A comparison of gait with solid, dynamic, and no ankle-foot orthoses in children with spastic cerebral palsy.

Children with spastic cerebral palsy (CP) frequently demonstrate pathological gait patterns involving abnormal joint motions,[1-5] muscle timing,[6-12] and temporal-distance characteristics such as decreased walking speed and stride length.[13,14] Abnormal joint motions include dynamic equinus or excessive ankle plantar flexion during the stance phase of gait, although no fixed contracture of the triceps surae muscle group is evident.[1-5] This gait deviation is often accompanied by excessive knee flexion or hyperextension; hip flexion, adduction, and medial (internal) rotation; and anterior pelvic tilt during the stance phase.[1,2,5]

The triceps surae muscle group have abnormal timing in the equinus gait pattern, with premature onset in late swing and continuous activity throughout stance.[6,7,9-11] The tibialis anterior muscle demonstrates curtailed activity during swing and prolonged activity into midstance.[6,7] The hamstring and quadriceps femoris muscle groups also exhibit prolonged activity during stance.[7,8,12]

Various types of lower-extremity orthoses have been used to correct an equinus gait pattern in children with spastic CP.[15] A polypropylene fixed or solid ankle-foot orthosis (AFO) has been the most commonly prescribed brace for reducing excessive ankle plantar flexion during stance.[16] The solid AFO is made of polypropylene that covers the entire posterior calf and the mediolateral borders and sole of the foot, with straps across the anterior upper tibia and front of the ankle.[17] It biomechanically controls the ankle by using a three-force system to prevent excessive ankle plantar flexion during stance.[17,18]

During the past 10 years, inhibitive or tone-reducing AFOs have been increasingly used as an alternative to the conventional solid AFO.[15,19,20] The rationale for the design, purpose, and use of the inhibitive AFO is proposed to be based on the inhibitive or tone-reducing, cast.[21-23] These casts are purported to decrease spasticity by prolonged stretch and pressure on the tendons of the triceps surae muscle and toe flexors and to inhibit or decrease abnormal reflexes in the lower extremity by protecting the foot from tactile-induced reflexes.[21,22] In addition, the casts are reported to prevent excessive ankle plantar flexion, improve lower-extremity muscle timing,[21] and normalize movements of the trunk, pelvis, and lower extremity in standing and during gait.[22-25] Studies have shown changes in the stretch sensitivities of the ankle plantar flexors,[26], increased ambulation ability,[22] improved passive ankle dorsiflexion and foot-floor contact during gait,[27] and improved stride length,[28,29] but no changes in bony alignment of the foot and ankle, in children with spastic CP who wore inhibitive casts.[30] No published studies, however, have examined the effects of inhibitive casts on muscle timing during gait in these children.

Several authors[15,19,22,31] have recommended that inhibitive casts should be replaced by polypropylene inhibitive AFOs when the child with CP begins ambulating. The inhibitive AFO is more flexible, lightweight, and easily worn with regular shoes than the cast[19] and is purported to have die same beneficial effects,[23] although no research has compared the effects of the inhibitive cast and AFO on the gait of children with spastic CP. Few published studies have examined the effectiveness of inhibitive AFOs among children with spastic CP. Studies using a single-subject design demonstrated increased standing duration[19] and improved knee motion during gait with inhibitive AFOs when compared with no AFOs.[32]

One type of inhibitive AFO is the dynamic ankle-foot orthosis (DAFO) with a plantar-flexion stop (Fig. 1).[20,23] The custom-contoured footplate of the DAFO is similar to the footplate of the inhibitive cast, with built-up areas under the toes, lateral and medial longitudinal arches, and transverse metatarsal arch and recessed areas under the metatarsal and calcaneal pad areas to provide support and stabilization to the arches of the foot and position the midtarsal and subtalar joints in a neutral position (Fig. 2).[20,23] The footplate is designed to reduce abnormal muscle activity and to effect biomechanical changes, including decreased excessive ankle plantar flexion and improved motions of the lower extremity, pelvis, and trunk during standing and gait.[19,20,23] The DAFO with a plantar-flexion stop is made of 2.4-mm-thick (3/32-in-thick) polypropylene that provides total contact by enclosing the forefoot and ankle with anterior trim lines at the center of the dorsum of the foot and covers about one third of the posterior calf (Fig. 1).[20,23] A toe loop stabilizes the first digit, an anterior forefoot strap increases total contact at the forefoot, and an ankle strap holds the heel in the orthosis.[23] The DAFO with a plantar-flexion stop is thinner, more flexible, and shorter than the conventional solid AFO.

The purpose of our study was to compare the effects of DAFOs with a plantar-flexion stop, polypropylene solid AFOs, and no AFOs on the gait of children with spastic CP who demonstrate a dynamic equinus gait pattern with excessive ankle plantar flexion during the stance phase. We hypothesized that there would be differences in the timing of lower-extremity muscle activity; joint motions in the lower extremity, pelvis, and trunk; and temporal-distance characteristics during ambulation with DAFOs, solid AFOs, and no AFOs. We also examined the DAFO's purported effects on improving joint motion and on producing more normal muscle timing during gait in children with spastic CP.

Method

Subjects

A sample of convenience consisting of 10 subjects with spastic CP was obtained from the outpatient clinic at Shriners Hospital for Children in San Francisco, Calif. The sample was limited to the first 10 subjects meeting the selection criteria who were recruited during the 2-year funded period for the study. Written informed consent was obtained from the parents of all subjects prior to participation in the study.

Four girls and six boys with spastic CP, with an average age of 6.5 years (SD=1.86, range=3.5-8.5), were recruited. Four subjects had spastic hemiplegia, and six subjects had spastic diplegia. All subjects were community ambulators who demonstrated (1) a plantigrade foot in weight bearing during standing, (2) excessive ankle plantar flexion during the stance phase of gait, (3) passive ankle dorsiflexion to 5 degrees or more with knees extended, (4) passive hip extension of -10 degrees or less as measured by the Thomas test, (5) passive hamstring muscle length of 60 degrees or more as measured by a straight leg raise, (6) mild to moderate spasticity of the lower extremities (a score of 1 or 2 on the Ashworth scale), (7) no use of assistive devices during ambulation, and (8) no orthopedic surgery during the past year and for the duration of the study. Five subjects wore solid AFOs, whereas the other five children used hinged AFOs for at least 1 year prior to participation in this study. Eight subjects received physical therapy for gait training ranging from once a month to twice a week. The type of physical therapy was not controlled in this study. Two subjects did not receive ongoing physical therapy.

Procedure

A repeated-measures design,[33] with subjects serving as their own controls, was used for the study. Each subject wore no orthoses for an initial 2-week period, solid AFOs for 1 month, no orthoses for an additional 2 weeks, and DAFOs with a plantar-flexion stop for 1 month. The six subjects with spastic diplegia wore orthoses on both lower extremities, and the four subjects with hemiplegia wore an orthosis on the involved lower extremity. The order of wearing solid AFOs prior to DAFOs was based on (1) scheduling constraints because subjects traveled 80 to 240 km (50-150 miles) to the clinic to be casted for DAFOs and (2) on coordinating the fabrication of solid AFOs and DAFOs by two different orthotists. The orthoses were worn for 1 month so that the subjects could gradually adapt to wearing them for the entire day. The shorter 2-week periods with no orthoses allowed the subjects to adapt to ambulating without orthoses, but prevented fixed contractures in the triceps surae muscles from developing. The solid AFOs and DAFOs were provided without cost to the subjects. The actual costs of both orthoses were similar, except that the DAFOs took about 1 additional hour to cast and fit.

The solid AFO was custom-made for each subject from a positive mold after casting by an orthotist. It was fabricated from 4.8-mm-thick polypropylene extending distally under the toes and on the mediolateral border of the foot and proximally on the posterior part of the leg to about 2.5 to 5 cm below the knee, with trim line, anterior to both malleoli and straps across the front of the ankle and anterior upper tibia.[17] The DAFO with a contoured footplate (Fig. 1) was made from 2.4-mm-thick polypropylene enclosing the dorsum of the fore-foot and ankle and covering the posterior part of the leg to about 5 to 7.5 cm above the malleoli, with straps across the ankle, forefoot, and first digit.[20,22,34] All subjects were casted for the DAFO using a prefabricated contoured footplatern[34] by a physical therapist consultant experienced in casting for this type of orthosis. The DAFO(*) was fabricated from a positive mold by an orthotist from Cascade Prosthetics and Orthotics (Bellingham, Wash).[34]

All subjects were tested in the Orthopaedic Biomechanics Laboratory at Shriners Hospital for Children at the end of each of the four interventions: barefoot with no orthoses for initial 2-week period, wearing solid AFOs for 1 month, barefoot with no orthoses for second 2-week period, and wearing DAFOs for 1 month. The subjects were tested barefoot for the two interventions with no orthoses because baseline gait studies conducted in this laboratory and in similar gait laboratories typically examine children ambulating without shoes.[35] The following gait measures were used: (1) surface electromyography (EMG) of the gluteus maximus, hamstring, quadriceps femoris, triceps surae, and tibialis anterior muscle groups to determine timing of these lower-extremity muscle groups during the stance phase, (2) use of contact-closing footswitches to obtain measurements of temporal-distance gait characteristics, including walking speed, stride length, and cadence, and (3) three-dimensional motion analysis to determine joint motions of the trunk, pelvis, hip, knee, and ankle at initial contact and mid-stance. Test-retest reliability of the gait measurements was determined by comparing the data obtained from the two interventions with no orthoses. Intraclass correlation coefficients (ICC[3,1])[33] for these two interventions were calculated for temporal-distance gait characteristics (ICCs=.96-.99), muscle timing (ICCs=.79-.98), and joint motions (ICCs=.29-.94) (Tab. 1).

Silver-silver chloride surface electrode pairs(dagger) measuring 5 mm in diameter and spaced approximately 2.5 cm apart were placed on the skin over the muscle belly after conventional skin preparation to gently abrade the surface area. The electrodes were applied longitudinal to the direction of the fibers of five muscle groups of the lower extremity with the greatest degree of excessive ankle plantar flexion during stance without orthoses, as determined by visual observation. The electrode sites were based on procedures described by Delagi et al[36] for the gluteus maximus, hamstring (long head of biceps femoris), quadriceps femoris (rectus femoris), triceps surae (lateral head of gastrocnemius), and tibialis anterior muscle groups. Each surface electrode pair was connected to an individual transmitter.(double dagger) Telemetered EMG signals were amplified with a gain factor of 3,000, sampled at a rate of 2,500 Hz, and recorded directly into the computer. The analog signal was collected with a bandwidth of 30 to 500 Hz and then digitally filtered with a bandwidth of 94 to 278 Hz to reduce low-frequency noise and motion artifact, typically found when recording during walking of young children. Although some signal power is contained below the lower cutoff, we believe that filtering at 94 Hz did not change the muscle timing information used for this study.

Electromyographic and footswitch data were recorded simultaneously with CODAS data-collection softwares(sections) as each subject ambulated on a 10-m walkway at a self-selected speed for at least two trials, with a rest period allowed between trials to prevent fatigue. Footswitch on-off signals for initial contact and toe-off were used as event markers to indicate stance and swing phases of the gait cycle. Data were also collected for one trial with the subject sitting or lying without moving and with all muscles at rest. The raw and digitally filtered EMG recordings for resting and gait for each muscle group were first visually inspected for noise and motion artifact.

The onset and cessation of muscle activity was determined by visual inspection by one of the authors (SAR) for all subjects and tests. The EMG recording at rest was used as the baseline measurement so that any recording for the muscle during gait that was larger than baseline measurement was identified as muscle activity. Although visual inspection of muscle activity is not as reliable as computerized analysis,[37] the reliability for the two periods with no orthoses in this study showed high ICCs of .79 to .98 (Tab. 1). The active periods for each muscle group for two trials (total of 10 gait cycles) were measured and normalized to the gait cycle as an average percentage of the stance phase. Electromyographic muscle timing, defined as the duration of muscle firing, of the five muscle groups was determined for each testing session.

Contact-closing footswitches were placed along the entire plantar surfaces of both feet and taped to the feet for tests without AFOs and to the shoes for tests with the orthoses. Footswitches were connected to a transmitter attached to the ankle. The on-off footswitch signals for initial contact and toe-off were transmitted by telemetry to a computer with CODAS data-collection software as the subject ambulated and interrupted the light-beam switches that were spaced 500 cm apart. Temporal-distance gait characteristics, including walking speed (distance over time [in meters per minute]), cadence (steps per minute), and stride length (distance [in centimeters] between two consecutive initial contacts on one foot), were determined from footswitch signals by a computer software program developed at the laboratory. Reliability for the footswitches for the two interventions with no orthoses showed high ICCs of .96 to .99 (Tab. 1) Walking speed, cadence, and stride length for two trials were averaged for each testing session.

A computerized, three-dimensional motion analysis system (Motion Analysis[TM][||]) was used to collect joint angle displacement data as each subject ambulated with a self-selected speed for at least two trials on a 10-m walkway. Gait data were collected for 4 to 6 seconds over a 1.5-m length of the walkway, which was calibrated prior to each test session. The motion analysis system consisted of six video cameras[#] with attached spotlights mounted on the wall in the front, back, and four comers of the laboratory, a video monitor,(**) a video processor (VP-320),[||] and a computer. Reliability for the two interventions with no orthoses showed ICCs greater than .50 for all joint motions, except hip rotation at initial contact and mid-stance and hip adduction and abduction and trunk rotation at mid-stance (Tab. 1).

Twenty-one retroreflective markers,[||] 2.54 cm (1 in) and 3.8 mm (1 1/2 in) in diameter using a modified Helen Hayes Hospital marker set,[38] were placed on each subject by one tester. Markers were attached directly to the skin with tape or straps over selected anatomical landmarks on the upper extremities, lower extremities, and pelvis, bilaterally, for the tests without AFOs. The ankle, heel, and toe markers were taped on the orthosis and shoe over anatomical landmarks for tests with orthoses. The larger markers were placed bilaterally on the lateral shoulder (aligned with the scapular acromial process), the lateral elbow (aligned with the humeral lateral epicondyle), the pelvic anterior superior iliac spines, and the superior aspect of the sacrum halfway between the pelvic posterior superior iliac spines. The smaller markers were positioned bilaterally on the dorsum of the wrist halfway between the ulnar and radial styloid processes, on the lateral femoral condyle of the knee, on the ankle above the tibial lateral malleolus, on the space between the second and third metatarsal heads of the foot, and on the middle of the posterior heel. Five-centimeter-long wands with the larger markers on the tips were secured with straps on the lateral midthigh and midshank.

The six video cameras recorded the images from the markers in the sagittal, coronal, and transverse planes at a sampling rate of 60 frames per second. The video processor produced digital outlines of the markers, which were transferred to three-dimensional paths by ExpertVision computer software.[||] The Orthotrak software package[||] used the three-dimensional path data to determine coordinate systems for each body segment and calculated anatomically correct joint angles for ankle, knee, hip, pelvis, and trunk motions.[38] These joint motions at initial contact and mid-stance during two trials (total of four to five gait cycles) were averaged for each testing session. Joint motion data from the same lower extremity with the greatest amount of excessive ankle plantar flexion in stance during ambulation without orthoses were used. This extremity was the same limb measured using surface EMG in all four tests.

Data Analysis

Descriptive statistics calculated were group means and standard deviations for temporal-distance gait characteristics; lower-xtremity, pelvis, and trunk joint angles at initial contact and mid-stance; and muscle timing for the five lower-xtremity muscle groups during the stance phase for the four interventions. A two-way analysis of variance (ANOVA) for repeated measures was used to test for the effects of the diagnosis, intervention, and interaction of diagnosis and intervention on temporal-distance gait characteristics, joint motions, and muscle timing. The adjusted alpha level for all ANOVA tests was set at .05/3=.02 for the three temporal-distance gait characteristics, at .05/11=.002 for the 11 joint motions at initial contact and mid-stance, and at .05/5=.01 for the timing of the five muscle groups during stance to reduce the probability of making a Type I error when multiple statistical analyses are conducted on the dependent variables.[33]

For all significant ANOVA tests, six post hoc pair-wise comparisons between interventions with Tukey's Honestly Significant Difference (HSD) test were used to determine differences at an alpha level of .05.[33] Clinical recommendations made by one of the authors (SRS) for the most appropriate orthoses for each subject based on clinical assessment of temporal-distance gait characteristics and joint motions were described. Two-way ANOVAs for repeated measures were conducted to examine the differences among subjects with the clinical recommendation of solid AFO, DAFO, or either orthosis on temporal-distance gait characteristics (P[is less than].02) and joint motions at initial contact and mid-stance (P[is less than].002).

Effect size and power were also calculated to examine die probability of making a Type II error.[33] For all nonsignificant dependent variables in the study, the effect sizes were less than .38 and the power values were less than .45 (Tab. 1). Because these values were low, inferences regarding the nonsignificant findings were prone to a Type II error. A larger sample size would be needed to increase the power of the tests. All statistical

[TABULAR DATA NOT REPRODUCIBLE IN ASCII] analyses were performed using the SYSTAT computer software package.[double dagger]

Results

Temporal-Distance Gait Characteristics

Group means and standard deviations for walking speed, stride length, and cadence for the four interventions are shown in Table 2. The two-way ANOVA for repeated measures revealed no differences at the P [is less than] .02 level between the diagnoses of spastic diplegia and hemiplegia for walking speed (F=1.21; df=1,8), cadence (F=0.18; df=1,8), and stride length (F=0.34; df=1,8). There were no differences at the P [is less than] .02 level among the interventions for walking speed (F=0.04; df=3,24). The effects of the interventions for stride length and cadence were significant (Tab. 3). There was no diagnosis X intervention interaction, indicating that the changes in stride length and cadence among the interventions were not dependent on diagnosis. Tukey's HSD post hoc pair-wise comparisons demonstrated differences at the P [is less than] .05 level in stride length and cadence for the following interventions: (1) initial 2-week period with no orthoses versus solid AFOs, (2) second 2-week period of no orthoses versus solid AFOs, (3) initial 2-week period with no orthoses versus DAFOs, and (4) second 2-week period with no orthoses versus DAFOs (Tab. 4). The mean stride length was increased and the mean cadence was decreased with both solid AFOs and DAFOs when compared with no orthoses (Tabs. 2 and 4).

Joint Motions

Group means and standard deviations for joint motions at initial contact and mid-stance for the four interventions are shown in Tables 5 and 6. The two-way ANOVA for repeated measures revealed no differences at the P [is less than] .002 level between the diagnoses of spastic diplegia and hemiplegia for joint motions of the lower extremity, pelvis, and trunk at initial contact and midstance (Tabs. 7 and 8). There were no differences at the P [is less than] .002 level in joint motions of the knee, hip, pelvis, and trunk at initial contact and mid-stance among the interventions (Tab. 7).

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

Table 7.

Results of Two-Way Analysis of Variance for Repeated Measures for Knee, Hip, Pelvic, and Trunk Motions at Initial Contact and Mid-stance (N = 10)

                                 F(a)

                                 Initial
Variable                         Contact   Mid-stance

Knee flexion/extension
  Diagnosis                      1.57      0.79
  Intervention                   0.89      2.06

Hip flexion/extension
  Diagnosis                      0.15      0.29
  Intervention                   0.20      0.38

Hip adduction/abduction
  Diagnosis                      0.43       0.01
  Intervention                   1.43       1.68

Hip rotation
  Diagnosis                      0.32       0.57
  Intervention                   0.18       0.44

Pelvic anterior/posterior tilt
  Diagnosis                      4.35       4.74
  Intervention                   0.12       0.36

Pelvic obliquity
  Diagnosis                      0.22       0.38
  Intervention                   0.31       0.10

Pelvic rotation
  Diagnosis                      0.01       0.02
  Intervention                   0.44       2.28

Trunk anterior/posterior lean
  Diagnosis                      0.19       0.24
  Intervention                   2.11       0.68

Trunk lateral lean
  Diagnosis                      0.17       0.53
  Intervention                   0.31       0.19

Trunk rotation
  Diagnosis                      0.23       0.64
  Intervention                   0.55       0.55

(a) All F values (df=1,8) for diagnosis and all F values (df=3,24) for intervention were not significant at P [is less than] .002.

Only the effects of the interventions for ankle motions at initial contact and mid-stance were significant at the P [is less than] .002 level (Tab. 8). There was no diagnosis X intervention interaction, indicating that the change in ankle motion at initial contact and mid-stance among the interventions was not dependent on diagnosis (Tab. 8). Tukey's HSD post hoc pair-wise comparisons showed differences at the P [is less than] .05 level for the following interventions: (1) initial 2-week period with no orthoses versus solid AFOs, (2) second 2-week period of no orthoses versus solid AFOs, (3) initial 2-week period with no orthoses versus DAFOs, and (4) second 2-week period with no orthoses versus DAFOs, (Tab. 9). The amount of ankle plantar flexion that occurred at initial contact and mid-stance in the interventions with no orthoses was reduced with both solid AFOs and DAFOs (Tabs. 5 and 6).

Table 8. Results of Two-Way Analysis of Variance for Repeated Measures for Ankle Joint Motions at Initial Contact and Mid-stance (N=10)

Variable                   df   SS        MS       F

Initial contact
  Diagnosis                 1     20.21    20.21    0.28
  Error                     8    571.47    71.43
  Intervention              3   1139.49   379.83   19.17(a)
  DiagnosisXintervention    3      5.59     1.86    0.96
  Error                    24    475.60    19.82

Mid-stance
  Diagnosis                 1     25.82    25.82     0.24
  Error                     8    853.01   106.63
  Intervention              3   1301.06   433.69    16.71(a)
  DiagnosisXintervention    3     53.13    17.71     0.68
  Error                    24    622.87    25.95

(a) Significant at P [is less than] .002.

Table 9. Tukey's Honestly Significant Difference Post Hoc Pair-wise Comparisons for Ankle Motions at Initial Contact and Mid-stance (N=10)

                                    Mean Difference

                                    Initial
Pair-wise Comparison                Contact    Mid-stance

No orthoses(a) and solid AFOs(b)    10.76(e)   11.69(e)
No orthoses(c) and solid AFOs       11.97(e)   11.31(e)
No orthoses(a) and DAFOs(d)          9.69(e)   12.79(e)
No orthoses(c) and DAFOs            10.89(e)   12.40(e)
DAFOs and solid AFOs                 1.07       1.09
No orthoses(a) and no orthoses(c)    1.21       0.38

(a) Initial 2-week period without orthoses. (b) AFO=ankle-foot orthosis. (c) Second 2-week period without orthoses. (d) DAFO=dynamic ankle-foot orthosis. (e) significant at P [is less than] .05.

Muscle Timing

Group means and standard deviations for timing of the five muscle groups during the stance phase for the four interventions (Tab. 10) showed that the duration of all muscle activity was prolonged when compared with that of children with no known pathology.[39] Muscle timing during stance in the majority of children without pathology was 0% to 43% for the tibialis anterior muscle, 0% to 33% for the quadriceps femoris muscle, 0% to 51% for the lateral hamstring muscle, 0% to 48% for the gluteus maximus muscle, and 19% to 79% for the triceps surae muscle.[39] In our study of children with CP, all muscles were active at initial contact, which is normal,[39] except for the triceps surae muscles, which fired prematurely (Tab. 10). The two-way ANOVA for repeated measures revealed no differences at the .01 level between the diagnoses of spastic diplegia and hemiplegia for timing of the tibialis anterior (F=0.17; df=1,8), triceps surae (F=3.55; df=1,8), hamstring (F=6.65; df=1,8), quadriceps femoris (F=0.12; d =1,8), and gluteus maximus (F=1.30; df=1,8) muscle groups during the stance phase. There were no differences at P [is less than] .01 among the interventions for timing of the tibialis anterior (F=1.19; df=3,24), triceps surae (F=0.53; df=3,24), hamstring (F=0.19; df=3,24), quadriceps femoris (F=1.94; df=3,24), and gluteus maximus (F=1.65; df=3,24) muscle groups during the stance phase.

[TABULAR DATA NOT REPRODUCIBLE IN ASCII]

Table 11. Results of Two-Way Analyses of Variance for Repeated Measures for the Effects of Clinical Recommendation of Orthoses on Temporal-Distance Gait Characteristics and Joint Motions

Variable                               F

Temporal-distance characteristics(a)
  Walking speed                        1.18
  Stride length                        0.16
  Cadence                              0.76

                                       Initial
Joint Motion(b)                        Contact   Mid-stance

Ankle dorsiflexion/plantar flexion     0.85      1.12
Knee flexion/extension                 0.14      0.23
Hip flexion/extension                  0.20      0.16
Hip adduction/abduction                4.25      0.01
Hip rotation                           1.99      1.89
Pelvic anterior/posterior tilt         0.17      0.06
Pelvic obliquity                       0.37      0.23
Pelvic rotation                        0.64      2.46
Trunk anterior/posterior lean          0.08      0.59
Trunk lateral lean                     0.12      0.55
Trunk rotation                         0.70      0.96

(a) All F values (df=2,7) were not significant at P [is less than] .02. (b) All F values (df=2,7) were not significant at P [is less than] .002.

Clinical Recommendations

Temporal-distance gait characteristics and joint motion data for each subject were reviewed by one of the authors (SRS) for the subjects, clinic appointment. The DAFO with a plantar-flexion stop was recommended for five subjects who showed improved knee, hip, and pelvic motions with this orthosis. Two of these subjects had improved walking speed and stride length with the DAFO, and three children had the same walking speed and stride length with both orthoses. The solid AFO was recommended for three subjects who showed improved knee and hip joint motions with this orthosis. Two of these subjects had improved walking speed and stride length with the solid AFO, and one child had the same walking speed and stride length with both orthoses. Either orthosis was recommended for two subjects who showed no differences for joint motions, walking speed, and stride length with the solid AFO or DAFO. The two-way ANOVA for repeated measures (Tab. 11), however, revealed no differences among the subjects with the clinical recommendation of solid AFO, DAFO, or either orthosis for temporal-distance gait characteristics (P [is less than] .02) and joint motions at initial contact and mid-stance (P [is less than] .002). Parents, subjects, and their physical therapists commented that the DAFO was lighter and more cosmetically appealing, but slightly more difficult for the children to initially learn to independently don and doff as compared with the solid AFO.

Discussion

Our study supports the benefits of orthotic use in children with spastic CP who demonstrate a dynamic equinus gait pattern with excessive ankle plantar flexion during the stance phase of gait. Children wearing either the conventional solid AFO or the DAFO with a plantar-flexion stop showed gait improvements, including a longer stride length, decreased cadence, and reduced excessive ankle plantar flexion at initial contact and mid-stance. Although both orthoses functioned as biomechanical restraints of excessive ankle plantar flexion, no change in the premature and prolonged activity of the triceps surae muscle group was found for either orthosis. No differences in the temporal-distance gait characteristics, joint motions, and muscle timing between the two orthoses were shown.

Children with spastic CP often demonstrate decreased walking speed, decreased stride length, and increased cadence when compared with children with no known pathology.[13,39] Previous studies[28,29] showed improved stride length with the use of inhibitive casts in children with CP. The design of the DAFO was proposed to be based on the inhibitive cast,[21,23] although no research has directly compared the effects of these two devices on gait. The finding of improved stride length with the DAFO as compared with no orthoses was consistent with the results of the inhibitive cast studies.[28,29]

No published studies have examined walking speed in children with spastic CP who wear inhibitive casts or orthoses. Walking speed can be increased by a longer stride length or a faster cadence.[13] The increased stride length found for both orthoses in our study was not enough to produce a corresponding increase in walking speed. Although the subjects walked with an increased stride length and decreased cadence with both orthoses, they also ambulated with increased cadence without the orthoses. This result produced no differences in walking speed when comparing the DAFO, solid AFO, and no orthoses.

The biomechanical effects on joint motions during ambulation were also compared between the two orthoses. The conventional solid AFO biomechanically prevents ankle plantar flexion using a three-force system at the calf, ankle, and foot.[17,18] Hylton[20,23] proposed that the DAFO's contoured footplate and total surface contact produces correct biomechanical alignment of the foot and ankle that improves distal stability and reduces compensatory, abnormal motions at the ankle and more proximal joints. Our findings showed that although both orthoses biomechanically controlled the ankle by decreasing the mild to moderate amount of excessive ankle plantar flexion at initial contact and midstance, there were no differences between the two orthoses. No changes were found in the proximal joint motions of the trunk, pelvis, hip, and knee for both orthoses at initial contact and mid-stance. Our study did not support the proposed effects of the DAFO with a plantar-flexion stop on the proximal joint motions during ambulation.

The findings of no change in knee motions at initial contact and mid-stance with the DAFO were not consistent with the results of a single-subject design study by Embrey et al.[32] They found improved knee motions in a child with CP who received physical therapy in conjunction with the use of a DAFO with free plantar flexion, which was a supramalleolar orthotic (SMO) design allowing plantar flexion.[15,20] This DAFO design with free plantar flexion used by Embrey et al[32] could account for the differences found when comparing their results with the findings of our study. The use of computerized, three-dimensional motion analysis that measures motions including rotational components[1] more accurately than the two-dimensional videographs used by Embrey et al[32] might also account for the discrepancy in the findings.

Children with spastic CP and an equinus gait pattern have abnormal timing of lower-extremity muscles in gait.[1,6-12 Carlson[21] proposed that inhibitive casts change muscle timing, with resultant improved functional movement during ambulation, although no research has examined this hypothesis. Because the design of the DAFO with a plantar-flexion stop was based on the inhibitive cast, neurophysiological changes as measured by muscle timing were hypothesized. No previous studies have examined the effects of the DAFO on muscle timing.

The proposed changes in the timing of the hamstring, quadriceps femoris, triceps surae, gluteus maximus, and tibialis anterior muscle groups during gait for the DAFO were not found in this study. Results showed no differences in muscle timing when comparing no orthoses, DAFOs, and solid AFOs. Although the abnormal premature and prolonged activity of the triceps surae muscle group in a dynamic equinus gait pattern was not changed by either the solid AFO or the DAFO, the excessive ankle plantar-flexion motion during initial contact and midstance was reduced with both orthoses.

Several factors could have affected the outcomes of this study. Subjects were tested barefoot for the two interventions without orthoses, whereas shoes were worn for all gait measurements with solid AFOs and DAFOs. Ambulation might have been affected by the use of shoes, although no published studies have compared gait with and without shoes in children with CP. Measurement errors from inconsistencies in the placement of the reflective markers on the subjects could have affected the reliability of joint angle measurements. The small sample size and the mild to moderate amount of excessive ankle plantar flexion during stance in the sample limited the generalization of the results and decreased the probability of finding a difference among the interventions. A crossover design for assigning the orthoses worn initially was not used because of scheduling constraints. There were, however, no carryover effects from the first orthoses, as no differences between the two interventions without orthoses were found. The variability in the physical therapy for gait training with the orthoses also could have affected the outcome, as this treatment was not controlled in the study.

Although no differences between the two orthoses were found, clinical examination of individual subjects' results showed that the DAFO was recommended for half of the children. No differences, however, were found among the subjects based on who received the DAFO or the solid AFO as a result of the individualized clinical assessment. Parents and subjects also found that the DAFO was lighter and more cosmetically appealing than the conventional solid AFO. Because this study showed that both orthoses improved stride length and ankle motion, the DAFO could be used instead of the solid AFO if preffered by the child and parents for cosmetic reasons. Additional factors such as orthotic cosmesis, durability, and cost; ease of donning and doffing the orthosis; and the effects on functional mobility such as sit-to-stand maneuvers or ambulation on uneven surfaces need to be considered when selecting the DAFO or solid AFO for children with spastic CP and an equinus gait pattern. Individual differences in children need to be addressed when clinical recommendations are made for orthoses because children with spastic diplegia and hemiplegia are a heterogeneous group showing variations in gait.[1,5]

Future studies are needed that include a larger sample size of children with spastic CP and moderate to severe amounts of dynamic equinus during ambulation who receive similar physical therapy for gait training with orthoses. A crossover design for assigning the orthoses worn initially would be beneficial. Measurements of joint kinetics, including hip, knee, and ankle joint moments and powers, should be compared during ambulation with solid AFOs and DAFOs. Because the rationale for the DAFO is based on the inhibitive cast, future studies should compare the effects of these two devices on gait in children with spastic CP. Future research should also examine the effects of solid AFOs, DAFOs with a plantar-flexion stop, and other orthoses such as SMOs or hinged AFOs on other functional activities such as transitions from a sitting to a standing position or from a supine position on the floor to a standing position, gait on uneven surfaces, and energy expenditure during ambulation.

Conclusion

Findings showed no differences in the temporal-distance gait characteristics, lower-extremity muscle timing, and joint motions of the lower extremity, pelvis, and trunk when comparing the solid AFO and DAFO. Both orthoses improved stride length, decreased cadence, and reduced the excessive ankle plantar flexion during stance. The DAFO's purported biomechanical effects on improving proximal joint motions and neurophysiological effects on producing more normal muscle timing during gait in children with spastic CP were not found. Other factors such as orthotic cost, cosmesis, and durability and individual differences need to be considered when selecting a DAFO with a plantar-flexion stop or a solid AFO for children with spastic CP and an equinus gait pattern. We suggest that this study be replicated with more children with CP having moderate to severe amounts of dynamic equinus during ambulation.

Acknowledgments

We thank all of the subjects and their parents for participating in the study; Patricia Taggart, PT, from Children's Hospital, Oakland, Calif, for casting subjects for the DAFOs, Don Buethorn, CPO, and the other staff from Cascade Prosthetics and Orthotics for fabricating the DAFOs; and Harry Brandt, BOC Orthotist, RTO, and the other staff in the orthotics department at Shriners Hospital for Children for fabricating the solid AFOs.

References

[1] Gage J. Gait Analysis in Cerebral Palsy. London, United Kingdom: MacKeith Press; 1991.

[2] Leonard C, Hirschfeld H, Forssberg H. The development of independent walking in children with cerebral palsy. Dev Med Child Neurol. 1991;38:567-577.

[3] Lin J, Brown J. Peripheral and central mechanisms of hindfoot equinus in childhood hemiplegia. Dev Med Child Neurol. 1992;34:949-965.

[4] Rose S, DeLuca P, Davis R, et al. Kinematic and kinetic evaluation of the ankle after lengthening of the gastrocnemius fascia in children with cerebral palsy. J Pediatr Orthop. 1993;13:727-732.

[5] Winter T, Gage J, Hicks R. Gait patterns in spastic hemiplegia in children and young adults. J Bone Joint Surg [Am]. 1987;69:437-441.

[6] Berger W, Quintern J, Dietz V. Pathophysiology of gait in children with cerebral palsy. Electroencephalogr Clin Neurophysiol. 1982;53:538-548.

[7] Brunt D, Scarborough N. Ankle muscle activity during gait in children with cerebral palsy and equinovarus deformity. Arch Phys Med Rehabil. 1988;69:115-117.

[8] Csongradi J, Bleck E, Ford W. Gait electromyography in normal and spastic children with special reference to quadriceps femoris and hamstring muscles. Dev Med Child Neurol. 1979;21:738-748.

[9] Etnyre B, Chambers C, Scarborough N, Cain T. Preoperative and postoperative assessment of surgical intervention for equinus gait in children with cerebral palsy. J Pediatr Orthop. 1993;13:24-31.

[10] Hoffer M, Perry J. Pathodynamics of gait alterations in cerebral palsy and the significance of kinetic electromyography in evaluating foot and ankle problems. Foot Ankle. 1983;4:128-134.

[11] Perry J, Hoffer M, Giovan P, et al. Gait analysis of the triceps surae in cerebral palsy. J Bone Joint Surg [Am]. 1974;56:511-520.

[12] Sutherland D, Davids J. Common gait abnormalities of the knee in cerebral palsy. Clin Orthop Rel Res. 1993;288:139-147.

[13] Norlin R, Odenrick P. Development of gait in spastic children with cerebral palsy. J Pediatr Orthop. 1986;6:674-680.

[14] Gage J, Fabian D, Hicks R, Tashman S. Preoperative and postoperative gait analysis in patients with spastic diplegia: a preliminary report. J Pediatr Orthop. 1984;4:715-725.

[15] Knutson L, Clark D. Orthotic devices for ambulation in children with cerebral palsy and myelomeningocele. Phys Ther. 1991;71:947-960.

[16] Rosenthal R. The use of orthotics in foot and ankle problems in cerebral palsy. Foot Ankle. 1984;4:195-200.

[17] Butler P, Nene A. The biomechanics of fixed ankle-foot orthoses and their potential in the management of cerebral-palsied children. Physiotherapy. 1991;77:81-88.

[18] Condie D, Meadows C. Ankle-foot orthosis. In: Bowker P, Condie D, Bader D, Pratt D, eds. Biomechanical Basis of Orthotic Management. Oxford, United Kingdom: Butterworth-Heinemann Ltd; 1993:99-123.

[19] Harris S, Riffle K. Effects of inhibitive ankle-foot orthoses on standing balance in a child with cerebral palsy. Phys Ther. 1986;66:663-667.

[20] Hylton N. Postural and functional impact of dynamic ankle-foot orthoses and foot orthoses in a pediatric population. Journal of Prosthetics and Orthotics. 1989;2:40-53.

[21] Carlson S. A neurophysiological analysis of inhibitive casting. Physical Therapy and Occupational Therapy in Pediatrics. 1984;4:31-42.

[22] Duncan W, Mott D. Foot reflexes and the use of inhibitive casts. Foot Ankle. 1983;4:145-148.

[23] Hylton N. Dynamic casting and orthotics. In: Glenn M, Whyte J, eds. The Practical Management of Spasticity in Children and Adults. Philadelphia, Pa: Lea & Febiger; 1990:167-200.

[24] Cusick B, Sussman M. Short leg casts: their role in the management of cerebral palsy. Physical Therapy and Occupational Therapy in Pediatrics. 1982;2:93-110.

[25] Sussman M, Cusick B. Preliminary report: the role of short leg tone-reducing casts as an adjunct to physical therapy of patients with cerebral palsy. Johns Hopkins Medical Journal. 1979;145:112-114.

[26] Otis J, Root L, Kroll M. Measurement of plantar flexor spasticity during treatment with tone-reducing casts. J Pediatr Orthop. 1985;5:682-686.

[27] Watt J, Sims D, Harckham F, et al. A prospective study of inhibitive casting as an adjunct to physiotherapy for cerebral palsied children. Dev Med Child Neurol. 1986;28:480-488.

[28] Bertoti D. Effect of short leg casting on ambulation in children with cerebral palsy. Phys Ther. 1986;66:1522-1529.

[29] Hinderer K, Harris A, Purdy A, et al. Effects of tone-reducing versus standard plaster casts on gait improvement of children with cerebral palsy. Dev Med Child Neurol. 1988;30:370-377.

[30] Ricks N, Eilert R. Effects of inhibitory casts and orthoses on bony alignment of foot and ankle during weightbearing in children with spasticity. Dev Med Child Neurol. 1993;35:11-16.

[31] Sussman M. Casting as an adjunct to neurodevelopmental therapy for cerebral patsy. Dev Med Child Neurol. 1983;25:804-805.

[32] Embrey D, Yates L, Mott D. Effects of neurodevelopmental treatment and orthoses on knee flexion during gait a single-subject design. Phys Ther. 1990;70:626-637.

[33] Portney L, Watkins M. Foundations of Clinical Research: Applications to Practice. East Norwalk, Conn: Appleton and Lange; 1993.

[34] Buethorn D. Dynamic Splinting Workshop Syllabus. Bellingham, Wash: Cascade and Prosthetics and Orthotics Inc; 1993.

[35] Rose S, Ounpuu S, DeLuca P. Strategies for the assessment of pediatric gait in the clinical setting. Phys Ther. 1991;71:961-980.

[36] Delagi E, Iazzetti J, Perotto A, Morrison D. Anatomic Guide for the Electromyographer: The Limbs. Springfield, Ill: Charles C Thomas Publisher Co; 1980.

[37] Di Fabio R. Reliability of computerized surface electromyography for determining the onset of muscle activity. Phys Ther. 1987;67:43-48.

[38] Kadaba M, Ramakrishnan H, Wootten M. Measurement of lower extremity kinematics during level walking. J Orthop Res. 1990;8:1-9.

[39] Sutherland D, Olshen R, Biden E, Wyatt M. The Development of Mature Walking. London, United Kingdom: MacKeith Press; 1988.