants
with more selective effects on the reuptake of NA and
5HT made it possible to test the simple amine hypotheses
in patients.
A clear comparison has been made in patients treated
with a NA uptake inhibitor, maprotiline, or with a 5HT
uptake inhibitor, zimelidine (7). According to the amine
deficit theory a deficiency in the serotonin system would
selectively be improved by a 5HT uptake inhibitor and
likewise a deficiency in the NA system would be benefitted
by treatment with a NA uptake inhibitor. In our study a
selective effect of these antidepressants could not be
seen. In 49 patients there was no relationship between
pretreatment levels of MHPG in the CSF and response to
treatment with maprotiline nor between pretreatment levels
of 5HIAA and treatment with zimelidine. Our CSF amine
metabolite results were consistent with other workers

407
in that we found a relationship between 5HIAA and HVA
and a history of suicidal attempts. The failure to
detect a selective antidepressant effect in a group
where the relative biochemical deficit had been
demonstrated casts doubt on the theory that there are
clinically selective antidepressants.
As with all negative findings it is possible that
there are other explanations. Zimelidine is not the
purest of the SHT uptake inhibitors available, merely
the earliest, and its metabolite, norzimelidine, has
some weak effects on the noradrenergic system.
Alternatively it may be that there is no such thing
as selectivity on these two amine systems and that
antidepressants working on either system affect both.
Most studies which have sought to demonstrate a
selective effect of antidepressants reveal a general
antidepressant effect despite reputed selectivity of
action. Certain symptoms such as sleep and appetite
may reflect pharmacological differences very early in
treatment but by the time the four to six week period has
elapsed when we look for antidepressant effect no
selectivity of antidepressant action may be seen.
The claims for selectivity have largely been based on
the biochemical theories and promoted without sufficient
validation in the clinic. The lack of selective clinical
effect of so-called selective antidepressants is consistent
\vith the experience of most clinicians who find that effect-
ive selective antidepressants appear to work in roughly
the same proportion and category of patients as effective
non-selective antidepressants. Caution is needed before
accepting conclusions about possible advantages of select-
ive antidepressants until there is adequate evidence from
the clinic. Theories which propose an antidepressant
effect which is independent of biochemical selectivity
become increasingly attractive.
REFERENCES
1 . M ft. s berg , P Thor en , L Tr as kman , L 8e r t i 1s son
and V Ringberger, "Serotonin Depression" -
A Biochemical Subgroup WithintheAffective Disorders?,
Science 191:478-480, (1976).
2. H M Van Praag, Toward a biochemical classification
of depression, Adv. Biochem. Psychopharmac.,
357-368, (1974).
3. M ~sberg, L Tr"askman, and P Thoren, SHIAA in the
cerebrospinal fluid - A biochemical suicide pred-
ictor?, Arch Gen. Psychiat. 33:1193-1197, (1976).

408
4. S A Montgomery and D Montgomery, Pharmacological
Prevention of Suicidal Behaviour, Journ. of Affect.
Disorders, 4:291-298, (1982).
5. G L Brown, J C Ballanger, MD Minichiello, and
F K Goodwin, Human aggression and its relationship
to cerebrospinal fluid 5-hydroxy-phenylglycol and
homovanillic acid, Psychopharmacology of Agression,
M Sandler, Raven Press, 131-148, (1979).
6. SA Montgomery, R McAuley, 0 Montgomery, S Dawling,
and R A Braithwaite, Plasma concentration of
clomipramine and desmethylclomipramine and clinical
response in depressed patients, Postgraduate Medical
Journal, 56:130-33, (1980).
7. SA Montgomery, S J Rani, R McAuley, D Roy and
D Montomery, The antidepressant efficacy of zimelidine
and maprotiline, Acta Psychiat. Scandin., Supp, 290,
63: 219-224, (1981).

409
THE SEROTONIN-NORADRENALINE LINK-HYPOTHESIS OF AFFECTIVE

DISORDERS

Fridolin Sulser

Vanderbilt University School of Medicine and

Tennessee Neuropsychiatric Institute
Nashville, TN 37232

Two main hypotheses of affective disorders have dominated re-

search strategies in Biological Psychiatry: The catecholamine
hypothesis (Schildkraut, 1965; Bunney and Davis, 1965) and the
indolealkylamine or serotonin hypothesis (Coppen, 1972). These
hypotheses are essentially deficiency hypotheses derived from
studies on acute pharmacological effects elicited by a number
of psychotropic drugs. The realization of the numerous inconsis-
tencies in biochemical metabolic data derived from either plasma,
urine or cerebral spinal fluid (Berger and Barchas, 1977) and of a
number of shortcomings of the original biogenic amine hypotheses
(Sulser, 1983), has triggered the search for alternate mechanisms.

New discoveries during the second half of the 1970's on the

regulation by antidepressants of receptor systems in brain have
shifted the research emphasis on affective disorders and on the
action of antidepressants from acute presynaptic neuronal effects
e.g. blockade of MAO, inhibition of uptake of noradrenaline (NA)
or serotonin (5HT) to alterations of postsynaptic receptor mediated
events following the administration of antidepressant drugs on a
clinically relevant time basis (for review see Sulser, 1983). I t
is by now well established that clinically effective antidepressant
treatments reduce the sensitivity of the NA receptor coupled adeny-
late cyclase in brain linked to a down-regulation of the number of
the beta adrenoceptor subpopulation. Contrary to earlier notions
that antidepressant treatments cause enhanced noradrenergic acti-
vity - which they do acutely - chronic treatment results in a net
deamplification of the NA signal despite the persistence of an in-
creased synaptic availability of the neurohormone caused by e.g.
blockade of neuronal uptake. This delayed net effect on noradren-
ergic transmission has been elegantly demonstrated by Heydorn et al.

411
(1982). Thus, chronic treatment with an MAO inhibitor or a tri-
cyclic antidepressant (which blocks the neuronal uptake of NA) re-
duced upon chronic administration the effect of both exogenously
administered isoproterenol or physiologically released NA on the
beta adrenoceptor mediated formation of melatonin in the pineal
gland.

Antidepressant treatments which have so far been shown to

unequivocally reduce the sensitivity of the NA sensitive adenylate
cyclase system and/or to down-regulate the density of its beta
adrenoceptor subpopulation are listed in Table 1.

On closer inspection, it becomes evident that this delayed biochem-

ical consequence following antidepressant treatments is not related
in any meaningful way to the neuronal uptake of biogenic amines.
Also, CNS active drugs devoid of clinical antidepressant activity
have been reported not to alter either the sensitivity of the NA
sensitive adenylate cyclase or the number of beta adrenoceptors
Table 1. Antidepressant treatments Which Cause
Subsensitivity of the NA-Sensitive Adenylate
Cyclase and/or Down-Regulate the Number of
RP.TR Adrenocentors
1. Drugs which block predominantly uptake of NA:
Desipramine
Nortriptyline
Oxaprotiline
2. Drugs which block more selectively uptake of serotonin
( SHT):
Zimelidine
3. Drugs which block uptake of SHT and NA:
Imipramine
Amitriptyline
Chlorimipramine
4. Drugs which block MAO:
Pargyline
Nialamid
Tranylcypromine
Clorgyline
5. Drugs which do not block uptake of either NA or 5HT:
Iprindole
Mianserin
Clenbuterol
6. Electroconvulsive treatment (ECT); rapid-eye-
movement sleep deprivation.

412
(Sellinger et al., 1980). Though antidepressants can exert effects
on other adrenergic receptor systems, these systems are not syste-
matically altered by antidepressants as has been shown for the
beta adrenoceptor system. The action of antidepressants on beta,
alpha1 and alpha2 adrenoceptor systems and its relevance for the
etiology and therapy of depression has been reviewed elsewhere
(Charney et al., 1981).

The earlier studies on the mechanisms involved in the regula-

tion of the sensitivity of the NA receptor coupled adenylate cy-
clase system and the down-regulation of beta adrenoceptors by anti-
depressants have shown that an unhindered synaptic availability of
the endogenous agonist NA at the receptor is necessary for desensi-
tization to occur. This evidence has been reviewed (Sulser et al.,
1983a). In relatively simple boilogical systems such as frog eryth-
rocytes, it has been demonstrated that desensitization of beta
adrenoceptor coupled adenylate cyclase systems involves the se-
questration of beta receptors away from other components of the
adenylate cyclase complex in internalized vesicles from where they
can recycle to the cell surface (Stadel et al., 1983). Whether or
not this scheme applies to neuronal beta adrenoceptors in brain
remains to be elucidated,

Recent studies have now shown that SHT is corequired with NA

for the down-regulation of the beta adrenoceptor system by tricyclic
antidepressants such as imipramine and desipramine (Brunello et al.,
1982a; 1982b; Janowsky et al., 1982). Moreover, beta adrenoceptors
are not only resistant to down-regulation by e.g. desipramine, but
in the absence of serotonergic neuronal input caused by destruction
of serotonergic neurons with 5,7-dihydroxytryptamine, the affinity
of beta adrenoceptors to agonists such as isoproterenol and also NA
is profoundly decreased (Manier et al., 1983). While desipramine
did not appreciably change the affinity of the agonist isoproterenol
for beta adrenoceptors in cortical tissue from normal animals, the
tricyclic antidepressant potentiated the decreased affinity of beta-
adrenoceptors for the agonist under conditions of impaired seroton-
ergic input (Table 2). We have now found that antidepressants with
a more selective action on serotonergic systems (e.g, zimelidine)
share with the predominantly noradrenergic drug desipramine the
shift in beta-adrenoceptor affinity in animals with impaired sero-
tonergic activity but not in normal animals (unpublished results
from this laboratory). More elaborate studies on the characteristics
of beta-adrenoceptors in the absence of serotonergic input have re-
vealed that these altered receptors display properties of "un-
coupled" receptors (Sulser et al., 1983b). It is noteworthy that
the reduction in the synaptic availability of SHT caused by p-
chlorophenylalanine (which inhibits the synthesis of SHT) reversed
within two days, despite the continuous administration of desipra-
mine, the down-regulated number of beta adrenoceptors but did not
reverse the desensitization of the NA sensitive adenylate cyclase

413
 Table 2. Change in Beta-Adrenoceptor Properties Following
Lesions of Serotonergic Neurones

(-)Isoproterenol
IC 50 ]1M ± SEM

Controls 0.43 ± 0.03 (12)

Controls plus DMI-treatment 0.51 ± 0.06 (6)

Lesioned with 5,7-DHT 2.19 ± 0.43* (11)

Lesioned with 5,7-DHT

plus DMI-treatment 11.9 ± 1.411~* (16)

Serotonergic neurons were lesioned with intraventricular

5,7-dihydroxytryptamine (5,7-DHT). 10 to 12 days following
lesioning, the animals received either desipramine (DMI:
15 mg/kg i.p.) or saline. The animals were sacrificed
7 days after the last injection. The ICso values re-
present the concentration of (-)-isoproterenol that dis-
places specific binding of [3H]-dihydroalprenolol by 50%
*p < 0.001 (Lesioned vs. Control) From Manier et al.
**p < 0.001 (Lesioned DMI vs. Lesioned) (1983)

(Sulser et al., 1983b). Cortical beta adrenoceptors from p-chloro-

phenylalanine treated animals like those from animals with lesions
of serotonergic neurons (Manier et al., 1983), displayed a marked
decrease in affinity for agonists as determined from competition
binding of (-)-isoproterenol for 3H-dihydroalprenolol (Sulser
et al., 1983b). These data are of considerable interest for two
reasons: 1) Assuming that the down-regulation of beta-adrenoceptors
by antidepressants reflects indeed a therapeutically relevant bio-
chemical adaptation that depends on the joint availability of NA
and 5HT, the data may provide the rationale for the puzzling clini-
cal finding that p-chlorophenylalanine nullified the therapeutic
activity of antidepressants in bipolar and unipolar endogenously
depressed patients (Shopsin et al., 1975; 1976). 2)The results
have heuristic value for the1neurobiologist interested in the molec-
ular mechanism of the regulation of both the recognition site and
the function of beta adrenoeeptors and the study of the fate of
beta-adrenoceptors after they can no longer be detected by radio-
ligands. The demonstrated co-requirement of 5HT for the NA receptor
inte~action to be effective in down-regulating beta adrenoceptors
might also provide an explanation for the reported potentiation of
the antidepressant effect of tricyclic antidepressants or MAO inhi-
bito~s by simultaneous administration of 5-hydroxytryptophan or

414
tryptophan (Coppen and Wood, 1982; Van Praag, 1982; Mendlewicz
and Youdim, 1980).

In the context of the known overlap throughout the neuraxis

of the terminal fields of serotonergic and noradrenergic projec-
tions arising from brain stem nuclei, the new experimental data
demonstrating a biomolecular linkage between serotonergic and
noradrenergic neurons, express~d functionally at the level of NA
receptors, provide the scientific basis for a "SEROTONIN-NORADREN-
ALINE LINK HYPOTHESIS" of affective disorders. The heuristic
value of this aminergic link hypothesis is considerable as it
makes possible new experimental approaches to study mechanisms of
the regulation of neuronal signal processing in brain. It is our
contention at the present time that the linked aminergic system
represents a fine tuning system that determines the intensity of
not only the NA signal transfer across the membrane but also the
sensitivity of other membrane receptor systems. The experimental
tools required for testing these ideas are available. The pursuit
of these studies should eventually contribute to our understanding
of the relationship of receptor structure, function and regulation
to the pathophysiology of affective and other CNS disorders.

ACKNOWLEDGEMENTS

The original research of this laboratory has been supported

by USPHS grant MH-29228 and by the Tennessee Department of Mental
Health and Mental Retardation.

REFERENCES

Berger, P.A. and Barchas, J.C., 1977, in: Psychopharmacology

(Barchas, J.C., Berger, P.A., Ciaranello, R.D. and
Elliot, G.R., eds.) pp. 151-173, Oxford University Press,
New York, NY.
Brunelle, N., Chunag, D.M., Costa, E., 1982; Use of specific
brain lesions to study the site of action of antidepres-
sants. Adv. Biosc. 40: 141.
Brunelle, N., Barbacc1a, M.L., Chuang, D.M. and Costa, E.,
1982, Down-regulation of S-adrenergic receptors following
repeated injections of desmethylimipramine: Permissive
role of serotonergic axons. Neuropharmacology 21,
1145.
Bunney, W.E. and Davis, J.M., 1965, Norepinephrine in depres-
sive reactions. Arch. Gen. Psyc~iat. 131: 483.
Charney, D.S., Menkes, D.B. and Hen1nger, G.R., 198la, Recep-
tor sensitivity and the mechanism of action of antide-
pressant treatment. Arch. Gen. Psychiatry 38, 1160.

415
Coppen, A., 1972, Indole amines and affective disorders. J.
Psvchiat. Res. 9, 163.
Coppen, A. and Wood, K., 1982, 5-Hydroxytryptamine in the
pathogenesis of affective disorders. Adv. Biochem.
Psvchouharmacol. 34, 249.
Heydorn, W.E., Brunswick, D.J. and Frazer, A., 1982, Effect
of treatment of rats with antidepressants on melatonin
concentrations in the pineal gland and serum. J. Pharma-
cal. Exp. Ther. 222, 534.
JanowsKy, A., OKada, F., Manier, D., Applegate, C.D. and
Sulser, F., 1982, Role of serotonergic input in the regu-
lation of the S-adrenergic receptor coupled adenylate
cyclase system. Science 218, 900.
Manier, D.H., Okada, F., Janowsky, A.J., Steranka, L.R. and
Sulser, F., 1983. Serotonergic denervation changes
binding characteristics of beta adrenoceptors in rat
cortex. European J. Pharmacal. 86, 137.
Mendlewicz, J. and Youdim, M.B.H., 1980, Antidepressant poten-
tiation of 5-hydroxytryptophan by 1-deprenyl in affective
illness. J. Affect. Disord. 2, 137.
Schildkraut, J.J., 1965, The catecholamine hypothesis of af-
fective disorders. Am. J. Psychiatry 122, 509.
Sellinger, M.D., Mendels, J. and Frazer, A., 1980, The effect
of psychoactive drugs on S-adrenergic receptor binding
sites in rat brain. Neurouharmacology 19, 447.
Shops in, B. , Friedman, E. and Gershon, 8. , -197 5, The use of
synthesis inhibitors in defining a role for biogenic amines
during imipramine treatment in depressed patients.
Psvchopharmacol. Comm. 1, 239.
Shopsin, B., Friedman, E. and Gershon, S., 1976. Parachloro-
phenylalanine reversal of tranylcypromine effects in de-
pressed patients. Arch. Gen. Psvch. 33, 811.
Stadel, M.M., Strulovici, B., Nambi, P., Lavin, T.N., Briggs,
M.M., Caron, M.G. and Lefkowitz, J., 1983, Desensitization of
the S-adrenergic receptor of frog erythrocytes. J. Biol.
Chern. 258, 3032.
Sulser, F., 1983, Mode of action of antidepressant drugs. J.
Clin. Psvchiat. 44 (5) Sec. 2, 14.
Sulser, F., Janowsky, A., Okada, F., Manier, D.H. and Mobley,
P.L., 1983a, Regulation of recognition and action function of
the norepinephrine receptor coupled adenylate cyclase system
in brain: Implications for the therapy of depression. Neuro-
pharmacology 22, 425.
Sulser, Y., Gillespie, D.G. and Manier, D.H., 1983b, Adrenoceptor
changes following repeated antidepressant treatment. Ann.
N.Y. Acad. Sci. (in press).
Van Praag, H.M., 1~82, Serotonin precursors in the treatment of
depression. Adv. Biochem. Psychopharmacol. 34, 259.

416
BIOCHEMISTRY OF MANIA

Robert M. Post, David R. Rubinow,

Phillip W, Gold and ~ffirkku Linnoila

Biological Psychiatry Branch, National Institute of

Mental Health and National Institute of Alcohol Abuse
and Alcoholism, Bethesda, Maryland, USA

INTRODUCTION

In this manuscript we briefly review the evidence for

neurotransmitter and peptide alterations in mania based on direct
measurements of these substances or their metabolites in the
cerebrospinal fluid (CSF) of man. The obvious difficulties of
Juaking inferences about central nervous system (CNS) function
from biochemical measurements in a global pool, such as the CSF,
require reiteration (Post et al., 1980) but are not further
discussed here.

CLASSICAL NEUROTRANSMITTERS

The several thousand noradrenergic cells located in the

locus coeruleus project diffusely throughout the brain and have
been linked to a variety of behaviors in animals that appear
altered in manic-depressive illness. It is thus of considerable
interest that norepinephrine has been found elevated in the CSF
of manic patients measured either by sensitive radioenzymatic
assay (Post et al., 1978) or by HPLC (unpublished observations,
1983). At the same time the norepinephrine metabolites, MHPG and
VMA, are not substantially elevated compared to values observed
in normal volunteers in our series, but the NIMH collaborative
study did find increases in MHPG in mania (Swann et al., 1983).
It is also most interesting that the biosynthetic enzyme convert-
ing dopamine to norepinephrine, dopamine beta-hydroxylase (UBH),
is significantly reduced in the CSF of manic patients (Post et
al., 1983). This dissociation of increases in CSF norepinephrine
in the face of decreases of its biosynthetic enzyme DBH in the
CSF suggests the possibilities that high levels of norepinephrine

417
inhibit DBH synthesis or release, that there is an endogenous
blockade of norepinephrine reuptake, that DBH is depleted in the
face of increased firing and release, or that the two substances
are differentially transported out of CSF in mania compared to
other clinical states. Further data enabling one to select among
these possibilites may help elucidate mechanisms of the
noradrenergic alterations observed in mania.

In contrast to the alterations in noradrenergic systems

in mania, there is little evidence for increases in dopamine
(unpublished observations, 1983) HVA or DOPAC in mania.
Similarly, baseline levels or probenecid-induced accumulations of
the serotonin metabolite 5-hydroxy-indoleacetic acid (SHIAA) are
not generally elevated in mania compared to euthymic and normal
states. While GABA deficiencies have been postulated in mania,
we found no evidence for this using direct CSF GABA measures
(Post et al., 1980) and in fact, in a rapidly cycling
manic-depressive patient, we have observed significantly higher
CSF GABA concentrations in the manic compared to the depressed
state. Moreover, in patients studied during depressed or
euthymic states, CSF GABA was significantly correlated with
number of episodes observed during NIH hospitalization,
suggesting that those with higher basal levels of GABA had more
rapidly cycling manic-depressive illness (Berrettini and Post,
1983).

ALTERATIONS IN CSF PEPTIDES IN MANIA

Measurement of peptides in the CSF of manic patients is of

particular interest because of the postulated role of these
substances in behavioral states such as mania; their presence in
CSF in concentrations that are often unrelated to those observed
in the periphery; the evidence that in many instances CSF
peptides are correlated with changes in brain function; and the
recent data that many of these substances coexist in the same
neurons with classical neurotransmitters (Post et al., 1982).

Although alterations in endogenous opiate substances have

been postulated in mania, we found no significant alterations in
either total CSF opioid binding activity or in immune reactive
beta-endorphin in manic patients compared to depressed patients
or normal volunteers (Post et al., 1983). Rubinow et al., 1983
reported that somatostatin in the CSF was significantly lower in
depressed patients, while values in the improved euthymic state
and in mania were not significantly different from those in
normal volunteers. These data parallel those of Gerner and
Yamada (1982). These investigators also failed to find
significant alterations in cholecystokinin (CCK) levels in the
CSF of manic patients compared to other groups, findings of
considerable interest in relation to the recent reports of the

418
coexistence of CCK and dopamine in the same neurons.

In contrast to these largely negative studies, Carman et

al. (1983) has found CSF calcitonin significantly lower in the
CSF of manic patients compared to other diagnostic subgroups and
normal controls, with a similar increase in the number of
nondetectable levels in the CSF of manics. These findings are of
interest in relation to the postulated behavioral and biochemical
effects of calcitonin (Carman et al., 1983), but remain to be
replicated and further investigated before a clear-cut
interpretation of the findings can be made.

The levels of the posterior pituitary hormones, arginine

vasopressin and oxytocin, have been studied by Gold et al. (1983)
and appear to show largely opposite alterations in mania compared
to depression. Values of arginine vasopressin are significantly
higher in manic compared to non-psychotic bipolar depressed
patients, but are not significantly different from normal
controls. In contrast, values of oxytocin are significantly
lower in manic patients compared to either depressed patients or
normal volunteers {Gold et al., unpublished observations). These
findings are of interest in relation to the postulated effect of
these substances on cognitive function (Gold et al., 1983) and
remain to be further investigated as they might relate to various
behaviors or symptoms in the manic syndrome.

CONCLUSIONS

Given the various clinical methodological and physiological

limitations of the measurements of substances in CSF of man {Post
et al., 1980), the initial data indicate that manic patients
display alterations in noradrenergic function, while there is
little evidence for alterations in the dopaminergic, serotonergic
and GABAergic neurotransmitter systems. CSF norepinephrine is
elevated, while CSF DBH is decreased in manic patients compared
to controls; the metabolites MHPG and VMA do not appear
substantially altered. These findings are suggestive that CSF
norepinephrine may be increased in the absence of an increase in
norepinephrine turnover and are at least partially supportive of
the original catecholamine hypothesis. However, it is of
particular interest that while there are increasing data for the
involvement of dopaminergic mechanisms in mania based on
pharmacological data {Post et al., 1980; Silverstone, this
symposium), there is little direct evidence from the CSF studies
to support this possibility. It is also possible that discrete,
regional alterations in dopamine metabolism that occur in areas
other than the striatum are obscured and not manifested in a
pooled source, such as CSF. Similar caveats hold for most of the
neurotransmitters measured in CSF reviewed here; positive
findings are of interest and require further exploration, while

419
negative findings even with relatively large Ns are less than
definitive and have to be interpreted with caution. Similarly,
the reliability and implications of the alterations in CSF
peptides in manic patients require further study.

References

Berrettini, w. H., and Post, R. M., 1983, GABA in affective

illness, in: "Neurobiology of Mood Disorders," R. M. Post
and J. C. Ballenger, eds., Williams and Wilkins,
Baltimore.

Carman, J., Wyatt, E., Smith, w., Post, R. M., and Ballenger,
J. c., 1983, Calcium and calcitonin in bipolar affective
disorder, in: Neurobiology of Mood Disorders," R. M. Post,
and J. C. Ballenger, eds., Williams and Wilkins,
Baltimore.

Gerner, R. H. and Yamada, T., 1982, Altered neuropeptide

concentrations in CSF of psychiatric patients. Brain
Res., 238:298.

Gold, P. w., Ballenger, J. c., Weingartner, H., Rubinow, D. R.,

Hoban, M. C., and Goodwin, F. K., 1983, Vasopressin in
affective illness: direct measurement, clinical trails,
and response to hypertonic saline, in: "Neurobiology of
Mood Disorders," R. M. Post and J. c. Ballenger, eds.,
Williams and Wilkins, Baltimore.

Post, R. M., Ballenger, J. c., and Goodwin, F. K., 1980,

Cerebrospinal fluid studies of neurotransmitter function in
manic and depressive illness, in: "Neurobiology of
Cerebrospinal Fluid, Vol. I," J. H. Wood, ed., Plenum
Press, New York.

Post, R. M., Ballenger, J. C., Hare, T. A., Goodwin, F. K., Lake,

C. R., Jimerson, D. c., and Bunney, w. E., Jr., 1980,
Cerebrospinal fluid GABA in normals and patients with
affective disorders, Brain Res. Bull., 5:755.

Post, R. M., Gold, P. w., Rubinow, D. R., Ballenger, J. c.,

Bunney, w. E., Jr., and Goodwin, F. K., 1982, Peptides in
the cerebrospinal fluid of neuropsychiatric patients: an
approach to central nervous system peptide function.
Life Sci., 31:1.

420
Post, R. M., Jimerson, D. c., Ballenger, J. C., Lake, c. R.,
Lerner, P., Uhde, T.W., and Goodwin, F.K., 1983, CSF
norepinephrine and its metabolites in manic-depressive
illness, in: "Neurobiology of Mood Disorders," R. M. Post
and J. C. Ballenger, eds., Williams and Wilkins,
Baltimore.

Post, R. M., Jimerson, D. C., Bunney, w. E., Jr., and Goodwin,

F. K., 1980, Dopamine and mania: behavioral and
biochemical effects of the dopamine receptor blocker
pimozide, Psychopharmacology, 67:297.

Post, R. M., Lake, c. R., Jimerson, D. C., Bunney, w. E., Jr.,

Wood, J. H., Ziegler, M. G., and Goodwin, F.K., 1978,
Cerebrospinal fluid norepinephrine in affective illness,
Am. J. Psychiatr~, 135:907.
Post, R. M., Pickar, D., Ballenger, J. c., Naber, D., and
Rubinow, D. R., 1983, Endogenous opiates in CSF:
relationship to mood and anxiety, in: "Neurobiology of
Mood Disorders," R. M. Post and J. C. Ballenger, eds.,
Willams and Wilkins, Baltimore.

Rubinow, D. R., Gold, P. w., Post, R. M., Ballenger, J. c.,

Cowdry, R., Ballenger, J. c., and Reichlin, S., 1983,
Somatostatin in affective illness, Arch. Gen. Psvchiatrv
40:403.

Swann, A., Secunda, s., Davis, J. M., Robins, E., Israel, H.,
Koslow, s., and Maas, J. w., 1983, CSF monoamine
metabolites in mania, Am. J. Psvchiatrv. 140:396.

421
THE PSYCHOPHARMACOLOGY OF MANlA: TOWARDS A UNIFYING HYPOTHESIS

Trevor Silverstone
Academic Unit of Human Psychopharmacology
St. Bartholomew 1 s Hospital, London, England

DOPAMINE
According to the original catecholamine hypothesis of affective
disorder [1], mania was thought to be due to an overactivity within
central noradrenergic (and possibly dopaminergic) pathways. But
in recent years dopamine (DA) has emerged as being of relatively
greater importance. The first hint that this might be the case
came from the report that fusaric acid, a DA hydroxylase inhibitor,
exacerbated the condition [2]. We went to examine this possibility
further by studying the effects of pimozide, a relatively specific
dopamine (DA) receptor blocking compound in five patients presenting
with a florid manic illness [3]. The results were encouraging;
all five patients improved within three days. We concluded that
our finding was consistent with the idea that mania results from
overactivity of central DA pathways. The same year Gerner et al
[4] came to a similar conclusion on the basis of their finding that
piribedil, aDA receptor agonist, precipitated manic episodes in
a predisposed individual, while pimozide ameliorated them.
In a subsequent review of the clinical pharmacological evidence
[5] I pointed out that manic symptoms generally appeared to be
improved by drugs which reduced DA activity, whereas it was
frequently precipitated by drugs which enhanced central DA
neurotransmission. At that time mania had been reported to be
relieved by alpha-methyl-paratyrosine, a compound which inhibits
the synthesis of DA; by a variety of antipsychotic drugs, all of
which inhibit post-synaptic DA receptors; and by lithium which,
among its many other actions, inhibits DA receptor-linked adenylate
cyclase (But more of lithium later). Drugs which had been reported
to precipitate mania included levodopa which enhances DA synthesis;
423
amphetamine which promotes DA release from the pre-synaptic neurone;
and bromocriptine and piribedil which act as direct DA receptor
agonists. All these observations when taken together led me to
conclude, "Any hypothesis which purports to explain the pathogenesis
of manic-depressive illness in neurochemical terms must find some
place in it for dopaminergic mechanisms." How well does this opinion
stand up five years later?
As far as the specific DA receptor blocking drug pimozide is
concerned, both the NIMH group [6] and ourselves [7] have fully
confirmed our earlier findings that pimozide is an effective anti-
manic drug. Post et al reported that the time course of response
to pimozide in seven of the eight patients treated with that drug
was comparable to that following treatment with chlorpromazine.
Similarly we observed that pimozide was at least as effective as
chlorpromazine in a double-blind controlled comparison of the two
drugs involving twenty-four manic patients.
More recently Nolen [8] has reported that the cis form of
clopenthixol, which has DA receptor blocking properties, is more
effective in the treatment of mania than trans-clopenthixol which
has much less DA receptor blocking activhY.
Further evidence to support the DA hypothesis of mania has
come from the report that bromocriptine can exacerbate mania, even
in seeemingly adequately-treated patients [9]. It has also induced
mania in acromegalic patients being treated with high doses of the
drug (Turner et al - unpublished observations).
However central the role of DA may be in the pathogenesis of
mania, it is highly unlikely to be acting in isolation.
Furthermore, there are other psychopharmacological findings which
point to the involvement of other neurotransmitters, particularly
acetylcholine (ACh) and gamma-aminobutyric acid (GABA).
ACETYLCHOLINE
[i] Physostiqmine
Janowsky et al [10] reported that intravenous administration
of physostigmine, an inhibitor of cholinesterase which crosses the
blood-brain barrier, improved such manic symptoms as elation,
grandiose delusions, flight of ideas, hypermotility and pressure
of speech, whereas no such improvement was noted after treatment
with neostigmine, a cholinesterase inhibitor which does not cross
the blood-brain barrier. This observation was confirmed by Davis
et al [11]. Assuming that the physostigmine-mediated inhibitor
of cholinesterase led to a functional enhancement of ACh activity,
they suggested that mania could be due to an imbalance between
cholinergic and dopaminergic transmitter pathways.

424
LITHIUM
Any adequate neurochemical theory of mania has to accomodate
the oft-repeated observation of the efficacy of lithium in this
condition, albeit in mainly the milder type of case. If we knew
how lithium worked we would be significantly further forward in
our understanding of the pathogenesis of mania itself. A number
of suggestions have been put forward, some of which involve an
action on one or other of the three neurotransmitter pathways we
have been considering.
[i] Dopamine
In experimental animals lithium appears capable of reducing
the denervation-type supersensitivity of striatal DA receptors which
follows long-term administration of the DA receptor blocking
compound, haloperidol [23]. A recent review emphasises that there
is, "an abundance of evidence [10/12 studies] suggesting that
lithium is capable of modulating the development of behavioural
supersensitive responses ... [24]. However, DA receptor binding when
measured directly appears to be unaffected by lithium [25].
[ii] Acetylcholine
In vitro studies of ligand binding by muscarinic receptors
derived from human caudate show that lithium reduces the affinity
of these receptors for a muscarinic receptor antagonist [26]. This
is in keeping with the possibility that lithium may have an overall
ACh enhancing action, and would fit in with the ACh-DA imbalance
hypothesis.
[iii] GABA
Patients with a history of bipolar affective disorder who were
euthymic and were receiving prophylactic lithium were found to have
higher levels of GABA in their CSF and plasma than similar patients
who were medication-free [27]. The levels in the medication-free
euthymic patients were below those of control subjects. Thus
lithium·appeared to raise CSF and plasma GABA to normal values,
but this may not be specific to GABA as CSF levels of other amino
acids such as taurine, aspartate and glycine have also been found
to be elevated by lithium [28].
CONCLUSIONS
Where have we got in our perambulation through the recent
literature on the pharmacology of mania?
Firstly, I think there is little doubt that DA plays a central

425
[ii] Lecithin
If increasing the availability of ACh in certain central
neurotransmitter pathways is the mechanism by which physostigmine
improves manic symptoms, then similar clinical effects might follow
an increase in central ACh activity produced by other means.
Lecithin is a precursor of ACh which in animals has been shown to
elevate brain ACh levels. A double-blind placebo-controlled trial
of lecithin in six manic patients revealed a clear superiority of
lecithin in five out of the six [12].
GABA
The possible involvement of GABA in mania originated from the
finding that the anti-convulsant drug valproate, which is thought
to act by increasing the availability of GABA in the brain was more
effective than placebo in relieving manic symptoms in four of the
five patients treated with the drug [13]. Three further patients
have been reported to have benefitted from valpr10ate [14]. On the
basis of these findings it was suggested that mania may represent
a state of GABA dysfunction.
Another anti-convulsant drug, carbamazepine, has also been
assayed as a treatment for mania. Following an earlier uncontrolled
study Okuma et al [15] undertook a double-blind comparison of
carbamazepine and chlorpromazine in sixty patients. They found
carbamazepine to be at least as effective as chlorpromazine.
Ballenger and Post [16] similarly reported that carbamazepine had
potential anti-manic activity; four out of the nine patients they
treated did well, and another patient did moderately well. It was
noted that carbamazepine was effective in certain patients who had
failed to respond to lithium, raising the possibility of two types
of mania, one type being responsive to carbamazepine, the other
to lithium. Folks et al [17] also found carbamazepine to be
effective in lithium non-responders. In this context the
combination of carbamazepine and lithium has been reported as being
more effective than either drug alone [18, 19] suggesting that they
may act on different but complementary neurochemical mechanisms.
It is not known how carbamazepine works; one suggestion is
that it is one of its metabolites, carbamazepine - 10, 11 epoxide,
which is the active antipsychotic agent [20]. This in turn has
been postulated to act by reducing the phenomenon of 'kindling'
within the limbic systems [21].
In keeping with a role for GABA in the pathogenesis of mania
is a case report that baclofen, which is a GABA agonist having an
inhibitory effect on mesolimbic DA neurones, led to mania on being
suddenly withdrawn [22].

426
role in its pathogenesis but probably only in close relationship
to other neurotransmitter and peptide modulator systems.
Acetylcholine and GABA in particular appear to be implicated,
largely, as we have seen on the basis of a rapidly growing body
of pharmacological evidence.
Can these seemingly independent findings be reconciled into
a unifying hypothesis? It is generally accepted that within the
nigro-striatal system there is a well-delineated interaction between
dopaminergic, cholinergic and GABAergic pathways; activation of
the DA neurones of the substantia nigra causes inhibition of the
cholinergic interneurones in the striatum, these, in turn, activate
GABA neurones, the axons of which pass back to the substantia nigra
and inhibit the firing of the DA neurones there [29].
Is there a similar interacting set of neurotransmitter pathways
in the mesolimbic system, upon which the pharmacological agents
we have been discussing act? Certainly there is an important DA
input to the limbic areas arising from the A10 area in the ventral
tegmentum [30]. Also, as in the striatal system, there are
cholinergic neurones within the limbic system [31]. Finally, there
are well-defined efferent pathways impinging back on the
ventrotegmental A10 DA area; and it is quite possible that GABA
may be acting as an inhibitory neurotransmitter in this tract,
thereby attenuating DA activity in the A10 neurones [32].
If such an interconnecting neurotransmitter system containing
dopaminergic, cholinergic and GABAergic elements, were indeed to
be involved in the pathogenesis of mania, it would adequately
account for the range of pharmacological findings reported in this
review. Or, putting it another way, if this turns out to be the
case, then the tools of clinical psychopharmacology would have been
shown to be of real value in defining the potential neurotransmitter
systems within the human brain concerned with the appearance of
that particular constellation of mental symptoms which, when taken
together, we call mania. That way lies progress.
REFERENCES
[1] J.J. Schildkraut, Catecholamine hypothesis of affective
disorders: Review of supporting evide~ce, A~ J Psvchiatrv, 122:
509-522 (1965). ~
[2] R.L. Sack and F.K. Goodwin, Inhibition of dopamine-~-hydroxylase
in manic patients, Arch Gen Psychiatry, 31: 649-654 (1974).
[3] J.C. Cookson ana 1. ~, 1verstone, ~-Hydroxytryptamine and
dopamine pathways in mania: A pilot study of fenfluramine and
pimozide, Br J Clin Pharmac, 3: 942-943 (1976).
[4] R.H. Gerner, R.M. Post and W.E. Bunney, A dopaminergic mechanism
in mania, Am J Psychiatry, 133: 1177-1180 (1976).

427
 [5] T. Silverstone, Dopamine, mood and manic-depressive psychosis,
in: "Depresssive Disorders", ed S. Garattini, Schattaner Verlag,
~uttgart, 419-430 (1978).
[6] R.M. Post, D.C. Jimerson, W.E. Bunney and F.K. Goodwin, Dopamine
and mania: Behavioural and biochemical effects of the dopamine
receptor blocker pimozide, Psychopharmacoloqy, 67: 297-305 (1980).
[7] J.C. Cookson, T. Silverstone and ~. Wei Is, A double-blind
clinical trial of pimozide and chlorpromazine in mania: A test of
the dopamine hypothesis, Acta Psvch Scand, 64: 381-397 (1981).
[8] W.A. Nolen, Dopamine and mania: The effects of trans and cis-
clopenthixol in a double-blind pilot study. J Affect Dis, 5: 91-
96 (1983).
[9] J.M. Johnson, Treated mania exacerbated by bromocriptine, Am
J Psvchiatry, 138: 980-982 (1981).
[10] D.S. Janowsky, M.K. El Yousef and J.M. Davis, Parasympathetic
suppression of manic symptoms by physostigmine, Arch Gen Psychiatry,
28: 542-547 (1973).
[11] K.L. Davis, P.A. Berger, L.E. Hollister and E. Defraites,
Physostigmine in mania, Arch Gen Psvchiatrv, 35: 119-122 (1978).
[12] B.M. Cohen, J.F. Lipinski ana R:I. Altesman, Lecithin in the
treatment of mania: Double-blind, placebo-controlled trials, Amer
J Psvchiatrv, 139: 1162-1164 (1982). --
"[13] H.M. Enirich, D. Von Zerssen, W. Kissling and H.J. Mollen, On
a possib.le role of GABA in mania. Therapeutic efficacy of sodium
valproate, Jlrlu RinrhPm P"vchonharmacol, 26: 287-296 (1981).
[14] D. Von-zefssen and H.M.-rmrich, New drugs for mania? Br J
Psvchiatry, 142: 532-533 (1983).
Tl5] T. OKuma, K. Inananga, S. Otsuki, K. Sarai, R. Takahashi, H.
Hazama, A. Mori and M. Watanabe, Comparison of the antimanic
efficacy of carbamazepine and chlorpromazine: A double-blind
controlled study, Psychopharmacoloqy, 66: 211-217 (1979).
[16] J.C. Ballenger and K.M. Post, Carbamazepine in manic-depressive
illness: A new treatment, Amer J Psvchiatrv, 137: 782-790 (1980).
[17] D.G. Folks, L. Douglas-Klng, S.B. Dowdy, W.N. Petrie, R.A.
Jack, J.C. Koonen, B.R. Swenson and P. Edwards, Carbamazepine
treatment of selected affectively disordered in-patients, Amer J
Psvchiatrv, 139: 115-117 (1982). ·
!18] J.F. Lipinski and H.G. Pope, Possible synergistic action
between carbamazepine and lithium carbonate in the treatment of
three acutely manic patients, Amer J Psvchiatry, 139: 948-949 (1982).
[19] R. Keisling, Carbamazepine and lithfum carbonate in the
treatment of-refractory affective disorders, Arch Gen Psvchiatrv,
40: 223 (1983). .
[20] R.M. Post, T.W. Uhde, J.C. Ballenger, D.C. Chatterji, R.F.
Greene and W.E. Bunney, Carbamazepine and its - 10, 11 - epoxide
metabolite in plasma and CSF. Arch Gen Psychiatry, 40: 673-676
( 1983). -
[21] R.M. Post, T.W. Uhde, F.W. Putnam, J.C. Ballenger and W.H.
Berrettini, Kindling and carbamazepine in affective illness, J Nerv
Ment Dis, 170: 717-731 (1982).

428
 [22] E.S. Arnold, S.M. Rudd and H. Kirshner, Manic psychosis
following rapid withdrawal from baclofen, Amer J Psychiatry, 137:
1466-1467 (1980). -
[23] A. Pert, J.E. Rosenblatt and C. Sivit, Long term treatment
with lithium prevents the development of dopamine receptor
supersensitivity, Science, 201: 171-173 (1978).
[24] W.E. Bunney and ~.L.- Garland, Possible receptor effects of
chronic lithi11m administration, Neuropharmacology, 22: 367-372
(1983).
·[25] F.E. Bloom, G. Baetge, S. Deyo, A. Ettenberg, L. Koda, P.J.
Magistretti, W.J. Shoemaker and D.A. Staunton, Chemical and
physiological aspects of the actions of lithium and antidepressant
drugs, Neurooharmacoloqy, 22: 359-365 (1983).
[26] G.D. Tollefson and S. Senogles, A cholinergic role in the
mechanism of lithium in mania, Bioloqical Psvchiatrv. 18: 467-479
(1982).
[27] W.H. Berrettini, J.I.Nurnberger, T.A. Hare, S. Simmons-Alling,
E.S. Gershon and R.M. Post, Reduced plasma and CSF gamma-amino-
butyric acid in affective illness: Effect of lithium carbonate,
Bioloqical Psvchiatrv. 18: 185-194 (1982).
[28] P.J. Goodnick, H.E. Evans, D.L. Dunner and R.R. Fieve, Lithium
and amino acids, Amer J Psvchiatrv, 139: 538-539 (1982).
[29] C.D. Marsden, ~asal ganglia d~sease, Lancet, 1141-1146 (1982).
[30] U. Ungerstedt, Stereotaxic mapping of the monoamine pathways
in the rat brain, Acta Psvchiat Scand, 367: 1-48 (1971).
[31] S.D. Iversen, Recent advances in the anatomy and chemistry
of the limbic system, to be published.
[32] D.P. Van Kammen, Gamma-amino-butyric acid (GABA) and the
dopamine hypothesis of schizophrenia, Amer J Psychiatry, 134: 138-
143 (1977).

429
PROLACTIN, GROWTH HORt·10NE AND CORTISOL IN MANIC ILLNESS

John C. Cookson
Consultant Psychiatrist
The London Hospital, 2a Bow Road, London E3 4LL

The secretion of anterior pituitary hormones is normally

controlled by neurotransmitters. A study of these hormones in
manic illness may therefore provide information about the role of
the neurotransmitters in the pathogenesis of mania and in the action
of anti-manic medication.
PROLACTIN
The secretion of prolactin (PRL) is under tonic inhibitory
control by dopamine (DA) released from tubero-infundibular neurones.
If overactivity of DA pathways contributes to the pathogenesis of
mania, it might be predicted that improvement in clinical ratings
during treatment with a neuroleptic drug would be paralleled by
a rise in circulating PRL levels. This was indeed found in a group
of 11 manic patients treated with pimozide, a selective DA receptor-
blocking drug [1]. After the first day of treatment PRL levels
rose gradually towards a plateau during two weeks of treatment;
clinical ratings improved in parallel with this.
However, this gradual rise in PRL level does not reflect a
gradual increase in the degree of blockade of pituitary DA
receptors, but represents a change in the responsiveness of the
pituitary. This conclusion is based on a later study of manic
patients treated with oral haloperidol and given intravenous doses
of haloperidol to test the respnnsiveness of the pituitary [2].
In untreated manic patients PRL levels are usually in the
normal range, thus providing no evidence that tubero-infundibular
DA neurones are overactive in mania.

431
GROWTH HORMONE
The secretion of human growth hormone (GH) has a more complex
control; both DA and noradrenaline (NA), through post-synaptic
alpha-2 receptors, act indirectly to cause release of the hormone.
Circulating levels of GH in a group of untreated manic patients
were not significantly different from levels in the same group of
patients during treatment with pimozide or at full recovery [1].
This finding is consistent with the view that DA pathways in
the hypothalamus are not overactive in mania, and it suggests that
alpha-2 receptors in the hypothalamus controlling GH secretion are
not overactive in mania.
CORTISOL
Circulating levels of cortisol reflect the secretion of
adrenocorticotrophic hormone (ACTH) from the pituitary. The
secretion of ACTH is normally controlled in part by noradrenergic
pathways with stimulation by alpha receptors [3]. Dopamine is not
thought to be important in the normal control of ACTH.
Cortisol levels are raised above normal more often in samples
taken at midnight than in morning samples in manic patients [4,
5].
Cortisol levels were said to be suppressed by dexamethasone
quite normally in mania. More recently, non-suppression has been
reported in about 50% of manic patients [7, 8]. It is not clear
whether non-suppression relates to the severity or duration of the
manic state or to the occurrence of depressive symptoms that are
common in mania.
In a study of more than 30 untreated manic patients we found
plasma cortisol levels to be raised above normal in all patients
in samples taken at midnight, but in only half the patients in
samples taken during the day. Patients with more severe clinical
ratings were more likely to have raised daytime cortisol levels
[9].
During treatment with pimozide, the raised levels of cortisol
returned towards normal in the course of two weeks, in parallel
with improvement in clinical ratings [10]. By contrast, during
treatment with haloperidol, cortisol levels were restored to normal
within 3 days, whereas clinical improvement continued for 2 weeks
[9].
The neurochemical mechanisms underlying the abnormal secretion
of cortisol in mania are unknown; it is possible that hypothalamic

432
noradrenergic pathways become activated in mania.
REFERENCES
[1] J.C. Cookson, T. Silverstone and L.H. Rees, Plasma prolactin
and growth hormone levels in manic patients treated with pimozide,
Br J Psvchiatrv, 140: 274-279 (1982).
LZJ J.C. Cookson, P.J.A. Moult, D. Wiles and G.M. Besser, The
relationship between prolactin levels and clinical ratings in manic
patients treated with oral and intravenous test doses of haloperidol,
Psvchol Med, 13: 279-285 (1982).
IJJ L. Rees, P.W.P. Butler, C. Gosling and G.M. Besser, Adrenergic
blockade and the corticosteroid and growth hormone response to
methylamphetamine, Nature, 228: 565-566 (1970).
[4] S.R. Platman anOIR.R. Fieve, Lithium carbonate and plasma
cortisol response in affective disorder, Arch Gen Psvchiatrv, 18:
591- 594 ( 196 8) •
[5] E.J. Sachar, Twenty-four hour cortisol secretory patterns in
depressed and manic patients, in: "Progress in Brain Research",
Vol 42, •Hormones, Homeostasis and the Brain•, W.H.B. Gispen, T.J.
Van Wimersima Greidanus, B. Bohus and D. De Wied eds, pp 81-91,
Elsevier, Amsterdam (1975).
[6] B.J. Carroll, Neuroendocrine function in mania, in: "Manic
Illness", B. Shopsin ed, pp 163-176, Raven Press, New)ork (1979).
[7] R.R. Krishnan, A.A. Maltbie and J.R. Davidson, Abnormal cortisol
suppression in bipolar patients with simultaneous manic and
depressive symptoms, Am J Psvchiatrv, 140: 203-205 (1983).
[8] P.M. Graham, J. Booth, G. Borange, S. Galhenage, C.M. Myers,
C.L. Teoh and L.S. Cox, The dexamethasone suppression test in mania,
J Affec Dis, 4: 201-211 (1982).
[9] J.C. Cookson, T. Silverstone, S. Williams and G.M. Besser-
in preparation.
[10] J.C. Cookson, T. Silverstone, G.M. Besser and S. Williams,
Plasma corticosteroids in mania: The effects of pimozide,
Neuropharmacology, 19: 1243-1244 (1980).

433
AMPHETAMINE INDUCED AROUSAL IN HUMAN SUBJECTS AS A

MODEL FOR MANIA

David Jacobs

Department of Postgraduate Education in

Psychiatry Unit, Selby Centre, Selby Street
Shenton Park 6008, Western Australia

One avenue for investigating the biological aspects

of mania, is to use a pharmacological model and in the
past a variety of animal models have been proposed.
These experiments however risk being an over-simplifica-
tion with the additional difficulty of extrapolation
from one species to another. Furthermore, animal models
allow no measure of the mental state, a necessity in·
research in the field of psychological illness.

A useful pharmacological model should demonstrate

close similarities symptomatically, psychophysiologically
and in the response to pharmacological challenge, with
the condition concerned, along with the drug having an
action relevant to the aetiological questions being
considered.

The response to amphetamine in human su£jects has

been proposed as a suitable model for mania . Ampheta-
mine has been shown in experimental animals to cause the
release of newly formed noradrenaline (NA) and dopamine
(DA) from 2 the presynaptic terminals and to prevent their
re-uptake • As both NA and DA have been implicated in
the aetiology of affective ill~ess, as in the catechola-
mine hypot~esis of Schildkriut , and the DA hypotheses
of Randrup and Silverstone , the pharmacological action
of amphetamine may be particularly appropriate a model.

Since its introduction the mood elevating, alerting,

activating, and sleep reducing effects of amphetamine in
humans have been well recognised; such a pattern of

435
 Table 1. Comparison of Mania and
Subjective Response to dAmphetamine.

Symptoms Mania dAMP

MOOD
Elation 1' 1*
Irritability 1' t*
HYPERACTIVITY
Arousal i f **
Restlessness i 1'*
Increased Energy t i **
Racing Thoughts 1' t **
Sleep J. ~ **

*P < .05 **P < .01

responses are similar to the symptoms occurring in mania.

In order to compare the features of mania with the

changes following administration of a low dose of dextro-
amphetamine (dAMP), a study in twenty-four healthy male
subjects was undertaken. Each subject was given either
20 mg dAMP or matching placebo at 1800 hours following
an eight hour fast. Before taking the drug and at half
hourly intervals for the next four hours subjects
completed visual analogue scales relating to elation,
irritability, alertness, restlessness, increased energy
and racing thoughts. In addition, they were given a
questionnaire the next morning concerning their sleep on
the night following dAMP or placebo. At least one week
later subjects returned and the procedure was repeated
with the alternative preparation. The order of adminis-
tration was random, and the whole procedure conducted
under strict double-blind conditions. See Table 1.

In addition, similar changes have been reported with

mania and following the administration of amphetamine
in some physical parameters and pharmacological
responses. See Table 2.

Thus it is clear there are many similarities

between normal subject's response to oral dAMP and
mania, rendering the amphetamine response by normal
human subjects a potentially useful and relevant
pharmacological model for mania.

436
Table 2. Other Reported Changes--Mania V Amphetamine

Parameter Mania Amphetamine

PHYSICAL
Pulse i) ( 5) i) ( 6)
Skin Conductance j) 't)

Stereotypy r (7) 1' (8)

NEUROENDOCRINE
Cortisol t ( 9) i ( 10)
PHARMACOLOGICAL RESPONSE
Methyl Paratyrosine l (11) l. ( 12)
Pimozide J ( 13) ,!,(14)
Lithium Carbonate l ( 15) J. ( 16)

One clear application of this model is to use the

known action of dAMP to explore the relative roles of
DA and NA in mania. In the catecholamine hypothesis
of affective illness Schildkraut reasoned that higher
levels of NA seem most likely to be cau~ing the
heightened mood and activity with mania . More recent
evidence 1 gives greater weight to an excitatory role of
dopamine which now seems to have eclipsed the role of
noradrenaline.

In order to investigate the possible roles of NA

and DA, an experiment using the amphetamine model of
mania as described earlier, was performed in two parts.
The first to investigate the 17ffects of the specific DA
blocking drug pimozide (PMZ) , and the second to inves-
tigati8the effects of a specific ~ NA blocker thymoxamine
(TMX) on the response to 20 mg of oral dAMP. Twelve
healthy male volunteers were used in each experiment.
Either 2 mg of PMZ or 4 mg of PMZ was given two hours
prior to dAMP. PMZ placebo was given two hours prior to
dAMP or dAMP placebo, giving four combinations in all.
Either 80 or 160 mg of TMX were used in a similar set of
combinations one hour prior to dAMP. The drug combina-
tions were randomised using a Latin square design and
given under double-blind conditions. Subjective change
was measured using visual analogue rating scales as
described in the first study. Psychophysiological
measures included pulse and blood pressure, skin conduct-
ance and rate of spontaneous fluctuations. Blood speci-

437
 Table 3. The Effect of PMZ and TMX Pretreatment on
the Response to dAMP.

Parameters
dAMP
alone
I PMZ + dAMP
V dAMP alone
TMX + dAMP
V dAMP alone

SUBJECTIVE CHANGES
Elation t * J. t
Irritability t * NC f *
Alertness 1' ** ·~ NC
Energy 1' ** i * NC
Restlessness t * .J. f **
Mental Speed 1' ** NC 1'
PHYSIOLOGICAL
MEASURES OF AROUSAL
Log sc t * J, * t *
Spontaneous
Fluctuations t ** ,L * t *
Pulse
Systolic BP
1' **
t **
+
NC
** NC
't *

Statistical difference comparing changes with pretreat-

ment of blocking drug with amphetamine alone.
NC = No change i or ! = Trend apparent
* = P(.05 ** = P<.Ol
mens were taken hourly for an estimation of serum dAMP
levels.

The serum dAMP levels were not significantly

influenced by either blocking drug. The results are
summarised in Table 3.

Although some of the individual _changes in subject-

ive measures fail to reach statistical significance,
taking the overall pattern of response, PMZ pretreatment
caused an attenuation of the mood and arousal response
caused by dAMP, whereas TMX caused an accentuation.

The finding that PMZ attenuated the euphoriant and

the activating effect of dAMP is consistent with previous
results and confirms the excitatory role of DA in dAMP-
induced arousal. The model in this regard relates well
to the role of dopamine in mania.

The finding that TMX pretreatment caused an

438
accentuation of the mood and activating effect of dAMP
was surprising, particularly in the light of the
previously held view that NA was 1 involved in inducing
states of activation and arousal~. These results suggest
an inhibitory rather than an excitatory role for the
~ NA system in states of increased agitation, and by
implication in mania. The evidence for an excitatory
role for NA in mania or in other states of arousal and
activation, on review appears quite conflicting. Assays
of NA metabolites in the CSF show an elevation with
mania, tho~gh it is not clear whether this is secondary
or primary . The published pharmacological studies are
difficult to interpret. Metabolic precursors of NA that
precipitate mania are precursors of DA as well. Giving
the DA- /3 -hydroxylase inhibitor fusaric acid which
blocks the format~8n of NA and increases available DA.
accentuates mania . Most of the drugs which reduce
manic symptoms and which act on catecholamine neuro-
transmitters exert a primary action through DA in£ib~tion,
as for example, alpha-methyl-paratyrosine and PMZ ' .

Experiments with drugs that influence specifically

the NA pathways have failed to yield a clear-cut answer
to the question of the involvement of NA in the patho-
genesis of mania. Clonidine caus2~ sedation and has been
reported to reduce manic symptoms , however it is
uncertain whether this action depends on a pre- or post-
synaptic alpha-adrenergic receptor activation. A NA
blockers in high doses can ameliorate manic symptoms but
as the non- /J NA blocking stereo-isomer of propanalol
also reduces manic symptoms, doubts have been raised
as to whether the antima~~c effect of propanalol is via
its ~ NA blocking action . The effects of NA receptor
blocking drugs on mania appears to be as yet unstudied.

Amphetamine, which we know is stimulatory and can

precipitate mania, presumably via its release of DA, 2~n
be sedating i~ 5 certain ar~gsed subjects (hyperkinesis .
schizophrenia and mania which raises the question
of whether the sedation is part of its NA activity.

There is increasing evidence for a more complex

role of NA in the central nervous system particu1~71y
from experimental animals. Antelman and Caguilla have
proposed that the function of NA varies according to the
state of arousal of the animal. In the normal animal NA
exerts either no or an excitatory action, but in the
aroused animal NA pathways inhibit the functioning of
the DA syszgm and DA-induced behaviour. Cools and
Van Rossum provideneurophysiological evidence for a

439
 neuromodulatory role for NA of DA in the nucleus
accumbens. Thus, it seems reasonable to propose on the
basis of the model described that at least part of the
function of NA pathways in mania may be to inhibit or
regulate the excitatory DA system. This in turn raises
further questions for research--a necessary attribute
of any potentially useful model.

REFERENCES

(1) T Silverstone and J Cookson, The Biology of Mania,

in: Recent Advances in Clinical Psychiatry, 4, edited
by Granville-Grossman, 201-241, (1962).
(2) J Scheelkruger, Some Aspects of the Mechanism of
Action of various Stimulant Amphetamine Analogues,
Psvchiat Neural Neurochir (Arnst), 75, 179-192, (1972).
(3) J Schildkraut, The Catecholamine Hypothesis of
Affective Disorders: A Review of Supporting Evidence.
Am J Psvchiatry, 122, 509-522, (1965).
(4) A Randrup, I Munkvad, R Fog, J Gerlach, L Molander,
B Kjellberg, J Scheelkruger, Mania, Depression and Brain
Dopamine, iQ: Current Developments Psychopharmacology,
2, Spectrum Publications, 205-248, (1975).
(5) D Hemsley and H Philips, Models of Mania: An Indivi-
dual Case Study, Brit J Psvchiatry, 127, 78-85, (1975).
(6) D Jacobs and T Silverstone, Pimozide Inhibits and
Thymoxamine Enhances D-Amphetamine Induced Arousal in
Human Subjects, 13th CINP Conoress_, ( 1982) .
(7) G Ashcroft, D Eccleston, L Murray, A Glen, T Crawford,
I Pullar, P Shields, D Walter, I Blackburn, J Connechan,
M Lonergan, Modified Amine Hypothesis for the Aetiology
of Affective Illness, Lancet, 573-577, (1972).
(8) G Ashcroft, D Eccleston, J Waddell, Recognition of
AmphetamineAddicts, BritMedJ, 1, 57, (1965).
(9) J Cookson, T Silverstone, G Besser, S Williams,
Plasma Corticosteroids in Mania: The Effects of Pimozide,
Neuropharmacology, 19, 1243-1244, (1980).
(10) G Besser, P Butler, J Landon, L Rees, Influence of
Amphetamines on Plasma Corticosteroid and Growth Hormone
Levels in Man, Brit Med J, 528-530, (1969).
(11) H Brodie, D Murphy, F Goodwin, W Bunney (Jnr), Cate-
cholamines and Mania: The Effect of Alpha-Methyl-Para-
tyrosine on Manic Behaviour and Catecholamine Metabolism,
Clinical Pharmacal Ther, 12(2), 218-224, (1971).
(12) L Jonsson, E Anggard, L-M Gunne, Blockade of IV
Amphetamine Euphoria in Man, Clin Pharmacal Ther, 12,
889-896, (1971).
(13) J Cookson, T Silverstone, B Wells, Double-Blind
Comparative Clinical Trial of Pimozide and Chlorpromazine
in Mania: A Test of the Dopamine Hypothesis, Acta Psych
Scand, 64, 381-397, (1981).

440
 (14) T Silverstone, J Fincham, B Wells, M Kyriakedes,
The Effects of the Dopamine Receptor Blocking Drug Pimo-
zide on the Stimulant and Anorectic Actions of Dextroam-
phetamine in Man, Neuropharmacoloqy, 19, 1235-1237, (1980).
(15) M Schou, Special Review: Lithium in Psychiatric
Therapy and Prophylaxis, J Psych Res, 6, 67-95, (1968).
(16) B Angrist and S Gershon, Variable Attenuation of
Amphetamine Effects by Lithium, Am J Psychiatry, 136(6),
806-8101 ( 1979) • .
(17) R Pinder, R Brogden, P Sawyer, T Speight, R Spencer,
G Avery, Pimozide--A Review of its Pharmacological Prop-
erties and Therapeutic Uses in Psychiatry, Druqs, 2, 1-40,
(1976).
(18) J Roquebert, P Demichel, A Malek, Anti-noradrenergic
Activity of Thymoxamine and its Metabolites in Rats,
Arch int Pharmacodvn, 249, 12-25, (1981).
(19) J schildkraut and s Kety, Biogenic Amines and
Emotion, Science, 156, 21-30, (1967).
(20) R Post, Biochemical Theories of Mania, in: Mania
An Evolving Concept, edited by Belmaker and Van Praag,
Spectrum Publications, New York, 217-265, (1980).
(21) R Sack and F Goodwin~ Inhibition of Dopamine B
Hydroxylase in Manic Patients, Arch Gen Psychiatr~, 31,
649-654, (1974).
(22) R Jouvent, Y Lecrubier, A Puech, P Simon, D Widlocher,
Antimanic Effect of Clonidine, Am J Psychiatry, 137(10),
1275-1276, (1980).
(23) H Emrich, D Vonzerssen, H Moller, W Kissling,
C Cording, J Schietschilt, E Riedel, Action of Propanol
in Mania: Comparison of the D and L Isomer, Pharmako-
I>Sychiat, 12(4), 294-304, (1979).
(24) L Arnold, P Wender, K McCloskey, S Snyder, Leva-
amphetamine and Dextroamphetamine: Comparative Efficacy
in the Hyperkinetic Syndrome, Arch Gen Psvchiatry, 27,
816-822, ( 1972) .
(25) D Van Kammen, W Bunney, J Docherty, S Marder,
M Ebert, J Rosenblatt, J Rayner, d-Amphetamine Induced
Heterogenous Changes in Psychotic Behaviour in Schizo-
phrenia, Am J Psychiatry, 139(8), 991-997, (1982).
(26) Von H Beckman and H Heineman, D-Amphetamine Beim
Manischen Syndrom, Arzheimittel Forsch, 26(6), 1185-
1186, (1976).
(27) S Antelman and A Caggiula, Norepinephrine -
Dopamine Interactions and Behaviour, Science, 195,
646-653, (1977).
(28) A Cools and J van Rossum, Multiple Receptors for
Brain Dopamine in Behavioural Regulation Concept of
Dopamine E and Dopamine I Receptors, Life Sciences,
27 ( 14)
1 1237-1253, (1980)
o '

441
TEN YEARS OF LITHIUM TRANSPORT RESEARCH

David G. Ostrow

Departments of Psychiatry and

Community Health & Preventive Medicine
Northwestern University Medical School
VA Lakeside Medical Center
Chicago, IL 60611 USA

The occasion of an international symposium on the subject of

lithium transport appropriately calls for a short review of some of
the history behind this area of research. We are fortunate to
have represented in this symposium a number of researchers from
Europe and the United States, all of whom have been working for
several years in this particular area of biological psychiatry.
The communication that symposiums such as this provide for those of
us working in this area will hopefully accelerate progress in
understanding the relevance of lithium transport to the clinical
practice of psychiatry. But first, a short review of the last ten
years of lithium transport research. This will be a personal view
of things reflecting primarily those events which have most influ-
enced my own research directions.

The earliest indication that there may be differences in the

membrane control of lithium distribution is in an early paper by
Lyttkens and collaborators {1). In 1973, these investigators
showed that in manic-depressive females treated with lithium the
distribution ratio of lithium ion between red blood cells and
plasma was approximately 0.5. This indicated a roughly two-fold
higher concentration of lithium on the outside of the red cell in
these patients relative to the inside of the red cell, and needs to
be explained by further understanding of the transport mechanisms
controlling the influx and efflux of lithium in red blood cells.
However, when normal control female subjects were likewise placed
on lithium carbonate, their intracellular to plasma lithium ratio
was even lower, nearly 0.39. This suggested that there might be
differences in the relative transport properties of lithium in red
blood cells {RBCs) between the two groups, namely manic-depressive
females and control females. While no control male group was
443
studied by Lyttkens and his collaborators, manic-depressive males
had lithium ratios similar to the control females, namely 0.38.
The physiological basis of this apparent sex difference will be
discussed in the following symposium. Other early research
efforts that heightened interest in RBC lithium transport came from
the Philadelphia investigators. In 1973, Drs. Mendels and Frazer
published results showing that the average lithium ratio appeared
to be higher in lithium responders when compared to non-responders,
and suggested that the lithium ratio might be a more valid index of
brain lithium than the usual serum or plasma level alone (2).
These investigators proposed a membrane theory of major affective
disorders, wherein they hypothesized a defect in membrane ion
transport. Drs. Dorus and Pandey subsequently demonstrated that
there were genetic mechanisms underlying the lithium ratio through
the use of twin studies (3,4). Most recently, the latter investi-
gators have shown that there is a correlation between the genetic
transmission of reduced levels of RBC lithium efflux and major
affective disorders in the families of bipolar patients (4).

The beginning of my own interest occurred when, as a junior

medical student, I had the opportunity to care for a patient with
an acute· lithium carbonate overdose. When monitoring of the
patient's blood lithium level by the standard clinical lab proved
impractical due to the 4-5 day delay in returning results, I turned
to the research laboratory of Dr. John Davis at Illinois State
Psychiatric Institute (ISPI) for the determination of both plasma

2.0

0.02 •
0 ~ ~ ~ ~ M ~ ~ ~
TIME (iloun)

Fig. 1. Kinetics of Plasma ••~-'"1•• and RBC 1e.,_-~•• Lithium During and
Following Peritoneal Dialysis for Acute Overdose
(from Sparber and Ostrow, 1973, unpublished observations)

444
and red cell lithium levels on an immediate turn-around basis. As
shown in Fig. 1, even under the conditions of acute lithium toxic-
ity, the plasma lithium levels in this patient always remained
considerably above RBC levels. This indicated to us that there
were mechanisms for the active removal of lithium from the RBC
operating at a rate matching or exceeding the rate of lithium
removed from the patient's extracellular fluid spaces during peri-
toneal dialysis. Shortly thereafter I joined Drs. Davis, Pandey,
and Tosteson in a collaborative investigation of lithium transport
pathways, facilitated by Dr. Tosteson having come to the University
of Chicago to serve as the Dean of the Division of Biological
Sciences just after he and Mark Haas had discovered a transport
mechanism explaining the low intracellular lithium level (5).
This transport pathway, now known as either lithium-sodium counter-
flow or sodium-dependent lithium efflux (6), is a sodium-dependent
movement of lithium across the RBC membrane in a direction opposite
that of the trans-membrane sodium gradient. The initial phases of
our collaborative research rapidly delineated the five major path-
ways of lithium transport in human red blood cells illustrated in
Fig. 2. The major lithium influx pathways are leakage of lithium
through the membrane and movement of lithium bicarbonates through
an anion flux channel (7). The efflux of lithium under normal
physiological conditions appears to be restricted primarily to
movement of lithium in an outward direction via the sodium-
dependent efflux pathway (8). Energy for this pathway comes
entirely from the trans-membrane sodium gradient, and is not
directly dependent on ATP generation. The finding that the in
vivo lithium ratio was inversely proportional to the rate of
lithium efflux via this pathway provides significant evidence that
this is the determining transport pathway in terms of the lithium
ratio (9).

Fig. 2. Pathways of Lithium Transport in Erythrocytes

445
 Once having determined the ion transport pathways responsible
for the lithium ratio. we are faced with the scientific problem
that numerous physiological factors may modulate the rate and
extent of lithium movement across the RBC membrane. In 1977. when
we were rapidly increasing our knowledge of lithium transport
physiology. we hypothesized circulating factors which could
increase the lithium ratio either by decreasing the rate of lithium
efflux or increasing the rate of lithium influx. As my talk. and
possibly others in this symposium. will show. we have in the last
three years obtained a growing appreciation that such putative
transport modulating factors do indeed exist. and may indeed play a
role in elevations or depressions of the lithium ratio in humans.
This in turn has lead to a growing appreciation of the interplay
between genetic. environmental. and physiological factors in
controlling a membrane process such as lithium transport. I feel
that this parallels at the level of basic membrane physiology our
growing _understanding of a similar complex interaction among
genetic, environmental. and physiological factors in regulating the
response of the human organism to stress and our conceptualizatio n
of psychiatric illness.

The speakers who have come to Vienna to participate in this

symposium at the Fifth World Congress of Psychiatry illustrate the
interdisciplinary nature of lithium transport research. Dr.
Jochen Duhm will speak to us about his important wor~ in studying

10
Fig. 3. Preservation of
Human RBCs for Lithium
Efflux Measurements
'D'
l
>
..
E
8

4
~
5
+ Lithium efflux rate over
;;;;
five days of cell preser-
vation in four subjects.

.. 8
'i .
i::: 8

-.•ED'

~2
RBC sodium content during
0'---'---'---L-__JL_ __JL__ preservation
2 3 4 5 (from Ref. 10)
Day

446
the basic mechanisms of RBC lithium transport and its possible
relationship to other ion transport mechanisms. Dr. David Garver
will describe the Cincinnati group's work in studying lithium
ratios in schizophrenia, and the role that membrane composition may
play in regulation of the ratio. Dr. Istvan Szentistvany will be
describing his and Dr. Janka's work in delineating lithium trans-
port pathways in cultured neurons. Finally, Dr. Julian Mendelwicz
will be discussing his work on the genetics of the lithium ratio in
manic-depressive patients.

Looking over this roster of investigators from four countries

brings us back to my initial point of the importance of interna-
tional symposiums in bringing together persons working in different
disciplines and different locales on similar problems. I hope that
today's symposium will stimulate the sharing of information that
will further our attempts to standardize methodology, resolve
differences, and look for similarities in our findings. A field
which is as technically oriented as lithium transport research
requires standardized diagnostic and laboratory methods before its
results can be brought into the arena of practical clinical
psychiatry. We need to be able to collaborate with other investi-
gators who may be interested in this work and have clinical prob-
lems worthy of investigation, but do not have sophisticated techno-
logical facilities. I would like to end this introduction there-
fore, by decribing a technique which we have recently developed for
the purpose of permitting and encouraging collaboration among
researchers and clinicians in various geographical areas. As shown
in Fig. 3, human RBCs can be preserved for up to 5 days in a media
containing KCl and still retain their relative sodium-dependent
lithium efflux rates (10). This technique allows researchers in
such disparate parts of the United States as California, w.
Virginia, and Philadelphia to ship RBC samples from patients under-
going a multi-centered clinical protocol to our laboratory in
Chicago. It also permits blind analysis of those samples by a
central laboratory, and thereby improves our ability to objectively
investigate the clinical potential of this measurement in psychiat-
ric diagnosis and treatment.

REFERENCES

1. L. Lyttkens, U. Soderberg, and L. Wetterberg, Increased

lithium erythrocyte-plasma ratio in manic-depressive
psychosis, Lancet 1:40 (1973).

2. J. Mendels, and A. Frazer, Intracellular lithium concentra-

tions and clinical response: Towards a membrane theory of
depression, J. Psychiatric Research 10:9-18 (1973).

447
3. E. Dorus, G. N. Pandey, A. Frazier, and J. Mendels, Genetic
determination of lithium ion distribution: I. An in vitro
monozygotic-dizygotic twin study, Arch. Gen. Psychiatry
31:463-465 (1974).

4. E. Dorus, G. N. Pandey~ R. Shaughnessy, M. Gaviria, E. Val,

s. Erickson, J. M. Davis, Lithium transport across red
cell membranes: A cell membrane abnormality in manic-
depressive illness, Science 205:932-934 (1979).

5. M. Haas, M. Schooler, and D. c. Tosteson, Coupling of

lithium to sodium transport in human red cells, Nature
258:425-427 (1975).

6. D. G. Ostrow, M. Trevisan, A. Okonek, R. Gibbons, R. Cooper,

J. M. Davis, Sodium dependent membrane processes in major
affective disorders, !!!: "Biological Markers in Psychiatry
and Neurology," I. Hanin and E. Usdin, eds., Pergamon Press,
Oxford (1982).

7. G. N. Pandey, B. Sarkadi, M. Haas, R. B. Gunn, J. M. Davis,

D. C. Tosteson, Lithium transport pathways in human red
blood cells, J. Gen. Physiology 72:233-247 (1978).

8. D. G. Ostrow, G. N. Pandey, J. M. Davis, S. S. Chang,

D. C. Tosteson, Clinical study of lithium transport: A
defect of phloretin-sensitive lithium transport in bipolar
illness, Proceedings of the 1977 American Psychiatric Asso-
ciation Annual Meeting 130:152-153.

9. D. G. Ostrow, A. Okonek, S. Sparks, S. Flagel, R. Cooper,

J. M. Da~is, RBC lithium efflux in major affective disorders,
Proceedings of the 1982 American Psvchiatric Association
Annual Meeti~g 135:180.

10. D. G. Ostrow, M. Canessa, S. Sparks, and J. M. Davis,

Preservation of human erythrocytes for sodium-dependent
lithium efflux rate measurements, Clin. Chim. Acta
129:39-44 (1983).

AcknowledR.ements

Dr. David G. Ostrow was the recipient of research support from

the Davis-Freedman Fou~dation and the VA Career Development Program
during the period of work described here. He would like to thank
Drs. Daniel C. Tosteson and Mitzy Canessa (Harvard Medical School),
and Dr. John M. Davis (ISPI) for their guidance and support during
the performance of these studies.

448
LITHIUM ION TRANSPORT MECHANISMS IN HUMAN ERYTHROCYTES

Jochen Duhm

Department of Physiology, University of Munich

Pettenkoferstr. 12, D-8000 Munchen-2, Germany

The observation of interindividual differences in the Li+

distribution rati~ between red cells and plasma (1) and the finding
of high ratios Li ./Li+ being characteristic for manic-depressive
(bipolar) patients~(2) ~uggested that disturbances in cell membrane
function may be somehow related to the pathophysiology of bipolar
illness. In st~dies on the basic mechanism~ underlying the interindi-
vidual differences in red cell membrane Li transport five pathways
for Li+ movements in human erythrocytes have been characterized.
These are listed in Fig. 1, together with inhibitors which have
been essential tools in their identification.

CELL MEDIUM TRANSPORT -S'f.lTEM INHIBITORS

~@} No'- Li' EXCHANGE }NEM

PHLORETIN
.
Li@
} ~·-K ' PUMP } OUABAIN

I} I
Lilil
Klil FUROSEtJj()E
Li@ ~·-K· COTRANSPORT PIRETANIOE
Klil BUMETANIOE

LEAK } OIPVRIOA.'>1(lE
}ANION EXCHANGE }SITS.DIOS
Fiqure 1: Li+ transport pathways in human erythrocytes

449
 In principle, all five pathways mediate Li+ transport in both
directions across the membrane. Only those directions are depicted
in Fig. 1 which+mediate net-movements in ~iv~ during maintenance Li+
therapy. For Li movements through the Na -K cotransport system both
directions are shown because it is not yet known which of the two pre-
dominates under physiological conditi~ns.
The most important routes for Li movements in vivo a;e i~dicated
by heavy lines in Fig. 1: outward transport through the Na -Li ex-
change system, and inward transport through the leak and the anion
exchange system, the latter two pr~ceeding at about equal rates under
physiological conditions (3-6). Li movements through the ~ther p~th­
ways are almost completely blocked at the physiological Na and K
concentrations in cells ~nd medium, respeftive!Y·
The steady-state Li distribution Li ./Li is thus the re~ult
of the balance between the relative rates 5f do~nhill inward Li move-
ments ($hrough the leak and the a~ion exchange system) and uphill out-
ward Li transport (by the+Na+-Li exchange system). The interindivi-
dual differences in the Li distribution ;atio are caused by differen-
ces in the maximum activity of the Na -Li exchange system (7).

THE Na+-Li+ EXCHANGE SYSTEM

In 1975-1978, the Na+-Li+ exchange or countertransport s~stem

has been identified to be responsible for the extrusion of Li
against its electrochemical gradient in h~man erythrocytes (8-13).
The system mediates a secondary active Li ~phill outward tra~sport
which is driven by the inwardly directed Na gradient. The Na gradi-
ent is ~ain$ained by the action of the ATP-dependent Na+-K+ pump.
Na -Li exchange is a previously unknown function of the ouabain-
resistant Na+-Na+ exchange system which was first described in 1947
by Ussing in frog skeletal muscle (14). Later, similar exchange sys-
tems were found to be present in almost all mammalian tissues, inclu-
ding nerve (15,16) and muscle cells (17) and erythrocytes (18).
The exfhange system has a 20-30-fold higher affinity for Li+
than for Na . The apparent K ~alues f~r internal and external Li+
range between 0.5 and 2 mM, ~a and Li competing with one another
when present at the same side of the membrane (11,13). The stoichi~­
me$ry ~f the exchange is 1:1 in the three possible modes of Na -Na ,
Na -Li and Li+-Li+ exchange, respectively (13,19).
+ + + +
PHYSIOLOGICAL FUNCTION OF THE Na -Li (Na -Na ) EXCHANGE

One possible physiological functi~n of the Na+-Na+ exchange sys-

tem m~y be the removal of the toxic Li cation from the cell interior.
If Li were passively distri~uted between the intra- and extracellular
water space, the cellular Li concentration should exceed 10-27 times
that prevailing in the interstitial fluid in cells with a membrane
potential of ~0-90 mV, as calculated from the Nernst equation. Today,

450
the plasma Li concentration in man ranges between 1 and 10 11M
(depending on the Li+ content of tap water and food) and reaches
values of 0.1-1 r~l in mineral waters an~ up to 10 mM in salt lakes.
A greater than 10-fold enrichment of Li in the cell interior due
to the cation attracting fo;r:ce of the inside negative potential can
thus well cause cellular Li to attain toxic concentrations. In
evolution it may have been essential for excitable ~ell~ with a
high membrane potential to maintain+the Na+-Na+ (Na -Li ) exchange
system functioning to serve as a Li detoxifying mechanism.
In erythrocytes, however, the 1.4-fold accumulation of Li+ resul-
ting from the low membrane potential of about 10 mV probably did
not exert harmful effects of cell function. Acc~rdingly, mutations
causing lowering or even deletion of the Na+-Li exchange system in
erythrocytes exerted no negative selective pressure and could thus
freely develop during evolution over billions of years. This hypo-
thesis (6) explains why the rPn rPll NA+-T.i+ exr.hanae svstem shows
a marked intraindividual variability in erythrocytes of man and
other species such as beef (19). There are al~o drastic interspecies
differences, the mean activity of red cell Na -Li+ exchange increa-
sing in the+order rat [<0.01] < man [1) < high-K sheep
[5] < low-K sheep [9] < horse [25) < beef [31] (relatlve rates in
brackets) (6,19). 2+
. + + + 2+ . + -
Other lOns such as K , Rb , Cs , Ca , Mg , chollne , Cl and
N0 3- do not interact with the exchange syst~m. ~he only exc~pti~n
are hydrogen ions which inhibit both the Na -Na and the Na -Li
exchange (18,19)~ In Fig. 2 the effect of the external pH on Li+
uptake by the Na -Li+ exchange system is assessed at different exter-
nal Li+ concentrations in beef erythrocytes. Obviously, hydrogen ions
inhibit inw~rd Li transport, the effect being more pronounced at low
external Li . Evaluation of the data in the Eadie-Hofstee plot shown
in Fig. 3 reveals that lowering the pH from 8.45 to 7.36 decreases
the apparent affinity of the system for external Li+ while the maximum
transport rate remain~ unaltered. Thi~ behaviour su~gests a competi-
tive inhibition of Li transport by H . At higher H activities the
inhibition appears to be of the mixed type with both the affinity and
the maximum rate decreasing. From the Dixon plot in Fig. 4 an ap~arent
inhibitory constant may be estimated in the range of 80-120 nM H
activity (corresponding to a pH of 7.10-6.85) .+A similar. inhibitory
constant of 120 nM H has been reported for Na release lnduced by
externa~ Na+ (20). Accordingly, the suggestion of Funder ~t ~1. (20)
that Na -Li+ exchanqe in erythrocytes is mediated by aNa -H exchanqe
system appears to be reasonable. A simil~r fonclusion has recently
been d;r:aw~ by Aronson from studies of Na -H exchange in kidney (2t).
The Na -H exchange system has the biological function to remove H
from cells with a high membrane potential (22). Acfordingly, it seems
unlikely that an interindividual variability of Na -Li+ exchange is
present in n~rve and muscle cells which is comparable to the variabili-
ty of Na+-Li exchange seen in erythrocytes.

451
 8 ----·-a45
I

/ ~. .~·-----·- 7.36
~
~ ..
·1 -------·-
<I>
u 6 /./
E
--:- 6.45
Ill

0
<I>
/. ------ pHe
E 4
::1, /.
w
~

~
a...
::::> 2
./·
+
:.:::i

0 2 3 4
EXTERNAL Li+ [mM)

;+ + .+
Fiaure ?.: L ~ uptake by the Na -L~ excha~ge system of beef erythro-
cytes in dependence of external pH and Li concentration (choline
media, 37°C)

;-,0~------------------------.
I
Km
.r::- • •
pH
I mM I
Vmax
u: [mMI
..!!! 8
Qj
u ~. 8.45 0.42 9.2
v·' 0.5 , / 0.5

E ·~
---.!..
~
~
6
.\. ·~.
I .
0.4

::1, 4

\.
7.36 0.58 9.1
w
~
~ 2
0... 6.45 1.49 7.7
:::>
+
/
:.:J
0 2 4 6 8 10
Li •uPTAKE · [ u;r,
-100 0 100 200 300 400
HYDROGEN ION COIIK:ENTRATION IJJM I

Fiqure 3: Eadie-Hofstee plot of Fiqure 4: Dixon plot of the

the data from Fig. 2. The appar- data from Fig. 2.
ent maximum transport rates (int-
ercepts at the ordinate) and Km
values (mM external Li+, obtained
from the slopes) are included in
the figure.

452
 + +
Na -Li EXCHANGE IN AFFECTIVE DISORDERS

The possibility that a low red cell Na+-Li+ exchange acti~ity

(associated with a high steady-state Li distribution ratio Li ./Li+ )
is characteristic for a subgroup of patients suffering from aff~ctiv~
disorders is dealt with in detail in the contributions of Drs. Grover,
Ostrov, Dorus, Mendlewicz and Davis in this volume. Our data indicat-
ing that there exist no essential differences between the mean in
vivo ratios of unipolar and bipolar patients and the mean in vitro
ratios of patients suffering from affective disorders and control
individuals have been published elsewhere (23) • Taken the informa-
tion avail~ble together it appears in summary as though a determina-
tion of Li transport properties of erythrocytes cannot serve as a
diagnostic or prognostic tool in psychiatry.

Na+-Li+ EXCHANGE IN ESSENTIAL HYPERTENSION

The reverse alteration, namely a twofold acceleration of Na+-Li+

exchange has been claimed to be present in erythrocytes from patients
suffering from essential hypertension (24). The initial report of
Canessa et al. was provocative in showing almost no overlap of the
data for normotensive and es~ential hypertensive patients (see Fig. 2
of Ref. 24). In addition, Na -Li exchange was found to be elevated in
normotensive first-degree relatives of essential hypertensive subjects
(24,25), and normal transport rates were seen in secondary hyperten-
sion (24). The diagnostic and prophylactic value of these data was
obvious.
However, in later reports of the same group (26) and of other
authors a much greater overlap (27) or even no difference (27,28)
between normotensives and essential hypertensive patients was obser-
ved. Accordingly, the idea of an increased red cell Na+-Li+ exchange
being typical for essential hypertension is as questionable as the
idea that the exchange is reduced in manic-depressive (bipolar)
disease.

Na+-Li+ EXCHANGE IN ESSENTIAL HYPERTENSIVE PATIENTS TREATED WITH Li+

Two retro~pective studies are available on the association of

the in vivo Li ratio with hypertension in patients treated with Li+
salts (Ref. 29, and W. Greil and J. Duhm, unp~lished observations).
In both studies the mean in vivo ratio Li+./Li was 0.45.
In the study of McCoy et al. (29) 4 (~%} o3t of 49 patients
were hypertensive. All of the hypertensives had ratios below 0.30.
Such low ratios were found only in 7 (14%} of the 49 patients, indi-
cating that the probability to have hypertension is significantly
increased in Li+-treated patients with an in vivo ratio below 0.30.
In our ~etrospective study reliable data for blood pressure from
80 of 148 Li -treated patients (23) were available. Of the 80 pa-
tients, 15 (19%} were hypertensive, no clinical signs indicating a
secondary hypertension. The mean distribution ratio was 0.43 ± 0.11

453
( ± 1 S.D.) in the hypertensive patients, as compared to
0.46 ± 0.12 in the non-hypertensive patients, the difference being
not statistically significant. 13 of the hypertensive patients were
females with a mean ratio of 0.43 ± 0.12 as compared to a mean
ratio of 0.48 ± 0.13 obtained in 100 females. A ratio below 0.30
was found in 1 (7%) of the 15 hypertensive patients, and in 8 {5%)
among 148 patients not selected according to blood pressure. Our
results thus indicate that the Na+-Li+ exch~nge activity is not
increased (as reflected by a low in vtvo Li ratio) in essential
hypertensive patients treated with Li salts.

IS THE RED CELL Na+-Li+ EXCHANGE ACTIVITY INHERITED?

There are several rep~rts indicating that the interindividual

differences in red cell Na -Li+ exchange activity are inherited (for
references see contributions of Drs. Dorus and Mendlewicz in this
volume). However, the Na -Li+ exchange activity has be~n reported
to be reduced in vi1o af a result of treatment with Li salts (30).
Furthermore, the Na -Li exchange was found to increase in late preg-
nancy (31). In addition, there is a positive correlatio~ b~$ween
the cholesterol content of the red cell membrane and Na -Ll exchange
activity and an acceleration of the exchange in cholostasis (32) .
Finally, the exchange rate was drastically reduced after hemodialysis
in patients with end-stage renal disease and the transport velocity
is possibly dependent on a dialyzable plasma factor (33). Possibly
the exchange is also changed in hyperthyreoidism (K.-H. Gless, perso-
nal communication). All these findings demonstrate that in addition
to genetic faftorf other (hormonal?) parameters participate in deter-
mining the Na -Li exchange activity of human erythrocytes.

IS A HIGH OR LOW RED CELL Na+-Li+ EXCHANGE ACTIVITY AN "ABNORMALITY"

OR A "DEFECT"?

In the literature, a high or low activity of the Na+-Li+ exchange

system of human erythrocytes is often regarded as an "abnormalitv" or
even ~sa "defect". However, the drastic interindividual variability
in Na -Li+ exchange discussed above is a characteristic property of
erythrocytes not only from apparently healthy man and patients suffer-
ing from affective disorders or essential hypertension, but also of
erythrocytes from other species, the greateft v$riability being seen
in beef (6,19). Accordingly, low or high Na -Li exchange activities
represent the lower and higher ends of the normal distribution and,
therefore, cannot be categorized as an "abnormality" (34). Further-
. + + + .+
more, t h e veloc1ty of 1:1 Na -Na o; Na -Ll exchange does not influ-
ence red cell function, and a Na+-H exchange (see above) does not
influence the red cell p~ sine~ the highly active anion exchange
system (mediating the Cl -Hco 3 exchange) rapidly reestablishes the
passive distribution of hydrogen ions between r~d c~lls and plasma
according to the Donnan potential. Thus, the Na -Li exchange system
of erythrocytes has no function vital for the cells. Accordingly, a

454
high of low exchange rate in erythrocytes cannot be considered as a
"defect" in the red cell membrane. At best it could be referred to
as a "variant". Similar arguments have bee!f r~ised concerning the
interindividual variability of red cell Na -K cotransport (35)
which is thought to be genetically determined (36) and to exhibit a
lowered activity in essential hypertension (37) .

ACKNOWLEDGEMENT

The work was supported in part by the Wilhelm-Sander-Stiftung.

REFERENCES

1) Maggs,R., in IVth World Congress of Psychiatry, Part 3 (J.J.Lopez,

ed.), Congress Series No 150, Excerpta Medica, Amsterdam, pp.
2211-2214 (1967)
2) Mendels,J., and A.Frazer, J. Psychiat. Res. 10: 9-18 (1973)
3) Ehrlich,B.E., and J.M.Diamond, Am.J.Physiol. 237: C102-C110 (1979)
4) Ehrlich,B.E., and J.M.Diamond, J.Membrane Biol. 52: 187-200 (1980)
5) Tosteson,D.C., Fed.Proc. 40: 1429-1433 (1981)
6) Duhm,J., in Basic Mechanisms in the Action of Lithium (H.M.Emrich,
J.B.Aldenhoff, and H.D.Lux, eds.), Excerpta Medica, Amsterdam,
pp. 1-20
7) Greil,W., F.Eisenried, B.F.Becker, and J.Duhm, Psychopharmacology
53: 19-26 (1977)
8) Haas,M., J.Schooler, and D.C.Tosteson, Nature (Land.) 258: 425-427
(1975)
9) Duhm,J., F.Eisenried, B.F.Becker, and W.Greil, Pflugers Arch.
364: 147-155 (1976)
10) Duhm,J., and B.F.Becker, Pflugers Arch. 367: 211-219 (1977)
11) Duhm,J., and B.F.Becker. Pflugers Arch. 370: 211-219 (1977)
12) Pandey, G.N., B.Sarkadi, M.Haas, R.B.Gunn, J.M.David, and
D.C.Tosteson, J.Gen.Physiol. 72: 233-247
13) Sarkadi,B., J.K.Alifimoff, R.B.Gunn, and D.C.Tosteson, J.Gen.
Physiol. 72: 249-265 (1978)
14) Ussing,H.H., Nature (Land.) 160: 262- (1947)
15) Szentistvanyi,I., Z.Janka, F.Joo, A.Rimanocza, A.Juhasz, and
L.Latzkovits, Neurosci.Lett. 13: 157-161
16) Reiser, G., and J.Duhm, Brain Res. 252: 247-258 (1982)
17) Smith,I.C.H., J.Physiol. (Land.) 242: 99P-101P (1974)
18) Motais,R., J.Physiol. (Land.) 233: 395-422 (1973)
19) Duhm,J., and B.F.Becker, J.Membrane Biol. 51: 263-286 (1979)
20) Funder,J., J.O.Wieth, H.A.Jensen, and K.K.Ibsen, In Recent
Advances in Hypertension Mechanisms (in press)
21) Aronson,P.S., New Engl.J.Med. 307: 317 (1982)
22) Roos, A., and W.F.Boron, Physiol.Rev. 61: 296-434 (1981)
23) Greil,W., B.F.Becker, and J.Duhm, In Lithium, Controversies and
Unresolved Issues (T.B.Cooper, S.Gershon, N.S.Kline, and
M.Schou, eds.), Excerpta Medica, Amsterdam, pp. 209-217 (1979)

455
24) Canessa,M., N.Adragna, H.S.Solomon, T.M.Connolly, and
D.C.Tosteson, New Engl.J.Med. 302: 772-776 (1980)
25) Woods,J.W., R.J.Falk, A.W.Pittman, P.J.Klemmer, B.S.Watson, and
K.Namboodiri, New Engl.J.Med. 306: 593-595 (1982)
26) Adragna,N.C., M.L.Canessa, H.Solomon, E.Slater, and
D.C.Tosteson, Hypertension 4: 795-804 (1982)
27) Ibsen,K.K., H.AE.Jensen, J.O.Wieth, and J.Funder, Hypertension
4: 703-709 (1982)
28) Duhm,J., B.O.Gobel, R.Lorenz, and P.C.Weber, Hypertension 4:
477-482 (1982)
29) McCoy,S.N., J.X.Thavundayil, G.Schwartz, and P.Etienne, Am.J.
Psychiatry 139: 247-248 (1982)
30) Meltzer,H.L., S.Kassir, D.L.Dunner, and R.R.Fieve, Psychopharma-
cology 54: 113-118 (1977)
31) Worley,R.J., W.M.Hentschel, C.Cormier et al., New Engl.J.Med.
307: 412-416 (1982)
32) Clegg,G., J.Gledhill, and D.B.Morgan, IRCS 10: 828-829 (1982)
33) Woods,J.W., J.C.Parker, and B.S.Watson, New Engl.J.Med. 308:
1258-1261 (1983)
34) Duhm,J., B.F.Becker, and W.Greil, In Biological Psychiatry
Today (J.Obiols et al., eds.), Elsevier, North-Holland Bio-
medical Press, pp. 1137-1142 (1979)
35) Stewart, A.D., Lancet i: 500-501 (1982)
36) Meyer,P., R.P.Garay, C.Nazaret, G.Dagher, M.Bellet, M.Broyer,
and J.Feingold, Br.Med.J. 282: 1114-1117 (1981)
37) Garay,R.P., G.Dagher, M.-G.Pernollet, M.-A.Devynck, and
P.Meyer, Nature (Lond.) 284: 281-283 (1980)

456
LI TRANSPORT AND THE PSYCHOSES

David L. Garver, Robert Hitzemann and

Jack Hirschowitz

University of Cincinnati College of Medicine

231 Bethesda Avenue, Cincinnati, Ohio 45267

INTRODUCTION

Previous studies have suggested that intracellular to

extracellular concentrations of Li across RBC membranes (Li ratios)
may be useful in identifying Li responsive affective illnesses.
Though agreement is not unanimous, 1 several authors have reported
that higher Li ratios obtained in RBC exposed to Li in vitro or in
vivo are associated with short term antimanic or antidepressant
effects of Li, or with successful prophylaxis from affective
episodes.2-5 Higher Li ratios, when determined by direct
measurements of RBC as well as plasma Li,6-10 as well as
impaired countertransport of Li from the RBC,ll appear to occur
with a high frequency among affectively disordered patients who
experience both manic and depressive episodes (bipolar disorder).

Intracellular RBC concentrations of Li and consequent intracell-

ular: extracellular Li ratios are predominantly the consequence of
extrusion of Li from cells by a sodium-lithium counterflow mechan-
ism.12 This membrane associated transport system moves Li
against its concentration gradient from an intracellular position
extracellularly. The energy required for such Li counterflow is
derived from the electrochemical gradient of high extracellular to
low intracellular sodium; the flow of sodium via the counterflow
mechanism into the cell is coupled to the counterflow of Li out-
ward.l3,14 Reduced activity of the counterflow channel,
as previously documented in some bipolar patients, 11 • 14
results in excessive accumulation of Li intracellularly and higher
intracellular to extracellular Li ratios. Increased activity of
the counterflow channel has been reported in some patients with
essential hypertension15,16 in whom such increased

457
extrusion and decline of the intracellular lithium concentration
may result in lower than normal intracellular to extracellular Li
ratios. Since activity rates of the counterflow system and conse-
quent Li ratios are relatively stable within subjects,9,11
and are under genetic control,17,18 Li ratios have been
used as a biologic marker for trait disturbances in ion
transport .17-19

Is the Li counterflow abnormality and, consequently, the Lith-

ium Ratio (LR) abnormality limited to affective disorder patients?
Such considerations prompted us to investigate the distribution of
the RBC lithium ratio in vitrq (LR) in 141 consecutively admitted
psychiatric patients.

METHODS

One hundred and forty-one consecutively admitted consenting

psychiatric patients and 38 age and sex matched controls underwent
LR studies concurrently. The controls (but not the patients) were
screened for the absence of hypertension and of psychiatric dis-
orders in themselves and in first degree relatives. None of the
patients or controls had received Li during the three months prior
to the study. A DSM-IH diagnosis could be made on the basis of a
SADS interview in 107 of the 141 acutely hospitalized patients.
Mean systolic and diastolic blood pressures were documented retro-
spectively from the hospital charts of 106 of the 141 acutely
hospitalized patients.

Li ratio (in vitro) assays were performed on the RBC from

patients who had been drug free for at least 48 hours and fasted
overnight. Blood for RBC assays was drawn 48 to 72 hours following
admission in 98 of the 141 in the survey cohort. LR determinations
were performed essentially according to the method of Pandey, et
al.20 The RBC were pre-loaded in a medium containing 60 mM
Li.Cl for one hour prior to a 24-hour incubation in a medium
containing 1.5 mM LiCl. Li concentrations in the RBC lysate and in
the 24-hour incubation media were assayed by atomic absorption
spectroscopy. Reported LR reflect the mean of triplicate Li
RBC/media ratio determinations for each subject, all done at one
time on the same blood sample. The mean (±SD) coefficient of
variation within assays was 4. 7 ± 0. 9%. Multiple measures of the
LR performed over a 2-year period in four control subjects show a
coefficient of variation over time of 3.9 ± 1.5%.

RESULTS

The distribution of the Li ratios in vitro (LR) of the 141

consecutively admitted hospitalized patients is shown in Figure
1a. The mean LR (±SD) was 0.35 ± 0.10 (n=141).

458
 0

.22 ~0 ~8 A6 ~4 .62
1(a) Li Ratio of 141 Consecutive Hospital Admissions

.22 .30 .38 .46

1(b) Expected Middle Component (0.34-0.04;n=82)
of Li Ratio Distribution of above

8 ,as,&, 88o8oo 8

.22 .30 .38 .46

1(c) Li Ratio of 38 Controls

Fig. 1. Li ratios in hospitalized psychiatric patients and in Li

treated psychotic patients; (a) Lithi11m rat io (in vitro) of
141 conse c11tively admitted, consenting psychi.atric patients;
(b) expected distr.ib11tion of the mi.ddle component (0.34 .±
0.04; n=82) of the trimodal distrib11Uon of Li ratios f r om
the consec11tively admitted patients; (c) Li ratio (in vitro)
of 38 controls screened for psychiatric illness in
themse lves as i.n t heir first degree relatives.

459
 The distribution of LR was tested for goodness of fit to a
normal distribution having the same mean and SD, utilizing a
freq•1ency interval of 0.03. The observed distribution differed
from normality at a high level of significance <x12 = 42.8, p <
.0001; n=l41).

By the method suggested by Bhattacharya,21 the distribu-

tion could be resolved into three components. The mean, SD and
freqnency for each component were as follows: 0.24 .± 0.03 (n=32);
0.34 .± 0.04 (n=82) (Figure lb); 0.51 .± 0.04 (n=27).

The expected frequencies for normally distributed variates

with these means, SDs and frequencies were calculated to obtain the
expected values for the total distribution. A chi-square goodness
of fit to this distribution yielded xy 2 = 3.68, p > .50 (n=l41).
Thus, the decomposition had yielded a good fit to the observed fre-
quencies. Roughly, 28% of cases fall into the lowest distribution,
58% into the middle distribution and 19% into the upper distribu-
tion.

LR values for 38 concurrently studied normal controls averaged

0.35 .± 0.06 (Fig•Ire 1c}. This mean value of the LR in controls did
not differ significantly from either the total group of the 141
consecutively admitted patients (0.35 .± 0.10), nor the middle
segment (0.34 .± 0.04; n=82), shown in Figure lb. However, the
magnitude of the variance for the total patient group (0.010) was
more than 3 times the variance of the controls (0.0036).

Using the estimated mean and SD from the middle distribution

just defined (0.34 .± 0.04), expected frequencies were obtained and
compared with the observed freq•1encies for the control group. The
resulting xt = 23.27 was highly significant (p < 0.001). The big-
gest difference occurred because of the excess of control subjects
(4.6) who had LR greater than 0.425. It, therefore, appears that
about 12% of controls also may belong to the population of individ-
uals with a high LR. Only about 6% could be estimated to have come
from the population of low LR.

Variables Potentiallv Associated with Altered LRs

The _lc!pper distribution of LR in the recently admitted psy-

chotic patients does not appear to be merely a consequence of prior
or concomitant antipsychotic drug medication:22,23eight
patients who had been antipsychotic drug-free for at least 120 days
had an elevated mean LR of 0.43. Neither did the presence of
hypertension24,25 (mean diastolic pressure greater than
90 mmHg) appear to be associated with the lowest LR distribution.
Patients with mean diastolic pressure greater than 90 mmHg had a
mean LR of 0.36 .± 0.09 (n=13), while normotensive patients averaged
0.35 .± 0.11 (n=93).

460
 There were no significant differences between the mean LR of
DSH-In26 schizophreniform disorders (0.36 _± 0.12; n=21) and
schizophrenics (0.36 _± 0.10; n=54). Smaller sample size prevented
clear evaluation of possible differences between these groups and
schizoaffectives (0.34 _± 0.06; n=7), manics with mood-incongruent
psychotic features (0.35 _± 0.08; n=5), manics without psychotic
features or with mood-congruent psychotic features (0.34 _± 0.08;
n=4), 1mipolar major depressive disorders with mood-incongruent
psychotic features (0.31 _± 0.08; n=3), non-psychotic unipolar major
depressive disorders (0.27 _± 0.07; n=4), and other diagnosis,
including atypical affective disorder, histrionic and borderline
personality disorders (0.42 _± 0.10; n=9).

REFERENCES

1. Hendlewicz J, Verbanek P: Lithium ratio and clinical response

in manic-depressive illness. Lancet 1:41, 1977.
2. Hendels J, Frazer A: Intracellular lithium concentration and
clinical response: Toward a membrane theory of depression. ~
Psvchiat. Res. 10:9-18, 1973.
3. Casper RC, Pandey N, Gosenfeld L, Davis JH: Intracellular
Lithium and Clinical Response. Lancet 2:418-419, 1976.
4. Flemenbaum A, Weddige R, Hiller J: Lithium Erythrocyte/Plasma
Ratio as a Predictor of Response. Am. J. Psvchiat. 135:336-338,
1978.
5. Delvecchio H, Farzati B, Haj H, Hinucci P, Kemali D: Cell
membrane predictors of response to lithium prophylaxis of affective
disorders. Neu~Q~~y~hobiology 7:243-247, 1981.
6. Cazullo CL, Smeraldi E, Sachetti E, Bottinelli S: Intracellu-
lar lithium concentration and clinical response. Br. J. Psvchiat.
126:298-300, 1975.
7. Lyttkens L, Soderberg U, Wetterberg L: Relationship between
erythrocyte and plasma lithium concentrations as an index of
psychiatric disease. Ups. J. Hed. Sci. 81:123-128, 1976.
8. Ramsey T, Frazer A, Dyson WL, Hendels J: Intracellular
lithium and clinical response. Br. J. Psych~Cit!_ 128:103-104, 1976.
9. Frazer A, Gottlieb J, Mendels J: Lithium ratio and clinical
response in manic-depressive illness. Lancet 1:41-42, 1977.
10. Ramsey TA, Frazer A, Mendels J, Dyson, WL: The erythrocyte
lithium-plasma lithium ratio in patients with primary affective
disorder • .Arch. Gen. Psvchi~t~ 36:457-461, 1979.
11. Ostrow DG, Halaris A, DeHet E, Gibbons R, Davis JH: Ion
transport and adrenergic function in major affective disorders.
Biol. Psvchiat. 17:971-980, 1982.
12. Grell W, Eisenried F, Becker F, Duhm J: Interindividual
differences in the NA-t-dependent Li+ countertransport system and in
the Li+ distribution ratio across the red cell membrane among Li+
treated patients. Psvchopharm •. 53:19-26, 1977.

461
13. Haas M, Schooler J, Tosteson DC: Coupli.ng of lithium to
sodium transport 1n human red cells. Nature 258:425-427, 1975.
14. Pandey GN, Sarkadi R, Gunn RB: Lithium transport pathways in
human red cells. Psvchooharm. Bull. 14:16-19, 1978.
15. Canessa M, Adragna N, Solomon HS, Connolly TM, Tosteson DC:
Increased sodium-lithil.Dil countertransport in red cells of patients
with essential hypertension. New Engl. J. Med. 302:772-776, 1980.
16. Ostrow DG, Trebesan M, Okonek A, Gibbons R, Cooper R, Davis
JM: Sodium dependent membrane processes in major affective
disorders, in Usdin E. and Hanin I. (eds.): Biological Markers in
Psychiatry and Neurology, Pergamon Press, New York, 1982 pp.
153-168.
17. Dorus E, Pandey N, Davis JM: Genetic determinant of lithium
ion distribution. Arch. Gen. Psvchiat. 32:1097-1102 1975.
18. Pandey GN, Ostrow DG, Haas M, Dorus E, Casper RC, Davis M,
Tosteson DC: Abnormal lithil.Dil and sodil.Dil transport in erythrocytes
of a manic patient and some members of his family. Proc._Natl.
~cad. Sci. USA. 74:3607-3611, 1977.
19. Dorus E, Pandey GN, Shaughnessy R, Davis J: Lithium transport
across the RBC membrane. Arch. Gen. Psychiat. 37:80-81, 1980.
20. Pandey GN, Baker J, Chang S, Davis JM: Prediction of in vivo
red cell/plasma lithil.Dil ratios in vitro methods. Clin. Pharm.
Ther. 24:343-349, 1978.
21. Bhattacharya CG: A simple method of resolution of a
distribution into Gaussian Components. Biometrics. 23:115-137,
1967.
22. Pandey GN, Goel I, Davis JM: Effect of neuroleptic drugs on
lithium uptake by the human erythrocyte. Clin. Pharmacol. Ther~
26:96-102, 1979.
23. Ostrow DG, Southam A, Davis JM: Lithium-drug interations
altering the intracellular lithhun level: An in vitro study.
Biol. Psvchiat~ 15:723-739, 1980.
l4. Canessa M, Adragna N, Solomon HS, Connolly TM, Tosteson DC:
Increased sodium-lithium countertransport in red cells of patients
with essential hypertension. New En2l. J. Med. 302:772-776, 1980.
25. McCoy JN, Thavundayil JX, Schwartz G, Etienne P: Lithium
ratio and hypertension in manic-depressed patients. Am. J~
Psychiat. 139:247-248, 1982.
26. American Psychiatric Association: Diagnostic and
Statistical Manual of Mental Disorders. Third Edition (DSM-Ili),
The American Psychiatric Association, Washington, DC, 1980.

462
LITHIUM TRANSPORT IN MAJOR AFFECTIVE DISORDERS: A MODEL SYSTEM FOR

MEMBRANE-PLASMA AND GENETIC-PHYSIOLOGIC INTERACTIONS

David G. Ostrow, Anna Okonek, Walter Dorus,

John M. Davis and Walter Kitt

Departments of Psychiatry and

Community Health & Preventive Medicine
Northwestern University Medical School
VA Lakeside Medical Center

Department of Research
Illinois State Psychiatric Institute
Chicago, IL 60611 USA

As mentioned in the introduction to this symposium and in Dr.

Duhm's paper, there appear to be both genetic and physiologic
influences on the rate of sodium-dependent lithium efflux in human
red blood cells (RBCs). We have been interested in biological
measurements for diagnosing and treating patients with major affec-
tive disorders, but are still quite far from the time when we will
be able to use this or any biological marker for definitively
diagnosing or predicting treatment response. Our investigations
have begun to unravel several physiologic and genetic influences on
this measurement, a necessary prerequisite for its clinical appli-
cation. Our findings of the observed influences on lithium trans-
port suggest its use as an example of a sodium-dependent membrane
function that can be influenced by both genetic and physiologic
parameters (Fig. 1). Genetic factors control the amount of
synthesized peptide, its primary structure, and hence its efficien-
cy in catalyzing the exchange of sodium and lithium across the RBC
membrane. Membrane factors not directly under genetic control are
the rate of cell membrane turnover and the composition of the non-
protein portion of the membrane. From the plasma side of the
membrane we Rce the influence of a variety of externally-associated
factors ranging in size from very small molecules, such as peptides
and steroid hormones, to very large lipoprotein complexes known to

463
be in active, dynamic equilibrium with the lipoprotein components
of the RBC membrane (1). Transfer proteins also regulate membrane
structure and function by controlling the rate of incorporation of
plasma components into the cell membrane (2). This model attempts
to describe the RBC membrane as a dynamic entity, which, although
it has certain stable genetically-controlled properties, is
susceptible to a number of time-dependent physiologic
perturbations that may either increase or decrease the relative
rate of ion transport, and hence affect intracellular to extra-
cellular ion distribution ratios. Using this model system we can
explore these interactions, their relationship to the ion balance
of the cell, and their potential clinical significance.

MODEL

.. l
MEMBRANE PLASMA

Regulatory Peptides
GENETIC
~ Hormones
·::_:":: Small Molecules

Protein Primary
Structure Transfer Proteins

NON-GENETIC

l
Cell/Membrane
Turnover @ Lipoproteins
~ @ @ Phospholipids
@ Sterols

Fig. 1. Model of Membrane-Plasma and Genetic-Physiologic

Effects on Rate of Red Cell Lithium Efflux

The genetic aspects of lithium transport are covered elsewhere

in this symposium, so we will only mention one family study that
demonstrates the degree of genetic control seen in this system.
In a young black male with manic-depressive illness we found a
complete absence of RBC lithium efflux and a lithium ratio of 1.0,
consistent with this pathway being the controlling factor in the
lithium ratio. When we examined his first-degree family members,
a similarly low level of lithium transport was correlated with
the presence of major psychiatric illness in the patient's father
and several siblings (3). This case also demonstrated the possi-
bility that there might be father-to-son transmission of both a
biological perturbation and major affective disorder, and contra-
dicted the X-linked model of manic-depressive illness transmission
that was popular when these results were first published (4).

464
Dr. Elizabeth Dorus has since expanded on this study and shown a
significant correlation between the lithium ratio and the presence
or absence of major affective illness in a large number of family
pedigrees (5).

The above example of a dramatic alteration in lithium trans-

port caused by a genetically controlled mechanism is not the usual
case in studies of large groups of subjects. We have observed" a
wide distribution of lithium ratios and efflux rates in both
patients and controls (6,7). Our current interest is in exploring
the question of whether or not some portion of this individual
variation may be related to psychiatric diagnosis. To assess this
we must first discern which individual physiologic differences
might account for variations in lithium transport rates, and
methodologically control for them before considering clinical
variables. Table 1 lists the variables that we have been studying.
We can summarize our findings by saying that all of the variables
shown on this list, with the exception of diet, have been shown by
us to have a significant influence on the rate of lithium efflux in
human RBCs. Age alone has no effect, but age-dependent physiolo-
gic factors, such as hypertension and obesity, do influence lithium
transport over time.

Table 2 shows the characteristics of our study group of 266

affective disorder patients. When lithium efflux rates of
patients, 220 controls, and total sample are examined, we consis-
tently observe a non-Gaussian, multi-modal distribution of values
(8,9). To determine whether or not these distributions are
composed of a mixture of normal component distributions, we have
used the approach of maximum likelihood mixture analysis according
to Day's formulas (10). Using this method, each group studied has
been found to contain three distinct normal component distribu-
tions, or modes, that fit the observed data (Fig. 2). The range of
lithium efflux values in the lowest mode is approximately 1 to 5
ueq/l.cells/min, that of the middle mode is from 5 to 8 ueq/
l.cells/ min, and values in the highest mode are those above 8 •

• Age • Diet
• Sex • Major Medical Illness
• Race Medication
• Blood Pressure • Alcohol Intake
• Body Mass (kg/m2) • Drug Abuse

465
 Table 2 Descriptive Statistics of Affective Disorder Group

Variable Mean + Standard Deviation (N=266)

Age 42.4 ;t 14.3
Height (in . ) 67.9 ± 3.7
Weight (lbs.) 162.9 ± 34.7
Body Mass Index 25.3 ± 4.5
Systolic BP 75.7 ± 10.8
Diastolic BP 117.2 ± 15.6
Lithium Efflux 4.9 ± 2.0
Race (W:B:O) 82:12:6
Sex (M:F) 67:33

20 DIITIIISUTION STATISTICS
TOTAL H 163

P Me•n STDOEY
16 1 0 .140 3 . 710 1.000

2 o.2eo e.uo o.eoo

:0. 0 . 100 1 . 150 1 .000
u 12 3
c
•
::ll

•...
0'

~
8

0
1.50 3.50 5 .50 7 .50 9.50 11.50 1 3 .50
Net Lithium Efflux

Fig. 2 Trimodal Distribution Pattern of RBC Lithium Efflux Rate

in Population of Major Affective Disorder Patients.

Squares represent the distribution of observed efflux values.

while shaded areas indicate the modeled distributions according
to Day's analysis (10).

Table 3 Modal Analysis of Physiolo~ic Variables

Variable Low (N:o:95) Middle (N=36) Hi2h (N=14)

Lithium Efflux 1.50 - 5.25 5.26 - 7.75 7.76 - 11.50
Height 68 ± 4 69 ± 4 68 ± 3
Weight 162 ± 33 166 ± 37 197 ± 27
Body Mass Index 25 ± 4 25 :t: 4 31 ± 6
Systol i c BP 116 ± 14 123 ± 13 127 ± 15
Diastol ic BP 73 ± 10 80 ± 1:0: 83 ± 12
Diagnosed
Hype rt e n s ion 7"1. 28~ 36%

466
 We carefully examined each mode of the patient group to see if
there were any physiologic factors that might explain the variation
in lithium efflux values. We found the most significant differen-
ces among groups related to blood pressure and body mass (Table 3).
Both blood pressure measurements and clinical diagnosis of hyper-
tension distinguish the middle and high group from the low mode,
and body mass index, a ratio of height to weight, differentiates
the high mode from both the low and middle groups. This is not
surprising, since there is much evidence suggesting that cardio-
vascular and metabolic disorders may alter ion transport mecha-
nisms. Sex and race also appear to affect transport rates, as
shown with data from our patient sample in Table 4. Males consis-
tently have higher RBC lithium efflux rates than females, and
whites have significantly higher levels than blacks. We found no
significant interaction of sex and race. A significant sex by
hypertension interaction was observed, showing a more pronounced
effect of blood pressure on lithium efflux in males than in females
(Table 5). Race and hypertension did not appear to significantly
interact, but larger sample sizes with more even racial distribu-
tions are needed to elucidate these higher order interactions.

Table 4 Lithium Efflux Values by Sex and Race in Patient Sample.

Mean ± S.D. N Student t

Males 5.44 ± 0.49 110

Females 4.31 ± 0.64 84 10.05, p < 0.002

Whites 5.08 ± 1. 95 228

Blacks 3.76 ± 1.61 28 3.43, p < 0.001

Table 5 Second-Order Interactions bv Analvsis of Variance

Whites Blacks

Males 5.61 (N•90) 3.82 (N=l3)

Females 4.41 (N=77) 2.99 (N•6) n.s.

Unmedicated Medicated
Normotensives Hypertensives Hypertensives

Males 4.97 (N=86) 7.08 (N=16) 7.19 (N•8)

Females 4.24 (N•75) 4.78 (N=5) 4.96 (N•4) p<0.05

Unmedicated Medicated
Normotensives Hypertensives Hypertensives

Whites 4.73 (N .. 139) 6.70 (N•18) 6.67 (N=10)

Blacks 3.37 (N=l5) 3.15 (N .. 2) 5.36 (N•2) n.s.

467
 After discerning the effects of physiologic factors on lithium
efflux values, we turned to clinical factors that might possibly
explain the wide degree of observed variation. Approximately 20%
of our affective disorder patients have a concurrent diagnosis of
alcohql abuse, which not only complicates the diagnostic picture,
but is known to be related to hypertension (11) and cell membrane
alterations as well. Table 6 shows that alcoholism does indeed
have a significant effect on the rate of lithium efflux in this
group. Analysis of covariance was performed on 183 cases after
multiple regression discerned the effects of sex, body mass, and
diagnosed hypertension on efflux measurements, and four contrasts
are shown. Patients with a concurrent diagnosis of alcoholism had
a significantly higher level of RBC lithium efflux than non-
alcoholics. Alcoholics in remission had an intermediate mean level
of efflux that was not significantly elevated over either current
or non-alcoholic groups.

Table 6 Effect of Alcoholism on Lithium Transport

In Patient Group (N•183) (Analysis of Variance)

No Current Alcoholism in
Alcoholism Alcoholism Remission

N•l44 N•25
X 4.79 6.07

I I I F=11.23, p <0.03
I I Fa7.26, p (0.008
I
------------------------------------
I
I
I
F•0.65, ns
F•5.92, ns

These results illustrate the necessity to examine in detail

each physiologic factor that may appear important in the overall
group. Before we can correlate lithium transport measurements with
diagnosis or treatment prediction, we must be able to determine
which physiologic factors need to be controlled for in a given
individual.

In an accompanying paper in this volume we present our results

of the possible diagnostic significance of lithium efflux. While
we were able to show that there are significant differences in the
corrected group mean lithium efflux values between unipolar and
bipolar patient groups, we must cautiously interpret this finding.
First, those studies were done before our recent appreciation of
the potential effects of alcohol use on lithium transport;
therefore this variable was not controlled for in the analysis.
Second, while mean differences are apparent in the patient groups
once the lithium transport rate has been corrected for some, but
not all, of the physiologic factors, these observations may not be
applicable to individual patients.

468
 Finally. we wish to briefly discuss our findings that suggest
circulating compounds are important in determining an individual's
level of RBC lithium efflux. We have shown (9,12) that female
patients with lithium-resistant rapid-cycling manic-depressive
illness show periodic variations in lithium ratio and transport
parameters that parallel their affective cycles. Plasma from such
patients can inhibit lithium efflux and stimulate lithium influx in
both patient and normal control RBCs. with the level of transport
modulating activity dependent on the patient's mood state. These
results again underline the need for careful physiologic assessment
in biological marker research 1 and suggest that there may be
circulating compounds as yet uncharacterized that are produced or
released in a time-dependent manner reflecting the mood state of
the individual.

Table 7 Factors Affecting Lithium Transport

Short Term (Minutes - Weeks)

Plasma Factors
Menstrual Cycle
Rapid Cycling Illness

Intermediate (Months - Years)

Membrane Factors
Lipid Composition
Alcohol
Medication
Pregnancy

Long Term
Hypertension
Obesity
Diabetes
Other?

In conclusion. we feel that these results illustrate the

dynamic nature of lithium transport and the presenc~ of a variety
of factors influencing its level in a particular individual over
time. The known factors and their approximate time-dependency for
affecting lithium efflux are summarized in Table 7. By exploring
the mechanism of these membrane-plasma and genetic-physiologic
interactions affecting membrane function. we hope to better under-
stand the mechanisms by which genes, environment. and stress inter-
act to influence the biological basis of psychiatric illness.

REFERENCES

1. B. Roclofsen, G. V. Meer. and J.A. F. Op Den Kamp 1 The lipids

of red cell membranes. Scan. J. Clin. Lab. Invest. 41 Suppl:
111-11 5 (1981 ) •

469
2. K. A. Muczynski, W. E. Harris, and W. L. Stahl, Altering
erythrocyte membrane composition with phospholipid exchange
protein, ~nt. J. Biochem. 13:959-962 (1981).
3. D. G. Ostrow, A. Okonek, s. Sparks, s. Flagel, R. Cooper,
and J. M. Davis, Red blood cell lithium efflux in major
affective disorders,Syllabus & Scientific Proceedings in
SummRrv Fnrm_ 1'lt:h MPPt:in<> AmPrican Psvchiat:ric Association
180 (1982).
4. J. Mendelwicz, J. Raines, Morbidity risk and genetic
transmission in manic-depressive illness, Am. J. Hum. Genet:.
26:692-701 (1974).
5. E. Dorus, N. J. Cox, R. D. Gibbons, R. Shaughnessy, G. N.
Pandey and C. R. Cloniger, Lithium ion transport and affec-
tive disorders within families of bipolar patients, Arch.
Gen. Psychiatry 40:545-552 (1983).
6. D. G. Ostrow, A. Halaris, E. DeMet, R. Gibbons, and J. M.
Davis, Ion transport and noradrenergic function in major
affective disorders, ~inlo2ical Psy~hiatry 17:971-980
(1982).
7. M. Trevisan, D. G. Ostrow, R. Cooper, K. Liu, S. Sparks, A.
Okonek, E. Stevens, J. Marquardt, and J. Stamler, Abnormal
red blood cell ion transport and hypertension: The Peoples
Gas Company Study, Hypertension 5:363-367 (1983).
8. D. G. Ostrow, A. Okonek, R. Gibbons, R. Cooper, and J. M.
Davis, Biological markers and antidepressant response, J.
Clin. Psychiatry (in press, 1983).
9. D. G. Ostrow, M. Trevisan, A. Okonek, R. Gibbons, R. Cooper,
and J, M. Davis, Sodium dependent membrane processes in
major affective disorders, in: "Biological Markers in
Psychiatry and Neurology," E. Usdin and I. Hanin, eds.,
Pergamon Press, Oxford (1982).
10. N. E. Day, Estimating the components of normal distribu-
tions, Biometrika 56:463-474 (1969).
11. G. S. Stokes, Hypertension and alcoholism: Is there a link?
J. Chronic Dis. 35:759-762 (1982).
12. D. G. Ostrow and J. M. Davis, Laboratory measurements in the
clinical use of lithium, Clinical Neuropharmacol~gy 5:317-
336 (1982).

Acknowledgem~~~~

Dr. Ostrow is the recipient of VA Merit and Career Development

Awards in support of these studies. In addition, this work has
been supported by grants from Coon Memorial and Earl M. Bane
Biomedical Research Funds of Northwestern University Medical
School, and the Chicago Community Trust. Mssrs. Steve Sparks and
Derald Halverson performed the lithium transport assays, while Drs.
Richard Ferm, Steven Flagel, Carlos Gonzales, Sharon Zimmerman, and
Richard Zweig performed the patient clinical evaluations.

470
LITHIUM TRANSPORT IN BRAIN CELLS

AND HUMAN ERYTHROCYTES

Istvan Szentistvanyi and Zoltan Janka

Department of Neurology and Psychiatry

University Medical School of Szeged
Szeged, P.O.Box 397, Hungary

INTRODUCTION

Many data have been accumulated, largely due to

the Chicago and Munich groups, over the last de~ade re-
garding the basic mechanisms of lithium ion /Li I trans-
port across the red cell membrane /Duhm and Becker,
1977; Pandey et al., 1978; Tosteson, 1981/. The impor-
tant discov~ry is that the steady state intra: extra-
cellular Li concent~ati~n ratio in the human blood is
determined by the Na -Li counterflow system /Haas et
al., 1975; Duhm et al., 1976/. Clinically and physiolo-
gically important new directions have been emanated
from studying this transport system in patients with
manic-depressive illness and their relatives /Ostrow et
al., 1978; Pandey et al., 1979; Ramsey et al., 1979;
Szentistvanyi and Janka, 1979 a, b; Mendlewicz and
Verbanck, 1981; Dorus et al., 1979, 1983/, in popula-
tions with other medical diseases believed to be
partially determined by genetic factors /Canessa et al.,
1980/ and also in the light of environmental and
pharmacological influences /Ostrow et al., 1980/.
Att~mpts to clarify the pathways for transmem-
brane Li movement in brain cells ha~e at least a tri-
ple significance: i./whether the Li transport machin-
ery being present in the red cell membrane exists in
neuron and glia, ii./ if additional pathways serve in
traversing of Li+ across the neural membranes and iii./
to what extent these transport mechanis~s contribute
to the therapeutic mode of action of Li .
Progress has been achieved in this field recently.
471
In several labs in America /Rochester, MN; Galveston,
TX/, in Germany /Wurzburg; Munich/ and in Hungary
I Szeged/, studies were performed in the search to char-
acterize Li+ transport pathways in different nerve
cell model systems.
Nerve tissue and cell cultures provide ample op-
portunities for studying the morphological, physiolog-
ical, biochemical and pharmacological properties of
neural cells in vitro /Nelson, 1975/. These techniques
have been particularly useful for extension of Li+
transport research to cellular elements of neural ori-
gin. As a resul$, data are available regarding membrane
transport of Li in nerve cells which we briefly
outline in the following section of this paper.
Li+ TRANSPORT PATHWAYS IN NERVE CELL MEMBRANES
Although the human red blood cell is an excellent
model for investigating basic mechanisms of membrane
transport processes, its difference compared to neuron
is prominent in several aspects: i./ red cell is nonex-
citable~ lacking voltage-dependent action potential Na+
/and Ca +1 channels /in this respect it is similar to
glia/, ii./ its membrane potential is -10 mv differing
markedly from neuron and glia /-60 mV/, iii./ it lacks
membrane receptors and receptor-coupled ion channels
and cyclic nucleotide system, iv./ it differs from neu-
rons in transport machinery for non-electrolytes. In
the view of point ii./, assuming a passive distribution
of Li+ between the intra: extracellular spaces in nerve
tissue, it follows from the Nernst equation that the
intra: extracellular Li+ concentration ratio should be
expected to be about 5-10. /This value is, however,
about 1.2 for the red cells./ No data indicate, however,
such a high Li+ ratio in neural cells. Not even in a
line of studies showing a considerable accumulation of
Li+ by rodent neuroblastoma and glioma cells /Gerkin
and Richelson~ 1979, 1981/. Nevertheless, other authors
have found Li ratio being below 1.0 in neuroblastoma x
glioma hybrid cells /Reiser and Duhm, 1982/ and in pri-
mary brain cell cultures /Szentistvanyi et al., 1979 a;
Janka et al., 1980 a, c/. This was also found by
measurements with Li+-selective microelectrodes in snail
neurons /Thomas et al., 1975/ and in frog spinal
motoneurons /Grafe et al., 1982/. +
Another point in the regulation of Li homeostasis
in the nervous system is the difference in Li+ trans-
port between neuron and glia. It was observed that gli-
oma cells accumulate more Li+ compared to neuroblastoma
cells /Gerkin and Richelson, 1979; Saneto and Perez-Polo,

472
 7

-
c:
Gl

~ 5
Q.

Cl
E 4
'E
~ 3
(")
0
~ 2

2 4 6 8 12 16 20
TIME (min I

Fig. 1. Time course of 22 N~ uptake by nerve cell

cultures during Li release. /Initial in-
tracellular content: 150 nmol/mg protein./

1982/. Our findings, however, revealed an opposite

tendency in primary cultures of neurons and glia /Janka
et al., 1980 b/. We also observed a pr~ferential
vulnerability of neurons exposed to Li in culture
/Janka et al., 1979, 1981/.
1. Na+-Li+ counterflow

+ Reducing the external Na++concentration increased

Li uptake and steady-state Li level in neural cells
/Sze¥tistvanyi et al.,+l979 a, b; Janka et al., 1980 c/.
A Na -dependency in Li efflux has also been shown re-
cently which was ouabain-insensit~ve phenomenon /Rei-
ser and Duhm, 1982/. Mor~over, Na uptake was signifi-
cantly accelerated in Li -loaded brain cells which is
demonstrated in Fig. 1. It appears on the basis of data
obtain~d o¥ trans-stimulation and cis-inhibition that
the Na -Li counterflow operates in the membranes of
neuroblastoma x glioma hybrid cells /Reiser and Duhm,
1982/ and ot chick brain cells /Janka et al., 1980 c/.
This NaT-Li exchange pathway is likely similar to
that described in erythrocytes.

473
 Lithium Transport in Brain Cells

Table 1. Volume densities of neuronal elements in

cultures exposed to neurotoxins in different
ionic media /mean + SEM/

Medium n Volume density

+
Na+ 21 0.253 + 0.024
Na+-veratridin e 21 0.502 + 0.034a
Na -veratridine-TT X 18 +
o. 306 0.037
Li+ 22 0.293 .±. 0.026b
Li!-veratridin e 22 0.515 + 0.035
Li -veratridine-TT X 13 0.334 + 0.037
Choline! 16 0.239 + 0.025
Choline -veratridine 15 0.219 + 0.022

Abbreviation: TTX, tetrodotoxin

Student's t test: +
~pc:::::O.OOl, compared to Na+-group
p~O.OOl, compared to Li -group

2. Voltage-sensi tive Na+ channe~

Since neurons are characterized by the generation

and conductance of action potentials which are based,
at least partly, on a transient i¥crease in the per-
meability of voltage-depend e¥t Na channels, the possi-
bility has arisen whether Li can use this pathway in
crossing the neuronal membrane.
Biochemical studies using cell culture methods
an~ neurotoxins as toots have revealed that opening of
Na channels allows Li permeation into neuroblastoma
/Richelson, 1977; Schlesinger et al., 1979/ and neuro-
blastoma x glioma hybrid cells ./Reiser et al.~ 1982;
Reiser and Duhm, 1982/. Furthermore, by applying differ-
ent types o~ ionic media and veratridine and tetrodo-
toxin as Na channel activator and blocker, respective-
ly, it was fo~nd by ultrastructura l and morphometric
means that Li +enters cultured cerebral neuronal cells
through the Na channels /Table 1.; Janka and Jones,
1982 a, b/.

3. Na + -K + pump

The role of Na+-K+ pump in Li+ uptake by nerve

cells is strongly dependent on the external ionic mili-
474
eu. Increasing the Na+ or K+ content in the medium
significantly decreases ouabain-sensitive Li+ uptake
/Szentistvanyi et al., 1979 a; Janka et al., 1980 c;
Saneto et al., 1980/. In a detailed study it was shown
that half-maximal inhibition of ouabain-sensitive Li+
uptake by neuroblastoma x glioma hybrid cells was at
about 0.2 mM external K+ /Reiser and Duhm, 1982/. un-
der physiolo~ical ionic conditions, Li+ uptake mediated
by the Na+-K pump is negligible, although a marked
pump-mediated Li+ uptake was reported for clonal human
neuroblastoma cells /Saneto et al., 1980/. It seems
that extrusion of Li+ from cultured nerve cells is
apparently not influenced by the Na+ pump /Szentistvanyi
et al., 1979 b; Reiser and Duhm, 1982/.
4. Other pathways,
Data suggest that residual and yet not well-char-
acterized pathways for Li+ transport can be found in
the nerve cell membrane. Neuroblastoma x glioma hybrid
cells showed a saturation kinetics in ouabain-resistant
Li+ uptake which is inhibitable by external Na+ and K+
/Reiser and Duhm, 1982/. Moreover, a second ouabain-
resistant Li+ efflux mechanism is assumed /Reiser and
Duhm, 1982/, as well as leak/s/ and other theoretical
ion translocating systems.
STUDIES OF ERYTHROCYTE Li+ TRANSPORT IN AFFECTIVE
PSYCHOSES
Our p;evious results are in accordance with the
data of Li transport research showing a lower rate of
Na+-Li+ counterflow in red cells and a higher Li+ ra-
tio in the blood of a subpopulation of bipolar affec-
tive patients /Ostrow et al., 1978; Pandey et al., 1979;
Ramsey et al., 1979; Mendlewicz and Verbanck, 1981;
Dorus et al., 1979, 1983/. We studied in simultaneous
investigations the Li+ ratio and the rates of Na+-
dependent uphill Li+ efflux in 76 diseased controls,
52 bipolar and 32 unipolar patients in repeated assays
/at least 5 times in each patient/. Intraindividual
differences were less than 20 % in each case. We found
si~nificantly higher Li+ ratio and lower rate of Na+-
Li counterflow in the bipolar affective group /Szent-
istvanyi and Janka, 1979 a; Szentist·.ranyi et al., 1980/.
In other series of experiments, we measured the in vivo
Li+ ratio and in vitro Na+-dependent downhill Li+
transport in 28 psychiatric controls, 22 bipolar and
17 unipolar patients. Results gave a similar dis-
tinction between the nosologic subtypes of affective

475
 Table 2. Rates of red cell Na+-dependent Li+
efflux in bipolar subtypes of affec-
tive disorders /mean + SD/

Family n Na+-Li+ counterflow

history /mrnol/1 red cell/4 h/

Positive 20 0.119 + 0.039a

Negative 26 0.153 + 0.051
Total 46 0.136 + 0.048

Student's t test:
apc:0.02, positive vs negative comparison

disorders /Szentistvanyi and Janka, 1979 b/. Correlation

analysis has shown that those patients having a hi~her
Li+ ratio possess a lower rate of Na+-dependent Li
transport in their red cells /Szentistvanyi, 1978;
Szentistvanyi and Janka, 1979 a/.
Our collaborative studies performed during the
last 7 ys repeatedly show that almost half of the bi-
polar affective patients examined can be characterized
by positive family history /Rihrner et al., 1982, 1983/.
Table 2. presents that patients with a positive family
history have a significantly lower rate of Na+-depend-
ent uphill Li+ transport in their red cells.
Data on the clinical and theoretical significance
and on unresolved issues of Li+ transport research
accumulated during the last 10 ys be found in the
papers presented by the contributors to this topic.

Acknowledaement-

Research reported was supported by Grant 4-25-

0303-02-0/SZ of Ministry for Health, Hungary. Thanks
are due to the staff of National Institute for Nervous
and Mental Diseases, Budapest, Hungary, for collab-
oration and to Prof. D.G. Jones, Department of Anat-
omy and Human Biology, University of Western Australia,
Perth, Australia, for the fellowship and support.

476
REFERENCES
Canessa,M.,et a1.,1980, N.Eng1.J.Med.,302:772
Dorus,E.,et a1.,1979, Science,205:934
Dorus,E.,et a1.,1983, Arch.Gen.Psychiat.,40:545
Duhm,J.,et a1.,1976, Pf1Ugers Arch.,364:147
Duhm,J. and Becker,B.F.,1977, Pf1Ugers Arch.,367:211
Gorkin,R.A. and Riche1son,E.,1979, Brain Res.,171:365
Gorkin,R.A. and Riche1son,E.,1981, Neuropharm.,20:791
Grafe,P.,et a1.,1982, Pf1Ugers Arch.,393:297
Haas,M.,et a1.,1975, Nature/Lond./,258:425
Janka,Z.,et a1.,1979, Acta Neuropathol.,46:117
Janka,Z.,et a1.,1980a, Experientia,36:1071
Janka,z.,et a1.,1980b, Neuropharmaco1.,19:827
Janka,Z.,et a1.,1980c, Psychopharmaco1.,71:159
Janka,Z.,et a1.,1981, Acta Neuropatho1.Supp1.,VII:44
Janka,Z. and Jones,DG,1982a, Brain Res.,237:261
Janka,Z. and Jones,DG,1982b, Neuroscience,7:2849
Mend1ewicz,J. and Verbanck,PMP.,1981, Am.J.Psychiat.,
138:119
Ne1son,P.G.,1975, Physio1.Rev.,55:1
Ostrow,D.G.,et a1.,1978, Am.J.Psychiat.,135:1070
Ostrow,D.G.,et a1.,1980, Bio1.Psychiat.,15:723
Pandey,G.N.,et al.,1978, J.Gen.Physiol.,72:233
Pandey,G.N.,et a1.,1979, Arch.Gen.Psychiat.,36:902
Ramsey,T.A.,et a1.,1979 1 Arch.Gen.Psychiat. 1 36:457
Reiser,G. and Duhm,J.,1982, Brain Res. 1 252:247
Reiser 1 G.,et a1.,1982 1 J.Neurochem.,39:228
Riche1son,E.,1977 1 Science,196:495
Rihmer,Z.,et a1.,1982 1 Psychiat.Res.,6:197
Rihmer 1 Z. 1 et a1. 1 1983 1 Psychiat.C1in. 1 /in press/
Saneto 1 R.P.,et a1.,1980, J.Neurochem. 1 34:1520
Saneto 1 R.P. and Perez-Po1o 1 J.R. 1 1982 1 J.Neurosci.Res.,
7:413
Sch1esinger,H.R.,et a1. 1 1979, Life Sci.,25:957
Szentistvanyi 1 I.,1978, Acta Bio1.Acad.Sci.Hung. 1 29:101
Szentistvanyi,I. and Janka,Z.,1979a,Bio1.Psychiat.,
14:973
Szentistvanyi 1 I. and Janka 1 Z. 1 1979b, Psychiat.Res.,
1:265
Szentistvanyi,I.,et a1.,1979a, Neurosci.Lett.,13:157
Szentistvanyi,I.,et a1. 1 1979b, Ce11.Mol.Biol. 1 25:315
Szentistvanyi 1 I.,et a1.,1980, Psychiat.C1in.,l3:57
Thomas, R.C. ,et al. 1 1975 1 Nature/Lend./ 1 258:754
Tosteson,D.C.,1981, Fed.Proc.,40:1429

477
THE GENETICS OF LITHIUM METABOLISM IN

MAJOR AFFECTIVE DISORDERS

Julien Mendlewicz

Department of Psychiatry, Erasme Hospital

University Clinic of Brussels
Free University of Brussels, 1070 Brussels

Various kinetic processes are involved in the fate of a

drug from its absorption to its binding to plasma
proteins or to various tissues, its biotransformation
and interaction with receptor sites, and finally to its
excretion. These five major processes take place
simultaneously, and genetically transmitted variations
can occur through DNA mutations inducing structural
alterations in the specific proteins directly involved
in these five processes. In a given person, more than
one pharmacologically significant mutation can arise and
such mutations may not be so rare, as illustrated by
phenomena responsible for genetic polymorphisms such as
the ABO or the HLA, the pseudocholinesterase and the
INH-acetyltransferas e polymorphisms. The unimodal
gaussian curve indicates genetic homogeneity or polyge-
nic control, while the bimodal (mendelian autosomal
dominant trait) or polymodal curve (autosomal recessive)
are suggestive of a mechanism controlled by genes at a
single locus.
The genetic and environmental factors involved in
lithium metabolism and lithium response can be studied
by various methods. The twin method, comparing identi-
cal and fraternal twins, makes it possible to evaluate
to what extent variations in lithium metabolism are
genetically influenced. Differences in lithium rates
between one-egg and two-egg twins can provide an est-
imate of the "heritability", a theoretical concept
expressing the-genotypic variance as a proportion of
the phenotypic variance. Family studies may also permit
differentiation of environmental and genetic sources of

479
variation in lithium response by establishing regression
of offspring values to averaged parent values and corn-
paring spouse values.
Genetic factors have been associated with various
patterns of lithium prophylaxis. Bipolar patients with
a positive family history of bipolar illness seem to
have a better long-term lithium response than patients
without family history (Mendlewicz et al, 1973;
Stallone et al, 1973). These results have now been
confirmed by several groups, (Zvolsky et al, 1974;
Prien et al, 1974; Taylor and Abrams, 1975; Kupfer et al
1975; Mendlewicz et al, 1978; Ananth et al, 1979;
Cazullo and Srneraldi, in press), although one study
(Misra and Burns, 1977) reported nonresponders to
lithium prophylaxis to have more relatives with affec-
tive illness than lithium responders. This association
was not observed however for the presence of unipolar
illness in relatives of bipolar patients, (Mendlewicz
et al, 1973; Stallone et al, 1973). These genetic
results indicate that there may be several distinct
genetic subgroups of bipolar illness showing different
treatment responses to lithium.
The genetic heterogeneity of long-term lithium response
in bipolar illness is further evidenced by twin studies.
A higher concordance rate for affective illness is
found in both monozygotic and dizygotic twin pairs when
the proband had experienced good long-term response to
lithium (Mendlewicz, 1979; Mendlewicz et al, 1978).
In those twin pairs where the proband was considered as
a lithium nonresponder, the concordance rate for affecti-
ve illness was much lower. Furthermore, concordant twin
pairs, with better lithium prophylaxis in the index
case, had higher morbidity risks for affective illness
than discordant pairs. The above family and twin
studies indicate that genetic factors are to be consi-
dered as useful biological predictors of long-term
lithium prophylaxis. These genetic data also underline
the importance of the concept of genetic heterogeneity
of bipolar illness in relation to long term-lithium
prophylaxis.
The histocompatibility antigenic system has also been
studied in an attempt to find immunogenetic markers
of lithium prophylaxis, (Perris et al, 1978). The HLA-
A3 antigen was more frequently found in bipolar patients
who were long-term lithium failures, while long-term
lithium responders showed a reduction in the HLA-B18
antigen (Perris et al, 1978). This is certainly an
interesting observation relevant to the membrane hypo-
thesis of affective illness, since some of the antigens
in the HLA system may be implicated in membrane trans-

480
port mechanisms, but this study needs further replica-
tion. The membrane hypothesis of affective illness has
recently stimulated considerable research into the
physiological mechanisms of lithium transport across the
red blood cell (RBC) . Claims were made that the lithium
RBC/plasma ratio may serve as a pharmacokinetic indica-
tor of short- and long-term lithium response in the
affective illness, (Mendels and Frazer, 1973; Mendels et
al, 1976; Frazer et al, 1978). Although these claims
have been disproved, (Mendlewicz et al, 1978; Carroll,
1979; Mendlewicz, 1979), very important knowledge has
been gained in the physiology of the transport of the
lithium ion across the red cell membrane (Greil et al,
1979). It is also now clear that genetic factors do
play an important role in the membrane lithium transport
(Schless et al, 1975; Dorus et al, 1975; Mendlewicz et
al, 1978; Mendlewicz, 1979).
The lithium RBC/plasma ratios have also been shown to
have an heterogeneous distribution (two separate sub-
groups) in patients who are nonresponders to lithium
prophylaxis. Patients responding well to lithium
prophylaxis clearly showed an homogeneous pattern of
lithium RBC/plasma ratio distribution (Mendlewicz, 1978;
Mendlewicz, 1979). More recently, Dorus et al, 1979,
have reported higher in vitro Li RBC/plasma ratios in
some relatives (secondary cases) of bipolar patients,
suggesting that higher Li RBC/plasma ratios may be
biological indicators of a membrane vulnerability to
bipolar illness. Similar results are found by us
(higher Li RBC/plasma ratios) in monozygotic and dizygo-
tic twins concordant for affective illness when compared
to discondant twins.
One investigation (Bert et al, 1977) has reported an
increase in Stages 3-4 slow wave sleep in long-term
lithium responders, a promising electrophysiological
predictor of lithium response if confirmed in later
studies. After many years of the clinical use of
lithium in psychiatry, although its therapeutic efficacy
has now been demonstrated both for acute stages of
affective illness as well as for affective prophylaxis,
the question of long term-therapeutic outcome and
prediction still remains poorly defined. A multidisci-
plinary approach combining genetic, pharmacokinetic,
neurophysiological, neuroendocrine and neurochemical
parameters along with well-designed, long-term clinical
outcome studies will be needed to shed more light on
one of the key issues in lithium treatment, that is,
the need for specific biological and clinical predictors
of lithium preventive treatment in affective illness.

481
REFERENCES

Ananth,J., Engelsmann, F., Kiriarkos, R. and Kolivakis,T

1979, Prediction of lithium response. Acta Psych.
Scand. 60:279-286.
Bert,J., Saier,J., Dufour,H., Scotto,J.C., Julien,R. and
Sutter,J.M., 1977, Modification du sommeil provoquee
par le lithium en administration aigue et en admi-
nistration chronique. Electroenceph. and Clinical
Neurophys. 43:745-748.
Carroll,B.J., 1979, Prediction of treatment outcome with
lithium. Arch. Gen. Psychiatry.36:870-878.
Cazullo,C.L., and Smeraldi,E. HLA system in psychiatry
and psychopharmacology. Progress in Neuropsycho-
pharmacology. In press.
Dorus,E., Pandey,G.N. and Davis,J.M., 1975, Genetic
determinants of lithium ion distribution. An in
vivo monozygotic-dizygotic twin study. Arch. Gen.
Psych. 32:1097-1100.
Dorus,E., Pandey,G.N., Shaughnessy,R., Gaviria,M., Val,
E., Ericksen,S. and Davis,J.M., 1979, Lithium trans-
port across red cell membrane: a cell membrane
abnormality in manic-depressive illness. Science
205:932-934.
Frazer,A., Mendels,J., and Brunswick,D., 1978, Erythro-
cyte concentrations of the lithium ion: clinical
correlates and mechanisms of action. Am. J. Psychia-
try 135:1065-1069.
Greil,W., Becker,B.F., and Duhm,J., 1979, "On the rele-
vance of the red blood cell/plasma lithium rati6",
in Cooper,T.B., Gershon,S., Kline,N.S. (eds), Lithium
Controversies and Unresolved Issues. Amsterdam:
Excerpta Medica.
Kupfer,D.J., Dickar,D., Himmelhoch,J.M., and Detre,T.P.,
1975, Are there two types of unipolar depression?
Arch. Gen. Psychiatry 32:866-871.
Mendels,J., and Frazer,A., 1973, Intracellular lithium
concentration and clinical response: Towards a
membrane theory of depression. J. Psych. Res. 10:
9-18.
Mendels,J., Frazer,A., and Baron,J., 1976, Intra-
erythrocyte lithium ion concentration and long-term
maintenance treatment. Lancet, i:966.
Mendlewicz,J., Fieve,R.R., and Stallone,F., 1973, Rela-
tionship between the effectiveness of lithium
therapy and family history. Am. J. Psychiatry 130:
1011-1013 . .
Mendlewicz,J., Verbanck,P., Linkowski,P. and Wilmotte,J.
1978, Lithium accumulation in erythrocytes of manic-
depressive patients:An in vivo twin study. Brit. J.
Psychiatry 133:433-444.

482
Mendlewicz,J., and Youdim,M.B.H., 1978, Anti-depressant
potentiation of 5-hydroxytryptophan by L-Deprenil,
an MAO "type B" inhibitor. J. of Neural Transmission
43:279-286.
Mendlewicz,J., 1979, "Prediction of treatment outcome:
family and twin studies in lithium prophylaxis and
the question of lithium red blood cell/plasma ratios"
in Lithium:Controversies and Unresolved Issues,
Cooper,T.B., Gershon,S., Kline,N.S. (eds). Amsterdam:
Excerpta Medica.
Mistra,P.C., and Burns,B.H., 1977, Lithium non responders
in a lithium clinic. Acta Psychiat. Scand. 55:32-40.
Perris,C., Stransman,E., and Wahlby,L., 1978, HLA antigens
and the response to prophylactic lithium. 11th C.I.N.P.
Congress, Vienna:9-14 In press.
Prien,R.F., Caffey,E.H., and Klett,C.J., 1974, Factors
associated with treatment success in lithium carbonate
prophylaxis. Arch. Gen. Psychiatry 31:189-192.
Schless,A.P., Frazer,A., and Mendels,J., 1975, Genetic
determination of lithium ion metabolism:II. An in vivo
study of lithium ion distribution across erythrocyte
membranes. Arch. Gen. Psychiatry 32:337-390.
Stallone,F., Shelley,E., Mendlewicz,J., 1973, The use of
lithium in affective disorders:III. A double blind
study of prophylaxis in bipolar illness. Am. J.
Psychiat. 130:1006-1010.
Taylor,M.A., and Abrams,R., 1975, Acute mania. Arch. Gen.
Psychiatry 32:863-865.
Zvolsky,P., Vinarova,E., Dostal,T., 1974, Family history
of manic-depressive and endogenous depressive patients
and clinical effect of treatment with lithium. Act.
Nerv. Sup. (Praha) 16:194-195.

483
PHARMACODYNAMICS AND PHARMACOKINETICS OF NEUROLEPTICS AND

NEUROLEPTIC ASSAY METHODOLOGY

Robert 0. Friedel

Professor and Chairman

Department of Psychiatry
Medical College of Virginia
Virginia Commonwealth University
Richmond, Virginia

The purpose of this review is to examine the pharmacodynamics

and pharmacokinetics of neuroleptics and to compare the advantages
and disadvantages of commonly used assay methods for determining
neuroleptic plasma levels.

The nature and degree of responses to neuroleptics depend on

pharmacodynamic and pharmacokinetic variables. Pharmacodynamics
is defined as the study of the biochemical and physiological ef-
fects of drugs and their mechanisms of action. Neuroleptics are
thought to exert their clinical effect primarily through their
antagonistic effects on specific CNS dopaminergic receptors.
Pharmacokinetics deals with the absorption, distribution, metabo-
lism, and excretion of drugs. These factors determine the blood
and brain levels of neuroleptics and their metabolites, levels
which affect neuroleptic pharmacodynamics and which vary markedly
among patients on similar neuroleptic dosages.

There are at least three possible relationships between the

clinical effects of neuroleptics and drug plasma levels. First,
there may be no relationship between drug plasma levels and clini-
cal response. This is often the case in drugs, such as MAO inhi-
bitors, which bind irreversibly with their receptors. However,
neuroleptics bind reversibly with dopaminergic receptors. There-
fore, in those studies which have failed to show a relationship
between neuroleptic plasma levels and clinical response, it is
possible that either a number of patients were selected who were
neuroleptic non-responders, which precludes establishing such a
correlation, or plasma levels of important active metabolites of

485
the neuroleptic used were not measured. For example, it is known
that numerous metabolites of phenothiazines are produced, some of
them having significant pharmacologic activity, e.g., hydroxy and
demethylated metabolites. If these are present in high concentra-
tions, they also must be determined if valid correlations with
clinical response are to be established.

The second possible relationship between neuroleptic plasma

levels and clinical response is a sigmoidal one (Figure 1). At
low neuroleptic plasma levels, no clinical response is observed.
As drug plasma levels increase, one observes an increase in clini-
cal response which finally levels off as plasma levels increase.
At that point, there is no increase in efficacy as a result of
increases in dosage or in drug plasma level.

The third possible relationship is an inverted U pattern in

which an initial therapeutic response occurs as in the sigmoidal
relationship, but after peaking decreases as drug plasma levels
increase (Figure 2). The drug plasma level range between onset
and termination of drug action has been referred to as a "thera-
peutic window," and has been demonstrated repeatedly with the tri-
cyclic antidepressant nortriptyline. The therapeutic window term
has been misused to describe a range of blood levels below which
no therapeutic effect occurs and above which toxic side effects
override clinical response. This might be more accurately termed
a "therapeutic-toxic" window.

Clinical
Response

Plasma level
Fig. 1. A sigmoidal relationship between clinical response and
plasma level.

486
 A
---"

Overall
Clinical
Response ---------~
Therapeutic Window

Plasma Level
Fig. 2. A "therapeutic window" relationship between overall cli-
nical response and plasma level.

PHARMACOKINETICS

Because of the clinical relevance of the relationship between

neuroleptic plasma levels and clinical response, it is important
to und·erstand the factors which determine drug plasma levels, and
subsequently drug brain levels. These factors, referred to as
pharmacokinetic variables, are drug absorption, distribution,
metabolism, and excretion (Figure 3).

Most patients are given drugs orally which readily are

absorbed through the gastrointestinal tract because of their high
lipid solubility. A considerable amount of drug metabolism occurs
as the neuroleptic passes from the G.I. tract through the venous
portal system to the liver, a phenomenon referred to as the "first
pass effect." The drug and its metabolites then pass into the
systemic circulation where a significant amount become bound to a
variety of blood elements. Ten to 15% of the drug remains free
and it is this free fraction which distributes throughout the body
tissues. Because of high lipid solubility, a significant fraction
of the drug is distributed to adipose tissue and much less to
muscle tissue and other tissues of high water content. The drug
exerts its clinical effect in the central nervous system, either
in its parent form or in the case of some neuroleptics, also in
the form of their demethylated and hydroxy metabolites.

487
ABSORPTION DISTRIBUTION

.,. .,. ~
I FREE
I '
I
I
~~
BOUND
t
I
EXCRETION Other
Organs
:#~Vi:.
) '
.,..~If
METABOLISM

Fig. 3. Absorption, distribution, metabolism, and excretion of

neuroleptic agents.

The rate and extent of absorption may vary considerably

depending on the route of administration of the neuroleptic and
its chemical form upon administration. For example, high neuro-
leptic blood levels may be achieved rapidly by intramuscular or
intravenous injection, thereby bypassing gastrointestinal absorp-
tion and first pass liver metabolism. However, absorption of phe-
nothiazines from intramuscular depot sites may be significantly
delayed if the drug is injected in the form of enanthate or deca-
noate esters, which first must be hydrolized at the injection
site.

The most important pharmacokinetic variables responsible for

the 10-20 fold interpatient variations in neuroleptic plasma
levels are drug metabolism and excretion which determine the
drug's half-life in the individual patient. Because of their
lipid solubility, in order to be excreted, neuroleptics must be
converted first to water soluble metabolites. This is
accomplished mainly by demethylation and hydroxylation of the
parent compound. The hydroxy metabolite is then conjugated to
form glucoronide conjugate which is readily excreted by the kid-
neys. The average half-life of most neuroleptics is approximately
18-24 hours and is dependent on interpatient variations in metabo-
lic rate, which in turn is dependent on genetic factors, age, and
the ingestion of drugs which alter the rate of liver metabolism,
such as barbiturates and alcohol.

488
COMPARISON OF NEUROLEPTIC ASSAY METHODS

There are at least five different methods for determining

neuroleptic plasma levels: gas chromatography; high performance
liquid chromatography; gas chromatography /mass spectrometry;
radioimmunoassay; and radioligand dopamine receptor binding
displacement (Table 1).

Gas chromatography is one of the most commonly used assay

methods and can be performed at a reasonable cost. The specifi-
city is good, that is, one can be fairly assured that only the
neuroleptic in question is being measured. However, other
substances which co-chromatograph with the neuroleptic if present,
may give falsely high readings. Sensitivity is also quite good
with 1 ng/ml representing the lower limit of detection. A major
difficulty with this particular assay is that polar metabolites,
such as the hydroxy metabolites of phenothiazines, are not readily
analyzed.

High performance liquid chromatography has characteristics

quite similar to gas chromatography and is commonly used. Cost
and sensitivity are about the same as gas chromatography, but it
is somewhat less sensitive with lower levels of detection usually
in the range of 5 ng/ml. The advantage of this technique is that
polar substances are more readily analyzed than by gas chro-
matography.

The most accurate assay method currently in use is gas

chromatography/mass spectrometry. This is an infrequently used
technique because of the very high instrumentation costs and the
high level of skill required to utilize this instrumentation. The
sensitivity is approximately the same as gas chromatography.
Because of its very high specificity, this technique is the stan-
dard against which other assays may be validated; it is the
reference assay. In the better laboratories, this assay is also
used to perform quality control checks on other assay techni~ues.
As is the case with gas chromatography, gas chromatography fmass
spectrometry is not as useful for analyzing polar compounds as
high performance liquid chromatography.

Radioimmunoassays have been reported for a few neuroleptics.

They require a radioactive, highly specific antibody to the neuro-
leptic in question. Once this antibody has been obtained, the
cost of the assays is reasonably low. However, the specificity of
these assays varies considerably depending on the specificity of
the antibodies to the substance being measured. The development
of such antibodies is difficult and limits the usefulness of this
technique. The sensitivity of this assay is usually excellent,
often ten times that of the other assays already mentioned.

489
,J:I.
co
0

Table 1. Currently Available Tests For Making Plasma Level Determinations

Adequacy for
Polar
Test Use Cost Specificity Sensitivity Substances

Gas chromatography Common Reasonable Good Good Inadequate

("'1 ng/ml)

High performance Common Reasonable Good Good Effective

liquid chroma- ("' 5 ng/ml)
tography

Gas chromatography/ Infrequent Very high Excellent Good Effective

mass spectrometry (~ 1 ng/ml)

Radioimmunoassay * Infrequent Low Varies Excellent

(0.1-1 ng/ml)

Radioligand dopamine Increasing Low Low Excellent

receptor binding (0.1-1 ng/ml)
displacement

*Requires labeled and highly specific antibody

 The radioligand dopamine receptor binding displacement assay
measures the quantity of neuroleptic and metabolites which
displace a radiolabelled ligand from dopamine receptors obtained
from homogenates of caudate tissue. Kits are commercially
available for performing this technique, and the use of standard
curves enables the estimation of the dopamine receptor displace-
ment activity in unknown samples of neuroleptics. The advantages
of this particular technique are high sensitivity, low cost, and
the theoretical capacity to measure antidopaminergic activity of
parent drug and all active metabolites. The major difficulty with
this technique is that it has not been well-validated against
other neuroleptic assays.

In summary, there is no single assay for neuroleptics or

their metabolites which combines high specificity and sensitivity,
capacity to evaluate parent compound and polar metabolites,
reliability, and low cost. It is advisable before using any
laboratory's assay for neuroleptics, or other psychotropic drugs,
to request that they provide information regarding the assay uti-
lized, the validity and reliability of the assay and how they were
determined, quality control measures used, and cost. Values
obtained from laboratories that do not provide this information
should be suspect until proven otherwise.

491
PLASMA LEVELS AND THE MANAGEMENT OF CHRONIC,

REFRACTORY SCHIZOPHRENIA

Leo E. Hollister

Stanford University School of Medicine

Veterans Administration Medical Center
Palo Alto, California

Thiothixene has a rather long clinical use as an antipsychotic

drug.! I t ranks with haloperidol and fluphenazine as being a
highly potent, somewhat specific neuroleptic. Drugs of this type
can be used in much higher antipsychotic doses than low-potency,
less specific drugs. Many patients are regularly treated with
doses of thiothixene in excess of 100 mg/day. Few data have been
accumulated in any systematic way about the effects of doses in
the 100 to 400 mg/day range in patients refractory to lower doses
of this or equivalent drugs. Therefore, we elected to use
thiothixene to treat refractory patients, using such doses if
needed.

Attempts to measure plasma concentrations of thiothixene have

been fraught with difficulty. 2, 3 One of the problems has been
that the drug is present in relatively low concentrations, so that
often measurements are accomplished within 2 to 3 hours after a
dose is given rather than during the post-distributive state.
Another is that the drug has both cis (active) and trans
(inactive) isomers, which have been very difficult to separate
with many of the techniques used. However, the true phar-
macokinetic parameters of the drug of any possible range of thera-
peutic plasma concentrations has been difficult to ascertain. We
decided to see if such measurements of plasma concentration of the
drug could be made in the context of an intensive treatment
program. To do this, we used the Wellcome Neuroleptic Assay Kit.
However, rather than calibrating results to a standard curve using
haloperidol, we adapted it to a standard curve using thiothixene.
Thus, results could be reported in terms of ng/ml thiothixene
neuroleptic activity.4

493
 Table 1. Background Data on Patients Selected for Intensive
Treatment

Duration
Duration Present Previous
Subject Age Illness Episode Hospitalizations

J.V.-1 22 >2 yrs 3 mo 3

J.V.-2 23 >2 yrs 1 yr 5
G.M. 30 18 mo <2 wks 3
B.B. 21 >2 yrs >2 yrs 5
G.A. 29 >2 yrs >2 yrs 5
S.P. 21 >2 yrs 3 mo 6
H.C. 26 >2 yrs 1 mo 4
D.M. 26 >2 yrs <2 wks 5
R.C. 19 12 mo 12 mo 1
M.L. 27 >2 yrs 1 yr 4
R.S. 25 >2 yrs 3 mo 3
J.F. 20 12 mo 1 mo 2

Subjects S.P., M.L., and J.V. (on second evaluation) were

diagnosed with schizophrenia, undifferentiated; all other subjects
were diagnosed with paranoid schizophrenia.

Selection of Patients

We selected a special group of eleven treatment-resistant

young schizophrenic patients, those who represent the greatest
challenge to present-day management of schizophrenia. The diagno-
sis of schizophrenia followed DSM-III criteria. Treatment
resistance was defined as patients who had not shown more than
minimal improvement on previously used neuroleptics, up to at
least 30 mg/day of thiothixene or the equivalent dose of other
drugs. The demographic aspects of the 11 patients are summarized
in Table 1.

Treatment

Initial doses of thiothixene varied from 15 to 120 mg/day.

Doses were chosen based on the patient's past experience with drug
therapy. Thus, the degree of known resistance of the patient to
drugs provided the basis for selecting the initial dose. Doses
were raised by increments of 10-20 mg/day every week, assuming

494
that by the end of that time the patient would have reached
steady-state plasma concentrations. The goal was to find the
point at which there would be evidence that the psychosis showed a
"break", using the analogy of the management of refractory asth-
matic patients. Final doses ranged from 30 to 480 mg/day of
thiothixene.

Patients were considered to be unequivocally responsive if

they showed a shift of three grades or more on the Clinical Global
Impression (CGI) Scale towards a more favorable course as judged
by the estimate of their current severity of illness (e.g., from
"severely ill" to "mildly ill"), confirmed by judgement of defi-
nite improvement in the weekly assessments prior to that level of
change. Such improvement was required to be sustained for a
period of at least three to four weeks. Confirmation was obtained
by a Basic Psychiatric Rating Scale (BPRS) showing a major impro-
vement from the originial rating of the patients. An equivocal
response to treatment was judged to have occurred if the
progression towards improvement was only manifested by a shift of
two grades on the judgement of severity of illness, despite other
criteria having been met. A shift of only one grade was construed
as no response.

Results

The data on treatment of patients are summarized in Table 2.

One patient (J.F.), treated with substantial doses for a reason-
able period of time, did not show enough improvement to be con-
sidered as having responded. Four other patients showed an
equivocal response by the criteria we applied, even though impro-
vement on the BPRS was cons:'derable in three of them. One of
these patients (R.C.) responded so promptly to such a modest dose
of drug that at least during this episode he could not be con-
sidered as treatment-resistant. Six other patients showed an un-
equivocal response, in each case well verified by the change in
BPRS scores. Patients in this group received larger doses of drug
than the others and attained higher plasma concentrations (24 to
57 ng/ml, median 31, compared with 9 to 43 ng/ml, median 11).

One patient (J. V.) was treated twice, after having relapsed
out of hospital when his medication was sharply reduced. On the
first occasion, his breaking dose was 270 mg/day after 7 weeks of
treatment with a plasma concentration of 24 ng/ml. On the second
course· of treatment, his breaking dose was 360 mg/day after 10
weeks of treatment with a plasma concentration of 28 ng/ml. Thus,
it appears that within a single patient, dose requirements, time
of treatment, and plasma concentrations are reasonably consistent.

495
~
co
0)

Table 2. Details of Intensive Treatment

Treatment Plasma
Initial Final Duration Concentration Initial Final Initial Final
Subject Dose (m~) Dose (mg} at Break (wks) of Dru~ (n~/ml) BPRS BPRS CGI CGI
UNEQUIVOCAL RESPONSE

J.V.-1 30 360 7 24 63 9 6 3
J.V.-2 60 390 10 28 62 20 6 2
G.M. 60 240 5 31 52 14 5 2
B.B. 60 210 4 23 78 9 7 3
G.A. 60 210 6 57 52 20 5 2
S.P. 90 320 10 45 75 32 7 3
H.C. 120 480 6 54 62 19 7 3
EQUIVOCAL RESPONSE

D.M. 30 60 4 9 62 10 4 2
R.C. 15 30 2 9 72 15 4 2
M.L. 30 90 6 11 48 21 4 2
R.S. 90 280 26 43 -- -- 5 3
NO RESPONSE

J.F. 20 210 6 18 -- -- 4 3
 Despite these rather large doses, side effects were not quali-
tatively or quantitatively different from those observed at lower
doses. Dry mouth, blurred vision, constipation, drowsiness,
headache and weight gain were prominent. Extrapyramidal reactions
such as akathisia, rigidity, tremor or other signs of Parkinson
syndrome were no more prevalent than what might have been expected
at lower doses, but all patients were also receiving antiparkinson
drugs concurrently.

CONCLUSIONS

The present study was not primarily to establish definitively

a range of therapeutic plasma concentrations for thiothixene. Its
purpose was to treat patients in a "naturalistic" fashion, just as
they might be treated in ordinary clinical practice. With only
one exception, patients who responded less well had plasma con-
centrations less than 18 ng/ml while those who responded best had
plasma concentrations between 24 and 57 ng/ml. These con-
centrations at best apply only to treatment-resista nt patients and
would undoubtedly be much lower for a group of newly admitted,
previously responsive patients. More and more it is evident that
ranges of therapeutic plasma concentrations of drug must be
denoted by the types of patients being treated.

REFERENCES

1. J. E. Overall, L. E. Hollister, J. Shelton, I. Kimbell and V.

Pennington, Broad-spectrum screening of psychotherapeuti c
drugs: thiothixene as an antipsychotic and antidepressant.
Clin. Pharmacal. Ther. 10:205 (1969).
2. D. C. Hobbs, W. M. Welch, M. J. Short, W. A. Moody and C. D.
van der Velde, Pharmacokinetics of thiothixene in man.
Clin. Pharmacal. Ther. 16:473 (1974).
3. J. A. Yesavage, J. Becker, P. D. Werner, M. J. Mills, C. A.
Holman and R. Cohn, Serum level monitoring of thiothixene
in schizophrenia: acute single-dose levels at fixed doses.
Am. J. Psvchiatr. 139:174 (1982).
4. L. Lombrozo, L. E. Hollister, Radioreceptor assay of thio-
thixene using the Wellcome Neuroleptic Assay Kit.
Submitted for publication.

497
PLASMA LEVELS AS PREDICTORS OF CLINICAL RESPONSE AND VIOLENT

BEHAVIOR

Jerome A. Yesavage

Assistant Professor of Medicine

Stanford University Medical School
Stanford, California

THE PROBLEM: HOW CAN SERUM LEVELS BE MEASURED IN AN ACUTE UNIT TO

ASSESS CLINICAL RESPONSE?

Steady state serum levels have been found to be useful meas-

urements when selecting dosage for medications. This has worked
well with such drugs as digoxin, aminophylline, and phenytoin
sodium, but work with neuroleptics has been less successful. A
number of problems hamper research in this field. Certain medica-
tions show extremely low serum levels; problems occur with enzyme
induction and multiple metabolites; and the assays are technically
complex. Many studies have used flexible dose schedules which can
obscure therapeutic results when additional medication is added to
the regimens of nonresponsive patients or doses are reduced be-
cause of side effects. Finally, the most difficult obstacle to
overcome is obtaining steady state levels in short-term acute care
clinical settings where rapid patient turnover and brief hospital-
ization are the rule.

THE SOLUTION: SINGLE-TEST DOSE METHODS

A solution to the problem seems to lie in kinetic studies

which have shown a direct correlation between serum levels taken
within a few hours after an initial, single dose of neuroleptic
and eventual steady state. Test doses of lithium carbonate have
been measured 24 hours after the dose as a guide to dosage. For
example, Chang and associates have shown that serum levels of
butaperazine 6 hours after an acute single dose correlate with
steady state levels.l We obtained acute single-dose levels 2.5

499
hours after a single dose and obtained acute and steady-state
levels for 8 subjects after 10 days of thiothixene. The correla-
tion between the two levels was .86 (p<.01).

This is a substantial correlation. A correlation of 1.0 is

perfect; zero is no relation; better than • 5 is good for psy-
chiatry.

We recently attempted to overcome these problems by using a

fixed-dose schedule with single test doses of thiothixene. 2 If
acute single-dose levels of drug correlate with steady state
levels, we hoped to develop an acute single-dose test that would
predict clinical response. We chose thiothixene because of its
high potency, widespread use, and availability of a reliable
fluorometric assay technique. Some have argued that this assay
overestimates the amount of thiothixene because it measures both
active cis-and inactive trans-thiothixen e. However, steady state
levels of thiothixene have been found to correlate with rating-
scale measures of clinical response in schizophrenia.3,4
We undertook our study at the psychiatric intensive care unit
of the VA Medical Center in Palo Alto, California. Extremely agi-
tated patients are usually given high doses of neuroleptics im-
mediately to forestall assault and self-injury. We have found
oral thiothixene, 20 mg q.i.d., to be a safe regimen.

The subjects chosen were 30 patients admitted to the service

who had a DSM-III diagnosis of schizophrenia, were on oral thio-
thixene therapy of 20 mg q.i.d. for no longer than three days, and
had taken no concomitant medications other than anticholinergics
or chloral hydrate. The Brief Psychiatric Rating Scale (BPRS)
was used to measure outcome. Ratings were completed on admission,
after 1 week of treatment, and then at 1-week intervals if the
patient was hospitalized for longer than 7 days.

The patients received 20 mg q.i.d of thiothixene and an acute

single-dose level was drawn 2 hours later. Dosage was increased
if the subject still had substantial symptoms. Serum concentra-
tions were determined by a modification of the fluorescence method
of Mjorndal and Oreland. 5 Patients were diagnosed with schizo-
phrenia, by DSM III criteria. The acute thiothixene level after
the 20 mg test-dose was significantly correlated with positive
changes in the total BPRS score after the first week of treatment
(r=.51, p<.005) at 80 mg/day total (Figure 1). Serum level and
age were also correlated positively (r=.43; p(.05). When con-
trolled for age, we still had a significant correlation between
improved BPRS rating and serum level; however, when controlled for
serum level, no significant correlation developed between age and
clinical improvement.

500
 ....
38-
•
34-
• •• •
28- ••
BPRS 24-
• •

-
Cheng• 18-
•
•••• •• • • •
14-
•
8-. •••
•
4- .~:· •
-1 -~--
• ...
•
-·~--.,~~~·T-I~I--~I~-r-1-~1-~1~1---l~~
3 8 15 21 27 33 38 45 51 57 83
lerUIII Tlllotlllx.... LMel, ng/RII
(20 lftl TMt DoH)
Fig.l. Relationship of serum thiothixene level and clinical
improvement.

Some interesting individual variations were noted. One

patient, whose serum level at 80 mg was 22 ng/ml, did not show
further improvement at higher doses of thiothixene whereas two
other patients who had low serum levels with little clinical
improvement, showed significant improvement with higher serum
levels when their dose was raised to 120 mg/day. The patient who
did not respond later was found to have substantially enlarged
ventricles on CT-scan. Another patient responded well to 120
mg/day and had a good serum level. When his condition began to
deteriorate after about 3 weeks, the dose was raised to 160 mg/day
but the serum level decreased. It was subsequently discovered
that he was holding the drug in his mouth without swallowing and
would later discard it. We did not see any serious side effects
nor EKG changes at any dose during treatment.

We feel that our study indicated a correlation between fixed

acute single-dose levels of thiothixene and clinical response.
Therefore, the need to wait several weeks before obtaining a valid
steady state determination can be avoided.

501
 It remained unclear as to which of the several available
assays should be used to determine serum levels. We compared
fluorescence spectrophotometry (FS) with gas chromatography (GC)
in relation to clinical response.

FS and GC assays were performed on 14 split samples and the

results compared. We found the two detection methods highly
correlated, both when the total GC fraction and when the active
cis fraction are independently considered (Table 1). The FS, GC
total, and GC cis levels are all equally correlated with clinical
response using. changes in BPRS as the outcome measure. We also
found a highly significant correlation between initial serum
levels and RBC levels of thiothixene. The ratio of RBC to serum
thiothixene was very consistent, with the first RBC fraction
representing an average of 27 percent of the serum level of thio-
thixene. Both the serum and RBC levels of thiothixene were sig-
nificantly correlated with the change in BPRS score; however,
RBC levels seemed to be superior in lower serum level areas
( <10 ng/ml [FS]) where serum levels were less well correlated.
These results, however, were achieved on a small number of sub-
jects with post hoc analysis.

Table 2 shows our comparison between the fluorometric and GC

assays for the entire sample. The GC is strongly correlated with
the fluorometric.

Table 1. Correlation Between Assays and Clinical Improvement

Assays

FS FS-RBC GC Total GC-cis

BPRS improvement .SO** .43 .56* .53

over one week N = 60 N = 14 N = 12 N = 9

*p<.OS
**p<.OO!
FS = Fluorescence spectrophotometry
GC = Gas chromatography

502
 Table 2. Correlations Between Assays

Fluorometric GC

Serum RBC Total cis

Mean(+) 8.2 (5.2)* 1.8 (1.6) 6.9 (4.8) 6.2 (4.3)

R with Fluorometric .88** .72** .73**

*Values in ng/ml
**p<.02, N=lS

OTHER APPLICATIONS

Effects of Age

It is surprising that there is such a lack of information on

serum levels of neuroleptics in the elderly. We know that aging
is associated with many physiologic changes which can affect the
pharmacologic activity of medications. Some of these changes are
reduction of total body water, lean body mass, cardiac output, and
renal function, and changes in the activity of drug metabolizing
enzyme systems and number and sensitivity of receptor sites. The
elderly often take several medications concomitantly which compli-
cates the prediction of their response to a new medication.

The relationship between aging and some other psychotropics,

however, have been studied. Roose et al7 found that lithium car-
bonate seems to be effective at lower dosages in the elderly than
in young patients. Braithwaite et al8 found that serum levels of
tricyclic antidepressants are positively correlated with age.

Given the widespread use of neuroleptics in the elderly popu-

lation and their potential side effects for which the elderly are
particularly susceptible, we undertook a study to correlate thio-
thixene serum levels with age. We applied the same method as we
had previously, i.e., acute single-dose test to a sample popula-
tion of schizophrenic patients but utilizing a wider range age.

Pilot data indicated that young schizophrenic subjects who

received test doses of 10 mg thiothixene had low serum levels two
hours later; therefore, we selected the 20 mg dose. However,

503
elderly patients (defined in this study as those over 50 years)
became sedated at this dosage. We then ran two experiments: one
at the 20 mg dose and the other at 10 mg. We started with the 20
mg study, and, when positive age correlations were found, we
completed the second group, which included more elderly subjects
at 10 mg.

In the 20 mg study, we obtained 28 acute serum levels (Figure

2). The patients had a mean age of 30 with a range of 22 to 53
years, and serum levels ranged from 3 to 45 ng/ml. We found a
significant correlation between serum levels and age (r=0.43;
p<0.02). Twenty-five patients in the 10 mg study had a mean age
of 41 years with a range of 22 to 62 years with serum levels rang-
ing from 2 to 19 ng/ml. Again the correlation was significant
(r=0.41; p(0.05).

We did not observe any serious side effects. Three cases of

sedation were found, but these patients had relatively low serum
levels.

Although our results show a significant correlation between

serum levels of thiothixene and age, this may not hold for all
types of neuroleptics. In the future, we want to concentrate on
subjects above 65 years even though these patients may have
deteriorated more than younger patients, and are less likely to
provide informed consent. But we 1 11 get a better idea of the

10
18-
52-
• •
48-
44-
Age 40-
• ••
.• ..
ae- •
• • •
32-

28- •• •
•
~,.
24-
•
••,--,,--r--,-,--,--r,--,,--r--,-,--,-1
20 +-,--,-.,....
3 • 15 21 27 33 31 45 51 57 13
Ierum TNotlllxeM Lewel, ng/1111
(20 1111 THt DoH)

Fig. 2. Correlation of thiothixene serum levels and age.

504
serum level-drug effect relationship for neuroleptics in the
elderly.

Inoatient Violence

We undertook a study to correlate neuroleptic serum levels

and inpatient violence in schizophrenic patients. 9 There are,
however, a number of problems involved in assessing this relation-
ship. For one, measurement of violent behavior is difficult.
Many workers have dissected the dangerousness-violence-anger-
hostility-aggression constellation into components. In our
study, we used the global term "danger-related measures'' to refer
to dangerous behavior and cited "violence" as the harmful act
itself.

Researchers have also used a diversified array of criteria

for and measures of violence. We were mostly concerned with
multiple danger-related measures, as measured on an inpatient
ward. Such behavior can be accurately measured with the "Lion
Scale" named for John Lion of the University of Maryland, which
utilizes behavioral counts of physical and verbal assaults, and
episodes of seclusion and restrain.10

In prior research we showed that core schizophrenic symptoms

were directly correlated to inpatient danger-related events on an
acute closed unit.ll Thus with treatment of schizophrenia with
neuroleptics, the incidence of danger-related events should sub-
side. Our study tried to answer the question, are patients with
lower levels of neuroleptics, specifically thiothixene, more
violent?

We analyzed two main sets of variables, thiothixene serum

level and measure of ward behavior, in 58 male schizophrenic
patients over a 12-month period. We again used an acute single-
dose method: 20 mg q.i.d with serum levels drawn two hours after
the first morning dose. We recorded instances of danger-related
behavior on the ward for the day of admission or from the first
day of thiothixene treatment and for seven successive days.

Rated behavior included episodes of physical assault against

other patients or staff, verbal assaults against other patients or
staff, days in which seclusion and four-point restraint were
required to protect others, and a summary score for the total
number of any danger-related event. We found significant correla-
tion between serum levels and physical assault (-0.48), verbal
assault (-0.37), and total score of danger-related events (-0.65).
Episodes of seclusion and restraint, however, did not correlate
with serum level. Table 3 shows the regression analysis performed
for these data. The most important prediction of physical assaults
and inpatient violent acts was thiothixene plasma level.

505
 Table 3. Regression Analysis

Dangerousness
Measure Predictors

Physical assaults Drug abuse history • 26

Thiothixene serum level .57*
Inpatient violent acts Thiothixene serum level .43*
Combat in Vietnam .76

*R is negative

The mean serum level for the group was 13.3 ng/ml. Eight of
the nine patients who committed assaults had serum levels that
were below the mean, and lower serum levels were associated with
higher scores on three of four inpatient danger-related measures.

Thiothixene treatment was able to control violent behavior in

the great majority of the patients studied. Serum levels corre-
lated strongly and inversely with three of four danger-related
measures. Thus, the implication is that some patients with low
serum levels may become violent because their core schizophrenic
symptoms were not well controlled. Perhaps violence in the ward
may be committed by schizophrenic patients who are not receiving
enough medication to control their symptoms. Further study should
be aimed at delivering adequate serum levels to each patient.
This can also be quantified quickly by means of this simple acute
blood test.

Comprehensive evaluation of inpatient violence

Prior studies have shown that in addition to low neuroleptic

serum levels, other predictors of inpatient violence include the
patient's legal status, degree of severity of psychotic symptoms,
the act leading to admission, Vietnam combat experience, childhood
discipline, history of drug abuse and family conflict, and demo-
graphics such as age and race. After three years of data collec-
tion, we were able to compare the relative imfortance of these
variables by performing a multivariate analysis. 2
Subjects were 70 male patients (mean age 32 years) who were
interviewed by two board-certified psychiatrists and met DSM III
criteria for schizophrenia. We included data on dependent mea-
sures of dangerous acts committed on the inpatient unit, neuro-

506
leptic serum levels, and psychometric and questionnaire informa-
tion gathered during a comprehensive interview. We recorded
instances of assault-related behavior, used the acute single test
dose of 20 mg of thiothixene to record serum levels, performed
BPRS ratings, rated the violence of the act leading to admission
on a 9 point scale developed by Rabiner, asked patients to recall
if they had been disciplined in each of a variety of ways differ-
ing in severity (e.g., verbal discipline, spanking, beating,
discipline resulting in serious injury) by each of their parents,
and, as a preliminary to studying the relationship between Vietnam
experience and violent behavior, we examined the reliability of
patients' reports that they were in Vietnam and had seen combat
duty. Twenty schizophrenic patients with experience in Vietnam
were examined and their data verified by their medical records and
examination of his military unit while in was in Vietnam.

We determined the relative contribution of each variable in

explaining danger-related behavior by using a stepwise multiple
regression which included each dependent danger-related measure.
This enabled us to control such demographic variables as age,
race, and duration of illness.

We found that the best correlates of inpatient assaults were

lower serum levels of thiothixene, BPRS schizophrenia factor, and
violence prior to admission, all of which accounted for 49% of the
variance of this dependent variable. Of particular note were the
serum levels of thiothixene which correlates strongly with two of
three inpatient measures and with total inpatient acts. These in-
clude physical and verbal assaults, and other acts. The exception
was the decision to seclude and restraint. This may be due to the
fact that this decision involves a complex nursing decision which
we did not take into account in this study.

We feel such studies as these are potential predictors of

inpatient violence which can be obtained by simple blood tests
(without having to wait for steady state levels to be reached) and
psychiatric interviews. In light of the fact that assaults by
patients are probably underreported and more frequent than
generally acknowledged, early use of such measures may prevent
injury to staff and patients.

SUMMARY

These studies have been done to fill the gap in our knowledge
of serum levels of neuroleptics and how they correlate with clini-
cal effect. We used thiothixene as our neuroleptic of choice
because of its widespread use and relative safety. By means of a
fixed dose schedule with single test doses of thiothixene, we have
found that single-dose acute levels of drug correlate with steady

507
state levels and consequently clinical outcome. We have been able
to apply this technique to the elderly and found that a signifi-
cant correlation can be shown between serum levels of thiothixene
and age. Finally, we found that serum levels of thiothixene are
our best single indicator of possible violence in the inpatient
schizophrenic population.

REFERENCES

1. S. S. Chang, H. Dekirmenjian, A. Walcynski, et al., Clinical

pharmacological studies of antipsychotics. Read before a
meeting of the Society of Biological Psychiatry, Atlanta,
May, 1978.
2. J. A. Yesavage, J. Becker, P.O. Werner, et al., Serum level
monitoring of thiothixene in schizophrenia: acute single-
dose levels at fixed doses, Am. J. Psychiatr. 139:2 (1982).
3. L. Jacobsson, L. Von Knorring, B. Mattsson, et al., Plenfluri-
dol and thiothixene: dosage, plasma level and changes in
psychopathology. Int. Pharmacopsychiatry 11:206 (1976).
4. L. Jacobsson, L. Von Knorring, B. Mattsson, et al., Controlled
trial of penfluridol and thiothixene in the maintenance
treatment of chronic schizophrenic syndromes, Acta.
Psvchiatr. Scand. 54:113 (1976).
5. T. Mjorndal and L. Oreland, Determination of thiothixenes in
plasma at therapeutic concentrations, Acta. Pharmacal.
Toxicol. (Copenh) 29:295 (1971).
6. J. A. Yesavage, C. A. Holman, R. Cohn and L. Lombrozo, Cor-
relation of initial thiothixene serum levels and clinical
response. Comparison of fluorometric, gas chromatographic,
and RBC assays, Arch. Gen. Psychiatr. 40:301 (1983).
7. S. P. Roose, S. Bone, C. Haidorfer, D. L. Dunne and R. R.
Fieve, Lithium treatment in older patients, Am. J.
Psychiatr. 136:843 (1Q70)
8. R. Braithwaite, S. Montgomery, and S. Dawling, Depression and
tricyclic antidepressant levels, in "Drugs and the "El-
derly," J. Crooks and I.H. Stevens, eds., University Park
Press, Baltimore (1979).
9. J. A. Yesavage, Inpatient violence and the schizophrenic pa-
tient: an inverse correlation between danger-related events
and neuroleptic levels, Biol. Psychiatr. 17 (no. 11):1331
(1982).
10. D. Student and J. R. Lion, Methodological Issues in Psycho-
pharmacological Research in Violent Individuals, Inter-
national Society for Research on Aggression, National
Institutes of Medicine.
11. J. A. Yesavage, P. D. Werner, J. M. T. Becker, C. A. Holman
and M. J. Mills, Inpatient evaluation of aggression in psy-
chiatric patients, J. Nervous Ment. Dis. 169:299 (1981).
12. J. A. Yesavage, Correlate£: of dangerous behavior by schizo-
phrenics in hospital (unpublished data, 1983).

508
GUIDELINES TO USE OF PLASMA LEVELS

FROM A CLINICAL PERSPECTIVE

Theodore Van Putten

Brentwood Veterans Medical Center

Los Angeles, California 90073

If plasma levels of antipsychotic drug could be related in

some way to clinical outcome, the pharmacotherapy of schizophrenia
might become less of a frontier profession. So far, all that ap-
pears to be firmly established is that plasma levels of antipsycho-
tic drugs vary enormously between patients.

Much of the past research on the relation between plasma levels

of antipsychotic drug and clinical change has been difficult to in-
terpret because of shortcomings with the assays or design deficien-
cies (lack of 1 f~xed-dose design, contamination with other treat-
ments, etc.). '

Studies that strive to relate plasma level to outcome should be

of fixed-dose design. If the clinician is allowed to vary the dosage
and if he is good at it, all plasma levels, assuming a strain of
schizophrenia sensitive to drug, will be in the therapeutic range and
there will be no relationship. If the sample contains schizophrenic
patients whose illness is not sensitive to drug, non-responders are
likely to develop high plasma levels as the clinician pushes the
dose. Fixed-dose studies, however, are demanding, both for the pa-
tient and the staff.

Recent 4~~ed (or consta9t) dose studies witR fluphenazine,3

haloperidol, thiothixene, and chlorpromazine do suggest a
relationship between plasma level and clinical change. If such a
relationship is threshold or linear, however, it really would have
little impact on prescribing practices; particularly with a linear
relationship (which is what clinicians unwittingly assume) one pushes
the dose until the patient, hopefully, improves. A curvilinear or

509
"therapeutic window" relationship is clinically important because it
would imply that the dose need be decreased in some non-responders.

The following studies performed by our laboratories address the

clinical usefulness of a neuroleptic plasma level:

Chl()rpromazine (CPZ)

Forty-eight newly readmitted schizophrenic patients were treat-

ed with a fixed conservative dose of CPZ (6.6 mg/kg or less). CPZ
levels ~ere measured by a gas chromatography mass spectrometry method
(GCMS).

Steady state plasma ~r saliva levels at 4 weeks did not appear

to be related to outcome. Rather, non-response appears more a
matter of illness sensitivity to CPZ and speed of response. Some
non-responders at 4 weeks did eventually improve on qu~te conserva-
tive plasma levels given another 4 weeks of treatment.

An important question is whether higher plasma levels are asso-

ciated with worsening behavior. After 4 weeks of treatment with a
fixed dose, dosage was increased (range 900-1800 mg) in the non-re-
sponders. Some became distinctly worse. Those who became worse had
higher plasma levels (>300 picomoles (or 100 ng) of CPZ) suggesting
behavioral toxicity.

Common findings in patients whose plasma levels eventually ex-

ceeded 300 picomoles were that each increase in dosage did not pro-
duce any noticeable change: that they remained severely ill: that
their illness, at least in retrospect, was not sensitive to CPZ: that
they were inaccessible. Each always replied that the medication was
beneficial, and either asked for or willingly took more. It is in
the inaccessible patient whose illness is only minimally, or not at
all, se~sitive to CPZ that a plasma level might be especially
useful.

If our findings with CPZ can be replicated and if they apply to

other antipsychotic drugs, psychotoxicity is probably much more com-
mon than currently realized.

Haloperidol (HPL)

Forty-four newly admitted schizophrenic patients were randomly

assigned to treatment with a fixed dose of HPL: either 5, 10 or 20
mg for 4 weeks at bedtime. After 4 weeks of treatment with fixed
dose, the dose was raised, side effects permitting, in the 10elative
non-responders. HPL was assayed by the radioimmuno assay.

The relationship between dosage and HPL level is a close one

(r=.95 over a dosage range of 5 to 40 mg). Such a close relationship

510
could be construed as a relative advantage for HPL, although inter-
patient variability at the 5 and 10 mg doses is such as to make a
plasma level clinically useful (other than just to rule out non-
compliance).

Plasma levels of responders and relative non-responders after

4 weeks of treatment showed complete overlap. Responders improved
over a wide range of plasma levels ranging from 2-30 ng/ml. Three
of the 5 who responded at <5 ng/ml had a type of schizophrenia that,
at least in the past, responded only to antipsychotic drug.

As plasma levels were raised in the relative non-responders,

some became distinctly worse. One man with a plasma level of 90
ng/ml developed a continuous "nervous cough"; another with a plasma
level of 61 ng/ml became more excited, angry and started to scream
at his hallucinations; another at 24 ng/ml developed "terror fits".
These exacerbations appeared drug-related in that they disappeared
as the dose was lowered.

Other non-responders, at higher plasma levels (10-30 ng), had

complex outcomes. Some were improved in the sense of chemical re-
straint. Others became less psychotic, but this gain seemed nulli-
fied by depression and despair. Higher plasma levels (10-30 ng/ml)
in relative non-responders pose complex risk-benefit issues and re-
quire more investigation.

Thiothixene

Thirty-four newly admitted schizophrenic patients were treated

with a fixed dose of thiothixene for 4 weeks at bedtime (0.44mg/kg
or less). Thiothixene and its active 'ltabolites were measured in
plasma by the radioreceptor technique. Dopamine receptor block-
ing activity of drug in the plasma samples was expressed as neuro-
leptic units (n.u.) (One n.u. of dopamine receptor blocking activity
is defined as activity produced by 100 ul of a 1 nM solution of halo-
peridol).

The relation between dose and plasma level is not a close one.
Above a dose of 25 mg, there was considerable variation in plasma
level so that the correlation (r) between plasma level and dose was
0.385.
The Figure shows the relationship between the BPRS 12 ratings
and plasma level. Improvement occurred over the entire range of re-
corded plasma levels, but the chance of substantial improvement ap-
peared greater above 40 n.u.·r The Global and NOSIE 1j ratings
show the same trend. Correlations between plasma levels and outcome
(both at 28 days) appear in Table 1. There was no relationship be-
tween dosage (either mg or mg/kg) and any of the four outcome mea-
sures.

511
 40

.. • •
E-<
z 30
101

•
::1!!
101
>
0
• • • •
••
11:1101

..•• ••••
P..O
::li!Z 20
""'<
= •
0
Cl.)
11:1
p.. 10 • • •
CQ

•• • •
0
•
·10
•
<15 1520 30 40 50 60 70 80 90 100 ~150
PLASMA THIOTHIXENE IN NEUROLEPTIC UNITS

Fig. 1. Change in BPRS vs. plasma level after 28 days of treatment.

There was no consistent relationship between side effects

(extrapyramidal symptoms or drowsiness) and plasma level.
The findings indicate that improvement after 28 days of treatment
is more likely at higher plasma levels (>40 n.u.) of thiothixene.
Many, however, improved at very low plasma levels. Indeed, two of
the six, who improved at plasma levels <15 n.u., had an illness
which.in the past, responded only to an antipsychotic drug.
The data do not support the notion of a "therapeutic window" in
that higher plasma levels were not associated with side effects or
clinical deterioration (although at extreme plasma levels this must
of course be so). In fact, an additional four patients who improved
markedly only on very high doses (60,120,120,and 160 mg of thiothi-
xine) had high plasma levels (65,78,170,and 140 n.u., respectively).

Table 1. Correlations (r) between 28-day plasma

levels of thiothixine and outcome

CGI CGI
Nurse Doctor NOS IE BRPS
r 0.260 0.358 0.363 0.289
E n.s. 0.025 0.025 n.s.

512
Again, higher plasma levels appear associated with improvement as
long as the patient can tolerate them!

Many relative non-responders, however, could not tolerate higher

doses. Akathisia that could not be suppressed by high doses of anti-
parkinson drug (trihexyphenidyl 12 mg daily or benztropine 8 mg
daily) and drowsiness were the most frequent dose-limiting side
effects. Thus, in 11 of 20 non-responders, dosage could not be in-
creased because of akathisia, drowsiness or akinesia, either singly
or in combinations.

A plasma level may be useful in the non-responder who cannot

tolerate a higher dose because of side effect. Time, however, is an
important issue in non-response; some "non-responders" with a low 8
plasma do eventually improve, turning out to be "slow responders".
When non-responders with dose-limiting side effects and a low plasma
level (<40 n.u.) were given thiothixene in the same or smaller dose
for another 28 days they did not improve. Thus, if a non-responder
with troublesome side effects has a low plasma level (<40 n.u.), it
would seem prudent to try another antipsychotic drug. If a non-re-
sponder with troublesome side effects has a high plasma level, dose
could be reduced (or kept the same) in the hope that improvement will
come with more time. In the non-responder without side effects, the
dose can be pushed as, at least to date, there is no indication of
psychotoxicity at higher plasma levels.

Conclusions

1. With CPZ, THX AND HPL there is a group of schizophrenics usually

referred to as "responders", who improve over a wide range of
plasma levels.

2. With CPZ, plasma levels >300 picomoles (or >100 ng/ml) appear to
be psychotoxic, but this requires replication by others.

3. With HPL three studies with different designs arrive at very

similar therapeutic windows (Smith et al,1982: 5-14 ng/ml; Extein
et al,1982: 5-15 ng/ml; Magliozzi et al,1981: 8-18 ng/ml). Our
findings are similar. The role of higher plasma levels (say
15-50 ng/ml) in relative non-responders is a complex and unset-
tled issue.

4. A fixed dose study with fluphenazine 3 found a therapeutic win-

dow relationship with presumed psychotoxicity above 2.4 ng/ml.
(However, there were only 3 patients >2.4 ng/ml.)

5. With THX higher plasma levels are associated with a greater

chance of improvement, side effects permitting. So far, 1 ~hl5e is
no indication of psychotoxicity at higher plasma levels. '

513
References

1. P. R. A. May and T. Van Putten, Plasma and saliva levels of

chlorpromazine in schizophrenia: a cri~ical review of the
literature. Arch. Gen. Psychiatr. 35:477, (1978).
2. J. M. Davis, s. Erickson and H. Dekirmenjian, Plasma levels of
antipsychotic drugs and clinical response, in "Psycho-
pharmacology: A Generation of Progress," M. A. Lipton, A.
DiMascio and K. F. Killam, eds., Raven, New York (1978).
3. M. W. Dysken, J. I. Javaid and S. S. Chang, Fluphenazine
pharmacokinetics and therapeutic response, Psychopharma-
cology 73:205 (1981).
4. J. R. Magliozzi, L. E. Hollister, 'K. V. Arnold and G. M.
Earle, Relationship of serum haloperidol levels to clinical
response in schizophrenic patients, Am. J. Psychiatr.
138:365 (1981).
5. I. Extein, K. A. Augusthy and M. S. Gold, Plasma haloperidol
levels and clinical response in acute schizophrenia,
~sychopharm. Bull. 18:156 (1982).
6. R. G, Smith, What is the clinical value of a neuroleptic blood
level? Presented at the American College of Neuropsycho-
pharmacology, Dec. 17, 1981.
7. T. Van Putten, P. R. A. May, S. R. Marder, J. N. Wilkens and
B. J. Rosenberg, Plasma levels of thiothixene by radio-
receptor assays: clinical usefulness, Psychopharmacology
79:40 (1983).
8. T. Van Putten, P. R. A. May and D. J. Jenden, Does a plasma
level of chlorpromazine help? Psycho!. Med. 11:729 (1981).
9. P. R. A. May, T. Van Putten, D. J. Jenden, C. Yale, and W. J.
Dixon, Chlorpromazine in blood and saliva levels and the
outcome of treatment in schizophrenic patients, Arch. Gen.
Psychiatr. 38,202 (1981).
10. R. T. Ruben, A. Forsman and J. Heykantz, Serum haloperidol
determinations in psychiatric patients: comparison of
methods and correlation with serum prolactin, Arch. Gen.
Psychiatr. 37:1069 (1980).
11. I. Creese, D. R. Burt and s. H. Snyder, A simple and sensitive
radioreceptor assay for antischizophrenic drugs in blood,
Nature 270:180 (1977).
12. J. E. Overall and D. R. Gorham, The brief psychiatric rating
scale, Psycho!. Rep. 10:799 (1962).
13. G. Honigfeld and C. J. Klett, The nurses 1 observation scale
for inpatient evaluation: a new scale for measuring im-
improvement in chronic schizophrenia J. Clin. Psycho!.
21:65 (1965)0

14. J. A. Yesavage, J. Becker and P. D. Werner, Serum level moni-

toring of thiothixine in schizophrenia: acute single-dose
levels at fixed doses, Am. J. Psychiatr. 139:174 (1982).
15. L. Hollister, Plasma levels and the management of chronic
refractory schizophrenics. Presented at VII World Congress
of Psychiatry, Vienna, 12 July, 1983.

514
OPTIMUM DOSAGE OF NEUROLEPTIC AGENTS BY ELECTRONIC

RECORDING OF EXTRAPYRAMIDAL FINE MOTORICITY IN HANDWRITING

(NTI HAASE)
Hans-J. Haase

Arztlicher Direktor
Pfalzklinik Landeck
Weinstr. 100, 6749 Klingenmunster F.R.G.

In the beginning of the fifties, working as an intern at

the University Psychiatric Clinic at Bonn, I was prompted
by a research assignment by the Psychological Institute
of the University of Bonn to investigate the relation
between drive and motoricity in patients suffering from
Parkinson's disease, even before neuroleptic agents had
been introduced. This was a peculiar coincidence,
committing me for a lifetime, as I was soon to discover.
Shortly thereafter, i.e. in 1953, I was given leave to
join the Psychiatric Hospital at Oberwill/Zug
(Switzerland), where numerous psychotic patients were then
treated with the ·first neuroleptic agent available at the
time, namely chlorpromazine (Megaphen, Largactil). When
it became evident, that a patient treated with
chlorpromazine shows a parkinsonian restriction of the
movement of his arms when walking, my respective letter
addressed to the scientific department of Messrs. BAYER/
Leverkusen was answered to the effect, that there was no
indication of a connection between this parkinsonism and
the effect of chlorpromazine. At that time, however,
evidence existed already in medical literature, that the
rauwolfia alkaloid reserpine, known in Indian folk
medicine, with a similar therapeutic effect on delusional
diseaies repeatedly evokes parkinsonism. Allowing for the
discrepancy between extrapyramidal coarse motoricity in
walking and fine motoricity in handwriting, I had dis-
covered during the a.m. investigation in Bonn, I embarked

515
on a systematic study of the fine motoricity of hand-
writing, first of patients treated with chlorpromazine,
and then of those receiving reserpine, as from the autumn
of 1953. I found out, that parkinsonian fine-motor
symptoms of inhibition in handwriting, but not the
subjectively unpleasant coarse-motor extrapyramidal
symptoms, give a basic indication of the required dosage
of drugs for the treatment of psychotic experience
productions, i.e. delusional diseases, drugs that were
soon referred to as neuroleptic agents by Delay and
Deniker.

A few years thereafter members of the research team of

Janssen (Belgium) were the first ones to recognize, that
neuroleptic agents basically inhibit animal motoricity,
and that the correlation between the data on neuroleptic
dosages derived from animal experiments and the result
reached by my team with fine-motor tests is closer than
is the case with any other method of investigation.

For didactic reasons I subdivided the target symptoms of

neuroleptic therapy into positive and negative symptoms.
Positive symptoms mean psychotic experience productions,
such as delusional ideas, hallucinations etc. as well as
psychotically induced states of severe excitation, inde-
pendent of whether they break out in the form of psycho-
motor agitation or whether they dam up inside (e.g. as
catatonic stupor).

By negative symptoms we understand the psychotically

induced diminution of psychic energy, which has to be met
by sociotherapeutic and psychotherapeutic measures.

In consideration of clinical experiments with about 1000

schizophrenic patients conducted by my team since 1953
with the aid of about 50.000 fine-motor evaluations of
handwriting specimens, all of the same text, and con-
sidering also the psychopathological findings, such as
adverse reactions and accompanying effects of 30
different neuroleptics, four terms have been worked out,
that that are of practical importance for the application
and dosage of neuroleptic agents:

1. the neuroleptic threshold

2. the neuroleptic potency, in relation to the neuro-
leptic threshold
3. the neuroleptic-therapeutical range
4. the disposition for the neuroleptic effect/threshold
i.e. the neuroleptic disposition

516
1. The neuroleptic threshold

As a neuroleptic threshold dosage I describe the dosage

of a neuroleptic agent, that causes an extrapyramidal
inhibition discerneble in fine-motor movements. It is
considered a minimum dosage, if it is a question of
obtaining the neuroleptic effect as such, above and beyond
a tranquilizing effect. The neuroleptic threshold dosage
is to be considered a minimum dosage in as far as the
psychotic positive symptoms, as mentioned above, are re-
duced and/or compensated during full vigilance. Experience
has shown, that especially with the more potent
neuroleptic agents, as far as they possess a cumulative
tendency, the neuroleptic threshold dosage drops, so that
in the course of prolonged treatment dosage reductions are
quite frequently required in order to avoid the appearance
of coarse-motor extrapyramidal symptoms of an overdosage,
particularly of dyskinetic reactions, acathisia and coarse-
motor parkinsonism.

2. Neuroleptic Potency
In consideration of the neuroleptic threshold dosages I
reduced the neuroleptic agents to a system on the basis of
chlorpromazine = 1, and designated those neuroleptic agents
as weakly potent, whose neuroleptic threshold dosage
exceeded that of chlorpromazine, and considered those as
medium highly potent, whose neuroleptic threshold dosages
were increasingly lower than that of chlorpromazine.

The weakly potent neuroleptics are used predominantly for

sedation, tranquilization, and anxiolysis as well as the
improvement of the tendency to fall asleep; they are
applied in particular, when internal uneasiness prevails,
i.e. at the beginning of hospital treatment or in patients
with a continuous disposition for restlessness, anxiety,
and also aggressiveness. An exception to the rule is
sulpiride (Dogmatil), which has no sedating effect below
the neuroleptic threshold dosage, and no marked stimula-
ting and antdrepressive action, while with an average
dosage of about 300 to 600 mg the neuroleptic threshold
is crossed and the drug may then be used as a neuroleptic
agent.

Important: An increasing neuroleptic potency does not mean,

that a drug compensates psychotic positive symptoms more
rapidly and frequently, it rather signifies, that the
neuroleptic threshold is reached with increasingly lower
dosages. There is one restriction, however, regarding
weakly potent neuroleptics: in patients with a low

517
neuroleptic disposition it may be impossible to cross the
neuroleptic threshold with a sufficient margin because of
the excessively high dosage, that would be required then,
and the respective anticholinergic side-effects.
Electronic recording of the extrapyramidal fine-motor
symptoms in handwriting.
In 1980 we began to introduce the electronic recording of
extrapyramidal fine-motor signs in handwriting in our
hospital (Pfalzklinik Landeck). For about two years
approximately one-hundred patients per day receiving
neuroleptic treatment in our hospital are subjected to my
handwriting test. The electronically obtained data are
recorded daily on the patient's chart and help physicians
to avoid neuroleptic over- or underdosages. In comparison
to the reference handwriting specimens taken before the
treatment a constriction by about - 13 % and more indi-
cates that the neuroleptic threshold has been reached, if
this trend is maintained during the next two or three days.
Summary of results
a) Indication for dosages in the neuroleptic threshold
range (NT)
1. At least 60 % of all acutely psychotic patients with
positive symptoms
2. Virtually 100 % of all patients requiring mainten-
ance treatment in the meaning of long-term therapy
b) Advantages of the NT dosage
1. Full maintenance of the psychomotor abilities.
Patients can drive cars and engage in sports, such
as tennis
2. Essential antipsychotic effect on positive symptoms
within about one week after exceeding the neuroleptic
threshold dosage in about two thirds of the patients
3. No pharmacogenic depression as a sequela of
subjectively unpleasant side-effects such as
tiredness or coarse-motor extrapyramidal symptoms
4. Appearance of coarse-motor extrapyramidal symptoms
only in about 10 % of the patients (with medium
dosages i.e. 2 to 10 times the neuroleptic threshold
dosage coarse-motor extrapyramidal symptoms in about
60 to 70 %. With high dosages, i.e. 10 to 30 times
the neuroleptic threshold dosage increasing fre-
quency of coarse-motor symptoms at the beginning of
treatment, but more pronounced sedation and
increasing appearance of coarse-motor symptoms
during prolonged treatment.

518
 5. During long-term treatment only light tardive
hyperRineslas in exceptional cases (about 3 %)
6. Avoidance of cumulative overdosages, especially
during treatment with important long-term
neuroleptics.
7. Avoidance of underdosages.

c) NTI = Neuroleptic threshold indicator

A handwriting diverging from the reference specimen in

three writing before the beginning of treatment by up to
minus 12 % is not neuroleptic as a rule. In exceptional
cases the neuroleptic threshold is already crossed with
a lesser degree of constriction.

Light neuroleptic handwriting symptoms = minus 13 % to

minus 20 %. With low values, e.g. minus 13 % or minus
14 % or even 15 % the constriction in handwriting is to
be considered neuroleptic only, if this trend is con-
firmed during the next £ew days. If the third verse shows
a distinct constriction as compared to the fi~st verse
(quotient as from minus 1%), this will indicate moderate
extrapyramidal fine-motor symptoms in the majority of
cases.

With a light neuroleptic constriction of handwriting there

are mostly no coarse-motor symptoms. With the exception
of light acathisia and maybe dyskinetic reactions mostly
in the neck muscle area in up to 10 %of the patients.

Up to minus 30 % moderate neuroleptic symptoms = in up

to one third of the cases appearance of coarse-motor
symptoms of the extrapyramdial type, that may be
subjectively unpleasant.

As from minus 30 % severe neuroleptic symptoms = with

increasing frequency subjectively disturbing coarse-
motor symptoms, especially pronounded akinesia, rigor,
acathisia. But even with more than minus 40 % frequently
no clinical discernible coarse-motor extrapyramidal
symptoms.

What is the clinician's task when establishing the

neuroleptic dosage by means of diagnosing extrapyramidal
fine-motor symptoms?

The clinician having the intention of using neuroleptics

with the objective of "as little as possible - as much
as necessary" can find out by individual incipient
dosage with fine-motor control, if and when the dopamine

519
receptors are blocked, which is considered a pre-
requisite for an incisive antipsychotic effect on
positive symptoms. As soon as the patient states
during examination, that he feels a motor constriction,
the dosage is frequently approaching overdosage, or
overdosage has already been reached.

Important:
The desired dosage with only light moderate extrapyra-
midal fine-motor inhibition is frequently not experienced
subjectively by the patient, at least not as psychomotor
inhibition. Therefore even the most subtle exploration is
no substitute for diagnosing the extrapyramidal fine-motor
symptoms in handwriting, just as little as a thorough
examination of the muscular tone or the patient's gait.

As soon as the neuroleptic threshold has been reached, at

least one week may pass without further dosage increases
in patients exhibiting no tendency of endangering them-
selves or others, because experience has shown, that the
essential antipsychotic effect on positive symptoms
appears within one week in two thirds of the cases.

The clinician establishing the dosage of neuroleptics,

of course, neither shall nor can be replaced by a
computerized method. On the other hand it is not possible
in view of the up to 16 fold interindividual differences
in disposition for the neuroleptic threshold to find
the individual dosage on a milligram basis, of no
indicator is used, showing that the neuroleptic threshold
has been reached.

520
COMPUTERIZED EVALUATION OF FINE-HOTOR SYMPTOMS

A. Dreher

Pfalzklinik Landeck
6749 Klingenmlinster
Federal Republic of Ge~many

The computerized evaluation of the handwriting test

according to Haase provides, for the first time, an
objective method of an individual dosage of neuroleptics
based on the severity of fine-motor inhibition. The
extrapyramidal-motor side-effects of neuroleptics are
caused by blocking the post-synaptic dopamine receptors
in the corpus striatum. At the beginning of treatment
dyskinetic reactions may occur, e.g. a neckface syndrome,
and later on the characteristic Parkinson syndrome with
akinesia, for instance. Our method aims at preventing
these coarse-motor side-effects.

The inhibition of fine motoricity, however, is desired;

it indicates the necessary blockade of dopamine
receptors, which is most easilv detected in hand-
writing. 1963 Haase designated the occurrence of an
extrapyramidal inhibition of movement, discernible ln
fine motoricity as "neuroleptic threshold".

The patients write a text of 12 lines, i.e. three times

the first verse of the song "Der Mai ist gekommen ... "
under standardized conditions; same time of day, same
seat, same pen or pencil, same deskpad; the patient
is instructed not to lift the pen when writing, all this
being conducive to a good reproducibility of the test.

Under neuroleptic influence the handwriting shows a

stiffening and diminution as well as a shortening of
the lines. The general impression of the handwriting lS
monotonous, the writing itself loses its sweep and
becomes stiff.
521
Our method of a computerized evaluation is a
quantitative test. The size of the handwriting and the
change it undergoes are determined in an objective
reproducible and rational manner. The handwriting lS
picked up by a black-and-white TV camera in an
illuminated box, and the .picture thus obtained is
transformed by the computer into digits. By vertical and
horizo~tal scanning the computer picks up the blackening
produced by the letters. The lengths of all twelve lines
are measured as well as the distances of the first and
fourth lines of each verse at the right-hand and left-
hand margins. From these data the areas covered by the
verses is computed. The three verses are then added
added up to the total area; this means, that the spaces
between the individual verses are eliminated.

The area of the reference handwriting measured and stored

before neuroleptic treatment is equalled to "1'', and the
percentage deviation (a diminution as a rule) of the test
writing, i.e. the handwriting under neuroleptic treatment,
is worked out. Apart from the total area as a main para-
meter the so-called quotient is calculated, by which we
understand the percentage deviation of the area of the
third verse from the area, covered by the first verse.
This is based on the observation, that a higher degree
of neuroleptic impact results in a deformation of the
handwriting in the meaning of a "pyramid turned upside
down", i.e. that the third verse becomes smaller in
relation to the first verse.

Establishing the Neuroleptic Th£~§_h()_ld:

Diminution of the handwriting (total'area in the

standardized handwriting test) as an expression of
neuroleptic hypokinesia with low (incipient) dosages
n = 15
total area:
percentage deviation from the reference specimen (which
equals 100)

Our investigations have shown that with a diminution of

the total area by -13 % and more in the standardized
handwriting test the neuroleptic threshold is crossed, if
the trend of the subsequent three days confirms this
specific diminution. By taking the trend into considera-
tion random deviations are excluded.

522
 days

LS: reference handwriting

NS: neuroleptic threshold
solid line: mean (arithmetic)
Figure 1.

With light neurolepsia, as well as in the case of normal

subjects the quotient (starting at - 1) indicates a
moderate to severe neuroleptic hypokinesia as a rule. The
inhibition of fine motoricity displays a biphasic
development.

At first here is a dynamic initial phase. An exponential

diminution of the size of the handwriting occurs,
modifications of the neuroleptic dosage result immediately
in reactions in the size of the handwriting. In this
period early dyskinesias and early acathisia occur as
undesired side-effects. This dynamic initial phase is
followed by the static phase, which may also be called
plateau phase because of the persistant diminution of
the handwriting.

A steady-state of the drug is now established; in the

case of a more pronounced neuroleptic inhibition this
also includes the extrapyramidal coarse motoricity, such
as akinesia or rigor. The influence of dosage reductions
on the size of the handwriting as a parameter of fine-
motor hypokinesia appears only in a tardy manner.

The dopamine receptors become sensitive, and even with

dosage reduction hypokinesia may be increased for a short
time. In this period, therefore, lower dosages of the

523
neuroleptic agent are sufficient to achieve a fine-motor
hypokinesia of the same degree than were necessary at the
beginning of treatment - in the dynamic initial phase.

If the neuroleptic agent is discontinued in this phase,

hypokinesia still persists in fine motoricity for weeks
and may be even months, depending on the duration of
pretreatment and the dosage in relation to individual
sensitiveness. There is an increased tendency to react
to a renewed neuroleptic treatment with extra-pyramidal
coarse-motor side-effects. This is of importance for
instance when planning a washout period in drug trials.

The drug-induced inhibition of fine motoricity measured

in the handwriting test may be of several degrees. A
diminution of the total area in the handwriting test by
-13 to -20% is considered "light", "moderate" would be
-21 to -30 %, while anything above 30 % would be termed
as "severe". In this latter range coarse-motor symptoms
occur with increasing frequency.

With medium dosages, i.e. twice to ten times the threshold

dosage, the range of light and moderate hypokinesia, i.e.
-13 to -30 %, is skipped. Anticholinergic agents such as
Akineton neutralize the inhibition in a dosage-dependant
manner. At the same time, however, they reduce the
antipsychotic neuroleptic effect and are therefore not
suitable for routine therapeutic use.

Of particular value is the method presented here for

planning a maintenance therapy with long-acting
neuroleptic agents. If the total area of the handwriting
is reduced by more than 30 %, caution is recommended when
using neuroleptic agents. They should only be administered
in reduced dosages. The evaluation of the size of the
handwriting with the aid of a computer will show,
whether fine-motor inhibition will be increased. In that
case the dosage interval should be lengthened and later
on the dosage should be reduced.

524
FINE-MOTOR CONTROL OF NEUROLEPTIC DOSAGE

I. Bitter
Psychiatric Clinic of the Semmelweis University
Budapest
Hungary

The early recognition of the side-effects of a drug ranks

amongst a physician's most important tasks. The hand-
writing specimen facilitates the early recognition
especially of the hypokinetic side-effects of neuroleptic
agents. A latent parkinson syndrome becomes discernible
by the strain of the fine-motor activity of writing on
the extrapyramidal system, i.e. during neuroleptic treat-
ment hypokinesia manifests itself earlier micrographically
in the handwriting than clinically. The handwriting
specimen as a reproducible test, i.e. handwriting test,
is of the same importance to the extrapyramidal system
as are the tests by Gierlich, Barre and Mingazzini to the
pyramidal system, because these tests may help us to
diagnose a hidden, latent paresis.
Amongst the evalution criteria for the handwriting the
total area covered by the three verses is of primary·
importance. We examined the variations in the toral area
covered by the three verses for 51 normal volunteers, who
copied a familiar verse three t~mes on the dame day; four
subjects exhibited variations ranging from 13 to 22 %
(Table 1). Schizophrenic patients of whom three samples
of the same text were taken, demonstrated more variation
than did the healthy subjects.
We performed the same test in a~coholics not treated with
drugs and not showing any withdrawal symptoms and found
the variation in the total area of the three verses so
considerable, that we should like to suggest handwriting
specimens of alcoholics by analysed only by experts with
525
a lot of practice in this particular field. For this
reason the handwriting specimens of alcoholics must be
excluded from computerized evaluation. When evaluating
the specimen with the aid of a computer these facts must
also be borne in mind in the interpretation of the
results, and the specimen may have to be visually
inspected by an expert, if necessary. In this case the
characteristics of modification described by HAASE (4)
will be considered: stiff, smaller, and more constricted
handwriting than before neuroleptic treatment.

By means of the handwriting test we can determine the

neuroleptic threshold. By neuroleptic threshold dosage
we understand the dosage of a neuroleptic agent eliciting
extrapyramidal symptoms discernible in handwriting. In
consideration of the neuroleptic in threshold dosages
HAASE (4) reduced the neuroleptic agents - the relation
being chlorpromazine = 1 - to a system, that became
known as the neuroleptic equivalent table.

I wish to demonstrate some of the possibilities of

applying the handwriting test in the clinical trial of
bromperidol. The investigation of bromperidol showed a
haloperidol-like effect, but bromperidol has a longer
elimination half-life than haloperidol (2, 3, 6, 7).
Eleven female alcoholics and 16 schizophrenic patients
were included in our study.

Our study confirms, that the neuroleptic threshold

dosage of bromperidol is similar to that of haloperidol,
amounting to about 3 mg, which means, that bromperidol
belongs to the very potent neuroleptics (1).

We tried to find clinical indicators for the neuroleptic

disposition.

As former investigations (3) in patients up to age 60

have shown, the disposition for the neuroleptic threshold
is independend of size, weight, age, sex, insulin and
alcohol tolerance. On the basis of our investigations
neither the severity of psychosis (measured by BPRS)
nor the improvement (measured by BPRS) nor the number
of previous admissions nor the administered doses of
the neuroleptic (in mg) may be interpreted as indicators
of the individual disposition for the neuroleptic
threshold (or extrapyramidal "sencitiveness"). Our
questions was: Is there a connection between dosage and
improvement or between dosage and side-effects? Since
the absolute dosage had no bearing on the question, we

526
should like to demonstrate the dosages applied in
relation to the individually established neuroleptic
threshold (Table 2). This table shows our practice of
incipient dosage, but the question arises: Was the
gradual increase in dosage the result of our impatience,
since the condition was stabilized after about 8 days?

There existed no statistical correlation between the

improvement and the administration of individually
determined neuroleptic threshold dosages or of a multiples
thereof, and no connection between the improvement and
the total dosages applied (Table 3).

The side-effect did not display a statistical

correlation with the total dosages applied and with the
neuroleptic threshold dosages, but there was a highly
significant correlation between side-effects and the
individually determined multiple threshold dosage
(Table 4). This means firstly: According to clinical
experience the administration of threshold dosages
does not elicit any important side-effects. It means
secondly: Only an individual determination of the so-
called "dyskinetic zone" as well as of the neuroleptic
threshold is of clinical importance. This correlation
between multiple neuroleptic threshold dosages and
side-effects provides, of course, no explanation for
the very important clinical question: Why do side-
effects in the course of therapy appear in a different
manner at the same dosage and in the same patients?

To answer these questions more studies are required,

in which also the pharmacokinetic and biochemical
aspects may prove to be essential.

527
DIAGNOSIS OF TARDIVE DYSKINESIA

Christopher G. Goetz

Rush-Presbyterian St. Lukes Medical Center

1725 S. Harrison Street
Chicago, IL 60612

The syndrome of tardive dyskinesia (TD) encompasses

a series of abnormal involuntary movements associated
with chronic exposure to neuroleptic agents. To es-
tablish the diagnosis with confidence, the history of
prolonged medication must accompany the movements that
characterize TD. One feature without the other neces-
sarily casts doubt on the diagnosis. Each diagnostic
feature will be discussed briefly followed by a section
on differential diagnosis.

ABNORMAL INVOLUNTARY MOVEMENTS

The movements that typify TD are usually choreic in

nature. The early description of this syndrome stressed
abnormal motor movements affecting the facl' the so-
called bucca-lingual masticatory syndrome. This con-
sisted of involuntary mouthing, chewing, sucking, and
licking movements of the tongue. Later, the syndrome
description was broadened to include a variety of ab-
normal muscular manifestations, including choreoathetoid
type movements of the fingers, hands, arms and feet, and
ballistic-type movements (particularly of the arms) , as
well as axial hyperkinesias, and diaphragmatic 2movement
resulting in grunting and difficult breathing. As the
condition was studied more, the mouthing movements were
further refined to include puckering, panting, smacking,
and tongue movements inside the mouth. The clinical
picture closely resembles L-dopa-induced dyskinesias and
other choreatic disorders, both in the character of the
movement abnorma~ity and in the diversity of the indi-
vidual symptoms.
529
 Special emphasis should be placed on the remarkable
diaphragmatic and truncal involvement that can lead to
severe respiratory distress and erratic, often unintelli-
gible speech in these patients. We have seen numerous
dyspneic patients in consultation who were originally
evaluated for myocardial infarctions, pulmonary emboli
and other primary cardio-respiratory conditions. The
accompanying generalized chorea or involuntary truncal
dyskinesias coincident with such fluctuating shortness of
breath suggested that the respiratory symptoms related to
the choreic disorder. The treatment with dopaminergic
depleting agents has succ~ssfully abated the respiratory
complaint and the chorea.

Speech abnormalities in TD have not received promi-

nent attention. In our center, 12 consecutive patients
with well established TD underwent detailed language
evaluation by trained speech pathologists who analyzed
oral reading and vowel production for deviation and
multiple speech dimensions. Six patients were identified
with abnormal speech. In this group temporal organi-
zation and voice production dimensions were the most
severely deviant and were highly related to the overall
intelligibility and bizarre quality of speech. Deviations
in articulation were less prominent. Evaluation of the
degree and distribution of these patient's chorea using
the abnormal and voluntary movement scale (AIMS) revealed
a significant correlation between speech abnormality and
truncal chorea. In contrast, lingual-facial-buccal and
total body chorea scores did not idffer between the two
groups. This study suggests that a specific subgroup of
choreic TD patients may be at high risk for speech com-
munication disorders. Furthermore, if speech therapy is
recommended to such patients, rehabilitative efforts
should be aimed at modifying the selective speech di-
mensions identified.

Although choreic movements are most characteristic

of TD, some patients show predominantly dystonic ab-
normalities ~n association with varying degrees of chorea.
Burke et al. described 23 patients treated with chronic
neuroleptic agents who developed pure dystonic reactions.
They termed this phenomenon tardive dystonia. Three
had experienced dystonic reactions with introduction of
neuroleptic therapy, and in four of nine patients whose
past history was detailed, abnorm31 birth or develop-
mental delay could be identified. The association of
psychiatric symptoms with dystonia remains a somewhat
bewildering and controversial topic; a definite cause and
effect relationship between chronic neuroleptic exposure

530
and progressive dystonia has not been established.
NEUROLEPTIC EXPOSURE
The second requirement for the diagnosis of TD is an
established history of chronic exposure to neuroleptic
medication. The definition of "chronic" is vague but a
minimum of 3 months exposure is generally accepted. The
provocative agents, structurally varied, all share the
common property of dopamine receptor site blockade. Most
usually, these medications are prescribed for psychiatric
reasons, although patients treated with these agents on a
long term basis for nausea or vomiting have also been
plagued with the disorder [prochlorperazine (compazine)
and metoclopramide (Reglan)]. The movements usually
appear as the dose of neuroleptic is decreased or the
treatment is interrupted, but may occur when the patient
continues a steady dose of the neuroleptic drug.

DIFFERENTIAL DIAGNOSIS

Because extrapyramidal movements and neuroleptic

drugs are often associated with one another, the differ-
ential diagnosis of TD is especially important. Tardive
dyskinesia usually begins after 1-2 years of chronic
neuroleptic treatment, and hence is temporally distinct
from other extrapyramidal disorders such as drug-induced
parkinsonism or dystonia. These latter conditions tend
to occur earlier in the course of neuroleptic therapy.
Drug-induced parkinsonism is felt to relate to dopa-
minergic receptor-site blockade in the striatum. The
clinical features of parkinsonism which include brady-
kinesia, resting tremor, rigidity, and loss of postural
reflexes appear to result whenever striatal receptor
sites do not receive appropriate dopaminergic stimu-
lation. Resting tremor is rarely misdiagnosed as TD,
except in cases where there is tremor in the jaw region.
Jaw tremor is also referred to in the psychiatric liter-
ature as "rabbit syndrome," since the patients have
rhythmic ~n4 gepetitive movements that resemble a rabbit's
munching. ' ' Since parkinsonism and TD are based on
very distinct pharmacologic alterations that require very
different management, the addition of another term,
regardless of how colorful it may be, tends to obscure
essential distinctions. Resting tremor, whether oral-
facial or limb, tends to occur coincident with commence-
ment or increase in neuroleptic medication and is associ-
ated with other parkinsonian features. It relates patho-
physiologically to decreased activity of dopaminergic
function at the striatal level and is therefore treated
in the same manner as other parkinsonian features.

531
 Neuroleptic induced dystonias are usually seen early
in the course of therapy and hence would not be confused
with TD. Dystonic manifestations can be quite diverse,
although the most dramatic clinical signs involve the
eyes and neck, leading to oculogyric crises and severe
torticollic spasms. As discussed earlier, recent reports
exist of a late onset, tardive dystonia associated with
neuroleptic exposure, and hence the presence of dystonia
or sustained abnormal and involuntary po~ture alone
cannot be used to exclude a TD syndrome.

An especially important form of dystonia that could

be confused with TD is Meige's syndrome or orofacial-
cervical dystonia. Because orofacial muscles are typi-
cally involved in both entities, the two might be thought
to resemble one another. However, the sustained charac-
ter of the dystonic contractions causing forced eye
blinking (blepharospasm) and maintained facial grimacing
are the hallmarks of Meige's syndrome. In contrast,
facial movements in TD are rapid and unsustained; irregu-
lar tongue protrusion and lip smacking are more charac-
teristic features. Pharmacologically Meige's syndrome is
distinct from TD in that there is no needed prior history
of neuroleptic exposure and anticholinergic drugs gener-
ally abate the dystonia. Both Meige's syndrome and TD
may improve with reserpine.

It is important to distinguish TD from other hyper-

kinetic syndromes, especially stereotypic mannerisms and
agitation related to psychotic behavior and akathisia
associated with early administration or recent elevation
in dose of neuroleptics. A wide variety of mannerisms
are seen in the psychotic population, and these peculiar
behavio9s may resemble oral-facial dyskinesias or limb
chorea. In the vast majority of cases, however, they
are highly ritualistic in character and consistent in
anatomic distribution, which disgintuishes them from the
unpredicatable irregularity of choreiform movements.
Furthermore, mannerisms are usually conscious and sym-
bolically significant to the patient, whereas TD is often
completely unrecognized or of little concern. The former
movements also are seen as part of the early psychotic
state and ~8 not usually emerge after prolonged drug
treatment. A psychotic or mentally deranged patient
with early onset chorea or dystonia should raise the
additional important diagnostic considerations of extra-
pyramidal disorders like Huntington's chorea, dystonia
musculorum deformans, and Wilson's disease.
An initial glance at TD patients might suggest that

532
they are agitated, since there is an abundance of move-
ment often with shifting in the chair, irregular tapping
of feet, or strange mouthing movements. While these
behaviors might suggest an overly nervous person, TD does
not resemble the hyperactivity seen in psychotic states
like manic-depressive illness or involutional psychosis.
In the latter cases, a generalized psychomotor activation
is seen characterized by overactivity, but with well
coordinated purposeful movements and often exaggerated
expressions of affect.

Acute akathisia can be differentiated from TD as a

state of motor restlessness that occurs following early
i~mi~if~ration or an augmentation of neuroleptic dose.
' ' It is not clear, however, that this effect
relates specifically to extrapyramidal dysfunction.
Akathisia does not occur naturally-except in very rare
occurrenci~ of Parkinson's disease-in any extrapyramidal
disorder. While it does occur coincident with neuro-
leptic administration, there are no data to confirm that
akathisia relates to neuroleptic effects on dopaminergic
activity or the extrapyramidal system. While it can be
mildly modified by the administration of anticholinergic
agents, the effect of this drug class on akathisia is
much le~s marked than in cases of drug induced parkin-
sonism. The hallmark of akathisia is the striking
complaint of inner tension, a need to move about and
resist inactivity. Usually, the patient feels forced to
shuffle his feet restlessly and may feel unable to sit or
lie down. Acute akathisia can be differentiated from TD
in its natural history, since the former occurs coinci-
dent with medication induction or dosage increase and is
not associated with fine involuntary movements so charac-
teristic of TD. The patient interview also helps since
the movements of akathisia are thought to be secondary to
the internal restlessness, whereas TD is not necessarily
associated with mental agitation of any sort. In fact,
often the patient is unaware of his TD.

Multifocal tics of childhood and Gilles de la

Tourette's syndrome can be confused with TD, especially
since these patients are often treated with neuroleptic
agents. The characteristic distribution of the abnormal
movements in the two disorders is different with lingual-
buccal movements more prominent in TD. These patients,
however, are often treated with neuroleptic drugs and
hence may be at a significant eventual risk for the
development of TD as a second involuntary movement dis-
order distinct from their tic disorder. The differential
diagnosis includes finally other forms of chorea, spe-

533
cifically Huntington's disease and a variety of drug or
metabolically-induced movement disorders. A dominant
family history of similar movements suggest Huntington's
chorea. These patients are treated with neuroleptic
agents and whether they develop eventual TD along with
their Huntington's chorea would be impossible to de-
termine. Hore perplexing, we have seen patients from
families with Huntington's disease who were treated with
neuroleptics for psychiatric adjustment problems. These
patients eventually develop chorea. Whether the psychi-
atric problem was the initial presentation of Huntington's
disease and therefore the chorea represented his genetic
disorder, or whether the psychiatric problem was an
environmentally related disorder and the chorea was
iatrogenic, is unknown. A history of chronic amphetamine
abuse, current hyperthyroidism, pregnancy, or birth
control pill ingestion should suggest chorea related to
these etiologies and be dis~~nguished from the neuro-
leptic related TD syndrome. Choreiform movements may
also occur spontaneously in the elderly population, a
condition termed senile chorea.

A recent report suggesting new diagnostic categories

for late onset drug related dyskinesias in~~uded "with-
drawal dyskinesia" and "covert dyskinesia. " The former
was defined as self limited TD, resolving within 12 weeks
after drug withdrawal; covert dyskinesia was defined as
dyskinesia seen with discontinuation or reduction of
neuroleptic dose without spontaneous abatement. These
distinctions, while of theoretical interest, are im-
practical. Since many patients develop dyskinesia after
medication withdrawal, the diagnosis would first be
withdrawal dyskinesia for 12 weeks and, then, if there
was no abatement, the diagnosis would be changed to
covert dyskinesia. The change in diagnosis gives no
insight into pharmacology or pathophysiology, and the
time of 12 weeks is at best arbitrary. As the authors of
the report admitted, these.distinctions may confuse
rather than clarify clinical thinking.

The accurate diagnosis of TD is essential to studies

of epidemiology, treatment and natural history. The
development of quantitative and qualitative rating scales
that can incorporate dystonic and choreic features and
yet distinguish them from parkinsonian movement abnor-
malities will permit more accurate reporting. The com-
plex description of TD and its often inadequate dis-
tinction from other extrapyramidal disorders have con-
tributed to the large and confusing literature that
plagues investigators in this field.

534
REFERENCES

1. L. Uhrbrand and A. Faurbye, Reversible and irreversible dys-

kinesia after treatment with perphenazine, chlorpromazine,
reserpine, E.C.T. therapy. Psvchopharmacologia 1: 408-418
(1960).
2. M. Bourgeois and J. Tignol, Les dyskinesies tardives des neuro-
leptiques. T.a Nov. Pres. Med. 6: 3649-3650 (1977).
3. C.G. Goetz and H.L. Klawans, Tardive Dyskinesia, in: "Psychi-
atric Aspects of Neurologic Disease II:" D.F. Benson, D. Blumer
eds., Grunne and Stratton, NY (1982).
4. W.J. Weiner, C.G. Goetz, P.A. Nausieda and H.L. Klawans, Res-
piratory dyskinesias: extrapyramidal dysfunction presenting as
shortness of breath. Ann. Int. Med. 88: 327-331 (1978).
5. R.E. Burke, S. Fahn and J. Jankovic, Tardive dystonia.
Neurology 31: 79 (1981).
6. K. Jus, A. Jus and J. Gautier, Studies on the action of certain
pharmacological agents on tardive. and on the rabbit syndrome.
Int. J. Clin. Pharmacol.. 9: 138-145 (1974).
7. R. Sovner and A. Diamascio, Effect of benztrophine in the rabbit
syndrome and tardive dyskinesia. Am. J. Psych. 134: 1301-1302
(1977).
8. A. Villeneuve, Rabbit syndrome-A peculiar extrapyramidal re-
action. Can. Psvchiatr. Assoc. J. 17: 69-72 (1972).
9. C.D. Marsden, D. Tarsy, R.J. Baldessarini, Spontaneous and drug-
induced movement disorders in psychiatric patients, in, "Psychi-
atric Aspects of Neurologic Disease," D.F. Benson, D. Blumer
eds., New York, Grune & Stratton (1975).
10. R. Degwitz, K.F. Binsack, H. Herkert, Zur problem der persist-
ierenden extrapyramidelen Hyperkinesen nach langfritiger an-
wendang von Neuroleptica. Nervenarzt 38: 170-174 (1967).
11. F.J. Ayd, A survey of drug-induced extrapyramidal reactions.
JAMA 175: 1054-1060 (1961).
12. F.J. Ayd, Neuroleptics and extrapyramidal reactions in psychi-
atric patients. Rev. Can. Bio1. 20: 451-459 (1961).
13. J.B. Martin, S. Lal, G. Tolis and H.G. Friesen, Inhibition by
apomorphine of prolactin secretion in patients with elevated
serum prolactin. J. Clin. Endocrin. and Metab. 39: 180-182
(1974).
14. G. Selvy, Parkinson's disease, in, "Handbook of Clinical
Neurology," Diseases of the Basal Ganglia, B. Vinken, G.W. Bruyn
eds., Amsterdam, Elsevier (1968).
15. C. Goetz, W.J. Weiner and H.L. Klawans, Treatment of the
choreas, in, "Disorders of Movement," A. Barbeau ed.,
Philadelphia, Lippincott (1981).
16. G. Gardos, J.O. Cole and D. Tarsy, D.: Withdrawal syndromes
associated with antipsychotic drugs. Am. J. Psvch. 135: 1321-
1324 (1978).

535
SPONTANEOUS DYSKINESIAS AMONGST PSYCHOI'IC PATIENTS

C. r::avid l\'larsden

university Department of Neurology

Institute of Psychiatry and King' s College
Hospital Medical School, De Crespigny Park
London S. E. 5, England

INTRODUcriON

The presence, nature and cause of abnormal involuntary

movements (dyskinesias) amongst untreated and drug-treated
psychotic patients has intrigued psychiatrist s and neurologists
but has caused great controversy.

The dopamine hypothesis of schizophreni a attributes the

illness to pathological overactivity of cerebral dopamine
mechanisms. So investigator s have hunted for evidence amongst
patients with untreated schizophreni a of those dyskinesias
which neurologists attribute to dopaminergic overactivity .
Whether they occur or not is hotly debated.

The advent of neuroleptic drugs, all of which block

dopamine receptors, was associated with the recognition of a
wide variety of dyskinesias amongst those so treated for lang
periods of time. Such tardive dyskinesias have been
attributed to the effect of chronic neuroleptic intake leading
to the emergence of dopamine receptor supersensitivi ty.
However, there are sane who do not believe this concept. They
still hold that such dyskinesias are not caused by chronic

537
neuroleptic intake, l::ut are due to the underlying psychotic
illness.

DYSKINF.SIAS AMON.;sr UNTREATED PSYCIDI'IC PATIENI'S

Kraepelin was quite specific in his description of

abnonnal novanents in schizophrenia in his nonograph en
Dementia Praecox. "Sane of them resanble novanents of
expression, wrinkling of the forehead, distorticn of the comers
of the IOOUth, irregular novanents of the tongue and lips,
twisting of the eyes, opening them wide, and shutting them
tight, in short, those novanents which we bring together under
the narre of ''making faces" or gr:imacing: they remind one of
the corresponding disorders of choreic patients. • •. Several
patients continually carried out peculiar sprawling, irregular,
choreifonn, outspreading rrovenents, which I think I can best
characterise by the expression "athetoid ataxia".

A ca.rualreading of this and the many other psychiatric

texts containing similar statements might suggest that those
dyskinesias familiar to neurologists as chorea and athetosis
occur in patients with schizophrenia. However, this is not. so.
Psychiatrists have used the ~rd chorea and athetosis to
describe what they see in schizophrenic patients, tut what.
they have seen has not been the sane as occurs in Hlmtingtan •·s
disease or dystonic Im.lsculomm defonnans.

Che reascn for canfusicn is that many of the series of

schizophrenic patients described in the earlier parts of this
century were contaminated with patients exhibiting a schizo-
phrenifonn psychosis secondacy to other brain disease. Syphilis
abounded, and in the late second and third decades, encepha-
litis lethargica swept the world. This illness left in its
wake many individuals with a canbination of a movement dis-
order and a psychotic illness.

A second cause of diagnostic difficulty has been the

recognition and definition of the various notor disorders that
do occur in schizophrenia, and their distinction fran neuro-
logical dyskinesias. The majority of the al:nonnal movements
seen in schizophrenia are described under the tenus stereo-
typies, or mannerisns. Kraepelin defined a stereotypy as the
persistence of the sane volitional action either as a con-
tinued position, or as a repetition of the sane movement.
He defined a mannerism as an adornment of volitional acticn,
causing unnatural and affected expression. Stereotypies thus

538
are purposeless repetitive motor acts, while mannerisms are
goal-directed acticns carried out in unusual and unnecessarily
bizarre ways. stereotypies and mannerisms frequently co-exist
and often roorge one with the other. 'Ihey may affect simple
bodily rrotor activity, speech, or ccmplex private or social
rituals. Sterotypies and mannerisms were cxmnonest in
catatcnic schizophrenia, in whan they were seen in as many as
a quarter of patients. However, catatonic schizophrenia is
now relatively infrequent, so the florid stereotypies and
mannerisms that illustrated earlier texts on schizophrenia
are rarely seen.

Confused tenninology is a plague in this area. Many

stereotypies and mannerisms, because of their repetitive
character, might be called tics by sane observers, particularly
if they can be suppressed by the patient temp:>rarily. A fixed
head tilt might be called a static stereotypy by one, or
torticollis by another, while Meige and the French school
classified torticollis as a tic. It is easy to see haw the
bizarre gait of a child with dystonia musculorum defonnans, or
the ann posture of an adult with dystonia writer's cramp,
might be tenood manneristic.

Yet, despite all these difficulties, it remains

generally true that no neurologist with experience in movement
disorders is convinced that neurological dyskinesias are
consistently present in patients with schizophrenia.

TARDIVE DYSKINESIAS

That al:nonnal involuntary IIOVerents (dyskinesias) are

provoked by chronic treatment with neuroleptic drugs (tardive)
now is accepted by nost, but not all, investigators. It is
mrth considering the evidence leading to this conclusion.

Originally, suspicion was raised by anecdote. Early

descriptions in the 1950's and 1960's were no more than case
reports of the appearance of dyskinesias in patients treated
chronically with such drugs. To begin with many doubted the
association. Identical dyskinesias were known to occur
arrongst patients never treated with neuroleptics, particularly
amongst survivors of the epidemic of encephalitis lethargica.
Neurologists, now witnessing these dyskinesias in those
treated with neuroleptics, were dubiuous that the drugs "Were
responsible. However, the weight of evidence began to tell

539
because the frequency of such dyskinesias amongst those on
chronic neuroleptic therapy increasingly appeared to exceed
anything seen before. In other TNOrds, epida:niological
evidence on the point prevalence of tardive dyskinesias was re-
ported to be in excess of that of similar spcntaneous dyskin-
esias amongst drug-free populations. Nevertheless, it became
apparent that spcntaneous abnonnal novement disorders -were not
uncamon, although their frequency was uncertain.

At this point, it becaire evident that the diagnostic

catego:ry of "tardive dyskinesia" was being used to encanpass
a variety of al:nonnal novements of different types.

The camonest category of tardive dyskinesias undoubtably

is the classical b.lcco-lingual-masticatory disorder of
chewing, lip-sucking, and tongue protruding IOO'VE!!'OOJ1ts. How-
ever, other patients on chronic neuroleptic treatment have
been described with dystoo.ia (tardive dystonia), simple tics
(tardive tics), generalised tics with vocalisation (tardive
'!burette) , and isolated truncal, vocal and other dyskinesias.
The assumptioo. that these la".:ter novements are due to neuro-
leptics cannot be sustained with coo.fidence en epidemiological
grounds, for their frequency amongst those not treated with
such drugs vrcis unknown.

Of course, the only way to prove the point is to conduct

extensive cross-sectioo.al estmates of point prevalence for
each dyskinesia amongst drug-treated and drug-free populatioo.s,
matched for age, sex, and other important variables. Yet this
is difficult to achieve for the rarer dyskinesias, so the
assumpticn that they are provoked by neuroleptics ranains anec-
dotal.

Another major problem is that epidemiological proof of a

higher incidence of a given no~ disorder in those treated
with such drugs does not prove that the drugs alone caused
the condition. An altemative hypothesis is that neuroleptic
administratioo. merely unearths an underlying neurological
abnonnality. Indeed, there is sane evidence that this may be
the case.

The· point prevalence of bucca-lingo-masticatory dyskinesia

amongst patients oo. chronic neuroleptic treatment has been
reported to vary fran ~.5 to 65%. Such wide variations
reflect differences in diagnostic criteria, as -well as

540
differences in patient populaticns. Many surveys have
employed standardised rating scales to document the presence
of a dyskinesia. For example, one widely used measure has
been the Abnonnal Involuntary l>bvement Scale (AIMS) , devised
by the Naticnal Institute of Mental Health. But this ordinal
scoring system cannot reliably distinguish be~ mild
nonnal lip-smacking. particularly in those with a dry IOOUth
and no teeth, and obvious unequivocal abnonnal IOOUth IOOVeirellts.
Other difficulties include the problem of distinguishing
between minor reversible degrees of tardive dyskinesia and
the more severe intractable cases, and the fact that continued
neuroleptic intake may mask mild tardive dyskinesias. There
also has been debate as to whether the prevalence of tardive
dyskinesias is greater arcongst chronic psychiatric in-patients
as against those managed as out-patients. ~r, both
groups appear susceptible. Final!y, there is the problem of
how long a period of neuroleptic intake is required to justify
the designation of "tardive". Most authorities have
settled on a minimum duration of treatnent for three rronths,
but this is an arbitary figure. Taking these various points
into account, it is now generally accepted that about 20% of
those on chronic neuroleptic therapy may exhibit a significant
tardive dy~cinesia.

'!his point prevalence figure must be set against the

prevalence of spontaneous bucco-lingucMnasticato:ry dyskinesias.
'!here is no doubt but that these are more frequent amongst the
elderly. However, the point prevalence of spontaneous
dyskinesias even in the old is unlikely to exceed 8%.
Accordingly, we may conclude that abnonnal involuntary IOOVe-
nents are likely to be tw::l to three tlires more prevalent in
neuroleptic-treated patients than in individuals not exposed
to such drugs.

541
ANIMAL MODELS OF TARDIVE DYSKINESIA

Harold L. Klawans, Paul M. Carvey, Caroline M. Tanner,

Christopher G. Goetz
Department of Neurological Sciences, Rush-Presbyterian St.
Luke's Medical Center, 1725 W. Harrison Street, Chicago, IL
60612, USA

For the last 12 years we have been studying the rodent model of tardive
dyskinesia (TD) in both rats and guinea pigs. All of these studies have
involved the observation of dopamine agonist-induced behavior following
chronic neuroleptic administration. The theoretical basis behind these
studies is the assumption that neuroleptic induced 1alterations of
dopaminergic function occur in both TD and these models. These models
have employed dopamine agonist-induced stereotyped behaviors as the
measure of presumed postsynaptic dopamine receptor site functional
sensitivity in the striatum.

In initial behavioral studies, groups of guinea pigs were given dally

subcutaneous injections of chlorpromazine (5 mg/kg for 3 weeks in one
group, 10 mg/kg for 4 weeks in the other group). Beginning 4 days after the
last injecti'2n of chlorpromazine amphetamine was given to the two groups
of animals. It was found that doses of amphetamine that were too small to
produce stereotyped behavior in control animals were effective in producing
stereotyped behavior in the chlorpromazine pretreated animals. The animals
were reevaluated weekly for 4 weeks, and it was found that the alteration in
threshold for stereotyped behavior persisted. These results were consistent
with our hypothesis that the neuroleptic pretreatment in some way altered
the sensitivity of the dopaminergic receptors in the striatum to the
dopamine release induced by the amphetamine. An alternative explanation
that the absolute amount of dopamine available to act at the striatum could
have been increased as a result of chlorpromazine pretreatment is unlikely,
since the concentration o~ ~opamine in the brain after chronic neuroleptic
therapy is not elevated • ' With such normal levels it is unlikely that
amphetamine is inducing release of increased amounts of dopamine in
chlorpromazine pretreated animals.
543
In a second study we used the same treatment schedule but used
apomorphine, a direct acting dopamine agonist, as the test drug and found a
persistent _fecrease in the threshold for apomorphine-induced stereotyped
behavior. Because apomorphine acts directly at dopaminergic receptor
sites in the striatum to elicit stereotyped behavior, these results support the
hypothesis of a postsynaptic, neuroleptic-induced alteration in sensitivity to
dopaminergic agents.

It appears from these initial studies that neuroleptic pretreatment altered

the sensitivity or the threshold of the dopaminergic receptors to' dopamine
and hence altered the threshold for the production of amine-induced
stereotyped behavior. Likewise, the striatal dopaminergic receptors in
patients undergoing neuroleptic therapy may be altered in such a way that,
following prolonged therapy, normal amounts of dopamine may trigger an
abnormal response.

Since this original description of chlorpromazine-induced behavioral

supersensitivity was reported, other investigators have reported simi~
findings with other neuroleptics in a variety of experimental animals. '
These studies all indicate that chronic administration of antipsychotic drugs
followed by drug withdrawal produce a behavioral state of increased
responsiveness to dopamine agonists, consistent with the development of
supersensitivity that lasts for at least several weeks.

In our more recent studies we have used this model to study several
epidemiologic issues in TD. In doing these studies we have focused on three
separate questions regarding neuroleptic-induced dopaminergic
hypersensitivity. (a) Is there a relationship between the daily neuroleptic
dose and the degree of eventual hypersensitivity produced? (b) Do
equipotent antipsychotic doses of different neuroleptics differ in their
ability to elicit hypersensitivity? (c) Is duration of treatment a risk factor
for the development of hypersensitivity? If the model is an accurate one of
TD in humans, the answers to these questions have theoretical significance,
but also very practical implications concerning the classes and doses of
neuroleptics that are safest in regard to prevention of TD.

Initially we administered different doses of four neuroleptics

(chlorpromazine, prochlorperazine, thioridazine, and trifluoperazine) to
groups of guinea pigs or rats for 3 weeks, and the degree of behavioral
supersensitivity was evaluated using the dopamine agonist apomorphine.
The resulting bar graph dose-response curves are shown in Fig. 1.
Chlorpromazine, prochlorperazine, and trifluoperazine showed a8 definite
dose effect, with larger doses causing greater hypersensitivity • At low
doses, no hypersensitivity was seen. Thioridazine induced no hypersensitivity
at the doses studied. These results suggest that exposure dose is a risk
factor for the development of dopaminergic striatal hypersensitivity at least
for some neuroleptics. If applied to TD in humans, the model suggests that
keeping the daily neuroleptic dose as low as possible may decrease the
incidence of the disorder, and, in fact, that very low doses may not be

544
 ?0 20

I~ 15
~
w
Vl 10 10
+I

~
..
c
0 5 5

..J
<I .. s ( 02 004 010 020 I 00 NS 0.10 0.20 050 100 500
PROCHLORPERAZINE THIORIDAZINE
~~

z
w
0::
w
u
u
c:
....
:5u 20 20
Vl
..J 15 15
<I
0::
Q
> 10 10
<I
X:
w
en
5 5

,. s 010 I 00 5.00 100 NS 002 004 010 040 2 00

CHLORPROMAZINE TRIFLUOPERAZINE
FIGURE 1: Behavioral hypersensitivity to a challenge doese of
apomorphine after three weeks of daily treatments with four
neuroleptics.

545
associated with any risk of hypersensitivity. The same data are graphed in
Fig. 2 after conversion from specific milligram doses to chlorpromazine
equivalents. This method allows the investigator to compare behavioral
effects-in this case, hypersensitivity-caused by different neuroleptics with
the milligram doses controlled for comparable antipsychotic potency. The
graph demonstrates that at equivalent antipsychotic doses. some drugs
appear more likely than others to induce behavioral hypersensitivity. The
direct extrapolation of these data to the incidence of TD in humans has not
been confirmed. Such human studies are difficult to perform, however, since
patients commonly receive many different neuroleptic agents over a
several-year period.

More recently, we have extended these studies to include the

nonphenothiazine neuroleptic haloperidol (Fig.3). Once again, after 3 weeks
of treatment the daily dose of haloperidol administred was directly related
to the increase in stereotypic behavioral response (hypersensitivity). It
appears from this study that the dose of haloperidol given daily, like that of
most phenothiazine neuroleptics, is a risk factor for behavioral
hypersensitivity in animals and perhaps by analogy for TD in man.

This study differed from the earlier studies in that the daily administration
of haloperidol was contihued for 6 and 9 weeks and the behavioral response
to apomorphine was observed three separate times (i.e., after 3, 6 and 9
weeks of continuous neuroleptic treatment) (Fig.3). This protocol allowed us
to look at both daily dose and duration of continuous therapy as risk factors
for behavioral hypersensitivity. The data demonstrate an increase in
behavioral response between 3 and 6 weeks, suggesting that for some period
of time the duration of therapy is a risk factor. This is not true for the last
time period, i.e., between 6 and 9 weeks. We have observed this same time
relationship with other neuroleptics-chlorpromazine and trifluoperazine
(authors' unpublished observations).

It thus appears that although duration of administration is a risk factor for

hypersensitivity over some time period, this risk factor does not extend as
long as the duration of therapy continues, but rather eventually ends. This is
most consistent with the human data. That duration is a risk factor for TD is
inherent in the definition (i.e., a complication of chronic as opposed to acute
exposure). On the other hand, the incidence may not be linear with duration
of therapy, and many patients despite extremely long periods of continuous
exposure never develop TD. These observations raise the possibility that the
major effect of duration is early in the course of therapy in man, and that at
some time much later in the course of continuous neuroleptic therapy the
duration of therapy is no longer a major risk factor. Unfortunately, it is
impossible from our data to suggest the exact time determinants in man.

Analogous experiments of 3, 6, and 9 week pretreatment periods have been

carried out with two other neuroleptics: fluphenazine and thioridazine.

The progressive behavioral hypersensitive response to apomorphine after 3,

546
 20.0

-
-~
r:::

-
Cl)
~ 15.0

Ci
....
Cl)

0
(.)
en
010.0
0
....
·:;;:
c
.c:.
m
Cl)
•
)(

THIORID

-1.0 -0.5 X 1.5

Log-Dose Chlorpromazine Equivalents

FIGURE 2: Behavioral hypersensitivity scores from Fig. 1 graphed in log
dose chlorpromazine equivalents.

CPZ Chlorpromazine

PROCHLO- Prochlorperazine

THIORID - Thioridazine

TRIFLUO - Trifluoperazine

547
01
~
(X)

38
withdraw 3 days
38 test with 0.75 mg;kg
34 Apomorphine
=E
w
u; 32

+I 30
u;
c
I)(
A.
:::;)
0
a:
0
ID
Cl)

NS 005 01 05 10 20 NS 005 01 05 10 NS 005 01 10

20 05 10

DOSE OF HALOPERIDOL (mg/kg)

3WEEKS &WEEKS 9WEEKS
FIGURE 3: Behavioral hypersensitivity following chronic haoperidol
treatment.
6, and 9 weeks of fluphenazine is shown in Figure 4. Analysis of variance on
behavioral score differentials was significant at all test points, 3 week F-
value 4.312 (p< 0.005) 6 week F-value 12.12 (p< 0.005) 6 week F-value 12.12
and 9 week f-value 12.676 (p< 0.0001). The mean group comparison to
control was significantly hypersensitive at both 6 (p<0.05) and 9 (p(0.05)
weeks. By 9 weeks, all individual dose treatment subgroups except the
lowest dose were significantly different from controls (p 0.05).

In contrast, thioridazine did not induce behavioral hypersensitivity (Figure

5). Analysis of variance on the behavioral score differentials demonstrated
no statistical change from baseline at 3 weeks, F-value 1.507 (NS), or 6
weeks, F-value 2.599 (NS). At 9 weeks, animals were subsensitive, F-value
3.051 (p( 0.05). Similarly, the mean group differentials demonstrated rpo
significant change at 3 or 6 weeks and subsensitivity (p <0.05) at 9 weeks ••

The study again demonstrated that all neuroleptics do not have the same
propensity to induce behavioral dopaminergic hypersensitivity. Our prior
studies with chlorpromazine, haloperidol, prochlorperazine, and
trifluoperazine indicated that these agents, like fluphe~ 0 zine, all induce
behavioral hypersensitivity to apomorphine by 9 weeks. Thioridazine is
unique among the neuroleptics we have studied in that it not only did not
induce hypersensitivity, but furthermore, induced subsensitivity to
dopaminergic striatal stimulation. This observation suggests that
thioridazine when given chronically has properties distinct from these other
neuroleptics. Clearly they all block dopamine receptors, and acutely at he
doses administered in our study inhibit apomorphine-induced stereotypic
behavior to a similar degree. Their chronic effects on the dopaminergic
system may, however, not be the same. Studies of dopaminergic antagonist
receptor binding in striatum after three weeks of neuroleptic treatment
have demonstrated that thioridazine differs from fluphenazine and
haloperidol in that the fon.ner drug effects no change where the latter two
enhance receptor binding. •

In our latest study, we have employed a protocol to test whether

thioridazine when added to haloperidol can prevent haloperidol induced
hypersensitivity. Our results suggest: 1) concurrent thioridazine can prevent
haloperidol hypersensitivity, 2) once haloperidol induced hypersensitivity is
established, continuation of haloperidol will either extend or increase it, but
both the addition of thioridazine to continued haloperidol and the
substitution of thioridazine for haloperidol allow the hypersensitivity to
reverse despite continuous neuroleptic therapy. These observations suggest
that thioridazine may play a role in neuroleptic induced movement
disorders.

This model can also be used to study the natural history of neuroleptic-
induced hypersensitivity. In these studies, animals were pretreated with
haloperidol for 3 weeks, and their behavioral responses to apomorphine were
observed sequentially for 6 weeks.

549
 ~2
wtlhdrow 3 days
tut ••lh 0 H m~/k~

Apomorphine
48

44
0

"'
+I

..
40

"'
IX 36
Q.
::>

..,a:
0
32

28

24

JII ••. 00101DlOJ0SI0 ·-·· 0.01 0.1 0.1' 0) 0.1 1.0 N.l. 0.01 01 01 O.J 01 10

3 WEEKS 6 WEEKS 9 WEEKS

DOSE OF FLUPHENAZINE (mQ/kQ)

FIGURE 4: Behavioral hypersensitivity following chronic fluphenazine

treatment.

550
 w•thdrow 3 days
38 te$1 w11h 0 75 mQ/kQ

H Apomorph1n1

Jr.

'5
0
34
(f)

tl 3l

(f) 32
<[

IX
Q_
::::>
0
0::
0

NS 01 02 0.5 100 NS 0.1 02 05 100 NS 01 02 05 100

3 WEEKS 6 WEEKS 9 WEEKS

DOSE OF THIORIDAZINE (mg/kg)

FIGURE 5: Behavioral hypersensitivity fol!owing chronic thioridazine
treatment.

551
552
Figure 6 shows the behavioral response of the haloperidol-treated guinea
pigs as compared with 1<2ontrols at various times after discontinuation of
haloperidol treatment. The animals exhibit an increased behavioral
response to subthreshold doses of the dopamine agonist apomorphine 1 week
and 3 weeks after haloperidol therapy, as can be seen from the increased
intensity and duration of the stereotyped behavior. However, this increased
behavioral responsiveness diminishes 6 weeks after the cessation of
haloperidol administration. These results indicate that the neuroleptic-
induced increased behavioral response to a dopamine agonist in animals is
reversible with time.

REFERENCES

1. Goetz, C.G., Weiner, W.J. Nausieda, P.A., Klawans, H.L.: Tardive

dyskinesia: pharmacology and clinical implications. Clin.
Neuropharmacol. 1982; 5: 3-22.

2. Rubovits, R, Patel, B.C., Klawans, H.L.: Effect of prolonged

chlorpromazine pretreatment on the threshold for amphetamine-
induced stereotypy: a model for tradive dyskinesias.
Adv. Neural. 1973; 1: 671-5.

3. Guldberg, H.C., Yates, C.M.: Effects of chlorpromazine on the

metabolism of catecholamines in dog brain. Br. J, Pharmacal. 1969;
36:535-7.

4. Nyback, H., Sedvall, G.: Effect of chlorpromazine on accumlation and

disappearance of catecholamines formed from tyrosine C-14 in brain.
J. Pharmacal. Exp. Ther. 1968; 162: 294-8.

5. Klawans, H.L., Rubovits, R.: An experimental model of tardive

dyskinesia. J. Neural Transm. 1972; 33: 235-46.

6. Tarsy, D., Baldessarini, R.J.: Pharmacologically-induced behavioral

supersensitivity to apomorphine. Nature (New Biol.) 1973; 245: 262.

7. Gianutsos, G.,Drawbaugh, R.B., Hynes, M.D., Lal, H.: Behavioral

evidence for dopaminergic supersensitivity after chronic haloperidol.
Life Sci. 1974; 14: 887-98.

8. K1awans, H.L., Carvey, P., Nausieda, P.A., Goetz, C.G., Weiner,

W.J.: Effect of dose and type of neuroleptic in an animal model of
tardive dyskinesia. Neurology (NY) 1980; 30: 95.

9. Goetz, C.G., Carvey, P.M., Tanner, C.M., Klawans, H.L.:

Neuroleptic-induced dopamine hypersensitivity. Life Sci. 1984; 34:
1475-1479.

10. Goetz, C.G., Klawans, H.L.: Controversies in animal models of

553
 tardive dyskinesias. In, Movement Disorders, ed., S. Fahn, C.D.
Marsden, pp. 250-263, Butterworth, London, 1982.

11. Snyder, S.H., Banerjee, S.P., Yamamura, H.I., Greenberg, D.: Drugs,
Neurotransmitters, and Schizophrenia: Phenothiazines,
amphetamines, and enzymes synthesizing psychotomimetic drugs aid
schizophrenia research. Science 184: 1243-1253, 1974.

12. Hitri, A., Carvey, P., Weiner, W.J., Klawans, H.L.: Biochemical and
behavioral studies of neuroleptic induced behavioral supersensitivity.
In: Fann, W.E., Smith, R.C., Davis, J.M., Domino, E.F., eds. Tardive
Dyskinesia: Research and Treatment. New York: Spectrum
Publications, 1980: 145-63

554
EXPERIMENTAL TARDIVE DYSKINESIA

Lars M. Gunne and Jan-Erik Haggstrom

Psychiatric Research Center
S-750 17 Uppsala, Sweden

In our laboratory we have contributed to the development of

animal models for tardive dyskinesia (TD) in Cebus apella monkeys
(Gunne and Barany 1976, Barany et al. 1979) and rats (Gunne et al.
1982). The abnormal movements observed in these monkeys were
either confined to the bucca-lingual area (with tongue protrusion,
athetotic tongue movements, and perioral twitchings) or spread to
neck, trunk and extremities. In the rats we only noticed a supra-
normal rate of vacuous chewing movements. All dyskinetic movements
persisted throughout the final drug withdrawal phase before sacri-
fice.
In both animal models chronic administration of haloperidol
or fluphenazine decanoate at monthly intervals has been applied on
intact animals. Instead of the last monthly injection of depot
neuroleptic a final injection of regular (non-depot) neuroleptic
drug formulation was substituted. All animals (both monkeys and
rats) were then sacrificed 2 months after this final dose. During
their long-term neuroleptic treatment the monkeys occassionally
received biperiden injections (70 ~g/kg) when attacks of acute
dystonia occurred. Both animal models were validated by pharmaco-
logical modification of the dyskinetic movements using various
transmitter agonists and antagonists (Barany and Gunne 1979,
Gunne et al. 1982). The primate model responded to dopamine agon-
ists and antagonists, acetylcholine agonists and antagonists and
GABA agonists like the majority of human subjects afflicted with
TO. The rodent model, on the other hand, showed a different re-

555
sponse to cholinergic agonists and antagonists, similar to what
has been reported in certain TD patients with an opposite, "para-
doxical" response to these drugs (Casey and Denney 1977, Jeste and
Wyatt 1979).
The monkey brains were subjected to cryosectioning vertical
to a line through the anterior and posterior commissures and at
right angles to the sagittal plane dividing both brain halves.
Defined brain areas were located, micropunched and taken for
analysis of brain monoamines and their metabolites, GABA, dynor-
phin, met-enkephalin and substance P and 3 different enzymes:
tyrosine hydroxylase (TH), choline acetyl transferase (CAT) and
glutamic acid decarboxylase (GAD). The present paper describes
only part of this ongoing study and deals mainly with results on
the regional distribution of GAD activity and the brain levels of
homovanillic acid (HVA) in dyskinetic animals and controls. GAD
activity was measured by a modification of Wu et al (1976) as
14 co formation from 1- 14 c-glutamic acid in the presence of pyri-
doxat-5-phosphate (Gunne and Haggstrom 1983). HVA was assayed by
massfragmentography according to Wiesel and Sedvall (1974).
The rat brains were divided in the following way: coronal
sections were made 1 and 2 mm in front of the optic chiasm. From
the 1 mm disc the cortex and white matter was removed and the rest
was subdivided into a striatal section and accumbens. Another 1 mm
section was made immediately in front of the pons and was sub-
divided into the nigral area (both compacta and reticulata) and
the superior colliculus area (the division was made by a horisontal
line half way down the upper surface and the aquaeduct). In addi-
tion a 5 x 5 mm area of the frontal cortex was taken for analysis.
Group comparisons were made by independent t-tests between 3
different groups: untreated controls, ch~onically neuroleptic-
treated animals without dyskinesia (neuroleptic-treated controls)
and dyskinetic animals.
Results
In most brain areas neuroleptic-treated controls (n=5) or
dyskinetic monkeys (n=6) did not differ significantly from un-
treated controls (n=7) with regard to GAD activity. However, there
were two exceptions. In the substantia nigra and in the globus
pallidus we noticed a reduction of GAD activity in monkeys with
persistent dyskinesia (fig. 1), but not in neuroleptic-treated
controls. This reduction of GAD activity was more prominent in

556
the substantia nigra (p<O.OOl) than in the pallidal region
{p<0.05). In a monkey with unilateral dyskinesia (in the left arm
and hand) there was a reduction of GAD only in the substantia nig-
ra of the opposite side. ~Jhen GAD activity levels for individual
monkeys were compared by chi-square tests the group of dyskinetic
monkeys was significantly lower (p<0.05) than untreated and neuro-
leptic-treated controls (table 1). Neostriatal HVA levels did
not differ between groups of monkeys.
In rats we noticed a corresponding reduction of nigral GAD
activity when vacuous chewing behavior had been induced after one
year of chronic haloperidol treatment. Table II shows that this re-
duction of GAD activity was confined to the substantia nigra, but
left other brain areas intact (globus pallidus was not investiga-
ted). When the depression of GAD activity was plotted against in-
dividual chewing behavior there was a significant (p<0.002) cor-
relation between these two phenomena (Gunne and Haggstrom 1983).

!-lkat/dm3
15
GAD D Untreated
controls
10 fZl Neuroleptic
treated
5 controls
• TO
0
Subst Globus
Nigra Pallidus

Fig. l. GAD activity levels (means~ S.E.) in substantia

nigra and globus pallidus. Comparison between 7 un-
treated controls, 5 neuroleptic-treated monkeys with-
out dyskinesia and 6 dyskinetic animals ( *p<0.05;
*** p<O. 00 l ) .

557
 Table I. Individual GAD activity levels (nmoles/mm~/h) in the
substantia nigra of 7 drug-naive Cebus apella monkeys,
5 chronically neuroleptic-treated monkeys without
dyskinesia and 6 animals with neuroleptic-induced
dyskinesia. The levels of each monkey represent the
mean of 2 - 4 samples. Untreated and neuroleptic-
treated c~ntrols were compared with dyskinetic group
using chi test.
Treatment n Dyskinesias GAD activity Chi 2 test

None 7 no 34.2 47.3 23.3 p<0.05

35.2 40.7 33.5
58.5
Chronic 5 no 36.5 42.1 60.0 p<0.05
neurolep- 45.1 37.6
tics
Chronic 6 yes 11.2 18.4 22.2
neurolep- 24.3 14.8 15.5
tics

Table II. GAD activity (nmoles/mm 3;h) in different brain

areas of rats given chronic haloperidol treatment
for fourteen months and untreated controls (means
- S.D.).
Area Untreated Chronic halo- Significance
controls (n=7) peri dol (n=14)
Franta 1 Cortex 17.8 +
+
2.9 16.2 + 2.3
+
N.S.
Striatum 14.0 + 2.3 19.4- 7.4 N.S.
Accumbens 33.5 - 9.7 35.6 !" 9.2 N.S.
Sup. Colliculus 33.3 +
+
9.4 37.8 +- 9.9 N.S.
Subst. Nigra 49.5 - 9.5 33.6 +- 5.0 p<0.001

558
Discussion
The present paper is a first demonstration in chronically neu-
roleptic-treated animals of a reduction of the GABA synthesizing
enzyme GAD and a direct relation between this phenomenon and a
persistent neuroleptic-induced movement disorder.
The reduction of GAD activity in both globus pallidus and
substantia nigra may indicate a reduced tissue density of GABA
neuron terminals which may be related to a dysfunction within the
striato-nigral and accumbens-pallidal neuron systems. These GABA-
ergic systems are known to be deeply affected during long-term ad-
ministration of neuroleptic drugs. Whereas intrinsic striatal GABA
neurons are disinhibited during neuroleptic-induced suppression
of dopamine transmission, these inter-nuclear GABA pathways are
inhibited (cf Gale and Casu 1981). In our experiments the suppres-
sion of pallidal and nigral GABA neurons was still manifest as low
GAD activity in dyskinetic animals but not in similarly neurolep-
tic-treated controls even two months after discontinuation of
treatment. Our findings demonstrate that there are individual dif-
ferences in the capacity of these GABA systems to regain their
original function after chronic depression by neuroleptic drug
treatment. These pathways may occasionally in susceptible indivi-
duals incur some as yet undefined lesion possibly through disuse
degeneration. Such a degeneration may underlie the low nigral and
pallidal GAD activity and could be an important biochemical back-
ground to the irreversible movement disorder induced by long-term
neuroleptic treatment, which has many traits in common with TO in
humans. We were unable to record any increase in neostriatal
levels of HVA in the dyskinetic animals. Such a rise might have
been expected in view of the regulatory role on nigrostriatal DA
neurons which is often ascribed to nigral GABA terminals.
Both in humans with TO and in our animal models of this dis-
order, GABA agonists tend to reduce the symptomatology (Barany and
Gunne 1979, Gunne et al. 1982). However, fairly large doses, in-
ducing mild ataxia, must be given to achieve such alleviation.
The present findings may not lead to major therapeutic improve-
ments, but an understanding of the mechanisms underlying TO may
facilitate the development of relevant screening tests for new
antipsychotics and may thus prove to be important for the pre-
vention of this drug-induced complication.

559
References
Barany, S. and Gunne, L.-M., 1976, Pharmacological modifica-
tion of experimental tardive dyskinesia, Acta Pharma-
cal. Toxicol., 45:107-111.
Barany, S., lngvast, A. and Gunne, L.-M., 1979, Development
of acute dystonia and tardive dyskinesia in cebus
monkeys, R~s. Commun. Chern. Pathol. Pharmacal.,
25:269-279.
Casey, D.E., and Denney, D., 1977, Pharmacological charcter-
ization of tardive dyskinesia, Psychopharmacology,
54:1-8.
Gale, K. and Casu, M., 1981, Dynamic utilization of GABA in
substantia nigra: regulation by dopamine and GABA in
the striatum, and its clinical and behavioral implica-
tion, Mol. Cel. Biochem., 39:369-405.
Gunne, L.-M. and Barany, s., 1976, Haloperidol-induced tar-
dive dyskinesia in monkeys, Psychopharmacology, 50:
237-240. . .
Gunne, L.-M., Growdon, J. and Glaeser, B., 1982, Oral dyski-
nesia in rats following brain lesions and reuroleptic
drug administration, Psvchooharmacology, 77:134-139.
Gunne, L.-M. and Haggstrom, J.-E., 1983, Reduction of nigral
glutamic acid decarboxylase in rats with neuroleptic-
induced oral dyskinesia, Psychopharmacology (sub-
mitted).
Jeste, D.V. and Wyatt, R.J., 1979, In the search of treatment
for tardive dyskinesia: Review of the literature,
Schizophrenia Bull:, 5:251-293.
Wiesel, F.A. and Sedval 1, G., 1974, Post-mortal changes of
dopamine and homovanillic acid levels in rat striatum
as measured by mass fragmentography, Brain Res., 65:
547-550.
Wu, J.U., Wong, E., Saito, K., Roberto, E. and Schousboe, A.,
1976, Properties of 1-glutamate decarboxylase from
brains of adult and newborn mice, J. Neurochem., 27:
653-659.

560
EPIDEMIOLOGY OF TARDIVE DYSKINESIA

Ross J. Baldessarini

Departments of Psychiatry and Neuroscience Program

Harvard Medical School; Mailman Research Center
McLean Hospital, Belmont, MA 02178

As emphasized by earlier speakers at this symposium on tardive

dyskinesia (TD), ambiguous definitions of this complex behavioral
syndrome limit certainty about its epidemiology (Baldessarini et
al., 1980; Jeste & Wyatt, 1982a; Kane & Smith, 1982; Tarsy &
Baldessarini, 1984). Reported prevalence of TD among patients
exposed to neuroleptics reflects the criterion of severity of
abnormal movements required for the diagnosis. Thus, Smith and
colleagues (1979a; 1979b) found a close, inverse correlation be-
tween the nominal prevalence rates of TD (ranging from less than
10%, to over 70%) in hundreds of inpatients and outpatients diag-
nosed as schizophrenic and the severity of dyskinetic movements
as rated by the NIMH Abnormal Involuntary Movement Scale (AIMS--
a common, internationally employed instrument for this disorder
(Baldessarini et al., 1980). There is also a need to correct raw
prevalence rates of TD for the risk of "spontaneous" or idiopathic
dyskinesias in similar populations not treated with a neuroleptic
agent. In our recent review of such studies, the mean rate (all
data are ±SEM) of spontaneous dyskinesias among 18 studies was 5.8
±1.0% (Smith & Baldessarini, 1980), and others have found even
higher rates (over 10%), especially among elderly patients (Kane
& Smith, 1982). Among 45 studies, we found a mean (raw) preva-
lence of TD-like manifestations among neuroleptic-treated patients
of 24.0 ±2.5%, and suggested a "best-estimate" for prevalence of
neuroleptic drug-associated TD, corrected for the spontaneous risk
rate, of 18.5 ±2.4%. To further complicate matters, a recent sur-
vey by Owens et al. (1982) found only slightly less dyskinesia
among a small group of chronically hospitalized schizophrenic pa-
tients who had not been treated with a neuroleptic than in a large
group exposed to ordinary antipsychotic chemotherapy. In part, this

561
similarity may reflect age-related spontaneous dyskinesias of older
persons as well as possibly increased risk of abnormal movements
other than TD among chronically psychotic patients (Owens, 1983).
Despite the uncertainty surrounding the amount of risk for the man-
ifestations of TD ascribable to neuroleptic treatment, there is
little doubt that such agents contribute to the risk (Baldessarini,
1974; Baldessarini et al., 1980). Data on the incidence of TD are
rare and remain inadequate. A recent study by Kane and associates
(1982) suggests that among schizophrenic patients followed over four
years of cumulative neuroleptic exposure at ordinary doses, the rate
of appearance of new cases of TD was about 12%. An impression is
that incidence rates are not linear over time (Kane & Smith, 1982),
but that there is a peak risk between six months and perhaps 2-5
years of neuroleptic treatment, and the data of Kane et al. (1982)
indicate a fairly steady incidence at 2-5% per year over 4 years of
treatment.

There are many unknown aspects of the natural history of TD:

the condition tends to emerge rapidly and not to be clearly pro-
gressive. Neuroleptic treatment can "mask" the appearance of dys-
kinesia, and it is common to observe some worsening of TD soon
after discontinuation of a neuroleptic agent; this effect is well
illustrated in a study by Branchey et al. (1981), in which there
was little emergence of TD with gradual discontinuation of ordinary
doses of loxapine, until the last 12% of the initial dose was fin-
ally replaced by an inactive placebo. Once established, TD does
not seem to be consistently progressive with continued neuroleptic
treatment (Kazamatsuri et al., 1973) although continued neuroleptic
or other anti-dopamine treatments seem essentially irrational and
their long-term safety is not established (Wyatt & Jeste, 1982b).
Current practice is to avoid continued use of a neuroleptic when-
ever possible once the diagnosis of TD is made (Baldessarini, et
al., 1980). Another clinically important insight is that the rate
of spontaneous remission of TD after discontinuation of a neuro-
leptic agent can be very slow, and may require many months or even
one to three years.

There has been considerable attention paid to the question of

whether certain agents, or treatment with high doses of a neurolep-
tic, might carry an increased risk of TD. The largely inconclusive
evidence on this point has been reviewed critically elsewhere
(Baldessarini et al., 1980; Jeste & Wyatt, 1982). One of the
largest collections of data on the question of type of agent is in
a US survey in the late 1970s involving over 10,000 VA patients
treated with various neuroleptic agents that cover the range of
potencies and chemical types (Gee & Mesard, 1979). In these data,
the prevalence rate of TD ranged only between 6% and 10%, with no
clear association with a specific type of agent, although flu-
phenazine ranked somewhat higher than other agents. While it is

562
reasonable to expect a relationship between dose and risk of TD,
this hypothesis has not been supported consistently by the available
literature (Baldessarini, et al., 1980). For example, Simpson et al,
(1978) found no association of TD prevalence with mean or total dose
of neuroleptic or associated antiparkinsonism agent. Despite the
uncertainty regarding relative risks of various agents, a clinically
prudent position for the present is to maintain a high index of sus-
picion toward all available neuroleptic agents, as virtually every
type in common clinical use has been associated with TD, regardless
of potency or chemical type.

Matters of drug type and dose are very difficult to evaluate

adequately as risk factors. In most surveys, only current medica-
tion is reported, but typically, chronically psychotic patients have
been exposed to a complex variety of agents, often with incomplete
documentation, over many years. Evaluation of an hypothesis that
potent, depot neuroleptics may carry added risk is complicated by
the question of delivery of drug by a reliable (injected) vs. an un-
reliable (oral) route of administration. There is also strong evi-
dence that neuroleptics of high-potency tend to be prescribed in
relatively higher daily chlorpromazine-equivalent doses than those
of low potency (Baldessarini et al., 1984). More generally, there
is growing evidence that typical maintenance regimens involve doses
of neuroleptics that are much greater than necessary for many pa-
tients, perhaps by a factor of five to ten (Baldessarini & Davis,
1980; Branchey, 1981; Kane et al., 1983; Baldessarini et al.,
1984). If this impression is correct, then dosing near the top of
a putative dose vs. risk curve would tend to overwhelm any chance
of observing such a relationship, although one might be hypothe-
sized to exist in the region of the clinical EDso of the agents,
and some preliminary data support that prediction (Kane et al.,
1983).

There is also a strong suspicion that endogenous factors con-

tribute to the risk of TD. Others in this symposium have addressed
the question of whether patients with chronic psychotic disorders
called schizophrenic have an increased risk of spontaneous dyskin-
esias or of neuroleptic-associated TD; this hypothesis has little
solid support (Marsden et al., 1975). Indeed, many cases of TD
have appeared in patients given neuroleptics who were not schizo-
phrenic. It has also been suggested that patients with affective
disorders may have an increased risk of TD (Rosenbaum et al.,
1977. Nevertheless, appropriate epidemiologic analyses of rela-
tive risk rates among diagnostic groups remain to be carried out.

Perhaps the most powerful predictor of risk for TD is age

(Smith & Baldessarini, 1980). The elderly are at a higher risk
for both spontaneous and neuroleptic-associated dyskinesias. Both
prevalence and severity ratings for TD increase as a factor of age,

563
virtually linearly between about 40 and 70 years. Moreover, the
chance of spontaneous remission after discontinuing a neuroleptic
appears to diminish with age throughout the entire adult life-cycle
(Smith & Baldessarini, 1980). While spontaneous remission rates
are apparently higher in adolescents and young adults, young pa-
tients are not immune to TD; indeed severe, crippling cases with
choreic and dystonic features have been observed in young men.

Perhaps neuroleptic treatment may not be a sufficient "cause"

of TD, but a contributing factor added to endogeneous pathogenic
changes in brain function among patients at risk, especially the
elderly. Studies of the effects of aging on forebrain systems im-
plicated in the pathophysiology of TD (Baldessarini & Tarsy, 1979;
1980) have led to complex and rather confusing findings. Based on
laboratory studies, older rodents are more sensitive to haloperi-
dol (Campbell & Baldessarini, 1981) a dopamine (DA) antagonist,
and to apomorphine, a direct DA agonist (Smith et al., 1978;
Campbell et al., 1984); both changes appear to involve pharmaco-
kinetic factors (Kapetanovic et al., 1982; Campbell et al., 1984).
Indeed, most of the laboratory evidence concerning DA receptor
abundance or function in aged rodents (Smith & Baldessarini,
1980; Campbell et al., 1984) suggests a decrease, rather than the
increase predicted by the DA-supersensitivity hypothesis in the
pathophysiology of TD (Baldessarini & Tarsy, 1979;1980; Tarsy &
Baldessarini, 1984). Comparable studies of the pharmacokinetics
and pharmacodynamics of neuroleptic agents in man are rare, but
suggest increased blood levels with advancing age (Yesavage et
al., 1981), as also occurs with many psychotropic agents (Triggs
& Nation, 1975).

In conclusion, the current epidemiology of TD includes many

areas of incomplete or inconclusive information. Prevalence rates,
corrected for the risk of spontaneous dyskinesias are about 15-20%
of those at risk (exposed regularly to virtually any neuroleptic
agent for more than a year). Incidence rates may be about 2-5%
per year for the first few years of neuroleptic maintenance. The
natural history of the illness is confusing, as it may not be con-
sistently progressive with further neuroleptic treatment. No class
of antipsychotic agents in common current clinical use is proven to
have a clearly lower or higher risk of association with TD than any
others. For the future, there is much interest in newer atypical
antipsychotic agents with little antidopaminergic action and little
risk for acute extrapyramidal reactions (such as clozapine and the
new agent, fluperlapine, discussed elsewhere at this Congress),
based on the hypothesis that such agents may be neurologically less
toxic than older, typical neuroleptics. The most important predic-
tor of risk for TD and for poor spontaneous recovery from it is
advancing age, especially above 50 years. As there is no satisfac-
tory treatment for TD (Jeste & Wyatt, 1982b), and even without the
backing of solid epidemiologic support for these recommendations,

564
it is reasonable to try to minimize risks of TD by thoughtful and
conservative long-term use of neuroleptic medications for clear
indications (chronic psychotic illness that responds to treatment),
and in the lowest effective doses, which may be as low as 10-20%
of those in common use in the US today (Baldessarini et al., 1980).

ACKNOWLEDGMENTS

Supported, in part, by NIMH awards MH-47370, MH-31154 and by

travelling funds generously provided by the Sandoz Corporation.

REFERENCES

Baldessarini, R.J., 1974, Tardive dyskinesia: A review of the etio-

logical association with neuroleptic therapy, J. Can. Psychiat.
Assoc., 19:551-554.
Baldessarini, R.J., Cole, J.O., Davis, J.M., Gardos, G., Preskorn,
S., Simpson, G.M., and Tarsy, D., 1980, Tardive Dyskinesia,
Task Force Report 18, Washington DC, American Psychiatric
Association, 205 pp.
Baldessarini, R.J. and Davis, J.M., 1980, What is the best mainte-
nance dose of neuroleptics in schizophrenia? Psvchiat. Res.
3:115-122.
Baldessarini, R.J., Katz, B., and Cotton, P., 1984 (in press), Com-
parison of doses of neuroleptic agents of high and low potency,
Am. J. Psychiatry.
Baldessarini, R.J. and Tarsy, D., 1979, Actions of neuroleptic
drugs and the pathophysiology of tardive dyskinesia, Int. Rev.
NPurobiol., 21:1-45.
Baldessarini, R. J .. , and Tarsy, D. , 1980, Dopamine and the pathophysi-
ology of dyskinesias induced by antipsychotic drugs, Ann. Rev.
Neurobiol., 3:23-41.
Branchey, M.H., Branchey, L.B., and Richardson, M.A., 1981, Effects
of neuroleptic adjustment on clinical condition and tardive
dyskinesia in schizophrenic patients, Am. J. Psychiatry 138:
608-612.
Campbell, A. and Baldessarini, R.J., 1981, Effects of maturation and
aging on behavioral responses to haloperidol in the rat, Psy-
chopharmacology, 73:219-222.
Campbell, A., Stoll, A., Maynard, P., and Baldessarini, R.J., 1984
(in press), Effect of age on behavioral responses and tissue
levels of apomorphine in rat, Neuropharmacology.
Gee, S. and Mesard, L., 1979, Psychiatric Drug Study, Part I Psy-
chiatric Ward/Unit Study, Office of the Controiler, Monograph
No. 9, Washington, DC, US Veterans Administration, 73 pp.
Jeste, D.V. and Wyatt, R.J., 1982a, Understanding and Treati~g
Tardive Dyskinesia, New York·, The Guilford Press, 3b3 pp.

565
Jeste, D.V. and Wyatt, R.J., 1982b, Therapeutic strategies against
tardive dyskinesia: Two decades of experience, Arch. Gen.
!sychiatrv, 39:803-816.
Kane, J.M., Rifkin, A., Woerner, M., Reardon, G., Sarantakos, S.,
Schiebel, D., and Ramos-Lorenzi, J., 1983, Low-dose neuroleptic
treatment of outpatient schizophrenics, Arch. Gen. Psychiatry,
40:893-896.
Kane, J.M. and Smith, J.M., 1982, Tardive dyskinesia: Prevalence
and risk factors, Arch. Gen. PsvC!_hi~l!_try, 39:473-481.
Kane, J.M., Woerner, M., WeinhoLd,~ •• Wegner, J., and Kinon, B.,
1982, A prospective study of tardive dyskinesia development:
Preliminary results, J. Clin. Psychopharmacol., 2:345-349.
Kapetanovic, I.M., Sweeney, D.J., and Rapoport, S.I., 1982, Age
effects on haloperidol pharmacokinetics in male, Fisher-344
rats, J. Pharmacal. Exp. Ther., 221:434-438.
Kazamatsuri, H., Chien, G.R~ and Cole, J.O. 1973, Long-term treat-
ment of tardive dyskinesia with haloperidol and tetrabenazine,
Am. J. Psychiatr~, 130:479-483.
Marsden, C.D., Tarsy, D., and Baldessarini, R.J., 1975, Spontaneous
and drug-induced movement disorders in psychotic patients, in
D.F. Benson and D.F. Blumer (eds.), Psychiatric Asoects of
Neurological Diseases, New York, Grune & Stratton, pp 219-265.
Owens, D.G., 1983, Personal communication, 19 May.
Owens, D.G., Johnstone, E.C., and Frith, C.D., 1982, Spontaneous in-
voluntary movement disorders, Arch. Gen. Psychiatry, 39:452-461.
Rosenbaum, K.M., Niven, R.G., Hanson, N.~. and Swanson, D.W., 1977,
Tardive dyskinesia: Relationship with primary affective dis-
order, Dis. Nerv. Svst., 38:423-427.
Simpson, G.M., Varga, E., Lee, J.H., and Zoubok, B., 1978, Tardive
dyskinesia and psychotropic drug history, Psvchopharmacologv,
58:117-124. .
Smith, J.M. and Baldessarini, R.J., 1980, Changes in prevalence,
severity, and recovery in tardive dyskinesia with age, A~ch. Gen.
Psychiatry, 37:1368-1373.
Smith, J.M., Kucharski, L.T., Eblen, C., Knutsen, E., and Linn, C.,
1979a, An assessment of tardive dyskinesia in schizophrenic
outpatients, Psvchopharmacolog~, 64:99-104.
Smith, J.M., Kucharski, L.T., Oswald, W.T.,, and Waterman, M.A., 1976b,
A systematic investigation of tardive dyskinesia in inpatients,
Am. J. Psychiatry, 136:918-922.
Smith, R.C., Leelavathi, D.E., and Lauritzen, A.M., 1978, Behavioral
effects of dopamine agonists increase with age, Commun. Psycho-
pharmacal., 2:39-43.
Tarsy, D. and Baldessarini, R.J., 1984 (in press) Tardive dyskinesia,
Ann. Rev. Med., 35:000-000.
Triggs, E.J. and Nation, R.L., 1975, Pharmacokinetics in the aged: A
review, J. Pharmacokin. BioPharm. 3:387-418.
Yesavage, J.A., Holman, C.A., and Cohn, R., 1981, Correlation of thio-
thixene serum levels with age • .Psychopharmacology, 74:170-172.

566
GABA-ERGIC TREATMENT OF TARDIVE DYSKINESIA

K. Heinrich, J. Tegeler, W. Woller, H. Quadbeck,

G. Arendt, E. Klieser, H. Lange and E. Lehmann
Psychiatrische Klinik der Universitat DUsseldorf
Bergische Landstr. 2, D 4000 DUsseldorf 12

While the exact pathophysiology of neuroleptic-induced

tardive dyskinesia is still unknown, dopaminergic super-
sensitivity and a disturbed balance between the dopamin-
ergic and cholinergic systems seem to play an important
role in this disorder. GABA, which exerts an inhibitory
influence on dopaminergic neurotransmission, is also sup-
posed to be involved.
A variety of pharmacological agents has been reported
to improve tardive dyskinesia: neuroleptics acting as
dopamine receptor blockers, dopamine depleting agents,
dopamine agonists, cholinergics, anticholinergics, GABA-
ergic substances, as well as several other drugs. However,
none of these therapeutic strategies proved to be satis-
factory. In many cases, therapeutic successes were only
transient and could not be replicated in subsequent stu-
dies. Recently, Jeste and Wyatt (9) reviewed 285 treatment
studies including more than 3000 patients with neuroleptic-
induced tardive dyskinesia. They found 19 studies on the
treatment with drugs acting via GABA-ergic mechanisms in-
cluding a total of 204 patients. Improvement was seen in
54 %, compared to an improvement rate of 67 % in studies
with neuroleptics, 46% in treatments with other dopamine-
antagonists, and 39% in pharmacological trials with chol-
inergic drugs. So, among nonneuroleptic drugs, GABA-ergic
compounds exhibited the highest improvement rates. On the
other hand, the small number of controlled studies, the
short duration of the evaluation periods, and the small
number of patients included suggest caution in interpreting
these findings.
567
 The following substances with putative GABA-facilit-
ating pharmacological activity have been investigated in
tardive dyskinesia: the benzodiazepines, baclofen, musci-
mol, THIP (tetrahydroisoxalolopyridinol), gamma-acetylenic-
GABA, gamma-vinyl-GABA, and sodium valproate. Several stu-
dies suggested that benzodiazepines may be effective in
tardive dyskinesia patients (7, 8, 10, 13, 15, 25, 26),
though the precise mechanisms by which benzodiazepines
exert their action on hyperkinetic movements are still
unclear. For example, diazepam was reported to be bene-
ficial in studies by Singh (25), Jus et al. (10) and Singh
et al. (26). Clonazepam reduced tardive dyskinesia sig-
nificantly in a study by Sedman (20), but a high incidence
of side-effects including marked drowsiness and confusion
was observed. Singh et al. (26), adding all the reports of
benzodiazepine treatment of tardive dyskinesia in the lit-
erature involving 98 patients, found an average improve-
ment rate of 85 %. Nevertheless, it must be emphasized,
that most of the studies in this area were preliminary
and uncontrolled. Moreover, the possibility should be con-
sidered that beneficial effects in tardive dyskinesia
could be produced by the sedative and muscle relaxant
properties rather than by the specific antihyperkinetic
effects of these substances. After all, it is not yet
clear whether the benzodiazepines exhibit specific anti-
hyperkinetic effects. Because of their sedative properties
and their marked dependence risk they are only of limited
value for a long-term treatment of tardive dyskinesia.
Baclofen, a GABA derivative, which is thought to be a
GABA agonist with a still unknown mode of action or a
substance P antagonist, has been investigated in tardive
dyskinesia, too. It was reported to be effective in 5
studies (1, 4, 5, 11, 22), whereas one study (16) did
not note any improvement of tardive dyskinesia. Possibly,
baclofen can provoke psychoses, as suggested by Simpson
et al. (23).
The complexity of biochemical mechanisms and the
lack of specifically antihyperkinetic properties of both
the benzodiazepines and baclofen caused investigators to
focus upon GABA-ergic substances with greater antihyper-
kinetic specificity and better understood modes of action.
Among the direct GABA agonists, which exert their
action directly upon GABA, muscimol seems to have the
greatest affinity to the GABA receptors. In patients with
tardive dyskinesia it caused significant and consistent
improvement of hyperkinetic movements, but as it was
found to provoke psychoses, it is of no value in the

568
treatment of tardive dyskinesia (27). THIP, another direct
GABA receptor agonist, is less toxic than muscimol and not
psychotogenic, but was found to be ineffective in tardive
dyskinesia.
The so-called indirect GABA mimetics act via diverse
mechanisms, such as inhibition of GABA uptake or by in-
hibition of GABA metabolism. Gamma-acetylenic-GABA was
shown to reduce tardive dyskinesia significantly (2).
Sodium valproate, a compound with predominantly antiepi-
leptic properties which is also supposed to increase brain
GABA by inhibiting GABA transaminase, has produced incon-
sistent results in tardive dyskinesia treatment studies.
Linnoila et al. (14) and Casey and Hammerstad (3) described
moderate improvement, while Nair (16) and Gibson (6) were
unable to confirm this finding.
Progabide is a specific non-toxic GABA receptor ago-
nist without an appreciable affinity for other neurotrans-
mitter receptors. Thus, at the molecular level, progabide
displaces specific ligands without effect on GABA synthe-
sis, uptake or metabolism. Via its GABA receptor acti-
vation, progabide can be predicted to have antiepileptic
activity. Also via its activation on GABA receptors, it
has been shown to regulate the activity of the nigrostri-
atal dopaminergic system. In this respect progabide re-
duces the dopamine release induced by neuroleptics, blocks
the neuroleptic-induced increase in dopamine neuron firing
and changes the development of dopamine receptor super-
sensitivity. These studies indicate that progabide may
have a therapeutic potential in tardive dyskinesia. During
an open trial a positive effect of progabide was observed
in 6 out of 10 patients with tardive dyskinesia (19). In
a double-blind cross-over trial of progabide versus placeso,
in which the duration of each treatment sequence was G
weeks, 5 out of 10 patients with tardive dyskinesia showed
a significant improvement under progabide (21).
In our study the efficacy and the safety of progabide
was evaluated in comparison with placebo. The study was
carried out in a double-blind cross-over design. After a
wash-out period of one week for patients with anti-par-
kinsonian agents, patients were randomly allocated either
to the progabide or to the placebo group and were treated
over a period of five weeks. Thereafter a wash-out period
of two weeks took place. Then the patients were crossed
over to the alternative treatment for a period of five
weeks. All patients were allowed to have only one neuro-
leptic in a constant dose during the course of the study.

569
The daily dose of progabide was 40 mg/kg (900 mg first day)
and then gradually increased over a period of one week up
to the maximum dose of 45 mg/kg. Tardive dyskinesias were
evaluated using the Abnormal Involuntary Movement Scale
(AIMS) and the Tardive Dyskinesia Rating Scale of Simpson
et al. (24). The mental state was documented with the BPRS
(18). Ratings were assessed on the day 0, 21, 35, 49, 70
and 84. Plasma concentrations of progabide and its main
metabolites were measured on the days 7, 35, 56 and 84.
At present we cannot but report preliminary results,
because only 6 patients have completed the study. In the
global assessment a positive therapeutic effect was obser-
ved under progabide in compaFison to placebo. Progabide
and placebo had a similar safety, no side-effects were
observed, and there was a high degree of conformity between
the judgment of investigator and patients. The results of
the AIMS total scores are shown in table 1.
Tab. 1 . A I M S (Abnormal Involuntary r-lovement
Scale)
-
X SD SEM N

Sequence 1 Baseline 14,00 4,76 2,38 4

(Progabide 5 weeks 9,50 6,40 3,20 4
Placebo) 7 weeks 9,50 7,33 3,66 4
12 weeks 8,25 6,85 3,42 4

Sequence 2 Baseline 16,00 2,83 2,00 2

(Placebo - 5 weeks 15,00 0,00 0,00 2
Progabide) 7 weeks 15,50 0,71 0,50 2
12 weeks 10,00 4,24 3,00 2

During the treatment with progabide there was an improve-

ment of the dyskinesias which remained constant under the
following application of placebo. In the group of patients
with the placebo - progabide sequence an improvement of
the abnormal movements was also observed under progabide.
Using the Simpson scale we found a similar improvement of
tardive dyskinesia scores under the treatment with pro-
gabide. During the treatment with progabide an improve-
ment was confirmed in the BPRS total scores. After crossing
over to placebo this improvement of mental state was not
observed.
It must be emphasized that at present only 6 patients
have completed the study. Regarding this small number of

570
patients it is difficult to discuss the results. We have
the impression that progabide may have a positive thera-
peutic effect in tardive dyskinesia. Possibly progabide
may have a triggering therapeutic effect in tardive dys-
kinesia because the improvement of the abnormal movements
remained unchanged switching from progabide to placebo.
The improvement of psychopathology rated on the BPRS may
be caused by a presumed ataractic effect of progabide.
Preliminary data suggest that progabide may also have a
slight anti-depressant effect. No side-effects were seen
under the treatment with progabide. These results encour-
age us to continue the study in order to gather further
experiences with progabide.
REFERENCES
1. Baldessarini, R.J., Tarsy, D., 1978, Tardive dyskinesia,
in: Psychopharmacology: A Generation in Progress,
Lipton, M.A., DiMascio, A., Killam, K.R., eds, Raven
Press, New York.
2. Casey, D.E., Gerlach, J., Magelund, G., Christensen, R.,
1980, Gamma-acetylenic-GABA in tardive dyskinesia,
Arch. Gen. Psvchiat., 37: 1376.
3. Casey, D.E., Hammerstad, J.P., 1979, Sodium valproate
in tardive dyskinesia, J. Clin. Psvchiat. 40: 483.
4. Frangos, E.A., Athanasenas, E., 1980, Lioresal in the
treatment of neuroleptic-induced tardive dyskinesia,
12-th CINP Congress Supplement to Progress in Neuro-
psychopharmacology, Pergamon Press.
5. Gerlach, J.: 1977, The relationship between parkinsonism
and tardive dyskinesia, Am. J. Psvchiat., 134: 781.
6. Gibson, A.C.,1978, Sodium valproate and tardive dyskin-
esia, Brit. J. Psvchiat. 133: 82.
7. Godwin-Austen, R.B., Clar·k, T., 1971, Persistent pheno-
thiazine dyskinesia treated with tetrabenazine, Brit.
MPn. .r. , 4: 25.
8. Itil, T~M., Unverdi, C., Mehta, D., 1974, Clorazepate
dipotassium in tardive dyskinesia, Am. J. Psvchiat..
131 : 1291 .
9. Jeste, D.V., Wyatt, R.J., 1982, Therapeutic strategies
against tardive dyskinesia, Arch. Gen. Psvchiat.,
39: 803. .
10.Jus, K., Jus, A., Gautier, J., Villeneuve, A., Pires,
P., Pineau, P., Villeneuve, R., 1974, Studies on the
action of certain pharmacological agents on tardive
dyskinesia and on the rabbit syndrome, Int. J.Phar-
macol. 9: 138.
11.Korsgaard, S., 1976, Baclofen (lioresal) in the treat-
ment of tardive dyskinesia, Acta Psvchiat. Scand.
54: 17.

571
 12. Korsgaard, S., Casey, D.E., Gerlach, J., 1981, Neuro-
peptides and GABA influences in tardive dyskinesia,
III. World Congress of Biological Psychiatry Stock-
holm.
13. Kruse, W., 1960, Persistent muscular restlessness after
phenothiazine treatment, Am. J. PsvchiAtrv 117: 112.
14. Linnoila, M., Viukari, V., Hietala, 0., 1976, Effects
of sodium valproate on tardive dyskinesia. Brit. J.
Psvchiat. 119: 114.
15. Mehta, D., Mehta, S., Matthew, P., 1976, Failure of
deanol in treating tardive dyskinesia, Am. J.Psvchiat.
133 : 1467.
16. Nair, N.P.V., Yassa, R., Ruiz-Navarro, J., Schwartz,
G., 1978, Baclofen in the treatment of tardive dys-
kinesia, Am. J. Psvchiat., 135: 1562.
17. O'Flanagan, P.N., 1975, Clonazepam in the treatment
of drug-induced dyskinesia, Brit. Med. J. 1: 269.
18. Overall, J.E., Gorham, D.R., 19b2, The Brief Psych-
iatric Rating Scale, Psvchol. Rep. 10: 799.
19. Rondot, P., Henry, J.F., 1952, ~tude preliminaire en
ouvertn du SL 76002 dans differentes affections neu-
rologiques, LERS Internal Report.
20. Sedman, G., 1976, Clonazepam in the treatment of tar-
dive oral dyskinesia. Brit. Med. ~· 2: 583.
21. Sevestre, P., Coloni, F., Lagrange, J., Seguier, N.,
1982, Effects neuropsychiatriques du progabide sur
les schizophrenies gravement deficitaires compli-
quees de dyskinesies tardives des neuroleptiques,
13-th CINP Congress Jerusalem.
22. Simpson, G.M., Branchey, M.H., Shrivastava, R.K.,
1978, Baclofen in schizophrenia. Lm1c~t I: ~66.
23. Simpson, G.M., Lee, J.H., Shrivastava, R.K., Branchey,
M.H., 1976, Baclofen in the treatment of tardive
dyskinesia and schizophrenia, Psvchonharmacol. Bull.
14: 16.
24. Simpson, G.M., Lee, J.H., Zoubak, B., Gardos, G.,
1979, A rating scale for tardive dyskinesia. Psvchq-
pharmacoloe:v. 64: 171 •
25. Singh, M.M., 1976, Diazepam in the treatment of tar-
dive dyskinesia, Int. Pharmacopsycb:lc:ti:_. 11: 232.
26. Singh, r'l.H., Becker, R.E., Pitman, !t.K., Nasrallah,
H.E., Lal, H., Dufresne, R.L., Weber, S.S., McCal-
ley-Whitters, M., 1982, Diazepam-induced changes
in tardive dyskinesia, ~iol. Psvchiat., 17: 729.
27. Tamminga, C.A., Crayton, J.W., Chase, T.N., 1979,
Improvement in tardive dyskinesia after muscimol
therapy, Arch. Gen. Psvchiat., 36: 595.

572
HISTORICAL REMARKS ON THE DISCOVERY OF ECT

Lothar B. Kalinowsky

New York Medical College

155 East 76th Street
New York, New York 10021

It was a strange coincidence that the first three treatments

which had a therapeutic impact on the functional psychoses, were
discovered within three subsequent years: 1933 Insulin Coma Treat-
ment, 1934 Convulsive Therapy, 1935 Psychosurgery. The only pre-
ceding successful treatment of a mental disorder had been Malaria
treatment for general paresis which, however, had been directed
only against the spirochetic invasion of the brain. When Wagner-von
Jauregg published his first report here in Vienna and later
received the Nobel Prize for it, this changed the nihilistic atti-
tude in psychiatry, and it was perhaps not surprising that Sakel's
Insulin Coma Treatment, although he had conceived the idea in
Berlin, also found its first clinical application here in Vienna.

If today Convulsive Therapy is the most widely used of the

first three treatments--all of which were introduced for schizo-
phrenia--it should not be forgotten that Insulin Coma Treatment in
1933 as well as Lobotomy introduced by Moniz in 1935, had unques-
tionable therapeutic results not seen in any previous therapeutic
attempts.

The first Convulsive Treatment was given by the Hungarian von

Meduna on January 23, 1934, and it was only four years later, in
1938, that Cerletti and Bini in Rome changed the pharmacological
induction of convulsions as used by Meduna, to the electric induc-
tion of seizures. Improvement of mental illness after spontaneous
seizures had been observed for a long time. The oldest report I
could find in the London Medical Journal of 1785 was given by a
Dr. W. Oliver who gave camphor to an insane because at that time
camphor was tried in many different medical conditions. Over the
years it had been frequently observed that patients in mental

573
hospitals who suffered spontaneous convulsions for one reason or
another, would show a sudden improvement of their mental condition.
In particular, catatonic patients who had been completely withdrawn
for years, would become communicative at least for a limited period
of time after one or two spontaneous seizures due to various causes
such as trauma or, most often, due to sudden discontinuation of hyp-
notics. Such clinical observations led to the thought that there
might be a certain antagonism between epilepsy and schizophrenia.

Meduna treated his first patient with camphor in oil which

turned out to be an unreliable and for the patient uncomfortable way
of inducing seizures. He, therefore, replaced camphor in oil with
the drug pentylenetetrazol which carries the brand name Cardiazol in
Europe and Metrazol in the United States. Injected intravenously
this led to generalized seizures within seconds but was accompanied
by an unpleasant praecordial pain before the patient lost conscious-
ness. In spite of some inconveniences of Meduna's treatment, it
should not be forgotten that the great impact that convulsive
therapy had on the treatment of schizophrenia throughout the world,
came about with this pharmacological convulsive treatment. Also, it
should be realized that for the first four years convulsive therapy
was applied exclusively to patients suffering from schizophrenia
which today is sometimes wrongly questioned as an indication for
convulsive therapy. In 1937, three years after the introduction of
convulsive therapy and four years after the initiation of Insulin
coma treatment an international meeting was convened by the Swiss
psychiatrist Max Muller in Munsingen, Switzerland. It was there
that Sake! claimed priority for the use of convulsions in the treat-
ment of schizophrenia because insulin hypoglycemia was often accom-
panied by convulsions. However, in his first publication Sake! had
considered convulsions a complication in insulin coma treatment and
recommended barbiturates against their occurrence. Therefore it
is unquestionable that Meduna was the first to recognize the value
of convulsive therapy.

It was also in Munsingen that Bini reported experiments on the

electric induction of seizures in animals which shortly after led to
the application of the first electric convulsive treatment by
Cerletti and Bini at the University Hospital in Rome. Cerletti had
worked with experimental epilepsy for many years, and when convul-
sive therapy was introduced, he immediately asked why Meduna had not
used the much simpler way of inducing seizures with electricity
rather than with pharmaca. Some misconceptions about the events
leading to Cerletti and Bini's discovery should be clarified.
Having been myself an associate of Cerletti at that time, I can
confirm an eye-witness report by Accornero published in 1970. It
was Bini who realized that the high percentage of animals dying in
experimental epilepsy, was due to the application of the electrodes
in mouth and anus whereby the current passed through the heart.
Applying both electrodes through the skull, he eliminated this

574
danger. It is an error that Cerletti got the idea of ECT from the
slaughterhouses in Rome. When he heard that the slaughterhouses
used electric shocks, they delayed the first treatment. Going
there, they found out that the current was only used to make the
animals unconscious before the actual killing.

At the first treatment given in April 1938 the patient received

a minimal voltage which did not make him unconscious. A subsequent
second stimulus led to unconsciousness but no convulsion. A third
stimulus elicited a full seizure after which the patient, a mute
schizophrenic, talked for the first time. He had several treat-
ments and a follow-up of two years showed him in good remission.

Like Meduna, also Cerletti introduced his method as a treat-

ment for schizophrenia. Only four years later it was noticed that
it was even more effective in depressions. It was in schizophrenia
that electric shock became a worldwide recognized treatment. The
best testimony for its effectiveness was that it was widely used in
spite of the complication of fractures. The elimination of this
complication is also of great historical interest. The American
psychiatrist A. E. Bennett who knew of the use of Curare by South
American hunters who wanted to catch animals alive, introduced
Curare to paralyze the muscles of patients before he subjected them
to seizures. Curare was discarded because fatalities occurred with
its use. It was Bovet who introduced succinyl choline to paralyze
the muscles prior to ECT. Since this also paralyzed the respira-
tory muscles and gave the patient respiratory distress, barbiturate
anesthesia was given before the injection of succinyl choline.

This historical survey should include a few remarks about

research. There was considerable research activity in the hope to
understand the mode of action in ECT. This was largely unproductive.
Only when basic science research in psychiatry was stimulated by the
introduction of psychopharmaca, research was again taken up in ECT.
Also, the Anti-Psychiatry movement which so much hampered the use of
ECT--in some countries more than in others--increased both, clinical
and basic science research. Clinical reports almost invariably prove
the efficacy of the treatment. Basic science research is promoted
in the hope that a better understanding of ECT may lead to simpler
new treatments in the future.

REFERENCES

Accornero, F. Testimonianza oculare sulla scoperta dell'Elettro-

shock. Pagine di Storia della Medicina 14:38, 1970.
Bennett, A. E. Preventing traumatic complications in convulsive
shock therapy by curare. JAMA 141:322, 1940.
Bini, L. Experimental researches on epileptic attacks induced by the
electric current. Am.J.Psychiatry 94(Supp) :172, 1938.

575
Cerletti, U., Bini, L. L'Elettroshock. Arch gen di neural psichiat,
e psichoanal 19:266, 1938.
Meduna, L. J. Die Konvulsionstherapie der Schizophrenia, Halle,
Carl Marhold, 1937.
Moniz, E. Tentatives operatoires dans le traitement de certaines
psychoses. Paris, Masson et Cie, 1936.
Muller, M. Insulin und Cardiazolschockbehandlung der Schizophrenia.
Fortschr Neural Psychiatr 9:131, 1937.
Sakel, M. The nature and origin of the hypoglycemic treatment of
psychoses. Am.J.Psychiatry 94(Supp) :24, 1938.
Wagner von Jauregg, J. Die Einwirkung der Malaria auf die progres-
sive Paralyse. Psychiat Neural Wochenschr 20:132, 1918.

576
ECT, ACETYLCHOLINE RECEPTORS, AND MEMORY

Bernard Lererl,2, Michael Stanley!, and Harvey Altmanl

Inept. of Psychiatry and Pharmacology

Wayne State University and Lafayette Clinic
Detroit, Michigan

2Jerusalem Mental Health Center

Jerusalem, Israel

Although almost fifty years have elapsed since its introduction,

the basic mechanisms whereby electroconvulsive therapy (ECT) exerts
its therapeutic effects, remain to be elucidated. The biological
basis of ECT-induced memory impairment, the major adverse effect of
anesthetic- and muscle relaxant-modified ECT, is also unknown.

Recent research on basic mechanisms of ECT has focused on the

effects of repeated electroconvulsive shock (ECS) on a number of
neurotransmitter-receptor systems in rodent brain. Effects of ECS
on adrenergic, serotonergic, and dopaminergic systems have been
widely studied (Grahame-Smith et al. 1978; Lerer and Belmaker,
1982; Lerer et al. 1983). ECS effects on central cholinergic
systems have been less extensively investigated in spite of evi-
dence linking depressive states and cholinergic supersensitivity
(Sitaram and Gillin, 1980) and a growing body of data regarding the
role of central cholinergic function in normal and impaired memory
(Sitaram et al. 1978; Davis et al. 1978; Bartus et al. 1982). ECT
is an effective antidepressant treatment (Kendell, 1981) but also
has deleterious effects on memory (anterograde as well as retro-
grade) (Squire, 1977). It is thus logical that the effects of ECS
on central cholinergic mechanisms be further studied.

ECS and Presynaptic Choliner£ic Mechanisms

Longoni et al. (1976) reported a significant decrease in

acetylcholine content in rat cerebral cortex immediately after

577
electrically induced seizures. Ictal and immediate post-ictal
decreases in ACh content have also been reported by Richter and
Crossland (1949), and Takahashi et al. (1961). Single ECS also
appears to increase choline acetyltransferase (ChAT) activity in
the cortex (Longoni et al. 1976). These effects of ECS may be
compatible with increased release of acetylcholine during the
seizure. This possibility is supported by human data showing in-
creased CSF acetylcholine levels following pentylenetetrazol-
induced or epileptic seizures and evidence for cholinergically-
related EEG slowing following ECT (Fink, 1966).

ECS and Cholinergic Receptors

Dashieff et al. (1982) studied the effect of 4 ECS daily over

4 days on [3H] QNB-labelled muscarinic cholinergic receptors in the
dentate and hippocampal gyri of rat brain. This ECS regimen induced
a highly significant 19-25% decrease in [3H] QNB binding in the above
hippocampal areas. Scatchard analysis of binding data showed the
decline in [3H] QNB binding to be due to a reduction in muscarinic
binding sites (B max) with no change in the affinity of the receptor
for the ligand (Kd).

Lerer et al. (1983) studied the effect of 7 consecutive daily

ECS (130 volts for 0.75 sec administered via earclip electrodes) on
[3H] QNB binding in rat cerebral cortex and hippocampus. The animals
were sacrificed 24 hours after the last ECS and the brains rapidly
dissected and frozen at -70°C until assay. All non-ECS treated
animals had earclip electrodes applied without current being passed
(Sham ECS). Binding of [3H] QNB was determined in accordance with
the methods described by Wastek and Yamamura (1978) at 25 pM
[3H] QNB concentration. Repeated ECS was found to induce a 15%
decrease in [3H] QNB binding in rat cerebral cortex and a 13% de-
crease in the hippocampus, both of which were statistically signi-
ficant. Animals sacrificed 24 hours after a single ECS did not
show a comparable reduction in cortical of hippocampal muscarinic
cholinergic receptor binding sites.

These findings confirm and extend those of Dashieff et al.

(1982), and point to an effect of repeated ECS to reduce muscarinic
cholinergic receptor binding sites in the brain areas studied.
This effect is further supported by the finding that concurrent
daily ECS blocked the significant increase in cortical [3H] QNB
binding induced by chronic administration of large doses of atropine
(10 mg per kg i.p. over 5 days) (Lerer et al. 1983). Previous
reports had found a less prominent effect of repeated ECS on corti-
cal and hippocampal [3H] QNB binding (Keller et al. 1981; Deakin
et al. 1981). Methodological factors, particularly in terms of
the concentrations of [3H] QNB used to measure binding, may explain
the differences.

578
Relevance to Antidepressant Mechanism of ECT

Reduction of muscarinic cholinergic receptor density by chronic

ECS may be functionally related to the antidepressant efficacy of
ECT. This possibility may be considered in the light of theories
linking depression to cholinergic predominance (Janowsky et al.
1972) as well as more recent findings of "supersensitivity" of
cholinergically-mediated REM sleep induction (Sitaram and Gillin,
1980) in depressed patients. Down-regulation of muscarinic choli-
nergic receptors by ECT may represent a correction of the hypothe-
sized cholinergic supersensitivity associated with depression.

Relevance to Amnestic Effects of ECT

Recent advances in delineating the role of cholinergic mechan-

isms in normal and imptired memory, suggest that a possible role
for central cholinergi1 dysfunction in ECT-induced amnesia, be
considered. A link betveen brain cholinergic activity and memory
is supported by studies in animals and man which show that centrally
acting anticholinergic csents impair memory functions while choli-
nergic precursors, choli tergic agonists or acetylcholine esterase
inhibitors enhance them 1Deutch and Rogers, 1979; Sitaram et al.
1978; Davis et al. 1978), Post mortem findings have strongly
linked the cognitive defi, its of Alzheimer's disease to degeneration
of cholinergic neurones h the nucleus basalis of Meynert and
depletion of choline acety .transferase activity and acetylcholine
levels in the cortex (Bartts et al. 1982). ECT-induced memory
impairment may result from a reversible down-regulation of musca-
rinic cholinergic receptors in the cerebral cortex and hippocampus
induced by repeated seizures. This muscarinic cholinergic subsen-
sitivity may imply a functional reduction in neurotransmission
through the affected cholinergic synapses.

An hypothesis linking ECT-induced amnesia to cholinergic re-

ceptor subsensitivity in cerebral cortex and hippocampus, is com-
patible with the transient nature of most of the ECT-induced memory
deficits. ECT-induced memory impairment is characterized by an
anterograde amnesia which is no longer present by 4-6 weeks after
the last treatment (Squire, 1977; Weeks et al. 1980) and by retro-
grade deficits which are either fully recovered or minimally
demonstrable after six months (Squire et al. 1981). A gradual
restoration of receptor number after cessation of the treatments
could account for the improvement in memory function which occurs.
Unlike Alzheimer's disease, where cholinergic neurotransmission is
thought to be permanently reduced because of degeneration of pre-
synaptic cholinergic neurones, restoration of function could be
expected to occur in patients with ECT-induced amnesia following
cessation of the treatments.

A recent study from our laboratory provides preliminary support

579
for this hypothesis. Groups of ECS and sham-treated rats underwent
passive-avoidance training 24 hours following the last of a series
of daily ECS for 7 days (training parameters: 0.5 mA for 1 sec.
inescapable shock). Retention of the original avoidance habit was
measured 7 days later by returning the animals to the lighted side
of the shuttle-box; after a brief accomodation period (10 sec.)
the door separating the lighted from the dark chamber was opened
and the animal's latency to cross (step-through latency) was re-
corded. Parallel groups of ECS- and sham-treated animals were
sacrificed 24 hours after the last treatment and [3H] QNB binding
was assayed in cortical homogenates as described above. Results of
the passive-avoidance testing and [3H] QNB binding studies are
presented in Table 1.

Highly significant impairment of delayed recall was demon-

strable in the ECS-treated group. Such impairment was not present
in rats trained 24 hours following a single ECS and tested 7 days
later. A significant reduction in cortical muscarinic cholinergic
receptor number (Bmax) is evident in the group of rats sacrificed
24 hours following the last ECS. The time-course of muscarinic
cholinergic receptor subsensitivity and memory impairment following
repeated ECS are thus parallel. Twenty four hours following the
last of a series of 7 ECS over 7 days, retention in a passive avoid-
ance training situation is impaired and muscarinic receptor subsen-

Table 1. Effect of Repeated ECS on Delayed Recall of an Aversive

Stimulus and on Muscarinic Cholinergic Receptor Number
in Rat Cerebral Cortex.

Step-Through Latency (Sec) Muscarinic Cholinergic Receptors

B Max Kd

Control 400.21 ± 77.06 885 ± 25.64 15.2 ± 1.55

(10) (9)

ECS 12.9 ± 3.47* 800 ± 26.49** 13.8 ± 1.54

(7) (7)

Rats were trained 24 hours following the last of a series of daily

ECS over 7 days and tested 7 days following training. A parallel
group of rats was sacrificed 24 hours following the last ECS for
[ 3H.] QNB binding studies. Number of animals given in parentheses.

* p < .001 (Student's t-test), ECS vs. control

** p < .05 (Student's t-test), ECS vs. control

580
sitivity is demonstrable. Neither were present in rats sacrificed
or trained 24 hours following a single ECS. Further experiments
are underway to evaluate the effect of cholinergic agonists and
antagonists on the ECS-induced memory deficit. Parallel studies
may be conducted in ECT-treated patients in whom antagonists would
be expected to further impair performance whereas agonists should
improve it.

Conclusions

Effects of electroconvulsive shock on acetylcholine receptors

represent a potentially fruitful avenue for elucidating the basic
mechanisms of ECT. The findings reported here suggest a link
between the effects of ECT on brain cholinergic systems and ECT-
induced memory deficits. A possible relationship to the therapeutic
mechanism of ECT should be studied in parallel. In addition to
further preclinical experiments, studies in humans are clearly
needed in order to evaluate whether these animal findings are
demonstrable in ECT-treated patients and to assess their clinical
relevance.

REFERENCES

Bartus, R.T., Dean, R.L., Beer, B., and Lippa, A.S., 1982, The
cholinergic hypothesis of geriatric memory dysfunction, Science
217:407-417.
Dashieff, R.M., Savage, D.D. and McNamara, J.O., 1982, Seizures
down-regulate muscarinic cholinergic receptors in hippocampal
formation, Brain Res., 235:327-334.
Davis, K.L., Mohs, R.C., Tinkelberg, J.R., Pfefferbaum, A.,
Hollister, L.E. and Koppel, B.S., 1978, Physostigmine: Im-
provement of long-term memory processes in normal humans,
Science, 201:272-274.
Deakin, J.F.W., Owen, F., Cross, A.J. and Dashwood, M.J., 1981,
Studies on possible mechanisms of action of electroconvulsive
therapy: Effects of repeated electrically induced seizurs on
rat brain receptors for monoamines and other neurotransmitters.
Psvchovharmacology 73:345-349.
Deutch, J.A. and Rogers, J.B., 1979, Cholinergic excitability and
memory: Animal studies and their clinical implications, in:
"Brain Acetylcholine and Neuropsychiatric Disease," K.L.
Davis and P.A. Berger, eds., Plenum Press, New York.
Fink, M.D., 1966, Cholinergic aspects of convulsive therapy. J.
Nerv. Ment. Dis., 142:475-484.
Grahame-Smith, D.G., Green, A.R., and Costain, D.W., 1978, Mechanism
of the antidepressant action of electroconvulsive therapy,
Lancet 1:245-256.
Janowsky, D.S., El-Yousef, M.K., Davis, J.M., and Sekerk, H.J.,
1972, A cholinergic-adrenergic hypothesis of mania and depres-
sion. Lancet 632-635.

581
Kellar, K.J., Cascio, C.S., Bergstrom, D.A., Butler, J.A. and
Iadarola, P., 1981, Electroconvulsive shock and reserpine:
Effects on beta-adrenergic receptors in rat brain, J. Neuro-
chem. 37:830-836. ·
Kendell, R.E., 1981, The present status of electroconvulsive ther-
apy. Brit J. Psvchiat. 189:265-283.
Lerer, B. and Belmaker, R.H., 1982, Receptors and the mechanism of
action of ECT. Bioi. Psychiat. 17:497-511.
Lerer, B., Weiner, R.D. and Belmaker, R.H. (eds.), 1983, "ECT:
Basic Mechanisms," John Libbey, London.
Lerer, B., Stanley, M., Demetriou, S., and Gershon, S., 1983, Effect
of electroconvulsive shock on 3H-QNB binding in rat cerebral
cortex and hippocampus, Journal of Neurochemistry, in press.
Longoni, R., Mulas, A., Novak, B.o., Pepeu, I.M. and Pepeu, G.,
1976, Effect of single and repeated electroshock applications
on brain acetylcholine levels and choline acetyltransferase
activity in the rat. ~europharmacology, 15:283-286.
Richter, D., and Crossland, J., 1Y49, Variation in acetylcholine
content of the brain with physiological data, Am. J. Phvsiol.
159:247-255.
Sitaram, N., Weingartner, H., and Gillin, J.C., 1978, Human serial
learning: Enhancement with arecoline and choline and impair-
ment with scopolamine. Science, 201:274-276.
Sitaram, N. and Gillin, J.C., 1980, Development and use of pharma-
cological probes of the CNS in man: Evidence of cholinergic
abnormality in primary affective illness. Biol. Psychiat.
15:925-955.
Squire, L.R., 1977, ECT and memory loss. Am. J. Psychiat. 134:997-
1005.
Squirek L.R., Slater, P.C. and Miller, P.L., 1981, Retrograde
amnesia and bilateral electroconvulsive therapy: Long-term
follow-up. Arch. Gen. Psvchiat. 38:89-95.
Takahashi, R., Nasu, T., Tamura, T., and Koriya, T., 1961, Relation-
ship of ammonia and acetylcholine levels to brain excitability.
J. Neurochem. 7:103-112.
Wastek, G.J. and Yamamura, H.I., 1978, Biochemical characterization
of the muscarinic cholinergic receptor in human brain:
Alterations in Huntington's Disease. Mol. Pharmacol.
Weeks, D., Freeman, C.P.L., and Kendell, R.E., 1980, ECT-fii:
Enduring cognitive deficits? Brit. J. Psychiat. 137:26-37.

582
MEMORY DISTURBANCE AFTER ECT IN LOW-PRESSURE NARCOSIS

Jan-Otto Ottosson and Kurt Widepalm

Department of Psychiatry
Sahlgren Hospital
Gothenburg, Sweden

INTRODUCTION

A correlation has been shown between the maximal systolic

blood pressure during ECT and the memory disturbance as measured
by Benton's visual retention test (Hamilton et al. 1979). The
purpose of this study was to examine if the memory disturbance
after ECT can be reduced by preventing an increase of the blood
pressure during the treatment.

MATERIAL AND METHODS

General Design

The design was an intraindividual cross-over comparison in

connection with the second and third treatments in the ECT series.
The first treatment was not utilized because it induces a signi-
ficantly longer seizure activity than the succeeding ones. One
half of the patients started with low-pressure narcosis while the
order was reversed in the other half, the order being randomized
(Fig. 1).

ECT Technique

The treatments were given in methohexital (Brevital) narcosis

and complete muscular relaxation with suxamethonium (Celocurin).
The blood pressure elevation was prevented with the ganglionic
blocking drug trimethaphan (Arfonad) wich was given for 3 minutes
after the methohexital injection (Fig. 2). Oxygen was administered
during the whole treatment. The seizure activity was recorded with
EEG with two electrodes on each half of the head. Cerebral

583
 Memory test Memory test
Immediate Delayed
Treatment reproduction ECT+trimethaphan reproduction
no. 2

1 h 2 h

Treatment ECT+saline
no. 3

Fig. 1. Cross over Design

El. stim. Re-

No. 2
~----+-~--------~--~~r---~~~---------·
Seizure!\l covery
/\~·~---­
Metho- Trimethaphan Metho- Suxa-
hexi- hexi- meta-
tal tal nium

No. 3 Saline t
~--+--------+----~--~~~----~~-+---
Fig. 2. ECT Procedure

Table 1. Patients

M 5
Sex
F 13

X 40
Age
Range 32-56

584
dominance was ascertained (d'Elia 1970) and the electrical stimul-
ation was applied unilaterally over the non-dominant hemisphere.
A Konvulsator 622 machine was used deliviering 5 msec unidirection-
al pulses. Threshold stimulations were given.

Memory Tests

The patients were given parallel forms of four memory tests

each with 30 items (d'Elia et al. 1977) in a retrograde design.
The tests were presented one hour before ECT and immediate memory
was recorded. Two hours after ECT the delayed memory was recorded.
The difference between immediate and delayed memory was used as a
measure of retrograde amnesia (Fig. 1).

Patients

18 testable depressive in-patients referred to ECT were studi-

ed. They were below 60 years of age and had no history of cardiac
or pulmonary diseases, and normal ECG. They had not received ECT
during the last six months or been tested with the test battery.
There were 13 women and 5 men with mean age 40 years, range 32-56
years (Table 1). The doses of drugs were constant during the study.
Bensodiazepines or barbiturates were not given.

Statistical Method

t-test for paired observations was used.

RESULTS

There were no differences in the amount of methohexital and

suxamethonium or time of electrical stimulation between the two
treatments. The blood pressure elevation which is regularly seen
during conventional treatment was prevented (Fig. 3). The seizure
duration was shortened to about 80% (Table 2). The retrograde
amnesia was similar in all of the four memory tests (Table 3)

DISCUSSION

There are good reasons to hypothesize that the increase of

the blood pressure as well as the increased cerebral blood flow
contribute to the memory disturbance after ECT. As shown by
Bolwig et al. (1977) the permeability of the blood brain barrier
increases during ECT which leads to extravasation of small mole-
cules. However, the seizure activity in ECT is not a necessary
condition since the same effect can be obtained by co 2 breathing.
There is a correlation between the blood pressure increase and a
low low performance in Benton's test. However, the present findings
give no support to the hypothesis that the blood pressure increase

585
 170

160

150

140

130 Systolic

120

llO

100

90
Diastolic

80

70 Narcosis Trimethaphan ECT 3 min. after

(Saline)
Fig. 3

586
Table 2. Seizure Duration

Without With Difference

trimethaphan

X 67 53 14.0 + 5.3
t 2.63
Range 26-94 29-101 p < 0.02

Table 3. Retrograde Amnesia

Memory Test Without With Difference

trimethaphan

30 Word pairs 8.0 6.8 1.2 + 1.3

30 Objects 4.3 3.4 0.9 + 0.9
30 Faces 5.3 8.1 -2.7 -+ 2.5
30 Geom.fig. 5.6 5.5 0.1 + 1.8

587
is crucial for the memory disturbance. Instead the blood pressure
together with the increased cerebral blood flow may be regarded as
compensatory mechanisms, by which the increased metabolic demands
of the brain are met during the treatment. The present findings
agree with the view that the seizure activity is of less importance
than direct effects of the electrical stimulation for the memory
disturbance after ECT (Ottosson 1960). The fact that the seizure
duration decreased to a submaximal level may imply a diminished
antidepressive efficiency.

CONCLUSION

ECT where the blood pressure increase was blocked with the
ganglionic blocking substance trimethaphan did not diminish the
memory disturbance. The elevated blood pressure during the treat-
ment is not a major causal factor for the memory disturbance.

REFERENCES
Bolwig, T.G., Hertz, M.M., Paulson, O.B., Spotoft, H., and
Rafaelsen, O.J., 1977, The permeability of the blood-brain
barrier during electrically induced seizures in man, Europ.
J. Clin. Invest., 7:87. .
d'Elia, G., 1970, Unilateral electroconvulsive therapy, Acta
Psvchiatr. Scand., Suppl. 215.
Hamilton, M., Stocker, M.J., and Spencer, C.M., 1979, Post-ECT
cognitive defect and elevation of blood pressure, Brit. J.
Psychiat., 135:77.
Ottosson, J.-0., 1960, Experimental studies of memory impairment
after electroconvulsive therapy, Acta Psychiatr. Scand.,
35:Suppl. 145:103.

588
THE INFLUENCE OF ECT ON REGIONAL CEREBRAL BLOOD FLOW (rCBF)

Tom G. Bolwig, Ralf Hemmingsen and Sissel Vorstrup

Department of Psychiatry, Rigshospitalet, Copenhagen

Department of Psychiatry, Hiller0d Hospital
Department of Neurology, Rigshospitalet, Copenhagen
Denmark

The question of the relative influence of the two cerebral hemi-

spheres on the development of depression and their function following
a course of ECT is the objective of the reported investigation.

Using a single-photon emission tomography technique (Tomomatic)

and EEG a study of patients undergoing ECT is started.

The single-photon emission tomography technique in which 133 xe-

non inhalation is used allows a 3-dimensional measurement of CBF,
which means that a study of subcortical structures can be made. Using
an 8-channel EEG the cortical electrophysiological activity of the
brain was measured.

In the following are presented the preliminary results of the

study of EEG and rCBF in two patients one week after the last of a
series of 12 ECT using mainly bilateral electrode application with
a MECTA-apparatus. The patients were right-handed and were diagnosed
as endogenously depressed.

The EEG findings showed a normal distribution of frequencies

over both hemispheres with no sign of asymmetry.
rCBF showed slight abnormalities as shown in Figure 1 and 2.

589
 Figure 1

These pictures show three slides 2 em thick. Slice 1 in which

the Xenon exhalation is seen as black is only used for determination
of cerebellum which is seen occipitally. Values given below are
estimated from Slide 2 in both patients. The pictures show that in
the first patient the mean cerebral blood flow was clearly decreased
being 43 ml/lOOg/min as compared to a large normal material showing
mean values of 56 ml/lOOg/min. In this patient a comparison of
symmetrical areas of the flow landscape in the two hemispheres shows
firstly a slight decrease in the left parieto-occipital region where

Figure 2

590
the values between left and right were 34 and 37 ml/lOOg/min,
respectively.
In the right temporal lobe of this patient CBF was 48 versus
left hemispheric CBF of 45 ml/lOOg/min. These differences are close
to being statistically significantly different.
In the second patient the mean hemispheric flow was 53 ml/lOOg/
min which is close to normal. The region of interest in this flow
landscape is the right parietal cortical region. When calculation
for the pixes in the right and left symmetrical side was made it was
seen that left-right values were 51/47 ml/lOOg/min, respectively.
No firm conclusions can be drawn from these preliminary
investigations, but it is remarkable that there seems to be regional
differences in a situation of normal EEG one week following the last
ECT.

Both patients were treated for endogenous depression, both were

clinically in a neutral mood, none of them had subjective or clinical-
ly observable cognitive disturbances, and both were discharged from
the department.

REFERENCE
Stokeley, E. M., Sveinsdottir, E., Lassen, N. A., and Rommer, P,
1980, Single-photon dynamic computer-assisted tomograph
(DECAl) for imaging brain function in multiple cross sections,
J Comput Assist Tomogr, 4: 230.

591
INFLUENCE OF RHEOLOGICAL THERAPY ON REGIONAL CEREBRAL
BLOOD FLOW IN PATIENTS WITH CEREBROVASCULAR DISEASE

Alexander Hartmann
Neurologische Univ.-Klinik Bonn
Sigmund-Freud-Str. 25, 5300 Bonn, FRG

Vascular diseases of the brain results in different neu-

rologic and psychiatric symptoms depending on the loca-
tion of blood flow reduction and on character of forma-
tion. Nowadays it is possible to measure the degree of
ischemia in the brain with different techniques:
Whereas Positron Emission Tomography allows estimation
of regional cerebral blood flow and cerebral metabolism
its application is limited to certain centers due the
considerable costs and enormous technical suppositions.
The stable Xenon techniaue - which uses the "cold"
anesthetic Xenon gas - permits measurement of local
blood flow in a three-dimensional way but is not appli-
cable to out-patients, since it possesses anesthetic
effects (depending on the concentration used). The
K~t;y-Schmidt-technique, with inhalation of nitrous oxide

measures blood flow and metabolism of the total brain

or one hemisphere respectively but does not produce re-
gional data. Focal defects are overlooked. The non-
traumatic Xenon 133-method allows repeated measurements
of regional cerebral blood flow (rCBF) of both hemi-

593
spheres. Since the method does not request any inva-
sive procedure it can be used repeatedly on an outpa-
tient basis. The technique involves the application of
small amounts of Xenon 133, which are inhaled over a
period of 60 sec, followed by 10 min of desaturation
(with inhalation of room air). During this period 32
detectors, placed over both aspects of the skull, re-
cord the activity from brain, bone,muscle etc. of the
head. The correction of these raw curves results in
calculation of flow data of the brain tissue itself.
Different modes of calculation permit the estimation
of several flow data, which represent the tissue per-
fusion of either the grey matter alone (F1, grey mat-
ter flow) or a mixture of both the grey and the white
matter (initial slope index - ISI - and mean flow-MF).

The advantage of this technique for measuring repeated-

ly blood flow of the brain in healthy and sick patients
can be used only if the actual CBF-course of an indivi-
dual patient or a patient group takes the following
observations into account:
1. In the middle age groups (up to the age of 60) rCBF
remains rather stable throughout the life (Fig. 1).
There is only a small decline from age group to age
group. However, thereafter rCBF declines significantly
for all modes of calculation. Mean rCBF of an indivi-
dual patient can be considered only as being patholo-
gic i f the actual flow value leaves the limit of the
normal age group. This is also true for all regional
data: the normal flow distribution over one hemisphere
with higher flow over the frontal areas and lower flow
over the parietal and temporal parts is preserved in
all age groups with high stability between correspon-

594
Fig. 1: Mean rCBF (mean of both hemispheres) in normal
volunteers. No history of cerebral vas cular disease or
diseases with increased intracranial pressure. Blood
flow- expressed as F1, ISI, MF (see text) -remains
rather stable between the age from 28 - 58. Thereafter
rCBF decreases for all modes of calculation.

ding regions of both hemispheres. Only if one re~ion

presents with a flow value which is beyond t he value of
t he normal flow t his can be considered as being pat ho-
logic.
2. Repeated measurements (Test-Retest-manoeuvres) in
patient s with vascular disease of the brain but without
focal perfusion deficit s have proven that mean rCBF
in t he subsequent measurements does not devi ate by mor e
than 3.1%+2.2 between 20 min and 24 hours. Only the
calculation of mean flow (MF) leads to a significantly
lower flow 20 min after the fi rst me asurement (Fig. 2).
This stable flow calculation in subsequent procedures
changes wi t h the presence of v as cular ob struction and

595
 CBF
~ steady state
0 20 min l ater

d hi
[ ] 3 hrs l ater
E ~. 1 day later
0>
0 p<O.OS
g
-
E

F 1 l SI MF

Fig. 2: Test-Retest-manoeuvres in patients with verte-

brobasilar insufficiency, who did not have any focal
rCBF-deficit. rCBF-measurements were repeated after
20 min, 3 hours, 24 hours. Except for MF in the 20
min-procedure blood flow calculation did not reveal
any significant difference.

acute ischemia. Here at least 60 min have t o elapse be-

fore the second measurement can be started.
Considering these methodological suppositions it be-
comes possible to judge the blood flow condition of an
individual patient during one time and repeated mea-
surements.
The aim of calculating rCBF in patients with cerebro-
vascular diseases is to judge the possible effect of
the therapeutic efforts. There are several possibili-
ties to improve tissue function in patients with vas-
cular disease of the brain:

596
1. Increase of vascular diameter
Reduction of cerebral vascular resistance by vasodila-
tion may lead to increase of rCBF. However, most of the
substances have failed to prove their beneficial effects
regarding blood flow enhancement. This might be due to
stiffness of arteriosclerotic vessels or to systemic
side effects of the drugs. Regarding the therapy of
chronic or acute ischemia of the brain due to arterio-
sclerosis the era of vasodilators has passed.

2. Alteration of metabolism
In patients in whom reduction of cerebral metabolism
leads to neurologic or psychiatric symptoms brain's
metabolism should be corrected. However, this is not
true for cerebrovascular diseases of the brain. In
these cases the primary cause of functional impairment
must be searched in the vessel wall, the vascular con-
tent or the forces which act on the vessel and the
propulsion of the blood through the organs. So far me-
tabolic therapy has only a supportive capacity.

3. Improvement of rheology
In acute and chronic ischemia of the brain rheological
condition of the blood are altered: viscosity might be
increased, platelets might possess increased tendency
to aggregate, erythrocyte might be stiffer than usual
(f. i. due to diabetes mellitus) or tend to aggregate.
These characteristics influence rCBF tremendously and
therefore play a decisive role for therapy. In addition
reduced blood flow to the brain - due to any cause -
increaseSviscosity by itself and closes the vitious
cycle of further rCBF-reduction. Improvement of blood
characteristics improves rCBF, prevenmischemia of the
brain in critical areas and avoids spread of ischemia
to focal areas in the neighborhood.

597
We like to present our experiences with improvement of
rCBF by rheological manoeuvres.
1. Nicergoline (N.)
Nicergoline - used as one of the drugs which is stated
to improve blood flow of the brain (2) and to reduce
platelet aggregation (3) - was tested in an open con-
trolled study: 20 outpatients with the diagnosis of
chronic cerebrovascular diseases and proven focally re-
ducend rCBF underwent a protocol, where rCBF was mea-
sured twice with a time of 4 weeks in between. Half of
the patients were not treated and served as controls,
the other 10 patients were treated with 3x 5 mg Nicer-
goline/day for 4 weeks. Looking at the total group mean
rCBF did not change significantly (p > 0.1) in both groups
(+0.9%±2 for the control group versus +2.1%±8.9 in the
N.-group). Only one out of 20 hemispheres (10 patients)
of the N.-group showed a significant increase of mean
rCBF. Looking at the regional data it became obvious
that in the N.-group more areas occurred with signifi-
cant rCBF-increase than in the control-group (n = 48/8).
On the other hand more areas in the N.-group presented
after therapy vli th a significant reduction of rCBF
(n = 18) compared to the control-group (n = 4). However,
the majoritY of brain tissue in the patients treated
with Nicergoline was not influenced by the therapeutic
manoeu.vres.

2. Low 11olecular Dextrane (U1D) and H:vdrox:veth:vl-starch

(HE)
LI1D improves blood viscosity (4) and is used for hyper-
and isovolemic hemodilution. HE is under research re-
garding its effect on rCBF and usefullness in cerebral

598
ischemia since it has beneficial action on viscosity
also. The improvement of global blood flow of LMD was
shown by Gottstein (5) and nowadays it is one of the
routine substances for therapy of ischemic infarction
in West Germany.
Little is known about the longtime effect of both sub-
stances on rCBF. We have measured rCBF in patients
with acute infarction of the brain who came to the me-
dical service on the day of onset of symptoms. rCBF
was measured three times, on the day of infarction
prior to start of therapy, 1 day and 7 days later.
Treatment with either LMD (500 ml/day) or HE (500 ml/
day) was performed throughout the protocol. The data
were compared to a control-group, who could not be
treated with hemodilution for medical reasons. This
group differs from both verum groups regarding age of
the patients but not in respect to diagnosis or severity
of disease.
Fig. 3 indicates the mean CBF-data in all three per-
formances. During steady state (day 1) there was no
difference between all 3 groups. The next day (day 2)
mean CBF in both treated groups was significantly
higher than in the control group, but there was no sig-
nificant difference between the patients treated with
either LMD or HE. The same was true for the measurements
on day 7.
It was concluded that both LMD and HE are able to im-
prove blood flow of the brain in ischemic infarction.
These results can be considered only as preliminary
data since the study was not performed in a double
blind manner and since the control group differed from
both verum groups regarding age of the patients.

599
 HYPERVOLEMIC HEMODILUTION

70 DEXTRAN 40 C500 ml/dle):

HAES CSOOml/dle>:
• •
(!)---(!)

CONTROLS .__ -4>

60

50

30

20
DAY
Cpre lnf.)

Fig.j: Mean CBF course of the infarcted hemisphere in

patients with acute unilateral ischemic cerebral in-
farction. CBF-measurements were done on day 1, 2, 7.
Both treated groups (either with low molecular dex-
trane-DEXTRAN- or with Hydroxyethyl starch-HAES) had
a higher mean CBF on day 2 and 7 compared to the con-
trol group. There was no difference between both
treated groups.

600
3. Pentoxifylline (P)
P reduces viscosity (4) thus improving microcirculation.
It has the advantage, that it might be given either par-
enterally or orally. Since the effect of a substance
in daily practical medicine is important in its orally
given form we have tested the effect on rCBF before
and after a 4-week period of the treatment with daily
dosage of 3x 400 mg. 10 patients with chronic cerebro-
vascular diseases and several focal tissue perfusion de-
ficits were selected for the protocol.
Fig. 4 presents the regional data: out of 311 regions
of interest (20 hemispheres) 07 ROI presented with re-
gional ischemia as compared to the pattern of normal
volunteers (Fig. 1). of these 207 ROI increased
rCBF after therapy in a manner that rCBF became normal.
Regarding the total group mean rCBF increased by 12.~fo
(p < 0.005) in all right and by 5. o/fo (p < 0.05) in all
left hemispheres. From these results it is concluded
that rCBF might be improved in patients with chronic
ischemia of the brain by administration of P. About
26% of all ischemic areas improve their flow up to a
normal level of tissue perfusion.
In conclusion, measurement of rCBF with the atraumatic
Xenon 133 design permits repeated controls of one of
the parameters which is important for the integrity
of the brain's tissue function. Our believe is that
improvement of rheological conditions which reduce
viscosity, enhance erythrocyte flexibility and de-
creases aggregation of both platelets and erythrocytes
improve regional cerebral blood flow and of signi-
ficant help in relieving neurologic and psychiatric
symptoms due to acute or chronic cerebrovascular dis-
orders.

601
 EFFECT OF PENTOXIFYLLINE ON F1-CBF
lj 10(
lml IOOg-' m•n-'

80 ·.
...•:
......:......
·": .·.
..., .. ·..
·:....
.. ';... . ......
.. .. ...
:...-::·._..: ·.·~=
~
60
·=·. ··::· ·.~
.
> \ . .. ·:·· .•
~:.

"cr . : ~::- ~:·:.~., '.j. ·::·

~ ..· . . ··:
~ .: ·.: :: ~ ·.,

:::0 40 ·.. '.

.
~

~ ·.
----~---L--~----li---~---------F1
40 60 80 100-
AFTER THERA_PY

Fig. 4: Effect of Pentoxifylline on rCBF. All regional

data of all patients prior to start of therapy are indi-
cated on the ordinate and are compared to the rCBF-data
after completion of the protocol (abscissa). The vertical
and horizontal cross indicates the mean value and standard
deviation of the normal CBF in this age group (Fl). The
following is indicated: 53 areas changed from hypoemia-
(prior to therapy) to normoemia or even hyperemia, 54
hypoemic areas did not reach normoemic flow, 17 areas were
normoemic before start of therapy and became hypoemic
and 38 areas were normoemic before and after completion
of the protocol.

602
References

1. A. Hartmann, R. v. Kummer
Fortschr. Neurol. Psychiat. 5, 57 - 68 (1968)

2. Iliff, L. D., Du Boulay, G. H. et al.

J. Neurol. Neurosurg. Psychiat. 40, 746 (1977)

3. Fregnan, G. B.
Acta Univ. Carol. Med., Monographia L III -IV,
263 (1972)

4. H. Heidrich, M. Ott
Herz Kreisl. 6, 542 - 546 (1974)

5. u. Gottstein
Arzneimittelforschung 31, 2028- 2032 (1981)

603
ORGANIC DEMENTIA: CLINICAL PICTURE RELATED TO REGIONAL

CEREBRAL BLOOD FLOW AND NEUROPATHOLOGICAL FINDINGS

Lars Gustafson, Arne Brun* and Jarl Risberg

Departments of Psychiatry and Pathology*

University Hospital
S-221 85 Lund, Sweden

INTRODUCTION

One of the most important problems in psychiatry is

the differentiation between organic and non-organic dis-
orders i.e.the differentiation between organic dementia
and pseudo-dementia secondary to psychiatric illness.
The influence of different organic and psychological
factors has to be evaluated with diagnostic tools such
as clinical ratings, psychometric testings and neuro-
physiological techniques.

The present paper summarizes results from our

research in patients with suspected organic brain syn-
dromes. The diagnostic differentiation based on regional
cerebral blood flow (rCBF) measurements will be compared
to diagnoses based on standardized clinical ratings and
to neuropathological evaluations.

MATERIALS AND METHODS

The patients pertained to a longitudinal study of

presenile dementia and dementia of old age. The patients
were selected not to include chronic psychosis, oligo-
phrenia, addiction, pre-existing epilepsy, severe somatic
disorders or apoplexy with severe neurological defects.
All patients underwent psychiatric, psychometric, neuro-
logical, neurophysiological (EEG and rCBF) and general

605
somatic examinations. At follow-up the patients were
studied with psychiatric rating, psychometric testing and
rCBF measurements. At death a neuropathological investi-
gation was performed in 33 cases.

The rCBF Measurement. The rCBF was measured with the

133-Xe inhalation technique (Obrist et al., 1975) that has
been used in our laboratory since 1975. With this technique
it is possible to measure rCBF in the cerebral cortex of
both hemispheres simultaneously. The isotope 133 xenon
mixed with air is inhaled by the patient for one minute
and thereafter eliminated during ten minutes of normal air
breathing. The flow parameter initial slope index (ISI)
presented here indicates blood flow in the grey matter
(Risberg et al., 1975). RCBF can be measured repeatedly
during rest, mental activation and treatment.

The Psychiatric Evaluation focussed upon the diffe-

rentiation between Alzheimer's disease (AD), senile demen-
tia of the Alzheimer type (SDAT), frontal lobe dementia of
the Pick type (PD) and multi-infarct dementia (MID). The
differentiation was based on the ischemic score developed
by Hachinski et al. (1975) and two rating scales for iden-
tification of Alzheimer's disease and of frontal lobe
dementia of the Pick type developed by Gustafson and
Nilsson (1982). The rating scale for diagnosis of AD con-
tained clinical features such as early spatial disorienta-
tion, apraxia, aphasia, agnosia, logoclonia and increase
of muscular tension. The rating scale for PD contained
e g early loss of insight, early signs of disinhibition,
echolalia, mutism and amimia. Patients who could not
clearly be classified as belonging to any of the groups
i e due to scant data were not included in the present
analysis. The separation between AD and SDAT was the time
of onset of dementia - before or after 65 years of age.

The Neuropathological Diagnosis was based on semi-

serial sections of whole hemispheres as described by Brun
and Gustafson (1976). With this technique it is also
possible to compare the rCBF and the regional distribution
of pathological changes in the brain.

RESULTS AND DISCUSSION

By the standardized clinical evaluation 76 patients

could be diagnosed as AD (19), SDAT (17), MID (21) or
PD (19) cases. The mean hemispheric flow values of these
diagnostic groups, a group of 32 non-demented depressed
patients and a reference group of 16 normal subjects are
presented in Fig. 1.

606
 Mean rCBF values in normal subjects and in pat1ents with
depression. multi-infarct dementia. Alz.heimer's disease.
senile dementia of the Alz.heimer type and Pick's disease.

rCBF 0 right hemisphere

(151) f:23 left hemisphere
60
50 T r
40 ~ 1 T 1 T

30
11
T t J1 ~
20
10

norm. depr MID AD SDAT PO

AGE 53:!:5 60:!:14 os:!:G 64:!:6 78:!:7 62:!:8
16
" 21 19 17 19

Fig. l.
The average flow values of MID, AD, SDAT and PD cases
are significantly lower (p < .01, t-test for uncorrelated
means) than those of depressives and normals. As shown by
Gustafson et al. (1981) not only non-demented depressives
but also manic patients with slight or no medication have
normal rCBF levels and patterns. Thus rCBF can improve the
differentiation between organic dementia and pseudo-demen-
tia due to affective disorders. The normal subjects in
Fig. 1. were, however, somewhat younger than the demented
cases and with increasing age there is a slight decrease
of rCBF.
31.9 t2y 40.1t3y 52.8t5y
(28. 35) (36·45) (46·63)

Rt ~
lSI~
-25 .. e +25%

pC02 41.7t3.0 40.0t2.3 37.8t3.4

Fig. 2. rCBF in three normal agegroups

607
 The mean hemispheric flow values are shown in the
boxes. The regional deviations are indicated by the clock
symbols. Black indicates a value above, striped a value
below the hemispheric average. A 25 % deviation from the
mean is equal to 90 degrees in the clock symbol. Twelve
o~clock indicates a regional value = hemispheric mean.

Normal subjects have during rest a symmetric hyper-

frontal flow pattern with rCBF values 5 to 15 % above the
hemispheric mean in frontal areas. The three reference
groups show a slight decrease of the hemispheric mean with
increasing age, while the normal flow pattern remains un-
changed.
The differentiation by means of rCBF between differ-
ent types of dementia is mainly based on the regional
distribution of the flow pathology. In presenile Alzhei-
mer~s disease there is a general flow decrease, which is
accentuated in the temporo-parieto-occi pital area. In
patients with frontal lobe dementia of the Pick type the
flow decrease is most marked in frontal and anterio-
temporal regions.

t4b Al z Mb Pick
AGE 64t6 AGE 62t8 n•19

Rt

Lt

Sign. levels (t·test l * pc .OS

** pc .01
*** pc .001

Fig. 3. Comparison between rCBF in Alzheimer's disease

and Pick-s disease (Risberg and Gustafson, 1983).

608
 Fig. 3. shows the regional flow pattern in patients
of a similar age with Alzheimer's disease and Pick's dis-
ease. The differences in flow pattern are highly signifi-
cant and in good agreement with the distribution of
cortical degeneration in these cases. The typical pattern
of degeneration in presenile AD is shown in Fig. 4. (Brun
and Gustafson, 1976; Brun and Englund, 1981).

In our experience the cortical degeneration in AD is

most pronounced in the medial temporal limbic areas, the
posterior cingulate gyrus, the posterior inferior temporal
areas and the adjoining portions of the parieto-occipital
lobes. On the other hand the sensory motor,calcarine and
anterior cingulate areas of the cortex are notably spared
(Brun and Gustafson, 1976). This pattern of degeneration
is in good agreement with the rCBF findings in AD. The
flow pathology is similar in AD and SDAT. In AD, however,
the flow decrease is more focalized, while the flow patho-
logy is more diffuse in SDAT. A typical finding in both
groups is the better preserved flow in Rolandic, occipital
and fronto-temporal structures.

Both from the rCBF and the neuropathological point

of view the multi-infarct group is more heterogeneous than
AD and PO cases. Therefore the rCBF average in the multi-
infarct group is a rather flat rCBF pattern (Risberg and
Gustafson, 1983). The typical finding in individual multi-
infarct cases was a general as well as spotty flow
asymmetry.

D n
w
17:1
LJ
no slight moderate severe total degen.
Fig. 4. Schematic representation of distribution of
degeneration in pre senile Alzheimer-s disease
(adapted from Brun and Gustafson, 1976).

609
 AD SDAT MID PO N

IT]
(/)
·u; 0
c(
0
c::
en
0
1-
cu c(
0 2 1
=a rn

12!1
I
0
0 1
...
m
~
.r:
cu 0
Q. 0.

...
0
~
Q)
z
z
IT]
Fig. 5. Neuropathological diagnosis compared to rCBF
diagnosis in 33 deceased patients. N, normal findings.

The correspondence between the diagnosis based on

rCBF findings and the neuropathological diagnosis in the
33 cases with a complete neuropathological investigation
is presented in Fig. 5.

The rCBF diagnosis was made without knowledge of the

neuropathological data and was correct in close to 90 %
of the cases. Only 4 patients showed disagreement, three
out of whom belonged to the SDAT group. Two of these
were diagnosed as multi-infarct dementia due to blood
flow asymmetries. In these cases vascular lesions were
found in addition to the Alzheimer encephalopathy. The
five cases with normal rCBF and no neuropathology
suffered from pseudo-dementia caused by depression and
other psychotic reactions. Thus there was a good corre-
spondence b~tween neuropathological diagnoses and rCBF
diagnoses.

CONCLUSION

The results presented here indicate the clinical

usefulness of the 133-Xe inhalation technique for differ-
ential diagnosis of dementia. Regional flow variations
correlate well with clinical and neuropathological
findings and can thus be used for diagnostic purposes.
The technique, which allows repeated measurements with
short intervals, has been used for evaluation of treat-
ment in dementia and in non-organic mental disorders. As
presented at this congress and elsewhere (Silfverskiold
et al., 1983) the 133-Xe inhalation technique has been
used to study depressive disorders during ECT.

610
Acknowledqements: This study was supported by the Swedish
Medical ReseaiCfi Council, grants No. 21X-4969, 21X-3950,
21P-5144 and The King Gustaf V and Queen Victoria~s
Foundation.

References

Brun, A., and Englund, E., 1981, Regional pattern of de-

generation in Alzheimer~s disease: neuronal loss and
histopathological grading, Histopathology, 5, 549-64.
Brun, A., and Gustafson, L., 1976, Distribution of
cerebral degeneration in Alzheimer~s disease. A
clinico-pathological study, Arch. Psychiat. Nervenkr.,
223, 15-33.
Gustafson, L., Risberg, J., and Silfverskiold, P., 1981,
Regional cerebral blood flow in organic dementia and
affective disorders, in: "Advances in Biological
Psychiatry," J. Mendlewicz and H.M. van Praag, eds,
s. Karger, Basel.
Gustafson, L., and Nilsson, L., 1982, Differential dia-
gnosis of presenile dementia on clinical grounds,
Acta psychiat. scand., 65:194-209.
Hachinski, V.C., Iliff, ~.D., Zihlka, E., du Boulay, G.H.,
McAllister, V.L., Marshall, J., Ross Russell, R.W.,
and Symon, L., 1975, Cerebral blood flow in dementia,
Arch. Neurol., 32, 632-37.
Obrist,- W. D., Thompson, H. K., Wang, H. S., and Wilkinson,
W., 1975, Regional blood flow estimated by 133-Xe
inhalation, Stroke, 6:245-56.
Risberg, J., Ali, z., Wilson, E.M., Wills, E.L., and
Halsey, J.H., 1975, Regional cerebral blood flow by
133-Xe inhalation. Preliminary evaluation of an
initial slope index in patients with unstable flow
compartments, Stroke, 6:142-48.
Risberg, J., and Gustafso'n, L., 1983, 133-Xe cerebral
blood flow in dementia and in neuropsychiatry re-
search, in: "Functional Radionuclide Imaging of the
Brain," 'P":"L. Magistretti, ed., Raven Press, New York.
Silfverskiold, P., Gustafson, L., and Risberg, J., 1983,
Changes of regional cerebral blood flow during
electroconvulsive therapy in depression, in: "ECT:
Basic Mechanisms," B. Lerer, R.D. Weiner, and R.H.
Belmaker, eds, John Libbey & Company Limited, London,
Paris.

611
REACTION TIME IN VASCULAR
(MULTIINFARCT) DEMENTIA

Gunther Ladurner, Marianne Tschinkel, Hedwig

Klebl, Helmut Lechner
Denartment of Psychiatry and Neurology of
the University Graz, Auenbruggerplatz 22
A-8036 Graz
INTRODUCTION
Reaction time (RT) is a physiological measurement
which has been found to increase with age and cerebral
disease (Benton 1977). For the increase in RT in cere-
bral disease (Klensch 1973), the particular localisation
of the lesion (Benton and Joynt, 1958; Howes and Boller
1975) as well as dementing processes have been consi-
dered to be responsible (Miller 1974; Pirozzolo et al.,
1981). However, it still remains unknown whether the RT
is related to diffuse or focal brain disease or is
caused by the loss of mental capabilities. To clarify
the situation it seems therefore of interest to inve-
stigate patients with cerebrovascular disease (CVD) in
whom all relevant parameters, focal lesions in the
non-dominant and dominant hemispheres as well as demen-
ting processes could be studied.
MATERIAL AND METHODS
95 patients have been investigated. They were sub-
divided according to their clinical and computertomo
graphical findings into two main groups:
1. Controls (31) with an average of 51 years and normal
psychiatric and neurological findings were studied
in whom there was also no known history of a
disease altering RT.

613
2. Cerebrovascular patients (64) with an average age of
59 years formed the second group. 45 patients had
focal neurological symptoms (hemiparesis, aphasia,
hemianopia etc.), 19 had non-focal neurological symp-
toms {diplopia, dysarthria etc.). 4 patients showed
no neurological findings on admission.
36 patients were demented; in 26 cases a psychologi-
cal test (WAIS) showed signs of loss of mental capa-
bilities. In 10 cases the diagnosis was based on
clinical findings only. For the comparison with the
controls the CVD group was age matched in patients
between 45 an 60 leaving 34 CVD patients. The inve-
stigation in the CVD group - comparing dominant ver-
sus non-dominant lesions etc - made use of data from
the whole CV group (64).
The localisation of the infarcts was based on compu-
tertomography (CT) in each patient. Atrophy was diagno-
sed by estimation only and no measurements have been
applied.
The reaction time (composed of movement and deci-
sion time) was measured with a system (Schuhfried) dur-
ing a set of 8 optic signals at random. The time from
the signal to leave a button (decision time) and the
time to move from the button to the knob to press
(movement time) was calculated and added to the reaction
time. This reaction time was measured for the dominant
as well as for the non-dominant hand. The discrimina-
tion dominant/non-dominant was obtained form the pa-
tient's history.

R E S U L T S
Controls
A significant correlation (pZO.OOl) could be
found between the age and the RT. No correlation, how-
ever, could be seen between RT and enducational level
aswell as IQ. Concerning dominant versus non-dominant
hemisphere, RT was not significantly different.
Cerebrovascular Disease
The same correlation with age was found also in
the CVD group. There was no correlation between the
IQ (measured in patients with the WAIS) and the RT.

614
On the other hand, the deterioration index (Baxa et al.,
1972), as a sign of lost mental capabilities, showed a
significant correlation with optical RT.
No correlation was found between the degree of a
paresis of the left of right side and the RT. It has to
be remembered, however, that the degree of the paresis
was mild to allow the patient to move his hand and fin-
gers and to press a button.

Comparison CVD/Controls
The comparison of RT in Controls and CVD showed a
significant difference (Table 1). When the CVD group was
divided with regard to their mental capabilities into
demented and non-demented patients (Table 2), it could
be seen clearly that the RT was without significant
difference in the Controls and CVD patients without de-
mentia (Table 3), whilst there was a significant diffe-
rence present between demented CVD patients and those
without dementia and the Controls.
~omparison CT/CVD
Comparing the patients with CVD with and without
atrophy in the CT (defined by experienced investigator
without measurements), a significant difference in RT,
it being higher with atrophy could be seen.
When the patients with a lesion in the non-domi-
nant or dominant hemisphere and patients with bilateral
lesions were compared, no significant difference could
be seen. There was also no difference with regard to
patients with CVD and normal CT's. Dividing the pa-
tients with normal CT's and those with bilateral in-
farcts into the demented and non-demented patients
showed again a significantly higher RT in the demen-
ted patients (Table 3).

D I S C US S I 0 N
The assessment of dementia in patients with CVD
is not always easy especially if the IQ is very low
and depressive mood changes are present (Nunn et al.,
1974; Ladurner et al., 1981). Even when psychological
testing is considered to distinguish between the
demented and the non-demented CVD patients, this can
sometimes only be decided in follow-up investigations.

615
Here a psychological test which is easily repeatable and
has no limitations over the whole IQ range would be of
great help. RT which has these two qualities has been
used in patients with organic brain disease (Klensch
1973; Benton 1977) and shown to be increased in associa-
tion with brain disease. This was also found in demen-
ting diseases (Miller 1974; Pirozollo et al., 1981) so
that the question arises if RT is related to brain di-
sease as a whole, to diffuse or localised alterations
in the brain or if the loss of mental capabilities is
the main factor. This question has not been solved so
far in that lesions in the non-dominant hemisphere
(Howes and Boller, 1975) have been considered respon-
sible whilst in other investigations there was no side
difference (Dee et al., 1972; Dee et al., 1973) for the
lesion but a significant increase in RT caused by le-
sions in the left versus the right hemisphere was found
with increasing task complexity.
From our investigation we could conclude that
there is an increase in RT following the CT criteria: no
lesion, unilateral non-dominant, dominant and bilateral
lesion. This has also been considered to be of impor-
tance in the same way for the appearance of dementia in
patients with CVD (Ladurner et al., 1982). Dividing then
the patients with CVD into demented and non-demented, it
could be clearly seen in our work that the cognitive
impairment in CVD is the key factor altering RT. There-
fore, we think that the differences found between na-
tients with brain disease and controls are related to
the fact that brain disease is usually to some extent
related to a loss of mental capabilities. That the de-
mentia is the most important factor for the increase
in RT would also be in agreement with findings in Alz-
heimers disease, showing a correlation between the
degree of dementia and the increase in RT (Pirozollo
et al., 1981). It can therefore be concluded that RT
is a good and simple diagnostic tool for the differen-
tiation of demented versus non-demented CVD patients
and could also be easily used for follow-up investiga-
tions.

616
 Table 1. Optical Reaction Time

Controls CVD
Dom DT 276 310++
Dom RT 402 480+++
NDom DT 268 306++
NDom RT 403 470++
++ +++
p < 0.01 p < 0.001

Table 2. Optical Reaction Time in Controls

and Cerebrovascular Disease

CVD
Control without with
Dementia
Dom DT 276 384
Dom RT 402 432
NDom DT 265 271
NDom RT 403 4 31
+++p < 0.001 (Control) ++p <0.01 (CVD without
Dementia)
+
p < 0. OS (CVD without Dementia)

Table 3. Optical Reaction Time in CVD and Localisa-

tion of Lesions versus Dementia

Normal CT Bilateral Infarct

without with without with
Dementia
Dom DT 279 323+'1' 300 383++
Dom RT 392 518+ 4 71 572
NDom DT 253 317++ 269 338+
NDom RT 390 522 4 11 528++
+ ++ p < 0.01
p < 0.05

617
 S U MMA R Y
95 patients have been investigated by optical re-
action time (RT). 31 were vontrols, 64 had cerebrovas-
cular disease and were investigated by CT.
CVD-patients showed a significantly higher RT than
controls. If the CVD group was divided into patients
with and without dementia, the demented only were sig-
nificantly increased in RT in comparison with the other
CVD patients and the controls. The localisation in the
CT had no significant influence on RT whilst the
appearance of atrophy was leading to a significant in-
crease in RT. Dividing patients with normal CT and
those with bilateral infarcts into patients with and
without dementia, a significantly higher RT was also
found in the demented versus the non-demented patients.

REFERENCES
Benton, 1977, Interactive Effects of Age and Brain Dis-
ease on Reaction Time, Arch. Neural., 34:369-
370
Klensch, H., 1973, Die diagnostische Valenz der Reak-
tionszeitmessung der verschiedenen zerebralen
Erkrankungen. Fortschr. Neural. Psychiat., 41:
575-581 -
Ladurner, G., Pieringer, W., Sager, W.D., 1981, Depres-
sive Syndromes in the Middle Age and Organic
Brain Disease. Psychiatria Clinica, 14: 105-111
Ladurner, G., Iliff, L.D., Lechner H., 1982, Clinical
factors associated with dementia in ischaemic
stroke. J. Neural., Neurosurgery and Psychiatry,
45: 97-101
Miller, E., 1984, Psychomotor performance in presenile
dementia. Psychological Medicine,4: 65-68
Nunn, C., Germann, K., Britton, P.G., ~orster, E.M.,
Hall, E.H., Kay, E.W.K., 1974, Intelligenre and
neurosis in old age. Br.J. Psychiatry, 124:
446-452 -
Pirozzolo, F.J., Christensen, K.J., Ogle, K.M., Hansch,
E.C., Thomson, G., 1981, Simple and Choice
Reaction Time in Dementia. Neurobiology of
Aging, 2: 113-117

618
COMPUTER AS A PROSTHESIS IN THE NEUROPSYCHOLOGICAL REHABILITATION

OF STROKE PATIENTS

Francisco I. Perez, George A. Brmvn, Michele Rusin'~ 1',

Dale Koehler and Victor Rivera

Baylor College of Medicine, Houston, Texas U.S.A.

Since the advent of small computers in the mid-1970's, there

has been explosive growth in microtechnology. Advances in miniatur-
ization and speed have produced small, inexpensive "desk top com-
puters" with the capabilities of larger machines which cost $20,000
to $50,000 ten years ago. The reduction in cost and the introduction
of an ever-increasing variety of microcomputers have led to a great
diversity in their applications (1). It has been reported that over
33% of American businesses now use small computers (2). The small
computer is also being implemented in various school systems where
our children are learning computer akills at an early age.

The relatively low cost of these systems and the enormous

variety of software make these machines an idea prosthesis in the
assessment and rehabilitation of the handicapped.

Small computers may be utilized in a very effective manner in

the areas of assessment of neurological deficit, cognitive and sen-
sory retraining, and as a prosthetic device (3).

Several studies (3, 4, 5, 6, 7) demonstrate that computer

assessment may be more sensitive to the behavioral deficits of
strokes than existing conventional techniques. Tasks can be devel-
oped to monitor the behavioral status of stroke patients from acute
onset through recovery and rehabilitation (4). The advantages of
computer assessment may be described as follows:
1. Problems in reading material or writing are eliminated.
2. Differences in examiner technique and bias are eliminated.

** Behavioral Medicine Consultants of Houston

619
 3. Automated testing allows for precise recording of the
behavior deficit under study.

After a brain injury, including CVA, a person may have to con-

tend with a variety of problems. Sometimes, these problems are not
readily apparent. A person may recover motor function but still be
left with cognitive or sensory dysfunction. The small computer can
be used to retrain a patient to compensate for his difficulty.
Patients exhibiting auditory and visual neglect have been success-
fully taught to compensate their difficulties when the computer
exhibits exact control over stimulus properties (5, 6, 7). Many
other investigators have developed specific programs for memory and
visuo-motor retraining.

We have found that the use of a special computer language

(LOGO) can be used to train CVA patients to enhance visual-perceptual
planning. LOGO's strength comes from the fact that the patient uses
the computer as a control device, LOGO users control an on-screen
cursor (called a turtle) to draw, create animated sequences, invent
video games or compose music. Drawing on the computer's video screen
is accomplished by simple commands as FORWARD, BACK, RIGHT, LEFT,
etc. These commands may be sequenced to enable the patient to draw
simple to complex figures. The patient learns problem-solving and
planning.

Small computers are also being used as prosthetic devices in

environmental control and as an aid in speech generation for those
who have lost speech ability. A number of manufacturing companies
offer hardware to allow an individual to control his environment by
having the computer read switches, close switches, initiate timing
functions. The small computer may be used to 1) answer or dial a
telephone, 2) control power, volume, and channels on a television,
3) operate appliances (lights, ovens, radio), 4) compose or type
letters, 5) run entertainment and educational programs, and 6) In-
teract with other computer systems via telephone lines.

Speech generation may be easily accomplished through digitized

speech or phoneme synthesis. This laboratory uses the VOTRAX person-
al speech system which is interfaced with an Apple II+ computer. The
VOTRAX is a self-contained text-to-speech synthesizer. This computer
peripheral produces unlimited speech output via literal English,
ASCII text input data, whether entered by a keyboard or ASCII
communication link. The synthesizer accesses basic sound units of
speech (phonemes) using a set of rules that describe English pronun-
ciations. The phonemes create speech output via a VOTRAX SC-01 chip.
Translations are performed independently of the host computer which
frees the computer for other tasks while speech is in progress.

Many of our patients generate speech by direct typing on a

keyboard. Programs may also be used to allow phrases or sentences

620
to be "spoken" by the patient striking one key.

As one may see, the small desktop computer may offer stroke
victims a way to control their environment, communicate with others,
and even retrain damaged cognitive processes. The low price of these
machines and the availability of software make these machines a very
practical choice for wide use with the handicapped.

REFERENCES

1. Barna, A., and Porat, D. Introduction to microcomputers and

microprocessors. John Wiley and Sons, New York, 1976.
2. Infoworld Weekly. Popular Computing, Inc. Framingham, MS.
September 13, 1983.
3. Perez, F. , Brown, G. A. and Lee, D. Memory Scanning in the
Dementias. Paper presented to the American Geriatric Society,
Dallas, 1978.
4. Perez, F., Brown, G.A., Cooke, N., and Grabois, M. Stroke
patients: A computerized-behavioral approach to cognitive
assessment and retraining. Archives of Physical Medicine and
Rehabilitation. Vl 61, October, 1980.
5. Perez, F., Brown, G.A. Auditory discrimination training in
patients with auditory neglect. Paper presented to the 4th
World Congress of the International Rehabilitation Medicine
Association. San Juan, Puerto Rico, April, 1982.
6. Madison, C., Brown, G.A. and Perez, F. Forced fixation as a
technique to retrain patients demonstrating visual neglect: A
computerized approach. Paper presented to the 4th World Congress
of the International Rehabilitation Medicine Association, San
Juan, Puerto Rico, April, 1982.
7. Madison, C., Brown, G.A., Damanoir, C., Perez, F. and Grabois,
M. Rehabilitation of visual neglect using a computerized shaping
procedure. Paper presented to the 58th Annual Session of the
American Congress of Rehabilitation Medicine, San Diego,
November, 1981.

621
NEUROPSYCHOLOGICAL ASPECTS IN CORTICAL BLINDNESS

DUE TO CEREBRAL INFARCTS

Victor M. Rivera*, G. R. Ravichandran**, Francisco I

Francisco I. Perez*, Wini Breitbach*,
Enrique Wulff* and Francisco Grajeda***

*Houston Neurological Center, Houston, Texas

**Texas Research Institute in Mental Sciences, Houston
***University of Guadalajara, Mexico

INTRODUCTION

Persistent cortical blindness following occipital infarcts is

commonly known to clinicians. The term "cortical blindness"
describes an inability to see due to lesions in the occipital lobes.
In these cases, the ocular globe structures are intact, as·are the
retina, optic nerves, chiasm and optic fibres. The duration and de-
gree of severity of this disabling neurological deficit is studied
in this report.

METHODS

Four adult patients who experienced cortical blindness due to

cardiac emboli to the occipital areas (one patient with an occipi-
tal infarction complicating a ruptured periventricular arterio-
venous malformation and one with thrombotic infarcts due to occlu-
sive disease of the basilar artery) were investigated with CAT scans,
VER, neuropsychological assessments, and one patient with Automated
Behavioral Analysis System (ABAS). This latter assessment system
consists of a computer-based interactive behavioral evaluator. This
is basically a visual recognition memory task and the objective is
to titrate recognition memory. Two tasks are used inducing a verbal
stimuli task (non sense syllable trigrams) and a spatial stimuli
task (random shapes).

RESULTS

Infarction of the occipital lobe, as well as hypoxic damage to

the cerebral cortex affecting the visual areas, can lead to

623
f jgure 1. Cat scan of brain showing bilateral enhancing is chemic
lesions. Embolic infarcts in case number 1.

blindne ss 1 • Tran si e nt cortical blindness has a l so been described

as a n unusual complication of co ronary angiography presumably re-
sulting in an embolus to the bifurcation of t he basilar artery2 .

In four of our cases, there was associative visual a nosognosi a

and de ni a l of blindness (Anton's syndrome) whi c h can be a str i king

624
feature in some cases of cortical blindness3. This phenomenon was
short-lived in two patients (Cases No. 1 and No. 2) but persisted
in a patient with dementia (Case No. 3) and in the patient with
basilar artery disease (Case No. 6). Visual anosognosia has been
described in patients with occipital lesions located close to the
parietal areas4. Case No. 1 also exhibited motor-visual ataxia in
conjunction with her visual motor coordination deficits5. Other
symptoms arising from occipital lobe dysfunction are the presence
of homonymous hemianopsia with preservation of optokinetic nystag-
mus (detected in Cases No. 2 and 4). The association of these
phenomena clinically suggests the presence of posteriorly located
cortical lesions and an occipital rather than a parietal locus6.
Occipital involvement in our patients was objectively confirmed by
CAT scan. (Figure 1) VER abnormalities, although not definitely
localizing, helped to assess occipital cortex dysfunction. Bilat-
eral extensive damage to areas 18 and 19 of the occipital cortex
produces grossly abnormal VER with suppressed amplitude and voltage
in the tracing. The greater the involvement of the occipital cdT-
tex, the more increasingly difficult it was to record activity from
the occiput stimulated by diffuse flashes inthe visual fields. VER
in cortical blindness has been found useful by other authors in
correlating the degree of electric disturbance with the extent of
patients with unilateral stroke and dementia8. This is basically
a visual recognition memory task and the objective is to titrate
recognition memory. ABAS can be used to evaluate residual visual
spatial processes in patients with cortical blindness. An embolus-
producing infarction in the calcarine cortex is seemingly the most
common cause of cortical blindness9,10. (Figure 2) Most cerebral
emboli originate from the heartll.

Case No. 3 had severe dementia, cortical atrophy and arterio-

sclerotic vascular disease, and was the most involved patient of
this group, having remained profoundly affected, while the less in-
volved first two patients had improved their neurological deficits
and visual function.

Patient No. 4 had a single embolus to the right occipital lobe,

but most likely, due to the "diaschisis" phenomenonl2, bilateral
occipital deficits took place temporarily.

Patient No. 5 suffered a left occipital infarct very likely

due to mechanical compromise of the left posterior cerebral artery
secondary to severe intraventricular hemorrhage and hydrocephalus.

Patient No. 6 had symmetrical occipital infarcts due to basi-

lar artery occlusive disease.

CONCLUSIONS

These observations show a proportional relationship between recovery

625
 Figure 2. Bilateral parieto-occipital low density lesions
(infarctions)

of a patient with cortical blindness and the presence of associ-

ated cortical defic its and dement ia . Prognosis is less promising
in the patient with more cortical involvement. Since cortical
blindness due to occipital infarctions is not an uncommon event,
the objective evaluation of this serious neurologic a l deficit . and
potentially permanent disability becomes paramount in the total

626
neurological assessment of the patient. The prognosis of this form
of ambliopia is relatively dependent upon the degree of involvement
of other associated areas in the cerebral hemispheres.

REFERENCES
1. A. B. Baker: Clinical Neurology Vol I New York, Harper and Row
Publishers, 1971.
2. Elder S. Duke: Text Book of Ophthalmology Vol VII London,
Henry Kimpton, 1971
3. W. M. Fischer, P.G. Gottschalk, D. N. Browell: Transient corti-
cal blindness. An unusual complication of coronary antiogra-
phy. Neurol 20: 353-355, 1970
4. S. M. James, Francois Boller: Associative visual agnosia and
its related deficits. The role of minor hemisphere in assign-
ing meaning to visual perception. Neuropsychologia 15 (2):
345-349, 1977
5. P. Rondot, J. DeRecondo, J. L. Ribadeau Dumas: Visuomotor
ataxia. Brain 100 (2): 355-376, 1977
6. A. B. Baker: Clinical Neurology Vol I New York, Harper and Row
Publishers, 1971
7. R. Sphelmann, A. G. Robert, Sam U Ho, E. L. Jan, A. N. Karyl:
Visual evoked responses and postmortem findings in a case of
cortical blindness. Trans Am Neurol Assoc 102: 157-160, 1977
8. F. I. Perez, N. A. Hruska, R. L. Stell, V. M. Rivera: Comput-
erized assessment of memory performance in dementia. Can J N
Sc 5: 307-312, 1978
9. C. Symonds, I. Mackenzie: Bilateral loss of vision from cere-
bral infarction. Brain 80: 415, 1957
10. P. J. Taugher: Visual loss after cardiopulmonary bypass. Am
J Oph 81 (3): 280-288, 1976
11. J. S. Meyer, z. C. Jonathan, V. M. Rivera, N. T. Mathews:
Cerebral embolization. Prospective clinical analysis of 42
cases. Stroke 2:541-554, 1971
12. J. S. Meyer, K. Tadashi, F. Yasuo, S. Kunio, W. D. Edward, D.E.
Arthur: Clinical prognosis correlated with hemisphere blood
flow in cerebral infarction. Stroke Vol 11: 383-394, 1971

627
CLINICAL AND MORPHOLOGICAL ASPECTS
IN DIAGNOSIS AND PROGNOSIS OF VASCULAR DEMENTIA

Gunther Ladurner and Goetz Bertha

Department of Psychiatry and Neurology of
the University Graz
Auenbruggerplatz 22, A.8036 Graz, Austria

INTRODUCTION
In the so called normal population dementia or
organic brain syndrome is present in 4.4-8% (Kay et al.
1970, Broe et al. 1976, Tower 1978). The percentage ri-
ses with age from 2.3% (65-69 years) to 22% (80 years
and older).
One of the key problems is the diagnosis of the
dementia at an early stage since the awareness of orga-
nic brain disease in old age is very low compared with
the appearance of neurological symptoms (Lechner et al.
1982). The second main problem is the diagnosis of the
diseases causing the dementia (Ladurner 1983). Whereas
in primary degenerative dementia (Alzheimer type) the
therapy is limited, in other diseases like in vascular
dementia there are several concepts for the treatment
or prevention of the disease. It is therefore important
to diagnose vascular dementia as early as possible, to
understand why some patients become demented after a
stroke and to know the way the dementia progresses.

DIAGNOSIS
1. Clinical factors
As has been shown in previous work (Ladurner
et al. 1982) the frequency of strokes in patients
with and without dementia was similar (Tab. 1).

629
 Table 1: Frequency of strokes
frequency of number of ischaemic stroke
strokes patients with without
dementia
36 21 15
2 24 12 12
3 7 5 2
4 (or more) 4 2 2
T o t a 1 71 40 31
not significant

The neurological symptomatology however, pointed to a

more common affection of the dominant hemisphere in the
demented than the non-demented patients. Bilateral neuro-
logical signs were also more frequent in the demented
group (Tab. 2).

2. Computertomography
At the first impression the loss of substance
either as atrophy or as infarcted tissue seems to
be important. If the CT findings are considered in

Table 2: Neurological symptoms - Hemispherical affec-

tion
neurological number of ischaemic stroke
symptoms patients with without
dementia
dominant
hemisphere 39 24+ l5
non-dominant 17 4 l3
bilateral 10 g++

T o t a 1 66 37 29
++
+
p < 0.05 p < 0.01

630
 more detail (Ladurner et al. 1982) not only the to-
tal loss of tissue is relevant but also the dhstri-
bution of infarcts. Here the bilateral infarcts - in
agreement with the importance of the bilateral neuro-
logical symptoms as well as the affection of the tha-
lamus - seem to be the main factors.

P R 0 GN0 S I S
1. Clinical
After a period of 33 months 33 % of the patients
with vascular dementia have died (Ladurner et al.
1984). The stroke incidence after diagnosis corre-
lates significantly with the course of the dementia
whilst the neurological status per se is not a
predictor of the outcome.
In contrast, in the group with psychiatric symp-
toms the appearance of confusional states corres-
ponds significantly with a poor prognosis.
2. Computertomog~aphy
The CT shows no occurrence of new infarcts in a
mean follow up period of 33 months in the patients
with stable symptoms.

In the patients with deterioration of the dementi~

atrophy and the number of infarcts increases significant-
ly (Tab. 3). There is also a significant correlation

Table 3: Prognosis of dementia and CT (infarcts)

dementia number of number of infarcts amount of atrophy
patients constant increased constant increased
stable 7 7 6
deteriora-
tion 16 3 13 4 12
T 0 t a 1 23 10 13 10 13

infarcts p < 0.01

atrophy p < 0. 05

631
between the clinical appearance of a new stroke and CT
findings of an increased number of infarcts. Some of the
infarcts however, appear as clinically silent strokes.

D I S C US S I 0 N
The appearance of dementia in patients with cerebro-
vascular disease does not only depend on the total loss
of parenchyma but also on the distribution and location
of infarcts (Ladurner et al. 1982). Sometimes, as in
bilateral thalamic infarction, the localisation is
obviously more important than the atrophy. In addition,
there is the problem of the mixed forms of Alzheimer and
vascular dementia which are clinically and by CT
usually diagnosed as vascular dementia. The clinical
outcome of these patients with a combination of diseases
(mixed Alzheimer and vascular) and the CT development
might be different from typical vascular dementia or
Alzheimer diseases which can be separated clinically
well (Hachinsky 1965, Harrison et al. 1979, Ladurner
et al. 1981).
It could be imagined that the progression of the de-
mentia and the atrophy without clinical and CT signs of
new strokes is mainly present in the mixed group giving
some possibilities of clinical differential diagnosis
of mixed Alzheimer and vascular versus pure vascular
dementia. This problem is not only of theoretical
interest but also of clinical importance for thera-
peutical trials on vascular dementia. Despite these
considerations of mixed Alzheimer and vascular demen-
tia the whole clinically defined group of vascular de-
mentia shows a good correlation between clinical and
morphological findings and the vascuiar disease for
the long term prognosis.

S U MMA R Y
The clinfcal findings show that the frequency of
strokes is less important for the appearance of demen-
tia than bilateral neurological signs or affection of
the dominant hemisphere. The CT findings are in agree-
ment with this.
Important factors for a poor prognosis are the appea-
rance of confusional states whilst the neurological
symptoms are of no significance.

632
 Important factors for a poor prognosis are the appea-
rance of confusional states whilst the neurological symp-
toms are of no significance.
The CT shows a good correlation between loss of sub-
stance and new infarcts with progression of dementia.

REFERENCES
Brae, Ga., AKHTAR Gr. (1976): Neurologic disorders in
the elderly at home . .T. Neural. 39: 362-366
Hachinski, V.C., Iliff, L.D., Du Boulay, G.H. et al.
(1975): Cerebral Blood Flow in dementia.
Arch. Neural. 32: 632-637
Harrison, M.J~., Thomas, D.J., Du Boulay, G., Marshall
J. (1979): Multiinfarct Dementia, J. Neurol.Sc.
40: 97-103
Kay D.W.K., Foster. E.M., Me Kennie, A.A., Roth, M.
(1970): Mental illness and hospital usage in the
elderly. Compr. Psychiatr. 11 : 26-35
Ladurner,G. (1983): Die Bedeutung der apparativen Diag-
nostik bei der aetiologischen Zuordnung demen-
tieller Prozesse. Nervenarzt 54: 171-180
Ladurner, G., Bertha G., Pieringer, W., Lytwin, H.,
Lechner, H. (1981): Klinische Untersuchungs-
kriterien bei vaskularer und primar degenera-
tiver Demenz. Nervenarzt 52, 401-404
Ladurner, G., Iliff, L.D., Sager, W.D., Lechner H.,
(1982): A clinical approach to vascular (Insult-
infarct) dementia. In: Hoyers S. (Ed.) Exp. Brain
Res., Supp. 5, The aging Brain Springer Berlin, ·
Heidelberg, N.Y. pp. 246-253
Ladurner, G., Bertha, G., Schneider, G., (1984): Clini-
cal and CT findings in the appearance and prog-
nosis of vascular (Multiinfarct) Dementia.
Egyptian J. Neural. and Psvchiat., in print
Lechner, H., Ladurner, G., Bertha G: 11982): Die Epi-
derminologie,Klinik und Langzeitprognose der vas-
kularen (Multiinfarkt) Demenz. Wr. Klin. Wsch. 94,
Suppl. 137, 13-16
Tower, B.A. (1978): Alzheimer disease - senile dementia
and related disorders. In: Katzman, R., Terry,
R.D., Bick K.L. (Eds.) Al~heimer's disease. Kaven
New York

633
COGNITIVE AND CONATIVE FUNCTION CHANGES
DEPEND! NG ON CVA AND TIA

A. Sever, A. Me sec and M. Cuk

Clinic Centre -Neurologic Clinic, Ljubljana

University of Edvard Kardelj, Ljubljana

INTRODUCTION
Cerebrovascular attack (CVA) is not only one of the leading causes
of death, but also a disabling disease which is becoming increasingly
frequent in the younger age groups. Thus, it affects people in the pro-
ductive period of life, when rehabilitation and vocational reorientation
constitute a most crucial issue. This adds to the importance of CVA
prevention, which is very difficult to carry out. For this purpose it is
necessary to identify all factors which might contribute to the develop-
ment of transitory ischaemic attacks (TIA) and CVA.

Many authors (G. Demeurisse et al. 1980, G. Tognola and L.A.

Vignolo 1980, K. York Haaland 1981 and others) are engaged in study-
ing the nature of the neuropsychological deficit produced by TIA, where
it is partially reversible, as well as by CVA, where it represents,
along with the motor deficit, the chief cause of invalidism. In addition
to physiological factors, also the personality and lifestyle are liekly
to play a role in CVA (Akiskal 1975). Therefore increasing attention
is being devoted to the personality of these patients and to their ways of
managing the essential spheres of life. Various authors (W. D. Gentry
et al 1979, R. Carasso 1981) speak of the "pressured" pattern of behavi-
our, marked by a more or less constant pressure to keep busy, a
feeling of being overburdened, the experience of an emotional pressure,
an increasing awareness of being unable to cope with the day-to-day
tasks and a persistent desire for advancement and recognition (Adler,
Mac Ritchie, Engel 1971). In the language of psychodynamics, these are
people for whom efficiency and esteem have an extraordinary and
exaggerated value.
635
 The results of different studies are controversial. Thus, Friedman
and Rosenman (1959) established a correlation between personality types
and the frequency of cardiovascular diseases. They found these diseases
to be more common among people exhibiting traits of Type A (coronary prone)
personality. R. Carasso (1981), in his study on 384 patients, noted that
the severity of CVA in patients without a history of cardiovascular prob-
lems, is influenced by two factors: life events and Type A personality.
A. Adler and associates (1971) established a similarity between the "pres-
sured" pattern of behaviour and Type A personality, suggesting an asso-
ciation between such behaviour and vulnerability to vascular disease ra-
ther than any specific symptomatic manifestation.

The aim of our study was to examine the mode of thinking and the
psychodynamic characteristics of patients following CVA and TIA.

Procedure

We selected a series of 20 patients with m.ild CVA (16) and TIA (4)
from the Slovene population suffering from these conditions. Patients
with consciousness disorders, severe right-sided hemiparesis, aphasia,
severe cardiac involvement and cerebral hemorrhage, as well as patients
over 60 years of age were excluded from the study. The subjects were
hospitalized in the University Department of Neurology during the second
half of the year 1982.

Neurological diagnostic assessment:

All the patient were subjected to a clinical neurological examina-
tion, eye field examination and laboratory studies for determining meta-
bolic risk factors (blood sugar and serum lipid levels). Theresults of .
other investigations (EEG, scintigraphy, CT) were not considered oo muse
they were not carried out in all the subjects.
Clinical psychological examination:
The clinical psychological evaluation included the Test of Copying
and Reproducing from Memory a Complex Geometric Design developed
by A. Rey, the Test of Mnestic Functions, adapted after A. L. Christensen
and A. R. Luria, and the Rorschach Test.
The mean age of male subjects (13) was 47. 7 years.
The mean age of female subjects (7) was 53. 8 years.

636
Results

Table 1. Test of Copying and Reproducing from Memory a Complex

Geometric Design after A. Rey:

Method of Work Score

I. II. III. /IV. VI X Cl
A 55% 10% 20% 15% 23,53 11,18
B 25% 20% 25% 30% 9, 83 8,13

A: copying
B: reproduction from memory
I, II, III/IV, VI: method of work with respect to the mental approach
pattern

The results were evaluated by means of Osterreith's Scoring System.

For the copying task 55% oft he patients used a systematic method of work,
whereas for reproducing the design only one fourth of the subjects emplo-
yed such a method. This suggests a decline in mental planning under dilfi-
cult conditions. The mean score achived on the copying task was 23. 53
out of a total of 36 points. Considering the patients- age, this indicates a
satisfactory capacity for visual analysis and integration of complex visual
information. Thesignificantly lower scores for reproduction suggest an
impairment of the visual memory function.

Table 2. Test of Mnestic Function - learning ability

1 2 3 4 5 6

5% 60% 9,4% 60% 30% 15%

1: scope of short-term memory

2: % of patients who were unable to memorize the entire series after
10 trials
3: number of trials required to memorize the entire series
4: perseverations evident in the course of learning
5: stereotypes
6: decline in activity in the course of learning

The patients' scope of memory is reduced (5%). Perseverations,

stereotypies and excessive fatiguability are evident in the course of learning.
A large percentage of patients were unable to memorize the series (60%).

637
 Table 3. Test of Mnestic Function - retention

1 2 3 4 5 6 7 8
x 1,75 1,20 0,45 0,90 2,35 1,05 1,65 5,1
G' 1,12 1,06 0,60 1,17 1,09 0,94 0, 88 3,40

1: visual recognition 5: verbal retention

2: visuaf retention 6: heterogeneous interference
3: acoustic retention 7: homogeneous interference
4: kinaesthetic retention 8: retrieval of sentences

Regarding individual sensory modalities, acoustic (0. 45) and kinae-

sthetic (0. 90) retention were most severely impaired, while verbal retention
was best preserved, showing an average score of 2. 35 of the possible
3 points. Our results indicate a decline in verbal retention under difficult
conditions (heterogeneous and homogeneous interference). The patients
achieved a remarkably low score on sentences retrieval. On average they
were able to recall only 5. 1 of the 46 units.

Table 4. Rorschach Test

Ind. x ~ Ind. x ()'
R 17,5 8,80 FC 0,35 1,26
w 48,15 4,51 CF 1,30 1,56
W% 45,95 25,17 c ~ ~
W+% 59,25 21,74 F(C) ~ ~
D 8, 08 7, 94 ChF 0,45 0,89
DW ~ ~ FCh 0,35 0, 81
s 0,70 0,92 H+Hd 1,60 1,70
Dd 0,55 0,76 A% 54, 85 22,40
Do 0,10 0,31 v 3,60 1,64
F+% 53,7 21,17 V% 24,40 14,42
F-indef ~ ~ Orig(O) 0, 80 2,26
M 0, 9 1,26 Pers 10,30 11,73
Time/sec 669,35 360,96

Table 4a.
Ind Color- Dark Succ-sligt succ- succ- EB C-D
+
Shock Shock undisc. disc. undisc. koart C-D
% 30 35 55 45 ~ 90 60

638
 Table 5. Clinical and laboratory data
Vis. Field Up. Extr. L. Extr. Sens. Atax. Vert. Fdi. RR BS S. LIP
R L R L R L R L
1 100% 90% 40 65 60 65 65 75 95 95 65 40 65 80
2 ~ ~ 60 35 35 35 35 15 5 5 35 40 35 20
3 ~ 10% ~ ~ 5 ~ ~ 10 ~ ~ ~ 20 ~ ~

1: no disorder
2: mild disorder
3: severe disorder

Intellectual functioning is at a low average level (R, W, W%, W+%, F,

M, A, V). The thinking is dull, less disciplined, showing a tendency to
perseverations and obvious disorders of mental concentration (Succ, F+%).
The reality control is still appropriate. The indicators (balance, colour
components) suggest a restricted emotional responsiveness, found in dull
and pedantic persons. Neurotic fears and environment-inducted dysphoria
(ChF) are present. The psychosocial contact is only formally adjusted
(V, V%, D, FC, H+Hd, EB, M).

Discussion
In the field of intellectual functioning, our patients exhibited a low
average level of intellectual performance with signs of psychological dys-
function of organic etiology at the time of testing. Our results support the
findings of those studies which demonstrate that the CNS responds to exten.,
sive lesions with a diffuse dysfunction. The patients' emotional and social
adjustment is only formal. They are emotionally dull and their responsive-
ness is reduced. They have great ambitions, which are out of proportion
with their ability for constructive assertion. They are meticulous and have
a remarkable capacity for reproduction, but they are intellectually insuf-
fi.;iently equipped to attain their aspirations. Thus, they are unable to ex-
perience self-assertion, and this leads them to frustration, because they
undertake tasks which they are not. equal to. This mode of life diminishes
their self-esteem and gives rise to severe neurotic fears and insecurity,
which cannot be neutralized by the defence mechanism of rationalization.
They are emotionally less involved in their psychosocial contacts, which
are thus often substitutional in nature.
Similar results were obtained by Gianturco in a group of patients with
a history of CVA and coronary disease. They exhibited great ambition and
perfectionism, which were the dominant traits of their "pressured" pattern
of behavrour. Gianturco also suggested that such a pattern of behaviour may
lead, through mechanisms related to elevation of lipid levels in serum,
accelerated coagulation or increased secretion of catecholamines, to the

639
development of vascular disease (notably coronary thrombosis). Our seri-
es does not support this contention. Hypertension and elevated sugar levels
in blood were quite frequent in our group of patients, but elevated lipid le-
vels in serum were significantly less pr valent. The problem of aggressi -
veness and ambition was emphasized also by Rolf Adl er and associates
(1971), who noted a tendency to consciously control the expression of aggres-
siveness in 87% of patients whit occlusive cerebral disease. We believe,
however, that in our group these processes occurred for the most part at
a subconscious level.
REFERENCES
Adler, R. , Mac Ritchie, K. , Engel, G. L. , Psychologic Processes and
~schemic Stroke (Occlusive Cerebrovascular Disease), Psvchosom. Med~ ,
33:1, 1971, 1-29
Akiskal, H. , Overview of recent research in depression, Archives of
General psvchiatrv, 32:1975, 285-293
Carasso, R. , Persona:ity type, life events and sudden cerebrovascular
attack, Intern. J. Neuroscience. 14:1981, 223-225
Caramazo, A., Sentence memory in aphasia, Neuropsychol., 16:6, 1978,
661 - 670
Demeurisse, G. , Quantitative evaluation of aphasia resulting from a
cerebral vascular accident, Neuropsychol, 17:1, 1979, 55-66
Gianturco, D. T., Breslin, M.S., Heyman, A., Gentry, W. D., Jenkins, C. D.,
Kaplan, B. , : Personality patterns and Life stress in Ischemic Cerbro-
vascular Disease. Psychiatric Findings, Stroke, 5:4, 1974, 453-460
Doyle, G. W., Jenkins, C. D., Kaplan, H. B., Heyman, A., Breslin, S.M.,
Gianturco, T. D., Type A behaviour pattern and ischemic cerebrovascular
disease, Heart & Lung, 8:6, 1979, 1113 - 1116
Mack, L., Levine, R.N., The basis of visual constructional disability in
patients with enilateral cerebral lesions, Cortex, 17:, 1981, 515 - 532

640
DEMENTIA AND ITS RELATION TO CEREBROVASCULAR DISEASE

s. Hoyer

Dept. of Pathochemistry and General

Neurochemistry, University of Heidelberg
D-6900 Heidelberg

According to Mayer-Gross, Slater and Roth (1), de-

mentia is based on a global deterioration of higher mental
functions and intellectual capacities. It may be defined
as a global disturbance of mental functioning in its cog-
nitive, intellectual and emotional aspects. McHugh and
Folstein (2) defined dementia as a deterioration of cogni-
tive functions including memory, abstract reasoning,
attention, language and perception without prominent
changes in consciousness. Disturbances in mood with an-
xiety and depression, in affectiveness, paranbld symptoms,
delusions, hallucinations, personality changes and cata-
strophic reactions were also described to be associated
with dementia. For clinical and pathophysiological reasons,
primary and secondary forms of dementia should necessarily
be separated. According to Sir Martin Roth (3), secondary
dementia symptoms will have been of relatively short
duration with rather abrupt onset. Their intellectual im-
pairment may be patchy and inconsistent and may show
marked fluctuations in severity as related to the severity
of the extracerebral disease. Clouding and confusion would
seem to be typical in secondary rather than primary de-
mentia. In this paper, only secondary dementias due to
circulatory disorders will be discussed.

Primary dementias may be compartmentalized into de-

generative and vascular forms. Degenerative dementia is
mostly represented by dementia of Alzheimer type which
predominates with 60 to 70% (4) , and which does not dis-
play any primary disturbances in the cerebral blood
vessels. Therefore, this dementia type will not be dis-

641
cussed further here. In this context, only primary de-
mentia of vascular type will be considered in more detail
(see below) .

It has been well documented in numerous investiga-

tions that, irrespective of this cause, dementia is clo-
sely related to changes in brain blood flow and oxidative
metabolism. It should be born in mind, however, that in
contrast to brain infarction, dementia - if at all - may
originate from chronic rather than acute circulatory dis-
orders. Such disorders may include chronic cardiac in-
sufficiency, general arteriosclerosis, arterial hypo- und
hypertension.

A reduction in cardiac output does not necessarily

lead to disturbances in brain blood flow and metabolism
and thus to dementia when there is no heart insufficiency.
If the latter exists then CBF and CMR-oxygen are found
to be decreased with symptoms of secondary dementia pre-
sent ( 5) •

Adam-Stokes syndrome with extreme bradycardia re-

sults in a fall of CBF but not in CMR-oxygen which has
found to be in a normal range (6). Disturbances in CBF
could thus have to be primarily expected before chang~s
in brain metabolism would occur.

On the other hand, in cardiac failure causing marked

secondary dementia symptoms, a mere decrease in CBF and
an unchanged CMR-oxygen was described (7).

When mean arterial blood pressure changes to arter-

ial hypo- or hypertension, an impairment of autoregula-
tion of cerebral blood flow may be impending. As a cause
of brain dysfunction in the elderly, a failure of auto-
regulation of cerebral blood flow has been reported. In
patients with postural hypotension, mean arterial blood
pressure fell 9 to 33 mmHg producing a reduction in CBF
by 32 to 67% (8). This would mean that the critical
treshold for an abolished autoregulation of CBF might be
shifted to higher mean arterial blood pressures than 50
to 70 mmHg.

Arterial hypertension is known to be an important

and top risk factor for cerebral arteriosclerosis and
stroke (9, 10). It may not implicitly vary brain blood
flow and metabolism and thus produce dementia (11). In
chronic arterial hypertension, autoregulation of cere-
bral blood flow may be assumed to be maintained. However,
the critical treshold for its abolition is shifted to

642
the right and similarly reset at a higher level of arter-
ial blood pressure as just mentioned, e.g. at 90, 110 or
more mmHg as compared to the lower normal treshold of 50
to 70 mmHg. In arterial hypertension, the upper treshold
of autoregulation is shifted to around 150 to 200 mmHg
(12). Hypertensive encephalopathy, i.e. the occurrence of
dementia symptoms, may be related to the so-called break-
through of autoregulation. Under these conditions, the
arterial vessels are maximally dilated and cerebral blood
flow is abnormally increased. No vasospasm occurs (13).

The overdilation may produce a brain edema (cyto-

toxic and vasogenic, respectively as deleterious conse-
quence of severe arterial hypertension. If the arterial
hypertension has become chronic, a reduced cerebral per-
fusion along with a decreased CMR-oxygen will produce
secondary dementia syptoms (14). In elderly patients with
arteriosclerotic changes such as elongation and tortuo-
sity of the aorta, calcification of the aorta, absent
peripheral pulses, or carotid siphon or retinal arterio-
sclerosis, brain blood flow and oxygen consumption were
found to be reduced when dementia symptoms were present
( 1 5) •

From these findings, one may tentatively conclude

that besides other extracerebral diseases, secondary de-
mentias may be caused by distinct circulatory disorders
provided they are long lasting enough to produce per-
manent disturbances in brain blood flow and metabolism.

Primary dementia of vascular type (DVT) amounts to

20 to 30% of major primary dementias (4). Clinically, it
is dominated by a sudden appearance of rather psycholo-
gical deficits such as emotional lability which may be
accompanied by depressive symptoms which mostly occur
later in the course of this dementia type. Intelligence
and personality are more commonly well-preserved over a
longer period of DVT, the course of which is mostly
intermittent and fluctuating ("stepwise downhill") (16).
These psychiatric symptoms in DVT are supposed to be
associated with morphological disturbances which are
characterized by multiple small infarcts scattered over
brain cortex and white matter (17). It may be hypothesi-
zed that the underlying process of this circumscribed
cell loss with gliosis may be related to disturbances in
cerebral microcirculation which lead to ischemic/anoxic
tissue lesions. These multiple small infarcts along with
distinct changes in oxidative brain metabolism (see
below) would primarily produce demetnia symptoms.

643
Moreover, the small infarcts may merge into larger ones
and this may be responsible for rather obscure neurolo-
gical deficits such as TIA or PRIND in the history along
with the pre-existing respective psychiatric symptoms. It
has as yet not become clear wether this type of primary
dementia and the multi infarct dementia (MID) character-
ized by large brain infarctions producing clear and more
distinct neurological deficits due to diseases of the
large brain arteries are brain diseases of the same no-
sological entity or wether they are different at least in
pathophysiological terms. At all events, the differences
in the clinical course would seem to be remarkable. In one
group, the brain disease starts with dementia symptoms and
were followed by more or less discrete neurological def-
icits and in another group, the neurological deficit is
first and dominates the clinical picture which is second-
arily completed by psychological deficits.

Cerebral blood flow does not uniformly vary in pri-

mary dementias, as was first demonstrated by O'Brien and
Mallett (18). They found a decreased cortical perfusion
rate in patients suffering from cerebrovascular disease
while in patients with primary neuronal degeneration, no
change of blood flow could be observed as compared to
dmentia-free controls. Based on the ischemic score out-
lined by Mayer- Grosset al (1), Hachinsky et al (19)
confirmed this pattern in CBF: a decrease in multi infarct
dementia, and normal flow rates in primary degenerative
dementia. Lower flow rates in the left cerebral hemis-
phere as compared to the right were described to be pre-
sent by Ladurner et al (20) whereas in primary degenrative
dementia, there was only a trend in this direction.
Different distribution curves of blood flow and oxygen
consumption of the brain could be found between DAT and
DVT: normal and slightly increased rates in DAT and re-
duced ones in DVT (21). However, these rather general
results were found not to be so characteristic for the
biological variations as could be demostrated in the
course of DVT.

In the beginning of DVT, CBF and CMR-oxygen were

found to be in a normal range, but CMR-glucose was ab-
normally enhanced. The disturbed balance between normal
cerebral oxygen consumption and abnormally increased
cerebral glucose uptake which runs opposite to DAT re-
sembles those changes which occur in acute anoxic brain
damage, e.g. after hypoxic hypoxia (22). In contrast to
hypoxia, no augmentation of cerebral blood flow and CMR-
lactate could be observed. The metabolic fate of the
"surplus" of glucose taken up by the brain is as yet un-

644
known. It may be speculated that the amount of glucose
not oxidized by the brain may be falsely metabolized and/
or abnormally stored in brain cells, i.e. in astrocytes
near the vessels as glucose polymers comparable to the
"glycogen" which was found by Klatzo et al (23) and Mo-
ssakowski et al (24) after acute experimental ischemia or
aspyxia. If a similar storage of glucose polymers in brain
cells occurs during the initial phase of DVT, it schould
be borne in mind that such a storage may lead to a di-
sturbance of glucose transport into the brain and to an
impairment of the metabolic pathway of glucose in the
brain. A decrease in CMR-glucose would have to be expected
as a consequence in the further course of DVT. Such a fall
in CMR-glucose along with reduced CBF and CMR-oxygen had
indeed beeb measured in the course of DVT. These biolo-
gical brain parameters approach to the same low functional
level as in DAT, thus demonstrating the reduced demands
of the diseased brain. No significant differences can be
observed between the two main dementia types with respect
to brain blood flow and oxidative metabolism in more ad-
vanced dementias (25, 26). Coupling between CBF, CMR-
oxygen and CMR-glucose in DVT is maintained between CBF
and CMR-oxygen on the one hand but not between the two
latter and CMR-glucose on the other hand. CMR-glucose is
abnormally enhanced in the onset of the disease. In
chronic DVT, however, this dissociation disappears.

It may thus be concluded that dementia symptoms pro-

duced by vascular disorders originate from different
extracerebral or obviously autochthonus brain disease. In
secondary dementias produced by cardio-vascular diseases,
brain blood flow and possibly CMR-oxygen are primarily
involved. In dementia of vascular type, primary metabolic
changes dominante. As a whole and as compared to dementia
of Alzheimer type, demetnias as related to cerebrovascular
disorders may represent only a minor part of mental dis-
ability.

REFERENCES

1. W. Mayer-Gross, E. Slater, M. Roth, CLinical Psychi-

atrie, 3rd ed. Bailliere, Tindall and Carssell,
London ( 1969)
2. P. R. McHugh and M. F. Folstein, Psychopathology of
dementia: Implications for neuropathology, in:Con-
genital and aquired cognitive disorders, R. Katz-
man, ed., Raven, New York (1979)
3. Sir Martin Roth, Diagnosis of senile and related forms
of dementia, in: Alzheiner's disease: Senile de-

645
 mentia and related disorders (Aging Vol.7), R.
Katzman, R. D. Terry, K. L. Bick, eds., Raven,
New York (1978)
4. B. E. Tomlinson, The structural and quantitative
aspects of the dementias, in: Biochemistry of de-
mentia, P. J. Roberts, ed., Wiley, Chichester
(1980)
5. u. Gottstein, A. Bernsmeier, H. Blamer, and w. Schim-
mler, Die zerebrale Haamodynamik bei Kranken mit
Mitralstenose und kombiniertem Nitralvitium, Klin.
Wschr. 38, 1025 (1960)
6. A. Bernsmeier, U. Gottstein, and W. Rudolph, Herz-
krankheiten als Ursache zerebraler Zirkulations-
storungen, Dtsch. Med. Wschr. 87, 16 (1962)
7. s. Eisenberg and W. Sensenbach, J. Clin. Invest. 35,
700 (1956)
8. L. Wollner, s. T. McCarthy, N. D. W. Soper, and D. J.
Macy, Failure of cerebral autoregulation as a cause
of brain dysfunction in the elderly, Br. Med. J.
I., 1117 (1979)
9. W. B. Kannel, P. A. Wolf, J. Verter, and P. M. Me Na-
mara, Epidemiologic assessment of the role of blood
pressure in stroke. The Framingham study, J. Amer.
Med. Ass. 214, 301 (1970)
10. u. Gottstein, Zur Pathogenese der Hirnischamie, unter
besonderer Berlicksichtigung der Risikofaktoren,
Internist 17, 1 (1976)
11 S. s. Kety, J. H. Hafkenschiel, W. A. Jeffers, J. H.
Leopold, and H. A. Shenkin, Blood flow, vascular
resistence and oxygen consumption of the brain in
essential hypertension. J. Clin. Invest. 27, 511
(1948)
12. s. Strandgaard, J. Olesen, E. Skinh¢j, and N. A.
Lassen, Autoregulation of brain circulation in
severe arterial hypertension. Brit. Med. J. I,
507 (1973)
13. N. A. Lassen and A. Agnoli, Upper limit of autore-
gulation of cerebral blood flow: on the pathoge-
nesis of hypertensive encephalopathy, Scand. J.
Clin. Lab. Invest. 30, 113 (1972)
14. H. A. Shenkin, P. Novack, B. Gohnboff, A.M. Soffe,
and L. Bortin, J. Clin. Invest. 32, 459 (1953)
15. R. N. Butler, D. K. Dastur, and S. Perlin, Relation-
ships of senile manifestations and chronic brain
syndromes to cerebral circulation and metabolism,
J. Psychiat. Res. 3, 229 (1965)
16. G. Blessed, Clinical aspects of senile dementia, in:
Biochemistry of dementia, P. J. Roberts, ed.,
Wiley, Chichester (1980)

646
17. J. A. N. Corsellis, The pathology of dementia, Br.
J. Hosp. Med. 3, 695 (1969)
18. M. D. O'Brien and B. L. Mallett, Cerebral cortex per-
fusion rates in dementia, J. Neurol. Neurosurg.
Psychiat. 33 497 (1970)
19. V. C. Hachinski, L. D. Iliff, E. Zilkha, G. H. Du
Boulay, V. L. McAllister, J. Marshall, R. W. Ross-
Russell, and L. Symon; Cerebral blood flow in de-
mentia, Arch. Neurol. 32, 632 (1975)
20. G. Ladurner, E. 0. Ott, P. J. Perry, P. Stix, H.
Schreyer, F. Wiedner, and H. Lechner, Bilateral
measurement of regional cerebral blood flow in
dementia, in:Cerebral Vascular Disease, eds.: J.
S. Meyer, H. Lechner, M. Reivich, Excerpta Medica,
Amsterdam, Oxford (1977)
21. S. Hoyer, K. Oesterreich, F. Weinhardt, and G. Kruger,
Veranderungen von Durchblutung und oxidativem
Stoffwechsel des Gehirns bei Patienten mit einer
Demenz, J. Neurol. 210, 227 (1975)
22. J. Hamer, K. Wiedemann, H. Berlet, F. Weinhardt, and
S. Hoyer, Cerebral glucose and energy metabolism,
cerebral oxygen consumption and blood flow in
arterial hypoxemia, Acta Neurochir. 44, 151 (1978)
23. J. Klatzo, E. Fargas-Bergeton, L. Guth, J. Miguel,
and Y. Olsson, Some morphological and biochemical
apects of abnormal glycogen accumulation in the
glia, in Proceed. VIth Int. Congr. Neuropathol.,
Masson, Paris (1970)
24. M. J. Mossakowski, D. M. Long, R. E. Myers, H. Rod-
riguez de Curet, and J. Klatzo, Early histo-chemi-
cal changes in perinatal asphyxia, J. Neuropathol.
Exp. Neurol. 27, 500 (1968)
25. s. Hoyer, Factors influencing cerebral blood flow,
CMR-oxygen and CMR-glucose in dementia patients,
in: Biochemsitry of dementia, P. J. Roberts, ed.,
Wiley, Chichester (1980)
26. W. D. Obrist, Noninvasive Studies of cerebral blood
flow in aging and dementia, in: Alzheimer's dis-
ease: Senile dementia and related disorders, R.
Katzman, R. D. Terry, K. L. Bick, eds., Raven,
New York (1978)

647
ORGANIC BRAIN SYNDROMES IN CEREBROVASCULAR DISEASE - EARLY AND
MILD CASES WITHOUT FOCAL NEUROLOGICAL SYMPTOMS AND SIGNS -

Gerd KrUger
Central Institute of Mental Health
J 5, P.O.Box 5970
D-6800 Mannheim 1 FRG

INTRODUCTION
The clinical differentiation was studied between organic
brain syndromes associated with cerebrovascular disease whose
clinical conditions were rather like cerebral arteriosclerosis
with small commonly unrecognized apoplexies (Alzheimer, 1899,
1902; Avarez, 1946) than stroke (Goodstein, 1983; Hachinski et al .,
1974) and those with other etiologic factors in young, middle-
aged and elderly patients (KrUger et al., 1981). The purpose was
to find diagnostically relevant features, especially in terms of
psychopathology.
Diagnostic differentiation between cerebrovascular and other
conditions is likely to prove impossible in early, mild, poten-
tially reversible and most eminently treatable cases unless clear-
cut focal neurological smyptoms and signs as stroke phenomena are
present. Diffuse (widespread) global involvement by multiple mini-
mal infarcts in brain may give rise to psychological dysfunction
through a variable constellation of pathogenic mechanisms.
Differential classification in psychiatry is implicitly based
on two distinct principles, the descriptive and the etiological
one (Roth, 1978). The first line of demarcation is that between
disorders exhibiting "functional" or "organic" features in terms
of psychopathology. The second line is the presence or absence
of brain disease or cerebral dysfunction arising from metabolic
disturbance. These two sets of criteria, the symptomatological
and the etiological ones, should be applied independently.

649
PATIENTS AND METHODS
155 consecutively admitted psychiatric patients (mean age 54
years, SEM: 1.25) with organic mental disorders of (alcohol) toxic,
degenerative, vascular and other etiology were studied. The clini-
cal assessment including mental state, neurologic and physical
examination was supported by electroencephalographic, neuroradio-
logic and other techniques to leave out cases of certain issue:
(1) "brain damage" due to gross, unilaterally and bilaterally focal
lesions by trauma, stroke, inflammatory disease etc.:;
(2) "treatable cases" obviously due to metabolic, toxic or in-
fectious conditions and cases with the diagnosis of stupor and
coma;
(3) seizure related conditions and delirium tremens in alcohol and
other drug withdrawal;
(4) advanced stages and severe degrees of Alzheimer's disease,
senile and multi-infarct dementia.
Psychopathology, autonomic and neurologic symptoms and signs were
evaluated by the AMDP-rating scale system. Cluster, discriminant
function and principal component analyses were performed on the
documented data (KrUger and Haubitz, 1980).

NUMBER

22
45
20

40

18

35
16

30
14

12
25

20
10
-

15

10

CEREBROVASCUlAR
l
PATIENTS (N•47l
1 " PERSONALITY SYNDROME
2 = AFFECTIVE SYNDROME
3 = DEMENTIA
4 = DELUSIONAL SYNDROME
5 = WITHDRAWAL SYNDROME
6 " DELIRIUM

Figure 1

650
RESULTS
Six organic brain syndromes were found: organic personality
syndrome (N=41), organic affective syndrome (N=42), dementia (N=
34), organic delusional syndrome (N=B), withdrawal syndrome (N=11)
and delirium (N=11). Eight patients were individual cases and
could not be classified. There was good separation for the diag-
nostic groups of patients found by discriminant function analysis.
Etiologic factors were chronic alcoholism (N=69), cerebrovascular
disease of multi-infarct type with diffuse (widespread, global)
patchy involvement of subcortical and cortical structures (N=47),
alcohol and barbiturate addiction, posttraumatic and postin-
fectious cerebral states, heavy metal intoxications (N=12) and
unknown "degenerative" causes (Alzheimer's) (N=19).
Figure 1 shows the distribution of t~e organic brain syndromes
within the cerebrovascular group of patients (N=47). Patients with
affective syndrome (N=23) predominate others with dementia (N=10),
delirium (N=7), organic personality (N=4) and delusional syndrome
(N=3). The distribution of the syndromes within the cerebrovascu-
lar group of patients is related to those in chronic alcoholism,
of unknown cause (?Alzheimer's) and other etiologic factors in
Figure 2.

NUMBER
28
40
26

24 ll

22

JO
20

18
21
16

14
20

12

10 II

10

N=147

1 = PERSONALITY SYNDR~E
0 OTHERS 2 ~ AFFECTIVE SYNDROME
3 = DEMENTIA
~ ~~~~~~R~~~~~ or 4 ~ DELUSIONAL SYNDROME
5 ~ WITHDRAWAL SYNDRO!o\E
[Ill ~~~~:~~VASCULAR 6 = DELIRIUM

~ CHRONIC ALCOHOLISM

Figure 2

651
They are compared to one another in per cent changes of the
syndromes found: delirium and affective syndrome in cerebrovascu-
lar disease are predominantly specified by their frequent
occurrence in this sample (see Figure 3).
Figure 4 shows the distribution of patients on both axes:
syndromes (horizontal plane) and etiology (vertical plane). A group
of patients was specified by the three dimensions of symptomatology,
etiology and frequency that proved to be different from the rest.
With other words, the specificity was valid in terms of N•s, syn-
drome and etiology within the given sample. See also Figure 4 for
N1 s in alcoholic personality (N=31), dementia (N=17) on the one
side and affective syndrome in cerebrovascular disease (N=23) on
the other.

100
95

90
85

80

75

70
65

60

55

50

45

40

35
30

20

15

10

N=147 t = PERSor.ALITY SYNDROME

2 = AFFECTIVE SYNDR()IIE
3 " DEMENTIA D OTHERS
4 • DELUSIONAL SYNDROME
5 = WITHDRAWAL SYNDROME ~ ~~~~~:R~l~~~ OF
6 = DELIRIUM
liD ~i~~:~~VASCULAR

II ~~~:b~tSM
Figure 3

652
 (N=41) (N=42) ( N=34) (N=8) (N=11) (N=11)
PERSONAL! TV AFFECTIVE DEMENTIA DELUSIONAL WITHDRAWAL DELIRIUM
SYNDROME SYNDROME SYNDROME SYNDROME

G
N=147

CHRONIC
ALCOHOLISM
(N=69) 0 8 0 0 0
CEREBROVASCULAR
DISEASE
(N=47)
0 G0 0 0
DETERIORATION OF
UNKNOWN CAUSE
(ALZHEIMER$)
(N=19)
0 0 0 0

OTHERS
(N=12)
0 0 0 0 0 0

Figure 4

Criteria for affective categories (''depression"), hypochondriasis

and for affective changes combined with disorders of motivation
and impulse control ("affective deficiency") are shown as discri-
minators in psychopathology in the following table 1:
6 of following AMDP-items: 3 of the following AMDP-items:
PERPLEXED RESTRICTED THINKING
DEPRESSED/SAD HYPOCHONDRIASIS
HOPELESS/DESPERATE FULL OF COMPLAINTS
ANXIOUS/FEARFUL
BETTER IN THE EVENINGS
FEELING OF BEING ILL
7 of the following AMDP-items:
BLUNTED AFFECT
ELATED/EUPHORIC
MOODY/IRRITABLE/DYSPHORIC
EMOTIONAL RIGIDITY
LACK OF FEELING OF BEING ILL
LACK OF INSIGHT INTO THE ILLNESS

653
 Table 2
Discriminatory function of hypochondriasis and affective changesl
affective syndrome personality syndrome dementia in chronic
in cerebrovascular disease vs. in chronic alcoholics vs. alcoholics
(N=23) (N=31) (N=17)

hypochondriasis 1.246 .376 .118

p(u)=.00057 p(u)=.00010
*** ***

depression .935 .183 .343

p(u)= .00~~~ p(u)=.00098
***

affective .210 1.452 2.294

deficiency p(u)=.OOOOO
p(u)=.OO~~~
***

1 typical for the class of patients with affective, personality and dementia
categories, regardless of the given etiology, containing typical specimens of all these patients

Discriminatory power is validated for hypochondriasis and

affective categories in organic affective syndrome of cerebro-
vascular patients and personality syndrome and dementia in chronic
alcoholics as shown in Table 2.

CONCLUSION
There were indications from earlier work that a sharper dis-
crimination and an improved diagnostic scale are to be achieved
with the aid of discriminant function techniques in 11 multi-infarct
dementia 11 (Roth, 1981). We found differential quantitative weights
due to the presence of hypochondriasis and depressive mood changes
and the absence of euphoria, irritability and emotional rigidity
in early and mild cases of cerebrovascular disease. The diagnostic
validity should be proved by further longitudinal studies.

REFERENCES
Alzheimer, A., 1899, Beitrag zur pathologischen Anatomie der Seelen-
storungen des Greisenalters, Neural. Ce~tralbl. 18, 95-96.
Alzheimer,A., 1902, Di._e Seelenstorungen auf arteriosklerotischer
Grundlage, Allg. Z. Psychiatr. 59, 695.
Alvarez, W.C., 1946, Cerebral arteriosclerosis with small common-
ly unrecognized apoplexies, Geriatrics 1, 189-216.

654
Goodstein, R.K., 1983, Overview: Cerebrovascular Accident and the
Hospitalized Elderly - A Multidimensional Clinical Problem,
Am J Psychiatry 140:2, 141-147.
Hachinksi, V.C., Lassen, N.A., Marshall, J., 1974, Multi-infarct
dementia, Lancet ii, 207-210.
KrUger, G., Haubitz, I., 1980, Classification of Organic Brain
Syndromes by Cluster Analysis, Arch.Psychiat.Nervenkr. 228,
299-315.
KrUger, G., Haubitz, I., Hoyer, S., 1982, Organic Psychiatric
Disease in Young Adults and Aged Patients, Experimental Brain
Research, Suppl. 5, Springer-Verlag Berlin, Heidelberg.
Rotn; M.; 1978, Psychiatric Diagnosis in Clinical and Scientific
Settings, in: Psychiatric Diagnosis: Exploration of Biological
Predictors, Akiskal, H.S., Webb, W.L. (eds.), New York.
Roth, M., 1981, The Diagnosis of Dementia in Late and Middle Life,
in: The Epidemiology of Dementia, Mortimer, J.A., Schumann, L.M.
\eds.), Oxford University Press, New York Oxford.

655
THE DIFFERENTIATION OF MULTI-INFARCT AND ALZHEIMER DEMENTIA

John Marshall
Institute of Neurology
National Hospital for Nervous Diseases
Queen Square, London WClN 3BG, England

The great decline in mortality from infective disease in the

Western industrialized countries as a result of the introduction
of antibiotics has greatly changed the age structure of the
population. About 14 per cent of people are now aged 65 years or
more. They become heir to the diseases of later life among which
is numbered dementia. It is estimated that about 1 in 20 of those
aged 65 years or more will become demented; among those aged over
80 years - the old old of geriatrics - the figure is nearer to
1 in 5. Dementia is therefore a new type of epidemic, not in the
sense of being passed from person to person but in the sense of
large numbers being afflicted at any one time.
There are many causes of dementia but two,Alzheimer's disease
and multi-infarct dementia account for the majority of cases. Our
understanding of both these conditions has advanced in recent years.
Vascular dementia has long been recognized but equally long
misunderstood. It was frequently referred to as cerebral arterio-
sclerosis, the assumption being that arteriosclerotic changes in
vessel walls prevented delivery of an adequate supply of blood
for the metabolic needs of the brain. This led to ischaemic cell
damage and so to dementia.
This has readily been shown not to be the case. Administration
of C02 to patients with vascular dementia causes a marked increase
in cerebral blood flow (CBF). The arteries are capable of supplying
more blood; the problem is there is not the demand for it. It is
true that C02 reactivity is not so great in the old as in the yoyng
but· is nevertheless quite sufficient to provide a great increase in
CBF when thus required.

657
 The important feature of vascular dementia is not the changes
in the arterial walls but the multiple small infarcts which are
found scattered throughout the brain - hence the term multi-
infarct dementia.l The infarcts arise mainly as a result of
changes in the small penetrating arteries of the brain produced by
hypertension. These changes are variously referred to as lipo-
hyalinosis or fibrinoid necrosis. They lead to occlusion of the
vessels and production of small infarcts.
More recently it has been appreciated that this condition can
arise in non-hypertensive subjects from multiple small emboli.
The source of the emboli is not always clear but is probably
small atheromatous plaques in the proximal parts of small cerebral
arteries. Thrombus forms on these and embolizes distally, much in
the same way as emboli from atheromatous lesions in larger arteries
such as the carotid artery embolize to the ophthalmic or middle
cerebral arteries.
There has been confusion also about the other major cause of
dementia - Alzheimer's dementia. This term was formerly reserved
for a certain kind of dementia arising in the presenium characterized
by the presence of senile plaques, neurofibrillary tangles and a
granular vacuolation of cells in the hippocampus. Dementia develop-
ing after the presenium - if not labelled vascular - was referred to
as senile dementia. The assumption in this case was that the
condition differed in some ill defined way from Alzheimer's disease.
It is now recognized that nonvascular dementia of the senium and
Alzheimer's disease of the presenium are the same condition showing
the same pathological features and differing only in the age of
onset.
There are therefore two major causes of dementia, multi-infarct
dementia and Alzheimer's disease. The former accounts for about
one-third of cases, the latter for about two-thirds. There is some
small degree of overlap - some elderly patients suffering from both
conditions - but the two conditions are quite distinct and the
majority of demented patients have one or the other.
Although there is as yet no specific treatment for either
condition it is important that every effort be made to distinguish
between them if progress is to be made. Fortunately the distinction
can be made clinically with a fairly high degree of reliability.
This is most easily done by seeking features of multi-infarct
dementia which if present justify this diagnosis. If they are
absent a diagnosis of Alzheimer's disease may be made.
The reason for this approach lies in the pathology described
above, namely the presence of small infarcts. These develop
abruptly though not all at the one time, hence give rise to a
clinical picture of a disease of abrupt onset, proceeding in a
step-wise fashion. The common sites of the infarcts - namely in

658
the territory of the deep penetrating areas of the brain
disrupting fronto-pontine connections - also gives rise to certain
features such as emotional incontinence and confusion. These
coupled with evidence of vascular disease elsewhere in the body
add up to a fairly typical picture.
Features of this kind have been assembled into what has come
to be known as the Hachinski score.2 A standard list of features
suggestive of multi-infarct dementia was drawn up, each item being
given a weighted score. A high score pointed to multi-infarct
dementia and a low score - involving as it did an absence of
vascular features - pointed to Alzheimer's disease. This score has
been found to be a useful diagnostic tool and has received clinico-
pathological validation.3
The advent of the CT scan has provided a further way in which
MID and Alzheimer's can be distinguished in life. Smaller lacunes
are beyond the resolution of the scan but larger lesions can be
seen and point to the diagnosis of MID. Enlargement of the
ventricles and cortical sulci are common to both conditions.
In similar fashion evidence of focal disturbance in the EEG
has been found to be more common in patients with MID than in
Alzheimers. This again reflects the focal nature of the lesions
in the former condition.
The non invasive measurement of CBF by a technique such as
intravenous 133xenon may also be of value. Both MID and Alzheimers
have reduced CBF, the degree of reduction being correlated with
the degree of dementia. However the degree of reduction is greater
in MID than in Alzheimers.
The two conditions can therefore be distinguished in life with
sufficient reliability to make it worth while. Hypertension is the
major factor in MID. Control of blood pressure before MID develops
is clearly called for. Whether reducing blood pressure once the
condition is established is of value is not certain but a knowledge
of the pathology suggests it may be helpful.
The deficiency in acetyl choline transferase now known to be
present in Alzheimer's disease opens up therapeutic possibilities.
Even though the administration of choline currently being subjected
to controlled clinical trial may prove unhelpful, identification of
a biochemical deficit must give rise to hope of advance.
Any trial of therapy for either condition must be based on
firm diagnosis with, if necessary, exclusion of doubtful cases.
Without this there is always the risk of a potentially valuable
treatment being missed because trial groups contain cases other
than the one to be treated. This need not happen in relation to
MID and Alzheimers because careful clinical appraisal together with
non invasive investigation can distinguish the two.

659
References
1. V.C. Hachinski, N.A. Lassen and J. Marshall, Multi-infarct
dementia A cause of mental deterioration in the elderly,
The Lancet 2: 207 (1974).
2. V.C. Hachinski, L.D. Iliff, E. Zilkha, G.H. du Boulay,
V.L. McAllister, J. Marshall, R.W. Ross Russell and L. Symon,
Cerebral blood flow in dementia, Arch. Neurol. 32: 632 (1975).
3. W.G. Rosen, R.D. Terry, P.A. Fuld, R. Katzman and A. Peck,
Pathological verification of ischemic score in differentiation
of dementias, Ann. Neurol .. 7: 486 (1980).

660
 DIFFE RENT IAL DIAGNOSIS BETWEEN

DEGENERATIVE AND VASCULAR DEMENTIA

carlo Loeb and Carlo Gand olfo

Depar tment of Neuro logy

Univ ersity of Genov a
Via De Toni, 5 - 16132 Genov a - Italy

KEYWORDS

Deme ntia - Mult i-inf arct Deme ntia - Senil e Deme

ntia
Alzh eime r's dise ase- Diffe renti al Diagn osis in
Deme ntia
CT in Deme ntia

INTRODUCTION

The two most frequ ently encou ntere d types of de-

ment ia are the senil e deme ntia of the Alzhe imer
type
(SDAT) , which accou nts for appro xima tely 50 %
of all pa-
tient s with deme ntia, and the so-ca lled mult i-inf
arct de-
ment ia (MID) , comp rising about 15-20 %of the
total .
The 15-20 % of deme nted patie nts have proba bly
disor ders (MIX) . Other less commo n disea ses accou both
nt for
the remai ning cases .

In the vast majo rity of cases (from 80 to 90 %)

,
the diffe renti al diagn osis is betwe en degen erativ
e senil e
deme ntia, mult i-inf arct deme ntia, and mixed forms
.
Hach inski et al. (1) selec ted 13 items as signs
of vas-
cular deme ntia. A point value was assig ned to
each fea-
ture, and the final summ ation of point s resul ted
in an
Ische mic Score (I.S. ).

In patie nts with deme ntia, a score of 4 or less

to a diagn osis of SDAT, while score s of 7 or highe point
r point
to a diagn osis of MID (2) (tabl e 1).

661
 Table 1. Hachinski's Ischemic Score

1. Abrupt onset 2
2. Stepwise deterioration 1
3. Fluctuation 2
4. Nocturnal confusion 1
5. Relative perservation of personality 1
6. Depression 1
7. Somatic complaints 1
8. Emotional lability 1
9. Hypertension 1
10. History of stroke 2
11. Focal symptoms 2
12. Focal signs 2
13. Other signs of arteriosclerosis 1

This paper aims at evaluating the accurancy of the

I.S. in a group of demented patients. The diagnosis of
dementia rested upon the obvious correlation between
history, clinical features, and the results of ancilla-
ry diagnostic procedures. In allocating each patient to
the SDAT or MID group, decisive stress was placed upon
the CT investigation.

MATERIAL AND METHODS

Out of 101 patients with dementia admitted to the

Department of Neurology at the University of Genoa from
January 1st, 1975 to February 1st, 1982, 94 were selec-
ted.

These were 51 men and 43 women with a mean age of

66 + 9. Seven patients with a dementia attributable to
causes other than SDAT, MID and MIX were excluded.

The patients entering this study (considered de-

mented on the basis of history, clinical features and
the results of the ancillary diagnosis procedures) were
subdivided, on the basis of CT features, into the follo-
wing subgroups: 1) CT-SDAT subgroup (65 cases), in which
the CT examination showed ventricular enlargement with
widening of the sulci. This subgroup of patients with
dementia without signs attributable to cerebral vascu-
lar involvement was considered as affected by SDAT; 2)
CT-MID subgroup (18 cases), in which CT showed the
existence of multiple areas of reduced density attri-
butable to ischemic lesions. 3) CT-VASC subgroup (11 ca-
ses), in which CT showed a single ischemic lesion and the

662
patient was considered as affected by a dementia of vas-
cular origin, possibly multi-infarct dementia.

RESULTS

There were no differences in age, sex, duration of

illness and general physical and laboratory examinations
among the three subgroups, with the exception of hyper-
tension, which prevailed in CT-MID with respect to CT-
SDAT (P < .05) and ECG alterations which prevailed in
CT-MID with respect to CT-SDAT (p < .05). The neurologi-
cal examination showed a significant frequency of focal
neurological signs in the CT-MID (p < .0005) and in the
CT-VASC (p < .001) subgroups, and of pseudobulbar signs
in the CT-MID (p < . 0005) and in the CT-VASC (p <.0005)
ones. The dementia score of Hachinski was significantly
lower in the CT-MID than in the CT-SDAT subgroups (p <
.025).

ISCHEMIC SCORE RATINGS

The total score in the three subgroups was as fol-

lows: i) the CT-SDAT (65 cases) subgroup showed an I.S.
equal or lower than 4 in 45 patients (69%), equal to
5-6 in 16 patients (25%) and equal or higher than 7 in
4 patients (6%); ii) the CT-MID (18 cases) subgroup had
an I.S. equal to 5-6 in 1 patient (6%), equal or higher
than 7 in 17 patients (96%); iii) the CT-VASC (11 cases)
subgroup showed in I.S. equal or lower than 4 in 2 cases
(18%), equal to 5-6 in 2 cases (18%), equal or higher
than 7 in 7 cases (63%).

In summary, nineteen patients (20%), of which 16 of

CT-SDAT, 1 of CT-MID and 2 of CT-VASC subgroups, had an
I.S. of 5-6 which we considered not discriminant to avoid
overlap of different types of dementia, thereby rending
the classification more flexible.

Moreover, 4 patients with CT-SDAT had an I.S. poin-

ting to a diagnosis of MID, and 3 patients with CT-VASC
had an I.S.< 4 pointing to a diagnosis of SDAT.

The statistical comparison of the various items of

I.S. among the three subgroups yielded significant dif-
ferences, since items 1 (abrupt onset), 10 (history of
stroke), 11 (focal neurological symptoms), 12 (focal
neurological signs), prevailed in CT-MID and CT-VASC
subgroups with respect to CT-SDAT subgroup (p <.0005).

663
COMMENT

The validation of I.S. by CT scan features was suc-

cessful, since 69% of patients with SDAT and 95% of pa-
tients with MID had an I.S. in agreement with the diagno-
sis established by the CT scan.

Nineteen patients with a non-discriminant score

(=5-6) and 9 patients with one hypodense lesion (CT-VASC
subgroup) had equivocal ischemic scores: the 19 cases
with a nondiscriminant score of 5-6, and the other 9 pa-
tients with score of different values, 2 cases pointing
to a SDAT and 7 cases to a MID diagnosis.

On the whole, 28 patients (30%) could not be con-

veniently classified according to the ischemic scale of
Hachinski.

Four features (i.e. abrupt onset, previous history

of stroke, focal symptoms and focal signs) were found
most relevant to a diagnosis of multi-infarct dementia.
We suggest therefore a modified version of the Ischemic
Score of Hachinski using 4 items of Hachinski~s I.S. plus
CT findings, as a suitable tool for the diagnosis of vas-
cular dementia. Such a modified ischemic score (MIS) could
comprise 5 items, that is, 4 relevant items of I.S. and
the CT features (isolated or multiple low density areas)
(table 2).

A full 9-10 point score points to the existence of

old or recent cerebrovascular lesions, simple or multiple.
If one employs such a MIS (3) in evaluating our 94 pa-
tients, the following is found; 1) out of 65 patients,
probable SDAT cases, 65 (95%) had a MIS of 0-2; 2 patients
of 3-4 (3%), 1 patient of 5 (2%); 2) out 18 patients,
proba~le MID cases, 18 had a MIS of 5-10 (100%); 3) out
of 11 patients, in whom vascular dementia was suggested,
3 had a MIS of 0-2, 2 of 3-4, and 6 of 5-9.

Therefore, a MIS from 0 to 2 points to a diagnosis

of SDAT, from 5 to 10 a diagnosis of MID, while a 3-4 MIS
score leaves the question insettled.

In 14 patients (15%) (3 patients of the SDAT subgroup

and 11 patients of the VASC subgroup) the diagnosis re-
mains doubtful. The I.S. could not differentiate between
MID and MIX, and in these cases mixed forms of dementia
could be present. In about 15% of the cases, the differen-
tial diagnosis is not feasible.

664
 Table 2. Modified Ischemic Score (M.I.S.)

1. Abrupt onset 2
2. History of stroke 1
3. Focal symptoms 2
4. Focal signs 2
5. CT-Low density areas
isolated 2
multiple 3
Maximum Score = 10

The proposed modified ischemic score (MIS), therefore

on the basis of our experience, respresents a diagnosis
tool which increases the discriminating power of the I.S.
of Hachinski in the differential diagnosis between SDAT
and MID.

REFERENCES

1. v. c.
Hachinski, L. Iliff, G.H. Duboulay, v. Me Alli
ster, J. Marshall, R.W. Ross Russell and L. Symon,
Cerebral blood flow in dementia, Arch. Neurol. 32,
632-637 ( 1975).
2. C. Loeb, Clinical Diagnosis of Multi-infarct Dementia,
in: "Aging of the Brain and Dementia" (Aging v 13),
ed. L. Amaducci et al., Raven Press, New York (1980)
3. c. Loeb and C. Gandolfo, Diagnostic evaluation of de-
generative and vascular dementia, Stroke 14: 385-
388 (1983).

665
LONG TERM CLINICAL OBSERVATION

OF MULTI INFARCT DEMENTIA

Helmut Lechner, Gotz Bertha and Erwin Ott

Department for Neurology and Psychiatry

University of Graz
Auenbruggerplatz 22, A-8036 Graz

INTRODUCTION

Since several years multi-infarct disease has been

recognized as a special form of cerebrovascular disease
(CVD) which may lead to dementia of the vascular type
(multi-infarct-dementia). Diagnosis of multi-infarct de-
mentia(MID) is based on various parameters such as compu-
terized tomography (CT), measurements of regional cerebral
blood flow (rCBF) , evaluation of vascular risk factors
(RF) and neuropsychological rating scales.

Since meager information exists about the prognosis

of MID it is the aim of this paper to report preliminary
results of a long term clinical observation of patients
with MID.

PATIENTS AND METHODS

There were 40 patients* (27 males, 17 females) with

a mean age of 62 years (range 38-75) who have been inclu-
ded in this study due to the clinical and neuropsycholo-
gical evidence of MID. Besides a detailed history and a
neuro-psychiatric examination all patients underwent an
extensive laboratory and neuropsychological examination
including CT, electrocardiogram (ECG), and determinations

*Originally 66 patients with MID have been selected.

22 patients died during the observation period.

667
of vascular RF. The neuro-psychological examination in-
cluded substests of the WAIS and the BENTON test; testing
of the vigilance has been performed using the Syndrom-
Kurztest. A social ability score (SAS) was established
considering the ability of the patients to manage the ne-
cessities of daily life. A compliance score considered
the readiness of the patients to visit their doctor regu-
larly and the intake of the drugs prescribed.
Mean observation time of this group was 31.3 months.
The values thus obtained have been compared to those of
13 patients (9 males, 4 females) with a mean age of 62
years (range 36-45) who have been investigated for sym-
ptoms due to multi-infarct disease. Neuropsychological
examination excluded signs of dementia in these patients.
Mean observation time of this group was 31.6 months.

All patients have been controlled at least twice and

have been ordered medical treatment including psychophar-
macological drugs if necessary.

RESULTS

1) Social _prognosis of MID ~o!N'ared to multi-infarct

disease (table 1):

The social prognosis of the patients with MID in-

vestigated has been established using a social ablility
score. Although not significant improvement of social
abilities was more frequent in non demented patients
(+18.5%) while deterioration was more frequent in demen-
ted patients (+24.6%).

Table 1. Social prognosis MID vs. M.I. disease

Social ability score

Unchanged im12roved worsened
1 N % N % N %
M.I.Dementia
(N=40) 14 35.0 10 25.0 16 40.0
M.I.Disease
(N=13) 6 46.1 5 38.5 2 15.4

1 statistical analysis not taking into consi-

deration the patients compliance

668
2) Clinical prognosis of MID (table 2) :

Originally 66 patients with MID have been considered

for this study, however, 22 patients died during the ob-
servation time and will be excluded from the further ana-
lysis of the data. Causes of death were recurrent stroke
in 5 patients (22.7%), cardiac insufficiency in 8 patients
(36.4%), pulmonary embolism in 2 patients (9.1%), pneumo-
nia in 6 patients (27.3%) and renal insufficiency in 1
patient (4.5%). Hence, cause of death was of vascular
nature in 68.2%.

Table 2. Long term clinical follow up in 62

patients with MID

dead unchanged improved worsened

22/62 14/62 10/62 16/62
(35.5%) (22.6%) (16.1%) (25.8%)

3) Frequency of RF in patients with MID vs. patients with

M.I. disease

According to their pathogenic importance in the de-

velopment of artherosclerotic changes of cerebral vessels
RF can be divided into those of the 1. and those of the
2. order. It can be noted from table 3 that hypertension
and diabetes mellitus were more frequent in patients with
MID while the distribution of other RF was similar in
both groups.

Table 3. Distribution of RF in patients with MID

vs. patients eith M.I. disease

M.I.Disease(N=13) M.I.Dementia(N=40)
N % N %
Hypertension 7 53.8 25 62.5
Diabetes 1 7.6 7 17.5
Hyperlipidaemia 7 53.8 17 42.5
Cor.Heart Disease 3 23.0 15 37.5
Hyperviscosity 9 69.2 30 75.0
Hyperfibrinogen 10 76.9 21 52.5
Enhanced PAG 10 76.9 25 62.5
Overweight 2 15.4 14 35.0

669
4) Compliance in patients with MID

Classification of compliance was based on the readi-

ness of the patients with MID to see their doctors regu-
larly and to follow their prescriptions. Thus, two groups
of patients with MID have been formed and the influence
of the compliance on the development of their social
abilities has been calculated. It can be seen from table 4
that improvement of social abilities was more frequent in
the poor compliance group. However, it must also be noted
that in the good compliance group 35.7% of the patients
showed a deterioration of their social abilities despite of
medical control which included also psycho-pharmacological
treatment.

Table 4. Importance of compliance for social

outcome of patients with MID

Social ablility score

Unchanged improved worsened
N % N % N %

Good compl. 10 35.7 8 28.6 10 35.7

(N=28)
Poor compl. 4 33.3 2 16.7 6 50.0
(N=12)

DISCUSSION

It has been demonstrated in this paper that a long

term clinical observation of patients with MID revealed
a lethal outcome in 35.5% whereby cause of death of vas-
cular nature in 68.2%. Moreover, in the surviving patients
deterioration of their social abilities occurred in 35.7%
despite controlled medical treatment which included also
psychopharmacological drugs. On the contrary deterioration
of social abilities was observed in 15.4% of patients
with multi-infarct disease but without evidence of de-
mentia signs.

Furthermore, hypertension and diabetes mellitus were

more frequent in patients with MID.

It has been shown that signs of dementia in patients

with multi-infarct disease may be present if bilateral
infarcts can be seen whereby bilateral involvement of the
thalamus predisposes to MID. In our patients with MID

670
bilateral thalamic involvement was present in 25%. In
general, neuropsychologic examinations of patients with
MID revealed predominantly disturbance of recent memory,
associative ablilities and attention. Similar psychopa-
thologic profiles may be found in long lasting hyperten-
sives and in our patients with MID hypertension together
with diabetes mellitus were the most frequent vascular RF.
The importance of these RF for the development of arthero-
sclerotic vessel disease are well recognized. Although
controlled medical treatment may favourably influence the
course of MID there is evidence from our results that the
progress of the vascular disease cannot be stopped. This
may be concluded from the fact that 35.7% of the "good
compliance group" showed deterioration of their social
ability score and that in 68.2% of the died patients with
MID cause of death was of vascular nature.

SUMMARY
In 40 patients with MID a long term clinical obser-
vation of 31.3 months revealed deterioration of their
social abilities in more than one third. Arterial hyper-
tension and diabetes mellitus have been identified as
leading vascular RF. There was a 50% incidence of clinical
deterioration in patients with poor compliance, however,
in patients with good compliance clinical deterioration
have been observed in 35%. From originally 62 patients
22 patients died during the observation period and in
68.2% cause of death was of vascular nature. Therefore,
it was concluded from the present results, that the under-
lying vascular process in MID might be favourably influ-
enced by controlled medical treatment but not stopped.

REFERENCES
F. Boller, B. Vrtunsky, J.L. Mack, 1977, Neuropsycholo-
logical correlates of hypertension, Arch. Neurol.,
34: 701-705
D.V. Cramon, J. Klihnlein, A. Wolfram, 1981, Die thalami-
sche Demenz. Fortschr. Neurol. Psychiat., 49: 129-
135
V.C. Hachinski, N.A. Lassen, J. Marshall, 1974, Multi-
infarct dementia. A cause of mental deterioration
in the elderly, Lancet 2: 207-210
V.C. Hachinsky, L.D. Iliff, E. Zilkha et al., 1975, Cere-
bral blood flow in dementia. Arch. Neurol., 32:
632-637
G. Ladurner, W.D. Sager, 1981, Morphologische Bedingungs-
konst llation der vaskularen Demenz. Fortschr.

671
 Neurol. Psychiat., 49: 53-55
G. Ladurner, L.D. Iliff, D. Sager, 1982, A clinical
approach to vascular (multi-infarct) dementia.
EXF· Brain Res. S~ppl., 5: 246-253
E. Ott, G. Bertha, K. Marguc et al., 1982, Klinische und
hamodynamische Aspekte des zerebralen Multiinfarkt-
geschehens. Nervenarzt, 53: 78-82
F.I. Perez, V.M. Rivera, J.S. Meyer et al., 1975, Analysis
of intellectual and cognitive performance in pa-
tients with multi-infarct dementia, vertebrobasilar
insufficiency with dementia and Alzheimer's disease.
J. Neurol. Neurosurg. Psychiat., 38: 533-540

672
GENERAL LIVING SYSTEMS THEORY

James Grier Miller

Neuropsychiatric Institute
University of California at Los Angeles
Los Angeles, California 90024

BASIC CONCEPTS

An increasing number of psychiatrists recognize the need

of psychiatry for an integrative theory to provide a basis for
their diagnostic and therapeutic actions. General living systems
theory is a conceptual framework within which the biological
and social approaches to the study of living things are logically
integrated with the physical sciences. This theoretical integra-
tion and the empirical testing of hypotheses which it requires,
can provide to psychiatry the theoretical and empirical support
that other medical specialties find in cellular biology, physio-
logy, and biochemistry.

Systems of the physical world, including all forms of life,

are concrete systems, that is, they are nonrandom accumulations
of matter and energy, each in a region of physical space-time.
Concrete systems share with systems of all kinds the characteris-
tics of having parts (units or components) that have some common
properties, are interdependent, and interact within the system.

An orderly progression in complexity can be observed in

the physical world. Its systems form a hierarchy of levels that
begins with the smallest system, which may be the atom, and pro-
gresses to the largest system of which we have knowledge, the
universe. In this hierarchy, systems at each level are made
up of systems at the level below except for the lowest level,
which has units that are not themselves systems. These units
and lower level systems are the system's components. Like the
nonliving systems of the physical universe, living systems occur
in a hierarchy of several levels, each more complex than the

673
one below. The increase in complexity is a result of the evolu-
tionary process by which the higher levels developed from the
lower. As larger systems evolved, more units with greater special-
ization were needed to carry out system processes. It is as
if each strand of a many-stranded rope had unraveled progressively
into more and more strings and threads so that instead of one
sort of component performing several distinct processes, each
process was carried out by a different specialized part. This
aspect of evolution we call "shred-out". As a result of its
greater complexity, each higher level has emergent characteristics
not found at the level below, a fact that is expressed in the
saying that "the whole is greater than the sum of its parts."

The Levels of Living Systems

Seven levels of living systems can be identified, each with

characteristic structure and processes, and each composed of
systems at the level below. The seven levels are: cell, organ,
organism, group, organization, society, and supranational system.
The structure, that is, the arrangement of their components in
space, and the functional and historical ~rocesses of systems
at each level are characteristically different from other levels,
although the levels grade into each other in such a way that
some systems may be difficult to classify. With some exceptions,
systems at each higher level are larger than those lower in the
hierarchy.

Living systems at all levels may include nonliving components

such as the variety of artifacts used in human societies or pros-
theses that substitute for missing parts of organisms or augment
the function of a defective organ.

Human organisms and higher-level systems interact with each

other; with living systems of other animal species; with plants,
fungi, and cells; and with nonliving components of the environment
in ecological systems. All these diverse components are linked
in the flows of matter, energy, and information in such systems.
Local ecological systems are components of larger ecological
systems which together form the total planetary system. Ecological
systems are mixed living and nonliving systems that resemble
and yet differ in significant ways from living systems.

The Subsystems of Living Systems

Living systems share many characteristics with nonliving

systems. In addition, they have distinctive additional attributes,
some of which are: (a) They are more complex in structure and
process than nonliving systems. (b) As open systems, they have
the unique ability to combat, for varying lengths of time, the
increase of entropy that leads to the dissolution of all physical

674
systems. They do this by taking in from the environment substances
of low entropy like organic molecules, fuels, and other complex
materials and returning to the environment substances of higher
entropy, more random organization, or less complex molecular
structure than their input. This makes it possible for them
to maintain or even increase their own molecular complexity,
carry on their activities, and reproduce. (c) In order to remain
alive and continue their species beyond a single generation,
all living systems must have a specific set of subsystem processes.
General living systems theory identifies 19 such processes in
complex organisms and higher level systems. Since some of these
emerged late in evolution, they are represented in many cells,
organs, and plant organisms by a variety of structural and environ-
mental adaptations. Each of these 19 processes is carried out
by a set of components. Systems that lack components to carry
out one or more subsystem processes may depend upon a parasitic
or symbiotic relationship with another system or exist in a favor-
able environment which provides it. (!!) Living systems process
more information than do most nonliving systems. Information
is patterning or order, as distinguished from randomness or dis-
order. It is conveyed in the shape of molecules that start or
stop cellular processes, in pheromones that influence insect
behavior, in bioelectric pulses along a neuron, or in the words
of a message.

Each subsystem processes a part of the matter-energy (matter

and/or energy) and information flows that are essential to systems.
Because they are systems at a lower level, they all process both
matter-energy and information. Within the larger system of which
they are a part, however, some process primarily matter-energy,
some process primarily information, and some process significant
amounts of both.

The boundary, located at the periphery of the system, holds

the components together, protects them from environmental stresses,
and prevents or permits entry or exit to various sorts of mat-
ter-energy and information. The reproduc~~ organizes matter-energy
according to information in the system's template or genetic
instructions, to give rise to other systems similar to the one
it is in.

The eight subsystems that process primarily matter-energy

are: The ingestor, which brings matter-energy into the system.
The distributor carries inputs from the environment or outputs
from subsystems to all parts of the system. The converter changes
certain inputs to the system into forms more useful for its special
processes. The producer synthesizes materials for growth of
the system, for repair and replacement of components, or for
providing energy for system processes of all kinds. Matter-energy
storage retains in the system, for different periods of time,

675
deposits of various sorts of matter-energy. The extruder transmits
products and wastes out of the system. The motor moves the system
or parts of it, or moves components of the environment. The
supporter maintains the proper spatial relationships among compo-
nents of the system so that they can interact without weighting
each other down or crowding each other.

Information flows in a comparable way from input to output,

processed successively by the information processing subsystems.
These are: The input transducer, the sensory subsystem which
brings information into the system on markers, such as radiant
energy or molecules. The internal transducer·, the sensory subsys-
tem which receives from subsystems or components markers about
internal states of the system, changing them to matter-energy
forms of a sort that can be transmitted within the system. The
channel and net is the subsystem over which markers bearing infor-
mation are transmitted to all parts of the system. The decoder
changes the code of information input to it from that used in
the environment to a "private" code used within the system. The
associator and memory each carry out part of the learning process.
The associator forms enduring associations among items of informa-
tion and the memory stores various sorts of information in the
system for different periods of time. The decider is the executive
subsystem that receives information from all other subsystems
and transmits to them information for guidance and control of
the entire system. The decider may be centralized in a single
component or decentralized to many components. This subsystem
process cannot be entirely performed for the system by any other
system. A system that loses its decider ceases to exist as a
system. The encoder alters the code of information input to
it from other information processing subsystems, from a "private"
internal code to a "public" code which can be interpreted by
other systems in the environment. The output transducer puts
out markers from the system, changing the markers on which informa-
tion has been transmitted within the system to others that are
suitable for transmission over environmental channels.

§ubsystem and System-Wide Processes

Each of a system's subsystems keeps a number of critical

variables in steady state range when the system is functioning
optimally. This is accomplished by adjustment processes that,
for example, increase or decrease the rate of processing, change
the number of components involved in a process, or filter inputs
or outputs at a subsystem boundary.

The matter-energy processing subsystems control such variables

as quantity, quality, and type of matter-energy ingested; changes
in distributing over time; and rate of producing. Variables
of the information processing subsystems include threshold, channel

676
capacity, rate of processing, lag, codes used in transmission,
and meaning of information processed. Measures or indicators
are available for most, if not all, variables of each subsystem
at each level, except the meaning variables for which no unit
of measurement is as yet generally accepted.

By varying the activities of one or more of its subsystems,

a living system can relieve strains that arise within it or are
caused by environmental stresses. Many subsystem and system-
wide adjustment processes are controlled by negative feedbacks.
Such feedbacks are conveyed by loops that connect various parts
of a system and relate the system to its environment. Each adjust-
ment process is carried out at some cost.

In addition to such reversible adjustment processes, a pattern

of irreversible historical processes occurs in the life cycle
of each living system as it grows, develops, passes through a
period of maturity, declines, and finally terminates. Each such
system is subject to different sorts of pathological processes
which reduce the ability of the system to function normally or
cause its termination.

BASIC RESEARCH

As is common for scientific theories, general living systems

theory depends upon empirical research to support or disconfirm
its assertions. A major research goal has been to find quantita-
tive formal identities at two or more levels of living systems.
My book Living Systems lists 173 cross-level hypotheses, some
of which have received such confirmation.! Since most scientists
are expert at only one level of living system, some of this has
been interdisciplinary group research.

To date, at least a dozen individual authors or groups have

either carried out studies or analyzed data bearing upon cross-
level hypotheses. The first full-scale experimental program
of this sort, on information input overload, was conducted by
the author and a group of colleagues in the late 1950's and early
1960's. We demonstrated that living systems at the levels of
cell, organ, organism, group, and organization yield comparable
output curves as information input rates increase. 2 The output
rate goes up to the channel capacity of the system, and then
declines toward zero as breakdown or a confusional state occurs.
The channel capacity at each higher level was lower than at the
previous level. In addition, a number of adjustment processes
to the increasing overload were found to be common to several
or all levels. Systems at more complex levels added new adjustment
processes while continuing to use those employed at lower levels.

Cross-level researches by other scientists have been concerned

677
with such topics as maximization of power (in nonliving and ecolog-
ical systems, cell, organism, society); effects of conflicting
commands on decision time (organism, group); threat-rigidity
effects (organism, organization); information flows in communica-
tion nets (organ, organization); monetary flows (society, supra-
national system); input-output relationships of input transducers
(cell, organism); and learning (organism, group, supranational
system).

APPLICATIONS

Systems analysis of any type and level of living system

involves examination of the structure and processes of each of
its subsystems and of its subsystem and systemwide adjustment
processes, to discover pathology and determine what intervention
is appropriate. Although such practical application is in its
early stages, hospitals, universities, corporations, urban systems,
and other organizations have been studied in this way. General
systems analysis has also been used in individual and group psycho-
diagnostics, psychotherapy, and combinations of drug therapy
and psychotherapy.3

Members of the staff of a psychiatric unit, including the

author, carried out such an analysis at levels of the organism
and group in the case of an anorectic boy. 4 His 19 subsystems
and his adjustment processes were examined in order to identify
pathology and make a physical and psychodynamic diagnosis. A
comparable procedure were carried out in terms of the 19 subsystems
and adjustment processes of his family as a total group system.
Coordinated therapies of the boy and his family were then conducted
as indicated by these diagnoses.

REFERENCES

1. J. G. Miller, "Living Systems," McGraw-Hill, New York,

(1978).
2. J. G. Miller, Information input overload and psycho-
pathology, Amer. J. Psychiat., 116:695 (1960).
3. J. G. Miller and J. L. Miller, General living systems
theory, in: "Comprehensive Textbook of Psychiatry,"
editions II, III, and IV, H. I. Kaplan, A. M. Freed-
man, and B. J. Sadock, eds., Williams and Wilkins,
Baltimore (1975, 1980, in press).
4. J. G. Miller and J. L. Miller, General living systems
theory and small groups, in: "Comprehensive Group Psycho-
therapy," H. I. Kaplan and B. J. Sadock, eds., Will-
iams and Wilkins, Baltimore (1983).

678
FROM DESCARTES TO PAVLOV TO ANOKHIN: THE EVOLUTION OF GENERAL
SYSTEMS CONCEPTS IN BIOMEDICAL SCIENCES IN EASTERN EUROPEl

Samuel A. Corson and Elizabeth O'Leary Corson

Department of Psychiatry and Department of Educational
Theory and Practice
Upham Hall
The Ohio State University
Columbus, Ohio 43210 USA

ABSTRACT
Rene Descartes developed a mechanistic notion of reflex action
in an attempt to describe automatic acts of "soulless" machine-like
"unfeeling" animals, in contrast to rational voluntary behavior of
humans directed by a "soul" via the pineal gland. In contradistinc-
tion to Cartesian dualism, Pavlov introduced the monistic concept
of an integrated organism (theory of nervism) and of conditional
reflexes enabling all animals to modify their behavior in response
to changing environmental situations. This presentation will discuss
the expansion of Pavlovian psychobiology by P.K. Anokhin to include
the application of systems concepts to psychophysiologic adaptation
and the notion of the functional system as the basic unit of neuro-
physiologic integration.

I. THE DEVELOPMENT OF PHILOSOPHICAL TRENDS IN BIOMEDICAL THINKING

Vitalism, one of the earliest concepts in biology, can be
traced to Plato's "eidos" or his proposition that only "ideas" are
real and that things which we can observe are only poor imitations
of "eidos" in the mind of God. Vitalism, as a form of philosophic
idealism, assumed that the functions of living organisms are due to
some "vital principle" which is distinct and independent of
lsupported in part by NIH grant HL 20861, the Ohio State University
Graduate School Biomedi£al Research Support Grant, and the
Psychiatric Research Foundation of Columbus.

679
physicochemical entities or forces. The disservice of such a concept
is that it removed the problem of life from the realms of scientific
investigation or experimental verification.
Descartes (1596-1650) attempted to apply principles of mathe-
matics and physics to biological thinking. Thus he can be considered
the major contributor to the introduction of the concept of material-
ism into biology. Paradoxically, while attempting to introduce the
scientific method into biological thinking, Descartes at the same
time reverted back to Platonic idealism when he came to analyze human
behavior and particularly the human mind. The science of physics at
the time of Descartes was primarily the science of mechanics and the
applicati.on of mechanics to various types of machinery, including
a variety of ingenious mechanical toys which excited the imagination
of laymen as well as scientists.
Descartes developed the concept of the reflex to explain the
behavior of animals as mechanical entities. He referred to animals
as automatons exhibiting "reactive" behavior without consciousness.
Reflexes also were involved in human behavior, but humans were
endowed with the ability to think and be conscious of their environ-
ment via the pineal gland, which allegedly was the connecting link
to the "soul". A human being, according to Descartes, is "some kind
of a union of two distinct things: a soul, or mind, and a body. The
body is part of mechanical nature, the mind, a pure thinking sub-
stance" (Macintyre, 1967).
Thus this Cartesian mechanistic materialism was coupled with
Cartesian mind-body dualism and in many ways has served as a deter-
rent to the development of scientific studies of the relationship
between the brain, behavior, and mental and emotional functions.
Before proceeding to an analysis of the contributions of the
Pavlovian and Anokhin schools to the development of a systems ap-
proach in biological thinking, it is worth mentioning briefly the
important contribution in this area by a man whose name is not gen-
erally encountered in the history of biological sciences. We are
referring to Baruch (Benedictus) Spinoza (1632-1677), a man who was
trained not as a scientist but as a rabbi, and who was excommunica-
ted from the Synagogue in Amsterdam for his profound independent
thinking.
Spinoza published an extensive critique of Descartes•s dualism
(~1aclntyre,1967). Spinoza proposed a monistic view of the world,
claiming that "Deus sive natura" ("God is nature"), and that "the
mind and the body are one and the same thing." He introduced a sys-
tems approach in philosophy. Thus he stated: "We must .9rasp the
system as a whole before we can hope to grasp the nature of the
parts, since the nature of the part is defined by its role in the
total system. Spinoza foreshadowed the concept of homeostasis by
11

680
stating that "the human being, like any other being, is a unity in-
Spired by a conatus in suo esse perseverandi (a drive toward self-
preservation)11. Further, he elaborated this concept by stating that
in a living organism "there is an inherent tendency to react to all
changes in a way that maintains its characteristic unity and equil-
ibrium" (Peters and ~1ace, 1967).
Until recently, biomedical thinking in the Western countries
and in the USA was largely dominated by a dualistic reductionist
orientation, a type of mechanistic materialism. For a number of
reasons (discussed elsewhere, Corson and O'Leary Corson, 1976),
Russian biomedical thinking during the nineteenth century had been
developing along integrative monistic concepts. Spinoza could have
been a connecting link between Cartesian mechanistic, dualistic ma-
terialism and the monistic integrative materialism developed by the
Russian biomedical "troika": the physiologist, I.M. Sechenov, the
clinician, S.P. Botkin, and the physiologist, I.P. Pavlov. Apparently
these scientists appeared to be unaware of Spinoza•s important con-
tributions to a integrative monistic view of the world.
The underlying theory developed by this trio is referred to as
"nervism" or the neurogenic theory. The essence of this theory is
as follows:
1. All living processes, including consciousness, can be ulti-
mately analyzed in terms of natural laws. However, the specific con-
figuration of materials forming living organisms endows them with
properties not possessed by the individual components. The proper-
ties of HzO or of HzOz do not represent a mere summation of the
properties of the component elements. A new quality is represented
in a molecule of HzO or of HzOz, this quality being determined by
the specific configuration of the elements. The appearance of a new
quality in a chemical compound does not compel the chemist to re-
sort to mysticism in explaining the properties of water or of a
tobacco mosaic virus molecule. By the same reasoning, a biologist
need not invoke mystical vitalism to explain the properties of liv-
ing matter. A biologist or a chemist could choose to resort to mys-
ticism, but this type of approach fails to provide either under-
standing, prediction, or control.
2. Nature contains many hierarchies of levels of organization
and configuration of matter, each level possessing characteristic
qualities. The quality we call "life" appeared at a particular
level of organization of certain material entities. Consciousness
and mind are qualities characteristic for a certain level of organ-
ization of neural elements.
3. Living organisms are characterized by the presence of a
hierarchy of different levels of integration, beginning with intra-
cellular integration and leading up to integration of organs and

681
organ systems by means of a central nervous system. In higher ani-
mals, the central nervous system itself is structured on the basis
of integrated hierarchic levels, the cerebral cortex assuming a pro-
gressively increasing dominant role in animals higher in the evolu-
tionary scale. Cerebrovisceral theory added the emphasis that this
increase in cerebral dominance in higher animals also holds true for
visceral and endocrine functions.
4. The progressively increasing dominance of the cerebral man-
tle in higher organisms, and especially in man, and the resulting
increase in the multiplicity of physiologic responses to symbols of
physicochemical stimuli, multiplied the possibilities for the im-
pingement of psychologic factors on somatic and visceral functions.
Consequently, the Sechenov-Botkin-Pavlov neurogenic theory led to
the emphasis of the importance of psychosocial factors in diagnosis,
therapy and prophylaxis. Thus was prepared the physiologic basis
for psychosomatic medicine.
One of the most far-reaching contributions to an integrative
systems approach in biomedical sciences was made by P.K. Anokhin
(1898-1974), one of Pavlov's most brilliant and imaginative stu-
dents. Having studied with Vladimir Bekhterev (at the Brain Research
Institute) before coming to Pavlov's laboratory, Anokhin attempted
to achieve a synthesis of Pavlov's basic conditioning studies with
Bekhterev's keen clinical psychiatric and neurologic observations
and his pioneering studies on behavior modification and group psy-
chotherapy methods. Anokhin was one of the first in the USSR to
initiate systematic utilizations of electrophysiologic techniques
in the investigation of conditional reflexes and their role in bio-
logical adaptation.
Anokhin had the unique distinction of initiating a systematic
dialogue and synthesis between the Pavlovian school and Western
neurophysiology, experimental psychology, and cybernetics. As early
as 1935, Anokhin published a paper in which he developed the con-
cept of the functional system as the basic unit of neurophysiologic
integration, incorporating into this concept the notion of return
11

afferentation thus foreshadowing the development of the concepts

11 ,

of feedback, and a systems approach in psychobiology long before

the publication of Wiener's Cybernetics in 1948. Anokhin's exten-
sive studies and original concepts are elegantly summarized in an
expanded English translation of his latest book (Anokhin, 1974).

REFERENCES
Anokhin, P.K., 1935, Problema tsentra i periferii v sovremennoi
fiziologii nervnoi deiatel'nosti (The problem of center and
periphery in the contemporary physiology of nervous activity),
in: Problema Tsentra i Periferii v Fiziologii Nervnoi Deia-
11

682
 tel'nosti (The Problem of Center and Periphery in the Physi-
ology of Nervous Activity), P.K. Anokhin, ed., Gosizdat,
Gorki, pp. l-70.
Anokhin, P.K., 1974, 11 Biology and Neurophysiology of the Conditioned
Reflex and Its Role in Adaptive Behavior, 11 Samuel A. Corson,
Scientific and Translation Editor, (International Series of
Monographs on Cerebrovisceral and Behavioral Physiology and
Conditioned Reflexes, Volume 3), Pergamon Press, Oxford.
Corson, Samuel A., and Corson, Elizabeth O'Leary, 1976, Philosophi-
cal and historical roots of Pavlovian psychobiology, in:
11 Psychiatry and Psychology in the USSR, 11 S.A. Corson,ed.,

Plenum Press, New York, pp. 19-58.

Descartes, R., 1897-1910, ( 11 0euvres de Descartes 11 ) , C. Adam and P.
Tannery, eds., 12 volumes, Paris.
t1acintyre, Alasdair, 1967, Spinoza, in: 11 Encyclopedia of Philosophy, 11
Volumes 7-8, MacHillan Publishing Co., New York, pp. 530-541.
Peters, R.S., and Mace, C.A., 1967, Descartes, in: 11 Encyclopedia of
Philosophy, 11 Volumes 7-8, ~1acmillan PublTShing Co., New York.
Spinoza, B., 1981, ( 11 The Ethics: Spinoza 11 ) , S.P.R. Charter, ed.,
Joseph Simon, Malibu, California.
Wiener, Norbert, 1948, 11 Cybernetics, 11 John Hiley and Sons, New York.

683
SYSTEMS THERAPY

Gottlieb Guntern

ISO, Institute for Systems Science

Postfach 523
3900 Brig, Switzerland

INTRODUCTION

Systems Therapy is the therapy of human systems such as indivi-

duals, couples, nuclear and extended families, and groups.

Systems Therapy is based upon the basic paradigm of systems

thinking and, therefore, constitutes a radical break with psychoana-
lysis, behaviorism and similar forms of reductionist theories and
therapies. Only a few basic concepts are presented here.

THE ORGANISM AS A SYSTEM

The organism is a system, i.e., an organized whole composed

of transacting functional subsystems and structural components.

Every exchange of matter-energy and/or information within the

organism or between the organism and its environment is called a
transaction. The organism is viewed as "being and becoming in trans-
actions"; it is a transactor participating in multidimensional, mul-
tidirectional, reciprocal exchanges of matter-energy and information
within an ecosystem. It co-determines the transactions of the ecosy-
stem and is co-determined by them in a continuous, multidimensional,
multidirectional, reciprocal process. It participates in something
which transcends the organism.

The organism may be described by an observational approach

called perspectivism (Von Bertalanffy), an epistemic result of

685
the principle of relativity which holds that there is no reality per
se; there are only realities as seen from a specific physical and/or
conceptual viewpoint of the observer (1). Each perspective yields
one aspect - not a structural "part" - of the organized whole. Each
perspective necessarily excludes other aspects of the organism, in-
cludes them or overlaps with them. Theoretically, only an infinite
set of perspectives yields a complete description of all aspects.
Historically, four aspects have been described in some depth: the
cognitive, the affective, the physio-chemical, and the behavioral
aspect.

It must be emphasized that systemic perspectivism definitively

discards the mind-body dichotomy which still pervades psychiatry and
medicine in general. Terms such as "somatic", "psychology", "psychia-
try", "psychosomatic", "psychodynamic", "psychoanalytic", "psychothe-
rapist", etc., make only sense if the mind-body dichotomy is accepted.
Within systems thinking they are meaningless relics of the history of
analytic-dualistic thinking prevailing since the days of Plato.

Aspects of Organismic Operation

The cognitive aspect of an organism's operation describes formal

properties and contents of neo-cortical information-processing. For
example: depression is a possible form of organismic operation. Its
cognitive aspect may be described as slow, circular thinking with
ideas of guilt and ruin etc.

The affective aspect of an organism's operation describes formal

properties and contents of mainly subcortical information-processing.
The affective aspect of a depressive organism may be described as
rigid emotions of anxiety and petrification, i.e., incapacity to
"feel" loneliness and melancholy etc.

The physio-chemical aspect of an organism's operation describes

formal properties and contents of matter-energy-processing. The
physio-chemical aspect of a depressive organism may be described as
lack of norepinephrine, serotonine and other neurotransmitters in
specific areas of the brain; as disturbed sleep-architecture; as
diminished speed of peristalsis etc.

The behavioral aspect of an organism's operation describes formal

properties and contents of verbal, paraverbal and non-verbal behavior.
The behavioral aspect of a depressive organism may be described as
disturbed verbal expression of cognition and emotions, slow, soft

686
speech and poor facial expression and body movement and stooped
posture.

Reductionisms such as "ultimately, depression is due to inade-

quate ideas or emotions", or "to a lack of norepinephrine and sero-
tonine" etc., are no longer justified yet still widely held.

Organismic Programs

Two types of organismic programs co-determine organismic operations.

The genetic pro~ram stored by the DNA of every cell's nucleus

contains the results of phylogenetic learning. It co-determines con-
stitution and operation of organismic structures and processes by
means of hard-programmed information (e.g., colour of eyes, overall
structure of "body" frame) and soft-programmed information (e.g.,
behavioral dispositions).

The syngenetic program (syngignomai = becoming, developing to-

gether) stored in the memory of the brain contains the results of
ontogenetic, individual learning (2). It co-determines organismic
operation by specifying global genetic "dispositions". For instance,
aggressive behavior is co-determined by a genetic disposition to
act aggressively under stress and by the specific rules of its
syngenetic program which specify how aggression is expressed. Unlike
the genetic program whose structure is given once and for all, the
syngenetic program is continuously built up, maintained, transformed
and dissolved.

Actual Operational State of an Organism

The actual operational state of an organism (e.g., consciousness,

sleep, exhaustion) is defined as the organized set of the numeric
values of all variables of an organism. It includes, amongst others,
adjustment (values of intraorganismic variables) and adaptation
(values of matter-energy and information variables between organism
and environment).

ECOSYSTEM

The basic unit of survival and development is not the organism-

as Darwin believed and many contemporary scientists still do - but
the ecosystem.

687
Constituents

The ecosystem is composed of a reference system (organism,

couple, family, group, etc.), a physical environment (geography,
meteorological conditions, buildings, etc.) and a biosocial environ-
ment (plants, animals, human beings).

Processes

Within and between the constituents of the ecosystem there is

a continuous exchange of matter-energy and information. The quanti-
tative and qualitative aspects of the input, throughput and output
co-determine the structure of the ecosystem as an organized whole.
The structural components and functional matter-energy- and infor-
mation-processing subsystems are described in detail by Miller (3).

Probabilistic Co-Determinism

The laws governing the physical environment (e.g., law of gravity,

Second Law of Thermodynamics), the laws governing the biosocial en-
vironment (e.g., principle of natural selection) and the laws and
principles governing reference systems (e.g., organismic programs)
co-determine every observable organismic transaction. Random events
also co-determine behavior and, thus, we may speak of probabilistic
co-determinism. The overall probabilities of an ecosystem are not
the result of the algebraic sum of individual probabilities but rather
the result of the organization of the individual probabilities.

Thus, organismic operation is probabilistically co-determined

by the physical enviroment, the biosocial environmen~, the organis-
~ic genetic program and the syngenetic program and its actual opera-
tional state.

TRANSACTIONAL FIELD AND TRANSACTIONAL PATTERNS

The Transactional Field and the Transactional Patterns

The quantity and quality of matter-energy and information trans-

actions within an ecosystem constitute what I call a transactional
field; within it a general matrix pattern (organization) and seven
subpatterns (episteme, autonomy, hierarchy, cooperation, conflict-
management, decision-making, strategic coping) are differentiated.
The transactional patterns are processes which are constantly built

688
up, maintained, transformed and dissolved.

Organization

Organization is the specific order of a transactional field and

its transactional patterns.

The basic model of organization (4) describes organization as

process, synchronically and diachronically composed of six subproces-
ses: evaluating status quo, establishing purposes and goals, computing
problems (difference between status quo and desired state of system),
choosing strategies and tactics, implementing strategies and tactics,
and controlling.the implementing.

The result of overall organization and, specifically, of con-

trolling produces morphogeneration (generation of new variables such
as structures, processes, relationships), morphostasis (maintenance
of existing variables), morphotransformation (transformation of ex-
isting variables; anatransformation towards higher complexity; iso-
transformation towards equal complexity and katatransformation towards
lower complexity) and morpholysis (dissolution of existing variables).
All four types may be informative if higher order results, or entropic
if lower order occurs.

Types of organization; a great number of types exist such as

se-auto-organization (a system autonomously organizes itself),
allo-organization (a system organizes something else), auto-organi-
zation (a system organizes), hetero-organization (a system is orga-
nized by another system).

The Seven Transactional Patterns

In every human system (individual, couple, family, group, com-

munity, social institution, society etc.) and in every ecosystem
seven transactional patterns may be described. This number is neither
exhaustive nor definitive although, for pragmatic reasons, it is fairly
sufficient to characterize a living system. Again, transactional
patterns are aspects of the transactional field as described by per-
spectivism. Adequate transactional patterns are necessary for ade-
quate survival and development.

Episteme is defined as a construct describing and explaining a

part or the whole of a human system or an ecosystem. An episteme
serves as an operational basis for purposeful transactions.

689
 Autonomy is defined as relative structural differentiation and
relative functional independence in transacting.

Hierarchy is defined as vertical order of living systems which

unequal distribution of authority (based on competence) or power
(based on access to and possession of resources).

Cooperation is defined as coordination and integration of pur-

poseful transactions.

Decision-making is defined as choice between alternatives. Every

living system is continuously confronted with numerous choice situa-
tions.

Conflict-management is defined as the dissolution of conflict.

Conflict exists wherever relative or complete autonomy exists.

Strategic coping is defined as the choice and implementing of

synchronic and diachronic combinations of purposeful transactions.

There are six levels of strategic coping. Depending upon cir-

cumstances and organismic programming human systems move from one
level to another whenever strategic coping within one level does not
solve its problems or permit to reach its purposes and goals.

Level 1: normal coping (e.g., working, thinking, communicating

normally). Level 2: neurotic coping (e.g., hysterical, obsessional
coping). Level 3: psychosomatic coping (e.g., stomach ulcers, asthma,
colitis ulcerosa). Level 4: addictive copi~g (e.g., alcoholism, drug
addiction). Level 5: psychotic coping (e.g., schizophrenia, depres-
sion). Level 6: destructive-aggressive coping (e.g., auto-aggression
such as auto-mutilation, suicide; allo-aggression such as torture or
homicide).

PURPOSES AND GOALS OF SYSTEMS THERAPY

The purposes and goals of Systems Therapy are informative morpho-

generation, morphostasis, morphotransformation and morpholysis of va-
riables in that part of the ecosystem which is accessible to therapy
(e.g., individual, couple, family, group).

PHASES OF THE THERAPEUTIC PROCESS OF SYSTEMS THERAPY

There are four major synchronic and diachronic phases of Systems

690
Therapywhlch in redundant cycles and metacycles constitute the the-
rapeutic process: morphogeneration, morphostasis, morphotransforma-
tion and morpholysis of a therapeutic system.

A therapeutic system is a system in which a-ll participants coope-

rate in the ,pursuit of therapeutic purposes and goals. Therapy is
done by a complex transactional process of strategies and counter-
strategies produced by the therapist and all the members of the
system in treatment.

OUTCOME OF SYSTEMS THERAPY

Few outcome studies of Systems Therapy exist because it is too

young a scientific discipline. Moreover, the methodic problems of
such studies are staggering indeed.

However, outcome studies of specific forms of Systems Therapy

such as the sex therapy of Masters and Johnson (5) and the struc-
tural family therapy of Minuchin et alia (6) indicate that their
therapeutic success by far exceeds that of pharmacotherapy, psy-
choanalysis, behavior modification therapy and other forms of
"psychodynamic", individual therapy.

REFERENCES

1. G. Guntern: 'Tourism, Social Change, Stress and Mental Health in

the Pearl of the Alps. Springer, Berlin-Heidelberg-New York
1979
2. G. Guntern: Das Syngenetische Programm: Verhaltenssteuerung,
Charaktertransformation und sozialer Wandel in der Perle der
Alpen. In: J.Duss-von Werdt, R. Welter-Enerlin (Hrsg.):
Der Familienmensch, Klett Stuttgart, pp. 57-64, 1980
3. J.G. Miller: Living Systems, Me Graw-Hill Book Co., New York 1978
4. G. Guntern: Auto-Organization in Human Systems. In: Behavioral
Science, Vol. 27, Nr. 4, pp. 323-337, 1982
5. W.H. Masters and V.E. Johnson : Human Sexual Inadequacy. Little,
Brown and Company, Boston 1970
6. S. Minuchin, B.L. Rosman, L. Baker: Psychosomatic Families: Ano-
rexia Nervosa in Context. Mass: Harvard University Press, Cam-
bridge 1978

691
HUMAN ANXIETY AND NORADRENERGIC FUNCTION: PRELIMINARY STUDIES

WITH CAFFEINE, CLONIDINE AND YOHIMBINE

T.W. Uhdel, J.-P. Boulengerl, B. Vittonel,

L.J. Siever2 and R.M. Postl

!Biological Psychiatry Branch, National Institute of

Mental Health, Bethesda, MD 20205 and 2Mt. Sinai School
of Medicine, New York, New York

INTRODUCTION

Although several neurotransmitter systems have been implicated

in the neurobiology of animal fear and human anxietyl, the focus
of the present report in this regard is limited to the noradrenergic
system. Several lines of evidence suggest an important role for
noradrenergic function in the neurobiology of fear, anxiety, alarm
and arousal (for review see Uhde et al. 2 • 3 ). Redmond and
Huang4 demonstrated that electrical and pharmacological activation
of the noradrenergic nucleus locus coeruleus (LC) in the monkey
produced fear behaviors similar to those occurring during exposure
to natural threat in the wild. In addition to these behavioral
manifestations of fear during LC activation, Redmond and
colleagues 4 •5 documented temporally-related increases in the
levels of noradrenaline and its metabolite, 3-methoxy-4-hydroxy-
phenylethylene glycol (MHPG), in the brain, cerebrospinal fluid
(CSF) and plasma.

In contrast, ablation and pharmacological inhibition of the LC

produced opposite behavioral and bioch~mical effects, e.g.,
decreased norepinephrine (NE) and MHPG • 5 • As reviewed
elsewhere2-4 , pharmacological agents which decrease (e.g.,
clonidine) and increase (e.g., methylxanthine, yohimbine) LC firing
in the monkey have been reported to have anxiolytic and anxiogenic
properties, respectively, in humans. Despite the poor relationship
between anxiety and many biochemical and physiological
variables 6 • 7 , these data suggest that NE and MHPG may be useful
correlates of noradrenergic activity and various forms of human

693
arousal. Consistent with this hypothesis, several investigators
have found an association between anxiety and correlates of
noradrenergic function. For example, elevations in plasma and CSF
NE have been found in psychiatric patients, including depressed
patients with pathological degrees of anxiety8, 9 • A similar
relationship between anxiety and plasmalO and CSFll levels
of MHPG has been reported in normal volunteers. In phobic-anxious
patients a significant correlation has also been reported between
plasma MHPG levels and ratings of anxiety during panic
attacksl2. Despite a number of nonspecific variables which in-
fluence the noradrenergic system13 , 14 , plasma MHPG may provide
a useful index of peripheral noradrenergic activity. Although con-
troversiallS-17, some investigators have suggested also that
plasma MHPG may be derived in sy~s~Bntial part from the CNsl8
and reflect central NE turnover , • Other research, too ex-
tensive to review here21 also suggests an important role for the
noradrenergic system in the neurobiology of anxiety.

In order to explore possible noradrenergic mechanisms of human

anxiety, the authors investigated and report in this paper the
behavioral and biochemical effects in humans of agents that either
increase (caffeine, yohimbine) or decrease (clonidine) noradrenergic
function in animals. Thus, the effects of caffeine and yohimbine
might represent "models" of anxiogenesis, while clonidine, an
alpha-2 adrenergic agonist that decreases LC firing, was investigat-
ed as a possible new anxiolytic agent.

METHODS

Under placebo-controlled conditions, single-dose challenges of

caffeine, a xanthine derivative (240-720 mg p.o.), yohimbine, an
alpha-2 adrenergic antagonist (20 mg p.o.), and clonidine, an
alpha-2 adrenergic agonist (2 ~g/kg i.v.), were blindly and randomly
administered to normal controls and patients whose symptoms met
Research Diagnostic Criteria (RDC) and Diagnostic and Statistical
Manual of Mental Disorders (DSM-III) criteria for panic disorder and
major affective illness, respectively. All patients and controls
were on a low monoamine diet for at least three days prior to the
challenges; in addition, those participating in the caffeine chall-
enge eliminated caffeine from their diets for at least one week
prior to the study. All affectively ill patients experienced patho-
logical degrees of anxiety concurrent with or separate from their
depressive episodes. State anxiety was assessed by the Spielberger
State-Trait Anxiety Inventory. Norepinephrine was assayed by high
performance liquid chromatography (HPLC) by Dr. Markku Linnoila at
NIMH. MHPG was assayed at NIMH in collaboration with Dr. David
Jimerson using his sensitive, specific gas-liquid chromatography-
mass spectrometry method22.

694
RESULTS

Caffeine: Three separate single-dose caffeine (240-720 mg

p.o.) challenges were administered to 8 normal subjects and 2 panic
disorder patients. In the combined group of patients and normal
controls, there was a significant dose-related anxiogenic effect
(ANOVA, p < .04) which was also apparent in the normal controls as a
separate group (ANOVA, p < .04). Although the 2 panic disorder
patients were more sensitive to caffeine, 2 normal controls devel-
oped unequivocal panic attacks after receiving caffeine at 720 mg.
These caffeine-related panic attacks were characterized in both nor-
mal controls by an acute fear of imminent death. Hyperventilation
with carpopedal spasm was prominent in one case. In a subgroup of 6
normal controls who received the maximum caffeine dose (720 mg),
there was a significant increase at 1-1/2 hours of plasma NE (paired
t = 3.78; p < .001).

Yohimbine: Six of 7 patients with panic disorder had severe

anxiety, including five patients who had panic attacks, after yohim-
bine, while none had anxiety to this degree or panic attacks while
receiving placebo as part of a separate challenge paradigm (McNamara
test, p < .OS). These yohimbine-related panic attacks were reported
by the patients to be similar to those attacks "triggered" by expos-
ure to phobic stimuli or occurring spontaneously. It is noteworthy
that one patient experienced alternating waves of depression and
panic anxiety after yohimbine. This same patient later had a drama-
tic response to an open trial of clonidine, an alpha-2 adrenergic
agonist which inhibits, rather than activates like yohimbine, LC
firing.

Clonidine: As predicted by the locus coeruleus "model" of

anxiety2 • 4 , clonidine produced a significant decrement in rat-
ings of anxiety in a combined group of anxious-depresse~ ~~tients
and normal controls (paired t = 2.98, p < .006, n = 37) • •
This anxiolytic effect was apparent in subgroups of depressed
(paired t = 2.56, p < .03, n = 18) and panic disorder (paired t =
2.59, p < .OS, n = 6) patients, but not in normal controls.
Although clonidine produced noteworthy drowsiness as measured by the
Stanford Sleepiness Scale, there was no correlation between the
anxiolytic and sedative properties of clonidine. For the entire
group of patients and normal controls, subjects with the highest
baseline levels of plasma MHPG had the greatest decrement in anxiety
after clonidine (r = 0.41, p < .03, n = 28).

DISCUSSION

Results of our ongoing research are the first to clearly demon-

strate that caffeine in sufficient doses may induce anxiety in norm-
al controls. In addition, these results are consistent with previ-
ous investigations 24 • 25 suggesting that yohimbine may have

695
anxiogenic properties in humans. Furthermore, some individuals may
be especially vulnerable to the anxiogenic effects of these
agents26. In contrast to caffeine and yohimbine, clonidine
appears to have anxiolyt.ic effects in a wide spectrum of cl.in.ical
condit.ions includ.ing major depressive illness 2 •3,l0, 23, panic
and generalized anxiety disorders 2 • 27 , and opiate w.ithdrawal
syndromes3,28,29. Consistent with a noradrenergic model of
anxiety, clonidine had its greatest antianxiety effects in those
subjects with the h.ighest baseline values of plasma MHPG.

Although our findings support and expand previous research

suggesting an important role for the noradrenergic system in the
regulation of arousal, alterations in benzodiazepine-GABA and adeno-
sine function30 also may be cr.it.ical in the neurobiology of both
normal and pathological anxiety.

REFERENCES

1. R. Hoehn-Saric, Neurotransmitters in anxiety, Arch. Gen.

Psychiatry 39:735 (1982).
2. T.W. Uhde, J.-P. Boulenger, L.J. Siever, R.L. DuPont, and R.M.
Post, Animal models of anxiety: implications for research in
humans, Psychopharmacol. Bull., 18:47 (1982a).
3. T.W. Uhde, L.J. Siever and R.M. Post, Clonidine: acute chall-
enge and clinical trial paradigms for the investigation and
treatment of anxiety disorders, affective illness and pain
syndromes, in: "Neurobiology of Mood Disorders", R.M. Post and
J.C. Ballenger, eds., Williams & Wilkins, Baltimore (1983).
4. D.E. Redmond, Jr., and Y.H. Huang, Current concepts II. New
evidence for a locus coeruleus-norepinephrine connection with
anxiety, Life Sci. 25:2149 (1979).
5. D.E. Redmond, Jr., Y.H. Huang, D.R. Snyder, and J.W. Maas,
Behavioral effects of stimulation of the nucleus locus coerul-
eus in the stump-tailed monkey Macaca arctoides, Brain Res.
116:502 (1976).
6. J.-P. Boulenger and T.w. Uhde, Peripheral biological correlates
of anxiety, L' Encephale 8:119 (1982a).
7. B.R. Morrow and A.H. Labrum, The relationship between psycho-
logical and physiological measures of anxiety, Psycho!. Med.
8:95 (1978).
8. R.J. Wyatt, B. Portnoy, D.J. Kupfer, F. Snyder and K. Engelman,
Resting plasma catecholamine concentrations in patients with
depression and anxiety, Arch. Gen. Psychiatry 24:65 (1971).
9. R.M. Post, C.R. Lake, D.C. Jimerson, W.E. Bunney, Jr., J.H.
Wood, M.G. Ziegler and F.K. Goodwin, Cerebrospinal fluid norep-
inephrine in affective illness, Am. J. Psychiatry 135:907
(1978).
10. T.W. Uhde, L.J. Siever, R.M. Post, D.C. Jimerson, J.-P.
Boulenger and M.S. Buchsbaum, The relationship of plasma free
MHPG to anxiety and psychophysical pain in normal volunteers,

696
 Psychopharmacol. Bull. 18:129 (1982b).
11. J.C. Ballenger, R.M. Post, D.C. Jimerson, C.R. Lake, P. Lerner,
W.E. Bunney, Jr., and F.K. Goodwin, Cerebrospinal fluid (CSF)
noradrenergic correlations with anxiety in normals, Sci. Proc.
Am. Psychiatr. Assoc. 134:235, 1981, Abstract #96.
12. G.N. Ko, J.D. Elsworth, R.H. Roth, B.G. Rifkin, H. Leigh, and
D.E. Redmond, Jr., Panic-induced elevation plasma MHPG levels
in phobic-anxious patients. Arch. Gen. Psychiatry 40:425
(1983).
13. D.R. Sweeney, J.W. Maas and G.R. Heninger, State anxiety,
physical activity and urinary 3-methoxy-4-hydroxyphenylethylene
glycol excretion, Arch. Gen. Psychiatry 35:1418, 1978.
14. R.M. Post and F.K. Goodwin, Studies of cerebrospinal fluid
amine metabolites in depressed patients: conceptual problems
and theoretical implications, in: "Biological Aspects of
Depression", J. Mende1s. ed., Spectrum, John Wiley & Sons, New
York (1975).
15. P.A. Blombery, I.J. Kopin, E.K. Gordon, S.P. Markey and M.H.
Ebert, Conversion of MHPG to vanillymandelic acid, Arch. Gen.
Psvchiatry 37:1095 (1980).
16. I.J. Kopin, E.K. Gordon, D.C. Jimerson and R.J. Polinsky,
Relation between plasma and cerebrospinal fluid levels of
3-methoxy-4-hydroxyphenylglycol, Science 219:73 (1983).
17. I.J. Kopin, P. Blombery, M.H. Ebert, E.K. Gordon, D.C.
Jimerson, J.P. Markey and R.J. Polinsky, Disposition and metab-
olism of MHPG-CD3 in humans: plasma MHPG as the principal
pathway.of norepinephrine metabolism and as an important deter-
minant of CSF levels of MHPG, in Proceedings of Nobel
Conference on "Frontiers in Biochemical and Pharmacological
Research in Depression", Skokloster, Sweden, June, 1982, E.
Usdin, ed., Raven Press, New York, 1983, in press.
18. J.W. Maas, S.E. Hattox, N.M. Greene and D.H. Landis, 3-Methoxy-
4-hydroxyphenethyleneglyco1 production by human brain in vivo,
Science 205:1025 (1979).
19. D.S. Charney, G.R. Heninger and D.E. Sternberg, Assessment of
a1pha-2 adrenergic autoreceptor function in humans: effects of
oral yohimbine, Life Sci. 30:2033 (1982).
20. J.R. Elsworth, D.E. Redmond, Jr. and R.H. Roth, Plasma and
cerebrospinal fluid 3-methoxy-4-hydroxyphenylethylene glycol
(MHPG) as indices of brain norepinephrine metabolism in
primates, Brain Res. 235:115 (1982).
21. J.A. Gray, The neuropsychology of anxiety, Br. J. Psychology
69:417 (1978).
22. D.C. Jimerson, S.P. Markey, J.A. Oliner, and I.J. Kopin, Simul-
taneous measurement of plasma 4-hydroxy-3-methoxyphenylethylene
glycol and 3,4-dihydroxyphenylethylene glycol by gas chromatog-
raphy mass spectrometry, Biomed. Mass. Spectrometry 8:256
(1981). .
23. L.J. Siever and T.W. Uhde, New studies and perspectives on the
noradrenergic receptor system in depression: effects of the

697
 alpha-2-adrenergic agonist clonidine, Biol. Psychiatry
February, (1984).
24. D.S. Charney, G.R. Heninger and D.E. Redmond, Jr., Yohimbine-
induced anxiety and increased noradrenergic function in humans:
effects of diazepam and clonidine, Life Sci. 33:19 (1982).
25. G. Holmberg and S. Gershon, Autonomic and psychic effects of
yohimbine hydrochloride, Psvchopharmacologia 2:93 (1961).
26. J.-P. Boulenger and T.w. Uhde, Caffeine consumption and anxie-
ty: preliminary results of a survey comparing patients with
anxiety disorders and normal controls, Psychopharmacol. Bull.
18:53 (1982b).
27. R. Hoehn-Saric, A.F. Merchant, M.K. Keyser and V.K. Smith,
Effects of clonidine on anxiety disorders, Arch. Gen.
Psychiatry 38:1278 (1981).
28. T.W. Uhde, D.E. Redmond, Jr. and H.O. Kleber, Clonidine supp-
resses the opioid abstinence syndrome without clonidine-with-
drawal symptoms: a blind inpatient study, Psychiatry Res.
2:37 (1980).
29. M.S. Gold, D.E. Redmond, Jr., and H.D. Kleber, Clonidine in
opiate withdrawal, Lancet 1:929 (1978).
30. J.-P. Boulenger, J. Patel, R.M. Post, A. Parma and P.J.
!1arangos, Chronic caffeine consumption increases the number of
brain adenosine receptors, Life Sci,- 32:1135 (1983).

ACKNOWLEDGMENTS

The authors would like to thank Drs. David L. Jimerson and

Markku Linnoila for their ongoing scientific collaboration with our
research.

698
 SYMPATHETIC-ADRENAL AND PITUITARY-ADRENAL

RESPONSE TO CHALLENGE*

Marianne Frankenhaeuser and Ulf Lundberg

Psychology Division, Department of Psychiatry and Psycho-

logy, Karolinska Institutet, and Department of Psychology
University of Stockholm, S-106 91 Stockholm, Sweden

INTRODUCTION

The sympathetic-adrenal medullary and pituitary-adrenal cortical

systems form the corner stones of modern stress research, with roots
in Cannon's and Selye's work. Both systems are controlled by the
brain. Hence, when a person perceives a change, or threat, or chal-
lenge in the environment, this triggers a chain of neuroendocrine
events. Messages go to the adrenal medulla, which secretes the cate-
cholamines adrenaline and noradrenaline, and to the adrenal cortex,
which secretes cortisol. Catecholamines and cortisol have several key
functions: as sensitive indicators of the stressfulness of person-
environment transactions, as regulators of vital bodily functions
and, under some circumstances, as mediators of bodily reactions
leading to disease. In short, the effects may be adaptive but they
may also be harmful, particularly in promoting cardiovascular
pathology (Kones, 1979).

What is known about the psychosocial conditions that activate

these two systems? Since most of the psychoneuroendocrinologists
engaged in stress research have restricted their work to one of the
two adrenal systems, knowledge about their relative sensitivity to
different demands has until recently been meagre. This is the main
topic of the present paper, but each of the two systems will first be
considered separately.

*The research reported in this paper has been supported by grants

from the Swedish Medical Research Council (Project No. 997), the
Swedish Council for Research in the Humanities and Social Sciences,
and the Swedish Work Environment Fund (Project No. 80/162).
699
SELECTIVE RELEASE OF ADRENALINE AND NORADRENALINE

In our laboratory we focused for a number of years on adrenal

medullary activation under different psychosocial conditions.
Experiments showed increased catecholamine output in subjects exposed
to a variety of stressors: noise, electric shock, parachute jumping,
radial acceleration, crowded trains, performance under time pressure,
cognitive conflict, etc. Results showed that it was the emotional
element in these situations, rather than the physical stimuli per
se, that induced the catecholamine increase. On the whole, the more
intense the emotional experience, whether pleasant or unpleasant,
the larger the catecholamine increase.

Habituation experiments are well suited for studying the rela-

tion between emotional intensity and hormone output. An early
experiment (Frankenhaeuser et al., 1962) showed that adrenaline
excretion was almost directly proportional to the degree of subjec-
tive stress (as measured by a ratio estimation technique) in subjects
habituating to the experience of riding a human centrifuge at 3G.
Noradrenaline excretion was markedly elevated in all sessions but,
unlike adrenaline, showed no tendency to decrease with repeated
exposure. The dissociation between the two adrenal-medullary hormones
in this particular situation appears consistent with their different
adaptive functions, noradrenaline serving to maintain cardiovascular
homeostasis during gravitational stress.

The fact that adrenaline and noradrenaline can be selectively

released by electric stimulation of specific areas in the
hypothalamus (Folkow and Euler, 1954) has raised the question whether
the two amines are differentially associated with different emotional
states. In the 1950's Funkenstein's work (1956) attracted a great
deal of attention. He suggested that adrenaline secretion was
specifically related to anxious reactions and noradrenaline secretion
to aggressive reactions. This idea appeared consistent with the
results of Ax (1953) who, using a number of polygraphic recordings,
found that laboratory situations designed to elicit fear vs. anger
gave rise to different cardiovascular response patterns, which he
described as adrenaline-like vs. noradrenaline-like. Later studies,
however, (e.g., Levi, 1965; Patkai, 1971) lend no support to the idea
that adrenaline and noradrenaline would be selectively released in
anxious and aggressive states. Instead, adrenaline secretion
increases in a variety of emotional states, including both anger and
anxiety, in accordance with the fight-flight function ascribed to
this hormone by Cannon.

Similarly, a rise in noradrenaline may occur in different affec-

tive states, but the threshold for noradrenaline release is generally
much higher than for adrenaline secretion. These conclusions are
based on a series of experiments in our laboratory, where both sub-
jective responses and hormone secretion have been measured (e.g.,

700
Frankenhaeuser, 1979). There is complete agreement between our
results and recent studies by Ward et al. (in press), showing that a
rise in plasma noradrenaline occurs in response to physical stressors
such as the cold-pressor test, whereas plasma adrenaline is much more
sensitive to mental demands. It is important to note that, while both
catecholamines are secreted by the adrenal medulla, the main part of
the circulating noradrenaline is released by the sympathetic nerve
endings.

CORTISOL RELEASE

Clinical and experimental studies show that cortisol levels may

be high in states of anxiety and depression (cf. Sachar, 1975). Pho-
bics, too, respond with increased cortisol secretion when exposed to
their phobic objects as shown in a recent experiment in our laborato-
ry (Fredriksen et al., in prep). Urinary cortisol was significantly
increased in phobic subjects exposed to slides of their phobic
objects (i.e., spiders, snakes, blood, and mutilation) compared to
neutral slides (flowers). The phobic-object slides also elicited more
intense feelings of distress and greater electrodermal activity.

In normal healthy subjects exposed to "everyday stress",

cortisol tends to increase in novel and unfamiliar situations which
evoke feelings of uncertainty and anxiety. This was the case in a
recent study (Lundberg et al., 1981) of 3-year old children who spent
a day in hospital with their parents for a medical check-up. In a
group of 38 children, 15 responded with anxiety when told that they
would be separated from their parents for a brief period of
psychological testing. Although these anxious children were allowed
to remain with their parents, the threat of separation induced a
significantly greater increase in cortisol, as well as in adrenaline
and noradrenaline, in these children than in the other children.
These results agree with those of Tennes et al. (1977), who found
positive correlations between separation anxiety and cortisol
excretion in one-year-old children who were distressed and agitated.

DIFFERENTIAL ACTIVATION OF THE SYMPATHETIC-ADRENAL AND

PITUITARY-ADRENAL RESPONSE

We shall now examine the psychological characteristics of the

different situations that activate either the sympathetic-adrenal or
the pituitary-adrenal system, or both systems. Factor analysis of
self-reports obtained from subjects in a large number of different
experimental situations, has led us to focus on two main components
of psychological arousal: effort and distress (Lundberg and
Frankenhaeuser, 1980).

The experience of effort involves elements of interest, engage-

ment and determination. This implies an active way of coping, a stri-
ving to gain and maintain control. Active coping tends to be accompa-

701
nied primarily by increased catecholamine secretion, whereas cortisol
output tends to remain low, or may even be actively suppressed.

The distress factor involves elements of dissatisfaction,

boredom, uncertainty and anxiety. It is often associated with a
passive attitude, being helpless, tending to give up. Such a state of
noncoping tends to be accompanied primarily by increased cortisol
secretion. Clinical studies (Sachar, 1975) show that depressed
patients often have very high cortisol levels.

How can one achieve the state of effort without distress? Our
data point to personal control as an important modulating factor in
this regard. A lack of control is almost invariably associated with
feelings of distress, whereas being in control tends to reduce
negative psychological feelings. Hence, personal control may act as a
buffer, changing the balance between sympathetic-adrenal and
pituitary-adrenal activity (cf. Frankenhaeuser, 1983).

LOW-CONTROL AND HIGH-CONTROL SITUATIONS

These principles will be illustrated by data from two laboratory

experiments (Frankenhaeuser et al., 1980; Lundberg and Forsman 1979).
One was a low-control situation, designed to induce both effort and
distress. The other was a high-control situation designed to induce
effort without distress.

In the low-control situation, the subjects performed a one-hour

vigilance task, which consisted in pressing a key in response to each
randomly occurring intensity change in a weak light signal. In the
high-control situation, the subjects performed a choice-reaction task
with a high degree of personal control. Every 5 min the subject was
given the opportunity to modify the stimulus rate so as to maintain
an optimal pace throughout the one-hour session. According to
self-reports, experimental arrangements were successful in creating a
work situation where each person felt pleasently challenged and
motivated to perform well.

Self-reports showed that the low-control task induced both

effort and distress. The high-control task induced effort, but dis-
tress fell below baseline, i.e., the work situation was experienced
as more pleasant and stimulating than the nonwork situation.

The neuroendocrine pattern differed in the two situations.

During the low-control task, which induced both distress and effort,
adrenaline as well as cortisol increased. During the high-control
task, which induced effort but no distress, adrenaline increased,
whereas cortisol decreased. In other words, the effort invested was
accompanied by adrenaline increase, whereas the lack of distress was
reflected in the decrease of cortisol. Thus, in this high-control
situation, the pituitary-adrenal cortical system was "put to rest",

702
while the sympathetic-adrenal medullary system was activated.

It is interesting that the same neuroendocrine pattern was seen

in a study (Johansson et al., in prep.) of control-room operators in
a Swedish steel factory, who were exposed to low-control and
high-control situations at work.

A point of special interest is the bidirectional nature of the

pituitary-adrenal response, cortisol either increasing or decreasing
relative to the baseline. In other words, environmental changes can
either elevate or suppress the secretion of cortisol, a phenomenon
first demonstrated in animal studies (Levine et al., 1979) and now
supported by studies of humans.

Our results concerning the dissociation between the functions of

the two neuroendocrine systems are consistent with those of Ursin et
al. (1978), who identified a "cortisol factor" and a "catecholamine
factor" by factor analysis of data from a study of parachute
trainees. In a general way, the results from these studies of human
beings fit the animal model proposed by Henry (Henry and Stephens,
1977), according to which the sympathetic-adrenal system is activated
when the organism is challenged in its control of the environment,
whereas the pituitary-adrenal system is associated with the
conservation-withdrawal response. Further evidence along these
lines has recently been presented by von Holst et al. (1983), who
found that dominant animals, in control of their territory, are high
in catecholamines and low in cortisol. In submissive, anxious animals
the hormone balance was reversed, i.e., the cortisol level was high,
the catecholamine level low.

REFERENCES

Ax, A., 1953, The physiological differentiation between fear and

anger in humans, Psychosom. Med., 15: 433.
Folkow, B., and Euler, U.S.v., 1954, Selective activation of noradre-
naline and adrenaline producing cells in the suprarenal gland
of the cat by hypothalamic stimulation, Circ. Res., 2:191.
Frankenhaeuser, M., 1979, Psychoneuroendocrine approaches to the
study of 'emotion as related to stress and coping, in:
"Nebraska Symposium on Motivation 1978", H.E. Howe and R.A.
Dienstbier, eds., University of Nebraska Press, Lincoln.
Frankenhaeuser, M., 1983, The sympathetic-adrenal and pituitary-adre-
nal response to challenge: comparison between the sexes, in:
havioral Bases of Coronary Heart Disease", T.M.Dembroski,
T.H.Schmidt and G. Blumchen, eds., Karger, Basel, New York.
Frankenhaeuser, M., Lundberg, U., and Forsman, L., 1980, Dissociation
between sympathetic-adrenal and pituitary-adrenal responses to
an achievement situation characterized by high
controllability: Comparison between Type A and Type B males
and females, Biol. Psychol., 10: 79.

703
Frankenhaeuser, M., Sterky, K., and Jarpe, G., 1962,
Psychophysiological relations in habituation to gravitational
stress, Percept. Mot. Skills, 15: 63.
Funkenstein, D.H., 1956, Nor-epinephrine-like and epinephrine-like
substances in relation to human behavior, J. Ment. Dis., 124:
58.
Henry, J.P., and Stephens, P.M., 1977, "Stress, Health, and the
Social Environment. A Sociobiologic Approach to Medicine",
Springer-Verlag, New York, Heidelberg & Berlin.
Holst, v.D., Fuchs, E., and Stohr, W., 1983, Physiological changes in
male Tupaia belangeri under different types of social stress,
in: "Biobehavioral Bases of Coronary Heart Disease", T.M.
Dembroski, T.H. Schmidt and G. Blumchen, eds., Karger, Basel,
New York.
Kones, R.J., 1979, Emotional stress, plasma catecholamines, cardiac
risk factors, and atherosclerosis. Angiology, 30: 327 ••
Levi, L., 1965, The urinary output of adrenaline and noradrenaline
during pleasant and unpleasant emotional states, Psychosom.
Med. 27: 80.
Levine, S., Weinberg, J., and Brett, L.P., 1979, Inhibition of
pituitary-adrenal activity as a consequence of consummatory
behavior, Psychoneuroendocrinol., 4: 275.
Lundberg, U., de Chateau, P., Winberg, J., and Frankenhaeuser, M.,
1981, Catecholamine and cortisol excretion patterns in three
year old children and their parents, J. Hum. Stress, 7: 3.
Lundberg, U., and Forsman, L., 1979, Adrenal-medullary and
adrenal-cortical responses to understimulation and
overstimulation: Comparison between Type A and Type B persons,
Biol. Psychol., 9: 79.
Lundberg, U., and Frankenhaeuser, M., 1980, Pituitary-adrenal and
sympathetic-adrenal correlates of distress and effort, J.
Psychosom. Res. , 24: 125.
Patkai, P., 1971, Catecholamine excretion in pleasant and unpleasant
situations, Acta Psychol., 35: 352.
Sachar, E.J., 1975, Neuroendocrine abnormalities in depressive
illness, in: "Topics in Psychoendocrinology", E.J. Sachar,
ed., Grune and Stratton, New York.
Tennes, T., Downey, K., and Vernandakis, A., 1977, Urinary cortisol
excretion rates and anxiety in normal 1-year old infants,
Psychosom. Med., 39: 178.
Ursin, H., Baade, E. and Levine, S., 1978, "Psychobiology of Stress.
A Study of Coping Men", Academic Press, New York, San
Francisco, and London.
Ward, M.M., Mefford, I.N., Parker, S.D., Chesney, M.A., Taylor, C.B.,
Keegan, D.L., and Barchas, J.D., Epinephrine and
norepinephrine responses in continuously collected human
plasma to a series of stressors, Psychosom. Med., in press.

704
PSYCHOPHYSIOLOGICAL RESPONSE PATTERNS IN ANXIETY

Rudolf Hoehn-Saric and Daniel R. McLeod

Department of Psychiatry and Behavioral Sciences

The Johns Hopkins University School of Medicine
Baltimore, Maryland 21205 USA

Prior to the re-classification of anxiety disorders in

DSM-III, 1 few attempts were made to distinguish subgroups of
patients classified under the broad umbrella of "anxiety neurosis."
The discovery that antidepressants prevent the recurrence of panic
attacks without affecting anticipatory anxiety or simple phobic
reactions 2 led to the separation of "panic disorder" from "genera-
lized anxiety disorder." In a survey of our anxiety clinic popula-
tion, we found that the two disorders differed in cardiopulmonary
symptomatology but not in muscular, gastrointestinal or psychic
symptoms.

Some patients who experience panic attacks at the onset of

their illness suffer later only from generalized anxiety, while
many, perhaps a majority, of those patients who develop panic
disorder have exhibited nervousness and somatic symptoms for years
prior to the first panic attacks. 4 Recent evidence suggests that
patients having generalized anxiety disorder may not form a homo-
geneous group. We have identified a subgroup of anxiety patients
who show predominantly hyperalertness and psychic manifestations
as well as high muscular tension, but are autonomically comparable
to a control population. 5

In this article we present some preliminary findings of a

larger study whose aim is to improve our understanding of autonomic
reaction patterns in generalized anxiety disorder.

705
HETHODS

Subjects

The population consisted of twelve patients diagnosed as having

generalized anxiety disorder. Excluded were patients with major
medical illnesses, history of psychoses, or history of recent
substance abuse. Subjects had to be sufficiently literate to fill
out questionnaires and they had to be off medications affecting the
central and autonomic nervous systems for at least one week. Of
the twelve patients seven were male and five female, eleven white
and one black. The average age was 41.2 (25-58) years with an
average education of 14.4 (12-17) years.

Procedures

All patients were tested in the morning in a semi-dark, temper-

ature-controlled room. Patients sat in a comfortable reclining
chair. Electrodes were attached to measure electromyographic
activity of the frontalis and right gastrocnemius muscles, cardiac
interbeat intervals, respiration, electrodermal activity (from the
left index and middle finger), and two channels of EEGs. Recordings
were processed through a Cyborg Biolab. Blood pressure and pulse
were recorded at intervals by an Infrasonde electronic blood pressure
monitor. Total recorded time was approximately 40 minutes, con-
sisting of a baseline period, startle and habituation to tones,
response to hyperventilation and response to a stressful mental task
(computerized version of the Stroop Test). After the session,
patients were placed on placebo capsules to be taken one capsule
three times a day. They were retested one week later precisely in
the same way as during the first examination.

RESULTS

Several correlations between autonomic measures were observed

in this study. During the first experimental session, five of the
twelve patients showed a drop in systolic blood pressure upon com-
pletion of the stress task (Stroop Test). During the second session,
one week later, only three patients showed such a drop. There were
no demographic differences or differences in anxiety ratings between
the two groups. For those who exhibited increases rather than de-
creases in blood pressure during both sessions following st~~ss~
the correlation between heart interbeat interval range during the
stress task in the two sessions was 0.90; when all twelve patients
were included in the analysis, the correlation was only 0.68.

During the first session, electrodermal activity was correlated

both with absolute systolic blood pressure changes during stress
and with interbeat interval variability during stress. The number

706
 Table 1. Correlations(N=l2)

~xperimental Session I (E 1)

EDA Fluctuations(St): Abs SBP Change(St) r= -.61 p<.025

EDA Fluctuation Ratio(St:Bl): Abs SBP Change(St) r= -.so p<.OS
EDA Fluctuation Ratio(St:Bl): IBI Range(St) r= +.53 p<.OS
EDA Fluctuation Ratio(St:Bl): IBI Range Ratio(St:Bl)r= +.62 p<.025

Experimen~al Session_ II (E£)

EDA Fluctuation Ratio(St:Bl): Abs SBP Change(St) r= .43 p<.lO

Experimental Session I vs. Experimental Session II (E1 :E 2)

IBI Range (St) E1:E2 {N=l2) r= .68 p<.Ol

IBI Range (St) E1:E2, excluding subjects whose SBP
dropped following St,(N=7)E 1 r= .90 p<.OOS

Table 2. Group Comparisons

Ex2erimental Session I
IBI Range(St)
Low(N=6) High(N=6) t sig.

EDA Fluctuation Ratio(St:Bl) 1.4 ± 0.98 2.6 ± 0.80 2.124 <.05

EDA Fluctuations (St) 25.2 ± 7.4 33.8 ± 4.3 2.264 <.025
Abs SBP Change (St) 19.2 ± 12.0 7.8 ± 5.6 1.916 <.OS

EDA Fluctuation Ratio(St:B~)

Low(N=6) High(N=6) t sig.

Abs SBP Change (St) 18.8 ± 12.2 8.2 ± 5.7 1. 767 <.10

Experimental Session II
EDA Fluctuation Ratio(St:Bl)
Low(N=6) High(N=6) t sig.
Abs SBP Change (St) 13.3 ± 9.1 7.3 ± 2.9 1.406 <.10
Heartrate Change (St) 5.2 ± 3.9 9.2 ± 4.8 1.439 <.10

Abbreviations: EDA:electrodermal activity; Abs SBP:absolute syst.BP;

IBI:interbeat interval; St:stress(Stroop Test); Bl: baseline

707
of fluctuations in skin conductance during the stress period and
the ratio of fluctuations during stress to those during baseline
(fluctuation ratio) correlated negatively with the absolute blood
pressure changes during the stress period. The ratio of interbeat
interval range during stress to that during baseline and the
interbeat interval range during stress both correlated positively
with the electrodermal fluctuation ratio. No strong correlations
were found during the second session, but electrodermal fluctuation
ratio was weakly correlated with absolute blood pressure change
during stress.

Patients were divided into two equal groups according to

electrodermal fluctuation ratio between stress and baseline. The
two groups differed on change of absolute systolic blood pressure
during stress in both sessions, and on heart rate change during
stress in the second session. Patients with larger electrodermal
fluctuation changes tended to have smaller systolic blood pressure
changes but greater heart rate changes under stress.

The patients were next divided according to their interbeat

interval range in the first session during the stress task. The
group with the smaller interbeat interval range showed significantly
fewer electrodermal fluctuations during stress, a lower fluctuation
ratio between stress and baseline and greater absolute blood pressure
changes following stress than the high variability group. In the
second session, interbeat interval range failed to divide patients
into distinct groups.

The division of patients according to electrodermal activity

and interbeat interval range failed to show differences on Hamilton
Anxiety Scale 0 , Spielberger State-Trait Anxiety Scale 7 , or Somatic
Symptoms ScaleS ratings.

CONCLUSIONS

While the presented sample is small, certain trends emerged.

Observed relationships between autonomic responses were more
pronounced during the first experimental session, possibly because
of a greater degree of stress in that session than during the repeat
session following a week of placebo. A larger number of patients
showed a decrease in systolic blood pressure at the end of the stress
task during the first session than during the second. This suggests
a stronger vagal rebound effect at the end of the first session.
Electrodermal activity changes were directly related to interbeat
interval changes, i.e., the higher the rate of spontaneous fluctua-
tions during stress and the greater the increase in number of
fluctuations from baseline to stress, the greater was the range of
the interbeat interval under stress. The reverse relationship was
found between electrodermal activity and systolic blood pressure

708
changes under stress; the group with fewer skin conductance
fluctuations showed a greater absolute systolic blood pressure
change. During the second session these relationships were much
weaker; only the correlation between electrodermal fluctuation
ratio and systolic blood pressure changes under stress approached
significance. Interbeat interval ranges from the first to second
session correlated highly, particularly in those patients whose
blood pressure tended to increase under stress during the first
session. However, neither electrodermal activity nor systolic
blood pressure changes correlated between sessions. This suggests
that these measures are more state-related than the interbeat
interval is.
In conclusion, these findings suggest that some autonomic
changes, such as electrodermal activity and blood pressure changes,
are more responsive to stress or novelty, while others, such as
heart interbeat interval, remain predictable over repeated situa-
tions. Further studies are necessary to establish the significance
of those autonomic response patterns with respect to short and long-
term manifestations of anxiety.

References:

1. American Psychiatric Association, "Diagnostic and Statistical

Manual of Mental Disorders, Third Edition, (DSM III),"
Washington, D.C. (1980).
2. D.F. Klein, Anxiety reconceptualized , in: "Anxiety, New Research
and Changing Concepts," D.F. Klein and J. Rabkin, Raven
Press, New York, (1981).
3. R. Hoehn-Saric, Comparison of generalized anxiety disorder with
panic disorder patients, .Psychopharmacol. Bull, 18:104(1982).
4. C.R. Cloninger, R.L. Martin, P. Clayton, A blind follow-up
and family study of anxiety neurosis: Preliminary analysis
of the St. Louis 500, in: "Anxiety, New Research and Changing
Concepts," Raven Press, New York, (1981).
5. R. Hoehn-Saric and B.J. Masek, Effects of naloxone on normals
and chronically anxious patients, Biol. Psychiatry, 16:1041
{1981). .
6. M. Hamilton, The assessment of anxiety states by rating,
Br. J. Med. Psvcho1.,32:50,(1 959).
7. C.D. Spielberger, R.L. Gorsuch and R.E. Lushene, "Manual for
the State-Trait Anxiety Inventory," Consulting Psychologists
Press, Palo Alto, California (1970).
8. R. Hoehn-Saric, Characteristics of chronic anxiety patients,
in: "Anxiety, New Research and Changing Concepts," D.F. Klein
and J. Rabkin, Raven Press, New York, (1981).

709
CLASSIFICATION OF ANXIETY DISORDERS

David V. Sheehan
Massachusetts General Hospital
Harvard Medical School, Boston, Hass.

Kathy E, Sheehan
Institute of Health Professions
Massachusetts General Hospital, Boston, Mass.

The aim of classification is to organize phenomena in a systemat-

ic way into groups where the members of each group have common fea-
tures. In psychiatric classification, the symptoms, signs, etiology,
natural history, family history, and response to treatment should be
similar within each disorder and would differ between disorders. A
good classification should enhance prediction of other clinical fea-
tures, complications and response to treatment. This reduction of many
phenomena to a few categories has heuristic merit in providing an e-
conomy of memory to clinicians and researchers and in enhancing their
ability to anticipate, manipulate and treat the disorder effectively.

Necessary Features of a good Classification

The following features are desirable in a good classification:
(1) Inherently mutually exclusive classes. ie there should be
minimal category overlap
(2) Each class should be internally homogenous
(3) Classification should be comprehensive ie. there should be a
category for every case encountered
(4) It should provide an economy of memory by being simple,clear
&precise & have as few manageable categories as possible.
(5) the characteristics in each group should be preponderantly
constant
(6) Each class should possess the property of naturalness to the
highest degree ie. the entities making up each category should
share the large majority of a large number of features
{7) the classification should permit optimal prediction.
(8) the criteria must be operationalised and have empirical
referents so that theJ can be reliably reapplied with consistency
from one centre to another.
711
 In evaluating the merits of any proposed classification some
of these criteria deserve priority of consideration. It is of
prime importance to have classes that are inherently mutually
exclusive, so that significant category overlap can be avoided.
Failure on this criterion is a serious handicap that seriously
detracts for any superiority on the other criteria.

Problems in Existing Classifi~ation

Anxiety, phobic, hysterical and related neurotic disorders
have attracted an extensive list of diagnostic labels over the
past century. Yet the majority of these labels were used for the
same cluster of symptoms.2 These labels usually reflected the
theoretical orientation or speciality interest of the diagnostician.
It was not until the recent criteria of Feighner et al3, SpitzeB
et al4, and the DSM III5 that operational criteria were precisely
defined in a way that allowed satisfactory interater reliability.
Although enlightened advances in diagnostic precision and nosolog~
they allowed the precision to be misplaced in sorting out the
diagnostic confusion in anxiety and phobic disorders.
The problems with the DSM III anxiety disorders classification
which has been reviewed elsewhere2,6, can be summarised briefly:
(1) There are 21 to 25 diagnoses in the DSM III to which a patient
with anxiety symptoms might reasonably be assigned.2,6. Any
one or a group of the patients symptoms might be selected and a
diagnosis assigned to that cluster.
(2) Some of these categories are allowed to coexist: All told 138
possible combinations of diagnoses are allowed. This is
unmanageable and fails to provide an economy of memory.l
(3) We have shown extensive category overlap2,6 among the above
groups. This suggests that as defined many of these categories
cannot be considered mutually exclusive.
(4) The symptom characteristics of many of the DSM III categories
are not preponderantly constant. For example many patients
with a diagnosis today of generalised anxiety disorder may in
months be diagnosed panic disorder and at a later point hypo-
chondriasis, social phobia or agoraphobia. 1
(5) Some categories do not possess the property of "naturalness"
For example hypochondriasis and depersonalisation may only be
an "unnatural" part of the single class of biological anxiety
disease.
(6) A diagnostic decision tree is provided in the DSM III to aid in
the differential diagnosis of the various groups. However the
decision tree does not allow for the cooexistence of diagnoses
as does the text. For example the text states that 'agoraphobia
or simple phobia may coexist with social phobia"5. The decision
tree logic clearly conflicts with this. This DSM III does not

712
 accomodate the natural fluctuations of this chronic pathological
anxiety disorder. If the DSM III is strictly followed the
diagnosis changes in these patients over time. In effect what
all the past classification have appeared to do is to view this
pathological anxiety disorder cross sectionally at different
points in time as it progressed through the stages of its
natural history. What is needed is a more parsimonious
classification that considers this one pathological anxiety
disorder in longitudinal perspective
A Heurestic Classification
An alternative approach towards a simple clear yet pragmatic
classification that has heurestic merit in guiding research efforts
and clinical management is offered.7 If a patient presents with
anxiety and phobic symptoms the first step is to rule out any medical
illness that could explain the symptoms. This completed, the next
step is to rule out psychotic disorders. The third step is a
pivotal one. The central question here is whether the patient has
or has had during the course of their current condition any spon-
taneous anxiety attacks or spontaneous attacks of some of this
cluster of symptoms eg lightheadedness, even in the absence of any
immediate obvious justifiable stress. If the patient has these
spontaneous attacks then the other diagnostic criteria for this
pathological anxiety disorder can be checked (see Reference 7).
The presence of these spontaneous anxiety (even panic) attacks
suggests a diagnosis of endogenous anxiety (anxiety spectrum
disorder). If such attacks never occur and never have but the
patient only becomes anxious in response to realistic justifiable
stress or in response to one phobic stimulus, this suggests a
diagnosis of exogenous anxiety. Exogenous anxiety may be acute or
chronic. Exogenous phobic anxiety is usually a monosymptomatic
phobia eg animal, insect phobia that occurs in the absence of any
history of spontaneous attacks. Exogenous anxiety is like the
anxiety of normal man. It occurs only in response to a trauma or
stress or is bound to one specific stimulus, but the symptoms do
not occur in other circumstances. The exogenous anxiety is not
diagnosed because psychosocial stressors or stimuli are identified.
Rather it is a residual category, to be used only when spontaneous
anxiety attacks are absent and when a patient fails to meet symptom
criteria for endogenous anxiety (anxiety spectrum disorder).

In contrast endogenous anxiety is associated with polysymptom-

atic anxiety attacks, and frequently followed by a flight response.
The accumulating evidence suggests that the endogenous anxiety cat-
egory may be a genetically inherited metabolic disease, in contrast
to the stress related or entirely stimulus bound exogenous anxiety.

713
Typically patients with endogenous anxiety describe passing through
several stages over several years in the natural history of their
disorder.
Stage I Subpanic symptom attacks
Stage 2 Spontaneous panic attacks
Stage 3 Hypochondriasis
Stage 4 Limited phobic avoidance
Stage 5 Social phobias
Stage 6 Extensive phobic avoidance/agoraphobia
Stage 7 Secondary Depression
The first manifestation of the disorder is frequently sudden surges
of one or two symptoms from the cluster eg tachycardia, shortness
of breath, lightheadedness, that occur without full blown panic
and in the absence of any justifiable provoking stress. If these
persist, eventually these attacks become polysymptomatic some of
the time and are associated with panic and flight response. Because
there is frequently no obvious justifiable stress many develop
illness fears, concerned that their symptoms are a manifestation
of some serious medical problem. With such preoccupations and
repeated help seeking behaviour they are often labeled as hypo-
chondriacal at this stage (Stage 3). If spontaneous panic attacks
persist, pa~ients soon start to fear and avoid situations they
associate with their worst attacks. At first they acquire a few
phobias (Stage 4), but if the spontaneous attacks persist intensely
and frequently, they inevitably progress to sociaL phobias, agor-
aphobia and extensive phobic avoidance behavior (Stages 5&6).
Because there is often no relief available from the relentless
attacks and progressive disability, many patients become increasingly
depressed (Stage 7). The number of phobias acquired appears to be
a function of the intensity, chronicity and frequency of the spon-
taneous attacks. The particular choice of phobias to a significant
(but not complete) extent appears to reflect the situations in
which the spontaneous attacks occured. Patients may progress rapid-
ly through the stages if the panic attacks are intense and frequent.
If the panic attacks remit at a certain stage, the patient usually
does not deteriorate further and may even retrogress to an early
stage of the disorder. With the next relapse of spontaneous attacks
the disorder progresses forward again through its spectrum of man-
ifestations or stages. This longitudinal perspective accomodates
a wide spectrum of manifestations of the disorder that were pre-
viously viewed as discrete diagnostic entities when the disorder
was analysed cross sectionally.

A study was carried out on 119 patients with spontaneous panic

attacks and multiple phobias as they entered two drug treatment
studies. They were asked to rank order the stages they had progressed
through from the first onset of their disorder to the present.
Definition of each stage were provided to operationalise them to be
as close as possible symptomatically with the DSM III definitions
of several anxiety disorders as follows: Stage I (generalised

714
anxiety disorderS), Stage 2 (panic disorder).S Stage 3 (hypochon-
driasis)S, Stage 4 (panic attacks with one or 2 phobias as in
simple phobiaS), StageS (social phobiaS), Stage 6 (agoraphobiaS),
Stage 7 (dysthymic disorderS, of at least 6 months duration).
Although there was variation in the way many rank ordered the stages,
the mode of the rankings was in the order outlined above. This
ranking was not done by blind raters but by the patient and eval-
uating psychiatrists reviewing the history jointly. This finding
is preliminary evidence that we may be dealing here with one dis-
order rather than several. Furthermore this single pathological
anxiety disorder we have labeled endogenous embraces a spectrum of
manifestations previously regarded as separate. For this reason
some may prefer the alternate name anxiety spectrum disorder to
endogenous anxiety. In the descriptive classification of this
disorder it may be reasonable to call the entire spectrum by one
diagnostic label and then assign a stage to the level the patient
has reached in its natural history.

There is preliminary evidence supporting this dichotomous

classification. The age of onset distribution , the sex distribu-
tion, and the treatment response are different in the two kinds
of anxiety.8,9 Exogenous anxiety usually responds to psychological
treatments, an exogenous phobia responds to behaviour therapy
especially exposure treatments. In contrast endogenous anxiety
responds best to antipanic drugs. Exposure treatments may be neces-
sary later to maximize extinction of any phobias that remain.

Final consensus on classification of these disorders is unlike-

ly until their etiology has been fully elucidated. In the meantime
we are forced to seek more or less useful predictors which permit
an economy of memory. The above alternative classification will
undoubtedly be refined, preferably by some type of biological
subtyping that will improve prediction of treatment response. In
the meantime it may offer a clearer starting point for further
biological investigation and a more pragmatic guide to clinical
management of anxiety disorders.

REFERENCES

1. Sokal,R.R., Sneath, P.R. Principles of Numerical Taxonomy.

W.H. Freeman and Company. San Fransisco. 1963. Chapter 7.
pp. 174.
2. Sheehan, D.V., Sheehan, K.E. The classification of anxiety and
hysterical states. Part I. Historical review and empirical
delineation. J. Clin. Psvchopharma~~~-1982: 2(14):23S-244.
3. Feighner,J.P., Robins, E., Guze,S.B., Woodruff, R.A.,
Winokur, G, Munoz, R. Diagnostic criteria for use in
psychiatric research ..Arch. Gen. Psychiat.1972:26:S7-63.

715
4. Spitzer; R.L., Endicott J., Robins,E. Research diagnostic
criteria (RDC) for a selected group of functional disorders.
2nd ed. Biometrics Research. New York State Psychiatric
Institute. Instrument 58. Nov 23,1975.
5. Diagnostic and Statistical Manual of mental disorders DSM III.
3rd ed. American Psychiatric Association.1980.
6. Sheehan, D.V., Sheehan, K.E. The classification of phobic
disorders. Int.J. Psvchiat. Med. 1982-83:12(4):243-264.
7. Sheehan,D.V., Sheehan, K.E. The classification of anxiety and
hysterical states. Part II: Towards a more heurestic
classification. J.Clin. Psvchopharmac .. 1982:2(6):386-393.
8. Sheehan,D.V., Ballenger, J., Jacobson, G. The treatment of
endogenous anxiety with phobic, hysterical and hypochondriacal
symptoms. Arch. Gen. Psychiatry. 1980:37:51-59.
9. Sheehan,D.V., Sheehan, K.E., Minichiello,W.E. Age of onset
of phobic disorders: a reevaluation. Compr. Psychiatry.
1981:22:544-553.

716
ANOREXIA NERVOSA - AN ADDICTION

Detlev Ploog

Max Planck Institute for Psychiatry

Munich, FRG

In a recent review on the aetiology of Anorexia nervosa G.

Hsu listed six main theories which are not mutually exclusive but
conceptually distinct. These are: the sociocultural theory, family
pathology theory, individual psychodynamic theory, the developmental
psychobiology theory, the primary hypothalamic dysfunction theory,
and finally the recently renewed contention that anorexia nervosa
may be related to affective disorders.

To these six a seventh point of view will be added here. This

is the hypothesis that anorexia nervosa is based on addictive
behavior. The disease possesses all characteristics of an addiction
as far as psychopathology, coping strategies and subjective feelings
as well as course and outcome are concerned. This hypothesis, first
offered in 1981, is not mutually exclusive from the others, since
several or all of the named factors may contribute to the develop-
ment of addiction, but it is certainly conceptually distinct and
may lead to a different concept of causation and treatment of
anorexia nervosa. The German diagnostic term Magersucht, which
literally means the addiction to become lean, places emphasis on
addictive behavior as the key mode of this disorder.

The purpose of this paper is to explain this concept ln more

detail.

First, the key role which feeding behavior plays in motivated,

drive-depending behavior and its neural control will be stressed.
It is important to recapitulate the fact that the releasing effect
of food stimuli varies with hunger and satiation, i.e., with the
internal state of the organism. The internal state of anorectics
is deviant from the regular states of hunger and satiation, and

717
consequently are the releasing effects of food stimuli subject to
change.

In primate neonates as well as in lower mammals the search for

the releasing stimulus, the nipple, is preprogrammed as is the per-
ception of the texture and taste of the stimulus. There are observ-
able and measurable signs of hunger and satiation as well as motor
patterns for the ejection of food, i.e., vomiting and spitting.
Feeding behavior of a simple kind can even be observed in rhomb-
encephalic and mesencephalic human infants. However, for goal-
directed behavior, i.e., searching for the nipple and getting
oriented on the mother's body, higher brain structures are neces-
sary. In brief, the basic behavioral patterns and neural control
mechanisms for feeding, food getting, food rejection, and even food
aversion are present at birth. This means that the most important
primary drive system (besides thermoregulation), which yields
pleasure by feeding and pain by hunger, is subject to change in
anorexia nervosa.

The change in the fully developed syndrome results in a

complete perversion of the drive: hunger causes a peculiar kind of
pleasure and eating causes unpleasant feelings in the stomach,
intestines and other parts of the body. The voluntary abstinence
from food causes changes in the central nervous state that may be
induced by changes in the production of gastrointestinal hormones
(e.g., peptides such as cholecystokinin) and their effect on vagus
nerve. The longer the fast continues, the more food is considered
an ambivalent matter - something repulsive and yet something that
completely occupies the mind of the patient. This extremely ambiv-
alent state is sometimes overridden by attacks of bulimia wherein
food stimuli evoke unrestrained voracious eating, in extreme cases
rotten, fetid meat and even the girl's own feces are devoured. This
motivational state is usually followed by vomiting which can also
occur after smaller meals.

If one compares signs and symptoms of anorectics with what is

known about lateral and medial hypothalamic functions and feeding
behavior in mammals, including monkeys, one can draw parallels
that may help to explain the brain and behavioral mechanisms of
this perverted behavior.

First, the well-known electric brain stimulation and self-

stimulation experiments in animals ought to be considered here.
Dopaminergic and noradrenergic neurons in various parts of the
hypothalamus mediate self-stimulation. Both of these catecholamine
systems are also responsible for the control of ingestive behavior.
Self-stimulation at sites related to the medial forebrain bundle is
typically rapid, may be accompanied by hyperactivity, and can be
modulated by hunger and thirst. Between the lateral hypothalamus,
which induces feeding behavior when stimulated, and the ventro-

718
medial hypothalamus which inhibits feeding when stimulated, is a
large transitional area where a mixture of positive and aversive
effects in the self-stimulating animal can be seen. According to
Olds and others, it is this in-between area where drive-targets
are changeable, either by the availability of the drive object,
as Valenstein has shown, or by conditional learning. That the
lateral hypothalamus takes part in the process of conditional
learning in regard to food objects has been firmly established by
Rolls et al. The discharge freQuency of a certain nerve cell pop-
ulation changes drastically if the hungry monkey sights food; the
cells cease firing when the monkey is satiated. If certain inedible
objects are temporarily associated with food objects, the cell
starts firing in the presence of the conditional stimulus. Neurons
that respond to food stimuli are also activated by electric stim-
ulation of those brain sites that yield reward in self-stimulation.
The brain stimulus mimics the effects caused by the sight of food.

The reward system in the brain, as determined by self-stim-

ulation experiments, and the direct hypothalamic involvement at
the cellular level in food-getting and food-rejecting behavior
serves as the neurobiological mechanism for developing and sus-
taining anorexia nervosa. Food stimuli and objects related to them
can take on an ambivalent or aversive meaning but, nevertheless,
excite the hypothalamus. The obsessive preoccupation with thoughts
about food, the compulsive evaluation of its content, and the
change in its significance as drive target point to the persistent
self-stimulation of the hypothalamus with mixtures of positive and
aversive effects. The bizarre and finicky eating habits, and the
swing from lustful fasting to craving for food are seen as conse-
QUences of a constant hypothalamic bombardment of nervous impulses
comparable to the effects of self-stimulation. Another irritation
of the hypothalamus may come through visceral afferent pathways.
Gastrointestinal peptides which may change in prolonged or chronic
hunger states not only affect adjacent endocrine cells but also
sympathetic and vagus fibers, including their afferent pathways to
the catecholamine systems originating in the brainstem. Yet an-
other indication for hypothalamic dysregulation and imbalance of
the transmitter systems is the disturbed thermoregulation, whose
intricate relationship with food intake has been carefully deter-
mined. The remarkable hyperactivity of many anorectics doing
physical exercises to the extreme may not only help to consume
caloric energy and produce heat but may be part of the self-stimu-
lation comparable to compulsive running. It has been assumed that
the elated mood of the so-called obligatory runners is caused by
high endorphin levels during and after running. The same could be
true for anorectics whose addicted motivational behavior has been
considered as an analog to obligatory running.

In concluding, the hypothesis is advanced that the behavior

of anorectics "runs" in a vicious circle which is similar to

719
addictive behavior. It results in perverted hedonic responses: food
stimuli are perceived as unpleasant and hunger as pleasant. The
craving for food and the stubborn determination not to eat it lead
to a continuous neural bombardment of the hypothalamus similar to
self-stimulation effects in animals. This may be brought about by
peptides of the gut under starvation and by persistent ideation of
food objects with ambivalent or repellent properties. The perverted
hedonic state can collapse whereupon food stimuli become unquali-
fied and overly attractive, and disinhibited eating occurs. This
may result in bulimic attacks which are experienced as punishment.
This anorectic vicious circle is very similar to other forms of
addiction, and so is the course and outcome of the disease. Chronic
cases show loss of zest, appear to be indifferent if not unhedonic
and live a vita minima. Although anorexia nervosa is psychogenic
in origin, unlike neurotic behavior it develops easily into a psycho-
organic state. As in other addictions, there is up to now no simple
causal explanation for the disease, nor is there a single effective
therapy which guarantees success. Treatment programs based on the
recognition that anorectics are addicts are required.
REFERENCES

Hsu, L.K.G., 1983, The aetiology of anorexla nervosa, Psychological

Medicine 13:231.
Olds, J., 1977, Drives and Reinforcements: Behavioral Studies of
Hypothalamic Functions, Raven Press, New York.
Ploog, D., 1981, Neuroethological aspects of anorexia nervosa, in:
Biological Psychiatry, C, Perris, B. Jansson, G. Struwe, eds.,
Elsevier/North-Holland Biomedical Press, Amsterdam.
Rolls, E.T., 1976, The neurophysiological basis of brain-stimulation
reward, in: Brain Stimulation Research, A. Wauquier,
E.T. Rolls, eds., North Holland/American Elsevier.
Valenstein, E.S., 1970, Stability and plasticity of motivational
systems, in: The Neurosciences: Second Study Program,
F.O. Schmitt, ed., Rockefeller University Press, New York.

720
DEPRESSION, ANOREXIA NERVOSA AND NUTRITION: EFFECTS OF STARVATION

ON ENDOCRINE FUNCTION AND MOOD

M.M. Fichter~·:, K.M. Pirke~'d:, F. HolsboerMo':,

W. Kemp in~·: and W. Weiss~·:

1: Dept. Psychiatr. Uni v. Munich, Nussbaumstr. 7 8

Munich 2, FRG
** Max-Planck-Inst.of Psychiat.,Kraepelinstr.10
8 Munich 40, FRG
***Dept. Psychiatr. Univ. Mainz, Langenbeckstr.1
65 Mainz, FRG

INTRODUCTION

A substantial body of literature has emerged concerning dis-

turbances in neuroendocrine functions in anorexia nervosa
(Vigersky, 1977; Beumont & Russell, 1982; Weiner, 1983). Although
there are some indications that many of these disturbances are
secondary to the reduced calorie intake, this notion has been
disputed by others. In order to clarify the effect of starvation
on neuroendocrine functions, an experimental study was conducted
with healthy volunteers. Results of this study also raise the
question as to what extent reduced calorie intake or weight loss
are responsible for mental endocrine disturbances in other mental
disorders associated with weight loss such as (endogenous) de-
pression or alcoholism.

METHOD

Five healthy subjects age 21 to 25 years (23.2 +/- 2.2) were

thoroughly screened by psychological and medical testing. They
were observed during a baseline phase (A), a three week period of
complete food abstinence (B), a phase of weight gain (C) and a
final baseline phase (D). The study was approved by the ethics
committee of the University of Munich. Subjects were healthy
volunteers and informed consent was obtained. During baseline

721
phase (A) body weight was kept stable at 103% +/- 2.7% ideal body-
weight (IBW) and at 102% +/- 2.3% of IBW during the final phase
(D). During the three week phase of complete starvation (B) the
subjects lost on the average 8.0kg of weight. At the end of the
fasting phase (B) they weighed on the average 49.6 +/- 4.1kg
(87.8 +/- 1.7% of IBW). At least twice in each of the experimen-
tal phases A, B, C and D the following endocrine assessments
were performed: 1.) Dexamethasone-Suppression-Test with applica-
tion of 1.5mg Dexamethasone at 11 pm and blood samples at 9 a.m.,
11 p.m. (day 1) and 9 a.m., 4 p.m. and 9 p.m. the following day
2.) Stimulation of TSH with 200ug Protirelin (TRH) and 3.) Stimu-
lation of growth hormone (HGH) with the alpha-adrenergic receptor
agonist clonidine (0.15mg i.v.). Before fasting, after fasting
and after weight gain blood samples were collected at 30 minute
intervals over 24 hours to determine the 24-hr-secretory pattern
of cortisol, luteinizing hormone (LH), growth hormone (HGH),
thyrotropin (TSH) and prolactin. Details of the radio-immuno-
assay procedure for determining hormone levels and results for
LH, HGH, TSH and prolactin secretion will be described in a
forthcoming paper.

For assessment of the mental state depression scores were

obtained weekly using the Depression Scale for Selfrating by von
Zerssen (1976) and the Hamilton Depression Scale (Hamilton,
1967); in the latter scale items related to weight were omitted
(7.4; 14; 16a & b; 17). In addition the Hamilton anxiety scale
(Hamilton, 1959), an adjective checklist concerning the general
feeling of wellbeing by von Zerssen (1976) and a lOcm bipolar

• 51 kg (g8'(. IBW)

250 o 46 kg (88 '/. IBW)

-
:::
..:-"' 200

...0
en 150
-1-
a:
0
u 10
c(
~
en 50
...
c(

a.
- ,r~------~21~o------2~~------~~-------~r-------,r,2------~,~

Fig. 1: 24-hr-plasma cortisol in subject 01 before fasting (51

kg) and after fasting (46 kg).

722
visual analog scale measuring the feeling of wellbeing were given
weekly to the subjects. Nonparametric statistical analyses were
performed (WILCOXON test for paired ranks), whenever N=5 and if
not indicated otherwise. The probability values reported in the
text are for onesided testing. Using the Simpson formula statis-
tical analyses for the 24-hr-plasma hormone levels were based on
the "area under the curve".

RESULTS

Hypothalamo-Pituitary-Adrenal-Function:
Dexamethasone-Suppression-Test: Figure 2 shows the results of
the dexamethasone tests, which were consecutively performed in
each place in a typical subject. Dexamethasone suppression was
abnormal already after minimal acute weight loss (97% IBW) and
normalized quickly within days with resumption of food intake.
Considering the results for all five subjects all 11 DST's per-
formed during the initial baseline phase were normal (cortisol
level below 5 ,ug/100 ml); half of the 14 DST's performed during
the fasting phase showed abnormally high cortisol values (above 5
pg/100 ml); with weight gain (C) following the fasting phase
normal suppression of cortisol was again observed in all 11
DST's. During the final baseline phase (D) 8 out of 9 DST's (17
out of 18 cortisol probes between 4 p.m. and 9 p.m.) were normal.

FASTING GAIN BASELINE

A, (104"1.18W)
A2 ( 104"1.1 BWl
I
_J I
QIS I
l/)
1-
a::
0
u
<t
:l:
l/)
<{ 2 (103"1.1BWl
...J
a..
9 23 9 16 23 9 23 9 16 23 9 23 9 16 23 9 23 9 16 23CLOCK TI~E

Fig. 2: Plasma cortisol levels following application of 1.5 mg

dexamethasone in subject 3 during initial baseline,
fasting, weight gain and final baseline.

723
 The increase of the cortisol levels during fasting were statisti
cally significa nt (df = 4; p (.05). Abnormal cortisol response s
to dexameth asone have also been reported for anorexia nervosa
(cf. Doerr et al., 1980).

The 24-hour- Cortisol Pattern showed basically the same

changes during starvatio n in healthy subjects as have been des-
cribed for anorexia nervosa in the acute stage of the illness
(Walsh et al., 1978; Doerr et al., 1980). The 24-hr.-pl asma
cortisol level (mean and standard deviation ) showed a rise from
7.33 +/- 4.2 ~g/100 ml at baseline (A) to 12.41 +/- 2.9 pg/100 ml
after fasting (B). After weight gain the cortisol levels had
normaliz ed to 6.75 +/- 4.2 pg/100 ml (df = 4; p (.05). Cortisol
half life in plasma showed a significa nt increase during the

'i
HAMA
ltASoE LilliE FAS.TI 6 AAIN Sl.~l

' '· sE•

' A 6 c 0
I

n
A
BASELINE I
HAMilTON
FASTING
DEPRESSION
B

SCALE
GAIN
C 0
BASELINE II
'
'
Ill. ' I
' ' '
. ' I
'
1. SEN

Dl!i.IRESS
II
56 l BEFINOLICHKEIT IBSIBS" von ZERSSENI 085
II, SEMI •es·

• ' l • • • • I. II l l l l I I
r
Fig. 3: Ratings in the Fig. 4: Ratings in the
Hamilton Depressio n Scale Hamilton Anxiety Scale, (HAMA) the
and Selfratin g of Depressio n Visual Analog Scale of Wellbein g
(mean and standard error) and the von Zerssen Distress
during the 4 phases of the Scale (mean and standard error).
study.

724
starvation phase (df = 4; p (.05). There were also indications of
an increase of cortisol production rate: The total amount of time
spent in secretory activity had increased after fasting (df = 4;
p <.05) and the number of secretory episodes in 24 hours showed
an increasing trend during starvation. The 24-hour-secretory
pattern was disturbed and there were secretions at unusual times.

Hypothalamo-Pituitary-Thyroid-Function:
While the 24-hr.-TSH-pattern did not show any major changes
after fasting, the basal TSH values before stimulation with
protirelin (TRH) were significantly lowered (df = 3; t-test for
dependent samples; p <.05). The basal TSH levels increased again
significantly with weight gain (df = 3; p <.05). The TSH incre-
ment after stimulation with TRH was significantly smaller after
fasting than before fasting (df = 3; p <.05) and after weight
gain (df = 3; p <.05). Only 30 minute values post-stimulation
were determined indicating a blunted TSH-response; we cannot
comment on a possible delayed response. Recently Casper & Frohman
(1982) have reported blunted TSH-responses to TRH in anorexia
nervosa, which may persist even after physical recovery. Our data
from starving healthy subjects shows that fasting by itself can
induce a blunted TSH-response to TRH. Basal TSH levels in
anorexia nervosa have been reported as normal, while we have
found indications for lowered basal TSH levels during starvation.
More refined and longitudinal studies with larger numbers of
anorexia nervosa patients may reveal lowered TSH levels as well.

Growth-Hormone-Response to Clonidine
The growth-hormone-response after stimulation with clonidine was
slightly elevated after fasting. A similar finding for anorexia
nervosa has been reported by Brambilla (1983). Most interesting
in our data with healthy volunteers was the significantly reduced
HGH-response to stimulation with clonidine after the subjects had
regained their normal weight ( df = 2; p<. 05; paired t-Test).

Mental State during Starvation

Our subjects were kept on our research ward as "inpatients"
during the 3 week fasting phase. This must be kept in mind in
interpreting the finding of an increase in the score of the
Hamilton and the von Zerssen Depression Scale (see fig. 3).
Figure 4 shows the increase in the Hamilton Anxiety Scale, the
Visual Analog Scale measuring the general feeling of wellbeing
and the General Distress Scale by von Zerssen. Our data on mental
changes during fasting are limited and to incomplete to allow
interpretation either as the result of starvation or as a conse-
quence of distress following a change in the environment.

DISCUSSION

Our data support the hypothesis that neuroendocrine

725
disturbances in anorexia nervosa are secondary to reduced food
intake and/or changes in body weight. Our results with healthy
subjects during starvation were in agreement with neuroendocrine
disturbances , which have previously been described for anorexia
nervosa and depression. "Poor appetite or significant weight loss
when not dieting" (DSM III) are usually considered typical symp-
toms of depression. If experimental starvation by itself can
induce the same neuroendocrine disturbances, which have previously
been described for endogenous depression or major affective dis-
orders, starvation must be considered as at least one major cause
of endocrine disturbances in depression.

Disturbances in the 24-hr-cortisol secretion have been reported

by Sachar et al. (1973) and many other authors. These changes in
depression are quite similar to those observed in anorexia nervosa
Doerr et al., 1980) and in starving healthy subjects. The cortisol
production rate and the number of secretory episodes are in-
creased. There are indications for an increase of cortisol plasma
half-life in anorexia nervosa (Doerr et al.) and in depression
(Sachar et al., 1973). Insufficient suppression of cortisol fol-
lowing application of dexamethasone has within the last years
been described as a biological marker for (endogenous) depression
in a vast number of reports (Carroll et al., 1981). Considering
the powerful effect, which reduced calorie intake and weight loss
appears to have on the pituitary adrenal axis, it is amazing that
only a few reports have considered this factor systematically
(Berger et al., 1982; Holsboer. et al., 1983). Reduced calorie
intake and weight loss will definitely contribute to neuroendo-
crine disturbances in those depressed patients who showed these
symptoms. We have observed in our healthy subjects that a normal
DST may become abnormal (and vice versa) very rapidly within
days. Abnormal DST's have also been reported for alcoholism,
where malnutrition may play a major part (Swartz & Dunner, 1982)
and bulimia (Gwirtsman et al., 1983; Hudson et al., 1983), which
is often associated with larger fluctuations in body weight.

Blunted TSH response to TRH has been reported in depressed

patients by Loosen & Prange (1980) and the TRH-test has been
suggested as a marker for endogenous depression. A blunted TSH
response was clearly established in our healthy subjects during
fasting and the same has been described for anorexia nervosa
(Casper & Frohman, 1982) and alcoholism (Loosen et al., 1979).

A blunted growth hormone response to stimulation with clonidine

has been described for endogenous depression by Matussek et al.,
(1980) and Checkley et al., (1981). In starving healthy subjects
we have also observed a diminished HGH response to clonidine in
experimental phase D, when subjects had already regained their
weight. The pituitary-adrenal and -thyroid axis appear to react
quickly (within days) to changes in calory intake, while other

726
hormonal axes show a timelag and slower adaptation to changes in
weight. A v~ry detailed history of calorie intake and body weight
is essential for endocrine research in depression. According to
our data (complete) starvation has major effects on neuroendo-
crine functions in healthy subjects. Calories intake and changes
in the weight definitely constitute one major factor, which can
account for disturbances in endocrine functions described for
depression and other disorders associated with changes in weight.

REFERENCES

Berger, M., Krieg, C. & Pirke, K.M.(l982), Is the Positive Dexa-

methasone in Depressed Patients a Consequence of Weight Loss.
Neuro Endocrinology Letters, Edition Medizin.
Beumont P.J.V. & Russell J.(l982), Anorexia Nervosa. In: Beumont
P.J.V. & Burrows, H. (Eds.) Handbook of Psychiatry and Endo-
crinology. Elsevier Biomedical Press, North Holland.
Brambilla F., Canagnini et al.(l983), Hypothesis on Neurotrans-
mitter Disturbances in Anorexia Nervosa. Paper Presented at the
VII. World Congress of Psychiatry, Vienna.
Carroll, B.J., Feinberg et al.(l981), A Specific Laboratory Test
for the Diagnosis of Melancholia. Arch. Gen. Psychiatry, 38, 15-
22.
Casper R.C. & Frohman L.A.(l982), Delayed TSH Release in Anorexia
Nervosa Following Injection of Thyrotropin-Releasing Hormone
(TRH). Psychoneuroendocrinology, 7, 59-68.
Checkley, S.A., Slade, A.P. & Shur, E.(l981), Growth Hormone and
other Responses to Clonidine in Patients with Endogenous
Depression. Brit. J. Psychiat., 138, 51-55.
Doerr P., Fichter, M.M. et al.(l980), Relationship Between Weight
Gain and Hypothalamic Pituitary Adrenal Function in Patients with
Anorexia Nervosa. J. Steroid Biochem. 13, 529-537.
Gwirtsman, H.E., Roy-Byrne, P., Yager J. & Gerner, R.-H. (1983),
Neuroendocrine Abnormalities in Bulimia. Am. J. Psychiat., 140,
559-563.
Hamilton, M.(l959), The Assessment of Anxiety States by Rating.
Brit. J. Med. Psychol., 32, 50-55.
Hamilton, M.(l967), Development of a Scale for Primary Depressive
Illness. Brit. J. Soc. Clin. Psychol.,6, 278-296.
Holsboer F., Steiger, A. & Maier, W. (1983), Four Cases of
Reversion to Abnormal Dexamethasone Suppression Test Response as
Indicator of Clinical Relapse: A Preliminary Report. Biol.
Psychiat., 18, 911-916.
Loosen P.T., Prange, A.J. & Wilson, I.C.(l979), TRH (Protirelin)
in Depressed Alcoholic Men. Arch. Gen. Psychiat., 36, 540-547.
Loosen. P.T. & Prange, A.J.(l980), Thyrotropin-Releasing Hormone
(TRH): A Useful Tool for Psychoneuroendocrine Investigation.
Psychoneuroendocrinol., 5, 63-80.
Matussek, N., Ackenheil, M.' et al.(l980), Effect of Clonidine on

727
Growth Hormone Release in Psychia~ric Patients and Controls.
Psychiatry Research, 2, 25-36.
Pirke, K.M., Fichter, M.M. et al.(1979), Twenty-Four-Hour Sleep-
Wake Pattern of Plasma LH in Patients with Anorexia Nervosa. Acta
Endocrinol., 92, 193-204.
Sachar, E.J., Hellmann, L. & Roffwarg, H.P.(1973), Disrupted 24-
hour Patterns of Cortisol Secretion in Psychotic Depression.
Arch. Gen. Psychiatry, .28, 19-24.
Swartz C.M. & Dunner, F.J.(1982), Dexamethasone Suppression
Testing of Alcoholics. Arch. Gen. Psychiat., 39, 1309-1312.
Vigersky, R.A. (Ed.)(1977), Anorexia Nervosa. New York, Raven
Press, Walsh, B.T., Katz, J.L. et al.(1978), The Production Rate
of Cortisol Declines during Recovery from Anorexia Nervosa. J.
Clin. Endocrinol. Metabol., 53, 203-205.
Weiner, H.(1983), Abiding Problems in the Psychoendocrinology of
Anorexia Nervosa. Report of the 4th Ross Conference on Medical
Research, Ross Laboratories, Columbus.
Zerssen, D.v. (unter Mitarbeit von Koeller, D.M.)(1976),
Klinische Selbstbeurteilungs-Skalen (KSb-S) aus dem MUnchner
Psychiatrischen Informationssystem. Beltz, Weinheim.

728
PERCEPTIONS OF THE BODY IN ANOREXIA NER VOSA

Paul E. Garfinkel and David M. Garner

Toronto General Hospital, Department of Psychiatry, tO 1

College Street, Toronto, Ontario M5J 1L7, Canada

The clinical features of anorexia nervosa are fairly similar, whether one
reads descriptions from Morton (1694) or Gull (1868) or one sees patients
today. Because of this clinical similarity, people have erred in looking for a
single pathogenesis to the illness. In the sense of shared symptomatology,
anorexia nervosa is clearly a discrete psychiatric syndrome, but as for other
syndromes this does not imply a single pathogenesis. Rather, in any
population there will be a group of individuals at risk for anorexia nervosa
because of the presence of a specific combination of predisposing factors. It
is the interaction and timing of these phenomena within a given individual
which are necessary for the person to become ill. In this sens, anorexia
nervosa is a final common pathway, the product of a group of interacting
forces.
We have previously described a variety of factors within the individual, the
family and culture which have been purported to play a role in the
pathogenesis or perpetuation of anorexia nervosa (Garfinkel and Garner,
1982). One such area within the individual, that of body perceptions, will be
examined in this paper by reviewing what is known of 2 major components of
this: body image and interoception.

Body Imag_e
Body image is a complex construct which has been approached from many
points of view (Garner and Garfinkel, 1981). These include the neural
representation determining bodily experience (Head, 1920), the mental
image that one has of one's body (Traub and Orbach, 1964), the feelings one
has about one's body (Secord and Jourard, 1953) and a personality construct
(Fisher and Cleveland, 1958; Kolb, 197 5; Schilder, 1935). Reference to a
disturbance in body image in anorexia nervosa dates back to Lasegue (1873):
"the patient, when told that she cannot live upon an amount of food that

729
 SELF ESTIMATION OF BODY SIZE

N=265
40
e e ANOREXIA NERVOSA
• • • • • •e NORMAL CONTROLS
...

.
•• ••
w
-'
30
:
••
·.••
.
0..
::>-
<(
·.••
••
·.
Vl (21.9)
LJ..
:
. .••
0
20 ·~
1-
z • •
••
w
()
.r • ••
••
••
w
0..
••
..••
•
10
,. .. ..•••• ••

. ••
••
••
•••• ••
0~&-~--~----~--~----~--~~---L_J~

+20 ·+15 +14 ·+10 +9 •+5 +4 ·0 ·1 ··5 ·6 •-10 -11 ··15 ·16~-20

OVERESTIMATION SELF ESTIMATION IN PERCENT UNDER ESTIMATION

(4 TRIALS)

Fig. 1.

730
would not support a young infant, replies that it furnishes sufficient
nourishment for her, adding that she is neither changed nor thinner". Bruch
(1962) however first recognized disturbed body image to be characteristic of
anorexia nervosa. While she considers it to be related to a more general
misperception of internal state, specifically it involves the patient's
inability to recognize her appearance as abnormal.
If a broad definition of the body image construct is used, the disturbance in
anorexia nervosa can be found to have different forms of expression which
may operate independently of conjointly. The first is "perceptual" and refers
to the degree to which the patient is not able to assess her size accurately.
The second involves cognitive and affective components without any obvious
signs of "perceptual" mediators, and refers to some patients who assess their
physical dimensions accurately but react to their bodies with extreme forms
of disparagment or occasionally aggrandizement.
The most perplexing, yet most commonly described abnormality is the
patient's apparent inability to recognize how thin she has become. Bruch
(1973) refers to this as "disturbed size awareness"; the patient simply does
not recognize that she has become emaciated. Some patients display a
variation of this phenomenon in which the overestimation seems to be
restricted to a particular part or parts of the body. Selected areas such as
the stomach or thighs are magnified and seen as disproportionate to the rest
of the body. These patients will acknowledge that in general they appear
emaciated, but that further dieting is necessary to eliminate their
"protruding stomach". Body image disturbance may be manifest in other
patients who perceive their sizes relatively accurately but who exhibit an
extraordinary loathing for all or parts of their body. This goes well beyond
the dissatisfaction with their appearance common for Western women
(Berscheid, et al, 1973), to the point of revulsion with one's shape. A further
expression of body image disturbance may occur either with normal size
perception or with overestimation. This involves an exaggerated pleasure
with, and overvaluation of, a thinner shape. These patients see their low
weight and thinness as an exceptional achievement. Later this pleasure in
low weight is replaced by a fear of any gain in weight.

Body Size Estimates as an Index of Body Image

A variety oTiechniques have been used to assess body image objectively;
these have recently been described elsewhere (Garner and Garfinkel, 1981;
Garfinkel and Garner, 1982) and will not be detailed here. In general,
disturbed body image in anorexia nervosa has been defined as an
overestimation of body parts or body size. We will confine this discussion to
the latter and review work that has been done using a distoring photograph
technique.
Glucksman and Hirsch (1969) had obese and normal subjects estimate their
body sizes using a projected photograph which could be distorted along the
horizontal axis. The image could be made to look anywhere from 20%
"thinner" to 20% "fatter" than its actual size. Glucksman and Hirsch (1969)
initially found that six dieting obese subjects overestimated their body sizes
in comparison with four controls. Garner et al (1976) applied this technique
to the study of body size perception in anorexic and obese patients and

731
 found overestimation tendencies in many anorexics. Further, this
overstimation of the body size was confined to the patient herself and did
not include her estimates of inanimate objects or other people. More
recently, Garfinkel et al (in press) have assessed body size estimates in the
parents of anorexics,the anorexics themselves and comparison non-anorexic
families (Table 1). Overestimation occurred only in the anorexic patients
and not their parents • Moreover, parents of anorexic, quite appropriately,
wished their daughers to be significantly larger (see Table 1). Mothers of
anorexics were quite accurate in their assessment of their daughters' sizes
but fathers' of anorexics underestimated their daughers' sizes relative to
controls, probably because of their concerns regarding the illness. Garfinkel
et al (in press) have also described a group of women with conversion
disorders and serious weight loss; these individuals had lost large amounts of
weight because of chronic vomiting. However, they did not display any of
the cardinal features of anorexia nervosa. Vomiting as not due to a drive for
thinness but rather usually expressed intense emotion, for example disgust
over conflicts relating to sexuality or other areas. In comparison with a
matched group of anorexics, the conversion group was found to lack body
image disturbance (Table 2) suggesting that such self-overestimates are not
merely due to weight loss or vomiting. Further evidence suggesting that
body size overestimation is not a. function of degree of weight loss comes
from a recent study of Garner et all (in press). They studied women who had
"bulimia" but who had never been emaciated. These "normal weight" bulimic
subjects had body size distortions which were comparable to a bulimic group
of anorexics. Other studies, using body part estimates of body image, have
found less specificity to the overestimation phenomenon; using other
techniques overestimation has been reported in many populations other than
anorexics (Hsu, 1982). There are several possibilities for these discrepant
findings:
I. the distorting photograph technique which measures general body size
estimates may measure something more fundamental to anorexic's
psychopathology than do techniques which assess body parts;
2. overestimation may relate to particular psychopathologic phenomena
which may be more common in anorexics, but are not unique to them. This
latter explanation will be discussed further, in terms of correlates of
marked overestimation of body size.
While body size overestimation may not be specific to anorexia nervosa and
while not all anorexics overestimate their sizes, anorexics as a group
overestimate their sizes relative to normal women on the distorting
photograph technique. Moreover studies using both this technique and those
estimates of body parts have found overestimation to be a strong predictor
of poor outcome (Garfinkel et al 1977; Casper et al 1979). The reliability of
this technique has been demonstrated in studies that assessed the same
patients at one week and one year intervals (Garfinkel et al 1978; Garfinkel
et al 1979). It has also been found to be stable in spite of such manipulations
as having subjects study their images in a mirror or ingesting meals that
connote high or low calorie foods (Garfinkel et al 1978).

732
 Table 1. Results on Body Size Estimation Tests in Parents and Children

Anorexics Controls Level of

(N =23) (N = 12) Statistical
mean+ S.D. mean+ S.D. Significanc e

Child
Visual Self-Percep tion
body image: (1% over or
under estimation) +8.1 + 8.1 +1.3 + 5.1 P-'0.004

Ideal Image of Self:

(%over or under-
estimation) -5.4 + 12.8 -7.5 + 6.9 N.S.

Mother:
Visual Self-Percep tion +4.9 + 7.2 +6.9 + 7.4 N.S.

Ideal Image of Self -2.4 + 8.7 -3.4 + 10.3 N.S.

Mother's estimate of
Child's Size +1.3 + 7.6 +6.2 + 6.9 N.S.

Mother's estimate of
Ideal Child's Size +9.7 + 8.4 +1.6 + 6.1 p<0.004

Father:
Visual Self-Percep tion +7.4 + 7.3 +6.5 + 5.6 N.S.

Ideal Image of Self +4.9 + 9.9 +1.3 + 9.7 N.S.

Father's estimate of
Child's Size -2.3 + 7.4 +5.5 + 4.9 p~ 0.003

Father's estimate of
Child's Ideal Size +15.2 + 5.1 +3.6 + 7.1 p~ 0.001

733
Table 2. Results of Self-Estimates of Body Size and Ideal Size in Anorexia
Nervosa and Conversion Disorder with Weight Loss

Conversion Anorexia Level of

Disorder Nervosa Statistical
(N = 20) (N = 20) Significance
mean+ S.D. mean+ S.D.

Visual Self-Perception
(body image):
%over or under-
estimation -3.2 + 8.3 +4.8 + 7.8 p<.0.05

Ideal Image of Self:

% over or under-
estimation +8.7 + 17.9 +4.8 + 11.7 p(0.05

734
"Marked" Overestimation of Body Size
bur group has studied 265 consecutive anorexic patients (using the criteria
of Garfinkel and Garner, 1982) with this technique and with an associated
battery of psychometric tests. We have found that anorexics as a group
overestimate their body size by over 4%, significantly more than for control
women of similar age who as a group are quite accurate (Figure 1).
Moreover, a large subgroup (40%) overestimate their body size by more than
10%, a degree of overstimation which is extremely uncommon in normal
women (Garner et al, 1976; Garfinkel et al, 1978). When we compared these
105 "marked overestimators" with the 160 anorexic patients who
overestimate only moderately or underestimate their body size, we found
that the "marked overestimators" weighed significantly less than the others
(73.4% vs 79.7% of average, p 0.001). Because this could have a significant
effect on other variables we selected subgroups of these of 88 subjects each
who were matched for weight. We have compared these two groups on a
number of clinical, historical and psychometric variables in order to learn
more about this group who markedly overestimate their sizes.
Table 3 displays data relating to the 2 groups on a variety of clinical
phenomena relating to weight and duration of illness. While the marked
overestimators tended to be individuals with bulimia, vomiting and laxative
misuse, these differences were not statistically significant. Marked
overestimators more commonly did not eat any breakfast (p 0.05), lunch
(p 0.01) and dinner (p 0.05). Important differences between the groups
related to depressive symptoms; the marked overestimators were described
as having labile moods (p 0.05), fragmented sleep (p 0.05) and early
morning awakening (p 0.01). They had alos mutilated themselves more
commonly than the comparison group (18% vs 7%, p 0.05).
Table 4 displays results comparing the 2 groups on psychometric measures of
attitudes toward eating, weight and their bodies. On each of the measures of
attitudes toward eating and weight and dieting (EAT and Restraint) the
marked overestimators displayed more extreme scores. Furthermores, on
the affective component of body image, body dissatisfaction, the
overestimators again revealed greater displeasure with their bodies; they
also reported significantly more anhedonia. Corresponding with these
findings, the marked overestimators reported wishing to be significantly
thinner on the distorting photograph technique, in spite of actually weighing
the same as the other anorexics.
Table 5 shows results of the marked overestimators on other psychometric
test. Their increased depression is evident on the Beck Depression Inventory,
as well as on the depression subscale of the Hopkins Symptom Checklist.
They also scored as more externally controlled, with lower self-esteem
(Janis-Field, 1959) and greater psychopathology in general. Therefore
marked overestimation is both a poor prognostic sign (Garfinkel et al, 1977)
and is also associated with psychopathology on a number of parameters.
A number of different theories have been offered as explanations of body
image disturbance in anorexia nervosa; these have been reviewed elsewhere
(Garner and Garfinkel, 1981). Of importance here, the data on the inter-
relationships between marked overestimation of size, body dissatisfaction,
anhedonia, depression and altered self-esteem, suggest that body perceptual

735
 Tabel 3. Comparison of "Marked Overestimators" and Other Anorexics
on Clinical and Historical Information

Marked Over- Underestimators Level of

estimators and Moderate Statistical
(N = 88) Overestimators Significance
(N = 88)
mean+ S.D. mean+ S.D.

Weight at testing (kg) 46.9 + 9.8 45.9 + 8.8 N.S.

% average weight 78.6 + 13.5 78.8 + 13.9 N.S.
Maximum weight (kg) 61.1 + 11.8 60.5 + 10.3 N.S.
% of average at maximum 102.7 + 17.6 103.1 + 15.2 N.S.
Minimum weight (kg) 40.4 + 6.9 41.1 + 6.9 N.S.
% of average at minimum 67.9 + 9.6 70.1 + 9.9 N.S.
Age (years) 22.1 + 5.6 21.7 + 5.1 N.S.
Duration of illness
(month) 49.3 + 54.9 41.2 + 41.1 N.S.
Social class 2.8 + 1.7 2.6 + 1.5 N.S.

736
 Table 4. Comparison of Marked Overestimators and Other Anorexics
on Attitudes to Eating, Weight and their Bodies

Marked Over- Underestimators Level of

estimators and Moderate Statistical
Overestimators Significance
mean+ S.D. mean+ S.D.

Eating Attitudes Test

(EAT) 45.3 + 14.7 33.3 + 15.6 p~ 0.001
(Garner and Garfinkel, 1979)

Restraint 25.2 + 5.1 22.4 + 6.2 p£0.002

(Herman and Polivy, 1975)

Body Satisfaction 80.4 + 18.3 69.6 + 14.6 p(0.001

(Berscheid et al, 1973)

Anhedonia 17.0 + 9.1 13.1 + 7.0 p(0.004

(Chapman et al, 1976)

Ideal Body Size Estimate -6.1 + 12.6 -0.1 + 13.5 p{0.003

(Garner et al, 1976)

737
 Table 5. Comparison of Marked Overestimators and Other Anorexics
on other Psychometric Tests

Marked Over- Underestimators Level of

estimators and Moderate Statistical
Overestimators Significance
mean+ S.D. mean+ S.D.

Beck Depression Inventory 30.4- + 13.9 21.0 + 12.5 p(0.001

(Beck, 1978)

Hopkins Symptom Checklist 136.9 + 32.0 116.8 + 29.4- p~ 0.001

(Total)
(Derogatis et al, 1974-)
Somatization 23.5 + 7.0 20.0 + 6.1 p< 0.001
Anxiety 17.0 + 5.3 14-.1 + 6.1 P< 0.001
Depression 29.1 +7.2 25.0 + 7.6 P< 0.001
Obsessive-Compulsive 18.9 + 6.1 15.8 + 5.4- p(0.002
Interpersonal
Sensitivity 17.4- + 4-.3 15.5 + 4-.5 p( 0.01

Locus of Control 18.6 + 6.6 16.0 + 5.8 p~0.01

(Reid and Ware, 1973)

Test for Self-Esteem 72.6- 14-.5 66.2 + 12.4- p<.0.05

(Janis and Field, 19 59)

738
abnormalities are closely related to other dimensions of negative self-
concept. This explanation may help explain why overestimation of size is
common in, but not unique to, anorexia nervosa. This is in agreement with
self-esteem consistency theory (Wylie, 1968) which holds that expectations
and perceptions will be determined by one's evaluation of general, non-
physical attributes. Further evidence for this comes from work comparing
women with anorexia nervosa with very weight preoccupied non-anorexic
women (Garner et al, in press). This latter group were selected because they
had a desire for thinness equal to the anorexics but they never developed the
clinical syndrome of anorexia nervosa. These weight preoccupied women
displayed dissatisfaction with their bodis equal to that of the anorexics and
far greater than in normal women. However, they lacked other psychological
parameters of anorexics. Therefore a strong drive for thinness may be
associated with a general dissatisfaction with one's body and negative sense
of self and this occurs in both anorexic and very weight preoccupied non-
anorexic women. Further studies on normal weight. women with "bulimia"
(Garner et al, in press) also support a link between body image disturbance
and these measures of psychopathology.
It has been asserted that in anorexia nervosa, self-worth becomes
concretized onto body shape (Bruch, 1973). Self-worth is largely determined
by external phenomena. Anorexics become highly sensitized to shape and
body fatness barnes an index by which non-physical qualities are evaluated.
If an individual views all aspects of herself negatively, as anorexics with a
dichotomous ("all or nothing") thinking style are prone to do (Garner and
Bemis, 1982), and if she also equates low self-esteem with fatness, she may
see herself as larger than her actual size. Low self-esteem and body
dissatisfaction may play a role in misperception of size in anorexia nervosa,
particularly when coupled with heightened shape concerns.

Distortions in Internal P~_!"~tions

According to Bruch (1962, 1973) the concept of "body awareness" or "body
identity" is not limited to body image but extends to the perception and
interpretation of interoceptive stimuli. More specifically, she postulates
that anorexia nervosa is fundamentally related to disturbed awareness of
inner processes which include misperception of hunger, satiety, and other
bodily sensations. She considers the lack of responsiveness to fatigue, cold,
and sexual feelings in anorexia nervosa to be examples of this disturbance.
Patients will often describe extreme confusion about their internal states
early in treatment or will appear devoid of thoughts and feelings which
reflect their personal experiences. These deficits range in depth from subtle
confusion in affective labeling to complete mistrust of one's internal state.
While variable in severity, this disturbance is very common in anorexia
nervosa (for further clinical descriptions of this see Bruch (1973) and
Garfinkel and Garner (1982).
Consistent with the clinical observations of interoceptive disturbances,
several lines of experimental inquiry suggest that these patients may not
experience their internal environment in the same way as individuals
without eating disorders.
In the early part of this century, Cannon (1912) claimed that gastric

739
 contractions, measured by an intragastric balloon, could be related to
hunger sensations in normal individuals. The relationship between gastric
activity and "hunger" has not been supported by more recent investigations
(Penick et al, 1967; Stunkard and Fox, 1971). However, the perception of
gastric contractions in anorexia nervosa has been examined by Silverstone
and Russell (1967) using an intragastic tube, and by Crisp (1967) using an
intragastric pressure telemetry pill. Both studies found no significant
differences in stomach motility between anorexic patients and normal
subjects. The anorexic patients were capable of recognizing contractions,
but interestingly, some did not interpret these as sensations of hunger.
Coddington and Bruch (1970) found that both anorexic and obese subjects
were le~s accurate than normals in perceiving the amounts of food
(Metrecal) that were directly indroduced into their stomachs. While only
three anorexic subjects were included in this study, it does support the
hypothesized deficit in the recognition of internal state. Other studies have
reported that subjects with juvenile onset obesity have difficulties
appropriately responding to internal satiety cues (Cabanac and Duclaux,
1970; Campbell et al, 1971) or that their eating behaviour may be largely
determined by external circumstances such as the availability, salience, and
palatability of food (Nisbett, 1972; Schachter, 1971). However, this external
responsiveness may be a phenomenon of dieting, rather than obesity
(Herman and Polivy, 1975) and therefore probably applies to anorexics as
well as the obese. The perception of hunger and satiety in anorexia nervosa
was investigated by Garfinkel (1974) using several self-report
questionnaires. Eleven anorexic patients reported distorted sensations of
satiety compared with 11 control subjects. After fasting for 12 hours, all
subjects were given a questionnaire inquiring about the experience of
hunger. Subjects were then given a standard meal followed by another
questionnaire assessing satiety. Except for an increased preoccupation with
food and a fear of eating, anorexics did not differ from normals in their
perception of hunger. However, in contrast to normals, the anorexic patients
reported disturbed sensations of satiety including bloating, absence of
stomach sensations, nausea, aches and pains. In a recent study, Dubois et al,
(1979) have shown that anorexic patients display delayed gastric emptying
rates for the anorexic patients tended to increase toward normal but were
still significantly less than in controls. These results may partially explain
the postprandial fullness, discomfort, and early satiety observed with most
patients.
In a recent series of studies Garfinkel et al (1978, 1978) have used taste
perception as an index of satiety in anorexia nervosa. This procedure was
derived from the work of Cabanac and Duclaux {1970), who found that obese
subjects do not experience any differences in the rated pleasantness of
sucrose tastes before versus after the ingestion of glucose. Normal subjects
in contrast, experience " satiety" or an aversion to the taste of sucrose after
glucose preloading. On the basis of these results Cabanac and Duclaux (1970)
suggested that obese subjects are less responsive than nonobese to internal
cues related to nutritional requirements. A replication of the Cabanac and
Duclaux steady (Underwood et al, 1973) reported an interesting but
unanticipated finding. One of the normal subjects who displayed an absence

740
of aversion to sucrose was later found to have anorexia nervosa.
Our first study of this demonstrated that disturbances in interoception
measured by the satiety-aversion-to-sucrose test were evident in anorexic
patients in contrast to normal controls. A modified version of the Cabanac
and Duclaux (1970) procedure was used - only a 20% sucrose solution was
employed since this produced the maximal differences between obese and
controls in the Cabanac and Duclaux (1970) studies. Rather than using a 50g
glucose load, we used two test lunches which contained 400 Cal. One of
these was designed to connote a high calorie content; it consisted of tuna
salad, cole slaw, and a large chocolate sundae. The second lunch was
identical in actual caloric content but was designed to connote a relatively
low calorie content, tuna salad and cole slaw both supplemented with
gluconal. Sweetness and pleasantness ratings were collected on all subjects
every three minutes for one hour after lunch. The anorexics, unlike the
controls, failed to develop an aversion to the sweet taste after the test
lunches. Follow-up of many of these patients one year later revealed similar
results.
Furthermore, in a large sample of our anorexic patients this interoceptive
disturbance was related to the tendency to overestimate body size (N=72,
r=0.026, p<0.02); it is the overestimators who fall to develop a normal
aversion. In contrast to Crisp and Kalucy's (1974) earlier report, we did not
find that the caloric connotation of the meal systematically influenced body
size estimation in either patients or controls. The lack of satiety aversion
was also independent of which meals was eaten. As in an earlier study
(Garfinkel, 1974) the results on a satiety questionnaire differentiated
anorexics from normal subjects on various parameters. These were
postprandial changes in mood, gastrointestinal sensations and willpower
required to stop eating. Using a 10 em analogue scale for "satiety",
anorexics displayed greater pre-meal fullness and their fullness persisted
longer after eating the meal that they believed to have more calories. This
latter finding suggests a substantial cognitive influence in the satiety
experience in anorexic patients.
While there is evidence for a cognitive component to the altered satiety
experience, recent work also suggests that the changes in gastrointestinal
physiology that accompany starvation may also play a role. Starving
anorexics have delayed gastric emptying (Dubois et al, 1979; Saleh and
Lebwohl, 1979; Holt et al, 1981) which likely contributes to many symptoms
of satiety following a meal. Dubois et al (1979) showed that bethanacol, a
cholinergic agonist, partially improved gastric emptying while Saleh and
Lebwohl (1980) demonstrated that metoclopromide returned gastric
emptying to normal. However the use of metoclopromide is limited by
neurological side effects and its propensity to produce depression. Recently
Russell et al (in press) have described an anorexic patient with delayed
gastric emptying and subjective post-prandial bloating. Domperidone, a
novel compound which enhances gastric peristalsis, accelerates gastric
emptying time and does not cross the blood brain barrier, improved both
gastric emptying and also the patient's satiety ratings. A controlled trial of
this medication is now in progress.
In summary, several lines of experimental inquiry have suggested that

741
patients with anorexia nervosa may misperceive internal experiences. It is
not clear whether these perceptual disturbances are determinants or
byproducts of the syndrome, and what their relationship is to complete
recovery. Since these aberrant experiences are clinically important with
these patients, more detailed study of potential m~chanisms will be of
considerable value.

742
References

1. Beck A. T. (1978) Depression Inventory. Center for Cognitive Therapy,

Philadelphia

2. Berscheid E., Walster E., Hohrnstedt G. (1973) The happy American

body: A survey report. Psychology Today November 119-31

3. Bruch H. (1962) Perceptual and conceptual disturbances in anorexia

nervosa Psycho! Med 24:187-194

4. Bruch H. (1973) Eating Disorders: Obesity, Anorexia Nervosa and the

Person Within. Basic Books New York

5. Cabanac M., Duclaux R. (1970) Obesity: Absence of satiety aversion to

sucrose. Science 168:496-497

6. Campbell R.G., Hashim S.A., Vanltallie T.B. (1971) Studies of food

intakte regulation in man: Responses to variations in nutritive density
in lean and obese subjects. N Engl J Med 285:1402-1407

7. Cannon W.B., Washburn A.L. (1912) An explanation of hunger. Am J

Physiol 29:441-454

8. Casper R.C., Halmi K.A., Goldberg S.C., Eckert E.D., Davis J.M.
(1979) Disturbances in body image estimation as related to other
characteristics and outcome in anorexia nervosa. Br J Psychiatry
134:60-66

9. Chapman L.F., Chapman J.P., Raulin M.L. (1976) Scales for physical
and social anhedonia. J Abnorm Psycho! 85:374-382

10. Coddington R.D., Bruch H. (1970) Gastric perceptivity in normal,

obese and schizophrenic subjects. Psychosomatics 11:571-579

11. Crisp A.H. (1967) The possible significance of some behavioural

correlates of weight and carbohydrate intake. J Psychosom Res
11:117-131

12. Crisp A.H., Kalucy R.S. (1974) Aspects of the perceptual disorder iri
anorexia nervosa. Br J Med Psycho! 47:349-361

13. Derogatis L., Lipman R., Rickels K., Uhlenhuth E.H., Covi L. (1974)
The Hopkins Symptom Checklist (HSCL): A self report symptom
inventory. Behav Sci 19:1-15

14. Dubois A., Gross H.A., Ebert M.H., Castell D.O. (1979) Altered gastric

743
 emptying and secretion in primary anorexia nervosa. Gastroenterology
77:319-323

15. Fisher S., Cleveland S.E. (1958) Body Image and Personality. Dover
Publications New York

16. Garfinkel P .E., Garner D.M. (1982) Anorexia Nervosa:

Multidimensional Perspective. Brunner Mazel New York

17. Garfinkel P.E. (1974) Perception of hunger and satiety in anorexia

nervosa. Psycho! Med 4:309-315

18. Garfinkel P .E., Moldofsky H., Garner D.M. (1977) Prognosis in anorexia
nervosa as influenced by clinical features, treatment and
selfperception. Can Med Assoc J 117:1041-1045

19. Garfinkel P .E., Moldofsky H., Garner D.M., Stancer H. C., Coscina D. V.
(1978) Body awareness in anorexia: Disturbances in body image and
satiety. Psychosom Med 40:487-498

20. Garfinkel P.E., Moldofsky H., Garner D.M. (1979) The stability of
perceptual disturbances in anorexia nervosa. Psycho! Med 9:703-708

21. Garfinkel P.E., Kaplan A.S., Garner D.M., Darby P.L. (in press) The
differentiation of vomiting and weight loss as a conversion disorder
from anorexia nervosa. Am J Psychiatry

22. Garfinkel P.E., Garner D.M., Rose J., Darby P.L., Brandes J.S.,
O'Hanlon J., Walsh N. (in press) A comparison of characteristics in the
families of patients with anorexia nervosa and normal controls.
Psycho! Med

23. Garner D.M., Garfinkel P.E. (1981) Body image in anorexia nervosa:
Measurement, theory and clinical implications. Inter J Psychiat Med
11:263-284

24. Garner D.M., .Garfinkel P.E., Stancer H.C., Moldofsky H. (1976) Body
Image disturbances in anorexia nervosa and abesity. Psychom Med
38:227-236

25. Garner D.M., Olmsted M., Garfinkel P.E. (in press) A comparison
between weight preoccupied wumen and anorexia nervosa. Int J Eating
Disorders

26. Garner D.M., Garfinkel P.E., O'Shaughnessy M. On press) A comparison

between bulimia with and without anorexia nervosa. New Directions in
Anorexia Nervosa. Report of the Fourth Ross Conference on Medical
Research

744
27. Garner D.M., Garfinkel P.E. (1979) The Eating Attitudes Test: An
index of the symptoms of anorexia nervosa. Psycho! Med. 9:1-7

28. Garner D.M., Bemis K. (1982) A cognitive-behavioral approach to

anorexia nervosa. Cognitive Therapy and Research 6:1-27

29. Glucksman M.L., Hirsch J. (1969) The response of obese patients to

weight reduction. Psychosom Med 31:1-7

30. Gull W.W. (1868) The address in medicine delivered before the annual
meeting of the BMA at Oxford Lancet 2:171

31. Head H. (1920) Studies in Neurology. Hodder Stoughton, London

32. Herman C.P., Polivy J. (1975) Anxiety, restraint and eating behaviour.
J Personal 84-:666-672

33. Holt S., Form M.J., Grant S. (1981) Abnormal gastric emptying in
primary anorexia nervosa. Br J Psychiatry 139:550-552

34. Hsu L.K. (1982) Is there a disturbance in body image in anorexia

nervosa. J Nerv Ment Dis 170:305-307

35. Janis I.L., Field P.B. (1959) Sex differences and personality factors
related to persuasibility. In: Houland C.I., Janis I.L. (Eds) Personality
and Persuasibility. Yale University Press, New Haven, p 55-68

36. Kolb C. (1975) Disturbances of body image. In: Arieti (Ed) American
Handbook of Psychiatry. Basic Books, New York, p 810-831

37. Lasegue (1873) De l'anorexie hysterique. Reprinted in Kaufman R.M.,

Heiman M. (Eds) Evolution of Psychosomatic concepts. Anorexia
Nervosa: A. Paradigm (1964). Internation Universities Press, New York

38. Morton R. (1694) Phthisiologica: Or a Treatise of Consuptions. Sam

Smith and Benjamin Wolford, London

39. Nisbett R.E. (1972) Eating behaviour and obesity in men and animals.
Adv Psychosom Med 7:173-193

40. Penick S.B., Smith G.P., Wieneke K. Jr., Hinkle L.E. Jr. (1967) An
experimental evaluation of the relationship between hunger and
gastric motility. Am J Physiol 205:421-426

41. Reid D. W., Ware E. E. (1974) Multidimensionality of internal versus

external control: addition of a third dimension and nondistinction of
self versus others. Can J Behav Sci 6:131-142

42. Russell D. MeR, Freedman M.L., Feiglin D.H.I., Jeejeebhoy K.N.,

745
 Swinson R.P., Garfinkel P.E. (in press) Delayed gastric emptying in
anorexia nervosa- improvement with domperidone. Am J Psychiatry

43. Saleh J. W., Lebwohl P. (1980) Metoclopramide-induced gastric

emptying in patients with anorexia nervosa. Am J Gastroenterol
74:127-132

44. Schacter S. (1971) Emotions, Obesity and Crime. Academic Press, New
York

45. Schilder P. (1935) Image and Appearance of the Human Body. Kegan,
Paul, Trench, Trubner and Company, London

46. Secord P.F., Jourard S.M. (1953) The appraisal of body cathexis: Body
cathexis and the self. Consult Psycho! 17:343-347

47. Silverstone J. T., Russell G.F.M. (1967) Gastric hunger in contractions

in anorexia nervosa. Br J Psychiatry 13:257-263

48. Stunkard A.J., Fox S. (1971) The relationship of gastric motility and
hunger: A summary of the evidence. Psychosom Med 33:123-134

49. Traub A.C., Orbach J. (1964) Psychological studies of body image: An

adjustable body distorting mirror. Arch Gen Psychiatry 11: 53-66

50. Underwood P.J., Belton E., Hulme P. (1973) Aversion to sucrose in

obesity. Proc Nutr Soc 32:94

51. Wylie R. (1968) The present status of self-theory. In: Borgatta E.G.,
Lambart W.W. (Eds). Handbook of Personality Theory and Research.
Rand McNaJly, Chicago

746
NALOXONE IN THE TREATMENT OF ANOREXIA NERVOSA: METABOLIC AND

OTHER EFFECTS

Ivor H. Mills and Linda Medlicott

Department of Medicine, University of Cambridge

Addenbrooke's Hospital, Cambridge CB2 2QQ
England

Anorexia nervosa is a compulsive behaviour disease and is one

of a number of such compulsive behaviour diseases. The reasons
which brought us to this conclusion are as follows:

(1) A number of young ladies are somewhat overweight prior to

the development of anorexia nervosa. They often start dieting in a
gentle way and succeed in losing weight slowly, with no effect on
their menstrual periods. After a period of gradual weight loss they
find that, as they say, "suddenly the diet took over". By this they
mean that something inside them drives them to more severe weight
loss with severe dieting. Usually from the moment of this change
they stop menstruating. During this time they find they have very
little ability to control their eating by their own intelligence.

(2) They find that thoughts of food, weight dieting, size and
size of clothes are more and more continuing in their mind. They
may think of food they would like to eat but they have lost their
normal appetite control of food intake. They frequently prepare
food for others but cannot bring themselves to eat it. Their only
escape from this everlasting thinking of food and related things is
by intense concentration on other things, frequently some type of
work. They may become exhausted by this, only to find that the
"record-player is stuck in the same groove" and plays even more
intensely.

(3) Sometimes a group of girls or young ladies decide that

they will all go on a diet together. After about a week or so, most
find that they cannot stand the hunger and stop dieting. In each
group there will be one or two who will not stop the food restriction.
They are the potential anorexics. They are self-selected for
747
qualities of perfectionism and determination and refuse to give in
just because they are hungry.

The motivation for such groups is the widely accepted slim

image for females which came in in the early nineteen sixties and
spread throughout the developed countries.

Not all perfectionist girls will succumb to the immediate

effect of dieting. They have to be in a high arousal state brought
about by intense dedication to some demanding task. Our research
some ten years ago (Mills et al, 1973) showed that in 75% of anorexic
young women they started their severe dieting when working for an
important examination, even though it might be many months away.
They tend to be pressurised by their own high ideals of achievement,
by parents, by teachers and by their peer group. As they increase
their own arousal they become aware of increased efficiency as
described by the Yerkes-Dodson inverted U-shaped curve. It is in
this state that they are vulnerable to the effect of dieting.

Starvation was found by Benedict in 1915 to cause gradual

increase in mental arousal which increased brain efficiency in
solving problems more accurately and in a shorter time. If the girl
is persuaded she has lost too much weight and re-feeds herself, her
arousal falls and she is less capable of work: she then returns
with new determination to severe dieting.

Of the 25% not working for examinations, most had social

problems that they were trying to solve. More and more we find
among such problems the break up of the family, with the girl
fighting to hold the two parents together. This is an unavailing,
open-ended commitment of the type they will never give up, with the
result that arousal and determination are brought to bear in the
same way as in the ones working for examinations.

(4) It is impossible to reason with the anorexic young woman.

The compulsive mechanism in her brain feeds into her consciousness
not only the thoughts of an anorexic kind but even complex forms of
behaviour to defeat the well-meaning parents and medical staff who
try to help her. They are not amenable to reason, especially when
they have in their brains an enlarged body image which they endeavour
to shrink by food restriction. They can only be treated by some
form of behavioural therapy employing rewards and punishments.

(5) About 20 to 30% of the anorexic girls swing over to

phases of compulsive eating. In its extreme form they eat regardless
of hunger and especially of carbohydrate foods that they normally
would reject. Sometimes vast quantities are consumed and may auto-
matically precipitate vomiting or the girl may be able to train
herself to vomit after such bingeing bouts. The vomiting itself may
become compulsive and occur after their normal small meals.

748
Compulsion Scoring

Having decided that anorexic behaviour is driven by a compulsive

mechanism in the brain we devised a coping questionnaire which was
scored and validated (Medlicott and Mills, 1983). From the 76
questions we selected those bearing on compulsive behaviour and
calculated the compulsion score. The values for pure anorexics,
anorexics who also have compulsive eating, people with other forms
of compulsive behaviour, a group of pure depressives and a control
group matched for age, sex, social class and responses to the
Eyesenck Personality Questionnaire, are shown in Table 1. Also in
this table are their responses to questions bearing on alcohol
dependence and a depression score.

Only the bingeing anorexics have a high alcohol score but all
three groups with some form of compulsive behaviour come out with
high compulsion scores. All four groups have significantly higher
depression scores than the control group.

Compulsive Mechanisms in the Brain

Having established the evidence of a compulsive mechanism in

the brain we sought to influence potential neuronal pathways with
drugs.

Table 1. Compulsion, Alcohol and Depression Scores (Mean± S.E.)

(from Medlicott and Mills, 1983)

Compulsion Alcohol Depression

Group n
Score Score Score

Controls 42 13.1 ± 1.1 0.7 ± 0.3 7.9 ± 0.9

Pure
Anorexics 34 32.1 ± 1.9** 0.6 ± 0.3 13.8 ± 0. 8'~'~

Bingeing
Anorexics 28 35.8 ± 1.9** 5.8 ± 1.2** 17 • 0 ± 0. 6*'~

Other
Compulsives 28 28.0 ± 2.2** 1.6 ± 0.6 11.1 ± 1.1'~

Depressed 30 14.8 ± 1.4 17.8 ± 0.7*':'

* p < 0.01 ** p < 0.005

749
 I
700
600 600
500 0 500
I
I
400 _J Q 400
..... 6 Q 'I

0 "(jl
300 _J
..... I
300
E I
I
200 :i.. I 200
100 <1
LL 100
0 w
z 0

8 10 12 14
A Weeks B Weeks

• 52~
I 5o~
~ 48~
1000
900
~ 46t
(l)

3
44
; •
700 I
_J
600 .....
42~r;.r
r.
ro
0 90 V"' 500 0E
6 80 :i..
:::J 70 9: Q p 300 LL <1

~ 60 r- Q),:(; \ : 200 w
z
:i.. 50 r- 0 : ¢ 100
~ 40~ ~ ?:P 0
6 30t' OJ.
6
'

(!)_ 2,0 fuj'

I
~--~~----~~----CUL-~- __j
0 4 8 12 16 20
c Weeks
Fig. 1 Changes in weight and serum levels of S-OHB and NEFA in rela-
tion to period of naloxone (and saline) infusions: A) in
patient 1; B) in patient 2; C) in patient 3 (from R. Moore,
I. H. Mills and A. Forster (1981) J. Roy. Soc. Med. 74., 129-
131).

750
 Antidepressants which block reuptake of noradrenaline, such as
nortriptyline and amitriptyline, help those girls who accept that
they must use their own intelligence to fight the compulsive drive.
Otherwise they become overtly depressed and then fail to maintain
the drive to eat. Noradrenaline potentiation does not appear to
turn off the compulsion to starve. Phenoxybenzamine also appeared
to give no help.

Blocking dopamine receptors with chlorpromazine is a well

accepted treatment but it fails to help the most depressed patients
and so does not break the compulsive mechanism. Stimulation of
dopamine receptors with bromocriptine was reported to help young
women with elevated prolactin levels to combat compulsive eating.
(High prolactin has been found to intensify stereotyped behaviour
in animals, Drago et al, 1981.) In those with normal prolactin
levels we found no benefit from bromocriptine.

The antidepressants used have potent anticholinergic actions

which might have produced some of the benefit of these drugs.
Indeed biperiden, an anti-cholinergic drug with no antidepressant
action, was reported to facilitate sleep in some of the patients
but it did not stop the anorexic thinking.

Potentiating 5-hydroxytryptamine with clomipramine in doses

up to 300 mgs per day had no obvious beneficial effect in anorexic
patients.

Possible Opiate Stimulation

We decided, therefore, that the compulsive mechanism in the

brain might be driven by an opiate mechanism and began treatment
with continuous infusion intravenously of the opiate antagonist
naloxone. It was given continuously by a mechanical pump in doses
between 1.6 m~s and 6.4 mgs per day for times up to 61 days. In
figure 1 are three patients' data in whom we measured plasma levels
of non-esterified fatty acid (NEFA) and S-hydroxybutyrate several
times a week in the fasting state in patients who had been persuaded
for some time to consume 3000 k calories per day. In each there
was a progressive and rapid fall in the NEFA and S-hydroxybutyrate
which rose again on cessation of naloxone. The fall did not occur
while saline was infused in the third patient (Moore et al, 1981).

The mean(± S.E.) of the weight gain prior to, during and after
naloxone infusion in 12 patients is shown in figure 2. It is clear
that naloxone had a very marked effect on weight gain. The effect
on NEFA and S-hydroxybutyrate suggest that it operates by inhibiting
lipolysis driven by an opiate mechanism.

With short term infusions it did not change the girl's way of
thinking but with long term infusions of three months or more many

751
 T
2 f-
I T
0 vs b p::: <0.001
d vs e P=<O.OI
c vs f p =<0.01

f-

l ~
T
0 r 1 I I
0 b c d e
week I week During I week I week 4 weeks
before N x ofterNx Nx infusion b€fore Nx after Nx after Nx.
started star ted .(mean ended ended ended
4.9 wks)

Fig. 2 Weight gain before, during and after naloxone infusion in 12

patients with anorexia nervosa (mean± S.E.) (from R, Moore,
I.H. Mills and A. Forster (1981) J. Roy. Soc. Med. 2±• 129-
131).

752
young women have been cured in the sense that all anorexic thinking
appears to have stopped despite, in some cases, up to ten years of
severe anorexia.

Endocrine Effects

Although profound effects of short-term treatment with very

large doses of naloxone on endocrine function in animals and man
have been reported, the effects of these relatively small doses in
the early stages were not marked. Prolactin and FSH values showed
the same scatter before and during naloxone infusion. However,
luteinising hormone values showed an average rise, though the
scatter was still wide.

Responses of LH and FSH to luteinising hormone releasing hormone

were however changed in four patients from the so-called hypothalamic
response with the highest value at 60 minutes, to a greater and normal
response during naloxone infusion (Table 2)

SUMMARY

Evidence suggests that anorexia nervosa is a compulsive

behaviour disease. We constructed a questionnaire from which it was
shown that pure anorexics had a mean(± S.E.) compulsion score of
32.1 ± 1.9, bingeing anorexics had a score of 35.8 ± 1.9 but
depressed patients' mean score was 14.8 ± 1.4 compared to a matched
control group's score of 13.1 ± 1.1. Each group had between 28 and
34 subjects but the control group had 42.

Drugs potentiating or inhibiting nerve transmission by

noradrenaline, acetyl choline, dopamine and 5-hydroxytriptamine had
no consistent effect on the compulsive anorexic behaviour.

Table 2. Responses to LHRH 100 ~gs I.V. before and during

naloxone in four patients (mean values)

Time
Naloxone
Status
0 20 min 60 min

LH Before 2.3 9.6 11.3

U/L During 2.4 28.3 18.8

FSH Before 2.9 7.6 14.1

U/1 During 3.5 13.5 11.0

753
 Treatment with the opiate antagonist naloxone in doses up to
6.4 mg per day, by continuous intravenous infusion for up to 61 days
had a marked effect on weight gain. In three subjects the naloxone
infusion was associated with a sharp fall in plasma non-esterified
fatty acid and S-hydroxy-butyrate, suggesting that the effect on
weight gain is due to inhibition of a lipolytic mechanism.

With prolonged infusion beyond three months the anorexic

thinking was apparently changed even in some chronic anorexics.

REFERENCES

Drago, F., Van Ree, J. M., Bohus, B., and De Wied, D., 1981,
Endogenous hyperprolactinaemia enhances amphetamine- and
apomorphine-induced stereotypy, Eur. J. Pharm., 72:249.
Medlicott, L., and Mills, I. H., 1983, In preparation.
Mills, I. H., Wilson, R. J., Eden, M. A.M., and Lines, J. G., 1973.
Endocrine and social factors in self-starvation amenorrho~d,
in: "Symposium- Anorexia Nervosa and Obesity,"
R. F. Robertson, ed., Royal College of Physicians of
Edinburgh, Edinburgh.
Moore, R., Mills, I. H., and Forster, A., 1981, Naloxone in the
treatment of anorexia nervosa: effect on weight gain and
lipolysis, J. Roy. Soc. Med., 74:129.

754
BENZODIAZEPINE RECEPTOR-MEDIATED "ANXIETY" IN PRIMATES

P. Skolnick, P. Ninan, T. Insel, J. Crawley and S. paul

NIADDK and NIMH, National Institutes of Health and

Alcohol, Drug Abuse and Mental Health Administration
Bethesda, MD 20205

During the past seven years, numerous investigations have

demonstrated that the benzodiazepine receptor [in concert with an
associated recognition site for GABA and a chloride ionophore] is
involved in the antianxiety (anxiolytic) actions of a number of
chemically unrelated substances such as the benzodiazepines, barb-
iturates, triazolopyridazines (e.g. CL 218,872), pyrazolopyridines
(e.g. SQ 65,396), and ethanol (Skolnick and Paul, 1981). During the
past three years, several compounds have been described which bind
to benzodiazepine receptors with high affinities, and antagonize one
or more of the pharmacologic actions of the benzodiazepines (Skol-
nick et al., 1982). The first of these benzodiazepine antagonists
described was the ethyl ester of beta carboline-3-carboxylic acid
(Tenen and Hirsch, 1980; Oakley and Jones, 1980). Although this
compound was first postulated to be an endogenous ligand of the
benzodiazepine receptor (Braestrup, et al., 1980), it is likely that
this compound as well as the related beta carbolines extracted from
human urine were formed during the isolation procedure employed
(Squires, 1981). Nonetheless, these compounds have proven invalu-
able in both better defining the physiologic role(s) of the benzo-
~iazepine rec=ptor complex and providing a robust, chemically
lnduced model of anxiety in primates.

Beta carboline-3-carboxylic acid ethyl ester (SCCE) was first

demostrated to antagonize the anticonvulsant actions of diazepam and
to possess a proconvulsant action, but did not appear to evoke any
gross behavioral changes in rodents at doses of up to 100 mg/kg,
i.p. (Tenen and Hirsch, 1980; Oakley and Jones, 1980). The lack of
any "intrinsic" action of this compound may be in part attributed to
pharmacokinetic factors (the t 112 of SCCE in plasma (in vitro) is
less than one minute) (Schweri et al., 1983) and in part due to the

755
lack of an appropriate screening paradigm for intrinsic actions
(Corda et al., 1983). Despite the apparent lack of intrinsic
actions by SCCE, it was subsequently demonstrated that the closely
related methyl ester of beta carboline-3-carb oxylic acid (SCCM) was
a potent convulsant in mice (Braestrup et al., 1982; Schweri et al.,
1982). These convulsions were effectively blocked by the benzo-
diazepine antagonists Ro 15-1788 (an imidazobenzodiaz epine) and CGS
8216 (a pyrazoloquinolino ne) as well as by diazepam. These observa-
tions strongly suggested that benzodiazepine receptors are involved
in the convulsant actions of SCCM. Furthermore, these observations
also suggested that the term "benzodiazepine antagonist" did not
adequately describe the spectrum of pharmacologic effects produced
by beta carboline-3-carb oxylate esters, Ro 15-1788, and CGS 8216.
We have termed beta carboline-3-carb oxylate esters such as SCCE and
SCCM "active antagonists" (Ninan et al., 1982) [the term "inverse
agonist" has also been proposed] (Mohler and Richards, 1983).
Compounds such as Ro 15-1788 and CGS 8216 have been termed "anta-
gonists" since they do not appear to possess marked "intrinsic"
actions over a relatively wide dose range (Hunkeler et al., 1981;
Bernard et al., 198la,b). Nonetheless, several reports have appeared
(File et al., 1982; Mendelson et al., 1983) suggesting these com-
pounds do possess "intrinsic" actions. Thus, the intrinsic actions
of both CGS 8216 and Ro 15-1788 appear to be very dependent on both
the paradigm and dose used.

The observation that SCCM elicited convulsions in rodents

suggested that other actions, best described as pharmacologicall y
"opposite" to the benzodiazepines could also be observed using this
or a related compound. We chose to examine the actions of SCCE in
chair adapted rhesus monkeys, since the behavioral effects of a
putative anxiety producing compound might be most easily detected in
a primate, and the possible accompanying somatic and endocrine .
changes would most readily be detected in a larger laboratory an1mal.

In preliminary experiments (Ninan et al., 1982), intravenous

injection of SCCE (2.5 mg/kg) elicited dramatic behavioral changes
including agitation (struggling in restraining chair), increased
head and body turning, periods of immobility, failure to eat or
drink for extended periods, defecation and urination immediately
after injection, and distress vocalization. Accompanying these
behavioral changes were marked increases in blood pressure, heart
rate, plasma cortisol, epinephrine, and norepinephrine (Ninan et
al., 1982). The behavioral, somatic and endocrine changes elicited
by SCCE could be blunted or completely antagonized by pretreatment
of animals with Ro 15-1788, suggesting that these effects were medi-
ated by benzodiazepine receptors. More recent studies (Insel et
al., 1984) have shown that a reproducible behavioral response can be
evoked in animals with 100 ~g/kg of SCCE (i.v.), 1/25 of the dose
used in our initial studies (Ninan et al., 1982)! Significant

756
increases in plasma ACTH, cortisol and beta endorphin have also been
observed with these doses of SCCE (Insel et al., 1984; and Ninan et
al., in preparation).

Although administration of "active antagonists" of the benzo-

diazepine receptor appears to evoke a syndrome reminiscent of fear
or anxiety [i.e. possessing a behavioral, somatic and endocrine
component] (Skolnick et al., 1983), any animal model attempting to
simulate a psychological or psychiatric state is incomplete since an
affective component is lacking. Although the behavior of the
monkeys in these studies ~pears to be fearful or anxious, and can
be blocked by clinically effective anxiolytic agents (Insel et al.,
1984; Skolnick et al., 1983), it should not be considered a valid
model until examined in humans. Recently, Dorow et al. (1983) have
administered FG 7142 (a beta carboline closely related to SCCE) to
normal volunteers. In two (of twelve) volunteers who had detectable
blood levels of this compound, a "motor unrest and a strong feeling
of inner tension described as a feeling of doom" was reported (Dorow
et al., 1983). These behavioral changes were accompanied by increases
in plasma cortisol, systolic blood pressure, and heart rate. In one
of the subjects, symptoms were immediately relieved by administration
of a benzodiazepine. Thus, there are striking similarities between
administration of SCCE to rhesus monkeys and a syndrome which appears
to be strongly reminiscent of generalized anxiety in humans. Further
investigations will be necessary to confirm the preliminary report
of an anxiety-like syndrome in humans. However, observations in
primates strongly suggest the benzodiazepine receptor may play a
physiologic role in mediating anxiety as well as a pharmacologic
role in the anxiolytic actions of a variety of compounds.

References

Bernard, P., Bergen, K., Sobiski, R., and Robson, R.D., 198la, CGS
8216 (2-phenylpyrazolo14,3-c]quinolin-3(5H)-one), an orally
effective benzodiazepine antagonist, Pharmacologist 23:150.
Bernard, P., Wilson, D.E., Sobiski, R., and Robson, R.D., 198lb,
Antagonism of diazepam-induced sedation and alcohol potenti-
ation by CGS 8216 (2-phenylpyrazolo[4,3-c]quinolin-3(5H)-one)
in the rat, Soc. Neurosci. Abstr. 7:862.
Braestrup, D., Nielsen, M., and Olsen, C.E., 1980, Urinary and brain
s:carboline-3-carboxylates as potent inhibitors of brain benzo-
diazepine receptors, Proc. Natl. Acad. Sci. USA 77:2288.
Braestrup, C., Schmiechen, R., Neef, G., Nielsen, M., and Petersen,
E., 1982, Interaction of convulsive ligands with benzodiazepine
receptors, Science 216:1241.
Corda, M., Blaker, W., Mendelson, W., and Guidotti, A., 1983, S-
Carbolines enhance the shock-induced suppression of drinking in
the rat, Proc. Natl. Acad. Sci. USA 80:2072.

757
 Dorow, R., Horowski, R., Paschelke, G., Arnin, M., and Braestrup, C.,
1983, Severe anxiety induced by FG 7142, a S-carboline ligand
for benzodiazepine receptors, Lancet 9:98.
File, S., Lister, R., and Nutt, D., 1982, lntrinsic actions of
benzodiazepine antagonists, Neurosci. Lett. 32:165.
Hunkeler, W., Mohler, H., Pieri, L., Pole, P., Bonetti, E., Cumin,
R., Schaffner, R., and Haefely, W., 1981, Selective antagonists
of benzodiazepines, Nature 290:514.
Insel, T., Ninan, P., Aloi, J., Jimerson, D., Skolnick, P., and
Paul, S., 1984, A new psychopharmacolo gic model of anxiety,
Arch. Gen. Psych., in press.
Mohler, H., and Richards, J., 1983, Receptors for anxiolytic drugs,
in: "Anxiolytics: Neurochemical, Behavioral and Clinical
Perspectives," J. Malick, S. Enna, and H. Yamamura, eds., Raven
Press, New York.
Ninan, P., Insel, T., Cohen, R., Cook, J., Skolnick, P., and Paul,
S., 1982, Benzodiazepine receptor-mediated experimental "anxiety"
in primates, Science 218:1332.
Oakley, N., and Jones, B., 1980, The proconvulsant and diazepam-
reversing effects of ethyl-S-carbolin e-3-carboxylate, Eur. J.
Pharmacol. 68:381.
Schweri, M., Cain, M., Cook, J., Paul, S., and Skolnick, P., 1982,
Blockade of 3-carbomethoxy-S -carboline induced seizures by
diazepam and the benzodiazepine antagonists, Ro 15-1788 and CGS
8216, Pharmacol. Biochem. Behav. 17:457.
Schweri, M., Martin, J., Mendelson, W., Barrett, J., Paul, S., and
Skolnick, P., 1983, Pharmacokinetic and pharmacodynamic factors
contributing to the convulsant action of S-carboline-3-ca rboxyl-
ate esters, Life Sci. 33:1505.
Skolnick, P., and Paul, S., 1981, Benzeodiazepine receptors, in:
"Annual Reports in Medicinal Chemistry, Vol. 16, H.J. Hess, ed.,
Academic Press, New York.
Skolnick, P., Hommer, D., and Paul, S., 1982, Benzodiazepine antago-
nists, in: "Pharmacology of Benzodiazepines, " E. Usdin, P. Skol-
nick, J. Tallman, D. Greenblatt, and S. Paul, eds., MacMillan
Publishing Co., London.
Skolnick, P., and Paul, S., 1983, New concepts in the neurobiology
of anxiety, J. Clin. Psychiatry 44:12.
Squires, R., 1981, GABA receptors regulate the affinities of anions
required for brain specific benzodiazepine binding, in: "GABA
and Benzodiazepine Receptors," E. Costa, G. di Chiara, and G.
Gessa, eds., Raven Press, New York.
Tenen, S., and Hirsch, J., 1980, S-Carboline-3-ca rboxylic acid ethyl
ester antagonizes diazepam activity, Nature 288:609.

758
BENZODIAZEPINE AND NON-BENZODIAZEPINE ANXIOLITICS

Harold Goldberg

Director
West-Ros-Park Mental Health Center
Roslindale, Massachusetts

'
Everyone experiences anxiety at some time in life. In fact,
a mild degree of anxiety is frequently constructive since it
motivates and often increases alertness, However, more severe
anxiety may be counterproductiv e, preoccupying the individual with
feelings of distress and frequently producing somatic symptoms.
When anxiety reaches this proportion, it can be classified as an
anxiety disorder.

Approximately 2% to 4% of the population has had an anxiety

disorder at some time. Generalized anxiety of a least one month's
duration, occurs with signs of motor tension, autonomic hyper-
activity, apprehensive expectation, and vigilance and scanning.

Subtypes of anxiety include: situational, free-floating, anti-

cipatory, traumatic anxiety and phobias. Anxiety can occur with
depression or secondary to ingesting a medication or a medical
condition.

Frequently, abnormal anxiety and depression will coexist,

producing numerous somatic symptoms as well as psychologic distress,

PHARMACOLOGICAL TREATMENT OF ANXIETY

Ethyl alcohol, This is the most common self-prescribed antianxiety

drug, For example, a person who is anxious before attending a
party or a new or unfamiliar event might be tempted to have a few
drinks to loosen up. Over the years I have treated a number of
individuals who were able to function at a marginal level, in spite
of the presence of severe agoraphobia, by drinking large amounts of
alcohol. Many of these patients, however, do develop alcoholism.
759
 The difficulties with alcohol are that it does not have con-
sistently positive effects on mood. Alcohol has a narrow thera-
peutic index, and may, in fact, increase anxiety, depression, and
dysphoria.

Barbiturates. Barbiturates quickly cause increase in mitochon-

drial enzyme production which speeds up the metabolism of many
psychotropic drugs, including antipsychotic and antidepressant
drugs and the barbiturates themselves. Thus, barbiturates induce
metabolic tolerance, leading to increased dosage. These drugs are
severely addicting, and addiction with severe withdrawal effects
can occur.

Barbiturates may aggravate depression and are the most commonly

used drugs by successful suicides. In elderly and arteriosclerotic
patients, barbiturates may produce a prolonged confused state or
have a paradoxical exitatory effect.

Nearly all studies comparing barbiturates with benzodiazepines

favored the benzodiazepines for the treatment of anxiety.

Diphenylmethane derivatives. The most important property of hydroxy-

zine and diphenhydramine is thought to be their antihistaminic
activity.

These drugs have not been readily accepted or used extensively

as the anxiolitic agents. In a study done by my research group,
hydroxyzine was less effective than placebo in the treatment of
moderate anxiety.l

Glycerol and Propanediol derivatives. In a review of studies, Green-

blatt and Shader found meprobamate superior to placebo in only five
of 26 studies, and superior to barbiturates in only one of ten
studies. 2
Antipsychotics. Over the years, numerous comparisons between the
benzodiazepines and neuroleptics in individuals who are not psychotic
have been done. However, because of the delayed toxic potential of
antipsychotic drug treatment, particularly tardive dyskinesia, an
antipsychotic agent should not be used to treat nonpsychotic anxiety
unless it is severe and unresponsive to other less toxic agents.

Antidepressants. Tricyclic antidepressants are useful in treating

anxiety commonly found in tricyclic responsive depression and may
be the drug of choice for those with anxiety and depression when the
latter is predominant.

Another major use is the treatment of panic attacks.

In both the agoraphobic and those with panic attacks, numerous

760
studies have revealed the utility of the tricyclic antidepressants,
most notably imipramine hydrochloride and the monoamine oxidase
inhibitors.3, 4

They seem to be most helpful in acute stress situations, such

as public speaking, where the possible psychomotor or intellectual
impairment produced by benzodiazepines is undesirable.

Benzodiazepines. The benzodiazepines are among the most widely

prescribed drugs in the world today, but recently adverse publicity
has caused many patients to fear taking them and prompted many
physicians to question prescribing them.

As a group, the benzodiazepines are clearly superior in anxio-

litic activity to the other older classes of antianxiety agents and
are much less toxic at clinically useful doses.

Benzodiazepines are often lipid soluble, assuring rapid gastro-

intestinal absorption within 15 to 20 minutes. Peak blood levels
are found 60 to 90 minutes after oral ingestion. Plasma albumin
binding is considerable. Redistribution to the fat stores accounts
for the extended half-lives of some of the drugs. The major metabo-
lite for many of the long half-life drugs is nordesmethyldiazepam,
an active metabolite with a half-life of more than 70 hours.

The safety of these drugs derives from the fact that their
therapeutic index is about 10 times that of the barbiturates. Two
hundred to a thousand times the average effective dose has been con-
sumed in suicide attempts with recovery. Over one hundred times the
average dose has been taken without inducing a deep coma.

Metabolism and Biotransformation. The benzodiazepines are metabo-

lized in two ways. The short-acting benzodiazepines such as loraze-
pam and oxazepam do not undergo oxidation and do not form active
metabolites. Instead, they are transformed by conjugation to water-
soluble glucuronides and excreted directly into the urine. The meta-
bolic products of glucuronidization are pharmacologically inactive.

The long-acting benzodiazpines are transtormed by oxidation.

The initial step involves demethylation in the liver followed by
aliphatic hydroxylation. These oxidative transformations occur very
slowly, and many of the metabolic products are active. The most
common active metabolite is N-desmethyldiazepam, which has a half-
life of greater than 70 hours.

When these drugs are given multiple dosing circumstances, half-

life becomes important in determining how long the drug remains in
the body. Steady-state conditions are achieved more quickly with
the short half-life drugs than with long half-life drugs.

761
 Withdrawal occurs more suddenly with the short half-life drugs,
since elimination after the last dose is quite rapid, taking place
in about two days. (Elimination of long-acting drugs takes over a
week.) Withdrawal with the short-acting drugs is more intense but
briefer than with long-acting drugs. Nightmares have been described
and seizures can occur one to four days after discontinuing the drug.
Over 50 instances of convulsions coinciding with the withdrawal have
been reported.

It should be noted that clorazepate and prazepam are precursors

of pro-drugs of nordiazepam. Both require the presence of gastric
acid for conversion into the active compound. Therefore, they should
not be used if the patient is taking antacids or has anacidity.

A major concern is the interaction of benzodiazepines with other

psychotherapeutic drugs, antihistamines, and alcohol. The effects
of benzodiazepines and alcohol are increased synergistically when
given in combination and the combination leads to excess sedation.
It has also been determined that cimetidine delays benzodiazepine
metabolism, so dose adjustment is necessary when these drugs are
combined. Tobacco may also reduce benzodiazepine effect.

Azaspirodecanediones - An even newer class of promising anxioselec-

tive agents has recently been described. The first and representa-
tive member is buspirone, one of a series of azaspirodecanedione
psychotro~ic agents prepared by Wu and Rayburn in the early
1970's.5, ,7 Buspirone does not bear an obvious structural resem-
blance to other anxiolytic agents that act either by inhibiting the
GBA recognition site or an associated chloride ionophore site.

Biochemically, buspirone is unlike benzodiazepine in that it

neither stimulates nor inhibits 3H benzodiazepine binding. It does
not affect the influence of GABA or halide anion on 3H benzodiazepine
binding, nor does it interfere with GABA binding or uptake. It
appears to interact only with the dopaminergic system and possesses
properties that are similar to both dopamine agonists and dopamine
antagonists.8 This implies that dopamine may play a role in the
ideology, expression, and pharmacotherapy of anxiety.

Buspirone lacks anticonvulsant activity, interacts minimally

with CNS depressants, and does not cause muscle relaxation. Its
tranquilizing activity is characterized by the ability to one, tame
aggressive rhesus monkeys; two, block the conditioned-avoidance
response in the rat; three, inhibit shock-induced fighting in the
mouse; and four, attenuate shock-induced suppression of drinking in
the rat.

A study by Moskowitz and Smiley9 comparing the effects of long-

term therapy with buspirone and diazepam on driving-related skills
showed that performance impairment associated with diazepam con-

762
trastea w1th improved performance with buspirone. Moreover, when
buspirone was combined with alcohol, the trend toward improvement
offset some of the impairment due to alcohol.

A study by Cole, et allO suggests that buspirone at higher doses

will not be abused recreationally if the compound is marketed for
use in anxiety states. The lower, 10 mg. dose, had no overt euphoria
or sedative effects compared to placebo. In contrast, diazepam at
20 mg, showed some euphoric effects compared to placebo.

A study by Rickelsll concluded that buspirone possessed anxio-

litic properties quite similar to those of diazepam. Diazepam seemed
to be slightly favored in treating somatic symptoms of anxiety,
while buspirone wws favored in treatment of symptoms indicative of
cognitive and interpersonal problems, such as anger and hostility.
Differences between the drugs were seen in terms of their side
effects. Diazepam produced significantly more sedation than buspi-
rone. Buspirone produced slightly more faintness or dizziness and
headaches. These results are similar to the first controlled study
of buspirone's anxiolitic effects reported by my Outpatient Psycho-
pharmacology Research Group.

Our group conducted three studies which consistently showed

buspirone to have excellent antianxiety activity, first in an open-
dose response to study, then in a double-blind comparative study
with diazepam and placebo, and, finally, in a third double-blind
study comparing.buspirone to clorazepate. Buspirone consistently
relieved anxiety and depression associated with .:i~xi~ty. There
was a low incidence of reported side effects, and sedation was
seen less often with buspirone than with diazepam or cloraze-
pate.l2,13

Studies by Wheatleyl4 support these findings, again showing

a higher incidence of drowsiness in the diazepam group than the
placebo group.

In a side effect profile study by Newtonl5 of approximately

700 patients receiving buspirone, sedation, lethargy, and depression
were comparable to placebo and were significantly less than with
the diazepam or clorazepate.

Feigner,l6 concluding his report of 100 patients on either

buspirone or diazepam, writes "Buspirone was slightly more effective
on many specific indices of improvement. Diazepam was associated
with a strikingly higher incidence of adverse reaction. On the
basis of these results one can reasonably conclude that buspirone
is a better choice in the treatment of generalized anxiety. In
addition, it appears that buspirone may be particularly indicated
for anxious patients who have a strong depressive component."

763
REFERENCES

1. Vistaril In the Treatment of Neuroses, Psychosomatics, Vol. XIV,

No,l. Presented at the 19th Annual Meeting of Academy of
Psychosomatic Medicine, San Diego, CA (1972)
2. R.J. Greenblatt and R.I. Shader., Pharmacotherapy of Anxiety
with Benzodiazepines and B-adrenergic Blockers, in Psycho-
pharmacology: A Generation of Progress. M.A. Lipton,
A. DiMascio, and K. Killam, eds., Raven Press, N.Y. (1974)
3. F. Donald Klein, C.H. Zitrin, and M. Woerner, Antidepressants,
Anxiety, Panic and Phobia, in: Psychopharmacology, A Genera-
tion of Progress, M.A. Lipton, A. DiMascio, and K.F. Killam,
eds., Raven Press, New York (1978)
4. D.U. Sheehan, Panic Attacks and Phobias, New England Journal of
Medicine, (1982)
5. Y.H. Wu, J.W. Rayburn, and L.E. Allen, U.S. Patent 3,717,634
(1973)
6. G.P. Casten, G.R. McKinney, and R.E. Newton, U.S. Patent
4,182, 763, (1980)
7. Y.H. Wu, J.W. Rayburn and L.E. Allen, Psychosedative Agents:
II. 5-(4-substituted 1-piperazinylaklyl -8-azaspiro (4.5)
decane-7, 9-diones. J.Med. Chem. 15:477, (1972)
8. L.A. Riblet, D.P. Taylor, M.S. Eison, Buspirone: Pharmacology
and Neurochemistry, J.Clin. Psychiatry, 43:12, (1982
9. H. Moskowitz, A. Smiley, The Effects of Chronically Administered
Buspirone and Diazepam on Driving-related Skills Performance.
Ibid.
10. J.O. Cole and M.H. Orzack, Assessment of the Abuse Liability of
Buspirone in Recreational Sedative Users, Ibid.
11. K. Rickels, K. Weisman and D.O. Norstrad, Buspirone and Diaze-
pam in Anxiety: A Controlled Study, Ibid.
12. H. Goldberg, R.J. Finnerty, Comparative Efficacy of Buspirone
and Diazepam in the Treatment of Anxiety. Am.J. Psychiatry,
136:1184, (1979)
13. H. Goldberg, R. Finnerty, Comparison of Buspirone in Two Sep.
Studies. J.Clin,Psychiatry (1982)
14. D. Wheatley, Buspirone, Multicenter Efficacy Study, Ibid.
15. R.W. Newton, G,P. Casion, D.R. Alms, The Side Effect Profile of
Buspirone in Comparison to Active Controls and Placebo, Ibid.
16. J.P. Feighner, A Double-blind Comparison of Buspirone and
Diazepam in Outpatients with Generalized Anxiety Disorder.
Ibid.

764
MITRAL VALVE PROLAPSE: A MIRROR FOR ANXIETY?

Harisios Boudoulas, Bernard D. King and Charles F. Wooley

The Ohio State University College of Medicine

Columbus, Ohio

The or1g1n of symptoms in patients with mitral valve prolapse

(MVP) remains an enigma. If one accepts the premise that many
patients with this condition were previously considered to have
irritable heart, soldier's heart, effort syndrome, neurocirculatory
asthenia, or anxiety neurosis, then physicians have been attempting
to fathom this phenomenon for more than a century.

The present report provides information from three separate

studies of patients with MVP performed by the authors at the Ohio
State University.

ANALYSIS OF SYMPTOMS

Eighty-six males and 227 females with auscultatory MVP refer-

red for evaluation from 1963 to 1977 were studied. Symptoms, age at
symptom onset, and initial symptom were analyzed.

Mean age of symptom onset was 30.6 years in men and 29.9 years
in women.

Chest pain was the initial complaint by 50% of men and 36% of
women. Conversely, more women (40%) presented initially with palpi-
tations than men (21%). Neurologic symptoms, including syncope, were
present initially in 10% of both men and women.

Chest pain was noted by 60% of men and women. Palpitations,

arrhythmia, fatigue, and dyspnea were more frequent in women (75%
vs 48%, 54% vs 34%, 48% vs 28%, and 34% vs 16% respectively p<O.Ol).
Neurologic symptoms including syncope were present in 15%. The
women had more symptoms than men.

765
 In summary, symptomatic MVP patients have clearly defined
symptoms. The mean age of initial symptom is 30 years for both
sexes. Chest pain is more often the initial symptom in men;
palpitations are more common initially in women. Fatigue, palpi-
tations, dyspnea and arrhythmias are more frequent in women. Chest
pain and neurologic events occur with the same frequency in both
sexes. Women have more symptoms than men.

METABOLIC STUDIES

Eighteen symptomatic females and two males, randomly selected

with auscultatory, phonocardiograph ic, and echocardiographi c find-
ings of MVP were admitted to the Clinical Research Center four days
to standardize the environmental stimulation. They were placed on
a caffeine-free diet.

Twenty-four hour urinary E, NE, and E plus NE were greater

(p < 0.01) in symptomatic patients with MVP compared to controls.

The day-to-day 24-hour E and NE excretion was not significantly

different in the study patients.

Symptomatic patients with MVP had a normal response to oral

glucose, but plasma glucose levels were higher than those of the
control group at two and three hours, plasma insulin levels were
higher three hours after glucose administration.

Thyroid function test results were similar in symptomatic MVP

patients and control patients.

The 24-hour urinary 17-KS and 17-0HCS and plasma cortisol were
also similar in symptomatic patients and the control patients. The
diurnal rhythm of cortisol secretion was normal in the MVP patients.

In summary, symptomatic MVP patients have normal thyroid func-

tion tests, normal plasma cortisol, normal diurnal variation of
cortisol and normal 24-hour urinary 17-KS and 17-0HCS excretion.
They have a normal response to oral glucose but higher glucose and
insulin levels than control patients. Symptomatic patients with
MVP have high adrenergic tone, manifested by the increased 24-hour
urinary catecholamine excretion.

ISOPROTERENOL INFUSIONS

Increased adrenergic tone in symptomatic MVP patients prompted

a study of adrenergic stimulation response.

Serial isoproterenol infusions were performed in 16 symptomatic

patients with MVP (12 females and 4 males). After 30 minutes of
recumbent rest, normal saline was infused for 30 minutes.

766
 Isoproterenol infusions produced a greater increase in heart
rate in symptomatic patients compared to control patients. Changes
in heart rate in symptomatic patients were significantly greater
compared to control patients throughout the infusion period.

Patients were not aware whether or not saline or isoproterenol

was being infused. Excluding palpitations, none of the control
patients developed symptoms during isoproterenol infusion. Iso-
proterenol infusion reproduced symptoms in 14 of 16 symptomatic
patients; saline infusion did not produce symptoms.

Symptom production during isoproterenol infusion was dose-

related. Three of sixteen patients developed symptoms with O.Spg/
min of isoproterenol infusion, five of thirteen patients developed
symptoms with 1.0 pg/min of isoproterenol infusion, and nine of
eleven patients developed symptoms with 2.0 pg/min of isoproterenol
infusion.

Isoproterenol infusion related symptoms included chest pain (7),

extreme fatigue (6), dyspnea (6), dizziness (4), numbness (2). Hyper-
ventilation was observed in only two patients. Four had cool hands
during isoproterenol infusion.

In summary, symptomatic MVP patients have high resting adrener-

gic tone and hypersensitivity to isoproterenol infusion, which sug-
gests that their symptoms may be catecholamine-related or mediated.

The relationship between the anatomic and the mitral valve pro-
lapse syndrome requires further definition. An independent, bio-
chemically-mediated anxiety state existing in an individual with
mitral valve prolapse is one potential explanation for the syndrome.
Equally plausible is that the anatomic entity is a specific marker
for the constitutional, neuroendocrine-cardiovascular process cur-
rently designated as the mitral valve prolapse syndrome. With better
understanding of the central and peripheral biochemistry of anxiety
and stress, the relationship between symptoms in mitral valve pro-
lapse and anxiety and tension will become more clear.

REFERENCES

1. C.F. Wooley, Where are the diseases of yesteryear? DaCosta's

Syndrome, Soldier's Heart, the Effort Syndrome, Neurocircula-
tory Asthenis and the Mitral Valve Prolapse Syndrome.
2. H. Boudoulas, J.C. Reynolds, E. Mazzaferri, C.F. Wooley,
Metabolic studies in mitral valve prolapse syndrome. A
neuroendocrine-cardiovascular process. Circ 61:1200 (1980).

767
 3. F.A. Gafney, E.S. Karlsson, W. Campbell, J.E. Schutte, J.V.
Nixon, J.T. Willerson, C.G. Blomqvist, Autonomic dysfunction
in women with mitral valve prolapse syndrome. Circ 59:894
(1979).
4. H.C. Coghlen, P. Phares, M. Cowley, D. Copley, T.N. James,
Dysantonomia in mitral valve prolapse. Am J Med 67:236 (1979).
5. A. Pasternac, T.F. Tubern, P.E. Puddu, R.B. Kral, J. Champlain,
Increased plasma catecholamine levels in patients with sympto-
matic mitral valve prolapse. Am J Med 73:783 (1982).
6. E.D. Frohlich, H.P. Dustan, R.C. Tarazi, Hyperdynamic beta-
adrenergic circulatory state. An overview. Arch Intern Med
126:1068 (1970).
7. S.F. Pariser, E.R. Pinta, B.A. Jones, Mitral valve prolapse
syndrome and anxiety neurosis/panic disorders. Am J
Psychiatry 135:246 (1978).
8. J.S. Kantor, C.M. Zitrin, S.M. Zeldis, Mitral valve prolapse
syndrome in agoraphobic patients. Am J Psychiatry 137:467 (1980).
9. R.R. Grove, D.L. Pauls, D.J. Slymen, R. Noyes, A family study of
anxiety neurosis. Morbidity risk in families of patients with
and without mitral valve prolapse. Arch Gen Psychiatry 37:77
(1980).
10. B. King, H. Boudoulas, M.E. Fontana, C.F. Wooley, Mitral valve
prolapse syndrome. Anthropometric, sexual, and clinical features.
Am J Cardiol 45:443 (1980)
11. H. Boudoulas, R.W. Clark, P. Geleris, H. Schmidt, S.F. Schaal,
R. P. Lewis, Autopometric characteristics, cardiac abnormalities
and adrenergic activity in patients with primary disorders of
sleep. J Medicine (in press).
12. R.B. Bervereux, T. Brown, R. Kramer-Fox, I. Sachs, Inheritance
of mitral valve prolapse: Effect of age and sex on gene
expression. Ann Int Med 97:826 (1982).
13. H. Boudoulas, C.F. Wooley, Chest pain associated with mitral
valve prolapse in Chest Pain, Problems in Differential Diagnosis,
edited by J. Hurst, H.M Spiro, New York, Biomedical Information
Corporation 5:1 (1979).
14. H. Boudoulas, B. King, J. Reynolds, E. Mazzaferri, M.E. Fontana,
C.F. Wooley, Mitral valve prolapse syndrome. A constitutional-
neuroendocrine-cardiovascular process. Clinical Research 27:
672A (1979).
15. D.E. Redmond, Jr, New and old evidence for the involvement of a
brain norepinephrine system in anxiety, in Phenomenology and
Treatment of Anxiety. W.E. Fann, edit. ~P. Medical and
Scientific Books, New York, London, (1979).
16. H. Boudoulas, J.C. Reynolds, E. Mazzaferri, C.F. Wooley, Mitral
valve prolapse syndrome: The effect of adrenergic stimulation.
J Amer Col Cardiol : Oct (1983).

768
COMPONENTS OF HUMAN STRESS RESPONSE:
COPING WITH HOSPITALIZATION IN ADOLESCENTS

Hans Steiner and Seymour Levine

Department of Psychiatry and Behavioral Sciences

Stanford University School of t1edicine
Stanford, CA

Human stress response (HSR) has been linked to a variety of

pathological processes (Ader, 1982) such as autoimmune and neo-
plastic diseases. By contrast the relationship of HSR to adaptation
is much less clear (Rutter, 1981). There does, however, seem to
be support for the assumption that less well-adapted individuals
are likely to have a more pronounced hormonal response (Miyabo,
1979) and can have inadequate psychological (Katz et al ., 1970)
response as well.
This study concerns the relationship between HSR and adaptation
in an adolescent population. Events surrounding hospitalization
(e.g., ·admission, surgery) have been shown to be stressful to a wide
range of subjects. We compared the impact of elective psychiatric
admission to elective orthopedic surgery on adolescent subjects.
We assessed arousal, defense and coping, social support, and plasma
cortisol levels.
There is evidence that less well-adapted individuals show
deficiencies in both arousal and defense and coping (Miyabo et al .,
1979; Katz et al ., 1970). These deficiencies were once thought
to be situational, but there is evidence suggesting that they may
be stable characteristics (Gorzynski et al., 1980), making them
even more interesting to study because they might serve as markers
for an individual's adaptation. Exposing the individual to a
stressor (such as hospitalization) might provide information about
reactions to other stressors.
Social support has been shown to contribute moderately, but
reliably to adaptation. It seems to exert its influence
predominantly by mitigating stress.
769
 The activity of the hypothalamic-pituitary-adrenal (HPA) axis
has repeatedly been shown to respond powerfully to psychological
stressors (Mason, 1975; Katz et al., 1970; Gorzynski et al., 1980;
Miyabo, 1979; Sakellaris et al ., 1975). Other facts make the HPA
axis a pertinent subject for study: ACTH has been shown directly
to influence Reurotransmission after only 15-30 minute exposure
(Hohnston et al., 1983). Adrenal corticoids have been shown to
influence anabolism and catabolism of monoamines (Yuwiler, 1982)
operating in such vital functions as mood maintenance, regulation
of hunger and hormone release itself. Corticosterone has been
shown to have direct influences on immunocompetence (Ader, 1982).
We postulated the following distinctions between normals and
psychiatric patients.
1) Normals would have lower baseline levels of cortisol during
hospitalization.
2) Normals would show a lesser cortisol response following a
stress interview about hospitalization (Miyabo et al ., 1979).
3) Normals would show higher defense ratings (Katz et al .,
1970).
4) Normals would show better coping in the hospital.
5) Normals would receive higher social support ratings and have
more congruent families.

SUBJECTS AND METHODS

We recruited a total of ten adolescent subjects for our study:
five from the Psychosomatic Inpatient and five from the Orthopedic
Service. They were comparable in terms of age (x=l5, range: 12-17),
sex, weight, height, socio-economic status, mean onset of sleep,
and number of previous hospitalizations. None of them were on
medications that would interfere with the activity of the HPA axis.
Their weight during hospitalization remained stable and none of
them had intercurrent febrile illnesses. Psychiatric diagnoses
were: dysthymia, adjustment reaction, conduct disorder, conversion
disorder, borderline personality.
The subjects were assessed in the following manner.
Arousal~ Defense and Coping
On the day of admission, all subjects were seen for a standard
psychiatric interview and mental status examination lasting 1-2
hours. Psychiatric diagnosis was by DSM-III criteria. During
this interview, the interviewer evaluated, in a semistructured
format, affect and defense according to the criteria developed
by Katz et al ., 1970}. This assessment was repeated on all days

770
when plasma cortisol was drawn. A brief 10-15 minute interview
preceded the blood drawing and covered stressful events of the
day.
We developed a 33-item coping scale which included the most
common items of self-care, and extended nursing care. Nurses as-
signed to the patient on days when plasma-cortisol was obtained
completed the scale. The resultant scores were expressed as per-
centage points achieved per applicable items, thus equalizing,
the effect of post-operative state.
Social Support Assessment
The interviewer rated the degree of social support present in
each family on a 12-point scale. Judged were a patient's description
of family and friends, observed interactions of the patient and
the family during the first week of admission.
All patients and their parents completed the Family Environment
Scale (FES) developed by Moos (1982). We calculated the incongruence
score and converted it to standard scores.
Cortisol Assessment
Plasma cortisol levels were drawn within 4 hours mean onset
of sleep, a period that has essentially been shown to be free of
pulsatile release. Only samples were accepted that were obtained
within 3 minutes from the start of venipuncture. Blood was drawn
into heparinized tubes and was immediately centrifuged. Super-
natent plasma was removed, coded and frozen for storage. Cortisol
levels were determi·ned by radioimmunoassay. Blood for plasma
cortisol levels was drawn on days 1, 3, 5, 7, and weekly thereafter
if necessary.

RESULTS
Table 1 summarizes our positive results. Defense ratings
were not significantly different in the two groups. Only the
affect rating component of this measure showed a strong trend in
the predicted direction; the psychiatric patients tending to
exhibit more negative affect than normals. However, our pre-
dictions about coping, social support and cortisol were all
supported.
The differences in mean cortisol levels between the groups
showed a strong trend in the predicted direction. On closer
examination, this was the function of seven peaks of cortisol
(ranging from 7.7 mg% to 18.1 mg%) which occurred in six patients:
one psychiatric patient had two peaks, four psychiatric patients

771
 Table 1. Positive Results
Norma 1 Psychiatric Mann-
Sample Sample Whitney U p

X SEM -
X SEt·1
Baseline cortisol 4 .1 ( 1 . 42) 6 .1 (1 .8) 4 NS (trend)
Peak post-stress 5.6 (2.9) 12.24 (4.2) 2 <0.05
cortisol
Affect rating 1 .0 (0.4) 1.25 (0.4) 4 NS (trend)
Coping rating 0.86 (0.2) 0.5 (0.2) <0.05
FES incongruence 12 .1 (4.3) 26.0 (6. 7) <0.05
Social support 10 .4 (0.9) 4.3 ( 1 . 9) <0.05
rating

had each one peak, but only one normal had one peak. By converting
these elevations into Z-scores and identifying as "responders"
those values above 1 SD, four of the five psychiatric patients
and the one normal could so be classified.
Topics of discussion associated with the elevations were:
1) pending or threatened discharge (4, all in psychiatric patients);
2) surgery (3, 1 in 1 normal and 2 in 1 psychiatric patient who had
two operations, each followed by an elevation). There were no
elevations associated with admission.

DISCUSSION
The response profile associated with psychiatric disability is
generally in keeping with our expectations. Psychiatric disability
seems to be associated with increased reactivity in the HPA system.
Sakellaris et al. (1975) have described increased rate of response
in the HPA axis in rats which had adapted to chronic stress. It
is possible that psychiatric disability is associated with chronic
stress due to deficient defense and coping.
Much to our surprise, admission was not associated at all with
cortisol response. Children's Hospital is indeed a most supportive,
friendly environment (Steiner, 1982) and seems to prepare patients
well for admission. Thus, our elective admissions might not be as
stressful as acute admissions to other settings. Discharge, by
contrast, was clearly an issue for four of the five psychiatric

772
patients, but none of the five normals. Return to a non-
supportive environment resulted in considerable elevations in the
patients. The characteristics of the families were clearly evident
to raters and on the FES. It was disappointing that our defense
ratings failed to distinguish between the two groups. It is
possible that the time allotted for assessment was too brief. The
trend for negative affect to distinguish between the two groups would
suggest that the use of standardized affect rating scales such as
the STAI might give better results. This could be complemented by
improved techniques of defense assessment which recently have become
available (Bond et al., 1983; Steiner et al., 1983).

REFERENCES
Ader, R., 1982, "Psycho Neuro Immunology," Academic Press, New York.
Bond, M., Gardner, S.T., Christian, J., and Sigal, J. J., 1983,
Empirical study of self-rated defense styles, Archives of
General Psychiatry, 40:333.
Gorzynski, J. G., -Holland, J., Katz, J. C., and Weiner, H., 1980,
Stability of ego defenses and endocrine responses in women
prior to breast biopsy and ten years later, Psychosomatic
Medicine, 42:323.
Johnston, M. F., Dravitz, E. A., Meiri, H., and Rahamimoff, R.,
1983, ACTH causes long-lasting potentiation of transmitter
release from frog motor nerve terminals, Science, 220:1071.
Katz, J. L., Ackman, P., Rothwax, Y., and Sachar, E. J., 1970,
Psycho-endocrine aspects of cancer of the breast, Psycho-
somatic Medicine, 32:1.
Mason, J. W., ~Emotion as reflected in patterns of endocrine
integration, in: "Emotions- Their Parameters and Measure-
ment," Raven Press, New York.
Miyabo, S., Asato, T., and Mitushima, N., 1979, Psychological
correlates of stress-induced cortisol and growth hormone
releases in neurotic patients, Psychosom Med, 41:515.
Moos, R.H., and Moos, B.S., 1982, "Family Environment Scale Manual,"
Consulting Psychologists Press, Palo Alto.
Rutter, M., 1981, Stress, coping and development: some issues and
some questions, J Child Psychol Psychiat, 22:323.
Sakellaris, P. C., Vernikos-Danellls, J., 1975, Increased rate of
response of the pituitary-adrenal system in rats adapted
to clinic stress, Endocrinology, 97:597.
Steiner, H., 1982, The soc1o-therapeutic environment of a child
psychosomatic ward, Child Psychiatry and Human Development,
1 3 : 71 . -
Steiner, H., and Anders, T. F., 1983, Speech correlates of stress
and coping (submitted for publication).
Syme, L. H., 1983, Sociocultural factors and disease etiology,
in: "Handbook of Behavioral Medicine," D. Gentry, ed.,
Guilford Press, New York (in press).

773
Yuwiler, A., 1982, Bio-behavioral consequences of experimental
early life stress, in: "Critical Issues in Behavioral
Medicine," L. J. West and M. Stein, eds., Lippincott,
Phil a del phi a.

774·
SPEECH CORRELATES OF STRESS AND COPING

Hans Steiner and Thomas F. Anders

Department of Psychiatry and Behavioral Sciences

Stanford University School of Medicine
Stan ford, CA

INTRODUCTION
After several decades of dedicated research, the precise
relationship between psychological and physiological components
of acute stress response is still not known (Elliott et al ., 1982).
The problem is of particular interest for three reasons:
(1) Psychological stimuli are among the most powerful activators
of physiological stress responses (Mason, 1975).
(2) Coping and defense mechanisms are able to protect the organism
from arousal (Lazarus, 1964) or even prevent lethal conse-
quences of hormonal stress response (MacLennan et al., 1983).
(3) Without proper knowledge of the relationship between these
variables, significant progress in stress response is unlikely.
Individual responses show a wide variability, impeding experi-
mental exploration. It is thought that this variability is at
least in part a function of the large number of unknown media-
tors in the response. And, in turn, a great number of those
mediators are psychological (Elliott et al ., 1982).
One of the major obstacles to resolving the problem is
the lack of adequate psychological instruments (Moos, 1974).
However, recent advances in psycholinguistics have made it
possible to measure psychological arousal, coping and defense
through speech, thus enabling us to develop our psycholinguistic
assessment protocol of psychological stress response.

775
BRIEF REVIEW OF THE LITERATURE
Any adequate assessment procedure must be able to distinguish
arousal, defense and coping as separate variables, because the
interplay between these components of stress response is to be
studied. Thinking, for example, of coping as the absence of arousal
as is frequently done in the physiologically oriented literature
has obvious weaknesses and is unacceptable. Assessment procedures
must also be sensitive enough to measure rapid changes in psycho-
logical stress response. Physiological stress response arises and
abates sometimes within minutes (Miyabo, 1979). We assume that
psychological stress response shares some of these characteristics;
however, adequate methods to measure such changes in the psycho-
logical stress response system do not exist. ·
Currently available instruments for the assessment of coping
and defense in human acute stress situations are mostly limited
to self-report questionnaires (e.g. Bond et al ., 1983) and observer
ratings (Vaillant, 1977). Both methods are subjective. Self-
report measures are further limited because of subjects report
on mechanisms which may operate largely outside of a person•s
awareness. Finally, repeated measurements needed in such stress
research are difficult to obtain using questionnaires because of
the danger of responses becoming stereotyped (Gottschalk, 1977).
In contrast, analysis of spontaneously produced speech seems more
appropriate to the task. Speech productions can be recorded,
transcribed and scored by independent raters. Speech productions
follow internal rules which are responsive to external circum-
stances (Scherer, 1981), yet the rules are, for the most part, not
under the individual •s conscious control, thus eliminating
subjective bias.
Three types of speech analysis are commonly used:
(1) Paralinguistic Analysis. Regarded as the most powerful
indicator of emotional arousal (Scherer, 1981).
(2) Content Analysis. Related to both emotional tone and
personality tra1ts (Gottschalk, 1977).
(3) Syntactical Analysis. Correlates stylistic elements of
speech, personality traits and behavior (Weintraub, 1981 ).
Emotional arousal in speech can be measured in a variety of
ways (Harper et al ., 1978). The best-established measure is
Mahl 's speech fluency disturbance rate (SFD/1000) (Mahl, 1963;
Scherer, 1981; Harper et al ., 1978). Flustered or confused speech
has been studied extensively by Mahl and his colleagues, and the
results have been replicated by at least four other independent
groups of investigators. Mahl's category of "non-ah" speech
disturbances correlates with situational anxiety (R;+.59), and
with galvanic skin response (R;-.38). Compared to content-analysis

776
approaches to the measurement of emotional states, Mahl •s procedure
has the advantage of measuring unavowed and unacknowledged emotions.
Unlike verbal content, speech dysfluencies are largely outside of
an individual •s conscious control.
For our defense and coping assessment, we selected the syntac-
tical approach of Weintraub (1981). Equating defense with style
of speech as reflected in the use of distinct syntactical categories
offers a convenient quantitative approach. Syntax, being free of
meaning, lends itself to easy recognition and largely operates
outside of an individual •s awareness. Both these characteristics
constitute distinct advantages over the content-analysis approaches
such as by Gottschalk (1977).
Empirical support for Weintraub's approach has been mostly
generated by his own group. He constructed fourteen scoring
categories, including the occurrence of pronouns such as I, we, me;
the use of negatives, retractors and nonpersonal clauses. In a
series of clinical studies, Weintraub compared six deviant popula-
tions to a control group of 46 normal subjects. Based on the
presence of impulsive, depressed, delusional, binge eating, com-
pulsive, and alcoholic behaviors predictions were made as to the
syntactical characteristics of free associative speech. In a
majority of the cases, Weintraub reported that deviant groups
differed significantly from control groups and from each other in
the predicted directions. His conclusions presume a link between
syntactical style and manifest behavior. In addition, Weintraub
studied small groups of normal subjects of different ages. He
succeeded in demonstrating a developmental course in speech patterns.
Although these results were obtained using small groups and need
to be replicated with more subjects, they seem promising.

PRELIMINARY STUDIES
We combined arousal and defense measures and applied them to
the study of 10-minute free associative speech samples produced
by subjects in a variety of stressful situations. We have used
our combined Verbal Analysis Protocol (VAP) with several samples.
In the first study, we investigated the relationship between
(1) linguistic performance on a l.O-minute free association task
about any topic, and (2) global clinical ratings of adaptation.
We studied four inpatients and one normal adolescent ranging
in age from 12-17 years. The I.Q.•s of all subjects were within
ten points of each other (range 115-124), and all were free of
expressive and receptive language disorders.
Our hypothesis was that better-adapted individuals would

777
demonstrate better ability than less well-adapted individuals to
cope with a 10-minute free association task by (1) being less
aroused during the task, reflected in lower SFD/1000, and (2) better
adapted individuals would have more defenses available in the ten
minutes reflected in a high DEF/SFD.
Three independent judges (two R.N.'s and one child psychiatrist)
ranked the five in terms of their global adaptation, agreeing
one hundred percent.
The subjects then were asked to speak about anything they
wished for ten minutes. Two independent scorers (who agreed
.79-.82 in all categories) scored the transcripts.
The hypothesis was partially confirmed. There was a moderate
but nonsignificant positive correlation between adaptational rank
and SFD/1000 words (Kendall's tau= +.40). However, the DEF/SFD
correlated significantly with the adaptational ratings (tau = +.80-
1.0), all significant at the p<0.05 level. Of the two measures,
the combined arousal-defense measure seems to be stronger at
predicting adaptation.
In a second study, we explored acute stress response and its
relationship to adaptation in more detail by exposing subjects to
a variety of acute stressors. The aim was to define differential
patterns of arousal and defense and their relationship to general
adaptation. The subjects were exposed to nonspecific stressors
(NSS) (entry into the study, imagined venipuncture, Wilkinson
Addition Test), one specific stressor (SS) (a 10-minute free
association task relating their most stressful life event), and a
recovery task (REC) (a 10-minute free association task after
debriefing, on any subject).
The hypotheses were:
(1) ASS results in increased arousal and defenses compared
to REC and NSS.
(2) A less well adapted subject reaches much higher levels
of arousal during both NSS and SS conditions than the normal
subject (Miyabo, 1979).
(3) A less well adapted subject would show lower order
defenses in general and earlier defensive failure.
We selected two adolescents for this study: one normal
adolescent and one anorexic inpatient whose weight and primary
anorexic symptoms had resolved. They were both 16 years old,
both female with I.Q.'s in the 120's, and their socioeconomic
status and academic performance equal. Both were free of speech
and language disturbances.

778
 The resulting three ten-minute speech segments per subject
were analyzed and compared by means of Kendall's tau and paired
T-tests.
Most of the hypotheses were indeed supported. The normal
subject showed considerably fewer SFD/1 000 than the anorexic
subject (paired two-tailed T-test, t=2.64, p<0.02). The anorexic
subject's arousal showed a trend toward increase from RECto NSS
to SS; however, the differences were not significant. This suggests
that the anorexic subject operated at a high level of arousal
regardless of the condition. The normal subject's SFD/1000
increased significantly from NSS toSS (two-tailed T-test,
p<O.Ol). Our prediction that the normal would not score as high
in level of arousal as the anorexic subject in the SS was not
confirmed. Both subjects had almost exactly the same number of
speech fluency disturbances.
Closer study of the pattern of arousal and defense activation
in all three conditions combined, however, shows differences in
the two subjects, reflecting diagnosis and presumably adaptive
style. The normal subject had a better ratio of total DEF/SFD.
This is even more pronounced in the category of mature defenses.
Thus, the normal adolescent scored higher on both mature and
primitive defenses, contrary to our expectations.
In order to assess the flexibility and failure of defenses in
the two subjects, we looked at each condition separately. Each
ten-minute transcript was examined in two-minute segments. Both
subjects performed similarly in the SS. The main difference
between their performances occurs in the REC and NSS. In the REC
speech sample, the normal subject shows a higher total defense
per SFD ratio. This is true for mature and primitive defenses.
In the NSS condition, the normal subject increases mature
defenses in response to various nonspecific stressors in order
to cope; the anorexic subject relies more on primitive defenses.
This tendency is clearly in line with our hypothesis. All these
differences were significant at the p<0.05 level.
To examine defensive failure, we evaluated the correlation
of each defense category with arousal in each two-minute segment
for each condition separately. Both subjects' defenses showed a
biphasic relationship to arousal: up to a critical value of
SFD/1000, defenses correlated positively (ranging from +.10 to
+.87). In the SS condition, most of the defenses (5/6) correlated
negatively (ranging from -.25 to -.92). However, the normal
subject did not reach her critical point until the SS condition,
while the anorexic did so in the NSS condition. These results
support our hypothesis.

779
 We conclude that psychiatri c morbidity seems associated with
a high level of arousal throughout stressful tasks, the mobiliza-
tion of more primitive defenses to cope with nonspecific stressors,
and earlier defensive failure.
In summary, we were able to develop an approach to arousal
and defense assessment that was successful in independently
identifyin g these two variables. Patterns of arousal and defense
can be examined in two-minute segments of free associativ e speech.
Thus our VAP provides us with the flexibilit y and sensitivit y we
think is so important for the study of acute stress response.
We are in the process of completing our own validity and
reliabilit y studies. Preliminary results are encouraging. We
are further planning combining our assessment with measures of
activity in the hypothalam ic-pituitary -adrenal axis.

REFERENCES
Bond, M., Gardner, S. T., Christian, J., and Sigal, J. J.,
1983, Empirical study of self-rated defense styles,
Archives of Gen Psychiatry , 40:333.
Elliott, G-:- R., and Eisdorfer, C., 1982, "Stress and Human Health,"
Springer, New York.
Gottschalk, L.A., 1977, Quantifica tion and psychological indicators
of emotions: the content analysis of speech and other objec-
tive measures of psychological states, in: "Psychosomatic
Medicine," Z. J. Lipowski, ed., Oxford University Press,
New York.
Harper, R. G., Wiens, A. N., and Matarazzo, J.D., 1978, "Nonverbal
Communication - The State of the Art," Wiley, New York.
Lazarus, R. S., and Alfert, E., 1964, Short circuiting of threat by
experimentally altering cognitive appraisal, J Abnorm Soc
Psychol, 69:195.
Mahl, G. F., 1963, The lexical and linguistic levels in the expres-
sion of the emotions, in: "Expression of the Emotions in
~1an," P. H. Knapp, ed., Internatio nal University Press,
New York.
Mason, J. W., 1975, Emotion as reflected in patterns of endocrine
integratio n, in: "Emotions - Their Parameters and Measure-
ment," L. Levi , ed. , Raven Press, New York.
Maclennan, J., and Maier, S. F., 1983, Coping and the stress-indu ced
potentiatio n of stimulant stereotypy in the past, Science,
21 9:1 091 .
Miyabo, S., Asato, T., and Mitushima, N., 1979, Psychological
correlates of stress-indu ced cortisol and growth hormone
releases in neurotic patients, Psychosom Med, 41:515.
Moos, R. H., 1974, Psychological techniques in the assessment of

780
 adaptive behavior, in: "Coping and Adaptation," G. V. Coelho,
ed., Basic, New YorK:
Scherer, K. R., 1981, Vocal indicators of stress, in: "Speech
Evaluation in Psychiatry," J. K. Darby, ed.-,Grune and
Stratton, New York.
Vaillant, G. E., 1977, "Adaptation to Life," Little, Brown and
Co., Boston.
Weintraub, W., 1981, "Verbal Behavior: Adaptation and Psycho-
pathology," Springer, New York.

781
SOCIAL ENVIRONMENT AND STRESS: A STUDY OF PSYCHIATRIC
AND PSYCHOSOMATIC PATIENTS AND THEIR FAMILIES

Hans Steiner, Cliff Mazer and Judy Sobieski

Department of Psychiatry and Behavioral Sciences
Stanford University School of Medicine
Stanford, CA 94305

There is mounting evidence to suggest that persons subject to

stress suffer less damaging effects if they receive social support.
The relationship between social support and stress has been most
clearly demonstrated in psychiatric disorders (Flaherty et al .,
1983;_ Syme, 1983). Concurrently, the application of family therapy
methods to treatment of psychiatric disorders has led several
groups of investigators (Minuchin, 1978) to examine the relation-
ship between certain family characteristics (i.e. lack of conflict
resolution, enmeshment, rigidity, etc.), and certain conditions
with psychosomatic components such as anorexia nervosa, diabetes
mellitus and bronchial asthma. Thus far empirical support for
these propositions is lacking and properly controlled studies are
absent from the literature.
In the present study we wished to examine further the relation-
ship between family structure and psychosomatic disease. We compare
the kind of social environment, i.e. the family structure, with
diagnostic categories on the one hand, and with severity of disorder
on the other. We guessed that severity of disorder regardless of
kind is related to dimensions of family structure. This idea would
be very much in line with the current social support literature
( Syme , 1983) .
Our specific hypotheses were that:
(1) The amount of organization present in a family is inversely
related to total number of hospitalizations (a simple and
convenient measure of severity of illness). Psychiatric and
psychosomatic patients would have the same results.

783
(2) The more severely ill patients with eating disorders have
families who appear less structured.

SUBJECTS
We evaluated 25 consecutive psychiatric inpatients and 35 con-
secutive anorexic inpatients. These families were comparable in
terms of male/female ratio of patients, mean age of patients, socio-
economic status, and number of previous hospitalizations. Only
intact families and those where the patient/parent tria~ completed
the assessment were considered for the study.

METHODS
All families were evaluated in the first week of their admission
to Children's Hospital at Stanford in order to avoid contaminating
influences of ongoing psychotherapy, For the assessment, we used
the Moos Family Environment Scale (FES) (1982). This is a self-
report measuring scale which produces sub ... scores on ten factors of
family environmental functioning and climate: Cohesion (Coh),
Expressiveness (Ex), Conflict (Con), Independence (Ind), Achievement
Orientation (AO), Intellectual Cultural Orientation (ICO), Active
Recreational Orientation (ARO), Moral Religious Emphasis (MRE),
Organization (Org), and Control (Ctrl). The FES displays a test-
retest reliability range on the ten sub-scores from .68 for
independence to .86 for cohesion. In addition, internal relia-
bilities range from .61 for independence to .78 for cohesion. The
internal consistencies of the device showed an acceptable range
according to the Kuder/Richardson formula 20.
The patients were diagnosed by their individual physicians,
according to DSM-III criteria. The number of hospitalizations
prior to the patients• present admissions was obtained by chart
review.
The symptom severity ratings of the anorexia nervosa patients
was performed according to the Global Clinical Score method by
Garfinkel et al. {1977). This is a composite score which takes
into consideration the patient's weight, eating habits, menstrua-
tion, social adjustment, e~ucational and vocational adjustment.
The ratings were obtained by two independent raters who agreed
R=+ .82 (p <0 .01). We also performed a rating of severity of
anorexia which was based on duration of illness, number of previous
treatment failures and AMA dispositions.
In order to test our first hypothesis, we compared psychiatric
and anorexic families whose children had no previous hospitaliza-
tions to those who did.

784
 In order to test our second hypothesis, we utilized only our
anorexic families, because we wanted to avoid difficulties with
different severity rating systems for psychiatric and anorexic
patients.
Statistical analysis included: one-way analysis of variance
and two-tailed student's T-tests as well as Kendall's Tau for
correlation.

RESULTS
Our first hypothesis was supported. Families of anorexic and
psychiatric patients with no hospitalizations scored significantly
higher on the FES Organization subscale (+=2.65, DF:48, p 0.02;
F test: NS) than families whose children had been hospitalized
previously.
Our second hypothesis was also supported. Families of anorexic
patients which had low scores on severity ratings (less than 11)
had high scores on the FES scales which were said to be reflective
of family structure (Coh, Ex, Ind, Org, ARO, ICO, MRE, AO) (+=2 .57,
DF:33, p <0.02).
Our own severity rating, based on duration of illness, number
of previous therapies and AMA discharges and dispositions correlated
significantly (R+0.52, p<O.Ol) with the Garfinkel Severity Rating.
Duration of illness alone correlated significantly (R=+0.47,
p<O.OOl) with the Garfinkel Severity Rating.

DISCUSSION
These findings were in line with our expectations and corre-
spond to the rest of the social support literature regarding other
illnesses such as depression (Flaherty et al., 1983), complications
of childbirth (Nuckolls et al., 1972), cardiac failure (Chambers
et al., 1958), discharge after surgery (Egbert et al., 1964), and
increased mortality from a variety of illnesses in persons deprived
of meaningful social contact (Syme, 1983). Severity of disorder
does seem to relate to degree of family structure present. By
contrast, our study does not support the model that specific kinds
of family characteristics are to be found in certain diagnostic
sub-groups. (Figure 1 summarizes these results.)
An objection to our study can be directed at the use of the
FES, a relatively insensitive instrument for the measurement of
transactional variables which perhaps reveal themselves only by
analysis of interaction patterns in families. Currently, though,

785
the measurement of observed transactions is notoriously difficult
and unreliable (Cromwell et al., 1976). The FES, with all its
limitations, is the best instrument available for the assessment
of family characteristics. Furthermore, despite its relative
insensitivity, the FES managed to produce significant results in
line with our hypotheses.

r-
..c: 0 Ol s...
0 X 0 u s... +>
u IJ.J c:( ..... 0 u
80 80

70 r- --- --- r--- r--- --- --- 1 - - - --- --- -- 70

Ul
Ul
M-
(!)
s... 60 r- - 60 ~
0
V)
u
4~~~\.
,_7'1\
.,"' '' .JP 0.
PJ
-~ ' ,t X

'
----v
/

t---- ·v··
50 50 ci.
"'0
s...
P"·. ' .. ( p..• -t r1
~ ·. d:r•• •• r p··
V)
n::l
"'0 ~{ ()

s:: 40 r- •• ·' - 40 ~
n::l
+> nr· ro
Ul

--
V)
__ ..!
30 r-- --- --- r--- -- --- -·-- -
I
---' 30

20 20

• Predictions o--o EDO Families

(Minuchin) N = 35
x--x Psychiatric 0··0 "Distressed"
Families Families
N = 25 (Moos, 1981)
N = 500
Figure 1

786
 Our study also does not establish the nature of the relation-
ship between severity and family structure. Thus, the question as
to whether it is the less structured environment which produces
the higher severity or vice versa, or even whether both of these
factors are a function of another unknown variable cannot be
answered. Other prospective studies in the field, though, have
suggested that it is indeed a lack of social support and meaning-
ful contacts which produces increased morbidity. The final
answer to this question can come from the prospective study of
populations at risk for eating disorders.
Further study of the disease/social support relationship is
clearly needed. In general, our data, as well as those of other
groups, consistently suggest that people from intact social
networks or families, people involved with other people in a
meaningful way, have lower rates of disease, have less severe forms
of specific diseases and recover more quickly from them (Syme,
1983). One implication from this data is that people somehow
benefit from the presence of others. 11 Social ties," "social
support 11 and meaningful "social contacts" are all concepts
attempting to capture the essence of this phenomenon. Their
operationalization is still in its infancy.
In summary, we compared family structure and diagnosis to
family structure and severity of anorexia nervosa. There seems
to be no justification for the assumption that anorexia nervosa
is produced by certain types of families. There was a consistent
relationship between family structural measures and severity of
anorexia nervosa, further supporting a link between social
environment and modification of disease process.

REFERENCES
Chambers, W. N., and Reiser, M. F., 1953, Emotional stress in the
precipitation of congestive heart failure, Psychosom Med,
15:38. -
Cromwell, R. E., Olson, D. H. L., and Fournier, D. G., 1976,
Tools and techniques for the diagnosis and evaluation in
marital and family therapy, Family Process, 15:1.
Egbert, L. D., Battit, G. E., Welch-,- C. E., and Bartlett, M. K.,
1964, Reduction of postoperative pain by encouragement
and instruction of patients, N Eng J Med, 270:825.
Flaherty, J. A., Gaviria, F. M., Black, E. M., -and Altman, E., 1983,
The role of social support in the functioning of patients
with unipolar depression, Am J Psychiatry, 140:473.
Garfinkel, P. E., Moldofsky, H., and Garner, D. M., 1977,
Prognosis in anorexia nervosa as influenced by clinical
features, treatment and self-perception, Can Med Assoc J,
117:1 041 .

787
Minuchin, S., Rosman, B. L., and Baker, L., 1978, "Psychosomatic
Fami 1 i es," Harvard, Cambridge.
Moos, R. H., and Moos, B. S., 1981, "Family Environment Seale
Manual," Consulting Psychologist s Press, Palo Alto.
Nuckolls, K. B., Cassel, J., and Kaplan, B. H., 1972, Psycho-
social assets, life crisis and the prognosis of pregnancy,
Am J Epidem, 95:431.
Syme, L. H., 1983, Sociocultura l factors and disease etiology,
in: "Handbook of Behavioral Medicine," D. Gentry, ed.,
Guilford Press, New York (in press).

788
THE LONG-TERM STRESS OF CHRONIC ILLNESS:
AN ILLUSTRATION FROM PEDIATRIC ONCOLOGY

Gregory K. Fritz and Judith Ro Williams

Department of Psychiatry and Behavioral Sciences
Stanford University School of Medicine
Stanford, CA 94305

Major advances in treatment over the past decade have dramati-

cally improved the outlook for children with cancer. r.1alignancy in
childhood now implies an arduous course of therapy and the very real
possibility of a cure. Increasing numbers of children are success-
fully completing treatment and entering the period of off-treatment
follow-up. Although termination of treatment brings release from
years of pain and anxiety for the child and family, our experience
suggests that the time after termination is a vulnerable period.
Many aspects of childhood cancer seem cruelly paradoxical.
The beginning symptoms can be banal, inocuous and long-overlooked,
and the grave diagnosis may come without warning. The treatment
and the side-effects can be more painful than the disease itself.
Children must believe that medications that make them deathly ill
will save their lives. Sophisticated as therapeutic techniques
have become, they are often still part of an experimental protocol.
Survival rates are remarkably improved but individual outcomes
remain uncertain. These facts, and our clinical experience with
oncology patients at the Children's Hospital at Stanford, led to
our interest in the period after treatment ends.
The experiences of children who reach the successful conclusion
of their cancer treatment have been given little attention so far,
primarily because the number of such children has been small. An
exception is the recent interest in the late effects of central
nervous system prophylaxis in leukemia: several reports have
pointed convincingly to decrfased intellectual abilities in children
who received such treatment. ,2 However, psychosocial issues were
not addressed in this work. The most extensive study of long-term
789
pediatric cancer survivors has been carried out at the Sidney
Farber Cancer Institute in Boston, recently presented in book form,
intriguingly titled The Damocles Syndrome.3 This study documented
a high rate of adjustment problems (59%) in the subjects, who were
seen to be at greater psychological risk for a wide spectrum of
difficulties than survivors of other chronic pediatric illnesses.
Prediction of outcome was complex, depending on a variety of
factors including age at diagnosis, the nature and course of the
specific illness, and family variables.
The major limitation of Koocher and O'Malley's work concerns
the applicability of their findings to today's pediatric oncology
population. Their subjects averaged 13 years since diagnosis
and termination occurred as long ago as 1947. A majority of their
114 subjects had experiences that differed from those of current
cancer patients. To determine the nature of the stresses facing
a youngster who survives cancer today and the family, and to
develop interventions that will enhance the long-term psychological
adjustment, we have undertaken an extensive three year research
project.
The work to date consists of a retrospective study to document
the psychological and social consequences of cancer treatment
termination at least two years after treatment was successfully
concluded. In videotaped interviews with the child, we obtain a
pre-illness history, a description of the illness and therapy
course, and then ask in detail about all aspects of the child's
life since treatment ended. We cover essentially the same ground
in interviews with the parents, most often the mother, and also
ask how their own lives have been changed personally, socially and
occupationally, Each interview is independently rated by two
members of the research team on a number of dimensions, including
a physical impairment scale; the children's depression rating
scale; scales of the patient's pre-illness adjustment, of siblings'
current adjustment and of marital adjustment, and a rating of ego
defenses. The child is administered subtests of the Wechsler
Intelligence Scale and fills out the Piers-Harris Self-Concept
Scale. The parents complete the Family Process Scale and a
sociodemographic information form.
The subjects of this report are 25 childhood cancer patients
and their families and comprise the first half of our retrospective
sample. Inclusion criteria were only that the child be between
7 and 18 years old at the end of successful treatment for a malignant
disease. Our patients were between 9 and 21 years old when inter-
viewed. Twenty-two had finished treatment between two and three
years ago; three others had been off treatment between five and
six years. Diagnostically, the group includes all forms of cancer.
The most prevalent diagnoses were leukemia {14) and lymphomas {6).
The two cases of osteosarcoma are both leg amputees.

790
 Socioeconomically, our families are a heterogeneous group with
incomes ranging from insufficient welfare payments to annual
salaries upward of $60,000. Occupationally they are equally varied,
and include specialized professionals, blue collar workers, upper
level managers and seasonal farm laborers. Six of the families are
Hispanic, and in two of these the parents were interviewed in
Spanish. The families' domiciles reflect the location of our
hospital: they come mainly from small towns, suburban or rural
communities, of which they have been surprisingly long-term resi-
dents. The mean number of years in their current homes is almost
ten. Another indication of the families' rootedness in their
communities is that 65% have extended family members 1iving nearby.
Two-thirds of the nuclear families are intact, a proportion that
is atypical of California, if not of the nation. Although many
marriages were strained by the child's cancer, only one divorce
was directly attributed to the illness; another long-troubled
marriage was hastened to dissolution by the illness.
The survivors we have evaluated so far are probably a group
with psychologically healthier outcomes than those we have not yet
seen. Although our sole criteria for selection were the child's
age and length of time since treatment ended, we have evaluated
those who have kept their clinic appointments, who could be reached
at their address and telephone, and who were willing to spend
additional time on the day of their medical visit or to come in
especially to talk to us. We have had only one outright refusal,
from a boy with severe cardiac complications consequent to his
cancer therapy, but there are others who have so far eluded us
because of repeatedly failed clinic appointments. Although we
have as yet had no formal research contact with these children and
families, many are known to the oncology team as patients who are
experiencing difficulties.
Perhaps the simplest and most straightforward indication of
the parents' response to the end of treatment is the admitted level
of worry about the child's health. More than one-third of the group
worry much if not all the time; less than one-fifth claim to worry
only a little. Initially, the worry is specific to the withdrawal
of medication: in the words of a father, "you just hope that the
levy has been built high enough so that the water won't come over
it"; an older sister describes the end of treatment as "one of the
worst things. I thought what have they done? I thought that without
medication he wouldn't stay well." Time alone does not allay
concerns, and two of the three mothers whose children have been off
treatment five years or more worry most of the time. "It's a keg
of dynamite," explained one mother. "You have a fuse in this hand
and a match in the other, and there is a strong wind blowing. A
year later it's like you blew out the match, and the second year
you are beginning to have a normal family life again." It's fair
to say that the keg of dynamite remains there for everyone, even if

791
 the match is blown out. Even for the few who describe themselves
as infrequently or little worried, worries nearly always escalate
in anticipation of a follow-up appointment and in response to minor
symptoms, especially if these replicate the symptoms of onset.
Since the interviews were usually conducted on the day of the
medical appointment and always on hospital grounds, we probably
tapped peak concerns in our discussions with the families.
The worries about the illness are not easy for families to talk
about: half of those interviewed either actively avoid discussing
the illness or do so reluctantly, and this is true even in families
who during the active stages were open among themselves and with
others. The reluctance to talk about the illness is quite frankly
linked by some to the incomplete knowledge about the illness and
the child•s uncertain state: is he or she now ill or well, and if
well, forever well? Although the justified lack of trust in the
child 1 s cure often remains unspoken, some of the family•s adaptations
betray that lack of trust. While the worries are not openly dis-
cussed, these adaptations clearly transmit the anxiety within the
family. There are three mothers who stopped working during the time
the child was in treatment and have not returned to work in case
the child should relapse; two others have resumed work but at jobs
that give them flexibility, like substitute teaching. One mother,
whose child is scheduled for check-ups at three-month intervals,
has arranged to have interim CBC and platelet counts by a physician
in her community so that the counts will never be more than six
weeks apart, even though she has been assured by the oncologist
that the more frequent counts are unnecessary. Another mother,
three years after treatment ended, describes herself as having 11 a
shield around me. If I take something to relax me, like just one
drink, I come apart. 11 In a religious family, the daily joint prayers
include an appeal for the maintenance of the son•s remission. Three
years after treatment ended successfully, a stepmother accused her
husband of treating his son as 11 if he were dying ... Two-thirds of
the parents say they continue to treat the surviving child more
leniently than siblings. Some admit they are not only more tolerant
of difficult behaviors but overtly more affectionate. For some
children, the leniency and indulgence are echoed by their larger
social world: extended family members, peers and teachers. There
are other parents who describe themselves as rigorous and even-
handed in their expectations but feel these are violated by indulgent
friends and relatives. The families that can create for the child
an unambiguous milieu of health are rare. In varying ways, most
of the children continue to live in a twilight world of ill and
well in which the end of treatment has not yet augured cure but
only a reprieve.
Overall, the responses of the young survivors themselves are
more varied and difficult to summarize than their parents. For
many, especially those who were very young at the time of diagnosis,

792
the understanding of the gravity of the illness has come subsequent-
ly. In our interviews with the children we inquired about their
current state of health and the restrictions--if any--the cancer
continues to impose. All but four have either no physical residua
or only minor ones that can be effectively covered, and all but four
say the cancer no longer causes interference or limitations in their
daily activities. Yet this predominately positive self-classifica-
tion of their current state of health underscores for many the
dilemma of having a disease that they now understand as life-
threatening without now being or looking ill. Our oldest subject,
the 21-year-old, in recounting the history of his illness, speaks
of "the weirdness" of playing football only hours before he learned
that his latest bone marrow showed he had relapsed. A leukemia
survivor for whom that "five-letter word 'fatal' is there all the
time" described her health as "excellent" on a college application
and reflected on the contradiction of having to worry about dying
without ever feeling or, for that matter, ever having felt or
looked ill.
While we did not formally rate the children's degree of worry,
they do not describe themselves as deeply or as frequently worried
as their parents. The children accurately perceive their parents
as more worried than they themselves profess to be and there are
some who say they manage never to think about their health anymore.
No parent made that claim. Contrary to their parents' perceptions,
many of the children see their parents not as lenient, but rather
as more restrictive, anxious and controlling, with many of the
restrictions and much of the anxiety attached to the child's physi-
cal activities. Those who use physical activity to emphatically
assert their health and normalcy are resentful of parental
restrictions.
The adaptation can go in the other direction as well. There
are four who describe themselves now as physically more timid and
cautious than they used to be. "I am afraid to do things" says a
14-year-old whose two older athletic brothers call him "chicken"
and who feels excluded by his peers as well. Two adolescent girls
have adopted invalid life styles: they are obese, have dropped
out of school, are home-bound and socially reclusive. A girl who
had been high achieving and socially successful prior to her
diagnosis, aligned herself with the school 's drop-outs and outcasts
when she returned to school after treatment ended. In these cases
the parents are bewildered as to how indulgent and accepting to
be or what pressures, if any, to exert.
The older survivors deliberately maintained one area of un-
certainty. All who had radiation therapy were informed that
barrenness or sterility could be possible sequelae but none had
so far sought out conclusive information and clarified their
reproductive status.

793
 In conclusion we believe some factors seem to reduce the stress
for the child and family after treatment ends, including: (1) a
relatively smooth illness course in which the remission was never
lost; (2) a sense of support, both emotional and practical, by the
nuclear family members and/or an extended network of family and
friends during the active stages of the illness, and for the child,
steadfastness and continuity of peer relationships; (3) minimal
disruptions in the family's life style and tasks during the illness,
e.g., the occupational roles of parents, the child's attendance in
school, the place of residence; (4) congruence of expectations of
the child within the family unit and between the family and the
child's significant others; (5) the age of the child: the children
who were younger and who completed their treatment before
adolescence, had a less troubled post-treatment course.
We would like to end by citing several suggestions the
subjects made as to what would have been helpful in the post-
treatment period: {1) explicit assurance by the responsible
physician at the time treatment ends that its ending is not
arbitrary; (2) as clear-cut an indication as possible from the
physician about what to expect in terms of physical and psycho-
logical consequences of treatment's end, and some continuing
guidance about cautions, symptoms and reasonable restrictions;
(3) an opportunity to meet with other families with surviving
childhood cancer patients.
REFERENCES
1 . A. T. Meadows, D. J. Massari , J. Fergusson, J. Gordon,
P. Littman, and K. Moss, Declines in IQ scores and
cognitive dysfunctions in children with acute lympho-
cytic leukemia treated with cranial irradiation, Lancet
1015-1018 (1981).
2. H. A. Moss, E. D. Nannis, and D. G. Poplack, The effects of
prophylactic treatment of the central nervous system on
the intellectual functioning of children with acute
lymphocytic leukemia, Am J Med 71:47-52 (1981).
3. G. P. Koocher and J. E. O'Malley, "The Damocles Syndrome,"
McGraw Hill , New York (1981).

794
ANOREXIA NERVOSA AND BULIMIA: STRESS SYNDROMES?

Hans Steiner and Iris F. Litt

Department of Psychiatry and Behavioral Sciences, and
Division of Adolescent t·1edicine
Stanford University School of Medicine
Stanford, CA 94305

In the past decade, research on anorexia nervosa (AN) and

bulimia (BUL) has shown an unprecedented increase in both quality
and quantity (Steiner, 1982). However, disappointingly little is
known about their etiology. In fact, the basic question: Are
eating disorders (EDO) qualitatively different from other states
of malnutrition or not, is not yet answered and can most likely
only be answered by prospective studies of a population at risk.
There is, as yet, no direct evidence for the popular belief
that constitutional vulnerability in the individual turns dieting
into the pathological process of AN. Indirect support derives from
two lines of investigation, both of which have significant meth-
odological problems: (1) Studies of psychiatric disorders in
relatives of anorectics which show that EDO, affective disorders
and possibly alcohol abuse occur more frequently in primary
relatives of patients with EDO, suggesting a relationship between
EDO and affective disorders (Cantwell, 1977; Winokur, 1980). The
constitutional vulnerability is thought to be expressed in
deficient regulation of both affect and eating behavior.
(2) Studies of the anorectic at the malnourished and weight-
rehabilitated state. Abnormalities of the anorectic•s physiology
which persist beyond weight restoration are then thought to be
biological markers of vulnerability. Several groups have identi-
fied abnormalities of the Hypothalamic-Pituitary-Ovarian (HPO) axis
(Katz and ~Ieiner, 1981), growth hormone secretion (~Jalsh, 1982),
and the Hypothalamic-Pituitary-Adrenal (HPA) axis (Halmi, 1982;
Gerner and Gwirtsman, 1981).
795
 Abnormalities persisting after weight restoration in AN can
be said to be weight independent but their true significance cannot
be established without prospective study prior to the occurrence of
malnutrition. Bulimics are somewhat different, tending to stay
within 10-15% of their ideal weight (Steiner, 1982). Therefore,
similar abnormalities in both EDO could identify significant
deficits antedating malnutrition.
Findings from both lines of research implicate the hypothalamus
as the seat of the dysfunction (Halmi, 1982; Doerr et al., 1980}.
The hypothalamus is involved in regulating affect, feeding and
satiety behavior, releasing hormones for the HPO and HPA axis, all
of which seem disturbed in EDO patients and in patients with
affective disorders. The search for a convenient and relatively
inexpensive test to reflect this dysfunction naturally leads one to
the dexamethasone suppression test (DST), an easy measure for hypo-
thalamic dysregulation (Carroll, 1981).
With this in mind, our study sought to investigate the useful-
ness of the dexamethasone suppression test (DST). We expected to
find abnormalities of suppression in response to dexamethasone,
which would persist beyond the original state of malnutrition in
AN, and be present without weight loss in BUL. These could then
be used in future prospective studies to more narrowly define the
population at risk together with other parameters. Rather than
looking at a general teen-age population, we could then limit our
prospective study to, for instance, prepubertal members of a dance
school without manifest malnutrition, but presenting with DST
abnormalities, some attitudinal changes on standardized question-
naires, and with a positive family history of depression, alcohol
abuse, or EDO. Data gathered from such a narrowly defined cohort
might indeed shed some light on the puzzling origin of EDO.
For our study, it was essential to accurately control dietary
intake, since it is well-known that malnutrition, especially protein
calorie malnutrition, alters the activity of the HPA axis and the
response to dexamethasone (Berger et al ., 1983}. Although failure
to suppress cortisol in response to dexamethasone after weight
gain beyond 80% from ideal has been reported in 3 of 7 patients
(43%) by Gwirtsman and Gerner, 1981), no detailed information
about the caloric intake and weight status at the time of the DST
is provided. In our own study, we combined mood assessment,
control of weight and caloric intake and the DST. We aimed at a
broader evaluation of mood consisting of arousal, anxiety and de-
pression measures since we had observed the common but by no means
ubiquitous nature of depressive symptoms in our patients. Some
seemed more anxious and stressed rather than depressed.
Our hypotheses were : {l) EDO patients would show a higher
percentage of nonsuppression at stable weight and caloric intake

796
than can be expected by chance (i.e. more than 10%). (2) Per-
sistent nonsuppressors would show elevated depression scores.

~1ETHODOLOGY

Subjects
We studied 37 consecutive patients from our Eating Disorders
Clinic at Stanford. There were 29 inpatients and 8 outpatients.
The male-female ratio was 4:33. Diagnosis was established according
to DSM-III criteria and by standardized self-rating instruments
(Garner and Garfinkel, 1982). Mean age for AN was 14.3 (1.8),
mean age for bulimics was 16.9 (1.2).
Methods
DST was performed on two occasions. The first suppression
test was done within one week of admission, after correction of
dehydration and other acute medical complications. It was repeated
after weight rehabilitation above 85% of ideal weight (90% of our
patients were above 90% ideal weight at this time). On the day of
the second DST, the patient was weighed and calories from the
previous day counted. Plasma cortisol was obtained only if
(1) weight either remained stable or increased on the day of DST,
(2) caloric intake equalled the calculated daily requirements.
The DST consisted of 1 mg dexamethasone at 11 P.M. and sub-
sequent 8 A.M. and 4 P.M. plasma cortisol levels the next day.
Cortisol assay was by a radioimmunoassay. Our criterion for non-
suppression was a conservative 6 mg %.
Mood Assessment
All patients were rated by their physicians as to the presence
or absence of depressive symptoms. In addition, they completed
our standard mood self-assessment battery, consisting of the ZUNG
depression scale, the STAI-X-1 and X-2, and the Visual Analogue
Arousal Scale.

RESULTS
Our first hypothesis was clearly supported. In the malnourished
state, 14/16 (88%) of our anorectics showed nonsuppression. After
weight rehabilitation, this abnormality persisted in 10/23 (44%).
Four of nine bulimics (44%) showed nonsuppression at a stable
weight.
Our second hypothesis, however, was not supported. There was

797
no clear relationship between persistent nonsuppression and de-
pressive symptoms as measured by either clinician ratings or
self-report. Both clinician ratings and ZUNG scores resulted in
a combined false negative/positive rate of 47%, thus discriminat-
ing suppressors from nonsuppressors at a level only slightly
better than chance.
By contrast, our arousal and anxiety measures fared much
better. Arousal produced only 22% combined false positives/
negatives, and the STAI's each 36%.

DISCUSSION
Our results suggest that a significant portion of EDO patients
have persistent abnormalities in their HPA axis. Most interestingly
the percentage of disorder was very similar in both AN and BUL,
and both percentages are directly comparable to those reported by
other groups in AN (Gwirtsman and Gerner, 1981) and BUL (Gwirtsman
et a1 . , 1983). Our finding of weight independent abnorma 1iti es
of DST again seems to implicate hypothalamic centers, especially
those under noradrenergic influencs (Halmi, 1982), since deficiencies
in noradrenalin transmission are thought to be responsible for
depressive symptoms and tonic inhibition of CRF. Contrary to our
expectations, the link between nonsuppression and depressive
symptoms was very weak. Three other groups have reported such
changes in DST as a function of changes in weight without concomitant
depressive mood changes (Gerner and Gwirtsman, 1981; Berger et al .,
1983; Edelstein et al., 1983). Instead, our data suggest a link
between nonsuppression and arousal and anxiety, both of which are
commonly associated with stress. Two ether observations argue
for the consideration of stress in the etiology of nonsuppression
and EDO: (1) 80% of our total clinic population of 75 patients
had clear-cut stressful precipitants for the onset of their disease
such as separation from the family, change of school setting.
(2) Most recently, Kaye et al. (1982) reported increased opioid
activity in the CSF of EDO patients. Both opioid and HPA systems
have been shown to react quite powerfully to stressful input.
To our knowledge, this is the first report where abnormalities
of cortisol suppression in patients with EDO are reported where
weight and caloric intake were carefully monitored at the time of
the DST. Our method of controlling weight and caloric intake at
the time of the second DST was adequate but by no means perfect.
The possibility remains that some of our patients could have mis-
represented weight and caloric intake and thus produced erroneous
results. Without further concomitant studies of catabolism such
as ketones, we cannot be certain of the validity of our results.
However, the following considerations seem to support the accuracy
of our findings: Doerr (1980) has reported normalization of

798
suppression after as little as 10% weight gain. Most of our
patients were well within 10% of their ideal weight and were
actually in the final weight maintenance phase of their program.
Their average daily caloric intake was around 2,000 calories,
and their average weekly weight gain ranged from 0.8 to 1 .2 kg.
In addition, our patients were quite young and as a rule did not
exhibit extensive deceptions commonly found in older, chronic
EDO patients. Furthermore, a large number of our patients were
hospitalized. Our experienced nursing staff had every opportunity
to supervise their activities.
Finally, our study is one of several recent ones casting
some doubt on the usefulness of the DST as a specific tool for
the diagnosis of depression, as su9gested by Carroll (1981).
DST abnormalities have now been found with increasing frequency
in disorders quite unrelated to depression (e.g. Dewan, 1982),
while studies of less severely ill depressives have shown a much
lower rate of nonsuppression than originally reported by Carroll.
It is quite possible that the DST may be a very nonspecific
indicator of the severity of a variety of syndromes (e.g. anorexia,
schizophrenia, depression, panic disorders, etc.), rather than
a highly specific diagnostic tool.

REFERENCES
Berger, M., Pirke, K. M., Doerr, P., 1983, Letter to the editor,
Arch Gen Psychiatry, 40:585.
Cantwell, D., Sturzenberger, S., Burroughs, J., 1977, Anorexia
nervosa. An affective disorder? Arch Gen Psychiatry,
34:1 087.
Carroll, B. J., Feinberg, M., Greden, J. F., 1981, A specific
laboratory test for the diagnosis of melancholia.
Arch Gen Psychiatry, 38:15.
Dewan, M. J., Pandurangi .~. K., Boucher, M. L., 1982, Abnormal
dexamethasone suppression test results in chronic
schizophrenic patients. Am J Psychiatry, 139:1501.
Doerr, P., Fichter, M., Pirke, K-:-14.~ T980, Re-lationship between
weight gain and hypothalamic pituitary adrenal function
in patients with anorexia nervosa, J Steroid Biochem,
13:529.
Edelstein, C. K., Roy-Byrne, P., Fawzy, F. 1., 1983, Effects of
weig~t loss on the dexamethasone suppression test.
Am J Psychiatry, 140:338.
Garfinkei,-·P. "'E., Garner, D. M., 1982, "Anorexia Nervosa: A
Multidimensional Perspective," Brunner/Mazel,
New York.
Gerner, R. H., Gwirtsman, H. E., 1981, Abnormalities of dexa-
methasone suppression test and urinary MHPG in anorexia
nervosa, Am J Psychiatry, 138:650.

799
Gwirtsman, H. E., Roy-Byrne, P., Yaeger, J., 1983, Abnormalities
in bulimia. Am J Psychiatry, 140:559.
Gwirtsman, H. E., Gerner, R. H., 1981, Neurochemical abnormalities
in anorexia nervosa: similarities to affective disorders,
Biological Psychiatry, 16:991.
Halmi, K. A., 1981--82"", Catec-holamine metabolism in anorexia
nervosa, Int J Psychiatry in Medicine, 11 :251.
Katz, J. L., Weiner, H.,l98r,-u-The aberrant reproductive
endocrinology of anorexia nervosa, in: 11 Brain, Behavior,
and Bodily Disease, 11 H. Weiner et al., ed., Raven,
New York.
Kaye, W. H., Pickar, D., Naber, D., 1982, Cerebrospinal fluid
opioid activity in anorexia nervosa, Am J Psychiatry,
139:643.
Steiner, H., 1982, Anorexia nervosa, Pediatrics in Review, 4:123.
Walsh, B. T., 1982, Endocrine disturbances in anorexia nervosa
and depression, Psychosomatic Medicine, 44:85.
Winokur, A., March, V., Mendels, J., 1980, Primary affective
disorder in relatives of patients with anorexia nervosa,
Am J Psychiatry, 137:695.

800
THE ROLE OF PSYCHO-SOCIAL STRESSES IN BRONCHIAL ASTHMA

C.M.B. Higgs and J.E. Harvey

Respiratory Department
Bristol Royal Infirmary
Bristol BS2 8HW, U.K.

Asthma is one of the commonest causes of disability in adults,

affecting 2 - 3% of the population in the U.K. However, the
pathogenesis is not fully understood and is certainly multifactorial.
The role of beta-adrenergic receptors, mediator release, vagal
stimulation, etc. has been extensively researched but what is the
role of the psyche and individuals' reactions to stress? These
may alter the perception of asthma and lead to inappropriate use
of standard asthma therapy. We describe a technique for studying
the perception of asthma and how the perception of asthma may be
affected by psychological status, including responses to stress,
and by drugs.
We have studied the perception of asthma by analysing the
relationship b~tween subjective assessment of asthma and objective
measurement of airflow obstruction. In outpatients this is achieved
by using diary cards and peak flow meters for home recordings of
(1) self-assessment of asthma on a 10 em. visual analogue scale
(VAS) and (2) peak expiratory flow rates (PEFR).
The first question to be answered is whether the knowledge of
PEFR recordings inherent in this technique affects the subjective
assessment of asthma. We therefore developed a coded peak flow
meter (the Codaflow). This is portable, reliable and has a so far
unbroken code. We studied 12 patients, 10 male and 2 female, aged
18 to 59, with unstable asthma. They were asked to record their
VAS asthma score and PEFR three times a day for four weeks, two
weeks with a coded meter and two weeks with an uncoded meter in
randomised order.
Analysis of results involves plotting VAS asthma score against

801
PEFR and by analysis of variance comparing the linear regressions
for the coded data against the uncoded data (see figure ld). From
this analysis we can compare (1) how closely subjective assessment
of asthma is correlated to PEFR (2) the slope, i.e., awareness of
changes in PEFR. A shift in position of the regression line
indicates a change in the subjective assessment of asthma.

100
+

80
+
+ -TO
•
+
.
++
.. 0
0 0
-•
~ 60
"tt-*'>
_,p:-++o o
:::
., 50
c

t.~
c
~ ~
:;; 40 :;; 40
c +lrt> c
: 30 .w-~ 0
~ 30
>

20
08~ 20

..
0
10 \....01-~
....
I0
i\, o'b ~ -fP o
I al Ib I o 0 <f' o~.. ~~+
00 100 200 300 400 SOD 600 00 100 200 300 400 500 600
PEFR 11/•lnl PEFR 11/mlnl

l DO 50

90 45

80 40

-TO

-•
- 35

......
E
E

~ 60 =
~ 30
+
+

,..
~50 ., 25 '+ :1:
c ' + + +
~+++
,...,...+:1:..
~
E
" 20
.
~ 40 o+io .. -
~ 30 0 .. : IS oCID+o+-~
> > 00 • 0 i-
0 00
g,
t
0 0
20 I0 0

10 +
:1: • +
Ic l
00 CDMtiMI ....... ++- + Id l
100 200 SOD 600 00 I 00 200 300 400 sao 600
PEFR I 1/mln I PEFR I 1/mln I

Fig. l. Examples of the relationship between VAS asthma score

and PEFR.
+ = uncoded data, o = coded data.
(ld) is the regression line for uncoded data,
is the regression line for coded data.

802
 Table 1. The number of patients showing a significant change
( p < 0. 05).

Uncod >Cod Cod> Uncod

Correlation 8 0
Slope 5 1
Shift in Position 5 1

Fig. 1 shows examples of the wide variation in perception of

asthma with (a) being an acute perceiver, (b) being a poor perceiver
and (c) being a "threshold perceiver". This last type of perception
is unsuitable for linear analysis, leaving 10 patients suitable for
analysis.

Table 1 shows that there is some effect of knowledge of peak

flow in nearly all the patients and this effect is quite consider-
able in half of them. Interestingly the 5 patients with the
greatest difference in slope and correlation between coded and
uncoded periods had the poorer perception of asthma, p < 0.05
(Mann-Whitney U). Perception of asthma (slope) was not related
to severity of asthma, age, sex or duration of asthma.
We have looked at the reproducibility of this technique by
repeating the two weeks with the coded meter in 6 patients. In no
patient was there a significant change in correlation or in slope
and in only 1 patient was there a significant shift in position.

To summarise thus far, we have found that the relationship

between VAS asthma score and PEFR varies widely between patients
and can be used as an index of perception of asthma. Knowledge
of the trends in PEFR from day to day can influence subjective
self-assessment of asthma. This effect is more pronounced in those
with poorer perception of asthma.
We then investigated the possibility of distinguishing patients
with greater perception of their asthma from those with poorer
perception in terms of psychological indices and responses to stress.
At this point we would like to acknowledge the important contribution
of Hans Steiner.

Each of the 12 patients was asked to fill in two questionnaires,

the Asthma Symptom Checklist and the Respiratory Illness Opinion
Survey. They also underwent a stress test in which they were
required to talk for 10 minutes about any stressful situation.
They and two observers filled in 12 VAS mood scales before and after
the 10 minute speech to assess their subjective and observed degree
of arousal. The 10 minute speeches were taped, coded and sent to

803
Hans Steiner for analysis. This is based on speech fluency
disturbances and syntactical correlates of defence and coping
machanisms. We analysed reported subjective arousal, observed
arousal and paralinguistic disturbances of speech in response to
this contrived acute stress. Subjects were ranked according to
ability to detect changes in their arousal, their use of low level
defences (denial) or high level defences (rationalisation), and
magnitude of their self-reported arousal. Statistical analyses
included Kendall's tau and Mann-Whitney U.
From the questionnaires we found that the ASC panic-fear
scores correlated closely with slope, 1" = 0.64 (p < 0.05). ASC panic-
fear scores also correlated with speech fluency disturbances,
'T = 0.72 (p<O.Ol).

One patient did not complete the stress test satisfactorily.

Analysis of the stress tests showed that:
(1) Those with difficulty in identifying changes in asthma
had difficulty in identifying emotional arousal. 'T = 0.78,
( p < 0.05).
(2) Those with poorer perception of asthma and arousal
employed predominently high level defences: rationalisation and
intellectualisation. 'T = 0. 72, ( p < 0. 05) •
(3) The use of low level defences (denial) correlated with
speech fluency disturbances. "f = 0. 51, ( p < 0. 05) •
(4) Those using predominantly high level defences were most
influenced by knowledge of PEFR. U = 0 (p< 0.05).
In summary, perception of asthma is closely linked to panic-
fear symptomatology, and is linked to perception of response to
stress. The use of denial appears to provide poor protection from
arousal of any form. Rationalisation seems to increase adaptation
to chronic illness.
We shall now briefly show that the techniques we have
described can be used to study the effects of drugs on the percept-
ion of asthma. We studied a further 12 patients with unstable
asthma. They were asked to use the diary cards to record assessment
of their asthma and PEFR (using a coded meter) while on inhaled
salbutamol alone for 2 weeks, then inhaled steroid or cromoglycate
for 2 weeks, followed by 1 week's washout period then 4 weeks on
increasing doses of an oral long-acting theophylline (with blood
level monitoring), followed by 1 week's washout period then a
further 2 weeks on inhaled cromoglycate or steroid.

Again we found that ASC panic-fear scores correlated with

slope. 1" = 0.57, (p<0.05).

804
 Table 2. The effect of 3 drugs on the perception of asthma.
The number of patients showing a significant change
from the baseline (salbutamol only) period (p<0.05).

Higher Lower Upward Downward

Slope Slope Shift Shift
Inhaled steroid 0 10 1 10
Oral theophylline 1 3 6 2
Inhaled cromoglycate 0 1 0 1
ASC panic-fear scores correlated with slope: 1" = 0.57, ( p < 0. 05) •

Somewhat to our suprise there was a significant effect of

inhaled steroid in reducing perception of asthma. This was
chieved without a significant change in PEFR. There was no effect
from inhaled cromoglycate. The effect of oral theophylline was
variable but in half the patients there was an increase in asthma
score without any change or in some cases even with an improvement
in PEFR.

In conclusion we feel that there is a wide spectrum of

perception of asthma ranging from 'over perceivers' through normal
to 'under perceivers'. Both ends of the spectrum should benefit
from normalisation of their perception. It is possible that severe
under perceivers are at greater risk from sudden death from asthma.
Our results indicate that making under perceivers more aware of
their PEFR, i.e. by home peak flow monitoring, may increase their
perception. Also, some drugs may be useful in reducing perception
of asthma. But it is important that the degree of perception can
be identified because the use of drugs, hypnosis, etc. to suppress
perception could be dangerous in 'under perceivers'. Our work
shows that the ASC panic-fear scale and responses to stress are
correlated with perception of asthma.

805
STRESS AND THE ADOLESCENT'S REPRODUCTIVE SYSTEM

Iris F. Litt

Department of Pediatrics
Stanford University School of Medicine
Stanford, California

Surprisingly little is known about the role of stress in the

pathogenesis of abnormalities in reproductive function among adol-
escents. A number of factors have contributed to the dearth of
information and include the fact that menstrual irregularity is
assumed to be normal among teenagers and often not investigated
as a result of this assumption; the fact that teenagers rarely
present to physicians for evaluation of infertility; that stress
is often assumed to be part of the definition of adolescence, and
finally, because studies of stress and reproduction systematically
exclude subjects under 18 years.

Of all studies of the reproductive system of adult women,

stress has been most frequently associated with amenorrhea.
Amenorrhea may be primary or secondary. Primary amenorrhea refers
to failure of menarche, while secondary amenorrhea describes ces-
sation of menses for three or more months in an individual with
previously normal menstrual cycles. In order to understand the
pathogenesis of amenorrhea in the adolescent, one must first under-
stand normal mechanisms for initiation and continuation of menses.
Menarche is thought to be initiated at the time of decrease in
sensitivity of the hypothalamus to feedback from the pituitary
gonadotrophins.

Menarche correlates well with the peak of the weight velocity

curve which generally occurs during Tanner Stage 3-4. The rela-
tionship of menarche to weight or total body fat is well estab-
lished.

Obese females have menarche earlier than the lean, yet the
mechanisms underlying this relationship remain to be elucidated.

807
Frisch has stated that 17% of body fat is necessary for menarche
and 20% for regular cycling. (1)

Primary amenorrhea may obviously be caused by chromosomal

abnormalities or structural anomalies of the reproductive system,
as well as by any condition capable of causing secondary amenor-
rhea. (2) Maintenance of normal menstrual cycles depends upon the
interaction of the negative and positive feedback systems of the
hypothalamus-pituitary-ovary such that a mid-cycle surge of LH
stimulates ovulation. Disruption in this feedback system occurs
as a result of increased secretion of other pituitary target
organs, pregnancy, ovarian failure or administration of drugs
which either stimulate prolactin production or diminish LH/FSH.
In a review of causes of amenorrhea among adolescents by Wentz,
20-40% of patients with primary amenorrhea had chromosomal or
morphologic end-organ anomalies, while chronic illness, endocrino-
pathies and polycystic ovaries explained the remainder. (3) Among
teenage patients with secondary amenorrhea, common diagnoses
included: pituitary failure, polycystic ovary disease, psycho-
genic causes and chronic illness. A word of caution concerning
these data - most studies antedated present study techniques
(releasing factors, laparoscopy, etc.) and more than one condition
may be present in the same patient (10%).

The well-known boarding school syndrome, in which teenagers

who leave home for the first time experience secondary amenorrhea,
and the finding of amenorrhea as an integral part of the syndrome
of anorexia nervosa, suggest that stress may also precipitate this
phenomenon. In both of these conditions, the coincidence of weight
loss complicates the picture, as the latter may, as we have seen,
be responsible for amenorrhea. In an attempt to disaggregate these
often related phenomenae, one looks to the literature for situa-
tions where stress occurs in the absence of weight loss. The
report of a single major traumatic event resulting in amenorrhea
is helpful in this regard. Loeser, in Lancet in 1943 reported
four cases of endometrial biopsy in women who experienced amenor-
rhea following a bombing in London. (4) In each case, the biopsy
showed arrest of the endometrium at the stage corresponding to the
day of the emotional shock, suggesting an immediate disruption of
release of pituitary gonadotrophic hormones.

Although most cases of amenorrhea in women incarcerated in

concentration camps occurred in those who were also starved, a
number of observations suggest that stress alone may have been
responsible. Sydenham's report of a 61% rate of amenorrhea among
324 women interned in Hong Kong during the 2nd World War concluded
that most developed it before the malnutrition of camp life. (5)
A lower overall percentage of amenorrhea was reported by Bass in
a study of the Theresienstadt concentration camp, but significant
differences in the rate emerged as a function of the degree of

808
stress. (6) In Group A, for example, composed of 3,000 women,
housed 70-80 per barrack, sleeping in three-tiered beds and ex-
periencing the constant threat of extermination, the rate of
amenorrhea was 54%. In Group B, composed of 120 employees living
with their families in small private rooms, on the same diet as
Group A women, but unthreatened by extermination, the amenorrhea
rate was only 25%. In 60% of all amenorrheic women, menses stopped
immediately after internment, before a change in nutritional
status; 25% became amenorrheic after one month, 2.5% after 3 months
and 2.5% every month thereafter. Most recovered menses within 6
months to a year after release, although malnutrition was often
still present. The only risk factor identified to predispose to
amenorrhea was age, which correlated inversely with the risk of
amenorrhea. Extensive ovarian endometrial and testicular atrophy
has been described at autopsy among prisoners indicted and later
executed for severe crimes by Stieve. (7) A review of the pre-
valence of amenorrhea under differing conditions of social dis-
ruption demonstrates that the rate of amenorrhea clearly increases
with increasing conditions of stress, regardless of nutritional
status.

While the presence of malnutrition in amenorrhea patients

with anorexia nervosa (A.N.) has precluded the possibility of a
clear delinieation of the role of stress in its pathogenesis, it
has been noted that even in A.N., amenorrhea may antedate sig-
nificant weight loss in one-third and one-third to one-half of
adolescents fail to resume menses when premorbid weight is re-
gained. (8) Recent studies demonstrating menstrual abnormalities
in bulimics without significant weight loss also lend credence to
the important contribution of emotional or hypothalamic factors
in its pathogenesis. (9) Moreover, a study of patients with poly-
cystic ovaries (Stein-Leventhal Syndrome), a syndrome characterized
by oligo-or amenorrhea and a distinctive pattern of endocrinologic
abnormalities (increased LH, normal FSH, and often increased
testosterone), showed an increased incidence of significant life
events among those suffering from the condition. (10) This obser-
vation may have particular significance, for adolescents as poly-
cystic ovaries are now being diagnosed with increasing frequency
among teenagers with irregular or absent menses. (2)

The association between stress and menstrual dysfunction has

been established by epidemiologic studies and the few clinical
comparative studies just reviewed. All of the limited evidence
now available suggests that humans do not respond uniformly to
stressors and that their responses are affected by a variety of
biological and psychosocial mediators. Different physiological
systems may respond differently to the same stimuli or may be
differentially sensitive to particular stressors. The system that
has been most extensively studied is that of the hypothalamic-
pituitary-adrenal axis. Extrapolation from data derived from

809
those studies suggests a hypothetical model for the potential
impact of stress on the reproductive system.

Stress is presumed to be telegraphed to the adrenal medulla

directly through neurons which stimulate the secretion of a number
of hormones, including epinephrine, norepinephrine and dopamine.
The latter two have been shown to exert a profound effect on the
reproductive system. Norepinephrine has a generally stimulating
effect on gonadotrophin production, while dopamine has the effect
of inhibiting production of these hormones and ovulation, as a
result. In addition, there is recent evidence to suggest that
certain neuroactive peptides including enkephalin are released by
the adrenal medulla at times of stress. (11) Other earlier research
has demonstrated a gonadotrophin-suppressin g effect of opioids,
suggesting still another potential mechanism by which stess-
induced anovulation may be mediated. This possibility is of
particular interest to me because of our early observation that
teenaged females who were using opioids, specifically heroin and
were not malnourished, as was the case with adult heroin addicts,
had, nonetheless, a much higher incidence of amenorrhea than did
adults. (12)

The known association of prolactin elevation with stress on

the one hand and amenorrhea on the other, must also'be explained.
These observations may not necessarily represent a cause and effect
relationship, but alternatively may both be the sequelae of stress,
mediated by a third factor. An additional observation of relevance
to this discussion of the role of stress in reproductive dysfunc-
tion is that of elevated prolactin levels in women with severe
premenstrual tension. Of interest was the finding that the pre-
menstrual tension was alleviated in 80-90% of patients following
suppression of prolactin.

Even less is known about the role of stress on the repro-

ductive system of the male. Lower primate data suggest that one
form of stress, that of crowding, may exert an inhibiting effect
on reproduction in the male adolescent, in that increasing the
number of adult males in a group modifies increases of testosterone
in pubertal males. (13)

Much work obviously remains in order to examine these findings

and hypotheses in humans and particularly in the more vulnerable
adolescent. The role of stress in initiating reproductive dys-
function is clear - the mechanisms of action of this association
are not. In evaluating an adolescent with amenorrhea it is
imperative that all factors potentially eventuating in amenorrhea
be sought, as psychological, physiologic and social stressors or
mediators may co-exist and the finding of one factor does not
preclude the presence of others.

810
1. R.E. Frisch and R. Revelle, Height and weight at menarche and
a hypothesis of critical body weights and adolescent events,
Science 169;397-99, 1970.
2. I.F. Litt, Menstrual problems in adolescence, Pediatrics in
Review, 4:203-12, 1983.
3. A.C. Wentz and G.S. Jones, Prognosis in primary amenorrhea,
Fertil Sterility 29:614, 1978.
4. A.A. Loeser, Effect of emotional shock on hormone release and
endometrial development, The Lancet Vol. 1:518-19, 1943.
5. A. Sydenham, Amenorrhea at Stanley Camp, Hong Kong during
internment, Brit Med J 2:159, 1946.
6. F. Bass, L~amenorrhae au camp de concentration de Terezin,
Gynaecologia 123:211, 1947.
7. F.L~Drew, The epidemiology of secondary amenorrhea J Chron Dis
14:396, 1961.
8. H. Steiner, Anorexia Nervosa, Pediatrics in Review, 4:123, 1982.
9. J.E. Mitchell, J.P.Bantle, Metabolic and endocrine investigations
in women of normal weight with the bulimia syndrome, Biol
Psychiatry, 18:355, 1983.
10. R.A.Lobe,-L.R. Granger, W.L. Paul, et al., Psychological stress
and increases in urinary norepinephrine metabolites, platelet
serotonin, and adrenal androgens in women with polycystic ovary
syndrome, Amer J Gynecol 145:496-503, 1983.
11. M. Schultzberg, J.M. Lundberg, T. Nokfelt, et al., Enkephalin-
like immunoreactivity in gland cells and nerve terminals of
the adrenal medulla, Neuroscience 3:1169-86, 1978.
12. I.F. Litt, S.K. Schonberg, Medical complications of drug abuse
in adolescents, Medical Clinics N.A. 59:1445-1452, Nov. 1975.
13. R.M. Rose, I. Bernstein, T.P. Gordon, et al., Changes in testos-
terone and behavior during adolescence in the male rhesus
monkey, Psychosomatic Med 40:60-70, 1978.

811
ANTI-ANXIETY DRUGS IN CARDIAC DISORDERS

David Wheatley
Psychopharmacology Research Group

Twickenham, TW2 SAX, U.K.

Although stress is an inescapable component of life and its

adverse effects in some individuals at some times are well
appreciated, these effects have never been systematically investi-
gated from a clinical point of view, In particular, the effects
of stress on organic illnesses, as for example, cardiovascular
disorders, are not fully appreciated.

Stress may take many forms, both pleasurable and obnoxious,

but not all individuals will react to such stresses in an adverse
manner. The release of catecholamines is one of the main effects
of stressful situations and these can produce profound effects on
the cardiovascular system, that may be particularly accentuated in
those already suffering from a cardiac lesion. The effects on
the psyche are mainly to produce symptoms of an anxiety state and
so the use of anti-anxiety drugs is very relevant to the problem
of stress and the heart.

INVESTIGATING STRESS FACTORS

We have evolved a system for uncovering stress factors by

means of patient and observer ratings, using standardised
questionnaires.

Social Stress

Under this heading the following are considered:

813
 SOCIAL STRESSES

L Habits 4. Harriage (etc)

2. Work 5. Relations c. Others

3. Recreations 6. Life Events

Of social habits, the most relevant are excessive consumption

of alcohol and smoking, both of which are potent risk factors in
coronary heart disease. However, many individuals use alcohol and
nicotine as tranquillizers, and if we are to deprive them of their
"natural" tranquillizers, should we not consider replacement with
a chemical substitute, provided that it is harmless? Frustration
at work may take a number of different forms: an excessive work-
load, standards of efficiency that may be difficult to attain and
maintain, abrasive relationships with others, particularly those
in authority above the individual. The regular use of anti-anxiety
drugs under these circumstances, although to be deprecated on
medical grounds, may nevertheless constitute the only means by
which individuals are enabled to carry on with their occupations.

Recreations may involve both mental and physical strain, such

as the excitement of watching a football match and the danger
involved in sports such as hang-gliding, etc. Drugs have little
part to play in this aspect of stress, and the patient would be
better advised to exchange the stressful recreation for a more
sedentary one such as reading, music, model making, etc. Relation-
ships with a consort, when they are unhappy, may be a potent stress
factor and if the concubinal situation is one that cannot be altered,
then once again judicious use of anti-anxiety drugs may be
indicated.

Relationships with other family members and friends come

under a similar category. Finally, life events can be either
adverse (i.e. death or illness of a loved one, financial troubles,
domestic disasters such as burst waterpipes and so on) or bene-
ficial (i.e. a happy family event such as a marriage, acquisition
of prized objects such as motor cars, attainment of some life goal
such as promotion at work, etc.). Such events, of both kinds, may
cause excitement and unrest during the daytime with consequent
difficulty in concentrating on work etc., and interfere with sleep
at night. Short-term use of anti-anxiety drugs must be fully
justified in such situations.

814
Psychiatric Stress

The affective disorders, anxiety states and depressive illness

are potent stress factors and will warrant appropriate treatment
with psychotropic drugs. Psychotic illness also often creates both
personal stress and stress on others who may be involved in looking
after the patient. Again, treatment is with appropriate psychotropic
drugs.

Sexual Stress

Sexual stresses are perhaps the most overlooked and yet the
most ubiquitous of mental stresses that afflict mankind. In
investigating sexual stress, the following factors are considered:

SEXUAL STRESSES

1. Mental 3. Female

2. Physical 4. Psychological

Under the heading of mental aspects must be considered the

degree of enjoyment arising from sexual intercourse and the quality
of the patient's libido. Physical aspects are concerned with any
difficulties that may arise in the actual performance of the sex
act such as: poor erection, premature ejaculation, frequency of
intercourse, partner's attitude (i.e. too passive or too demanding),
partner's technique and problems with contraceptives (i.e. distaste
for mechanical methods, worry over oral c-c's). Peculiar to women
are dysparunia and fear of possible pregnancy which again can be
stress factors. Finally, psychological stresses are produced when
illicit sex relationships are involved, guilt feelings, fear of VD,
worry about "abnormal" practices, feelings of sexual inadequacy and
jealousy concerning one's partner.

These aspects must be fully investigated and anti-anxiety drugs

may have a part to play as an adjunct to sex therapy.

Stress and Sleep

Lack of sleep is a potent stress factor and needs to be fully

investigated considering both the aetiology and type of insomnia,
together with daytime effects on normal activities. Depending on
the type of insomnia present and the causative factors, an
appropriate hypnotic drug may be indicated, For example, pro-

815
longed latency (time taken to fall asleep) may be best treated by
a short-acting benzodiazepine, whilst awakenings during the night
and early morning awakening will need longer acting preparations.
Daytime activities must be considered in this context, since some
residual daytime tranquillization may be beneficial for some
individuals but disastrous for others.

Geriatric Stress

The elderly are particularly vulnerable to many stresses and

not least of these is the sense of isolation from other members of
the community, occasioned not only by their social circumstances
but by their failing senses. Thus, for example, failing eyesight
and hearing can effectively isolate them from normal contacts with
other people and even with the media via. newspapers, TV and radio.
Tranquillizing drugs may be particularly useful in the elderly,
bearing in mind the altered pharmacokinetic problems, since daytime
sedation may not be adverse for many elderly people who may be
confined to their homes.

Stresses Peculiar to Women

Just as the incidence of affective disorders is much greater

in women than in men, so is there no period of a woman's life that
may not be affected by psychiatric symptoms associated with her
reproductive functions. The psychiatric disorders associated with
menstruation and pregnancy are wellknown and even when the mani-
festations of her reproductive capacity begin to reach their
termination in the menopause, rather than escaping from the psycho-
logical hazards that may accompany them, a woman becomes even more
vulnerable to the development of such symptoms, both during and
after the climacteric. Standardised questionnaires for premenstrual
and menopausal symptomatologies are used to elicit these factors
and in many cases psychotropic drugs may be indicated in treatment.

CORONARY HEART DISEASE

Our group has undertaken a number of double-blind clinical

trials to compare anti-anxiety drugs to placebo, added to standard
cardiac medication in patients suffering from coronary heart disease.
These have yielded conflicting results as illustrated from the
results of a trial to compare clorazepate to placebo in patients
who had suffered their first myocardial infarction (Fig.l)l and
diazepam to placebo in patients being treated with the calcium
blocking drug, verapamil (Fig.2)2,

816
 25
[ill Clorazepat e

~ 20
0 Placebo
<1)
......
..0
(.!. SEM)
Cll
+J 15
'-!-<
0
I-<
<1)
,.0 10
3
z

2 Wks 4 Wks 6 '~s 8 Wks 10 Wks 12 \•Tks

Fig. 1. Mean GTN requiremen ts per 2-week period,
clorazepat e compared to placebo.

~ Verapamil + Diazepam D Verapamil + Placebo

Anginal GTN
1.5 Attacks Requiremen ts

.5

Control 3-4 h'ks 7-8 Wks Control 3-4 Wks 7-8 Wks
Fig. 2. Mean daily anginal attack rate and GTN requiremen ts,
verapamil + diazepam compared to verapamil + placebo.

817
 In the first study statistically significant differences were
demonstrated in both the anginal attack rate and requirements of
glyceryl trinitrate (GTN) to relieve attacks (illustrated), whereas
in the second study, no significant differences were demonstrated
between the anti-anxiety drug, diazepam, and placebo, Differences
in the patient population in the two studies may account for these
anomalous results and it is also possible that the dose of diazepam
used in the second study was too low (6 mg. daily).

One of the most important aspects of treating the anxiety

factor in CHD is the prevention of further attacks of myocardial
infarction. Combining the results from all our trials of anti-
anxiety drugs in CHD, yields a total of 81 treated with such drugs
as compared to 77 treated with placebo, There were no cases of
myocardial infarction in the patients treated with the anti-anxiety
drugs but five such cases in those treated with placebo (p<0.05).
This therefore is an aspect of treatment with anti-anxiety drugs
that needs further elucidation by conducting further clinical trials
on larger numbers of cases and over longer periods of time. The
development of new non-benzodiazepine anxiolytics such as buspirone,
with a reduced incidence of daytime sedation and dependence-potential,
may be of particular interest in this context.

REFERENCES

1. D.Wheatley, Treatment of the anxiety factor, in: "Stress and

the Heart" (2nd edition), D.Wheatley, ed. ,Raven Press,
New York. 39-62 (1981).
2. D.Wheatley. The value of anti-anxiety drugs in the management
of cardiac disease, in: Acta Med Scand (Suppl) 660:
219-230 (1982).

ACKNOWLEDGEMENTS

Figure 1 is reproduced from "Stress and the Heart", 2nd edition,

by kind permission of the publishers, Raven Press of New York.
Figure 2 is reproduced from "Acta Medica Scandinavica" by kind
permission of the Editor of that journal.

818
PSYCHOGENIC HYPERTENSION: EXPERIMENTAL REVIEW

Richard Friedman and James A. McCaughran, Jr.

Department of Psychiatry and Behavioral Science

State University of New York
Stony Brook, New York 11794 U.S.A.

Exposure to environmental situations which are perceived as

stressful can result in profound cardiovascular alterations. The
relationship between such exposure and acute elevations in blood
pressure and heart rate have been well documented. More
problematic is the establishment of the relationship between
repeated and/or prolonged exposure to stress and the pathogenesis
of chronic hypertension. The purpose of this manuscript is the
presentation of a series of fundamental experimental observations
concerning this relationship. The authors are well aware of the
difficulties associated with extrapolating data derived from
laboratory studies in animals to the human condition. However,
clinically oriented mental health professionals interested in the
relationship between psychological functioning and hypertension
should, at least in the authors' opinion, be familiar with basic
laboratory results.

Although experimental efforts attempting to implicate stress

as an etiological or exacerbational agent in hypertension have
been reported for decades, the results have been equivocal (1).
An important research opportunity emerged in the 1960's when
scientists began breeding rats genetically predisposed to develop
hypertension. Although none of these rat models were developed to
specifically study the relationship between psychological stress
and high blood pressure, all were eventually exploited to this
end. One such model, the Dahl rat model is particularly well
suited for such studies. Dahl and his colleagues were primarily
concerned with the relationship between excess salt ingestion and
hypertension. Believing that genetic factors play a role, Dahl
selectively bred male and female rats which evinced the most
dramatic salt-induced blood pressure elevations and also
819
selectively bred those rats which exhibited no blood pressure rise
in response to a high salt diet.

After several generations of selective breeding according to

this phenotype, two distinct lines had been produced (2). One,
the sensitive or OS line, rapidly and predictably developed severe
hypertension in response to excess salt ingestion. The other, the
resistant or DR line, exhibited no blood pressure elevations when
fed the same diet. It became apparent that the genetically
determined sensitivity of the OS rats was not restricted to salt.
Indeed, it appears as though OS rats have a generalized
sensitivity to a variety of putative hypertensinogenic stimuli
such as renal artery constriction and adrenal regeneration (see 3
for a review). The DR rats on the other hand are generally
resistant to those insults. Because of their unique sensitivity,
Dahl and his co-workers exposed OS rats to stressful environmental
conditions including repeated electrical shocks, crowding and cold
temperatures. None of these situations resulted in chronic blood
pressure elevations in the OS rats leading Dahl to conclude that
stress is probably unimportant in the pathogenesis of hypertension
( 4) •

This conclusion may have been unwarranted. Several years

later, a series of studies on OS rats established an important
principle concerning the relationship between stress and
experimental hypertension. Specifically, the exposed organisms
interaction with the stress importantly determines the magnitude
of blood pressure change. In the original Dahl stress study, rats
were passively exposed to stress. Their behaviors had no effect
on the presentation of the stressful stimuli. In a subsequent
study, Friedman and Dahl (5) constructed five experimental groups
to explore the issue of control. In one condition, (conflict),
rats were exposed to a food-shock conflict six hours daily for
months which resulted in food deprivation and frequent electrical
shocks. Specifically, they were required to press a lever for
food according to a random reinforcement schedule. The same
lever-press response also resulted in the electrification of the
grid floor of the experimental chamber on an independent random
schedule. Lever pressing was the only means by which rats could
obtain food throughout the entire study. In a second group,
(yoked food and shock), rats received the identical pattern of
food and shock as the conflict-exposed rats. However, in the
conflict condition the reinforcement and the punishment were
contingent on subjects' behaviors while the yoked rats had no
contingent control. For these subjects lever pressing was without
effect. A third group consisted of rats that received yoked shock
according to the behavior of conflict rats but who received no
food deprivation. These rats could lever press for food without
contingent shocks. A fourth group was yoked for food
reinforcement to the conflict rats but received no electrical

820
 TABLE 1

SYSTOLIC BLOOD PRESSURE (rnm/Hg)

GROUP DS DR

Conflict 170 121

Yoked food and shock 157 118
Yoked shock 145 127
Yoked food 140 115
Control 144 119

shocks. A fifth group consisted of free feeding unshocked

controls. The left side column of Table 1 presents the mean
systolic blood pressures for the five groups after 12 weeks
exposure. This representative week indicates the most significant
result of the study. Conflict exposed DS rats exhibited blood
pressures significantly higher than any other group. Particular
mention of the difference between the conflict and the yoked food
and shock group is warranted. The two groups were exposed to
exactly the same amount and pattern of food and shock presentation
and differed only on the requirement to make responses which would
result in food and/or shock. This psychological demand; i.e.,
having to make the decision to respond or not to respond, resulted
in more severe hypertension. Clearly, the yoked food and shock
group, passively suffering both food deprivation and frequent
electrical shocks could be described as experiencing·a stressful
environment. However, the way in which the organism is required
to interact with the environment is an important consideration.

Another fundamental point was demonstrated in a subsequent

study. In the above experiment, only DS rats were used as
subjects. Friedman and his colleagues repeated the above study
using both DS and DR rats (6). The right hand column in Table 1
presents mean systolic blood pressures for DR rats after 12 weeks
exposure to the five treatment protocols. Obviously exposure to
these environments resulted in no elevations in blood pressure.
The ability of stress to induce hypertension, therefore, depends
to a large extent on the genetic predisposition of the exposed
organism as well as to the nature of the stress. It is also worth
discussing the behavioral response of DS and DR rats exposed to
these stress conditions. During the daily conflict exposure rats
of both lines continually approached, then withdrew, from the
lever. These rats repeatedly chewed and clawed at the floor and
walls of the chamber and were extremely reactive to handling.
Careful observation failed to detect systematic differences in the
behavior of DS and DR rats exposed to conflict. The actual number
of lever presses per day was also similar in OS and DR rats.
Hence, despite similar behavioral reactivity to this stressful

821
environment, the cardiovascular reactions, genetically determined,
were quite different. At least in these rats then, and perhaps in
humans as well, observation of stress-induced behavior change is
not always correlated with stress-induced physiological change.
Obviously, the genetic predisposition to hypertension cannot
be precisely determined in humans. In order to maximize clinical
success, it would be useful if individuals likely to develop
hypertension exhibited premorbid characteristics which
distinguished them from individuals unlikely to develop
hypertension. Animal models have been used to suggest such
distinguishing characteristics. An example using the Dahl rats is
illustrative. Normotensive DS and DR rats were implanted with
cardiac electrodes and heart rate was monitored in the home cage
for several days. During this baseline period DS rats had lower
heart rates than DR rats. However, as indicated in Table 2 upon
exposure to immobilization stress, the increase in heart rate was
greater and more prolonged in DS compared to DR rats (7). Greater
stress-induced tachycardia in animals genetically predisposed to
hypertension vis-a-vis appropriate controls has been reported for
other animal models as well (e.g. 8). Furthermore, there is
evidence indicating that in humans, the normotensive offspring of
hypertensive parents, upon exposure to psychological stress,
evince greater acute increases in heart rate than do the
normotensive offspring of normotensive parents (9). The specific
point to emphasize here is that relatively great stress-induced
tachycardia may be a distinguishing characteristic of organisms at
risk for hypertension. The more general point is that a
characteristic, or constellation of characteristics, is being
experimentally pursued which might allow identification of
individuals most at risk for hypertension and most likely to
profit from early intervention.

Finally, there is a considerable body of experimental work

examining central and peripheral neural mechanisms which might
account for the inter-relationships among stress, genetic
predisposition and hypertension. It is not possible to even

TABLE 2

BASAL HEART RATES AND STRESS-INDUCED HEART RATE CHANGES (b.p.m.)

DS DR

Basal (Home cage) 356 379

Immobilization: 1 minute* 149 115
Immobilization: 5 minutes 140 113
Immobilization: 15 minutes 113 91
Immobilization: 30 minutes 99 75

*Immobilization readings are increases from basal measurement

822
briefly review this work here although extensive reviews are
available (10). There are, however, consistent reports indicating
that ~nimals genetically predisposed to hypertension exhibit
hyper-reactivity of the sympatho-adrenal medullary system (11),
and have central nervous system receptor site characteristics
which distinguish them from animals without such genetic
predisposition. For example, in the authors' laboratory, we have
demonstrated that OS rats have a greater number of cholinergic
receptor binding sites than DR rats in the cortex, hypothalamus
and medulla (12). Furthermore, stimulation of central cholinergic
sites results in a greater acute pressor response in OS than DR
rats (13) while chronic cholinergic blockade attenuates the
development of salt-induced hypertension in OS rats (14). The
density of alpha 1- and alpha 2- adrenoceptors has also been
investigated in OS and DR rats. OS rats maintained on a high salt
diet have fewer alpha 1- and alpha 2- receptors in the
hypothalamus and medulla than similarly treated DR rats (15).
These results, combined with the results of others, have led to
the speculation that in the Dahl rat model the central cholinergic
system overmodulates the adrenergic system and results in greater
sympathetic discharge. The generalizability of this speculation
to other animal models and to human essential hypertension awaits
further study.

In summary, we have presented several experimental points

which are relevant for those interested in the relationship
between psychological factors and hypertension: (1) The
cardiovascular pathogenecity of stress depends at least in part on
the nature of the exposed organism's interaction with the stress~
(2) The ability of chronic stress to cause chronic blood pressure
elevations depends to a great extent on the genetic predisposition
of the exposed organism~ (3) Premorbid characteristics such as
relatively great acute stress-induced heart rate increments may
distinguish organisms sensitive to hypertension from organisms
resistant to hypertension and; (4) Hyper-reactivity of the
sympatho-adrenal medullary system as well as central adrenergic
and cholinergic receptor site differences may be involved in
genetic predisposition to hypertension.

REFERENCES

1. R. Friedman, Experimental Psychogenic Hypertension, in:

"Stress and the Heart," o. Wheatley, ed., Raven Press, New
York, ( 1981).
2. L. K. Dahl, M. Heine, and L. J. Tassinari, Effects of chronic
excess salt ingestion: Evidence that genetic factors play
an important role in susceptibility to experimental
hypertension, Journal of Experimental Medicine. 115:1173-6
(1962).

823
 3. J, p, Rapp, Dahl salt-susceptible and salt-resistant rats. A
review, Hvoertension. 4:753-763 (1982).
4. L. K. Dahl, K. D. Knudsen, M. Heine, and G. Leitl,
Hypertension and stress, Nature. 219:735-736 (1968).
5. R. Friedman, and L. K. Dahl, ~he effect of chronic conflict
on the blood pressure of rats with a genetic susceptibility
to experimental hypertension, Psychosomatic Medicin~.
37:402-416 (1975).
6. R. Friedman, and J. Iwai, Genetic predisposition and
stress-induced hypertension, Science. 193:161-162 (1976).
7. R. Friedman, M. McCann, R. Leder, and J. Iwai, Genetic
predisposition to hypertension and stress-induced
alterations in heart rate, Behavioral and Neural Biology.
35:426-431 (1982).
8. M. Hallback, and B. Folkow, Cardiovascular response to acute
mental "stress" in spontaneously hypertensive rats, Acta
Physiolo~ica Scandanavica. 90:684-698 (1974).
9. B. Falkner, B. Onesti, L. T. Angelakos, M. Fernandes, and
c. Langman, Cardiovascular response to mental stress in
normal adolescents with hypertensive parents, Hypertension.
1:23-30 (1979).
10. M. J, Brody, J, R. Haywood, and K. B. Touw, Neural mechanisms
in hypertension, Annual Review of Physiology. 42:441-453
( 1980) •
11. R. McCarty, and I. J, Kopin, Sympatho-adrenal medullary
activity and behavior during exposure to footshock stress:
A comparison of seven rat strains, Physiology and Behavior.
21:567-572 (1978).
12. E. Edwards, J, A. McCaughran, Jr., R. Friedman, w. McNally,
and N. Schechter, Cholinergic receptor site binding, choline
acetyltransferase, and acetylcholinesterase activity in the
forebrain and brainstem of the Dahl rat model of essential
hypertension, Clinical and ExEerimental Hypertension. In
Press. (1983).
13. J, A. McCaughran, Jr., D. Murphy, N. Schechter, and
R. Friedman, Participation of the central choLinergic system
in blood pressure regulation in the Dahl rat model of
essential hypertension, Journal of Cardiovascular
Pharmacolo~. In press, (1983).
14. J, A. McCaughran, Jr., c. Manetta, s. Gionet, and
R. Friedman, Attenuation of salt-induced hypertension in the
Dahl rat model by chronic cholinergic blockade. Submitted
for editorial review.
15. J, A. McCaughran, Jr., E. Edwards, R. Friedman, and
N. Schechter, Adrenoceptor binding in the forebrain and
brainstem of the Dahl rat model of essential hypertension.
Submitted for editorial review.

The authors wish to thank Joyce Manolakes for her expert

assistance in the preparation of this manuscript.

824
THE RELAXATION RESPONSE AND REDUCED NOREPINEPHRINE REACTIVITY

Richard Friedman*•** and Herbert Benson*

*Division of Behavioral Medicine
Department of Medicine
Beth Israel Hospital
Harvard Medical School
Boston, Mass, 02215
**Department of Psychiatry and Behavioral Sciences
State University of New York
Stony Brook, N.Y. 11793

INTRODUCTION
There is growing appreciation both within and outside of the
medical community that psychological stress is importantly involved
in a variety of disease processes. The interaction of psychological
factors and pathophysiology has been most vigorously pursued in
relationship to the cardiovascular diseases. Because of this
interest, and the accumulating evidence substantiating the link
between psychology and cardiovascular diseases, mental health
professionals are being increasingly asked to therapeutically
intervene in patients with manifest disease or in patients
deemed at high risk for such diseases. It is important to
acknowledge that there is no implication that patients so
referred are in need of traditional psychiatric or psychologic
therapy. Rather, there is an implicit assumption that repeated
or prolonged exposure to environmental demands, which are
perceived as stressful, results in reactions that have an
adverse affect on the cardiovascular system. It is reasonable
to suggest that some effort at reducing such exposure either
by altering environmental characteristics, altering the individual's
perception, or altering reactions is worthwhile and justified.
Attending to these considerations. therapeutic interventions have
been specifically devised .. These interventions usually involve
825
combinations of four strategies: 1) attempts to modify dietary
habits, caloric reductions generally, with special efforts to reduce
lipids and salt; 2) attempts to modify exercise; 3) attempts to
train individuals to relax and or meditate and; 4) attempts at
cognitive behavioral therapy. Although all four can be construed
as psychological or behavioral interventions, the latter two are
more traditionally considered the provience of mental health
profession a1s.
The authors are currently involved in such an intervention
program for patients with benign essential hypertension.
The purpose of this presentation is a description of one
particular procedure; elicitation of the relaxation response (1).
This procedure is the foundation of the authors• intervention
program and has been adopted into many other stress management
programs. Further, a study which suggests a mechanism by which
regular elicitation of the relaxation response may have salubrious
effects on diseases such as hypertension will be described.

THE RELAXATION RESPONSE

The relaxation response is defined by a set of integrated,
physiological changes that are elicited when the subject; 1)
assumes a relaxed position in a quiet environment; 2) closes
his or her eyes; 3) engages in a repetitive mental action and;
4) passively ignores distracting thoughts (2). The relaxation
response is believed to be the counterpart of the emergency reaction
or the fight or flight response, originally described by Canon (3),
'which is associated with increased sympathetic nervous system
activity. Both the relaxation response and the fight or flight
response are probably mediated through activation of hypothalamic
areas of the brain (4).
Elicitation of the relaxation response has been associated
with physiological changes that include; decreased oxygen con-
sumption; decreased heart rate; decreased arterial blood pressure;
decreased respiratory rate; and decreased arterial blood lactate.
Furthermore, there are slight increases in skeletal muscle blood
flow (2,5,6). The changes just described are different from those
reported during sleep or during quiet sitting (7). Although the
physiological changes associated with elicitation of the relaxation
response are consistent with decreased sympathetic nervous system
activity, direct measurement of plasma norepinephrine levels have
been reported to be either unchanged (8) or variable (9}. The
following study was intended to elucidate the mechanism by which

826
elicitation of the relaxation response may result in the above
described physiological changes.
EXPERIMENTAL PROTOCOL
In this study, the late John Hoffman, Herbert Benson, and their
colleagues examined the sympathetic nervous system reactivity of two
groups of normotensive subjects (10). In the experimental group
(n=lO), subjects were instructed to practice a technique which
elicits the relaxation response. In the control group (n=9),
subjects were instructed to practice the control technique of
sitting quietly. The primary distinguishing characteristic of the
two instructional sets were the use of a repetitive word or phrase
and the adoption of a passive attitude in the experimental group.
Subjects in both groups were instructed to practice for 20 minutes
twice each day for 30 days. Each subject was asked to record their
compliance in a diary.
Biochemical and cardiovascular reactivity was tested prior to
practicing the techniques (Session 1) and following the 30 day
trial (Session 2). In both cases, the procedures were identical.
On the evening prior to testing subjects were admitted to the
Clinical Research Center of Boston•s Beth Israel Hospital.
During the evening, maximum hand grip tension levels were
established. Subjects were awakened the next morning and an intra-
venous catheter was inserted into the antecubital vein of the non-
dominant arm 30 minutes prior to testing. Sequential blood samples
were drawn for plasma norepinephrine (NE) analysis (11) at five
different times. These times were labelled as minus 10 minutes,
zero, plus 5 minutes, plus 10 minutes and plus 15 minutes. When
the first two samples were taken, the subjects were supine. The
subjects then stood up and after five minutes of standing, the plus
five blood sample was drawn. For the next five minutes, the
subjects remained standing and also gripped at a tension
corresponding to 30% of the previously established maximum. At
the end of this five minute period, the plus ten sample was
drawn. In the subsequent five minutes subjects remained standing
and gripped at 100% of the previously established maximum grip
tension. The plus fifteen sample was then drawn. Prior to each
blood drawing heart rate (HR) was measured by palpation of the
radial artery and blood pressure (BP) was measured by auscultation.
Following session 2, six subjects from the control group were
asked to practice the experimental procedure for 30 days and return
for a third testing session conducted as described above. Subjects
in the experimental group were not asked to practice the control
technique since it has been our experience that once the relaxation
response has been learned, it is difficult to sit quietly and not
elicit it.

827
EXPERIMENTAL RESULTS
All of the relevant experimental results are presented in
Table 1.
Table 1. Mean blood pressure, heart rate, and plasma norepinephrine
response to graded levels of stresses at minutes; -10 (supine),
0 (supine), +5 (standing), +10 (standing+ 30 percent hand grip),
and +15 (standing and 100 percent hand grip).
Control Experimental Control-crossover
grou~ grou~ grou~

Min Ses 1 Ses 2 Ses 1 Ses 2 Ses 2 Ses 3

Systolic Blood Pressure (rrrnHg)
-10 106 106 120 108 106 101
0 106 106 117 108 106 103
+5 110 112 111 110 112 106
+10 118 122 126 119 122 115
+15 138 134 133 131 134 136
Heart Rate (beats/min)
-10 60 64 60 57 64 65
0 59 62 57 57 62 66
+5 74 76 78 72 76 81
+10 84 81 84 79 81 85
+15 95 90 94 87 90 93
Plasma Norepinephrine (pg/ml)
-10 200 225 210 375 250 250
0 210 200 200 360 225 275
+5 400 450 450 595 500 500
+10 400 575 700 900 575 800
+15 800 700 825 1210 700 1000

During session 1 there was no significant difference

between groups in supine or graded stress plasma NE levels.
In both groups, plasma NE increased significantly as a function
of the three graded stresses.
After eliciting the relaxation response for 30 days, the
experimental group displayed significantly greater NE concentrations
with graded stress than on session 1. Further, the experimental
and control groups had significantly different patterns of NE
responses on the two sessions and differed significantly at each

828
level of stress. At +15, the mean NE concentrations increased from
the first to the second session in the experimental group. Eight
of the ten experimental subjects exhibited increased NE concentra-
tions. No significant differences in the mean NE concentrations
in the control group between sessions 1 and 2, were observed and
only one of the nine control subjects displayed increased NE.
In the crossover condition, the mean plasma NE response to the
graded stress was significantly greater on session 3 than on session
2. Four of the six original control subjects who subsequently
elicited the relaxation response, exhibited this change.
In both control and experimental groups, systolic and diastolic
(not shown in Table 1) BP increased progressively with graded stress.
On session 1, BP did not differ between groups in any condition.
The significant augmentation of plasma NE release exhibited by the
experimental group in session 2 was not associated with a parallel
increase in BP. In session 2 for the experimental group and in
session 3 for the crossover group, systolic and diastolic BP tended
to be lower than in previous sessions but the results failed to
achieve statistical significance.
Graded stresses also resulted in progressive increases in HR.
The two groups were similar on session 1. On session 2, HR levels
were not distinguishable from those of session 1 in either group.
There were also no changes when sessions 2 and 3 were compared in
the crossover group.
CONCLUSIONS AND IMPLICATIONS
Subjects who regularly elicited the relaxation response for 30
days exhibited an augmented plasma NE response to graded stress while
subjects who practiced a control technique of sitting quietly
exhibited no change. Despite this difference in NE reactivity BP
and HR response to the graded stress was similar in both groups.
These results were replicated when subjects who initially practiced
the control technique subsequently elicited the relaxation response
for 30 days.
In this study, and as previously reported by others (8,9)
regular elicitation of the relaxation response did not result in
a change in plasma NE under low-stress conditions such as supine
posture. However, when exposed to higher-stress conditions such
as upright posture or isometric stress, subjects who had regularly
elicited the relaxation response exhibited augmented plasma NE
responsivity. The sympathetic nervous system is primarily
responsible for mediating the cardiovascular adjustments to these
stress conditions. In this study, the index of sympathetic nervous
activity, plasma NE, increased disproportionately over BP and HR.
These results suggest that in order to achieve the normal

829
compensatory increases in HR and BP required by the postural and
isometric stress, subjects eliciting the relaxation response
required more NE.
A reduction in adrenergic and organ responsivity may result
from elicitation of the relaxation response although the mechanism
responsible for such a change is not clearly identified (12). The
data from this study are consistent with a previous report (13)
suggesting that elicitation of the relaxation response may be
associated with reduced reactivity to stress.
The relaxation response is a very popular and widely used
technique. It has been demonstrated that regular elicitation of
the relaxation response can be successfully employed in the treat-
ment of disorders in which increased sympathetic nervous system
activity is undesirable. Examples of such disorders are: hyperten-
sion, cardiac arhythmias, anxiety, and pain. The evidence cited
here establishes for the first time, a mechanism to account for the
fact that elicitation of the relaxation response once or twice a day
has a positive effect on stress reactivity throughout the entire
24 hour cycle.
REFERENCES
1.H. Benson, The Relaxation Response," Morrow, New York, (1975).
11

2.R. K. Wallace, H. Benson and A. F. Wilson, A wakeful

hypometabolic physiologic state, American Journal o~
.!'hvs i o1oqv, 221 :795-799 (1971 ) .
3. W. B. Cannon, The emergency function of the adrenal medulla
in pain and the major emotions, American Journal of
Phvsioloqv. 33:356-372 (1914).
4. W.' R. Hess, -.. The Functional Organization of the Diencephalon,"
Grune and Straton, New York (1957).
5. R. K. Peters, H. Benson and J. M. Peters, Daily relaxation
response breaks in a working population, American Journal
of Public Health. 67:954-959 (1977).
6. H. Benson, Systemic hypertension and the relaxation response,
New Enqland Journal of Medicine. 296:1152-1156 (1977).
7. J. F. Beary and H. Benson, A simple psychophysiologic technique
which elicits the hypometabolic changes of the relaxation
response, Psvchosomatic Medicine. 36:115-120 (1974).
8. R. R. Michaeis, M. J. Huber and D. S. McCann, Evaluation of
transcendental meditation as a method of reducing stress,
Science! 192:1242-1244 (1976).
9. R.· Lang, K. Dehof, K. A. Meurer and W. Kaufman, Sympathetic
activity and transcendental meditation, Journal of Neural
Transmission. 44:117-135 {1979}. ·~· ·

830
10. J. W. Hoffman, H. Benson, P. A. Arns, G. L. Stainbrook,
L. Landsberg, J. B. Young and A. Gill, Reduced sympathetic
nervous system responsivity associated with the relaxation
response, Science. 215:190-192 (1982).
11. C.R. Lake, M. G. Ziegler and I. J. Kopin, Use of plasma
norepinephrine for evaluation of sympathetic neuronal function
in man, Life Science. 18:1315-1326 (1976).
12. P. E. Cryer, Physiology and pathophysiology of the human
sympathoadrenal neuroendocrine system, New Enqland Journal
of Medicine. 303:436-444 (1980).
13. R. K. Michaels, J. Parra, D. S. McCann and A. J. Vander,
Renin, cortisol~ and aldosterone during transcendental
meditation, _Psychosomatic MediCine. 41:50-54 (1979).

*The experimental work presented here was originally

reported by Hoffman et al., 1982. See reference 10.

831
REDUCTION OF CARDIAC RISK FACTORS BY AUTOGENIC TRAINING
AND PHYSICAL TRAINING

Malcolm Carruthers
Director of Clinical Laboratory Services
Maudsley &Bethlem Royal Hospital
Denmark Hill, London SE5 8AZ

Physical training has long been regarded as one of the ways of

balancing up the physical responses induced by stress, and the
craze for jogging has become a mania to the extent that psychi-
atrists and psychologists, being unable to beat the craze, have
joined it and now run alongside their patients as part of the
treatment of depression (1). Not only is this regarded as
having a powerful influence in reducing cardiac risk factors
(2) and preventing (3) and treating (4) coronary heart disease,
but there are encouraging signs that by regularly increasing the
level of sympathetic activity by periods of vigorous but not
violent physical activity, there is a beneficial rebound
enhancement of parasympathetic activity in the resting state.
This is shown by the lower blood pressure, slower pulse rate,
and greater. sinus arrhythmia effect seen in individuals who
undertake physical training. As shown by the research at the
City Gymnasium in London reported by myself and Alistair
Murray in our book "F/40: Fitness on 30 Minutes a Week" (5).
the amount of exercise needed to achieve a significant and
useful cardia-protectiv e training effect is fortunately
relatively small and can be controlled safely by attention to
maximum pulse rate and perceived exertion. We are also developing
wrist watches which will automatically signal when either heart
rate or blood pressure exceed predetermined safety levels.
Mental trffining t relatively poorly evaluated as a way of
reducing cardiovascular risk factors. Pioneer work in this
field has been done by Dr Chandra Patel, and in a joint study
carried·-out in her general practice population we were able to
show that biofeedback-aided meditation produced significant
reductions in several lipid fractions, including cholesterol.

833
as well as in blood pressure in both normal and hypertensive
subjects (6, 7).
It was with considerable interest therefore that we learned
from Dr Luthe that similar effects could be obtained by Autogenic
Training (8). Working at the Autogenic Training Centre in
London, we applied a cardiovascular monitoring system known as
Chemofeedback before and after the two month standard course
in Autogenic Training, and regularly observed reductions in
blood cholesterol and in blood pressure in both normo-tensive
and hypertensive individuals. In one particular case of a top-
ranking airline pilot who had been on a low fat diet and a wide
variety of drugs for several years to lower his blood cholesterol,
we observed a reduction in cholesterol from 410 to 275 mm/100ml.
Also the majority of patients were able to progressively reduce,
and then stop, a wide variety of medications including tran-
quillizers, sleeping pills, and blood pressure lowering agents
during the course and practise of the training.
However, as in the majority of studies so far reported,
this work was uncontrolled and a further study was designed
to overcome this objection. 48 senior managers of the National
Coal Board kindly agreed to take part in this study during their
six week residential course at the Staff College in the beautiful
setting of Chalfont St Giles, Buckinghamshire. They were
randomly allocated to either the control, physical training,
or Autogenic Training groups during the first and last weeks
of the test. The following psychological, physiological, and
biochemical tests were carried out:-
a) Psychological Tests
Eysenck's personality questionnaire (adult)
Goldberg general health questionnaire for the detection
of psychiatric illness in the outpatient population
Three measures of Rosenman and Friedman's now generally
accepted coronary prone Type A behaviour pattern
Bortner questionnaire
Framingham questionnaire
Carruthers 10 item listing
b) Physiological Measures
Resting heart rate and blood pressure measured with
a conventional sphygmomanometer

834
 c) Biochemical Measures
Fasting cholesterol, triglyceride, free fatty acids,
high density lipoprotein, glucose, uric acid

Results
All subjects completed the study except for three in the
exercise group who dropped out early on because one injured
his foot in a cricket match, and two others because of minor
muscle and tendon injuries. Several people in both training
groups commented that they found the sessions pleasantly
relaxing after the intense mental activity involved in the
advanced management course. As instructed, the control group
later confirmed that they kept their level of physical
activity and food intake to the same as they would at home.

Psychological Measures
Analysis of the EPQ showed that they were low on all the four
scores, i.e. psychoticism, extraversion, neuroticism, and
lying. From this they appeared to be a very stable group of
self-selected individuals who would be good at both public and
private interpersonal relationships. The information on Type A
behaviour is still being analysed and will be reported separately.
Physiological Measures
Paired student T tests showed significant reductions in pulse
rate and systolic and diastolic blood pressures in the Autogenic
Training group with smaller reductions in pulse rate and
diastolic blood pressure in the physical training group, and
n.o change in the control group. (Figure 1)
Biochemical Measures
There were reductions in cholesterol in both the AT and PT
group. Triglycerides and free fatty acids were reduced in
the AT group, but not in the PT group. There was a slight
increase in triglyderide in the control group, but no changed
in cholesterol or free fatty acids. (Figure 2)
High density lipoprotein increased in the PT group, but was
unchanged in the other two. Similarly uric acid was
decreased in the PT group, but unchanged in the other two.
Plasma glucose was decreased after training in both the AT
and PT group with again no change in the control group.
(Figure 3)

835
 Control AT PT
Pre Post Pre Post Pre Post

BPM PULSE RATE

70 ~r-;-l ~,0.01
'
'' i !· . ::~.'-I
60

mmlig SYSTOLIC BLOOD PRESSURE

0 f, '
130 1-

''

120
0.05'',1
~---~~--I
mmlig DIASTOLIC BLOOD PRFllSURE

80 1- 1---I l,'
' L
NS o.oo1·J ·~.01

70 1-
·····I
Fig. 1 Pulse rates and blood presssures before and after six
weeks Autogenic Training (AT 16 men) and Physical
Training (PT 14 men) compared with a control group of
16 men.

836
 Control AT PT
Pre Post Pre Post Pre Post

mmol/1 CHOLESTEROL
7.0 r- l
'' 1..
I I ' \
6.) ~ NS 0.001' ·'?.001

''t
6.0 r- ·y

1.4 ~ TRIGLYCERIDE:>

r-
?
1.2 NS
}..._0:..05
-
1.0

0.8 - '""' J······ ··I

0.8 r- FREE FATTY ACIIS

0.7 r-
0.6 1- }, ,0.001 I··;;;---I
0.5 1- ~ '
''I
0.4 1-

Fig. 2 Lipid levels before and after six weeks Autogenic

Training (AT 16 men) and Physical Training (PT 14 men)
compared with a control group of 16 men.

837
 Control AT PT
Pr e Post Pre Post Pre Post
mmol/1
6 - GLUCOSE

l, I,
5
1-;-r ''
'
0.001'!
.......
o.oo"lt-f

4
0.45 ~
URIC ACID

0.40 ~~ NS l~.05....
0.35 ~
t---1 '•i
HDL
1.6 ~ .. .·
1.4 ~
NS
I--~--I
ri)_., ..··

1.2 ~
Fig. 3. Glucose, uric acid and high density lipoprotein (HDL)
levels before and after six weeks Autogenic Training
(AT 16 men) and Physical Training (PT 14 men) compared
with a control group of 16 men.

838
Conclusions
In this controlled trial there were significant reductions
in a wide range of cardiovascular risk factors. As these risk
factors don't merely add up, but multiply up, (9), this multi-
factorial approach to "dividing and conquering heart disease"
should have a powerful net effect. It is an inexpensive and
life-enhancing technique by which the individual can gain a
measure of volunatry control over the involuntary nervous
system, and has a good adherence rate as judged by the results
of a questionnaire administered to the AT and PT groups six
months later on a follow-up course. Also both short-term and
long-term there appears to be a lower drop out rate with AT
than with PT. Ideally however one would encourage a "Positive
Health Programme" where both methods of training are combined
to have what should be a synergistic action.

REFERENCES
1. Griest, J.H., Klein, M.H., Eischens, R.R. Running through
your mind. J. Psychosomatic Research, 22: 259-294, 1978
2. Durbeck, D.C., Heinzelmann, F., Schacter, J. et al. The
National Aeronautics and Space Administration - US Public
Health Service Evaluation and Enhancement Program. Am. J.
Cardiology, 30:784-790, 1972
3. Morris J.N., Everitt, M.G., Pollard, R. et al. Vigorous
exercise in leisure time: protection against coronary
heart disease. Lancet 2:i207-1210, 1980
4. Kavanagh, T., Kennedy, J., Qureshi, S. Prognosis in myo-
cardial infarction - the benefits of exercise as seen in
non-randomized trials. Brit. J. Sports Med. 15:6-16, 1981
5. Carruthers M. and Murray A., F/40: Fitness on Forty
Minutes a Week. Futura Books, London, 1976
6. Patel C. and Carruthers M. Coronary risk factor reduction
through biofeedback-aided relaxation and meditation.
J. Royal College Practitioners 27:401-405, 1977
7. Patel, C., Marmot, M.G., Terry, D.J., Controlled trial of
biofeedback-aided behavioural methods in reducing mild
hypertension. Brit. Med. J. 282:2005-2008, 1981
8. Luthe, W. Autogenic Therapy, vol 4, Research and Theory,
Grune and Stratton, New York, 1969
9. Carruthers, M. Log-a-Risk: An aid to advising the
coronary candidate. J. Royal College General Practitioners,
25:30-223, 1975

839
HIGHER NERVOUS ACTIVITY IN PSCHIATRIC PATIENTS

Christian Astrup
Gaustad Hospital
Oslo, Norway

ABSTRACT
A review is presented of studies of higher nervous
activity in psychiatric patients. In neuroses and
reactive psychoses the basic pathology appears to be
centered around the psychogenic complex structures. In
addition, neuroses as well as reactive psychoses reveal
general disturbances of higher nervous activity.
In the schizophrenic and manic-depressive-psychoses
there are indications of disturbances in deep-lying
brain structures. Both types of psychoses are
heterogeneous groups of clinical conditions. An
important task for future experimental studies is to
establish the types of disturbances of higher nervous
activity in well defined clinically homogeneous groups.
1. NEUROSES
Neuroses are conceived of as psychogenic disorders.
In accordance with this there are strong effects of
complex structures. More unexpected is the general
impairment of higher nervous activity. Neurotics are
markedly slower in word responses and respond with a
poorer quality than normal controls. They also often
fail in complex motor tasks. The neurotics show less
autonomic respones to complex words and other stimuli
than normal controls. All this suggests that neurotics
tend to have inhibitions at various levels of higher
nervous activity. Accordingly the neuroses are probably
841
not only centered around psychological mechanisms, but
also involve general neurophysiological mechanisms. In
Berlevag project out-patient neurotics revealed a
considerable inhibition of conditional skin conductance
responses. The neurotics also had general disturbances
of word associations, but not a great as in the Gaustad
Hospital material. T~is is understandable, because only
severe neurotic conditions were admitted to our hospital
(Astrup, 1983; Smith-Meyer et al., 1976).
For treatment of neurotics, the author has
elaborated a method of flooding in hypnosis. With
flooding there apprears to occur some kind of
"exhaustion" of pathologically exited and inert complex
structures, which diminishes anxiety (Astrup, 1975,
1978). The effects of flooding appear to be best when
the general disturbances of higher nervous activity are
slight, so that the pathology is dominated by the
complex structures. The same may very well turn out to
be the case for the psychodynamic therapies.
2. SCHIZOPHRENIA
My first material of schizophrenics was studied
during the years 1955-1958. The studied population made
up 489 cases, and the chronic cases were subdivided in
the 19 groups described by Leonard (Astrup, 1962). The
patterns of disturbances of higher nervous activity vary
very much for different subgroups. Certain paranoid
cases perform nearly as well as normal controls, while
severly deteroirated catatonics performed like 2- to 4-
year-old children. Inhibition of unconditional reflexes
appear to play a great role. The same was the case for
dissociative responses. In a series of 122
schizophrenics studied in the 1960's incoherent reponses
with the word association rest especially separated the
schizophrenics from other clinical groups (Astrup &
Flekkoy, 1968).
Prognostic models were calculated by the computer
for followed up functional psychoses. Experimental data
could predict the long-term clinical outcome as well as
coded clinical symptomatology (Astrup & Noreik, 1966).
With improved experimental techniques, there should be
good prospects of predicting long-term outcome better
from higher nervous activity data, than from the
clinical symptomatology.
Clinical follow-ups demonstrate that neuroleptics
treatment can prevent severe schizophrenic deterioration
to a great extent. The effects of drugs can be

842
illustrated by much less pathology of word associations
in chronic schizophrenics in the 1970's than in 1955-1957
(Flekkoy et al, 1975). A hypothesis has been put forward
that neuroleptics act through strengthening protective
inhibition (Astrup, 1962).
A recent review of Soviet studies points out that
there are characteristic patterns of cognitive
dysfunction in schizophrenics (Zavarin et al., 1982).
Venables (1978) thinks that deficits in the more complex
cognitive processes were probably due to disturbances of
attention and perception rather than of thought itself.
According to Venables (1981) about 55% of schizophrenics
are non-responders or hyporesponders as measured with
autonomic components of the orienting reflex. More
controversial is an inverse finding of hyper-
responsibility among some schizophrenics.
Kubryashev (1979) finds variations of cardivascular
responses for different types of schizophrenia. Horvath
and Meares (1979) find that non-paranoid schizophrenics
have lost the normal inverse relationship between
habituation and level of arousal as manifested in the
rate of spontaneous skin conductance fluctuation. In a
tachistoscope study Magaro and Chamrad (1983) found left
hemisphere dysfunction in non-paranoid schizophrenics.
Rest et al. (1981) studied autonomic conditioning in
schizophrenics. Their findings possibly reflected a
heightened sensitivity of the cardiovascular system.
Schwartz and Winstead (1982) observed visual processing
deficits in acute and chornic schizophrenics. There is
an increasing amount of electrophysiological studies
suggesting.that schizophrenics are abnormal with regard
to sleep patterns, hemispheric and brain stem functions
(Buchsbaum, 1975, Gruzelier and Venables, 1974, Holzman
et al., 1976, Roth 1980, Shagass et al, 1976). Sem-
Jacobsen and Astrup are currently using a new
electrophysiological technique to investigate how the
deeper lying structures of the brain are affected in
different types of schizophrenia. The preliminary
results show that among 27 cases of chronic
hallucinatory and delusional schizophrenics only one had
a pathological EEG, while 11 or 27 nuclear
schizophrenics had pathological responses.
Schizophrenia is probably a heterogeneous group of
psychoses, ranging from psychogenic developments over
functional disturbances of biological systems to organic
brain lesions. With computer tomography Luchins et al.
(1982) finds cerebral asymmetry in schizophrenics.
Stevens (1982) thinks there are neuropathological

843
abnormalities in three fourths of the brains from
schizophrenic subjects. The nature and distribution of
the findings suggest previous or low-grade inflammation.
3. MANIC DEPRESSIVE PSYCHOSES
Comparatively few experimental studies have been
carried out in manic depressives. Common to
schizophrenia and manic depressive psychoses is a
tendency to inhibition of the unconditional defensive
reflexes. This may suggest that changes of deep-lying
brain structures are common for the endogenous
psychoses. An important experimental basis of
differential diagnoses between manic depressives and
schizophrenics is the presence of incoherent word
associations in the latter illness (Astrup and Flekkoy,
1968).
4. REACTIVE PSYCHOSES
In the reactive psychoses marked effects of complex
structures indicate the pathogenetic importance of
psychodynamic factors, just as in the neuroses. In
contrast to manic depressive psychoses there apprears to
be no inhibition of unconditional defensive reflesex. In
experimental studies the various subgroups of reactive
psychoses have not been well differentiated from each
other (Astrup and Flekkoy, 1968).
REFERENCES
Astrup, C. (1962) Schizophrenia: Conditional reflex
studies. Thomas, Springfield, Illinois.
Astrup, C. (1965) Pavlovian psychiatry: a new
synthesis. Thomas, Springfield, Illinois.
Astrup, C. and Noreik, K. (1966) Functional psychoses:
Diagnostic and prognostic models. Thomas,
Spingfield, Illinois.
Astrup, C. and Flekkoy, K. (1968) Association
experiments in psychiatric patients and normal
controls. Activ. nervos. Superior. 10. 373-381.
Astrup, C. and Flekkoy, K. (1969) Verbal and electrical
stimulation in pschiatric patients and normal
controls. Activ. Nervos, Superior, 11. 1-10.
Astrup, C. (1975) The psychological mechanisms of
flooding (implosion) therapy. Psychoter. Psychoson.
25, 63-68.
844
Astrup, C. (1978) Physiological mechanisms of flooding
(implosion) therapy. Pavlov, J. Biol, sci. 4, 195-
198.
Astrup, C. (1979) The Chronic schizophrenias.
Columbia Un. Press, New York.
Astrup, C. and Vatten, L. (1983) Effect of the
benzodiazepine derivate estazolam in schizophrenia.
Biol. Psychiat. In press.
Berger, N., Doerr, P., Lund, R., Bronich, T. and
Zerssen, D. von (1982) Neuroendocrinological and
neurophysiological studies in major depressive
disorders. Are there biological models for the
endogenous subtype? Biol. Pschiat. 17. 1217-1242.
Bruno, R.L., Myers, S.J. and Glassmann, A.H. (1983) A
correlational study of cardiovascular autonomic
functioning and unipolar depression. Biol.
Psychiat. 18, 227-235.
Buchbaum, M. (1975) average evoced response, augmenting/
reducing in schizophrenia and affective disorders.
In Friedman, D.X. (ed.) Biology of the major
psychoses. Raven, New York.
Flekkoy, K. and Astrup, C. (1968) Motor conditional
reflexes in psychiatric patients and normal
controls. Activ. Nervos. Superior. 10, 105-111.
Flekkoy, K., Lund, J. and Astrup, C. (1975) A prolonged
clinical and experimental follow-up of hospitalized
schizophrenics. Neuropsychobiol. 1. 47-58.
Gruzelier, J.H. and Venables, P.H. (1974) Bimodality and
lateral asymetry of skin conductance orienting
activity in schizophrenics. Biol. Psychiat. 8, 55-
73.
Heimann, H. (1979) Psychophysiologie endogener
Psychosen. Schweiz. Arch. Neurol. Neurochir. 125,
231-252.
Holzman, P.S., Levy, D.L. and Proctor, L.R. (1976)
Smooth-pursuit eye movements, attention and
schizophrenia. Arch. Gen. Psychiat. 33, 1415-1420.
Horwath, T. and Meares, R. (1979) The sensory filter in
schizophrenia. A study of habituation, arousal, and
the dopamine hypothesis. Brit. J. Psychiat. 134,
39-45.
845
James, A.L. and Barry, R.J. (1980) A review of
psychophysiology in early onset psychosis. Schiz.
Bull. 6, 506-525.
Kostandov, E.A. (1980) Status and perspectives of
physiological investigations in psychiatry (Rus.)
Z. Neuropatol. Psichiat. Korsakoff 80, 552-560.
Kudriashev, V.E. (1979) Possible system approach to the
study of regulation of the cardiovascular system in
schizophrenia (Rus.) Z. Neuropathol. Psichiat.
Korsakoff, 79, 592-596.
Leonhard, K. (1979) Classification of the endogenous
psychoses. Wiley, New York.
Luchins, D.J., Weinberger, D.R. and Wyatt, R.J. (1982)
Schizophrenia and cerebral asymmetry detected by
computer tomography. Amer. J. Psychiat. 139, 753-
757.
Magaro, P.A. and Chamrad, D.L. (1983) Information
processing and lateralization in schizophrenia.
Biol. Psychiat. 18, 29-44.
Monakhov, K.K. and Kamenskaya, V.M. (1980) The status of
neurophysiological investigations in psychiatry
(Rus.) Z. Neuropathol. Psichiat. Korsakoff, 80,
560-564.
Rist, F., Baumann, W. and Cohen R. (1981) Effects of
awareness and motor involvement on autonomic
conditioning in chronic schizophrenics. Pavlov. J.
Biol. Sci. 16, 8-17.
Rofth, W.T., Horvath, T.B., Pfefferbaum, A. and
Kopell, B.S. (1980) Event-related potentials in
schizophrenics. Electroencephal. Clin.
Neurophysiol. 48, 127-139.
Schulz, H. and Tetzloff, W. (1982) Distribution of REM
latencies after sleep interruption in depressive
patients and control subjects. Biol. Psychiat. 17,
1367-1376.
Schwartz, B.D. and Winstead, D.K. (1982) Visual
processing deficits in acute and chronic
schizophrenics. Biol. Psychiat. 17, 1377-1387.
Shragass, c., Raemer, R.A. and Amadeo, M. (1976) The
eye-bracking performance and engagement of
attention. Arch. Gen. Psychiat. 33, 121-125.
846
Stevens, J.R. (1982) Neuropathology of schizophrenia.
Arch. Gen. Psychiat, 39, 1131-1139.
Venables, P.H. (1978) Cognitive disorder. Jn. Wing, J.K.
(ed.) Schizophrenia: Towards a new synthesis.
Academic Press, London.
Venables, P.H. (1981) Psychophysiology of abnormal
behaviour. Brit. Med. Bull. 37, 199-203.
Zavarin, V., Tonkonogy, J. and Ostwald, P. (1982)
Cognitive processes in schizophrenia and related
disorders. Pavlov, J. Biol. Sci. 17, 188-203.

847
THE TWO MODELS OF CONDITIONING OF NEUROTIC ANXIETY

Joseph Wolpe
Temple University
Department of Psychiatry
3300 Henry Avenue
Philadelphia, Pa. 19129 U.S.A.

Classically Conditioned and Cognitively Based Fears

Behavioral scientists who have been nurtured in the conditioning
tradition have naturally thought of unadaptive (neurotic) fears as
based, like adaptive fears, on classical conditioning. In recent
years a zealous coterie of mentalistic psychologists (e.g. Beck,
1976; Ellis, 1962) have vigorously articulated an opposed position -
that neurotic fears are always due to the belief that there is danger
where in fact there is none. However, many behavioral clinicians
(e.g. Wolpe, 1958, 1983; Rachman, 1977) have recognized that not all
clinical fears are classically conditioned - that some do depend on
unfounded beliefs that are either verbally communicated or inferred
from the fearful behavior of others. An early formal demonstration
of this dichotomy was actually provided by Bridger & Mandell (1964)
and then ignored for two decades.
The original basis of the sympathetic-dominated pattern of auto-
nomic responses that constitutes fear is to be found in the young
child's unconditioned responses to noxious, i.e. painful, stimulation
as, since the time of John Watson (1913) has been generally recognized.
The sympathetic arousal by noxious stimulation is very obvious when
one uses electrical stimulation to produce experimental neuroses
(Masserman, 1943; Wolpe, 1952; Smart, 1965).
From a very early age, "neutral" stimuli making impact on a
child at any time when pain evokes fear become conditioned stimuli
to fear. For example, if a mother's voice acquires a certain tone
whenever she slaps her child, the autonomic responses that the pain-
ful slap produces will come to be conditioned to that tone of voice.
849
In a similar way, the burning of a child's hand conditions in him
a fear of the stove. Such conditioning can theoretically occur with
any kind of stimulation that happens to be contagious with the
autonomic arousal - sounds, things, people, words, ideas - anything
seen or heard or felt. Later, by second order conditioning, fear
may be spread from one stimulus to another. However, as is now
well-known, on the basis of the work of Seligman (1970) and of
6nman and his colleagues (1975), not all stimuli are equally condi-
tionable to fear.
Once classical conditioning has established a repertoire of
conditioned fear responses in a child, a foundation exists for
further spread of fear through the agency of language. Language
enables fear stimuli to be cognitivel~y grouped together under
verbal labels that center on the idea of "danger". This labeling
process amounts to cognitive learning in which new stimulus-response
connections are established at the highest level of neural activity.
We fear lightning or contaminated food, or being driven by a reckless
driver - not because of classical conditioning, but because of cog-
nitive learning. These are "normal" fears of real dangers.
Apart from the testimony of everyday experience, cognitive
transmission of fear has been systematically recorded in both ani-
mals and man. Miller, Murphy &Mirsky (1959) showed that fear can be
acquired by an observer monkey who watches a fearful model monkey or
even pictorial representations of fearful monkeys. In humans,
Grinker &Spiegel (1945) described combat airmen whose fears orig-
inated from observing a crewmate expressing intense fear. Similar
reports have been provided by Hagman (1932) and Solomon (1942).
Classically conditioned fears that are neurotic (i.e.,
unrealistic) begin like their "normal" counterparts in relation to
actual fear arousal -often a solitary experience (Wolpe, 1981).
For example, a severe claustrophobia began from one terrifying exper-
ience in a cave; and fears of public scrutiny not infrequently
originate from a single overwhelming embarrassment in a classroom.
Unrealistic fears of a cognitive kind similarly parallel "normal"
cognitive fears in their mode of acquisition. Fears of masturbation
are frequently the re~ult of the false belief that it leads to
physical or mental harm; and fears of sexual arousal in a woman may
come from observing her mother's repugnance. A person may fear
worms, flying insects, doctors or hospitals as a result of repeatedly
observing a parent show fear of these things.
The Corresponding Treatment Modalities
The reciprocal inhibition principle (the use of anxiety-
inhibiting responses) was a major advance in the treatment of class-
ically conditioned fears. The best known technique is systematic

850
desensitization in which relaxation-induced calmness competes with
anxiety. Outcome studies (Leitenberg, 1976; Paul, 1966; 1969) have
shown desensitization to be both clinically and statistically more
effective than insight and no treatment control groups. Paul's
(1966) study on subjects with severe fears of public speaking was
particularly noteworthy because the therapists were all psycho-
analytically oriented and nevertheless were significantly more
successful with systematic desensitization than with their own
accustomed therapy or a control procedure. Other classical condi-
tioning methods include assertiveness training and flooding.
Another modern advance in the treatment of inappropriate anxiety
has been cognitive therapy (Beck, 1976; Ellis, 1962) which is de-
signed to eliminate irrational thoughts, faulty logic and mistaken
beliefs which are often the basis of unadaptive fear. Cognitive
therapy involves the identification and elimination of maladaptive
cognitions which lead to fear. After determining the patterns,
misconceptions, and beliefs that are the basis of the individual's
fear, treatment takes the form of providing corrective information.
There are many clinical papers that provide evidence of the efficacy
of cognitive therapies (e.g. Bandura, Blanchard and Ritter, 1969;
Mahoney, 1974; and Meichenbaum, 1977). Unfortunately, they are all
confounded by the fact that none of the authors has paid attention
to the diagnostic differentiation between classically conditioned
and cognitively based anxieties. The only statistical data that is
directly relevant is in a still unpublished study of my own, in
which 13 of 14 cases diagnosed as cognitive overcame their anxieties
through cognitive correction.
Distribution and Diagnosability of _the Subclasses of Fear
The clinical distribution of the two fear bases is a matter of
considerable importance. In a retrospective survey of 40 cases
(Wolpe, 1981), I found that the fear of 24 was exclusively classi-
cally conditioned; in 14 it was exclusively cognitively based, and
in 2 both bases were present. Obviously, much more extensive work
is necessary.
It is, of course, important for clinicians in general, and
behavior therapists in particular, to be able to distinguish between
the two fear bases. Otherwise the distinction has little practical
value. Two recent studies in our clinic have shown that it is not
at all difficult to learn how to make the distinction. In the first
study, 3 groups of clinical students at the beginning of their
academic year were respectively given 1, 2, or 3 seminars on making
the distinction. Then each group was given 10 separate examples of
transcripts of the decisive part of an interview between an exper-
ienced therapist and a patient. The scores were as shown on the
slide. In the group that had had one practicum the range of correct
responses was 6-10, in the groups that had had two and three the

851
ranges were 7-10 and 8-10 respectively. the last having the extremely
significant Kappa coefficient of 0.876.
The second study was designed to assess to what extent inde-
pendent assessors would agree with an 11 expert's 11 diagnostic dis-
tinction. Two psychologists from the third group of the afore-
mentioned study were blind assessors of cases whose diagnostic
subclass had already been decided by an expert. In the first 8
cases, all three assessments were in aqreement. The binomial
probability of this, given a base rate chance agreement rate of
sb-50, is p ~ .004.
It is actually often obvious from the outset that a particular
case must be classically conditioned and that another must be cogni-
tively based. For example, a man of 51 had for many years had anxiety
at the sound of the telephone. He was not aware of any danger at
the sound, so there was no wrong conception to correct. He responded
as expected rapidly to systematic desensitization. By contrast, a
woman of 29 years who had severe agoraphobia was afraid to venture
out alone because she believed that her frequent attacks of
dizziness and tingling of the hands were early signs of 11 going crazy 11 •
She had an aunt who had been in a mental hospital and two cousins
who were 11 peculiar 11 , and had thus concluded that she was hereditarily
unstable. I demonstrated to her that her dizziness and tingling
were due to hyperventilation, and strongly contested any possibility
of her going crazy. She rapidly lost her fearfulness. After 7
sessions she was able· to go freely anywhere on her own. Systematic
desensitization was not considered, since it could not have been
expected to overcome her fear as long as she held this wrong belief
of imminent mental destabilization.
The Relationshio Between Diaanostic Subclass and Effective Treatment
Any behavior therapist who gets into the habit of making the
distinction between classically conditioned and cognitively based
anxiety neuroses, and who bases his choice of method on the distinc-
tion, is soon aware of a marked increase in his therapeutic efficacy.
However, there is no systematic data. The only exception is a small
scale study that I reported in 1979. Of 25 cases of neurotic
depression, 19 were found due predominantly to classically conditioned
anxiety and 6 were cognitively based. Seventeen of the 19 responded
markedly to reconditioning; and 5 of the 6 to cognitive correction.
To obtain further data we are planning the following study.
Two common fears: (1) claustrophobia, and (2) fear of physical
sensations will be the subject matter. Cognitively based and emo-
tionally conditioned cases will be diagnosed by clinical interview.
Cases of each type will be treated either bv coanitive correction or
by emotional reconditioning procedures. The effects of treatment
will be evaluated by three forms of assessment: (1) subjective

852
 Table 1
Mean Score Kappa
(Possible 10) Ranqes Coefficient

Group I (10 subjects) 8.2 6-10 0.419

Group II ( 5 subjects) 9.2 7-10 0.676
Group III (10 subjects) 9.6 8-10 0.876

anxiety ratings, (2) physiological arousal, and (3) the degree of

alleviation of the disruption of life functioning caused by the fear.
It is expected that fears which are the result of cognitive habit
will respond to cognitive correction, while those which are a matter
of emotional conditioning will be significantly more responsive to
emotional reconditioning techniques.
If the predictions of this study are upheld, there will no
longer be grounds for anyone to contend that all neurotic fears are
due to faulty thinking on the one hand or to classical conditioning
on the other. It will be unequivocally clear that there are two
different fear bases and that to overcome one requires a mode of
treatment that is specific to it and distinct from the kind of
treatment required to overcome the other. Psychotherapeutic activity
will be seen to consist of two broad categories of methods, each
applicable to only one category of case -to the classically condi-
tioned or to the cognitively based.
These scores show that the diaqnostic distinction is quite easily
communicated, and that the ability to make it from qiven data in-
creases rapidly with training.

References
Bandura, A., Blanchard, E.B. and Ritter, B., 1969, Relative efficacy
of desensitization and modelling approaches for inducing behav-
ioral, affective, and attitudinal changes. J. Pers. and Soc.
Psychol., 13:173.
Beck, A.T., 1976, Cognitive Therapy and the Emotional Disorders,
11 11

International U. Press, New York.

Bridger, W.H. and Mandell, I.J., 1964, A comparison of GSR fear
responses produced by threat and electric shock. J. of
Psychiat. Research, 2:31-41.

853
Ellis, A., 1962, "Reason and Emotion in Psychotherapy," Lyle Stuart,
New York.
Grinker, R. and Spiegel, J., 1945, "Men Under Stress," Blakiston,
Philadelphia.
Hagman, C., 1932, A study of fears in pre-school children. J. of
ExoPr. P~vrhnl., 1:110.
Leitenberg, J., 1976, "Handbook of Behavior Modification and Behav-
ior Therapy," Prentice Hall, Englewood Cliffs.
Mahoney, M.J., 1974, "Cognition and Behavior Modification,"
Ballinger; Cambridge.
Masserman, J.H., 1943, "Behavior and Neurosis," Univ. of Chicago
Press, Chicago.
Meichenbaum, D.H., 1977, "Cognitive-Behavior Modification," Plenum,
New York.
Miller, R., Murphy, J., and Mirsky, I., 1959, Non-verbal communica-
" tion of affect. Journ. of Clin. Psvchol., 15:155.
Ohman, A., Erixon, G. and Lofberg, I., 1975, Phobias and prepared-
ness·: Phobic versus neutral pictures as conditioned stimuli
for human autonomic responses. J. Abn. Psychol., 84:41.
Paul, G.L., 1966, "Insight Versus Desensitization 1n· Psychotherapy,"
Stanford Univ. Press, California.
Paul, G.L., 1969, Outcome of systematic desensitization. In
Franks, C., "Behavior Therapy: Appraisal and Status." McGraw-
Hill, New York.
Rachman, S., 1977, The conditioning theory of fear-acquisition:
A critical examination. Behav. Res. &Ther., 15:375.
Seligman, M., 1970, On the generality of the laws of learning.
Psvchol. Rev. 77:406.
Smart, R.G., l96b, Conflict and conditioned aversive stimuli in the
development of experimental response. Canad. J. Psvchol.
19:208.
Solomon, J., 1942, Reactions of children to blackouts. Amer. J. of
Orthoosvchiat., 1:110.
Watso~ J., 1913, ·Psychology as the behaviorist views it. Psvchol.
Rev. 20: 158.
Wolpe, J., 1952, Experimental neurosis as learned behavior, Brit. J.
Psvchol., 43:243a.
Wolpe, J., 1958, "Psychotherapy By Reciprocal Inhibition," Stanford
Univ. Press, California.
Wolpe, J., 1981, The dichotomy between classical conditioned and
cognitively learned anxiety. J. Behav. Ther. & Exp. Psychiat.,
12:35. -
Wolpe, J., 1983, "The Practice of Behavior Therapy," 3rd edition,
Pergamon Press, New York.

854
CONDITIONING OF AUTONOMIC FUNCTIONS, SCHIZOKINESIS, AND
PSYCHOSOMATIC MEDICINEl

Samuel A. Corson and Elizabeth O'Leary Corson

Department of Psychiatry and Department of Educational
Theory and Practice
The Ohio State University
Columbus, Ohio 43210, USA

ABSTRACT
In Pavlovian conditioning experiments with aversive reinforce-
ment, the animal is faced with a no-solution problem and thus cannot
achieve a consummatory response. Gantt reported that in dogs, clas-
sical cardiac conditioning (to aversive stimuli) occurs more rapid-
ly and extinction occurs more slowly than conditional motor defense
responses. He referred to this phenomenon as schizokinesis. Such
somatovisceral dichotomy was observed by us in regard to many other
conditional visceral reactions, particularly in certain breeds of
dogs. Some dogs (e.g., wirehair fox terriers, border collies, Ger-
man shepherds, cocker spaniels, and some mongrels) exhibited highly
generalized, almost inextinguishable conditional tachycardia, polyp-
nea, profuse salivation, high energy metabolism, high urinary vaso-
pressin and catecholamines, and increased muscle tension. Other dogs
(most beagles and other hounds and some mongrels) exhibited very
little or only temporary schizokinesis. We postulate that individu-
als exhibiting marked and persistent schizokinesis may be considered
at risk for developing psychosomatic disorders. Suggestions for
possible prophylactic therapeutic interventions will be discussed.

lsupported in part by NIH, USPHS research grants HL 20861, MH

12089, MH 18098, The Ohio State University Graduate School Bio-
medical Science Research Support Grant, and the Psychiatric
Research Foundation of Columbus.
855
I. WHY STUDY CONDITIONING OF AUTONOt-11C FUNCTIONS?
Some three decades ago in the course of making renal hemodynam-
ics measurements on unanesthetized dogs, we were puzzled by the fact
that in some dogs in carefully controlled experiments we often could
not secure reproducible data. As well-trained reductionist physiolo-
gists, we made sure that we carefully controlled all chemical and
physical factors.
It slowly dawned on us that we had failed to control the psycho-
logic factors in our laboratory. Our experiments were conducted in a
laboratory shared with several other researchers. In the course of
the experiments, students and staff members would come and go, open
and close doors, talk freely, ask questions, run water in the sink,
answer a ringing telephone, etc.
During the same period we also came across another biologic
phenomenon which puzzled us. We were trying to measure blood volume
in some of our experimental dogs by the Evans blue dilution method.
This involved the intravenous injection of a nontoxic dye, Evans
blue, and subsequent determination of the blood volume by measuring
the concentration of the dye in a sample of blood plasma. \~hat
puzzled us was that in some dogs the plasma sample was so full of
chylomicrons (microscopic lipid particles) that it was impossible to
measure the dye concentration by spectroscopic methods. Since this
occurred in dogs which had not been fed for some 16-18 hours, the
source of lipids in the blood again suggested the possibility of
psychologic influences.
In our search for methods of investigating the influences of
psychologic factors on renal and other visceral functions, we came
to the conclusion that Pavlovian conditioning methods would lend
themselves to such investigations. Bykov and Kurtsin (1966) demon-
strated the usefulness of such methods in the study of gastrointes-
tinal functions. Gantt (1953) utilized such techniques in the study
of cardiac functions.

II. EXPERIMENTAL METHODS

The fact that different dogs ir1 our renal experiments exhibited
different sensitivities to psychologic factors suggested to us the
importance of investigating the role of constitutional differences
in psychophysiologic reactions. We also observed that some dogs
habituated rapidly to the psychologic variables in our laboratory,
whereas other dogs failed to habituate. We therefore decided to
conduct longitudinal studies on mongrels as well as on several pure
breeds of dogs in order to elucidate the interaction of genetic and
psychosocial factors in reactions of dogs to psychologically stress-
ful situations.
Our experimental design involved the study of the same dogs
856
repeatedly in 1) a neutral control room where baseline behavioral
and physiologic parameters could be recorded; and 2) a psychologic-
ally aversive room where neutral tones were reinforced with electro-
cutaneous stimuli, eventually leading to the development of Pavlovian
conditional motor defense responses.
The Pavlovian term "conditional motor defense reflex" is rather
misleading. The fact is that in a Pavlovian paradigm the conditional
leg lift response does not have any defense function, inasmuch as
lifting the leg does not prevent the application of the electric
stimulus. Therefore, the animal is not in a position to develop a
consummatory adaptive response. This is in contrast to Pavlovian ex-
periments with alimentary reinforcement where conditional salivation
does represent an adaptive reaction.
Pavlovian conditioning with electrocutaneous reinforcement is
essentially equivalent to a "no-solution" problem situation and may
therefore be conducive to the development of frustration, anxiety,
and neurotic manifestations. This is in contrast to electrocutaneous
reinforcement in an operant paradigm where the animals are permitted
to develop avoidance responses.
Detailed descriptions of our Pavlovian conditioning techniques
have been published (Corson, 1971; Corson and O'Leary Corson, 1976).
Therefore, only a brief outline of the methods will be presented
here. After a standard period of food and water withdrawal, the dogs
were given a water load (25 ml/kg) via stomach tube or gastric fis-
tula at the beginning of each experiment. The purpose of the stan-
dard water load was to enable us to measure, among other variables,
the influence of psychologic stress on the time course of a water
diuresis and on the osmolality and chemical composition of the urine.
The introduction of a large water load is equivalent to a temporary
diabetes insipidus. This enabled us to study the effects of emotion-
al stress on the dynamics of vasopressin release.
The Pavlovian conditioning experiments (with electrocutaneous
reinforcement) were conducted by a different experimenter in a con-
ditioning room that was distinctly different externally and inter-
nally from the neutral control room. As in the control room, the
experiments were of two hours duration and were composed of six
20-minute data collection periods.

III. GENETIC FACTORS IN SCHIZOKINESIS: PERSISTENT GENETICALLY IN-

FLUENCED DIFFERENCES IN PSYCHOPHYSIOLOGIC REACTIONS OF DOGS
TO AN AVERSIVE PAVLOVIAN CONDITIONING ROOM
We observed two major types of psychophysiologic (genetically
influenced) reactions to repeated exposure to the psychologically
stressful environment represented by the Pavlovian aversive condition-

857
ing room. Some dogs (e.g., beagles, most hounds, and some mongrels)
exhibited rapid psychophysiologic adaptation (high adaptation dogs
= HA). Other dogs (e.g., wirehair fox terriers, cocker spaniels, some
border collies, German shepherds, and some mongrels) exhibited a
persistent, almost inextinguishable integrated octet of psychophysio-
logic reactions to the entire conditioning room complex: tachycardia;
polypnea; profuse salivation; increases in energy metabolism, rectal
temperature, and electromyographic reactions; vasopressin release;
and increased urinary catecholamines (low adaptation dogs =LA).
In contrast to the conditional motor defense reflexes which
were stimulus-bound, well differentiated, and easily extinguished,
the psychophysiologic visceral reactions in LA dogs were highly
generalized and exhibited poor extinction. This phenomenon of somata-
visceral dichotomy is comparable to the principles of schizokinesis
reported by Gantt (1953) for differences in the rates of develop-
ment and extinction between conditional motor defense responses and
conditional cardiac reactions.
Our studies suggest that: a) schizokinesis is characteristic
primarily for LA dogs; and b) schizokinesis involves not only cardiac
reactions but a great variety of visceral responses. These observa-
tions have some relevance to the problem of the target organ in
psychosomatic medicine and to differences in susceptibility to psy-
chosomatic disorders in human subjects who may be exposed to similar
stressful situations.
The phenomenon of somatovisceral dichotomy or the discordance
between overt behavior and psychophysiologic reactions in our LA
dogs may be comparable to similar discordance described by Eliot and
Buell (1983) in human subjects. These authors reported that some
individuals are 11 hot 11 reactors, i.e., they exhibit exaggerated car-
diovascular and endocrine stress reactions, whereas other individu-
als are 11 coo,.. reactors, i.e., they exhibit normal cardiovascular
stress reactions.
We should emphasize at this point that the general overt be-
havior of our dogs (in the aversive conditioning room or in their
own kennels) cannot be used as a predictor of the type of psycho-
physiologic stress reactions these animals may exhibit. In many dogs
(LA or HA) we observed a discordance between the overt behavior and
the covert psychophysiologic stress reactions. Other dogs exhibited
a concordance between overt and covert reactions.
It is instructive that Eliot and Buell reported that in humans
classified as exhibiting types A orB overt behavior, some individu-
als may show concordance between overt behavior and covert psycho-
physiologic reactions, whereas others may exhibit discordance, i.e.,
a person could be externally a 11 cool 11 reactor and internally a "hot"
reactor, and vice versa.

858
 Our studies on dogs and those of Eliot and Buell on humans
suggest that the predictive power of types A and B classification
could be significantly enriched and strengthened by incorporating
psychophysiologic measurements is the classification of human
behavior repertoires.

REFERENCES
Bykov, K.~1., and Kurts in, I. T., 1966, "The Corti co viscera 1 Theory
of the Pathogenesis of Peptic Ulcer," S.A. Corson, trans-
lation ed., Pergamon Press, London, New York.
Corson, S.A., 1971, Pavlovian and operant conditioning techniques
in the study of psychosocial and biological relationships,
in: "Society, Stress and Disease," Vol. 1, L. Levi, erl.,
Oxford University Press, pp. 7-21.
Corson, S.A., and O'Leary Corson, E., 1976, Constitutional differ-
ences in physiologic adaptation to stress and distress, in:
"Psychopathology of Human Adaptation," Serban, ed., Plenum
Press, New York, pp. 77-94.
Eliot, R.S., and Buell, J.C., 1983, The role of the CNS in cardio-
vascular disorders, Hospital Practice, May 1983:189-199.
Gantt, W.H., 1953, Principles of nervous breakdown- schizokinesis
and autokinesis, Ann. N. Y. Acad. Sci. 56(2):143.

859
BIOFEEDBACK AND ANXIETY STATES

Melinee Agathon and Jean Bernard Mazel

Clinique des Maladies Mentales et de L'Encephale

(Pr. P. PICHOT) 100, rue de la Sante
75014 Paris, France

INTRODUCTION

Insofar there had not been much evidence that electromyographic

biofeedback (EMG BFB) has an effect on general anxiety in non-clini-
cal populations such as dental phobics (Miller et al., 1975) in which
various methods of relaxation have the same effect as BFB on the
sedation of anxiety, flight phobics (Reeves and Mealia, 1975) and
test anxious students (Romano and Cabianca, 1978). In controlled
studies BFB did not reduce anxiety more then did relaxation (Leboeuf
and Lodge, 1980) although it had more effect than placebo. Burish
(1981) comes to the conclusion that EMG BFB is an alternate, not a
superior, method of treating people with anxiety disorders and that
it is not demonstrated that its effect is superior to placebo.

In psychiatric patients, especially in anxiety states in which

anxiety leads sometimes to panic attacks and for which pharmacological
treatment, relaxation training and/or psychotherapy were previously
tried unsuccessfully, biofeedback has been used since the last ten
years. Its effect is nethertheless difficult to estimate because
systematic measures of physiological level are not reported in most
studies and also because some kind of behavior therapy, e.g. syste-
matic desensitization, is frequently at the same time and the spe-
cific effect of BFB remains unclear.

In our laboratory, after we demonstrated that operant condi-

tioning of heart rate could be obtained in a control group of adults
(Agathon, 1973), heart rate was tried to reduce crises in an anxiety
neurosis (Agathon, 1977). The fact that the patient could be made
aware of the slowing of this cardiac activity seemed to have a be-
neficial effect. Cardiac activity was reduced from 100 to 80 beats

861
per minute with biofeedback. During follow-up, he did not report any
more anxiety crisis. Electromyographic biofeedback (EMG BFB) was also
found efficient in a 39 years female patient suffering since age 22
anxiety neurosis (Agathon et al., 1979).

EXPERIMENTAL STUDY

The present study investigates the effect of EMG BFB on patients

with panic attacks.

Apparatus

EMG biofeedback is given in a sound-proof darkened room. The

patient, resting alone, is seated in a reclined armchair. Three
silver electrodes are placed on his forehead, two of them on the
frontalis muscle, 1,5 em above eyebrows, the neutral electrode being
between them. They are connected to the BFB apparatus (EMG 120)
which filters and amplifies the EMG signal and translates it into
an auditory feedback consisting of clicks. The frequency of these
clicks is proportional to the EMG activity. They reach the experi-
mental room by mean of loudspeaker.

The EMG BFB apparatus is completed by a digital integrator which

provides digital readout of the average value of the EMG activity
calculated on periods of 30 seconds. A printing counter is relied to
the integrator.

Patients are asked to listen to the clicks in order to be aware

of their muscular activity and try to reduce their frequency.

Experiment

Previous to biofeedback, each patient has, with the behavior

therapist, two interviews. During these interviews, behavior analysis
is done. Then, the patient is informed about hypothesis on the con-
ditioning of physiological reactions of anxiety in panic attacks.
Some explanations on biofeedback are also given. It is then suggested
that he would have 10 EMG BFB sessions.

In each session, after a few minutes of rest, baseline is

recorded for 5 minutes (phase I). In phase II, feedback (acoustical
signal) is available for the patient during 10 minutes. It is
supressed in phase III (5 minutes) which is presented as a rest
period, and given during the last phase which lasts 10 minutes.
Before and after session, the patient has a few minutes interview
with the therapist during which he describes his present state but
no desensitization for his specific symptoms is tried. Between the
weekly sessions, he is asked to practice every day 15 minutes rest,
trying to imagine his physiological state during BFB.

862
Patients

18 of the patients referred to us for strong anxiety states

were treated by EMG BFB without adjunction of desensitization nor
of any classical technique of relaxation.

Patients (10 males, 8 females) were aged 24 to 71 (mean age

37,22, sigma 12,01). They were outpatients except for a woman who
had to be hospitalized because she lived far away from the hospital.

Diagnosis included, according to the French I.N.S.E.R.M.*

classification 5 anxiety neuroses
2 hysterical neuroses
3 phobic neurones
obsessionnal neuroses
4 unclassified neuroses
2 manic depressive patients
hebephrenic schizophrenic patient

All patients had, once a day or more frequently, panic attacks

with chronic anxiety since more than a year. Previously they had
been given various pharmacological or psychotherapeutic treatments.
These treatments, when prescribed, were not discontinued.

Criteria

The effect of biofeedback was estimated by comparing the mean

values of the level of muscular tension in frontalis in each of the
4 phases of sessions. When Ss acquire BFB learning, this level is
lowered with introduction of acoustic feedback (clicks). Slowering
the rate of clicks can be considered as a positive reinforcement for
a patient informed that it might reduce his anxiety level. After
acquisition of this learning, the effect generalizes in the no-feed-
back period.

One can consider that EMG BFB learning is acquired when the
level of frontalis tension keeps lowered without these reinforcements.

RESULTS

EMG-BFB

11 patients showed evidence of BFB learning as defined in the

above criteria : the mean level of muscular tension had lowered from
11 to 40% of its initial value.
4 patients did not show any evidence of this learning.
Results were unclear for the 3 other patients.

* I.N.S.E.R.M. : Institut National de la Sant~ et de la Recherche

M~dicale

863
Clinical evolution

The clinical state at the end of the sessions was rated by the
psychiatrist on a 5 points scale : no symptom, improvement, no change,
agravation, severe agravation.
6 patients had "no symptom"
7 patients were "improved"
5 patients were "stationary"
No patient was rated "agravated" or "severely agravated''

Correlation between BFB and clinical rating

5 of the 6 patients rated "no symptom" had acquired the BFB

training. So did 5 of the "improved" patients and 2 of the "statio-
nary" ones. As expected there was neither BFB acquisition nor clini-
cal change in the hebephrenic schizophrenic patient.

In summary, the diminution of the frontalis EMG activity and

the BFB acquisition are related to clinical evolution in 69 % of the
cases.

Anxiety level

Before BFB all patients had a high anxiety level score, above
7 (global score) at Cattel's scale (IPAT Anxiety Scale).

The score was lowered after the sessions for 44 % of the pa-
tients who showed clinical improvement. It was lowered for only
25 % of the patients with stationary clinical state.
This mesure of anxiety agreed significantly with the estimation
of clinical state (r = .52, p = .02).

DISCUSSION

Follow-up studies, 6 to 20 months after the BFB sessions,

showed no change in clinical improvement. These results are in
agreement with previous experiments made in order to compare EMG
BFB with progressive relaxation : for Canter et al. (1975) 12 to 14
patients (85 %) improved with EMG BFB and only 50 % with progressive
relaxation whilst Leboeuf and Lodge (1980) found 8 to 13 patients
(61 %) improved with BFB as well as with progressive relaxation.
Their data showed that EMG BFB was much more effective in reducing
frontalis EMG activity than relaxation.

Clinical results might explained by the different diagnoses

of the patients : Canter has more improvement in patients with
anxiety and panic attacks than with patients who do no complain of
such attacks. Leboeuf and Lodge carried their study on patients with
generalized anxiety. It seems that BFB gives better results for
patients with generalized anxiety states and panic attacks corres-

864
pending to the French diagnosis of "n~vrose d'angoisse". In the pre-
sent study, four of the five patients with this diagnosis are evalu-
ated by the psychiatrist as free of symptom, the fifth patient being
also clinically improved although he had not met our criteria of
BFB acquisition.

In all cases, it is not clear whether ratings of anxiety made

either by the patient e.g. with Cattell's scale or with a state-trait
anxiety inventory (Towsend et al., 1975) or by a psychiatrist with
Hamilton's scale (e.g. in Leboeuf and Lodge's study) are good mea-
sures of the physiological states reported by the patient in anxiety
states.

The high level of muscular tension which is often concomitant

with anxiety states is considered in EMG BFB as a physiologically
learned response of the organism. It can be unlearned when the pa-
tient becomes informed, through feedback provided by the electronic
device, when this level decreases. As Yates (1980) states, feedback
has a main role of information and of secondary reinforcement. Many
cognitive and motivational factors are necessary for clinical use of
biofeedback training : the patient must understand what Yates calls
the "servomechanical model of biofeedback control". He must be moti-
vated for obtaining the secondary reinforcement and the anxious pa-
tient, when he has no other explanation of panic attacks and no
other mean to control them, is more likely to seek help in a device
offering his some mean of regulating his physiological reactions,
whether cardiac or muscular.

Still, as most research workers have already pointed out, mecha-

nism of biofeedback remains unclear and effects of biofeedback, alone,
on anxiety state are not yet demonstrated. But, once more, researches
on conditioning have proved, in psychopathology, to be an experimental
model for the study of higher nervous activity. With no doubt, bio-
feedback derives from "the scientific fields of psychopathology,
learning theory and the experimental basic analysis of behavior"
(Katkin and Goldband, 1979). Nowadays there is some evidence that
by reinforcement procedures it is possible to modify or control some
autonomic responses.

BIBLIOGRAPHY

Agathon, M., Delcros, M., Mazel, J.B. and Ruschel, S., 1979, Un cas
d'angoisse trait~ par la r~troaction ~lectromyographique,
J. Ther. Comportement. Lang. Franc;aise, I, 2: 107.
Agathon, M., Olivier-Martin, R., Kaprinis, G. and Pichot, P., 1977,
Tentative d'application de la r~troaction biologique dans le
traitement d'une n~vrose d'angoisse, Ann. m~d.-pychol.,
I, 2 : 255.

865
Agathon, M., and Roussel, A., 1973, Etude preliminaire sur le condi-
tionnement operant cardiaque. Son interet en psychopathologie,
Psychol. Med., 5, 7: 1377.
Burish, T.G., 1981, EMG biofeedback in the treatment of stress-
related disorders, in : "Medical Psychology", C.K. Prokop,
and L.A. Bradley, ed., Academic Press, New-York.
Canter, A., Kondo, J.R., and Knott, J.R., 1975, A comparison of EMG
feedback and progressive relaxation training in anxiety
neurosis, Brit. J. Psychiat., 127 : 470.
Katkin, E.S., and Goldband, S., 1979, The placebo effect and bio-
feedback, in : "Clinical application of biofeedback : appraisal
and status", R.J. Gatchel, and K.P. Price, ed., Pergamon Press,
New-York.
Leboeuf, A., and Lodge, J., 1980, A comparison of frontalis EMG bio-
feedback training and progressive relaxation in the treatment
of chronic anxiety, Brit. J. Psychiat., 137 : 279.
Miller, M.P., Murphy, P.J., and Miller, T.P., 1978, Comparison of
electromyographic feedback and progressive relaxation training
in treating circumscribed anxiety stress reactions, J. Consult.
Clin. Psychol., 46 : 1291.
Reeves, J.L., and Mealia, W.I., 1975, Biofeedback-assisted and con-
trolled relaxation for the treatment of flight phobics,
J. behav. Ther. exp. Psychiat., 6 : 105.
Romano, J.G., and Cabianca, W.A., 1978, EMG biofeedback training
versus desensitization for test anxiety reduction, J. Consult.
Psychol., 25 : 8.
Stoyva, J.M., 1979, Guidelines in the training of general relaxation,
in : "Biofeedback : principles and practice", J.V. Basmadjan,
ed., Williams and Wilkins, Baltimore.
Townsend, R.E., House, J.R., and Addario, D., 1975, A comparison of
biofeedback-mediated relaxation and group therapy in the
treatment of chronic anxiety, Amer. J. Psychiat., 132, 6 :
598.
Yates, A.J., 1980, "Biofeedback and the modification of behavior",
Plenum Press, New-York.

866
BIOFEEDBACK TREATMENT IN PSYCHOSOMATIC ILLNESS

Daniel E. Sheer

Department of Psychology
University of Houston
Houston, Texas, USA

In spite of the fact that the field of biofeedback is now well

into its second decade, few studies have documented the long-term
effectivenes s for reducing clinical symptomatolo gy in psychosomati c
illness. Integration of those that have been conducted is compli-
cated by the varying treatment modalities, lengths of training, and
lengths of follow-up periods employed by the investigator s.
Moreover, the definition of "success" is necessarily an arbitrary
and subjective one, susceptible to the influence of both the
patient's and the evaluator's expectations . Recognizing these
potential sources of error exist, one can nevertheless make some
statements regarding long-term effectivenes s.

Overall, the percentages of successfully treated patients

appear to exceed that which might be attributed to chance or pro-
longed placebo effects. Patients with vascular headaches or
primary idiopathic Raynaud's disease were more often successful,
with rates generally greater than 70%. Patients with mixed head-
aches did slightly less well (about 60% successful), and patients
with tension or muscle contraction headaches showed significant
improvement among roughly 50% of the cases. Long-term results for
a relatively smaller number of patients with the irritable bowel
syndrome and secondary Raynaud's phenomenon showed success rates of
roughly 60% and 40%, respectively . Biofeedback treatment was least
effective for patients with essential hypertension , where generally
only slight decreases in diastolic blood pressure were revealed at
follow-up. Some patients, however, were able to decrease the amount
of antihyperten sive medication following training.

867
 Since a number of the reported training procedures combined
relaxation techniques with biofeedback, it was not always possible
to evaluate the effectiveness of biofeedback training per se.
Several studies, however, directly compared biofeedback versus other
forms of training. The results consistently showed that biofeedback
techniques were equivalent to other forms of relaxation training for
treatment of tension headache, vascular headache, and primary
Raynaud's disease. In the one instance where different biofeedback
modalities were compared for treatment of vascular headache (Cohen,
et al., 1980), all were found to be equivalent, although the success
rates with any of the pure biofeedback conditions were apparently
lower than others reported. For treatment of hypertension, blood
pressure or EMG feedback was found to be equivalent to relaxation
training. In no instance was biofeedback, either alone or in com-
bination with relaxation techniques, shown to be more effective than
the relaxation techniques alone. An implication is that there are
factors in operation which transcend the mode of treatment, such as
the patient's expectations for relief, the acquisition of general
relaxation skills, and the cognitive restructuring whereby the
perceived sense of self-control is developed. The primary advantage
of biofeedback is that it is congruent with the belief structures
of the technologically minded sector of the population which may be
resistant to other treatment modalities, such as traditional
medicine, hypnosis, and psychotherapy. It provides an effective
treatment alternative for many of those people who might otherwise
not seek relief, or who have become dissatisfied with other
approaches they have tried.

Studies in which biofeedback alone has been compared to

relaxation alone have offered no clear-cut evidence that biofeedback
has an advantage over relaxation training. It has been suggested
that biofeedback is an elaborate means of teaching relaxation. In
terms of cost-effectiveness of treatment, this is a critical issue
for the practicing clinician. The combination of relaxation and
biofeedback may provide the patient with more effective treatment,
but the interrelationships are as yet unclear and may be unique to
the individual.

Non-specific factors other than relaxation have been implicated

in biofeedback treatment. The majority of outcome investigations
have focused on the treatment of headache, and these may be sub-
divided with respect to type of headache--muscle contraction or
vascular. A line of evidence against a modality-specific model of
biofeedback treatment for headache are those studies which have
found that both muscle-contraction and vascular or migraine head-
aches respond well to both EMG and temperature biofeedback training.
Coping strategies based on Meichenbaum's work (1977) implicating
a shift in cognitive self-statements from a helplessness stance to
a sense of self-control have been investigated as the active element
in the biofeedback process.

868
 In an elegant demonstration of the effectiveness of the
single-case study design, Kremsdorf, Kochanowicz and Castell (1981)
demonstrated in two single-case studies that cognitive coping skills
training was the active component in reducing headache activity in
a coping skills-biofeedback treatment package. The design for the
first subject was A-C-CB-C with A representing the baseline re-
cording period, B representing biofeedback training, and C
representing cognitive skills training. The design for the second
subject was A-B-BC-B-BC. EMG level and hourly headache activity
were the dependent variables. In the first study, cognitive skills
training alone resulted in diminished headache activity although
EMG levels remained high. A combination of EMG training and
cognitive skills training continued to diminish headache activity
and lowered EMG levels. When EMG training was eliminated, headache
activity continued to diminish while EMG levels rose slightly. In
the second design, EMG training alone resulted in lowered EMG
levels but did not diminish headache activity. A combination of
EMG training and cognitive skills training resulted in lowered EMG
levels and diminished headache activity. When cognitive skills
training was removed, headache activity increased slightly while
EMG levels remained low. When the combination treatment was re-
instituted, both dependent variables were reduced to lowest levels.
These studies again emphasize the conclusion that cognitive
variables are a significant component of biofeedback training. A
combination of biofeedback and cognitive coping skills training may
well be the most effective treatment package.

Identification of the active components in the biofeedback

training procedure is but one of several issues that the biofeedback
clinician needs to consider in the development of an effective
biofeedback treatment package. Other treatment issues include the
role of home practice, how many sessions are necessary, the interval
between sessions, and the need for follow-up visits.

Of the above variables, the role of home practice has been

most often investigated. Since the first studies of biofeedback,
home practice has been implicated as an important factor in
maintaining success in biofeedback training. In her review,
Tarler-Benlolo (1978), concluded that although it cannot be
determined whether home practice is critical because it cannot be
controlled experimentally, there is convincing evidence that home
practice may play an important role in maintaining results.
Although the above evidence indicates that it would be unwise for
future therapy or research programs to ignore the variable of
continuing practice, the nature of the relationship between home
practice and treatment success remains unclear.

Another treatment parameter known to vary widely is the number

of treatment sessions required per individual. The length of
treatment is related not only to the number of sessions, but also

869
to the between-session interval. Most biofeedback programs are
conducted on an out-patient basis with the patient having two or
three sessions per week. However, due to the need to establish
a program for out-of-town patients or to provide service to those
whose medical insurance will only cover in-patient hospital
expenses, several short but intensive pJograms have been developed
across the country. The effectiveness of these programs has only
begun to be investigated.

Caution must be advised to the clinician who develops a

clinical biofeedback program based on the experimental biofeedback
literature as it exists today. The lack of follow-up studies in
this area is a serious deterrent to drawing inferences from experi-
mental procedures. Biofeedback treatment success on a short-term
basis has been attributed to placebo effect and regression toward
the mean. No matter what the technique used, follow-up studies
are necessary for firm conclusions about treatment effectiveness.

References

Cohen, M. J., McArthur, D. L., and Rickles, W. H., 1980,

Comparison of four biofeedback treatments for migraine
headache: Physiological and headache variables,
Psychosomatic Medicine, 42(5), 463-479.
Kremsdorf, R. B., Lochanoweiz, N. A., and Castell, S., 1981, Cognitive
skills training vs. EMG biofeedback in the treatment of
tension headache, BFB Self Reg., 6(1): 93-101.
Meichenbaum, D., Cognitive-Behavior Modification--An Integrative
Approach, 1977, Putnam Press, New York.
Tarler-Benlolo,' 1., 1978, The role of relaxation in biofeedback
training: A critical review of the literature, Psych. Bull,
85: 727-755.

870
EXTINCTION FAILURE IN CLASSICAL CONDITIONING AS A MECHANISM OF

PSYCHOSOMATIC ILLNESS

Archie Levey*, Irene Martin**, Robert Blizard** and

Matthew Cobb**

*MRC Applied Psychology Unit, Cambridge CB2

**Institute of Psychiatry
De Crespigny Park, Denmark Hill, London SE5 8AF

BACKGROUND
Momentary phasic alterations of visceral activity are a normal
accompaniment of external stimulation. A particular organ system
which is hyper-reactive to stimulation will readily become condition-
ed to the contexts and antecedents which are associated with that
stimulation. If the resultant conditioned responding persists,
that is if it fails to extinguish, the organ system is at risk for
psychophysiological disorder.

These three propositions form the basis for a preliminary

model of psychosomatic illness based on classical conditioning
principles (Levey and Martin, 1981). Our previous work has
identified a small group of individuals who failed to extinguish
the conditioned eyelid response (32 of 144 normal subjects). When
compared with subjects who conditioned to the same level in
acquisition but who extinguished normally they were found to be
hyper-reactive in the eyelid effector system but low on measures
of general arousal, namely psychogalvanic activity, body tempera-
ture and subjective measures of boredom (Martin and Levey, 1980).
They also complained of physical symptoms consistent with psycho-
somatic disorder (Levey, 1981). These findings lead us to suggest
that low general arousal may be an important determinant of extinc-
tion failure when it forms the setting for hyperactivity in one or
more organ systems. The present paper reports some preliminary
results of a large scale study designed to test this prediction
and the model underlying it.

AIMS
The current study aims to extend our previous findings, based
on eyelid activity, to response measures which have greater relevance

871
to psychosomatic disorder, viz: respiration, gastric motility,
pulse volume and heart rate. It aims to identify the subgroup of
subjects who fail to extinguish and to compare them with subjects
who extinguish normally, usiPg measures of general arousal and of
health status. In preparation for the main investigation a pilot
study was undertaken to assess the feasibility of the main study
and this paper reports results of an exploratory analysis of that
study.

METHODS
Twenty-five young adult males, recruited by newspaper adver-
tisement and screened for psychiatric disorder were tested using the
following schedule of reinforced (R) and unreinforced (U) trials:
R-R-U-R-U-U-U-U. The conditioned stimulus was a 60db tone of 30
sec. duration. The unconditioned stimulus was a tone of 110 db
which overlapped the termination of the CS. The inter-trial
interval averaged 45 sees. Physiological measures, recorded on
an Elema Schonander Mingograph, included respiratory cycle moni-
tored by a thoracic strain gauge, electrocardiogram, pulse volume
monitored by a digital plethysmograph, gastric motility monitored
from surface electrodes, and galvanic skin response (GSR). Subjects
completed the Spielberger State Anxiety Scale and the Lader Mood
Sc~le before and after the experimental session. This simple
conditioning schedule was used to examine acquisition and extinction
parameters in the four response systems.

Since one purpose of the pilot study was to familiarize our-

selves with the response characteristics of each of the systems
and to determine optimum time intervals for automatic scoring, the
chart records were hand-scored by the two senior authors (AL,IM)
using the customary division into first, second and third responses.
Four categories of response were scored on the following simple
scale: no response (scored 0), doubtful or minimal response (1),
average response (2), conspicuous or exaggerated response (3). A
set of six records was scored jointly in order to standardise the
categories and intervals for each system. The records were then
divided at random and scored individually to yield scores for each
trial in each of the three time intervals corresponding to the
orienting response (OR), in the first interval, the anticipatory
conditioned response (CR) in the second interval and the uncondition-
ed response (UCR) in the third interval. On unreinforced trials the
third interval response was scored by the same criteria as on
reinforced trials.

Scoring was based on visual estimation of the ongoing inter-

trial activity and represented departures from it. Arrows were
used to code responses in terms of increased or decreased activity
in the response systems other than GSR. This was particularly
necessary in the case of the electrogastrogram (EGG) for which
responses in any of the three intervals represented changes in the

872
normal rhythmic activity of the gut. Similarly, changes in
respiratory cycle were scored as increase or decrease in rate
and/or volume. This scoring system, while apparently crude,
enabled us to achieve high inter-scorer agreement, to establish
the feasibility of the 30 sec. CS interval and to identify response
characteristics in each of the three intervals.

RESULTS
All subjects showed evidence of conditioning in at least one
response system and each of the response systems produced condition-
ing. Of the visceral systems, heart-rate showed the highest level
of conditioned responding across subjects and respiration the lowest
but the differences were not large. There was a tendency for those
subjects who demonstrated GSR conditioning to produce fewer condi-
tioned visceral responses, and no subject conditioned in all five
systems.

The aim of the analysis was to identify those subjects who

failed to extinguish in order to compare results with the previous
eyelid conditioning study. Our preliminary model does not deal
with subjects who respond differently in different systems. We
propose that the category of extinction failure represents a typ-
ology which identifies individuals for whom failure of extinction
is characteristic, and who are consequently at risk for psycho-
physiological disorder. In the present study about one third of
the subjects showed a mixed pattern of extinction and non-extinc-
tion in separate systems.

In order to identify the "true" non-extinguishers within our

sample the following procedure was adopted. Within each of the
visceral response systems subjects were arranged in order of the
number of CRs in acquisition, ignoring the direction of responding.
Those who had conditioned in acquisition but who gave no response
on the last three trials were designated as extinguishers. Each
of these was paired with a subject who had given the same number
of responses in acquisition but who had responded on the last trial
or on the second last if no match were found. In most cases the
non-extinguishers had responded on both of the last trials. The
four systems were then compared and subjects who had consistently
extinguished were placed in the extinction (EXT) subgroup. Subjects
who had consistently failed to extinguish (N=5) were placed in the
non-extinction group (NON). Subjects who appeared on both lists
(N=8) were placed in a mixed group (MIX). The remaining subjects
(N=6) who had conditioned too weakly to meet either criterion were
placed in a non-conditioning residual group (NCR).

The purpose of this rather elaborate procedure was to identify

unequivocally two groups having characteristic tendencies to
extinguish rapidly or to fail to extinguish and to keep them
separate from those who showed a mixed performance. The two

873
criterion groups thus formed showed roughly equal distributions across
each of the response systems, as well as being equally distributed
in terms of number of CRs in acquisition within response systems.
We regard this symmetry of distribution as one of the important
findings of the pilot study since it shows that the dimension of
extinction failure is not determined by specific response systems
or by level of conditioning but represents some individual difference
parameter. We also found it interesting that the number of subjects
(N~5) in the EXT and NON groups fell remarkably close to the figure
of 22% found in the original eyelid conditioning study.

Across all systems the magnitude of the third interval response,

the UCR, to trial 1 predicted the level of subsequent conditioning
significantly. This finding, which is familiar in the condition-
ing literature, confirms the first prediction of our model: that
system reactivity is positively related to strength of conditioning.
For confirmation of the second prediction, that low general arousal
is conducive to extinction failure we compared the criterion groups
on GSR performance. Total magnitude of responding in each interval
was used as the index of GSR reactivity with the following results.
For the first interval response the mean for NON and EXT groups
were 5.4 and 12.0 respectively (F J.,8 =' 6.35) significantly confirm-
ing the prediction that non-extinguishers are lower in arousal.
For the second interval response the means of 3. 0 and 8. 2 ( F 1, 8 '"
8.89) carried the same implication. The difference between the
two groups for the third interval response was in the same direction
but not significant.

The role of anxiety in facilitating conditioning has often

been discussed. The two criterion groups, NON and EXT showed
significantly higher levels on the Spielberger scale pre-test
(42.0 and 44.8) than the two remaining groups, MIX and NCR (34.4
and 29.0) but did not differ from one another. This is consistent
with the fact that both the NON and EXT groups conditioned well
in acquisition. Interestingly, the anxiety level measured post-
test did not differ significantly among groups partly because the
NON group had increased in anxiety ( -1-4.4) while the EXT group had
slightly decreased (-1.4).

The individual response systems did not show clear cut results
in terms of extinction and non-extinction and this supports our
view that consistent failure of extinction is a characteristic of
individuals rather than of systems. However, there were some
interesting reciprocities among systems. Individuals who showed
decelerative orienting responses in heart rate, for example, were
significantly less likely to condition in the gastric system.
Those whose first interval response was accelerative (N=6) were
significantly more likely to fall in the non-conditioning group.
Conversely a decrease in respiratory rate and volume, as either the
CR or UCH, decreased the likelihood of conditioning heart rate

874
though this trend did not reach significance. The present set of
preliminary data do not comfortably support analyses of this type
however, and these and other questions will be further explored
in the analyses of the main study.

DISCUSSION
This study, though concerned with small numbers and employing
severely restrictive criteria to identify the EXT and NON groups,
showed rather striking parallels with the original eyelid study
in which the groups were first identified. We feel justified in
concluding that the conditioning model of psychosomatic disorder
gains some support. While the third prediction, that non-extinguishers
are at risk for psychophysiological disorder was not directly tested,
it follows from the persistent maladaptive responding which extinc-
tion failure implies. This is being tested by health questionnaires
in the main study.

The fact that no subject failed to condition in at least one

system, though some conditioned only weakly, raises a question
which has not been sufficiently considered in the conditioning
literature. If the present study had been concerned only with
heart rate we would have concluded that an appreciable number of
our subjects had "failed" to "condition". Whatever reason we
might have found to account for this "failure" would have been
misleading if it devolved on the conclusion that the individuals
themselves had not conditioned rather than the conclusion that the
cardiac system had failed to demonstrate the conditioning which had
manifestly occurred. Clearly the concept of specific organ system
reactivity is useful in approaching this problem.

However, it raises another issue which is relevant to our

formulation, The recent literature in both animal and human studies
is increasingly sympathetic to an information processing view of
conditioning consistent with the swing towards cognitive models of
behaviour. In these terms, the example just given should be re-
phrased to say that the fact that conditioning occurred at some
level in every subject is evidence that the stimulus information
was adequately processed, even though it was not reflected in
every organ system. These and other considerations have led us
to re-evaluate the role of general arousal in our model. We
originally proposed that a low level of general arousal is
conducive to extinction failure. We now suggest that a low level
of information processing is consistent with a low level of arousal
and that this factor may account for failure of extinction. The
non-extinguisher responds as if he has not received or has failed
to process the information that the CS-UCS contingency has been
altered.

Three scraps of evidence in the data support this view.

l) The fact that the NON group increased in anxiety while the EXT

875
group decreased suggests that they were processing less efficiently
the information that would reassure them that the highly noxious
loud tone had been withdrawn. 2)The fact that the third interval
GSR to the UCS did not distinguish between the groups suggests that
the discriminating responses to the CS, both orienting and anticipa-
tory, reflect active processing of the CS information. 3) If
information processing is involved in detecting the change in
CS-UCS contingency then the UCS omission response should occur
earlier in the EXT than in NON group. Examination of the GSR
third interval response to the first three unreinforced trials
(3, 5 and 6), showed this to be the case though the trend would
require larger numbers of achieve significance.

Paralleling the swing toward cognitive models of behaviour the

recent literature in two different areas, relaxation/meditation
and stress response studies has turned toward the concept of
anticipatory arousal and has suggested that the criterion of
adaptive responding is not the initial level of activation response
to threat of stress but the facility with which the activation level
returns to normal limits (e.g. Burchfield, 1979; Christie et al,
1979; Lehrer et al., 1980). This alternative language bears a
close resemblance to the acquisition-extinction formulation which
we propose. Indeed, a similar notion was expressed more than a
decade ago by Roessler in his studies of ego strength (Es) and
psychophysiology: aEs persons are more responsive in skin conduct-
ance to complex stimuli of high symbolic value as well as to simple
ones . . . (they) appear to be persons who are pervasively more
responsive to their environment and to changes in it." (Roessler
et al., 1970, p. 738).

We might ask, therefore, if there is any advantage in retaining

the concepts and terminology of the conditioning laboratory. We
would suggest that there is a considerable advantage not only in
terms of precision but in terms of the relevance of conditioning to
psychosomatic illness. The two important questions in this area are:
Why ~his particular target individual? and Why this particular target
organ? While the cognitive models may answer the first of these
questions, only the conditioning model, coupled with the concept
of specific organ system reactivity addresses itself to the second.
It is one of the strengths of the main study that it is being
conducted by investigators trained in the methods of behavioral
genetics (RB,MC), using a large sample of twin pairs, balanced for
zygosity, in order to examine the origin of differences in
reactivity, whether constitutional or acquired. On this basis
we hope in the near future to provide interesting answers to
some of the problems broached in this preliminary study.

REFERENCES

Burchfield, S.R., 1979, The stress response: a new perspective,

Psychosom. Med., 41: 661.

876
Christie, M.H. and McBrearty, E.M., 1979, Stress response and
recovery, in: "Response to Stress: Occupational Reports",
C. MacKay and T. Cox, Eds., International Publishing Company,
London.
Lehrer, P.M., Schoickett, S., Carrington, P. and Woolfolk, R., 1980,
Psychophysiological and cognitive responses to stressful
stimuli in subjects practising progressive relaxation and
clinically standardised meditation, Behav. Res. and Therap.,
18: 293.
Levey, A.B., 1981, Learning theory and psychosomatic disorder,
J. Psychosom. Res., 25: 475.
Levey, A.B. and Martin, Irene, 1981, The relevance of classical
conditioning to psychosomatic disorder, in: "Foundations
of Psychosomatics", M.J. Christie and P. Mellet, Eds.,
John Wiley and Sons, Chichester.
Martin, Irene and Levey, A.B., 1980, Failure of extinction in
classical conditioning, in: "Functional States of the Brain:
Their Determinants", M. Koukkou, D. Lehman, and J. Angst,
Eds., Elsevier, Amsterdam.
Roessler, R. and Collins, F., 1970, Personality correlates of
physiological responses to motion pictures, Psychophysiol.,
6: 732.

877
COMPUTER ASSISTED EEG - ANALYSIS AS AN APPROACH TO THE

EXPLORATION OF MENTAL PROCESSES

Helmut Pockberger, Peter Rappelsberger and

Hellmuth Petsche

Institut fUr Neurophysiologie

Universitat Wien
A-1090 Wien, AUSTRIA

INTRODUCTION

Since HANS BERGER's ingenious discovery of the

EEG, a vigorous exploration of mental processes by
measuring the brain's electrical activity has been
started. However, before the computer era little has
been achieved, mainly because the multitude of data,
recorded during single EEG sessions, could not be
handled. The problem of a proper quantification of
EEG activity by measuring amplitude and phase relation-
ships of the Alpha-activity between different locations
on the scalp has been approached by several authors
(particularly by Motokawa) even before the introduction
of computers. Their studies however had been largely
neglected since.

With the introduction of computers for averaging

transients a new era began: instead of analysing spon-
tanous EEG activity event-related-potent ials have been
thought to be one clue to a better understanding of
higher mental activities. Apart from this main stream
in EEG research, Livanov (1977) has studied the quant-
ification of spontanous EEG activities. He demonstrated
by correlation analysis that spontanous EEG activities
change during different states of mental activity.

879
 During the last 10 years a quantification of
spontanous EEG activities has been reconsidered as an
appropriate method in psycho-physiology. Several authors
gave some hints that spectral analysis of spontanous
EEG activities during various tasks show remarkable
differences, thus indicating the close connections
between neocortical functioning and electrical activity
(Sklar et al. 1972, Busk and Galbraith 1975, Shaw et
al. 1976, Grabow et al. 1979).

In the following, the approach of our group to the

exploration of higher mental activities by means of
computer - assisted EEG analysis is briefly described.

METHODS

Since topographical aspects of EEG activities are

considered as most important for the exploration of
neocortical activities, the full set of 19 scalp elect-
rodes according to the international 10-20 system is
considered indispensible for such studies. All record-
ings are made against linked mastoid electrodes. The
19 EEG signals are amplified (TC 0.1, FB 30~ by a
Schwarzer 32-channel EEG machine, written out on paper
and simultanously stored on a 19-channel analogue tape
for off-line processing in a digital computer (HP 1000
F- processor).

After visual inspection, up to 15 epochs of two

seconds duration are chosen for a Fast Fourier Trans-
formation (FFT). In a first step of our analysis the
topographical distribution of power spectra is evaluated.
The second step consists in the computation of squared
coherence spectra for EEG signals, recorded from homo-
topic areas (interhemispheric coherence spectra) and
adjacent areas (intrahemispheric coherence spectra).
Furthermore confidence intervals are computed to find
out statistical significant differences between coher-
ences obtained during various mental activities and
rest.

RESULTS AND DISCUSSION

The examples given below show some typical obser-

vations, made in healthy and cooperative volunteers.
The first figure shows the power distribution during
three different conditions: the upper row demonstrates
the power distribution of different frequency bands of
an EEG record taken during rest and eyes closed. •ryp-
ically highest power is seen in the Alpha-band in the

880
....
..........

Fig. 1 Distribution of power in different frequency

bands during rest- eyes closed (upper row),
rest - eyes open (middle row) and reading (lower
row). Each line connects locations o f equal power
- contour-line spacing 10 ~V2 •

occipital regions. The power distributions shown in the

middle row demonstrate the changes, taking place as soon
as the volunteer opens the eyes. The lowe st row finally
shows the power distributi o n s during reading. Again
significant changes in the power distibution of the
EEG during "eyes open - rest" and "reading" are observed.
However, we observed that power distributions in any
frequency band of the EEG may considerably differ from
one to the o the r person. This is due to the different
thi ckness of the skull on the various neocortical a r eas,
so that the EEG signals are attanuated heterogenously.
This applies especially for t he higher frequency bands
(Beta activity). Therefore this example cannot be con-
sidered a r epresentat i ve distributi on for all v ol unteers
examined.

Coherence spectra can b e interpreted as a measure

o f " info rmat i on transfer" between different neocortical
r e gions . Thus low cohe rence betwe en two EEG signa ls from

881
different neocortical areas may be interpreted as a low
"information transfer" between these two areas. In con-
trast high coherence indicates a strong interplay betwe-
en these two regions.

Figure 2 shows the intra- and interhemispheric

coherence spectra for two different mental states. The
solid lines represent the coherence spectra for the EEG

H. P. 160583/14-15
EYES CLOSED RELAXED - LISTENING TO MUSIC

L
ct::t::t:
LEH
b·t:·b·lci RIGHT

COI-IU2
btJct:
B 2.4 HZ t:

t: h b

LL
!6/
' 1&,. I....,
t:

8 24 HZ

Fig. 2 Intrahemispheric coherence spectra between ad-

jacent electrodes (upper drawing) and inter-
hemispheric coherence spectra between signals
recorded on homotopic areas. The solid lines
represent the coherence during rest - eyes closed
the broken line during eyes closed - listening
to music. Note the changes in the temporal
regions of both sides.

882
activities during rest and eyes closed, the broken lines
show the coherence spectra during listening to music and
eyes closed. Vertical bars in each of the diagrams in-
dicate the significant differences between coherences in
the different frequency bands (95% confidence intervals).
During rest with eyes closed, interhemispheric coherence
is characteristically highest in the frontal areas and
lowest in the temporal areas.

In contrast to these observations, interhemispheric

coherences increase considerably in the temporal regions
during listening to music. On the other hand intrahemis-
pheric coherence values, measured in the transversal dir-
ection (i.e. FP1- FP2, F3- F7, etc.) decrease in the
mid-temporal regions of both hemispheres, particularly
in the Beta band.

These findings may be interpreted as increased in-

terrelationship between the electrical activities within
the temporal regions of both hemispheres during listening
to music. Simultanously however, activities within the
temporal regions themselves become less coherent.

Although one should be aware that only some aspects

of neocortical electrical activity can be examined (it
should be remembered that approximately 40% of the neo-
cortical activity cannot be "seen" with scalp electrodes)
several applications may be envisaged. Firstly this
method may give some new insight into changes of cerebral
functioning during altered conditions of mental activity
and thus may become a tool as a psychiatric diagnostic.
Secondly this method may have importance in the eval-
uation of drug effects and thus help to better under-
stand different states of higher nervous activity.

REFERENCES

Busk J. and Galbraith G.C., EEG correlates of visual-

motor practice in man, Electroenceph.Clin.Neurophysiol.
38, 415-422 (1975)

Grabow J.D., Aronson A.E., Green K.L. and Offord K.P.,

A comparison of EEG activity in the left and right
cerebral hemispheres by power-spectrum analysis during
language and non-language tasks, Electroenceph.Clin.
Neurophysiol. 47, 693-703 (1979)

883
Livanov M.N., Spatial Organization of Cerebral Processes
John Wiley & Sons, New York 1977

Motokawa K. and Tuziguti K., Die Phasendifferenzen der

Alpha-Wellen und Lokalunterschiede der elektrischen
Aktivitat der GroBhirnrinde des Menschen, The Japanese
Journal of Medical Sciences, Vol. 10 (1), 23-38 (1944)

Shaw J.C., O'Connor K. and Ongley C., The EEG as a

measure of cerebral functional organization,
Br.J.Psychiatry 130, 260-264 (1976)

Sklar B., Hanley J. and Simmons W.W., An EEG experiment

aimed toward identifying dyslexic children, Nature (Lond)
240, 414-416 (1972)

884
GOAL DIRECTED BEHAVIOUR AND MOVEMENT RELATED BRAIN
MACROPOTENTIALS.

* **
C.L. Cazzullo and G.A. Chiarenza
*Istituto di Clinica Psichiatrica e **Istituto di
Neuropsichiatria Infantile, Universita degli Studi
di Milano, Via G.F. Besta 1 20161 Milano - Italy

Goal directed behaviour is teleologically purposive. It

often seems to be a SP.arch for a goal previously defined by a model
or idea in the brain 1. Exactly how a goal is achieved, however,
can vary. Formulation of strategies, their realization through
actions, evaluation of results and comparisons with past expe-
riences can be explained in different ways. Fundamentally,
there are two approaches to account for purposive behaviour. One
is the cybernetic approach, which views behaviour as homeostatic
and largely reflexive 2. According to this model, an organism is
endowed with innate patterns of behaviour explained as reflexes
triggered by the stimulus or as the reduction of drives.
Numerous observations have established the great power of this
approach to account for many complex as well as simple behaviours
in humans and other mamma 1s, as we 11 as in insects, fishes and
birds. As we ascend the philogenetic scale, the cybernetic appro-
ach becomes insatisfactory. Behaviours emerge that cannot be expla-
ined plausibly as innated or conditioned reflexes. For example,
response learned to a specified stimulus can be elicited by genera-
lization to a novel stimulus that activated very different affe-
rent pathways; learned responses can be executed by using muscles
that achieve the desired purpose but which were never before used
for that behaviour; animals and humans can learn new skills by
watching the behaviour of another individual.
As one tries to explain such behaviours, another approach must
885
be considered that assumes that higher animals possess conscious-
ness, have ideas and can think about the significance of the infor-
mation from the environment.
The newborn individual can survive only by the action of spe-
cies characteristic reflexes and homeostatic processes. As
the individual develops, other mechanisms may serve to distribute
information to additional regions of the brain.
In this manner, multisensory and multivariate transactions
begin to modulate genetically specific processes that were ini-
tially more simply determined and a cognitive model of the
environment is built gradually, incorporating features of indivi-
dual experience as well as species characteristic features.
One can view behaviour as resulting from a cognitive process
which involves an interaction between neural events representing
the previous experience, the present state of the individual and
the occurrence of particular features in the environment. Such be-
haviour consists of the attempt to match nP.w experience against
an idea reflecting past experiences. It is cognitive rather than
reflective, involving thinking rather than activation of specific
neural pathways constituting stimulus response circuits 3.
Aim of this paper is to present a method with which goal direc-
ted behaviour can be observed and to provide a tentative neurophy-
siological model of cognitive processes by recording the Movement
Related Brain Macropotentials (MRBMs) both in children and
adults in normal and pathological conditions.
The most fundamental problems in building such a model are
what types of information are essential for goal directed behavi-
our and where and how are such informations represented in
the brain.
At least three kinds of information can be identified:
a) information about the presence of a goal in the environment;
b) information about behaviours that might obtain that goal, and
c) information constituting the idea to seek some previously
defined goal 3.
These conditions are realized when we investigate the behavioural
and neurophysiological phenomena that determine a voluntary motor
action, that is the movement, where the muscles are the means to
carry out the project elaborated by the mind.
When it is the subject that begins a voluntarily action to
modify the surrounding environment and these modifications can be

886
evaluated and compared with the preprogrammed strategies and past
experiences, we have all the necessary and sufficient conditions
to observe the interactions between the environment and the human
being and how the relative informations are represented in the
brain. When a subject is engaged in a motor perceptua 1 task in
order to achieve a preset goal and receives a visual feedback in
rea 1-t i me about his motor performance, a characteristic sequence
of brain macro potentia 1s can be recorded from the sea 1p both in
adults and in children 4.5.6.
The motor perceptual task consists in initiating a single
sweep of the asci 11 oscope trace by a thumb press of the button
held in the left hand, and in stopping the sweep within a defined
central area of the oscilloscope screen by a thumb press of the
other button held in the right hand 5. The goal of the subject is
to stop the sweep within this area, i.e. 40-60 ms. after its ini-
tiation. The spot velocity is 10 ms/cm. These performance are
referrd to as self-paced performance reaction time. It is obvious
that such a test has to be totally preprogrammed and in order to
be performed correctly the subject has to monitor the results of
his actions using the on-line provided feedback.
Such a paradigm allows the recording of the performance and
the brain potentials preceding, during and following the motor
actions.
On the basis of the spatia-temporal characteristic of these
potentials and their relationship to electromyographic activity
and performance it has been proposed that there are four succes-
sive time-periods during a motor perceptual task 7. A premotor
period that precedes the EMG activities is characterized by the
Bereitschaftspotential (BP) 8. Its amplitude is higher during
skilled and goal-oriented tasks than during a skilled and not pur-
posive ones 7. Its scalp distribution is prevalent in the central
and precentral areas. During unimanual actions the BP is predomi-
nant on the contralateral emisphere; while during bimanual skilled
action is simmetric in right hand subjects but not in the left
hand subjects 9. This potentia 1 seems to reflect the strategic
organization of the ideokinetic elements necessary to achieve the
goa 1 . The sensory motor period starts from the onset of the EMG
activities and lasts for 80 ms. after the peak of the EMG.
It is during this period that behaviour is manifest. During this
period electric cerebral activity is dominated by a further nega-
tive potentia 1 , Motor Cortex Potentia 1 ( MCP) 7. This potentia 1
seems to be an index of the reafferent peripheral activity.

887
The motor completion period is characterized by the decline of the
EMG activities and by a positive peak with a 1atency of 200 ms. :
P200 10. During this period the visual evoked potential,
evoked by the sweep of the oscilloscope, during a motor perceptual
task, is suppressed in the precentral areas. The post-motor period
is characterized by the return of the electromyographic activities
to the preceding rest conditions and by a presence of a 1arge
positive potential with a latency of 460 ms. denominated Skilled
Performance Positivity (SPP) 4.5.7. This potential has a central
and parietal scalp distribution. It is observed in healthy
subjects, when they are asked to perform a motor perceptual
task that requires precision and improvement of performance level
by providing adequate real time feedback information on the
outcome. The SPP is absent during an unskilled task and when ade-
quate feed-back is not provided to the subject 5.11. From a chro-
nological standpoint it may be said that SPP clashes with the
subject's being aware of the success or failure of his performance.
This positivity appears when the subject looks for information
about his outcome that is to say information relevant to the effi-
ciency of his psychomotor preprogrammed organization. The knowled-
ge of the outcome is likely to be used to influence future actions.
The various stages of organization of goal directed behaviour
can be appreciated following the developmental features of the
Movement Related Brain Macropotentials from childhood to adult
life. Figure 1 shows the sequence and scalp distribution of
these potentia 1s during ski 11 ed performance in a child 10 years
old.
In children of six-seven years of age MCP and P200 are present;
they represent the sensory information coming respectively from
the subject and from the environment. Therefore, in spite of the
counsciousness of the movement carried out, the BP, characteristic
of the preparatory period, is absent in these young subjects.
The BP starts to be recordable as negative potentials in children
of 9 or 10 years of age 6. It is just in this age that the child
is able to interiorize actions on the objects that can be carried
out mentally without being carried out phisically. These operations
are oriented towards concrete things and events in the immediate
present, so movement towards the non present, or potentia 1, is
limited. These interpretations also give an account for the absence
of the SPP in the children of 9-10 years of age 6. During the
time of appearance of this potential another one is recorded with
a negative polarity denominated Post-action Negativity (PAN) most

888
 -~I /1 MCP

~
---- I

-2¥ I _.,. I

Fig. 1 Movement Related Brain Macropotentials during skilled per-

formance. The vertical bar indicates the computer trigger.
For abbreviations see the text.

889
 prominent in the front a1 and centra 1 regions, while the SPP has
a more central-parietal distribution 6.
This negativity characteristic of children between 6-9 years
of age might indicate the cognitive processes of children of the
preconceptua 1 stage of the representation a1 i nte 11 i gence in
which the rea 1i ty depends and extinguishes in the moment of the
immediate perception 12. In these children the surprise or the
novelty about the outcome of the performance predominates. 6
This potential decreases with age; in parallel during 10-11
, years of age, the BP and SPP appear. During this period the chil-
dren acquire the adult capacity for abstract thought. They conceive
many possible ways in which they could operate and many alternati-
ve ways in which they could be better. We suppose that these
different possibilities, these strategic qualities, i.e. the inten-
tionality of actions to achieve a goal, are reflected in the BP
amplitude. Furthermore after eleven years of age thinking is propo-
sitional 12. In the period of formal operations the children
take the outcome of their performance, put them in sentence form,
and begin to find re 1at i onshi ps among the sentences. According
to this view, the emergence of SPP corresponds to a new way of
braiA functioning; the outcome and the knowledge of the results
are used to match them with the projects and past experiences in
order to improve the goal-directed behaviour.
The all pattern of these potentials as have been described
here, is characteristic of the adult age. The effect of ageing on
these potentials have been also described 13. In particular the
BP is reduced in amp 1itude and the MCP disappears. On the other
side SPP is homogeneously and consistently present both in ampli-
tude and in latency.
The goal-directed behaviour as is manifested with the MRBMs
can be also observed during psychopathological conditions in
which it is profoundly altered 13.14.15. For example in chronic
schizophrenics the disruption of cognitive processes is characteri-
zed by concrete thinking, palaelogic ideation, desymbolization
and motor dysphunctions. Therefore, utilizing such a paradigm in
these subjects it should be expected that the MRBMs were altered.
In a recent study, Chiarenza et al. 16, comparing normal sub-
jects with chronic schizophrenics, have demonstrated that the BP
amplitude of the chronic schizophrenics was consistently and signi-
ficantly reduced. It could he suggested that the neurophysiologi-

890
cal correlates of the impairment of schizophrenics in their
capability to organize a purposeful action is expressed through
the reduced BP amplitude. The MCP is also significantly reduced.
The poor performance of chronic schizophrenics could be attributed
to this difficulty in processing peripheral informations, as indi-
cated by the reduction of the MCP amplitude. In chronic schizo-
phrenics the SPP is absent in frontal, central and precentral
areas; it is present a1though reduced in its amp 1i tude in the
parietal region. The absence of the SPP in schizophrenic subjects
may mean that they do not use informations related to their per-
formance outcome in order to achieve a greater number of correct
performance; that it is the effect of the event -whether the
goal is or isn't achieved- on conceptual or cognitive responses
in schizophrenic patients is different fro the effect in normal
subjects.
All these results, as rewieved here, are in agreement
with the psychological theories that place at determined levels
of age the acquisition of the capacities of abstract thought and
with the functional neuroanatomical studies according to which a
biological maturation is necessary for learning processes and the
acquired level of learning is necessarily in relationship with
the maturation level of the cerebral structures. When these
structures due to different conditions are not able to accomplish
adequately their functions, the behaviour is always affected
during its strategic stage or its realization or its evaluation
period as revealed by the Movement Related Brain Macropotentials.

References
1. R. Granit, "The Purposive Brain", Mit Press, Cambridge,
Mass. (1977).
2. N. Wiener, Cybernetics or control and communication in the
animal and the machine. Mit Press, Cambridge, Mass. (1971).
3. E. R. John, A neurophysiological model of purposive behaviour,
in: "Neural Mechanisms of goal-directed behaviour and
learning", R.F. Thompson, L.H. Hicks, V.B. Shvyrkov,
ed., Academic Press, New York (1980).
4. D. Papakostopoulos, Electrical activity of the brain associated
with skilled performance, ~: "~lultidisciplinary Perspecti-
ves in Event-Related Brain Potential Research", D.A.Otto,
ed., U.S. Environmental Protection Agency office of Research
and Development, Washington, D.C. (1978).

891
 5. D. Papakostopoulos, A no stimulus, no response. Event-Related
Potential of the human cortex., Electroenceph. Clin.
Neurophysiol. 48:622 (1980).
6. G. A. Chiarenza, D. Papakostopoulos, F. Giordana, A. Guareschi
Cazzullo, Movement Related Brain Macropotentials during
skilled performances. A developmental study., Electroenceph.
Clin. Neurophysiol. 56:373 (1983).
7. D. Papakostopoulos, The serial order of self-paced movement in
terms of brain macropotentials in man., J. Physiol. (Land.)
280:70 (1978).
8. H. H. Kornhuber, und L. Deecke, Hi rn potentia 1anderungen bei
Willkurbewegungen und passiven Bewegungen des Memschen:
Bereitschaftspotential und reafferente Potentiale.,
Pfluqers Arch. ges. Physiol. 284:1 (1965).
9. D. Papakostopoulos, The Bereitschaftspotential in left and
right-handed subjects., in:"Progress in brain research,
vol. 54, ~lotivation, motor and sensory processes of
the brain: electrical potentials, behaviour and clinical
use", H. H. Kornhuber and L. Deecke, ed., Elsevier /North-
Holland, Amsterdam (1980).
10. G. H. Vaugham jr., L. D. Costa, and W. Ritter, Topography of
the human motor potential., Electroenceph. Clin. Neuro-
physiol. 25:1 (1968).
11. D. Papakostopoulos, MRBM during skilled performance with and
without feedback of knowledge of results, in:"Preliminary
Poster Reports, VIIth Int'l Conference on ERPs of the Brain"
E.P.I.C. VII -in press-, Firenze (1983).
12. B. Inhelder, and J. Piaget, The Growth of logical thinking
from childhood to adolescence., Basic Book Inc., New
York (1958).
13. C. L. Cazzullo, G. A. Chiarenza, S. Scarone, L. Caresia,
L. Bellodi, and F. Giordana, Clinical and Neurophysiologi-
cal assessment of higher nervous activity in obsessive
compulsive disorders, in:"Biological Psychiatry 1981",
C. Perris, G. Struwe & B. Jansson, eds., Elsevier North
Holland Biomedical Press, Amsterdam, 1127 (1981).
14. M. Timsit-Berthier, CNV, slow potentials and motor potential
studies in normal subjects and psychiatric patients, in:
"Human neurophysiology, psychology, psychiatry : Average
evoked responses and their conditioning in normal subjects
and psychiatric patients", A. Fessard and G. Lelord,
eds, Inserm, Paris (1973).

892
15. M. Timsit-Berthier, J. Delaunoy, and J. C. Rousseau, Slow po-
tential changes in psychiatry. II Motor potential.,
Electroenceph. Clin. Neurophysiol. 35:363 (1973 b).
G. A. Chiarenza, D. Papakostopoulos, M. Dini, and C. L. Cazzullo
Neurophysiological correlates of psychomotor activity in
chronic schizophrenics, Electroenceph. Clin. Neurophysiol.
(submitted for publication).

893
CARDIORESPIRATORY ORIENTING RESPONSES AS PREDICTORS OF
INDIVIDUALS AT RISK FOR PSYCHOVISCERAL DISORDERSl

Samuel A. Corson and Elizabeth O'Leary Corson

Department of Psychiatry and Department of Educational

Theory and Practice
The Ohio State University
Columbus, Ohio 43210, USA

ABSTRACT
Utilizing a combination of idiographic and nomothetic research
designs with repeated measures in several breeds of dogs, we dis-
covered stable constitutional differences in psychophysiologic
reactions to repeated exposure to psychologically stressful situa-
tions. Some dogs showed high psychophysiologic adaptation (HA dogs).
Other dogs developed persistent psychophysiologic reactions to the
psychologically aversive environment: tachycardia, polypnea, pro-
fuse salivation, high energy metabolism and muscle tension, and
high urinary vasopressin and catecholamines (low adaptation or LA
dogs). Retrospective analysis of the development and extinction of
cardiac and respiratory orienting reflexes (OR) in these two types
of dogs revealed that the LA dogs exhibit higher frequency and more
intense, persistent, and hi]hly fluctuatinq (poorly modulated)
cardiac and respiratory OR than the HA dogs, which show rapid OR
habituation and good modulation. Insofar as one may extrapolate
these data to human subjects, recording dynamics of development and
habituation and degree of modulation of visceral OR may facilitate
early detection of individuals at risk for cardiorespiratory and
other psychovisceral disorders.

Tsupported in part by NIH, USPHS research grants HL 20861, MH

12089, MH 18098, The OSU Graduate School Biomedical Science
Research Support Grant~ and the Psychiatric Research Foundation
of Columbus.

895
I. THE OBJECT OF THIS STUDY
In this report we describe animal models which may make it
possible to investigate psychophysiologic substrates of types A and
B behaviors described by Friedman and Rosenman (1959) in human sub-
jects. According to these authors, type A behavior was characterized
by time urgency, chronic impatience, intense striving for achieve-
ment, overcommitment to work, excessive aggressive drive, competi-
tiveness, and abruptness of gesture and speech. Type B individuals
were described as more relaxed, patient, easygoing, and lacking
aggression, competitiveness, and time urgency. The authors postula-
ted that individuals exhibiting type A behavior patterns are more
likely to develop coronary heart disease than individuals with type
B behavior. On the basis of prospective investigations by the Wes-
tern Collaborative Study Group, Rosenman et al. {1976) reported
confirmation of the suggestion that type A behavior patterns do
represent an important risk factor for coronary heart d1sease, in-
dependent of other risk factors. Kornitzer et al. (1981) confirmed
a relationship between type A behavior and the prevalence of coron-
ary heart disease in the Belgian Heart Disease Prevention Project.
Classification of human subjects into types A or B is based on
a structured interview (SI) developed by Rosenman and Friedman
(1977), the Jenkins Activity Survey Questionnaire (JAS) (1978),
a Short Rating Scale developed by Bortner (1969), and the Framing-
ham Type A Scale as described by Haynes et al. (1980). In all cases,
the classification is based on the judgment of trained interviewers
and observers of overt behavior, or self-evaluation responses.
The Review Panel on Coronary-Prone Behavior and Coronary Heart
Disease (1981) convened by the National Heart, Lung, and Blood In-
stitute, suggested 11 the need for improved techniques for assessing
type A behavior and systematic investigation of mechanisms by
which type A behavior affects disease status ... Other recommendations
included the need for the elucidation of intervening physiologic
mechanisms and the role of genetic factors. This presentation
describes animal models that may help to find answers to some of
these questions.
Our experimental data on a prospective study of dogs with dif-
ferent genetic programs also suggest possibilities for the utiliza-
tion of autonomic orienting reactions as aids in the psychophysio-
logic classification of individuals and possible prediction of in-
dividuals at risk for psychophysiologic disorders.
II. EXPERIMENTAL DESIGN
Because of our interest in the interaction of genetic and
psychosocial factors in emotional stress reactions, we selected dogs

896
as our experimental animals in view of the availability of a great
variety of standard breeds and mongrels and a wide spectrum of emo-
tional reactions in different dogs. ·
Our experimental design involved concurrent recording of be-
havioral, physiologic, endocrine, and biochemical parameters in the
same dogs studied repeatedly for a period of several years (in some
dogs for as long as 10 years) in four psychologically distinct
rooms maintained at a temperature of 21-23 C: (1) a neutral control
room where baseline data were recorded; (2) an aversive condition-
ing room where Pavlovian motor defense reflexes (leg lift) were
elaborated by reinforcing neutral tones with unavoidable electrocu-
taneous stimuli; (3) an operant conditioning room where similar neu-
tral tones were reinforced with electrocutaneous stimuli, but where
the dogs were permitted to develop discriminated avoidance respon-
ses; (4) a room where Pavlovian conditional salivary reflexes were
elaborated, using food reinforcement only. Details of our experimen-
tal methods were described elsewhere (Corson, 1971; Corson and Cor-
son, 1976).

III. CONSTITUTIONAL DIFFERENCES IN PSYCHOPHYSIOLOGIC REACTIONS OF

DOGS TO AN AVERSIVE PAVLOVIAN CONDITIONING R00~1
As can be seen in table 1, we observed two major types of psy-
chophysiologic (genetically influenced) reactions to repeated ex-
posure to the psychologically stressful environment represented by
the Pavlovian aversive conditioning room (Corson, 1971; Corson and
Corson, 1976). Some dogs (e.g., beagles, most hounds, and some
mongrels) exhibited rapid psychophysiologic adaptation (high
adaptation dogs= HA). Oth~r dogs (e.g., wirehair fox terriers,
cocker spaniels, some border collies, German shepherds, and some
mongrels) exhibited a persistent almost inextinguishable integrated
octet of psychophysiologic reactions to the entire conditioning room
complex: tachycardia; polypnea; profuse salivation; increases in
energy metabolism, rectal temperature, and electromyographic reac-
tions; vasopressin release; and increased urinary catecholamines
(low adaptation dogs = LA).
These psychophysiologic reactions in the LA dogs were invaria-
bly triggered by the entrance of the dogs into the Pavlovian con-
ditioning room, even in extinction sessions or in the absence of any
conditional or unconditional signals. Some of the dogs were studied
by us for as long as 10 years during which period the Pavlovian
conditioning room continued to trigger these psychophysiologic stress
reactions. In contrast to the conditional motor defense reflexes
which were stimulus-bound, well differentiated, and easily extin-
guished, the psychophysiologic visceral reactions in LA dogs were
highly generalized. This phenomenon of somatovisceral dichotomy is
comparable to the principles of schizokinesis reported by Gantt
(1953).

897
 TABLE I
CORSON TYPOLOGY
Based on Reactions to Repeated Exposure
To a Psychologically Stressful Environment
LOW ADAPTATION DOGS HIGH ADAPTATION DOGS
Tachycardia Normal heart rate
Polypnea Normal respiratory rate
Copious psychogenic thermo- No psychogenic salivation
regulatory salivation
Marked increase in: Little or no increase in:
Rectal temperature Rectal temperature
02 consumption 02 consumption
Antidiuretic hormone Vasopressin release
(vasopressin release)
Urinary catecholamines Urinary catecholamines
(norepinephrine, epinephrine)
*Pavlovian weak type of nervous *Pavlovian strong type of
system nervous system
*As determined by Pavlovian salivary reflex method, using the
caffeine test.

Physiologically, the LA dogs respond to the psychologically

aversive environment as though they were engaged in intensive muscu-
lar effort associated with Walter B. Cannon•s "fight or flight"
reactions, although many of these animals generally do not struggle
in the conditioning room. These physiologic reactions of the LA
dogs would represent appropriate adaptive responses under conditions
where the animals were able to engage in "fight or flight" activi-
ties. However, in a Pavlovian paradigm with aversive reinforcement,
these dogs are not able to achieve an adaptive consummatory response.
Therefore, this incessant triggering of visceral arousal in the ab-
sence of actual aversive stimuli represents the transformation of an
adaptive reaction complex into a maladaptive pattern of psychovis-
ceral pathology. This interpretation of the psychophysiologic reac-
tions of the LA dogs to the Pavlovian room is comparable to the con-
cept developed by Glass for type A behavior in humans "as a style of
coping with uncontrollable stressful events" (1977).
It is important to note that we found it impossible to predict
from the overt behavior of a dog whether this animal would exhibit
low adaptative or high adaptative psychophysiologic reactions to a
stressful situation. Some dogs may exhibit a concordance between
overt behavior and covert psychophysiologic reactions. In other
dogs there is a dissonance between autonomic reactions and external

898
behavior. Similar observations were reported by Eliot and Buell
(1983) in human subjects.

IV. CARDIO-RESPIRATORY ORIENTING RESPONSES AS PREDICTORS OF

INDIVIDUALS AT RISK FOR PSYCHOVISCERAL PATHOLOGY
Before the initiation of classical conditioning experiments, it
is customary first to record behavioral and physiologic reactions of
the experimental subjects to the neutral signals which will eventu-
ally become reinforced with unconditional stimuli and thus develop
into conditional signals. The responses to the neutral signals have
been referred to by Pavlov (1928) as orienting reflexes (OR). If the
orienting stimuli are truly neutral, i.e., have no positive or nega-
tive biological significance for the organism, then repeated presen-
tation of these signals will eventually lead to extinction or
habituation.
A preliminary analysis of our data revealed that the dynamics
of the development and extinction of cardiac and respiratory OR could
be used as predictors of high or 1ow ada pta ti on to the psychologic-
ally aversive environment.
At this point we can summarize some of the major characteris-
tics of visceral orienting responses of the LA dogs which may serve
as indices of discovering individuals at risk for psychovisceral
pathology.
1. Intensity of Visceral Orientin~_g Reactions: The cardiac and res-
piratory OR in the LA dogs were of a higher intensity than in the HA
dogs, often being of magnitude comparab·le to reactions to uncondi-
tional or conditional aversive stimuli.
2. Extinction Rate of Orienting Responses: There is a marked delay
in the habituation of orienting responses in the LA dogs, suggesting
decreased central inhibitory modulation.
3. High Lability of OR: One of the most dramatic characteristics of
the LA dogs is the high fluctuation in cardiac and respiratory par-
ameters, particularly in responses to orienting stimuli, again sug-
gesting poor central damping or the occurrence of large errors in the
central regulatory homeostatic mechanisms. It is instructive that
Lader (1971) reported that in human subjects 11 Slow habituation is
related to many fluctuations (high arousal), much overt anxiety, and
high self-assessment of anxiety. 11 Wolf (1972) summarized clinical
data suggesting that 11 disease states are often associated with la-
bility of the involved system. 11
If these findings should prove to be applicable to humans, the
OR method may turn out to be a useful component for a psychophysic-

899
logic personality assessment, and as a method for discovering indi-
viduals at risk for psychophysiologic pathology.
REFERENCES
Bortner, R.W., 1969, A short rating scale as a potential measure of
pattern A behavior, J. Chron. Dis. 22:87-91.
Corson, S.A., 1971, Pavlovian and operant conditioning techniques in
the study of psychosocial and biological relationships, in:
11 Society, Stress and Disease, .. Vol.
1, L. Levi, ed., Oxford
University Press, pp. 7-21.
Corson, S.A., and Corson, E.O'L., 1976, Constitutional differences in
physiologic adaptation to stress and distress, in: 11 Psycho-
pathology of Human Adaptation, .. Serban, ed., Plenum Press,
New York, pp. 77-94.
Eliot, R.S., and Buell, J.C., 1983, The role of the CNS in cardio-
vascular disorders, Hospital Practice, May 1983:189-199.
Friedman, M., and Rosenman, R.H., 1959, Association of specific overt
behavior pattern with blood and cardiovascular findings, J.
Am. Med. Assoc. 169:1286-1295.
Glass, D.C., 1977, 11 Behavior Patterns, Stress, and Coronary Disease, ..
Lawrence Erlbaum Associates, Hillsdale, NJ.
Haynes, S.G., Feinleib, M., and Kannel, W.B., 1980, The relationship
of psychosocial factors to coronary heart disease in the Fra-
mingham study, Am. J. Epidemiol. 111:37.
Jenkins, C.D., 1978, A comparative review of the interview and ques-
tionnaire methods in the assessment of the coronary-prone be-
havior pattern, in: 11 Coronary-Prone Behavior, 11 T.M. Dembroski
et al., eds., Springer-Verlag, New York, pp. 71-88.
Kornitzer, M., Kittel, F., DeBacker, G., and Dramaix, M., 1981, The
Belgian heart disease prevention project: type "A" behavior
pattern and the prevalence of coronary heart disease, Psycho-
som. Med. 43(2):133-145.
Lader, M.H., 1971, The responses of normal subjects and psychiatric
patients to repetitive stimulation, in: 11 Society, Stress and
Disease, 11 Vol. 1, Levi, ed., Oxford Univ. Press, pp. 417-430.
Pavlov, I.P., 1954, 11 0Tipakh Vysshei Nervnoi Deiatel'nosti i Eksperi-
mental'nykh Nevrozakh (Types of Higher Nervous Activity and
Experimental Neuroses) 11 , Medzig, Moscow.
Rosenman, R.H., and Friedman, M., 1977, Modifying type A behavior
pattern, J. of Psychosom. Res. 21:323-331.
Rosenman, R.H., Brand, R.J., Sholtz, R.I., and Friedman, M., 1976,
Multivariate prediction of coronary heart disease during 8.5
year follow-up in the western collaborative group study, Am.
J. Cardiol. 37:903. --
The Review Panel on Coronary-Prone Behavior and Coronary Heart Dis-
ease, 1981, Coronary-prone behavior and coronary heart dis-
ease: a critical review, Circulation 63(6):1199-1215.
Wolf, s., 1972, Sudden cardiac death, in: "The Artery and the Pro-
cess of Arteriosclerosis t·1easurement and· f·1odification, 11 Wolf,
ed., Plenum Press, New York.

900
THE ORIENTING REFLEX AND FUNCTIONAL

STABILITY OF THE NERVOUS SYSTEM

H. D. Kimmel

University of South Florida, Department of Psychology

Tampa, Florida 33612

The primary thesis of this presentation is that V.D.

Nebylitsyn's concept of functional stability of the central nervous
system (CNS) (Nebylitsyn, 1964) may provide a potential neuro-
biological basis for the intellectual dimension of human behavior.
Evidence is presented in support of the proposition that individuals
whose CNS is more functionally stable have higher measured IQs,
while those whose CNS is less functionally stable have lower IQs.

Nebylitsyn (1964) defines the functional stability of the

nervous system dually, in terms of its capacity to withstand
repetitive or very intense stimulation coupled with its insensi-
tivity to very weak stimulation. The "temporal" measure of CNS
strength refers to the ability of the system to continue responding
under repeated stimulation. Resistance to the development of
inhibition with reinforcement (Kimmel and Burns, 1975) is an
example of temporal strength. Resistance to simple habituation is
another example. The "stimulus-intensity" measure of CNS strength
refers to the system's capacity to respond to exceedingly intense
stimulation, i.e., to resist transmarginal (protective) inhibition.
A strong CNS continues to respond over a greater number of
repetitions of stimulation than does a weaker CNS, and it tolerates
more intense stimuli without shutdown than does a weaker system.
Sensitivity, on the other hand, refers basically to the lower
absolute threshold of excitability of the system. The more
sensitive CNS may be changed from a state of rest to one of
catabolic, or destructive activity by extremely weak stimuli,
while the less sensitive CNS does not yield as readily to weak
stimuli.

Strength and sensitivity are negatively correlated. The

greater the strength of the system, e.g., the degree to which it
901
resists transmarginal inhibition under caffeine stimulation or
the degree to which it resists inhibition with reinforcement, the
higher its auditory and visual absolute thresholds. According to
Nebylitsyn, this observed negative correlation between CNS strength
and sensitivity is due to the fact that strength and sensitivity
both are reflections of the same underlying property of the CNS,
" . the extent to which the (system) is susceptible to function-
ally destructive influences" (Nebylitsyn, 1964, p. 440).

We and a number of other investigators have reported that a

negative relationship exists between habituability of the orienting
reflex (OR) and measured IQ (e.g., Grings, Lockhart, and Dameron,
1962; Kimmel and Deboskey, 1978; Kimmel, Pendergrass, and Kimmel,
1967). Grings, et al (1962) found that intellectually subnormal
young adults' skin conductance responses (SCR) to pure tones were
initially larger in magnitude and more persistent over repeated
trials than those of more severely retarded young adults. Kimmel,
Pendergrass & Kimmel (1967) compared the elicited SCRs of severely
retarded and normal children using geometrically shaped visual
stimuli. Again, the lower IQ group made smaller initial SCRs
which habituated more quickly over trials. In a more recent study
Kimmel and Deboskey (1978) found that a group of intellectually
gifted children (IQ > 130) made initially larger and more persistent
SCRs to visual stimuli than did intellectually normal controls
(IQ 90-110). On the basis of these findings, Kimmel (1981)
concluded that a positive correlation exists between the strength
and persistence of the OR in children and measured intelligence,
and that this relationship holds across the full practical range
of measured IQ (i.e., 50-170).

On the assumption that rate of OR habituation (habituability)

provides an inverse meas.ure of CNS strength, analogous to Nebylitsyn' s
use of inhibition with reinforcement, it thus, may also be an
inverse measure of the functional stability of the CNS. The OR
elicited by the clearly suprathreshold auditory and visual stimuli
employed in the studies summarized above ceases to occur after
only a small number of.repetitions in children and young adults
with very low measured IQs, but it continues to occur over many
repetitions in subjects with high IQs.

Hendrickson and Hendrickson (1980) have proposed that the

average evoked potential (AEP) of highly intelligent individuals
should differ from those of individuals with low intelligence,
with the higher IQ subjects showing increasingly more identifiable
peaks and troughs than the lower IQ subjects as a function of time
following the eliciting stimulus •.

Using data from 10 high IQ and 10 low IQ children (Ertl and

Schafer, 1969), the Hendricksons measured the total continuous
length of all of the upward and downward excursions in each

902
record. This was achieved by placing pins along an enlarged picture
of the waveform and placing a string alongside of the pins. The
actual total excursion distance was then obtained by measuring the
length of the string. This is referred to as the "string measure."
The correlation between these AEP string measures and IQ in Ertl
and Schafer's 10 high and low IQ children (combined) was r = 0.773,
Supporting the proposition that high IQ subjects' AEPs are more
complex than those of low IQ subjects receives some support from
these data. A follow-up investigation of the relationship between
the AEP string measure and intelligence utilized 37 adults. The
Hendricksons report that the correlation between the string measure
and the Raven's Advanced Matrices test in this follow-up was
r = 0.47, but this value may have been attenuated somewhat by
restriction of range in the intelligence measures.

The use of habituabil ity and magnitude of the OR to identify

differences in CNS strength is straightforward. Human adults'
SCRs in response to simple auditory and visual stimuli display a
wide range of habituability .. Kimmel and Goldstein (1967) found
that 4 subjects in a sample of 31 reached a habituation criterion
of two successive nonresponses within only 4 trials (including
the 2 nonresponse trials), while 2 of these 31 subjects failed to
meet the habituation criterion within 50 trials. The mean number
of trials to habituation of all 31 subjects was just under 20
trials. The size of elicited SCRs, similarly., varies widely
among subjects. Some make no measurable response even on the
first stimulus presentation, while others make responses so large
that grr:>atly reduced amplification is required to keep them "on
the record." Although the use of OR habituability and, secondarily,
OR magnitude as indices of CNS strength is tentative, both are
positively correlated with measured IQ.

The variability of repeatedly evoked (auditory) potentials

may tentatively be interpreted to be an index of CNS sensitivity.
On the assumption that heightened vulnerability to the influence
of random error is analogous to excessive reactivity to very weak
stimulation, the resulting observable reduction in complexity of
the AEP may be assumed to reflect reduced functional stability of
the CNS. Thus, invariance should be negatively correlated with OR
measures of CNS strength. This relationship remains to be confirmed
empirically.

The proposition that measured intelligence in humans is based

upon the functional stability of the CNS, a complex neurobiological
concept that embodies both strength and sensitivity aspects, appears
to have potential utility. A significant proportion of the
variance in IQ scores has. indeed. been shown to be attributable
to variation in strength and persistence of elicited ORs, and to
variation in AEPs.. If the present assumption that these
psychophysiological measures capture the essence of Nebylitsyn's

903
concept of the functional stability of the nervous system, a more
than adequate stimulus for further research on its relationship
to IQ has been provided.
REFERENCES
Ertl, J.P., and Schafer, E.W.P. Brain response correlates of
psychometric intelligence. Nature, 1969, 223, 421-422.
Grings, W.W., Lockhart, R.A., and Dameron, L.W. Conditioning
autonomic responses of mentally subnormal individuals.
Psychological Monographs, 1962, 76, Whole No. 558.
Hendrickson, D.E., and Hendrickson, A.E. The biological basis of
individual differences in intelligence. Journal of
Personality and Individual Differences, 1980, 1.
Kimmel, H.D. Intelligence and the orienting reflex.- In M.P.
Friedman, J.P. Das, and N. O'Connor (Eds.), Intelligence
and Learning. New York: Plenum, 1981.
Kimmel, H.D., and Burns, R.A. Adaptational aspects of
conditioning. In W.K. Estes (Ed.), Handbook of Learning
and Cognitive Processes. Vol. 2. Hillsdale, New Jersey:
Lawrence Erlbaum Associates, 1975.
Kimmel, H.D., and Debosky, D. Habituation and conditioning of
the orienting reflex in intellectually gifted and average
children. Physiological Psychology, 1978, 6, 377-380.
Kimmel, H.D., and Golds-tein, A.J. Retention of habituation of the
GSR to visual and auditory stimulation. Journal of
Experimental Psychology, 1967, 73, 401-404.
Kimmel, H.D., Pendergrass, V.E., and Kimmel, E. Modifying
children's orienting reactions instrumentally. Conditional
Reflex, 1967, 2, 227-235.
Nebylitsyn, V.D. AD investigation of the connection between
sensitivity and str:e.ngth of the nervous system. In J.A.
Gray (Ed.), Pavlov's Typology. New York: MacMillan, 1964.

904
ANXIETY AND ATTENTION

J.F. Orlebeke, R.J.M. Somsen and M.W. van der Molen

Free University
Department of Experimental Psychology
De Boelelaan 1115
1081 HV Amsterdam

One of the central issues in psychophysiology with regard to

anxiety is the effect of anxiety on attentional processes. Both
unvoluntary attention, manifesting itself in the orienting reflex
(OR) and its habituation, as well as voluntary attention, occurring
e.g. during anticipation of task relevant or noxious stimuli, will
be considered in this paper.

HABITUATION OF THE OR (UNVOLUNTARY ATTENTION)

The OR - or, following Pavlov, the "what is it reflex" - is

a response to change in the environment, especially change that 1s
not expected. Turning of the head, pricking the ears (in some
animals), inhibition of ongoing behavior, EEG a-blocking, pupil
dilatation, increased sweating in hand palms and foot soles,
peripheral vasoconstriction. These are some of the most charac-
teristic components of the OR.
Stimulus repetition produces a diminishing of OR amplitude. This
is called habituation. Habituation can be considered to be the
most elementary form of learning and can be conceived as learning
to neglect irrelevant stimuli. In their classic series of experi-
ments in 1966, Lader & Wing compared speed of habituation in
anxious patients (anxiety not being a byproduct of other complaints)
with that of a healthy control group. After a rest period of 10
minutes subjects received a series of 20 auditory stimuli (tones
of 1000 cps and 100 dB) of I second and an interstimulus interval
between 45 and 85 sec. The dependent variable was the amplitude
of the electrodermal response (EDR). Two characteristic differences
between patients and control subjects emerge. First the OR ampli-

905
tude to the first stimulus is smaller in patients than in controls
and second patients habituate slower than controls. This result -
that has been replicated several times by others - indicates that
anxiety produces a sustained attention for irrelevant environmental
events. Or, in other words, anxiety produces an inability to in-
hibit attention to irrelevant stimuli.
Several authors (e.g. Hart, 1974) have argued that the slowly
habitu~ting EDR in anxious subjects actually is not an OR, i.e. a
response of increased attention and thus enhanced input, but a
manifestation of the defensive reflex (DR}, i.e. a response aimed
at a reduction of environmental input. Thus accordingly to this
point of view the anxious subjects in the Lader & Wing experiment
are predominantly reacting with DR's and the control subjects
with OR's.
The problem is that EDR-amplitude does contain information that
makes it possible to differentiate between OR and DR. There are
however other possibilities. Let me briefly rehearse the ways OR
and DR can be distinguished:

-Constriction of cephalic vessels in case of a DR and dilatation

in case of an OR (Sokolov, 1963).

-Phasic heart rate acceleration in case of a DR and deceleration

in case of an OR (Graham & Clifton, 1966; Orlebeke & Passchier,
19761.
-EDR amplitude, measured in the palm of the hand is larger than
that measured simultaneously at the back of the hand in case of a
DR; the reverse is true in case of an OR (Ohman et al., 1978).
-The recovery of the EDR to prestimulus level is longer in case
of a DR than in case of an OR (Edelberg, 1972).

THE RELATIONSHIP WITH ANXIETY

There are several indications that personality traits, such

as "anxiety" and "need for stimulation" (sensation seeking) can
be described in terms of input modulation. The distinction between
OR and DR might be the biological substratum of that input
modulation continuum. This could hold not only for traits but also
for momentary states of the organism.
Pribram (Pribram & Melges, 1969) has pointed out that the organism,
being confronted with some unexpected environmental change,
"inspects" its behavioral repertoire (to so called "plans") with
regard to their usefulness to cope with the environmental change.
If the perception of the organism is such that the plans are
unappropriate to cope with the situation, then they (the plans)
can be extended and adapted to the input or the input can be
adapted to the plans. The former process is labelled by Pribram
with the term "participation" and the latter wi·th "preparation".

906
 Figure I. EDR and recovery time (ERT)

Anxiety could be conceived as an example of preparation, i.e.

continuously paying unnecessary and with other processes inter-
fering attention, resulting in an increased input. Coping takes
place by exaggerated attention, even if that is not functional.
What evidence do we have for such a statement when we look to the
OR-DR differentiation?
First of all, I want to draw your attention to the difference
between high and low anxious subjects with regard to the electro-
dermal recovery time (ERT). This ERT is defined as the time
constant of a EDR elicited by a stimulus. This is visualized in
figure I.
Concerning the ERT three types of relationship have been demon-
strated repeatedly. First it appears that -across subjects -
ERT of EDR's to aversive stimuli is longer than that to non
aversive stimuli (Edelberg, 1972; Klarman et al., 1975, 1977).
Second it has been found that stimuli that have a certain meaning
(signal stimuli) for the subject - such as is for example the
case with a CS in a classical conditioning situation and with a
warning signal in a reaction time task- evoke EDR's with a
shorter ERT in comparison with so called non-signal stimuli
(Edelberg, 1970, 1972).
Third it has been demonstrated several times (Hinton, 1981;
Levander et al., 1980) that ERT in anxious subjects is shorter
than that of normal controls and especially shorter than the ERT
of psychopathic subjects. The latter group is generally considered

907
to be low anxious relative to the population in general.
Congruent with these findings is the interpretation of the ERT by
Venables (1975) that a fast ERT reflects open attention towards
environmental input.
All this ERT evidence strongly suggests that slow habituation of
the EDR in anxious subjects regards the OR and not the DR.
Another indication for the idea that overattention for irrelevant
stimuli is a correlate of anxiety stems from the observation that
80 dB t;m,nal stimuli (this is about the OR-DR shift intensity)
evoke cardiac deceleration in high anxious and acceleration in low
anxious subjects (Orlebeke & Feij, 1979). See figure 2.

ANTICIPATION TO SIGNAL STIMULI (VOLUNTARY ATTENTION)

During anticipation to aversive stimuli or to stimuli to

which the subject has to make an overt response, a typical
cardiac response can be measured: a fast and small heart rate
deceleration, maximal about one sec after the warning signal (SI),
followed by a more pronounced acceleration and again a relatively
strong deceleration, reaching its maximum at the beginning of the
imperative signal (S2).

• low emotional
o high emotional

4 10 11
lime in half seconds

Figure 2. Phasic heart rate responses to auditory non-signal

stimuli of 80 dB for high and low emotional subjects.
(Reproduced by permission from Orlebeke & Feij, 1979.)

908
 w
0
z
~
I
u

SECONDS

Figure 3. Typical heart rate response in a SI-S2 choice reaction

time task (averaged across subjects and trials).

S2 can be either an (aversive) stimulus not requiring the subject

to give an overt motor response or some signal to which the subject
has to respond (often as fast as possible). The cardiac response
just described (averaged across subjects and trials) has been
presented in figure 3.

The first accelerative component (often labelled as AI) appears

to be smaller in anxious subjects in comparison with non-anxious
controls (Lykken et al., 1972; Orlebeke & Van Doornen, 1977).
The same has been found in a group of subjects high at risk for
coronary heart diseases (CHD) in comparison with persons who have
low CHD risk (Van Doornen et al., 1980). This can be mediated by
a difference in anxiety between the two groups.
A typical group, characterized by low anxiety is formed by
psychopaths. They show - in comparison with non-psychopa th
controls- a rather strong A)-component (Hare et al., 1978).

An important question is of course, which psychologica l inter-

pretation should be assigned to AI. An experiment from our own
laboratory (Somsen et al., 1983) tried to answer that question
by manipulating task parameters. The authors compared four

909
 SI-S2 situations, the lSI being
PLACEBO
PROPRANOLOL
5 sec.

I. Avoidance RT; by g~v~ng a very

£ast disjunctive reaction to S2
<auditory signal) the subject could
avoid an electrical shock.

2. Normal RT; as I but without

60
threat of shock.

3. RT or shock; S2 could be a shock

w
0
or a tone. The subject was uncer-
z
<
0
tain about the nature of S2 during
~ the SI-S2 interval.
<

60
4. Unavoidable shock; the subject
knew that S2 was a shock and that
he could not avoid it.

~
a:
5. Control; as 2 but no motor
~
< response was required.
"a:z
0

Trials belonging to these five

J conditions were randomly presented
60

to all subjects, SI indicated the

type of trial. The experiment was
65 ~
carried out twice: with and without
"0
~ pharmacological blockade of sym-
"'a: pathetic heart innervation with
0
:r 40 mg of propranolol.
60
Figure 4 shows that Al-amplitude
is highest in the "avoidance RT"
task, then in the "normal RT" task
followed by the "RT or shock" task.
Because RT appears to be shortest
in the "avoidance RT" task and

Figure 4. Averaged phasic heart

rate responses in dif-
60 ~
I
ferent Sl-S2 tasks with
\ ,..._ I and without propandol.
~· 'I I
(After Somsen, Van der
I '
\I ,
/ Molen and Orlebeke, 1983).
55 \.../
SECONDS

910
longest in the "RT or shock" task, RT's in the "normal RT" task
being intermediate, one may conclude that the AI component of the
heart rate response measured in a SI-S2 paradigm represents
"motorpreparation".
Across several SI-S2 situations it appears that anxious subjects
give heart rate responses as people in general do when they anti-
cipate an aversive event. Such as is the case in the "unavoidable
shock" task. The AI component has disappeared almost completely
in that condition. So anxious subjects tend to inhibit antici-
pated motorpreparation. This can be considered as an elementary
form of "freezing". One could even speculate that this coping style
is biological related to what is called "Totstellreflex" in the
German biological literature.
As can be seen from figure 4 the morphology of the heart
rate responses remains unaffected by the administration of
propranolol,indicating the dominant vagal control of this
phenomenon.

In conclusion one can say that anxiety is characterized

by a relative inability to inhibit unvoluntary attention for
irrelevant stimuli and a rather strong inhibition of voluntary
anticipated motorpreparation.

REFERENCES

Edelberg, R., I970, The information content of the recovery limb

of the electrodermal response, Psychophysiol., 6:527.
Edelberg, R., I972, Electrodermal recovery rate, goal orientation
and aversion, Psychophysiol., 9:512.
Graham, F.K., Clifton, R.K., I966, Heart rate change as a component
of the orienting response, Psycho!. Bull., 65:305.
Hare, R.D., Frazelle, J., Cox, D.N., I978, Psychopathy and physio-
logical responses to threat of an aversive stimulus.
Psychophysiol., I5:165.
Hart, J.D., 1974, Physiological responses of anxious and normal
subjects to simple signal and non-signal auditory stimuli,
Psychophysiol., I1:443.
Hinton, J.W., I98I, Can different types of anxiety states be dif-
ferentiated electrodermally? Paper presented at the
Conference on "Stress" in Karlovy Vary.
Klorman, R., Wiesenfeld, A.R., Austin, M.L., I975, Autonomic
responses to affective stimuli, Psychophysiol., 12:553.
Klorman, R., Weissberg, R.P., Wiesenfeld, A.R., I977, Individual
differences in fear and autonomic reaction to affective
stimulation, Psychophysiol., I4:45.
Lader, M.H., Wing, L., I966, Physiological measures, sedative
drugs and morbid anxiety, London, Oxford University
Press.

911
Levander, S.E., Schalling, D.S., Lidberg, L., Bartfai, A.,
Lidberg, Y., 1980, Skin conductance reco,rery time and per-
sonality in a group of criminals, Psychophysiol., 17:105.
Lykken, D.T., Macindie, J., Tellegen, A., 1972, Preception: Auto-
nomic response to shocks as a function of predictability
in time and locus, Psychophysiol., 9:318.
Ohman, A., Frederiksen, M., Hughdahl, L., 1978, Orienting and
defensive responding in the electrodermal system: palmar-
dorsal differences and recovery rate during conditioning
to potentially phobic stimuli, Psychophysiol., 15:93.
Orlebeke, J.F., Passchier, J., 1976, Organismic, stimulus and task
determinants of phasic and tonic heart rate and skin
conductance changes, Biol. Psychol., 4:173.
Orlebeke, J.F., Feij, J.A., 1979, The orienting reflex as a
personality correlate, in: The orienting reflex in humans,
H.D. Kimmel, E.H. van Olst, J.F. Orlebeke, eds., Erlbaum,
Hillsdale, N.J.
Orlebeke, J.F., Van Doornen, L.J.P., 1977, Preception (UCR diminu-
tion) in normal and neurotic subjects, Biol. Psychol.,
5:15.
Pribram, K.H., Melges, F.T., 1969, Psychophysiological basis of
emotion, in: Handbook of clinical neurology, vol. 3, Dis-
orders of~igher nervous activity, P.J. Vinken,
G.W. de Bruyn, eds., Amsterdam, North Holland.
Sokolov, E.N., 1963, Perception and the conditioned reflex.
Pergamon, Oxford.
Somsen, R.J.M., Vander Molen, .M.W., Orlebeke, J.F., 1983,
Phasic heart rate changes in reaction time, shock avoidance
and avoidable shock tasks: Are hypothetical generalizations
about different S1-S2 tasks justified?, Psychophysiol.,
20:88-94
Van Doornen, L.J.P., Orlebeke, J.F., Somsen, R.J.M., 1980,
Coronary risk and coping with aversive stimuli,
Psychophysiol., 17:598.
Venables, P.R., 1975, The recovery limb of the skin conductance
response in high-risk research, in: Genetics, environ-
ment and psychopathology, S.A. Mednick, F. Schulsinger,
J. Higgins, B. Bell, eds., Amsterdam, North Holland.

912
 SOME CHARACTERISTICS OF ORIENTING REFLEXES

Irving Maltzman
UCLA - Department of Psychology
405 Hilgard Avenue
Los Angeles, California 90024

"My principal thesis today will be that the input is never

into a quiescent or static system, but always into a system which
is already actively excited and organized. In the intact organism
behavior is the result of interaction of this background of excita-
tion with input from any designated stimulus. Only when we can
state the general characteristics of this background of excitation,
can we understand the effects of a given input" (Lashley, 1951,
p.l12).

This insightful statement made more than 30 years ago by

Lashley in opening his classic paper on the problem of serial order
in behavior is equally pertinent today. It is especially pertinent
for a discussion of the OR in health and sickness. A statement
concluding Lashley's paper is also relevant: "In spite of its
present inadequacy, I feel that the point of view which I have
sketched here holds some promise of a better understanding of
cerebral integration. Attempts to express cerebral function in
terms of the reflex arc, or of associated chains of neurons, seem
to me doomed to failure because they start with the assumption of
a static nervous system. Every bit of evidence available indicates
a dynamic, constantly active system, or rather a composite of many
interacting systems ••• " (Lashley, 1951, p.135)

We may agree with Lashley that the old reflex arc notion he
attacks is incapable of dealing with the complexities of behavior
and the integrative activity of the nervous system, including
problems of serial order and a dynamic, constantly active system.

913
But surely, the current cognitive psychophysiology with its informa-
tion processing computer metaphors is no better equipped to deal
with the problems of a dynamic organism than the old reflex arc
notion.

It may be remembered that Lashley offered several suggestions

concerning fruitful directions leading towards a solution of
problems of serial order in behavior. One such suggestion was the
examination of the implications of an old and relatively neglected
conception, one introduced by the Wurzburg school, the determining
tendency or set. It is interesting that Lashley was unaware that
the notion of determining tendency or set had already been coordi-
nated with a physiological conception of great importance, but
unfortunately unduly neglected in the West. I am referring, of
course, to the notion of the dominant focus, a conception intro-
duced by Ukhtomsky which has been under experimental and theoret-
ical investigation for many years (e.g., Bykov, 1958; Bechterev,
1932; Maltzman, 1979a; 1979b, Razran, 1961; Rusinov, 1973).

The dominant focus and set

According to Rusinov (1973), "The dominant is defined as the

prevailing reflex in an animal at each moment of its activity. More
precisely, the dominant is the temporarily prevailing reflex system
which directs the activity of n~ural centers at a given moment .••
Under normal conditions the dominant is a functional combination of
neural centers consisting of a relatively mobile cortical component
and subcortical, autonomic, and humoral component" (p.1).
My thesis is that our understanding of the OR is limited in the
same manner as any other biobehavioral process. Only when we can
state the general characteristics of the biobehavioral background or
state of the organism at the moment can we adequately develop the
lawful relations of the OR. In other words, in order to adequately
understand the characteristics of the OR we must know, state, con-
trol, understand, the characteristics of the dominant focus, or as
I have called it, cortical set. (Maltzman, 1979a, 1979b). Tech-
niques for examining this relationship are available, although they
unfortunately have not been used extensively for the purpose of
studying the particular relationship under consideration. I have
in mind use of a toposcope, multiple EEG leads, regional cerebral
blood flow, PET scans and other similar techniques that obtain
measures of activity in specificable regions or areas of th~ brain,
measure functional constellations of activity in the brain, where
these measures reflect ongoing activity rather than a phasic, rela-
tively monentary response to a discrete stimulus without regard to
brain states present at the moment. In the absence of the general
use of these techniques in conjunction with indices of the OR, it
is nevertheless possible to begin to explain, predict and control,
the OR by the indirect manipulation of cortical sets or dominant

914
foci. Two ways in which the OR may be varied by varying the set
of the individual are (1) the use of instructions, the most obvious
technique and the one with the longest history, one that goes back
to the Wurzburg school (Humphrey, 1951), and, (2) the use of
different kinds of drugs.

It is unfortunate that the preponderance of research and

theorizing concerning the OR has been relatively limited to the
effects of novelty of one kind or another, the effects of what
Berlyne (1960) has designated as collative variables. From the
very earliest appearance of reports of biobehavioral studies
of the OR conducted in the USSR, the importance of significance
as a determiner of the OR has been apparent. Significance
was reported as both learned and unlearned and even species specific
in kind (Razran, 1961; Voronin, Leontiev, Luria, Sokolov, &
Vinogradova, 1965). One of the most stimulating, and neglected,
early papers on the OR, the one by Luria and Vinogradova (1959),
emphasized the role of significance as a determiner of the OR.

We raise the problem of significance in the present context

because it is intimately related to problems of integrative
activity posed by Lashley. Significance of a stimulus and there-
fore whether or not an OR may occur or if it occurs, is enhanced,
depends upon the state of the organism at the time the stimulus
is presented. Occurrence of an OR depends upon the dominant
focus or cortical set present at the moment, not simply stimul~s
change or novelty. Significance is not a property of the stimu-
lus. Nor is significance a unique process to be inferred from
within the organism. Significance is a consequence of an interac-
tion between an organism and a stimulus where the organism is in
a particular state that heightens the OR. Instrumentally, signi-
ficance of a stimulus may be manifested by approach or avoid-
ance behavior. With intact normal humans, evaluative ratings on
Osgood's (Osgood, Suci, & Tannenbaum, 1957) semantic differential
may serve as an independent measure of significance (Maltzman,
1979a; Maltzman, Raskin, & Wolff, 1979).

Fortunately, it is relatively simple to induce differential

dominant foci or cortical set in normal adult subjects by means
of task instructions. Effects of some of these procedures were
reported by Luria and Vinagradova (1959). One of the first
experiments on the OR conducted in my laboratory enlarged upon their
findings by using different kinds of task instructions (Maltzman &
Raskin, 1965). Significance as a determiner of the OR in these
habituation experiments is a consequence of the establishment of a
dominant focus or cortical set induced by task instructions impar-
ting signal value to an innocous stimulus. Occurrence of the desig-
nated task stimulus evokes an OR as a consequence of a form of
verbal conditioning even though the stimulus lacks novelty. No new
principles are involved in the theoretical analysis of significance.

915
Just as classical conditioning occurs as a consequence of the estab-
lishment of conditioned reflexes and dominant foci, so evocation of
an OR under the present circumstances is a consequence of a verbally
conditioned OR and a dominant focus or cortical set induced by ·the
task instructions.

References

Bechterev, V. M. General principles of human reflexology. New York:

International Publishers, 1932. (Republished: New York: Arno
Press, 1973).
Berlyne, D. E. Conflict. arousal and curiosity. New York: McGraw-
Hill, 1960.
Bykov, K. M. (Ed.). Text-book of physiology. Moscow: Foreign
Languages Pu~lishing House, 1958.
Humphrey, G. Thinking. London: Methuen, 1951.
Lashley, K. S. The problem of serial order in behavior. In L. A.
Jeffress (Ed.), Cerebral mechanisms in behavior. New York:
Wiley, 1951.
Luria, A. R., & Vinogradova, 0. S. An objective investigation of
the dynamics of semant~c systems. British Journal of Psycho-
logy, 1959, 50, 89-105.
Maltzman, I. Orienting reflexes and classical conditioning in
humans. In H. D. Kimmel, E. H. van Olst, and J. F. Orlebeke
(Eds.), The orienting reflex in humans. New York: Erlbaum,
1979. (a)
Maltzman, I. Orienting reflexes and significance; A reply to
O'Gorman. Psychophysiology, 1979, 16, 274-283. (b)
Maltzman, I., & Raskin, D. C. Effects of individual differences
in the orienting reflex on conditioning and complex processes.
Maltzman, I., Raskin. D. C., & Wolff, C. Latent inhibition of the
GSR conditioned to words. Physiological Ps~chology, 1979. l•
193-203.
Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. The measurement of
meaning. Urbana, Ill.: University of Illinois, 1957.
Razran, G. The observable unconscious and the inferable conscious
in current Soviet psycho-physiology: Interoceptive conditioning,
semantic conditioning, and the orienting reflex. Psychological
Review, 1961, 68, 81-147.
Rusinov, V. S. The-dominant focus. New York: Consultants Bureau,
1973.
Voronin, L. G., Leontiev, A. N., Luria, A. R., Sokolov, E. N., &
Vinogradova, 0. S. (Eds.), Orienting reflex and exploratory
behavior. Washington, D. C.: American Institute for Biological
Sciences, 1965.

916
THE DIAGNOSTIC POTENTIAL OF THE ORIENTING REFLEX

Joseph Wolpe
Temple University
Department of Psychiatry
3300 Henry Avenue
Philadelphia, Pa. (19129)

Corson, Corson and Andrysco (1981) showed that the character of

a dog's cardiorespiratory orienting responses is a reliable pre-
dictor of the effects that repeated aversive, electrical stimulation
will have on the animal. It has lonq been known that such stimula-
tion of spatially restricted animals often produces long-lasting
autonomic and motor response complexes that Pavlov (1927) referred
to as "experimental neuroses". These are characterized by sympa-
thetic-dominated responses such as mydriasis, tachycardia and
polypnea, which are especially arousable by the stimuli of the con-
ditioning room.
The innovative step of the Corsons was to study each dog's
cardiorespiratory orienting reflexes before subjecting the animal
to aversive electrical stimulation. They found that the animals in
which these orienting reflexes were strong and persistent were the
ones that developed experimental neuroses, and they labelled them
low adaptation dogs. Animals in which the orientation reflexes were
weaker and less persistent did not develop experimental neuroses
(high adaptation dogs).
The Corsons have inferred from their observations that strong
and persistent orientation reactions should distinguish those human
beings who are especially liable to coronary heart disease, en-
abling preventive measures to be taken. They postulate a basic
similarity between low adaptation dogs and people who are especially
subject to coronary heart disease- who have what has in recent years
come to be called the Type A personality profile (Friedman &
Rosenman, 1959; Glass, 1977). So convinced are they that the para-
llel exists that they sometimes allude to their low adaptation dogs
as type A dogs.
917
 The cardinal features of the Type A profile are time urgency,
chronic impatience, intense striving for achievement, over-
commitment to work, excessive agressive behavior, competitiveness
and abruptness of gesture and speech. While in some individuals
the existence of this combination of features is obvious, in others
it is obscured - for example, by their having learned to control,
conceal or disguise characteristics like impatience or aggressive-
ness. It would obviously be very useful to have a convenient
objective test on the basis of which the Type A personality could
be identified.
In actuality, there is no justification for equating Type A
people with animals susceptible to experimental neurosis. On the
contrary, there are weighty reasons for believing that the experi-
mental neurosis is the model of human neuroses (Wolpe, 1952, 1958,
1967, 1982), the keynote of both being elicitation of anxiety re-
actions by situations in which they are inappropriate. The neurotic
person or animal is fearful in response to such harmless stimuli
as flashes of light or earthworms. The reactions of the clinical
and the experimental subjects also resemble each other in manner
of causation, generalization, second order conditioning, persist-
ence, and modes of extinction (Wolpe, 1967).
If animal and human neuroses are parallel phenomena, it is
reasonable to expect strong and persistent orientation responses
in people who are susceptible to neurosis. In fact, Lader (1971)
found such responses in anxiety patients. It seems unlikely that
they would also be found in Type A individuals, since they are so
different from neurotics. The neurotic person tends to be retiring
and submissive in contrast to the aggressive Type A personality.
The realities will however not be established by logic but by
empirical investigation. It might also turn out that weak and
transient orienting reflexes are pathognomic of the Type A person-
ality! In that event, the orienting reflex would be a predictor,
at opposite extremes, of both Type A personality and susceptibility
to neurosis -giving the Corson's test even greater utility than
they had envisaged, despite the erroneousness of their specific
prediction.
The ability to identify the Type A personality is of practical
value only if we have the capacity beneficially to modify it.
There is evidence that routine behavioral techniques, such as re-
laxation training and systematic desensitization, can be success-
fully used to overcome excessive habits of anger and other over-
reactions to frustration (Roskies, 1979)- in much the same way as
they are used for overcoming neurot1c anxiety.

918
Corson'•s observation. lends itself to facilitating. future studies of
experimental neuroses. The experimental neurosis is the most
direct and economical way of obtaining information relevant to
human neuroses. The successful therapy of animal neuroses has been
the model upon which a whole range of treatments for clinical
neuroses has been based {Wolpe, 1982). It is to be hoped that
studies of the experimental neuroses, long in abeyance, will once
again be a prominent feature of the animal laboratory; and it will
certainly be a great help - at least when dogs are the experimental
animals - to be able to predict beforehand which animals are likely
to be susceptible.

References
Corson, S.A., Corson, E.O., and Andrysco, R.M., 1981, Cardiac and
respiratory orienting reflexes as indices of susceptibility
to psychosomatic cardio-respiratory disorders. Proceedings
of the International Symposium on "Psychophysiological Risk
Factors in Cardiovascular Diseases," Karlovy Vary,
Czechoslovakia, September 7-11.
Friedman, M. and Rosenman, R.H., 1959, Association of specific
overt behavior pattern with blood and cardiovascular findings.
J. of the A.M.A., 169:1286.
Glass, D.C., 1977, Stress, behavior patterns, and coronary disease.
Amer. Scient., 65:177.
Lader, M.H., 1971, The responses of normal subjects and psychiatric
patients to repetitive stimulation, in "Society, Stress and
Disease," Vol. 1: "The Psychosocial Environment and Psycho-
somat1c Diseases," pp. 417-430. Levi, ed., Oxford University
Press, London.
Roskies, E., 1979, Considerations in developing a treatment program
for the coronary-prone {Type A) behavior pattern, in: "Behavioral
Medicine: Changing Lifestyles," P.D. Davidson and S.M.
Davidson, ed., Bruner/Mazel, New York.
Pavlov, I. P. , 1927, "Conditioned Reflexes," G. V. An rep {Trans. ) ,
Liveright, New York.
Wolpe, J., 1952, Experimental neuroses as learned behavior. Brit.
J. of·Psythol., 43:243.
Wolpe, J., 1958, "Psychotherapy By Reciprocal Inhibition," Stanford
University Press, California.
Wolpe, J., 1967, Parallels between animal and human neuroses. In
"Comparative Psychopathology," J. Zubin &H.F. Hunt, eds.,
Grune and Stratton, New York.
Wolpe, J., 1982, "The Practice of Behavior Therapy," {3rd Ed.),
Pergamon Press, New York.

919
ORIENTING REFLEXES IN MEDICAL PRACTICE

Walter Knopp

Ohio State University College of Medicine

Department of Psychiatry
Columbus, Ohio 43210

INTRODUCTION

About thirty years ago, unbeknownst to each other, two

scientific teams encountered in some of their subjects unexpected
phenomena which appeared to be related to stressful stimuli. One
team was headed by a biophysicist physiologist (Corson 1961). The
other team by two cardiologists (Friedman et al. 1958). Both teams
pursued their leads against all odds and peer advice. Meanwhile,
Corson, using dogs established a physiologically well defined
typology based on longitudinal studies of reactions (including
orienting reflexes - O.R.) to a psychologically stressful
environment (Corson and Corson, 1976; Corson and Corson, 1980).
His findings are supported by refined human studies (Eliot and
Buell 1983). In the latter, the two authors have used different
measures and came to similar conclusions, namely that there are
two kinds of responders diametrically opposed to each other.

Friedman et al. (1981) defined the Type A and Type B reaction

pattern and found Type A to be a relevant risk factor for coronary
heart disease (CHD). Their earlier work was rightfully criticized
for lack of sophistication in the behavioral assessment of their
subjects. With the help of a team of behavioral scientists they
succeeded to better define the Types A and B behaviors and devise
more effective therapeutic methods of dealing with Type A behavior
(Friedman et al., 1982; Thoresen et al., 1981; Price, 1983).
Their work found support in a recent study by Williams (1981 and
1983). Physicians with high hostility scores had a coronary heart
disease and death rate of 12% compared to 3% in physicians with low
hostility scores. The fact that myocardial infarction also occurs

921
in Type B persons and no infarction may occur in Type A persons
is still one of the sources of criticism and controversy (Bass
and Wade, 1982; Stamler, 1980 and Medical Tribune, 1983). It was
recently addressed by Eliot and Buell (1983) who found that the
overt behavior typified as Type A is not necessarily correlated
with the covert endocrime metabolic and other physiological
reactions. They found subjects who overtly were "hot reactors"
and covertly were "cool reactors" and vice-versa. Thus they
introduced a new differential concept of concordance and dis-
cordance between overt and covert behavior into the study of this
problem. All biobehaviorally· oriented teams emphasize individuali-
zation, not only in the behavioral but also in the physiological
sphere of their diagnostic and therapeutic endeavors.

ORIENTING REFLEXES AND PHYSIOLOGIC AROUSAL

In his review of Kimmel's book on the "Orienting Reflex in

Humans" Corson (1981) states that some of the contradictory and
confusing aspects of O.R. studies can probably be ascribed to
(1) differences exhibited by different physiological and be-
havioral parameters and species differences as well as
constitutional differences in the same species, and (2) due to
the confusion in the use of the term "physiologic arousal".
Corson differentiates the "high physiological arousal" as evidenced
by increased and poorly controlled activity in the psychomotor
and/or autonomic nervous (visceral) systems from the "high physio-
logic arousal" of higher cortical structures that may imply better
integrated goal directed psychomotor behavior and better inhibitory
modulation of visceral functions. Hakerem's and Knopp's work
using electronic pupillography (Lowenstein, 1959) supports Corson's
view. One of the measures, namely the pupillary response to light,
can be compared to one obtained via an "electrode" implanted in the
mesencephalic region (Edinger- Westphal- Nucleus). It is a
measure of supranuclear inhibition presumably via higher cortical
and subcortical structures. The other measure, namely the
spontaneous fluctuations of the dark adapted pupillary diameter is
comparable to that obtained via an "electrode" implanted in the
hypothalamus. The former seems to monitor the mesencephalic level
of psychomotor integration subserved by the neurotransmitter dopa-
mine, while the latter seems to monitor the hypothalamic level of
autonomic nervous system (visceral) integration subserved mainly
nor-epinephrine and serotonin.

In a double blind study of Gilles-de-la Tourette's syndrome

(Knopp, 1967) in which four weeks of no drug or placebo were
followed by ten weeks of haloperidol treatment, the extent of
pupillary contraction to a standard light stimulus under standard

922
conditions (presumably reflecting the mesencephalic modulation of
pupillary activity) displayed a 300% increase, while the dark
adapted diameter of the pupil (presumably reflecting the hypo-
thalamic level of modulation) showed no change whatsoever when
haloperidol (a predominantly dopamine blocker) was introduced. In
the same study a similar distinction was seen between the frequency
of tics and mannerisms (psychomotor behavior) which was haloperidol
dependent while the emotional (visceral) behavior i.e., hypothala-
mically modulated behavior dominated by tension, anxiety, guilt
and dread did not show such a relationship. Similar findings were
observed in a group of 25 acute schizophrenics who were treated by
Trifluoperazine (Knopp, 1970). The neuroleptic items (hallucinations,
delusions, i.e., meso-limbic integration) showed an early response
to the neuroleptic drug whereas the emotional items did not show
such a response at all. In Hakerem's study (Knopp and Hakerem
1972) two families of pupillary reactions to light were
superimposed. The normals as a group displayed a larger average
extent of contraction than did the schizophrenics as a group.
Sin~e smaller extent of contraction is associated with increased
arousal (Loewenfeld, 1973), it is of interest that Goldstein and
Sugerman (1969) and Marjerrison et al. (1967) found EEG signs of
hyper-arousal in schizophrenic patients. Also, one of the theories
of the schizophrenias emphasizes dopamine increase in the basal
ganglia and in the mensolimbic region in schizophrenics. Hakerem's
schizophrenics, who were supposedly more aroused than the normals
and would have been expected to have larger dark adapted diameters,
showed surprisingly a significantly smaller average dark adapted
diameter than did the normals. This discrepancy can only be
explained by a split (or uncoupling) between the two modulating
systems, namely, the noradrenergic-serotonerg ic modulation of the
hypothalamus which seems responsible for the dark adapted pupil at
stimulus impact, and the dopaminergic modulation of the Edinger-
Westphal nucleus via the nigro-striatal system and reticular
formation which seems responsible for the extent of contraction of
the light reflex.

Hakerem (1973) also studied the dark adapted pupillary diameters

simultaneously with evoked vertex potentials. In response to a
gambling situation, there was a high intra-individual reliability
over time and a very great inter-individual variability between
subjects. Under conditions of certainty of the outcome or when
the subject was encouraged not to guess but only to report what he
heard the responses were very different from the condition of
uncertainty. Although Hakerem assumes that he had ruled out the
orienting responses in these cases and that the differences are a
result of the cognitive processes going on in a betting situation,
this finding seems of interest in this context.

923
CONCLUSIONS

These examples seem to emphasize Corson's assertion that

some of the contradictory and confusing aspects of the orienting
reflex studies may be due to semantic troubles, lack of under-
standing of the different levels of central autonomic nervous
system integration, differential arousals within these levels and
the extreme complexity of the integrated psychophysiologic
response that is simply called the Orienting Reflex.

If it were possible to combine Corson's methodology with

those used by Friedman-Rosenman, Eliot-Buell and Williams the
hypothesis could be tested that O.R. patterns may be potential
simple physiological predictors of persons at risk for CHD and
other psycho-visceral disorders.

REFERENCES

Bass, C., Wade, C., 1982 "Type A behaviour: not specifically

pathogenic?" Lancet II:1147.
Corson, 1961 "Personal Communication (Observed 1948).
Corson, S.A. and Corson, E.O'L., 1976, Constitutional
differences in physiologic adaptation to stress and
distress, .:f.n: "Psychopathology of human adaptation,"
G. Serban, ed., Plenum Press, New York.
Corson, S.A., Corson, E.O'L., Lanese, R.A. and Andrysco, R.M.,
1980, "Psychophysiologic personality assessment on the
basis of cardiorespiratory orienting responses,"
Pav. J. Biol. Sci., 15(2):86.
Corson, S.A. and Corson, E.O'L., 1981, "Review of orienting
reflex in humans," Pav. J. Biol. Sci., 16:176-177.
Eliot, R.S. and Buell, J.C., 1983, "The role of CNS in
cardiovascular disorders," Hospital Practice, 18(5):
189-199.
Friedman, M., Rosenman, R.H., Carroll, V., 1958, "Changes in
the serum cholesterol and blood clotting time in men
subjected to cyclic variation of occupational stress."
Circulation, 12:852.
Friedman, M., Thoresen, C.E., and Gill, J.J., 1981, "Type A
behavior: Its possible role, detection, and alteration
in patients with ischemic heart disease," in: J.W. Hurst,
ed., "Heart Update V," New York, N.Y., McGraw-Hill.
Friedman, M., Thoresen, C.E., Gill, J.J., Ulmer, D.,
Thompson, 1., Powell, L., Price, V., Elek, S.R.,
Rabin, D.D., Breall, W.S., Piaget, G., Dixon, T.,
Bourg, E., Levy, R.A., Tasto, D.L., 1982, "Feasibility
of altering type A behavior pattern after myocardial
infarction," Circulation 66(1):83-92.

924
Goldstein, L., Sugerman, A., 1969, "EEG correlates of
psychopathology," in: Neurobiological aspects of psycho-
pathology," J. Zubin and C. Shagass, editors, Grunne &
Stratton, New York, London.
Hakerem, G., 1973, "The effect of cognitive manipulation on
pupillary diameter and evoked vertex potentials,"
in: "Normal and disturbed pupillary movements," Dodt
and Schrader, Editors, Verlag, J.F. Bergmann, Munich.
Knopp, w., 1967, "Formal discussion," of Challas, G.,
Chapel, J. and Jenkins, R., "Tourette's disease: control
of symptoms and its clinical course," Internat. J.
Neuropsychiat. 3 (Suppl. No. 1), 106-109.
Knopp, w., 1970, "Man's tripartite brain and psychosomatic
medicine," in: Recent research in psychosomatics,
R.A. Pierloot, ed., Psychother. Psychosom. 18:130-136,
S. Karger, Basel/New York.
Knopp, W., Hakerem, G., 1973, "Pupillography, bioamines and
behavioral pathology (in German)," in: "Normal and
disturbed pupillary movements," Dodt and Schrader,
Editors, Verlag J.F. Bergmann, Munich.
Loewenfeld, I., 1973, "Supranuclear inhibition (in German),"
in: "Normal and disturbed pupillary movements," Dodt and
Schrader, Editors, Verlag J.F. Bergmann, Munich.
Lowenstein, 0., 1959, "Electronic pupillography, why, how and
when?" The E.E.N.T. Monthly, 38:549-558.
Marjerrison, G., Krause, A., Keogh, R., 1967, "Variability of
the EEG in schizophrenia: quantitative analysis with a
modulus voltage integration," Electroenceph. Clin. Neuro-
physiol. 24:35-41.
Medical Tribune, May 11, 1983, "Avert 15,000 2nd Mis a year
behaviorally," pp. 108, "MRFIT study challenges A, B
Typology; Dr. Friedman calls Findings Ridiculous,"
pp. 8-20, "Ventilation, discovery, change: the route to
type A reform," pp. 8-20.
Price, V.A., 1982, "Type A behavioral pattern- a model for
research practice," Academic Press, New York, London.
Stamler, J., 1980, "Type A behavior pattern: An established
major risk factor for coronary heart disease? The
key life-style trait responsible for the coronary
epidermic?" in: "Current controversies in cardio-
vascular disease," E. Rapaport, ed., W.B. Saunders,
Philadelphia, London, Toronto.
Thoreson, C.E., Friedman, M., Gill, J.K., Ulmer, D.K., 1982,
"The recurrent coronary prevention project. Some
preliminary findings." Acta Med. Scand. (Suppl)
660:172-192.

925
 Williams, R.B., Jr., 1981, "Behavioral factors in cardio-
vascular disease: an uptdate," J.W. Hurst, ed,
"Heart Update V," New York, N.Y., McGraw-Hill.
Also: "Hostility linked to disease, death rate,"
American Medical News, p. 42, January 21, 1983.

926
PSYCHOPHARMACOLOGICAL IMPLICATIONS OF THE ORIENTING REFLEX

Constant H. Vranckx

Psychiatrisch Hospitaal der Broeders Alexianen

39 Veldbornstraat, Tienen, B-3300, Belgie

The foregoing communications not only point to the complexity

of orienting reflexes in animals as well as in humans, but they
also disclose seemingly contradictory interpretations and
conclusions.

We might have a tendency to consider the stable, "diuretic" dogs

as described by Samuel Corson (which extinguish easily their
orienting reflexes and which seem to do so by discriminating
more efficiently novel stimuli) as being more "intelligent".

However Herbert Kimmel offers an attractive alternative. It lS

interesting to note that there exists a good correlation between
persisting orienting reflexes and high IQ values.

Higher vasopressin release has been found in dogs with persisting

orienting reflexes. Vasopressin improves short-term and long-
term memory.

Superior adaptation does not mean superior intelligence, but

higher IQ values do not necessarily lead to better thinking
either (E. de Bono).

The distinction between "overt" (striated muscle) and "covert"

(visceral and neuroendocrine reactivity) orienting behaviour
opens perspectives which have to be explored in humans.

It has not been shown that Samuel Corson's unstable antidiuretic

dogs have been more liable to disease in spite of the poor modu-
lation of their orienting reflexes.

927
We still do not know whether in humans persisting orienting re-
flexes and covert (visceral) reactivity correlate with "anxiety
neurosis" and/or other neurotic patterns .of reactivity (phobias,
obsessions ••• ) or rather with type A personalities (Rosenman)
or with other typologies.

As long as we do not have better insight in the significance of

orienting reflexes in human patients or volunteers, "psycho-
pharmacological implications" remain speculative.

However, it is interesting to note, the effects of lysinvaso-

pressin (as a nasal spray : which improves the penetration into
the brain) in human patients. Lysinvasopressin improves memory
function in posttraumatic amnesia and in memory disturbances of
aging people. In a few studies the beneficial effects of lysin-
vasopressin in some schizophrenic patients (type 2, i.e. deterio-
rating and presenting with negative symptoms) has been shown,
thus confirming the earlier work of Lorant Forisz (1952) with
Pitressin. The significance of these clinical results of vaso-
pressin in schizophrenic patients should be related to the
differences in orienting reflexes in schizophrenic patients as
reported by Wagner Bridger. Its effects in neurotic patients
or in type A personalities has not yet been studied systemati-
cally. Oxytocin has not been studied sufficiently either.

Another, rather unexpected, drug effect which becomes less

"paradoxical" in the light of the work of Samuel Corson and his
collaborators, is the normalizing effect of small doses of
amphetamines on the "visceral turmoil" of some patients in anti-
cipation of - even pleasurable - events, such as going to a
party. This "turmoil" leads to disturbing reactions like
vomiting, diarrhea, o.a., it appears in the hours preceding
the anticipated and desirable activity. In some cases it does
not respond to benzodiazepine-treatment of a sufficient range
of dosages. (Vranckx 1982 unpublished data).

Samuel Corson had shown that amphetamine treatment of hyper-

kinetic and agressive dogs leads to a dramatic change. At
first such dogs will no longer bite nor react with agitated
motor behaviour, but only as long as they are under the effect
of amphetamines. After several weeks of training sessions with
amphetamines, while they respond to "tender loving care", they
learn to behave in. the desired social way even without the drug.
They have learned something and this experience changes their
behaviour. Would it be out of place from now on to be more
hopeful in the case of so-called sociopathic and antisocial
children?

928
As to the use of benzodiazepines, I would like to suggest to use
them in a similar way : in specific "training sessions", on a
basis of behaviour-therapy rather than as a "continuous" correc-
tor of maladapted reactivity. In this way the dangers of these
useful! drugs could be minimized.

929
INDICATIONS FOR PSYCHOSURGERY:

SOME THEORETICAL ASPECTS

P.K. Bridges and J.R. Bartlett

The Geoffrey Knight Psychosurgical Unit

The Brook General Hospital
Shooters Hill Road, London, SE18 U.K.

Ideas about the indications for psychosurgical operations

depend upon the wider issue of hypotheses concerning the classi-
fication of psychiatric illnesses. A recent report by Hansen and
colleagues (1982) gave quite numerous diagnoses for 65 patients who
had been treated by stereotactic psychosurgery at one or more of
three cerebral sites. Their paper tends to employ neuroanatomical
concepts, especially involving the limbic system, to relate the
diagnoses to the indicated operation sites. We suggest that the
problem is better considered from the aspect of clinical psychiatry
and not neuroanatomy.

The most important controversy in psychiatry remains the

aetiological relationship between the neuroses and the psychoses.
Psychodynamic theories tend to emphasise a continuum and, if this
were so, it would be difficult to justify the use of physical
treatments, especially psychosurgery. However, if the psychoses
are aetiologically distinct from the neuroses in that the former
involve a metabolic abnormality, then physical treatments would be
specifically indicated for the functional psychoses (schizophrenia
and the affective disorders). There is some evidence for this
difference. The functional psychoses are not invariably more
severe than are neuroses, as psychodynamic theory implies. The
psychoses are likely to be inherited and are associated with
specific symptoms, not found among the neuroses, such as delusions,
hallucinations, thought-disorder and diurnal variation of mood.
There is now much psychopharmacological evidence to suggest that
both groups of functional psychoses seem to involve inherited
neurotransmission disorders presenting with prominent psychological
symptoms. In this respect they need not be fundamentally different
from other neurotransmission disorders such as Huntingdqn's Chorea,

931
hypertension and Parkinson's disease. Neurosurgery was appropriate
for Parkinson's disease and therefore it should not be surprising
that neurosurgery may be used for the neurotransmission abnormality
of refractory endogenous depression. Persuasive clinical evidence
in favour of this approach is that treating one of these disorders
with appropriate medication tends to produce another as a side-
effect; for example, giving a neuroleptic to a schizophrenic
patient may cause symptoms of Parkinsonism, giving methyl-dopa
for hypertension may produce endogenous depression.

It is therefore suggested that the various presentations of

endogenous (or psychotic) depression are appropriate for psycho-
surgery when intractable; and the neuroses, with no evidence at
present of being associated with metabolic defects, are not
indications. There will be problems with selection because the
diagnostic differentiation is at present necessarily clinical
Therefore atypical presentations will produce diagnostic uncertainty.
Responses to previous physical treatments, which can be seen to have
some diagnostic significance, are important in the assessment for
psychosurgery which should not be considered as an isolated treat-
ment for very atypical cases. Psychosurgery is at the end of a
progression of physical treatments appropriate for affective
disorders, which will initially involve the use of antidepressant
medication and then perhaps electroconvulsive therapy. If both
come to fail to help then psychosurgery should be considered and
it would be unusual for this Unit to accept a patient for psycho-
surgery who has never shown a response to either antidepressants
or ECT.

While this failure to respond to other physical treatments

is clearly an important aspect there may be doubt as to what dose
of antidepressant should be tried before therapeutic failure is
assumed, and this problem has been discussed by Bridges (1983).
In any case it is clear now that 'maintenance ECT' is contra-
indicated because modern refined psychosurgery can be expected
to be permanently effective without the risk of dysmnesia in such
cases.

In clinical terms, then, the assessment for psychosurgery need

not be unduly complicated. At its simplest, selection might
proceed thus:

1. Age of onset - the relevant illnesses tend to begin in

the second half of life
2. Normal personality function (not necessarily a robust
personality) apparent before the onset and between
attacks of illness - this shows that the illness
tends to be independent of a personality abnormality.

932
3. Earlier response to antidepressants andjor ECT - shows
that physical treatments have been effective in the
past.
4. The presence of symptoms of endogenous depression such
as diurnal variation of mood and guilty delusions -
shows that the appropriate illness is present
5. Positive family history - implies the presence of the
appropriate inherited affective disorder.

It is concluded that only a limited number of psychiatric

diagnoses respond to psychosurgery but the wide-ranging presenta-
tions of affective disorders often cause problems with clinical
recognition. For further clinical details as to indications
see Bartlett et al. (1981).

REFERENCES

Bartlett, J.R. Bridges, P.K. and Kelly, D., 1981,

Contemporary indications for psychosurgery
Brit. J. Psychiat., 38:507
Bridges, P.K., 1983, . . . and a small dose of an
antidepressant might help, Brit. J. Psychiat.,
142:626
Hansen, H., Andersen, R., Theilgaard, A. and Lunn, V.
1982, Stereotactic Psychosurgery. Acta Psychiat.
Scand., Suppl. 301:66

933
FOLLOW-UP STUDY OF OBSESSIVE-COMPULSIVE PATIENTS TREATED BY

PSYCHIATRIC SURGERY IN BELGIUM AND THE NETHERLANDS FROM 1971 TO

1981
P. Cosyns, W.P. Haaijman, J.A.M. Ceha and N. Sijben

Univ.Psychiatrisch Centrum St.Jozef

Leuvensesteenweg 517
B - 3070 Kortenberg (Belgium)

I. COMMITTEE ON PSYCHIATRIC SURGERY

Psychiatric surgery is a rather infrequent procedure in the

treatment of psychiatric patients in Belgium and the Netherlands.

This is exactly the reason why various medical professionals

decided to get together since 1971 in Utrecht (Netherlands) in
order to concentrate all available experience and to evaluate in
a multi-disciplinary fashion the possible indications and contra-
indications for this procedure.

Psychiatrists, neurosurgeons and neuro-physiologists from both

countries meet since then on a regular basis to evaluate potential
psychosurgery candidates as they are referred by their psychia-
trists or hospital treatment teams. This self-composed "Committee
on Psychiatric Surgery" now reviews all such patients in the
Netherlands and the majority of Belgian candidates for psychosur-
gery. In the period of 1971 through 1981 a total of 71 cases was
submitted to this Committee by various therapists for evaluation.

II. FOLLOW-UP 1971-1981 OF OBSESSIVE-COMPULSIVE PATIENTS

These cases include 36 obsessive-compulsive patients; 15 men-

tal retardates with impulse control disorder (automutilation or
aggressive behavior) and 20 patients with various problems inclu-
ding affective disorders, anxiety disorders and personality dis-
orders. All diagnoses were made according to the now prevalent
DSM III standards. This study is exclusively concerned with the
36 obsessive-compulsive patients.

935
 The Committee would consider in its evaluation all patients
referred by their treatment psychiatrists for possible psychosur-
gery. However one of the most important criteria for psychiatric
surgery was a severe, long-standing and disabling disorder, not
responding to the usual available therapies. Thus, in effect, most
patients had a history of 10 to 15 years of chronic unremitting,
incapacitating, therapy-resistent obsessive compulsive disorder,
a rather negative basis for selection as psychosurgical candidates.

In order to receive a positive advice from this Committee

various other factors were considered, such as: the availability
of intensive clinical follow-up treatment after surgery; who would
be helped most by surgery (the patient, his surroundings or the
treatment team) and what changes will be necessary in the patients
life if surgery proved to be successful?

The attitude of his surroundings (significant others) was

considered to be an essential factor in these evaluations. The
patient himself was usually the last, but the most important,
person to be informed about plans for psycho-surgical treatment.
This was necessary in order to avoid undue pressure by the patient
on the Committee. Informed consent by the patient was the last
necessary step in this process.

SEVERITY COMPULSIVE BEHAVIOR S+ SURGERY

SCORE S- No SURGERY -------
100
S+
90

80
S+
70

60
-20 %
50

40

30
PATIENT'S EVALUATION THERAPIST'S EVALUATION
20

10

BEFORE ~.T FOLLO\HIP BEFORE AT FOLLOW-UP

Figure l

936
 The afore mentioned 36 obsessive-compulsive patients were
reached by questionnaire and through their current or last thera-
pists.

This group -with an average follow-up period of 6 1/2 years -

consisted of surgically treated patients (21 cases) and a group of
non-operated patients (15 cases). 'The no-surgery group consisted
of obsessive -compulsive patients not meeting the Committee's in-
clusion criteria (8 cases) and 7 cases that refused surgery in
spite of a positive recommendation by the Committee. The mean age
- for both groups - was rather similar (43 years for the operated
patients, 46 years for the non-operated group); 25 women versus 11
men were evaluated, but in the surgery group were eventually 12
women and 9 men. During follow-up a total of 6 deaths were registe-
red, 3 in the operated group (of natural causes) and 3 in the no-
surgery group (2 of these by suicide). Patients and their thera-
pists reported by return questionnaire as to their main symptoms
now and before surgery or - for the no-surgery group - compared to
about one year ago. As to the symptoms of compulsive behavior, ob-
sessional thoughts, anxiety and depression the figures 1 through
4 given the composite results. A remarkable and sterotyped cross-
over design is visualised: in regards to these 4 main symptoms all
surgically treated patients and their therapists (the solid lines
on the Fig.l to 4) report improvement, varying from 20 to 39;
SEVERITY
SCORE OBSESSIONAL IDEAS, THOUGHTS, IMAGES OR IMPULSES

S+ SuRGERY - - -

S+
s- No SURGERY ------

s-... ---
------.-13%
-33%

-39%

PATIENT'S EVALUATION THERAPIST'S EVALUATION

HEFORE AT FOLLOW-UP BEFORE AT FOLLOW-UP

Figure 2

937
SEVERITY ANXIETY
SCORE
100 S+ SURGERY <--1
S- No SURGERY(------)
90

80

70

60

50

40
PATIENT'S EVALUATION THERAPIST'S EVALUATION
30

20

10

BEFORE AT FOLLOW-UP BEFORE AT FOLLOW-UP

Figure 3

SEVERITY DEPRESSIVE MOOD S+ SURGERY - -

SCORE S- No SURGERY -----
100

90

80

-- --
S+ - _ ..+ 10 %
70

60

50 -26 %

40

30

PATIENT'S EVALUATION THERAPIST'S EVALUATION

20

10

BEFORE AT FOLLOW-UP BEFORE AT FOLLOW-UP

Figure 4

938
whereas the no-surgery group, by patient's subjective and thera-
pist's reporting (the interrupted lines on Fig.lto 4), indicate a
maximal improvement of only 3% and a worsening of symptoms of 21%.
In other words the combination of these data gives an impressive
illustration of the effects of psychosurgical intervention.

III. PSYCHOSURGICAL TECHNIQUES

Crow et al. reported in 1961 the fractionated leuco-coagula-

tion by means of implanted electrodes (the so called "chronic
method"). Knight et al. (1972) described the bifrontal stereotactic
tractotomy in the substantia innominata (the so called "acute
method").

In our groups, 9 patients underwent the chronic method while

12 had the acute procedure. As to the results in relation to the
type of surgery the following is evident: patients treated acutely
report a greater improvement in 3 main symptom areas as compared
to the patients of the chronic treatment method (mean is 42 versus
18!). The therapists report much less difference, but as far as
obsessional ideation is concerned, they tend to favor the chronic
method.

IV. DISCUSSION

All these results, taken together, lead us to conclude:

- the mean improvement score at follow-up reported both by patient
and therapist is definetely higher for the surgery group than
for the no-surgery group.
- this improvement is evident for the 4 main symptoms. The repor-
ted levels of distress were higher in all areas initially for
the surgical group, and decreased following surgery and subse-
quent therapy below any level of distress of the no-surgery group,
even though this group continued to receive their usual previous
therapies.
- many treatment resistent obsessive-compulsive patients apparent-
ly respond to various therapies after psychosurgery while these
same therapies proved to be ineffectual before. This applies to
psychotherapeutic procedures as well as to psychopharmacological
impact.
- the follow-up results of the no-surgery group, regardless of the
fact whether they were approved on disapproved by the Committee
for psychosurgery, indicate a specific and important influence
of the psychosurgical intervention. The no-surgery group, 1n
spite of continued therapies, showed no improvement.
- in spite of the reported positive results for the surgery group
as a whole we must conclude that a certain amount of psychiatric
disability remains: most patients were socially under-productive
and dependent on disability payments. Yet subjective improvement

939
 remains a very much worthwhile treatment goal.
the better results of the "acute" method lead us to decide against
future use of the "chronic" method. In this way we will save the
patient a lot of distress and reduce the time and energy spent
by neurosurgeons, neurophysiologists and various therapeutic
personel during these chronic procedures.

V. BIBLIOGRAPHY

-Crow H.J., Cooper R., Phillips D.G.: Controlled multifocal

frontal leucotomy for psychiatric illness. J.Neurol.Neurosurg.
Psychiatry, 24 (1961), 353-360.
-Knight G.G.: "Bifrontal sterotaxic tractotomy in the substantia
innominata". In: Psychosurgery, ed. Hitchcock E., Laitinen L.
and Vaernet K., Thomas Ch. 1972.

940
AUTHOR INDEX

Ackenheil, M., 21, 215 Breitbach, W., 623

Ackermann, R. H., 293 Bretschneider, S., 389
Agathon, M., 861 Breuer, H., 243
Albus, M., 215 Brick, I. , 347
Alevizos, B., 279 Bridges, P. K., 931
Alpert, N. M., 293 Brown, G. A., 619
Altmann, H., 577 Bruinvels, J., 193
Anders, T. F., 775 Brun, A., 605
Angst, J., 67 Bunney, W. E., Jr., 43
Arato, M., 187 Busse, E. W., 9
Arendt, G., 567
Aschauer, H., 249 Caraceni, T., 273
Astrup, C., 841 Carruthers, M., 833
Carvey, P.M., 543
Baer, L., 293 Cazzullo, C. L., 885
Baldessarini, R. J., 561 Ceha, J. A.M., 935
Banki, C. M., 187 Chiarenza, G. A., 885
Barry, D. I. , 111 Christensen, N. J., 111
Bartlett, J. R. , 931 Cobb, M., 871
Bedard, P., 237 Comar, D., 289
Benkert, 0., 223, 229 Cookson, J. C., 431
Benson, H., 825 Cooper, R., 159
Berger, M., 383 Correia, J. A., 293
Bergstrand, G., 37 Corsino, G. U., 233
Bertha, G., 629, 667 Corson, S. A. , 679, 855, 895
Bitter, I., 525 Cosyns, P., 935
Bjerkenstedt, L., 37 Crawley, J. , '755
Bliestle, A., 389 Crouzel, M., 289
Blizard, R., 871 Cuk, M., 635
Blomquist, G., 305
Bolwig, T. G., 589 Darragh, A., 347
Bondy, B. , 215 Davis, J. M., 159, 463
Bossert, s., 389 DePaulis, T., 305
Boudoulas, H., 765 Del Zompo, M., 233
Boulenger, J.-P., 693 DiPaolo, T., 237
Bourgoin, S., 341 Dirlich, G., 367

941
Doerr, P., 367, 383 Haefely, W., 105
Dorus, W., 463 Haeggstroem, J.-E., 555
Dreher, A., 521 Hall, M., 341
Duhm, J., 449 Hamon, M., 341
Hantraye, P., 289
Ehrin, E., 305 Hartmann, A., 593
El Mestikawy, S., 341 Harvey, J. E., 801
Emrich, H. M., 367 Hedstroem, C. G., 305
Erikson, L., 305 Heinrich, K., 567
Hemmingsen, R., 589
Farde, L., 305 Higgs, C. M. B., 801
Fehm, H. F., 389 Hippius, H., 21
Fichter, M. M., 721 Hiramatsu, K.-I., 232
Flaherty, J. A., 177 Hirschowitz J., 457
Frackowiak, R.S.J., 299 Hitzemann, R., 457
Frankenhaeuser, M., 699 Hoehn-Saric, R., 705
Friedel, R. 0., 485 Hollister, L. E., 493
Friedman, R., 819, 825 Holsboer, F., 223, 229, 721
Fritz, G. K., 789 Hoyer, S., 641

Gabriel, E., 183 Ingvar, D. H., 305

Gallhofer, B., 299 Insel, T., 755
Gandolfo, C., 661 Itoh, K., 321
Garfinkel, P.E., 729
Garland, B., 43 Jacobs, D., 435
Garner, D. M., 729 Janka, Z., 471
Garver, D. L., 457 Jarrett, D. B., 377
Gaviria, F. M., 165, 171, 177 Jellinger, K., 203
Gibbons, R. , 159 Jones, T., 299
Gillman, M. A., 397
Giovannini, P., 273 Kaijima, M., 289
Gierris, A., 111, 115, 197 Kalinowsky, L. B., 573
Goetz, C. G., 529, 543 Kameyama, T., 321
Gold, P. W., 417 Kempin, F., 721
Goldberg, H., 759 Kenny, M., 347
Gozlan, H., 341 Keshavan, M.s., 249
Grajeda, F., 623 Kimmel, H. D. , 901
Greitz, T., 305 King, B. D., 765
Griffith, J. L., 293 Kitt, W., 463
Gruzelier, J., 313 Klawans, H. L., 543
Guibert, B., 289 Klebl, H., 613
Gunne, L. M., 555 Klieser, E., 567
Gunthern, G., 685 Knopp, W., 921
Gustafson, L., 605 Koehler, D., 619
Koester, G., 243
Haaijman, w. P., 935 Koinig, G., 249
Haase, H.-J., 515 Kreuz, A., 229
Hackett, T. P., 293 Krieg, J.-Ch., 383

942
Krueger, G., 649 Mills, I. H., 747
Kryspin-Exner, K., 15 Montgomery, S. A., 405
Kupfer, D. J., 377 Mueller, E. E., 273
Mueller-Spahn, F., 215
Laakmann, G., 257 Muench, U., 215
Labrie, F. , 237
Ladurner, G., 613, 644 Naber, D., 215
Lahmeyer, H. W., 165, 171 Naquet, R., 289
Lambe, R., 347 Naylor, G. J., 353
Lange, H., 567 Nilsson, J. L. G., 305
Langer, G., 249 Ninan, P., 755, 221
Lechner, H., 613, 667 Niwa, S.-I., 321
Lehmann, E., 567 Nyba·eck, H. , 37
Lenehan, T., 347
Lerer, B., 577 O'Leary Corson, E., 679, 855, 895
Levey, A., 871 Okonek, A., 159, 463
Levine, s., 769 Orlebeke, J. F., 905
Lichtigfeld, F. J., 397 Ostrow, D. G., 159, 443, 463
Linkowski, P., 263 Ott, E., 667
Linnoila, M., 221, 417 Ottosson, J.-0., 583
Liparaki~ M., 279 Oxenstierna, G., 37
Lipowski, Z. J., 1
Litt, I. F., 795, 807 Papadimitriou, G., 279
Litton, J.-E., 305 Papp, Z., 187
Loeb, C., 661 Parati, E., 273
Lund, R., 367, 371 Parnas, J., 85
Lundberg, U., 699 Paul, S., 755
Peet, M., 361
Maltzmann, I., 913 Pepplinkhuizen, L., 193
Mann, L. s., 221 Perez, F. I., 619, 623
Marsden, C. D., 537 Petsche, H., 879
Marshall, J., 657 Pfohl, B., 77
Martin, B., 353 Piccardi, M. P., 233
Martin, I. , 871 Pichat, L. , 341
Matussek, N., 397 Pirke, K. M., 367, 721, 383
Maulet, C., 347 Ploog, D., 717
Maze!, B., 861 Pockberger, H., 879
Mazer, C., 783 Post, R. M., 417, 693
Maziere, M., 289 Prasad, B. R., 171, 165
McCaughran, J. A., Jr., 819 Prenant, C., 289
McLeod, D. R., 705
Medlicott, L., 747 Quadbeck, H., 571
Mednick, S. A., 85
Meier, L., 229 Rafaelsen, 0. J., 57, 111, 115,
Meloni, M. , 233 197
Mendlewicz, J., 95, 263, 479 Rappelsberger, P., 879
Mesec, A., 635 Ravichandran, G. R., 623
Miller, J. G., 673 Resch, F., 249

943
Reynolds, G. P., 209 Val, E., 165, 171
Risberg, J., 605 Van Kammen, D. P., 221
Rivera, V. M., 619. 623 Van Kammen, W. B., 221
Robinow, D. R., 417 Van Praag, H. M., 61, 121, 327
Rusin, M., 619 Van Puten, T., 509
Rymar, K., 321 Vander Molen, M. W., 905
Vittone, B., 693
Saitoh, 0., 321 Voigt, K. H., 389
Sastre, J., 289 Volk, s., 371
Scheinin, M., 221 Von Zerssen, D., 367, 383
Schepank, H., 101 Vorstrup, S., 589
Schoenbeck, G., 249 Vranckx, C. H., 927
Schulsinger, F., 85 Weiler, M., 171
Schulsinger, H., 85 Weiss, W., 721
Schultz, H., 371 Wheatley, D., 813
Scigliano, G., 273 Wiesel, F.-A., 37, 305
Sedvall, G., 37, 305 Widen, L., 305
Sever, A., 635 · Widepalm, K., 583
Sheehan, D. V., 711 Wik, G., 305
Sheehan, K. E., 711 Williams, J. R., 789
Sheer, D. E., 867 Winokur, G., 77
Sherman, B. , 77 Wise, R.J.S., 299
Siever, L. J., 693 Woeller, W., 571
Sijben, N., 935 Wolpe, J., 849, 917
Silverstone, T., 423 Wooley, C. F., 765
Skolnick, P., 755 Wulff, E., 623
Sobieski, J., 783
Soldatos, C., 279 Yesavage, J., 499
Somsen, R. J. M., 905
Stanley, M., 577
Stefanis, C. N., 279
Steiner, H., 769, 775, 783,
795
Stone-Elander, s., 305
Sulser, F. , 411
Surman, 0. s., 293
Szentistvany, I., 471

Takahashi, R., 51
Tanner, C. M., 543
Teasdale, Th. W., 85
Tegler, J., 571
To lis, G., 279
Trimble, M.R., 299
Tschinkel, M., 613
Turner, P. , 151

Uhde, T. W., 693

944
SUBJECT INDEX

Acethylcholine, 423, 577 Asthma, 801

ACTH secretion, 77 Attention, 293, 905
Acute psychoses, 193 Atypical depression, 115
Acute single-test dose Auditory evoked potentials, 321
method, 499 Auditory feedback, 861
Adjustment disorder, 187 Autogenic training, 833
Adolescents, 769 Autonomic functions, 855
Adrenaline, 111,115,197
Affective disorders, 67, 95, Beclomethasone, 801
171, 449, 463, 471, 479 Behavior genetics, 855
Aging brain, 561 Behavior therapy, 861
Agoraphobia, 711 Benzodiazepine receptor, 289, 755
Alcohol dependence, 187, 397 Benzodiazepines, 927
Alzheimer's disease, 605, 657, Beta-adrenoceptors, 411
661 Biofeedback training, 867
Amines, 111 , 115 , 197 Biological aspects, 57, 197
Amino acids, 9 Biological correlates, 187
Amitriptyline, 361 Biological markers, 57, 159, 197
Amphetamine, 435, 927 Biological rhythms, 43
Amygdala, 209 Bipolars, 263
Angina, 813 Blood pressure, 705, 833
Animal models, 555 Body temperature, 367
Anorexia nervosa, 717, 721, Borderline personality, 165, 171
729, 747 BPRS, 499
Anterograde anmesia, 577 Brain, 111
Anthropology, 685 Brain atrophy, 203
Anticholinergic, 347, 361 Brain cells, 471
Antidepressant, 105, 111, 115, Brainstem gliosis, 203
197, 347, 353, 361, 405 Buspirone, 341, 813
Antidepressant treatment
response, 159 Cancer survivors, 789
Antipsychotic drugs, 561 Cardiac interbeat interval, 705
Anton's syndrome, 623 Cardiovascular alterations, 819
Anxiety, 341, 693, 711, 755, Cat scan, 623
759, 765, 813, 849, 861, Catalepsy, 237
905 Cerebral blood flow, 299, 589

945
Cerebral hypoxia, 203 Couple therapy, 685
Cerebral lateralization, 321 Covert behavior, 895
Cerebral metabolic rate, 299 Covert reactions, 855
Cerebrospinal fluid, 111, 417 CSF, 237
Cerebrovascular attack, 635 CSF 5HIAA, 187
Cerebrovascular disease, 593, CSF cortisol, 187
641, 649 CSF HVA, 187
Chemical messengers, 9 CSF tryptophan, 187
Childhood cancer, 789 CT in dementia, 661
Chlorimipramine, 257 CT scan, 657
Chlorpromazine (CPZ), 509 Cushing's syndrome, 383, 389
Cholesterol, 833 Cycloid psychoses, 67
Chorea, 293
Chronic mental illness, 561 Danger-related measure, 499
Chronic schizophrenia, 493 Defense mechanisms, 775
Chronobiology, 367 Definition of depression, 67
Circadian pacemaker, 367 Degenerative disorders, 9
Circadian rhythm, 371 Delirium, 1
Circulation of cerebrospinal Dementia, 1, 15, 613, 629, 641,
fluid, 37 657' 661
Classification, 67, 711 Depression, 57, 61, 77, 121, 151,
Clinical response, 509 159, 177' 187' 197' 233,
Clonidine, 693 249, 353, 371, 377, 383,
Clorazepate, 813 389, 397, 405, 711, 721
Cognitive coping, 867 Desimipramine, 257
Cognitive deficits, 15, 321 Deuteranopia, 95
Complex structures, 841 Dexamethesone suppression, 77,
Complexity, 921 721' 795
Compulsive disorder, 717, 747, Diagnosis, 457, 711
935 Diagnostic specificity, 187
Computed tomography, 37 Diazepam, 257, 813
Computer assisted EEG, 879 Differential arousal, 921
Conditioning, 849, 861, 871 Differential diagnosis, 605
Conditional reflexes, 841 Differential diagnosis in
Clonidine-test, 721 dementia, 661
Constitutional difference, 855, Disactivation hypothesis, 249
895 Dominant focus, 913
Continuum of depression, 67 Dopamine, 43, 209, 423, 435
Coping mechanisms, 775 Dopamine receptors, 209, 233, 237,
Coping with stress of chronic 305
illness, 789 Dopamine system, 51
Coronary heart disease, 813, Dopaminergic neurons, 717
921 Dose-effect relationship, 561
Coronary-prone behavior, 895 Drug-induced switches, 43
Corson typology, 927 DSM3, 711
Cortical blindness, 623 Dysfunction of GABA neurons, 555
Cortisol, 367, 431, 721 Dyskinesia, 237, 537, 561
Cortisol/ACTH, 257

946
Eating disorders, 795 Fluorescence spectrophotometry,
Eco-anthrophology, 685 499
Ecosystem, 698 Focal neurological symptoms, 649
ECT, 573 Forced expiratory flow rates, 801
EEG, 165 Functional psychopathology, 61
Electroconvulsive therapy,
577, 583, 589 G-6-PD, 95
Electrodermal skin activity, Gaba, 423
705 GABA-ergic treatment, 567
Electromyographic biofeedback Gas chromatography, 485, 499
(EMG BFB)t 861 Gastrointestinal hormones, 717
EMG biofeedback, 867 Generalized anxiety disorder, 705
Emotional stress, 383 Genetic-physiologic interaction,
Endocrine dysfunction, 721 463
Endocrinology, 223 Genetics, 67, 101, 479
Endogenous, 711 Geriatrics, 813
Endogenous depression, 229, Glucocorticoid feed-back, 389
257, 367 11-C glucose, 305
Endorphine, 717 Goal-directed behaviour, 885
Entrainment, 367 Gonadal steroids, 243
Environment, 85 Group therapy, 685
Epidemiology, 561 Growth hormone, 257, 397, 431, 721
Epilepsy, 299
Erythrocytes, 449 Habituation, 855, 905
Essential hypertension, 449 Hachinski score, 657
Estrogen-Vit B treatment, 229 Haloperidol (HPL), 509
Estrones, 237 HDL, 833
Etiology, 795 Healthy subjects, 257
Event-related potentials, 321 Heart rate, 861
Exogenous, 711 Heredity, 85
Experimental neuroses, 917 High performance liquid
Extinction failure, 871 chromatography, 485
Extrapyramidal disorders, 273 Higher nervous activity, 841
Extrapyramidal fine motoricity, Hippocampus, 341
515 Homotypology, 67
Homovanillic acid (HVA), 237
Family adjustment, 789 Hospitalization, 769
Family history, 37 Human brain, 209
Family therapy, 685 Human ecology, 685
Feed-back inhibitory control, Huntington's disease, 273
341 Hypercortisolism, 389
Feeding behaviour, 717 Hypertension, 657
Feighners research criteria Hypochondriasis, 711
for schizophrenia, 183 Hypophysectomy, 237
FFA, 833 Hypothalamus, 9, 717
Fine motorsysmptoms, 521, 525 Hypothalamus-pituitary axis, 273
Fixed dose, 509
Idealism, 679

947
Indalpine, 347, 353, 361 Mood disturbances , 795
Individual therapy, 698 Motor activity, 367
Information processing, 321 Movement and ageing, 885
Integrative materialism, 679 Movement and development, 885
Isolated symptoms, 187 Movement and goal-directed
Isoprotereno l, 765 behavior, 885
Movement and psychopatholo gy, 885
Lacunes, 657 Movement disorder, 293
Laterality, 313 Movement related brain
Left hemispheric dysfunction, macropotenti als, 885
321 Multi infarct dementia, 605, 657,
Life events, 177 661, 667
Lithium, 43, 443, 449, 457 Multidiagnos tic approach, 183
Lithium metabolism, 479 Muscarinic cholinergic receptors,
Lithium ratio, 449, 457 577
Lithium transport, 443, 463, Muscular tension, 861
471 Myocardial infarction, 813
Low-pressure narcosis, 583
Luteinizing hormone, 397 NA subsensitivi ty and anti-
depressants, 411
Major psychoses, 203 Na+-H+ exchange, 449
Malnutrition , 383, 721 Na+-Li+ exchange, 449
Mania, 43, 417, 423, 431, 435, Na+-Na- exchange, 449
457 Naloxone, 397, 747
Manic depressive psychoses, Nervous system, 901
443, 841 Neuroendocri nology, 215, 243, 383,
MAO, 115 389
MAO-inhibito rs, 411 Neurogenic theory, 679
Mechanistic materialism, 679 Neurohormones, 9
Melatonine, 263 Neuroleptics , 209, 237, 249, 299,
Membrane, 449 493, 515, 521, 525, 529,
Membrance-plasma interaction, 537, 543, 567
463 Neuropatholo gy, 203, 605
Memory disturbance, 583 Neuropeptide s, 9
Menses, 813 Neuropsychol ogic tests, 635
Mental process, 879 Neuropsychol ogical rehabilitatio n,
Mesolimbic system, 209 619
Methysergide , 257 Neurophysiol ogical testing, 623
Mianserine, 353 Neuropsychology, 293
Mitral value prolapse, 765 Neuroradiolo gy, 203
Mode of thinking, 635 Neuroses, 711, 841, 861
Modulation, 921 Neurotic depression, 257
Monkey, 305, 555 Neurotransm itter, 9, 57, 151, 197,
Monoamine metabolites, 37 243, 417
Monoamine oxidase activity, 51 Nigral and pallidal GAD reduction,
Monoamineoxidase inhibitors, 555
111, 115, 197 Nitrous oxide, 397
Monoamines, 9, 105 Nomifensine, 257

948
Noradrenaline, 111, 115, 197, Plasma level, 499, 509
411' 435 Platelet, 347
Noradrenergic function, 693 Pleasure centers, 717
Noradrenergic neurones, 717 Polypeptides, 9
Norepinephrine reactivity, 825 Positron emmission tomography,
Novelity, 913 289, 293, 299, 305
Nucleous basalis of Meynert, Post mortem studies, 183
203 Post-mortem, 209
Postsynaptic 5-HT receptors, 341
Obsessive, 935 Prediction, 921
Occipital infarcts, 623 Presenile dementia, 605
Oestrogen, 233 Presynaptic 5-HT autoreceptors,
Operant conditioning, 867 341
Opiates, 747 Presynaptic effects, 151
OR, 913 PRL, 263
Organic brain syndromes, 1, 15, Prognosis, 841
629, 649 Prolactin, 237, 257, 431
Organic psychoses, 203 Propranolol, 257
Organismic view, 685 Protanopia, 95
Orienting reflex, 901, 913, Pseudo-dementia, 605
917' 927 Psychedelic symptoms, 193
Orienting responses, 895 Psychiatric surgery, 935
Overt behaviour, 855, 895 Psychiat!'y, 711
Psychobiology, 43
Panic, 711 Psychodinamic pattern, 635
Panic attacks, 861 Psychoendocrinology, 795
Paranoid-hallucinatory Psychogenic diseases, 101
syndrome, 249 Psychogenic hypertension, 819
Parkinson's disease, 273 Psychologic stress, 855
Passive-avoidance training, 577 Psychological testing, 801
Pathochemistry, 203 Psychomotor, 921
Pavlovian conditioning, 855 Psychoneuroendocrinology, 389, 699
Peptide, 57, 197 Psychoneuroses, 101
Peptide alterations, 417 Psychopathology, 183
Perception, 801 Psychopharmacology, 61, 561
Personality, 833 Psychophysiologic disorders, 895
Personality disorders, 101 Psychophysiologic lability, 855,
Pharmacodynamics, 485 895
Pharmacokinetics, 485 Psychophysiological response
Pharmacological model, 435 patterns, 705
Phentolamine, 257 Psychoses, 299, 457, 537, 561
Phobias, 711 Psychosomatic illness, 867, 871
Physical training, 833 Psychosomatic patients, 783
Pick's disease, 605 Psychotropic effect, 229
Pimozide, 435
Pituitary, 237 Radioimmunoassay, 485
Pituitary gland, 9 Radioreceptor assay, 493
Plasma concentrations, 493 Range of oscillation, 367

949
Rats, 555 Speech, 775
Reaction time, 613 Spiroperidolbind ing in brain
Reactive psychoses, 841 tissue, 183
Reactivity, 928 Starvation, 721
Receptor binding, 215 Steady state, 499
Red blood cells, 471 Stress, 367, 795, 813
Regional cerebral blood flow, Stress research, 699
593, 605 Stress response, 769, 775, 783,
Relaxation methods, 867 801, 807, 819, 825
Relaxation response, 825 Striatum, 341
REM latency, 371 Stroke patient, 619
REM sleep, 371, 377 Substituted benzamides, 305
Reproductive systems, 807 Sulpirid/haldol, 257
Rheological therapy, 593 Switch process, 43
Sympathetic-adre nal medullary
Salbutamol, 801 system, 699
Saliva, 361 System analysis, 673
Salivation, 367 Systems biology, 679
Scanning, 299 Systems science, 679, 685
Schizoaffective psychoses, 67 Systems theory, 673
Schizophrenia, 37, 51, 85, Systems therapy, 685
183, 187, 203, 209, 215,
299 ' 305' 313' 321 ' 457 ' Tardive dyskinesia, 529, 543, 555,
529, 561, 573, 841 561, 567
Schizophreniform disorder, 457 Testosterone, 223
Selective attention, 321 Theophylline, 801
Selfstimulation, 717 Therapeutic mechanisms, 249
Senile dementia, 657 Therapeutic window, 485, 509
Serine-glycine metabolism, 193 Thermal biofeedback, 867
Serotonin, 121, 347, 411 Thiothixene (THX), 493, 509
Serotonin (5-hydroxytryptam ine, Thymoxamine, 435
5-HT), 341 Thyrotropin-relea sing-hormone
Serotonin-noradr enaline link- (TRH) test, 249
hypothesis, 411 Transitory ischemic attack, 635
Sex, 813 Transport, 449
Sex ratio, 67 TRH-test, 721
Sexual behaviour, 341 Tricyclic antidepressants, 411
Sexual dysfunction, 223 Triglycerides, 833
Significance, 913 Trimethaphan, 583
Sleep, 371, 377, 813 TSH, 263
Sleep disturbances, 367 Type A, 833
Sleep-recording, 165
Social aspects, 177 Unconditional reflexes, 841
Social sciences, 685 Unipolars, 263
Somatovisceral dichotomy, 895 Uric acid, 833
Social environment, 783
Specific high affinity Validation, 67
binding, 341 Vascular risk factors, 667

950
Vasoactive intestinal poly
peptide, 57, 197
Vasopressin, 927
Verapamil, 813
Vienna research criteria for
schirzophrenia, 183
Visual anosognosia, 623
Visual evoked responses, 623
Vitalism, 679
Word association, 841

X-linkage, 95
133-Xe inhalation, 605

Yohimbine, 257, 693

951
 SUMMARY

CONTENTS OF

VOLUMES l-8

953
 VOLUME I

CLINICAL PSYCHOPATHOLOGY

NOMENCLATURE AND CLASSIFICATION

Nomenclature and Classification

Reasons for Choosing the Axes in Multiaxial Classification in Psychiatry
International Problems of Communication about Diagnoses and
Symptomatology in Epidemiological Psychiatry
The Polydiagnostic Approach in Psychiatric Research
Comprehensive Psychopathological Scales and Systems
Translations of the AMDP System
Strategies and Tactics in Psychiatric Research
(Strategies et tactiques dans la recherche en psychiatrie)
Plus and Minus Symptomatology ih Schizophrenia
Newer Concepts of Basic Disorders and Basic Symptoms in Endogenous
Psychoses
Actual Problems in Paranoid Psychoses
Aetiology of Depression in Schizophrenia
Chronicity and Personality in Affective Disorders
Joining Together the Clinical Assessment Systems for Depression
Psychopathology of Organic Psychosyndromes
Concepts of Neurotic and Personality Disturbances
Panic and Other Anxiety Disorders: Diagnosis and Treatment
The Delusional Misidentification Syndromes
Psychopathology of Self-Destructive Behaviour
Research in Suicidology
Methodological Problems in Therapeutical Trials
(Problemes methodologiques dans l'essai therapeutique)
Evaluation of Treatment Methods
Systematic and Systemic Psychopathology
(Systematische und systemische Psychopathologie)
Psychopathology of Expression
Psychopathology in the Crisis Situation. Third German-Japanese-Austrian
Encounter
(Psychopathologie in der Krise? 3. Deutsch-japanisch-1:isterreichisches
Treffen)

954
 VOLUME 2

BIOLOGICAL PSYCHIATRY

HIGHER NERVOUS ACTIVITY

BIOLOGICAL PSYCHIATRY
Biological Aspects - Organic Brain Syndromes
Biological Aspects - Functional Psychoses
Genetic Aspects of Psychiatry
New Prospects in the Treatment of Depression
Clinical and Research Aspects of Affective Disorders
Pathochemical Markers in Major Psychoses
Steroids in Psychiatry
Frontiers in Psychoneuroendocrinology
Positron Emission Tomography I+ II
Laterality and Psychopathology
Serotonin and Disturbances of Mood
(Serotonine et troubles de l'humeur)
Psychobiology of Depression -Recent Findings and Theoretical Models
The Old Amine Theory and New Antidepressants
Biology of Mania
Lithium Transport Research: From Cellular Membrane to Clinical Practice
Clinical Applications of Plasma Levels in the Management of Schizophrenia
Dosing Neuroleptic Medication with the Help of Electric Registration of
Extrapyramidal Fine Motoricity
(Dosierung der Neuroleptika mit Hilfe der elektronischen Registrierung
der extrapyramidalen Feinmotorik)
Movement Disorders and Tardive Dyskinesia
ECT: Background and Current Research
Psychic Changes in Patients with c;erebrovascular Diaseases I + II
Systems Science and Systems Therapy
Psychobiology of Anxiety
Psychobiology of Anorexia Nervosa
Physiological Basis of Anxiety
The Development of Human Stress Response: Research Findings and Clinical
Application
Stress and the Heart
HIGHER NERVOUS ACTIVITY
Physiological Investigations of Psychological Processes in Health
and and Psychiatric Diseases
Orienting Reflexes in Psychophysiological Health and Disease

955
 VOLUME 3

PHARMACOPSYCHIATRY

Pharmaco EEG and Psychometry in Early Psychopharmacology

Phase-IV Research in Psychiatry: Methodology and Objectives
Drug Strategies for Altering Cognitive Functions
Predictors - New Antidepressants
Treatment of Therapy-Resistant Depression
Depression and its Treatment in the Problem Patient - Experiences with
Bupropion, a New-Gerneration Antidepressant
Early Onset of Action and Safety in Antidepressant Therapy with Trazodone
- New Clinical Data
Pirlindole in the Management of Depressed Patients in Clinical Practice
Psychopharmacology of Suicidal Behaviour
Psychotic Disorders Responsive to Lithium
Lithium Withdrawal Studies, Treatment of Non-Responders
The Use of Anticonvulsants in Affective Disorders
Benzodiazepine Research in the 1980's
Alternatives to Treatment with Minor Tranquillisers
Recent Developments in Substituted Benzamide Drugs
Mondern Trends in the Chemotherapy of Schizophrenia
Special Patient Populations and their Unique Treatment Demands
Depot Medication in Chronic Schizophrenia
International Workshop "NB 106-698 - A Non-Classical Neuroleptic"

956
 VOLUME4

PSYCI-IOTHERAPY

PSYCI-IOSOMATIC MEDICINE

PSYCHOTHERAPY
New Paradigms in Psychotherapy
Short-Term Psychotherapies
Non-Verbal Aspects and Techniques of Psychotherapy
Initiating and Developing the Process of Family Therapy
Psychotherapy of Depression
Comprehensive Management of Mood and Emotion
Psychotherapy for the Developing World
Psychoanalytic Hospital Treatment
Group Psychotherapy in Psychiatry
(Gruppenpsychotherapie in der Psychiatrie)
Group Psychotherapy and the Educational Process
Uses and Abuses of Behavioral Therapy
(Uses y abases de Ia terapfe de Ia conducta)
Assertive Behaviour: An Overview and Conclusion
New Developments in Obsessive-Compulsive Disorder
Overview and Current Research in Relaxation and Imaging Techniques:
Some Current Applications of These Techniques in the Practice of
Psychiatry
Neurolinguistic Programming
Occupational Therapy
PSYCHOSOMATIC MEDICINE
Psychosomatic Pathology as a Developmental Failure: A Model for Research
Environmental Stress Factors and Their Psychosomatic Correlates
Relationship between Psychiatry and Psychosomatic Medicine
Issues in Liaison Psychiatry: A Model for Education, Research, and Patient
Care
Psychophysiological Risk Factors of Somatic Disorders: Methodological and
Transcultural Aspects
The Psychosomatic Medicine of the Year 2000

957
 VOLUME 5

CHILD AND ADOLESCENT PSYCHIATRY

MENTAL RETARDATION

GERIATRIC PSYCHIATRY

CHILD AND ADOLESCENT PSYCHIATRY

Child Psychiatry
Training in Child Psychiatry - Paediatric Neuropsychiatry
Treatment in Child Psychiatry
Psychopharmaceutics in Child and Adolescent Psychiatry
(Psychopharmaka in der Kinder- und Jugendpsychiatrie)
Social and Transcultural Child Psychiatry
The Immigrant Child in Canadian Society
Puerperal Mental Illness
MENTAL RETARDATION
Mental Retardation I + II
GERIATRIC PSYCHIATRY
Dementia in Late Life
Recent Advances in Psychophysiological Disorders of Old Age
Neurotransmitters in the Dementias of Old Age and Their Implications for
Treatment
Dementias as Multi-Stage Processes
Mental Deterioration in the Aged - Aspects of Recognition, Diagnosis, and
Management
Psychopharmacology in Late Life
Services to the Elderly - Evaluation of Effectiveness
Day Hospital Care for the Elderly

958
 VOLUME 6

DRUG DEPENDENCY AND ALCOHOLISM

FORENSIC PSYCHIATRY

MILITARY PSYCHIATRY

DRUG DEPENDENCY AND ALCOHOLISM

The Challenge of Addiction and Substance
Use Disorders: Are they the Number One Health Problem for
Psychiatry?
Diagnosis of Alcoholism
Biological Psychiatry of Chronic Alcoholism and of the Postwithdrawal
Syndrome
FORENSIC PSYCHIATRY
New Horizons in Psychiatry and Legislation
The Role of the Psychiatrist as an Expert in Court Procedures: Changing .
Patterns and New Trends
The Psychiatrist's Response to Repression
Mental Health Needs of Victims of Violence
The Role of Psychiatry in the Treatment of Delinquents in a Correctional
Setting
Infanticide and Incest: Cross-Cultural Perspectives
Terrorists and Terrorism
Mass Murder
MILITARY PSYCHIATRY
International Studies in Military Psychiatry I + II

959
 VOLUME 7

EPIDEMIOLOGY AND COMMUNITY PSYCHIATRY

Sociological Aspects of Psychiatry

Mental Health Service Research
Supportive Networks and the Incidence of Mental Disorders
General Hospital Psychiatry
International Trends and National Progress in Psychosocial Rehabilitation
Organisational Problems and New Trends in Psychiatric Services
Comparative Approaches to Initiating Mental Health Care in Primary
Settings
Community Mental Health Services in Urban Areas
The Politics of Community Care
Primary Care and Mental Health Services
New Trends in In- and Out-Patient Care
Manpower Requirements in Psychiatric Hospitals
Schizophrenia: On Current Treatment, Research, and Social Considerations
Helping the Family to Cope with Schizophrenia
Antidepressants and Active Life
(Antidepresseurs et vie active)
The Year After- Results of a Follow-up Study on Newly Admitted
Psychiatric Patients
(Das Jahr danach - Ergebnisse einer Uingsschnittuntersuchung an
ersteingewiesenen psychiatrischen Patienten)
Results of the Internal Reform in Psychiatry
(Ergebnisse der inneren Psychiatriereform)

960
 VOLUME 8

HISTORY OF PSYCHIATRY

NATIONAL SCHOOLS

EDUCATION

TRANSCULTURAL PSYCHIATRY

HISTORY OF PSYCHIATRY
The History of Mental Hospitals
The History of Psychiatry
Psychiatry and Ethics
Psychiatry and World Affairs
NATIONAL SCHOOLS
National Schools and World Psychiatry
Psychiatry in Spanisch and Portuguese Speaking Countries
(La Psiquiatrfa en paises de habla espanola y portuguesa)
EDUCATION
Graduate Education: The Road to International Certification and
Accreditation in Psychiatry
Basic Psychiatric Education for All Medical Students
Education in Psychiatry of the Non-Psychiatrists
The Use of Videotape in Psychiatric Education
Psychiatry and Audiovisual Means
(Psiquiatrfa y medias audiovisuales)
TRANSCULTURAL PSYCHIATRY
Training in Cultural Psychiatry
Mental Health Care in the Context of Differing Cultural Needs and Settings
Cultural Psychiatry - An Update
Aspects of Treating Hispanic Patients in the USA
Conflicts and Mental Health Problems of Migrants and Their Families
Immigrants' Acculturation
Cultural Aspects of Family Assessment and Therapy
Neurasthenia, Somatic Complaint Syndromes, and Depression: A
Transcultural Exploration of Semantics in Psychiatry
Traditional Healers and Folk Therapies
Sexology in the Next 25 Years
Sports and Psychiatry
(Sport et psychiatrie)

961