Motor unit firing frequency of lower limb muscles during an incremental slide board
skating test
*Tatiane Piucco1,2, Rodrigo Bini3, Masanori Sakaguchi1, Fernando Diefenthaeler1,2, Darren
Stefanyshyn1

1- Roger Jackson Centre for Health and Wellness Research, Faculty of Kinesiology, University
of Calgary, Calgary, AB, Canada.
2- Physical Education Department, Sports Centre, Federal University of Santa Catarina,
Florianópolis, SC, Brazil.
3- Centre of Physical Training of the Army, School of Physical Education of the Army, Rio de
Janeiro, Brazil

*Corresponding author
University of Calgary, Calgary, Canada
Faculty of Kinesiology – Human Performance Laboratory
University of Calgary, 2500 University Drive NW
Calgary, AB, Canada T2N 1N4
Phone: +1 (403) 220-6472
Fax: 403-284-3553
Email: tpiucco@mtroyal.ca

The authors acknowledge the funding support of CAPES agency Brazil, and the athletes to be
willing to participate in this study.
1

2

Abstract
This study investigated how the combination of workload and fatigue affected the frequency
components of muscle activation and possible recruitment priority of motor units during skating
to exhaustion. Ten male competitive speed skaters performed an incremental maximal test on
a slide board. Activation of six muscles from the right leg was recorded throughout the test. A
time-frequency analysis was performed to compute overall, high, and low frequency bands
from the whole signal at 10, 40, 70, and 90% of total test time. Overall activation increased for
all muscles throughout the test (p < 0.05 and ES > 0.80). There was an increase in low frequency
(90% vs 10%, p = 0.035, ES = 1.06) and a decrease in high frequency (90% vs 10%, p = 0.009,
ES = 1.38, and 90% vs 40%, p = 0.025, ES = 1.12) components of gluteus maximus. Strong
correlations were found between the maximal cadence and vastus lateralis, gluteus maximus
and gluteus medius activation at the end of the test. In conclusion, the incremental skating test
lead to an increase in activation of lower limb muscles, but only gluteus maximus was sensitive
to changes in frequency components, probably caused by a pronounced fatigue.
Word count: 199.

Keywords: fitness test, electromyography, fatigue, frequency band analysis.

Introduction

Studies have described the intermuscular coordination patterns of speed skaters of
different levels of performance (De Koning, De Groot, & Van Ingen Schenau, 1991) and
analysed muscle activation patterns at different skating velocities (Chang, Turcotte, & Pearsall,
2009). However, the spectral properties of surface electromyography (EMG) signals during
skating have been poorly investigated. Only one study (Buckeridge, LeVangie, Stetter, Nigg,
3

& Nigg, 2015) investigated the principal muscle activation patterns during hockey skating, in
order to assess differences between elite and recreational players, but no results about the
recruitment priority of motor units during skating are available.
It is thought that an increase in muscle activation represents the recruitment of additional
motor units and/or the increase in firing rate of the active motor units to compensate for a
decrease in contractility of impaired or fatigued motor units for a given force production
(Edwards & Lippold, 1956). It has been hypothesised that a reduction in the percentage of high
frequency motor unit recruitment results in the shift of mean power frequency of surface
electromyograms to lower frequencies during fatigue (De Luca, 1997).
Assessment of muscle activation during motion presents some challenges when
processing the resulting EMG signals. The most important challenge results from the nonstationarity profile induced in the signal during dynamic contraction, because of the movement
of muscle fibres in relation to the recording electrodes, changes in fibre length, and alterations
in the number of active motor units and their firing rates (Bonato, Gagliati, & Knaflitz, 1996).
Therefore, the fast Fourier transform (FFT) and other traditional signal processing methods may
not be appropriate for the analysis of myoelectrical signals during dynamic contractions.
During incremental fatiguing tests, surface EMG of working muscle is thought to be
influenced by a combination of increased workload and muscle fatigue. During graded exercise
tests, some muscles present a threshold in activation profile (Hug, Faucher, Kipson, & Jammes,
2003), and the changes in priority in motor unit recruitment could be affected by the mechanical
function of the muscles and workload level (Hodson-Tole & Wakeling, 2009; Priego Quesada,
Bini, Diefenthaeler, & Carpes, 2015). It has been suggested that during incremental exercises
an increase in higher frequency components can occur because of recruitment of faster motor
units (Van Boxtel & Schomaker, 1984) and this might mask a decrease of median frequency
components. However, combined effects from workload and fatigue, during incremental
4

cycling protocol to exhaustion, lead to an increase in low frequency content for the biceps
femoris, without alteration in high frequency components for this muscle (Priego Quesada et
al., 2015).
The validation of fitness tests specific for speed skaters are scarce, particularly because
skating is challenging to simulate in the laboratory (Foster, Thompson, & Synder, 1993).
Therefore, incremental load cycling tests are frequently used to evaluate speed skaters, but
several studies have shown that substantial biomechanical and physiological differences exist
between cycling and skating activities (Foster, Rundell, & Snyder, 1999; Rundell, 1996;
Snyder, O'Hagan, Clifford, Hoffman, & Foster, 1993). Skating on a slide board elicits similar
physiological and biomechanical responses compared to real skating (Kandou et al., 1987), and
it could be an alternative laboratory-based method to evaluate performance of speed skaters
(Piucco, O’Connell, Stefanyshyn, & De Lucas, 2016).
During skating, a combination of the crouched position, the relatively long isometric
gliding phase and high intramuscular forces results in reduced blood flow to the working
muscles (Foster et al., 1999; Rundell, 1996). The deoxygenation of the working muscles may
lead to fatigue, reflected by the relatively high blood lactate concentrations of skaters compared
with those in other sports (Beneke & von Duvillard, 1996; Foster et al., 1999). Subsequently, a
decrease in recruitment of the fast-twitch fibres may occur, since the decrease in intracellular
pH leads to a decrease in muscle fibre conduction velocity, and as a consequence, a decrease of
median frequency (Basmajian & De Luca, 1985; Brody, Pollock, Roy, De Luca, & Celli, 1991).
However, to date, no study assessed if these changes in metabolic state of lower limb muscles
in speed skaters could be translated into changes in EMG spectral properties (e.g. reduced high
frequency content from reduced recruitment of fast-twitch fibres).
This study aimed to investigate how combined effects of increased workload and fatigue
state affected the activation of lower limb muscles and potential recruitment priority of motor
5

units, during a maximal incremental skating protocol performed on a slide board. We
hypothesised that the activation of the main lower limb muscles, responsible for the skating
movement pattern, would increase with workload/fatigue due to greater contribution from low
(but not high) frequency components in analysed muscles.

Methods

Participants
Ten male competitive long track ice speed skaters (event distances between 500 m and
5000 m) participated in this study. The skaters’ mean and standard deviation values for age,
body mass, percentage of body fat and height were respectively 18.9 ± 5.3 years, 72.1 ± 7.8 kg,
12.8 ± 1.5 %, and 1.78 ± 4.6 m. All skaters were participating in a systematic training program
with a volume of eight hours per week, for at least two years. The participants signed an
informed consent and all testing complied with University of Calgary Institutional Ethics
Committee regulations regarding the use of human participants. Skaters were tested at the end
of the off-season, while participating in a structured training program. All participants had
experience with slide board skating and with maximal incremental effort tests.

Procedures
The participants were instructed to refrain from heavy intensity training, to maintain a
regular diet, and to abstain from the ingestion of any stimulant (caffeine drink, nicotine) or
alcohol 24h prior to testing. Each participant performed a maximal incremental skating test to
exhaustion on a slide board (2.0 m × 0.6 m × 0.025 m, kinetic friction coefficient of 0.13 ±
0.05), which was instrumented with optical sensors placed at the slide board extremities to
determine the skaters’ cadence during the test (Figure 1).
6

FIGURE 1 AROUND HERE

The slide board skating protocol started at a cadence of 30 push-offs per minute (ppm),
with progressive increments of 3 ppm every minute until voluntary exhaustion, or until they
were no longer capable of maintaining the cadence of the corresponding stage (within a range
of 10% per push-off, controlled by the software). The participants were asked to maintain a
constant skating posture, with free movement of their arms during the test. The reliability of
this protocol was previously determined (ICC > 0.9 and typical error of measure < 3.5%)
(Piucco, dos Santos, de Lucas, & Dias, 2015).

Data acquisition
Gas exchanges were measured breath-by-breath during the test using a gas analyser
(K4b2 Cosmed®, Rome, Italy), calibrated according to the manufacturer’s instructions prior to
each test. Peak oxygen uptake (VO2peak) was considered to be the highest value averaged over
a 15-s period. Blood samples were collected from participants’ fingertip at minute one, three,
and five following the tests conclusion, to assess the maximal blood lactate concentration
([Lac]max). [Lac] were assessed using a Lactate Pro® analyser (Arkay Inc., New South Wales,
Australia), calibrated according to the manufacturer’s recommendations before each analysis.
Maximal cadence (CADmax) was defined as the maximal number of push-offs per minute
reached during the slide board tests. If the final stage was not completed, the CADmax was
calculated according to the following equation adapted from Kuipers, Verstappen, Keizer,
Geurten, and van Kranenburg (1985):
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝐶𝐶𝐶𝐶𝐶𝐶𝑓𝑓 +

𝑡𝑡

𝑠𝑠

∗3
7

(1)

where CADf is the cadence of the final stage completed, t the uncompleted stage time, s
represents duration of the stage in seconds, and 3 the cadence increment per stage.
Muscle activation was monitored during the last 15-s of each stage of the incremental
test, using a Biovision EMG system (Biovision Inc., Wehrheim, Germany) and Ag-AgCl
bipolar surface electrodes (Myotronics-Noromed Inc., Tukwila, USA), placed on the skin over
the vastus lateralis (VL), vastus medialis (VM), biceps femoris (BF), gluteus maximus (GM),
gluteus medius (GMd), and adductor magnus (AM) of the right leg. After shaving and cleaning
the skin, the electrodes were placed over the belly of the muscles, parallel with the muscle fibre
orientation (Hermens, Freriks, Disselhorst-Klug, & Rau, 2000) and taped to the skin to
minimise movement artefact. A reference electrode was placed over the tibial tuberosity. EMG
electrodes were directly connected to a data acquisition unit (Biovision Inc., Wehrheim,
Germany) which was housed inside a portable backpack. Data were recorded at 2400 Hz per
channel with 12-bit resolution. A triaxial accelerometer (1000 Hz of sampling frequency and
+/- 200 g of measurement range, model ADXL377G, Analog Devices, Inc., Massachusetts,

USA) was fixed on the heel of the right shoe and connected to the EMG data acquisition unit,
in order to synchronise the EMG signals within the skating cycle by using the impact peak of
the foot on the lateral bumper of the slide board.

Data analysis

EMG signals were separated into three complete strokes and averaged, defined at each
impact on the right lateral bumper, registered by the accelerometer (Figure 2). 10% time
windows were created for analysis (10 to 20%, 40 to 50%, 70 to 80% and 90 to 100%) of each
skater. Each time window was filtered using a fifth order Butterworth digital filter at nine bandpass frequencies (Table 1), divided in high (146.95-300.80 Hz) and low (26.95-75.75 Hz)
8

frequency bands. Each EMG signal of the nine frequency bands was then converted into its root
mean square values (RMS), computed using moving average windows of 40 ms (Neptune,
Kautz, & Hull, 1997).

FIGURE 2 AROUND HERE

TABLE 1 AROUND HERE

The sum of the RMS signal of the nine averaged frequency bands was calculated to
define the overall RMS of each muscle (i.e. activation of all frequency bands of the EMG
signal). The fifth, sixth, and seventh bands were averaged to compute the RMS from high
frequency components of the signals, which would potentially represent the response of larger
motor units (Wakeling & Horn, 2009). The first and the second bands were averaged to compute
the RMS from low frequency components of the signals, which would potentially represent the
response of smaller motor units (Wakeling & Horn, 2009).
The overall RMS values at 40, 70 and 90% of the test time were normalised to the RMS
values at the start of the test (first 10% of the total time), when the skaters were considered not
to be fatigued, in order to reduce between-participant variability in EMG data (Quesada et al.,
2014). High and low frequency RMS values, at 40%, 70% and 90%, were then normalised by
the respective overall RMS, in order to provide a percentage of the overall signal from each
frequency content during the test. All signals were processed in MATLAB® (Mathworks Inc.,
Natick, Massachusetts, USA).

Statistical analysis
9

Overall RMS, as well as RMS of high and low frequency bands were averaged across
participants. Each instant of the total time of the test (10, 40, 70, and 90% of total time) was
compared for the overall activation, as well as high, and low frequency bands for each muscle
using Cohen’s effect size and Student’s t test for paired samples. High and low frequency bands
were compared between muscles at each instant of time, normalised by the overall RMS
activation. The comparison of the overall activation between muscles could not be made,
because the data were not normalised by the maximal voluntary contraction of each muscle.
Effect sizes (ES) of each pair of comparisons were categorized as small (ES < 0.20), moderate
(ES > 0.20 - 0.8) or large effects (ES > 0.8) (Cohen, 1988). The correlation coefficient between
overall RMS at each instant of time (10, 40, 70 and 90%) and CADmax obtained during the
incremental test was determined using Pearson’s correlation test. These coefficients were
ranked following methods from Dancey and Reidy (2004) [i.e. r = 1.0 indicates perfect
association, r between 0.9 and 0.69 indicates strong association, r between 0.4 and 0.69
indicates moderate association and r smaller than 0.39 indicates small to no association].
Statistical significance was defined at p < 0.05 and ES greater than 0.8. Statistical analysis was
conducted using Statistical Package for Social Sciences (SPSS Inc.v.17.0, Chicago, USA) and
Excel (Microsoft, USA).

Results

The mean and standard deviation values for CADmax, VO2peak, maximal heart rate
(HRmax) and [Lac]max obtained in the incremental slide board test were 59.5 ± 5.1 ppm, 46.4
± 3.6 ml/kg/min, 190 ± 9 bpm and 10.03 ± 1.92 mmol/l, respectively.
Overall RMS activation increased significantly throughout the test (% of time) for all
muscles analysed (p < 0.05 or ES > 0.80) (Figure 3).
10

FIGURE 3 AROUND HERE

The proportion of the low frequency components was larger compared to the high
frequency components for all muscles during all instants of the test (p<0.05, Figure 4). Low
frequency components were higher for VL and VM muscles, followed by GM and AM. High
frequency components were higher for GMd muscles, followed by VL and VM and BF muscles.
Low frequency components from GM activation were higher at 90% of the test (43.2 ± 3.8%
of overall activation) in comparison to 10% of the total time (39.8 ± 2.5%, p = 0.035 and ES =
1.06), whereas high frequency components from GM were lower at 90% (6.2 ± 1.3% of overall
activation) in comparison to 40% (8.0 ± 1.9% p = 0.025 and ES = 1.12) and 10% of the test (8.5
± 2.1% p = 0.009 and ES = 1.38) (Figure 4).

FIGURE 4 AROUND HERE

Significant correlations between overall activation and CADmax were found for VL at
70% (r = 0.87, p < 0.01) and 90% (r = 0.74, p = 0.015) of the total time, for GM at 70% (r =
0.86, p < 0.01) of the total time and for GMd at 70% of the total time (r = 0.74, p = 0.009).
Figure 5 illustrates the relationships between CADmax, VL, GM, and GMd overall activation,
at 70% and 90% of total time.
FIGURE 5 AROUND HERE

Discussion and Implications
11

This study examined the effects of workload and fatigue on frequency components of
muscle activation during an incremental skating protocol performed on a slide board. The main
findings were that overall activation progressively increased for all muscles analysed (Figure
3), and an increase in low and a decrease in high frequency components of GM were observed
(Figure 4). The increase in overall activation was expected to occur because of the combined
effect of fatigue and workload elicited during the incremental test. The increase in overall
muscle activation reflects the recruitment of additional motor units and an increase in motor
unit rate coding to compensate for the deficit in contractility resulting from impairment of
fatigued motor units, as required for increments in muscle force (MacDonald, Farina, &
Marcora, 2008).
Previous studies showed an increase (Kuznetsov, Popov, Borovik, & Vinogradova,
2011), no change (Jansen, Ament, Verkerke, & Hof, 1997; MacDonald et al., 2008) or even an
initial increase and further decrease (Gamet, Duchêne, & Goubel, 1996) in discharge rate
(median frequency) with fatigue during incremental cycling to exhaustion. However, during
skating, a low crouched posture is adopted in order to reduce air friction forces and obtain a
better push-off angle, and a long isometric contraction occurs in order to sustain the posture (de
Koning et al., 1991; Noordhof, Foster, Hoozemans, & de Koning, 2013). All these
characteristics are related to a restriction in muscle blood flow to the lower limbs and high
intramuscular forces during speed skating, which lead to higher muscle oxygen desaturation
and lower oxygen uptake during skating when compared to cycling (Foster et al., 1999; Rundell,
1996; Snyder et al., 1993). The deoxygenation of the working muscles may lead to a faster
fatigue process, reflected by the relatively high blood lactate concentrations of skaters
compared with those in other sports (Beneke & von Duvillard 1996; Foster et al., 1999).
Consequently, a decrease in recruitment of the fast-twitch fibres (i.e. decrease in high frequency
12

components) may occur, since the decrease in intracellular pH determines the decrease of
muscle fibre conduction velocity (Basmajian, & De Luca, 1985; Brody et al., 1991; Hummel et
al., 2005).
Despite the evidence suggesting that very rapid movements require preferential
recruitment of type II over type I fibres (Smith, Betts, Edgerton, & Zernicke, 1980; Van Boxtel
& Schomaker, 1984), the increase in high frequency components of muscle activation with
faster skating cadence, toward the end of the incremental test, was not observed in this study.
This finding may indicate that the increase in cadence was not sufficient to result in significantly
increased recruitment of type II fibres, over a larger decrease in intracellular pH from peripheral
fatigue.
During speed skating, hip power is mainly produced by hip extensor monoarticular
muscles, such as the GM, which remains active until 100 ms before the end of the stroke (de
Boer, Vos, Hutter, de Groot, & van Ingen Schenau, 1987). This contribution explains the high
overall activation level of the GM found in this study. The important contribution from the GM
muscle during skating can explain why this muscle was the only one to show significant
changes in frequency components. The changes in spectral properties are probably related to a
decrease in contractility of impaired or fatigued motor units. This could be related to a greater
proportion of type I fibre motor units reported (52.4%) compared to type II fibres (47.6%) for
the GM (Johnson, Polgar, Weightman, & Appleton, 1973). The decrease in high frequency
components of GM activation may reflect a decrease in fast-twitch fibre recruitment with
fatigue, while the increase in low frequency components reflects the contribution of type I fibres
as an attempt to improve force production, and sustain the required workload.
VM and VL were also highly recruited during the incremental skating test, consistent
with the fact that the knee extensor muscles are responsible to generate power across the knee
joint during skating. However, differently from GM, frequency components of VL and VM
13

muscles did not change during the test. VL and VM presented a greater contribution from low
frequency components, which could be related to a greater proportion of type I fibres (64.9%)
reported on the VL muscle of elite ice speed skaters (Ahmetov et al., 2011). Also, Green (1978)
found a preferential utilization of type I or slow-twitch muscle fibres before a progressive
depletion of type II fibres of the VL during skating. This characteristic allows the VL to endure
longer and to delay any changes in spectral properties, which could also explain the unchanged
low/high frequency components towards the end of the incremental test.
The greater contribution from high frequency components of the GMd muscle can be
related to the critical contribution of the abductors/adductors muscle activity and its
coordination to generate high skating speed and therefore skating performance (Chang et al.
2009). At the end of the incremental cadence test, the skaters reached high stroke frequency
(~70 ppm), and a positive relationship between GMd and CADmax was observed. VM and GM
were also correlated with CADmax, probably because these muscles are mainly responsible for
the push-off during skating.
Studies with cyclists suggested that antagonist muscle activation is reduced with fatigue,
to minimise co-contraction and improve pedalling effectiveness during cycling (Hautier et al.,
2000). This effect was not observed for BF and AM muscles, since both muscles showed a
significant increase in activation towards the end of the incremental test. However, the BF
muscle can also play a role in hip extension and knee flexion during skating. A significant
activation of the BF during the gliding phase (until approximately 200 ms before the end of
stroke) of skating occurs, to lock the knee joint and pre stretch the knee extensor muscles for a
shorter and more explosive extension of the knee joint during the push-off phase (de Boer et
al., 1987). The continuous increase of AM activity throughout the incremental test highlights
the important function of the adductors during skating. The adductor magnus muscle activity
and its coordination are related to the generation of higher skating speed and, therefore, skating
14

performance (Chang et al., 2009). Anatomically, adductor muscles (both adductor magnus and
adductor longus) appear to contribute to extension and flexion of the hip joint, respectively
(Pressel & Lengsfeld, 1998), which could also contribute to the skating push-off. Nonetheless,
the slight hip adduction that occurs during the beginning of the stride would lengthen the GM,
potentially increasing the passive contribution from tendons and aponeurosis to the following
force production.
This study had some limitations, mostly related to the use of surface EMG to capture
muscle activation signals. The spectral properties of the surface action potentials are influenced
by the distance between the active muscle fibres and the detection point. Moreover, as with
estimates of amplitude, spectral features are influenced by modifications in volume conductor
properties and the shift of electrodes with changes in joint angle, which result in changes during
movement being attributable to both geometrical and physiological effects (Farina, 2006).
Another limitation related to surface EMG is perspiration under the surface electrodes, which
can reduce the amplitude of the EMG signal (Abdoli-Eramaki et al (2012). Future studies
should investigate the neuromuscular fatigue mechanisms, together with local muscle
deoxygenation parameters, to better understand the neuromuscular and physiological
adaptations during skating.

Conclusion

In conclusion, the present study demonstrated that the combined effects of fatigue and
workload during an incremental slide board skating test to exhaustion led to an increased overall
muscle activation for all muscles. However, only gluteus maximus was sensitive to changes in
motor unit firing frequency, represented by an increase in low and a decrease in high frequency

15

components. The pattern of muscles activation investigated in this study was similar to those
described during ice or treadmill skating.

Acknowledgements
The authors acknowledge the funding support of CAPES agency Brazil, and the athletes to be
willing to participate in this study.

References
Abdoli-Eramaki, M., Damecour, C., Christenson, J., & Stevenson, J. (2012). The effect of
perspiration on the sEMG amplitude and power spectrum. Journal of Electromyography
and Kinesiology, 22, 908-913.
Ahmetov, I. I., Druzhevskaya, A. M., Lyubaeva, E. V., Popov, D. V., Vinogradova, O. L., &
Williams, A. G. (2011). The dependence of preferred competitive racing distance on
muscle fibre type composition and ACTN3 genotype in speed skaters. Experimental
Physiology, 96, 1302-1310. doi:10.1113/expphysiol.2011.060293
Basmajian, J.V., & De Luca, C. J. (1985). Muscles alive. Baltimore: Williams & Wilkins, 201222.
Beneke, R., & von Duvillard, S. P. (1996). Determination of maximal lactate steady state
response in selected sports events. Medicine and Science in Sports and Exercise, 28,
241-246.
Bonato, P., Gagliati, G., & Knaflitz, M. (1996). Analysis of myoelectric signals recorded during
dynamic contractions: A time-frequency approach to assessing muscle fatigue. IEEE
Engineering and Medicine in Biology, 15, 102-111.
Brody, L. R., Pollock, M. T., Roy, S. H., De Luca, C. J., & Celli, B. (1991). pH-induced effects
on median frequency and conduction velocity of the myoelectric signal. Journal of
16

Applied Physiology (1985), 71, 1878-1885.
Buckeridge, E., LeVangie, M. C., Stetter, B., Nigg, S. R., & Nigg, B. M. (2015). An on-ice
measurement approach to analyse the biomechanics of ice hockey skating. PLoS One,
10(5), e0127324. doi:10.1371/journal.pone.0127324
Chang, R., Turcotte, R., & Pearsall, D. (2009). Hip adductor muscle function in forward skating.
Sports Biomechanics, 8, 212-222. doi:10.1080/14763140903229534
Dancey, C. P., & Reidy, J. (2004). Statistics without maths for psychology: Using SPSS for
Windows TM. Pearson: Prentice Hall, 168-211.
de Boer, R. W., Vos, E., Hutter, W., de Groot, G., & van Ingen Schenau, G. J. (1987).
Physiological and biomechanical comparison of roller skating and speed skating on ice.
European Journal of Applied Physiology and Occupational Physiology, 56, 562-569.
de Koning, J. J., de Groot, G., & van Ingen Schenau, G. J. (1991). Coordination of leg muscles
during speed skating. Journal of Biomechanics, 24, 137-146.
De Luca, C. J. (1997). The use of surface electromyography in biomechanics. Journal of
Applied Biomechanics, 13, 135-163.
Diefenthaeler, F., Bini, R. R., & Vaz, M. A. (2012). Frequency band analysis of muscle
activation during cycling to exhaustion. Brazilian Journal of Kinanthropometry and
Human Performance, 14, 243-53.
Edwards, R. G., & Lippold, O. C. (1956). The relation between force and integrated electrical
activity in fatigued muscle. The Journal of Physiology, 132, 677-681.
Farina, D. (2006). Interpretation of the surface electromyogram in dynamic contractions.
Exercise and Sport Sciences Reviews, 34, 121-127.
Foster, C., Thompson, N. N., & Synder, A. C. (1993). Ergometric studies with speed skaters:
evolution of laboratory methods. Journal of Strength and Conditioning Research, 7,
193-200.
17

Foster, C., Rundell, K. W., Snyder, A. C., Stray-Gundersen, J., Kemkers, G., Thometz, N., . . .
Knapp, E. (1999). Evidence for restricted muscle blood flow during speed skating.
Medicine and Science in Sports and Exercise, 31, 1433-1440.
Gamet, D., Duchêne, J., & Goubel, F. (1996). Reproducibility of kinetics of electromyogram
spectrum parameters during dynamic exercise. European Journal of Applied Physiology
and Occupational Physiology, 74, 504-510.
Hautier, C. A., Arsac, L. M., Deghdegh, K., Souquet, J., Belli, A., & Lacour, J. R. (2000).
Influence of fatigue on EMG/force ratio and cocontraction in cycling. Medicine and
Science in Sports and Exercise, 32, 839-843.
Hermens, H. J., Freriks, B., Disselhorst-Klug, C., & Rau, G. (2000). Development of
recommendations for SEMG sensors and sensor placement procedures. Journal of
Electromyography and Kinesiology, 10, 361-374.
Hodson-Tole, E. F., & Wakeling, J. M. (2009). Motor unit recruitment for dynamic tasks:
current understanding and future directions. Journal of Comparative Physiology B, 179,
57-66.
Hug, F., Faucher, M., Kipson, N., & Jammes, Y. (2003). EMG signs of neuromuscular fatigue
related to the ventilatory threshold during cycling exercise. Clinical Physiology and
Functional Imaging, 23, 208-214.
Hummel, A., Läubli, T., Pozzo, M., Schenk, P., Spillmann, S., & Klipstein, A. (2005).
Relationship between perceived exertion and mean power frequency of the EMG signal
from the upper trapezius muscle during isometric shoulder elevation. European Journal
of Applied Physiology, 95, 321-326. doi:10.1007/s00421-005-0014-7
Jansen, R., Ament, W., Verkerke, G. J., & Hof, A. L. (1997). Median power frequency of the
surface electromyogram and blood lactate concentration in incremental cycle
ergometry. European Journal of Applied Physiology and Occupational Physiology, 75,
18

102-108. doi:10.1007/s004210050133
Johnson, M. A., Polgar, J., Weightman, D., & Appleton, D. (1973). Data on the distribution of
fibre types in thirty-six human muscles. An autopsy study. Journal of Neurological
Sciences, 18, 111-129.
Kandou, T. W., Houtman, I. L., vd Bol, E., de Boer, R. W., de Groot, G., & van Ingen Schenau,
G. J. (1987). Comparison of physiology and biomechanics of speed skating with cycling
and with skateboard exercise. Canadian Journal of Sport Science, 12, 31-36.
Kautz, S. A., & Hull, M. L. (1993). A theoretical basis for interpreting the force applied to the
pedal in cycling. Journal of Biomechanics, 26, 155-165.
Kuipers, H., Verstappen, F. T., Keizer, H. A., Geurten, P., & van Kranenburg, G. (1985).
Variability of aerobic performance in the laboratory and its physiologic correlates.
International Journal of Sports Medicine, 6, 197-201. doi:10.1055/s-2008-1025839
Kuznetsov, S. I., Popov, D. V., Borovik, A. S., & Vinogradova, O. L. (2011). Wavelet analysis
of m. vastus lateralis surface EMG activity in incremental tests till exhaustion using
bicycle and knee extension exercises. Fiziologiia Cheloveka, 37, 129-136.
Loras, H., Ettema, G., & Leirdal, S. (2009). The muscle force component in pedaling retains
constant direction across pedaling rates. Journal of Applied Biomechanics, 25, 85-92.
Macdonald, J. H., Farina, D., & Marcora, S. M. (2008). Response of electromyographic
variables during incremental and fatiguing cycling. Medicine and Science in Sports and
Exercise, 40, 335-344. doi:10.1249/mss.0b013e31815b491e
Neptune, R. R., Kautz, S. A., & Hull, M. L. (1997). The effect of pedaling rate on coordination
in cycling. Journal of Biomechanics, 30(10), 1051-1058.
Noordhof, D. A., Foster, C., Hoozemans, M. J., & de Koning, J. J. (2013). Changes in speed
skating velocity in relation to push-off effectiveness. International Journal of Sports
Physiology Performance, 8, 188-194.
19

Piucco, T., dos Santos, S.G., de Lucas, R.D., Dias, J.A. (2015). A novel incremental slide board
test for speed skaters: Reliability analysis and comparison with a cycling test. Apunts
Medicina de l’Esport, 50, 57-63.
Piucco, T., O'Connell, J., Stefanyshyn, D., & de Lucas, R. D. (2016). Incremental testing design
on slide board for speed skaters: comparison between two different protocols. Journal
of Strength and Conditioning Research. doi:10.1519/JSC.0000000000001392
Pressel, T., & Lengsfeld, M. (1998). Functions of hip joint muscles. Medical Engineering and
Physics, 20, 50-56.
Priego Quesada, J. I., Bini, R. R., Diefenthaeler, F., & Carpes, F. P. (2015). Spectral properties
of muscle activation during incremental cycling test. Journal of Science and Cycling, 4,
7-13.
Rundell, K. W. (1996). Compromised oxygen uptake in speed skaters during treadmill in-line
skating. Medicine and Science in Sports and Exercise, 28, 120-127.
Smith, J. L., Betts, B., Edgerton, V. R., & Zernicke, R. F. (1980). Rapid ankle extension during
paw shakes: selective recruitment of fast ankle extensors. Journal of Neurophysiology,
43, 612-620.
Snyder, A. C., O'Hagan, K. P., Clifford, P. S., Hoffman, M. D., & Foster, C. (1993). Exercise
responses to in-line skating: comparisons to running and cycling. International Journal
of Sports Medicine, 14, 38-42. doi:10.1055/s-2007-1021143
van Boxtel, A., & Schomaker, L. R. (1984). Influence of motor unit firing statistics on the
median frequency of the EMG power spectrum. European Journal of Applied
Physiology and Occupational Physiology, 52, 207-213.
von Tscharner, V. (2002). Time-frequency and principal-component methods for the analysis
of EMGs recorded during a mildly fatiguing exercise on a cycle ergometer. Journal of
Electromyography and Kinesiology, 12, 479-492.
20

Wakeling, J. M., & Horn, T. (2009). Neuromechanics of muscle synergies during cycling.
Journal of Neurophysiology, 101, 843-854. doi:10.1152/jn.90679.2008

21

Table 1. Frequency bands selected for band-pass filtering of EMG signals (von Tscharner
2002).
Band

1

2

3

4

5

6

7

8

9

High

48.45

75.75

110.00

149.00

193.45

244.45

300.80

363.80

431.65

Low

26.95

48.45

74.80

108.00

146.95

191.75

242.20

297.40

359.35

22

Figure 1. Schematic slide-board instrumentation. 1- Photoemitter; 2- Photoreceptor.

Figure 2. VL activation and accelerometer signal during skating on the slide board. The arrows
indicate the impact peak, registered by the accelerometer, of the foot on the lateral bumper of
the slide board (i.e. beginning of push-off phase) used to separate each stride (dashed line).

23

Figure 3. Mean overall RMS activation at each instant of time of the incremental test. VL=
vastus lateralis; BF= biceps femoris; GMd= gluteus medius; AM= adductor magnus; VM=
vastus medialis; GM= gluteus maximus.

24

Figure 4. Mean low (Panel A) and high (Panel B) frequency bands of each muscle at each
instant of the test, normalised by the overall activation at each instant of the test. *p<0.05. VL=
vastus lateralis; BF= biceps femoris; GMd= gluteus medius; AM= adductor magnus; VM=
vastus medialis; GM= gluteus maximus.

25

Figure 5. Relationship between CADmax, VL activation at 70% and 90% of the test, GM and
GMd at 70% of total time.

26