JOURNAL OF CHILD AND ADOLESCENT PSYCHOPHARMACOLOGY

Volume 7, Number 3, 1997

Mary Ann Liebert, Inc.
Pp. 145-155

Buspirone Treatment of Psychiatrically Hospitalized

Prepubertal Children with Symptoms of Anxiety and
Moderately Severe Aggression
CYNTHIA R. PFEFFER, M.D., HONG JIANG, M.P.H., and LISA J. DOMESHEK, B.A.

ABSTRACT

Open-label buspirone was studied in 25 prepubertal psychiatric inpatients (age 8.0 ± 1.8
years, 76% boys) presenting with anxiety symptoms and moderately aggressive behavior.
Patients with severe aggression, requiring rapid treatment with mood stabilizers or neu-
roleptics, were excluded. A 3-week titration (maximum 50 mg daily) preceded a 6-week main-
tenance phase at optimal dose. Buspirone was discontinued in 6 children (25%): 4 devel-
oped increased aggression and agitation, and 2 developed euphoric mania. For the 19 patients
who completed the study, mean optimal dose was 28 mg daily. Among completers, depres-
sive symptoms were reduced 52% by Week 6 on Children's Depression Inventory (p < 0.001).
Decreased aggressivity was reflected in a 29% reduction on Measure of Aggression, Vio-
lence, and Rage in Children [MAVRIC] ratings (p :S 0.02) and in 86% less time in seclusion
or physical restraints (p < 0.02). Clinical Global Assessment scores improved (CGAS 41 vs.
54, p < 0.01). Only 3 children improved sufficiently to continue buspirone after the study.
Residual aggressivity and global functioning remained problematic. Buspirone may pose be-
havioral risks in treating moderate aggressivity in 24% of children with anxiety; in the oth-
ers, the therapeutic effects on aggression, anxiety, and depression were limited but signifi-
cant.

INTRODUCTION

ANXioLYTic, is an azapirone that exerts effects on the serotonin neu-

Buspirone, an NONBENZODiAZEPiNE
rotransmitter system by acting as an agonist at 5-HTla presynaptic receptors and as a partial agonist at
postsynaptic 5-HTla receptor sites (Robinson et al. 1990; Sramek et al. 1996; Wheatley 1982). Buspirone
acts to reduce the outflow of serotonin from the presynaptic neuron, and it binds to the postsynaptic re-
ceptor to further reduce serotonergic activity in the central nervous system (CNS). Buspirone has little or
no addicting and withdrawal effects (Robinson et al. 1990).
Several controlled studies suggest that buspirone is efficacious for the treatment of generalized anxiety

Department of Psychiatry, The New York Hospital-Cornell Medical Center, Westchester Division, White Plains,
New York.

145
 PFEFFER ET AL.

disorder in adults. In acomparative double-blind placebo-controlled study of 54 adults with generalized

anxiety disorder, buspirone was superior to placebo and comparable to diazepam at a mean dose of bus-
pirone 20 mg daily (Goldberg and Finnerty 1982). In another comparative double-blind controlled study of
130 adults with generalized anxiety disorder, buspirone was comparable to clorazepate in reducing symp-
toms of anxiety over a 4-week period at a mean dose of buspirone 16 mg daily (Goldberg and Finnerty
1982); buspirone had a low number of side effects, similar to placebo. In a well controlled study of 162
adults with comorbid mild symptoms of depression (Hamilton Rating Scale for Depression at baseline ranged
from 12 to 17) and generalized anxiety disorder (Hamilton Rating Scale for Anxiety baseline rating below
18), Sramek and colleagues (1996) reported that buspirone (mean 29.5 mg daily, 6 weeks) had significantly
greater efficacy than placebo in reducing symptoms of anxiety and depression; at least one adverse event
(nausea, dizziness, somnolence, or sweating) was reported by 97% of the buspirone-treated adults, signifi-
cantly more than the 93% treated with placebo (Sramek et al. 1996). In a 13-week drug withdrawal study
of adults whose lorazepam was discontinued after at least 6 weeks of treatment, buspirone 15 mg daily was
significantly more effective in preventing benzodiazepine-withdrawal syndrome and controlling symptoms
of generalized anxiety disorder in 13 adult patients compared to increased anxiety in 14 adult patients on
placebo (Chiaie et al. 1995). Other placebo-controlled studies validate the efficacy of buspirone in reduc-
ing symptoms of generalized anxiety disorder in adults (e.g., Robinson et al. 1990).
Buspirone may have efficacy in reducing symptoms of major depressive disorder in adults. In a study of
199 placebo-treated and 183 drug-treated adult patients who initially scored above 18 on the Hamilton Rat-
ing Scale for Depression, buspirone (median 40 mg daily for 8 weeks) had a significantly greater reduction
of symptoms of major depressive disorder (Robinson et al. 1990). These patients also had symptoms of
generalized anxiety disorder and had baseline scores above 18 on the Hamilton Rating Scale for Anxiety.
These reports of buspirone treatment in adults describe its safety and efficacy in relieving anxious-
depressive symptomatology.
Small series and case reports have noted the usefulness of buspirone for treating a range of psychiatric
symptoms in children and adolescents; however, controlled research is minimal on buspirone treatment of
anxiety disorders in child and adolescent psychiatric patients. Symptom relief was noted in a 13-year-old
adolescent with generalized anxiety disorder treated for over 8 weeks with buspirone 5 mg twice daily
(Kranzler 1988). Buspirone augmentation of fluoxetine was reported in an 11-year-old child with obses-
sive-compulsive disorder who received buspirone 10 mg three times daily during a 3-week period (Markovitz
et al. 1990). Buspirone (mean 20 mg daily) reduced "symptoms of social phobia in a 16-year-old male, psy-
chiatrically hospitalized with cormorbid mixed personality disorder, who experienced symptoms of mild
headache, nausea, and light-headedness for the first 3 days of buspirone treatment (Zwier and Rao 1994).
In a 4-week open treatment study of buspirone prescribed at a mean daily dose of 19 mg in 15 children
with anxiety disorders, Simeon and colleagues (1994) reported a significant improvement in clinical global
impression ratings and in scores for anxiety, depression, behavior problems, and psychomotor excitation as
assessed on the Brief Psychiatric Rating Scale for Children. All of these improvements were independent
of the psychiatric diagnosis. Mild transient side effects of buspirone included sleep disturbances and day-
time tiredness (53%), nausea and stomach pains (33%), and headaches (27%). These case reports of re-
duction in anxiety during buspirone treatment highlights the need to validate such findings with controlled
buspirone treatment trials in children and adolescents.
The possibility that buspirone might reduce aggressive symptoms has been described in a few case se-
ries and single case reports. Ratey and colleagues (1980) described 14 developmentally disabled adults with
anxiety, aggression, and self-injurious behavior and who responded to buspirone at doses of 15 to 45 mg
daily during a 4-week period. Among children and adolescents, Gross (1995) reported that 88% of 50 psy-
chiatric outpatients with attention-deficit/hyperactivity disorder (ADHD) who were treated with buspirone
15 to 60 mg daily (as an adjunct to other medications used to treat ADHD symptoms) showed improve-
ment in irritability, aggression, and temper outbursts. Four autistic children, treated with buspirone (up to
15 mg daily for 4 weeks) and then switched to methylphenidate or fenfluramine (for 4 additional weeks),
showed improvement in hyperactivity and aggressivity but only while treated with buspirone (Realmuto et
al. 1989). Quiason (1991) reported successful treatment of an 8-year-old boy with unprovoked aggression
and assaultive behavior using buspirone up to 15 mg three times daily during a 10-day period. Consider-

146
 BUSPIRONE FOR AGGRESSIVE CHILDREN WITH ANXIETY

ing these reports, systematic research is required to determine whether buspirone is effective in modifying
symptoms of aggression in children and adolescents.
Especially in view of the serotonergic properties of buspirone, we conducted a open-label study to ex-
plore the safety, tolerability, and possible effectiveness of buspirone in treating concurrent symptoms and
moderate aggression in prepubertal psychiatric inpatients.

METHODS

Sample
Prepubertal psychiatric inpatients, aged 5 through 12 years, were eligible for the study if they met
DSM-III-R criteria (American Psychiatric Association 1987) for at least one of the disruptive behavior dis-
orders (ADHD disorder, oppositional-defiant disorder, conduct disorder) and were free of acute or chronic
medical illness as determined by history, physical and neurological examination, and laboratory procedures.
Comorbid psychiatric disorders such as major depression, dysthymia, separation anxiety, overanxious disor-
der, post-traumatic stress disorder, and obsessive-compulsive disorder did not result in exclusion. Children
were excluded if they had a psychotic, organic mental, or substance-related disorder. Concomitant treatment
with psychostimulant medication was acceptable, but children receiving other medications were excluded.
Another exclusion criterion was a history of multiple side effects or allergic response during previous trials
with buspirone or other psychotropic medications, except psychostimulants, within 4 weeks of this study. In
accordance with our Institutional Review Board, children entered the study after written informed consent
was obtained from the parent or legal guardian and written assent was obtained from the child.
This study was conducted from April 1994 to April 1996 and involved 3 phases. Baseline evaluations
(Phase I) were conducted each week during the first 2 weeks of hospitalization and involved (a) evaluation
of psychiatric symptoms using standard observational and self-report research instruments for children, (b)
physical and neurological examinations completed by a pediatrician and a child neurologist, (c) laboratory
procedures involving blood test assessments of hematological, thyroid, electrolyte, liver, and renal status,
(d) electrocardiogram (EKG), and (e) electroencephalogram (EEG). DSM-III-R psychiatric disorders were
determined by the inpatient staff within the first week of hospitalization by discussing information from the
psychiatric interviews and behavioral observations of the children.
The following standard research instruments were used:
1. Child Depression Inventory (CDI, Kovacs 1980) is a reliable and valid self-report questionnaire com-
pleted by the children each week to rate the severity of depressive symptoms. A total score of 18 or
more indicates clinically significant symptoms of depression.
2. Revised Children's Manifest Anxiety Scale (RCMAS, Reynolds and Richmond 1985) is a reliable and
valid self-report questionnaire completed by the children each week to rate the severity of anxiety symp-
toms. A total T score of 63 ± 10 or more indicates clinically significant symptoms of anxiety.
3. Measure of Aggression, Violence, and Rage In Children (MAVRIC, Bass et al. 1993) is a reliable 19-
item self-report questionnaire completed by the child each week to rate aggressive symptoms, predom-
inantly directed toward others. A total score of 10 or more indicates clinical significance.
4. Suicidal and Assaultive Behavior Scales (SABS, Pfeffer, 1986, Pfeffer et al. 1986) are administered dur-
ing semi-structured interviews with the children each week to rate severity of suicidal and assaultive acts
and ideation of the child within the prior week. Interviews were conducted by trained research assistants
who achieved excellent reliability (Kappa 0.80, Cohen 1960) before administering this interview to the
children. Ratings for suicidal or assaultive behavior were scored as nonsuicidal or nonassaultive (rated
1), suicidal ideation or assaultive ideation (rated 2), suicidal or assaultive threats (rated 3), mild suici-
dal or assaultive acts (rated 4), and serious suicidal or assaultive acts (rated 5). These scales discrimi-
nate between prepubertal psychiatric inpatients, outpatients, and nonpatients (Pfeffer et al. 1986).
5. Overt Aggression Scale (OAS, Silver and Yudofsky 1991) is a reliable observational measure to rate ag-
gressive behavior of the child. It was completed hourly each day from 7 am to 11 pm by the inpatient
unit nursing staff. It includes a total score, which is the sum of 5 subscores that rate verbal aggression,

147
 PFEFFER ET AL.

physical aggression against objects, physical aggression against self, physical aggression against other
people, and types of intervention used to control aggression. Scores in each subsection are weighted with
respect to degree of seriousness of aggression. Mean total and mean subscale scores each week are in-
dicated as the mean of the daily scores each week.
6. Children's Global Assessment Scale (CGAS, Shaffer et al. 1983) reliably rates overall severity of psy-
chological, social, and school disturbance of prepubertal children; it reflects the lowest level of func-
tioning. Ratings are made on a continuum from 100 (superior functioning) to 1 (needs constant super-
vision). In this study, the research team completed the CGAS at baseline and at the end of buspirone
treatment by considering all available information for each child.
7. The UKU Side Effects Rating Scale (Scandinavian Society of Psychopharmacology, Committee of Clin-
ical Investigations 1986) rates the presence and severity of side effects of medications. In this study, it
was administered to the children in an interview format by a registered nurse just before the adminis-
tration of each dosage of buspirone to the child. The registered nurse was trained to interview in lan-
guage appropriate for prepubertal children.

Medication treatment period

The medication treatment period of this study was divided into 2 phases: a 3-week dose titration period
(Phase II) and a 6-week dose maintenance period (Phase III). Children who rated high on anxiety
(RCMAS >: 50) and aggression (MAVRIC > 10) at either of the two baseline evaluations were eligible to
be treated with buspirone in this study. During the titration phase, buspirone was started at a dose of 5 mg
daily and titrated upward by 5 to 10 mg every 3 days to a maximum dose of 50 mg daily.
Medications were prescribed by the 3 inpatient unit child psychiatrists who were supervised by the
principal investigator (C.R.P.). Doses of buspirone greater than 5 mg daily were administered in divided
doses on a twice daily basis at 8 am and 4 pm. Throughout the buspirone treatment, side effects were
rated and the blood pressure and pulse were measured by a registered nurse who interviewed the child
before administering each morning and afternoon dose. A buspirone dose was not given if the pulse rate
was greater than 110 beats per min or if the systolic blood pressure was less than 80 mm Hg or greater
than 140 mm Hg. Buspirone dose was determined by evaluation of the degree of symptom reduction and
the tolerability of medication side effects. Buspirone dose was reduced if side effects were severe or if
mild to moderate side effects were unremitting over the course of 3 days. Each week, psychiatric symp-
toms were rated with the same self-report and observational research instruments used at the initial as-
sessment.
Duringthe maintenance phase, the optimum daily dose of buspirone was continued. Dose adjustments
were made during this phase to achieve the maximum reduction of psychiatric symptoms with minimum
side effects. Psychiatric symptoms were evaluated weekly throughout the maintenance phase using the same
previously used observational and self-report research instruments. At the end of the 6-week maintenance
period, patients who improved were allowed to continue on buspirone. Those patients who had a negligi-
ble clinical response and/or problematic side effects were tapered off buspirone during a 10-day period. At
the end of the maintenance period, all patients were re-evaluated with the same psychiatric symptom as-
sessment and medical evaluation procedures used at initiation of the study.

RESULTS

Sample
Among the 110 prepubertal psychiatric patients admitted to the inpatient unit from April 1994 to April
1996 (length of stay 3.4 ± 1.6 months), 22 children (20%) had to be excluded from this study because con-
sent for treatment was not granted by the Department of Social Services. This agency of the state govern-
ment had legal custody of those children and made a policy of excluding all custodial children from med-
ical research. In contrast, all parents, who were the legal guardians of their children, consented for their
child to participate and receive treatment in this study.

148
 BUSPIRONE FOR AGGRESSIVE CHILDREN WITH ANXIETY

In addition, 30 children (27% of the total sample) met eligibility criteria but could not enter the titration
phase because their of
degree aggression was so severe that medications such as a neuroleptic or lithium
were indicated and started within 2 weeks of admission to the hospital. An additional 33 children (30% of
the total sample) were excluded because of criteria described under Methods.
Thus, 25 of the 110 psychiatrically hospitalized prepubertal children (23%) met eligibility criteria for this
clinical treatment trial of buspirone and entered the titration phase of the study. The age was 8.0 ± 1.8
(range 5 to 11) years, and 19 children (76%) were boys. The predominance of boys is representative of pre-
pubertal psychiatric inpatient samples. The distribution of racial/ethnic backgrounds was 40% African-Amer-
ican, 28% Hispanic, 20% Caucasian, and 12% other. Socioeconomic class (Hollingshead and Redlich 1958)
reflected backgrounds of the low range in 16 children (64% of the treatment sample), middle level in 5 chil-
dren (20%), and high status in 4 children (16%). This demographic distribution was not different from the
demographic distribution of all children admitted to our psychiatric unit during the 2-year study period.
Table 1 shows the DSM-III-R diagnoses of the 25 children who entered the titration phase of study. The
most prevalent disorders were oppositional-defiant disorder (52%), ADHD, (36%), and specific develop-
mental disorders. Comorbidity predominated in this sample of inpatient children; only 10% of the children
had one psychiatric disorder. Approximately half of the 25 children had 3 comorbid disorders, 35% had 2
comorbid disorders, and 5% had 4 comorbid disorders. This was the first psychiatric hospitalization for all
25 patients. Most children had no previous treatment with psychotropic medications.

Buspirone dose and length of treatment

Among the 25 children who started on buspirone, 4 patients (16%) were discontinued because of per-
sistent behavioral toxicity, specifically agitation and increased aggressivity. Discontinuation of buspirone
led to a reduction of these symptoms, and no withdrawal symptoms were observed. Subsequent neurolep-
tic or lithium treatment was successful in reducing the psychiatric symptoms in these children. All 4 chil-
dren showed their increased aggressive or agitated behavior during the second week of buspirone treatment.
Of the 25 children who entered the titration phase, 21 patients continued into the buspirone maintenance
phase. Two children (8%) were discontinued during the first week of the maintenance phase (at buspirone
15 mg daily) because they developed distinct severe euphoric symptoms, increased impulsivity, and out-
of-control behavior. Both children subsequently were treated successfully with either lithium or another
mood stabilizer (valproate or carbamazepine). Of the 21 children who entered the maintenance phase, 19
children (76%) completed the treatment trial.
The duration of buspirone treatment (titration and maintenance phases) for the 19 children who com-
pleted the 9-week study was 8.2 ± 4.6 weeks (range 5-9 weeks, median 9 weeks). Two children were treated
for less than 9 weeks (i.e., less than 3 weeks in titration phase or 6 weeks in maintenance phase) because
they were discharged from the hospital before completing the maintenance phase. The daily dose of bus-

Table 1. Psychiatric Disorders of 25 Children Who Started Buspirone Trial

DSM-III-R disorder Number of patients Percentage of patients

Mood disorder 9 36
Major depressive disorder 5 20
Dysthymic disorder 4 16
Disruptive behavior disorder 21 84
Attention-deficit hyperactivity disorder 9 36
Oppositional-defiant disorder 13 52
Conduct disorder 4 16
Anxiety disorder 8 32
Separation anxiety disorder 1 4
Overanxious disorder 2 8
Post-traumatic stress disorder 5 20
Specific developmental disorder 9 36

149
 PFEFFER ET AL.

pirone during the last week of each child's treatment was 28 ± 11 mg (range 10 to 50 mg daily, median
30 mg). Two children were being treated with methylphenidate 20 mg daily at the time of entry to the study,
and they continued to receive this medicine throughout the study.

Buspirone side effects

None of the 19 children who completed treatment with buspirone reported severe side effects. Two chil-
dren complained of mild transient headaches in the first week of buspirone treatment. Pulse, blood pres-
sure, respiration, temperature, EKG, and EEG remained stable throughout treatment with buspirone.

Change in psychiatric symptoms during the course of study

Table 2 shows the mean baseline scores, calculated as the average of the 2 weeks of initial ratings, for
specific psychiatric symptoms and the endpoint scores for the same symptoms in the last week of buspirone
treatment for the 19 children who completed this study. Table 2 indicates the significance of change, cal-
culated with paired t tests, between the baseline and endpoint scores for each psychiatric symptom.
There was a significant reduction (p < 0.001) in the CDI score from 19 ± 8.2 at baseline to 9.2 ± 7.5
at the end of the study. This reduction, from a baseline level of severe depression to a nondepressed (or
nonclinical) level, can be considered a clinically meaningful improvement in the symptoms of depression.
This change represented a 52% decrease in the baseline CDI score, a level of improvement that was at-
tained in the third week of the maintenance phase (See Fig. 1). It is notable that 7 of the 10 completers
(70%) who had a baseline CDI score of at least 18 (the cut-off for clinically significant depression) had an
endpoint CDI score that was less than 12 (the cut-off for a nonclinically significant level of depression).
There was a significant lowering in the number of seclusions and/or physical restraints daily (89%, p <
0.01) and the number of hours that children spent in seclusion and/or restraints daily (86%, p < 0.02). Com-
pared to baseline, the number of hours daily spent in seclusion and/or physical restraints was 50% lower
by the end of the third week of the titration phase (See Fig. 2).
In general, though, children remained quite aggressive and still required intervention with seclusion or
physical restraints to control aggressive behavior at the end of the study. There was a significant reduction
(29%, p < 0.02) in MAVRIC scores, but children expressed clinically significant levels of aggression at
the end of study (MAVRIC baseline 18 ± 7, endpoint 13 ± 9).
There was a significant reduction (16%, p ^ 0.04) in social anxiety (rated as a factor on the RCMAS),
despite the lack of significant change in general anxiety (rated as the total RCMAS score). Children were
more eager to become involved in interpersonal interactions and appeared more assured in initiating activ-
ities with others.
Clinical global functioning (assessed by CGAS) was significantly improved by the end of the study (base-

Table 2. Mean Baseline and Endpoint Scores for Psychiatric

Symptoms for 19 Cihldren Who Completed the Study

Symptom variable Baseline Endpoint t df p

Depression (CDI) 19.03 ± 8.67 9.15 ± 7.50 6.28 18 0.001

Suicide Spectrum (SABS) 2.58 ± 1.24 1.58 ± 1.17 2.10 11 0.06
Social Anxiety Scale (RCMAS) 11.18 ± 2.79 9.42 ± 2.93 2.26 18 0.04
Aggression
Seclusion and/or restraints—Hours 1.97 ± 3. 0.28 ± 0.44 2.58 18 0.02
Seclusion and/or restraints—Number 1.58 ± 2.21 0.45 ± 0.85 2.73 18 0.01
Aggression, violence, rage in children 18.27 ± 6.71 13.00 ± 8.97 2.74 10 0.02
(MAVRIC)
Global Assessment (CGAS) 40.68 ± 10.49 54.47 ± 14.18 -3.80 18 0.01

Scores are means ± SDs. Note: n 19 for all variables except MAVRIC (n = 11) and SABS (n 12). =

MAVRIC: Measure of Aggression, Violence, and Rage in Children. CGAS: Children's Global Assessment Scale.
-

150
 BUSPIRONE FOR AGGRESSIVE CHILDREN WITH ANXIETY

line 41 ± 10, endpoint 54 ± 14, p < 0.01), reflecting a 34% enhancement of global functioning. However,
the endpoint level of global functioning was still quite impaired.
The level of suicidal tendencies, rated on the SABS, was lower at the end of this treatment trial (base-
line 3 ± 1, endpoint 2 ± 1); this change almost reached a significant level (p < 0.06). There were no sig-
nificant changes in scores on the Overt Aggression Scale, but this may have been related to the lack of con-
sistent hourly ratings by the inpatient staff.
Subsequent to the end of the buspirone treatment trial, 3 children (16%) who completed the medication
trial remained on buspirone throughout the remainder (3.6 ±1.1 weeks) of their hospitalizations. Other chil-
dren who completed the study were discontinued from buspirone and then started trials with other med-
ications, primarily mood stabilizers and neuroleptics.

DISCUSSION

This 9-week open-label study of buspirone, a nonbenzodiazapine anxiolytic medication, aimed to iden-
tify treatment effects in 25 prepubertal psychiatric inpatients with symptoms of both anxiety and moderate
aggression. The sample had a high rate of comorbid psychiatric disorders (90%), predominantly consisting
of mood, attention-deficit hyperactivity, and other disruptive behavior disorders. The results suggested that
buspirone might be helpful in reducing anxiety and especially aggression in these young patients.
The results should be interpreted considering the limitations of any uncontrolled open treatment trial. The
findings regarding the type and degree of improvement, adverse effects, and buspirone regimen are limited
by the lack of a comparison group of children who received placebo and the lack of blinding and random-
ization. Specifically, the improvements reported in this study may be related to the effects of buspirone,
placebo response, other modalities of psychiatric hospital treatment, observer bias, sample heterogeneity re-
sulting from the use of clinical rather than standardized procedures for assessing diagnosis, or the natural
course of the children's psychopathologies. The results also may have been affected by the limited range
of buspirone doses that were used and the duration of the treatment trial.
Behavioral toxicity, specifically involving agitation and increased aggressivity, was evident in 4 of the

30 -,

25 A

20

CDI SCORES 15
.

10

0
WEEK OF STUDY: 0123456789 10 il
Entry
I_1 I_I I_ _J
PHASE: BASELINE TITRATION MAINTENANCE

FIG. 1. Change in weekly CDI scores (Mean ± SD) of 19 children who completed the study.

151
 PFEFFER ET AL.

25 children (16%) who entered the titration phase. It was necessary to discontinue buspirone in these chil-
dren within the first 3 weeks of treatment, before they entered the maintenance phase. In addition, 2 of the
25 children (8%) developed symptoms suggestive of childhood mania, and these children were discontin-
ued from buspirone treatment in the first week of the maintenance phase (fourth week of drug treatment).
These 2 children met criteria for DSM-III-R bipolar disorders and were successfully treated with lithium
or an anticonvulsant medication. The fact that one-quarter of these anxious and aggressive children could
not be continued on buspirone, because of adverse behavioral effects, suggests that its usefulness in this
population might be limited and that an empirical trial of buspirone may pose risks.
The mean dose of buspirone for the 19 children who completed this study was 28 mg daily, adminis-
tered in two divided doses at 8 am and 4 pm. These children tolerated buspirone well with minimal com-
plaints of side effects, such as headaches (10%). There were no buspirone-induced changes in the EKGs,
EEGs, or laboratory blood test findings.
Our results regarding adverse effects are generally similar to one double-blind placebo-controlled study
in adults with generalized anxiety disorder that found a low rate of side effects (Chiaie et al. 1995, Gold-
berg and Finnerty 1982, Robinson et al. 1990) but another double-blind study of adults suggested more side
effects, such as nausea (34% vs. 13%), dizziness (64% vs. 12%), and somnolence (19% vs. 7%), with bus-
pirone than placebo (Sramek et al. 1996). The occurrence of behavioral side effects, such as manic symp-
toms, and the recurrence of impulsivity and out-of-control behaviors observed in our study, suggests that
children with vulnerability to bipolar disorder should be identified before treatment with buspirone is in-
stituted.
Significant changes in specific symptoms were identified in the 19 children who completed this treat-
ment trial. The reduction in aggressive behavior was especially notable: 89% decrease in number of hours
that children spent in seclusion and/or physical restraints, an 86% reduction in the number of acute seclu-
sions and/or physical restraints that were used to interrupt uncontrolled aggressive behavior of the children,
and a 29% decrease in self-reported aggressivity on the MAVRIC. The OAS did not indicate any signifi-

HOURS OF SECLUSION
AND/OR RESTRAINT 4

WEEK OF STUDY: 0 12 10 11
Entry
PHASE: BASELINE TITRATION MAINTENANCE

FIG. 2. Change in weekly number of hours of seclusion and/or physical restraints for 19 children who completed the
study (Mean ± SD).

152
 BUSPIRONE FOR AGGRESSIVE CHILDREN WITH ANXIETY

cant change in aggressiveness, perhaps because the nursing staff members who made the ratings were in-
consistent in their observations. Although these findings together suggest a significant improvement in ag-
gressivity symptoms, the children retained serious aggressive symptoms at the end of the 9-week buspirone
treatment trial. These children required continuing use of intensive interventions, including the occasional
use of seclusions and physical restraints.
About 75% of children who started the buspirone trial were able to complete it; among the completers,
self-reported symptoms of depression (as rated on the CDI) significantly decreased by 50% in those chil-
dren, and 70% of them had a reduction of depressive symptoms from a clinically severe to a clinically non-
significant level. This improvement in depression appeared to be particularly meaningful changes because
most of the children no longer had pathological levels of depressive symptoms by the end of the 9-week
trial. The changes in self-rated depression is consistent with the trend toward a reduction of severity of sui-
cidal tendencies. These findings concur with reports that buspirone is effective in reducing symptoms of
major depressive disorder in adults (Robinson et al. 1990).
Self-reported anxiety during social interactions decreased significantly with buspirone treatment, although
minimal change was noted for total RCMAS scores for symptoms of anxiety. These results add to the var-
ied results previously reported for buspirone in treating social anxiety. For example, beneficial effects of
buspirone in social phobia have been described in uncontrolled reports on an adolescent (Zwier, Rao 1994)
and on 21 adults (Schneier et al. 1993), whereas a 12-week double-blind placebo-controlled study of 30
adults reported no therapeutic effect of buspirone for symptoms of social phobia (van Vliet et al. 1997).
CGAS scores showed clear improvements, but global functioning remained quite impaired at the end of
this clinical trial (CGAS 55). For this reason, over 80% of the 19 children who completed the 9-week trial
were discontinued from buspirone at the end of the study, and other medications were tried to reduce the
still impairing symptoms.

Clinical and research implications

Despite significant improvements in symptoms of anxiety, depression, and aggression among the 19 chil-
dren who completed this study, 27% of the prepubertal psychiatric inpatients (30 of 110) who met eligi-
bility criteria for this study were not treated with buspirone because their severe aggressivity required rapid
intervention with lithium or neuroleptics, medications that have been shown in double-blind controlled treat-
ment studies of prepubertal children to be effective in decreasing severe aggression (Campbell et al. 1984,
Campbell et al. 1995). Thus, our sample of buspirone-treated patients did not include the 25% of children
with the most severe aggressive symptoms on psychiatric hospital units. Buspirone may still be useful to
treat children with mild to moderate levels of aggressivity and who suffer from concurrent symptoms of
anxiety (and, speculatively, mild to moderate aggressivity associated with major depression). Double-blind
controlled treatment studies are essential to assess the possible clinical value of buspirone in treating pre-
pubertal children with anxiety and mild to moderate levels of aggression.
The 2 children who were treated with methylphenidate for ADHD during this buspirone trial tolerated
this combined medication treatment with no apparent adverse effects. Research is needed to determine if
combining buspirone with psychostimulant treatment is effective in reducing symptoms of moderate ag-
gressivity in children who have ADHD.
Finally, an unwanted medical barrier posed by a current public policy became evident during the con-
duct of this study. Approximately 20% of our child psychiatric inpatients were precluded from entry into
this study because of a policy of the government in New York State: The Department of Social Service
does not give approval for clinical research involving any child for whom it has legal guardianship. Be-
cause of this policy, a sizable proportion of children with severe psychopathology are unable to receive the
potential advantages of participating in clinical research, even if it may identify beneficial approaches to
enhance the quality of life of the child. Although this policy may be effective in keeping the Department
of Social Service from making responsible complex decisions on behalf of children and in protecting the
Department from certain types of criticism, it is doubtful that this policy uniformly serves the interests of
children in their care with severe pathology. These issues require review from the perspective of research
and service delivery policy for children.

153
 PFEFFER ET AL.

ACKNOWLEDGMENT

This study was funded by an unsolicited grant from Bristol-Myers Squibb Company (1994-1996) and a
fund established in The New York Community Trust by DeWitt-Wallace.

REFERENCES

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised
(DSM-III-R). Washington, D.C., American Psychiatric Association, 1987
Bass JW, Geenens DL, Popper C: Measure of Aggression, Violence, and Rage in Children (MAVRIC) [abstract]. Pro-
ceedings of the Annual Meeting of the American Academy of Child and Adolescent Psychiatry, San Francisco (CA),
October, 1993
Campbell M, Adams PB, Small AM, Kafantaris V, Silva RR, Shell J, Perry R, Overall JD: Lithium in hospitalized ag-
gressive children with conduct disorder: A double-blind and placebo-controlled study. J Am Acad Child Adolesc Psy-
chiatry 34:445^153, 1995
Campbell M, Small A, Green WH, Jennings SJ, Perry R, Bennett WG, Anderson L: Behavioral efficacy of haloperi-
dol and lithium carbonate: A comparison in hospitalized aggressive children with conduct disorder. Arch Gen Psychi-
atry 41:650-656, 1984
Chiaie RD, Pancheri P, Casacchia M, Stratta P, Kotzalidis GD, Zibellini M: Assessment of the efficacy of buspirone
in patients affected by generalized anxiety disorder, shifting to buspirone from prior treatment with lorazepam: A
placebo-controlled, double-blind study. J Clin Psychopharmacol 15:12-19, 1995
Cohen J: A coefficient of agreement for nominal scales. Educational Psychological Measures 20:37^16, 1960
Goldberg HL, Finnerty R: Comparison of buspirone in two separate studies. J Clin Psychiatry 43:87-91, 1982
Gross MD: Buspirone in ADHD with ODD [letter]. J Am Acad Child Adolesc Psychiatry 34:1260, 1995
Hollingshead HM, Redlich P: Social Class and Mental Illness. New York, Wiley, 1958
Kovacs M: Rating scales to assess depression in school-age children. Acta Paedopsychiatrica 46:305-315, 1980
Kranzler HR: Use of buspirone in an adolescent with overanxious disorder. J Am Acad Child Adolesc Psychiatry
27:789-790, 1988
Markovitz PJ, Stagno SJ, Calabrese JR: Buspirone augmentation of fluoxetine in obsessive compulsive disorder. Am
J Psychiatry 147:798-800, 1990
Pfeffer CR: The Suicidal Child. New York, Guilford Press, 1986
Pfeffer CR, Plutchik R, Mizruchi MS, Lipkins R: Suicidal behavior in child psychiatric inpatients, and outpatients and
in nonpatients. Am J Psychiatry 43:733-738, 1986
Quiason N, Ward D: Buspirone for aggression [letter]. J Am Acad Child Adolesc Psychiatry 30:1026, 1991
Ratey JJ, Sovner R, Mikkelsen E, Chielinsky HE: Buspirone therapy for maladaptive behavior and anxiety in devel-
opmental^ disabled persons. J Clin Psychiatry 50:382-384, 1980
Realmuto GM, August GJ, Garfinkel BD: Clinical effect of buspirone in autistic children. J Clin Psychopharmacol
9:122-125, 1989
Reynolds CR, Richmond BO: Revised Children's Manifest Anxiety Scale (RCMAS) Manual. Los Angeles (CA), West-
ern Psychological Services, 1985

Robinson DS, Rickels K, Feighner J, Fabre L, Gammans R, Shrotriya R, Alms D, Andary J, Messina M: Clinical ef-
fects of the 5-HT(la) partial agonists in depression: A composite analysis of buspirone in the treatment of depression.
J Clin Psychopharmacol 10 (Suppl) 67S-76S, 1990
Scandinavian Society of Psychopharmacology, Committee of Clinical Investigations (UKU): The UKU Side Effects
Rating Scale: Scale for the Registration of Unwanted Effects of Psychotropics Manual. Oslo, Norway, 1986
Schneier FR, Saoud JB, Campeas R, Fallón BA, Hollander E, Copian J, Lieboowitz MR: Buspirone in social phobia.
J Clin Psychopharmacol 13:251-256, 1993

154
 BUSPIRONE FOR AGGRESSIVE CHILDREN WITH ANXIETY

Shaffer D, Gould MS, Brasic J, Ambrosini P, Fisher P, Bird H, Aluwahlia S: A Children's Global Assessment Scale
(CGAS). Arch Gen Psychiatry 40:1228-1231, 1983
Silver IM, Yudofsky SC: The Overt Aggression Scale: Overview and guiding principles. I Neuropsychiatry 3
(Suppl):22-29, 1991
Simeon JG, Knott VI, Dubois C, Wiggins D, Geraets I, Thatte S, Miller W: Buspirone therapy of mixed anxiety dis-
orders in childhood and adolescence: A pilot study. I Child Adolesc Psychopharmacol 4:159-170, 1994
Sramek JJ, Tansman M, Suri A, Hornig-Rohan M, Amsterdam ID, Stahl SM, Weisler RH, Cutler NR: Efficacy of bus-
pirone in generalized anxiety disorder with coexisting mild depressive symptoms. I Clin Psychiatry 57:287-291, 1996
van Vliet IM, den Boer JA, Westenberg HGM, Pian KLH: Clinical effects of buspirone in social phobia: A double-
blind placebo-controlled study. J Clin Psychiatry 58:164-168, 1997
Wheatley D: Buspirone: Multicenter efficacy study. J Clin Psychiatry 43:92-94, 1982
Zwier, KJ, Rao U: Buspirone use in an adolescent with social phobia and mixed personality disorder (Cluster A type).
J Am Acad Child Adolesc Psychiatry 33:1007-1011, 1994

Address reprint requests to:

Cynthia R. Pfeffer, M.D.
21 Bloomingdale Road
White Plains, NY 10605

155