See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8061148

Mechanics and energetics of swinging the human leg

Article  in  Journal of Experimental Biology · March 2005

DOI: 10.1242/jeb.01408 · Source: PubMed

CITATIONS READS
217 431

3 authors, including:

Max Donelan
Simon Fraser University
55 PUBLICATIONS   5,490 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Gait optimization and energetic cost View project

All content following this page was uploaded by Max Donelan on 22 January 2014.

The user has requested enhancement of the downloaded file.

The Journal of Experimental Biology 208, 439-445 439
Published by The Company of Biologists 2005
doi:10.1242/jeb.01408

Mechanics and energetics of swinging the human leg

Jiro Doke1,*, J. Maxwell Donelan2 and Arthur D. Kuo1
1
Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109-2125 USA and
2
Department of Physiology, University of Alberta, Edmonton, AB T6G 2H7 Canada
*Author for correspondence (e-mail: jdokeh@umich.edu)

Accepted 25 November 2004

Summary
We measured how much metabolic energy is expended (N=12) swung one leg at frequencies 0.5–1.1·Hz and fixed
to swing a human leg. A previous dynamical model of amplitude. Rate of mechanical work ranged from
walking predicted that increasing metabolic costs for 0.02–0.27·W·kg–1 over these frequencies. Net metabolic
walking with step length and step frequency trade-off rate for leg swinging (subtracting that for quiet standing)
against each other to determine the optimum step increased from 0.41–2.10·W·kg–1, approximately with the
combination at a given speed. Simple pendulum dynamics fourth power of frequency (R2=0.92) and in proportion to
indicate that the cost of walking at high step frequencies a hypothesized cost of force production for short
could be associated with driving the legs back and forth durations. The costs of producing force and work could
relative to the body, at a rate increasing approximately account for the increase. In a crude comparison, moving
with the fourth power of frequency, possibly due to the the legs back and forth at a typical stride frequency of
low economy of producing muscle force for short 0.9·Hz, might consume about one-third of the net energy
durations. A similar cost would be expected for isolated (2.8±0.8·W·kg–1) needed for walking at 1.3·m·s–1.
swinging of a leg at faster than its natural frequency. We
constructed an apparatus to measure work performed on Key words: metabolic energy, locomotion, biomechanics, muscle,
the leg, and measured metabolic cost as human subjects force production.

Introduction
Swinging of the leg is an important part of human estimated that 26% of the energy used by the limbs was for
locomotion. The back-and-forth motion of the legs has been the swing phase.
likened to that of a pendulum, and evidence suggests that The cost of driving the legs back and forth relative to the
pendular dynamics may be responsible for much of the swing body may be significant during human walking. Griffin et al.
phase of gait (Mochon and McMahon, 1980). A pendulum can (2003) proposed that little energy is needed to move the legs
move at its natural frequency with minimal energy input, but during walking because human energy expenditure increases
mechanical force and work requirements increase sharply with in proportion to carried load. But again, expenditure increases
frequency of motion. Swinging the leg at high frequencies several times more when the limbs are loaded (e.g. Soule and
might therefore cost metabolic energy. The metabolic cost of Goldman, 1969) as opposed to the trunk, therefore implying
walking also increases with step frequency, and this increase the opposite. Another argument for a cost for moving the legs
might partially be explained by the cost of moving the legs. is derived from a mathematical model of walking (Kuo, 2001).
Here we test whether fast leg swinging is metabolically costly. We hypothesized that mechanical work must be performed to
The metabolic cost of active leg motion has previously been redirect the center of mass from the pendular arc dictated by
considered mostly in running animals. Taylor et al. (1980) the stance leg, especially in the transition from one step to the
argued that the cost of running is dominated by the generation next (Kuo, 2002). Metabolic energy is needed to perform this
of force to support body weight for short durations of ground work (Donelan et al., 2002a,b), which itself could be
contact (Kram and Taylor, 1990), rather than to move the legs minimized by taking short steps. But the actual metabolic
(Taylor, 1994). But loading the limbs of running humans minimum for a given speed occurs at a longer step length,
(Myers and Steudel, 1985) or dogs (Steudel, 1990) causes indicating a separate cost for short but fast steps (Kuo, 2001).
metabolic cost to increase more than for the same loads To account for this trade-off, our model required a metabolic
applied to the center of mass, suggesting that moving the cost for walking at high step frequencies increasing roughly
limbs does require energy. Marsh et al. (2004) used more with the fourth power of step frequency. The force and work
direct measurements of blood flow to estimate energy needed to move the legs relative to the body might explain this
expenditure in the swing-phase muscles of guinea fowl. They proposed cost of high step frequencies. In particular, a cost for

THE JOURNAL OF EXPERIMENTAL BIOLOGY

440 J. Doke, J. M. Donelan and A. D. Kuo
producing force over short durations – applied to fast leg T(t) = A (–ω2 + ω2n) cos ωt . (3)
motions and called the ‘force/time hypothesis’ here – could
The amplitude, T0, therefore behaves according to
potentially account for the trade-off against high step
frequencies. Regardless of the specific dependency on T0
| ω2 – ω2n | . (4)
magnitude or duration of work or force, our model predicted a
The most obvious potential metabolic cost is for performing
substantial cost for moving the legs relative to the body,
work on the limb, increasing with the cube of swing frequency.
associated with neither center of mass redirection nor body
The mean rate of work performed on the pendulum is
weight support.
A simple approach for estimating this cost is to study leg 2ω π 2ω
W= T ⋅ θ dt
swinging itself without walking. The loading conditions must π 0

necessarily differ between isolated leg swinging and actual Each cycle of pendulum swing involves both positive and
walking, but the range of hip torques and leg angles can negative work, both exacting a positive metabolic cost (Hill,
roughly be matched. The metabolic cost of leg swinging could 1938). Substituting Equations (2) and (3) into (5), the mean
potentially depend on the work produced by muscles, and rate of positive work, W (+), for the pendulum model is
possibly even the force/time cost predicted for walking. But A2
regardless of the particular cause, both of these possibilities W (+) = ω ω 2 − ω2n
π
predict a sharply increasing metabolic cost for swinging at
higher than the leg’s natural frequency. with an equal magnitude of negative work.
The purpose of the present study was to measure the Another possible contributor to metabolic cost is the
mechanics and metabolic energetics of swinging the human leg force/time cost derived from our model of walking, increasing
by itself. We tested how much metabolic rate increases with with the fourth power of swing frequency. This exponent was
frequency of swinging and considered possible contributions required to explain the preferred step length vs speed
to the cost of moving the legs. relationship. In our formulation of the force/time hypothesis,
when muscle force is produced in bursts, metabolic cost
increases with force but with an economy inversely
Materials and methods proportional to burst duration time (Kuo, 2001). The force in
We constructed a simple apparatus for measuring torque and question is proportional to hip torque amplitude, and the burst
displacement of a single swinging leg. We collected data from duration is proportional to swing period τ = 1/f. The proposed
human subjects at a variety of frequencies above the natural metabolic cost, labeled ‘rate of force/time’ or Fτ, is
pendular frequency, and at constant amplitude. Potential T0
Fτ ∝ ⋅f
contributors to metabolic cost include rate of work and the τ
force/time hypothesis. Before describing the experiments
Substituting Equations (2)–(4) into (7) yields the model
themselves, we use a simple pendulum model to quantify the prediction
predictions arising from these possible contributors.
Fτ
| ω4 – ω2  ω2n | . (8)
Model
This fourth power prediction only applies to fast motions above
A simple pendulum model (Fig. 1) illustrates the mechanics the leg’s natural frequency, because of the high force
of moving the leg. Employing a small angle approximation, requirements and short durations. For slower motions, work
and measuring the angular displacement θ from vertical, the will likely dominate metabolic cost.
equations of motion are The mechanics and metabolic energetics of isolated leg
θ + ω n2 θ = Τ (1) swinging can be measured experimentally. It is straightforward
to compare leg-swinging mechanics against the pendulum
where ω n is the pendular natural frequency and T is the model, but it is more difficult to differentiate between
moment of applied muscle force normalized by leg inertia. contributions to metabolic cost, because metabolic rate may
∆
Natural frequency ω n (in rad·s–1; in Hz, fn =ω n /2π) depends depend not only on work and force but also other factors not
mainly on inertial properties, such as the location of the leg considered. Rather than attempting to test for one potential cost
center of mass, but may also be affected by parallel elastic against another, we tested for an increase in metabolic rate
elements, such as from passive tissue compliance about the hip. proportional to the larger (fourth power) component in a
The pendulum is assumed to be driven approximately combined cost including both rate of work and rate of
∆
sinusoidally with fixed amplitude A and frequency ω=2πf, force/time.
θ(t) = A cos ωt . (2) Experimental procedure
Active movement of the leg requires muscle force or torque, We measured mechanical and metabolic costs of swinging
increasing with the square of swing frequency. Combining (1) the leg at different frequencies in twelve young adults. All
and (2), the torque is subjects (six males, six females; body mass M=64.8±8.3·kg,

THE JOURNAL OF EXPERIMENTAL BIOLOGY

 Mechanics and energetics of swinging the human leg 441

T
Leg angle θ

40°
Support
frame
Optical
encoder Leg torque
Fig.·1. Isolated leg swinging was modeled as a simple pendulum. Leg
angle θ was defined relative to vertical, and torque T due to muscle
Splint
force was defined as positive in the same direction as θ. We assumed

50 Nm
a relatively constant moment arm for muscle force. Torque and rate
of work requirements increase with the square and cube, respectively, θ
of swing frequency f (Hz) or ω (rad·s–1) above the natural frequency,
fn or ωn respectively. 1s
Force
leg length l=0.88±0.07·m, mean ± S.D.) were healthy and plate
exhibited no clinical gait abnormalities. They gave their
informed consent to participate in this study prior to the Fig.·2. Experimental apparatus. Subjects performed leg swinging
while attached to a rigid frame, with weight supported by both arms
experiment. Seven different swing frequencies, f, ranging from
and one leg. Subjects were strapped to the metal frame, with leg angle
approximately 0.5–1.1·Hz, were tested on each subject’s left θ measured by optical encoder. Force plate underneath the frame
leg, all with a peak-to-peak amplitude 2A of approximately measured ground reaction forces, used to compute leg torque
45°. The trial order was randomized to reduce fatigue effects. produced at hip (representative data shown).
We constructed an apparatus for estimating work performed
on the limb through measured leg displacement and reaction the torque and work performed on the leg. The average rate of
forces (see Fig.·2), effectively acting as an ergometer. This positive mechanical work performed on the swing leg, W (+),
apparatus consisted of a metal frame, to support the upper was found by integrating the half-wave rectified mechanical
body, mounted atop a force platform. Subjects stood inside the power (product of T and θ), and dividing by the entire data
frame with one leg on a raised block so that the other leg could collection time for the trial. An equal magnitude of negative
swing freely. Their upper bodies were strapped securely to the mechanical work was also performed, because there was zero
frame, with their weight distributed between one leg and two net work performed on the leg. We also computed the
armrests. A lightweight knee splint was used to keep the amplitude of the first fundamental component of the hip torque,
swinging leg straight. An optical encoder, rotating about an T0, after first applying a low-pass filter with a cut-off frequency
axis through the hip and attached to the swinging leg, was used of twice the swing frequency. The force/time was calculated
to measure the leg angle with respect to vertical. This angle from Equation (7).
was also displayed to the subject on a computer screen, along We measured the metabolic cost of swinging the leg using
with visual targets showing 45° peak-to-peak amplitude. an open circuit respirometry system (VMax29, SensorMedics
Subjects were asked to swing the leg between the targets, in Corp., Yorba Linda, CA, USA). Each trial was 6·min long,
time with a metronome set to twice the swing frequency, so with the first three minutes used to allow oxygen consumption
that there was an audible cue for swinging of the leg in each to reach steady state, followed by 3·min of data collection. The
direction. Subjects typically learned to follow the metronome first trial for each subject was used to measure the resting
after a few minutes’ practice. metabolic rate with no leg motion, with the subject standing
Reaction forces were measured using a force platform quietly on one leg while attached to the ergometer frame. This
(AMTI Biomechanics Platform Model OR6-5; Watertown, resting rate was used as a baseline that was subtracted from
MA, USA) underneath the ergometer frame. The platform each subsequent data set to yield net metabolic rate. Trials 2–8
collected forces and moments simultaneously at a sampling were leg swinging trials conducted in random order, with a
rate of 120·Hz. Assuming that the swing leg alone was in short resting period given between each. Finally, we conducted
motion, the ground reaction forces were equal to the hip a walking trial, in which the subject walked at 1.3·m·s–1 on a
reaction forces. We used anthropometric measurements and motorized treadmill.
regression equations (Yeadon and Morlock, 1989) to determine Net metabolic rate, E, was computed from the average rate
the distance from the hip to the leg center of mass, r, as well of oxygen consumption, subtracting the rate for quiet standing.
as the leg moment of inertia. These were used, with measured We assumed a rate of 20.9·W for 1·ml·s–1 O2. Although we did
leg kinematics, in inverse dynamics equations (Kuo, 1998) for not measure lactate concentration, we did monitor the
a single rigid leg to calculate the hip reaction torque, T. respiratory exchange ratio (RER). In all trials, RER was <0.9,
We used the kinematic and reaction force data to calculate indicating that the exercise was primarily aerobic.

THE JOURNAL OF EXPERIMENTAL BIOLOGY

442 J. Doke, J. M. Donelan and A. D. Kuo
To account for differences in subjects’ body size, we of positive mechanical work (see Fig.·3). Hip torque amplitude,
performed our analysis using dimensionless variables, with M, T0, increased approximately with the square of swing
g, and l as base units. For example, energy rates W (+) and E frequency (Fig.·3a), fitting well with Equation (9), R2=0.96.
were made dimensionless by the factor A2Mg1.5l 0.5, and The fitted model had a minimum equivalent to 1.47·N–m,
force/time Fτ by AMg1.5l 0.5. For convenience of data occurring at a frequency of fn=0.64·Hz, which was not
presentation, frequencies are presented in units of Hz, and significantly different from the natural frequency derived from
energy rates additionally in the more commonly used units of leg inertia properties, 0.64±0.02·Hz (S.D.) (P=0.51, t-test). Rate
W·kg–1, converted back by an appropriate mean non- of positive mechanical work performed at the hip, W (+), also
dimensionalizing factor. increased sharply but with the cube of swing frequency
We first tested the degree to which isolated leg swinging (Fig.·3b) as in Equation (10), R2=0.93. The minimum W (+)
resembled forced motion of a pendulum. We used regression equivalent to 0.02·W·kg–1 also occurred at the natural
tests to compare measured T0, W (+) and Fτ with the amounts frequency, rising to a maximum of 0.27·W·kg–1 at the highest
given by Equations (4), (6) and (8), respectively. In terms of frequency. The regression coefficients for T0, converted to
frequencies in Hz, the regressions were dimensional units, were CT=53.63±2.36·Nm·s–2 (95%
confidence interval, CI), DT=0.47±0.94·Nm (CI). The
T0 = CT  | f 2 – fn2 | + DT (9)
coefficients for W (+) were CW=0.31±0.02·W·s–3·kg–1 (95%
confidence interval, CI), DW=0.02±0.01·W·kg–1 (CI).
W(+) = CW  | f 3 – f  fn2 | + DW (10)
The rate of force/time Fτ, increased with fourth power of
swing frequency (Fig.·3c), fitting well with Equation (11),
Fτ = CF  | f 4 – f 2  fn2 | + DF , (11)
R2=0.95. The dimensionless regression coefficients were
each with a constant of proportionality C, and a constant offset CF=1.52±0.07 (CI), DF=0.001±0.002 (CI).
D. The offset is ideally zero; for example in Equation (6), Examining the metabolic rate, we found it increased
theoretically, no work is needed to swing a rigid leg at its substantially with swing frequency (Fig.·4a), fitting well with
natural frequency. However, in practice, we found that subjects the force/time hypothesis (Equation 12), R2=0.92. The data
performed some work at all frequencies, perhaps because of ranged from 0.41±0.26·W·kg–1 (S.D.) to 2.10±0.31·W·kg–1. The
energetic losses in the leg and experimental apparatus, and regression coefficients were CE=2.04±0.23·W·s–4·kg–1 (CI),
because some control is needed to follow the beat of the with offset DE=0.30±0.22·W·kg–1 (S.D.). The second regression
metronome, probably resulting in force and/or work related to demonstrated that E increased approximately linearly with Fτ
small corrective actions. We therefore included the possibility (Fig.·4b), R2=0.85. The coefficients were CEF=6.62±0.74 (CI),
of a non-zero offset in Equations (9)–(11). with offset DEF=0.015±0.005 (S.D.).
We next compared the measured net metabolic rate E against When walking at 1.3·m·s–1, subjects chose a stride frequency
the force/time hypothesis using two regressions. The first was of about 0.9·Hz and consumed metabolic energy at net rate
of the form 2.8±0.8·W·kg–1. Work loops for a typical subject showed a
range of hip torques and angles for comparison with walking
E = CE  | f 4 – f 2  fn2 | + DE (12)
(Fig.·5).
to determine how metabolic cost increased as a function of
swing frequency. In this test, we found that subjects exhibited
varying offsets in their metabolic rate. We therefore allowed Discussion
for different individual offsets in the overall regression for We sought to determine the degree to which isolated
Equation (12). The second regression was a linear fit between swinging of a leg is metabolically costly. The mechanics of a
E and Fτ: pendulum predict that sharply increasing force and work are
required to move the leg quickly. Production of muscle work
E = CEF  Fτ + DEF . (13)
and force for short durations (force/time hypothesis) might
Because the force/time hypothesis only applies to fast both require metabolic energy.
swinging of the leg, these tests were performed only for The observed increases in mechanical (Fig.·3) and metabolic
frequencies above 0.7·Hz. (Fig.·4) measurements confirm that pendulum mechanics are
We also compared leg swinging with walking. In terms of relevant to the leg motion. Even though no net work is
mechanics, we compared a typical subject’s hip work loops at performed over an entire cycle, force and work are necessary
all swing frequencies against published data for normal within each cycle to move the leg faster than its natural
walking at 1.3·m·s–1 (Whittle, 1996). Metabolic rates were frequency. The sharp increase in mechanical demands must be
compared between leg swinging and walking at 1.3·m·s–1. met by muscle, probably causing the fourfold increase in
metabolic cost.
Mechanical work is probably responsible for much of the
Results metabolic cost. Some active work is almost certainly
The mechanics of leg swinging agree well with the performed by muscle fibers, but not necessarily as much as the
pendulum model, in terms of both peak-to-peak torque and rate measured work performed on the limb. Series elasticity, in

THE JOURNAL OF EXPERIMENTAL BIOLOGY

 Mechanics and energetics of swinging the human leg 443
tendon and other structures, can also perform work on the limb actively. In addition, the highest forces occur at the extremes
passively (Kuo, 2001), thus lowering the proportion performed of leg motion when speed is lowest, so that muscle fibers can
be nearly isometric. The actual muscle fiber work during leg
50
swinging and its true contribution to metabolic cost are
A Hip torque amplitude, T0
0.08 unknown here.
Experimental data Metabolic cost may also increase with short durations of
40

Dimensionless torque
Theoretical model force production, as indicated by the observed correlation to
0.06 Fτ, consistent with the force/time hypothesis. Others have
Torque (Nm)

30 Model fit previously observed increases in metabolic rate with short

Eqn. (9)
0.04 force durations in situations as disparate as supporting body
20 weight in running (Roberts et al., 1998) and propelling the
Natural
frequency body in cross-country skiing (Bellizzi et al., 1998), although
10 fn 0.02 with varying formulations. The cost might be attributed to
increasing recruitment of less economic fast-twitch muscle
R2=0.96 fibers for short durations of force production (Kram and
0
0.5 0.6 0.7 0.8 0.9 1.0 1.1 Taylor, 1990), but this explanation alone is probably

B Rate of mechanical mork, W(+)

0.3 Dimensionless rate of work

2.5 A Metabolic rate vs frequency

Rate of work (W kg–1)

0.01 10
Experimental data
0.2 2.0 Theoretical model
Model fit 8
Eqn. (10) Model fit
1.5 Eqn. (12)
6
0.1
1.0

Dimensionless metabolic power (×10–2)

4
R2=0.93
Net metabolic power (W kg–1)

0 0 0.5 2
0.5 0.6 0.7 0.8 0.9 1.0 1.1
R2=0.92
0 0
Rate of force/time, Fτ 0.5 0.6 0.7 0.8 0.9 1.0 1.1
C
12 Frequency (Hz)
Dimensionless force/time

10 2.5 B Metabolic rate vs force/time

10
8
2.0
Model fit 8
6 Eqn. (11)
1.5
4 Model fit 6
Eqn. (13)
2 1.0 4
R2=0.95
0 0.5
0.5 0.6 0.7 0.8 0.9 1 1.1 2
Frequency (Hz)
0 R2=0.85 0
Fig.·3. Mechanics of leg swinging as a function of frequency f, in 0 2 4 6 8 10 12 14
terms of torque, work, and force/time, were modeled reasonably well Dimensionless force/time (×10–3)
by a forced pendulum (Equations 9–11). (a.) Hip torque amplitude,
T0, increased approximately with f 2 above natural frequency fn Fig.·4. Metabolic rate increased over fourfold with frequency of
(R2=0.96). (b.) Rate of mechanical work, W (+), increased isolated leg swinging, for motion faster than the leg’s natural
approximately with f 3 (R2=0.93). (c.) Rate of force/time, Fτ, increased frequency. (A) E vs frequency of leg swinging f, showing metabolic
approximately with f 4 (R2=0.95). Metabolic cost is hypothesized to rate increasing approximately with f 4 as predicted by force/time
increase with both rate of work and force/time, for frequencies above hypothesis (Equation 12). Data shown are for all frequencies applied,
the natural frequency fn=0.64·Hz. Data fits were performed using but the curve fit was only performed on data for fast leg swinging.
dimensionless variables (right-hand axis) with body mass, (B) E vs Fτ, showing metabolic rate increasing approximately linearly
gravitational constant, and leg length serving as base units; with the hypothesized force/time cost (Equation 13). Rate of
conventional units are shown (left-hand axis) for convenience. Data mechanical work is also likely to contribute, but cannot accurately be
shown are mean ± S.D. distinguished from force/time in overall metabolic cost.

THE JOURNAL OF EXPERIMENTAL BIOLOGY

444 J. Doke, J. M. Donelan and A. D. Kuo
insufficient to explain the fourfold increase in cost seen here. degree of freedom to the fit of Equation (12) without
Isolated muscle measurements also demonstrate a substantial explaining the many other factors that might affect metabolic
increase in energetic cost of producing force for shorter cost. Further differentiation of costs might require experiments
durations, even when fiber type recruitment is relatively fixed that manipulate force and duration as separate independent
(Hogan et al., 1998). A possible explanation is the energetic variables.
cost of activation–deactivation dynamics, in particular The conditions of isolated leg swinging were only roughly
calcium pumping associated with the sarcoplasmic reticulum comparable to human walking in terms of hip torque and
(Bergstrom and Hultman, 1988; Hogan et al., 1998; Verburg amplitude, but with a lower frequency. Our experiment sought
et al., 2001). to separate the problem of moving the legs relative to the body
Our results suggest that metabolic rate depends on work and from that of loading and propelling the center of mass. But
force/time, but cannot distinguish between their relative lacking the actual loads of walking, isolated leg swinging
contributions. The fit to the larger fourth-power cost (Fig.·4a) cannot simultaneously match both the forces and kinematics of
means that the force/time hypothesis cannot be excluded. But walking. At f=1.08·Hz, the ranges of hip torques and angles are
neither can the contribution of rate of work be excluded. In the roughly similar to that of normal walking at 1.3·m·s–1 (Fig.·5),
absence of a hypothesis for how these two costs might sum – but not to the stride frequency of about 0.9·Hz. Swing and
most likely in a nonlinear fashion – it would not be fruitful to stride frequency could alternatively be matched, f=0.9·Hz, but
add another term for work. Such a term would only add a third the torques would then be considerably lower than for walking.
By using similar torque and amplitude but lower frequency,
our conditions conservatively involved less work than that
A Hip joint work loops performed on the leg during walking.
–60
Flex

We also used a splint to restrict greatly the knee motion that

0.58 Hz
would normally occur in walking. If the knee were instead
–40 0.75 Hz
0.83 Hz
allowed to bend, the leg could theoretically be moved at higher
0.96 Hz frequency for the same hip torque. But as noted above, the
–20
Hip torque, T

1.08 Hz same hip torque at a higher speed would not be expected to

require less metabolic power, and if active torque were
0 produced at the knee it might also exact a metabolic cost.
Allowing knee motion could alternatively enable the same
20 Faster
swinging swing frequency at lower hip torque and, therefore, lower rate
of work and lower metabolic cost. But our intent was to
40 approximate the hip torques used in walking, and doing so with
Ext

a knee splint is likely to produce an underestimate of the cost

60 of moving the leg during walking because of the lower speed.
A more significant limitation of this experiment was that
B Work loop compared with walking
subjects exerted considerable effort to hold their bodies
Flex

–60
immobile during leg swinging, despite our efforts to strap their
–40 bodies to the rigid measurement apparatus. In walking,
Leg swing reciprocal leg motion allows reaction forces to be produced
1.08 Hz
–20 between the legs and against each other rather than an external
Hip torque, T

frame. In terms of reaction forces, isolated swinging of a single

0 leg might be roughly comparable to moving both legs during
Walking
0.9 Hz walking.
20 The most conservative walking comparison is, therefore,
with the condition of isolated leg swinging at f=0.9·Hz, which
40 would cost approximately 0.95·W·kg–1, or about one-third of
Ext

the rate for walking at the same stride frequency. It is

60 conservative because it likely underestimates the hip torque
–30 –20 –10 0 10 20 30 and work needed to move both legs in walking, it does not
Flex Hip angle, θ Ext include possible costs of moving the knee, and it assumes that
the reaction forces for swinging a single leg are similar to those
Fig.·5. Work loops of hip torque vs angle for a typical subject: (A) as
a function of frequency; and (B) compared with normal walking. (A)
for moving both legs during walking.
Work loops varied mostly in terms of torque rather than amplitude or In actual walking, it is difficult to experimentally isolate
contained area. (B) Compared with normal walking at 1.3·m·s–1 (data moving the legs from supporting body weight or redirecting
from Whittle, 1996), isolated leg swinging occurred at a comparable the center of mass. When a leg is in contact with the ground,
range of torques and angles, although walking occurs at a lower stride it is somewhat arbitrary to assign muscle activity to one
frequency of 0.9·Hz and with more work at the hip. function or another. But there might be a significant metabolic

THE JOURNAL OF EXPERIMENTAL BIOLOGY

 Mechanics and energetics of swinging the human leg 445
cost even if moving the legs were narrowed to the swing phase Bergstrom, M. and Hultman, E. (1988). Energy cost and fatigue during
alone, as evidenced by the findings of Marsh et al. (2004). intermittent electrical stimulation of human skeletal muscle. J. Appl.
Physiol. 65, 1500-1505.
Swing-phase activity of guinea fowl hip muscles accounted for Donelan, J. M., Kram, R. and Kuo, A. D. (2002a). Mechanical work for
about one-fourth of metabolic cost even at slow speed of step-to-step transitions is a major determinant of the metabolic cost of
0.5·m·s–1. Assuming a hip height of 0.2·m and a stride human walking. J. Exp. Biol. 205, 3717-3727.
Donelan, J. M., Kram, R. and Kuo, A. D. (2002b). Simultaneous positive
frequency of 1.5·Hz (Gatesy and Biewener, 1991), the one- and negative external mechanical work in human walking. J. Biomech. 35,
fourth estimate applies to a slightly slower dimensionless 117-124.
speed and stride frequency than humans (speed 0.36 vs 0.44, Fedak, M. A., Heglund, N. C. and Taylor, C. R. (1982). Energetics and
mechanics of terrestrial locomotion. II. Kinetic energy changes of the limbs
stride frequency 0.21 vs 0.27, guinea fowl vs human, and body as a function of speed and body size in birds and mammals. J.
respectively). Legs of guinea fowl are also a slightly lower Exp. Biol. 97, 23-40.
proportion of body mass than in humans, approximately 10% Gatesy, S. M. and Biewener, A. A. (1991). Bipedal locomotion: Effects of
speed, size and limb posture in birds and humans. J. Zool. Lond. 224, 127-
vs 16% per leg (Fedak et al., 1982). A comparable cost for the 147.
human swing phase alone appears reasonable, both in relation Griffin, T. M., Roberts, T. J. and Kram, R. (2003). Metabolic cost of
to these data, and to our one-third estimate for moving the legs generating muscular force in human walking: insights from load-carrying
and speed experiments. J. Appl. Physiol. 95, 172-183.
relative to the body including both swing and stance phases. Hill, A. V. (1938). The heat of shortening and the dynamic constants of
We do not interpret the one-third cost estimate to conflict muscle. Proc. R. Soc. Lond. B 126, 136-195.
with Mochon and McMahon’s (1980) notion that much of the Hogan, M. C., Ingham, E. and Kurdak, S. S. (1998). Contraction duration
affects metabolic energy cost and fatigue in skeletal muscle. Am. J. Physiol.
swing phase is ballistic. These authors noted that muscle 274, E397-E402.
activity occurs mostly at the start and end of swing, with the Kram, R. and Taylor, C. R. (1990). Energetics of running: a new perspective.
motion being largely passive in between. For fast leg motions, Nature 346, 265-267.
Kuo, A. D. (1998). A least-squares estimation approach to improving the
the highest torques are needed only at the extremes of motion, precision of inverse dynamics computations. J. Biomech. Eng. 120, 148-
and there could be substantial intervals of time that need 159.
Kuo, A. D. (2001). A simple model of bipedal walking predicts the preferred
minimal muscle activity (Kuo, 2002). Even with a high speed-step length relationship. J. Biomech. Eng. 123, 264-269.
metabolic cost, the swing phase appears to be governed by Kuo, A. D. (2002). Energetics of actively powered locomotion using the
pendulum dynamics. simplest walking model. J. Biomech. Eng. 124, 113-120.
Marsh, R. L., Ellerby, D. J., Carr, J. A., Henry, H. T. and Buchanan, C.
We originally hypothesized a cost for moving the legs to I. (2004). Partitioning the energetics of walking and running: swinging the
explain why humans do not walk with short steps to minimize limbs is expensive. Science 303, 80-83.
the costs of transitioning between inverted pendulum arcs Mochon, S. and McMahon, T. A. (1980). Ballistic walking. J. Biomech. 13,
49-57.
(Kuo, 2001). Large increases in metabolic cost for leg Myers, M. J. and Steudel, K. (1985). Effect of limb mass and its distribution
swinging may be sufficient to explain the increasing cost of on the energetic cost of running. J. Exp. Biol. 116, 363-373.
walking with step frequency. The approximate fourth power Roberts, T. J., Kram, R., Weyand, P. G. and Taylor, C. R. (1998).
Energetics of bipedal running. I. Metabolic cost of generating force. J. Exp.
exponent of frequency is also consistent with the optimum Biol. 201, 2745-2751.
combination of step length and frequency preferred by humans. Soule, R. G. and Goldman, R. F. (1969). Energy cost of loads carried on the
Low economy of force production for short durations is a head, hands, or feet. J. Appl. Physiol. 27, 687-690.
Steudel, K. (1990). The work and energetic cost of locomotion. II. Partitioning
potential explanation for this cause, but more detailed the cost of internal and external work within a species. J. Exp. Biol. 154,
physiological studies would be necessary to test this further. 287-303.
Taylor, C. R. (1994). Relating mechanics and energetics during exercise. In
Comparative Vertebrate Exercise Physiology: Unifying Physiological
This research was supported in part by an NSERC Principles, vol. 38A, pp. 181-215. New York: Academic Press.
fellowship to J.M.D., and NIH grant DC0231201 to A.D.K. Taylor, C. R., Heglund, N. C., McMahon, T. A. and Looney, T. R. (1980).
The authors thank R. Kram for assistance in pilot studies for Energetic cost of generating muscular force during running: a comparison
of large and small animals. J. Exp. Biol. 86, 9-18.
this experiment, and N. B. Alexander for assistance with Verburg, E., Thorud, H. M., Eriksen, M., Vollestad, N. K. and Sejersted,
laboratory facilities. O. M. (2001). Muscle contractile properties during intermittent nontetanic
stimulation in rat skeletal muscle. Am. J. Physiol. Reg. Integr. Comp.
Physiol. 281, R1952-R1965.
Whittle, M. W. (1996). Gait Analysis: An Introduction. Oxford, UK:
References Butterworth-Heinemann.
Bellizzi, M. J., King, K. A., Cushman, S. K. and Weyand, P. G. (1998). Yeadon, M. R. and Morlock, M. (1989). The appropriate use of regression
Does the application of ground force set the energetic cost of cross-country equations for the estimation of segmental inertia parameters. J. Biomech.
skiing? J. Appl. Physiol. 85, 1736-1743. 22, 683-689.

THE JOURNAL OF EXPERIMENTAL BIOLOGY

View publication stats