International Journal of

Environmental Research
and Public Health

Article
Neuromuscular Fatigue Responses of Endurance- and
Strength-Trained Athletes during Incremental Cycling Exercise
Maciej Jurasz 1 , Michał Boraczyński 2, * , Zbigniew Wójcik 3 and Piotr Gronek 4

1 Department of Sport Medicine and Traumatology, Poznan University of Physical Education,

61-871 Poznań, Poland; [email protected]
2 Faculty of Health Sciences, Collegium Medicum, University of Warmia and Mazury in Olsztyn,
10-719 Olsztyn, Poland
3 Faculty of Health Sciences, Olsztyn University, 10-283 Olsztyn, Poland; [email protected]
4 Laboratory of Healthy Aging, Department of Dance and Gymnastics, Poznan University of Physical
Education, 61-871 Poznań, Poland; [email protected]
* Correspondence: [email protected]; Tel.: +48-533-101-720

Abstract: This study explored the development of neuromuscular fatigue responses during pro-
gressive cycling exercise. The sample comprised 32 participants aged 22.0 ± 0.54 years who were
assigned into three groups: endurance-trained group (END, triathletes, n = 10), strength-trained
group (STR, bodybuilders, n = 10) and control group (CG, recreationally active students, n = 12). The
incremental cycling exercise was performed using a progressive protocol starting with a 3 min resting
measurement and then 50 W workload with subsequent constant increments of 50 W every 3 min
until 200 W. Surface electromyography (SEMG) of rectus femoris muscles was recorded during the
final 30 s of each of the four workloads. During the final 15 s of each workload, participants rated
their overall perception of effort using the 20-point rating of the perceived exertion (RPE) scale. Post
hoc Tukey’s HSD testing showed significant differences between the END and STR groups in median
Citation: Jurasz, M.; Boraczyński, M.;
Wójcik, Z.; Gronek, P. Neuromuscular
frequency and mean power frequency across all workloads (p < 0.001 and p < 0.01, respectively).
Fatigue Responses of Endurance- and Athletes from the END group had significantly lower electromyogram amplitude responses than
Strength-Trained Athletes during those from the STR (p = 0.0093) and CG groups (p = 0.0006). Increasing RPE points from 50 to 200 W
Incremental Cycling Exercise. Int. J. were significantly higher in the STR than in the END group (p < 0.001). In conclusion, there is a
Environ. Res. Public Health 2022, 19, significant variation in the neuromuscular fatigue profiles between athletes with different training
8839. https://doi.org/10.3390/ backgrounds when a cycling exercise is applied. The approximately linear trends of the SEMG and
ijerph19148839 RPE values of both groups of athletes with increasing workload support the increased skeletal muscle
Academic Editors: Paul B. recruitment with perceived exertion or fatiguing effect.
Tchounwou and Martin Burtscher
Keywords: neuromuscular fatigue; muscle bioelectrical activity; rating of perceived exertion;
Received: 31 May 2022
cycling exercise
Accepted: 19 July 2022
Published: 21 July 2022

Publisher’s Note: MDPI stays neutral

with regard to jurisdictional claims in 1. Introduction
published maps and institutional affil-
Scientific interest in muscle stimulation (myoelectric activity) and fatigue in athlete
iations.
performance has increased in recent years [1]. Neuromuscular fatigue can occur via central
and/or peripheral sites along the motor pathway of force production. In general, central or
neural fatigue (affecting the proximal motor neurons of the brain and spinal cord) involves
Copyright: © 2022 by the authors.
insufficient neural drive to the muscles, whereas peripheral fatigue, where the motor units
Licensee MDPI, Basel, Switzerland. (MUs) of the peripheral nerves, motor endplates and muscle fibers are affected, involves
This article is an open access article changes beyond the neuromuscular junction [2]. Several previous cross-sectional studies
distributed under the terms and have reported that specific power, strength and endurance modes of training influence
conditions of the Creative Commons respective neuromuscular fatigue (NMF) profiles in trained athletes [2,3]. Moreover, within
Attribution (CC BY) license (https:// endurance-trained athletes, the type of locomotion may vary with the specific sport and
creativecommons.org/licenses/by/ further influence the NMF profile [2]. There are a number of mechanical (e.g., decreased
4.0/). countermovement jump height) and metabolic (e.g., lactate accumulation) measures of

Int. J. Environ. Res. Public Health 2022, 19, 8839. https://doi.org/10.3390/ijerph19148839 https://www.mdpi.com/journal/ijerph
Int. J. Environ. Res. Public Health 2022, 19, 8839 2 of 13

fatigue [4,5]. In addition, psychophysiological variables such as the rating of perceived

exertion (RPE),which is related to physiological markers of the stress response to exercise,
are widely used. RPE has been used to investigate the mechanisms of fatigue as well as
to prescribe exercise protocols in numerous populations [6–8]. Another common, non-
invasive and objective method to measure neuromuscular fatigue is electromyography
(EMG). While EMG can be assessed both via surface electrodes and intramuscularly, for
the purposes of this paper we will focus on conventional bipolar surface EMG (SEMG). In
general, SEMG provides a global view of skeletal muscle function as well as the activity of
individual MUs [9]. The analysis of multielectrode recordings via SEMG sensors, which
display muscle activation patterns (the algebraic summation of all MU action potentials),
has been extensively used in the fields of sports science and medical research.
Of interest is that the RPE responses during resistance and endurance performance
have been associated with the degree of skeletal muscle recruitment measured by SEMG [10].
Multiple studies have shown a linear increase in the SEMG and RPE during exhaustive
constant- and progressive-load exercise [11,12], which allows the estimation of their in-
creasing rates and helps to estimate the effect of engaged muscle mass on the magnitude
of performance fatigability. Increased SEMG activity is a good indicator of a change in
muscular neural drive [13] and has been used to interpret both dysfunctional and func-
tional muscle recruitment patterns related to cycling activity [14]. For high-intensity cycling
exercise, a relatively consistent magnitude of peripheral fatigue has been shown in various
exercise protocols [15]. However, this consistency of peripheral fatigue (especially at ex-
ercise termination), which reflects a centrally mediated “individual critical threshold”, is
probably task- or training-dependent [16,17]. To the best of our knowledge, NMF profiles
and specific adaptive differences between cardiovascular endurance and muscular strength
modes of training have not been adequately investigated in moderately experienced ath-
letes. Hence, the question arises whether differences in muscle activity, subjective fatigue
perception and associated decrements in cycling performance exist during a submaximal
incremental cycling test in athletes trained in triathlon and bodybuilding, and recreationally
active men. Because changes in muscle activation patterns and kinematics under dynamic
fatigue have been previously observed, identifying any specific training-dependent differ-
ences would be beneficial for determining the likelihood of musculoskeletal injury and for
designing appropriate motor control training protocols.
This study was designed to investigate and compare the development of exercise-
induced neuromuscular and perceptual fatigue responses using SEMG and RPE measure-
ments during an incremental submaximal cycling exercise. We hypothesized that the
specific training background (cardiovascular endurance versus muscular strength) would
influence the neuromuscular fatigue profiles, as expressed by SEMG activity and RPE-based
subjective fatigue perception in endurance- and strength-trained athletes and recreationally
active men.

2. Materials and Methods

2.1. Participants
A total of forty-seven male participants (mean ± SD: age 22.0 ± 0.5 years, range
19–25 years) from Poznań in the western region of Poland were recruited to take part in a
cross-sectional observational study. The recruitment process was conducted by academic
website information, email and recruitment flyers. The thirty-two eligible participants were
placed into their three respective groups based on their training history. The endurance-
trained group (END; triathletes, n = 10, age 20.3 ± 0.6 years, body mass 75.1 ± 2.9 kg, body
height 179.8 ± 2.1 cm, body mass index 23.2 ± 0.64 kg·m–2 ) reported training regularly for
3–7 years. Their weekly training volumes averaged 14.5 ± 1.3 h, with 1–2 daily exercise
sessions in the three triathlon disciplines (4.9 ± 0.3 h for swimming, 5.7 ± 0.6 h for cycling,
3.1 ± 0.3 h for running and 0.8 ± 0.1 h for specific preparation). Their average training
intensity was 15–17 on the RPE scale.
Int. J. Environ. Res. Public Health 2022, 19, 8839 3 of 13

The strength-trained athletes (STR; bodybuilders, n= 10, age 22.4 ± 0.4 years, body
mass 83.9 ± 2.7 kg, body height 181.9 ± 2.4 cm, body mass index 25.4 ± 0.62 kg·m–2 )
reported training 5.1 ± 0.2 sessions/week (2.4 ± 0.3 h/session) with 4–7 years experience
in consistent bodybuilding exercise (minimum of three sessions per week lifting workouts).
Their training intensity rating was 15–16 on the RPE scale. Typically, training programs
reported by the participants in STR group consisted of multi-joint weightlifting exercises
to promote sufficient stimuli for gaining muscle size and strength in all muscles involved
in the exercise. The bodybuilders performed 2–3 circuits of 10–12 exercises (total-, upper-,
and lower-body muscles), with each set consisting of 6–12 repetitions.
The recreationally active control individuals (CG, n = 12, age 23.2 ± 0.4 years, body
mass 78.2 ± 2.6 kg, body height 181.6 ± 1.8 cm, body mass index 23.7 ± 0.64 kg·m–2 )
consisted of students participating in a variety of recreational activities. Participants from
CG group reported the frequency and duration of different types of activity: vigorous (i.e.,
heavy lifting, performing high-intensity exercises, using treadmill); moderate (i.e., carrying
light loads and bicycling at a regular pace); walking activities and the average time spent
sitting on a weekday. According to the IPAQ scoring recommendations, these participants
were classified as moderately active (≥600 MET–minutes/week). Their mean total MET
score was 954 ± 287.3 MET–minutes/week.

2.2. Eligibility Criteria

The inclusion criteria for END and STR groups were: male, aged 18–30 years, self-
reported regular engagement in endurance or strength training, at least 3 years of com-
petitive experience, no current sustained experience in the other modes of training, non-
smokers, not taking performance-enhancing drugs and reporting no medical contraindica-
tions to participate in exercise testing (e.g., knee injury or neuromuscular disorder). For
the CG group the inclusion criteria included: male, aged 18–30 years, non-smokers, no
reported medical contraindications to participate in exercise testing, a BMI < 25 kg/m2 and
active lifestyle (>600 MET) confirmed by an adapted version of short-form of International
Physical Activity Questionnaire (IPAQ-SF). Participants from CG group were defined as
recreationally active if they performed regular exercise (moderate-to-vigorous exercise
2–3 times per week), but did not train or compete within a sport at a competitive level.

2.3. Ethical Approval

All participants gave written informed consent and verbal assent regarding their
participation in the study. The research and protocol were approved by the Scientific
Ethical Committee of the Karol Marcinkowski University of Medical Sciences in Poznań.
The procedures performed in this study involving human participants were in accordance
with the ethical guidelines of the 2013 World Medical Association Declaration of Helsinki
(2013: seventh revision, 64th meeting, Fortaleza, Brazil).

2.4. Design and Procedures

Data collection was performed over a 2-week period, in which participants took part
in laboratory tests on two separate occasions with a minimum of 48 h between sessions.
During the first visit to the laboratory, anthropometric data were collected, and all par-
ticipants underwent a non-invasive pre-participation cardiovascular screening (based on
12-lead ECG). During the second visit, participants had a familiarization session (cycle
ergometer and the testing procedure) and performed incremental cycling test. Testing was
conducted in the morning hours (9:00–11:00). Participants performed tests in condition-
controlled laboratory room with a fan (average temperature: 22.5 ± 0.5 ◦ C, relative humid-
ity: 35.0 ± 5.0%, barometric pressure: 760–770 mmHg).

2.5. Pre-Test Preparation

Participants were required to refrain from strenuous exercise during the 48 h prior to
the incremental cycling test session. No alcohol, caffeine and tobacco were allowed for 24 h
Int. J. Environ. Res. Public Health 2022, 19, 8839 4 of 13

prior to testing. In addition, participants were asked to avoid using any ergogenic aids for
at least 48 h before the testing session. They were also instructed to refrain from eating for
2-h prior to testing and only drink water.

2.6. Anthropometry
Body height (BH) was measured to the nearest 0.1 cm using a patient weighing scale
with a height rod (Seca 217, Hamburg, Germany). Body mass (BM, after removal of shoes
and heavy clothing) was measured to the nearest 0.1 kg.

2.7. Incremental Cycling Test

The exercise test was performed using the electromagnetically braked cycle ergometer
(Ergometrics 900 S, Ergo-line GmbH, Bitz, Germany). Before the study, the cycle ergometer
was calibrated for power outputs of 25–1000 W at different cadences and was found to
be within 1% of a true value. The cycle ergometer was set according to each participant’s
anatomy: the participant adjusted saddle height and the handlebars to his own cycling
posture. However, the trunk forward inclination angle did not exceed 20◦ , and the knee
flexion angle was 5◦ with the extended lower limb. Three-minute standardized warm-up
protocol was followed prior to testing. The latter was performed at a constant work-
load (1.5–2.0 W·kg–1 of BM) and pedaling cadence (70 rotations·min−1 ) interspersed with
2 all-out sprints of 2–3 s (~90 rotations·min−1 ) to elicit HR between 150 and 160 beats·min−1 .
Participants then rested for 15 min prior to incremental cycling test. This graded test was
performed using a progressive protocol according to Dufour et al. [18]. The test started
with a 3min resting measurement in the sitting position and then 50 W workload with
3 subsequent increments of 50 W every 3 min (100, 150 and 200 W). The cadence was set at
70 rotations·min−1 and was maintained via the use of a metronome.

2.8. Ratings of Perceived Exertion (RPE)

During the familiarization session participants were given standard instructions and
anchoring procedures for overall perception of effort using the 6–20-point RPE scale [19].
During the final 15 s of each workload, participants were asked to rate their overall
perception of effort using any number on a large RPE scale displayed in front of them. A
rating of 6 was associated with “no exertion at all” (rest) and a rating of 20 was considered
to be a “maximal exertion” of effort and associated with the most stressful exercise ever
performed. RPE is widely used in exercise tests using cycle ergometers (incremental aerobic
exercise).

2.9. Surface Electromyography (SEMG)

Bipolar SEMG is a well-established method for evaluating muscular load. During
the laboratory visit, pre-gelled surface electrodes (Ag/AgCl, AccuSensor, Lynn Medical,
Wixom, MI, USA) were placed in a bipolar arrangement (20-mm center-to-center) on the
dominant rectus femoris (RF) muscles (27 participants were right-legged and 5 left-legged).
Dominant limb was determined using methods consistent with those described by Hebbal
and Mysorekar [20]. The recording electrodes were fixed lengthwise over the anatomic
reference points of the muscle according to the Surface Electromyography for a Non-
invasive Assessment of Muscles (SENIAM) recommendations [21]. In detail, the surface
electrodes were placed as follows: distal to the midpoint along the line connecting the
superior side of the patella and anterior superior iliac spine. Before placing the Ag/AgCl
adhesive electrodes, the skin was shaved, cleaned with ether–alcohol solution and antiseptic
cotton, and dried to control a low impedance at the skin–electrode interface (Z< 5 kΩ).To
ensure the reliability of the SEMG signals, the same researcher positioned the electrodes on
all the participants. Throughout the incremental cycling exercise, an electromyographic
recording system—the Muscle TesterME3000Pwith MegaWin software version 1.41 (Mega
Electronics Ltd., Kuopio, Finland)—was used to display and analyze SEMG signals in
the RF muscles. The obtained data were directly stored on a disk of a personal computer
 Int. J. Environ. Res. Public Health 2022, 19, 8839 5 of 13

for further offline analysis. Electrical activity of muscles (quantified in microvolts) was
recorded four times: in the last 30 s of the set workload values. The EMG signals were
analogically amplified with gain of 50×, band-pass filtered at 10–500 Hz and sampled
through a 12-bit analog-to-digital (A/D) converter with a sampling frequency of 1000 Hz.
The EMG signal was calculated for 0.5-s segments.

2.10. Statistical Analysis

Statistica 13.3 software package [22] and IBM SPSS Statistics ver. 26.0.0.1 [23] were
utilized for statistical analysis. All the results considered were reported as mean ± SD.
Kolmogorov–Smirnov normality test revealed that all the data were normally distributed.
An analysis of variance (ANOVA) model for repeated measures was used to assess
differences between means. A two-way mixed design was applied to test for main and
interaction effects: group (END, STR, CG) was used as a fixed factor, while workload level
(50, 100, 150 and 200 W) was used as a repeated factor. The sphericity of variance for
repeated measures was tested using Mauchly’s test [24]. In the case of lack of sphericity
of variance, the epsilon Greenhouse–Geisser correction was applied. Detailed differences
between means were tested post hoc using the Tukey HSD test [25]. A significance level
was set at α=0.05. The effect size (ES) for each ANOVA effect was presented as a partial eta
squared (η2 p ). The interpretation of ES was based on benchmarks established by Cohen [26]
where d = 0.01 indicates a small effect, 0.06 a medium effect and 0.14 a large effect.

3. Results
A total of 32 electromyograms were recorded for RF muscles of a quadriceps perform-
ing dynamic work during concentric contractions (collected during the last 30 s of each
J. Environ. Res. Public Health 2022, 19,4xmin
FORworkload).
PEER REVIEW Figures 1–4 show the comparisons between the mean values ± standard
deviation (mean ± SD) of SEMG variables (MF, MPF, AEMG) and RPE obtained (1) at the
subsequent workloads (50, 100, 150 and 200 W) and (2) for different groups: endurance-trained
athletes (END), strength-trained athletes (STR) and control group (CG).

Figure 1. The changes in median frequency (MF) in the research groups (END—triathlon;
Figure 1. The changes
STR—bodybuilder; in median
CG—recreationally frequency
active) (MF) in
during progressive the exercise
cycling researchtest atgroups
subse- (END
STR—bodybuilder; CG—recreationally
quent workloads; point active) during progressive cycling exercise te
and whiskers—mean ± SD.
quent workloads; point and whiskers—mean ± SD.
 Figure 1. The changes in median frequency (MF) in the research groups (END—triathlon;
STR—bodybuilder; CG—recreationally active) during progressive cycling exercise test at subse-
Int. J. Environ. Res. Public Health 2022, 19,workloads;
quent 8839 point and whiskers—mean ± SD. 6 of 13

Int. J. Environ. Res. Public Health 2022, 19, x FOR PEER REVIEW 7 of 13
Int. J. Environ. Res. Public Health 2022, 19, x FOR PEER REVIEW 7 of 13
Figure 2. The
Figure changes
2. The in mean
changes power power
in mean frequency (MPF) in(MPF)
frequency the research
in thegroups(END—triathlon;
research groups(END—triathlon;
STR—bodybuilder; CG—recreationally active) during progressive
STR—bodybuilder; CG—recreationally active) during progressive cyclingcycling
exerciseexercise
test at subse-
test at subsequent
quent workloads; point and whiskers—mean ± SD.
workloads; point and whiskers—mean ± SD.

Figure 3. The changes in average amplitude of EMG (AEMG) in the research groups
(END—triathlon;
Figure
Figure 3. The STR—bodybuilder;
Thechanges
changesinin
average
average CG—recreationally
amplitude of EMG
amplitude active)
(AEMG)
of EMG during
(AEMG) progressive
in theinresearch cycling
groups
the research exer-
(END—triathlon;
groups
cise test at subsequent
(END—triathlon;
STR—bodybuilder; workloads; point
STR—bodybuilder;
CG—recreationally and whiskers—mean
CG—recreationally
active) ± SD.
duringactive) during cycling
progressive progressive cycling
exercise testexer-
at subsequent
cise test at subsequent workloads; point and whiskers—mean ± SD.
workloads; point and whiskers—mean ± SD.

Figure 4.4. The

Figure Thechanges in ratings
changes of perceived
in ratings exertion
of perceived (RPE)(RPE)
exertion in the in
research groups (END—triathlon;
the research groups
(END—triathlon;
Figure
STR—bodybuilder; STR—bodybuilder;
4. The changes in ratings CG—recreationally
CG—recreationally of active)
perceived active)
exertion
during duringin
(RPE)
progressive progressive cycling
the exercise
cycling research exer-
groups
test at subsequent
cise test at subsequent
(END—triathlon; workloads; point
STR—bodybuilder; and whiskers—mean
CG—recreationally ± SD.
active) during progressive cycling exer-
workloads; point and whiskers—mean ± SD.
cise test at subsequent workloads; point and whiskers—mean ± SD.
3.1. Median
3.1. MedianFrequency
Frequency (MF)
(MF)
3.1. Median
There Frequency
was no (MF)
interaction effect for mean MF; however, we found
There was no interaction effect for mean MF; however, we the
found effect
theforeffect
the for the
There level
workload was no interaction
(F(3,87) = 15.61, effect for mean
p < 0.001, MF;
η2p = 0.350,
2 however,
power = we found
0.999). Thethelevel
effect
of for
MFthe in-
workload level (F(3,87) = 15.61, p < 0.001, η p = 0.350, power = 0.999). The level of MF
workload level (F(3,87) =at15.61,
creased significantly p < 0.001, η
the subsequent 2p = 0.350,
workloads power50= to
from 0.999).
150 WThein level of MF (for
each group in-
creased significantly
consecutive workload at comparisons;
the subsequent workloads
Tukey’s HSD from 50 top <150
test; MF: W inp <each
0.001, group
0.0002, (for
respec-
consecutive
tively). The workload
value of MF comparisons; Tukey’s
did not differ HSD test;
significantly at MF: p < workloads
higher 0.001, p < 0.0002, respec-
(HSD Tukey’s
tively).
test: 150The value
W vs. 200ofW,MF p =did not The
0.483). differ significantly
ANOVA at highera significant
demonstrated workloads group
(HSD effect
Tukey’sfor
test:
MF 150 W=vs.
(F(2,29) 200 pW,
13.92, p = 0.483).
< 0.001, η2p = The ANOVA
0.490, power demonstrated a significant
= 0.996). A significant groupwas
difference effect for
noted
Int. J. Environ. Res. Public Health 2022, 19, 8839 7 of 13

increased significantly at the subsequent workloads from 50 to 150 W in each group

(for consecutive workload comparisons; Tukey’s HSD test; MF: p < 0.001, p < 0.0002,
respectively). The value of MF did not differ significantly at higher workloads (HSD
Tukey’s test: 150 W vs. 200 W, p = 0.483). The ANOVA demonstrated a significant group
effect for MF (F(2,29) = 13.92, p < 0.001, η2 p = 0.490, power = 0.996). A significant difference
was noted between the END and STR groups (Tukey’s HSD test: p < 0.001). However, no
significant difference was found between the END and CG groups (p = 0.505) (Figure 1).

3.2. Mean Power Frequency (MPF)

A significant group × workload interaction effect was observed in the SEMG variable,
MPF (F(6,87) = 8.98, p < 0.001, η2 p = 0.243, power = 0.984). The two-way ANOVA demon-
strated a main effect for group (F(2,29) = 4.65, p < 0.001, η2 p = 0.382, power = 0.958) showing
that significant differences in MPF across all workloads were found between the END and
STR groups (Tukey’s HSD test: END vs. STR, p = 0.0054) as well as the STR and CG groups
(Tukey’s HSD test: STR vs. CG, p = 0.0022). Conversely, no significant main effect for
workload was observed (p = 0.441) (Figure 2).

3.3. Electromyogram Amplitude (AEMG)

The significant group × workload interaction effect was observed in AEMG
(F(6,87) = 11.75, p < 0.001, η2 p = 0.448, power = 0.999). The ANOVA using AEMG showed
in a significant main effect for group (F(2,29) = 10.65, p < 0.001, η2 p = 0.423, power = 0.981).
Furthermore, a main effect for workload was also observed (F(3,87) = 173.02, p < 0.001,
η2 p = 0.856, power = 1). In general, AEMG tended to increase at subsequent workloads
in every group. Post hoc analyses demonstrated that the athletes from the END group
had significantly lower AEMG responses than those from the STR group (p = 0.0093) and
CG group (p = 0.0006). Interestingly, at the highest workload (200 W) there was no sig-
nificant difference between the END and STR groups (p = 0.805), while AEMG increased
substantially in the CG group and was significantly different compared to the END and
STR groups (both p< 0.001) (Figure 3).

3.4. Ratings of Perceived Exertion (RPE)

Significant main effects for RPE were demonstrated for both group (F(2,29) = 116.84,
p < 0.001, η2 p = 0.890, power = 1) and for workload (F(3,87) = 116.84, p < 0.001,
η2 p = 0.887, power = 1). There were also significant interactions between groups
(F(6,87) = 15.32, p < 0.001, η2 p = 0.514, power = 1), indicating that the successively in-
creasing RPE points from 50 to 200 W were significantly higher in the STR than in the END
group (Figure 4). As expected, participants from the CG group reported that each workload
completed was more exhaustive as compared to the groups of athletes (all p < 0.001).

4. Discussion
A number of differences were observed between the END and STR groups with
regards to their neuromuscular responses as demonstrated by SEMG and RPE. Amongst the
significant differences found were a larger increase in MF and MPF for the RF portion of the
quadriceps of the END athletes (triathletes) compared with the STR athletes (bodybuilders).
In contrast, a more pronounced increase in AMG and RPE were observed in the STR group
as compared to the END group. However, MF, MPF and RPE responses were the highest in
the CG group (recreationally active participants).
According to our assumptions, the neuromuscular responses were significantly differ-
ent between the END and STR groups, indicating that both central (neural) and peripheral
(muscular) function is associated with the training background. However, various neuro-
muscular factors, including altered patterns in neural recruitment [27] and the intrinsic
and extrinsic influences (e.g., thickness of subcutaneous tissue and/or distribution of MU
conduction velocities) on the EMG signal could affect the NMF profiles of the study partici-
pants [28,29]. In comparison to isometric muscle contraction, diverse biochemical processes
Int. J. Environ. Res. Public Health 2022, 19, 8839 8 of 13

(e.g., depletion of high-energy phosphates including creatine phosphate, accumulation of

lactate, hydrogen ion [H+ ] and inorganic phosphate) associated with the muscle fatigue
changes cause the interpretation of SEMG signals from dynamic contractions to be much
more difficult [30]. Since muscle activity has been shown to be repeatable during incre-
mental cycling exercise [31] and we did not use any method of normalization, it could be
assumed that the changes found in muscle activity during each cycling workload were
due to the variation in muscle activity caused by fatigue or from the intrinsic and extrinsic
influences on the EMG signal.
While the same muscle (e.g., RF) of different types of athletes may exhibit various
percentages of MUs, these muscles will probably show a different SEMG activity during
progressive cycling exercise. Indeed, the tested groups were significantly different in all
SEMG variables as well as RPE. Furthermore, endurance- and strength-trained athletes
present a heterogeneous make-up of different fiber types (muscle fiber phenotypic profiles)
with task-specific function [32]. It is well recognized that highly-trained endurance athletes
possess significantly higher slow-twitch muscle fiber (type I, oxidative) percentages (as
high as 90–95%) relative to resistance-trained (60–80%) or untrained individuals (approx.
50%) [33,34]. Since endurance training increases the amount of type I fibers [35], the
differences in muscle fiber type distribution between the subjects may be potential factors
that influence SEMG responses. In addition, in terms of morphological structure, different
muscles within a quadriceps represent diversified muscle fiber phenotypic profiles; the
proportion of fast-twitch muscle fibers (type II, glycolytic) is highest in RF (70.5%) compared
to vastus lateralis (VL) and vastus medialis (VM) (67.3% and 56.3%, respectively) [36].
Nevertheless, an important practical aspect of this research is that RF muscles appear to be
most sensitive to fatigue [37].
The comparative analysis of changes in lower extremity muscle activities during cy-
cling in laboratory settings is limited because of the variable experimental study designs,
participant training characteristics and/or specific muscles or muscle groups. The majority
of studies recorded the activity patterns in VL, which proved to be highly reproducible,
as well as it being considered the most active quadriceps component during cycling ex-
ercise [38–40]. However, RF (a biarticular muscle) and other muscle groups within the
quadriceps i.e., VL and VM (monoarticular muscles), and medial and lateral gastrocnemius
(MG and LG), are moderately sensitive in tracking changes in workload when a cycling
mode is applied [41]. Although during the incremental cycling exercise the integrated
EMG (iEMG) signal was consistently lower in RF compared with VL and VM muscles as
presented in Chin et al.’s [42] study, RF activity does track changes in workload to the same
extent as other muscles do (i.e., VL, MG and LG) [41]. Indeed, Baum and Li [43] proved
that lower extremity muscle activities during cycling are influenced by workload and also
by frequency. In their study, a workload effect was observed from the onset of RF and
the offset of RF muscles. Mileva et al. [44], reported that RF muscles demonstrated the
largest fatigue index change across the 15 repetitions of bilateral knee extension exercise
(80 ± 50% vs. 283 ± 89%; endurance vs. strength group). The average decrease in the
median EMG frequency values for the strength and endurance groups was significantly
different (−21 ± 5% and −9 ± 7%, respectively). A recent study conducted on power and
endurance athletes showed that a significant difference (p < 0.05) was found in the RF
muscle of the dominant leg for several EMG features [45].
From a traditional point of view, muscle fatigue can be identified by a decrease in the
frequency components of the EMG signal; for instance, detected from the MF of the SEMG
power spectrum [46]. We know that there is an association between fiber type distribution
and muscle fiber conduction velocity (MFCV) and/or spectral parameters such as MF and
MPF of type II fibers [47]. Moreover, it has been shown that slowdown of MFCV leads to a
decrease in MPF [48]. Therefore, it is likely that during cycling exercise the fatigue-induced
shift to lower values of MPF are indicative of the differences in the MFCV characteristics
(i.e., fiber type) of recruited muscle fibers [49]. In this study, during the progressive cycling
exercise, we observed (in all groups) a shift towards lower EMG frequencies (i.e., a low-
Int. J. Environ. Res. Public Health 2022, 19, 8839 9 of 13

frequency fatigue) together with an increased AEMG. This relationship is considered to be

an indicator of NMF [50]. Other studies have confirmed this mechanism, based on a higher
AEMG signal toward the end of the exercise [51]. We observed that the rate of increase in
AEMG during the submaximal cycling exercise was the highest in the CG group, whereas
the STR group achieved significantly higher values of this parameter in relation to the
END group. A significant reduction in MF in the END and CG groups (between 150 and
200 W) suggest that this SEMG parameter may be considered as a suitable indicator of
muscle fatigue, which is in line with multiple previous studies [44,52,53]. Interestingly,
such a change did not occur in the STR group.
In the present study, MPF decreased significantly at the peak workload (200 W) only in
the CG group, while there was no change and even an increase between 150 and 200 W in
both the STR group and the END group. These outcomes suggest decreased muscle activity
in the CG group and no remarkable signs of NMF in either groups of athletes. When we
focus on the groups of athletes, mean MF, but not the MPF variable, was significantly higher
in the course of the cycling exercise in the END compared to the STR group (p = 0.038).
However, the values of both were mostly similar in all groups during the last workload
(200 W). Trends in these spectral parameters observed in the END and STR groups are
not consistent with previous observations as higher values of the EMG power spectrum
parameters (MF and MPF) have been observed for muscles with a greater percentage
of type II fibers or greater relative area of type II fibers [54]. On this basis we should
expect higher values of MF and MPF in strength-trained athletes, but these were higher in
endurance-trained athletes (except MPF at 150 W). During submaximal, fatiguing exercise,
these responses can be attributed to fatigue-induced increases in muscle activation and MU
recruitment. It is probable that power and strength training improves MU recruitment and
synchronization more than endurance training. Moreover, the greater NMF in the END
group than in STR group could represent a specific defense mechanism to prevent any
extensive peripheral fatigue as has been recently exposed in well-trained male cyclists [52].
RPE is considered a factor of subjective fatigue, also termed a psychological fatigue,
and is highly correlated with physiological responses, muscle activity and workloads
during both continuous incrementaland constant cycling [55,56]. Previous studies have
shown a corresponding increase in the RPE and SEMG variables during cycling exercise
using different protocols [11,57]. For example, the EMG and RPE responses have been
shown to increase linearly until exhaustion during high-intensity cycling exercise, although
a constant-load exercise protocol was used [58]. A similar observation was also present in
this study when RPE and SEMG values gradually increased in accordance with the increase
in exercise workload in each of the study groups. The results are clear, showing the highest
RPE at exercise termination (200 W) in non-athletes (CG group), where peripheral fatigue
was also the highest. However, significantly lower RPE was reported in the END group
compared to other groups. If we refer to the previous studies of aerobic exercise anchored
by RPE [59,60], we can conclude that the neuromuscular responses during submaximal
exercises expressed by RPE were reflective of the fatigue-related physiological mechanisms.
This mechanism corroborates our findings, and therefore RPE can be regarded as a useful
indirect measure of muscle activity during progressive cycling exercise.

Limitations
The main limitation of this study is the lack of normalizing SEMG measurement, which
is the most appropriate for examining muscle activity during cycling for one-off measure-
ments (i.e., sprint method). In the SEMG measurement we also did not include the upper
limbs, which is a limitation because these muscle groups play a key role during balancing
on the cycle ergometer. Furthermore, athletes from the END group were triathletes, so
they routinely performed cycle exercise training. Hence, the type of exercise test was more
specific for them, while for the athletes from the STR group it was the non-familiar mode.
The study also involved young men, so the ability to generalize the results/conclusions
to individuals of other genders and ages is limited. Lastly, it could be possible that the
Int. J. Environ. Res. Public Health 2022, 19, 8839 10 of 13

duration of the cycling test, according to our protocol (not to exhaustion), may not fit the
time needed to obtain in-depth analysis for exercise-induced neuromuscular fatigue.

5. Conclusions
In conclusion, this study demonstrates that there is a significant variation in neuromus-
cular fatigue profiles between athletes with different training backgrounds when a cycling
exercise is applied. The study confirmed a significant shift of the frequency spectrum of
SEMG towards a lower frequency, together with the rising AEMG, when a participant is
fatigued during progressive cycling exercise. Confirmation of this are the changes in MPF
and MF between workloads of 150 and 200 W, especially in the END group, which suggest
the onset of exercise-induced fatigue. The observed differences might be attributed to the
fiber type specific characteristics of endurance- and strength-trained athletes. Probably due
to the greater percentage of type II fibers and/or greater relative area of type II fibers in
strength-trained athletes, they performed exercise cycling with better motor unit recruit-
ment and synchronization compared to endurance-trained athletes. The effects comprised
higher values of MF and MPF in endurance-trained athletes (except MPF at 150 W).
Furthermore, the approximately linear trends of the SEMG and RPE values of both
groups of athletes (END and STR) with increasing workload support the increased skeletal
muscle recruitment with perceived exertion or fatiguing effect. Thus, RPE may be useful to
control intensity during progressive cycling exercise, providing an indirect measure of the
muscle activity. However, greater increases in most SEMG parameters for the RF muscles
of the quadriceps of endurance-trained athletes (triathlon) compared with strength-trained
athletes (bodybuilding) suggest that the neuromuscular profile depends on the participant’s
training background.

Author Contributions: Conceptualization, M.J. and P.G.; methodology, M.J. and M.B.; software, M.J.;
validation, M.J., M.B. and P.G.; formal analysis, M.J. and P.G.; investigation, M.J., M.B., P.G. and
Z.W.; resources, M.J. and P.G.; data curation, M.J. and M.B.; writing—original draft preparation, M.J.,
M.B. and Z.W.; writing—review and editing, M.J., M.B., P.G. and Z.W.; visualization, M.B. and Z.W.;
supervision, M.J., M.B. and P.G.; project administration, M.J. and P.G.; funding acquisition, M.B. and
Z.W. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: The study was conducted in accordance with the Declaration
of Helsinki, and approved by the Ethics Committee of the Karol Marcinkowski University of Medical
Sciences in Poznań, Poland (decision no. 492/14).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
Data Availability Statement: Data are available upon request from the corresponding author.
Acknowledgments: The authors would like to thank the dedicated group of athletes who made this
study possible. We also thank the late Ryszard Grucza, who supervised the research process.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Kamali, A.M.; Saadi, Z.K.; Yahyavi, S.S.; Zarifkar, A.; Aligholi, H.; Nami, M. Transcranial direct current stimulation to enhance
athletic performance outcome in experienced bodybuilders. PLoS ONE 2019, 14, e0220363. [CrossRef]
2. Garrandes, F.; Colson, S.S.; Pensini, M.; Legros, P. Time course of mechanical and neuromuscular characteristics of cyclists and
triathletes during a fatiguing exercise. Int. J. Sports Med. 2007, 28, 148–156. [CrossRef] [PubMed]
3. Sleivert, G.G.; Backus, R.D.; Wenger, H.A. Neuromuscular differences between volleyball players, middle distance runners and
untrained controls. Int. J. Sports Med. 1995, 16, 390–398. [CrossRef] [PubMed]
4. Sanchez-Medina, L.; González-Badillo, J.J. Velocity loss as an indicator of neuromuscular fatigue during resistance training. Med.
Sci. Sports Exerc. 2011, 43, 1725–1734. [CrossRef] [PubMed]
5. Garrett, J.; Graham, S.R.; Eston, R.G.; Burgess, D.J.; Garrett, L.J.; Jakeman, J.; Norton, K. A novel method of assessment for
monitoring neuromuscular fatigue in Australian rules football players. Int. J. Sports Physiol. Perform. 2019, 14, 598–605. [CrossRef]
Int. J. Environ. Res. Public Health 2022, 19, 8839 11 of 13

6. Astokorki, A.H.; Mauger, A.R. Tolerance of exercise-induced pain at a fixed rating of perceived exertion predicts time trial cycling
performance. Scand. J. Med. Sci. Sports 2017, 27, 309–317. [CrossRef]
7. Glace, B.W.; Kremenic, I.J.; McHugh, M.P. Sex differences in central and peripheral mechanisms of fatigue in cyclists. Eur. J. Appl.
Physiol. 2013, 113, 1091–1098. [CrossRef]
8. Kremenic, I.J.; Glace, B.W.; Ben-Avi, S.S.; Nicholas, S.J.; McHugh, M.P. Central fatigue after cycling evaluated using peripheral
magnetic stimulation. Med. Sci. Sports Exerc. 2009, 41, 1461–1466. [CrossRef]
9. Medved, V.; Medved, S.; Kovač, I. Critical appraisal of surface electromyography (sEMG) as a taught subject and clinical tool in
medicine and kinesiology. Front. Neurol. 2020, 11, 560363. [CrossRef]
10. Graef, J.L.; Smith, A.E.; Kendall, K.L.; Walter, A.A.; Moon, J.R.; Lockwood, C.M.; Beck, T.W.; Cramer, J.T.; Stout, J.R. The rela-
tionships among endurance performance measures as estimated from VO2PEAK, ventilatory threshold, and electromyographic
fatigue threshold: A relationship design. Dyn. Med. 2008, 7, 15. [CrossRef]
11. Fontes, E.B.; Smirmaul, B.P.; Nakamura, F.Y.; Pereira, G.; Okano, A.H.; Altimari, L.R.; Dantas, J.L.; de Moraes, A.C. The relationship
between rating of perceived exertion and muscle activity during exhaustive constant-load cycling. Int. J. Sports Med. 2010, 31,
683–688. [CrossRef] [PubMed]
12. Keller, J.L.; Housh, T.J.; Hill, E.C.; Smith, C.M.; Schmidt, R.J.; Johnson, G.O. Self-regulated force and neuromuscular responses
during fatiguing isometric leg extensions anchored to a rating of perceived exertion. Appl. Psychophysiol. Biofeedback 2019, 44,
343–350. [CrossRef] [PubMed]
13. Gabriel, D.A.; Kamen, G.; Frost, G. Neural adaptations to resistive exercise. Sports Med. 2006, 36, 133–149. [CrossRef] [PubMed]
14. Srinivasan, J.; Balasubramanian, V. Low back pain and muscle fatigue due to road cycling—An sEMG study. J. Bodyw. Mov. Ther.
2007, 11, 260–266. [CrossRef]
15. Amann, M. Central and peripheral fatigue: Interaction during cycling exercise in humans. Med. Sci. Sports Exerc. 2011, 43,
2039–2045. [CrossRef]
16. Rossman, M.J.; Garten, R.S.; Venturelli, M.; Amann, M.; Richardson, R.S. The role of active muscle 684 mass in determining the
magnitude of peripheral fatigue during dynamic exercise. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2014, 306, R934–R940.
[CrossRef]
17. Thomas, K.; Elmeua, M.; Howatson, G.; Goodall, S. Intensity-dependent contribution of neuromuscular fatigue after constant-load
cycling. Med. Sci. Sports Exerc. 2016, 48, 1751–1760. [CrossRef]
18. Dufour, S.P.; Lampert, E.; Doutreleau, S.T.; Lonsdorfer-Wolf, E.; Billat, V.L.; Piquard, F.R.; Richard, R.U. Eccentric cycle exercise:
Training application of specific circulatory adjustments. Med. Sci. Sports Exerc. 2004, 36, 1900–1906. [CrossRef]
19. Borg, G. Ratings of perceived exertion and heart rates during short-term cycle exercise and their use in a new cycling strength
test. Int. J. Sports Med. 1982, 3, 153–158. [CrossRef]
20. Hebbal, G.V.; Mysorekar, V.R. Evaluation of some tasks used for specifying handedness and footedness. Percept. Mot. Skills 2006,
102, 163–164. [CrossRef]
21. Hermens, H.J.; Freriks, B.; Disselhorst-Klug, C.; Rau, G. Development of recommendations for SEMG sensors and sensor
placement procedures. J. Electromyogr. Kinesiol. 2000, 10, 361–374. [CrossRef]
22. Statistica (Data Analysis Software System), version 13; TIBCO Software Inc.: Palo Alto, CA, USA, 2017. Available online: http:
//statistica.io (accessed on 7 August 2021).
23. IBM. IBM Manuals; IBM ®SPSS ®Statistics Algorithms; IBM Corporation: Armonk, New York, USA, 2020.
24. Mauchly, J.W. Significance test for sphericity of a normal n-variate distribution. Ann. Math. Statist. 1940, 11, 204–209. [CrossRef]
25. Sokal, R.R.; Rohlf, E.J. Biometry, 3rd ed.; W.H. Freeman and Company: San Francisco, CA, USA, 1995.
26. Cohen, J. Statistical Power Analysis for the Behavioural Sciences, 2nd ed.; Academic Press: New York, NY, USA, 1988.
27. Gergley, J.C. Comparison of two lower-body modes of endurance training on lower-body strength development while concurrently
training. J. Strength Cond. Res. 2009, 23, 979–987. [CrossRef]
28. Nelson, N.L.; Churilla, J.R. A narrative review of exercise-associated muscle cramps: Factors that contribute to neuromuscular
fatigue and management implications. Muscle Nerve 2016, 54, 177–185. [CrossRef] [PubMed]
29. Twomey, R.; Aboodarda, S.J.; Kruger, R.; Culos-Reed, S.N.; Temesi, J.; Millet, G.Y. Neuromuscular fatigue during exercise:
Methodological considerations, etiology and potential role in chronic fatigue. Neurophysiol. Clin. 2017, 47, 95–110. [CrossRef]
30. Besomi, M.; Hodges, P.W.; Clancy, E.A.; Van Dieën, J.; Hug, F.; Lowery, M.; Merletti, R.; Søgaard, K.; Wrigley, T.; Besier, T.;
et al. Consensus for experimental design in electromyography (CEDE) project: Amplitude normalization matrix. J. Electromyogr.
Kinesiol. 2020, 53, 102438. [CrossRef]
31. Laplaud, D.; Hug, F.; Grélot, L. Reproducibility of eight lower limb muscles activity level in the course of an incremental pedaling
exercise. J. Electromyogr. Kinesiol. 2006, 16, 158–166. [CrossRef]
32. Plotkin, D.L.; Roberts, M.D.; Haun, C.T.; Schoenfeld, B.J. Muscle fiber type transitions with exercise training: Shifting perspectives.
Sports 2021, 9, 127. [CrossRef]
33. Bergh, U.; Thorstensson, A.; Sjödin, B.E.; Hulten, B.O.; Piehl, K.A.; Karlsson, J. Maximal oxygen uptake and muscle fiber types in
trained and untrained humans. Med. Sci. Sport 1978, 10, 151–154.
34. Fry, A.C.; Schilling, B.K.; Staron, R.S.; Hagerman, F.C.; Hikida, R.S.; Thrush, J.T. Muscle fiber characteristics and performance
correlates of male Olympic-style weightlifters. J. Strength Cond. Res. 2003, 17, 746–754.
Int. J. Environ. Res. Public Health 2022, 19, 8839 12 of 13

35. Harber, M.P.; Konopka, A.R.; Undem, M.K.; Hinkley, J.M.; Minchev, K.; Kaminsky, L.A.; Trappe, T.A.; Trappe, S. Aerobic exercise
training induces skeletal muscle hypertrophy and age-dependent adaptations in myofiber function in young and older men.
J. Appl. Physiol. 2012, 113, 1495–1504. [CrossRef] [PubMed]
36. Johnson, M.; Polgar, J.; Weightman, D.; Appleton, D. Data on the distribution of fibre types in thirty-six human muscles: An
autopsy study. J. Neurol. Sci. 1973, 18, 111–129. [CrossRef]
37. Pincivero, D.M.; Gandhi, V.; Timmons, M.K.; Coelho, A.J. Quadriceps femoris electromyogram during concentric, isometric and
eccentric phases of fatiguing dynamic knee extensions. J. Biomech. 2006, 39, 246–254. [CrossRef]
38. Bercier, S.; Halin, R.; Ravier, P.; Kahn, J.F.; Jouanin, J.C.; Lecoq, A.M.; Buttelli, O. The vastus lateralis neuromuscular activity
during all-out cycling exercise. J. Electromyogr. Kinesiol. 2009, 19, 922–930. [CrossRef]
39. Hug, F.; Decherchi, P.; Marqueste, T.; Jammes, Y. EMG versus oxygen uptake during cycling exercise in trained and untrained
subjects. J. Electromyogr. Kinesiol. 2004, 14, 187–195. [CrossRef]
40. Singh, V.P.; Kumar, D.K.; Djuwari, D.; Polus, B.; Fraser, S.; Hawley, J.; Giudice, S.L. Strategies to identify muscle fatigue from
SEMG during cycling. In Proceedings of the 2004 Intelligent Sensors, Sensor Networks and Information Processing Conference,
Melbourne, Australia, 14–17 December 2004; pp. 547–551.
41. Albertus-Kajee, Y.; Tucker, R.; Derman, W.; Lambert, M. Alternative methods of normalising EMG during cycling. J. Electromyogr.
Kinesiol. 2010, 20, 1036–1043. [CrossRef]
42. Chin, L.M.; Kowalchuk, J.M.; Barstow, T.J.; Kondo, N.; Amano, T.; Shiojiri, T.; Koga, S. The relationship between muscle
deoxygenation and activation in different muscles of the quadriceps during cycle ramp exercise. J. Appl. Physiol. 2011, 111,
1259–1265. [CrossRef]
43. Baum, B.S.; Li, L. Lower extremity muscle activities during cycling are influenced by load and frequency. J. Electromyogr. Kinesiol.
2003, 13, 181–190. [CrossRef]
44. Mileva, K.N.; Morgan, J.; Bowtell, J. Differentiation of power and endurance athletes based on their muscle fatigability assessed
by new spectral electromyographic indices. J. Sports Sci. 2009, 27, 611–623. [CrossRef]
45. Sulaiman, M.; Azaman, A.; Salleh, N.A.; Yahya, A. Lower Limb Muscles Activation of Endurance and Power Athletes. In 11th
Asian-Pacific Conference on Medical and Biological Engineering. IFMBE Proceedings; Shiraishi, Y., Sakuma, I., Naruse, K., Ueno, A.,
Eds.; Springer: Cham, Germany, 2021; Volume 82, pp. 111–117.
46. Chang, K.M.; Liu, S.H.; Wu, X.H. A wireless sEMG recording system and its application to muscle fatigue detection. Sensors 2012,
12, 489–499. [CrossRef]
47. Nasrabadi, R.; Izanloo, Z.; Sharifnezad, A.; Hamedinia, M.R.; Hedayatpour, N. Muscle fiber conduction velocity of the vastus
medilais and lateralis muscle after eccentric exercise induced-muscle damage. J. Electromyogr. Kinesiol. 2018, 43, 118–126.
[CrossRef] [PubMed]
48. Calder, K.M.; Stashuk, D.W.; McLean, L. Physiological characteristics of motor units in the brachioradialis muscle across fatiguing
low-level isometric contractions. J. Electromyogr. Kinesiol. 2008, 18, 2–15. [CrossRef] [PubMed]
49. Matsuura, R.; Ogata, H.; Yunoki, T.; Arimitsu, T.; Yano, T. Effect of blood lactate concentration and the level of oxygen uptake
immediately before a cycling sprint on neuromuscular activation during repeated cycling sprints. J. Physiol. Anthropol. 2006, 25,
267–273. [CrossRef] [PubMed]
50. Liu, X.; Li, Z. Influence mechanism of running sportswear fatigue based on BP neural network. Eurasip. J. Adv. Signal Process.
2021, 61, 1–5. [CrossRef]
51. Luomala, M.J.; Oksa, J.; Salmi, J.A.; Linnamo, V.; Holmér, I.; Smolander, J.; Dugué, B. Adding a cooling vest during cycling
improves performance in warm and humid conditions. J. Therm. Biol. 2012, 37, 47–55. [CrossRef]
52. García-García, O.; Cuba-Dorado, A.; Riveiro-Bozada, A.; Carballo-López, J.; Álvarez-Yates, T.; López-Chicharro, J. A maximal
incremental test in cyclists causes greater peripheral fatigue in biceps femoris. Res. Quar. Exerc. Sport 2020, 91, 460–468. [CrossRef]
53. Wang, L.; Wang, Y.; Ma, A.; Ma, G.; Ye, Y.; Li, R.; Lu, T. A comparative study of EMG indices in muscle fatigue evaluation based
on grey relational analysis during all-out cycling exercise. Biomed. Res. Int. 2018, 2018, 9341215. [CrossRef]
54. Gerdle, B.; Karlsson, S.; Crenshaw, A.G.; Elert, J.; Fridén, J. The influences of muscle fibre proportions and areas upon EMG
during maximal dynamic knee extensions. Eur. J. Appl. Physiol. 2000, 81, 2–10. [CrossRef]
55. Balagué, N.; Hristovski, R.; García, S.; Aguirre, C.; Vázquez, P.; Razon, S.; Tenenbaum, G. Dynamics of perceived exertion in
constant-power cycling: Time-and workload-dependent thresholds. Res. Quar. Exerc. Sport 2015, 86, 371–378. [CrossRef]
56. Green, J.M.; McIntosh, J.R.; Hornsby, J.; Timme, L.; Gover, L.; Mayes, J.L. Effect of exercise duration on session RPE at an
individualized constant workload. Eur. J. Appl. Physiol. 2009, 107, 501–507. [CrossRef]
57. Macdonald, J.H.; Farina, D.; Marcora, S.M. Response of electromyographic variables during incremental and fatiguing cycling.
Med. Sci. Sports Exerc. 2008, 40, 335–344. [CrossRef] [PubMed]
58. Moritani, T.O.; Takaishi, T.E.; Matsumoto, T.A. Determination of maximal power output at neuromuscular fatigue threshold. J.
Appl. Physiol. 1993, 74, 1729–1734. [CrossRef] [PubMed]
Int. J. Environ. Res. Public Health 2022, 19, 8839 13 of 13

59. Cochrane-Snyman, K.C.; Housh, T.J.; Smith, C.M.; Hill, E.C.; Jenkins, N.D.; Schmidt, R.J.; Johnson, G.O. Inter-individual variability
in the patterns of responses for electromyography and mechanomyography during cycle ergometry using an RPE-clamp model.
Eur. J. Appl. Physiol. 2016, 116, 1639–1649. [CrossRef]
60. Lander, P.J.; Butterly, R.J.; Edwards, A.M. Self-paced exercise is less physically challenging than enforced constant pace exercise of
the same intensity: Influence of complex central metabolic control. Br. J. Sports Med. 2009, 43, 789–795. [CrossRef] [PubMed]