Keywords
hippotherapy - gross motor function - cerebral palsy
Introduction
Cerebral palsy (CP), a neurological disorder caused by nonprogressive and permanent
brain damage, can result in impairments of movement, posture, sensation, perception,
cognition, and motor control. Physical therapy (PT) is a common intervention for children
with CP to practice functional movements, improve motor control, reduce impairments,
and learn strategies that compensate for lost function.[1]
The World Health Organization's International Classification of Functioning, Disability
and Health: Child and Youth Version (ICF-CY) framework is based on a biopsychosocial
model that covers functioning and disability with its components, namely, body structure,
body functions, activities, and participation, and identifies the need to consider,
represented by personal and environmental factors. The ICF-CY defines activity as
“the execution of a task or action by an individual” and participation as “involvement
in a life situation.”[2] Diverse evidence indicates that physical activity has a beneficial effect on all
children, including children with disabilities.[3] Children with CP participate in leisure-time physical activities less often, with
less intensity and reduced diversity than their peers without disabilities.[4]
[5] Reduced activity and participation cause deterioration in body function and structure,
such as cardiovascular disease risk, bone density, metabolic disturbance risk, obesity,
and elevated blood pressures in children with CP.[6] However, many traditional physical therapies for children with CP focus on impairments
within the child at the domain of body function and structure. Damiano[7] stated that “treatments cannot be justified unless they produce a change in activity,
participation, or, more elusively, health-related quality of life for the person receiving
the intervention.” Furthermore, she stated that hippotherapy for muscle symmetry and
activities is regarded as a “good” treatment approach for children with CP.
Equine-assisted activities and therapies (EAATs) have been widely used as leisure
activities for children with CP since the 1960s.[8] Hippotherapy was developed to maximize the therapeutic effect of EAATs. The term
hippotherapy refers to how occupational therapy, physical therapy, and speech-language
pathology for professionals use evidence-based practice, and clinical reasoning in
the purposeful manipulation of equine movement to engage sensory, neuromotor, and
cognitive systems to achieve functional outcomes. The three-dimensional smooth, rhythmic,
and repetitive movements of the walking horse produce normalized pelvic movement similar
to the pelvic movement during ambulation.[9] The warmth and rhythmic movement of the horse are estimated to decrease hypertonicity
and increase range of motion. While adjusting the dynamic base of support from the
horse, children can improve trunk strength, balance, and motor planning.[10] Most importantly, children with CP recognize EAATs as pleasant activities rather
than therapies. Children with CP, like other children, want to be physically active,
have fun, and enjoy the sensation of speed.[11]
EAATs have been shown to improve gross motor function, postural control, balance,[10]
[12] gait,[13] and functional performance[14] in children with CP. Several meta-analyses and systematic reviews have shown that
EAATs positively affect gross motor function in children with CP.[10]
[12]
[15] Gross motor function measure (GMFM) is the most widely used outcome measure for
assessing changes in gross motor function over time in children with CP. Although
many studies have reported significant improvements in GMFM scores, a recent meta-analysis
showed a nonsignificant increase in GMFM scores after EAATs in children with CP.[16] In fact, the improvement in the GMFM score varies widely among individuals. However,
little research has been conducted on the factors that influence the gross motor outcome
of EAATs in children with CP. Therefore, this study aimed to identify the individual
factors influencing gross motor outcome in children with CP after hippotherapy.
Methods
Study Design
This was a retrospective study investigating children with CP who participated in
8 weeks of a hippotherapy program (30 minutes per session, twice a week, for a total
of 16 sessions) at one local riding arena, between October 2009 and August 2016. Clinical
information, including GMFCS (gross motor function classification system) level, age,
sex, CP distribution, CP type, gross motor function measure-88 (GMFM-88), GMFM-66,[17] and pediatric balance scale (PBS) were collected retrospectively. GMFM-88, GMFM-66,
and PBS have been used to measure the functional outcomes in participants after hippotherapy
since 2009 for quality control. Since a clinically meaningful improvement in GMFM-66
was reported to be an average of 1.58,[18] we regarded children with an increase in GMFM-66 of 2.0 points as good responders
to hippotherapy. Furthermore, we analyzed factors that affected good responders.
Participants
All participants voluntarily received hippotherapy, in addition to the conventional
physiotherapy that they had participated in. All participants were screened and evaluated
before and after the hippotherapy by one pediatric physiatrist who had worked at a
tertiary medical center located in the Seoul metropolitan. Furthermore, the medical
records of all patients were reviewed by the same physiatrist. The inclusion criteria
were: (1) diagnosis of CP, (2) body weight < 35 kg, and (3) age, 3 to 10 years. The
exclusion criteria were: (1) botulinum toxin injection within 6 months, (2) selective
dorsal rhizotomy or orthopedic surgery within 1 year, (3) moderate to severe intellectual
disability (when the participant could not follow the one-step obey command), (4)
uncontrolled seizure, and (5) poor visual or hearing acuity. These criteria have been
used for selecting participants in our center since 2008. Participants who were absent
for the lessons more than four times were excluded from the final analysis. All the
parents agreed on behalf of the participants to refrain from botulinum toxin injection
during the course of the therapy.
Hippotherapy
Samsung's Riding for the Disabled Program (RD-Samsung) in an 18 × 27 m indoor riding
arena located in Gunpo, Kyunggido, Republic of Korea provided hippotherapy classes.
Sessions were conducted by two physical therapists who held hippotherapy clinical
specialist certificates. For safety, two volunteer side walkers walked with the horse
and all participants wore helmets. A soft saddle made of wool was used to maximize
contact between the participants and the pony. The four ponies were trained by the
staff to participate in hippotherapy (average height, 135 cm; average weight, 294
kg). The ponies and participants were matched for size and function as much as possible.
Two participants were grouped together for each session, and each was assigned a separate
therapist for private lessons. Our hippotherapy treatment protocol was based on the
study by McGibbon et al[19] which included muscle relaxation; sustenance of optimal postural alignment of the
head, trunk, and lower extremities and independent sitting; and active exercises (stretching,
strengthening, dynamic balance, and postural control) directed by the therapist. The
participants were encouraged to maintain postural alignment during all the activities.
The intensity of the exercises and degree of assistance were individualized according
to the participants' ability to control their posture.
Ethics
The study protocol was reviewed and approved by the institutional review board at
the Samsung Medical Center (registry no. SMC 2017-12-081). Informed consent was waived
due to the retrospective nature of the study.
Outcomes Measured
Primary Outcome: GMFM-66
Assessment of GMFM-66 was performed by two pediatric physical therapists before and
after the 8-week hippotherapy program. The change in GMFM-66 is a reliable tool for
detecting changes in gross motor function in children with CP.[17] GMFM-66 consists of 66 items through Rasch's analysis, excluding 22 items with repeated
difficulty levels.
Secondary Outcome: GMFM-88
The GMFM-88 consists of 88 items in the following five dimensions: (A) lying and rolling;
(B) sitting; (C) crawling and kneeling; (D) standing; and (E) walking, running, and
jumping.[17]
Secondary Outcome: PBS
The PBS is a modified version of the Berg's balance scale and used in children with
spastic CP. The PBS consists of 14 items and evaluates functional balance in everyday
tasks. The items assess the functional activities within the home, school, or community,
including sitting balance, standing balance, sit to stand, stand to sit, transfers,
stepping, reaching forward, reaching the floor, turning, and stepping on and off through
an elevated surface. Each item is scored on a four-point scale. The validity and inter-/intrarater
reliability of PBS have been demonstrated in children with CP.[20]
Statistical Analysis
We regarded children with an increase in GMFM-66 of 2.0 points as good responders
to hippotherapy as mentioned above. For sex, age, and distribution (only spastic type)
subgroups, the Chi-square test was used to analyze the proportion of good responders.
For CP type, Fisher's exact test was used to analyze the proportion of good responders.
For GMFCS level, the linear-by-linear association method was used to analyze the proportion
of good responders. The t-tests were used to compare GMFM-66, GMFM-88, and PBS changes according to sex, age,
and distribution (only spastic type). For the CP type and GMFCS level, Kruskal–Wallis
test (nonparametric method) was used. For GMFCS level, Mann–Whitney test was used
as a posttest.
The variables used for univariate logistic regression were age, sex, distribution,
CP type, GMFCS level, baseline GMFM A/B/C/D/E, and baseline PBS. For analysis purposes,
GMFCS levels were stratified into three groups, namely, levels I and II, III, and
IV. The possibility of self-ambulation and the necessity of using walking aids are
the most important factors in classifying children with CP according to their motor
function. Baseline GMFM-66 and baseline GMFM-88 were strongly correlated with each
other. Therefore, baseline GMFM-88 was excluded from the selection of variables. Variables
with p-values less than 0.2 in the univariate logistic regression analysis were selected
for multivariate logistic regression analysis to identify the factors that affected
the efficacy of hippotherapy. The significance level was set at < 0.05. All analyses
were performed with SPSS version 24.0 software (IBM Corp., Armonk, New York, U.S.A.).
Results
We enrolled 155 participants who met the inclusion and exclusion criteria. Nine children
were excluded from the final analysis owing to frequent absences (> four times). Finally,
the data of 146 participants were analyzed retrospectively. The attendance rate of
the participants was 94.0 ± 8.0%.
The mean age of the participants was 5.78 ± 1.72 years. Of the 146 participants, 82
(56.2%) were boys; 128 (87.7%), 10 (6.8%), and eight (5.5%) participants were spastic,
dyskinetic, and ataxic CP, respectively. Of the 128 children with spastic CP, 12 had
a unilateral distribution; 24 (16.4%), 48 (32.9%), 44 (30.1%), and 30 (20.5%) participants
were GMFCS levels I and II, III and IV ([Table 1]), respectively.
Table 1
Analysis of the ratios of good responders to hippotherapy among subgroups
|
Subgroups
|
Total
n (%)
|
Good responders
n (%)
|
Poor responders
n (%)
|
p-Value
|
|
Sex
|
|
|
|
|
|
Boys
|
82 (56.2)
|
39 (47.6)
|
43 (52.4)
|
0.770
|
|
Girls
|
64 (43.8)
|
32 (50.0)
|
32 (50.0)
|
|
|
Age (y)
|
|
|
|
|
|
3 ≤ age < 6
|
77 (52.7)
|
37 (48.1)
|
40 (51.9)
|
0.883
|
|
6 ≤ age < 11
|
69 (47.3)
|
34 (47.9)
|
35 (50.7)
|
|
|
CP type, n (%)
|
|
|
|
|
|
Spastic
|
128 (87.7)
|
62 (48.4)
|
66 (51.6)
|
1.000
|
|
Dyskinetic
|
10 (6.8)
|
5 (50.0)
|
5 (50.0)
|
1.000
|
|
Ataxic
|
8 (5.5)
|
4 (50.0)
|
4 (50.0)
|
1.000
|
|
Distribution (spastic type)
|
|
|
|
|
|
Unilateral
|
12 (9.4)
|
7 (58.3)
|
5 (41.7)
|
0.471
|
|
Bilateral
|
116 (90.6)
|
55 (47.4)
|
61 (52.6)
|
|
|
GMFCS level
|
|
|
|
|
|
I
|
24 (16.4)
|
15 (62.5)
|
9 (37.5)
|
< 0.001
|
|
II
|
48 (32.9)
|
31 (64.6)
|
17 (35.4)
|
|
|
III
|
44 (30.1)
|
18 (40.9)
|
26 (59.1)
|
|
|
IV
|
30 (20.5)
|
7 (23.3)
|
23 (76.7)
|
|
Abbreviations: CP, cerebral palsy; GMFCS, gross motor function classification system.
Seventy-one participants (48.6%) were good responders to hippotherapy. There was no
difference in the proportion of good responder in sex, age, CP type and distribution
subgroups. However, as a result of linear-by-linear association analysis, we found
that as the GMFCS level increased, the proportion of good responder decreased ([Table 1]).
GMFM-88 changes in children with unilateral CP were significantly greater than in
those with bilateral CP (p = 0.001). There was no statistical difference of GMFM-66, GMFM-88 and PBS changes
according to CP types. There were significant differences in GMFM-66, GMFM-88, and
PBS changes according to the GMFCS level. Results are described in detail in [Table 2].
Table 2
Changes of GMFM-66, GMFM-88, and PBS after hippotherapy
|
Subgroups
|
GMFM-66 change
|
GMFM-88 change
|
PBS change
|
|
Mean (SD)
|
p-Value
|
Mean (SD)
|
p-Value
|
Mean (SD)
|
p-Value
|
|
Sex
|
|
|
|
|
|
|
|
Boys
|
2.27 (1.77)
|
0.561
|
2.81 (1.92)
|
0.415
|
3.40 (3.25)
|
0.843
|
|
Girls
|
2.45 (2.10)
|
|
3.16 (2.96)
|
|
3.30 (3.13)
|
|
|
Age (y)
|
|
|
|
|
|
|
|
3 ≤ age < 6
|
2.30 (1.79)
|
0.768
|
2.90 (2.32)
|
0.762
|
3.47 (3.01)
|
0.657
|
|
6 ≤ age < 11
|
2.40 (2.07)
|
|
3.02 (2.56)
|
|
3.23 (3.39)
|
|
|
CP type
|
|
0.818
|
|
0.208
|
|
0.523
|
|
Spastic
|
2.32 (1.94)
|
|
2.91 (2.54)
|
|
3.29 (3.24)
|
|
|
Dyskinetic
|
2.39 (1.50)
|
|
3.52 (1.64)
|
|
4.30 (3.10)
|
|
|
Ataxic
|
2.78 (2.23)
|
|
3.01 (1.19)
|
|
3.25 (2.98)
|
|
|
Distribution
|
|
|
|
|
|
|
|
Unilateral
|
2.46 (2.17)
|
0.787
|
1.56 (1.16)
|
0.001
|
3.83 (3.30)
|
0.544
|
|
Bilateral
|
2.30 (1.92)
|
|
3.05 (2.60)
|
|
3.23 (3.25)
|
|
|
GMFCS level
|
|
|
|
|
|
|
|
I
|
3.33 (2.26)
|
0.001[a]
|
1.73 (1.20)
|
0.001[c]
|
2.88 (2.13)
|
< 0.001[d]
|
|
II
|
2.82 (2.07)
|
< 0.001[b]
|
2.86 (2.16)
|
|
4.42 (3.61)
|
|
|
III
|
1.98 (1.57)
|
|
3.76 (2.95)
|
|
3.57 (3.41)
|
|
|
IV
|
1.36 (1.17)
|
|
2.94 (2.36)
|
|
1.73 (2.00)
|
|
Abbreviations: CP, cerebral palsy; GMFCS, gross motor function classification system;
GMFM, gross motor function measure; PBS, pediatric balance scale; SD, standard deviation.
a Significant difference of GMFM-66 between GMFCS I and IV.
b Significant difference of GMFM-66 between GMFCS II and IV.
c Significant difference of GMFM-88 between GMFCS I and III.
d Significant difference of PBS between GMFCS II and IV.
Variables with p-values less than 0.2 in the univariate logistic regression analysis were: the GMFCS
level; the baseline GMFM B, C, D, and E scores; and the baseline PBS score. In the
multivariate logistic regression analysis, GMFCS level of I and II compared with IV
(p = 0.046) and III compared with IV (p = 0.045), baseline GMFM E score (p = 0.031), were the factors that had a positive influence on good response to hippotherapy.
Baseline GMFM B score (p = 0.046) had a negative influence on good response to hippotherapy. Sex, age, CP
type and distribution were not factors influencing gross motor outcome of hippotherapy
according to our analysis ([Table 3]).
Table 3
Factors that influence the therapeutic effect of hippotherapy
|
Factors
|
Univariate logistic analysis
|
Multivariate logistic analysis
|
|
OR (95% CI)
|
p-Value
|
OR (95% CI)
|
p-Value
|
|
Sex
|
Boys
|
Reference
|
–
|
–
|
–
|
|
Girls
|
1.10 (0.57–2.12)
|
0.770
|
–
|
–
|
|
Age (y)
|
3 ≤age< 6
|
Reference
|
–
|
–
|
–
|
|
6 ≤age< 11
|
1.05 (0.55–2.01)
|
0.883
|
–
|
–
|
|
Distribution
|
Unilateral
|
Reference
|
–
|
–
|
–
|
|
Bilateral
|
0.65 (0.20–2.16)
|
0.485
|
–
|
–
|
|
GMFCS level
|
I or II
|
5.81 (2.20–5.38)
|
< 0.001
|
6.83 (1.03–45.09)
|
0.046
|
|
III
|
2.28 (0.81–6.42)
|
0.121
|
4.45 (1.03–19.16)
|
0.045
|
|
IV
|
Reference
|
–
|
Reference
|
–
|
|
CP type
|
Spastic
|
Reference
|
–
|
–
|
–
|
|
Dyskinetic
|
1.07 (0.29–3.86)
|
0.924
|
–
|
–
|
|
Ataxic
|
1.07 (0.26–3.86)
|
0.932
|
–
|
–
|
|
Baseline
gross motor
|
Baseline GMFM A
|
0.93 (0.69–1.25)
|
0.626
|
–
|
–
|
|
Baseline GMFM B
|
1.04 (0.99–1.09)
|
0.124
|
0.93 (0.87–0.999)
|
0.046
|
|
Baseline GMFM C
|
1.06 (1.02–1.11)
|
0.006
|
–
|
–
|
|
Baseline GMFM D
|
1.05 (1.02–1.08)
|
0.001
|
–
|
–
|
|
Baseline GMFM E
|
1.04 (1.02–1.05)
|
<0.001
|
1.05 (1.01–1.11)
|
0.031
|
|
Baseline PBS
|
1.04 (1.02–1.06)
|
<0.001
|
–
|
–
|
Abbreviations: CI, confidence interval; CP, cerebral palsy; GMFCS, gross motor function
classification system; GMFM, gross motor function measure; PBS, pediatric balance
scale; OR, odds ratio.
Discussion
Here, we found that the children with CP, GMFCS levels I and II or III, with relatively
poor postural control in sitting might have a greater chance to improve their GMFM-66
scores through hippotherapy. To our knowledge, this study is the first to analyze
the factors that affect the therapeutic effect of hippotherapy. Conflicting evidence
exists regarding the improvement of GMFM scores after hippotherapy. Despite this,
many studies reported significant improvements: the most recent meta-analysis revealed
an insignificant increase in GMFM score.[16] Contributing to these contradictory results may be the limitations of small sample
sizes, individual variability (gross motor functions, age, and CP types and distribution),
types of intervention (hippotherapy vs. therapeutic horseback riding and private lesson
vs. group lesson), and insufficient intervention duration. This study involves a relatively
large number of subjects (146 participants with CP). Furthermore, the intervention
is relatively uniform over a sufficient period (single-center study for 8 weeks).
As we also had observed that the degree of improvement in GMFM scores after hippotherapy
had individual differences, we would like to analyze factors (individual variables)
that affect the gross motor outcome measured using GMFM-66. As the minimal clinically
important difference in GMFM-66 was reported to be 1.58, we regarded the children
with GMFM-66 scores increased by 2.0 points as good responders to hippotherapy. Moreover,
we analyzed the factors that affected good responders.
Several studies have been conducted that investigate the factors are affecting the
response of children with CP to physical therapy in a hospital setting. Chen et al
reported that the PT efficacy of young (< 3 years old) children with CP was better
and that the GMFM-66 score improved at 8 years of age. Furthermore, the improvement
differed among the five GMFCS levels (p < 0.001), and GMFCS level II had a faster progression in Chen et al's study.[20] Similar results were confirmed in another study on factors that affected the response
to intensive therapy in children with CP or developmental delay.[21] However, we could not find any additional effects of age on GMFM improvement after
hippotherapy, which is not consistent with previous studies.[20]
[21] This result is thought to be due to the fact that approximately half of the participants
were aged 6 years or older who already reached the plateau of the GMFM-66 score. This
can be paradoxically interpreted that hippotherapy can be an effective therapy for
some school-aged children with little potential of further improvement in GMFM scores.
Like conventional PT, hippotherapy was more effective in the ambulatory group than
in the nonambulatory group. The participants with GMFCS levels I and II and III were
significantly more likely to be good responders than those with GMFCS level IV ([Table 3], odds ratio [OR] = 6.83, 4.45). The most marked change in GMFM-88 was noted in GMFCS
level III and the most marked change of PBS was noted in GMFCS level II. A recent
study that used ICF-CY checklist confirmed that the effect of hippotherapy was distinct
from GMFCS levels, and most improvements were present in children with GMFCS levels
I–III.[22]
The GMFM-66 curves appear to reach plateaus by about age 7 years. Children, on average,
reach approximately 90% of their motor function (as measured by the GMFM-66) by around
age ≤ 5 years, depending on their GMFCS level. Children with GMFCS levels IV and V
are expected to achieve 90% of their potential GMFM score at the age of 3.5 and 2.7
years, respectively.[23] Children with GMFCS levels IV and V exhibited significant limited motor function
and potential to improve.[24] However, this does not mean that hippotherapy should be performed only for children
with GMFCS levels I and II, or III. The previous randomized controlled trial (RCT)
study showed that the gross motor function of children with GMFCS level IV was also
significantly improved as compared with that of the control group after hippotherapy.[25] Hsieh et al reported that the limitations of activities and participation domain
of the ICF-CY checklist (d235–managing one's behavior, and d315–communicating with
or receiving nonverbal messages) were reduced after hippotherapy in children with
GMFCS levels IV and V. They thought that this improvement may have been due to an
increased awareness of the child's ability to communicate and perform general tasks
and respond to demands. The authors stated that hippotherapy may provide these children
with a modifying activity that ensures their participation in building a relationship
with a horse.[22]
The GMFM D and E areas have a decisive impact on the GMFM-66 score. All 13 GMFM-88
D items and all 24 GMFM-88 E items are included in GMFM-66. Therefore, high GMFM D
and E scores can affect further improvement in GMFM-66 score. We found that baseline
GMFM E score (walking, running, and jumping) was a positive predictive factor of hippotherapy.
However, unexpectedly, we found that a lower B score (a low sitting function) predicts
a better outcome of hippotherapy. During hippotherapy, children do various activities
to improve postural control in the sitting position. Previous meta-analysis studies
demonstrated that EAATs are indicated to improve postural control and balance in children
with CP.[10] The present results and previous meta-analysis results suggest that hippotherapy
is a context-focused therapy to improve postural control in sitting. Children with
CP and GMFCS levels I and II or III who have relatively poor postural control in sitting
might have a greater chance to improve their GMFM-66 scores through hippotherapy.
This result also supports the recent study which suggests that hippotherapy shows
distinct therapeutic strengths with regard to promoting upright stand and gait in
children with cerebral palsy.[26]
We did not find any influence of CP type and distribution on the increase in GMFM-66
score after hippotherapy. Relatively fewer ataxic (5.5%) and dyskinetic (6.8%) CP
cases than spastic cases were enrolled in this study. Spastic CP is the most common
type of CP, accounting for 72 to 91% of CP cases.[27] Furthermore, children with dyskinetic CP were reported to have more severe cognitive
and motor impairments than children with bilateral spastic CP.[28] Therefore, they were likely to have been excluded from the therapy according to
our inclusion/exclusion criteria. There is a big difference in gross motor function
between the children with unilateral CP and those with bilateral CP; 98% of hemiplegia,
66.7% of diplegia, and only 3.9% of quadriplegia were reported as GMFCS levels I–II.[29] The greater change of GMFM-88 in unilateral CP might be attributed to the high percentage
of GMFCS levels I and II in this group.
Children's participation in activities at home, school, and community is one of the
most important outcomes of rehabilitation interventions, because it is fundamental
to their health and development. Most traditional interventions for children with
CP focus primarily on factors in the domain of body function and structure. EAATs
seem to be attractive to all children with CP and their family members, as it is a
fun sport that improves their body functions, activities, and participation.[22] Emerging therapy interventions for children with physical disabilities suggest focusing
changes in the environment and/or the activity demands (i.e., context-focused therapy)
rather than on directing change to the child's abilities (i.e., child-focused therapy)
as a useful approach to improve function and participation.[30] Furthermore, the recently introduced treatment models influenced by the dynamic
systems theory consider task or activity completion as the goal, with less emphasis
on remediation or “normalization” of movement components.[31] From this point of view, EAATs can be a context-focused, goal-based therapy for
children with CP. During hippotherapy, the rider must perform various activities and
tasks on the horse in real-life environments outside the therapy gym (barn, arena,
nature, etc.). Moreover, human-horse interaction is a powerful motivation for engaging
children's participation in the therapy.[32] Novak stressed the importance of “child-active approaches” where the child is actively
practicing real-life tasks during intervention (usually in real-life environments)
to gain or consolidate real-life skills that they want to learn.[33]
Limitations
This study has several limitations. First, this is a retrospective study without a
control group. A prospective RCT should be conducted to confirm our results. Our results
indicate that walking ability is a determinant factor influencing motor outcome of
hippotherapy in children with CP. Therefore, in designing future studies, the GMFCS
level of children should be considered as an inclusion/exclusion criterion. Hippotherapy
is currently classified as a yellow-light intervention which had either lower-level
evidence supporting their effectiveness or inconclusive evidence. Novak stated that
“when yellow-light interventions are used, it is imperative that clinicians utilize
a sufficiently sensitive outcome measure to confirm whether or not the intervention
is working and if it is helping the child to achieve their family's goals.”[33] From this point of view, the Canadian occupational performance measure and goal
attainment scaling (GAS) are recommended as outcome measures in EAATs or hippotherapy
research. Both measures work well within a family-centered approach because they encourage
family-led goal setting and facilitate individualization that is critical to heterogeneous
conditions, such as CP. The therapeutic riding assessment of impact network proposed
using GAS as a common outcome measure at four therapeutic riding centers since 2014.
Further several previous RCTs on EAATs or hippotherapy examined mostly body structure
and function, such as muscle asymmetry, spasticity, gait, gross motor function, and
balance, using various laboratory and clinical measures, such as surface electromyography,[34] modified Ashworth scale,[35] GMFM,[25]
[36] and PBS.[25] Only two RCTs[26]
[36] used the CP quality of life questionnaire for children, child health questionnaire,
and the KIDSCREEN-27 that represent quality of life. When planning EAAT or hippotherapy
clinical trials, we recommend standardized outcome measures that meet corresponding
individual's goals based on the ICF model (body structure/function, activity, and
participation). The complexity of EAATs or hippotherapy has an important influence
on RCT design. The key issues to consider include the definition of each intervention
and their components, the person providing the intervention (therapists, instructors,
and side-walkers), and where and how the interventions will be standardized and monitored
during the trial. Second, the proportion of children significantly differed with CP
type. Supplementing the proportion of children with ataxic/dyskinetic CP and hemiplegic
CP may enable further analysis. Third, factors related to the long-term effects of
hippotherapy remain unknown. Furthermore, this retrospective study did not clarify
the effects of the individual language ability, cognitive level, and attention on
GMFM improvement after hippotherapy. Fourth, the examiners were not blind to the intervention
and therefore, the postintervention ratings could have a positive bias.
