Secondary Logo

Journal Logo


Effect of Three Styles of Custom Ankle Foot Orthoses on the Gait of Stroke Patients While Walking on Level and Inclined Surfaces

Lewallen, James CO; Miedaner, Jim MS PT; Amyx, Scott CO; Sherman, Jack PhD

Author Information
JPO Journal of Prosthetics and Orthotics: April 2010 - Volume 22 - Issue 2 - p 78-83
doi: 10.1097/JPO.0b013e3181d84767
  • Free

One of the goals of stroke rehabilitation is safe, efficient, and comfortable ambulation. An ankle foot orthosis (AFO) is often used to aid in foot positioning and toe clearance during hemiplegic gait. Teasell et al.,1 in 2001, concluded that the AFO plays an important role in the rehabilitation of stroke patients by reducing the likelihood of falls and improving ambulation. A study by Nolan et al.2 came to the conclusion that an AFO can be used to improve functional ambulation and in hemiplegic adults. Rao et al.3 found improvements in velocity, cadence, step, and stride length in individuals with hemiparesis using a GaitRite to collect data.

De Wit et al.,4 in 2004, found that the effect of an AFO on walking ability to be statistically significant but was not clinically significant. Their study concluded that self-confidence plays a role in the motivation behind the use of an AFO.

The hemiplegic gait of stroke patients is typically characterized by excessive plantar flexion and inversion at the ankle and inadequate hip and knee flexion. Perry5 noted that this extensor synergy pattern of the plantar flexors causes problems with toe clearance during swing phase, a shortened contralateral step length, hyperextension of the knee, and little or no tibial advancement. These characteristics lead to a gait that is asymmetrical, less energy efficient, and slower. In addition, hemiplegic individuals are often prone to trips and falls. Ambulation on inclines and declines such as hills and wheelchair ramps can be exceptionally difficult. During downhill walking, step lengths are typically shorter due to increased friction demands at heelstrike.6 The increased friction demand is a result of an individual's weight being directed downhill at heelstrike as opposed to perpendicular to the walking surface as it is in level ground walking.

Although the use of AFOs in the rehabilitation of stroke patients is common medical practice, a variety of different types of AFOs can be used and few studies have looked at their potentially distinct effects on gait in different walking environments. Tyson et al.,7 in 1998, examined the effect of an articulated thermoplastic AFO on the gait of stroke patients by collecting ink footprints on paper walkways. Although the sample studied was limited to four cases, all four participants showed significant improvement in velocity, stride, and step length in both the affected and sound limbs. Three subjects showed significant improvement in step symmetry. In a face-to-face questionnaire, all subjects viewed the articulated AFO favorably and stated that they felt the AFO improved their walking.

In a similar but more comprehensive study in 2001, Tyson and Thornton8 looked at the effect of articulated AFOs on gait on level ground and included functional ambulation categories as a measure of disability. This study of 25 subjects found articulated AFOs improved functional ability, stride length of weak and sound legs, velocity, and cadence. There was neither improvement in step length of the weak or sound limb nor any improvement in symmetry. Twenty-four of the 25 subjects felt favorably about the AFO. A majority of patients felt the AFO lifted their toes, helped their leg swing forward, took weight off the foot, improved confidence, improved safety, increased the distance they walked, and increased their speed.

Hesse et al.,9 in 1996, compared gait of stroke patients while barefoot, wearing shoes only, and a solid AFO—a conventional AFO with single metal upright. Eight of the subjects were affected by clonus. The AFO condition, but not the other conditions, showed improvements in walking velocity, stride length, initial double stance duration, cadence, and deviation from a center gait line.

Other studies have examined the effect of tone-reducing AFOs. Mueller et al.10 found increased foot area contact and decreased stance duration in hemiparetic patients while wearing a tone-reducing AFO in a study of foot loading patterns using a plantar pressure analysis system. Dieli et al.,11 in 1996, compared the Dynamic supra malleolar orthosis to a prefabricated AFO with a posterior leaf spring (PLS) design and barefoot condition using three individual case studies of cerebrovascular accident (CVA) patients. This study found the Dynamic supra malleolar orthosis to be an acceptable alternative to the prefabricated AFO. Burdett et al.,12 in 1998, used videotape and footprint analysis to study stroke patients' gait. They analyzed their gait while wearing a metal AFO that prevented plantar flexion but allowed free dorsiflexion. Burdett et al. found that plantar flexion at initial contact and mid stance was decreased and excessive inversion was prevented at terminal stance and mid swing. The only temporal spatial parameter affected was the affected limb's step length, which was increased.

Smiley et al.,13 in 2002, performed a study comparing solid AFOs, articulated AFOs, PLS AFOs, and shoes-only in 14 children with spastic diplegic cerebral palsy. The study examined temporal spatial, kinematic, and energy expenditure data. It found significant differences in kinematic data, with shoes providing the most normal amount of dorsiflexion. There was no significant difference in temporal spatial measurements of velocity, cadence, and stride length measurements. Nor were there significant differences in energy expenditure with the different orthoses. When subjects were asked which brace they preferred after the conclusion of the study, eight chose the articulated AFO, six chose the PLS AFO, and no subject chose the solid AFO.

Radtka et al.14 compared solid, hinged, and shoes-only condition. The study involved healthy adults and looked at the kinematic and kinetic effects while walking on stairs. It found that the solid AFO resulted in greater kinetic and kinematic compensations resulted from the use of a solid AFO versus the hinged AFO.

Wening et al.15 authored a article that also used the GaitRite system to analyze velocity, cadence, stride length, and step length of both chronic and acute stroke patients walking with and without a prescribed AFO. They found that AFOs improved the gait parameters of both chronic and acute stroke patients. Specifically acute patients showed improvements in velocity and cadence but not step length or stride length. Chronic patients showed improvement in velocity, cadence, and stride length. This article did not involve incline walking and the type of AFO used by each patient was not controlled nor disclosed in the article.

In summary, the review of literature on the efficacy of orthoses in improving ambulation of stroke patients reveals an inconsistent array of findings. Importantly, different studies used different study populations, different orthoses, and different gait measures making it difficult to meaningfully compare types of orthoses used across a specific population. Moreover, no one study examined the interaction of specific orthoses and walking surface inclines, an important consideration relevant to real-world ambulation. It is possible that an orthosis could cause gait deviations when a subject is walking on an incline that are not present when walking on level ground.

The aim of this study was to provide a direct comparison of the most common types of custom AFOs used to assist chronic stroke victims during ambulation. It was designed to compare the effect of the different AFOs on temporal spatial parameters across varied walking conditions of inclined and declined surfaces and to determine if AFOs and surface conditions interact to affect ambulation.



Thirteen subjects with hemiparesis secondary to unilateral cerebral vascular accident were tested. Subjects were selected from our university hospital by a sample of convenience. The 10 men and 3 women were of similar age (mean = 58, SD, = 83.40, range, 44–79). Seven subjects had right cerebral lesions and six subjects had left cerebral lesions. Stroke duration was a minimum of 8-month post-CVA. To qualify for this study, hemiparetic subjects were required to have 1) sufficient cognitive and communication abilities to understand the tasks, 2) no known progressive neurological deficits, 3) absence of injury or musculoskeletal deformity of lower limb affecting gait, 4) pain free ambulation, 5) lower limb modified Ashworth score less than 3, 6) no falls in past 3 months, 7) passive dorsiflexion of greater than or equal to zero degrees with knee extended, 8) no lower limb deformity or skin lesions preventing proper fitting of AFOs, 9) vision within functional limits by patient report, 10) ability to ambulate more than 500 feet indoors and outdoors with or without walking aids, and 11) functional ambulation profile equal to or greater than 4. The functional ambulation category is a 1 to 5 ambulation rating scale designed for patients following stroke. This scale has been demonstrated to be reliable16 and rates ambulation skills based on amount of supervision or support needed to walk. A rating of one indicates the need for assistance of two people. A rating of five is completely independent anywhere.

At the time of the data collection, two participants wore an articulated AFO, two wore a solid AFO, one wore a PLS AFO, and eight wore no orthoses (Table 1). In instances when subjects were wearing an AFO design included in the study, to maintain consistency in fit of orthoses, the investigator fabricated a new AFO of the same design. Six participants used a cane and seven walked without a cane or walker. Individuals were allowed to use a cane during the study.

Table 1:
Patient characteristics

Therapists and orthotists at the University of Wisconsin Hospitals and Clinics recruited all subjects. The study protocol was approved by the University of Wisconsin-Madison's Health Sciences Instititutional Review Board. Signed informed consent was obtained from each of the subjects after explanation of the test procedures and patient rights.


Participants were seen twice. During the initial visit casts were taken using plaster bandage and a foot plate with no heel height. Three styles of AFOs were fabricated with the ankle positioned at 90 degrees before the pouring of a positive mold. All impressions, fabrication, and fittings were done by the investigating orthotist. During the second visit, the primary investigator fit each AFO and collected gait data.

The solid AFO was fabricated with trimlines anterior to the malleoli to prevent ankle dorsiflexion and plantar flexion; the articulated AFO was fabricated with Tamarack ankle joints and a plantar flexion stop at 90 degrees allowing free dorsiflexion and no plantar flexion; and the PLS AFO was fabricated with trimlines posterior to the malleoli to allow graded amounts of both plantar flexion and dorsiflexion during stance phase. All three AFOs had trimlines that ended 1 centimeter distal to the neck of the fibula proximally, and each AFO had a ¾ length footplate ending proximal to metatarsal heads. Each AFO had two straps: a two-inch strap at the proximal aspect and a one-inch strap at the instep.

All temporal spatial data were collected using the GAITRite system during the second visit. The GAITRite system has demonstrated acceptable intrarater reliability in previous studies.17 The GAITRite system consists of a 12-feet-long mat walkway with two rows of 256 pressure-activated sensors that record and process footfalls to be displayed on a computer screen as temporal spatial data. Each subject walked the length of the GAITRite mat three times on level ground for each of the three orthotic conditions and wearing shoes only. They then walked the length of the GAITRite mat after it was placed on an incline. Similar to the trials on level ground, the subjects walked up and down the ramp three times for each of the four orthotic conditions. The ramp was 10 feet long and 48 inches wide and set at a 10 degree incline with a platform on the top that was 44 inches long and 48 inches wide. Data were collected while subjects walked up the ramp and while they walked down. The order in which the three AFOs were worn was determined randomly. Subjects were asked to walk at a “comfortable pace” and were allowed to take rest breaks as needed. Subjects were allowed to have AFOs adjusted for comfort at any time during the study. Two clinicians guarded the subjects against falls during all incline trials.

At the ends of the level ground and incline trials, the subjects identified which type of AFO they preferred. Shoes-only was not offered as a choice of preferred orthotic condition.


First, each of the four dependent variables, affected step length (ASL), unaffected step length (USL), speed (SPD), and single support time (SST), were statistically analyzed employing a 4 (orthotic condition: shoes-only, solid AFO, articulated AFO, and PLS AFO) × 3 (surface: level, incline, and decline) within-subject analysis of variance. Second, for each dependent variable for which a statistically significant main effect of orthotic condition or surface was found, subsequent pair-wise comparisons were performed to clarify which treatment condition contributed to the difference within a factor. Because no statistically significant interaction was obtained between orthotic condition and surface for any of the four dependent variables, results for each factor are presented separately averaged over the levels of the other condition. Pair-wise comparisons of means for one factor were averaged across all levels of the other factor. Because of the exploratory nature of this study, alpha level for statistical significance was set at p < 0.05. All analyses were conducted with the SPSS 15.0 statistical package.


For all dependent measures except SST, there was a statistically significant main effect of either orthotic condition or surface. Results are presented for each dependent variable, first for orthotic condition and then surface.


For the orthotic factor the mean scores for the effect of each condition for each dependent measure is presented (Figure 1).

Figure 1.:
Orthotic condition averaged across surfaces.


Orthotic condition yielded a statistically significant overall effect on SPD [F(3,33) = 13.38, p < 0.0001]. Subsequent pair-wise comparisons showed that only the mean SPD of the solid AFO significantly differed from each of the other means [ts(1,11) ≥ 3.49, ps ≤ 0.005]; no significant effect of orthotic upon velocity was found between the PLS AFO, articulated AFO, and shoes only condition. Thus, the solid AFO condition yielded significantly slower mean times than each of the other orthoses.

Affected Step Length

Orthotic condition yielded a statistically significant effect on ASL [F(3,33) = 10.02, ps < 0.0001]. Subsequent pair-wise comparisons showed that the mean ASL of solid AFO condition differed from each of the other mean ASLs [ts(1,11) ≥ 3.01, p ≤ 0.01]. The solid AFO condition yielded statistically significant shorter step lengths in the affected leg than each of the other orthotic conditions.

Unaffected Step Length

Orthotic condition yielded a statistically significant effect on USL [F(3,33) = 2.92, p < 0.05]. Subsequent pair-wise comparisons showed that the solid AFO condition resulted in shorter mean step lengths in the unaffected limb than the PLS AFO, articulated AFO, and shoes-only conditions [ts(1,11) > 1.76, p < 0.035].

Single Support Time

There was no statistically significant main effect of orthotic condition on SST [F(3,33) = 1.76, p > 0.17].



Surface condition yielded a statistically significant effect on SPD [F(2,22) = 26.07, p < 0.001]. Pair-wise comparisons showed that each surface yielded a statistically significant effect on SPD [ts(1,11) ≥ 2.98, ps ≤ 0.012]. Such that mean SPD was higher on a level surface than on inclined and declined surfaces, and the decline condition yielded an SPD lower than that observed on inclined and level surfaces.

Affected Step Length

Surface condition yielded a statistically significant effect on ASL [F(2,22) = 17.911, p < 0.001]. Pair-wise comparisons showed that each surface yielded a statistically significant effect on ASL [ts(1,11) ≥ 2.51, ps ≤ 0.029]. Thus, ASLs were longest on level ground and shortest on a decline. The ASL during incline walking is statistically longer than decline walking and shorter than level ground walking.

Unaffected Step Length

Surface condition yielded a statistically significant effect on USL [F(2,22) = 12.92, p < 0.001]. Pair-wise comparisons showed that each surface condition had a statistically significant effect on USL [F,(1,11) ≥ 2.487, ps ≤ 0.030]. The USL was longest on level ground and shortest on a decline with the incline condition resulting in USLs statistically longer than the decline condition and shorter than the level ground condition.

Single Support Time

There was no statistically significant main effect of surface condition on SST [F(2,22) = 2.11, p > 0.05]. The subjects were asked to choose an AFO on level ground and an AFO on the incline that they felt was the most comfortable or their “favorite.” The responses were totaled and shown in Table 2.

Table 2:
Orthosis means across surfaces


The aim of this study was to provide a direct comparison of the most common types of AFOs used to assist stroke victims during ambulation. It was designed to compare the effect of the different AFOs on temporal spatial parameters across varied walking conditions of inclined and declined surfaces and to determine if AFOs and surface conditions interact to affect ambulation.

The solid AFO resulted in reduced SPD, reduced ASL, and reduced USL than the other three conditions. The solid AFO, when compared with the other three orthotic conditions, negatively impacted all temporal spatial parameters of gait studied except SST, on which it had no effect. The solid AFO was also the only AFO that was not chosen by a subject as his or her preferred orthosis (Table 3). The preference results (Table 3) are similar to those found by Smiley et al.13 in that no subject chose the solid AFO; however, in study of Smiley et al. no significant difference among gait parameters was found, possibly attributable to the different populations studied in their research. That this study found an adverse effect of solid AFO and no effect of other commonly used AFOs is at odds with several studies finding a positive outcome with orthoses.

Table 3:
Patient AFO preference

The study of Hesse et al.,18 in1996, found that subjects wearing a solid AFO showed improved gait when compared to barefoot and shoes-only conditions. The subject inclusion criteria, specifically the average modified Ashworth score of 3.7 versus our study's average of 1.4, may have resulted in the difference in results between Hesse's study and ours. The difference in average modified Ashworth score indicates their study had a more involved patient population.

Our study did not replicate the results found by Tyson and Thornton8 in which subjects walked better with an articulated AFO than without based on our dependent variables (Table 1). In general, the patient population in our study had a greater time between stroke onset and participation in the study and higher functional ambulation profiles than the population in Tyson's. Eighty-eight percent of the patients in Tyson's study required supervision when walking without an orthosis, in contrast to our study in which only 50% needed supervision. The average functional ambulation category of Tyson's patients was three with 60% having a score of one or two. All the subjects in this study had a functional ambulation category score of either 4 or 5.

In addition to those mentioned another difference between our study and Hesse's 1996 study and Tyson's 2001 study was the average time poststroke of each study's patients. A study by Patterson et al.19 in 2007 showed patients 2 to 5-month post-CVA had an average walking velocity of 23 to 43 cm/sec and subjects 40- to 49-month post-CVA had an average walking velocity of 73 to 77 cm/sec. Patients in our study had an average velocity of 71 cm/sec, an average stride length of 100 cm, and were an average of 62-month post-CVA. Tyson's study subjects had an average SPD of 25 cm/sec, an average stride length of 44 cm, and were an average of 8-month post-CVA and patients and Hesse's study subjects had an average barefoot walking velocity of 33 cm/sec. It would seem reasonable that higher walking velocities and longer stride lengths could be negatively affected with a solid AFO. This may account for the difference in findings concerning the solid AFO clearly warranting further study.

No interaction was found between AFO style and surface condition. In other words, performance trends walking up and down the slope were consistent irrespective of the type of orthosis the subjects were wearing. Specifically, gait was the slowest on the decline and fastest on level surface. However, performance levels did vary when one orthosis was compared with another. Typically subjects walked faster and with longer steps with orthoses that offered some movement at the ankle joint. This study considered only temporal spatial parameters and a brief preference assessment. Further studies that incorporate kinematic analysis, energy expenditure, and a more in-depth questionnaire could be useful in drawing conclusions about the interaction of AFOs and walking surface conditions. For example, the preference assessment might include shoes only as a preference.

In general, gait under the solid AFO condition was the most impaired compared to the other orthotic conditions based on our dependent variables. It was under the solid AFO condition that SPD was the slowest at 0.555 m/sec and affected and USLs were the shortest at 39.8 cm and 39.4 cm, respectively. This may be significant considering solid AFOs are used to assist stroke patients. Our inclusion criteria eliminated subjects that were less than 6-month post-CVA and could suggest that new AFOs fabricated for adults less than 6-month post-CVA could be made as a solid AFO to provide early stability with the option of adding articulating joints. Further study is needed to determine if clinicians should give consideration to modifying the type of AFO as a patient progresses during rehabilitation for a CVA.

SPD, USL, and ASL were all less under the decline condition than on level ground and inclined walking. Level ground appears to be easiest for subjects to navigate because their velocity was highest on level ground and step lengths of both affected and unaffected limbs were longest on level ground (Table 4).

Table 4:
Surface means across orthosis

A limitation to this study was the small sample size. The study also limited its assessment to temporal spatial parameters and a questionnaire; other variables such as foot clearance and knee position could be considered when recommending a specific type of AFO for patients. Our study, when compared with and combined with previous research on AFOs used among stroke patients, indicates that patient characteristics of velocity and functional ambulation category can assist in selecting the most effective AFO. Our study did not allow for an accommodation period during which subjects could adapt to each AFO. We did not incorporate an accommodation period because of time constraints. The results of this study are only one consideration in orthotic decisions.

This study found that the examined temporal spatial parameters of chronic CVA patients with characteristics similar to our inclusion criteria are not statistically different for the various AFOs studied with the exception of the solid AFO. The solid AFO was shown to result in the most compromised gait when considering SPD, step length, and SST. For chronic stroke patients with a functional ambulation category of 4 or 5, these potential negative effects of the solid AFO should be kept in mind when recommending its use in rehabilitation.


1. Teasell R, McRae M, Foley N, Bhardwaj A. Physical and functional correlations of ankle-foot orthosis use in rehabilitation of stroke patients. Arch Phys Med Rehabil 2001;82:1047–1049.
2. Nolan KJ, Savalia KK, Lequerica AH, Elovic EP. Objective assessment of functional ambulation in adults with hemiplegia using ankle foot orthotics after stroke. PMR 2009;1:524–529.
3. Rao N, Chaudhuri G, Hasso D, et al. Gait assessment during the initial fitting of an ankle foot orthosis in individuals with stroke. Disabil Rehabil Assist Technol 2008;3:201–207.
4. de Wit DC, Buurke JH, Nijlant JM, Ijzerman MJ, Hermens HJ. The effect of an ankle-foot orthosis on walking ability in chronic stroke patients: a randomized controlled trial. Clin Rehabil 2004;18:550–557.
5. Perry J. The mechanics of walking in hemiplegia. Clin Orthop Relat Res 1969;63:23–31.
6. Sun J, Walters M, Svensson N, Lloyd D. The influence of surface slope on human gait characteristics: a study of urban pedestrians walking on an inclined surface. Ergonomics 1996;39:677–692.
7. Tyson S, Thornton H, Downes A. The effect of a hinged ankle-foot orthosis on hemiplegic gait: four single case studies. Physiother Theory Pract 1998;14:75–85.
8. Tyson S, Thornton H. The effect of a hinged ankle foot orthosis on hemiplegic gait: objective measures and users' opinions. Clin Rehabil 2001;15:53–58.
9. Hesse S, Luecke D, Jahnke M, Mauritz KH. Gait function in spastic hemiparetic patients walking barefoot, with firm shoes, and with ankle-foot orthosis. Int J Rehabil Res 1996;19:133–141.
10. Mueller K, Cornwall M, Mcpoil T, et al. Effect of a tone-inhibiting dynamic ankle-foot orthosis on the foot-loading pattern of a hemiplegic adult: a preliminary study. J Prosthet Orthot 1992;4:86–92.
11. Dieli J, Ayyappa E, Hornbeak S. Effect of dynamic AFOs on three hemiplegic adults. J Prosthet Orthot 1997;9:82.
12. Burdett R, Borello-France D, Blatchly C, Potter C. Gait comparison of subjects with hemiplegia walking unbraced, with ankle-foot orthosis, and with air-stirrup brace. Phys Ther 1988;68:1197–1203.
13. Smiley S, Johnston R, Jacobsen F, et al. A comparison of the effects of solid, articulated, and posterior leaf-spring ankle-foot orthoses and shoes alone on gait and energy expenditure in children with spastic diplegic cerebral palsey. Orthopedics 2002;24:411–416.
14. Radtka SA, Oliveira GB, Lindstrom KE, Borders MD. The kinematic and kinetic effects of solid, hinged, and no ankle-foot orthoses on stair locomotion in healthy adults. Gait Posture 2006;24:211–218.
15. Wening J, Huskey M, Hasso D, et al. The effect of an ankle-foot orthosis on gait parameters of acute and chronic hemiplegic subjects. The Academy Today 2009:5.
16. Collen F, Wade D, Bradshaw CM. Mobility after stroke: reliability of measures of impairment and disability. Int J Diabil Stud 1990;12:6–9.
17. Mcdonough A, Batavia M, Chen F, et al. The validity of reliability of the GAITRite system's measurements: a preliminary evaluation. Arch Phys Med Rehabil 2001;82:419–425.
18. Hesse S, Luecke D, Jahnke M, Mauritz KH. Gait function in spastic hemiparetic patients walking barefoot, with firm shoes, and with ankle-foot orthosis. Int J Rehabil Res 1996;19:133–141.
19. Patterson S, Forrester L, Rodgers M, et al. Determinants of walking function after stroke: differences by deficit severity. Arch Phys Med Rehabil 2007;88:115–119.

ankle foot orthosis; hemiplegic gait; AFO; brace; stroke gaitrite; hemiplegia; CVA; orthosis

© 2010 American Academy of Orthotists & Prosthetists