Muscle fatigue can be defined as a reduction in the maximum force or power output and is known to be governed by both peripheral and central mechanisms (13). Peripheral changes in the muscle due to fatigue involve processes such as a modulation in excitability of muscle fibers and a reduction in the excitation–contraction coupling processes (11), whereas central changes in spinal reflex pathways and descending corticospinal pathways alter the spinal motoneurons recruitment and firing rates (13). In this context, transcranial magnetic stimulation (TMS) provides a noninvasive method of investigating changes within the central nervous system during fatigue (13). By stimulating the motor cortex at suprathreshold levels, the corticospinal pathway can be activated to elicit short latency muscle responses known as motor-evoked potentials (MEP). Several studies have used suprathreshold TMS and cervicomedullary stimulation (a method to directly stimulate the corticospinal axons) to show that motor cortical excitability increases while spinal excitability decreases during fatiguing maximal and submaximal single joint contractions (7,20,25,36). In contrast to these isometric single joint studies, we recently provided evidence that during 30 min of steady-state sustained cycling exercise, the responsiveness of the motor cortex and motoneurons was similar to baseline. Because changes in cortical and spinal responsiveness seem to differ between sustained locomotor versus sustained isometric single joint contractions, it seems that there are task-dependent differences in cortical and spinal responses to fatigue, possibly linked to differences in both systemic and local (i.e., muscular) changes between the two types of exercises.
Although studies monitoring central responsiveness via evoked potentials are reliant on synchronous activity in a large population of corticospinal cells and motoneurons, TMS at an intensity that is subthreshold for the generation of a descending volley (i.e., no MEP or facilitation present in the EMG) allows the measurement of intracortical inhibition with little disruption to the ongoing muscle activity (6,8). The low-intensity stimulus activates low-threshold intracortical inhibitory neurons that have direct and indirect projections onto the corticospinal cells (9). If the corticospinal cells make direct contribution to exciting the motoneurons involved in muscle activation, a decrease in output from these cells via activation of inhibitory interneurons can cause a suppression in the ongoing EMG (6,8). The technique is commonly used when performing isometric contractions (6,8); however, it has also been used during multijoint dynamic contractions in tasks such as walking (3,27), hopping (40), and cycling (32). It is important to note that the technique can only be used to infer intracortical effects when the stimulus intensity is below the threshold to activate descending projections. Because it will only be possible to find such a stimulus intensity when the threshold for activating inhibitory intracortical interneurons is substantially lower than for activating excitatory intracortical interneurons, and because these relative excitability levels seem to differ between people, it is typical that this method can only be applied in approximately half of any given pool of volunteers for experiments of this kind.
So far, only one study has used subthreshold TMS technique during a single joint submaximal isometric contraction held to task failure and showed a greater suppression in the ongoing EMG at task failure when compared with the start of the contraction (30). The data showed that the excitability of intracortical inhibitory circuits increases during a sustained isometric contraction. This effect is similar to an increase in silent period after an MEP (24,36), the latter part of which is known to be influenced by an increase in intracortical inhibition (33) during sustained single joint contractions. Because corticospinal cell excitability increases during sustained single joint contraction (20,23,25,36), it seems that changes in the excitability of intracortical inhibitory interneurons does not necessarily parallel changes in excitability of corticospinal cells during sustained isometric contractions. However, no previous study has looked at changes in excitability of intracortical inhibitory interneurons during sustained locomotor exercise.
In the current study, we used subthreshold TMS to investigate the extent to which intracortical inhibition via the suppression of EMG changed during 30 min of sustained cycling exercise. We tested the hypothesis that intracortical inhibition would increase during sustained cycling exercise as occurs during sustained single joint contraction.
Sixteen healthy subjects (13 men and 3 women), with a mean age of 24.5 ± 1.1 yr, body mass of 70.2 ± 1.5 kg, and height of 176.1 ± 1.9 cm, were recruited for the study. Each subject gave written informed consent before the study and completed a health-screening questionnaire for participation in studies involving TMS and endurance exercise. The experiment procedures were approved by the local university ethics committee.
EMG and ECG recordings
EMG recordings were taken from the right knee extensor (vastus lateralis [VL]), knee flexor (long head of biceps femoris [BF]), and dorsiflexor (tibialis anterior [TA]). ECG was used to monitor and record the HR throughout the study. Areas under all electrodes were shaved, abraded, and cleansed with alcohol swabs to obtain accurate EMG and ECG recordings. EMG was recorded via bipolar configurations (Ag-AgCl, 10-mm diameter, 2-cm interelectrode distance; Kendall, Medi Trace, Canada) positioned over the muscle bellies. ECG was also recorded via bipolar configuration with the anode and cathode positioned approximately 10 cm apart across the chest. To minimize movement artifacts, electrode cables were fastened to the subjects’ muscles using medical adhesive tape and wrapped with elastic bandage. The signals were amplified (200–1000 times; Neurolog Systems, Digitimer Ltd., Hertfordshire, England, UK), band-pass filtered (50–2000 Hz; NL844; Digitimer Ltd.), and converted from analog to digital at a sampling rate of 2000 Hz using a 16-bit Micro 1401 mk-II and Spike 2 data collection software via custom written scripts (Cambridge Electronic Design, Cambridgeshire, England, UK).
Cycle ergometer setup
Subjects sat on a cycle ergometer with the seat height adjusted to an optimal level and their feet fastened securely to the pedals. They rested their arms on a custom-made adjustable frame mounted to the front of the ergometer. A chin rest attached to the frame was used to ensure that the upper body and head were kept stable during stimulations (Fig. 1A). This setup enabled a consistent delivery of TMS and has been used previously (32). The crank angle was monitored continuously via a calibrated linear potentiometer coupled to the pedal end of the crank.
Each subject participated in two sessions, an incremental workload session for the determination of maximum workload followed by a sustained cycling session, separated by at least 48 h. Subjects indicated their perceived exertion value every 3 min during fatigue cycling in both sessions using a Borg scale of 6–20 (6 = no exertion, 20 = maximal exertion). Throughout the study, subjects were required to keep their cycling cadence constant at 80 rpm via self-monitoring on a cadence meter in front of the ergometer (see Fig. 1A).
The cycling exercise began with a standard warm-up at 1 W·kg−1 body weight for 5 min followed by 3 min of rest. Afterward, they were required to perform an incremental cycle exercise test to exhaustion on a mechanically braked cycle ergometer (Lode Rehcor; Lode BV, the Netherlands) for the determination of maximum workload (Wmax) to set target power outputs (i.e., 75% Wmax) for the following sustained cycling session. The incremental exercise began at 100 W for 3 min. Output was then increased by 20 W (female) and 30 W (male) every 3 min until volitional exhaustion. In the latter stage of this test, subjects were verbally encouraged to exert themselves maximally. Wmax was taken as the highest 3 min average of power output (mean group Wmax = 221.9 ± 11.4 W).
Sustained cycling session
The session began with a 2-min warm-up, with the subject cycling at the prescribed workload (i.e., 75% Wmax). The EMG obtained during warm-up cycling was rectified before it was averaged with respect to a “position” signal from the top dead center (i.e., 0°) on the crank cycle during a 20-s segment midway through the 2-min bout of cycling. This was used to determine the relationship between crank angle and quadriceps EMG activity and also served to identify a point in the crank cycle where EMG activity in VL was increasing and was at approximately 50% of its maximum. This point was then targeted for all stimulations during cycling (mean position = 332.3 ± 3.1 on the crank cycle).
Transcranial magnetic stimulation
The optimal location to elicit motor-evoked responses in the right leg muscles was determined using a magnetic stimulator (model 2002; The Magstim Company Ltd., UK) with a concave, double cone coil (130-mm diameter). The junction of the double cone coil was aligned tangentially to the sagittal plane, with its center 1–2 cm (mean optimal position = 1.75 ± 0.1 cm) to the left of the vertex. The optimal coil position (with posterior to anterior induced current flow within the cortex) to elicit the largest response in VL during a small tonic contraction was determined before the experiment while subjects were quietly seated on the cycle ergometer. The position was marked directly on the scalp for accurate replacement throughout the session (i.e., between each control cycling bout and from the end of control cycling to the start of sustained and recovery cycling but kept stable throughout the 30-min sustained and recovery cycling).
Subthreshold TMS intensity and control cycling
Initially, 10 TMS pulses at an intensity above active motor threshold (mean intensity = 31% ± 1% of maximum stimulator output) and 10 nonstimulated (i.e., control) pulses were applied in random order at the selected crank angle during cycling at 75% of maximum workload. An average EMG trace was determined (with respect to a trigger signal) for a 100-ms period beginning 20-ms before each stimulation. The average rectified EMG was monitored online (from both TMS and control pulses) and overlaid for determining the effect of the stimulation on EMG amplitude. The latency of the MEP was recorded from the average EMG trace and used to help predict the latency at which inhibition would develop, which has been shown to be within 10 ms from the latency of facilitation (32). The stimulator intensity was then decreased by approximately 5%, and 10 more TMS and control pulses were applied during cycling. This procedure was repeated until no facilitation was observed in the EMG trace at the predicted latency. Once the subthreshold intensity was determined (mean intensity = 18.5% ± 0.8% of maximum stimulator output), a minimum of 50 TMS and 50 control pulses were applied during repeated 1-min bouts of cycling at 75% of Wmax (with at least 3 min of rest in between trials to minimize the effects of fatigue). Although VL was targeted for inhibition, the selected crank position coincided with small levels of EMG in the other leg muscles (i.e., BF and TA), consequently allowing inhibition to develop in these muscles for some subjects. The same protocol and intensity was used to collect the same number of TMS and control pulses during cycling at a reduced workload (i.e., 37.5% Wmax). This was performed to compare the effects of subthreshold TMS at lower level of cycling EMG (see Fig. 1B).
Sustained cycling and recovery
After the completion of the control bouts, subjects were required to cycle at 75% of Wmax for 30 min. Throughout the exercise, subthreshold TMS and control pulses were applied with a pseudorandomly determined interstimulus interval of between 3 and 5 s. At least 300 stimulations were elicited during the 30-min cycling. At the completion of the cycling, the workload was reduced to 37.5% Wmax, and subjects were required to cycle for an additional 5 min while subthreshold TMS and control pulses were elicited at the same intensity as during the 30-min cycling to monitor recovery.
Data analysis was performed via custom-written scripts using Spike 2 software. Specifically, the reference point for waveform averaging was taken as the same point on the crank angle used for eliciting stimulations. Averaging was done for all stimulations 1) during each of the two control cycling workloads, 2) every 5 min during the sustained cycling, and 3) during the 5-min recovery. The waveform average of the rectified EMG with subthreshold TMS pulses was overlaid with the waveform average of rectified EMG with no stimulations (control trace) to see the effects of the stimulations on ongoing cycling EMG.
EMG facilitation was defined as when the mean amplitude of the stimulated trace was larger than the amplitude of the control trace at the expected MEP latency. EMG suppression was determined via visual inspection and defined as any period in which the average EMG for the stimulated condition was less than control EMG for at least 4 ms between 20 and 50 ms for upper leg muscles (VL and BF) (32) and 30–60 ms for the lower leg muscle (TA) (27). The end of the suppression was defined as the point when the stimulated trace was above the control trace after the start of the suppression. A suppression of the ongoing EMG was only quantified in trials where no facilitation was observed. This method of quantifying suppression has been used in recent articles using the same technique (21,32,40).
In some subjects, facilitation (with and/or without subsequent suppression) developed over the course of the 3-min sustained cycling probably because of an increase in cycling EMG over the course of sustained cycling. An increase in EMG activity over the course of sustained exercise is reflective of an increase in the number of units recruited and/or their firing rates to counteract a lack of force production of the exercising muscle (5,12). This observation suggests, as expected, that the development of facilitation is sensitive to increases in EMG. However, because the EMG suppression method can only be validly applied when the stimulation is subthreshold for eliciting facilitation, subjects in whom facilitation developed over the course of the sustained exercise were excluded from the analysis (approximately 50% of the 16 subjects).
To address our main question as to whether intracortical inhibition increases during sustained cycling and for simplification of the statistical analysis, we compared the amount of EMG suppression in the initial 5 min with the last 5 min of the sustained cycling. Comparisons of the overlaid EMG averages were made between: 1) control cycling at 75% Wmax versus 37.5% Wmax, 2) first 5 min (0–5 min) versus last 5 min (25–30 min) of the 30-min sustained cycling exercise at 75% Wmax, and 3) control cycling at 37.5% Wmax versus recovery cycling at 37.5% Wmax. For comparisons between conditions, the start and the end of the longer duration of the suppression were used to set cursors and quantify suppression. For example, during sustained cycling, if suppression obtained during the first 5 min was 8 ms in duration while suppression obtained during the last 5 min was 5 ms in duration, then the suppression obtained in the first 5 min was used as the reference for the curser settings and quantifying suppression in the two conditions. Because suppression was occasionally absent during one of the measurement times, small negative values in the amount of suppression were also possible. To quantify the amount of suppression, the mean area of the EMG activity between the two cursors was determined for both the stimulated and the control trace. The amount of suppression was expressed as a percentage of the area of the stimulated trace to the control trace. The latency and the duration of suppression were also quantified. Average cycling EMG was measured for a 50-ms time window (50 ms after the selected point of stimulation) from the control trace of the waveform average.
Paired t-tests were used to test for differences in HR, RPE, cycling EMG, and amount of suppression (%) between 1) two control bouts at 75% and 37.5% Wmax, 2) first and last 5 min of sustained cycling at 75% Wmax, and 3) control and recovery cycling at 37.5 % Wmax. HR was not recorded in one of the subjects during control cycling, and two of the subjects did not complete recovery cycling. For each comparison, data were analyzed for all subjects in whom a suppression (as defined earlier) was elicited without prior facilitation at either time point. Data from all other subjects were excluded. As this protocol relies on objective criteria, it involves no opportunity for experiment-wise error. Further, any random effects would have equal likelihood of being expressed before and after fatigue. Thus, we are confident that this approach can reveal genuine differences in intracortical inhibition despite the high exclusion rate necessitated by characteristics of this method. All data are reported as mean ± SEM. Statistical significance was set at P ≤ 0.05.
Psychophysiological responses during sustained cycling
All subjects successfully completed the prescribed sustained exercise (75% Wmax for 30 min; 165.8 ± 8.3 W). Table 1 shows the HR and RPE during control, sustained, and recovery cycling. HR was significantly greater at a higher workload during the control bouts (t14 = 7.3, P ≤ 0.05; Table 1). During the last 5 min of sustained exercise, HR was significantly greater than during the first 5 min (t15 = 12.8, P ≤ 0.05) and had reached 94% of maximum recorded at task failure of cycling during the incremental session. Although HR during recovery was significantly greater than that during control cycling at 37.5% Wmax (t13 = 8.0, P ≤ 0.05), it was comparable with the first 5 min of sustained exercise (t13 = 0.9, P = 0.4). Similar to the pattern of HR changes, RPE during the last 5 min of exercise was significantly greater than the first 5 min (t15 = 9.3, P ≤ 0.05; Table 1) and had reached 87% of maximum recorded at task failure during the incremental session, indicating that all subjects had reached a high level of subjective fatigue. During recovery, RPE was significantly lower than during the first 5 min of sustained cycling (t13 = 4.0, P ≤ 0.05).
Characteristics of suppression
Table 2 shows the latencies and durations of the suppression and the number of subjects in whom suppression could be obtained in each of the three conditions under comparison. The suppression of EMG was seen in approximately 50% of the 16 subjects during control and sustained cycling. An exception was during recovery in TA, where relatively fewer subjects (n = 3) exhibited suppression. In all muscles and conditions (i.e., control, sustained, and recovery cycling), the onset of inhibition occurred approximately 10 ms after the expected MEP latency (i.e., ∼30 ms for VL and BF and ∼40 ms for TA) and lasted for approximately 7 ms (for raw data from single subjects representative of group data, see Fig. 2). It is important to note that the absolute magnitude of the suppression is not comparable across the three conditions contrasted in Figure 3 (i.e., control, fatigue, and recovery).This is because suppression values from different subjects contributed to each contrast because suppression was only calculated in the absence of an initial facilitation. For example, some subjects who demonstrated inhibition in the control trials developed facilitation during the sustained fatigue (i.e., 5 of 10 subjects). Thus, data from these subjects did not contribute to the mean suppression value in the first 5 min in the fatigue contrast, and the absolute magnitude of the suppression seems lower despite similar exercise conditions.
Despite significant differences in cycling EMG amplitude between the two workloads in all muscle groups (t4–9 > 3.1, P ≤ 0.05; Fig. 3A), the amount of suppression during 75% Wmax was not different from that during 37.5% Wmax in all muscles (t4–9 < 1.4, P > 0.1; Fig. 4A).
In all the muscles, EMG during the last 5 min of the total 30 min of cycling was not different to the first 5 min of cycling (t7–8 < 2.1, P > 0.05; Fig. 3B). In some subjects, the suppression of EMG was notseen during the first 5 min of sustained cycling, but it developed and became apparent during the last 5 min (for representative subject data, see Fig. 2B). Compared with the first 5 min, the amount of suppression increased by 10%–15% on average during the last 5 min in all muscle groups tested (t7–8 > 3.4, P ≤ 0.05; Fig. 4B).
Cycling EMG during the recovery phase at 37.5% Wmax was similar to control cycling at 37.5% Wmax in all the muscles (t2–7 < 0.8, P > 0.4; Fig. 3C). The amount of suppression during the recovery phase was also similar to that obtained during control cycling at 37.5% Wmax in all muscle groups (t2–7 < 1.2, P > 0.2; Fig. 4C).
The main findings show that the TMS-evoked suppression of cycling EMG increased significantly for more than 30 min of sustained cycling exercise. The data suggest an increased responsiveness of the intracortical inhibitory interneurons under conditions of increased psychophysiological demands of the locomotor exercise. The increase in intracortical inhibition during the sustained cycling exercise is comparable with that observed in a previous study during a sustained single joint isometric contraction (30), indicating that the intracortical inhibitory interneurons respond similarly to the two types of exercises.
Subthreshold TMS measures intracortical inhibition
The technique of subthreshold TMS is based on the principle that at very low stimulation intensities, low-threshold intracortical inhibitory interneurons are activated, which via their synaptic input reduce corticospinal cell excitability and therefore decrease the neural drive that produces voluntary movement (26,30).The reduction of drive can be observed as a reduction in amplitude of the ongoing EMG. There is also evidence that the observed suppression in the EMG is specifically due to a reduction in output from the corticospinal cells. First, the stimulation of the corticospinal tracts caudal to the motor cortex at subthreshold intensities failed to induce the suppression (8,27), and second, the observation of simultaneous inhibition in agonist–antagonists pairs (27,32,40) excludes the involvement of spinal reciprocal inhibitory interneurons. Furthermore, it has been shown via single motor unit recordings that the suppression in the EMG can be obtained within 2 ms from the time of first arrival of a descending volley from the cortex, indicating that it involves the most direct corticospinal projections (6).
In the current study, a lack of suppression of EMG without prior facilitation was observed in 50% of the subjects who participated. A similar percentage has been observed in other studies using this method, and it is generally accepted that this method of quantifying intracortical inhibition is valid despite the high exclusion rate (3,30,40). As previously suggested, the inability to induce suppression in the absence of prior facilitation would be expected if the thresholds for activation of the inhibitory versus excitatory intracortical interneurons are similar (32). In the current study, the excitability of both inhibitory and excitatory interneurons might have been affected by fatigue, such that a stimulus intensity that was subthreshold for EMG facilitation at baseline may have become suprathresold after fatigue in some subjects. Because it is not possible to use the method whenever a descending volley is induced by TMS, we were unable to quantify intracortical inhibition in cases such as this. However, because we compared the same subjects across time, via the same time window for EMG measurement, we are confident that our experiment provides valid information indicating that intracortical inhibition increases because of sustained cycling exercise.
Modulation of intracortical Inhibition with locomotor exercise
The data from the current study provides new evidence that the excitability of the intracortical inhibitory interneurons increases during sustained cycling exercise. We recently used suprathreshold TMS and stimulation at the cervicomedullary junction to investigate cortical and spinal changes during a locomotor cycling exercise of similar intensity as in the current study. The results of that study provided evidence that during the 30-min steady-state sustained cycling exercise, the responsiveness of the motor cortex and motoneurons was similar to baseline (31). We suggested that the apparent lack of change in cortical responsiveness was most likely mediated by cortical processes such as an increased excitability of intracortical inhibitory interneurons. The data from the current study lend support to our previous suggestion. Although cycling exercise was not performed to task failure, during the last 5 min of the sustained cycling exercise, psychophysiological demands of the exercise were high, as shown by increased HR and high levels of subjective exertion. Although the current study did not have an objective measure of fatigue, previous studies have demonstrated a reduction in maximum muscle force immediately postcycling exercise of similar intensity and duration (4,39). At the end of the sustained exercise, when subjects were required to continue cycling at a lower workload, the psychophysiological responses and the amount of suppression were at baseline levels. Thus, we have shown a greater suppression of EMG during the last 5 min of sustained cycling exercise compared with the initial 5 min when subthreshold TMS was applied under conditions of increased psychophysiological stress. Although the current data suggest that the intracortical inhibitory interneurons were not more excitable immediately after exercise (i.e., recovery stage), it is also possible that there may be short-lasting effects that are not apparent with methods that require averaging over many stimulations.
Cycling EMG during the last 5 min of the sustained exercise was not different from the first 5 min of exercise. Indeed, this lack of an increase in EMG activity is in contrast to that seen during submaximal sustained isometric single joint contractions whereby EMG increases over the progression of the contraction (15,22). However, a similar pattern of no change in cycling EMG has been reported previously during constant workload cycling exercise of similar intensity and duration (10,31). In any case, the increase in the amount of suppression during the last 5 min of sustained exercise in all muscle groups is argued to be independent of changes in EMG per se because our control data suggest that changes in workload and cycling EMG have little effect on the amount of EMG suppression. The observations within the current study support the argument of an increased excitability of intracortical inhibitory interneurons under conditions of sustained locomotor exercise.
Possible mechanisms of increased intracortical inhibition
Given that the increase in the amount of suppression during sustained cycling exercise in the current study is similar to that reported during a sustained isometric contraction (30), it seems that the responsiveness of the intracortical inhibitory interneurons to the two types of tasks is similar. However, this similarity in intracortical inhibitory responses contrasts with the difference in the responsiveness of the corticospinal cells during the two types of exercises. On one hand, corticospinal cell excitability is enhanced during sustained single joint contraction (20,25). On the other hand, we recently provided evidence of a lack of an increase in the responsiveness of motor cortical cells during sustained locomotor exercise (31). The responsiveness of the corticospinal cells to fatiguing exercise therefore seems to be task specific (38), which might be to be due to differences in systemic and local (i.e., muscular) responses to locomotor versus single joint exercises (31). Indeed, during locomotor exercises involving multiple limb muscles, cardiorespiratory demands are higher than that during single joint exercise (1,2,17), and factors such as temperature, blood glucose, catecholamines, and cerebral oxygenation are likely to become more important for the regulation of homeostasis (16,18,28,29). These factors may have differentially influenced the responsiveness of the corticospinal cells.
Although the mode of exercise seems to have different effects on the responsiveness of corticospinal cells, the intracortical inhibition measured via subthreshold TMS seems to increase with both single joint and locomotor fatiguing exercise. A possible factor underlying the increased intracortical inhibition observed in the current study and previously during sustained single joint contraction (30) is intramuscular metabolic changes that are common to fatiguing contractions of any type (11,35). There is evidence that when feedback from sensory neurons was blocked via spinal anesthesia, cortical silent period was the same, immediately postfatiguing single joint exercise, whereas it increased in control conditions (19). There is also evidence via the cortical twitch interpolation technique using TMS (13) to show that the output from the motor cortical cells is suboptimal during fatiguing single joint contractions (34,36) and with reference to elbow flexors during exhaustive cycling exercise (28). In addition, the extent of this activation was shown to remain diminished when a muscle was held ischemic postsingle joint exercise (14). The evidence from these studies suggests that increased activity of group III and IV afferents is likely to play a role in increasing the amount of intracortical inhibition and reducing cortical drive from at or above the motor cortical cells during sustained exercise (30,34,37).
The current results show that the sustained cycling exercise induced considerable psychophysiological stress and increased the excitability of the intracortical inhibitory interneurons. Immediately after exercise, when workload was reduced by half, the resumption of both the psychophysiological variables and the amount of suppression to baseline suggest that the effects of the exercise were not long lasting.
The authors thank Kirill Vaynberg for his assistance during experiments and Dr David Lloyd for his help in producing the model for the experimental setup in Figure 1.
No conflict of interest, financial or otherwise, is declared by the authors. No funding was received for this study.
The results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. Amann M, Blain GM, Proctor LT, Sebranek JJ, Pegelow DF, Dempsey JA. Implications of group III and IV muscle afferents for high-intensity endurance exercise performance in humans. J Physiol
. 2011; 589 (Pt 21): 5299–309.
2. Amann M, Romer LM, Subudhi AW, Pegelow DF, Dempsey JA. Severity of arterial hypoxaemia affects the relative contributions of peripheral muscle fatigue to exercise performance in healthy humans. J Physiol
. 2007; 581: 389–403.
3. Barthelemy D, Nielsen JB. Corticospinal contribution to arm muscle activity during human walking. J Physiol
. 2010; 588 (Pt 6): 967–79.
4. Bentley DJ, Smith PA, Davie AJ, Zhou S. Muscle activation of the knee extensors following high intensity endurance exercise in cyclists. Eur J Appl Physiol
. 2000; 81: 297–302.
5. Bigland-Ritchie B, Furbush F, Woods JJ. Fatigue of intermittent submaximal voluntary contractions: central and peripheral factors. J Appl Physiol
. 1986; 61 (2): 421–9.
6. Butler JE, Larsen TS, Gandevia SC, Petersen NT. The nature of corticospinal paths driving human motoneurons during voluntary contractions. J Physiol
. 2007; 584 (2): 651–9.
7. Butler JE, Taylor JL, Gandevia SC. Responses of human motoneurons to corticospinal stimulation during maximal voluntary contractions and ischemia. J Neuroscience
. 2003; 23 (32): 10224–30.
8. Davey NJ, Romaiguere P, Maskill DW, Ellaway PH. Suppression of voluntary motor activity revealed using transcranial magnetic stimulation of the motor cortex in man. J Physiol
. 1994; 477 (2): 223–35.
9. Di Lazzaro V, Restuccia D, Oliviero A, et al.. Magnetic transcranial stimulation at intensities below active motor threshold activates intracortical inhibitory circuits. Exp Brain Res
. 1998; 119 (2): 265–8.
10. Dorel S, Drouet JM, Couturier A, Champoux Y, Hug F. Changes of pedaling technique and muscle coordination during an exhaustive exercise. Med Sci Sports Exerc
. 2009; 41 (6): 1277–86.
11. Fitts RH. Cellular mechanisms of muscle fatigue. Physiol Rev
. 1994; 74: 49–94.
12. Fuglevand AJ, Winter DA, Patla AE. Models of recruitment and rate coding organisation in motor unit pools. J Neurophysiol
. 1993; 70 (6): 2470–88.
13. Gandevia SC. Spinal and supraspinal factors in human muscle fatigue. Physiol Rev
. 2001; 81 (4): 1725–89.
14. Gandevia SC, Allen GM, Butler JE, Taylor JM. Supraspinal factors in human muscle fatigue: evidence for suboptimal output from the motor cortex. J Physiol
. 1996; 490: 529–36.
15. Garland SJ, Enoka RM, Serrano LP, Robinson GA. Behavior of motor units in human biceps brachii during a submaximal fatiguing contraction. J Appl Physiol
. 1994; 76 (6): 2411–9.
16. Gonzalez-Alonso J. Hyperthermia impairs brain, heart and muscle function in exercising humans. Sports Med
. 2007; 37 (4–5): 371–3.
17. Goodall S, Ross EZ, Romer LM. Effect of graded hypoxia on supraspinal contributions to fatigue with unilateral knee-extensor contractions. J Appl Physiol
. 2010; 109 (6): 1842–51.
18. Hasegawa H, Piacentini MF, Sarre S, Michotte Y, Ishiwata T, Meeusen R. Influence of brain catecholamines on the development of fatigue in exercising rats in the heat. J Physiol
. 2007; 586: 141–9.
19. Hilty L, Lutz K, Maurer K, et al.. Spinal opioid receptor-sensitive muscle afferents contribute to the fatigue-induced increase in intracortical inhibition in healthy humans. Exp Physiol
. 2011; 96 (5): 505–17.
20. Hoffman BW, Oya T, Carroll TJ, Cresswell AG. Increases in corticospinal responsiveness during a sustained submaximal plantar flexion. J Appl Physiol
. 2009; 107 (1): 112–20.
21. Lauber B, Leukel C, Gollhofer A, Taube W. Time to task failure and motor cortical activity depend on the type of feedback in visuomotor tasks. PLoS One
. 2012; 7 (3): e32433.
22. Löscher WN, Cresswell AG, Thorstensson A. Electromyographic responses of the human triceps surae and force tremor during sustained submaximal isometric plantar flexion. Acta Physiol Scand
. 1994; 152 (1): 73–82.
23. Martin PG, Smith JL, Butler JE, Gandevia SC, Taylor JL. Fatigue-sensitive afferents inhibit extensor but not flexor motoneurons in humans. J Neurosci
. 2006; 26 (18): 4796–802.
24. McKay WB, Stokie DS, Sherwood AM, Vrbova G, Dimitrijevic MR. Effect of fatiguing maximal voluntary contraction in excitatory and inhibitory responses elicited by transcranial magnetic motor cortex stimulation. Muscle Nerve
. 1996; 19: 1017–24.
25. McNeil CJ, Martin PG, Gandevia SC, Taylor JL. The response to paired motor cortical stimuli is abolished at a spinal level during human muscle fatigue. J Physiol
. 2009; 587 (Pt 23): 5601–12.
26. Petersen NC, Butler JE, Taylor JL, Gandevia SC. Probing the corticospinal link between the motor cortex and motoneurones: some neglected aspects of human motor cortical function. Acta Physiol (Oxf)
. 2010; 198 (4): 403–16.
27. Petersen NT, Butler JE, Marchand-Pauvert V, et al.. Suppression of EMG activity by transcranial magnetic stimulation in human subjects during walking. J Physiol
. 2001; 537 (2): 651–6.
28. Rasmussen P, Nielsen J, Overgaard M, et al.. Reduced muscle activation during exercise related to brain oxygenation and metabolism in humans. J Physiol
. 2010; 588 (Pt 11): 1985–95.
29. Secher NH, Seifert T, Van Lieshout JJ. Cerebral blood flow and metabolism during exercise: implications for fatigue. J Appl Physiol
. 2008; 104: 306–14.
30. Seifert T, Petersen NC. Changes in presumed motor cortical activity during fatiguing muscle contraction in humans. Acta Physiol (Oxf)
. 2010; 199 (3): 317–26.
31. Sidhu SK, Cresswell AG, Carroll TJ. Motor cortex excitability does not increase during sustained cycling exercise to volitional exhaustion. J Appl Physiol
. 2012; 113 (3): 401–9.
32. Sidhu SK, Hoffman BW, Cresswell AG, Carroll TJ. Corticospinal contributions to lower limb muscle activity during cycling in humans. J Neurophysiol
. 2012; 107 (1): 306–14.
33. Siebner HR, Dressnandt J, Auer C, Conrad B. Continuous intrathecal baclofen infusions induced a marked increase of the transcranially evoked silent period in a patient with generalized dystonia. Muscle Nerve
. 1998; 21 (9): 1209–12.
34. Søgaard K, Gandevia SC, Todd G, Petersen NT, Taylor JL. The effect of sustained low-intensity contractions on supraspinal fatigue in human elbow flexor muscles. J Physiol
. 2006; 573 (2): 511–23.
35. Spriet LL, Soderland K, Bergstrom M, Hultman E. Anaerobic energy release in skeletal muscle during electrical stimulation in men. J Appl Physiol
. 1987; 62 (2): 611–5.
36. Taylor JL, Allen GM, Butler JE, Gandevia SC. Supraspinal fatigue during intermittent maximal voluntary contractions of the human elbow flexors. J Appl Physiol
. 2000; 89: 305–13.
37. Taylor JL, Butler JE, Allen GM, Gandevia SC. Changes in motor cortical excitability during human muscle fatigue. J Physiol
. 1996; 490: 519–28.
38. Taylor JL, Gandevia SC. A comparison of central aspects of fatigue in submaximal and maximal voluntary contractions. J Appl Physiol
. 2008; 104: 542–50.
39. Theurel J, Lepers R. Neuromuscular fatigue is greater following highly variable versus constant intensity endurance cycling. Eur J Appl Physiol
. 2008; 103 (4): 461–8.
40. Zuur AT, Lundbye-Jensen J, Leukel C, et al.. Contribution of afferent feedback and descending drive to human hopping. J Physiol
. 2010; 588 (Pt 5): 799–807.