Share this article on:

Effects of Noradrenaline and Dopamine on Supraspinal Fatigue in Well-Trained Men


Medicine & Science in Sports & Exercise: December 2012 - Volume 44 - Issue 12 - p 2299–2308
doi: 10.1249/MSS.0b013e318265f356
Basic Sciences

Purpose: Prolonged exhaustive exercise induces a failure of the nervous system to activate the involved muscles maximally (i.e., central fatigue). Part of central fatigue may reflect insufficient output from the motor cortex (i.e., supraspinal fatigue), but the cause is unresolved. To investigate the potential link between supraspinal fatigue and changes in brain concentration of dopamine and noradrenaline in temperate environment, we combined neurophysiological methods and pharmacological manipulation of these two neurotransmitters.

Methods: Changes in performance of a cycling exercise (time trial [TT]) were tested after oral administration of placebo (Pla), dopamine, or noradrenaline reuptake inhibitors (methylphenidate and reboxetine [Rebox], respectively) in well-trained male subjects. Changes in voluntary activation, corticospinal excitability, and muscle contractile properties were tested in the knee extensors using transcranial magnetic stimulation and motor nerve electrical stimulation before and after exercise. A psychomotor vigilance task (PVT) was also performed.

Results: Compared with Pla, methylphenidate did not affect exercise performance (P = 0.19), but more time was needed to complete the TT after administration of Rebox (approximately 9%, P < 0.05). For the latter condition, the reduced performance was accompanied by a central/supraspinal fatigue (5%–6%, P < 0.05) and worsened PVT performance (7%, P < 0.05). For the three conditions, corticospinal excitability was unchanged, and peripheral fatigue was similar. Because the ingestion of Rebox induced a greater decrease in voluntary activation and PVT performance after the TT than Pla, with no modification in corticospinal excitability, the noradrenaline reuptake inhibitor likely affected supraspinal circuits located before the motor cortex.

Conclusion: These results suggest that noradrenaline, but not dopamine reuptake inhibition, contributes to the development of central/supraspinal fatigue after a prolonged cycling exercise performed in temperate conditions.

1Laboratory of Applied Biology, Université Libre de Bruxelles, Brussels, BELGIUM; 2Department of Human Physiology and Sports Medicine, Vrije Universiteit Brussel, Brussels, BELGIUM; 3Fonds Wetenschappelijk Onderzoek—Vlaanderen, Research Foundation Flanders (FWO), Brussels, BELGIUM; and 4Viper, Department of Behavioral Sciences, Royal Military Academy, Brussels, BELGIUM

Address for correspondence: Romain Meeusen, Ph.D., FACSM, Department of Human Physiology and Sports Medicine, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium; E-mail:

Submitted for publication March 2012.

Accepted for publication June 2012.

Prolonged exercise or sustained contraction of a muscle group at low intensity induces a failure of the nervous system to activate the muscle at its maximal capacity (18,36,37). The magnitude of voluntary activation is commonly estimated by inducing a supramaximal electrical stimulus to the motor nerve during an isometric maximal voluntary contraction (MVC [40]). When extra force is evoked by the “superimposed” stimulus, voluntary activation is considered as submaximal. A progressive deficit in voluntary activation usually develops during a prolonged exercise or contraction and is classically called central fatigue (9,40). Furthermore, studies applying transcranial magnetic stimulation (TMS) to motor cortical areas during an MVC suggested that when the stimulation evokes an additional torque increment, part of the central fatigue can be due to an insufficient output from the motor cortex (9,17,36–38,40). This deficit is usually called supraspinal or cortical fatigue.

Although the underlying mechanisms are not yet understood, it has been suggested that central fatigue might be related to a change in the synthesis and metabolism of brain monoamines, such as serotonin, dopamine, and noradrenaline (for review, see Ref. [35]). These neurotransmitters play a key role in the modulation of various brain functions such as motivation, arousal, attention, stress responses, and motor control (2). For example, it has been suggested from animal studies that an increased serotonergic activity (29) and the ratio of serotonin to dopamine brain content during exercise (6) may be important for the development of central fatigue. However, several pharmacological studies performed in humans failed to alter the exercise capacity through changes in serotonergic neurotransmission (32), suggesting that the role of serotonin is often overestimated and dependent on drug dosages and receptor that is targeted (25). Other investigations reported that pharmacologically induced inhibition of the reuptake of dopamine and noradrenaline led to changes in performance in both normal (12,33) and high ambient temperatures (30,33,34). However, to our knowledge, no studies have investigated whether the effects of these drugs on the performance of a prolonged exercise are related to changes occurring at central/supraspinal level.

In humans, it is impossible to directly measure the effects of specific brain neurotransmitters on motor cortical excitability and output. Therefore, we pharmacologically manipulated the concentration of dopamine and noradrenaline to investigate the potential effects of these two brain neurotransmitters on central/supraspinal fatigue. Depending on the research field, supraspinal fatigue (38,40), sometimes called brain or cortical fatigue (22), has been reported as an alteration affecting not only the motor cortex (38,40) but also a frontoparietal network involved in cognitive functions (22). These two aspects of fatigue can be investigated by measuring voluntary activation (9,17,38,40) and assessing cognitive performance (22,24) in the same experiment. Accordingly, electrophysiological methods (voluntary activation tested by TMS) and a vigilance task (psychomotor vigilance task (PVT) [22]) were used to explore further the supraspinal component of central fatigue shortly after a prolonged cycling exercise performed in temperate environment. We hypothesized that if there is a supraspinal component in central fatigue, the effects would affect not only the motor output but also the cognitive function.

Back to Top | Article Outline


Back to Top | Article Outline


Ten well-trained cyclists or triathletes (age, 27 ± 6 yr; height, 1.81 ± 0.05 m; mass, 75.7 ± 8.0 kg; maximal power output (Wmax), 367 ± 23 W; V˙O2max, 67 ± 7 mL·min−1·kg−1) participated in this investigation. In the present study, a fatiguing exercise involving most of the lower limb muscles was preferred to an isometric contraction of an isolated muscle group because it induces a much longer depression in voluntary activation (27,36). The time trial (TT) protocol applied (see description in the succeeding part of this article) has been shown to be much more reliable and reproducible than a constant load exercise sustained to exhaustion, especially in well-trained subjects (14). Therefore, the subjects selected for this experiment had an experience of at least 5 yr of regular cycling training and competition that ensured a high degree of reproducibility of performance across the different trials (14,32–34). Because of potential issues with the menstrual cycle and hormonal variations in female subjects, experiments were only performed on a male population.

Before the beginning of the study, all volunteers received written information regarding the nature and purpose of the experimental protocol. After an opportunity to ask questions, a written statement of consent was signed. The experimental protocol was approved by the Ethics Committee of the Vrije Universiteit Brussel (Brussels, Belgium), and experiments were performed in accordance with the Declaration of Helsinki.

Back to Top | Article Outline

Cycling exercise.

Before the experimental sessions, all subjects completed a maximal exercise test that consisted of a continuous incremental cycling exercise. Subjects started the maximal exercise test at a resistance of 80 W; every 3 min, 40 W was added until volitional exhaustion, i.e., the moment that subjects could not maintain 80 rpm. This test was used to determine the power output required to elicit 55% and 75% Wmax and V˙O2max. The experimental cycling protocol was identical with the one used previously (33,34). Subjects completed a familiarization trial, followed by three experimental trials, all separated by at least 5 d to ensure full recovery. The first part of each trial consisted of 60 min of fixed intensity exercise at a workload corresponding to 55% Wmax and was followed by a TT to measure performance. There was a 1- to 2-min delay between the end of the constant load exercise and the beginning of the TT to program the ergometer (Lode Excalibur Sport, Groningen, The Netherlands). The TT required the subjects to complete a predetermined amount of work equal to 30 min at 75% Wmax as quickly as possible (14). Subjects began the TT at a workload corresponding to 75% Wmax but were free to increase or decrease their power output as desired from the outset. During the TT, a computer program displayed a bar indicating the percentage of total work completed to give the subject an indication of his progress. Throughout the protocol, no feedback was provided regarding the time lapsed, power output, pedal cadence, or heart rate. During exercise, subjects had ad libitum access to plain water. RPE (scale from 6 to 20 [4]) and heart rate were assessed during the TT. Core temperature was measured before and during the cycling exercise.

Back to Top | Article Outline


The subjects were randomly assigned to one of three drug treatments. The first treatment consisted of 8 mg of reboxetine (Rebox, noradrenaline reuptake inhibitor) on the night before the experiment. On arrival in the laboratory, subjects ingested another 8 mg of Rebox. In the second condition, subjects ingested a placebo (Pla) (lactose, 20 mg) on the night before the experiment and 40 mg of methylphenidate (Mph, dopamine reuptake inhibitor) on arrival in the laboratory. Mph is an amphetamine-like stimulant that is widely administered to treat children with attention deficit hyperactivity disorder (31). It is currently on the doping list of the World Anti-Doping Agency, indicating athletes can only use it under the condition of therapeutic use exemption. The third treatment was the Pla; both the night before and the day of the experimental trial, subjects ingested 20 mg of lactose. The treatment was randomized and administered in a double-blind cross-over manner. Rebox, Mph, and Pla capsules were prepared by an independent pharmacy to appear indistinguishable concerning dimensions and color.

Back to Top | Article Outline

Ergometric device.

The neuromuscular parameters were recorded while the subjects sat on an adjustable chair such that the hip and knee joints were at 100° (angle between trunk and thigh) and 80° (full extension = 0°), respectively. The chair had a long backrest to provide full back support. The head of the subjects was secured in a custom-made headrest to ensure a stable position during the experiment. The right leg of the subjects was connected to a force transducer placed approximately 2 cm above the lateral malleolus using velcro straps. To minimize the movement of the trunk during contractions, the subject was attached to the chair by a racing harness. The position of the headrest and force transducer as well as the harness adjustments were recorded for each subject to ensure identical position in each experimental session.

Back to Top | Article Outline

Torque and electromyographic recordings.

The isometric torque developed by the knee extensor muscles was measured by a force transducer (linear range, 0–2500 N; U2000 load cell; Maywood Instruments Ltd, Basingstoke, UK) connected to the leg and the chair by a rigid system. The signal from the force transducer was amplified by a custom-made amplifier. Voluntary and electrically evoked EMG activities (aEMGs) were recorded from the rectus femoris, the vastus medialis, and the long head of the biceps femoris using self-adhesive bipolar electrodes (Ag–AgCl, 10-mm diameter, interelectrode distance approximately 5 cm). The EMG electrodes were placed over the muscle belly of each muscle, approximately 10 cm proximal to the superior border of the patella for the rectus femoris and on the midpart of the muscle belly, determined by manual testing, for the vastus medialis and the long head of the biceps femoris. The ground electrodes were located over the lateral condyle of the tibia. The position of the electrodes was marked with indelible ink to ensure a similar location on the different visits. All EMG signals were amplified (×1000) and filtered (10 to 1 kHz) by a custom-made differential amplifier. The EMG and torque signals were acquired on a personal computer at a sampling rate of 2 kHz with a data acquisition system and analyzed offline with AcqKnowledge analysis software (model MP 100; BIOPAC Systems, Santa Barbara, CA).

The potential effect of EMG crosstalk between knee extensor muscles and biceps femoris was evaluated in four subjects by recording the aEMG of biceps femoris, vastus medialis, and rectus femoris muscles during a maximal electrical stimulation of the latter one at its motor point. To quantify crosstalk, the ratio between the amplitude of the signal recorded over biceps femoris or vastus medialis and the stimulated muscle was then computed. The EMG amplitudes of the biceps femoris and vastus medialis were ≤4% and 6%, respectively, of the rectus femoris amplitude, indicating that crosstalk between synergist and antagonist muscles was negligible in our experimental conditions.

Back to Top | Article Outline

Electrical stimulation.

The knee extensors were stimulated by single supramaximal electrical pulses (0.2-ms duration) delivered by a constant current stimulator (DS7A; Digitimer, Welwyn Garden City, UK) that was triggered by a digital timer (Master-8; AMPI, Jerusalem, Israel). The stimuli were delivered to the femoral nerve through two self-adhesive electrodes (Ag–AgCl, 10-mm diameter). The cathode was positioned over the nerve in the femoral triangle, and the anode was placed midway between the greater trochanter and the iliac crest (36). The precise location of the cathode was determined at rest by weak electrical stimulation, whereas the maximal stimulus intensity was adjusted during submaximal contractions (approximately 50% MVC). The level of stimulation was then set 30% above that required to induce an Mmax during the submaximal contractions.

Back to Top | Article Outline


A double-cone coil (130-mm outer diameter) was positioned over the cortex to elicit motor-evoked potentials (MEPs) in the right knee extensors with a Magstim 200 stimulator (Magstim, Dyfed, UK). The junction of the double-cone coil was positioned 1–2 cm to the left of the vertex. The direction of current flow in the coil preferentially activated the left motor cortex. At the beginning of each experiment, the coil position was determined to stimulate optimally the rectus femoris and vastus medialis muscles, and this position was marked on the scalp. The stimulator output (30%–60% of maximum) was set to evoke large MEP in both rectus femoris and vastus medialis (>80% of Mmax area at 50% MVC), minimal MEP in the biceps femoris, and the greatest superimposed twitch during the different target torques used in the protocol (see the succeeding part). The custom-made headrest ensured a stable location of the coil on the scalp during the experiment.

Back to Top | Article Outline


The psychomotor vigilance test was used as a sustained attention task in the present study (22). For a detailed description of the task, readers are referred to the original study by Dinges and Powell (8). Briefly, a laptop computer displayed a visual stimulus (fireball, always in the middle of the screen) at a variable interval (between 2 and 10 s). Subjects were required to respond by pressing the space bar with the index of the dominant hand. The distance between the subject’s eyes and the screen was approximately 40 cm. Subject’s reaction time was displayed on a computer screen for 1 s after space bar pressing. If the subject did not respond within 500 ms, the trial was stored as a lapse, and a warning message appeared to the participant.

The PVT was performed once before and once 15 min after the end of the cycling exercise when the subject was sitting on the chair used for the recording of the neuromuscular parameters (see description given previously). The PVT ran for a total period of 10 min (100 stimuli). The reaction time of correct responses (<500 ms) and number of lapses were stored.

Back to Top | Article Outline

Experimental protocol.

As already mentioned, all subjects participated in a familiarization session and three experimental sessions (Pla, Rebox, and Mph). The familiarization trial was undertaken to ensure that subjects were accustomed to the procedures used during the investigation and to minimize any potential learning or anxiety effects. During this session, the subjects practiced the isometric contractions and experienced the electrical motor nerve stimulation and TMS. The familiarization trial was identical with the experimental trials in all respects.

Experimental trials were undertaken in temperate (18°C) conditions with relative humidity maintained between 50% and 60%. Subjects were instructed to record dietary intake and physical activity during the 2 d before the first trial and to replicate this in the 2 d before the subsequent experimental trials. No exercise, caffeine, or alcohol consumption was permitted in the 24 h before each trial. For each subject, the four sessions were performed at the same time of the day, either at the end of the morning or in the afternoon. For all trials, subjects were dressed in cycling shorts, socks, and shoes.

On arrival in the laboratory, subjects inserted a rectal thermister (Gram Corporation LT-8A, Saitama, Japan) 10 cm beyond their anal sphincter for the measurement of core temperature and heart rate was recorded telemetrically by a cardio frequency meter (Polar Accurex Plus, Kempele, Finland). They were then placed in the experimental chair to record the neuromuscular parameters. After a warm-up period consisting in a few submaximal contractions, subjects performed two to three brief MVCs (approximately 2–3 s) to determine their baseline maximal torque. The torque of the greatest MVC was then measured, and two submaximal target torques (50% and 75% MVC) were determined and displayed on a computer screen used as visual feedback. Thereafter, preexercise recordings were started. Voluntary activation was tested by TMS and motor nerve stimulation. To that end, subjects performed sets of brief contractions. Each set was composed of a brief MVC followed by contractions at 50% and 75% MVC and ended by an MVC. TMS was delivered during the first three contractions, whereas motor nerve stimulation was delivered during the last MVC. Subjects were asked to maintain the required contraction level during and immediately after TMS or motor nerve stimulation. The resting potentiated twitch was recorded in response to motor nerve stimuli delivered 3–5 s after the last MVC (Fig. 1). The order of the contractions at the different intensities (50%, 75%, and 100% MVC) and of the two stimulation types (TMS and electrical stimulation) was randomized across subjects but was kept constant for each subject across experimental sessions. For the preexercise recordings, three sets of contractions were at least performed (Fig. 1). In each set, the contractions were separated by 10 s of rest, and sets were separated from each other by 4 min of rest. The TMS protocol, originally introduced by Todd et al. (40), has been shown to produce reliable estimations of resting twitch and voluntary activation in the knee extensor muscles (11,36).

After the cycling effort, two sets of contractions and stimulations similar to the preexercise condition were performed at 10, 30, and 45 min of recovery (R10, R30, and R45, respectively; Fig. 1). For all contraction levels and stimulation types, the recording was repeated when the torque level did not remain constant during or just after the time at which the stimulation was delivered. The PVT was performed 15 min after the end of the TT.

Back to Top | Article Outline

Data analysis.

The duration of the TT and the time course of the power output during the TT were measured. For the trial that produced the greatest MVC torque, the maximum torque and associated average aEMG of rectus femoris, vastus medialis, and biceps femoris, recorded before and after the cycling exercise, were determined for a 500-ms period during the MVC torque plateau. The peak-to-peak amplitude and area of the MEP and the Mmax recorded during the contractions at different intensities were also measured. Because neither MEP amplitude nor MEP area changed after the cycling exercise, only areas are reported. To avoid the influence of activity-dependent changes in the muscle fiber membrane during the fatiguing contraction and to assess the corticospinal adjustments, MEP area was normalized, at each time point, to the corresponding Mmax area recorded during the MVC. The MEP latency and duration of the silent period (SP) after TMS were measured in both the rectus femoris and vastus medialis. Latency was defined as the time between the stimulus artifact and the first major deflection in the EMG signal. The SP duration was determined in the contracting muscle as the interval from the stimulus artifact to the return of continuous aEMG. The superimposed twitch amplitude, recorded in response to TMS or motor nerve stimulation, was measured as the difference between the superimposed peak torque minus the MVC torque. Peak torque, time-to-peak torque, and half-relaxation time of the resting twitch, in response to motor nerve stimulation, were also measured. For each subject and time point, parameters recorded from multiple sets were averaged. The amplitude of the resting twitch evoked by TMS was estimated rather than measured directly (40). For each subject, a linear regression of the amplitude of the superimposed twitch evoked by TMS against voluntary torque was performed for the 50%, 75%, and 100% MVC contractions. The y-intercept was taken as the amplitude of the estimated resting twitch (ERT [40]). For the two techniques, voluntary activation level (% of maximum) was calculated according to the following equation: (1 − superimposed twitch/resting twitch (or ERT)) ×100 (36,40). For the PVT, reaction time and lapses averages before and 15 min after the cycling exercise were compared for each subject.

Back to Top | Article Outline

Statistical analysis.

Before comparing each dependent variable, the normality of the data was confirmed with the Kolmogorov–Smirnov test. To evaluate differences in time necessary to complete the TT and in mean power output during the TT in the three conditions (Rebox, Mph, and Pla), a one-way ANOVA with repeated measures was used. Data collected over time before, during, and after the cycling exercise were analyzed using a two-factor (drug × time) ANOVA with repeated measures. When a significant main effect was found, a Tukey post hoc test was used to identify the significant differences among the selected means. The relationship between the subjects’ change in MVC torque and voluntary activation tested by TMS 10 min after the TT was fitted by a linear regression using the least-squares method. For all comparisons, statistical significance was accepted at P < 0.05. Data are reported as means ± SD within the text and means ± SEM in the figures.

Back to Top | Article Outline


Back to Top | Article Outline

TT performance.

All subjects completed the three experimental trials. They took 9.4% more time to complete the TT in the Rebox condition (33.66 ± 3.61 min) compared with the Pla condition (30.78 ± 2.08 min; P < 0.05). Despite a 1-min or 3.4% faster TT performance, the time needed to complete the TT was not significantly different between Mph (29.73 ± 1.40 min) and Pla trials (P = 0.19). The mean power output generated during the TT was significantly (P < 0.05) lower in the Rebox trial compared with the Pla trial (252 ± 31 vs 271 ± 26 W), whereas no difference (P = 0.18) was observed between Mph (279 ± 28 W) and Pla trials. In Pla and Mph conditions, the power output remained constant throughout the TT, whereas it decreased progressively for the Rebox trial (ANOVA drug × time interaction, P < 0.01; Fig. 2A).

The subjects’ RPE at the end of the TT was similar between Pla (18.0 ± 1.4), Rebox (18.1 ± 1.1), and Mph (18.7 ± 1.1) conditions (P = 0.18). Heart rate was higher for Mph (185.4 ± 7.3 bpm) than for Pla (174.4 ± 9.2 bpm, P = 0.05) at the end of the TT, but there was no significant effect for Rebox (168.8 ± 16.6 bpm, P = 0.40). At the completion of the TT, core temperature was significantly greater (P < 0.001) compared with preexercise values for all conditions (37.1°C ± 0.2°C, 37.1°C ± 0.3°C, and 37.3°C ± 0.2°C for Pla, Rebox, and Mph, respectively). It was slightly greater after Mph administration (39.4°C ± 0.5°C, P < 0.05) but was not significantly influenced by Rebox (38.8°C ± 0.3°C, P = 0.64) when compared with Pla (38.9°C ± 0.2°C). Core temperature remained significantly elevated 10 min after the end of the TT for Pla (38.5°C ± 0.5°C), Rebox (38.4°C ± 0.4°C), and Mph (38.9°C ± 0.5°C; post hoc tests, P < 0.001) and did not return to control values until the end of the recovery period (post hoc tests, P < 0.01).

Back to Top | Article Outline

MVC torque and aEMG.

Before the cycling exercise, the average MVC torque was similar (P = 0.87) for Pla, Rebox, and Mph trials and corresponded to 233.3 ± 65.5, 236.0 ± 73.3, and 232.8 ± 67.4 N·m, respectively. When recorded 10 min after the end of the TT, the MVC torque was significantly reduced for the three conditions (post hoc tests, P < 0.05; Fig. 2B). This reduction tended to be slightly greater for Rebox as compared with Pla and Mph trials (P = 0.15, Fig. 2B). For the three conditions, the deficit in MVC torque was still present after 45 min of recovery (post hoc tests, P < 0.05).

As for the MVC torque, the corresponding aEMG for the agonist and antagonist muscles was also similar in preexercise conditions regardless of the drug administered (P > 0.37). The aEMG values in Pla, Rebox, and Mph conditions were 709.6 ± 234.1, 716.1 ± 245.4, and 693.2 ± 299.0 µV, respectively, for rectus femoris, 778.6 ± 448.6, 771.9 ± 369.1, and 753.2 ± 366.1 µV, respectively, for vastus medialis and 149.2 ± 60.0, 126.0 ± 49.5, and 159.2 ± 64.9 µV, respectively, for biceps femoris. There was no statistically significant interaction between drug and time for the aEMG change after the cycling exercise (P = 0.19). The aEMG of the agonist muscles averaged across the recovery period decreased slightly for the three conditions (91.7% ± 8.2%, 88.8% ± 9.1%, 95.9% ± 9.2% of control values for Pla, Rebox, and Mph, respectively). However, this decrease was significant for Pla and Rebox conditions (P < 0.05) but not for the Mph condition (P = 0.28). A similar tendency was observed for the aEMG of biceps femoris (93.8% ± 18.1%, 89.2% ± 24.6%, 94.1% ± 25.5% of control values for Pla, Rebox and Mph, respectively), but the reduction did not reach statistical significance (post hoc tests, P > 0.18).

Back to Top | Article Outline

MEP parameters.

The MEP area (%Mmax) recorded in rectus femoris and vastus medialis during the MVC before the cycling exercise was similar in the three conditions (P = 0.25, Table 1). Regardless of the drug administered, it was unchanged in both rectus femoris (ANOVA time effect, P = 0.78) and vastus medialis (ANOVA time effect, P = 0.81) when measured 10 min after the end of the TT. The MEP recorded from the biceps femoris was very small compared with that from the knee extensors and similar for the three conditions (9.8% ± 4.2%, 9.6% ± 5.5%, 9.3% ± 3.3% of knee extensors MEP values for Pla, Rebox and Mph, respectively; P = 0.95). It did not change after the cycling exercise (ANOVA time effect, P = 0.56).

The MEP latency and SP duration, measured in rectus femoris and vastus medialis, were similar during preexercise recordings for the three conditions (P > 0.8) and did not change after the cycling exercise (ANOVA time effect, P > 0.25; Table 1). Regardless of the drug administered, the Mmax area recorded during the MVC was not statistically different before exercise for rectus femoris (P = 0.32) and vastus medialis (P = 0.11), and it was unchanged when measured 10 min after the cycling exercise compared with control values (ANOVA time effect, P = 0.53 and P = 0.49, respectively; Table 1).

Back to Top | Article Outline

Voluntary activation.

During preexercise recordings, voluntary activation measured with TMS was not statistically different for the three conditions and represented 96.9% ± 3.0%, 94.3% ± 8.2%, and 96.6 ± 2.8% of maximal activation for Pla, Rebox, and Mph, respectively (P = 0.34, Fig. 3C). There was a significant ANOVA interaction between drug and time (P < 0.05), and post hoc tests indicated that voluntary activation was significantly reduced in the Rebox condition (5.4%, P < 0.001) but unchanged for Pla and Mph conditions (P = 0.7 and P = 0.9, respectively) when measured 10 min after the end of the cycling exercise (Fig. 3C). For the Rebox trial, this reduction persisted until 45 min of recovery (P = 0.2).

A significant linear association (y = 0.83x + 16.56, r2 = 0.29) between subjects’ voluntary activation tested by TMS and change in MVC torque 10 min after the end of the TT was found when data from the three conditions were plotted together (P < 0.01).

During preexercise recordings, voluntary activation tested by motor nerve stimulation was similar (P = 0.96) for Pla, Rebox, and Mph trials (94.2% ± 2.6%, 94.4% ± 3.9%, and 94.6% ± 5.0%, respectively; Fig. 3D). For the three conditions, it was slightly reduced 10 min after the end of the cycling exercise, but again, the difference was only statistically significant for the Rebox condition (4.7%, P < 0.05; Fig. 3D). Although still present after 30 min of recovery, the deficit in voluntary activation was no more statistically different (P = 0.07) from control value.

Back to Top | Article Outline

Twitch contractile properties.

The peak torque of the twitch induced by motor nerve stimulation at rest did not differ statistically between the three conditions during preexercise recordings (P = 0.11, Table 2). Compared with control values, it was significantly decreased shortly after the TT in the three conditions (post hoc tests, P < 0.05; Table 2). There was, however, no significant ANOVA interaction between drug and time (P = 0.4). Twitch torque was still significantly lower than control values after 45 min of recovery (post hoc tests, P < 0.05). As for peak torque, the twitch time-to-peak torque and half-relaxation time were similar for the three conditions during preexercise recordings (P = 0.18 and 0.73, respectively; Table 2). Ten minutes after the end of the TT, all these values were reduced for Pla, Rebox, and Mph (post hoc tests, P < 0.05; Table 2). The twitch time-to-peak torque recovered during the following resting period and did not differ statistically from control value after 30 min of recovery for Mph (P = 0.21) and after 45 min for Pla (P = 0.07) and Rebox (P = 0.63). Half-relaxation remained below the initial values at the end of the recovery period for Pla (P < 0.01) and Mph (P < 0.05) but not for Rebox (P = 0.27).

As for the electrically induced twitch, the ERT was similar before exercise in the three conditions (34.5 ± 14.6, 36.9 ± 18.7, and 37.0 ± 22.4 N·m for Pla, Rebox, and Mph, respectively; P = 0.6) and was also similarly reduced (post hoc tests, P < 0.05) at the end of the TT in the three conditions (70.5% ± 19.9%, 65.5% ± 23.0%, and 69.9 ± 21.7% of control values for Pla, Rebox, and Mph, respectively). There was no ANOVA interaction between drug and time (P = 0.78). ERT was still significantly lower than control values after 45 min of recovery (post hoc tests, P < 0.05).

Back to Top | Article Outline


The control values for the reaction time and number of lapses were similar regardless of the drug administered (P = 0.73 and 0.40 for the reaction time and number of lapses, respectively; Table 3). For the reaction time, the ANOVA showed a significant interaction between drug and time (P < 0.05). It increased significantly for the Rebox condition (P < 0.01), did not change for the Pla trial (P = 0.97), and decreased for the Mph trial (P < 0.01, Table 3). In contrast, the number of lapses did not change significantly after the cycling exercise for any condition (ANOVA time effect, P = 0.48; Table 3).

Back to Top | Article Outline


The objective of the current study was to investigate the potential link between neural fatigue and changes in brain concentration of noradrenaline and dopamine by combining neurophysiogical methods and pharmacological manipulation of these two neurotransmitters in a temperate environment. To that purpose, we tested the effects of oral administration of noradrenaline and dopamine reuptake inhibitors on acute cycling performance and central/supraspinal fatigue. The main novel finding is that the decrease in cycling performance induced by the noradrenaline reuptake inhibition was accompanied by a greater central/supraspinal fatigue and a worsened PVT performance. In contrast, the dopamine reuptake inhibition did not influence the TT performance and central/supraspinal fatigue differently than the Pla in the current environmental conditions.

Back to Top | Article Outline

TT performance, MVC torque, and twitch contractile properties.

TT performance was decreased in Rebox condition, whereas Mph had no significant influence on performance. These results are in agreement with previous studies performed in our laboratory during similar cycling exercise and environmental conditions (33,34). Whereas the TT performance was not influenced by Mph in normal ambient temperature (18°C), it was improved by 16% in the heat (30°C [34]). The lack of effect of Mph in temperate conditions was confirmed, even though the dose of Mph has been doubled (40 vs 20 mg) compared with our previous investigation (34).

The decline in MVC torque measured 10 min after the end of the TT was lower than what is usually reported for moderately trained subjects after similar cycling efforts (18,36) but comparable to well-trained cyclists or triathletes (20,28). The mean decrease in MVC torque was slightly greater for Rebox than Pla trial. However, this difference did not reach statistical significance. This observation may be partly explained by the modest fatigue-related decline in MVC torque in the three conditions and the relatively high intersubject variability. However, when individual data of the three conditions were plotted together, a significant linear association (r2 = 0.29, P < 0.01) was observed between subjects’ voluntary activation tested by TMS 10 min after the TT and changes in MVC torque. This observation suggests that the level of voluntary activation measured shortly after the TT, which is influenced by the drug administered, contributed partly to reduce MVC torque. The relatively greater reduction in TT performance compared with the loss of MVC torque under Rebox condition could be due to the delay required (10 min) to replace very precisely the subject on the experimental set-up used to test the electrophysiological parameters after the end of the cycling exercise. During such period, some recovery may have occurred. Furthermore, the contribution of neural factors (see the succeeding part) to the alteration in performance may be more pronounced for a multijoints task such as cycling than for a single-joint isometric MVC that requires less complex muscular synergy.

The decrease of the electrically induced resting twitch and ERT observed in our study indicates that impairment of the excitation–contraction coupling contributes to the fatigue observed after prolonged cycling (18,20,36). The similar decrease in twitch torque and time course for the three conditions suggests that peripheral fatigue induced by the cycling exercise was not worsened by the noradrenaline reuptake inhibitor. Consequently, the negative influence of Rebox on TT performance should be mainly due to neural impairments. The shortened twitch time course (time-to-peak torque and half-relaxation time) was most probably related to the increase in core temperature (5) and the relative interaction between fatigue and postactivation potentiation (10).

Back to Top | Article Outline

Neural contribution to muscle fatigue.

Preexercise voluntary activation values measured by supramaximal electrical stimulation of the motor nerve and TMS were similar in the three conditions. The cycling effort did not significantly reduce the voluntary activation level tested by electrical stimulation for Pla and Mph trials. The absence of central fatigue in the Pla condition is surprising because, contrary to sustained isometric fatiguing contractions of an isolated muscle group after which voluntary activation usually recovers within the first minutes (15), a long-lasting central and supraspinal fatigue has been reported after prolonged exhaustive exercises (27,36). Our results may however be explained by the high training level of our subjects because it has been reported that well-trained athletes did not display significant central fatigue when tested within a 30-min period after prolonged cycling exercise (20,28).

In contrast to the Pla trial, a significant central fatigue was observed in the Rebox condition. Indeed, we observed a similar deficit in voluntary activation when tested 10 min after the TT by TMS and motor nerve stimulation. Although both TMS and motor nerve stimulation measure the failure of the nervous system to drive the muscle maximally during fatigue, TMS probes more specifically the supraspinal component of central fatigue (36,38,40). A direct quantitative comparison of the two techniques is not straightforward because they can activate synergistic muscles differently and the shapes of the voluntary activation–force relationships for motor cortical and motor nerve stimulation differ slightly (38,40); however, the reduction in voluntary activation tested by TMS indicates that a long-lasting (30 min) supraspinal fatigue was present after the cycling exercise. The presence of a deficit in activation due to supraspinal impairment shortly after the TT in the Rebox condition suggests that noradrenaline is an important neurotransmitter in the modulation of central/supraspinal fatigue under the present environmental conditions. The absence of significant effect of Mph on central/supraspinal fatigue in our experimental conditions does not exclude, however, a role for dopamine in central fatigue in high ambient temperature (30,34). Previous studies have indeed shown that mechanisms involved in central fatigue might be temperature dependent (30). Together, this hypothesis and the current results indicate that future research should focus on the effects of the different neurotransmitter systems in high ambient temperature that would increase the thermal stress to a much greater extent than that in the current study.

Back to Top | Article Outline

Location of neural impairments and possible mechanisms.

Because the normalized MEP area recorded during the MVC did not change after the cycling exercise, the responsiveness of neurons in the pathway from the motor cortex to the muscle was presumably not impaired (21,36). Similarly, because the SP duration was not affected for the three conditions when tested shortly after the cycling exercise, a change in intracortical (13) and/or motor neuronal inhibition is unlikely (38).

The PVT is considered as a reliable method to measure “cognitive fatigue,” defined as “vigilance decrement,” or “time on-task effect” (24), consequent, for example, to sleep deprivation or during taxing cognitive tasks (22). Our observation of a greater PVT performance after exercise in the Mph condition is consistent with studies showing that this drug can improve performance in vigilance tasks in healthy subjects and patients with attention deficit disorder (31,39). In contrast, Rebox had a negative effect on PVT performance after exercise. The changes in reaction time observed after the end of TT in Rebox and Mph conditions cannot be ascribed to a change in motor time because MEP latency and twitch time-to-peak torque displayed a similar time course in the three conditions. Because Rebox in resting conditions has no significant effect on reaction time during psychometric tests in healthy subjects (16), it is suggested that the noradrenaline reuptake inhibitor and exercise interact at the supraspinal level and affect brain centers located before the motor cortex.

Animal studies have shown that the locus coeruleus, which contains the neurons synthesizing noradrenaline, provides numerous projections to different parts of the CNS involved in behavioral, attention, arousal, and vigilance regulation (for a review, see Ref. [3]). Recent works indicate that noradrenaline contributes to maintain forebrain neuronal activity states appropriate for sensory information integration (e.g., waking) and modulates the collection and processing of salient sensory information (3) through its projections to cortical and subcortical sensorimotor, attention, and memory circuits (3,19). In addition, iontophoretically applied noradrenaline in anesthetized animals can enhance both excitatory and inhibitory responses of somatosensory cortical neurons to afferent synaptic input (42). This last effect was selective for noradrenaline because dopamine does not exhibit a similar property (41).

The noradrenaline reuptake inhibitor we administered may thus have affected the integration of sensory information and/or motor cortical sensitivity to these sensory afferents and consequently altered the generated motor output (1). This discussion is consistent with the observation that PVT activates a frontoparietal network involved in attention, motivation, sensory information integration and motor programming (22) that may have been affected by the ingestion of the noradrenaline reuptake inhibitor. However, we cannot completely rule out that Rebox has influenced other neurotransmitter systems (26) and that they have contributed to the effect of Rebox on central fatigue. Although Rebox is known to have a much higher potency for noradrenaline compared with dopamine and serotonin (26), in brain areas where noradrenergic terminals are in abundance (such as the medial prefrontal cortex), elevated dopamine concentrations occur because of the inhibition of dopamine reuptake by the noradrenaline transporters (23). Thus, dopamine and noradrenaline are intrinsically linked via chemical pathways, in that hydroxylation of the former yields the latter (7).

In conclusion, the alteration of the cycling performance due to the ingestion of a noradrenaline reuptake inhibitor was associated with a greater decrease in voluntary activation and PVT performance than that in the Pla condition in well-trained male subjects. In contrast, dopamine reuptake inhibition had no additional effect than Pla in temperate conditions. These original findings suggest that noradrenaline, but not dopamine reuptake inhibition, contributes to the development of supraspinal fatigue observed after a prolonged cycling exercise performed in temperate environment. Because corticospinal excitability was not modified after the cycling exercise, noradrenaline appears to affect more specifically the supraspinal centers located before the motor cortex. As a consequence, the output from the motor cortex becomes insufficient to drive the muscles optimally and thereby likely contributes to affect motor performance.

Both authors contributed equally to this article. This study was supported by a grant from the Fonds National de la Recherche Scientifique (FRS-FNRS, FC82578) (M. Klass), the Fund for Scientific Research Flanders (FWO) (B. Roelands), a research funding from the Vrije Universiteit Brussel (1235), and the Research Council of the Université Libre de Bruxelles. Malgorzata Klass and Bart Roelands contributed equally to this work. The authors have no professional relationships with companies or manufacturers. The authors report no conflict of interest. The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Amann M, Proctor LT, Sebranek JJ, Pegelow DF, Dempsey JA. Opioid-mediated muscle afferents inhibit central motor drive and limit peripheral muscle fatigue development in humans. J Physiol. 2009; 587 (Pt 1): 271–83.
2. Benarroch EE. The locus ceruleus norepinephrine system: functional organization and potential clinical significance. Neurology. 2009; 73 (20): 1699–704.
3. Berridge CW, Waterhouse BD. The locus coeruleus-noradrenergic system: modulation of behavioral state and state-dependent cognitive processes. Brain Res Rev. 2003; 42 (1): 33–84.
4. Borg GA. Psychophysical bases of perceived exertion. Med Sci Sports Exerc. 1982; 14 (5): 377–81.
5. Davies CT, Young K. Effect of temperature on the contractile properties and muscle power of triceps surae in humans. J Appl Physiol. 1983; 55 (1 Pt 1): 191–5.
6. Davis JM, Bailey SP. Possible mechanisms of central nervous system fatigue during exercise. Med Sci Sports Exerc. 1997; 29 (1): 45–57.
7. Del Campo N, Chamberlain SR, Sahakian BJ, Robbins TW. The roles of dopamine and noradrenaline in the pathophysiology and treatment of attention-deficit/hyperactivity disorder. Biol Psychiatry. 2011; 69 (12): e145–e57.
8. Dinges DF, Powell JW. Microcomputer analyses of performance on a portable, simple visual RT task during sustained operations. Behav Res Methods. 1985; 17 (6): 652–5.
9. Gandevia SC, Allen GM, Butler JE, Taylor JL. Supraspinal factors in human muscle fatigue: evidence for suboptimal output from the motor cortex. J Physiol. 1996; 490 (Pt 2): 529–36.
10. Garner SH, Hicks AL, McComas AJ. Prolongation of twitch potentiating mechanism throughout muscle fatigue and recovery. Exp Neurol. 1989; 103 (3): 277–81.
11. Goodall S, Romer LM, Ross EZ. Voluntary activation of human knee extensors measured using transcranial magnetic stimulation. Exp Physiol. 2009; 94 (9): 995–1004.
12. Heyes MP, Garnett ES, Coates G. Central dopaminergic activity influences rats ability to exercise. Life Sci. 1985; 36: 671–7.
13. Inghilleri M, Berardelli A, Cruccu G, Manfredi M. Silent period evoked by transcranial stimulation of the human cortex and cervicomedullary junction. J Physiol. 1993; 466: 521–34.
14. Jeukendrup A, Saris WH, Brouns F, Kester AD. A new validated endurance performance test. Med Sci Sports Exerc. 1996; 28 (2): 266–70.
15. Keller ML, Pruse J, Yoon T, Schlinder-Delap B, Harkins A, Hunter SK. Supraspinal fatigue is similar in men and women for a low-force fatiguing contraction. Med Sci Sports Exerc. 2011; 43 (10): 1873–83.
16. Kerr JS, Powell J, Hindmarch I. The effects of reboxetine and amitriptyline, with and without alcohol on cognitive function and psychomotor performance. Br J Clin Pharmacol. 1996; 42 (2): 239–41.
17. Klass M, Lévénez M, Enoka RM, Duchateau J. Spinal mechanisms contribute to differences in the time to failure of submaximal fatiguing contractions performed with different loads. J Neurophysiol. 2008; 99 (3): 1096–104.
18. Kremenic IJ, Glace BW, Ben-Avi SS, Nicholas SJ, McHugh MP. Central fatigue after cycling evaluated using peripheral magnetic stimulation. Med Sci Sports Exerc. 2009; 41 (7): 1461–6.
19. Lee S-B, Beak SK, Park SH, Waterhouse BD, Lee HS. Collateral projection from the locus coeruleus to whisker-related sensory and motor brain regions of the rat. J Comp Neurol. 2009; 514 (4): 387–402.
20. Lepers R, Maffiuletti NA, Rochette L, Brugniaux J, Millet GY. Neuromuscular fatigue during a long-duration cycling exercise. J Appl Physiol. 2002; 92 (4): 1487–93.
21. Lévénez M, Garland SJ, Klass M, Duchateau J. Cortical and spinal modulation of antagonist coactivation during a submaximal fatiguing contraction in humans. J Neurophysiol. 2008; 99 (2): 554–63.
22. Lim J, Wu W-C, Wang J, Detre JA, Dinges DF, Rao H. Imaging brain fatigue from sustained mental workload: an ASL perfusion study of the time-on-task effect. Neuroimage. 2010; 49 (4): 3426–35.
23. Linnér L, Endersz H, Ohman D, Bengtsson F, Schalling M, Svensson TH. Reboxetine modulates the firing pattern of dopamine cells in the ventral tegmental area and selectively increases dopamine availability in the prefrontal cortex. J Pharmacol Exp Ther. 2001; 297 (2): 540–6.
24. Mackworth JF. Vigilance, arousal, and habituation. Psychol Rev. 1968; 75 (4): 308–22.
25. Meeusen R, Watson P, Hasegawa H, Roelands B, Piacentini MF. Central fatigue: the serotonin hypothesis and beyond. Sports Med. 2006; 36 (10): 881–909.
26. Miller DK, Wong EHF, Chesnut MD, Dwoskin LP. Reboxetine: functional inhibition of monoamine transporters and nicotinic acetylcholine receptors. J Pharmacol Exp Ther. 2002; 302 (2): 687–95.
27. Millet GY, Lepers R. Alterations of neuromuscular function after prolonged running, cycling and skiing exercises. Sports Med. 2004; 34 (2): 105–16.
28. Millet GY, Millet GP, Lattier G, Maffiuletti NA, Candau R. Alteration of neuromuscular function after a prolonged road cycling race. Int J Sports Med. 2003; 24 (3): 190–4.
29. Newsholme EA. Amino acids, brain neurotransmitters and a function link between muscle and brain that is important in sustained exercise. In: Benzi G, editor. Advances in Myochemistry. London: John Libbey Eurotext; 1987. pp. 127–33.
30. Nybo L. CNS fatigue provoked by prolonged exercise in the heat. Front Biosci. 2010; 2: 779–92.
31. Pietrzak RH, Mollica CM, Maruff P, Snyder PJ. Cognitive effects of immediate-release methylphenidate in children with attention-deficit/hyperactivity disorder. Neurosci Biobehav Rev. 2006; 30 (8): 1225–45.
32. Roelands B, Goekint M, Buyse L, et al.. Time trial performance in normal and high ambient temperature: is there a role for 5-HT? Eur J Appl Physiol. 2009; 107 (1): 119–26.
33. Roelands B, Goekint M, Heyman E, et al.. Acute norepinephrine reuptake inhibition decreases performance in normal and high ambient temperature. J Appl Physiol. 2008; 105 (1): 206–12.
34. Roelands B, Hasegawa H, Watson P, et al.. The effects of acute dopamine reuptake inhibition on performance. Med Sci Sports Exerc. 2008; 40 (5): 879–85.
35. Roelands B, Meeusen R. Alterations in central fatigue by pharmacological manipulations of neurotransmitters in normal and high ambient temperature. Sports Med. 2010; 40 (3): 229–46.
36. Sidhu SK, Bentley DJ, Carroll TJ. Locomotor exercise induces long-lasting impairments in the capacity of the human motor cortex to voluntarily activate knee extensor muscles. J Appl Physiol. 2009; 106 (2): 556–65.
37. 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 (Pt 2): 511–23.
38. Taylor JL, Todd G, Gandevia SC. Evidence for a supraspinal contribution to human muscle fatigue. Clin Exp Pharmacol Physiol. 2006; 33 (4): 400–5.
39. Theunissen EL, Elvira Jde L, van den Bergh D, Ramaekers JG. Comparing the stimulant effects of the H1-antagonist fexofenadine with 2 psychostimulants, modafinil and methylphenidate. J Clin Psychopharmacol. 2009; 29 (5): 439–43.
40. Todd G, Taylor JL, Gandevia SC. Measurement of voluntary activation of fresh and fatigued human muscles using transcranial magnetic stimulation. J Physiol. 2003; 551 (Pt 2): 661–71.
41. Waterhouse BD, Moises HC, Woodward DJ. Noradrenergic modulation of somatosensory cortical neuronal responses to iontophoretically applied putative neurotransmitters. Exp Neurol. 1980; 69 (1): 30–49.
42. Waterhouse BD, Woodward DJ. Interaction of norepinephrine with cerebrocortical activity evoked by stimulation of somatosensory afferent pathways in the rat. Exp Neurol. 1980; 67 (1): 11–34.


©2012The American College of Sports Medicine