771824

research-article2018
CRE0010.1177/0269215518771824Clinical RehabilitationLintanf et al.

CLINICAL
Original Article REHABILITATION

Clinical Rehabilitation

Effect of ankle-foot orthoses on gait, 1­–14

© The Author(s) 2018
Reprints and permissions:
balance and gross motor function sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0269215518771824
https://doi.org/10.1177/0269215518771824

in children with cerebral palsy: a journals.sagepub.com/home/cre

systematic review and meta-analysis

Mael Lintanf1,2 , Jean-Sébastien Bourseul1,2,

Laetitia Houx1,3,4, Mathieu Lempereur1,4,
Sylvain Brochard1,2,3,4 and Christelle Pons1,2,3,4

Abstract
Objective: To determine the effects of ankle-foot orthoses (AFOs) on gait, balance, gross motor function
and activities of daily living in children with cerebral palsy.
Data sources: Five databases were searched (Pubmed, Psycinfo, Web of Science, Academic Search
Premier and Cochrane Library) before January 2018.
Review methods: Studies of the effect of AFOs on gait, balance, gross motor function and activities of
daily living in children with cerebral palsy were included. Articles with a modified PEDRO score ≥ 5/9 were
selected. Data regarding population, AFO, interventions and outcomes were extracted. When possible,
standardized mean differences (SMDs) were calculated from the outcomes.
Results: Thirty-two articles, corresponding to 56 studies (884 children) were included. Fifty-one studies
included children with spastic cerebral palsy. AFOs increased stride length (SMD = 0.88, P < 0.001) and gait
speed (SMD = 0.28, P < 0.001), and decreased cadence (SMD = –0.72, P < 0.001). Gross motor function scores
improved (Gross Motor Function Measure (GMFM) D (SMD = 0.30, P = 0.004), E (SMD = 0.28, P = 0.02),
Pediatric Evaluation of Disability Inventory (PEDI) (SMD = 0.57, P < 0.001)). Data relating to balance and
activities of daily living were insufficient to conclude. Posterior AFOs (solid, hinged, supra-malleolar, dynamic)
increased ankle dorsiflexion at initial contact (SMD = 1.65, P < 0.001) and during swing (SMD = 1.34, P < 0.001),
and decreased ankle power generation in stance (SMD = –0.72, P < 0.001) in children with equinus gait.
Conclusion: In children with spastic cerebral palsy, there is strong evidence that AFOs induce small improvements
in gait speed and moderate evidence that AFOs have a small to moderate effect on gross motor function. In
children with equinus gait, there is strong evidence that posterior AFOs induce large changes in distal kinematics.

Keywords
Child rehabilitation, gait, cerebral palsy, orthoses, mobility

Received: 10 August 2017; accepted: 18 February 2018

1Physical and Rehabilitation Medicine Department, University 4LaTIM—INSERM UMR1101, Brest, France
Hospital of Brest, Brest, France
2University of Western Brittany, Brest, France Corresponding author:
3Pediatric Physical and Rehabilitation Medicine Department, Mael Lintanf, Physical and Rehabilitation Medicine Department,
Fondation Ildys, Brest, France University Hospital of Brest, Brest 29609, France.
Email: lintanfmael@hotmail.fr
2 Clinical Rehabilitation 00(0)
Introduction
An important goal for the management of children
with cerebral palsy is to improve gait and gross
motor function, because of their impact on auton-
omy and participation.1 Ankle-foot orthoses
(AFOs) are often prescribed to correct abnormal
gait and facilitate gait training and functional activ-
ities.2,3 A recent French survey found that more
than 20% of children with cerebral palsy wear
AFOs during the day.4
Systematic reviews evaluating the effectiveness
of AFOs on gait in children with cerebral palsy
concluded that they improve gait parameters; how-
ever the size of this effect was not quantified5–7 and
the most recent literature was not taken into
account. Furthermore, the impact of AFOs on cer-
tain gait parameters (velocity, cadence, energy con-
sumption, hip kinematics), as well as on gross
motor function, balance or participation is still
under debate.5,8
Many different types of orthoses have been
developed, each with specific mechanical proper-
ties.9 Many prospective studies have also been car-
ried out but the population sizes were small and the
type of orthoses varied. It is, thus, difficult to draw
conclusions with regard to the effects of each
orthosis on the different types of cerebral palsy.5
One systematic review of the literature, which only
evaluated solid and hinged ankle-foot orthoses,
supported the use of hinged AFOs to increase peak
dorsiflexion and gait speed and decrease energy
expenditure.6 Although studies involving large
samples have been carried out, the quality of the
conclusions is often limited by their retrospective
methodology.10–13
The primary aim of this study, therefore, was to
evaluate the general effectiveness of AFOs, regard-
less of their nature, on gait parameters, balance,
gross motor function and activities of daily living
in ambulant children with all types of cerebral
palsy. The secondary objectives were (a) to evalu-
ate the effect of each type of orthosis on gait in
children with cerebral palsy and (b) to evaluate the
effectiveness of orthoses on gait as a function of
the unilateral or bilateral nature of the cerebral
palsy. A systematic literature review was carried
out with meta-analysis when possible.
Methods
Search strategy
Five databases were analyzed: Pubmed, Psychinfo,
Web of Science, Academic Search Premier and The
Cochrane Library. The following combinations of
keywords were used: “Cerebral Palsy or Hemiplegia
or Diplegia” AND “Gait or Walk or
Electromyography or Muscle Activation or
Activity of Daily Living or Motor Function or
Balance or Participation” AND “Ankle-foot
orthoses or Orthosis or Orthotic or Brace,” with no
limits (Supplementary Material 1). Two investiga-
tors independently performed the searches for each
database.
Articles were included if (a) they were written
in English or French, (b) were published before 13
January 2018 and (c) if the main results reported
the effects of AFOs compared with a control condi-
tion (barefoot or with shoes) in children with cere-
bral palsy (aged 0–18 years). All orthoses that
encompassed the ankle joint and the whole or part
of the foot were considered as AFOs.14 Articles that
evaluated functional electrical stimulation devices,
elastic derotation devices, the rehabilitation effect
of AFOs (evaluation of motor tasks without the
AFO) and shoes that support the ankle but were not
AFOs were thus excluded. Retrospective articles,
case reports and case series were also excluded.
The references of the selected articles were then
screened to complete the review process.
References from previous systematic reviews on
the same topic were also checked.5–7 The titles,
abstracts and whole texts of the articles identified
by the search were independently evaluated by the
two examiners. Any disagreements were resolved
by discussion with a third examiner.
The methodological quality of each of the
selected articles was evaluated by two independent
examiners using a modified PEDro scale15
(Supplementary Material 2). The three criteria
regarding study blinding were removed16 and one
criterion (description of the orthoses) was
added.17,18 For the latter criterion, 1 point was
awarded if at least three of four parameters were
specified: material, ankle plantarflexion angle,
anatomical limits and manufacturer. The maximum
Lintanf et al.
score was 9 points. Any disagreement on these
points was resolved with a third examiner. Articles
with a score of 0–4 were considered to be of “poor
quality,” a score of 5 or 6 was considered to be of
“moderate quality,” a score of 7 or 8 was “good”
and 9 was “excellent.” Data from articles with a
score of less than 5 were not reported
(Supplementary Material 3).
For each article, the two independent examiners
collected information regarding the sample, meth-
odology, type of orthosis and the parameters stud-
ied (gait, balance, gross motor function and
activities of daily living). For the evaluation of
gait, because of the multiplicity of gait parameters,
14 parameters were considered as clinically perti-
nent and were defined a priori. These included 8
kinematic, 1 kinetic, 1 energetic, 1 electromyo-
graphic and 3 spatio-temporal parameters (Table
1). Articles that did not provide any of these data
were excluded38 The following types of AFOs were
identified: solid, hinged, supra-malleolar, dynamic
(flexible posterior orthosis that encompasses the
whole calf and foot, made with a material that
allows plantarflexion) and ventral shell. Solid,
hinged, supra-malleolar and dynamic orthoses are
collectively termed “posterior orthoses” hereafter.
Any inconsistencies in the data collected by the
examiners were resolved by discussion and, if nec-
essary, the authors of the original study were con-
tacted for clarification, to obtain quantitative data
for the meta-analysis and/or to complete missing
data points.
When possible for the meta-analyses, continu-
ous outcomes were pooled according to their stand-
ardized mean differences (SMDs) (dimensionless
value) and 95% CIs (RevMan version 5.3). In order
to limit heterogeneity, only the results of the arti-
cles in which posterior AFOs were evaluated in
children with spastic cerebral palsy and equinus
gait were pooled in the meta-analysis. However,
because of the remaining variability of both chil-
dren with cerebral palsy and of orthoses used, a
distribution of effects was expected. A random-
effects model was therefore used. Consistency was
evaluated using I2. Publication bias was assessed
using funnel plots. An SMD of between 0.2 and 0.3
was considered as small, of 0.5 was moderate and
3
above 0.8 was regarded as large. Values of P < 0.05
were considered as significant. Subgroup analyses
were conducted for each type of orthosis and cere-
bral palsy type (unilateral and bilateral). The results
which could not be pooled in the meta-analysis
were reported in the systematic review. When an
article tested several orthoses, the analysis was car-
ried out as if each evaluation was independent. In
this review, the word “study” is used to describe
the comparison between one orthosis and the con-
trol condition.
Results
The search identified 4866 articles. Thirty-two arti-
cles that fulfilled the selection and quality criteria
were included (Figure 1). Twenty-three of the arti-
cles (38 studies) were of moderate qual-
ity,19–24,29–31,33,37,39–50 nine articles (18 studies) were
of good quality25–28,32,34,36,51,52 and no article was
classified as excellent quality (Supplementary
Material 2). Fifteen of the 32 articles reported
multiple follow-up assessments or different study
cohorts,19–21,24–32,34,39,40,50 resulting in a total of 56
different studies. Data from 35 studies could be
pooled in the meta-analysis and data from the 21
remaining studies were included in the systematic
review. The funnel plots showed no publication bias.
The methodological characteristics of the arti-
cles are presented in Table 2. The 32 articles
included 544 children (mean age = 7.9 years,
SD 2.1) and encompassed 56 studies with a
total of 884 children. In 51 studies, only
children with spastic cerebral palsy were inclu-
ded.1,19–29,31–34,36,37,39–41,43–45,47,48,51 In one study of
28 children, 2 had ataxic cerebral palsy and 26 had
spastic cerebral palsy.42 In 4 studies, the type of
cerebral palsy was not reported.49,50,52 Thirty stud-
ies included children with bilateral cerebral
palsy,19,20,24,26–29,31,35,36,39,41,43,44,46,47,51,52 18 included
children with unilateral cerebral palsy,21–
23,30,32,34,40,48 and 7 included children with unilateral
and bilateral cerebral palsy.25,33,42,45,49,50 In all the
studies, the children served as their own controls.
The gait pattern was described in 27 studies.
Control conditions were barefoot in 32 stud-
ies,22–25,27,28,30,32,33,36,38–40,43–45,48–50 with shoes in 20
4 Clinical Rehabilitation 00(0)

Table 1. Results of pooled analysis of the effects of AFOs on gait.

Outcomes References No. of studies No. of subjects SMD (95% CI) P-value I2 (%)
Ankle dorsiflexion 19–28 28 384 1.65 <0.001 83
at initial contact (1.47, 1.83)
Ankle dorsiflexion 20,21,23,26,28 16 226 1.34 <0.001 74
in swing phase (1.13, 1.56)
PODCI sports and 24 2 30 0.91 0.001 81
physical function (0.36, 1.46)
Stride length 19–32 32 444 0.88 <0.001 82
(0.73, 1.03)
Cadence 19–31,33–35 33 468 –0.72 <0.001 77
(–0.86;–0.59)
Peak power 20,21,23,26–29 18 248 –0.72 <0.001 72
generation in (–0.91;–0.53)
stance phase
PEDI mobility 21,25 6 135 0.57 <0.001 0
(0.33, 0.81)
Hip flexion at 20,26,28,34–36 12 110 0.33 0.02 0
initial contact (0.06, 0.60)
GMFM D standing 21,24,25,37 9 188 0.30 0.004 26
(0.10, 0.51)
Speed 19–31,33–35 33 468 0.28 <0.001 76
(0.14, 0.41)
GMFM E walking 21,24,26,37 9 188 0.24 0.02 0
running jumping (0.04;0.45)
PODCI transfers 24 2 30 0.20 0.47 89
(–0.33, 0.73)
Hip flexion in swing 20,25,34 5 50 0.15 0.45 0
phase (–0.24, 0.55)
Tibialis anterior 19,26,28,34 4 44 0.11 0.59 0
activity in stance (–0.31, 0.54)
phase
Knee flexion in swing 19,20,22,23,25,26,34 14 181 0.09 0.38 0
phase (–0.12, 0.3)
Knee extension in 19,20,23,34 8 103 –0.04 0.77 0
stance phase (–0.31, 0.23)
Hip extension in 20,26 5 44 –0.03 0.9 0
stance phase (–0.45, 0.39)
O2 consumption 21,23,25 7 146 0.00 1.0 52
(–0.23, 0.23)
Knee flexion at initial 19,20,22,23,25,26,28 18 211 0.00 0.98 53
contact (–0.20, 0.19)

AFO: ankle-foot orthosis; SMD: standardized mean difference; CI: confidence interval; PODCI: Pediatric Outcomes Data Collec-
tion Instrument; PEDI: Pediatric Evaluation of Disability Inventory; GMFM: Gross Motor Function Measure.
Values in bold represent statistically significant result (P < 0.05). Results are classified by decreasing effect size.

studies,19–21,26,29,31,41,46,47,52 barefoot and shoes in 2 AFOs,21,24,26,31,32,34,35,40,50 18 evaluated hinged

studies34 and no information was given in 2 stud- AFOs,19–24,26,27,30,31,36,39–41,43,48,50 11 evaluated solid
ies.37,42 Ten studies evaluated dynamic AFOs,19,21,25–31,39,40 5 evaluated supra-malleolar
Lintanf et al.
Figure 1. Flowchart.
orthoses,20,25,28,29,37 2 evaluated ventral shell
AFOs44,46 and nine evaluated the children’s own
AFOs (the orthosis usually worn by the child, out
with the study).
Forty-six studies reported results for gait over
level ground, 15 reported the scores of functional
scales,21,24,26,37,40,41,50 2 reported standing balance
outcomes,36,44 3 reported sit-to-stand transfer
parameters,39,43 1 study was about gait over level
ground with obstacles,42 3 studies reported on stair
climbing40 and 2 studies were concerned with
activities of daily living.46,52
Global effects on level ground walking
Although one study found a decrease in speed45
and 4 studies found no change in the non-pooled
data, there was an increase in speed with AFOs
according to the other non-pooled data and the data
pooled in the meta-analysis (SMD = 0.28 (0.14,
0.41), P < 0.001, n = 468, 33 studies).19–33,35,50,51
AFOs decreased cadence (SMD = –0.72 (–0.86;
–0.59), P < 0.001, n = 468, 33 studies)19–33 and
increased stride length (SMD = 0.88 (0.73, 1.03),
5
P < 0.001, n = 444, 32 studies).19–32 There was no
significant difference in O2 consumption with
AFOs according to the data pooled in the meta-
analysis (SMD = 0.00 (–0.23, 0.23), P = 1.0, n = 146,
7 studies).21,23,26 The results of the non-pooled
studies varied: 3 found no change,31 3 found a
decrease in energy cost, 1 found a tendency toward
a decrease and 1 found an increase in energy cost.45
Global effects on functional scales
Different scales were used: 10 studies used the
GMFM (Gross Motor Function Measure),21,24,26,37,41
9 used the PEDI (Pediatric Evaluation of Disability
Inventory),21,26,40 and 2 used the PODCI (Pediatric
Outcomes Data Collection Instrument).24 Although
one study found no change in the GMFM (sections
D: standing and E: walking/running/jumping) with
AFOs,41 the pooled analysis found improvements
in sections D (SMD = 0.30 (0.10, 0.51); P = 0.004,
n = 188, 9 studies)21,24,26,37 and E (SMD = 0.24
(0.04, 0.45); P = 0.02, n = 188, 9 studies).21,24,26,37
The pooled analysis showed an improvement in the
mobility score of the PEDI scale (SMD = 0.57
 6

Table 2. Description of the articles.

Motor Task Author year Population AFO Intervention Outcome measures

Type Side Gait pattern Number Age (SD) GMFCS Type Usually Study Control Duration
affected of children wears an no. condition
(male/female) AFO?

Level ground
walking
Mossberg et al. Spastic Bilateral – 18 (10/8) 8.3 (2.8) – Own AFO Y 1 BF Test day Energy cost (at self-
199051 selected speed)
Radtka et al. Spastic Uni-/ Equinus 10 (6/4) 6.5 (1.9) – SAFO; Y 2 BF 1 Mth Spatio-temporal,
199725 bilateral SMO kinematic, EMG
Carlson et al. Spastic Bilateral Equinus/ 11 (6/5) 6.9 (0.7) – SAFO; Y 2 S 4 Mth Spatio-temporal,
199729 crouch SMO kinematic, kinetic
Brunner et al. Spastic Unilateral – 14 (6/8) – SAFO; – 2 BF Test day Spatio-temporal,
199830 HAFO kinematic, kinetic
Rethlefsen et al. Spastic Bilateral – 21 (13/8) 9.1 (2.2) – SAFO; Y 2 S 4–6 Wk Kinematic, EMG
199919 HAFO
Crenshaw et al. Spastic Bilateral – 8 (5/3) 8.4 (1.9) – 2 HAFOs; Y 3 S 4 Wk Spatio-temporal,
200020 SMO kinematic, kinetic
Maltais et al. Spastic Bilateral – 10 (8/2) 9 (2.1) 9 I/1 II HAFO Y 1 S 4–6 Wk Energy cost (at
200141 self-selected speed),
functional
Buckon et al., Spastic Unilateral – 30 (21/9) – SAFO; Y 3 S 3 Mth Spatio-temporal,
200121 HAFO; kinematic, kinetic,
DAFO energy cost (at self-
selected speed/fast
walking), functional
Dursun et al. Spastic Uni-/ Equinus 24 (10/14) 6.7 (0.7) – own AFO Y 1 BF Test day Spatio-temporal
200233 bilateral
Smiley et al. Spastic Bilateral – 14 (8/6) – SAFO; – 3 S Test day Spatio-temporal,
200231 HAFO; kinematic, energy
DAFO cost (at self-selected
speed)
Buckon et al. Spastic Bilateral – 16 (10/6) 8.4 (2.3) 4 I/12 II SAFO; Y 3 S 3 Mth Spatio-temporal,
200426 HAFO; kinematic, kinetic,
DAFO energy cost (at self-
selected speed/fast
walking), functional
Lam et al. Spastic Bilateral Equinus 13 (7/6) 5.9 (1.8) – SAFO; – 2 BF Test day Spatio-temporal,
200528 SMO kinematic, kinetic,
EMG
Clinical Rehabilitation 00(0)
Table 2. (Continued)

Motor Task Author year Population AFO Intervention Outcome measures

Type Side Gait pattern Number Age (SD) GMFCS Type Usually Study Control Duration
Lintanf et al.

affected of children wears an no. condition

(male/female) AFO?

Radtka et al. Spastic Bilateral Equinus 12 (6/6) 7.5 (3.8) – SAFO; Y 2 BF 2 Wk Spatio-temporal,
200527 HAFO kinematic, kinetic,
EMG
Desloovere et al. Spastic Unilateral Equinus 5.86 (1.8) – 2 DAFOs Y 2 BF + S Test day Spatio-temporal,
200634 kinematic, kinetic
Bjornson et al. Spastic – – 4.3 (1.5) I/II/III SMO Y 1 – Test day Functional
200637
Romkes et al. Spastic Unilateral Equinus 10 (4/6) 9.7 (1.6) 8 I/2 II HAFO Y 1 BF Test day Spatio-temporal,
200622 kinematic, EMG
Balaban et al. Spastic Unilateral Equinus 11 (7/4) 7.2 (1.2) – HAFO Y 1 BF Test day Spatio-temporal,
200723 kinematic, kinetic,
energy cost (at 0.5 m/
sec)
Smith et al. Spastic Bilateral Jump gait 7.5 (2.9) 15 I HAFO; – 2 BF 4 Wk Spatio-temporal,
200924 DAFO kinematic, kinetic,
functional
Vanwala et al. Spastic Uni-/ – – – – 1 BF Test day Spatio-temporal,
201445 bilateral energy cost
Kerkum et al. Spastic Bilateral Crouch 15 (11/4) 10 (2) 2 I/11 VSAFO Y 1 S 12– Spatio-temporal,
201646 II/2 III 20 Wk kinetic, energy cost,
activity of daily living
Bjornson et al. – Bilateral Equinus/ 1 I/9 own AFO Y 1 S 4 Wk Activities of daily
201652 jump/crouch II/1 III living
Wren et al. – Uni-/ Equinus/ 10 (4/6) 7.5 (2.18) 6 I/4 III DAFO; – 2 BF 4 Wk Spatio-temporal,
201550 bilateral crouch restricted kinematic, kinetic,
HAFO center of pressure
velocity and
displacement
Schmid et al. Spastic Unilateral – 10 (9/1) 12 (1.4) 9 I/1 II HAFO – 1 BF Test day Spatio-temporal,
201648 kinematic
Bhise Swati 201747 Spastic Bilateral – 10.0 (2.67) – own AFO Y 1 S Test day Spatio-temporal,
energy cost (at self-
selected speed)
Tavernese et al. Spastic Unilateral Equinus/ 8 (4/4) 8.0 (2.2) 8 II flex Y 2 BF Test day Spatio-temporal,
201732 recurvatum- DAFO/ kinematic, kinetic
crouch own AFO

(Continued)
7
 8

Table 2. (Continued)

Motor Task Author year Population AFO Intervention Outcome measures

Type Side Gait pattern Number Age (SD) GMFCS Type Usually Study Control Duration
affected of children wears an no. condition
(male/female) AFO?

Tavernese et al. Spastic Unilateral Equinus/ 7 (4/3) 7.2 (2.2) 4 I/3 II stiff AFO/ Y 2 BF Test day Spatio-temporal,
201732 recurvatum- own AFO kinematic, kinetic
crouch
Galli et al. 201649 – Unilateral 11 (7/4) Range = 4–13 I/II own AFO Y 1 BF Test day Gait profile score and
(SMO- gait variable score
HAFO)
Galli et al. 201649 – Uni-/ 10 (5/5) Range = 5–14 I/II own AFO Y 1 BF Test day Gait profile score and
bilateral (SMO- gait variable score
HAFO)
Quiet standing
Rha et al. 201036 Spastic Bilateral Equinus 21 (11/10) 6.1 (1.1) 4 I/13 HAFO – 1 BF Test day Center of pressure’s
II/4 III velocity and
displacement
Bahramizadeh et Spastic Bilateral Crouch 8 (2/6) 8.1 (2.4) I/II VSAFO Y 1 BF Test day Center of pressure’s
al. 201244 velocity and
displacement
Sit to stand transfer
Wilson et al. Spastic Bilateral Equinus – SAFO; – 2 BF Test day Temporal, kinematic
199739 HAFO
Park et al. 200443 Spastic Bilateral – 3.7 (1.1) – HAFO – 1 BF Test day Temporal, kinematic,
kinetic
Gait with obstacle crossing
Kott and Held Spastic/ Uni-/ – 28 (20/8) 10.6 (4.5) 19 I/9 II own AFO Y 1 – Test day Functional
200242 Ataxic bilateral/
other
Stair, climbing/descent
Sienko-Thomas et Spastic Unilateral – 19 (11/8) – SAFO; Y 3 BF 3 Mth Spatio-temporal,
al. 200240 HAFO; kinematic, functional
DAFO

AFO: ankle-foot orthosis; BF: barefoot; SAFO: solid ankle-foot orthosis; SMO: supra-malleolar ankle-foot orthosis; HAFO: hinged ankle-foot orthosis; DAFO: dynamic ankle-foot orthosis; VSAFO:
ventral shell ankle-foot orthosis; Y: yes; N: no; EMG: electromyography; S: with shoes; Mth: Month(s); Wk: Week(s).
Bold represents articles pooled in the meta-analysis.
Clinical Rehabilitation 00(0)
Lintanf et al.
(0.33, 0.81), P < 0.001, n = 135, 6 studies).21,26
Improvements in the PODCI score (sports and
physical function items) were also found for the
pooled results of 2 studies (SMD = 0.91 (0.36,
1.46), P = 0.001, n = 30, 2 studies);24 however, there
was no effect on PODCI score for transfers
(SMD = 0.20 (–0.33, 0.73), P = 0.47, n = 30, 2
studies).24
Global effects on balance
Four studies found no significant effect of AFOs on
standing postural control in children with bilateral
cerebral palsy.36,44,50 Use of AFOs did not change
the pediatric balance scale score in one study.42
Global effects on sit to stand transfers
Two studies found a significant decrease in sit-to-
stand transfer time as well as kinematic changes in
young children with bilateral cerebral palsy with
AFOs.39,43
Global effects on gait with obstacle
crossing and stair climbing
One study evaluated obstacle crossing in children
with different types of cerebral palsy who were
wearing their own orthoses and found no signifi-
cant effect of the orthoses.42 Use of AFOs did not
reduce stair climbing ability and they improved
foot contact position during stair ascent.40
Activities of daily living
AFOs did not change the total daily strides in aver-
age hour of walking per day or the peak activity
index.46,52
Effect of each type of orthosis on gait
For children with equinus gait, use of a posterior
AFO increased ankle dorsiflexion at initial contact
(SMD = 1.65 (1.47, 1.83), P < 0.001, n = 384 chil-
dren, 28 studies)19–28,32,34,48 and during swing
(SMD = 1.34 (1.13, 1.56), P < 0.001, n = 226, 16
studies)20,21,23,26,28,32,34 compared with the control
9
conditions (Supplementary Material 4). The meta-
analysis found no significant difference in knee
flexion at initial contact (SMD = 0.00 (–0.20, 0.19),
P = 0.98, n = 211, 18 studies),19,20,22,23,25,26,32 during
swing phase (SMD = 0.09 (–0.12, 0.3), P = 0.38,
181 children, 14 studies)19,20,22,23,25,26,28 and in knee
extension during stance phase (SMD = –0.04
(–0.31, 0.23), P = 0.77, n = 103, 8 studies).19,20,23,28
The results of 5 non-pooled studies in children with
unilateral cerebral palsy did not show any real
change in knee kinematics with AFOs.21,34 Hip
flexion at initial contact increased (SMD = 0.33
(0.06, 0.60), P = 0.02, n = 110, 12 stud-
ies).20,25,27,28,32,34 No other kinematic changes were
found for the hip in the meta-analysis.
There was a decrease in peak power generation
in stance in the studies included in the meta-analy-
sis (SMD = –0.72 (–0.91; –0.53), P < 0.001, n = 248,
18 studies) and in the non-pooled studies included
in the systematic review.20,21,23,26–29,32,34 One study
found a decrease in tibialis anterior muscle activity
at initial contact using AFOs in the systematic
review.22 However, neither the other studies nor the
pooled analysis found a change in tibialis anterior
activation time (SMD = 0.11 (–0.31, 0.54), P = 0.59,
n = 44, 4 studies of tibialis anterior activation in
stance phase).19,25,27,28
More specifically, hinged AFOs improved gait
parameters significantly more than other orthoses,
particularly gait speed (SMD = 0.35 (0.11, 0.59),
P = 0.004, n = 158, 11 studies).19–24,26,27,30,31
Dynamic AFOs produced the greatest effect sizes
on stride length (SMD = 1.47 (1.12, 1.82), P < 0.001,
6 studies, n = 89),21,24,26,31 cadence (SMD = –1.16
(–1.50; –0.82), P < 0.001, 6 studies, n = 89),21,24,26,31
and ankle kinematics at initial contact (SMD = 2.13
(1.68, 2.58), P < 0.001, 5 studies, n = 75).21,26,32
Supra-malleolar orthoses did not reduce peak ankle
power generation in stance phase (SMD = 0.00
(–0.5, 0.49), P = 0.99, 3 studies, n = 32),20,28,29 pro-
duced the smallest effect sizes on the ankle at ini-
tial contact (SMD = 0.85 (0.38, 1.32), P < 0.001, 4
studies, n = 42)20,25,28 and during swing phase
(SMD = 0.81 (0.16, 1.46), P = 0.02, 2 studies,
n = 21)20,28 and also on cadence (SMD = –0.44
(–0.88, 0.00), P = 0.05, 4 studies, n = 42)20,25,28,29
Solid AFOs reduced peak power generation with a
10
large effect size (SMD = –1.24 (–1.61; –0.86),
P < 0.001, 4 studies, n = 69).21,26–29 None of these
orthoses induced significant kinematic changes at
the hip or knee.
Ventral shell AFOs reduced knee flexion in
stance by 2.4°, and a tendency toward a reduction
in energy cost was found in one study.46
Effects of AFOs on gait as a function of
type of cerebral palsy
In the pooled results, gait speed only increased sig-
nificantly with AFOs in children with unilateral
cerebral palsy (SMD = 0.45 (0.21–0.69), P < 0.001,
n = 174, 13 studies) (Supplementary Material 5).21–
23,25,26,30,32,34 The effect sizes for the parameters that
changed significantly with AFOs (stride length,
cadence and ankle dorsiflexion at initial contact
and during swing) were greater in children with
unilateral cerebral palsy than in those with bilateral
cerebral palsy. The decrease in peak ankle power
generation was similar in children with unilateral
and bilateral cerebral palsy. Oxygen consumption
decreased significantly only in children with bilat-
eral cerebral palsy (SMD = –0.51 (–0.91; –0.10),
P = 0.02, n = 48, 3 studies),26 a similar trend was
found in four of the eight studies included in the
systematic review.41,46,47,51 Regarding kinematic
scores, two studies reported that the gait profile
score described by Baker35 was significantly
reduced in children with unilateral cerebral palsy
and unchanged in children with bilateral cerebral
palsy.49
Discussion
This systematic review and meta-analysis reports
current evidence and quantified effectiveness of
AFOs in children with cerebral palsy, according to
the International Classification of Functioning dis-
ability and health (ICF) body function and struc-
ture (mechanical) and activity categories. The
studies found included a large majority of children
with spastic cerebral palsy, and most evaluated
posterior AFOs. The results showed a strong level
of evidence for small-to-large improvements in
spatio-temporal parameters using AFOs, and a
Clinical Rehabilitation 00(0)
moderate level of evidence for small-to-moderate
improvements in gross motor function. There was
a low level of evidence (from non-pooled data) that
AFOs affected neither balance nor activities of
daily living. A strong level of evidence was found
for small-to-large changes in biomechanical
parameters during level ground walking using pos-
terior AFOs in children with equinus gait (increased
ankle dorsiflexion at initial contact and during
swing). Some specific effects of type of orthosis as
well as the effects in children with unilateral and
bilateral cerebral palsy were highlighted.
There are some limitations to the conclusions
drawn by this review. The strength of the conclu-
sions depends on the completeness of the data
reported. Non-bibliographic database sources were
not searched and articles in languages other than
English and French were not included, so some
articles may have been missed. In order to be as
complete as possible, studies reported within the
same article were analyzed separately, which may
have increased the weighting of some of the data.
However, the funnel plots for the meta-analysis
showed no publication bias. The articles included
were of moderate to good quality, but, apart from
one study,24 the number of subjects needed for sta-
tistical power was not calculated, which reduced
the strength of conclusions that could be drawn
from the results. For future studies, randomization
and calculation of the number of subjects needed
would improve robustness and the reliability and
quality of evidence.16 Despite the exclusion of arti-
cles evaluating ventral shell AFOs, and those arti-
cles where it was impossible to determine if the
children had an equinus gait in the meta-analysis,
the I2 results from 9 of 19 articles (Table 1) for the
total study population suggested that the study
designs and results were heterogeneous, a fact
which can limit the generalization of results. This
heterogeneity persisted in the subgroup analysis
for each type of cerebral palsy and orthosis, prob-
ably because of the diversity of the population of
children with cerebral palsy, different gait patterns,
tests conditions and designs of orthoses. More spe-
cific subgroup analysis, for example, of children
with the same gait pattern and the same AFO would
have provided more homogeneous results but were
Lintanf et al.
not possible because of a lack of description of
study samples and orthoses. Most articles provided
partial information regarding gait patterns, but not
clear kinematic data. Similarly, the description of
each orthosis should have included its goal, type,
material, full details of design, stiffness and ankle
angle;13,17,18,53 however, this was rarely provided. It
is thus not possible to conclude regarding the bio-
mechanical effect of the different orthoses in sub-
groups with specific gait patterns.
This meta-analysis and systematic review
reports evidence of a positive effect of AFOs on
gross motor function (ICF: activity) in children
with cerebral palsy, as shown by the D and E sub-
sections of the GMFM, the mobility subsection of
the PEDI and the sports and physical function sub-
section of the PODCI. The results showed AFOs
produce a large increase in stride length, and a mod-
erate decrease in cadence, inducing a small increase
in gait speed. In children with equinus gait who
used posterior AFOs, these effects were largely due
to large increases in ankle dorsiflexion in the stance
phase and moderate increases in dorsiflexion in the
swing phase. No moderate or large changes in hip
or knee kinematics were found. Some parameters
that could have improved gross motor function
(ICF: activity) were not modified or were actually
worsened with AFOs (balance, oxygen consump-
tion, peak ankle power generation). Thus, it seems
reasonable to attribute the small-to-moderate
improvement in gross motor function to the distal
biomechanical effects of AFOs in those children.
There was a positive relationship between the bio-
mechanical effect and the effect on activity with
AFOs. Most of the studies in this review evaluated
gait during level ground walking, therefore more
studies are needed to evaluate the biomechanical
effect of AFOs in different conditions (e.g. stair
climbing and running) to increase understanding of
the improvements in gross motor function.
Interestingly, only two recent articles evaluated
activities of daily living as an outcome. Given the
paucity of data in this area and low level of evi-
dence regarding the effects of AFOs on activities of
daily living and participation, we believe that more
studies should be carried out.
11
The main results of this study were reported
for ambulant children with spastic cerebral palsy.
These data are useful for global decision making
in children with spastic cerebral palsy. Because
of the large majority of children with spastic cer-
ebral palsy included, results cannot be general-
ized to children with dyskinetic and ataxic
cerebral palsy. Subgroup analyses were con-
ducted to determine specific effects in children
with unilateral and bilateral cerebral palsy, as
well as for posterior AFOs and other types of
orthosis, data which are more pertinent for clini-
cal practice. The results suggested AFOs improve
gait parameters (step length, ankle kinematics
and gait speed) more in children with unilateral
cerebral palsy than with bilateral cerebral palsy.
Reducing energy cost may be a better target for
children with bilateral cerebral palsy. Regarding
the type of orthosis, hinged AFOs increased gait
speed, dynamic AFOs had a large effect on ankle
kinematics at initial contact and stride length,
supra-malleolar orthoses did not alter peak ankle
power generation and solid AFOs had a large
negative effect on propulsion. These results from
the meta-analysis help to understand the effects
of AFOs and can be used to guide the prescription
of orthoses in a specific subgroup of children
with cerebral palsy, depending on the improve-
ment sought. For example, the results suggest
that if the aim is to improve the gait speed in a
child with unilateral cerebral palsy, a hinged AFO
would be the most appropriate. In a child with
either unilateral or bilateral cerebral palsy, if the
aim is to improve ankle kinematics, a dynamic
AFO would be appropriate. Because of the lack
of studies of ventral shell AFOs, no conclusions
can be drawn regarding their efficacy. The analy-
sis of other specific subgroups such as different
gait patterns was limited by the lack of descrip-
tion in the articles included. Further robust stud-
ies are required to provide high levels of evidence
for different subgroups and specific gait patterns.
Such studies would increase understanding of the
biomechanical effects of AFOs, as well as the
relationships between body function and struc-
ture (mechanical) and activity.
12
Clinical Messages
•• Ankle-foot orthoses improve slightly gait
speed and small to moderately gross
motor function in children with spastic
cerebral palsy.
•• Posterior orthoses induce large changes
in ankle kinematics in children with equi-
nus gait.
•• Few data on the effect of ankle-foot
orthoses on activities of daily living are
available.
Acknowledgements
The authors sincerely thank Johanna Robertson for her
help with the English translation of this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest
with respect to the research, authorship and/or publica-
tion of this article.
Funding
The author(s) received no financial support for the
research, authorship and/or publication of this article.
Supplementary material
Supplementary material is available for this article
online.
ORCID iD
Mael Lintanf https://orcid.org/0000-0003-1489-7456
References
1. Bjornson KF, Zhou C, Stevenson RD, et al. Relation of
stride activity and participation in mobility-based life hab-
its among children with cerebral palsy. Arch Phys Med
Rehabil 2014; 95(2): 360–368.
2. Wingstrand M, Hägglund G and Rodby-Bousquet E.
Ankle-foot orthoses in children with cerebral palsy: a
cross sectional population based study of 2200 children.
BMC Musculoskelet Disord 2014; 215: 327.
3. Graham HK, Rosenbaum P, Paneth N, et al. Cerebral
palsy. Nat Rev Dis Primers 2016; 2: 15082.
4. Sacaze E, Garlantezec R, Rémy-Néris O, et al. A survey
of medical and paramedical involvement in children with
cerebral palsy in Britanny: preliminary results. Ann Phys
Rehabil Med 2013; 56(4): 253–267.
Clinical Rehabilitation 00(0)
5. Figueiredo EM, Ferreira GB, Moreira RC, et al. Efficacy
of ankle-foot orthoses on gait of children with cerebral
palsy: systematic review of literature. Pediatr Phys Ther
2008; 20(3): 207–223.
6. Neto HP, Grecco LA, Galli M, et al. Comparison of articu-
lated and rigid ankle-foot orthoses in children with cer-
ebral palsy: a systematic review. Pediatr Phys Ther 2012;
24(4): 308–312.
7. Aboutorabi A, Arazpour M, Bani MA, et al. Efficacy of
ankle foot orthoses types on walking in children with cer-
ebral palsy: a systematic review. Ann Phys Rehabil Med
2017; 60: 393–402.
8. Morris C, Bowers R, Ross K, et al. Orthotic management
of cerebral palsy: recommendations from a consensus
conference. Neurorehabilitation 2011; 28(1): 37–46.
9. Kerkum YL, Brehm M-A, Buizer AI, et al. Defining
the mechanical properties of a spring-hinged ankle
foot orthosis to assess its potential use in children with
spastic cerebral palsy. J Appl Biomech 2014; 30(6):
728–731.
10. Van Gestel L, Molenaers G, Huenaerts C, et al. Effect of
dynamic orthoses on gait: a retrospective control study in
children with hemiplegia. Dev Med Child Neurol 2008;
50(1): 63–67.
11. Ries AJ, Novacheck TF and Schwartz MH. The efficacy
of ankle-foot orthoses on improving the gait of children
with diplegic cerebral palsy: a multiple outcome analysis.
PM R 2015; 7: 922–929.
12. Lucareli PR, Mde O, Lucarelli JG, et al. Changes in joint
kinematics in children with cerebral palsy while walk-
ing with and without a floor reaction ankle-foot orthosis.
Clinics 2007; 62(1): 63–68.
13. Brehm M-A, Harlaar J and Schwartz M. Effect of ankle-
foot orthoses on walking efficiency and gait in chil-
dren with cerebral palsy. J Rehabil Med 2008; 40(7):
529–534.
14. BS 7313-1.1:1990. Prosthetics and orthotics. Terminology.
Glossary of general terms relating to external limb pros-
theses and external orthoses.
15. Maher CG, Sherrington C, Herbert RD, et al. Reliability
of the PEDro scale for rating quality of randomized con-
trolled trials. Phys Ther 2003; 83(8): 713–721.
16. Tyson SF, Sadeghi-Demneh E and Nester CJ. A system-
atic review and meta-analysis of the effect of an ankle-foot
orthosis on gait biomechanics after stroke. Clin Rehabil
2013; 27(10): 879–891.
17. Ridgewell E, Dobson F, Bach T, et al. A systematic
review to determine best practice reporting guidelines for
AFO interventions in studies involving children with cer-
ebral palsy. Prosthet Orthot Int 2010; 34(2): 129–145.
18. Eddison N, Mulholland M and Chockalingam N. Do
research papers provide enough information on design
and material used in ankle foot orthoses for children with
cerebral palsy? A systematic review. J Child Orthop
2017; 11(4): 263–271.
19. Rethlefsen S, Kay R, Dennis S, et al. The effects of
fixed and articulated ankle-foot orthoses on gait patterns
Lintanf et al.
in subjects with cerebral palsy. J Pediatr Orthop 1999;
19(4): 470–474.
20. Crenshaw S, Herzog R, Castagno P, et al. The efficacy of
tone-reducing features in orthotics on the gait of children
with spastic diplegic cerebral palsy. J Pediatr Orthop 2000;
20(2): 210–216.
21. Buckon CE, Thomas SS, Jakobson-Huston S, et al.
Comparison of three ankle-foot orthosis configurations
for children with spastic hemiplegia. Dev Med Child
Neurol 2001; 43(6): 371–378.
22. Romkes J, Hell AK and Brunner R. Changes in mus-
cle activity in children with hemiplegic cerebral palsy
while walking with and without ankle-foot orthoses. Gait
Posture 2006; 24(4): 467–474.
23. Balaban B, Yasar E, Dal U, et al. The effect of hinged
ankle-foot orthosis on gait and energy expenditure in
spastic hemiplegic cerebral palsy. Disabil Rehabil 2007;
29(2): 139–144.
24. Smith PA, Hassani S, Graf A, et al. Brace evaluation in
children with diplegic cerebral palsy with a jump gait pat-
tern. J Bone Joint Surg Am 2009; 91(2): 356–365.
25. Radtka SA, Skinner SR, Dixon DM, et al. A comparison
of gait with solid, dynamic, and no ankle-foot orthoses
in children with spastic cerebral palsy. Phys Ther 1997;
77(4): 395–409.
26. Buckon CE, Thomas SS, Jakobson-Huston S, et al.
Comparison of three ankle-foot orthosis configurations
for children with spastic diplegia. Dev Med Child Neurol
2004; 46(9): 590–598.
27. Radtka SA, Skinner SR and Johanson ME. A compari-
son of gait with solid and hinged ankle-foot orthoses in
children with spastic diplegic cerebral palsy. Gait Posture
2005; 21(3): 303–310.
28. Lam WK, Leong JCY, Li YH, et al. Biomechanical and
electromyographic evaluation of ankle foot orthosis and
dynamic ankle foot orthosis in spastic cerebral palsy. Gait
Posture 2005; 22(3): 189–197.
29. Carlson WE, Vaughan CL, Damiano DL, et al. Orthotic
management of gait in spastic diplegia. Am J Phys Med
Rehabil 1997; 76(3): 219–225.
30. Brunner R, Meier G and Ruepp T. Comparison of a stiff
and a spring-type ankle-foot orthosis to improve gait in
spastic hemiplegic children. J Pediatr Orthop 1998;
18(6): 719–726.
31. Smiley SJ, Jacobsen FS, Mielke C, et al. A comparison of
the effects of solid, articulated, and posterior leaf-spring
ankle-foot orthoses and shoes alone on gait and energy
expenditure in children with spastic diplegic cerebral
palsy. Orthopedics 2002; 25(4): 411–415.
32. Tavernese E, Petrarca M, Rosellini G, et al. Carbon
modular orthosis (Ca.M.O.): an innovative hybrid modu-
lar ankle-foot orthosis to tune the variable rehabilitation
needs in hemiplegic cerebral palsy. Neurorehabilitation
2017; 40(3): 447–457.
33. Dursun E, Dursun N and Alican D. Ankle-foot orthoses:
effect on gait in children with cerebral palsy. Disabil
Rehabil 2002; 24(7): 345–347.
13
34. Desloovere K, Molenaers G, Van Gestel L, et al. How can
push-off be preserved during use of an ankle foot ortho-
sis in children with hemiplegia? A prospective controlled
study. Gait Posture 2006; 24(2): 142–151.
35. Baker R, McGinley JL, Schwartz MH, et al. The gait pro-
file score and movement analysis profile. Gait Posture
2009; 30(3): 265–269.
36. Rha D, Kim DJ and Park ES. Effect of hinged ankle-foot
orthoses on standing balance control in children with
bilateral spastic cerebral palsy. Yonsei Med J 2010; 51(5):
746–752.
37. Bjornson KF, Schmale GA, Adamczyk-Foster A, et al.
The effect of dynamic ankle foot orthoses on function
in children with cerebral palsy. J Pediatr Orthop 2006;
26(6): 773–776.
38. Swinnen E, Baeyens J-P, Van Mulders B, et al. The influ-
ence of the use of ankle-foot orthoses on thorax, spine, and
pelvis kinematics during walking in children with cerebral
palsy. Prosthet Orthot Int 2018; 42: 208–213.
39. Wilson H, Haideri N, Song K, et al. Ankle-foot orthoses
for preambulatory children with spastic diplegia. J Pediatr
Orthop 1997; 17(3): 370–376.
40. Sienko-Thomas S, Buckon CE, Jakobson-Huston S, et al.
Stair locomotion in children with spastic hemiplegia: the
impact of three different ankle foot orthosis (AFOs) con-
figurations. Gait Posture 2002; 16(2): 180–187.
41. Maltais D, Bar-Or O, Galea V, et al. Use of orthoses low-
ers the O(2) cost of walking in children with spastic cer-
ebral palsy. Med Sci Sports Exerc 2001; 33(2): 320–325.
42. Kott KM and Held SL. Effects of orthoses on upright
functional skills of children and adolescents with cerebral
palsy. Pediatr Phys Ther 2002; 14(4): 199–207.
43. Park ES, Park CI, Chang HJ, et al. The effect of hinged
ankle-foot orthoses on sit-to-stand transfer in children
with spastic cerebral palsy. Arch Phys Med Rehabil 2004;
85(12): 2053–2057.
44. Bahramizadeh M, Mousavi ME, Rassafiani M, et al. The
effect of floor reaction ankle foot orthosis on postural
control in children with spastic cerebral palsy. Prosthet
Orthot Int 2012; 36(1): 71–76.
45. Vanwala J, Diwan S, Shah S, et al. Effect of ankle foot
orthosis on energy expenditure index and gait speed in
spastic cerebral palsy children: an observational study. Int
J Contemp Pediatr 2014; 1(1): 17–19.
46. Kerkum YL, Harlaar J, Buizer AI, et al. An individual
approach for optimizing ankle-foot orthoses to improve
mobility in children with spastic cerebral palsy walking
with excessive knee flexion. Gait Posture 2016; 46: 104–
111.
47. Bhise S. Comparison between physiological cost index in
healthy normal children as against ambulatory spastic diple-
gic cerebral palsy (with and without orthosis) in the age
group 6 to 8 years. Int J Physiother 2017; 3(4): 395–400.
48. Schmid S, Romkes J, Taylor WR, et al. Orthotic correc-
tion of lower limb function during gait does not imme-
diately influence spinal kinematics in spastic hemiplegic
cerebral palsy. Gait Posture 2016; 49: 457–462.
14
49. Galli M, Cimolin V, Rigoldi C, et al. Quantitative evalu-
ation of the effects of ankle foot orthosis on gait in chil-
dren with cerebral palsy using the gait profile score and
gait variable scores. J Dev Phys Disabil 2016; 28(3):
367–379.
50. Wren TAL, Dryden JW, Mueske NM, et al. Comparison
of 2 orthotic approaches in children with cerebral palsy.
Pediatr Phys Ther 2015; 27(3): 218–226.
51. Mossberg KA, Linton KA and Friske K. Ankle-foot
orthoses: effect on energy expenditure of gait in spastic
Clinical Rehabilitation 00(0)
diplegic children. Arch Phys Med Rehabil 1990; 71(7):
490–494.
52. Bjornson K, Zhou C, Fatone S, et al. The effect of ankle-
foot orthoses on community-based walking in cerebral
palsy: a clinical pilot study. Pediatr Phys Ther 2016;
28(2): 179–186.
53. Harlaar J, Brehm M, Becher JG, et al. Studies examining
the efficacy of ankle foot orthoses should report activity
level and mechanical evidence. Prosthet Orthot Int 2010;
34(3): 327–335.