- Email: [email protected]
Are stance ankle plantar flexor muscles necessary to generate propulsive force during human gait initiation?
Neuroscience Letters 325 (2002) 139–143 www.elsevier.com/locate/neulet
Are stance ankle plantar flexor muscles necessary to generate propulsive force during human gait initiation? V. Michel*, M.C. Do Laboratoire de Physiologie du Mouvement, INSERM U483, Universite´ Paris-Sud, Bat 441, Orsay, France Received 11 April 2001; received in revised form 11 March 2002; accepted 11 March 2002
Abstract The study examined whether the generation of the forward propulsive force (PF) during gait initiation resulted mainly from the electromyogram activity of stance ankle plantar flexor muscles (APF) which ‘push’ on the ground as is generally claimed in the literature. Six unilateral above-knee amputees performed a specific gait initiation protocol, i.e. they were asked to walk as fast as possible from an upright posture. Data from a force platform were collected and processed to obtain gait parameters (centre of mass (CoM) acceleration, anteroposterior (A/P) progression velocity, step length, etc.). The results showed that the A/P CoM velocity at the time of foot-off differed depending on the state of the lower limb (sound or prosthetic limb) performing the step. However, the A/P velocity of the CoM reached at the time of foot contact was similar whatever the state of the lower limb initiating the gait. Thus, the absence of ankle and knee muscles did not affect the velocity of body progression, i.e. the generation of the PF in gait initiation. Furthermore, the comparable slopes of the A/P velocity between the stance sound limb and the stance prosthetic limb suggest that the organization of the motor synergy underlying the production of the PF remained the same and did not directly involve the APF. However, other mechanisms could explain PF generation. PF could be generated by the swing leg oscillation, by the trunk movement, or by other mechanisms such as the energy transfer and the exchange of gravity potential energy into kinetic energy. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Propulsive force; Progression velocity; Push; Ankle plantar flexors; Gait initiation
The displacement of a system, for example the body propulsion during gait, requires the application of propulsive force (PF) which can be measured by ground reaction force. According to Winter et al. [20–22], the PF results from a ‘push’ on the ground by ankle plantar flexor muscles (APF) during the mid-late stance phase. This interpretation was based on the increase of the electromyogram (EMG) activity of the APF with the progression velocity, which is representative of the forces applied at the centre of mass (CoM) [14,20–22]. However, the role of pushing devoted to the APF remains controversial. Mann et al. [11] and Simon et al. [18] proposed that the force produced by the EMG activity of the APF would control the body’s forward momentum during its progress forwards. For Murray et al. [15] and Sutherland et al. [19], the EMG activity of APF of the support limb would restrain the forward rotation of the tibia on the talus. Furthermore, according to Dillingham et
* Corresponding author. Tel.: 133-1-69-15-58-63; fax: 133-169-15-58-69. E-mail address: [email protected] (V. Michel).
al. [9], generation of the PF results mainly from the oscillation of the swing limb. The aim of this article was to contrast various hypotheses on the role of APF in the PF generation. In this study, we analyzed the CoM progression velocity in unilateral above-knee amputee subjects initiating gait with the sound limb then with the prosthetic limb. With regards to the ‘pushing’ hypothesis, the gait initiation paradigm is a demanding experimental situation when the subject is required to walk as fast as possible from an initial upright posture. Thus, we compared the anteroposterior (A/ P) CoM velocity reached at foot contact, the slope of progression velocity development and the forward velocity ‘gain’ measured during the swing phase. If the stance APF ‘pushed’ on the ground, one should observe a greater progression velocity when gait was initiated with the prosthetic limb than when gait was initiated with the sound limb. Furthermore, the slope of CoM progression velocity would also be different. In contrast, if A/P CoM velocity and the slope of the progression velocity showed no difference whatever the limb
0304-3940/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 0 2) 00 25 5- 0
140
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143
executing the step, then the ‘pushing’ hypothesis by the APF should be re-examined. Six unilateral above-knee amputees—two men and four women—with a mean age of 32 ^ 10 years and mean height of 1.65 ^ 0.15 m performed the experiment with their informed consent. The subjects stood upright on the force plate and were instructed to walk as quickly as they could following a beep triggered by the experimenter. Each subject performed two series of ten trials: gait initiated with the prosthetic limb (P condition); and with the sound limb (S condition). A wooden walkway extended the force platform to allow the subject to take at least four steps for each recording. Identical prostheses which do not allow energy restitution were used by all subjects. The time of amputation dated at least 6 years, the level of amputation was comparable for all the subjects, and the prosthesis was used regularly by the amputee subjects. No pain or discomfort was reported by the subjects before, during or after the experimental session. A large force plate (AMTI, 0.6 m £ 1.20 m) was used to collect ground reaction forces and moments which were transformed to obtain CoM acceleration and the co-ordinates of centre of foot pressure (see Fig. 1). CoM acceleration was directly calculated from Newton’s first law: m £ D ¼ W 1 Rwhere m is the subject’s mass, D is the CoM acceleration, W is the subject’s weight and R is the ground reaction force. The instantaneous progression velocity of the CoM was obtained by integration of the acceleration. In this study, we were interested only in the phenomena occurring in the A/P axis (progression axis). The kinematic parameters analyzed were (see Fig. 1; Table 1): time of footoff (FO); A/P CoM velocity at foot-off (VFO); time of foot contact (FC); A/P CoM velocity at foot contact (VFC); swing phase duration (Swg); the maximum A/P CoM velocity reached at the end of the first step (Vm); the gain A/P CoM velocity (ga); i.e. the difference between (VFC) and (VFO); and the slope of the A/P CoM velocity (Sp), Sp ¼ ga=Swg. The latter parameter was allowed to determine how CoM progression velocity varied during the single support phase. The statistical analysis used analysis of variance with repeated measures to test the difference between the S vs. P conditions. The threshold of significant difference was P , 0:05. Qualitatively, the biomechanical traces of the gait initiated with the sound limb and with the prosthetic limb were roughly similar, and also similar to normal subjects (see Fig. 1), although one difference was often observed (see Fig. 1). We observed the absence of braking in A/P CoM acceleration (i.e. absence of change in the sign of X 00 g after foot contact of the swing limb) when above-knee amputee subjects initiated gait with the prosthetic limb. This phenomenon occurred in approximately half of all trials (all subjects combined). When there was an absence of braking, there was an absence of peak in the CoM progression velocity trace; in this case, the peak of progression velocity was measured at the time of minimal A/P CoM
acceleration, i.e. at the time of the second toe-off (FO2; see Fig. 1, right traces, X 00 g trace). Otherwise, the peak of progression velocity at the end of first step was easy to determine (see Fig. 1, left traces, X 00 g and X 0 g). As in normal gait initiation [2–5], subjects showed a postural adjustment period prior to step execution, commonly named anticipatory postural adjustment. When the gait was initiated with the sound limb, the time of footoff, i.e. onset time of step execution (FO) occurred 0.77 s (SD ¼ 0:13 s) after the onset of variation of the biomechanical traces vs. 0.45 s (SD ¼ 0:10 s) when the gait was initiated with the prosthetic limb (P , 0:001). A/P CoM velocity at foot-off (VFO) was statistically higher for the S condition than the P condition (P , 0:002), the means were 0.56 (0.15 m/s) and 0.15 m/s (0.09 m/s), respectively. Foot contact time (FC) for a gait initiating with the sound limb and the prosthetic limb occurred 1.06 (0.13 s) and 0.94 s (0.08 s), respectively (P . 0:46) after the onset of biomechanical traces. A/P CoM velocity at the foot contact (VFC) when the subjects initiated the gait with the sound limb was 0.97 m/ s (0.17 m/s), and 0.85 m/s (0.21 m/s) when they initiated the gait with the prosthetic limb. There was no statistical difference (P . 0:20). Swing phase duration (Swg) was significantly shorter for a step executed with the sound limb than with the prosthetic limb; mean values were 0.29 (0.06 s) and 0.50 s (0.07 s), respectively (P , 0:0002). The average A/P CoM velocity (Vm) when subjects initiated gait with the sound limb was 1.15 m/s (0.18 m/s), and 1.15 m/s (0.24 m/ s) when they initiated the gait with the prosthetic limb. There was no statistical difference (P . 0:99). The absolute gain of A/P CoM velocity (ga) during the swing phase was lower for the S condition than the P condition and the means were 0.41 (0.11 m/s) and 0.70 m/s (0.19 m/s), respectively (P , 0:03). The slope of A/P CoM velocity (Sp) showed no significant difference between the S condition and the P condition Table 1 Mean and standard deviation of the parameters of gait initiation in above-knee amputees subjects executing a step with the prosthetic limb a or the sound limb b,c
FO (s) VFO (m/s) FC (s) VFC (m/s) Swg (s) Vm (m/s) ga (m/s) Sp (m/s/s) a
S
P
P value
0.77 ^ 0.13 0.56 ^ 0.15 1.06 ^ 0.13 0.97 ^ 0.17 0.29 ^ 0.06 1.15 ^ 0.18 0.41 ^ 0.11 1.49 ^ 0.72
0.45 ^ 0.10 0.15 ^ 0.09 0.94 ^ 0.08 0.85 ^ 0.21 0.50 ^ 0.07 1.15 ^ 0.24 0.70 ^ 0.19 1.41 ^ 0.32
,0.001 ,0.002 .0.46 ,0.20 ,0.0002 .0.99 ,0.03 .0.63
P condition. S condition. c FO, time of foot-off; VFO, A/P centre of gravity velocity at footoff; FC, time of foot contact; VFC, A/P centre of gravity velocity at foot contact; Swg, swing phase duration; Vm, maximal A/P centre of gravity velocity reached at the end of the first step; ga, absolute gain A/P centre of gravity velocity; Sp, slope of A/P centre of gravity velocity. b
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143
(P , 0:63). Mean values were 1.49 (0.72 m/s/s) and 1.41 m/ s/s (0.32 m/s/s), respectively. Our results showed that, whichever the limb initiating the gait—sound or prosthetic—the A/P CoM at foot contact (VFC) was similar. It was the same for the slope of the progression velocity as a function of the time during the single support phase. Hence, the lack of numerous lower limb muscles, and particularly the APF, did not affect the generation of the PFs. This suggests that the muscular activ-
141
ity of the stance APF was not necessary to generate the PF. However, two opposing hypotheses remained in competition. As regards to the ‘push’ hypothesis, it is possible that when the APF are present, they directly contribute to the generation of the PF by ‘pushing’ on the ground, and when they are absent, their role is supplied or replaced by the proximal remaining muscles, such as hip flexors [1]. The second hypothesis is that PF could be generated by other mechanisms than a push on the ground by APF.
Fig. 1. Recording of the biomechanical parameters of gait initiation in above-knee amputee subjects initiating step with the sound limb (S condition) and the prosthetic limb (P condition). X 00 g, X 0 g, acceleration and velocity of the centre of gravity along the A/P axis; Z 00 g, acceleration of the centre of gravity along the vertical axis; Xp, Yp, A/P and lateral displacement of the centre of pressure; to, time of onset of the first mechanical phenomenon; FO, time of foot-off; FC, time of foot contact; FO2, time of the second foot-off; APA, anticipatory postural adjustment phase; Swg, swing phase duration; VFO, A/P centre of gravity velocity at foot-off; VFC, A/P centre of gravity velocity at foot contact; Vm, A/P maximal centre of gravity velocity.
142
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143
Our results and some data in the literature do not support the first hypothesis. The similar slopes of the PFs as a function of the time during the single support phase, whichever the side carrying out the step, suggest that the organization of the motor synergy underlying the production of the PF remained the same and did not involve the APF as a direct generator of PF. This interpretation is supported by comparative data of APF EMG activity elicited in normal gait initiation as compared with gait induced by a forward fall [7] and in normal walking as compared with treadmill walking [1,16]. In gait induced by a forward fall [7], the triceps surae muscle EMG activity was greater than in normal gait initiation, for comparable progression velocity. In treadmill walking as compared with normal walking at a similar speed, the APF EMG activity was comparable [1,16]. In terms of PF control, the PF is generated in forward fall by the release of the initial inclined posture and the EMG of stance APF would not be related to PF generation but would be devoted mainly to reacting to the fall and to maintaining postural equilibrium. In treadmill walking, as the body remains roughly in place, low or no PF is applied to the CoM, and the EMG of stance APF would be devoted to controlling the postural equilibrium during the single support phase. In normal gait initiation, the subject has to generate PF and also control postural equilibrium during the single support phase as in treadmill walking. The higher EMG activity in forward fall could be explained by the more demanding constraint of postural equilibrium. So, the EMG activity in stance APF would not be related to the direct mechanism of generation of PF, but rather to controlling postural equilibrium. This interpretation of the functional role of control of postural equilibrium assumed for the EMG activity of APF is compatible with the interpretation of restraining the body’s forward momentum [11,18], restraining the forward rotation of the tibia on the talus [15,19] or increasing leg energy and accelerating the leg into the swing phase [13]. So, by which mechanism is the PF generated? The generation of PFs by the oscillation of the swing limb [9] is in line with our results. Our results showed that when the step was initiated with the prosthetic limb, VFO was slower than when the step was initiated with the sound limb. However, VFC was comparable whatever the state of the limb executing the step. The greater gain (ga) of progression velocity during the swing phase of the prosthetic limb is compatible with a contribution of the swing limb to generating PF. This hypothesis does not disallow the participation of the trunk movement in the PF proposed by McGibbon et al. [12]. These authors have shown that in the old subject the control of lower trunk movement during gait would preserve walking speed. So, it is likely that lower trunk movement also contributes to FP generation. Two other hypotheses about FP could be taken into account. According to Gitter et al. [10] and Meinders et al. [13], forward body progression during gait would result
in energy transfer from the swing lower limb to the trunk. A comparable mechanism would occur in gait [10,17] and running [8] in amputee subjects. Nevertheless, the exchange of gravitational potential energy and kinetic energy, shown by Cavagna et al. [6], during level walking that occurs during the swing phase in gait initiation could generate FP. In conclusion, APF would not generate PFs by a push on the ground during gait initiation. The generation of the PF could come from the oscillating swing limb, trunk movement, energy transfer from swing leg to the trunk and exchange of potential energy into kinetic energy. The authors would like to thank Dr P. Sautreuil and Dr Ph. Thoumie for recruiting the subjects. [1] Arsenault, A.B., Winter, D.A. and Marteniuk, R.G., Treadmill vs. walkway locomotion in human: an EMG study, Ergonomics, 29(5) (1986) 665–676. [2] Brenie`re, Y. and Do, M.C., When and how does state gait movement induced from upright posture begin? J. Biomech., 19 (1986) 1035–1040. [3] Brenie`re, Y. and Do, M.C., Modifications posturales associe´ es au lever de talon dans l’initiation du pas de la marche, J. Biophys. Biomec., 11 (1987) 161–167. [4] Brenie`re, Y., Do, M.C. and Sanchez, J., A biomechanical study of the gait initiation process, J. Biophys. Med. Nucl., 5 (1981) 197–205. [5] Brenie`re, Y., Do, M.C. and Bouisset, S., Are dynamic phenomena prior to stepping essential to walking? J. Mot. Behav., 19 (1987) 62–76. [6] Cavagna, G.A., Heglund, N.C. and Taylor, C.R., Mechanical work in terrestrial locomotion: two basic mechanisms for minimizing energy expenditure, Am. J. Physiol., 233(5) (1977) 243–261. [7] Chong, R.K.Y., Michel, V. and Do, M.C., Is soleus activity necessary for forward body progression during gait initiation (Abstract), Society of Neurosciences, 2000. [8] Czerniecki, J.M., Gitter, A.J. and Beck, J.C., Energy mechanism as a compasatory strategy in below knee amputee runners, J. Biomech., 29(6) (1996) 717–722. [9] Dillingham, T.R., Lehmann, J.F. and Price, R., Effect of lower limb on body propulsion, Arch. Phys. Med. Rehabil., 73 (1992) 647–651. [10] Gitter, A., Czerniecki, J. and Meinders, M., Effect of prosthetic mass on swing phase work during above-knee amputee ambulation, Am. J. Phys. Med. Rehabil., 76 (1997) 114– 121. [11] Mann, R.A., Hagy, J.L. and Simon, S.R., Push off phase of gait, Abbott Proc., 5 (1974) 85. [12] McGibbon, C.A. and Krebs, D.E., Age-related changes in lower trunk coordination and energy transfer during gait, J. Neurophysiol., 85(5) (2001) 1923–1931. [13] Meinders, M., Gitter, A. and Czerniecki, J.M., The role of the plantar muscle work during walking, Scand. J. Rehabil. Med., 30 (1998) 30–46. [14] Milner, M., Basmajian, J.V. and Quanbury, A.O., Multifactorial analysis by electromyography and computer, Am. J. Phys. Med., 50 (1971) 235. [15] Murray, M.P., Gutten, G.N., Sepic, S.B., Garder, G.M. and Baldwin, J.M., Function of the triceps surae during gait, J. Bone Jt. Surg., 60-A (1978) 473–476. [16] Murray, M.P., Spurr, G.B., Gardner, G.M. and Mollinger, L.A., Treadmill vs. floor walking: kinematic, electromyo-
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143 gram, and heart rate, J. Appl. Physiol., 59(1) (1985) 87– 91. [17] Seroussi, R.E., Gitter, A., Czerniecki, J.M. and Weaver, K., Mechanical work adaptation of above-knee amputee ambulation, Arch. Phys. Med. Rehabil., 77(11) (1996) 1209– 1214. [18] Simon, S.R., Mann, M.D., Hagy, J.L. and Larsen, L.J., Role of the posterior calf muscles in normal gait, J. Bone Jt. Surg., 60-A (1978) 465–472. [19] Sutherland, D.A., Cooper, L. and Daniel, D., The role of the
143
ankle plantar flexors in normal walking, J Bone Joint Surg., 62-A (1980) 354–363. [20] Winter, D.A., Overall principle of lower limb support during stance phase of gait, J. Biomech., 13 (1980) 923. [21] Winter, D.A., Energy generation and absorption at the ankle knee during fast natural, slow cadences, Clin. Orthop., 1758 (1983) 147–154. [22] Winter, D.A. and Sienko, S.E., Biomechanics of below-knee amputee gait, J. Biomech., 21(5) (1988) 361–367.
Are stance ankle plantar flexor muscles necessary to generate propulsive force during human gait initiation? V. Michel*, M.C. Do Laboratoire de Physiologie du Mouvement, INSERM U483, Universite´ Paris-Sud, Bat 441, Orsay, France Received 11 April 2001; received in revised form 11 March 2002; accepted 11 March 2002
Abstract The study examined whether the generation of the forward propulsive force (PF) during gait initiation resulted mainly from the electromyogram activity of stance ankle plantar flexor muscles (APF) which ‘push’ on the ground as is generally claimed in the literature. Six unilateral above-knee amputees performed a specific gait initiation protocol, i.e. they were asked to walk as fast as possible from an upright posture. Data from a force platform were collected and processed to obtain gait parameters (centre of mass (CoM) acceleration, anteroposterior (A/P) progression velocity, step length, etc.). The results showed that the A/P CoM velocity at the time of foot-off differed depending on the state of the lower limb (sound or prosthetic limb) performing the step. However, the A/P velocity of the CoM reached at the time of foot contact was similar whatever the state of the lower limb initiating the gait. Thus, the absence of ankle and knee muscles did not affect the velocity of body progression, i.e. the generation of the PF in gait initiation. Furthermore, the comparable slopes of the A/P velocity between the stance sound limb and the stance prosthetic limb suggest that the organization of the motor synergy underlying the production of the PF remained the same and did not directly involve the APF. However, other mechanisms could explain PF generation. PF could be generated by the swing leg oscillation, by the trunk movement, or by other mechanisms such as the energy transfer and the exchange of gravity potential energy into kinetic energy. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Propulsive force; Progression velocity; Push; Ankle plantar flexors; Gait initiation
The displacement of a system, for example the body propulsion during gait, requires the application of propulsive force (PF) which can be measured by ground reaction force. According to Winter et al. [20–22], the PF results from a ‘push’ on the ground by ankle plantar flexor muscles (APF) during the mid-late stance phase. This interpretation was based on the increase of the electromyogram (EMG) activity of the APF with the progression velocity, which is representative of the forces applied at the centre of mass (CoM) [14,20–22]. However, the role of pushing devoted to the APF remains controversial. Mann et al. [11] and Simon et al. [18] proposed that the force produced by the EMG activity of the APF would control the body’s forward momentum during its progress forwards. For Murray et al. [15] and Sutherland et al. [19], the EMG activity of APF of the support limb would restrain the forward rotation of the tibia on the talus. Furthermore, according to Dillingham et
* Corresponding author. Tel.: 133-1-69-15-58-63; fax: 133-169-15-58-69. E-mail address: [email protected] (V. Michel).
al. [9], generation of the PF results mainly from the oscillation of the swing limb. The aim of this article was to contrast various hypotheses on the role of APF in the PF generation. In this study, we analyzed the CoM progression velocity in unilateral above-knee amputee subjects initiating gait with the sound limb then with the prosthetic limb. With regards to the ‘pushing’ hypothesis, the gait initiation paradigm is a demanding experimental situation when the subject is required to walk as fast as possible from an initial upright posture. Thus, we compared the anteroposterior (A/ P) CoM velocity reached at foot contact, the slope of progression velocity development and the forward velocity ‘gain’ measured during the swing phase. If the stance APF ‘pushed’ on the ground, one should observe a greater progression velocity when gait was initiated with the prosthetic limb than when gait was initiated with the sound limb. Furthermore, the slope of CoM progression velocity would also be different. In contrast, if A/P CoM velocity and the slope of the progression velocity showed no difference whatever the limb
0304-3940/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S03 04 - 394 0( 0 2) 00 25 5- 0
140
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143
executing the step, then the ‘pushing’ hypothesis by the APF should be re-examined. Six unilateral above-knee amputees—two men and four women—with a mean age of 32 ^ 10 years and mean height of 1.65 ^ 0.15 m performed the experiment with their informed consent. The subjects stood upright on the force plate and were instructed to walk as quickly as they could following a beep triggered by the experimenter. Each subject performed two series of ten trials: gait initiated with the prosthetic limb (P condition); and with the sound limb (S condition). A wooden walkway extended the force platform to allow the subject to take at least four steps for each recording. Identical prostheses which do not allow energy restitution were used by all subjects. The time of amputation dated at least 6 years, the level of amputation was comparable for all the subjects, and the prosthesis was used regularly by the amputee subjects. No pain or discomfort was reported by the subjects before, during or after the experimental session. A large force plate (AMTI, 0.6 m £ 1.20 m) was used to collect ground reaction forces and moments which were transformed to obtain CoM acceleration and the co-ordinates of centre of foot pressure (see Fig. 1). CoM acceleration was directly calculated from Newton’s first law: m £ D ¼ W 1 Rwhere m is the subject’s mass, D is the CoM acceleration, W is the subject’s weight and R is the ground reaction force. The instantaneous progression velocity of the CoM was obtained by integration of the acceleration. In this study, we were interested only in the phenomena occurring in the A/P axis (progression axis). The kinematic parameters analyzed were (see Fig. 1; Table 1): time of footoff (FO); A/P CoM velocity at foot-off (VFO); time of foot contact (FC); A/P CoM velocity at foot contact (VFC); swing phase duration (Swg); the maximum A/P CoM velocity reached at the end of the first step (Vm); the gain A/P CoM velocity (ga); i.e. the difference between (VFC) and (VFO); and the slope of the A/P CoM velocity (Sp), Sp ¼ ga=Swg. The latter parameter was allowed to determine how CoM progression velocity varied during the single support phase. The statistical analysis used analysis of variance with repeated measures to test the difference between the S vs. P conditions. The threshold of significant difference was P , 0:05. Qualitatively, the biomechanical traces of the gait initiated with the sound limb and with the prosthetic limb were roughly similar, and also similar to normal subjects (see Fig. 1), although one difference was often observed (see Fig. 1). We observed the absence of braking in A/P CoM acceleration (i.e. absence of change in the sign of X 00 g after foot contact of the swing limb) when above-knee amputee subjects initiated gait with the prosthetic limb. This phenomenon occurred in approximately half of all trials (all subjects combined). When there was an absence of braking, there was an absence of peak in the CoM progression velocity trace; in this case, the peak of progression velocity was measured at the time of minimal A/P CoM
acceleration, i.e. at the time of the second toe-off (FO2; see Fig. 1, right traces, X 00 g trace). Otherwise, the peak of progression velocity at the end of first step was easy to determine (see Fig. 1, left traces, X 00 g and X 0 g). As in normal gait initiation [2–5], subjects showed a postural adjustment period prior to step execution, commonly named anticipatory postural adjustment. When the gait was initiated with the sound limb, the time of footoff, i.e. onset time of step execution (FO) occurred 0.77 s (SD ¼ 0:13 s) after the onset of variation of the biomechanical traces vs. 0.45 s (SD ¼ 0:10 s) when the gait was initiated with the prosthetic limb (P , 0:001). A/P CoM velocity at foot-off (VFO) was statistically higher for the S condition than the P condition (P , 0:002), the means were 0.56 (0.15 m/s) and 0.15 m/s (0.09 m/s), respectively. Foot contact time (FC) for a gait initiating with the sound limb and the prosthetic limb occurred 1.06 (0.13 s) and 0.94 s (0.08 s), respectively (P . 0:46) after the onset of biomechanical traces. A/P CoM velocity at the foot contact (VFC) when the subjects initiated the gait with the sound limb was 0.97 m/ s (0.17 m/s), and 0.85 m/s (0.21 m/s) when they initiated the gait with the prosthetic limb. There was no statistical difference (P . 0:20). Swing phase duration (Swg) was significantly shorter for a step executed with the sound limb than with the prosthetic limb; mean values were 0.29 (0.06 s) and 0.50 s (0.07 s), respectively (P , 0:0002). The average A/P CoM velocity (Vm) when subjects initiated gait with the sound limb was 1.15 m/s (0.18 m/s), and 1.15 m/s (0.24 m/ s) when they initiated the gait with the prosthetic limb. There was no statistical difference (P . 0:99). The absolute gain of A/P CoM velocity (ga) during the swing phase was lower for the S condition than the P condition and the means were 0.41 (0.11 m/s) and 0.70 m/s (0.19 m/s), respectively (P , 0:03). The slope of A/P CoM velocity (Sp) showed no significant difference between the S condition and the P condition Table 1 Mean and standard deviation of the parameters of gait initiation in above-knee amputees subjects executing a step with the prosthetic limb a or the sound limb b,c
FO (s) VFO (m/s) FC (s) VFC (m/s) Swg (s) Vm (m/s) ga (m/s) Sp (m/s/s) a
S
P
P value
0.77 ^ 0.13 0.56 ^ 0.15 1.06 ^ 0.13 0.97 ^ 0.17 0.29 ^ 0.06 1.15 ^ 0.18 0.41 ^ 0.11 1.49 ^ 0.72
0.45 ^ 0.10 0.15 ^ 0.09 0.94 ^ 0.08 0.85 ^ 0.21 0.50 ^ 0.07 1.15 ^ 0.24 0.70 ^ 0.19 1.41 ^ 0.32
,0.001 ,0.002 .0.46 ,0.20 ,0.0002 .0.99 ,0.03 .0.63
P condition. S condition. c FO, time of foot-off; VFO, A/P centre of gravity velocity at footoff; FC, time of foot contact; VFC, A/P centre of gravity velocity at foot contact; Swg, swing phase duration; Vm, maximal A/P centre of gravity velocity reached at the end of the first step; ga, absolute gain A/P centre of gravity velocity; Sp, slope of A/P centre of gravity velocity. b
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143
(P , 0:63). Mean values were 1.49 (0.72 m/s/s) and 1.41 m/ s/s (0.32 m/s/s), respectively. Our results showed that, whichever the limb initiating the gait—sound or prosthetic—the A/P CoM at foot contact (VFC) was similar. It was the same for the slope of the progression velocity as a function of the time during the single support phase. Hence, the lack of numerous lower limb muscles, and particularly the APF, did not affect the generation of the PFs. This suggests that the muscular activ-
141
ity of the stance APF was not necessary to generate the PF. However, two opposing hypotheses remained in competition. As regards to the ‘push’ hypothesis, it is possible that when the APF are present, they directly contribute to the generation of the PF by ‘pushing’ on the ground, and when they are absent, their role is supplied or replaced by the proximal remaining muscles, such as hip flexors [1]. The second hypothesis is that PF could be generated by other mechanisms than a push on the ground by APF.
Fig. 1. Recording of the biomechanical parameters of gait initiation in above-knee amputee subjects initiating step with the sound limb (S condition) and the prosthetic limb (P condition). X 00 g, X 0 g, acceleration and velocity of the centre of gravity along the A/P axis; Z 00 g, acceleration of the centre of gravity along the vertical axis; Xp, Yp, A/P and lateral displacement of the centre of pressure; to, time of onset of the first mechanical phenomenon; FO, time of foot-off; FC, time of foot contact; FO2, time of the second foot-off; APA, anticipatory postural adjustment phase; Swg, swing phase duration; VFO, A/P centre of gravity velocity at foot-off; VFC, A/P centre of gravity velocity at foot contact; Vm, A/P maximal centre of gravity velocity.
142
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143
Our results and some data in the literature do not support the first hypothesis. The similar slopes of the PFs as a function of the time during the single support phase, whichever the side carrying out the step, suggest that the organization of the motor synergy underlying the production of the PF remained the same and did not involve the APF as a direct generator of PF. This interpretation is supported by comparative data of APF EMG activity elicited in normal gait initiation as compared with gait induced by a forward fall [7] and in normal walking as compared with treadmill walking [1,16]. In gait induced by a forward fall [7], the triceps surae muscle EMG activity was greater than in normal gait initiation, for comparable progression velocity. In treadmill walking as compared with normal walking at a similar speed, the APF EMG activity was comparable [1,16]. In terms of PF control, the PF is generated in forward fall by the release of the initial inclined posture and the EMG of stance APF would not be related to PF generation but would be devoted mainly to reacting to the fall and to maintaining postural equilibrium. In treadmill walking, as the body remains roughly in place, low or no PF is applied to the CoM, and the EMG of stance APF would be devoted to controlling the postural equilibrium during the single support phase. In normal gait initiation, the subject has to generate PF and also control postural equilibrium during the single support phase as in treadmill walking. The higher EMG activity in forward fall could be explained by the more demanding constraint of postural equilibrium. So, the EMG activity in stance APF would not be related to the direct mechanism of generation of PF, but rather to controlling postural equilibrium. This interpretation of the functional role of control of postural equilibrium assumed for the EMG activity of APF is compatible with the interpretation of restraining the body’s forward momentum [11,18], restraining the forward rotation of the tibia on the talus [15,19] or increasing leg energy and accelerating the leg into the swing phase [13]. So, by which mechanism is the PF generated? The generation of PFs by the oscillation of the swing limb [9] is in line with our results. Our results showed that when the step was initiated with the prosthetic limb, VFO was slower than when the step was initiated with the sound limb. However, VFC was comparable whatever the state of the limb executing the step. The greater gain (ga) of progression velocity during the swing phase of the prosthetic limb is compatible with a contribution of the swing limb to generating PF. This hypothesis does not disallow the participation of the trunk movement in the PF proposed by McGibbon et al. [12]. These authors have shown that in the old subject the control of lower trunk movement during gait would preserve walking speed. So, it is likely that lower trunk movement also contributes to FP generation. Two other hypotheses about FP could be taken into account. According to Gitter et al. [10] and Meinders et al. [13], forward body progression during gait would result
in energy transfer from the swing lower limb to the trunk. A comparable mechanism would occur in gait [10,17] and running [8] in amputee subjects. Nevertheless, the exchange of gravitational potential energy and kinetic energy, shown by Cavagna et al. [6], during level walking that occurs during the swing phase in gait initiation could generate FP. In conclusion, APF would not generate PFs by a push on the ground during gait initiation. The generation of the PF could come from the oscillating swing limb, trunk movement, energy transfer from swing leg to the trunk and exchange of potential energy into kinetic energy. The authors would like to thank Dr P. Sautreuil and Dr Ph. Thoumie for recruiting the subjects. [1] Arsenault, A.B., Winter, D.A. and Marteniuk, R.G., Treadmill vs. walkway locomotion in human: an EMG study, Ergonomics, 29(5) (1986) 665–676. [2] Brenie`re, Y. and Do, M.C., When and how does state gait movement induced from upright posture begin? J. Biomech., 19 (1986) 1035–1040. [3] Brenie`re, Y. and Do, M.C., Modifications posturales associe´ es au lever de talon dans l’initiation du pas de la marche, J. Biophys. Biomec., 11 (1987) 161–167. [4] Brenie`re, Y., Do, M.C. and Sanchez, J., A biomechanical study of the gait initiation process, J. Biophys. Med. Nucl., 5 (1981) 197–205. [5] Brenie`re, Y., Do, M.C. and Bouisset, S., Are dynamic phenomena prior to stepping essential to walking? J. Mot. Behav., 19 (1987) 62–76. [6] Cavagna, G.A., Heglund, N.C. and Taylor, C.R., Mechanical work in terrestrial locomotion: two basic mechanisms for minimizing energy expenditure, Am. J. Physiol., 233(5) (1977) 243–261. [7] Chong, R.K.Y., Michel, V. and Do, M.C., Is soleus activity necessary for forward body progression during gait initiation (Abstract), Society of Neurosciences, 2000. [8] Czerniecki, J.M., Gitter, A.J. and Beck, J.C., Energy mechanism as a compasatory strategy in below knee amputee runners, J. Biomech., 29(6) (1996) 717–722. [9] Dillingham, T.R., Lehmann, J.F. and Price, R., Effect of lower limb on body propulsion, Arch. Phys. Med. Rehabil., 73 (1992) 647–651. [10] Gitter, A., Czerniecki, J. and Meinders, M., Effect of prosthetic mass on swing phase work during above-knee amputee ambulation, Am. J. Phys. Med. Rehabil., 76 (1997) 114– 121. [11] Mann, R.A., Hagy, J.L. and Simon, S.R., Push off phase of gait, Abbott Proc., 5 (1974) 85. [12] McGibbon, C.A. and Krebs, D.E., Age-related changes in lower trunk coordination and energy transfer during gait, J. Neurophysiol., 85(5) (2001) 1923–1931. [13] Meinders, M., Gitter, A. and Czerniecki, J.M., The role of the plantar muscle work during walking, Scand. J. Rehabil. Med., 30 (1998) 30–46. [14] Milner, M., Basmajian, J.V. and Quanbury, A.O., Multifactorial analysis by electromyography and computer, Am. J. Phys. Med., 50 (1971) 235. [15] Murray, M.P., Gutten, G.N., Sepic, S.B., Garder, G.M. and Baldwin, J.M., Function of the triceps surae during gait, J. Bone Jt. Surg., 60-A (1978) 473–476. [16] Murray, M.P., Spurr, G.B., Gardner, G.M. and Mollinger, L.A., Treadmill vs. floor walking: kinematic, electromyo-
V. Michel, M.C. Do / Neuroscience Letters 325 (2002) 139–143 gram, and heart rate, J. Appl. Physiol., 59(1) (1985) 87– 91. [17] Seroussi, R.E., Gitter, A., Czerniecki, J.M. and Weaver, K., Mechanical work adaptation of above-knee amputee ambulation, Arch. Phys. Med. Rehabil., 77(11) (1996) 1209– 1214. [18] Simon, S.R., Mann, M.D., Hagy, J.L. and Larsen, L.J., Role of the posterior calf muscles in normal gait, J. Bone Jt. Surg., 60-A (1978) 465–472. [19] Sutherland, D.A., Cooper, L. and Daniel, D., The role of the
143
ankle plantar flexors in normal walking, J Bone Joint Surg., 62-A (1980) 354–363. [20] Winter, D.A., Overall principle of lower limb support during stance phase of gait, J. Biomech., 13 (1980) 923. [21] Winter, D.A., Energy generation and absorption at the ankle knee during fast natural, slow cadences, Clin. Orthop., 1758 (1983) 147–154. [22] Winter, D.A. and Sienko, S.E., Biomechanics of below-knee amputee gait, J. Biomech., 21(5) (1988) 361–367.