619
ORIGINAL ARTICLE
Effects of a Postoperative Strength-Training Program
on the Walking Ability of Children With Cerebral Palsy:
A Randomized Controlled Trial
Dimitrios Patikas, PhD, Sebastian I. Wolf, PhD, Katrin Mund, PT, Petra Armbrust, PT,
Waltraud Schuster, MTA, Leonhard Döderlein, MD
ABSTRACT. Patikas D, Wolf SI, Mund K, Armbrust P,
Schuster W, Döderlein L. Effects of a postoperative strength-
training program on the walking ability of children with cere-
bral palsy: a randomized controlled trial. Arch Phys Med
Rehabil 2006;87:619-26.
Objective: To investigate the effect of a postoperative
strength-training program on the walking of children with
cerebral palsy (CP).
Design: Randomized controlled trial.
Setting: Hospital rehabilitation department.
Participants: Thirty-nine children with CP (age range, 6⫺
16y). After orthopedic surgery, the control group (n⫽20) fol-
lowed a conventional physiotherapy (PT) program, and the
strength-training group (n⫽19) followed a strength-training
program in addition to the conventional PT. Twenty-nine age-
matched healthy children were used as references.
Intervention: A 9-month strength-training program.
Main Outcome Measures: Spatiotemporal, kinematic, and
kinetic parameters during gait analysis were analyzed before
(E0) and 1 year after (E1) the surgery. For 22 children, a 2-year
postoperative gait analysis (E2) took place as well.
Results: At E1, several kinematic and kinetic parameters
improved, although there was no significant difference between
the groups. Spatiotemporal parameters showed a worsening at
E1 and a recovery to preoperative values at E2.
Conclusions: The examined parameters may be more sub-
stantially influenced by factors such as the surgery outcome
and the variability of pathologic characteristics than by the
strength-training program per se. However, a more significant
effect of the strength-training may appear if more intense and
short-term training protocols are used, considering factors such
as patients’ motivations, ages, and postoperative statuses.
Key Words: Cerebral palsy; Exercise; Gait; Rehabilitation;
Surgery.
© 2006 by the American Congress of Rehabilitation Medi-
cine and the American Academy of Physical Medicine and
Rehabilitation
From the Department of Orthopaedic Surgery, University of Heidelberg, Germany.
Presented in part to the European Society for Movement Analysis of Adults and
Children, 2004, Warsaw, Poland.
Supported by the Department for Science Research and Arts of the State of
Baden-Wuerttemberg and Stifterverband für die Deutsche Wissenschaft.
No commercial party having a direct financial interest in the results of the research
supporting this article has or will confer a benefit upon the authors or upon any
organization with which the authors are associated.
Reprint requests to Dimitrios Patikas, PhD, Dept of Orthopaedic Surgery, Univer-
sity of Heidelberg, Schlierbacher Landstr 200a, 69118 Heidelberg, Germany, e-mail:
Dimitrios.Patikas@ok.uni-hd.de.
0003-9993/06/8705-10527$32.00/0
doi:10.1016/j.apmr.2006.01.023
EREBRAL PALSY (CP) IS a disorder of movement and
C posture caused by a nonprogressive defect or lesion of the
1
immature brain. The movement deviation of children with CP
is attributed to abnormal motor control, restricted range of
motion (ROM), and insufficient force production caused by
selective muscle atrophy.2
Surgical intervention is often used as treatment in patients
with CP, especially during the developmental ages. Excessive
hip and knee flexion (crouch gait), hip adduction and internal
rotation, and equinus feet are typical symptoms of children
with CP that are often indications for surgical intervention.3
The aim of surgery is to improve the joint ROM and to adjust
skeletal deformities for a more beneficial function during ev-
eryday activities. Furthermore, to restore normal muscle length
and improve muscle balance, myotomies and tendon lengthen-
ings and transfers may be undertaken, which would further
weaken the already weak muscles.4,5
Despite the traditional belief that strengthening muscles
causes increased spasticity levels,6 after strength-training pro-
grams children with CP can improve in strength without any
significant negative effects on the spasticity level of the mus-
cles.7 There are numerous studies8-13 that have shown muscle
strength gains in children with CP. Additionally, increasing
muscular strength correlates with improvement in several func-
tional parameters.5,7,14-19 Therefore, strength training has been
suggested as an additional treatment option for this patient
group.
Several training methods have been recommended for
children with CP. Training protocols using isokinetic devices
permitting a well-controlled and optimal loading of the joints
through the whole ROM20 have been shown to be beneficial in
children with CP.7,10,13,15 However, isokinetic devices are not
accessible for the vast population of patients. Furthermore, the
isokinetic movement rarely relates to any functional move-
ment and therefore, according to the training specificity prin-
ciple,21 it is more unlikely for training benefits to be transferred
into activities such as walking. For this reason more recent
studies have concentrated on home-based strength-training pro-
grams without the requirement of excessive and costly equip-
ment.5,8,14 Moreover, most training programs lasted 6 to 8
weeks and it has been argued that a 6-week training program
for children with CP may be not long enough to induce func-
tional adaptation during walking.9 To our knowledge, there is
no long-term study on the effects of a functional strength-
training program as an additional postoperative rehabilitation
tool that was aimed at shortening the rehabilitation time and
that has resulted in functional improvement and correcting gait
pattern. The purpose of this study was to examine the effect of
a postoperative long-term strength-training program in addition
to physiotherapy (PT) in children with CP, with minimal re-
quirements in equipment and costs and involving exercises and
movements that are important in everyday life.
Arch Phys Med Rehabil Vol 87, May 2006
620 POSTOPERATIVE EFFECT OF STRENGTH TRAINING, Patikas

METHODS used as normative references (mean age, 10.1⫾2.8y; height,

Participants
Patients aged between 6 and 16 years old, diagnosed with
spastic diplegia, and having an indication for surgical treatment
were included in the present study. All participants were out-
patients of the Department of Orthopaedic Surgery (University
of Heidelberg, Germany) and their parents agreed to their
participation in the study by signing the informed consent. The
study was approved by the local ethics committee.
After the operation patients were randomly divided (sealed
envelope method without replacement) into 2 groups (fig 1):
(1) the control group, which followed a typical PT program
after the operation and (2) the strength-training group, which
followed a strength-training program in addition to PT. The PT
program was performed 2 to 3 times a week and included exer-
cises to improve gait and to restore the ROM of the lower-limb
joints. Twenty children were evaluated for the control group
(mean age ⫾ standard deviation [SD], 8.9⫾1.9y; height, 128⫾
11cm; body mass, 28.5⫾8.9kg) and 19 for the strength-training
group (mean age, 10.6⫾3.2y; height, 135.6⫾15.5cm; body
mass 32.5⫾12.5kg). Before surgery, 2 children from the
strength-training group and 7 from the control group required
external assistance for walking (walker or crutches). The gait
analysis data of 29 healthy children from 6 to 16 years old were
Assessed for eligibility
(n=47)
Enrollment
142.3⫾16.3cm; body mass, 38.1⫾13.7kg).
Experimental Design
Two to 4 days after surgery, all patients started performing a
regular PT program to re-establish the desired ROM of the
lower-limb joints. At the same time, the strength-training pro-
gram was begun explicitly for the strength-training group.
During the hospital stay (duration, 1.8⫾0.5mo), physiothera-
pists supervised the strength-training program. At home, the
children performed the exercises with their parents. Children
and parents filled in personal diaries to determine the frequency
of the training sessions. The physiotherapists who treated pa-
tients during the regular PT session were well informed about
the procedures of the training and were able to answer any
potential questions that the patients had.
All patients were examined before the surgery, and this
examination (E0) was used as reference. At E0, 3-dimensional
gait analysis (kinematics, kinetics), video and clinical exami-
nation (including evaluation of the spasticity level using the
Modified Ashworth Scale [MAS]), gross motor function mea-
surement (using the Gross Motor Function Measure [GMFM]),
and estimation of oxygen consumption and walking efficiency
were assessed. Nine months after hospital discharge, which
designated the end of the strength-training program for the
strength-training group (.93⫾.06y postsurgery), all patients
were evaluated using the same tests (E1). Additional gait analysis
and video and clinical examinations were performed 2.14⫾
0.30 years after the surgery (E2). Between E1 and E2, the
strength-training– group children did not perform any strength-
training sessions. The decision for the E2 measurements was

retrospective, and therefore it was possible to measure only 22

Excluded (n=4) of the 39 patients. Fourteen of these patients were able to walk
• Mentally incapable to without support at E0, E1, and E2.
follow the study (n=3)
• Did not consent (n=1) Surgical Treatment
All patients underwent a single-event multilevel surgical
Randomized (n=43) treatment on both legs.22 The decision for operation was made
after analyzing data from video, clinical examination, 3-dimen-
sional gait analysis, and dynamic electromyography. The sur-
gery included 271 soft-tissue (control group, n⫽144; strength-
Allocation

Control Group Strength Training

(CG: n=21) Group (SG: n=22) training group, n⫽127) and 121 bony procedures (control
• Received regular • Received regular group, n⫽56; strength-training group, n⫽65). A more detailed
Physiotherapy Physiotherapy and description of the procedures and content of the surgery is
Strength training available from the authors (D Patikas et al, unpublished data,
2006).
Excluded
• Intraoperative Training Protocol
Follow-up

E1 follow-up vessel injury (n=1) The training protocol consisted of 7 exercises involving the
unattended
• Parents argued
hip, knee, and ankle extensors and flexors. If children suc-
Drop outs ceeded in overcoming the resistance against gravity, elastic
that the testing • Discontinued the
protocol was too intervention due to bands were used to increase resistance.a Two sets of 5 repeti-
stressful (n=1) unplanned tions were performed for each exercise with a low velocity to
obligations (n=2) permit movement control and eccentric activation of the mus-
cles. A more detailed description of the training protocol con-
Analyzed (n=20) Analyzed (n=19) Group tent is available from the authors (D Patikas et al, unpublished
• Missing kinetic data • Missing kinetic data comparison data, 2006).
during
Analysis

(n=11): not free wal- (n=4): not free walker The training sessions started 3 to 4 weeks after surgery.
ker during E0 or E1 during E0 or E1 E0 - E1 After hospital discharge, children in the strength-training group
were trained for 40.3⫾0.4 additional weeks, until E1. Each
session lasted approximately 30 to 45 minutes and was per-
Analyzed (n=22, CG: n=9 , SG: n=13)
E0 - E1 - E2 formed on average 3.2⫾0.3 times a week.
• Missing kinetic data (n=8, SG: n=4, CG: n=4): not comparison
free walker during E0 or E1 or E2 Gait Analysis
A 3-dimensional gait analysis was performed using the Helen
Fig 1. Flow diagram of the randomization procedure. Hayes marker set.23 Data were captured with a 9-camera

Arch Phys Med Rehabil Vol 87, May 2006

 POSTOPERATIVE EFFECT OF STRENGTH TRAINING, Patikas
3-dimensional motion analysis systemb using reflective mark-
ers (diameter, 14mm) applied on the skin. The children were
asked to walk over a 9-m line at self-selected speeds. Kinetic
data were captured using 2 forceplatesc and only in case a
patient required no external body support during walking.
Gross Motor Function Measurement
GMFM scores24 were assessed at the E0 and E1 examina-
tions. The overall and individual scores of the 5 GMFM sec-
tions (A, lying and rolling; B, sitting; C, crawling and kneeling;
D, standing; E, walking, running, and jumping) were evaluated
and the results were expressed as a percentage of the maximum
possible score. A mean of all sections represented the total
score, and sections D and E were analyzed separately.5,8
Walking Efficiency
The heart rate, breathing frequency, and oxygen consump-
tion (V̇O2) were measured online.d Patients walked twice for 5
minutes around a 25-m–long pathway (8⫺12 rounds for each
5-min walk, depending on walking speed). After each 5-minute
walk, children rested while sitting for 2 to 5 minutes, until the
resting heart rate was reached. The resting V̇O2 was subtracted
from the V̇O2 during walking and this value was normalized to
the body mass and distance that each patient walked. Further-
more, the Energy Expenditure Index (EEI) was calculated
using the following formula:
EEI⫽
walking velocity
冉
HRwalking⫺HRrest heart beats ⁄ min
m ⁄ min
冊 One year after the surgery (E1), the spatiotemporal param-
where HRwalking and HRrest are mean heart rates in beats per
minute during walking and at rest, respectively.25 The strong
linear relation between oxygen uptake and heart rate in differ-
ent velocities supports the use of the EEI as a walking efficiency
indicator.25-27
Data and Statistical Analysis
Means and SDs of the averages from both sides were cal-
culated for all dependent variables. The spatiotemporal, kine-
matic and kinetic parameters and the normalcy index28 were
processed with tools developed in Matlab.e To eliminate the
effect of differences between the groups before surgery, a
univariate analysis of covariance (ANCOVA) was assessed for
each dependent variable to determine the differences between
the groups at E1, using as a covariate the E0 measurement.
One-way analysis of variance (ANOVA) for repeated measure-
ments was used to determine changes between E0, E1, and E2.
The significance level was set at .05, and the 95% confidence
interval of the difference between the compared groups or
measurements was calculated. All statistical analyses were
performed using SPSS.f
RESULTS
Clinical Data, GMFM, and Walking Efficiency
No difference between the groups was observed in the spas-
ticity level at E0 and E1, as measured with the MAS (table 1).
However, a significant decrease in spasticity between E1 and
E0 was observed for both groups.
The GMFM score of section D (related to standing) im-
proved significantly at E1, and no difference between the
groups was observed when E0 intergroup differences were
taken into account.
From the 13 control-group and 16 strength-training– group
children who performed the walking efficiency test, the V̇O2
621
Table 1: Results of the Statistical Analysis for the MAS, GMFM,
V̇O2, and EEI for the 2 Groups Before and 1 Year After Surgery
95% Confidence Interval for Differences
Between
Measures CG–SG* E1–E0†
MAS score (0⫺4 nominal
scale) ⫺0.102 to 0.205 ⴚ0.598 to ⴚ0.363
GMFM section D score
(% of maximum) ⫺11.5 to 1.4 0.4 to 10.9
GMFM section E score
(% of maximum) ⫺13.9 to 3.7 ⫺7.4 to 1.6
GMFM total score (% of
maximum) ⫺6.6 to 0.7 ⫺1.3 to 3.2
V̇O2 (mL·kg⫺1·m⫺1) ⫺0.125 to 0.083 ⫺0.051 to 0.070
EEI (heart beats/m) ⫺0.617 to 0.120 0.050 to 0.428
NOTE. Significant differences are in boldface.
Abbreviations: CG, control group; SG, strength-training group.
*Univariate ANCOVA with the E0 measurement as covariate.
†
One-way ANOVA for repeated measurements (E0 and E1).
did not change significantly between the E0 and E1 measure-
ments or between groups. The EEI increased significantly at
E1, and this change was not group dependent.
Gait Parameters
eters of both groups deviated more from the healthy children as
compared with the preoperative measurements (E0) (fig 2).
This fact, combined with the increased EEI at E1, led us to
decide to measure as many children as possible 2 years after the
operation. Because we observed no significant differences be-
tween the groups in the spatiotemporal, kinematic, and kinetic
parameters (table 2), the 2 groups were merged to calculate the
differences between E0, E1, and E2. Walking speed, duration
of the stance phase, and stride duration improved significantly
during the time period between E1 and E2.
The normalcy index was significantly reduced at E1, and a
further decrease followed at E2 (fig 3). The control group had
a higher normalcy index before the surgery (see table 2). When
we adjusted the preoperative differences between the groups,
we observed no significant difference in the normalcy index
between the groups.
Concerning the hip joint, 59% of the examined lower ex-
tremities did not reach full hip extension at E0, and this deficit
was retained after the operation for both groups. Preopera-
tively, the minimum knee flexion angle at terminal swing was
increased for both groups with respect to the normative values.
One year postsurgery, both groups had a significant decrease in
the knee flexion angle, approaching the normative values. This
decrease remained significant at E2, despite the observed in-
crease between E1 and E2. The ankle joint during the stance-
swing transition period was in the equinus position, with the
control group having a higher plantarflexion angle at E0. The
increased plantarflexion preoperatively was drastically reversed at
E1 for both groups and tended to increase toward the normative
values at E2.
Concerning the kinetic data (fig 4), it was possible to capture
data from 9 control-group and 15 strength-training– group chil-
dren at E0 and E1 and from 14 children at E0, E1, and E2 (see
also fig 1). The power absorption at the hip joint was decreased
at E0 and increased gradually toward the normative values at
E1 and E2 for both groups. We observed no significant differ-
ence between the groups or the examinations for the maximum
knee power absorption during the loading response. Concern-
Arch Phys Med Rehabil Vol 87, May 2006
622 POSTOPERATIVE EFFECT OF STRENGTH TRAINING, Patikas

Fig 2. (A—D) Spatiotemporal parameters for the control (CG) and strength-training (SG) groups, before (E0), 1 year after (E1), and 2 years
after (E2) the surgery. The normative values with SD are represented by the continuous horizontal line and the grey zone (nⴝ29). NOTE.
Values are mean ⴞ SD. *Significant differences between the measurements for the merged group (CGⴙSG) that were examined at E0, E1,
and E2.

ing the ankle joint kinetics, we observed an improvement only
between E1 and E2. There was a trend to decrease between E0
and E1, whereas the changes in moment and power of the ankle
joint were not group dependent.
DISCUSSION
It is well documented that the gait patterns of patients with
CP deviate from normal, taking into consideration gait anal-
ysis data.29 For instance, decreased power production from the
plantarflexors of children with hemiplegia during walking has
been previously reported,30 and our study supports this as well.
Despite the improvement in the ankle joint kinematics at E1, the
moment and power of the plantarflexors tended to decrease. This
could be attributed to the decreased walking speed, which affects
the power production.31 In addition, the increased duration of the
double-support phase could also explain why the ankle moment—
and consequently power— decreased at E1. The earlier-mentioned
parameters recovered between E1 and E2. This shows that al-
though joint ROM was restored at E1, it took almost 2 years to
restore the preoperative power when walking. Furthermore, our
study showed that 1 year after the surgery the EEI increased.
Consequently, an interval longer than 1 year postsurgery is re-
quired to describe the long-term outcome of the surgery. This is in
line with a very recent report32 that suggests that the gait evalu-
ation should be performed at least 3 years after surgery to give the
most predictive outcome of the treatment. The same research
group had previously reported that triceps surae lengthening led to
significant increase in the plantarflexor moment 3 years after the
operation.33
The group that followed a specific strength-training program
in addition to regular PT during the rehabilitation period failed
to show any significant advantages compared with the group
that had only conventional PT. This contrasts with previous
studies that have documented well the advantages of strength
training in patients with CP. However, there are several expla-
nations for these contradictory conclusions.

Arch Phys Med Rehabil Vol 87, May 2006

 POSTOPERATIVE EFFECT OF STRENGTH TRAINING, Patikas 623

Table 2: Results of the Statistical Analysis for the Gait Analysis Data
95% Confidence Interval for Differences Between

Measures CG–SG* E1–E0† E2–E0† E2–E1†

Walking speed (cm/s) ⫺13.3 to 21.9 ⫺20.7 to 51.2 ⫺7.5 to 0.1 6.0 to 51.2
Stance phase duration (% of gait cycle) ⫺4.51 to 2.65 0.21 to 4.96 ⫺3.38 to 0.79 ⴚ5.65 to ⴚ2.11
Stride duration (s) ⫺0.49 to 0.29 0.04 to 0.55 ⫺0.10 to 0.23 ⴚ0.44 to ⴚ0.02
Stride length (cm) ⫺7.1 to 14.8 ⫺7.8 to 5.6 ⫺3.3 to 11.7 ⫺1.2 to 11.8
Normalcy index ⫺91 to 170 ⴚ168 to ⴚ27 ⴚ227 to ⴚ106 ⴚ129 to ⴚ10
Maximum hip extension (deg) ⫺3.4 to 5.0 ⫺4.3 to 1.9 ⫺5.4 to 1.6 ⫺3.6 to 2.2
Minimum knee flexion during terminal swing (deg) ⫺2.9 to 8.4 ⴚ23.2 to ⴚ12.8 ⴚ18.8 to ⴚ10.3 0.4 to 6.6
Maximum plantarflexion during stance-swing
transition (deg) ⫺3.0 to 5.7 9.1 to 20.0 5.7 to 17.4 ⫺6.3 to 0.3
Maximum hip power absorption (W/kg) ⫺0.240 to 0.192 ⫺0.271 to 0.041 ⴚ0.592 to ⴚ0.128 ⴚ0.419 to ⴚ0.072
Maximum knee power absorption during loading
response (W/kg) ⫺0.217 to 0.392 ⫺0.240 to 0.115 ⫺0.492 to 0.215 ⫺0.319 to 0.167
Maximum plantarflexion moment (Nm/kg) ⫺0.131 to 0.161 ⫺0.215 to 0.003 ⫺0.080 to 0.087 0.032 to 0.186
Maximum plantarflexion power generation (W/kg) ⫺0.880 to 0.050 ⫺0.609 to 0.088 ⫺0.210 to 0.620 0.166 to 0.765

NOTE. Significant differences are in boldface.

*Univariate ANCOVA with the E0 measurement as covariate.
†
One-way ANOVA for repeated measurements (E0 and E1).

Most of the studies in the past have used more intense and
short-term training protocols, with regular supervision,12,14 which
nonetheless required children to make scheduled visits to the
rehabilitation center and in some cases to use special equip-
ment, such as isokinetic dynamometers.7,10,13,15 This kind of
approach is probably ideal for research purposes, but because
of time and cost limitations most patients are not able to follow
such a training program for a long or even short period. For this
reason, and in accordance with more recent studies,5,8,14 a
home-based training program was suggested in the present
study. However, when practicing at home the research team has
less control de facto over the frequency and intensity of train-
ing. Dodd et al8 indicated that the lack of changes in function-
ally important gait parameters, such as walking speed, after
strength training may be attributed to the fact that training
sessions were assessed at home and that the minimized thera-
pist supervision could have resulted in a lower-than-optimal
training intensity. Our results support this statement.
The strength-training protocols in most of the previous stud-
ies, which have been effective for children with CP, had
durations of 6 to 8 weeks. Damiano et al9 recommended that
the strength-training duration should be longer than 6 weeks to
be effective. Studies of patients with spinal cord injuries also
suggest long-term strength training.34 In the present study, the
training program was long term (⬇40wk) to cover the time
period after the operation when children show a loss of muscle
force production (D Patikas et al, unpublished data, 2006).
However, there are some objective difficulties in making such
a home-based protocol as intense as possible during this long
time period, and therefore it cannot be ruled out that the moti-
vation of some children may have decreased while performing
the same training protocol. This cannot rule out the possibility
that subjects were “underdosed” to strength training and did not
get stronger. Furthermore, the children’s mental maturity may
also influence the outcome of training. The age of the children
in the present study ranged from 6 to 16 years old. This
imposes a noteworthy variability in the way they understand
the aspect of strength training (as a part of the treatment or as
an obligation for their parents). Such factors may increase the
variance within the strength-training group, suppress the con-
trast between strength-training and control groups, and make
comparisons between different studies difficult. Methods to
increase children’s motivation (like exercises in the form of
games) are necessary to enhance the tolerance of the strength
training. It is probable that if in the present study the children
had been able to choose into which group they wanted to be
placed, the effects of strength training might have been more
pronounced. Moreover, it should not be ruled out that some
children belonging to the control group could be physically
more active than some children that belonged to the training
group. It was not possible to control such factors, because it
would have been ethically inappropriate to force the children
who participated in the study as the control group to be phys-
ically inactive.
Another factor that might have limited the training intensity
in the present study is the nature of the performed exercises,
which involved mainly more than 1 joint and were associated
with functionally important everyday activities. Although the
principle of specificity is an important issue for training in the
healthy population, low levels of motor control may not permit
the training effects to be transferred to functional tasks like
gait.9 Execution of these exercises by patients with CP requires
an ordinary level of motor control that may not be available.
Previous studies increased the homogeneity of the examined
groups by excluding more severely affected patients and taking
into consideration specific gait parameters.8 In the current
study no such exclusion criterion took place, probably resulting
in the fact that some patients were not able to execute their
training programs as efficiently as others. Another consequence
of using multijoint exercises in patients with CP is that the use
of greater load to increase the training intensity is restricted,
compared with studies that used single-joint movements with
maximal resistance.10 In our study, the possibility of increasing
training intensity after the surgery was limited, because most
children were able to execute the exercises moving against
gravity at the beginning of the training period (sometimes even
with assistance) and gradually were able to use external resis-
tance with elastic bands.
Only a limited number of studies examined the effect of
strength training in children compared with a control group that
did not perform strength training.8,10 In the most recent ran-
domized controlled study,8 there was only a trend for the
strength-training group to improve in the walking speed and
GMFM parameters when compared with the control group.

Arch Phys Med Rehabil Vol 87, May 2006

624 POSTOPERATIVE EFFECT OF STRENGTH TRAINING, Patikas

Fig 3. Normalcy index and kinematic parameters for the control and strength-training groups, prior to, 1 year after, and 2 years after surgery.
(A) Normalcy index, (B) minimum knee flexion during terminal swing, (C) maximum hip extension, and (D) maximum plantarflexion during
stance to swing transition. The normative values with SD are represented by the continuous horizontal line and the grey zone (nⴝ29). NOTE.
Values are mean ⴞ SD. *Significant differences between the measurements for the merged group (CGⴙSG) that were examined at E0, E1,
and E2.

Although the GMFM is sensitive enough to detect changes
caused by therapeutic interventions in children with CP,35 this
was not confirmed in the present study, neither for the GMFM
scores nor for the other examined parameters. However, the
strength-training group in the study of Dodd et al8 had lower—
although statistically insignificant (P⫽.07)—GMFM scores
than the control group during the baseline measurement, which
possibly subdued the likelihood that the strength-training group
would improve. In the current randomized study, this situation
was unluckily reversed (the control group had lower GMFM
scores at E0), and this might partially explain the lack of
differences between the groups.
Although all participants in the present study were diagnosed
with diplegic spastic CP, we observed a considerable variance
on several clinical and gait parameters, and this necessitated
specialized surgical treatments, depending on the clinical status
of each patient. Consequently, the surgery outcome could be
neither standardized nor predicted. According to the presented
data, the impact of the applied strength-training program was
not sufficient to overcome the radical changes that occur in the
musculoskeletal system after surgery. This can be verified by
the fact that some temporal parameters that showed worsening
at E1 recovered at E2, although none of the patients followed
a strength-training program between E1 and E2.
CONCLUSIONS
The outcome of a specific strength-training protocol applied
in children with CP may be affected by several factors such as
the optimal duration, intensity, and frequency of the training
session. These factors are still to be defined for optimization of
the training effectiveness, although it is more likely that these
parameters should be determined individually for each patient.
The lack of differences between the groups in the present study
could be attributed to the immense effect that the surgery had
on both groups and should not discourage the application of
strength training in children with CP. Apart from the psycho-

Arch Phys Med Rehabil Vol 87, May 2006

 POSTOPERATIVE EFFECT OF STRENGTH TRAINING, Patikas 625

Fig 4. Kinetic parameters for the control and strength-training groups prior to, 1 year after, and 2 years after surgery. (A) Maximum hip power
absorption, (B) maximum plantarflexor moment, (C) maximum knee power absorption during loading responses, and (D) maximum ankle
power generation. The normative values with SD are represented by the continuous horizontal line and the grey zone (nⴝ29). NOTE. Values
are mean ⴞ SD. *Significant differences between the measurements for the merged group (CGⴙSG) that were examined at E0, E1, and E2.

logic benefits that children have after such training,11,36 a special-
ized training protocol targeted to well-motivated patients may
have the potential to counteract the obligatory muscle weak-
ening after the operation.
Acknowledgments: We acknowledge Sven Schneider, PhD (De-
partment of Orthopaedic Surgery, University of Heidelberg) for his
support concerning the statistical analysis of the data and Tim The-
ologis, MD (Nuffield Department of Orthopaedic Surgery, University
of Oxford, UK) for editing and commenting on the manuscript.
References
1. Nelson KB, Swaiman KF. Cerebral palsy. In: Swaiman KF, editor.
Pediatric neurology: principles and practice. 2nd ed. Toronto:
Mosby; 1993. p 471-88.
2. Castle ME, Reyman TA, Schneider M. Spasticity of spastic muscle in
cerebral palsy. Clin Orthop Relat Res 1979;Jul-Aug(142):223-32.
3. Gage JR. Gait analysis in cerebral palsy. Oxford: Mac Keith Pr;
1991.
4. Wiley ME, Damiano DL. Lower-extremity strength profiles in
spastic cerebral palsy. Dev Med Child Neurol 1998;40:100-7.
5. Damiano DL, Vaughan CL, Abel MF. Muscle response to heavy
resistance exercise in children with spastic cerebral palsy. Dev
Med Child Neurol 1995;37:731-9.
6. Bobath B. Motor development: its effect on general development
and application to the treatment of cerebral palsy. Physiotherapy
1971;57:526-32.
7. MacPhail HE, Kramer JF. Effect of isokinetic strength-training on
functional ability and walking efficiency in adolescents with ce-
rebral palsy. Dev Med Child Neurol 1995;37:763-75.
8. Dodd KJ, Taylor NF, Graham HK. A randomized clinical trial of
strength training in young people with cerebral palsy. Dev Med
Child Neurol 2003;45:652-7.
9. Damiano DL, Kelly LE, Vaughan CL. Effects of quadriceps
femoris muscle strengthening on crouch gait in children with
spastic diplegia. Phys Ther Sports 1995;75:658-71.

Arch Phys Med Rehabil Vol 87, May 2006

626 POSTOPERATIVE EFFECT OF STRENGTH TRAINING, Patikas
10. McCubbin JA, Shasby GB. Effects of isokinetic exercise on
adolescents with cerebral palsy. Adapt Phys Activity Q 1985;2:
56-64.
11. Darrah J, Wessel J, Nearingburg P, O’Connor M. Evaluation of a
community fitness program for adolescents with cerebral palsy.
Pediatr Phys Ther 1999;11:18-23.
12. O’Connell DG, Barnhard R. Improvement in wheelchair propul-
sion in pediatric wheelchair users through resistance training: a
pilot study. Arch Phys Med Rehabil 1995;76:368-72.
13. Healy A. Two methods of weight-training for children with spas-
tic type of cerebral palsy. Res Q Exerc Sport 1958;29:389-95.
14. Damiano DL, Abel MF. Functional outcomes of strength training
in spastic cerebral palsy. Arch Phys Med Rehabil 1998;79:119-25.
15. Kramer JF, MacPhail HE. Relationships among measures of walk-
ing efficiency, gross motor ability and isokinetic strength in ado-
lescents with cerebral palsy. Pediatr Phys Ther 1994;6:3-8.
16. Bohannon RW. Relative decreases in knee extension torque with
increased knee extension velocities in stroke patients with hemi-
paresis. Phys Ther 1986;67:1218-20.
17. Bohannon RW. Selected determinants of ambulatory capacity in
patients with hemiplegia. Clin Rehabil 1989;3:47-53.
18. Nakamura R, Hosokawa T, Tsuji I. Relationship of muscle
strength for knee extension to walking capacity in patients with
spastic hemiparesis. Tohoku J Exp Med 1985;145:335-40.
19. Nakamura R, Watanabe S, Handa T, Morohashi I. The relation-
ship between walking speed and muscle strength for knee exten-
sion in hemiparetic stroke patients: a follow-up study. Tohoku J
Exp Med 1988;154:111-3.
20. Osterning LR, Caster BL, James CR. Contralateral hamstring
(biceps femoris) coactivation patterns and anterior cruciate liga-
ment dysfunction. Med Sci Sport Exerc 1995;27:805-8.
21. Sale DG, MacDougall JD. Specificity in strength training: a re-
view for the coach and the athlete. Can J Appl Sport Sci 1981;6:
87-92.
22. Zwick EB, Saraph V, Strobl W, Steinwender G. Single event
multilevel surgery to improve gait in diplegic cerebral palsy—
a prospective controlled trial. Z Orthop Ihre Grenzgeb 2001;139:
485-9.
23. Kadaba MP, Ramakrishnan HK, Wootten ME, Gainey J, Gor-
ton G, Cochran GV. Repeatability of kinematic, kinetic, and
electromyographic data in normal adult gait. J Orthop Res
1989;7:849-60.
24. Russell D, Rosenbaum PL, Cadman DT. The gross motor function
measure: a means to evaluate the effects of physical therapy. Dev
Med Child Neurol 1989;31:341-52.
25. Rose J, Medeiros JM, Parker R. Energy cost index as an estimate
of energy expenditure of cerebral-palsied children during assisted
ambulation. Dev Med Child Neurol 1985;27:485-90.
Arch Phys Med Rehabil Vol 87, May 2006
26. Rose J, Gamble JG, Lee J, Lee R, Haskell WL. The energy
expenditure index: a method to quantitate and compare walking
energy expenditure for children and adolescents. J Pediatr Orthop
1991;11:571-8.
27. Rose J, Gamble JG, Medeiros JM, Burgos A, Haskell WL. Energy
cost of walking in normal children and in those with cerebral
palsy: comparison of heart rate and oxygen uptake. J Pediatr
Orthop 1989;9:276-9.
28. Schutte LM, Narayanan U, Stout JL, Selber P, Gage JR, Schwartz
MH. An index for quantifying deviation from normal gait. Gait
Posture 2000;11:25-31.
29. Perry JP. Gait analysis: normal and pathological function. Thoro-
fare: Slack; 1992.
30. Olney SJ, MacPhail HE, Hedden DM, Boyce WF. Work and power
in hemiplegic cerebral palsy gait. Phys Ther 1990;70:431-8.
31. Lelas JL, Merriman GJ, Riley PO, Kerrigan DC. Predicting peak
kinematic and kinetic parameters from gait speed. Gait Posture
2003;17:106-12.
32. Saraph V, Zwick EB, Auner C, Schneider F, Steinwender G,
Linhart WE. Gait improvement surgery in diplegic children. How
long do the improvements last? J Pediatr Orthop 2005;25:263-7.
33. Saraph V, Zwick EB, Uitz C, Linhart W, Steinwender C. The
Baumann procedure for fixed contracture of the gastrosoleus in
cerebral palsy. Evaluation of function of the ankle after multilevel
surgery. J Bone Joint Surg Br 2000;82:535-40.
34. Hicks AL, Martin KA, Ditor DS, et al. Long-term exercise train-
ing in persons with spinal cord injury: effects on strength, arm
ergometry performance and psychological well-being. Spinal Cord
2003;41:34-43.
35. Bower E, McLellan DL. Effect of increased exposure to physio-
therapy on skill acquisition of children with cerebral palsy. Dev
Med Child Neurol 1992;34:25-39.
36. McBurney H, Taylor NF, Dodd KJ, Graham HK. A quantitative
analysis of the benefits of strength training for young people with
cerebral palsy. Dev Med Child Neurol 2003;45:658-63.
Suppliers
a. Thera-Band; Hygenic Corp, 1245 Home Ave, Akron, OH 44310.
b. Vicon 612; Oxford Metrics, 14 Minns Estate, West Way, Oxford,
Oxfordshire, OX2 0JB.
c. Kistler Instruments, 75 John Glenn Dr, Amherst, NY 14228-2171.
d. K4b2, version 3.2; COSMED USA, 2211 N Elston Ave, Ste 305,
Chicago, IL 60614.
e. Version 6.5.1; The Mathworks Inc, 3 Apple Hill Dr, Natick, MA
01760-2098.
f. Version 12.0; Apache Software Foundation, 1901 Munsey Dr, For-
est Hill, MD 21050-2747.