The study design was a retrospective radiographic review. The study was
approved by our hospital's institutional review committee. The cases to be
used for the study were identified by searching a computerized database of
children who had been evaluated in our institution's Motion Analysis
Laboratory between March 2004 and August 2005. As part of the comprehensive
motion-analysis evaluation, all patients who utilize an orthotic device are
studied while walking barefoot and while wearing the orthosis in order to
determine the effects and benefits of the brace during gait. In this setting,
standing anteroposterior and lateral radiographs of the foot as well as a
standing anteroposterior radiograph of the ankle are typically made with the
child barefoot and with the child wearing the orthosis. Only patients with
complete sets of radiographs in the barefoot and orthotic conditions were
included in the review.
The study cohort consisted of 102 patients (160 feet) with a mean age of
10.0 years (range, 5.1 to 18.7 years) who met the inclusion criteria. Of the
102 patients, fifty-eight were male and forty-four were female. The diagnoses
included spastic monoplegia (one patient), spastic diplegia (forty-two),
spastic hemiplegia (forty-one), spastic triplegia (seven), spastic
quadriplegia (seven), and a mixed variety (four). With use of the Gross Motor
Function Classification
system21, twelve
patients were classified as grade 1, sixty-seven were classified as grade 2,
twenty were classified as grade 3, and three were classified as grade 4. Of
the 160 feet, eighty-three were on the right side and seventy-seven were on
the left. Of the 160 braces, 109 were solid ankle orthoses, fourteen were
posterior leaf spring orthoses, thirty-one were articulated orthoses, and six
were ground-reaction orthoses. All were provided by our institution's in-house
Prosthetics and Orthotics Laboratory, with the primary goal of optimizing
overall foot shape and alignment in the fabrication process. Surgical
correction to improve foot alignment and to facilitate bracing was performed
prior to the fabrication of the orthoses for rigid malaligned feet.
Radiographic Technique
All of the foot and ankle radiographs were made with a standardized
technique. The standing anteroposterior radiograph of the foot was made with
the film cassette placed beneath the foot and the machine tube angled 30°
from the vertical axis and aimed from distal to proximal (from the toes to the
heel) at a distance of 40 in (102 cm) from the foot. The standing lateral
radiograph of the foot was made with the film cassette placed parallel to the
hindfoot and the machine tube parallel to the horizontal axis at a distance of
40 in (102 cm) from the foot. The standardized standing anteroposterior
radiograph of the ankle was made with the appropriate-size film cassette
placed behind the hindfoot, the ankle aligned to neutral by palpation of the
malleoli, and the machine tube parallel to the floor at a distance of 40 in
(102 cm) from the ankle. Standard radiographs with the patient barefoot were
made initially, followed immediately by radiographs with the patient wearing
the orthosis (Figs. 1-A,
1-B,
1-C and 1-D).
Radiograph Measurement Technique
Radiographs were measured with use of the technique as described in the
literature from which the normative data were
generated20,22-24.
In order to minimize bias, the radiographs representing the barefoot condition
were measured before the radiographs representing the orthotic condition were
viewed. For the purposes of the present study, measurements from the
anteroposterior and lateral radiographs of the foot were included in the
analysis. Measurements from the lateral radiograph of the foot were obtained
initially, followed by measurements from the anteroposterior radiograph of the
foot. Standardized measurement techniques included the use of a single
goniometer and a sharpened x-ray marking pencil. All measurements were made by
a pediatric orthopaedic surgeon (D.E.W.) and an orthopaedic resident (J.C.S.).
For the present study, two measurements of hindfoot alignment (the calcaneal
pitch angle and the talocalcaneal angle), two measurements of midfoot
alignment (the naviculocuboid overlap and the talonavicular coverage angle),
and two measurements of forefoot alignment (the lateral talus-first metatarsal
angle and anteroposterior talus-first metatarsal angle) were obtained for each
set of radiographs.
The values of each measurement corresponding with the barefoot and orthotic
conditions were compared with normal age-appropriate values from the
literature20. The
normal range for each measurement was defined as the mean normal value plus or
minus one standard deviation. These normal values were previously obtained in
a similar study involving the use of standard radiographs of sixty normal feet
in patients ranging from five to seventeen years of age. An abnormally high
value was defined as a value falling above the normal range for each
measurement. An abnormally low value was defined as a value falling below the
normal range for each measurement.
Categorical descriptors were also established for each segment of the foot
(Table I). In the hindfoot, an
abnormally high value of the calcaneal pitch measure was defined as a
calcaneus deformity, an abnormally low value of the calcaneal pitch measure
was defined as an equinus deformity, an abnormally high value of the
talocalcaneal angle was defined as a valgus deformity, and an abnormally low
value of the talocalcaneal angle was defined as a varus deformity. In the
midfoot, an abnormally high value of the naviculocuboid overlap was defined as
a pronation deformity, an abnormally low value of the naviculocuboid overlap
was defined as a supination deformity, an abnormally high value of the
talonavicular coverage angle was defined as an abduction deformity, and an
abnormally low value of the talonavicular coverage angle was defined as an
adduction deformity. In the forefoot, an abnormal elevation in the value of
the lateral talus-first metatarsal angle was defined as a planus deformity, an
abnormally low value of the lateral talus-first metatarsal angle was defined
as a cavus deformity, an abnormal elevation in the value of the
anteroposterior talus-first metatarsal angle was defined as an abduction
deformity, and an abnormally low value of the anteroposterior talus-first
metatarsal angle was defined as an adduction deformity.
Mean values and standard deviations for the radiographic measurements
corresponding with the barefoot and orthotic conditions were compared with
normal values from the literature. To prevent potential washout of data from
offsetting deformities, comparisons between the barefoot and orthotic
conditions were performed with use of the absolute value of change for each
measurement. For each measurement, the difference between the barefoot value
and orthotic value was calculated. This difference was then converted to its
absolute value. The absolute values were then averaged to determine the mean
absolute value change.
Statistical Analysis
Measurements corresponding with the barefoot and orthotic conditions were
compared with use of the paired t test. All tests were two-sided, with the
level of significance set at p = 0.05. The analyses were performed with use of
SAS software (SAS Institute, Cary, North Carolina). Power analysis was
performed on the basis of a difference of 5° (a clinically meaningful
difference) between the measurements corresponding with the barefoot and
orthotic conditions with use of the paired t test, an alpha level of 0.05, and
two-sided tests (nQuery Advisor version 4.0; Statistical Solutions, Saugus,
Massachusetts). With 160 feet, the current study had 99% power to detect an
effect size difference of =0.3°.
Barefoot Condition Compared with Normal Values
The mean value of calcaneal pitch for the barefoot condition was 10.1°,
compared with a normal value of 17°; this difference was significant (p
< 0.05). The mean value of the talocalcaneal angle for the barefoot
condition was 43.7°, compared with a normal value of 49°; this
difference was also significant (p < 0.05). In the forefoot, the mean value
of the lateral talus-first metatarsal angle for the barefoot condition was
17.6°, compared with a normal value of 13°; this difference was
significant (p < 0.05). The remaining measure of forefoot alignment and the
two measures of midfoot alignment showed no significant differences when the
values for the barefoot condition were compared with normal values. Although
significant differences were demonstrated in three of the six measures,
clinically important differences were not appreciated. The greatest magnitude
of the differences between all measures when the barefoot condition values
were compared with the normal values was seen in the measure of calcaneal
pitch, with a difference of <7°.
Orthotic Condition Compared with Normal Values
The mean value of calcaneal pitch for the orthotic condition was 10.2°,
compared with a normal value of 17°; this difference was significant (p
< 0.05). The mean value of the talocalcaneal angle for the orthotic
condition was 42.3°, compared with a normal value of 49°; this
difference was also significant (p < 0.05). The only other measure to show
a significant difference was the anteroposterior talus-first metatarsal angle,
with a mean value of 5.4° for orthotic condition as compared with a normal
value of 10° (p < 0.05). The remaining measures of midfoot and forefoot
alignment for the orthotic condition demonstrated no significant differences
from normal values. Although significant differences were demonstrated in the
measures of hindfoot alignment, clinically important differences were not
appreciated. The greatest magnitude of the differences between all measures
when the orthotic condition was compared with the normal values was seen in
the measure of calcaneal pitch, with a difference of <7°.
Barefoot Condition Compared with Orthotic Condition
(Table II)
In the hindfoot, the absolute value change was 2.3° for calcaneal pitch
and 3.6° for the talocalcaneal angle; both differences were significant
but not clinically important (p < 0.05).
In the midfoot, the absolute value change was 9.4% for the naviculocuboid
overlap and 5.5° for the talonavicular coverage angle; again, both of
these differences were significant but not clinically important (p <
0.05).
In the forefoot, the absolute value change was 4.9° for the lateral
talus-first metatarsal angle and 5.3° for the anteroposterior talus-first
metatarsal angle; once again, both of these differences were significant but
not clinically important (p < 0.05).
Categorical Assessment
(Table
III)
Hindfoot
Five (3.1%) of the 160 feet demonstrated calcaneus deformity in the
barefoot condition, and one (20%) of these five feet was corrected to the
normal range in the orthosis. Conversely, eighty-five (53.1%) of the 160 feet
demonstrated an equinus deformity in the barefoot condition and twelve (14.1%)
of these eighty-five feet were corrected to the normal range in the orthosis.
Sixteen (10.0%) of the 160 feet demonstrated a hind-foot valgus deformity in
the barefoot condition, and seven (43.8%) of these sixteen feet were corrected
to the normal range in the orthosis. Seventy (43.8%) of the 160 feet
demonstrated a hindfoot varus deformity in the barefoot condition, and four
(5.7%) of these seventy feet were corrected to the normal range in the
orthosis.
Midfoot
Fifty-nine (36.9%) of the 160 feet demonstrated a pronation deformity in
the barefoot condition, and fourteen (23.7%) of these fifty-nine feet were
corrected to the normal range in the orthosis. Fifty (31.3%) of the 160 feet
demonstrated midfoot supination in the barefoot condition, and nine (18%) of
these fifty feet were corrected to the normal range in the orthosis.
Fifty-eight (36.3%) of the 160 feet demonstrated midfoot abduction in the
barefoot condition, and fifteen (25.9%) of these fifty-eight feet were
corrected to the normal range in the orthosis. Thirty-five (21.9%) of the 160
feet demonstrated midfoot adduction in the barefoot condition, and five
(14.3%) of these thirty-five feet were corrected to the normal range in the
orthosis.
Forefoot
Seventy (43.8%) of the 160 feet demonstrated planus deformity (an elevated
lateral talus-first metatarsal angle) in the barefoot condition, and
twenty-five (35.7%) of these seventy feet were corrected to the normal range
in the orthosis. Twenty-seven (16.9%) of the 160 feet demonstrated cavus
deformity in the barefoot condition, and five (18.5%) of these twenty-seven
feet were corrected to the normal range in the orthosis. Forty (25.0%) of the
160 feet demonstrated forefoot abduction (an elevated anteroposterior
talus-first metatarsal angle) in the barefoot condition, and ten (25.0%) of
these forty feet were corrected to the normal range in the orthosis.
Forty-eight (30.0%) of the 160 feet demonstrated forefoot adduction in the
barefoot condition, and only four (8.3%) of these forty-eight feet were
corrected to the normal range in the orthosis.
The goal of the current study was to assess the impact of ankle-foot
orthoses on the static, standing segmental alignment of the foot and ankle in
children with cerebral palsy. The only previous study to consider this issue
compared standing weight-bearing radiographs of the foot and ankle in a small
series of children wearing inhibitory casts, articulated orthoses, or solid
ankle orthoses19.
In that study, lateral radiographs of the foot were made with and without the
cast or ankle-foot orthosis after one to three weeks of wear time. Five
different measures on the lateral radiograph were obtained. The only
significant difference was found in the measure of the calcaneal inclination
("calcaneal pitch") while an articulated ankle-foot orthosis was
worn. The authors concluded that there was no significant change in the
osseous alignment of the foot and ankle during weight-bearing in children with
cerebral palsy when ankle-foot orthoses or inhibitory casts were worn.
Other studies have reviewed the radiographic changes associated with shoe
inserts in neurologically normal individuals. Kuhn et al., in a study of
patients ranging in age from six to fifty-seven years, reviewed the effects of
custom-made flexible shoe orthoses on flexible pes planus on the basis of
standing anteroposterior and lateral radiographs of the foot that were made
with and without the orthosis in
place25. Three
radiographic measurements (anteroposterior talocalcaneal angle, lateral
talocalcaneal angle, and lateral talar pitch) were obtained for each foot.
Although significant differences were found for each measure, the changes were
small, and clinically important differences in foot alignment were not
noted.
In contrast to Kuhn et al., Penneau et al. reviewed the effects of four
different shoe modifications on flexible pes planus in
children26. In that
study, ten children with bilateral flexible pes planus were evaluated
radiographically while bare-foot and while wearing a shoe modification or an
orthotic device (an orthopaedic shoe with a medial heel wedge, a longitudinal
arch support, a University of California Biomechanics Laboratory shoe insert,
or a Gillette foot orthosis). Radiographic measures included the
anteroposterior talocalcaneal angle, the lateral arch ratio, and the posterior
tibialcalcaneal alignment. No significant improvement in foot alignment was
found with the use of any of these devices. Wenger et al. prospectively
evaluated the treatment of flexible flatfoot in young children with corrective
shoes or inserts6.
Treatment groups were clinically and radiographically compared with a
nontreatment control group after three years. Both groups demonstrated
comparable clinical and radiographic improvement irrespective of treatment
with orthotics. Although orthoses are often prescribed to improve foot
alignment or to prevent deformity with growth, minimal support for this
treatment method is found in the literature.
The current study evaluated a much larger cohort of children with cerebral
palsy who were able to walk. All had ankle-foot orthoses that were fabricated
in a single laboratory where all efforts were made to improve foot and ankle
alignment through the orthotic fabrication process. Compared with other
reported studies, a more comprehensive radiographic analysis was performed,
with evaluation of the alignment of three segments of the foot in two planes
and comparison of these measurements with age-appropriate normal
values20. When the
barefoot foot alignment in the group of children with cerebral palsy was
compared with normal values, only three of the six radiographic measurements
were found to be significantly different. This does not mean that the
alignment of the foot and ankle in the children with cerebral palsy was near
normal in most cases, but rather reflects the problem of analyzing a mean
value when the possibility of quantitatively opposite malalignment values
(e.g., varus and valgus, adduction and abduction, supination and pronation,
and cavus and planus) may exist within the cohort. In such instances, a
categorical classification paradigm may provide more clinically relevant
information. When the barefoot condition was compared with the orthotic
condition for the children with cerebral palsy, significant differences were
identified in all six radiographic measures of foot alignment, although these
changes were small (<6° or <10°) and we do not believe that they
were clinically important.
When a categorical classification system was utilized, the most common
hindfoot malalignments in the barefoot condition were equinus and varus.
Correction of the hindfoot to within the normal range in the orthosis was most
likely for valgus malalignment; this occurred in seven (44%) of sixteen feet.
Midfoot malalignment was present in both directions in both the lateral plane
(i.e., pronation and supination) and the anteroposterior plane (i.e.,
abduction and adduction). Midfoot pronation was typically coupled with midfoot
abduction, whereas midfoot supination was typically coupled with midfoot
adduction. Correction of the midfoot to within the normal range in the
orthosis was most likely in cases of pronation (fourteen of fifty-nine feet;
24%) and abduction (fifteen of fifty-eight feet; 26%). The most common
forefoot deformity in the lateral plane was planus. In the anteroposterior
plane, malalignment was present in both directions (i.e., adduction and
abduction). Correction of the forefoot to within the normal range in the
orthosis was most likely in cases of planus (twenty-five of seventy feet; 36%)
and abduction (ten of forty feet; 25%). When the three segments are considered
together, the coupled malalignment of equinoplanovalgus (clinical flat-foot)
showed radiographic correction of at least one segment to within the normal
range in 24% to 44% of feet when an orthosis was used. The coupled
malalignment of equinocavovarus (clinical high arched foot) showed
radiographic correction of at least one segment to within the normal range in
5% to 20% of feet when an orthosis was used. These findings are consistent
with the clinical observation that equinoplanovalgus, because of the
biomechanical coupling between segments that facilitates both shock absorption
and lever function in normal feet, is typically a more flexible malalignment
pattern than equinocavovarus, which is typically a more rigid malalignment
pattern4,27,28.
Improvement in static foot alignment (clinically and radiographically)
theoretically leads to improved foot function and improved gait. Although all
patients underwent a complete instrumented motion-analysis study, the gait
data were not included in the current investigation. The goals of the present
study were to determine if any improvement in static alignment of the foot and
ankle occurs with the use of a brace. We did not seek to determine if
improvements during gait (specifically during stance phase) correlated with
radiographic evidence of improvement in foot alignment secondary to use of the
orthosis.
The present study shows that ankle-foot orthoses failed to improve the
static foot alignment in the majority of feet of children with cerebral palsy
who were able to walk. The functional benefits of an ankle-foot orthosis are
likely due to its ability to provide external support and stability and to
limit range of motion, effectively restoring the normal lever arm function of
the foot in stance phase and preventing dynamic deformities and deviations
related to muscle weakness or imbalance, while improving foot clearance and
prepositioning of the foot during the swing phase of gait. ?