Differences in the gait characteristics of people with diabetes and transmetatarsal amputation compared with age-matched controls

Transcription

1 Gait and Posture 7 (1998) Differences in the gait characteristics of people with diabetes and transmetatarsal amputation compared with age-matched controls Michael J. Mueller a, *, Gretchen B. Salsich a, Amy J. Bastian a,b a Program of Physical Therapy, Washington Uni ersity School of Medicine, Box 8502, 4444 Forest Park Bl d, St. Louis, MO 63110, USA b Department of Anatomy and Neurobiology, Washington Uni ersity School of Medicine, Box 8502, 4444 Forest Park Bl d, St. Louis, MO 63110, USA Accepted 3 February 1998 Abstract Although qualitative reports in the surgical literature suggest that people with diabetes mellitus (DM) and transmetatarsal amputation (TMA) walk well with regular shoes and a toe-filler, recent data indicates that this population has multiple complications and difficulty with functional mobility. A thorough description of their gait characteristics may provide insights to the cause of these difficulties. The purpose of this study was to compare selected gait characteristics of people with DM and TMA to age-matched controls. We studied 15 subjects with DM and a TMA, and 15 age-matched controls with an overall mean age of years. Data were collected with computer assisted video as subjects walked across a force platform. Range-of-motion (ROM), moments, and power were estimated at the ankle, knee, and hip in the sagittal plane using standard link-segment methods. People with DM and TMA had decreased ROM excursion, peak moments, and peak power at the ankle. At the hip, people with DM and a TMA had decreased ROM excursion, an earlier onset of the hip flexor moment, but no differences in peak moments or peak power. Since people with DM and TMA have reduced ability to generate plantar flexor power at the ankle, they appear to rely more heavily on pulling their leg forward from the hip using their hip flexor muscles. This compensation is not complete, however, as people with DM and a TMA take shorter steps and walk slower than controls. Additional research is needed to determine methods to improve or better compensate for these gait deviations during late stance phase Published by Elsevier Science B.V. All rights reserved. Keywords: Walking; Foot; Amputation; Diabetes; Human 1. Introduction When surgical criteria can be met, most authors and patients believe a transmetatarsal amputation (TMA) is preferable to a transtibial amputation because it saves the ankle and a portion of the foot, and provides a distal weight-bearing residuum [1 5]. Miller and coworkers report that 93% of a group of 160 patients with TMA were able to resume an independent gait, a functional result superior to that following transtibial amputation [2]. * Corresponding author. Fax: ; muellerm@medicine.wustl.edu Despite the benefits of a TMA compared to a transtibial amputation, patients with TMA face multiple challenges during walking. Studies indicate that skin breakdown, wound failure, or higher amputation can arise in 17 44% of patients, with an average rate of about 30% [1 5]. Recent studies have documented that peak plantar pressures, which contribute to skin breakdown, are higher on the TMA residuum than on the contralateral foot [6,7]. In addition, people with DM (diabetes mellitus) and TMA have slower walking speed (51 vs 75 m/min), shorter functional reach (19.1 vs 31.5 cm), and lower scores on a physical performance test (18.7 vs 24.1, 28 points possible) than age-matched controls [8]. These reports have selected subjects with /98/$ Published by Elsevier Science B.V. All rights reserved. PII S (98)

2 M.J. Mueller et al. / Gait and Posture 7 (1998) TMA and DM because most TMAs (70 77%) are performed in patients with DM [1 5]. There are minimal data in the literature describing specific gait characteristics of patients with TMA and DM. Qualitative descriptions in the surgical literature suggest that patients with a TMA have few problems walking. Durham and co-workers report that successful healing of the open TMA results in a durable stump with no need of a prosthetic device; only a shoe filler or a shoe modification incorporating a steel shank into the sole of the shoe will allow normal toe push off and will prevent excessive dorsiflexion [1]. In describing patients with TMA and an Achilles tenotomy for equinus deformity, Sanders and Dunlap report that although gait was not propulsive on the affected side, casual observation of these patients walking into and out of the clinic gave no clue that a TMA had been performed [5]. Garbalosa and co-workers report that patients with DM and TMA showed less dorsiflexion ROM during walking on the TMA side as compared to the intact side [7]. Force plate analysis of subjects with a TMA secondary to trauma without DM showed reduced stance duration and deficient forward propulsion on the amputated side [9]. No studies, however, have investigated kinetic and kinematic variables at the entire lower extremity of people with DM and a TMA, or compared the gait characteristics to age-matched controls. The purpose of this study was to compare selected gait characteristics of people with DM and TMA to age-matched controls. A thorough description of their gait characteristics may provide insights to the cause of their functional limitations and help identify treatment options. Because of the shortened foot, we hypothesized that people with DM and a TMA would show the following specific changes in gait characteristics compared to the age-matched controls: 1. Decreased peak plantar flexor range-of-motion (ROM), moment, and power at the ankle. 2. Earlier onset of the hip flexor moment (we speculated that patients with a TMA would have to pull their leg forward using hip flexor muscles to compensate for the lack of push-off from the ankle). 3. Shorter step length and slower walking speed. Based on a previous gait study in patients with DM and peripheral neuropathy, we did not believe patients would be able to compensate for the decreased ankle power by increasing power at the hip or knee [10]. 2. Methods 2.1. Subjects We studied 15 patients with diabetes mellitus (DM) and transmetatarsal amputation (TMA) and 15 agematched controls; the overall mean age was years. There were six females and nine males in each group. Inclusion criteria for the TMA group were a history of DM, a history of a TMA with healed incision site, and ability to walk independently without an assistive device. Only subjects with TMA and DM were selected because most TMAs occur as a complication to DM [1 5], and we wished to exclude factors such as pain which often accompanies TMA secondary to trauma [9]. The mean duration of DM was years, and the mean time since TMA was months. Ten patients had severe peripheral neuropathy (loss of protective sensation) as evidenced by their inability to sense the 5.07 Semmes Weinstein monofilament on the plantar surface of their foot [11,12]. There were four patients with bilateral transmetatarsal amputation. Subjects in the control group were recruited from family members of the patient group and from employees of the medical center. Inclusion criteria for the control group was the ability to walk independently. Exclusion criteria were a history of DM, TMA, or other severe neurological or orthopedic problem which noticeably compromised their function. There were no differences in age, height, weight, or gender between the groups (Table 1) Procedure The procedures of the study were explained to all subjects and subjects signed an approved informed consent statement prior to testing. A brief history was taken and demographic data including birth date, sex, and significant medical problems were recorded. Patients were questioned regarding history of DM, type of DM (noninsulin or insulin dependent DM) reported duration of DM, and duration of TMA Gait testing Subjects wore shorts and shoes with heels of 2.54 cm (1 inch) or less. Patients with TMA wore regular shoes with a toe-filler. If subjects did not have footwear which fit their residuum, we supplied them with standard tennis shoes that included a toe-filler. Patients were not allowed to wear therapeutic footwear because we wanted to determine their gait characteristics independent of footwear style, and because regular footwear with a toe-filler has been recommended as an adequate footwear condition [1]. Although only motion in the sagittal plane was analyzed, three-dimensional (3D) measurements of the involved thigh, leg, and foot were collected with a 3D motion analysis system sampling at 60 Hz. The spatial movement of the lower extremity segments was determined from the position of reflective

3 202 M.J. Mueller et al. / Gait and Posture 7 (1998) spherical markers, 25 mm in diameter, which were secured to the skin and shoes with doubled-sided adhesive tape. The location of markers, coordinate system, and direct linear transformations have been documented in detail elsewhere [13]. Briefly, three markers were placed on each lower extremity segment as follows; lateral thigh at the greater trochanter, mid-anterior femur, lateral epicondyle of femur (for thigh segment), lateral fibular head, mid-anterior leg, distal lateral fibula (for leg segment), superior navicular, lateral calcaneus, and lateral fifth metatarsal head (for foot segment). A tenth marker was placed on the lateral pelvis superior to the greater trochanter. This marker, along with the markers at the greater trochanter and the lateral epicondyle, was used to obtain kinematics for the hip joint in the sagittal plane (ie hip flexion and extension ROM). The nine markers on the lower extremity were measured with respect to an estimated joint center. Distance from the estimated joint center to each marker was measured with a tape measurer and used in direct linear transformation methods [13]. Positions of the lateral joint centers were the greater trochanter for the hip, the lateral epicondyle of the femur for the knee, and the distal lateral malleolus for Table 1 Subject characteristics Characteristic Group TMA No TMA Overall Number Age (years) X NS a S.D Range Sex Female NS Male Height (meters) X NS S.D Weight (kg) X NS S.D Loss of protec tive sensation Duration since TMA (months) X 31.2 S.D Duration of DM (years) X 20.7 S.D Significance a NS=not significant at the 0.05 alpha level. b Number of patients who were unable to sense the 5.07 Semmes-Weinstein Monofilament on the plantar surface of their foot. the ankle. Data were collected as subjects walked normally across a 6.8 meter walkway. A cm AMTI lg6-2-1 force platform embedded at floor level midway down the walkway sampled the force exerted by the foot on the floor during stance at 1200 Hz. Subjects were allowed to practice walking on the walkway. No effort was made to control their speed. Once the subject was accustomed to walking on the walkway, data were collected. Five trials were recorded for each subject. Subjects were allowed to rest as needed. The raw data were filtered using a Butterworth filter with a cutoff frequency at 6 Hz. Joint angles and moments at the ankle, knee, and hip in the sagittal plane (Figs. 1 and 2) were calculated using standard link-segment equations and Kintrak software (Motion Analysis Corporation). The peak plantar flexor moment, peak hip extensor moment, and peak hip flexor moment were used for analysis. Power was calculated by taking the product of the moment and the joint angular velocity. There were three power peaks that were selected for analysis; the peak ankle power during late stance (Fig. 3, A4) which represents concentric contraction of the plantar flexor muscles, the peak hip power during early stance (Fig. 3, H1) which reflects concentric contraction of the hip extensor muscles, and the peak hip power during late stance (Fig. 3, H3) which reflects concentric contraction of the hip flexor muscles [14,15]. These power variables were selected because of their importance in the control of gait. For each kinematic and kinetic variable of walking, a mean of three trials was used to enhance reliability [16]. Kinetic variables (moments and power) were normalized by body mass to reduce intersubject variability. Student t-tests were used to determine if there was a difference between groups for each of the gait characteristics. Alpha level was set at 0.05 for all apriori hypotheses [17]. To protect against a Type I error, a Bonferroni correction factor was used for analysis of all remaining variables tested for significant differences. 3. Results A complete comparison of gait characteristics for the two groups is provided in Table 2. The joint angles, moments, and power of a representative control subject and a subject with DM and a TMA are illustrated in Figs. 1 3, respectively. Consistent with hypothesis 1, the people with DM and a TMA showed decreased plantar flexion ROM ( vs , P= 0.004, Fig. 1), peak plantar flexor moment ( vs Nm/kg, P 0.001, Fig. 2), and peak plantar flexor power ( vs W/kg, P 0.001, Fig. 3) during late stance phase.

4 M.J. Mueller et al. / Gait and Posture 7 (1998) Fig. 1. Joint angles during stance for a representative control subject (A) and a subject with DM and a TMA (B). Positive values are extension motion, negative values are flexion motion. The patients with DM and a TMA showed a reduction of joint angle excursion at each joint, but particularly at the ankle into plantar flexion motion (1). The onset of the hip flexor moment occurred earlier in stance phase for the TMA subjects than for the control subjects ( vs %, P 0.02, Fig. 2), consistent with hypothesis c2. Patients with TMA walked slower ( vs m/s) and took shorter steps ( vs m) compared to controls (P 0.002) consistent with hypothesis c3. Patients with TMA showed no difference in peak hip moments or peak hip power (P 0.05, Figs. 2 and 3). Patients with TMA did show, however, decreased hip ROM during stance compared to the controls ( vs , P=0.002, Fig. 1). There were no differences in peak knee moments between the groups, but patients with TMA showed less knee ROM excursion and walked slower than controls (Table 2). 4. Discussion Greatest difference between groups were noted in ankle ROM, moments, and power during late stance phase (Table 2 and Figs. 1 3), as expected. Peak ankle power during late stance phase (Fig. 3, A4) was the most sensitive indicator of differences between groups; the TMA and DM group was less than 25% of the age-matched controls (Table 2 and Fig. 3). These changes are most likely due to the shorted foot and plantar flexor lever arm. Although there were no differences in magnitudes of moments or power at the hip or knee, there was a significant difference in timing of the hip flexor moment. People with a TMA began their hip flexor moment at a mean time of 47% of stance compared to 61% of stance for control subjects (Table 2 and Fig. 2). Compared to age-matched controls, people with a TMA appeared to rely more heavily on advancing their leg using the hip flexor muscles rather than the plantar flexor muscles which have a shortened lever arm. Other studies have reported that various patient populations with biomechanical constraints at the foot and ankle walk using a similar pattern. Patients with DM and peripheral neuropathy showed less ankle mobility, ankle moments and power, but no difference in knee or hip moments or power when compared to age-matched controls. Mueller and co-workers [10] speculated that patients with peripheral neuropathy pulled their legs forward primarily using hip flexor muscles (hip strategy) rather than pushing the leg forward primarily using plantar flexor muscles (ankle strategy). A similar pattern with greater deviations is seen in this current study, as one might expect given the greater biomechanical deficit at the foot (i.e. TMA). Although meth-

5 204 M.J. Mueller et al. / Gait and Posture 7 (1998) Fig. 2. Joint moments during stance for a representative control subject (A) and a subject with DM and a TMA (B). Positive values are extension moments, negative values are flexion moments. The patients with DM and a TMA showed an earlier onset of the hip flexor moment (1) and a reduction in the peak plantar flexor moment (2). ods of the two studies were not identical and preclude statistical analysis, this study of people with DM and TMA showed decreased ankle moments (0.93 vs 1.03 Nm/kg), decreased ankle power (0.43 vs 1.05 W/kg), and decreased walking speed (0.86 vs 1.06 m/s) compared to people with DM and peripheral neuropathy [10]. Interestingly, hip and knee moments and power were similar to the respective control values in either study. Patients with DM and a TMA typically have multiple medical complications [3] and we speculate they could not increase their hip or knee activity above that of healthy controls. In other gait studies, patients with below-knee amputations and children with cerebral palsy showed similar walking patterns as the people with DM and a TMA. Patients with below-knee amputation, however, showed increased hip extensor moments and power to compensate for a reduction in plantar flexor moments and power [18]. Children with cerebral palsy produced most of their work at the hip during walking rather than the typical pattern of power generated at the ankle [19]. This study supplies more evidence to a previous gait study [9] to indicate that people with TMA have substantial gait deviations. As expected, the most significant deviations occur during late stance phase when the forefoot alone would be in contact with the ground. Custom-made full-length shoes, with a total contact insert, and a rocker-bottom-sole have been shown to improve function [20] and decrease peak plantar pressures [6] in people with DM and a TMA. The rockerbottom sole is designed to help the leg roll over the foot during late stance phase and help compensate for the loss of the metatarsal phalangeal joint. Although this footwear enhances function, significant deficits persist [6]. Research is needed to help identify how patients can compensate better for the loss of foot length and toes through other orthotic components or different walking strategies. There are a number of limitations to consider in this study. First, it is difficult to determine clearly which gait changes are secondary to DM and which are secondary to TMA. Studying groups of subjects with (a) a TMA but no DM, and (b) DM but no TMA might help to clarify the influence of these individual factors on gait characteristics. The incidence of TMA without DM is low [1 5], however, and we were unable to recruit such a group. Additionally, people with TMA secondary to trauma often have considerable pain, an additional confounding variable, which limits their walking [9]. Although people with DM and peripheral

6 M.J. Mueller et al. / Gait and Posture 7 (1998) Fig. 3. Joint power during stance for a representative control subject (A) and a subject with DM and a TMA (B). Positive values represent concentric contraction while negative values generally represent eccentric contractions. H1, H3, and A4 represent peak hip extensor power, hip flexor power, and ankle plantar flexor power respectively. Patients with DM and a TMA showed considerable reduction in peak plantar flexor power (A4). neuropathy (but no TMA) have been studied in earlier reports [10,21], this study did not include such a group. Gait deviations appear to increase with the additional impairment of a TMA in people with DM and peripheral neuropathy as described earlier, but additional research would be needed to determine the exact contribution of each factor. Another potential confounding variable of this study is that ten people with DM and TMA had severe peripheral neuropathy. Decreased somatosensory input may have had an influence on their walking characteristics [10,21]. Also, four subjects had bilateral TMA. There were no significant differences, however, between people with unilateral or bilateral TMA, or those with and without sensory neuropathy in a post-hoc analysis of the gait characteristics. All subjects were included because we believed they were representative of the population of patients with DM and a TMA that would be encountered in the clinic. In addition, there was a significant difference in walking velocity between the groups and a change in velocity is known to affect gait characteristics [22]. We believe the change of walking velocity between the groups, however, was a result of the DM, shortened foot, and inability to generate power at the ankle of the TMA group, and was not the cause of the differences in gait characteristics. A change in walking velocity alone would be expected to show linear changes in all gait variables rather than a change in pattern and preferential change in magnitude at the foot and ankle as seen in this study. Finally, walking is performed in three dimensions, whereas this study only considered motion in the sagittal plane of the lower extremity on one side. Additional research is needed to determine the interaction of movement at the rest of the body, particularly the other lower extremity. Research also is required to study movement in the frontal and transverse plane because interesting compensations may be occurring that were not identified by this study. In conclusion, people with DM and TMA showed less ROM excursion, peak moments, and power at the ankle, and an earlier onset of the hip flexor moment compared to age-matched controls. The people with DM and a TMA had limited push-off power (25%) and appeared to rely more heavily on pulling their leg forward from the hip using their hip flexor muscles. Additional research is needed to determine if use of movement strategies (i.e. increased reliance on hip or knee), strengthening exercises, and/or footwear to en hance the plantar flexor moment arm can improve or better compensate for these gait deviations during late stance phase.

7 206 M.J. Mueller et al. / Gait and Posture 7 (1998) Table 2 Summary of kinetic and kinematic data from gait trials Group Significance (P value) TMA (n=15) Control (n=15) Variables with apriori hypotheses Peak plantar flexion angle ( ) Peak plantar flexion moment, normalized (Nm/kg) Peak ankle power (A4, w/kg) Onset of hip flexion moment (% stance phase) Walking speed (m/s) Step length (meters) Variables without apriori hypotheses Hip ROM excursion ( ) Peak hip extension moment normalized (Nm/kg) (NS) Peak hip flexion moment normalized (Nm/kg) (NS) Peak hip power, H1 (w/kg) (NS) Peak hip power, H2 (w/kg) (NS) Knee excursion ( ) Peak knee extension moment normalized (Nm/kg) (NS) Peak knee flexion moment normalized (Nm/kg) NS=not significant, P (NS) Acknowledgements This work was supported by the National Institutes of Health grant HD , National Center for Medical Rehabilitation Research. The authors thank Tina Elsner for her extensive help in data collection and data reduction. References [1] Durham JR, McCoy DM, Sawchuk AP, Meyer JP, Schwarcz TH, Eldrup-Jorgensen J, Flanigan DP, Schuler JJ. Open transmetatarsal amputation in the treatment of severe foot infections. Am J Surg 1989;158: [2] Miller N, Dardik H, Wolodiger F, Pecoraro J, Kahn M, Ibrahim IM, Sussman B. Transmetatarsal amputation: The role of adjunctive revascularization. J Vasc Surg 1991;13: [3] Mueller MJ, Allen BT, Sinacore DR. Incidence of skin breakdown and higher amputation after transmetatarsal amputation: Implications for rehabilitation. Arch Phys Med Rehabil 1995;76:50 4. [4] McKittrick LS, McKittrick JB, Risley TS. Transmetatarsal amputation for infection or gangrene in patients with diabetes mellitus. Ann Surg 1949;130: [5] Sanders LJ, Dunlap G. Transmetatarsal Amputation: A successful approach to limb salvage. J Am Pod Med Assoc 1992;82: [6] Mueller MJ, Strube MJ, Allen BT. Therapeutic footwear can reduce plantar pressures in patients with diabetes and transmetatarsal amputation. Diabetes Care 1997;20: [7] Garbalosa JC, Cavanagh PR, Wu G, Ulbrect JS, Becker MB, Alexander IJ, Campbell JH. Foot Function in diabetic patients after partial amputation. Foot and Ankle Int 1996;17:43 8. [8] Mueller MJ, Salsich GB, Strube MJ. Functional limitations in patients with diabetes and transmetatarsal amputation. Phys Ther 1997;77: [9] Hirsch G, McBride ME, Murray DD, Sanderson DJ, Dukes I, Menard MR. Chopart prosthesis and semirigid foot orthosis in traumatic forefoot amputation: Comparative gait analysis. Am J Phys Med Rehabil 1996;75: [10] Mueller MJ, Minor SD, Sahrmann SA, Schaaf JA, Strube MJ. Differences in the gait characteristics of patients with diabetes and peripheral neuropathy compared with age- matched controls. Phys Ther 1994;74: [11] Birke JA, Sims DS. Plantar Sensory Threshold in the Hansen s Disease Ulcerative Foot. Read at the Proceedings of the International Conference on Biomechanics and Clinical Kinesiology of Hand and Foot, Madras, India, December 16-18, l985. [12] Rith-Najarian SJ, Stolusky T, Gohdes DM. Identifying diabetic patients at high risk for lower-extremity amputation in a primary health care setting: A prospective evaluation of simple screening criteria. Diabetes Care 1992;15: [13] Nigg BM, Fisher V, Ronsky JL. Gait characteristics as a function of age and gender. Gait and Posture 1994;2: [14] Winter DA. Biomechanics of normal and pathological gait: Implications for understanding human locomotor control. J Motor Behavior 1989;21: [15] Winter DA, Eng P. Kinetics: our window into the goal and strategies of the central nervous system. Behavior Brain Research 1995;67: [16] Mueller MJ, Norton BJ. Reliability of equipment to measure kinematics of rearfoot motion: A technical report. Phys Ther 1992;72: [17] Rosenthal R, Rosnow RL. Contrast analysis: Focused comparisons in the analysis of variance. New York: Cambridge University Press, [18] Winter DA, Sienko SE. Biomechanics of below-knee amputee gait. J Biomechanics 1988;21: [19] Olney SJ, MacPhail HE, Hedden DM, Boyce WF. Work and power in hemiplegic cerebral palsy gait. Phys Ther 1990;70: [20] Mueller MJ, Allen BT, Strube MJ. Therapeutic footwear: Enhanced function in people with diabetes and transmetatarsal amputation. Arch Phys Med & Rehabil 1997;78: [21] Courtemanche R, Teasdale N, Boucher P, Flueury M, Lajoiye Y, Bard C. Gait problems in diabetic neuropathic patients. Arch Phys Med & Rehabil 1996;77: [22] Andriacchi TP. Walking speed as a basis for normal and abnormal gait measurements. J Biomechanics 1977;10:261 8.