Trans-tibial amputee gait: time-distance parameters and EMG activity

Prosthetics and Orthotics International, 2000, 24, 216-220 Trans-tibial amputee gait: time-distance parameters and EMG activity E. ISAKOV*, O. KEREN* and N. BENJUYA** *Kinesiology Laboratory, Loewenstein Hospital, Sackler School of Medicine, Tel Aviv University, Ra'anana, Israel **Biomechanics Laboratory, Kaye College of Education, Beer-Sheva, Israel Abstract Gait analysis of trans-tibial (TT) amputees discloses asymmetries in gait parameters between the amputated and sound legs. The present study aimed at outlining differences between both legs with regard to kinematic parameters and activity of the muscles controlling the knees. The gait of 14 traumatic TT amputees, walking at a mean speed of 74.96 m/min, was analysed by means of an electronic walkway, video camera, and portable electromyography system. Results showed differences in kinematic parameters. Step length, step time and swing time were significantly longer, while stance time and single support time were significantly shorter on the amputated side. A significant difference was also found between knee angle in both legs at heel strike. The biceps femoris/vastus medialis ratio in the amputated leg, during the first half of stance phase, was significantly higher when compared to the same muscle ratio in the sound leg. This difference was due to the higher activity of the biceps femoris, almost four times higher than the vastus medialis in the amputated leg. The observed differences in time-distance parameters are due to stiffness of the prosthesis ankle (the SACH foot) that impedes the normal forward advance of the amputated leg during the first half of stance. The higher knee flexion at heel strike is due to the necessary socket alignment. Unlike in the sound leg, the biceps femoris in the amputated leg reaches maximal All correspondence to be addressed to Dr Eli Isakov, Department of Orthopaedic Rehabilitation, Loewenstein Hospital, Ra'anana 43100, Israel. 216 activity during the first half of stance, cocontracting with the vastus medialis, to support body weight on the amputated leg. The obtained data can serve as a future reference for evaluating the influence of new prosthetic components on the quality of TT amputee's gait. Introduction A good prosthesis in basically fit patients enables trans-tibial (TT) amputees to ambulate freely during the entire day. Nevertheless, gait analysis in TT amputees has shown conspicuous leg asymmetry, as reflected by various measured gait parameters (Bagley and Skinner, 1991; Isakov et al, 1997). The gait of healthy subjects is characterised by a high degree of symmetry between timedistance parameters for both legs. The highest values for symmetry observed are for the step length (ratio: 0.98), the stance time (ratio: 0.96) and the double-limb support time (ratio 0.90) (Hirokawa, 1989). Muscle activity during normal gait has also been studied and some characteristics have been established. For example, the triceps surei muscles are active for approximately 10 to 50% of the gait cycle while their antagonist, the tibialis anterior, is active during the entire swing and in the first 15% of the gait cycle following initial contact. The vasti muscles, active mainly during loading response, begin their activity at 90% of the gait cycle, nearly at the end of the swing phase, The hamstrings muscles, acting eccentrically to decelerate the passively extending knee, reach maximal activity from mid to terminal swing (75% of gait cycle) (Perry, 1992; Craik and Oatis, 1995).

The amputated limb of TT amputees is less active in the functions of standing and walking. Evaluation of standing balance activity of both limbs in TT amputees showed that the footground reactive forces generated by the amputated limb are smaller when compared to the sound leg (Geurts and Mudler, 1992; Isakov et al., 1992). Reduced activity and subsequent atrophy of the thigh muscles of the amputated limb have also been demonstrated from thigh circumference measurements, muscle biopsies and muscle strength measurements. Peak torque in isokinetic concentric and eccentric contractions and maximal torque in isometric contractions are significantly smaller in the quadriceps and hamstrings muscles of the amputated leg (Renstrom et al., 1983; Isakov etal., 1996). In the present study attention was focused on kinematic parameters and on muscle activity controlling the knees during ambulation of TT amputees. Special emphasis was placed on events that precede and succeed heel strike, e.g. swing deceleration, initial contact, and loading response. Subjects Fourteen (14) males, 5 with left and 9 with right trans-tibial traumatic amputation volunteered to participate in this study. The ages of the subjects ranged from 32 to 57 years (mean age: 45.07±7.1 years), their heights from 1.60 to 1.85m (mean height: 1.7378±.0059m), and their masses from 54 to 87kg (mean mass: 73.07±7.2kg). The time lapse between the date of amputation and the time of testing ranged from 8 to 32 years, with a mean time lapse of 16.5±8.9 years. All prostheses were patellartendon-bearing (PTB) with a solid-anklecushion-heel (SACH) feet. All subjects were excellent walkers who used their prostheses on a regular basis and were leading an active normal life. Methods Before testing, all subjects were assessed by a technician to ensure optimal fit and function of the prosthesis. None of the subjects had stump problems (blisters, sores, swelling or pain) on the study day. All subjects ambulated without any other device apart from their prosthesis. Subjects were instructed to ambulate at their most comfortable speed along a walkway of Gait analysis of trans-tibial amputees 217 3.6m length. Time-distance parameters were measured by means of an electronic walkway (GaitRite). This system consists of a roll-up carpet (active area: 0.6m width, 3.6m length) that contains sensors activated by the subjects' feet while they ambulate across the walkway. The number of activated sensors determines the area of contact, the distance between the activated sensors and the time of activation/deactivation. The electronic walkway transfers this information to a PC via an interface cable. The application software serves as a control to indicate adequate walkway function, processes the raw walkway data into footfall patterns and computes the temporal and spatial parameters. Parameters measured included step, stance, swing, single support time and length of each step. Kinematic activity was assessed from data collected by filming the legs during ambulation. A single high speed (loohz) video camera was used to measure 2-dimensional kinematics, especially of the knee angle (Winanalyze by Mikromak, GmbH). A full gait stride was obtained three times from each subject. The kinematics was obtained from the averages of these three strides. The position of both knees was measured at selected gait events such as heel strike, loading response, toe off, and maximal swing flexion. A portable wireless EMG system was used to record muscle activity signals during ambulation. Surface electrodes were placed over the quadriceps (vastus medialis-vm) and hamstrings (biceps femoris-bf) of both the amputated and non-amputated thighs. The skin, over the antero-medial aspect of the thigh and above the patella, was shaved and cleaned with isopropyl alcohol 95%. The surface EMG signals were recorded in a bipolar fashion using silver/silver-chloride monopolar surface electrodes (Medicotest N-00-S 30x22mm, Olstykke, Denmark). The surface electrodes were positioned 3cm centre-to-centre distance over the muscle bellies while the ground electrode was placed over the medial condyle of the femur (Basmajian and DeLuca, 1985). The EMG signals were recorded continuously on a portable data logger MEGA ME 3000 (MEGA Electronics Ltd, Kuopio, Finland). The raw EMG signals were first treated by the preamplifiers located on the electrode leads and then filtered (15-500Hz, CMMR HOdB and gain of 412) and digitized (12 bit with sampling

218 E. Isakov, O Keren andn. Benjuya Table 1. Mean valuies ands SD of time-difference parameters of the amputated and sound legs. Gait parameters Amputated leg Sound leg Step time (s) 0.582±.04 0.569+04.001 Swing time (s) 0.438±.04 O.407±.O3.007 Stance time (s) 0.708±.05 0.744±.06.045 Single support time (s) 0.407±.03 O.438±.O4.007 Step length (m) 0.7379±.058 0.6900±.063.009 rate of looohz). The raw EMG signals were rectified and averaged over averaging time of 0.1s by means of the ME3000P Multisignal software (MEGA Electronics Ltd, Kuopio, Finland). Integrated EMG (iemg) of the VM and BF were calculated for the first and second halves of stance time and swing time. Ratios of hamstrings/quadriceps iemg were calculated as follows: Firstly, time of stance and swing phases was divided into two halves. Secondly, iemg of each muscle was calculated for the periods of stance and swing. Thirdly, ratios of BF/VM iemg activity were calculated for each halves of stance and swing. A statistical difference was determined by using the Student's paired t-test. Results were judged to be statistically significant at p<0.05. Results Mean speed of gait of all subjects was 74.96m/min and the mean cadence was 106.04 steps/min. Significant differences in time-distance parameters between the amputated and sound leg were found for most parameters (Table 1). In the amputated leg, step time, swing time, and step length were longer while stance time and single support time were shorter. A significant difference was found between knee angle at heel strike since the knee of the amputated leg was flexed at 7.50±3.6 and the knee of the sound leg was flexed at 4.36±3.4. No significant differences were found between knee angle at loading response, toe off, and maximal swing peak (Table 2). ratios of BF/VM iemg activity during the first and second half of stance and swing time were calculated in each leg (Table 3). The BF/VM mean of ratios in the amputated leg, during the first half of stance phase (3.8±2.6), was significantly higher when compared to the BF/VM mean of ratios in the sound leg (2.0±1.2). This difference is due to a number of reasons. Firstly, the mean iemg activity is almost twice as much in the amputated leg. Secondly, the mean iemg activity of the VM is almost equal in both legs. During the second half of stance phase, the amputated leg mean iemg activities of both BF and VM, are more than twice as much, and therefore the BF/VM mean of ratios are similar in both legs. During swing phase, mean ratios of BF/VM iemg activity in both legs was similar as well. Discussion Lower limb activity during normal gait is highly symmetrical, with insignificant differences between the dominant and nondominant sides of the body (Hirokawa, 1989). Quality of TT amputee gait depends on many factors, which include a pain-free stump, optimally fitted socket, acceptable alignment of the prothesis, and a good physical condition of the amputee. The present study investigated the gait of traumatic TT amputees who used their prostheses daily without any limitation. The results indicate that parameters of stance time and single support time are significantly shorter in the amputated leg while parameters of swing Table 2. Mean values and SD of knee angle of the amputated and sound legs measured at selected gait events. Knee angle at: Amputated leg ( ) Sound leg ( ) Heel strike Loading response Toe off Max. swing flexion 7.50±3.6 14.55±5.0 50.27±6.8 62.27±6.9 4.36±3.4 13.00±3.9 48.42+6.0 58.94+5.7.035.155.137.122

Gait analysis of trans-tibial amputees 219 Table 3. Absolute mean values of iemg and mean of ratios of biceps femoris (BF)/vastus medialis (VM) activity in amputated and sound legs calculated for the two halves of stance time and swing time. Gait phase Amputated leg (BF/VM) Sound leg (BF/VM) iemg ratios iemg ratios Stance time 1st half 2nd half 26.85+17.3 8.45+5.2 8.65+6.1 15.60+15.4 3.8±2.6 1.1±1.2 15.19±12.1 9.75+8.4 3.75±2.3 6.13+4.1 2.0±1.2 0.7±0.3.042.197 Swing time 1st half 2nd half 9.20+9.5 6.75+3.9 13.26±9.8 4.84+2.8 1.8±2.0 2.9+2.7 8.13±6.4 6.44±2.8 9.94±5.2 5.63±3.7 1.3±0.9 2.5±1.6.156.148 time, step time and step lengths are significantly shorter in the sound leg. This short step length of the sound leg is most probably due to the exclusive use of the SACH foot in this study. Indeed, it has been shown (Snyder et al., 1995) that the step length of the sound leg can be significantly improved by providing the amputee with a prosthetic foot design different from the SACH foot. Indeed, stride length increased from 1.25+0.16m with use of the SACH foot to 1.35+0.19m when the Flex-Foot was used (Lemaire et al., 1993). Also, walking with a Flex-Foot significantly improved the range of dorsiflexion at the ankle joint (14.3+3.8 with SACH foot, 21.8+3.6 with Flex-Foot). Whenever the sound leg swings to take a step, the rigid SACH foot cannot allow a dorsiflexion moment and the sound leg is lowered earlier. As a result, stance time is shorter in the amputated leg while swing time and step length are shorter in the sound leg. It has therefore been assumed that energy-storing feet (such as Flex-Foot) improve, to a certain extent, the lack of ankle moment generator encountered when using the rigid SACH foot. In TT amputees, the prosthesis socket is routinely positioned in a 5 flexion which is said to improve patellar tendon weight bearing. This explains significantly higher flexion in the amputated leg (7.5±3.6 ) in comparison with the sound leg (4.3±3.4 ), when measured at heel strike. All other knee angle measurements between the two legs showed no significant differences in this study. It is well known that the strength of the thigh muscles in TT amputees decreases with time (Renstrom et al., 1983). Indeed, strengthening of the quadriceps and hamstrings was shown to improve gait quality in these amputees, as stabilisation of the knee during stance time is achieved by co-contraction of these muscles (Winter and Sienko, 1988). The EMG data presented in this study indicate that during the first half of stance, the hamstrings iemg activity in the amputated leg is more than three times higher than the quadriceps iemg activity (hamstrings/quadriceps mean of ratios 3.8). The sound leg also showed a higher hamstrings iemg activity but less than twice that of the quadriceps iemg activity (hamstrings/quadriceps mean of ratios 2.0). The differences between the obtained mean of ratios were found to be significant. These findings are in agreement with other researchers (Winter and Sienko, 1988) who also monitored hyperactivity of the hip extensors in early and mid-stance but without absolute values or ratios of EMG activity. These researchers explained that the above normal activity of the hamstrings resulted in an above normal knee flexor moment. But, during gait of TT amputees, the knee flexor moment is counterbalanced by the cocontracting quadriceps. Unlike in stance, swing phase iemg activity and hamstrings/quadriceps means of ratios in TT amputees were similar in both legs. In conclusion, the present study displays the time-distance parameters and the activity of the

220 E. Isakov, O Keren and N. Benjuya muscles controlling the knees of the amputated and the sound leg. Differences between the indicated gait parameters and muscle activity can serve as references in future studies directed to evaluate the influence of new prosthetic components on gait symmetry in TT amputees. ISAKOV E, MIZRAHI J, RING H, SUSAK Z, HAKIM N (1992). Standing sway and weight-bearing distribution in people with below-knee amputations. Arch Phys Med Rehabil 73, 174-178. ISAKOV E, BURGER H, GREGORIC M, MARINCEK C (1996). Isokinetic and isometric strength of the thigh muscles in below-knee amputees. Clin Biomech 11, 233-235. REFERENCES BASMAJIAN JV, DELUCA C (1985). Muscle alive: function revealed by electromyography. Baltimore, MD: Williams & Wilkins. BAGLEY AM. SKINNER HB (1991). Progress in gait analysis in amputees: a special review, Crit Rev Phys Rehabil Med 3(2), 101-120. CRAIK RL, OATIS CA (1995). Gait analysis: theory and application. - St Louis, MI: Mosby Year Book. GEURTS ACH, MULDER TW (1992). Reorganisation of postural control following lower limb amputation: theoretical considerations and implications for rehabilitation. Physiotherapy Theory Practice 8, 145-157. HIROKAWA S (1989). Normal gait characteristics under temporal and distance constrains. J Biomed Eng 11, 449-456. ISAKOV E, BURGER H, KRAJNIK J, GREGORIC M, MARINCEK C (1997). Double-limb support time and step-length differences in below-knee amputees. Scand J Rehabil Med 29, 75-79. LEMAIRE ED, FISHER FR, ROBERTSON DGE (1993). Gait patterns of elderly men with trans-tibial amputations. Prosthet Orthot Int 17, 27-37. PERRY J (1992). Gait analysis: normal and pathological function. - Thorofare, NJ: Slaek. RENSTROM P, GRIMBY G, LARSSON E (1983). Thigh muscle strength in below-knee amputees. Scand J Rehabil Med Suppl No.9, 163-173. SNYDER RD, POWERS CM, FONTAINE C, PERRY J (1995). The effect of five prosthetic feet on the gait and loading of the sound limb in dysvascular below-knee amputees. J Rehabil Res Dev 32, 309-315. WINTER D, SIENKO SE (1988). Biomechanics of belowknee amputee gait. J Biomech 21, 361-367. ISPO-BRAZIL The Society is pleased to announce the formation of a new National Member Society in Brazil. Following is a list of its officers: Chairman: Mario Cesar Alves de Carvalho Honorary Treasurer: Ana Claudia A.C. Freitas Vice-Chairman: Dra. Eliane Machado de Araujo Secretary: Elgson Dimas Ribeiro, Jnr. Address: ISPO-Brazil Rua Dom Francisco Aquiro Correa 225 13075-080 Campinas Sao Paulo BRAZIL