University of Southern California Dissertations and Theses

At what age are gait characteristics mature? Evaluation of gait kinematics, kinetics, and intersegmental dynamics in 7 year-old children

(USC Thesis Other)

At what age are gait characteristics mature? Evaluation of gait kinematics, kinetics, and intersegmental dynamics in 7 year-old children

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content AT WHAT AGE ARE GAIT CHARACTERISTICS MATURE?
EVALUATION OF GAIT KINEMATICS, KINETICS, AND INTERSEGMENTAL
DYNAMICS IN 7 YEAR-OLD CHILDREN
by
Kathleen Jodell Ganley
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOKINESIOLOGY)
May 2003
Copyright 2003 Kathleen Jodell Ganley
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 3103890
UMI
UMI Microform 3103890
Copyright 2003 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, Ml 48106-1346
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES, CALIFORNIA 90089-1695
This dissertation, written by
K ath leen J o d e ll G a n l e y _________________
under the direction o f h er dissertation committee, and
approved by all its members, has been presented to and
accepted by the Director o f Graduate and Professional
Programs, in partial fulfillment o f the requirements fo r the
degree o f
DOCTOR OF PHILOSOPHY
Director
Date May 16. 2003
Dissertation Committee
Chair
4X4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Kathleen Ganley Christopher Powers
ABSTRACT
AT WHAT AGE ARE GAIT CHARACTERISTICS MATURE?
EVALUATION OF GAIT KINEMATICS, KINETICS, AND INTERSEGMENTAL
DYNAMICS IN 7 YEAR-OLD CHILDREN
The age at which adult-like gait patterns are expressed has not been
definitively established. The objective of this dissertation was to determine whether
gait patterns of children are mature by 7 years. First, a method for obtaining age-
specific anthropometric data was established. It was demonstrated that dual energy
x-ray absorptiometry (DXA) can be used to quantify lower extremity
anthropometries (segment mass, center of mass location, and moment of inertia) in-
vivo. Next, DXA-derived anthropometric measurements were obtained from 7-13
year-old children (n=50) and compared to cadaver-based estimates. Additionally,
lower extremity net joint moments were calculated for 3 children during gait using a)
DXA-derived, and b) cadaver-based estimates. Statistically significant differences
were identified for DXA-derived and cadaver-based anthropometric values.
However, net joint moments calculated with the two sources of anthropometric
parameters were very similar. Next, joint angles, moments, and power obtained
during level walking in 7 year-old children (n=15) were compared to data from
adults (n=15) using DXA-derived anthropometric data for kinetic calculations. For
most of the variables examined 7 year-olds were similar to adults, providing
evidence that gait kinematics and kinetics of 7-year old children are comparable to
those of adults, at least during the stance phase of gait. Lastly, a detailed analysis of
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the swing phase of gait was performed to test the hypothesis that compared to adults,
7 year-old children would demonstrate a relatively larger contribution of muscle
moment and a smaller contribution of interaction moment to the net knee joint
moment during the swing phase of gait. Net knee joint moments as well as the
contributions of muscle, interaction, and gravity moments to the net joint moment
were compared between 7 year-old children (n=10) and adults (n=10) during self
selected free walking. The knee kinematics and net joint moments, as well as the
contributions of the muscle and interaction moments to the net joint moments, were
similar between the two groups, providing evidence that the control of limb
intersegmental dynamics during swing is similar in 7 year-olds and adults during
level walking. Overall, the results of this dissertation support the premise that
walking patterns in 7 year-old children are mature.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DEDICATION
This work is dedicated to all of the children whom I have had the opportunity
to work with over the past 15 years.
ii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ACKNOWLEDGEMENTS
This dissertation could not have been completed without the guidance of
support of many people. I would like to first acknowledge financial support from the
Pediatric Section of the American Physical Therapy Association, the Whitaker
Foundation, and the Ackerberg Foundation.
I would like to thank my graduate advisor, Dr. Christopher Powers, for giving
me the opportunity to pursue my research interests and for modeling the dedication,
work ethic, and professionalism required of an academic. I would also like to thank
the other members of my dissertation committee, Dr. James Gordon, who is as
committed to the educational process as anyone I have ever known, Drs. Jill McNitt-
Gray and George Salem, whose expertise in biomechanics I depended on, and Dr.
Carolee Winstein, whose exceptional teaching was the foundation of my graduate
work.
I would like to thank the faculty of the Department of Biokinesiology and
Physical Therapy at the University of Southern California for the academic
preparation I have received, and I would like to acknowledge my colleagues in the
Musculoskeletal Biomechanics Research Laboratory: Judy Bumfield, Yu-Jen Chen,
Sean Flanagan, Matt Sandusky, Susan Sigward, and Sam Ward. I especially thank
Susan Sigward for her friendship, Sam Ward for being the first person I go to with
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
any academic question, and Matt Sandusky who has his fingerprint on every piece of
work that leaves the lab.
I would like to thank the following people for their friendship and personal
support as I pursued this goal, my mother who stands behind every decision I make,
Lisette Ackerberg who reminds me that it is “making a difference” that is important,
Dr. Beth Fisher who has taught me how to tell the story, Suzi Zwick who generously
shares her humor, Allison Whiteside who keeps me focused on the clinical
importance of research, and Shelly Watkins who is always there to add perspective.
Finally, I would like to acknowledge Dr. Carl DeRosa and the faculty of Northern
Arizona University’s Physical Therapy Department for introducing me to the
challenges and rewards of physical therapy education.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
DEDICATION................................................................................................................ ii
ACKNOWLEDGEMENTS..........................................................................................iii
LIST OF TABLES...................................................................................................... viii
LIST OF FIGURES........................................................................................................ix
ABSTRACT................................................................................................................... xi
CHAPTER I: OVERVIEW.............................................................................................1
Specific Aims......................................................................................................3
CHAPTER II: LITERTURE REVIEW........................................................................ 4
Development of Gait.......................................................................................... 4
Intersegmental Dynamics...................................................................................6
The Role of Intersegmental Dynamics in Multi-Segment
Limb Movements...................................................................................7
Intersegmental Dynamics in Development..........................................8
The Role of Intersegmental Dynamics in Adult Gait..........................9
Anthropometries................................................................................................12
Summary........................................................................................................... 15
CHAPTER III: DETERMINATION OF LOWER EXTREMITY
ANTHROPOMETRIC PARAMETERS USING DUAL ENERGY X-RAY
ABSORPTIOMETRY...................................................................................................16
Introduction....................................................................................................... 16
Methods............................................................................................................. 18
Subjects.................................................................................................18
Dual Energy X-Ray Absorptiometry................................................. 19
Center of Mass Location........................................................ 21
Moment of Inertia...................................................................21
Reliability of DXA Measurements.........................................22
Cadaver-Based Estimates....................................................................23
Gait Analysis........................................................................................23
Statistical Analysis...............................................................................25
Results............................................................................................................... 25
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Dual Energy X-Ray Absorptiometry and Cadaver-Based
Estimates.............................................................................................. 25
Gait Analysis........................................................................................29
Discussion.........................................................................................................32
Conclusion........................................................................................................35
CHAPTER IV: ANTHROPOMETRIC PARAMETERS IN CHILDREN:
A COMPARISON OF VALUES OBTAINED FROM DUAL ENERGY
X-RAY ABSORPTIOMETRY AND CADAVER-BASED ESTIMATES............ 36
Introduction.......................................................................................................36
Methods............................................................................................................. 37
Subjects.................................................................................................37
Dual Energy X-Ray Absorptiometry................................................. 38
Center of Mass Location........................................................ 40
Moment of Inertia...................................................................40
Cadaver-Based Estimates.................................................................... 41
Gait Analysis........................................................................................41
Statistical Analysis.............................................................................. 43
Results............................................................................................................... 43
Dual Energy X-Ray Absorptiometry and Cadaver-Based
Estimates.............................................................................................. 43
Net Joint Moments...............................................................................47
Discussion.........................................................................................................51
Conclusion........................................................................................................53
CHAPTER V: GAIT KINEMATICS AND KINETICS OF 7
YEAR-OLD CHILDREN: A COMPARISON TO ADULTS USING
AGE-SPECIFIC ANTHROPOMETRIC DATA........................................................55
Introduction.............. 55
Methods............................................................................................................. 57
Subjects.................................................................................................57
Instrumentation....................................................................................58
Procedures............................................................................................ 58
Data Management................................................................................59
Statistical Analysis...............................................................................60
Results............................................................................................................... 61
Discussion.........................................................................................................67
Conclusion........................................................................................................70
CHAPTER VI: INTERSEGMENTAL DYNAMICS DURING THE SWING
PHASE OF GAIT: A COMPARISON OF KNEE KINETICS BETWEEN
7 YEAR-OLD CHILDREN AND ADULTS..............................................................72
Introduction.......................................................................................................73
Methods............................................................................................................. 75
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Subjects.................................................................................................... 75
Instrumentation....................................................................................76
Procedures............................................................................................ 76
Data Management................................................................................77
Statistical Analysis...............................................................................79
Results............................................................................................................... 79
Discussion.........................................................................................................83
Conclusion........................................................................................................86
CHAPTER VII: SUMMARY AND CONCLUSIONS.............................................. 87
REFERENCES.............................................................................................................. 92
APPENDIX A ............................................................................................................... 96
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF TABLES
Table 3-1. Physical Characteristics of Study Participants......................................... 18
Table 3-2. Boundaries Used to Identify Lower Extremity Segments
from DXAScans............................................................................................................21
Table 3-3. DXA-Derived and Cadaver-Based Anthropometric Proportions........... 22
Table 3-4. Equations Used to Calculate Subject-Specific Anthropometric
Values from Cadaver-Based Data................................................................................23
Table 4-1. Physical Characteristics of Study Participants.........................................38
Table 4-2. DXA-Derived and Cadaver-Based Anthropometric Proportions
Used for the Calculation of Net Joint Moments during Gait.................................... 41
Table 5-1. Age-Specific Anthropometric Proportions Used for the
Calculation of Net Joint Moments during Gait...........................................................60
Table 5-2. Comparison of Peak Joint Angles Between 7-Year-Olds
and Adults..................................................................................................................... 61
Table 5-3. Comparison of Peak Net Joint Moments Between
7-Year-Olds and Adults............................................................................................... 63
Table 5-4. Comparison of Peak Power Between 7-Year-Olds and Adults.............. 65
Table 6-1. Comparison of Kinematics and Kinetics Between
7-Year-Olds and Adults................................................................................................80
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Figure 3-1. Method to Calculate the Distance from the Proximal
Segment Boundary to the Center of Mass of the Shank............................................20
Figure 3-2. Comparisons of Segment Mass for the Thigh, Shank,
and Foot as determined by DXA and Cadaver-Based Estimates.............................. 26
Figure 3-3. Comparisons of the Distance form the Proximal
Segment Boundary to the Center of Mass for the Thigh, Shank,
and Foot as Determined by DXA and Cadaver-Based Estimates............................. 27
Figure 3-4. Comparisons of the Moment of Inertia About the Center
of Mass for the Thigh, Shank, and Foot as Determined by DXA and
Cadaver-Based Estimates............................................................................................. 28
Figure 3-5. Net Ankle Joint Moments Calculated with Anthropometric
Data Derived from DXA and Cadaver-Based Estimates...........................................29
Figure 3-6. Net Knee Joint Moments Calculated with Anthropometric
Data Derived from DXA and Cadaver-Based Estimates...........................................30
Figure 3-7. Net Hip Joint Moments Calculated with Anthropometric
Data Derived from DXA and Cadaver-Based Estimates...........................................31
Figure 3-8. Comparison of Average RMSE Scores of the Net Hip,
Knee, and Ankle Joint Moments During Stance and Swing......................................32
Figure 4-1. Comparisons of Segment Mass for the Thigh, Shank,
and Foot as determined by DXA and Cadaver-Based Estimates.............................. 44
Figure 4-2. Comparisons of the Distance form the Proximal Segment
Boundary to the Center of Mass for the Thigh, Shank, and Foot as
Determined by DXA and Cadaver-Based Estimates................................................. 45
Figure 4-3. Comparisons of the Moment of Inertia About the Center
of Mass for the Thigh, Shank, and Foot as Determined by DXA and
Cadaver-Based Estimates............................................................................................. 47
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4-4. Net Ankle, Knee, and Hip Joint Moments Obtained from
a 7-Year-Old.................................................................................................................. 48
Figure 4-5. Net Ankle, Knee, and Hip Joint Moments Obtained from
a 10-Year-Old................................................................................................................ 49
Figure 4-6. Net Ankle, Knee, and Hip Joint Moments Obtained from
a 13-Year-Old................................................................................................................ 50
Figure 5-1. Comparison of Joint Angles at the Hip, Knee, and Ankle
Between 7-Year-Olds and Adults................................................................................62
Figure 5-2. Comparison of Net Joint Moments at the Hip, Knee, and
Ankle Between 7-Year-Olds and Adults..................................................................... 64
Figure 5-3. Comparison of Joint Power at the Hip, Knee, and Ankle
Between 7-Year-Olds and Adults................................................................................66
Figure 5-4. Schematic of the Foot in Late Stance Illustrating the
Relationship Between the Vertical Component of a Ground Reaction
Force and its Lever A rm .............................................................................................. 69
Figure 5-5. Correlation Between Peak Ankle Moment and Foot Length
for All Subjects............................................................................................................. 69
Figure 6-1. Knee Angles, Angular Velocity and Acceleration Obtained
from 7-Year-Old Children and Adults........................................................................ 81
Figure 6-2. Net Joint Moments, Interaction Moments, Gravity Moments,
and Muscle Moments from Adults and 7-Year-Old Children...................................82
Figure 6-3. The Contribution of Muscle Moments to Net Joint Moments
in 7-Year-Olds and Adults and Interaction Moments to Net Joint
Moments in 7-Year-Olds and Adults.......................................................................... 83
x
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I
OVERVIEW
An understanding of the development of normal ambulation is essential for
appreciating how pathology might influence gait patterns. Compared to the plethora
of publications relating to adults, the literature describing gait in developing children
is sparse. Furthermore, studies involving children report conflicting results,
especially with regards to the age at which "mature" gait patterns are expressed.
Further investigation is necessary to determine how gait differs between adults and
children and when adult-like gait characteristics are established. This information is
an essential first step toward improving our ability to identify abnormalities,
understand the cause(s) of such abnormalities, and, ultimately, design more effective
rehabilitation strategies for children with gait-related disabilities.
Current literature suggests that during gait, 7-year-old children produce
kinematic patterns that are indistinguishable from adults [37]. However, the
underlying kinetic patterns may differ [8]. To date, most kinetic analyses have
focused on the examination of net joint moments and power during stance.
Considering that children often acquire the ability to coordinate non-muscular and
muscular torques gradually with experience, a more detailed analysis of
intersegmental dynamics during swing is necessary in order to determine if and how
gait kinetics of 7 year-olds differ from those of adults.
Before accurate age-related kinetic comparisons can be made, a
methodological concern must be addressed. Cupp et al. postulated that the kinetic
1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
differences identified in their study might actually be artificial, in that they could
have resulted from the use of cadaver-based anthropometric data that do not
accurately represent the body segment parameters of children [8]. While there is
support for this hypothesis [17], it is yet to be tested because directly measured, age-
specific anthropometric data have not been available.
The overall objective of this dissertation was to determine whether gait
patterns in 7 year-old children are similar to those in mature adults. The focus was on
7 year-old children as existing literature suggests that kinetics are still emerging at
this age even though adult-like kinematics are well established. This dissertation
involved 4 experiments, each of which is described in a separate chapter. Chapter III
describes a method of determining lower extremity anthropometric measurements of
segment mass, center of mass location, and moment of inertia using DXA. Chapter
IV addresses the question of whether DXA-derived anthropometric values in
children differ from widely used cadaveric data and describes the extent to which
such values influence kinetic calculations during gait analysis. Chapter V compares
stance phase kinetics (net joint moments; joint power) between 7 year-old children
and adults using DXA-derived, age-specific anthropometric data. Finally, Chapter VI
compares the relative contributions of net joint moment components (i.e. gravity,
muscle, and interaction) in 7-year-old children and adults during the swing phase of
gait.
The results of these studies will enhance our ability to characterize “normal
gait” in children. Such information is critical for the identification and remediation of
gait disabilities in children. Furthermore, this work has major implications for
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
children with disabilities, who often are not privileged with abundant opportunities
to explore their environments through movement.
SPECIFIC AIMS
The principle aims of this study were to:
1. Develop a method of determining age-specific anthropometric measures in-vivo
using DXA (Chapter III).
2. Compare data obtained from 7-year-old children to widely used cadaveric data
(Chapter IV).
3. Compare lower extremity kinematics, net joint moments, and joint power
between 7-year-old children and adults during the stance phase of gait using age
specific anthropometric data for kinetic calculations (Chapter V).
4. Compare the relative contributions of the net joint moment components (i.e.
gravity, muscle, interaction) between 7-year-olds and adults during the swing
phase of gait using age-specific anthropometric data for kinetic calculations
(Chapter VI).
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER II
LITERATURE REVIEW
DEVELOPMENT OF GAIT
The age at which an adult-like pattern of gait is typically observed has not
been definitively established. Even though most children develop the ability to walk
independently within the first 1 to 1 1/2 years of life, many features of gait continue
to mature over the course of several years. Sutherland and colleagues have shown
that kinematic parameters, such as angular rotations and ratios of single- and double
limb support are mature by 3 1/2 to 4 years of age [37]. Similarly, walking velocity,
cadence, and step length, when scaled to leg length, are comparable to adults by
about 4 years of age [39],
Although kinetic analyses have become the "gold standard" in gait analyses
of adults [40], they have been employed in only a few gait studies involving healthy
children and have been limited to analysis of net joint moments and joint power
[8,25]. Furthermore, the results of these studies are inconclusive. In a study designed
to establish a single, normative kinetic database, Ounpuu et al. examined net joint
moments of the hip, knee, and ankle during gait in 5-16 year-old children (n=31).
The authors concluded that kinetic patterns were adult-like by 5 years of age. This
conclusion, however, was based on qualitative comparisons of averaged data from 5-
16 year-old children with previously reported adult data [19]. No between-age
comparisons were reported [25].
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In contrast, recent studies involving more detailed comparisons of children
and adults have revealed significant age-related kinetic differences that become
progressively more adult-like with age [8]. Cupp et al. compared the kinetics of 4-5
(n=6), 6-7 (n=7), and 8-10 (n=10) year-old children to a control group of 18-21 year-
olds 01=5). Statistical analyses were performed using a time-series, point-by-point
paired t-test to identify differences in moment and power curves at various regions of
the gait cycle. Despite similar walking velocities and normalization of moments to
body weight, all groups of children < 7-years demonstrated distinct differences
compared to adults. For example, amplitudes of knee flexion and ankle
plantarflexion moments during stance were diminished, and knee flexion moments
during pre-swing were delayed. Furthermore, at the knee, power generation, rather
than absorption, was observed as the knee was flexing in preparation for swing. This
may suggest that, in contrast to adults, children under the age of 7 years
accomplished initial knee flexion with concentric activity of knee flexor
musculature. In fact, McFadyen et al. observed such a pattern in 7-9 year old
children [24]. Although their study was designed primarily to explore the control of
obstacle avoidance, normal walking trials were used as a basis for comparison. It was
noted that during these control trials, children often accomplished initial knee flexion
during pre-swing with active use of the hamstrings, and the usual antagonistic
activity of the knee extensors was decreased.
Together, these results suggest that while gait kinematics mature relatively
early in childhood, the kinetics producing them are not completely adult-like until
sometime after 7 years of age. McFadyen and colleagues reached a similar
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
conclusion regarding obstacle avoidance in 7-9 year-old children compared to adults
[24], They, too, found that while kinematic patterns were consistently adult-like,
kinetic patterns appeared less mature. These authors postulated that the kinetic
patterns observed in children represent variable strategies of coordinating the
movement of linked segments. The kinetics associated with the movement of linked
segments can only be examined with a level of analysis beyond that of net joint
moments and joint power.
INTERSEGMENTAL DYNAMICS
The term, intersegmental dynamics, refers to the mechanical influence of one
linked segment on another. For gait analysis, the lower extremity is typically
modeled as 3 linked segments, the thigh, shank, and foot. Displacement of any one
of these segments results in rotational forces that act on the other two. These forces
are termed interaction moments and in addition to gravity and muscle moments,
comprise net joint moments. Using Newtonian equations and the inverse dynamics
approach, the relative contributions of interaction, muscle and gravity moments to
any lower extremity net joint moment can be calculated.
A classic illustration of the information that can be gleaned from the study of
intersegmental dynamics involved the description of the roles of gravity, muscle, and
interaction moments in controlling the paw-shake response of a cat [14]. The paw-
shake response was characterized by rapid oscillations at the hip, knee, and ankle, as
well as a distinctive pattern of muscle activation. By decomposing the net moments,
it was shown that at the ankle, muscle moments dominated and functioned primarily
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to generate paw acceleration. However, m uscle moments at the hip and knee
accounted for the forces caused by acceleration of the distal segments. For example,
at the knee, muscle moments controlled limb intersegmental dynamics by
counterbalancing the large interaction moments generated by angular acceleration of
the paw. Therefore, rather than contributing directly to angular rotations at the knee,
the muscle moments functioned to slow and reverse joint motions caused by
interaction moments.
The Role of Intersegmental Dynamics in Multi-Segment Limb Movements
A movement involving multiple limb segments or joints can be produced by
a variety of motor patterns. This is because there are many degrees-of-freedom, or
number of dimensions that can independently vary, within the motor system. It has
been postulated that a desirable motor pattern is determined based on the efficiency
with which it accomplishes the task [29]. Bernstein proposed that movement of one
segment might be accomplished with greater efficiency through the mechanical
interaction of a moving segment that is linked to it, or intersegmental dynamics, than
through direct muscle activity [3].
Schneider et al. demonstrated that refined organization of intersegmental
dynamics characterized improvements in performance of a complex reaching
movement [32], In this study, subjects reached around an obstacle to a vertically
positioned target and returned to the start position with instructions to move rapidly.
Performance improvements, as determined by reductions in movement time and
smoother trajectories, corresponded with differences in intersegmental dynamics at
7
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the elbow, such that there was a greater contribution of interaction moments to the
net moment, with muscle moments functioning to counterbalance and keep the
interaction moments 'in-check' rather than to generate joint rotations. These results
are consistent with Bernstein's hypothesis that exploitation rather than elimination of
interaction torques characterize efficient use of intersegmental dynamics [3].
Intersegmental Dynamics in Development
There are few studies that describe the role of intersegmental dynamics in
limb movements of children, but taken together underscores the necessity of practice,
or experience, and implies that an adult-like pattern of intersegmental dynamics
develops gradually in childhood. Schneider and colleagues described the
intersegmental dynamics associated with spontaneous kicking in 3-month-old infants
[33]. These kicks were highly variable in terms of speed, intensity, and amplitude.
The reversals from hip flexion to extension during relatively non-vigorous kicks
were primarily the product of gravity, and muscle and interaction torques produced
small flexion moments. Since more vigorous kicks involved greater range of motion,
gravity and large interaction torques created a hip flexion moment. Therefore, the
reversals of these kicks were attributed to moments produced by the hip extensors.
This study established that even in infancy, sensori-motor experiences occur that
may shape internal models of limb dynamics.
Other studies, primarily involving reaching activities, have addressed the
effects of experience on the re-organization of muscle and interaction moments for
the production of effective and efficient movements. With age, appropriately timed
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
interaction moments that contribute to the effectiveness of reaching are exploited
while those that are extraneous and not productive are dampened [38]. Building on
these results, Konczak et al. analyzed the intersegmental dynamics of the shoulder
and elbow joints in spontaneous and target-directed reaches of infants [23]. They
determined that improvements in reaching performance were not merely
accomplished by regulating force amplitude, but by learning to modulate the correct
timing of muscular and interaction torques. These authors suggested that the
performance variability observed in this longitudinal study represented an
"unsupervised learning" process whereby a child gains experience coordinating
internal and external forces, thereby acquiring information about limb dynamics and
improving their performance.
The hypothesis that children require experience, or practice, to develop
internal models and refine motor programs is consistent with the schema theory of
motor learning proposed by Schmidt [30], This theory predicts that limb movements
are initiated in accordance with a motor program that represents a "rule" associating
the task demands with the necessary output but that a certain amount of experience is
essential to establish a robust rule.
The Role of Intersegmental Dynamics in Adult Gait
Intersegmental dynamics have a particularly important role during the swing
phase of gait, when ground reaction forces are absent and interaction torques make a
relatively high contribution to the net joint moment. Simulation and
electromyographic studies demonstrate that swing phase involves precise
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
coordination of non-muscular and muscular torques generated by linked lower
extremity segments [27,28]. For example, Phillips et al. quantified the non-muscular
reactions between two adjacent segments using a model that simulated thigh and
shank motions during running [27]. In this study, the linear kinematics of the hip
joint and the muscle moment acting on the thigh were specified throughout the
motion. The only other inputs were anthropometric constants and the initial
configurations of the thigh and shank. The muscular forces and moment acting at the
knee joint were considered to be zero, and the simulated motion was compared to
observed running kinematics. Three different aspects of the gait cycle were
examined: 1) initial hip angular rotation; 2) rapid hip angular rotation; 3) maximum
negative hip angular acceleration. The results suggested that proximal thigh motion
significantly influenced distal shank motion through intersegmental reactions.
Specifically, thigh angular rotations resulted in initial knee flexion for swing and
knee extension in mid- and late swing. The simulated motion was markedly different
from real motion, however. For example, without precisely timed and scaled knee
muscle moments, thigh rotation resulted in excessive knee flexion in early swing,
delayed and incomplete knee extension in preparation for terminal swing, and knee
hyperextension in terminal swing.
Dynamic electromyography confirms that knee angular rotation during swing
depends largely on the mechanical interaction of the lower limb segments but that
precisely timed and scaled muscle activity is necessary for the production of
consistent kinematic patterns [26]. A substantial proportion (about 66%) of total
knee flexion necessary for swing is accomplished without the influence of muscles
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
acting directly at the knee. This occurs as the center of pressure is transferred to the
metatarsophalangeal joints and the hip is rapidly flexing. Frequently, rectus femoris
activity is necessary at the onset of initial swing to counteract these "passive" forces
and prevent excessive knee flexion. The short head of the biceps femoris becomes
active during initial swing and produces a small amount of knee flexion to ensure
foot clearance. Similarly, no direct muscle action is needed at the knee during mid
swing as knee extension is accomplished by inertial forces acting on the tibia
secondary to continued hip angular rotation. During terminal swing, the quadriceps
are active to produce knee extension, but as hip flexion motion is reversed, excessive
knee extension is prevented by eccentric activity of the hamstrings.
To date, there is only indirect evidence that children differ from adults in
terms of their use of interaction torques during gait. For example, where initial knee
flexion in the adult is accomplished primarily through forward thigh advancement
with little to no muscle activity at the knee, McFadyen et al. found that during
normal walking trials, some 7-9 year-old children accomplished initial knee flexion
with active use of the hamstrings [24]. Similarly, Cupp et al. found that children < 7
years demonstrated power generation rather than absorption at the knee during initial
swing. This, too, suggests that concentric muscle activity at the knee was necessary
to produce knee flexion [8].
Further kinetic analysis, on the level o f intersegmental dynamics, m ay be
helpful in answering the question of whether the gait patterns of 7 year-olds are
similar to adults. However, prior to extending kinetic analysis to studies of children,
a key methodological issue must first be addressed. The anthropometric data [10]
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
used in the inverse dynamics equations to calculate kinetics in each of the
experiments described previously may not be representative of the body segment
parameters of children. While the potential limitations associated with the use of
these data are recognized [8], no experiments have been conducted to determine the
extent to they influence kinetic calculations during gait, primarily because age-
specific body segment parameters are not readily attainable. In order to accurately
compare the kinetics of gait in children and adults, the issue of the use of
inappropriate anthropometric data must be addressed.
ANTHROPOMETRICS
In biomechanics research, inverse dynamics equations are routinely used to
calculate net joint moments. The accuracy of these equations is dependent upon the
input data, namely kinematic trajectories, external forces, and anthropometric data
(e.g. segment mass, center of mass location, and moment of inertia). While precise
measures of kinematics and ground reaction forces can be obtained in most motion
analysis laboratories, anthropometric data are difficult to measure in-vivo. Instead,
they are typically estimated from tables or regression equations derived from
cadaveric studies involving older adults [5,7,10,40]. For example, anthropometric
data compiled by Dempster, which is commonly used in kinetic analyses involving
subjects o f all ages, was collected from 8 male cadavers with a mean age and mass o f
68.5 years and 60 Kg, respectively [10].
Whether or not the segment proportions determined from cadavers are
appropriate for use in kinetic studies involving children is a concern [8,16,17].
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Several investigators have recognized this limitation, and, using mathematical
models, have attempted to develop regression equations that adjust cadaveric
anthropometric data based on age [13,17,18,31]. For example, Jensen reported on a
9-year longitudinal study involving 8 males between the ages of 4 and 20 years [17].
Using segment dimensions and assumed segment densities, mass, center of mass
locations, and principal moments of inertia were calculated for 16 body segments.
Accuracy was determined by comparing the calculated total body mass with actual
total body mass. The results suggest that re-distributions of mass between segments
occur with age. Regression equations based on age were formulated.
Jensen utilized these equations to determine if the use of age-specific
anthropometric data would impact the results and interpretation of kinetic
calculations by examining the kinetics associated with a jumping activity in a 6-year-
old child [17]. Net joint moments calculated from data derived from cadavers were
compared to those calculated from data that had been adjusted for age using
regression equations. Resulting moment curves were qualitatively similar, but
distinct with regard to amplitude. These results support the notion that
anthropometric data derived from adult cadavers do not accurately represent the
body proportions of children and that use of these data for kinetic calculations, at
least those involving high velocity movements, could produce misleading results.
One limitation o f age-based regression equations is that the m odels used to
determine them do not account for possible variations in tissue density within or
between segments. This could result in inaccurate determinations of segment mass
and/or mass distributions. It is possible that these errors could balance each other
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and, therefore, not be reflected in accuracy measurements, such as comparisons of
calculated and measured total body mass. It follows that age-specific anthropometric
data determined directly from in-vivo measurements are necessary for accurate
kinetic calculations that involve children. These data, however, are currently not
available in the literature.
One tool that has the apparent capability of quantifying anthropometric data
in-vivo is dual energy x-ray absorptiometry (DXA). DXA is commonly used to
assess bone mineral density or body composition for research and diagnostic
purposes, and is capable of accurately measuring mass of the entire body or of a
defined region of interest [22,34], Therefore, the mass of thigh, shank, and foot
segments, as well as the mass of slices within these segments, can be measured
directly. Center of mass location and principle moments of inertia can then be
calculated using equations described by Winter [40],
With age-specific anthropometric data, comparisons can be made to
determine if they differ from cadaver-based estimates. Also, use of such data for
kinetic calculations will answer the question of whether or not previously identified
kinetic differences between children < 7 years and adults were representative of true
differences or if they merely reflected the use of inappropriate anthropometric data.
Finally, the net joint moments can be inspected in detail to determine if the use of
intersegmental dynamics differs between 7-year-old children and adults during the
swing phase of gait.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SUMMARY
The age at which gait matures is unclear. While kinematic patterns are similar
to those of adults are apparent by 3 1/2 to 4 years of age, kinetic differences have
been reported in children 7 years and younger. Previous investigators have
postulated that these differences are merely the product of the use of inappropriate
anthropometric estimates for kinetic calculations. While there is evidence to support
this hypothesis, it has not been directly tested. Alternatively, the reported kinetic
differences might be reflective of gait patterns that are not fully mature in 7 year-old
children. In order to sufficiently confirm or refute this, examination of the
intersegmental dynamics associated with swing should be performed in addition to
the more routine examination of net joint moments and power during stance.
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER III
DETERMINATION OF LOWER EXTREMITY ANTHROPOMETRIC
PARAMETERS USING DUAL ENERGY X-RAY ABSORPTIOMETRY
One limitation of previous work involving kinetic analysis of gait in children
has been the use of anthropometric data collected from cadavers of aged adults in
which the anthropometric values may not be representative of children. Since no
anthropometric data have been reported for the age group of children involved in the
present study, a practical method for determining age-specific anthropometric data
was needed. This chapter will describe a method of determining lower extremity
anthropometric data using dual energy x-ray absorptiometry (DXA). Additionally,
values obtained with this method were compared to predictions from cadaver-based
estimates, and the extent to which these data affect the calculation of net joint
moments during walking in adults were quantified.
INTRODUCTION
Inverse dynamics equations are routinely used to calculate intersegmental
forces and net joint moments during gait. A potential source of error in such
calculations can be the estimation of segment anthropometric parameters (e.g. mass,
center of mass location, and moment of inertia about the center of mass) [2,21].
While precise measures of kinematics and ground reaction forces can be obtained in
most motion analysis laboratories, anthropometric data are typically estimated from
cadaveric studies involving limited samples of older adults [5,7,10].
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Imaging studies using magnetic resonance imaging, computer aided
tomography, and gamma-ray scanning have demonstrated that cadaveric data may
not accurately represent certain populations, such as children or college-age athletes
[6,15,42], However, because of inaccessibility, high radiation doses, and/or financial
expense, it is generally not practical for investigators to determine population-
specific anthropometric values. To account for variability in subject age, size,
gender, and body dimensions, several mathematical methods have been developed to
modify cadaver-based anthropometric estimates [13,17,18,31], These models,
however, rely on assumed segment densities and geometries and, therefore, may
misrepresent segment mass and mass distribution.
Dual energy x-ray absorptiometry (DXA) is commonly used to assess bone
mineral density or body composition for research and diagnostic purposes.
Compared to other imaging methods, it is relatively inexpensive, and the associated
x-ray dose is substantially less than normal background sources [20]. Since DXA is
capable of accurately measuring mass of the entire body or of a defined region of
interest [22,34], segment mass (SM), as well as the mass of slices within these
segments, can be measured directly. With this information, center of mass location
(COM) and moment of inertia about the COM ( I c o m ) can be calculated [11,40].
Given the capability of DXA to quantify anthropometric measures in-vivo,
the purpose o f this study was threefold: 1) develop a method for obtaining SM,
COM, and I c o m for the thigh, leg, and foot, 2) compare values obtained with this
method to predictions based on widely used cadaver-based estimates [10], and 3)
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
quantify the extent to which these two sources of anthropometric values affect the
calculation of net joint moments during walking in healthy adults.
METHODS
Subjects
Twenty healthy adults (10 male, 10 female) between the ages of 23 and 50
years participated in this study. The average age, height, and mass of this group was
31+9 years, 171 + 10 cm, and 74.7 ± 15.2 kg, respectively (Table 3-1). Subjects
were screened to rule out the presence of metal implants and the use of medication
known to affect bone or muscle tissue. A sub-set of ten subjects (5 male, 5 female)
participated in the gait analysis portion of this study. None of these individuals had
any neurologic or orthopedic disorder that might influence gait. The average age,
height, and mass of this subgroup was 32 ± 8 years, 168 ± 10 cm, and 70.0 ± 15.1 kg,
respectively (Table 3-1). Prior to participation, all procedures were explained to each
subject and informed consent was obtained, as approved by the Institutional Review
Board of the University of Southern California, Health Sciences Campus.
Table 3-1
Physical Characteristics Of Study Participants
DXA Scanning
(10 females, 10 males)
Gait Analysis
(5 females, 5 males)
Mean SD Range Mean SD Range
Age (yrs) 31.0 8.6 23-50 32.2 7.5 25-48
Height (cm) 171.3 10.0 154.9-188.0 167.6 9.8 154.9-180.3
Mass (kg) 74.7 15.2 45.1-106.6 70.0 15.1 45.1-89.7
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Dual Energy X-Ray Absorptiometry
Full-body DXA scans were performed within the Clinical Exercise Research
Center at the University of Southern California using a Hologic QDR-1500
densiometer and analyzed with version 7.2 software (Hologic, Waltham, MA). The
densiometer generates a narrow x-ray beam at two alternating frequencies (140 kVp;
70 kVp), and, based on differential attenuation, proportions of bone, fat, and lean
tissue are identified on a pixel-by-pixel basis (pixel size = 0.27 cm ). Mass is then
quantified using assumed densities for bone (2.5-3.0 g/cc), fat (0.9 g/cc), and lean
(1.08 g/cc) tissue. Accuracy of the densiometer was assessed through use of a
phantom with calibration error being less than 1.5% as specified by the
manufacturer.
Prior to scanning, participants were asked to remove shoes, jewelry, and any
clothing with metal. Scans were performed with subjects lying supine with the lower
extremities in slight external rotation and maximum plantarflexion. Total scan time
was approximately 10 minutes for each subject.
For each DXA scan, frontal plane boundaries for each lower extremity
segment (thigh, shank, and foot), as defined by Dempster [10], were manually
identified (Table 3-2). Total body mass, as well as the length and mass of each
segment and the mass of sequential 3.9 cm horizontal sections within these segments
were recorded (Figure 3-1). Pilot studies revealed that SM, COM, and I c o m values
calculated using 3.9 cm sections were very similar to those calculated using 1.3 cm
sections (coefficients of variation ranging from 0-0.03 across variables, with a mean
of 0.008).
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m.
com
m ,
Figure 3-1. Method to calculate the distance from the proximal segment boundary to
the center of mass (COM) of the shank (X) based on mass distribution, where m, is
the mass of the zth section and x, is the distance from the proximal segment boundary
to the center of the z'th section (height of each section = 3.9 cm).
With respect to the calculation of SM, COM, and I c o m the following
assumptions were made:
• Each segment is symmetrical about its anteroposterior and mediolateral
axes.
• The center of mass of each 3.9 cm horizontal section is located as a point
mass in the geometric center of that section.
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
• In computing the mass moment of inertia, a segment can be modeled as a
set of point masses spaced along the longitudinal axis.
• The mass moment of inertia is negligible about the longitudinal axis.
Table 3-2
Boundaries Used To Identify Lower Extremity Segments From DXA Scans
Segment Proximal Boundary Distal Boundary
Thigh Superior aspect greater
trochanter*
Lateral epicondyle
Shank Lateral epicondyle Inferior aspect lateral
malleolus
Foot Inferior aspect lateral
malleolus
5th metatarsal head
*Mass of the pelvis not included
Center o f Mass Location
COM from the proximal segment boundary was calculated using Equation 1:
COM = (Em,x,)/SM (1)
where m; is the mass of the zth section and x, is the distance from the proximal
segment boundary to the center of the zth section (Figure 3-1).
Moment o f Inertia
The mass moment of inertia about the proximal segment boundary (I pr0x) was
calculated using Equation 2:
Ip rox — YjniXi (2)
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The parallel axis theorem was then used to calculate I c o m using Equation 3:
I c o m = Iprox - S M * C O M 2 ( 3 )
To allow for comparisons with cadaver-based estimates, the radius of
gyration about the center of mass (Rg ) was determined for each segment using
Equation 4.
Rg=[(IcoM/SM)1/2] (4)
For the calculation of the net joint moments, anthropometric data obtained
from DXA (SM, COM, and Rg ) were converted to proportions of total body mass or
segment length, and then averaged across subjects (Table 3-3).
Table 3-3
DXA-Derived And Cadaver-Based Anthropometric Proportions
Segi
Total
nent M
Body
lass/
Mass
Center of Mass/
Segment Length
(proximal boundary)
Radius of Gyration/
Segment Length
Thigh Shank Foot Thigh Shank Foot Thigh Shank Foot
DXA (n=20) 0.113 0.045 0.010 0.461 0.412 0.497 0.251 0.266 0.254
Cadaver*
(Dempster, 1955) 0.100 0.047 0.015 0.433 0.433 0.500 0.323 0.302 0.475
* Adapted from Winter [40]
Reliability o f DXA Measurements
Reproducibility of DXA-derived anthropometric values was assessed by
repeating the analysis of 5 scans on two separate days. A coefficient of variation was
calculated for 5 subjects then averaged across subjects. The mean coefficient of
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
variation for all parameters of all segments was 0.020 ± 0.017 (range = 0.004 to
0.06), indicating excellent repeatability.
Cadaver-Based Estimates
To allow for statistical comparisons with commonly used cadaveric data,
cadaver-based proportions were used to estimate SM, COM, and I c o m based on each
subject’s height and mass (Table 3-4) [40].
Table 3-4
Equations Used To Calculate Subject-Specific Anthropometric Values From
Cadaver-Based Data
Segment
Mass
Center of Mass Location
(proximal boundary)
Moment of Inertia
(about center of mass)
Thigh 0.100*
Body Mass
0.433 * (0.245 * Body Ht) (0.323 * 0.245 *BodyHt)2
* SM
Shank 0.0465 *
Body Mass
0.433 *(0.246 * Body Ht) (0.302* 0.246* Body Ht)2
* SM
Foot 0.0145 *
Body Mass
0.50 *(0.152 * Body Ht) (0.475*0.152*BodyHt)2
* SM
Adapted from Winter [40]
Gait Analysis
The gait portion of this study was performed in the Musculoskeletal
Biomechanics Research Laboratory at the University of Southern California and
occurred within 2 weeks of the DXA scans. Three-dimensional lower extremity
kinematics were recorded at 60 Hz using a six-camera Vicon motion analysis system
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Oxford Metrics Ltd., Oxford, England). Ground reaction forces were recorded at a
rate of 600 Hz using 3 AMTI force plates (OR6-5-1000; Advanced Mechanical
Technology, Inc., Newton, MA) camouflaged in the middle of a 10-meter walkway.
Reflective markers (20 mm spheres) were placed at anatomical landmarks of
each lower extremity (posterior superior iliac spine, anterior superior iliac spine,
lateral thigh, femoral epicondyle, lateral tibia, lateral malleolus, calcaneous, 5th
metatarsal head, and the dorsum of the foot). After 2-3 practice trials, kinematics and
ground reaction forces were recorded simultaneously as subjects walked barefoot
along the walkway at a self-selected speed. Photoelectric sensors positioned at either
end of the walkway triggered the start and end of data collection. Three trials per
subject were recorded. Trials in which the entire foot did not contact a force plate or
gait appeared unnatural (e.g. force plate targeting or exaggerated step lengths) were
excluded.
One representative stride per subject was selected for analysis. Reflective
markers were manually labeled and digitized in order to locate embedded coordinate
systems for the pelvis, thigh, shank, and foot (Vicon Clinical Manager, software
version 3.5; Oxford Metrics Ltd., Oxford, England). Trajectories were filtered using
a Woltring filtering process [41].
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Inverse dynamics equations were employed to calculate net joint moments at
the ankle, knee, and hip using a) DXA-derived, and b) cadaver-based anthropometric
proportions (Table 3-4) [4], For the purposes of this study, only moments in the
sagittal plane were reported. Moments were normalized to body weight and
expressed as a percentage of the gait cycle.
Statistical Analysis
To determine if DXA-derived measurements differed from cadaver-based
estimates, a 2 x 3 (measurement method x segment) analysis of variance (ANOVA)
with repeated measures was performed. This analysis was repeated for SM, COM,
and I c o m - To compare differences in moment curves obtained using DXA-derived
anthropometric values and cadaver-based estimates, a root mean square error
(RMSE) score was calculated for the stance and swing phases at the hip, knee, and
ankle for each subject. Comparisons involved a 2 x 3 (gait phase x joint) ANOVA
with repeated measures. Post hoc testing was performed using t-tests. A P-value of <
0.05 was used to determine statistical significance. All statistical tests were
performed using SPSS software (Chicago, II).
RESULTS
Dual Energy X-Ray Absorptiometry And Cadaver-Based Estimates
For DXA-derived and cadaver-based SM values, a significant method x
segment interaction (F2, i8= 28.76; P < 0.05) was identified. Post-hoc testing revealed
that, compared to cadaver-based estimates, DXA-derived measurements were greater
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
at the thigh (8.4 ± 1.7 kg vs. 7.5 ± 1.5 kg) and less at the foot (0.7 ± 0.2 kg vs. 1.1 ±
0.2 kg) (Figure 3-2).
■ DXA
Cadaver-based (Dempster, 1955)
Thigh Shank Foot
Figure 3-2. Comparison of segment mass for the thigh, shank, and foot as
determined by dual energy x-ray absorptiometry (DXA) and cadaver-based
estimates. * indicates statistically significant differences (P < 0.05).
For DXA-derived and cadaver-based COM values, a significant method x
segment interaction (F2,i8 = 53.04; P < 0.05) was identified. Post-hoc testing
revealed that, compared to cadaver-based estimates, DXA-derived measurements of
the distance from the proximal segment boundary to the COM were greater at the
thigh (0.20 ± 0.02 m vs. 0.18 ± 0.01 m), but less at the shank (0.17 ± 0.01 m vs. 0.18
± 0.01 m) and the foot (0.11 ± 0.01 m vs. 0.13 ± 0.01 m) (Figure 3-3).
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DXA
Cadaver-based (Dempster, 1955)
Shank
Figure 3-3. Comparison of the distance from the proximal segment boundary to the
center of mass for the thigh, shank, and foot as determined by dual energy x-ray
absorptiometry (DXA) and cadaver-based estimates. * indicates statistically
significant differences (P < 0 .0 5 ).
For DXA-derived and cadaver-based Icom values, a significant method effect
(Fi,i9 = 1 4 3 .9 7 ;P < 0 .0 5 ) and a significant method x segment interaction (F2 ,i8 =
5 8 .5 6 ; P < 0 .0 5 ) were identified. Post-hoc testing revealed that DXA-derived Icom
values were significantly less than cadaver-based estimates at the thigh (0 .0 9 8 ±
0 .0 2 9 kg-m2 vs. 0 .1 4 4 ± 0.041 kg-m2 ), shank (0 .0 4 2 ± 0 .0 1 2 kg-m2 vs. 0 .0 5 7 ± 0 .0 1 8
kg-m2 ), and foot (0 .0 0 9 ± 0 .0 0 6 kg-m2 vs. 0 .0 1 7 ± 0 .0 0 5 kg-m2 ) (Figure 3 -4 ).
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
■ DXA
M Cadaver-based (Dempster, 1955)
Thigh Shank Foot
Figure 3-4. Comparison of the moment of inertia about the center of mass for the
thigh, shank, and foot as determined by dual energy x-ray absorptiometry (DXA) and
cadaver-based estimates. * indicates statistically significant differences (P < 0.05).
Gait Analysis
In general, the net joint moment curves calculated with DXA-derived
anthropometric measurements were similar to those calculated with cadaver-based
estimates at the ankle, knee, and hip (Figures 3-5, 3-6, 3-7, respectively). For the
average RMSE scores during the stance and swing phases, a significant phase effect
(Fi;9 = 88.64; P < 0.05), a significant joint effect (F2,8 = 65.30; P < 0.05), and a
significant phase x joint interaction (F2,b = 49.01; P < 0.05) were identified. Post-hoc
testing revealed that the RMSE values during swing were greater than those during
stance at the hip (0.100 ± 0.031 N-m/kg body weight vs. 0.030 ± 0.008 N-m/kg body
weight), knee (0.053 ± 0.014 N-m/kg body weight vs. 0.012 ± 0.003 N-m/kg body
weight), and ankle (0.009 ± 0.002 N-m/kg body weight vs. 0.005 ± 0.001 N-m/kg
body weight). Additionally, when collapsed across phases RMSE scores were greater
at the hip (0.065 ± 0.043 N-m/kg body weight) than at the knee (0.033 ± 0.023
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
N-m/kg body weight) and ankle (0.007 ± 0.003 N-m/kg body weight). Similarly,
RMSE scores were significantly greater at the knee than at the ankle (Figure 3-8).
£
D £
s
t ,
d
o
E
o
1.5 i
1.0 -
0.5
0.0
-0.5
Plantarflexor
------- DXA
/ \
- - - Cadaver-based
/ \
(Dempster, 1955)
Dorsiflexor
i i i i i i i i r
10 20 30 40 50 60 70
% Gait Cycle
80 90 100
Figure 3-5. Net ankle joint moments from a representative subject during gait,
calculated with anthropometric data derived from dual energy x-ray absorptiometry
(DXA) and cadaver-based estimates. Vertical line distinguishes stance
(approximately 0-60% of the gait cycle) from swing phase (approximately 61-100%
of the gait cycle).
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.0
DXA
Extensor
£
.a
w o 0.5
£ 6
|
o.o
■ " Cadaver-based
(Dempster, 1955)
-0.5
Flexor
- 1.0
10 20 30 40 50 60 70 80 90 100 0
% Gait Cycle
Figure 3-6. Net knee joint moments from a representative subject during gait,
calculated with anthropometric data derived from dual energy x-ray absorptiometry
(DXA) and cadaver-based estimates. Vertical line distinguishes stance
(approximately 0-60% of the gait cycle) from swing phase (approximately 61-100%
of the gait cycle).
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DXA
1.0
Extensor
- - Cadaver-based
(Dempster, 1955)
a
< D
o -0.5
s
Flexor
- 1.0
0 20 40 60 80 100
% Gait Cycle
Figure 3-7. Net hip joint moments from a representative subject during gait,
calculated with anthropometric data derived from dual energy x-ray absorptiometry
(DXA) and cadaver-based estimates. Vertical line distinguishes stance
(approximately 0-60% of the gait cycle) from swing phase (approximately 61-100%
of the gait cycle).
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-G
•S f
'3
Kfl
0.15 n
0.10 -
-G
o
p q
o d 0.05
0.00
f t
Hip
* *
I Swing
I Stance
*
T
Knee
•kkk
___
Ankle
Figure 3-8. Comparison of average RMSE (n=10) scores of the net hip, knee, and
ankle joint moments during stance and swing. Significant differences (P < 0.05)
were as follows: *RMSE during swing is significantly greater than that during
stance; **RMSE hip > RMSE knee and ankle; ***RMSE knee > RMSE ankle.
DISCUSSION
This paper describes a method by which DXA can be used to quantify lower
extremity anthropometries. Although DXA-derived values of SM, COM, and I c o m
differed statistically from cadaver-based estimates, the magnitudes of these
differences were quite small. Furthermore, for the population examined, these
differences only affected the calculation of net joint moments during the swing phase
of gait.
Given that the present study involved younger adults, the differences between
DXA-derived and cadaver-based estimates of SM, COM, and I c o m may be due to
age-related differences in mass and mass distribution. In support of this, DXA-
derived anthropometric measurements were found to be comparable to those
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
reported from recent studies which also involved younger subjects. When averaged
across subjects, DXA-derived values of segment mass were 11.3%, 4.5%, and 1.0%
of total body mass for the thigh, shank, and foot, respectively. From subjects with a
mean age of 26 years, Cheng et al. reported values of 13.6%, 4.4%, and 2.0% (MRI)
[6], and from subjects with a mean age of 23.8 years, Zatsiorsky and Seluyanov
reported values of 14.2%, 4.3%, and 1.4% (gamma-ray scanning) for the thigh,
shank, and foot, respectively [42]. The higher thigh mass value reported by
Zatsiorsky and Seluyanov can likely be attributed to a more proximal segmentation
of the thigh from the trunk; specifically, the superior border of the thigh segment was
near the femoral head versus the greater trochanter.
With regards to COM distance from the proximal segment boundary, the
present study reported values of 46.1%, 41.2%, and 49.7% of segment length for the
thigh, shank, and foot, respectively. Similarly, using imaging techniques, Zatsiorsky
and Seluyanov reported proportions of 45.5%, 40.5%, and 55.9%, while Cheng et al.
reported proportions of 44.7%, 44.2%, and 54.0% [6,42].
DXA-derived measures of segment mass distribution can be compared to the
existing literature by examining Rg values. DXA-derived Rg values for the thigh,
shank, and foot (25.1%, 26.6%, and 25.4% segment length, respectively) closely
approximate those reported by De Leva about the mediolateral axis (26.7%, 27.5%,
and 24.5%) [9],
This study also demonstrated the extent to which differences between DXA-
derived and cadaver-based anthropometric values influence the calculation of net
joint moments during gait. As indicated by RMSE scores, the resulting moment
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
curves were similar overall, with essentially no differences during stance. The
observed differences between stance and swing can be explained by the contributing
components of the net joint moment. During stance, ground reaction forces dominate
the moment calculation, and inertial terms add very little. During swing, however,
inertial terms, which are greatly influenced by anthropometric values, dominate the
moment calculation. Given as such, it is not surprising that RMSE scores were
greater during swing.
RMSE scores during both stance and swing were greater at the hip and knee
compared to the ankle. This finding can be attributed to the relatively large
differences in DXA-derived and cadaver-based SM, COM, and I c o m values at the
thigh. Also, methods for using inverse dynamics equations begin with calculations at
the foot and work proximally; therefore, moment errors from the ankle and knee
would be readily transferred to the hip.
Three potential limitations to the anthropometric portion of the present study
should be considered. First, each segment was assumed to be symmetrical, such that
the moments of inertia and radii of gyration were equal about the anteroposterior and
mediolateral axes. Support for this assumption is found in data reported by De Leva,
where the reported radius of gyration values about the anteroposterior axis for the
thigh, shank, and foot (26.7%, 28.1%, and 25.7% segment length, respectively) were
quite similar to those about the mediolateral axis (26.7%, 27.5%, and 24.5% ) [9].
Next, while it is possible that modeling each segment as a set of point masses spaced
along the longitudinal axis may have underestimated I c o m , any such underestimation
would be quite small and unappreciable once converted to Rg. Lastly, anthropometric
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
values for the foot were determined in the frontal plane, therefore some error
associated with the out-of-plane component of the foot axis was unavoidable.
However, with the ankle positioned in maximal plantarflexion, the assumed foot axis
(ankle to 5th metatarsal head) approached the frontal plane, thus minimizing the
magnitude of the error.
CONCLUSION
Based on the findings of this study, DXA can be used to obtain population-
specific anthropometric data. Although significant differences between DXA and
cadaver-based estimates of SM, COM, and I c o m were observed, these differences
had a small influence on the calculation of sagittal plane moments at the ankle, knee,
or hip.
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER IV
ANTHROPOMETRIC PARAMETERS IN CHILDREN: A COMARISON OF
VALUES OBTAINED FROM DUAL ENERGY X-RAY ABSORPTIOMETRY
AND CAD AYER-BASED ESTIMATES
Anthropometric data used in gait analysis for children are often estimated
from older adult cadavers. The extent to which these estimates accurately represent
children or affect the calculation of net joint moments during gait analysis is
unknown. Using the methods described in Chapter III, this chapter compared DXA-
derived lower extremity anthropometric data to cadaver-based estimates in 7-13
year-old children. Secondly, this chapter determined the extent to which DXA-
derived anthropometric data influenced the calculation of net joint moments during
gait in children.
INTRODUCTION
Using inverse dynamics equations, representative body segment parameters
(mass, center of mass location, and moment of inertia about the center of mass) are
necessary for the accurate calculation of net joint moments. These parameters are
often estimated from data derived from older cadavers [5,7,10]. Cadaver-based
estimates are com m only used for kinetic analyses involving children [8,25], This is a
concern as cadaver-based estimates may not accurately represent the body
proportions of children [1,16,17]. For example, Jensen provided evidence that use of
cadaveric data resulted in substantial error when calculating net joint moments
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
during jumping for a 6 year-old child [17]. The extent to which the use of cadaveric-
based estimates of body segment parameters affects the calculation of net joint
moments during gait has not been reported.
Several mathematical methods have been developed to modify cadaver-based
anthropometric estimates based on a child’s age, size, and body dimensions
[1,13,16,17]. These models use densities determined from cadaveric studies that are
assumed to be constant throughout a segment. Therefore, segment mass and mass
distribution may be misrepresented.
Dual energy x-ray absorptiometry (DXA) has been shown to be a valid
method to obtain subject-specific anthropometric parameters [11]. Using DXA,
segment mass (SM) and mass of slices within each segment can be accurately
measured [22,34]. With this information, center of mass location (COM) and
moment of inertia about the COM ( I c o m ) can be calculated [11,40].
The purpose of this study was to: 1) compare DXA-derived anthropometric
parameters to cadaver-based estimates in 7-13 year-old children, and 2) determine
the extent to which DXA-derived anthropometric data influence the calculation of
net joint moments during gait in children.
METHODS
Subjects
Fifty healthy children (33 male, 17 female), ages 7-8 (n=21), 9-10 (n=16),
and 11-13 (n=13) years participated in the DXA portion of this study (Table 4-1).
Subjects were screened to rule out the presence of metal implants and the use of
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
medication known to affect bone or muscle tissue. In addition, 3 male children (7,
10, and 13 years-old) participated in the gait analysis portion of this study (Table 4-
1). These subjects were screened to rule out any neurologic or orthopedic disorders
that would influence gait. Prior to participation, all procedures were explained to
each subject (and parents) and informed consent was obtained as approved by the
Institutional Review Board of the University of Southern California, Health Sciences
Campus.
Table 4-1
Physical Characteristics Of Study Participants
DXA Analysis Gait Analysis
Age (years) Height
(cm)
Weight
(kg)
Age
(years)
Height
(cm)
Weight
(kg)
Mean SD Mean SD
7-8 (n=21; 9 males,
12 females)
129.0 0.06 33.4 10.5 7 (n=l) 132.1 26.8
9-10 (n=16; 13
males, 3 females)
138.9 0.08 40.2 9.6 10 (n=l) 139.7 33.6
11-13 (n=13; 11
males, 2 females)
154.0 0.09 46.4 9.6 13 (n=l) 160.0 44.1
Dual Energy X-Ray Absorptiometry
Full-body DXA scans were performed within the General Clinical Research
Center on the Health Science Campus of the University of Southern California using
a Hologic QDR-4500W densiometer (Hologic, Waltham, MA). The densiometer
generates an x-ray beam at two alternating frequencies (140 kVp; 70 kVp), and
based on differential attenuation, proportions of bone, fat, and lean tissue are
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
• < « • • • • 7
identified. Mass is then quantified on a pixel-by-pixel basis (pixel size = 0.27 cm )
using assumed densities for bone (2.5-3.0 g/cc), fat (0.9 g/cc), and lean (1.08 g/cc)
tissue. Accuracy was ensured through use of a phantom with calibration error being
less than 1.5% as specified by the manufacturer.
Prior to scanning, participants removed shoes and any clothing or jewelry
with metal. Scans were performed with the participants lying supine with the lower
extremities in slight external rotation and maximum plantarflexion. Total scan time
was approximately 5 minutes for each subject.
For each DXA scan, lower extremity segment boundaries, consistent with
cadaveric studies [10], were identified (Table 3-2). Total body mass, as well as the
length and mass of the thigh, shank, and foot segments and the mass of sequential
3.9-cm horizontal sections within these segments were recorded (Figure 3-1). Pilot
studies revealed that SM, COM, and I c o m values calculated from 3.9-cm sections did
not statistically differ from those calculated from 1.3-cm sections (coefficients of
variation ranging from 0-0.03%).
With respect to the calculation of SM, COM, and I c o m the following
assumptions were made:
• Each segment is symmetrical about its anteroposterior and mediolateral
axes.
• The center of mass of each 3.9 cm horizontal section is located as a point
mass in the geometric center of that section.
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
• In computing the mass moment of inertia, a segment can be modeled as a
set of point masses spaced along the longitudinal axis.
• The mass moment of inertia is negligible about the longitudinal axis.
Center o f Mass Location
COM from the proximal segment boundary was calculated using Equation 1:
COM = (Em;x;)/SM (1)
where m, is the mass of the z'th section and x, is the distance from the proximal
segment boundary to the center of the z'th section (Figure 3-1) [40],
Moment o f Inertia
The mass moment of inertia about the proximal segment boundary (Ipr0x) was
calculated using Equation 2:
Iprox ~ Yjn^i (2)
The parallel axis theorem was then used to calculate I c o m using Equation 3:
I c o m = I p r o x-SM*X2 (3)
where X is the distance from the proximal segment boundary to the COM [40].
For the calculation of the net joint moments, the radius of gyration about the
center of mass (Rg ) was calculated for each segment using Equation 4 [40].
Rg=[(IcoM/SM )1/2] (4)
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In addition, anthropometric data obtained from DXA (SM, COM, and Rg )
were converted to proportions of total body mass or segment length, and group
averages were determined (Table 4-2).
Table 4-2
DXA-Derived And Cadaver-Based Anthropometric Proportions Used For
The Calculation Of Net Joint Moments During Gait
Segment Mass/Body
Mass
COM/Segment
Length (from
proximal segment
boundary)
Rg/Segment Length
(about COM)
DXA-derived Thigh Shank Foot Thigh Shank Foot Thigh Shank Foot
7-8 yrs (n=21) 0 . 1 1 0 0.046 0.0137 0.463 0.415 0.482 0.256 0.274 0.259
9-10 yrs (n=16) 0.114 0.047 0.0149 0.465 0.416 0.488 0.252 0.274 0.259
11-13 yrs (n=13) 0.117 0.048 0.0149 0.468 0.415 0.483 0.256 0.276 0.259
Cadaver-based* 0.100 0.047 0.0145 0.433 0.433 0.500 0.323 0.302 0.470
* Adapted from Winter [40]
Cadaver-Based Estimates
To allow for statistical comparisons with commonly used cadaveric data,
cadaver-based proportions were used to estimate SM, COM, and I c o m based on each
subject’s height and weight (Table 3-4) [10].
Gait Analysis
The gait portion of this study was performed in the Musculoskeletal
Biomechanics Research Laboratory on the Health Science Campus of the University
of Southern California. Three-dimensional lower extremity kinematics were recorded
41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
at 60 Hz using a six-camera Vicon motion analysis system (Oxford Metrics Ltd.,
Oxford, England). Ground reaction forces were recorded at a rate of 1200 Hz using 3
AMTI force plates (OR6-5-1000; Advanced Mechanical Technology, Inc., Newton,
MA) camouflaged within a 10-meter walkway.
Reflective markers (20 mm spheres) were placed at anatomical landmarks of
each lower extremity (posterior superior iliac spine, anterior superior iliac spine,
th
lateral thigh, femoral epicondyle, lateral tibia, lateral malleolus, calcareous, 5
metatarsal head, and the dorsum of the foot). After 2-3 practice trials, kinematics and
ground reaction forces were recorded simultaneously as subjects walked barefoot
along the 10-meter walkway at a self-selected speed. Photoelectric sensors
positioned at either end of the walkway automatically triggered the start and end of
data collection. Three trials per subject were recorded. Only trials in which the entire
foot made contact with the force-plate without targeting were included in the
analysis.
One representative stride per subject was selected for analysis. Reflective
markers were manually labeled then digitized using Vicon Clinical Manager,
software version 3.5 (Oxford Metrics Ltd., Oxford, England). These markers were
used to locate embedded coordinate systems for the pelvis, thigh, shank, and foot.
Trajectories were filtered using a Woltring filtering process [41].
Inverse dynamics equations were used to calculate net joint moments at the
ankle, knee, and hip using a) DXA-derived and b) cadaver-based anthropometric
proportions (Table 4-3) [10]. For the purposes of this study, only moments in the
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sagittal plane were reported. Moments were normalized to body weight and
expressed as a percentage of the gait cycle.
Statistical Analysis
To determine if DXA-derived measurements differed from cadaver-based
estimates, a three-way (measurement method x segment x age) analysis of variance
(ANOVA) with repeated measures was performed. This analysis was repeated for
SM, COM, and I c o m - Within age group, paired t-tests were used for post-hoc testing.
A P-value of < 0.05 was used to determine statistical significance. All statistical tests
were performed using SPSS software.
RESULTS
Dual Energy X-Ray Absorptiometry And Cadaver-Based Estimates
When comparing DXA-derived and cadaver-based SM values, a significant
main effect for measurement method (F = 1 1 6 .6 1 , 4 7; P < 0 .0 1 ) was found.
Additionally, significant measurement method x age (F = 1 2 . 12,47; P < 0.01),
measurement method x segment (F = 5 1 . 2 2,46; P < 0.01), and measurement method x
segment x age (F = 1 .9 4 ,94; P < 0.05) interactions were identified. Post-hoc testing
revealed that at the thigh, DXA-derived SM values were greater than cadaver-based
estimates for 7-8 year-olds (3.7 ± 1.2 kg vs. 3.3 ±1.1 kg), 9-10 year-olds (4.6 +1.2
kg vs. 4.0 ± 1.0 kg), and 11-13 year-olds (5.5 ± 1.4 kg vs. 4.6 ± 1.0 kg) (Figure 4-1).
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
8.0 ■ DXA ^Cadaver-Based (Dempster, 1955)
Shank
Figure 4-1. Comparison of segment mass for the thigh, shank, and foot as
determined by dual energy x-ray absorptiometry (DXA) and cadaver-based estimates
for each age group. * indicates statistically significant differences (P < 0.05).
When comparing DXA-derived and cadaver-based COM values, a significant
main effect for measurement method (F = 22.9i,47; P < 0.01) was found.
Additionally, measurement method x segment (F = 3 8 0 .9 6 2 , 4 6; P < 0.01) and
measurement method x segment x age (F = 3 .3 6 4 ,94; P < 0.05) interactions were
identified. Post-hoc testing revealed that at the thigh, DXA-derived values of the
distance from the proximal segment boundary to the COM were greater than
cadaver-based estimates for 7-8 year-olds (0.145 ± 0.008 m vs. 0.136 ± 0.008 m), 9-
10 year-olds (0.161 ± 0.011 m vs. 0.150 ± 0.009 m), and 11-13 year-olds (0.183 ±
0.013 m vs. 0.169 ± 0.012 m) (Figure 4-2). At the shank, DXA-derived values of the
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
distance from the proximal segment boundary to the COM were less than cadaver-
based estimates for 7-8 year-olds (0.124 ± 0.008 m vs. 0.129 ± 0.009 m), 9-10 year-
olds (0.135 ± 0.008 m vs. 0.141 ± 0.010 m), and 11-13 year-olds (0.151 ± 0.013 m
vs. 0.158 ± 0.014 m) (Figure 4-2). Similarly, at the foot, DXA-derived values of the
distance from the proximal segment boundary to the COM were less than cadaver-
based estimates for 7-8 year-olds (0.047 + 0.005 m vs. 0.049 ± 0.006 m), 9-10 year-
olds (0.051 ± 0.006 m vs. 0.052 ± 0.007 m), and 11-13 year-olds (0.055 ± 0.006 m
vs. 0.057 ± 0.006 m) (Figure 4-2).
0.3 n I DXA ^Cadaver-Based (Dempster, 1955)
a 5 0.0
a
4)
U
7-8 9-10 11-13
Thigh
7-8 9-10 11-13 7-8 9-10 11-13
Shank Foot
Figure 4-2. Comparison of the distance from the proximal segment boundary to the
center of mass for the thigh, shank, and foot as determined by dual energy x-ray
absorptiometry (DXA) and cadaver-based estimates for each age group. * indicates
statistically significant differences (P < 0.05).
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
When comparing DXA-derived and cadaver-based I Co m values, a significant
main effect of measurement method (F = 399.71,47; P < 0.01) was found.
Additionally, measurement method x age (F = 9 .3 2 , 47; P < 0.05), measurement
method x segment (F = 226.82,46; P < 0.01), and measurement method x segment x
age (F = 6 .5 6 4 ,9 4; P < 0.05) interactions were identified. Post-hoc testing revealed
that at the thigh, DXA-derived I c o m values were less than cadaver-based estimates
for 7-8 year-olds (0.025 ± 0.012 kg-m2 vs. 0.035 ± 0.015 kg-m2 ), 9-10 year-olds
(0.036 ± 0.012 kg-m2 vs. 0.051 ± 0.016 kg-m2 ), and 11-13 year-olds (0.056 ± 0.019
kg-m2 vs. 0.076 ± 0.023 kg-m2 ) (Figure 4-3). At the shank, DXA-derived I c o m values
were also less than cadaver-based estimates for 7-8 year-olds (0.010 ± 0.004 kg-m2
vs. 0.013 ± 0.006 kg-m2 ), 9-10 year-olds (0.015 ± 0.005 kg-m2 vs. 0.019 ± 0.007
kg-m2 ), and 11-13 year-olds (0.023 ± 0.008 kg-m2 vs. 0.027 ± 0.009 kg-m2 ) (Figure
4-3). Similarly, at the foot, DXA-derived I c o m values were less than cadaver-based
estimates for 7-8 year-olds (0.0005 ± 0.0004 kg-m2 vs. 0.0011 ± 0.0004 kg-m2 ), 9-10
year-olds (0.0005 ± 0.0002 kg-m2 vs. 0.0015 ± 0.0006 kg-m2 ), and 11-13 year-olds
(0.0006 ± 0.0002 kg-m2 vs. 0.0020 ± 0.0007 kg-m2 ) (Figure 4-3).
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.10
s
*
M
0.08
« 0.06 -
< u
a
h H
o
• f j
a
< U
=
o
0.04 -
0.02
0.00
I DXA H Cadaver-Based (Dempster, 1955)
*
7-8 9-10 11-13
Thigh
7-8 9-10 11-13
Shank
* * *
7-8 9-10 11-13
Foot
Figure 4-3. Comparison of the moment of inertia about the center of mass for the
thigh, shank, and foot as determined by dual energy x-ray absorptiometry (DXA) and
cadaver-based estimates for each age group. * indicates statistically significant
differences (P < 0.05).
Net Joint Moments
Net joint moment curves calculated with DXA-derived anthropometric
measurements were qualitatively and quantitatively similar to those calculated with
cadaver-based estimates for the 7, 10, and 13 year-old subjects (Figures 4-4, 4-5, 4-
6). For each subject, the maximum difference between moments calculated with
cadaver-based and DXA-derived anthropometric proportions occurred at the hip
during the swing phase of gait. The magnitude of this difference, however, was quite
small, ranging from 0.03 Nm/kg body weight in the 10 year-old to 0.06 Nm/kg body
weight in the 7 year-old.
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a) Ankle
JS
b t
T 3
O
M
o n
a
sz
1.5
1
0.5
0
-0.5
-1
-1.5
1 Plantarflexor
- - - 'Cadaver-Based
\
--------- DXA-Derived
Dorsiflexor
> s a
.0 1 5
* 3
£
£
s
3 3
B O
0 4
a
£
1.5
1
0.5
0
-0.5
-1
-1.5
o a
b t
• p M
o >
£
-d
o
3 3
b t
04
a
£
1.5
1
0.5
0
-0.5
-1
-1.5
20
b) Knee
40 60
% Gait Cycle
80 100
" 1 Extensor
- - - 'Cadaver-Based
Flexor
- 1 1 i I i
20 40 60
% Gait Cycle
80 100
c) Hip
Extensor ■ - - 'Cadaver-Based
'A
---------DXA-Derived
/ t
\ J
Flexor
- .....- 1 i 1 1 1
20 40 60
% Gait Cycle
80 100
Figure 4-4. Net ankle (a), knee (b), and hip (c) joint moments obtained from a 7-
year-old while walking at a self-selected velocity. Moments calculated using DXA-
derived and cadaver-based anthropometric body segment parameters. Vertical line
distinguishes stance from swing phase.
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4 3
OJD
■ p H
4 >
£
T J *
0
c a
ex
4 4
1
Z
1.5
1
0.5
0
-0.5
-1
-1.5
a) Ankle
" 1
- - - 'Cadaver-Based
Plantarflexor/ \
\
--------- DXA-Derived
Dorsiflexor
i i i -----1 ---- 1 — i i i i i
xi
two
•P H
4 J
£
£
o
P P
ex
4 4
Z
43
ex
• pH
< u
£
£
o
p a
ex
44
z
1.5
1
0.5
0
-0.5
-1
-1.5
1.5
1
0.5
0
-0.5
-1
-1.5
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
b) Knee
1
- ' ' 'Cadaver-Based
Extensor
--------- DXA-Derived
Flexor
0 10 20 30 40 50 60 70 80 90 100
c) Hip
% Gait Cycle
' - - 'Cadaver-Based
Extensor
--------- DXA -Derived
Flexor
T ---- 1 ---- 1 ---- 1 ---- 1 ---- 1 ---— i---- 1 ---- 1---- 1 ---- r
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
Figure 4-5. Net ankle (a), knee (b), and hip (c) joint moments obtained from a 10-
year-old while walking at a self-selected velocity. Moments calculated using DXA-
derived and cadaver-based anthropometric body segment parameters. Vertical line
distinguishes stance from swing phase.
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a) Ankle
Plantarflexor 'Cadaver-Based
DXA-Derived
W > -0.5 -
Dorsiflexor
-1.5
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
b) Knee
Cadaver-Based
Extensor
DXA-Derived
on -0.5 -
-1.5
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
c) Hip
Cadaver-Based
Extensor
d o
DXA-Derived
0.5 -
M -0.5 -
Flexor
-1.5
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
Figure 4-6. Net ankle (a), knee (b), and hip (c) joint moments obtained from a 13-
year-old while walking at a self-selected velocity. Moments calculated using DXA-
derived and cadaver-based anthropometric body segment parameters. Vertical line
distinguishes stance from swing phase.
50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DISCUSSION
This study compared DXA-derived anthropometric parameters to cadaver-
based estimates in 7-13 year-old children. Although most of the DXA-derived SM,
COM, and Icom values differed statistically from cadaver-based estimates, the
magnitudes of these differences were quite small and had a negligible effect on the
calculation of net joint moments during gait.
When averaged across age groups, DXA-derived SM values for the thigh
were 0.5 kg greater than cadaver-based predictions. These differences equate to only
a 1-2% difference when SM is expressed as a proportion of total body mass, as is
customary for moment calculations. There were no differences in SM values at the
shank or foot between the two methods. Maximum differences between DXA-
derived and cadaver-based COM values were 0 .0 0 2 m at the foot and 0.01 m at the
shank and thigh. When expressed as a proportion of segment length, only a 1-3%
difference was observed. The greatest percent differences between DXA-derived
anthropometric parameters and cadaver-based predictions occurred in Icom (1 3 -
70% ). However, the actual magnitudes of these differences were minimal. For
example, the maximum difference between measurement methods was 0.001 kg-m2
at the foot, 0 .0 0 4 kg-m2 at the shank, and 0 .0 2 kg-m2 at the thigh. When Rg as a
proportion of segment length was considered, a 7-22% difference was noted.
The few studies that have reported body segment parameters for children 7-
13 years of age have not provided mean data for specific ages. Nonetheless, gross
comparisons indicate that the results of the present study compare favorably to
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
previous reports. For example, using elliptical zone modeling techniques, Ackland et
al. reported SM (% total body mass) as 12.7% and 5.4% for the thigh and shank,
respectively, in children with a mean age of 13.7 years [1]. Similarly, regression
equations developed by Jensen predict thigh and shank SM to be 9.2% and 4.7% for
7 year-olds and 11.0% and 5.3% for 12 year-olds [16]. In comparison, the present
study reported thigh and shank mass as 11.0% and 4.6% of body mass for 7-8 year-
olds and 11.7% and 4.8% of body mass for 11-13 year-olds.
With regards to COM (expressed as a percentage of segment length from the
proximal segment boundary), Ackland reported values of 43.6% and 41.8% for the
thigh and shank [1], while Jensen’s equations predict values of 45.6% and 43.2% for
7 year-olds and 46.2% and 41.7% for 12 year-olds [16]. In comparison, the present
study reported thigh and shank COM values of 46.3% and 41.5% of segment length
for 7-8 year-olds and 46.8% and 41.5% of segment length for 11-13 year-olds.
I c o m values for the thigh (0.0036 ± 0.0019 kg-m2 ) and shank (0.015 ± 0.008
kg-m2) reported in the present study for 11-13 year-olds were substantially less than
9 9
those reported by Ackland (0.008 kg-m ; 0.037 kg-m ) [1]. However, the radius of
gyration proportions predicted by Jensen’s equations for the thigh, shank, and foot
(0.272, 0.284, and 0.239, respectively) [16] are comparable to those reported in the
present study (0.256, 0.276, and 0.259, respectively).
The resulting moment curves, which were calculated with the two sources of
anthropometric data, were quite similar in all cases. The only discernible differences
occurred at the hip. This can probably be attributed to the differences in thigh mass
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and mass distribution between cadaver-based and DXA-derived proportions and the
fact that calculations from inverse dynamics equations begin distally. Hence any
“error” at the foot, shank, or thigh will affect the net joint moment calculation at the
hip. Furthermore, the greatest differences between moment curves occurred during
the swing phase of gait when inertial terms, which are greatly influenced by
anthropometric values, dominate the moment calculation.
Three potential limitations to the anthropometric portion of the present study
should be considered. First, each segment was assumed to be symmetrical, such that
the moments of inertia and radii of gyration were equal about the anteroposterior and
mediolateral axes. Based on studies of mass distribution in adults [9], this is a
reasonable assumption. Next, I c o m may have been underestimated as each segment
was modeled as a set of point masses spaced along the longitudinal axis. However,
any such underestimation would be small and unappreciable once converted to Rg.
Lastly, anthropometric values for the foot were determined in the frontal plane,
therefore, some error associated with the out of plane component of the foot axis was
unavoidable. However, with the ankle positioned in maximal plantarflexion, the
assumed foot axis (ankle to 5th metatarsal head) approached the frontal plane, thus
minimizing the magnitude of the error.
CONCLUSION
Although DXA-derived anthropometric parameters obtained from 7-13 year-
old children differed statistically from those predicted from cadaver-based estimates,
absolute and relative differences were minimal. The similarity between net joint
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
moments calculated using these two sources of anthropometric data suggests that
cadaver-based estimates are satisfactory for use in gait analysis of children.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER V
GAIT KINEMATICS AND KINETICS OF 7 YEAR-OLD CHILDREN: A
COMPARISON TO ADULTS USING AGE-SPECIFIC ANTHROPOMETRIC
DATA
Previous comparisons of adult and children’s gait have been inconclusive
with respect to the age at which mature kinetic patterns are expressed. Furthermore,
previous investigators have used cadaver-based anthropometric estimates for kinetic
calculations in children as well as adults. Using age-specific anthropometric data
from Chapters III and IV, this chapter determined if sagittal plane gait kinematics
and kinetics of 7 year-old children differ from adults. Joint angles, moments, and
power obtained during level walking in 7 year-old children (n=15) were compared to
data from adults (n=15).
INTRODUCTION
An understanding of age-related changes in ambulation patterns is essential
for diagnosing and treating pathological gait in children. In a review paper on the
development of mature gait, Sutherland posed two yet-to-be-answered questions;
how does gait in children differ from that in adults, and at what age do children
achieve adult-like patterns [35]? In addition, Sutherland emphasized the importance
of identifying factors controlling the acquisition of adult-like gait characteristics. To
date, the literature describing typical gait patterns in developing children is relatively
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sparse, and comparisons are often made for diagnostic and intervention purposes to
adult normative data.
Gait kinematics of children have been reported to be similar to those of adults
by 3 V 2 - 4 years of age [37]. However, with regards to underlying kinetics,
discrepant results have been reported. Ounpuu et al. examined net joint moments of
the hip, knee, and ankle during gait in 5-16 year-old children (n=31) and concluded
that kinetic patterns were adult-like by 5 years of age [25]. This conclusion,
however, was based on qualitative comparisons of averaged data from 5-16 year-old
children with previously reported adult data [19], and no between-age comparisons
were reported. In contrast, Cupp et al. compared the kinetics of 4-5 (n=6), 6-7 (n=7),
and 8-10 (n=10) year-old children to a control group of 18-21 year-old adults (n=5),
and despite similar walking velocities and normalization of moments to body weight,
all groups of children < 7 years demonstrated distinct kinetic differences compared
to adults [8]. Specific differences in the sagittal plane included diminished peak knee
extensor and ankle plantarflexor moments, as well as differences in ankle and knee
power bursts.
Age-related differences in the gait profiles of children and adults have been
interpreted as evidence that children lack the neuromuscular maturation to support an
adult-like gait pattern [35]. An alternate hypothesis is that the kinetics of gait differ
between adults and children because of differences in body size and body segment
proportions. For example, it has been postulated that previously identified kinetic
differences between adults and children during gait may be artificial, in that they
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
may result from the use of anthropometric data that do not accurately represent the
body proportions of children [8]. While there is support for this hypothesis [17], it
has yet to be tested as direct methods to obtain age-specific anthropometric data for
children have not, until recently, been available [11].
Using age-specific anthropometric data obtained from dual energy x-ray
absorptiometry, the purpose of this study was to determine if sagittal plane gait
kinematics and kinetics of 7 year-old children differ from those of adults. 7 year-old
children were selected because existing literature suggests that mature kinetics are
still emerging at this age even though adult-like kinematics are well established. The
identification of age-related differences would imply that comparisons of children’s
data to adult data for diagnostic and intervention purposes might not be appropriate.
As such, this study is a first step toward assessing the need for a pediatric database of
age-specific, normative gait data.
METHODS
Subjects
Fifteen 7 year-old children (8 female, 7 male) and fifteen adults (8 females, 7
males) participated in this study. The average age, height, and mass of the children
was 7.4 + 0.4 years, 121 + 5.0 cm, and 24.8 ± 6.6 kg, respectively. The average age,
height, and mass of the adults was 31.8 + 6.8 years, 167.2 ± 9.3 cm, and 69.7 ± 13.0
kg, respectively. All subjects were screened to rule out any neurologic or orthopedic
disorders that would influence gait. Prior to participation, all procedures were
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
explained to each subject and his/her parent(s), and informed consent was obtained,
as approved by the Institutional Review Board of the University of Southern
California, Health Sciences Campus.
Instrumentation
Three-dimensional lower extremity kinematics were recorded at 60 Hz using
a six-camera Vicon motion analysis system (Oxford Metrics Ltd., Oxford, England).
Ground reaction forces were recorded at a rate of 1200 Hz using 3 AMTI force plates
(OR6-5-1000; Advanced Mechanical Technology, Inc., Newton, MA) camouflaged
within a 10-meter walkway.
Procedures
All testing was performed in the Musculoskeletal Biomechanics Research
Laboratory on the Health Science Campus of the University of Southern California.
Reflective markers (20 mm spheres) were placed at anatomical landmarks of each
lower extremity (posterior superior iliac spine, anterior superior iliac spine, lateral
thigh, femoral epicondyle, lateral tibia, lateral malleolus, calcareous, 5th metatarsal
head, and the dorsum of the foot). After 2-3 practice trials, kinematics and ground
reaction forces were recorded simultaneously as subjects walked barefoot along the
10-meter walkway at a self-selected speed. Photoelectric switches positioned at
either end of the walkway automatically triggered the start and end of data
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
collection. Three trials per subject were recorded. Only trials in which the entire foot
made contact with the force-plate without targeting were included in the analysis.
Data Management
Reflective markers were manually labeled then digitized using Vicon
Workstation, software version 4.4 (Oxford Metrics Ltd., Oxford, England).
Reflective markers were used to locate embedded coordinate systems for the pelvis,
thigh, shank, and foot. All trajectories were filtered using a Woltring filtering process
[41].
Marker trajectories were fit with a cubic spline and 100 equally spaced points
were output. Following calculation of segment orientation and axes, segment motion,
defined as rotation of the distal segment about a fixed point on the proximal segment,
was determined. Euler angles were used to determine segmental motion, and joint
angles were determined based on relative motion of the segments comprising the
joint of interest.
Sagittal plane net joint moments (hip, knee, and ankle) were calculated using
inverse dynamics equations [4] and age-specific anthropometric proportions (Table
5-1) derived in previous studies using dual energy x-ray absorptiometry (DXA)
[11,12], Briefly, from full body DXA scans (Hologic QDR-1500), the masses of the
thigh, shank, and foot, as well as the masses of slices within these segments were
measured. Using established equations, segment mass, center of mass location, and
moment of inertia of the thigh, shank, and foot were determined and converted to
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
proportions of total body mass and segment length [40]. Age-specific proportions for
7 year-olds (n=20) and adults (n=20) were generated from averaged group data.
Table 5-1
Age-Specific Anthropometric Proportions Used For The Calculation Of Net
Joint Moments During Gait [11] [12]
Segment Mass/Body
Mass
COM/Segment Length
(proximal axis)
Rg /Segment 1
(about CC
^ength
>M )
Thigh Shank Foot Thigh Shank Foot Thigh Shank Foot
7 Year-
Olds 0.112 0.047 0.013 0.459 0.417 0.482 0.262 0.275 0.257
Adults 0.113 0.045 0.010 0.459 0.411 0.497 0.251 0.266 0.254
Joint power was calculated as the dot product of the net joint moment and
joint angular velocity. Joint moments and power were normalized to body weight,
and all kinematic and kinetic data were normalized to 100% of the gait cycle.
Dependent variables included peak flexion and extension angles, peak flexor
and extensor moments, and peak joint power absorption and generation. For each
variable, peaks were averaged across 3 trials for each subject. Group means were
then generated.
Statistical Analysis
Independent Mests were used to compare each dependent variable between
groups. A Bonferroni correction was made for the multiple comparisons (23),
resulting in a significance level of 0.002.
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
RESULTS
The average walking velocity for both groups was identical (1.3 ±0.1 m/sec),
however the 7 year-olds achieved this speed utilizing a significantly greater cadence
(146.2 ± 8.6 steps/min vs. 120.2 ± 6.9 steps/min) and significantly shorter step length
(0.5 ± 0.1 m vs. 0.7 ± 0.1 m) compared to adults.
No significant differences in joint kinematics were observed between 7 year-
olds and adults at the hip, knee, or ankle (Table 5-2, Figure 5-1). While there was a
trend toward decreased peak dorsiflexion angles in 7 year-olds (stance: 8.3 ± 3.1° vs.
10.9 ± 1.9°; swing: 0.9 ± 3.5° vs. 1.8 ± 1.3°), statistical significance was not reached.
Table 5-2
Comparison Of Peak Joint Angles Between 7-Year-Olds And Adults
(All values in degrees)
7 Year-Olds Adults
P-value Mean SD Mean SD
Hip Flexion 32.6 4.8 33.7 3.8 0.49
Hip Extension 17.0 5.7 13.6 6.4 0.13
Knee Flexion Swing 60.5 4.3 59.5 5.9 0.61
Knee Flexion Stance 7.1 4.9 4.9 3.0 0.40
Knee Extension -0.6 4.2 -0.4 2.8 0.92
Ankle Dorsiflexion Stance 8.3 3.1 10.9 1.9 0.01
Ankle Dorsiflexion Swing 0.9 3.5 1.8 1.3 0.01
Ankle Plantarflexion 22.9 8.8 20.2 6.2 0.34
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a) Hip
7 Year-olds Adults
Flexion
30
-10
Extension
-20
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
b) Knee
Flexion
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
c) Ankle
Dorsiflexion
M
t u
Q -10
-15
-20
-25
Plantar flexion
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
Figure 5-1. Comparison of joint angles at the hip (a), knee (b), and ankle (c)
between 7 year-olds and adults. All data normalized to 100% of the gait
cycle. Vertical line distinguishes stance from swing phase.
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
No between-group differences were noted in peak net joint moments at the
hip or knee. However, peak plantarflexor moments of 7 year-olds (1.15 ± 0.18
N-m/kgbody weight) were significantly smaller than those of adults (1.56 ± 0.11
N'm/kgbody weight) during late stance (P < 0.002; Table 5-3, Figure 5-2).
Table 5-3
Comparison of Peak Net Joint Moments Between 7-Year-Olds and
Adults (All values in N-m/kg body weight)
7 Year-Olds Adults
P-value Mean SD Mean SD
Hip Extensor 0.92 0.18 0.97 0.26 0.54
Hip Flexor 0.61 0.20 0.73 0.21 0.12
Knee Extensor 0.38 0.21 0.43 0.21 0.54
Knee Flexor 0.38 0.12 0.43 0.11 0.26
Ankle Plantarflexor 1.15 0.18 1.56 0.11 P< 0.0001
Ankle Dorsiflexor 0.12 0.06 0.13 0.08 0.76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a) Hip
Adults 7 Year-olds
1 .0
!xtensor
0.0
£ -0.5
Flexor
- 1.0
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
b)Knee
0.5
Extensor
0.0
ex
Flexor
-0.5
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
c) Ankle
2.0 Plantarflexor
1.5
1.0
0.5
0.0
Dorsiflexor
-0.5
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
Figure 5-2. Comparison of net joint moments at the hip (a), knee (b), and ankle (c)
between 7 year-olds and adults. All data normalized to body weight and 100% of the
gait cycle. * indicates P < 0.002. Vertical line distinguishes stance from swing phase.
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Similarly, no between-group differences in joint power were observed at the
hip or knee. However, peak power absorption at the ankle during midstance was
significantly less in 7 year-olds compared to adults (-0.56 ± 0.25 watts/kg vs. -1.05 ±
0.22 watts/kg vs.), as was peak power generation during late stance (2.79 ± 0.44
watts/kg vs. 3.46 ± 0.55 watts/kg) (P < 0.002; Table 5-4, Figure 5-3).
Table 5-4
Comparison of Peak Power Between 7-Year-Olds and Adults (all values in
Watts/kg body weight). Positive values indicate power generation, and negative
values indicate power absorption. See Figure 3 for power phases by joint
7 Year-O ds Adults
P-value
Power Phase
By Joint
Mean SD Mean SD
Ankle 1 (Al) -0.56 0.25 -1.05 0.22 P< 0.0001
Ankle 2 (A2) 2.79 0.44 3.46 0.55 P< 0.001
Knee 1 (Kl) -1.12 0.74 -0.99 0.85 0.67
Knee 2 (K2) 0.68 0.36 0.67 0.55 0.95
Knee 3 (K3) -0.62 0.23 -0.64 0.21 0.80
Knee 4 (K4) -0.76 0.25 -0.63 0.12 0.08
Hip 1 (HI) 0.79 0.45 0.82 0.32 0.82
Hip 2 (H2) -0.82 0.39 -0.77 0.38 0.73
Hip 3 (H3) 0.95 0.24 1.23 0.33 0.02
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
5 2
D £
• PM
O
£ 1
■ o
C 0
0 3
6 * . .
^-1
£ '2
-3
a) Hip
......... A dults------- -7 Year-olds
Generation
H3
HI
V 1 ^
H2
Absorption
1 1 1 1 " " T ... r • •' l...............i ---------1 ------------r
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
b) Knee
£ 2
.2?
‘ 3
£ i
> >
© 0
p a
© £
=*-1
C3 -
£ '2
-3
Generation
K2
\
K1
Absorption
K3 K4
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
c) Ankle
4
Generation A2
A1
Absorption
-3
0 10 20 30 40 50 60 70 80 90 100
% Gait Cycle
Figure 5-3. Comparison of joint power at the hip (a), knee (b), and ankle (c) between
7 year-olds and adults. All data normalized to body weight and 100% of the gait
cycle. * indicates P < 0.002.
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DISCUSSION
The purpose of the present study was to determine if sagittal plane gait
kinematics and kinetics of 7 year-old children differed from adults when age-specific
anthropometries were used. The results demonstrated that the gait patterns of 7 year-
olds were similar to those of adults for most of the variables examined, however,
some differences were observed. Specifically, during late stance, 7 year-olds
presented with diminished peak plantarflexor moments and less peak power
absorption and generation at the ankle.
The results of the current study, for the most part, are consistent with those
reported previously. The lack of age-related differences in joint kinematics support
the findings of Sutherland and colleagues who reported that joint angles during gait
are similar to adults in children by 3'/2-4 years of age [37]. The differences in ankle
net joint moments identified in the present study are consistent with the data of Cupp
and colleagues who reported diminished magnitudes of peak plantarflexor moments
in 6-7 year-olds (1.0 N-m/kg body weight) when compared to adults (1.5 N-m/kg
body weight) [8]. Examination of the data reported by Ounpuu and colleagues
suggests that peak plantarflexor moments in late stance also were around 1.0 N-m/kg
body weight for children; however, age-related statistical comparisons were not
made in their study [25]. Both the present study and Cupp et al. found that children
absorbed less power during midstance and generated less ankle power during late
stance. In contrast to Cupp and colleagues [8], however, we did not find age-related
differences in knee kinetics.
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The observed differences in kinetic variables cannot be explained by
anthropometric proportions as age-specific values were used in the current study.
Similarly, they cannot be explained by velocity since both groups walked at the same
speed (1.3 m/sec). It has been suggested that age-related kinetic differences reflect
neuromuscular immaturity in children. For example, Sutherland, hypothesized that
children may have to rely more on hip power and less on ankle power during
terminal stance because of maturational factors associated with the plantarflexors
that result in decreased torque and power producing capabilities [36], However,
examination of our data revealed that the identified differences in ankle kinetics
between 7 year-olds and adults could be explained by factors unrelated to
neuromuscular maturation.
The primary factor contributing to the net joint moment at the ankle during
stance is the vertical ground reaction force and its spatial relationship to the axis of
rotation of the foot segment. The further the ground reaction force is from the axis of
rotation, the greater the lever arm, and, thus, the greater the moment. During late
stance, when peak plantarflexor moments occur, the center of pressure is located
anterior to the axis of rotation, typically at or near the metatarsal heads. Therefore,
the perpendicular distance from the center of pressure to the axis of rotation would
appear to be dictated by foot length (Figure 5-4). Post-hoc analysis of our data
revealed that peak plantarflexor moments and foot length were strongly correlated (r
= 0.80), such that 64% of the variance in plantarflexor moments could be explained
by foot length (Figure 5-5).
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5-4. Schematic of the foot in late stance illustrating the relationship between
the vertical component of a ground reaction force (Fy) and its lever arm (/), creating a
moment (M) about the segment center of mass.
* Adults ■ 7 Year-Olds
W O
u
©
X
* E s
O S £
^ w
fl
e s
P h
P M
S3
E
o
1.5
0.5
■ 1
150
r = .80
250 200
Foot Length (mm)
Figure 5-5. Correlation between peak ankle moment and foot length for all subjects.
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The trend towards decreased dorsiflexion in 7 year-olds during late stance
also could have contributed to the kinetic differences observed at the ankle.
Diminished dorsiflexion during terminal stance would result in less of a need for
power absorption as the plantarflexors control the rate and extent of forward tibial
rotation. Similarly, reduced dorsiflexion range during late stance could result in a
decreased plantarflexor moment as the anterior displacement of the body’s center of
mass, and therefore the center of pressure, relative to the foot segment center of mass
would be limited.
By definition, net joint power is the product of the net joint moment and joint
angular velocity, both of which probably contributed to decreased peak power
absorption and generation at the ankle during late stance in the 7 year-olds. For
example, decreased peak ankle power was observed at the same time as the
decreased peak plantarflexor moment. Also, the diminished rate of change of tibial
advancement in stance suggests a lower ankle angular velocity in 7 year-olds (Figure
5-1 c). Taken together, these factors likely explain the observed decreases in ankle
power in the 7 year-old group.
CONCLUSION
The results of this study provide evidence that gait kinematics and kinetics of
7 year-old children approximate those of adults when age-specific anthropometric
data were used. Although significant differences in ankle kinetics were observed,
physical factors, rather than neuromuscular maturation could account for this
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
finding. For example, foot length explained a large portion of the variance in peak
ankle plantarflexor moment. This suggests that consideration should be given to
normalizing ankle kinetics to foot length.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VI
INTERSEGMENTAL DYANMICS DURING THE SWING PHASE OF GAIT:
A COMPARISON OF KNEE KINETICS BETWEEN 7 YEAR-OLD
CHILDREN AND ADULTS
Mature movement patterns are characterized by the ability to exploit
interaction moments (those generated through the movement of linked segments)
while minimizing muscle moments [33]. In order to examine joint kinetics at this
level, a detailed analysis of the contributions of the interaction and muscle moments
to the net joint moment is necessary. These relationships are especially important
during the swing phase of gait, when ground reaction forces are absent and
interaction torques have a relatively high contribution to the net joint moment. In
order to determine whether gait patterns of children are mature by 7 years of age, the
use of interaction moments during the swing phase of gait was compared in 7 year-
old children and adults. Specifically, the contributions of muscle and interaction
moments to the net knee joint moment were examined in adults (n=10) and 7 year-
old children (n=10). Age-specific anthropometric data from Chapters III and IV were
used for kinetic calculations.
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
INTRODUCTION
An understanding of the development of normal walking, as well as
knowledge regarding the factors underlying the acquisition of adult-like gait, is
necessary for the diagnosis and treatment of gait disabilities in children. Most
children develop the ability to ambulate independently within the first 1 to 1 1/2
years of life, however many features of gait continue to mature over the course of
several years [37,39]. For example, children demonstrate kinematic patterns that are
similar to adults by 3-4 years [37], however the underlying kinetics may differ from
adults until at least 7 years of age [8]. Historically, kinetic differences between
children and adults have been attributed to neuromuscular immaturity [35].
Kinetic evaluation of pediatric gait has typically included the analysis of net
joint moments and power, and decisions regarding what is adult-like have been based
on the timing and magnitudes of these variables. However, it has been suggested that
the ability to exploit interaction moments (those generated through the movement of
linked segments) while minimizing muscle moments determines whether movement
patterns are mature [33]. In order to examine joint kinetics at this level, a detailed
analysis of the contributions of the interaction and muscle moments to the net joint
moment is necessary.
The focus of previous investigations comparing children to adults has been
on the stance phase of gait. In general, during stance, net joint moments are
dominated by muscle moments, and the contribution of interaction moments is
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
minimal [40]. However, the intricate relationship between interaction and muscle
moments during gait is most evident during swing, when ground reaction forces are
absent and angular accelerations are high [27].
The development of adult-like kinetic patterns, especially those involving
non-muscular interaction moments, occurs with time and experience [23,24,38]. For
example, it has been reported that the ability to exploit interaction moments, which
contribute to the effectiveness of upper extremity reaching tasks, is a product of age
and experience [23,38]. Similarly, McFadyen and colleagues demonstrated that
adult-like patterns of obstacle avoidance tasks develop with experience. These
authors attributed the "less mature" kinetic patterns observed in children to varying
strategies of coordinating the movement of linked segments [24].
To date, there is only indirect evidence that children may differ from adults in
terms of their use of interaction moments during gait. Adults accomplish knee
flexion for swing phase primarily through forward thigh advancement with little to
no muscle activity at the knee [26,27,28], However, McFadyen et al. found that
during normal walking trials, some 7-9 year-old children accomplished swing phase
knee flexion using the hamstrings [24], Similarly, Cupp and colleagues found that
children < 7 years demonstrated power generation rather than absorption at the knee
during initial swing. Taken together these findings suggest that interaction moments
generated by linked lower extremity segments may be the primary cause of swing
phase knee flexion in adults while children < 7 years may employ concentric muscle
activity to achieve the same goal.
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The purpose of this paper was to compare the use of interaction moments
during the swing phase of gait in 7 year-old children and adults. Specifically, I
sought to determine whether the contributions of muscle and interaction moments to
the net knee joint moment would differ between these two age groups. The knee was
examined for two reasons: 1) literature suggests that differences in knee kinetics
between children and adults; 2) the knee represents the intermediate joint of the
lower extremity, and it is the intermediate joint where the effects of interaction
torques are often most apparent. It was hypothesized that 7 year-olds would
demonstrate net joint moments similar to adults, but that they would demonstrate a
relatively larger contribution of muscle moment to the net joint moment and a
smaller contribution of interaction moment to the net joint moment compared to
adults. The focus of this investigation was on 7 year-olds because existing literature
suggests that adult-like kinetics are still emerging at this age. Results from this study
will help to determine at what age gait patterns in children are adult-like, thereby
providing a benchmark for assessing pathology (i.e. children with disabilities who
often are not privileged with abundant opportunities to explore their environments
through movement).
METHODS
Subjects
Ten 7 year-old children (4 female, 6 male) and ten adults (5 females, 5 males)
participated in this study. The average age, height, and mass of the children was 7.5
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
± 0.4 years, 123 ± 4.3 cm, and 23.7 ± 2.2 kg, respectively, while the adults averaged
32 + 8 years, 168 ± 10 cm, and 70.0 ± 15.1 kg. These subjects were screened to rule
out any neurologic or orthopedic disorders that would influence gait. Prior to
participation, all procedures were explained to each subject and his/her parent(s), and
informed consent was obtained (as approved by the Institutional Review Board of
the University of Southern California, Health Sciences Campus).
Instrumentation
Three-dimensional lower extremity kinematics were recorded at 60 Hz using
a six-camera Vicon motion analysis system (Oxford Metrics Ltd., Oxford, England).
Ground reaction forces were recorded at a rate of 1200 Hz using 3 AMTI force plates
(OR6-5-1000; Advanced Mechanical Technology, Inc., Newton, MA) camouflaged
within a 10-meter walkway.
Procedures
All testing was performed in the Musculoskeletal Biomechanics Research
Laboratory on the Health Science Campus of the University of Southern California.
Reflective markers (20 mm spheres) were placed at anatomical landmarks of each
lower extremity (posterior superior iliac spine, anterior superior iliac spine, lateral
thigh, femoral epicondyle, lateral tibia, lateral malleolus, calcareous, 5th metatarsal
head, and the dorsum of the foot). After 2-3 practice trials, kinematics and ground
reaction forces were recorded simultaneously as subjects walked barefoot along the
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
10-meter walkway at a self-selected speed. Photoelectric switches positioned at
either end of the walkway automatically triggered the start and end of data
collection. Three trials in which the entire foot made contact with the force-plate
without targeting were recorded for each subject.
Data Management
Reflective markers were manually labeled and digitized using Vicon
Workstation, software version 4.4 (Oxford Metrics Ltd., Oxford, England).
Reflective markers were used to locate embedded coordinate systems for the pelvis,
thigh, shank, and foot. All trajectories were filtered using a Woltring filtering process
and fit with a cubic spline. [41]
Following calculation of segment orientation and axes, segment motion was
determined. Euler angles were used to determine sagittal plane knee motion. Knee
angular velocities and accelerations were derived from the kinematic data.
Net joint moments (sagittal plane) were calculated using inverse dynamics
equations [4] and age-specific anthropometric proportions derived from previous
studies using dual energy x-ray absorptiometry (DXA) (Table 5-1) [11,12],
The components of the net knee joint moment (gravity, interaction, and
muscle moments) were calculated using Newtonian equations of motion (Appendix
A) [14]. The gravity moment was defined as the moment created by the force of
gravity acting at the center of mass of the shank, while the interaction moment was
defined as the moment generated at the knee by the movement of the thigh and shank
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
segments. The muscle moment represented the moment created by contractile and
non-contractile internal structures and was defined as the difference between the net
moment and the sum of the gravity and interaction moments (Appendix A). For each
component, a moment that produced knee extension was designated as being positive
while a moment that produced flexion was negative. All moment data were
normalized to body weight and to 100% of the swing phase.
For each subject, mean curves (average of 3 trials) were generated for the
following variables: knee flexion angle, knee angular velocity, knee angular
acceleration, net knee joint moment, gravity moment, muscle moment, and
interaction moment. The minimum and maximum values for each of these values
were identified and used for statistical analysis.
Root-mean-square-error (RMSE) scores were used to establish the relative
contribution of the interaction and muscle moments to the net joint moment. Two
separate RMSE scores were calculated during swing from the mean curves for each
subject, a net-to-interaction RMSE and a net-to-muscle RMSE. The net-to-
interaction RMSE represented the average difference between the interaction
moment and the net joint moment during the swing phase, while the net-to-muscle
RMSE represented the average difference between the muscle moment and the net
joint moment. Group means were then generated for the net-to-interaction RMSE
and the net-to-muscle RMSE values.
78
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Statistical Analysis
Independent /-tests were used for between-group comparisons of minimum
and maximum knee flexion angles, knee angular velocities, knee angular
accelerations, net knee joint moments, gravity moments, muscle moments, and
interaction moments. To determine if the relative contributions of interaction and
muscle moments to the net knee joint moment differed between 7 year-olds and
adults, /-tests were used to compare a) net-to-interaction RMSE scores and b) net-to-
muscle RMSE scores between groups. For all comparisons, a P-value of < 0.05 was
used to determine statistical significance. SPSS statistical software was used for all
analyses.
RESULTS
There were no statistically significant differences in knee kinematic variables
between the two groups (Table 6-1; Figure 6-1). The average minimum and
maximum knee flexion angles during swing in 7 year-olds closely matched those of
adults as did the average minimum and maximum knee angular velocities and
accelerations (Table 6-1).
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 6-1
Comparison of Kinematics and Kinetics between 7 Year-Old Children
and Adults
7 Year-Olds Adults
Mean SD Mean SD P-value
Knee Flexion Angles (°)
Min 5.9 5.4 6.0 3.9 0.93
Max 60.1 3.7 60.2 6.3 0.98
Angular Velocity (°/sec)
Min 2.7 1.9 1.9 1.4 0.45
Max 323.9 38.4 337.6 27.2 0.65
Angular Acceleration
(7sec2 ) -4287.3 722.3 -4549.4 674.7 0.44
Min 2975.8 553.1 3423.0 675.9 0.14
Max
Net Joint Moment
(N-m/kg body weight)
Min -0.20 0.02 -0.20 0.04 0.99
Max 0.08 0.03 0.10 0.03 0.18
Muscle Moment
(N-m/kg body weight)
Min -0.18 0.02 -0.18 0.04 0.74
Max 0.08 0.02 0.10 0.02 0.37
Interaction Moment
(N-m/kg body weight)
Min -0.08 0.019 -0.08 0.020 0.62
Max 0.02 0.006 0.03 0.010 0.11
Gravity Moment
(N-m/kg body weight)
Min -0.04 0.005 -0.04 0.008 0.51
Max 0.07 0.004 0.08 0.01 0.43
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
a)
Adults - - - -7 Year-Olds
75
50
25
25 50 75 100 0
k) 100 % Swing Phase
400
0 50 25 75 100
100% Swing Phase
5000
-5000
50 0 25 75 100
100% Swing Phase
Figure 6-1. Knee angles (a) angular velocity (b) and acceleration (c), normalized
100% of the swing phase, obtained from 7 year-old children (n=10) and adults
(n=10) while walking at a self-selected velocity. All data normalized to 100% of
swing phase.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The net joint knee moments during swing in 7 year-olds were similar to those
of adults (Table 6-1; Figure 6-2). Similarly, the contributions of muscle and
interaction moments to the net joint moment were comparable between the two
groups (Table 6-1; Figure 6-2).
The net-to-interaction RMSE scores of the 7 year-olds (0.12 ± 0.01 N-m/kg
body weight) did not differ statistically from those of adults (0.13 ± 0.02 N-m/kg
body weight) (Table 6-1; Figure 6-3). Similarly, net-to-muscle RMSE scores of the 7
year-olds (0.030 ± 0.003 N-m/kg body weight) did not differ statistically from those
of adults (0.032 ± 0.006 N-m/kg body weight) (Table 6-1; Figure 6-3).
a) Net Joint Moment
0.2
E x t e n s i o n
A d u l t s
« - - 7 Y e a r - O l d s
M -0 .1
1 - 0.2
F l e x i o n
-0.3
0.2
0.1
0
ffi -0.1
O JO
1 - 0 - 2
Z
- 0 . 3
100 %Swing P h a s e
b) Interaction Moment
£
E x t e n s i o n A d u l t s
- - - - 7 Y e a r - O l d s
F l e x i o n
1 0 0 %Swing P h a s e
c) Muscle Moment
0.2 y 1-------------------------------------
E x t e n s i o n
A d u l t s
•1 Y e a r - O l d s
-o.i
1 - 0.2
F l e x i o n
-0.3
0.2
■ m ° -l
J S
.2°
I 0
■ o
e
CO ‘
100%Swing P h a s e
d) Gravity Moment
E x t e n s i o n
- A d u l t s
• 7 Y e a r - O l d s
- 0.1
1: -0.2
Z
-0.3
F l e x i o n
100 %Swing P h a s e
Figure 6-2. Net joint moments (a), interaction moments (b), gravity moments (c), and
muscle moments (d) from adults (n=10) and 7 year-old children (n=10) normalized to
body weight and 100% of the swing phase.
82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
7 Year-Olds
a)
Adults
b)
0.2
Net Joint Moment
Extension
Muscle Moment
-0.1
o n
- 0.2
Flexion RMSE = 0.030 ±0.003
- 0.3
100% Swing Phase
c)
Extension
Net Joint Moment
Interaction Moment
Flexion
RMSE = 0.12 ±0.01
100% Swing Phase
0.2
Extension
Net Joint Moment
0.1
Muscle Moment
0
-o .i
- 0.2
Flexion RMSE = 0.032 ±0.006
- 0.3
100% Swing Phase
d)
0.2
Extension
N et Joint M om ent
Interaction M om ent
o r 0 . 1
^ -0.2
Flexion
RMSE = 0.13 ± 0.02
100% Swing P hase
Figure 6-3. The contribution of muscle moments to net joint moments in 7 year-olds
(a) and adults (b) and of interaction moments to net joint moments in 7 year-olds (c)
and adults (d). All moments normalized to body weight and 100% of the swing
phase.
DISCUSSION
The purpose of this chapter was to test the hypothesis that compared to
adults, 7 year-old children would demonstrate a relatively larger contribution of
muscle moment and a smaller contribution of interaction moment to the net knee
joint moment during the swing phase of gait. This hypothesis was not supported by
our data as the net joint moments as well as the relative contributions of muscle and
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
interaction moments to the net joint moment during the swing phase of gait were
very similar between groups. These results provide evidence that the use of
interaction moments at the knee in 7 year-olds is the same as in adults during swing.
Kinematic and kinetic patterns of 7 year-olds were similar to those of adults
and comparable to published data [26,37]. In both groups, a net extension moment
functioned to decelerate knee flexion in early swing, and a net flexion moment
resulted in deceleration of knee extension in late swing. In general, the muscle
moments followed the same pattern and closely resembled the net joint moment, as
illustrated by the small net-to-muscle RMSE scores for both groups. This indicates
that the muscle moment was largely responsible for the net joint moment. In contrast,
the interaction moment generally opposed the net joint moment, tending to cause
knee flexion in early swing and extension in late swing. This was illustrated by the
larger net-to-interaction RMSE scores. These data suggest that the muscle moment
functioned to dampen the effects of the interaction moment during the swing phase
of gait.
A certain amount of experience is thought to be critical for the acquisition of
adult-like patterns of intersegmental dynamics [23,24,38]. The lack of differences
observed in this study suggest that by 7 years of age, typically developing children
have attained the experience necessary for the production of adult-like patterns of
intersegmental dynamics. Support for this premise can be found in the existing
literature, in that previous reports of the development of intersegmental dynamics
were based on studies of children performing novel tasks. For example, most of the
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
reaching studies were performed with infants early in the first year of life, when this
task was relatively new. Also, the 7-9 year-old children who demonstrated immature
kinetic patterns during obstacle avoidance tasks most likely had limited prior
experience of avoiding surprise obstacles, such as those presented in the laboratory
setting. Future studies are necessary to explore the use of intersegmental dynamics in
children lacking experience with gait, such as younger children or children with
motor delays.
An interesting finding of this study was that neither group appeared to exploit
interaction moments in the manner expected. Based on background literature [33], it
was anticipated that the interaction moment would dominate the net joint moment
and the muscle moment would complement the interaction moment. However, this
was not the case for either group. Instead, the muscle moment dominated the net
joint moment and functioned, essentially, to overcome the interaction moment. Such
a finding suggests that control of the knee during the swing phase of gait is
dependent on muscle action and is consistent with previous EMG studies that have
reported significant muscle action at the knee (particularly in late swing) [26],
One limitation of this study relates to the sample size. Since only 10 children
were evaluated, care must be made in extrapolating the results of this study to all 7
year-old children. However, the similarities between groups suggest that a very large
sample size would be necessary to identify any between-group differences if they did
indeed exist. A second limitation of this study was that only free walking speeds
were evaluated. It is possible that activities involving greater angular accelerations,
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
such as running or kicking, would require greater use of interaction moments and
therefore be more likely to demonstrate age-related differences.
CONCLUSION
The data presented in the current investigation provide evidence that, at least
by 7 years of age, limb intersegmental dynamics during the swing phase of gait are
similar to adults. Furthermore, the absence of differences observed in this study
refutes our hypothesis that children of this age use interaction moments differently
from adults. Future studies examining the development of gait, especially those
focused on the control of intersegmental dynamics, should involve children lacking
opportunities for practice, such as younger children or children with motor delays,
and include activities with higher angular accelerations than those observed in free
walking.
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER VII
SUMMARY AND CONCLUSIONS
Knowledge of normal ambulation in children is essential for recognizing the
effects of pathology and designing effective rehabilitation strategies for children with
gait-related disabilities. Because literature describing gait in children is sparse,
comparisons, for diagnostic and intervention purposes, are often made to adult
normative data. Furthermore, the studies that have looked specifically at children
report conflicting results, especially with regards to the age at which "mature" gait
patterns are expressed. The objective of this dissertation was to determine whether
gait patterns of children are mature by 7 years. In order to address this objective, four
studies designed to analyze the stance and swing phases of gait in 7 year-olds and
adults were conducted.
One limitation of previous work involving kinetic analysis of gait in children
has been the use of anthropometric data collected from older cadavers that may not
be representative of children. Since no data have been reported for the age group of
children involved in the present study, a method for determining age-specific
anthropometric data had to be developed. Chapter III of this dissertation described a
method by which dual energy x-ray absorptiometry (DXA) could be used to quantify
lower extremity anthropometries in-vivo. Using DXA scans obtained from twenty
healthy adults, the mass, center of mass, and moment of inertia of each lower
extremity segment were determined and compared to cadaver-based estimates. In
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
addition, gait analysis was performed on 10 of these individuals. Net joint moments
(sagittal plane) were calculated at the ankle, knee, and hip using anthropometric data
obtained from a) DXA, and b) cadaver-based estimates. Small but statistically
significant differences were identified for DXA-derived and cadaver-based segment
mass values at the thigh and foot and center of mass and moment of inertia values at
the thigh, shank, and foot. However, the similarity between moment curves
computed using the two data sets indicated that the differences in anthropometric
values had little influence on the calculation of gait kinetics. It was concluded that
DXA is an appropriate method for obtaining subject-specific anthropometric data.
In Chapter IV, the DXA method developed in Chapter III was used to
quantify age-specific anthropometric data for 7 year-old children, which were used
in subsequent studies within this dissertation (Chapters V and VI). Specifically, the
purpose of this study was: 1) to compare dual energy x-ray absorptiometry (DXA)-
derived anthropometric parameters to cadaver-based estimates in children, and 2) to
determine if DXA-derived anthropometric data altered gait kinetics in children.
Using DXA, the mass, center of mass location, and moment of inertia of the foot,
shank, and thigh were obtained from 7-13 year-old children (n=50) and compared to
cadaver-based estimates. Additionally, lower extremity net joint moments were
calculated (inverse dynamics equations) for 3 children during gait using a) DXA-
derived, and b) cadaver-based estimates of anthropometric parameters. Statistically
significant differences were identified for DXA-derived and cadaver-based segment
mass values at the thigh and center of mass and moment of inertia values at the thigh,
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
shank, and foot. However, these differences did not appreciably change the
calculation of gait kinetics, as net joint moments calculated with the two sources of
anthropometric parameters were very similar. The results of this study indicated that
cadaver-based anthropometric estimates are satisfactory for use in gait analysis of
healthy children. They also suggest that the use of cadaver-based anthropometric
data do not explain the kinetic differences found in previous investigations of
pediatric gait.
The purpose of the study presented in Chapter V was to determine if sagittal
plane gait kinematics and kinetics of 7 year-old children differed from adults when
age-specific anthropometries were used in the calculation of net joint moments. Joint
angles, moments, and power obtained during level walking in 7 year-old children
(n=15) were compared to data from adults (n=15). Kinetic calculations were
performed using age-specific anthropometric data obtained from dual energy x-ray
absorptiometry. For most of the variables examined, 7 year-olds were similar to
adults; however children demonstrated a diminished peak plantarflexor moment and
less peak power absorption and generation at the ankle during late stance. Post-hoc
analyses revealed that 64% of the variance in the plantarflexor moment could be
explained by foot length, suggesting that physical factors, rather than neuromuscular
immaturity, may explain differences at the ankle. These results provide evidence that
gait kinematics and kinetics of 7-year old children are comparable to those of adults;
however consideration should be given to normalizing ankle moments by foot length
as this physical parameter may influence ankle kinetics (including power).
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The purpose of the study presented in Chapter VI was to test the hypothesis
that compared to adults, 7 year-old children would demonstrate a relatively larger
contribution of muscle moment and a smaller contribution of interaction moment to
the net knee joint moment during the swing phase of gait. Net knee joint moments as
well as the contributions of muscle, interaction, and gravity moments to the net joint
moment were compared between 7 year-old children (n=10) and adults (n=10)
during self-selected free walking. The knee kinematics and net joint moments were
similar between the two groups. Also, the contributions of the muscle and interaction
moments to the net joint moment were similar (P > 0.05), suggesting that the ability
to employ interaction torques during the swing phase of gait did not differ between
the two groups. These results provide evidence that, at least by 7 years, the control of
limb intersegmental dynamics during swing is similar to adults. Furthermore, these
results discount the theory that children of this age lack the neuromuscular maturity
to produce an adult-like gait pattern.
In conclusion, the results of this dissertation support the premise that walking
patterns in 7 year-old children are mature. This was reflected in the kinematic and
kinetic similarities between 7 year-old children and adults. For example, it was
demonstrated that 7 year-olds produced kinematic patterns that were similar to those
produced by adults during the stance and swing phases of gait. Furthermore, through
the use of age-specific anthropometric data for kinetic calculations, it was
demonstrated that overall kinetic patterns (net joint moments and power) of 7 year-
old children were similar to those of adults during the stance phase of gait. Lastly, a
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
detailed kinetic analysis of the swing phase of gait revealed that the coordination of
lower limb intersegmental dynamics were similar between 7 year-old children and
adults. Together, these results provide evidence that adult-like gait patterns are
present, and not still emerging, in children 7 years of age.
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
REFERENCES
[1] Ackland TR, Blansksby BA, Bonjour JP. Inertial characteristics of adolescent
male body segments. Journal of Biomechanics 1988;21:319-327.
[2] Andrews JG, Mish SP. Methods for investigating the sensitivity of joint
resultants to body segment parameter variations. J Biomech 1996;29:651-654.
[3] Bernstein N. The Coordination and Regulation of Movement. 1967;
[4] Bresler B, Frankel JP. The forces and moments in the leg during level walking.
Trans ASME 1950;72:27-36.
[5] Chandler RF, Clauser CE, McConville JT, Reynolds HM, Young JW.
Investigation of the inertial properties of the human body. 1975;
[6] Cheng CK, Chen HH, Chen CS, Lee CL, Chen CY. Segmental inertial
properties of Chinese adults determined from magnetic resonance imaging.
Clinical Biomechanics 2000;15:559-566.
[7] Clauser CE, McConville JT, Young JW. Weight, volume and center of mass of
segments of the human body. 1969;60-70.
[8] Cupp T, Oeffinger D, Tylkowski C, Augsburger S. Age-related kinetic changes
in normal pediatrics. J Pediatr Orthop 1999;19:475-478.
[9] De Leva P. The Mass and Inertia Characteristics of the Main Segments of the
Human Body. Journal of Biomechanics 1996;29:1223-1230.
[10] Dempster WT. Space requirements of the seated operator. 1955;55-159.
[11] Ganley KJ, Powers CM. The use of dual energy x-ray absorptiometry in
determining subject-specific anthropometric measures for kinetic analysis
during gait. Gait and Posture 2001;13:271-272.
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[12] Ganley KJ, Powers CM. Are cadaver-based estimates of body segment
parameters appropriate for calculating net joint moments in children? Gait and
Posture 2002;16:S69-S70.
[13] Hatze H. A mathematical model for the computational determination of
parameter values of anthropomorphic segments. J Biomech 1980;13:833-843.
[14] Hoy MG, Zemicke RF. Modulation of limb dynamics in the swing phase of
locomotion. J Biomech 1985;18:49-60.
[15] Huang HK, Suarez FR. Evaluation of cross-sectional geometry and mass
density distributions of humans and laboratory animals using computerized
tomography. Journal of Biomechanics 1983;16:821-832.
[16] Jensen RK. Body segment mass, radius and radius of gyration proportions of
children. J Biomech 1986;19:359-368.
[17] Jensen RK. Changes in segment inertia proportions between 4 and 20 years. J
Biomech 1989;22:529-536.
[18] Jensen RK, Fletcher P. Distribution of mass to the segments of elderly males
and females. J Biomech 1994;27:89-96.
[19] Kadaba MP, Ramakrishnan HK, Wootten ME. Measurement of lower
extremity kinematics during level walking. J Orthop Res 1990;8:383-392.
[20] Kelly TL, Berger N, Richardson TL. DXA Body Composition: Theory and
Practice. Applied Radiation Isol 1998;49:511-513.
[21] Kingma I, Toussaint HM, De Looze MP, Van Dieen JH. Segment inertial
parameter evaluation in two anthropometric models by application of a
dynamic linked segment model. J Biomech 1996;29:693-704.
[22] Kohrt WM. Preliminary evidence that DEXA provides an accurate assessment
of body composition. Journal of Applied Physiology 1998;84:372-377.
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[23] Konczak J, Borutta M, Topka H, Dichgans J. The development of goal-directed
reaching in infants: hand trajectory formation and joint torque control. Exp
Brain Res 1995;106:156-168.
[24] McFadyen BJ, Malouin F, Dumas F. Anticipatory locomotor control for
obstacle avoidance in mid-childhood aged children. Gait Posture 2001;13:7 -
16.
[25] Ounpuu S, Gage JR, Davis RB. Three-dimensional lower extremity joint
kinetics in normal pediatric gait. J Pediatr Orthop 1991;11:341-349.
[26] Perry J. Gait Analysis: Normal and Pathological Function. Thorofare, NJ:
Slack Inc.,1992.
[27] Phillips SJ, Roberts EM, Huang TC. Quantification of intersegmental reactions
during rapid swing motion. J Biomech 1983;16:411-417.
[28] Piazza SJ, Delp SL. The influence of muscles on knee flexion during the swing
phase of gait. J Biomech 1996;29:723-733.
[29] Rosenbaum D. Human Motor Control. San Diego, CA: Academic Press, 1991.
[30] Schmidt R. A schema theory of discrete motor skill learning. Psychol Review
1975;82:225-260.
[31] Schneider K, Zemicke RF. Mass, center of mass, and moment of inertia
estimates for infant limb segments. J Biomech 1992;25:145-148.
[32] Schneider K, Zemicke RF, Schmidt RA, Hart TJ. Changes in limb dynamics
during the practice of rapid arm movements. J Biomech 1989;22:805-817.
[33] Schneider K, Zemicke RF, Ulrich BD, Jensen JL, Thelen E. Understanding
movement control in infants through the analysis of limb intersegmental
dynamics. Journal of Motor Behavior 1990;22:493-520.
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
[34] Slosman DO, Casez JP, Pichard C, Rochat T, Fery B, Rizzoli R, Bonjour JP,
Morabia A, Donath A. Assessment of whole-body composition with dual
energy x-ray absorptiometry. Radiology 1992;185:593-598.
[35] Sutherland DH. The development of mature gait. Gait and Posture 1997;6:163-
170.
[36] Sutherland DH, Cooper L, Daniel D. The role of the ankle plantar flexors in
normal walking. J Bone Joint Surg Am 1980;62:354-363.
[37] Sutherland DH, Olshen R, Cooper L, Woo SL. The development of mature
gait. J Bone Joint Surg [Am ] 1980;62:336-353.
[38] Thelen E, Corbetta D, Kamm K, Spencer JP, Schneider K, Zemicke RF. The
transition to reaching: mapping intention and intrinsic dynamics. Child Dev
1993;64:1058-1098.
[39] Todd FN, Lamoreux LW, Skinner SR, Johanson ME, St Helen R, Moran SA,
Ashley RK. Variations in the gait of normal children. A graph applicable to the
documentation of abnormalities. J Bone Joint Surg Am 1989;71:196-204.
[40] Winter DA. Biomechanics and Motor Control of Human Movement. 2n d ed.
New York: John Wiley & Sons, Inc., 1990.
[41] Woltring HJ. A FORTRAN package for generalized cross-validatory spline
smoothing and differentiation. Advances in Engineering Software 1986;8:104-
113.
[42] Zatsiorsky V, Seluyanov V. The Mass and Inertia Characteristics of the Main
Segments of the Human Body. Biomechanics 1983 ;VIII B: 1152-1159.
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix A: Equations of Motion
Net Joint Moment = la
Interaction =
[B4c o s ( © s - © t ) + B 5 c o s ( © s - © t ) + B ic o s ( © f - © t ] © t (aaT )
+[B4sin(©s-©i)+B5sin(©s-©T)+ Bisin(©F -©T )]©T 2 (avT )
+ [B 6+ B 2c o s ( © f - © s ) ] © s (aas)
+[B2sin(©F -©s)]©s2 (avT )
+[IF +(mF rF 2 )+B2 cos(©F-©s)]©F (aaF )
-[B2sin(©p-©s)]©F 2 (avj)
-[(B3sin©F+B7 sin©s)X"]-[(B3 cos©F +B7 cos©s)Y] (hla)
Gravity =
-[B3COS©F+B7COS©s]g
Muscle = Net Joint Moment - Interaction - Gravity
Terminology:
aai, aas, aaF angular acceleration (thigh, shank, foot)
avj, avs, avF angular velocity (thigh, shank, foot)
hla hip linear acceleration
i u t , ms, mF mass (thigh, shank, foot)
r F . rs, rf distance from the proximal joint center to center of
mass (thigh, shank, foot)
It, Is, If length (thigh, shank, foot)
I t , I s , If moment of inertia at center of mass (thigh, shank, foot)
© t , © s, © f orientation angles at proximal end of segment from
right horizontal (thigh, shank, foot)
© t, © s, © f angular velocity of © t, © s, © f
© t , © s, © i angular acceleration of © t, © s, © f
X, Y linear acceleration of hip joint center in anterior (X)
and vertical (Y) directions
g gravitational constant (9.8 m/sec2)
Bi mF rF lT
B 2 mF rF ls
B 3 mF rF
B 4 m [1 s 1 ' i
B 5 msrslT
1 ^ 6 nipls2
B 7 mF ls+ msrs
Bs mslT2
B 9 mF lT 2
Bio m^T + m s^ + mF lT
B i 1 u l s 1 t + m p lT + m TrT
96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Human and environmental factors contributing to slip events during walking

PDF

Biomechanical and neuromuscular aspects of non-contact ACL injuries: The influence of gender, experience and training

PDF

Factors contributing to patellofemoral joint stress: a comparison of persons with and without patellofemoral pain

PDF

Assessing the cost implications of combined pharmacotherapy in the long term management of asthma: Theory and application of methods to control selection bias

PDF

Control and dynamics during horizontal impulse generation

PDF

Assessment of fatigue as a late effect of therapy among survivors of childhood leukemia

PDF

Characterization of neural mechanisms involved in hypoglycemic detection at the portal vein

PDF

Cardiovascular fitness, exercise, and metabolic disease risk in overweight Latino youth

PDF

Bimanual coordination of goal-directed multi-joint rapid aiming movements following stroke

PDF

The role of the vasti in patellar kinematics and patellofemoral pain

PDF

Cigarette smoking, metabolic syndrome, and subclinical atherosclerosis

PDF

Identification of oncogenes cooperating in murine mammary tumorigenesis and transgenic mouse models of breast cancer

PDF

A study of employee health plan choice and medical cost: Panel data probit regression and sample selection model

PDF

Expression and regulation of B7 on antigen presenting cells

PDF

Design considerations in the application of membrane reactors for use in steam-carbon dioxide reforming environments

PDF

Endogenous regulatory factors in the inhibition and down-regulation of immune responses

PDF

A comparison of the play performance of boys with autism and that of boys without disabilities in Taiwan

PDF

Impact of housing arrangements on social support and health status among Chinese American elderly

PDF

Functional brain correlates for premovement planning and compensatory adjustments in rapid aimed movement

PDF

Facing the camera: Dickens, photography, and the anxiety of representation

Asset Metadata

Creator Ganley, Kathleen Jodell (author)

Core Title At what age are gait characteristics mature? Evaluation of gait kinematics, kinetics, and intersegmental dynamics in 7 year-old children

School Graduate School

Degree Doctor of Philosophy

Degree Program Biokinesiology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag health sciences, recreation,health sciences, rehabilitation and therapy,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Powers, Christopher (committee chair), [illegible] (committee member), Gordon, James (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-363808

Unique identifier UC11334984

Identifier 3103890.pdf (filename),usctheses-c16-363808 (legacy record id)

Legacy Identifier 3103890.pdf

Dmrecord 363808

Document Type Dissertation

Rights Ganley, Kathleen Jodell

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA