Turk J Phys Med Rehab 2019;65(4):361-370

DOI: 10.5606/tftrd.2019.3078
©Copyright 2019 by Turkish Society of Physical Medicine and Rehabilitation - Available online at www.turkishjournalpmr.com

Original Article

Comparison of the effectiveness of partial body weight-supported

treadmill exercises, robotic-assisted treadmill exercises,
and anti-gravity treadmill exercises in spastic cerebral palsy
Berke Aras1, Evren Yaşar2, Serdar Kesikburun2, Duygu Türker2, Fatih Tok2, Bilge Yılmaz2
1
Department of Physical Medicine and Rehabilitation, Kastamonu Physical Therapy and Rehabilitation Center, Kastamonu, Turkey
2
Department of Physical Medicine and Rehabilitation, Health Sciences University Gülhane Faculty of Medicine, Ankara, Turkey

Received: April 16, 2018 Accepted: October 17, 2018 Published online: November 22, 2019

ABSTRACT
Objectives: This study aims to compare the effectiveness of the partial body weight-supported treadmill exercise (PBWSTE), robotic-assisted
treadmill exercise (RATE), and anti-gravity treadmill exercise (ATE) in children with spastic cerebral palsy (CP).
Patients and methods: Between December 01, 2015 and May 01, 2016, a total of 29 children (18 males, 11 females; mean age 9.3±2.3 years;
range, 6 to 14 years) with spastic CP were included in the study. The patients were randomly divided into three groups as the PBWSTE group
(n=10), RATE group (n=10), and ATE group (n=9). Each group underwent a total of 20 treadmill exercise sessions for 45 min for five days a
week for a total of four weeks. The patients were assessed using three-dimensional gait analysis, open-circle indirect calorimeter, six-minute
walking test, and Gross Motor Functional Measurement (GMFM) scale before and after treatment and at two months of follow-up.
Results: No significant change compared to baseline was found in the walking speed on gait analysis among the groups after the treatment.
There was no statistically significant difference among the groups in terms of the GFMF-D, GMFM-E and six-minute walking test (p>0.05).
There was a significant improvement in the oxygen consumption in the ATE group (p>0.05) and RATE group (p>0.05), but not in the
PBWSTE group (p<0.05).
Conclusion: Our study findings indicate that all three treadmill exercises have a positive impact on walking, and RATE and ATE can be used
more actively in patients with spastic CP.
Keywords: Anti-gravity treadmill exercises, cerebral palsy, gait analysis, robotic gait training.

Cerebral palsy (CP) is a non-progressive incidence was found to be 4.4% in children aged two
neurodevelopmental disorder which affects the to 16 years.[4]
child's motor and sensory system, and consequently Several factors such as selective motor control
posture and walking.[1] The underlying etiology is disturbance, muscle-tendon shortness, joint
considered multifactorial, and no specific etiological contractures, muscle weakness, spasticity, and
factor has been blamed. While it may be seen in the involuntary movements have been suggested to play
prenatal, neonatal, and postnatal period, prenatal a role in the development of gait disturbances in
factors account for 70 to 80% of all cases.[2] Although children with CP. Since one of the major problems in
the incidence of CP widely varies worldwide, it CP is gait disturbance, technological developments
has been often estimated as 1.5 to 2.5/1,000.[3] In targeting gait problems in CP treatment have also
an epidemiological study conducted in Turkey, the increased in recent years.

Corresponding author: Berke Aras, MD. Kastamonu Fizik Tedavi ve Rehabilitasyon Merkezi, Fiziksel Tıp ve Rehabilitasyon Bölümü, 37200 Kastamonu, Türkiye.
e-mail: drberkearas@gmail.com
Cite this article as:
Aras B, Yaşar E, Kesikburun S, Türker D, Tok F, Yılmaz B. Comparison of the effectiveness of partial body weight-supported treadmill exercises, robotic-assisted treadmill exercises, and anti-gravity
treadmill exercises in spastic cerebral palsy. Turk J Phys Med Rehab 2019;65(4):361-370.
Presented at the 5th Medical Rehabilitation Congress, November 03-06, 2016, Ankara, Turkey
362 Turk J Phys Med Rehab

In this context, repetitive and task-specific the effectiveness of these three treadmill exercises in
therapies are used, such as partial body weight- the treatment of gait disorder secondary to CP.
supported treadmill exercises (PBWSTE), robotic-
assisted treadmill exercises (RATE), and anti-gravity PATIENTS AND METHODS
treadmill exercises (ATE). The PBWSTE reduces the
amount of weight riding in the lower extremities using In this prospective study, a comprehensive
a postural control system consisting of a parachute chart review of the children who were followed
assembly. The main goal of PBWSTE is to provide in the pediatric rehabilitation unit of Gaziler
a symmetrical walking pattern to patients with Physical Therapy and Rehabilitation Research and
rhythmic and normal physiological walking training.[5] Training Hospital, Ankara, Turkey was performed
Current theories of motor learning have suggested that to screen eligible subjects for the study between
task-specific repetitive practices improve functional December 01, 2015 and May 01, 2016. Inclusion criteria
activities such as walking in patients with CP.[6] The were as follows: age between six to 14 years; having
RATE is used in the field of rehabilitation to recover a diagnosis of CP, having Level II-III according
or improve walking ability in cases where walking to the Gross Motor Function Classification System
is impaired due to neurological disorders.[7] It is (GMFCS), and (being able to ambulate at least 10 meters
with or without an assistive device (i.e., walker, or
performed on a treadmill in the standing position with
tripod). Exclusion criteria were as follows: having
a body-supported suspension system. With the help of
botulinum toxin injection within the last six months,
the robot, it has been claimed that the movement of the
having prior orthopedic surgery within the last one
legs in the normal walking cycle, send signals to the
year, having a lower limb length discrepancy of
brain with sensorial perception from the legs and help
>2 cm, having a fixed contracture or joint deformity,
develop the ability of adaptation in the brain called
receiving oral medication (i.e., baclofen, tizanidine, or
plasticity.[8] The ATE consists of treadmill limited to
trihexyphenidyl) or baclofen pump to adjust the muscle
a positive pressure environment. The body weight
tone, and having concomitant cardiovascular diseases.
support depends on the amount of air entrapped
Of 48 children who met the inclusion criteria, 18 were
in the pack and creating lift force. Patients are able
excluded as 15 of them received botulinum toxin
to comfortably use the upper limbs during walking
injections within the last six months, and the parents
exercises and hold on to the parallel bars on both sides,
of three children did not give consent to participate
if necessary.[9]
in the study. A total of 30 children were enrolled
The cost-effectiveness of these rehabilitative for the study. Although there was no loss through
technologies has not been well-documented, yet. In the study and follow-up in the PBWSTE and RATE
addition, there are conflicting results regarding their groups, one child in the ATE group was excluded
use in CP. In the present study, we aimed to compare due to non-completion of the treatment sessions.

Assessed for eligibility (n=48)

Excluded (n=18)
• Previous botulinum toxin
Randomized (n=30)
injection (n=15)
• Rejected by parents (n=3)

Allocated to partial
Allocated to robotic supported Allocated to antigravity
body weight supported
treadmill exercise treadmill exercise
treadmill exercise

Analyzed (n=10) Analyzed (n=10) Analyzed (n=9)

• Withdrawn due to
disruption of the treatment
(n=1)

Figure 1. Study flow chart.

Comparison of the effectiveness of the different treadmill exercises in spastic cerebral palsy 363

Finally, a total of 29 children with spastic CP (18 males, calorimeter (Vmax 29c, SensorMedics, Yorba Linda,
11 females; mean age 9.3±2.3 years; range, 6 to 14 years) CA, USA) system. After a proper resting period with
were analyzed. The study flow chart is shown in the face mask, the test was carried out with a walking
Figure 1. A written informed consent was obtained time of five min at the running speed of the treadmill
from each parent. The study protocol was approved at its own walking speeds (average 0.70 m/s). It is
by the local Research Ethics Committee of Gulhane known that children with CP reach a balanced plateau
Military Medical Academy. The study was conducted following a two-min submaximal walk.[10] Therefore,
in accordance with the principles of the Declaration the first two min taken to reach the steady state during
of Helsinki. a five-min walk was not included in the calculation.
Procedures The 6MWT was used to assess changes in the
walking capacity of the patients. In the test developed
The patients were randomized into three groups
by Balke in 1963, the distance the patient took in
using the closed envelope method. All patients took
six min was measured in meters.[11] Before the test,
a 20-session walking exercise program for 45 min
each patient rested for 15 min at the start of the
for five days a week for a total of four weeks using
course. The patient was informed about the test and
the PBWSTE (Visio Gait, Bama Technology, Turkey),
was, then, asked to walk for six min at their own
RATE (Lokomat, Hocoma AG, Switzerland), and ATE
walking speed.
(Alter-G Inc., Fremont, CA, USA). Each treatment
session was performed under the supervision of a Statistical analysis
physiotherapist. The body weight support was started
To determine the sample size, the study conducted
at 60% and gradually decreased to a level which
by Arya et al.[12] on children with CP was used. The
prevented the collapse of the knee in flexion during
primary outcome measure of this study was walking
the stance phase. The treadmill speed was initiated at
speed on the gait analysis and a 0.12 m/sec difference
the average walking speed according to the children's
was considered clinically significant. Accordingly, the
walking pattern, weight, and endurance and, then,
sample size was calculated to be at least eight for each
increased to the highest level tolerated.
group with an error margin of 0.05 and a beta value
Outcome measures of 0.20 at 80%. A total of 10 patients were required
for each group, taking a dropout rate of 25% into
The primary outcome measure was walking
consideration.
speed on three-dimensional (3D) gait analysis data.
Temporospatial parameters including walking Statistical analysis was performed using SPSS
speed, cadence, stride time, stride length, step for Windows version 15.0 software (SPSS Inc.,
time, step length, and single-double support Chicago, IL, USA). Descriptive data were expressed
phase were measured using the 3D gait analysis. in mean ± standard deviation (SD) for continuous
Open-circuit indirect calorimetry, the six-minute variables, and number and frequency for categorical
walking test (6MWT), the Gross Motor Function variables. Conformity to normal distribution was
Measure-66 (GMFM-66) part D (standing) and assessed using the Shapiro-Wilk test. Non-parametric
part E (walking-running) were the secondary tests were used for the data not showing normal
outcome measures. All measurements were distribution. Pre-treatment, post-treatment, and
performed before the treatment, immediately after two-month follow-up results were assessed with
the treatment, and three months after the treatment. the Friedman test and the results were considered
statistically significant at a p value of <0.05. In cases
Three-dimensional gait analysis of the patients was
where statistically significant changes were found
performed in the gait and motion analysis laboratory of
within the group, two-up comparisons were made with
Gaziler Physical Therapy and Rehabilitation Research
the Bonferroni corrected Wilcoxon test and the results
and Training Hospital. Gait analysis was applied by a
were considered statistically significant at a p value of
single physician for each patient. The 3D gait analysis
<0.017. In the intergroup comparisons, the changes
test was performed using the Vicon 512 (Oxford
in the parameters examined were evaluated with
Metrics, Oxford, United Kingdom) motion analysis
the Kruskal-Wallis test. When statistically significant
system. Data were related to the temporospatial
changes were found between the groups, two-up
characteristics of the walk.
comparisons were made with the Bonferroni corrected
The energy consumption of the children was Mann-Whitney U test and the results were considered
measured using a mask with an open circuit indirect statistically significant at a p value of <0.017.
364 Turk J Phys Med Rehab

Table 1. Demographic and clinical characteristic of patients

RATE group PBWSTE group ATE group
n % n % n % p
Sex 0.94
Male 6 60.0 6 60.0 6 66.7
Female 4 40.0 4 40.0 3 33.3
Cerebral palsy type 0.98
Hemiplegic 3 30.0 3 30.0 3 33.3
Diplegic 7 70.0 7 70.0 6 66.7
GMFCS level 0.41
Level II 9 90.0 7 70.0 8 88.9
Level III 1 10.0 3 30.0 1 11.1
RATE: Robotic assisted treadmill exercise; PBWSTE: Partial body weight supported treadmill exercise; ATE: Antigravity
treadmill exercise; GMFCS: Gross motor functional classification system.

RESULTS No statistically significant difference in the baseline

data of gait analysis data, oxygen consumption,
There was no statistically significant difference
6MWT, GMFM-D and E was detected between the
in the baseline characteristics of the groups (p>0.05) groups (p>0.05). Walking speed on gait analysis
(Table 1). The CP type was diplegic in 20 (69%) and showed an increase in all treatment groups, but did
hemiplegic in nine patients (31%). The GMFCS was not reach a statistical significance in any treatment
Level II in 24 (82.8%) and Level III in five patients group (Figure 2). In the ATE group, the increase in
(17.2%). Twenty-five children (86.2%) could walk cadence, stride length, and stride time were statistically
without an assistive device and four (13.8%) could walk significant, while the increases in the RATE and
with an assistive device (three with walkers, one with PBWSTE groups were not statistically significant
tripod). Walking orthotics were used in 24 children (p>0.05). The decrease in double support phase was
(82.8%). statistically significant in the RATE and ATE groups

Pre-treatment
Post-treatment
Follow-up
1.25

1.00
Walking speed (m/sn)

0.75

1.00

0.25

0.00

RATE group PBWSTE group ATE group

Figure 2. Walking speed changes of the treatment groups.

RATE: Robotic-assisted treadmill exercise; PBWSTE: Partial body weight supported treadmill exercise; ATE: Antigravity
treadmill exercise.
Comparison of the effectiveness of the different treadmill exercises in spastic cerebral palsy 365

Table 2. Intra-group comparison of gait analysis data

Pre-treatment Post-treatment Follow-up
n Mean±SD Mean±SD Mean±SD p
Walking speed (m/sn)
RATE 10 0.5±0.3 0.6±0.3 0.7±0.4 0.172
PBWSTE 10 0.6±0.3 0.7±0.3 0.7±0.4 0.119
ATE 9 0.6±0.3 0.6±0.3 0.7±0.3 0.117
Cadans (step/min)
RATE 10 92.8±28.5 102.0±25.3 102.5±26.4 0.061
PBWSTE 10 98.0±38.2 98.1±42.5 95.7±41.0 0.794
ATE 9 101.2±28.7* 108.3±29.5* 112.0±32.1 0.017
Stride length (m)
RATE 10 0.6±0.2 0.7±0.3 0.7±0.3 0.122
PBWSTE 10 0.7±0.2 0.7±0.3 0.7±0.3 0.387
ATE 9 0.6±0.3 0.7±0.3 0.7±0.2 0.042
Step length (m)
RATE* 10 0.3±0.1 0.3±0.1 0.4±0.1 0.052
PBWSTE*† 10 0.4±0.2 0.4±0.2 0.4±0.1 0.323
ATE† 9 0.3±0.1 0.4±0.2 0.4±0.1 0.062
Single support
RATE 10 0.4±0.1 0.4±0.1 0.5±0.1 0.256
PBWSTE 10 0.5±0.2 0.4±0.1 0.5±0.1 0.898
ATE 9 0.4±0.2 0.4±0.1 0.4±0.1 0.531
Double support
RATE 10 0.5±0.4* 0.4±0.3* 0.3±0.2 0.011
PBWSTE 10 0.6±0.6 1.0±1.6 0.9±1.6 0.479
ATE 9 0.5±0.6 a 0.5±0.6* 0.4±0.4 0.019
Step time
RATE* 10 0.7±0.2 0.6±0.2 0.6±0.2 0.087
PBWSTE*† 10 0.7±0.4 0.9±0.9 0.9±0.9 0.337
ATE† 9 0.7±0.4 0.6±0.3 0.6±0.3 0.107
Stride time
RATE* 10 1.4±0.4 1.3±0.4 1.3±0.4 0.255
PBWSTE*† 10 1.5±0.8 1.8±1.8 1.8±1.8 0.607
ATE† 9 1.4±0.9 1.2±0.7 1.2±0.6 0.013
SD: Standard deviation; RATE: Robotic assisted treadmill exercise; PBWSTE: Partial body weight supported treadmill exercise;
ATE: Antigravity treadmill exercise; * The binary difference between the first and second measurements was statistically
significant (p<0.017); † The binary difference between the first and third measurements was statistically significant (p<0.017).

(p<0.05), although the change in the PBWSTE group A significant improvement in the GMFM was
was not significant (p>0.05). Stride length, step time, detected in all three groups in both GMFM-D and
and single support phase changes were not significant GMFM-E after treatment (p<0.05) (Table 4). There
in any groups (Table 2). Statistically significant changes was no statistically significant difference among the
were observed among the group measurements in groups in terms of GFMF-D and GMFM-E (p>0.05)
terms of step length, step time, and stride time after (Table 5).
treatment (p<0.05). In pairwise comparisons, the
RATE and ATE groups were found to be significantly There was a significant decrease in the oxygen
superior compared to the PBWSTE group (p<0.017) consumption in the RATE and ATE groups after
and there was no statistically significant difference treatment (p<0.05), although the decrease in the
between the RS and the ATE group (p>0.017) (Table 3). PBWSTE group was not statistically significant
366 Turk J Phys Med Rehab

Table 3. Post-treatment and follow-up changes of gait analysis data

Pre-treatment Follow-up
Mean±SD p Mean±SD p
Cadans (step/min)
RATE 9.3±11.4 Ï 9.7±10.4 Ï
Ô Ô
PBWSTE 0.2±14.9 Ô˝ 0.122 -2.2±8.6 Ô˝ 0.055
ATE 7.1±8.4 10.8±12.6
Ï Ï

Stride length (m)

RATE 0.0±0.1 Ï 0.1±0.1 Ï
Ô Ô
PBWSTE -0.1±0.2 Ô˝ 0.097 0.0±0.2 Ô˝ 0.354
ATE 0.1±0.2 0.1±0.1
Ï Ï

Step length (m)

RATE* 0.0±0.1 Ï 0.1±0.1 Ï
Ô Ô
PBWSTE*† -0.1±0.1 Ô˝ 0.240 -0.1±0.2 Ô˝ 0.020
Ï Ï
ATE† 0.0±0.1 0.0±0.1
Single support
RATE 0.0±0.1 Ï 0.0±0.1 Ï
Ô Ô
PBWSTE -0.0±0.1 Ô˝ 0.839 0.0±0.1 Ô˝ 0.360
Ï Ï
ATE -0.1±0.1 -0.0±0.1
Double support
RATE -0.1±0.2 Ï -0.2±0.2 Ï
Ô Ô
PBWSTE 0.4±1.2 Ô˝ 0.151 0.3±1.2 Ô˝ 0.192
Ï Ï
ATE -0.0±0.1 -0.2±0.2
Step time
RATE* -0.1±0.2 Ï -0.1±0.1 Ï
Ô Ô
PBWSTE*† 0.2±0.5 Ô˝ 0.286 0.2±0.5 Ô˝ 0.022
Ï Ï
ATE† -0.0±0.1 -0.1±0.2
Stride time
RATE* -0.1±0.3 Ï 0.1±0.2 Ï
Ô Ô
PBWSTE*† 0.3±1.1 Ô˝ 0.129 0.4±1.1 Ô˝ 0.047
ATE† -0.2±0.2 -0.2±0.4
Ï Ï

SD: Standard deviation; RATE: Robotic assisted treadmill exercise; PBWSTE: Partial body weight supported treadmill exercise;
ATE: Antigravity treadmill exercise; * The binary difference between the first and second measurements was statistically
significant (p<0.017); † The binary comparison between PBWSTE and RATE was statistically significant (p<0.017).

(Figure 3; Table 4). There was a statistically significant DISCUSSION

difference in oxygen consumption among the groups
The results of the present study of PBWSTE, RATE,
after treatment (p<0.05). The RATE and ATE groups
and ATE treatments showed positive gains in the
showed statistically significant superiority compared
gross motor function (GMFM) and walking capacity
to the PBWSTE group (p<0.017). However, there was
(6MWT) in children with spastic CP. A reduction
no statistically significant difference between the
in oxygen consumption was seen as an advantage in
RATE and ATE groups (p>0.017) (Table 5).
the RATE and ATE groups. Based on temporospatial
In the 6MWT, a significant improvement was data, positive effects were only seen in the ATE group
observed in all three groups after treatment (p<0.05) in cadence, stride length, and stride time parameters.
(Table 4). However, there was no statistically The double support phase demonstrated a positive
significant difference among the groups (p>0.05) impact on the RATE and the ATE groups. Although
(Table 5). an increase in walking speed was observed in all the
Comparison of the effectiveness of the different treadmill exercises in spastic cerebral palsy 367

Table 4. Intra-group comparison of GMFM, oxygen consumption and 6MWT

Pre-treatment Post-treatment Follow-up
Mean±SD Mean±SD Mean±SD p
GMFM-D
RATE 73.8±17.2*† 77.4±15.9* 77.4±15.9† <0.001
PBWSTE 75.3±16.6 80.0±13.5 80.0±13.5 <0.001
ATE 80.9±7.5 84.4±6.1 84.4±6.1 <0.001
GMFM-E
RATE 77.4±16.9*† 79.8±15.51* 80.0±15.5† <0.001
PBWSTE 75.2±14.4*† 77.9±13.4* 77.9±13.4† <0.001
ATE 79.2±8.6*† 82.9±7.5* 82.9±7.5† <0.001
Oxygen consumption
RATE 10.0±2.6*† 8.6±2.1* 9.0±2.0† 0.030
PBWSTE 10.8±3.0 10.7±2.9 10.5±2.9 0.479
ATE 9.6±2.53*† 8.4±2.1* 8.5±1.9† 0.001
6MWT (m)
RATE 263.9±132.5*† 303.5±134.4* 309.1±134.2† 0.001
PBWSTE 253.7±151.6*† 291.3±164.3* 302.3±156.3† 0.001
ATE 281.3±115.1*† 329.7±110.8* 339.6±123.3† 0.001
6MWT: Six-minute walking test; GMFM: Gross Motor Functional Measurement; SD: Standard deviation; RATE: Robotic assisted
treadmill exercise; PBWSTE: Partial body weight supported treadmill exercise; ATE: Antigravity treadmill exercise; * The binary
difference between the first and second measurements was statistically significant (p<0.017); † The binary difference between the
first and third measurements was statistically significant (p<0.017).

Table 5. Post-treatment and follow-up changes of GMFM, oxygen consumption and 6MWT
Post-treatment Follow-up
Mean±SD p Mean±SD p
GMFM-D
RATE 3.6±2.5 Ï 3.6±2.5 Ï
Ô Ô
PBWSTE 4.6±4.6 Ô˝ 0.732 4.6±4.6 Ô˝ 0.732
ATE 3.5±2.5 3.5±2.5
Ï Ï

GMFM-E
RATE 2.4±2.0 Ï 2.6±1.8 Ï
Ô Ô
PBWSTE 2.6±1.7 Ô˝ 0.293 2.6±1.7 Ô˝ 0.301
ATE 3.7±1.9 3.7±1.9
Ï Ï

Oxygen consumption
RATE* -1.4±1.4 Ï -1.0±1.5 Ï
Ô Ô
PBWSTE*† -0.1±1.0 Ô˝ 0.040 -0.3±0.7 Ô˝ 0.120
Ï Ï
ATE† -1.2±0.9 -1.2±0.9
6MWT (m)
RATE 39.6±40.4 Ï 45.2±44.4 Ï
Ô Ô
PBWSTE 37.6±20.2 Ô˝ 0.324 48.6±37.8 Ô˝ 0.240
Ï Ï
ATE 48.3±25.1 58.2±22.9
6MWT: Six-minute walking test; GMFM: Gross Motor Functional Measurement; SD: Standard deviation; RATE: Robotic assisted
treadmill exercise; PBWSTE: Partial body weight supported treadmill exercise; ATE: Antigravity treadmill exercise; * The binary
difference between the first and second measurements was statistically significant; † The binary comparison between PBWSTE
and RATE was statistically significant (p<0.017).
368 Turk J Phys Med Rehab

1.25
Pre-treatment
Post-treatment
1.00 Follow-up
Oxygen consumption (VO2)

0.75

1.00

0.25

0.00

0.00

RATE group PBWSTE group ATE group

Figure 3. Oxygen consumption changes of the treatment groups.
RATE: Robotic-assisted treadmill exercise; PBWSTE: Partial body weight supported treadmill exercise; ATE: Antigravity
treadmill exercise.

treatment groups, compared to the baseline values, this The small number of patients and different scales used
increase was not significant. in the studies might have reduced the reliability of the
studies. The second was therapeutic heterogeneity. The
In a previous study[13] investigating the effects of
intensity of walking exercises, the weekly frequency,
RATE on standing and walking in children with CP,
the number and duration of sessions, the percentage
CP patients with mild motor impairment (GMFCS I-II)
of partial weight-supported, and the walking speed
were more successful than those with severe motor
vary in different studies and do not provide sufficient
impairment (GMFCS III-IV). In a retrospective study,
information for clinicians. The third limitation was
Schroeder et al.[14] found that pre-treatment GMFM
heterogeneity in the patient population. The presence
values were the most important factors affecting the
of wide age range or inclusion of patients at different
prognosis of treatment which examined patient-related
GMFCS levels did not provide reliable information
factors associated with therapeutic effects of RATE. It
about which treatment should be given in which
was concluded that age, sex, CP type, and botulinum
functional state. Therefore, the treatment protocol
toxin injections were not found to be effective factors.
in walking exercises varies between studies in the
That is why only GMFCS II-III children were included
literature. The session frequency ranges from two to
in the current study.
five per week, and the duration of treatment ranges
The PBWSTE reduces the amount of weight in from two weeks to five months. In most studies, the
the lower extremities using a postural control system treatment duration is between 20 to 30 min. There has
consisting of a parachute set, which brings the child’s not been enough evidence to show which treatment
walk to a normal physiological walking pattern protocol is superior. In the present study, 20 sessions
with task-specific, repetitive applications. Although of walking for 45 min, five days a week for four weeks,
Chrysagis et al.[15] reported that 12 sessions (four weeks, were applied, which was described by Drubicki et al.[18]
three sessions per week) of PBWSTE were superior as the most reasonable time for functional activity.
to conventional physiotherapy in patients with CP
One of the repetitive, task-specific therapies used
of GMFM-D and E, Willoughby et al.[16] advocated
in CP treatment and which improves the walking
that PBWSTE was not superior to conventional
patterns of the children is RATE. In rehabilitation
physiotherapy in GMFM III-IV.
practice, it is also used in other neurological diseases
Zwicker and Mayson[17] presented three basic other than CP, such as stroke, spinal cord injury,
problems related to the literature in a systematic traumatic brain injury and multiple sclerosis.[18] Due
review of the use of PBWSTE in children with CP. to conflicting data in literature, the superiority of
The first problem was methodological heterogeneity. RATE has been questioned, and it has been suggested
Comparison of the effectiveness of the different treadmill exercises in spastic cerebral palsy 369

to be an expensive treatment option and has, per week) of 30-min anti-gravity walking exercises
therefore, not received as much interest as expected in were administered to nine CP children at GMFM level
adult rehabilitation.[19] However, RATE is considered II-IV. There was a significant increase in the walking
a promising option in CP due to the extension of speed, step length and stride phase, the step width;
the exercise period with frequent step repetition, however, no statistically significant change was found
symmetric walking patterns acquired by motor in the stride step length, and cadence at the end of the
learning and central nervous system adaptation. As study. In the present study, the increase in the walking
in the PBWSTE, there are similar problems about speed was not statistically significant in the ATE
the use of RATE in the treatment of CP (i.e., lack of a group. This finding indicates a significant increase in
randomized controlled study, fewer patients, different the cadence, stride length, and stride time. According
sessions and duration, and different GMFCS levels). to the current knowledge, this study can be considered
of value, as it is the second scientific study where ATE
In the study investigating the long-term effects of
was used in CP patients.
RATE on CP of Borggraefe et al.,[20] 14 participants
took 12 sessions (three weeks, four days a week) of In the present study, the energy and calories
treatment and were evaluated with GMFM-D and consumed by children were measured by open
GMFM-E, the 6MWT and the 10-meter walking test. circuit indirect calorimetry method. In the literature,
All parameters improved after the treatment and the the 6MWT is usually used for the measurement
development was observed to continue during six- of endurance in the treadmill exercises. Although
month follow-up. In the present study, a slight decrease this test indicates the cardiopulmonary capacity in
was observed after the treatment. However, the gains cardiopulmonary diseases, it does not fully represent
were sustained through three months (GMFM-D, the cardiopulmonary capacity in diseases which may
GMFM-E, 6MWT, oxygen consumption). cause gait disturbances due to loss of muscle strength,
balance disorders, and spasticity such as CP.[25,26] In a
There have been only two randomized controlled study[27] which measured peak oxygen consumption
trials investigating the effects of RATE in children (VO2) with the 6MWT in children with CP, there
with CP. In the first by Smania et al.[21] in 2011, Gait was a weak correlation between the two parameters,
Trainer GT-I was used in 18 patients who received and the 6MWT showed walking performance rather
10 sessions (two weeks, five days a week) 30 min of than cardiopulmonary capacity. In the present study,
robotic walking and 10 min of strengthening exercises indirect calorimetry levels were used for the evaluation
in the study group and 40 min of conventional of cardiopulmonary capacity.
physiotherapy in the control group. The study group
showed an improvement in 10-meter walking test, a Nonetheless, there are some limitations to the
6MWT, a WeeFIM scale, and gait analyses after the present study. Small sample size is the main limitation.
treatment. However, the main limitation was that the To be able to make more definitive conclusions about
age of the patients included in this study which was the effectiveness of these treatment protocols, there
between 10 and 18 years. It is particularly difficult in is a need for further studies with larger samples. In
this age group to exclude the effects of rapid growth addition, the lack of a control group to which no
during puberty. That is why the current study did not walking exercises were applied is another limitation.
include children over the age of 14 years. In the study In conclusion, to the best of our knowledge, this is
of Druzbicki et al.,[18] the study group received RATE the first study in the literature to compare the effects
for 20¥45-min sessions (five sessions a week for four of three different walking exercises (PBWSTE, RATE,
weeks) and 20 sessions of conventional exercises were and ATE) used in CP rehabilitation. Our study showed
applied to the control group. As a result of this study, significant changes in the gross motor function and
a slight increase in the walking speed was found walking performance in children of all three groups
in both groups, although the difference was not after the treatment. In addition, positive changes in
statistically significant within the groups or between the PBWSTE group were less than those in the other
the groups. two treatment groups. Based on these findings, we
Although ATE has been shown to be effective in believe that PBWSTE is relatively more dependent on a
certain conditions such as stress fracture,[22] Achilles physiotherapist compared to RATE and ATE.
tendon injuries,[23] and muscular dystrophy,[24] there Declaration of conflicting interests
has been only one study of its use with CP patients. In a The authors declared no conflicts of interest with respect
study by Kurz et al.,[9] 12 sessions (six weeks, two sessions to the authorship and/or publication of this article.
370 Turk J Phys Med Rehab

Funding determinants of responsiveness to robot-enhanced

The authors received no financial support for the research treadmill therapy in children and adolescents with cerebral
palsy. Dev Med Child Neurol 2014;56:1172-9.
and/or authorship of this article.
15. Chrysagis N, Skordilis EK, Stavrou N, Grammatopoulou E,
Koutsouki D. The effect of treadmill training on gross motor
REFERENCES function and walking speed in ambulatory adolescents with
cerebral palsy: a randomized controlled trial. Am J Phys
1. Rosenbaum PL, Walter SD, Hanna SE, Palisano RJ, Russell
Med Rehabil 2012;91:747-60.
DJ, Raina P, et al. Prognosis for gross motor function in
cerebral palsy: creation of motor development curves. 16. Willoughby KL, Dodd KJ, Shields N, Foley S. Efficacy of
JAMA 2002;288:1357-63. partial body weight-supported treadmill training compared
with overground walking practice for children with
2. Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M,
cerebral palsy: a randomized controlled trial. Arch Phys
Damiano D, et al. A report: the definition and classification
Med Rehabil 2010;91:333-9.
of cerebral palsy April 2006. Dev Med Child Neurol Suppl
2007;109:8-14. 17. Zwicker JG, Mayson TA. Effectiveness of treadmill training
in children with motor impairments: an overview of
3. Drougia A, Giapros V, Krallis N, Theocharis P, Nikaki A,
systematic reviews. Pediatr Phys Ther 2010;22:361-77.
Tzoufi M, et al. Incidence and risk factors for cerebral palsy
18. Druzbicki M, Rusek W, Snela S, Dudek J, Szczepanik M,
in infants with perinatal problems: a 15-year review. Early
Zak E, et al. Functional effects of robotic-assisted locomotor
Hum Dev 2007;83:541-7.
treadmill thearapy in children with cerebral palsy. J Rehabil
4. Serdaroğlu A, Cansu A, Ozkan S, Tezcan S. Prevalence of
Med 2013;45:358-63.
cerebral palsy in Turkish children between the ages of 2 and
19. Dobkin BH, Duncan PW. Should body weight-supported
16 years. Dev Med Child Neurol 2006;48:413- 6.
treadmill training and robotic-assistive steppers
5. Dodd KJ, Foley S. Partial body-weight-supported treadmill
for locomotor training trot back to the starting gate?
training can improve walking in children with cerebral Neurorehabil Neural Repair 2012;26:308-17.
palsy: a clinical controlled trial. Dev Med Child Neurol
20. Borggraefe I, Kiwull L, Schaefer JS, Koerte I, Blaschek A,
2007;49:101-5.
Meyer-Heim A, et al. Sustainability of motor performance
6. Barbeau H. Locomotor training in neurorehabilitation: after robotic-assisted treadmill therapy in children: an
emerging rehabilitation concepts. Neurorehabil Neural open, non-randomized baseline-treatment study. Eur J Phys
Repair 2003;17:3-11. Rehabil Med 2010;46:125-31.
7. Jezernik S, Colombo G, Keller T, Frueh H, Morari M. 21. Smania N, Bonetti P, Gandolfi M, Cosentino A, Waldner
Robotic orthosis lokomat: a rehabilitation and research tool. A, Hesse S, et al. Improved gait after repetitive locomotor
Neuromodulation 2003;6:108-15. training in children with cerebral palsy. Am J Phys Med
8. Neckel N, Wisman W, Hidler J. Limb alignment and Rehabil 2011;90:137-49.
kinematics inside a Lokomat robotic orthosis. Conf Proc 22. Tenforde AS, Watanabe LM, Moreno TJ, Fredericson M.
IEEE Eng Med Biol Soc 2006;1:2698-701. Use of an antigravity treadmill for rehabilitation of a pelvic
9. Kurz MJ, Corr B, Stuberg W, Volkman KG, Smith N. stress injury. PM R 2012;4:629-31.
Evaluation of lower body positive pressure supported 23. Saxena A, Granot A. Use of an anti-gravity treadmill in the
treadmill training for children with cerebral palsy. Pediatr rehabilitation of the operated achilles tendon: a pilot study.
Phys Ther 2011;23:232-9. J Foot Ankle Surg 2011;50:558-61.
10. Unnithan VB, Dowling JJ, Frost G, Bar-Or O. Role of 24. Berthelsen MP, Husu E, Christensen SB, Prahm KP, Vissing
cocontraction in the O2 cost of walking in children with J, Jensen BR. Anti-gravity training improves walking
cerebral palsy. Med Sci Sports Exerc 1996;28:1498-504. capacity and postural balance in patients with muscular
11. Balke B. A simple field test for the assessment of physical dystrophy. Neuromuscul Disord 2014;24:492-8.
fitness. REP 63-6. Rep Civ Aeromed Res Inst US 1963:1-8. 25. Cahalin L, Pappagianopoulos P, Prevost S, Wain J, Ginns L.
12. Arya BK, Mohapatra J, Subramanya K, Prasad H, Kumar R, The relationship of the 6-min walk test to maximal oxygen
Mahadevappa M. Surface EMG analysis and changes in gait consumption in transplant candidates with end-stage lung
following electrical stimulation of quadriceps femoris and disease. Chest 1995;108:452-9.
tibialis anterior in children with spastic cerebral palsy. Conf 26. Zugck C, Krüger C, Dürr S, Gerber SH, Haunstetter
Proc IEEE Eng Med Biol Soc 2012;2012:5726-9. A, Hornig K, et al. Is the 6-minute walk test a reliable
13. Borggraefe I, Schaefer JS, Klaiber M, Dabrowski E, Ammann- substitute for peak oxygen uptake in patients with dilated
Reiffer C, Knecht B, et al. Robotic-assisted treadmill therapy cardiomyopathy? Eur Heart J 2000;21:540-9.
improves walking and standing performance in children 27. Slaman J, Dallmeijer A, Stam H, Russchen H, Roebroeck
and adolescents with cerebral palsy. Eur J Paediatr Neurol M, van den Berg-Emons R. The six-minute walk test cannot
2010;14:496-502. predict peak cardiopulmonary fitness in ambulatory
14. Schroeder AS, Von Kries R, Riedel C, Homburg M, adolescents and young adults with cerebral palsy. Arch Phys
Auffermann H, Blaschek A, et al. Patient-specific Med Rehabil 2013;94:2227-33.