View Options - Click to expand
Print Options - Click to expand
E-Mail Options - Click to expand

Effects of Three Types of Ankle-Foot Orthoses on the Gait and Bicycling of a Patient with Charcot-Marie-Tooth Disease

Ray G. Burdett, PhD, PT, CPed, and
Gavin Hassell, BS, CO, CPed


A patient with Charcot-Marie-Tooth disease was referred for ankle-foot orthoses for gait assistance. He was an avid bicyclist and wanted an assessment of the effects of orthoses on his bicycling, as well as his gait. The effects of three ankle-foot orthoses on joint angles during walking and bicycling, ankle torque and power production during walking, and heart rate during bicycling were measured. The quantitative effects of each AFO on walking and bicycling mechanics are discussed relative to the patient's preferences.

Charcot-Marie-Tooth (CMT) disease is a hereditary neuropathic disease resulting in progressive atrophy of the muscles distal to the knee and frequently leads to gait limitations.1 Ankle-foot orthoses (AFOs) have been shown to have positive effects on the gait of individuals with ankle muscle weakness, including restoration of heelstrike, improved control of plantar flexion after footstrike, normalization of heel rise, increased propulsion during push-off, stabilization of the knee during stance, and decrease in abnormal hip and knee flexion during swing. 1-5 AFO design may also influence gait in individuals with CMT disease. In a single-subject study of a patient with CMT disease, modification of the patient's AFO resulted in a decrease in the energy consumption during walking. 6

A literature search located a single article reporting the effects of AFOs on bicycling mechanics. In that study, 7 immobilization of the ankle with AFOs for subjects with normal muscle function resulted in reduced power transmission during the downstroke and enhanced power transmission during the upstroke.

In the current case study, we present the effects of three types of AFOs on joint angles during walking and bicycling, ankle torque and power production during walking, and heart rate during bicycling for a patient with CMT. We also discuss the relationship among these effects and the patient's preference for AFO type for walking and bicycling.


The patient, a 40-year-old man with CMT disease, had been referred by his physician for AFOs. At this time he was walking without assistive devices, but he thought he might benefit from the use of AFOs. He was also an avid recreational bicyclist, and he wanted to know what effect an AFO would have on his bicycling. He agreed to try three different AFOs to determine their effects on his gait and bicycling and to determine his preference for type of AFO for walking and bicycling.


The patient typically walked with tightly laced high-top athletic shoes. He walked more slowly than normal and had excessive hip and knee flexion during swing and excessive knee flexion and ankle dorsiflexion during stance. He had muscle atrophy distal to the knee but well-developed thigh muscles. In barefoot standing, he had bilateral cavus feet, tibial varum, calcaneal inversion, excessive dorsiflexion, and excessive knee flexion. When cycling on a Monark exercise bicycle (Monark Exercise AB, Varberg, Sweden), he preferred to have the pedal axis closer to under his ankle joint center than is typical. Even with this foot position, a large amount of dorsiflexion occurred during the downward stroke.


His dorsiflexors were graded 3-/5 bilaterally, his plantar flexors were graded 3/5 bilaterally, and he had no detectable inversion or eversion strength bilaterally. His ankle joint passive ranges of motion were: 50 degrees of plantar flexion bilaterally, 5 degrees of dorsiflexion bilaterally with knees flexed, and 0 degrees of dorsiflexion bilaterally with knees extended. At the subtalar joint, his passive ranges of motion were: right side, 20 degrees of inversion and lacking 4 degrees of eversion; left side, 15 degrees of inversion and lacking 12 degrees of eversion.



Because of the patient's size (1.78 meters height, 79.5 kilograms), activity level, and tendency to invert during gait, 4.8-mm co-polymer was used to construct the solid ankle AFO. A moderate rocker bottom was added to a pair of his high-top athletic shoes to allow for a smoother toe-off action. The ankle was set at neutral with accommodation for the 9.5-mm heel height of his shoe. Standard anterior trim lines and a sulcus length footplate were used. This device also provided resistance to inversion/eversion.


A posterior trim line AFO, also called a spring leaf AFO, was fabricated from 4.8-mm co-polymer with a posterior reinforcement to stiffen its plantar flexion resistance feature. The trim lines were very posterior, the ankle was set at neutral, and the footplate was cut to sulcus length. This AFO resisted plantar flexion but offered little resistance to dorsiflexion or inversion/eversion.


A prefabricated AFO (ToeOFF, Camp Health Care, Jackson, MI) was fit to the patient. This AFO is made of carbon and Kevlar (Dupont, Wilmington, DE) and consists of a footplate, an anterior shell, and a lateral strut. According to the manufacturer, this AFO is designed to control foot drop and to return stored energy at toe-off. We also thought that it may be effective in resisting dorsiflexion during pedaling. Because of the patient's inverted ankle position, the lateral strut pressed slightly against the lateral side of his ankle.


After construction and fitting of the AFOs were completed, the patient was told to use them as much as possible for both walking and bicycling and to determine which of the braces he preferred for each activity. After approximately 1 month of use, the patient's gait was analyzed with no AFO and with the three AFOs. Two weeks later, the patient's bicycling was analyzed with no AFO and with the three AFOs.


Gait speed is often used as a simple clinical measure of gait function. 8 To determine the effect of the AFOs on his walking speed, the patient walked over a 10-meter section of floor at a self-selected speed three times wearing each AFO and high-top athletic shoes and wearing the high-top athletic shoes only. Speed of walking was calculated for the middle 6.5 meters for each trial. The average of three trials was used to determine his self-selected speed for each condition.


To control for the effect of speed, joint angles were measured while the patient walked at a fixed speed on a treadmill. A comfortable speed of treadmill walking with minimal hand support was determined for the no-brace condition. This speed (0.54 m/sec) was used for all conditions. The right side of the patient was videotaped and a Peak Performance Technologies, Inc. (Englewood, CO) two-dimensional motion analysis system was used for data analysis. The patient wore bicycling shorts and the appropriate AFOs or shoes. Reflective markers were placed on his right greater trochanter, lateral knee joint line, lateral malleolus, lateral side of the heel of the shoe, and the lateral side of the shoe at the fifth metatarsal-phalangeal joint. The treadmill was slowly brought to speed, and after the designated speed was reached, the patient was videotaped for 1 minute. Joint angles of the thigh relative to the vertical, knee flexion, ankle dorsi/plantar flexion, and foot relative to the horizontal were measured for two gait cycles. Joint angles at right footstrike, right toe-off, and maximum joint angles during swing phase were averaged for the two gait cycles for each condition.


To calculate ankle torque and power generated during stance phase, the right leg and foot of the patient were videotaped while he walked at a self-selected speed over an AMTI (Advanced Mechanical Technology, Inc., Watertown, MA) force plate. Vertical and anterior-posterior ground reaction forces and position of the center of pressure were measured, and the position of the lateral malleolus and ankle joint angular velocity were determined from the videotape. The center of pressure and motion coordinates were transformed to the same coordinate system and synchronized in time, and then torques about the lateral malleolus were calculated for stance phase. Ankle joint power was calculated as the product of angular velocity and torque. Data for three trials for each condition were collected, normalized to 100 percent stance time, averaged for each condition, and normalized by body mass in kilograms.


The patient's preferred seat height, foot position on the pedals, and pedaling cadence were determined while he pedaled a Monark exercise bicycle using his own cycling shoes without AFOs. The work load was increased until the patient felt he was pedaling at his typical exertion level for his usual 22-kilometer ride. This seat height, pedal position, cadence (72 pedal cycles per minute), and work load (2 kilopond-meters) were used for all conditions. Reflective markers were attached as was done for the gait trials. For each condition, the patient bicycled for 5 minutes, and the last minute of each trial was videotaped. The same joint angles as for the gait trials were measured for two complete pedal cycles. Joint angles at the end of the downward and upward portions of the cycles were averaged for the two pedal cycles for each condition.


Heart rate during cycling has been shown to be linearly related to work rate. 9 Therefore, to obtain a measure of energy cost, for each AFO condition the patient's heart rate was monitored at 1-minute intervals for 5 minutes. His resting heart rate was initially determined to be 69 beats per minute. Between conditions, while the AFOs were being changed, the patient's heart rate was observed, and in all cases it returned to 90 beats per minute or lower.




Speed of walking over ground was fastest for the no-AFO condition (1.09 m/second), followed by the solid ankle AFO condition (1.04 m/second), and then the posterior trim and prefabricated AFOs (1.01 m/second). Normal self-selected adult walking speed ranges from 1.33 m/second to 1.51 m/second,10 so his walking speed was slower than normal. The differences in speed among the three AFO conditions and the no-brace condition are small, but use of the AFOs may have decreased his walking speed slightly.


The primary difference in joint angles occurred at the ankle joint during swing phase with the use of the solid ankle and posterior trim AFOs. The prefabricated AFO had little effect on the ankle joint angles compared with the no-AFO condition.

During swing phase, there was a pronounced foot drop of 23 degrees of plantar flexion without an AFO ( Table 1 ), and use of the prefabricated AFO only decreased this to 16 degrees. Use of either the solid ankle AFO or posterior trim AFO decreased the amount of foot drop considerably to almost neutral. The maximum amount of hip and knee flexion during swing was also decreased with either the solid ankle AFO or the posterior trim AFO ( Table 1 ). Thus, these two AFOs were effective in controlling foot drop and decreasing the excessive hip and knee flexion used to compensate for the foot drop during swing.

At footstrike, without an AFO and with the prefabricated AFO, the patient's ankle was plantar flexed, and there was only a slight upward angulation of the foot relative to the floor ( Table 1 ). With the posterior trim AFO and the solid ankle AFO, the ankle joint was in dorsiflexion at footstrike with a more normal heelstrike.

At toe-off, the posterior trim AFO and the solid ankle AFO each held the ankle close to neutral ( Table 1 ). Without an AFO and with the prefabricated AFO, the ankle joint was in a more plantar flexed position. This plantar flexed position of the ankle also resulted in more hip and knee flexion at toe-off, which persisted through swing phase.


nternal ankle torque generation is a combination of muscle activity, passive stretch of tissue, and resistance provided by the AFO and shoe. Compared with normal gait, all conditions resulted in much lower torque generation for both dorsiflexion and plantar flexion ( Figure 1 ). The solid ankle and posterior trim AFO conditions both resulted in dorsiflexion torques of longer duration after footstrike when compared with the no-AFO and prefabricated AFO conditions. These results are consistent with the ankle joint angle results, which showed more of a heelstrike with the solid ankle and posterior trim AFOs. The solid ankle AFO and the posterior trim AFO were both designed to be resistant to external plantar flexion torques, and they were more effective in controlling the ankle at footstrike.

During the midstance and propulsion phases of gait, the no-AFO and posterior trim AFO conditions resulted in production of lower values of plantar flexion torque compared with the solid ankle AFO or the prefabricated AFO conditions. During the midstance phase of gait, the plantar flexors normally provide resistance to forward progression of the tibia, whereas the foot remains planted on the ground, and during the propulsion phase the plantar flexors normally act to bring about heelrise and plantar flexion. The solid ankle AFO and the prefabricated AFO were more effective in providing plantar flexion torque (resistance to dorsiflexion) during these phases. The posterior trim AFO was designed to provide resistance to plantar flexion but not to dorsiflexion, so it was not as effective in providing plantar flexion torque during these phases.


Figure 2 shows ankle power generation for normal gait compared with that generated by the patient during the four AFO conditions. Normal gait has a period of power absorption by the plantar flexors to approximately 70 percent of stance phase because they provide resistance to the dorsiflexion that is occurring as the tibia progresses over the fixed foot, followed by a large peak of power generation by the plantar flexors during late stance as the ankle plantar flexes. This results in a net positive work done by the ankle muscles. In contrast, all the AFO conditions and the no-AFO condition resulted in a net negative work done by the ankle muscles of the patient. There was a period of low-level power absorption during the first 30 percent to 40 percent of stance time and then a period of almost no absorption or generation until about 60 percent of stance time. At approximately 60 percent of stance time, the ankle had a period of absorption of power caused by the increase in plantar flexion torque at this time and the continuation of dorsiflexion of the ankle. The final portion of stance phase resulted in a period of power generation, similar to that of normal gait, but the power generation occurred much later in the stance phase and was much lower in magnitude.



Knee flexion ranges of motion were similar for all AFO conditions, with the solid ankle AFO and prefabricated AFO conditions resulting in slightly less knee motion ( Table 2 ). However, dorsiflexion varied considerably among the conditions. With no AFO, the ankle joint was in a considerable amount of dorsiflexion throughout and went through a range of approximately 7 degrees of dorsiflexion. With the posterior trim AFO, the ankle was in slightly less overall dorsiflexion but still went through approximately the same amount of dorsiflexion range. The prefabricated AFO positioned the ankle in the least amount of overall dorsiflexion, but the ankle still went through approximately 8 degrees of dorsiflexion range. The solid ankle AFO provided the most resistance to dorsiflexion and the ankle went through the smallest range of dorsiflexion.

We were not able to directly measure ankle power, but ankle dorsiflexion speed ( Table 2 ) during the downward stroke is an indication of power absorption at the ankle. With no AFO and with the posterior trim AFO, the ankle reached a high dorsiflexion speed during this period, indicating that the ankle joint may be absorbing power during this portion of the power stroke. With the solid ankle AFO and the prefabricated AFO, the ankle did not reach as high a dorsiflexion speed, which may indicate less power absorption at the ankle joint during the power portion of the pedal stroke when compared with use of the posterior trim AFO or no AFO.


Heart rate was consistently higher during the final 2 minutes of cycling with no AFO, compared with all of the AFO conditions ( Table 2 ). However, heart rates did not vary appreciably among the AFO conditions. From the heart rate data alone, no recommendation of superiority of any of the AFOs can be made.



Both the solid ankle and posterior trim AFOs were effective in decreasing the excessive plantar flexion, knee flexion, and hip flexion during swing phase and in restoring a heelstrike. The solid ankle AFO, being the most resistant to dorsiflexion, was the most effective AFO in assisting with production of plantar flexion torque. No AFO was very effective in decreasing power absorption at the ankle during stance phase.

The patient felt that all of the AFOs improved his walking ability, especially during swing phase. However, he felt that this improvement was primarily cosmetic, and he used his AFOs mainly in situations in which he wore dress shoes. For everyday use, he continued to walk with high-top athletic shoes and no AFOs. With his dress shoes, he preferred to use the posterior trim AFO, with the solid ankle AFO being his second preference. He preferred the posterior trim AFO to the solid ankle AFO because he said that it was lighter in weight and easier to don with his shoes. He also stated that use of the solid ankle AFO without his rocker bottom shoes produced a noticeable resistance to a smooth forward progression at the end of stance phase. He reported that the prefabricated AFO was not comfortable for extended use because it pressed against the lateral side of his ankle.

Because of the weakness of the plantar flexors, the patient walked with excessive dorsiflexion and knee flexion during stance phase compared with normal gait, and this was not affected substantially by use of any of the AFOs. It is possible that a more plantar flexed alignment of the AFOs would decrease this excessive knee flexion during stance. Possible changes in AFO design, including different alignment, different materials, or different trim lines could have resulted in even greater improvements in his gait, particularly in controlling excessive knee flexion during stance phase.


For cycling, this patient preferred the solid ankle AFO because he felt he could apply more force to the pedals with less dorsiflexion. Indeed, our results showed that the solid ankle AFO was the most effective in limiting dorsiflexion range of motion and dorsiflexion speed during the downward portion of the stroke. This could affect cycling efficiency directly and indirectly. As a direct effect, less dorsiflexion velocity during the downward stroke would result in less power absorption at the ankle. As an indirect effect, less dorsiflexion during the power stroke may enable this patient to assume a more efficient cycling position. He could move the attachment point of the pedal more anteriorly on the shoe, which would allow him to raise his seat height. Because seat height has been shown to affect cycling efficiency,11 these changes may have a positive effect on his efficiency. However, in our testing of the various AFOs, seat height and pedal position were held constant.

As measured by heart rate, use of the AFOs seemed to have small positive effects on his cycling efficiency, but none of the AFOs seemed superior to the others. Perhaps if the seat height and foot position had been modified to take advantage of the dorsiflexion-resistant qualities of the solid ankle AFO, we might have detected a difference among the various AFO conditions. Although the patient felt that he could apply more power to the pedals with the solid ankle AFO, he did not notice any difference in the overall time of completion of his weekly 22-kilometer ride on his road bike while wearing the solid ankle AFO as compared with wearing no AFO. However, he had not made any changes in his pedal position or seat height on his road bike to accommodate the less dorsiflexed position induced by the solid ankle AFO.


Use of both the posterior trim and solid ankle AFOs during walking decreased the excessive hip and knee flexion and ankle plantar flexion during swing phase, but there were no major changes during stance phase. The patient was aware of these positive changes in his gait with use of the AFOs, but these changes were not significant enough to cause him to use AFOs in all situations, and he used them only in situations in which the appearance of his gait was important to him. When he did use AFOs for walking, he preferred the posterior trim AFO to the solid ankle AFO because it fit better into his dress shoes.

During cycling, use of the various AFOs decreased the tendency toward excessive dorsiflexion during the downward stroke, with the solid ankle AFO being the most effective. The patient preferred to use the solid ankle AFO during cycling, but he did not notice any difference in the speed of his cycling on his regular road bicycle when he used this AFO.

None of the AFOs proved to be superior in walking or cycling, and the patient preferred to use different AFOs for these tasks. More study would be needed, incorporating different AFO designs, materials, and alignments, to determine if it is possible to develop one AFO that optimizes both his walking and bicycling efficiency. Additional outcome measures, including oxygen consumption and ankle power measurements during cycling, may help to better distinguish among differences in performance attributable to the use of different AFOs.

RAY G. BURDETT, PhD, PT, CPed, is an associate professor in the School of Health and Rehabilitation Sciences, University of Pittsburgh, Pittsburgh, Pennsylvania.

GAVIN HASSELL, BS, CO, CPed, is the Director of Orthotics for De La Torre Orthotics and Prosthetics, Inc., Pittsburgh, Pennsylvania.


  1. Njegovan ME, Leonard EI, Joseph FB. Rehabilitation medicine approach to Charcot-Marie-Tooth disease. Clin Podiatr Med Surg 1997;14:99-116.
  2. Lehmann JF, Condon SM, de Lateur BJ, et al. Gait abnormalities in peroneal nerve paralysis and their corrections by orthoses: a biomechanical study. Arch Phys Med Rehabil 1986;67(6):380-386.
  3. Lehmann JF, Condon SM, de Lateur BJ, et al. Ankle-foot orthoses: effect on gait abnormalities in tibial nerve paralysis. Arch Phys Med Rehabil 1985;66(4):212-218.
  4. Lehmann JF. Push-off and propulsion of the body in normal and abnormal gait. Correction by ankle-foot orthoses. Clin Orthop 1993;288:97-108.
  5. Tyson SF, Thornton HA. The effect of a hinged ankle foot orthosis on hemiplegic gait: objective measures and users' opinions. Clin Rehabil 2001;15(1):53-8.
  6. Bean J, Walsh A, Frontera W. Brace modification improves aerobic performance in Charcot-Marie-Tooth disease: a single-subject design. Am J Phys Med Rehabil 2001;80(8):578-582.
  7. Pierson-Carey CD, Brown DA, Dairaghi CA. Changes in resultant pedal reaction forces due to ankle immobilization during pedaling. Journal of Applied Biomechanics 1997;13(3):334-346.
  8. Andriacchi TP, Ogle JA, Galante JO. Walking speed as a basis for normal and abnormal gait measurements. J Biomech 1977;10:261-268.
  9. Hesser CM, Linnarsson D, Bjurstedt H. Cardiorespiratory and metabolic responses to positive, negative, and minimum-load dynamic leg exercise. Respir Physiol 1977;30(1-2):51-67.
  10. Perry J, ed. Gait Analysis: Normal and Pathological Function. Thorofare, NJ: Slack Incorporated, 1992.
  11. Nordeen-Snyder KS. The effect of bicycle seat height variation upon oxygen consumption and lower limb kinematics. Med Sci Sports 1977;9(2):113-117.