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

Mechanical Gait Analysis of Transfemoral Amputees: SACH Foot Versus the Flex-Foot

Pamela A. Macfarlane, PhD
David H. Nielson, PhD, PT
Donald G. Shurr, MA, PT, CPO

ABSTRACT

The purpose of this study was to identify gait mechanics that might explain the physiological benefits found when transfemoral amputees walked over ground using a Flex-FootR compared to a SACH foot fastened below a hydraulic knee joint. Five active traumatic unilateral transfemoral amputees were videotaped as they walked overground at five controlled speeds ranging from 1.5 to 3.5 mph on a continuous path.

Analysis of the videotape data was conducted on the mean step length, early and late swing and stance phases, and double- and single-support phases for each speed. Since there were no speed-by-foot-type statistical interactions, the foot-type comparisons could be made across all speeds. The only significant differences due to foot type were found in the uninvolved double- and uninvolved single-support phases and the late stance-phase ratio (involved late stance phase/uninvolved late stance phase). These differences appear to be caused by a delay in the involved toe-off while using the Flex-Foot.

The Flex-Foot is designed to deform during weight acceptance and reform giving a "pushoff" during late stance. The additional time needed to reform the Flex-Foot could explain the phase differences, and this pushoff mechanism could explain the physiological benefits of walking with the Flex-Foot compared to the SACH foot. With a pushoff, the inertia of the involved limb may be overcome, and the prosthetic limb may be recovered with less upper-body movement and thus less energy would be expended, resulting in a more efficient, symmetrical gait.

Key Words: Above-Knee Amputee; Gait Analysis; Flex-FootR; SACH Foot.

Introduction

People with transfemoral amputations have been shown to walk slower using significantly more energy and adopting an asymmetrical gait compared to people without impairments. The ideal prosthetic leg design for transfemoral amputees would allow optimal gait biomechanics while using as little energy as possible. The hip, knee and ankle joints must coordinate to allow efficient walking mechanics.

Energy cost and gait mechanics are interrelated, suggesting any comprehensive analysis of gait should include both physiological and biomechanical variables. The purpose of the present study was to conduct a comprehensive analysis of gait in unilateral transfemoral amputees during walking over a functional range of speeds using two types of prosthetic feet (the SACHa foot and the Flex-FootRb). The study included both physiological and mechanical gait parameters. The physiological results are presented in a separate article (1).

The purpose of this article is to present the mechanical gait analysis. The primary objective was to document the foot type and walking speed effects on the gait mechanics of transfemoral amputees' ambulation in addition to identifying potential contributing factors that could explain the observed physiological differences reported in the related article (see pages 138-143). Table A contains a list of definitions used in the study.

Review of Literature

Subjects with lower-limb amputations walk slower and use a less efficient, asymmetrical gait compared to normal subjects with nonpathological gait (2-4). More proximal amputations exacerbate these physiological and kinematic complications (5). Walking dynamics in normal subjects, transtibial amputees and transfemoral amputees have been studied previously with a considerable amount of interest focused on the effects of variations in joint components and prosthetic designs. On flat, level surfaces various knee joint designs were reported to have little effect on transfemoral amputee gait (6), suggesting the distal components may be more important determinants of the gait mechanics. In this context the design of the ankle and foot has prompted considerable attention.

Of particular interest has been the comparison of the conventional type of prosthetic foot with the newer dynamic elastic response foot (7), sometimes referred to as the energy-storing/releasing prosthetic foot. The solid-ankle cushioned heel (SACH) is an example of a conventional foot that does absorb some force at heelstrike due to its cushioned heel but that is otherwise rigid. The Flex-Foot, which has a flexible pylon that deforms during weight acceptance and reforms to give a "pushing off" action during unweighting, is an example of the dynamic elastic response foot design. A more detailed description of these two types of prosthetic feet is provided in the physiology article preceding this article (1).

In normal walking the knee is extended at heelstrike; it then flexes and extends again prior to heeloff. This mechanism, controlled by eccentric and concentric contraction of the quadricep femoris muscle group, serves to limit vertical trunk displacement while the body passes over the weighted foot. The prosthetic leg of the transfemoral amputee does not show this knee flexion motion due to the lack of active joint control subsequent to lost quadricep muscle function (8). During weightbearing, the knee must be fully extended, and often hyperextended, to prevent it from buckling into flexion due to loss of quadricep control (9). Absence of the knee flexion mechanism and a solid, stiff keel foot during stance cause an increase in the vertical trunk displacement during the walking cycle. A greater vertical displacement of the center of mass during gait is associated with an increase in mechanical work (10), which has been shown to correlate highly with the energy cost or oxygen uptake (r > 0.94) in transtibial amputees during multiple-speed level walking (11).

Research has been conducted on gait differences using various dynamic elastic response feet (12-14); in some studies, comparisons also were made to the conventional, non-energy-storing prosthetic foot (4,15,16). The Flex-Foot has been shown to improve the gait symmetry in active transtibial amputees compared to a conventional prosthetic foot (15). The late stance phase of transtibial amputees is shorter on the involved side than on the uninvolved side when using the conventional foot, but the difference is less when the Flex-Foot is used compared to the conventional foot, thus improving gait symmetry (16). Similarly, Menard et al. (13) found the Flex-Foot afforded greater gait symmetry compared to the Seattle foot in transtibial amputees.

Walking velocity, cadence, stride length and stride time in traumatic transfemoral amputees are two standard deviations below normal mean values and one standard deviation below normal mean values in traumatic transtibial amputees (2). Comparative data at self-selected walking velocity from male normal subjects (S-SWV) (18), transtibial amputees and transfemoral amputees (19) are as follows: S-SWV: 82, 68 and 55 m•min-1; cadence: 108, 98 and 86 steps•min-1; and stride length: 1.51, 1.38 and 1.26 m•stride-1, respectively. As the self-selected walking velocity and walking confidence decrease, subjects tend to spend less time in single support and more time in double support.

Unilateral leg amputation results in an asymmetrical gait, which has been linked not only to chronic degenerative changes in the lumbar spine and knees (20) but also to the increased energy cost of walking. In a study of 30 "good gait" transfemoral amputees, Zuniga et al. (21) reported a significantly shorter involved single-support phase (34-percent stride time) compared to the uninvolved single-support phase (42-percent stride time). The source of this asymmetry could be the limited time the subject is willing to bear full body weight on the involved leg compared to the uninvolved leg or it could be the involved leg is more difficult to swing forward in recovery, necessitating a longer single-support phase on the uninvolved leg. Hale (22) found increasing the mass of the involved lower leg in transfemoral amputee subjects did not change the involved swing time, suggesting it is not the weight of the involved leg that limits its mobility.

As the level of amputation becomes more proximal, the interstep variability increases (22), and amputees who have not adapted to their prostheses do not walk as effectively as they might with accommodation (23). To gather reliable gait data, transfemoral amputee subjects must be given adequate time to adapt to the prosthesis and become oriented to the specific test procedures and facilities.

Overall, compared to the non-energy-storing designs, dynamic elastic response foot designs appear to be beneficial for active transtibial amputees. It is not known if transfemoral amputees also might benefit from dynamic elastic response foot designs. In this study it was hypothesized the gait would be more symmetrical using the Flex-Foot when compared to the SACH foot, and the mechanical gait characteristics would differ from normal in the same direction as seen in transtibial amputee gait but to a greater extent.

Methods

Design

The independent variables in this study were the type of involved foot and the walking velocity. A repeated measures design was used to compare SACH to Flex-Foot ambulation over a functional range of walking velocities: 40.2, 53.6, 67.1, 80.5 and 93.9 m•min-1 (1.5, 2.0, 2.5, 3.0 and 3.5 mph), respectively.

The kinematic dependent variables all are related to foot placement viewed from the side. Figure 1 is a temporal diagram showing the phase variables that were studied. The variables include 13 phase variables (involved and uninvolved leg: early and late swing, early and late stance, single and double support; and stride time) and two step length variables (involved and uninvolved leg). All gait variables were calculated for each subject, each contralateral foot, at each speed and each foot type. Seven symmetry variables (early and late stance, early and late swing, single and double support, and step length) were calculated as the ratio of the involved leg value to the uninvolved leg value.

Procedures

The five active male midfemoral unilateral traumatic amputee subjects and general procedures for collecting data have been described in the physiology article from this study (1). Subject group descriptive statistics are presented in Table B . All subjects judged by the prosthetist were good walkers with both the Flex-Foot and the SACH foot. The same individual fitted all subjects with a prosthesis designed to accept both the SACH foot and the Flex-Foot below an SNS (swing and stance) hydraulic knee joint, which is designed for physically active individuals.

To ensure acclimation to the prosthetic foot for testing purposes, subjects wore that prosthesis continuously for the week prior to testing. Subjects were instructed to wear the same footwear and keep the same hydraulic unit settings throughout the testing period. Participation involved three sessions: an orientation session where they practiced walking under testing conditions and two test sessions, one using the Flex-Foot and one with the SACH foot. Foot-type condition was in random order.

Subjects walked continually around a 50-m indoor walkway that had long straight sides and a large semicircular path at each end. This pathway allowed the subjects to maintain their walking velocity. The velocity was controlled by an assistant walking alongside the subject holding a speedometer walking cane (24). Subjects walked continually for at least four minutes.

A VHS videocamera (Panasonic AG-190-P) was positioned with the recording plane parallel to a straight section of the walkway, far enough back to record four strides of each subject's walking each time he or she passed in front of the camera. At least seven trials were recorded for each subject under each condition. Prior to each subject's test a 1-m linear scale was videotaped in the plane of walking for later use in transforming the recorded linear measures to actual measures. A time code was inserted onto the videotape using a time code generator (Bio-Electronics, SMPTE TC-3) so each of the 60 video fields per second was encoded. Data from the video image were extracted and analyzed using a method described elsewhere (25). All gait values were calculated as the mean of three consecutive strides from one selected trial. Criteria for trial selection included the last trial at each speed in which at least three complete gait cycles were visible and the subject appeared to walk consistently.

Swing-, stance-, double- and single-support phase variables, shown in Figure 1 , were calculated as the mean time taken from the instance initiating the phase to the instance ending the phase for three consecutive strides. Step length values were converted to centimeters using a scale factor calculated from the recorded linear scale. Symmetry ratios for all of the phase and distance variables were calculated using the values for the involved leg divided by those from the uninvolved leg.

Data Analysis

Descriptive statistics were calculated on the dependent variables. A two-way repeated measures analysis of variance (ANOVA) was used to test for main effects on walking speed and type of prosthetic foot as well as for any interaction between speed and foot type. To decrease the chance of type I errors associated with the testing of multiple dependent variables, a Bonferroni adjustment was made to an alpha level of 0.05, resulting in an adjusted p value of <0.003 for the phase and distance gait variables (0.05/15) and <0.007 (0.05/7) for the symmetry variables to reach significance (26).

Results

One subject could not walk at 3.5 mph using his SACH foot; no data for this subject at 3.5 mph were included in the analysis. Gait data for one subject walking at the slowest speed with his Flex-Foot was missing due to camera malfunction. All other data were included.

The mean values for the gait variables are graphically presented in Figures 2-5. As seen in Figure 2 , the early and late stance phases for all walking conditions appeared to systematically decrease with increases in walking velocity. The late stance-phase decreases were more dramatic. Contralateral leg comparisons indicated longer early stance phases for the uninvolved leg. Similar but less pronounced differences were seen during late stance. Foot-type differences were negligible.

As seen in Figure 3 , the involved and uninvolved late swing phases tended to decrease with increases in walking velocity. Walking velocity appeared to have little effect on the early swing phases. At all speeds the swing phase for the involved leg was longer than for the uninvolved leg. Foot-type differences were negligible.

The support-phase data tended to decrease with increases in walking velocity (see Figure 4) , with more noticeable decreases in double support. Contralateral leg comparisons showed longer uninvolved versus involved differences. There did appear to be foot-type differences.

The step length results (see Figure 5) indicated that as walking speed increased, step length increased regardless of leg or foot type. Contralateral leg- and foot-type comparisons revealed no systematic differences.

Figure 6 is a schematic diagram showing the means (and standard errors) of all variables summed across speed and foot type. From this figure it can be seen that the phases of the gait that occur after the involved leg heelstrike and before the uninvolved leg heelstrike are of shorter duration than the contralateral phases. The mean duration of the sum of these phases across foot type and speeds was 58.39 sec•100-1 versus 71.92 sec•100-1 (a ratio of 0.81).

There was no foot type by speed interaction for any of the gait variables studied. This allowed the foot-type analysis to be conducted across all speeds. Table C presents the ANOVA results for the foot-type phase and step-length comparisons. The only significant foot-type differences found were in the uninvolved double support

(p = 0.002) and uninvolved single support (p = 0.003) phases. Compared to the SACH foot, Flex-Foot walking was associated with significantly longer uninvolved double-support and significantly shorter uninvolved single-support phases.

The ANOVA results for the foot-type comparisons for the seven ratio variables are presented in Table D . The late stance-phase ratio was the only gait symmetry variable significantly affected by foot type. Summed across speeds, the mean late stance-phase ratio while walking using the Flex-Foot (0.98) differed significantly (p = 0.004) from this ratio while using the SACH (0.94). Across all speeds Flex-Foot walking was associated with more symmetrical late stance phase ratios.

Discussion

The mechanical gait analysis was used to identify gait characteristics that could help explain the physiological differences observed between Flex-Foot and SACH foot walking in this group of subjects previously described (1). Across speeds, Flex-Foot walking was associated with significantly lower percentages of age-predicted maximum heart rate, lower energy expenditure and improved gait efficiency compared to SACH foot walking. The only statistically significant biomechanical gait differences due to foot type were in the uninvolved double-support phase, the uninvolved single-support phase and the involved late-stance phase ratio. Delaying the involved toe-off while walking with the Flex-Foot would appear to account for these phase differences.

The major design function of the dynamic elastic response foot is to compress during stance and recoil during unweighting. Due to this "ankle" recoil, the Flex-Foot would be expected to stay in contact with the ground longer than the SACH foot, which cannot compress or recoil. The time taken to reform the Flex-Foot could account for a delay in the involved toe-off. During this time the dynamic elastic response foot pushes against the floor prior to being lifted and swung forward. The pushoff may serve to overcome the inertia of the prosthesis.

Using ground-reaction force-plate data collected during walking with the Flex-Foot, Menard et al. (13) reported a "kick" very late during stance. These researchers did not think this kick was useful in propelling the body or the prosthetic limb. This reaction force might, however, facilitate the transfemoral amputee who, unlike the transtibial amputee, does not have an active hamstring muscle to lift the lower segment of the prosthesis. Rotating the pelvis about a longitudinal axis through the weighted leg facilitates the forward swinging of the lower limb in transfemoral amputee gait. This pelvic rotation, which is accompanied by compensating upper-body movements, is associated with an increased energy cost (27). If, unlike the SACH foot, the Flex-Foot does provide a "pushoff" during late stance as claimed by the designers, then less energy would be needed to rotate the pelvis and swing the leg forward, and the gait would be more symmetrical.

While it was hypothesized that transfemoral amputee walking would deviate from normal subject walking in a similar but more asymmetrical manner than transtibial amputee walking, this was not found to be the case. While the key to the asymmetrical gait in transtibial amputees has been shown to be the limited time the subject spends in late single stance on the prosthetic foot (15), the key to the transfemoral amputee gait asymmetry appears to be during the involved leg's late swing and early stance phases. The slowing down of the recovering prosthetic leg and the initial weight acceptance onto that leg would explain the asymmetry.

As the involved leg swings forward, the thigh and shank segments swing like a pendulum. During this swing phase, the shank portion of the prosthesis must first be controlled then slowed so the prosthetic knee can be locked into full extension prior to heelstrike and weight acceptance. Without quadricep muscle action to control the knee extension or the hamstring muscles to slow the knee extension, the prosthetic late swing phase is delayed compared to the intact side that does have a muscle-controlled knee motion. As walking speed increased, the involved late swing phase and step rate both changed in the same direction as seen in normal walking. This suggests that even in lieu of lost knee joint control the transfemoral amputees in this study still were able to control shank movement and achieve timely knee extension via the SNS knee unit with its design bias.

This finding appears to be in contrast to other reports on transfemoral amputees using the Otto Bock-4R20 knee unit. Explained by restricted flexion-extension prosthetic knee motion, Boonstra et al. (28) indicated the amputee is forced to vary walking speed only by means of the sound leg. Jaegers et al. (29) suggested difficulty in controlling the swing speed of the knee mechanism influences the spontaneous choice of step rate.

If the lack of muscle did limit the control of the involved late swing phase, adding weight to the limb would be expected to exacerbate this problem; however, Hale (22) did not find this when weight was added to the shank segment in a group of transfemoral amputees. During shank deceleration Hale found significant increases in hip muscular effort. It appears the transfemoral amputee can control the prosthetic limb swing, apparently using muscular effort from the hip, but this control increases the duration of the involved late swing phase. The hip control of the involved limb during the swing phase needs to be investigated further.

Immediately after involved heelstrike, the transfemoral amputee shifts weight forward quickly over the involved leg so the weight vector can pass anterior to the knee, preventing it from flexing during weight acceptance. In the uninvolved leg the quadriceps muscle controls the knee flexion during stance. Compared to the involved early stance phase, the uninvolved early stance phase is longer because it is during this phase that the forward recovery of the contralateral involved leg has to be initiated with upper-body and pelvic rotation. The uninvolved early and late swing phases are short compared to the involved swing phases possibly to minimize the time the involved leg has to support the entire body. The involved early swing phase is short due to the difficulty the transfemoral amputee has in initiating the forward motion of the involved leg. By the time of involved leg toe-off, the inertia of the prosthesis already is overcome, and the leg moves quickly to the midswing position. From the above information it would seem the asymmetrical gait is caused by protecting the involved leg from excessive weightbearing during early stance by limiting the control of the swinging involved leg during late swing or by a combination of the two.

In conclusion, it was found transfemoral amputee walking is far more taxing physiologically and far more asymmetrical than transtibial amputee or normal subject walking. Across a functional range of walking speeds, it took transfemoral amputees longer to clear the prosthetic foot off the ground when walking with the Flex-Foot than with the SACH foot. This may have been a result of the time needed for the Flex-Foot to reform after being compressed during weight acceptance. The differences in walking when using a Flex-Foot compared to a SACH foot resulted in a more symmetrical late stance phase and a decrease in the physiological requirement of walking. The Flex-Foot does appear to benefit active individuals with midfemoral amputations; although these benefits are small, they are statistically significant. Even small benefits that can be afforded transfemoral amputees may have practical significance in light of the large biomechanical and physiological stresses they endure due to their gait impairment. (5)


References:

  1. Macfarlane PA, Nielsen DH, Shurr DG, Meier K, Clark R, Kerns J, Moreno M, Ryan B. Transfemoral amputee physiological requirements: comparisons between SACH foot walking and Flex-Foot walking. JPO 1997;9:4:138-143.
  2. Waters RL, Perry J, Antonelli D, Hislop H. Energy cost of walking of amputees: the influence of level of amputation. JBJS 1976;58(A):42-6.
  3. Skinner HB, Effeney DJ. Special review: gait analysis in amputees. Am J Phys Med 1985;64:2:82-9.
  4. Nielsen DH, Shurr DG, Golden JC, Meier K. Comparison of energy cost and gait efficiency during ambulation in below-knee amputees using different prosthetic feet-a preliminary report. JPO 1989; 1:1:24-31.
  5. Waters RL, Yakura JS. The energy expenditure of normal and pathological gait. Phys & Rehab Med 1989;1:3:183-209.
  6. Godfrey CM, Jousse AT, Brett R, Butler JF. A comparison of some gait characteristics with six knee joints. Orth Pros 1975; 29:3:33-8.
  7. Michael J. The energy-storing feet: a clinical comparison. Clin Pros Orth 1987, 11:154-68.
  8. Suzuki KM, Takahama M, Mizutani Y, Arai M, Iwai A. Locomotive mechanics of normal adults and amputees. In: Matsui H, Kobayashik K (eds.). Biomechanics VIII-A. Champaign, Ill.: Human Kinetics Publishers, 1983, 380-5.
  9. Perry J. Gait analysis: normal and pathological function. Englewood Cliffs, N.J.: Slack Inc., 1992.
  10. Hay JG. The biomechanics of sports techniques. 3rd ed. Englewood Cliffs, N.J.: Prentice Hall Inc., 1985.
  11. Macfarlane PA. Biomechanical, physiological and subjective parameters in walking with two types of below-knee prostheses. Unpublished doctoral dissertation, The University of Iowa, 1989.
  12. Torburn MS, Perry J, Ayyappa E, Shanfield SL. Below-knee amputee gait with dynamic elastic response prosthetic feet: a pilot study. J Rehab Res & Devel 1990;27:4:369-84.
  13. Menard MR, McBride ME, Sanderson DJ, Murray DD. Comparative biomechanical analysis of energy-storing prosthetic feet. Arch Phys Med & Rehab 1992;73:451-8.
  14. Barth DG, Schumacher L, Sienko S. Gait analysis and energy cost of below-knee amputees wearing six different prosthetic feet. JPO 1992;4:2:63-75.
  15. Macfarlane PA, Nielsen DH, Shurr DG, Meier K. Gait comparisons for below-knee amputees using a Flex-Foot versus a conventional prosthetic foot. JPO 1991; 3:4:150-61.
  16. Wagner J, Sienko S, Supan T, Barth D. Motion analysis of SACH foot vs Flex-Foot in moderately active below-knee amputees. Clin Pros & Orth 1987;11:1:55-62.
  17. Macfarlane PA, Nielsen DH, Shurr DG, Meier K. Perception of walking difficulty by below- knee amputees using a conventional foot versus the Flex-Foot. JPO 1992;3:31:114-9.
  18. Waters RL, Lunsford BR, Perry J, Byrd R. Energy-speed relationship of walking: standard tables. J Orthop Res 1988; 6:2:215-22.
  19. Bagley AM, Skinner HB. Progress in gait analysis in amputees: a special review. Clin Rev in Phys & Rehab Med 1991; 3:2:101-20.
  20. Brouwer BJ, Allard P, LaBelle H. Running patterns of juveniles wearing SACH and single-axis foot components. Arch Phys Med & Rehab 1989;70:128.
  21. Zuniga EN, Leavitt LA, Calvert JC, Canzoneri J, Peterson CR. Gait patterns in above-knee amputees. Arch Phys Med & Rehab 1972;53:373-82.
  22. Hale SA. Analysis of the swing phase dynamics and muscular effort of the above-knee amputee for varying prosthetic shank loads. Pros Orth Intl 1990;14:125-35.
  23. Zahedi MS, Spence WD, Solomonidis SE, Paul JP. Repeatability of kinetic and kinematic measurements in gait studies of the lower-limb amputee. Pros Orth Intl 1987;11:55-64.
  24. Edelstein JE. Current choices in prosthetic feet. Crit Rev in Phys & Rehab Med 1991; 2:4:213-26.
  25. Macfarlane PA, Nielsen DH. Gait analysis: a cost- and time-efficient video method for use in the clinic or classroom. Clin Kines 1996;49:4:99-105.
  26. Huberty CJ, Morris JD. Multivariate analysis versus multiple univariate analyses. Psych Bull 1989;105:2:302-8.
  27. Cappozzo A, Figura F, Gazzani F, Leo T, Marchetti M. Angular displacements in the upper body of AK amputees during level walking. Pros Orth Intl 1982;6:131-8.
  28. Boonstra AM, Fidler V, Eisma WH. Walking speed of normal subjects and amputees: aspects of validity of gait analysis. Pros Orth Intl 1993;17:78-82.
  29. Jaegers SMHJ, Vos LDW, Rispens P, Hof AL. The relationship between comfortable and most metabolically efficient walking speed in persons with unilateral above-knee amputation. Arch Phys Med & Rehab 1993;74:521-25.
  30. Nielsen DH, Gerleman DG, Amundsen LR, Hoeper, DA. Clinical determination of energy cost and walking velocity via