Journal of Sports Sciences

ISSN: 0264-0414 (Print) 1466-447X (Online) Journal homepage: http://www.tandfonline.com/loi/rjsp20

Electromyographic analysis of thigh muscles

during track cycling on a velodrome

Kohei Watanabe, Takayuki Sato, Takahiro Mukaimoto, Wataru Takashima,

Michio Yamagishi & Tetsunari Nishiyama

To cite this article: Kohei Watanabe, Takayuki Sato, Takahiro Mukaimoto, Wataru
Takashima, Michio Yamagishi & Tetsunari Nishiyama (2015): Electromyographic analysis
of thigh muscles during track cycling on a velodrome, Journal of Sports Sciences, DOI:
10.1080/02640414.2015.1114135

To link to this article: http://dx.doi.org/10.1080/02640414.2015.1114135

Published online: 16 Nov 2015.

Submit your article to this journal

Article views: 32

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=rjsp20

Download by: [Chinese University of Hong Kong] Date: 22 November 2015, At: 05:40
 JOURNAL OF SPORTS SCIENCES, 2015
http://dx.doi.org/10.1080/02640414.2015.1114135

Electromyographic analysis of thigh muscles during track cycling on a velodrome

Kohei Watanabea, Takayuki Satob, Takahiro Mukaimotoc, Wataru Takashimad, Michio Yamagishie
and Tetsunari Nishiyamae
a
Laboratory of Neuromuscular Biomechanics, School of International Liberal Studies, Chukyo University, Nagoya, Japan; bGunma National College
of Technology, Maebashi, Japan; cTokyo University of Science, Tokyo, Japan; dHachinohe Gakuin University, Aomori, Japan; eLaboratory of Sports
Biomechanics, Nippon Sport Science University, Tokyo, Japan

ABSTRACT ARTICLE HISTORY

We aimed to investigate neuromuscular activation of thigh muscles during track cycling at various Accepted 24 October 2015
speeds. Eight male competitive cyclists volunteered to participate in this study. Surface electromyo-
KEYWORDS
Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015

graphy of the vastus lateralis, biceps femoris and adductor magnus muscles of the bilateral legs was Track cyclist; quadriceps
recorded during track cycling on velodromes with a 250-m track. The participants were instructed to femoris muscles; hamstrings
maintain three different lap times: 20, 18 and 16 s. The average rectified value (ARV) was calculated muscles; hip adductor
from the sampled surface electromyography. Significantly higher ARVs were observed in the right muscles
compared to left leg for the biceps femoris muscle during both straight and curved sections at 18- and
16-s lap times (P < 0.05). In the biceps femoris muscle, significant changes in ARVs during the recovery
phase with an increase in speed were seen in the right leg only (P < 0.05). There were no significant
differences in ARVs between the straight and curved sections for all three muscles (P > 0.05). From our
findings, it was suggested that during track cycling on a velodrome the laterality of the biceps femoris
muscle activity is a key strategy to regulate the speed, and fixed neuromuscular strategies are adopted
between straight and curved sections for thigh muscles.

Introduction the world championship and Olympic games are indoors and
mostly constructed with wood to minimise influence of the
Electromyographic analysis of lower extremity muscles during
wind on the cycling track performance and the rolling resis-
cycling has been reported by numerous researchers (Dorel,
tance. Therefore, the rider and bicycle are inclined in the
Couturier, & Hug, 2008; Duc, Bertucci, Pernin, & Grappe,
frontal plane during half of the track cycling event. A cyclist
2008; Ericson, Nisell, Arborelius, & Ekholm, 1985; Hug,
riding a bicycle around a curve generates a centrifugal accel-
Drouet, Champoux, Couturier, & Dorel, 2008a; Li & Caldwell,
eration vector and must lean inwards to compensate for the
1998; Takaishi, Yamamoto, Ono, Ito, & Moritani, 1998). From
outer acceleration (di Prampero & Narici, 2003). We thus
these studies, the effects of cadence, posture, workload, fati-
hypothesised that inconsistent joint power and muscle recruit-
gue, the training status and other factors on the neuromus-
ment are needed for pedalling movement on a banked curve.
cular activation of recruited muscles for pedalling movement
However, little attention has been focused on neuromuscular
were investigated by surface electromyography. These find-
activation of lower extremity muscles during track cycling.
ings provide us with an opportunity to improve cycling per-
Also, since the rider and bicycle are mostly oriented vertically
formance and neuromuscular mechanics during pedalling. All
in the frontal plane during cycling on an ergometer/roller and
the data reported have been obtained during cycling on
in a road race, neuromuscular activation of a unilateral leg has
ergometers or bicycles with a roller. This experimental condi-
been investigated in previous studies based on a symmetric
tion simulates cycling performance during a road race, since
pedalling strategy. During pedalling on a banked curve, an
the rider and bicycle are mostly oriented vertically in the
asymmetric muscle recruitment strategy may be utilised.
frontal plane during a road race. However, as far as we
From the various skeletal muscles, we must select key
know, no studies applied electromyographic analysis or other
muscles when investigating a specific motion or sporting
physiological measurements when the rider and bicycle were
activity. Hug et al. (2006) reported selective hypertrophy of
not oriented vertically in the frontal plane, i.e., on a curve
the vastus lateralis, biceps femoris, and adductor magnus
during a road race or track cycling.
muscles in professional cyclists. The first two muscles, i.e.,
Track cycling competitions, such as time trials, individual
the vastus lateralis and biceps femoris muscles, were exten-
and team pursuits/sprints, keirin, scratch races and point races,
sively observed in many previous studies and contribute to
are held on banked tracks or velodromes. The modern
joint motion in the sagittal plane, such as knee extension/
Olympic games and world championships use velodromes
flexion and hip extension, which generate major joint torque
with a 250-m track, which consists of two straight and two
on pedalling (Gregor, Cavanagh, & LaFortune, 1985; Neptune,
banked curved section. Also, the velodromes that are used for
Kautz, & Hull, 1997; Neptune, Kautz, & Zajac, 2000). On the

CONTACT Kohei Watanabe wkohei@lets.chukyo-u.ac.jp

© 2015 Taylor & Francis
 2 K. WATANABE ET AL.
other hand, there are few quantitative data on neuromuscular
activation of the adductor magnus muscle during pedalling. It
has been recognised that applying the surface electromyogra-
phy technique to hip adductor muscles, including the adduc-
tor magnus muscle, is difficult because they are deeper
muscles. Recently, we established a methodology to record
surface electromyography of the adductor magnus muscle,
which employs ultrasonographic guidance to correctly define
the location of the muscle under the skin (Watanabe,
Katayama, Ishida, & Akima, 2009). Since this muscle may con-
tribute to joint motion in the frontal plane, such as hip adduc-
tion (Pressel & Lengsfeld, 1998), electromyographic analysis of
the adductor magnus muscle could provide important infor-
mation to help understand muscle activation during track
cycling.
The present study aimed to investigate neuromuscular
Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015
activation of thigh muscles of the bilateral legs during track
cycling at various speeds. For the above-mentioned reasons,
electromyographic analysis was applied to the vastus lateralis,
biceps femoris and adductor magnus muscles of the bilateral
legs. In this study, electromyographic responses to an increase
in speed were characterised to understand the neuromuscular
strategy of the thigh muscles during track cycling. Based on
joint motion generated by the muscles, we hypothesised that
the adductor magnus muscle is utilised in inconsistent neuro-
muscular activation patterns between the legs during the
banked curves, and that neuromuscular activation patterns
of other muscles are consistent between the legs during over-
all track cycling.
Methods
Participants
Eight male competitive cyclists (more than three years of
racing experience) volunteered to participate in this study
(age: 20.7 ± 1.3 years, height: 170.0 ± 5.3 cm, body mass:
66.0 ± 4.7 kg). Their peak performance time for the 4000-m
individual pursuit was 290.242 ± 14.2 s. Dominant leg during
soccer and pedalling was identified. Seven of eight partici-
pants selected right leg as kicking leg for soccer. In pedalling,
four of eight participants prefer the use of right leg during
down-stroke phase. The participants often trained in the velo-
dromes used in this study and familiar with track cycling. They
gave informed consent for the study after receiving a detailed
explanation of the purposes, potential benefits and risks asso-
ciated with participation. All participants were healthy, with no
history of any musculoskeletal or neurological disorders. All
procedures used in this study were in accordance with the
Declaration of Helsinki and research code of ethics of Chukyo
University.
Experimental design
The experiment was conducted on an indoor velodrome with
a 250-m track, consisting of two straights and two banked
curves (Izu Velodrome, Shizuoka, Japan) (Figure 1(A)).
Participants individually performed track cycling at three dif-
ferent speeds, controlled by the lap time: 20, 18 and 16 s per
250 m (participants performed trials with the three lap times
in this order). Since at least two laps were needed for analysis
at each speed, approximately three laps at a constant speed
were set. Each participant was told their half lap time for every
half lap in the middle of the straight (Figure 1(A-a)). Rest
periods among the trials at the different speeds were
≥10 min; the mean rest period was 19.7 min and the longest
was 37 min.
Material and data collection
During track cycling, a digital video was taken at 60 Hz with
panning on the rider from the centre of the velodrome. From
the video, time points were extracted when the rider passed
the pursuit lines (Figure 1(A-a)) to measure the lap time for
estimating the speed, and the borders between the straight/
curve and curve/straight (Figure 1(A-b)) to determine the
straight or curve periods. At the beginning of the video, a
light emitting diode with an electrical pulse was used for
synchronisation with surface electromyography and electrical
signals for the crank angle.
Participants were instructed to pedal with a dropped pos-
ture at a constant speed and run between measuring and
sprinter lines (Figure 1(B)). Because track cycling was per-
formed anticlockwise, right and left legs were located on the
outside and inside of the bank, respectively (Figure 1(B)). Each
participant rode a standard track-racing bicycle (gear ratio of
front/rear: 50/14) with a magnet switch mounted to identify
the top dead centre of the crank cycle and an optical rotary
encoder with a resolution angle of 7.3° to determine the crank
angle. In this study, the highest pedal position (top dead
centre) was defined as 0°. The handle bars, saddle and cycling
shoes were changed to each participant’s own items and
adjusted to their individual cycling style for optimal pedalling
performance.
Surface electromyography signals were recorded from the
vastus lateralis, biceps femoris and adductor magnus muscles
of the bilateral legs. Electrode specifications in this study were
as follows: amplification, single differential; inter-electrode dis-
tance, 2 cm; contact sensors, two 0.1 × 1-cm silver bars; pre-
amplifier gain, 400-fold; input impedance, >200 MΩ; common
mode rejection ratio, >110 dB. The main amplifier unit feature
was a gain of 100-fold and frequency response of 5–500 Hz
(FA-DL-141, 4 assist, Tokyo, Japan). Surface electromyography
data were sampled at 1000 Hz and recorded by a compu-
terised data logger (FA-DL-3100, 4 assist, Tokyo, Japan). Prior
to attaching the electrodes, the skin was shaved, abraded and
cleaned with alcohol. The electrodes for the vastus lateralis
and biceps femoris muscles were placed at the midpoint
between the head of the greater trochanter and inferior
edge of the patella and the midpoint between the ischial
tuberosity and lateral epicondyle of the tibia, respectively
(Watanabe et al., 2009). These electrodes were placed parallel
to the estimated muscle fibres. For the adductor magnus
muscle, the electrode location was determined under the
guidance of ultrasonographic imaging (FAZONA CB; FujiFilm,
Tokyo, Japan). This procedure was fully described in our pre-
vious study (Watanabe et al., 2009). A common reference
electrode was attached to the iliac crest. After the placement

Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015
Figure 1. Identification of straight and curved sections based on video and camera locations on the velodrome (A). A participant at the middle point of a straight (A-
a). A participant at the transition from a curve to straight (A-b). A participant at a curve (A-c). Illustration of a participant during cycling on a curve. Participants were
instructed to run between measuring and sprinter lines (B).
of all electrodes, we confirmed a validation of electrode loca-
tions and SEMG responses with manual muscle testing.
Data analysis
From the video, three continuous curves and straights at a
constant speed were selected for analysis (Figure 2(A)). Three
crank cycles were defined by electrical signals from the mag-
net switch at TDC for each curve or straight (Figure 2(B)).
Therefore, nine crank cycles were selected for a straight or
curved section for a lap time or speed. Electromyography
signals recorded during the selected times were filtered with
a 4th order Butterworth filter (band width was 20–400 Hz) and
the average rectified value (ARV) was calculated. ARVs from
nine crank cycles for a straight or curve at a specified speed
were averaged every 7.3° of the crank angle (Figure 2(C))
because the encoder signal separated a crank cycle into 49
parts. The mean ARV during a crank cycle was calculated (total
ARV) for analysis. A crank cycle was divided into four sectors:
sector 1, 302–2°; sector 2, 2–122°; sector 3, 122–182°; sector 4,
182–302°, based on a previous study (Hug, Drouet, Champoux,
Couturier, & Dorel, 2008b) (Figure 2(D)). These sectors were
determined from electromechanical delay and mechanical
data and a functional aspect; sectors 1, 2, 3 and 4 correspond
to around top dead centre, propulsive, around bottom dead
centre and recovery phases, respectively (Hug et al., 2008b).
JOURNAL OF SPORTS SCIENCES 3
The total ARV and ARV during each sector for each muscle and
each participant were calculated and normalised by the peak
ARV for each participant.
To identify the muscular activation phase, the onset and
offset of surface electromyography were assessed. Phases of
muscular activation were defined by the period when the
signal was above a threshold of 25% of the peak ARV for
each muscle and participant (Duc et al., 2008; Li & Caldwell,
1998). When multiple bursts were noted during a crank cycle,
onset and offset were calculated from the burst including
peak activation during a crank cycle.
Statistics
All data are shown as the mean and standard deviation (SD).
Non-parametric tests were used in this study since the sample
size was not large (n = 8) and the distribution of data was
partly non-Gaussian. The Shapiro–Wilk test was used to test
for the normality of data. Friedmann and Wilcoxon rank sum
tests were performed to test changes with speed and to
compare sections (straight/curve), legs (left/right) for total
ARV, ARV during each sector for each muscle, onset and offset
of surface electromyography and cadence of cycling. The level
of significance was set at 0.05. Statistical analyses were per-
formed using MATLAB (version 7, Math Works GK, Tokyo,
Japan) and SPSS software (version 15.0; SPSS, Tokyo, Japan).
 4 K. WATANABE ET AL.
Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015

Figure 2. Raw surface electromyographic signals for three selected thigh muscles of left and right legs during one and a half laps for a participant with a 16-s lap
time (A). Raw surface electromyographic signals of the vastus lateralis muscle of the right leg during a curve and straight of a participant (B). The average rectified
value (ARV) of surface electromyography during a crank cycle of the vastus lateralis muscle of the right leg of a participant. Solid and dashed lines indicate straight
and curved sections, respectively (C). Definition of crank angle and phases. VL, vastus lateralis muscle; BF, biceps femoris muscle; AM, adductor magnus muscle; R,
right leg; L, left leg; TDC, top dead centre.

Results Mean electromyography patterns during a crank cycle

General results
Mean lap times were 20.1 ± 0.2, 18.1 ± 0.2 and 16.0 ± 0.3 s for
the tasks set at lap times of 20, 18 and 16 s, respectively.
Speeds for each trial calculated from the lap times were
44.7 ± 0.5, 49.7 ± 0.5 and 56.2 ± 1.0 km · h−1 for the 20, 18
and 16 s, respectively. These speeds were 90.1 ± 4.1,
100.1 ± 5.1 and 113.3 ± 5.5% of the participants’ average
race speed for the 4000-m individual pursuit, respectively.
Mean cadences during straight and curved sections were
100.5 ± 1.7 and 102.9 ± 2.1 for a 20-s lap time, 111.6 ± 1.1
and 114.8 ± 1.4 for an 18-s lap time, and 123.8 ± 3.7 and
129.0 ± 3.4 for a 16-s lap time, respectively. With an increase in
speed, cadences were significantly increased for both straight
and curved sections (P < 0.05). Also, cadences during curves
were significantly higher than those during straights for all lap
times (P < 0.05).
for the three thigh muscles and speeds are shown in
Figure 3. Increases in the ARV were mainly observed in
the sectors 1 and 2 for the vastus lateralis muscle, sectors
2 and 3 for the biceps femoris muscle and sectors 1 and 2
for the adductor magnus muscles, respectively. While simi-
lar patterns were seen in the left and right legs and
straight and curved sections for the vastus lateralis and
adductor magnus muscles, a relatively higher ARV was
demonstrated from 90° to 270° of the crank angle in the
right compared to left leg for straight and curved sections
in the biceps femoris muscle.
ARV for each sector
Total ARVs in the vastus lateralis and biceps femoris muscles
for both legs and sections were significantly changed with an
increase in speed (P < 0.05) (Figure 4). In the vastus lateralis
 JOURNAL OF SPORTS SCIENCES 5
Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015

Figure 3. The mean normalised average rectified value (ARV) during a crank cycle. Solid, dashed, thick solid, and thick dashed lines indicate the right leg during
straight, right leg during curved, left leg during straight and left leg during curved sections. VL, vastus lateralis muscle; BF, biceps femoris muscle; AM, adductor
magnus muscle.

Figure 4. The mean normalised average rectified value (ARV) during a complete crank cycle. VL, vastus lateralis muscle; BF, biceps femoris muscle; AM, adductor
magnus muscle; R, right leg; L, left leg; S, straight; C, curve. *P < 0.05 significant change with an increase in speed. #P < 0.05 significant difference between left and
right legs.

muscle, total ARVs for the left leg were significantly higher
than that for the right leg during curves with 18- and 16-s lap
times and during straights with 16-s lap time (P < 0.05)
(Figure 4). In the biceps femoris muscle, total ARVs for the
left leg were significantly higher than for the right leg during
straights with a 20-s lap time (P < 0.05) (Figure 4).
For the vastus lateralis muscle, ARVs in sectors 1 and 2 for
both legs were significantly higher with an increase in speed
during straight and curved sections (P < 0.05) (Figure 5). In
sectors 3 and 4, significant changes with an increase in speed
were observed only in the right leg (P < 0.05) (Figure 5).
Significantly higher ARVs were shown in the left leg compared
to right leg during curves with 18- and 16-s lap times in sector
2, during curves with all lap times in sector 3, and during
curves with 18- and 16-s lap times and during straights with
an 18 s lap time in sector 4 (P < 0.05) (Figure 5). In sector 2,
the ARV in the right leg with a 16-s lap time was significantly
higher during straight than curved section (P < 0.05)
(Figure 5).
For the biceps femoris muscle, ARVs in sectors 1 and 2
for both legs and sections were significantly higher with an
increase in speed (P < 0.05) (Figure 6). In sectors 3 and 4,
significant changes with an increase in speed were
observed only in the right leg (P < 0.05) (Figure 6). In
sectors 3 and 4, ARVs in the left leg during straights were
significantly higher than in the right leg with all lap times
(P < 0.05) (Figure 6).
For adductor magnus muscle, ARVs in sectors 1 and 2 were
significantly higher with an increase in speed only in the left
leg (P < 0.05) (Figure 7).
 6 K. WATANABE ET AL.
Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015
Figure 5. The mean normalised average rectified value (ARV) for the vastus lateralis muscle during sectors 1–4. VL, vastus lateralis muscle; R, right leg; L, left leg; S,
straight; C, curve. *P < 0.05 significant change with an increase in speed. #P < 0.05 significant difference between left and right legs. +P < 0.05 significant difference
between a straight and curve.
Onset and offset of surface electromyography
The offset of surface electromyography was significantly chan-
ged with an increase in speed for both legs and sections in the
vastus lateralis muscle (P < 0.05) (Figure 8). With all lap times,
the offset of surface electromyography in the vastus lateralis
muscle of the right leg appeared earlier in the crank cycle
during curves. For the biceps femoris muscle, significant
changes in the onset of surface electromyography were
observed in the right leg during both sections and in the left
leg during the straight (P < 0.05) (Figure 8). Also, a significant
change in the offset of surface electromyography was noted in
the right leg during curves in the biceps femoris muscle
(P < 0.05) (Figure 8). The offset of surface electromyography
in the biceps femoris muscle of the right leg appeared earlier
in the crank cycle during straights with all lap times (P < 0.05)
(Figure 8). For the biceps femoris muscle, the onset of surface
electromyography in the right leg with a 20-s lap time during
straights appeared earlier in the crank cycle than during
curves (P < 0.05) (Figure 8). For the adductor magnus muscle,
the offset of surface electromyography was significantly chan-
ged with an increase in speed for both legs in curves and for
left leg during straight (P < 0.05) (Figure 8). Also, the offset of
surface electromyography in the right leg of the adductor
magnus muscle was detected earlier in the crank cycle than
that in the left leg during straights with a 16-s lap time
(P < 0.05) (Figure 8).
Discussion
The present study investigated neuromuscular activation of
the bilateral thigh muscles at various speeds during track
cycling. To our knowledge, surface electromyography patterns
of muscles while cycling on a velodrome were identified for
the first time in the present study. The main findings of this
study were that asymmetric surface electromyography
responses with an increase in speed were observed in all
three tested muscles, with no stereotyped behaviours among
the muscles.
From onset and offset timing and surface electromyogra-
phy-crank angle relationships, we could confirm that activa-
tion patterns of the vastus lateralis and biceps femoris muscles
were similar with those during cycling on an ergometer or a

Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015
Figure 6. The mean normalised average rectified value (ARV) for the biceps femoris muscle during sectors 1–4. BF, biceps femoris muscle; R, right leg; L, left leg; S,
straight; C, curve. *P < 0.05 significant change with an increase in speed. #P < 0.05 significant difference between left and right legs.
roller in previous studies (Baum & Li, 2003; Duc et al., 2008;
Ericson et al., 1985; Hug & Dorel, 2009; Hug et al., 2008a; Li
& Caldwell, 1998; Takaishi et al., 1998; Watanabe et al.,
2009). On the other hand, for the adductor magnus muscle,
the surface electromyography pattern during a crank cycle
was inconsistent with a previous study (Watanabe et al.,
2009). Our previous study showed two different electromyo-
graphy patterns among participants: a single peak during
the propulsive and two peaks during the propulsive and
recovery phases in the adductor magnus muscle (Watanabe
et al., 2009). The present study also demonstrated two
different electromyography patterns: a single peak during
sector 2 (n = 4) and a major peak during sector 2 with
additional peaks during sectors 3 and 4 (n = 4) (Figure 9).
While direct comparison between our previous study
(Watanabe et al., 2009) and the present study is difficult,
inter-individual differences in surface electromyography–
crank angle relationships were detected in the adductor
magnus muscle in both studies. There are many differences
in experimental conditions, i.e., cyclist/non-cyclist, erg-
ometer/bicycle on a velodrome or dropped/upright pos-
tures, between the present and previous studies. Although
JOURNAL OF SPORTS SCIENCES 7
these variations could lead to changes in activation patterns
of the recruited muscles, similar patterns were observed
between the present and previous studies in the vastus
lateralis and biceps femoris muscles. This suggests that the
activation pattern during cycling could be sensitive to ped-
alling conditions and/or experience regarding the adductor
magnus muscle, but not the vastus lateralis and biceps
femoris muscles. In order to clarify the characteristics of
the muscle activation pattern while cycling on a velodrome
in detail, further studies in which pedalling conditions and
experience are strictly controlled are needed.
Right and left legs are located on the outside and inside
of the bank, respectively, during cycling on a velodrome
(Figure 1(B)). On both straight and curve, the sprinter’s
lane, which the participants of this study used, is inclined
horizontally. We thus suggest that the left and right legs play
different roles during track cycling. In the present study,
electromyographic responses to an increase in speed were
characterised to understand the neuromuscular strategy
between left and right legs during track cycling. Bilateral
legs changes in normalised ARVs with an increase in speed
were not uniform among the tested three muscles. Although
 8 K. WATANABE ET AL.
Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015

Figure 7. The mean normalised average rectified value (ARV) for the adductor magnus muscle during sectors 1–4. AM, adductor magnus muscle; R, right leg; L, left
leg; S, straight; C, curve. *P < 0.05 significant change with an increase in speed.

Figure 8. Mean onset and offset timing of surface electromyography during a crank cycle. VL, vastus lateralis muscle; BF, biceps femoris muscle; AM, adductor
magnus muscle; R, right leg; L, left leg; S, straight; C, curve. Circle and triangle indicate onset and offset timings, respectively. §P < 0.05 significant change with an
increase in speed for onset timing. *P < 0.05 significant change with an increase in speed for offset timing. #P < 0.05 significant difference between left and right
legs. +P < 0.05 significant difference between straight and curve.

we assessed relative changes in ARVs, a significant change in
the ARV with an increase in speed reflects changes in the
amount of muscle activation. For the vastus lateralis muscle,
ARVs during sectors 1 and 2, where this muscle activates
increased, for both legs with an increase in speed during
curved and straight sections (P < 0.05) (Figure 5). For the
biceps femoris muscle, while a significant rise in ARVs with
an increase in speed was observed in both legs during
sectors 1 and 2, these patterns were seen only in the right
leg during sectors 3 and 4 (Figure 6). For the adductor
magnus muscle, a significant increase in ARVs was found
only in the left leg during sectors 1 and 2, where this muscle

Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015
Figure 9. Example of two patterns of averaged rectified values (ARVs) during a crank cycle for the adductor magnus muscle of the left leg during a straight. AM,
adductor magnus muscle. Participants 5 and 3 showed single major peak and major peak activation with additional bursts, respectively.
activates (P < 0.05) (Figure 7). These results suggest that
biceps femoris and adductor magnus muscles of left and
right legs play different roles during track cycling.
In a previous study (Craig & Norton, 2001), it was stated that
the power output and speed fluctuations are proportional to
the cyclist’s position on the track, i.e., higher power output and
speed are seen during straights and curves, respectively. The
positional relationship between the body and bank could be
different between straights and curves. It was estimated that a
participant riding a bicycle along a curved path generates a
centrifugal acceleration vector and must lean inwards to com-
pensate for the outer acceleration (di Prampero & Narici, 2003).
Thus, it can be hypothesised that different strategies are used
between straight and curved sections for the recruited muscles.
In the present study, a different form of left–right asymmetry in
ARV responses to an increase in speed was observed between
straights and curves in the vastus lateralis and biceps femoris
muscles. For example, higher normalised ARVs in the vastus
lateralis muscle of the left leg in sectors 2 and 3 were detected
during the curve (Figure 5). In the biceps femoris muscle,
normalised ARVs in sectors 3 and 4 in the left leg were higher
than those in the right leg during straights, but not during
curves, except for the curves for the lowest speed (20 s)
(Figure 6). From the results of offset timing, these higher nor-
malised ARVs in the left leg can be explained by the longer
duration of activation due to a later offset timing during a crank
cycle in the left leg (Figure 8). These results suggest that vastus
lateralis and biceps femoris muscles are regulated in asym-
metric and the section-dependent manners during track
cycling. In the vastus lateralis muscle, a longer activation in
the left leg appeared at around 90–100° of the crank angle
(Figures 3 and 6). In this phase, contraction of the vastus
lateralis muscle could contribute to produce a downward
crank force or higher power during the downstroke. Since
cyclists on a curve need to incline their body and bicycle
inwards (di Prampero & Narici, 2003), the left leg, which is
located on the inside of bank, would load greater body weight
than right leg during curve. We thus suggested that this body
weight loading may induce greater and/or longer activation of
the vastus lateralis muscle of the left leg during the curve.
JOURNAL OF SPORTS SCIENCES 9
For track cycling, very few reports are currently available in
the published literature on performances of cyclists on velo-
dromes. While some studies reported the power output of
cyclists during track cycling (Broker, Kyle, & Burke, 1999;
Craig & Norton, 2001), there were no data on physiological
or biomechanical responses. Here we present, for the first
time, a part of neurophysiological responses during track
cycling. However, further experiments are needed to under-
stand the physiological or biomechanical responses during
track cycling.
The trials with three different lap times were performed in a
fixed order. Although the rest periods were provided, we did
not assess fatigue development in participants. Therefore,
trials with a 16-s lap time, the final test, may include an effect
of fatigue on neuromuscular activation. This is one of the
limitations of this study. We should also note another limita-
tion in the calculation processes of surface electromyography.
The total ARV and ARV in each sector were normalised by the
peak ARV during all trials for each muscle of each leg and each
participant. Since the peak ARV is not detected during max-
imum contraction, the normalised ARV in the present study
should be interpreted as relative changes with an increase in
speed. Moreover, our statistical analysis, such as the non-para-
metric test for the small sample (n = 8), may lead to type II
error. Further studies with larger samples are needed to clarify
this issue. Finally, we should consider limited muscles from
which we recorded surface electromyography. We selected
three thigh muscles based on the muscle volume, function
and morphological data of cyclists in a previous study (Hug
et al., 2006). However, these muscles are only some of the
recruited muscles during cycling. Other muscles shown to be
active during cycling are gluteus and calf muscles (Dorel et al.,
2008; Hug & Dorel, 2009). To understand the muscle recruit-
ment pattern during track cycling, surface electromyography
recording of other lower extremity muscles would be
necessary.
In conclusion, we demonstrated SEMG of three selected
thigh muscles: the vastus lateralis, biceps femoris and adduc-
tor magnus muscles, during cycling on a velodrome in order
to investigate the neuromuscular strategy for track cycling.
 10 K. WATANABE ET AL.
There were right–left differences in surface electromyography
responses with an increase in speed in all tested muscles. Also,
asymmetric surface electromyography responses were selec-
tively noted during curved or straight sections in the vastus
lateralis and biceps femoris muscles. These results suggest
that the vastus lateralis and biceps femoris muscles are regu-
lated in asymmetric and section-dependent manners.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This research was supported in part by a research grant from Nippon
Sports Science University (PI: Nishiyama T.).
Downloaded by [Chinese University of Hong Kong] at 05:40 22 November 2015
References
Baum, B. S., & Li, L. (2003). Lower extremity muscle activities during cycling
are influenced by load and frequency. Journal of Electromyography and
Kinesiology, 13(2), 181–190. doi:10.1016/S1050-6411(02)00110-4
Broker, J. P., Kyle, C. R., & Burke, E. R. (1999). Racing cyclist power require-
ments in the 4000-m individual and team pursuits. [Comparative
Study]. Medicine & Science in Sports & Exercise, 31(11), 1677–1685.
doi:10.1097/00005768-199911000-00026
Craig, N. P., & Norton, K. I. (2001). Characteristics of track cycling. [Review].
Sports Medicine, 31(7), 457–468. doi:10.2165/00007256-200131070-
00001
di Prampero, P. E., & Narici, M. V. (2003). Muscles in microgravity: From
fibres to human motion. [Lectures]. Journal of Biomechanics, 36(3), 403–
412. doi:10.1016/S0021-9290(02)00418-9
Dorel, S., Couturier, A., & Hug, F. (2008). Intra-session repeatability of lower
limb muscles activation pattern during pedaling. Journal of
Electromyography and Kinesiology, 18(5), 857–865.
Duc, S., Bertucci, W., Pernin, J. N., & Grappe, F. (2008). Muscular activity
during uphill cycling: Effect of slope, posture, hand grip position and
constrained bicycle lateral sways. Journal of Electromyography and
Kinesiology, 18(1), 116–127. doi:10.1016/j.jelekin.2006.09.007
Ericson, M. O., Nisell, R., Arborelius, U. P., & Ekholm, J. (1985). Muscular
activity during ergometer cycling. Scandinavian Journal of
Rehabilitation Medicine, 17(2), 53–61.
Gregor, R. J., Cavanagh, P. R., & LaFortune, M. (1985). Knee flexor moments
during propulsion in cycling-a creative solution to Lombard’s Paradox.
Journal of Biomechanics, 18(5), 307–316. doi:10.1016/0021-9290(85)
90286-6
Hug, F., & Dorel, S. (2009). Electromyographic analysis of pedaling: A
review. Journal of Electromyography and Kinesiology, 19(2), 182–198.
doi:10.1016/j.jelekin.2007.10.010
Hug, F., Drouet, J. M., Champoux, Y., Couturier, A., & Dorel, S. (2008a).
Interindividual variability of electromyographic patterns and pedal
force profiles in trained cyclists. European Journal of Applied
Physiology, 104(4), 667–678. doi:10.1007/s00421-008-0810-y
Hug, F., Drouet, J. M., Champoux, Y., Couturier, A., & Dorel, S. (2008b).
Interindividual variability of electromyographic patterns and pedal
force profiles in trained cyclists. [Research Support, Non-U.S. Gov’t].
European Journal of Applied Physiology, 104(4), 667–678. doi:10.1007/
s00421-008-0810-y
Hug, F., Marqueste, T., Le Fur, Y., Cozzone, P. J., Grélot, L., & Bendahan, D.
(2006). Selective training-induced thigh muscles hypertrophy in profes-
sional road cyclists. European Journal of Applied Physiology, 97(5), 591–
597. doi:10.1007/s00421-006-0218-5
Li, L., & Caldwell, G. E. (1998). Muscle coordination in cycling: Effect of
surface incline and posture. Journal of Applied Physiology, 85(3), 927–
934.
Neptune, R. R., Kautz, S. A., & Hull, M. L. (1997). The effect of pedaling rate
on coordination in cycling. Journal of Biomechanics, 30(10), 1051–1058.
doi:10.1016/S0021-9290(97)00071-7
Neptune, R. R., Kautz, S. A., & Zajac, F. E. (2000). Muscle contributions to
specific biomechanical functions do not change in forward versus
backward pedaling. Journal of Biomechanics, 33(2), 155–164.
doi:10.1016/S0021-9290(99)00150-5
Pressel, T., & Lengsfeld, M. (1998). Functions of hip joint muscles. Medical
Engineering & Physics, 20(1), 50–56. doi:10.1016/S1350-4533(97)00040-4
Takaishi, T., Yamamoto, T., Ono, T., Ito, T., & Moritani, T. (1998).
Neuromuscular, metabolic, and kinetic adaptations for skilled pedaling
performance in cyclists. Medicine & Science in Sports & Exercise, 30(3),
442–449. doi:10.1097/00005768-199803000-00016
Watanabe, K., Katayama, K., Ishida, K., & Akima, H. (2009).
Electromyographic analysis of hip adductor muscles during incremental
fatiguing pedaling exercise. European Journal of Applied Physiology, 106,
815–825. doi:10.1007/s00421-009-1086-6