| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
SKELETAL MUSCLE AND EXERCISE |
1 Rehabilitation and Physical Medicine Unit, Université catholique de Louvain, Avenue Mounier 53, 1200 Brussels, Belgium
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
(Received 10 January 2007;
accepted after revision 20 April 2007;
first published online 26 April 2007)
Corresponding author F. Massaad: Rehabilitation and Physical Medicine Unit, Université catholique de Louvain, Tour Pasteur 5375, Avenue Mounier 53, 1200 Brussels, Belgium. Email: firas.massaad{at}gmail.com
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Humans walk straight, but even the apes – our closest relatives – rarely walk on two legs, and when they do, their legs bend in a compliant pattern (Alexander & Jayes, 1978; Alexander, 1992). Compliant gait is paramount for arboreal locomotion because it reduces ground reaction forces and involves little oscillations of the CM (i.e. flat gait), thus enhancing stability and allowing primates to maintain loosely stabilized joints (Schmitt, 2003). Recent studies suggest that our common ancestors with apes, probably 
7 million years ago (Gibbons, 2002), might also have walked flexed and flat, which would have facilitated their transition to terrestrial bipedality when they had not completely forsaken arboreal locomotion (Schmitt, 2003). Hence, the important locomotor innovation characterizing the human lineage was likely to be the development of straight-legged walking rather than upright walking per se (Alexander, 1992; Schmitt, 2003). When did this occur? The australopithecines, as represented by the famous 3.2-million-year-old ‘Lucy’, were short with poorly stabilized limbs, sacrum and vertebral joints (Stern & Susman, 1983; Ward, 2002). Therefore, scientists believed that they may have walked flexed and flat, and then the shift to a more robust skeleton in Homo erectus
1.8 million years ago may have reflected the adoption of straight-legged walking (Schmitt, 2003). However, recent simulation studies suggest that Lucy already walked straight and moved up and down (Nagano et al. 2005; Sellers et al. 2005), while its primitive anatomical features were rather associated with a limited ability for endurance running (Bramble & Lieberman, 2004). Hence, straight-legged walking with its inherent vertical CM movement may have appeared well before the genus Homo.
The peculiar relatively straight-legged walking in modern humans, and the energy saving that it may induce, may suggest that the straighter we walk the less energy we consume. However, straight legs may induce a bouncy walking characterized by larger vertical CM displacement and abrupt changes in CM path from one arc to the next. This would imply large forces to redirect the CM and to raise the body against gravity during each step, thus increasing energy cost. Hence, Saunders et al. (1953) postulated that humans walk as flat as possible, by using some of the flexed joint characteristics seen in our ancestors' compliant gaits, to decrease the energy cost. Recent human experiments have shown that, contrary to Saunders' hypothesis, flat walking costs more energy than normal walking (Ortega & Farley, 2005). On the other hand, a recent breakthrough in a computer optimization model of human locomotion demonstrated that all optimization simulations converged towards a normal walking style but never to a flat gait (Srinivasan & Ruina, 2006). Nonetheless, the physiological reasons that underlie the increased energy cost in flat walking remain unknown, in addition to the questions of why our natural mode of walking is the optimal one in comparison with flat or bouncy walking and where we stand between flatness and bounce.
We set out to test the mechanics, energetics, muscle efficiency and optimization of human walking with regard to the CM displacement and to assess the benefits from our inherent vertical movement. This study will provide insights into the physiological determinants of energy consumption in human walking and the reasons for nature to have chosen our natural modes of locomotion.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Mechanical, metabolic and electromyographic data were acquired from six healthy adult volunteers (5 female and 1 male, age 24 ± 1 years, weight 60.3 ± 11 kg, height 1.7 ± 0.1 m (mean ± S.D.)). All were moderately active, but none were specialists in any particular sport. Before the experiment, the subjects signed an informed consent approved by the local ethics committee after the nature of the study was explained. The experiments involved no discomfort and were performed according to the Declaration of Helsinki.
The subjects walked barefoot on a custom-built motorized treadmill mounted on four 3-D strain-gauge force transducers (Dierick et al. 2004). All six subjects walked at all six speeds (1, 2, 3, 4, 5 and 6 km h–1). For each speed, they walked in three different conditions: normally, with minimum vertical CM displacement (flat walking), and with maximum vertical CM displacement (bouncy walking). In all, 18 walking trials were carried out in a random order to eliminate any possible order effects and in two testing sessions on two different days to avoid physical fatigue. Before testing sessions, the subjects walked on the treadmill in a practice session during which they were given at least 30 min to become accustomed to the treadmill and the three types of walking at different speeds. This practice session exceeded the minimum treadmill habituation time of 10 min recommended in the literature for healthy subjects (Wall & Charteris, 1981; Van de Putte et al. 2006). The habituation time was similar for both flat and bouncy walking and a resting time was given between the practice and the testing session.
Metabolic data were continuously measured during the experiments. A baseline standing metabolic rate was taken at the beginning of each measurement session, for at least 3 min. Each walking trial was maintained as long as necessary to obtain a steady metabolic state for at least 3 min, during which we simultaneously collected metabolic, kinematic, EMG and ground reaction force data of 10 consecutive walking strides (i.e. the time between two successive instants of initial floor contact of the same foot). After each walking trial, a resting time was given until metabolic data and pulse rate returned to initial standing values ± 10% and showed a steady state for at least 3 min.
Subjects were provided with real-time visual feedback to help them modulate the vertical displacement of their CM. A marker over the mid-sacrum was videotaped from a posterior view and projected onto a screen projector in front of the subject. The position of the marker was chosen according to the literature, which has shown that a sacral marker is accurate and sensitive to estimate the changes in vertical CM displacement (Thirunarayan et al. 1996). The camera was calibrated so that 1 cm displacement of the marker appeared as 10 cm on the screen. The only instruction given was to modulate the vertical displacement of the marker regardless of the strategy used. At the beginning of the practice session, the participants tried several techniques to achieve this goal. However, by the end of the practice session they all showed a similar strategy to modulate the marker displacement and similar changes in the mechanical, metabolic and EMG variables as reflected later by the relatively small standard deviations in the results.
Positive mechanical work and CM displacement over a walking stride
Segmental kinematics was measured with an Elite system (Elite V5, BTS, Italy) at a sampling rate of 100 Hz. Six infrared cameras measured the 3-D bilateral co-ordinates of reflective markers positioned on specific anatomical landmarks to compute angular displacements as described in detail by Davis et al. (1991). The internal work, Wint, i.e. the work required to move the limbs relative to the CM, was computed from kinematic data following the method described by Willems et al. (1995) and Detrembleur et al. (2003). The body was divided into seven rigid segments: head–arm–trunk (HAT), and both thighs, shanks and feet. The internal mechanical energy of the body segments corresponded to the sum of rotational and translational energies of these segments due to their movements relative to the CM. For each lower limb, the internal mechanical energy–time curves of the thigh, shank and foot were summed. The Wint of each lower limb and HAT segment were then calculated separately as the sum of the increments of the respective internal mechanical energy curves. Finally, the Wint during gait corresponded to the sum of the Wint done to move the lower limbs and the HAT segment and was expressed per kilogram body mass and per metre travelled.
Vertical, forward and lateral ground reaction forces were recorded by means of the strain gauges of the force measuring treadmill at a sampling rate of 100 Hz and were digitized synchronously with the Elite system. The external work, Wext, i.e. the work performed to lift and accelerate the CM relative to the surroundings, was computed following the method described in detail by Cavagna (1975) and Willems et al. (1995). The 3-D accelerations of the CM were computed from the vertical, lateral and forward components of the ground reaction forces and the mass of the subject. The mathematical integration of the 3-D accelerations gave the velocity changes of the CM in all three directions (Vv, Vf, Vl). From the instantaneous Vv, Vf and Vl and the body mass (M), we computed the instantaneous vertical (Ekv
= 1/2MVv2), forward (Ekf
= 1/2MVf2) and lateral (Ekl
= 1/2MVl2) kinetic energies of the CM. A second mathematical integration of Vv was performed to determine the vertical CM displacement (Sv) of the CM. The amplitude of vertical CM displacement (Av) was measured as the peak-to-peak amplitude on the Sv curve over a stride. A stride was selected for analysis only when the sum of the increments in Vv, Vf and Vl changes did not differ by more than 25% from the sum of the decrements, thus revealing a relatively constant average height and speed per stride (Cavagna, 1975). The instantaneous gravitational potential energy (Ep
=
MgSv) was computed from the vertical displacement (Sv) of the CM, the body mass and the gravity constant (g). The total external mechanical energy (Eext) of the CM was calculated as the sum of Ekf, Ekv, Ekl and Ep. The increments of the Ekf, Ekv, Ekl and Ep curves represented, respectively, the positive work (Wekf, Wekv, Wekl and Wep) necessary to accelerate the CM in the three directions and to lift the CM during a stride. Wext was obtained by summing the increments of Eext over a stride. Wekf, Wekv, Wekl, Wep and Wext were expressed per kilogram body mass and per distance travelled. The ‘recovery’, quantifying the percentage of mechanical energy saved via a pendulum-like exchange between gravitational potential energy and kinetic energy of the CM, i.e. an index reflecting the effectiveness of the pendulum-like mechanical mode of walking, was calculated as:
|
| (1) |
The total mass-specific muscular work per distance travelled (Wtot) was calculated as the sum of Wext and Wint and expressed in J kg–1 m–1.
Metabolic cost and efficiency of positive work production
Breath-by-breath rate of oxygen consumption, rate of carbon dioxide production and heart rate were measured with an ergospirometer (Quark b2, Cosmed, Italy). Values were automatically converted by a software program (Cosmed, Quark b2 win, v. 7.6.a) to standard temperature, pressure and dry oxygen consumption. After steady state was reached, metabolic data were collected. The mass-specific gross energy consumption rate (W kg–1) was obtained from the oxygen consumption rate using an energy equivalent of oxygen, taking into account the measured respiratory exchange ratio (RER) (i.e. the ratio between the rate of carbon dioxide production and the rate of oxygen consumption) (McArdle et al. 1996). Only trials with RER 
1.0 were retained and analyzed. The mass-specific net energy consumption rate (W kg–1) was calculated from the energy consumption attributed to the walking per se, i.e. the energy consumption rate while walking minus the energy consumption rate while standing. Finally, the mass-specific net cost of transport, Cnet (J kg–1 m–1), was calculated by dividing the net energy consumption by the walking speed in m s–1.
Efficiency of positive work production by the muscles and tendons during walking (
) was calculated as the ratio of Wtot to Cnet and expressed as a percentage (Cavagna & Kaneko, 1977).
Electromyographic activity
The EMG activity of the vastus lateralis (VL), biceps femoris (BF), tibialis anterior (TA) and medial gastrocnemius (MG) muscles in both lower limbs was recorded by a telemetry system (Telemg, BTS, Milan, Italy). Circular surface electrodes were used with a diameter of 1 cm and an interelectrode distance of 2 cm (Medi-Trace, Graphic Controls Corporation, NY, USA) after hair shaving, skin abrasion and application of alcohol and ether so that the interelectrode resistance was below 5 k
. For each muscle, a pair of recording electrodes were placed close to each other, parallel with the muscle fibres and close to the motor point (Perotto, 1994), and care was taken not to place the electrodes over the peripheries of the muscles to minimize EMG cross talk between muscles (Winter, 2005).
The EMG signal was digitized at 1000 Hz, full-wave rectified and filtered (bandwidth 25–300 Hz). The onset and cessation of muscle activity were both visually (Basmajian, 1979) and mathematically determined by computing the EMG threshold voltage as described in detail by Van Boxtel et al. (1993). By combining the visual and mathematical methods, we obtained the EMG packets (onset to cessation) for each muscle. The strides were normalized to 100% in time before averaging the EMG activity of each muscle (Winter, 1991). Finally, the index of co-contraction between the BF and VL muscles (BF – VL) and between the TA and MG muscles (TA – MG) was temporally quantified as the percentage of the gait stride during which these antagonistic muscles were simultaneously activated (Yang et al. 1998).
Statistical analysis
For each walking trial, data obtained during 10 strides were averaged and the mean values were used for statistical analysis. The effects of walking conditions (normal, flat and bouncy) and speed on metabolic, mechanical and EMG variables were tested using a two-way repeated measures analysis of variance (ANOVA) (SigmaStat version 2.0, SPSS Sciences Software GmbH, Erkrath, Germany) and Tukey's post hoc tests where appropriate. The assumption of homoscedasticity was verified for all the comparisons, and the significance level alpha was set at 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In normal walking, the amplitude of vertical CM displacement, Av, increased linearly with speed from about 0.015 ± 0.003 m (mean ± S.D.) at the lowest speed to 0.034 ± 0.004 m at the highest speed (Fig. 1A). In flat walking, the subjects were able to significantly decrease Av to a constant amplitude for all speeds (0.015 ± 0.002 m, P < 0.001) corresponding to the half of Av measured during normal walking at an intermediate speed of 4 km h–1. To walk flat, all the subjects adopted a flexed-legged pattern although no specific instructions were given regarding the strategy to use (Fig. 1B and C).
|
, from 27.91 ± 6.4% in normal to 14.34 ± 4.8 J kg–1 m–1 in flat walking (P < 0.001) (Fig. 3B).
|
|
|
|
In bouncy walking, the subjects were able to increase Av from an average of 0.024 ± 0.01 m in normal to 0.094 ± 0.02 m in bouncy walking (P < 0.001), corresponding to up to 8 times greater than normal at 1 km h–1 and 3.5 times greater than normal at 4 km h–1 (Fig. 1A). The relatively small standard deviations in the kinematics also show the limited variability in the strategy between the subjects (Fig. 1B and C).
Increasing Av led to a dramatic increase in Wep from 0.35 ± 0.04 J kg–1 m–1 in normal to 1.41 ± 0.6 J kg–1 m–1 in bouncy walking (P
= 0.003). It also increased Wek from 0.30 ± 0.1 to 0.46 ± 0.2 J kg–1 m–1 (P < 0.001) (Fig. 2). A dramatic increase also occurred in Wext from 0.28 ± 0.02 J kg–1 m–1 in normal to 1.21 ± 0.7 J kg–1 m–1 in bouncy walking (P
= 0.004) associated with a significant decrease in the recovery from 52.49 ± 16.1% in normal to 40.08 ± 21.3% in bouncy walking (P
= 0.002). On the other hand, Wint remained close to normal (P
= 0.121) (Fig. 3A). Consequently, a dramatic increase occurred in the total mechanical work, Wtot, particularly at the lowest speed from 0.55 ± 0.1 J kg–1 m–1 in normal to 1.51 ± 0.6 J kg–1 m–1 in bouncy walking (P
= 0.004) as well as in the net energy cost, Cnet, from 2.11 ± 0.4 J kg–1 m–1 in normal to 6.23 ± 3.6 J kg–1 m–1 in bouncy walking (P
= 0.004). However, the efficiency, , remained close to normal walking values (P
= 0.591) (Fig. 3B).
In bouncy walking, all muscles showed increased activation timing relative to the walking stride time (Table 1, Fig. 4). Muscle co-contraction timing also increased significantly in shank muscles from 17.78 ± 15.9% in normal to 43.81 ± 16.2% in bouncy walking (P = 0.001) but not in thigh muscles (P = 0.194).
Optimization of human walking
The optimal combination of Av and walking speed, which minimizes Cnet, was 3 cm for Av at 4 km h–1 corresponding to Av at the spontaneously adopted speed in normal human walking (Fig. 5). In bouncy walking, Cnet increased only a little with a quite large increase of Av from the optimum, particularly at high speeds, where in flat walking, Cnet was highly sensitive to reductions of Av. For instance, at around 4 km h–1 the subjects were able to reduce Av by nearly half (from about 0.025 to 0.015 m) which nearly doubled their Cnet (from about 1.7 to 3.2 J kg–1 m–1); however, they were able to increase Av up to 3–4 times higher (from about 0.025 to 0.088 m), but their Cnet only doubled (from about 1.7 to 4.2 J kg–1 m–1).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Relation to previous studies
Our study showed that the exceptionally flat walk is never optimal as is also the case with an exceptional increase in CM undulations, which confirms the prediction of Srinivasan and Ruina's computer optimization model that showed that normal walking is the most optimal (Srinivasan & Ruina, 2006). However, our study showed the physiological grounds of this optimization and revealed an additional physiological complexity to the simple mathematical model insofar as the conditions in which the muscles provided mechanical work efficiently differed with respect to CM displacement (Fig. 3). A part of our results is also in agreement with the recent human experiments done by Ortega & Farley (2005). In fact, they showed that subjects spontaneously adopted a flexed-legged pattern to walk flat, but contrarily to the long-lasting theory of Saunders (Saunders et al. 1953), flat walking costs more, even twice more, than normal walking although the external work necessary to move the CM remains normal. However, the physiological reasons that underlie this increased energy cost remained unknown because they assessed neither muscle efficiency, as they only measured the external work that is only a part of the total mechanical work, nor muscle activity to examine the role of co-contraction in increasing the energy cost. We have also examined the effect of the increase in the vertical CM displacement above normal and directly compared all three walking styles at a large range of speed, which enabled us to see the effect of the inherent vertical movement seen in normal human walking and to find that bouncy walking is even better than flat walking for muscles to provide efficient work. Thus, we were able to deduce that not only do humans move up and down in normal walking to save energy via a pendulum-like mechanism (Cavagna et al. 1976; Ortega & Farley, 2005) but also to make their muscles work efficiently.
Factors influencing the energy cost and muscle efficiency in flat and bouncy walking
An efficient locomotion is one where most of the metabolic energy input is transformed into mechanical work (Minetti, 2004). The efficiency in our study takes into account the efficiency of the positive work derived from the contractile engine and that derived from energy storage and release by muscles and tendons (Cavagna & Kaneko, 1977). Consequently, it mainly depends on the shortening velocity of the muscles, stretching of elastic musculotendon structures before contraction and frictional losses within the body (Cavagna, 1978; Heglund & Cavagna, 1985). Isolated muscle experiments in vitro showed that muscle efficiency is low when the velocity of contraction is too low or too high, whereas the maximal efficiency of unstretched human muscles is about 25% (Dickinson, 1929; Heglund & Cavagna, 1985). However, stretching muscles and tendons before work production increases the efficiency above 25% by providing positive work, free of cost, during the recoil of the previously stretched structures, which is taken as evidence of the pre-stretching mechanism in vivo (i.e. during human walking) (Cavagna & Kaneko, 1977; Heglund & Cavagna, 1985; Willems et al. 1995). Recent studies have indeed shown the in vivo spring-like behaviour of the stretched musculotendon structures such as the tendon in the calf muscles in particular (Fukunaga et al. 2001; Ishikawa et al. 2005). On the other hand, in vivo, frictional losses mainly occur within antagonistic muscles doing work on each other as measured during co-contraction, which would decrease the efficiency by using metabolic energy no apparent work produced (Cavagna, 1978; Heglund & Cavagna, 1985; Minetti, 2004).
In our study, muscle activation increased in both flat and bouncy walking as exemplified by the vastus lateralis being likely to counter an increased knee flexion in flat walking and to straighten the knee to increase the CM amplitude in bouncy walking by raising the CM further when it has to be at a maximum (Figs 1 and 4). The increase in muscle activation associated with an increase in muscle co-contraction may also have occurred because flat and bouncy walking were unfamiliar tasks. However, our study included two practice sessions that exceeded the recommended treadmill habituation time, and the constant strategy that subjects showed suggests that they had become relatively habituated with both walking types. Nonetheless, any remaining habituation effect would have influenced flat and bouncy walking in a similar way, as both had a similar habituation time. Unexpectedly, the increase in muscle co-contraction did not seem to be a determining factor influencing the muscle efficiency, as this latter variable remained normal in bouncy walking. This is similar to recent experiments in our laboratory that also showed that muscle efficiency may remain normal, for instance in hemiparetic spastic patients pedalling with one leg on an ergometer (Stoquart et al. 2005) or walking with a pathological increase in the vertical CM displacement (Detrembleur et al. 2003), although muscle co-contraction timing was increased. The less determinant role of muscle co-contraction timing may be due to the fact that, as measured at present, it does not necessarily reflect the amount of useless force, as EMG activity is not necessarily correlated to muscle tension during walking (Winter, 2005) and, consequently, is not directly correlated to the energy consumed by the muscles. However, to our knowledge, it is not yet possible to measure the force developed and the energy consumed by individual muscles in vivo to better explain the relation between the energy cost, the mechanical work and the muscular efficiency during exercise.
Nonetheless, other factors seem to have influenced muscle efficiency. In fact, in bouncy walking, the efficiency reached a maximum (35%) at intermediate speeds and then decreased with higher speeds, which is expected when the positive mechanical work is mainly derived from the transformation of chemical energy by the contractile machinery of muscles, as in cycling (Dickinson, 1929) and normal walking (Cavagna & Kaneko, 1977), according to the force–velocity relationship and the measurements of the ‘initial efficiency’ of muscle (Hill, 1964). Hence, in bouncy walking, the role played by the contractile machinery prevails over that of the elastic component, although this last element cannot be neglected. In flat walking, the efficiency increased continuously with speed, which is expected when the positive work is mainly due to the recoil of elastic structures and to a lesser extent to the active shortening of the contractile machinery, as in running (Cavagna & Kaneko, 1977), in accordance with the fact that the useful effect of stretching continuously increases with the speed of stretching (Cavagna et al. 1968). However, the low efficiency in flat walking, below 25%, suggests that energy losses other than co-contraction may also have contributed to mask this underlying elastic mechanism, such as the chemical energy spent during isometric contractions and that to overcome internal frictions (Willems et al. 1995), for which to our knowledge, no attempt has yet been made to precisely quantify their contribution during walking.
Interestingly, humans also walk flat to achieve higher walking speeds in competition, but contrary to what we have found at common walking speeds, flat walking is intriguingly more efficient and less costly than normal walking in competition walkers at high speeds above 6–7 km h–1. The efficiency in competition walking also shows continuous increase at high speed while the pendular recovery decreases steeply, as in our study, indicating a smaller utilization of the pendulum-like mechanism characteristic of walking and a greater elastic storage of mechanical energy (Cavagna & Franzetti, 1981). This suggests that athlete walkers may have discovered an efficient way to walk flat, which may be due to the fact that they are constrained by the race rules to keep their legs straight, while using a peculiar hip movement to flatten their CM displacement (Alexander, 1992), contrary to the spontaneously flexed-legged pattern adopted by common subjects to walk flat.
Significance to the evolution of human locomotion
Even if 7 million years or more separate us from our common ancestor with apes, we are still close to a flat CM displacement relative to our current ability to change this displacement (Fig. 5). Hence, a reduced vertical CM displacement in normal walking is indeed important to minimize energy cost but only to certain limits beyond which energy cost increases dramatically due to a sharp decrease in muscle efficiency. In other words, tracing human walking back to its origin is much more costly, although the track is much shorter.
Our experiments showed that an efficient straight-legged gait does not waste as much energy as a primitive flat and flexed-legged gait does. This was probably a primary objective of natural selection, which is likely to have been achieved if Lucy walked straight. This may explain in part why its minimum energy cost was estimated to be close to that of humans although it occurred at lower speeds (around 2 J kg–1 m–1 at 2 km h–1) (Sellers et al. 2005). This minimum would probably have been doubled for similar muscle work if Lucy had walked flexed and flat. However, Lucy may have showed a much larger energy cost with increasing speeds (Heglund & Schepens, 2003; Sellers et al. 2005). Thus, the stasis of Lucy's anatomy over 3 million years may be explained in part by this efficient gait that is not necessarily economical at high speeds for a species living in less-open habitats. The increase in leg length and modern anatomy seen in H. erectus, being probably the first hominid to leave Africa to feed its voracious larger brain (Leonard, 2002), would have made it possible to cover longer distances more economically (Steudel-Numbers & Tilkens, 2004; Wang et al. 2004). Hence, natural selection may have maximized muscle efficiency 3 million years before reducing energy costs at higher speed to prepare H. erectus for its legendary travel, which was probably vital for our large-brained species.
Future perspectives and other applications
Modern humans take advantage of both primitive flat and modern straight-legged walking styles. Apart from their ability to shortcut in a few seconds millions of years of evolution by walking smoothly while, for instance, carefully holding their brimful cup of boiling coffee, humans walk flat to achieve higher walking speeds in competition as mentioned above. On the other hand, recent backpack design generates electricity from our natural vertical CM displacement to power an increased number of portable devices (Rome et al. 2005). It is also interesting to note that a similar decrease in the amount of pendulum recovery is seen in reduced gravity such as on Mars (0.4 g) (Cavagna et al. 2000) to that in flat walking in our study. This is probably due to the fact that decreasing the vertical CM displacement by half would have had a similar effect on the potential energy as a similar decrease in the constant gravity g.
Finally, humanoid robots have gone through an evolution similar to that of humans. Although some robots walk flat while others move up and down, efficiency and energy consumption remain the main challenges. Some state-of-the-art robots use sophisticated motor control to walk smoothly while demonstrating appalling inefficiency with excessive energy cost, whereas recent robots rival human efficiency by relying on passive dynamic mechanisms with much simple active control of actuators (Collins et al. 2005), which echoes what we have found as crucial physiological determinants of optimal human walking.
In conclusion, our study showed that not only do humans move up and down in normal walking to save energy via a pendulum-like mechanism (Cavagna et al. 1976; Ortega & Farley, 2005) but also to make their muscles work efficiently. In fact, walking flat makes the muscles work in unfavourable conditions that waste energy. The experience of million of years of marriage between extreme solutions in search for optimal bipedalism, which seems to have managed the compromise between muscle work and efficiency and come to a halt halfway between flatness and bounce, may provide a lesson in walking physiology for future robots.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Alexander RMcN & Jayes AS (1978). Vertical movements in walking and running. J Zool 185, 27–40.
Basmajian JV (1979). Muscles Alive, edn 4. Williams & Wilkins, Baltimore.
Bramble DM & Lieberman DE (2004). Endurance running and the evolution of homo. Nature 432, 345–352.[CrossRef][Medline]
Cavagna GA (1975). Force platforms as ergometers. J Appl Physiol 39, 174–179.
Cavagna GA (1978). Aspects of efficiency and inefficiency of terrestrial locomotion. In International Series on Biomechanics, vol 2A, pp. 3–22. University Park Press, Baltimore.
Cavagna GA, Dusman B & Margaria R (1968). Positive work done by a previously stretched muscle. J Appl Physiol 24, 21–32.
Cavagna GA & Franzetti P (1981). Mechanics of competition walking. J Physiol 315, 243–251.
Cavagna GA & Kaneko M (1977). Mechanical work and efficiency in level walking and running. J Physiol 268, 467–481.
Cavagna GA, Thys H & Zamboni A (1976). The sources of external work in level walking and running. J Physiol 262, 639–657.
Cavagna GA, Willems PA & Heglund NC (2000). The role of gravity in human walking: pendular energy exchange, external work and optimal speed. J Physiol 528, 657–668.
Collins S, Ruina A, Tedrake R & Wisse M (2005). Efficient bipedal robots based on passive-dynamic walkers. Science 307, 1082–1085.
Davis RB III, Õunpuu S, Tyburski D & Gage JR (1991). A gait analysis data collection and reduction technique. Hum Mov Sci 10, 575–587.[CrossRef]
Detrembleur C, Dierick F, Stoquart G, Chantraine F & Lejeune T (2003). Energy cost, mechanical work, and efficiency of hemiparetic walking. Gait Posture 18, 47–55.[Medline]
Dickinson S (1929). The efficiency of bicycle-pedalling, as affected by speed and load. J Physiol 67, 242–255.
Dierick F, Penta M, Renaut D & Detrembleur C (2004). A force measuring treadmill in clinical gait analysis. Gait Posture 20, 299–303.[CrossRef][Medline]
Fukunaga T, Kubo K, Kawakami Y, Fukashiro S, Kanehisa H & Maganaris CN (2001). In vivo behaviour of human muscle tendon during walking. Proc R Soc Lond B Biol Sci 268, 229–233.[Medline]
Gibbons A (2002). In search of the first hominids. Science 295, 1214–1219.
Heglund NC & Cavagna GA (1985). Efficiency of vertebrate locomotory muscles. J Exp Biol 115, 283–292.
Heglund NC & Schepens B (2003). Ontogeny recapitulates phylogeny? Locomotion in children and other primitive hominids. In Vertebrate Biomechanics and Evolution, pp. 283–295. BIOS, Scientific Publishers, Oxford.
Hill AV (1964). The efficiency of mechanical power development during muscular shortening and its relation to load. Proc R Soc Lond B Biol Sci 159, 319–324.[Medline]
Ishikawa M, Komi PV, Grey MJ, Lepola V & Bruggemann G (2005). Muscle–tendon interaction and elastic energy usage in human walking. J Appl Physiol 99, 603–608.
Leonard WR (2002). Food for thought: dietary change was a driving force in human evolution. Sci Am 287, 106–115.[Medline]
McArdle WD, Katch FI & Katch VL (1996). Exercise Physiology: Energy, Nutrition, and Human Performance, edn 4. Williams & Wilkins, Baltimore.
Minetti AE (2004). Passive tools for enhancing muscle-driven motion and locomotion. J Exp Biol 207, 1265–1272.
Nagano A, Umberger B, Marzke MW & Gerritsen KG (2005). Neuromusculoskeletal computer modeling and simulation of upright, straight-legged, bipedal locomotion of Australopithecus afarensis (A.L. 288–1). Am J Phys Anthropol 126, 2–13.[CrossRef][Medline]
Ortega JD & Farley CT (2005). Minimizing center of mass vertical movement increases metabolic cost in walking. J Appl Physiol 99, 2099–2107.
Perotto AO (1994). Anatomical Guide for the Electromyographer. The Limbs and Trunk, 3rd edn. Charles C Thomas, Springfield.
Rome LC, Flynn L, Goldman EM & Yoo TD (2005). Generating electricity while walking with loads. Science 309, 1725–1728.
Saunders JB, Inman VT & Eberhart HD (1953). The major determinants in normal and pathological gait. J Bone Joint Surg Am 35, 543–558.
Schmitt D (2003). Insights into the evolution of human bipedalism from experimental studies of humans and other primates. J Exp Biol 206, 1437–1448.
Sellers WI, Cain GM, Wang W & Crompton RH (2005). Stride lengths, speed and energy costs in walking of Australopithecus afarensis: using evolutionary robotics to predict locomotion of early human ancestors. J R Soc Interface 2, 431–441.[CrossRef][Medline]
Srinivasan M & Ruina A (2006). Computer optimization of a minimal biped model discovers walking and running. Nature 439, 72–75.[CrossRef][Medline]
Stern J & Susman R (1983). The locomotor anatomy of Australopithecus afarensis. Am J Phys Anthropol 60, 279–317.[CrossRef][Medline]
Steudel-Numbers KL & Tilkens MJ (2004). The effect of lower limb length on the energetic cost of locomotion: implications for fossil hominins. J Hum Evol 47, 95–109.[CrossRef][Medline]
Stoquart GG, Detrembleur C, Nielens H & Lejeune TM (2005). Efficiency of work production by spastic muscles. Gait Posture 22, 331–337.[CrossRef][Medline]
Thirunarayan MA, Kerrigan DC, Rabuffetti M, Croce UD & Saini M (1996). Comparison of three methods for estimating vertical displacement of center of mass during level walking in patients. Gait Posture 4, 306–314.[CrossRef]
Van Boxtel GJ, Geraats LH, Van den Berg-Lenssen MM & Brunia CH (1993). Detection of EMG onset in ERP research. Psychophysiology 30, 405–412.[CrossRef][Medline]
Van de Putte M, Hagemeister N, St-Onge N, Parent G & de Guise JA (2006). Habituation to treadmill walking. Biomed Mater Eng 16, 43–52.[Medline]
Wall JC & Charteris J (1981). A kinematic study of long-term habituation to treadmill walking. Ergonomics 24, 531–542.[Medline]
Wang W, Crompton RH, Carey TS, Gunther MM, Li Y, Savage R & Sellers WI (2004). Comparison of inverse-dynamics musculo-skeletal models of AL 288-1 Australopithecus afarensis and KNM-WT 15000 Homo ergaster to modern humans, with implications for the evolution of bipedalism. J Hum Evol 47, 453–478.[CrossRef][Medline]
Ward CV (2002). Interpreting the posture and locomotion of Australopithecus afarensis: where do we stand? Am J Phys Anthropol 45, 185–215.
Willems PA, Cavagna GA & Heglund NC (1995). External, internal and total work in human locomotion. J Exp Biol 198, 379–393.[Medline]
Winter DA (1991). The Biomechanics and Motor Control of Human Gait: Normal, Elderly and Pathological, edn 2. University of Waterloo Press, Canada.
Winter DA (2005). Biomechanics and Motor Control of Human Movement, edn 3. John Wiley & Sons, New Jersey.
Yang JF, Stephens MJ & Vishram R (1998). Infant stepping: a method to study the sensory control of human walking. J Physiol 507, 927–937.
|
Acknowledgements |
|---|
Related Article
J. Physiol. 2007 582: 474.
This article has been cited by other articles:
|
|
 |
|
  R. M. Alexander Flat and bouncy walking J. Physiol., July 15, 2007; 582(2): 474 - 474. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
as a percentage of walking stride time in normal versus flat and bouncy walking
