Zeevi Dvir

Zeevi Dvir
Clinical Biomechanics
NATIONAL INSTITUTES OF HEALTH

NIH LIBRARY
r"“
DEC 2 3 2000
J
BLDG 10, 10 CENTER DR.
BETHESDA, MD 20892-1150
Zeevi Dvir, PhD, LLB
Department of Physical Therapy
Sackler Faculty of Medicine
Tel Aviv University
Ramat Aviv
Israel
Churchill Livingstone
A Harcourt Health Sciences Company

New York Edinburgh London Philadelphia
CHURCHILL LIVINGSTONE
^ . A Harcourt Health Sciences Company
The Curtis Center

Independence Square West
) ^■•^T^Philadelphia, Pennsylvania 19106
JYV
Library of Congress Cataloging-in-Publication Data
Clinical biomechanics / [edited by] Zeevi Dvir.
p. cm.
ISBN 0-443-07945-5
1. Orthopedics. 2. Human mechanics. 3. Musculoskeletal

system—Mechanical properties. I. Dvir, Zeevi. [DNLM:
1. Biomechanics. WE 103 C6403 2000]
RD732. C575 2000 616.7—dc21
DNLM/DLC ■37691
Acquisitions Editor: Andrew Allen

Manuscript Editor: Amy Norwitz
Production Manager: Natalie Ware
Illustration Specialist: Francis Moriarty
CLINICAL BIOMECHANICS
ISBN 0-443-07945-5
Copyright © 2000 by Churchill Livingstone.
Churchill Livingstone® is a registered trademark of Harcourt, Inc.

— is a trademark of Harcourt, Inc.
All rights reserved. No part of this publication may be reproduced or transmitted in any form or
by any means, electronic or mechanical, including photocopy, recording,. or any information storage
and retrieval system, without permission in writing from the publisher.
Printed in the United States of America.
Last digit is the print number: 9 8 7 6 5 4 3 2 I

In loving memory of my father.
Contributors
Dan L. Bader, PhD Walter Herzog, PhD

Professor of Medical Engineering, Department of Professor, Human Performance Laboratory,
Engineering, Queen Mary and Westfield University of Calgary, Calgary, Alberta,
College, University of London, London, UK Canada
Biomechanics of Soft Tissues Biomechanics of Manual Therapy
JOHN BEHIR1, PHD Judith Laprade, BSc(PT), MSc

Senior Lecturer, Department of Materials, Queen Department of Anatomy and Cell Biology,
Mary and Westfield College, University of Faculty of Health Science, Queen’s University,
London, London, UK Kingston, Ontario, Canada
Biomechanics of Bone Biomechanics of the Shoulder Complex
Richard W. Bohannon, EdD, PT, NCS Michael Lee, M Biomed Eng

Professor, Department of Physical Therapy, Lecturer, School of Exercise and Sport Science,
School of Allied Health, University of and Department of Aeronautical Engineering,
Connecticut, Storrs, Connecticut; Senior University of Sydney, Sydney, New South
Scientist, Institute of Outcomes Research, Wales, Australia
Biomechanics of Manual Therapy
Hartford Hospital, Hartford, Connecticut
Biomechanics of Neurologic Treatment
Stuart M. McGill, PhD

Professor of Spinal Biomechanics, Department of
Carlijn Bouten, PhD Kinesiology, Faculty of Applied Health
University Lecturer, Department of
Sciences, University of Waterloo, Waterloo,
Computational and Experimental Mechanics,
Ontario, Canada
Eindhoven University of Technology,
Biomechanics of the Thoracolumbar Spine
Eindhoven, The Netherlands
Biomechanics of Soft Tissues
Joseph Mizrahi, DSc

Professor, Department of Biomedical
Elsie Culham, PT, PhD Engineering, Julius Silver Institute of
Associate Professor, School of Rehabilitation Biomedical Engineering Sciences,
Therapy, Faculty of Health Science, Queen’s Technion-Israel Institute of Technology, Haifa,
University, Kingston, Ontario, Canada Israel
Biomechanics of the Shoulder Complex Biomechanics of Unperturbed Standing Balance
Zeevi Dvir, PhD, LLB Joseph J. Sarver, BSME, MS Biomed Eng

Professor and Chairman, Department of Physical Teaching Assistant, School of Biomedical
Therapy, Sackler Faculty of Medicine, Tel Engineering, Science and Health Systems,
Aviv University, Ramat Aviv, Israel Drexel University, Philadelphia, Pennsylvania
Biomechanics of Muscle Measurements in Biomechanics
Julianna Gal, PhD Rami Seliktar, PhD

Senior Lecturer, School of Applied Sciences, Professor of Biomedical Engineering and
University of Glamorgan, Pontypridd, Wales, Mechanical Engineering, Drexel University,
UK Philadelphia, Pennsylvania
Biomechanics of Manual Therapy Measurements in Biomechanics
VII
viii Contributors
JOHN Stallard, BTech, CEng, FIMECH E, Deepak Vashishth, be (Hons), MSME, PhD
FIPEM Assistant Professor, Department of Biomedical
Technical Director, Orthotic Research and Engineering, Rensselaer Polytechnic Institute,
Locomotor Assessment Unit, Robert Jones & Troy, New York
Agnes Hunt Hospital, Oswestry, Shropshire, Biomechanics of Bone
UK
Lower Limb Orthotics
Preface
Back in 1975, as a first-year PhD student at the Bioengineering Unit, Strathclyde

University in Glasgow, I was introduced to Biomechanics by one of the world leaders
in this branch of science and engineering, Professor Robert M. Kenedi. It took “Bobby,”
as he was known in the professional and social circles, no less than five lecture sessions
to encompass the vast domain of biomechanics. With the elegance of a great artist and
the mastery of a distinguished conductor, he moved from components to systems, from
tissue to organ and whole-body level, and from the most basic science concepts to the
most applied. In so doing, he managed to instill in us the excitement as well as the
humility with which one should view and approach this fascinating interdisciplinary
subject. In years to come, we became very close, and in the course of many visits I paid
to the Kenedis’ home in beautiful Milngavie near Glasgow, his scientific curiosity,
particularly with respect to several biomechanical aspects of the aging body, was as
robust as ever. During my last meeting with him in February '98, I was so extremely
pleased to learn that he would gladly write the preface to this book, but this plan was
sadly frustrated by his untimely death a few months later. This book is a tribute to an
absolutely outstanding teacher, scientist, and friend.
In formulating the guidelines to this book, I was advised that it should (1) be clinically
oriented, particularly toward motor dysfunctions and rehabilitation; (2) aim at the
postgraduate level; (3) serve more as a reference than as a textbook; and (4) be written
by an international team of researchers who were involved with various aspects of
musculoskeletal biomechanics. This joint authorship has inevitably led to a measure of
heterogeneity, but I believe that it is a worthwhile price.
The opening chapter of the book introduces the reader to basic concepts and measure¬
ment methods; Chapters 2 through 4 relate to tissue biomechanics; Chapters 5 and 6
deal with two major joint-muscle systems, the spine and the shoulder; and Chapters 7
through 10 present the use of biomechanics in various clinical applications: treatment of
patients with neurologic disorders, balance measurement and analysis, spinal manipula¬
tion, and orthotics. It was felt that these chapters presented an intellectual as well as
practical challenge. On the other hand, it was decided not to include an independent
chapter on gait analysis because this subject is well covered by other texts.
It is a great pleasure to offer my sincere appreciation and thank all contributors for a
very fine piece of work. I am also indebted to Professor Otto Payton from the Medical
College of Virginia, who paved the way for this book, as well as to the publishing team
at WB Saunders for their persistent effort and kind help in bringing this project to a
successful completion.
Zeevi Dvir
NOTICE
Physical therapy is an ever-changing field. Standard safety precautions must be
followed, but as new research and clinical experience broaden our knowledge,
changes in treatment and drug therapy become necessary or appropriate. Readers
are advised to check the product information currently provided by the manufac¬
turer of each drug to be administered to verify the recommended dose, the method
and duration of administration, and the contraindications. It is the responsibility of
the treating physician, relying on experience and knowledge of the patient, to
determine dosages and the best treatment for the patient. Neither the publisher nor
the editor assumes any responsibility for any injury and/or damage to persons
or property.
The Publisher
Contents
I Measurements in Biomechanics I
Rami Seliktar and Joseph J. Sarver
2 Biomechanics of Soft Tissues 35

Dan L. Bader and Carlijn Bouten
3 Biomechanics of Bone 65
John Behiri and Deepak Vashishth
4 Biomechanics of Muscle 83
Zeevi Dvir
5 Biomechanics of the Thoracolumbar Spine 103

Stuart M. McGill
6 Biomechanics of the Shoulder Complex 141

Elsie Culham and Judith Laprade
7 Biomechanics of Neurologic Treatment 165

Richard W. Bohannon
8 Biomechanics of Unperturbed Standing Balance 189

Joseph Mizrahi
9 Biomechanics of Manual Therapy 209

Michael Lee, Julianna Gal, and Walter Herzog
10 Lower Limb Orthotics 239

John Stallard
Index 267
?
CHAPTER
Measurements in Biomechanics
Rami Seliktar and Joseph J, Sarver
Overview some point they appear gray. Nothing physically

changes in the object; rather, it is our perception
that is biased by the lighting conditions at dusk.
Measuring weight, length, speed, temperature,
This example helps to clarify the measurement
voltage, and other variables has become part of
process, which typically involves at least two
our modem lifestyle. We often question the accu¬
components, namely a sensor (transducer) and a
racy of our measurement devices by asking such
conditioner. Our eyes (sensors or transducers)
questions as, Is the $5 garden thermometer really
transform the light energy into neural signals,
showing the correct temperature? Isn’t the bath¬
whereas our brain conditions these neural signals
room scale reading a little high? Are we really
so that we perceive a color.
going 65 miles per hour? Our doubts are perfectly
So far, we have used several terms, such as
legitimate because these devices are not precision
accuracy, precision, error, bias, and sensor, that
instruments, and their readings involve a consider¬
have distinct definitions in the measurement jar¬
able margin of error. It is possible to increase the
gon. Other terms commonly used to describe char¬
precision of the measuring instruments at some
acteristics of measurement systems and measur-
extra cost, but the measurement will never be
ands in general include noise, sensitivity,
absolute, and this is the first thing that we have
discrimination, repeatability, reproducibility, hys¬
to understand when we study the principles of
teresis, resolution, linearity, drift, distortion, cali¬
measurement. Although it is important to perform
bration, cross-talk, analog, discrete, and sampling.
a correct measurement, the price of greater accu¬
At least some of these terms sound familiar be¬
racy may be high, whereas the benefit may be
insignificant; therefore, the objective of the mea¬ cause they are borrowed from expressions used
surement should determine the desired margin of daily. For example, the size of the setting sun
error. For example, for determining weather condi¬ appears larger than the midday sun, meaning there
tions, a thermometer that can read temperature is obviously a perception (measurement) error.
with an accuracy of one thousandth of a degree Lifting a heavy object biases our sense of force;
Celsius is of no advantage to us over a thermome¬ consequently, a lighter object may feel much
ter that reads only up to one degree. lighter. When we bake a cake, we weigh our
To this end, we need to have an appreciation of ingredients to achieve good reproducibility. For¬
measurements and can begin by looking at our mal definition of these terms is important because
own sensory system and its measuring capabili¬ they describe the properties of the measuring sys¬
ties. The belief that “what you see is what the tem and the measurands.
object actually is” is quite common and quite A key to good measurement is understanding
incorrect. People tend to believe that what they the process of measurement and the tools used to
see, feel, or sense is absolute, that is, represents perform the measurement. More than anything
the actual size, shape, color, and weight of an else, however, it is important to know what to
object. Several aspects of measuring, however, expect. Without this basic understanding, mea¬
affect how a person perceives a given object or a surement errors may never be noticed and there¬
stimulus; this results in different perceptions fore may be assumed correct. For instance, most
among different people. For example, as the sun people can tell that if a speedometer in a car reads
sets, the colors of objects appear to fade, until at a speed of 250 miles per hour, it must have false
I
2 Clinical Biomechanics
calibration. Without prior information about the

range of speeds to expect when driving a car, this
measurement error would not have been noticed
by the operator. On the other hand, if a person
jumps from a chair and an acceleration gage
shows that the acceleration of his head at landing
is 50 m/s2, few could tell whether that measure¬
ment is right or wrong.
The point of this chapter is to discuss the in¬
strumentation commonly used when measuring
biomechanical variables. Within this overview
section, we provide a brief introduction to bio¬
mechanics and the variables of interest as well
as a review of the different measurement terms
mentioned earlier. A more detailed discussion of
some biomechanical variables follows, and a more
in-depth description of the instrumentation most
commonly used to measure these variables is then
• If a ball hits the player’s head, the impulse is
presented, including a review of some clinical and
transmitted through the skull to the brain,
research applications of such equipment.
creating conditions that may result in brain
injury. This may be caused by shear of the
WHAT IS BIOMECHANICS? brain tissue, compression on the impact side, or
cavitation on the opposite side. Flow do we
With the latest surge of interest in human motion determine what are the force and impulse
and gait analysis, particularly on the part of nonen¬ thresholds that will cause injury?
gineering professionals, misconceptions about bio¬
Although there is no formal convention for clas¬
mechanics have developed. For example, one can
sifying different biomechanical activities, the fol¬
frequently find references to gait analysis as “bio¬
lowing broad categories encompass most of the
mechanics” in discussion forums as well as pub¬
research disciplines:
lished works. Although gait analysis is a part of
biomechanics, it is by no means representative of Mechanics of human performance, in which one
the spectrum of activities in biomechanics. A brief of the objectives is to study the performance
clarification of the definition, topics, and disci¬ itself or, alternatively, to study the
plines within biomechanics is therefore needed. performance to determine the internal
Because our existence and interaction with our (musculoskeletal) forces
environment is primarily mechanical, most of us Mechanics of biologic tissues (including biofluid
develop an intuition with regard to “what happens mechanics), in which the objective is to study
if?” The following examples depict some aspects the material behavior of the system’s elements
of the science of biomechanics: (e.g., bone, cartilage, tendons, body fluids).
• While swinging a tennis racquet, as in Figure BIOMECHANICAL VARIABLES OF INTEREST
1-1, the tennis player’s muscles need to exert
forces that move his joints so that they reach The basic variables that need to be measured in
the required positions. any mechanical analysis and that therefore are
• While swinging his racquet, synergistic of interest in a biomechanical analysis are force,
muscles in the player’s torso, shoulder, arm, position (and displacement), mass, and time. Other
and wrist act in concert to execute the motion, variables of interest, such as moments, moments
not only in terms of positioning but also in of inertia, velocity, acceleration, and stress, can
temporal precision and with the appropriate be derived from these basic variables. More com¬
velocity and acceleration for impact with the pounded variables, such as work-energy, power,
ball (kinematics). impulse, and internal muscle and joint forces, can
• Consequently, overexertion of the shoulder, subsequently be computed once these fundamental
arm, wrist, or lumbar spine can create variables have been measured.
excessive stress in the muscle, tendons,
ligaments, joint cartilage, or even bone and can MEASUREMENT TERMS
either cause acute failure or initiate a
As mentioned earlier, measurement is a process
cumulative trauma disorder.
involving at least two components: a sensor/trans-
I Chapter I MEASUREMENTS IN BIOMECHANICS 3
error; however, instrument A is obviously less

accurate than instrument B.
It is interesting to apply some basic principles
of statistics and probability to the example in
Figure 1-2 that can help us to understand the
nature of measurements. If we were to compare
the mean measurements from instruments A and
B, we would see that indeed instrument A’s mean
measurement is significantly larger than that of B
as well as the true value. The bias of instrument
A could then be defined as the difference between
its mean measurement and the true measurement.
Furthermore, the standard deviations of either in¬
FIGURE 1-2 Depiction of accuracy and precision.
strument give an indication of the precision error,
that is, the larger the standard deviation, the larger
the precision error. In fact, the equation that de¬
ducer and a conditioner. Either one of these com¬ fines precision error is as follows:
ponents can be responsible for distortion of the
measurement, which results in a production of er¬ %ER = ^ X 100%
rors. R
An error, defined as the difference between the
measured value and true value, can occur in either where cr is the standard deviation and R is the
the sensor/transducer or the conditioner. The cor¬ range (maximum minus minimum) of the measur¬
responding accuracy (A) of a measurement is de¬ ing device.
fined by the ratio of the error to the true value, The sensitivity, defined as the ratio between the
as follows: change in output (Y) for the given change in input
(X), or S = AY/AX (the slope of the line in Fig.
1-3), can sometimes be too low or too high. A
_ (Measured value — true value)
sensitivity that is too low results in minimal output
True value
for a significant change in input, whereas a sensi¬
tivity that is too high results in a large change in
Notice that both accuracy and error require knowl¬ output for an insignificant change in input. Notice
edge of a measurand’s true value. Unfortunately, in Figure 1-3 that the data (hollow circles) follow
it is impossible to know the true value of any a nearly straight line. In such a case, the instru¬
given variable for several reasons, one of which ment is said to have a relatively low linearity
is noise. Noise is defined as a random variation of error, defined by the following equation:
a given value that is of no interest to the measure¬
ment itself. For example, when one is listening
to music (i.e., sound at specific frequencies and %(eL)max = x 100%
amplitudes), a jackhammer being used outside

would add noise, altering the perceived (mea¬ where (eL)max is the maximum expected error be¬
sured) sound. Noise is unavoidable and is present tween the measurement and a linear approxima-
in every measurement, making the true value of a
given measurand intangible.
To minimize the effects of errors resulting from
various sources such as noise, it is advisable to
first calibrate the measuring instrumentation by
performing repeated measurements of the variable
while in a controlled system. Upon doing so, we
encounter a distinction between accuracy and pre¬
cision (repeatability).
In Figure 1-2, for example, we see 10 random
measurements from instruments A and B (filled
and hollow circles, respectively) of the same vari¬
able, whose true value is seen as a line. Clearly,
both instruments have repeatable (reproducible)
measurements and therefore have a low precision FlGURE 1-3 Example of sensitivity.
pies is to use multiple instruments to measure the

same variable simultaneously. Some disadvantages
to this method are the cost and the impracticality,
given the nature of the measuring device. Another
way to increase the number of samples is to mea¬
sure the same variable repeatedly given approxi¬
mately the same conditions. The problem with this
method is that the variable being measured must
be independent of time (unfortunately, most bio¬
mechanical variables are time dependent). Indeed,
recording sample measurements from a nonsta¬
FIGURE 1-4 Example of drift. tionary (i.e., time-dependent) system leads to an¬
other measurement issue, namely sampling fre¬
quency.
tion of the measurement system’s behavior (in Fig. The sampling frequency (time between samples)
1-3 the difference between the circle values and one uses must be small enough so that no relevant
the solid line approximation). portions of the signal are lost. The simplest way
A particular error commonly encountered is to understand this is to examine a nonstationary
drift. Drift, particularly zero drift, can occur to signal such as the one seen in Figure 1-5. This
instrumentation for a variety of reasons, one of signal has only one frequency of 1 Hz. If we were
which is temperature. In the example shown in to take sample measurements of this signal at the
Figure 1-4, we see that a high shift results in zero same frequency (square sample points), plot our
output at nonzero input. Drift, then, is described measurements, and attempt to reconstruct our orig¬
by the shifting of the linear model of the measur¬ inal signal (dashed line), we would obviously have
ing system. Other common errors, such as cross¬ an inaccurate reconstruction. If we sampled at
talk, are discussed in a later section of this chapter. twice the system’s frequency (2 Hz), our recon¬
Because a “true” measurement is realistically structed signal might be better than before, and at
nonexistent, it is common practice to measure a four times the system’s frequency (4 Hz), our
given variable repeatedly in hope of reducing er¬ reconstructed signal is even better. According to
rors associated with either noise or the measuring the Nyquist theorem, the minimum frequency at
device itself. The measurement (X) is then typi¬ which one can sample is at least two times the
cally reported as a mean with a certain confidence system’s frequency. However, the Nyquist fre¬
interval, or X = p, ± c. A variable’s confidence quency is not always the best choice. For example,
interval can be increased by increasing the number in Figure 1-5, if we began our sampling at t =
of measurements according to the central limit 0.25 s instead of t = 0 s, the reconstructed signal
theorem. One way to increase the number of sam- would have been fairly inaccurate.
Example nonstationary signal Reconstructed nonstationary
FjGURE_1 ^ Example of a nonstationary signal and sampling

In reality, a measured signal is unlikely to have

only one frequency component, and therefore de¬
termining a system’s frequency and the corres¬
ponding sampling frequency is not as simple as in
the previous example. In fact, there is an entire
science dedicated to measuring and analyzing non¬
stationary signals (digital signal processing), and
although many of the details involved in signal
processing, such as filtering or analog-to-digital
conversion, are beyond the scope of this chapter,
one concept useful to mention here is a signal’s
distribution.
A signal’s, or in more general terms, a vari¬ FIGURE 1-6 Position of the tennis ball (point P) relative to
able’s distribution allows for determination of the the origin point (point O) located in the floor.
minimum number of samples, or sampling fre¬
quency. Most commonly, variables are assigned
a normal (gaussian) distribution, but this is not If, for example, we are interested in the ballistic
necessarily the case for all variables. Other com¬ motion of an athlete performing a long jump, we
mon distributions are even (white), log-normal, can model the athlete’s body as if the entire mass
Poisson, Weibull, and student-t. The student-t dis¬ of the person were concentrated at a single point
tribution is commonly used when dealing with (the center of mass). Such a model is referred to
relatively few samples (10 < number of samples as a particle model, in which all forces act at the
< 50). Unfortunately, it is not always possible to center of mass and the motion trajectory is the
know the distribution of a given variable; in such motion of the center of mass. Determining the
cases, nonparametric statistics can be used to esti¬ location of the center of mass of an athlete, how¬
mate the number of samples needed. ever, is by no means simple because as the jumper
The review of the measurement jargon pre¬ changes his or her segmental orientation, the loca¬
sented here is simplified and is meant to provide tion of the athlete’s center of mass also changes.
a familiarity with the terminology, not a detailed Therefore, one cannot assume that the center of
description. For a more detailed description of mass is fixed at the level of the navel, regardless
measurements, the reader is urged to examine of the configuration of the body segments.
some of the measurement literature provided at A simpler example of a particle model is given
the end of this chapter. in Figure 1-6, in which the tennis ball is approxi¬
mated by a particle with its center of mass at point
P. Although it might seem that modeling the ball
MECHANICS OVERVIEW
as a particle is obviously appropriate, intrinsic
It is generally appropriate to assume that, in bio¬ to the particle model is the assumption that the
mechanical analyses, we use mechanical tools and orientation of the object being modeled is unim¬
methods to model and analyze biologic systems. portant or unrelated to the objectives. In other
Our subsequent discussion, therefore, focuses on words, the particle model does not account for
the mechanical modeling of the human body, its rotational motion of the ball during its flight,
response to the application of forces, and its per¬ which may prove very important when consider¬
formance as an active force-generating system. A ing aerodynamic effects! To account for such rota¬
brief discussion of applicable mechanical vari¬ tions, it is necessary to model the object as a body.
ables and methods follows. In general, a body has a specific mass, geome¬
The first step in analyzing any mechanical sys¬ try, and material properties. As is commonly the
tem is to create a model that describes the physical case when studying human performance, however,
system’s behavior with a set of simplified ele¬ we often simplify our analysis by ignoring the
ments. These elements can in turn be subjected to compliance of the body and assume that it is rigid.
the corresponding laws of physics and formulated Furthermore, it is possible to combine several
in a mathematical way to derive conclusions about rigid bodies into a multilink or segment model,
such as the model of the tennis player’s various
the system’s behavior.
In mechanical modeling, we describe physical body segments and racquet as depicted in Figure
objects that have specific geometries, masses, and 1-7.
material properties. The type of mechanical model The implication of the rigid body assumption
we choose is primarily a function of the informa¬ in the case of human performance analysis would
tion we hope to extract from the physical system. be that the various body segments are assumed to
ysis assuming either particles or rigid bodies. We

then briefly present the variables representing the
material behavior factors relevant to this text.
The Variables
CLASSIC MECHANICS VARIABLES
As an introduction to our discussion of biome¬

chanical variables, we review the fundamentals of
classic mechanics and its variables, including
mass, position, force, and time. Most of us are
familiar with the concepts of these variables; how¬
ever, a more strict “mechanical” definition is
needed before we can proceed. It is important to
note at this point that units of measurement have
to be specified when dealing with quantitative
maintain a constant shape and mass distribution
values of variables. The most common system of
throughout the task studied. However, our body
units in mechanics is the SI system, by which
segments change shape and mass distribution sig¬
force is measured in Newtons (N), position in
nificantly as a result of muscle contraction and
meters, mass in kilograms, and time in seconds.
forces applied to the soft tissue. Even acceleration
alone can cause substantial deformation to body
parts and shift the positions of internal organs. Mass
The compromise involved in assuming rigidity of
the body segments while dynamic performance is An object has mass, which is defined as some
studied is relatively insignificant. On the other impedance (resistance) to motion. For example, in
hand, if the objective were to study the compli¬ Figure 1-8, the tennis ball has a mass and there¬
ance of the tissue, such as in the case of the fore resists any change to its motion, meaning that
stump-socket interface in limb prostheses or in the player must exert a force to alter the motion
the study of bed sores, no rigidity assumption of the ball. Mass is a scalar quantity; that is, it
would be legitimate. Instead, we would need to can be described by a single number. For example
apply methods employed in analyzing mechanics we would say that the tennis ball has a mass of
of biologic tissues. 0.1 kg. Mass can also be viewed as a material
After formulating a satisfactory mechanical
model, we can construct a free body diagram
of our modeled system. Next, we can apply the y
condition of equilibrium, if the system is static, to
determine unknown forces. For example, as a per¬
son stands (neglecting postural sway or adjust¬
ments), we can consider the person static. We can
therefore apply the static equilibrium condition
and determine such forces as the ground reaction
force. If the system is dynamic, we formulate
the dynamic equations of motion to analyze the
interactive forces (kinetics), to determine the na¬
ture of motion, (kinematics), or both (dynamics).
For example, knowing its mass, we can use the
acceleration and velocity of the tennis ball to
determine the force exerted by the racquet. With
the multisegment model of the tennis player, on
the other hand, knowing the ground reaction force,
we can determine the moment exerted at the dif¬
ferent anatomic joints.
In the subsequent section, we will discuss the
Figure 1 8 Example depicting orientation of a rigid body
variables of interest in a human performance anal¬
(tennis racquet).
property, which is usually measured indirectly

through its weight. The weight of the mass, how¬
z
ever, may change in accordance with the geo¬
graphic position where the weighing takes place
owing to changes in the gravitational field.
Position (and Its Time Derivatives)
The following section may sound complicated to

readers who do not have the appropriate mathe¬
matical background. However, understanding the
concept of describing position and viewing the
same position from different frames of reference
is extremely important when dealing with motion
measurement and analysis and with musculoskele¬
tal forces. A somewhat detailed discussion is
therefore necessary, but this is in no way sufficient
to provide the necessary skills to perform such
analyses; our intention is merely to make the
FIGURE 1-9 The racquet as a rigid body and its
reader aware of some of the issues involved.
orientation defined relative to the global coordinates.
Position describes the location and orientation
of an object relative to some reference frame. For
example, in Figure 1-6, we describe the position
of the ball (modeled as a particle with its center the three global coordinates (e.g., the racquet’s z
of mass at point P) relative to the reference frame unit vector has three angles relative to the global
located in the floor at point O. Notice that to [0zX, 0zY, 0zZ], as in Fig. 1-9), then it seems
describe the position of point P, we need to have like nine values are needed to define an orientation
an array of three measurements in the X, Y, and of one coordinate system relative to another. These
Z directions; position is a vector quantity, and by nine values, however, contain a substantial redun¬
convention, we describe the position vector of a dancy of information, and it can be shown that
point (P) as Rp. The dimensionality of the position only three angles are necessary for such descrip¬
variable, however, is dependent on the modeling tion. In other words, only three independent coor¬
tools used (particle versus rigid body). dinates (degrees of freedom) are sufficient to de¬
The position of a particle is indeed a three- fine the orientation of a body (or for that matter,
dimensional (X, Y, Z) quantity; however, when a coordinate system) in space.
modeling the racquet as a rigid body, we need six In fact, these nine values (of which only three
dimensions (coordinates) to describe its are independent) are used to define the 3X3
position—three to describe the location of some directional cosines (DC) matrix [C], which is used
reference point on the body and three to describe to describe mathematically the orientation of a
its orientation, such as seen in Figure 1-8. We rigid body (or local coordinate system) relative to
may choose to use a set of coordinates describing a global coordinate system (see Fig. 1-9). Trans¬
the position of a point at the location of the palm forming from one coordinate system to another is
of the hand (RH) (X, Y, Z) and a set of angles to accomplished by multiplying the position of a
describe the orientation of the racquet relative to point by the DC matrix [C], It is important to
the XYZ global coordinates. Such angles describe note, however, that such transformations can be
the orientation of the tennis racquet’s (local) coor¬ performed between any two coordinate systems,
dinate system (lowercase x, y, and z in Fig. 1-8) regardless of whether one of the systems is fixed
relative to the global coordinate system (X, Y, Z). in space (global) or both coordinate systems are
Keep in mind that the local coordinate system is moving.
fixed to the racquet and is hence fully representa¬ Typical to the subject of this discussion is, for
tive of the position of the racquet as a rigid body. example, the case of viewing a body with one
Coordinate systems of the kind discussed here camera and trying to determine how the body
are usually represented by a set of three orthogo¬ would appear to a second camera that is displaced
nal vectors whose dimensions are one unit (unit and rotated relative to the first. This process of
vectors). Considering that each unit vector repre¬ either description of an object relative to a coordi¬
senting the axes of the local coordinates x, y, and nate system or transformation of position informa¬
z can be specified by three angles with respect to tion from one frame of reference to another is
Px
Rr ■Py
RPz
We can calculate the position of the ball in terms

of the tennis racquet’s reference frame according
to the following equation:
Rr*
R,P(x,y,z) Rpy — [C]RP(XYZ)
Rp.
1 0 0 'Rr*'
0 COS0 sin0 < RPy
0 — sin0
R,Px
COS0
kJ
FIGURE 1 10 Example of racquet rotation. RPy cos0 + RPz sin0
_ - RPy sin0 + RPz cos0
bidirectional and is mathematically done by inver¬ This, however, is an extremely simple example in
sion of the DC matrix. The inversion of a matrix which the rotation occurred around only one axis;
may be a somewhat tedious process, but in the in general, rotations occur around all three axes,
case of the direction cosines, the matrix belongs making the transformation matrix [C] much more
to a class of orthonormal matrices in which the complicated. Regardless of the complexity of the
inverse equals the transpose: [C]_l = [C]T. What directional cosine [C], the method is the same as
that means is that inversion of the matrix can be that described previously.
obtained by exchanging rows with columns. We have concluded, therefore, that it takes six
For example, if we want to transform the posi¬ coordinates to define the position of a rigid body
tion of the ball, whose motion is originally given in space: three regular cartesian coordinates to
in terms of the global coordinates (X, Y, Z), into define translation and three angular coordinates to
the racquet’s local coordinate system (x, y, z), we define orientation (rotation). Unfortunately, unlike
take the product of the position vector and the cartesian position coordinates, angular coordi¬
transformation matrix [C], as follows: nates are not vectors. Angular displacements do
not obey the law of commutativity, nor can they
Rp(xyz) — [C] RP(XYZ) be added or subtracted by using a parallelogram
law. To state this severe restriction in a simpler
The advantage of such representation may be in language: we cannot sequentially add angular dis¬
viewing the approach of the ball from the perspec¬ placements that occur in a three-dimentional space
tive of the racquet. The matrix [C] contains the to obtain a unique resulting orientation. Changing
angles associated with the orientation of the rac¬ the sequential order changes the result obtained.
quet’s local coordinate system relative to the Conversely, we cannot resolve angular motion into
global system. For the purpose of simplicity of cartesian components as we do with linear posi¬
illustration, let us assume that the racquet was tion vectors. An exception to this rule is planar
rotated only around the X-axis by angle 0 as in motion, in which case only one axis of rotation is
Figure 1-10. possible, or an isolated motion around a single
In this case, the matrix [C] would be calculated axis, such as the spinning motion of a propeller
as follows: of a fan. When the angular displacements are very
small, however, they do obey the commutativity
1 0 0 law and can therefore be added as if they were
[C] 0 cos0 sin0 vectors. The main relief resulting from this excep¬
0 - sin0 cos0 tion is that angular velocities are vectors because
they are derived from infinitely small angular dis¬
placements.
The global position of the ball is calculated as fol¬
lows: To summarize, the position of an object can be
treated as a three-dimensional or six-dimensional
variable, depending on the modeling tools se¬

lected. When using a particle model, in which an
object is represented as a point-concentrated mass,
such as the tennis ball, the orientation of the
particle is meaningless and the location (X, Y, Z)
of a single point on the body is sufficient to
describe its position. Otherwise, if the rigid body
model is used, then both the location (X, Y, Z)
and orientation need to be specified, and hence six
coordinates are necessary per rigid body (seg¬
ment).
FIGURE 1-12 Change in velocity from A to B.
Displacement and Velocity

vector without the need to derive it from the
If we imagine that the tennis ball continues to
orientation coordinates.
travel after the player strikes it and we then track
the motion of the ball, we will see a trajectory as Acceleration
in Figure 1-11, where the ball moves from point P
to point A and finally to point B. The displacement Finally, as the ball follows the trajectory seen in
vector that describes the motion from point P to Figure 1-12, the velocity of the ball changes
point A is given by ATa/p and is equal to the as it moves from point A to point B. Note that the
difference between the position vectors at points velocity, being a vector, can change in magnitude
as well as in orientation. The change in velocity
A and P, or AT a/p= Ta —TP. Furthermore, a dis¬
from point A to point B is defined as the differ¬
placement serves to define velocity, which is the
ence between the velocity vectors at points B and
rate of change of the position vector:
A, or AVb/a = VB - VA, and can be seen in Fig¬
ure 1-12. We refer to the change in velocity versus
- _ dT _ lim / ATa/p\ time as acceleration, which is defined as follows:
A - dt ~~ At -»• 0 \ At /
_* _ dV _ lim / AVb/a\
The displacement and velocity vectors, therefore, aB ~ dt ~ At—> 0 \ At )
lie in the same direction because, like mass, time
is a scalar value and the direction of a vector that The acceleration of the tennis ball at point B
is divided by a scalar remains unchanged. (aB) will have the same direction as the change
Considering the restrictions imposed on the an¬ in velocity vector AVb/a. Thus, position and its
gular coordinate as discussed earlier, one cannot corresponding time derivatives (velocity and ac¬
simply differentiate angles to obtain angular ve¬ celeration) are vector variables. Note, however,
locity (w) or, conversely, integrate angular veloc¬ that unlike displacement and velocity, which are
ity to obtain position. Such simple differentiation always in the direction of motion, acceleration is
can be performed only if the angular position not! An object can move in a curved path at a
(orientation) is represented by a set of three inde¬ constant speed (absolute value of the velocity),
pendent angular coordinates, such as Euler s and yet its acceleration is not zero. This is because
angles. For dynamic analysis purpose, however, the change of direction of the velocity vector by
we are often able to define the angular velocity itself, even without change of magnitude, consti¬
tutes a change in the velocity vector. In this exam¬
ple, if the speed does not change, the acceleration
Z is perpendicular to the direction of motion.
In analogy to linear acceleration, angular accel¬
eration is defined as the rate of change of the
angular velocity vector:
dco
Force
Force is defined as the ability to accelerate a_mass
FIGURE 1-11 Displacement of the tennis ball.
or according to Newton’s second law: F =
ma, where F is force, m is mass, and a is acceler¬

ation. More specifically, Newton’s second law
states that the sum of all forces (2F) acting on a
body of a given mass (m) equals the mass times
acceleration (a) of that body, or 2F = ma. Thus,
the fact that the ball is accelerating as depicted in
Figure 1-12 implies that all external forces acting
on the ball sum to provide a vector in the same
direction as the ball’s acceleration ("a B). For exam¬
ple, consider the forces exerted on the ball by the
racquet (Fraq) and by gravity (Wball) as depicted in
Figure 1-13. The force vector resulting from sum¬
mation of these two forces (£Fball) has the same
direction as the acceleration of the ball. FIGURE 1-14 Depiction of moment.
In fact, the summation of forces seen in Figure
1-13 is an example of one of the most fundamen¬
tal aspect of mechanical modeling, a free body force (R) acting on the foot in Figure 1-15. If we
diagram (FBD). want to calculate the moment of this forcejiround
the knee joint, we have to determine the M = r
DERIVED VARIABLES X F product, which could be expressed as the
following determinant:
The basic mechanical variables described pre¬ A A A
1 j k
viously are used to develop further relations that
Mk = r X R r* ry rz
lead to other variables of interest in a biomechani¬
cal analysis. The following discussion is con¬ Rx Ry Rz
cerned with the most commonly used (in biomech¬
where i, ], and k are the unit vectors of the coordi¬
anics) class of derived variables.
nate system where and R are described.
Moment Momentum
Moment is the product of position and force. Be¬ Momentum is defined as the product of mass and
cause both force and position are vectors, we have velocity (G = mv) and is representative of the
to define the nature of the product, which in this persistence (inertia) of the object in its motion.
case is a cross-product, yielding a vector, or M
= r X F. It is important to note that 7
is a position vector originating at a reference point
(O) and leads to any point on the line of action of
the force F as depicted in Figure 1-14.
Consider, for example, the ground reaction
FIGURE 1—15 Depiction of ground reaction moment.

I Chapter I MEASUREMENTS IN BIOMECHANICS I I
Impulse
Impulse is defined as the time integral of the force

and equals the difference of momenta,
I = /Fdt = A(mv)
Angular Momentum
Angular momentum is defined as H = Sr X mv,

where the summation sign S represents an overall
contribution of all elements (particles) toward
such momentum, r is a position vector of any
such element of the system, and mv is the linear
momentum of such element. An actual detailed
discussion of the specific formulation of the angu¬
lar momentum is beyond the scope of the present
discussion but can be found in any advanced text¬
FIGURE 1-16 Demonstration of the leg inertia properties.
book of engineering dynamics.1
Moment and Product of Inertia ical analyses, although not always with sufficient
justification.
The moment of inertia is a rotational inertia prop¬ To illustrate the meaning of these inertial prop¬
erty that is somewhat equivalent to mass in trans¬ erties, let us consider the leg described in Figure
lational motion. Unlike mass, however, which is a 1-16. Without getting involved with the intricacies
scalar variable, the moment of inertia is a much of the dynamic equations of motion, a moment
more complex variable, called a tensor, which is exerted at the knee joint about the z axis causes
made up of nine components. The inertia tensor the shank to swing forward. The greater the mo¬
is described by the following matrix: ment of inertia of the shank around the knee Izz,
the greater the moment required to accelerate the
I -I xz shank. Considering now the possible existence of
*y
-Iyz products of inertia, such as Ixz or Iyz, such inertia
[I] -I yx O-yy
components cause the shank to swing (rotate)
—I -I zy
around the x axis and the y axis, merely as a
byproduct of its rotation around the z axis. Such
The definition of the different terms of the matrix a cross-effect is often referred to as dynamic im¬
is mathematical and depends on the coordinate balance.
system selected. The terms designated by equal The elimination of such cross-effects by elimi¬
suffices are called moments of inertia, and the nating the products of inertia is referred to as
ones with the mixed suffices are called products diagonalizing of the inertia matrix, so that it ob¬
of inertia. tains the following form:
Moments of inertia and products of inertia are
defined mathematically as follows:
[I] 0 Iyy 0
Ixx = / (x2 + y2)dm 0 0 Izz
and Work
Ixy = f xydm Work is defined analytically as W = / F • dr,

where W designates work, F is the force under
The products of inertia are more difficult to mea¬ consideration, and dr is the displacement that the
sure or assess and are often ignored in biomechan- force has undergone. Note that the force has to
move to produce work. It is also important to

understand that work is a scalar and therefore that = Fx i + Fy j + Fz k
there is no directionality attached to it.
hence:
Energy
Energy is a rather complicated variable, and only W = / F • d? = — A V

some basic features are discussed here. In mechan¬
ics, we distinguish between two types of energy:
Since we have shown above that
kinetic energy and potential energy. Both kinetic
and potential energy are usually assessed on a
differential basis, namely, the change in kinetic or
W = f F • dr - - A(KE)
potential energy is the actual variable being as¬
sessed and not the energy level as such.
and for a conservative force system that
Kinetic Energy. The kinetic energy is derived
directly from the definition of work and the appli¬ W = AV
cation of Newton’s second law. The change in
kinetic energy of a particle can be expressed in we can write
the following form:
W = A(KE) = -AV
2
1
or
A(KE) + AV = 0
= A = A(KE)
which is the principle of conservation of mechani¬

Kinetic energy of a rigid body is based on the cal energy.
same principle but constitutes a summation pro¬ To verify that a system is conservative and that
cess over all the particles that make up the rigid the principle of conservation of energy can be
body. This summation yields an expression of the used, one has to examine the force system and
form (note that for simplicity, only moments of ensure that the nonconservative forces are not
inertia have been included; i.e., the inertia matrix doing work. For example, applying force to stretch
is diagonalized): a spring stores potential energy in the spring.
When the force is released, the work done to
stretch the spring is recovered almost completely,
A(KE) = a(-^) + a(^) through utilization of the stored energy, if the
spring is elastic. On the other hand, when sliding
a box on the ground against friction (a nonconser¬
Potential Energy. Potential energy is the nega¬
vative force), the entire work is dissipated and
tive of the work done exclusively by “conserva¬
cannot be recovered.
tive” forces. The principle of conservation is
In a biomechanical context, consider a person
again a complicated one, and the objective of
jumping from some height and hitting the ground
this discussion is to prevent mistaken assumptions
with taught gastrocnemius and quadriceps mus¬
about conservation of energy. To minimize com¬
cles. As the person approaches the ground, he or
plication of the discussion, we use certain terms
she loses potential energy of gravitation and ac¬
or definitions without getting too analytical.
quires kinetic energy. When the person comes in
A simple definition of the characteristics of a
contact with the ground, knees slightly flexed and
conservative force is that its work does not depend
ankle plantar-flexed, the muscles and connective
on the path, or the work done by such force when
tissues act as a combined spring and damper. The
performing a closed path equals zero. Analyti¬
spring elements store some of the energy, but most
cally, a conservative force is defined as one that
of it is dissipated by the damping components,
can be derived as a gradient of a potential func¬
tion (V): and therefore the jumper’s rebound is minimal. If
the muscles acted as fully conservative springs,
the person would probably rebound to the same
Tf _ 8V _ 8V 8V ->
"consv — 5 1 — T J — —— k height that he or she jumped from. It is to our
§X 8y 8z
benefit, however, to dissipate such impact energy.
Power In the example depicted in Figure 1-17, the

ground reaction force (GRF) and the vector nor¬
Power is defined as the work done per unit of mal to the surface area Af (in this case, a vector
time. Power is also a scalar and is useful in as¬ pointing in the negative z direction) combine to
sessing the instantaneous energy consumption dur¬ produce a stress tensor acting on the surface of
ing performance activities. the shoe:
MECHANICS OF MATERIALS (STRESS AND

•xy Txz
STRAIN)
Q\ Tyz
yy
Overview •yz
It is important to consider the mechanical behavior

Looking at this stress tensor, we see that it has
of materials when subjected to forces for the pur¬
nine components; however, owing to the direction
pose of understanding both the behavior of bio¬
of Af and the symmetric nature of the stress tensor,
logic tissues and some operational principles of
only three are of interest: the axial stress (crzz) and
force-measuring instruments. Mechanics of mate¬
two shearing stresses (tzx and Tzy). The axial stress
rials can be studied on different levels, depending
is given by the z component of the GRF according
on the complexity of the material properties and
to the following equation:
the complexity of the loading conditions as well
as on the geometry of the object under consider¬
ation. Our discussion focuses on the most simple GRFz
relation between stresses acting on materials and
the corresponding deformation that results from
the application of such stresses.
On the other hand, the shearing stresses (tzx and
Tzyare given by the x and y components of the
Stress GRF respectively, as follows:
Imagine the underside of the tennis player’s right
shoe as seen in Figure 1-17. The ground reaction GRFx _ GRFy
force acting on the bottom of the shoe produces a
stress on the surface area of the shoe in contact
with the ground, labeled Af. Stresses are forces
The situation described in Figure 1-17 is a simple
applied on an area; therefore, unlike force, which
case in which only one load was applied at the
is a vector, and time, which is a scalar, stress is
center of the surface area Af. If, however, an
specified through another quantity, called a tensor.
object is subjected to bending in a typical three-
Unlike scalars and vectors, tensors are represented
point situation such as depicted in Figure 1-18,
by a combination of two vectors, and in the case
compressive (negative) stresses develop in the up¬
of the stress tensor, one vector is the force and the
per layers of the object shown, and tensile (posi¬
other is normal to the area on which such force
tive) stresses develop in the bottom layers. As a
is acting. result, we anticipate a deformation of the kind
depicted on the right of the figure.
z
Strain
Because in real loading conditions the stresses

vary throughout the material, deformations are
considered on a local basis. Furthermore, deforma¬
tions are usually expressed as nondimensional lo¬
cal changes in the size of small elements. Typi¬
cally, if we mark a very short line on the surface
of the material under consideration and subject
the material to its loading, the line will change
its dimensions. The relative deformation, namely
FIGURE 1-17 Depiction of the stresses acting on the strain, is defined as the ratio of the change in
tennis player’s shoe. length and the original length:
■+ + + + + + + + + + + + +I
r—-_ _-—n f
FIGURE 1-18
1 1 -
Stress and deformation in a beam in three-point loading.
i
A1 knee joint can be studied by using an optical
motion-analysis system. However, if we were in¬
terested in only the flexion and extension of the
Linear Elasticity knee, we might be better off using a goniometric
linkage that directly measures the joint angle and
Often, material properties are expressed in terms its variation. On the other hand, to study the
of the relation between stress and strain. In solid accelerations acting on a person riding an agricul¬
materials such as metals, wood, and bone, one tural tractor, no motion-analysis system is sensi¬
simple characterization of the material can be tive enough to produce a differentiable motion
made by the stress-strain relation by assuming data representing the higher frequencies of vibra¬
linear (elastic) behavior. Under such an assump¬ tions that may be particularly harmful to the spine.
tion, we can refer to a material’s modulus of In such an event, the choice would clearly be to
elasticity (E), which is the ratio of stress to strain: use accelerometers.
To study repetitive exposure to musculoskeletal
stresses of a postal worker sorting mail and depos¬
iting it into boxes, clearly an optoelectronic mo¬
tion system would be preferred. Supplemental
This relation is the most simple manner in which force information can be acquired from GRF
mechanical behavior of materials can be repre¬ transducers, and effects of muscle fatigue can be
sented. Otherwise, like stress, strain is also repre¬ detected by an appropriate electromyogram
sented by a 3 X 3 tensor, and the general relation (EMG) system. Moreover, an EMG may be re¬
between stress and strain can be described by the quired to facilitate modeling of the musculoskele¬
relation between each component of the stress tal system by reducing the computational redun¬
tensor and corresponding components of the strain dancy, namely identifying the active muscles.
tensor. Such relations, however, are beyond the Clearly, the measuring tools employed depend
scope of the present discussion. Because the ob¬ on the application, and there is definitely no dupli¬
jective of this chapter is to describe the application cation or redundancy in the existence of the differ¬
of mechanics in biomechanical measurements and ent kinematic and dynamic measuring tools.
not to teach applied mechanics, it is confined to
simple examples. MEASURING FORCE
Force can be measured through the effects that it

Measurement Instrumentation creates. According to Newton’s law, force causes
acceleration. If we know the mass of the object
OVERVIEW that is subjected to such force, and if we could
measure the acceleration of the object, we could
As concluded earlier, the basic variables of biome¬ compute the force from the equation F = ma.
chanical measurement are position, mass, force, Unfortunately, single forces seldom appear in iso¬
and time. Theoretically, once these basic variables lation and therefore may not necessarily cause
have been defined or measured, we can proceed motion because other forces are concurrently op¬
to derive other kinematic and dynamic variables, posing them. For example, an object resting on
such as velocity, acceleration, jerk, moments, mo¬ the ground is subjected to its own gravitational
mentum, impulse, work, and energy. In practice, force (weight). At the same time, the ground ap¬
however, such derivation may not be simple; plies a reactive force to the object (according to
hence, the primary variables and measuring tools Newton s third law) that is equal in magnitude
that we select depend on the designated use of the and opposite in direction, and the object remains
information. For example, the kinematics of the in static equilibrium, as illustrated in Figure 1-19.
FIGURE 1-21 Beam balance.
example of such electronic force transducers that

are commonly used is based on strain gage tech¬
nology.
As a result, a more convenient effect that can Strain Gages

be used is the deformation caused to the object by
the force. For example, when we apply a force to The basic principles behind strain gage technology
a spring, the spring compresses, and the amount are discussed first, and the applications of strain
of its compression is measurable. Likewise, many gages in measuring force, acceleration, and other
solid engineering materials exhibit elastic proper¬ related variables are reviewed subsequently. A va¬
ties similar to the spring. If the spring or the solid riety of sensors referred to as strain gages exist,
material has linear force displacement characteris¬ and for simplicity, we discuss the most common
tics, its deflection is directly proportional to the type of resistance strain gage while keeping in
force applied. Technically, the relative change in mind that other types exist and use similar con¬
length is referred to as strain, and therefore many cepts of measurement.
force transducers act as strain sensors. The classic strain gage is an electrically conduc¬
The most common form of force measuring is tive material that, when deformed, undergoes a
weighing. Traditionally, weighing was performed change in its resistance that is directly propor¬
by using some type of a spring mechanism (Fig. tional to the change in its length (strain). For
1-20) or by comparing between the moments of example, pulling on either end of a wire causes a
the weights of masses by balancing a beam on a positive strain and increases the resistance of the
fulcrum, as illustrated in Figure 1-21. Although wire. By passing an electrical current through the
such force-measuring equipment has lasted wire, the change in resistance can be measured as
through the centuries, the spring and beam bal¬ a change in voltage. Furthermore, by choosing
ances have limitations when it comes to measuring appropriate materials, we can develop a linear
dynamic forces. As a result, alternative technolog¬ relationship between the pulling force and the
ies for measuring force have evolved and are corresponding change in voltage.
based on force transducers that can be electroni¬ Specific materials are chosen to serve as gages
cally monitored, thereby allowing the measured based on several material properties, one of which
information to be stored and further processed. An is the linearity between strain and resistance. In
addition, the strain values that allow this linearity
are typically rather small, meaning that the force
(or other loading condition) that can be applied to
such strain gages is also relatively small. As a
result, strain gages are typically mounted to metal
bases so that forces (or loading conditions), which
are the magnitude of interest, can be evaluated. In
such a system, the strain on the surface of the
base is passed on to the gage, and the material of
the base is chosen so as not to exceed strains
outside the linear range of the gage under the
expected loading conditions.
The resistor and semiconductor (or piezoresis-
tive) strain gages are the most commonly used in
biomechanical measurements. The resistor strain
gage is more linear, is less sensitive to temperature
changes, and can measure higher values of strain,
but it is by far less sensitive than the semiconduc-
y _ E RcRa ~f RcRd RdRb RcRd

Equation 1
(Rb + Rc)(Ra + Rd)
If all resistors were identical, V would be zero

and the bridge would be said to be balanced. This,
E however, is unlikely to happen in practice because
of small manufacturing differences in resistance
and the high sensitivity of the bridge. A variable
resistor is typically added in a calibration circuit
to counteract such effects and facilitate balancing
of the bridge. Of importance when measuring
FIGURE 1-22 Schematic description of the Wheatstone force, acceleration, or other variables while using
bridge. strain gages is the linearity of both the gage and
the material to which the gage is mounted. With
appropriate materials, it is possible to develop a
tor strain gage. Typically the gage factor, repre¬
linear relation between the applied force, accelera¬
senting the relation between strain and change in
tion, and so forth and the change in resistance of
resistance, is about 2 for a resistor gage and 125
the conductive strain gage and corresponding out¬
for a semiconductor strain gage.
put voltage of the Wheatstone bridge. For exam¬
Strain is usually measured in nondimensional
ple, the case presented in detail in the following
units of microstrain, which describes the relative
subsection describes the use of strain gages
change of dimensions in the underlying material
attached to a base used to measure force, which is
(or strain gage). In most applications, the strain is
conventionally known as a load cell.
kept well below the yield value of the metals
serving as the structural component of the load
cell. For example, mild steel begins to yield at
Showing the Relation Between
about 1400 microstrains, whereas some other ma¬
Resistance and Force
terials can obtain up to 40,000 microstrains.
Therefore, the more common useful range of
strain gages designed to be mounted on metal is The intent of the load cell depicted in Figure 1-23
within 10-5 to 10-2 of relative resistance change. is to measure the axial force P. Strain gages are
Because of this small change in relative resis¬ mounted on the top and bottom surfaces of the
tance, strain gages are typically placed into an metal base so as to increase sensitivity while re¬
electrical circuit called a Wheatstone bridge. ducing measurement errors, such as those resulting
Strain gages are connected in either one or several from temperature changes, as will be shown
of the positions described by the resistors Ra, Rb, within the following analysis. Furthermore, these
Rc, and Rd in Figure 1-22. A bridge containing strain gages are placed into a Wheatstone bridge
only two strain gages as active resistors is referred in the positions of resistors Ra and Rd. As de¬
to as a half-bridge configuration, whereas a bridge scribed in the variables section of this chapter, the
containing four active resistors is referred to as a loading of the metal base, as seen in the figure,
full bridge. results in compressive stresses on the top and
tensile stresses on the bottom surfaces, as follows:
The Wheatstone Bridge
Equations
Assuming that the bridge’s input voltage E is o'top — — cra - ab Equation 2

constant, the output voltage V can be described
as follows:
abot = — cra + crb Equation 3
Strain gages
FIGURE 1-23 Depiction of a load cell.

Loaded members with strain gages.
Thus, common to both strain gages is the compres¬ In equation 9, we see that the axial strains (i.e.,
sive axial stress (cra), whereas the bending stresses the common strains) cancel each other, while the
(crb) have opposite signs. Furthermore, recall from bending strains sum together, such that we get the
our discussion of stress and strain that for a lin¬ following:
early elastic element such as the metal base in our
example load cell, stress is linearly related to E
strain as follows: V = — — Gfeb Equation 10
a = E•e
Thus, any strain that is common to both surfaces,
such as the axial stresses, or the effects of temper¬
Substituting this relation into equations 2 and 3,
ature are canceled with the Wheatstone bridge in
we have the relation between the strains on the
this configuration. Finally, it can be shown that
top (Ra) and bottom (Rd) strain gages:
the bending strain (eb) is linearly dependent on the
applied load P, as follows:
etop = - (- ea - eb) Equation 4
V = —E— P Equation 11
2 bh2
^bot = ” (+ ea +eb) Equation 5
11
where c, b, and h are constants associated with
Finally, we link our mechanical analysis with the the load cell’s geometry (see Fig. 1-23), Gf is a
Wheatstone bridge by chosing a linear gage that constant associated with the gage, and E is the
has the resistance-strain properties described as constant input voltage to the Wheatstone bridge.
follows: Thus, the output voltage of the bridge, V, is lin¬
early dependent on the applied force, P.
8R
— = Gf€ Equation 6
Measuring Force in General
where Gf is the manufacturer’s gage factor, 8R is The previous example of the load cell is only a
the change in gage resistance associated with an generic description of such a device. Specific load
applied strain, and R is the resistance of the gage cells are designed to measure forces in unidimen¬
with zero strain. sional, two-dimensional, and three-dimensional
We now revisit the original bridge circuit equa¬ configurations, as discussed in the following sec¬
tion with the basic assumption that before any tion. Although most force transducers are made
strain is applied, the bridge is balanced, or Ra from either resistive strain gages or piezoelectric
Rb =» Rc Rd = R (recall that Ra and Rd are crystals, it is possible to use other elements, such
the strain gages). We can then simplify as follows: as conductors or fiberoptics. Regardless of the
nature of the transducer used within the load cell,
one major problem is how to design the cell so
V = E-—-—- Equation 7
2(2R + 5Rd + 8Ra) that force measured in one direction is not affected
by forces acting in other directions, a phenomenon
However, we know that the change in the gage’s referred to as cross-talk. To facilitate automation
resistance associated with the applied strain (8R) of the collection, processing, and storage of data,
will be much smaller than the original resistance as well as to compensate for known transducer
in both Ra and Rd, or mathematically, 8Ra, 8Rd errors, it is customary to place the cell into a data
<< R. We can therefore simplify equation 7 as acquisition system, as seen in the block diagram
follows: of Figure 1-24.
E 8Rd - 8Ra The Dynamic Forceplate

Equation 8
4 ’ R
The dynamic forceplate is a force transducer de¬
Recalling that Ra was the strain gage on the top signed to measure forces applied to the ground by
surface and Rd on the bottom, and combining an individual while walking or while performing
equation 6 with equation 8, we have the following: other athletic, occupational, or leisurely activities.
Traditionally, forceplates consist of a rectangular
plate mounted on four pillar transducers, one at
V = |Gf[(-ea -€„)-(-£. + e„)] Equation 9
each comer of the plate. Each transducer measures
FIGURE 1 24 Block diagram including load cell, wheatstone bridge, signal conditioning, and appropriate
data acquisition equipment.
three force components, one vertical and two ity and compact load cells, whereas their major
shear, as depicted in Figure 1-25. For simplicity, drawback is the cost of the electronics.
only one such set of forces is illustrated. Considering the forces measured by the four
A variety of forceplate models were built over load cells, if all four vertical components are
the years using either strain gage or piezoelectric summed, the total vertical force is obtained: Fvertica)
technology. Generally, the piezoelectric transduc¬ = £Fz = Fzl + Fz2 + Fz3 + Fz4. The same
ers were treated with skepticism because of the summation procedure applies to Fy and Fx. Typi¬
relatively poor response and stability at the low cally, the person walks in the y direction, and
frequency range. However, good charge amplifiers therefore Fy represents the anteroposterior ground
can provide excellent stability and reliability of force, and Fx represents the transverse or mediolat-
the signal and a very good response to low fre¬ eral force. These three components (Fx, Fy, and Fz)
quency signals up to a full directional cosine re¬ can be reassembled into a single vector, represent¬
sponse (constant force). The major advantage of ing the total foot-to-ground force.
the piezoelectric forceplates is their high sensitiv- In addition to the force itself, the foot applies a
moment around the z axis, which can be computed
from the 12 force components described. The z
moment, however, is also influenced by the eccen¬
tricity of the total ground force with respect to the
Y
center of the forceplate; therefore, the location of
the insertion point of the resultant ground force
must be identified. Once again, these x and y
coordinates of the insertion point (or center of
pressure) can be computed from the 12 measured
force components. The actual formulas for the
determination of the position coordinate and the z
moment are not given here, but the general con¬
cept can be stated as follows: the forceplate mea¬
sures a total of six components: three forces and
three moments. The three force components are
the actual forces exerted by the foot to the ground
(or by the ground to the foot). The z moment
(after processing) is an actual moment exerted by
the foot on the ground; however, the moments
around the x and y axes are not actual moments
FIGURE 1-25 Schematic description of forceplate applied to the ground but rather are merely mo¬
illustrating the sensitivity axes and the shear force ments of the resultant force owing to its offset
components acting on one corner transducer.
from the sensitive axes located (theoretically) at
the geometric center of the forceplate. These mo¬ analysis, such as in human locomotion,
ments are therefore used to determine the coordi¬ athletics, and ergonomics.
nates of the center of pressure. In summary, each
In all of these classes, we are usually interested in
forceplate provides information on Fx, Fy, and Fz
more than just position measurements, and we
and on Mz, X, and Y of the center of pressure of
often compute or measure velocities and accelera¬
the foot.
tions. For example, we may want to study the
It is important to be aware of the following:
behavior of a given biologic tissue, such as ten¬
• When the objective is to study the dynamics of don, while it is subjected to different rates of
the human body, the forces to be considered stretching; the variation of angular velocity at a
are the GRFs, namely the forces exerted on the given joint; or the speed at which a tennis player
person and not those exerted by the person on swings his racquet. Moreover, derivation of the
the ground. dynamic equations of motion require that we
• For any computational purpose involving the know the acceleration and velocities. Thus, the
ground forces, one should be aware of the higher derivatives of the position variable are es¬
location of the sensitive axes of the forceplate. sential and can be calculated based on the re¬
• Because the forceplate is a highly constrained corded position measurements or can be measured
transducer, there is a high probability for cross¬ independently.
talk. Usually, cross-talk values are provided by When data processing, such as differentiation,
the manufacturers, but it is highly integration, or transformation into frequency do¬
recommended to perform calibration tests. main, is being considered, it is almost impractical
• Cross-talk is particularly detrimental when to perform such operations manually. Particularly
accuracy of the position coordinates of the when considering measurement of fast movements
center of pressure is sought. such as those occurring in most human perfor¬
• At least two forceplates are needed to monitor mance tasks, including normal locomotion, there
the ground forces during gait. Measuring forces is an obvious advantage to automating the collec¬
on one foot at a time is not a good practice tion and storage of position data. It is therefore
because gait can be significantly biased owing desirable to convert position measurements into
to “aiming” and environmental disturbances; electric signals, which in turn can be recorded
therefore, data gathered from different walk on a computer and made available for further
cycles may be inconsistent. processing.
• The forceplates can be placed either one beside The following section discusses the fundamen¬
the other or one in front of the other with 50% tal principles behind different position measuring
overlap. We prefer the latter. systems as they apply to measuring the different
• For the forces to be measured, the walker classes of biomechanical position measurements.
needs to “hit” the forceplate. This is often
found to be biasing and distorting to the gait Intersegmental Motion
patterns. The question of whether to mark the
location of the forceplates and ask the walker Goniometers
to aim at them, or to completely conceal the
A goniometer is a device that typically converts
forceplates, has been a controversial issue.
the angles between two rigid bodies connected by
Discussion of the advantages and limitation of
some joint into an electrical signal. The most
the aiming approach can be found elsewhere.2^
common form of a goniometer is a rotary potenti¬
ometer (a variable resistor), which is attached to
MEASURING POSITION two arms (calipers). One arm is attached to its
base and the other to its shaft. When the arms
In biomechanics, we are concerned with several
rotate one relative to the other, the potentiometer
classes of position measurements:
changes its resistance in a manner that is propor¬
1. Small displacements, such as those tional to the angular displacement and in this
measured when subjecting a biologic tissue way provides a measure of the angle between the
to stresses to study its mechanical calipers. In biomechanics, goniometers are placed
properties. Such small displacement over an anatomic joint (such as the knee), and the
measurements are considered irrelevant to calipers are fastened to the adjacent body seg¬
the present text and are not discussed here. ments (e.g., the thigh and shank). A simple electric
2. Relative movement between body circuit can be wired to generate voltages propor¬
segments, typically angular displacements. tional to the potentiometer’s resistance, which is
3. General purpose motion documentation and then calibrated to read the angle between these
two segments. The output voltage can be readily

recorded by a data acquisition system.
A variety of methods exist for the transduction
of angular position into an electrical signal. Per¬
haps the most common in biomechanics is the
potentiometer described previously. There are dif¬
ferent types of potentiometers; the two most com¬
mon are the wound wire and the conductive plas¬
tic. The wound wire type is a goniometer in which
the conductive element is made of a thin long
wire that is wound around a circular core and a
slider that slides on the coil transversely and con¬
tacts the coils sequentially. The disadvantage of
such a potentiometer is that its resolution is lim¬
ited to the resistance of a single coil (there are
jumps in readings between the coils). However,
the precision and linearity of such measurement
are much better than those of the conductive plas¬
tic. Conductive plastic, on the other hand, gives
a smooth and continuous measurement (usually
classified as infinite resolution), but its accuracy
and linearity depend on the manufacturing process
and hence its cost. Good-quality precision potenti¬
ometers can be very expensive. FIGURE 1-26 Schematic illustration of a knee goniometer.
Another popular transducer uses the Hall effect,
whereby a magnet within the goniometer is ori¬
knee joint around the other two axes causes further
ented in such a way that as the joint moves, inconvenience and interference.
the magnet moves. Hall effect sensors within the
Goniometric linkages range in complexity from
goniometer produce a voltage directly propor¬ measuring one angular rotation to measuring all
tional to the angle of the magnet; thus, a change three angles. Multiaxial goniometric systems are
in joint angle is seen as a change in voltage. In mechanically complicated and are seldom used
fact, Hall sensors have been developed that use because of their high cost and inconvenience. At
the earth’s magnetic field rather than a magnet 3 degrees of freedom, linkage that fully separates
within the goniometer. In such a system, the the rotations and aligns itself with the anatomic
change in the orientation in the wires relative to axes is not only complicated but also cumbersome
the fixed magnetic field produces voltage signals and interferes with the naturality of the human
that are readily recordable. Although such a sys¬ performance. Such complexity is detrimental to its
tem can be configured to detect angular displace¬ viability as a clinical or even a research tool.
ments, it is beyond what is considered the tradi¬ Although most applications of potentiometers
tional goniometer. are in measuring intersegmental motion, there are
Goniometers are intended primarily to measure other forms of applications. In one such applica¬
single joint motion and not to serve as a whole- tion, an open chain of links is formed from rods
body kinematic system. A major difficulty associ¬ interconnected by multiaxial joints equipped with
ated with the design and mounting of a goniomet- potentiometers. Knowing the relative positions of
ric system is in the fact that all skeletal joints are the links relative to each other can facilitate the
either multiaxial or polycentric. As a result, one determination of the position of the end point of
needs to develop fairly complicated linkages that the device. Such equipment is not suitable for
allow precise measurement of the desired motion monitoring free motion of a point in space but is
while minimizing the constraining effects of the useful for any quasistatic motion, or in particularly
linkage. In other words, consider a knee joint and for describing trajectories and anatomic configu¬
a pair of calipers mounted on top of it, so as to rations. One typical application is to describe the
measure exclusive flexion and extension (see Fig. curvature of a scholiotic spine.
1-26). If the center of articulation of the knee
General Purpose Motion Documentation
does not coincide with the hinge of the calipers,
there will be resistance to the knee flexion, which Optical Imaging Systems
will force the calipers to slide on the thigh and Historically, the use of optical images to measure
the shank. Moreover, the rotation of the anatomic position and subsequently motion dates back to
the mid 1880s and the work of Muybridge,5 who age resolution and significantly lower shutter
captured a series of photographic images of hu¬ speeds. Again, extracting information from video
mans performing various locomotive tasks. Al¬ tape is not a simple matter, but software that can
though advances in both the film and digital im¬ perform such image processing is available (in
aging technologies have progressed rapidly in the some cases on the Internet). Finally, video and
past century, the basic traditional principles re¬ digital cameras that use a variety of optoelectronic
garding machine and image calibration are still principles can also be used to capture motion.
valid. To begin, the image capturing device, usu¬ Such cameras can produce either analog or digital
ally a camera, needs to be calibrated. output signals, which can be readily recorded by
a computer.
Camera Accuracy. Regardless of the image A digital camera’s resolution is described by
medium (film, video, or digital), several aspects several features, including the number of pixels
of camera operation have an effect on the accuracy that capture the light, the type of photoelectric
of any position measurements that might later be equipment being used (e.g., charged couple device
extracted from recorded images. For example, [CCD]), and the bit resolution of each pixel. For
with some systems, the cameras first need to be example, a common resolution is a 256 X 256 8-
focused on the subject being studied. Although bit CCD camera, which is to say that there are
this might seem trivial, consider capturing frontal 65526 CCD pixels, each with an 8-bit resolution.
plane images of a subject walking. It is necessary The clearest advantage of digital cameras is
that the cameras stay focused on the subject as he the ease in image storage and processing. One
or she walks away or toward the cameras. By disadvantage of the system is that most digital
decreasing the aperture of the camera, one can cameras require a separate computer for storing
extend the camera’s depth of field, thus allowing the images, although some cameras have been
the subject to remain in focus for most of the designed with small internal computers. Although
recorded motion, at the cost of reducing the relatively inexpensive digital cameras have a mod¬
amount of light reaching the recording medium as est resolution (8-bit 256 X 256) and sampling
well as measurement accuracy. rates as high as 500 frames per second, the rapid
The time between recorded images and length progress of computer technology is likely to im¬
of exposure (e.g., shutter speed, frames per sec¬ prove, making the storage, resolution, and sam¬
ond, sampling frequency) onto any of the media pling rate deficiencies of current systems a thing
(film, video, or digital) must be appropriately ad¬ of the past.
justed so that no significant motion is lost between
images. As discussed earlier, choosing the appro¬ Image Calibration. One aspect of using im¬
priate sampling frequency for a specific task de¬ ages to measure position that is not likely to
pends on a detailed analysis of the frequency change in the near future is the need for image
components of the given motion. For example, the calibration. Several different techniques for cali¬
frequency components of a hand tremor (8 to 12 brating images are available, but each method
Hz) are significantly different from those in a gait essentially establishes the distance between cam¬
cycle, and therefore the sampling frequency (for era and subject. Without this calibration, the posi¬
tremor at least 24 Hz, but typically 50 Hz) de¬ tions of the various body segments extracted from
pends on the recorded motion. Imperfections in the image have no units and therefore cannot be
the curvature of the lens lead to distortion of the compared with results of other studies, even those
image and must therefore be taken into account with the same camera system.
when extracting measurements. Careful examina¬ The most rudimentary form of calibration is to
tion of both the motion to be recorded and the have the subject move in front of some back¬
proposed camera equipment is needed. ground of known dimension. By combining the
background’s height, anthropometric measure¬
Recording Media. Intrinsic to the choice of a ments of the subject, and camera focal length, the
camera system is the image medium. Cinemato¬ images can be calibrated. More recently developed
graphic medium has been used for many years, digital tracking systems are supplied with a cali¬
and cinematographic analysis has the advantage of bration frame of known dimension. By placing
high image resolution and shutter speed. However, the calibration frame in view of the camera, the
extracting information from cine images is com¬ system’s software performs a self-calibration.
plicated and time-consuming, and the costs of When the distance between camera and subject
developing and storing film can be high. High- has been determined, it is possible to extract fea¬
quality video cassettes can also be used to record tures, such as the position of the knee throughout
human motion, although with relatively poor im¬ a gait cycle, from the recorded images. When
analyzing human motion, the features being ex¬ of an image. In the early cinematography work,
tracted are the positions of specific points on the feature extraction, or digitizing, was done by iden¬
various body segments. tifying bony landmarks on the body segments of
interest from each exposure. The positions of these
Conversion From Two to Three Dimen¬ various bony landmarks, in combination with the
sions. Image media are currently limited to two calibration information, were used to extract the
dimensions; therefore, extracting the three-dimen¬ positions of the various body segments. This is a
sional position of any given point on a body tedious and inaccurate means of measuring posi¬
segment requires at least two cameras. Thus, when tion. By adding markers on the bony landmarks,
using multiple cameras, each camera, as well as the accuracy of film digitizing is increased, but the
the combination of all cameras, must be calibrated. feature extraction process is still time-consuming.
This calibration is by no means simple and relies With the advent of computers and of video and
heavily on statistical averaging to estimate the digital cameras, image storage and manipulation
three-dimensional position of any point. As a re¬ (such as feature extraction) became readily avail¬
sult, using more than two cameras has the benefit
able. Systems typically in use today have taken
of reducing these estimation errors owing to the
several different approaches to both feature extrac¬
increased number of samples. Although it is true
tion and storage. For example, some systems store
that “the more cameras, the more accurate,” the
all the digital images captured by each camera.
decrease in error associated with the addition of
Later, during analysis, the user places “virtual”
another camera becomes less significant with each
markers on the various body segments and can
additional camera, making configurations with
then track the position of these markers (and there¬
more than six cameras impractical. Furthermore,
fore the body segments) through time. These sys¬
having multiple cameras allows for occlusion of a
tems have the advantage of not requiring any
given marker from any one camera’s point of
markers to be placed on a subject. Because the
view—meaning that throughout a recorded mo¬
digital cameras typically capture light from the
tion, at least two (typically more) cameras can
view all body segments. visible spectrum, they can be used in almost any
Reconstruction of the three-dimensional posi¬ environment. However, because these systems
tion of a point based on the combination of images store entire images, the demands on computer
from several cameras has several sources of errors memory for data storage are high, especially if
that affect the accuracy of the position’s measure¬ one considers a multiple camera system with a
ment. For example, by increasing the volume of high sampling rate. For example, if one were to
space from which we want to record a given record with a single 256 X 256 8-bit camera for
motion, we significantly reduce the accuracy of 10 seconds at a sampling rate of 50 Hz, the entire
the position calculation. The sources of these er¬ recording session would produce more than 32
rors are far too numerous and complex to discuss megabytes of data!
here, but it is important to note that the larger the Other systems (known as tracking systems) rely
recording volume (i.e., depth of field), the greater on markers placed on the body at specific loca¬
the error. According to system specifications on tions. The system then records (tracks) the posi¬
certain equipment, it is possible to achieve accura¬ tions of the markers through time. Typically, these
cies to the nearest 0.5 mm for a 1.5-m cubic systems use multiple cameras (four or five) to
volume, or less than 0.05%! Such accuracy, how¬ capture the same image simultaneously. Through
ever, is highly dependent on several site-specific some image processing done through either hard¬
issues, including the position of the various cam¬ ware or software in real time, it is possible to
eras, the lighting of the room, and objects in the extract the three-dimensional coordinates of each
field of view. Such issues have been investigated marker. The information of interest is usually not
thoroughly by numerous investigators; however, the position of a particular point on a body seg¬
there appears to be no decisive conclusion about ment, but rather the position and orientation of the
which aspect of capturing human motion is more entire segment (recall the discussion of the posi¬
correct for every situation, and the photogramme- tion variable). For this reason, when capturing the
try literature should be reviewed before deciding three-dimensional motion of body segments, it is
what camera configuration is best for a particular necessary to place at least three markers on any
task. given segment. Although many image-capturing
systems designed for analyzing human motion
Feature Extraction. To be able to extract posi¬ have software packages that, when provided suf¬
tions of body segments from images, one must be ficient information, can calculate the orientation
able to define a body segment by some features of body segments, it is beneficial to describe the
z
ence in that the transformation results in angles
like flexion and extension, varsus-valgus, and in¬
ternal and external rotation of the shank. The
matrix [CT] can be calculated using linear algebra
if, for example, we use the position of point B
relative to point A in both the global (Ra/b) and
local (ta/b) coordinates:
r A/B [Ct] * Ra/b —^ r A/B * Ra/B 1 — [CT]
The solution from [CT] will be the angles of rota¬

tion of the thigh relative to the global coordinate
system; however, of more interest to us is the
angular rotation of the shank relative to the thigh.
FIGURE 1-27 Determining the six-dimensional
We can determine this by first transforming the
coordinates of a segment using kinematic data where A, B, global positions of points D, E, and F into the^thigh’s
and C are located on the thigh and D, E, and F are on the coordinate system, or TD> E> F = [CT] • RD> E> F .
shank. Next, we determine the transformation matrix of
the shank relative to the thigh, [Cs/T], in the same
way we calculated the thigh’s transformation ma¬
founding principles behind such analyses. Imagine trix. Within the matrix [Csn-] will be the flexion
the situation as depicted in Figure 1-27. and extension, varus-valgus, and internal and ex¬
ternal rotation angles of the shank relative to the
Example of Extracting the Six-Dimensional thigh, which is typically how clinicians refer to
Position of Body Segments. The vector describ¬ joint rotations.
ing the position of C relative to B, that is
Rc«, describes the x direction of the thigh’s local Markers. Several different techniques have
coordinate system and can be calculated using been developed based solely on the type of mark¬
simple vector arithmetic: ers used in any given tracking system. Reflective,
or passive, markers are not likely to interfere with
the motion being measured; however, because
(Rex — Rbx)
each marker is the same, it is not possible to
Rc/r — Rr Rr ' (Rcy — Rby) ’ distinguish between markers based solely on one
„(Rcz — Rbz), image. Instead, image processing must track the
position of the marker for identification purposes.
Light-emitting, or active, markers, on the other
In addition, the vector Rc/B, which describes the z hand, are labeled by a central module and can be
direction of the thigh’s local coordinate system, readily identified by, for instance, sequencing their
can also be calculated. Finally the y direction of ignition. This sequencing or labeling control re¬
the thigh’s coordinate system can be calculated by quires some form of wiring, which renders the
taking the cross-product of the z (Ra/b) and x method cumbersome, especially for use in moni¬
(Rob) directions, or y direction = R^ X Rc/b- toring athletic performance. Because tracking sys¬
With these three vectors, it is possible to deter¬ tems automatically determine the position of the
mine the orientation of the thigh’s local coordinate markers, efforts have been made to minimize the
system relative to the global coordinates. This can error associated with extracting the three-dimen¬
be done by first remembering the discussion of sional coordinates of a marker’s position. Some
multiaxis position from the variables section, researchers have used the infrared light spectrum
where it was shown that it is possible to transform to detect markers, thereby reducing problems as¬
a vector in the global space into a vector in the sociated with reflections within the visible-light
local space using the matrix [CT]. spectrum. Furthermore, software and hardware ex¬
The matrix [CT] contains the angles with which ist that use pattern recognition, whereby the pat¬
the local coordinate system has been rotated rela¬ tern of interest is the image of a marker. The
tive to the global coordinate system. Although camera is then automatically adjusted so that only
several transformations are acceptable (e.g., Car- the marker’s positions are recognized, signifi¬
dan-Euler or helical), the joint coordinate system cantly reducing the amount of information stored
is most readily understandable to the clinical audi¬ for each image.
Nonoptical Imaging Systems
Another way of measuring body motion from a

global reference frame that uses magnetic fields
has been developed. These systems use electri¬
cally conductive coil markers placed on the sub¬
ject. The flux from the magnetic field (either gen¬
erated or natural) induces a current in the marker’s
coils, which is proportional to the angle between
each coil’s sensitive axis and the magnetic flux
lines. By placing three coils with sensitive axes in
orthogonal directions within a given marker, it FlGURE 1-28 Differentiating the y-component dimension
is possible for such systems to record the six¬ of a position vector.
dimensional positions of various markers. These
magnetic systems have the advantage that the hu¬
man body is “invisible” to the detector, meaning
that no marker occlusion occurs. Systems of this entiation. Differentiation of experimental data,
type have the problem that large ferrous objects however, may be a “noisy” process and is not
in the room interfere with the detection of marker always as simple as it sounds; therefore, alterna¬
position as a result of the metal’s interaction with tive methods that measure velocity and accelera¬
the magnetic flux, and thus the operating environ¬ tion directly are sometimes considered. For exam¬
ment is limited. ple, in Figure 1-28, the derivative of y with
respect to time is described as follows:
Marker Systems in General
In any of the marker-based motion-analysis sys¬ dy _ lim /Ay\

tems, the placement of the markers is crucial to dt ~~ At -»• 0 \ At /
the accuracy of the measurements recorded. Most
complete motion analysis systems provide detailed
instructions for placing the markers on a subject. If the curve in Figure 1-28 represents the best fit
to the experimental data points (circles) and hence
One particular problem with all such marker-based
the actual process monitored, then obviously, as
measurement systems is the fact that all positions
being measured are truly positions on the skin, seen from the figure, the slope described by two
not necessarily of the bones underneath the skin. adjacent points is in no way representative of the
Particular applications that require a more precise actual derivative of the curve at the same point.
measure of bone position might use x-ray or mag¬ Moreover, if we take the second and third points
netic resonance imaging technology; however, as and draw the connecting line to determine the
of yet these imaging systems are not designed for slope (derivative), we can see that there is a sub¬
obtaining rapid motion images, and needless to stantial difference between the slopes of the two
say the costs of acquiring multiple images from lines. A graph representing such derivatives based
these systems is prohibitively high. on computation of tangents of adjacent data points
would therefore appear to be a very dispersed one.
Hence, performing a second process of differentia¬
Various Applications
tion to estimate acceleration would yield results
As one can well imagine, the applications of the that are totally useless. A rather complicated
previously described measuring devices to the smoothing process has to be applied to obtain a
field of biomechanics in the evaluation of human close determination of the actual derivatives of Y.
performance are nearly endless. Various research¬
ers have used systems to monitor athletic activi¬
VELOCITY
ties, examine workplace tasks, perform gait analy¬
ses, and evaluate pretreatment and posttreatment
of various musculoskeletal conditions. Direct measurements of velocity have not been
popular in biomechanics because velocity trans¬
ducers are relatively complicated when multivari¬
Position Time Derivatives able human performance is being considered. Ve¬
locity can be measured directly by moving a
The kinematic variables of velocity, acceleration, permanent magnet in the vicinity of a fixed con¬
and jerk (if needed) can be derived through differ- ductive coil. The motion of the flux lines of the
permanent magnetic field through the coil generate

electric current that can be measured. The voltage
77/7777777
generated is proportional to the velocity, but to Strain \
obtain meaningful readings, the geometry of the gages K

system has to be predetermined and the motion
must be highly constrained.
Alternatively, with ultrasound, the Doppler ef¬
5
fect can be used to measure velocity. The Doppler
effect is based on the change of frequency of a
beam generated by a stationary ultrasonic emitter
and reflected from a moving object. Such mea¬
<7 X /' X z z z z z zl
FIGURE 1-29 Example of an accelerometer.
surements are therefore practical with respect to a
translating object when only a single motion vari¬
able is being measured. Alternatively, when multi¬
dimensionality is required, emitters can be accelerate a mass. Figure 1-29 depicts a typical
mounted on the moving object and can be picked configuration of an accelerometer.
up by stationary receivers. Here again, reconstruc¬ The mass is mounted on a cantilever beam, and
tion of a three-dimensional motion data can be a strain gages are mounted on both sides of the
complicated process. beam. As the accelerometer accelerates in the di¬
Of course, both the electromagnetic and the rection of the sensitive axis (indicated by the
ultrasonic Doppler concept can be adapted to mea¬ arrow), the mass lags behind owing to its inertia
sure multidimensional movement, but the cost can and causes the beam to bend. This bending causes
then become prohibitive, accuracy may be com¬ strain in the cantilever beam, which is detected by
promised, and the technical arrangement may be¬ the gages. This strain is related to the force exerted
come constraining to the moving object. Both by the mass on the beam and is therefore propor¬
methods have their advantages and disadvantages, tional to the acceleration of the accelerometer.
but neither has been very popular in human mo¬ As in many other types of dynamic equipment,
tion analysis. such excitation of the beam causes it to oscillate
One rather crude but useful method of velocity around the nominal deflection caused by accelerat¬
measurement is based on conversion of linear mo¬ ing force, and to obtain a correct and stationary
tion into angular velocity. For example, a string reading, these oscillations need to be damped.
wound on a spool when unwound will cause the Efficient damping, under such conditions, may be
spool to spin. The angular motion is easier to achieved by filling the box with viscous fluid.
sense because it can be measured with an electro¬ One feature of the accelerometer that is im¬
magnetic tachometer. Moreover, unlike linear mo¬ portant to understand is the fact that its effective
tion, angular motion generates centripetal acceler¬ measurement is the force applied on the mass by
ation, which in turn can be sensed by a force the beam, and vice versa. As a result, when the
transducer. A speed transducer of this kind had accelerometer is placed in the gravitational field,
been in use at the gait laboratory of Moss Rehabil¬ the beam deflects, even if the box remains station¬
itation Hospital in Philadelphia and was used to ary owing to the weight of the mass, as seen in
document walking speed of test subjects. The Figure 1-30.
string was tied to the waist of the subject and its The effect of gravity is an advantage as well
other end was wrapped around a spring biased as a disadvantage. The disadvantage is that the
spool attached to the tachometer. Of course, such accelerometer will always sense a false accelera¬
a system measures a scalar value velocity at the tion value equal to 1 g, as illustrated in Figure
point of attachment of the string, and special atten¬ 1-30. If the sensitive axis is aligned with the
tion should be paid to the concurrence between direction of gravity, then it will produce a signal
the direction of the string and the direction of equivalent to 1 g (about 9.81 m/s2). If the sensitive
the velocity. axis is rotated at an angle with respect to gravity
line, then the accelerometer will read 1 g
ACCELERATION multiplied by the cosine of such an angle. If the
sensitive axis is once again aligned with gravity
The fact that acceleration is related to force but pointing in the opposite direction, then the
through the second law of Newton is of major accelerometer will read ( —)1 g.
advantage to its measurement. Acceleration mea¬ The advantage, however, is that this effect can
surements are usually performed by measuring the be used for a relatively simple static calibration of
strain that is exerted on a cantilever beam to the accelerometer by flipping it from positive to
cept of inertial navigation in air and space applica¬

tions. However, inertial navigation equipment is
by far less restricted than motion-analysis systems
in terms of weight, volume, and cost. To define
the motion of one segment of the human body in
space, at least six uniaxial accelerometers need to
be placed on the segment such that the accelerom¬
eters’ sensitive axes are aligned with the seg¬
ments’ local coordinate system. The solution of
the resulting six kinematic equations, to obtain the
translation and orientation position variables, is
complicated and computation intensive and is not
presented here. In navigation systems, gyroscopes
g are often used to maintain constant attitude of the
FIGURE 1-30 Effect of gravity.
accelerometric platform (axes); thus, the integra¬
tion of the accelerations remains uncoupled at
all times. Inertial platforms of the kind used for
negative around the horizontal axis. Doing so navigation are extremely costly, heavy, and gener¬
should produce a difference of 2 g provided that ally impractical for application to human motion.
the sensitive axis is kept vertical. Moreover, one complete inertial set is needed per
Although accelerometers in general are highly body segment, thus rendering this concept virtu¬
accurate, placing an accelerometer on the skin to ally useless for three-dimensional motion monitor¬
measure the acceleration of the bone can lead ing.
to rather large errors, depending on the motion.
Imagine placing an accelerometer on the skin over Oscillatory Motion (Vibrational
the triceps and asking the subject to wave his or Effects)
hand frantically. Obviously, the muscle and skin
are compliant and move differently than the under¬ The purpose of measuring oscillatory motion of
lying bone, and therefore extracting the accelera¬ the human body is varied. In diagnostics, vibra¬
tion of the humerus is rather circumspect. In addi¬ tional properties of bone have been studied for the
tion, because most accelerometers have cabled purpose of monitoring the healing process of long
power supply and signal transmission, too many bones and loosening of prosthetic implants. In
accelerometers on a subject can become cumber¬ pathogenesis, vibrations are believed to bear some
some. Furthermore, although it is possible to inte¬ responsibility for the development of degenerative
grate the acceleration signal to determine velocity changes in bone and cartilage (osteoarthritis). In
and then position, this is rarely done because of any of the applications, the understanding of the
the rather large errors resulting from the summa¬ oscillatory properties, that is, the natural frequency
tion process of integration, which tends to accu¬ of the musculoskeletal elements and their trans-
mulate errors resulting from drifts or constant off¬ missibility, were of primary concern.
sets of the signal. Natural frequency is the frequency of oscillation
of an object, such as in the case of a swing or a
How Are Accelerometers Used to Monitor pendulum of a clock or the oscillation of a spring
Biomechanical Phenomena? mass system, as shown in Figure 1-31.
The importance of knowing the natural fre¬
Most of the applications of accelerometers in bio¬
quency or of generally understanding the oscilla¬
mechanics are in the measurement of impulsive or
tory behavior of a system is for the determination
oscillatory motion, in which either the axes of the
of the harm that can be sustained by the musculo¬
body’s fixed coordinates are easily identifiable or
directionality is not of major importance. Often, skeletal system when subjected to such an envi¬
accelerometers are used as expensive event mark¬ ronment. Mechanical systems have a tendency to
ers to identify the onset of an event, such as heel-
resonate when subjected to vibrations whose fre¬
strike during locomotion. quency is close to their natural frequency. The
mechanical energy that enters the system during
Monitoring Human Motion With such excitation gets trapped there, at least in part,
Accelerometers and the disturbance to the mechanical equilibrium
of the system continues to grow and may ulti¬
The idea of using accelerometers for the measure¬ mately produce stresses that are damaging to its
ment of human motion was inspired by the con¬ constituents.
Bone Fracture Healing
An accelerometer is mounted at one end of the

bone, while the other end is subjected to a me¬
chanical impulse such as by a hammer strike (Fig.
1-32). The impulse causes the bone to oscillate at
its natural frequency, which is picked up by the
accelerometer. As the union between the fractured
bone pieces progresses, it is expected that a shift
in both the natural frequency of the bone and the
transmissibility of the oscillation through the bone
will occur. Some problems associated with this
method include the following:
m
• Because the method is noninvasive, both the
FIGURE 1-31 Illustration of an oscillatory system. excitation and the measurement of the
acceleration are conveyed through soft tissue,
which impedes and filters the signal.
• If the excitation is produced by a vibration
It has been determined with a reasonable degree generator, one has to watch carefully that the
of confidence that most humans can be adversely resonant frequencies of the generator are not
affected when chronically exposed to vibrations perceived as those of the bone.
ranging from 0 to 16 Hz. Although the low fre¬ • Despite the natural appeal of the method,
quency range, 0 to 1 Hz, primarily affects the because of its noninvasiveness, it never reached
vestibular organs (causing seasickness) without a clinical level of applicability.
much known cumulative damage, portions of the
human spine have a natural frequency that ranges Vehicle Vibration
between 4 and 16 Hz and can sustain permanent
damage if subjected to vibrations on that level. Vehicles are complex multibody dynamic systems
The International Standardization Organization whose different components are connected by con¬
(ISO) has developed some guidelines with regard straining elements, such as bearings, springs, and
to such exposure. For example, a driver operating shock absorbers. Because of this composition, the
a truck is chronically exposed to vibrations gener¬ vehicle has a whole spectrum of natural frequen¬
ated by the machine (motor plus components). If cies that are excited either by the motor or by the
the vibrations transmitted to the driver are within humps of the terrain. From the biomechanical
the frequency range of 4 to 16 Hz, cumulative standpoint, of major importance are the vibrational
trauma disorder to the spine may result. In this components transmitted to the driver. One should
situation, the critical values pertaining to injury keep in mind that vibrations can be filtered even
susceptibility are the frequency and amplitude of if they are generated. One type of filter is the
the accelerations. Although the frequency of oscil¬ suspension of the vehicle. Another type of filtering
lation of the environment can be obtained through that is more localized is the seat cushion or the
either displacement or velocity measurement, the seat suspension. To study the exposure of a vehicle
variable that most closely represents force is the operator to vibrations, accelerometers should be
acceleration. placed on the seat surface, and measurements
Accelerometer
FIGURE 1-32 Schematic

illustration of an accelerometer
used to monitor healing.
should be performed while the operator sits in his ers, on human cadavers, to relate as well as possi¬
or her normal position and posture. ble the measured accelerations and injury patterns.
A vast amount of relevant literature can be found
Human Response to Vibrations primarily in publications issued by the Society for
Automotive Engineering (SAE).
Various researchers attempted to study or simulate
the response of the human system to vibrational Selection of Equipment
excitation. In such an event, a seated or standing
person is subjected to oscillations produced by a Because the accelerations differ substantially be¬
base on which he or she sits or stands. Accelerom¬ tween the different body locations, one has to be
eters are mounted at different levels on the body, careful with the choice of the mounting site and
in the proximity of the spinal column. Ideally, one mounting method. For example, when the head
wants to determine what is the transmissibility impacts the windshield, accelerations at the front
at the different levels of the spine because the end of the skull reach hundreds of g’s. Placement
intervertebral disks are the most vulnerable ele¬ of an accelerometer at the rear or top of the
ments. In vivo invasive measurements on humans, cranium produces much lower acceleration values.
in which accelerometers are physically affixed to Moreover, which portion of this acceleration is
the bone, are prohibited, and thus the data must transmitted to the frontal lobes of the brain can
be obtained from gages mounted on the surface of only be estimated or measured indirectly using
the body. Such surface measurements are obvi¬ pressure transducers.
ously influenced by soft tissue compliance and Acceleration values as high as 200 g are com¬
have to be treated with great caution. mon in traumatic incidents, such as body impact
On the other hand, to obtain information on against the vehicle interior during automobile ac¬
transmissibility through the higher levels of the cidents. The duration of such acceleration peaks
spinal column up to the head, one can mount an is short, however, and is measured in milliseconds.
accelerometer to a “biting block” (Fig. 1-33). If the peak value occurs for a very short period of
Such an attachment presents a reliable measure¬ time, it may have no injurious consequences. For
ment of the vibrations reaching the head, and one example, in cases of head trauma, peak values of
can compare the ratio of the head acceleration acceleration of 200 g extending over a period of
amplitude with that of the excitation. This, how¬ 2 ms or less are often considered harmless.
ever, does not solve the problem of studying the Of course, for such measurements one needs to
lumbar spine, which is most likely to sustain select the appropriate characteristics of the accel-
such injuries. erometric equipment. The accelerometer must be
capable of measuring values of the order of hun¬
Impulsive Motion dreds of g’s and respond to frequencies of the
order of 1000 Hz without much distortion. On the
One of the most common uses of accelerometers other hand, one should ensure that the natural
is in impulsive motion, often referred to as high frequency of the accelerometer remains outside
acceleration environment. In the automotive in¬ the frequency range of the event being measured.
dustry, human surrogates (dummies) are used to
assess the biomechanical consequences of colli¬
JERK
sions. The dummies are equipped with load cells
and accelerometers that measure the force and Jerk is the time derivative of acceleration or, alter¬
acceleration patterns at different anatomic sites. natively, can be defined as the third time derivative
Concurrently, studies are performed by research¬ of the position vector. Of course, the position
vector can be differentiated any number of times,
but such derivatives may totally lack physical sig¬
nificance. Jerk is a borderline variable in biomech¬
anics and is seldom discussed. To try to visualize
the physical meaning of jerk, one can imagine
riding a vehicle that negotiates a turn and concur¬
rently bounces from side to side (such as roller
coaster). Although in a steady turn, the person
Accelerometer senses a constant centripetal acceleration and
therefore is made to lean away from the center of
curvature. The added lateral bounces affect his or
FIGURE 1-33 Measurement of vibration transmissibility by her acceleration (increasing and decreasing it) and
an accelerometer mounted on a biting block. in turn causes added discomfort. One way of de-
scribing such jerky motion is as the degree of the velocity of point A, Vc is the velocity of point
riding discomfort. Because jerk is not directly
C, a) is the angular velocity vector of the rigid
measurable, one has to measure acceleration and
differentiate it. body, and VB/A and VC/A are the corresponding
Another example of the consideration of jerk in relative velocities between the points.
biomechanics is from the area of control of human Resolution of these vector equations into scalar
motion. In the determination of the strategies em¬ equations produces six such scalar equations, but
ployed by the human nervous system for the syn¬ because of redundancy, only three of the equations
thesis of motion trajectories, it has been hypothe¬ are independent, and their solutions renders three
sized that one of the control objectives is to values for cox, co^ and coz.
minimize jerk. This may sound like a somewhat After the angular velocity vector has been de¬
far-reaching interpretation, but if we consider that fined, the angular acceleration can be determined
jerk, or its absence, reflects a degree of smooth¬ either in a similar procedure by relating linear
ness of the path, it does make sense. Moreover, accelerations of three points or by direct differenti¬
jerk is also representative of the derivative of the ation of the angular velocity.
force characteristic; hence, such minimization of
MEASUREMENT OF MOMENTS AND
jerk is consistent with maintenance of an unper¬
PRODUCTS OF INERTIA
turbed muscle tone.
Moments and products of inertia can be computed,
ANGULAR VELOCITY AND ANGULAR assessed, or measured. For geometrically defined
ACCELERATION MEASUREMENT shapes of rigid bodies, the integration as defined
earlier can be performed. If the body has a compli¬
Unlike linear (translational) velocity and accelera¬ cated geometry, its volume can be divided into a
tion, angular velocity is not usually derived by finite number of geometric elements by forming a
differentiation of the position coordinates, owing mesh, the simplest being a parallel piped, and then
to the restrictions imposed by the noncommutativ¬ summing the contribution of these mass elements
ity of the angular displacement coordinates. An toward the total moment or product of inertia.
exception to this concept is planar motion. The Actual measurement of the inertia matrix com¬
angular acceleration, however, can be derived by ponents can be performed by performing a dy¬
direct differentiation of the angular velocity vec¬ namic test, such as mounting the rigid body on a
tor. In general, dealing with angular motion in a pendulum and measuring the forces and moments
three-dimensional space is a complicated subject generated by this motion with the aid of load cells.
that may require the use of representation of the The measured forces and moments can then be
vectors in terms of rotating coordinate systems. substituted into the dynamic equations of motion,
Such representation complicates the relevant dif¬ with the moments of inertia as the unknown vari¬
ferentiation criteria and is not discussed here. The ables. The solution of these equations will produce
basic concept of measuring angular velocity and the inertia values. The actual equations are com¬
angular acceleration is important, however, and is plicated and are not discussed here.
discussed in an implicit way. Because our primary interest is the biomechani¬
As discussed earlier, six coordinates are needed cal application of the moments of inertia and our
to describe the state of a rigid body in space. The limb segments are not detachable, the above
angular velocity and angular acceleration, there¬ method is hardly practical for in vivo testing.
fore, require the identification of six such indepen¬ However, a number of cadaver tests have been
dent coordinates (degrees of freedom). The motion performed for such definition of the inertia proper¬
of three points on the rigid body have to be mea¬ ties of the human body segments.
sured to facilitate determination of all the degrees A compromised alternative to cadaver testing is
of freedom. After such measurement has been the experimental swinging of a distal limb seg¬
performed, two equations describing the relative ment, such as the shank of the lower limb, and
motion between the points are written and solved monitoring its motion. Substituting the kinematic
simultaneously to obtain a solution for the corres¬ variables into the corresponding dynamic equa¬
ponding angular velocity coordinates. The form of tions of motion and solving these equations can
these equations is as follows: produce some values for the moments of inertia.
Unfortunately, this method can be applied only to
VB = VA + w X RB/A distal segments.
PRESSURE MEASUREMENTS
Vc = VA + w X RC/a
Pressure is used extensively in engineering pri¬
where VB is the velocity of point B, VA is the marily to describe phenomena related to fluid (and
gas) mechanics. It is a less popular term in solids, on optical monitoring of interference fringes cre¬
where our preference is to talk about stress rather ated by polarized light reflection from an appro¬
than pressure. Biomechanical application of pres¬ priate polymer due to impression of rigid cylin¬
sure sensors is primarily for the purpose of moni¬ ders. Although no cables are needed to acquire
toring stresses applied on the interface of soft the data, quantification of this technique involves
tissue. Because of its nature of describing force optical imaging and a considerable processing of
distribution over contact area, pressure is often the image information (per single frame), which
used to describe the way the GRF is distributed renders it computation intensive and cost prohibi¬
throughout the contact area of the foot or the shoe. tive in dynamic applications.
Qualitatively speaking, we can appreciate the A method that gained considerable popularity
amount of pressure we exert as we walk by exam¬ is based on capacitance sensing technology. A
ining the imprints left by our feet in the sand conductive mesh is imprinted between layers of
as we stroll along the beach. Clearly, the sand’s dielectric compressible polymer in such a way that
displacement is proportional to the pressure ap¬ the pattern of one conductor is perpendicular to
plied by one’s feet. Although the materials used in the other. Basically, a matrix of rows and columns
modem pressure transducers are somewhat more is created, forming N X M sensing points. The
complicated, reliable, and expensive than sand, information is acquired by only N + M wires,
the basic concept of measuring the displacement however, which is a substantial saving over N X
of various points on a surface is still valid. M, which would be necessary under ordinary wir¬
Industrial pressure transducers are typically ing conditions. Some systems of this type have
load cells that use the strain gage or piezoelectric reached a full commercial application and are used
transduction technology, and the electrical output primarily in foot pressure monitoring. Several
is used for direct monitoring or further processing problems associated with this technology are simi¬
and control. Such sensors, however, are seldom lar to those listed for conductive plastics. The
suitable to measure biomechanical pressure distri¬ considerable advantage of this technology is in
bution. In biomechanics, usually an array of force the multiplexed nature of the output signals, which
transducers is arranged in a specific geometry to economizes considerably on the number of output
conform to the surface area on which pressure channels. The materials are not sufficiently devel¬
is to be measured. Conventional (strain gage or oped at this point to comply with complicated
piezoelectric) pressure transducers are by far too surfaces, such as the interior of a prosthesis or the
bulky and complicated to form such arrays. More¬ shape of the body in contact with a seat. Foot
over, their power supply and signal conditioning pressure is measured adequately, however, pro¬
systems use conducting cable connections, which vided that proper calibration is performed.
renders such systems impractical. For example, Another concept of pressure distribution mea¬
using ordinary strain gage pressure transducers to surement involves pressure sheets whereby micro¬
measure pressure on the surface of the residual capsules of dye enclosed between translucent
limb of an amputee while using a prosthesis can sheets burst and form stain patterns corresponding
be done only at several sample locations rather to the pressure intensity. A similar method of
than at the entire interface. Using more than 10 staining loose microcapsules was popular in as¬
such points makes the system heavy and cumber¬ sessing excessive pressure on residual limb-socket
some as a result of cabling. Furthermore, the pro¬ interfaces. Of course, this method is quantitative
cedure is invasive, transducers have to be installed and most likely highly nonlinear in the calibration
into the material, and the prosthesis can be used of transparency versus pressure of the film.
only experimentally.
A variety of other attempts to measure pressure
More biomechanically suitable methods of mea¬ have been made.8
suring distributed force or pressure have been ex¬ The major deficiency of the methods of pressure
plored extensively with some degree of success. distribution measurement described here is that
Conductive polymers constitute one group of ma¬
they are genuine “bearing stress” measuring sys¬
terials studied in connection with such applica¬ tems and do not provide information on shearing
tions, but their suitability is still questionable. Sev¬ stresses at the interface. Such shearing stresses
eral problems are associated with conductive may be crucial to certain applications.
polymers: usually only measurement that is rela¬
tive to a calibrated value can be made, and the
measurement has poor repeatability, considerable Summary
hysteresis, and drift of the signal; also, the poly¬
mers are sensitive to high temperatures.
This chapter has presented several concepts rang¬
Another concept has been explored that is based ing in complexity from mass (a scalar measure-
• I Chapter I MEASUREMENTS IN BIOMECHANICS 31
ment) to the six-dimensional position of a rigid Warandale, PA, SAE Publications (SAE PT-43),
body in space and finally to the stress tensor. 1993, p 1200.
Note, however, that the discussion in the variables 7. Backaitis SH (ed): Biomechanics of impact injury
section was not intended to provide absolute defi¬ and injury tolerance of the thorax-shoulder
complex. Warandale, PA, SAE Publications (SAE
nitions. Rather, the point of this section was to
PT-45), 1994, p 1306.
illustrate to the reader that such variables are com¬
8. Doebelin EO: Measurement Systems, Applications
plex and require a detailed analysis. Several ap¬ and Design, 4th ed. New York, McGraw-Hill Inc.,
propriate references can be found at the end of 1990.
this chapter that will aid the reader in obtaining a
more fundamental understanding of the biome¬ Suggested Reading
chanical variables presented in this text.
Allard P, Stokes IAF, Blanchi JP: Three-Dimensional
We then examined some of the instrumentation
Analysis of Human Movement. Champaign, IL,
used to measure variables of interest in many
Human Kinetics, 1995.
biomechanical analyses. Again, a detailed analysis Amirouche FML: Modeling of human reactions to
into the acquisition and underlying theory of many whole-body vibration. Trans AMSE 109:210, 1987.
of the presented transducers was beyond the scope Andriacchi TP, Ogle JA, Glanate JO: Walking speed
of this text. Instead, we examined in detail some as a basis for normal and abnormal gait
simplified examples of different instrumentation measurements. J Biomechanics 10:261, 1977.
to illustrate the transduction nature of measure¬ Asada H, Slotine J-JE: Robot Analysis and Control.
ment systems. New York, John Wiley Interscience, 1985, p 266.
Throughout this chapter, we have repeatedly Brash JI, Skorecki J: Determination of the modulus of
elasticity of bone by a vibration method. Med Biol
seen that the choice of measuring equipment is
Engineering 8:389, 1970.
primarily dependent on the type of biomechanical
Braune CW, Fisher O: Der gang des menschen. i teil
analysis of interest. Thus, as has been stated, de¬ versuche unbelasten und belasten menschen.
termining the appropriate measuring equipment Abhandl J Math-Phys Cldk Sachs Gesellsch 21:153,
requires a detailed analysis of the variables of 1985.
interest as well as an understanding of the measur¬ Bur AJ: Measurements of the dynamic piezoelectric
ing devices themselves. Although manufacturer properties of bone as a function of temperature and
and product analysis literature exists, when pur¬ humidity. J Biomech 9:495, 1976.
chasing equipment for any biomechanical study, it Butler DL, Stouffer DC: Tension-torque characteristics
of the canine anterior cruciate ligament. II.
is beneficial to contact informed authorities about
Experimental observations. Trans ASME 105:160,
the appropriateness of certain equipment in meas¬
1983.
uring specific tasks. With the advances of the Campbell JN, Jurist JM: Mechanical impedance of the
Internet, worldwide web, and so forth, such re¬ femur: A preliminary report. J Biomechanics 4:319,
sources are readily available to clinicians world¬ 1971.
wide. Chesnin KJ, Besser MP, Selby-Silverstein L, et al: A
fourteen segment geometric-based multiple linear
References regression model for calculating segment masses.
Proceedings of the ASB Conference, Atlanta, 1996.
1. Greenwood DT: Principles of Dynamics, 2nd ed. Corradini ML, Gentilucci M, Leo T, Rizzolatti G:
Englewood, NJ, Prentice Hall, 1988, p 552. Motor control of voluntary arm movements:
2. Seliktar R, Mizrahi J, Vachranukunkeit T, et al: Kinematic and modelling study. Biol Cybem
Human performance with prosthetic devices and 67:347, 1992.
surgically modified skeletal elements. Automedica Czerniecki JM, Gitter A, Munro C: Joint moment and
11:145, 1989. muscle power output characteristics of below knee
3. Seliktar R, Susak Z, Najenson T, Solsi P: Dynamic amputees during running: The influence of energy
features of standing and their correlation with storing prosthetic feet. J Biomech 24:1, 1991.
neurological disorders. Scand J Rehabil Med 10:59, Doebelin EO: Measurement Systems Application and
1978. Design, 4th ed. New York, McGraw-Hill, 1990, p
4. Seliktar R: Integrated information approach to 960.
clinical analysis of gait. In Johnson CP (ed): Dorlot JM, Ait Ba Sidi M, Tremblay GM, Drouin G:
Proceedings of the Third ASCE/Engineering Load elongation behavior of the canine anterior
Mechanical Division Speciality Conference, Austin, cruciate ligament. Trans ASME 102:190, 1980.
TX, 1979, p 586. Doud JR, Walsh JM: Muscle fatigue and muscle
5. Muybridge E: The male and female figure in length interaction: Effect on the EMG frequency
motion. Mineola, NY, Dover Publications Inc., components. Electromyogr Clin Neurophysiol
1984. 35:331, 1995.
6. Backaitis SH (ed): Biomechanics of impact injury Elftman H: A cinematic study of the distribution of
and injury tolerance of the head and neck complex. pressure in the human foot. Anat Rec 59:481, 1934.
Elftman H: Experimental studies on the dynamics of fluxes in skeletal muscle fatigue. Can J Physiol
human walking. N Y Acad Sci Ser 26:1, 1943. Pharmacol 69:246, 1991.
Elftman H: Scientific apparatus and laboratory Loeb GE, Gans C: Electromyography for
methods: The measurement of external force in experimentalists. Chicago, University of Chicago
walking. Science 88:2276, 1938. Press, 1986.
Esquenazi A, Talaty M, Silktar R, Hirai B: Dynamic Mansour JM, Lesh MD, Nowak MD, Simon SR: A
EMG during walking as an objective measurement three dimensional multi-segmental analysis of the
of lower limb orthotic alignment. Basic Applied energetics of normal and pathological human gait. J
Myology 7(2): 103, 1997. Biomech 15:1, 1982.
Evaluation of human exposure to whole-body Marey EJ, Demeny: Etudes experimentales de la
vibration. 1. General requirements. International locomotion humain. Compt Rendu Acad d Sc
Organization for Standardization. ISO 2631/1-1095 105:544, 1873.
(E). Mederios J: Automated measurement systems for
Figiola RS, Beasley DE: Theory and Design for clinical motion analysis. Phys Ther 64:12, 1984.
Mechanical Measurements. New York, John Meijer GA, Westerterp KR, Verhoeven FM, et al:
Wiley & Sons, 1991. Methods to assess physical activity with special
Foux A, Seliktar R, Valeiro A: Effects of lower body references to motion sensors and accelerometers.
negative pressure on the distribution of the body IEEE Trans Biomed Eng 38:3, 1991.
fluid. J Appl Physiol 41:5, 1976. Morris JRW: Accelerometery: A technique for the
Gage JR: The clinical use of kinetics for evaluation of measurement of human body movements. J
pathological gait in cerebral palsy. J Bone Joint Biomech 6:729, 1973.
Surg 76-A:4, 1994. Morrison JB: Bioengineering analysis of force actions
Galante J: Human Motion Research. Proceedings of transmitted by the knee joint. Biomed Eng 3:164,
the 48th Annual Meeting of the American Academy 1968.
of Orthopaedic Surgeons, Las Vegas, 1981. Nigg BE, Herzog W: Biomechanics of the Musculo¬
Grassino AE, Clanton T: Mechanisms of muscle skeletal System. New York, John Wiley & Sons,
fatigue. Monaldi Arch Chest Dis 48:1, 1993. 1995.
Griffin MJ: Handbook of Human Vibrations. New Noyes DH, Clark JW, Watson CE: Mechanical input
York, Academic Press, 1990, p 988. impedance of human teeth in vivo. Med Biol Eng
Guzelsu N, Saha S: Diagnostic capacity of flexural 6:487, 1968.
waves in wet bones. ASME Biomechanics Ozgirgin N, Bolukbasi N, Beyazova M, Orkun S:
Symposium, 1983, p 197. Kinematic gait analysis in hemiplegic patients.
Guzelsu N, Saha S: Electro-mechanical wave Scand J Rehabil Med 25:51, 1993.
propigation in long bones. J Biomechanics 14:19, Padgaonkar AJ, Krieger KW, King AI: Measurement
1981.
of angular acceleration of a rigid body using linear
Holman JP: Experimental Methods for Engineers, 6th accelerometers. J Appl Mech 42:3, 1975.
ed. New York, McGraw-Hill, 1994, p 616. Patriarco AG, Mann RW, Simon SR, Mansour JM: An
Isacson J, Brostrom LA: Gait in rheumatoid arthritis: evaluation of the approaches of optimization models
An electrogoniometric investigation. J Biomech in the prediction of muscle forces during human
21:6, 1988. gait. J Biomech 14:8, 1981.
Judge G: Measurement of knee torque during the Pope MH, Wilder DG, Jomeus L: The response of the
swing phase of gait. Engineering in Medicine 4:3, seated human to sinusoidal vibration and impact. J
1976.
Biomech Eng 109:279, 1987.
Katz K, Susak Z, Seliktar R, Najenson T: End bearing Rosenstein AD, McLardy-Smith PD, Cunningham JL,
characteristics of patellar-tendon bearing prosthesis. Turner-Smith AR: The differentiation of loose and
Bulletin on Prosthetic Research 10-32:55, 1979. secure femoral implants in total hip replacement
Kettelkamp DB, Chao EY: A method for quantitative using a vibrational technique: An anatomical and
analysis of medial and lateral compression forces at pilot clinical study. Proc Inst Mech Eng 203'77
the knee during standing. Clin Orthop 83:202, 1972. 1989.
King AI, Mital NK: Computation of rigid-body Roy SH, DeLuca CJ, Casavant DA: Lumbar muscle
rotation in three-dimensional space from body-fixed fatigue and chronic lower back pain. Spine 14 9
linear acceleration measurements. J Appl Mech 1989.
46:4, 1979.
Ryker NJ, Bartholomew SH: Determination of
Kirsch RF, Rymer WZ: Neural compensation for acceleration by use of accelerometers. Advisory
muscular fatigue: Evidence for significant force Committee on Artificial Limbs National Research
regulation in man. J Neurophysiol 57:6, 1987. Council 11:7, 1951.
Krebs DE, Wong D, Jevsevar D, et al: Trunk
Saha S, Lakes RS: The effect of soft tissue on wave-
kinematics during locomotor activities. Phys Ther propagation and vibration tests for determining the
72:7, 1992.
in vivo properties of bone. J Biomech 10:393, 1977.
Lewis JL: A dynamic model of healing fractured long Sarver JJ, Smith BT, Seliktar R: A quantitative
bone. J Biomech 8:17, 1975.
analysis of shoulder motion used for controlling
Lindinger MI, Heigenhauser GJF: The role of ion
FES systems in adolescents with C4 level SCI.
Proceedings of the RESNA Conference, Pittsburgh, characteristics of the canine anterior cruciate
1997. ligament. 1. Theoretical framework. Trans ASME
Seliktar R, Mizrahi J: Gait characteristics of below- 105:154, 1983.
knee amputees and their reflection in the ground Susak Z, Katz K, Seliktar R: Persistence of waddling
reaction forces. Eng Med 15:1, 1986. gait after total hip replacement for congenital
Seliktar R, Mizrahi J: Partial immobilization of the dislocation of the hip. Scand J Rehabil Med 12:113,
ankle and talar joints complex and its effects on the 1980.
ground-foot force characteristics. Eng Med (UK) Sutherland DH, Hagy JL: Measurement of gait
13:1, 1984. movements from motion picture film. J Bone Joint
Seliktar R, Susak Z, Najenson T, Solsi P: Dynamic Surg 54-A:4, 1972.
features of standing and their correlation with Threshold Limit Values for Chemical Substances and
neurological disorders. Scand J Rehabil Med 10:59, Physical Agents Biological Exposure Indices.
1978. ACGIH Worldwide, second printing, 1996, p. 138.
Seliktar R, Susak Z: Limits of applicability of force Vaughan CL, Davis BL, O’Connor JC: Dynamics of
plates in clinical examination of human movement. Human Gait. Champaign, IL, Human Kinetics
Agressology 24:2, 1983. Publishers, 1992.
Seliktar R: Biomechanics of prosthetic gait. In Webster G: Medical Instrumentation: Application and
Esquenazi A (ed): Physical Medicine and Design. Princeton, NJ, Houghton Mifflin, 1991.
Rehabilitation: State of the Art Review. Wiker SF, Chaffin DB, Langolf GD: Shoulder posture
Philadelphia, Hanley & Belfus, 1994, p 89. and localized muscle fatigue and discomfort.
Seliktar R: The use of impulse momentum for Ergonomics 32:211, 1989.
quantification of gait disorders. In DeLuc C (ed): Winter DA: Biomechanics and Motor Control of
Proceedings of the 4th International Congress of the Human Movement, 2nd ed. New York, John
International Society for Electrophysiological Wiley & Sons, 1990.
Kinesiology, Boston, 1979, p 224. Winter DA: Biomechanics of normal and pathological
Sonstegard DA, Mathews LS: Sonic diagnosis of bone gait: Implications for understanding human
fracture healing: A preliminary study. J Biomech locomotor control. J Motor Behav 21:4, 1989.
9:689, 1976. Yang L, Condie DN, Granat MH, et al: Effects of joint
Stokes IAF, Medlicott PA, Wilder PA: Measurement of motion constraints on the gait of normal subjects
movement in painful intervertebral joints. Med Biol and their implications on the further development of
Eng Comput 18:694, 1980. hybrid FES orthosis for paraplegic persons. J
Stouffer DC, Butler DJ, Kim H: Tension-torque Biomech 29:2, 1996.
'
*
CHAPTER 2
Biomechanics of Soft Tissues
Dan L. Bader and Carlijn Bouten
Many noncalcified skeletal connective tissues, novial joints is generally termed articular carti¬
such as articular cartilage, intervertebral disk, liga¬ lage. Articular cartilage is the soft tissue that cov¬
ments, and tendons, have a biomechanical func¬ ers the articulating ends of the bones that
tion in response to the dynamic mechanical loads terminate at a synovial joint (Fig. 2-1). In this
applied in a repetitive manner during normal phys¬ position, cartilage is subjected to the forces that
iologic activity. Each of these tissue structures pass through the joint. Thus, the main functions
constitutes a biologic composite, which contains of articular cartilage are to reduce the contact
the ubiquitous biologic macromolecule collagen in stresses to safe values, protect the subchondral
fibrous form. The building block of collagen fibers bone, and provide the joint with low-friction, low-
is the tropocollagen molecule, 300 nm long and wear surfaces. Indeed, articular cartilage in con¬
1.5 nm wide, with a molecular weight of about junction with its lubricant, synovial fluid, produces
300,000 daltons. Its molecular form is ideally de¬ a coefficient of friction in healthy joints that is
signed to support tensile loads, to which the struc¬ lower than can be achieved for any manmade
tural composites are subjected. Its precise role in engineering system.
a biologic composite depends, however, on the The thickness of cartilage varies from joint to
nature of loading present when the individual soft joint and by location on a joint surface. Cartilage
tissues perform their various functions. This role thickness in major load-bearing joints is generally
is illustrated in this chapter in relation to a series between 1.5 and 3.5 mm, although cartilage lo¬
of soft tissues. cated on the lateral facet of the patella can be 5
The vascularity of most soft tissues is generally mm thick. Young tissue is usually white or bluish
poor and in some cases is nonexistent. Thus, the white and translucent, but with increasing age,
natural self-healing response of soft tissues is lim¬ articular cartilage tends to become yellowish and
ited. Impairment of these soft tissue structures as opaque.
a result of mechanical overload, such as repetitive Normal human articular cartilage does not con¬
strain injury, and of low back pain or disease, tain blood vessels. Thus, pathways must exist to
such as osteoarthritis, often seriously affects body supply its cells with the vital nutrients to maintain
posture and movement and hence normal daily viability and remove waste products. In addition,
functioning. These conditions, which affect people articular cartilage is devoid of nerve endings.
of all ages, are commonly presented in hospital Therefore, any perceived pain from its location
clinics. Indeed, it has recently been estimated that within the joint is probably a direct result of
soft tissue replacements constitute about 35% of abnormal bone contact or other structural damage
the world market for all medical devices.1 from within the joints.
STRUCTURE AND COMPOSITION

Articular Cartilage
Articular cartilage consists of cells known as
GENERAL FEATURES
chondrocytes and an abundant extracellular matrix
Three types of cartilaginous soft tissue are found that contains three major constituents, water, colla¬
in the body: elastic cartilage, fibrocartilage, and gen fibers, and proteoglycans, in the approximate
hyaline cartilage. Hyaline cartilage located in sy¬ proportions given in Table 2-1. The properties of
35
Superficial Collagen
zone parallel to surface
Oblique fibers
Mid zone —
Random fibers
Radial fibers
Deep zone —
Tidemark
Calcified —
zone
Bone —
FIGURE 2 1 Structural variation through the thickness of articular cartilage showing zonal arrangement of
collagen fibers and chondrocytes.
these extracellular components and their interac¬ depth from the articular surface. In normal adult
tions determine the physical and mechanical prop¬ articular cartilage, four cellular zones can be iden¬
erties of the tissue. tified under the light microscope:
Chondrocytes • A superficial zone beneath the articular surface

in which the cells are flattened and orientated
The chondrocytes are responsible for the synthesis parallel to the surface
and maintenance of the extracellular constituents • A transition zone in which the cells are more
of cartilage. The chondrocytes are on average be¬ ellipsoidal, with their axes oblique to the
tween 10 and 30 p,m in diameter and are contained articular surface
within spaces called lacunae. The ratio of cell • A deep zone, sometimes called the proliferative
volume to tissue volume is lower in chondrocytes zone, which contains ellipsoidal or spherical
than in most tissues, accounting for between 1% cells in groups of four to eight arranged in
and 10% of the tissue volume, with a reported columns perpendicular to the articular surface
mean chondrocyte density of 14,000 cells/mm3.2 • A calcified zone beneath the uneven basophilic
The chondrocytes are most numerous near the line, known as the tidemark, marking the
articular surface. transition to calcified cartilage
The characteristics of chondrocytes change with
It is the cells in the transition zone and deep
zone, with relatively large cytoplasmic volumes
containing well-developed endoplasmic reticulum
TABLE 2-1 Relative Proportion of Noncellular
and Golgi complexes that produce the main syn¬
Components in Adult Human Articular Cartilage
thetic activity in cartilage.
Component Wet Weight (%) The chondocytes in articular cartilage are sup¬
Collagen 15-20
plied by nutrients in the synovial fluid, which are
Ground substance 3-15 transported across the cartilage surfaces. Clearly,
Water 65-80 the permeability of cartilage affects nutrition by
Noncollagenous proteins and lipids 1 means of fluid transport; thus, structural changes
Cartilage oligomeric matrix protein that occur with age and disease certainly affect
(COMP) these pathways.
The extracellular matrix (ECM) surrounding
I Chapter 2 BIOMECHANICS OF SOFT TISSUES 37
each chondrocyte can also be divided into three the articular surface. Zones were referred to earlier
regions, known as the pericellular, territorial, and when describing chondrocyte variation according
interterritorial regions. These regions differ in to depth. The superficial zone, including the artic¬
both their proximity to the chondrocytes and their ular surface, extends to a depth of about 200 |xm
structure and organization. The interterritorial re¬ below the surface and contains fibrils about 300
gion is the largest and thus the main contributor nm in diameter. The fibrils are arranged in sheets
to the mechanical properties of the matrix. that lie parallel to the articular surface, and in each
sheet, the direction of the fibrils is approximately
Collagen constant. The fiber orientation varies, however, in
subjacent sheets. Fibrils and fiber bundles are
A large proportion of the nonaqueous ECM of closely packed with little intervening proteogly¬
cartilage is composed of a network of collagen cans when compared with the deeper zones.4 At
fibrils. The building blocks of collagen, the tropo- the surface of the superficial zone, there is a thin
collagen molecules, consist of three polypeptide layer composed of fine fibrils, about 5 nm in
chains, designated a chains, coiled together in a diameter, associated with abundant proteogly¬
right-handed helical structure. Each a chain con¬ cans.4' 5
tains about 1000 amino acid residues. With the Immediately subjacent to the superficial zone is
exception of the short nonhelical sequences at the the transition zone, occupying up to 45% of the
end of each chain, one third of the amino acid matrix volume, in which the collagen fibrils vary
residues are composed of the small molecule gly¬ in diameter between 30 and 60 nm. There are
cine. The remainder are composed of large quanti¬ fewer fiber bundles, and the interfibrillar spacing
ties of proline and hydroxyproline, both of which is increased with larger amounts of proteoglycans.
have inherent rigidity in their ring structure and The fibrils in this zone, which is about 50 p,m
hence confer stability to the triple helix. The pre¬ thick, are more randomly disposed, although there
cise sequence of amino acids determines the type is evidence to suggest a principally radial array.6
of collagen present. The deep zone of cartilage occupies the remain¬
Collagens are conventionally classified ac¬ der of the uncalcified tissue thickness. The colla¬
cording to the structures of the three a chains. gen fibers surround the columns of cells and thus
Articular cartilage contains type II collagen, in also tend to be orientated perpendicularly to the
which each of three identical polypeptide chains interface between the cartilage and the subchon¬
contains a specific content of hydroxylysine resi¬ dral bone.7- 8 In the calcified region, there are
dues and linked prosthetic groups. In the past few increasing deposits of calcium salts in the matrix
years, other types of collagen have been reported surrounding the few cells. The collagen fibrils are
within cartilage, namely types VI, IX, X, XI, and radially aligned in this region, which occupies
XIV. These may have important roles in the over¬ between 5% and 10% of the matrix volume.
all structure of the tissue.3
Collagen fibrils undergo a slow maturation Proteoglycans
process during which intermolecular and intramo¬
lecular cross-links are formed. The formation of The high-molecular-weight macromolecules of
these cross-links requires a degree of overlap be¬ connective tissue are complexes of carbohydrates
tween adjacent molecules, which is consistent and proteins in which the carbohydrate consists of
with the 64-nm periodicity observed in the fibrils repeating disaccharide units. The disaccharide unit
(Fig. 2-2). Intermolecular cross-links may also be is composed of an amino sugar and a uronic
formed between residues located in the helical acid or galactose. Three main glycosaminoglycans
regions of the adjacent molecules. The reducible exist in cartilage: chondroitin sulfate (CS), keratan
cross-links formed in immature cartilage are inter¬ sulphate (KS), and hyaluronan, in proportions that
mediate processes in the formation of more stable, vary with the type of cartilage and with age.
nonreducible cross-links. There is also an increase The term proteoglycan describes macromole¬
in the intermolecular covalent cross-links with age cules of connective tissues composed of a protein
in human tissues. Such observations are consistent core to which a large number of glycosaminogly-
with the marked increase in chemical stability of can chains are attached laterally. A unique feature
the collagen fibers in mature adult cartilage. of cartilage proteoglycans is their ability to form
large multimolecular aggregates, named aggrec-
ans, with molecular weights on the order of 50
Collagen Microstructure
million daltons. The aggrecan consists of a core
Both the diameter and orientation of collagen fi¬ protein to which both CS and KS chains and O-
bers in articular cartilage vary with depth below linked and N-linked oligosaccharides are attached
N-terminal C-terminal
region (nonhelical) region (nonhelical)
Helical /
region /
Tnnrinnnr
Tropocollagen
molecule
Principal covalent
intermolecular cross-links
(Fig. 2-3).9 The core protein has three different

Proteoglycans are polyanions as a result of the
globular domains: the Gl, G2, and G3 domains.
negatively charged groups associated with KS and
The Gl domain is situated at the amino-terminal
CS along their molecular chains. These properties
end of the core protein and binds the aggrecan
control the ability of proteoglycans to bind water
to hyaluronan (see Fig. 2-3). This interaction is
osmotically, and the polyelectrolyte gels may be
stabilized by a link protein that binds to hyaluro¬
considered osmotic systems. The gel behaves like
nan as well as to the Gl region of the aggrecan.
a solution in that it tends to dilute itself with
The aggrecan-hyaluronan interaction retains the
solvent. In cartilage, this process is limited by the
aggrecan within the matrix. When there is a large
extensibility of the collagen fibers with which the
excess of proteoglycan over hyaluronan, as in car¬ proteoglycans are enmeshed.
tilage, a large number of proteoglycan molecules
The proteoglycan content is inhomogeneously
interact with a single chain of hyaluronan, as
distributed throughout the depth of cartilage. In
shown in Fig. 2-4.9
the superficial zone, there is little hyaluronan and,
Other smaller proteoglycans exist in articular
as a consequence, only small amounts of aggrecan.
cartilage, including biglycan, decorin, and fibro-
Instead, the proteoglycans present in this zone
modulin. Their functions are largely unknown, but
are strongly associated with the densely packed
the latter two small proteoglycans are thought to
collagen fibers. The content of proteoglycans in¬
be involved in collagen fibrillogenesis.
creases significantly with distance from the articu-
_E2
Hyaluronan 1

representation of an aggrecan
molecule bonded to hyaluronan by
means of a link protein. (From
Knight MM: Deformation of Isolated
Articular Chondrocytes Cultured in
Agarose Constructs. PhD Thesis.
London, University of London, Link protein _
1997.) Glycosaminoglycan side chains
lar surface, attaining a maximum value in the Only a small proportion of the total water in
middle and deep zones. cartilage is intracellular. It is now well established
that the interstitial water can exist in two distinct
Interstitial Water compartments. About 30% of the interstitial water
is strongly associated with the collagenous net¬
Water readily enters articular cartilage because of work. The remaining water is associated with the
the hydrophilic nature of the proteoglycans. The proteoglycan domains and is freely exchangeable
highest water content is found near the articular during joint loading and unloading. It is this
surface, and the amount of water decreases toward movement that is important in joint lubrication
the subchondral bone, attaining a value of about and in maintenance of chondrocyte viability. Fur¬
65% in the deep zone.10 thermore, the tissue fluid-containing mobile-
FlGURE 2-4 Model of

aggrecan-hyaluran complex in
articular cartilage. (From Knight
MM: Deformation of Isolated
Articular Chondrocytes Cultured in
Agarose Constructs. PhD Thesis.
London, University of London,
1997.)
charged ions, such as calcium, sodium, and hydro¬ forces, and it is clear that the mechanical proper¬
gen, can establish streaming potentials in cartilage ties of cartilage in both compression and tension
that can alter cell metabolism. are important to the transmission of joint forces by
the tissue. The magnitude of the resultant forces
BIOMECHANICAL CONSIDERATIONS regularly reaches five times body weight, equiva¬
lent to about 3500 newtons (N), corresponding to
Load Carriage the period just after the heel-strike of the gait
cycle. In some cases involving vigorous sporting
The physicochemical interaction of the various activities, joint forces can exceed 10,000 N. Tak¬
components of the ECM is responsible for the ing into account the surface area of the supporting
mechanical properties of the healthy tissue. For
cartilage in major load-bearing joints, contact
example, in articular cartilage, there is a physico¬
stresses of the order of 5 megapascals (MPa) are
chemical equilibrium between the osmotic swell¬ common in normal walking.
ing pressure (Pswelling) of the proteoglycan gel,
The behavior of the ECM may be modeled in
which is balanced by the hydrostatic pressures
terms of a biphasic material incorporating a fluid
(Peptic), owing to the tensile stresses generated
phase of interstitial water and a solid phase, the
within the collagen fiber network. This balance
collagen-proteoglycan organic solid matrix, to
exists even in unloaded articular cartilage. It is
represent the fluid-filled porous-permeable me¬
altered when the tissue is loaded in compression
dium.11 12 Other phases have been introduced to
by an applied hydrostatic pressure (Papplied), re¬
accommodate the presence of interstitial ions
sulting in a net pressure differential (Ap) and fluid
within the matrix. These models are discussed
flow away from the compressed tissue. This may later.
be represented as follows:
Biomechanical Studies
Papplied T Pelastic Pswelling
Articular cartilage has been tested in vitro using a

All of these terms may be a function of time.
range of loading modes, including compression,
This fluid flow results in an increased proteogly¬
tension, shear, and torsion. This section concen¬
can concentration within the tissue and a change
trates on the studies that have examined compres¬
in the relative magnitudes of the stresses in the sion and tension.
two solid components of articular cartilage. If the
compressive load remains constant, the rate of
Compression
fluid flow decreases with time and eventually re¬
duces to zero at a new state of equilibrium, as fol¬
Specific test methods have been developed for
lows:
testing cartilage in compression. The two most
common methodologies are the indentation test,
P' swelling = Papplied + P'
A elastic which uses an indenter to apply compression to
discrete areas of the cartilage surface in situ, and
This time-dependent creep behavior is charac¬ the uniaxial compression test, which involves the
teristic of all viscoelastic soft tissues. If the ap¬
use of a platen or large indenter to compress
plied force is suddenly released, the cartilage re¬
isolated cylinders of cartilage perpendicular to the
coils almost instantaneously to a limited extent articular surface. This test has been performed
and then recovers gradually to its original thick¬
with the cartilage specimen in either a confined
ness. This time-dependent component of recovery state or an unconfined state. The former test state
occurs as water is reimbibed into the cartilage
employs porous platens to apply compression to
matrix, until the original unloaded equilibrium po¬ cylinders whose sides are confined.
sition is reached. Cyclic loading produces similar
A critique of the two test methods is provided
behavior, although the extent of the recovery
in Table 2-2. Both methods, however, produce a
within each cycle depends on the form and fre¬
similar response after the application of a rapid
quency of the loading pattern.
load to the cartilage surface, using a damped me¬
The resultant forces transmitted through the ma¬
chanical testing system.12- 13 Figure 2-5 demon¬
jor load-bearing joints of the lower limb are
strates the characteristics of a typical viscoelastic
largely compressive, and the cartilages, with their material.
inherently low frictional properties, are thus sub¬
Many of these studies estimated the creep mod¬
jected mainly to compression perpendicular to the
ulus 2 seconds after load application and related
articular surface. In resisting these compressive
it to the chemical constituents of cartilage.13 They
forces, the collagen fibers experience tensile
found that this parameter, which as its name sug-
TABLE 2-2 Appraisal of Testing Methods of Articular Cartilage in Compression
Method Advantages Disadvantages
Indentation 1. Provides normal physiologic constraints 1. It is difficult to align the indenter

on deformation and fluid perpendicular to the cartilage surface
2. Facilitates assessment of properties 2. Complex stress distribution under the
across joint surface indenter makes it difficult to determine
material constants
Uniaxial compression 1. Unidirectional stress allows the 1. Boundary conditions and fluid flow are
(confined and unconfined) calculation of compressive stiffness different from those of intact joint
2. Specimen size permits controlled tests
with surrounding media of different
concentrations or incubation with
biochemicals
gests includes a degree of creep, was significantly study18 addressed this response by using an under¬
correlated to total glycoaminoglycan content but damped mechanical system to test articular carti¬
was independent of the natural variation in the lage in unconfined compression. The resulting ini¬
collagen content. Stress-relaxation tests have en¬ tial transient oscillatory response of cartilage, as
abled a relaxation equilibrium modulus or relax¬ seen in Figure 2-5, could be used to uncouple the
ation modules (HA) to be calculated. This parame¬ elastic stiffness and the viscous damping in the
ter reflects the intrinsic stiffness of the solid tissue. This investigation examined the effects of
components of the ECM. A summary of represen¬ two enzymes on the short-term compressive prop¬
tative values of derived stiffness for the articular erties of cartilage. It demonstrated that the colla¬
cartilage is presented in Table 2-3.12-15 This table gen component of cartilage largely influenced the
also shows some data from cartilage areas where elastic stiffness, whereas the damping term was
the damage ranged from slight surface fibrillation influenced by the proteoglycan constituents. Ex¬
to gross ulceration. amination of the data related to individual carti¬
In the studies cited in Table 2-3, the short-term lage specimens revealed an inverse relation be¬
response of the tissue was partly determined by tween the elastic stiffness of cartilage and the
the dynamic characteristics of the test apparatus. specimen thickness, t. This relation is consistent
This limitation was recognized by Woo and col¬ with the expression for the stiffness of an elastic
leagues,16 who found that the initial elastic stiff¬ prism, k, of constant cross-sectional area, A, and
ness of the articular cartilage continued to increase constant Young’s modulus, E, as follows:
at very high loading rates and that its measurement
is thus limited ultimately by the response of the k = EA/t
measuring equipment. Indeed, some studies re¬
ported that when loaded rapidly in compression,
cartilage displayed a transient oscillatory response The result implies that the initial “elastic” modu¬
that decayed to a steady creep less than 1 second lus of cartilage in compression is fairly constant in
after application of load.16-18 The more recent healthy joints, yielding a value of about 10 MPa.
TABLE 2-3 Stiffness Values for Adult Human Articular Cartilage
Stiffness Modulus
Joint Surface (MPa) Poisson’s Ratio
Test Model
Femoral head 1.9-14.4 0.50

Indentation
2-Second creep13 Normal areas (mean) 7.0-8.7
Linear elastic Damaged areas (mean) 3.1-6.4
Compression unconfined Femoral condyles 6.1-14.9 —
0.6 second14 Femoral head 5.3-19.7

Compression confined Patella 0.79 < HA < 1.91
Biphasic12
Femoral condyles 0.59 < Ha < 0.70 0.07 < v < 0.10
Indentation
Biphasic15
FIGURE 2-5 Typical response of

articular cartilage to a compressive
step-load input, with, transient
response indicative of an undamped
mechanical loading system.
Tension
gen fibers tend to become aligned in the direction
Collagen fibers are the main tension-resistant ele¬ of the tensile force. The tensile tangent modules
ments in connective tissues. Their presence in in this phase depend on the initial orientation of
articular cartilage suggests that tensile stresses are the collagen fibers and the effective resistance of
present, even though the tissue is loaded predomi¬ the proteoglycan gel for their alignment. With
nantly in compression perpendicular to the articu¬ increasing alignment of the collagen fibers, the
lar surface. component of the applied stress along the fibers
increases, and the stress—strain curve increas¬
Investigators have examined the tensile proper¬
ingly reflects the mechanical properties of the
ties of thin slices of articular cartilage. When an
fibers. With increasing stress, therefore, the tan¬
isolated dumbbell-shaped specimen of cartilage is
gent modulus increases until specimen fracture
subjected to a tensile force in a plane parallel to
occurs. The dependence on the relationship be¬
the articular surface, a nonlinear relationship is
tween the two main structural components and
produced. This response can be regarded as the
tensile properties can be demonstrated by the incu¬
continuous progression between three phases of
bation of cartilage specimens in proteolytic en¬
behavior. Initially, at low levels of stress, the colla¬ zymes.14' 19
There is also considerable variation in the ten¬ of the fatigue processes within the articular carti¬
sile properties with orientation of the specimen lage. Other manifestations of the disease condition
with respect to the predominant direction of the include the formation of microfractures and subar-
superficial collagen fibers. Parallel-orientated ticular cysts in the subchondral bone and gross
specimens yield higher tensile stiffness and geometric changes.
strength when compared with perpendicular-orien¬ Recent interest has developed in devices de¬
tated specimens. Tensile properties also decrease signed to measure cartilage stiffness during clini¬
with respect to distance from the articular surface. cal investigations. One such instrument used under
These results demonstrate both the anisotropic and arthroscopic control aims to measure cartilage
nonhomogeneous nature of articular cartilage. stiffness objectively and to facilitate diagnosis of
In two tensile fatigue studies, Weightman and the softening of articular cartilage at an early
colleagues20-21 demonstrated that human articular stage.23 If such an instrument proves reliable, it
cartilage is prone to tensile fatigue failure in vitro. will enable the assessment of cartilage repair
Tests on specimens from human femoral head after surgery.
cartilage revealed that the fatigue resistance de¬
creased with age at a rate that was faster than that Treatment
which could be explained by normal usage. The
decrease in fatigue resistance was not related to Articular cartilage has limited intrinsic repair po¬
either of the two major solid constituents of the tential. Many options have been proposed for the
specimens. By combining the observations, the repair of the soft tissues. These generally involve
following model was produced: either synthetic or biologic materials. Synthetic
materials have the advantage of providing a struc¬
S = 23 - 0.1a - 1.83 log (N) ture that is immunologically acceptable, with the
mechanical integrity required of load-bearing
where S is the stress in megapascals, a is the age structures. The polymeric nature of most synthetic
in years, and N is the number of cycles to failure. options, however, leads to inherent instability in
the body and thus poor long-term performance.
AGING AND CLINICAL FEATURES Biologic solutions traditionally involve autografts,
allografts, and xenografts. Each of these options
During aging, changes occur in the composition has proved to be far from ideal; for example,
of articular cartilage, including a decrease in total autographs have led to donor site morbidity.
water content and an increase in the concentration Other common solutions involve augmentation
of proteoglycan.22 There are also alterations in the devices incorporating both synthetic structures and
composition and size of the aggrecan, in particular biologic grafts. These devices depend on tissue
a decrease in the proportion of the CS-rich region ingrowth and regeneration induced by the success¬
of the monomer and an increase in the proportion ful transfer of stress from the synthetic material
of the binding region.22 In addition, there is a to the natural tissues. This stress transfer process
reported increase in the diameter of collagen fibrils would need to change with time after implantation
with age. The stiffness of articular cartilage inevi¬ as the tissue regenerates. Clinical reports are not
tably decreases with age and degeneration.14 generally convincing because of, for example, the
Osteoarthritis is a degenerative condition that inadequate initial performance of the synthetic
affects the synovial joint. The prevalence of the component of the augmentation device.
condition increases with age, reaching a maximum The relative failure of many synthetic and graft
in the sixth and seventh decades. However, there solutions has led to the growing interest in the
is a growing number of younger patients who development of cell-seeded repair systems for
develop the disease subsequent to traumatic dam¬ solving a number of clinical problems related to
age of major load-bearing joints, such as the knee. connective tissues, including articular cartilage,
Some of the earliest detectable changes leading menisci, and ligaments. These systems have also
to osteoarthritis occur in the articular cartilage. been called tissue-engineered repair systems. Typ¬
Features of degenerative cartilage are a gradual ically, autologous or allogenic cells are isolated
softening and disruption of the articular cartilage from a tissue biopsy removed from an undamaged
from the bone surfaces. Small fissures appear in site remote to the injury. The cells are expanded
the superficial zone; the subsequent exposure and in cell culture and seeded in suitable three-dimen¬
disruption of the collagen fibers is termed fibrilla¬ sional resorbable scaffold material, which when
tion. Not all cartilage fibrillation leads to a pro¬ implanted into the defect site elicits a biologic
gressive form of degeneration, however—some repair. The potential advantages of these systems
mechanical weakening with age is a direct result include a complete repair that effectively elimi-
nates the problems of synthetic wear particles or positions, the major load on the spine is axial and
prosthetic failure. In addition, the biologic nature mainly produces compression of the intervertebral
of the system may ensure maintenance and turn¬ disk. In normal daily activities, however, the disk
over of the repaired tissue in the long term. Suc¬ is often loaded dynamically and in a complex
cess depends, however, on the ability of the cells manner, owing to combined bending, compres¬
to synthesize a functional matrix at a rate suffi¬ sion, and torsion of the spine. Consequently, the
cient to balance the loss of mechanical integrity actual load on the intervertebral disk is much
of the resorbing scaffold material. Additionally, higher and produces increased compressive as
on implantation, those systems are subjected to well as tensile and shear stresses in the disk.
normal physiologic forces. Thus, mechanotrans- There are 23 intervertebral disks in the spinal
duction pathways are activated that alter cell re¬ column, amounting to one fourth to one third of
sponse in a manner that cannot be predicted in an the total height of the spinal column. No disks are
unstrained cell culture environment. It is vital, found between the axis, atlas, and occiput (the
therefore, to understand the effect of dynamic base of the skull) or between the fused vertebrae
mechanical strain on cells within tissue-engi¬ of the pelvic region of the spine. The individual
neered systems to predict the success of the device disks are roughly cylindrical, with a diameter of
in vivo. In addition, it has been proposed that 2 to 4 cm and a thickness of about 1 cm.30 Disk
the application of strain in vitro using cellular size and shape vary from cranial to caudal, with
engineering techniques may be used to optimize the largest disks being in the lumbar region and
device performance after implantation.24 the most rounded ones in the cervical region of
A number of reports have described in vitro the spine. The ratio of disk height to the height
and in vivo evaluation of cell-engineered systems of the vertebral body varies from 1:3 in the lumbar
for repair of load-bearing cartilaginous tissues.25-27 region to 1:5 in the thoracic region and 2:5 in the
One such approach involving the injection of au¬ cervical region. Because of the greater propor¬
tologous chondrocytes beneath a periosteal graft tional height of the disks, the lumbar and cervical
has been evaluated with some success after a few regions of the spine have greater mobility than the
years in clinical study. Most of the postimplanta¬ thoracic spine. In the lumbar and cervical regions,
tion analysis has involved histologic and biochem¬ the disks are also thicker anteriorly than posteri¬
ical analysis of repair tissue, with only a few orly, resulting in anteriorly convex curvatures of
reports assessing its biochmechanical integrity.26-27 the spine, known as the lumbar and cervical lordo¬
This integrity of the constructs both before and ses. The thoracic spine, on the other hand, exhibits
after implantation is crucial to assessing their a posteriorly convex curvature, known as thoracic
long-term performance. kyphosis.
The mechanical function and load-bearing ca¬
pacity of the intervertebral disk are dependent on
Intervertebral Disk its anatomy and chemical composition and are
influenced by disk degeneration as a result of
GENERAL FEATURES aging or disease. When a disk is loaded beyond
its load-bearing capacity, structural damage may
The intervertebral disk constitutes the primary occur. Usually, damage occurs as a consequence
connection between two adjacent vertebrae in the of high-level, short-term loads that exceed load
spinal column. It allows for motion in all direc¬ tolerance. Submaximal long-term and repeated
tions while being guided and constrained by the loads may also cause disk failure. To understand
uncovertebral and facet joints as well as the verte¬ the mechanisms of disk failure, a basic knowledge
bral ligaments and muscles. Its main mechanical of the mechanical behavior of intervertebral disk
function is the transmission of loads from the tissue under normal and extreme loading condi¬
upper body to the pelvis. This involves the trans¬ tions is crucial.
mission of the weights and resulting moments of
the head and trunk and any additional masses
STRUCTURE AND COMPOSITION
lifted by the upper body. In addition, the disk
serves as a shock absorber in rapidly increasing The intervertebral disk is an avascular, fibrocarti¬
loads. By measuring the intradiskal pressure, it laginous structure consisting of a gelatinous cen¬
has been determined that the loads acting on a ter, the nucleus pulposus, surrounded by the con¬
lumbar disk in sitting and standing with straight centrically arranged fibrous layers or lamellae of
back amount to twice the whole body weight and the anulus fibrosus, and two cartilaginous end
as much as three times the body weight if a plates that cover the top and bottom of the nucleus
weight of 20 kg is held in the hands.28’29 In these and the anulus (Fig. 2-6).
Nucleus pulposus
Vertebral Anulus fibrosus

bodies
Annulus laminate
FIGURE 2-6 Sagittal section through intervertebral disk.
As a result of the presence of both randomly either singly or in groups of four to six. The
and aligned fiber networks, the structure of inter¬ average density of cells is 4000 cells/mm3.2 Like
vertebral disk tissue is anisotropic. Like articular articular cartilage, most of the collagen in the
cartilage, it consists of cells, at a low density of nucleus is type II, but types IV, IX, and XI are
about 5800 cells/mm3- embedded in ECM. Most also found.35 The nucleus forms 25% to 50% of
of the disk volume is taken up by the ECM, which the sagittal cross-area of the disk. It is roughly
consists mainly of water, collagen fibers, and pro¬ located at the center of the disk, except in the
teoglycans. The relative proportions of the ECM large lumbar disks, where the nucleus is more
components change with location in the disk; for posterior than central.
example, the water content is highest in the gelati¬ Because of the high proteoglycan content, the
nous nucleus, whereas the collagen content is nucleus can attract water by means of osmosis and
highest in the fibrous anulus (Table 2-4).31-34 The can mechanically be considered a highly hydrated,
different structures in the intervertebral disk are flattened, incompressible sphere that exerts pres¬
most conveniently described in separate sections. sure in all directions. In the unloaded disk, the
hydrostatic intradiskal pressure is 70 kilopascals
Nucleus Pulposus (kPa) and balances the tensile stresses in the anu¬
lus fibers and longitudinal spinal ligaments.29
The gel-like nucleus pulposus is the stress-ab¬
sorbing and load-distributing center of the inter¬ Anulus Fibrosus
vertebral disk. It consists of highly hydrated pro¬
teoglycan ground substance with a few cells and Although there are significant differences in their
randomly oriented loose collagen fibers. The cells structural organizations, there is no clear boundary
are rounded with a spheroidal nucleus and occur between the nucleus and the anulus: the nucleus
Table 2-4 Relative Proportions of Noncellular Components in the Three Structures of the Adult Human
Intervertebral Disk
Nucleus Pulposus Anulus Fibrosus End Plates
Wet Dry Wet Dry Wet Dry

Weight Weight Weight Weight Weight Weight
Component (%) (%) (%)_W_W_^
5 25 20 65 28 60
Collagen
10 50 3-6 10-20 2-8 5-18
Proteoglycan ground substance
80 * 70 * 55 *
Water
*Values not available.

Data from references 3 I to 3-4.
FIGURE 2-7 A, The nucleus pulposus surrounded by the laminated structure of the anulus fibrosus. Part
of the anulus is removed to show the opposite direction of collagen fibers in the succeeding lamellae. B, The
fibers run parallel at an angle of -30 to 30 degrees to the plane of the disk. (Adapted from Panjabi MM,
White AA: Physical properties and functional biomechanics of the spine. In White AA, Panjabi MM [eds]:
Clinical Biomechanics of the Spine. Philadelphia, JB Lippincott, 1990, p 3.)
simply merges into the anulus in the transitional whereas the outer ones attach directly to the bone
zone or inner anulus. The outer anulus consists of and are known as Sharpey’s fibers. This latter
15 to 25 concentric laminated layers of collagen connection is substantially firmer. On the outside,
fibers, ground substance, a few elastin fibers, and the lamellae are mutually connected by the ante¬
narrow, elongated fibroblast-like cells (9000 cells/ rior and posterior longitudinal ligaments (Fig. 2—
mm3). The elastin fibers play a role in recovery of 8). From the transitional zone outward, the lamel¬
the anulus after deformation. lae increase in thickness, and there is a steady
In the circumferential direction, the anular lay¬ increase in the proportion of collagen. The type
ers, or lamellae, terminate, join, or split into two of collagen also varies in the radial direction: the
layers but do not form complete rings.36 In the inner anulus contains mainly type I collagen that
individual lamellae, the collagen fibers run parallel predominates in tendons and fibrous cartilage,
at an average direction of 30 degrees to the disk whereas the outer anulus contains mainly type II
plane. In adjacent lamellae, however, they run in collagen. Other collagen types found in the anulus
opposite directions and are therefore oriented at are types III, V, VI, IX, and XI.38 Contrary to
120 degrees to each other (Fig. 2-7).37 This ar¬ the collagen content, the proteoglycan, and hence
rangement is optimal for absorbing the centrifugal water content, decreases from the inner to the
stresses generated by the nucleus pulposus in axial outer anulus.39’40
loading of the disk and is believed to play a role
in restricting axial rotation of the spine. End Plates
The inner lamellae of the anulus fibrosus merge The end plates separate the nucleus and anulus
into the cartilage end plates of the vertebrae, from the vertebral bodies. They are mainly com-
Ligamentum flavum
Intertransverse
ligament
Facet capsular
Posterior longitudinal ligament
ligament
Cortical shell Interspinous

ligament
Supraspinous
ligament FIGURE 2-8 Motion segment. To show all
Anterior the ligaments, the posterior arc of a third
longitudinal vertebra has been added. (Adapted from
ligament Panjabi MM, White AA: Physical properties
and functional biomechanics of the spine. In
Vertebral body Facetjoint White AA, Panjabi MM [eds]: Clinical
Biomechanics of the Spine. Philadelphia, JB
Intervertebral disk Lippincott, 1990, p 3.)
posed of hyaline cartilage, which lacks the distinc¬ arrangement of the anulus fibrosus. Thus, its me¬
tive fibrous appearance of the anulus and is more chanical behavior is influenced by the type and
translucent. The plates have a mean thickness of direction of the applied load.
0.6 mm and are weakly attached to the vertebral As emphasized by Panjabi and White,37 it is
bodies by a layer of calcified material.34 At the important to distinguish between the global, exter¬
side of the disk, they connect to the lamellae that nal loads applied to the disk as a whole and
form the inner one third of the anulus. The colla¬ the internal stresses and strains within the disk
gen fibers of these lamellae continue in the end structures. Normal axial loading produces com¬
plate, thereby enveloping the nucleus pulposus. plex stresses and a distribution of forces within
Within the end plate, the collagen fibers are the disk. As a result, the anular fibers are radially
densely packed and run parallel to the surface. and circumferentially stretched as they are pushed
The radial proteoglycan and water profiles show outward (see Figs. 2-2 to 2-4). During compres¬
the same tendency as those in the nucleus and sion, fluid is expelled from the disk in a manner
anulus: ratios are highest at the center and de¬ similar to that with articular cartilage. As a result
crease toward the periphery of the end plate. The of the loss of water, the proteoglycan content
end plates are of major importance to the nourish¬ and hence osmotic pressure increase, whereas the
ment of the intervertebral disk. Small capillaries tension in the collagen-proteoglycan network de¬
stretch from the bone marrow into the end plates creases. The fluid flow continues until the external
and provide essential nutrients. This pathway ac¬ load is balanced. Removal of the compressive load
counts for 70% of the metabolic requirements; the results in an uptake of fluid by the tissue, leading
other 30% is provided by the blood vessels on top to a decrease in the osmotic pressure and an in¬
of the exterior lamellae of the anulus fibrosus. crease in tension of the network. It has been exper¬
Nutritional conditions and biochemistry of the in¬ imentally determined that the intranuclear fluid
tervertebral disk are crucial to its viability and pressure during axial compression on the disk is
mechanical functioning. Therefore, even minor 1.5 times the externally applied load. The tensile
disturbances in end plate structure and composi¬ stresses are much higher: in the dorsal part of the
tion may lead to disk failure. anulus, they can reach values up to 5 times the
applied load.42,43
Motion Segment During flexion, extension, and lateral bending
of the spine, one part of the disk is subjected to
Clarification of the biomechanical behavior and compression while the other part is loaded in
injury mechanisms of the intervertebral disk is tension (Fig. 2-9). There is associated bulging of
largely facilitated by the introduction of the mo¬ the anulus fibrosus on the pressure side and retrac¬
tion segment, first described in 1951.41 A motion tion on the traction side. As a consequence, part
segment, or functional spinal unit (FSU), is a 6- of the anular fibers are stretched, whereas other
degrees-of-freedom structure that comprises two parts are relaxed. Axial rotation of the spine pro¬
adjacent vertebrae with the surrounding soft tis¬ duces torsional loads that result in shear stresses
sues: intervertebral disk, uncovertebral joints, inside the disk. Pure shearing forces acting on the
facet joints, and ligaments (see Fig. 2-8). Its bio¬ disk as a whole do not occur in vivo. Usually,
mechanical behavior is similar to that of the entire they are combined with flexion and rotation.
spinal column, which may be considered as a
series of connecting motion segments. Biomechanical Studies
The mechanical properties of the intervertebral

BIOMECHANICAL CONSIDERATIONS
disk are generally estimated from the viscoelastic
Load Carriage behavior during load-displacement experiments.
Typically, such experiments have concentrated on
Similar to articular cartilage, the intervertebral uniaxial loading because mechanical characteriza¬
disk exhibits time-dependent properties, such as tion during more physiologic simulations of com¬
viscoelasticity, characterized by load-rate sensitiv¬ bined, or coupled, movements (e.g., bending and
ity, hysteresis, creep, and relaxation, to deal with rotation) is extremely difficult to achieve. Test
both short-term high-amplitude loads and long¬ specimens are usually single or multisegmental
term low-amplitude loads. Its mechanical behavior cadaveric motion segments. Isolated disks, or
depends largely on the intrinsic viscoelasticity of specimens from disk tissue, are less frequently
the collagen network but is also highly influenced studied.
by the fluid flow inside the disk. In addition, the The properties of test specimens are presented
disk possesses anisotropy owing to the lamellar by means of structural, load-displacement, stiff-
c
FIGURE 2-9 Schematic representation of the intravertebral disk in unloaded condition (A); in pure axial
loading (B); and in flexion or extension (C). During axial loading, the anulus fibrosus is pushed outward and
the anular fibers are stretched. Bending causes bulging of the disk on the pressure side and retraction on
the traction side. Some of the anular fibers are stretched, whereas others are relaxed.
ness-flexibility, or material (i.e., stress-strain) re¬ The origin of a three-dimensional coordinate sys¬
lationships. The load-displacement curve of the tem has been placed at the geometric center of the
intervertebral disk has the typical nonlinear, sig¬ upper vertebral body. The signs for the directions
moidal shape of viscoelastic materials. At low of force (F) and moment (M) and the correspond¬
loads, the disk provides little resistance and re¬ ing translations (t) and rotations (4>) need to be
mains relatively flexible (neutral zone). At higher defined to reflect the different mechanical behav¬
loads, it becomes stiffer and more resistant, pro¬ ior for positive and negative displacements around
viding greater stability (elastic zone). Owing to and along the same axis. Assuming symmetry with
this viscoelasticity, as well as to the anisotropic respect to the midsagittal plane, the mechanical
structure of the disk, load-displacement curves and behavior for left and right lateral shear ( ± Fy), left
related mechanical properties vary with strain rate
and loading direction. In addition, the nonlinear
nature of the load-displacement curve gives rise
to difficulties when mechanical properties, such as
stiffness or relaxation, are represented by a unique
value.44 Hence, differences in experimental condi¬
tions and setup, as well as the representation of
experimental data, may lead to considerable varia¬
tions in mechanical properties. These factors, to¬
gether with the inherent biologic variation of disk
properties and the limited accuracy of the compli¬
cated experiments, are responsible for the large
differences between mechanical properties re¬
ported in the literature.
For clarity, only average values and general
conclusions are presented here, but the obvious
variation in mechanical properties should be kept
in mind when interpreting the quoted values.
Mechanical Properties of Motion Segments
FIGURE 2-10 Three-dimensional coordinate system

The experimental arrangement for the mechanical
associated with the center of the upper vertebral body of a
characterization of a motion segment is shown in
motion segment. (Adapted from Panjabi MM, White AA:
Figure 2-10. The lower vertebra is fixed to a test
Physical properties and functional biomechanics of the spine.
frame, whereas the upper vertebra is free to move In White AA, Panjabi MM [eds]: Clinical Biomechanics of the
in all directions in response to the applied loads. Spine. Philadelphia, JB Lippincott, 1990, p 3.)
and right lateral bending (± Mx), and left and right Several authors have presented an overview of
axial rotation (± Mz) is identical. stiffness coefficient, or reciprocal flexibility coef¬
In most studies, the motion segment is placed ficients, of motion segments for different modes
in a universal testing machine and subjected to of loading.37’44 On average, stiffness increases
controlled loads, whereas load and deformation from the cervical to lumbar regions of the spine
are recorded simultaneously. Generally, loads are for all loading directions; the cervical spine is
applied statically, with the load increasing in small more flexible, whereas the lumbar spine is more
increments, after which the specimen is given ade¬ resistant to the relatively high-amplitude loads in
quate time to relax before the displacement is this region. A summary of average values and
measured. Consequently, viscous effects are mini¬ ranges reported from a selected number of studies
mized, and the relaxed elastic characteristics are is provided in Table 2-5.45,46,50-58
obtained. In few cases, loads are applied quasi- It is clear that there is little difference in stiff¬
statically or dynamically, with the load continu¬ ness for bending, flexion and extension, and axial
ously increasing at prescribed rates and the visco¬ rotation of the spine. Also, the resistance to shear
elastic characteristics subsequently estimated. is about equal in all directions. Shearing forces
Almost all research on the mechanical proper¬ are less well resisted than forces in the axial
ties of motion segments has been carried out on direction, the stiffness being about 10% of the
lumbar segments. Many of these properties, how¬ resistance to compression. The highest stiffness of
ever, are relatively consistent throughout the spine the disk is seen during compression, probably
and are merely dependent on loading direction because of the increased nuclear pressure seen
rather than spinal level. Therefore, mechanical with compression of the disk.
properties of motion segments, whether in the After periods of sustained loading of the motion
cervical, thoracic, or lumbar spine, are included segment, the intervertebral disk shows creep. In
as average values or ranges with respect to loading quasi-static compression, disk creep is five to
direction. Properties that might be expected to seven times higher in magnitude than creep in the
vary with spinal level are described separately. bony structures of the segment.59 A 3-hour period
The displacement of a motion segment in the of compressive loading at 1200 N, comparable to
flexible neutral zone of the load-displacement unsupported sitting with anterior flexion, results
curve is on average 0.3 mm for tension and com¬ in a 10% decrease in disk height and a 5% to 13%
pression and 0.8 mm for shear.45 It ranges between increase in sagittal disk diameter,60 which is of the
1.7 and 3.2 degrees for lateral bending, flexion, same order of magnitude as that for diurnal varia¬
extension, and axial rotation.46 The physiologic tion seen in vivo. Higher compressive loads pro¬
range of motion, covering the neutral zone and duce greater deformation and a higher creep rate.
the subsequent linear elastic zone of the load- A significant fraction of creep under compression
displacement curve, ranges between 0.1 and 1.9 is due to fluid loss,61 but creep and recovery char¬
mm for tension, compression, and shear,45' 47 acteristics of the disk are also influenced by the
whereas values ranging from 1 to 17 degrees have viscoelasticity of the anulus fibrosus.62 Creep
been reported for lateral bending, flexion, exten¬ causes the intranuclear pressure to fall and the
sion, and axial rotation.46-48'49 anular stresses to rise. The disk becomes stiffer in
TABLE 2-5 Mean Stiffness Values and Ranges of Spinal Motion Segments
Name Mean Stiffness Stiffness Range

Direction of Load
N/mm N/mm
Anterior shear 105 34-183

+ Fx
Posterior shear 105 49-189
Fx
Lateral shear 150 53-385
+ Fy
Tension 417 53-1000
+ FZ
Compression 1065 141-2500
Fz
N m/deg N m/deg
Lateral bending 1.6 0.01-7.7

+ Mx
Flexion 1.2 0.1-2.2
+ Mr
Extension 1.2 0.1-2.8
My
Axial rotation 1.3 0.2-5.0
±MZ
Data from references 45, 46, and 50 to 58.

compression and more flexible in bending.63 Thus, exception of investigations involving the anulus
the loading history is important for both the me¬ fibrosus39-67-68 and the cartilage end plates.69 Be¬
chanical behavior and the load-bearing capacity of cause of the anisotropy and inhomogenicity of its
the disk. structure, the mechanical properties of the anulus
Hysteresis of disk tissue in motion segments fibrosus vary with region and direction of applied
was first described by Virgin,58 who noted that load. To date, experiments have been performed
hysteresis was most marked in the lower lumbar to determine mechanical properties of single or
spine and increased with the applied load. Hyste¬ multilayer samples of anulus fibrosus. Cylindrical
resis decreased when the same segment was or dumbbell-shaped samples are cut from different
loaded for the second time. This may explain why regions of the anulus and loaded in compression
the intervertebral disk is less resistant to repetitive or tension, respectively. Samples are mounted and
or vibrational loads than to sustained or single¬ tested in universal testing machines to study ten¬
impact loads. The number of load repetitions that sile properties. Compressive properties are studied
can be tolerated before disk damage occurs has in uniaxial confined or unconfined compression,
been extensively studied in fatigue tests of motion the characteristics of which are discussed later.
segments.51- 61- 64- 65 Although there is little Figure 2-11 shows an example of a confined
agreement between the different studies and ex¬ compression experiment.70 A cylindrical sample is
perimental outcomes, it is generally concluded uniaxially confined with a load of 0.2 MPa, and
that risk of disk damage increases with load and its height variation is plotted as a function of time.
number of cycles. Furthermore, major damage is To stabilize the hydration of tissue, the sample is
inflicted by cyclic axial compressive loads, bathed in physiologic saline solution.
whereas cyclic flexion and extension and axial In general, studies on anular mechanical proper¬
rotation may not be capable of producing any ties have shown that axial compressive stiffness
serious damage to the disk because of the protec¬ varies with location, being higher in outer regions
tion provided by the ligaments and associated and the posterior part than in the inner regions
muscles.66
and anterior part.39 Tensile stiffness shows higher
values in the anterior and posterior parts as com¬
Mechanical Properties of Disk Tissue pared with the lateral parts and outer regions,
which are stiffer than the inner regions.68-71 Thus,
Relatively few studies have been completed on the central region near the nucleus pulposus ap¬
the mechanical properties of disk tissue, with the pears to be the weakest area of the anulus, whereas
5 10 15
Time (h)
FIGURE 2-11 Sample height versus time for a 15-h confined compression
experiment on anulus fibrosus tissue. The sample was taken from the outer region of
the anulus and compressed with a load of 0.2 MPa. The gradual increase in sample
height at the beginning of the experiment represents the swelling of the sample when
it was placed in saline solution. The solid line indicates the mechanical behavior during
the compression phase as predicted by a finite element model of the tissue (Adapted
from Houben GB, Drost MR, Huyghe JM, et al: Nonhomogeneous permeability of
canine anulus fibrosus. Spine 22:7, 1997.)
the strongest parts are found on the anterior and a stress higher than the ultimate failure stress is
posterior sides of the disk. With respect to the generated at a given point. The mechanism of
orientation of the anular samples tested in tension, failure during long-term, low-amplitude loads and
Galante42 has demonstrated that anulus layers are repetitive loading of relatively low magnitude is
weakest when loaded in the axial direction and entirely different and is due to fatigue failure: a
strongest when loaded along the major fiber direc¬ tear develops at a point where the nominal stress
tion. is relatively high and enlarges, eventually leading
Setton and colleagues69 studied creep behavior to complete disk failure. Although exact load tol¬
of cartilaginous end plates in confined compres¬ erance levels cannot be given, especially in the
sion experiments. They found that creep rates of case of fatigue failure, it is clear that they are
the end plates were significantly higher than those highly dependent on the chemical and structural
observed for articular cartilage. Possible changes composition of the disk, which changes with age
in mechanical behavior with relation to biochemi¬ and disease. Age-related changes in structure and
cal composition and location within the end plate composition of the intervertebral disk are respon¬
were not studied. sible for a more than two-fold increase in the
incidence of disk-related back problems in people
Modeling 45 to 65 years of age, compared with those 18 to
44 years of age.79 In addition, diseases such as
To achieve more insight into the mechanical prop¬ osteoporosis and metabolic disorders such as
erties of the intervertebral disk, and in particular ochronosis accompanied by disk calcification in¬
into the internal stresses and strains inside the fluence the composition of the disk and hence
disk, various mathematical models have been de¬ increase its sensitivity.59
veloped. The recent application of biphasic mix¬ With increasing age, the metabolic exchange in
ture models, describing a solid component and a the tissues of the intervertebral disk retards, so that
fluid component in their formulation, is popularly the tissues simultaneously change in biochemical
used to predict the mechanical behavior of the quality and fluid content. The water content of the
disk.39,69’72 These models are based on assump¬ nucleus decreases from more than 80% in young
tions of elastic or viscoelastic material behavior adults to about 70% after the fifth decade.80 The
and incorporate the movement of interstitial fluid. proteoglycan concentration and thus the water¬
Hence, they are capable of predicting flow-depen¬ binding capacity in the nucleus also decrease; by
dent mechanical behavior. A third or fourth phase the age of 60 years, these measure one half the
or component is added to reflect further the com¬ value of those in the young adult.32 On the other
plicated interaction between the biochemical con¬ hand, the concentration and diameter of collagen
stituents and the mechanical behavior of the fibers increase, accompanied by an increase in
disk.73’74 With the increasing complexity of these collagen-proteoglycan binding. The relative con¬
models, however, experimental validation is also centration of type I fibers to type II fibers de¬
more difficult.75 Furthermore, considering the in¬ creases. In contrast, the concentration and diame¬
creasing number of assumptions and simplifica¬ ter of collagen fibers in the anulus fibrosus
tions, care should be taken to avoid overgeneraliz¬ decreases,36 and the ratio of type I to type II
ing the numerical solutions of the models. collagen increases.81 As a result, the parallel fiber
Nevertheless, mathematical modeling is a helpful arrangement is gradually lost, and the distinction
and necessary tool to study the biomechanics and between the anulus and the nucleus vanishes.
related injury mechanisms of the intervertebral There is also evidence of a decrease in proteogly¬
disk. can and water content of the anulus, but the influ¬
ence of these changes on both the swelling pres¬
sure and the mechanical behavior of the anulus is
CLINICAL FEATURES AND AGING
unclear.68 Age changes in the fibrocartilaginous
Many spinal injuries are associated with disk fail¬ end plates result in irregularly arranged cartilage
ure; for example, up to 30% are directly or indi¬ tissue, which disappears with time and is replaced
rectly related to disk herniation.76,77 In addition, by bone.82
disk ruptures or degenerative changes in nonhemi- All these changes together lead to a biochemical
ated disks are associated with low-back pain and and thus biomechanical compromise to the inter¬
whiplash injuries, which may lead to severe limi¬ vertebral disk, which eventually limits its ability
tation in activity and constitute the most common to perform the required mechanical function. As
cause of sick leave in industrialized countries.78 disks degenerate, they become stiffer and less
Short-term, high-amplitude loads cause irreparable viscoelastic, with associated decreases in both
structural damage to the intervertebral disk when creep and hysteresis.62,83 Consequently, the disks
exhibit a diminished tolerance to normal loading. posterior sides being the strongest parts. Values as
Load tolerance levels have been experimentally high as 1.4 MPa have been reported as the maxi¬
determined for whole-body segments under com¬ mal stresses tolerated by these areas, whereas the
pression, tension, shear, and axial rotation. These central region of the disk can tolerate tensile
values are summarized in Table 2-684 for different stresses up to 0.3 MPa.51 Under normal physio¬
spinal levels. Generally, stress tolerance levels re¬ logic loading conditions, the disk is seldom sub¬
sult from quasi-static, uniaxial tests and therefore jected to pure tensile loads, but local tensile
cannot be extrapolated to more physiologic condi¬ stresses can occur in the anulus fibrosus in various
tions. From Table 2-6, it can be concluded, how¬ conditions. Damaging tensile stresses are associ¬
ever, that the larger lumbar motion segments can ated with the hyperextension, hyperflexion, and
tolerate higher loads than the smaller and more lateral inclination of the spine. Under these condi¬
cranial motion segments.
tions, the disk ruptures near the end plate, and
From the experiments on motion segments, it disk fragments can extrude from the interverte¬
has been generally accepted that pure compressive bral space.
loads do not directly cause disk failure. Virgin58 With regard to axial rotation, the intervertebral
observed that after compressing lumbar disks un¬ disk is more resistant than the vertebral bones.
der loads severe enough to deform the anulus Farfan and colleagues52 tested the torsional charac¬
permanently, there was no evidence of herniation teristics of lumbar motion segments. They were
of nuclear material. He found that disks could unable to produce any radial tears in the disk,
support more than 4400 N. Even when an incision
but circumferential separations of annular lamellae
was made in the postolateral part of the anulus were observed. About 16 degrees of rotation was
and the disk was loaded in compression, there was generally required to produce this type of damage.
still no herniation. Brown and colleagues51 found
For degenerated disks, the angle of failure was
that the first component to fail under axial com¬
somewhat smaller, that is, 14.5 degrees. The aver¬
pression was the vertebra, in particular involving
age failure torque for nondegenerated disks was
a central fracture of the end plates. Under further
25% higher than that for the degenerated disks.
loading, the gelatinous material of the nucleus
When the tolerance for rotational loads was ex¬
pulposus or even parts of the anulus fibrosus can
ceeded, ruptures in the external layer of the
enter the vertebral body.
annulus fibrosus were observed near the end plate.
The degeneratively altered disk shows another
Such tears may be the precursors of dorsolateral
injury mechanism during compression. The dehy¬ extrusion of the intervertebral disk.59
drated nucleus cannot absorb and laterally distrib¬
With shear, the periphery of the disk always
ute any load. Hence, the axial load is transmitted
experiences the greatest stress, and disk damage
by the anulus alone, so that the outer zones of the
is usually seen in this area. It is relatively rare,
end plates are stressed relatively more than the
however, for the disk to fail clinically from pure
central parts. Fractures of the outer zones of the
shear loading. Similarly, disk failure due to pure
end plate, therefore, become more probable with
compressive, tensile, or rotational loading is a
increasing degeneration of the intervertebral disk.
rare phenomenon. Clinical evidence of disk failure
Tension experiments have demonstrated a rela¬
implies that the intervertebral disk most likely
tively high vulnerability of the disk. Under tensile
fails because of some combination of bending,
loading, the intervertebral disk is damaged before
tension, and torsion. For instance, pure compres¬
the adjacent vertebrae suffer any injury. As men¬
sion experiments indicate that some factor other
tioned previously, mechanical properties and
than compression must be involved in the produc¬
hence tensile strength of the intervertebral disk
tion of acute herniation in healthy disks. It has
vary with the location in the disk, the anterior and
been suggested that a combination of bending and
torsional loads is necessary to produce this type Ligaments have a number of gross forms. In
of injury.85 Similarly, rupture of the disk has been the hip and shoulder joints, some ligaments are
associated with coupled movements, namely, local reinforcements of the joint capsule that pri¬
flexion or extension, both with a rotational compo¬ marily act to stabilize the joint. Extraarticular liga¬
nent. ments, such as the medial collateral ligament
Only a few studies have investigated the fatigue (MCL) of the knee, or intraarticular ligaments,
tolerance levels of the intervertebral disk. By such as the anterior cruciate ligament (ACL) in
definition, fatigue tolerance may result from cyclic the knee, on the other hand, are distinct structures
loads of relatively low amplitude, well within that guide movements and restrain the joint from
physiologic limits.37 By subjecting lumbar motion abnormal excursions. In general, extraarticular and
segments to cyclic compressive loads, up to 80% intraarticular ligaments are relatively long, cord¬
of the ultimate failure load, at 0.5 Hz, Liu and like structures, whereas capsular ligaments are
others64 reported end-plate failure after about 1000 broad, bandlike structures, in which different fiber
cycles. From a compression fatigue study,86 it can bundles are loaded during different types of move¬
be concluded that the number of cycles to failure ments. Because of the high incidence of injuries
is inversely related to the applied compressive to the ligaments in the knee, such as the ACL or
cycle stress raised to a power of 14. Thus, even a MCL, these structures have been studied inten¬
small increase in stress considerably decreases the sively and have been described widely in the liter¬
number of cycles needed to produce disk failure. ature.
As a result of the lack of adequate capillary
perfusion, the intervertebral disk has a limited STRUCTURE AND COMPOSITION
capacity for regeneration, even after extensive
therapy and other clinical interventions. Extensive Most ligaments consist of nearly parallel-fibered
studies are being performed to develop interverte¬ bundles of collagen. There are two exceptions: the
bral disk prostheses as a replacement for nonheal¬ ligamentum nuchae and the ligamenta flava along
ing and irreparable disks. So far, these studies the spinal column, which contain large amounts
have focused on fully artificial disks, mainly con¬ (almost 75% of the dry weight) of elastic fibers.
sisting of fiber-reinforced composites,87 and on the In normal ligaments, however, collagen predomi¬
development of cell-seeded structures. The latter nates and constitutes between 70% and 80% of
alternative offers great possibilities considering the dry weight.88 Normal ligaments contain more
biocompatibility, whereas the former can be more than 90% type I collagen with less than 10% type
easily adapted to the required mechanical proper¬ III collagen. Type IV collagen is present in smaller
ties. However, complete experimental validation quantities. Other constituents of ligament tissue
must be performed before these alternatives be¬ are fibroblasts (3% to 6%), proteoglycans, water,
come available for application in the clinical field. and elastin. The noncellular composition of liga¬
ments is included in Table 2-7.89
Both collagen and elastic fibers influence the
Ligaments mechanical function of ligaments. The structure
and mechanical behavior of collagen fibers was
GENERAL FEATURES
described previously. Contrary to collagen fibers,
Skeletal ligaments are passive bands of fibrous elastic fibers can be easily stretched to a consider¬
tissue that connect the bony parts of articulating able length—up to 200% of their unloaded
joints at two ligament-bone junctions. There is length—at relatively low loads. At higher loads,
minimal stress concentration at the insertions of they rupture suddenly. Elastic fibers are of great
the ligament into the stiffer bone, owing to the importance to the recovery of tissues after loading.
presence of functionally graded materials, involv¬
ing fibrocartilage with varying degrees of mineral¬
ization. The ligaments are constructed to resist TABLE 2-7 Proportion of Noncellular Components
tensile forces without considerable elongation. in a Normal Ligament
Hence, they are composed of densely packed fiber Wet Dry
bundles, which run mostly parallel along the lon¬ Weight Weight
Composite (%) (%)
gitudinal axis of the tissue. The primary functions
of the passive ligaments are to stabilize the joints Collagen 25-32 70-80
and to guide and restrict articular movements. In Elastin 1-4 2-9
addition, they function as a signal source for re¬ Proteoglycans 0.3 1
flexes and provide proprioceptive information Water 60-65
about movements and postures.
Their function in recovery, however, reduces to¬ alters the ultrastructure, chemistry, and physical
ward maximal loading levels because their maxi¬ properties of the ligament. In this way, ligaments
mum strength is about five times lower than that are capable of hypertrophy, atrophy, and healing
of collagen.90 responses. Carefully conducted training protocols
Elastic fibers are composed mainly of elastin, a may increase tensile strength and stiffness of liga¬
hydrophobic, nonglycosylated protein composed ments,93 whereas controlled, functional stresses
of about 830 amino acid residues. Elastin mole¬ applied to a damaged ligament stimulate the heal¬
cules are secreted by the fibroblasts in the ECM, ing process by increasing fibril diameters.94 During
where they form filaments and sheets in which the immobilization or disuse, on the other hand, liga¬
elastin molecules are highly cross-linked to one ment structure deteriorates, and mechanical prop¬
another to generate an extensive network. The erties are drastically decreased.95 Although this
cross-links are formed between lysine residues by effect is not irreversible, it may be present for a
the same mechanism that operates in cross-linking considerable period of time. In primates immobi¬
collagen molecules. Elastin molecules are unlike lized for 8 weeks, the ultimate tensile strength of
most other proteins in that their function requires the ACL fell to 60% of the original strength. After
their polypeptide backbone to remain unfolded as 5 months of recovery, only 80% of the original
“random coils.” It is the cross-linked, random- strength was restored, and 12 months of recovery
coil structure of the elastic fiber network that were needed to regain 90% of the original
allows the network to stretch and recoil. Sheaths strength.96
constituted by type I, III, and IV collagens sur¬
round elastic and collagen fibers, fiber bundles, BIOMECHANICAL CONSIDERATIONS
and whole ligaments. The individual fibers attach
either to the periosteum or directly to the bone, Like other soft connective tissue, ligaments dis¬
where they join the collagen of the bone in Shar- play time and loading history-dependent visco¬
pey’s fibers. elastic behavior.97,98 These mechanical characteris¬
The structural organization of fiber bundles in tics are influenced by the structural organization of
ligaments is less well organized than in tendinous the fiber bundles; the relative amount of collagen,
tissue. In most cordlike ligaments, nearly all fibers elastin, and proteoglycans; and the interactions
run in parallel. Some fibers, however, run criss¬ between tissue constituents.
cross through the ligament (Fig. 2-12). Therefore, The primary role of ligaments is to resist tensile
not all fiber bundles are stretched when loaded loading; hence, their mechanical properties are
along the main fiber axis. Consequently, ligaments commonly measured in tensile loading experi¬
are less strong than tendons. In broad bandlike ments. Usually, isolated cadaveric bone-ligament-
ligaments, different portions of the structure have bone complexes of human or animal origin, such
different fiber directions, with the main fiber direc¬ as the rabbit femur-MCL-tibia complex, are used
tion dependent on the function and working range for this purpose. The ligament is tested with the
of the particular portion. In the cruciate ligaments bones affixed to clamps to produce uniform stress
of the knee, fiber organization is even more com¬ distributions across the ligament. Load can be
plex. The fibers are twisted between 90 and 180 measured, for instance, with load cells99 or omega¬
degrees from the femoral to the tibial attachment shaped strain gauges,100 whereas elongation can be
sites, and as a result, different portions of the determined using displacement transducers101 or
cruciate ligaments function at different rotation optical techniques, such as the video dimensional
and knee flexation angles.91,92 analyzer system.102 The resulting load-elongation
As with other tissues in the human body, liga¬ relations and parameters, such as stiffness, ulti¬
ments exhibit remodeling behavior. The metabo¬ mate load, ultimate elongation, and energy ab¬
lism of the fibroblasts within the tissue responds sorbed to failure, are representative of the whole
to environmental stimuli and functional needs and bone-ligament-bone complex. Geometric cross-
-V-/--:\ , > ^
-■
S \
^ V ^ itr \
J
!
'l
-x.— -- r ^ t-
- ^ r £==□
Tendon Ligament
FIGURE 2-12 Structural organization of fibers in a normal tendon compared with that of fibers in a
ligament.
~70
FIGURE 2-13 Typical stress-strain curves of ligaments loaded to failure. A, Ligament with normal
composition. B, Ligament with high concentration of elastin.
sectional data for the ligament tissue and strain data from a range of studies that have focused
data are required to obtain a stress-strain curve of on stress-strain properties of ligament tissue as
the ligament alone, which can be used to calculate, determined from the uniaxial tension tests on
for instance, modulus, ultimate tensile strength, bone-ligament-bone complexes.107-114 As can be
and ultimate strain of the ligament tissue.103 Indi¬ observed, the reported mechanical properties of
rect methods using a mass measure divided by ligaments have been widely divergent. For exam¬
specimen length, as well as direct methods such ple, ultimate strains at failure range from 5% to
as laser micrometer systems, have been employed more than 40%. One study reports strain to failure
to determine the cross-sectional shape and area of as high as 93%.115 As with the other soft tissues,
a ligament.104'105 In the laser micrometer method, the variability of data can be attributed to speci¬
the widths of the tissue are measured as the speci¬ men considerations such as age, species, and ana¬
men is rotated about its longitudinal axis. tomic location, and to experimental factors, such
A typical stress-strain curve for most ligaments as testing method, loading history, and accuracy
in uniaxial tension is shown in Figure 2-13A. This of applied techniques.
curve starts with a nonlinear toe region (region I), Mechanical properties of ligaments are gener¬
where the resulting stress per unit strain gradually ally determined in vitro because the measurement
increases. The toe region is believed to reflect of loads, ligament length, and cross-section in
uncrimping of collagen fibers and is followed by vivo is extremely difficult. Tipton and col¬
a rather linear segment of the curve (region II) leagues116 determined the uniaxial load-elongation
when the fiber bundles are optimally aligned in properties of the femur-MCL-tibia complex of
parallel and the collagen triple helix is stretched. anaesthetized rats in vivo and compared the results
Toward the end of the linear segment, interfibrillar with those obtained in vitro. Linear stiffness and
slippage occurs, and individual fibers start to rup¬ ultimate load at failure for the in vivo specimens
ture.106 Finally, the curve begins to bend toward were 6% and 4% lower, respectively, whereas the
the strain axis (region III) and reaches a point of ultimate elongation was 12% higher. The major
ultimate stress; the stress then decays while more drawback of these in vivo experiments116 is that
and more fibers rupture.96 The stress-strain curve during uniaxial tensile testing of the bone-
of ligaments with a high concentration of elastin, ligament-bone complexes, nearly all fibers along
like the ligamenta flava, is entirely different and the loading axis are stretched simultaneously. This
resembles the stress-strain curve of elastin (Fig. situation is very unlike the normal physiologic
2-13B). Here, considerable elongation is possible situation and may provide a false impression of
before stiffness increases. At this point, the stiff¬ the functional properties of ligament. Thus, care
ness increases sharply, and the ligament ruptures should be taken when interpreting the properties
suddenly thereafter.90 listed in Table 2-8.
Numerous studies have been performed to de¬
termine the biomechanical properties of various Modeling
ligaments and to assess the effects of several phys¬
iologic factors, including aging and exercise, on Various mathematical and numerical models have
the biomechanical properties. Table 2-8 provides been developed to study mechanical properties of
TABLE 2-8 Tensile Properties of a Range of Ligaments From Various Sources
Strain
Ultimate Ultimate Modulus* Rate Reference
Ligament and Source Stress (MPa) Strain (%) (MPa) (%/S) No.
Monkey ACL 66.1 ± 8.4 47 ± 5.6 186 ± 26 66 115

Human young adult ACL 66.1 ± 8.4 44.3 ± 8.5 III ±26 100 115
Human lateral collateral ligament 32-44 10-17 360—410 100 116
Human ACL 30-40 14-16 280-330 100 116
Human PCL 35—43 14-18 280-440 100 1 16
Human lateral collateral ankle 24.2-46.2 13-17 217-512 0.13 117
Human medial collateral ankle 15.6-34 10-33 100-321 0.13 117
Sheep ACL 60-124 45-93 180—437 (8.3 mm/s) 107
Rabbit mature MCL 75.2 ± 2.4 9.5 ± 0.5 700 ± 70 0.01 1 18
106.7 ± 6.8 13.0 ± 1.0 760 ± 160 222
Human wrist joint ligaments — — 22.6-119.3 66 119
Rabbit ACL 42 8 711 0.016 120
50 10 674 1.3
68 9 930 381
Human older PCL 24-36 18.0-19.5 145-248 50 121
Human coracoacromial ligament 25.3 ± 8.7 5.1 ± 1.2 658.4 ± 261.3 0.1 122
The modulus was generally estimated from the linear region of the stress-strain curve.
ACL, anterior cruciate ligament; MCL, medial collateral ligament; PCL, posterior cruciate ligament.
ligaments. For example, Fung117-118 used advanced the acute ligament ruptures in athletics, pitching,
theories of viscoelasticity to characterize time- and racquet sports, and swimming being responsi¬
history-dependent properties of a range of soft ble for most of the overload injuries. Tensile fail¬
tissues. Numerical models have also been devel¬ ure of a ligament, resulting either in acute rupture
oped, using combinations of ideal elastic, viscous, or in microtears, is a result of the weakest link in
and plastic elements in parallel and in series with the bone-ligament-bone complex. It can occur at
each other.119 Three-dimensional, computer-simu¬ the insertion sites, which is known as avulsion;
lated models of whole joints (e.g., the knee joint) within the ligament itself (substance tear); or as a
permit the analysis of the relations between various combination of both. Some investigators reason
ligament properties and ligament function.120-121 In that the failure mode is strain rate dependent in
addition, computer-simulated conditions may pro¬ that avulsion failures result primarily from slow
vide clinically relevant data that cannot be ob¬ extension rates, whereas ligament substance tears
tained otherwise. For example, application of result from high extension rates.107- 124 This sug¬
computer models to ACL replacement can help gests that with increasing strain rate, the increase
determine the ideal anatomic insertion points for in tissue strength is higher at the bone—ligament
reconstruction.122 Experimental validation of junction than in the ligament itself. Other investi¬
model predictions, however, is difficult owing to gators have shown that failure modes are indepen¬
lack of realistic in vivo data. dent of strain rate and merely depend on the
maturation of the loaded bone—ligament—bone
CLINICAL FEATURES OF AGING complex. In a study of rabbits grouped by age
ranging from 1.5 to 15 months, Woo and associ¬
The most common injuries to the ligaments are ates125 showed that the ligament substance matured
acute ruptures caused by high strain rate supra¬ earlier than the ligament-bone junction. There¬
maximal tensile loads. For example, ACL and fore, in the immature animals, ligaments failed by
MCL ruptures account for as much as 77% of all avulsion, whereas in the mature animals, liga¬
acute injuries to the knee.123 Other ligament inju¬ ments failed by substance tear.126 Ligament me¬
ries are overload injuries due to repetitive submax - chanical properties change considerably with mat¬
imal loading. These injuries result in microtears, uration and aging. Stiffness, strength, and energy
followed by inflammation reactions and some¬ required to rupture the tissue peak during late
times calcification of the ligament tissue. Both maturity and then decline.127-128 In the ACL, for
types of injuries frequently occur as a conse¬ example, stiffness and ultimate load of bone-
quence of sports activities, with American foot¬ ligament-bone specimens obtained from subjects
ball, soccer, and skiing accounting for many of at old age are two to three times lower than
ACL specimens obtained from young adults.107,129 ment may lead to a new generation of ligament
Similar to mature bone-ligament-bone com¬ replacements with optimized morphology and
plexes, aged bone-ligament-bone complexes fail properties.
by ligament substance tear.
Tendons
Ligament Healing and Repair
GENERAL FEATURES
It has been recognized that extraarticular liga¬
ments, such as the MCL, can generally heal effec¬ Tendons, like ligaments, are highly fibrous regular
tively when injured,130, 131 whereas intraarticular connective tissues, in which the fibers are regu¬
ligaments, such as the ACL, do not heal effec¬ larly orientated with respect to one another to
tively.132 Ruptured ACL ligaments do not heal form thicker bundles. Tendons connect muscles to
because the tom ends do not meet and new tissue structural elements such as bone, adopting the
does not fill the gap,133 even if the ends are apposi- form of cords or straps, and are round, oval, or
tioned by sutures or wires.134 Therefore, surgical elongated in cross-section. Tendons are highly re¬
repair of the ACL usually aims at reconstruction sistant to extension but are relatively flexible and
of the ruptured ligament by grafts. Most com¬ can therefore be angulated around bone surfaces
monly used are autografts, such as the patellar or deflected beneath retinacula to alter the direc¬
tendon or semitendinosus tendon, but in recent tion of the muscular pull. Tendons consist of fasci¬
years, allografts or synthetic polymeric materials cles of collagen fibers largely running parallel to
have also been used.89 In the United States, about the long axis of the tendon. Their surface is gener¬
50,000 ligament reconstructions are performed ally smooth, although longitudinal ridging due to
each year, most of which are performed to recon¬ the arrangement of coarse fasciculi is common in
struct a ruptured ACL.135 The aim of recon¬ larger tendons. Bundles of collagen fibrils, each
structive surgery is to restore normal passive func¬ 0.02 to 0.20 |xm in diameter, assemble into fibers
tion of the ligaments and the laxity characteristics 1 to 20 |xm in diameter. These fibers bundle into
of the knee joint. Whether a reconstruction is the fascicles, within which the fibers are more or
successful depends on the mechanical properties less longitudinally orientated. Within the fascicles,
of the graft, its pretension,136 and its position in the fibers are bound by a small amount of an
the knee.91,136 amorphous mucoprotein cement. It is well estab¬
Functional and mechanical tests on recon¬ lished that the fibers demonstrate a planar crimp¬
structed ligaments indicate' that grafts that mimic ing, of wavelength between 100 and 300 |xm.
closely the stmcture and mechanics of the natural The areolar connective tissue that permeates the
knee ligament have a higher chance of success. tendon between its fascicles provides a route for
For example, the use of two distinct segments vessels and nerves. The blood supply of tendons
with separate insertion sites and pretension, unlike is provided by a relatively sparse array of small
a graft consisting of one individual bundle, gives arterioles, which run longitudinally from the adja¬
more stability during flexion of the knee.137 Ex¬ cent muscular tissues. These arterioles ramify in
tensive evaluations of reconstructed ACLs, how¬ the interfascicular spaces, where they intercommu¬
ever, show that the rate of success is less than nicate freely and are accompanied by veins. The
perfect.138, 139 Generally, there is a marked, pro¬ longitudinal network in the tendon is augmented
longed postoperative weakening of ligament re¬ by small vessels from the surrounding areolar con¬
placements. After 1 year, autografts have less than nective tissue or, where present, the synovial
80% of the tensile strength and 50% of the ulti¬ sheets. Because vascular networks are of low den¬
mate load that they had before transfer.140 This sity, tendons appear white.
weakening is caused partly by the inability to The nerve supply of tendons appears to be
reproduce in the graft the stresses and strains of mainly afferent, with no clear evidence of vaso¬
the collagen fibers found in the original ligament. motor control. Specialized afferent receptors, neu-
Moreover, ligament replacements may not achieve rotendinous endings known as Golgi tendon or¬
normal strength and normal morphology, includ¬ gans, exist particularly near the tendinomuscular
ing collagen fiber diameter, because of the abnor¬ junction. Each is about 500 |xm long and 100 p.m
mal loading states in those replacements.122 As in diameter and consists of a small bundle of
with other connective tissues, tissue-engineered tendon fibers enclosed in a delicate capsule. When
equivalents of ligaments are being developed for tendons are stretched, these neurotendinous end¬
repair and healing of irreversibly damaged struc¬ ings activate and initiate myotactic reflexes, which
tures or as a model to study environmental factors inhibit the development of excessive tension dur¬
on ligament healing.141,142 With time, this develop¬ ing muscular contraction.
STRUCTURE AND COMPOSITION cific muscle-tendon unit produces a cyclic force

with a constant maximal value, which is propor¬
Tendons may be considered unidirectional fiber- tional to the external load. Because of muscle
reinforced composites, as indicated in Figure 2- tone, a small amount of tension is always present,
12. The building block of the fibrous phase con¬ ensuring that the muscle-tendon units are taut
sists of tropocollagen molecules, which reinforce even when the muscle is relaxed. Moreover, any
a matrix made up of a hydrated proteoglycan gel. possible slackening due to creep of the tendon
Collagen constitutes at least 30% of the wet would be eliminated by a reduction in the length
weight of tendons. The collagen fibers may be of the muscle. Therefore, the in vivo repetitive
considered largely independent, although stress loading pattern of tendons in the lower limbs may
can be transferred from one fiber to its neighbor be broadly classified as a tension-tension square
by shear in the matrix or by more specific cross¬ wave, as observed in the in vivo tensile load
links. The tendons are surrounded by some loose pattern in the Achilles tendon during various
connective tissue, the epitendon, which facilitates forms of locomotion.145
gliding of the tendon over extratendinous tissues
as well as through synovial sheaths at the digits.
Biomechanical Studies
The epitendon contains tendon cells that are physi¬
cally joined to tendon fibroblasts inside the ten¬ The load-bearing function of functional collage¬
don. Both cell types are subjected to a combina¬ nous tissues prompted a plethora of biomechanical
tion of shear stresses during gliding as well as studies over several decades. The uniaxial tensile
tension as a result of muscle contraction. In gen¬ testing of animal and human tendons has been
eral, the cellular material occupies about 20% of performed over many decades.147-149 In vitro test¬
the tissue volume, whereas the ECM accounts for ing is fraught with practical difficulties associated
the remaining 80%. with specimen preparation, gripping, and local
strain measurements. When these are overcome,
BIOMECHANICAL CONSIDERATIONS however, the resulting stress-strain curves for ten¬
dons are nonlinear, with a characteristic toe-in
Load Carriage region followed by a linear region before failure.
This is very similar to the form for ligaments (see
Tendons act as linking elements connecting mus¬ Fig. 2-13A), which can be related to the structural
cles to bones, thus transmitting muscular pulls and aspects of the tissue, with the toe-in region at
external loads. In slow concentric activities, in low stress levels producing a straightening of the
which the inertial effects may be neglected, the crimped collagen fibers with an associated in¬
maximal load to which the tendon may be sub¬ crease of the tangent modulus. A quasi-static test
jected is the muscle isometric tetanic contraction, produces an ultimate tensile strength (UTS) of
about equal to 0.35 N/mm2 of muscle belly.143 As about 100 MPa and a failure strain of about 15%.
an example, the isometric stress in the patellar The material properties are clearly dependent on
tendon can be estimated at 29 MPa, based on the the rate of tensile testing, with values for elastic
cross-sectional areas of 125 mm2 and 10,300 mm2 modulus quoted within the range of 1 to 2
for the tendon and the quadriceps femoris muscle, giga-pascals (GPa). More recently, dynamic char¬
respectively. Most daily activities, however, con¬ acterization of human tendons has been
sist of quick eccentric movements, in which iner¬ performed.150'151 These tests emphasize the visco¬
tial effects play an important role. Experimental elastic nature of these collagenous tissues.
studies have indicated that these dynamic activi¬ A recent study addressed the in vitro fatigue
ties are associated with tendon stresses in the behavior of human tendons.152’153 A linear model
range of 42 to 110 MPa, the latter exceeding the based on the median value, equivalent to a 50%
established values of the ultimate tensile probability of failure (or survival), was employed
strength.144-145 In most cases, however, human ten¬ as illustrated in Figure 2-14. The relation of stress
dons are so thick that there is no possibility of level against the median fatigue life produced a
rupture by a single application of maximum load. statistically significant linear model of the follow¬
Healthy people are estimated to walk about 1 ing form:
to 1.5 million strides per year.21'146 During loco¬
motion, the musculoskeletal system is subjected S = 101.25 - 14.83 log (N)
to a constant external load, the body weight.
Movement is achieved by muscular contraction where S is the normalized stress expressed as a
and the consequent production of torque around percentage of the UTS, and N is the number of
the joints of the lower limbs. Evidently, any spe¬ cycles to failure.
mated to be about 30% to 50% of all sports-

related injuries in the United States.155 Tendon
injuries are common examples of such injuries
because during physical activity, much of the force
is focused on the tendon component of the mus¬
cle-tendon-bone unit. Thus, rapid elongation of
this unit with maximum uncoordinated muscle
contraction may rupture healthy tendons.
Adult tendons have a low metabolic rate and
vascularity. Increasing intratendinous tension
causes a decrease in blood perfusion. As a result
of persistent hypoxia, parts of the tendon may
undergo fibrocartilaginous metaplasia.156 In these
cases, chondrocytes may mediate calcium deposi¬
tion, which may proceed to the formation of dis¬
crete nodules within the tendons. This alters the
biomechanical properties of the whole tendon.
Log number of cycles Tendon Healing and Repair

FIGURE 2-14 Fatigue life of human tendons. Normalized
stress versus logarithm of median number of cycles to failure. Tendons, like ligaments, remodel in response to
UTS, ultimate tensile strength. (From Schechtman H: the mechanical demands placed on them.157 This
Mechanical Characterisation of Fatigue Failure in Human manifests as an increase in strength and stiffness
Tendons. PhD Thesis. London, University of London, 1995.) when subjected to enhanced activity levels such
as occur during physical training. By contrast,
reducing the stress below normal levels, such as
by stress shielding or immobilization, produces a
This model predicted a static strength of 101.3
decrease in strength and stiffness of the tendons.
MPa, which was clearly within 1 standard devia¬
It is well established that tendons are not well
tion of the experimental data obtained in the qua¬
vascularized or well innervated; therefore, the
sistatic tensile tests. The linear model also sug¬
healing response is limited. Much effort has been
gested the absence of an endurance limit. Some
directed toward understanding the mechanisms as¬
analysis that takes into account in vivo healing
sociated with the stimulation of tendon healing.
was proposed. This analysis was able to explain
One such study has investigated the potential for
the presence of intact tendons throughout the life¬
mechanical stimulation in conjunction with
time of a patient. The results implied a finite
growth factors to stimulate tendon cell activity in
fatigue behavior with no apparent endurance limit.
vitro. Banes and colleagues158 applied loading by
To explain the presence of intact tendons through¬
means of a commercial strain unit (Flexercell,
out the lifetime of a patient, the authors postulated
Flexcell International, McKeesport, PA), in which
the importance of remodeling processes in vivo.
cells were attached to a collagen substrate at the
base of culture plates. Negative pressure was ap¬
CLINICAL FEATURES AND AGING plied in a cyclic manner to provide deflection of
the base of the plate. Results highlighted the com¬
During aging, changes occur in the composition plex mechanisms involved in the mechanical con¬
of tendon. In particular, the collagen content de¬ trol of cell division and maintenance of matrix in
creases with age, with an associated decrease in tissue required to resist high tensile strains.
the mechanical strength, stiffness, and ability to
withstand deformation.
Tissue trauma or injury can result from two Conclusions
basic mechanisms. A single-impact macrotrauma,
such as a blow to a leg or a twisting injury of a Certain features are common to the discussion of
joint, injures bone, muscle, tendon, ligament, and the biomechanics of all types of soft tissues:
even neurovascular elements. The other mecha¬
nism of injury is repetitive microtrauma caused by • The complex microstructure, which varies with
the repeated exposure of tissue to low-magnitude site, age, and so forth, and the lack of standard
force, which itself does not result in tissue in¬ test methods have led to the diversity of
jury.154 The frequency of overuse injuries is esti¬ biomechanical data in the literature. To date,
there has been little consensus regarding a set cartilage. In Freeman (ed): Adult Articular
of material constants that can uniquely Cartilage. London, Pitman, 1979, p 281.
characterize the properties of healthy soft 14. Bader DL: The relationship between the
tissues. This presents a major obstacle in mechanical properties and the structure of adult
human articular cartilage. PhD Thesis.
producing a design template for replacement
Southampton, UK: University of Southampton,
tissues.
1985.
• Biomechanical changes are apparent in soft
15. Athanasiou KA, Rosenwasser MP, Buckwalter
tissues as a result of aging and disease. In JA, et al: Interspecies comparisons of in situ
addition, deviation from normal biomechanical intrinsic mechanical properties of distal femoral
loading, such as with joint immobilization, cartilage. J Orthop Res 9:330, 1991.
affects the structure and biomechanical 16. Woo SL-Y, Simon BR, Kuei SC, Akeson WH:
integrity of the tissues. Quasi-linear viscoelastic properties of normal
• Synthetic options for the repair of soft tissues articular cartilage. J Biomech Eng 102:85, 1980.
have limited success. More recent options, 17. Coletti JM, Akeson W, Woo SL-Y: A comparison
incorporating cell-engineering principles, might of the physical behaviour of normal articular
be effective in the future in creating tissue cartilage and the arthroplasty surface. J Bone
Joint Surg 54(A): 147, 1972.
implants with mechanical integrity.
18. Bader DL, Kempson GE: The short-term
compressive properties of adult human articular
References cartilage. Biomed Mater Eng 4:245, 1994.
19. Bader DL, Kempson GE, Barrett AJ, Webb W:
1. Materials Technology Foresight in Biomaterials. The effects of leukocyte elastase on the
London, Institute of Materials, 1995. mechanical properties of adult human cartilage in
2. Maroudas A, Stockwell RA, Nachemson A, tension. Biochim Biophys Acta 677:103, 1981.
Urban JP: Factors involved in the nutrition of the 20. Weightman B: Tensile fatigue of human articular
human lumbar intervertebral disc: Cellularity and cartilage. J Biomech 9:192, 1976.
diffusion of glucose in vitro. J Anat 120:113, 21. Weightman B, Chappell DJ, Jenkins EA: A
1975. second study of tensile fatigue properties of
3. Mayne R: Cartilage collagens. Arthritis Rheum human articular cartilage. Ann Rheum Dis 37:58,
32(3):241, 1989. 1978.
4. Weiss C, Rosenberg L, Helfet AJ: An 22. Wachtel E, Maroudas A, Schneiderman R: Age-
ultrastructure study of normal young adult human related changes in collagen packing of human
articular cartilage. J Bone Joint Surg 50A:663, articular cartilage. Biochim Biophys Acta 1243,
1968. 239, 1995.
5. Balazs EA, Bloom GD, Swann DA: Fine 23. Lyrra T, Jurvelin J, Pitkanen P, et al: Indentation
structure and glycosaminoglycan content of the instrument for the measurement of cartilage
surface layer of articular cartilage. Fed Proc Fed stiffness under arthroscopic control. Med Eng
Am Soc Exp Biol 25:1813, 1966. Phys 17:395, 1995.
6. Broom N: The altered biomechanical state of 24. Lee DA, Bader DL: Compressive strains at
human femoral osteoarthritic articular cartilage. physiological frequencies influence the
Arthritis Rheum 27:1028, 1984. metabolism of chondrocytes seeded in agarose. J
7. McCall JG: Scanning electron microscopy of Orthop Res 15(2): 181, 1997.
articular surface. Lancet 2:1194, 1968. 25. Brittberg M, Lindahl A, Nilsson A, et al: The
8. Bullough P, Goodfellow JG: The significance of treatment of deep cartilage defects in the knee
the fine structure of articular cartilage. J Bone with autologous chondrocyte transplantation. N
Joint Surg 50B:852, 1968. Engl J Med 331:889, 1994.
9. Knight MM: Deformation of isolated articular 26. Kim WS, Vacanti JP, Cima LG, et al: Cartilage
chondrocytes cultured in agarose constructs. PhD engineered in predetermined shapes employing
Thesis. London, University of London, 1997. cell transplantation on synthetic biodegradable
10. Maroudas A: Physicochemical properties of polymers. Plast Reconstr Surg 94(2):233, 1994.
articular cartilage. In Freeman MAR (ed): Adult 27. Paige KT, Cima LG, Yaremchuk MJ, et al: De
Articular Cartilage. London, Pitman, 1979, novo cartilage generation using calcium alginate-
p 215. chondrocyte constructs. Plast Reconstr Surg
11. Mak AF, Lai WM, Mow VC: Biphasic 97(1): 168, 1996.
indentation of articular cartilage. I Theoretical 28. Andersson GBJ, Ortengren R, Nachemson A,
solution. J Biomech 20:703, 1987. Elfstrom G: Lumbar disc pressure and
12. Armstrong CG, Mow VC: Variations in the myoelectric back muscle activity during sitting.
intrinsic mechanical properties of human articular I. Studies on an experimental chair. Scand J
cartilage with age, degeneration and water Rehabil Med 3:104, 1974.
content. J Bone Joint Surg 64(A):88, 1982. 29. Nachemson A, Morris JM: In vivo measurement
13. Kempson GE: Mechanical properties of articular of intradiscal pressure: Discometry, a method for
I Chapter 2 Biomechanics of Soft Tissues 61
the determination of pressure in the lower lumbar GBJ: Load-displacement properties of lower
discs. J Bone Joint Surg 46A:1077, 1964. cervical spine motion segments. J Biomech
30. Hukins DWL: Disc structure and function. In 21:769, 1988.
Ghosh P (ed): The Biology of the Intervertebral 48. Goel VK, Goyal S, Clark CR, et al: Kinematics
Disc, vol 1. Boca Raton, FL, CRC Press, 1988, of the whole lumbar spine: Effect of discectomy.
P 1. Spine 10:543, 1985.
31. Eyre DR: Biochemistry of the intervertebral disc. 49. Goel VK, Clark CR, Harris KG, Schulte KR:
Int Rev Connect Tissue Res 8:227, 1979. Kinematics of the cervical spine: Effect of
32. Gower WE, Pedrini V: Age related variations in multiple total laminectomy and facet wiring. J
protein-polysaccharide from human nucleus Orthop Res 6:611, 1988.
pulposus, annulus fibrosus, and costal cartilage. J 50. Berkson MH, Nachemson A, Schultz AB:
Bone Joint Surg 51 A: 1154, 1969. Mechanical properties of human lumbar spine
33. Lyons G, Eisenstein SM, Sweet MB: motion segments. 2. Responses in compression
Biochemical changes in intervertebral disc and shear, influence of gross morphology. J
degeneration. Biochim Biophys Acta 673:443, Biomech Eng 101:53, 1979.
1981. 51. Brown T, Hanson R, Yorra A: Some mechanical
34. Roberts S, Menage J, Urban JP: Biochemical and tests on the lumbo-sacral spine with particular
structural properties of the cartilage end plate reference to the intervertebral disc. J Bone Joint
and its relation to the intervertebral disc. Spine Surg 39A:1135, 1957.
14:166, 1989. 52. Farfan HF, Cossette JW, Robertson GH, et al:
35. Ghosh P, Bushell GR, Taylor TKF, Akeson WK: The effects of torsion on the lumbar
Collagens, elastin and non-collagenous protein of intervertebral joint: The role of torsion in the
the intervertebral disc. Clin Orthop Relat Res production of disc degeneration. J Bone Joint
129:124, 1977. Surg 52A:468, 1970.
36. Marchand F, Ahmed AM: Investigation of the 53. Goel VK, Clark CK, McGowan D, Goyal S: An
laminate structure of lumbar disc annulus in-vitro study of the kinematics of the normal,
fibrosus. Spine 15:402, 1990. injured and stabilized cervical spine. J Biomech
37. Panjabi MM, White AA: Physical properties and 17:363, 1984.
functional biomechanics of the spine. In White 54. Hirsch C, Nachemson A: A new observation on
AA, Panjabi MM (eds): Clinical Biomechanics of the mechanical behavior of lumbar disc. Acta
the Spine. Philadelphia, JB Lippincott 1990, p 3. Orthop Scand 23:254, 1954.
38. Eyre DR: Collagens of the disc. In Ghosh P (ed): 55. Miller JAA, Schultz AB, Andersson GBJ: Load-
The Biology of the Intervertebral Disc, vol 1. displacement behaviour of sacroiliacal joints. J
Boca Raton, FL, CRC Press, 1988, p 171. Orthop Res 5:92, 1987.
39. Best BA, Guilak F, Setton LA, et al: 56. Shea M, Edwards WT, White AA., Hayes WC:
Compressive mechanical properties of the human Variations in stiffness and strength along the
annulus fibrosus and their relationship to human cervical spine. J Biomech 24:95, 1991.
biochemical composition. Spine 19:212, 1994. 57. Tencer A, Ahmed A, Burke D: Some static
40. Johnstone B, Urban JPG, Roberts J, Menage J: mechanical properties of the lumbar
The fluid content of the human intervertebral intervertebral joint, intact and injured. J Biomech
disc. Spine 17:412, 1992. Eng 104:193, 1982.
41. Junghanns H: Die functionelle Pathologie der 58. Virgin W: Experimental investigations into
Zwischenwirbelscheiben als Grundlage fur physical properties of intervertebral disc. J Bone
klinische Betrachtungen. Arch Klin Chir Joint Surg 33B:607, 1951.
267:393, 1951. 59. Junghanns H: Clinical Implications of Normal
42. Galante JO: Tensile properties of the human Biomechanical Stresses on Spinal Function.
lumbar annulus fibrosus. Acta Orthop Scand Aspen, CO, 1990.
Suppl 100:1, 1967. 60. McNally DS, Adams MA: Internal intervertebral
43. Nachemson A: Lumbar intradiscal pressure. Acta disc mechanics as revealed by stress
Orthop Scand Suppl 43:1, 1960. profilometry. Spine 17:66, 1992.
44. De Jager M: Mathematical modelling of the 61. Adams MA, Hutton WC: The effect of posture
human cervical spine: A survey of the literature. on the fluid content of intervertebral discs. Spine
In Proceedings of the International Conference 8:665, 1983.
on the Biomechanics of Impacts (IRCOBI). 62. Broberg KB: Slow deformation of intervertebral
1993,p 213. discs. J Biomech 26:501, 1993.
45. Panjabi MM, Summers DJ, Pelker RR, et al: 63. Adams MA, Dolan P, Hutton WC, Porter RW:
Three-dimensional load-displacement curves due Diurnal changes in spinal mechanics and their
to forces on the cervical spine. J Orthop Res clinical significance. J Bone Joint Surg 72:266,
4:152, 1986. 1990.
46. Panjabi MM, Lydon C, Vasavada A, et al: On the 64. Liu YK, Njus G, Buckwalter J, Wakano K:
understanding of clinical instability. Spine Fatigue response of lumbar intervertebral joints
19:2642, 1994. under axial cyclic loading. Spine 6:857, 1983.
47. Moroney SP, Schultz AB, Miller JAA, Andersson 65. Goel VK, Voo L-M, Weinstein JN, et al:
Response of the ligamentous lumbar spine to 83. Markoff KL: Stiffness and damping
cyclic bending loads. Spine 13:294, 1988. characteristics of the thoracic lumbar spine. In
66. Goel VK, Weinstein JN: Time-dependent Proceedings of the NIH Workshop on
biomechanical response of the spine. In Goel Bioengineering Approaches to the Problems of
VK, Weinstein JN (eds): Biomechanics of the the Spine, 1970.
Spine: Clinical and Surgical Perspective. Boca 84. Sonoda T: Studies on the strength for
Raton, FL, CRC Press, 1990. compression, tension and torsion of the human
67. Drost MR, Willems P, Snijders H, et al: Confined vertebral column. J Kyoto Prof Med 71:659,
compression of canine annulus fibrosus under 1962.
chemical and mechanical loading. J Biomech 85. Farfan HF: Hypothesis of degenerative process.
Eng 117:390, 1995. In Farfan HF (ed): Mechanical Disorders of the
68. Ebara A, Iatridis JC, Setton LA, et al: Tensile Low Back. Philadelphia, Lea & Febiger, 1973, p
properties of nondegenerate human lumbar 201.
annulus fibrosus. Spine 21:452, 1996. 86. Hansson TH, Keller TS, Spengler DM:
69. Setton LA, Zhu WB, Mow VC: Compressive Mechanical behaviour of the lumbar spine. II.
properties of the cartilage endplate of the baboon Fatigue strength during dynamic compressive
lumbar spine. J Orthop Res 11:228, 1993. loading. J Orthop Res 5:479, 1987.
70. Houben GB, Drost MR, Huyghe JM, et al: 87. Lee CK, Langrana NA, Parsons JR, Zimmerman
Nonhomogeneous permeability of canine annulus MC: Development of a prosthetic intervertebral
fibrosus. Spine 22:7, 1997. disc. Spine 16:S253, 1991.
71. Skaggs DL, Weidenbaum M, Iatridis JC, et al: 88. Amiel D, Frank C, Harwood F, et al: Tendons
Regional variation in tensile properties and and ligaments: A morphological and biochemical
biochemical composition of the human lumbar comparison. J Orthop Res 1:257, 1984.
annulus fibrosus. Spine 19:1310, 1994. 89. Silver FH: Biomaterials, medical devices, and
72. Simon BR, Gaballa M: Poroclastic finite element tissue engineering. London, Chapman & Hall,
models for the spinal motion segment in 1994.
including ionic swelling. In Spilker RL, Simon 90. Nachemson AL, Evans JH: Some mechanical
BR (eds): Computational Methods in properties of the third human lumbar interlaminar
Bioengineering. ASME, 1988, p 93. ligament (ligament flavium). J Biomech 1:211,
73. Snijders J, Huyghe J, Janssen J: Triphasic finite 1968.
element model for swelling porous media. Int J 91. Blankevoort L, Huiskes R, DeLange A:
Num Meth Fluids 20:1039, 1997. Recruitment of knee-joint ligaments. J Biomech
74. Huyghe JM, Janssen JD: Quadriphasic mechanics Eng 113:94, 1991.
of swelling incompressible porous media. Int J 92. Amis AA, Dawkins GPC: Functional anatomy of
Eng Sci 35:793, 1997. the anterior cruciate ligament-fibre bundle actions
75. Oomens CWJ, Heus HJ, Huyghe et al: Validation related to ligament replacements and injuries. J
of the triphasic mixture theory for a mimic of Bone Joint Surg 73B:260, 1991.
intervertebral disk tissue. Biomimetics 3:171, 93. Tipton CM, Vailas AC, Matthes RD:
1995. Experimental studies on the influences of
76. Kalfas I, Wilberger J, Goldberg A, Protsko ER: physical activity on ligaments, tendons, and
Magnetic resonance imaging in acute spinal cord joints: A brief review. Acta Med Scand Suppl
trauma. Neurosurgery 23:295, 1988. 771:157, 1986.
77. Mirvis SE, Geisler FH, Jelinek JJ, et al: Acute 94. Williams IF, McCullagh KG, Silver IA: The
cervical spine trauma: Evaluation with 1.5 T MR distribution of types I and III collagen and
imaging. Radiology 166:807, 1988. fibronectin in the healing equine tendon. Connect
78. Andersson GBJ; Epidemiology. In Weinstein JN, Tissue Res 12:211, 1984.
Rydevik BL, Sonntag VKH (eds): Essentials of 95. Viidik A: Structure and function of normal and
the Spine. New York, Raven 1995, p 1. healing tendons. In Mow VC, Ratcliffe A, Woo
79. Praemer A, Fumer S, Rice DP: Musculoskeletal SL-Y (eds): Biomechanics of Diarthrodial Joints,
conditions in the United States. Park Ridge, IL, vol 1. New York, Springer-Verlag, 1990, p 3.
American Academy of Orthopaedic Surgeons, 96. Noyes FR: Functional properties of knee
1992, p 26.
ligaments and alterations induced by
80. Panagiotacopulos ND, Pope MH, Block R, Krag immobilization. Clin Orthop 123:210, 1977.
MH: Water content in human intervertebral discs. 97. Johnson GA, Tramaglini DM, Levine RE, et al:
II. Viscoelastic behaviour. Spine 12:918, 1987. Age related changes in the tensile and
81. Pearce RJ: Morphologic and chemical aspects of viscoelastic properties of the human patellar
aging. In Buckwalter JA, Goldberg VM, Wood tendon. J Orthop Res 12:796, 1994.
SLY (eds): Musculoskeletal Soft-Tissue Aging: 98. Weisman G, Pope MH, Johnson RJ: The effect
Impact on Mobility. Rosemont, IL, American of cyclic loading on knee ligaments. Trans
Academy of Orthopaedic Surgeons, 1993, p 363. Orthop Res Soc 4:24, 1979.
82. Bemick S, Cailliet R: Vertebral end-plate 99. Amiel D, Woo SL-Y, Harwood FL, Akeson WH:
changes with aging of human vertebrae. Spine The effect of immobilization on collagen
7:97, 1982.
turnover in connective tissue: A biochemical-
biomechanical correlation. Acta Orthop Scand 116. Tipton CM, Matthes RD, Sandage MS: In situ
53:325, 1982. measurement of junction strength and ligament
100. Durselen L, Claes L: Ligament strain elongation in rats. J Appl Physiol 37:758, 1974.
measurements using strain gauge equipped 117. Fung YCB: Elasticity of soft tissues in simple
omega transducers. Workshop on Biomechanics elongation. Am J Physiol 213:1544, 1967.
of Human Knee Ligaments. Ulm, Germany, 118. Fung YCB: Biomechanics: Mechanical Properties
1987. of Living Tissues. New York, Springer-Verlag,
101. Arms SW, Pope MH, Boyle JB: Knee medial 1981.
collateral ligament strain. Trans Orthop Res Soc 119. Viidik A: Mechanical properties of parallel-fibred
7:47, 1982. collagenous tissues. In Viidik A, Vuust J (eds):
102. Woo SL, Gomez MA, Seguchi Y, et al: Biology of Collagen. London, Academic Press,
Measurement of mechanical properties of 1980, p 237.
ligament substance from a bone-ligament-bone 120. Huiskes R: Mathematical modeling of the knee.
preparation. J Orthop Res 1:22, 1983. In Finerman GAM, Noyes FR (eds): Biology of
103. Woo SL, Chan SS, Yamaji T: Biomechanics of the Traumatized Synovial Joint: The Knee as a
knee ligament healing, repair and reconstruction. Model. Park Ridge, IL, American Academy of
J Biomech 30:431, 1997. Orthopaedic Surgeons, 1992, p 419.
104. Lee TQ, Woo SL-Y: A new method for 121. Hirokawa S: Biomechanics of the knee joint: A
determining cross-sectional shape and area of critical review. Crit Rev Biomed Eng 21:79,
soft tissues. J Biomech Eng 110:110, 1988. 1993.
105. Woo SL, Danto MI, Ohland KJ: The use of a 122. Finerman GAM, Noyes FR (eds): Biology of the
laser micrometer system to determine the cross- Traumatized Synovial Joint: The Knee as a
sectional shape and area of ligament: A Model. Park Ridge, IL, American Academy of
comparative study with two existing methods. J Orthopaedic Surgeons, 1992.
Biomech Eng 112:426, 1990. 123. Hirschman HP, Daniel DM, Miyasaka K: The
106. Mosler E. Folkhard W, Knorzer E, et al: Stress- fate of unoperated knee ligament injuries. In
induced molecular rearrangement in tendon Daniel DM, Akeson WH, O’Conner JJ (eds):
collagen. J Mol Biol 182:589, 1985. Knee Ligaments: Structure, Function, Injury and
107. Noyes FR, Grood ES: The strength of the Repair. New York, Raven, 1990, p 481.
anterior cruciate ligament in humans and rhesus 124. Crowinsheild RD, Pope MH: The strength and
monkeys. J Bone Joint Surg 58B: 1074, 1976. failure characteristics of rat medial collateral
108. Butler DL, Kay MD, Stouffer DC: Comparison ligaments. J Trauma 16:99, 1976.
of material properties in fascicle bone units from 125. Woo SL-Y, Orlando CA, Gomez MA, et al:
human patellar tendon and knee ligaments. J Tensile properties of the medial collateral
Biomech 19:425, 1986. ligaments as a function of age. J Orthop Res
109. Siegler A, Block J, Schneck CD: The mechanical 4:133, 1986.
characteristics of the collateral ligaments of the 126. Peterson RH, Gomez MA, Woo SL-Y: The
human ankle joint. Foot Ankle 8:234, 1988. effects of strain rate on the biomechanical
110. Woo SL-Y, Peterson RH, Ohland KJ, et al: The properties of the medial collateral ligament: A
effects of strain rate on the properties of the study of immature and mature rabbits. Trans
medial-collateral ligament in skeletally immature Orthop Res Soc 12:127, 1987.
and mature rabbits: A biomechanical and 127. Viidik A: Age-related changes in connective
histological study. J Orthop Res 8:712, 1990. tissues. In Viidik A (ed): Lectures on
111. Savelberg HHCM, Kooloos JGM, Huiskes R, Gerontology; vol I: On Biology of Aging.
Kauer JMG: Stiffness of the ligaments of the London, Academic Press, 1982, p 173.
human wrist joint. J Biomech 25:369, 1992. 128. Woo SL-Y, Ohland KJ, Weiss JA: Aging and
112. Danto MI, Woo SL-Y: The mechanical properties sex-related changes in the biomechanical
of skeletally matured rabbit anterior cruciate properties of the rabbit medial collateral
ligament and patellar tendon over a range of ligament. Mech Ageing Dev 56:129, 1990.
strain rates. J Orthop Res 11:58, 1993. 129. Woo SL-Y, Hollis JM, Adams D, et al: Tensile
113. Race A, Amis AA: The mechanical properties of properties of human femur-anterior cruciate-tibia
the two bundles of the human posterior cruciate complex: The effects of specimen age and
ligament. J Biomech 27:13, 1994. orientation. Am J Sports Med 19:217, 1991.
114. Soslowsky LJ, An CH, Johnson SP, Carpenter 130. Fetto JF, Marshall JL: Medial collateral ligament
JE: Geometric and mechanical properties of the injuries of the knee: A rationale for treatment.
coracoacromial ligament and their relationship to Clin Orthop 132:206, 1978.
rotator cuff disease. Clin Orthop Relat Res 131. Jones RE, Henley MB, Francis P: Nonoperative
304:10, 1994. management of isolated grade III collateral
115. Rogers GJ, Milthorpe BK, Muratore A, ligament injury in high school football players.
Schindhelm K: Measurement of the mechanical Clin Orthop 213:137, 1986.
properties of the ovine anterior cruciate bone- 132. O’Donoghue DH, Rockwood CA, Frank GR, et
ligament-bone complex: A basis for prosthetic al: Repair of the anterior cruciate ligaments in
evaluation. Biomaterials 11:89, 1990. dogs. J Bone Joint Surg 48A:503, 1966.
133. Hawkins RJ, Misamore GW, Merritt TR: EMG during a vigorous movement in man.
Followup of the acute nonoperated isolated Scand J Rehabil Med 10:99, 1978.
anterior cruciate ligament tear. Am J Sports Med 145. Komi PV, Fukashiro S, Jarvinen M:
14:205, 1986. Biomechanical loading of Achilles tendon during
134. O’Donaghue DH, Frank GR, Jeter GL, et al: normal locomotion. Clin Sports Med 11:521,
Repair and reconstruction of the anterior cruciate 1992.
ligament in dogs. J Bone Joint Surg 53A: 710, 146. Wallbridge N, Dowson D: The walking activity
1971. of patients with artificial hip joints. Eng Med
135. Daniel DM, Teiege RA, Grana WA: Knee and 11:95, 1982.
leg: Soft-tissue trauma. In Orthopaedic 147. Abrahams M: Mechanical behavior of tendon in
Knowledge Update. Park Ridge, IL, American vitro: A preliminary report. Med Biol Eng 5:433,
Academy of Orthopaedic Surgeons, 1990, p 557. 1967.
136. Fleming B, Beynnon BD, Johnson RJ, et al: 148. Benedict JV, Walker LB, Harris EH: Stress-strain
Isometric versus tension measurements: A characteristics and tensile strength of
comparison for the reconstruction of the anterior unembalmed human tendons. J Biomech 1:53
cruciate ligament. Am J Sports Med 21:82, 1993. 1968.
137. Radford WJP, Amis AA: Biomechanics of a 149. Butler DL, Grood, ES, Noyes FR, et al: Effects
double prosthetic ligament in the anterior of structure and strain measurement technique on
cruciate deficient knee. J Bone Joint Surg the material properties of young human tendons
72B:1038, 1990. and fascia. J Biomech 17:579, 1984.
138. Daniel DM, Stone ML, Dobson BE, et al: Fate 150. Bennett MB, Ker RF, Dimery NJ, Alexander
of the ACL-injured patient: A prospective RM: Mechanical properties of various
outcome study. Am J Sports Med 22: 632, 1994. mammalian tendons. J Zool Lond (A) 209-537
139. Gillquist J: Repair and reconstruction of the 1986.
ACL: Is it good enough? Arthroscopy 9:68, 151. Schechtman H, Bader DL: Dynamic
1993. characterisation of human tendons. Eng Med
140. Clancy WG, Narechania RG, Rosenberg TD, et 208:241, 1994.
al: Anterior and posterior cruciate ligament 152. Schechtman H, Bader DL: In vitro fatigue of
reconstruction in rhesus monkeys. J Bone Joint human tendons. J Biomechanics 30:829, 1997.
Surg 63A:1270, 1981. 153. Schechtman H: Mechanical characterisation of
141. Eastwood M, Ray field EJ, McGrouther DA, fatigue failure. In Human Tendons. PhD Thesis.
Brown RA: Cellular alignment and collagen London, University of London, 1995.
production in ligament equivalent constructs. In 154. Micheli LJ: Lower extremity overuse injuries.
Proceedings of the 3rd International Conference Acta Med Scand Suppl 711:171, 1986.
on Cellular Engineering. San Remo, Italy, 155. Jarvinen M: Epidemiology of tendon injuries in
September 1997. sports clinics. Sports Med 11:493, 1992.
142. Woo SL-Y, Smith DW, Allen CR, et al: 156. Uhthoff HK, Sarkar K, Maynard JA: Calcifying
Engineering and healing of the rabbit medial tendinitis: A new concept of its pathogenesis.
Clin Orthop Relat Res 118:164, 1976.
collateral ligament. In Proceedings of the 3rd
157. Hayashi K: Biomechanical studies of the
International Conference on Cellular Engineering.
San Remo, Italy, September 1997. remodelling of knee joint tendons and ligaments.
J Biomech 29:707, 1996.
143. Guyton AC: Textbook of Medical Physiology,
158. Banes AJ, Tsuzaki M, Hu P, et al: PDGF-BB,
7th ed. Philadelphia, WB Saunders, 1986.
IGF-1 and mechanical load stimulate DNA
144. Wahrenberg H, Lindbeck L, Ekholm J: Knee
synthesis in avian tendon fibroblasts in vitro. J
muscular moment, tendon tension force and Biomech 28:1505, 1995.
CHAPTER 3
Biomechanics of Bone
John Behiri and Deepak Vashishth
Because bone is the major component of the skele¬ tive summary rather than a comprehensive litera¬
ton, associated with sustaining and distributing a ture review and is concerned with cortical or
range of applied forces to the supporting surface compact bone, which constitutes the outer, load-
of the body, the mechanical behavior of this sys¬ bearing, shell of the major support bones.
tem has become one of the most extensively inves¬
tigated when compared with those of any other
biologic tissue material, with both its mechanical STRUCTURE OF BONE
properties and its structure providing a topic of
active research interest for more than a century. Bone is a complex system of structure and func¬
More recently, research in this area has been stim¬ tion on the macroscopic, microscopic, and ultrami-
ulated by the almost routine use of various syn¬ croscopic levels, with numerous and varied chemi¬
thetic materials as replacements for bone in a cal and physiologic interrelations. Bone is a
variety of surgical procedures, ranging from bone dynamically adaptable and metabolically active
grafts to total hip replacement, because there is a tissue whose form is continuously undergoing sub¬
constant need to define precisely the mechanical tle remodeling to conform to its functions. It is a
properties of the natural tissue being replaced. highly vascular tissue composed of interconnected
The mechanical behavior of a structure such as cells in an intercellular substance, with an out¬
bone depends not only on the size and shape of standing capacity for self-repair. Bone forms the
the structure but also on the mechanical properties framework or skeleton of most vertebrates and has
of the materials of which it is composed, and the following functions:
knowledge of the properties of bone tissue is es¬
sential to understanding the behavior of whole • Supports the body
bones. Furthermore, apart from having some • Protects vital organs
unique properties, bone is self-repairing and can • Provides kinematic links and attachment sites
alter its properties and configuration in response for muscles
• Facilitates muscle action and body movement
to changes in mechanical demand. For example,
decreases in bone density are commonly observed In biologic terms, bone can be described as a
after long periods of disuse, and increases are “specialized” form of connective tissue. Connec¬
observed after periods of greatly increased func¬ tive tissue connects body parts and is an aggrega¬
tion. Changes are also observed in bone shape tion of similarly specialized cells that are united
during fracture healing and after certain postopera¬ to perform a particular function. Typical forms of
tive bone procedures.1 connective tissue other than bone include hyaline
This chapter is concerned with the structure and cartilage, found only in a few places in the body,
mechanical properties of cortical or compact bone and dense fibrous tissue, such as tendons and
and its behavior under different loading condi¬ ligaments. Hyaline cartilage is the avascular soft
tions. Any direct comparisons of data are often tissue that forms the supporting structures of the
misleading, and the inclusion of all published ex¬ larynx (or voice box), attaches the ribs to the
periments on the mechanical properties of bone in breastbone, and covers the ends of bones where
any review would generally leave anyone discour¬ they form joints. Tendons and ligaments are the
aged. This chapter, therefore, is more of a selec¬ soft, ropelike structures made of dense fibrous
65
tissue; tendons attach skeletal muscle to bone, and A Amino acid sequence
ligaments connect bone to bone at joints. gly—x—y— gly—pro—hypro—gly—x—y
Certain characteristics differentiate bone from
these other forms of connective tissue. The most
B Dimensions of individual collagen chain
striking difference is that bone is hard—only den¬ k 15 Ao
tin and enamel in teeth are harder. This hardness T

-S- *1
3,000 A
is the result of the deposition within a soft organic
matrix (collagen) of a complex mineral substance C Triple helix
known as calcium hydroxyapatite, which is com¬
posed of calcium, phosphorus, sodium, magne¬
sium, fluoride, and other ions in trace amounts.
Bone, therefore, not only forms the framework or D Quarter stagger
skeleton of most vertebrates, providing support
and protection to the body, but also acts as a
mineral reservoir, playing a vital role in the main¬
tenance of steady-state levels of ions in the body. E Banded pattern
Apart from collagen and mineral, bone also con¬
tains small amounts of mucopolysaccharides (also
known as ground substance) and water (Table
3-1).
FIGURE 3-1 Collagen structure. A, Amino acid
To understand the properties and behavior of sequence—repeating tripeptide. B, Molecular dimensions. C,
bone, it is necessary to consider the various levels Triple helix—three alpha chains. D, Quarter stagger of
of structural organization in bone, namely the ul- collagen molecules. £, Banded pattern resulting from
trastructural, microstructural, and macrostructural alternating overlap and hole zones. (From Owen R,
levels, which interact in providing bone with its Goodfellow J, Bullough P: Scientific Foundations of
characteristic properties. Orthopaedics and Traumatology. Philadelphia, WB Saunders,
1980, sec I, p 30.)
ULTRASTRUCTURE
At the ultrastructural level, bone is unique in the primary building unit of collagen is the tropocolla-
body in that it contains both the mineral compo¬ gen molecule (Fig. 3-1), which is a long, rigid
nent, hydroxyapatite, and the organic component, molecule (300 X 1.5 nm) and is composed of
collagen. The mineral is dispersed in microcrystal¬ three spiral chains of peptides (monomers of pro¬
line form throughout the collagen matrix, and de¬ teins, i.e., amino acids), known as a chains, bound
spite numerous investigations into the nature of together in a triple helix. Different types of colla¬
bonding between these two phases, their relation¬ gen for the function in question can be obtained
ship still remains to be fully elucidated. depending on the amounts and sequence of the
Collagen is the most abundant protein found in amino acids on the chain. Type I collagen is typi¬
the body and can be found in places such as cally found in bone and skin, type II in cartilage,
the skin, cartilage, tendons, and blood vessels. In and type III in blood vessels.
cartilage, collagen opposes the swelling pressure The tropocollagen molecules are aligned in a
of proteoglycans to yield a shock-absorbing mate¬ quarter-stagger array (Fig. 3-2), with an overlap
rial; in skin and blood vessels, it is mixed with
elastin to combine strength and elasticity; and in
tendons, where only tensile strength is of primary J L
importance, it is almost the only component. The J—i J
TABLE 3-1 Composition of Adult Cortical Bone
Dry
Weight
Wet
Weight
1U4 Overlap zone
.
Component (%) (%) \ Hole zone

Collagen 18.5 15.5 FIGURE 3-2 Two-dimensional quarter stagger of collagen
Mineral 70.0 59.9 molecules with alternating overlap and hole zones. (From
Ground substance 3-3 2.8 Owen R, Goodfellow J, Bullough P: Scientific Foundations of
Water 8.2 22.7 Orthopaedics and Traumatology. Philadelphia, WB Saunders,
1980, sec I, p 31.)
• I Chapter 3 BIOMECHANICS OF BONE 67
zone of 26.5 nm and a hole zone of 37.5 nm

between subsequent molecules. A collection of
tropocollagen molecules forms a collagen fibril,
and bundles of fibrils form fibers that have diame¬
ters ranging from 0.2 to 12 |xm. Under an electron
microscope, the fibrils appear to have a banded
pattern, which repeats every 64 nm as a direct
consequence of the quarter-stagger array.
As previously stated, the most important me¬
chanical properties of collagen are its tensile stiff¬
ness and strength. The only information available
about collagen’s mechanics as a tensile material is
that about the properties of vertebrate tendon and
ligament (collagen is the major component of ten¬
don, making up 70% to 80% of the dry weight).
Tendon is the structure that provides the rigid
attachment of muscle to bone, and as such, it must FIGURE 3-3 Stress-strain behavior of collagen.
transmit the muscle force with a minimum of loss.
This is achieved through the parallel arrangement
of the collagen fibers to form ropelike structures portion of the curve indicates that the force is
with a high modulus of elasticity and high tensile being applied directly to extended collagen fibers.
strength. Table 3-2 compares some of the proper¬ Surrounding the collagen fibrils in a highly or¬
ties of collagen with other materials. ganized geometric arrangement are small crystal¬
Figure 3-3 illustrates the stress-strain behavior lites of hydroxyapatite (Ca10[PO4]6OH2), the major
of collagen and shows that collagen can be ex¬ inorganic phase of bone, which are 20 to 40 nm
tended reversibly to strains of about 4%, but long and about 5 nm thick. It is not known with
strains above this level result in irreversible any certainty how the crystals and the collagen are
changes (under normal working conditions, colla¬ bound together, although some form of bonding is
gen is strained within the limits of 0% to 3%). certainly present. The mineral is apparently ini¬
The collagen fibers fracture at strains of about 8% tially deposited in the holes between the “heads”
to 10%, which on examination reveal that it is a and “tails” of the tropocollagen molecules (Fig.
consequence from the pulling apart of adjacent 3-4). More mineral is later deposited all over and
molecules rather than from the fracture of the within the collagen fibrils, with the long axis of
rodlike molecules. It also indicates that intermo- the mineral plates being fairly well aligned with
lecular cross-linking is an important factor in me¬ the collagen fibrils.
chanical properties. As a material on its own, hydroxyapatite is
In the case of Young’s modulus, this increases quite stiff compared with collagen (Table 3-3). Its
gradually over the first few percent of extension Young’s modulus (E = 130 GPa) is well over
and then reaches a value of about 1 giga-pascal half that of steel (E = 210 GPa) and is greater
(GPa). Experiments with polarized light have than that of aluminium (70 GPa), but it fractures
shown that collagen fibers, particularly in tendon, easily and consequently possesses a low fracture
are crimped and that on extension, the fibers toughness. This superimposition of two quite di¬
straighten and become parallel. The initial verse materials with significantly different me¬
rounded portion of the stress-strain curve reflects chanical properties results in an ultrastructural
the straightening of these crimps, and the linear composite.
MICROSTRUCTURE
TABLE 3-2 Comparison Between Some of the
Properties of Collagen and Other Engineering On the microscopic level, above the level of the
Materials
collagen fibril and its associated mineral (and de¬
Tensile
pendent on the manner in which the fibers are laid
Young's
Modulus Strength down), bone exists in two forms: woven bone and
Material (GPa) (MPa) lamellar bone. Woven bone is characterized by an
isotropic nonlamellar pattern of collagen fibers,
Collagen 1-2 50-1000
150
which are usually laid down very quickly. This
Aluminum 70
220 700 type of bone is most characteristically found in
Steel
the fetus and is also the first type of bone to form
FIGURE 3-4 Distribution of apatite crystallites in

collagen matrix showing orientation of the C axis. (From
Lees S, Davidson CL: The role of collagen in the elastic
properties of calcified tissues. J Biomechanics 10:473-486,
1977. Copyright 1977, reprinted with permission from
Elsevier Science.)
during healing from a fracture. The collagen in thick with a high mineral content and little colla¬
woven bone is fine fibered, 0.1 |xm in diameter, gen between the pairs of lamellae. Lamellar bone
and randomly oriented. It contains cells (osteo- can extend for many millimeters and has three
cytes) and blood vessels, and the spaces sur¬ major patterns: osteonal, circumferential, and in¬
rounding the blood vessels are extensive. terstitial.
In contrast to woven bone, lamellar bone is laid In the case of osteonal lamellar bone, lamellae
down more slowly and has a preferred orientation. can be seen formed into densely packed concentric
The collagen and associated mineral are arranged lamellar structures, which combine concentrically
in sheets about 5 p,m thick. It has been suggested around vascular or haversian canals to build up
that the collagen fibrils in a lamella are all oriented the next hierarchical structure (i.e., the haversian
in the same direction, but this is probably not the canal, or osteon). This is essentially a cylindrical
case. In many lamellae, the fibrils are in small structure (Fig. 3-5) whose long axis courses some¬
domains; within a domain, the fibril orientation is what irregularly along the long axis of the bone.
constant, but it changes from one domain to the It contains thick walls and a narrow lumen or
next and from one lamella to another. The colla¬ haversian canal about 22 to 110 p,m in diameter
gen in lamellar bone is thicker than in woven bone surrounded by 4 to 20 concentrically arranged
(about 2 to 3 pan). The division between one lamellae, with each lamella being 3 to 7 |xm thick
lamella and the next is abrupt, and there appears and having its collagen fibers oriented parallel to
to be a sheet of interlamellar bone about 0.1 p.m each other. Each adjacent lamellar layer has a
different orientation of collagen fibers, and as a
result, the successive lamellations appear as alter¬
TABLE 3-3 Comparison of the Individual nating bright and dark layers. Circumscribing the
Components of Bone outermost concentric lamella of the haversian sys¬
Property
tem is a narrow zone known as the cement line,
Collagen Mineral Bone
which contains calcified mucopolysaccharides and
E (GPa) 1 130 8-30 is devoid of collagen. It is 1 to 2 p,m thick and is
UTS (MPa) 50 100 120 the weakest constituent of bone. These concentric
lamellar structures bonded to one another by
E, Youngs modulus; UTS, ultimate tensile strength.
means of these cement lines form a fiber-rein-
I Chapter 3 BIOMECHANICS OF BONE 69
FIGURE 3-5 Detailed structure of Apatite

an osteon. (Adapted from Park JB: Haversian Concentric
mineral crystals
Biomaterials. New York, Plenum canal lamella
(200-400 A long)
Press, 1979, p 105.) (3-7 pm)
forced composite on the microstructural level. De- mm3 of bone can contain as many as 26,000
velopmentally, two distinct types of osteons exist, lacunae and 10 canaliculi, which indicates that
namely primary and secondary. Primary osteons they occupy a major proportion of the bone ma¬
are formed by the deposition of lamellae around a trix. Osteoblasts lie at the end farthest from the
previously existing vascular canal and are differ¬ bone surface and are actively engaged in the man¬
ent morphologically and mechanically from sec¬ ufacture and secretion of components of bone ma¬
ondary osteons. Primary osteons do not have ce¬ trix. They are mononucleated cells, vary consider¬
ment lines because they are not the product of ably in size (15 to 80 pm), and can take different
bone remodeling, and they have smaller vascular shapes, from ovoid to columnar to cuboidal. Dur¬
canals and fewer lamellae than secondary osteons. ing bone formation, osteoblasts become trapped
In circumferential bone, lamellae can be seen within the calcified tissue and differentiate into
uninterrupted around the circumference of the osteocytes.3 Osteoclasts are large, 20- to 100-pm,
shaft, located on the external surfaces of cortical multinucleated cells (2 to 100 cells) found on or
bone immediately beneath the periosteum and on near bone surfaces and are associated with bone
the internal surface adjacent to the endosteum resorption. They are particularly abundant in areas
(Fig. 3-6). of active bone resorption.
In interstitial lamellae, fragments of previous
concentric and circumferential lamellae fill the MACROSTRUCTURE
gaps between haversian systems. They are contin¬
uous with the osteons, and their makeup is identi¬ At the macroscopic level, bone is composed of
cal to that of the osteons, apart from their geomet¬ two types: cortical, or compact, bone and cancel¬
ric form. As in the case of osteons, no point in lous, or trabecular, bone (see Fig. 3-7). Compact
the interstitial lamellae is farther than 100 pm bone appears solid; the only microscopic spaces
from its blood supply. present are haversian and Volkmann’s canals, lacu¬
Associated within or on the matrix can be found nae, canaliculi, and resorption sites, and these are
three characteristic cell types: the osteocyte, the invisible to the naked eye. Cancellous bone, on
osteoblast, and the osteoclast, each of which is the contrary, is a porous material and is composed
associated with specific functions. Osteocytes are of trabeculae that join to form a network. It is
contained in small disk-shaped cavities, known as generally accepted that cortical and cancellous
lacunae (Fig. 3—7), that lie along the boundaries bone are similar in their material and morphologic
of each layer, or lamella. A number of theories characteristics,4-5 and the main distinction between
exist as to the function of osteocytes, but it is the two types of bone lies in their porosity. Corti¬
generally believed that they assist in the exchange cal bone has a porosity of 5% to 30%; porosity of
of materials between tissue fluids and bone matrix. cancellous bone is 30% to 90%.
The lacunae interconnect with each other and with A typical long bone such as the tibia or femur
the haversian canal through a matrix of tiny chan¬ is composed of (1) the diaphysis, which is the
nels, called canaliculi, that are about 0.35 pm in main central cylindrical shaft and contains walls
diameter. According to calculations of Martin,2 1 of dense compact bone; (2) the epiphyses, the
Haversian
system
Inner (osteon)
circumferential
lamellae Outer
circumferential
lamellae
Volkmann's
canal
Endosteum Haversian Periosteum

canal
FIGURE 3-6 Microstructural features of cortical bone. (Redrawn from Junqueira LC, Carneiro J,
Kelley RO: Basic Histology, 9th ed. Norwalk, CT, Appleton & Lange, 1999, p 136.)
extreme ends of the bone that are composed of region encloses the medullary or marrow-filled
cancellous bone, covered at the surface by a thin cavity, which is lined with a fibrous membrane
layer of compact bone; and (3) the metaphyses, known as the endosteum. This cavity contains
the two intermediate conelike regions connecting yellow bone marrow and communicates freely
the shaft and articular ends (see Fig. 3-7). The with that of the trabecular bone.
articular regions and the metaphyses are spongy The exterior bone surface is enclosed by the
in form, consisting of weblike trabeculae that di¬ periosteum, a membranous tissue whose vascular
vide the interior volume of the metaphyses and network directly communicates with the underly¬
epiphyses into intercommunicating pores filled ing bone, except at the joints, where articular
with marrow. The sizes of the pores vary consider¬ cartilage forms the covering.
ably throughout the bone interior and thus present
a structure of variable porosity. The arrangement
of the trabeculae is functional. Their orientation BIOMECHANICAL
closely parallels the trajectories of maximum CONSIDERATIONS
stress (Fig. 3-8) and thus gives the skeleton maxi¬
mum rigidity and resistance to mechanical stresses From the description given previously, it is clear
and strains. The compact bone of the diaphysis that bone may be regarded as a two-phase hierar-
FIGURE 3-7 Structure of a long bone.
FIGURE 3-8 Diagram of the lines of

O Compression
stress in the upper femur. (From Margel-
o Tension
Robertson DR: Studies of Fracture in
• Compression Bone. PhD Thesis. Stanford, CA, Stanford
and tension
University, 1973.)
chical composite. As a result of the way it is built STRESS-STRAIN BEHAVIOR

and the functional requirements placed on it, bone
can exhibit a variety of mechanical properties. Influence of Orientation
Furthermore, it can adapt in vivo to suit the load¬
ing environment. A typical stress-strain curve for bone in tension
The present state of knowledge of the mechani¬ is shown in Figure 3-9 and illustrates that bone
cal properties of bone has been well documented.1, behaves in a manner similar to that of engineering
^10 Two of the most important mechanical proper¬ materials in that it consists of an initial elastic
ties of bone are its strength and stiffness. At the region followed by yielding, namely the onset of
ultrastructural level, these and other properties are nonelastic deformation, before failure. The non¬
dominated by the contribution of collagen fibers elastic portion of the curve, often referred to as
and hydroxyapatite crystals and the bonding be¬ the plastic region, represents irreversible micro¬
tween them. When compared with conventional damage in the bone, whereas the linear slope of
engineering composites, bone has its fibrous struc¬ the curve provides a direct measure of Young’s
ture, collagen, as the matrix, as opposed to the modulus (E) of bone. Because the structure of
strong, reinforcing element of the composite. At bone is dissimilar in the transverse and longitudi¬
the microstructural level, bone again exhibits a nal directions, it exhibits different mechanical
composite nature consisting of lamina of tissue properties when loaded along different axes, a
arranged in a number of organized patterns, pre¬ characteristic known as anisotropy. For example,
viously described and separated by weak inter¬ Young’s modulus does not have a unique value
faces. Even at the macrostructural level, bone can but varies with orientation, with a maximum value
be considered as a composite, consisting of dense parallel to the long axis of the bone and a mini¬
cortical bone to form the shafts and peripheries mum value (a factor of about 2 smaller) perpen¬
and trabecular bone at the extremities to provide dicular to the long axis.11 The absolute values
a means of distributing applied loads. The stresses measured for E depend on the precise nature of the
produced in bone are complex. They comprise bone specimen and its microstructure but normally
both normal and shear stresses that vary topo¬ range between 7 and 30 GPa.
graphically in magnitude and direction. To deter¬ Some of the most comprehensive examinations
mine the behavior of materials, they must be ex¬ of the effects of orientation on the ultimate proper¬
amined under simple stress conditions initially and ties of human cortical bone can be found in the
the results then applied in more complex condi¬ work of Reilly and colleagues12, 13 and Bonfield
tions. Standard mechanical tests, such as simple and O’Connor,14 who investigated the tensile and
tension, compression, shear, torsion, impact, and compressive behavior of cortical bone at different
fatigue tests, can supply mechanical constants and orientations. Reilly and Burstein13 reported elastic
information that can in turn be used to determine modulus values of 17.4 GPa for bone loaded in
the behavior of a material under conditions of the longitudinal direction compared with 12.7 GPa
normal use. This section reviews some of the in the transverse direction. A similar story holds
major findings with respect to the properties that in the case of ultimate tensile and compressive
describe the mechanical behavior of bone.
strength, for which it was found that bone loaded
Figure 3-9 Stress-strain curves for wet and

dry cortical bone. (From Black J: Orthopaedic
Biomaterials in Research and Practice. New York,
Churchill Livingstone, 1988, p 116.)
FIGURE 3-10 Variation of fracture strength

as a function of angle between the longitudinal
and radial axes. (From Pope MH, Outwater
JO: Mechanical properties of bone as a
function of position and orientation. J Longitudinal Radial
Biomechanics 7:61-66, 1974. Copyright 1974, 1 I I I I I I
reprinted with permission from Elsevier 0 15 30 45 60 75 90
Science.) Angle (degrees)
in the transverse direction was significantly Viscoelasticity

weaker in both tension (53 MPa) and compression
As a consequence of bone’s viscoelastic nature
(133 MPa) compared with bone loaded in the
(mainly due to the significant amount of fibrous
longitudinal direction, for which values of 135
proteins present), the stress-strain curves depicted
MPa and 193 MPa have been reported.13 Pope
in Figures 3-7 and 3-8 are influenced by the rate
and Outwater15 considered the variation in the
at which bone is loaded. This effect was originally
mechanical properties of long bones as a function
demonstrated in a convincing manner by McElha-
of position and orientation, with respect to the
ney17 and is shown in Figure 3-12. Young’s modu¬
longitudinal axis of the bone. Their results indi¬
lus, ultimate compressive strength, and yield
cated there was a distinct change of strength and
strength increased with increasing rate of loading;
elasticity of bone as a function of distance from however, the failure strain and the fracture tough¬
the epiphysis, which was directly related to the ness of the bone reached a maximum and then
manner in which the osteons were oriented. At decreased, implying that there is a critical rate
the mid-diaphysis, the osteons are mostly oriented of loading. Strain rate effects have since been
parallel to the longitudinal axis, and the bone is confirmed by other workers, including Bonfield
relatively much stronger, as shown in Figure 3-10. and Clark18 and Currey.19 A more recent and com¬
At the epiphysis, the orientation of the osteons prehensive study was performed by Wright and
changes, and the bone becomes better equipped Hayes,20 in which an increase in Young’s modulus,
for its role as a bearing surface rather than a col¬ by a factor of 2 (about 17 to 40 GPa), was reported
umn. with increases in strain rate from 5 X 10_4s_1 to
The contribution of collagen and mineral to the 2.4 X 102s_1. An increase of E with strain rate of
elastic-plastic properties was examined by this magnitude appears to represent most of the
Burstein and associates,16 who found that progres¬ data gathered in various investigations either as a
sive surface decalcification of bone specimens function of strain rate or at particular strain rates.
with dilute hydrochloric acid produced progres¬ The loading rate is clinically significant because
sive decrease in the tensile yield stress and ulti¬ it influences the fracture pattern and the amount
mate stress but caused no change in the yield of soft tissue damage at fracture. Depending on
strain or ultimate strain unless decalcification was how rapidly the load is applied, the crack may
completed (Fig. 3-11). They suggested that the remain a single crack or may bifurcate. The more
mineral contributes the major portion of the tensile rapidly the load is applied, the more the crack
yield strength of bone, whereas the magnitude of bifurcates and the greater the number of resulting
the plastic modulus is a function only of the prop¬ bone fragments. When a bone fractures, the stored
erties of collagen and is not a major contributor energy is released. At low rates of loading, the
to the tensile yield strength of bone. energy is dissipated through the formation of a
FIGURE 3-11 Average stress-strain curves for specimens of bovine bone treated to
create increasing degrees of decalcification. (From Burstein AH, ZikaJM, Heiple KG, Klein L:
Contribution of collagen and mineral to the elastic-plastic properties of bone. J Bone Joint
Surg 57A:958, 1975.)
single crack. Under rapid loading, bone dissipates Young’s modulus in normal and pathologic human
the large amount of energy through the creation cortical bone in relation to bone mass and density,
of additional free bone surfaces and the creation showed increases in Young’s modulus from 7 to
of a large number of cracks. The number of bone 21 GPa for similar density ranges.
fragments can therefore be directly related to the
velocity of the impact. Influence of Age
Aging is associated with significant changes in the

Influence of Density
morphology, composition, and density of cortical
Although density has received limited attention in bone, resulting in changes in the mechanical prop¬
the literature, its importance has been recognized, erties. Previous investigations23,24 have shown that
with experimental studies showing that the elastic both Young’s modulus and yield strength increase
modulus correlates well with the cube of the den¬ to maximum values in the third and fourth decades
sity and that the ultimate fracture strength corre¬ of life (Fig. 3-14) and then slowly decrease with
lates well with the square of the density of bone increasing age, whereas the elongation to failure
tissue21 (Fig. 3-13). Wall and coauthors22 carried decreases throughout life. Possible improvements
out tensile strength and density measurements on may result from physical activity, whereas possi¬
small samples of cortical bone taken from the ble negative effects may result from hormonal
femoral diaphysis and showed increases in ulti¬ changes (e.g., due to menopause) or from seden¬
mate tensile strength from 58 to 88 MNm'2 for tary lifestyle. In a more recent investigation,25
density increases ranging between 1.8 and 2 average deterioration rates of 5% per decade for
Mgm~3. Abendschein and Hyatt,21 investigating ultimate stress, 9% per decade for ultimate strain,
FIGURE 3-13 Influence of density on the compressive properties of compact and trabecular bone. Shown
are human (triangles, filled circles, filled squares) and bovine (open circles, open squares) bone tissue.
(Modified from Carter DR, Hayes WC: Bone compressive strength: The influence of density and strain rate.
Srionce 194'I 174-1 176, 1976. Copyright 1976 American Association for the Advancement of Science;
Galante J, Rostoker W, Ray RD: Physical properties of trabecular bone. Calcif Tissue Res 5:236-246, 1970;
and McElhaneyJH, Byars EF; ASME Publ. 65-WA/HUF09, 1965.)
FIGURE 3-14 Summarized results of

the effect of age on mechanical
properties of bone. (From Burstein AH,
Reilly DT, Martens M: Ageing of bone
tissue: Mechanical properties. J Bone
Joint Surg S8A:82-86, 1976.)
and 12% per decade for energy absorption were exhibits only a small total elongation to fracture
reported based on 235 subjects ranging in age (0.5% to 3.5%) with limited nonelastic deforma¬
from 20 to 102 years.
tion, an inevitable consequence of specimen prep¬
aration is the presence of random surface flaws.
FRACTURE MECHANICS OF BONE This effect was well appreciated by Evans,6 whose
careful surveys of ultimate tensile strength in¬
The fracture of bone is a complex process that has
volved tests on many specimens to produce statis¬
received considerable attention over the years.
tically significant results.
One of the difficulties in characterizing bone frac¬
Because the energy to fracture a brittle solid is
ture is the variability in the associated microstruc¬
significantly affected by the presence of cracks,
ture among bones, and indeed for different loca¬
however, an alternative approach to define bone
tions in the same bone, as well as the differences
fracture is by testing precracked specimens with a
related to age and sex. Moreover, because the
single characterized crack or notch from which
microstructure in the major support bones is aniso¬
fracture will propagate. Bonfield and Li26 first
tropic, the resultant fracture properties also depend
demonstrated that an introduced surface crack sig¬
on the orientation.
nificantly reduced the energy absorbed during
The pioneering work of Evans1-6 was of consid¬
fracture of bovine cortical bone specimens in both
erable value in distinguishing the relative effects
the longitudinal and transverse directions. A series
of microstructural parameters, strain rate, and tem¬
of significant investigations followed in which the
perature on bone fracture from measurements of
fracture energy of precracked or notched cortical
the ultimate tensile and compressive strength.
bone was determined for transverse and longitudi¬
Given that cortical bone is a brittle solid that
nal directions in various bones.15-27-28 Piekarski’s
• I Chapter 3 BIOMECHANICS OF BONE 77
study27 showed that at low strain rates, a crack

follows the weak interfaces of bone (i.e., the ce¬
ment lines), whereas at high rates of strain, a crack
propagates in a brittle manner without regard to
the microstructure. The cement lines have also
been shown to be a weakness by Evans and
Bang,29 Dempster and Coleman,30 and Behiri and
Bonfield.31
For a given bone, the considerable scatter (vari¬
ability) in values in the work (energy) required
to cause fracture, illustrated that the relationship
between the geometry of the introduced crack
and the energy or stress associated with fracture
required determination if an absolute measure of
bone fracture was to be established. The recogni¬
tion of the dependence of fracture stress on the
nature of introduced cracks provided the basis for
a more precise evaluation of the fracture of bone
in terms of its fracture toughness, defined as a
measure of the resistance of a material to crack
propagation. This approach utilized the concepts
of linear elastic fracture mechanics,32 developed
for engineering materials, to determine values of
the critical stress intensity factor (Kc, or fracture
toughness) and the critical strain energy release
rate (Gc) of bone. Such an approach to the failure
of bone was adopted by Wright and Hayes32 and
Behiri and Bonfield.31,33,34 Wright and Hayes32
Log velocity (ms 3)
demonstrated the significant effect of an increase
in bone density in producing an increase in both FIGURE 3-15 Dependence of critical stress intensity
Kc and Gc and established the important link factor (Kc) of bovine tibia cortical bone for different density
between fracture mechanics and microstructure. groups within the normal range. (From Behiri JC, Bonfield
W: Fracture mechanics of bone: The effects of density,
In an independent approach, Behiri and Bon¬
specimen thickness and crack velocity on longitudinal fracture
field31 conducted a comprehensive study of longi¬
J. Biomechanics 17:25-34, 1984. Copyright 1984, reprinted
tudinal fracture in bone and found that crack ve¬ with permission of Elsevier Science.)
locity increases from 1.75 X 10-5ms-1 to 23.6
X 10_5ms_1 produced increases in Kc and Gc
from 4.46 to 5.38 Mnmr3/2 and 1736 to 2796
Jm“2, respectively. At higher crack velocities, and fracture toughness values were shown to in¬
fracture became catastrophic. Hence, the concept crease from 3.2 MNm_3/2 for fracture in the longi¬
of a critical crack velocity associated with a maxi¬ tudinal direction to 6.3 MNm_3/2 for fracture in
mum in the fracture toughness of bone was estab¬ the transverse direction, again clearly illustrating
lished and shown to represent a transition from the anisotropic nature of bone. Age was also dem¬
controlled to catastrophic crack propagation. In onstrated to have a significant effect on the fracture
later work, which embraced the effects of density toughness of bone,35 with decreases in toughness
and crack velocity, this transition between micro- from 4.5 Mnm_3/2 at 25 years to 2.5 Mnm'3'2
structure-dependent fracture and catastrophic at 90 years (Fig. 3-17).
crack propagation associated with a maximum in The application of fracture mechanics tech¬
fracture toughness is shown in Figure 3-15. For niques for an evaluation of the fracture of bone
controlled, slow crack velocities, the investigators has developed to the stage at which meaningful
correlated fracture with crack-intersecting osteon results can be achieved that are consistent with
interactions, with osteon pullout similar to that the protocol followed for engineering materials.
noted by Piekarski.27 The dependence of fracture This approach has two advantages: first, tests can
on microstructure was not significant at higher be performed on a relatively small number of
crack velocities. The fracture characteristics at a specimens, which is helpful when the supply of
range of orientations under controlled slow crack bone is limited, as in the case of human bone; and
propagation were also investigated34 (Fig. 3-16), second, a more systematic approach to prosthesis
8-i
7-
6-
5-
4-
<D
CT>
to
1
—
<D
3-
>
<
2- FIGURE 3-16 Dependence of
critical stress intensity factor (Kc)
on orientation between 0 and 90
degrees. (From Behiri JC, Bonfield
W: Orientation dependence of the
fracture mechanics of cortical bone.
T "1 i i i i-1-1-1-r
0 10 20 30 40 50 60 70 80 90 J Biomechanics 22:863-872, 1989.
Copyright 1989, reprinted with
Orientation (degrees) permission of Elsevier Science.)
FIGURE 3-17 Critical stress intensity factor (Kc) associated with controlled, slow fracture
of human tibia cortical bone specimens as a function of age (both male and female, clinically
normal). (From Bonfield W, Behiri JC, Charalembides B: Orientation and age-related
dependence of the fracture toughness of cortical bone. In Perren SM, Schneider E [eds]:
Biomechanics: Current Interdisciplinary Research. Dordrecht, The Netherlands, Martinus
Nijhoff Publishing, 1985, pp 185-189. Reprinted with kind permission from Kluwer Academic
Publishers.)
design can be made with a precise knowledge of that if the stress levels are kept below a certain
the limiting fracture toughness of the natural tis¬ level, theoretically, the material will remain intact
sues as well as of the engineering materials con¬ regardless of the number of cycles of loading and
sidered. unloading. Seireg and Kempke36 conducted this in
vivo fatigue test and found that the S-N curve did
FATIGUE FRACTURE not exhibit a constant decline at lower loads but
had a distinct knee region at 42% of the static
In addition to bone fractures produced by a single failure load after 5 X 103 cycles. This was inter¬
monotonic load at stresses that exceed the ultimate preted as the “endurance limit” of the bone be¬
strength, fractures in bone may also occur by the yond which it did not fatigue. This is not the case
repeated or cyclic application of a lower load. for bone when tested in vitro. Initial investigations
These types of fractures, known as fatigue frac¬ by Swanson and colleagues37 and later studies by
tures and sometimes referred to clinically as stress other investigators did not detect any endurance
fractures, are usually sustained during continuous limit during in vitro fatigue testing. Carter and
strenuous physical activity. Such fractures occur coauthors, in a series of studies,38-41 carried out a
in bones of the lower limb, including metatarsals, comprehensive investigation of the influence of
calcaneus, tibia, fibula, femur, and pelvis, and stress amplitude, temperature, bone density, and
occurrence of such fractures in a particular bone microstructure on the fatigue life of compact bone.
varies with the activities of the individual. For They found that (1) decreasing the stress ampli¬
example, fatigue fractures of tibiae and metatar¬ tude from 108 to 65 MPa caused a four-fold in¬
sals are more common among athletes, ballet crease in the fatigue life, (2) decreasing the tem¬
dancers, and military recruits. perature from 45° to 21°C caused a three-fold
Fatigue of any material, including bone, is de¬ increase in the fatigue life, and (3) a 5% increase
fined as the progressive loss of strength or stiff¬ in density doubled the fatigue life.
ness under cyclic loads resulting in complete fail¬ In general, investigations into the cyclic load
ure of the material at stresses below the static behavior of bone in vitro have established that
failure levels. Tests to determine fatigue character¬ cyclic loading produces microstructural damage
istics are conducted by subjecting standardized that accumulates with each loading cycle and that
subjects to a constant amplitude loading and de¬ damage accumulates faster at higher intensities of
termining the total number of cycles to failure. cyclic loading.38 Furthermore, continued fatigue
Tests are performed at various stress levels, and loading of bone results in loss of strength and
the results are used to plot stress against the num¬ stiffness, which has been suggested as being due
ber of cycles to failure (S-N curve). Figure 3-18 to the accumulation of small microcracks. Because
is a compilation of fatigue data obtained from a living bone is a dynamic and self-repairing mate¬
variety of bone studies. In the case of bone tested rial, a fatigue fracture occurs only when the re¬
in vivo, the fatigue curve is asymptotic, indicating modeling processes outpace the fatigue processes,
and consequently, any comparisons between in reconstruction and mineral content on some
vivo and in vitro fatigue data should be viewed mechanical properties of bovine bone. J Biomech
with some caution. 8:81-86, 1975.
20. Wright TM, Hayes WC: Tensile testing of bone
over a wide range of strain rates: Effects of strain
References rate, microstructure and density. Med Biol Eng
14:671-679, 1976.
1. Evans FG: Stress and Strain in Bones. Springfield, 21. Abendschein W, Hyatt GW: Ultrasonics and
IL, Charles C Thomas, 1957. selected physical properties of bone. Clin Orthop
2. Martin RB: Porosity and specific surface of bone. 69:294-301, 1970.
Crit Rev Biomed Eng 10:179-222, 1984. 22. Wall JC, Jatterji SK, Jeffery JW: Age related
3. Pritchard JJ: The osteoblast. In Bourne GH (ed): changes in the density and tensile strength of
The Biochemistry and Physiology of Bone, vol. 1. human femoral cortical bone. Calc Tissue Int
New York, Academic Press, 1972, pp 21-43. 27:105-108, 1979.
4. Carter DR, Spengler DM: Mechanical properties 23. McCalden RW, McGeough JA, Barker MB, Court-
and composition of cortical bone. Clin Orthop Brown CM: Age-related changes in the tensile
135:192-217, 1978. properties of cortical bone. J Bone Joint Surg
5. Schaffler MB, Burr DB: Stiffness of compact 75(A-8): 1193-1199, 1993.
bone: Effect of porosity and density. J Biomech 24. Burstein AH, Reilly DT, Martens M. Ageing of
21:13-16, 1988. bone tissue: Mechanical properties. J Bone Joint
6. Evans FG: Mechancial Properties of Bone. Surg 58A:82-86, 1976.
Springfield, IL, Charles C Thomas, 1973. 25. Carter DR, Hayes WC: The compressive
7. Kraus H: On the mechanical properties and behaviour of bone as a two phase porous
behaviour of human compact bone. In Levine SN structure. J Bone Joint Surg 59A:954-962, 1977.
(ed): Advances in Biomedical Engineering and 26. Bonfield W, Li CH: Deformation and fracture of
Medical Physics, vol 2. New York, Interscience, bone. J Appl Physiol 37:869-875, 1996.
1968, pp 169-204. 27. Piekarski K: Fracture of bone. J Appl Physiol
8. Swanson SAV: Biomechanical characteristics of 41:215-233, 1970.
bone. In Kenedi RM (ed): Advances in 28. Moyle DD, Welbom JW, Cooke FW; Work to
Biomedical Engineering, vol 1. New York, fracture of canine femoral bone. J Biomech
Academic Press, 1971, pp 137-187. 11:435-440, 1978.
9. Currey JD: The mechanical properties of bone. 29. Evans FG, Bang S: Physical and histological
Clin Orthop 73:210-231, 1970. differences between human fibular and femoral
10. Herrmann G, Liebowitz H: Mechanics of bone compact bone. In Evans FG (ed): Studies on the
fracture. In Liebowitz H (ed): Fracture: An Anatomy and Function of Bone and Joints.
Advanced Treatise, vol 7. New York, Academic Heidelberg, Germany, Springer-Verlag, 1966,
Press, 1972, pp 771-840. p 142.
11. Dempster WT, Liddicoat RT: Compact bone as a 30. Dempster WT, Coleman RF: Tensile strength of
non isotropic material. Am J Anat 91:331-62, bone along and across the grain. J Appl Physiol
1952. 16:355, 1961.
12. Reilly DT, Burstein AH, Frankel VH: The elastic 31. Behiri JC, Bonfield W: Crack velocity dependence
modulus for bone. J Biomech 7:271, 1974. of longitudinal fracture in bone. J Mater Sci
13. Reilly DT, Burstein AH: The elastic and ultimate 15:1841, 1980.
properties of compact bone tissue. J Biomech 32. Wright TM, Hayes WC: Fracture mechanics
8:393, 1975. parameters for compact bone: Effects of density
14. Bonfield W, O’Connor P: Anelastic deformation and specimen thickness. J Biomech 10:419-430
and friction stress of bone. J Mater Sci 1977.
13:202-207, 1978. 33. Brown W, Srawley J: Plain strain crack testing of
15. Pope MH, Outwater JO: The fracture high strength metallic materials. ASTM-STP-410.
characteristics of bone substance. J Biomech Philadelphia, ASTM, 1966.
5:457-465, 1972. 34. Behiri JC, Bonfield W. Fracture mechanics of
16. Burstein AH, Zika JM, Heiple KG, Klein L: bone: The effects of density, specimen thickness
Contribution of collagen and mineral to the and crack velocity on longitudinal fracture. J
elastic-plastic properties of bone. J Bone Joint Biomech 17:25-34, 1984.
Surg 57a:956, 1975. 35. Bonfield W, Behiri JC, Charalambides B.
17. McElhaney JH: Dynamic response of bone and Orientation and age-related dependence of the
muscle tissue. J Appl Physiol 21:1231-1236 fracture toughness of cortical bone. In Perren SM,
1966.
Schneider E (eds): Biomechanics: Current
18. Bonfield W, Clark EA: Elastic deformation of
Interdisciplinary Research. Dordrecht, The
compact bone. J Mater Sci 8:1590-1594, 1973.
Netherlands, Martinus Nijhoff Publishers, 1985
19. Currey JD: The effects of strain rate, pp 185-189.
36. Seireg A, Kempke W: Behaviour of in vivo 39. Carter DR, Hayes WC: Compact bone fatigue
bone under cyclic loading. J Biomech damage: A microscopic examination. Clin Orthop
2:455-462, 1969. 127:265-274, 1977.
37. Swanson SAV, Freeman MAR, Day WH: The 40. Carter DR, Hayes WC: Fatigue life of compact
fatigue properties of human cortical bone. Med bone. I. Effects of stress amplitude, temperature
Biol Eng 9:23-32, 1971. and density. J Biomech 9:27-34, 1976.
38. Carter DR, Hayes WC: Compact bone fatigue 41. Carter DR, Hayes WC, Schurman DJ: Fatigue life
damage. I. Residual strength and stiffness. J of compact bone. II. Effects of microstructure and
Biomech 10:325-337, 1977. density. J Biomech 9:211-218, 1976.
.. '
.
■
I
f
Biomechanics of Muscle
Zeevi Dvir
Motion is perhaps the most conspicuous sign of • Muscle action vector and moment arm
life—so much so that someone described as lying • Muscle moment and its relationship to joint
completely motionless is presumed to be either angular position
unconscious or dead. In the latter case, however, • Statically determinate systems
examination of other movement-related parame¬ • Statically indeterminate systems and common
ters, such as pulse, would be required. These spe¬ methods of solution
cial manifestations, in particular the full spectrum
of normal or pathologic biomotion, from simple
reflexes to stereotyped activities to the most com¬ Muscle Structure and
plex intentional movements, are the exclusive re¬ Architecture
sult of the capacity of muscles to generate tension.
From a control systems point of view, skeletal
Both of these general aspects of skeletal muscle
muscle presents a unique feature in that it operates
are now fairly well documented. In this text, struc¬
as both an effector and a sensor. In other words,
ture refers to the arrangement of the morphologic
it possesses the dual abilities of producing tension,
units that consist, in order of decreasing size, of
on one hand, and detecting and transmitting sig¬
the muscle as an integral unit, the compartment
nals relevant to its functioning, on the other. More¬
(where applicable), the fascicle, the muscle fiber,
over, as discussed in Chapter 5, some muscles
the myofibril, and the sarcomere, which contains
may be intended for positional sensing more than
the basic contractile proteins myosin and actin.
for the actual sharing in supporting the load. The
Muscle architecture refers to the pennation type,
dual abilities of skeletal muscle are without com¬
fiber length, and physiologic cross-sectional area
parison among the other sense organs. Consider, (PCSA).
for instance, production and sensing of sound; the
former is the responsibility of the vocal cords,
whereas the latter is that of the ear; neither organ STRUCTURE
can fulfill both roles.
The subdivisions of skeletal muscle are depicted
This distinguishing quality is but one of many
in Figure 4-1. Within the muscle proper, an intri¬
that have rendered skeletal muscle one of the most
cate latticework of connective tissue enables parti¬
researched organ tissues in the body. Among the
tioning of the contractile machinery into the pre¬
various fields of muscle research, biomechanics
viously mentioned units. This tissue is made up
naturally occupies a special niche. Consequently,
predominantly of collagen fibers enmeshed with a
survey of all related aspects would be beyond the
relatively much smaller number of elastin fibers.
scope of this chapter; instead, focus is given to a
As seen in Figure 4-2, the epimysium covers the
number of pertinent aspects, including the follow¬
integral unit and invades the inside of the muscle,
ing:
connecting with the perimysium, which subdivides
• Muscle structure and architecture the muscle into fascicles. Lining the fibers is the
• Mechanism of tension generation and endomysium, a dense mesh of delicate collagen
transmission fibers that may provide a physical continuation
• Force-length (F-L) relationship between the perimysium on one side and the outer
• Force-velocity (F-V) relationship (basement) membrane of the fiber on the other.1
83
Terminal Transverse Sarcoplasmic Sarcolemma

cistern tubules reticulum (muscle fiber membrane)
-/■ \
Myofibrils
Filament Mitochondrion
H ' A I /
Band Z Band Band
I—| disk
Myofibril
B
Thin filament (F-actin)
Figure 4-1 A, Three-dimensional reconstruction of a sector of a muscle fiber. B, Individual myofibril

showing the dark bands that correspond to the regions of actin-myosin overlap. C, Schematic cross-
section of an individual sarcomere. Note the stem and globular double-head configuration of the
myosin molecule. (From Kandell EC, Schwartz JH, Jessel TM: Principles of Neural Science. Englewood
Cliffs, NJ, Prentice Hall, 1991. Reproduced with permission of The McGraw-Hill Companies.)
In the context of gross structure, some muscles sarcomere, which consists of parallel actin and
may consist of several compartments divided by myosin filaments. Under the electron microscope,
transverse fibrous bands. This means that bundles a finer subdivision of each sarcomere into several
of short fibers may be arranged in series,2 so that bands is visible (see Fig. 4-1). The centrally lo¬
the overall shortening capacity and, hence, variety cated dark A band corresponds to the presence of
of contraction may be considerably enhanced. myosin. Two less refractory I bands correspond to
Each compartment has its own nerve supply and the presence of actin and insert into the thin Z
is centrally represented. In humans, the sartorius disks, which also serve as the borders of the sar¬
has four compartments, whereas the gracilis has comere. In a noncontracting muscle, the H zone,
two.3 To what extent this layout depends on the which occupies the middle portion of the A band,
complexity of function or on the sheer size of the corresponds with the nonoverlapping region of the
muscle is not absolutely clear. As suggested by filaments. This region also contains a thin central
Windhorst and colleagues,4 however, the control M region, a latticework of filaments supporting
of some muscles as single units poses significant the myosin filaments while maintaining an inter¬
problems that could be solved more effectively filament distance. This support system, which con¬
using compartmentalization. sists of a number of proteins,1 is essential for
The basic functional unit of the muscle is the structural integrity as well as for the proper func-
I Chapter 4 BIOMECHANICS OF MUSCLE 85
FIGURE 4-2 Schematic cross-

section of a muscle showing the
three types of connective tissue
sheath. (From McComas AJ: Skeletal
Muscle Form and Function.
Champaign, IL, Human Kinetics,
1996. Courtesy of Dr. John
Maguire.)
tioning of the sliding mechanism. In addition, be¬ determine, to a considerable extent, the force and
cause of their intimate relationship with the con¬ velocity profile of the individual muscle. However,
tractile filaments, these noncontractile proteins the individual architectural profile of a given mus¬
contribute to the stiffness properties of the muscle. cle should not be confused with the motion profile
The arrangement of the actin and myosin fila¬ of the segment it acts on. As in most instances,
ment is revealed by the electron microscope. In motion depends on the activity of a number of
principle, each myosin filament is surrounded by muscles. For instance, the soleus and gastrocne¬
a hexagonal array of six actin filaments (Fig. 4—3). mius form the core of the plantar-flexor group, yet
About 250 myosin molecules make up 1 myosin architecturally, they are quite different.7
filament whose length is about 1.6 nm. These In humans, skeletal muscle fibers may reach a
molecules are distinguished by a Z-disk-oriented length of up to 160 mm (the semitendinosus), but
head and M-region-oriented tail portions, known in most limb muscles, fiber length ranges typically
as heavy and light meromyosin, respectively. between 40 and 100 mm.8'10 The distribution of
Along the major axis of the myosin filament, the fiber length, together with several other architec¬
heads of the myosin molecules, which act as the tural parameters in selected lower extremity mus¬
binding site during the cross-bridge formation, are cles, is outlined in Table 4-1. Data for most mus¬
arranged in pattern. The basic block is of six cles are based on three cadavers and are presented
heads, configured as three opposing heads lying at in terms of the range (in parentheses) and aver¬
an offset angle of 60 degrees to each other. The age values.8
distance between each pair is 14.3 nm.5 Thin fil¬ Assuming that all sarcomeres have the same
aments are made up of braided polymerized actin length and length variation (shortening, lengthen¬
monomers and have a length of 1.27 nm.6 Inter¬ ing), longer fibers, which by definition have a
twined between them are actin-binding tropomyo¬ larger number of sarcomeres, also possess a larger
sin and troponin molecules. absolute range of excursion. It also follows that
because the maximum contraction velocity of a
ARCHITECTURE muscle depends on the number of sarcomeres,
Among the architectural parameters, fiber length longer fibers can produce higher velocity. On the
and PCSA are the most essential because they other hand, shorter fibers are characterized by a
i<-
I H zone
w !
1 i
FIGURE 4-3 Actin-myosin

hexagonal array. (From Huxley HE,
in Bourne GH: The Structure and
Function of Muscle. Copyright ©
1972, by Academic Press,
Thick filament Thick filament reproduced by permission of the
(myosin) (actin) publisher.)
more dominant ability to produce force. In a study and ankle plantar-flexors are primarily designed
of the specific mechanical nature of long and for the generation of high forces, whereas the knee
short fibers in an animal model, Woittiez and flexors and ankle dorsiflexors are more suitable
associates11 have proposed the use of a ratio for greater excursions.7
termed index of architecture (ia), which is equal to Another essential architectural parameter is the
the fiber length divided by the muscle belly length. PCSA of the muscle. The importance of the PCSA
Figure 4-4 depicts the major trends based on lies in the fact that it is linearly related to the
modeling and experimental findings that reveal maximal force output of the muscle under isomet¬
that the ia is positively and linearly related to the ric conditions and at optimal length. Illuminating
maximal velocity of contraction but negatively findings come from a study of human cadaveric
related to the maximal force produced. In princi¬ material in which a number of architectural param¬
ple, the situation with human muscles may be eters were measured in muscles that compose the
similar. The value of ia for the tibialis anterior lower limb.8 Figure 4-5 depicts some of these
muscle is 0.26, compared with 0.14 for the medial findings, which clearly support the general dichot¬
gastrocnemius, whereas that of the semimembra¬ omy into predominantly “force” or “velocity”
nosus is 0.24, compared with 0.17 in the vastus muscles. By way of example, the gracilis and
lateralis. These variations are reflected in what is sartorius have the lowest PCSA-to-muscle weight
perceived as the more dominant functional feature ratio but the highest ia. Although a geometric
of these muscles, namely, that the knee extensors property, estimation of the PCSA factors in the
TABLE 4-1 Architectural Features of Selected Lower Extremity Muscles
Femur Pennation
Muscle Fiber Length Length-to-Muscle Angle
Muscle Length (cm) (cm) Length Ratio (%) (degrees) PCSA (cm2)
Rectus femoris 31.6 (30.5-32.4) 6.6 (6.3-6.8) 21 (21-21) 5 (5-5) 12.7 (8.9-15.2)
Vastus lateralis 32.5 (30.8-35.3) 6.6 (6.4-67) 20 (19-22) 5 (5-5) 30.6 (21—42.9)
Vastus rnedialis 33.5 (31.4—36.3) 7.3 (6.4-7.5) 21 (20-22) 5 (5-5) 21.1 (13.3-28.0)
Semimembranosus 26.2 (26.0-26.5) 6.3 (5.4—7.0) 24 (21-26) 15 (10-20) 16.9 (13.9-18.6)
Biceps femoris* 34.2 (32.8-36.9) 8.5 (7.8-9.5) 25 (21-29) 0 (0-0) 12.8 (7.4-16.4)
Semitendinosus 31.7 (31.3-32.1) 15.8 (15.6-16.0) 50 (50-50) 5 (5-5) 5.4 (4.4-6.3)
Sartorius 50.3 (45.8-55.2) 45.5 (41.9—48.2) 90.6 (87-93) 0 (0-0) 1.7 (1.1-2.1)
Adductor magnus 30.5 (28.8-32.7) 1 1.5 (10.6-13.1) 38 (36-40) 0 (0-0) 18.2 (13.6-20.8)
Adductor longus 22.9 (21.0-25.2) 10.8 (10.5-1 1.2) 48 (43-50) 6 (5-8) 6.8 (4.0-10.5)
M. gastrocnemius 24.8 (23.7-26.8) 3.5 (3.2-3.9) 14.3 (13-16) 16. 6 (15-25) 32.4 (29.3-38.7)
L. gastrocnemius 21.7 (19.5-22.9) 5.1 (4.0—5.9) 23.3 (21-26) 8. 3 (5-10) —
Soleus 30.9 (30.8-31.1) 1.95 (1.9-2.0) 6 (6-6) 25 (20-30) 58.0f

Tibialis anterior 29.8 (28.4-32.2) 7.7 (6.9-9.3) 25.6 (24-29) 5 (5-5) 9.7 (7.9-12.7)
Peroneus brevis 23 (20.8-25.2) 3.9 (3.4—4.6) 17 (16-18) 5 (5-5) 5.7 (4.5-77)
Peroneus longus 28.6 (25.8-31.8) 3.9 (3.3—4.4) 13.7 (12-16) 10 (10-10) 12.3 (7.4-17.3)
*Long (ischial) head.

fSingle measurement.
PCSA, physiologic cross-sectional area.
mass of the muscle belly and its density, as fol¬ effect of pennation angle on the force transmitted
lows: by the muscle to the tendon, consider Figure 4—6,
in which a is the angle of pennation. In this
PCSA = (mcosa) / (fiber length X muscle density) configuration, the force developed in the tendon
(F) and the force developed in the muscle (FM)
where m is the mass of muscle belly, a is the are related as follows:
angle of pennation, and the muscle density is
1.056 kg-3.12 There is some controversy regarding F = FMcosa
the representative value of the slope of the above
linear relationship. Generally speaking, it has been Also note that the total length of the musculo¬
estimated to range from 20 to 35 N/cm243 but tendinous unit (Lmt) is computed as follows:
factors such as fiber type distribution and its asso¬
ciated biochemical variations may have resulted Lmt = Lr + LMct
in quotation of different values. For instance,
whereas Gregor14 quotes 22.5 N/cm2 as the value, where Ma is the “co-linear” muscle length. As¬
Pierrynowski15 suggests 35 N/cm2 as more reason¬ suming that w, the width of the unit, remains
able. constant, a is 30 degrees, and maximal fiber short¬
The third architectural factor, pennation, relates ening is 50%, a situation may be reached in which
to the arrangement and direction of the fibers in the fibers are orthogonal to the tendon and there¬
terms of the insertion regions. In humans, most fore cannot transmit any force. However, both
muscles have their fibers oriented along the longi¬ conditions may not be realized in the human body.
tudinal axis, resulting in the so-called fusiform Pennation angles rarely exceed 25 degrees,8
type. In this context, note that the “longitudinal whereas full shortening of the fibers is hampered
axis” does not necessarily imply a straight line by external constraints, such as the joint range
between the insertions, because, as discussed later of motion.
in this chapter, muscles may more accurately be
modeled as piecewise linear actuators.15 The other
way fibers may be configured is feather-like. Thus, Mechanism of Tension Generation
in pennated muscles, fibers are essentially at an and Transmission
angle to the major axis. Variants of this include the
unipennate, bipennate, and multipennate muscles. The prevailing model for muscular tension genera¬
Clearly, the higher the index of architecture, the tion is based on a series of studies by Andrew
lower the pennation angle. To gain insight into the Huxley and Hugh Huxley, begun almost half a
Maximum
velocity/force
Velocity
Force
A 0 1.0
Angle(°)
Fiber length
(%)
FIGURE 4-4 A, Relationship between the index of

Thickness architecture (ia) and maximal velocity of shortening and
(%) maximal force. B, Measured values of some morphologic
parameters of three medial gastrocnemius muscles (asterisk,
125 - triangle, circle) in combination with constructed (model) values
relative to 3 ias. The goodness-of-fit of the model is indicated
100 -
by high correlation coefficients with the raw data: in a, r =
0.96; in b, r = 0.99; and in c, r = 0.88. (A and B, Adapted
75 -
from Woittiez RD, Huijing PA, Boom H, Rozendal RH: A three
dimensional muscle model: A quantified relation between form
and function of skeletal muscles. J Morphol 182:95-113, 1984.
—i-Length (%)
Copyright © 1984. Reprinted by permission of Wiley-Liss, Inc.,
B 75 100 125
a subsidiary of John Wiley & Sons, Inc.)
Fiber length (mm)

SR GR
350
180
J
A
-
160 ■
ST
140 -
120 -
AM
AL
PC AB
100 -
FIGURE 4-5 Fiber lengths. Physiologic

cross-sectional area (PCSA) in lower limb EH BF
muscles. SR, sartorius; GR, gracilis; ST,
80 - ED
semitendinosus; AM, adductor magnus; AL, TA
adductor longus; AB, adductor brevis; PC, VM
VI
pectineus; EH, extensor hallucis longus; ED, RF VL
SM
extensor digitorum longus; TA, tibialis anterior;
60 -
BF, biceps femoris; RF, rectus femoris; SM,
semimembranous; VM, vastus medialis; VI,
vastus intermedius; VL, vastus lateralis; PT,
plantaris; PB, peroneus brevis; PL, peroneus 40 - PT PB PL
MG
longus; FH, flexor hallucis longus; FD, flexor FH
digitorum longus; PO, popliteus; TP, tibialis PO
FD TP SL
posterior; MG, medial gastrocnemius; SL,
soleus. (Data from Wiekiewicz Tl, Roy RR, 20 I —l- —i- Hh
Powell P, Edgerton VR: Muscle architecture of 500 1000 1500 2000 2500 3000 3500 6000
the human lower limb. Clin Orthop
179:275-283, 1983.)
PCSA (mm^)
century ago.16-18 According to this model, the pro¬ hence, the name of the model is the sliding fila¬
duction of force is effected by a sequential forma¬ ment theory. The model predicts that because each
tion of cross-bridges between the respective active cross-bridge acts as an independent force genera¬
sites of actin and myosin filaments. This process tor, the sliding, which introduces a linearly grow¬
translates into the relative sliding of the filaments; ing number of cross-bridges, would, within a cer¬
tain excursion, be reflected in a corresponding
increase in the total force produced by the fiber.
This hypothesis was indeed validated by Gordon
and colleagues19 in an experiment whose major
results are depicted in Figure 4—7. It was indicated
that with sarcomere length exceeding 3.65 |xm,
filament overlapping could not be reached, and
hence no tension was produced. As the fiber ends
were allowed to approximate, tension increased
FIGURE 4-6 Geometric relation between the pennation
linearly, reaching a peak at a sarcomere length
angle and other linear length parameters in muscle. FT, force
between 2 and 2.25 |xm. The length (L0) at which
developed in tendon; LT, tendon’s length; LMT, total length of
musculotendinous unit; LMo, colinear muscle length, w,
peak force (P0) was reached corresponded to opti¬
tendon-to-tendon vectorial distance. mal filament overlapping. Further approximation
Sarcomere length (pm)

pertinent points should be highlighted, however.
0.4 3.5 3.0 2.5 2.0 1.5 1.0
i_i_i_i_i_i_i First, actin and myosin filaments have a strong
chemical affinity that, in the absence of adenosine
triphosphate (ATP), results in their binding to¬
gether.22 The actin-myosin complex can therefore
be described as a physiologic “default” situation,
characteristic of such phenomena as rigor mortis.
As soon as ATP is available on site, however, it
enters a specific segment of the myosin molecule,
where it causes weakening of the actin-myosin
complex. Second, immediately after, the ATP en¬
trenches within this segment, enabling a detach¬
ment of the complex, a relative movement of 5 N
m, and reattachment. Third, after a series of chem¬
ical reactions, the myosin, through its head, per¬
forms the power stroke and advances another 5
N m. Fourth, after other reactions, the cycle is
concluded by freeing the ATP site and reformation
of the actin-myosin complex. Experiments have
shown that the total unresisted excursion covered
within one such cycle, which lasts about 50 ms,
is on average 11 N m.23 The actual binding time,
however, is 2 ms. This, of course, must be inter¬
preted in terms of whole muscle, as should the
force generated by a single myosin head: 3 to 4
X 10_12N.
Once generated within its fibers, muscle tension
is transmitted to the bones. This is normally done
FIGURE 4-7 Relationship between sarcomere length and through tendons, and owing to the passive proper¬
the relative tension it develops. (Data from Gordon AM, ties of the latter, consideration is given to the
Huxley AF, Julian FJ: The variation in isometric tension with
combined musculotendinous unit. When muscle
sarcomere length in vertebrate muscle fibers. J Physiol
exerts force on tendons, a deformation results that
184:170-192, 1966.)
reflects two independent properties: one that is
structural and another that is purely mechanical.
Structurally, tendons are made up of a woven
resulted in a reduction of force, probably owing mesh of collagen fibers, which under no tension
to electrostatic repulsion between adjacent and have a wavy appearance. Mechanically, tendons
opposing actin filaments. In vivo studies indicate are viscoelastic materials, but the elastic compo¬
that the length-tension behavior of whole muscle nent is by far the more dominant.24 Thus, when
is similar, provided that the involved muscle’s force acts on the tendon, the initial effect is that
lever arm does not change during movement of of straightening out the waviness (slackness). Suf¬
the mobile relative to the stationary segment. A ficiently high force further results in elongation of
good approximation of the in vitro findings may the fibers (deformation). This elongation is lin¬
be found in a study by Sale and associates,20 in early proportional to the stress (force/unit area of
which isometric movement records of calf muscles tendon material); that is, the stiffness of the tendon
under various stimulation conditions were em¬ is kept almost constant with a tangent modulus of
ployed. These relationships may also be derived 1.2 giga-pascal (GPa).25 From the functional point
from an isokinetic strength curve by dividing the of view, tendon elasticity may serve a protective
resulting moment by the corresponding length of role: momentary loads exerted by the muscle are
the lever arm.21 Further discussion of the length- not transmitted directly to the bone. Even more
tension relationship is provided later in this chap¬ important, however, is the fact that energy dissipa¬
ter.
tion in the tendons, as indicated by the area inside
A detailed description of the activation and de¬ the hysteresis loop, is small.26 This means that
activation of myosin-actin cross-bridges is beyond tendons manifest a high degree of elastic recoil,
the scope of this chapter, and interested readers that is, that energy injected into the system may
are referred to the work by McComas.1 Some in part be recovered.
• I Chapter 4 BIOMECHANICS OF MUSCLE 9!
Force-Length Relationship relationship takes the form of an inverted U shape

(see Fig. 4-7). When a whole muscle specimen is
The F-L relationship is one of the most discussed considered, it is predominantly the effect of the
in muscle mechanics. As the term implies, the passively resisting viscoelastic elements that con¬
main issue is the way in which muscle force varies tributes to the change in the total force recorded
with its length under zero-velocity (isometric) and the way it varies with the muscle’s length
conditions. A general Hill-type model is com¬ (Fig. 4-8). In vivo analysis of the F-L relationship
monly used to explain the major mechanical inter¬ in humans is, of course, a rarity. Measurements
actions. In this model, which precludes the tendon, carried out on some arm muscles in volunteer
the active contractile element (CE) operates in a cineplastic amputees, however, have shown that
series with the series elastic element (SEE), which muscles behave in quite the same way.
represents stiffness residing in the cross-bridges Passive muscles whose length is equal to or
and in parallel with the passive element (PE). below the resting length offer little resistance to
However, other than in muscles operating through pull. Once past its resting length, unstimulated
short tendons, or unless contrary to the accepted muscle records increasing resistance to deforma¬
theory,17-27 positive indications for Z-disk extensi¬ tion whose main source could be myofibrillar elas¬
bility exist, and the energy stored in SEE is con¬ ticity.28 This passive resistance does not vary with
sidered to be negligible compared with that stored the level of muscular activation, as may be seen
by the tendon; hence, the SEE may be ignored. in Figure 4-8. On the other hand, a lower activa¬
As previously indicated, muscle fibers reach tion level leads to lower total tension and, conse¬
their maximal force when the sarcomere length is quently, to lower net (active) force. Because the
2 to 2.25 p,m. Moreover, the fiber-specific F-L active force cannot physically be divorced from
a(t) = 1
a (t) = 0.5
FIGURE 4-8 Nominal static and dynamic properties of muscle tissue. A, Isometric force-length (F-L)
relationship with passive (dotted) and active (dashed) components. FM, muscle force; F0, peak force; l0, optimal
length. 8, Less than fully activated muscle develops less force, but the passive component is unaffected. a(t),
level of activation C, Force-velocity (F-V) relationship in fully activated tissue when measured at l0. Vm.
maximal shortening velocity. D, When activated at a level lower than its maximum, muscle may contract at a
lower rate. — vM, maximal velocity.
the passive force, biomechanical calculations of length, meaning the shortest length at which the
joint forces equilibrium refer to the total (active muscle exerts resistance, 83% for the gastrocne¬
plus passive force) as the muscle force. mius and 71% for the semimembranosus, and the
The shape of the F-L relationship depends on passive stiffness in relation to the muscle optimum
the architectural properties of fiber length and length, which in the gastrocnemius is almost twice
pennation. Two variants of the F-L curve, which that in the semimembranosus.
are based on an animal study,11 are shown in The above parameters may vary as a result of
Figure 4-9, where the semimembranosus (SM, ia length changes imposed on skeletal muscle. Using
= 0.7) is compared with the gastrocnemius (GM, an animal model, Williams and Goldspink29 were
ia = 0.3). The active tension of both muscles is able to shift the F-L curve, this shift being depen¬
of the inverted U shape form, where the maximal dent on the age of the animal. Compared with the
value is reached at the normalized length of 100%. typical control curve, the F-L relationship, after
Likewise, in both, the passive resistance can be imposed lengthening, shifted to the right in adult
closely approximated by an exponential curve. On animals and to the left in young people (Fig.
the other hand, the falloff toward both ends is 4-10A and B, respectively). As evident from the
much sharper in the gastrocnemius owing to its adults’ curves, although the L0-based active force
greater pennation and hence significantly more was higher in the lengthened position than in the
limited range of excursion. This difference affects control group, at an original muscle length of 95%
the normalized length at which active muscle L0 and below, it was lower than that measured
force can be produced, which is greater in the in the control group. Gossman and associates30
gastrocnemius. It is also apparent with regard to speculated that this finding, if allowed to be ap¬
other interconnected parameters: the active slack plied in live human muscle, would explain the
Force (%)
Length (%)
Force (%) 0.35
Figure 4-9 Force-length relationship in semimembranosus (SM) and medial gastrocnemius (GM).
Numbers on the curves refer to the index of architecture. (From Woittiez RD, Huijing PA, Boom H,
Rozendal RH: A three dimensional muscle model: A quantified relation between form and function of
skeletal muscles. J Morphol 182:95-113, 1984. Copyright © 1984. Reprinted by permission of Wiley-Liss
Inc., a subsidiary of John Wiley & Sons, Inc.)
Chapter 4 BIOMECHANICS OF MUSCLE 93
Tension (gram unit) is highlighted in some pathologic states amenable

to intervention.30 In the present context, special
reference is made to the treatment of children with
cerebral palsy using tone-reducing casts. The aim
of using the cast is to elongate hypoextensible
muscles to improve function. This intervention is
effected by imposing a neutral foot position where
equinus exists and then firmly casting for a period
of several weeks. Assessment of improvement
may be conducted using a force or movement
measurement instrument applied at the same ankle
joint positions before and after casting. Tardieu
and associates32 indicated that patients whose mus¬
cles could adapt well to the new imposed length
Tension (gram unit)
obtained better compliance (passive torque) after
cast removal.
Force-Velocity Relationship
The F-V relationship is the second mechanical

function of skeletal muscle that has been the focus
of numerous studies over the years. Similar to
the studies concentrating on the F-L relationship,
0 60 70 80 90 100 110 120
research has been directed at the muscle fiber
B Muscle length of control (%)
level, at a whole muscle under stimulation, or at
FIGURE 4-10 A, Length-associated changes in adult animal muscle in the live body. In the last case, the
muscles following immobilization in lengthened and shortened measured quantity in in vivo studies is the torque
positions relative to control. B, Length-associated changes in rather than the actual force developed by the mus¬
young animal muscles following immobilization in lengthened
cle.
and shortened positions relative to control. (A and 8,
The general description of the F-V relationship
Adapted from Williams PE, Goldspink G: Changes in
was intensively explored by Hill in the late 1930s.
sarcomere length and physiological properties in immobilised
muscle. J Anat 127:459-468, 1978. Reprinted with the
The experiments were based on maximal stimula¬
permission of Cambridge University Press.) tion of a muscle that was allowed to shorten or
lengthen against a series of discrete loads, render¬
ing the situation truly isotonic. The graphical pre¬
sentation of the findings (see Fig. 4-8) clearly
phenomenon of stretch weakness, namely, the ten¬
evidences two branches: the concentric and the
dency of muscles that are stretched because of
eccentric. During the concentric contractions, F
postural malalignment to test stronger at the new
and V were inversely related according to the
L0 but weaker at the natural length. However
following equations:
tempting this theory is, because of the variation in
the muscle’s lever arm and because of inherent v = b(F — F0)/F + a)
methodologic problems, conclusive proof of this
speculation with respect to active muscle contrac¬ where v is muscle velocity, F0 is maximal muscle
tion may not be possible. In young people, length¬ force at isometric conditions and optimal sarco¬
ening of muscle results in its actual weakening. mere length, F is muscle force, and a and b are
The difference may be due to the different mecha¬ constants derived experimentally. By rearranging,
nism operating: in adults, there is an increase in this hyperbolic relationship could be solved for F
the sarcomere number, whereas in children, it is as follows:
the tendon that grows disproportionately in length,
placing the muscle belly in a shortened position, F = (F0b — av)/(b + v)
which in turn leads to a reduction in the sarcomere
length and total force.29'31-32 When shortening is Likewise, one may obtain the maximal velocity
imposed in animals, the result in both adult and (v0) by setting F = 0, which leads to the following
young animals is a reduction in the number of equation:
sarcomeres and the force produced.
The clinical significance of the F-L relationship v0 = bFo/a
Power, the rate at which work is done by muscle Muscle action vector and
(dW/dt), may also be derived from the previous Moment Arm
equations because W = Fds and therefore dW/dt
— Fds/dt = Fv. In Figure 4-11, the power curve
is obtained by multiplying the velocity by the Analysis of muscle action in terms of the forces
respective force value. Power reaches its peak it generates and the reactions produced in the
at about 0.3vo, and its value there is equal to joints and associated structures necessitates further
about 0.1F0v0.5 information that cannot be obtained from the pre¬
These relationships do not apply in the eccentric viously described physiologic and mechanical pa¬
branch of the curve. In vitro findings demonstrate rameters: muscle and fiber length, muscle belly
that the force produced is generally independent versus tendon, angle of pennation, and PCSA.
of the velocity at which the muscle lengthens and Specifically, because the dominant expression of
that theoretically its maximum may not be higher muscle activity is rotational motion in the joints it
than 1.8F0. The former observation is amply sup¬ spans, knowledge of factors such as the instanta¬
ported by in vivo (isokinetic) studies of eccentric neous direction of the force vector and the mo¬
contractions. ment arm vector is essential. To determine these
Similar to the passive resistance to elongation, factors, the human body must be represented spa¬
characteristic of the F-L curve, in the F-V relation¬ tially as a system of rigid links (segments) whose
ship there is a velocity-dependent response that positions relative to each other are known. Once
reflects the inherent viscoelastic makeup of mus¬ link position is known, a process of scaling is
cles and is therefore not related to a reflexive undertaken in which fitting of the so-called norma¬
component. Furthermore, of definite clinical sig¬ tive databases to the patient’s anthropometric pa¬
nificance is the question of whether velocity sensi¬ rameters (height, weight, skin-fold thickness, so-
tivity varies in patients with chronic spasticity. A matotype, and segmental length and girth) is
recent study3’ compared the nonreflexive resistive performed. The location of the insertion regions
torque (NRT) of the plantar-flexors in spinal cord of the muscles (and other structures) relative to
injury patients and in control subjects using isoki¬ the links can then be calculated.
netic dynamometry. It was revealed that in both Although ostensibly a straightforward proce¬
groups, NRT increased positively with the veloc¬ dure, scaling is not without some severe limita¬
ity, which was varied over a range of 5 to 180 tions.15 For example, cadaveric material may not
degrees/s. Moreover, although the NRT was gener¬ be sufficiently representative because of the actual
ally larger in patients than in control subjects, it preservation method, and because cadavers are
tended to plateau in the former group at 60 de¬ normally older, their dimensions may not be di¬
grees/s, whereas a proportional linear trend was rectly applicable to the younger population. Be¬
maintained in the control group throughout the cause no better approach is available, the largely
tested velocity spectrum. It was suggested that accepted theory of elastic similarity deems that
spasticity and disuse could alter the tensile proper¬ scaling is basically a linear process, measuring
ties of muscle and that this change should be increase or decrease proportionately in all dimen¬
accounted for, namely, by avoiding rapid vigorous sions.34’ 35
and intensive stretching techniques. There are two major issues concerning the di-
Moment
Power
Con
^ • « Angular velocity
FIGURE 4-11 Schematic strength and power relationships. Con, in concentric contraction; Ecc, in
eccentric contraction.
• I Chapter 4 BIOMECHANICS OF MUSCLE 95
rection of the muscle action vector. One is associ¬ where MA is the corrected moment arm, r(ijk) is
ated with the fact that in many cases, muscles the distance from the disk centroid, F(ijk) is the
connect insertion areas rather than insertion points. unit vector of the fiber’s line of action, and U(ijk)
The other, which is more difficult to tackle, relates is the unit vector of the orthopedic axis. However,
to the curvature, or pulley design, of muscles, that although magnetic resonance imaging provides an
is, that some of them cannot be treated as straight accurate platform for measuring both the disks
lines and hence their action vector does not simply and muscles, this method depends on the position
point from the origin to the insertion (or vice assumed by the subject during imaging. Moreover,
versa). As for the former issue, the common as pointed out by McGill and colleagues,37 lumbar
method is to represent the insertion area by its muscle boundaries are not easy to delineate, nei¬
centroid, which is basically the geometric average ther with respect to other soft tissues nor with
of the area, for example, the center of a circle or respect to adjacent muscles.
an ellipse. With regard to the latter, a few ap¬ On a different level and with respect to these
proaches have been adopted for representing the muscles, it would be erroneous to assume that the
action vector. In the simplest case, muscle may be force contributed by individual muscles can be
represented as a straight line using the scaled determined using a single section (lumbar level)
coordinates of the origin (X°, Y°, Z°) and insertion because the PCSA of some muscles can change
(X1- Y1' Z1) to obtain the direction cosines of the dramatically from one vertebral level to another.
vector, its length, and its describing equation. The These and other factors render the derivation of
equation may then be employed to solve for the an exact moment an almost impossible task in
instantaneous length of the moment arm given the some cases. In cases in which the geometry of the
coordinates of the instantaneous center or axis of muscles and joints concerned is less complex,
rotation of the relevant joints. however, reasonable accuracy may be established,
In a detailed analysis of 47 human lower ex¬ leading to improved prediction of the force and
tremity muscles, more than half were represented load-sharing mechanisms operating in a specific
by nonjointed straight lines.15 Prominent among joint territory.
these were various components of the gluteus min¬
imus and medius. On the other hand, a more
complex course indicated by a pronounced change
Muscle Moment and Its
of direction, at least once, required three anchor
Relationship to Joint Angular
points (two sections). One such example is the Position
superficial part of the gluteus minimus. Highly
Muscle moment, which is obtained by the vector
deflected muscles such as those acting on the toes
product of the muscle force and its moment arm,
(the extensor and flexor digitorum longus and the
is the mechanical equivalent of the physiologic
extensor and flexor hallucis longus) required no
strength. This is a typical in vivo entity as it
less than six anchor points and hence five linear
relates to the turning effect of the muscle when it
segments for their representation.
operates on the joints it spans. Both the force and
If the muscle has a truly curved trajectory, curve
the moment arm vary with the change in the
fitting better serves the purpose. This may be done
angular position of the joint. Although variation
using parametric cubic polynomials, namely, of
in the former is largely dictated by the F-L rela¬
the form F(m) = am + bm2 + cm3 + d, where
tionship, the latter is a function of the degree of
m is the parameter and the values of a, b, c, and
sliding, spinning, and rolling taking place in the
d are related to two points and their corresponding
joint. These distinct types of motion are in turn
tangent vectors through which the muscle passes.15
a result of a number of elements, such as the
In vivo estimation of action vectors currently em¬
ligamentous-capsular configuration and its laxity,
ploys imaging techniques and is based on sequen¬
congruence of the articular surfaces, and the joint¬
tial sections of the specific territory.36 For exam¬
loading profile, all of which manifest intimate
ple, the magnetic resonance imaging-based
angle dependency. Consequently, the moment gen¬
approximated muscle moment arms of the lumbar
erated by the muscle must be angle dependent.
musculature with respect to the disk centroid were The determination of the in vivo L0 is to a large
calculated by McGill and colleagues37 using the extent arbitrary. On the other hand, it is reasonable
following matrix: that for most uniarticular muscles, the ends of the
range of motion are likely to coincide with the
ri rj rk shortest or longest muscle belly length. Because
Fi Fj Fk the F-L relationship is a monotonically decreasing
Ui Uj Uk function from L0 toward the shortest possible
length, and muscles may generally exceed L0 by ited to animal models39 and even then may prove
more than 10%, it is reasonable to expect that the to be a partial solution only. For example, instru¬
muscle moment-joint angular position relationship mentation of the patellar tendon can yield informa¬
will basically reflect variations in the moment arm. tion about the level of tension transmitted by the
The moment-joint angular position relationship heads of the quadriceps but may hardly solve the
obtained in maximal contractions may be derived problem of intervasti force distribution. Im¬
by either static (isometric) or dynamic (low-veloc¬ planting a number of transducers might enhance
ity isokinetic) testing. our understanding but at the same time would
introduce serious technical problems. For many
reasons, employment of electromyography (EMG)
Statically Determinate Systems is still not a viable solution for assessing muscle
forces. Although much less invasive than sensor
Up to this point, several parameters associated implantation, this technology, notwithstanding
with the structure and basic mechanical descrip¬ some major developments, has not yet been per¬
tors of the skeletal muscle have been presented. It fected for purposes of dynamic contraction analy¬
is now time to consider the variable force that sis as much as it is effective for static analysis.
may be generated in the course of purposeful On the other hand, it should be conceded that
contraction. This information is essential in clini¬ EMG may be used effectively for either endorsing
cal instances involving the effect of the force on analytic solutions or guiding the selection of a
the relevant joints, ligaments, capsule, disks, or more appropriate solution set. At any rate, the
muscle itself. For example, the magnitude of bone- governing approach in modern biomechanical
to-bone forces depends intimately on the level of analysis of musculoskeletal problems is by mathe¬
muscular forces. Thus, after joint replacement or matical modeling. Such analysis extends from the
resurfacing, close monitoring of the level of per¬ simplistic methods discussed in this section to the
missible forces is crucial. For example, quadriceps advanced methods reviewed in the next section.
force is the dominant factor in the tension gener¬ Both methods model a “force-movement unit,”
ated within the anterior cruciate ligament. Conse¬ which consists of the interaction of two basic
quently, in rehabilitation of the conservatively components: the biologic component, comprising
treated or surgically reconstructed anterior cruciate the joint, its associated passive structures, and the
ligament, extreme care should be taken in design¬ muscles; and the external force system, which is
ing clinically adequate exercises. Additionally, for the resistance the body has to overcome (or re¬
the muscle’s own sake, it is crucial to know the strain) (that is, objects lifted). Among the main
magnitude of contraction (relative to the maximal tenets of the model are the following:
tension) that a given muscle may be allowed to 1. Muscles generate only those tensile forces
exert after, for example, partial tear of its tendon. whose location with respect to an arbitrary
Had the situation been that each distinct muscle coordinate system can be determined.
was responsible for only one motion (e.g., in the 2. Ligaments (and capsule) exert tensile forces
case of the elbow joint: brachialis for flexion, only.
triceps for extension, supinator for supination, and 3. Joints are frictionless and transmit
pronator for pronation), simple mathematical mod¬ compressive forces.
els of the kind described below would have suf¬ 4. Joints have a well-defined axis of rotation
ficed to find the force exerted by the muscles for with respect to which the moment arms of
a specific activity. Biomotion, however, is distin¬ all forces may be calculated.
guished by a significant extent of redundancy; that
According to the simpler approach, biomechani¬
is, there are normally more muscles than what is
cal force systems have a unique solution that can
strictly necessary, although in some pathologies,
be arrived at by reducing the number of un¬
the opposite could be true.38 This creates a cardinal
knowns, mostly muscle forces, to the number
and fundamental biomechanical problem whose of equipollence equations. This can be done by
essence is the determination of the exact contribu¬ lumping several muscles into one “working
tion each muscle is making to the total joint load group,”40-42 by ascribing specific force proportions
system.
between muscles,43 or by ignoring certain muscles
An attempt to solve it using direct (in vivo) whose contribution to the moment is considered
measurements of muscle force proves to be a negligible.44 These procedures result in a statically
formidable task. Although it is possible to implant determinate system in which the number of un¬
a force transducer and measure the force exerted knowns equals the number of equations.
by the muscle, this intervention is obviously lim¬ The most basic statically determinate system
requires no reduction. Consider the simple planar it is operating against a “solid” plate of articular
situation in which segment 2 is to be held at a cartilage (with or without disk or bursae, when
certain position relative to segment 1 with which relevant). Depending on the geometry of the joint,
it shares a common joint (Fig. 4-12). In principle, however, passive tension of a ligamentous or a
one muscle (m) would suffice to support the capsular nature would be required to offset the
counter (resisting) moment provided it can gener¬ shear component, which invariably tends to sublux
ate enough tension. Assuming that the magnitude or dislocate the joint.
of the resisting moment (Mext) can be calculated, Problems involving motion of the segment are
the force generated by the muscle (Fm) can readily solved likewise. For instance, consider the case in
be derived using the following equilibrium equa¬ which segment 2 moves radially relative to seg¬
tion: ment 1. Providing that its moment of inertia and
its kinematics are known, the following equations
X M = rm X Fm + Mex, = 0 are equally applicable and yield a unique solution:
where rm is the vector representing the muscle’s

moment arm. In this hypothetical example, sup¬ X F = 0
pose that m has a moment arm of 40 mm, and the
external moment is 30 newton-meters (N m). The XM = 0
force (Fm) is therefore equal to 30 divided by 0.04,
or 750 N. If another muscle, q, is operating along where the relevant components of la are added.
with m, and it is assumed (e.g., owing to their If, on the other hand, owing to the inertia imparted
respective PCSAs) that Fm = 2Fq, the following by muscle m, segment 2 approaches the end of
equation can readily be solved for Fm and then Fq: range too quickly and antagonistic activity is initi¬
—> ated in muscle n (see Fig. 4-12) in such a way
X M = rm X Fm + r q X Fq + Mext - that m and n do not co-contract, Fn can be derived
rm X Fm + 0.5rq X Fm + Mext =0 in a similar fashion.
Inherent in these simple models is the assump¬
To solve for the reactive forces, the force equi¬ tion that neither synergism (m and q) nor antago¬
librium equation is used, as follows: nistic coactivation (m and n or m, q, and n) takes
place. This assumption is obviously erroneous and
therefore may be used only when there is good
X F = 0
grounds for its existence. On the other hand, re¬
The forces acting on the joint are then com¬ duction methods can sometimes yield a reasonable
monly resolved along the compression and shear approximation to the real situation, hence their
axes of the joint. Clearly, the compressive compo¬ applicability.
nent does not require additional balancing because
Statically Indeterminate Systems

and Common Methods of Solution
From a purely mathematical viewpoint, the situa¬

tion changes dramatically when the assumptions
given previously are no longer valid, that is, when
the number of muscles (or other load-transmitting
elements) is larger than the number of equations.
This results in redundancy or the so-called distri¬
bution problem,45 both of which relate to the deter¬
mination of the force magnitude developed by P
muscles based on R equations, where P is greater
than R. Because in principle there is an infinite
number of solutions to such a system, it is not
FIGURE 4-12 Schematic muscle-joint configuration
possible to determine a unique solution, and thus
system. Moment arms are denoted by respective r’s. F„, force
the system becomes statically indeterminate.
vector of muscle n; Fm force vector of muscle m; Fq, force
Such systems are the rule in musculoskeletal
vector of muscle q; rm, moment arm of muscle m; rq,
moment arm of muscle q; rn, moment arm of muscle n; Mext,
biomechanics, and their solution poses one of the
external moment. main challenges in modern studies. For example,
in the foregoing problem, suppose that m, q, and during the past 20 years, and some have proved
n are simultaneously contracting, that is, that there highly successful in predicting temporal activity
is an agonist-antagonist coactivation in addition patterns when compared with EMG traces. The
to synergy. The equation to be solved thus would magnitude of the forces, however, could not be
be as follows: validated. The fact that temporal patterns are pre¬
dictable proves, at least partially, that the body
2 M = rm X Fm + rq X Fq + rn X Fn + Mext = 0 works according to some optimization principle.
On the other hand, because the nature of the
If the method of simple reduction is ruled out, optimization is yet unknown, it may well be that
this equation must be supplemented by other con¬ more than one principle is operating. Furthermore,
ditions or equations to be solved. To that aim, cost functions that reasonably operate in intact
there are two general approaches. One approach and normal musculoskeletal systems may not nec¬
that has attracted considerable attention, optimiza¬ essarily apply in other instances. In this context,
tion, is based on the arbitrary yet physiologically one could speculate about the significance of spe¬
reasonable assumption that activation of muscles cific cost functions in movements provoking joint-
is done according to some criterion. Furthermore, related pain or the need to exercise or when a
this criterion, which is known by the terms cost, need arises for extremely delicate movements. It
objective, or penalty function, is to be minimized is therefore not surprising that the specific selec¬
or maximized—for instance, muscular stresses or tion of the optimization technique is paramount in
muscular endurance, respectively. The other ap¬ deciding the sort of solutions at which one ar¬
proach is based on testing all possible combina¬ rives.48 The evolution of optimization methods in
tions within a solution space and selecting those musculoskeletal biomechanics started with the use
that are physiologically viable or meet certain of linear models.49 This approach enjoys the rela¬
accepted criteria.46 Methodologically, the second tive ease of linear programming, but its results are
approach is reminiscent of the reduction tech¬ not always consistent physiologically. The nonlin¬
nique, although it is by far more sophisticated. In ear approach pioneered by Pedotti and associates50
this chapter, a more detailed presentation is given and Crownshield and Brand45 is more complicated
to the first (optimization) method, although it mathematically but yields physiologically inter¬
should by no means be construed as the better of pretable solutions.
the two. To demonstrate the elegance of the optimization
One of the first formulations of a cost function technique as well as its drawbacks, the following
in this kind of problem was by MacConaill,47 relatively simple example, taken from a study by
who proposed the minimal total muscular force Dul and coworkers,44 was selected. The problem
principle according to which “no more total force to be solved was that of finding the load distribu¬
than is both necessary and sufficient to maintain a tion among the muscles performing planar qua¬
body posture or perform a motion is used.”45 sistatic knee flexion in the seated position (Fig.
Other cost functions (see later) have been tested 4-13) while the force exerted by subjects in-
FlGURE 4-13 Schematic muscle-joint configuration for solving the optimization problem. FLH, long
hamstring; FSH, short hamstring; FG, gastrocnemius. (Adapted from Dul J, Townsend MA, Shiavi R, Johnson
GE. Muscular synergism. I. On criteria for load sharing between synergistic muscles. J Biomechanics
17:663-673, 1984, with permission from Elsevier Science.)
creased from 30 to 275 N. The problem was first where xmax refers to the maximal force or stress of
reduced to three muscle groups by the following the i-th muscle; in the present case, 30 N/cm2 was
lumping procedure: selected as the constraint.
Long hamstrings comprising the long biceps

Xi = KXj (i # j)
femoris, semimembranosus, and
semitendinosus
where K is a force ratio that dictates that when
Short hamstrings comprising the short biceps muscle X; is contracting, so does muscle x,; that
femoris is, there is a synergistic pattern. However, the
Gastrocnemius comprising the lateral and medial force (stress) in i is K times greater than that in j.
heads The results of this problem subject to several
different cost functions are presented next. For the
The small popliteus and plantaris muscles were sake of simplicity, the joint reactions were not
ignored owing to a small motion arm and small included in the figures.
PCSA, respectively. Also, no consideration was In the simplest situation, the design variables
given to ligamentous (e.g., posterior longitudinal) are muscle forces, p = 1; that is, optimization is
forces. There were three joint equipollence equa¬ done using the Simplex method, and the cost func¬
tions: two relating to the planar forces and one to tion is the sum of forces (2 F,). No upper force
the moment. On the other hand, the system in¬ limit is imposed, and synergy is not assumed. The
cluded five unknown forces (three muscle forces solution takes the form depicted in Figure 4—14A,
and two joint reactions), hence the optimization. which shows that the only muscle operating is the
In the second stage, the design variables and long hamstring because it has the longest motion
constraint functions were determined. The design arm and hence is the least expensive (minimal
variables are those elements that are included in force) to use. As pointed out before, this may not
the cost function and that are systematically varied be a realistic solution because coactivation takes
(normally by an iterative process) until the cost place. Another linear optimization is illustrated in
function is minimized. In the present context, they Figure 4-145, in which the cost function is the
are invariably the magnitudes of the muscular same as in the former example, but this time, the
forces.51 The constraint functions are those mathe¬ previous two equations are imposed, so that the
matical relationships aimed at restricting the solu¬ long and short heads of the hamstrings are con¬
tion to certain boundary conditions. tracting simultaneously. Moreover, the stress de¬
In formulating the optimization problem, the veloped in each of these muscles is equal (K =
following equations were applied: 1); that is, FLH = 4.9FSH. As a consequence, FLH
and Fsh depart together from the x axis, but the
Minimize: 2iXf (i = 1-3) slope is different, reflecting the force proportion.
Nonlinear optimization of the sum of cubed
where x( is a decision variable that stands for forces (2 F;3; p = 3) imposing the same con¬
either the i-th muscle force (N) or stress (N/cm2), straints as in the second example is shown in
the latter being defined as F/PCSA of the individ¬ Figure 4-14C. Note that in this case, the FLH-FSH
ual muscle; p is the power of the decision variable force pattern is strikingly similar, but FG follows
to be minimized. The equation p = 1 renders the a different recruitment pattern; that is, instead of
system linear; otherwise, it is nonlinear, subject to being delayed until the external force reaches
the following: about 130 N, it is contracting parallel to the ham¬
strings. Finally, another cost function, the sum of
x;>0 (i = 1-3) the cubed stresses (2 [F./PCSA,3]), is used under
the same constraints, and the results are depicted
in Figure 4-1 AD. Note the significant change in
where x; is the force or stress.
the order of strength occurring between FLH and
Fg.
SCjXi = M These examples emphasize that the solution to
the distribution problem is dependent on the
where c; (in cm or cm3) is a constant derived choice of the cost function and on the decision
from the moment equation and M is the resultant and constraint variables; that is, the solution is
moment (in Ncm). arbitrary. Nevertheless, such selection can be ren¬
dered physiologically sound and therefore can
N — Xmax yield a reasonable solution.
Muscle forces (N)
>
Muscle forces (N)
FIGURE 4-14 Solutions to different cost functions. Squares, long hamstring; circles, short hamstring;
triangles, gastrocnemius.
References thin filaments periods in skeletal muscle fibers of

the rhesus monkey. Anat Rec 178:63-82, 1973.
1. McComas AJ: Skeletal Muscle Form and 7. Lieber RL, Bodine-Fowler SC: Skeletal muscle
Function. Champaign, IL, Human Kinetics, 1996. mechanics: Implications for rehabilitation. Phys
2. Coers C, Woolf AL: The Innervation of Muscle. Ther 73:844-856, 1993.
Oxford, UK, Blackwell Scientific Publications, 8. Wiekiewicz TI, Roy RR, Powell PL, Edgerton
1959. VR: Muscle architecture of the human lower limb.
3. Barrett B: The length and mode of termination of Clin Orthop 179:275-283, 1983.
individual muscle fibers in the human sartorius 9. Lieber RL, Fazeli BM, Botte MJ: Architecture of
and posterior femoral muscles. Acta Anat selected wrist flexor and extensor muscles. J Hand
48:242-251, 1962. Surg (Am) 14:244-250, 1990.
4. Windhorst U, Hamm TM, Stuart DG: On the 10. Lieber RL, Jacobson MD, Fazeli BM:
function of muscle and reflex partitioning. Behav Architecture of selected muscles of the arm
Brain Res 12:629-681, 1989. and forearm: Anatomy and implications for
5. Herzog W: Muscle. In Nigg BM, Herzog W (eds): tendon transfer. J Hand Surg (Am) 17:787-798
Biomechanics of the Musculoskeletal System. 1992.
New York, John Wiley & Sons, 1994. 11. Woittiez RD, Huijing PA, Boom H, Rozendal RH:
6. Walker SM, Schrodt GR: I-segment lengths and A three dimensional muscle model: A quantified
relation between form and function of skeletal 29. Williams PE, Goldspink G: Changes in sarcomere
muscles. J Morphol 182:95-113, 1984. length and physiological properties in immobilized
12. Lieber RL, Blevins FL: Skeletal muscle muscle. J Anat 127:459^468, 1978.
architecture of the rabbit hindlimb: Functional 30. Gossman MR, Sahrmann SA, Rose SJ: Review of
implications of muscle design. J Morphol length-associated changes in muscle. Phys Ther
199:93-101, 1989. 62:1799-1808, 1982.
13. Alexander RMcN, Vernon A: The dimensions of 31. Tabray JC, Tabray C, Tardieu C: Physiological
knee and ankle muscles and the forces they exert. and structural changes in the cat’s soleus muscle
J Hum Movement Studies 1:115-123, 1975. due to immobilization at different lengths by
14. Gregor RJ: Skeletal muscle mechanics and plaster casts. J Physiol (Lond) 224:231-244, 1972.
movement. In Grabiner MD: Current Issues in 32. Tardieu C, Tabray JC, Tabray C: Comparison of
Biomechanics. Champaign, IL, Human Kinetics, the sarcomere number adaptation in young and
1993. adult animals: Influence of tendon adaptation. J
15. Pierrynowski MR: Analytic representation of Physiol (Paris) 73:1045-1055, 1977.
muscle line of action and geometry. In Allard P, 33. Lamontagne A, Malouin F, Richards C, Dumas F:
Stokes IAF, Blanchi J-P (eds): Three Dimensional Impaired viscoelastic behavior of spastic
Analysis of Human Movement. Champaign, IL, plantarflexors during passive stretch at different
Human Kinetics, 1995. velocities. Clin Biomech 12:508-515, 1997.
16. Huxley AF, Niedergerke R: Structural changes in 34. Gunther B: Dimensional analysis and theory of
muscle during contraction: Interference biological similarity. Physiol Rev 55:659-699,
microscopy of living muscle fibers. Nature 1975.
173:971-973, 1954. 35. McMahon TA: Muscles, Reflexes and
Locomotion. Princeton, NJ, Princeton University
17. Huxley AF, Simmons RM: Proposed mechanism
of force generation in striated muscle. Nature Press, 1984.
36. Koolstra JH, van Eijden TMGJ, Weijs WA: An
233:533-536, 1971.
iterative procedure to estimate muscle lines of
18. Huxley HE, Hanson J: Changes in the cross
action in vivo. J Biomech 22:911-920, 1989.
striations of muscle during contraction and stretch
37. McGill SM, Santaguida L, Stevens J:
and structural interpretations. Nature 173:973-976,
Measurement of the trunk musculature from T4 to
1954.
L5 using MRI scans of 15 young males corrected
19. Gordon AM, Huxley AF, Julian FJ: The variation
for muscle fiber orientation. Clin Biomech
in isometric tension with sarcomere length in
8:171-178, 1993.
vertebrate muscle fibers. J Physiol 184:170-192,
38. Dul J: The biomechanical prediction of muscle
1966. forces. Clin Biomech 1:27-30, 1986.
20. Sale DG, Quinlan J, Marsh E, et al: Influence of 39. Whitting QC, Gregor RJ, Roy RR, Edgerton VR:
joint position on ankle plantarflexion in humans. J A technique for estimating mechanical work of
Appl Physiol 52:1636-1642, 1982. individual muscles in the cat during treadmill
21. Dvir Z: Isokinetics: Muscle Testing, Interpretation locomotion. J Biomech 17:685-691, 1984.
and Clinical Applications. Edinburgh, Churchill 40. Paul JP: Bioengineering studies of the forces
Livingstone, 1995. transmitted by joints. In Kenedi RM (ed):
22. Rayment, I, Holden HM, Whittaker M, et al: Engineering Analysis, Biomechanics and Related
Structure of actin-myosin complex and its Bioengineering Topics. Oxford, UK, Pergamon
implications for muscle contractions. Science Press, 1965, pp 369-380.
261:58-65, 1993. 41. Morrison JB: The mechanics of the knee joint in
23. Finner JT, Simmons RM, Spudich JA: Single relation to normal walking. J Biomech 3:51-71,
myosin molecule mechanics: PicoNewton forces 1970.
and nanometre steps. Nature 368:113-119, 1994. 42. Procter P: Ankle Joint Biomechanics. Unpublished
24. Latash ML, Zatiorsky VM: Joint stiffness: Myth PhD thesis, Strathclyde University, Glasgow, UK,
or reality? Hum Movement Sci 12:653-692, 1993. 1980.
25. Zajac EF: Muscle and tendon: Properties, models, 43. Nicol A: Biomechanics of the Elbow Joints.
scaling and application to biomechanics and motor Unpublished PhD thesis, Strathclyde University,
control. Crit Rev Biomed Eng 17:359^411, 1989. Glasgow, UK, 1977.
26. Alexander RMcN: Elastic mechanisms in animal 44. Dul J, Townsend MA, Shiavi R, Johnson GE:
movement. Cambridge, Cambridge University Muscular synergism. I. On criteria for load sharing
Press, 1988. between synergistic muscles. J Biomech
27. Ford LE, Huxley AF, Simmons RM: Tension 17:663-673, 1984.
responses in sudden length change in stimulated 45. Crownshield RD, Brand RA: The prediction of
frog muscle fibers near slack length. J Physiol forces in joint structures: distribution of
(Lond) 269:441-449, 1977. intersegmental resultants. Exer Sport Sci Rev
28. Magid A, Law DG: Myofibrils bear most of the 9:159-181, 1981.
resting tension in frog skeletal muscle. Science 46. Collins JJ: The redundant nature of locomotor
optimization laws. J Biomech 28:251-268, 1995.
230:1280-1282, 1985.
47. MacConnail MA: The ergonomic aspect of evaluation of forces in lower extremities of the
articular mechanics. In Evans FG (ed): Studies on musculoskeletal systems. J Biomech 6:313-326,
the Anatomy and Function of Bones and Joints. 1973.
Berlin, Springer, 1967. 50. Pedotti A, Krishnan VV, Starke L: Optimization of
48. An K-N, Kaufman K, Chao EYS: Estimation of muscle-force sequencing in human locomotion.
joint and muscle forces. In Allard P, Stokes IAF, Math Biosci 38:57-76, 1978.
Blanchi J-P (eds): Three Dimensional Analysis of 51. Herzog W, Binding P: Mathematically
Human Movement. Champaign, IL, Human indeterminate systems. In Nigg BM, Herzog W
Kinetics, 1995. (eds): Biomechanics of the Musculoskeletal
49. Seireg A, Arkivar RJ: A mathematical model for System. New York, John Wiley & Sons, 1994.
CHAPTER 5
Biomechanics of the
Thoracolumbar Spine
Stuart M. McGill
Most clinicians reading this chapter will already The trabecular arrangement within the cancellous
have studied basic biomechanics and anatomy. bone is aligned with the trajectories of stress to
The study of biomechanics becomes an exercise which it is exposed. Three orientations
in basic science if the clinical relevance is ignored. dominate—one vertical and two oblique2 (Fig. 5-
The intent of this chapter is to describe some 2). This is a special architecture in terms of how
normal biomechanics of the thoracolumbar spine, the vertebral bodies bear compressive loading and
then to describe the injury process, and finally how they fail under excessive loading. Although
to revisit some anatomic-biomechanical features, the walls of the vertebrae appear to be rigid on
possibly in a way not previously considered, and compression, the nucleus of the disk pressurizes
to relate these to function, the reduction of the (see the classic works by Nachemson3,4) and
risk of low-back injury or reinjury, and the design causes the cartilaginous end plates of the vertebrae
of optimal rehabilitation programs. The profes¬ to bulge inward, seemingly to compress the can¬
sional challenge is to make wise decisions from cellous bone.5 In fact, under compression, it is the
the blending of laboratory and empirical evidence cancellous bone that fails first,6 making it the
with clinical experience. determinant of failure tolerance of the spine (at
least when the spine is not positioned at the end
range of motion). It is difficult to injure the disk
Normal Biomechanics of the anulus this way (anular failure is discussed
Thoracolumbar Spine later).
Although this notion is contrary to the concept
that the vertebral bodies are rigid, the functional
ANATOMIC FEATURES AND THEIR CLINICAL
interpretation of this anatomy suggests the pres¬
RELEVANCE
ence of a clever shock-absorbing and load-bearing
system. Farfan7 proposed that the vertebral bodies
Vertebrae
act as shock absorbers of the spine, although he
based this on vertebral body fluid flow and not
The Body
end-plate bulging. Because the nucleus is incom¬
It is assumed that the reader knows there are 12 pressible, bulging end plates suggest fluid expul¬
thoracic and 5 lumbar vertebrae. The construction sion from the vertebral bodies, specifically blood
of the vertebral bodies themselves may be likened through the perivertebral sinuses.8 This would sug¬
to a barrel whereby the round walls are formed gest protective dissipation on quasistatic and dy¬
with relatively stiff cortical bone (Fig. 5-1). The namic compressive loading of the spine. The ques¬
top and bottom of the barrel are formed with a tion is: how do the end plates bulge inward into
more deformable cartilage plate (end plate) that is seemingly rigid cancellous bone? The answer ap¬
about 0.6 mm thick but is thinnest in the central pears to be in the architecture of the cancellous
region.1 The end plate is porous for passage of bone, which is dominated by the system of col¬
nutrients such as oxygen and glucose, whereas the umns of bone (shown in Fig. 5-2) with much
inside of the barrel is filled with cancellous bone. smaller transverse bony ties. On axial compres-
103
FIGURE 5-1 The parts of a typical lumbar vertebra. VB, vertebral body; P, pedicle; TP, transverse process;
SP, spinous process; L, lamina; SAP, superior articular process; LAP, inferior articular process; saf, superior
articular facet, iaf, inferior articular facet; MP, mamillary process; AP, accessory process; vf, vertebral
foramen; RA, ring apophysis; NA, neural arch. (From Bogduk N, Twomey LT: Clinical Anatomy of the
Lumbar Spine. New York, Churchill Livingstone, 1987.)
I Chapter 5 BIOMECHANICS OF THE THORACOLUMBAR SPINE 105
its original structure and function. It would appear

that cancellous fracture could heal quickly when
damaged given the small amount of osteophyte
activity needed, at least compared with the length
of time needed for repair of collagenous tissues.
Both the disk and the vertebrae deform while
supporting spinal loads. Under excessive compres¬
sive loading, the bulging of the end plates into the
vertebral bodies also causes radial stresses in the
end plate sufficient to cause fracture in a stellate
pattern. These fractures, or cracks, in the end
plate are sometimes sufficiently large to allow the
nucleus of the disk to squirt through into the
vertebral body,10 resulting in the formation of the
classic Schmorl’s node (Fig. 5^1). This type of
injury is associated with compression of the spine
when the spine is not at the end range of motion
(i.e., not flexed, bent, or twisted). I believe that
this type of common compressive injury is often
misdiagnosed as a herniated disk owing to the
flattened interdiscal space seen on planar radio¬
graphs. The anulus of the disk, however, does
remain intact. It is simply a case of the nucleus’
leaving the disk and progressing through the end
plate into the cancellous core of the vertebrae.
The Posterior Elements
The posterior elements of the vertebrae (pedicles,

laminae, spinous processes, and facet joints) have
a shell of cortical bone but contain a cancellous
bony core in the thicker parts. The transverse
processes project laterally together with a superior
pair and inferior pair of facet joints (see Fig. 5-1).
On the lateral surface of the bone that forms the
superior facets are the accessory and mamillary
FIGURE 5-2 The trabeculae are aligned (according to processes, which, together with the transverse
Gallois and Japoit, 19252) with the dominant trajectories of process, are major attachment sites of the longissi-
stress. A, Vertical trabecular system. B, Inferior and superior mus and iliocostalis extensor muscle groups (de¬
oblique systems. C, Oblique systems viewed from above.
scribed later). The facet joints are typical synovial
(From Dupuis PR, Kirkaldy-Willis WM: The spine: Integrated
joints in that the articulating surfaces are covered
function and pathophysiology. In Cruess RL, Rennie WRJ
with hyaline cartilage and are contained in a cap¬
[eds]: Adult Orthopaedics, vol 2. New York, Churchill
Livingstone, 1984, p 683.)
sule. Fibroadipose enlargements or miniscoids are
found around the rim of the facet, although mostly
at the proximal and distal poles,11 which have
sion, as the end plates bulge into the vertebral been implicated as a possible structure that could
bodies, these columns experience compression and “bind” and lock the facet joint (Fig. 5-5).
appear to bend in a buckling mode; under exces¬ The neural arch in general (pedicles and lami¬
sive load, they buckle as the smaller bony trans¬ nae) appears to be somewhat flexible. In fact,
verse ties fracture, as documented by Fyhrie and Bedzinski12 demonstrated flexibility of the pars
Schaffler9 (Fig. 5-3). In this way, the cancellous during flexion and extension of cadaveric spines,
bone can rebound to its original shape (at least whereas Dickey and associates13 documented up
95% of the original unloaded shape) when the to 3-degree changes of the right pedicle with re¬
load is removed, even after suffering fracture and spect to the left pedicle during mild daily activities
delamination of the transverse ties. This architec¬ using pedicle screws in vivo. Failure of these
ture appears to afford superior elastic deformation, elements, together with facet damage, leading to
even after marked damage, and then heal to regain spondylolisthesis, is often blamed exclusively on
FlGyRE_5^3 Under compressive loading, bulging of the end plate causes buckling stresses in the
vertical trabeculae (A) that, when excessive, cause damage in the transverse trabeculae (B). (From Fyhrie
DP, Schaffler MB: Failure mechanisms in human vertebral cancellous bone. Bone I5[l]: 105-109, 1994,
with permission from Elsevier Science.)
4 Stellate-patterned end-plate fracture. B, Intrusion of nuclear material (shown at the tip of

the scalpel) into the vertebral body from compressive loading of a spine in a neutral posture These are
porcine specimens.
FIGURE 5-5 Lateral view of the facet face revealing the fibroadipose meniscoids (FM) and the adipose
tissue pad (AP), which have been implicated in joint binding. CT, connective tissue rim. (From Bogduk N,
Twomey LT: Clinical Anatomy of the Lumbar Spine. New York, Churchill Livingstone, 1987, p 31.)
anteroposterior shear forces. A case could be made higher load rates produced wedge fractures and
from epidemiologic evidence, however, that the facet damage.
damage to these posterior elements may also be
associated with full range of motion in athletes Intervertebral Disk
such as gymnasts and Australian cricket bowlers.14
It would appear that injury to the posterior bony The disk comprises three major components: the
elements in these sorts of activities is a fatigue nucleus pulposus, the anulus fibrosus, and the end
injury caused by cyclic full flexion and extension, plates. The nucleus is a gel-like substance with
fatiguing the arch with repeated bending. On the collagen fibrils suspended in a base of water and
other hand, there is no doubt that excessive shear various mucopolysaccharides, giving it both vis¬
forces also cause injury to these elements. Poste¬ cosity and some elastic response when perturbed
rior shear of the superior vertebrae can lead to in vitro. Although there is no distinct border with
ligamentous damage but also failure in the verte¬ the anulus, the lamellae of the anulus become
brae itself as the end plate avulses from the rest more distinct, moving radially outward. The colla¬
of the vertebral body (Fig. 5-6). Anterior shear of gen fibers of each lamina are obliquely oriented;
the superior vertebrae has been documented to the obliquity runs in the opposite direction in each
cause pars and facet fracture, leading to spondylo¬ concentric lamella. The ends of the collagen fibers
listhesis with a typical tolerance of an adult lum¬ anchor into the vertebral body with Sharpey’s fi¬
bar spine of about 2000 N.15 Although similar bers in the outermost lamellae, whereas the inner
injury mechanisms and tolerance values were ob¬ fibers attach to the end plate (discussed earlier).
served in young porcine spine specimens,6 the The disks in cross-section resemble a rounded
type of injury appeared to be modulated by load¬ triangle in the thoracic region and an ellipse in
ing rate. Specifically, anterior shear forces pro¬ the lumbar region, suggesting anisotropic facilita¬
duced undefinable soft tissue injury at low load tion of twisting and bending.
rates (100 N/s), but fractures of the pars, facet The disk appears to be a hydrostatic structure
face, and vertebral body were observed at higher that allows 6-degrees-of-freedom motion between
load rates (7000 N/s). Posterior shear forces ap¬ vertebrae, but its ability to bear load is dependent
plied at low load rates produced undefinable soft on its shape and geometry, as determined by the
tissue failure and vertebral body fracture, whereas adjacent vertebrae. Because of the orientation of
support compressive load as the nucleus pressur¬

izes, applying hydraulic forces to the end plates
and to the anulus and causing the anulus collagen
fibers to bulge outward, which places them under
tension.
Many years ago, Markolf and Morris16 elegantly
demonstrated that a disk with the nucleus removed
lost height but preserved properties of axial stiff¬
ness, creep, and relaxation rates. It would appear
that the nucleus is required to preserve disk height,
which has implications on facet loading, shear
stiffness, ligament mechanics, and so forth. Con¬
sideration of progressive disk injury is in order
here. If little hydrostatic pressure is present, per¬
haps the nucleus has been lost through end-plate
fracture or herniation; when the disk is com¬
pressed, not only does the outer anulus bulge
outward but also the inner anulus bulges inward
(Fig. 5-7). This double-convex bulging causes the
laminae of the anulus to separate, or delaminate,
and has been hypothesized to form a pathway for
nuclear material to leak through the lamellar lay¬
ers and finally extrude, creating a frank herniated
disk.17
FIGURE 5-6 Shear injuries include fracture of the facet
From a review of the literature, one can make
base and end-plate avulsion from the vertebrae.
three general conclusions about anulus injury and
resulting bulging or herniation. First, it appears
that the disk must be bent to the full end range of
the collagen fibers within the concentric rings of motion to herniate,18 and herniations tend to occur
the anulus, with half of the fibers oblique to the in younger spines19 (those with higher water con¬
other half, the anulus is able to resist loads in tent20 and more hydraulic behavior). Second, disk
twist; however, only half of the fibers are able to herniation is associated with extreme deviated
support this mode of loading, and the other half posture, fully flexed, and the risk is higher with
become disabled, resulting in a substantial loss of repeated loading of at least 20,000 or 30,000
strength or ability to bear load. When the disk is times, highlighting the role of fatigue as a mecha¬
subjected to bending and to compressive load, it nism of injury.21-22 Third, epidemiologic data link
has been argued that the anulus and the nucleus herniation with sedentary occupations and the sit-
^'GURE 5-7 lf the nucleus loses Pressure (*•§•■ ^ a result of an end-plate fracture) upon compression U
the anulus compresses, resulting in radial bulging both outward and inward, causing delaminating stresses.
When the nucleus is contained in a healthy disk, these stresses are minimal (B).
ting posture.23 In fact, Wilder and associates24 doc¬ overlying the L4 extensors are seen. Raw muscle
umented anular tears in young calf spines from relative physiologic cross-sectional areas and mo¬
prolonged simulated sitting postures and cyclic ment arms are provided in Tables 5-1 to 5-3,30
compressive loading (i.e., simulated truck driv¬ whereas areas corrected for oblique lines of action,
ing). Older spines appear not to exhibit the classic for some selected muscles at several levels of
extrusion of nuclear material but rather are charac¬ the thoracolumbar spine, are shown in Table 5-4.
terized by delamination of the anulus layers and Guidelines for estimating true physiologic areas
by radial cracks, which appear to progress with are provided by McGill and associates.29
repeated loading (see the review by Goel and Moment arms of the abdominal musculature
colleagues25). reported in CT- and MRI-based studies have re¬
cently been shown to underestimate true values
Muscles by 30% as a result of the supine posture adopted
in MRI and CT scanners, which causes the abdom¬
Most textbooks present the major thoracic and inal contents to collapse under gravity.31 In real
lumbar musculature from a posterior view. How¬ life, the abdominals are pushed away from the
ever, many of the functionally relevant aspects are spine with the visceral contents when a person is
better viewed in the sagittal plane (see the synop¬ standing.
sis of the sagittal plane lines of action presented
by Bogduk and colleagues26-28). Furthermore, The Rot at ores and
there is a tendency to obtain a mechanical appreci¬ Intertransversarii
ation of function by simply interpreting the lines
of action, region of attachment, and lines of pull Many anatomic textbooks describe the small rota¬
of the musculature, which may be misleading. tor muscles of the spine, which attach adjacent
Together with knowledge of muscle morphology, vertebrae, as fulfilling the role of creating axial
knowledge of activation of the musculature in a twisting torque (Fig. 5-9). Similarly, the inter¬
wide variety of movement and loading tasks is transversarii are assigned the role of lateral
required to understand the function and purpose flexion. There are several problems with these
of each muscle and how the motor control system proposals. First, these small muscles are of such
activates the musculature to support external small physiologic cross-sectional area that they
loads. Therefore, this section provides an anatomic can generate only a few newtons of force, and
description of the musculature together with the second, they work through such a small moment
results of various electromyography (EMG) stud¬ arm that their total contribution to rotational axial
ies to help interpret function. twisting and bending torque is minimal. It would
appear that they have some other function. There
Muscle Size is evidence to suggest that these muscles are
highly rich in muscle spindles (4.5 to 7.3 times
The physiologic cross-sectional area of muscle more rich than multifidus32), such that they would
determines the force-producing potential, whereas be involved as length transducers or vertebral po¬
the line of action and moment arm determine sition sensors at every thoracic and lumbar joint.
the effect of the force in moment production, In some indwelling EMG experiments per¬
stabilization, and so forth. It is erroneous to esti¬ formed on ourselves a couple of years ago, we
mate force based on muscle volume without ac¬ placed some electrodes very close to the vertebrae.
counting for fiber architecture or from taking In one case, we strongly suspected that the elec¬
transverse scans to measure anatomic cross-sec¬ trode was in a rotator. Isometric twisting efforts
tional areas.29 Muscle forces are often underesti¬ with the spine untwisted (or in a neutral posture)
mated because a large number of muscle fibers were attempted in both directions, which produced
are not “seen” in a single transverse scan of a no EMG activity from the rotator—only the usual
pennated muscle, and magnetic resonance imaging activity in the abdominal obliques and so forth.
(MRI) or computed tomography (CT) scans must When nonresisted twisting was attempted in one
be corrected for architecture and scan plane obliq¬ direction there was no response, whereas in the
uity.30 Transverse scans of one subject show the other direction there was major activity. It ap¬
changing shape of the torso muscles over the peared that this particular rotator was not activated
thoracolumbar region (Fig. 5-8), highlighting the through torque development but rather acted in
need to obtain fiber architecture data from dissec¬ response to position change. Thus, its activity re¬
tion. In this example, the thoracic extensors seen sulted as a function of twisted position; it was not
at T9 provide extensor moment at L4 even though consistent with the role of creating torque to twist
they are not seen in the L4 scan; only their tendons the spine. From a clinical perspective, it is likely
I 10 Clinical Biomechanics
FIGURE 5-8 Transverse scans of one subject (supine) at the levels of T9 (A), LI (B), L4 (C), and SI (D)
showing the musculature in cross-section.
that these structures are effected during therapeu¬ and pars thoracis, and iliocostalis lumborum pars
tic manipulation with the joint at end range of mo¬ lumborum and thoracis. These two functional
tion.
groups (pars lumborum and pars thoracis) form
quite a marvelous architecture for several reasons
The Extensors: Longissimus, and are discussed with this distinction (i.e., lumbar
Iliocostalis, and Multifidus versus thoracic).
Groups The pars thoracis components of these two mus¬
cles attach to the ribs and vertebral components
The major extensors of the thoracolumbar spine and have relatively short contractile fibers with
are the longissimus, iliocostalis, and multifidus long tendons that run down the spine, to their
groups. The longissimus and iliocostalis groups origins over the posterior surface of the sacrum
are often separated in anatomy books, although it and medial border of the iliac crests (Fig. 5-10).
may be more productive to recognize the thoracic Their basic line of action is parallel to the com¬
portions of both muscles separately from their pressive axis of the spine. Furthermore, their line
lumbar portions because they are architecturally26 of action over the lower thoracic and lumbar re¬
and functionally different.33 Even fiber-typing gion is superficial, such that forces in these mus¬
studies have noted differences between the lumbar cles have the greatest possible moment arm and
and thoracic sections; the thoracic sections contain therefore produce the greatest amount of extensor
about 75% slow twitch, whereas the lumbar sec¬ moment with a minimum of compressive penalty
tions are generally evenly mixed.34 Bogduk26 parti¬ to the spine. When seen on a transverse MRI or
tioned the lumbar and thoracic portions of these CT scan at a lumbar level, their tendons have
muscles into longissimus thoracis pars lumborum the greatest extensor moment arm, overlying the
VO - v
T (N O vO
fS N N N
N vfl ro la — 1^
o ^ i\ c-. co
TT lo N sO nO CO
CM CM ON
N CM O' x—v
CO CM 00 LD
^ ^ ^ O'
^ m n o\ — o
i^. — r^. h*.
CO O t'-'- I""- £ +
— CM
O' O' m t
00 00 'O - s
cs — — —
T N CM O' c-. cm
vO O' T N O' 00
N 'O CO 00
ON — — o
lo oo o o
— CM CM —
— CM On On
00 CO M" CM
LO LO O —
^ o
CM NO
— ON
CS CS C2^ CH — NO^
cm lo oo r". co — CO CM
- LO LO - - hv CO CO
CO O 'T "T 'T 'T on
Po O^ lo
o co co lo — CM LO
Raw Cross-Sectional Areas (mm2) Measured Directly From Magnetic Resonance Imaging Scans
ON LO CO CM CM CM — C'-
— CO 00 O' O CM LO —
LO LO NO CO On CM — LO
CO CO CO CM nO N o o
,—> CM
s O — NO CM LO ’<r
O' — CONONOOO — ODCO CM CM
'T CM CM — 00 00 CM CO CS CM — h>
COCOM)-NtCSCM-
rocoLOLOcocMLOOco^T
O ON LO LO CO CO CM — 00 O
— — — CM
CM M- CO O nOnOnO — CO LO — OnOn.
CM CO CO t vO N CM fO LO LO CM — — i
LO 00 — LO O' (N t o^r — ocmono —

O' LO CM CO 00 — — \D — O CO \D fS (N T CO
O CM O' 00 CM CO O ON'O'OOtCOCOCMN
CM CM — CO
OO'^-'ON^—vx-vOOlOOOOnOOnOlOOn
CMCO-t'OiO'ONOOCMCMiOtCMCOCO
CM — fOCMO'0'CM(SttCOCMCMCMCMNO
CO O — COOtNfSiOCOOCOCMtt —
t CO 00 LO O' CM — 00 — CM — 00 O' O co O'
CM — — — cMcoNnOnONlO^CO-^-COO
— — — — CM CM — O'
Êrector mass includes longissimus thoracis, iliocostalis lumborum, and multifidus.
-(-Abdominal wall includes external and internal oblique and transverse abdominis.
O' O CM CM CO CM . CO Op vOO — — CO CO N CM — P M) m CO CM O'

CO NT CM CO N f OONOLONLOOOONNLOONONrLOOOONONOO
fS (N fS CM 3- C2^ — tJ-lOCOCM — — — — CM — LO CM CM — —
CM 00 00 — lO N vO M5 — — LO O' Nf O CO M) O' fS t CO N — CM 'f CM t}-

— nT LO LO lo CM O' ’t lo — rôoOLo t nO CM N m fO N — LO — CO
CO
CO
00
— CO O O LONOO — — O — — ro CO t co CO CO — CM lo 'O
CM CM — — — LO
ro h- — COIÔOLOCMOOCMON — ro — O' CM P CO CM O' — cm

M N N — — — — — fSvO'OcMtCMN'OO' — lO 00 nO O' —
— — — CM ro ro — — nJ- -sf- — — ro cm^ cM^ cM^ —_ CM, ^ co, co^ CM^ CM
OrO'OLOL/lT'O'OONONNfNCOinlNCMCM'O — CO 'T ro —
r>. O' co — fSiON'OCMTCO'OONtNwncooO'O'P
'O'OCMcoinTfococifoNNcocottcicMCOO'LniflS
— — — — ro ro — — CM cM — —
Fn — O' 00 v lo v —s 00 lo O' vO P CM O' O' O M"

O 00 O' ro — CM 00 — h- roNtNOTNCO
mtfOCMCMcMfMO'
CM — — CM 00 3 CO S S S
O'OLOCMroOf^'^rcMO — tj- — o lo lo O' ro

^Standard deviation is given in parentheses.
LO NT — O' O O ro CM — CM LO CO NO CM CM CM LO —
NNO'ONO'O'rMOlT'}' -CMCOOONnO'TCO
CM CM — — — —
CM CM
^ ro — CO oo t NO cm
CM N O' N CO CO O' c- —
CM LO CO CO CO CO — CM CM o^
hv CM lo NO no O O CM
O 00 O O' nO —
00 O CO O'
r^. oo O' O' NO LO
— CM
<D CL) 
O O o ■§ "Oo -oo nrt t;ri 5 t
-9.
5-1
3 3 5 5 E E
* * —!! ■5 I s a
oo ir i/i
txO fTn P miss w) ~U “o P
I 'TDi ‘g> g> g 8 a C .</> k _ „
TABLE
<l) <L)

— — t m n n
Raw Lateral Distances (mm) Between Muscle Centroids and Intervertebral Disk Centroid
—„ ^ — fS rN O' — tj- ro
O' 'O — N CO fS O' O' TT O

(N ro o no — (N (N n Tf-
l l l “ 7 l
; £7" o-T oT — FZ ro Tf ro fN — lo
O t N O — O' CO fN O <N <N NO o
ro ro sO vO — — <N (N t ro n t t
I I I - "T I I I
00 —
— — LtT FvT (-o' rS 3s S' ^
CO LO ro — (N N 't
'O \D — — (N — Tf*
I I-7
O' O' O' O' , ro CDO't'Oriroioioro
ro ro vO vO
Êrector mass includes longissimus thoracis, iliocostalis lumborum, and multifidus.
fAbdominal wall includes external and internal oblique and transverse abdominis.
iiiilll I I
co^ ^ 6? ^
oo" oo^ <S hT co' fC" t Ifl t 'O — oo — t iA ro n >0 t t
ro oo O ifl'O(NrlNO't(NO'(NC0-tfS
cn — — — o — — <n — lo tj- — — o
iC fC ! O' — — o^ — —
00 vO lo O 0s ro CO — rOLO
rororsirsoooo — —
*Standard deviation is given in parentheses.
- - CO O' vO lo TT LD
<N — 'f "T

rs rs lo lo
i i
w U1 E
.zi «/>
y *o ?
.£•£«« E
0 —r3 |I
El.?? E E
g S’ « -Q £ If-B I
iq u> — O
Table 5-2
XI
! I 0 o -QO _£3 <D —
O 1/1 ^ c
■E .E rt E E
:n E55 3 -o
X -o
00 <*z
E .E
1
Z o v 'n So
sss
(U V Si
v _ VU rt
g
tJ L
2-8 S’ «■ 8
rrt-LJ(~l^~UO-;—(*-*
2 -D £
p a) o o _S _
rt m
ot a: w lij < t 2 3 i 11 i £ <l £ a o1 s
ai J oi j oi _j a: j oi j a: j oi
I 12
lo m m ro
I I I
o
fO N t Ifl
CM — vO CM
— — Tf rj-
I I I I
vO CD t t
I I I I
g
N0
I I I I
SSSss
Is-. CM O CM —
t fM — LO LO
I I I I I
^SSss
Raw Anterioi^Posterior Distances (mm) Between Muscle Centroids and Intervertebral Disk Centroid
O^
O' n t m t rM o
t PM fM LO
I I I I I I
t t — CS ^ iC- cT
_ S'
M)fMh-'ONNfMCOt(M
LOLOLOLO'^^CrOCMLOLO
I I I I I I I
00 RCo ro 00 00 i-O LO CM — vO 00
o O' O' CO — O O' nO 1^- ^ — — CM’^r

\DLOLOLOLOLororoLOLn — — rororM
i i i i i i i i i i i i i i
RfT m -5}- so lo 2 R LO ^ Q ^ ^ UT,
'Ô'O — — i/l T CO
tj- io >o — — ro m —
I I I I I I I ! I I I I
^ S' oT
- - - -
f'T' (o'
- —
,—..
nO LO
v
— — S' (o' S' (o (o' (o' — — (o' (o' (o' S' Co (o' S'
OCMOOOOnOO'O'^J'O — CM CO — — LO 'O N 'D — (S O' 03 hv vO ON
O'O'tSfMOtfDtcOfO'D'O'O'Ouiifltt'O'O I | rr
I I I I I I I I I I I
(Erector mass includes longissimus thoracis, iliocostalis lumborum, and multifidus.
fAbdominal wall includes external and internal oblique and transverse abdominis.
O t t — y—CM — O (S CM /T . ^ nO S' -—
— — — — 0s — — — — — nO LO r— — — lo lo lo t 'O 'O co
OOCnlo-OCMCoIÔ — — fN N LO O l^ n — — N M) N 'f —
-3- vO nO
s 00 CM — CM (N CM CM — rMvOvOLOLOLOLO'T'
I I
I I I I I I I I I I
ttOCONN-tfIM LO LO LO O' LO O'
ro ro lo CM — — 00 O — CM — — — CM 'O — CM
Nlsfoottfimnfn sO nO CO CO I
II II
^Standard deviation is given in parentheses.
^ (o' S'
o lo — — nO lo O' 00
I I
S 3 § s> -8
_ C S 6 s
■ ±r
U
o o-^-Q
5? <d« -5 -£ § E
■n„ EP
u
E
3
O O
|!
E E
* £ (/J — -
rt rt C 2 •- .S'! 55
5-3
o o o -§ 8 £ E c
LO £ £
I I S S3 | | E 1 P i-
!« <11
rt
2111 S §!
L0
E
O
E
O M W) g P ±J T3 .«/> i/>
L0
_=
X “O L.
b <D o5 q3 ^ 8 3 g «
TABLE
c x) -a
CD <u « § 3 2 2 £ O' o 3
cL aL l2 i2 £ _£ £
S_iS_iS_iS_iS_i _J o' _i I 2
113
any anterior-reaction shear forces of the upper

vertebrae produced as the upper body is flexed
forward in a typical lifting type of posture (this
possible injury mechanism is addressed later).
The multifidus muscles perform a different
function, particularly in the lumbar region, where
they attach posterior spines of adjacent vertebrae
or span two or three segments. Their line of action
tends to be parallel to the compressive axis or, in
some cases, runs anteriorly and caudad in obliq¬
uity. The major mechanically relevant feature of
multifidus, however, is that because it spans only
a few joints, its forces affect only local areas of
the spine. Therefore, the multifidus muscles are
involved in producing extensor torque, but they
provide the ability for corrections or moment sup¬
port only at specific joints that may be foci of
stresses. In a later section, we propose an injury
mechanism involving inappropriate neural activa¬
tion signals to multifidus, using an example of
injury observed in the laboratory.
The Abdominal Wa 11
Although many classic anatomy texts consider the

abdominal wall to be an important flexor of the
6 trunk, it appears that the rectus abdominis is the
FIGURE 5-9 Short muscles of the spine have been major trunk flexor (and the most active during sit-
erroneously described as performing the duty of creating ups and curl-ups35). It is interesting to consider
axial twisting torque. I, Levator costarum longus; 2, Levator why the rectus abdominis is partitioned into sec¬
costarum brevis; 3, intertransversarius medialis; 4, tions, rather than being a single long muscle, given
intertransversarius lateralis; 5, rotator longus; 6, rotator that the sections share a common nerve supply
brevis, 7, interspinalis. (From Dupuis PR, Kirkaldy-Willis WH:
and that a single long muscle would have the
The spine: Integrated function and pathophysiology. In Cruess
advantage of broadening the force-length relation¬
RL, Rennie WRJ [eds]: Adult Orthopaedics, vol 2. New York,
Churchill Livingstone 1984, p 70S.)
ship over a greater range of length change. Per¬
haps a single muscle would bulk upon shortening,
compressing the viscera, or be stiff and resistant
to bending. Not only does the sectioned rectus
lumbar bulk—often by more than 10 cm29-30 (see abdominis limit bulking upon shortening, but also
Fig. 5-8). the sections have a “bead effect,” which allows
The lumbar components of these muscles (ilio- bending at each tendon to facilitate torso flexion
costalis lumborum pars lumborum and longissi- and extension or abdominal distention or contrac¬
mus thoracis pars lumborum) are very different tion as the visceral contents change volume.36
anatomically from their thoracic namesakes. They The three layers of the abdominal wall (external
connect to the mamillary, accessory, and trans¬ oblique, internal oblique, transverse abdominis)
verse processes of the lumbar vertebrae and origi¬ perform several functions. They are involved in
nate, once again, over the posterior sacrum and flexion and appear to have their flexor potential
medial aspect of the iliac crest. Each vertebra is enhanced as a result of their attachment to the
connected bilaterally with separate laminae of linea semilunaris (Fig. 5-12),37 which redirects the
these muscles (Fig. 5-11). Their line of action is oblique muscle forces down the rectus sheath to
not parallel to the compressive axis of the spine; increase their flexor moment arm. The obliques
rather, they have a posterior caudal obliquity that are involved in torso twisting38 and lateral bend39
causes them to generate posterior shear forces and appear to play some role in lumbar stabiliza¬
together with extensor moment on the superior tion because they increase their activity, to a small
vertebrae. These posterior shear forces support degree, when the spine is placed under pure axial
I Chapter 5 BIOMECHANICS OF THE THORACOLUMBAR SPINE I 15
TABLE 5-4 Examples of Corrected Cross-Sectional Areas and Anteroposterior and Lateral Moment Arms
Perpendicular to the Muscle Fiber Line of Action
Values*
Cross-Sectional Area Anteroposterior Lateral Moment Arm

Muscle Vertebral Level (mm)2 Moment Arm (mm) (mm)
Longissimus pars lumborumf L3-L4 644 51 17

Quadratus lumborum LI-L2 358 31 43
L2-L3 507 32 55
L3-L4 582 29 59
L4-L5 328 16 39
External oblique L3-L4 1 121 17 1 10
Internal oblique L3-L4 1 154 20 89
*Data are derived from the cosines listed in McGill and colleagues.30 These are the values that should be used in biomechanical models, rather than
the uncorrected values obtained directly from scan slices.
-(Longissimus pars lumborum at the L4-L5 level would have been listed here by virtue of their cosines but were not because they could not be
distinguished on all scan slices.
FIGURE 5-10 A bundle of longissimus thoracis pars thoracis has been isolated (inserting on
the ribs at T6). Their tendons, lifted by the probes, course over the full lumbar spine to their
sacral origin. They have a very large extensor moment arm.
FIGURE 5-1 1 lliocostalis lumborum pars lumborum and longissimus thoracis pars lumborum
originate over the posterior surface of the sacrum, follow a superficial pathway, and then dive
obliquely to their vertebral attachments. The compressive axis is shown (C). They create
posterior shear forces and extensor moment on each successive superior vertebra.
compression40 (this functional notion is developed vation level with each increase in hand load more
later in this chapter). than any other muscle. After measuring the activa¬
tion of the psoas, the obliques, the extensors, and
The Special Case of Qua dr at us the quadratus lumborum, it was clearly the quadra¬
Lumborum and Psoas tus lumborum that the major muscle incrementally
activated to stabilize the spine in this special situa¬
Although the psoas has often been claimed to be tion in which only compressive loading was ap¬
a good stabilizer of the lumbar spine, I think plied to the spine in the absence of any bending
that this is unlikely and that, rather, quadratus moments. In addition, the architecture of the quad¬
lumborum is the major stabilizer, particularly of ratus lumborum suits a stabilizing role by at¬
the lower thoracic and lumbar regions. It is true taching each transverse process (therefore, bilat¬
that the psoas complex attaches to T12 and to eral vertebral buttressing) with the more rigid
every lumbar vertebra on its course over the pelvic pelvis and rib cage.
ring, but its activation profile (see Juker and col¬
leagues35 for indwelling EMG data of psoas and Ligaments
McGill and associates40 for quadratus lumborum)
is not consistent with that of a spine stabilizer, The column formed by the vertebrae is joined
instead indicating that the role of psoas is purely with two ribbon-like ligaments, the anterior longi¬
as a hip flexor. In contrast, it appears that quadra¬ tudinal and posterior longitudinal ligaments,
tus lumborum is the major stabilizer of the lumbar which assist in restricting excessive flexion and
spine, for two reasons. First, during flexion-domi¬ extension (Fig. 5-13). Both have bony attach¬
nant, extensor-dominant, or lateral-bending tasks, ments to the vertebral bodies and to the anulus.
the quadratus lumborum is always active (e.g., Posterior to the spinal cord is the ligamentum
12% of maximum voluntary contraction (MVC) flavum, which is characterized by a composition
during bent-knee sit-ups, 74% during heavy lifts, of about 80% elastin and 20% collagen, signifying
42% during standing isometric twists, 54% during a special function for this ligament. It has been
side-supported isometric lateral-bending holds40). proposed that this highly elastic structure, which is
Second, in a task in which the subjects stood under pretension throughout all levels of flexion,
upright but held buckets in both hands, and in appears to act as a barrier to material that would
which load was incrementally added to each otherwise encroach on the cord in some regions
bucket, the quadratus lumborum increased its acti¬ of the full range of motion. Furthermore, this
Furthermore, with its oblique line of action, it

protects against posterior shearing of the superior
vertebrae and is implicated in an injury scenario
described later on in this chapter. The supraspi¬
nous ligament, on the other hand, is aligned more
or less parallel to the compressive axis of the
spine, connecting the tips of the posterior spines,
and it appears to provide resistance against exces¬
sive forward flexion.
The facet capsule consists of connective tissue
with bands that restrict joint flexion as well as
distraction of the facet surfaces resulting from
axial twisting. Other ligaments in the thoracolum¬
bar spine include the intertransverse ligaments,
which span the transverse processes and have been
argued by Bogduk and Twomey42 to be sheets of
connective tissue rather than true ligaments. These
authors suggested that the intertransverse liga-
FIGURE 5-12 The oblique muscles transmit force along

their fiber lengths and then redirect force along the rectus
abdominis, through their attachment to the linea semilunaris,
to enhance their effective flexor moment arm. EO, anterior
portion of external oblique; IO, anterior portion of internal
oblique; T, transverse abdominis. (From McGill SM; A revised
anatomical model of the abdominal musculature for torso
flexion efforts. J Biomech 29[7]:973-977, 1996, with
permission from Elsevier Science.)
prestretched elastic structure prevents any sort of

buckling folds that would otherwise impinge the
cord.
The interspinous and supraspinous ligaments
are often classed as a single structure in most
anatomy texts, although functionally they appear
to have different roles. The interspinous ligaments
connect adjacent posterior spines but are not ori¬
ented parallel to the compressive axis of the spine.
Rather, they have a large angle of obliquity (Fig.
5-14).41 Although many anatomy textbooks sug¬
FlGURE 5-13 Major lumbar ligaments. ALL, anterior
gest that this ligament serves to protect against
longitudinal ligament; PLL, posterior longitudinal ligament;
excessive flexion, I disagree. Heylings41 suggested SSL, supraspinous ligament; ISL, interspinous ligament; v,
that the ligament acts like a collateral ligament ventral part; m, middle part; d, dorsal part; LF, ligamentum
similar to the knee, whereby the ligament controls flavum, viewed from within the vertebral canal, and in sagittal
the vertebral rotation to follow an arc throughout section at the midline. (From Bogduk N, Twomey LT: Clinical
the flexion range, which in turn assists the facet Anatomy of the Lumbar Spine. New York, Churchill
joints to remain in contact, gliding with rotation. Livingstone, 1987, p 34.)
ments closest to the disk, causing the earlier stud¬

ies to suggest that the capsular and longitudinal
ligaments may be more important in resisting
flexion than is actually true in vivo. The work of
Sharma and coauthors43 has shown that the major
ligaments for resisting flexion are those making
up the supraspinous complex.
Mechanical failure of the ligaments is a topic
worthy of consideration. King21 noted that soft
tissue injuries are much more common during
high-energy traumatic events, such as automobile
collisions. Our own observations on pig and hu¬
man specimens loaded at slow rates in bending
and shear suggests that excessive tension in the
longitudinal ligaments usually results in avulsion
or bony failure as the ligament pulls some bone
away near its attachment. Noyes and colleagues44
noted that slower strain rates (0.66% per second)
produced more ligament avulsion injuries,
whereas faster strain rates (66% per second) re¬
sulted in more ligamentous failure to the fiber
bundles (in the middle region of the ligament), at
least in monkey knee ligaments. It is interesting,
however, to interpret the clinical report by Rissa-
nen45 that about 20% of cadaveric spines pos¬
sessed visibly ruptured lumbar interspinous liga¬
FIGURE 5-14 The interspinous ligament runs obliquely to ments (in their middle, not at their bony
the compressive axis and thus has limited capacity to check attachment) and that dorsal and ventral portions
flexion rotation of the superior vertebrae. On the contrary, of interspinous, together with supraspinous, liga¬
the interspinous ligament may act as a collateral ligament, ments remained intact.
controlling vertebral rotation and imposing anterior shear Given the oblique fiber direction of the interspi¬
forces on the superior vertebrae. The compressive axis (C) nous complex (see Fig. 5-65), a likely scenario
is indicated, together with the zygopophyseal joint (ZG) and
to damage this ligament would be slipping and
the posterior spines of LI, L2, and L3. (From Heylings DJ:
falling and landing on one’s behind, driving the
Supraspinous and interspinous ligaments of the human lumbar
spine. J Anat I25[I]:I29, 1978. Reprinted with the
pelvis forward on impact, creating a posterior
permission of Cambridge University Press.) shearing of the lumbar joints when the spine is
fully flexed. The interspinous ligament is a major
load-bearing tissue in this example of high-energy
loading, in which anterior shear displacement is
ment-membrane forms a septum between the ante¬ combined with full flexion. Given the available
rior and posterior musculature that is an embryo- data, I believe that damage to the ligaments of the
logic holdover from the development of these two spine during lifting or other normal occupational
sections of muscle. activities, particularly to the interspinous complex,
Determining the roles of ligaments has involved is more uncommon than common. Rather, it ap¬
qualitative interpretation using their attachments pears much more likely that ligament damage oc¬
and lines of action, together with functional tests curs during a more traumatic event, particularly
in which successive ligaments were destroyed and landing on one’s behind during a fall, which then
the joint motion reassessed. Early studies to deter¬ leads to joint laxity and acceleration of subsequent
mine the relative contribution of each ligament to arthritic changes. As has been often said in refer¬
restricting flexion in particular were performed on ence to the knee joint, “ligament damage marks
cadaveric preparations that were not precondi¬ the beginning of the end.”
tioned before testing. This suggests that early data
that described the relative roles of various liga¬
ments were incorrect. For example, upon death, The Lumbodorsal Fascia
the disks, because they are hydrophilic, increase
their water content and consequently their disk Although a functional interpretation of the lumbo¬
height. The “swollen” disks in cadaveric speci¬ dorsal fascia (LDF) is provided later in this chap¬
mens resulted in an artificial preload on the liga¬ ter, a short anatomic description is given here.
First, the transverse abdominis and internal suspect pathology at a specific location but is
oblique muscles obtain their posterior attachment simply noticing normal anatomic asymmetry.
to the fascia, as does the latissimus dorsi over the Joint stiffness values convey the amount of
upper regions of the fascia. The fascia forms a translational and rotational deformation of a spine
compartment around the lumbar extensors section under the application of force or moment.
(multifidus and pars lumborum groups of iliocos- The average stiffness values (Table 5-6)49 docu¬
talis and longissimus) and has been implicated in ment the stiffness of the spine in a neutral posture,
compartment syndrome.46-47 Some have suggested indicating the greater stiffness under compression
that the abdominals work through their fascia at¬ loads than shear loads and the greater stiffness in
tachments to create extension of the spine48; how¬ axial torsion than in rotation around the other two
ever, this notion is highly questionable. Perhaps axes of flexion and extension and lateral bend.
the most tenable explanation for the role of the Generally, range of motion decreases with age,
fascia is that of a large extensor retinaculum to but certain injuries, particularly disk injuries, can
constrain the long tendons of the thoracic and increase the range of motion in bending and shear
lumbar extensors throughout all levels of lordosis. translations,50 a situation that has often been impli¬
cated in subsequent facet joint derangement.51
KINEMATICS AND KINETIC PROPERTIES OF
THE THORACOLUMBAR SPINE Loads on the Low Back During Lifting and
Walking
The ranges of thoracic and lumbar segmental mo¬
tion around the three principal axes (Table 5-5) Low-back loads during lifting result from the mus¬
demonstrate the greater flexion, extension, and cle and ligament tension required to support the
lateral bending capability of the lumbar region posture and facilitate movement. This is why lift¬
and the relatively greater twisting capability of the ing technique is so important to reduce low-back
thoracic region. Although the segmental ranges moment demands and the risk of excessive load¬
shown in the table are population averages, there ing. The following example demonstrates this con¬
is a great amount of variability between subjects cept.
and between segments in a single person. Specifi¬ The components of muscular moment genera¬
cally, there are individual asymmetries in bending tion are detailed in Table 5-7 for the period of
to the right and left, for example, and twisting peak loading in a sample squat lift of 27 kg, which
clockwise and counterclockwise; these are of great provided a reaction moment in the low back of
importance to the clinician, who may sometimes 450 newton-meters (N m) and a total compressive
load of more than 7000 N. The individual muscle
forces, subsequent joint moment, and components
TABLE 5-5 Range of Motion of Each Spine Level of compression and shear that are imposed on
the joint are useful information. In this particular
Lateral Axial example, the lifter avoided full spine flexion, min¬
Level Flexion Extension Bending Twist
imizing ligament and other passive tissue tension
TI-T2 4 6 9 and relegating the moment restoration responsibil¬
T2-T3 4 6 8 ity to the musculature.
T3-T4 4 6 8 Compressive and especially shear components
T4-T5 4 6 8 of muscular force have been greatly neglected
T5-T6 4 6 8 during assessment of injury mechanisms. The very
T6-T7 5 6 8 large magnitude of force in the pars lumborum
T7-T8 6 6 8
laminae results from their large individual cross-
T8-T9 6 6 7
sectional area. These forces produce a large pro¬
T9-TI0 6 6 4
2
portion of the extensor moment. Negative mo¬
TIO-TI 1 9 7
12 9 2 ments observed in Table 5-7 correspond to the
Tl I-TI2
TI2-LI 12 8 2 flexor contributions of abdominal co-contraction.
LI-L2 8 5 6 2 The abdominal co-contraction in this lifting exam¬
L2-L3 10 3 6 2 ple, and in most sagittal plane lifting tasks, was
L3-L4 12 1 8 2 small at the instant of peak extensor moment.
L4-L5 13 2 6 2 The compression penalty from even mild ab¬
L5-SI 9 5 3 5 dominal activity can be observed from the individ¬
ual muscle forces shown in Table 5-7. To meet
All data are from White and Panjabi,108 except for flexion and
extension lumbar data, which are from Pearcy and colleagues109 and
the requirements of the net moment, additional
Pearcy and Tibrewal."0 extensor activity is necessary to offset the flexor
TABLE 5-6 Average Stiffness Values for the Adult Human Spine*
Shear Bending
Axial
Spine Level Compression Anterior/Posterior Lateral Flexion/Extension Lateral Torsion
TI-TI2 1250 86/87 101 155/189 172 149

LI-L5 667 145/143 132 80/166 92 395
L5-SI 1000 78/72 97 120/172 206 264
♦Shear values are given in Newtons per millimeter and bending and axial torsion in newton-meters per radian.
Data on Tl—T2 are from White and Panjabi108; data on LI— L5 are from Schultz and colleagues'" and Berkson and associates"2; data on L5—SI are
from McGlashen and colleagues.113
moment produced by the abdominals. However, observed often reveal that the lifters feel that it
this creates a double contribution to joint compres¬ stiffens the trunk to prevent buckling of the spine.
sion: compression from abdominal activity and This idea has been tested by Cholewicki and
compression from the additional extensor forces. McGill52 and is discussed in a later section of
Even so, when all of the component forces are this chapter.
summed, the total predicted joint compression is Negative shear forces from the muscles (shown
less than what would have been predicted by a in Table 5-7) correspond to L4 shearing posteri¬
simple analysis using a single equivalent muscle orly on L5. Hence, a powerful anti-anterior shear
model (5-cm extensor moment arm has been used mechanism is observed, in the tabulated forces,
in the past). Obviously, abdominal activity would owing to the obliquity of the pars lumborum ex¬
result in a shorter equivalent moment arm. The tensors. These muscles help to offset the anterior
ability of an individual to reduce compression reaction shear force from lifting a load when they
appears to be determined by the degree of abdomi¬ are activated presumably to contribute extensor
nal activity. As is often observed in elite lifters, moment.33 The implication of these forces is a
however, the abdominals are not completely unin¬ reduced load on the facet joints. Some subjects
volved, exhibiting varying degrees of activity. whom we have tested have offset the reaction
This suggests that they are sacrificing minimum shear force almost completely, depending on the
compression for some other benefit. Interviews forward inclination of the disk (and trunk) and on
with some elite lifters about why co-contraction is the magnitude of force in these obliquely orien-
TABLE 5-7 Musculature Components for Moment Generation of 450 N m During Peak Loading
for a Squat Lift of 27 kg
Muscle Force (N) Moment (N m)* Compression (N) Shear(N)*
Rectus abdominis 25 -2 24 5
External oblique 1 45 1 39 24
External oblique 2 43 -2 30 31
Internal oblique 1 14 1 14 -2
Internal oblique 2 23 - 1 17 - 16
Longissimus thoracis pars lumborum (L4) 862 35 744 -436
Longissimus thoracis pars lumborum (L3) 1514 93 1422 -518
Longissimus thoracis pars lumborum (L2) 1342 121 1342 0
Longissimus thoracis pars lumborum (LI) 1302 NO 1302 0
lliocostalis lumborum pars thoracis 369 31 369 0
Longissimus thoracis pars thoracis 295 25 295 0
Quadratus lumborum 393 16 386 74
Latissimus dorsi (L5) 1 12 6 79 -2
Multifidus 1 136 8 134 18
Multifidus 2 226 8 189 124
Psoas (LI) 26 0 23 12
Psoas (L2) 28 0 27 8
Psoas (L3) 28 1 27 6
Psoas (L4) 28 1 27 5
♦Negative moments correspond to flexion, whereas negative shear corresponds to L4 shearing posteriorly on L5.
tated pars lumborum fibers.53 The clinical implica¬ of this area was 3.8 cm anterior to the center of
tions of these shear forces are discussed in a the T12 disk (compare with a 511-cm2 pelvic floor
subsequent section of this chapter. and 465-cm2 diaphragm, together with moment
During walking every day, thousands of low- arm distances up to 11.4 cm, which were outside
level loading cycles are endured by the spine. of the chest in most subjects used in other studies).
Although the small loads in the low back during During squat lifts, it appears that the net effect of
walking suggest a noninjurious activity, walking the involvement of the abdominal musculature and
has been found to provide relief to some persons IAP is to increase compression rather than allevi¬
but is painful to others—particularly fast walking ate joint load. (A detailed description and analysis
as opposed to strolling. The compressive loads of of the forces can be found in McGill and Nor¬
about 2.5 times body weight and the shear forces man63). This predicted finding agrees with experi¬
are well below any known in vitro failure load. mental evidence of Krag and colleagues,60 who
Strolling, however, reduces spine motion and pro¬ used EMG, and with Nachemson and associates,62
duces static loading of tissues, whereas faster who documented increased intradiscal pressure
walking with arm swinging causes cyclic loading with an increase in IAP.
of tissues,54 which may begin to explain the relief The generation of appreciable IAP during load¬
experienced by some patients who undertake this handling tasks is well documented, but the role of
activity. IAP is not. Farfan7 has suggested that IAP creates
a pressurized visceral cavity to maintain the hoop¬
ANATOMIC CONSISTENCY IN EXAMINING like geometry of the abdominals. Recent work
THE ROLE OF INTRAABDOMINAL PRESSURE measuring the distance of the abdominals to the
spine (their moment arms) was unable to confirm
It has been claimed for many years that intraab¬ substantial changes in abdominal geometry when
dominal pressure (IAP) plays an important role in activated in a standing posture.31 The compression
support of the lumbar spine, especially during penalty of abdominal activity, however, cannot be
strenuous lifting. Anatomic accuracy has been in¬ discounted. It appears that the spine prefers to
fluential in this debate. This issue has been consid¬ sustain increased compression loads if intrinsic
ered in lifting mechanics for years and, for some, stability is increased. An unstabilized spine buck¬
has formed a cornerstone for prescription of ab¬ les under extremely low compressive load (e.g.,
dominal belts to industrial workers and has moti¬ about 20 N).64 The geometry of the musculature
vated various abdominal strengthening programs. suggests that individual components exert lateral
Some research reports suggest that IAP may be a and anteroposterior forces on the spine that per¬
mechanism to reduce lumbar spine compression haps can be thought of as guy wires on a mast to
directly.55'56 Some, however, have indicated that prevent bending and compressive buckling.52 In
they believe the role of IAP in reducing spinal addition, activated abdominals create a rigid cylin¬
loads has been overemphasized.57-59 In fact, some der of the trunk, resulting in a stiffer structure.
experimental evidence suggests that somehow, in Thus, it appears that increased IAP, commonly
the process of building up IAP, the net compres¬ observed during many activities, including lifting,
sive load on the spine is increased! Increased low- as well as in patients with back pain, does not
back EMG activity with higher IAP was noted by have a direct role in reducing spinal compression
Krag and coauthors60 during voluntary Valsalva but rather is an agent used to stiffen the trunk and
maneuvers. Nachemson and colleagues61’ 62 prevent tissue strain or failure from buckling.
showed an increase in intradiscal pressure during
a Valsalva maneuver, indicating a net increase WHAT IS THE ROLE OF THE LUMBODORSAL
in spine compression with an increase in IAP, FASCIA?
presumably a result of abdominal wall muscula¬
ture activity. Recent studies have attributed various mechanical
In our own investigations, which used an ana¬ roles to the LDF. In fact, there have been some
tomically detailed modeling approach, an evalua¬ attempts to recommend lifting postures based on
tion of the net spine compression benefit and pen¬ LDF hypotheses. Suggestions were originally
alty to build up IAP and produce concomitant made that lateral forces generated by internal
abdominal activity was performed. The size of oblique and transverse abdominis are transmitted
the cross-sectional area of the diaphragm and the to the LDF through their attachments to the lateral
moment arm used to estimate force and moment border, claiming that the fascia could support sub¬
produced by LAP have a major effect on conclu¬ stantial extensor moments.48 This lateral tension
sions reached about the role of IAP.63 The dia¬ was hypothesized to increase longitudinal tension,
phragm surface area was 243 cm2, and the centroid from Poisson’s effect, pulling in the direction of
the posterior midline of the lumbar spine and require lower loads to reduce the risk, in other
causing the posterior spinous processes to move sedentary occupations, the risk can be better re¬
together, resulting in lumbar extension. This pro¬ duced with more loading and with varying the
posed sequence of events formed an attractive nature of the loading. To decide which is best,
proposition because the LDF has the largest mo¬ the clinician must understand the biomechanics
ment arm of all the extensor tissues. As a result, of injury.
any extensor forces within the LDF would impose
the smallest compressive penalty to vertebral com¬ MECHANICAL LOADING AND THE PROCESS
ponents of the spine.
OF INJURY
This hypothesis was examined by three studies,
all published about the same time, which collec¬ A generic scenario for injury is presented first, and
tively questioned its viability: Tesh and col¬ references for injury from repeated and prolonged
leagues,65 who performed mechanical tests on ca¬ loading to specific tissues are provided in the
daveric material; Macintosh and associates,66 who following section. The purposes of this section are
recognized the anatomic inconsistencies with the to motivate consideration of the many factors that
abdominal activation; and McGill and Norman,67 modulate the risk of tissue failure and to generate
who tested the viability of LDF involvement with hypotheses to probe the causes of injury.
latissimus dorsi as well as with the abdominals. Injury, or failure of a tissue, occurs when the
Regardless of the choice of LDF activation strat¬ applied load exceeds the failure tolerance or
egy, the LDF contribution to the restorative exten¬ strength of the tissue. For the purposes of this
sion moment was negligible compared with the chapter, injury is defined as the full continuum
much larger low-back reaction moment required from the most minor of tissue irritations (but mi¬
to support the load in the hands. crotrauma nonetheless) to the grossest of tissue
Although the LDF does not appear itself to be failures, for example, vertebral fracture or liga¬
a significant active extensor of the spine, it is a ment avulsion. I proceed on the premise that such
strong tissue with a well-developed lattice of col¬ damage generates pain. Obviously, a load that
lagen fibers. Its function may be that of an exten¬ exceeds the failure tolerance of the tissue, applied
sor muscle retinaculum.68 The tendons of longissi- once, produces injury (e.g., the snowmobiler, air¬
mus thoracis and iliocostalis lumborum pass under borne and about to experience an axial impact
the LDF to their sacral and ilium attachments. with the spine fully flexed, is at risk of posterior
Perhaps the LDF provides a form of “strapping” disk herniation on landing). This injury process is
for the low-back musculature. Hukins and coau¬ depicted in Figure 5-15, in which a margin of
thors,69 on theoretical grounds only at this time, safety is observed in the first cycle of subfailure
have proposed that the LDF acts to increase by up load. In the second loading cycle, the applied load
to 30% the force per unit cross-sectional area that increases in magnitude, simultaneously decreasing
muscle can produce. They suggest that it does this the margin of safety to zero, and injury occurs.
by constraining bulging of the muscles when they Although this description of low-back injury is
shorten. This contention remains to be proved. common, particularly among medical community
Tesh and colleagues65 have suggested that the LDF members who are required to identify an event
may be more important for supporting lateral when completing injury reporting forms or work¬
bending. No doubt, this notion will be pursued in ers’ compensation reports, I believe that relatively
the future. Given the confused state of knowledge few low-back injuries occur in this manner. (More
about the role, if any, of the LDF, the promotion detail on the types of loads which create injury is
of movement strategies based on intentional LDF provided in the next section.)
involvement, for either low-back pain patients or Some more likely scenarios that result in injury,
healthy people, cannot be justified at this time. when considering occupational and athletic en¬
deavors, involve cumulative trauma from subfail¬
ure magnitude loads. In such cases, injury is the
Biomechanics of Low-Back Injury result of accumulated trauma produced by either
the repeated application of a relatively low load
Reducing the risk of low-back injury, as perceived or the application of a sustained load for a long
by many clinicians, engineers, and ergonomists, is duration (as in sitting). Figure 5-16 shows a per¬
thought to involve the reduction of applied loads son loading boxes on a pallet, repeatedly loading
to the various anatomic components at risk of the tissues of the low back (several tissues could
injury. This is an overly simplistic view—optimal be at risk) to a subfailure level and causing a
tissue health requires an envelope of loading, not slow degradation of their failure tolerance (e.g.,
too much or too little. Although some occupations vertebrae5'l9). As the margin of safety approaches
Force
FIGURE 5-15 A, The snowmobile driver is about to experience an axial compressive impact load to a
fully flexed spine. One-time application of load can reduce the margin of safety to zero as the applied load
exceeds the strength or failure tolerance of the supporting tissues (small arrow in B). (From McGill SM:
Biomechanics of low back injury: Implications on current practice and the clinic. J Biomech 30[5]:465—475,
1997.)
zero, this person experiences low-back injury. Ob¬ riod of time, are loading the posterior passive
viously, the accumulation of trauma is more rapid tissues and initiating changes in disk mechanics.
with higher loads70; however, at least with bone, The sustained load causes a progressive reduction
fatigue failure occurs with fewer repetitions when in the margin of safety, whereby injury is associ¬
the applied load is closer to the yield strength. ated with the nth percentage of tissue strain. Analy¬
Yet another way to produce injury with a sub¬ sis of injury is further complicated by the interac¬
failure load is to induce stresses over a sustained tion between the various tissues in the low back.
period of time. For example, rodmen (Fig. 5-17), For example, a prolonged, stooped posture loads
with their spines fully flexed for a prolonged pe¬ the posterior ligaments of the spine and posterior
FIGURE 5-16 Repeated subfailure loads lead to tissue fatigue (A), reducing the failure tolerance
and resulting in failure on the nth repetition of load (B) (or box lift in this example). (From McGill
SM: Biomechanics of low back injury: Implications on current practice and the clinic. J Biomech
30[5]:465—475, 1997.)
FIGURE 5—17 A, These “rodmen” are loading posterior passive tissues for a long duration, reducing
the failure tolerance leading to failure at the nth percentage of tissue strain (arrow in B indicates where
the margin of safety has reduced to zero and the injury occurs). (From McGill SM: Biomechanics of
low back injury: Implications on current practice and the clinic. J Biomech 30[5]:465-475, 1997.)
fibers of the intervertebral disk, causing creep AN EXAMPLE OF INJURY

deformation, possibly to the point of microfail¬
ure.71-72 This could initiate another chain of events: The work that many have reported over recent
“stretched” ligaments increase joint laxity, in¬ years has demonstrated the extraordinary magni¬
creasing the risk of hyperflexion injury (to the tudes of forces within the various components of
disk) as well as the risk of local instability, leading the trunk musculature, even during nonstrenuous
tasks. Although these forces have been interpreted
to injury of unisegmental structures and to ever-
for their mechanical role, clinicians have ex¬
increasing shearing and bending loads on the neu¬
pressed interest in their potential to cause injury.
ral arch. It would appear that the most appropriate
Damage to bony attachments remains a possibility
injury intervention strategies must appreciate the
complexities of tissue overload. that may be wrongfully attributed to alternative
mechanisms. One such example follows.
The objective of injury avoidance strategies is
Pain in the sacroiliac region is common and
to ensure that tissue adaptation, stimulated by ex¬
often attributed to disorders of the sacroiliac joint
posure to load, keeps pace with, and ideally ex¬
or to the iliolumbar ligament.73’74 For this reason,
ceeds, the accumulated tissue damage. Thus, expo¬
the role of the musculature may have been ne¬
sure to load is necessary, but in the process of
glected. A large proportion of the extensor muscu¬
accumulating microtrauma, the applied loads must
lature has its origin in the sacroiliac and posterior
be removed to allow the healing and adaptation
superior iliac spine region.26 The area of tendon-
process to increase the failure tolerance gradually
periosteum attachment and extensor aponeurosis
to the necessary level. Tissue loading and the risk
is relatively small in relation to the volume of
of injury form an optimum U-shaped relationship,
muscle in series with the tendon complex. From
whereby the determination of the safety optimum
this, a hypothesis evolved that the seeming mis¬
for individual tissue loading encompasses both the
match of large muscle tissue to small attachment
art and science of medicine and biomechanics.
area for connective tissue places the connective
In summary, the injury process need be associ¬
tissue at high risk of sustaining microfailure, re¬
ated not only with very high loads but also with
sulting in pain.75 Knowledge of the collective mus¬
relatively low loads that are repeated or sustained,
cle force-time histories enables speculation about
justifying the need for consideration of injury and
one-time failure loads and cumulative trauma. For
tissue loading experienced by an individual for
example, if the forces of muscles that originate in
substantial periods of time before the culminating
the sacroiliac region are tallied for the trial illus¬
injury event. Simply focusing on a single variable,
trated in Table 5-7, the total force transmitted to
such as one-time load magnitude, may not result the sacroiliac region during peak load exceeds 5.6
in a successful index of risk of injury, particularly kiloNewtons (kN). Such a load would lift a small
across a wide variety of activities. car off the ground!
The failure tolerance of these connective tissues tissues and muscles that ultimately modulates the
is not known, which makes speculation about the risk of several types of injury. For example, if a
potential for microfailure difficult. No doubt, the subject holds a load in the hands with the spine
risk of damage must increase with the extremely flexed enough to achieve myoelectric silence in
large loads observed in the extensor musculature the extensors (reducing their tension), and with all
and with the frequency of application. Task analy¬ joints held still so that the low-back moment re¬
sis of many industrial tasks has documented that mains the same, the recruited ligaments appear to
lifting three containers in excess of 18 kg per add to the anterior shear to levels well over 1000
minute during an 8-hour day is not an unusual N, which is of great concern from an injury risk
task, suggesting there is significant potential for viewpoint (Fig. 5-18). However, a more neutral
cumulative trauma. lordotic posture is adopted, and the extensor mus¬
This mechanical explanation may account for culature is responsible for creating the extensor
local tenderness on palpation associated with most moment and at the same time will support the
cases of sacroiliac syndrome. In addition, muscle anterior shearing action of gravity on the upper
strain and spasm often accompany sacroiliac pain. body and hand-held load. Disabling the ligaments
Nonetheless, treatment is often directed toward greatly reduces shear loading (Table 5-8). In this
the articular joint despite the extreme difficulty in example the spine is at much greater risk of sus¬
diagnosing the joint as the primary source of pain. taining shear injury (>1000 N) than compressive
Although reduction of spasm through conven¬ injury (3000 N), simply because the spine was
tional techniques would reduce the sustained load flexed, or in a position at the end range of motion
on the damaged connective fibers, patients should (for a more comprehensive discussion see Potvin
be counseled on techniques to reduce internal and colleagues,53 McGill and Norman,76 and
muscle loads through effective lifting mechanics. McGill and Kippers77). This example also illus¬
This example of sacroiliac syndrome, just one of trates the need to consider more loading modes
many, of which there may be several, illustrated than simple compression—in this case the real
how knowledge of individual muscle force-time risk was anteroposterior shear load.
histories suggested a mechanism for injury for
which a specific treatment modality would be pre¬ CHANGES IN THE FUNCTION OF THE SPINE
scribed. THROUGHOUT THE DAY
The task of reducing injury becomes more in¬

Using Biomechanics for Better volved as the biomechanical function of the spine
Prevention of Low-Back Injury changes throughout the day depending on the pre¬
vious tasks and the time since rising from bed.
LUMBAR POSTURE: A GENERAL The diurnal variation in spine length together with
CONSIDERATION FOR INJURY AVOIDANCE the ability to flex forward has been well docu¬
mented. Losses in sitting height over a day have
A generalization that appears to have justification been measured at up to 19 mm by Reilly and
based on knowledge of how injury occurs is that coworkers,78 who also noted that approximately
a neutral spine (at least avoiding end range of 54% of this loss occurred in the first 30 minutes
motion) reduces the risk of many of the injuries after rising. Over the course of a day, and de¬
listed previously. Many injuries are associated pending on the task history, hydrostatic pressures
with the spine at end range of motion. cause a net outflow of fluid from the disk, re¬
The following example demonstrates the shifts sulting in narrowing of the space between the
in tissue loading, predicted from our modeling vertebrae, which in turn reduces tension in the
approach, which has dramatic effects on shear ligaments. When a person lies down at night,
loading of the intervertebral column and the resul¬ osmotic pressures in the disk nucleus exceed the
tant injury risk. First, the dominant direction of hydrostatic pressure, causing the disk to expand
the pars lumborum fibers of longissimus thoracis (and the spine to lengthen). Adams and col¬
and iliocostalis lumborum have been noted earlier leagues79 noted that the range of lumbar flexion
(see Figs. 5-10 and 5-11), producing a posterior increased by 5 degrees throughout the day. The
shear force on the superior vertebra. In contrast, increased fluid content after a person rises from
the interspinous ligament complex generates bed causes the lumbar spine to be more resistant
forces with the opposite obliquity to impose an to bending, while the musculature does not appear
anterior shear force on the superior vertebra (see to compensate by restricting the bending range.
Fig. 5-14). This is one example of how spine Adams and colleagues estimated that disk bending
posture determines the interplay between passive stresses increased by 300% and ligament stresses
n 0 1 2 3 4 5
Time (s)
FIGURE 5-18 The fully flexed spine (A) is associated with myoelectric silence in the back extensors
and strained posterior passive tissues and high shearing forces on the lumbar spine (from both reaction
shear on the upper body and interspinous ligament strain; see Fig. 5-14). A more neutral spine posture
B) recruits the pars lumborum muscle groups (large activity shown in C) to support the reaction shear
U$ redUCin§ t0ta' j°int Shear (t° about 200 N in this example) shown in D. (From
°f 'OW b3Cl< in'Ury: lmplications on current Pîce and the clinic, j Biomech
126
TABLE 5-8 Individual Muscle and Passive Tissue Forces and the Associated Lumbar Moments, Compression,
and Shear Forces*
Fully Flexed Lumbar Spine

Neutral
Moment (N m) Shear (N)
Lumbar Spine
Force Compression
(N) Flexion Lateral Twist (N) Anteroposterior Lateral Force(N)
Muscle
R. Rectus abdominis 16 -2 1 1 15 5 -4 39
L. Rectus abdominis 16 -2 - | - 1 15 5 4 62
R. External oblique 1 10 - 1 1 1 8 7 -3 68
L. External oblique 1 10 - 1 - 1 -1 8 7 3 40
R. External oblique 2 7 - 1 1 0 6 2 -3 62
L. External oblique 2 7 - 1 - 1 0 6 2 3 31
R. Internal oblique 1 35 0 3 -2 21 - 19 20 130
L. Internal oblique 1 35 0 -3 2 21 - 19 -20 102
R. Internal oblique 2 29 -2 2 -3 8 - 17 21 1 16
L. Internal oblique 2 29 -2 -2 3 8 - 17 -21 88
R. Pars lumborum (LI) 21 2 1 0 21 6 2 253
L. Pars lumborum (LI) 21 2 - 1 0 21 6 -2 285
R. Pars lumborum (L2) 27 2 1 0 26 8 2 281
L. Pars lumborum (L2) 27 2 - 1 0 26 8 -2 317
R. Pars lumborum (L3) 31 1 1 0 29 -4 6 327
L. Pars lumborum (L3) 31 1 -I 0 29 -4 -6 333
R. Pars lumborum (L4) 32 1 1 0 30 -7 6 402
L. Pars lumborum (L4) 32 1 - 1 0 30 -7 -6 355
R. Iliocostalis lumborum 58 5 4 1 57 14 - 1 100
L. Iliocostalis lumborum 58 5 -4 -1 57 14 1 137
R. Longissimus thoracis 93 7 4 0 91 23 -6 135
L. Longissimus thoracis 93 7 -4 0 91 23 6 179
R. Quadratus lumborum 25 1 2 0 25 - 1 i 155
L. Quadratus lumborum 25 1 -2 0 25 -1 -1 194
R. Latissimus dorsi (LS) 15 1 1 0 14 - 1 -6 101
L. Latissimus dorsi (LS) 15 1 - 1 0 14 - 1 6 1 15
R. Multifidus 1 28 1 1 1 26 6 9 80
L. Multifidus 1 28 1 - I -1 26 6 -9 102
R. Multifidus 2 28 1 1 0 28 6 0 87
L. Multifidus 2 28 1 - I 0 28 6 0 90
R. Psoas (LI) 25 1 2 0 24 0 6 61
L. Psoas (LI) 25 - 1 -2 0 24 0 -6 69
R. Psoas (L2) 25 - 1 2 0 24 0 6 62
L. Psoas (L2) 25 0 -2 0 24 0 —6 69
R. Psoas (L3) 25 0 1 0 24 0 7 62
L. Psoas (L3) 25 0 - 1 0 24 0 -7 69
R. Psoas (L4) 25 0 1 1 24 0 8 61
L. Psoas (L4) 25 0 - 1 -1 24 0 -8 69
Ligament
Anterior longitudinal 0 0 0 0 0 0 — 0
Posterior longitudinal 86 2 0 0 261 44 — 0
Ligamentum flavum 21 1 0 0 21 2 3
R. Intertransverse 14 0 0 0 13 3 — 0
L. Intertransverse 14 0 0 0 13 3 — 0
74 2 1 1 65 40 — 0
R. Articular
74 2 -1 -1 65 40 — 0
L. Articular
R. Articular 2 103 3 2 2 84 -3 — 0
103 3 -2 -2 84 -3 — 0
L. Articular 2
301 18 0 0 273 142 — 0
Interspinous 1
345 14 0 0 233 268 0
Interspinous 2
298 10 0 0 194 238 0
Interspinous 3
592 41 0 0 591 79 u
Supraspinous
R. Lumbodorsal fascia 122 8 1 0 109 — I
122 - 1 0 109 - 1 0
L. Lumbodorsal fascia
Passive tissue
— 9 0 0 — — — 1
Disk
0 0 1
Gut, etc. — 1 1
^During full flexion, with just the forces in a more neutral lumbar posture demonstrating the shift from muscle to passive tissue and the resultant
effects on joint compression and shear. The extensor moment with full lumbar flexion was 171 N m, producing 3 145 N of compression and 954
N of anterior shear; in the more neutral posture of 170 N m, the forces produced 3490 N of compression and 269 N of shear.
increased by 80% (owing to the swollen disks prolonged full flexion may cause the posterior
and longer spine) and concluded that there is an ligaments to creep, which may allow damaging
increased risk of injury to these tissues when one flexion postures to go unchecked if lordosis is not
bends forward when first arising. This risk is a controlled during subsequent lifts. The data of
variable that is modulated with disk hydration McGill and Brown,82 in a study of posterior pas¬
level, which is directly linked to the activity his¬ sive tissue creep while sitting in a slouched pos¬
tory and the time since rising from bed. ture, showed that during the 2 minutes after 20
minutes of full flexion, subjects regained only half
MEMORY OF THE SPINE of their intervertebral joint stiffness, and even after
30 minutes of rest, some residual joint laxity re¬
The function of the spine is modulated by a pa¬ mained. Thus, the spine has a memory. This is of
tient’s activity history in terms of disk hydration particular importance for those people whose work
(explained in the previous paragraph) as well as is characterized by cyclic bouts of full end range-
viscoelastic creep in the disk and ligaments. In of-motion postures followed by exertion. Before
effect, the loading of any activity determines disk lifting exertions after a stooped posture, or after
hydration (and therefore the size of the disk space prolonged sitting, a case could be made for stand¬
and disk geometry), which, in turn, modulates ing or even consciously extending the spine for a
ligament rest length, joint mobility, stiffness, and short period. Allowing the nuclear material to
load distribution. It has been proposed that the “equilibrate” or move anteriorly to a position
nucleus within the anulus “migrates” anteriorly associated with normal lordosis may decrease
during spinal extension and posteriorly during forces on the posterior nucleus. Ligaments will
flexion.80 McKenzie’s program of passive exten¬ regain some protective stiffness.
sion of the lumbar spine (which is presently popu¬ In conclusion, the anatomy and geometry of the
lar in physical therapy) was based on the supposi¬ spine is not static. Much remains to be accom¬
tion that an anterior movement of the nucleus plished before we will understand the importance
pulposus would decrease pressure on the posterior of tissue loading history on subsequent biomecha¬
portions of the anulus fibrosus, the most problem¬ nics, rehabilitation therapies, and injury mechan¬
atic site of herniation. Because of the viscous ics.
properties of the nuclear material, such reposition¬
ing of the nucleus does not occur immediately
THE ANATOMIC FLEXIBLE BEAM AND TRUSS:
upon postural change but rather takes time.
MUSCLE CO-CONTRACTION AND SPINE
Krag and colleagues81 demonstrated anterior STABILITY
movement, albeit minute, from an elaborate exper¬
iment that placed radiopaque markers in the nu¬ The ability of the joints of the lumbar spine to
cleus of cadaveric lumbar motion segments. bend in any direction is accomplished with large
Whether this observation reflects simply the redis¬ amounts of muscle coactivation. These coactiva¬
tribution of the centroid of the wedge-shaped nu¬ tion patterns are counterproductive to generating
clear cavity moving forward with flexion or migra¬ the torque necessary to support the applied load
tion of the whole nucleus remains to be seen. in a way that minimizes the load penalty imposed
Nonetheless, hydraulic theory would suggest on the spine from muscle contraction. Several
lower bulging forces on the posterior anulus if the ideas have been postulated to explain muscular
nuclear centroid moved anteriorly during exten¬ coactivation: the abdominals are involved in the
sion. If compressive forces were applied to a disk generation of intraabdominal pressure83 or in pro¬
in which the nuclear material was still posterior viding support forces to the lumbar spine through
(as in lifting immediately after a prolonged period the LDF.48 These ideas, however, have not been
of flexion), a concentration of stress would occur without opposition.63
on the posterior anulus, modulating the risk of in¬ It appears that another explanation for muscular
jury.
coactivation is tenable. A ligamentous spine will
Although this specific area of research needs fail under compressive loading in a buckling
more development, there does appear to be a time mode, at about 20 N64; in other words, a bare
constant (indicating viscosity) associated with this spine is unable to bear compressive load. The
redistribution of nuclear material, suggesting that spine can be likened to a flexible rod in that under
it would be unwise to lift an object immediately compressive loading, it will buckle. If the rod has
after prolonged flexion—such as when sitting, or guy wires connected to it, like the rigging on a
stooping, as would a stooped gardener who may ship mast, the rod ultimately experiences more
stand erect and lift a heavy object. Furthermore, compression but is able to bear much more com¬
it was suggested by Adams and coauthors79 that pressive load because it is stiffened and more
resistant to buckling. The co-contracting muscula¬ error that results in a temporary reduction in acti¬
ture of the lumbar spine (the flexible beam) can vation to one of the intersegmental muscles—for
perform the role of stabilizing guy wires (the example, a lamina of longissimus, iliocostalis, or
truss) to each lumbar vertebra bracing against multifidus—could allow rotation at just a single
buckling. Work by Crisco and Panjabi84 has begun joint to the point where passive, or other, tissue
to quantify the influence of muscle architecture could become irritated or even more traumatically
and the necessary coactivation on stability of the injured. Cholewicki noted that the risk of such an
lumbar spine. The architecture of the lumbar erec¬ event is greatest when there are high forces in the
tor spinae is especially suited for this role.26 To large muscles with simultaneous low forces in the
invoke this antibuckling and stabilizing mecha¬ small intersegmental muscles (a possibility in the
nism when lifting, one could justify lightly co¬ case of the power lifter) or when all muscle forces
contracting the musculature to minimize the po¬ are low, such as during a low-level exertion. Thus,
tential of spine buckling. a mechanism is proposed, based on motor control
error resulting in temporary inappropriate neural
MORE ON STABILITY AND MOTOR activation, that explains how injury might occur
CONTROL: HOW DO PEOPLE HURT THEIR during extremely low load situations, for example,
BACKS WHILE PICKING UP A PENCIL? picking up a pencil from the floor after a long day
at work performing a very demanding job.
Although injury from large exertions is under¬
standable, explanation of how people injure their PREVENTING INJURY: WHAT DOES THE
backs performing rather benign-appearing tasks is WORKER NEED TO KNOW?
more difficult—but the following is worth consid¬
ering. Continuing the considerations about stabili¬ Workers are often told to bend the knees and keep
zation from the previous paragraph, a number of the back straight in an effort to minimize the risk
years ago, we were investigating the mechanics of of injury. This demonstrates an overly simplistic
power lifters’ spines while they lifted extremely view of injury prevention because very few jobs
heavy loads using videofluoroscopy for a sagittal can be performed this way and it is physiologi¬
view of the lumbar spine.85 The range of motion cally costly.86 Rather, work may be designed to
of the power lifters’ spines was calibrated and incorporate some of the principles developed in
normalized to full flexion by first asking them to the previous sections of this chapter; for example,
flex at the waist and support the upper body avoid end range of lumbar motion, design work
against gravity with no load in the hands. During to vary so that loads are rotated among the various
their lifts, although they outwardly appeared to supporting tissues to minimize the risk of accumu¬
have a flexed spine, in fact, the lumbar joints lated deformation, allow time for tissues to restore
were 2 to 3 degrees per joint from full flexion, their unloaded-rested geometry after the applica¬
explaining how they could lift such magnificent tion of prolonged loads when creep has occurred
loads (up to 210 kg) without sustaining the injuries before performing demanding tasks, avoid pro¬
that we suspect are linked with full lumbar flexion. longed sitting, and keep the loads close to the low
During the execution of a lift, however, one lifter back. (A much more developed list may be found
reported discomfort and pain. On examination of in McGill and Norman.87)
the videofluoroscopy records, one of the lumbar
joints (specifically, the L4-L5 joint) reached the
full flexion calibrated angle, while all other joints Using Biomechanics to Build
maintained their static position (2 to 3 degrees
Better Rehabilitation Programs
from full flexion). This is the first report in the
FOR PATIENTS WITH LOW-BACK
scientific literature that we know of documenting
proportionately more rotation occurring at a single Injury
lumbar joint, and it would appear that this occur¬
rence was due to an inappropriate sequencing of Rehabilitation of the injured low back involves
muscle forces (or a temporary loss of motor con¬ exercise prescription to stress both damaged tissue
trol wisdom). and healthy supporting tissues and to foster repair;
This motivated the work of our colleague and the key is to avoid excessive loading, which can
former graduate student, Dr. Jacek Cholewicki, to exacerbate existing structural weakness. Once
investigate and continuously quantify stability of again, blending the understanding of the biome¬
the lumbar spine throughout a reasonably wide chanics of injury with clinical art assists the clini¬
variety of loading tasks.52 Generally speaking, it cian in choosing the optimal load—neither too
appears that the occurrence of a motor control much nor too little.
It is important to understand the applications, bent knees; compression, 3230 N with straight
and conversely the limitations, of scientific labora¬ legs, 3410 N with bent knees; shear, 260 N with
tory approaches for investigating tissue loading in straight legs, 300 N with bent knees). Compres¬
vivo and in vitro. Because the low-back system is sive loads in excess of 3000 N certainly raise
an extremely complex mechanical structure, and questions of safety. This type of quantitative anal¬
direct measurement of tissue forces in vivo is not ysis is necessary to demonstrate that the issue of
feasible, the only tenable option for tissue load performing sit-ups using bent knees or straight
prediction is to use sophisticated modeling ap¬ legs is probably not as important as the issue of
proaches. Several issues must be addressed, how¬ whether to prescribe sit-ups at all! There are better
ever, including the need for anatomic detail, a ways to challenge the abdominals.
method to solve for the inherent indeterminacy Several exercises are required to train all the
from so many unknown forces among the signifi¬ muscles of the lumbar torso, and the exercises that
cant load-bearing structures, and development of best suit the individual depend on a number of
methods that enable the prediction of loads in variables, such as fitness level, training goals, and
deep (and inaccessible) muscles and supporting history of previous spinal injury. Depending on
ligaments. Although these issues are outside the the purpose of the exercise program, however,
scope of this chapter, the interested reader is urged several principles apply. For example, an individ¬
to consult Juker and colleagues,35 McGill,39 and ual beginning a postinjury program is better ad¬
Cholewicki and McGill52 for a description of the vised to avoid loading the spine throughout the
scientific methods used to develop the following range of motion, whereas a trained athlete may
program.
indeed achieve higher performance levels by do¬
ing so. Selection of the following exercises was
TOWARD DEVELOPING SCIENTIFICALLY biased toward safety—minimizing spine loading
JUSTIFIED LOW-BACK REHABILITATION during muscle challenge. Therefore, a neutral
EXERCISES spine (neutral lordosis) is emphasized while the
spine is under load—neither hyperlordotic or hy-
Exercises in this section have been selected and polordotic. A general rule is to preserve the nor¬
evaluated based on tissue-loading evidence and mal low-back curve (similar to that of upright
the knowledge of how injury occurs to specific standing) or some variation that minimizes pain.
tissues (described in the original scientific publica¬ Although in the past performing a “pelvic tilt”
tions35 88 89). In fact, some integrated components when exercising has been recommended, this is
of this section were adapted from my chapter in not justified because the pelvic tilt increases spine
The American College of Sports Medicine text¬ tissue loading when the spine is no longer in
book Resource Manual for Guidelines for Exercise static-elastic equilibrium; therefore, it is probably
Testing and Prescription.90
unwise to recommend the pelvic tilt when chal¬
Choosing exercises has not always been a scien¬ lenging the spine.
tific endeavor. The following example illustrates
the need for quantitative analysis for evaluating
ISSUES OF FLEXIBILITY
the safety of certain exercises. We have all been
aware of the principle to perform sit-ups and other Training to optimize spine flexibility depends on
flexion exercises with the knees flexed—but on the person’s injury history and exercise goal. Gen¬
what evidence? Several hypotheses have sug¬ erally, for the patient with a back injury, spine
gested that this disables or changes the line of flexibility should not be emphasized until the
action of the psoas. Recent MRI-based data91 dem¬
spine has stabilized and has undergone strength
onstrated that the psoas line of action does not
and endurance conditioning—some patients may
change as a result of lumbar or hip posture (except
never reach this stage! Despite the notion held by
at L5-S1) because the psoas laminae attach to
some, there are little quantitative data to support
each vertebra and “follow” the changing orienta¬
a major emphasis on trunk flexibility to improve
tion of spine. There is no doubt, however, that the
back health and lessen the risk of injury. In fact,
psoas is shortened with the flexed hip, modulating
some exercise programs that have included load¬
force production. But the question remains: is
ing of the torso throughout the range of motion
there a reduction in spine load with the legs bent?
(in flexion and extension, lateral bend, or axial
In a recent study,92 I examined 12 young men,
twist) have had negative results,93’94 and greater
with the laboratory technique described pre¬
spine mobility has been, in some cases, associated
viously, and observed no major difference in lum¬
with low-back trouble.95’96 Furthermore, flexibility
bar load as the result of bending the knees (aver¬
of the spine has been shown to have little pre¬
age moment of 65 N m in both straight legs and
dictive value for future low-back trouble.94-97 The
FIGURE 5-19 The “cat stretch" is performed by slowly cycling through full spine flexion (A) to full
extension (B). Spine mobility is emphasized rather than “pressing” at the end range of motion. This exercise
provides motion for the spine with very low loading of the intervertebral joints.
most successful programs appear to emphasize through full flexion and extension in a slow,
trunk stabilization through exercise with a neutral smooth motion (Fig. 5-19). Hip and knee flexibil¬
spine98 but emphasize mobility at the hips and ity may be achieved with the following maneu¬
knees. Bridger and colleagues99 demonstrate ad¬ vers, emphasizing a neutral spine throughout: hip
vantages for sitting and standing, whereas McGill mobility—standing hip extension, standing hip
and Norman87 outline advantages for lifting. flexion; hip mobility, strength, and endurance—
For these reasons, specific torso flexibility exer¬ slow lunges (Figs. 5-20 and 5-21).
cises should be limited to unloaded flexion and
extension for those concerned with safety but per¬
haps not for those interested in specific athletic ISSUES OF STRENGTH AND ENDURANCE
performance (of course, spine flexibility may be
of greater desirability in athletes who have never The link between lower muscle strength and en¬
suffered back injury). The spine may be cycled durance performance in patients with previous
FIGURE 5-20 Hip mobility is enhanced with standing flexion and extension positions. A, The correct
neutral spine. B, An incorrect, flexed spine.
FIGURE 5-21 Hip mobility, strength,

and endurance are challenged with slow
lunges. A, The torso remains upright
throughout the lunge effort. B, Emphasis
is placed on a neutral spine during hip
exercises to focus the stretch over the
hip and knee joints. C, The incorrect,
flexed spine.
back injuries is well documented. In fact, few Aerobic Exercise

longitudinal studies have linked reduced strength
and endurance with the risk of a subsequent first¬ The mounting evidence supporting the role of
time low-back injury. The few studies available aerobic exercise both in reducing the incidence of
suggest that endurance has a much greater prophy¬ low-back injury101 and in treating low-back injury
lactic value than strength.100 Furthermore, it would patients102 is compelling. Recent investigation into
appear that emphasis placed on endurance should loads sustained by the low-back tissues during
precede specific strengthening exercise in a gradu¬ walking54 confirm very low levels of supporting
ated progressive exercise program (i.e., longer- passive tissue load coupled with mild, but pro¬
duration, lower-effort exercises).
longed, activation of the supporting musculature.
Epidemiologic evidence also sheds light on the were characterized by larger psoas activation than
effects of aerobic exercise. A large study103 exam¬ straight-leg sit-ups, through to the highest psoas
ined age-related changes to the lumbar spines of activity observed during leg raises and hand-on-
elderly people as a function of lifelong activity knee flexor isometric exertions. Note that the
level; those who were runners had no differences “press-heels” sit-up, which has been hypothesized
in spine changes measured from MRI images, to activate hamstrings and neurally inhibit psoas,
whereas weight lifters and soccer players were actually increased psoas activation. Normalized
characterized with more disk degeneration and EMG data in Table 5-9 are provided for compara¬
bulges. tive purposes. (Some athletes intentionally wish to
train psoas and will find these data informative;
The Abdominals (Anterior and Lateral) and low-back injury patients must be more selective.)
Quadratus Lumborum One exercise not often performed but appearing
to have merit is the horizontal side support be¬
There is no single abdominal exercise that chal¬ cause it challenges the lateral obliques without
lenges all of the abdominal musculature— high lumbar compressive loading.89 In addition,
requiring the prescription of more than a single this exercise produces high activation levels in
exercise. Calibrated intramuscular and surface the quadratus lumborum, which appears to be a
EMG evidence35-92 suggests that the various types significant stabilizer of the spine40 (as previously
of curl-ups challenge mainly the rectus abdominis noted). Graded activity in the rectus abdominis
because psoas and abdominal wall (internal and and each of the components of the abdominal wall
external oblique, transverse abdominis) activity is change with each of these exercises, demonstra¬
low. Sit-ups (both straight-leg and bent-knee) are ting that there is no single best task for the collec¬
characterized by higher psoas activation and tive abdominal muscles. Clearly, curl-ups excel
higher low-back compression, whereas leg raises at activating the rectus abdominis but produce
cause even higher activation and also spine com¬ relatively lower oblique activity. Several other
pression (Table 5-9). clinically relevant findings from these two data
Several relevant observations were made re¬ sets include notions that psoas activation is domi¬
garding abdominal exercises in our investigations. nated by hip flexion demands and that psoas activ¬
The challenge to psoas is lowest during curl-ups, ity is not consistent with either lumbar sagittal
followed by higher levels during the horizontal moment or spine compression demands. We ques¬
isometric side support, whereas bent-knee sit-ups tion the often-cited notion that psoas is a lumbar
TABLE 5-9 Subject Averages of Electromyogram Activation Normalized to 100% of Maximum Voluntary
Contraction*
Abdominal Tasks
Quadratus
Lumborum Psoas I , Psoas 2 , EOi lOi TAi RA_RF
Straight-leg sit-ups 15 (12) 24 (7) 44 (9) 15(15) II (9) 48 (18) 16 (10) 4(3)
Bent-knee sit-ups 12(7) 17(10) 28 (7) 43 (12) 16(14) 10(7) 55 (16) 14(7) 6(9)
Press-heel sit-ups 28 (23) 34(18) 51 (14) 22 (14) 20 (13) 51 (20) 15(12) 4(3)
Bent-knee curl-up II (6) 7(8) 10 (14) 19 (14) 14 (10) 12(9) 62 (22) 8(12) 6(10)
Bent-knee leg raise 12(6) 24(15) 25 (8) 22 (7) 8(9) 7(6) 32 (20) 8(5) 6(8)
Straight leg raise 9(2) 35 (20) 33 (8) 26 (9) 9(8) 6(4) 37 (24) 23 (12) 7(11)
Isometric hand-to-knee
16 (16) 16(8) 68 (14) 30 (28) 28 (19) 69 (18) 8(7) 6(4)
LH-RK
56 (28) 58 (16) 53 (12) 48 (23) 44 (18) 74 (25) 42 (29) 5 (4)
RH-LK
Cross curl-up
6(4) 5(3) 4(4) 23 (20) 24 (14) 20 (1 1) 57 (22) 10 (19) 5(8)
RS—across
6(4) 5 (3) 5(5) 24(17) 21 (16) 15(13) 58 (24) 12 (24) 5(8)
LS—across
Isometric side support 54 (28) 21 (17) 12(8) 43 (13) 36 (29) 39 (24) 22 (13) II (II) 24(15)
26 (18) 13(5) 44 (16) 42 (24) 44 (33) 41 (20) 9(7) 29 (17)
Dynamic side support
4(1) 24 (19) 12(5) 29 (12) 10 (14) 9(9) 29 (10) 10(7) 3(4)
Push-up from feet
Push-up from knees 14(11) 10(7) 19 (10) 7(9) 8(8) 19(11) 5(3) 3 (4)
*Mean and standard deviation (in parentheses) are given. Note psoas channels, external oblique (EO), internal oblique (IO), and transverse
abdominals (TA) are intramuscular electrodes (i), whereas rectus abdominis (RA), rectus femoris (RF), and erector spinae (ES) are surface
electrodes.
LH-RK, left hand-right knee; RH-LK, right hand-left knee; RS, right shoulder; LS, left shoulder.
—CURE 5-22 A, In the curl-up, the head and shoulders are raised off the ground with the hands under
the lumbar region to help stabilize the pelvis and support the neutral spine. B, A variation is to bend only
one leg while the other, straight leg assists in pelvic stabilization and preservation of a neutral lumbar curve.
spine stabilizer; quadratus lumborum activity is dominal exercises are shown—the curl-up with
consistent with lumbar sagittal moment and com¬ the hands on the low back to stabilize the pelvis
pression demands, suggesting a larger role in sta¬ and assist in preservation of a neutral lordosis
bilization; and psoas activation is relatively high (lumbar curvature), and the horizontal isometric
(greater than 25% MVC) during push-ups, sug¬ side support (again with the spine in a neutral
gesting cautious concern for the low-back injury posture) using either the knees or the feet for
patient.
support (Figs. 5-22 and 5-23).
A wise choice for abdominal exercises in the
early stages of training or rehabilitation would
The Back Extensors
consist of several variations of curl-ups for the
rectus abdominis, and isometric, horizontal side We have been searching for methods to activate
support (with the body supported by the knees the extensors with minimal spine loading,88 given
and upper body supported by one elbow on the that most traditional extensor exercises are charac¬
floor) exercises to challenge the abdominal wall terized by high spine loads, which result from
in a way that imposes minimal compressive pen¬ externally applied compressive and shear forces
alty on the spine. The level of challenge with the (from either free weights or resistance machines).
isometric, horizontal side support exercises can be It appears that the single leg extension hold, while
increased by supporting the body with the feet
on the hands and knees (Fig. 5-24), minimizes
rather than the knees. Specific recommended ab¬ external loads on the spine but produces spine
Figure 5-23 The horizontal isometric side support.

Supporting the lower body with the knees on the floor
reduces the demand further for those who are more
concerned with safety. Supporting the body with the feet
increases the muscle challenge but also the spine load.
Progression of challenge is indicated, with the lowest in A
and highest in C.
FIGURE 5-24 A, Single-leg extension holds, while on the hands and knees, produce mild extensor activity
and lower spine compression (<2500 N). B, Raising the contralateral arm increases extensor muscle activity
but also spine compression to levels higher than 3000 N. (From Callaghan J, Gunning J, McGill SM:
Relationship between lumbar spine load and muscle activity during extensor exercises. Phys Ther 78[I ]:8— 18,
1998.)
extensor moment (and small isometric twisting interspinous ligament (noted earlier as an injury
moments), activating the extensors (one side of the mechanism).
lumbar extensors about 18% MVC). Activation is
sufficiently high on one side of the extensors to Should Abdominal Belts Be Worn?
facilitate training, but the total spine load is re¬
duced because the contralateral extensors are pro¬ The average patient must be confused when ob¬
ducing lower forces (lumbar compression is less serving both elite lifting athletes and back injury
than 2500 N). Switching legs trains both sides of patients wearing abdominal belts. Several years
the extensors. ago, I conducted a review of the effects of belt
In total, seven tasks were analyzed to facilitate wearing104 and summarized the following: those
comparison of various extensor tasks.88 Simultane¬ who have never had a previous back injury appear
ous leg extension with contralateral arm raise in¬ to have no additional protective benefit from wear¬
creases the unilateral extensor muscle challenge ing a belt; it would appear that those who have
(about 27% MVC in one side of lumbar extensors had an injury while wearing a belt risk a more
and 45% MVC in the other side of thoracic exten¬ severe injury; belts appear to give people the per¬
sors) but also increases lumbar compression to ception they can lift more and may in fact enable
well more than 3000 N (see Fig. 5-24). The often- them to lift more; and belts appear to increase
performed exercise of lying prone on the floor and intraabdominal pressure and blood pressure; belts
raising the upper body and legs off the floor is appear to change the lifting styles of some people
contraindicated for anyone at risk of low-back to either decrease the loads on the spine or in¬
injury or reinjury (Fig. 5-25). In this task, the crease the loads on the spine. In summary, given
lumbar region pays a high compression penalty to the assets and liabilities of belt wearing, they are
a hyperextended spine (about 4000 N or higher), not recommended for routine exercise participa¬
which transfers load to the facets and crushes the tion.
Notes for Exercise Prescription
The exercise professional must design exercise

programs to meet a wide variety of objectives.
The following is a list of general caveats to assist
in achieving the best prescription (this list was
adapted from my chapter in the American College
of Sports Medicine publication90):
FIGURE 5-25 This often-prescribed extensor exercise is
1. Although there is a common belief among
contraindicated for most people because of the extended
some experts that exercise sessions should
posture (and facet and anulus loading) and high spine
be performed at least three times per week,
compression load of 4000 to 6000 N. (From Callaghan J,
Gunning J, McGill SM: Relationship between lumbar spine
it appears that low-back exercises have the
load and muscle activity during extensor exercises. Phys Ther most beneficial effect when performed
78[ I ]:8— 18, 1998.) daily.105
2. The “no pain, no gain” axiom does not Acknowledgments: I wish to acknowledge the contribu¬
apply when exercising the low back, tions of several colleagues to the collection of works
particularly when applied to weight reported here: Daniel Juker, MD; Craig Axler, MSc;
training, and scientific and clinical wisdom Jacek Cholewicki, PhD; Michael Sharratt, PhD; John
would suggest the opposite is true. Seguin, MD; Vaughan Kippers, PhD; and in particular,
Robert Norman, PhD. Financial support from the Natu¬
3. Although specific low-back exercises have
ral Science and Engineering Research Council, Canada
been rationalized in this chapter, general
has made this series of work possible.
exercise programs that also combine
cardiovascular components (such as
walking) have been shown to be more References
effective in both rehabilitation and injury
1. Roberts S, Menage J, Urban JPG: Biochemical
prevention.102 The exercises shown here
and structural properties of the cartilage end-
constitute only a component of the total
plate and its relationship to the intervertebral
program.
disc. Spine 14:166, 1989.
4. Diurnal variation in the fluid level of the 2. Gallois J, Japoit T: Architecture interieure des
intervertebral disks (disks are more vertebres du point de vue statique et
hydrated early in the morning after rising physiologique. Rev Chir (Paris) 63:688, 1925.
from bed) changes the stresses on the disk 3. Nachemson AL: Lumbar interdiscal pressure.
throughout the day. It would be unwise to Acta Orthop Scand Suppl 43, 1960.
perform full-range spine motion while 4. Nachemson A: The load on lumbar discs in
under load, shortly after rising from bed.17 different positions of the body. Clin Rel Res
45:107, 1966.
5. Low-back exercises performed for
5. Brinckmann P, Biggemann M, Hilweg D:
maintenance of health need not emphasize
Prediction of the compressive strength of human
strength. High-load, low-repetition tasks,
lumbar vertebrae. Clin Biomech 4(suppl 2),
rather than more repetitions of less 1989.
demanding exercises, will assist in the 6. Yingling VR, McGill SM: Mechanical properties
enhancement of endurance and strength. and injuries resulting from anterior and posterior
There is no doubt that back injury can shear loading of the spine at different load rates.
occur during seemingly low-level demands In Proceedings of the American Society for
(such as picking up a pencil) and that the Biomechanics, Atlanta, October 17-19, 1996,
risk of injury from motor control error can pp 267-268.
occur. Although it appears that the chance 7. Farfan HF: Mechanical disorders of the low
back. Philadelphia, Lea & Febiger, 1973.
of motor control errors, resulting in
8. Roaf R: A study of the mechanics of spinal
inappropriate muscle forces, increases with
injuries. J Bone Joint Surg 42B:810, 1960.
fatigue, there is also evidence documenting 9. Fyhrie DP, Schaffler MB: Failure mechanisms in
the changes in passive tissue loading with human vertebral cancellous bone. Bone
fatiguing lifting.106 Given that endurance 15(1): 105-109, 1994.
has more protective value than strength,100 10. McGill SM: Low back exercises: Prescription for
strength gains should not be the healthy back and when recovering from
overemphasized at the expense of injury. In American College of Sports Medicine:
endurance. Resource Manual for Guidelines for Exercise
6. There is no such thing as an ideal set of Testing and Prescription, 3rd ed. Baltimore,
Williams & Wilkins, 1998.
exercises for all patients. An individual’s
11. Bogduk N, Engel R: The menisci of the lumbar
training objectives must be identified (e.g.,
zygapophyseal joints: A review of their anatomy
rehabilitation, injury risk reduction,
and clinical significance. Spine 9:454, 1984.
optimization of general health and fitness, 12. Bedzinski R: Application of speckle photography
or maximization of athletic performance) methods to the investigations of deformation of
and the most appropriate exercises chosen. the vertebral arch. In Little EG (ed):
Although science cannot evaluate the Experimental Mechanics. New York, Elsevier
optimal exercises for each situation, the 1992.
combination of science and clinical 13. Dickey JP, Pierrynowski MR, Bednar DA:
experience must be used to enhance low- Deformation of vertebrae in vivo: Implications
back health. for facet joint loads and spinous process pin
instrumentation for measuring sequential spinal
7. Patience and perseverance are important.
kinematics. Presented at the Canadian
Increased function and reduction of pain
Orthopaedic Research Society, Quebec City Mav
may not occur for 3 months.107 25, 1996.
14. Hardcastle P, Annear P, Foster D: Spinal scans of supine postures for use in standing
abnormalities in young fast bowlers. J Bone Joint postures. J Biomech 29:643-646, 1996.
Surg 74B(3):421, 1992. 32. Nitz AJ, Peck D: Comparison of muscle spindle
15. Cripton P, Berlemen U, Visarino H, et al: concentrations in large and small human epaxial
Response of the lumbar spine due to shear muscles acting in parallel combinations. Am
loading. In Injury Prevention Through Surg 52:273-277, 1986.
Biomechanics. Detroit, Wayne State University, 33. McGill SM, Norman RW: Effects of an
May 4-5, 1995, p 111. anatomically detailed erector spinae model on
16. Markolf KL, Morris JM: The structural L4/L5 disc compression and shear. J Biomech
components of the intervertebral disc. J Bone 20(6):591, 1987.
Joint Surg 56A(4):675, 1974. 34. Sirca A, Kostevc V: The fibre type composition
17. Adams MA, Dolan P: Recent advances in lumbar of thoracic and lumbar paravertebral muscles in
spinal mechanics and their clinical significance. man. J Anat 141:131, 1985.
Clin Biomech 10(1):3, 1995. 35. Juker D, McGill SM, Kropf P, Steffen T:
18. Adams MA, Hutton WC: Prolapsed intervertebral Quantitative intramuscular myoelectric activity of
disc: A hyperflexion injury. Spine 7:184, 1982. lumbar portions of psoas and the abdominal wall
19. Adams MA, Hutton WC: Gradual disc prolapse. during a wide variety of tasks. Med Sci Sports
Spine 10:524, 1985. Exerc 30(2):301-310, 1998.
20. Adams P, Muir H: Qualitative changes with age 36. Belanger M: Personal communication. University
of proteoglycans of human lumbar discs. Ann of Quebec at Montreal, 1996.
Rheum Dis 35:289, 1976. 37. McGill SM: A revised anatomical model of the
21. King AI: Injury to the thoraco-lumbar spine and abdominal musculature for torso flexion efforts. J
pelvis. In Nahum AM, Melvin JW (eds): Biomech 29(7):973, 1996.
Accidental Injury, Biomechanics and 38. McGill SM: Electromyographic activity of the
Presentation. New York, Springer-Verlag, 1993. abdominal and low back musculature during the
generation of isometric and dynamic axial trunk
22. Gordon SJ, Yang KH, Mayer PJ, et al:
torque: Implications for lumbar mechanics. J
Mechanism of disc rupture: A preliminary report.
Orthop Res 9:91, 1991.
Spine 16:450, 1991.
39. McGill SM: A myoelectrically based dynamic 3-
23. Videman T, Nurminen M, Troup JDG: Lumbar
D model to predict loads on lumbar spine tissues
spinal pathology in cadaveric material in relation
during lateral bending. J Biomech 25(4):395,
to history of back pain, occupation and physical
1992.
loading. Spine 15(8):728, 1990.
40. McGill SM, Juker D, Kropf P: Quantitative
24. Wilder DG, Pope MH, Frymoyer JW: The
intramuscular myoelectric activity of quadratus
biomechanics of lumbar disc herniation and the
lumborum during a wide variety of tasks. Clin
effect of overload and instability. J Spinal Disord
Biomech 11(3): 170, 1996.
1(1): 16, 1988. 41. Heylings DJ: Supraspinous and interspinous
25. Goel VK, Monroe BT, Gilbertson LG, ligaments of the human lumbar spine. J Anat
Brinckmann P: Interlaminar shear stresses and
125:127, 1978.
laminae—separation in a disc: Finite element 42. Bogduk N, Twomey LT: Clinical anatomy of the
analysis of the L3-L4 motion segment subjected lumbar spine. New York, Churchill Livingstone,
to axial compressive loads. Spine 20(6):689, 1987.
1995. 43. Sharma M, Langrama NA, Rodriguez J: Role of
26. Bogduk N: A reappraisal of the anatomy of the ligaments and facets in lumbar spine stability.
human lumbar erector spinae. J Anat 131 (3):525, Spine 20(8):887, 1995.
1980. 44. Noyes FR, De Lucas JL, Torvik PJ:
27. Macintosh JE, Valencia F, Bogduk N, Munro Biomechanics of ligament failure: An analysis of
RR: The morphology of the human lumbar strain-rate sensitivity and mechanisms of failure
multifidus. Clin Biomech 1:196, 1986. in primates. J Bone Joint Surg 56A:236, 1994.
28. Macintosh JE, Bogduk N: The morphology of 45. Rissanen PM: The surgical anatomy and
the lumbar erector spinae. Spine 12(7):658, 1987. pathology of the supraspinous and interspinous
29. McGill SM, Patt N, Norman RW: Measurement ligaments of the lumbar spine with special
of the trunk musculature of active males using reference to ligament ruptures. Acta Orthop
CT scan radiography: Duplications for force and Scand Suppl 46, 1960.
moment generating capacity about the L4/L5 46. Carr D, Gilbertson L, Frymoyer J, et al: Lumbar
joint. J Biomech 21(4):329, 1988. paraspinal compartment syndrome: A case report
30. McGill SM, Santaguida L, Stevens J: with physiologic and anatomic studies. Spine
Measurement of the trunk musculature from T6 10:816, 1985.
to L5 using MRI scans of 15 young males 47. Styf J: Pressure in the erector spinae muscle
corrected for muscle fibre orientation. Clin during exercise. Spine 12:675, 1987.
Biomech 8:171, 1993. 48. Gracovetsky S, Farfan HF, Lamy C: Mechanism
31. McGill SM, Juker D, Axler CT: Correcting trunk of the lumbar spine. Spine 6(1):249, 1981.
muscle geometry obtained from MRI and CT 49. Ashton-Miller JA, Schultz AB: Biomechanics of
the human spine and trunk. In Parrdolf KB (ed): 66. Macintosh JE, Bogduk N, Gracovetsky S: The
Exercise and Sport Science Reviews, vol 16. biomechanics of the thoracolumbar fascia. Clin
American College of Sports Medicine Series. Biomech 2:78, 1987.
New York, MacMillan, 1988. 67. McGill SM, Norman RW: The potential of
50. Spencer DL, Miller JAA, Schultz AB: The lumbodorsal fascia forces to generate back
effects of chemonucleolysis on the mechanical extension moments during squat lifts. J Biomed
properties of the canine lumbar disc. Spine Eng 10:312, 1988.
10:555, 1985. 68. Bogduk N, Macintosh JE: The applied anatomy
51. Kirkaldy-Willis WH, Burton CV: Managing low of the thoracolumbar fascia. Spine 9:164, 1984.
back pain, 3rd ed. New York, Churchill 69. Hukins DWL, Aspden RM, Hickey DS:
Livingstone, 1992.
Thoracolumbar fascia can increase the efficiency
52. Cholewicki J, McGill SM: Mechanical stability of the erector spinae muscles. Clin Biomech
of the in vivo lumbar spine: Implications for 5(1):30, 1990.
injury and chronic low back pain. Clin Biomech 70. Carter DR, Hayes WC: The comprehensive
11(1): 1, 1996.
behavior of bone as a two-phase porous
53. Potvin J, Norman RW, McGill S: Reduction in
structure. J Bone Joint Surg 58A:954, 1977.
anterior shear forces on the L4/L5 disc by the
71. Adams MA, Hutton WC, Stott JRR: The
lumbar musculature. Clin Biomech 6:88, 1991.
resistance to flexion of the lumbar intervertebral
54. Callaghan JP, Patla A, McGill SM: 3D Analysis
joint. Spine 5:245, 1980.
of Spine Loading During Gait. Atlanta, American
72. McGill SM, Brown S: Creep response of the
Society for Biomechanics, October 17-19, 1996.
lumbar spine to prolonged lumber flexion. Clin
55. Troup JDG, Leskinen TPJ, Stalhammear HR,
Biomech 7:43, 1992.
Kuorinka IA: A comparison of intra-abdominal
73. Dontigny RL: Function and pathomechanics of
pressure increases, hip torque, and lumbar
the sacroiliac joint. Phys Ther 65:35-44, 1985.
vertebral compression in different lifting
74. Resnick D, Niwayama G, Georgen TG:
techniques. Hum Factors 25(5):517, 1983.
Degenerative disease of the sacroiliac joint.
56. Thomson KD: On the bending moment capability
Invest Radiol 19:608, 1975.
of the pressurized abdominal cavity during
75. McGill SM: A biomechanical perspective of
human lifting activity. Ergonomics 31(5):817,
1988. sacro-iliac pain. Clin Biomech 2:145-151.
76. McGill SM, Norman RW: Partitioning of the L4/
57. Beam JG: The significance of the activity of the
L5 dynamic moment into disc, ligamentous and
abdominal muscles in weight lifting. Acta Anat
45:83, 1961. muscular components during lifting. Spine
11 (7):666, 1986.
58. Grew ND: Intraabdominal pressure response to
loads applied to the torso in normal subjects. 77. McGill SM, Kippers V: Transfer of loads
Spine 5(2): 149, 1980. between lumbar tissues during the flexion
59. Ekholm J, Arborelius UP, Nemeth G: The load relaxation phenomenon. Spine 19(19)-2190
1994.
on the lumbosacral joint and trunk muscle
activity during lifting. Ergonomics 25(2): 145 78. Reilly T, Tynell A, Troup JDG: Circadian
1982. variation in human stature. Chronobiol Int 1121
60. Krag MH, Byrne KB, Gilbertson LG, Haugh LD: 1984.
Failure of intraabdominal pressurization to 79. Adams MA, Dolan P, Hutton WC: Diurnal
reduce erector spinae loads during lifting tasks. variations in the stresses on the lumbar spine
In Proceedings of the North American Congress Spine 12(2): 130, 1987.
on Biomechanics, Montreal, August 25-27, 1986 80. McKenzie RA: Prophylaxis in recurrent low back
p 87. pain. N Z Med J 89:22, 1979.
61. Nachemson AL, Morris JM: In vivo 81. Krag MH, Seroussi RE, Wilder DG, Pope MH:
measurements of intradiscal pressure. J Bone Internal displacement distribution from in vitro
Joint Surg 46A:1077, 1964. loading of human thoracic and lumbar spinal
62. Nachemson A, Andersson GBJ, Schultz AB: motion segments: Experimental results and
Valsalva manoeuvre biomechanics: Effects on theoretical predictions. Spine 12(10): 1001, 1987.
lumbar trunk loads of elevated intra-abdominal 82. McGill SM, Brown S: Creep response of the
pressure. Spine 11(5):476, 1986. lumbar spine to prolonged lumber flexion. Clin
63. McGill SM, Norman RW: Reassessment of the Biomech 7:43, 1992.
role of intraabdominal pressure in spinal 83. Davis PR: The causation of hemiae by weight¬
compression. Ergonomics 30(11):1565, 1987. lifting. Lancet 2:155, 1959.
64. Lucas D, Bresler B: Stability of the ligamentous 84. Crisco JJ, Panjabi MM: Postural biomechanical
spine. Tech. Report No. 40, Biomechanics stability and gross muscular architecture in the
Laboratory, University of California, San spine. In Winters J, Woo S (eds): Multiple
Francisco, 1961.
Muscle Systems. New York, Springer-Verlag
65. Tesh KM, Dunn J, Evans JH: The abdominal 1990, p 438.
muscles and vertebral stability. Spine 12(5) 501 85. Cholewicki J, McGill SM: Lumbar posterior
1987.
ligament involvement during extremely heavy
lifts estimated from fluoroscopic measurements. J with body position and hip muscle length. Int J
Biomech 25(1): 17, 1992. Ind Ergonom 9:235, 1992.
86. Garg A, Herrin G: Stoop or squat: A 100. Luoto S, Heliovaara M, Hurri H, Alaranta M:
biomechanical and metabolic evaluation. Am Inst Static back endurance and the risk of low back
Ind Eng Trans 11:293, 1979. pain. Clin Biomech 10(6):323, 1995.
87. McGill SM, Norman RW: Low back 101. Cady LD, Bischoff DP, O’Connell ER, et al:
biomechanics in industry: The prevention of Strength and fitness and subsequent back injuries
injury. In Grabiner MD (ed): Current Issues of in firefighters. J Occup Med 21 (4):269, 1979.
Biomechanics. Champaign, IL, Human Kinetics, 102. Nutter P: Aerobic exercise in the treatment and
1992. prevention of low back pain. State Art Rev
88. Callaghan J, Gunning J, McGill SM: Occup Med 3:137, 1988.
Relationship between lumbar spine load and 103. Videman T, Sama S, Crites-Battie M, et al: The
muscle activity during extensor exercises. Phys long term effects of physical loading and
Ther 78(1):8-18, 1998. exercise lifestyles on back-related symptoms,
89. Axler CT, McGill SM: Low back loads over a disability, and spinal pathology among men.
variety of abdominal exercises: Searching for the Spine 20(b):669, 1995.
safest abdominal challenge. Med Sci Sports 104. McGill SM: Abdominal belts in industry: A
Exerc 29(6):804-811, 1997. position paper on their assets, liabilities and use.
90. McGill SM: Low back exercises: Prescription for Am Ind Hyg Assoc J 54(12):752, 1993.
105. Mayer TG, Gatchel RJ, Kishino N, et al:
the healthy back and when recovering from
Objective assessment of spine function following
injury. In American College of Sports Medicine:
industrial injury: A prospective study with
Resource Manual for Guidelines for Exercise
comparison group and one-year follow up. Spine
Testing and Prescription, 3rd ed. Baltimore,
10(6):482, 1985.
106. Potvin JR, Norman RW: Can fatigue compromise
91. Santaguida L, McGill SM: The psoas major
lifting safety? In Proceedings of the Second
muscle: A three-dimensional mechanical
North American Congress on Biomechanics,
modelling study with respect to the spine based
August 24-28, 1992, p 153.
on MRI measurement. J Biomech 28(3):339,
107. Manniche C, Hesselsoe G, Bentzen L, et al:
1995. Clinical trial of intensive muscle training for
92. McGill SM: The mechanics of torso flexion: chronic low back pain. Lancet Dec. 24/31:1473,
Situps and standing dynamic flexion manoeuvres.
1988.
Clin Biomech 10(4): 184, 1995. 108. White AA, Panjabi MM: Clinical Biomechanics
93. Malmivaara A, Hakkinen U, Aro T, et al: The of the Spine. Philadelphia, JB Lippincott, 1978.
treatment of acute low back pain: Bed rest, 109. Pearcy MJ, Portek J, Shepherd J: Three
exercises, or ordinary activity? N Engl J Med dimensional x-ray analysis of normal
332:351, 1995. measurement in the lumbar spine. Spine 9:294,
94. Nachemson A: Newest knowledge of low back 1984.
pain: A critical look. Clin Orthop 279:8, 1992. 110. Pearcy MJ, Tibrewal SB: Axial rotation and
95. Biering-Sorensen F: Physical measurements as lateral bending in the normal lumbar spine
risk indicators for low back trouble over a one measured by three-dimensional radiography.
year period. Spine 9:106, 1984. Spine 9:582, 1984.
96. Burton AK, Tillotson KM, Troup IDG: Variation 111. Shultz AB, Warwick DN, Berkson MH,
in lumbar sagittal mobility with low back Nachemson AL: Mechanical properties of human
trouble. Spine 14(6):584, 1989. lumbar spine motion segments. I. Response in
97. Battie MC, Bigos SJ, Fischer LD, et al: The role flexion, extension, lateral bending and torsion. J
of spinal flexibility in back pain complaints Biomech Eng 101:46, 1979.
within industry: A prospective study. Spine 112. Berkson MH, Nachemson AL, Shultz AB:
15(8):768, 1990. Mechanical properties of human lumbar spine
98. Saal JA, Saal JS: Nonoperative treatment of motion segments. II. Responses in compression
herniated lumbar intervertebral disc with and shear: Influence of gross morphology. J
radiculopathy: An outcome study. Spine Biomech Eng 101:53, 1979.
14(4):41, 1989. 113. McGlashen KM, Miller JAA, Shultz AB,
99. Bridger RS, Orkin D, Henneberg M: A Anderson GBJ: Load displacement behaviour of
quantitative investigation of lumbar and pelvic the human lumbosacral joint. J Orthop Res
postures in standing and sitting: Interrelationships 5:488, 1987.
'■
j|
Biomechanics of the
Shoulder Complex
Elsie Culham and Judith Laprade
The shoulder complex consists of the humerus, the clavicular surface is concave anteroposteriorly
clavicle, and scapula and the joints that link them and convex cephalocaudally.1,2 The superior por¬
into a functional articulated unit. Although gleno¬ tion of the medial end of the clavicle extends
humeral stability is provided by the capsular liga¬ above the manubrium and forms the border of the
mentous structures and the dynamic action of the sternal notch (Fig. 6-1).
rotator cuff musculature surrounding the shoulder, An intraarticular disk runs downward from the
the mobility is governed largely by the structure nonarticular superomedial aspect of the clavicle to
of the shallow glenoid, which affords a large de¬ the first costal cartilage near its junction with the
gree of motion. Ultimately, the three bones and sternum. The circumference of the disk is attached
four joints of the shoulder complex work interde- to the capsule, completely dividing the joint into
pendently to position the hand in space and allow two separate cavities.2 The inferior portion of the
intricate gross and skilled functions to be per¬ disk is thinner than the superior aspect, and its
formed. The four functional articulations that con¬ attachment is weaker and easily tom.1'3
tribute to this essential joint motion are the scapu- The sternoclavicular joint is inherently stable
lothoracic joint, the sternoclavicular joint, the and rarely dislocates. However, the relative contri¬
acromioclavicular joint, and the glenohumeral bution of the joint capsule, ligaments, and disk to
joint. The coracoacromial arch and the head of the joint stability has not been extensively studied.
humerus compose a fifth functional articulation, The joint is enclosed by the fibrous capsule, which
which plays an important role in shoulder function is reinforced on all sides by ligaments. The supe¬
and is commonly included in the description of rior capsule is reinforced by the interclavicular
the complex. It is important to understand the ligament, which connects the clavicles, and ante-
structure, mechanics, and function of each compo¬
nent of the complex before the integrated dynamic
function of the shoulder complex can be fully
appreciated.
Sternoclavicular Joint
JOINT STRUCTURE
The sternoclavicular joint is the only articulation

connecting the upper limb and the axial skeleton.
This is a synovial joint between the inferior por¬
tion of the bulbous medial end of the clavicle and
the superolateral aspect of the manubrium and the
cartilage of the first rib. The articular surfaces are
covered with fibrocartilage and are saddle shaped; FlGURE 6-1 Sternoclavicular joint and interposed disk.
141
rior and posterior sternoclavicular ligaments rein¬

force the anterior and posterior capsule. Beam3
determined, based on cadaver experimentation,
that stability of the joint is primarily due to the
strong joint capsule and supporting ligaments. The
posterior superior aspect of the capsule was the
thickest portion and was found to be the primary
mechanism responsible for maintaining the normal
10-degree upward angulation of the clavicle in the
coronal plane. Division of the joint capsule, leav¬
ing the disk intact, resulted in depression of the
lateral end of the clavicle below that of the medial
end.3 The addition of a load, equivalent to the
weight of the arm, to the lateral end of the clavicle
resulted in rupture of the inferior attachment of
the disk.
The costoclavicular ligament runs superiorly
and laterally from the superior surface of the first
rib and costal cartilage to the inferior surface of
the clavicle. It has two components: an anterior
part, which passes upward and laterally, and a
posterior part, which runs upward and medially.
The ligament strongly binds the medial end of
the clavicle to the first rib and prevents medial
displacement of the clavicle over the sternum.4
This ligament also limits elevation and protraction
of the clavicle at the sternoclavicular joint during
arm elevation.
KINEMATICS
Scapulothoracic motion occurs primarily at the

sternoclavicular joint.5 Although the articular sur¬
faces are saddle shaped, the joint functions as a
ball-and-socket joint with 3 degrees of freedom
(Fig. 6-2). Movements are limited by ligaments
and compression of the joint surfaces.1 Elevation
and depression occur primarily between the me¬
dial end of the clavicle and the articular disk. It is
estimated that 45 degrees of elevation and 5 de¬
grees of depression occur,6 the latter movement
being limited by the first rib. In the movements of
protraction and retraction, the disk and clavicle
move as a unit on the sternum. Moseley6 estimated
that 15 degrees of movement occur in each direc¬
Figure 6-2 Kinematics of the sternoclavicular joint. A,
tion, whereas Abbott and Lucas4 stated that 35
Anteroposterior rotation around a vertical axis. B, Elevation
degrees of clavicular protraction occurs at the ster¬ and depression around an anteroposterior axis. C, Rotation
noclavicular joint. The clavicle is also capable of around the long axis of the clavicle.
rotating around its long axis, with dorsal axial
rotation occurring during the final stages of upper
extremity elevation. Abbott and Lucas4 stated that
Acromioclavicular Joint
50 degrees of rotation are possible; however,
Beam3 reported 25 degrees of rotation based on
JOINT STRUCTURE
cadaver experimentation. Dempster1 also reported
25 to 30 degrees of axial rotation of the clavicle A small oval convex articular facet on the lateral
at the sternoclavicular joint.
end of the clavicle articulates with a concave facet
I Chapter 6 BIOMECHANICS OF THE SHOULDER COMPLEX 143
/
on the acromion process of the scapula at the distraction. The acromioclavicular ligament was
acromioclavicular joint.7 The joint line is oblique, found to be primarily responsible for resisting
running inferiorly and medially, and the lateral posterior displacement of the distal clavicle on
end of the clavicle rises above the acromion. Thus, the acromion and posterior axial rotation. This
there is a tendency for the acromion to be driven ligament was also responsible for restraining supe¬
under the lateral end of the clavicle when forces rior displacement of the distal clavicle at small
are applied through the joint. Fibrocartilage covers loads; the conoid ligament became increasingly
the joint surfaces,2 and an intraarticular disk may important as the load increased. Similarly, anterior
partially or completely divide the joint cavity into displacement was resisted by both the acromi¬
two components.6 oclavicular and conoid ligaments. The conoid liga¬
Stability is provided by superior and inferior ment provided the most constraint to anterior
acromioclavicular ligaments, which reinforce the movement of the sternal end of the clavicle on
joint capsule. The coracoclavicular ligament is a the acromion, whereas all three ligaments resisted
primary stabilizing structure of the joint and the posterior movement. The acromioclavicular and
major suspensory ligament of the scapula (Fig. conoid provided resistance to upward rotation of
6-3). This ligament lies medial to the joint and the sternal end of the clavicle, with the conoid
has two distinct portions, often separated by a assuming a greater role as the load increased. The
bursa or fat pad. The trapezoid ligament, the an¬ trapezoid ligament provided significant restraint
terolateral component, runs laterally and slightly only to axial compressive loading.
upward from the superior aspect of the coracoid
process to the trapezoid line on the inferior surface KINEMATICS
of the clavicle.2 The conoid ligament is posterior
and medial to the trapezoid. Its fibers run more The acromioclavicular joint also functions as a
vertically from the superior aspect of the coracoid ball-and-socket joint and has 3 degrees of freedom
process to the conoid tubercle located on the infe¬ (Fig. 6-4). Movements occurring around a vertical
rior surface of the clavicle posteriorly.2 axis through the joint result in rotation of the
Fukuda and associates8 studied the relative con¬ scapula around the lateral end of the clavicle, a
tribution of the ligaments to stability of the acro¬ motion termed scapular winging. If conventional
mioclavicular joint in cadavers using load-displace¬ terminology for motion at the glenohumeral joint
ment tests combined with sequential sectioning were applied, as advocated by Dempster,1 this
of the ligaments. Experiments involved anterior, motion would be equivalent to internal and exter¬
posterior, and superior movement of the sternal nal rotation. Abbott and Lucas4 described 15 de¬
end of the clavicle on a fixed acromion; anterior, grees of motion around a vertical axis. Rotation
posterior, and superior displacement of the distal around a sagittal axis results in upward and down¬
clavicle on the acromion; axial rotation of the ward rotation of the scapula on the distal clavicle,
clavicle around its long axis; and compression and a movement termed abduction and adduction us¬
ing Dempster’s terminology. Movement of the
scapula around a frontal axis through the acromi¬
Coracoclavicular oclavicular joint causes tipping or tilting of the
scapula in the sagittal plane, or flexion and exten¬
Coracoacromial
ligament sion of the scapula. The amount of movement in
this plane is limited by the contact between the
scapula and thorax.
Coracohumeral
ligament Although significant motion has been described
around these axes in the acromioclavicular joint
in cadaver experiments, little motion occurs at this
joint in vivo.5'9 The clinical significance of this is
that the acromioclavicular joint, unlike the sterno¬
clavicular joint, can be fused without loss of upper
extremity function.5
Acromioclavicular joint separations are com¬
mon, accounting for about 12% of all injuries to
the shoulder girdle.10 They occur most often in
FIGURE 6-3 Ligaments arising from the coracoid process:
young athletic males, with the most common
the acromioclavicular, coracohumeral, and coracoclavicular, a
mechanism of injury being a fall in which the
primary stabilizer of the acromioclavicular joint. (From Neer
CS II: Shoulder Reconstruction. Philadelphia, WB Saunders,
adducted shoulder strikes the ground or a firm
1990. With permission from Robert Demarest.) object.11 With this type of fall, the scapula is
driven in an inferior direction relative to the clavi¬

cle.11 Injuries of this type, in which only the
acromioclavicular joint ligaments are either
sprained or disrupted, are classified as type I and
type II injuries, respectively.10 In type III injuries,
both the acromioclavicular and coracoclavicular
ligaments are ruptured, and the acromion moves
inferiorly in relation to the clavicle. Types IV and
V injuries are more severe and result in damage
to the trapezius and deltoid muscles.
Types I and II injuries can generally be treated
conservatively with a reduction in activity and
stabilization with a sling.12- 13 Treatment of type
III injuries is more controversial, but there is some
evidence that outcomes are as good or better with
conservative versus surgical management.12-13
Types IV and V injuries require open reduction
and surgical repair with stabilization of the acro¬
mioclavicular joint with or without repair of the
acromioclavicular and coracoclavicular liga¬
ments.11 Alternatively, the lateral clavicle is fixed
to the coracoid process with screws, wire loops,
or other materials.
In rare instances, a force directed to the distal
clavicle may drive it inferiorly such that it lodges
under the acromion or coracoid process.11 The
acromioclavicular and coracoclavicular ligaments
are ruptured in this type of injury.10 This is classi¬
fied as a type VI dislocation and requires surgical
repair.
SCAPULOTHORACIC COMPLEX
SCAPULAR POSITION
The scapula is a thin, flat triangular bone that lies

on the posterolateral aspect of the thorax over ribs
two to seven.14-15 It has no bony or ligamentous
attachment to the axial skeleton, other than
through the acromioclavicular and sternoclavicular
joints. The axioscapular muscles, including the
trapezius, serratus anterior, and rhomboid major
and minor, and the levator scapulae play a role
in maintaining scapular position.2 The concave
anterior surface of the scapula is separated from
the convex external surface of the thorax by the
subscapularis and serratus anterior muscles, which
glide over one another during movement.
The plane of the scapula is approximately at
right angles to the plane of the glenoid,16 and at
rest it lies obliquely, 30 to 45 degrees anterior to
FIGURE 6-4 Kinematics of the acromioclavicular joint. A, the coronal plane.14-17-18 Objective measures of the
Rotation around an anteroposterior axis (scapular abduction resting position of the scapula, in the plane of
and adduction). B, Rotation around a vertical axis (scapular the scapular spine, have been obtained using an
winging). C, Rotation around a mediolateral axis (scapular electromagnetic system.19 The scapular spine, rep¬
tilt).
resented by a line joining the angle of the aero-
mion and the root, formed a mean angle of 30.2 in standing. The average downward tilt, measured
degrees to the coronal plane in healthy women in 52 healthy male subjects, was 5.29 degrees.22
younger than 40 years of age. The coronal plane Similarly, Poppen and Walker18 reported an aver¬
was defined as a line connecting the roots of the age downward tilt of 4.7 degrees.
right and left scapular spines.19
The scapula is tilted forward slightly in the KINEMATICS
sagittal plane.19,20 The average angle of forward
inclination of the medial border of the scapula The scapula can rotate and move linearly on the
was 9.0 degrees in healthy women younger than thorax from its resting position. Motions generally
40 years of age.19 do not occur independently of one another, and
In the coronal plane, the medial border of the all motion must occur at the sternoclavicular or
scapula is oriented vertically. The mean lateral acromioclavicular joint or both. Generally, three
angle formed between the medial border of the rotatory motions and two translatory motions of
scapula and a horizontal axis was 91.3 degrees in the scapula are described. Upward and downward
women younger than 40 years of age19 (Fig. 6-5). movements of the scapula on the rib cage are
Basmajian and Bazant21 stated that when the scap¬ translatory movements termed elevation and de¬
ula is correctly oriented on the thorax, the glenoid pression. Translation of the scapula toward and
fossa faces anteriorly, laterally, and upward. These away from the vertebral column has been called
authors further suggested that the upward orienta¬ adduction and abduction, respectively, by some
tion of the glenoid fossa provided a “locking authors.16
mechanism,” which prevented downward sublux¬ Rotation of the scapula around a sagittal axis
ation of the humeral head on the glenoid. Subse¬ results in upward or downward tilt of the glenoid
quent radiographic studies, however, have con¬ fossa. This movement has been labeled both ab¬
firmed that in the relaxed standing position, the duction or adduction1 and upward or downward
glenoid fossa has a slight downward inclination in rotation.16 Full elevation of the upper extremity
most normal shoulders.18'22 Glenoid tilt, the angle requires about 60 degrees of scapular rotation in
of a line connecting the superior and inferior gle¬ this plane, which results in progressive upward tilt
noid tubercles relative to the vertical, was meas¬ of the glenoid fossa, thus providing a base on
ured from radiographs taken in the scapular plane which the humeral head moves.23 From 0 to about
80 degrees of arm elevation, scapular rotation is
predominantly a result of elevation of the clavicle
at the sternoclavicular joint. The middle phase of
arm elevation, described by Bagg and Forrest,23
occurs between 80 and 140 degrees of total arm
elevation. This is described as the most stressful
phase of arm elevation and is the phase when
most scapular rotation occurs.20’23 The scapular
rotation is due to continued elevation of the clavi¬
cle at the sternoclavicular joint and limited rota¬
tion of the scapula on the clavicle at the acromi¬
oclavicular joint.23
In the final phase of arm elevation in the scapu¬
lar plane, clavicular elevation becomes increas¬
ingly limited by tension in the costoclavicular
ligament16, 24 as the powerful scapular muscles
continue to rotate the scapula upward. The amount
of movement of the scapula on the clavicle at the
acromioclavicular joint is limited by the coraco-
clavicular ligament.25 The tension generated in this
ligament, as the coracoid process moves away
FIGURE 6-5 Scapular position in the plane of the scapula. from the clavicle, causes a dorsal rotation of the
The mean angle between the medial border of the scapula clavicle around its long axis.4,24,26’27 Because the
and the horizontal (a) was 91.3 degrees in women younger clavicle is crank shaped, this long axis rotation
than the age of 40 years (n = 34). (From Culham E, Peat M.
causes the acromial end of the clavicle to elevate
Functional anatomy of the shoulder compex. J Orthop Sports
further, carrying the scapula with it.24’26’28 Ac¬
Phys Ther 18:342, 1993, with permission of the Orthopaedic
cording to Neer,29 rupture of the coracoclavicular
and Sports Sections of the American Physical Therapy
ligaments in acromioclavicular separation injuries
Association.)
can disrupt this mechanism. If these ligaments are KINETICS OF SCAPULAR MOTION
not intact, the scapula may not rotate sufficiently
during arm elevation, resulting in impingement of Axioscapular Musculature
the greater tuberosity against the acromion proc¬
ess. van der Helm30 postulated that absence of the Trapezius and Serrat us Anterior
conoid ligament would lead to increased torsional Muscles
stress in the acromioclavicular joint and poten¬
tially to joint damage. Upward rotation of the scapula during arm eleva¬
The instant center of rotation of the scapula tion is brought about by the upper and lower
(ICR) varies throughout arm elevation, its position portions of trapezius combined with the serratus
indicating where the motion is primarily oc¬ anterior muscle, particularly the powerful lower
curring. Based on radiographic evidence, Dvir and digitations.23’ 31> 32 These muscles constitute the
Berme24 reported that the ICR was near the root scapular force couple (Fig. 6-6). The upper trape¬
of the scapular spine during the first 90 to 100 zius, forming the superior component of the force
degrees of arm elevation, and at the acromiocla¬ couple, has a supportive as well as a rotatory
vicular joint during the final phase of elevation. function. The lower fibers of trapezius and lower
Similarly, Bagg and Forrest23 found the ICR of digitations of serratus form the inferior compo¬
the scapula to be at or near the root of the scapular nent. The serratus anterior, owing to its large mo¬
spine during the first 80 degrees of arm elevation. ment arm, is capable of producing large moments
Thereafter, the ICR moved along the scapular around the sternoclavicular joint and is considered
spine toward the acromioclavicular joint as the the most important component of the force cou¬
arm was elevated between 80 and 140 degrees, ple.30
and reached the acromioclavicular joint in the final Electromyographic (EMG) activity has been
phases of elevation. Poppen and Walker18 reported monitored in these muscles during elevation in
a somewhat different pattern of movement of the both the coronal and sagittal planes.27-33-34 Both
ICR. In this study, the ICR was found to be trapezius and serratus anterior are active through¬
initially located near the center of the scapula, out arm elevation in both planes, reaching peak
moving superiorly and medially during the first activity at maximal elevation. Upper trapezius has
60 degrees of arm elevation and thereafter moving
superiorly and laterally toward the acromiocla¬
vicular joint.
Rotation of the scapula in other planes has been Upper
less extensively studied. Poppen and Walker18 de¬
scribed rotation of the scapula around a coronal
axis during arm elevation in the scapular plane
that resulted in movement of the superior angle of
the scapula away from the body wall. Forty de¬
grees of motion were measured from radiographs
during arm elevation.18 van der Helm30 also de¬
scribed upward and posterior movement of the
clavicle and ventral movement of the inferior
angle of the scapula during coronal plane abduc¬
tion in a finite element model. This movement is
generally described as scapular tipping or tilting.
Protraction and retraction occur around a verti¬
cal axis. Protraction, defined as forward move¬
ment of the scapula around the thoracic wall,
combines linear translation away from the verte¬
bral column, rotation of the scapula around the
end of the clavicle (winging), and anterior move¬
ment of the lateral end of clavicle.2 The reverse
of these combined motions results in scapular re¬
traction. Abbott and Lucus4 describe 50 degrees
of scapular protraction, 35 degrees due to anterior
movement of the clavicle at the sternoclavicular
joint and 15 degrees due to scapular rotation on
Figure 6-6 Scapular force couple consisting of the upper
the clavicle at the acromioclavicular joint. and lower trapezius and the lower serratus anterior muscles.
been found to be more active in coronal plane

abduction than in sagittal plane elevation35 and
was reported to be particularly active in the first
60 degrees of coronal plane abduction. This com¬
ponent of the force couple has a large moment
arm around a sagittal axis through the sternocla¬
vicular joint during this phase of coronal plane
elevation.30 Lower trapezius was found to be the
more active during abduction than during flexion,
whereas serratus anterior is more active during
sagittal plane flexion.27,30,33 It was proposed that
the decrease in lower trapezius activity during
flexion allowed the scapular protraction needed
for elevation in this plane. Both Inman and col¬
leagues27 and Weidenbauer and Mortenson33 re¬
ported that the lower trapezius was most active
during the latter stages of elevation in both planes.
Similarly, van der Helm30 reported minimal activ¬
ity in lower trapezius below 90 degrees of eleva¬
tion in the coronal plane.
Bagg and Forrest31 studied EMG activity of the
scapular rotators during elevation in the scapular
plane. The upper trapezius and lower serratus an¬
terior were active throughout arm elevation, with
activity gradually increasing, particularly in the
upper ranges of elevation. Little activity was ob¬
served in the lower trapezius until after 90 degrees
of elevation, with a sharp rise in activity thereafter.
These authors proposed that migration of the ICR
of the scapula toward the acromioclavicular joint FIGURE 6-8 Military press exercise. (From Townsend H,
during the middle phase of elevation would result Jobe F, Pink M, Perry J: Electromyographic analysis of the
in an increase in the moment arm of lower trape¬ glenohumeral muscles during a baseball rehabilitation
zius, thus improving its mechanical advantage. program. Am J Sports Med 19:264, 1991.)
The lower trapezius is therefore better able to
assist upper trapezius and serratus anterior during

the phase when the most scapular rotation is oc¬
curring.
Activity in these muscles has also been studied
in healthy subjects performing rehabilitation exer¬
cises to determine which exercises most effec¬
tively recruited these muscles.34 Rowing (Fig. 6-7)
and the military press (Fig. 6-8) best recruited the
upper trapezius. Coronal plane abduction, rowing,
horizontal abduction in prone with the humerus in
a neutral or external rotation (Fig. 6-9), and sagit¬
tal plane flexion all resulted in significant activity
in the lower trapezius. All exercises were done
with a light weight in the hand. High levels of
activity occurred in the middle and lower fibers of
serratus anterior during elevation of the arm in the
sagittal, coronal, and scapular planes. The press-
FIGURE 6-7 Rowing exercise. (From Townsend H, Jobe F,
up (Fig. 6-10) and the push-up exercises (Fig.
Pink M, Perry J: Electromyographic analysis of the
6-11) also resulted in high levels of activity in
glenohumeral muscles during a baseball rehabilitation
program. Am J Sports Med 19:264, 1991.) this muscle.
in internal rotation. The same exercise was performed with

the arm in neutral and external rotation. (From Townsend H,
Jobe F, Pink M, Perry J: Electromyographic analysis of the
program. Am J Sports Med 19:264, 1991.)

Middle Trapezius and Rhomboid

Muscles
EMG studies demonstrate that the middle trape¬

zius is active throughout elevation of the upper
extremity in both the coronal27 and scapular
planes.31 Peak activity during coronal plane abduc¬
tion occurred at 90 degrees of elevation, with
activity reaching a plateau or decreasing slightly
beyond that level.27 Conversely, during scapular
plane abduction, low levels of activity were re¬
ported below 90 degrees; the activity level in¬
creased beyond 90 degrees and peaked at maxi¬
mum elevation.31 Much lower levels of middle
trapezius activity were recorded during elevation
of the upper extremity in the sagittal plane.27-30
Inman and colleagues27 suggested that the middle
trapezius functions to stabilize the scapula during
coronal plane abduction. Activity must be reduced
during sagittal plane elevation to allow protraction
of the scapula around the thorax.27
The rhomboids, like the middle trapezius, are
most active during coronal plane abduction, with
lower levels of activity during sagittal plane
flexion.27-30 van der Helm30 stated that during ele¬
Figure 6-10 Press-up exercise. (From Townsend H, Jobe vation in the coronal plane, rhomboid activity acts
F, Pink M, Perry J: Electromyographic analysis of the to counterbalance activity in the upper serratus
anterior, with both muscles acting to stabilize the
root of the scapular spine on the thorax. In both
the coronal and sagittal planes, peak activity oc¬ Glenoid retroversion
curred at the end of range.27

In the study of rehabilitation exercises con¬
ducted by Moseley and coauthors,34 horizontal ab¬
duction with the humerus in neutral rotation (see
Fig. 6-9) resulted in high levels of activity in both
the rhomboid and middle trapezius muscles. High
levels of activity were also found in the middle
trapezius during humeral extension in prone posi¬
tion (Fig. 6-12) and during rowing exercise (see
Fig. 6-7). Rhomboids were recruited during eleva¬
tion in the coronal and scapular planes (light
weight in hand) and during the rowing exercise.
Glenohumeral Joint
ARTICULAR SURFACES
FIGURE 6-13 Humeral and scapular retrotorsion angles.
The glenohumeral joint is a synovial, multiaxial (From Neer CS II: Shoulder Reconstruction. Philadelphia, WB
ball-and-socket joint between the head of the hu¬ Saunders, 1990. With permission from Robert Demarest.)
merus and glenoid cavity of the scapula. The head
of the humerus is much larger than the glenoid
cavity, and thus only a portion of the head articu¬
lates with the fossa in any position of the joint. and the humeral shaft in the frontal plane, varies
The glenohumeral index, calculated by dividing from 130 to 150 degrees.14 In the transverse plane,
the maximal transverse diameter of the glenoid by the humeral head and neck are rotated posteriorly
the maximal transverse diameter of the humeral 25 to 40 degrees with respect to an axis through
head, was found to average 57.5 in 50 normal the humeral epicondyles36,37 (Fig. 6-13). This is
shoulders.36 Lower values of this index have been called the angle of torsion or retrotorsion,37 It has
reported in shoulders with recurrent anterior dislo¬ been suggested that a high retrotorsion angle may
cation; however, the difference was due to erosion contribute to recurrent anterior dislocation.39 Ret¬
of the anterior glenoid margin and labrum re¬ rotorsion angle, however, measured radiographi¬
sulting from recurrent trauma rather than to devel¬ cally and with computed tomography, did not dif¬
opmental dysplasia of the glenoid fossa.36,37 fer in shoulders with anterior instability compared
The humeral head forms a portion of a sphere, with normal shoulders.36,37
with deviations from sphericity of less than 1 % of The glenoid fossa has been described as pear
the radius.38 The head faces medially, superiorly, shaped.40 The superoinferior dimensions of the
and posteriorly with respect to the shaft of the glenoid fossa averaged 39 mm in 140 cadaver
bone.15,16 The angle of inclination, the angle be¬ shoulders.40 The average anteroposterior dimen¬
tween an axis through the humeral neck and head sion of the lower portion of the glenoid was 29
mm.40 The articular cartilage of the glenoid is
thicker peripherally than centrally, whereas the
cartilage on the humeral head is slightly thicker
centrally.38 Soslowsky and colleagues38 reported
that the glenoid fossa with articular cartilage
formed a portion of a sphere with a radius of
curvature within 3 mm of that of the correspond¬
ing humeral head. Thus, although the bone surface
of the glenoid is flatter than that of the correspond¬
ing humeral head, the cartilage surface is not, and
the mating surfaces of the humeral head and gle¬
noid are more congruent than the bony architec¬
ture alone would indicate.38
FIGURE 6-12 Humeral extension in prone lying. (From
Several studies have confirmed that the glenoid
Townsend H, Jobe F, Pink M, Perry J: Electromyographic
analysis of the glenohumeral muscles during a baseball
fossa is retrotilted with respect to the plane of the
rehabilitation program. Am J Sports Med 19:264, 1991.) scapula in most normal shoulders36,37,39 (see Fig.
6-13). Saha reported an average of 7.4 degrees of twice that resisted in an anteroposterior direction.
retrotilt of the fossa in relation to the plane of the The degree of stability was related to the depth of
scapula.39,41 This author39 also suggested that a the concavity and to the degree of compression
ventral tilt or anteversion of the glenoid was asso¬ between the humeral head and the glenoid. Re¬
ciated with anterior instability of the joint. How¬ moval of the labrum resulted in a 20% reduction
ever, no cases of anteversion were found in either in the resistance to translatory forces in both direc¬
stable or unstable shoulders at any level of the tions.48
glenoid in a computed tomography study by Ran- Dislocation of the glenohumeral joint most
delli and Gambrioli.36 Brewer and associates42 commonly occurs in an anterior direction.49 The
found that excessive retroversion was the primary most common position of dislocation is with the
cause associated with nontraumatic posterior in¬ humerus in maximal external rotation with the
stability of the shoulder. arm in an overhead position. Acute dislocation
results in avulsion of the anterior labrum and
GLENOID LABRUM joint capsule from the rim of the glenoid (Bankart
lesion) and, less commonly, in a fracture of the
The glenoid labrum is a rim of fibrous tissue anterior rim of the glenoid.49
attached around the margin of the glenoid fossa.
Its inner surface is covered with synovium; the JOINT CAPSULE
outer surface is continuous with the periosteum of
the scapular neck.28 It is a flexible structure, The fibrous capsule originates from the labrum
allowing adaptation of its shape to accommodate and from the margin of the glenoid fossa beyond
rotation of the humeral head. Cooper and coau¬ the labrum. Laterally, it is attached to the circum¬
thors,43 in an anatomic and histologic study, found ference of the anatomic neck of the humerus.
that the labrum consisted of primarily fibrous tis¬ Inferiorly and medially, the attachment descends
sue, confirming findings of an earlier study by about Vi inch onto the shaft of the humerus.2 The
Moseley and Overgaard.44 Fibrocartilage bridged capsule forms a continuous structure connecting
the region between hyaline articular cartilage and the two bones, except in the area of the biceps
the fibrous labrum.43 Differences were identified groove, where the biceps tendon exits the joint.50
in the superior and inferior aspects of the labrum. If the capsule is intact, little distraction of the
The superior and superoanterior labrum was joint surfaces is possible because of the negative
loosely attached to the glenoid rim by thin elastic intraarticular pressure within the joint.50 Up to 2
connective tissue. The most superior portion in¬ cm of distraction could be obtained when the
serted into the biceps tendon distal to its insertion capsule of cadaver shoulders was opened or
onto the supraglenoid tubercle. The inferior por¬ vented.50 Similarly, puncture of the capsule in ca¬
tion of the labrum consisted of inelastic fibrous daver shoulders was found to result in significant
tissue, which was firmly attached to the glenoid subluxation of the humeral head51 and increased
margin, resembling an extension of the articular anteroposterior translation of the humeral head on
cartilage.43 the glenoid during movement of the arm, empha¬
It is generally accepted that the glenoid labrum sizing the importance of negative intraarticular
deepens the glenoid cavity.7,45,46 This hypothesis pressure to joint stability.52
was supported by a study of 25 cadaver shoulders The anterior, superior, and posterior aspects of
in which the labrum was found to contribute about the capsule are reinforced by the tendons of the
half of the total depth of the fossa.47 The labrum rotator cuff. Fibers of these tendons insert into the
and glenoid margin combined to create a socket capsule, and separation of the tendons from the
that was 9 mm deep on average in a superoinferior capsule is difficult.53 It has been suggested that
direction and 5 mm deep in an anteroposterior this tendinous insertion into capsule helps in re¬
direction.47 Lippitt and Matsen48 reported lower tracting redundant portions of capsule during arm
values of 4.9 mm on average in the superoinferior movements.50 There is no reinforcement by the
plane and 2.2 mm in an anteroposterior direction. musculotendinous cuff in the interval between the
These authors also investigated the ability of the anterior border of supraspinatus tendon and the
glenoid concavity to resist translatory forces ap¬ superior border of the subscapularis tendon.53 This
plied to the humeral head in a cadaver study.48 triangular space, with the coracoid process form¬
Compressive forces were applied to the humeral ing the base, is called the rotator interval.54 The
head, followed by increasing tangential forces, capsule was found to be thickest in this region in
until the humeral head dislocated over the glenoid a histologic study by Clark and associates.50
rim. Average maximal translatory forces that could The inferior capsule is lax when the arm is
be resisted in the superoinferior direction were adducted. Although this portion of the capsule is
not crossed by tendons, Clark and associates50 the biceps tendon passes.58 Others describe inser¬
found that fibers of subscapularis and teres minor tion of the ligament into the capsule in the rotator
muscles inserted into the anterior and posterior interval and into supraspinatus and subscapularis
aspects of the inferior capsule. A muscular portion tendons.50'57 Neer and colleagues57 found that in
of the long head of the triceps was also found to most of the cadaver shoulders studied, the liga¬
arise from the medial side of the inferior capsule.50 ment did not have a distinct insertion into bone.
In mechanical stress tests of glenohumeral joint Cooper and coauthors53 stated that the coracohu¬
capsule, Kaltsas55 found the anteroinferior capsule meral ligament is a fold of anterior superior cap¬
to be the weakest area of capsule because this was sule that is more prominent when the humeral
the first region to rupture. head is translated inferiorly and much less distinct
with superior translation of the humeral head.
LIGAMENTS The smaller superior glenohumeral ligament
originates from the base of the coracoid process
Coracohumeral and Superior Glenohumeral and labrum adjacent to the supraglenoid tubercle
Ligaments and inserts into the humerus just above the lesser
tuberosity. Although the ligament is readily identi¬
The coracohumeral and superior glenohumeral lig¬ fied in most shoulders,58-60 its size is variable.58,60
aments reinforce the joint capsule in the rotator Cadaver studies indicate that the structures in
interval56 (Fig. 6-14). The coracohumeral ligament the rotator interval limit inferior translation of the
has a broad origin from the lateral aspect of the humeral head on the glenoid fossa in the adducted
base of the coracoid process54’57 (see Fig. 6-3). It arm.54,56,58 Bowen and Warren,61 based on selec¬
runs obliquely downward and laterally. Ferrari58 tive cutting and static translation experimentation
stated that the ligament divides into two compo¬ in cadavers, concluded that the superior glenohu¬
nents, which insert into the greater and lesser meral ligament was the primary restraint to infe¬
tuberosities, thus creating a tunnel through which rior translation and that the coracohumeral liga-
Posterior
capsule
Posterior
band
IGHL
FlGURE 6-14 The glenohumeral

ligaments. SGHL, Superior
glenohumeral ligament; MGHL, middle
glenohumeral ligament; IGHL, inferior
glenhumeral ligament. (From Culham
E, Peat M: Functional anatomy of the
shoulder complex. J Orthop Sports
Phys Ther 18:342, 1993, with
permission of the Orthopaedic and
Sports Sections of the American
Physical Therapy Association.)
ment played little or no role. Conversely, Ferrari58 the axillary pouch. More recently, O’Brien and
and Nobuhara and Ikeda56 found that the coraco- associates60 described a thickened posterior band
humeral ligament was a major restraint to inferior in addition to the anterior band and axillary pouch.
translation.56'58 Histologically, the anterior and posterior bands
Insufficiency of the rotator interval structures consisted of organized bundles of collagen fibers
also leads to anteroposterior instability of the gle¬ oriented in the coronal plane. The axillary pouch
nohumeral joint. Cadaver studies have demon¬ was thicker in comparison, but the collagen fibers
strated that the capsule and ligaments in the rotator were less organized.
interval must be sectioned, along with the poste¬ All three components of the inferior glenohu¬
rior capsule, to produce posterior dislocation of meral ligament are important to joint stability.
the adducted arm.62-63 The coracohumeral ligament The anterior or superior band and axillary pouch
also functions to limit external rotation of the tighten with elevation and external rotation, pro¬
adducted humerus57 as well as external rotation viding a broad buttress-like support for the ante¬
between 0 and 60 degrees of arm elevation and rior and inferior aspects of the joint in the middle
therefore is important in providing anterior stabil¬ and upper ranges of abduction.59’ 60 Deficiency in
ity during this phase of arm elevation.58 this part of the ligament may result in anterior
subluxation and dislocation in this part of the
Middle Glenohumeral Ligament range. Similarly, the posterior band and axillary
pouch provide support for the humeral head poste¬
The middle glenohumeral ligament is the most riorly when the abducted arm is internally ro¬
variable of the glenohumeral ligaments (see Fig. tated.60
6-14). O’Brien and associates60 reported that it
was deficient or absent in 4 of the 11 shoulders CORACOACROMIAL ARCH
examined. It arises from the anterosuperior aspect
of the labrum and margin of the glenoid fossa The coracoacromial arch is formed by the acro¬
inferior to the superior glenohumeral ligament. mion process, the coracoid process, and the cora¬
The ligament passes laterally and inserts into the coacromial ligament, the triangular ligament that
anterior aspect of the anatomic neck and lesser runs upward, laterally, and slightly posteriorly
tuberosity of the humerus. It lies under the tendon from the coracoid to the acromion process2,67 (see
of the subscapularis muscle and is intimately Fig. 6-3). Together, these structures form an im¬
attached to it near its insertion into the lesser portant protective arch over the glenohumeral joint
tuberosity.58- 64 The middle and superior glenohu¬ and provide a secondary restraining socket for the
meral ligaments are separated by the subscapu¬ humeral head, preventing upward dislocation. The
laris bursa.58 space between the arch and the proximal humerus
The middle glenohumeral ligament and sub¬ is called the subacromial space or the supraspi-
scapularis tendon contribute to anterior stability of natus outlet (Fig. 6—15). This space is occupied
the glenohumeral joint in the lower to middle by the supraspinatus tendon, the tendon of the
ranges of abduction.58'59-63'65'58 Based on anatomic long head of biceps, and the subacromial bursa.
dissections of 100 cadaver shoulders, Ferrari58 One of the most common causes of chronic
found that the middle glenohumeral ligament was shoulder pain is impingement of the structures
a major anterior stabilizer between 60 and 90 within the subacromial space in people with fre¬
degrees of elevation. Attenuation or absence of quent overhead use of the upper extremity. The
this ligament could lead to anterior instability in supraspinatus tendon and the tendon of the long
this part of the range. head of the biceps are the structures most com¬
monly affected. Neer68 described three progressive
Inferior Glenohumeral Ligament phases of impingement: stage I, edema and hemor¬
rhage; stage II, thickening and fibrosis of the bursa
The inferior glenohumeral ligament is crucial to and tendinitis of the cuff; and stage IB, rotator
shoulder stability, particularly in the upper ranges cuff tears, biceps rupture, and bone changes.
of elevation.60-66 The ligament attaches to the ante¬ Neer68 recognized that impingement occurred
rior, inferior, and posterior margins of the glenoid against the anterior edge and undersurface of the
labrum and neck. It inserts into the neck of the anterior third of the acromion and under the
humerus just below the articular margin.59 Turkel acromioclavicular joint, rather than under the lat¬
and coauthors59 described a thickened anterosuper¬ eral acromion as previously thought (Fig. 6-16).
ior portion of the ligament, which they called Most upper extremity functional activities occur
the superior band (see Fig. 6-14). The remaining with the arm in or anterior to the scapular plane.
inferoposterior part of the ligament was termed When elevation occurs in these planes, the supra-
underwent surgical repair of rotator cuff tears,

70% had a type III acromial shape.69
The slope of the acromion has also been impli¬
cated in outlet impingement syndrome.29,71 Ante¬
rior acromial spurs and a narrower supraspinatus
outlet were associated with a flatter acromial
slope71 (Fig. 6-18).
Anterior acromial spurs at the attachment of
the coracoacromial ligament can also narrow the
supraspinatus outlet. These spurs may be due to
chronic impact of the humerus on the ligament,
resulting in development of a traction spur on the
acromion. A prominent acromioclavicular joint or
osteophytes on the undersurface the acromiocla¬
vicular joint also lead to narrowing of the outlet
and can contribute to impingement.29
OSTEOKINEMATICS
The glenohumeral joint is a ball-and-socket syno¬

vial joint with 3 degrees of freedom. Much of the
early research investigated glenohumeral motion
in the anatomic planes of the body or, in other
words, described humeral movement relative to
the trunk rather than to the scapula. Internal and
external rotation occur in the transverse plane, and
the range available is dependent on the degree of
elevation of the arm. Flexion and extension occur
FIGURE 6-15 The supraspinatus outlet is the space
in the sagittal plane, and flexion is accompanied
beneath the anterior acromion, coracoclavicular ligament, and
by some medial rotation.72,73 According to Gagey
acromioclavicular joint. Narrowing of this space is the most
common cause of impingement syndrome. (From Neer CS II:
Shoulder Reconstruction. Philadelphia, WB Saunders, 1990.
Arc of elevation
With permission from Robert Demarest.)
spinatus and biceps tendons pass under the ante¬

rior acromion, the acromioclavicular joint, and the
coracoclavicular ligament.
Impingements have been classified as either
outlet or nonoutlet in type. Outlet impingements
are by far the more common and are a conse¬
quence of factors that result in a narrowing of the
supraspinatus outlet. Variations in the shape and
slope of the acromion, anterior acromial spurs,
and a prominent acromioclavicular joint all can
contribute to outlet impingement.29 Three distinct
acromial shapes were identified by Bigliani and
associates69’70 in both cadaver and in vivo studies
(Fig. 6-17). A type I or flat acromion was identi¬
fied in 18% of 200 patients studied, a type II or
curved acromion was found in 41%, and a type FIGURE 6-16 The functional arc of elevation is in front of,
III or hooked acromion was identified in 41%.69 rather than lateral to, the body, and impingement occurs
A similar distribution was found in 140 cadaver more commonly against the anterior one third of the
shoulders. A type III acromion was more com¬ acromion and undersurface of the acromioclavicular joint.
monly associated with rotator cuff tears in both (From Neer CS II: Shoulder Reconstruction. Philadelphia, WB
cadaver and living subjects. Of 35 patients who Saunders, 1990.)

and 120 degrees when the humerus is allowed to

externally rotate.14-45-75
It has long been recognized that the final hu¬
meral position is the same, regardless of the plane
of elevation. The humeral epicondyle faces for¬
ward, and the humerus is in the plane of the
scapula.73-76 This is the position of maximal osseo-
ligamentous stability and greatest congruency be¬
tween the articular surfaces. Little or no active
rotation is possible with the humerus in this close-
packed position.
Elevation of the humerus occurs naturally in
the plane of the scapula (30 to 40 degrees anterior
to the coronal plane) (Fig. 6-19). Johnston73 ar¬
gued that humeral motion should be described
relative to the scapula rather than to the trunk,
and the terms scapular plane abduction and true
abduction were adopted to describe elevation in
this plane. True flexion and extension occur in a
plane at right angles to the plane of the scapula.
In scapular plane abduction, the deltoid and
supraspinatus muscles are optimally aligned to
abduct the humerus,18- 73 and the glenohumeral
joint capsule is not twisted. Little or no rotation
of the humerus is required when elevation occurs
in this plane. In the final stages of elevation in
either the coronal or sagittal plane, increasing ten¬
sion in components of the capsule and ligamen¬
tous complex occurs, necessitating movement of
the humerus into the scapular plane to achieve full
FIGURE 6-17 Three distinct acromion shapes, flat (A),

curved (B), and hooked (C), as described by Bigliani and
colleagues.69 70 (From Ticker JB, Bigliani LU: Impingement
pathology of the rotator cuff. In Andrews JR, Wilk KE [eds]:
The Athletic Shoulder. New York, Churchill Livingstone,
1994, p 124.)
and colleagues,74 the medial rotation occurs be¬

cause of increasing tension in the coracohumeral
ligament as the humerus elevates in this plane.
Abduction occurs in the coronal plane, and it is
generally accepted that this movement must be
accompanied by external rotation of the humerus.
This external rotation was thought to be necessary
for the greater tuberosity to pass under the acro¬
mion. More recent research suggests that the ex¬
ternal rotation occurs because of increasing ten¬
sion in the inferior capsule and glenohumeral
ligament. Tension in this ligament limits coronal
FIGURE 6-18 A flat acromial angle may result in a
plane abduction to 60 to 90 degrees when the
narrower supraspinatus outlet and cause impingement
humerus is maintained in internal rotation during
syndrome. (From Neer CS II: Shoulder Reconstruction.
elevation in this plane.45 74-75 The range of coronal
Philadelphia, WB Saunders, 1990. With permission from
plane abduction range increases to between 90 Robert Demarest.)
•I Chapter 6 BIOMECHANICS OF THE SHOULDER COMPLEX 155
FIGURE 6-19 The arm is naturally

elevated in the plane of the scapula rather
than in the sagittal or coronal plane of the
body. The capsule is not twisted when
elevation occurs in this plane, and external
humeral rotation is not needed for the
greater tuberosity to clear the acromion.
Flexion and extension occur in a plane at
right angles to the scapular plane.
elevation. The range of elevation possible in the on the glenoid with movement into either flexion
glenohumeral joint during scapular plane elevation or internal rotation from this position. The maxi¬
varies from 100 to 115 degrees.20,22,23 mal extension and external rotation range was
less in subjects with anterior instability, and the
ARTHROKINEMATICS posterior translation of the humeral head did not
occur. Seven of the 12 subjects with instability
Translatory motion of the humeral head on the demonstrated anterior translation of the head on
glenoid during upper extremity motion has been the glenoid during maximal horizontal abduction
studied by several researchers.18,52,77 Poppen and without rotation and in submaximal abduction
Walker18 measured humeral translation on the gle¬ with full external rotation.
noid in a radiographic study of 12 healthy subjects Similarly, Harryman and associates52 reported
and 15 patients with abnormalities of the shoulder. posterior translation of the humeral head on the
During the first 30 to 60 degrees of arm elevation glenoid with extension and lateral rotation and
in the scapular plane, the humeral head moved anterior translation with flexion and medial rota¬
upward on the glenoid about 3 mm in the healthy tion of the glenohumeral joint in a cadaver study.
subjects. The authors hypothesized that this up¬ The degree of anterior translation on flexion in¬
ward motion during early elevation served to cor¬ creased after operative tightening of the posterior
rect the initial depression of the head in the depen¬ capsule.
dent position. Through the remainder of elevation These studies52,77 suggest that translation of the
range, the humeral head was relatively centered humeral head on the glenoid is related to tight¬
on the glenoid, moving upward or downward an ening of the joint capsule as movement approaches
average of 1.09 mm. Excessive upward translation the end of range of motion. Harryman and co¬
occurred in 7 subjects who had experienced a workers52 suggested that as the capsule becomes
previous dislocation, had a rotator cuff tear, or had taut, it forces the humeral head to translate in the
shoulder pain associated with previous trauma. direction of humeral movement. A taut anterior
Howell and colleagues77 studied glenohumeral capsule in abduction and external rotation thus
mechanics during movement in the horizontal forces the humeral head to translate posteriorly on
plane in 20 healthy subjects and 12 patients who the glenoid. That posterior translation did not oc¬
had anterior instability. Measurements were made cur in subjects with anterior capsular laxity lends
from radiographs taken with the subject in supine- support to this hypothesis.
lying position with the arm in varying positions The data presented in these studies appear to
of horizontal abduction and external rotation. In contradict the translation or glide predicted by the
the healthy subjects, the humeral head remained concave-convex rule, which states that when a
centered on the glenoid during all horizontal plane convex surface moves on a concave surface, the
movement, except when the arm was in maximal direction of the glide that accompanies the roll
extension and external rotation (cocking phase of must be in the opposite direction to the roll. Ac¬
pitching). Movement into this position was accom¬ cordingly, the humeral head should slide inferiorly
panied by a 4-mm posterior translation of the during abduction, anteriorly during external rota¬
humeral head on the glenoid. The head recentered tion, and posteriorly during internal rotation. The
research suggests that the direction of the glide is Inman and coauthors27 found that supraspinatus
dictated by tautness in the capsule, at least as end EMG activity increased during elevation in both
range is approached. The direction of mobilization planes, reaching a peak between 80 and 100 de¬
for stiffness may need to be reconsidered in view grees. Activity decreased beyond this point in the
of these findings. range. Saha39 reported a similar pattern of supra¬
Excessive translation of the humeral head on spinatus activity during elevation in the coronal
the glenoid in patients with glenohumeral instabil¬ plane. Peak activity occurred sooner when eleva¬
ity can lead to secondary impingement syndrome. tion was performed in the sagittal plane.27
Laxity of the static stabilizers (capsule and liga¬ The deltoid has the largest cross-sectional area
ments) is common in people who perform activi¬ of the scapulohumeral muscles and also has a
ties requiring repetitive overhead motion, as a greater moment arm than the supraspinatus. Thus,
result of the repetitive physical demands placed the deltoid exerts the largest moments around the
on the shoulder. This laxity can result in superior glenohumeral joint and is more effective in eleva¬
translation of the humeral head and impingement tion.30 Wuelker and associates85 demonstrated
of the tendons of the cuff and biceps under the greater effectiveness of the deltoid in arm eleva¬
coracoacromial arch. Muscular fatigue of the hu¬ tion in cadaver experimentation. The supraspi¬
meral head depressors is also postulated to con¬ natus produced less rotational torque but high
tribute to an increase in superior humeral head compression forces, necessary for stabilization of
translation and impingement syndrome in swim¬ the humeral head on the glenoid.85 van der Helm30
mers.78 also stated that the moments and forces exerted
by the supraspinatus are small because of the
KINETICS small cross-sectional area and small moment arm
and concluded that the supraspinatus contributed
Scapulohumeral Musculature little to the abductor moment when elevation oc¬
curred in the coronal plane. Conflicting informa¬
Deltoid and S up r a s p inatu s tion was reported by Howel and colleagues.81 In
vivo studies of paralysis of either the suprascapu¬
The deltoid and supraspinatus muscles are gener¬ lar or axillary nerves resulted in a 50% reduction
ally considered the prime movers of humeral ele¬ in torque produced during elevation in the scapu¬
vation regardless of the plane in which elevation lar plane and sagittal plane flexion. The authors
occurs.79-82 Inman and coauthors27 demonstrated concluded that the deltoid and supraspinatus were
that both muscles are active throughout the range responsible for 100% of the torque production in
of elevation in both the coronal and sagittal these planes of motion and that the supraspinatus
planes. Deltoid activity increased throughout ele¬ and deltoid contributed equally to torque produc¬
vation to about 100 degrees of elevation and pla- tion.81 The infraspinatus was not considered to
teaued. The level of activity was greater when have a role in abduction. Otis and associates,86
movement was performed in the coronal plane. however, concluded that the infraspinatus can con¬
Similarly, Scheving and Pauly83 demonstrated that tribute to abduction of the arm in the scapular
all three components of the deltoid were active plane based on the results of measures of moment
during coronal and sagittal plane elevation, with arm length in cadavers. Similarly, van der Helm30
the greatest amplitude occurring for the compo¬ reported that the infraspinatus could contribute to
nent of the muscle actually producing the motion elevation in the sagittal plane and to the first 60
(anterior deltoid in flexion and middle deltoid in degrees of coronal plane abduction.
abduction). It was hypothesized that simultaneous Although the deltoid is the prime mover for
contraction of the other components assisted in abduction, its removal because of tumor was
stabilization of the humeral head on the glenoid found to result in only minimal loss of range of
fossa. McCann and colleagues84 also reported pos¬ arm elevation in five patients.87 Muscle strength
terior deltoid activity during elevation in the coro¬ measurements in both flexion and abduction re¬
nal plane but no contribution from this component vealed no greater than a 40% reduction compared
during elevation in the scapular plane. Activity in with the contralateral limb in any of the positions
both the middle deltoid and supraspinatus was measured.87 Thus, other muscles, possibly includ¬
considerably reduced when scapular plane eleva¬ ing the biceps, coracobrachialis, and clavicular
tion was performed with the elbow flexed rather portion of the pectoralis major, are capable of
than extended, and this exercise was recom¬ assuming the role of the deltoid in arm elevation.
mended during the early postoperative period after
Townsend and colleagues88 studied 17 exercises
total shoulder arthroplasty to reduce stress on heal¬
used in a rehabilitation program for baseball play¬
ing tissues.84
ers using indwelling wire electrodes. The purpose
• I Chapter 6 BIOMECHANICS OF THE SHOULDER COMPLEX 157
of the study was to determine which exercises

most effectively elicited activity in the shoulder
musculature and therefore would be the best exer¬
cises for strengthening these muscles. Activity was
measured as a percentage of that obtained during
a maximal isometric contraction. Abduction in the
scapular plane with the arm in internal rotation
(thumb-down position) and a light weight in the
hand resulted in high levels of activity in both the
supraspinatus and middle deltoid muscles. The
military press (see Fig. 6-8) resulted in the highest
level of EMG activity in the supraspinatus muscle.
Infraspinatus, Teres Minor, and

Sub s c apularis
The rotator cuff is the musculotendinous complex

formed by the attachment to the capsule of the
supraspinatus muscle superiorly, the subscapularis
muscle anteriorly, and the teres minor and infra¬ FIGURE 6-20 Force couple consisting of the deltoid, the
spinatus muscles posteriorly. These tendons blend primary abductor of the humerus, and the muscles of the
intricately with the fibrous capsule. They are ma¬ musculotendinous cuff, which function to depress the
jor active stabilizers of the glenohumeral joint and humeral head and oppose the upward translatory force of
can be considered true dynamic ligaments.27 89 The the deltoid.
capsule is less well protected inferiorly because
the tendon of the long head of the triceps brachii
muscle is separated from the capsule by the axil¬ degrees of elevation and disappear beyond 120
lary nerve and the posterior circumflex humeral ar¬ degrees. Weakness or loss of the humeral head
tery.2 depressors may lead to increased superior migra¬
With the arm at the side, the directional force tion of the humeral head during arm elevation,
of the deltoid muscle is almost vertical.45,64 Thus, contributing to impingement of the supraspinatus
most of the deltoid force will cause upward trans¬ and possibly the biceps tendons against the acro¬
lator motion of the humeral head, which if unop¬ mion process.29 Thus, instability often coexists
posed would cause the humeral head to contact with impingement of subacromial structures.
the coracoacromial arch, resulting in impingement Strengthening of the rotator cuff muscles has been
of soft tissues.90 The action line of the infraspi¬ advocated to improve their efficiency as humeral
natus, subscapularis, and teres minor muscles are head depressors.91
such that each tends to have a rotatory component Inman and coauthors27 studied the EMG activity
as well as a compressive force.64,90 Each also has of the subscapularis, infraspinatus, and teres minor
a downward translator component that offsets the muscles during elevation of the arm in the coronal
upward translation force of the deltoid32, 45, 85,91 and sagittal planes. All three muscles were found
(Fig. 6-20). to work continuously throughout arm elevation in
Wuelker and associates,85 based on cadaver ex¬ both planes. Teres minor and subscapularis activ¬
perimentation, found that the supraspinatus did ity peaked between 100 and 120 degrees of eleva¬
not contribute to the depressive forces on the tion and declined through the remaining elevation
humeral head. The infraspinatus, teres minor, and range. Subscapularis activity was greater when
subscapularis thus form a force couple with the elevation was carried out in the coronal plane.
deltoid and act to stabilize the humeral head on Infraspinatus activity increased throughout the en¬
the glenoid fossa, allowing the deltoid to act to tire motion in the coronal plane. Greater activity
abduct the humerus.76,91 The compressive and was observed in this muscle during sagittal plane
downward translation actions of these muscles are flexion, with two peaks of activity occurring at 60
thought to be most important in the middle range and 120 degrees of elevation.27
of elevation when the capsule and ligaments are van der Helm30 predicted the contribution of
relatively lax.92 the infraspinatus, teres minor, and subscapularis
In studies on a mechanical model, Comtet and muscles to glenohumeral elevation using a finite
colleagues79 determined that the depressor forces element model. The model predicted that the teres
would be at their maximum between 60 and 80 minor would not contribute to either coronal or
sagittal plane elevation. The infraspinatus contrib¬ The biceps brachii is postulated to contribute to
uted to sagittal plane elevation and the first 60 both superior and anterior stability of the glenohu¬
degrees of coronal plane abduction. The subscapu- meral joint. Kumar and coauthors,95 in an in vitro
laris was found to be the second most important study, measured the acromial humeral distance
(after deltoid) contributor to coronal plane abduc¬ from radiographs with tension applied to one or
tion, owing to its large cross-sectional area and both heads of the biceps, before and after division
useful moment arm, but would not contribute to of the tendon of the long head. With the long head
sagittal plane elevation. divided, application of tension to the short head
The EMG activity of these muscles has also resulted in upward migration of the humeral head
been studied during performance of rehabilitation on the glenoid. The authors concluded that the
exercises in healthy subjects using indwelling wire long head was an important stabilizer of the shoul¬
electrodes. Exercises that generated at least half der during activities involving elbow flexion and
the amount of activity obtained during a maximal supination. Action of the cuff musculature was
isometric contraction were considered to be suit¬ not simulated in this study, and it is unclear
able choices for a strengthening program.88 The whether normal functioning of these muscles in
most effective exercise for recruitment of the sub- vivo would compensate for loss of the long head
scapularis was abduction in the scapular plane of biceps. Neer29 also postulated that rupture of the
with the arm in internal rotation (thumb-down long head of the biceps could result in elevation of
position) and a light weight in the hand. The the humeral head during arm movements and a
exercise that generated the greatest EMG activity nonoutlet type of impingement of the humeral
in the infraspinatus and teres minor was horizontal head against the acromion process.
abduction of the arm in the prone-lying position The contribution of the long and short heads
with the humerus in external rotation and a light of the biceps to anterior stability has also been
weight held in the hand. The same exercise with investigated in cadavers.96,97 Anterior translation
the arm in internal rotation also generated high forces were applied to the humeral head with the
levels of EMG activity in these muscles (see Fig. arm in 90 degrees of abduction and 60, 90, or 120
6-9). degrees of external rotation with and without loads
McCann and coauthors84 reported moderate to applied to either the long or short head. Tension
high levels of EMG activity in the external rota¬ applied to both the long and short heads was
tors during resistive exercises performed with an found to decrease anterior translation of the head
elastic band below shoulder level and advocated on the glenoid at 60 and 90 degrees of external
this technique for strengthening after shoulder re¬ rotation. The degree of anterior translation that
construction. Similarly, Harms-Ringdahl and col¬ occurred when anterior translation forces were ap¬
leagues93 advocated resisted external rotation exer¬ plied at 120 degrees of external rotation was mini¬
cises using a pulley apparatus to strengthen the mal, possibly because of the extreme tautness of
infraspinatus muscle and determined the optimal the capsule in this position. Loading of the biceps
patient positioning to maximize muscle strength¬ tendons decreased anterior translation at this angle
ening without excessive joint loading. of external rotation only when a Bankart lesion
(sectioning of the anteroinferior capsule) was pres¬
Biceps Brachii ent. The authors concluded that the biceps is an
important anterior stabilizer, particularly when the
The tendon of the long head originates on the anteroinferior capsule is compromised, and sug¬
supraglenoid tubercle of the scapula, crosses over gested that strengthening of the biceps was indi¬
the anterolateral aspect of the glenohumeral joint, cated for patients with chronic anterior shoulder
and passes through the intertubercular groove. The instability.96,97 Glousman’s98 observation that
tendon lies within the joint capsule within its own EMG activity of the biceps was greater during the
synovial sleeve. The short head originates from acceleration phase of pitching in subjects with
the coracoid process. Furlani,94 in an EMG study unstable shoulders also suggests that the biceps
of healthy men, demonstrated that both heads of may help to compensate for anterior instability.
the biceps are active during free and resisted
flexion of the shoulder with the elbow extended.
Axiohumeral Musculature
Little activity was seen in either the long or short
head during free or resisted shoulder abduction, Two muscles, latissimus dorsi and pectoralis ma¬
and neither component contributed in medial or jor, connect the axial skeleton and humerus. The
lateral rotation movements. The short head was latissimus dorsi also has an attachment on the
active during resisted adduction in half of the 30 scapula and is capable of scapular adduction and
subjects tested.94
depression as well as glenohumeral extension, ad-
duction, and medial rotation.83 The latissimus pectoralis major, and latissimus dorsi were postu¬
dorsi is particularly active in closed kinetic chain lated to contribute to this compression in some
activities in which body weight is supported by shoulder positions.
the upper extremities, for example, when using The magnitude and direction of forces at the
crutches or doing a push-up. Contraction of the glenohumeral joint have been estimated by several
muscle during these activities results in elevation researchers.27- 90- 100 Inman and coauthors27 esti¬
of the trunk in relation to the humerus.32-88 The mated the compressive forces due to activity in
sternal portion of pectoralis major also contributes the deltoid and rotator cuff musculature during
to depression of the shoulder complex during up¬ arm abduction in the coronal plane. A maximum
per extremity weight-bearing activities.88 Both compressive force of 50% body weight was esti¬
muscles were found to have high levels of EMG mated to occur at 90 degrees of abduction. Poppen
activity during a press-up exercise88 (see Fig. 6- and Walker90 estimated the compressive, sheer,
10) and during the pull-through phase of the free¬ and resultant forces at the glenohumeral joint dur¬
style swim stroke." ing scapular plane abduction with neutral rotation.
In open kinetic chain motions, both the clavicu¬ The resultant force increased with elevation,
lar and sternal portions of the pectoralis major reaching a peak of 0.89 times body weight at 90
assist the anterior deltoid in elevation of the hu¬ degrees of elevation. The upward sheer force of
merus in the sagittal plane, with the clavicular the humeral head on the glenoid peaked at 0.42
portion being the more active.83 The pectoralis times body weight at 60 degrees of elevation. The
major also contributes to adduction and resisted authors suggested that the greater compressive
medial rotation of the humerus in open kinetic forces, compared with those reported by Inman
chain movements.83 Both latissimus dorsi and pec¬ and coauthors,27 may have been related to a differ¬
toralis major may also contribute to the depressor ence in assumptions regarding the resultant line of
forces on the humeral head.88 action of the depressor cuff muscle. Karlsson and
Peterson,100 using a computer model, predicted a
FORCES AT THE GLENOHUMERAL JOINT resultant glenohumeral contact force of 600 N (0.8
times body weight) between 60 and 90 degrees of
Lippitt and Matsen48 discussed the concept of con¬ scapular plane elevation.
cavity compression in maintaining glenohumeral Three-dimensional biomechanical models have
joint stability throughout the range of motion. The also been used to predict muscle forces around
concavity component related to the depth of the the glenohumeral joint.101- 102 Predicted forces in
glenoid fossa and its surrounding labrum, as pre¬ the subscapularis, middle and anterior deltoid, su-
viously discussed. The compressive component praspinatus, and infraspinatus ranged from 45 N
referred to the degree to which forces compressing (supraspinatus) to 167 N (middle deltoid) when
the humeral head into the glenoid fossa contribute the arm was held in 90 degrees of scapular plane
to stability of the articulation. The degree to which abduction with the elbow extended. The same
compressive forces could limit translation of the muscles were active during a maximal abduction
humeral head on the glenoid was investigated us¬ effort, at 90 degrees of elevation, with forces
ing fresh-frozen cadaver shoulders, with the sur¬ increasing by 260%. The greatest forces were pre¬
rounding musculature resected. The magnitude dicted for the middle and anterior deltoids (323 N
and direction of applied compressive and transla¬ and 434 N, respectively), whereas the posterior
tor forces were measured using magnetic spatial deltoid was inactive in both unloaded and loaded
sensors and force transducers.48 Compressive conditions. Subscapularis force was greatest dur¬
forces of 50 and 100 N were applied at various ing a resisted internal rotation effort with the arm
points in the range of motion, followed by the in an abducted position (1725 N). Resisted exter¬
application of translatory forces in each position. nal rotation in an abducted position generated
The higher compressive load resulted in increased forces of 175 N and 723 N for the supraspinatus
resistance to translatory forces, leading the authors and infraspinatus, respectively.
to conclude that the greater the compression, the Maximal teres minor forces were predicted dur¬
more stable the joint. The authors suggested that, ing a maximal external rotation effort with the
in vivo, the rotator cuff muscles and the biceps arm in an adducted position. Force predictions for
brachii are aligned such that they contribute to the supraspinatus were 150% greater during maximal
compressive force of the humeral head on the external rotation exertions than during maximal
glenoid. It was suggested that weakened or dam¬ abduction exertion, and the authors suggested that
aged cuff muscles might be less able to withstand reduction of internal rotation loading be consid¬
translatory forces, leading to instability during ered in the prevention and rehabilitation of rotator
movement.48 The outer muscular sleeve, deltoid, cuff impairments. The posterior deltoid was active
during internal rotation efforts in both the ab¬ the 180 degrees described by Inman and coau¬
ducted and adducted positions and during external thors.27 Total scapular rotations of 58.620 and
rotation efforts with the arm adducted. The force 63.823 degrees have been reported. Elevation at
predicted for the middle deltoid during a maximal the glenohumeral joint varied from 103 to 113
internal rotation effort in the abducted position degrees.20,22,23
exceeded that predicted for the resisted abduction The ratio of glenohumeral to scapulothoracic
effort at 90 degrees of scapular plane elevation. motion, measured during elevation in the scapular
Muscle forces have also been predicted at 15- plane, was found to vary throughout the range and
degree intervals between 0 and 120 degrees of to be highly variable among subjects.20, 22,23, 104
scapular plane elevation with a 1-kg weight in the Hogfors and colleagues104 reported that there was
hand. Force levels of up to 150 N were predicted considerable variation in scapulohumeral relation¬
for the middle and anterior deltoids, supraspinatus, ship among subjects; however, the rhythm within
infraspinatus, subscapularis, clavicular portion of a subject was stable and was unaffected by small
the pectoralis major, biceps, and coracobrachialis loads held in the hand.
during the movement. Greatest forces were pre¬ Glenohumeral motion predominates during ini¬
dicted for the middle deltoid (150 N) at about 100 tial arm elevation. A glenohumeral-to-scapulotho-
degrees of elevation. Zero or low levels of activity racic ratio as high as 7.29:1 has been reported
were predicted in the subscapularis during the during the first 30 degrees of elevation.20 A ratio
early phase of elevation and for the supraspinatus of 3.29:1 was reported by Bagg and Forrest23
and infraspinatus at higher elevation angles. These between 20 and 80 degrees of elevation. Scapular
muscles demonstrated EMG activity throughout motion predominates between 80 and 140 degrees
scapular plane elevation, and it was suggested that of total arm elevation. Glenohumeral-to-scapulo-
the activity was related to their role as humeral thoracic ratios of 0.79:120 and 0.71:123 have been
head stabilizers. reported during this phase of elevation. Glenohu¬
meral motion again predominates above 140 or
Integrated Function of the 150 degrees of total arm motion.20, 23 Bagg and
Shoulder Complex Forrest23 reported a glenohumeral-to-scapulotho-
racic ratio of 3.49:1 during this phase in subjects
SCAPULOHUMERAL RHYTHM demonstrating the most typical pattern of move¬
ment.
Full elevation of the upper extremity results from Alteration in the normal scapulohumeral rhythm
the combination of scapular rotation, such that has not been extensively studied in subjects with
the glenoid fossa tilts progressively upward, and shoulder pathology. It has been postulated that
elevation of the humerus at the glenohumeral weakness or fatigue of the scapular rotators, par¬
joint. Codman103 introduced the term scapulohu¬ ticularly the serratus anterior, may contribute to
meral rhythm to describe this integrated move¬ shoulder instability related to a failure of these
ment at the glenohumeral and scapulothoracic muscles to maintain a stable glenoid base for
joints during upper extremity elevation. Glenohu¬ humeral movement.98,99, 105, 106 It has also been
meral and scapular motions were studied during suggested that dysfunction of these muscles could
arm elevation in the coronal and sagittal planes lead to insufficient upward movement of the gle¬
by Inman and coauthors.27 Glenohumeral and noid during overhead activities, resulting in sec¬
scapular rotation were found to contribute a maxi¬ ondary impingement of structures in the sub¬
mum of 120 and 60 degrees, respectively, to total acromial space.98, 99 Glousman and associates105
arm motion. An initial scapular “setting phase” have indicated that the trapezius and serratus ante¬
was described, in which the scapula moved medi¬ rior may be more prone to fatigue than other
ally or laterally or remained fixed. This phase muscles near the shoulder during activities requir¬
constituted the first 30 degrees of abduction or ing repetitive overhead motion, lending some sup¬
first 60 degrees of flexion. Most of the motion port to these hypotheses.
during the setting phase occurred at the glenohu¬ Abnormal scapular motion has been identified
meral joint. The overall ratio of glenohumeral to in subjects with inferior and multidirectional insta¬
scapular thoracic rotation throughout full elevation bility of the glenohumeral joint using cineradiog¬
was 2:1.27
raphy.107 Scapular motion and humeral motion
Scapulohumeral rhythm during arm elevation were measured through arm elevation in the scap¬
has been examined more recently during scapular ular plane in 11 subjects with instability and in 30
plane elevation.20,22,23 Total range of upper ex¬ normal shoulders. The scapula did not rotate up¬
tremity elevation measured in these studies varied ward to the same degree in the subjects with
between 168 and 172 degrees, somewhat less than instability as in those with normal shoulders.107
Similarly, Nobuhara and Ikeda56 reported abnor¬ 3. Beam JG: Direct observations on the function of
mal scapular motion in subjects with instability the capsule of the sternoclavicular joint in
due to lesions of the capsule and ligaments in the clavicular support. J Anat 101:159, 1967.
4. Abbott LC, Lucas DB: The function of the
rotator interval, and Poppen and Walker18 found an
clavicle: Its surgical significance. Ann Surg
abnormal glenohumeral-to-scapulothoracic ratio in
140:583, 1954.
9 of 15 subjects with shoulder pathology.
5. Flatow EL: The biomechanics of the
Alterations in scapular mechanics and EMG acromioclavicular, sternoclavicular, and
activity in the scapular muscles has been identified scapulothoracic joints. Instr Course Lect 42:237,
in swimmers with shoulder pain.108 Decreased ac¬ 1993.
tivity in the serratus anterior was reported 6. Moseley HF: The clavicle: Its anatomy and
throughout the free-style stroke in swimmers with function. Clin Orthop 58:17, 1968.
painful shoulders. A decrease in activity in the 7. Moore KL: Clinically Oriented Anatomy.
anterior and middle deltoids, rhomboids, and up¬ Baltimore, Williams & Wilkins, 1980.
per trapezius was also reported to be associated 8. Fukuda K, Craig EV, An K, et al: Biomechanical
study of the ligamentous system of the
with a more medial hand placement during the
acromioclavicular joint. I Bone loint Surg
early pull-through phase. It is unclear whether the
68A:434, 1986.
altered mechanics and EMG activity contributed 9. Rockwood CA Ir: Dislocations about the
to the shoulder pathology or were related to an shoulder. In Rockwood CA Jr, Green DP (eds):
alteration in the stroke because of the pain. Fractures. Philadelphia, JB Lippincott, 1975, p
Warner and colleagues106 studied symmetry of 624.
scapulothoracic motion in 22 healthy subjects, 22 10. Herscovici D, Sanders R, DiPasquale T, Gregory
patients with anteroinferior instability, and 7 sub¬ P: Injuries of the shoulder girdle. Clin Orthop
jects with impingement syndrome using a Moire 318:54, 1995.
topographic analysis technique. Static tests in¬ 11. Richards R: Acromioclavicular joint injuries.
volved holding a 4.5-kg weight in each hand with Instr Course Lect 42:259, 1993.
12. Larson E, Bjerg-Nielsen A, Christensen P:
both arms held in a 90-degree forward flexed
Conservative or surgical treatment of
position with elbows extended for 5 seconds. For acromioclavicular dislocation: A prospective,
the dynamic tests, subjects lifted the same weight controlled randomized study. J Bone Joint Surg
through 0 to 120 degrees of forward flexion bilat¬ 68A:552, 1986.
erally with elbows flexed to 60 degrees. Sixty- 13. Taft TN, Wilson FC, Oglesby JW: Dislocation of
four percent of subjects with instability and 100% the acromioclavicular joint: An end-result study.
of those with impingement demonstrated some J Bone Joint Surg 69A:1045, 1987.
abnormality in the scapulothoracic motion during 14. Steindler A: Kinesiology of the Human Body
the dynamic flexion test. In the static test, the Under Normal and Pathological Conditions.
scapula was lower on the side of the unstable Springfield, IL, Charles C Thomas, 1955.
15. Kapandji IA: The Physiology of the Joints, vol 1:
shoulder in the instability group compared with
Upper Limb, 2nd ed. Edinburgh, E & S
the contralateral side but was higher on the side
Livingstone, 1970.
of the lesion in the impingement group relative to 16. Norkin CC, Levangie PK: Joint Stmcture and
the contralateral side. The differences in both the Function, 2nd ed. Philadelphia, FA Davis, 1992.
static and dynamic tests were attributed to weak¬ 17. Bechtol CO: Biomechanics of the shoulder. Clin
ness of the serratus anterior and trapezius muscles. Orthop 146:37, 1980.
Whether the alterations are primary or secondary 18. Poppen NK, Walker PS: Normal and abnormal
is unknown. motion of the shoulder. J Bone Joint Surg
In summary, full elevation and normal mechan¬ 58A:195, 1976.
ics of the shoulder complex is dependent on simul¬ 19. Culham E, Peat M: Spinal and shoulder complex
posture. I. Measurement Using the 3Space
taneous, coordinated motions of the scapulotho¬
Isotrak. Clin Rehabil 7:309, 1993.
racic, sternoclavicular, acromioclavicular, and
20. Doody SG, Freedman L, Waterland JC: Shoulder
glenohumeral joints. Disruptions of any of the
movements during abduction in the scapular
normal joint mechanics or altered joint structure plane. Arch Phys Med Rehabil 51:595, 1970.
can result in the development of clinical patholo¬ 21. Basmajian JV, Bazant FJ: Factors preventing
gies and disruption of function. downward dislocation of the adducted shoulder
joint. J Bone Joint Surg 41A:1182, 1959.
References 22. Freedman L, Munro RR: Abduction of the arm
in the scapular plane: Scapular and glenohumeral
1. Dempster WT: Mechanisms of shoulder
movement. Arch Phys Med Rehabil 46:49, 1965. movements. J Bone Joint Surg 48A:1503, 1966.
2. Warwick R, Williams P: Gray’s Anatomy, 37th 23. Bagg SD, Forrest WJ: A biomechanical analysis
ed. London, Longman Group, 1989. of scapular rotation during arm abduction in the
scapular plane. Am J Phys Med Rehabil 67:238, non-traumatic posterior instability of the
1988. shoulder. J Bone Joint Surg 68A:724, 1986.
24. Dvir Z, Berme N: The shoulder complex in 43. Cooper DE, Amoczky SP, O’Brien SJ, et al:
elevation of the arm: A mechanism approach. J Anatomy, histology, and vascularity of the
Biomech 11:219, 1978. glenoid labrum. J Bone Joint Surg 74-A:46,
25. Singleton MC: Functional anatomy of the 1992.
shoulder: A review. J Am Phys Ther Assoc 44. Moseley HF, Overgaard B: The anterior capsular
46:1043, 1966. mechanism in recurrent anterior dislocation of
26. Ljunggren AE: Clavicular function. Acta Orthop the shoulder: Morphological and clinical studies
Scand 50:261, 1979. with special reference to the glenoid labrum and
27. Inman VT, Saunders JB, Abbott LC: the gleno-humeral ligaments. J Bone Joint Surg
Observations on the function of the shoulder 44B:913, 1962.
joint. J Bone Joint Surg 26:1, 1944. 45. Lucas DB: Biomechanics of the shoulder joint.
28. Peat M: Functional anatomy of the shoulder Arch Surg 107:425, 1973.
complex. Phys Ther 66:1855, 1986. 46. Perry J: Anatomy and biomechanics of the
29. Neer CS: Shoulder Reconstruction. Philadelphia, shoulder in throwing, swimming, gymnastics, and
WB Saunders, 1990. tennis. Clin Sports Med 2:247, 1983.
30. van der Helm FCT: Analysis of the kinematic 47. Howell SM, Galinat BJ: The glenoid-labral
and dynamic behavior of the shoulder socket: A constrained articular surface. Clin
mechanism. J Biomech 27:527, 1994. Orthop 243:122, 1989.
31. Bagg SD, Forrest WJ: Electromyographic study 48. Lippitt S, Matsen F: Mechanisms of
of the scapular rotators during arm abduction in glenohumeral joint stability. Clin Orthop 291:20,
the scapular plane. Am J Phys Med 65111 1993.
1986.
49. Zarins B, McMahon MS, Rowe CR: Diagnosis
32. Perry J: Normal upper extremity kinesiology. and treatment of traumatic anterior instability of
Phys Ther 58:265, 1978.
the shoulder. Clin Orthop 291:75, 1993.
33. Wiedenbauer MM, Mortenson OA: An 50. Clark J, Sidles JA, Matsen FA III: The
electromyographic study of the trapezius muscle. relationship of the glenohumeral joint capsule to
Am J Phys Med 31:363, 1952.
the rotator cuff. Clin Orthop 254:29, 1990.
34. Moseley JB, Jobe FW, Pink M, et al: EMG 51. Kumar VP, Balasubramaniam P: The role of
analysis of the scapular muscles during a
atmospheric pressure in stabilising the shoulder:
shoulder rehabilitation program. Am J Sports An experimental study. J Bone Joint Surg
Med 20:128, 1992. 67B:719, 1985.
35. Mathiassen SE, Winkel J: Electromyographic 52. Harryman DT II, Sidles JA, Clark JM, et al:
activity in the shoulder-neck region according to
Translation of the humeral head on the glenoid
arm position and glenohumeral torque. Eur J
with passive glenohumeral motion. J Bone Joint
Appl Physiol 61:370, 1997. Surg 72A:1334, 1990.
36. Randelli M, Gambrioli PL: Glenohumeral
53. Cooper DE, O’Brien SJ, Warren RF: Supporting
osteometry by computed tomography in normal
layers of the glenohumeral joint. Clin Orthop
and unstable shoulders. Clin Orthop 208T 51 289:144, 1993.
1986.
54. Harryman DT, Sidles JA, Harris SL, Matsen FA
37. Cyprien JM, Vasey HM, Burdet A, et al:
III: The role of the rotatory interval capsule in
Humeral retrotorsion and glenohumeral
passive motion and stability of the shoulder. J
relationship in the normal shoulder and in Bone Joint Surg 74A:53, 1992.
recurrent anterior dislocation (scapulometry).
55. Kaltsas DS: Comparative study of the properties
Clin Orthop 175:8, 1983.
of the shoulder joint capsule with those of other
38. Soslowsky J, Flatow EL, Bigliani LU, Mow VC:
joint capsules. Clin Orthop 173:20, 1983.
Articular geometry of the glenohumeral joint.
56. Nobuhara K, Ikeda H: Rotator interval lesion.
Clin Orthop 285:181, 1992.
Clin Orthop 223:44, 1987.
39. Saha AK: Dynamic stability of the glenohumeral
57. Neer CS, Satterlee CC, Dalsey RM, Flatow EL:
joint. Acta Orthop Scand 42:491, 1971.
The anatomy and potential effects of contracture
40. Iannotti JP, Gabriel JP, Schneck SL, et al: The
of the coracohumeral ligament. Clin Orthop
normal glenohumeral relationships. J Bone Joint 280:182, 1992.
Surg 74A:491, 1992.
58. Ferrari DA: Capsular ligaments of the shoulder:
41. Saha AK: Mechanics of elevation of
Anatomical and functional study of the anterior
glenohumeral joint: Its application in
superior capsule. Am J Sports Med 18:20, 1990.
rehabilitation of flail shoulder in upper brachial
59. Turkel SJ, Panio MW, Marshall JL, Girgis FG:
plexus injuries and poliomyelitis and in
Stabilizing mechanisms preventing anterior
replacement of the upper humerus by prosthesis.
dislocation of the glenohumeral joint. J Bone
Acta Orthop Scand 44:668, 1973.
Joint Surg 63A: 1208, 1981.
42. Brewer BJ, Wubben RC, Carrera GF: Excessive
60. O’Brien SJ, Neves MC, Amoczky SP, et al: The
retroversion of the glenoid cavity: A cause of
anatomy and histology of the inferior
glenohumeral ligament complex of the shoulder. 80. deLuca CJ, Forrest WJ: Force analysis of
Am J Sports Med 18:449, 1990. individual muscles acting simultaneously on the
61. Bowen MK, Warren RF: Ligamentous control of shoulder joint during isometric abduction. J
shoulder stability based on selective cutting and Biomech 6:385, 1973.
static translation experiments. Clin Sports Med 81. Howell SM, Imobersteg AM, Seger DH, Marone
10:757, 1991. PJ: Clarification of the role of the supraspinatus
62. Schwartz E, Warren RE, O’Brien SJ: Posterior muscle in shoulder function. J Bone Joint Surg
shoulder instability. Orthop Clin North Am 68A:398, 1986.
18:409, 1987. 82. Colachis SC, Strohm BR: Effect of suprascapular
63. Ovesen J, Nielsen S: Anterior and posterior and axillary nerve blocks on muscle force in
shoulder instability: A cadaver study. Acta upper extremity. Arch Phys Med Rehabil 52:22,
Orthop Scand 57:324, 1986. 1971.
64. Sarrafian SK: Gross and functional anatomy of 83. Scheving LE, Pauly JE: An electromyographic
the shoulder. Clin Orthop 173:11, 1983. study of some muscles acting on the upper
65. Ovesen J, Nielsen S: Stability of the shoulder extremity of man. Anat Rec 135:239, 1959.
joint: Cadaver study of stabilizing structures. 84. McCann PD, Wootten ME, Kadaba MP, Bigliani
Acta Orthop Scand 56:149, 1985. LU: A kinematic and electromyographic study of
66. Bigliani LU, Pollock RG, Soslowsky LJ, et al: shoulder rehabilitation exercises. Clin Orthop
Tensile properties of the inferior glenohumeral 288:179, 1993.
ligament. J Orthop Res 10:187, 1992. 85. Wuelker N, Plitz W, Roetman B, Wirth CJ:
67. Rothman RH, Marvel JP, Heppenstall RB: Function of the supraspinatus muscle. Acta
Anatomic considerations in the glenohumeral Orthop Scand 65:442, 1994.
joint. Orthop Clin North Am 6:341, 1975. 86. Otis JC, Jiang CC, Wickeiwicz TL, et al:
68. Neer CS: Impingement lesions. Clin Orthop Changes in the moment arms of the rotator cuff
173:70, 1983. and deltoid muscles with abduction and rotation.
69. Bigliani LU, Ticker JB, Flatow EL, et al: The J Bone Joint Surg 76A:667, 1994.
relationship of acromial architecture to rotator 87. Markhede G, Monastyrski J, Stener B: Shoulder
cuff disease. Clin Sports Med 10:823, 1991. function after deltoid removal. Acta Orthop
70. Bigliani LU, Morrison DS, April EW: The Scand 56:242, 1985.
morphology of the acromion and rotator cuff 88. Townsend H, Jobe FW, Pink M, Perry J:
impingement. Orthop Trans 10:228, 1986. Electromyographic analysis of the glenohumeral
71. Aoki M, Ishii I, Usui M: The slope of the muscles during a baseball rehabilitation program.
acromion and rotator cuff impingement. Orthop Am J Sports Med 19:264, 1991.
Trans 10:228, 1986. 89. Brostrom L, Kronberg M, Nemeth G: Muscle
72. Blakely RL, Palmer ML: Analysis of rotation activity during shoulder dislocation. Acta Orthop
accompanying shoulder flexion. Phys Ther Scand 60:639, 1989.
64:1214, 1984. 90. Poppen NK, Walker PS: Forces at the
73. Johnston TB: The movements of the shoulder- glenohumeral joint in abduction. Clin Orthop
joint: A plea for the use of the “plane of the 135:165, 1978.
scapula” as the plane of reference for 91. Sharkey NA, Marder RA: The rotator cuff
movements occurring at the humero-scapular opposes superior translation of the humeral head.
joint. Br J Surg 25:252, 1937. Am J Sports Med 23:270, 1995.
74. Gagey O, Bonfait H, Gillot C, et al: Anatomic 92. Speer KP: Anatomy and pathomechanics of
basis of ligamentous control of elevation of the shoulder instability. Clin Sports Med 14:751,
shoulder (reference position of the shoulder 1995.
joint). Surg Radiol Anat 9:19, 1987. 93. Harms-Ringdahl K, Arborelius UP, Ekholm J, et
75. Cailliet R: Shoulder Pain. Philadelphia, FA al: Shoulder externally rotating exercises with
Davis, 1966. pulley apparatus. Scand J Rehabil Med 17:129,
76. Saha AK: Mechanism of shoulder movements 1985.
and a plea for the recognition of “zero position” 94. Furlani J: Electromyographic study of the m.
of the glenohumeral joint. Clin Orthop 173:3, biceps bracii in movements at the glenohumeral
1983. joint. Acta Anat 96:270, 1976.
77. Howell SM, Galinat BJ, Renzi AJ, Marone PJ: 95. Kumar VP, Satku K, Balasubramaniam P: The
Normal and abnormal mechanics of the role of the long head of biceps brachii in the
glenohumeral joint in the horizontal plane. J stabilization of the head of the humerus. Clin
Bone Joint Surg 70A:227, 1988. Orthop 244:172, 1989.
78. Allegrucci M, Whitney SL, Irrgang JJ: Clinical 96. Itoi E, Kuechle DK, Newman SR, et al:
implications of secondary impingement of the Stabilizing function of the biceps in stable and
shoulder in freestyle swimmers. J Orthop Sports unstable shoulders. J Bone Joint Surg 75B:546,
Phys Ther 20:307, 1994. 1993.
79. Comtet JJ, Herberg G, Naasan IA: 97. Itoi E, Newman SR, Kuechle DK, et al: Dynamic
Biomechanical basis of transfers for shoulder anterior stabilizers of the shoulder with the arm
in abduction. J Bone Joint Surg 76B:834, 1994.
paralysis. Hand Clin 5:1, 1989.
98. Glousman R: Electromyographic analysis and its Biomechanical model of the human shoulder
role in the athletic shoulder. Clin Orthop 288:27, joint. II. The shoulder rhythm. J Biomech
1993. 24:699, 1991.
99. Nuber GW, Jobe FW, Perry J, et al: Fine wire 105. Glousman R, Jobe F, Tibone J, et al: Dynamic
electromyography analysis of muscles of the electromyographic analysis of the throwing
shoulder during swimming. Am J Sports Med shoulder with glenohumeral instability. J Bone
14:7, 1986. Joint Surg 70A:220, 1988.
100. Karlsson D, Peterson B: Towards a model for 106. Warner JJP, Micheli LJ, Arslanian LE, et al:
force predictions in the human shoulder. J Scapulothoracic motion in normal shoulders and
Biomech 25:189, 1990. shoulders with glenohumeral instability and
101. Karlsson D, Peterson B: Towards a model for impingement syndrome. Clin Orthop 285:191,
1992.
force predictions in the human shoulder. J
107. Ozaki J: Glenohumeral movements of the
Biomech 25:189, 1992.
involuntary inferior and multidirectional
102. Hughes RE, An K: Force analysis of rotator cuff
instability. Clin Orthop 238:107, 1989.
muscles. Clin Orthop 330:75, 1996.
108. Scovazzo ML, Browne A, Pink M, et al: The
103. Codman EA: The Shoulder. Boston, Thomas
painful shoulder during freestyle swimming: An
Todd, 1934.
electromyographic cinematographic analysis of
104. Hogfors C, Peterson B, Sigholm G, Herberts P: twelve muscles. Am J Sports Med 19:577, 1991.
CHAPTER 7
Biomechanics of Neurologic
Treatment
Richard W. Bohannon
INTRODUCTION length-tension curve to the left (Fig. 7—l ).2-3 In

mature mammals, the changes involve the shorten¬
For an individual to function effectively, he or ing of the muscle belly secondary to a loss of
she must be able to generate forces that are of sarcomeres in series.4 This loss occurs in both
appropriate magnitude and rapidity, are applied in innervated and denervated muscles.5 In growing
the appropriate direction, and are coordinated. children, another source of increased stiffness is
This chapter is concerned primarily, but not exclu¬ possible.6 As children grow in stature, their mus¬
sively, with the ability of patients with neuro¬ cles or associated tendons may not grow at a rate
logic disorders to generate forces (and therefore comparable to that of the bones to which they are
torques) of sufficient magnitude and rapidity to attached. As a consequence, the muscle belly or
function. Prior to discussing the treatment of force tendon of the unit may become short relative to
production problems, the determinants, nature, and its associated bones.
implications of such problems will be discussed. The active component of muscle force reflects
Topics relevant to the biomechanics of neurologic motor unit activity. The number of motor units
treatment but covered elsewhere in detail (e.g., activated, the frequency of motor unit activation,
posture and balance, and orthoses) will receive and the size of activated motor units are the major
limited attention. determinants of the force generated by a pool of
motor units over a period of time. All of these
factors can be altered in the presence of neurologic
Force Production by Patients disorders. Lesions of either the central or periph¬
With Neurologic Disorders eral nervous system can decrease the total volume
of motor drive through a reduction in the number
DETERMINANTS OF FORCE PRODUCTION of motor units that can be activated7-10 and the
frequency with which they are activated.11-13 These
Muscles are the means by which the vast majority reductions can result from a destruction of motor
of forces are generated within humans. The forces nerves, a block of the conduction of motor axons,
have both a passive and an active element. or a slowing of the velocity at which they conduct
The passive component of muscle force reflects impulses. Fortunately, the body has means for
the contribution of the musculotendinous unit’s compensating for the loss of motor unit drive. At
elasticity and is reflected in the unit’s passive the level of the motor unit, the body sometimes
length-tension curve. As the musculotendinous compensates for a loss of functioning units by
unit is elongated (stretched), its passive force in¬ expanding the territory of remaining motor units9'
creases curvilinearly.1 After neurologic insults, 14-17 or by increasing the ratio of motor unit firing
particularly those involving the central nervous per unit of force.18'19 Another way in which the
system, the musculotendinous unit can become body compensates is by prolonging the duration
increasingly resistant to stretch. This resistance, that a pool of motor units is activated. In the
which is largely the consequence of changes in presence of an inadequate magnitude of muscle
the musculotendinous unit, results in a shift of the drive, the duration of drive can be extended to
165
associated with spasms can affect the motor be¬

havior in patients with central nervous system
lesions.24-26
Spasticity, if defined as a velocity-dependent
response to stretch, can be a problem only if an
action is of sufficient velocity to evoke a stretch
reflex. Although this is certainly possible,27 it is
probably less frequent than seems to be suggested
by some authorities.28 Research points to excessive
coactivation as the more likely source of restraint.
Although exaggerated coactivation is well docu¬
mented in patients with central nervous system
lesions,29-31 it is not always present.32-34 It is more
likely to occur during concentric dynamic efforts,
particularly those that are of a high speed or
reciprocal nature.30-35 Overall, however, it appears
that failures of patients with neurologic disorders
to bring sufficient force to bear on the environ¬
ment can be attributed primarily to reduced ago¬
nist drive rather than to excessive antagonist re¬
straint.32- 33-36-37
NATURE OF MUSCLE FORCE PRODUCTION

FIGURE 7-1 Passive length tension curves derived from
the stretching of the gastrocnemius musculotendinous units Although an array of signs and symptoms can
of 14 patients with spastic hemiplegia. The left (open circles) accompany neurologic disorders, particularly
curve represents the plegic side. The right (closed circles) those of the brain, force production impairments
curve represents the nonplegic side. The tension associated are among the most prevalent. Particularly com¬
with a given length (e.g., 44.5 cm) is greater on the plegic mon are reductions in the maximum force that
side. (Adapted from Halar EM, Stolov WC, Venkatesh B, et
a person can bring to bear on the environment
al: Gastrocnemius muscle belly and tendon length in stroke
(strength). Associated with these reductions are
patients and able-bodied persons. Arch Phys Med Rehabil
59:476, 1978.)
decreases in the rate of force generation and speed
of movement. The ability to sustain adequate force
production over time (endurance) can also be re¬
duced.
generate the force necessary to move a mass
through a required distance.20'21
Impairments in Muscle Strength
The forces that muscles are able to bring to
bear on the environment depend on the combined Impairments in the force that patients with neuro¬
effects of passive and active components and rep¬ logic disorders are able to bring to bear on the
resent the resultant effects of agonist and antago¬ environment are apparent during both clinical test¬
nist muscles. The torques of any action, therefore, ing and functional activities. The results of studies
are the summed effect of mobilizing torques gen¬ describing these impairments in some of the more
erated by agonists minus restraining torques gen¬ common neurologic disorders follow.
erated by antagonists. Although the agonist activa¬ Stroke is among the most prevalent of neuro¬
tion associated with any muscle action tends to be
logic diagnoses accompanied by impairments in
accompanied by a concomitant activation of the muscle strength. During clinical testing, these im¬
antagonist, the latter is typically of minimal mag¬
pairments are most overt in the limb muscles
nitude.22’ 23 In patients with central nervous system
contralateral to the lesion, with the upper extrem¬
lesions, antagonist restraint is often increased. The ity of that side thought to be more affected than
restraint can be from the passive or active ele¬ the lower extremity.38-39 In the upper extremity of
ments. Restraint originating from active elements
the side contralateral to the lesion, the actions of
can be the consequence of spasms, spasticity, or
the hand are usually weaker than the actions that
excessive coactivation. Spasms, which are short-
are generated more proximally (e.g., elbow flexion
duration involuntary muscle contractions, can be
or shoulder abduction).40 A lesser but well-docu¬
provoked by a muscle stretch but can result from
mented weakness is often present in the limb
other stimuli as well. The involuntary contractions
muscles of the side ipsilateral to the lesion as
• I Chapter 7 BIOMECHANICS OF NEUROLOGIC TREATMENT 167
well. This weakness appears to be greater in more which impairments in muscle strength are associ¬
proximal limb actions (e.g., shoulder abduction or ated.5^53 The distribution of these impairments is
hip flexion).38 The muscles of the trunk are also a major factor in the classification of the disorder
affected by stroke. Impairments have been docu¬ (e.g., diplegia versus hemiplegia). There is no
mented in lateral and forward flexion and in rota¬ published evidence that the impairments tend to
tion.41' 42 Although the weakness after stroke tends ameliorate naturally over time.
to be greater in some actions than in others, there Spinal cord injuries represent another major di¬
is a degree of generalizability to the weakness.43 agnostic group with which impairments in muscle
Thus, a patient who is weaker than another patient strength are an expected occurrence, the magni¬
in one action is likely to be weaker in other tude of the impairments being dependent on the
actions as well. Strength impairments diminish neurologic level and completeness of the lesion.54
spontaneously with time, with the greatest im¬ Strength increases somewhat after the acute stage
provements occurring in the first several months of the injury, but the ultimate strength demon¬
after stroke44; the ultimate strength realized is strated by patients with spinal cord injury is de¬
strongly related to the strength retained very soon pendent on their strength soon after injury.55-57
after the stroke.39'45 Improvements in strength oc¬ Both multiple sclerosis58-60 and amyotrophic lat¬
cur bilaterally in the limbs39, 46 as well as in the eral sclerosis61 are characterized by muscle weak¬
trunk.47 Figure 7-2 illustrates the relationship be¬ ness that worsens with time. The deterioration in
tween the knee-extension strengths of the paretic strength in multiple sclerosis is variable, whereas
and nonparetic sides obtained on initial and final the deterioration accompanying amyotrophic lat¬
assessment. eral sclerosis tends to be precipitous and predict¬
Not as common as stroke but also depressing able.
limb muscle strength are other intracranial lesions, Of the several forms of muscle disease that
such as traumatic brain injury and intracranial result in muscle weakness, Duchenne’s muscular
tumors.48'49 Both, like stroke, have the potential to dystrophy has probably been studied the most.
impair the strength of limb muscles contralateral Unlike healthy boys whose strength increases as
and ipsilateral to the brain lesion, with this poten¬ they mature and gain height and weight, boys
tial documented in patients with traumatic brain with Duchenne’s muscular dystrophy tend to get
injuries.48 Although strength tends to increase over weaker as they age.62-64 Consequently, they dem¬
time after a traumatic brain injury, the strength of onstrate a growing gap in strength relative to age-
patients with intracranial tumors can increase or matched boys as they get older.
decrease depending on the nature of the tumor Acute cases of poliomyelitis are essentially non¬
and interventions directed at it. Cerebral palsy is existent in the developed world today. There are,
another neurologic disorder of the brain with nonetheless, numerous survivors of the polio epi¬
demics that occurred earlier in this century. Pa¬
tients who appeared to have regained most or
all of their premorbid strength are demonstrating,
decades later, emerging deficits in muscle perfor¬
mance.65, 66 Particularly susceptible to such deficits
(postpolio syndrome) are older patients whose
original poliomyelitis was more severe.66 Al¬
though postpolio syndrome is an unfortunate real¬
ity for many polio survivors, strength does not
necessarily decrease inexorably, at least not over
a period of several years.67 Decreases in strength
tend to be insidious.
Muscle weakness is a hallmark sign of Guillain-
Barre syndrome. Fortunately, most patients af¬
fected by this disease recover substantially or
completely. Most realize their maximum strength
impairment within 3 weeks. The extent of and
time required for recovery is predictable from
“muscle weakness at maximum.”68
Parkinson’s disease is but one neurologic disor¬
FIGURE 7-2 Scatterplot illustrating the relationship
der in which the primary impairment is not typi¬
between paretic (squares) and nonparetic (circles) knee
cally considered to be weakness. Nevertheless,
extension force measurements obtained on initial and final
patients with Parkinson’s disease frequently dem-
assessment from patients with stroke.
onstrate muscle weakness in the limbs and ance, however, depends in part on the manner in
trunk.69-73 That weakness is apparently related to which their endurance is characterized. Studies
some of the other more stereotypical motor signs in which a fatigue index (percentage decline in
(e.g., bradykinesia).72 maximum force or torque) has been used to de¬
scribe endurance are variable in their results.
Impairments in Rate of Force Generation and Some investigations have shown a greater percent¬
Speed of Movement age decline in neurologic patients than in healthy
subjects.83-85 Others have not shown greater de¬
Impairments in the rate of force production have clines.86-90 What is more certain is that patients
been documented in patients with stroke,74-75 Par¬ who are weak tend to decline to a required force
kinson’s disease,72-74-76 and multiple sclerosis.58 more quickly. Whether or not their rate of decline
Such impairments, however, are probably not lim¬ differs from that of healthy subjects, patients who
ited to patients with these diagnoses. The relation¬ are weak reach the required threshold of force
ship between rate of force generation and strength more quickly because they are closer to that
in patients with a variety of neurologic disorders threshold to start with (Fig. 7—3). Being weaker,
suggests that impairments in the two force vari¬ the patients have less reserve separating their ini¬
ables are reflecting a common underlying limita¬ tial maximum and threshold forces.91
tion.72- 74-75
Patients with neurologic disorders as diverse
IMPLICATIONS OF MUSCLE FORCE
as stroke,34-77-79 Parkinson’s disease,80-81 cerebral
PRODUCTION
palsy,50-52 and spinal cord injury82 have been
shown to demonstrate decreases in speed of move¬
Impairments in muscle force production are im¬
ment. Research results are inconsistent but provide
portant because of their implications for the per¬
some evidence that these speed impairments are
formance of functional activities. Historically,
related to impairments in muscle strength, at least
these implications were questioned by authorities
in patients with stroke.34-37 77 78 Thus, patients who
who suggested that muscle weakness is not a
have greater muscle weakness are likely to have
central cause of dysfunction in patients with cer¬
lower maximum movement velocities.
tain neurologic disorders.28 Nevertheless, the im¬
plications should be self-evident. Although most
Impairments in Muscle Endurance functional activities do not require a great deal
of strength, they do require force levels that are
Patients with weakness secondary to neurologic sufficient to accelerate or decelerate the mass of
disorders are often unable to sustain their perfor¬ the body and its segments or to respond appropri¬
mance at activities requiring muscle force produc¬ ately to external forces. In the absence of such
tion over an extended time. Whether these patients muscularly generated forces, the person becomes
are judged to be truly lacking in muscle endur¬
dependent on assistance from other people or
Figure 7-3 Illustration of the relationship between muscle strength and endurance.
,- i—auciigui diiu enaurance. Note
onfftb ronilt-r in ~ .. _J I
that greater strength results in a relative reduction in demand, which results in greater
absolute endurance. (Reprinted by permission from Sale DG: Testing strength and power In
MacDougaH p, Wenger HA, Green HJ [eds]. Physiological Testing of the High-Performance
Athlete, 2nd ed. Champaign, IL, Human Kinetics, 1991, pp 22-24.)
equipment. The importance of muscle force pro¬ 90

duction to function has been confirmed by studies
80
of statistical relationships involving clinical mea¬
surements and by biomechanical investigations. A _ 70
o
brief review of this research follows. CD
45 60
E
Studies of Statistical Relationships ■3 50
T3
O A
Numerous studies have revealed that muscle force cd 40
Q_
production, particularly when expressed as (/) __
strength, is related statistically to functional per¬ ~ 30
cC
formance. Such studies have employed an array ° 20
of muscle strength-measuring procedures and a
variety of functional measures, either multicompo¬ 10
nent or specific. Although research on relation¬
ships does not prove cause and effect, it does .0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.
provide evidence for the validity of strength mea¬ Paretic knee extension force
surements among patients with neurologic disor¬
FIGURE 7-4 Scatterplot illustrating the relationship of
ders. paretic knee extension force (expressed as a ratio of normal)
Multicomponent indexes characterizing the and comfortable gait speed (r = 0.74) in 35 patients with
functional limitations of patients with neurologic stroke.
disorders have been shown repeatedly to correlate
significantly with the patients’ muscle strength.
This relationship has been pointed out most often
able to generate have been linked to their perfor¬
among patients with stroke; their strength has been
mance at specific tasks. Chief among these tasks
found to correlate with the Barthel Index,92'98
are the sit-to-stand maneuver, standing, and walk¬
Frenchay Arm Test,44 and Sickness Impact Profile
ing.
scores.99 In one study involving the Barthel In¬
dex,92 the correlations between initial measures of
strength and final measures of function (r = 0.53 Sit to Stand
and 0.58, respectively) were particularly remark¬
Studies of the sit-to-stand maneuver have provided
able; they were only slightly less than the correla¬
much insight into how people go about making
tions between initial and final measures of func¬
a transition from one position to the other. The
tion (r = 0.64 and 0.65, respectively).92 The
transition involves a movement of the person’s
muscle strength of patients with spinal cord inju¬
center of mass in both the anterior and vertical
ries has also been noted to correlate with func¬
directions121 (Fig. 7-5). The movement, of course,
tional index scores, for example, with the Barthel
is driven by muscular forces. Research on patients
Index, Functional Independence Measure, and
with stroke reveals findings that might be expected
Quadriplegia Index of Function.57-I00-101
of people with lower extremity weakness that is
Patient performance at specific functional activ¬
greater on one side than the other. That is, the
ities can also be explained by measurements of
patients tend to take longer to attain standing than
muscle strength. This has been demonstrated most
do matched control subjects.121"123 When attaining
often in regard to various aspects of gait. Whether
standing in a natural fashion, the patients typically
patients are weak as a consequence of stroke102 108
bear more weight through their nonparetic lower
(Fig. 7—4), muscular dystrophy,62 myelomeningo¬
extremity.121"123 Although it is logical that they
cele,109 spinal cord injury,110"112 poliomyelitis,113 or
would favor a “better” extremity during spontane¬
cerebral palsy51 does not seem to matter; their
ous standing, they are able to increase their sym¬
weakness is predictive of walking performance.
metry with attention to it and training.124-125 The
Muscle weakness also explains limitations in pa¬
increase in symmetry, however, is not necessarily
tient performance at transfers,114- 115 bed mobil¬
retained upon the cessation of training.126
ity,116 dressing,115-117 stair climbing,118 and wheel¬
chair propulsion.115- U9-120
Standing
Biomechanical Investigations
Like studies of the sit-to-stand maneuver, most
Primarily through research involving forceplates studies of force during standing have been con¬
or weighing scales, the forces that patients are ducted on patients with stroke. These studies also
25
20
15
Dzquit
FIGURE 7—5 Movement of center of mass during

0 sit-to-stand by a healthy subject (solid line) and a
patient with hemiparesis (broken line). Tsoe, seat off;
Dxsoe, anterior displacement before seat off;
-5
Dxsoe, vertical displacement before seat off; Dxquit,
anterior displacement after seat off; Dzquit, vertical
-10 displacement after seat off; Tbegin, beginning of
rising; Tquit, end of rising. (From Hesse S, Schauer
10 20 30 [cm] M, Malezic M, et al: Quantitative analysis of rising
from a chair in healthy and hemiparetic subjects.
Anterior Scand J Rehabil Med 26:161, 1994.)
show a preference by patients for weight bearing Schroeder and colleagues140 concluded that “ab¬
through the nonparetic lower extremity during normal gait was due to difficulty in moving the
comfortable standing.127-132 This asymmetry does body over the unstable limb.” Pai and associates141
not appear to be fixed; the patients are able to found that in subjects they tested, 50% of the
adjust their weight bearing both laterally and in weight transfers to the nonparetic side were unsuc¬
an anteroposterior direction.102-133-136 Maximum cessful, and 80% of the weight transfers to the
weight bearing through or weight shifting toward paretic side were unsuccessful (from a failure ei¬
the paretic lower extremity, however, is also re¬ ther to hold or to shift far enough).
duced.102, 133-136
Other Aspects of Walking. Biomechanical re¬

Wa Iking search reveals far more about gait than its initia¬
tion and problems with weight shifting. The re¬
Weight Shifting. A person who is standing and search explains why patients with neurologic
wishes to initiate walking by taking a step must disorders walk slowly, and it also describes me¬
first shift weight away from the extremity that is chanical specifics of their ambulatory behavior.
to be lifted and advanced. Research involving
forceplates has shed light on this process. The
Determinants of Gait Speed. A biomechanical
research has shown that in healthy subjects, most
explanation of the decreased walking speed dem¬
of the lateral force (about 4% of body weight)
onstrated by patients with neurologic disorders
that shifts weight away from the uplifting lower
begins with factors determining walking speed
extremity and onto the stance lower extremity is
in people who are apparently healthy. Studies of
generated by the uplifting extremity before lift¬
healthy subjects demonstrate that muscle activity
off.137- 138 In patients with stroke, most of the lat¬
(as measured by electromyography) increases with
eral shifting force is generated by the uplifting
increased speed.142-143 Net measures of moments,
lower extremity only when that extremity is the
work, and power at the hip, knee, and ankle have
nonparetic lower extremity. Most of the lateral
also been shown to co-vary with gait speed. Win¬
shifting force preceding the lifting of the paretic
ter144 documented work and power bursts that de¬
lower extremity is generated by the nonparetic
creased progressively as gait cadence went from
lower extremity, which remains on the ground.
fast to natural to slow. Winter and White145 found
Thus, in patients with stroke, the nonparetic lower
a correlation of -0.94 between peak knee power
extremity is responsible for most of the lateral
and walking or jogging velocity. This led them
shifting force accompanying gait initiation regard¬
to conclude the existence of a cause-and-effect
less of which extremity is lifted first.139- 140 The
relationship of neurologic origin. They wrote that
ability of patients with stroke to shift weight effec¬
“power or moment profiles represent the final
tively is important but often compromised. Von
output of the neuromuscular system, and therefore
I Chapter 7 BIOMECHANICS OF NEUROLOGIC TREATMENT 171
any discussion must take into account the organi¬ zen in their gait, the amplitude and duration of the
zation of the CNS.” It should not be surprising, curves diminishes markedly as they cease shifting
therefore, that patients with pathologic71,102,104,110, their weight completely from one foot to the other.
U3, we, 147 or experimental148 lesions resulting in Given the decreased strength, reduced walking
muscle weakness walk slower than normal and speed, and relationship between force-associated
that the walking speed of patients with stroke variables and gait speed present in patients with
tends to correlate significantly with the moments, neurologic disorders, it should not be surprising
work, and power generated at the lower extremity that the patients demonstrate reduced joint mo¬
joints.149,150 ments, work, and power during gait itself. In pa¬
tients who have experienced stroke, the joint mo¬
Mechanical Specifics of Walking. In addition ments, work, and power are reduced on the
to gross deficits in speed, specific deviations or affected compared with the unaffected side. Spe¬
deficits related to force are present in the gait of cifically different are positive moments, work, and
patients with neurologic disorders. These devia¬ power at the ankle, negative work at the knee, and
tions have been examined most thoroughly in pa¬ both positive and negative work at the hip.149,150
tients with stroke.
Weight loading and ground reaction forces have
been the focus of numerous studies. Able-bodied Interventions for Improving
subjects tend to demonstrate periods of lower ex¬ Force Production
tremity loading (stance) that are symmetric be¬
tween sides. That is, the duration of left and right Force production can be increased by interven¬
stance are essentially equal. After neurologic dis¬ tions that have an immediate or nearly immediate
orders that result in primarily unilateral weakness effect or by training interventions that increase
(e.g., stroke), the stance duration of the unaffected force production over a prolonged course of ther¬
side is notably prolonged relative to the stance apy. Both categories of interventions are addressed
duration of the affected side and relative to the here. Interventions that might be considered physi¬
stance duration of able-bodied subjects.147,151 Dou¬ cal, however, are emphasized.
ble stance is also markedly prolonged.147 Such
behaviors might be expected to accompany a pa¬ INTERVENTIONS CAPABLE OF PRODUCING
tient’s reliance on the stronger of two lower ex¬ IMMEDIATE EFFECTS
tremities. In able-bodied people, the stance phase
of gait is associated with a double-peaked vertical Muscle Stretch
ground reaction force curve. The first peak of the
curve, which accompanies weight acceptance, is Chief among the techniques that can have a
distinct from the second peak, which accompanies prompt effect on muscle force are those involving
push-off.152, 153 The magnitude of the vertical muscle stretch. Three such techniques are men¬
ground reaction force, which exceeds body tioned here.
weight, increases with an increased speed of walk¬ The first technique involves merely putting a
ing. Among patients with stroke, the ground reac¬ muscle in an elongated position so that it can
tion pattern is often changed; neither the double generate more force. Although this behavior exists
peaks nor the interpeak trough is necessarily pres¬ in able-bodied subjects,156,157 it has particular im¬
ent.152, 154 The ground reaction force associated portance to patients with neurologic disorders.158
with weight acceptance is greater on the affected By adjusting the position of the joints that a mus¬
side than on the unaffected side, but the force cle crosses, the muscle can be lengthened and its
associated with push-off is significantly less on force increased. Although there are many circum¬
the affected than on the unaffected side. The dif¬ stances in which this helps neurologic patients to
ference between sides in the magnitude of the function, I will describe only three. First, a person
ground reaction force is less in patients with with a weak grasp after a spinal cord injury may
greater motor recovery after stroke. Morita and realize a useful augmentation of grip strength
coauthors154 reported a correlation of 0.731 be¬ when the long finger flexors are put on stretch by
tween the stage of motor recovery and a measure splinting the wrist in extension. Second, a person
of vertical ground reaction force symmetry. Pa¬ who has difficulty using a lower extremity to
tients with Parkinson’s disease also demonstrate propel a wheelchair after a stroke may gain a
abnormal vertical ground reaction force curves. beneficial supplementation of knee flexion
When patients are ambulating with a shuffling strength by leaning slightly forward rather than
gait, their force curves are characterized by nar¬ slouching back in the wheelchair seat. Doing so
row, single peaks. When the patients become fro¬ puts a greater stretch on the hamstrings, allowing
the patient to pull back more vigorously with the preceding hip extension (which accompanies lift¬
propelling lower extremity. Third, a patient who off) with forward motion of the trunk, the ham¬
is unable to lift a trailing lower extremity onto a strings are activated eccentrically before being ac¬
bed while the lower extremity already on the bed tivated concentrically to extend the hips.162
is fully extended (Fig. 7-6A) may be able to do A third technique involves a quick stretch of a
so if the contralateral hip and knee are first flexed muscle involved in an activity. Therapists some¬
(Fig. 7-65). This maneuver tilts the pelvis posteri¬ times provide such stretches manually in an effort
orly159 and flattens the back. As a consequence, to facilitate a desired movement.163 The stretch
the hip flexors are lengthened and are enabled to can be applied by a quick pull on a segment that
create more force. is counter to its intended direction of movement
A second technique involving stretch calls on or by one or more taps on the tendon or belly
the stretch-shortening cycle. When a muscle un¬ of a muscle involved in a movement. That such
dergoes active shortening (a concentric contrac¬ “facilitation” techniques influence movement in
tion) immediately after it has undergone active patients with central nervous system lesions is
lengthening (an eccentric contraction), the concen¬ easily verified. Tapping may even have a small
tric force is greater than if not preceded by the short-duration residual effect on a muscle’s force-
eccentric action.160 The benefits of the stretch¬ producing capability.164 The problem is that stretch
shortening cycle are readily apparent in able-bod¬ facilitation involves an external input by another
ied subjects, for example, during jumping. In pa¬ person. Patients themselves can employ muscle-
tients who are weak, however, the use of the elongating postures and the stretch-shortening
stretch-shortening cycle is probably more cru¬ cycle.
cial.161 One important use of the stretch-shortening
cycle is during the sit-to-stand maneuver. A pa¬
tient who has difficulty attaining standing from Other Interventions
sitting when the task is performed very slowly
may be more independent when the trunk is flexed Several other interventions can affect muscle force
forward quickly before rising is attempted. By while they are being applied. Of particular note
FIGURE 7-6 Line drawings from

photographs of a patient with
hemiparesis lifting his paretic lower
extremity onto a bed. He was unable
when the contralateral hip was
extended (A) but able when the
contralateral hip was flexed (B).
are electrical stimulation, vibration, and cutaneous to cutaneous stimulation provided by application
stimulation. of orthokinetic cuffs to the patient’s paretic upper
Functional electrical stimulation can markedly extremity.182 Orthokinetic cuffs have also been ap¬
augment the force generated by the muscles of plied to patients with cerebral palsy and poliomy¬
patients with central nervous system lesions. The elitis with positive results.183,184
augmentation has been demonstrated during nu¬
merous activities, both in patients with stroke and INTERVENTIONS PRODUCING LASTING
in those with spinal cord injury. Among patients IMPROVEMENTS IN MUSCLE FORCE
with stroke, tetanizing current has been used pri¬ PRODUCTION
marily to augment muscle force during gait. The
ankle dorsiflexors are probably the muscle group Interventions aimed at increasing the forces that
most often stimulated during gait. Their stimula¬ muscles can bring to bear on the environment
tion can increase foot clearance during swing, typically focus on the amelioration of strength
counteract excessive inversion during swing and impairments through voluntary exercise. Other in¬
stance, increase gait speed, and decrease the en¬ terventions target antagonist restraint. These thera¬
ergy demand of walking.165-167 By activating nu¬ pies can be applied alone or in combination and
merous different muscle groups in a coordinated can be supplemented with other treatments.
sequence, multichannel stimulation can affect nu¬
merous aspects of gait favorably (e.g., inadequate Voluntary Exercise for Increasing Muscle Force
hip and knee flexion during swing, excessive knee Production
flexion during stance, and inadequate push-off).168
Electrical stimulation has also been used to im¬ It is now beyond question that the muscle force-
prove muscle function in the upper extremity of producing capacity of able-bodied subjects can
patients with stroke. Stimulation of the supraspi- be increased when they participate in exercise
natus and posterior deltoid can be used to reduce regimens of adequate intensity, frequency, and du¬
shoulder subluxation.169 Merletti and colleagues170 ration. That voluntary exercise programs can pro¬
found that multichannel stimulation of the upper mote improved motor performance in patients
extremity allowed performance of tasks “that with a variety of neurologic conditions is also
were impossible without stimulation.” Among pa¬ supported by considerable evidence.
tients with spinal cord injury, electrical stimulation Historically, authorities on the rehabilitation of
has been used most often to elicit contractions of patients with brain lesions did not recommend the
the quadriceps femoris. Stimulation of the quadri¬ training of strength among such patients.28 The
ceps has been used to assist patients with spinal results of research, however, suggest that the pa¬
cord injury to stand,171-173 walk,174'175 and cycle.176 tients can benefit from exercise programs of suf¬
Other muscles, however, have been stimulated in ficient frequency, intensity, and duration. Numer¬
conjunction with or in lieu of the quadriceps femo¬ ous positive outcomes have been described among
ris to help the patients to cycle and walk.175-177 patients with stroke. More than 20 years ago,
Electrical stimulation has also been employed to Inaba and colleagues185 found that patients who
activate the upper extremity muscles of patients participated in progressive resistance exercise (iso¬
with tetraplegia.178 tonic leg presses) increased significantly more in
Vibration, much like tapping, can result in the strength than patients undergoing an active exer¬
potentiation of voluntary force production by mus¬ cise regimen or training in activities of daily liv¬
cles that are paretic after central nervous system ing. The resistance-trained patients were also sig¬
lesions.179 Although there is some short-term nificantly more likely after 1 month to improve in
carry-over of vibration,179 patients who are treated activities of daily living than patients in active
over time with vibration do not demonstrate exercise or activities of daily living training
strength increases beyond those realized by pa¬ groups. More recently, repetitive resisted flexion
tients not treated with vibration.180 and extension of the fingers has been reported by
Cutaneous stimulation applied by at least two Biitefisch and coauthors186 to lead to increased
techniques has been shown to facilitate muscular strength, velocity, and acceleration in the hand.
activity in patients with hemiparesis. Cutaneous Sharp and Brouwer187 found that isokinetic train¬
stimulation through rapid brushing has been re¬ ing of the hemiparetic knee resulted in improve¬
ported by Matyas and associates181 to increase ments in both knee muscle strength and gait veloc¬
isometric knee extension and flexion forces but ity.
not isometric dorsiflexion force. Neeman and Nee- Engardt and coworkers188 also demonstrated
man182 have attributed increases in the active el¬ that isokinetic training of the knee muscles is
bow flexion range of a patient with hemiparesis beneficial. Whether trained concentrically or ec-
centrically, their subjects who had experienced a conditions, the training should probably be closed
stroke (on average) more than 2 years previously chain.204'205 Similarly, if the goal is to increase
improved in paretic knee extension strength. Only muscle performance under ballistic conditions, the
the eccentric training group, however, increased training should at least be oriented b al 1 i s tic ally.206
in both concentric and eccentric strength on the
paretic side compared with the nonparetic side. Interventions Targeting Antagonist Restraint
Only the eccentrically trained group realized a
“nearly symmetrical body weight distribution” Because antagonist restraint can contribute to de¬
between lower extremities when achieving stand¬ creases in the force of muscle actions, interven¬
ing from sitting. Electromyographic studies of pa¬ tions aimed at reducing the restraint are sometimes
tients with spastic paresis offer an explanation for employed in an effort to improve the motor perfor¬
the apparent superiority of eccentric training.189 mance of patients with neurologic disorders. Posi¬
Compared with concentric efforts, eccentric ef¬ tive effects have been demonstrated for such inter¬
forts are accompanied by a higher level of agonist ventions.
activation and a lower level of antagonist activa¬ One way that antagonist restraint can be re¬
tion. Knutsson and associates189 concluded that duced is by the lengthening of the antagonist. This
“the stretch imposed on antagonists in concentric is often accomplished surgically, particularly in
contractions dampens voluntary activation of ago¬ children with cerebral palsy.3-207’208 The effects of
nists through reciprocal inhibition” and that such lengthening on voluntary muscle force have
“stretch imposed upon agonists in eccentric con¬ been quantified by Reimers.209'210 He showed that
tractions facilitates voluntary activation.” ankle dorsiflexion strength increased more than
Research supportive of the use of resistance 200% in 52 patients with cerebral palsy who un¬
training is not limited to adults with brain lesions. derwent surgical elongation of the triceps surae a
Patients with cerebral palsy have also been shown median of 14 months before and that knee exten¬
to respond favorably to resistance exercise.51’19(M94 sion strength increased by a median 22% in 38
Investigations of their response demonstrate in¬ patients with cerebral palsy who underwent surgi¬
creases in strength and, in separate studies, im¬ cal elongation of the hamstring muscles. Antago¬
proved gait193 and wheelchair propulsion.194 nists can also be lengthened by casting and
Strength training has also been established to stretching. Casting of the ankle to stretch the plan¬
be effective for patients with neurologic disorders tar-flexors markedly increases passive dorsiflexion
or problems outside the brain. Among such disor¬ and reduces resistance to dorsiflexion in patients
ders are spinal cord injury, neuromuscular disease, with brain lesions.6, 211> 212 Prolonged stretching of
and poliomyelitis. Resistance exercises have been muscles can lead to notable increases in passive
shown to be effective for increasing the strength joint range of motion.213-215 Stretching of the hip
of patients with spinal cord lesions194"195 and for adductor muscles has also been reported to result
improving their wheelchair propulsion perfor¬ in increased voluntary hip abduction in adults with
mance.194-196 Isotonic exercise has been found to spastic paraparesis.215 Odeen reported an average
augment the strength of multiple actions in pa¬ increase of 255% in voluntary abduction after a
tients with neuromuscular disorders.197 Postpolio¬ mean of about 4 months of 30-minute stretches
myelitis patients also appear also to respond favor¬ performed two to five times daily.215
ably and without adverse effects to nonfatiguing Other ways of reducing antagonist restraint, at
resistance training.198"200 Even patients with myosi¬ least in cases in which the restraint is due to
tis may benefit from progressive resistance exer¬ spasticity or excessive coactivation, are through
cise.201
the administration of pharmacologic agents and
Regardless of the neurologic disorder, the issue nerve sections. Pharmacologic agents can be ad¬
of action specificity should be taken into account ministered systemically or through injections into
when enrolling patients in a strengthening regi¬ the antagonists or the nerves supplying them. The
men. Not all actions entail the same degree of latter include injections of phenol or alcohol, and
muscle activation, and training a muscle in one more recently botulinum toxin. In addition to re¬
action may not optimize its improvement in an¬ ducing tone, these injections have been shown
other. Electromyography can clarify the degree repeatedly to improve passive range of motion
to which muscles are activated during specific and in some cases to increase active range of
activities.202"203 Although the research related to motion.216-223 Keenan,216 for example, reported that
specificity of training has typically involved patients increased an average 25 degrees in wrist
healthy subjects, it points to the importance of the extension after the motor points of their spastic
goal of training. If, for example, the goal is to
wrist flexors were injected with phenol.216 Pierson
increase muscle performance under closed-chain and colleagues218 described a mean gain in active
range of motion of 17 degrees in an array of documented that isometric knee extension torque
different upper and lower extremity joints affected increased 479% in a combination treatment group
by spasticity. but 166% in a control group. Bowman and coau¬
The results of research addressing the functional thors232 described significant increases in both
consequences of pharmacologic injections are wrist extension torque and selective range of mo¬
equivocal.216-223 The reduction of antagonist re¬ tion. The torque increased 280% when measured
straint has been accomplished through two types with the wrist at 30 degrees of extension.
of nerve sections: selective peripheral neurotomy Evidence supporting the application of electri¬
and selective posterior rhizotomy. An example of cal stimulation in cases of cerebral palsy is also
the first is the partial section of the mixed tibial available but limited. A few studies have provided
nerve that Feve and associates224 described. That support for very-low-intensity “therapeutic elec¬
section resulted in a reduction in ankle plantar- trical stimulation.”233,234
flexor muscle tone and a significant (mean, 7.5 Considerable research has been conducted on
degrees) increase in active ankle dorsiflexion the response of spinal cord injury patients to the
range of motion.225, 226 All patients’ gaits im¬ longitudinal application of functional electrical
proved. The second involved the partial sectioning stimulation. Much of this research, like that of
of posterior (sensory) roots. Steinbok and coau¬ Barr and coworkers235 and Rabischong and
thors225 compared patients with spastic cerebral Ohanna,236 has specified how functional electrical
palsy who underwent the procedure and physical stimulation can increase evoked muscular output
therapy with a group that received only physical in patients with intact peripheral innervation and
therapy. The patients whose treatment included motor complete paraplegia. Functional electrical
rhizotomy showed greater reductions in spasticity stimulation, however, has also been shown to in¬
and greater increases in passive range of motion. crease the voluntary strength of patients with mo¬
Muscle strength did not improve differentially be¬ tor incomplete cord injuries.178
tween treatments, but the Gross Motor Function Although most of the research supporting the
Measure did; at 9 months, it had increased 11.3% use of electrical stimulation has focused on pa¬
in the patients receiving the rhizotomy but only tients with lesions of the brain and spinal cord, its
5.2% in the patients receiving physical therapy efficacy with other types of neurologic patients
alone. Thomas and colleagues226 reported that the has been demonstrated. One particularly interest¬
gait of children with spastic diplegia was im¬ ing application has been to cases of Bell’s palsy.
proved a year after dorsal rhizotomy. By using eutrophic electrical stimulation, Far-
ragher and colleagues237 reported remarkable re¬
Supplements to Voluntary Exercise Interventions sponses in patients with “apparently intractable”
Bell’s palsy. They employed a facial paralysis
Although they can be used in isolation, electrical recovery index to document improved muscle per¬
stimulation, acupuncture, and biofeedback are typ¬ formance in the patients.
ically used as supplements to voluntary exercise There is considerable support for the use of
interventions. As such, they can be used individu¬ acupuncture to treat patients with neurologic dis¬
ally or in combination. orders, particularly stroke. Naeser and associ¬
Electrical stimulation, which can be employed ates238 showed that patients with stroke receiving
for its immediate effects, can also be employed acupuncture treatment for hand paresis experi¬
over an extended period to improve muscle perfor¬ enced a “good response defined as improvement
mance. It has been found in numerous studies to in at least four of six hand tests after 20 or 40
be efficacious when so employed with patients acupuncture treatments.” They observed signifi¬
who were paretic after stroke. Large strength in¬ cant improvements in the patients’ grip and pinch
creases have been reported by Levin and Hui strength and timed hand dexterity. The positive
Chan (ankle dorsiflexion force increased up to results of acupuncture appear to be limited by the
820%)227 and Merletti and colleagues (ankle dorsi¬ extent of brain damage accompanying the
flexion moment increased about 300%).228 Sub¬ stroke.238,239 The benefits, however, extend to pos¬
stantial increases in voluntary movement ampli¬ tural control and function239-242 (Fig. 7-7).
tude have been noted by Smith (multiple limb Researchers performing meta-analyses of the
joint movements increased 67% to 480%)229 and effectiveness of electromyographic biofeedback
Lagasse and Roy (elbow extension range in¬ for patients with stroke, those to whom the proce¬
creased 28 to 105 degrees).230 Electrical stimula¬ dure is probably most often applied, have come to
tion in combination with positional feedback has differing conclusions. Schleenbaker and Mai-
been shown to increase muscle performance at nous243 suggested that such feedback “is an effec¬
the knee and wrist. Winchester and associates231 tive tool for neuromuscular reeducation in the
cle weakness, patients with neurologic disorders

may have to employ techniques that differ from
those employed by people with normal strength.
The use of interventions that rely on muscle
strength have been mentioned previously. Beyond
these alternatives, patients can use methods that
rely on the muscle strength that they retain and
maneuvers that optimize the effects of leverage.
Numerous such methods exist. Three, however,
provide particularly useful examples of the effec¬
tiveness of altered technique. The first involves
rolling in patients with stroke. Patients with stroke
have difficulty rolling toward their nonplegic side.
Bohannon115 showed that a significant number of
patients with stroke increased in their indepen¬
Month dence at this task after being taught to use lever¬
FIGURE 7-7 Line graph illustrating the greater recovery of age (i.e., specific placement of their extremities)
function (Barthel scores) following stroke in a group of to their advantage. The second example of using
patients treated with acupuncture compared with a control altered technique can be found in patients with
group. (Graph generated from data reported by Johansson tetraplegia who lack sufficient triceps brachii
and colleagues.242)
strength to extend the elbow against gravity and
can employ the anterior deltoid and upper pectoral
hemiplegic stroke patient.” Glanz and cowork- muscles. These muscles can be used to lock the
ers,244 on the other hand, concluded that “results elbow while the hand is fixed and enable trans¬
of pooling available randomized control trials do fers.246, 247 Third, patients with muscle disease and
not support the efficacy of biofeedback in restor¬ other neurologic disorders resulting in lower ex¬
ing the range of motion of hemiparetic joints.” tremity weakness sometimes rely on the upper
Since the publication of these two analyses, Intiso extremities and altered sequences of lower extrem¬
and colleagues245 completed a study that led them ity and trunk maneuvers to arise from sitting.248
to conclude that “electromyographic biofeedback One such maneuver, seen often in patients with
technique increases muscle strength and improves Duchenne’s muscular dystrophy, is Gowers’ ma¬
recovery of functional locomotion in patients with neuver, in which patients push with their upper
hemiparesis and foot-drop after cerebral ische¬ extremities on their thighs (Fig. 7-8).
mia.” Although their sample size was small, their
findings are compelling. If electromyographic bio¬ MODIFICATION OF DEMANDS
feedback is to be applied, research suggests that it
can be employed to augment agonist activation Because force equals mass times acceleration and
without concern for reducing the activation of the the force-producing capacity of the muscles of
antagonist.36,37 patients with neurologic disorders is frequently
reduced, the demand for force from the muscles
must often be reduced if the patients are to manage
Compensations for Reduced functionally. One way of reducing the demand is
Force Production to reduce the body’s mass,249,250 which has a ten¬
Whether or not the force production of neurologic dency to increase in some patients with neurologic
patients can be increased immediately or over disorders. Another way of reducing the demand
time, compensations or adjustments can be em¬ for force is to provide a means of reducing the
ployed to their advantage. These compensations force requirements on weak muscles. Assistive
take three major forms: altered techniques, the devices, discussed in the next section, are one
modification of demands, and the provision of way of reducing force requirements. In addition,
assistive devices. Such compensations have been assistance from other people and equipment serves
mentioned in passing earlier; they are elaborated this purpose. Clearly, there are many patients who
on here. can manage functionally if assistance is provided
to move a body segment (e.g., the lower extremity
ALTERED TECHNIQUES off the bed when going from supine to sitting) or
to supplement their efforts against gravity (e.g.,
There are many alternative techniques for accom¬
during sit to stand or the unilateral stance phase of
plishing functional tasks. In the presence of mus¬
gait). An example of demand-reducing equipment
FIGURE 7-8 Line drawing of sit-

to-stand technique in which the
patient uses the upper extremities
on her thighs to achieve standing.
(From Butler PB, Nene AV, Major
RE: Biomechanics of transfer from
sitting to the standing position in
some neuromuscular diseases.
Physiotherapy 77:521, 1991.)
employed lately with good success, both during neurologic pathology. Even when patients possess
application of the equipment and afterward, is a enough strength to walk with an assistive device
weight-unloading harness during gait on a tread¬ and orthoses, a wheelchair may be safer or more
mill.251-255 The buoyancy of water can also supple¬ efficient.265- 266 Patients with severely impaired
ment patients’ antigravity efforts.256’257 During sit lower extremity and trunk strength but adequate
to stand, the demand on the lower extremities can upper extremity strength can use their upper ex¬
be reduced considerably by elevating the height tremities to propel a wheelchair.267'268 Patients with
of the seat from which a patient is rising.258-261 insufficient extremity strength on one side (e.g.,
Arborelius and associates261 found that the mean after stroke) can use the extremities of their
maximum knee extension moment was 60% lower stronger side to propel the chair, although for
when rising from a higher rather than a standard- some hemiparetic patients this is too difficult or
height seat. Weiner and colleagues260 reported that impossible.269’270 For patients with weakness of all
when the chair height increased from 17 to 22 four extremities and the trunk, a powered chair
inches, the number of nursing home residents in may be best, even if they can manage a standard
their study who were able to rise successfully chair.271,272
nearly doubled. The demand on the lower extremi¬
ties during sit to stand can also be reduced through Gait-Assistive Devices
the use of arm rests.261-262 Pneumatic or spring-
loaded seating surfaces, although not used exten¬ Gait-assistive devices (i.e., walkers, crutches, and
sively, are capable of reducing the load on the canes) are the most frequently prescribed of pa¬
lower extremities during sit to stand.263'264 tient aides.273 They can assist patients with neuro¬
logic disorders in two ways: they can increase
stability by expanding a patient’s base of support,
PROVISION OF ASSISTIVE DEVICES and they can enable a patient to reduce the load
placed through a weak lower extremity during
A variety of assistive devices are available to stance.
increase the independence or performance level of That walkers, crutches, and canes can increase
patients during functional activities. Entire books a patient’s stability should be self-evident. Need¬
have been written on this subject, which has been less to say, gait-assistive devices can increase sta¬
touched on elsewhere in this chapter and in Chap¬ bility only if they are on the ground and the
ter 10. This section addresses in a cursory fashion forces applied through them by a patient’s upper
only three categories of devices that can compen¬ extremity or extremities are properly directed. Pa¬
sate for the reduction in muscle force production tients with considerable postural instability may
that so often accompanies neurologic pathology. benefit from a device that does not have to be
The categories are wheelchairs, gait-assistive de¬ lifted to be advanced, that is, a wheeled walker.
vices, and lower extremity orthoses. For patients lacking sufficient upper extremity
strength to handle a walker (e.g., those with hemi-
paresis), a platform can sometimes be attached for
Wheelchairs
the weaker upper extremity (Fig. 7-9). Provision
Wheelchairs are often prescribed for patients with of such a walker for a patient of mine with ataxia
extremity and trunk muscle weakness secondary to and hemiparesis of 30 years’ duration resulted
for unloading is not too great, crutches or canes

can be used.276,277 When crutches or canes are used
unilaterally, better unloading is achieved when the
device is placed on the side contralateral to the
weak lower extremity.276 Of course, in many cases,
it is not possible to do otherwise. After a stroke,
for example, many patients would lack sufficient
strength in the upper extremity of the weak side
to employ a cane or crutch. The degree to which
an assistive device can be used to unload a weak
lower extremity can be affected by the gait pattern
with which it is used.276
Lower Extremity Orthoses
Clearly, orthoses are a practical way of compen¬

sating for weakness in patients with neurologic
disorders. Although not limited in application to
the lower extremities, it is there that they are
employed most often. Orthoses are used either to
stabilize structures with inadequate muscle sup¬
port or to enable patients to complete or improve
patterns of movement with which they are having
difficulty. A multicomponent orthosis may accom¬
plish both and can be supplemented with interven¬
tions already addressed, such as electrical stimula¬
tion.
An obvious example of stabilization is provided
by long leg braces, which prevent the knees from
Figure 7-9 Line drawing of a patient with hemiparesis giving way while a patient with lower extremity
using a rolling platform walker.
weakness stands or walks.266’278-280 An ankle that
is unstable due to muscle weakness or imbalance
can be stabilized by an orthosis.281,282 In patients
in her ambulating more than 200 meters without with instability or inadequate control of both the
assistance. She was unable to walk even a meter knee and ankle after stroke, an orthosis with both
without assistance when using a hemiwalker or knee and ankle components can be used to reduce
quad cane. Crutches and canes can also increase genu recurvatum.283
stability.274’275 One research study, however, was The benefits of orthoses for the completion or
able to verify a stabilizing effect only while pa¬ improvement of movement patterns is probably
tients stood with canes; increased stability could most apparent in gait. During the swing phase
not be verified during walking.274 In their investi¬ of normal gait, the non-weight-bearing extrem¬
gation of patients with stroke, Milczarek and col¬ ity is functionally shortened and advanced by
leagues275 found that canes resulted not only in active hip flexion, knee flexion, and ankle dorsi-
reduced mediolateral and anteroposterior sway but flexion. One of the most common applications
also in a shift in the center of pressure toward the of orthoses is to assist ankle dorsiflexion during
cane side. Contrary to what might be expected, a swing among patients with stroke. Such assis¬
four-footed cane was no more stabilizing than a tance improves the mechanics of swing and ini¬
single joint cane.
tial stance as well as cadence, step length, gait
Some patients with neurologic disorders are not speed, ambulatory distance, and oxygen con¬
able to bear full weight through a weak lower sumption during gait.281-282-284-289
extremity; it lacks sufficient extensor strength to
support the weight.102’136 Consequently, they can¬
not advance their contralateral lower extremity. Summary
One possible solution to this problem is for the
patients to use the upper extremities and an as¬
Although biomechanics is relevant to the treat¬
sistive device to unload the weak extremity. Walk¬
ment of patients with a variety of pathologies and
ers can be used for this purpose, and if the demand
disorders, the muscle weakness that is so prevalent
among patients with neurologic problems makes CF (eds): Clinical Electromyography, 2nd ed.
biomechanics an issue of particular importance in Boston, Butterworth-Heinemann, 1993, p 95.
14. McComas AJ, Sica REP, Campbell MJ, Upton
their treatment. This chapter has provided a cur¬
ARM: Functional compensation in partially
sory review of relevant background information
denervated muscles. J Neurol Neurosurg
as well as a basis for specific interventions to
Psychiatry 34:453, 1971.
correct or compensate for the weakness that af¬ 15. Tollback A, Borg J, Borg K, Knutsson E:
fects the function of patients with neurologic dis¬ Isokinetic strength, macro EMG and muscle
orders. biopsy of paretic foot dorsiflexors in chronic
neurogenic paresis. Scand J Rehabil Med 25:183,
1993.
References 16. Rodriguez AA, Agre JL, Harmon RL, et al:
Electromyographic and neuromuscular variables
1. Gajdosik RL, Guiliani CA, Bohannon RW: in post-polio subjects. Arch Phys Med Rehabil
Passive compliance and length of the hamstring 76:989, 1995.
muscle of healthy men and women. Clin 17. Stalberg E, Grimby G: Dynamic
Biomech 5:23, 1990. electromyography and muscle biopsy changes in
2. Halar EM, Stolov WC, Venkatesh B, et al: a 4-year follow-up: Study of patients with a
Gastrocnemius muscle belly and tendon length in history of polio. Muscle Nerve 18:699, 1995.
stroke patients and able-bodies persons. Arch 18. Miller RG, Sherratt M: Firing rates of human
Phys Med Rehabil 59:476, 1978. motor units in partially denervated muscle.
3. Tardieu G, Tardieu C: Cerebral palsy: Neurology 28:1241, 1978.
Mechanical evaluation and conservative 19. Reiners K, Herdmann J, Freund H-J: Altered
correction of limb joint contractures. Clin Orthop mechanisms of muscular force generation in
219:63, 1987. lower motor neuron disease. Muscle Nerve
4. Tabary JC, Tabary C, Tardieu C, et al: 12:647, 1989.
Physiological and structural changes in the cat’s 20. Angel RW: Electromyographic patterns during
soieus muscle due to immobilization at different ballistic movement of normal and spastic limbs.
lengths by plaster casts. J Physiol 224:231, 1972. Brain Res 99:387, 1975.
5. Goldspink G, Tabary C, Tabary JC, et al: Effect 21. Hallett M, Khoshbin S: A physiological
of denervation on the adaptation of sarcomere mechanism of bradykinesia. Brain 103:301, 1990.
number and muscle extensibility to the functional 22. Solomonow M, Guzzi A, Baratta R, et al: EMG
length of the muscle. J Physiol 236:733, 1974. force model of the elbows antagonistic muscle
6. Tardieu G, Tardieu C, Colbeau-Justin P, pair. Am J Phys Med 65:223, 1986.
Lespargot A: Muscle hypoextensibility in 23. Amiridis IG, Martin A, Morion B, et al: Co¬
children with cerebral palsy. II. Therapeutic activation and tension-regulating phenomena
implications. Arch Phys Med Rehabil 63:103, during isokinetic knee extension in sedentary and
1982. highly skilled humans. Eur J Appl Physiol
7. Dantes M, McComas A: The extent and time 73:149, 1996.
course of motoneuron involvement in 24. Little JW, Micklesen P, Umlauf R, Britell C:
amyotrophic lateral sclerosis. Muscle Nerve Lower extremity manifestations of spasticity in
14:416, 1991. chronic spinal cord injury. Am J Phys Med
8. Bromberg MB, Forshew DA, Nau KL, et al: Rehabil 68:32, 1989.
Motor unit number estimation, isometric strength, 25. Kawamura J, Ise M, Tagami M: The clinical
and electromyographic measures in amyotrophic features of spasms in patients with a cervical
lateral sclerosis. Muscle Nerve 16:1213, 1993. cord injury. Paraplegia 27:222, 1989.
9. Lenman JAR: Quantitative electromyographic 26. Fowler HL, Warmoth J, Nabizadeh A, Schein-
changes associated with muscular weakness. J berg L: Rapid conversion from extensor to fixed
Neurol Neurosurg Psychiatry 22:306, 1959. flexor posture of the lower extremities in
10. McComas AJ, Sica REP, Upton ARM, Aguilera multiple sclerosis. J Neurol Rehabil 2:159, 1988.
N: Functional changes in motoneurons of 27. Corcos DM, Gottlieb GL, Penn RD, et al:
hemiparetic muscles. J Neurol Neurosurg Movement deficits caused by hyperexcitable
Psychiatry 36:183, 1973. stretch reflexes in spastic humans. Brain
11. Edstrom L, Grimby L, Hannerz J: Correlation 109:1043, 1986.
between recruitment order of motor units and 28. Bobath B: Adult Hemiplegia: Evaluation and
muscle atrophy pattern in upper motor neuron Treatment, 3rd ed. Oxford, Heinemann Medical
lesion significance of spasticity. Experentia Books, 1997.
29. Hammond MC, Fitts SS, Kraft GH, et al: Co¬
29:560, 1973.
12. Jakobsson F, Grimby L, Edstrom L: Motoneuron contraction in the hemiparetic forearm
activity and muscle fibre type composition in quantitative EMG evaluation. Arch Phys Med
hemiparesis. Scand J Rehabil Med 24: 115, 1992. Rehabil 69:348, 1988.
13. Brown WF: Negative symptoms and signs of 30. Knutsson E, Maartensson A: Dynamic motor
capacity in spastic paresis and its relation to
peripheral nerve disease. In Brown WF, Bolton
prime mover dysfunction, spastic reflexes and hemiparetic stroke patients: A follow up study.
antagonist co-activation. Scand J Rehabil Med Tohoku J Exp Med 154:111, 1988.
12:93, 1980. 47. Bohannon RW: Recovery and correlates of trunk
31. Levin MF, Hui-Chan C: Ankle spasticity is muscle strength after stroke. Int J Rehabil
inversely correlated with antagonist voluntary Research 18:162, 1995.
contraction in hemiparetic subjects. 48. Smutok MA, Grafman J, Salazar AM, et al:
Electroenceph Clin Neurophysiol 34:415, 1994. Effects of unilateral brain damage on
32. Tang A, Rymer WZ: Abnormal force-EMG contralateral and ipsilateral upper extremity
relations in paretic limbs of hemiparetic human function in hemiplegia. Phys Ther 69:195, 1989.
subjects. J Neurol Neurosurg Psychiatry 44:690 49. Walker GC, Cardenas DD, Guthrie MR, et al:
1981. Fatigue and depression in brain-injured, patients
33. Fellows SJ, Kaus C, Ross HF, Thilmann AF: correlated with quadriceps strength and
Agonist and antagonist EMG activation during endurance. Arch Phys Med Rehabil 72:467,
isometric torque development at the elbow in 1991.
spastic hemiparesis. Electroenceph Clin 50. Brown JK, van Rensburg F, Walsh G, et al: A
Neurophysiol 93:106, 1994. neurological study of hand function of
34. Fellows SJ, Kaus C, Thilmann AF: Voluntary hemiplegic children. Dev Med Child Neurol
movement at the elbow in spastic hemiparesis. 29:287, 1987.
Ann Neurol 36:397, 1994. 51. Damiano DL, Vaughan CL, Abel MF: Muscle
35. Duncan P: The effect of a prior quadriceps response to heavy resistance exercise in children
contraction on knee flexor torque in normal with spastic cerebral palsy. Dev Med Child
subjects and multiple sclerosis patients with Neurol 37:731, 1995.
spastic paraparesis. Physiother Pract 3:11, 1987. 52. Brown JK, Rodda J, Walsh EG, Wright GW:
36. Gowland C, de Brain H, Basmajian JV, et al: Neurophysiology of lower-limb function in
Agonist and antagonist activity during voluntary hemiplegic children. Dev Med Child Neurol
upper-limb movement in patients with stroke. 33:1037, 1991.
Phys Ther 72:624, 1992. 53. Van den Berg-Emons RJG, van Baak MA, de
37. Wolf SL, Catlin PA, Blanton S, et al: Barbanson DC, et al: Reliability of tests to
Overcoming limitations in elbow movement in determine peak aerobic power, anaerobic power
the presence of antagonist hyperactivity. Phys and isokinetic muscle strength in children with
Ther 74:826, 1994. spastic cerebral palsy. Dev Med Child Neurol
38. Bohannon RW, Andrews AW: Limb muscle 38:1117, 1996.
strength is impaired bilaterally after stroke. J 54. Maynard FM, Bracken MB, Creasey G, et al:
Phys Ther Sci 7:1, 1995. International standards for neurological and
39. Bohannon RW: Muscle strength changes in functional classification of spinal cord injury.
hemiparetic stroke patients during inpatient Spinal Cord 35:226, 1997.
rehabilitation. J Neurol Rehabil 2:163, 1988. 55. Schwartz S, Cohen ME, Herbison GJ, Shah A:
40. Colebatch JG, Gandevia SC: Distribution of Relationship between two measures of upper
muscular weakness in upper motor neuron extremity strength: Manual muscle test compared
lesions affecting the arm. Brain 112:749, 1989. to hand-held myometry. Arch Phys Med Rehabil
41. Bohannon RW, Cassidy D, Walsh S: Trunk 73:1063, 1992.
muscle strength is impaired multidirectionally 56. Herbison GJ, Isaac Z, Cohen ME, Ditunno JF:
after stroke. Clin Rehabil 9:47, 1995. Strength post-spinal cord injury: Myometer vs
42. Tanaka S, Hachisuka K, Ogata H: Trunk rotatory manual muscle test. Spinal Cord 34:543, 1996.
muscle performance in post-stroke hemiplegic 57. Saboe LA, Darrah JM, Pain KS, Guthrie J: Early
patients. Am J Phys Med Rehabil 76:366, 1997. predictors of functional independence 2 years
43. Bohannon RW: Internal consistency of after spinal cord injury. Arch Phys Med Rehabil
dynamometer measurements in healthy subjects 78:644, 1997.
and stroke patients. Percept Motor Skills 58. Armstrong LE, Winant DM, Swasey PR, et al:
81:1113, 1995.
Using isokinetic dynamometry to test ambulatory
44. Sunderland A, Tinson D, Bradley L, Langton patients with multiple sclerosis. Phys Ther
Hewer R: Arm function after stroke: An 63:1274, 1982.
evaluation of grip strength as a measure of
59. Chen W-Y, Pierson FM, Burnett CN: Force-time
recovery and a prognostic indicator. J Neurol
measurements of knee muscle functions of
Neurosurg Psychiatry 52:1267, 1989.
subjects with multiple sclerosis. Phys Ther
45. Bohannon RW, Smith MB: Upper extremity 67:934, 1987.
strength deficits in hemiplegic stroke patients:
60. Ponichtera TA, Rodgers MM, Glaser RM, et al:
Relationship between admission and discharge
Concentric and eccentric isokinetic lower
and time since onset. Arch Phys Med Rehabil
extremity strength in persons with multiple
68:155, 1987.
sclerosis. J Orthop Sports Phys Ther 16114
46. Nakamura R, Watanabe S, Handa T, Morohashi 1992.
I: The relationship between walking speed and
61. Andres PL, Finison LJ, Conlon T, et al: Use of
muscle strength for knee extension in
composite scores (megascores) to measure deficit
in amyotrophic lateral sclerosis. Neurology 79. Wyke M: Effect of brain lesions on the rapidity
38:405, 1988. of arm movement. Neurology 17:1113, 1967.
62. Scott OM, Hyde SA, Goddard C, Dubowitz V: 80. Baroni A, Benvenuti F, Fantini L, et al: Human
Quantitation of muscle function in children: A ballistic arm abduction movements: Effects of L-
prospective study in Duchenne muscular dopa treatment in Parkinson’s disease. Neurology
dystrophy. Muscle Nerve 5:291, 1982. 34:868, 1984.
63. Allsop KG, Ziter FA: Loss of strength and 81. Godaux E, Koulischer D, Jacquy J: Parkinsonian
functional decline in Duchenne’s dystrophy. Arch bradykinesia is due to depression in rate of rise
Neurol 38:406, 1981. of muscle activity. Ann Neurol 31:93, 1992.
64. Griggs RC, Moxley RT, Mandell JR, et al: 82. Wierzbicka MM, Wiegner AW: Effects of weak
Prednisone in Duchenne dystrophy: A antagonist on elbow flexion movements in man.
randomized controlled trial defining the time Exp Brain Res 91:509, 1992.
course and dose response. Arch Neurol 48:383, 83. Milner-Brown HS, Miller RG: Increased
1991. muscular fatigue in patients with neurogenic
65. Amundsen LR, Graves JM: Testing knee muscle weakness: Quantification and
extensor muscles of survivors of poliomyelitis. J pathophysiology. Arch Phys Med Rehabil
Hum Muscle Perform 1:25, 1991. 70:361, 1989.
66. Agre JC, Rodriguez AA: Neuromuscular 84. Lenman AJR, Tulley FM, Vrbova G, et al:
function: Comparison of symptomatic and Muscle fatigue in some neurological disorders.
asymptomatic polio subjects to control subjects. Muscle Nerve 12:938, 1989.
Arch Phys Med Rehabil 71:545, 1990. 85. Miller RG, Green AT, Moussavi RS, et al:
67. Munin MC, Jaweed M, Staas WE, et al: Post Excessive muscular fatigue in patients with
poliomyelitis muscle weakness: A prospective spastic paraparesis. Neurology 40:1271, 1990.
study of quadriceps strength. Arch Phys Med 86. Checchia GA, Gionnone F, Miccoli B, et al: Iso¬
Rehabil 72:729, 1991. kinetic testing of muscular function and fatigue
68. Andersson T, Siden A: A clinical study of the in patients with multiple sclerosis. Isokinetics
Guillain-Barre syndrome. Acta Neurol Scand Exercise Science 3:101, 1993.
66:316, 1982. 87. Bohannon RW: Relative dynamic muscular
69. Roller W, Kase S: Muscle strength testing in endurance of patients with neuromuscular
Parkinson’s disease. Eur Neurol 25:130, 1986. disorders and of healthy matched control
70. Pedersen SW, Oberg B: Dynamic strength in subjects. Phys Ther 67:18, 1987.
Parkinson’s disease. Eur Neurol 33:97, 1993. 88. Zelaschi F, Felicetti G: Functional isokinetic
71. Pedersen SW, Oberg B, Larsson L-E, Lindval B: parameters in the hemiparetic patient: Training
Gait analysis, isokinetic muscle strength efficacy. Isokinetics Exercise Science 5:25, 1995.
measurement in patients with Parkinson’s 89. Cameron T, Calancie B: Mechanical and fatigue
disease. Scand J Rehabil Med 29:67, 1997. properties of wrist flexor muscles during
72. Corcos DM, Chen C-M, Quinn NP, et al: repetitive contractions after cervical spinal cord
Strength in Parkinson’s disease: Relationship to injury. Arch Phys Med Rehabil 76:929, 1995.
rate of force generation and clinical status. Ann 90. Walker GC, Cardenas DD, Guthrie MR, et al:
Neurol 39:79, 1996. Fatigue and depression in brain-injured patients
73. Bridgewater KJ, Sharpe MH: Trunk muscle correlated with quadriceps strength and
training and early Parkinson’s disease. Physiother endurance. Arch Phys Med Rehabil 72:469,
Theory Pract 13:139, 1997. 1991.
74. Bohannon RW, Walsh S: Nature reliability, and 91. Sale D: Testing strength and power. In
predictive value of muscle performance measures MacDougall JD, Wenger HA, Green HJ (eds):
in patients with hemiparesis following stroke. Physiological Testing of the High-Performance
Arch Phys Med Rehabil 73:721, 1992. Athlete, 2nd ed. Champaign, IL, Human
75. Tsuji I, Nakamura R: The altered time course of Kinetics, 1991, pp 22-24.
tension development during the initiation of fast 92. Olsen TS: Arm and leg paresis as outcome
movement in hemiplegic patients. Tohoku J Exp predictors in stroke rehabiliation. Stroke 21:247,
Med 151:137, 1987. 1990.
76. Wierzbicka MM, Wiegner AW, Logigian EL, 93. Friedman PJ: The star cancellation test in acute
Young RR: Abnormal most-rapid isometric stroke. Clin Rehabil 6:23, 1992.
contractions in patients with Parkinson’s disease. 94. Friedman PJ, Davis G, Allen B: Semi-
J Neurol Neurosurg Psychiatry 54:210, 1991. quantitative SPECT scanning in acute ischaemic
77. Bohannon RW, Walsh S: Relationship between stroke. Scand J Rehabil Med 25:99, 1993.
isometric torque and velocity of knee extension 95. Stone SP, Patel P, Greenwood RJ: Selection of
following stroke with hemiparesis. J Hum acute stroke patients for treatment of visual
Muscle Performance 1:40, 1991. neglect. J Neurol Neurosurg Psychiatry 56:463,
78. Davies JM, Mayston MJ, Newhan DJ: Electrical 1993.
and mechanical output of the knee muscles 96. Logigian MK, Samuels MA, Falcover J, Zagar
during isometric and isokinetic activity in stroke R: Clinical exercise trial for stroke patients. Arch
and healthy adults. Disabil Rehabil 18:83, 1996. Phys Med Rehabil 64:364, 1983.
97. Friedman PJ: Spatial neglect in acute stroke: the relationship of lower extremity strength and gait
line bisection test. Scand J Rehabil Med 22:101, parameters in patients with post-polio syndrome.
1990. Arch Phys Med Rehabil 74:165, 1993.
98. Abdulwahab SS: Physical disability in patients 114. Bohannon RW: Determinants of transfer capacity
with hemiparesis. Int J Rehabil Res 19:157, in patients with hemiparesis. Physiother Can
1996. 40:236, 1988.
99. Segboer SHAM, Terpstra-Lindeman E, 115. Bohannon RW: Rolling to the nonplegic side:
Buitenhuis M, Adam JJ: Spastic paresis. I. Influence of teaching and limb strength in
Relationship between impairment and disability. J hemiplegic stroke patients. Clin Rehabil 2:215,
Rehabil Sci 6:3, 1993. 1988.
100. Lazar RB, Yarkony GM, Ortolano P, et al: 116. Welch RD, Lobley SJ, O’Sullivan SB, Freed
Prediction of functional outcome by motor MM: Functional independence in quadriplegia:
capability after spinal cord injury. Arch Phys Critical levels. Arch Phys Med Rehabil 67:235,
Med Rehabil 70:819, 1989. 1986.
101. Marino RJ, Huang M, Knight P, et al: Assessing 117. Bohannon RW: Determinants of patient gown
selfcare status in quadriplegia: A comparison of donning performance soon after stroke. Eur J
quadriplegia index of function (QIF) and the Phys Med Rehabil 2:70, 1992.
functional independence measure (FIM). 118. Bohannon RW, Walsh S: Association of paretic
Paraplegia 31:225, 1993. lower extremity muscle strength and balance
102. Bohannon RW: Relationship among paretic knee with stair climbing ability in patients with stroke.
extension strength, maximum weight bearing, J Stroke Cerebrovasc Dis 1:129, 1991.
and gait speed in patients with stroke. J Stroke 119. Tupling SJ, Davis GM, Pierrynowski R, Shepard
Cerebrovasc Dis 1:65, 1991. RJ: Arm strength and impulse generation:
103. Nakamura R, Hosokawa T, Tsuji I: Relationship Initiation of wheelchair movement by the
of muscle strength for knee extension to walking physically disabled. Ergonomics 29:303, 1986.
capacity in patients with spastic hemiparesis. 120. Shima H: A study of key muscles for wheelchair
Tohoku J Exp Med 145:335, 1995. driving in quadriplegics. J Phys Ther Sci 9:43,
104. Suzuki K, Nakamura R, Yamada Y, Handa T: 1997.
Determinants of maximum walking speed in 121. Hesse S, Schauer M, Malezic M, et al:
hemiparetic stroke patients. Tohoku J Exp Med Quantitative analysis of rising from a chair in
162:337, 1990. healthy and hemiparetic subjects. Scand J
105. Bohannon RW: Knee extension power, velocity, Rehabil Med 26:161, 1994.
and torque: Relative deficits and relation to 122. Yoshida K, Iwakura H, Inoue F: Motion analysis
walking performance in stroke patients. Clin in the movements of standing up from and sitting
Rehabil 6:125, 1992. down on a chair. Scand J Rehabil Med 15:133
106. Bohannon RW: Knee extension force 1983.
measurements are reliable and indicative of 123. Engardt M, Olsson E: Body weight-bearing while
walking speed in stroke patients. Int J Rehabil rising and sitting down in patients with stroke.
Res 12:193, 1989. Scand J Rehabil Med 24:67, 1992.
107. Bohannon RW, Andrews AW: Correlation of 124. Engardt M, Ribbe T, Olsson E: Vertical ground
knee extensor muscle torque and spasticity with reaction force feedback to enhance stroke
gait speed in patients with stroke. Arch Phys patients’ symmetrical body-weight distribution
Med Rehabil 70:330, 1990. while rising/sitting down. Scand J Rehabil Med
108. Bohannon RW: Selected determinants of 25:41, 1993.
ambulatory capacity in patients with hemiplegia. 125. Fowler V, Carr J: Auditory feedback: Effects on
Clin Rehabil 3:47, 1989.
vertical force production during standing up
109. Agre JC, Findley TW, McNally MC, et al:
following stroke. Int J Rehabil Res 19:265, 1996.
Physical activity in children with 126. Engardt M: Long term effects of auditory
myelomeningocele. Arch Phys Med Rehabil feedback training on relearned symmetrical body
68:372, 1987.
weight distribution in stroke patients. Scand J
110. Waters RL, Adkins R, Yakura J, Vigil D: Rehabil Med 26:65, 1994.
Prediction of ambulatory performance based on 127. Bohannon RW, Larkin PA: Lower extremity
motor scores derived from standards of the
weight bearing under various standing conditions
American Spinal Injury Association. Arch Phys in independently ambulatory patients with
Med Rehabil 75:756, 1994.
hemiparesis. Phys Ther 65:1323, 1985.
111. Waters RL, Yakura JS, Adkins R, Barnes G:
128. Waldron RM, Bohannon RW: Weight distribution
Determinants of gait performance following
when standing: The influence of a single point
spinal cord injury. Arch Phys Med Rehabil
cane in patients with stroke. Physiother Pract
70:811, 1980. 5:171, 1989.
112. Waters RL, Yakura JS, Adkins RH: Gait
129. Caldwell C, MacDonald D, MacNeil K, et al:
performance after spinal cord injury. Clin Orthop
Symmetry of weight distribution in normals and
288:87, 1993.
stroke patients using digital weigh scales.
113. Perry J, Mulroy SJ, Ren wick SE: The
Physiother Pract 2:109, 1986.
130. Arcan M, Brail MA, Najenson T, Solzi P: FGP et al: Gait pattern in the early recovery period
assessment of postural disorders during the after stroke. J Bone Joint Surg 78-A:1506, 1996.
process of rehabilitation. Scand J Rehabil Med 148. Sutherland DH, Cooper L, Daniel D: The role of
9:165, 1977. ankle plantar flexors in normal walking. J Bone
131. Dickstein R, Nissan M, Pillar T, Scheer D: Foot- Joint Surg 62A:354, 1980.
ground pressure pattern of standing hemiplegic 149. Olney SJ, Griffin MP, Monga TN, McBride ID:
patients. Phys Ther 64:19, 1984. Work and power in gait of stroke patients. Arch
132. Bohannon RW, Waldron RM: Weightbearing Phys Med 72:309, 1991.
during comfortable stance in patients with stroke: 150. Olney SJ, Griffin MP, McBride ID: Temporal,
Accuracy and reliability of measurements. Aust J Kinematic and kinetric variables related to gait
Physiother 37:19, 1991. speed in subjects with hemiplegia: A regression
133. Bohannon RW, Tinti-Wald D: Accuracy of approach. Phys Ther 74:872, 1994.
weightbearing estimation by stroke versus 151. Hesse SA, Jahnke MT, Schreiner C, Mauritz
healthy subjects. Percept Motor Skills 72:935, K-H: Gait symmetry and functional walking
1991. performance in hemiparetic patients prior to and
134. Goldie P, Evans O, Matyas T: Performance in the after a 4-week rehabilitation programme. Gait
stability limits test during rehabilitation following Posture 1:166, 1993.
stroke. Gait Posture 4:315, 1996. 152. Carlsoo S, Dahlof A-G, Holm J: Kinetic analysis
135. Turnbull GI, Charteris J, Wall JC: Deficiencies in of the gait in patients with hemiparesis and in
standing weight shifts by ambulant hemiplegic patients with intermittent claudication. Scand J
subjects. Arch Phys Med Rehabil 77:356, 1996. Rehabil Med 6:166, 1974.
136. Goldie PA, Matyas TA, Evans OM, et al: 153. Nilsson J, Thorstensson A: Ground reaction
forces at different speeds of human walking and
Maximum voluntary weightbearing by the
running. Acta Physiol Scand 136:217, 1989.
affected and unaffected legs in standing
154. Morita S, Yamomoto H, Furaya K: Gait analysis
following stroke. Clin Biomech 11:333, 1996.
of hemiparetic patients by measurement of
137. Brant D, Lafferty MJ, McKeon A, et al:
ground reaction force. Scand J Rehabil Med
Invariant characteristics of gait initiation. Am J
27:37, 1995.
Phys Med Rehabil 70:206, 1991.
155. Ueno E, Yanagisawa N, Takami M: Gait
138. Rogers MW, Pai YC: Dynamic transitions in
disorders in Parkinsonism: A study with floor
stance support accompanying leg flexion
reaction forces and emg. Adv Neurol 60:414,
movements in man. Exp Brain Res 81:398, 1990.
1993.
139. Rogers MW, Hedman LD, Pai Y-C: Kinetic
156. Lunnen JD, Yack J, LeVeau BF: Relationship
analysis of dynamic transitions in stance support
between muscle length, muscle activity and
accompanying voluntary leg flexion movements
torque of the hamstring muscles. Phys Ther
in hemiparetic adults. Arch Phys Med Rehabil
61:190, 1981.
74:19, 1993. 157. Bohannon RW, Reed ML, Gajdosik R:
140. Von Schroeder HP, Courts RD, Lyden PD, et al: Electrically evoked knee flexion torque increases
Gait parameters following stroke: A practical with increased pelvifemoral angles. Clin
assessment. J Rehabil Res Devel 32:25, 1995. Biomech 5:23, 1990.
141. Pai Y-C, Rogers MW, Hedman LD, Hanke TA: 158. Bohannon W: Decreased isometric knee flexion
Alterations in weight-transfer capabilities in torque with hip extension in hemiparetic patients.
adults with hemiparesis. Phys Ther 74:647, 1994. Phys Ther 66:521, 1986.
142. Perry J, Ireland ML, Gronley J, Hoffer MM: 159. Bohannon RW, Gajdosik RL, LeVeau BF:
Predictive value of manual muscle testing of gait Relationship of pelvic and thigh motions during
analysis in normal ankles by dynamic unilateral and bilateral hip flexion. Phys Ther
electromyography. Foot Ankle 6:254, 1986. 65:1501, 1985.
143. Miyashita M, Matsui H, Miura M: The relation 160. Komi PV: Physiological and biomechanical
between electrical activity in muscle and speed correlates of muscle function: Effects of muscle
of walking and running. Medicine and Sport: structure and stretch-shortening cycle on force
Biomechanics II 6:192, 1971. and speed. Exerc Sport Sci Rev 12:81, 1984.
144. Winter DA: Energy generation and absorption at 161. Svantesson U, Sunnerhagen KS: Stretch¬
the ankle and knee during fast, natural and slow shortening cycle in patients with upper motor
cadences. Clin Orthop 175:147, 1983. neuron lesions due to stroke. Eur J Appl Physiol
145. Winter DA, White SC: Cause-effect correlations 75:312, 1997.
of variables of gait. In Jonsson B (ed): 162. Wretenberg P, Arborelius UP: Power and work
Biomechanics X—A. Champaign, IL, Human produced in different leg muscle groups when
Kinetics, 1987, p 363. rising from a chair. Eur J Appl Physiol 68:413,
146. Siegler KL, Stanhope SJ, Caldwell GE: 1994.
Kinematic and kinetic adaptations in the lower 163. Voss DE, Ionta MK, Myers BJ: Proprioceptive
limb during stance in gait of unilateral femoral Neuromuscular Facilitation, 3rd ed. Philadelphia,
neuropathy patients. Clin Biomech 8:147, 1993. Harper & Row, 1985, pp 293-294.
147. Kramers de Quervain IA, Simon SR, Leurgans S, 164. Bohannon RW: Hemiparetic elbow flexion force
production before and after muscle belly tapping. 181. Matyas TA, Galea MP, Spicer SD: Facilitation of
Neurol Report 11:75, 1987. the maximum voluntary contraction in
165. Liberson WT, Holmquest HJ, Scott D, Dow M: hemiplegia by concomitant cutaneous
Functional electrotherapy: Stimulation of the stimulation. Am J Phys Med 65:125, 1986.
peroneal nerve synchronized with the swing 182. Neeman RL, Neeman M: Efficacy of orthokinetic
phase of the gait of hemiplegic patients. Arch orthotics for post-stroke upper extremity
Phys Med Rehabil 42:101, 1961. hemiparetic motor dysfunction. Int J Rehabil Res
166. Granat MH, Maxwell DJ, Ferguson ACB, et al: 16:302, 1993.
Peroneal stimulator: Evaluation for the correction 183. Neeman RL, Neeman HJ, Neeman M:
of spastic drop foot in hemiplegia. Arch Phys Application of orthokinetic orthoses in
Med Rehabil 77:19, 1996. habilitation of a person with upper extremity in
167. Burridge JH, Taylor PN, Hagan SA, et al: The coordination secondary to spastic quadriplegia
effects of common peroneal stimulation on the due to cerebral palsy. Can J Rehabil 1:145, 1988.
effort and speed of waling a randomized 184. Blashy MRM, Fuchs-Neeman RL: Orthokinetics:
controlled trial with chronic hemiplegic patients. A new receptor facilitation method. Am J Occup
Clin Rehabil 11:201, 1997. Ther 13:226, 1959.
168. Stanic U, Acimovic-Janezic R. Gros N, et al: 185. Inaba M, Edberg E, Montgomery J, Gillis MK:
Multichannel electrical stimulation for correction Effectiveness of functional training, active
of hemiplegic gait. Scand J Rehabil Med 10:75, exercise, and resistive exercise for patients with
1978. hemiplegia. Phys Ther 53:28, 1973.
169. Faghri PD, Rodgers MM, Glaser RM, et al: The 186. Biitefisch C, Hummelsheim H, Denzler P,
effects of functional electrical stimulation on Mauritz K-H: Repetitive training of isolated
shoulder subluxation, arm function recovery, and movements improves the outcome of motor
shoulder pain in hemiplegic stroke patients. Arch rehabilitation of the centrally paretic hand. J
Phys Med Rehabil 75:73, 1994. Neurol Sci 130:59, 1995.
170. Merletti R, Acimovic R, Grobelnik S, Cvilak G: 187. Sharp SA, Brouwer BJ: Isokinetic strength
Electrophysiological orthosis for the upper training of the hemiparetic knee: Effects on
extremity in hemiplegia: Feasibility study. Arch function and spasticity. Arch Phys Med Rehabil
Phys Med Rehabil 56:507, 1975. 78:1231, 1997.
171. Bajd T, Kralj A, Turk R: Standing-up of a 188. Engardt M, Knutsson E, Jonsson M, Stemhag M:
healthy subject and a paraplegic patient. J Dynamic muscle strength training in stroke
Biomech 15:1, 1982. patients: Effects on knee extension torque
172. Ewins DJ, Taylor PN, Crook SE, et al: Practical electromyographic activity, and motor function.
low cost stand/sit system for mid-thoracic Arch Phys Med Rehabil 76:419, 1995.
paraplegics. J Biomed Eng 10:184, 1988. 189. Knutsson E, Gransberg L, Maartensson A:
173. Jaeger RJ, Yarkony GM, Smith RM: Standing Facilitation and inhibition of maximal voluntary
the spinal cord injured patient by electrical contractions by the activation of muscle stretch
stimulation: Refinement of a protocol for clinical reflexes in patients with spastic paresis.
use. TEEE Trans Biomed Eng 36:720, 1989. Electroenceph Clin Neurophysiol 70:37P, 1988.
174. Marsolais EB, Kobetic R: Functional walking in 190. MacPhail HEA, Kramer JF: Effect of isokinetic
paralyzed patients by means of electrical strength-training on functional ability and
stimulation. Clin Orthop 175:30, 1983. walking efficiency in adolescents with cerebral
175. Bajd T, Kralj A, Turk K, et al: The use of a palsy. Dev Med Child Neurol 37:763, 1995.
four-channel electrical stimulator as an 191. Horvat M: Effects of a progressive resistance
ambulatory aid for paraplegic patients. Phys Ther training program on an individual with spastic
63:1116, 1983.
cerebral palsy. Am Corr Ther J 41:7, 1987.
176. Petrofsky JS, Heaton H, Phillips CA: Outdoor 192. McCubbin JA, Shasby GB: Effects of isokinetic
bicycle for exercise in paraplegics and exercise on adolescents with cerebral palsy.
quadriplegics. J Biomed Eng 5:292, 1983. Adapt Phys Act Q 2:56, 1985.
177. Bajd T, Stafancic M, Matjacic Z, et al: 193. Damiano DL, Kelly LE, Vaughn CL: Effects of
Improvement in step clearance via calf muscle quadriceps femoris muscle strengthening on
stimulation. Med Biol Eng Comput 35:113, 1997. crouch gait in children with spastic diplegia
178. Shropshire BM, Broton JG, Cameron TL, Klose Phys Ther 75:658, 1995.
KJ: Improved motor function in tetraplegics 194. O’Connell DG, Barnhart R: Improvement in
following neuromuscular stimulation assisted arm wheelchair propulsion in pediatric wheelchair
ergometry. J Spinal Cord Med 20:49, 1997. users through resistance training: A pilot study.
179. Hagbarth KE, Eklund G: The muscle vibrator: A Arch Phys Med Rehabil 76:368, 1995.
useful tool in neurologic therapeutic work. Scand 195. Gersten JW: Isometric exercises in the paraplegic
J Rehabil Med 1:26, 1969.
and in the patient with weakness of quadriceps
180. Knutsson E, Lindblom U, Odeen I: Reflex
and hamstrings. Arch Phys Med Rehabil 42'498
facilitation by muscle vibration in the treatment 1961.
of spastic hemiparesis. Scand J Rehabil Med 196. Davis GM, Shephard RJ: Strength training for
2-3:110, 1970.
wheelchair users. Br J Sports Med 24:25, 1990.
197. McCartney N, Moroz D, Gamer SH, McComas adults with traumatic head injuries. Phys Ther
AJ: The effects of strength training in patients 77:240, 1997.
with selected neuromuscular disorders. Med Sci 213. Bohannon RW, Larkin PA: Passive ankle
Sports Exerc 20:362, 1988. dorsiflexion increases in patients after a regimen
198. DeLorme TL, Schwab RS, Watkins AL: The of tilt table-wedge board standing. Phys Ther
response of the quadriceps femoris to progressive 65:1676, 1985.
resistance exercises in poliomyelitic patients. J 214. Light KE, Nuzik S, Personius W, Barstrom A:
Bone Joint Surg 30A:834, 1948. Low-load prolonged stretch vs high-load brief
199. Filly aw MJ, Badger GJ, Goodwin GD, et al: The stretch in treating knee contractures. Phys Ther
effects of long-term non-fatiguing resistance 64:330, 1984.
exercise in subjects with post-polio syndrome. 215. Odeen I: Reduction of muscular hypertonus by
Orthopedics 14:1253, 1991. long-term muscle stretch. Scand J Rehabil Med
200. Agre JC, Rodriguez AA, Franke TM, et al: Low- 13:93, 1981.
intensity, altemate-day exercise improves muscle 216. Keenan MA: Management of the spastic upper
performance without apparent adverse affect in extremity in the neurologically impaired adult.
postpolio patients. Am J Phys Med Rehabil Clin Orthop 233:116, 1988.
75:50, 1996. 217. Petrillo CR, Knoploch S: Phenol block of the
201. Spector SA, Lemmer JT, Koffman BM, et al: tibial nerve for spasticity: A long-term follow-up
Safety and efficacy of strength training in study. Int Disabil Studies 10:97, 1988.
patients with sporadic inclusion body myositis. 218. Pierson SH, Katz DI, Tarsy D: Botulinum toxin
Muscle Nerve 20:1242, 1997. in the treatment of spasticity: Functional
202. Wolf SL, Edwards DI, Shutter LA: Concurrent implications and patient selection. Arch Phys
assessment of muscle activity (CAMA): A Med Rehabil 77:717, 1996.
procedural approach to assess treatment goals. 219. Bhakta BB, Cozens JA, Bamford JM,
Phys Ther 66:218, 1986. Chamberlain MA: Use of botulinum toxin in
203. Bohannon RW: Electromyographic activity of the stroke patients with severe upper limb spasticity.
quadriceps femoris muscles during four activities J Neurol Neurosurg Psychiatry 61:30, 1996.
in stroke rehabilitation. Int J Rehabil Res 13:80, 220. Reiter F, Danni M, Ceravolo MG, Provinciali L:
1990. Disability changes after treatment of upper limb
204. Augustsson J, Esko A, Thomee R, Svantesson U: spasticity with botulinum toxin. J Neurol Rehabil
Weight training of the thigh muscles using closed 10:47, 1996.
versus open kinetic chain exercises: A 221. Corry IS, Cosgrove AD, Walsh EG, et al:
comparison of performance enhancement. J Botulinum toxin A in the hemiplegic upper limb:
Orthop Sports Phys Ther 27:3, 1998. A double-blind trial. Dev Med Child Neurol
205. Sale DG, Martin JE, Moroz DE: Hypertrophy 39:185, 1997.
without increased isometric strength after weight 222. Dunne JW, Heye N, Dunne SL: Treatment of
training. Eur J Appl Physiol 64:51, 1992. chronic limb spasticity with botulinum toxin A. J
206. Behm DG, Sale DG: Intended rather than actual Neurol Neurosurg Psychiatry 58:232, 1995.
movement velocity determines velocity-specific 223. Hesse S, Friedrich H, Domasch C, Mauritz K-H:
training response. J Appl Physiol 74:359, 1993. Botulinum toxin therapy for upper limb flexor
207. Olney BW, Williams PF, Meneleus MB: spasticity: Preliminary results. J Rehabil Sci
Treatment of spastic equinus by aponeurosis 5:98, 1992.
lengthening. J Pediatr Orthop 8:422, 1988. 224. Feve A, Decq P, Filipetti P, et al: Physiological
208. Matsuo T, Hara H, Tada S: Selective lengthening effects of selective tibial neurotony on lower
of the psoas and rectus femoris and preservation limb spasticity. J Neurol Neurosurg Psychiatry
of the iliacus for flexion deformity of the hip in 63:575, 1997.
cerebral palsy patients. J Pediatr Orthop 7:690, 225. Steinbok P, Reiner AM, Beauchamp R, et al: A
1987. randomized clinical trial to compare selective
209. Reimers J: Functional changes in the antagonists posterior rhizotomy plus physiotherapy with
after lengthening the agonists in cerebral palsy. I. physiotherapy alone in children with spastic
Triceps surae lengthening. Clin Orthop 253:30, diplegic cerebral palsy. Dev Med Child Neurol
1990. 39:178, 1997.
210. Reimers J: Functional changes in the antagonists 226. Thomas SS, Aiona MD, Buckon CE, Piatt JH:
after lengthening the agonists in cerebral palsy. Does gait continue to improve 2 years after
II. Quadriceps strength before and after distal selective dorsal rhizotomy? J Pediatr Orthop
hamstring lengthening. Clin Orthop 253:35, 17:387, 1997.
227. Levin MF, Hui Chan CWY: Relief of hemiparetic
1990.
211. Conine TA, Sullivan T, Mackie T, Goodman M: spasticity by TENS is associated with
Effect of serial casting for the prevention of improvement in reflex and voluntary motor
equinus in patients with acute head injury. Arch functions. Electroenceph Clin Neurophysiol
Phys Med Rehabil 71:310, 1990. 85:131, 1992.
212. Moseley AM: The effect of casting combined 228. Merletti R, Zelaschi F, Latella D, et al: A control
with stretching on passive ankle dorsiflexion in study of muscle force recovery in hemiparetic
patients during treatment with functional Electromyographic biofeedback for

electrical stimulation. Scand J Rehabil Med neuromuscular reeducation in the hemiplegic
10:147, 1978. stroke patient: A meta analysis. Arch Phys Med
229. Smith LE: Restoration of volitional limb Rehabil 74:1301, 1993.
movement of hemiplegics following patterned 244. Glanz M, Klawansky S, Stason W, et al:
functional electrical stimulation. Percept Motor Biofeedback therapy in post stroke rehabilitation:
Skills 71:851, 1990. A meta-analysis of the randomized controlled
230. Lagasse PP, Roy M-A: Functional electrical trials. Arch Phys Med Rehabil 76:508, 1995.
stimulation and the reduction of co-contraction in 245. Intiso D, Santilli V, Grasso MG, et al:
spastic biceps brachii. Clin Rehabil 3:111, 1989. Rehabilitation of walking with electromyographic
231. Winchester P, Montgomery J, Bowman B, Hislop biofeedback in footdrop after stroke. Stroke
H: Effects of feedback stimulation training and 25:1189, 1994.
cyclical electrical stimulation on knee extension 246. Marciello MA, Herbison GJ, Cohen ME,
in hemiparetic patients. Phys Ther 63:1096, Schmidt R: Elbow extension using anterior
1983. deltoids and upper pectorals in spinal cord-
232. Bowman BR, Baker LL, Waters RL: Positional injured subjects. Arch Phys Med Rehabil 76:426,
feedback and electrical stimulation: An 1995.
automated treatment for the hemiplegic wrist. 247. Bergstrom EMK, Frankel HL, Galer IAR, et al:
Arch Phys Med Rehabil 60:497, 1979. Physical ability in relation to anthropometric
233. Hazlewood ME, Brown JK, Rowe PJ, Salter PM: measurements in persons with complete spinal
The use of therapeutic electrical stimulation in cord lesion below the sixth cervical segment. Int
the treatment of hemiplegic cerebral palsy. Dev Rehabil Med 7:51, 1985.
Med Child Neurol 36:661, 1994. 248. Butler PB, Nene AV, Major RE: Biomechanics of
234. Pape KE, Kirsch SE, Galil A, et al: transfer from sitting to the standing position in
Neuromuscular approach to the motor deficits of some neuromuscular diseases. Physiotherapy
cerebral palsy: A pilot study. J Pediatr Orthop 77:521, 1991.
13:628, 1993. 249. Griffiths RD: Controlling weight in muscle
235. Barr FMD, Moffat B, Bayley JIL, Middleton disease to reduce the burden. Physiotherapy
FRI: Evaluation of the effects of functional 75:190, 1989.
electrical stimulation on muscle power and 250. Edwards RHT, Round JM, Jackson MJ, et al:
spasticity in spinal cord injured patients. Clin Weight reduction in boys with muscular
Rehabil 3:17, 1989. dystrophy. Dev Med Child Neurol 26:384, 1984.
236. Rabischong E, Ohanna F: Effects of functional 251. Pillar T, Dickstein R, Smolinski Z: Walking
electrical stimulation (FES) on evoked muscular reeducation with partial relief of body weight in
output in paraplegic quadriceps muscle. rehabilitation of patients with locomotor
Paraplegia 30:467, 1992. disabilities. J Rehabil Res 28:47, 1991.
237. Farragher D, Kidd GL, Tallis R: Eutrophic 252. Visintin M, Barbeau H: The effects of parallel
electrical stimulation for Bell’s Palsy. Clin bars, body weight support and speed on the
Rehabil 1:265, 1987.
modulation of the locomotor pattern of spastic
238. Naeser MA, Alexander MP, Stiassny-Eder D, et paretic gait: A preliminary communication.
al: Acupuncture in the treatment of hand paresis Paraplegia 32:540, 1994.
in chronic and acute stroke patients: 253. Hesse S, Helm B, Krajnik J, et al: Treadmill
Improvement observed in all cases. Clin Rehabil training with partial body weight support:
8:127, 1994.
Influence of body weight release on the gait of
239. Naeser MA, Alexander MP, Stiassny-Eder D, et hemiparetic patients. J Neurol Rehabil IT 15
al: Real versus sham acupuncture in the 1997.
treatment of paralysis in acute stroke patients: A 254. Hesse S, Bertelt C, Jahnke MT, et al: Treadmill
CT scan lesion site study. J Neurol Rehabil training with partial body weight support
6:163, 1992.
compared with physiotherapy in nonambulatory
240. Magnusson M, Johansson K, Johansson BB: hemiparetic patients. Stroke 26:976, 1995.
Sensory stimulation promotes normalization of 255. Hesse S, Bertelt C, Schaffrin A, et al:
postural control after stroke. Stroke 25:1176, Restoration of gait in nonambulatory hemiparetic
1994.
patients by treadmill training with partial body
241. Kjendahl A, Sallstrom S, 0sten PE, et al: A one weight support. Arch Phys Med Rehabil 75:1087
year follow-up study on the effects of 1994.
acupuncture in the treatment of stroke patients in 256. Harrison R, Bulstrode S: Percentage of
the subacute stage: A randomized, controlled weightbearing during partial immersion in the
study. Clin Rehabil 11:192, 1997.
hydrotherapy pool. Physiother Pract 3:60, 1987.
242. Johansson K, Lindgren I, Widner H, et al: Can 257. Harrison R, Hillman M, Bulstrode S: Loading of
sensory stimulation improve the functional
the lower limb when walking partially immersed
outcome in stroke patients? Neurology 43:2189
implications for clinical practice. Physiotherapy
1993. 78:164, 1992.
243. Schleenbaker RE, Mainous AG:
258. Burdett RG, Habasevich R, Pisciotta J, Simon
SR: Biomechanical comparison of rising from 274. Lu CL, Yu B, Basfored JR, et al: Influences of
two types of chairs. Phys Ther 65:1177, 1985. cane length on the stability of stroke patients. J
259. Finlay OE, Bayles TB, Rosen C, Milling J: Rehabil Research Devel 34:91, 1997.
Effects of chair design, age and cognitive status 275. Milczarek JJ, Kirby RL, Harrison ER, MacLeod
on mobility. Age Ageing 12:329, 1983. DA: Standard and four-footed canes: Their effect
260. Weiner DK, Long R, Hughes MA, et al: When on the standing balance of patients with
older adults face the chair-rise challenge. J Am hemiparesis. Arch Phys Med Rehabil 74:281,
Geriatr Soc 41:6, 1993. 1993.
261. Arborelius UP, Wretenberg P, Lindberg F: The 276. Baxter ML, Allington RO, Koepke GH: Weight
effects of armrests and high seat heights on distribution variables in the use of crutches and
lower limb joint load and muscular activity canes. Phys Ther 49:360, 1969.
during sitting and rising. Ergonomics 11:1377, 277. Dickstein R, Abulaffio N, Pillar T: Vertical force
1992. loaded on walking canes in hemiparetic patients.
262. Wretenberg P, Lindberg F, Arborelius UP: Effect Gait Posture 1:113, 1993.
of armrests and different ways of using them on 278. Cybulski GR, Jaeger RJ: Standing performance
hip and knee load during rising. Clin Biomech of persons with paraplegia. Arch Phys Med
8:95, 1993. Rehabil 67:103, 1986.
263. Wretenberg P, Arborelius UP, Lindberg F: The 279. Mikelberg R, Reid S: Spinal cord lesions and
effects of a pneumatic stool and a one-legged lower extremity bracing: An overview and
stool or lower limb joint load and muscular follow-up study. Paraplegia 19:379, 1981.
activity during sitting and rising. Ergonomics 280. Coughlan JK, Robinson CE, Neumarch B,
36:519, 1993. Jackson G: Lower extremity bracing in
264. Wretenberg P, Arborelius UP, Weidenhelm L, paraplegia: Follow-up study. Paraplegia 18:25,
Lindberg F: Rising from a chair by a spring- 1980.
loaded flap seat: A biomechanical analysis. 281. Burdette RG, Borello-France D, Blatchly C,
Scand J Rehabil Med 25:153, 1993. Potter C: Gait comparison of subjects with
265. Thomas L, Fewell E, Walker JM: Energy cost of hemiplegia walking unbraced, with ankle-foot
walking and of wheelchair propulsion by orthosis, and with Air-Stirrup brace. Phys Ther
children with myelodysplasia: Comparison with 68:1197, 1988.
normal children. Dev Med Child Neurol 25:617, 282. Mojica JAP, Nakamura R, Kobayashi T, et al:
1983. Effect of ankle-foot orthosis (AFO) on body
266. Hawran S, Biering-Sorensen F: The use of long sway and walking capacity of hemiparetic stroke
leg calipers for paraplegic patients: A follow-up patients. Tohoku J Exp Med 156:395, 1988.
study of patients discharged 1973-82. Spinal 283. Isakov E, Mizrahi J, Onna I, Susak Z: The
Cord 34:666, 1996. control of genn recurvatum by comgining the
267. Harbum LK, Spaulding SJ: Muscle activity in Swedish knee-cage and an ankle-foot brace.
spinal cord-injured during wheelchair ambulation. Disabil Rehabil 14:187, 1992.
Am J Occup Ther 40:629, 1986. 284. DeVries J: Evaluation of lower leg orthosis use
268. Mulroy SJ, Gronley JK, Newsam CJ, Perry J: following cerebro-vascular accident. Int J Rehabil
Electromyographic activity of shoulder muscles Res 14:239, 1991.
during wheelchair propulsion by paraplegic 285. Yamamoto S, Miyazaki S, Kubota T:
persons. Arch Phys Med Rehabil 77:187, 1996. Quantification of the effect of the mechanical
269. Blower P: The advantages of the early use of property of ankle-foot orthoses on hemiplegic
wheelchairs in the treatment of hemiplegia. Clin gait. Gait Posture 1:27, 1993.
Rehabil 2:323, 1988. 286. Dieli J, Ayyappa E, Hombeak S: Effect of
270. Blower PW, Carter LC, Sulch DA: Relation dynamic AFOs on three hemiplegic adults. J
between wheelchair propulsion and independent Prosth Orthot 9:82, 1997.
walking in hemiplegic stroke. Stroke 26:606, 287. Lehmann JF, Condon SM, Price R, deLateur BJ:
Gait abnormalities in hemiplegia: Their
1995.
271. Butler C: Effects of powered mobility on self- correction by ankle-foot orthoses. Arch Phys
initiated behaviors of very young children with Med Rehabil 68:763, 1987.
locomotor disability. Dev Med Child Neurol 288. Hesse S, Luecke D, Jahnke MT, Mauritz KH:
Gait functin in spastic hemiparetic patients
28:325, 1986.
272. Butler C: Effects of powered mobility on self- walking barefoot, with firm shoes, and with
initiated behaviors of very young children with ankle-foot orthosis. Int Rehabil Res 19:133,
locomotor disability. Dev Med Child Neurol 1996.
289. Corcoran PJ, Jebsen RH, Brengelmann GL,
28:325, 1986.
273. Smith ME, Walton MS, Garraway WM: The use Simons BC: Effects of plastic and metal leg
of aids and adaptations in a study of stroke braces on speed and energy cost of hemiparetic
ambulation. Arch Phys Med Rehabil 51:69, 1970.
rehabilitation. Health Bull 39:98, 1981.
'•
'
CHAPTER 8
Biomechanics of Unperturbed
Standing Balance
Joseph Mizrahi
The Dynamic Nature of Balance dent (CVA) patients, craniocerebral injury (CCI)
patients, and amputees wearing their prostheses,
Bipedal standing is an unstable position that ne¬ as soon as they become able to stand without sup¬
cessitates a continuous regulation process, involv¬ port.
ing periodic contractions of muscles in the lower This chapter discusses the parameters that de¬
limbs and trunk.12 The direct result of this contin¬ scribe standing balance of able-bodied subjects
uously acting stabilizing system is the existence of and of subjects with neuromuscular and skeletal
body sway in the sagittal, coronal, and transversal deficiencies. A special emphasis is given to the
planes. The ability to maintain a stable upright measurement of the foot—ground reaction forces
position depends on the integrity of the musculo¬ in standing still. These are easily measurable
skeletal and neurologic systems. Specifically, the quantities that provide a natural input to biome¬
visual,3 vestibular,4 and proprioceptive5 systems chanical models on human standing. At the same
play significant roles in controlling human bal¬ time, they can form the basis for the clinical
ance. In different pathologies caused by disease characterization of standing posture in subjects
or trauma, one or more of these sensitive sources with balance pathologies.
can be damaged, resulting in impaired stability in
standing.
Measurement of postural balance has been used
Methods of Balance Measurement
in the past to assess stability in bipedal standing
and walking. The body situations in which balance
ability is needed are, however, diverse: from the UNPERTURBED STANDING TESTS
easiest ones, in which the body’s center of gravity
is absolutely within the base of support and in A commonly used method in the measurement of
which no motion of the supporting surface takes standing balance deals with the variations of the
place (e.g., lying supine), to the most demanding center of pressure of the vertical foot-ground reac¬
and exacting ones, such as maintaining balance tion forces obtained from fixed forceplates.6-11 In
while treading a rope on one leg, or during ice using this method, however, special care should
skating. Specifically, standing in the upright posi¬ be taken not to confuse the measured trajectory of
tion in the absence of external perturbations is a the center of pressure and that of the body’s center
position of fundamental importance. Postural sway of mass.12 In fact, earlier, Valk-Fai13 showed the
tests while standing still are relatively easy to existence of a difference between the center of
carry out, demanding minimal cooperation and pressure and center of mass trajectories, and Gur-
effort from the tested subjects or patients. These finkel6 estimated this difference for various values
tests also provide valuable and reliable informa¬ of the swaying frequency. Later, Shimba14 pro¬
tion on the mechanisms of unperturbed balance in posed a method for the estimation of the trajectory
normal and pathologic conditions. This particu¬ of the body’s center of gravity in single-limb
larly applies to patients with deficiencies in the stance from the rate of change of the moment of
locomotor system, such as cerebrovascular acci¬ momentum of the whole body, as obtained from
189
forceplate data. This method was later expanded separation line, toward the direction of the sound
to double-limb stance by Levin and Mizrahi.15 leg. These results clearly demonstrate the differ¬
Typical trajectories of the center of gravity with ence between the trajectory of the center of pres¬
respect to the platforms’ plane, as reported by sure and that of the center of gravity.
Levin16 for four able-bodied subjects and two sub¬ Elaborate methods of measurement of postural
jects with musculoskeletal pathologies, are shown sway have included measurement of all the com¬
in Figure 8-1. The pathologies included a right ponents of the actual reactive foot-ground forces,
below-knee amputation (BKA) and paralysis of including those in the tangential plane.17”20 Addi¬
the left plantar-flexor muscles. The average posi¬ tional methods, applied separately or in conjunc¬
tion of the center of gravity trajectories is indi¬ tion with the previously described methods, have
cated on the surface of the platforms. It can be included accelerometry of the trunk,17 goniometry
seen that the average position of the center of of the ankle joint,21 and measurement of the mo¬
gravity trajectories of the able-bodied subjects tion of various parts of the body by one of the
falls around the separation line between the plat¬ following means: mechanical pen writer,22 photog¬
forms, outside the contact area between each of raphy,12 light-emitting diodes,23 television and
the feet and the ground. Conversely, the average light spots,24 spark microphones,25 and variable-
position of the center of gravity trajectories of capacitance sway meter.26
the pathologic subjects is shifted away from the A major advantage of measuring the actual re-
AR [Right BKAl EL [N'l
left platform
■ V
02 1 y.
17^ x
Y
1 Oi
\ ■
i
15cm j 15cm
/ " L ^^
33 34 32 33 16
~lk-URE r TyP'Cal traieCt°ry °f the center of Sravity 'n the transverse plane for four able-bodied
subjects (N) and two subjects with pathology: right below-knee amputation (BKA) and paralysis of the left
plantar-flexor muscles (FMP). The average position of the center of gravity trajectories is pointed on the
Sâ“ , P k 'S n°ted that the avera«e Positions of the center of gravity trajectories of the
able-bodied subjects fall around the separation line (SL) between the platforms. Conversely, the average
positions of the center of gravity trajectories of the subjects with pathology are shifted away from the SL
toward the direction of the sound leg. EL, JL, TZ, and OR, normal subjects; AR, below-knee amputee; DD
bject with FMP, O,, 02, centers of platform, respectively; X, Y, anteroposterior and mediolateral axes
respectively. ’
I Chapter 8 BIOMECHANICS OF UNPERTURBED STANDING BALANCE 191
active forces is that the results obtained can be discussed by many investigators, for example,
easily related to the acceleration of the center of Gantchev and Popov,35 Hlavacka and
mass of the body, through the equations of motion. Litvinenkova,36 and Seidel and Brauer.37 In
Note, however, that in most of the studies dealing fact, in most investigations on postural
with force measurements, only one force platform balance, visual feedback was of central
was used, thus measuring the sum of the reactive interest.8-9-18> 22 It was reported that in the
forces, acting on both feet together. The impor¬ absence of vision, balance becomes less
tance of double forceplate measurements was later stable.8,21 Furthermore, Paulus and colleagues3
pointed out in conjunction with both bilateral char¬ showed that a decrease in visual acuity is
acterization of the swaying motion and multiseg- usually accompanied by an increase in postural
mental modeling of the human body.15-27 instability.
Other factors. Of special interest were the effects
TEST PARAMETERS IN UNPERTURBED on postural balance stability of factors such as
physical fatigue,29'38 environment of acoustic
STANDING
noise,29 anesthetics, drugs and alcohol,32 the
Studies of healthy subjects of different ages have effect of equilibrium lesions,39^12 the effect of
attempted to establish ranges of norms for this ischemic blocking of the leg,43 and the effect
population. Variability of the reported results was, of other general pathologies.9'40'44-48
however, very large, mainly because of the differ¬
ent methodologies used. An attempt to overcome PERTURBED STANDING TESTS
this problem was made by Kapteyn and col¬
The balance tests described previously were made
leagues28 by setting principles for the standardiza¬
in unperturbed standing; in some of the more
tion of postural sway forceplate measurements.
recent reports, dynamic testing conditions were
Significant test parameters in postural sway
also treated. One method of dynamic testing con¬
measurements include the following: sists of moving the measuring platform, either
Duration of test. The duration of swaying tests linearly or angularly, to create an effect of either
ranges in the literature from a few seconds29 to unexpected or continuous perturbation.9- 29> 49-53
a few minutes17-19-30; in most studies, however, Another method of perturbed testing is moving
the typical duration was within the range of 20 the visual scene in front of the tested subject.21 In
to 80 seconds.10- 20,31-33 either of these dynamic methods, the role of visual
feedback was shown to be much more significant
Footwear. Differences between swaying tests than that found when testing in static condi¬
made with or without shoes were not referred
tions.49' 51
to in most studies on this subject, with the The relative role of each of the visual, vestibu¬
exception of those of Harris and colleagues10 lar, and proprioceptive sensors has been described
and Shimba.20 in several studies, using both fixed and sway-
Positioning of and spacing between feet. The referenced supports.54-55 Particularly, the relative
effect of foot positioning was specifically weighting of the sensors of patients with vestibu¬
addressed by Stribley and colleagues,18 who lar deficits using sway-referenced supports and
reported that posture, especially in the vision was examined.56, 57 The adaptability and
mediolateral direction, was more stable with a relative contribution of each of the sensors in
wide opening between the feet compared with young children was reported.58 Fixed and sway-
standing with the feet parallel and touching.31 reference supports were also compared in conjunc¬
Similar findings were also reported by Nayak tion with chronic low-back dysfunction.59
and coauthors11 and, more recently, by Kirby The electrical activity in five lower limb mus¬
and associates.34 Another view is that no cles was investigated in dynamic standing bal¬
restrictions should be made on the foot ance.53 The spatial and temporal parameters were
placement, except that they should be measured separately in the sagittal and frontal
comfortably positioned.19 The influence of foot planes,60 using an instrumented stabilometer.61 It
positioning and spacing between the feet is was found that the total angular displacement of
especially important when bilateral the tilting platform and its fundamental frequency
measurements of postural balance are to be were strongly correlated. Additionally, spatiotem-
taken.27 poral symmetry was shown to indicate subjects
performance and to monitor the learning progress
Visual feedback. The importance of visual
feedback in the regulation of posture has been of the dynamic balance tasks.62
The Nature of Swaying Motion in In the more simple approach, the restoring com¬
Standing Still ponents in the system are treated as purely elastic
and passive.6,66 Later studies have demonstrated
Swaying motion while standing still has been re¬ that this treatment is erroneous.17, 19,24 For one
ported to be periodic, containing waves that follow thing, the inverted pendulum representation failed
no recognizable pattern.17'63 Although the cycles to predict correlations that should emerge from
of this motion are irregular in amplitude and fre¬ such a model between sway magnitudes and phy¬
quency, large low-frequency primary waves are sique variables, such as body height and weight,19
identifiable. Superimposed on these waves are sec¬ due to lack of symmetry of the actual body motion
ondary, smaller waves of higher frequencies. The around the vertical axis.21
higher-frequency waves were reported to be of Additionally, Thomas and Whitney17 and Les-
greater regularity, with regard to both cycle dura¬ tienne and colleagues21 have suggested that pos¬
tion and amplitude. Bizzo and associates64 deter¬
tural movements are maintained by continuous
mined that the whole frequency range in stabi-
muscular action, and not as a simple pendular
lometry was 0 to 10 Hz. Previous studies17- 63
motion with passive elasticity. Thus, the existence
reported swaying frequencies exceeding the upper
of continuous muscle action with relatively small
limit of this range. Altogether, three typical fre¬
forces required in the ankle, as well as the other
quency groups were identified and reported: (1)
rotating joints, makes it evident that a multiseg-
above 5 Hz17'19- 27- 43-45'63; (2) 0 5 to 5 Hz6-13> 14-19'
20, 27. 29. 63, 65. (3) bdow (£5 Hz >2, 17, 21, 27, 45 mental model is more adequate for describing
body sway. This suggested modification of the
frequency groups were found in both kinematic
body sway model also sheds light on the regula¬
(swaying) and dynamic (forceplate data) tests. The
tion mechanism of postural sway, through the de¬
existence of the high-frequency group was dissoci¬
tailed action of the leg muscles.
ated from any passive elasticity effects and attrib¬
Figure 8—2 illustrates the main muscle groups
uted instead to muscle tremor. According to
that are potentially involved in maintaining equi¬
Thomas and Whitney,17 this tremor originates from
librium during standing still. Under the assump¬
the asynchronous firing of individual motor units,
which accompanies all muscular activity. tion of a locked-knee position, the muscle groups
In most of the previously mentioned investiga¬ shown in the part A of this figure act around
tions, not all the frequency groups were studied. the ankle and hip joints. A three-segment model
The reason could be either that the investigators representation of the swaying body in the sagittal
have put their emphasis on the more easily detect¬ plane is presented in part B. In this model, the
able frequency group or that, owing to their own tibialis anterior and gluteus maximus muscles act
methodology, some frequencies simply could not as torque generators around the ankle and hip
be identified. For example, if the swaying data are joints, respectively. Note that the more simplistic
sampled at the frequency of 5 Hz,21 no oscillation inverted pendulum model6-66 treats the body in
frequencies above 2.5 Hz can be detected. On the terms of two rigid segments, the foot and the
other hand, a power spectrum analysis of force remainder above it, with a purely elastic recoil at
traces sampled at 50 Hz has demonstrated the the ankle-joint level.
above frequency groups: two rather broad ranges Several factors indicate that the mechanism in¬
of local peaks were detected at the two lower volving the regulatory muscle action may act inde¬
frequencies, that is, about 0.1 and 1 Hz, and a pendently in the left and right legs. These factors
sharp peak was detected at the higher frequency, are listed as follows:
corresponding to about 7 Hz.27
1. The existence of relative motion between
the body segments, which contributes to
Modeling of Swaying Motion body corrections,13 contradicts the
previously mentioned bisegmental model.
Relating to modeling of postural sway, the in¬ The amplitudes of these compensatory body
verted pendulum bisegmental model has gained a movements, which are minimized by the
significant amount of attention, particularly in the influence of visual information,36 are related
sagittal plane motion. In this motion, the body is to the position of each leg. It was further
considered rigid and hinged above the ankle joint. suggested that the relative and
This representation complies with the conception
compensatory movements occur in such a
of using one force platform only for measuring
way that the ground reaction forces are not
the reactive forces.
predictably invoked.19
• I Chapter 8 BIOMECHANICS OF UNPERTURBED STANDING BALANCE 193
FIGURE 8-2 Simple modeling of the human body while standing still in the sagittal plane. A, The main
muscle groups involved in maintaining equilibrium during standing with locked knees. 8, The tibialis anterior
and gluteus maximus muscles are shown to act as torque generators around the ankle and hip, respectively,
in a three-segment model. Joint forces and torques are denoted as F, and T,, respectively, for each joint.
Reactive foot—ground forces are not shown. HAT, head and arms and trunk; Fgm» force in the gluteus
maximus; FTA, force in the tibialis anterior.
2. In addition to the vestibular organs, the the specific activity of forces on each leg is re¬
mere rotation of the ankles was pointed out quired, each of the supporting limbs should be
as the major source of information of the represented as a segment. It is obvious that pelvic
body’s orientation.29-50 In comparison to the attachment of the lower limbs requires that each
vestibular organs, ankle rotation, which is of these limb segments be connected to an upper
segment, representing the trunk of the body. The
based on the stretch and long-loop reflexes
practical implication of a multi segmental model is
of the plantar-flexing posterior muscles,63
that measurement of body sway should be made
was reported to be more significant in the
by means of two force platforms, one for each
higher sway frequencies.50-67
leg, and that the force data obtained from such a
3. The source of existence of the high-
system should be used as input values for this
frequency sway (6 Hz) is associated with
model.
muscle tremor,17 which possibly originates
A five-segment model of the human body for
from the asynchronous firing of the
bipedal standing-still postural sway is shown in
individual motor units and which
Figure 8-3. The reaction forces FL and FR are
accompanies all muscular activity. This
measured bilaterally from both feet, using two
may indicate that each leg acts
force platforms. A block diagram of the iterative
independently with an individual postural
calculation procedure16 for this model is shown
activity pattern. in Figure 8-4. Apart from the forceplate data,
It is evident that, to relate the externally mea¬ information of the individual anthropometry of the
sured foot-ground reaction forces to the mecha¬ tested subject is essential. However, a thorough
nisms of stabilization and posture regulation, data evaluation of the kinematics and dynamics of the
collected from a single force platform on which model is bounded by the uncertainty of the exact
the subject stands, thus treating the swaying body positioning of the joints in space and by the an¬
as a single-supported structure, cannot provide thropometry of the subject. Consequently, the
sufficiently detailed information on the individual model is constrained by including a condition that
activity of each of the supporting legs. Such infor¬ compares model results to actually measured data.
mation is essential to understand the mechanism The first step in the calculation procedure is the
involved in normal postural sway and, even more estimation of the trajectory of the center of gravity,
so, to describe and evaluate postural deficiencies as evaluated from forceplate measurements.15
Kinematics of the segments are then estimated
existing in pathologic cases68-69 (e.g., center or
from the center of gravity trajectory, followed by
gravity trajectory; see Fig. 8-1). In particular, if
METHODOLOGY OF NATURAL SWAY

MEASUREMENTS
Bilateral measurements of the natural swaying

motion in standing are made by means of two
collaterally installed force platforms for the adja¬
cent positioning of the left and right feet during
standing. The tested subject is requested to stand
still during the test, and the measured quantities
include the force components and the coordinates
of the center of pressure. A distinction should be
made in the vertical and mediolateral directions
between the level of forces (dc component) and
the force oscillations around this level (ac compo¬
nent). The force oscillations is the component of
interest in studying the dynamics of body sway. A
strong correlation was found between the amount
of opening between the feet and the level of medi-
solution of the forces and moments acting on the

distal segments of the model by means of inverse
dynamics methods. Differences between model
and measured forces of the right foot are mini¬
mized by readjusting the positioning of the joints
and the anthropometric properties of the segments.
After minimizing the calculation errors, the itera¬
tion procedure of the kinematics and dynamics of
the segments is concluded.
Bilateral Measurements of
Natural Body Sway
It has been argued that in using only one force

platform for monitoring postural sway, the lateral
forces on both feet are vectorially summed, re¬
sulting in a small net lateral force that is within
the noise level of the measurements.27 In using
two force platforms, the force components and
centers of pressure on each foot can be measured
separately. As discussed in the previous section,
this is essential if standing sway is to be modeled
FlGURE 8-4 Block diagram of the iterative calculation
as a multi segmental model,70’71 which accounts for
procedure of the five-segment model presented in Figure
possible asymmetries. The significance of bilateral 8-3. The first step is the estimation of the trajectory of the
forceplate measurements in postural sway has center of gravity (CG). as evaluated from forceplate
been reported for able-bodied populations27 as measurements. Kinematics of the segments are thereafter
well as for populations with neurologic72-76 or or¬ estimated, from which forces and moments on the distal
thopedic77- 78 pathologies. segments are resolved using inverse dynamics. Differences
It is expected, for example, that in CVA pa¬ between model and measured forces are minimized for the
tients, because muscular activity may differ con¬ right foot by readjusting the positioning of the joints and
siderably from one side of the body to the other, anthropometric properties of the segments. After minimizing
the calculation errors, the iteration procedure of the
information on the forces acting on each of the
kinematics and dynamics of the segments is concluded. FL,
legs in maintaining equilibrium would be of
left reaction force; FR, right reaction force; £, error of
great interest.
estimation; moment about the ith axis.
olateral forces measured, implying that spacing anteroposterior, mediolateral, and vertical force
between the feet should be carefully controlled.27 components, the following parameters were de¬
The duration of the measuring tests is typically fined:
30 to 60 seconds, of which the central two thirds 1. Relative sequence of the tangential force
are considered for data processing. This duration vectors (RSTFV). As a result of body sway,
is of the same order of magnitude taken in most the horizontal component of the
previous investigations.10'18-20,24’31,32 foot-ground reaction vector obtained for
Early works dealing with bilateral measurement each leg was found to oscillate periodically
of sway were confined to studying weight-bearing and diagonally backward and forward along
imbalance between the legs, that is, through the a given direction. It was also found that the
vertical component of the foot-ground reaction horizontal components of the foot-ground
forces. Using foot-ground pressure measurements, reaction vectors of both feet generally have
a striking asymmetry of 75% in weight bearing the same oscillation frequency, although
between the feet in post-CVA patients was re¬ they may appear in various relative
ported.44 Similar findings were also reported when sequences.27'73,74 For instance, the
using two force platforms and when concentrating anteroposterior force components on both
on the magnitude of the vertical force component legs can be either concurrent or opposite to
only and the coordinate of its center of pressure.45 each other. The same is true for the
In subsequent studies, bilateral force measure¬ mediolateral components. Thus, different
ments on the supporting limbs were made to pro¬ configurations may occur, all resulting in
vide a new representation of postural sway, which time-synchronized diagonal oscillations of
was implemented to evaluate able-bodied subjects the tangential reactive force on both feet.
as well as patients suffering from disorders in the This synchronous appearance discloses the
musculoskeletal and neurologic systems.72-78 bilateral dynamics of the reactive forces
Parameters, including frequency amplitude involved in maintaining equilibrium during
force sequence between the feet, force activity on standing. The RSTFV can be established
each foot, and asymmetry, were defined, pre¬ either by visually inspecting the relative
sented, and compared between the different course of the force traces in the time
groups of subjects studied. domain or by using the following
computerized procedure: The tangential
force components in the x and y directions
DEFINED PARAMETERS of the two legs are plotted one against the
other in different combinations, disclosing
The force traces obtained included a transient, the correlations existing between them.
slow oscillation (about 0.1 Hz), above which more From these correlations, the RSTFV can be
rapid oscillations (1 Hz and higher) were superim¬ resolved, as demonstrated in the Poincare
posed. Tremor oscillations (about 6 Hz) also pre¬ plots shown in Figure 8-6. Previous results
sented, but with much smaller amplitudes (Fig. indicate that the RSTFV has a common
8-5). a procedure was established to measure pattern in most able-bodied subjects
the peak-to-peak amplitudes of the middle (1 Hz) tested.27 Typical deviations from this pattern
oscillations and to compute their averages.27-77 are found in defined pathologies, such as in
From these amplitude averages obtained for the CCI and CVA patients.

demonstration of force oscillations in
standing sway. A frequency analysis
discloses the existence of three major
frequency ranges of about 0.1, I, and
6 Hz.
FIGURE 8-6 Poincare plots of the

FXR US FXL FXR US FYR FXR US FYL FXL US FYL force components illustrating the
procedure to establish the relative
i sequence of the tangential force
A A X vectors (RSTFV). A, The tangential
Y force components in the x and y
Y L R directions of the two legs plotted
against one another in different
combinations. B, Resolution of the
N
/ RSTFV from the correlations between
X X = anteroposterior the components. C, Definition of the
' Y = mediolateral axes for the two forceplates used.
R = right VS, versus; FX, force in the
anteroposterior direction; FY, force in
L = left
the mediolateral direction; R, right; L,
B C left.
2. Weight-bearing imbalance (WBI), in the ESTIMATION OF THE TRAJECTORY OF THE

vertical direction, is defined as follows: CENTER OF GRAVITY FROM BILATERAL
WBI = absolute value of vertical force REACTIVE FORCES
difference between the feet normalized to
body weight. Thus, WBI expresses the An iterative model for the estimation of the center
difference between the average vertical of gravity trajectory by using foot-ground reaction
forces supported by each of the legs. force and center of pressure data was developed.15
3. Total sway activity is defined in the The iterative procedure was used to examine pos¬
horizontal plane as the added force sible effects of the rate of change of angular
amplitude averages in the anteroposterior momentum on the estimated trajectory. It was con¬
and mediolateral directions (Fig. 8-7). This cluded that the contribution of the time-rate
quantity combines the force activity of both change of the angular momentum is negligible in
legs and may therefore be directly related the iteration procedure of estimating the center of
to the added muscle activity of both legs gravity trajectory. The method was illustrated on
and represents the accumulated efforts
invested in maintaining standing
equilibrium. Note that single-platform
measurements would not enable us to ASYM
obtain such an integrated measure because
the measured forces in this case represent
the net external force acting on the body.
4. Asymmetry, also defined in the horizontal
plane, is the subtracted force amplitude
averages in the anteroposterior and
mediolateral directions (see Fig. 8-7). This
quantity is related to the difference in
activity existing between both legs, and in
the presence of ideal symmetry it should
have a zero value. Asymmetry thus
indicates the measure in which the force xl+xr
activity required during standing is shared
FIGURE 8-7 Vectorial definition of total sway activity (SA)
between the two legs. and asymmetry (ASYM).
a group of 11 able-bodied subjects and 2 subjects studied. The average age mass and height were 41
with musculoskeletal pathologies. years, 66 kg, and 168 cm, respectively.
Symmetric standing can be defined in terms of
the forces applied by each of the feet or by the
FREQUENCIES OF SWAYING MOTION
location of the average position of the center of
gravity trajectory. Under the condition of symmet¬
In all the subjects tested, the force traces corres¬
ric positioning of the feet on the forceplates, the
ponding to both feet were synchronous with each
separation line between the forceplates can be
other and thus had the same basic frequencies.
used as a reference for the average position of the
These frequencies might, however, differ from test
center of gravity trajectory. The latter was found
to test, as they might in the presence of (as op¬
to be considerably displaced from the separation
posed to in the absence of) visual feedback. Al¬
line for the two pathologic subjects, whereby the
though the cycles of the periodic swaying motion
displacement was found to be significantly large
are irregular in frequency and amplitude, primary
(P < .005) compared with that of the able-bodied
waves are easily identifiable. Superimposed on
subjects. Similar results on the effect of pathologic
these waves are secondary smaller waves of
disorders on the shift of the center of gravity
higher frequencies. The higher-frequency waves
trajectory were reported by Snijders and Verduin.8
(middle- and high-frequency oscillations) were re¬
Nevertheless, in this study, the data were related
ported to be of greater regularity with regard to
to a single forceplate.
both cycle duration and amplitude.17,63 This has
A different reported methodology for the pre¬
also been demonstrated in the power spectrum
diction of the center of gravity from forceplate
analysis of the traces.27
measurements during gait is based on a Fourier
Average values of the waveform frequencies for
representation of the acceleration components of
all the subjects studied are summarized in Figure
the center of gravity, whereby the fundamental
8-8. The orders of magnitude frequencies for the
harmonic of the Fourier series is one cycle of
oscillations detected in this study were 6 Hz for
gait.79 The problem in applying such a methodol¬
the rapid oscillation, 1.2 Hz for the middle oscilla¬
ogy in standing still is the difficulty of identifying
tion, and 0.1 Hz for the slow oscillation. Because
one basic harmonic of the standing-still posture
of their regularity, the middle-frequency oscilla¬
sway. As previously mentioned, spectral analysis
tions (on the order of 1.2 Hz) were the most
of forceplate data has in the past identified at least
easily measurable and were, in fact, the only ones
three main frequency ranges.27
referred to by several investigators.19,20
In the literature, the frequency values reported
for postural sway vary considerably. Seliktar and
Balance in Able-Bodied Subjects colleagues45 reported on three distinct ranges
around the following frequencies: 5, 0.5, and 0.01
SUBJECTS TESTED Hz. Lakes and associates32 reported on two fre¬
quency bands of 2.9 and 0.56 Hz, and Thomas
The postural balance in quiet standing of 23 adult and Whitney17 found two frequency components:
able-bodied subjects (13 men and 10 women) was the high frequency around 10 Hz and the low one
RAPID MIDDLE SLOW
n
N
N I ■ Closed
£ 4
>>
o □ Open
FIGURE 8-8 Average frequencies of c
0)
oscillations in standing sway in the 3
cr
three major frequency ranges for three 0) 2
population groups: able-bodied
(normal), cerebrovascular accident
(CVA), and craniocerebral accident
(CCI) patients. The vertical bars denote
Normal CVA CCI
I standard deviation.
around 0.2 Hz. In other studies, however, only cases analyzed. The vectors in this pattern are
one frequency was reported, presumably that concurrent and run somewhat parallel to each
which was the most easily measurable or which other from the left anterior to the right posterior
was judged to be the most significant. direction of the body.
Smith,63 Thomas and Whitney,17 and Kapteyn24 The vector diagrams presented in Table 8-1
emphasized in their reports the 10-Hz frequency. shed light on the force patterns acting on the feet
Other authors19'20 focused mainly on the 1-Hz- and their relative sequences. Each vector schemat¬
frequency oscillation. The effect of visual feed¬ ically describes the direction of the horizontal
back on the waveform frequencies indicates a reactive force acting on the foot. The synchronous
slight increase in the absence of visual feedback. appearance of the vectors, as described separately
Particularly, the frequency of the middle oscilla¬ for each foot, thus discloses the dynamics of the
tion increased by about 7% (from 1.21 to 1.29 Hz) reactive forces involved in maintaining equilib¬
when the visual feedback was eliminated (eyes rium during standing. The diagonal patterns, with
covered, as compared with eyes open). diverging vectors from the center of the body to
the front of the body, were found to be most
RELATIVE SEQUENCE OF THE TANGENTIAL frequent, indicating that while the anteroposterior
FORCE VECTORS forces run concurrently, the mediolateral forces
normally run in opposite directions on both feet.
The recordings obtained indicate that the antero¬
Despite the observed different activities on the
posterior forces on both feet were always concur¬
legs, however, the fact that the tracings on both
rent with each other. The mediolateral forces,
feet were synchronous with each other tends to
however, were in most cases opposite to each
indicate that the muscle forces in both legs are
other.
controlled by one central source.
The RSTFV on both feet is summarized in
Table 8—1. Three vectorial patterns were found
AMPLITUDES
in the normal population group tested. The most
frequent pattern is shown in the first column and Amplitudes of the medium oscillations are ex¬
was found in 77% of the cases analyzed. In this pressed separately for the dominant and opposite
pattern, also shown in Figure 8-6, the vectors run legs. Average values of the amplitudes for the
synchronously, diverging diagonally from each normal population, expressed in terms of body
other in the front side of the body and converging weight, are given in Figure 8-9. The magnitudes
in the rear side of the body. In 18% of the cases on both legs were comparable. Also, the differ¬
analyzed (column 2), the pattern was somewhat ence in the presence or absence of visual feedback
similar, except that the intersecting point was was not significant.
shifted toward the front side of the body. It is also In the vertical (z) direction, weight-bearing im¬
noted that the roles of the left and right feet are balance was 5.4% and 2.1% of body weight, with
interchanged in these two patterns. It may be of and without visual feedback, respectively (Fig.
interest to note that all cases having the second 8-10). Oscillations of the vertical forces in the
pattern involved women. The least frequent pat¬ same condition amounted to 1.5% of the body
tern among the healthy subjects tested is shown weight. Here, again, the effect of visual feedback
in column 3 and was noted in only 5% of the was not significant (see Fig. 8-9).
Table 8-1 Relative Sequence of the Tangential Force Vectors (RSTFV) on Both Feet*
Healthy subjects 77 18.0 5.00

Subjects with
cerebrovascular
accident 29.4 1.6 9.7 25.2 5.10 19.2 0.3 0.50
Subjects with
craniocerebral
injury 62.0 2.9 5.8 5.7 14.7 8.8
*The figures represent percentage occurrence from the recordings obtained for the populations tested.
Chapter 8 BIOMECHANICS OF UNPERTURBED STANDING BALANCE 199
0.7 H Closed left -II-

0 Open right (sound for CVA)
■ B Closed right -II-
0.6
sz
cp
0) 0.5
>>
■D
O 0.4
CD 0.3
O
0.2 -
0.1 ■
0-
Normal CCI Normal CVA CCI
-1-fy-
.c
g>
CD
>
T3
O
jQ
FIGURE 8-9 Average values of forces in standing CD

O
i_
sway for three population groups: normal, O
LL
cerebrovascular accident (CVA), and craniocerebral
accident (CCI) patients. The forces, expressed in
percentage body weight, are presented in three
components: anteroposterior (Fx), mediolateral (FY),
and vertical (Fz). The vertical bars denote I standard
deviation.
The values of total sway activity and asymme¬ 1. The oscillations of the forces in the
try, as defined earlier, are given in Figure 8-11. tangential plane are approximately 0.2% of
The total sway activity values were about 0.5% of body weight, corresponding to about 1.2 N
body weight. Asymmetry values were lower than on each leg.
0.1% of body weight, indicating that for the nor¬ 2. The oscillations of the forces in the
mal population, this parameter was negligibly low. anteroposterior (x) direction are slightly
The most common order of magnitude found in higher than those in the mediolateral (y)
the literature for the force amplitudes was about 1 direction.
N.17-19 Shimba20 reported on different force magni¬ 3. In the vertical (z) direction, the oscillations
tudes in the various directions measured: 0.5 N of the forces were of a much higher
in the anteroposterior direction, 0.35 N in the amplitude than in the horizontal directions
mediolateral direction, and 1.18 N in the vertical (1.57% of body weight, corresponding to
direction. Note that these values correspond to the about 9.35 N).
resultant forces acting on both feet together. Our 4. Absence of visual feedback has no
results on the amplitudes of the middle frequency, significant effect on the forces measured,
presented separately for the dominant and the op¬ except in the anteroposterior (x) direction,
posite legs, indicate the following: where a slight increase was noticed.
WBI been discussed in the literature. Thomas and Whit¬

70 ney17 refer to the high-frequency tremor during
sway to criticize the inverted pendulum model,
which is based on passive elastic terms only, and
60 □ Open
to justify the existence of continuous muscular
1 Closed
action. Smith63 states further that the stabilization
50
of the ankle joint during standing is due to two
forces: the first originating from the passive tis¬
sues resisting dorsiflexion, and the second due
JJ
'5 40
5 to the posterior muscles that are responsible for
dorsiflexion of the joint. In this context, the sig¬
m 30 nificance of sway activity and asymmetry parame¬
vP
0s ters, as seen in Figure 8-11, is easily understood.
Beyond the force magnitudes reported separately
20
for each leg in Figure 8-9, sway activity repre¬
sents the vector summation of the absolute values
10 of these forces and can therefore be related to the
overall muscle activity of both legs. Typical values
0 of sway activity were found to be 3 N, correspond¬
Normal CVA CCI BKA ing to 0.5% body weight. Asymmetry is related to
FIGURE 8-10 Average values of weight-bearing imbalance the difference in activity existing between both
(WBI), expressed in percentage body weight, in standing sway legs, and in the presence of ideal symmetry, this
for four population groups: normal, cerebrovascular accident parameter should be zero. Asymmetry in the
(CVA), craniocerebral accident (CCI), and below-knee healthy subjects tested in this study was about 0.4
amputation (BKA) patients. The vertical bars denote I N (0.07% body weight).
standard deviation.
Balance in Patients With

To compare these results with those found in Cerebrovascular Accident
the literature, vector summations were done and
the resultants further evaluated. It was concluded PATIENTS TESTED
that the results presented by the single-support
method are erroneous if the forces on each leg are The postural balance in quiet standing was studied
to be known. The relationship between postural in 16 CVA patients. The average age was 60.9
sway and muscular activity around the joint has ± 10.5 years, and the average time from stroke to
O)
§
>.
"D
2
O
CO
1
FIGURE 8-1 1 Average values of total sway
activity (SA) and asymmetry (ASYM), expressed in
percentage body weight, for four population
groups: normal, cerebrovascular accident (CVA),
craniocerebral accident (CCI), and below-knee
Normal CVA CCI BKA amputation (BKA) patients. The vertical bars
denote I standard deviation.
first test was 11.8 ±8.6 weeks. Six able-bodied ably higher than those found in the control group.
subjects from the group of healthy subjects (aver¬ The difference between the forces on the sound
age age, 64.6 ±8.1 years) served as controls. Sex leg in the hemiplegic population and those on any
and age differences were checked between the leg in the control population were statistically
patients and the control groups using the chi- significant (P < .01) in the mediolateral direction,
square test, and the results showed that no adjust¬ and even more so in the anteroposterior direction.
ment was required for age or sex. The force differences between the legs within the
hemiplegic population were significant (P < •01);
WAVEFORM FREQUENCIES however, this was not the case within the control
population. It may be seen from Figure 8-9 that
Average values of waveform frequencies for the absence of visual feedback considerably increased
CVA patients are given in Figure 8-8. The average the force amplitudes in the hemiplegic group;
frequencies of the medial and rapid oscillations however, this increase was not statistically sig¬
were in the same range as in the control popula¬ nificant. Also, in the vertical (z) direction, ampli¬
tion. On the other hand, slow oscillations were of tudes were higher in the hemiplegic compared
a considerably higher frequency in the hemiplegic with the healthy group for both the weight-bearing
population. The frequencies tended to be slightly imbalance (see Fig. 8-10) and the force oscilla¬
higher in the absence of visual feedback than tions (see Fig. 8-9). The effect of absence of
when visual feedback was present. visual feedback was similar in this direction as in
the previously described horizontal plane, that is,
RELATIVE SEQUENCE OF THE TANGENTIAL an increase in amplitude in the hemiplegic popula¬
FORCE VECTORS tion.
Values of total sway activity and asymmetry
The RSTFV on both feet is summarized in Table are given in Figure 8-11. Both parameters had
8-1. Of the various patterns found in the CVA significantly higher values in the hemiplegic group
population, the diagonal pattern corresponding to than in the control group (P < .04 and P < .001,
concurrent anteroposterior forces, but opposite to respectively), indicating that the former group re¬
mediolateral forces on both legs (column 1), was quires a higher muscular activity for maintaining
found the most frequently, in more than 29% of postural equilibrium.
the tests. Both of the next two common patterns in It is obvious from the results presented in Fig¬
the CVA population had concurrent mediolateral ure 8-11 that the sound leg’s contribution to this
forces. The anteroposterior components were, process is much greater than that of the plegic leg.
however, opposite in one of these patterns (25.2% Asymmetry was defined to express the difference
of the tests, column 4) and concurrent in the in activities between both legs. Although in the
second pattern (19.2% of the tests, column 6). control subjects tested, this parameter was very
Other patterns, with various degrees of occur¬ nearly zero, it amounted in the hemiplegic popula¬
rence, are shown in Table 8-1. The presence or tion to 0.27% of the body weight. Absence of
absence of visual feedback did not affect the visual feedback resulted in an increase in those
RSTFV. two parameters. However, the only statistically
These results show that the patterns in the hemi¬ significant difference was the increase in asymme¬
plegic population are different from those in the try, occurring in the absence of, as opposed to the
control subjects. The most frequent pattern in the presence of, visual feedback.
control population was found in only 29.4% of The above results on the amplitudes of the
the CVA patients. Other patterns, in which the middle frequency, presented separately for the
mediolateral and anteroposterior forces are con¬ plegic and sound legs in the CVA population, can
current, were also frequent (see columns 3 to 6 in be summarized as follows:
Table 8-1) and can be considered equivalent.
When grouping them together, an occurrence in 1. The forces were significantly higher than
about 29% of the tests is obtained, approaching those of the control group (P < .01),
the occurrence of the most frequent pattern. amounting to double the value in the
anteroposterior component on the sound
leg.
AMPLITUDES
2. The force oscillation during sway was
Amplitudes of the medial frequency oscillations higher in the anteroposterior (x) direction
are expressed separately for the sound and plegic than in the mediolateral direction (y)
legs in the hemiplegic patients (see Fig. 8-9). The (supporting previous findings20). Although
amplitudes in the hemiplegic group were invari¬ in the healthy subjects this difference was
minor, in the hemiplegic group it was and asymmetry values, which tended to be clearly
considerably higher, especially in the forces reduced during rehabilitation.
on the sound leg.
3. The force magnitudes were always higher
in the sound leg than in the plegic leg (P < Balance in Patients With
.01). The most significant difference was Craniocerebral Injury
found in the anteroposterior (x) directions
both with the eyes open and with them PATIENTS TESTED
closed. The increased force magnitudes on
the sound leg in comparison with the plegic Ten CCI patients (9 men and 1 woman), aged 19
leg points to an increased activity of the to 43 years (28.5 ± 10.3 years), with no previous
sound let to compensate for the relative history of neurologic illness, drug abuse, or alco¬
inactivity of the plegic leg, to regain the holism, took part in the postural balance tests.74
ability to maintain the standing position. Eight patients suffered from blunt head injuries
Remember than in the control subjects, no due to road accidents, and two patients suffered
differences were found between the legs. from penetrating lesions. All the patients were in
4. In the vertical (z) direction, the force a coma for at least 24 hours (range, 1 to 20
oscillations were of much higher amplitudes days; mean, 7.1 ± 4.9 days). Neurobehavioral
than in the horizontal directions. Both the disturbances were evaluated by previously de¬
amplitude oscillations and the weight¬ scribed methods.80-81
bearing imbalance were considerably higher Eleven healthy subjects from the group of nor¬
in the hemiplegic compared with the mal subjects (average age, 29.9 ± 6.6 years)
control population. served as controls. All the subjects included were
5. Although absence of visual feedback did free from neurologic, orthopedic, vestibular, and
not have noticeable effects in the control visual impairment and had no recent history of
subjects, it appeared to increase the skeletal or muscular injury. The two above groups
amplitudes of force oscillations in the were found to be similar regarding sex and age
hemiplegic group. When comparing the (t = —0.36; P — .7), therefore requiring no
effect of eye closure for both groups adjustment for these factors.
separately, the only significant parameter
was asymmetry in the normal population. WAVEFORM FREQUENCIES
Eye closure did not cause any statistically
significant effect in the group of Average values of waveform frequencies in the
hemiplegics, which may suggest that this CCI patients are presented in Figure 8-8. The
population does not extract helpful average frequencies of both the low- and high-
information from the visual input in the frequency oscillations were lower in the CCI pa¬
process of maintaining equilibrium during tients than in the control subjects. In the middle-
standing. Moreover, because the more frequency oscillations, the average was higher in
significant differences between the groups the CCI group. The differences in frequencies
were obtained with the eyes open rather between the two population groups, however, were
than closed, it may be concluded that in not statistically significant. The effect of visual
hemiplegics, visual input may provide feedback on the oscillation frequencies tended to
conflicting information from the balance be altogether minor in the CCI patients, as it was
point of view. also for the control subjects.
TIME EFFECT AND LOCOMOTOR OUTCOME RELATIVE SEQUENCE OF THE TANGENTIAL

FORCE VECTORS
No single parameter or array of parameters of a The RSTFV on both feet is summarized in Table
single dimension—clinical, physiologic, or 8-1. The most frequent pattern was the diagonal
biomechanical—was found to predict the locomo¬ pattern corresponding to concurrent anteroposter¬
tor outcome of a patient.75-76 Nevertheless, some ior (x) forces but opposite mediolateral (y) forces
significant findings were reported for the patients on both legs (62%). The next most common pat¬
with the better locomotor outcome. These in¬ tern in the CCI group consisted of opposite forces
cluded (1) the RSTFV patterns, which became in both the anteroposterior and mediolateral direc¬
remarkably similar to those of healthy people tions on the two legs in relation to each other (8.8
without hemiplegia; and (2) the total sway activity +14.7 = 23.5%; see columns 7 and 8 in Table
8-1). None of the control subjects exhibited this closed that their sway disturbances, as revealed by
pattern. The pattern corresponding to concurrent the magnitudes of total sway activity and asymme¬
forces in both the anteroposterior and mediolateral try, tended to be related to the severity of neurobe-
directions on the two legs had an occurrence of 2.9 havioral disturbances,74 the correlation coefficients
+ 5.7 = 8.6% in the CCI patients (see columns 3 being 0.97 and 0.60, respectively, for sway activity
and 6 in Table 8-1). The least frequent pattern in and asymmetry.
the CCI group, with concurrent mediolateral but
opposite anteroposterior forces, is shown in col¬
umn 5. Presence or absence of visual feedback or Balance in Patients With Below-
of a curtain drawn between the patient and site of Knee Amputation
measurement did not affect the RSTFV found for
any patient. THE QUESTION OF PROPRIOCEPTIVE
DEFICIT
AMPLITUDES
Published results about the standing stability in
Our results on the amplitudes of the middle fre¬ people who have had amputations are controver¬
quency are presented in Figure 8-9 and indicate sial. An increased postural sway in people with
the following: above-knee and below-knee amputations was
demonstrated.82 On the other hand, it has been
1. The forces were significantly higher in the
stated that sway in people with below-knee ampu¬
CCI patients than in the control subjects (P
tation is comparable to that of a matched control
< .05), amounting to double the value.
group.83 Additional questions about standing sta¬
2. The force oscillation during sway was
bility of BKA patients relate to the effects of
higher in the anteroposterior (x) direction visual feedback and of compensation and adapta¬
than in the mediolateral direction (y), tion on improved stability in these subjects. It has
supporting previous findings.20 been assumed that in subjects with BKA, unlike
3. The differences between left and right legs in able-bodied subjects, when the visual and ves¬
within each of the two groups tested were tibular systems are intact, the main cause for inse¬
not statistically significant. The high cure standing is a quantitative decrease in proprio¬
asymmetry values obtained for the CCI ceptive inflow due to the loss of foot and leg
patients (see Fig. 8-11) indicate, however, muscles. The proprioexteroceptive information is
that within this group, at the individual transmitted by different receptors localized in the
level, differences are found between the skin, subcutis, joint capsule, ligaments, tendons,
forces on both legs. and muscles.
4. The force oscillations were much higher in Intense responses to afferent inputs arising
the vertical (z) direction than in the within the environment are mediated primarily by
horizontal directions. Both the amplitude exteroceptive inputs.5 Muscular responses re¬
oscillations (see Fig. 8-9) and the weight¬ sulting from proprioceptive inputs are mild. There¬
bearing imbalance (see Fig. 8-10) were fore, the muscle spindles, which are highly sensi¬
considerably higher in the CCI group than tive proprioceptors, evoke the stretch reflex and a
in the control group. muscular response even after small disturbances.
5. Absence of visual feedback caused no As a result, postural instabilities should increase
noticeable effects. only when inputs from muscle spindle afferents
Values of the sway activity parameter were sig¬ are eliminated. Nevertheless, published results re¬
nificantly higher in the CCI population studied garding the relationship between visual inflow and
than in control subjects (P < .02), indicating that postural stability are still controversial. As far as
the former group requires a higher muscular activ¬ we know, no investigations have been performed
ity for maintaining postural equilibrium. Asymme¬ to test whether lower limb amputation results in
try amounted in the CCI population to 0.15% proprioceptive deficit. In fact, it has been found
body weight, nearly double the value for the con¬ that amputation above the knee affects postural
trol subjects. sway only when the subject’s eyes are closed.82
Surprisingly, it has also been found that the pos¬
tural sway of people with BKA was significantly
CORRELATION BETWEEN MECHANICAL
greater than it was for people with above-knee
AND CLINICAL PARAMETERS
amputations when the patients’ eyes were either
An attempt to grade the patients according to open or closed. Bilateral measurements were thus
ascending magnitude of neurologic deficits dis¬ undertaken to address these questions.
SUBJECTS found between closed eyes and open eyes when

comparing subjects with BKA with the able-bod¬
Eleven men who had BKA, with an average age ied controls, that the influence (or lack of influ¬
of 64.8 ± 9.2 years (range, 49 to 80 years), were ence) of vision is similar in both groups tested.
included in this study. The level of amputation To address the issue of the role of proprioception
was below the left knee in seven men and below deficit in subjects with BKA, the differences ob¬
the right knee in four men. In all patients, amputa¬ tained from tests of subjects in both groups with
tion was performed because of gangrene second¬ closed eyes and open eyes were compared. These
ary to diabetes mellitus. Nine able-bodied men results show that the differences were significantly
from the group of control subjects (mean age, higher in subjects with BKA for the sway activity
65.6 ± 8.6 years) served as controls. Patients or and asymmetry parameters in the late testing pe¬
subjects with partial blindness, neurologic prob¬ riod. The difference was less pronounced for sway
lems, or any disease known to affect equilibrium activity in subjects with BKA in the early testing
were excluded from the study. To establish integ¬ period. These differences have no effect on WBI.
rity of equilibrium in the test subjects, a prelimi¬ Therefore, the obvious difference in values ob¬
nary Romberg test82 was conducted. Those who tained from tests of subjects with BKA with eyes
failed to maintain equilibrium during the test were open and eyes closed indicates that they are less
excluded from the study. stable and less secure, probably owing to proprio¬
Subjects with BKA were tested after exercising ceptive deficit. This is further emphasized by the
with their new prostheses. The first test was done fact that the differences remain in people with
1 to 2 days after the subjects with BKA received BKA throughout rehabilitation.
their prostheses. They were tested again 3 to 4 It has also been demonstrated that in people
weeks later on completion of their prosthetic reha¬ who have recently experienced BKA, there is a
bilitation program.
continuous process of compensation and adapta¬
tion to the new situation of a partial limb loss. In
AMPLITUDES OF SWAY fact, with prosthesis use for 3 to 4 weeks, signifi¬
cant improvement was seen in sway activity (eyes
Values of sway activity, asymmetry, and weight¬ open) and WBI (with eyes closed). Other parame¬
bearing imbalance (WBI) (expressed in percentage ters also improved, but to a lesser extent.
of body weight) obtained in subjects with BKA In conclusion, subjects with BKA were found
are shown in Figures 8-10 and 8-11. All three to be significantly less stable than able-bodied
quantities were substantially higher in the BKA people during standing. Deficits in proprioceptive
patients than in control subjects: 3.8-, 13.1-, and inflow due to limb amputation contribute even
3.84-fold higher for sway activity, asymmetry, and more to this difference. There is a process of
WBI, respectively. The differences were statisti¬ habituation and compensation achieved by means
cally significant. Results for the second testing of prosthetic training and use. Future studies on
period for the subjects with BKA indicate that subjects with BKA should be conducted to test
improvement was achieved in all three parameters. the effects of longer periods of prosthetic use.
In the second test, performed on completion of
the prosthetic rehabilitation program, there was a
INFLUENCE OF PROSTHESIS ALIGNMENT ON
reduction in sway activity, asymmetry, and WBI
STANDING BALANCE
by 11%, 28%, and 16%, respectively. There was
a significant reduction for sway activity with eyes Alignment of a prosthesis is established by the
open (P = .01) and for WBI with eyes closed (P relative geometric position and orientation of the
= .027). When visual feedback was present, there various prosthetic components, such as the socket,
was a significant decrease in asymmetry in the shaft, joints, and foot. Optimal alignment is a
subjects with BKA, both in the early and in the crucial factor for the successful rehabilitation of
late testing periods. Total sway activity decreased the amputee. No less important are the quality of
in the late testing period only. the fit of the socket, quality of suspension, mass
There was no significant improvement in WBI properties of the prosthesis, and cosmesis.84”87
in either group when the test was performed with Several investigators have attempted to estab¬
eyes open or closed. Comparison between results lish the optimal range of alignment for the BKA
obtained by subtracting values for closed eyes patient, through subjective impression of several
from values for open eyes indicates that the differ¬ prosthetists and direct feedback from the patient.88
ences obtained in the early test and the late test Other approaches to investigate the effect of
were not significant. It may be assumed, in those changing prosthetic alignment on gait have in¬
cases in which no significant differences were cluded indices of symmetry between limbs.89>90 It
was shown that maximum symmetry of kinematic compared with those in the optimal position. In
parameters does not always correspond to a sub¬ the amputated limb, the anteroposterior and me¬
jectively determined optimal alignment. There¬ diolateral forces in varus tilting and the antero¬
fore, gait symmetry should not be the only goal posterior force in valgus tilting were significantly
of the clinician when aligning a prosthesis. higher than in the optimal position. In the contra¬
The standing BKA patient adapts to the prosthe¬ lateral limb, no significant differences were found
sis alignment through accommodation by the con¬ between the optimal and other alignments. A pos¬
tralateral limb. Therefore, use of unilateral mea¬ sible explanation is that an inert prosthesis is inca¬
surement techniques91,92 has the disadvantage of pable of compensating for the missing function of
not considering interaction of the contralateral the joints of an anatomic foot and ankle. This
limb, and monitoring of both limbs simultaneously imposes on the musculature of the contralateral
is essential to determine the combined effect of limb an additional balancing activity in the antero¬
alignment on standing balance. posterior direction. This increased level of activity
Two relevant questions thus arise: (1) What are of the contralateral limb prevails in the optimal
the foot-ground force patterns obtained by BKA alignment as well as in the modified alignments
patients when standing while the prosthesis is op¬ in the anteroposterior direction. The amputated
timally aligned? (2) Do changes in prosthesis limb activity increased significantly, however, in
alignment influence the bilateral foot-ground the anteroposterior and mediolateral directions
force patterns and the standing sway activity? when the prosthesis was tilted into varus and in
To address these questions, the following bilat¬ the anteroposterior direction only when tilted into
eral sway measurements were made. valgus.
Sway activity obtained with optimally aligned
Subjects and Tests prostheses was compared with sway activity in
the different measured alignments. Changing the
We assessed three volunteers, one woman and two prosthesis alignment did not significantly influ¬
men, with traumatic BKA. The sides of amputa¬ ence the sway activity during standing. Asymme¬
tion for the three subjects respectively were right, try and WBI were also verified; neither differed
left, and right: ages 45, 30, and 52 years; masses significantly between the optimal and the other
75, 78, and 90 kg; heights 1.62, 1.80, and 1.80 tested alignment configurations, except for asym¬
meters; and time from amputation 15, 19, and 25 metry in the valgus tilting, which was significantly
years. All were excellent walkers who used their higher in this position. It can therefore be con¬
prostheses on a regular basis and were conducting cluded that the standing, well-trained BKA patient
an active normal family life. is able to adapt to drastic alignment changes while
The test was conducted while the subject was preserving a well-balanced position.
wearing his or her regular prosthesis. All subjects
used a modular patellar-tendon-bearing prosthesis
with belt suspension and a solid ankle-cushioned Summary
heel foot. Because the subjects were excellent
walkers and were satisfied with their prostheses, Postural sway is the result of the minute move¬
the existing prosthesis alignment in each of the ments of the different body segments while stand¬
three amputees tested was considered optimal and ing still. Forceplate measurements are only one
was therefore taken as the reference position. expression of this phenomenon; they can, how¬
The optimal alignment was changed by tilting ever, be very significant if properly made. Mea¬
the pylon in the anterior, posterior, medial, and surement of postural sway, as reported in the liter¬
lateral directions. The maximal tilt in each direc¬ ature, has for many years been based on
tion, obtained by maneuvering the coupler, was monitoring the variations of the center of pressure
9 degrees. of both feet together when standing on one force
platform. Most of the sway studies erroneously
interpreted the trajectory of the center of pressure
Results
to reflect the trajectory of the center of mass of
The results indicate that the anteroposterior force the body. In our work, we have corrected this
differences between the amputated and contralat¬ misconception in the following two respects: (1)
eral limbs were found to be highly significant in we measure the actual forces between foot and
all alignment positions. The measured mediolat- ground, rather than their center of pressure; and
eral force did not differ significantly. (2) we make our measurements bilaterally, that is,
For each leg and each alignment position, the in each leg separately.
anteroposterior and mediolateral forces were also Normal posture by nature has some systemati-
zation, although intricate, allowing quantification from bilateral reactive force measurements in
and delineation of possible patterns of associated standing sway. Gait Posture 4:89-99, 1996.
postural adjustments. In patients with orthopedic 16. Levin O: Dynamic parameters in human standing
or neurologic disorders, motor and sensory deficits posture. M.Sc. Thesis, Technion, Israel Institute of
may influence the resulting swaying mechanisms Technology, Haifa, Israel, March 1994.
17. Thomas DP, Whitney RJ: Postural movements
and balance outcome. Although some of these
during normal standing in man. J Anat (London)
features are difficult to assess and quantify accu¬
93:524-539, 1959.
rately, however, common standing patterns can 18. Stribley RF, Albers JW, Tourtellotte WW, Cockrell
often be identified in the different groups of pa¬ JL: A quantitative study of stance in normal
tients. The possibility of characterizing the sway¬ subjects. Arch Phys Med Rehabil 55:74-80, 1974.
ing features of the various pathologies is of major 19. Soames RW, Atha J: The validity of physique-
importance for both diagnostic and follow-up pur¬ based inverted pendulum models of postural sway
poses. behaviour. Ann Human Biol 7:145-153, 1980.
20. Shimba T: Ground reaction forces during human
Acknowledgment: This study was supported by the standing. Eng Med 12:177-182, 1983.
Segal Foundation, Tel Aviv, Israel. 21. Lestienne F, Soechting J, Berthoz A: Postural
readjustments induced by linear motion of visual
scenes. Exp Brain Res 28:363-384, 1977.
References
22. Walsh EG: Standing man, slow rhythmic tilt,
importance of vision. Agressologie 14C:79-85
1. Basmajian JV: Muscles Alive: Their Functions
1973.
Revealed by Electromyography. Baltimore,
23. Poliak VA, Wyss UP: A simple and inexpensive
technique for the measurement of head sway. J
2. Winter DA: A.B.C. of Balance During Standing
Biomech 16:349-350, 1983.
and Walking. Waterloo, Ontario, Canada, Waterloo
24. Kapteyn TS: Afterthought about the physics and
Biomechanics, 1995.
mechanics of the postural sway. Agressologie
3. Paulus WM, Straube A, Brandt T: Visual
14C:27-35, 1973.
stabilization of posture: Physiological stimulus
25. Gueguen G, Leroux J: Identification d’un modele
characteristics and clinical aspects. Brain
107:1143-1163, 1984. representant les deplacements du centre de gravite
de Fhomme. Agressologie 14C:73-77, 1973.
4. McClure JA: Vertigo and imbalance in the elderly.
26. Lee DN, Lishman JR: Visual proprioceptive
J Otolaryngol 15:248-252, 1985.
control of stance. J Hum Movement Stud 1:87-95
5. Evarts EV: Sherrington’s concept of
1975.
proprioception. Trends Neurosci 4:44-46, 1981.
27. Mizrahi J, Susak Z: Bi-lateral reactive forces
6. Gurfinkel EV: Physical foundations of
patterns in postural sway activity of normal
oscillography. Agressologie 14C:9-13, 1973.
subjects. Biol Cybernet 60:297-305, 1989.
7. Hirashawa Y: Study of human standing ability.
28. Kapteyn TS, Bles W, Njiokiktjien ChJ, et al:
Agressologie 14C:37^13, 1973.
Standardization in platform stabilometry being a
8. Snijders CJ, Verduin M: Stabilograph, an accurate
part of posturography. Aggressologie
instrument for sciences interested in postural
24C:321-326, 1983.
equilibrium. Agressologie 14C.T5-20, 1973.
29. Era P, Heikkinen E: Postural sway during standing
9. Herman R, MacEwen GD: Idiopathic scoliosis: A
and unexpected disturbance of balance in random
visuo-vestibular disorder of the central nervous
samples of men of different ages. J Gerontol
system? In Zorab PA (ed): Sixth Symposium on
40:287-295, 1985.
Scoliosis. New York, Academic Press, 1980, p 61.
30. Kataoka J, Sakamoto K, Hara T, Hayami A:
10. Harris GF, Knox TA, Larson SJ, et al: A method
Principal component analysis of spontaneous
for the display of balance platform center of
physical movements in sustained standing posture
pressure data. J Biomech 15:741-745, 1982.
of children. J Hum Ergol 10:61-71, 1981.
11. Nayak VSL, Gabell A, Simons MA, Isaacs B:
31. Yamomoto T: Changes in postural sway related to
Measurement of gait and balance in the elderly. J
age. J Phys Fitness Jpn 28:249-256, 1979.
Am Geriatrics Soc 30:516-520, 1982.
32. Lakes RS, Korttila K, Eltoft D, et al: Instrumented
12. Koozekanani SH, Stockwell CW, McGhee RB,
force platform for postural sway studies. IEEE
Firoozmand F: On the role of dynamic models in
Trans Biomed Eng 28:725-729, 1981.
quantitative posturography. IEEE Trans Biomed
33. Hashizume K, Ito H, Marruyama H, et al: Age-
Eng 27:605-609, 1980.
related changes of stability in standing posture.
13. Valk-Fai T: Analysis of the dynamical behaviour Jpn J Geriatr 23:85-92, 1986.
of the body whilst “standing still.” Agressologie
34. Kirby RL, Price NA, MacLeod DA: The influence
14C:21-25, 1973.
of foot position on standing balance. J Biomech
14. Shimba T: An estimation of center of gravity from 20:423-427, 1987.
force platform. J Biomech 17:53-60, 1984.
35. Gantchev G, Popov: Quantitative evaluation of
15. Levin O, Mizrahi J: An iterative model for
induced body oscillations in man. Agressologie
estimation of the trajectory of center of gravity 14C:91-94, 1973.
36. Hlavacka F, Litvinenkova V: First derivative of balance in the frontal plane. Agressologie
the shabilogram and posture control in visual 23:71-73, 1982.
feedback conditions in man. Agressologie 54. Booth JB, Stockwell CW: A method for
14C:45^19, 1973. evaluating vestibular control of posture.
37. Seidel H, Brauer D: Effects of visual information, Otolaryngology 86:93-97, 1978.
conscious control and low-frequency whole-body 55. Ishida A, Imai S, Fukuoka Y: Analysis of the
vibration on postural sway. Agressologie posture control system under fixed and sway-
20C: 189-190, 1979. referenced support conditions. IEEE Trans Biomed
38. Yamamoto T: Changes in postural sway related to Eng 44:331-336, 1997.
fatigue. J Phys Fitness Jpn 28:18-24, 1979. 56. Nashner LM, Block FO, Wall C: Adaptation to
39. Tokita T, Miyata H, Matsuoka T, et al: Correlation altered support and visual conditions during
analysis of the body sway in standing posture. stance: Patients with vestibular deficits. J Neurosci
Agressologie 17B:7-14, 1975. 2:536-544, 1982.
40. Flufschmidt A, Dichgans J, Mauritz KH, 57. Allum JHJ, Pfaltz CR: Visual and vestibular
Hufschmidt M: Some methods and parameters of contributions to pitch sway stabilization in the
body sway quantification and their neurological ankle muscles of normals and patients with
applications. Arch Psychiat Nervenkr bilateral peripheral vestibular deficits. Exp Brain
228:135-150, 1980. Res 58:82-94, 1985.
41. Nashner LM, Black OF, Wall C III: Adaptation to 58. Forssberg H, Nashner LM: Ontogenetic
altered support and visual conditions during development of postural control in man:
stance: Patients with vestibular deficits. J Neurosci Adaptation to altered support and visual
conditions during stance. J Neurosci 2:545-552,
2:536-544, 1981.
42. Black OF, Wall C III, Nashner LM: Effects of 1982.
59. Dvir Z, Daniel-Atrakci R, Mirovski Y: The effect
visual and support surface orientation references
of frontal loading on static and dynamic balance
upon postural control in vestibular deficient
reactions in patients with chronic low back
subjects. Acta Otolaryngol 95:199-210, 1983.
dysfunction. Basic Applied Myol 7:91-96, 1997.
43. Dietz V, Mauritz KH, Dichgans J: Body
60. Dvir Z, Trousil T: A multiple parameter study of
oscillations in balancing due to segmental stretch
dynamic balance in the frontal and sagittal planes.
reflex activity. Exp Brain Res 40:89-95, 1980.
Agressologie 22:129-134, 1981.
44. Arcan M, Brull MA, Najenson T, Solzi P: FGP
61. Dvir Z, Trousil T: Instrumented stabilometer for
assessment of postural disorders during the
dynamic balance studies. Med Biol Eng Comput
process of rehabilitation. Scand J Rehabil Med
20:19-22, 1982.
9:165-168, 1977.
62. Trousil T, Dvir Z: Dynamic balance: A learning
45. Seliktar R, Susak Z, Najenson T, Solzi P:
strategy. Hum Movement Sci 2:211-218, 1983.
Dynamic features of standing and their correlation
63. Smith JW: The forces operating at the human
with neurological disorders. Scand J Rehabil Med
ankle joint during standing. J Anat (London)
10:59-64, 1978. 91:545-564, 1957.
46. Taguchi K: Spectral analysis of the movement of 64. Bizzo G, Guillet N, Patat A, Gagey PM:
the center of gravity in vertiginous and ataxic Specifications for building a vertical force
patients. Agressologie 19B:69-72, 1978. platform designed for clinical stabilometry. Med
47. Mauritz KH, Dichgans J, Hufschmidt A: Biol Eng Comput 23:474-476, 1985.
Quantitative analysis of stance in late cortical 65. Kodde L, Caberg HB, Mol JMF, Massen CH: An
cerebellar atrophy of the anterior lobe and other application of mathematical models in
forms of cerebellar ataxia. Brain 102:461—468, posturography. J Biomed Eng 4:44-48, 1982.
1979. 66. Magnusson M, Johannson R: Characteristic
48. Dichgans, Mauritz KH, Allum JHJ, Brandt Th: parameters of anterior-posterior body sway in
Postural sway in normals and atactic patients: normal subjects. In Amblard B, Berthoz A, Clarac
Analysis of the stabilizing and destabilizing F (eds): Posture and Gait: Development,
effects of vision. Agressologie 17C: 15-24, 1976. Adaptation and Modulation. Excerpta Medica
49. Bles W, DeWit G: Study of the effects of optic International Congress Series 812. Amsterdam,
stimuli on standing. Agressologie 17C:l-5, 1976. Elsevier, 1988, p 177.
50. Nashner LM: Adapting reflexes controlling the 67. Nashner LM: A model describing vestibular
human posture. Exp Brain Res 26:59-72, 1976. detection of body sway motion. Acta Otolaryngol
51. Soechting JF, Berthoz A: Dynamic role of vision 72:429-436, 1971.
in the control of posture in man. Exp Brain Res 68. Nashner LM: Analysis of movement control in
37:551-561, 1979. man using the movable platform. Adv Neurol
52. Vidal PP, Berthoz A, Millanvoye M: Difference 39:607-619, 1983.
between eye closure and visual stabilization in the 69. Nashner LM, Shumway-Cook A, Marin O: Stance
control of posture in man. Aviation Space Environ posture control in select groups of children with
Med 53:166-170, 1982. cerebral palsy: Deficits in sensory organization
53. Dvir Z, Trousil T: EMG study of several lower and muscular coordination. Exp Brain Res
limb muscles during maintenance of dynamic 49:393-409, 1983.
70. Fioretti S, Leo T, Maurizi M, et al: Functional Characterization of gait of young adult females by
evaluation of multiple sclerosis patients at an early means of body center of mass oscillations derived
stage. In Woollacott M, Horak F (eds): Posture from ground reaction forces. Gait Posture
and Gait: Control Mechanisms, vol 2. Xlth 1:61-68, 1993.
International Symposium of the Society for 80. Stem B, Stern FM: Neuropsychological outcome
Postural and Gait Research. Portland, OR, during late stage of recovery from brain injury: A
University of Oregon Books, 1992, p 63. proposal. Scand J Rehabil Med Suppl 12:27-30,
71. Fioretti S, Leo T, Maurizi M, Pieroni R: A 1985.
CAMA system for the functional evaluation of 81. Stem JM, Melamed S, Silbey S, et al: Behavioral
posture maintenance. In Woollacott M, Horak F disturbances as an expression of severity of
(eds): Posture and Gait: Control Mechanisms, vol cerebral damage. Scand J Rehabil Med Suppl
2. Xlth International Symposium of the Society 12:36—41, 1985.
for Postural and Gait Research. Portland, OR, 82. Fernie GR, Holling PJ: Postural sway in amputees
University of Oregon Books, 1992 p 67. and normal subjects. J Bone Joint Surg (Am)
72. Mizrahi J, Najenson T, Nissel R: Asymmetry and 60:895-898, 1978.
total activity analysis of postural sway option in 83. Vittas D, Larsin TK, Jansen EC: Body sway in
cerebral vascular accident patients. Proc 14th Int below-knee amputees. Prosthet Orthot Int
Conf on Med and Biol Eng Espoo, Finland. Med 10:139-141, 1986.
Biol Eng Comput 23(Suppl 1):418-419, 1985. 84. Porter D, Roberts VC: A review of gait
73. Mizrahi J, Solzi P, Ring H, Nisell R: Postural assessment in the lower limb amputee. 2. Kinetic
stability in stroke patients: Vectorial expression of and metabolic analysis. Clin Rehabil 3:157-168,
asymmetry, sway-activity and relative sequence of 1989.
reactive forces. Med Biol Eng Comput 85. Nissen SJ, Newman WP: Factors influencing
27:181-190, 1989. reintegration to normal living after amputation.
74. Mizrahi J, Groswasser Z, Susak Z, Reider- Arch Phys Med Rehabil 73:548-551, 1992.
Groswasser I: Standing posture of craniocerebral 86. Burgess EM: Amputations. Surg Clin North Am
injured patients: Bi-lateral reactive force pattern. 63:749-770, 1983.
Clin Phys Physiol Meas 10:25-37, 1989. 87. Skinner HB, Effeney DJ: Gait analysis in
75. Ring H, Mizrahi J: Bilateral postural sway in amputees. Am J Phys Med 64:82-89, 1985.
stroke patients: New parameters for assessing and 88. Zahedi MS, Spence WD, Solomonidis SE, Paul
predicting locomotor outcome. J Neurol Rehabil JP: Alignment of lower limb prostheses. J Rehabil
5:175-179, 1991. Res Dev 23:2-19, 1986.
76. Ring H, Mizrahi J: Biomechanical sway 89. Hannah RE, Morrison JB, Chapman AE:
parameters in the evaluation of stroke patients. Prostheses alignment: Effect on gait of persons
Neurorehabilitation 2:27-35, 1992. with below-knee amputations. Arch Phys Med
77. Isakov E, Mizrahi J, Ring H, et al: Standing sway Rehabil 65:159-162, 1984.
and weight bearing distribution in people with 90. Andres RO, Stimmel SK: Prosthetic alignment
below-knee amputations. Arch Phys Med Rehabil effects on gait symmetry: A case study. Clin
73:174-178, 1992. Biomech 5:88-96, 1990.
78. Isakov E, Mizrahi J, Susak Z, et al: Influence of 91. Jones D, Paul J: Analysis of variability in pylon
prosthesis alignment on the standing balance of transducer signals. Prosthet Orthot 2:35-50, 1973.
below-knee amputees. Clin Biomech 9:258-262 92. Wilson AB Jr, Pritham C, Cook T: Force-line
1994. visualisation system. Prosthet Orthot 3:85-87
79. Crowe A, Schiereck P, deBoer R, Keessen W: 1979.
Biomechanics of Manual
Therapy
Michael Lee, Julianna Gal, and Walter Herzog
Manipulative procedures have been used for hun¬ with manipulation. In addition to the loads being
dreds, perhaps thousands, of years.1 This century, applied slowly, mobilizations are distinguished by
especially since World War II, detailed descrip¬ a large number of loading cycles and a much
tions of the techniques have been published, and longer duration of loading than manipulations. In
the training of practitioners of manual therapy this chapter, the two classes of techniques are
has become increasingly formalized. Despite the treated separately.
recent formal codification of the procedures in¬ Although the term manual therapy strictly ap¬
volved, there has been little attempt to develop a plies to any therapy delivered with the prac¬
comprehensive analysis of the mechanics of man¬ titioner’s hands, the use of the term in this chapter
ual therapy techniques using realistic assumptions is confined to that group of procedures in which
about the nature of joint movements. A complete there is an attempt to focus the treatment directly
mechanical analysis would describe the physical on the joints—either the joints of the vertebral
effects on all relevant tissues, would take into column or the peripheral joints. The objective of
account the effects of tissue abnormalities associ¬ this chapter is to review the current state of knowl¬
ated with disease or trauma, and could be vali¬ edge of the mechanics of manual therapy tech¬
dated by experimental data. Further, the conse¬ niques. This information is potentially valuable
quences of variations in therapeutic technique for practitioners who are attempting to understand
would be considered, to allow prediction of the the possible mechanisms involved in manual ther¬
treatment methods most likely to produce the de¬ apy and to develop theories that may optimize the
sired outcome. Such an analysis has not yet been use of manual techniques.
presented. Certain elements of the manual therapy
process have been examined, but the development
of a complete analysis requires research in many Joint Movements and Loads:
areas. Description and Quantification
There are two main classes of manual tech¬
niques, categorized on the basis of the rate of
strain and number of loading phases. One class is Because we are considering a therapy that is di¬
commonly known as manipulation. In a manipula¬ rected at joints, we need to be able to describe the
tion, the manual force is usually delivered using a mechanical effects of the therapy on joints. This
single high-speed thrust. The other class of tech¬ description process essentially involves two ele¬
niques is often referred to as mobilization and ments. First, we need to describe the joint
is associated with relatively slow loading rates, kinematics—the movements that occur between
ranging from almost static loading to cyclic load¬ the bones that form the target joint and any other
ing at rates as high as 5 to 6 Hz. Mobilization, joints that may be of interest. Second, we need
however, is generally performed at less than 2 Hz, to describe the joint dynamics—the forces and
and in the case of higher frequencies (5 Hz), the moments applied to the bones involved and acting
amplitude of oscillation is probably quite small, across the joints of interest, and their role in caus¬
so that the tissue strain rates are small compared ing joint movements.
209
KINEMATICS the application of moments to one bone. These

moments may be required to stabilize one bone to
Movement Produced by the Clinician: prevent movement or to produce rotation of a
“Physiologic” and “Accessory” Movements bone. The long bones allow forces to be applied
at two points some distance apart, allowing a large
In manual therapy, the clinician applies a moment,
moment to be applied. In contrast, smaller bones,
tending to produce rotation at a joint, or a force,
such as the vertebrae and the bones of the foot
tending to translate one bone relative to another.
and wrist, are most likely to be subject to manual
The rotational movements, simulating the move¬
force application that can be represented as acting
ments produced by the patient’s muscles, are
through a single point. In the latter case, variation
known as physiologic movements. Isolated transla¬
in the amount of joint rotation can be produced
tion between bones cannot normally be produced
by altering the point of application and orientation
by the patient through voluntary activation of
of the applied manual force in relation to the other
muscles, and these movements are often known
forces applied to the bone by tissues, such as
as accessory movements. The degree to which a
ligaments, disk, and joint cartilage, that are in
manual therapy procedure produces physiologic or
contact with the bone.
accessory movements is not necessarily evident,
especially when spinal joints are involved. Al¬
though the clinician may be attempting to produce Body Segment Movements and Joint Surface
a pure accessory translation between two bones, Movements
some rotation may also be produced. Similarly,
The relationship between bone movements and
the clinician’s attempt to produce a physiologic
joint surface movements may be of interest to the
movement may not produce the same pattern of
clinician who believes that there is a limitation of
rotation and translation that would result from
relative movement between the joint surfaces that
activation of the patient’s own muscles.
is amenable to manual therapy treatment proce¬
dures. Consider again the typical joint shown in
Joint and Bone Structures
Figure 9-1. Let us first examine the case of an
idealized version of this joint, in which the two
A typical synovial joint can be characterized as
joint surfaces have the same constant radius form¬
shown in Figure 9-1. There is one concave and
ing circular arcs around the same center, C. Imag¬
one convex joint surface, but the radius of curva¬
ine that the ligaments, joint capsule, and muscles
ture is not constant along the joint surface and
hold the surfaces in close contact. Physiologic
not exactly the same on the concave and convex
movement of either bone will occur in a pattern
surfaces. For any given joint position, a point,
that involves rotation of all points on the moving
such as point C, at the center of the arc that forms
bone around the center of the joint surface arc, C.
the convex joint surface at the point of contact,
Hence, if a given joint conforms to this ideal
can be located. Because the joint surface does not
form a perfect circle, the point C will be in a structure, the bone movements, including joint
different location for each different point of con¬ surface movements, can be predicted from the
tact. structure. Each point on the bone, including the
joint surface, would move in a circular path
The size and shape of the bones that compose
the joint are important variables for the clinician. around the joint center, C. As rotation occurs, the
Part of the manual therapy process often involves joint surfaces slide past each other, the “gliding”
movement described by Kaltenbom.2 It can be
noted that in this ideal case, the amount of linear
movement (glide) between the two joint surfaces
at the interface is s = rB, where 0 is the angular
movement between the bones (measured in radi¬
ans), and r is the radius from the surface to point
C, the center of the radius of curvature. The hip
joint is perhaps the joint that conforms most with
this ideal morphology in three dimensions,3 while
the elbow joint is a joint with a two-dimensional
FIGURE 9-1 Typical joint, showing point P, the point of profile that is close to ideal.4
contact between the two joint surfaces; point C, the center In the case of a nonideal joint, with two joint
of the arc describing the bone 2 joint surface at the point of surfaces that do not have the same radius and are
contact; and the direction that is arbitrarily designated not held in close congruence, “rolling” of one
extension.
joint surface on another may occur. Pure rolling
• I Chapter Biomechanics of Manual Therapy 211
may occur when there is no gliding between the space, around which the moving body segment
joint surfaces at the point of contact, but in most rotates (Fig. 9-2). Because of the irregularity of
cases, rolling and gliding occur together. The rela¬ human joints, the axis is almost never exactly
tive amounts and directions of the rolling and perpendicular to a cardinal plane, such as the
gliding determine the amount and direction of sagittal plane. Further, the axis does not usually
movement of the point of contact between the two stay fixed in space during an entire movement
bones. Rolling that would move the point of con¬ through the available range of motion. Fully de¬
tact in one direction can be combined with gliding scribing a physiologic movement pattern involves
that would move the point of contact in the oppo¬ giving the location of the axis at as many points
site direction, to maintain the contact point in in the movement as practical. It is possible that
one location. during a physiologic movement, there is also a
small amount of translation along the axis of mo¬
Description of Joint Movements tion. For example, if a person produced knee
flexion in a sitting position, this translation would
One of the major aims of a manual therapy tech¬ correspond to translation in an approximately
nique is to produce joint movements; that is, an mediolateral direction. Obviously, in normal
attempt is made to produce relative movement joints, this along-axis translation would be small,
between two bones. In many cases, the two bones but its presence in abnormal amounts may be a
are readily identifiable, and the success of the sensitive indicator of certain injuries.
technique can thus be partly measured by describ¬ In some situations, the HAM is simplified by
ing the relative motion between the two bones. representing it as a single point, with the implicit
In the next two sections, methods of describing assumption that the axis is perpendicular to the
physiologic and accessory movement will be plane on which the point is located. This point,
given that can be used for this purpose. known as a center of rotation, can be a valid way
In some cases, there are more than just two of representing the movement, provided the HAM
bones whose relative movement must be de¬ really is perpendicular to the plane concerned, and
scribed. For example, the kinematic outcome of provided there is negligible movement of the body
manual therapy techniques in the foot may be segment out of the plane (along the axis).
given by describing the relative motion between If we perform a manual therapy technique that
the tibia and the calcaneus, between the talus involves a passive physiologic movement, we can
and the calcaneus, between the talus and the first use the HAM to describe the way the joint moves
metatarsal, or between many other bone pairs. in response to that technique. The HAM data for
There may not be just a single anatomic joint this movement could be compared with the HAM
between the bones whose relative movements are data for both voluntary active movements and
of interest. The spine presents a similar situation. passive physiologic movements performed on sub¬
When we apply a manual force to a vertebra, we jects with no abnormalities, to establish the degree
are usually interested in relative movement in at of normality of the passive physiologic movement.
least two intervertebral joints, proximal and distal An alternative approach to movement descrip¬
to the target vertebra, as well as between other tion is to divide the total movement into main and
bones. Therefore, the complete description of the coupled movements, where a main movement is
movement response may involve multiple applica¬ one that occurs in the direction of the externally
tions of the methods described next. applied force (or moment) delivered by the clini¬
cian. The vertebra’s movement in response to the
clinician’s force is influenced by the restraints
Methods Used for Describing
offered by the connecting tissues. Many different
Physiologic Movements
tissue forces are applied to a vertebra, as well as
Because normal voluntary joint movements pre¬ the clinician’s force, and the vertebral movement
dominantly involve rotations of one body segment may therefore take place with components of
relative to another, the most appropriate method movement in any direction. Movements that occur
for describing those movements is a method that concomitantly with the main movement, but in
focuses on the rotation. Rotational movements are directions other than that of the clinician’s force,
best defined by describing the helical axis of mo¬ are known as coupled movements. As shown in
tion (HAM) for the motion. This method is capa¬ Figure 9-3, there can be translations in three direc¬
ble of giving a full three-dimensional description tions (1, 2, 3; the directions of the x, y, and z
of the movement of the segment. The key element axes) and rotations in three directions (4, 5, 6;
of the description process is the defining of the about lines parallel to these three axes). Hence,
helical axis itself. The axis is an imaginary line in there are six possible movement components. In
FIGURE 9-2 Helical axes of

motion corresponding to eight
steps of knee flexion from 0 to 109
degrees, shown in an
anteroposterior view of axes
projected onto the frontal plane.
(Adapted from Huiskes R, van Dijk
R, de Lange A, et al: Kinematics of
the human knee joint. In Berme N,
Engin AE, Correia da Silva KM
[eds]: Biomechanics of Normal and
Pathological Human Articulating
Joints. Dordrecht, The Netherlands,
Martinus Nijhoff, with kind
permission of Wolters Kluwer
Academic Publishers, 1985, p 165.)
other words, in addition to the main movement, instantaneously located along the line of contact
there can be five coupled movement components. between the cylinder and the table. In an end view
For a complete movement description, just as the of the cylinder, the HAM is seen as a point at
HAM components must be given for as many the bottom of the cylinder. The movement of the
small steps as practical throughout the movement, cylinder means that some time later, the HAM has
use of the main and coupled movement approach moved parallel to its original position, to a new
requires that the amount of these main and cou¬ line of contact between the cylinder and the table.
pled movements be given for each small portion The direction of movement from the initial HAM
of the total range of movement. to the final HAM is perpendicular to the axis.
Therefore, if an accessory movement in which
Methods Used for Describing there is a mixture of rotation and translation is
Accessory Movements produced, the movement can be adequately de¬
scribed using the HAM approach.
In contrast to physiologic movements, the aim If we are primarily interested in translations,
of performing accessory movements is usually to however, we may wish to choose an approach that
produce translations of the body segment in¬ focuses more on translation. If the manual therapy
volved. Therefore, if we are to describe the re¬ technique involves a force applied in a particular
sulting movements, there must be an emphasis on direction, then the variable in which we are proba¬
translation, although some rotation may also oc¬ bly most interested is the amount of translation in
cur. In the HAM method, translation along the the direction of the applied force. Therefore, it is
axis is explicitly described, whereas translation in most common to describe the movement response
a direction perpendicular to the axis is accommo¬ by partitioning into main and coupled movements.
dated by movement of the axis location in time. The sizes of the coupled displacements may be
For example, if a cylinder is rolling along a table, expressed in absolute terms or as a proportion of
then the HAM that describes this movement is the main displacement. In a clinical context, the
I Chapter 9 BIOMECHANICS OF MANUAL THERAPY 213
movement, and perhaps coupled movements. For

example, a flexion torque may be applied to an
C> elbow joint, and the joint’s response can be char¬
acterized by showing the variation of flexion angle
y with applied torque. For anteroposterior accessory

movement at the shoulder, the joint’s response
is characterized by the relationship between the
applied anteroposterior force and the anteroposter¬
ior movement of the head of the humerus. In
clinical practice, the clinician attempts to feel only
the main movement, but laboratory studies10 have
described the variation of coupled movements as
well as the main movement.
This approach can be applied to both physio¬
logic and accessory movements. The joint me¬
chanical behavior can be measured, or in clinical
practice, the clinician performs a manual evalua¬
tion. When a manual assessment is performed,
the documented joint behavior is often called a
movement diagram11 (Fig. 9-4). The movement
diagram shows the clinician’s perception of the
amount of movement as a proportion of the per¬
FIGURE 9-3 Coordinate system located at the center of
ceived range of available movement, in relation to
the vertebral body, showing the six possible movements in
the resistance to movement, presumably the force
relation to the x, y, and z axes. (From Panjabi MM, Brand RA,
White AA: Three-dimensional flexibility and stiffness
opposing the clinician’s applied force. In a move¬
properties of the human thoracic spine. J Biomech 9:185, ment diagram, the movement concerned is the
1976, with permission from Elsevier Science.) absolute movement of the point at which the clini¬
cian’s load is applied over the target bone.
To understand the clinical significance of the
main displacement is usually assessed manually, movement diagram more fully, however, we
often while assuming that the coupled displace¬ would also need to know the relationship between
the target bone movement and the movement of
ments are negligible.
other relevant bones. For example, in the case of
the anteroposterior movement at the shoulder, we
KINETICS
might like to know how much of the anteroposter-
Joint Loads During Manual Therapy
Ideally, the clinician would like to know the loads

experienced by each relevant anatomic structure
as a result of the manual force being applied. In
practice, the net force and net moment acting Resistance to
across a joint may be more readily estimated, joint movement
using the techniques of inverse dynamics. Alterna¬ during
mobilization
tively, direct measurement of tissue forces5 and
strains6 is possible, and tissue loads can also be (arbitrary units)
predicted using theoretical models.7 The loads ap¬
plied by the clinician can be measured by interpos¬
ing a transducer between the hands and the pa¬
tient’s skin,8 or they may be estimated by placing
either the clinician or the patient on a forceplate Amount of joint movement
to measure the ground reaction force.9 (% ROM for a “normal" joint)
FIGURE 9-4 The “movement diagram” is the clinician’s

Mechanical Behavior of Joints perception of the variation of resistance with displacement
(with displacement expressed as a proportion of the normal
The mechanical behavior of a joint is commonly range of motion [ROM]). Maitland" has stated that the
characterized by describing the relationship be¬ perceived pattern of resistance could take a number of
tween an applied force (or moment) and the main forms, including curves such as the three curves shown.
ior movement is occurring at the joint of primary is a commonly quoted variable,13 whereas the am¬
interest, the glenohumeral joint. We may also like plitude of force variation14 is relevant to cyclic
to know the amount of movement of the scapula loading conditions, as occur during mobilization.
in relation to the rib cage and movement at the During both cyclic and single-thrust loading cases,
acromioclavicular joint. It is an implicit assump¬ there is often a base level of force that is applied
tion of many manual therapy techniques that a before the main manual therapy force.8 The mag¬
substantial contribution to the absolute movement nitude of this base level of force is also of interest.
of the target bone comes from the joint that is of
primary interest, but for most joints, there have Displacement
been no studies of this issue.
The amount of relative displacement (linear and
angular) between the bones that make up the target
The Manual Therapy Process joint is a variable in which clinicians are likely to
be interested. Manual therapy is often directed at
The aim of manual therapy treatment is usually to increasing the amount of relative displacement at
relieve pain; however, there is said to be a link a joint. Therefore, the joint kinematic behavior
between lack of joint mobility and joint pain.12 during the manual therapy intervention, as detailed
Also, some patients describe problems that appear previously, can be used as a measure of the me¬
to be primarily due to joint hypomobility in one chanical input to the joint.
or more directions. Therefore, many clinicians use
treatments aimed at increasing the mobility of Energy
joints.
The manual therapy intervention can be quanti¬ The energy transferred to the joint can be calcu¬
fied in terms of the mechanical input to the patient lated as the area under the force-displacement
or in terms of the outcome of the treatment. In curve during loading. During unloading, much of
traditional medical language, the nature of the this energy is returned to the clinician, with the
treatment input can be described as the dose. The energy returned being the area under the unloading
manual therapy dose can largely be measured in curve. Because, in the case of biologic materials,
mechanical variables, but the treatment outcome the unloading curve follows a different (lower)
measures often include measures of pain and other path than the loading curve, there is a net transfer
symptoms as well as physical measures. In this of energy into the tissues. The amount of energy
chapter, only mechanical outcome measures that transferred is equal to the area between the loading
are assessed by manual application of force to and unloading curves.15
a joint are considered. Other types of outcome
measurement tools, such as pain scales or range- Sounds
of-motion measures, may be valid as indicators of
the patient’s status but are outside the scope of It has been observed that when manual forces are
this chapter. applied to joints, a clicking noise sometimes arises
from the joint. This noise is more commonly asso¬
POSSIBLE METHODS FOR MEASURING DOSE ciated with high-speed manipulation than is the
case when forces are applied slowly. Some prac¬
A number of approaches have been suggested for titioners regard the occurrence of the click as a
quantifying the dose involved in manual therapy. criterion for the “success” of the manipulation. A
discussion of the possible causes and significance
Force of the click is given later in the section on manipu¬
lation.
The variation of applied manual force with time
during the manual therapy procedure is one of
OUTCOME EVALUATION BY MANUAL
the most commonly measured mechanical dosage
TECHNIQUES
characteristics. The area under the curve showing
applied force versus time is the impulse applied Many manual therapists use manual methods to
by the manual force.8 The impulse may be a more evaluate the force-displacement or torque-angle
important variable than the magnitude of the force characteristics of a joint, as described previously.
alone. The instantaneous or short-term average These characteristics are assessed before and after
rate of change of force can also be computed to the treatment to indicate the effect of the manual
describe the behavior of the applied manual treatment on the mechanical behavior of the joint.
force.12 The maximum value of the applied force One method of describing the manual findings is
to represent the perceived variation of resistance which depends on the “grade” of treatment. Mait¬
(force or torque) on a movement diagram.11 The land11 defined a series of grades of mobilization
accuracy of such subjective representations has in terms of the location of the movement within
not been evaluated. An alternative is to use clini¬ the range of motion of the joint. Based on Mait¬
cal jargon to express the nature of the mechanical land’s definitions, a grade I mobilization is a
behavior, often without unambiguously defining small-amplitude movement near the beginning of
those terms. For example, the range of joint move¬ range; a grade II mobilization is a large-amplitude
ment has been divided into movement up to the movement that occurs in a part of the range that
“first stop” and movement between the “first does not involve significant resistance; grade III
stop” and the “final stop,” without defining these mobilizations are performed with a large ampli¬
terms.2 The mechanical behavior in the latter part tude and do involve significant resistance; and
of the movement between the first stop and the grade IV mobilizations are small-amplitude move¬
final stop, the “end-feel” of the joint, has been ments performed near the end of range (Fig. 9-5).
described in terms such as “soft,” “firm,” and Different clinicians show different interpretations
“hard.” It has also been implied that a relationship of the levels of force corresponding to these
exists between the type of end-feel and the tissue grades,9 probably because of the subjective inter¬
that prevents further movement.2 As pointed out pretation of concepts such as “significant resis¬
by Riddle,16 there are few data to support the idea tance” and “end of range.”
that there is a consistent relationship between the Performing a Maitland11 grade II mobilization
tissue that limits movement and the nature of the of the L3 vertebra, a group of experienced physi¬
end-feel. Further, the terms soft, firm, and hard cal therapists applied an average maximum verti¬
are subjective and cannot be unambiguously trans¬ cal component of force of 33.3 N. Using a similar
lated into measurable parameters. Therefore, in measurement method, and also applying grade II
this chapter, we consider only methods and de¬ mobilization, Matyas and Bach9 found that the
scriptive terms that can be understood as measur¬ peak forces when mobilizing four thoracic verte¬
able physical variables. brae varied from 7.6 N to 87.1 N among a group
of seven specialist manual therapists. They found
high variability in peak applied force among the
MOBILIZATION clinicians tested.9 In another study using the L3
vertebra,14 the average maximum force applied to
There has been little systematic study of the re¬ a group of 18 healthy subjects by an experienced
sponse of joints to mobilization. Those data that physical therapist performing grade IV mobiliza¬
are available are mostly experimental results refer¬ tion was 92.5 N. Threlkeld,19 using two experi-
ring to the lumbar spine. In the future, similar
experimental methods could be applied to periph¬
eral joints to allow an analysis of the effects of
relevant variables to be quantified. By examining
the current lumbar spine data, together with infor¬
mation about the mechanical behavior of joints
and isolated tissues, however, some principles may Amount and
be established that could, to some extent, be ap¬ location of joint
plied to joints that have not yet been studied. movement
during
MOBILIZATION DOSE: THE APPLIED mobilization
MANUAL FORCE
(% ROM)
The manual force is usually applied with a magni¬
tude and pattern that is varied in a graded manner
and dependent on the patient’s problems and the
aims of treatment.11, 17 The force applied during
mobilization has been measured in a number of
studies, all involving the spine. The forces have
been measured during mobilization of healthy sub¬ Grade of mobilization
jects,9, 18, 19 patients with low back pain,20 and FIGURE 9-5 The grades of mobilization for mobilization of
physical models of patients.21 a normal joint with a “hard end-feel.” The amount of
Perhaps the most commonly recorded parameter movement is expressed as a proportion of the total range of
related to force has been the maximum value, motion (ROM). (Data from Maitland.")
FIGURE 9-6 Force-time graph for mobilization of

the L4 vertebra during assessment of stiffness of a
subject lying prone. The range is defined as the
difference between the maximum force and the
minimum force during a set of oscillations. The
magnitude of the applied force shown is the sum of
the two major components: the vertical component
and the horizontal component. (Data from Viner
and associates.45)
enced physical therapists to mobilize the mid- ing cyclic loading has been reported much less
thoracic spine, found peak vertical forces of 140 frequently than the magnitude of the peak force.
and 206 N in grade I, and 232 and 500 N in grade We can define the range of forces as the difference
IV. Simmonds and colleagues21 performed an in¬ between the maximum force and the minimum
vestigation that used a physical model of the spine force during oscillations (Fig. 9-6). Threlkeld19
based on a spring. In part of the experiment, a reported an average range of about 39% of the
spring with a stiffness of 22.2 N/mm was used to peak force in grade I thoracic mobilization, and
produce stiffness similar to a high level of lumbar 27% of the peak force in grade IV. Petty14 found
posteroanterior stiffness.22 Ten physical therapists a mean range of about 10% of the peak force for
applied average maximum forces of 58 N, 86 N, a grade IV mobilization at L3. Maitland11 de¬
116 N, and 108 N during mobilization at grades scribed the different grades of mobilization move¬
I, II, III and IV, respectively, according to the ment as having either a small or large amplitude
Maitland11 grading system.21 All these studies (grades I and IV have small amplitude, grades II
found that there is high variability among clini¬ and III have large amplitude), but the rationale for
cians and that peak mobilizing forces ranged from using a particular amplitude of force is not clear.
10 N to 500 N depending on joint location, pa¬ One factor may be the desire to keep the amount
tient, clinician, and grade of mobilization. of soft tissue stretch at a level that is beyond the
Different grades of mobilization are used for toe region, the region corresponding to low loads
different purposes, with relatively large forces be¬ and strains (Fig. 9-7). It has been argued that
ing used in cases in which the clinician is at¬ if the load is at a level beyond the toe region,
tempting to produce increased movement at a microfailure is more likely to occur.19 As pre¬
joint, whereas smaller forces are often used when viously stated, however, the relationship between
pain is the primary problem.19 The relationship applied force and the tension within each resisting
between applied manual force and particular tissue tissue is unknown.
forces has not been established for any spinal or The frequency of the mobilizing force varies
peripheral joints. Although it would be expected among individual clinicians19 and, within one as¬
that greater applied forces would produce greater sessment session, may be varied by a clinician to
tissue forces, the pattern of load sharing among alter his or her perception of the nature of the
the tissues would be expected to vary with the resistance to movement.11 In a small number of
level of applied load, owing to the nonuniformity cases, the mobilizing frequency has been mea¬
among tissues and nonlinearity of the tissue be¬ sured, and frequencies in the range of 1.5 to 5.5
havior.23
Hz have been observed.14'19 Greater frequency of
The amplitude of the applied manual force dur¬ loading is associated with increased stiffness of
to produce one of the components of joint surface

movement without the presence of the other
movements (e.g., translation without the main ro¬
tation), a kinematic analysis is needed to examine
whether this objective is being met.
Movements of Vertebrae
The movements of the many vertebrae involved

in spinal mobilization are potentially more com¬
plex than bone movements during peripheral mo¬
bilization, in which only two bones are usually
involved. A major issue to be considered is how
the various vertebrae move when the mobilizing
force is applied to one point on a single target
vertebra. The behavior of the most commonly
studied region, the lumbar spine, can be seen in
terms of principles that are also applicable to other
Skin surface displacement (mm)
parts of the spine, although their application to a
FIGURE 9-7 Force-displacement curve for thoracic particular region must take into account the spe¬
posteroanterior loading showing a nonlinear region below an cial characteristics of that region.
applied load of about 20 N.
One of the first to study the vertebral move¬
ments arising from manual force applied to the
lumbar spine was Ward (cited in Lee and Mose¬
the overall response (see later) and may also be ley,25 p. 155). Ward reported a single case study
linked to more localized movement,24 although of the application of a static 100-N force to the L3
this hypothesized relationship has not yet been spinous process. Qualitative comparison of lateral-
proved. view radiographs made in the unloaded and loaded
The dose of mobilization therapy could also be cases showed that the force produced a general
measured by reference to the amount of displace¬ increase in lumbar extension. The individual inter¬
ment produced, energy transferred to the patient, vertebral movements were small. In a subsequent
or whether clicking sounds were released. For experimental investigation,24 cyclic loading at 0.5
mobilization, however, these methods do not ap¬ and 1.0 Hz in the vertical direction was applied
pear to have been used. over the L3 spinous process. Measurement of skin
surface movements showed that at both frequen¬
PATIENT RESPONSES TO MOBILIZATION cies, there was a measurable response as far away
as T8. Another study of L3 mobilization, using a
FORCES
slower mobilization frequency, found that pelvic
Bone Movements in Peripheral Joint Mobilization anterior rotation also occurred. There were about
2 degrees of pelvic anterior rotation for every 100
There have been few published studies of the N of applied L3 force.26
mechanical behavior of peripheral joints of living An approach that allows more detailed descrip¬
subjects during actual or simulated mobilization tion of the vertebral movements is to use a mathe¬
procedures. The investigations that have been per¬ matical model to predict the response to the man¬
formed have not involved detailed kinematic anal¬ ual force. Such a model was used by Lee and
ysis. Most peripheral joint mobilization proce¬ associates27’28 to simulate posteroanterior forces
dures performed involve accessory movements, being applied to the lumbar spinous processes, as
with the clinician almost always attempting to occurs during mobilization. A validation study
produce translations at the joint surface. Questions found that the model behaved in a way that was
to be answered by a kinematic analysis would reasonably representative of an average healthy
include the following: What joint surface move¬ subject.27 The model predicted that the interverte¬
ments are produced during the mobilization? How bral rotation and translation at each joint were
do the movements produced during mobilization small—generally less than 1 degree or 1 mm per
compare with the movements that occur during 100 N of applied force (Fig. 9-8).
active movement? Does the “fixed” bone (stabi¬ Figure 9-8 shows that the greatest degree of
lized by the clinician or the patient) actually re¬ extension was predicted to occur in the joints
main stationary? If an aim of the mobilization is adjacent to the target bone, with the joints above
Extension
Anterior,
superior
FIGURE 9—8 Predicted response to 100-N

posteroanterior force applied to the spinous
process of the L3, L4, or L5 vertebra. A, Angular
displacement in the sagittal plane. B, Linear
displacement in the shear and longitudinal
directions. (From Lee M, Kelly DW, Steven GP:
Lumbar spine manual therapy: Effect of choice of
target vertebra. In Proceedings of the First
Australian Biomechanics Conference, Sydney, 1996,
p 104.)
and below the target bone showing similarly sized vertebra. The model predicted that lumbar mobili¬
rotations. Shear displacements were predicted to zation is not a localized technique. Many joints
be maximal around the thoracolumbar junction move concomitantly, in contradistinction to the
and to be small in the joints adjacent to the target textbook image of isolated intervertebral move-
ments at joints immediately cephalic and caudal Tissue Resistance to Mobilization

to the target vertebra.11,29 Another important aspect of spinal mobilization
Because lumbar mobilization produces move¬ dynamics is the pattern of resistance to the mobi¬
ment at many joints and deformation of a number lizing force, as perceived by the clinician. Because
of anatomic structures,11 the total amount of move¬ the manual force is applied slowly, the clinician
ment of the skin surface where the clinician ap¬ is able to perceive the nature of the mechanical
plies the force can be quite large, on the order of response. This response is thought to be indicative
10 mm of displacement under 100 N of applied of the mechanical behavior of relevant spinal
force.30 The relation between the joint mechanical joints, although there are few data to support this
properties of interest to the practitioner and the assumption. The pattern of resistance, shown as a
total movement under the practitioner’s hands is movement diagram, gives the relationship between
not clear. For lumbar spine mobilization, it has applied manual force and body surface displace¬
been implied that intervertebral shear displace¬ ment. Throughout this chapter, the gradient of
ments are the major focus and that these move¬ this force-displacement curve is described as the
ments are reflected in the movement felt by the stiffness of the mobilization movement. In the
clinician during mobilization.11, 29 Although mod¬ terminology of Latash and Zatsiorsky,31 this could
eling studies indicate that the intervertebral shear be described as the apparent quasi-stiffness be¬
displacements are likely to be less than 10% of cause the physical nature of the resistive forces is
the total posteroanterior movement,28 it is possible not known, and it is not measured at equilibrium.
that the intervertebral joint properties could have It should be emphasized that the movement stiff¬
a palpable influence on the overall movement. ness, measured in this way, cannot necessarily be
The general patterns of movement described directly related to purely elastic behavior and the
storage of elastic potential energy.
previously for lumbar spine mobilization may also
The force-displacement curves have been found
apply similarly to other spinal regions, although
to have two phases.24,30 The first phase is a region
kinematic studies have not yet been performed.
in which the stiffness (the gradient of the force-
The movements occurring in the thoracic spine
displacement relation) begins at a low level and
are likely to reflect the additional constraints of¬ then increases rapidly, producing a nonlinear re¬
fered by the rib cage, whereas the cervical spine sponse. The nonlinear region continues up to
movements are influenced by the high segmental about 20 to 30 N and is followed by a linear
mobility and the variable degree of constraint pro¬ region.24,30 A number of investigations have found
vided over its length. The manner in which the this region to be linear in the range 30 to 100
patient’s head is supported on the treatment couch N,24,30 but a recent case study found that when a
may be an important variable. large range of forces was considered (about 250
N), there was observable nonlinearity32 (Fig. 9-9).
There was found to be a modest stiffening of the
Tissue Loads movement with increasing force, so that if a large
range of forces was considered, the force-
The loads sustained by various anatomic elements displacement relationship would need to be repre¬
sented by a nonlinear function, such as a parabola.
during mobilization are important determinants of
Because most investigations have employed a
the effects on the tissues.19 In relation to spinal
moderate range of forces (less than 100 N), the
mobilization, there is little direct evidence about
region above 30 N has been satisfactorily repre¬
the load distributions. Lee and Evans7 used a sim¬
sented as linear and has been fitted with a straight-
ple model of the lumbar spine to predict the pat¬ line relationship between force and displacement.
tern of posteroanterior shear force and extension In such cases, the entire force-displacement rela¬
moment variation under an L4 posteroanterior tionship has been characterized by the length of
manual force of 150 N. They predicted anterior the toe region up to 30 N (D30) and the slope of
shear loads of about 120 N at joints caudal to the the subsequent linear region (stiffness coefficient
point of loading and posterior shear loads of about K).33 Average values of K for normal young sub¬
30 N at cephalic joints, with extension moments jects have been found to be about 15 N/mm for
of 2 N m and 6.5 N m at L5-S1 and L4-L5, L3 loads, with the normal values ranging from
respectively, and varying from 5 N m at L1-L2 to about 8 to 29 N/mm.22,30,34
8 N m at L3-L4 at lumbar joints cephalic to the
Factors Associated With an Alteration in Lumbar
target vertebra. The extent of simplifying assump¬
Mobilization Stiffness
tions and the lack of validation studies mean that
these predictions may not be good indicators of Many manual therapists believe that one sign of
the true load distribution. joint disease or dysfunction is the presence of
Displacement (mm)
Figure 9-9 Force-displacement relationship up to 275 N of applied force, for forces applied to the L3
vertebra at 0.5 Hz. Data for two subjects are shown. (From Lee M, Latimer J, Maher C: Normal response to
large postero-anterior lumbar loads: A case study approach. J Manipulative Physiol Ther 20:369, 1997.)
altered stiffness during manual assessment using the total displacement. Referring to the conceptual
mobilization procedures. As described previously, model in Figure 9-10, the displacement through
it is likely that many tissues contribute to the space of a lumbar vertebra under a manual load,
movement felt by the clinician during mobiliza¬ A, is partly dependent on the intervertebral joint
tion. The lumbar spine can be seen as a flexible, stiffness at nearby intervertebral joints. Flexible
segmented complex that is suspended between two intervertebral joints allow greater movement of
supporting structures: the pelvis and the rib cage.35 the target vertebra in relation to its neighbors.
A conceptual model of this structure is given in The degree to which the mobilization stiffness is
Figure 9-10. The posteroanterior movement of the affected by the stiffness of any single interverte¬
target vertebra is produced by many intervertebral bral joint is unknown. A number of factors that are
movements. Further, the deformation of the rib associated with significant increases in stiffness
cage and movement of the pelvis also contribute to during lumbar mobilization have been identified.
E
FIGURE 9 10 Simple schematic model of application of manual force to spine (not to scale). The head
(G), pelvis (C), and vertebrae (such as J) can be considered rigid bodies. The spine may be represented as a
stiff segmented beam comprising the vertebrae and the interconnecting tissues (M). The spine is supported
by the head, by a series of springs (K, representing the rib cage), and by the pelvis. The pelvis in turn is
supported by springs restraining vertical (E) and horizontal (D) movement and a spring that resists pelvic
rotation around the axis (F). The manual force (A) is applied through a spring representing the dorsal soft
tissues (B) to the spinous process (H), and the patient lies on a bed (L). (Modified from Lee M, Steven GP,
Crosbie J, et al: Variations in posteroanterior stiffness in the thoraco-lumbar spine: Preliminary observations
and proposed mechanisms. Phys Ther 78:1277, 1998. With permission of the American Physical Therapy
Association.)
Muscle Activity age increases in stiffness are observed when

changing frequency from 0.5 to 1 Hz.24 The stiff¬
One factor that has the potential to have a large ness of the posteroanterior responses observed by
influence on the stiffness during lumbar mobiliza¬ Lee and Liversidge22 under very slow loading
tion is the degree of activity of lumbar muscles. were highly correlated with the stiffness values
When a person with no low-back pain lies in a observed at 0.5 Hz, with correlation coefficients
relaxed prone position, it is normal for there to be of 0.82 at L3, 0.91 at L4, and 0.97 at L5. The
no activity of the lumbar extensors.36'37 It has been tissues that are the source of this increasing stiff¬
reported that patients with low-back pain do have ness with increasing frequency are not known
variable amounts of muscle activity present when because many tissues contribute to the stiffness
lying in a position that is used for manual ther¬ felt by the clinician, and all show properties that
apy,36 although much information on the presence depend on strain rate.22
of muscle activity is based on unsubstantiated
clinical observation. When healthy subjects maxi¬ Location of Target Bone
mally activated their spinal extensors, with the
pelvis and thorax restrained, there was a 350% Within the spine, the location of the target vertebra
increase in mean posteroanterior stiffness at L3 is an important variable in determining the stiff¬
during mobilization at 0.7 Hz.34 ness during mobilization.22 There are a number of
It has also been found that spinal extensor acti¬ possible reasons for this, but it has been proposed
vation at levels below maximum produces in¬ that the proximity of the target vertebra to the
creases in posteroanterior stiffness that are nearly relatively rigid pelvis is a major factor.22
proportional to the level of activation of the lum¬ As indicated in the conceptual model of Figure
bar erector spinae.38 A hypothesized mechanism 9-10, deformation of the rib cage and displace¬
for this effect relates to the orientation of the ment of the pelvis can allow movements of the
major lumbar extensors, the erector spinae entire lumbar spine, adding to intervertebral move¬
group.34 It has been noted that the erector spinae ments within the lumbar spine to produce the total
pass posteroinferiorly from their origin on the movement of the target vertebra. Imagine that the
lumbar vertebrae to their insertion on the pelvis, lumbar spine is a rigid beam connected between
so that their action produces a posterior shear two flexible supports, so that no intervertebral
force on the lumbar vertebrae.39 Therefore, be¬ movement occurs. A load is applied at various
cause posteroanterior mobilization produces ante¬ points along the beam. It can be shown that if the
rior movement of the lumbar vertebrae relative to flexibility of the two supports is different, the
the pelvis, the activated erector spinae tend to movement at the point of loading would be great¬
oppose this movement. Mobilization tends to est when the load was applied at the more flexible
stretch these muscles, and their stiffness is nearly support. The movement would be least under a
proportional to the level of activation,40 thus pro¬ load applied at the stiffer support. For loads ap¬
ducing mobilization stiffness that depends on the plied along the beam at points between the sup¬
level of muscle activation. The question remains ports, the movement would be intermediate be¬
open about whether patients with low-back pain tween the values at the two ends, with the
might have increased stiffness during mobilization movement increasing as the load moved from the
as a result of this mechanism or whether the stiff support toward the flexible support.
actual levels of activation are too low to produce In the human body, the rib cage would be ex¬
a significant effect. In this regard, it should be pected to provide a less rigid support to the lumbar
noted that patients with low-back pain were found spine than would the pelvis, owing to the relative
to have 8% higher posteroanterior stiffness than flexibility of the ribs and costal cartilages. If the
when they had little or no pain.33 target vertebra is close to the relatively rigid sup¬
port of the pelvis, then the posteroanterior dis¬
placement of the target vertebra would be ex¬
Mobilization Frequency
pected to be less. Hence, the posteroanterior
Mobilization is usually performed at frequencies mobilization stiffness would be greater. Conse¬
of up to 3 to 5 Hz. In isolated tissues, resistance to quently, the lumbar posteroanterior stiffness could
stretch increases as the rate of stretch increases.41 decrease as the point of load application moves
During lumbar mobilization, this stiffening effect away from the sacrum toward the rib cage, as has
is also apparent. The posteroanterior stiffness has been observed.22, 42 This simple model is not di¬
been observed to increase by an average of about rectly applicable to the spine because the lumbar
25% when the frequency of mobilization changes spine itself is not rigid. Hence, the actual variation
from very slow loading to 0.5 Hz.22 Smaller aver¬ of posteroanterior movement would depend on
the intervertebral joint stiffness as well as on the changed. For example, both the angle between
stiffness of the supports. the applied force and the target vertebra, and the
Another possible factor that could be involved pressure distribution on the anterior body surface
in the variation of posteroanterior stiffness be¬ may have changed concomitantly with the degree
tween target vertebrae is the orientation of the of lordosis.
mobilizing force. The mobilizing force tends to be
oriented in a direction that takes into account Mobilizing Force Magnitude
the sagittal lumbar curve,42 and hence the force
direction changes in association with change of The magnitude of mobilizing force varies with the
the target vertebra. Altering the force direction grade of mobilization.11 As mentioned previously
results in application of different components of (see Fig. 9-9), the gradient of the force-
load to the intervertebral joints. For example, a displacement relation, the movement stiffness, de¬
more caudally applied force at lower lumbar levels pends to some extent on the level of applied force.
is associated with a relatively higher longitudinal A case study has shown that the linear phase of
component of force and changes the moment of the force-displacement response may actually be
force at many intervertebral joints. Changing the nonlinear if a large range of forces is considered.32
direction of applied force also changes the mo¬ Further, depending on the grade of mobilization,
ment of force around the center of pressure some clinicians apply only small forces that may
through which the pelvis is supported by the treat¬ involve responses in the nonlinear region below
ment bed and thus may alter the amount of pelvic 30 N.9’21 Therefore, in some cases of mobilization
rotation. These factors could account for some in which the applied force is below 30 N, or if
changes in stiffness due to change in force direc¬ the range of forces is large, the mobilization stiff¬
tion per se. ness is dependent on the level of applied force. In
Some support for this hypothesis comes from a such cases, the overall stiffness coefficient gives
modeling study. Lee and colleagues27 predicted an indication of the patient’s response, but a more
that at the lower lumbar levels, where the mobiliz¬ appropriate approach to evaluation of stiffness
ing force is inclined somewhat caudad, a change could be used. First, a nonlinear function such as
in direction involving a more caudad inclination a parabola could be fitted to the data, allowing the
would result in greater stiffness. Hence, the tangent stiffness to be calculated at any given
change in force direction alone with change in point on the force-displacement relation. Parabo¬
the target vertebra may be responsible for some las have been successfully fitted to the nonlinear
increase in stiffness. Another possibility is that region below 20 N24 as well as to the region
higher posteroanterior stiffness at inferior levels between 30 and 275 N,32 although there has been
reflects a greater stiffness of the more caudal inter¬ no documented attempt to fit a single nonlinear
vertebral joints. It is unclear to what extent this function to the entire force-displacement relation¬
factor might account for the systematic changes in ship. A second approach could be to represent
posteroanterior stiffness that have been observed22 the force—displacement relationship as piecewise
because the lumbosacral joint stiffness appears to linear. That is, the force-displacement relationship
be lower than at superior joints under anteropos¬ could be broken into a number of segments, each
terior and posteroanterior shear but higher under of which can be reasonably fitted with a straight
flexion and extension moments.43 line. If such an approach is taken with the data for
subject 2 in Figure 9-9, the variation in stiffness is
Degree of Lor do sis more readily apparent (Table 9-1).
The latter method of analysis would be appro¬
The sagittal curve of the lumbar spine varies priate if it was known that a particular interval of
among patients and, during therapy, can be altered force with a linear force-displacement relationship
by the clinician by adjusting the treatment couch was the focus of the mobilization. If stiffness
or by placing pillows under parts of the patient. coefficients are to be compared between patients
One experiment has been conducted in which the or between occasions, it is necessary to use the
lordosis was increased by upwardly tilting the same range of forces in each case.
upper trunk while maintaining the position of the
pelvis and lower limbs. The more lordotic posture
Breathing
produced in this way was found to be associated
with greater posteroanterior stiffness.25 It is un¬ During inspiration, the lumbar spine moves in
clear whether it was the increase in lordosis per an anteroposterior direction. Therefore, there is
se that was responsible for this change in stiffness potential for breathing to produce artefacts as well
because a number of other variables were as to produce real changes in the posteroanterior
• I Chapter 9 BIOMECHANICS OF MANUAL THERAPY 223
TABLE 9-1 Calculated Stiffness Coefficients for A similar mechanism may also operate on the
Force Intervals in the Force-Displacement anterior surface of the patient to produce direct
Relationship for Subject 2* effects on stiffness. The thickness of the anterior
skin and subcutaneous fat may be important in
Force Interval (N) Stiffness Coefficient (N/mm)
providing a compliant element that allows move¬
0-10 6.8 ment of the bony elements, giving additional dis¬
10-30 13.8 placement under the clinician’s hands. For exam¬
30-100 17.5 ple, more anterior soft tissue might allow more
100-170 21.5 pelvic rotation and hence may increase the target
170-240 23.4
bone displacement and reduce the posteroanterior
stiffness. A model study suggested that the mobili¬
*See Figure 9-9.
Data from Lee M, Latimer J, Maher C: Normal response to large zation stiffness would be very sensitive to the
postero-anterior lumbar loads—a case study approach. J Manip manner in which the pelvis was constrained,
Physiol Ther 20:369, 1997
which is likely to be related to the mechanical
properties of the anterior soft tissues.27 The con¬
straint provided by the anterior soft tissues is
stiffness when evaluation of the mechanical be¬
likely to be applied in at least two ways: to resist
havior of the spine is attempted. One investigation
both anterior displacement and anterior rotation of
found that, when measuring the posteroanterior
the pelvis. Greater thickness of anterior superficial
stiffness of healthy subjects at L3 during 0.5-Hz
soft tissues tends to be associated with lower con¬
mobilization, posteroanterior stiffness was lower
straints to pelvic movements and hence low mobi¬
when associated with breath-holding at the end of
lization stiffness.27 Although recent observations45
a normal expiration than when the stiffness was
have demonstrated a strong negative correlation
measured during breathing.44 This result can prob¬
between soft tissue thickness and lower lumbar
ably be explained by the subjects’ timing of their
posteroanterior stiffness, the hypothesized mecha¬
breathing so that inspiration tended to coincide
nisms to produce this relationship remain uncon¬
with the loading phase of the mobilization, thus
firmed.
increasing the stiffness as a result of the action of
the inspiratory muscles. Clinicians attempting to
Treatment Couch Padding
evaluate stiffness during mobilization need to be
aware of the possible influence of this variable.
Through a mechanism similar to that described for
skin compression, the thickness and compliance of
S kin- F o Id Thickness and Body
the padded surface of the treatment couch on
Mass Index
which the patient is lying can affect the stiffness
It has been observed that both skin-fold thickness felt by the clinician during mobilization.47 A stan¬
(measured over the anterior superior iliac spine) dard treatment couch gives lumbar posteroanterior
and body mass index (mass divided by height stiffness values that are on average 19.2% lower
squared) are significant predictors of the lumbar than those perceived if the patient is assessed on
posteroanterior stiffness in the linear region during an unpadded surface,47 although the change in
mobilization, accounting for up to half of the stiffness with the addition of padding varies sub¬
variance in the stiffness at lower lumbar levels 45 stantially among subjects.
People with small skin-fold thickness or a low
body mass index tend to have greater stiffness. At Other Variable s
least two mechanisms are likely to be responsible
for this relationship. The skin and subcutaneous Many other variables may have an influence on
fat overlying the target bone are compressed by the characteristics of the responses to mobiliza¬
the mobilizing force. Because of the nonlinear tion, including diurnal variation in tissue behavior,
behavior of these tissues,46 most of the compres¬ various pathologic processes within and outside
sion takes place at relatively low loads. Nonethe¬ the spine, postural variation among patients, and
less, beyond the toe region, some compression physiologic range of joint movement. However,
of these compliant superficial soft tissues occurs, there is currently no direct evidence to link these
providing greater displacement under the clini¬ factors with the response to mobilization.
cian’s hands than would otherwise take place. If There is some evidence that the posteroanterior
there is a large initial thickness of the superficial movement response is similar in the thoracic
tissues, there are relatively larger amounts of soft spine,48'49 and it is likely that the factors that
tissue compression and hence more overall dis¬ increase the posteroanterior stiffness in the lumbar
spine are also applicable in the thoracic region.
placement.
Few data, however, are available to support this of resistance. The only joint that has been studied
suggestion, and there is even less information to any great extent appears to be the knee. The
about the responses in the cervical spine. anteroposterior force-displacement relationship at
the tibiofemoral joint has been studied because of
RESISTANCE TO MOBILIZATION AT its demonstrated relationship to cruciate ligament
PERIPHERAL JOINTS integrity.51 In the normal tibiofemoral joint, the
anteroposterior force-displacement relationship
A number of tests of joint responses have been shows only a small zone of low-resistance move¬
made to applied moments to establish the normal ment, but this zone enlarges when there is loss of
pattern of resistance to physiologic movements. restraints at the joint (Fig. 9-12). A similar pattern
Such responses may indicate the general pattern of normal resistance in the intact joint has been
of resistance to mobilizations that employ physio¬ demonstrated for the humeral anteroposterior
logic movements, even though the data were ob¬ movement.52 Logically, this pattern also applies to
tained without any specific attempt to simulate the most other peripheral joints because the structures
loading pattern of mobilization. The general na¬ around the joint are intended to allow movements
ture of the responses of peripheral joints to physio¬ that are controlled by muscles (physiologic move¬
logic moments is shown in a typical curve for ments) but to prevent movements that cannot be
the hip joint in Figure 9-11. The response is controlled by the muscles (accessory movements).
characterized by a region with minimal resistance Factors associated with an increase in stiffness
close to the neutral position and by a gradual, of accessory movements are likely to be similar
nonlinear increase in resisting moment toward the to those established for the spine, as explained
end of range of motion.50 previously. In particular, the degree of activity
This characteristic pattern of responses can also in muscles crossing the joint; the thickness of
be seen in the behavior of other joints, including superficial soft tissue; the frequency of mobiliza¬
the knee, shoulder, and ankle,10 although some tion; the location, direction, and magnitude of
aspects of the response vary depending on the applied force; and the degree of restraint offered
joint and the direction of movement. For example, by the treatment couch could be all expected to
for shoulder movements in most directions, there affect the pattern of the patient’s response.
is a large amount of movement with low resistance
close to the neutral position, whereas ankle flexion
and extension may not show any substantial resis¬ Manipulation
tance-free zone.10
Mobilization using accessory movements is Treatment with manipulation is commonly used
more common than using physiologic movements, for disorders involving the spine. Although most
but little information is available about the pattern often associated with the chiropractic profession,
Knee angle = 0.0 Degrees _

Knee angle = 33 Degrees .
Knee angle = 51 Degrees-
Figure 9-11 Pattern of resistance of the hip

joint to flexion moments. The flexion direction is
positive. One subjects loading and unloading curves
for knee flexion angles of 0, 33, and 51 degrees are
shown. (From Yoon YS, Mansour JM: The passive
elastic moment at the hip. J Biomech 15:905, 1982,
with permission of Elsevier Science.)
Anterior Despite the considerable success of SMT in

alleviating pain and restoring functionality,53'55-61
the mechanisms (mechanical, physiologic, psycho¬
logical, or combinations thereof) by which SMT
continues to achieve the observed beneficial re¬
sults are, as yet, unknown. Because SMT is clearly
a mechanical procedure, several groups of re¬
searchers have endeavored to implement a biome¬
chanical approach in an attempt to understand and
characterize the application and efficacy of SMT.
Biomechanically speaking, the clinician deliv¬
ers a force of specific magnitude, direction, and
time-history to a specific location on the patient.
As such, the clinician is free to alter only a finite
number of variables with respect to the delivery
of a manual treatment. The text that follows sum¬
marizes the salient features of the current knowl¬
edge base regarding the biomechanics of SMT
FIGURE 9-12 Pattern of resistance to anterior and
from the perspective of the external observer (e.g.,
posterior forces at the knee in cases of intact ligaments and
after trauma has occurred. (From Huiskes R, van Dijk R, de
the clinician or scientist).
Lange A, et al: Kinematics of the human knee joint. In Berme
N. Engin AE, Correia da Silva KM [eds]: Biomechanics of FORCE MAGNITUDES AND TIME HISTORIES
Normal and Pathological Human Articulating Joints.
Dordrecht, The Netherlands, Martinus Nijhoff, with kind The earliest attempts at measuring the forces ex¬
permission of Wolters Kluwer Academic Publishers, 1985, p erted by clinicians during SMT were made by
165.) Adams and Wood.62'63 By using a “treatment sim¬
ulator,” they were able to characterize the force-
time-histories exerted by clinicians of varying
spinal manipulative therapy (SMT) is also used by levels of experience, during mock sacroiliac treat¬
groups of physical therapists and other prac¬ ment sessions. Although it was recognized that
titioners. In particular, the site-specific, short- treatment simulations are of limited use for the
lever, high-velocity, low-amplitude spinal adjust¬ understanding of the mechanics of true clinician-
ment is the most commonly used manipulative patient interactions, the results that they obtained
technique. This method of SMT involves the rapid were important in their clear demonstration of the
application of a thrustlike force to the vertebral variable nature of SMT. Later, Triano and
processes (spinal, transverse, or mamillary), the Schultz64 used an inverse dynamics approach to
iliac and ischial spines, or the mastoid, in a spe¬ estimate indirectly the forces exerted by clinicians
cific direction.53'54 onto the cervical spines of subjects. Despite the
Clinicians typically administer SMT in response fact that this approach requires the quantification
to a number of problematic scenarios, as indicated of accelerations of body segments and is thus
primarily by manual diagnoses (palpation) and susceptible to errors when high-velocity, low-am¬
radiography. In summary, these diagnostic scenar¬ plitude movements are involved, their results rep¬
ios include vertebral malposition (a vertebra does resent the first attempt at quantifying the mechan¬
not appear to be in the correct orientation or ics at the clinician-patient interface.
location; that is, it is a question of static position¬ By innovative use of a pressure pad, originally
ing or alignment); abnormal vertebral motion (a designed for measuring foot and insole pressure
vertebra does not appear to be moving in the (EMED Inc., Munich, Germany), Hessel and col¬
correct way; that is, it is a question of dynamic leagues8 were the first to measure directly the
range of motion or direction of motion); abnormal forces exerted by clinicians onto patients during
joint play or end-feel (a purely subjective determi¬ SMT. They clearly showed the distinctive nature
nation of passive motion, elasticity or “give” in of the treatment: the preload force, in which the
a joint within its physiologic range); soft tissue clinician aims to move the targeted vertebral body
abnormalities (palpable changes in paraspinal soft through to the end-point of its physiologic range
tissue, such as tenderness, muscle spasm, and trig¬ of movement; the treatment thrust; and a termina¬
ger points); and muscle contraction or imbalance tion of clinician-patient contact, wherein the ex¬
(palpable spinal contracture in synergistic or op¬ ternally applied forces ultimately drop to zero
posing muscle groups).53 (sometimes referred to as resolution; Fig. 9-13).
Force (N)
Time (s)
FIGURE 9-13 Force-time-histories of spinal manipulative therapy using the Thompson technique on the
sacroiliac joint of a single subject. The solid line and dashed line represent chiropractor I and chiropractor
2, respectively. (Adapted from Hessel BW, Herzog W, Conway PJ, McEwen MC. Experimental measurement
of the force exerted during spinal manipulation using the Thompson technique. J Manipulative Physiol Ther
13[8]:448—453, 1990.)
Since the study by Hessel and colleagues,8 sev¬ To what extent, however, could this mechanism
eral studies employing the same pressure pad sys¬ of manual feedback be expected to be capable of
tem have been conducted (Table 9-2). Compari¬ transducing changes in joint stiffness during the
sons of data among the various studies involving rapid application of the treatment thrusts? Herzog
homologous vertebral sites show that, in general, and associates65 investigated the relationships
preload and peak forces and treatment times tend among the preload forces, the peak forces, and the
to overlap in widely ranging bands. This suggests rates at which those peak forces were achieved
that there may be considerable variation among during SMTs to the cervical, thoracic, and sacroil¬
clinicians treating the same patient (as illustrated iac regions of the spine of a large subject pool.
in Fig. 9-13) and within a particular clinician’s They found, as illustrated in Figure 9-16, that a
approach to treating similarly diagnosed patients significant variation in the peak forces achieved
(Fig. 9-14).
during SMT could be explained by the magnitudes
Specifically, SMTs delivered to the cervical re¬ of the corresponding preload forces applied. As
gion were done faster and used lower peak forces such, they suggested that perhaps it was during
than SMTs delivered to the thoracic or sacroiliac the preload phase of the SMT that the clinician
regions, as illustrated in Figure 9-15. Preloads was interpreting most of the manual feedback as¬
were often so small that they were virtually imper¬ sociated with the physiologic and mechanical con¬
ceptible by the pressure pad. Thus, experienced dition of a particular joint, when the time frame
clinicians appear to be able to judge, through was slow enough that changes in applied force or
direct manual contact, the amount of force re¬ contact area could be made. Because it appeared
quired to achieve the desired result. The more that the change from preload to peak thrusting
compliant joints of the cervical region require less force was effectively brought about by a prepro¬
force than the more massive and stiffer joints of grammed movement, they concluded that control¬
either the thoracic or sacroiliac region. ling the preload force precisely could be an easy
0 h*.
^ Z LO ■'f
5 < </>
“ 2 2 c +1 +1 +1 +1 +1
o 00 Is-* (N O'
F0-*' LO 00 O' r^>
m in _ o o o
(N O
ro
<N
—
LO
f | I? I
LO 2 i d>
2 ^ — o o
(N
o
(N
H
I-
z
rLU UJ 5
7
H 5 2 — o +1
2 *- z oo
(N <N fN
oc O
H
Ul
U O
“ z z
LU
:
O O
LO LO
LO
d)
O
^r
O'
££3 LO (N O'
(N fS
oC —1 O'
nO CM 00 — sO
CO — oC _i — NO LO
z -
+1
2 o 2 z +1 +1 +1 +1 ro — +1 +1 +1
<N —
“-£z 00 O' — — 00 (N o
LO rN O' — O
LO
(N <N ro ro
Summary of Selected Force-Time Characteristics of Spinal Manipulative Therapy
LU LU o
00
U O ^
* z z
5
(N
££~
Q 00 VO
^r <N LO
< “J Z
r-s
0 +1
+1
_i y is Z +1 +1 +1 +1
LU 0 T
cc Li. 2. 00 O' <N O LO O'
00 ro (N <N
CL
h a> 0)

i_
L_ L- L. i_ L_
Z LU D D D C 3 D D D
to
UJ D to to
to to to
amics
to t/) to 00
plate
Z O cu 
w U C3 D X> Q -a D ~U 0> C Q -o Q -D Q -o Q
O "2
< LU ZD LU £ LU £ LU C3
Cl.
u
i_
>> LU
T3
g_£ « w g_m LU g_
£ z z z o z
z1- CO LU LU LU LL. LU
Q) 
0) _c Q)
II
C c u L. “O
LU U o o
to ■5 $

I- H CL i- L. to
E E E >> i 0 C 2
o o o *-> 0 0
_c o CS
_1 U aL
CL |2
oi <N (N (N
LU u U u U
H
l7> U u u u u
E D
r3
X) _C
15 *
TABLE 9-2
>
Q <
D
P 08 _ O
QJ co
txo ro60 ro JZD —3 m
-i >0
(/> ~o
o
<U
to
O'
O'
O£4 O'
O O'
O' O
O O'
O' oo'uSCuSjuS? u x?
£4 O' c O ^ ^ — ^ —
0)
to _
<U <u -r- rci ro (d rt
I I I £ ^ ^ ^ ^
227
Force (N)
FIGURE 9-14 Selected force-time-histories of

spinal manipulative therapy on the thoracic spine
(T4). Treatments were performed by one physician
on three different patients. Treatments were
performed in a posteroanterior direction using a
reinforced hypothenar contact on the left transverse
process of T4. (Adapted from Herzog W, Conway
PJ, Kawchuk GN, et al: Forces exerted during spinal
manipulative therapy. Spine 18:1206, 1993.)
method of potentially controlling the ultimate important clinical skill, particularly with respect
magnitude of the peak thrusting force. to minimizing potentially dangerous manipulative
In a study aimed at investigating the effects of outcomes to which the cervical vertebrae are most
preload and initial positioning of the head on the susceptible.
subsequent movements of the head and thorax
during two distinct methods of cervical SMT, Tri- FORCE DIRECTIONS
ano and Schultz66 came to a similar conclusion:
that clinicians can successfully modify their appli¬ The discussions of force magnitudes thus far have
cation of procedures to control the amplitude and been referring only to the forces directed perpen¬
direction of body segment displacements that arise dicularly to the measuring surface. For example,
during SMT. They concluded that this was an the pressure pad used by Herzog and coworkers67
Force (N)
• I Chapter 9 BIOMECHANICS OF MANUAL THERAPY 229
Peak force (N)

1000
Preload (N)
FIGURE 9-16 Correlation between preload force and peak force for treatments performed on the
sacroiliac joint. (Adapted from Herzog W, Kawchuk GN, Conway PJ: Relationship between preload and peak
spinal manipulative treatments. J Neuromusculoskeletal Syst 1:52, 1993. Reprinted by permission of Journal
of the Neuromuscular Skeletal System, Data Trace Publishing Company.)
was oriented in such a way as to record the poste- diately (within 1 or 2 seconds), clinicians consis¬
rior-to-anterior forces delivered during SMTs to tently delivered the second “satisfactory”
the left transverse process of T4. Even though this treatment faster and to a higher peak force.
particular technique of force delivery appeared to Similarly, Conway and colleagues69 noted that
be distinctly unidirectional, it is entirely possible during an investigation aimed at exploring the
that force vectors other than those measured by cracking sounds often acknowledged to be an indi¬
the pressure pad may play an important role in cation of a successful SMT, the participating clini¬
the outcome of a successful SMT. In an effort to cian chose to follow a noncracking treatment im¬
characterize more specifically the force-time- mediately with another SMT. The force-time
histories of several different types of cervical profile of the second SMT (which did result in a
SMT, Kawchuk and associates68 readily acknowl¬ cracking sound) differed from the first attempt in
edged the likelihood of underestimating the shear that it (again) was delivered more rapidly and to
or laterally directed forces, particularly in rota¬ a higher peak force. Thus, it would appear that
tory-type techniques. The development and imple¬ while a clinician is not likely capable of changing
mentation of a “three-dimensional” force meas¬ the course of a treatment thrust once it has begun,
urement device would be useful in determining he or she is quite adept at effectively perceiving
the specific force components associated with the the mechanical response of the patient and, if
variety of available SMT techniques and would necessary, repeating the treatment with corres¬
be invaluable in the teaching of prospective clini¬ pondingly altered mechanical input, ultimately to
achieve the desired goal.
cians.
MULTIPLE SUCCESSIVE SMT CRACKING SOUNDS
Insofar as a clinician is able to differentiate a The audible cracking sound that so often accompa¬
successful SMT from an unsuccessful one, the nies SMT is likely one of the most universally
option may be taken to repeat the treatment imme¬ and clearly perceived mechanical responses of the
diately. Kawchuk and associates68 asked the partic¬ patient. Varying levels of significance have been
ipating clinicians to identify whether SMTs that ascribed to this sound, ranging from absolute un¬
they had administered were satisfactory or unsatis¬ importance to absolute necessity, for successful
factory. When allowed to repeat a treatment imme¬ manipulative treatment.70
Amplitude (g)
Amplitude (g)
FIGURE 9-17 Acceleration versus time signal for

posteroanterior spinal manipulative therapy at T4,
which caused cavitation. The graphs depicted in A
and B are identical, except for the time scale on the
horizontal axis. (Adapted from Herzog W, Zhang Y,
Conway P, et al: Cavitation sounds during spinal
manipulative treatments. J Manipulative Physiol Ther
16:523, 1993.)
Investigations aimed at understanding the phys¬ similar to the cracking sounds associated with
ical mechanisms of articular sounds were tradi¬ cervical spine manipulation.
tionally conducted using metacarpophalangeal Herzog and colleagues73 showed that by taping
(MCP) joints.71 In his review, Sandoz70 pointed a uniaxial accelerometer to the skin, the vibrations
out some of the features of MCP joint-cracking associated with the joint crack during SMT could
sounds, namely that they were associated with the be measured directly. Figure 9-17 illustrates the
appearance of radiolucent spaces on x-ray film vibrations recorded for a typical posteroanterior
and that an MCP joint could not produce a crack¬ SMT to T4. The lower-frequency triphasic wave
ing sound a second time until a minimum refrac¬ represents the vibration associated with the manip¬
tory period had elapsed. More recently, Meal and ulative thrust. Superimposed on this lower-fre¬
Scott72 have shown that when the finger is sub¬ quency wave is the higher-frequency vibration of
jected to increasing tensile forces, an audible the cracking sound. When a crack was not heard,
cracking sound is associated with a significant the corresponding higher-frequency acceleration
reduction in the resistive tension across the MCP wave was absent. Later, Conway and associates
joint. Further, by using microphones, they showed (unpublished data, 1998) were able to confirm
that the cracking sounds of the MCP joint are
that the higher-frequency accelerometric signals
indeed corresponded to cracking sounds recorded of the treatments, they were unable to attribute
by highly sensitive microphones. Herzog, Con¬ the cracking phenomenon to a single force-time
way, and their colleagues73 concluded not only variable. They suggested that a complex interac¬
that the cracking sound associated with SMT was tion of a number of these variables was responsi¬
clearly perceived by clinicians but also that it ble for eliciting the cracking response and that
was a vibration phenomenon distinct from the unraveling the hierarchy of the potentially im¬
manipulation itself, warranting further study. portant influences would require a more syste¬
Conway and associates69 conducted a study matic investigation. Interestingly, in a more recent
aimed at determining what specific force-time investigation, Reggars and Pollard74 showed that
variables were responsible for triggering the the cracking sound associated with cervical rota¬
cracking sound during SMT. By combining the tory SMT was recorded significantly more often
use of a pressure-sensitive pad with uniaxial accel- on the ipsilateral side of the rotation movement,
erometry, both the posteroanterior forces and ac¬ that is, opposite to the side of the thrust. Their
celerations were synchronously recorded for a se¬ study, and in fact all of the studies previously
ries of manipulative treatments to a current patient mentioned in which joint cavitation had been re¬
pool. In 8 of the 10 trials, the cracking sound and corded by external sound-vibration methodolo¬
corresponding higher-frequency vibration oc¬ gies, suffers from the inability to identify precisely
curred just before the peak manipulative force, as which joint had cracked.
shown in Figure 9-18. In the remaining 2 trials, Although the joint crack has been demonstrated
the crack occurred just after peak thrusting force to be a physical phenomenon that is intimately
had been achieved. During one test, an SMT that associated with SMT, its mechanistic importance
did not result in an audible crack was followed in general and its importance to the success of
almost immediately by a second SMT that did SMT specifically has only recently been investi¬
produce a crack. The second SMT was applied gated and challenged. Traditionally, the audible
more rapidly and to a greater peak force than cracking sound has been thought to be the result of
the initial thrust. In considering all of their data cavitation in the spinal facet joints. The proposed
pertaining to the preloads, peak forces, rates at sequence of events first involves an increase in
which peak thrusts were delivered, and impulses synovial joint space (or volume). The concomitant
Force (N)
FIGURE 9-18 Representative example of force¬

time and cavitation-time signals from one patient
subjected to posteroanterior spinal manipulative
therapy at T4. Here, cavitation occurred before
peak force was achieved. (Adapted from Conway PJ,
Herzog W, Zhang Y, et al: Forces required to cause
cavitation during spinal manipulation of the thoracic
spine. Clin Biomech 8:210, 1993, with permission of
Elsevier Science.)
reduction in synovial joint pressure allows gases milliseconds after the initiation of the manipula¬
to come out of the synovial fluid, in the form of tion, compared with 30 to 200 ms for the normal
a bubble. At the same time, fluid flows into the rapid SMT (see Table 9-2). A uniaxial accelerom¬
area of low pressure, causing the gas bubble to eter was taped to the skin over the spinous process
collapse. The energy that is released when the of the target vertebra to record the cavitation
bubble collapses is perceived as the audible sounds. Finally, the electromyographic (EMG) ac¬
crack.70 Brodeur75 offered an alternative mecha¬ tivity of back muscles just opposite to the area of
nism for the cracking sound. He suggested that treatment was recorded using a pair of standard
the reduction in synovial joint pressure would silver-silver chloride bipolar surface electrodes.
cause the capsular ligaments to invaginate and that For the “fast” SMTs, the investigators consis¬
the cracking sound was caused by the snapping tently observed a significant EMG independent of
back of the capsular ligaments away from the any record of a cavitation response. In contrast,
synovial interface. Although neither of these ex¬ the slowly applied forces were never accompanied
planations for the origins of synovial joint crack¬ by an EMG response, even when a cavitation
ing sounds has been conclusively demonstrated, sound was clearly perceived by the clinician and
both rely on considerable deformation of the recorded by the accelerometer. It appeared that the
joint capsule. audible crack did not (by itself) evoke muscle
activation or a joint proprioceptive reflex re¬
REFLEX MUSCLE CONTRACTION sponse, as had been speculated in the literature.
Suter and coworkers80 conducted a similar but
Deformation of the synovial joint capsule appears larger study, which included more EMG and ma¬
to be at the heart of many hypotheses for the nipulation sites and a larger pool of asymptomatic
mechanisms by which SMT reduces back pain subjects. Fast and slow manipulations were ad¬
and restores functionality. It is thought, for exam¬ ministered to T3, T6, and T9, and electrode pairs
ple, that interarticular adhesions may be broken or were positioned over the back muscles just oppo¬
that entrapped synovial folds may be released dur¬ site the three preselected locations. Additionally,
ing manipulative treatments and that these could a fourth electrode pair was positioned laterally
lead to enhanced intervertebral mobility. More¬ over the posterior inferior serratus muscle. Their
over, it has been speculated that articular facet results showed that reflex muscle contraction was
mechanoreceptors and muscle spindles may be observed only during the regular fast SMTs, about
suitably stimulated during SMT, thereby causing 50 to 200 ms after the onset of the thrusting force.
reflex inhibition of spastic muscles in the treat¬ These bursts of EMG activity would typically last
ment area. This increase in sensory input is also for about 120 ms, well beyond the time at which
thought to reduce the transmission of nociceptive peak force was observed, but clearly disappearing
activity, resulting in a decrease in perceived before the manipulative forces dropped back to
pain.76-78
preload levels. These bursts were observed not
Brodeur75 suggested that the cavitation process only in the back musculature directly opposite to
provided a simple means for initiating a reflex the site of each fast treatment but also in the most
activation of muscles in the area of treatment lateral electrode pair. By considering the timing
application and that without the cavitation process, of the onset of each EMG response, the duration
it would be difficult to generate the forces in of each response, and the fact that the response
the appropriate tissues without causing muscular was consistently observed at all four electrode
damage. Thus, although it appeared that reflex locations, these authors concluded that the most
muscle contraction could be an important aspect likely origin of the observed muscle activity was
of a successful SMT, it was not clear how the associated with the excitation of type II articular
reflex contractions observed were elicited and mechanoreceptors, located in the articular joint
what role, if any, was played by spinal joint cavita¬ capsules of the vertebrae. The ideas proposed by
tion during the manipulative process. Brodeur75 concerning the relationship between
Herzog and colleagues79 conducted a pilot in¬ joint cavitation and muscle reflexes, and the pro¬
vestigation aimed at solving this particular prob¬ tective nature thereof, were not supported by these
lem: Was joint cavitation a prerequisite for the new experimental data.81 Thus, the clinical sig¬
onset of muscle reflex activity? A pressure-sensi¬ nificance of the manipulative joint crack remains
tive pad was used to record the posteroanterior somewhat of a mystery.
force-time-history of two distinct types of manip¬
ulation: the rapid SMT as described earlier, and a MOVEMENTS OF VERTEBRAE
slowly increasing type of force in which peak
The assumption that vertebrae move relative to
magnitudes were achieved several thousands of
one another during SMT has virtually become
manipulative therapy dogma. Certainly, with re¬ nipulative thrusts were as similar to genuine clini¬
spect to the diagnostic scenarios introduced at the cian-patient force-time-histories as possible. Two
beginning of this section, manipulative therapy is electronically synchronized high-speed cine cam¬
thought to be most efficacious for the improve¬ eras were used to record the movements of the
ment or elimination of vertebral pathomechanics, embedded bone pins. This allowed for the calcula¬
that is, for the restoration of normal vertebral tion of posteroanterior translation, lateral transla¬
position and normal intervertebral mobility.53’82> 83 tion, axial rotation, and sagittal rotation of T10,
Yet despite its implicit importance, few investiga¬ Til, and T12 when the SMT was executed at
tions have specifically addressed the problem of each of those vertebral sites (Fig. 9-20).
quantifying the relative movements of vertebral In general, significant relative movements
bodies during SMT. Indeed, the few studies that tended to be small and target specific, such that
do exist have been primarily concerned with quan¬ during an SMT to T10, for example, statistically
tifying spinal deformation during the mobilization significant relative movements were observed be¬
type of manipulative therapy and were thus con¬ tween T10 and Til but not between Til and T12.
sidered earlier in this chapter. When it is remembered that during the preload
As previously described (see Table 9-2), SMT phase of each SMT, the clinician endeavors to
thrusting forces are typically applied rapidly. As move the target vertebra through its physiologic
such, the subsequent vertebral movements could range of motion, the mean relative translations
be expected to be equally rapid. This would pre¬ and rotations observed by Gal and associates84 of
clude the use of bone imaging techniques such as about 0.4 to 1.1 mm, and 0.2 to 1.6 degrees,
radiography, fluoroscopy, and magnetic resonance respectively, take on new significance. Relative
imaging, which are useful when studying static sagittal rotation was particularly interesting in that
situations or when movements take place slowly. it appeared that SMT consistently evoked a hyper¬
Thus, by using an invasive approach with unem¬ extension response between T10, Til, and T12,
balmed post-rigor mortis human cadavers and regardless of which vertebral body was the target
high-speed cinematography (Fig. 9-19), Gal and of the manipulation (Fig. 9-21). Moreover, the
associates84 were able to investigate whether sig¬ hyperextended condition of T11-T12 persisted
nificant relative movements occurred between ver¬ even as the thrusting forces returned to their pre¬
tebrae subjected to SMT. load levels.
Bone pins were embedded in T10, Til, and Although this latter observation tends to support
T12 to project into the sagittal and transverse the hypothesis that SMT may result in realignment
plane. Posteroanterior SMTs were delivered to of vertebrae from unfavorable positions to more
T10, Til, and T12 by a clinician using a rein¬ favorable ones, the duration of this hyperextension
forced hypothenar contact. The posteroanterior remains unknown. It appeared that successive ma¬
forces were recorded using the pressure pad sys¬ nipulations tended to evoke similar magnitudes
tem previously described, to ensure that the ma¬ and directions of absolute translations and rota-
Sagittally oriented camera
Transversely oriented camera
FIGURE 9-19 The experimental setup used by Gal and colleagues84 to record the movements of
vertebrae during posteroanterior spinal manipulative therapy to TIO, Tl I, or TI2 of unembalmed post-rigor
mortis cadavers. (Adapted from Gal JM, Herzog W, Kawchuk GN, et al: Movements of vertebrae during
manipulative thrusts to unembalmed human cadavers. J Manipulative Physiol Ther 20:30, 1997.)
ABC
Figure 9-20 An example of the absolute translations and rotations of TIO (dashed line), Tl I (dotted line),
and TI2 (solid line) for one posteroanterior spinal manipulative therapy to Tl I (A), TI2 (B), and TIO (C) of
an unembalmed post-rigor mortis cadaver. (Adapted from Gal JM, Herzog W, Kawchuk GN, et al:
Movements of vertebrae during manipulative thrusts to unembalmed human cadavers. I Manipulative Physiol
Ther 20:30, 1997.) 7
tions, which suggests that any cumulative effects one another during SMT, even after the vertebrae
of multiple manipulations (separated by about 10 had been forced to the end ranges of physiologic
minutes each) on the mechanics of the cadaveric movement by the applied preload forces. A fully
spines were minimal. Thus, it was concluded that three-dimensional analysis of all six degrees of
the realignment of vertebrae after SMTs was per¬ freedom, using a larger number of unembalmed
haps a short-term observation. Nonetheless, this cadavers, was suggested as the next logical step
study was the first to confirm the long-suspected in attempting to clarify the relationships between
notion that vertebrae do indeed move relative to the externally applied thrusting forces and the
Force
Rotation
tion^r- | Rotation
-^^Rota
FIGURE 9-21 A schematic of the force application during a thoracic treatment to TI I, and the
corresponding rotation of the target vertebra around its longitudinal axis (axial rotation, left) and the
vertebrae immediately adjacent to the target vertebra around their transverse axes (sagittal rotation, right).
(Adapted from Herzog W: Mechanical, physiologic, and neuromuscular considerations of chiropractic
treatments. In Advances in Chiropractic, vol 3. St Louis, Mosby-Year Book, 1996, p 269.)
observed relative movement patterns of vertebrae treatment can be optimized using a process that
during SMT. involves measurement of the force-time-history of
the treatment procedure and correlating aspects of
this with the resulting clinical outcome.
The second major area of progress has been in
Conclusions
the description of movements of the target bones,
largely during manual therapy of the spine. During
To understand the possible mechanisms of action
some types of both SMT and spinal mobilization,
of both manipulation and mobilization techniques,
the intervertebral movements have been estimated.
we need to understand the kinematics and kinetics
The situations that have been studied are some¬
of the procedures. Clinicians need to know the
what limited at this stage, but there is clear evi¬
movement patterns of the bones to which the
dence that a major component of the movement
manual force is applied, including the relative
response to posteroanterior force techniques is lo¬
movements at the joint surfaces and between other
calized intervertebral extension. The sensitivity of
key points on the bones as well as the loads that
the movement responses to variations in technique
are applied through manual therapy to particular
tissues. The major advances in manual therapy or differences between patients has not been well
biomechanics during the past 10 years have studied. In addition, the question of whether the
movements produced by the manual therapy have
largely been in three areas.
First, there has been substantial progress in de¬ a desirable effect in patients with spinal pain has
scribing the applied manual force and its variation not yet been explored.
with time during a single dose of treatment. This The third aspect of manual therapy biomecha¬
type of information is useful for the training of nics that has been examined recently is related
manual therapy practitioners and also contributes mostly to mobilization. Practitioners who use this
to the understanding of how the techniques might form of therapy may use their perception of the
work. For example, a “successful” treatment can displacement that occurs in response to the manual
be characterized in terms of its force-time behav¬ force to attempt to evaluate the stiffness of the
ior, so that practitioners being trained in manual movement response. Recent research has estab¬
therapy can be taught to apply similar levels of lished that certain variables have systematic ef¬
force. It is also possible that aspects of manual fects on the stiffness of the spinal mobilization
movements. Although many of these variables are Articulating Joints. Dordrecht, Martinus Nijhoff
of clinical interest, the mechanisms that are in¬ Publishers, 1985, p 137.
volved in determining the resistance to spinal mo¬ 11. Maitland GD: Vertebral Manipulation, 5th ed.
bilization are not known. Therefore, the interpreta¬ London, Butterworths, 1986.
tion of variations in the stiffness of the 12. Twomey LT: A rationale for the treatment of back
pain and joint pain by manual therapy. Phys Ther
mobilization response is somewhat obscure at this
72:885, 1992.
stage. In particular, we do not know the degree to
13. Herzog W, Nigg BM, Read LJ: Quantifying the
which the response reflects the behavior of any
effects of spinal manipulations on gait using
one intervertebral joint. patients with low back pain. J Manip Physiol Ther
Manual therapy continues to alleviate pain and 11:151, 1988.
restore musculoskeletal functionality. Mechanistic 14. Petty NJ: The effect of posteroanterior
descriptions of precisely how this beneficial effect mobilisation on sagittal mobility of the lumbar
is accomplished remain elusive. We are confident, spine. Man Ther 1:25, 1995.
however, that substantial progress will be made in 15. Loebl WY: The assessment of mobility of
this area of research using technologies that will metacarpophalangeal joints. Rheum Phys Med
11:365, 1972.
allow for more accurate quantification of, and
16. Riddle DL: Measurement of accessory motion:
ultimately the prediction of, the changes in the
Critical issues and related concepts. Phys Ther
internal mechanics of the patient.
72:865, 1992.
17. Kaltenbom FM: Orthopedic manual therapy for
Acknowledgment: Drs. Julianna Gal and Walter Herzog physical therapists. Nordic system: OMT
would like to acknowledge the support of the College Kaltenbom-Evjenth concept. J Manual Manip Ther
of Chiropractors of Alberta, the Chiropractic Founda¬ 1:47, 1993.
tion for Spinal Research, and the Killam Foundation. 18. Lee M, Moseley A, Refshauge K: Effect of
feedback on learning a vertebral joint mobilization
References skill. Phys Ther 70:97, 1990.
19. Threlkeld J: The effects of manual therapy on
connective tissue. Phys Ther 72:893, 1992.
1. Schiotz EH, Cyriax J: Manipulation Past and
20. Goodsell M, Lee M: The effect of spinal
Present. London, William Heinemann Medical
Books, 1975. mobilisation on pain and range of motion in
patients with low back pain. In Proceedings of
2. Kaltenbom FM: Mobilization of the Extremity
12th International Congress of World
Joints. Oslo, Olaf Norlis Bokhandel
Confederation for Physical Therapy. Washington
Universitetsgaten, 1980.
DC, 1995, p 945.
3. Singleton MC, LeVeau BF: The hip joint:
21. Simmonds M, Kumar S, Lechelt E: Use of a
Structure, stability, and stress, a review. Phys Ther
spinal model to quantify the forces and motion
55:957, 1975.
that occur during therapists’ tests of spinal motion.
4. London JT: Kinematics of the elbow. J Bone Joint
Phys Ther 75:212, 1995.
Surg 63-A.529, 1981.
22. Lee M, Liversidge K: Posteroanterior stiffness at
5. Vleeming A, Pool-Goudzwaard AL,
three locations in the lumbar spine. J Manip
Hammudoghlu D, et al: The function of the long
Physiol Ther 17:511, 1994.
dorsal sacroiliac ligament: Its implication for
23. Lee M: Dynamics of the Human Body, 4th ed.
understanding low back pain. Spine 21:556, 1996.
Sydney, Zygal, 1997.
6. Buttermann GR, Janevic JT, Lewis JL: Description
24. Lee M, Svensson NL: Effect of frequency on
and application of instrumented staples for
response of the spine to lumbar posteroanterior
measuring in vivo bone strain. J Biomech
forces. J Manip Physiol Ther 16:439, 1993.
27:1087, 1994.
25. Lee M, Moseley A: Dynamics of the Human
7. Lee R, Evans J: Towards a better understanding of
Body, 2nd ed. Sydney, Zygal, 1991.
spinal posteroanterior mobilisation. Physiotherapy
26. Lee M, Lau T, Lau H: Sagittal plane rotation of
80:68, 1994.
the pelvis during lumbar posteroanterior loading. J
8. Hessel BW, Herzog W, Conway PJ, McEwen MC:
Manip Physiol Ther 17:149, 1994.
Experimental measurement of the force exerted
27. Lee M, Kelly DW, Steven GP: A model of spine,
during spinal manipulation using the Thompson
ribcage and pelvic responses to a specific lumbar5
technique. J Manip Physiol Ther 13(8):448^453
manipulative force in relaxed subjects. J Biomech
1990.
28:1403, 1995.
9. Matyas TA, Bach TM: The reliability of selected
28. Lee M, Kelly DW, Steven GP: Lumbar spine
techniques in clinical arthrometrics. Aust J
manual therapy: Effect of choice of target
Physiother 31:175, 1985.
vertebra. In Proceedings of the First Australian
10. Engin AE: Passive and active resistive force
Biomechanics Conference. The University of
characteristics in major human joints. In Berme N, Sydney, Sydney, 1996, p 104.
Engin AE, Correia da Silva KM (eds):
29. Paris SV: Mobilization of the spine. Phys Ther
Biomechanics of Normal and Pathological Human 59:988, 1979.
30. Lee M, Svensson NL: Measurement of stiffness of skin and underlying tissues in vivo.
during simulated spinal physiotherapy. Clin Phys Biomaterials 4:305, 1983.
Physiol Meas 11:201, 1990. 47. Latimer J, Holland M, Lee M, et al: Plinth
31. Latash ML, Zatsiorsky VM: Joint stiffness: Myth padding and measures of posteroanterior lumbar
or reality? Hum Movement Sci 12:653, 1993. stiffness. J Manip Physiol Ther 20:315, 1997.
32. Lee M, Latimer J, Maher C: Normal response to 48. Lee M, Latimer J, Maher C: Manipulation:
large postero-anterior lumbar loads—a case study Investigation of a proposed mechanism. Clin
approach. J Manip Physiol Ther 20:369, 1997. Biomech 8:302, 1993.
33. Latimer J, Lee M, Adams R, et al: An 49. Panjabi MM, Brand RA, White AA: Three-
investigation of the relationship between low back dimensional flexibility and stiffness properties of
pain and lumbar posteroanterior stiffness. J Manip the human thoracic spine. J Biomech 9:185, 1976.
Physiol Ther 19:587, 1996. 50. Yoon YS, Mansour JM: The passive elastic
34. Lee M, Esler M-A, Mildren J, et al: Effect of moment at the hip. J Biomech 15:905, 1982.
extensor muscle activation on response to lumbar 51. Huiskes R, van Dijk R, de Lange A, et al:
posteroanterior forces. Clin Biomech 8:115, 1993. Kinematics of the Human Knee Joint. In Berme
35. Lee M, Steven GP, Crosbie J, et al: Variations in N, Engin AE, Correia da Silva KM (eds):
posteroanterior stiffness in the thoraco-lumbar Biomechanics of Normal and Pathological Human
spine: Preliminary observations and proposed Articulating Joints. Dordrecht, Martinus Nijhoff
mechanisms. Phys Ther 78:1277, 1998. Publishers, 1985.
36. Shirley D, Lee M: A preliminary investigation of 52. Lee M: Assessment of the glenohumeral joint
the relationship between lumbar postero-anterior through passive accessory movements. In Current
mobility and low back pain. J Manual Manip Ther Concepts in the Management of Shoulder Region
1:22, 1993. Dysfunction. Sydney, Manipulative Therapists
37. Pauly JE: An EMG analysis of certain movements Association of Australia, 1988, p 41.
and exercises. I. Some deep muscles of the back. 53. Haldeman S: Spinal manipulative therapy: A
Anat Rec 155:223, 1966. status report. Clin Orthop Rel Res 179:62, 1983.
38. Shirley D, Lee M: Muscle activity as a possible 54. Buerger AA: A non-redundant taxonomy of spinal
contributing factor to lumbar posteroanterior manipulative techniques suitable for physiologic
stiffness, p. 145 In Proceedings of the 9th explanation. Man Med 1:54, 1984.
Biennial Conference of the Manipulative 55. Greenland S, Reisbord LS, Haldeman S, Buerger
Physiotherapists Association of Australia, AA: Controlled clinical trials of manipulation: A
Manipulative Physiotherapists Association of review and a proposal. J Occup Med 22:670,
Australia, Gold Coast, 1995. 1980.
39. Bogduk N, Twomey LT: Clinical anatomy of the 56. Deyo RA: Conservative therapy for low back
lumbar spine, 2nd ed. Melbourne, Churchill pain. JAMA 250:1057, 1983.
Livingstone, 1991. 57. Herzog W: Biomechanical studies of spinal
40. Crisco JJ, Panjabi MM: Postural biomechanical manipulative therapy. J Can Chiropract Assoc
stability and gross muscular architecture in the 35:156, 1991.
spine. In Winters JM, Woo SL-Y, eds: Multiple 58. Boline PD, Kassek K, Bonfort G, Nelson C:
Muscle Systems: Biomechanics and Movement Spinal manipulation vs. amitriptyline for the
Organization. New York, Springer-Verlag, 1990, p treatment of chronic tension-type headaches: A
438. randomized clinical trial. J Manip Physiol Ther
41. Fung YC: Biomechanics: Mechanical properties of 18:148, 1995.
living tissues, 2nd ed. New York, Springer-Verlag, 59. Dabbs V, Lauretti WJ: A risk assessment of
1993. cervical manipulation vs. NSAIDs for the
42. Viner A, Lee M: Direction of manual force treatment of neck pain. J Manip Physiol Ther
applied during assessment of stiffness in the 18:530, 1995.
lumbosacral spine. J Manip Physiol Ther 18:441, 60. Nilsson N, Christensen HW, Hartvigsen J: Lasting
1995. changes in passive range of motion after spinal
43. McGlashen KM, Miller JAA, Schultz AB, et al: manipulation: A randomized blind controlled trial.
Load displacement behavior of the human lumbo¬ J Manip Physiol Ther 19(3): 165—168, 1996.
sacral joint. J Orthop Res 5:488, 1987. 61. Vernon H: Spinal manipulation for chronic low
44. Beaumont A, McCrum C, Lee M: The effects of back pain: A review of the evidence. J Can
tidal breathing and breath-holding on the postero¬ Chiropract Assoc 40:180, 1996.
anterior stiffness of the lumbar spine. Proceedings 62. Adams A, Wood J: Comparison of forces used in
of the 7th Biennial Conference of the selected adjustments of the low back: A
Manipulative Physiotherapists Association of preliminary study. The Research Forum. Palmer
Australia, 1991, p 244. College of Chiropractic. 1:5, 1984.
45. Viner A, Lee M, Adams R: Postero-anterior 63. Wood J, Adams A: Forces used in selected
stiffness in the lumbosacral spine: The relationship chiropractic adjustments of the low back: A
of stiffness between adjacent vertebral levels. preliminary study. The Research Forum. Palmer
Spine 22:2724, 1997. College of Chiropractic. 1:16, 1984.
46. Bader DL, Bowker P: Mechanical characteristics 64. Triano JJ, Schultz AB: Cervical spine
manipulation: Applied loads, motions and 75. Brodeur R: The audible release associated with
myoelectric responses. Proceedings of the 14th joint manipulation. J Manip Physiol Ther 18:155,
Annual Meeting of the American Society of 1995.
Biomechanics, Miami, vol 14. 1990, p 187. 76. Raftis K, Warfield CA: Spinal manipulations for
65. Herzog W, Kawchuk. GN, Conway PJ: back pain. Hosp Pract 15:89, 1989.
Relationship between preload and peak forces 77. Zusman M: Spinal manipulative therapy: Review
during spinal manipulative treatments. J of some proposed mechanisms, and a new
Neuromusculoskeletal Sys 1:52, 1993. hypothesis. Aust J Physiother 32:89, 1986.
66. Triano JJ, Schultz AB: Motions of the head and 78. Gillette RG: A speculative argument for the
thorax during neck manipulations. J Manip coactivation of diverse somatic receptor
Physiol Ther 17:573, 1994. populations by forceful chiropractic adjustments:
67. Herzog W, Conway PJ, Kawchuk GN, et al: A review of the neurophysiological literature. Man
Forces exerted during spinal manipulative therapy. Med 3:1, 1987.
Spine 18:1206, 1993. 79. Herzog W, Conway PJ, Zhang Y, et al: Reflex
68. Kawchuk GN, Herzog W: Biomechanical responses associated with manipulative treatments
characterization (fingerprinting) of five novel on the thoracic spine: A pilot study. J Manip
methods of cervical spine manipulation. J Manip Physiol Ther 18:233, 1995.
Physiol Ther 16:573, 1993. 80. Suter E, Herzog W, Conway PJ, et al: Reflex
69. Conway PJ, Herzog W, Zhang Y, et al: Forces response associated with manipulative treatment of
required to cause cavitation during spinal the thoracic spine. J Neuromusculoskeletal Sys
manipulation of the thoracic spine. Clin Biomech 2:124, 1994.
8:210, 1993. 81. Herzog W: On sounds and reflexes. J Manip
70. Sandoz R: The significance of the manipulative Physiol Ther 19:216, 1996.
crack and of other articular noises. Ann Swiss 82. Cassidy JD, Potter GE: Motion examination of the
Chiropract Assoc 4:47, 1969. lumbar spine. J Manip Physiol Ther 2:151, 1979.
71. Roston JB, Haines RW: Cracking in the 83. Carrick FR: Treatment of pathomechanics of the
metacarpophalangeal joint. J Anat 81:165, 1947. lumbar spine by manipulation. J Manip Physiol
72. Meal GM, Scott RA: Analysis of the joint crack Ther 4:173, 1981.
by simultaneous recording of sound and tension. J 84. Gal JM, Herzog W, Kawchuk GN, et al:
Manip Physiol Ther 9:189, 1986. Movements of vertebrae during manipulative
73. Herzog W, Zhang Y, Conway P, et al: Cavitation thrusts to unembalmed human cadavers. J Manip
sounds during spinal manipulative treatments. J Physiol Ther 20:30, 1997.
Manip Physiol Ther 16:523, 1993. 85. Kawchuk GN, Herzog W, Hasler EM: Forces
74. Reggars JW, Pollard HP: Analysis of generated during spinal manipulative therapy of
zygapophyseal joint cracking during chiropractic the cervical spine: A pilot study. J Manip Physiol
manipulation. J Manip Physiol Ther 18:65, 1995. Ther 15:275, 1992.
CHAPTER 10
Lower Limb Orthotics
John Stallard
Historically considered a last resort, to be used often is named for the person who is alleged
only when all else fails, orthoses have in recent to have pioneered the design. To overcome this
years become a more respected treatment option potentially troubling problem, the American Acad¬
for many conditions. Orthoses have increasingly emy of Orthopaedic Surgeons proposed that ortho¬
seen as having the potential to provide a positive ses should be defined by the joints of the body
contribution to the rehabilitation of severely hand¬ they cross. Thus, a long-leg caliper is described
icapped patients, and orthotists are now commonly as a knee-ankle-foot orthosis (KAFO). A knee
included in the clinical team. Many enlightened brace is called a knee orthosis (KO), a wrist splint
practitioners1-2 advocate therapy regimens in a wrist-hand orthosis (WHO), a below-knee brace
which the elements of surgery, physical therapy, an ankle-foot orthosis (AFO), and so on. This
and orthotics are integrated at the beginning of system has now been widely adopted and is
treatment to ensure the optimum functional out¬ accepted as an International Standard (ISO
come for the patient. An understanding of the 8549-3:1989 E/F: Prosthetics and orthotics-
potential of orthotics to contribute to the well¬ Vocabulary-Part 3: Terms relating to external or¬
being of a patient is therefore an important part of thoses). The adoption of this system is becoming
the armory of skills for modem clinicians, what¬ ever more common, and this has certainly been a
ever their discipline. boost to international exchange of information,
Confusion about the role of orthotics does still which in turn is fostering an improvement in treat¬
exist. It is therefore necessary to have a clear ment in most countries.
understanding of its scope. Being a comparatively There are surprisingly few standards governing
modem profession, there are still a number of the design and manufacture of orthotic devices.
definitions that are applied. Two of these give a This is more a reflection of the paucity of data on
good indication of the role of the orthotist and the the loading applied to these devices than of the
scope of orthotics: lack of need for such documents. In prosthetics
(external artificial limbs), international coopera¬
“An external device designed to apply or remove tion has led to funding for research to establish
forces to or from the body with the object of performance specifications for lower limb devices,
supporting, correcting or for an anatomical and an ISO Standard3 has recently been published.
deformity or weakness, however caused. It The loads on orthoses are much less easily de¬
may be applied with the additional object of fined, and the research to establish an equivalent
assisting, allowing or restricting movement of standard would therefore be considerably more
the body” (UK DHSS definition). expensive. This has inhibited the financing of
“An externally applied device used to modify the work that is accepted as being as necessary as that
structural or functional characteristics of the achieved in the field of prosthetics. Initiatives are
neuro-muscular-skeletal system” (International
being pursued, and it is hoped that international
Standards Organisation definition). collaboration will one day make performance
There is a wide range of orthotic devices, and specifications available to professionals working
they are given a variety of names. This can cause in the field of orthotics.
great confusion because local custom frequently To make the most of an orthosis, it is important
determines the description of an orthosis, which to understand the principles on which it works
239
and the potential benefits it can provide for the permits a sharper focus on the problem areas and
patient receiving rehabilitation care. This means a better grasp of the mechanical controls that must
that, ideally, not only must the function of an be applied to achieve the most practical outcome
orthosis and the objective of the forces it applies for an individual patient.
or resists be appreciated, but also the long-term Normal ambulation is an effective compromise
outcome for the patient should be identified. In a that enables people to function in a wide range of
young profession such as orthotics, there is not conditions. It is generally observed while subjects
yet a wealth of research data establishing long¬ are walking on a flat surface that allows good
term outcomes for particular treatments. However, grip, but it adapts well to slopes, rough ground,
the scientific background is growing rapidly, and and different amounts of friction. Gait has evolved
clinical research is commonly regarded as a neces¬ to minimize energy expenditure, consistent with
sary requirement to underpin new developments. maintaining its inherent flexibility, to permit adap¬
Consequently, it is now possible to establish evi¬ tation to different conditions. To achieve overall
dence of efficacy for some orthoses, and it should efficiency, the normal gait cycle is arranged to
be expected that this body of information will minimize excursions of the general body center
grow exponentially as the professions allied to of gravity. Other mechanisms of normal gait are
medicine increasingly turn their attention to the invoked to smooth impact loads by absorbing en-
need to produce proven findings. ergy (e.g., in muscles and tendons), some of which
The contributions that orthotics can make to can be recovered later in the gait cycle.
the treatment of patients fall under a number of It can be seen, therefore, that walking consists
headings. Although there is no formally adopted of complex and sophisticated interactions of a
system of clinical categorization of orthoses, a number of simultaneous mechanisms that demand
commonly adopted practice is to group them ac¬ a high degree of control. Whenever neurologic
cording to the area of the body they are intended control mechanisms are compromised, gait inevi¬
to treat. This logical approach is a convenient tably increases in difficulty and decreases in effi¬
means of structuring an examination of the sub¬ ciency. The greater the degree of the control defi¬
ject. Lower limb orthoses perform a wide range cit, the more extensive the handicap. Lower limb
of functions, almost all of them related to ambula¬ orthoses can compensate for lack of subtle neuro¬
tion for handicapped patients. Because all lower muscular control through the use of cruder, pure
limb devices have biomechanical principles that mechanical controls. Although these have the ca¬
are related to the functional requirements, an es¬ pacity to permit ambulation, this inevitably has a
sential starting point is a knowledge of the funda¬ lower efficiency than normal walking4 because
mentals of walking. these orthoses do not permit the subtle interplay
of finely controlled biomechanics.
It is a great advantage in identifying the most
The Walking Cycle
appropriate mechanical controls for handicapped
Lower limb orthoses are prescribed mainly to treat gait to have a knowledge of the normal gait cycle.
walking disorders. The degree of handicap and the The normal gait cycle (Fig. 10-1) is generally
commensurate complexity of the biomechanical considered to have six distinct major events (on
both sides), as follows:
solution can vary from the simple to the highly
complex. Successful treatment clearly demands an 1. Heel-strike
accurate evaluation of the problems faced by the 2. Foot-flat
patient, so that the appropriate biomechanics to 3. Mid-stance
ensure the most effective outcome can be 4. Heel-off
achieved. Analysis of handicapped gait is en¬ 5. Toe-off
hanced by an understanding of normal gait. It 6. Mid-swing
Chapter 10 LOWER LIMB ORTHOTICS 241
This cycle repeats with heel-strike. The tempo¬ stages of the stance phase and reverses this during
ral relationships between these events lead to rec¬ the later stages, while the pelvis rotates downward
ognizable phases of the gait cycle and have an slightly around the stance hip in mid-stance.
important influence on performance. In the sagittal Hardly any deviation occurs in hip or knee joints
plane, viewed from the right side of the body, in the coronal plane, although supination and pro¬
these relationships are as follows: nation of the foot respond to unevenness of the
Heel-strike to foot-flat. At right heel-strike the
ground.
leg is in front of the trunk, with the hip at its Vertical movements of the trunk inevitably oc¬
cur as lower limbs flex and extend around the
maximum degree of flexion (while walking),
hip joint. These vertical deviations are attenuated,
the knee slightly flexed, and the ankle in the
however, by knee flexion and slight hip adduction
neutral position. Immediately after the heel
in mid-stance to reduce the peak rise, and foot
contacts the ground, the ankle joint absorbs
plantar-flexion at the end of stance phase lessens
energy by plantar-flexing under eccentric
the drop of the trunk. These subtle movements
muscle control until the foot is flat on the
minimize the excursion of body center of gravity
ground, while the knee flexes slightly and the
(Fig. 10-2) and thereby lessen the dynamic forces
hip maintains its flexed position. This phase
required to raise the mass of the body, reducing
takes about 10% of the cycle and is also
the energy cost of ambulation.
known as the ankle rocker.
Any form of walking requires that the body has
Mid-stance. In this phase of the cycle, the foot internal stabilization to prevent it from collapsing
remains flat on the ground as the rest of the under its own weight, external stabilization so that
body progresses forward over the ankle joint, it does not topple over, propulsive forces, and
which moves progressively from plantar- control of the stabilization and propulsive forces
flexion to dorsiflexion while the hip joint to ensure organized progression. The subtle inter¬
simultaneously moves from flexion into play of the different phases and events in the
extension. The knee initially flexes but begins normal gait cycle demands an excellent control
to extend before the lower limb and trunk system that organizes the muscles to provide inter¬
reach vertical alignment (at which point it is nal stabilization (by balancing moments generated
still noticeably flexed), and the left leg at its around the joints of the lower limbs), the posture
mid-swing passes the right and continues to do so that body center of gravity is always within an
so until the phase is completed. This phase effective (usually very small) support area, and the
occurs from about 10% to 45% of the gait injection of propulsive forces to provide forward
cycle. progression. Absence of effective control or mus¬
Heel-off to toe-off. During this phase (sometimes cle power inhibits walking ability and requires
known as push-off), the right ankle plantar- compromises to be made (involuntary or con¬
flexes more rapidly while flexion of both the scious), with the inevitable consequence that effi¬
right hip and knee occurs, and the left leg ciency is diminished. Proper analysis of handi¬
achieves heel-strike. This phase occurs from capped gait is essential if the most effective
about 45% to 60% of the gait cycle.
Swing phase. This phase commences as the right
toe leaves the ground. The thigh accelerates
from extension into flexion, imparting forward
motion to the limb. Initially, the knee
continues to maximum flexion while the ankle
dorsiflexes from plantar-flexion to neutral to
ensure ground clearance as the left leg reaches
mid-stance with the right leg passing it
halfway through swing phase. As flexion at the
hip develops, the thigh decelerates, and the
knee moves into extension in preparation for
heel-strike, which completes the cycle. This
phase of gait occurs from approximately 60%
FIGURE 10-2 Vertical excursion of the body’s center of
to 100% of the cycle.
gravity (C of G). (From ORLAU [Orthotic Research and
There are also some subtle relationships oc¬ Locomotor Assessment Unit], The Robert Jones and Agnes
curring in the coronal plane. The trunk moves Hunt Orthopaedic and District Hospital, Oswestry,
slightly toward the support side during the early Shropshire, England SYI0 7AG.)
compromise is to be achieved for each patient. those joints. To permit a more immediately ob¬
This requires not only a good understanding of servable indication of the aberrant moments (turn¬
normal gait but also the means of establishing the ing effects) that may be present in a patient, the
temporal, force, and phasic relationships occurring Video Vector Generator (VVG)5 has been devel¬
in handicapped ambulation through appropriate oped; it converts the force platform signals into a
methods of evaluation. line on a video screen, accurately aligned to the
image of the subject as he or she walks over
the platform. This constitutes a vector of ground
Methods of Evaluation reaction force, the length of which is proportional
to the magnitude and direction to the alignment.
One of the problems of judging the effectiveness The video image, therefore, gives an immediate
of orthotic treatment has been the lack of evalua¬ indication of the effect of ground reaction forces
tion techniques or instrumentation that can readily on the lower limb. In normal walking (Fig. 10-4),
be applied in routine clinical situations. These are the ground reaction force in the sagittal plane
becoming available, and it seems likely that it will remains close to the hip and knee joints through¬
be increasingly possible to make valid scientific out the stance phase of gait. The system shows
observations of the effect of applying an orthosis any deviations from this due to pathologic condi¬
to ameliorate the problems created by biomechani¬ tions and allows an evaluation of their effects on
cal abnormalities. Three methods serve as an ex¬ an individual patient.
ample of the means of monitoring patients to
establish outcomes of orthotic treatment: MONITORING RELATIVE ENERGY
EXPENDITURE
• Measuring ground reaction forces to permit
qualitative and quantitative assessment of
Many orthotic treatments are intended to ease the
biomechanical alignment relative to the joints
burden of walking for severely handicapped pa¬
of the lower limbs
tients. It is not always easy to determine subjec¬
• Evaluating relative energy expenditure by
tively whether that objective has been realized.
measuring speed of walking and heart rate to
Patients, for a variety of reasons, are not necessar¬
produce a physiologic cost index (PCI)
ily reliable reporters in this respect. It is a valuable
• Classification of handicapped gait through
adjunct to the rehabilitation process to have a
allocation of appropriate categories that have
quantitative analysis of relative effort for an indi¬
specific determinants
vidual patient walking before and after treatment.
The scientifically preferred method of measuring
MONITORING GROUND REACTION FORCES oxygen uptake is not practical for many handi¬
capped patients because the apparatus required
Ground reaction forces are monitored by force (which incorporates tight-fitting face masks and
platforms. These are devices placed within a walk¬ cumbersome instrumentation) affects performance
way that follow the rapidly changing forces oc¬ to an unacceptable degree. A simpler method for
curring between the foot and the ground during the patient is to monitor speed and heart rate. This
the stance phase of gait. The output from the can be done by picking up the electrocardiogram
force platform is given as force in three mutually signal through two electrodes on the chest and
perpendicular directions, referred to as Fx, Fy, Fz transmitting it to instrumentation that processes it
(Fig. 10-3A):
to give a frequently updated reading of heart rate.
• The vertical (FJ Speed can be calculated over an appropriate num¬
• The horizontal in the line of walking (Fy) ber of timed circuits of a suitable course de¬
• The horizontal at right angles to the line of pending on the level of handicap (e.g., a figure-of-
walking (Fx) eight 25-meter circuit) to establish a comparative
standard. The raw data can be converted into a
These three planar outputs for a healthy subject single figure of PCI by dividing the increase in
walking at his or her usual pace are shown in heart rate from the resting rate by the speed of
Figure 10-35, as a percentage of body weight. walking,^ which gives a reading in beats per
Although used widely in biomechanics, they are meter walked.
difficult to interpret for pathologic conditions be¬ This method is comparatively simple to apply,
cause they give little indication of the effects that does not create unacceptable difficulties for the
abnormalities have on the alignment of forces patient, and, provided that the selected test is used
relative to the joints of the lower limb and the rigorously before and after treatment, gives an
turning effects they therefore generate around outcome that is easy to interpret. Although there
Chapter 10 Lower Limb Orthotics 243
g Time (s)
FIGURE 10-3 Ground reaction forces. A, The three force components of gait. B, Ground reaction forces
in normal walking. (A, from ORLAU [Orthotic Research and Locomotor Assessment Unit], The Robert
Jones and Agnes Hunt Orthopaedic and District Hospital, Oswestry, Shropshire, England SYIO 7AG.)
FIGURE 10-4 Video vector display of ground reaction force in a healthy subject during stance phase
range.
is a wide spectrum of PCIs for healthy subjects proaches have not be identified. In doing so, it is
(ranging from 0.2 to 0.6 beats/min7), patients have essential to ensure that the test has good repeat¬
repeatable results. The test is not a good way of ability and measures effectively the parameter of
comparing different patients but is a valid method interest without interference from other factors. It
of monitoring change within a single subject as a is therefore best, if possible, to use a previously
result of rehabilitation treatment. validated test procedure.
CLASSIFICATION OF HANDICAPPED GAIT

Engineering Considerations
A simple but effective way of monitoring change
in handicapped patients is to classify their walking BIOMECHANICS
ability according to well-defined categories. Hof-
fer and colleagues9 defined four categories for a The most common objective of most orthotic de¬
series of myelomeningocele patients for whom vices is to resist moments generated around unsta¬
they were providing treatment. These categories, ble joints. Each unstable joint requires a minimum
which have subsequently been widely used and of three forces to generate an equal and opposite
slightly adapted, are now generally accepted as moment to that which is causing the instability.
follows: For example, in Figure 10-5, body weight acting
Chairbound patients: patients who never walk vertically produces a moment around the knee
and use a wheelchair as their main means of joint of 700 N times 0.15 m (i.e., 105 N m). This
locomotion is resisted in the flail limb by the so-called three-
point fixation of the caliper, which produces a
Therapeutic walkers: patients who use apparatus
balancing moment through the action of the three
to walk primarily for therapeutic purposes and
forces, Fls F2, and F3, acting through their respec¬
do so generally under supervision
tive moment arms. The magnitude of the forces
Household walkers: patients who use apparatus can be minimized by increasing the moment
to walk independently for functional purposes arms to the maximal practical distance available
within the protected environment of home or on the caliper.
special school
Community walkers: patients who use apparatus
to walk independently within the community
None of these categories imply that patients
have given up the use of a wheelchair, which
remains the most efficient form of moving around
on a flat surface for most severly handicapped pa¬
tients.
Hoffer’s classification9 is a simple means of
providing broad categories of ambulation for mon¬
itoring long-term change within an individual pa¬
tient or for making general comparisons of out¬
come for different treatments. More refined
systems of comparison have grown out of this
simple (but effective) approach, and Garret and
Meehan,10 among others, have extended the con¬
cept for more sophisticated evaluation.
SUMMARY
FIGURE 10-5 Line of force relative to a flexed knee and
These are but three examples of the means of the equal and opposite moment generated by three-point
making clinical comparisons to assess the effec¬ fixation applied by a knee-ankle-foot orthosis (dotted area).
tiveness of rehabilitation treatments in which or- Knee moment = W X L (W = body weight; L = moment
arm). F| = F2 + F3; together they produce the equal and
thotics may play a part. Others exist to suit differ¬
opposite moment to W X L (F„ F2, and F3 are the forces
ent situations, and an appropriate literature search
applied by the orthosis). (From ORLAU [Orthotic Research
could reveal more appropriate methods for partic¬
and Locomotor Assessment Unit], The Robert Jones and
ular situations. Alternatively, tests can be designed Agnes Hunt Orthopaedic and District Hospital, Oswestry,
to cover comparisons for which suitable ap¬ Shropshire, England SYI0 7AG.)
•I Chapter 10 LOWER LIMB ORTHOTICS 245
Minimizing the forces is advantageous in that are more likely to produce overloads than those
it reduces the loads applied directly to the patient who are more sedentary, even if the latter are
through the orthosis. High interface pressures be¬ heavier.
tween the orthosis and the patient can lead to
discomfort or, more seriously, particularly in anes¬
thetic skin, to pressure sores. To reduce the risk, MATERIALS
it is necessary not only to minimize the forces by
creating the longest possible moment arms but An objective of prescribing orthoses is usually
also to spread the load over as large an area to improve the function of the patient. Success
as possible. inevitably leads to the generation of higher dy¬
Little research has been done to establish the namic forces from increased activity, and it would
loads applied to orthoses during patient activities. be counterproductive to attempt to limit the func¬
It is clear that the highest loads are likely to occur tion of the patient to protect the orthosis. It must
in lower limb orthoses as patients ambulate. In the be recognized that there is the possibility for over¬
absence of real research data, there is a temptation loads to occur on orthotic structures, and this must
to imagine these as being related merely to patient greatly influence the choice of material. One of
body weight. This is to ignore the potential of the most important properties of a material is the
dynamic loading during vigorous activity. When¬ way in which it fails, and there are essentially two
ever a body changes speed or direction, forces modes of failure: brittle and ductile.
over and above those required to maintain the The characteristics of these two types of mate¬
status quo are required to produce those changes. rial failure are discernible from stress-strain dia¬
These are known as dynamic forces and can be grams (Fig. 10-6A, brittle, and Fig. 10-65, duc¬
very significant. tile). A straight-line relationship between stress
During normal walking, the vertical ground re¬ and strain implies an elastic material that stores
action force rises to a maximum of 20% to 30% strain energy. As stress increases (i.e., load is
above body weight (see Fig. 10-35). More vigor¬ applied), equal increments of stress produce equal
ous activity causes the dynamic forces to rise increments of strain (i.e., deformation). When the
dramatically. Running, for example, creates verti¬ load is removed, both stress and strain return to
cal ground reaction forces of about three times zero (back down the straight line), and there is no
body weight. Unexpected incidents can lead to net deformation.
even higher forces, which are likely to lead to Brittle materials are essentially purely elastic.
accidental overloads on orthotic structures. It is They follow the straight-line relationship as load
well within the bounds of possibility that loads is increased until fracture occurs, which happens
10 times body weight could occur, either from suddenly and with no warning. The area under a
unexpected patient activity (e.g., jumping from a load deformation graph (i.e., stress-strain) repre¬
moving bus or abseiling for a hobby) or from sents the energy required to cause failure. With
unforeseeable accidents, and it is wise to bear brittle materials (which are purely elastic), this
these possibilities in mind when assessing the or¬ energy is released at fracture, and fragments are
thotic needs of a patient. Highly active patients imparted with kinetic energy, which, in orthotic
Fracture Absorbed Fracture
FIGURE 10-6 Stress-strain relationships for elastic materials. A, Brittle material. B, Ductile material.
devices that are overloaded, creates the potential poses. Brittle materials have low tolerance of
for patient injury. overload and are therefore inherently unsafe. Duc¬
With ductile materials, the straight-line portion tile materials, on the other hand, can ensure that
of the curve is followed by a change in character¬ patients receive maximum protection from the po¬
istic. These materials reach a point at which they tentially catastrophic consequences of a fracture
are no longer able to sustain elastic behavior, in a material that may be very close to vulnerable
but they do not fracture when this occurs. Equal areas of the body.
increments in stress produce greater increments of When materials are cyclically loaded, as many
strain. The point at which this happens is known orthotic devices are (particularly those applied to
as the yield point, and the slope of the graph the lower limb), fatigue failure becomes a possi¬
reduces dramatically. After yield has occurred, the bility. This means that the material fails at load
deformation becomes permanent and is known levels below those it can safely accommodate
as plastic deformation. In this phase, energy is when it is new or when the load is not cyclically
absorbed within the material (mostly converted applied. The higher the loads relative to the
to heat). strength of the device, the lower the number of
The stress point at which a material fractures is load cycles to cause fatigue failure. This is an
known as the ultimate strength of the material. important failure mechanism throughout the field
For ductile materials, the stress point at which of orthopedic engineering but has particular rele¬
yield occurs is the yield strength, and this is the vance to lower limb orthoses. Failure commences
point at which failure begins to occur. at a point of high stress concentration, which is
A further important characteristic of materials caused by sharp changes of section or holes within
is stiffness. The slope of the straight-line portion the device, scratches on the surface, or damage
of the graph for elastic materials is known as the from general wear or manufacturing processes.
modulus of elasticity. It is a direct measure of the The localized failure due to fatigue causes a crack
amount of deformation that occurs under applied in the material that does not lead to immediate
stress and therefore defines the stiffness of the fracture of the device. Propagation of the crack
material. When deformation must be resisted, a occurs as the cyclic load is continued, and this
high modulus of elasticity is necessary (i.e., a accelerates as the general strength of the device
steep slope on the graph). For flexibility, a lower diminishes in response to the growth of the fa¬
modulus of elasticity (less vertical slope) is re¬ tigue-induced flaw, eventually leading to cata¬
quired. Achieving an appropriate compromise for strophic failure. Functional orthotic devices should
structural stiffness in an orthosis is often a chal¬ be inspected regularly to ensure that they remain
lenge faced by designers. Material stiffness (as safe for the patient to use, and an examination for
well as section modulus in the structure) plays an signs of fatigue should be a routine part of such
important role in ensuring the required characteris¬ a process.
tics in the device.
Figure 10-6A and B shows that for materials STRUCTURAL PROPERTIES
with the same ultimate strength and stiffness, a
ductile material demands much greater energy to As already indicated, the most common loads ap¬
cause fracture than does a brittle material (al¬ plied to orthotic structures are those that occur in
though permanent deformation occurs earlier in bending. This is because an orthosis is usually
the ductile material than does fracture in the brittle expected to resist the moments generated around
material). However, only the same amount of elas¬ joints that do not have the muscle strength and
tic strain energy is released when fracture finally power to do this.
does occur—the remainder having been safely ab¬ The ability of a structural section to resist bend¬
sorbed by plastic deformation. ing is related to the distribution of material around
Typical examples of brittle materials are chalk the bending axis. This principle is illustrated by
and glass. Such materials require comparatively bending a ruler around its long axis (Fig. 10-7).
small amounts of energy to cause fracture and With the ruler flat (see Fig. 10-7A) it is easy to
typically have jagged, sharp edges at the point of bend, whereas in the upright position (see Fig.
fracture. A good example of the properties of a 10-77?) it is difficult. The same amount of material
ductile material is shown in a paper clip. It is is involved in both cases, but in the upright posi¬
comparatively easy to cause deformation and ef¬ tion, the distribution of material is more effec¬
fective failure of the device but much more diffi¬ tively spaced from the bending axis.
cult to cause the material to fracture. A useful fact that helps to reinforce the impor¬
It is important to ensure maximum patient tance of this principle is that doubling the thick¬
safety when providing orthoses for functional pur¬ ness of a section increases bending resistance by
(KO), the knee-ankle-foot complex (KAFO), and

the hip-knee-ankle-foot complex (HKAFO). Note
that foot orthoses are treated as a separate entity
and are not dealt with in this chapter. It is theoreti¬
cally possible to have a hip orthosis, but these are
rarely used and tend to be accepted as a special
category.
To achieve the required control of each joint
in the lower limb, a three-point force system is
necessary to provide the equal and opposite mo¬
ment to that being applied to the joint, which
cannot be resisted by the patient’s compromised
control system. In general, an orthosis needs to be
as long as possible so that the moment arms
through which the orthosis applies its forces are
at a maximum, reducing the magnitude of those
forces.
ANKLE-FOOT ORTHOSES
AFO devices are the most commonly prescribed

orthoses. They are intended to control the ankle
and subtalar joints as well as the biomechanics
of the knee or hip joint by ensuring appropriate
alignment of ground reaction forces relative to
FIGURE 10-7 Bending a ruler in the flat (A) and upright those joints. A wide range of functions can be
(B) planes to indicate the difference in bending resistance performed by AFO devices. Among the most com¬
made by distribution of material around the bending axis. mon are the following:
• Resisting ankle plantar-flexion-dorsiflexion
a factor of eight if the width remains the same. • Controlling a drop-foot condition
By the same token, halving the thickness reduces • Controlling the range of ankle plantar-
the bending resistance by a factor of eight. Figure flexion-dorsiflexion
10-8, in which there are two beams of the same • Resisting valgus and varus at the ankle
material, width and length, with one being twice
the thickness, shows how the thinner beam de¬ Resisting Plantar-Flexion-Dorsiflexion
flects eight times as much under the same loading
conditions. The most structurally demanding function for an
AFO is to resist plantar-flexion-dorsiflexion mo¬
ments around the ankle joint. During single stance,
Lower Limb Orthoses gait forces in excess of body weight are routinely
generated. In conditions in which vaulting using
Lower limb orthoses are subdivided into groups the stance leg occurs, such forces can be consider¬
that cover the ankle-foot complex (AFO), the knee able and may cause large moments around the
FIGURE 10-8 Deflexion of two beams under the same load (the
beams are identical other than one being twice as thick as the other). The
deflexion of the thinner beam is eight times greater. W, weight. (From
ORLAU [Orthotic Research and Locomotor Assessment Unit], The
Robert Jones and Agnes Hunt Orthopaedic and District Hospital,
Oswestry, Shropshire, England SYIO 7AG.)
Wm
FIGURE 10-9 Ankle-foot orthoses (AFO) to resist plantar-flexion and dorsiflexion. A, Twin side-
member AFO (note flat spurs to resist moments around the ankle joint). B, Molded plastic AFO (note
depth of section at the ankle to resist moments around that joint).
ankle joint. Resisting those large moments may A good example of the use of an AFO to control
require a metal structure (Fig. 10-9A) to achieve knee mechanics is in cerebral palsy patients with
sufficient stiffness to apply an effective three-point a dynamic equinus condition. Although on exami¬
fixation system. Many patients are reluctant to nation the ankle may be correctable to neutral
wear such a device because of the inevitably poor (or dorsiflexion), during walking, aberrant control
cosmesis that results from this type of structure. leads to an equinus during stance phase and a
An alternative plastic structure (Fig. 10-95) can consequent ground reaction force considerably
be used in the less structurally demanding situa¬ ahead of the knee. Figure 10-10 shows the use of
tions involving pediatric, light-weight, or dynami¬ a VVG5 to provide a video vector display of this
cally nonchallenging patients. Note that a very condition (the on-screen line represents a vector
deep section around the ankle joint is required to of ground reaction force). This gives an indication
produce sufficient bending resistance, even with of the excessive knee extending moment that re¬
the lower demands of these patients. sults. The subject is a boy about 400 N in weight.
Controlling ankle plantar-flexion-dorsiflexion In this figure, the moment on the knee can be
inevitably influences alignment of ground reaction calculated as follows:
forces relative to the knee and hip joints. Some
patients may be able to compensate their gait to Ground reaction force X moment arm
ensure that potentially adverse effects are limited (knee center to ground reaction force)
when the objective is primarily to control an un¬
stable ankle. However, the potential of fixed ankle By inspection, the ground reaction force is about
orthoses to influence alignment of ground reaction one third of the subject s body weight at maximum
forces can, in patients with lower limb control moment arm position (as illustrated in Fig. 10-
problems, either create unmanageable moments 10), and the moment arm is about 500 mm. On
for the joints above the ankle or be put to good that basis, the moment is calculated as follows:
effect in achieving the desired alignment through (400 N v 3) X 500 mm = 133 N
careful tuning of the AFO shoe complex.
X 0.5 m = 67 N m
Abnormal stresses on the knee of the magnitude

implied by this load are likely to lead to the
development of a recurvatum unless the condition
is alleviated. The use of a fixed-ankle AFO to
prevent the equinus from developing might be
expected to control the alignment problem. Expe¬
rience shows, however, that achieving this objec¬
tive requires careful tuning of the ankle-shoe com¬
plex by adjusting alignment with heel or sole
raises on the shoe. Figure 10-11A shows the use
of an AFO before tuning. It can be seen that
surprisingly little effect has been achieved simply
by preventing the equinus. After tuning with an
appropriate heel raise (Fig. 10-1 IB), however, the
alignment of the vector is much closer to the knee,
and the excessive knee extending moment has
been eliminated.
Long-term follow-up has been studied in a se¬
FIGURE 10-10 Dynamic equinus: typical ground reaction ries of six cerebral palsy patients with a dynamic
force alignment at mid-stance. equinus while walking.11 Each of eleven limbs
with the condition showed a similar pattern of
forward-facing vector relative to the knee joint
when the subject walked without an AFO. In none
Without an AFO, this moment is resisted only
of the cases was corrected alignment achieved
by the skeletal structures around the knee, where
with an AFO before tuning of the system. By
the maximum moment arm would be about 50
monitoring the patients using a VVG, appropriate
mm. Thus, 67 N m equals the force in the knee
alignment was achieved for each of the limbs by
structures times 0.05 meters (i.e., 50 mm). The
tuning the AFO-shoe complex with heel or sole
potential force in the knee, therefore, is raises. Long-term follow-up showed that correct¬
ing the alignment had an effect on the patients’
67 N m/0.05 m = 1333 N control systems. Over a period of 6 months or
more, consistent use of the system led to an ability
Bearing in mind that 1 kg is almost 10 N, it can to maintain appropriate alignment when not using
be seen that this force (equivalent to 133 bags of the orthosis. Some patients have been followed for
sugar at 1 kg each) has great destructive potential. more than 3 years after abandoning the orthosis
FIGURE 10-1 1 Provision of an orthosis to compensate for dynamic equinus. A, Before tuning the ankle-
foot orthosis. R, After tuning the ankle-foot orthosis (using a video vector generator).
following achievement of corrected control, and cope with that condition. The moment generated
they have maintained their improvement. This around the ankle joint by the weight of the foot
strongly suggests that mechanical orthoses can during swing phase is resisted by the structural
interact with a patient’s control system and under¬ section of the orthosis at the back of the ankle in
lines the need to understand the mechanics of the vicinity of the Achilles tendon. Because there
orthotic control of joints in individual cases. is also a need to permit some dorsiflexion of the
The dynamic equinus condition is perhaps the ankle during stance phase, this section requires a
most dramatic example of the potential to control subtle compromise. Ideally, it should be stiff
the joints above the ankle with an AFO. Crouch enough to prevent significant deformation from
gait can also be influenced beneficially by ad¬ the moment caused by the weight of the foot in
justing alignment, so that the ground reaction swing phase, but flexible enough to permit an
force passes closer to the knee and hip joints, appropriate range of dorsiflexion from the greater
reducing the flexing moments that the patient must moments around the ankle joint under the influ¬
resist with knee and hip extensors. Whenever a ence of body weight during stance phase. If the
fixed-ankle AFO is used, it influences alignment; orthosis is too stiff to permit an adequate dorsi¬
it is therefore appropriate to monitor the effects of flexion range, it can be adjusted by trimming the
this to ensure that optimal mechanics are achieved. section of the orthosis in the area adjacent to the
Achilles tendon. When this is done, the sensitivity
Controlling a Drop-Foot Condition of a bending section to a reduction in thickness is
of great importance, and the general rule that, for a
Some pathologic conditions, most notably stroke, rectangular section, halving the thickness reduces
can lead to a condition in which the patient has resistance by a factor of eight should caution
no effective dorsiflexors. The weight of the foot against overambitious trimming.
during the swing phase of gait creates a plantar-
flexing moment around the ankle. Because of the Controlling the Range of Ankle Plantar-
loss of dorsiflexors, the patient is unable to gener¬ Flexion-Dorsiflexion
ate the necessary balancing moment, and the foot
drops into an equinus position. The consequence Limiting the range of flexion and extension of the
is that during swing phase, the patient needs to ankle joint is a function that can be performed by
flex the knee to a greater extent than normal to an AFO. To achieve this ideally requires an or¬
create sufficient clearance for the foot to avoid thotic hinge in line with the ankle and stops
contact with the ground. This not only leads to an attached to the structure to prevent the ankle joint
abnormal gait pattern but also increases the energy from flexing or extending beyond the clinically
cost of walking. specified range.
A drop-foot orthosis (Fig. 10-12) is intended to Because of the high dorsiflexing moment ap¬
plied to the ankle joint during the stance phase of
gait, the structural demands on an orthosis de¬
signed to achieve limitation of range can be con¬
siderable. Metal structures, with their greater
strength and stiffness, are therefore frequently re¬
quired for this function when patients are unable
to provide balancing moments from their own
musculoskeletal system.
In patients who are of light weight, have low
levels of dynamic function, or can provide some
of the biomechanical function themselves, the use
of a plastic orthosis with flexion and extension
stops becomes a much more practical option.
Resisting Valgus and Varus at the Ankle
To influence valgus or varus, an AFO must apply

three-point fixation around the os calcis. A plastic
fixed-ankle AFO can be molded to apply three-
point fixation forces in both sagittal (to control
FIGURE 10-12 An ankle-foot orthosis designed to control
plantar-flexion-dorsiflexion) and coronal planes to
a drop-foot condition.
control ankle valgus or varus. This total blocking
of the ankle-subtalar complex is frequently con¬ control valgus or varus at the ankle, so that a
sidered unacceptable, and alternative orthoses that range of ankle flexion and extension is maintained.
permit ankle plantar-flexion-dorsiflexion range are This makes it necessary either to have a hinge in
therefore necessary. line with the ankle joint (Fig. 10-13B) or to have
The weight of the patient, degree of dynamic sufficient structural flexibility in the sagittal plane
activity, and extent of valgus and varus deformity to permit flexion and extension.
determine the appropriate design of the orthotic
structure to achieve the required control. When DYNAMIC ANKLE-FOOT ORTHOSES
any of these is excessive, it may be necessary to
resort to a metal structure to achieve sufficient A new category of orthosis that seeks to control
structural stiffness to permit the required degree the behavior of the ankle joint (and by reference
of control. Typically, such a device (Fig. 10-13A) the other joints of the lower limb) by influencing
has a metal side-member with a spur that fits into neurologic behavior is an interesting develop¬
a socket insert in the heel of the patient’s shoe. A ment.12 These AFOs seek to achieve this through
T strap or Y strap applies the center control force rectification of the positive mold to produce stimu¬
at the level of the malleoli, with the counterforces lating impressions on the orthosis relative to key
applied by the shoe through the spur and socket neurologic receptor sites on the plantar surface of
and by the calf band of the orthosis. Not surpris¬ the foot. Some versions of these orthoses extend
ingly, this device is not considered cosmetically into and above the ankle joint region, whereas
acceptable by most patients and is increasingly others remain below it. This suggests that those
used only in the most extreme biomechanical cir¬ that do extend into the ankle region are intended
cumstances. to produce a combination of mechanical and neu¬
The use of plastic molded orthoses to control rologic effect, whereas those that remain below
valgus and varus is now common. Although they the ankle rely only on neurologic stimulation.
may be biomechanically more compromised than The diminished nature of the orthotic structure
the alternative metal structures, their greater ac¬ that the design allows provides major cosmetic
ceptability to patients ensures that they are more benefits that could be an important encouragement
likely to be regularly used. To achieve maximum for patients to persevere with the treatment. How¬
effectiveness requires careful molding and cast ever, there is little published on the principles by
rectification to ensure appropriate application of which these devices achieve their intended aims
three-point fixation to the os calcis, and such de¬ and a dearth of clinical research to support the
vices therefore demand a high level of profession¬ concept. In some clinical centers, the devices are
alism in casting, manufacture, and fitting. used enthusiastically, suggesting that the evidence
A common requirement of an orthosis is to will accumulate and be published in the future. If
FIGURE 10-13 Orthoses designed to control valgus-varus at the ankle. A, Metal design. 8, Plastic design.
(From ORLAU [Orthotic Research and Locomotor Assessment Unit], The Robert Jones and Agnes Hunt
Orthopaedic and District Hospital, Oswestry, Shropshire, England SYIO 7AG.)
the principles are elucidated clearly and supported

by clinical evidence of success, these devices
could revolutionize clinical practice.
Plastic molded AFO devices are now the most
common type of orthosis for most forms of ankle
control. Their attraction to patients in comparison
with metal structures is obvious. They do, how¬
ever, have some features that require careful con¬
sideration. Direct contact of nonpermeable materi¬
als with the patient can lead to overheating and
perspiration. The molding process does not always
give good control of material thickness in areas of
high stress, and the overall strength of the struc¬
ture can be compromised by this and possible
adverse affects from overheating or underheating
during the molding process. Not all plastic materi¬
als have good fatigue resistance, and inappropriate
selection from ostensibly similar materials can
cause premature failure.
Provided that these matters are effectively ac¬
commodated in the design of AFO devices, plastic
materials have the potential to achieve most of
their objectives. It is likely that technologic devel¬
opments will result in new, more effective mold-
able materials that improve the position further,
and to this extent, the future looks promising.
KNEE ORTHOSES
Knee orthoses are devices that span the knee joint

only and do not seek to control other joints. They
generally fall into one of two categories in terms
of the type of patient they are intended to address:
(1) vigorous patients who have suffered knee
trauma and (2) more frail patients who are suffer¬
ing a degenerative condition of the knee. Although
the category of patient may influence the choice
of orthosis, the principles through which such de¬
FIGURE 10-14 Ground reaction forces in the coronal
vices achieve their objectives remain the same.
plane (as indicated by the video vector generator). A, Patient
The following forms of control can be applied by with osteoarthritis of the knee. B, Healthy subject.
knee orthoses:
• Resisting valgus and varus moments

• Limiting flexion and extension range in the coronal plane (from a VVG analysis) on
• Resisting anteroposterior relative movement an osteoarthritic patient who has a severe valgus
between femur and tibia deformity of the knee. As a result of the degenera¬
• Limiting rotation of the femur on the tibia tive process, the patient has a large valgus-de-
forming moment (calculated as being about 10 N
The direct loads applied to the knee by the m about the medial condyle) in comparison with
more frail patients are comparatively low. It is a healthy subject (Fig. 10-145), in whom it is
known that patients with severe osteoarthritis, for effectively zero. Pathologic conditions of this kind
example, generate vertical ground reaction forces present a severe challenge to knee orthoses.
barely above body weight. Degenerative disease, The example of a frail patient with an osteoar¬
however, usually leads to deformity of the joint, thritic deformity demonstrates the potential for
which in turn can mean that ground reaction force high moments to be generated around the knee
falls outside normal alignment, thereby producing joint even with comparatively low ground reaction
excessively high moments around the joint. Figure forces inherent in such conditions. There is clearly
10-14A shows alignment of ground reaction force ' a strong possibility for vigorous patients to gener-
ate large dynamic loads, and these might also be must bear the lateral moments required to achieve
combined with abnormal alignment to produce the required function. They must therefore have
even higher moments. Thus, the potential exists low friction under load and have strength and
for very large moments around the knee joint to rigidity commensurate with the rest of the struc¬
be generated in a variety of situations. Knee ortho¬ ture.
ses have short moment arms through which to In regard to the magnitude of valgus and varus
produce resisting moments. It is therefore unrealis¬ moments around the knee, informal measurements
tic to expect that excessive moments can be ac¬ using a VVG have registered 10 N m at the lateral
commodated by an orthosis of this type because condyle for a frail, light-weight patient with a
the large forces that would be required through severe valgus deformity and 15 to 20 N m for an
the short moment arms would be intolerable for average adult male subject indulging in fairly light
most patients. badminton activity while thrusting off sideways to
Nevertheless, knee orthoses can deliver valu¬ make a shot. Such moments are not inconsiderable
able functions, provided that their prescription is and indicate that the structural demands on a knee
limited to objectives and applications that are real¬ orthosis for a valgus and varus condition that is
istic. expected to do anything other than realign the
joint during walking are high and therefore require
Resisting Valgus and Varus Moments sound engineering design.
The line of force in the coronal plane during Limiting Flexion and Extension Range
normal walking passes through the medial condyle
of the knee (see Fig. 10-145). Any deviation In normal walking, the ground reaction force pas¬
from this produces an abnormal moment. When a ses close to the knee joint throughout the stance
patient has a correctable deformity, an orthosis phase of gait, thereby producing very low flexing
can be used to realign the knee joint. If this is or extending moments. Pathologic conditions that
achieved, the moment can be reduced to near¬ result in a flexion (or less commonly, an exten¬
normal levels, and the orthosis will not have to sion) deformity can result in malalignment of
bear excessive moments, except when aberrant ground reaction forces and lead to large moments
alignments of force occur as a result of an accident around the knee in the sagittal plane. Sports fre¬
or unusual activity. quently require vigorous use of the quadriceps to
Noncorrectable severe deformity produces un¬ resist knee flexing moments generated in response
acceptable stresses on the knee joint. Some relief to the athletic demands of the recreational activity.
of those stresses can be provided through the use A knee orthosis can be used to resist any flexion
of an orthosis to transfer some of the loads to and extension of the knee (usually in a pathologic
other parts of the affected limb. It is not realistic, condition) or to limit the range of flexion and
however, to expect a complete relief of the applied extension to that which is clinically acceptable.
moment because no acceptable orthotic structure The biomechanical demand on such an orthosis
could achieve this. The interface pressure created depends on the type of patient and the required
at the reaction points is also a cause of great function. Holding the knee in a fixed position by
concern, particularly in patients whose pathology resisting any flexion or extension requires that the
compromises their skin condition. orthosis be capable of sustaining the full flexion or
Sports activity can produce large valgus and extension moment around the joint in the position
varus moments around the knee joint in the lateral determined by deformity or clinical need. The
plane (of the order 30 N m), particularly when greater the angle of the knee from neutral (in the
side thrusts on the leg are necessary, such as occur sagittal plane), the larger the potential moment
when tennis or football players change direction that will have to be resisted. Knee orthoses with
suddenly. Patients with sports injuries are more short moment arms are unsuitable for applications
able to tolerate interface pressures and accommo¬ in which there is excessive knee flexion because
date the type of structure that would be necessary the high moment demands large orthotic forces.
to produce large moments than are those with The knee should be kept as straight as possible
degenerative disease. Even in this situation, how¬ for bracing within the clinical requirements of the
ever, it is not realistic to expect complete relief of treatment regimen being followed, to minimize
stress on the knee joint in the coronal plane. the orthotic forces.
When the only objective of the treatment is to For sports or other trauma applications, it is
resist valgus and varus moments, it is necessary likely that the objective is to limit the range of
for the structure to incorporate freely hinged knee movement to that considered clinically necessary
joints. For the orthosis to be effective, these joints or desirable. In general use, the function of the
orthosis is to act as a reminder to the patient not Depending on which way the forces are applied,
to exceed the prescribed limit. In those circum¬ posterior or anterior drawer can be resisted, so
stances, used properly, loads are applied by the that the required alignment of femur and tibia can
orthosis only when the patient inadvertently at¬ be achieved. Such an approach does not inhibit
tempts to exceed the limit. There is the potential, other functions, provided that the system is prop¬
however, for very high shock loads if the patient erly applied. This is important because the feature
loses control over the knee range, and an orthosis is often incorporated in an orthosis intended to
needs to reflect the requirement to accommodate achieve other objectives at the same time.
that situation. Knee-flexing moments of the order During normal walking, horizontal ground reac¬
160 N m have been monitored by a VVG for a tion forces of the order of 160 N have been regis¬
healthy male subject during single-leg landing tered by VVG monitoring, and it is these that have
after playing a badminton shot. The impact this to be resisted to counteract drawer. Clearly, more
could create on the flexion end stop of a range- vigorous activity would bring commensurate in¬
limiting orthosis is potentially high, and it is likely creases in the magnitude of this force.
that the prescribed range would be exceeded in
those circumstances.
Resisting Rotation of the Femur on the Tibia
Resisting Anteroposterior Relative Movement Most activities that engage the lower limb involve
the application of torque around the long axis of
When knee laxity is present, undesirable relative the bones. In gait, there are both internal and
movement of the femur on the tibia may occur external moments around the axis of the lower
(this is frequently referred to as drawer). During limb during each full cycle. Joint laxity can lead
any form of gait, horizontal components of force to unacceptable rotations in response to these axial
are applied to the lower limb. Depending on the torques, and a function required of a knee orthosis
direction of these, the femur will have a tendency can be to resist the turning moment along the axis
to slide forward or backward on the tibia. of the limb. This is perhaps the most difficult of
Some orthoses attempt to deal with this problem all the functions expected of a knee orthosis.
by providing general resistive support around the There are no obvious solid reaction points around
knee. Although this simple approach can be effec¬ which to produce equal and opposite torque, and
tive in coping with drawer, it can also have a systems intended for this purpose have to rely on
resistive effect on other functions, such as permit¬ friction between the orthosis and the patient to¬
ting free flexion of the knee. An alternative ap¬ gether with resilient strapping across the knee.
proach is also adopted in which counter moments Although this approach can provide some resis¬
are applied by the orthosis to the leg above and tance, because these systems have to operate on a
below the knee joint. This is achieved by so- short moment arm (i.e., the width of the knee),
called four-point fixation of the knee (Fig. 10-15). the force to achieve a full compensatory torque is
Above-knee
couple
Figure 10-15 Four-point fixation of a knee joint

to control drawer. (From ORLAU [Orthotic
Research and Locomotor Assessment Unit], The
Robert Jones and Agnes Hunt Orthopaedic and
District Hospital, Oswestry, Shropshire, England
SYIO 7AG.)
inevitably very high. Because the reaction point is Control of Knee Flexion and Extension
resilient, it must deform to a large degree to gener¬
ate that force and thus allow a degree of rotation In normal walking, the line of force passes
of the knee. That is not to say that support of this through or very close to the center of the knee
kind is worthless, but a cautionary note should be joint for most of the stance phase of gait. This
sounded about expecting too much of this or- ensures minimal demands on the muscles around
thotic function. the knee to balance the small moments, which
During normal walking, the torque around the nevertheless inevitably develop. When there is no
long axis of the lower limb is about 10 N m. control of the quadriceps, it is essential to ensure
Fairly gentle badminton playing increases this to that at no time is there a flexing moment because
about 30 N m. A professional tennis player, crick¬ it is impossible for the patient to generate the
eter, football or soccer player, or baseball pitcher equal and opposite moment. In those circum¬
could well exceed 100 N m during peak perfor¬ stances, it is necessary to use a KAFO to provide
mance. This gives an indication of the task when the requisite control at the knee (and also at the
an orthosis is to be used in sporting applications. ankle) through three-point fixation to prevent col¬
lapse of the patient. The three points of force
application (see Fig. 10-5) are at the back of the
KNEE-ANKLE-FOOT ORTHOSES
thigh through a thigh band, over the knee through
KAFO devices encompass control of both the a knee band (which usually incorporates a releas¬
knee and the ankle and therefore have the potential able fastening), and at the bottom of the shank,
to incorporate all of the elements of AFOs and most commonly through the shoe (with spur and
KOs. Because of their much more comprehensive socket or in-shoe AFO section).
nature, KAFOs are intended to deal with greater
levels of handicap and are inevitably more obtru¬ Control of Knee Valgus and Varus Moments
sive, which makes them cosmetically less accept¬
able to patients. That these devices extend the Excessive valgus or varus moments around the
whole length of the lower limb gives them the knee require resistance by three-point fixation in
advantage of long moment arms through which to the lateral plane. These are applied laterally
provide the controlling forces, and this enables through the orthosis at the thigh and lower shank
them to cope with situations for which KOs would levels, with the opposing force applied through
be impractical. the knee strap at the side of the knee joint (medial
The primary functions of KAFO devices are for valgus, lateral for varus). Depending on the
to provide control of knee flexing and extending degree of deformity and the weight of the patient,
moments, knee valgus and varus moments, and valgus and varus stabilizing forces can be consid¬
any combination of these, along with weight relief erable, and the orthotic structure must have ade¬
of the lower limb. Any of the functions of AFO quate strength and stiffness in the lateral plane to
devices can also be incorporated, if necessary, by cope with these.
attaching to the lower end of the structure the
appropriate orthotic components. Additional knee Combination of Orthotic Function
controls can be applied as described in the section
on KOs, but the greater level of pathologic insult Patients usually need control of both flexion and
implied by the use of a KAFO makes this a less extension and valgus and varus of the knee (to¬
frequently applied option. gether with controls of the ankle joint). When a
The primary indications for the prescription of KAFO has an adequate structure, these functions
KAFO devices are loss of control of the knee joint can be readily combined into one orthosis. In this
(as in paraplegia), severe weakness of the knee situation, attention must be paid to the patient
joint muscles, gross instability of the knee joint support strap arrangements to ensure that all the
combined with excessive deformity, and structural necessary forces are applied in the appropriate
weakness requiring weight relief. directions.
In most applications, KAFOs resist knee flexion
and extension. To be practical, these devices need Weight-Relieving Orthoses
to incorporate releasable knee joints so that pa¬
tients can sit with their knees in a flexed position. A function that is occasionally required of KAFO
Selection of these requires careful consideration devices is the relief loading on the lower limb.
of several important points, including joint speci¬ This requires that the thigh section of the device
fication, method of release, and location of the extend to the level at which loads can be transmit¬
ted through the patient’s ischial tuberosities, and
joint center.
the shank section should incorporate a ground they will be). In stabilizing orthoses (i.e., those
reaction support plate under the patient’s shoe. To seeking only to resist moments generated around
ensure that no ground reaction forces are transmit¬ the knee), the patient’s skeletal structure is in¬
ted through the leg, there must be clearance be¬ tended to carry the ground reaction forces. Inevita¬
tween the shoe and the support plate. Such an bly, because of friction at the orthosis location
arrangement demands a raise on the opposite side points, an indeterminable proportion of the loads
to compensate. along the long axis of the lower limb are carried
Transmitting loads through the ischial tuberosi¬ by the KAFO. This makes it necessary to have a
ties causes discomfort to the patient, and the addi¬ generous safety margin in the design of these
tional raise is also an unwelcome inconvenience. devices (particularly because there are also sig¬
Because the orthosis, in these circumstances, is nificant fatigue implications), which is in conflict
required to carry all the dynamic ground reaction with the patient requirements of light weight and
forces in the stance phase of gait, the structure inobtrusiveness. A careful compromise, therefore,
must be robust and therefore comparatively heavy. has to be struck to achieve the solution most
The musculoskeletal system of the lower limb (if acceptable to the patient in line with adequate
intact) can also continue to function and generate function and safety within the structure of the
loads internally, minimizing the relief of skeletal orthosis. Fig. 10-16A shows a typical metal
stress. It is for these reasons that this orthotic KAFO.
function is only occasionally invoked. For light-weight patients with no significant
knee flexion contracture, it is possible to use a
Structural Design Considerations plastic molded structure incorporating a thigh cuff
and in-shoe AFO connected by metal knee joints
The magnitude of the stabilizing forces that the riveted to the above-knee and below-knee compo¬
orthosis is required to generate (and therefore the nents (Fig. 10-16B). This is widely known as a
design that needs to be adopted) depends on the cosmetic caliper13 and has obvious attractions to
weight of the patient, the degree of knee flexion many patients. When used appropriately, this de¬
deformity, the level of dynamic activity under¬ vice is an eminently satisfactory solution to the
taken by the patient, and the spacing of the stabi¬ design dilemma. The temptation to use it in situa¬
lizing forces (the farther apart they are, the lower tions in which it is too close to the limit of its
mm
IS . u_
. ——— “““
FIGURE 10-16 Knee-ankle-foot orthoses. A, Metal design. B, Plastic design (with metal knee joints) for
improved cosmesis.
capabilities is great, however, and can lead to its should ideally be ductile in nature, so that energy
being prescribed unwisely. Before a decision is can be absorbed before fracture occurs. When a
taken to opt for this type of design, an investiga¬ joint does have a brittle failure mode, the orthotic
tion of the activities of the patient should be structure should incorporate a “mechanical fuse”
undertaken to ensure they are compatible with the that will fail in a safe manner before the joint is
structural specification of the orthosis. Cases have overloaded.
occurred in which ostensibly ideal patients experi¬
enced frequent structural failure of the orthosis.
Method of Release
Subsequent discussion with the patients usually
identified a hobby, such as abseiling, to be the
To enable patients to achieve full independence, it
likely cause. Although not always feasible, gait
is essential to provide a practical means of releas¬
analysis can also confirm that the line of ground
ing the lock on an orthotic knee joint. Two com¬
reaction force is not producing excessive knee-
monly used solutions to this problem are the ring
flexing moments.
lock and the bale lock.
When a patient is heavy or indulges in vigorous With the ring lock (Fig. 10-17A), a sliding ring
dynamic activity, it is essential to use substantial on the upper section of the side-member of the
orthotic structures to produce a viable device capa¬ KAFO overlaps a tongue on the lower section of
ble of safely sustaining the very high bending the side-member above the hinge, thus preventing
loads that will be applied to control knee flexion the hinge from rotating until the ring is raised
and extension. This approach usually requires the above the tongue. Simple methods of retaining the
use of standard metal sections to form the side- ring in the lower (locked) or upper (unlocked)
members of the orthotic structure. The choice of position on the KAFO side-member are usually
material is typically stiff, inexpensive steel or employed. This system demands conscious opera¬
light-weight, lower-ductility aluminum alloy. In tion by the patient for both locking and unlocking
selecting the material, a major influencing factor (although simple “automatic” gravity locking can
is the likelihood of the patient’s overloading the be used if means of retaining the ring in the upper
orthosis. A frail elderly woman with a low level position are not employed).
of activity, for example, would need the advantage In bale lock arrangements (Fig. 10-175), the
of the light-weight structure made possible by stop is operated using a lever extending upward
aluminum alloy. An active, young, heavy man from the rear of the joint hinge. The lever is
could be expected to have a propensity to overload usually spring biased to the locked position, so
the orthotic structure in some unexpected manner. that as the joint is straightened, the lock closes
The high level of ductility possessed by steel and automatically. Unlocking can then be effected
the consequently high level of energy it absorbs manually by the patient or automatically as the
before fracturing (after the yield point has been lever comes into contact with a seat edge when
exceeded) makes it the ideal choice for this situa¬ the patient descends into the seated position.
tion, even though the greater weight of the mate¬ To operate either of these two types of orthotic
rial is a necessary penalty. For patients between knee joint, it is necessary for the patient to relieve
those two extremes, the choice becomes increas¬ the knee-flexing moment on the joint when he or
ingly difficult as they approach the midpoint of she releases the lock. An inability to do this will
the compromise range. Fatigue performance of the compromise the independence of the patient.
two materials is fairly similar, although steel
would be expected to be less likely to suffer this
mode of failure, all other things being equal. Location of Orthotic Joint Center
Orthotic joints are normally aligned as closely as

Joint Specification
possible to the appropriate anatomic joint of the
Patient safety is of paramount importance, and the patient. This ensures that there is minimal relative
design of an orthotic knee joint used in KAFO movement between the patient and the orthosis as
devices must reflect that requirement. Bending is the two joints move through their permitted range.
clearly the most demanding mode of loading that Severe malalignment can lead to interface prob¬
will be applied, and the joint strength must obvi¬ lems from rubbing of the orthosis on skin (particu¬
ously be sufficient to withstand that which can larly when there is insensitivity) or excess inter¬
reasonably be anticipated during any function the face pressure due to migration of the orthosis from
patient might be expected to undertake. Because its optimal position. Orthoses that cross more than
overload is an ever-present possibility, the failure one joint are particularly prone to raising interface
mode of the joint (when locked) in all planes pressures when malalignment leads to effective
FIGURE 10-17 Commonly used knee joint designs. A, Ring lock joint. B, Bale lock joint.
shortening of the distance between the orthotic Menelaus,1 2 Rose,14 Carroll,15 and others have all
joints relative to anatomic joints. cited clinical experience suggesting that ambula¬
tion can achieve these objectives. More recently,
HIP-KNEE-ANKLE-FOOT ORTHOSES Mazur and colleagues16 showed in a study of
matched groups that nonwalkers have five times
Orthoses of this generic type can achieve all of the number of pressure sores and twice the number
the functions available in AFO and KAFO devices of bone fractures. Additionally, walkers were able
together with required controls of the hip joint. to perform a series of activities of daily living in
Three-point fixation of the hip is achieved by significantly greater numbers than the nonwalkers
providing forces at the thoracic and sacral levels and were three times more likely to be indepen¬
and on the thigh in the relevant directions for the dently mobile within the community. The case for
selected control. The main purpose of HKAFO providing orthoses to enable such patients to walk
devices is to facilitate ambulation for severely is therefore strong, provided that a practical, fi¬
handicapped patients who have inadequate or ab¬ nancially viable system that is acceptable to pa¬
errant control of the hip joint. Typically, patients tients is available. What is practical and acceptable
have complete spinal lesions at LI or above, varies among patients and generally is judged
which implies that there is no motor control of against the following factors:
the hip joint at all. It is not a realistic expectation
that these patients will achieve as efficient loco¬ Walking style
motion on a flat surface as they can anticipate in Energy cost of walking
a wheelchair, and it is therefore necessary to iden¬ Ease of applying and removing the system
tify the justification for prescription of such de¬ Overall cost of all elements of treatment,
vices. Those who advocate walking for the se¬ including ongoing service
verely handicapped have two main objectives:
As indicated earlier, any form of walking re¬
1. To provide therapeutic benefit quires control of both internal and external stabili¬
2. To increase independence and allow zation forces. Severe handicap inhibits the neces¬
patients to undertake activities not possible sary control, and an HKAFO is therefore required
in a wheelchair
to compensate. In particular, an orthosis provides
appropriate control at the hip joint for the style of In most applications, unlockable joints at the
ambulation in question, which (together with the hips and knees are required to permit the patient
support for other lower limb joints) provides inter¬ to sit while wearing the orthosis. The structural
nal stabilization and a larger area of support demands on these (and other elements of the or¬
(sometimes in combination with other walking thosis) are dictated by the deformities of the pa¬
aids) than is necessary for a healthy subject to tient. When there are large contractures at the
ensure good external stabilization. Propulsive anatomic joints, the moments that the orthosis
forces are injected by the upper limbs (by an must resist are correspondingly great, and the
additional walking aid) or through inertial forces structure needs to reflect the increased stresses
generated by truncal movement. that result.
Many different walking systems have evolved, Although swing-through gait is probably the
and the means of hip control (and control of other simplest means of providing walking for severely
distal joints) is a fundamental determining factor handicapped patients, it is physically demanding
in the style of walking they provide and their and is the least efficient of the available options.
consequent ability to satisfy the needs of patients. The conventional orthotic structures generally
The other factors are dependent on detail design used for swing-through gait (Fig. 10-18A) are
of individual orthotic systems. Overall effective¬ difficult to put on and take off, and this inhibits
ness of service delivery (e.g., patient assessment, independence. Although more modern designs of
production of orthosis, fitting and training the pa¬ orthosis address some of the difficulties17 (Fig.
tient, ongoing regular review) are of primary im¬ 10-185), many patients find the challenges of
portance in ensuring a practical outcome for any adopting this type of ambulation too great. A
form of walking system for severely handicapped review of the literature18-20 reveals that the average
patients. energy cost for swing-through gait in patients with
thoracic lesions is about 30 J/kg/m, which is al¬
Walking Style most 10 times that for normal walking.
Several different walking styles are employed by Swivel Walking

patients with complete absence of control in the
lower limbs. Those most commonly adopted in¬ Swivel walking requires an orthosis that provides
clude the following: internal stabilization of the patient by resisting
Swing-through gait moments around all the joints of the lower limbs
Swivel walking (i.e., at hips, knees, and ankles) to avoid collapse
Reciprocal walking of the skeletal structure. Swivel walking also in¬
corporates footplates to give an adequate support
The choice of style depends on patient and
area for external stabilization, to prevent the pa¬
clinician preference, the physical characteristics
tient from falling over.
and functional capabilities of the patient, and the
The footplates also need a mechanism that per¬
physical therapy, orthotic, and financial resources
mits them to swivel forward alternately on the
available. Apart from swivel walking, an addi¬
grounded footplate as the patient rocks from side
tional walking aid such as walking frame or
to side. Ambulation is achieved by the patient’s
crutches is also necessary.
injecting an inertial force into the orthosis by
thrusting the trunk sideways (Fig. 10-19A). This
Swing-Through Gait
leads to several interactive mechanical reactions,
For swing-through gait, the patient’s hips and as indicated in Fig. 10-195. Being slightly angled
knees are held in a fixed position by the HKAFO (dihedral angle), the footplates rest on their inner
(the ankles may be either fixed or allowed to flex border, and the orthosis rocks onto the footplate
and extend, depending on the pathology of the coincident with the direction of inertial thrust (I).
ankle). This enables a walking frame or crutches The bearing center of the footplates are arranged
to be used to raise the whole weight of the body to be behind the patient’s center of gravity, so that
(and orthosis) from the ground using the upper the sideways inertial thrust, If, normally produces
limbs. The legs are then swung forward in space a forward rotating moment around the grounded
to land a complete stride forward. The perfor¬ footplate to advance the swivel walker in the gen¬
mance of the patient depends on his or her erally preferred direction. To ambulate backward,
strength and control and is further influenced by the inertial thrust, Ir, must be arranged to lie be¬
the weight of the orthosis and its ability to provide hind the footplate bearing center. This is more
the required stability, particularly in the lateral difficult (but not impossible) for the patient to
plane, during the stance phase. achieve.
leads to forward rotation of the swivel walker if

the footplate is held on the ground (e.g., by step¬
ping on it), provided that the center of gravity of
the patient is forward of the footplate bearings.
A typical example of a swivel walker is shown
in Figure 10-20, including the four primary fixa¬
tion points necessary to achieve the required inter¬
nal stabilization of the patient.
Swivel walking is performed without the use of
additional walking aids such as crutches, but as a
consequence, it is limited to flat, smooth surfaces,
which are only common indoors. It is compara¬
tively easy to perform, requiring coordination of
upper body movements rather than strength, and
is therefore suited to very young patients (it can
be adopted from the age of 1 year) and to those
with upper limb involvement that prevents the use
of crutches.
The structural design that can be adopted for
this type of orthosis permits a more ready accom¬
modation of the increased moments generated by
stabilizing joints with contractures. This enables
ambulation to be provided for much greater levels
of physical handicap than any other form of walk¬
ing. In addition, it is possible to build in a high
degree of structural rigidity, and this influences
the ease of walking by increasing the efficiency of
transferring the inertial forces required to produce
ambulation.
Several manifestations of this type of orthosis
exist, and they generally fall into one of two
categories:
1. Those providing hip joints (and knee joints

in some cases) to permit sitting21'22
2. Those designed to ease putting on and
taking off, so that the orthosis can be
readily discarded for sitting23'24
For patients who can transfer easily from a
wheelchair while wearing a cumbersome orthosis,
there are some potential advantages to the first
category if ambulation is interspersed with periods
of sitting. Most patients are likely to use a swivel
walker for set periods of ambulation, and in this
situation the second category offers more comfort
Figure 10-18 Hip-knee-ankle orthoses for swing-through and safety when sitting. Careful analysis of the
gait. A, Conventional orthosis. B, Parapodium. (B, From most appropriate compromise is needed to ensure
ORLAU [Orthotic Research and Locomotor Assessment that patients are provided with the best solution
Unit], The Robert Jones and Agnes Hunt Orthopaedic and for their individual needs.
District Hospital, Oswestry, Shropshire, England SYIO 7AG.)
Reciprocal Walking
In addition to the inertial forces, there is a small Swing-through gait and swivel walking do have
component of gravitational force (K) because of significant disadvantages. Both have a style that
the angle of the body relative to the ground when emphasizes disability; the former is difficult and
the swivel walker is on one footplate. This can be has a high energy consumption, and the latter is
useful when setting up a swivel walker because it slow and limited to flat surfaces. More recent
I Chapter 10 LOWER LIMB ORTHOTICS 26
FIGURE 10-19 The mechanics of ambulation in a swivel

walker. A, Side-thrust to achieve inertial reaction. 6, Swivel
walker force components, eg, center of gravity; k, lateral
component of gravitational force; w, weight of patient; l„
inertial reaction forward of c of g; l„ inertial reaction to the
rear of c of g. (From ORLAU [Orthotic Research and
Locomotor Assessment Unit], The Robert Jones and Agnes
Hunt Orthopaedic and District Hospital, Oswestry,
Shropshire, England SYI0 7AG.)
Grounded
footplate
Grounded
footplate
injected through walking frame or crutches, and

these also serve to provide increased external sta¬
bilization by increasing the support area.
The essential features of reciprocal walking are
that at least one leg is on the ground supporting
the weight of the subject at any stage of the gait
cycle and that the trunk advances over the stance
leg while the swing leg moves from a relatively
extended to a flexed position before landing to
initiate a double support phase.
For healthy subjects, this means of progression
is the most efficient form of ambulation, and its
translation to handicapped patients held the hope
of more practical solution to walking. This has
proved to be the case for those with good upper
limb function, limited deformities, lack of obesity,
and good coordination. The main principles of
orthotic reciprocal walking are as follows:
Swing-leg clearance is achieved by tilting or

lifting the body through the walking frame (or
crutches). Tilting is easier but requires a high
degree of lateral rigidity in the HKAFO.
Swing-leg progression by hip flexion is initiated
FIGURE 10-20 Typical swivel walker with four points of by gravity on a freely hinged orthotic hip
force required to provide internal stabilization for a joint, or through a cross-linkage from the
paraplegic patient. (From ORLAU [Orthotic Research and
opposite orthotic hip joint to allow progression
Locomotor Assessment Unit], The Robert Jones and Agnes
on the stance side to drive this requirement.
Hunt Orthopaedic and District Hospital, Oswestry,
Shropshire, England SYIO 7AG.) Advancing the trunk over the stance leg is
achieved through the use of latissimus dorsi to
pull the trunk forward by the arms, via the
orthotic developments have sought to overcome crutches or walking frame reacting on the
those difficulties by making reciprocal walking ground.
possible. This style mimics much more closely These principles are indicated diagrammatically
normal walking by allowing flexion and extension in Fig. 10-21 for one practical manifestation of
of the hip joint under orthotic control while main¬ orthotic reciprocal walking.
taining stabilization of the other joints of the lower A number of reciprocal walking systems are
limbs. Because there is no active power in the routinely available. These fall into two main cate¬
paralyzed hip joint, propulsive forces have to be gories:
FIGURE 10-21 Reciprocal ambulation

mechanics (as in the Parawalker). (From
ORLAU [Orthotic Research and
Locomotor Assessment Unit], The
Robert Jones and Agnes Hunt
Orthopaedic and District Hospital,
Oswestry, Shropshire, England SYIO
7AG.)
Those with linkages between the two orthotic hip showing a measurable increase in efficiency of
joints, constraining them to move in unison, ambulation. Each of three patients with long expe¬
but with opposite rotation. Typical rience in the use of the system significantly re¬
manifestations of this approach have been duced the energy cost of walking31 in the stiffer
described by Douglas and associates,25 orthosis. Although this is the only direct evidence
Lissons26 and Motloch.27 of the theoretical benefit of lateral rigidity, there
Those with freely hinged hip joints with a is much other evidence that the orthosis with the
controlled range of flexion and extension. This greatest stiffness provides the most efficient walk¬
approach has been proposed by Rose,28 and an ing. It cannot be ruled out that other features of
this orthosis (which has freely hinged hip joints
orthosis that uses this mechanism is shown in
with limited flexion range) also contribute to ease
Figure 10-22.
of walking for the patient.
The linked hip joint mechanisms are intended
to ensure that the anatomic hip joints remain under
orthotic control at all times, whereas a limited free Energy Cost of Walking
range of the orthotic hip joints provides gravita¬
tional drive of the swing phase and greater flexi¬ In a review of six patients using two types of
bility of use of the anatomic hip joints. Little orthoses, Bowker and colleagues32 showed that the
published evidence is available to support the energy cost in the orthosis with the lower lateral
competing claims of the benefits of the two ap¬ rigidity was on average significantly greater. Banta
proaches to the hip joint mechanisms. and colleagues,33 using a different methodology
Other aspects of overall design have been exam¬ on four other patients, reported almost identical
ined in some detail. The theoretical case for lateral quantitative findings.
rigidity of the orthotic structure has been proposed An important factor in the practical value of a
by Stallard and Major.29 The ORLAU Parawalker30 walking orthosis is the additional walking aids
is widely acknowledged as the reciprocal walking required to make the system viable. Crutches are
orthosis with the greatest lateral rigidity, and a considerably less cumbersome than a walking
modification of this device, which brought about frame and much more convenient to transport;
an increase in stiffness of 10%, has lent further they are more difficult to use, however, and rely
support to the importance of that property by more heavily on the orthotic structure to supple¬
ment the patient input. In a study of several types
of reciprocal walking orthoses, Bemardi and co¬
workers34 found that for the orthosis with the
greatest lateral rigidity, all patients used crutches,
whereas for those with less stiffness, many pa¬
tients used walking frames.
Although rigidity does confer higher levels of
efficiency in walking, the structural design neces¬
sary to achieve this is judged by some observers
to carry a penalty in terms of cosmesis. Others
consider that overall functional performance (in¬
cluding an ability to put on and take off the
orthosis as well as efficiency of walking) deter¬
mines the degree to which walking orthoses can
provide therapeutic and independence benefits. In
the final analysis, it is the regular, ongoing use of
the walking system that holds the key to success.
Three studies of the Parawalker on different
groups of patients35-37 showed that 60% of subjects
or more continued to use the system for 3 years
or considerably longer. It was concluded that an
important factor in the superiority of these results
to those reported in similar studies of equivalent
patients using conventional orthoses was the rou¬
tine regular review of patients using the Para¬
walker. Nevertheless, the level of compliance re¬
FIGURE 10-22 A reciprocal walking orthosis (Parawalker).
ported in different groups gives encouragement to
the concept that reciprocal walking is a viable 2. Rose GK: Surgical/orthotic management of spina
option for severely handicapped patients. bifida. In Murdoch G (ed): The Advance of
As with other types of lower limb orthoses, Orthotics. London, Edward Arnold, 1976, pp
contractures of hips, knees, or ankles increase the 403—413.
moments that the orthosis must resist to ensure 3. ISO 15032: Structural testing of hip joints, 1996.
internal stabilization for the patient. They may 4. Major RE, Butler PB: Discussion of segmental
also inhibit the available range of hip flexion and stability with implications for motor leaning. Clin
Rehabil 9:167, 1995.
extension and thereby limit the capacity for effec¬
5. Tait JH, Rose GK: The real time video vector
tive reciprocal walking. Ease of ambulation and
display of ground reaction forces during
strength of the orthotic structure have an important ambulation. J Med Eng Technol 3:252, 1979.
bearing on the practicality of a system for these 6. MacGregor J: The evaluation of patient
patients, and extra caution should be exercised performance using long term ambulatory
when assessing patients who fall into this cate¬ monitoring technique in the domiciliary
gory. environment. Physiotherapy 67:30, 1981.
7. Butler PB, Englebrecht M, Major RE, et al:
Realism of Treatment Objectives Physiological cost index of walking for normal
children and its use as an indicator of physical
Important advances have been made in the field handicap. Dev Med Child Neurol 26:607, 1984.
of orthotics in the whole range of devices, from 8. Nene AV, Jennings SJ: Physiological cost index of
AFOs to HKAFOs. When they are used in the paraplegic locomotion using the ORLAU
context of a multidisciplinary team assessing the Parawalker. Paraplegia 30:246, 1992.
most appropriate treatment options for an individ¬ 9. Hoffer M, Feiwell E, Perry R, et al: Functional
ual patient, they do have the potential to provide ambulation of patients with myelomeningocele. J
Bone Joint Surg 55A:137, 1973.
effective and long-term benefit. The limitations of
10. Garrett M, Meehan C: Classification of walking
orthoses must always be borne in mind when they
handicap in the spinal cord injured population: A
are a treatment option being considered against pilot study. In Pedotti A, Ferrarin M (eds):
other systems that may also be appropriate. To be Restoration of Walking for Paraplegics: Recent
effective, an orthosis must always interface with Advances and Trends. Oxford, UK, IOS Press
the patient, and the loads that have to be applied 1992, pp 343-349.
to achieve the treatment objectives must be practi¬ 11. Butler PB, Thompson N, Major RE:
cal for both the patients and the orthosis. Improvements in walking performance of children
Orthoses that can achieve subtle effects by in¬ with cerebral palsy: Preliminary results. Dev Med
fluencing neurologic plastic change are being de¬ Child Neurol 34:567, 1992.
veloped, but this demands accurate alignment of 12. Hylton N: The use of dynamic ankle foot orthoses
and their impact on balance and upper body
forces relative to specific joints. In those circum¬
function. Neurol Rep 13:15, 1989.
stances, it is essential to have instrumentation 13. Tuck W: The Stanmore cosmetic caliper. J Bone
available to permit tuning of the orthosis-patient Joint Surg 56B:115, 1974.
relationship. The increasing availability of clinical 14. Rose GK: Orthoses for the severely handicapped:
gait laboratories makes it more practical to con¬ Rational or empirical choice? Physiotherapy
sider this option. Without that facility, it is not 66:76, 1980.
possible to be certain that the required effects are 15. Carroll N: The orthotic management of the spina
being achieved, and inappropriate outcomes would bifida child. Clin Orthop 102:108, 1974.
be likely to occur. 16. Mazur JM, Shurtleff D, Menelaus M, Colliver J:
New materials are being developed that hold Orthopaedic management of high-level spina
bifida. J Bone Joint Surg 71A:56, 1989.
much promise for important advances in the field
17. Motloch W: The Parapodium: An orthotic device
of orthotics. Greater strength and stiffness can
for neuromuscular disorders. Artif Limbs 1536
lead to less obtrusive devices with yet further 1971. ' ’
improvements in patient performance. In consider¬ 18. Clinkingbeard JR, Gesten JW, Hoehn D: Energy
ing these new opportunities, designers must not cost of ambulation in traumatic paraplegia Am J
lose sight of the potential for accidental overload Phys Med 43:157, 1964.
with handicapped patients who have compromised 19. Huang CT, Kuhlemeier KV, Moore MB, Fine PR:
control and should consider carefully the failure Energy cost of ambulation in paraplegic patients
mode of the materials they specify. using Craig Scott braces. Arch Phys Med Rehabil
60:595, 1979.
References 20. Merkel KD, Miller NE, Westbrooke PR, Merritt
JL. Energy expenditure of paraplegic patients
1. Menelaus MBD: Progress in the management of
standing and walking with two knee ankle foot
the paralytic hip in myelomeningocele. Orthop
orthoses. Arch Phys Med Rehabil 65:121, 1984.
Clin North Am 11:17, 1987.
21. Griffiths JC, Henshaw JT, Heywood OB, et al:
I Chapter 10 LOWER LIMB ORTHOTICS 265
Clinical applications of the paraplegic swivel for paraplegic patients. Physiotherapy 73:393,
walker. Biomed Eng 2:250, 1980. 1987.
22. Motloch WM, Elliott J: Fitting and training 31. Stallard J, Major RE: The influence of orthosis
children with swivel walkers. Artif Limbs stiffness on paraplegic ambulation and its
Autumn:27, 1966. implications for functional electrical stimulation
23. Edbrooke H: Clicking splint. Physiotherapy (FES) walking systems. Prosthet Orthet Int
56:148, 1970. 19:108, 1995.
24. Butler PB, Farmer IR, Poiner R, Patrick JH: Use 32. Bowker P, Messenger N, Ogilvie C, Rowley D:
of the ORLAU Swivel Walker for the severely The energetics of paraplegic walking. J Biomed
handicapped patient. Physiotherapy 68:324, 1982. Eng 14:344, 1992.
25. Douglas R, Larson P, D’Ambrosia R, McCall RE: 33. Banta JV, Bell KJ, Muik EA, Fezio J: Parawalker:
The LSU Reciprocating Gait Orthosis. Energy cost of walking. Eur J Pediatr Surg
l(Suppl I):7, 1991.
Orthopedics 6:834, 1983.
34. Bemardi M, Canale I, Castellano V, et al: The
26. Lissons MA: Advanced Reciprocating Gait
efficiency of walking of paraplegic patients using
Orthosis in paraplegic patients. ISPO World
a reciprocating gait orthosis. Paraplegia 33:409,
Congress, Chicago, June 28 to July 3, 1992.
1995.
27. Motloch W: Principles of ortho tic management for 35. Moore P, Stallard J: A clinical review of adult
child and adult paraplegia and clinical experience paraplegic patients with complete lesions using the
with the Isocentric RGO. In Zupko P (ed): ORLAU Parawalker. Paraplegia 29:191-196,
Proceedings of the 7th World Congress of the 1991.
ISPO, Chicago, June 28 to July 3, 1992, p 28. 36. Major RE, Patrick JH, Stallard J: A review of
28. Rose GK: The principles and practice of hip adult paraplegic patients using the ORLAU
guidance articulations. Prosthet Orthot Int 3:37, Parawalker. Proceedings of ISPO World Congress,
1979. Melbourne, April 2-7, 1995.
29. Stallard J, Major RE: The case for lateral stiffness 37. Stallard J, Major RE, Patrick JH: The use of the
in walking orthoses for paraplegic patients. Part ORLAU Parawalker by adult myelomeningocele
H. J Eng Med 207:1, 1993. patients: A seven year retrospective
30. Butler PB, Major RE: The Parawalker: A rational study—preliminary results. Eur J Pediatr Surg
approach to the provision of reciprocal ambulation 5(Suppl I):24, 1995.
i I
Index
Note: Page numbers in italics refer to illustrations; page numbers followed by t refer
to tables.
A band, of sarcomere, 84, 84 Amplitudes of oscillations (Continued) Balance (Continued)

Abdominal belts, use of, for low-back in craniocerebral-injured patients, 203 proprioceptive deficit in, 203
injury, 135 Amputation, below-knee, balance after, subjects tested for, 204
Abdominal muscle(s), 114, 116, 117 203-205 dynamic nature of, 189
Abdominal muscle exercise, for low- amplitudes of sway in, 204 in able-bodied subjects, 197-200
back injury, 133t, 133-134, 134 influence of prosthesis alignment amplitudes of oscillations in, 198,
Abduction, of scapula, 143 on, 204-205 199-200
Able-bodied subjects, balance in, proprioceptive deficit in, 203 frequency of swaying motion in,
197-200 subjects tested for, 204 197, 197-198
amplitudes of oscillations and, Amyotrophic lateral sclerosis, muscle relative sequence of tangential
198-200, 199-200 strength impairment associated force vectors and, 198, 198t
frequency of swaying motion and, with, 167 in cerebrovascular accident patients,
197, 197-198 Angular acceleration, defined, 9 200-202
relative sequence of tangential Angular momentum, defined, 11 amplitudes of oscillations in, 201-
force vectors and, 198, 198t Anisotropy, 72 202
Acceleration, angular, measurement of, Ankle plantar-flexion-dorsiflexion, relative sequence of tangential
29 orthosis for, 250 force vectors and, 201
as performance variable, 9, 9 Ankle-foot orthosis, 247-252 time effect and locomotor outcome
defined, 9 dynamic, 251-252 and, 202
measurement of, 25-28, 25-28 for ankle plantar-flexion-dorsiflexion, waveform frequencies and, 201
selection of equipment in, 28 250 in craniocerebral-injured patients, am¬
Accelerometer, configuration of, 25 for drop-foot condition, 250, 250 plitudes of oscillations and, 203
effect of gravity on, 25, 26 for resisting plantar-flexion- correlation between mechanical and
in fracture healing, 27, 27 dorsiflexion, 247-250, 248-249 clinical parameters in, 203
measurement of vibration transmissi- for resisting valgus and varus mo¬ relative sequence of tangential
bility by, 28, 28 ments, 250-251, 251 force vectors and, 202-203
monitoring motion with, 26 plastic molded, 252 waveform frequencies and, 202
Accessory movements, of joints, 210, to control knee mechanics, in cerebral measurement of, 189-191
212-213 palsy patient, 248-249 perturbed standing tests in, 191
Accuracy, in measurement, 3, 3 Antagonist restraint, targeting, to unperturbed standing tests in, 189—
Acromial angle, flat, 153, 154 improve muscle force production, 191, 190
Acromial spurs, 153 174-175 parameters of, 191
Acromioclavicular joint, injuries to, 144 Anteroposterior relative moment, postural, measurement of, 189
kinematics of, 143-144, 144 resisting, knee orthosis for, 254, swaying motion in, bilateral measure¬
structure of, 142-143, 143 254 ment of, 194-197, 195-196
Acromium, shapes of, 153, 154 Anulus fibrosus, of intervertebral disk, modeling of, 192-194, 193-194
Actin filaments, 85, 86 44, 45-46, 45-46, 107 nature of, 192
Acupuncture, to improve muscle force noncellular components of, 45t Bale lock joint, in knee-ankle-foot
production, 175, 176 Arthrokinematics, of glenohumeral joint, orthosis, 257, 258
Adduction, of scapula, 143 155-156 Ball-and-socket joint. See specific joint,
Adductor muscles, architectural features Articular cartilage. See Cartilage, e.g.. Acromioclavicular joint.
of, 87t articular. Ballistic motion, 5
Aerobic exercise, for low-back injury, Asymmetry, defined, 196
Barthel Index, 169
132-133 Augmentation devices, for articular
Beam balance, in measurement of force,
Age (aging), influence of, in stress- cartilage repair, 43
15, 15
strain behavior of bone, 75-76, 76 Avulsion, of bone-ligament-bone
Below-knee amputation, balance after,
of articular cartilage, 43-44 complex, 56
203-205
of intervertebral disks, 51-53, 52t Axiohumeral musculature, in
amplitudes of sway in, 204
of ligaments, 56-57 glenohumeral joint kinetics,
influence of prosthesis alignment
of tendons, 59 158-159
on, 204-205
Aggregan-hyaluronan, in articular Axioscapular musculature, in scapular
proprioceptive deficit in, 203
cartilage, 38, 39 motion, 146-149, 146-149
subjects tested for, 204
Aggregans, of cartilage proteoglycans,
Bending resistance, of orthotic devices,
37-38, 39 246-247, 247
a chains, of tropocollagen molecule, 37 Back, low, injury to. See Low-back
injury. Biceps brachii tendon, in glenohumeral
Ambulation. See Walking.
loads on, during lifting and walk¬ joint kinetics, 158
American Academy of Orthopaedic
ing, 119-121, 120t Biologic tissues, mechanics of, 2
Surgeons, orthosis defined by, 239
Balance, 189-206 Biomechanic analysis, 1-31
Amplitudes of oscillations, in able-
after below-knee amputation, 203- instrumentation in, 14-24. See also In¬
bodied subjects, 198-200, 199-200
205 strumentation.
in below-knee amputation subjects,
amplitudes of sway in, 204 for measuring force, 14-19
204 for measuring position, 19-24
in cerebrovascular accident patients, influence of prosthesis alignment
on, 204-205 meaning of, 2, 2
201-202
268 Index
Biomechanic analysis (Continued) Brittle materials, for orthotic devices, Connective tissue, 65-66. See also
measurement terms used in, 2-5, 3—4 245, 245-246 specific type, e.g., Tendon(s).
mechanics of materials in, 13-14, Contraction(s), reflex muscle, in spinal
13-14 manipulative therapy, 232
linear elasticity and, 14 Calcified zone, in articular cartilage, 36,
Coracoacromial arch, 152-153, 153-154
strain and, 13 36, 37
Coracoid process, ligaments arising
stress and, 13, 13-14 Calcium hydroxyapatite, in bone, 66, 67
from, 143, 143
mechanics overview of, 5-6, 5-6 Caliper, cosmetic, 256, 256
Cosmetic caliper, 256, 256
performance variable(s) in, 6-13. See Camera accuracy, in optical imaging
Costoclavicular ligament, 142
also Performance variable)s). systems, 21
Couch padding, as factor in lumbar
classic, 6-10 Canaliculi, in bone, 69, 71
mobilization stiffness, 223
derived, 10-13 Canes, 177-178
Coupled movement, of joints, 211
position time derivatives in, 24—30 Capacitance sensing technology, 30
Cracking sounds, in spinal manipulative
acceleration and, 25-26, 25-28 Cartilage, articular, 35—44
therapy, 229-232, 230-231
angular velocity and angular accel¬ aging of, 43^44
Craniocerebral injury, balance in
eration measurement and, 29 biomechanical studies of, 40
patients with, amplitudes of
jerk and, 28-29 chondrocytes in, 36-37
oscillations and, 203
moments of inertia/products of iner¬ collagen in, 36, 37, 38
correlation between mechanical and
tia measurement and, 29 components of, 36t
clinical parameters and, 203
pressure measurement and, 29-30 compression of, 40-41, 411
relative sequence of tangential
velocity and, 24-25 general features of, 35, 36
force vectors and, 202-203
variables of interest in, 2 interstitial water in, 39^40
load carriage of, 40 waveform frequencies and, 202
Bipedal standing, 189
Creep behavior, of end plates, 51
Body, stabilization of, in walking cycle, proteoglycans in, 37-39, 39
repair of, 43-44 of intervertebral disk, 49-50
241
structure and composition of, Critical stress intensity factor, 77, 77
Body mass index, as factor in lumbar
35-40 associated with controlled, slow frac¬
tension of, 42^43 ture, 77, 78
Body segments, six-dimensional
Cell-seeded repair systems, for articular dependence of, on orientation be¬
position of, in optical imaging
cartilage, 43—44 tween 0 and 90 degrees, 77, 78
systems, 23
Cellular zones, in articular cartilage, 36, Cruciate ligament, anterior, features of,
Body sway, natural, bilateral
36, 37 53
measurement of, 194-197
Cement line, of haversian system, 68, rupture of, 56
center of gravity in, estimation
69 surgical repair of, 57
of trajectory of, 196-197
Center of gravity, trajectory of, 190, 190 Crutches, 177-178
defined parameters in, 195-196,
estimation of, from bilateral reac¬ Cutaneous stimulation, to improve
195-196
tive forces, 196-197 muscle force production, 173
methodology of, 194-195
Center of rotation, in joint movement,
Bone, 65-80. See also named bone,
211
e.g., Femur.
Cerebral palsy, muscle strength Deep zone, in articular cartilage, 36, 36,
biomechanical studies of, 70, 72-80 37
cell types in, 69, 71 impairment associated with, 167
Cerebrovascular accident. See also Deltoid muscle, in glenohumeral joint
characteristics of, 66
Stroke. kinetics, 156-157
components of, comparison between,
balance in patients with, 200-202 Density, influence of, in stress-strain
67, 68t
amplitudes of oscillations and, behavior of bone, 75, 75
cortical, composition of, 66t
201-202 Depression, of scapula, 145
fracture mechanics of, 76-77, 77-78,
relative sequence of tangential Deviations, in standard measurement, 3
79. See also Fracture(s).
force vectors and, 201 Dimensions, in optical imaging systems
function of, 65
time effect and locomotor outcome 22
lamellar, 68-69, 69-71
and, 202 Disk, intervertebral. See Intervertebral
long, stress lines in, 70, 71
waveform frequencies and, 201 disk.
structure of, 69-70, 71
Chairbound patient, 244 Dislocation, of glenohumeral joint, 150
macrostructure of, 69-70, 71
Chondrocytes, in articular cartilage, Displacement, as performance variable,
mechanical properties of, 72
36-37 9, 9
microstructure of, 67-69, 69-71
Chondroitin sulfate, in cartilage, 37, 39 in manual therapy, 214
movements of, in peripheral joint mo¬
Clicking sounds, in manual therapy, 214 Dose, in manual therapy, measurement
bilization, 217
Collagen, in bone, 66-67 of, 214
spinal. See Vertebrae.
in cartilage, fibrillation of, 43 mobilization, applied manual force
stress-strain behavior of, 72-76
mechanical properties of, 67, 67t and, 214-217, 215-217
influence of age in, 75-76, 76
microstructure of, 37 Drawer, control of, knee orthosis for
influence of density in, 75, 75
molecules of, 66, 66-67 254, 254
influence of orientation in, 72-73,
stress-strain behavior of, 67, 67-68 Drift, 4, 4
72-74
structure of, 66, 66 Drop-foot condition, ankle-foot orthosis
viscoelasticity in, 73, 74, 75
tension of, 42-43 for, 250, 250
ultrastructure of, 66-67, 66-68, 67t,
tropocollagen molecules of, 37, 38 Duchenne’s muscular dystrophy, muscle
68t
Collateral ligament, medial, features of, strength impairment associated
woven, 67-68 53 with, 167
Brain injury, traumatic, muscle strength rupture of, 56 Ductile materials, for orthotic devices
impairment associated with, 167 Community walkers, 244 245, 246
Breathing, as factor in lumbar Compression, of articular cartilage,
Dynamic forceplate, in measurement of
mobilization stiffness, 222-223 40-41, 41t, 42
force, 17-19, 18
Index 269
Dynamic forces, in orthosis, 245 Fibrocartilage, 35. See also Cartilage. Glenohumeral joint (Continued)
Dynamic imbalance, 11 Flexibility issues, in rehabilitation dislocation of, 150
programs, for low-back injury, forces at, 159-160
130-131, 131-132 glenoid labrum of, 150
ECM. See Extracellular matrix (ECM). Flexion range, limitation of, knee kinetics of, axiohumeral musculature
Elastic cartilage, 35. See also Cartilage. orthosis for, 253-254 in, 158-159
Elasticity, linear, 14 Fluid phase, of interstitial water, 40 scapulohumeral musculature in,
modulus of, of orthotic materials, 246 Force(s), as performance variable, 9-10, 156-158
Electrical stimulation, to improve 10 ligaments of, 151, 151-152
muscle force production, 173, 175 defined, 9-10 osteokinematics of, 153-155, 155
Elevation, of scapula, 145 direction of, in spinal manipulative scapulohumeral rhythm of, 160
End plates, 44, 46-47, 107 therapy, 228-229 Glenoid fossa, 149-150
creep behavior of, 51 dynamic, in orthosis, 245 Glenoid labrum, 150
fracture of, 105, 106, 108 equilibrium of, 15 Golgi tendon organs, 57
noncellular components of, 45t ground reaction, in stroke patient, 171 Goniometers, 19-20, 20
Endurance issues, for low-back injury, monitoring of, 242, 243 Gravity, center of, trajectory of, 190,
131-135 in joint mobilization, frequency of, 190
Energy, as performance variable, 12 216-217 estimation of, from bilateral reac¬
transfer of, in manual therapy, 214 grading of, 215, 215 tive forces, 196-197
Energy cost, of walking, with hip-knee- range of, 216, 216-217 effect of, on accelerometer, 25, 26
ankle-foot orthosis, 263-264 manual, application of, to lumbar Ground reaction forces, in stress tensor,
Energy expenditure, relative, monitoring spine, 220, 220 13
of, in orthotic treatments, 242, 244 measurement of, 14—19 in stroke patient, 171
Euler’s angles, 9 beam balance in, 15, 15 monitoring of, 242, 243
Exercise(s), for low-back injury, 130 spring balance in, 15, 15 Guillain-Barre syndrome, muscle
abdominal muscle, 133t, 133-134, strain gauges in, 15-17, 16 strength impairment associated
134 propulsive, in walking cycle, 241 with, 167
aerobic, 132-133 rate of change of, in manual therapy,
extensor muscle, 134—135, 135 214
for thoracolumbar spine, 135-136 relation between resistance and, 16, H band, of sarcomere, 84, 84
rehabilitation, for scapulothoracic 16-17 HAM (helical axis of motion), in
complex, 147, 147-148 Force-displacement relationship, of accessory movements of joints, 212
voluntary, to improve muscle force applied force to vertebrae, 219, 220 in physiologic movements of joints,
production, 173-174 Force-length relationship, of muscle 211-212, 272
supplements to, 175-176, 176 mechanics, 91-93, 91-93 Handicapped gait, Hoffer’s classification
Extension range, limitation of, knee Forceplate, dynamic, in measurement of of, 244
orthosis for, 253-254 force, 17-19, 18 Haversian canal, 68-69, 69, 70
Extensor muscle(s), 110, 114, 115-116 Force-time-histories, in spinal Heel-off to toe-off phase, of gait cycle,
Extensor muscle exercise, for low-back manipulative therapy, 225-226, 240, 241
injury, 134-135, 135 226, 2211, 228, 228-229 Heel-strike to foot-flat phase, of gait
Force-velocity relationship, of muscle cycle, 240, 241
Extracellular matrix (ECM), of
cartilaginous tissue, 40 mechanics, 93-94, 94 Helical axis of motion (HAM), in •
Fracture(s), end plate, 105, 106, 108 accessory movements of joints, 212
of intervertebral disk, 45
fatigue (stress), 79, 79-80 in physiologic movements of joints,
surrounding chondrocytes, 36-37
healing of, accelerometer in, 27, 27 211-212, 272
Extremity(ies), lower, muscles of,
mechanics of, 76-77, 77-78, 79 Hip joint, pattern of resistance of, to
architectural features of, 87t
strength of, variation in, 73, 73 flexion moments, 224, 224
orthotic devices for, 239-264. See
toughness of. See also Critical stress Hip-knee-ankle-foot orthosis, 258-264
also Orthosis; specific type,
intensity factor. walking style with, 259-263
e.g., Ankle-foot orthosis.
defined, 77 energy cost of, 263-264
upper, elevation of, glenohumeral mo¬
Frenchay Arm Test, 169
tion in, 160 reciprocal walking as, 260, 262-
Functional Independence Measure, 169
scapular plane in, 154—155, 155 263, 262-263
scapular rotation in, 145-146 principles of, 262
Gait, handicapped, classification of, 244 swing-through gait as, 259, 260
swing-through, with hip-knee-ankle- swivel walking as, 259-260, 261-
Fascia, lumbodorsal, 118-119 foot orthosis, 259, 260 262
role of, 121-122 Gait cycle, phases of, 240, 240-241 Hoffer’s classification, of handicapped
Fatigue fracture, 79, 79-80 Gait speed, determinants of, 170-171 gait, 244
Fatigue life, of tendons, 58, 59 Gait-assistive devices, as compensation Horizontal abduction exercise, for
Fatigue tolerance levels, of for reduced muscle force scapulothoracic complex, 148
intervertebral disk, 53 production, 177-178, 178 Household walkers, 244
Feature extraction, in optical imaging Gastrocnemius muscle, architectural Humerus, elevation of, 154—155, 155
systems, 22-23, 23 features of, 871 Hyaline cartilage, 65. See also
Femur, rotation of tibia on, resisting, force-length relationship in, 92, 92 Cartilage, articular.
knee orthosis for, 254—255 Glenohumeral joint, 149-160 Hyaluronan, in cartilage, 37, 39
Fiber bundles, in ligament and tendon, arthrokinematics of, 155-156 Hysteresis, of disk tissue, in motion
organization of, 54, 54 articular surfaces of, 149, 149-150 segments, 50
Fiber length, in muscle architecture, capsule of, 150-151
85-86, 88 coracoacromial arch in, 152-153,
Fibrillation, of collagen fibers, 43 153-154 I band, of sarcomere, 84, 84
270 Index
Iliocostalis muscle, 110, 114, 116 Joint(s). See also named joint, e.g., Knee-ankle-foot orthosis (Continued)
Image calibration, in optical systems, Glenohumeral joint. combination function of, 255
21-22 manual therapy for, 209-236 for control of valgus-varus moments,
Imbalance, dynamic, 11 kinematics of, 210, 210-213, 212- 255
Impingement, of glenohumeral joint, 213 joint location in, 257-258
152-153, 153 kinetics of, 213, 213-214 joint specification in, 257
Impulse, defined, 11 process of, 214-215 release of, method of, 257, 258
Impulsive motion, 28 technique(s) of, manipulation as, structural design considerations in,
Indentation test, of cartilage 209, 224—235. See also Spinal 256, 256-257
compression, 40, 41t, 42 manipulative therapy (SMT). weight-relieving, 255-256
Industrial pressure transducers, 30 mobilization as, 209, 215-224. Kyphosis, 44
Inertia, moments and products of, See also Mobilization.
mathematically defined, 11 mechanical behavior of, 213, 213-214
measurement of, 29 movements of, 211-213 Lacunae, chondrocytes in, 36
Infraspinatus muscle, in glenohumeral accessory, 210, 212-213 in bone, 69, 71
joint kinetics, 157, 157-158 physiologic, 210, 211-212, 212- Lamellae, circumferential, 69, 70
Injury, craniocerebral, 202-203 213 interstitial, 69, 70
defined, 122 osteonal, 68-69, 69
orthotic, location of, 257-258
low-back. See Low-back injury. Lamina, of thoracolumbar vertebrae,
specification of, 257
spinal cord, electrical stimulation for, 104, 105
stiffness of, 119, 120t. See also Lum¬
173 Latissimus dorsi muscle, in
bar mobilization stiffness.
muscle strength impairment associ¬ glenohumeral joint kinetics,
surfaces of, linear movement (glide)
ated with, 167 158-159
between, 210
traumatic brain, muscle strength im¬ Lifting, loads on low back during,
rolling of, 210-211
pairment associated with, 167 119-121, 120t
Joint capsule, glenohumeral, 150-151
Instrumentation, 14-24. See also Ligament(s), 53-57. See also named
Joint loads, during manual therapy, 213
Measurement. ligament, e.g., Cruciate ligament,
Joint motion, goniometer for measuring,
for measuring force, 14—19, 15 anterior.
20, 20
dynamic forceplate and, 17-19, 18 aging of, 56-57
strain gauge technology and, 15- arising from coracoid process, 143,
17, 16 143
Keratan sulfate, in cartilage, 37, 39
for measuring position, 19-24 as connective tissue, 65-66
Kinematics, of acromioclavicular joint,
Interclavicular ligament, 141-142 biomechanical considerations for,
143-144, 144
Interspinous ligament, 117, 118 54-56
of manual joint therapy, 210-213
Interstitial water, in articular cartilage, general features of, 53
body segment and joint surface
39-40 healing and repair of, 57
movements in, 210-211
fluid and solid phases of, 40 mechanical properties of, mathemati¬
joint and bone structures in, 210,
Interterritorial region, of extracellular cal models for, 55-56
matrix, 37 210
joint movements in, 210, 211-213 of glenohumeral joint, 151, 151-152
Intertraversarii muscle, 109-110, 114
accessory, 212-213 stress-strain curve for, 55, 55
Intervertebral disk, 44—53
physiologic, 211-212, 212-213 structure and composition of, 53t, 53-
aging of, 51-53, 52t
54, 54
anulus fibrosus of, 44, 45t, 45^16, of scapulothoracic complex, 145-146
of sternoclavicular joint, 142, 142 tensile properties of, 55, 56t
45-46, 107
of thoracolumbar spine, 119t, 119— thoracolumbar, mechanical failure of,
biomechanical studies of, 47^-8
121, 120t 118
creep behavior of, 49-50
Linear elasticity, 14
degeneratively altered, 52 Kinetic energy, 12
Load carriage, of articular cartilage, 40
end plates of, 44, 45t, 46-47, 107 Kinetics, of glenohumeral joint,
of intervertebral disks, 47, 48
creep behavior of, 51 axiohumeral musculature in,
of tendons, 58
fracture of, 105, 106, 108 158-159
Load tolerance values, of motion
noncellular components of, 45t scapulohumeral musculature in,
segment of disk, 52, 52t
fatigue tolerance levels of, 53 156-158
Longissimus muscle, 110, 114, 115-116
general features of, 44 of manual joint therapy, 213, 213-214
Longitudinal ligament(s), lumbar,
load carriage of, 47, 48 of scapular motion, 146-149, 146-
116-117, 117
mechanical properties of, 50, 50-51 149
Lordosis, 44
motion segment of, 46, 47 Knee, flexed, line of force relative to,
degree of, as factor in lumbar mobili¬
average load tolerance values of, 244
zation stiffness, 222
52, 52t pattern of resistance of, to anterior Low back, loads on, during lifting and
hysteresis in, 50 and posterior forces, 224, 225 walking, 119-121, 120t
mechanical properties of, 48, 48- Knee mechanics, control of, ankle-foot Low-back injury, biomechanics of,
50, 49t orthosis for, 248-249 122-125
nucleus pulposus of, 44, 45, 45t, 45- Knee orthosis, 252-255 example of, 124-125
46, 107 for limiting flexion and extension prevention of, 125-129
stresses and strains of, mathematical range, 253-254 activity history in, 128
models for, 51 for resisting anteroposterior relative biomechanics in, 125-129
structure and composition of, 44-47 moment, 254, 254 changes in spinal function and,
thoracolumbar, 107-109, 108 for resisting rotation of femur on 125, 128
Intraabdominal pressure, role of, in tibia, 254-255 knowledge in, 129
thoracolumbar spine support, 121 for resisting valgus-varus moments, motor control in, 129
Intracranial tumors, muscle strength 252, 253 muscle co-contraction and spinal
impairment associated with, 167 Knee-ankle-foot orthosis, 255-258 stability in, 128-129
Index 271
Low-back injury (Continued) Measurement (Continued) Muscle(s) (Continued)

objective strategies in, 124 standard deviations in, 3 cross-section of, 84
posture in, 125, 126, 127t Mechanical loading, and process of low- force-length relationship of, 91-93,
process of, mechanical loading and, back injury, 122-124, 123-124 91-93
122-124, 123-124 Mechanical modeling, 5-6, 5-6 force-velocity relationship of, 93-94,
rehabilitation programs for, 129-135. Mid-stance phase, of gait cycle, 240, 94
See also Rehabilitation pro¬ 241 functional unit of, 84, 84—85
grams, for low-back injury. Military press exercise, for scapulothora- length-associated changes in, 92-93,
Lower extremity muscles, architectural cic complex, 147 93
features of, 87t Mobilization, 209, 215-224 movement of, sliding filament theory
Lower extremity orthosis, 239-264. See applied forces in, frequency of, 2lb- of, 89
also Orthosis; specific type, e.g.. 217 physiologic strength of, 95
Ankle-foot orthosis. grading of, 215, 215 static and dynamic properties of, 91,
Lumbar mobilization stiffness, factor(s) patient response(s) to, 217-224 91-92
associated with alteration in, alteration in lumbar stiffness as, statistically determinate systems of,
219-224, 220, 223t 219-224, 223t 96- 97, 97
breathing as, 222-223 breathing and, 222-223 statistically indeterminate systems of,
couch padding as, 223 couch padding in, 223 97- 99, 98, 100
degree of lordosis as, 222 degree of lordosis in, 222 structure of, 83-85, 84-86
location of target bone as, 221-222 location of target bone in, tension generation and transmission
mobilization frequency as, 221 221-222 in, mechanism of, 87, 89-90, 90
mobilization frequency in, 221 thoracolumbar, 109-116
mobilizing force magnitude as,
mobilizing force magnitude in, abdominal wall, 114, 116, 117
222, 223t
extensor, 110, 114, 115-116
muscle activity as, 221 222, 223t
negative shear forces from, 120t,
other variables as, 223-224 muscle activity in, 221
120-121
skin-fold thickness and body mass other variables in, 223-224
psoas, 116
index as, 223 skin-fold thickness and body
quadratus lumborum, 116
Lumbar posture, in prevention of low- mass index in, 223
rotatores and intertraversarii, 109-
back injury, 125, 126, 127t bone movements as, 217
110, 114
Lumbar spine. See also Thoracolumbar tissue loads as, 219
size of, 109, 110, lilt, 112t, 113t,
spine. tissue resistance as, 219, 220
115t
application of manual force to, 220, vertebral movements as, 217—
Muscle action, analysis of, 94—95
220 219, 218 Muscle activity, as factor in lumbar
stiffness of, 219-224. See also Lum¬ range of, 216, 216-217
bar mobilization stiffness. frequency of, as factor in lumbar stiff¬ Muscle architecture, 83, 85-87, 87t,
Lumbodorsal fascia, 118-119 ness, 221 88-89
role of, 121-122 resistance to, at peripheral joints, 224, fiber length in, 85-86, 88
224-225 index of, 86, 88
Mobilizing force magnitude, as factor in pennation in, 87, 89
M region, of sarcomere, 84, 84 lumbar stiffness, 222, 223t physiologic cross-sectional area in,
Main movement, of joints, 211 Modulus of elasticity, of orthotic 86-87, 89
Manual force, application of, to lumbar materials, 246 Muscle contraction(s), reflex, in spinal
spine, 220, 220 Moment, as performance variable, 10, manipulative therapy, 232
Manual therapy, 209-236 10 Muscle endurance, impairments in, 168,
dose in, methods for measurement of, Moments of inertia, 11 168
214 measurement of, 29 Muscle fiber, 83-84, 84
kinematics of, 210, 210-213, 212- Momentum, angular, defined, 11 Muscle force, active component of,
213 as performance variable, 10-11 165-166
kinetics of, 213, 213-214 defined, 10-11 decreased speed of movement in, neu¬
process of, 214—215 Motion, ballistic, 5 rologic disorders and, 168
technique(s) of, manipulation as, 209, impulsive, 28 passive component of, 165, 166
224—235. See also Spinal manipu¬ jerky, 28-29 Muscle force production, determinants
lative therapy (SMT). oscillatory, 26-27, 27 of, 165-166, 166
mobilization as, 209, 215-224. See Motion segment, of intervertebral disk, during standing, 169-170
also Mobilization. 46, 47 during walking, 170-171
outcome evaluation of, 214-215 average load tolerance values of, intervention(s) for improvement of,
Marker systems, for position measuring, 52, 52t 171-176
24 hysteresis in, 50 cutaneous stimulation as, 173
in optical imaging, 23 mechanical properties of, 48, 48- electrical stimulation as, 173, 175
Mass, as performance variable, 6, 6-7 50, 49t immediate effects seen in, 171-173
defined, 6 Motor control, in prevention of low- lasting effects seen in, 173-176
Measurement, 2-5. See also back injury, 129 muscle stretch as, 171-172, 172
Instrumentation. Multifidus muscle, 110, 114 targeting antagonist restraint as,
accuracy and precision in, 3, 3 Multiple sclerosis, muscle strength ^ 174-175
components of, 2-3 impairment associated with, 167 vibration as, 173
corresponding accuracy of, defined, 3 Muscle(s), 83-100. See also named voluntary exercise as, 173-174
muscle, e.g.. Gastrocnemius supplements to, 175-176, 176
drift in, 4, 4
muscle. nature of, 166-168, 167
errors in, 3^1
co-contraction of, and spinal stability, reduced, compensation(s) for, 176—
sampling frequency in, 4-54
128-129 178
sensitivity in, 3, 3-4
272 Index
Muscle force production (Continued) Orthosis (Continued) Pedicle, of thoracolumbar vertebrae,

altered techniques as, 176, 177 as compensation for reduced muscle 104, 105
modification of demands as, force production, 178 Pennation, in muscle architecture, 87,
176-177 engineering consideration(s) in, bio¬ 89
provision of assistive devices as, mechanics as, 244, 244-245 Performance, mechanics of, 2
177, 177-178 materials as, 245, 245-246 Performance variable(s), 6-13
sit-to-stand maneuver in, 169, 170 structural properties as, 246-247, angular momentum as, 11
statistical relationships in, studies of, 247 classic, 6
169, 169 evaluation of, 242-244 displacement and velocity as, 9, 9
Muscle moment, 95 classification of handicapped gait energy as, 12
relationship of, to joint angular posi¬ in, 244 force as, 9-10, 10
tion, 95-96 monitoring ground reaction forces impulse as, 11
Muscle strength, impairments of, in, 242, 243 mass as, 6, 6-7
166-168, 167 monitoring relative energy expendi¬ moment as, 10, 10
Muscle stretch, to improve muscle force ture in, 242, 244 moments of inertia/products of inertia
production, 171-172, 172 hip-knee-ankle-foot, 258-264. See as, 11, 11
Muscle-joint system, determinate, also Hip-knee-ankle-foot ortho¬ momentum as, 10-11
96-97, 97 sis. position as, 7-8, 7-9
indeterminate, 97-99, 98, 100 knee, 252-255. See also Knee ortho¬ power as, 13
Myofibril, 84 sis. work as, 11-12
Myosin filaments, 85, 86 knee-ankle-foot, 255-258. See also Pericellular region, of extracellular
Knee-ankle-foot orthosis. matrix, 37
principles and benefits of, 239-240 Peroneus muscles, architectural features
Nerve supply, to tendons, 57
role of, 239 of, 87t
Neural arch, of thoracolumbar vertebrae,
treatment objectives of, 264 Physiologic cross-sectional area
104, 105
types of, 247-264 (PCSA), in muscle architecture,
Neurologic disorder(s), 165-171. See
Orthotic device(s),'brittle materials for, 86-87, 89
also specific disorder, e.g., Stroke.
muscle endurance impairment in, 168,
245, 245-246 Physiologic movements, of joints, 210,
168 ductile materials for, 245, 246 211-212, 212-213
range of, 239 Plantar-flexion-dorsiflexion, resisting,
muscle strength impairment in, 166-
168, 167 Orthotic joints, location of, 257-258 ankle-foot orthosis for, 247-250,
specification of, 257 248-249
rate of force generation and speed of
movement impairment in, 168 Oscillations, amplitudes of, in able- Plastic deformation, of orthotic
Neurologic therapy, 165-179 bodied subjects, 198-200, materials, 246
muscle force production in, 165-171. 199-200 Plastic region, of stress-strain curve, 72
See also Muscle force produc¬ in below-knee amputation subjects, Poliomyelitis, muscle strength
tion. 204 impairment associated with, 167
determinants of, 165-166, 166 in cerebrovascular accident pa¬ Position(s), measuring of, 19-24
implications of, 168-171, 169-170 tients, 201-202 goniometers in, 19-20, 20
interventions for improvement of, in craniocerebral-injured patients, marker systems in, 24
171-176 203 nonoptical imaging systems in, 24
nature of, 166-168, 167-168 in standing sway, 194, 195, 195 optical imaging systems in, 20-23
reduced, compensations for, 176— natural frequency of, 26, 27 Position of object, as performance
178, 177 Oscillatory motion, 26-27, 27 variable, 7-8, 7-9
Nonoptical imaging systems, for Osteoarthritis, 43 mathematical equations for, 8
position measuring, 24 Osteoblasts, 69 Position time derivative(s), 24, 24—30
Nucleus pulposus, of intervertebral disk, Osteoclasts, 69 acceleration and, 25-28, 25-28
44, 45, 45-46, 107 Osteocytes, 69, 71 angular velocity and acceleration mea¬
noncellular components of, 45t Osteokinematics, of glenohumeral joint, surements and, 29
153-155, 155 jerk and, 28-29
Osteons, 68-69, 69, 70 moment and product of inertia mea¬
Optical imaging systems, camera
surements and, 29
accuracy in, 21
pressure measurements and, 29-30
conversion from 2 to 3 dimensions in, Parkinson’s disease, muscle strength velocity and, 24-25
22 impairment associated with, Postural balance, measurement of, 189
feature extraction in, 22-23, 23 167-168 Posture, lumbar, in prevention of low-
for position measuring, 20-23 Patient, chairbound, 244 back injury, 125, 126, 127t
image calibration in, 21-22 Patient response(s), to mobilization Potential energy, 12
markers in, 23 forces, alteration in lumbar Power, as performance variable, 13
recording media in, 21 stiffness as, 219-224, 223t Precision, in measurement, 3, 3
six-dimensional position of body seg¬ bone movements as, 217 Press-up exercise, for scapulothoracic
ments in, 23 tissue loads as, 219 complex, 148
Orientation, influence of, in stress-strain tissue resistance as, 219, 220 Pressure, measurement of, 29-30
behavior of bone, 72-73, 72-74 vertebral movements as, 217-219, Pressure transducers, industrial, 30
Orthosis, 239-264 218 Products of inertia, 11, 77
American Academy of Orthopaedic PCSA (physiologic cross-sectional area), measurement of, 29
Surgeons definition of, 239 in muscle architecture, 86-87, 89 Propulsive forces, in walking cycle, 241
and walking cycle, 240-241, 240-242 Pectoralis major muscle, in Prosthesis alignment, influence of, on
ankle-foot, 247-252. See also Ankle- glenohumeral joint kinetics, balance, after below-knee
foot orthosis. 158-159 amputation, 204-205
Index 273
Proteoglycans, in articular cartilage, Scalar quantity, of mass, 6 Spinal manipulative therapy (SMT)
37-39, 39 Scapular motion, kinetics of, 146-149, (Continued)
Protraction, of scapula, 146 146-149 cracking sounds in, 229-232, 230-
Psoas muscle, 116 trapezius and rhomboid muscles in, 231
Push-up exercise, for scapulothoracic 148-149, 149 force directions in, 228-229
complex, 148 trapezius and serratus anterior mus¬ force-time-histories in, 225-226, 226,
cles in, 146-147, 146-148 227t, 228, 228-229
Scapular position, 144—145, 145 multiple successive, 229
Quadratus lumborum muscle, 116 Scapular rotation, 145-146 reflex muscle contraction in, 232
Quadriplegia Index of Function, 169 Scapular tilting, 146 vertebral movements in, 232-235,
Scapular tipping, 146 233-235
Scapular winging, 143 Spine, bones of. See Vertebrae.
Range of motion, of thoracolumbar
Scapulohumeral musculature, in lumbar, application of manual force
spine, 119t
glenohumeral joint kinetics, to, 220, 220
Reciprocal walking, with hip-knee- stiffness of, 219-224. See also Lum¬
156-158
ankle-foot orthosis, 260, bar mobilization stiffness.
Scapulohumeral rhythm, of glenohu¬
262-263, 262-263 stability of, muscle co-contraction
meral and scapulothoracic joints,
principles of, 262
160 and, 128-129
Recording media, in optical imaging
Scapulothoracic complex, 144—149 thoracolumbar, 103-136. See also Tho¬
systems, 21 racolumbar spine.
kinematics in, 145-146
Rectus femoris muscle, architectural
scapular position in, 144—145, 145 Spring balance, in measurement of
features of, 87t
scapulohumeral rhythm of, 160 force, 15, 15
Reflex muscle contraction, in spinal
Schmorl’s node, 105, 106 Stability, spinal, muscle co-contraction
manipulative therapy, 232
Semimembranosus muscle, architectural and, 128-129
Rehabilitation exercises, for
features of, 87t Stabilization of body, in walking cycle,
scapulothoracic complex, 147,
force-length relationship in, 92, 92 241
147-148 Stance phase, ground reaction force
Semitendinosus muscle, architectural
Rehabilitation programs, for low-back
features of, 87t during, 242, 243
injury, 129-135
Sensitivity, in measurement, 3, 2-4 Standing, bipedal, 189
exercises in, 130
Serratus anterior muscle, in scapular studies of force during, 169-170
abdominal muscle, 133t, 133—
motion, 146-147, 146-148 symmetric, defined, 197
134, 134
Sharpey’s fibers, 46, 107 Standing balance. See Balance.
aerobic, 132-133
Shear forces, negative, from Standing still, modeling of body while,
extensor muscle, 134—135, 135
thoracolumbar muscles, 120t, 192-194
flexibility issues in, 130-131, 131-
120-121 in sagittal plane, 193
132 swaying motion while, nature of, 192
Shearing stresses, calculation of, 13
strength and endurance issues in,
Shoulder complex, 141-161. See also Standing tests, perturbed, 191
131-135
specific joint, e.g., Acromioclavicu¬ unperturbed, 189-191, 190
use of abdominal belts in, 135
lar joint. parameters in, 191
Relative energy expenditure, monitoring
integrated function of, 160-161 Sternoclavicular joint, kinematics of,
of, in orthotic treatments, 242, 244
Sickness Impact Profile scores, 169 142, 142
Relative sequence of tangential force
Sit-to-stand maneuver, studies of, 169, structure of, 141, 141-142
vectors (RSTFVs), 195, 196
170 Sternoclavicular ligament, 142
in able-bodied subjects, 198, 198t
technique of, in reduced muscle force Stiffness, joint, 119, 120t. See also
in cerebrovascular accident patients,
production, 176, 177 Lumbar mobilization stiffness.
201 Stiffness values, of articular cartilage,
Six-dimensional position, of body
in craniocerebral-injured patients,
segments, in optical imaging 41, 41t
202-203 of spinal motion segments, 49t
systems, 23
Resistance, relation between force and, of thoracolumbar spine, 120t
Skin-fold thickness, as factor in lumbar
16, 16-17 Stimulation, to improve muscle force
Retrotorsion angles, humeral and production, cutaneous, 173
Sliding filament theory, of muscle
scapular, 149, 149 electrical, 173, 175
movement, 89
Rhomboid muscle, in scapular motion,
SMT. See Spinal manipulative therapy Strain, 13
148-149, 149 defined, 13
(SMT).
Ring lock joint, in knee-ankle-foot Strain gauges, in measurement of force,
Soft tissue, 35-60. See also specific
orthosis, 257, 258 15-17, 16
tissue for details,
Rotator interval, of joint capsule, 150 Strength, ultimate, of orthotic materials,
cartilaginous, 35-44
Rotatores muscle, 109-110, 114 246
intervertebral disc as, 44-53
Rowing exercise, for scapulothoracic Strength issues, in rehabilitation
ligamentous, 53-57
complex, 147 programs, for low-back injury,
tendinous, 57-59
Rupture, of ligaments, 56 131-135
Soleus muscle, architectural features of,
87t Stress, 13, 13-14
Solid phase, of interstitial water, 40 as function of strain rate, for compact
Sampling frequency, in measurement, 4, Spasticity, defined, 166 bone, 73, 74
Spinal cord injury, electrical stimulation buckling, of thoracolumbar vertebrae,
4-5
for, 173 105, 106
Sarcomere, 84, 84
muscle strength impairment associ¬ Stress fracture, 79, 79-80
length of, relationship between devel¬
ated with, 167 Stress lines, in long bone, 70, 71
oped tension and, 89, 90
Spinal manipulative therapy (SMT), Stress-relaxation tests, of articular
Sartorius muscle, architectural features
209, 224-235 cartilage, 41, 41t
of, 87t
274 Index
Stress-strain behavior, of bone, 72-76 Tensile properties, of ligaments, 55, Tumor(s), intracranial, muscle strength
influence of age in, 75-76, 76 56t impairment associated with, 167
influence of density in, 75, 75 Tensile strength, ultimate, of tendons,
influence of orientation in, 72-73, 58
72-74 Tension, generation and transmission of, Ultimate strength, of orthotic materials,
viscoelasticity in, 73, 74, 75 in muscle, mechanism of, 87, 246
of collagen, 67, 67-68 89-90, 90 Uniaxial compression test, for cartilage,
Stress-strain curve(s), for bovine bone, of collagen fibers, 42-43 40, 4It, 42
72, 74 Tensor, inertia, 11 Upper extremity elevation,
for cortical bone, 72, 72 stress, 13 glenohumeral motion in, 160
for ligaments, 55, 55 Teres minor muscle, in glenohumeral scapular plane in, 154-155, 155
plastic region of, 72 joint kinetics, 157, 157-158 scapular rotation in, 145-146
Stroke. See also Cerebrovascular Territorial region, of extracellular
accident. matrix, 37
electrical stimulation for patient with, Therapeutic walkers, 244 Valgus-varus moments, control of, knee-
173 Thoracolumbar spine, 103-136. See also ankle-foot orthosis for, 255
muscle strength impairment associ¬ Lumbar spine. resisting, ankle-foot orthosis for, 250-
ated with, 166-167, 167 anatomic features of, clinical rele¬ 251, 251
sit-to-stand maneuvers in patient with, vance of, 103-119 knee orthosis for, 252, 253
169, 169 exercise prescription for, 135-136 Vastus muscles, architectural features of,
standing studies in patient with, 169— function of, changes in, 125, 128 87t
170 injury to. See Low-back injury. Vehicle vibration, 27
walking studies in patient with, 170- kinematics and kinetic properties of, Velocity, angular, measurement of, 29
171 119t, 119-121, 120t as performance variable, 9, 9
weight loading and ground reaction ligaments of, 116-118, 117-118 defined, 9
forces in patient with, 171 mechanical failure of, 118 measurement of, 24-25
Subacromial space, in coracoacromial Iumbodorsal fascia of, 118-119 Vertebrae, applied force to, force-
arch, 152 role of, 121-122
displacement relationship of, 219,
Subscapularis muscle, in glenohumeral memory of, 128 220
joint kinetics, 157, 157-158 muscles of, 109-117. See also Mus-
movements of, in joint mobilization,
Superficial zone, in articular cartilage, cle(s), thoracolumbar. 217-219, 218
36, 36, 37 normal biomechanics of, 103-122
in spinal manipulative therapy,
Supraspinatus muscle, in glenohumeral range of motion of, 119t
232-235, 233-235
joint kinetics, 156-157 stability of, muscle co-contraction
thoracolumbar, 103-109
Supraspinatus outlet, in coracoacromial and, 128-129
body of, 103, 104, 105
arch, 152, 153 stiffness value of, 120t
buckling stresses in, 105, 106
Supraspinous ligament, 117 support of, role of intraabdominal
disks between, 107-109, 108
Swaying motion, bilateral measurement pressure in, 121
end plate fracture of, 105, 106, 108
of, 194-197 vertebrae of, 103-109. See also Verte¬
posterior elements of, 105, 107,
center of gravity in, estimation of brae, thoracolumbar.
107-108
Tibia, rotation of femur on, resisting,
trajectory of, 196-197 trabecular arrangement of, 103, 105
knee orthosis for, 254—255
defined parameters in, 195-196, Vertical foot-ground reaction forces, in
Tibialis anterior muscle, architectural
195-196 measurement of balance, 189
features of, 87t
methodology of, 194-195 Vertical movement of trunk, in walking
Tidemark, in articular cartilage, 36,
modeling of, 192-194, 193-194 cycle, 240, 241
36
nature of, 192 Vibration(s), human response to, 28, 28
Tissue, soft. See Soft tissue.
Swing phase, of gait cycle, 240, 241 to improve muscle force production,
Tissue loads, during joint mobilization,
Swing-through gait, with hip-knee- 173
219
ankle-foot orthosis, 259, 260 Tissue mechanics, 2 vehicle, 27
Swivel walking, with hip-knee-ankle- Tissue resistance, to joint mobilization, Viscoelasticity, in stress-strain behavior
foot orthosis, 259-260, 261-262 219, 220 of bone, 73, 74, 75
Synovial joint. See also specific joint, Tissue-engineered repair systems, for Voluntary exercise, to improve muscle
e.g., Sternoclavicular joint. articular cartilage, 43^14 force production, 173-174
structure of, 210, 210 Total sway activity, defined, 196 supplements to, 175-176, 176
Trabeculae, arrangement of, in
thoracolumbar vertebrae, 103,
Target bone, location of, as factor in 105 Walker(s), community, 244
lumbar mobilization stiffness, household, 244
Transition zone, in articular cartilage,
221-222 platform, 177, 178
36, 36, 37
Tendon(s), 57-59. See also named swivel, 261, 262
Transverse process, of thoracolumbar
tendon, e.g., Biceps brachii tendon. vertebrae, 104, 105 therapeutic, 244
aging of, 59
Trapezius muscle, in scapular motion, Walking, 240-242
as connective tissue, 65-66 lower, 146-147, 146-148 determinants of gait speed in, 170-
biomechanical studies of, 58-59, 59 middle, 148-149, 149 171
exertion of muscle force on, 90 Trauma. See Injury. loads on low back during, 119-121
general features of, 57 Tropocollagen molecules, in bone, 66 120t
healing and repair of, 59 66-67 mechanical specifics of, 171
load carriage of, 58 in cartilage, 37, 38 normal gait in, 240, 240-241
structure and composition of, 58 Trunk, vertical movement of, in walking vertical movement of trunk in, 240
Tensile fatigue studies, 43 cycle, 240, 241 241
Index 275
Walking (Continued) Water, interstitial, fluid and solid phases Work, as performance variable, 11-12
weight shifting during, 170 of, 40 defined, 11
with hip-knee-ankle-foot orthosis, in articular cartilage, 39—40
259-263 Weighing, in measurement of force, 15,
energy cost of, 263-264 15 Y derivative, of position vector, 24
reciprocal walking as, 260, 262- Weight loading, in stroke patient, 171 with respect to time, 24
263, 262-263 Weight-bearing imbalance, defined, 196 Yield point, of orthotic materials,
principles of, 262 Weight-relieving orthosis, 255-256 246
swing-through gait as, 259, 260 Wheatstone bridge equations, 16
swivel walking as, 259-260, 261- Wheelchairs, as compensation for reduced
262 muscle force production, 177 Z disc, of sarcomere, 84, 84
I
UiKAfty
Amazing Help.
http://nihHbrary.nlh.,
i.gov
i»„ti T°Ce"*erD rive

heldarMD 20892-1750
Zeevi Dvir, PhD, LLB
Written by an interdisciplinary team of internation

this reference is the most practical, clinically orient
of its kind. In a clear, easy-to-read style it presents
guidance you need to apply biomechanical principl
practice.
This superb book ...

• Explains the fundamental concepts of biomechani
• Analyzes the mechanical interactions of various ti
I
systems, including bones, ligaments, tendons, an
• Examines the complex muscle-joint interactions i
and shoulder to illustrate principles such as muse
and joint load distribution.
• Demonstrates the specific applications of biomec
various clinical areas, such as body equilibrium a
I
treatment of patients with neurologic disorders, sp
manipulation, and orthotics.
• Simplifies complex concepts with over 200 illustrations.
Clinical Biomechanics explores the role that biomechanics
plays in both dysfunction and rehabilitation. Its diverse group
of authors offers you critical information about applying
biomechanical principles to more effectively treat a full range of
conditions.
yj>
Churchill Livingstone
A Harcourt Health Sciences Company

Zeevi Dvir - Clinical Biomechanics-Churchill Livingstone (2000)

Uploaded by

Copyright:

Available Formats

Zeevi Dvir - Clinical Biomechanics-Churchill Livingstone (2000)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Zeevi Dvir - Clinical Biomechanics-Churchill Livingstone (2000)

Uploaded by

Copyright:

Available Formats

NATIONAL INSTITUTES OF HEALTH

A Harcourt Health Sciences Company

The Curtis Center

Library of Congress Cataloging-in-Publication Data

Clinical biomechanics / [edited by] Zeevi Dvir.

1. Orthopedics. 2. Human mechanics. 3. Musculoskeletal

RD732. C575 2000 616.7—dc21

Acquisitions Editor: Andrew Allen

Copyright © 2000 by Churchill Livingstone.

Churchill Livingstone® is a registered trademark of Harcourt, Inc.

Printed in the United States of America.

Last digit is the print number: 9 8 7 6 5 4 3 2 I

Dan L. Bader, PhD Walter Herzog, PhD

JOHN BEHIR1, PHD Judith Laprade, BSc(PT), MSc

Richard W. Bohannon, EdD, PT, NCS Michael Lee, M Biomed Eng

Stuart M. McGill, PhD

Joseph Mizrahi, DSc

Zeevi Dvir, PhD, LLB Joseph J. Sarver, BSME, MS Biomed Eng

Julianna Gal, PhD Rami Seliktar, PhD

Back in 1975, as a first-year PhD student at the Bioengineering Unit, Strathclyde

2 Biomechanics of Soft Tissues 35

5 Biomechanics of the Thoracolumbar Spine 103

6 Biomechanics of the Shoulder Complex 141

7 Biomechanics of Neurologic Treatment 165

8 Biomechanics of Unperturbed Standing Balance 189

9 Biomechanics of Manual Therapy 209

10 Lower Limb Orthotics 239

Overview some point they appear gray. Nothing physically

calibration. Without prior information about the

error; however, instrument A is obviously less

amplitudes), a jackhammer being used outside

pies is to use multiple instruments to measure the

Example nonstationary signal Reconstructed nonstationary

FjGURE_1 ^ Example of a nonstationary signal and sampling

In reality, a measured signal is unlikely to have

ysis assuming either particles or rigid bodies. We

CLASSIC MECHANICS VARIABLES

As an introduction to our discussion of biome¬

property, which is usually measured indirectly

Position (and Its Time Derivatives)

The following section may sound complicated to

We can calculate the position of the ball in terms

variable, depending on the modeling tools se¬

Displacement and Velocity

ma, where F is force, m is mass, and a is acceler¬

FIGURE 1—15 Depiction of ground reaction moment.

Impulse is defined as the time integral of the force

Angular momentum is defined as H = Sr X mv,

Ixy = f xydm Work is defined analytically as W = / F • dr,

move to produce work. It is also important to

Energy is a rather complicated variable, and only W = / F • d? = — A V

which is the principle of conservation of mechani¬

Power In the example depicted in Figure 1-17, the

MECHANICS OF MATERIALS (STRESS AND

It is important to consider the mechanical behavior

Because in real loading conditions the stresses

Force can be measured through the effects that it

FIGURE 1-21 Beam balance.

example of such electronic force transducers that