Muscle fatigue
in middle-distance running
By Ari Nummela, Karen Heath, Leena Paavolainen, Mike Lambert, Alan St Clair Gibson,
Heikki Rusko, Timothy Noakes
ABSTRACT
This study investigated fatigue-induced changes in neuromuscular and stride characteristics during
and immediately after a hard middle-distance running effort. Eighteen well-trained male distance runners performed a maximal
20m sprint test and maximal voluntary contractions (MVC) on a leg press machine before and immediately after a 5000m time
trial. In all the tests the EMG of five lower limb muscles was measured. The results showed that muscle fatigue measured in
maximal exercises is not related to the fatigue induced changes during the time trial. The fatigue in the 20m sprint test
was related to the maximal20m pre-test velocity. but the velocity loss during the time trial was inversely related to 5000m
performance and training volume. The authors conclude that the fatigue measured at maximal effort both pre- and post-time
trial is more related to sprint performance than endurance performance and that the fatigue measured during the time trial
is related to endurance performance and factors affecting pacing strategy. The findings support the idea that pacing strategy
is regulated in an anticipatory manner by a central governor, which ensures that physiological reserves are maintained. This
article was originally published in the International Journal of Sport Medicine under the title "Fatigue during a 5-km
running time trial"
AUTHORS
Ari Nummela, PhD, is a researcher at the Research Institute for Olympic Sports, Jyvaskyla, Finland.
Karen Heath, PhD, is a researcher at the University of Cape Town and the Sports Science Institute of South Africa, Cape Town,
South Africa. Leena Paavolainen, PhD, is a researcher at the Research Institute for Olympic Sports, Jyvaskylti, Finland. Mike
Lambert, PhD, is a Professor at the University of Cape Town and the Sports Science Institute of South Africa, Cape Town, South
Africa. Alan St Clair Gibson, PhD, is a Professor at the University of Cape Town and the Sports Science Institute of South
Africa, Cape Town, South Africa. Heikki Rusko, PhD, is a Director at the Research Institute for Olympic Sports and a Professor
in the Department of Biology of Physical Activity, University of Jyvaskyla, Finland. Timothy Noakes, MBChB, MO, DSc, is a
Professor at the University of Cape Town and the Sports Science Institute of South Africa, Cape Town, South Africa. He is
a member of NSA's International Scientific Advisory Board.
Introduction
Muscle fatigue is a complex phenomenon that can be defined as
an inability to maintain a level of force production or as a reduction in the maximum force that a muscle can exert. After
exercise involving long-term running, the isometric strength loss is related in a non-linear way to the exercise duration".
A decrease in maximal running speed and an increase in ground contact time also occur after long running exercises.
Muscle fatigue may arise not only because of peripheral changes at the level of the muscle, but also because the central nervous
system fails to drive the motor neurons adequately. It has been suggested that during long-duration exercise central fatigue
may develop. Furthermore, metabolic and structural changes may be related to muscle fatigue after long-duration exercise,
It has also been suggested that in an exercise such as running, which repetitively evokes the stretch shortening cycle, performance
may be impaired partly from alterations in stiffness regulation.
A decrease in integrated electromyography
(EMG) activity during maximal voluntary contraction and during maximal velocity running has been recorded for lower limb muscles
after prolonged runs. The effects of fatigue on EMG activity in submaximal exercises are not as clear as the reduced EMG activity
in maximal efforts. NICOL et al. observed that the integrated EMG of the gastrocnemius muscle increased during the push-off
phase at sub-maximal running velocity after the marathon, AVOGADRO et al. and PAAVOLAINEN et al. measured EMG activity of
lower limb muscles during a fatiguing running exercise, They did not observe any changes in integrated EMG values at submaximal
running velocities, BORRANI et al. observed a progression in the average frequency of the motor unit discharge toward the
high frequencies, which coheres with the hypothesis of the progressive recruitment of fast-twitch fibres during
the
VO2 slow component.
Self-selected pacing strategy during a time trial plays a key role in a complex
regulatory system, in which an integrated neural control regulates running intensity to ensure that any damage to cellular
function does not occur. Optimal pacing strategy differs for distances shorter than 800m compared to distances 1500 to 10,000m.
In the optimal short-term races «800m), the highest running velocities are achieved after the acceleration phase, and
thereafter the running velocities decreased gradually to the end of the race. In longer duration events, pacing strategy is
regulated to ensure that a reserve is maintained, and therefore the highest velocities occur at the end of the race",
It has been proposed that the brain uses an "internal clock" to generate knowledge of the distance still to be covered,
so that running velocity and metabolicrate can be altered appropriately during the time trial. Therefore analyzing the pacing
strategy in time trial events may allow insight into the muscle fatigue and related physiological and regulatory processes,
In the present study a 5000rn time trial was selected, since in 5000m world record runs acing strategy has been remarkably
consistent. Either the first kilometre or the final kilometre is always the fastest while the middle part of the race, from
2 - 4km, has never been faster than either the
first or the final kilometre in any world record performance.
Most studies have examined muscle fatigue and related neuromuscular alterations by looking at the difference in maximal performance
between pre- and post-exercise but little is known about the development of fatigue during the exercise, Therefore the aim
of the present study was to compare the changes in EMG activity, force production and running stride characteristics in maximal
short-term performance pre- and post-exercise to the respective changes developed during a 5000rn running time trial. We hypothesized
that the changes in maximal short-term performance (MVC and 20rn sprint running) after a fatiguing endurance exercise are
not completely related to the respective changes during the self paced endurance exercise.
Materials and Methods
Subjects
Eighteen well-trained male distance runners were recruited to participate in this study. Subjects were included if
they were able to complete 10km in under 38 minutes. Each subject signed an informed consent form at the beginning of the
study. The study was performed in Cape Town and approved by both the Ethics and Research Committee of the Faculty of Health
Sciences, University of Cape Town, South Africa and the Ethics Committee of the University of Jyvaskyla, Finland.
Experimental design
The subject runners were required to visit the laboratory on three separate occasions over
a ten-day period. They were asked to maintain their regular physical activity pattern for the duration of the study and were
requested not to exercise on the morning prior to their testing. On their first visit to the laboratory, the subjects were
given the opportunity to become familiar with equipment and testing protocols that would be used during the trial. This familiarization
was performed in an attempt to reduce error associated with subjects performing unaccustomed exercise. The volume of training
(km-week) and racing history during the preceding three months was also obtained from each subject.
On the first testing day, the subjects performed the 5000m time trial and two maximal performance tests before and immediately
after the time trial. The maximal performance tests were a 20m maximal sprint with a running start of 15m and maximal voluntary
contraction (MVC) with knee extensor muscles.
On their next visit to the laboratory, no more than
seven days later, peak treadmill running speed (PTRS) and VO2peak were measured using a continuous, incremental running protocol
on a horizontal, motor driven treadmill.
Electromyographic
activity and stride parameter measurements
Before the start of testing, each subject had
bipolar EMG electrodes (Beckman miniature skin electrodes, Illinois, USA) placed onto the vastus lateralis (VL), vastus medialis
(VM), rectus femoris (RF), biceps femoris (BF) and gastrocnemius (GA) muscles of the right leg. The skin was shaved, rubbed
with sandpaper and cleaned with alcohol. The electrodes were positioned longitudinally on the belly of each muscle and carefully
taped. All EMG data were recorded telemetrically (Biomes 2000, Glonner, Germany) during each running test on the track with
a laptop computer using Labview 5.1 (National Instruments, Texas, USA). The telemetric system was carried on the waist of
the subject with a specially designed belt. The EMG signal was amplified (band-pass cut off frequency was 10 - 550Hz) and
digitized at the sampling frequency of 1000Hz by an on-line computer system. The non-smoothed EMG signals were full wave rectified,
integrated (J.LVxs) and time normalized (Average EMG, AEMG) for the two phases of running: preactivation (100 ms before ground
contact) and total ground contact time.
In order to measure stride parameters, a photocell contact
mat and two photocell gates connected to an electronic timer (Newtest, Ltd, Oulu, Finland) were placed on the final straight
of the track. During the 5000m time trial, average velocity, ground contact times and flight times were measured with the
photocell contact mat from a 20m section at every other lap simultaneously with EMG from the VL, VM, RF, BF and GA. Stride
frequency (strides-s') was calculated by using contact times (CT) and flight times (FT) as (CT + FT)'. Stride length was calculated
by dividing the
average velocity by the stride frequency. Both EMG and stride parameter data collected during each stride
were averaged for the number of strides taken along the 20m straight during each running test.
Maximal 20m sprint
test
The subjects performed three to five maximal 20m sprints on the indoor track. They
were able to accelerate 15m before the start of the sprint to ensure a normal and maximal running gait throughout the 20m.
Each sprint was separated by a 1 - 2 min recovery period during which the runners returned to the start of the sprint course.
The sprints were measured using two photocell gates connected to an electronic timer (Newtest Ltd, Oulu, Finland). The fastest
performance of each subject was chosen for all subsequent data analysis.
Maximal voluntary contractions
Ten minutes after completing the final 20m sprint, subjects performed three five-second maximal
voluntary contractions (MVC) separated by five seconds rest on a custom-built seated leg press machine (Hur Ltd., Kokkola,
Finland) with a knee flexion angle of 70°. Subjects were asked to exert maximal force with both legs against the footplate
for five seconds during which standardized verbal encouragement was provided by the investigators. The single five-second
contraction that produced the highest force was taken as the non-fatigued, pre-5000m time trial value, and was thus used for
all subsequent data analysis. The EMG activity was analyzed from the VL, VM, RF and BF muscles of the right leg in the MVC.
The non-smoothed EMG signals were full wave rectified, integrated and time normalized (AEMG) for the period of two seconds
during the plateau phase of the force. Immediately after finishing the 5000m time trial, the subjects went directly to the
leg press machine where they performed one five-second MVC. This test occurred within ten seconds of finishing the time trial.
5000m time trial
To reduce the possible fatigue effects induced by the
leg press and 20m sprint tests, the subjects rested for 20minutes after these tests. Thereafter, they performed a 5000m time
trial on a 144m indoor track. They were instructed to run as fast as possible and were provided with verbal encouragement
during the run. Times were recorded for each lap and the split times at each kilometre were given to the athletes during the
time trial. During the final lap, subjects repeated the maximal 20m sprint test down the straight section on the track. The
signals from the EMG, photocell gates and photocell contact mat were analyzed from the fastest lap at the beginning of the
5000m (228 - 545m). from the slowest lap at the end of the 5000m (3972 - 4856m) and from the last lap (4980 - 5000m) when
the runners entered the appropriate 20m section of track. The average EMG values were calculated for all the muscles during
the ground contact and the 100 ms pre-activity phase.
Incremental exhaustive treadmill test
After a brief warm up, the runners
began running at 12 krn-h'. The speed was increased by 0.5 km-h' every 30 sec thereafter. Oxygen consumption (Cosmed K4 RQ,
Rome, Italy) and heart rate (Vantage XL Polar Electro, Finland) were measured continuously during the test. The test continued
until the runner was unable to maintain the pace of the treadmill. VO2peak was defined as the highest oxygen consumption during
the test over a 60 sec period. The PTRS was defined as the fastest running speed the subject could maintain for 30 sec.
Data
analysis
All the statistical analyses were done using SPSSWIN 13.0 (SPSS, Inc., Chicago,
ILL, USA). Analysis of variance (AN OVA) for repeated measures was used to evaluate changes before and after the 5000m time
trial and changes between the beginning and end of the time trial. Pearson's product moment correlation coefficient determined
relationships between variables. Values are expressed as mean ± standard deviation or standard error. Statistical significance
was accepted as p < 0.05.
Results
The descriptive data of the runners are shown in Table 1.
In Table 2
it can be seen that the maximal 20m velocity decreased 16.3 ± 5.2% after the 5000m time trial (p < 0.001). The relative
decrease in maximal velocity after the fatiguing time trial was similar to the decrease in average force (15.1 ± 14.9%)
in the maximal leg press (p < 0.01). In the time trial the decrease in the velocity from the highest velocity at the beginning
to the slowest velocity at the end of the 5000m was 9.9 ± 6.3% (p < 0.001). The velocity loss during the time trial
did not correlate with either the decrease in maximal 20m velocity (r = 0.24) or the decrease in the average force of the
MVC (r = 0.03). Furthermore, no significant correlation existed between the decrease in maximal 20m velocity and the decrease
in the average force of the MVC (r = 0.43).
The changes in stride characteristics and AEMG in the
maximal 20m sprint, in the MVC and during the 5000m time trial are also shown in Table 2. Figure 1 shows that the decrease
in
velocity was correlated significantly with the increase in the ground contact time both in the maximal 20m sprint
(r = -0.84, p < 0.001) and during the time trial (r = -0.92, p < 0.001). Although the AEMG during the pre-activation
and total ground contact phase decreased in the maximal 20m sprint and during the 5000m time trial, the decreases in the AEMG
were not related to the decrease in velocity. However, as shown in Figure 2, the changes in the AEMG during the pre-activation
were related to the changes in ground contact time (r = -0.80, P < 0.001) and stride frequency (r = 0.79, P < 0.001)
in the maximal 20m sprint. Similar relationships, although not significant, were observed during the 5000m time trial (Figure
2). Furthermore, the decreased average force in the MVC was related to the decrease in the AEMG (r = 0.82, P < 0.001).
The correlation coefficients between the selected variables of training and performance characteristics
of the runners show that the fatigue in the maximal 20m sprint was related to the maximal running velocity but not to the
volume of training or the performance in the 5000m time trial (Table 3). The fatigue during the 5000m time trial was related
to the relative velocity (% of PTRS) in the beginning of the time trial, the volume of training and the performance in the
time trial (Table 3). The fatigue in the MVC was only related to the relative velocity (% PTRS) in the beginning of the time
trial but not to any training or performance characteristics of the subjects (Table 3). As shown in Figure 3, the correlation
coefficient between the average velocity in the 5000m time trial and average lap velocity increased gradually during the first
kilometre of the time trial and remained above r = 0.90 from 900m until it decreased to 0.61 during the final lap.
Discussion
These data suggest that this group of athletes displayed a relatively wide range of both anthropometrical
and performance variables. Several studies have measured knee extensor muscle
isometric strength loss after prolonged
exercise. After running exercise for longer than two hours the loss of strength increases as a non-linear function of the
exercise duration". The isometric
strength loss in the present study was 15%, which is somewhat lower than that
in running exercises of longer duration shown in a previous study and therefore it was in accordance with the finding of MILLET
& LEPERS. We observed a decrease in velocity (16%) in the maximal 20m sprint after the 5000m time trial similar to the
decrease in force in the MVC (15%). In the previous study, the decrease in the maximal velocity was 23% after a 10,000m time
trial in the well-trained endurance athletes. The first interesting and remarkable finding of the present study was that the
changes in the average force of the MVC, the maximal velocity of the 20m sprint test and the decrease in velocity during the
5000m time trial were not related to each other. The strongest relationship (r = 0.43) existed between the changes in maximal
exercises before and after the 5000m, but this did not decrease in force in the MVC (15%). In the previous study, the decrease
in the maximal velocity was 23% after a 10,000m time trial in the well-trained endurance athletes.
The first interesting and remarkable finding of the present study was that the changes in the average force of the MVC, the
maximal velocity of the 20m sprint test and the decrease in velocity during the 5000m time trial were not related to each
other. The strongest relationship (r = 0.43) existed between the changes in maximal exercises before and after the 5000m,
but this did not reach the level of significant. In the MVC and maximal 20m sprint tests the force or velocity loss was not
related to the performance in the 5000m. This finding supports the results of PAAVOLAINEN et al." who found that the
decrease in velocity in the maximal running test after a 10,000m time trial was not related to the performance in the time
trial. However, the decrease in velocity from the beginning to the end of the 5000m time trial was related to the performance
in the 5000m. An interpretation of these results is that the type of muscular contraction and type of test (i.e. MVC vs. 20m
sprint vs. 5000m self paced) determine the reduction in strength and power. Since the fatigue results of MVC were not related
to any measured performance abilities, it is questionable to use an isometric test as an indicator of fatigue for runners.
As shown in Table 3, the fatigue in the MVC was not related to any measured performance abilities.
The velocity decrease in the maximal 20m sprint correlated positively with the maximal 20m velocity before the 5000m and negatively
with the VO2max suggesting that the maximal running velocity decreased most in the athletes who had the best sprinting abilities
and the lowest maximal oxygen uptake. This confirms earlier findings, which have shown that sprint and middle-distance runners,
who had higher power output and lower endurance abilities, showed greater fatigability than long-distance runners. This could
be explained at least partly by fibre type composition, since athletes with high percentage of fast-twitch fibres are more
susceptible to fatigue than athletes with a high percentage of slow-twitch fibres.
The decrease in
velocity during the 5000m was positively correlated with the initial velocity of the 5000m, negatively correlated to the training
volume (recorded for the three months preceding the tests) and to the VO2max suggesting that initial velocity and endurance
abilities affect the velocity loss or fatigue resistance during the 5000m. The present results suggest that the fatigue in
the maximal 20m sprint test was more dependent on the fast-twitch muscle fibres, utilization of anaerobic energy sources
and ability to produce power than the velocity loss during the 5000m, which was more dependent on the endurance characteristics.
This may also explain the shape of the velocity curve and the remarkable increase of velocity (34%) at the end of the time
trial,
when the runners should have been most fatigued. This suggests that the runners had the ability to increase their
velocity at any time during the 5000m but their pacing strategy controlled this to prevent premature exhaustion.
The pacing strategy in the present study was freely chosen by the athletes. They were only asked to run 5000m as fast as possible
and to run the last 20m at their maximal effort. Furthermore, the laps were counted for the athletes and split times at each
kilometre were given to the athletes during the time trial. The average velocity curve of the athletes was typical in the
5000m run, which is characterised by fast starts, a period of slower running during the middle of the race, and followed by
a significant increase in speed towards the end. This is similar to the pacing strategies adopted by elite rowers during 2000m
rowing races and time trials. ST CLAIR GIBSON et al. have stated that the most important factor allowing the establishment
of the pacing strategy is knowledge of the endpoint. However, the pacing strategy is also affected by the environmental conditions,
the motivation of the athlete, the knowledge and experience of the athlete, and each athlete's particular physiological capacity.
The environmental conditions were constant in the present study, since the tests were performed on an indoor track with controlled
environmental conditions. The motivation of the runners was not measured but they were all verbally encouraged during the
whole run. The knowledge of split times at each kilometre may also affect motivation and pacing strategy of the athletes,
but this seems to have only a minor effect since the typical velocity profile for 5000m was observed in the present study.
The velocity at the beginning of the 5000m was affected by the experience of the runners in the present
study. The experience of the runners can be evaluated by the level of 5000m performance and training volume so that more experienced
runners had better distance running performance and trained more than the runners with less experience. We observed that the
runners who had better 5000m performance and higher training volume during the three months preceding the tests started the
5000m at relatively lower velocity (% PTRS) and they had more constant pacing during the 5000m, since a significant correlation
was observed between the high initial velocity and velocity loss during the time trial. The correlation coefficients between
the lap velocities and 5000m performance in Figure 3 supported this finding, since the initial velocity was poorly correlated
with the 5000m performance in this group of athletes.
In order to maintain and alter the running
pace during the 5000m time trial, the brain must process an enormous quantity of data from the external conditions and from
the various physiological systems of the body. From the beginning of the run, this afferent information is used to set and
alter the appropriate pace control algorithm for the rest of the time trial as suggested by ST CLAIR GIBSON et al. An example
of the runner's physiological limits and successful pacing algorithm can be seen in the middle part of the 5000m time trial
(Figure 3) where there were high correlation
coefficients (r > 0.90) between the lap velocities and 5000m performance.
The reduced EMG activity and velocity during the 5000m time trial supports the idea that in longer
duration events the pacing strategy is regulated to ensure that a reserve is present at the end. Therefore running speed can
be increased consequent to an increase in skeletal muscle recruitment at the end of the run. In the present study, this was
shown as an increased EMG activity at the end of the 5000m time trial. The phenomenon seems to be different in short -duration
runs like the 400m and long-duration runs like the triathlon and marathon. In 400m time trials, the running velocity decreased
progressively even though motor unit activation (measured as EMG activity) was able to increase to compensate for the apparent
failure of contractility. This could be explained by the differences in pacing strategy and by the use of anaerobic energy
sources. In optimal pacing strategy for short-term maximal exercises like the 400m, the velocity decreased throughout the
exercise and there is no need for a reserve at the end of the run. A possible reason for the difference between 5000m and
marathon running is that muscle fatigue and strength
loss is greater in longer events, as suggested by MILLET &
LEPERS, and athletes are not able to increase their running velocity at the end of the marathon to the same extent as they
do at the end of the 5000m. This greater strength loss in a marathon requires that the athletes must compensate by increasing
their motor unit activation.
Another important finding of this study was that the AEMG during both
the pre-activation and the ground contact phase of the lower limb muscles decreased during the 5000m time trial. These findings
are similar to those after a 400m sprint, a 10,000m time trial the MVC decreased after the 5000m time trial. Force and EMG
have also decreased after fatiguing maximal leg extensions, high intensity cycling, and after a 65km footrace. These studies
have concluded that the reduction in EMG and the corresponding decrease in isometric force represent a form of central fatigue.
However, this explanation is not absolutely established, since the sarcolemmal excitability can also be modified. The reduced
EMG activity in the maximal situation can be explained by spinal and supraspinal factors. The reduced activity can be caused
by a failure to generate output from the motor cortex or by an alteration in input from muscle spindle, Golgi tendon organ,
and group III and IV muscle afferents innervating the fatiguing muscle, or both.
The contribution
of the spinal and supraspinal changes to fatigue was not measured in this study, but we observed a clear EMG reduction in
the pre-activation phase during running. AVELA & KOMF observed that the reduced EMG activity in the soleus and vastus
medialis muscles were more pronounced in the pre-activation and eccentric phases compared to the concentric phase
after
the fatiguing marathon run. It has been agreed that the pre-activation phase is strongly pre-programmed and controlled by
higher supraspinal centers. Thus our findings may suggest that some level of supraspinal fatigue was present. HORITA et al.
have found that in drop jump exercises muscle pre-activity correlated with muscle stiffness during the initial phase of contact.
This relationship suggests that pre-activation regulates the force production in drop jumps as demonstrated by the high series
elastic component stiffness, the short contact time and more economical
work. In the present study, the decreased pre-activation
and a significant correlation between the decreased pre-activation and increased contact times suggest that changes in stiffness
regulation may playa role in reducing maximal running speed during and after the 5000m time trial. Similar findings were also
observed in the fatigue that developed after a 10,000m time trial, The authors found a decreased pre-activation and a significant
relationship between the decreased pre-activation and decrease of horizontal ground reaction force in the braking phase.
In conclusion, the results of the present study showed that muscle fatigue measured in maximal exercises
like a 20m sprint and MVC are not related to the velocity loss during a 5000m time trial. The results of fatigue depend on
the type of the measurements, and whether the fatigue is measured pre- and post-exercise or during the fatiguing exercise.
The fatigue in 5000m running measured pre- and post-exercise at maximal effort is more related to sprint performance rather
than endurance performance, but the fatigue measured during 5000m running is related to
endurance performance and factors
affecting pacing strategy. We propose that the present findings support the idea that pacing strategy is regulated in an anticipatory
manner by a
central governor, which ensures that physiological reserves are maintained.
FROM: IAAF NSA 4-2009
