J Appl Physiol 122: 446 –459, 2017.

First published December 22, 2016; doi:10.1152/japplphysiol.00942.2016.

RESEARCH ARTICLE

Muscle metabolic and neuromuscular determinants of fatigue during cycling

in different exercise intensity domains
Matthew I. Black,1,2 Andrew M. Jones,1 Jamie R. Blackwell,1 Stephen J. Bailey,1,2 Lee J. Wylie,1
Sinead T. J. McDonagh,1 Christopher Thompson,1 James Kelly,1 Paul Sumners,3 Katya N. Mileva,3
Joanna L. Bowtell,1 and Anni Vanhatalo1
1
College of Life and Environmental Sciences, St. Luke’s Campus, University of Exeter, Exeter, United Kingdom; 2School of
Sport, Exercise and Health Sciences, Loughborough University, United Kingdom; and 3Sport and Exercise Science Research
Centre, School of Applied Sciences, London South Bank University, London, United Kingdom
Submitted 24 October 2016; accepted in final form 16 December 2016

Black MI, Jones AM, Blackwell JR, Bailey SJ, Wylie LJ, critical power; gas exchange threshold; neuromuscular fatigue; mus-
McDonagh STJ, Thompson C, Kelly J, Sumners P, Mileva KN, cle metabolism; cycling exercise
Bowtell JL, Vanhatalo A. Muscle metabolic and neuromuscular
determinants of fatigue during cycling in different exercise intensity
domains. J Appl Physiol 122: 446 – 459, 2017. First published De- INTENSE AND/OR PROLONGED excitation of muscle leads to a
cember 22, 2016; doi:10.1152/japplphysiol.00942.2016.—Lactate or reversible decline in its force-generating capacity and rate of
gas exchange threshold (GET) and critical power (CP) are closely contraction, commonly known as fatigue (21–23, 27, 55). This
associated with human exercise performance. We tested the hypoth- temporary reduction in muscle performance may be attributed
esis that the limit of tolerance (Tlim) during cycle exercise performed to central factors that limit the neural drive for muscle con-
within the exercise intensity domains demarcated by GET and CP is traction, and to peripheral factors that occur at or distal to the
linked to discrete muscle metabolic and neuromuscular responses. neuromuscular junction and that often involve metabolic and
Eleven men performed a ramp incremental exercise test, 4 –5 severe- ionic perturbations that reduce the muscle’s ability to respond
intensity (SEV; ⬎CP) constant-work-rate (CWR) tests until Tlim, a to neural stimulation (2, 3, 25, 30, 41).
heavy-intensity (HVY; ⬍CP but ⬎GET) CWR test until Tlim, and a The extent of the muscle metabolic and ionic, and blood
moderate-intensity (MOD; ⬍GET) CWR test until Tlim. Muscle acid-base and respiratory perturbations experienced during
biopsies revealed that a similar (P ⬎ 0.05) muscle metabolic milieu exercise is dependent on exercise intensity, which can be
(i.e., low pH and [PCr] and high [lactate]) was attained at Tlim
categorized into three distinct domains demarcated by physio-
(approximately 2–14 min) for all SEV exercise bouts. The muscle
metabolic perturbation was greater at Tlim following SEV compared
logical thresholds (32, 72). The upper limit of the moderate
with HVY, and also following SEV and HVY compared with MOD exercise-intensity domain is indicated by the lactate threshold
(all P ⬍ 0.05). The normalized M-wave amplitude for the vastus [LT, which is often estimated using the gas exchange threshold
lateralis (VL) muscle decreased to a similar extent following SEV (GET)], and the boundary between the heavy and severe
(⫺38 ⫾ 15%), HVY (⫺68 ⫾ 24%), and MOD (⫺53 ⫾ 29%), (P ⬎ exercise-intensity domains is given by the critical power (CP).
0.05). Neural drive to the VL increased during SEV (4 ⫾ 4%; P ⬍ Using 31P-magnetic resonance spectroscopy, it has been dem-
0.05) but did not change during HVY or MOD (P ⬎ 0.05). During onstrated that severe-intensity, single-leg knee-extension exer-
SEV and HVY, but not MOD, the rates of change in M-wave cise is associated with a progressive loss of muscle homeosta-
amplitude and neural drive were correlated with changes in muscle sis with time {i.e., progressive reductions in muscle phospho-
metabolic ([PCr], [lactate]) and blood ionic/acid-base status ([lactate], creatine concentration ([PCr]) and pH, and an increase in
[K⫹]) (P ⬍ 0.05). The results of this study indicate that the metabolic inorganic phosphate concentration ([Pi])} (9, 31, 33, 69). In
and neuromuscular determinants of fatigue development differ ac- contrast, heavy- and moderate-intensity small-muscle-mass ex-
cording to the intensity domain in which the exercise is performed. ercise is associated with much more limited muscle metabolic
NEW & NOTEWORTHY The gas exchange threshold and the perturbation with new steady-state values of [PCr], pH, and
critical power demarcate discrete exercise intensity domains. For the [Pi] being achieved within a few minutes of initiation of
first time, we show that the limit of tolerance during whole-body exercise (33, 48, 67). These intensity-related differences in
exercise within these domains is characterized by distinct metabolic muscle metabolic, as well as related blood acid-base and
and neuromuscular responses. Fatigue development during exercise respiratory gas exchange responses to exercise (33, 51, 68, 73),
greater than critical power is associated with the attainment of con- likely underpin the close relationships reported between these
sistent “limiting” values of muscle metabolites, whereas substrate threshold phenomena (LT/GET, and CP) and human exercise
availability and limitations to muscle activation may constrain per- performance (8).
formance at lower intensities. The role of exercise intensity in defining the extent and
dynamics of muscle metabolic perturbation implies that exer-
cise intensity may also influence the nature of neuromuscular
Address for reprint requests and other correspondence: A. Vanhatalo,
College of Life and Environmental Sciences, St. Luke’s Campus, Univer-
fatigue development (3, 22, 24, 39, 41, 52, 53). The peripheral
sity of Exeter, Heavitree Road, Exeter, EX1 2LU, United Kingdom (e-mail: component to fatigue, as estimated noninvasively using surface
[email protected]). electromyography (EMG), electrical muscle stimulation, trans-
446 Licensed under Creative Commons Attribution CC-BY 3.0: © the American Physiological Society. ISSN 8750-7587. http://www.jappl.org
Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al. 447
cranial magnetic stimulation, or a combination of these appears associated risks, and potential benefits of participation had been
to be especially important during high-intensity exercise (45, explained.
64, 66), whereas central fatigue may be more prominent during
prolonged, low-intensity exercise (38, 45, 57, 61, 66). The Subjects
intensity-dependent interaction between peripheral and central Eleven healthy recreationally active men (age, 21.8 ⫾ 1.9 yr;
components of fatigue is thought to be modulated by changes height, 1.79 ⫾ 0.05 m; body mass, 78.2 ⫾ 8.1 kg; means ⫾ SD)
in afferent feedback arising from the muscle metabolic milieu. volunteered to participate in this study, 8 of whom volunteered to
Consistent with this, the critical torque (CT; analogous with provide muscle tissue samples. One subject who volunteered for the
CP) for small-muscle-mass exercise has been shown to repre- biopsy procedure withdrew from the study after completing only the
sent a threshold in the development of neuromuscular fatigue severe-intensity exercise trials. That subject’s data were excluded
(10), such that severe-intensity knee-extensor contractions from statistical difference tests, but included in the correlational
analysis. All subjects were in good health and had no known history
(⬎CT) were associated with elevated motor unit recruitment of neurological or motor disorder. Subjects were instructed to report
and a disproportionate increase in the rate of neuromuscular to all testing sessions in a rested and fully hydrated state ⱖ3 h
fatigue development relative to heavy-intensity contractions postprandial, and to avoid strenuous exercise and refrain from caffeine
(⬍CT). and alcohol in the 24 h before testing. Each subject started each
It is presently unclear whether the determinants of neuro- experimental trial at the same time of day (⫾2 h). All trials were
muscular fatigue development during whole body exercise performed on the same electronically braked cycle ergometer (Lode;
such as cycling differ according to the intensity domain in Excalibur, Groningen, The Netherlands).
which the exercise is performed. Previous studies have as-
sessed neuromuscular fatigue before and after self-paced max- Experimental Design
imal time-trial cycle exercise (66) and during constant work Each subject visited the laboratory on approximately seven occa-
rate (CWR) cycling performed ostensibly within the severe- sions over a 6-wk period with each visit separated by a minimum of
intensity domain (65). These studies suggested that, in contrast 24 h. A minimum recovery of 7 days was provided following the
to knee extension exercise (10), the level of peripheral fatigue heavy- and moderate-intensity exercise tests. After the completion of
at exhaustion for cycling may also be intensity-dependent a ramp incremental test (visit 1), subjects performed four to five CWR
above CP (65). Compared with small-muscle-mass exercise, severe-intensity exercise tests to define the power-duration relation-
whole body exercise is associated with greater rates of pulmo- ship, a heavy-intensity CWR test, and a moderate-intensity CWR test,
completed in a randomized order (Fig. 1), except that the severe-
nary ventilation and gas exchange (58, 74), differences in
intensity tests always preceded the heavy-intensity test. Pulmonary
cardiac output and muscle perfusion (12, 46, 58), and greater gas exchange was measured continuously during all tests, except the
activity of type III/IV muscle afferents that may modulate moderate-intensity test, during which it was measured periodically for
central drive (52, 53). It is possible that these factors affect the 10-min intervals, with the midpoint of collection coinciding with
relationship between muscle metabolic changes and neuromus- blood sample collection and femoral nerve stimulation (see below).
cular fatigue development during exercise. We encouraged the subjects to continue exercising during the mod-
To date, the physiological and neuromuscular responses to erate-intensity test to enable 10 min of gas exchange data to be
whole body exercise and their possible interrelationship has not collected immediately before exercise cessation. EMG data were
been assessed within distinct exercise intensity domains. The obtained continuously from the vastus lateralis (VL) and vastus
purpose of this study therefore was to evaluate possible differ- medialis (VM) muscles throughout the exercise period with stimula-
tion of the femoral nerve delivered at regular intervals (Fig. 1) to
ences in the muscle metabolic and systemic responses to
quantify the neuromuscular changes occurring during the exercise
different, well-defined intensities of exercise, with the aim of protocols. Venous blood samples were obtained before and during
elucidating whether the exercise intensity domain influences exercise for the moderate-, heavy-, and for three of the severe-
the determinants of neuromuscular fatigue. Based on earlier intensity exercise tests. In addition, muscle tissue was obtained at rest
studies investigating small-muscle-mass exercise (33, 69), we and immediately following the moderate-, heavy-, and three of the
tested the hypotheses that 1) a consistent muscle metabolic severe-intensity exercise tests (Fig. 1). The severe-intensity tests were
milieu ([ATP], [PCr], [lactate], pH) and neuromuscular re- performed at three to five different work rates spanning 60%⌬ to
sponses (muscle excitability and neural drive) will be attained V̇O2peak (where ⌬ refers to the work rate difference between GET and
at the limit of tolerance (Tlim) during severe-intensity exercise V̇O2peak). Three of these severe-intensity tests (including short
(⬎CP); 2) severe-intensity exercise will be associated with 85 ⫾ 5%⌬, intermediate 75 ⫾ 5%⌬, and long 65 ⫾ 5%⌬) were
grouped and compared for differences in muscle, neuromuscular, and
greater muscle metabolic perturbation compared with heavy-
blood responses within the severe-intensity domain.
and moderate-intensity exercise; and 3) the rate of neuromus-
cular fatigue development will be greater during severe- com- Incremental Test
pared with heavy- and moderate-intensity exercise due to
greater muscle metabolic and ionic perturbations. During the first laboratory visit, subjects completed a ramp incre-
mental test for determination of the V̇O2peak and GET. The ergometer
seat height and handlebars were adjusted for comfort, and the same
METHODS
settings were replicated for each subsequent test. Initially, subjects
Ethical Approval completed 3 min of baseline cycling at 20 W, after which the work
rate was increased by 30 W/min until volitional exhaustion. Subjects
The protocols were approved by the Research Ethics Committee of cycled at a constant self-selected pedal rate [80 revolutions per minute
Sport and Health Sciences (University of Exeter) and conducted in (rpm), n ⫽ 9, 90 rpm, n ⫽ 2), which was recorded and reproduced in
accordance with the code of the ethical principles of the World subsequent tests. The test was terminated when the pedal rate fell
Medical Association (Declaration of Helsinki). Subjects gave written more than 10 rpm below the preferred value for more than 5 s despite
informed consent to participate after the experimental procedures, strong verbal encouragement. Breath-by-breath pulmonary gas ex-

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
448 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al.

350

300
Fig. 1. Schematic of the exercise protocol.
Group mean work rates are shown for the
severe- (solid line), heavy- (dotted line), and 250

Power output (W)

moderate- (dashed line) intensity trials. All
trials were started with a 3-min warm-up
phase at 20 W, followed by an immediate 200
step increase to the required work rate. Sub-
jects were encouraged to continue exercising
150
as long as possible. Dashed arrows indicate
collection of venous blood and femoral nerve
stimulation. Solid arrows indicate collection
100
of muscle tissue. For clarity, the resting mus-
cle sample obtained before the first trial is
not shown.
50

0
-3 0 3 6 9 12 15 20 30 40 50 60 120 180 240 300 360
Time (min)

change data were collected continuously throughout the test and The work rate for the heavy-intensity CWR trial was equal to the
recorded as a 10-s moving average for data analysis. V̇O2peak was lower bound of the 95% confidence limit in the CP parameter (33).
determined as the highest mean V̇O2 during any 30-s period, and GET The moderate-intensity CWR trial was performed at a work rate
was determined as previously described (5, 68). corresponding to 90% of GET. Subjects were permitted to ingest
water ad libitum during the heavy- and moderate-intensity tests.
CWR Tests
Pulmonary Gas Exchange
All CWR tests started with 3 min of cycling at 20 W, followed by
a step increase to the required work rate. Subjects were instructed to Breath-by-breath pulmonary gas exchange and ventilation were
remain seated and to maintain their preferred pedal rate for as long as measured continuously during all exercise tests, except the moderate-
possible. Strong verbal encouragement was provided, but subjects intensity test, when it was measured at discrete time points. Subjects
were not informed of either the work rate or the elapsed time. Tests wore a nose clip and breathed through a mouthpiece and impeller
were terminated when pedal rate fell more than 10 rpm below the turbine assembly (Jaeger Triple V; Jaeger, Hoechberg, Germany). The
preferred value for more than 5 s. Tlim was recorded to the nearest inspired and expired gas volume and concentration signals were
second. continuously sampled at 100 Hz (Oxycon Pro; Jaeger) via a capillary
The parameters of the power-duration relationship (CP and W=) line connected to the mouthpiece. The gas analyzers were calibrated
were estimated by completion of four to five severe-intensity exercise before each trial with gases of known concentration, and the turbine
tests (4 trials, n ⫽ 9; 5 trials n ⫽ 2) at different work rates (~60%⌬, volume transducer was calibrated using a 3-liter syringe (Hans Ru-
70%⌬, 80%⌬, and 100% V̇O2peak) resulting in Tlim ranging between dolph, Kansas City, MO). The volume and concentration signals were
~2 and 14 min. If the standard errors associated with the CP and W= time aligned by accounting for the delay in capillary gas transit and
exceeded 5 and 10%, respectively, after four exercise tests had been the analyzer rise time relative to the volume signal.
completed, a fifth test was performed. Any tests in which the end-
exercise V̇O2 was ⬍95% of the individual’s ramp test determined Blood Analyses
V̇O2peak were excluded from the modeling of the power-duration
relationship. Venous blood samples were drawn into 5-ml heparinized syringes
CP and W= (the amount of work performed above the CP) param- (Terumo, Leuven, Belgium) from a cannula (Insyte-W; Becton Dick-
eters were estimated using three models: the hyperbolic P-Tlim model inson, Madrid, Spain) inserted into the subject’s antecubital vein.
(Eq. 1); the linear work-time (W-Tlim) model, in which the total work Blood was analyzed for [lactate] within ~5 min of collection (YSI
performed (W) is plotted against time (Eq. 2); and the linear inverse- 2300; Yellow Springs Instruments, Yellow Springs, OH). The remain-
of-time (1/Tlim) model, in which power output is plotted against the ing whole blood was then centrifuged at 4,000 rpm for 7 min (Hettich
inverse of time (Eq. 3): EBA 20; Germany) before plasma was extracted and analyzed for
[K⫹] (9180 Electrolyte Analyzer; F. Hoffman-La Roche, Basel, Swit-
Tlim ⫽ W⬘⁄ 共 P – CP兲 (1) zerland).
W ⫽ CP · Tlim ⫹ W⬘ (2)
Neuromuscular Function
P ⫽ W⬘共1 ⁄ Tlim兲 ⫹ CP (3)
EMG was used to continuously record VL and VM activity during
The standard errors of the estimate associated with CP and W= exercise using active bipolar bar electrodes with single differential
were expressed as coefficients of variation (CV%; that is, relative configuration (DE2.1; DelSys, Boston, MA), positioned over the
to the parameter estimate). For each subject, the “best fit” model muscle belly [surface EMG for noninvasive assessment of muscles
associated with the lowest CV% for CP and W= was used for further (SENIAM) guidelines]. The ground electrode was positioned on the
analyses (7). patella. Double-sided adhesive interfaces and hypoallergenic medical

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al. 449
tape were used to keep the EMG sensors in place and to reduce skin Muscle Biopsy
impedance. The leads connected to the electrodes were secured using
hypoallergenic medical tape to minimize artifacts due to movement of The biopsy site was prepared on the alternate thigh to the EMG and
the leads. The skin area underneath each electrode was shaved, peripheral nerve stimulation setup. Local anesthesia was applied (2–3
abraded, and cleaned with alcohol swabs before electrode placement ml of 20 mg/ml lidocaine) and an incision was made in the medial
to minimize skin impedance. The EMG signal was considered of good region of the VL. Muscle samples were obtained using needle biopsy
quality when the average rectified EMG baseline level for each with suction (6). Resting muscle samples were obtained before any
muscle was below 2 ␮V (18). The EMG signals were preamplified exercise on the first laboratory visit and postexercise biopsies were
(1,000⫻), band-pass filtered (20 – 450 Hz, Bagnioli-8; DelSys), and taken within ~10 s of cessation of each exercise test with the subject
digitized at a sampling rate of 2,000 Hz and resolution of 16 bits using supported on the ergometer. Muscle tissue was rapidly frozen in liquid
a Power 1401 mk-II analog-to-digital converter and Spike 2 data nitrogen.
collection software run by custom-written sampling configuration
(CED; Cambridge Electronic Design, UK). Muscle Tissue Analysis
The location of the optimal site for transcutaneous femoral nerve Frozen muscle samples from each biopsy were weighed before and
stimulation was determined while subjects were positioned on the after freeze-drying to determine water content. After freeze-drying,
cycle ergometer. Using an adhesive cathode (Boots UK, Nottingham, the muscle samples were dissected free from blood, fat, and connec-
England) placed ~2 cm medial of the femoral pulse, and an adhesive tive tissue. Before muscle metabolite analysis, 200 ␮l of 3 M per-
anode (Boots UK) placed at the anterior aspect of the iliac crest, single chloric acid was added to ~2.5 mg of dry weight (dw) muscle. The
electrical pulses generated by a constant current stimulator (DS7 A; solution was then centrifuged and placed on ice for 30 min. It was
Digitimer, UK) were delivered. The cathode was systematically subsequently neutralized to pH 7.0 with 255 ␮l of cooled KHCO3 and
moved vertically and horizontally, and the amplitude of the compound centrifuged (10,000 g). The supernatant was analyzed for PCr, ATP,
muscle action potential (i.e., M-wave) was monitored to identify the and lactate by fluorometric assays (35). An aliquot containing 1–2 mg
optimal position of the cathode for attaining maximal peak-to-peak of dw muscle was extracted in 1 M HCl and hydrolyzed at 100°C for
M-wave amplitude during the cycling trials. 3 h before glycogen content was determined using the hexokinase
Following attachment of EMG and stimulation electrodes, the method (35). Muscle pH was measured using a glass electrode
crank angle at which stimulation was to be delivered during the trials following the homogenization of 1–2 mg of dw of muscle in a
was determined for each subject. Subjects were positioned on the nonbuffering solution containing (in mM) 145 KCl, 10 NaCl, and 5
cycle ergometer and cycled at a moderate work rate (20 W below iodoacetic acid.
GET) for 1 min. EMG activity obtained during this period was
rectified and averaged for 20 complete crank revolutions. The duration Statistical Analyses
of each revolution was determined by a custom-made magnetic switch
that generated an event marker signal on each occasion that the crank One-way ANOVA with repeated measures was used to assess
passed top dead center (i.e., 0°). For each subject, the crank angle at differences between severe-intensity exercise tests in V̇O2peak, muscle
which the rectified VL EMG activity was maximal was determined, [ATP], [PCr], [pH], [lactate], and [glycogen], M-wave amplitude,
and as performed by Sidhu et al. (56), stimulations were delivered at M-wave area, voluntary EMG amplitude and RMS/M, and blood and
the identified crank angle for all subsequent trials for that participant plasma variables at Tlim. The data from the severe-intensity tests were
(65 ⫾ 5° relative to the top dead center). A custom-written sequencer subsequently averaged for each subject for comparison with the
script triggered three stimulations, with at least 1 and up to 10 pedal heavy- and moderate-intensity tests. Differences in V̇O2peak, muscle
revolutions between stimuli. The intervals were randomly determined [ATP], [PCr], [pH], [lactate], and [glycogen] between the severe-,
using a random number generator incorporated within the sequencer heavy-, and moderate-intensity tests were assessed using a one-way
script. This was designed to prevent participants from anticipating the ANOVA. Two-way repeated-measures ANOVA (condition ⫻ time) was
stimulus delivery, which may have affected the evoked response. used to analyze differences in M-wave amplitude and area, and voluntary
A standard M-wave recruitment curve protocol was completed EMG amplitude for VL and VM muscles, and blood and plasma
during each laboratory visit. Subjects cycled at 20 W below GET variables at common time points (baseline, 1 min, 3 min, and Tlim)
throughout the recruitment curve protocol. A single-pulse electrical among the severe-, heavy-, and moderate-intensity tests. Significant
stimulation (200 ␮s) was delivered at the individually identified crank interaction and main effects were followed up with a Bonferroni post
angle as described above. The current was increased in 20-mA hoc test. Relationships between the rates of change of metabolic and
increments until the M-wave amplitude reached a plateau at the neuromuscular variables were assessed using Pearson’s product-mo-
maximal M-wave amplitude (Mmax). A pulse of 130% Mmax current ment correlation coefficients. Statistical significance was set at P ⬍
was applied during the exercise tests (mean stimulation intensity, 0.05 and data are presented as means ⫾ SD.
350 ⫾ 50 mA).
EMG signals from the VL and VM were processed using a RESULTS
custom-written script to measure peak-to-peak M-wave amplitude and
M-wave area. The root mean square (RMS) of the EMG signal (an The V̇O2peak measured in the ramp incremental test was
index of the power of the signal) was calculated as the mean over a 4.32 ⫾ 0.46 l/min (56 ⫾ 8 ml·kg⫺1·min⫺1) and the peak work
25-ms prestimulation period at each stimulation time point. The EMG rate was 385 ⫾ 50 W. GET occurred at 2.33 ⫾ 0.34 l/min and
RMS amplitudes and the M-wave parameters were normalized to the 137 ⫾ 24 W.
corresponding values attained after 1 min of exercise during each trial
to evaluate temporal changes in the voluntary muscle activation level
(i.e., EMG RMS amplitude) and the peripheral neuromuscular excit- Physiological Responses Within the Severe-Intensity Domain
ability (i.e., M-wave amplitude and area). In addition, the voluntary
EMG RMS amplitude was normalized to the M-wave amplitude
Tlim values in the severe-intensity CWR exercise tests
recorded at that time point to assess changes in neural drive (RMS/M; ranged from 2.2 to 13.9 min. There were no differences
42). The rates of change in M-wave and EMG parameters from between the three models (Eqs. 1–3) in the CP or W= estimates
baseline cycling to Tlim were calculated for each exercise to quantify (P ⬎ 0.05; Table 1). The CP from the best-fit model corre-
the rate of neuromuscular fatigue development in each intensity sponded to 64 ⫾ 7% of ramp test peak work rate and
domain. 45 ⫾ 11%⌬.

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
450 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al.

Table 1. Parameter estimates derived from Eqs. 1–3 and the “optimized fit” model
R2 CP, W SEE, W CV% W= kJ SEE, kJ CV%

W-Tlim model 0.993–1.000 253 ⫾ 54 6⫾3 2.6 ⫾ 1.4 22.5 ⫾ 5.3 2.3 ⫾ 1.0 11.0 ⫾ 6.2
1/Tlim model 0.939–0.999 252 ⫾ 52 7⫾4 3.0 ⫾ 2.3 20.7 ⫾ 5.2 1.9 ⫾ 1.1 9.5 ⫾ 5.6
P-Tlim model 0.919–1.000 248 ⫾ 52 5⫾3 2.2 ⫾ 1.4 22.4 ⫾ 3.8 2.5 ⫾ 1.8 11.3 ⫾ 9.4
Optimized fit model 0.944–1.000 250 ⫾ 53 5⫾2 2.0 ⫾ 1.2 22.5 ⫾ 6.1 1.8 ⫾ 0.8 8.3 ⫾ 4.5
CP, critical power; CV%, coefficient of variation; SEE, standard error of estimate; W= finite work capacity above the CP.

V̇O2peak values during the shorter (~85%⌬, 4.43 ⫾ 0.50 intermediate (5.8 ⫾ 1.1 mM), and longer (5.7 ⫾ 0.6 mM) se-
l/min), intermediate (~75%⌬, 4.49 ⫾ 0.47 l/min), and longer vere-intensity tests (P ⬎ 0.05).
(~65%⌬, 4.41 ⫾ 0.47 l/min) severe-intensity tests were not
different from V̇O2peak achieved during the ramp incremental Physiological Responses During Severe-, Heavy-, and
test (all P ⬎ 0.05). Moreover, no significant differences Moderate-Intensity Exercise
were observed at Tlim among the three severe-intensity tests
for any of the muscle tissue variables or for blood [lactate] Pulmonary V̇O2, blood [lactate], and plasma [K⫹] during
(all P ⬎ 0.05; Fig. 2). There were also no differences in moderate-, heavy-, and severe-intensity exercise are illustrated
plasma [K⫹] at Tlim among the shorter (5.6 ⫾ 0.6 mM), in Fig. 3. The Tlim for heavy-intensity exercise (231 ⫾ 56 W)

A B
18 * 60 *
Muscle [ATP] (mmol·kg-1 DW)

Muscle [PCr] (mmol·kg-1 DW)

15 50

12 40
16
9
12
6
8

3 4

0 0

C 160 D
Muscle [La] (mmol·kg-1 DW)

7.25 *
Fig. 2. Muscle metabolic responses ([ATP]
(A), phosphocreatine ([PCr]) (B), pH (C), 7.00 120
[lactate] (D), [glycogen] (E), and blood
Muscle pH

[lactate] (F) at the limit of tolerance (Tlim)

were not different following exhaustive 6.75
exercise at three different severe-intensity 80
work rates. R, rest; S1, short trials at ~85%⌬ 6.50
(Tlim ⫽ 224 ⫾ 41 s); S2, intermediate trials at
~75%⌬ (Tlim ⫽ 333 ⫾ 131 s); and S3, long 40
trials at ~65%⌬ (Tlim ⫽ 475 ⫾ 145 s). *Dif- 6.25
ferent from S1, S2, and S3 (P ⬍ 0.05). *
0.00 0

600
E 12
F
[Glycogen] (mmol·kg-1 DW)

500 *
Blood [La] (mmol·L-1)

400 8

300

200 4

100
*
0 0
R S1 S2 S3 R S1 S2 S3

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al. 451

A ab

5.0 .
Pulmonary VO2 (L.min-1)

4.5 a Ramp test VO2peak

4.0

3.5
.
3.0
bc
2.5 GET

2.0

1.5
0 10 20 30 40 50 60 120 180 240 300 360

B
10 Fig. 3. Pulmonary V̇O2 (A) during severe
ab (solid circle, open circle, solid triangle)-,
Blood [lactate] (mM)

8 heavy (open triangles)-, and moderate (solid

squares)-intensity exercise. End-exercise VO2
values in the three severe-intensity trials were
6 not different from the ramp test VO2peak.
a
Blood [lactate], (B), and plasma [K⫹] (C) re-
4 sponses to severe- (solid circles), heavy- (open
circles), and moderate- (solid triangles) intensity
2 bc exercise. To aid clarity error bars have been
omitted from all but the final data point. aDif-
ferent from moderate-intensity P ⬍ 0.05; bdif-
0 ferent from heavy-intensity P ⬍ 0.05; cdifferent
0 10 20 30 40 50 60 120 180 240 300 360
from severe-intensity (P ⬍ 0.05).

C
6.5 ab
a

6.0
Plasma [K+] (mM)

5.5 bc

5.0

4.5

4.0

3.5
0 10 20 30 40 50 60 120 180 240 300 360
Time (min)

was 43.5 ⫾ 16.2 min (range, 20.5 to 67.4 min), and V̇O2 at Tlim exercise (all P ⬍ 0.05). The [K⫹] continued to rise throughout
(3.78 ⫾ 0.53 l/min; 87 ⫾ 4% of V̇O2peak) was lower than the severe-intensity exercise, although it stabilized during heavy-
ramp test V̇O2peak (P ⬍ 0.05). Tlim for the moderate-intensity intensity exercise beyond 6 min (Fig. 3C). During moderate-
exercise (113 ⫾ 19 W) was 211.1 ⫾ 57.0 min (range, 180 to intensity exercise, blood [lactate] did not change from baseline
360 min) and the V̇O2 at Tlim (2.22 ⫾ 0.38 l/min, 52 ⫾ 8% of (P ⬎ 0.05), although plasma [K⫹] was elevated above resting
V̇O2peak) was not different from GET (P ⬎ 0.05). In 9 of 11 baseline at 1 min (P ⬍ 0.05), with no further increase thereafter
subjects V̇O2 remained below GET throughout the moderate- (all time points P ⬎ 0.05).
intensity exercise bout. Muscle metabolic variables at rest and at Tlim following
During severe-intensity exercise, blood [lactate] increased moderate-, heavy-, and severe-intensity exercise are illustrated
rapidly until Tlim and was significantly greater than baseline in Fig. 4. For severe- and heavy-intensity exercise, muscle
values after 3 min (P ⬍ 0.05). During heavy-intensity exercise, [ATP], [PCr], and pH were lower, and muscle [lactate] was
the rate of blood [lactate] increase was slower than during greater at Tlim relative to rest (all P ⬍ 0.05). There was no
severe-intensity exercise such that blood [lactate] did not differ significant muscle [glycogen] depletion during severe-intensity
from baseline until after 10 min (P ⬍ 0.05), and no further exercise relative to rest (P ⬎ 0.05), but there was a tendency
increase was observed between 10 min and Tlim (P ⬎ 0.05) for glycogen depletion during heavy-intensity exercise (P ⫽
(Fig. 3B). Plasma [K⫹] was elevated above baseline at all 0.06). In contrast, for moderate-intensity exercise, muscle
measurement time points during heavy- and severe-intensity [PCr] at Tlim was greater than for severe- and heavy-intensity

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
452 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al.

A 25 B 60 abc

50
20

[ATP] (mmol·kg-1dw)

[PCr] (mmol·kg-1dw)
abc
*bc 40
15
30
10 *ac
*bc
20
*ab
5 *ac
10
*ab
0 0
0 10 20 30 40 50 60 120 240 360 0 10 20 30 40 50 60 120 240 360

C 7.4
bc
D 160
*ab

[Lactate] (mmol·kg-1·dw)
140
7.2
120
Fig. 4. Muscle [ATP] (A), [PCr] (B), [pH] (C), 7.0
[lactate] (D), and [glycogen] (E) at rest (open 100
triangles), and following severe- (solid circles),
pH

6.8 80
heavy- (open circles), and moderate-intensity
c *ac
exercise (solid triangles). *Different from rest 60
6.6
P ⬍ 0.05. aDifferent from moderate-intensity *c
P ⬍ 0.05; bdifferent from heavy-intensity P ⬍ 40
0.05; cdifferent from severe-intensity P ⬍ 0.05. 6.4 *ab 20 bc bc

6.2 0 10 20 30 40 50 60 120 240 360 0

0 10 20 30 40 50 60 120 240 360
Time (min)
E 600
[Glycogen] (mmol·kg-1dw)

ab
500

400

300 a

200 a

*bc
100

0
0 10 20 30 40 50 60 120 240 360
Time (min)

exercise (all P ⬍ 0.05), and muscle [glycogen] was both lower differences were observed between trials in the neural drive to
than at rest and lower than at Tlim for heavy- and severe- VL and VM during cycling at 20 W below GET.
intensity exercise (all P ⬍ 0.05). Muscle [pH] and [lactate] did Neuromuscular excitability: M-wave amplitude and M-wave
not change significantly from rest during moderate-intensity area. M-wave characteristics at Tlim for the three severe-
exercise (P ⬎ 0.05). intensity exercise tests, and for moderate-intensity, heavy-
intensity, and the mean of the severe-intensity exercise tests are
Neuromuscular Responses During Severe-, Heavy-, and shown in Fig. 5, A–D. Peripheral neuromuscular excitability at
Moderate-Intensity Exercise Tlim, indicated by M-wave amplitude and M-wave area, did not
The coefficients of variation (CV%) between trials during differ among the severe-intensity tests (all P ⬍ 0.05) (Fig. 5, A
unloaded cycling were 25% (VL) and 35% (VM) for peak-to- and C). M-wave amplitude and M-wave area at Tlim were
peak M-wave amplitude, and 32% (VL) and 32% (VM) for greater during severe-intensity exercise compared with both
M-wave total area. CV% between stimulations during un- heavy- and moderate-intensity exercise in the VM (P ⬍ 0.05),
loaded cycling was 11% (VL) and 9% (VM) for peak-to-peak and the M-wave area at Tlim was also greater in severe- than in
M-wave amplitude, and 10% (VL) and 9% (VM) for M-wave heavy-intensity exercise in the VL (P ⬍ 0.05) (Fig. 5, B and
total area. Mean Mmax amplitudes measured during cycling at D). Differences in M-wave characteristics between severe-,
20 W below GET (VL, 2.77 ⫾ 1.43; VM, 0.99 ⫾ 1.18 mV) heavy-, and moderate-intensity exercise at each measurement
were not different between visits (all P ⬎ 0.05). No significant time point are shown in Fig. 6, A–D.

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al. 453

Severe 1 Moderate
Severe 2 Heavy
Severe 3 Severe (mean)

Muscle excitability at Tlim

180
A 180
B
Normalised M-wave

160 160
amplitude (%)

a
140 140
b
120 120
100 100
c
80 80 c
60 60
40 40
20 20
0 0
VL VM VL VM

160
C 160
D a
Fig. 5. Compound muscle action potential (M-wave) am-
Normalised M-wave

140 140 b plitude and M-wave area (normalized to maximum M-wave

120 120 during baseline pedalling) (group means ⫾ SD) indicating
area (%)

100 100 peripheral neuromuscular excitability (A–D); voluntary

b c c
electromyographic root mean square (EMG RMS) ampli-
80 80 tude (normalized to M-wave amplitude at 1 min of exer-
c
60 60 cise) indicating muscle activation level (E and F); and
40 40 RMS/M-wave (normalized to corresponding M-wave am-
plitude at each measurement time point) indicating central
20 20 fatigue (G and H) at the limit of tolerance (Tlim) for
0 0 moderate-, heavy-, and severe-intensity exercise (B, D, F,
VL VM VL VM and H) and for three work rates (severe 1 ~85%⌬, severe 2
~75%⌬, and severe 3 ~65%⌬) within the severe-intensity
Voluntary activation at Tlim domain (A, C, E, and G). There were no significant differ-
ences among the severe-intensity work rates in muscle
E F
2.0 2.0 a
excitability (A and C) or indices of central fatigue (E and
b G). VL, vastus lateralis muscle; VM, vastus medialis mus-
cle. aDifferent from moderate-intensity P ⬍ 0.05; bdifferent
amplitude (%)

1.5 1.5 b from heavy-intensity P ⬍ 0.05; cdifferent from severe-

EMG RMS

b intensity P ⬍ 0.05.
a
1.0 1.0 c
c c

0.5 0.5

0.0 0.0
VL VM VL VM

Neural drive at Tlim

RMS / M-wave amplitude

20 20
18 G H a
16
15
14
12
(%)

a
10 10 a
8 c
b
6 c
5
4
2
0 0
VL VM VL VM

Voluntary activation and neural drive. Voluntary muscle at Tlim for severe-intensity compared with heavy- and moder-
activation level, measured as EMG RMS amplitude, and neural ate-intensity exercise in VM (P ⬍ 0.05) (Fig. 5F). In VL, the
drive, as indicated by RMS/M-wave amplitude, did not differ EMG RMS at Tlim was also greater for severe- than for
at Tlim among the severe-intensity exercise tests (all P ⬍ 0.05) heavy-intensity exercise, and RMS/M was greater for severe-
(Fig. 5, E and G). Both EMG RMS and RMS/M were greater than for moderate-intensity exercise (both P ⬍ 0.05) (Fig. 5, F

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
454 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al.

140
A VL
140
B VM

Normalised M-wave
cd
120 120

amplitude (%)
100 100
†¥ a a
80 80
60 60
40 40
20 20
0 0
0 3 6 9 60 120 180 240 300 360 0 3 6 9 60 120 180 240 300 360

C D
140 140
VL cd VM

Normalised M-wave
120 120
100 100

area (%)
c e a
Fig. 6. Normalized M-wave amplitude (A and B), 80 80 a
M-wave area (C and D), voluntary EMG RMS
60 a 60
amplitude (E and F), and RMS/M-wave amplitude
(G and H) during severe- (solid circles), heavy- 40 40
(open circles), and moderate-intensity (solid trian- 20 20
gles) exercise in vastus lateralis (VL) and vastus
medialis (VM) muscles. M-wave amplitude and 0 0
area were normalized to maximum M-wave during 0 3 6 9 60 120 180 240 300 360 0 3 6 9 60 120 180 240 300 360
baseline pedaling, EMG RMS was normalized to
M-wave amplitude at 1 min of exercise, and RMS/ 2.0
E VL
2.0
F VM
Normalised EMG RMS

cd
M-wave was normalized to corresponding M-wave
amplitude (%)

amplitude at each measurement time point. Error 1.5 1.5

c
bars have been omitted from all but the final data c
point to aid clarity. aDifferent from rest; bdifferent
1.0 1.0 b
from severe-intensity (P ⬍ 0.05); cdifferent from b
heavy-intensity (P ⬍ 0.05); ddifferent from mod-
erate-intensity (P ⬍ 0.05); and etrend for differ- 0.5 0.5
ence from heavy-intensity (P ⫽ 0.055).
bd
0.0 0.0
0 3 6 9 60 120 180 240 300 360 0 3 6 9 60 120 180 240 300 360

d
G H
16 16
VL VM
RMS / M-wave
amplitude (%)

12 12
cd d
d d
8 8
b
b
bc
4 4
b b

0 0
0 3 6 9 60 120 180 240 300 360 0 3 6 9 60 120 180 240 300 360
Time (min) Time (min)

and H). The only difference in neuromuscular variables ob- related to high plasma [K⫹] and low muscle [PCr], and high
served at Tlim between moderate- and heavy-intensity exercise muscle [lactate] and [glycogen] (Table 2). During moderate-
was a significantly greater EMG RMS in VL (Fig. 5F). Dif- intensity exercise, the M-wave amplitude was inversely corre-
ferences in EMG RMS and RMS/M during severe-, heavy-, lated with the reduction in [PCr] (Table 2).
and moderate-intensity exercise at each measurement time
point are shown in Fig. 6, E–H. DISCUSSION

Relationships Between Physiological and Neuromuscular To our knowledge, the present study is the first to combine
Variables muscle biopsy, blood analyses, and measurements of neuro-
muscular excitability and neural drive (via electrical stimula-
During severe-intensity exercise, M-wave amplitude de- tion of the femoral nerve during exercise) to assess the muscle
creased in parallel with [PCr] depletion and plasma K⫹ metabolic, acid-base, and neuromuscular responses to cycling
accumulation (Table 2). Moreover, increased neural drive performed within discrete exercise intensity domains (32). The
(RMS/M) was correlated with high blood [lactate] and plasma data presented herein provide novel insight into the in vivo
[K⫹] and low muscle [PCr], and high muscle [lactate] and relationships between exercise intensity, muscle metabolic per-
[glycogen] (Table 2). During heavy-intensity exercise, the turbation, and neuromuscular function, and support the notion
reduction in M-wave amplitude was related to low muscle that LT/GET and CP separate exercise intensity domains
[PCr] and high plasma [K⫹], and increased neural drive was within which exercise tolerance is limited by discrete fatigue

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al. 455
Table 2. Correlation coefficients between the rate of change that, at least for small-muscle-mass exercise, the utilization of
in blood and muscle tissue variables and rate of change in this finite energy store (W=) coincides with depletion of muscle
neuromuscular variables measured in vastus lateralis muscle PCr and accumulation of fatigue-related metabolites (i.e., Pi,
H⫹) until a consistent, presumably “limiting” value is attained
M-Wave Amplitude Voluntary EMG Neural Drive
(33, 69). The findings of the current study indicate that,
Severe irrespective of work rate or exercise duration (approximately
Blood [lactate]b ⫺0.30 0.57a 0.47a 2–14 min), Tlim during severe-intensity exercise is associated
Plasma [K⫹]b ⫺0.39a 0.68a 0.64a
[PCr]c 0.59a ⫺0.80a ⫺0.80a
with the attainment of consistently low values of muscle [PCr]
[lactate]c ⫺0.40 0.44a 0.55a (~23% of resting value), [ATP] (~76% of resting value), and
[glycogen]c ⫺0.22 0.46a 0.56a pH (~6.56), and consistently high values of muscle [lactate]
[pH]c ⫺0.13 0.36 0.37 (~1,382% of resting value) and blood [lactate] (~838% of
[ATP]c 0.21 ⫺0.60a ⫺0.59a resting value). It should be noted that the observed muscle
Heavy
Blood [lactate]d ⫺0.42 0.13 0.49 metabolite and substrate changes are reflective of the homog-
Plasma [K⫹]d ⫺0.88a ⫺0.29 0.86a enate muscle sample, and therefore reflect the mean values for
[PCr]e 0.93a ⫺0.28 ⫺0.72a that muscle portion. It is known that the depletion of muscle
[lactate]e ⫺0.25 0.63 0.66 [PCr] during exercise displays significant regional heterogene-
[glycogen]e ⫺0.15 0.53 0.77a
[pH]e 0.13 0.78a 0.27
ity (13, 54). It is therefore possible that the eventual failure of
[ATP]e ⫺0.26 0.32 0.63 subjects to maintain the requisite power output was caused by
Moderate attainment of sufficiently low values of [PCr] and, perhaps,
Blood [lactate]d 0.08 0.05 0.10 [ATP], and/or sufficiently high values of muscle metabolites
Plasma [K⫹]d 0.12 0.18 0.49 ([Pi], [ADP], [H⫹], and their sequelae) within some of the
[PCr]e ⫺0.67a ⫺0.36 0.58
[lactate]e ⫺0.44 ⫺0.34 0.04 recruited muscle fibers [(51); see also (3, 23, 24)]. Clearly,
[glycogen]e ⫺0.10 0.43 0.23 subjects either could not, or would not, tolerate this “critical
[pH]e 0.19 0.06 ⫺0.30 combination” of substrate and metabolite concentrations, but it
[ATP]e 0.09 0.59 0.24 is not possible to ascertain whether this was related to direct
a
P ⬍ 0.05; bn ⫽ 33; cn ⫽ 24; dn ⫽10; en ⫽ 7. effects of the muscle metabolic milieu on contractile function
(17) or to the attainment of some individual sensory “critical
fatigue threshold” that might constrain central motor drive
mechanisms. In classical terms, when exercise intensity ex- and muscle activation via feedback from type III/IV neural
ceeds CP, the oxidation of fat and carbohydrate cannot keep afferents (4). The appreciable metabolic perturbation we
pace with required ATP turnover, and the rate of pyruvate observed during severe-intensity exercise was associated
production from glycolysis exceeds the capacity of the Krebs with a concomitant decrease in M-wave amplitude in both
cycle, resulting in progressive increase in intramuscular lactate VL and VM muscles. A strong inverse correlation was
and H⫹ concentrations. We demonstrated that a similar muscle observed between both the voluntary EMG RMS amplitude
metabolic milieu (i.e., [ATP], [PCr], [lactate], and pH) was and neural drive, and changes in [ATP] and [PCr] (Table 2).
attained at Tlim irrespective of work rate within the severe- This is consistent with there being greater engagement of
intensity domain. The muscle metabolic perturbation was central neural mechanisms (e.g., muscle fiber recruitment
greater (i.e., lower [ATP] and pH, and higher [lactate]) at Tlim and firing frequency modulation) to compensate for periph-
following severe- and heavy-intensity exercise compared with eral fatigue development.
moderate-intensity exercise. In contrast, more extensive mus- We have proposed that changes in muscle metabolic status
cle glycogen depletion occurred during moderate- compared that occur concomitantly with expenditure of W= drive the
with both severe- and heavy-intensity exercise. development of the V̇O2 slow component during severe-inten-
However, although the results indicate that CP represents a sity exercise (8, 33, 70). Thus, exercise intolerance in this
critical threshold for both muscle metabolic control and neu- intensity domain is associated with the complete utilization of
romuscular fatigue development, the importance of GET in W=, attainment of some “critical” combination of muscle
separating exercise intensity domains was less obvious; unlike substrate and/or metabolite concentrations, and achievement of
some muscle metabolic, pulmonary gas exchange, and blood V̇O2peak (8, 14, 47, 70). In the present study, we observed a
[lactate] responses, neuromuscular indices of fatigue develop- reduction in muscle excitability in parallel with increased
ment were not strikingly different between moderate-intensity metabolic stress. The reduction in muscle membrane excitabil-
and heavy-intensity exercise. ity is likely mediated, at least in part, by changes in plasma
Fatigue During Severe-Intensity Exercise [K⫹] (Table 2), which may reflect a rise in interstitial [K⫹]
within the t-tubule, which weakens propagation of the action
Tlim during the severe-intensity exercise tests ranged from potential along the surface membrane. Increased extracellular
2.2 to 13.9 min, and in all cases, subjects achieved V̇O2peak. [K⫹] impairs force generation due to depolarization of the cell
Historically, the amount of work that can be performed above membrane, resulting in a reduced amplitude of the action
CP (i.e., the curvature constant of the power-duration relation- potential (11, 40). This process attenuates Ca2⫹ release from
ship, W=), and therefore the cause or causes of exercise the sarcoplasmic reticulum, reducing cross-bridge formation
intolerance within the severe-intensity domain, has been linked and the force-generating capacity of the myocyte (36). In our
to the depletion of high-energy phosphates and a source related study, increased plasma [K⫹] was accompanied by a transient
to anaerobic glycolysis, along with a finite amount of stored O2 increase in neural drive that was brought about via a preser-
(43, 44). Consistent with this, recent studies have demonstrated vation of the EMG amplitude with reduced M-wave amplitude.

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
456 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al.

It was notable that reductions in M-wave amplitude and M- example, at the end of exercise, muscle [PCr] had fallen to
wave area in the VM muscle during exhaustive severe exercise ~76% of the baseline value, and pH had fallen by 0.1 unit from
were less pronounced compared with moderate and heavy the resting value, and blood [lactate] and plasma [K⫹] were
exercise (Fig. 5, B and D), suggesting that muscle excitability also largely unchanged (Figs. 3 and 4). There was, however, a
was preserved to a greater extent than at lower exercise large reduction (– 83%) in muscle [glycogen] (1, 29, 59, 60). It
intensities. It is important, however, to consider this finding in is therefore likely that the development of peripheral fatigue
the context of increasing neural drive during severe exercise within the moderate-intensity domain is related to depletion of
(Fig. 6, G and H), which implies that exercise cessation was muscle glycogen and impairment in neuromuscular excitability
not due to central fatigue. Low muscle pH attained during and transmission (15, 28, 49, 50, 62). In addition to being an
severe exercise may attenuate the reduction in muscle mem- essential substrate for regeneration of ATP, it has been dem-
brane excitability (3, 24). Furthermore, muscle glycogen con- onstrated that under conditions in which [ATP] is held high,
tent, a key regulator of sarcoplasmic Ca2⫹ release rate and thus low muscle [glycogen] can impair muscle function (49, 62).
muscle excitability (15, 50), did not fall significantly during The association between low muscle [glycogen] and impaired
severe exercise. Precisely how utilization of W= and the asso- muscle function can be attributed to the modulatory role of
ciated alterations in muscle substrate and metabolite concen- glycogen in the release of Ca2⫹ from the sarcoplasmic
trations and ionic changes influence muscle excitability war- reticulum (15, 19, 20, 28, 49, 50). In keeping with the role
rants further investigation. of glycogen in excitation-contraction coupling, individuals
with glycogen phosphorylase deficiency (McArdle disease)
Fatigue During Heavy-Intensity Exercise do not experience a considerable drop in pH; rather, they
demonstrate an earlier decline in the M-wave amplitude
Heavy-intensity exercise was maintained for an average of
during exercise (16). Furthermore, glucose administration
43.5 min (Tlim ranged from 20.5 to 67.4 min) and, in contrast
during exercise has been shown to partially restore both the
to severe-intensity exercise, no subject achieved V̇O2peak at Tlim
M-wave amplitude and muscle contractility (34, 37, 63),
(~87% V̇O2peak). Consistent with our second hypothesis, the
thus supporting the notion that carbohydrate availability
muscle metabolic perturbation experienced following heavy-
modulates muscle excitability and contractile function. The
intensity exercise was less than that observed following severe-
findings of the present study show that moderate-intensity
intensity exercise, but was greater than that observed following
exercise (⬍GET) can be sustained for a long duration with
moderate-intensity exercise. At Tlim, significant reductions
little change in muscle metabolites and indicate that muscle
were observed in muscle [PCr] (~66%), [ATP] (~12%), [pH]
glycogen depletion is the likely mechanism responsible for
(~97%]), and [glycogen] (~59%), and there was a significant
the decline in neuromuscular function and exercise intoler-
increase in muscle [lactate] (~447%) relative to resting values.
ance in this domain.
Similarly, blood [lactate] and plasma [K⫹] displayed greater
Most research investigations of neuromuscular fatigue de-
perturbation relative to moderate-intensity exercise, but less
velopment during exercise have focused on small muscle
perturbation relative to severe-intensity exercise (Fig. 3). It is
groups and have been limited to the assessment of neuromus-
of interest that the decrease in muscle excitability from rest to
cular function before exercise and as soon as possible (usually
Tlim was greater during heavy-intensity than during severe-
within 2–3 min) after exercise. Considering the task-specific
intensity exercise (Fig. 5). Following the onset of exercise,
nature of neuromuscular fatigue development and the rapid
plasma [K⫹] increased rapidly to attain a peak value at 10 min,
recovery in muscle function (within 2 min) after high-intensity
which was sustained until Tlim; the reduction in M-wave
cycle exercise (26), it is possible that previously reported
amplitude followed a similar temporal profile. It is therefore
changes in neuromuscular function from before exercise to
likely that the initial reduction in M-wave amplitude was a
after exercise underestimate fatigue development during exer-
result of plasma [K⫹] accumulation, which reduced the release
cise. Recently, Sidhu et al. (56) adopted an approach that uses
of Ca2⫹ from the sarcoplasmic reticulum, thus impairing ex-
the motor compound action potential (M-wave) for assessment
citation-contraction coupling (36, 71). As heavy-intensity ex-
of changes in neuromuscular function during cycle exercise.
ercise continued, it is possible that the combined metabolic and
By adopting a similar approach (56), we found large reductions
ionic perturbation, coupled with the ~60% decrease in muscle
in M-wave amplitude and M-wave area in both VL and VM
[glycogen], may have further impaired Ca2⫹ release and cross-
muscles during exercise to Tlim in each discrete exercise
bridge formation (2, 3, 23, 24, 36, 40, 41) and/or the sensitivity
intensity domain. This suggests that changes in muscle excit-
of the myofilaments to Ca2⫹ (17). Although fatigue develop-
ability linked to the fatigue process can occur consequent to a
ment during heavy-intensity exercise appears to be more com-
wide range of perturbations in muscle and blood chemistry,
plicated than it is for severe-intensity exercise, it is related to
with limited differentiation between exercise intensity do-
the combined influence of ionic changes in muscle membrane
mains. The consistency of indices of neuromuscular fatigue
excitability, muscle metabolite accumulation, and the decrease
during severe-intensity cycling exercise in our study contrasts
in energy substrate, which act collectively to impair excitation-
with a recent report by Thomas et al. (65) in which peripheral
contraction coupling.
fatigue, assessed after exercise using electrical stimulation
Fatigue During Moderate-Intensity Exercise during isometric contractions, was greater at higher work rates
within the severe-intensity domain. It is possible that this
Moderate-intensity exercise performed at a work-rate of 20 reflects differences in the experimental techniques employed,
W below GET was continued for an average of 211 min, with and underlines the importance of accounting for the task
subjects working at ~52% V̇O2peak at Tlim. Muscle metabolic specificity of fatigue and the dynamics of muscle recovery after
perturbation was relatively slight in this domain (Fig. 3). For exercise (10).

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al. 457
Conclusion 8. Burnley M, Jones AM. Oxygen uptake kinetics as a determinant of
sports performance. Eur J Sport Sci 7: 63–79, 2007. doi:10.1080/
This study employed a novel and rather comprehensive 17461390701456148.
combination of invasive and noninvasive techniques that en- 9. Burnley M, Vanhatalo A, Fulford J, Jones AM. Similar metabolic
perturbations during all-out and constant force exhaustive exercise in
abled simultaneous assessment of metabolic, ionic, systemic, humans: a 31P magnetic resonance spectroscopy study. Exp Physiol 95:
and neuromuscular factors that define muscular performance. 798 –807, 2010. doi:10.1113/expphysiol.2010.052688.
Although direct measures of the contribution of central factors 10. Burnley M, Vanhatalo A, Jones AM. Distinct profiles of neuromuscular
to fatigue were not employed, peripheral nerve stimulation fatigue during muscle contractions below and above the critical torque in
permitted elucidation of their relative importance in neuromus- humans. J Appl Physiol 113: 215–223, 2012. doi:10.1152/japplphysiol.
00022.2012.
cular fatigue development during exhaustive cycle exercise 11. Cairns SP, Hing WA, Slack JR, Mills RG, Loiselle DS. Different effects
performed within each of the well-defined exercise intensity of raised [K⫹]o on membrane potential and contraction in mouse fast- and
domains. This study is consistent with the notion that GET and slow-twitch muscle. Am J Physiol Cell Physiol 273: C598 –C611, 1997.
CP demarcate exercise intensity domains within which fatigue 12. Calbet JA, Jensen-Urstad M, van Hall G, Holmberg HC, Rosdahl H,
is mediated by distinct mechanisms. Exercise intolerance Saltin B. Maximal muscular vascular conductances during whole body
upright exercise in humans. J Physiol 558: 319 –331, 2004. doi:10.1113/
within the severe-intensity domain (⬎CP) was associated with jphysiol.2003.059287.
the attainment of a consistent critical muscle metabolic milieu 13. Cannon DT, Howe FA, Whipp BJ, Ward SA, McIntyre DJ, Ladroue
(i.e., low [PCr] and pH). In contrast, moderate-intensity exer- C, Griffiths JR, Kemp GJ, Rossiter HB. Muscle metabolism and
cise (⬍GET) was associated with more significant depletion of activation heterogeneity by combined 31P chemical shift and T2 imaging,
muscle [glycogen]. The causes of fatigue during heavy-inten- and pulmonary O2 uptake during incremental knee-extensor exercise. J
Appl Physiol 115: 839 –849, 2013. doi:10.1152/japplphysiol.00510.2013.
sity exercise (⬎GET, ⬍CP) were more obscure with interme- 14. Chidnok W, Fulford J, Bailey SJ, Dimenna FJ, Skiba PF, Vanhatalo
diate changes in muscle metabolic perturbation and glycogen A, Jones AM. Muscle metabolic determinants of exercise tolerance
depletion being apparent. These results are consistent with the following exhaustion: relationship to the “critical power”. J Appl Physiol
notion that both GET and CP demarcate exercise intensity 115: 243–250, 2013. doi:10.1152/japplphysiol.00334.2013.
domains characterized by distinct respiratory and metabolic 15. Chin ER, Allen DG. Effects of reduced muscle glycogen concentration on
force, Ca2⫹ release and contractile protein function in intact mouse
profiles. Strikingly, CP represents a boundary above which
skeletal muscle. J Physiol 498: 17–29, 1997. doi:10.1113/jphysiol.1997.
both metabolic and neuromuscular responses conform to a sp021838.
consistent ceiling or nadir irrespective of work rate and exer- 16. Cooper RG, Stokes MJ, Edwards RH. Myofibrillar activation failure in
cise duration. McArdle’s disease. J Neurol Sci 93: 1–10, 1989. doi:10.1016/0022-
510X(89)90156-1.
DISCLOSURES 17. Debold EP, Fitts RH, Sundberg CW, Nosek TM. Muscle fatigue from
the perspective of a single crossbridge. Med Sci Sports Exerc 48: 2270 –
No conflicts of interest, financial or otherwise, are declared by the authors. 2280, 2016. doi:10.1249/MSS.0000000000001047.
18. De Luca CJ. The use of surface electromyography in biomechanics. J
AUTHOR CONTRIBUTIONS Appl Biomech 13: 135–163, 1993. doi:10.1123/jab.13.2.135.
19. Duhamel TA, Green HJ, Perco JG, Ouyang J. Effects of prior exercise
M.I.B., A.M.J., S.J.B., K.J.M., J.L.B., and A.V. conceived and designed and a low-carbohydrate diet on muscle sarcoplasmic reticulum function
research; M.I.B., A.M.J., J.R.B., S.J.B., L.J.W., S.T.J.M., C.T., J.K., and during cycling in women. J Appl Physiol 101: 695–706, 2006. doi:10.
A.V. performed experiments; M.I.B., J.R.B., L.J.W., C.T., J.K., K.J.M.,
1152/japplphysiol.00052.2006.
and J.L.B. analyzed data; M.I.B., A.M.J., J.R.B., S.J.B., P.S., K.J.M.,
20. Duhamel TA, Perco JG, Green HJ. Manipulation of dietary carbohy-
J.L.B., and A.V. interpreted results of experiments; M.I.B. and A.V.
drates after prolonged effort modifies muscle sarcoplasmic reticulum
prepared figures; M.I.B. and K.J.M. drafted manuscript; M.I.B., A.M.J.,
responses in exercising males. Am J Physiol Regul Integr Comp Physiol
L.J.W., K.J.M., J.L.B., and A.V. edited and revised manuscript; M.I.B.,
291: R1100 –R1110, 2006. doi:10.1152/ajpregu.00858.2005.
A.M.J., J.R.B., S.J.B., L.J.W., S.T.J.M., C.T., J.K., P.S., K.J.M., J.L.B., and
21. Enoka RM, Duchateau J. Muscle fatigue: what, why and how it influ-
A.V. approved final version of manuscript.
ences muscle function. J Physiol 586: 11–23, 2008. doi:10.1113/jphysiol.
2007.139477.
REFERENCES 22. Enoka RM, Stuart DG. Neurobiology of muscle fatigue. J Appl Physiol
1. Ahlborg B, Bergstrom J, Ekelund LG, Hultman E. Muscle glycogen 72: 1631–1648, 1992.
and muscle electrolytes during prolonged physical exercise. Acta Physiol 23. Fitts RH. Cellular mechanisms of muscle fatigue. Physiol Rev 74: 49 –94,
Scand 70: 129 –142, 1967. doi:10.1111/j.1748-1716.1967.tb03608.x. 1994.
2. Allen DG. Fatigue in working muscles. J Appl Physiol 106: 358 –359, 24. Fitts RH. The cross-bridge cycle and skeletal muscle fatigue. J Appl
2009. doi:10.1152/japplphysiol.91599.2008. Physiol 104: 551–558, 2008. doi:10.1152/japplphysiol.01200.2007.
3. Allen DG, Lamb GD, Westerblad H. Skeletal muscle fatigue: cellular 25. Fowles JR, Green HJ, Tupling R, O’Brien S, Roy BD. Human neuro-
mechanisms. Physiol Rev 88: 287–332, 2008. doi:10.1152/physrev.00015. muscular fatigue is associated with altered Na⫹-K⫹-ATPase activity
2007. following isometric exercise. J Appl Physiol 92: 1585–1593, 2002. doi:
4. Amann M, Dempsey JA. Ensemble input of group III/IV muscle afferents 10.1152/japplphysiol.00668.2001.
to CNS: a limiting factor of central motor drive during endurance exercise 26. Froyd C, Millet GY, Noakes TD. The development of peripheral fatigue
from normoxia to moderate hypoxia. Adv Exp Med Biol 903: 325–342, and short-term recovery during self-paced high-intensity exercise. J
2016. doi:10.1007/978-1-4899-7678-9_22. Physiol 591: 1339 –1346, 2013. doi:10.1113/jphysiol.2012.245316.
5. Beaver WL, Wasserman K, Whipp BJ. A new method for detecting 27. Gandevia SC. Spinal and supraspinal factors in human muscle fatigue.
anaerobic threshold by gas exchange. J Appl Physiol 60: 2020 –2027, Physiol Rev 81: 1725–1789, 2001.
1986. 28. Gejl KD, Hvid LG, Frandsen U, Jensen K, Sahlin K, Ørtenblad N.
6. Bergstrom J. Percutaneous needle biopsy of skeletal muscle in physio- Muscle glycogen content modifies SR Ca2⫹ release rate in elite endurance
logical and clinical research. Scand J Clin Lab Invest 35: 609 –616, 1975. athletes. Med Sci Sports Exerc 46: 496 –505, 2014. doi:10.1249/MSS.
doi:10.3109/00365517509095787. 0000000000000132.
7. Black MI, Jones AM, Bailey SJ, Vanhatalo A. Self-pacing increases 29. Gollnick PD, Piehl K, Saltin B. Selective glycogen depletion pattern in
critical power and improves performance during severe-intensity exercise. human muscle fibres after exercise of varying intensity and at varying
Appl Physiol Nutr Metab 40: 662–670, 2015. doi:10.1139/apnm-2014- pedalling rates. J Physiol 241: 45–57, 1974. doi:10.1113/jphysiol.1974.
0442. sp010639.

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
458 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al.

30. Green HJ. Cation pumps in skeletal muscle: potential role in muscle 51. Poole DC, Ward SA, Gardner GW, Whipp BJ. Metabolic and respira-
fatigue. Acta Physiol Scand 162: 201–213, 1998. doi:10.1046/j.1365- tory profile of the upper limit for prolonged exercise in man. Ergonomics
201X.1998.0300f.x. 31: 1265–1279, 1988. doi:10.1080/00140138808966766.
31. Hogan MC, Richardson RS, Haseler LJ. Human muscle performance 52. Rossman MJ, Garten RS, Venturelli M, Amann M, Richardson RS.
and PCr hydrolysis with varied inspired oxygen fractions: a 31P-MRS The role of active muscle mass in determining the magnitude of peripheral
study. J Appl Physiol 86: 1367–1373, 1999. fatigue during dynamic exercise. Am J Physiol Regul Integr Comp Physiol
32. Jones AM, Poole DC. Introduction to oxygen uptake kinetics and histor- 306: R934 –R940, 2014. doi:10.1152/ajpregu.00043.2014.
ical development of the discipline. In Oxygen uptake kinetics in sport, 53. Rossman MJ, Venturelli M, McDaniel J, Amann M, Richardson RS.
exercise and medicine (Jones AM, Poole DC, eds.). London: Routledge, Muscle mass and peripheral fatigue: a potential role for afferent feedback?
2005, p. 2–35. Acta Physiol (Oxf) 206: 242–250, 2012. doi:10.1111/j.1748-1716.2012.
33. Jones AM, Wilkerson DP, DiMenna F, Fulford J, Poole DC. Muscle 02471.x.
metabolic responses to exercise above and below the “critical power” 54. Sahlin K, Söderlund K, Tonkonogi M, Hirakoba K. Phosphocreatine
assessed using 31P-MRS. Am J Physiol Regul Integr Comp Physiol 294: content in single fibers of human muscle after sustained submaximal
R585–R593, 2008. doi:10.1152/ajpregu.00731.2007. exercise. Am J Physiol Cell Physiol 273: C172–C178, 1997.
34. Karelis AD, Péronnet F, Gardiner PF. Glucose infusion attenuates 55. Sejersted OM, Sjøgaard G. Dynamics and consequences of potassium
muscle fatigue in rat plantaris muscle during prolonged indirect stimula- shifts in skeletal muscle and heart during exercise. Physiol Rev 80:
1411–1481, 2000.
tion in situ. Exp Physiol 87: 585–592, 2002. doi:10.1113/eph8702391.
56. Sidhu SK, Cresswell AG, Carroll TJ. Motor cortex excitability does not
35. Lowry OH, Passonneau JV. A flexible system of enzymatic analysis. New
increase during sustained cycling exercise to volitional exhaustion. J Appl
York: Academic Press, 1972.
Physiol 113: 401–409, 2012. doi:10.1152/japplphysiol.00486.2012.
36. MacIntosh BR, Holash RJ, Renaud JM. Skeletal muscle fatigue–
57. Smith JL, Martin PG, Gandevia SC, Taylor JL. Sustained contraction
regulation of excitation-contraction coupling to avoid metabolic catastro-
at very low forces produces prominent supraspinal fatigue in human elbow
phe. J Cell Sci 125: 2105–2114, 2012. doi:10.1242/jcs.093674. flexor muscles. J Appl Physiol 103: 560 –568, 2007. doi:10.1152/
37. Marcil M, Karelis AD, Péronnet F, Gardiner PF. Glucose infusion japplphysiol.00220.2007.
attenuates fatigue without sparing glycogen in rat soleus muscle during 58. Smith JR, Ade CJ, Broxterman RM, Skutnik BC, Barstow TJ, Wong
prolonged electrical stimulation in situ. Eur J Appl Physiol 93: 569 –574, BJ, Harms CA. Influence of exercise intensity on respiratory muscle
2005. doi:10.1007/s00421-004-1273-4. fatigue and brachial artery blood flow during cycling exercise. Eur J Appl
38. Martin V, Kerhervé H, Messonnier LA, Banfi JC, Geyssant A, Bon- Physiol 114: 1767–1777, 2014. doi:10.1007/s00421-014-2905-y.
nefoy R, Féasson L, Millet GY. Central and peripheral contributions to 59. Sjøgaard G. Electrolytes in slow and fast muscle fibers of humans at rest
neuromuscular fatigue induced by a 24-h treadmill run. J Appl Physiol and with dynamic exercise. Am J Physiol Regul Integr Comp Physiol 245:
108: 1224 –1233, 2010. doi:10.1152/japplphysiol.01202.2009. R25–R31, 1983.
39. Matkowski B, Place N, Martin A, Lepers R. Neuromuscular fatigue 60. Sjøgaard G. Water and electrolyte fluxes during exercise and their
differs following unilateral vs bilateral sustained submaximal contractions. relation to muscle fatigue. Acta Physiol Scand Suppl 556: 129 –136, 1986.
Scand J Med Sci Sports 21: 268 –276, 2011. doi:10.1111/j.1600-0838. 61. Søgaard K, Gandevia SC, Todd G, Petersen NT, Taylor JL. The effect
2009.01040.x. of sustained low-intensity contractions on supraspinal fatigue in human
40. McKenna MJ. The roles of ionic processes in muscular fatigue during elbow flexor muscles. J Physiol 573: 511–523, 2006. doi:10.1113/
intense exercise. Sports Med 13: 134 –145, 1992. doi:10.2165/00007256- jphysiol.2005.103598.
199213020-00009. 62. Stephenson DG, Nguyen LT, Stephenson GM. Glycogen content and
41. McKenna MJ, Bangsbo J, Renaud JM, Muscle K. Muscle K⫹, Na⫹, excitation-contraction coupling in mechanically skinned muscle fibres of
and Cl⫺ disturbances and Na⫹-K⫹ pump inactivation: implications for the cane toad. J Physiol 519: 177–187, 1999. doi:10.1111/j.1469-7793.
fatigue. J Appl Physiol 104: 288 –295, 2008. doi:10.1152/japplphysiol. 1999.0177o.x.
01037.2007. 63. Stewart RD, Duhamel TA, Foley KP, Ouyang J, Smith IC, Green HJ.
42. Millet GY, Lepers R. Alterations of neuromuscular function after pro- Protection of muscle membrane excitability during prolonged cycle exer-
longed running, cycling and skiing exercises. Sports Med 34: 105–116, cise with glucose supplementation. J Appl Physiol 103: 331–339, 2007.
2004. doi:10.2165/00007256-200434020-00004. doi:10.1152/japplphysiol.01170.2006.
43. Monod H, Scherrer J. The work capacity of a synergic muscle group. 64. Tomazin K, Morin JB, Strojnik V, Podpecan A, Millet GY. Fatigue after
Ergonomics 8: 329 –338, 1965. doi:10.1080/00140136508930810. short (100-m), medium (200-m) and long (400-m) treadmill sprints. Eur J
44. Moritani T, Nagata A, deVries HA, Muro M. Critical power as a Appl Physiol 112: 1027–1036, 2012. doi:10.1007/s00421-011-2058-1.
measure of physical work capacity and anaerobic threshold. Ergonomics 65. Thomas K, Elmeua M, Howatson G, Goodall S. Intensity-depen-
24: 339 –350, 1981. doi:10.1080/00140138108924856. dent contribution of neuromuscular fatigue after constant-load cycling.
45. Morris MG, Dawes H, Howells K, Scott OM, Cramp M, Izadi H. Med Sci Sports Exerc 48: 1751–1760, 2016. doi:10.1249/MSS.
Alterations in peripheral muscle contractile characteristics following high 0000000000000950.
and low intensity bouts of exercise. Eur J Appl Physiol 112: 337–343, 66. Thomas K, Goodall S, Stone M, Howatson G, St Clair Gibson A,
Ansley L. Central and peripheral fatigue in male cyclists after 4-, 20-, and
2012. doi:10.1007/s00421-011-1970-8.
40-km time trials. Med Sci Sports Exerc 47: 537–546, 2015. doi:10.1249/
46. Mortensen SP, Damsgaard R, Dawson EA, Secher NH, González-
MSS.0000000000000448.
Alonso J. Restrictions in systemic and locomotor skeletal muscle perfu-
67. Vanderthommen M, Duteil S, Wary C, Raynaud JS, Leroy-Willig A,
sion, oxygen supply and VO2 during high-intensity whole-body exercise
Crielaard JM, Carlier PG. A comparison of voluntary and electrically
in humans. J Physiol 586: 2621–2635, 2008. doi:10.1113/jphysiol.2007. induced contractions by interleaved 1H- and 31P-NMRS in humans. J Appl
149401. Physiol 94: 1012–1024, 2003. doi:10.1152/japplphysiol.00887.2001.
47. Murgatroyd SR, Ferguson C, Ward SA, Whipp BJ, Rossiter HB. 68. Vanhatalo A, Black MI, DiMenna FJ, Blackwell JR, Schmidt JF,
Pulmonary O2 uptake kinetics as a determinant of high-intensity exercise Thompson C, Wylie LJ, Mohr M, Bangsbo J, Krustrup P, Jones AM.
tolerance in humans. J Appl Physiol 110: 1598 –1606, 2011. doi:10.1152/ The mechanistic bases of the power-time relationship: muscle metabolic
japplphysiol.01092.2010. responses and relationships to muscle fibre type. J Physiol 594: 4407–
48. Newham DJ, Cady EB. A 31P study of fatigue and metabolism in human 4423, 2016. doi:10.1113/JP271879.
skeletal muscle with voluntary, intermittent contractions at different 69. Vanhatalo A, Fulford J, DiMenna FJ, Jones AM. Influence of hyper-
forces. NMR Biomed 3: 211–219, 1990. doi:10.1002/nbm.1940030504. oxia on muscle metabolic responses and the power-duration relation-
49. Nielsen J, Schrøder HD, Rix CG, Ortenblad N. Distinct effects of ship during severe-intensity exercise in humans: a 31P magnetic reso-
subcellular glycogen localization on tetanic relaxation time and endurance nance spectroscopy study. Exp Physiol 95: 528 –540, 2010. doi:10.
in mechanically skinned rat skeletal muscle fibres. J Physiol 587: 3679 – 1113/expphysiol.2009.050500.
3690, 2009. doi:10.1113/jphysiol.2009.174862. 70. Vanhatalo A, Poole DC, DiMenna FJ, Bailey SJ, Jones AM. Muscle
50. Ørtenblad N, Nielsen J, Saltin B, Holmberg HC. Role of glycogen fiber recruitment and the slow component of O2 uptake: constant work rate
availability in sarcoplasmic reticulum Ca2⫹ kinetics in human skeletal vs. all-out sprint exercise. Am J Physiol Regul Integr Comp Physiol 300:
muscle. J Physiol 589: 711–725, 2011. doi:10.1113/jphysiol.2010.195982. R700 –R707, 2011. doi:10.1152/ajpregu.00761.2010.

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.
 Metabolic and Neuromuscular Correlates of Fatigue • Black MI et al. 459
71. Westerblad H, Allen DG. Cellular mechanisms of skeletal muscle fa- 73. Whipp BJ, Wasserman K. Oxygen uptake kinetics for various intensities
tigue. Adv Exp Med Biol 538: 563–570, 2003. doi:10.1007/978-1-4419- of constant-load work. J Appl Physiol 33: 351–356, 1972.
9029-7_50. 74. Wüthrich TU, Eberle EC, Spengler CM. Locomotor and diaphragm
72. Whipp BJ, Ward SA. Pulmonary gas exchange dynamics and the muscle fatigue in endurance athletes performing time-trials of different
tolerance to muscular exercise: effects of fitness and training. Ann Physiol durations. Eur J Appl Physiol 114: 1619 –1633, 2014. doi:10.1007/
Anthropol 11: 207–214, 1992. doi:10.2114/ahs1983.11.207. s00421-014-2889-7.

J Appl Physiol • doi:10.1152/japplphysiol.00942.2016 • www.jappl.org

Downloaded from journals.physiology.org/journal/jappl at Pontificia Univ Catolica De Chile (146.155.210.019) on August 24, 2022.