- Email: [email protected]
Energy cost of walking measurements in subjects with lower limb amputations: A comparison study between floor and treadmill test
Gait & Posture 27 (2008) 70–75 www.elsevier.com/locate/gaitpost
Energy cost of walking measurements in subjects with lower limb amputations: A comparison study between floor and treadmill test Marco Traballesi *, Paolo Porcacchia, Tiziano Averna, Stefano Brunelli Unita` Operativa D, Fondazione Santa Lucia, Scientific Institute for Research, Hospitalization and Healthcare, Via Ardeatina 306, 00179 Roma, Italy Received 4 July 2006; received in revised form 23 January 2007; accepted 27 January 2007
Abstract Measuring the energy cost of walking (ECW) is a valid way of assessing the walking efficiency of subjects who were prosthetic users following lower limb amputation. The aim of this study was to determine whether, in these subjects, treadmill and floor ECW measurements are comparable. We tested 24 subjects who had undergone unilateral lower limb amputations for vascular diseases as they walked at a self-selected comfortable speed on the floor and on a treadmill. The tests were conducted at the end of rehabilitative treatment to fit prosthesis. Eight subjects underwent transtibial and 16 transfemoral amputation. The measurements were taken with a portable gas analyzer. The self-selected comfortable speed on the treadmill was significantly lower than that on the floor, where the patients adopted the aid they normally used for walking; oxygen consumption was the same in the two tests. Therefore, for both transtibial and transfemoral patients, ECW was greater during walking on the treadmill. Steady-state heart rate did not differ in the two tests. The data show that the ECW values of the amputated subjects obtained on the treadmill at the end of rehabilitation did not correspond with those they obtained on the floor. The floor test is the one that may better reflect walking with prostheses and aids in everyday life, in subjects with dysvascular lower limb amputation, using the prosthesis for a short time. # 2007 Elsevier B.V. All rights reserved. Keywords: Amputees; Walking; Oxygen consumption; Human locomotion
1. Introduction Energy cost measurement, which is a functional evaluation method adopted for studying the physiology of physical exercise, is used in rehabilitation to determine the effect of disability on walking [1]. In the literature there are several reports on energy cost of walking (ECW) measurement for functional evaluation of subjects with unilateral amputation and who use prostheses [2–10]. Measuring ECW in the amputated subject is an established method for quantifying the actual effort exerted [2,3,6–8,11–13] and for comparing the effectiveness of different prosthetic devices [4,5,9,10,14–16]. This measure* Corresponding author. Tel.: +39 0651501840; fax: +39 0651501919. E-mail addresses: [email protected] (M. Traballesi), [email protected] (P. Porcacchia), [email protected] (T. Averna), [email protected] (S. Brunelli). 0966-6362/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2007.01.006
ment has significant clinical importance, since ECW affects the subject’s ability to use the prosthesis and acquire the necessary motor abilities, thus influencing level of autonomy and quality of life. Reports in the literature show that ECW is greater in the amputated individuals compared to healthy controls [1,6,7] and increases with a higher level of amputation [2,6]. Transtibial and transfemoral vascular disease amputees have higher ECW than traumatic amputatees [2,7,12,13]. In previous studies, measurements were taken on the treadmill [5,10,12,14–18], or during overground walking [7–9,11,19], as in field tests or methods were used in the same study [14]. Patients who use aids to walk on flat surface use the equipment supports on the treadmill instead, because of difficulty keeping their balance. They also acquire adequate levels of coordination and specific motor abilities for walking on the treadmill only after consistent training. In the literature
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
the studies that compared the energy cost of walking on the floor and on the treadmill were conducted with nonamputated subjects [20]. We are not aware of any previous studies on this topic conducted in lower limb amputatees. Therefore, the aim of this study was to verify whether ECW tests on a treadmill and during free walking are really equivalent, or whether there are measurement differences in patients with lower limb amputations who use prostheses. 2. Materials and methods 2.1. Subjects The subjects recruited for the study had to meet the following inclusion criteria: (i) unilateral amputation of a lower limb for peripheral vascular disease due to diabetes or atherosclerosis; (ii) no pathological stump conditions that could impede prosthesis fitting (open surgical wound, ulcers or infections); (iii) no functional impairment of the sound limb; (iv) no cognitive disorders or other significant medical conditions. We recruited 24 consecutive patients admitted to our rehabilitation unit who were referred from surgical wards. Eight patients had undergone transtibial amputation (3 related to diabetes, 5 to atherosclerosis) and 16 transfemoral amputation (8 related to diabetes, 8 to atherosclerosis). Table 1 reports the characteristics of the population examined. 2.2. Rehabilitation treatment A physician led the rehabilitation team, which developed a program based primarily on practical locomotor skills with prostheses. Subjects underwent physiotherapy for 180 min once a day, 5 days a week, for 2 months. All above-knee amputees used a modular prosthesis with quad socket, polycentric knee joint and S.A.C.H. foot. Below-knee amputees used a modular patellar tendon bearing hard socket and energy storing foot. When the patients were able to walk independently on the floor with an aid they also began to exercise on the treadmill. The subjects walked on the device at a comfortable speed, progressively increasing the length of the sessions until they were able to walk on it continuously for at least 15 min. Training on the treadmill was carried out without aids, using the available supports and patients were trained to adjust velocity by themselves. 2.3. Measurement of energy cost of walking For all patients ECW measurement was obtained at the end of rehabilitation. In the same testing session each patient was Table 1 Subjects’ characteristics
TT TF
No. of subjects
M/W
Age
Height (cm)
Weighta (kg)
8 16
6/2 11/5
56 17 61 11
170 13 169 7
77 18.9 65 11.6
TT, transtibial amputation; TF, transfemoral amputation; M, man; W, woman. a Weight was measured on the day of the test, without the prosthesis.
71
Fig. 1. Breath-by-breath oxygen consumption (VO2) and carbon dioxide production (VCO2) during floor test. The patients were asked to walk at their own self-selected comfortable speed. The test lasted at least 7 min to enable them to reach and maintain steady-state (SS) conditions. Data obtained during the last 2 min were assumed to be the steady state and mean value of these 2 min was used. Verification that the patients had reached steady state occurred in real time by means of data acquisition and graphic visualization, in telemetry (this figure is example for a single subject).
measured first on the floor and then on the treadmill, with a rest interval that lasted as long as necessary to return to basal heart rate and oxygen consumption. 2.3.1. Floor measurement The test was conducted indoors, in a hallway with a regular floor surface. We chose a 61 m rectilinear course the subjects had to walk back and forth on. The patients were asked to walk at their own selfselected comfortable speed, with the aid they normally used. The test lasted at least 7 min to enable them to reach and maintain steady-state conditions. Data obtained during the last 2 min were assumed to be the steady state and mean value of those was used [21] (Fig. 1). Steady state was defined as a condition in which, after some minutes of exercise at a constant and sub-maximal workload, the rate of oxygen consumption reaches a level sufficient to meet energy demands. In this condition oxygen consumption and other physiological parameters (cardiac output, heart and respiratory rate) maintain a plateau. Verification that the patients had reached steady state occurred in real time by means of data acquisition and graphic visualization, in telemetry (Fig. 1). 2.3.2. Treadmill measurement A Technogym treadmill (RUNRACE model), adapted for rehabilitation needs (the lowest speed was .1 m/s), was used. The test was carried out without aids and with a treadmill inclination of 08. The subject, with upper limbs forward, grasped the bar at the height of the treadmill display and could easily push the button to change speed. The speed indicator was covered and patients chose their own comfortable speed without knowing the speed indicated on the treadmill. Once the speed was chosen, the test lasted for at least 7 min. The parameters recorded on both tests were the following: walking speed, steady-state oxygen consumption, resting heart rate and, finally, steady-state respiratory exchange ratio (carbon dioxide production/oxygen consumption). Oxygen consumption was expressed in milliliters per minute per kilogram of body weight (ml/min kg), and speed was expressed in meters per second (m/s). Energy cost (ECW), expressed in milliliters of oxygen consumed per meter walked for kilogram of body weight (ml/m kg), was calculated using the formula: ‘‘oxygen consumption/speed’’. Breath-by-breath gas exchange was measured using the portable system K4b2 (COSMED) [22], and heart rate using a POLAR heart rate monitor applied to the patient’s thorax.
72
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
2.4. Statistical analysis To verify normal distribution of the data, Kolmogorov–Smirnov test of normality was used. Values were expressed as means stanstandard deviation. Sample median values were calculated too for variables without normal distribution (comfortable walking speed, steady-state heart rate, steady-state respiratory exchange ratio). To compare each patient’s floor and treadmill values, Student’s t-test for paired data (for normal variables) and Wilcoxon signed ranks test (for not normal variables) were performed. A probability value of .05 was set as the level of statistical significance. A regression line equation and the Bravais–Pearson correlation coefficient were calculated to compare energy cost on the treadmill and on the floor. All subjects gave their written informed consent to participate in the study, according to the guidelines established by the local ethics committee, which approved this study.
Fig. 2. Correlation between energy cost of walking (ECW) during treadmill and floor tests. Regression analysis shows that treadmill ECW is approximately 2.5 times floor ECW. Regression line equation: y = treadmill ECW; x = floor ECW; r, Bravais–Pearson correlation coefficient.
equation shows that ECW on the treadmill was about 2.5 times greater than on the floor. The correlation coefficient (r) was equal to .74.
3. Results At the end of the rehabilitative training, lasting 2 months, the patients with transtibial amputation walked with the following aids: one patient did not use aids, one walked with a cane, one with one crutch, four with two crutches and one patient used a walker. Of the patients with transfemoral amputation, two did not use aids, one used a cane, three used one crutch, nine used two crutches and one used a walker. As shown in Table 2, for all subjects the comfortable speed chosen was on average .22 m/s on the treadmill (median value .17), i.e., significantly less ( p < .001) than the .52 m/s chosen on the floor (median value .42). Steadystate oxygen consumption was similar in the two tests: 12.8 ml/min kg on the treadmill and 13.3 ml/min kg on the floor; therefore, the energy cost of locomotion was significantly greater when measured on the treadmill than on the floor. In fact, a cost of 1.26 ml/m kg was calculated on the treadmill and a cost of .49 ml/m kg on the floor ( p < .001). Steady-state heart rate (HR) was the same in the two trials as was steady-state respiratory exchange ratio (R). All subjects reached steady-state conditions in 90–140 s. After the firs test, the subjects took a mean of 12 min to restore basal conditions (in terms of heart rate, oxygen consumption and subject-perceived sensations). Rest time interval between floor and treadmill tests lasted at least 20 min. Fig. 2 presents the energy cost values of walking on the treadmill compared to those on the floor. The regression line
3.1. Amputation level In Table 3 the subjects were separated on the basis of amputation level, i.e., transtibial (TT) and transfemoral (TF). Statistical analysis produced results similar to those found for the entire group of subjects. The self-selected comfortable speed was significantly less on the treadmill than on the floor for both TT and TF ( p < .001). However, steady-state oxygen consumption in the two groups was not significantly different on the treadmill and the floor. ECW was higher for both TT ( p < .05) and TF ( p < .001) when measured on the treadmill rather than on the floor. Fig. 3 shows each subject’s speed in the two tests. Subjects were classified on the basis of aids used to walk on the floor and their speed. The ratio between speed on the floor and speed on the treadmill was 2.5 for the subjects who did not use aids, 2.7 for those who used one aid, 2.6 for those who used two aids and 2.9 for those who used a walker.
4. Discussion The goal of rehabilitative treatment in subjects with lower limb amputation is efficient locomotion with prostheses, and ECW measurement is a useful instrument for evaluating walking efficiency.
Table 2 Functional measures during floor and treadmill tests
F T
HR rest (bpm)
V (m/s)
SS oxygen consumption (ml/min kg)
ECW (ml/m kg)
SS HR (bpm)
SS R
84 11 86* 13
.52 .22 .22** .14
13.3 2.8 12.8 3.1
.49 .18 1.26** .63
110 18 108 19
.82 .11 .79 .05
F, floor measure; T, treadmill measure; HR, heart rate, in beats per minute (bpm); V, velocity; SS, steady state; ECW, energy cost of walking; R, respiratory exchange ratio. Median values for variables without normal distribution: floor V = .42 m/s; treadmill V = .17 m/s; floor SS HR = 105 bpm; treadmill SS HR = 103; floor R = .8; treadmill R = .79. *p < .05; **p < .001 (statistical analysis matches treadmill and floor measures).
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
73
Table 3 Functional measures during floor and treadmill tests based on amputation level HR rest (bpm) TT
F T
TF
F T
84 17 85 20 85 8 87* 8
V (m/s)
SS oxygen consumption (ml/min kg)
.66 .26 .29** .18
13.5 2.4 12.3 2.5
.45 .17 .19** .11
13.2 3.1 13 3.5
ECW (ml/m kg)
SS HR (bpm)
SS R
.40 .17 1.00* .68
110 27 106 28
.80 .04 .79 .03
.54 .17 1.40** .59
110 13 108 14
.82 .13 .80 .06
TT, transtibial amputation; TF, transfemoral amputation; F, floor measure; T, treadmill measure; HR, heart rate, in beats per minute (bpm); V, velocity; SS, steady state; ECW, energy cost of walking; R, respiratory exchange ratio. *p < .05; **p < .001 (statistical analysis matches treadmill and floor measures).
In this study, we aimed to determine whether, in subjects with unilateral amputation of a lower limb, ECW would be similar when measured during walking on the floor or a treadmill. On both the floor and the treadmill all patients walked at their own self-selected comfortable speed. It emerged that heart rate and steady-state oxygen consumption values were the same in the two tests, indicating that they had the same internal load. Heart rate, oxygen consumption and respiratory exchange ratio values indicated a sub-maximal degree of intensity of effort, and these parameters are representative of the motor task requested, indicating that neither the first nor the second test induced significant fatigue. In fact, during the rehabilitation treatment, subjects walked at their comfortable self-selected speed and walking time was progressively increased. Therefore they were able to walk, even on treadmill, from 15 min to 1 h continuously. This can explain how a 7 min test did not represent a difficult effort for them. However, the comfortable speed chosen on the treadmill was significantly different from that on the floor, i.e., on the treadmill the comfortable speed was much slower. Differences in walking speed, between floor and treadmill, can be explained as follows. Subjects did not use on the treadmill their usual aids; their posture on the device was not their usual and they required greater coordination. Even though our subjects were capable of walking on the treadmill, we cannot exclude that fear of falling can lead to a more cautious and slower walk. Higher ECW was obtained for measurements on the treadmill independently of the level of amputation or type of
aid used on the floor. As can be observed in Fig. 2, walking on the treadmill was about 2.5 times more costly for all subjects (as energy spent per meter covered) than walking on the floor. Further, the value of the linear correlation coefficient, i.e., .74, indicates a certain degree of dispersion of the values around the line, showing that each subject’s behavior was different; therefore, ECW on the treadmill cannot be predicted based on ECW on the floor and vice versa. A limitation of this study, in comparing floor and treadmill test, is the lack of randomization. With regard to level of amputation and in agreement with the data reported in the literature [1–3,6,7], the subjects with TF amputation showed slower walking speed and greater energy cost than the subjects with TT amputations. Both groups presented the same differences between measurements taken on the treadmill and on the floor: on the treadmill there was a reduction of the comfortable speed and an increase in ECW. With regard to the aids used, speed was always higher on the floor than on the treadmill. This was also the case in the various subgroups reported in Fig. 3 (0, 1 or 2 aids, walker). Casillas et al. [14] studied a group of 12 patients with transtibial amputations for vascular disease with a mean age of 73 years. These subjects did not perform the test on the treadmill because it was judged to be too difficult for them. Their findings walking on a flat surface at a self-selected comfortable speed were similar to ours. The results of study, compared with those in the literature that studied ECW on the treadmill and overground walking separately, showed slower self-selected speed. Therefore
Fig. 3. TT, transtibial amputation; TF, transfemoral amputation. Walking aids: (0) no walking aids; (1) one cane or one crutch; (2) two crutches; (3) walker. All subjects were ordered on the basis of the aid they used for floor walking and then by floor speed, in decreasing order. Floor/treadmill mean speed ratio is 2.5 for subjects using no walking aids, 2.7 for subjects using one aid, 2.6 for subjects using two aids, 2.9 for subjects using a walker.
74
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
ECW values higher [1,2,7,9,11,18], even when the internal load was the same [7,9]. The slow walking speed can be explained when considering the time spent using the prosthesis, the cause of amputation, the age of our subjects and the walking aids used. Our patients had been using prostheses for only a short period of time (on average, 2 months) compared to other studies, where the minimum period of use was 6 months [2,11]. Subjects in previous studies did not use walking aids [11], while the majority of our subjects did. All of our patients underwent amputations for peripheral vascular disease and their mean age was at the upper limits of those reported in the literature. This can affect physical fitness, walking speed and ECW. It is meaningful to measure ECW in the patients’ usual walking conditions and, in the older prosthetic users, with their aids, when appropriate. In a 5-year follow-up study by Gauthier-Gagnon et al. [23], 85% of the TFs used aids in outdoor locomotion. This supports our view that assessment of ambulation with aids is important. For all the above reasons, it may not be appropriate to extrapolate our results to a younger and fitter population. Previous studies compared treadmill and floor walking in non-amputee subjects and found a trend similar to our study’s. Greig et al. [24] found that, in healthy elderly people, heart rate response to treadmill walking was less representative of the ‘‘real life’’ situation than in younger adults. Swerts et al. [25] reported that, in patients with chronic obstructive pulmonary disease, corridor walking was more efficient than treadmill walking. Finally, dynamic and biomechanical studies [26,27] demonstrated significant differences in step length and ground reaction forces when comparing treadmill and floor walking. A study by Pearce et al. [20] is the only one in the literature that compares ECW in treadmill and overground walking. The energy cost of locomotion in normal subjects was measured on a flat surface and on a treadmill to obtain equations which would predict oxygen consumption based on speed. Their results were opposite to ours: both at comfortable and at higher speed normal subjects showed higher oxygen consumption on the floor compared to the treadmill. However, the self-selected comfortable speed of the normal subjects was much higher than that of the subjects with lower limb amputations. For this reason and because of the different step patterns [28,29], we believe that the equations obtained from normal subjects should not be used for subjects with lower limb amputation.
5. Conclusion Amputees tested on treadmill demonstrated a significantly slower self-selected comfortable speed compared to overground walking with similar oxygen consumption. This indicates an increase in ECW with treadmill walking.
Therefore, treadmill testing results in lower limb unilateral amputees may not represent walking efficiency in a reliable way. The treadmill test overestimates ECW in subjects with unilateral lower limb amputations above or below the knee who have only been using the prosthesis for a short period of time. Energy cost measurement in these subjects should, therefore, be carried out on the ground and with the aids used in daily life.
Conflict of interest statement No support in the form of grants, equipment, drugs or any of these was drawn on. No commercial entity with a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated.
References [1] Waters RL, Mulroy S. The energy expenditure of normal and pathologic gait. Gait Posture 1999;9:207–31. [2] Waters RL, Perry J, Antonelli D, Hislop H. Energy cost of walking of amputees: the influence of level of amputation. J Bone Joint Surg 1976;58A:42–6. [3] Traugh GH, Corcoran PF, Reyes RL. Energy expenditure of ambulation in patients with above-knee amputation. Arch Phys Med Rehabil 1975;56:67–71. [4] Boonstra AM, Schrama J, Fidler V, Eisma WH. Energy cost during ambulation in transfemoral amputees: a knee joint with mechanical swing phase control vs a knee joint with a pneumatic swing phase control. Scand J Rehabil Med 1995;27(2):77–81. [5] Datta D, Heller B, Howitt J. A comparative evaluation of oxygen consumption and gait pattern in amputees using intelligent prostheses and conventionally damped knee swing-phase control. Clin Rehabil 2005;19(4):398–403. [6] Gonzalez EG, Corcoran PJ, Reyes RL. Energy expenditure in belowknee amputees: correlation with stump length. Arch Phys Med Rehabil 1974;55:111–9. [7] Gailey RS, Wenger MA, Raya M, Kirk N, Erbs K, Spyropoulos P, et al. Energy expenditure of trans-tibial amputees during ambulation at selfselected pace. Prosthet Orthot Int 1994;18(2):84–91. [8] Pagliarulo MA, Waters R, Hislop HJ. Energy cost of walking of belowknee amputees having no vascular disease. Phys Ther 1979;59(5): 538–42. [9] Torburn L, Powers CM, Guiterrez R, Perry J. Energy expenditure during ambulation in dysvascular and traumatic below-knee amputees: a comparison of five prosthetic feet. J Rehabil Res Dev 1995;32(2): 111–9. [10] Schmalz T, Blumentritt S, Jarasch R. Energy expenditure and biomechanical characteristics of lower limb amputee gait: the influence of prosthetic alignment and different prosthetic components. Gait Posture 2002;16(3):255–63. [11] Nowrozzi F, Salvanelli ML. Energy expenditure in hip disarticulation and hemipelvectomy amputees. Arch Phys Med Rehabil 1983;64: 300–3. [12] Pinzur MS, Gold J, Schwartz D, Gross N. Energy demands for walking in dysvascular amputees as related to the level of amputation. Orthopedics 1992;15(9):1033–6 [discussion 1036–7].
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75 [13] Jaegers S, Vos L, Rispens P. The relationship between comfortable and most metabolically efficient walking speed in persons with unilateral above-knee amputation. Arch Phys Med Rehabil 1993;75:521–5. [14] Casillas JM, Dulieu V, Cohen M, Marcer I, Didier JP. Bioenergetic comparison of a new energy-storing foot and SACH foot in traumatic below-knee vascular amputations. Arch Phys Med Rehabil 1995;76(1): 39–44. [15] Huang GF, Chou YL, Su FC. Gait analysis and energy consumption of below-knee amputees wearing three different prosthetic feet. Gait Posture 2000;12(2):162–8 [Huang GF. Retraction. Gait Posture 2003;18(3):124]. [16] James KB, Stein RB. Improved ankle-foot system for above-knee amputees. Am J Phys Med 1986;65(6):301–14. [17] Gailey RS, Nash MS, Atchley TA, Zilmer RM, Moline-Little GR, Morris-Cresswell N, et al. The effects of prosthesis mass on metabolic cost of ambulation in non-vascular trans-tibial amputees. Prosthet Orthot Int 1997;21(1):9–16. [18] Taylor MB, Clark E, Offord EA, Baxter C. A comparison of energy expenditure by a high level trans-femoral amputee using the Intelligent Prosthesis and conventionally damped prosthetic limbs. Prosthet Orthot Int 1996;20(2):116–21. [19] Czerniecki JM, Gitter A, Weaver K. Effect of alterations in prosthetic shank mass on the metabolic costs of ambulation in above-knee amputees. Am J Phys Med Rehabil 1994;73(5):348–52. [20] Pearce ME, Cunningham DA, Donner AP, Rechnitzer PA, Fullerton GM, Howard JH. Energy cost of treadmill and floor walking at selfselected paces. Eur J Appl Physiol 1983;52:115–9.
75
[21] Bernardi M, Macaluso A, Sproviero E, Castellano V, Coratella D, Felici F, et al. Cost of walking and locomotor impairment. J Electromyogr Kinesiol 1999;9:149–57. [22] Fatone S, Boyd R, Rodda J, Gallagher D, Cullis E, Nattrass G, et al. Evaluation of the Cosmed K4 in a clinical gait analysis laboratory. Gait Posture 1997;5(1):79–179. [23] Gauthier-Gagnon C, Grise´ MC, Potvin D. Enabling factors related to prosthetic use by people with transtibial and transfemoral amputation. Arch Phys Med Rehabil 1999;80:706–13. [24] Greig C, Butler F, Skelton D, Mahmud S, Young A. Treadmill walking in old age may not reproduce the real life situation. J Am Geriatr Soc 1993;41(1):15–8. [25] Swerts PM, Mostert R, Wouters EF. Comparison of corridor and treadmill walking in patients with severe chronic obstructive pulmonary disease. Phys Ther 1990;70(7):439–42. [26] Vogt L, Banzer W, Bayer I, Schmidtbleicher D, Kerschbaumer F. Overground and walkway ambulation with unilateral hip osteoarthritis: comparison of step length asymmetries and reproducibility of treadmill mounted force plate readings. Physiother Theory Pract 2006;22(2):73–82. [27] White SC, Yack HJ, Tucker CA, Lin HY. Comparison of vertical ground reaction forces during overground and treadmill walking. Med Sci Sports Exerc 1998;30(10):1537–42. [28] Saunders JB, Inman VT, Eberhart HD. The major determinants in normal and pathological gait. J Bone Joint Surg 1953;35A(3):543–57. [29] Fisher SV, Gullickson Jr G. Energy cost of ambulation in health and disability: a literature review. Arch Phys Med Rehabil 1978;59(3): 124–33.
Energy cost of walking measurements in subjects with lower limb amputations: A comparison study between floor and treadmill test Marco Traballesi *, Paolo Porcacchia, Tiziano Averna, Stefano Brunelli Unita` Operativa D, Fondazione Santa Lucia, Scientific Institute for Research, Hospitalization and Healthcare, Via Ardeatina 306, 00179 Roma, Italy Received 4 July 2006; received in revised form 23 January 2007; accepted 27 January 2007
Abstract Measuring the energy cost of walking (ECW) is a valid way of assessing the walking efficiency of subjects who were prosthetic users following lower limb amputation. The aim of this study was to determine whether, in these subjects, treadmill and floor ECW measurements are comparable. We tested 24 subjects who had undergone unilateral lower limb amputations for vascular diseases as they walked at a self-selected comfortable speed on the floor and on a treadmill. The tests were conducted at the end of rehabilitative treatment to fit prosthesis. Eight subjects underwent transtibial and 16 transfemoral amputation. The measurements were taken with a portable gas analyzer. The self-selected comfortable speed on the treadmill was significantly lower than that on the floor, where the patients adopted the aid they normally used for walking; oxygen consumption was the same in the two tests. Therefore, for both transtibial and transfemoral patients, ECW was greater during walking on the treadmill. Steady-state heart rate did not differ in the two tests. The data show that the ECW values of the amputated subjects obtained on the treadmill at the end of rehabilitation did not correspond with those they obtained on the floor. The floor test is the one that may better reflect walking with prostheses and aids in everyday life, in subjects with dysvascular lower limb amputation, using the prosthesis for a short time. # 2007 Elsevier B.V. All rights reserved. Keywords: Amputees; Walking; Oxygen consumption; Human locomotion
1. Introduction Energy cost measurement, which is a functional evaluation method adopted for studying the physiology of physical exercise, is used in rehabilitation to determine the effect of disability on walking [1]. In the literature there are several reports on energy cost of walking (ECW) measurement for functional evaluation of subjects with unilateral amputation and who use prostheses [2–10]. Measuring ECW in the amputated subject is an established method for quantifying the actual effort exerted [2,3,6–8,11–13] and for comparing the effectiveness of different prosthetic devices [4,5,9,10,14–16]. This measure* Corresponding author. Tel.: +39 0651501840; fax: +39 0651501919. E-mail addresses: [email protected] (M. Traballesi), [email protected] (P. Porcacchia), [email protected] (T. Averna), [email protected] (S. Brunelli). 0966-6362/$ – see front matter # 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.gaitpost.2007.01.006
ment has significant clinical importance, since ECW affects the subject’s ability to use the prosthesis and acquire the necessary motor abilities, thus influencing level of autonomy and quality of life. Reports in the literature show that ECW is greater in the amputated individuals compared to healthy controls [1,6,7] and increases with a higher level of amputation [2,6]. Transtibial and transfemoral vascular disease amputees have higher ECW than traumatic amputatees [2,7,12,13]. In previous studies, measurements were taken on the treadmill [5,10,12,14–18], or during overground walking [7–9,11,19], as in field tests or methods were used in the same study [14]. Patients who use aids to walk on flat surface use the equipment supports on the treadmill instead, because of difficulty keeping their balance. They also acquire adequate levels of coordination and specific motor abilities for walking on the treadmill only after consistent training. In the literature
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
the studies that compared the energy cost of walking on the floor and on the treadmill were conducted with nonamputated subjects [20]. We are not aware of any previous studies on this topic conducted in lower limb amputatees. Therefore, the aim of this study was to verify whether ECW tests on a treadmill and during free walking are really equivalent, or whether there are measurement differences in patients with lower limb amputations who use prostheses. 2. Materials and methods 2.1. Subjects The subjects recruited for the study had to meet the following inclusion criteria: (i) unilateral amputation of a lower limb for peripheral vascular disease due to diabetes or atherosclerosis; (ii) no pathological stump conditions that could impede prosthesis fitting (open surgical wound, ulcers or infections); (iii) no functional impairment of the sound limb; (iv) no cognitive disorders or other significant medical conditions. We recruited 24 consecutive patients admitted to our rehabilitation unit who were referred from surgical wards. Eight patients had undergone transtibial amputation (3 related to diabetes, 5 to atherosclerosis) and 16 transfemoral amputation (8 related to diabetes, 8 to atherosclerosis). Table 1 reports the characteristics of the population examined. 2.2. Rehabilitation treatment A physician led the rehabilitation team, which developed a program based primarily on practical locomotor skills with prostheses. Subjects underwent physiotherapy for 180 min once a day, 5 days a week, for 2 months. All above-knee amputees used a modular prosthesis with quad socket, polycentric knee joint and S.A.C.H. foot. Below-knee amputees used a modular patellar tendon bearing hard socket and energy storing foot. When the patients were able to walk independently on the floor with an aid they also began to exercise on the treadmill. The subjects walked on the device at a comfortable speed, progressively increasing the length of the sessions until they were able to walk on it continuously for at least 15 min. Training on the treadmill was carried out without aids, using the available supports and patients were trained to adjust velocity by themselves. 2.3. Measurement of energy cost of walking For all patients ECW measurement was obtained at the end of rehabilitation. In the same testing session each patient was Table 1 Subjects’ characteristics
TT TF
No. of subjects
M/W
Age
Height (cm)
Weighta (kg)
8 16
6/2 11/5
56 17 61 11
170 13 169 7
77 18.9 65 11.6
TT, transtibial amputation; TF, transfemoral amputation; M, man; W, woman. a Weight was measured on the day of the test, without the prosthesis.
71
Fig. 1. Breath-by-breath oxygen consumption (VO2) and carbon dioxide production (VCO2) during floor test. The patients were asked to walk at their own self-selected comfortable speed. The test lasted at least 7 min to enable them to reach and maintain steady-state (SS) conditions. Data obtained during the last 2 min were assumed to be the steady state and mean value of these 2 min was used. Verification that the patients had reached steady state occurred in real time by means of data acquisition and graphic visualization, in telemetry (this figure is example for a single subject).
measured first on the floor and then on the treadmill, with a rest interval that lasted as long as necessary to return to basal heart rate and oxygen consumption. 2.3.1. Floor measurement The test was conducted indoors, in a hallway with a regular floor surface. We chose a 61 m rectilinear course the subjects had to walk back and forth on. The patients were asked to walk at their own selfselected comfortable speed, with the aid they normally used. The test lasted at least 7 min to enable them to reach and maintain steady-state conditions. Data obtained during the last 2 min were assumed to be the steady state and mean value of those was used [21] (Fig. 1). Steady state was defined as a condition in which, after some minutes of exercise at a constant and sub-maximal workload, the rate of oxygen consumption reaches a level sufficient to meet energy demands. In this condition oxygen consumption and other physiological parameters (cardiac output, heart and respiratory rate) maintain a plateau. Verification that the patients had reached steady state occurred in real time by means of data acquisition and graphic visualization, in telemetry (Fig. 1). 2.3.2. Treadmill measurement A Technogym treadmill (RUNRACE model), adapted for rehabilitation needs (the lowest speed was .1 m/s), was used. The test was carried out without aids and with a treadmill inclination of 08. The subject, with upper limbs forward, grasped the bar at the height of the treadmill display and could easily push the button to change speed. The speed indicator was covered and patients chose their own comfortable speed without knowing the speed indicated on the treadmill. Once the speed was chosen, the test lasted for at least 7 min. The parameters recorded on both tests were the following: walking speed, steady-state oxygen consumption, resting heart rate and, finally, steady-state respiratory exchange ratio (carbon dioxide production/oxygen consumption). Oxygen consumption was expressed in milliliters per minute per kilogram of body weight (ml/min kg), and speed was expressed in meters per second (m/s). Energy cost (ECW), expressed in milliliters of oxygen consumed per meter walked for kilogram of body weight (ml/m kg), was calculated using the formula: ‘‘oxygen consumption/speed’’. Breath-by-breath gas exchange was measured using the portable system K4b2 (COSMED) [22], and heart rate using a POLAR heart rate monitor applied to the patient’s thorax.
72
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
2.4. Statistical analysis To verify normal distribution of the data, Kolmogorov–Smirnov test of normality was used. Values were expressed as means stanstandard deviation. Sample median values were calculated too for variables without normal distribution (comfortable walking speed, steady-state heart rate, steady-state respiratory exchange ratio). To compare each patient’s floor and treadmill values, Student’s t-test for paired data (for normal variables) and Wilcoxon signed ranks test (for not normal variables) were performed. A probability value of .05 was set as the level of statistical significance. A regression line equation and the Bravais–Pearson correlation coefficient were calculated to compare energy cost on the treadmill and on the floor. All subjects gave their written informed consent to participate in the study, according to the guidelines established by the local ethics committee, which approved this study.
Fig. 2. Correlation between energy cost of walking (ECW) during treadmill and floor tests. Regression analysis shows that treadmill ECW is approximately 2.5 times floor ECW. Regression line equation: y = treadmill ECW; x = floor ECW; r, Bravais–Pearson correlation coefficient.
equation shows that ECW on the treadmill was about 2.5 times greater than on the floor. The correlation coefficient (r) was equal to .74.
3. Results At the end of the rehabilitative training, lasting 2 months, the patients with transtibial amputation walked with the following aids: one patient did not use aids, one walked with a cane, one with one crutch, four with two crutches and one patient used a walker. Of the patients with transfemoral amputation, two did not use aids, one used a cane, three used one crutch, nine used two crutches and one used a walker. As shown in Table 2, for all subjects the comfortable speed chosen was on average .22 m/s on the treadmill (median value .17), i.e., significantly less ( p < .001) than the .52 m/s chosen on the floor (median value .42). Steadystate oxygen consumption was similar in the two tests: 12.8 ml/min kg on the treadmill and 13.3 ml/min kg on the floor; therefore, the energy cost of locomotion was significantly greater when measured on the treadmill than on the floor. In fact, a cost of 1.26 ml/m kg was calculated on the treadmill and a cost of .49 ml/m kg on the floor ( p < .001). Steady-state heart rate (HR) was the same in the two trials as was steady-state respiratory exchange ratio (R). All subjects reached steady-state conditions in 90–140 s. After the firs test, the subjects took a mean of 12 min to restore basal conditions (in terms of heart rate, oxygen consumption and subject-perceived sensations). Rest time interval between floor and treadmill tests lasted at least 20 min. Fig. 2 presents the energy cost values of walking on the treadmill compared to those on the floor. The regression line
3.1. Amputation level In Table 3 the subjects were separated on the basis of amputation level, i.e., transtibial (TT) and transfemoral (TF). Statistical analysis produced results similar to those found for the entire group of subjects. The self-selected comfortable speed was significantly less on the treadmill than on the floor for both TT and TF ( p < .001). However, steady-state oxygen consumption in the two groups was not significantly different on the treadmill and the floor. ECW was higher for both TT ( p < .05) and TF ( p < .001) when measured on the treadmill rather than on the floor. Fig. 3 shows each subject’s speed in the two tests. Subjects were classified on the basis of aids used to walk on the floor and their speed. The ratio between speed on the floor and speed on the treadmill was 2.5 for the subjects who did not use aids, 2.7 for those who used one aid, 2.6 for those who used two aids and 2.9 for those who used a walker.
4. Discussion The goal of rehabilitative treatment in subjects with lower limb amputation is efficient locomotion with prostheses, and ECW measurement is a useful instrument for evaluating walking efficiency.
Table 2 Functional measures during floor and treadmill tests
F T
HR rest (bpm)
V (m/s)
SS oxygen consumption (ml/min kg)
ECW (ml/m kg)
SS HR (bpm)
SS R
84 11 86* 13
.52 .22 .22** .14
13.3 2.8 12.8 3.1
.49 .18 1.26** .63
110 18 108 19
.82 .11 .79 .05
F, floor measure; T, treadmill measure; HR, heart rate, in beats per minute (bpm); V, velocity; SS, steady state; ECW, energy cost of walking; R, respiratory exchange ratio. Median values for variables without normal distribution: floor V = .42 m/s; treadmill V = .17 m/s; floor SS HR = 105 bpm; treadmill SS HR = 103; floor R = .8; treadmill R = .79. *p < .05; **p < .001 (statistical analysis matches treadmill and floor measures).
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
73
Table 3 Functional measures during floor and treadmill tests based on amputation level HR rest (bpm) TT
F T
TF
F T
84 17 85 20 85 8 87* 8
V (m/s)
SS oxygen consumption (ml/min kg)
.66 .26 .29** .18
13.5 2.4 12.3 2.5
.45 .17 .19** .11
13.2 3.1 13 3.5
ECW (ml/m kg)
SS HR (bpm)
SS R
.40 .17 1.00* .68
110 27 106 28
.80 .04 .79 .03
.54 .17 1.40** .59
110 13 108 14
.82 .13 .80 .06
TT, transtibial amputation; TF, transfemoral amputation; F, floor measure; T, treadmill measure; HR, heart rate, in beats per minute (bpm); V, velocity; SS, steady state; ECW, energy cost of walking; R, respiratory exchange ratio. *p < .05; **p < .001 (statistical analysis matches treadmill and floor measures).
In this study, we aimed to determine whether, in subjects with unilateral amputation of a lower limb, ECW would be similar when measured during walking on the floor or a treadmill. On both the floor and the treadmill all patients walked at their own self-selected comfortable speed. It emerged that heart rate and steady-state oxygen consumption values were the same in the two tests, indicating that they had the same internal load. Heart rate, oxygen consumption and respiratory exchange ratio values indicated a sub-maximal degree of intensity of effort, and these parameters are representative of the motor task requested, indicating that neither the first nor the second test induced significant fatigue. In fact, during the rehabilitation treatment, subjects walked at their comfortable self-selected speed and walking time was progressively increased. Therefore they were able to walk, even on treadmill, from 15 min to 1 h continuously. This can explain how a 7 min test did not represent a difficult effort for them. However, the comfortable speed chosen on the treadmill was significantly different from that on the floor, i.e., on the treadmill the comfortable speed was much slower. Differences in walking speed, between floor and treadmill, can be explained as follows. Subjects did not use on the treadmill their usual aids; their posture on the device was not their usual and they required greater coordination. Even though our subjects were capable of walking on the treadmill, we cannot exclude that fear of falling can lead to a more cautious and slower walk. Higher ECW was obtained for measurements on the treadmill independently of the level of amputation or type of
aid used on the floor. As can be observed in Fig. 2, walking on the treadmill was about 2.5 times more costly for all subjects (as energy spent per meter covered) than walking on the floor. Further, the value of the linear correlation coefficient, i.e., .74, indicates a certain degree of dispersion of the values around the line, showing that each subject’s behavior was different; therefore, ECW on the treadmill cannot be predicted based on ECW on the floor and vice versa. A limitation of this study, in comparing floor and treadmill test, is the lack of randomization. With regard to level of amputation and in agreement with the data reported in the literature [1–3,6,7], the subjects with TF amputation showed slower walking speed and greater energy cost than the subjects with TT amputations. Both groups presented the same differences between measurements taken on the treadmill and on the floor: on the treadmill there was a reduction of the comfortable speed and an increase in ECW. With regard to the aids used, speed was always higher on the floor than on the treadmill. This was also the case in the various subgroups reported in Fig. 3 (0, 1 or 2 aids, walker). Casillas et al. [14] studied a group of 12 patients with transtibial amputations for vascular disease with a mean age of 73 years. These subjects did not perform the test on the treadmill because it was judged to be too difficult for them. Their findings walking on a flat surface at a self-selected comfortable speed were similar to ours. The results of study, compared with those in the literature that studied ECW on the treadmill and overground walking separately, showed slower self-selected speed. Therefore
Fig. 3. TT, transtibial amputation; TF, transfemoral amputation. Walking aids: (0) no walking aids; (1) one cane or one crutch; (2) two crutches; (3) walker. All subjects were ordered on the basis of the aid they used for floor walking and then by floor speed, in decreasing order. Floor/treadmill mean speed ratio is 2.5 for subjects using no walking aids, 2.7 for subjects using one aid, 2.6 for subjects using two aids, 2.9 for subjects using a walker.
74
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75
ECW values higher [1,2,7,9,11,18], even when the internal load was the same [7,9]. The slow walking speed can be explained when considering the time spent using the prosthesis, the cause of amputation, the age of our subjects and the walking aids used. Our patients had been using prostheses for only a short period of time (on average, 2 months) compared to other studies, where the minimum period of use was 6 months [2,11]. Subjects in previous studies did not use walking aids [11], while the majority of our subjects did. All of our patients underwent amputations for peripheral vascular disease and their mean age was at the upper limits of those reported in the literature. This can affect physical fitness, walking speed and ECW. It is meaningful to measure ECW in the patients’ usual walking conditions and, in the older prosthetic users, with their aids, when appropriate. In a 5-year follow-up study by Gauthier-Gagnon et al. [23], 85% of the TFs used aids in outdoor locomotion. This supports our view that assessment of ambulation with aids is important. For all the above reasons, it may not be appropriate to extrapolate our results to a younger and fitter population. Previous studies compared treadmill and floor walking in non-amputee subjects and found a trend similar to our study’s. Greig et al. [24] found that, in healthy elderly people, heart rate response to treadmill walking was less representative of the ‘‘real life’’ situation than in younger adults. Swerts et al. [25] reported that, in patients with chronic obstructive pulmonary disease, corridor walking was more efficient than treadmill walking. Finally, dynamic and biomechanical studies [26,27] demonstrated significant differences in step length and ground reaction forces when comparing treadmill and floor walking. A study by Pearce et al. [20] is the only one in the literature that compares ECW in treadmill and overground walking. The energy cost of locomotion in normal subjects was measured on a flat surface and on a treadmill to obtain equations which would predict oxygen consumption based on speed. Their results were opposite to ours: both at comfortable and at higher speed normal subjects showed higher oxygen consumption on the floor compared to the treadmill. However, the self-selected comfortable speed of the normal subjects was much higher than that of the subjects with lower limb amputations. For this reason and because of the different step patterns [28,29], we believe that the equations obtained from normal subjects should not be used for subjects with lower limb amputation.
5. Conclusion Amputees tested on treadmill demonstrated a significantly slower self-selected comfortable speed compared to overground walking with similar oxygen consumption. This indicates an increase in ECW with treadmill walking.
Therefore, treadmill testing results in lower limb unilateral amputees may not represent walking efficiency in a reliable way. The treadmill test overestimates ECW in subjects with unilateral lower limb amputations above or below the knee who have only been using the prosthesis for a short period of time. Energy cost measurement in these subjects should, therefore, be carried out on the ground and with the aids used in daily life.
Conflict of interest statement No support in the form of grants, equipment, drugs or any of these was drawn on. No commercial entity with a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the author(s) or upon any organization with which the author(s) is/are associated.
References [1] Waters RL, Mulroy S. The energy expenditure of normal and pathologic gait. Gait Posture 1999;9:207–31. [2] Waters RL, Perry J, Antonelli D, Hislop H. Energy cost of walking of amputees: the influence of level of amputation. J Bone Joint Surg 1976;58A:42–6. [3] Traugh GH, Corcoran PF, Reyes RL. Energy expenditure of ambulation in patients with above-knee amputation. Arch Phys Med Rehabil 1975;56:67–71. [4] Boonstra AM, Schrama J, Fidler V, Eisma WH. Energy cost during ambulation in transfemoral amputees: a knee joint with mechanical swing phase control vs a knee joint with a pneumatic swing phase control. Scand J Rehabil Med 1995;27(2):77–81. [5] Datta D, Heller B, Howitt J. A comparative evaluation of oxygen consumption and gait pattern in amputees using intelligent prostheses and conventionally damped knee swing-phase control. Clin Rehabil 2005;19(4):398–403. [6] Gonzalez EG, Corcoran PJ, Reyes RL. Energy expenditure in belowknee amputees: correlation with stump length. Arch Phys Med Rehabil 1974;55:111–9. [7] Gailey RS, Wenger MA, Raya M, Kirk N, Erbs K, Spyropoulos P, et al. Energy expenditure of trans-tibial amputees during ambulation at selfselected pace. Prosthet Orthot Int 1994;18(2):84–91. [8] Pagliarulo MA, Waters R, Hislop HJ. Energy cost of walking of belowknee amputees having no vascular disease. Phys Ther 1979;59(5): 538–42. [9] Torburn L, Powers CM, Guiterrez R, Perry J. Energy expenditure during ambulation in dysvascular and traumatic below-knee amputees: a comparison of five prosthetic feet. J Rehabil Res Dev 1995;32(2): 111–9. [10] Schmalz T, Blumentritt S, Jarasch R. Energy expenditure and biomechanical characteristics of lower limb amputee gait: the influence of prosthetic alignment and different prosthetic components. Gait Posture 2002;16(3):255–63. [11] Nowrozzi F, Salvanelli ML. Energy expenditure in hip disarticulation and hemipelvectomy amputees. Arch Phys Med Rehabil 1983;64: 300–3. [12] Pinzur MS, Gold J, Schwartz D, Gross N. Energy demands for walking in dysvascular amputees as related to the level of amputation. Orthopedics 1992;15(9):1033–6 [discussion 1036–7].
M. Traballesi et al. / Gait & Posture 27 (2008) 70–75 [13] Jaegers S, Vos L, Rispens P. The relationship between comfortable and most metabolically efficient walking speed in persons with unilateral above-knee amputation. Arch Phys Med Rehabil 1993;75:521–5. [14] Casillas JM, Dulieu V, Cohen M, Marcer I, Didier JP. Bioenergetic comparison of a new energy-storing foot and SACH foot in traumatic below-knee vascular amputations. Arch Phys Med Rehabil 1995;76(1): 39–44. [15] Huang GF, Chou YL, Su FC. Gait analysis and energy consumption of below-knee amputees wearing three different prosthetic feet. Gait Posture 2000;12(2):162–8 [Huang GF. Retraction. Gait Posture 2003;18(3):124]. [16] James KB, Stein RB. Improved ankle-foot system for above-knee amputees. Am J Phys Med 1986;65(6):301–14. [17] Gailey RS, Nash MS, Atchley TA, Zilmer RM, Moline-Little GR, Morris-Cresswell N, et al. The effects of prosthesis mass on metabolic cost of ambulation in non-vascular trans-tibial amputees. Prosthet Orthot Int 1997;21(1):9–16. [18] Taylor MB, Clark E, Offord EA, Baxter C. A comparison of energy expenditure by a high level trans-femoral amputee using the Intelligent Prosthesis and conventionally damped prosthetic limbs. Prosthet Orthot Int 1996;20(2):116–21. [19] Czerniecki JM, Gitter A, Weaver K. Effect of alterations in prosthetic shank mass on the metabolic costs of ambulation in above-knee amputees. Am J Phys Med Rehabil 1994;73(5):348–52. [20] Pearce ME, Cunningham DA, Donner AP, Rechnitzer PA, Fullerton GM, Howard JH. Energy cost of treadmill and floor walking at selfselected paces. Eur J Appl Physiol 1983;52:115–9.
75
[21] Bernardi M, Macaluso A, Sproviero E, Castellano V, Coratella D, Felici F, et al. Cost of walking and locomotor impairment. J Electromyogr Kinesiol 1999;9:149–57. [22] Fatone S, Boyd R, Rodda J, Gallagher D, Cullis E, Nattrass G, et al. Evaluation of the Cosmed K4 in a clinical gait analysis laboratory. Gait Posture 1997;5(1):79–179. [23] Gauthier-Gagnon C, Grise´ MC, Potvin D. Enabling factors related to prosthetic use by people with transtibial and transfemoral amputation. Arch Phys Med Rehabil 1999;80:706–13. [24] Greig C, Butler F, Skelton D, Mahmud S, Young A. Treadmill walking in old age may not reproduce the real life situation. J Am Geriatr Soc 1993;41(1):15–8. [25] Swerts PM, Mostert R, Wouters EF. Comparison of corridor and treadmill walking in patients with severe chronic obstructive pulmonary disease. Phys Ther 1990;70(7):439–42. [26] Vogt L, Banzer W, Bayer I, Schmidtbleicher D, Kerschbaumer F. Overground and walkway ambulation with unilateral hip osteoarthritis: comparison of step length asymmetries and reproducibility of treadmill mounted force plate readings. Physiother Theory Pract 2006;22(2):73–82. [27] White SC, Yack HJ, Tucker CA, Lin HY. Comparison of vertical ground reaction forces during overground and treadmill walking. Med Sci Sports Exerc 1998;30(10):1537–42. [28] Saunders JB, Inman VT, Eberhart HD. The major determinants in normal and pathological gait. J Bone Joint Surg 1953;35A(3):543–57. [29] Fisher SV, Gullickson Jr G. Energy cost of ambulation in health and disability: a literature review. Arch Phys Med Rehabil 1978;59(3): 124–33.