Address reprint requests to Dario Acuña-Castroviejo, Centro de Investigación Biomédica, Parque Tecnológico de Ciencias de la Salud, Avenida del Conocimiento s/n, 18100 Armilla, Granada, Spain. E-mail: firstname.lastname@example.org
Abstract: The aim of this review is to update the reader as to the association between physical exercise and melatonin, and to clarify how the melatonin rhythm may be affected by different types of exercise. Exercise may act as a zeitgeber, although the effects of exercise on the human circadian system are only now being explored. Depending on the time of the day, on the intensity of light, and on the proximity of the exercise to the onset or decline of the circadian production of melatonin, the consequence of exercise on the melatonin rhythm varies. Moreover, especially strenuous exercise per se induces an increased oxidative stress that in turn may affect melatonin levels in the peripheral circulation because indole is rapidly used to combat free radical damage. On the other hand, melatonin also may influence physical performance, and thus, there are mutually interactions between exercise and melatonin production which may be beneficial.
Structure and function of the human circadian system
Most organisms, from cyanobacteria to humans, display rhythms in their physiology and behavior that are synchronized to environmental cycles of 24 hr. These rhythms are known as circadian rhythms. To generate this rhythmicity, organisms are provided with a time measuring device, known as the circadian timekeeping system. In humans, the circadian system is composed of a hierarchical network of structures responsible for the generation of circadian rhythms which keep cellular clocks synchronized with the daily environmental cycles (Fig. 1), especially the photoperiod.
In mammals, the main circadian pacemaker is located in the suprachiasmatic nuclei (SCN) of the hypothalamus where each individual neuron can independently generate a self-sustained circadian rhythm. These rhythms are generated by a molecular clock based on complex cycles of transcription, translation, protein–protein interaction, phosphorylation, nuclear translocation, and protein degradation, all of which impose a delay to create a coordinated molecular cycle that matches the 24-hr environmental LD period .
The basic loop of the molecular clock is composed of two positive elements, CLOCK and BMAL1, which dimerize to activate rhythmic transcription of Per and Cry genes by binding to specific promoter elements. After translation, PER and CRY proteins dimerize and undergo nuclear translocation to inhibit the coupling of CLOCK:BMAL1, resulting in decreased transcription of Per and Cry genes. This molecular clock also regulates the expression of several clock-controlled genes (3–15%), which are not a part of the main clock machinery but are responsible for the generation of circadian rhythmicity in many physiological processes .
Recently, the molecular clock has been also reported in many peripheral tissues and organs such as, adipose tissue, liver, heart, suprarenal, pancreas, all of them known as peripheral circadian oscillators [3, 4]. Currently, it is thought that circadian oscillators are functional in most, if not all, body cells . Under normal physiological conditions, peripheral oscillators are synchronized by the activity of the central pacemaker, the SCN. However, under certain situations, these oscillators can desynchronize from the SCN control generating a situation of circadian misalignment, known as chronodisruption [5–7].
When isolated from environmental cues, both the central pacemaker and the peripheral oscillators exhibit circadian rhythms with periods slightly different from 24 hr; thus, they need regular periodic input from environmental time cues (zeitgebers or synchronizers) to maintain their activity entrained to 24 hr. Among these inputs, the 24-hr light–dark cycle is the main zeitgeber for the SCN; however, humans can also use other nonphotic zeitgebers, including scheduling of rest and activity, meal time and social contacts as useful ‘time-giver’.
The synchronized activity of SCN cells communicates circadian phase information through neural and humoral signals to every cell in the organism. The autonomic nervous system and the melatonin and core body temperature rhythms are well-known output signals from SCN to generate circadian rhythms in most physiological and behavioral variables. However, as humans can modify voluntarily some of these output rhythms, for example, meal time, rest activity or physical exercise, these outputs can also act as input signals which, in turn, modify the activity of the central and peripheral clocks. Moreover, the central pacemaker and peripheral oscillators respond to different zeitgebers; for example, the liver is synchronized by food restriction but not the SCN, which responds mainly to light. These features of human circadian system make the interpretations of results from human studies more complicated.
Physical exercise as a synchronizer of the circadian system
Contrary to light, the mechanisms involved in the synchronization by nonphotic time-cues, such as physical exercise, are less well understood. In recent years, there has been a growing interest in physical exercise as a nonphotic signal to entrain the human circadian clock. In addition to its clear health benefits, physical exercise is a preferred method, over pharmaceutical interventions, to synchronize the circadian system. Under certain conditions, scheduled physical exercise significantly phase delays the human circadian pacemaker [8–10], which may help to facilitate circadian adaptation to schedules requiring a delay in the sleep–wake cycle such as in clock-wise rotating shift work [11, 12]. In addition, some studies show that physical exercise can accelerate the re-entrainment of the human sleep–wake rhythm to an acutely shifted sleep–wake cycle [13, 14], supporting its application as a treatment for circadian rhythm misalignments resulting from jet-lag or shift-work as an example.
Animals studies demonstrate that vigorous wheel running exercise, even when it is unscheduled, prevents the loss of rhythmicity in rats maintained under constant light (a condition which induces arrhythmicity in this species) and accelerates the emergence of circadian pattern in arrhythmic animals moved to dim light. Physical exercise promotes stronger coupling in the multioscillatory circadian system . In addition, in some diurnal rodent species such as Octodon degus and Arvicanthis niloticus, the availability of a wheel in the cage can induce a phase inversion, from diurnal to nocturnal behavior. The mechanism by which physical exercise induces nocturnalism should be located downstream from the clock [16–18].
The effects of physical exercise on the circadian system, however, require medium-to-long-term repeated training. In a pioneering study, Van Someren et al.  administered a 3-month training program to healthy elderly men. He reported that fitness training induced a significant reduction in the fragmentation of their rest-phase rhythm. Fragmented rhythms are characteristics of aging and neurodegenerative diseases such as Alzheimer or Parkinson diseases; thus, scheduled physical exercise can be used to improve some negative symptoms associated with these debilitating conditions .
The actual usefulness of physical exercise as a chronobiotic in humans is still unclear and results are controversial. One reason is that studies investigating the effects of exercise on circadian rhythmicity are hampered by the lack of controls for competing zeitgebers, such as light and athletic status (sedentary individuals to Olympic athletes) . In fact, there are reports showing that rhythm amplitudes of physically fit subjects tend to be higher than in unfit individuals .Thus, standardization of these factors is necessary before any firm conclusion can be drawn regarding the effects of exercise on the human circadian system. Nevertheless, there is growing evidence that physical exercise of varied duration and intensity can cause phase shifts that are independent of those produced by light .
In summary, a review of the references on rest-activity schedules and physical exercise suggests that a combination of both of them can be used to potentiate the entrainment of the circadian system. Rest-activity schedules alone can entrain the SCN but in a very narrow range of periods around 24 hr. Physical exercise can induce larger effects and, as a consequence, it might expand this range. However, the fact that only some blind people, without circadian photoreception, remain entrained to 24 hr suggest that nonphotic cues, including physical activity, should be considered as weak zeitgebers compared with light–dark cycle .
The effect of physical exercise on melatonin production
One of the main disadvantages in assessing the synchronizing effect of exercise on the human circadian system is the inability to directly measure its phase-shifting effects of the central pacemaker. Instead, the levels of one of the main output signals of the clock, melatonin, are commonly used to report the phase-shifting effects of exercise on the circadian clock. In addition, acute and chronic physical exercise also modifies plasma melatonin levels.
In this regard, there is some controversy about the effects of physical activity on the endogenous profile of melatonin secretion. It has been shown that melatonin levels increase [24–26], decrease [27, 28], or remain unaffected by exercise [13, 29]. Such conflicting findings may be due to differences in lighting conditions and the time of day at which the study subjects exercised . Moreover, there is a mixture of applied and basic research on this topic. Applied researchers are interested in whether exercise can help people during rhythm disturbances; but this question still remains to be answered . There is now evidence that exercise of quite varied durations and intensities can mediate phase shifts in rhythms in secretory products, independent from those of light, in populations differing widely in athletic status and age .
Effects of physical exercise on the phase of melatonin rhythm
Three markers of phase from the diurnal profile of melatonin are currently used: the onset of the rise (dim light melatonin onset, DLMO), the time of peak, and the offset of melatonin secretion. Although melatonin is a good marker of clock phase (currently considered as the ‘gold’ standard marker of circadian system), it is difficult to eliminate confounding variables that can affect its synthesis and release following physical exercise. For example, it is well known that nocturnal melatonin synthesis is greatly inhibited by light . So, the time of day (presence or absence of light) at which the effect of exercise on phase shift is assessed is important. These ‘masking’ effects can produce false interpretation of the results and/or inconsistencies between studies, but research strategies are now in place to minimize these confounding factors.
Despite the inconsistencies mentioned earlier, there is a general consensus that night-time exercise, whether of moderate or high intensity, under a constant routine results in phase delays in melatonin onset [8–10]. Barger et al.  also reported this phase-delay effects of nocturnal exercise on the melatonin rhythm when exercise was performed in the dark or under very dim light. There is also agreement that, on average, the difference in phase shift between the nocturnal exercise and control conditions was similar between the young versus older participants, suggesting that age does not influence the sensitivity of circadian rhythms to exercise .
Reports of exercise-induced phase advances in the melatonin rhythm are more rare. Of the few studies that have been published, phase advances were reported when exercise was performed in the morning compared with evening . These studies have been criticized as the phase advance observed in these subjects may be due to morning sunlight exposure (light suppressing effects on melatonin synthesis), and are not because of exercise carried out at this time. Recently, a clear phase advance in response to a single bout (1 hr) of exercise has been reported . In this study, five groups [(morning, afternoon, evening, night, and control (no-exercise)] were investigated. The mean timing of the onset of melatonin secretion was 22:45 ± 01:11 hr. Analysis of the phase-response curve (PRC) of melatonin showed a crossover point between delays and advances at approximately 12 hr after melatonin onset . Unlike with exposure to bright light, phase advances in melatonin rhythm were found when exercise was done between 18 and 21 hr after melatonin onset (from noon to evening). The question as to why exercise and light have different phase response curve (PRC) deserves further research.
Most results reporting exercise as a zeitgeber have come from studies in which participants undertook a single bout of exercise followed by a single estimate of clock phase. Buxton et al.  measured the melatonin rhythm in three circadian cycles after bouts of exercise were performed in the evening, and found that apparent phase advances of 30 ± 15 min between day 1 and 2 become delays of 66 ± 9 min after day 3. This shows the complex relationship between exercise and phase shifting, and strengthens the argument that although shifts are initially small, cumulatively they can become substantial if exercise is performed regularly.
Influence of physical exercise on melatonin levels
Accumulating evidence suggests that in addition to its phase-shifting effects, exercise can acutely alter melatonin levels as well . Although most studies show that plasma melatonin levels increase shortly and transiently after exercise [34–36], a decrease or no change in melatonin secretion after exercise have been also reported [27, 28]. This variability in exercise-induced acute response on melatonin release could be dependent on the circadian phase at which exercise was undertaken . Inter-study differences in light conditions (presence or absence of bright light), and the type, duration, intensity of exercise, as well as melatonin sampling sites (saliva versus plasma) and its measurement methods (RIA versus ELISA) also impinged on the outcomes .
A clearer observation seems to be that the acute increase in circulating melatonin that occurs after an exercise bout is attenuated by regular and vigorous training. In woman subjected to a conditioning (running) program, the acute peak of melatonin in response to treadmill exercise decreased by 52% as training progressed . Although the exact physiological relevance of the acute rise in plasma melatonin induced by exercise is not known, it may be advantageous to the organism given that, at least strenuous exercise generates oxidative stress and melatonin is a potent antioxidant capable of protecting against potential molecular damage [38, 39].
In normal volunteers, it was found that melatonin level increased immediately after exercise and returned to pre-exercise levels 1 hr after physical exertion . In training subjects, morning 6-sulfatoxymelatonin levels before competition started in professional male cyclists were higher than those collected in the evening [9, 40]. However, evening levels of 6-sulfatoxymelatonin during 3 wk of a tour race were higher than morning levels. On the other hand, both morning and evening 6-sulfatoxymelatonin levels decreased during the 3 wk of the tour race . It was recently reported that after 4 days of competition, well-trained cyclists show an adaptative response to physical overloads, regulating efficiently their oxidative stress, and increasing the diurnal melatonin levels . Interestingly, urinary excretion of 6-sulfatoxymelatonin did not change significantly. It seems that whereas training does not cause any chronic change in melatonin secretion, physical exercise increases melatonin in the blood temporarily . Together, these data suggest that the melatonin rise in response to extenuated exercise is metabolized through oxidative pathways to metabolites others than 6-sulfatoxymelatonin, for example, to N1-acetyl-methyl-N2-formyl-5-methoxykynuramine . Thus, urinary 6-sulfatoxymelatonin excretion may be not useful to evaluate melatonin changes in exercise. Working with sled dogs, a dog strain that can be considered elite athletes, Dunlap et al.  reported a reduction in melatonin levels after exercise in both winter and summer seasons in dogs raised in Alaska, but only a reduction in winter in sled dogs raised in New York. These data also account for seasonal variations in exercise-dependent melatonin changes.
Further complexities in assessing the melatonin response to exercise arise from the hypothesis that changes in melatonin levels might depend on the initial concentration of systemic melatonin. That is, exercise conducted when melatonin levels are already high results in a further increase in its levels. For example, results from Monteleone et al.  show that in late evening (22:40 hr) exercise, in the rising phase of melatonin secretion, may blunt melatonin levels, whereas Buxton et al.  reported that high-intensity exercise during the night-time (01:00 hr) period, when melatonin levels are already elevated, was associated with an acute increase in melatonin concentrations. However, again there is contrasting evidence showing inconsistent  or no  immediate effects of 3 hr moderate intensity exercise on melatonin concentrations at the same circadian phases. A later study  also failed to show an acute increase in melatonin secretion following morning and afternoon exercise. Although the mechanism underpinning the possible acute alteration in melatonin concentration following exercise remained unknown, a recent report confirmed that the absolute rise in melatonin levels following exercise was greatest and more pronounced in the morning than in the afternoon .
It is known that subjecting animals to stress results in an alteration in endocrine physiology. This stress could induce an elevation in plasma melatonin in humans, especially in athletes during performance or endurance training. Bullen et al.  suggested that these exercise-enhanced melatonin levels may have severe health consequences, contributing to impairment of reproductive function in women engaging in endurance sports. This seems unlikely as suppression of reproductive physiology by melatonin in humans has not been reliably documented. Other studies indicate that professional road cyclist display an adaptive response to the physical overloads during competitive events . It is believed that this adaptive strategy allows them to efficiently regulate intracellular oxidative stress and prevent an exacerbation of pro-inflammatory cytokine induction. After competition, an increase in plasma melatonin concentration is also detected. Because melatonin is known to have anti-oxidative properties , an increase in its level can provide a modulatory role in adaptive stress responses, offering protective actions against free radical-mediated damage.
Coupling mechanisms between exercise and melatonin
Exercise is known to impact on many of the body’s homeostatic systems, including the stress response. Daily variation in melatonin synthesis is controlled by norepinephrine secreted by the postganglionic sympathetic nerves that innervate the pineal gland . Exercise caused a marked increase in the activity of the sympathetic nervous system and catecholamine secretion . This increased activity in the sympathetic nervous system could potentially modulate melatonin secretion. In the human, this rise in melatonin secretion could cause a net phase-shifting effect, via the pineal gland, by acting directly on SCN cells, which express receptors for melatonin .
Besides this stress-mediated effect of exercise on the pineal gland, two other input signals induced by physical exercise could be responsible for the entraining effect of physical exercise on the SCN and, thus, on the melatonin rhythm . The first pathway allows the integration of photic (mediated by the retinohypothalamic tract) and nonphotic inputs originating in the midbrain raphe nuclei (MRN). Physical activity stimulates the raphe which in turn activates serotoninergic input to the intergeniculate leaflets (IGL). The IGL communicates this signal to SCN via neuropeptide Y release. The second important afferent input to the master pacemaker comes directly from MRN. The activity of serotoninergic neurons from the MRN is arousal dependent as serotonin levels in the SCN follow the daily pattern of locomotor activity, both in nocturnal and diurnal rodents .
Exercise has positive effects on mood and anxiety [51–54]. How daily exercise affects mood is not known exactly. Some hormones, for example, glucocorticoids and several neuro-transmitters, for example, serotonin, norepinephrine, which are altered during both psychological and physical stresses, may play an important role in mood after exercise . Exercise may be involved in mood improvement by resetting the master circadian pacemaker via these sero-tonergic inputs. Serotonin provides both direct and indirect inputs to the SCN and it is involved in the effects of physical activity on the clock .
The molecular mechanisms coupling acute exercise or serotoninergic receptor activation during the mid-subjective day in nocturnal rodents are associated with the suppression of Per expression. This suppression is associated with large phase advances when serotonin agonists are administered during subjective day. However, the levels of mRNA of Per 1, Per2, Rev-Erbα and β and Rorα and β are not modified by serotoninergic activation within the SCN of a diurnal rodent, Arvicanthis ansorgei . Further research is needed to understand the mechanism by which physical exercise can induce circadian system synchronization and modifies melatonin levels.
Melatonin and physical performance
Any one undertaking exercise needs to provide a sustained high oxygen delivery to the working muscles, including those related to respiration. In elite athletes, this is achieved to a large extent by the ability to increase cardiac output, together with adaptations in the trained muscles and amelioration of the deleterious effects of exercise-related oxidative stress. Herein, we will describe melatonin effects that could improve physical performance.
Effects on cardiovascular system
The increase in cardiac output can be attributed to an rise in maximal stroke volume enabled by having an enlarged left ventricle as a result of prolonged periods of endurance training . In addition, it is totally accepted that the onset of exercise is associated with changes in the cardiovascular system (increase in the heart rate and vasoconstriction except in working muscles) that are a consequence of a withdrawal of cardiac vagal tone  and an increasing effect of cardiac sympathetic acceleratory nerves . These changes are activated by central motor command and by afferent fibers from exercising muscles sensitive to distortion or stretch (type III) or to metabolic products (type 4; review in ).
As it has been pointed out earlier, a relationship between melatonin and exercise is supported by the fact that catecholamine secretion is markedly elevated during physical exercise . Cardiovascular effects of exogenous and endogenous melatonin could favor exercise performance. Thus, intravenous administration of melatonin in the baboon increases the cardiac output and left ventricular ejection fraction, which implies a positive inotropic action on the heart by melatonin that could increase cardiac output . Measurements of salivary melatonin concentration during and after exercise show that the gradient of the heart rate–melatonin relationship is steeper in the morning than in the afternoon . Taking into account that the absolute increase in melatonin was greater in the morning when the circulating levels of the indoleamine are higher, these results raise the possibility that circulating melatonin favors the increase in cardiac output during exercise and that the time of day alters the relationships between exercise-mediated sympathetic neural activity and melatonin secretion.
In humans at rest, melatonin administration decreased heart rate and blood pressure suggesting that melatonin increases cardiac vagal tone in the supine position in awake men. Melatonin administration also may exert suppressive effects on sympathetic tone as evidenced by a decrease in catecholamine and dopamine levels . This effect of melatonin could reinforce the exercise-induced rise in vagal resting tone, which improves exercise performance.
Acute myocardial infarction (AMI) is a pathological condition which is associated with marked neuroendocrine dysfunction in addition to cardiac damage. Clinical studies have reported an increased incidence of AMI, sudden death and ischemia when there is a rapid withdrawal of vagal activity and an increase in sympathetic tone . This condition is related to a rise in proinflammatory agents and a decrease in the serum levels of anti-inflammatory and antioxidant compounds including melatonin .
Physical exertion acts as a trigger of sudden death and AMI in susceptible individuals , and current evidence suggests that prolonged intense exercise causes the heart to become transiently dysfunctional, with a reduction in left systolic and diastolic function . In a rat model, it has been reported that melatonin administration prior to acute exercise reduces elevated creatine kinase (CK) and cardiospecific isoenzyme of CK (CK-MB) activities in blood and myeloperoxidase levels in cardiac muscle. In addition, exercise was associated with a significant rise in TNF-a, IL-1 and IL-6 mRNA levels in heart, suggesting an inflammatory response that is also supported by the elevation of intercellular adhesion molecule-1 (ICAM-1), inducible nitric oxide synthase (iNOS), cyclooxygenase-2 (COX-2) and a significant activation of nuclear factor kappa B (NF-κB) . These effects were totally or partially prevented by melatonin administration, which suggests that melatonin administration has potent protective effects against damage caused by acute exercise in rat heart. These findings are in keeping with the beneficial effects reported for melatonin on age-related oxidative stress and inflammation in mouse hearts . It is therefore reasonable to conclude that melatonin supplementation would reduce the risk of AMI and myocardial damage associated with acute exercise. Indeed, in animal studies, it is well documented that melatonin reduces damage to heart muscle after ischemia/reperfusion injury .
Effects on the skeletal muscle and motor end-plate
One important factor in physical performance is the adaptative capability of skeletal muscle to different types of exercise training which depends on rearrangements in the contractile apparatus, mitochondria, other fiber organelles and neuromuscular junctions. Endurance exercise, commonly performed at a moderate intensity with continuous or interval-based repetitions, results in increased mitochondrial number, capillary density, enhanced capacity of the oxidation of carbohydrates and lipids [68–70] and the structural–functional relationships between the energy system and contractile machinery in muscle fibers . Resistance exercise, which is commonly performed at high intensities for shorter durations of time, enhances muscle size by increasing the synthesis of contractile and structural proteins and, as a result, the muscle is often larger and also more powerful .
There is not much information related to melatonin effects on nerve physiology and remodeling after exercise. Remodeling of the sciatic nerve because of heavy exercise caused an increase in rheobase and chronaxie, and a reduction in the maximum depolarization, total area under the compound action potential (CAP), fall-down phase of CAP kinetics and speed of the intermediately conducting group. Pretreatment with melatonin protected the sciatic nerve from exercise-induced damage, which increased the contribution of intermediate nerve fibers to muscle excitation .
Nicotinic AChRs (nAChRs) are ligand-gated ion channels formed by pentameric arrangements of α and β subunits, which based on their sensitivity to α-BTX and nicotine can be subdivided in two large families . Endurance training increases the abundance of endplate-associated nAChRs that are sensitive to α-BTX . Data collected in several tissues and under different conditions indicate that the nocturnal melatonin surge increases the number and/or potentiates the effects mediated by nAChRs sensitive to α-BTX (reviewed in ). Thus, it is be possible that endogenous or exogenous melatonin enhances the response of skeletal fibers to alpha motoneurons, increasing exercise performance. This hypothesis has not been tested to date.
Effects on exercise-related metabolism
Carbohydrate, fat, and amino acid metabolic pathways provide a substrate for muscular energy metabolism . During prolonged exercise, both glucose and free fatty acids are utilized as fuels , which results in a significant hypoglycemia and increased plasma levels of lactate and β-hydroxybutyrate, together with a significant reduction of glycogen in muscle and liver . Skeletal and liver glycogen can provide much of the carbohydrate required to perform endurance activities. However, the longer the exercise duration, the larger is the contribution of liver glucose arising from glycogenolysis and gluconeogenesis as an energy substrate . The ability to store and maintain muscle glycogen has long been considered to be the most important limiting factor in the successful performance of submaximal endurance events . It has been shown that melatonin supplementation before exercise preserves glycogen stores through changes in carbohydrate and lipid utilization maintaining glycemia and reducing plasma and liver lactate and plasma β-hydroxybutyrate [79, 81, 82]. Metabolic effects of melatonin in exercised rats are not related to a decrease in nitric oxide because blockade of nitric oxide synthase by L-NAME did not mimic those of melatonin. A role of GH, although suggested by the authors, is not supported by the effects of the cholinergic agonist pyridostigmine . Melatonin also increases glucose uptake into skeletal muscle via the stimulation in phosphorylation of insulin receptor substrate-1 (IRS-1) and the activity of phosphoinositide 3-kinase (PI3-kinase) , which would help to increase myotube glucose content during exercise.
Effects on muscle oxidative stress
It has been known since the late 1970s that exercise promotes oxidative stress  with skeletal muscle being one of the major source of free radicals and ROS generation . The available evidence suggests that contracting muscles produce ROS from several cellular locations: mitochondria, NADPH oxidase (located in sarcoplasmic reticulum, transverse tubules, and sarcolemma), PLA2-dependent processes and xanthine oxidase (reviewed in ). It is well established that low levels of ROS are a requirement for normal force production  and that a small increase in ROS in skeletal fibers promotes an elevation in force generation, whereas high ROS concentrations reduce force production  and contribute to muscular fatigue during prolonged and intense exercise . Animals and human studies have shown that scavenging of ROS with the use of antioxidants delays muscle fatigue and melatonin, owing to its antioxidant activity, reduced free radical–mediated muscle damage resulting from exercise [89, 90]. The role of melatonin as an effective antioxidant agent and a free radical scavenger in all tissues is well known. This effect is initiated by oxidative cleavage of the pyrrole ring when it donates an electron [38, 39, 91, 92] and it is enhanced by its metabolites cyclic 3-hydroxymelatonin (C3-OHMel) and N1-acetyl-N2-formyl-5-methoxykyinuramine (AFMK) and N1-acetyl-5-methoxykynuramine (AMK) [93–96]. Taking this into account, it is highly probable that melatonin treatment can reduce muscle fatigue improving physical performance.
Repeated bouts of endurance exercise increases aerobic capacity and improves mitochondrial function of the skeletal muscle. Proteomic analysis of human skeletal muscle mitochondria after training has shown that proteins related to enhanced capacity for adenosine triphosphate generation were differentially expressed . In addition, mitochondrial biogenesis is promoted via activation of the transcriptional co-activator PGC-1α , and free radicals were suggested as the possible signal for mitochondrial biogenesis . However, it has also been reported that strenuous or excessive exercise causes fatigue and damage to muscle accompanied by a decrease in PGC-1α and a rise in autophagy and mitochondrial fission in skeletal muscle. This was reverted after treatment with the antioxidant hydroxytyrosol, which also favored mitochondria fusion and elevated mitochondrial complex I and II activities . Although melatonin effects on mitochondrial function have not yet been explored in exercise, the beneficial effects of melatonin in the restoration or preservation of mitochondrial function after ischemia/reperfusion injury in skeletal muscle is well documented [101, 102]. Our group has also shown that melatonin counteracts sepsis-induced mtNOS induction and respiratory chain failure in the diaphragm and skeletal muscle by restoring the redox status [103, 104]. Mitochondria from iNOS knockout (iNOS−/−) mice were unaffected by either sepsis or melatonin treatment, indicating a role for NO in the deleterious effects of sepsis and that melatonin can protect against mtNOS-mediated mitochondrial failure. More extensive reviews on the effects of melatonin in mitochondrial function can be found elsewhere [105, 106].
Effecst on physical performance
Based on these findings, we would expect an increase in physical performance after melatonin usage. There is little research on this topic, and the results found are not promising. Some studies looked for the hypothermic effects of melatonin to improve endurance performance in hot environments and the results are not consistent. It has been reported that melatonin has no influence on rectal or skin temperature responses at rest or during walking at 40°C when a dose of 1 mg melatonin was evaluated, but 5 mg caused a significant decrease in core temperature during rest and during walking in an environment at 23°C, but no effects were found when trials were performed at 40°C . It was recently reported that 2.5 mg of melatonin moderated the increase of rectal temperature, and amplified the skin blood flow and hypotension responses to exercise without any side effects on alertness .
There are also some reports suggesting that when melatonin (10–80 mg) was administered in the daytime to men, their reaction time was slowed, subjective alertness decreased and the number of correct responses on a vigilance task dropped . In a different study, 5 mg of melatonin did not increase performance and ratings of perceived exertion during a 4-km cycling time trial, but reduced alertness, short-term memory, eight-choice reaction time and intra-aural temperature . Similarly, melatonin administration during daytime did not have any acute effects either on the maximal jumping ability or on strength in spite of the reported decrease in alertness by melatonin .
Regarding exercise performance, a study designed to determine whether melatonin could overcome the decline in the performance caused by jet-lag reported no effects on static nor dynamic physical performance, whereas caffeine slightly recovered static physical performance after jet lag . Considering the reported findings, melatonin may have a greater effect on cognitive performance variables than on physical performance . More detailed studies in animals and humans are needed to unveil the effects of melatonin on exercise performance.
Conclusions and perspectives
From the data reviewed here, it is clear that the exercise–melatonin interplay may have a favorable influence over many systems in the body. More information on exercise and melatonin, however, is necessary to explain some contradictory findings in the literature. If the exercise–melatonin relationships are clarified, through a careful control of time of exercise and light experimental conditions, a melatonin phase response curve for exercise may be clarified. Moreover, additional research is required to fully understand the relationship between exercise and melatonin, because some of the effects of exercise on human biology may be mediated by melatonin. In this regard, melatonin reduces oxidative stress and inflammation in cardiac and skeletal muscle induced by different conditions such as sepsis, aging, exercise, etc. [104, 113–118]. Thus, the levels of melatonin detected after exercise do not correspond to the melatonin produced by exercise itself, but also depend on the degree of utilization of melatonin in resisting oxidative damage. Finally, the results of future studies may yield important information on the melatonin–exercise interactions, which are relevant in the clinical practice [119–121].
This study was partially supported by grants from the Instituto de Salud Carlos III, Spain (RD06/0013/0008, RD06/0013/1012, and RD06/0013/0019, RETICEF); from the Ministerio de Educación y Ciencia, Spain (BFU 2010-21945-CO1, and BFU 2007-60563), and from the Consejería de Innovación, Ciencia y Empresa, Junta de Andalucía, Spain (CTS-101).
All authors have directly contributed to the preparation and drafting of the manuscript, its critical revision, and approval of the final form of this review.