Vilas-Boas, Machado, Kim, Veloso (eds.)
Biomechanics in Sports 29
Portuguese Journal of Sport Sciences
11 (Suppl. 2), 2011
LOWER-LIMB BIOMECHANICAL ASYMMETRY IN MAXIMAL VELOCITY SPRINT
RUNNING
Timothy Exell, David Kerwin, Gareth Irwin and Marianne Gittoes
Cardiff School of Sport, University of Wales Institute, Cardiff, United Kingdom
Asymmetry analyses have provided valuable insight into submaximal running and
walking gait. Knowledge of asymmetry in sprint running is limited due to traditional
unilateral methods of data collection. The aims of the study were to develop asymmetry
measures that included intra-limb variability and to investigate asymmetry of sprint
running in an ecologically valid environment. Asymmetry was quantified for a group of
sprint runners through the development of novel multifactorial asymmetry scores. The
largest kinematic asymmetry values (7%) were smaller than the corresponding kinetic
values (90%). The presence of significant athlete asymmetry suggested unilateral
analyses may overlook important information. Information about individual athletes’
asymmetry may also help to inform the coaching process.
KEYWORDS: variability, bilateral, symmetry angle, inverse dynamics analysis.
INTRODUCTION: The analysis of biomechanical asymmetry has proved to be useful from
performance (Vagenas & Hoshizaki, 1991), injury (Jacobs et al., 2005) and sports technology
(Buckley, 2000) perspectives. Significant differences have been reported between values for
left and right sides of the body for kinematic (Vagenas & Hoshizaki, 1991) and kinetic (Munro
et al., 1987) variables in submaximal running. However, limited information is available
relating to asymmetry during sprint running. From a coaching perspective, knowledge of
asymmetry may inform the nature of an athlete’s training based on bilateral performance
differences. Information about asymmetry in sprint running also has implications for
biomechanical research. Previous biomechanical studies of sprint running have collected
unilateral, spatio-temporal data due to constraints on data collection, such as the positioning
of cameras or scanners (Gittoes & Wilson, 2010). In the event of a large amount of
asymmetry being present during sprint running, a unilateral analysis may provide an
incomplete description of technique and important kinematic and kinetic factors could be
overlooked if occurring in the limb that was not chosen for analysis. Dufek et al. (1995)
discussed potential problems with group-based analyses due to the creation of ‘mythical
average’ performance data. If asymmetry exists in sprint running, similarly misleading
‘mythical average’ data could result from combining data of both sides of the body. Vagenas
and Hoshizaki (1991) noted individual joint asymmetry varies within a limb and highlighted
the need to include individual joint analyses when investigating asymmetry.
Zifchock et al. (2008) proposed the ‘symmetry angle’ (θSYM) as a method of quantifying
asymmetry, which did not suffer from the artificial inflation associated with other methods
such as the symmetry index (Robinson et al., 1987). The θSYM provides values ranging from
0% (no asymmetry) to 100% (perfect asymmetry) and allows quantification of the difference
between left and right values. However, the θSYM does not include the important
consideration of intra-limb variability. Giakas and Baltzopoulos (1997) noted that for
asymmetry to be significant, the difference between values for left and right limbs must be
larger than the intra-limb variability. Therefore, the aims of this investigation were to develop
methods for quantifying kinematic and kinetic asymmetry that included the previously
neglected intra-limb variability and to use these methods to gain understanding of asymmetry
during sprint running.
METHODS: Data collection and processing: Ethical approval was gained from the
University’s Research Ethics Committee prior to commencement of the study. Eight male
sprint trained athletes performed 9-12 maximal 60 m sprint runs (mean velocity = 9.03 m∙s-1).
Athletes mean age, mass and stature were 21.75 [±4.83] years, 73.99 [±8.72] kg and 1.79
ISBS 2011
491
Porto, Portugal
Vilas-Boas, Machado, Kim, Veloso (eds.)
Biomechanics in Sports 29
Portuguese Journal of Sport Sciences
11 (Suppl. 2), 2011
[±0.07] m, respectively. Three-dimensional positional data were collected from an 8 m
section of each run, centred on the 40 m mark, using an automated motion analysis system
(CODA) operating at 200 Hz. Twelve active cx1 markers were connected in pairs to ‘twinmarker drive boxes’ and attached to athletes using adhesive tape. Markers were positioned
lateral to the fifth metatarsal-phalangeal joint, lateral malleolus, lateral condyle of the tibia,
greater trochanter, iliac crest and greater tubercle for both sides of the body. Kinetic data
were collected via two piezoelectric force plates (Kistler 9287BA), mounted end to end along
the direction of the running lane. Force plates were mounted in recessed customised
housings and covered with running track identical to that covering the rest of the lane.
Kinematic and kinetic data were filtered using a fourth-order Butterworth filter, with optimum
cutoff frequencies determined using the autocorrelation method (Challis, 1999). Twodimensional inverse dynamics analyses were performed to calculate net joint moments at the
ankle, knee and hip joints. Eight kinematic variables were selected for analysis based on
association with successful technique (Hunter et al., 2004) and identification by expert sprint
coaches (Thompson et al., 2009). Following tests for normality, parametric statistics were
used to test for significant (p<0.05) differences between left and right limbs for each variable.
Significant differences were also tested for between the magnitude of asymmetry present in
step velocity and the other kinematic variables. Seven discrete kinetic variables were
selected for analysis due to their association with successful sprint running and the kinematic
variables analysed. The inclusion of tests for significant differences between limbs meant
that intra-limb variability was included in the asymmetry measures.
Calculation of asymmetry scores: Left and right values were combined to calculate the
θSYM for each variable using the method of Zifchock et al. (2008). A composite kinematic
asymmetry score (KMAS) was calculated for each athlete by multiplying each variable’s θ SYM
by 0, 1 or 2 (representing neither, one or both tests indicating significance, respectively). A
kinetic asymmetry score (KAS) was also calculated for each athlete by summing two values.
The first was the event asymmetry score, calculated by summing the θSYM values for the
discrete variables that displayed a significant difference between left and right limbs; the
second was the profile asymmetry score, calculated from asymmetry present in the
magnitude, phase, and duration of power profiles for the ankle, knee and hip joints.
RESULTS: Kinematic θSYM values (Table 1) were all <10%, with the largest value (6.68%)
being touchdown distance for Athlete 4. Touchdown distance displayed the most frequent
significant differences (7 athletes), while minimum hip height displayed the least (2 athletes).
Table 1
Kinematic variables contributing to the kinematic asymmetry score of eight athletes
Athlete
1
2
3
4
5
6
7
8
SV
0.79*
0.62*
0.32*
0.18
0.22
0.39*
0.25
0.25
SL
1.28*
1.16*
0.79
1.33*#
1.01
1.04*
0.62
0.58
SF
1.13
1.68*#
0.81
1.44*#
1.12#
1.38*#
0.65
0.65
zHMIN
0.62
0.43
0.69
0.34
0.47*
0.70*
0.23
0.58
zKMAX
1.04*
0.92*
0.81
0.66
0.56
1.44#
0.81
1.78*#
θKFLEX
3.72*#
1.60
1.81*#
4.15*#
3.54*#
3.53#
1.39#
1.52
θHEXT
0.69
0.92*
0.69*
0.41*
0.61#
0.55
0.25
1.24*#
yTD
2.63
3.76#
2.59#
6.68*#
1.79#
2.56
3.13#
2.60#
KMAS
10.53
10.73
7.22
27.60
11.07
9.86
4.52
8.64
SV = step velocity, SL = step length, SF = step frequency, zHMIN = minimum hip height during contact,
zKMAX = maximum knee lift during contact, θKFLEX = minimum knee angle during swing, θHEXT =
maximum hip angle at end of contact, yTD = touchdown distance, * = significant difference between left
#
and right values, = significantly larger asymmetry compared to SV for the other variables.
Kinetic θSYM values, shown in Table 2, ranged from 0.06% (net vertical impulse) to >90% (net
ankle joint work). Net ankle joint work displayed the largest amount of significant differences
(4 athletes), whilst net vertical impulse and mean support moment displayed the least (1
athlete). There did not appear to be any relationship between kinematic and kinetic
asymmetry or between either kinematic or kinetic asymmetry and step velocity (Figure 1).
ISBS 2011
492
Porto, Portugal
Vilas-Boas, Machado, Kim, Veloso (eds.)
Biomechanics in Sports 29
Portuguese Journal of Sport Sciences
11 (Suppl. 2), 2011
Table 2
Kinetic variables contributing to the kinetic asymmetry score of eight athletes
Table
Table 2
2
Athlete Kinetic
IMPH variables
IMPV contributing
FzMAX MSUP
WANET asymmetry
WKNET score
WH of eight
PRO
KAS
to
the
Kinetic variables contributing to the kinetic
kinetic
asymmetry
scoreNET
of eight athletes
athletes
1
25.07*
1.27
2.14
3.54
42.95*
8.48
5.47
124.89
193.50
Athlete
IMP
IMP
Fz
M
WA
WK
WH
PRO
KAS
H
V
MAX
SUP
NET
NET
NET
Athlete
IMP
IMP
Fz
M
WA
WK
WH
PRO
KAS
H
V
MAX
SUP
NET
NET
NET
2
2.99
0.73
0.38
4.59
11.64
76.94*
11.28
209.76
286.70
1
25.07*
1.27
2.14
3.54
42.95*
8.48
5.47
124.89
193.50
1
25.07*
1.27
2.14
3.54
42.95*
8.48
5.47
124.89
193.50
3
13.44*
1.97
2.32
3.48
6.07
23.23
21.63
159.17
173.16
2
2.99
0.73
0.38
4.59
11.64
76.94*
11.28
209.76
286.70
2
2.99
0.73
0.38
4.59
11.64
76.94*
11.28
209.76
286.70
4
9.38
0.79
3.01*
5.06
21.57*
42.67
3.42
49.04
73.62
3
13.44*
1.97
2.32
3.48
6.07
23.23
21.63
159.17
173.16
3
13.44*
1.97
2.32
3.48
6.07
23.23
21.63
159.17
173.16
5
1.55
0.06
1.12
5.30*
23.74
23.82*
24.25
40.49
69.61
4
9.38
0.79
3.01*
5.06
21.57*
42.67
3.42
49.04
73.62
4
9.38
0.79
3.01*
5.06
21.57*
42.67
3.42
49.04
73.62
6
0.18
0.83
0.90
2.68
14.54*
22.86
13.83
48.00
62.54
5
1.55
0.06
1.12
5.30*
23.74
23.82*
24.25
40.49
69.61
5
1.55
0.06
1.12
5.30*
23.74
23.82*
24.25
40.49
69.61
7
10.25
1.84
0.71
3.99
41.25*
56.43
66.43
28.00
69.25
6
0.18
0.83
0.90
2.68
14.54*
22.86
13.83
48.00
62.54
6
0.18
0.83
0.90
2.68
14.54*
22.86
13.83
48.00
62.54
8
2.39
5.95*
4.33*
7.47
93.23
79.56
44.99*
67.65
122.92
7
10.25
1.84
0.71
3.99
41.25*
56.43
66.43
28.00
69.25
7 H = net 10.25
0.71
41.25*
69.25
IMP
horizontal1.84
impulse,
IMPV = 3.99
net vertical
impulse,56.43
FzMAX = 66.43
maximum28.00
vertical force,
MSUP =
8
2.39
5.95* 4.33*
4.33*
7.47
93.23
79.56
44.99* 67.65
67.65
122.92
8
2.39
5.95*
7.47
93.23
79.56
44.99*
hip joints, *
mean support moment, WANET, WKNET and WHNET = net work around the ankle, knee and 122.92
IMP
net horizontal impulse,
IMPV =
net vertical
impulse, Fz
= maximum vertical force, MSUP =
H =
IMP
= significant
difference impulse,
between IMP
left and
rightvertical
values.impulse, FzMAX
H = net horizontal
V = net
MAX = maximum vertical force, MSUP =
,, WK
WH
net
work
around
the
mean
NET and
NET =
WK
and
WH
=
net
work
around
the ankle,
ankle, knee
knee and
and hip
hip joints,
joints, **
mean support
support moment,
moment, WA
WANET
NET
NET
NET
=
significant
difference
between
left
and
right
values.
= significant difference between left and right values.
1
3
200
200
150
3
38
150
150
100
50
50
(a)
-1)-1(m∙s
Mean
MeanMean
Velocity
Velocity
Velocity
(m∙s
(m∙s
) -1)
KAS
KASKAS
2
2
250
250
200
100
100
50
10.50
2
300
300
250
7
7
0 7
0
0
8
8
1
1
5
6
4
5
6 10 5
6
20
10
10
20
20
KMAS
KMAS
4
4
30
30
30
10.50
10.50
10.50
10.00
6
6
6
10.00
10.00
9.50
8
9.50
9.50
9.00
9.00
9.00
8.50
8.50
8.50
(b)
38
8
7 3
3
5
7
0 7
2
5
5
1
2
2
1
10 1
20
0
0
10
10
20
20
KMAS
KMAS
4
-1-1
Mean
MeanMean
Velocity
Velocity
Velocity
(m∙s
(m∙s
(m∙s
) ) -1)
300
10.50
10.50
10.00
6
6
6
10.00
10.00
9.50
8
5
9.50
9.50
9.00
8
8
5
5
7
9.00
9.00
8.50
7
7
4
3
3 1
3
2
2
2
4
100
150 1
30
1 200 250 300
4
8.50
8.50
KAS
100 150 200 250 300
30
(c) 50
50 100 150 200 250 300
30
KAS
KAS and mean
mean velocity (b) and KAS
4
4
50
KMASand
KMAS
Figure 1: Comparisons
of KMAS and KAS (a), KMAS
(a)
(b)
(c)
(a) (c) for Athletes 1-8.
(b)
(c)
velocity
Figure
1:
Comparisons
of
KMAS
and
KAS
(a),
KMAS
and
mean
velocity
Figure 1: Comparisons of KMAS and KAS (a), KMAS and mean velocity (b)
(b) and
and KAS
KAS and
and mean
mean
velocity
(c)
for
Athletes
1-8.
velocity
(c)
for
Athletes
1-8.
DISCUSSION: The aims of this study were to develop methods for quantifying kinematic and
kinetic asymmetry
that
included
intra-limb
variability methods
and to use
these methods
to gain
DISCUSSION:
The
aims
study
were
to
for
kinematic
and
DISCUSSION:
The
aims of
of this
this
study
wererunning.
to develop
develop
methods
for quantifying
quantifying
kinematic
and
understanding
of
asymmetry
during
sprint
Novel
composite
asymmetry
scores
were
kinetic
asymmetry
that
included
intra-limb
variability
and
to
use
these
methods
to
gain
kinetic asymmetry
that included
intra-limb
variability
and to the
usedifference
these methods
tolimbs.
gain
developed
that
included
intra-limb
variability
when
quantifying
between
understanding
of
during
sprint
running.
Novel
composite
asymmetry
were
understanding
of asymmetry
asymmetry
during
sprint
running.
Noveland
composite
asymmetryofscores
scores
were
Calculating
overall
scoresintra-limb
along
with
detailed
kinematic
kinetic
measures
asymmetry
developed
that
included
variability
when
quantifying
the
difference
between
limbs.
developed
that
included
intra-limb
variability
when
quantifying
the
difference
between
limbs.
allowed
identification
of thealong
specific
mechanisms
underpinning
an athlete’s
asymmetry,
whilst
Calculating
overall scores
scores
with
detailed kinematic
kinematic
and kinetic
kinetic
measures
of asymmetry
asymmetry
Calculating
along with
detailed
and
measures
of
providing
a overall
method that
allowed
inter-athlete
comparisons
to be
made.
Largest
kineticwhilst
θSYM
allowed
identification
of
the
specific
mechanisms
underpinning
an
athlete’s
asymmetry,
allowed
identification
of thenine
specific
mechanisms
underpinning
an
athlete’s θasymmetry,
whilst
values
were
more
than
times
larger
than
the
largest
kinematic
values.
The
SYM
providing
a
that
allowed
inter-athlete
comparisons
to
be
Largest
kinetic
θ
providing
a method
method
that
allowed
inter-athlete
comparisons
to (e.g.
be made.
made.
Largest
kinetic
θSYM
SYM
inclusionwere
of
measures
that
incorporated
values
close
tolargest
zero
net joint
work)
reinforced
values
more
than
nine
times
larger
than
the
kinematic
θ
values.
SYM values. The
values
were
more
than
nine times
larger
than the largest
kinematic
θSYM
The
the
use
of
θ
as
a
successful
measure
of
asymmetry,
due
to
the
artificially
inflated
results
SYM
inclusion
that
incorporated
values close
to
(e.g.
joint
reinforced
inclusion of
of measures
measures
that
incorporated
to zero
zeroindex
(e.g. net
net
joint work)
work)
reinforced
characteristic
of other
methods,
such values
as
theclose
symmetry
(Zifchock
et al.,
2008).
the
use
of
θ
a
successful
measure
of
asymmetry,
due
to
the
artificially
inflated
results
SYM as
the
use
of
θ
as
a
successful
measure
of
asymmetry,
due
to
the
artificially
inflated
results
SYM
Comparing
KMAS
and
KAS
between
athletes
indicated
that
there
was
no
relationship
characteristic
of
methods,
such as
symmetry
index
(Zifchock
et
al.,
2008).
characteristic
of other
other
methods,
as the
thesimilarly
symmetry
index
(Zifchock
etand
al., KAS
2008).
between theKMAS
two
scores.
Athlete
7such
displayed
lowthat
scores
for was
KMAS
in
Comparing
and
KAS
between
athletes
indicated
there
no
relationship
Comparing
KMAS
and
KAS
between
athletes
indicated
that
there
was
no
relationship
relation
to
the
other
athletes,
whereas
Athlete
2
displayed
a
large
KAS
and
a
moderate
between
the
Athlete 7
similarly
scores
for
KAS
between
the two
two scores.
scores.
7 displayed
displayed
similarly low
lowwas
scores
for KMAS
KMAS and
andkinematic
KAS in
in
KMAS intocomparison
to theAthlete
other
athletes.
No relationship
apparent
relation
the other
other athletes,
athletes,
whereas
Athlete
2 displayed
displayed
a large
large
KASbetween
and a
a moderate
moderate
relation
to
the
whereas
Athlete
2
a
KAS
and
or
kinetic
asymmetry
and
sprint
performance.
Athletes
1
and
4
displayed
similar
mean
KMAS
to the
athletes.
No relationship
was
between
kinematic
KMAS in
in comparison
comparison
the-1)other
other
athletes.
relationship
was apparent
apparent
between
kinematic
velocities
(8.64
& 8.58toand
m∙s
but the
KMASNo
calculated
for 1Athlete
1 displayed
(10.53%)
was
lessmean
than
or
kinetic
asymmetry
sprint
performance.
Athletes
and
4
similar
or kinetic
asymmetry
and
performance.
Athletes
1 7and
4 displayed
similar
mean
-1sprint
half
the
magnitude
of
Athlete
4
(27.60%).
Athletes
6
and
displayed
similar
KAS
values
-1) but the KMAS calculated for Athlete 1 (10.53%) was less than
velocities
(8.64
&
8.58
m∙s
-1
velocities
(8.64 & 8.58
m∙s
) butlarge
the KMAS
calculated
for step
Athlete
1 (10.53%)
was
lessm∙s
than
(62.54
& magnitude
69.25%)
whilst
having
differences
in mean
velocity
(10.14
&KAS
8.68
).
half
the
of
Athlete
4
(27.60%).
Athletes
6
and
7
displayed
similar
values
half
the
magnitude
of
Athlete
4
(27.60%).
Athletes
6
and
7
displayed
similar
KAS
values
-1
The
variables
displaying
significant
asymmetry
were
different
for
all athletes.
Step
velocity
-1).
(62.54
&
69.25%)
whilst
having
large
differences
in
mean
step
velocity
(10.14
&
8.68
m∙s
(62.54 & 69.25%)
whilst
having
differences
in mean that
step asymmetry
velocity (10.14
& 8.68
m∙s
).
asymmetry
was
low
(<1%)
forlarge
all asymmetry
athletes,
indicating
present
invelocity
other
The
variables
displaying
significant
were
different
for
all
athletes.
Step
The
variables
displaying
significant
asymmetry
were
different
for
all
athletes.
Step
velocity
kinematic
and kinetic
variables all
may serve toindicating
compensate
for a strength
imbalance
or
asymmetry
low
(<1%)
that
asymmetry inwas
was
low variable
(<1%) for
for
all athletes,
athletes,
indicating
that asymmetry
asymmetry present
present in
in other
other
asymmetry
another
(Vagenas
& Hoshizaki,
1991).
kinematic
and
kinetic
variables
may
serve
to
compensate
for
a
strength
imbalance
kinematic and kinetic variables may serve to compensate for a strength imbalance or
or
asymmetry
asymmetry in
in another
another variable
variable (Vagenas
(Vagenas &
& Hoshizaki,
Hoshizaki, 1991).
1991).
ISBS 2011
493
Porto, Portugal
Vilas-Boas, Machado, Kim, Veloso (eds.)
Biomechanics in Sports 29
Portuguese Journal of Sport Sciences
11 (Suppl. 2), 2011
From a coaching perspective, asymmetry evident in kinematics and kinetics of sprint running
could influence sprint training by informing the coaching-biomechanics interface (Kerwin &
Irwin, 2008). Asymmetry present in some kinetic variables was associated with asymmetry in
corresponding kinematic variables. For example, the asymmetrical mean support moment
shown by Athlete 5 was linked with asymmetry in mimumum hip height. Inter-athlete
differences in KMAS and KAS and the contributing variables reinforced the importance of
individual analyses, as discussed by Dufek et al. (1995). Asymmetry was present in the
kinetics of all joints analysed; however, net joint work was only significantly different between
limbs for one of the three joints for each athlete, supporting the need for individual joint
asymmetry analyses (Vagenas & Hoshizaki, 1991).
From a data collection perspective, asymmetry was found to be inconsistent between
variables and between athletes. For example, if touchdown distance data were collected
unilaterally from Athlete 4, the difference of 0.06 m observed between left and right legs
would have been lost. Conversely, touchdown distance was not significantly asymmetrical for
Athlete 1; however, maximum knee lift, which was not significantly different between sides for
Athlete 4, displayed a significant difference of 0.04 m for Athlete 1. The inconsistency of
asymmetry between athletes indicated that bilateral analyses may be required to ensure
athlete-specific bilateral differences are not overlooked.
CONCLUSION: The New asymmetry scores have highlighted bilateral differences that exist
in sprint runners, which could provide coaches with information about individual athletes’
asymmetry and inform future methods of data collection. Future research could investigate
the robustness of the new asymmetry scores for a wider population and extend the new
scores to investigate asymmetry in other forms of running, such as amputee sprinting.
REFERENCES:
Buckley, J.G. (2000). Biomechanical adaptations of transtibial amputee sprinting in athletes using
dedicated prostheses. Clinical Biomechanics, 15(5), 352-8.
Challis, J.H. (1999). A procedure for the automatic determination of filter cutoff frequency for the
processing of biomechanical data. Journal of Applied Biomechanics, 15(3), 303-17.
Dufek, J.S., Bates, B.T., Stergiou, N. and James, C.R. (1995). Interactive effects between group and
single-subject response patterns. Human Movement Science, 14(3), 301-23.
Giakas, G. and Baltzopoulos, V. (1997). Time and frequency domain analysis of ground reaction
forces during walking: an investigation of variability and symmetry. Gait and Posture, 5(3), 189-97.
Gittoes, M.J.R. and Wilson, C. (2010). Intralimb joint coordination patterns of the lower extremity in
maximal velocity phase sprint running. Journal of Applied Biomechanics, 26(2), 188-95.
Hunter, J.P., Marshall, R.N. and McNair, P.J. (2004). Interaction of step length and step rate during
sprint running. Medicine and Science in Sports and Exercise, 36(2), 261-71.
Jacobs, C., Uhl, T.L., Seeley, M., Sterling, W. and Goodrich, L. (2005). Strength and fatigability of the
dominant and nondominant hip abductors. Journal of Athletic Training, 40(3), 203-6.
Kerwin, D.G. and Irwin, G. (2008). Biomechanics for coaches. In R. L. Jones, M. Hughes and K.
Kingston (Eds.), An introduction to sports coaching (pp 87-100). London, UK, Routledge.
Munro, C.F., Miller, D.I. and Fuglevand, A.J. (1987). Ground reaction forces in running – a
reexamination. Journal of Biomechanics, 20(2), 14-55.
Robinson, R.O., Herzog, W. and Nigg, B.M. (1987). Use of force platform variables to quantify the
effects of chiropractic manipulation on gait symmetry. Journal of Manipulative Physiological Therapy,
10(4), 172-6.
Thompson, A., Bezodis, I.N. and Jones, R.L. (2009). An in-depth assessment of expert sprint coaches’
technical knowledge. Journal of Sports Sciences, 27(8), 855-61.
Vagenas, G. and Hoshizaki, B. (1991). Functional asymmetries and lateral dominance in the lower
limbs of distance runners. International Journal of Sports Biomechanics, 7(4), 311-329.
Zifchock, R.A., Davis, I., Higginson, J. and Royer, T. (2008). The symmetry angle: a novel robust
method of quantifying asymmetry. Gait and Posture, 27(4), 622-7.
Acknowledgement
This work was funded by EPSRC grant number EP/D076943.
ISBS 2011
494
Porto, Portugal