J Appl Physiol 123: 884 –893, 2017.
First published June 29, 2017; doi:10.1152/japplphysiol.00536.2016.
RESEARCH ARTICLE
Factors contributing to lower metabolic demand of eccentric compared with
concentric cycling
Luis Peñailillo,1 Anthony J. Blazevich,2 and X Kazunori Nosaka2
1
Exercise Science Laboratory, School of Kinesiology, Faculty of Medicine, Universidad Finis Terrae, Santiago, Chile; and
2
Centre for Exercise and Sports Science Research, School of Medical and Health Sciences, Edith Cowan University,
Joondalup, Western Australia, Australia
Submitted 16 June 2016; accepted in final form 26 June 2017
Peñailillo L, Blazevich AJ, Nosaka K. Factors contributing to
lower metabolic demand of eccentric compared with concentric cy-
cling. J Appl Physiol 123: 884 – 893, 2017. First published June 29,
2017; doi:10.1152/japplphysiol.00536.2016.—This study compared
muscle-tendon behavior, muscle oxygenation, and muscle activity
between eccentric and concentric cycling exercise at the same work
output to investigate why metabolic demand is lower during eccentric
cycling than with concentric cycling. Eleven untrained men
(27.1 ⫾ 7.0 y) performed concentric cycling (CONC) and eccentric
cycling (ECC) for 10 min (60 rpm) at 65% of the maximal concentric
cycling power output (191 ⫾ 45 W) 4 wk apart. During cycling,
oxygen consumption (V̇O2), heart rate (HR), vastus lateralis (VL)
tissue total hemoglobin (tHb), and oxygenation index (TOI) were
recorded, and muscle-tendon behavior was assessed using ultrasonog-
raphy. The surface electromyogram (EMG) was recorded from VL,
vastus medialis (VM), rectus femoris (RF), and biceps femoris (BF)
muscles, and cycling torque and knee joint angle during each revo-
lution were also recorded. Average V̇O2 (⫺65 ⫾ 7%) and HR
(⫺35 ⫾ 9%) were lower and average TOI was greater (16 ⫾ 1%)
during ECC than CONC, but tHb was similar between bouts. Positive
and negative cycling peak crank torques were greater (32 ⫾ 21 and
48 ⫾ 24%, respectively) during ECC than CONC, but muscle-tendon
unit and fascicle and tendinous tissue length changes during pedal
revolutions were similar between CONC and ECC. VL, VM, RF, and
BF peak EMG amplitudes were smaller (24 ⫾ 15, 22 ⫾ 18, 16 ⫾ 17,
and 18 ⫾ 9%, respectively) during ECC than CONC. These results
suggest that the lower metabolic cost of eccentric compared with
concentric cycling was due mainly to a lower level of muscle activa-
tion per torque output.
NEW & NOTEWORTHY This study shows that lower oxygen
consumption of eccentric compared with concentric cycling at the
same workload is explained by lower muscle activity of agonist and
antagonist muscles during eccentric compared with during concentric
cycling.
metabolic cost; lengthening contractions; muscle efficiency; stretch-
shortening cycle; muscle-tendon behavior
MUSCLES ARE ACTIVATED during lengthening in eccentric con-
tractions, whereas they are activated during shortening in
concentric contractions. The classic study of Abbott et al. (1)
reported 41– 67% lower oxygen consumption (V̇O2) during
eccentric cycling when compared with concentric cycling for
the same work, and this finding was confirmed by others (7, 17,
Address for reprint requests and other correspondence: L. Peñailillo, School
of Kinesiology, Universidad Finis Terrae, 1509 Pedro de Valdivia Av.,
Providencia, Santiago Chile (e-mail: lpenailillo@uft.cl).
884 8750-7587/17 Copyright © 2017 the American Physiological Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
47, 50). For example, Bigland-Ritchie and Woods (7) reported
that muscles consumed 79% less oxygen in eccentric than
concentric cycling for the same work output and speculated
that this was due to fewer muscle fibers being recruited to
perform eccentric vs. concentric cycling. Indeed, it has been
shown using surface electromyogram (EMG) recordings and
the interpolated twitch technique that muscle activation is
lower and force output is higher during maximal eccentric than
during maximal concentric contractions (5). Peñailillo et al.
(48) reported that vastus lateralis (VL) EMG amplitude was
38% lower during eccentric than concentric cycling performed
at the same power output (⬃175 W). However, it is not
necessarily clear how the same work can be performed with
lower muscle fiber activation in continuous (cyclic) eccentric
exercise than in concentric exercise.
Although an action of passive elastic muscle structures (i.e.,
aponeurosis, tendons) has been postulated to explain the
greater force generation and low metabolic cost during eccen-
tric contractions (39), a recent study has shown that titin, one
of the sarcomeric proteins responsible for elasticity, actively
contributes to this capacity due to the ability to change its
stiffness in the presence of calcium (59). Furthermore, titin
appears to be involved in the capacity of the muscle to produce
greater isometric force after it is actively lengthened in com-
parison with a purely isometric contraction performed at that
length, i.e., the residual force enhancement phenomenon (46,
59). This contributes to the lesser ATP utilization in eccentric
contractions (34) and increased neuromuscular efficiency
(force/EMG) (33). In addition, although muscle fascicles
shorten during concentric contractions, fascicles may remain
quasi-isometric, whereas the myotendinous region and aponeu-
rosis of the muscle are lengthened during eccentric contrac-
tions, suggesting that the tendinous tissue elongates signifi-
cantly (31).
In fact, previous studies have shown a quasi-isometric be-
havior of muscle fascicles during maximal eccentric isokinetic
contractions of medial gastrocnemius (10) and tibialis anterior
(53). For example, Chino et al. (10) showed that tendinous
tissues contributed 40 – 65% of the ankle flexor length change
during maximal eccentric contractions. However, some authors
have reported that muscle-tendon behavior is different between
muscles such that fascicle and tendinous tissue stretching in
vastus lateralis are greater than those in medial gastrocnemius
muscles during a drop jump (32), and this is also affected by
the intensity of the contraction (i.e., less fascicle stretch at
higher intensities) (30, 32). Furthermore, during activities that
http://www.jappl.org
 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.
involve significant stretch-shortening cycle (SSC) patterns
such as walking, running, and jumping, where the muscle-
tendon unit stretches to store elastic energy, skeletal muscle
length change may be minimal, whereas the tendon undergoes
significant strain (8, 9, 13). This cyclic performance of the
muscle-tendon unit during terrestrial locomotion differs from
the eccentric contraction-dominant eccentric cycling task, in
which SSC is minimal, as we have shown previously (49). In
eccentric cycling, a large number of repeated pure eccentric
contractions are performed without active concentric contrac-
tions; thus, when compared with traditional concentric cycling,
the opportunity is provided to compare characteristics of pure
eccentric vs. concentric exercise.
Using near-infrared spectroscopy (NIRS), Denis et al. (12)
showed that the VL oxygenation index (TOI: %of oxygen in
myoglobin) decreased during maximal knee extensor isokinetic
(60°/s) eccentric contractions to exhaustion but was not differ-
ent from the changes observed during maximal concentric
contractions of the same muscle. In contrast, Muthalib et al.
(44) reported that TOI of biceps brachii was 50% greater
during maximal eccentric than during concentric contractions
of the elbow flexors despite 30% greater torque being gener-
ated during the eccentric contractions. These authors specu-
lated that this resulted from a lower central activation of the
muscle and consequently lower ATP demand at the muscle
level during eccentric compared with concentric contractions.
Furthermore, Ryschon et al. (57) estimated that the ATP
production rate per work output was 20% greater during
submaximal eccentric vs. concentric contractions of tibialis
anterior in humans, using 31P-nuclear magnetic resonance
imaging. Beltman et al. (6) reported that energy cost (i.e.,
high-energy phosphate consumption) was 70% lower during
maximal eccentric vs. concentric contractions of medial gas-
trocnemius of rats. Thus, it seems that eccentric contractions
have a lower muscular ATP demand, which has been typically
shown indirectly through the measurement of muscle tissue
oxygen uptake and ATP synthesis/resynthesis markers.
Taken together, it has been proposed that the lower energy
cost of eccentric contractions could be explained by 1) the
nonadenosine triphosphate (ATP)-dependent mechanical
detachment of the actin-myosin cross-bridges (28, 40); 2)
the spring-like properties of the muscle-tendon unit allow-
ing the storage and reuse of elastic potential energy (39),
and more recently sarcomeric structural proteins, that en-
hance force production and which may not be dependent on
ATP hydrolysis (27); and 3) the lower muscle fiber recruit-
ment at a given level of muscle force (14, 43). However, no
previous study has systematically investigated these possible
mechanisms simultaneously. Therefore, in the present study,
we compared the muscle-tendon behavior, muscle oxygenation
of VL, and muscle activity of VL and its agonists and one
antagonist (biceps femoris) between eccentric and concentric
cycling to better understand the factors contributing to the
lower metabolic cost of eccentric compared with concentric
cycling. It was assumed that SSC contribution would be
minimal in cycling, and thus the comparison between eccentric
and concentric cycling would be ideal to understand the char-
acteristics of eccentric vs. concentric contractions better. It was
hypothesized that eccentric cycling would elicit smaller
changes in muscle oxygenation, lesser muscle activity, and
greater strain of the tendinous tissue than concentric cycling
J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
885
and that a combination of these factors would explain the lower
metabolic cost of eccentric cycling.
METHODS
Participants. Approval was obtained from the Institutional Human
Research Ethics Committee before the study. Eleven healthy, physi-
cally active young men who had not performed lower limb resistance
training regularly in the 6 mo before the study and who reported no
history of neurological disorder or orthopedic lower limb injuries
were recruited from the university community. Their average (⫾ SD)
age, height, body mass, and body mass index were 27.1 ⫾ 7.0 yr,
178.4 ⫾ 8.2 cm, 73.6 ⫾ 6.2 kg, and 23.1 ⫾ 1.3 kg/m⫺2, respectively.
They completed an informed written, consent form and a medical
questionnaire before participating in the study. The sample size was
calculated on the basis of an ␣-level of 0.05 and a power (1-␤) of 0.8,
with an estimated 20% difference in VL fascicle length change
between concentric and eccentric maximal isokinetic contractions
based on a previous study (21). Participants were instructed not to
perform any exercise, take anti-inflammatory medication, or apply
any treatments (e.g., massage, stretching) from 2 days before each
cycling bout.
Study design. The present data were collected in a larger study, and
physiological and muscle damage profiles during two bouts of eccen-
tric cycling have been published previously (49). Participants reported
to the laboratory on four occasions: the first visit for an incremental
exercise test in a recumbent cycle ergometer to determine their power
output-V̇O2 relationship during concentric cycling, the second visit for
a concentric cycling bout (CONC) 2–3 days later, and the third and
fourth visits for an eccentric cycling bout separated by 2 wk. The
participants were more familiarized with eccentric cycling in the
second eccentric bout than in the first, and fewer coordination prob-
lems were observed for the second than for the first bout. Thus, the
second eccentric cycling (ECC) bout was used in the present study for
comparison with CONC. It was assumed that the initial eccentric
cycling bout familiarized the participants with the eccentric ergometer
and eccentric cycling. The participants performed 10 min of cycling at
the same work output for CONC and ECC cycling, and oxygen
consumption (V̇O2) and heart rate (HR) were assessed during the
cycling. Furthermore, muscle oxygenation and hemodynamics, VL
fascicle behavior and the surface electromyogram (EMG), crank
torque, and knee joint angle were recorded as detailed below. Because
of the limited number of probes available, VL muscle-tendon behavior
was examined and assumed to be representative of the mono-articular
vasti muscles in the quadriceps, as described previously (49).
Cycling exercise. An incremental cycling test to determine V̇O2peak
was performed using a recumbent ergometer (Tunturi F30R). The test
started at a power output of 50 W for 4 min and increased 25 W every
minute until volitional exhaustion. Cadence was kept at 60 rpm, and
the participants received verbal encouragement during the test. The
concentric cycling (CONC) was performed in the same recumbent
ergometer as that for the incremental cycling test for 10 min (60 rpm)
at 65% of the maximal power output (POmax) obtained during the
V̇O2peak test. The eccentric cycling (ECC) was performed using a
recumbent cycle ergometer fitted with a motor that moved the cranks
of the ergometer backward at a selected cadence (Eccentric Trainer,
Metitur, Finland). Subjects were given the instruction to “smoothly
resist the backward movement of the pedals,” and thus eccentric
contractions were performed mainly by the knee extensor muscles.
Both ergometers were equipped with flat pedals, and thus the subjects
could not actively pull upward during the exercise. A 5-min familiar-
ization period was performed at a low power output (50 W), and this
was completed ⬃3 min before each 10-min cycling bout.
Synchronization process. Fascicle length, crank torque, EMG, knee
joint angle, and NIRS data were collected simultaneously. Both cycle
ergometers were equipped with a switch (crank angle of 270°) that
sent a pulse to the ultrasound apparatus and the PowerLab system,
886 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.
allowing the synchronization of ultrasound, EMG, crank torque, knee
joint angle, and NIRS data. The data collected at different frequencies
(ultrasound collected at 47 Hz, crank torque collected at 477 Hz, and
EMG and knee joint angle collected at 2,000 Hz) were resampled at
200 Hz to allow multiple calculations over each revolution, and data
presentation as done in a previous study (21). After synchronization,
ultrasound video clips, EMG, knee joint angle, and crank torque data
for the first three full revolutions at 1 and 10 min of cycling were
analyzed. No significant differences were found between 1- and
10-min results for all variables. Thus, 1- and 10-min values (200 data
points over an averaged revolution) were averaged as representation
of the whole corresponding cycling bout for further comparisons.
Near-infrared spectroscopy. Near-infrared spectroscopy (NIRS)
measures O2-dependent absorption of hemoglobin (Hb) in the microcir-
culation blood vessels (i.e., arterioles, capillaries, and venules) and
myoglobin in the muscle cytoplasm. A NIRO-200 oximeter (Hamamatsu
Photonics, Hamamatsu, Japan) shows concentration changes in oxy-
genated-Hb (O2Hb), deoxygenated Hb (HHb), and total Hb volume
(tHb ⫽ O2Hb ⫹ HHb) expressed in ␮M/cm as well as an absolute
measure of O2Hb saturation represented as tissue oxygenation index
(TOI ⫽ O2Hb/tHb ⫻ 100) expressed in percent. TOI reflects the dy-
namic balance of O2 supply by the muscle microcirculation and O2
consumption/demand of the muscle, and the changes in tHb can be
considered as an indirect measure of changes in local muscle blood
flow/O2 supply (20).
The NIRO-200 optical probe unit consists of one emitter (with 3
laser-emitting diodes of 775, 810, and 850 nm) and one detector (with
2 silicon photodiodes separated by a 6-mm center-to-center distance)
separated by a center-to-center distance of 4 cm. The probe was
supported by a black rubber shell and firmly attached to the skin at the
VL muscle belly of the right leg parallel to the major axis of the thigh
by double-sided adhesive tape and covered by a soft black cloth, and
all wires were taped down to the skin to minimize movements during
exercise. Before each testing session, an initialization procedure on
the NIRO-200 was carried out, which set each laser power automat-
ically to establish the optimum measurement conditions. The zero-set
procedure was also adopted to return the tHb to the zero value, but this
did not affect the TOI values. The NIRO-200 sampling frequency was
6 Hz, and NIRS data were collected simultaneously with EMG and
ROM data (see below) into a PowerLab system (ADInstruments) and
stored on the computer for later analysis. The baseline of TOI and tHb
was determined as the mean value ⬎6 s before the start of the cycling.
Measurements are presented as a 1-min average of the TOI absolute
values (TOI; %) and difference from the baseline in tHb (⌬tHb).
Statistical analyses were made based on the 1-min average of TOI and
tHb during cycling.
Crank torque and knee range of motion. Crank torque was mea-
sured continuously by a KIP power meter bicycle crank (Breakaway
Innovations) implemented to the left crank of the eccentric and
concentric bike, and sampling at a frequency of 477 Hz. The data were
stored in a data logger, and power output was instantaneously calcu-
lated and displayed on a power meter device (Rider 50, Bryton,
Taiwan). Knee joint range of motion was continuously measured by
an electronic goniometer (SG; Biometrics) and collected at 2,000 Hz
in the PowerLab system.
Electromyography. Surface electromyography (EMG) was re-
corded from vastus lateralis (VL), rectus femoris (RF), vastus medi-
alis (VM), and biceps femoris (BF) muscles during cycling. The skin
was shaved and cleaned with alcohol swabs, and electrodes were
placed at two-thirds of the distance from the anterior superior iliac
spine (ASIS) to the patella for VL, at 50% of the distance between
ASIS and the patella for RF, at 80% of the distance between ASIS and
the patella for VM, and at 50% of the distance between the ischial
tuberosity and the lateral condyle of tibia according to the SENIAM
guidelines. A Bagnoli-8 desktop EMG system (Delsys) was used to
record EMG data using a bipolar electrode configuration (DE-2.1
SEMG sensor) with a 10-mm interelectrode distance. Sampling fre-
J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
quency was set at 2,000 Hz, and a 10- to 450-Hz digital band-pass
filter was applied. EMG data were smoothed using a root mean square
(RMS) algorithm with a 50-ms window to produce a linear envelope
according to a previous study (41). Values are presented as the
percentage of EMG peak amplitude obtained during a maximal
voluntary isometric contraction of the knee extensors and flexors
performed at 70° of knee flexion on an isokinetic dynamometer
(Biodex, System 3) before each cycling bout. Before maximal volun-
tary isometric contraction (MVC) torque of the left knee extensors
was measured, each participant performed a 5-min warmup on a cycle
ergometer (Monark 828E) with a load of 1 kp at 60 rpm. Once seated
in the chair of the dynamometer, each participant performed three
submaximal contractions at 50, 50, and 80% of the perceived MVC
for 3 s each, with 1 min of rest between contractions, followed by
three MVCs with a 1-min rest between contractions. All participants
were instructed to contract as fast and hard as possible, and visual
feedback was provided as a real-time display of the torque output on
a computer screen. The maximum value of the three measurements
was used for further analysis.
Muscle and tendon behavior. An ultrasound probe (60 mm, Model
UST-5712, 7.5 MHz, B-mode, Aloka SSD ␣-10) was attached to the
left midthigh proximal to the VL EMG electrode (explained above) by
surgical tape, and the probe was supported by a custom-made flexible
plastic cast (size: 10 ⫻ 20 cm, surrounding the leg and secured by
Velcro straps). Echo absorptive tape was used to ensure that no probe
displacement occurred during the 10-min cycling bouts. The ultra-
sound video clips were captured at a sampling rate of 47 Hz (i.e., 47
frames/pedal revolution) and analyzed offline with video analysis
software (Tracker 4.7, free download: http://www.cabrillo.edu/
~dbrown/tracker/). The superior and inferior aponeuroses and a
clearly identifiable fascicle (i.e., visible echo from the fascicle inter-
spaces that could be followed across the image) were identified to track
the same fascicle throughout the movement and digitized frame by frame
assuming a linear continuation of the fascicles. For each subject, the
entire length of the VL fascicle (Lf) was estimated using the trigonomet-
ric method (21) because Lf could not be visualized throughout the
pedaling cycle. The error for estimating Lf with this method has been
reported to be 2–7% (21, 22), and the methodology of the measure has
previously been published elsewhere (49). Since no significant difference
in the change in the fascicle length was evident, and considering the time
constraints of manual frame-by-frame fascicle length analysis, the aver-
age of three consecutive revolutions was used to represent the average
length change at 60 rpm. The fascicle length and angle data were filtered
using a smoothing spline method (61).
The length of tendinous tissue (LTT) was defined as the sum of the
proximal and distal tendinous structures and aponeuroses (29). Length
change in the tendinous tissue (tendon and aponeurosis; LTT) was
calculated using the formula LTT ⫽ LMTU – Lf ⫻ Cos ␪, where LMTU
is the MTU length and ␪ is the angle between fascicle and deeper
aponeurosis, as reported in previous similar studies for VL (29, 49).
LMTU changes were estimated using previously derived models on the
basis of the knee joint position (angle) data and the limb length of
each subject using the Hawkins and Hull (25) model.
Statistical analysis. A paired t-test was used to compare the average
power output, V̇O2, and HR during cycling between CONC and ECC.
Because no significant differences in fascicle length changes, EMG,
crank torque, or knee angle were evident between 1 and 10 min of
cycling by a two-way repeated-measures analysis of variance
(ANOVA), values at 1 and 10 min were averaged to represent the
whole exercise to compare between CONC and ECC. After synchro-
nization of crank torque, EMG, muscle fascicle length, and tendinous
tissue (TT) length data, the average pattern of changes (means ⫾ SD)
within the pedal revolution was obtained (see Fig. 3). Paired t-tests
were used to compare the changes in fascicle and tendinous tissue
length, average fascicle angle, peak cycling torque, and peak and
mean EMG amplitudes (normalized to the EMG measured during
MVC) between CONC and ECC. A two-way repeated-measures
 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al. 887
ANOVA was used to compare changes in tHb and TOI over time
(immediately before and 1-min average during the 10 min of cycling)
between CONC and ECC. If a significant main effect was found, a
Fisher’s least significant difference test was followed. The signifi-
cance level was set at P ⬍ 0.05. All statistical analyses were
performed with PASW Statistics 21 software for Mac (SPSS, IBM).
Data are presented as means ⫾ SD.

RESULTS

Cycling power output, oxygen consumption, and heart rate.

In the incremental cycling test, the participants’ average
(⫾ SD) maximal concentric power output, peak heart rate, and
peak oxygen consumption were 286.4 ⫾ 62.6 W, 188.9 ⫾ 7.0
beats/min, and 3.3 ⫾ 0.6 l/min, respectively. During the 10-
min cycling bout, the average cycling power output was not
different (P ⫽ 0.07) between ECC (194.0 ⫾ 48.5 W) and
CONC (188.6 ⫾ 40.8 W). As shown in Fig. 1, the average
oxygen consumption during ECC (0.96 ⫾ 0.24 l/min) was
64.8 ⫾ 7.4% lower (P ⬍ 0.001) than that during CONC
(2.81 ⫾ 0.53 l/min). The average heart rate during ECC
(106.9 ⫾ 17.9 beats/min) was 34.7 ⫾ 8.9% lower (P ⬍ 0.001)
than during CONC (163.7 ⫾ 15.4 beats/min).
Muscle oxygenation and hemodynamics. Figure 2 shows
changes in tHb and TOI during the 10-min cycling bouts. tHb
decreased from baseline similarly for ECC and CONC soon
after the onset of cycling (P ⬍ 0.001) without a significant
difference between conditions (P ⫽ 0.13). TOI decreased only
during CONC and remained on average 16.3 ⫾ 6.1% lower
during CONC than during ECC (P ⬍ 0.001). Fig. 2. Changes (1-min average) in total hemoglobin (tHb) from preexercise
Knee joint range of motion and peak crank torque. The knee values (A) and absolute changes in total oxygenation index (TOI) from
joint range of motion during cycling was not different between preexercise level (Pre) during 10-min concentric (CONC) and eccentric (ECC)
ECC and CONC (P ⫽ 0.28) and ranged from 17.2 ⫾ 8.9 to cycling.
119.2 ⫾ 5.1° (0° representing full extension). Figure 3A shows
the average crank torque during a complete cycle revolution for
CONC and ECC. Torque increased during the pushing phase 0.31 ⫾ 0.14 mV, P ⫽ 0.62), RF (0.28 ⫾ 0.11 and 0.28 ⫾ 0.16
and reached the peak at ⬃70° knee flexion for ECC and 40° mV, P ⫽ 0.98), VM (0.25 ⫾ 0.14 and 0.27 ⫾ 0.10 mV, P ⫽
knee flexion for CONC and decreased in the recovery phase. 0.38), or BF (0.22 ⫾ 0.15 and 0.23 ⫾ 0.11 mV, P ⫽ 0.81).
Positive peak torque generated during ECC (75.2 ⫾ 19.4 Nm) Figure 4 shows changes in EMG amplitude normalized to the
was 32.3 ⫾ 20.5% greater (P ⫽ 0.006) than that during CONC peak EMG amplitude during MVC for a complete revolution
(50.2 ⫾ 10.6 Nm), and negative peak torque generated during for ECC and CONC. Muscle activity increased during the
ECC (14.4 ⫾ 4.4 Nm) was 47.5 ⫾ 24.3% greater (P ⫽ 0.002) revolution with increasing torque, and EMG amplitude peaked
than that during CONC (7.3 ⫾ 3.7 Nm), as shown in Fig. 3B. at ~70° knee flexion for VL, RF, and VM during ECC and at
EMG. Isometric MVC torques were not different between ⬃100° for CONC, followed by gradual decreases during the
CONC and ECC for the knee extensors (263.6 ⫾ 55.7 and recovery phase. EMG peak amplitude during cycling (relative
275.3 ⫾ 47.9 Nm, respectively, P ⫽ 0.4) or knee flexors to the amplitude during MVC) was significantly lower (P ⬍
(117.3 ⫾ 24.3 and 122.5 ⫾ 26.5 Nm, respectively, P ⫽ 0.6). 0.001) for ECC (VL: 17.3 ⫾ 8.7%; VM: 15.4 ⫾ 4.4%; RF:
Peak EMG amplitude during MVC before exercise was not 13.7 ⫾ 7.3%) than for CONC (VL: 41.1 ⫾ 17.6%; VM:
different between CONC and ECC for VL (0.29 ⫾ 0.13 and 37.7 ⫾ 23.6%; RF: 29.4 ⫾ 20.8%). BF EMG peak amplitude
was also significantly lower (P ⬍ 0.001) for ECC (4.6 ⫾ 2.6%)
than for CONC (22.8 ⫾ 13.1%). Similarly, mean EMG ampli-
tude during cycling (relative to the amplitude during MVC)
was significantly lower in ECC for most of the muscles
assessed (VL: 5.2 ⫾ 2.9%, P ⫽ 0.003; VM: 5.6 ⫾ 1.4%, P ⫽
0.004; RF: 5.4 ⫾ 2.9%, P ⫽ 0.3) than in CONC (VL:
10.8 ⫾ 4.7%; VM: 15.6 ⫾ 7.2%; RF: 7.9 ⫾ 6.5%). BF EMG
mean amplitude was also significantly lower (P ⬍ 0.001) in
ECC (1.3 ⫾ 1.3%) than in CONC (8.8 ⫾ 5.2%).
Fascicle and tendinous tissue behavior. The coefficient of
Fig. 1. Average oxygen consumption (A) and heart rate (B) during concentric
variation of the fascicle length for the six revolutions used for
(CONC) and eccentric (ECC) cycling. *Significantly different from CONC. the analyses was 5.6 ⫾ 1.4% (range: 4.6 – 9.7%), which was
small; thus the average of the six revolutions was used for

J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
888 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.

Fig. 3. Changes (means ⫾ SD) in crank torque

(A), muscle-tendon unit length (C), vastus
lateralis (VL) fascicle length (E), and tendi-
nous tissue (TT) length (G) in relation to knee
joint angle during 1 pedal revolution and pos-
itive and negative peak crank torque (B), max-
imal change in muscle-tendon unit length (D),
and maximal change in fascicle (F) and TT
(H) lengths during concentric (CONC) and
eccentric (ECC) cycling; 0° of knee flexion
represents knee full extension. The standard
deviations (SD) of CONC and ECC are shown
by dark and light gray shadows, respectively,
in A, C, E, and G. *Significantly (P ⬍ 0.05)
different from CONC.

further analyses. Fascicle length increased as the knee joint
angle increased, resulting in maximum fascicle length at
112.9 ⫾ 4.6 and 117.8 ⫾ 7.0° of knee flexion during ECC and
CONC, respectively, without a significant difference between
the two (P ⫽ 0.1). No significant difference between ECC and
CONC was evident for the knee joint angle at which the
fascicle length was maximal (P ⫽ 0.23), and fascicle length
changed from 8.5 ⫾ 1.9 to 15.8 ⫾ 3.9 cm during ECC and
from 10.1 ⫾ 2.8 to 17.8 ⫾ 3.9 cm during CONC, with a
significant difference between ECC and CONC at minimum
and maximum lengths (P ⫽ 0.01 and P ⫽ 0.01, respectively).
The magnitude of fascicle length change (i.e., the difference
between the minimum and maximal fascicle lengths) was
similar (P ⫽ 0.54) between ECC and CONC, as shown in
Fig. 3F. Peak torque was achieved at a greater (P ⬍ 0.001)
knee joint angle in ECC (71.6 ⫾ 10.4°) than in CONC
(45.3 ⫾ 6.9°). Peak torque was generated at similar (P ⫽
0.7) fascicle length during ECC (13.5 ⫾ 3.1 cm) than
CONC (12.7 ⫾ 2.8 cm). Fascicle angle changes were simi-
lar (P ⫽ 0.65) between ECC and CONC and increased
significantly from 7.8 ⫾ 3.1 to 13.3 ⫾ 4.6° during ECC
(P ⬍ 0.001) and from 7.8 ⫾ 1.6 to 13.1 ⫾ 2.7° during
CONC (P ⬍ 0.001).
TT length increased as the knee joint moved toward maxi-
mum flexion, as shown in Fig. 3G. The maximum TT length
changes were not significantly different (P ⫽ 0.45) between
ECC (10.1 ⫾ 3.3 cm) and CONC (10.3 ⫾ 2.5 cm), as shown in
Fig. 3H.

J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.
DISCUSSION
The present study tested the hypothesis that muscle oxygen-
ation changes and muscle activity of the knee extensors and
flexors would be less and the use of the tendinous tissue greater
during eccentric than during concentric cycling. The results
revealed a 65% lower oxygen consumption in ECC than in
CONC, which was associated with a 16% smaller muscle
deoxygenation and 16 –24% lower muscle activity of VL, RF,
VM, and BF during ECC than during CONC. However, the
magnitudes of fascicle and tendinous tissue length changes
were similar between ECC and CONC. Thus, muscle-tendon
behavior differences do not appear to explain the different
metabolic demands of ECC and CONC. Hence, the hypothesis
was only partly supported, with the lower metabolic demand of
eccentric cycling resulting primarily from the smaller knee
extensor and flexor muscle activations and lesser muscle de-
oxygenation.
Oxygen consumption and heart rate were 65 and 35% lower,
respectively, during ECC compared with CONC, which is in
line with several previous studies reporting a lower metabolic
demand during eccentric vs. concentric cycling (7, 48, 50). The
smaller difference found in heart rate (35%) vs. oxygen con-
sumption (65%) has been also reported in a previous study
(16). The heart rate response during ECC may have been
influenced by inputs from sympathetic (catecholamine) and
parasympathetic systems as well as peripheral mechanical
(joints and muscles) and chemical receptors (metaboreceptors)
(2). However, such regulation did not specifically affect the
oxygen consumption, since this is a measure of the oxygen
utilized by the working muscles and is thus less influenced by
afferent inputs (2). Dufour et al. (16) speculated that the
relatively higher HR during eccentric vs. concentric cycling
performed at a similar metabolic power could originate from an
elevated muscle temperature that induced HR-mediated ele-
vated cardiac output either by thermoregulation activation or
J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
889
Fig. 4. Average muscle activity[electromyogram
(EMG) amplitude] of vastus lateralis (A), vastus
medialis (B), rectus femoris (C), and biceps femoris
(D) during 1 full cycle revolution (x-axis; knee joint
range of motion) for concentric (CONC; dashed
line) and eccentric (ECC; solid line) cycling as
%EMG activity measured during maximal volun-
tary isometric contraction (MVC); 0° of knee flex-
ion represents knee full extension.
by thermal response of group III and IV muscle afferents. It is
important to note that heart rate does not necessarily represent
metabolic demand during eccentric cycling, and oxygen utili-
zation is a more reliable measure of skeletal muscle metabo-
lism and energetic demand during exercise.
In the present study, the total muscular hemoglobin (tHb)
level, an indirect measure of blood flow, was similar during
ECC and CONC (Fig. 2A). As shown also by Clifford and
Hellsten (11), the similar decrease in tHb during ECC and
CONC suggests that intramuscular pressure during the contrac-
tion phase was sufficiently high as to occlude blood vessels and
reduce the volume of blood in the muscle under both condi-
tions, and the circulation returns to the muscle during the
subsequent relaxation phase. Ruiter et al. (56) showed that
muscle reoxygenation was stopped during isometric contrac-
tion at 35% of MVC. The level of muscle contractions pro-
duced during cycling in the present study was ⬃30% of MVC,
as shown in a previous study (49). Thus, it is possible that
CONC and ECC cycling progressively reduced the reoxygen-
ation capacity of VL throughout the 10-min cycling bout.
However, the dynamic contractions performed during cycling
may have less of an effect on blood occlusion. Muscle tissue
oxygenation (i.e., TOI) was 16% greater during ECC than
during CONC (Fig. 2B), which is in line with a previous study
(44) showing a lower muscle tissue deoxygenation during
maximal isokinetic eccentric vs. concentric contractions of the
elbow flexors. It is generally accepted that the use of oxygen by
the muscle reflects ATP synthesis during submaximal exercise
(58). Thus, the maintained muscle oxygenation for the same
blood supply during muscle contraction in the eccentric cycling
may indicate a lesser ATP utilization at the muscle level. In
fact, Ortega et al. (45) have shown that ATP cost of force
production in concentric contractions is ⬎50% greater than
that of eccentric contractions of the first dorsal interosseous. It
appears that the difference in the whole body oxygen consump-
890 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.
tion (65%) between ECC and CONC reflects the difference in
ATP cost. However, the difference in TOI between ECC and
CONC was only 16%. It is speculated that NIRS measurements
in VL reflect only a portion of the oxygen utilization by the
whole knee extensor group during cycling. Elmer and Martin
(18) showed that quadriceps muscle produced 54% of the force
generated during eccentric cycling and that hip extensors and
ankle flexors were also involved in this exercise. Thus, it is
likely that these muscle groups also contributed to the greater
differences in whole body oxygen consumption. Furthermore,
it may be that other regions of the body such as knee flexor
muscles also worked more during CONC than during ECC
(Fig. 4). If all of these muscles are considered, the difference in
TOI between ECC and CONC may explain the 65% difference
in oxygen consumption.
Although the average power output was similar between
ECC (194.0 ⫾ 48.5 W) and CONC (189.7 ⫾ 43.2 W), the peak
positive and negative crank torques produced per revolution
were 32 and 48% greater during ECC than during CONC,
respectively (Fig. 3, A and B). It was surprising to find this
difference since the participants were asked to maintain a
constant workload that was matched between ECC and CONC
(65% of concentric POmax). Because EMG amplitude did not
change over time, and power output is the product of the force
applied to the pedal and the change in angular distance per unit
of time, it is possible that participants resisted harder the
backward movement of the pedals (greater positive work) to
overcome the negative torque produced during ECC and,
therefore, to perform similar targeted power output (65%
power output of CONC POmax). This may explain the similar
average torque during ECC and CONC cycling (Fig. 3A). It is
known that muscles can generate greater force during eccentric
than concentric, such that MVC force is 28 – 40% greater in
eccentric than concentric contractions of the knee extensors
(41). LaStayo et al. (38) reported that eccentric cycling could
produce four to seven times greater power output compared
with concentric cycling when performed at the same relative
intensity (i.e., 65% HRpeak). These greater force and power
production capacities in eccentric contractions have been pos-
tulated to be related to the spring-like properties of structural
proteins of the sarcomeres and tendinous tissue (26) and
specific neural control strategies (14, 19). From our data, it
seems that the positive torque is greater during eccentric
cycling, although average power output is matched to concen-
tric cycling. This could influence muscle mass and strength
gains shown after eccentric cycling training due to greater
mechanical stimuli.
It is well known that energy is stored in the lengthened
muscle-tendon unit (as potential elastic energy) during eccen-
tric actions and subsequently utilized (recoil) in the subsequent
propulsive movement phase (i.e., concentric) during activities
that involve a stretch-shortening cycle, such as running and
jumping (8). This would allow the muscle to function more
efficiently (consume less oxygen) and generate greater muscle
force, especially when fast eccentric contractions are per-
formed (62). Furthermore, previous studies showed that muscle
fascicles behaved quasi-isometrically during maximal isoki-
netic (10, 53), submaximal, isotonic (60) eccentric contractions
in humans and maximal eccentric contractions in animal mod-
els (3, 54). However, such behavior was not observed during
ECC in the present study, which was possibly due to eccentric
J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
cycling not having a stretch-shortening cycle as in other ex-
perimental models; therefore, it may be considered as a purely
eccentric exercise. Furthermore, it should be noted that eccen-
tric contractions during ECC were submaximal, as indicated by
EMG activities that were ⬍20% of those of maximal voluntary
contraction and probably ⬃30% of the force production mea-
sured during maximal eccentric contractions (49). This could
explain that muscles were lengthened during ECC in a similar
extent to that of CONC, thus resulting in similar changes in
fascicle and TT lengths for the same range of motion (Fig. 3,
C–H). It should be noted that the data of ECC analyzed in the
present study was from the second ECC bout after the partic-
ipants performed the same ECC 2 wk earlier. We have reported
previously that muscle fascicles were elongated less during the
second than during the first ECC (49). It is possible that
dynamic and/or passive muscle stiffness increased after the
first bout of ECC, which contributed to the smaller minimum
and maximum fascicle lengths for ECC than for CONC (Fig.
3E). It is possible that the current ultrasound technique was not
sensitive enough to detect small changes in fascicle (and thus
tendon) behavior, and we cannot provide information on po-
tential differences at the sarcomere level. Nevertheless, the
results of the present study indicate that the utilization of the
stored energy in the tendon was not a major factor influencing
the lower energy cost of ECC. However, the possibility exists
that some energy storage during the lengthening phase of the
muscle contraction could occur within muscle tissue. Recently,
some research groups (e.g., see Refs. 27 and 42) have sug-
gested that the titin protein works as a molecular spring to
enhance muscular force production during active contractions
without additional requirement for ATP, and it is possible that
the role of titin may indeed partly explain the energetic effi-
ciency of eccentric contractions.
It should be noted that a low level of EMG activity (mean
and peak EMG amplitude) was observed throughout the full
pedal stroke in ECC cycling (Fig. 4). Thus, it is possible that
stretching of active muscles occurred during the pushing phase,
and recoil of it contributed to mechanical work during the
phase when no positive torque was produced. However, this
may provide only a small force due to the relatively small
number of activated cross-bridges. Although the storage of
elastic energy would require some energy to be retained by
active cross-bridge formation, and any detachment of cross-
bridges would dissipate energy from the muscle-tendon unit
(55), actomyosin cross-bridge detachment occurs with only
small changes in muscle length (e.g., ⬍2% strain; see Ref. 52),
and the force required to trigger detachment is lower under
conditions of slower stretch velocity (23, 51, 56). Thus, it
seems possible that force generated during eccentric contrac-
tions may come from the stretch of sarcomeres. In the present
study, muscle fascicles were lengthened ⬃70% during eccen-
tric cycling (from ⬃9 to 15 cm; Fig. 3). Under slow-velocity,
low-intensity eccentric contractions, the majority of muscle
length change should occur through active cross-bridge cycling
rather than forced muscle lengthening, and thus elastic energy
storage might be expected to be minimal. It is possible that
mechanisms contributing to a residual force enhancement
(RFE) were associated with the lower energy requirement
during ECC. The RFE phenomenon has been observed in
studies on single sarcomeres as well as whole muscles (59),
and it enhances force following stretch while the metabolic
 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.
energy consumption per unit force is small (34). Some of the
mechanisms proposed to explain RFE include Ca2⫹-dependent
increases in titin stiffness as well as increases in the number of,
and force produced by, active cross-bridges (59). RFE has also
been observed during submaximal voluntary contractions (46),
so it is possible that titin plays a role in the energetic efficiency
of eccentric contractions.
A lower muscle activity during eccentric actions is also
postulated to explain the lower metabolic cost of eccentric vs.
concentric cycling (7). Enoka (19) showed that eccentric con-
tractions require a unique motor unit recruitment pattern,
which may require a different input strategy from the central
nervous system when compared with concentric contractions.
We found that mean EMG amplitudes of knee extensor (ago-
nist) muscles were 6 (VL) and 10% (VM) smaller during ECC
than during CONC. These results are consistent with those of
Beltman et al. (5), who examined muscle activity during
maximal isokinetic contractions (60°/s) using the interpolated
twitch technique and found that knee extensor voluntary acti-
vation was significantly lower during eccentric (79%) com-
pared with concentric (92%) contractions despite torque pro-
duction being 30% greater for eccentric contractions. More-
over, in the present study, the antagonist (i.e., coactivated) BF
muscle was 7.5% less active during ECC than during CONC
cycling (Fig. 4D), indicating that maximal coactivation by BF
ranged between 4 and 5% of MVC. Coactivation levels have
been reported to range from 10 to 55% of the EMG activity of
the hamstring muscle when working as an antagonist in iso-
metric (24) and dynamic contractions (36). Furthermore, EMG
amplitude has often been reported to be lower during maximal
eccentric contractions than during maximal concentric and
isometric contractions (7, 35). However, some studies have not
found any differences in EMG amplitude between eccentric
and concentric contractions (4, 15). These discrepancies may
be related to the muscle length, joint angular velocity, and type
of contraction (isotonic, isokinetic, etc.) assessed or to the
phase of the movement used for analysis. For instance, we have
recently shown that EMG amplitude was different between an
initial and a secondary bout of eccentric exercise during the
recovery phase, but not in the propulsive phase of eccentric
cycling pedal stroke (49). The smaller muscle activity (peak
and mean EMG amplitudes) in each revolution in ECC vs.
CONC indicates that muscle activation during ECC was much
less than in CONC. Thus, the lower muscle activation of
antagonist muscles during eccentric contractions may have also
contributed to the reduced metabolic demand of eccentric
cycling.
The present study was not exempt from limitations. More
precise measurement of movement patterns of all joints in-
volved could add information regarding movement coordina-
tion and smoothness of the cycling technique of subjects.
When comparing with CONC, ECC requires more practice to
rotate pedals smoothly. In the present study, we have analyzed
an eccentric cycling bout after a short familiarization session
plus a full eccentric exercise session, as reported by Peñailillo
et al. (49); evidence that subjects were more familiar in this
second full session includes the lower EMG and metabolic
parameters shown in our previous study. However, it is possi-
ble that even two practice bouts were insufficient for all
subjects to perform eccentric cycling smoothly, and thus an
investigation of muscle function in experts might prove in-
J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
891
sightful in the future. In fact, some concentric contractions
indicated by the negative torque produced in the recovery
phase during ECC were observed, but the level of muscle
activity was low. This was likely due to the difficulty of muscle
coordination during ECC even during the second bout. We
speculate that when participants become expert at eccentric
cycling, no concentric contractions would be performed, and
this may result in greater differences between eccentric and
concentric cycling than those presented in the current study.
In conclusion, the lesser activity of agonist and antagonist
muscles in addition to the lower muscle tissue oxygen utiliza-
tion underpin the lower metabolic cost of eccentric compared
with concentric cycling, whereas changes in the muscle and
tendon lengthening appear to have a negligible effect. Future
research should utilize more direct muscle energetic measure-
ments from muscle samples or through the use of nuclear
magnetic resonance spectroscopy to confirm these findings,
which would allow a more detailed understanding of the
differential muscle adaptations to eccentric vs. concentric cy-
cling training. It has been shown previously that eccentric
cycling training produces greater adaptations to skeletal mus-
cles than concentric cycling training (38) or weight training
(37); however, mechanisms underpinning the adaptations have
yet to be elucidated. A better understanding of the physiolog-
ical and biomechanical characteristics of eccentric cycling will
shed light on these mechanisms.
ACKNOWLEDGMENTS
We acknowledge Drs. Julien Louis and Eduardo Saez-Saez de Villareal for
their help in data collection and analysis.
GRANTS
This work was partially funded by a research grant awarded to L. Peñailillo
(no. 11150293) by FONDECYT of Chile.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the authors.
AUTHOR CONTRIBUTIONS
L.P. conceived and designed research; L.P. performed experiments; L.P.
analyzed data; L.P., A.J.B., and K.N. interpreted results of experiments; L.P.
prepared figures; L.P., A.J.B., and K.N. drafted manuscript; L.P., A.J.B., and
K.N. edited and revised manuscript; L.P., A.J.B., and K.N. approved final
version of manuscript.
REFERENCES
1. Abbott BC, Bigland B, Ritchie JM. The physiological cost of nega-
tive work. J Physiol 117: 380 –390, 1952. doi:10.1113/jphysiol.1952.
sp004755.
2. Amann M, Blain GM, Proctor LT, Sebranek JJ, Pegelow DF, Demp-
sey JA. Group III and IV muscle afferents contribute to ventilatory and
cardiovascular response to rhythmic exercise in humans. J Appl Physiol
(1985) 109: 966 –976, 2010. doi:10.1152/japplphysiol.00462.2010.
3. Azizi E, Roberts TJ. Geared up to stretch: pennate muscle behavior
during active lengthening. J Exp Biol 217: 376 –381, 2014. doi:10.1242/
jeb.094383.
4. Babault N, Pousson M, Ballay Y, Van Hoecke J. Activation of human
quadriceps femoris during isometric, concentric, and eccentric contrac-
tions. J Appl Physiol (1985) 91: 2628 –2634, 2001.
5. Beltman JG, Sargeant AJ, van Mechelen W, de Haan A. Voluntary
activation level and muscle fiber recruitment of human quadriceps during
lengthening contractions. J Appl Physiol (1985) 97: 619 –626, 2004.
doi:10.1152/japplphysiol.01202.2003.
6. Beltman JG, van der Vliet MR, Sargeant AJ, de Haan A. Metabolic
cost of lengthening, isometric and shortening contractions in maximally
892 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.
stimulated rat skeletal muscle. Acta Physiol Scand 182: 179 –187, 2004.
doi:10.1111/j.1365-201X.2004.01338.x.
7. Bigland-Ritchie B, Woods JJ. Integrated electromyogram and oxygen
uptake during positive and negative work. J Physiol 260: 267–277, 1976.
doi:10.1113/jphysiol.1976.sp011515.
8. Cavagna GA. Storage and utilization of elastic energy in skeletal muscle.
Exerc Sport Sci Rev 5: 89 –129, 1977. doi:10.1249/00003677-197700050-
00004.
9. Cavagna GA, Franzetti P, Heglund NC, Willems P. The determinants
of the step frequency in running, trotting and hopping in man and other
vertebrates. J Physiol 399: 81–92, 1988. doi:10.1113/jphysiol.1988.
sp017069.
10. Chino K, Mitsukawa N, Kobayashi K, Miyoshi Y, Oda T, Kanehisa H,
Fukunaga T, Fukashiro S, Kawakami Y. The influence of fascicle
behavior on the lack of velocity dependence in eccentric joint torque in
humans: in vivo observation. J Appl Biomech 25: 111–118, 2009. doi:10.
1123/jab.25.2.111.
11. Clifford PS, Hellsten Y. Vasodilatory mechanisms in contracting skeletal
muscle. J Appl Physiol (1985) 97: 393–403, 2004. doi:10.1152/
japplphysiol.00179.2004.
12. Denis R, Bringard A, Perrey S. Vastus lateralis oxygenation dynamics
during maximal fatiguing concentric and eccentric isokinetic muscle
actions. J Electromyogr Kinesiol 21: 276 –282, 2011. doi:10.1016/j.
jelekin.2010.12.006.
13. Dewolf AH, Peñailillo LE, Willems PA. The rebound of the body during
uphill and downhill running at different speeds. J Exp Biol 219: 2276 –
2288, 2016. doi:10.1242/jeb.142976.
14. Duchateau J, Baudry S. Insights into the neural control of eccentric
contractions. J Appl Physiol (1985) 116: 1418 –1425, 2014. doi:10.1152/
japplphysiol.00002.2013.
15. Duclay J, Martin A. Evoked H-reflex and V-wave responses during
maximal isometric, concentric, and eccentric muscle contraction. J Neu-
rophysiol 94: 3555–3562, 2005. doi:10.1152/jn.00348.2005.
16. Dufour SP, Doutreleau S, Lonsdorfer-Wolf E, Lampert E, Hirth C,
Piquard F, Lonsdorfer J, Geny B, Mettauer B, Richard R. Deciphering
the metabolic and mechanical contributions to the exercise-induced cir-
culatory response: insights from eccentric cycling. Am J Physiol Regul
Integr Comp Physiol 292: R1641–R1648, 2007. doi:10.1152/ajpregu.
00567.2006.
17. Dufour SP, Lampert E, Doutreleau S, Lonsdorfer-Wolf E, Billat VL,
Piquard F, Richard R. Eccentric cycle exercise: training application of
specific circulatory adjustments. Med Sci Sports Exerc 36: 1900 –1906,
2004. doi:10.1249/01.MSS.0000145441.80209.66.
18. Elmer SJ, Martin JC. Joint-specific power loss after eccentric exer-
cise. Med Sci Sports Exerc 42: 1723–1730, 2010. doi:10.1249/MSS.
0b013e3181d60ead.
19. Enoka RM. Eccentric contractions require unique activation strategies by
the nervous system. J Appl Physiol (1985) 81: 2339 –2346, 1996.
20. Ferrari M, Mottola L, Quaresima V. Principles, techniques, and limi-
tations of near infrared spectroscopy. Can J Appl Physiol 29: 463–487,
2004. doi:10.1139/h04-031.
21. Finni T, Ikegawa S, Lepola V, Komi PV. Comparison of force-velocity
relationships of vastus lateralis muscle in isokinetic and in stretch-short-
ening cycle exercises. Acta Physiol Scand 177: 483–491, 2003. doi:10.
1046/j.1365-201X.2003.01069.x.
22. Finni T, Komi PV, Lepola V. In vivo muscle mechanics during locomo-
tion depend on movement amplitude and contraction intensity. Eur J Appl
Physiol 85: 170 –176, 2001. doi:10.1007/s004210100438.
23. Flitney FW, Hirst DG. Cross-bridge detachment and sarcomere ‘give’
during stretch of active frog’s muscle. J Physiol 276: 449 –465, 1978.
doi:10.1113/jphysiol.1978.sp012246.
24. Grabiner MD, Campbell KR, Hawthorne DL, Hawkins DA. Electro-
myographic study of the anterior cruciate ligament-hamstrings synergy
during isometric knee extension. J Orthop Res 7: 152–155, 1989. doi:10.
1002/jor.1100070122.
25. Hawkins D, Hull ML. A method for determining lower extremity muscle-
tendon lengths during flexion/extension movements. J Biomech 23: 487–
494, 1990. doi:10.1016/0021-9290(90)90304-L.
26. Herzog W. Mechanisms of enhanced force production in lengthening
(eccentric) muscle contractions. J Appl Physiol (1985) 116: 1407–1417,
2014. doi:10.1152/japplphysiol.00069.2013.
27. Herzog W. The role of titin in eccentric muscle contraction. J Exp Biol
217: 2825–2833, 2014. doi:10.1242/jeb.099127.
J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.
28. Huxley AF. Muscle structure and theories of contraction. Prog Biophys
Biophys Chem 7: 255–318, 1957.
29. Ishikawa M, Finni T, Komi PV. Behaviour of vastus lateralis muscle-
tendon during high intensity SSC exercises in vivo. Acta Physiol Scand
178: 205–213, 2003. doi:10.1046/j.1365-201X.2003.01149.x.
30. Ishikawa M, Komi PV. Effects of different dropping intensities on
fascicle and tendinous tissue behavior during stretch-shortening cycle
exercise. J Appl Physiol (1985) 96: 848 –852, 2004. doi:10.1152/
japplphysiol.00948.2003.
31. Ishikawa M, Komi PV, Grey MJ, Lepola V, Bruggemann GP. Muscle-
tendon interaction and elastic energy usage in human walking. J Appl
Physiol (1985) 99: 603–608, 2005. doi:10.1152/japplphysiol.00189.2005.
32. Ishikawa M, Niemelä E, Komi PV. Interaction between fascicle and
tendinous tissues in short-contact stretch-shortening cycle exercise with
varying eccentric intensities. J Appl Physiol (1985) 99: 217–223, 2005.
doi:10.1152/japplphysiol.01352.2004.
33. Jones AA, Power GA, Herzog W. History dependence of the electro-
myogram: Implications for isometric steady-state EMG parameters fol-
lowing a lengthening or shortening contraction. J Electromyogr Kinesiol
27: 30 –38, 2016. doi:10.1016/j.jelekin.2016.01.008.
34. Joumaa V, Herzog W. Energy cost of force production is reduced after
active stretch in skinned muscle fibres. J Biomech 46: 1135–1139, 2013.
doi:10.1016/j.jbiomech.2013.01.008.
35. Kay D, St Clair Gibson A, Mitchell MJ, Lambert MI, Noakes TD.
Different neuromuscular recruitment patterns during eccentric, concentric
and isometric contractions. J Electromyogr Kinesiol 10: 425–431, 2000.
doi:10.1016/S1050-6411(00)00031-6.
36. Kellis E, Baltzopoulos V. Muscle activation differences between eccen-
tric and concentric isokinetic exercise. Med Sci Sports Exerc 30: 1616 –
1623, 1998. doi:10.1097/00005768-199811000-00010.
37. LaStayo PC, Ewy GA, Pierotti DD, Johns RK, Lindstedt S. The
positive effects of negative work: increased muscle strength and decreased
fall risk in a frail elderly population. J Gerontol A Biol Sci Med Sci 58:
M419 –M424, 2003. doi:10.1093/gerona/58.5.M419.
38. LaStayo PC, Pierotti DJ, Pifer J, Hoppeler H, Lindstedt SL. Eccentric
ergometry: increases in locomotor muscle size and strength at low training
intensities. Am J Physiol Regul Integr Comp Physiol 278: R1282–R1288,
2000.
39. Lindstedt SL, LaStayo PC, Reich TE. When active muscles lengthen:
properties and consequences of eccentric contractions. News Physiol Sci
16: 256 –261, 2001.
40. Lombardi V, Piazzesi G. The contractile response during steady length-
ening of stimulated frog muscle fibres. J Physiol 431: 141–171, 1990.
doi:10.1113/jphysiol.1990.sp018324.
41. McHugh MP, Tyler TF, Greenberg SC, Gleim GW. Differences in
activation patterns between eccentric and concentric quadriceps contrac-
tions. J Sports Sci 20: 83–91, 2002. doi:10.1080/026404102317200792.
42. Monroy JA, Powers KL, Gilmore LA, Uyeno TA, Lindstedt SL,
Nishikawa KC. What is the role of titin in active muscle? Exerc Sport Sci
Rev 40: 73–78, 2012. doi:10.1097/JES.0b013e31824580c6.
43. Moritani T, Muramatsu S, Muro M. Activity of motor units during
concentric and eccentric contractions. Am J Phys Med 66: 338 –350, 1987.
44. Muthalib M, Lee H, Millet GY, Ferrari M, Nosaka K. Comparison
between maximal lengthening and shortening contractions for biceps
brachii muscle oxygenation and hemodynamics. J Appl Physiol (1985)
109: 710 –720, 2010. doi:10.1152/japplphysiol.01297.2009.
45. Ortega JO, Lindstedt SL, Nelson FE, Jubrias SA, Kushmerick MJ,
Conley KE. Muscle force, work and cost: a novel technique to revisit the
Fenn effect. J Exp Biol 218: 2075–2082, 2015. doi:10.1242/jeb.114512.
46. Oskouei AE, Herzog W. Observations on force enhancement in submaxi-
mal voluntary contractions of human adductor pollicis muscle. J Appl
Physiol (1985) 98: 2087–2095, 2005. doi:10.1152/japplphysiol.01217.
2004.
47. Peñailillo L, Blazevich A, Nosaka K. Energy expenditure and substrate
oxidation during and after eccentric cycling. Eur J Appl Physiol 114:
805–814, 2014. doi:10.1007/s00421-013-2816-3.
48. Peñailillo L, Blazevich A, Numazawa H, Nosaka K. Metabolic and
muscle damage profiles of concentric versus repeated eccentric cy-
cling. Med Sci Sports Exerc 45: 1773–1781, 2013. doi:10.1249/MSS.
0b013e31828f8a73.
49. Peñailillo L, Blazevich AJ, Nosaka K. Muscle fascicle behavior during
eccentric cycling and its relation to muscle soreness. Med Sci Sports Exerc
47: 708 –717, 2015. doi:10.1249/MSS.0000000000000473.
 Low Metabolic Demand of Eccentric Cycling • Peñailillo L et al.
50. Perrey S, Betik A, Candau R, Rouillon JD, Hughson RL. Comparison
of oxygen uptake kinetics during concentric and eccentric cycle exercise.
J Appl Physiol (1985) 91: 2135–2142, 2001.
51. Piazzesi G, Francini F, Linari M, Lombardi V. Tension transients
during steady lengthening of tetanized muscle fibres of the frog. J Physiol
445: 659 –711, 1992. doi:10.1113/jphysiol.1992.sp018945.
52. Pinniger GJ, Ranatunga KW, Offer GW. Crossbridge and non-cross-
bridge contributions to tension in lengthening rat muscle: force-induced
reversal of the power stroke. J Physiol 573: 627–643, 2006. doi:10.1113/
jphysiol.2005.095448.
53. Reeves ND, Narici MV. Behavior of human muscle fascicles during
shortening and lengthening contractions in vivo. J Appl Physiol (1985) 95:
1090 –1096, 2003. doi:10.1152/japplphysiol.01046.2002.
54. Roberts TJ, Azizi E. The series-elastic shock absorber: tendons attenuate
muscle power during eccentric actions. J Appl Physiol (1985) 109:
396 –404, 2010. doi:10.1152/japplphysiol.01272.2009.
55. Roberts TJ, Konow N. How tendons buffer energy dissipation by
muscle. Exerc Sport Sci Rev 41: 186 –193, 2013. doi:10.1097/JES.
0b013e3182a4e6d5.
56. Ruiter CJ, Didden WJ, Jones DA, Haan AD. The force-velocity
relationship of human adductor pollicis muscle during stretch and the
effects of fatigue. J Physiol 526: 671–681, 2000. doi:10.1111/j.1469-7793.
2000.00671.x.
893
57. Ryschon TW, Fowler MD, Wysong RE, Anthony A, Balaban RS.
Efficiency of human skeletal muscle in vivo: comparison of isometric,
concentric, and eccentric muscle action. J Appl Physiol (1985) 83: 867–
874, 1997.
58. Salin K, Auer SK, Rey B, Selman C, Metcalfe NB. Variation in the link
between oxygen consumption and ATP production, and its relevance for
animal performance. Proc Biol Sci 282: 20151028, 2015. doi:10.1098/
rspb.2015.1028.
59. Seiberl W, Power GA, Hahn D. Residual force enhancement in humans:
Current evidence and unresolved issues. J Electromyogr Kinesiol 25:
571–580, 2015. doi:10.1016/j.jelekin.2015.04.011.
60. Spanjaard M, Reeves ND, van Dieën JH, Baltzopoulos V, Maganaris
CN. Gastrocnemius muscle fascicle behavior during stair negotiation in
humans. J Appl Physiol (1985) 102: 1618 –1623, 2007. doi:10.1152/
japplphysiol.00353.2006.
61. Stafilidis S, Karamanidis K, Morey-Klapsing G, Demonte G, Brügge-
mann GP, Arampatzis A. Strain and elongation of the vastus lateralis
aponeurosis and tendon in vivo during maximal isometric contraction. Eur
J Appl Physiol 94: 317–322, 2005. doi:10.1007/s00421-004-1301-4.
62. Turner AN, Jeffreys I. The stretch-shortening cycle: Proposed mecha-
nisms and methods for enhancement. Strength Condit J 32: 87–99, 2010.
doi:10.1519/SSC.0b013e3181e928f9.

J Appl Physiol • doi:10.1152/japplphysiol.00536.2016 • www.jappl.org

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (190.153.136.020) on October 13, 2018.
Copyright © 2017 American Physiological Society. All rights reserved.