Endocrine-Hypertension Division Brigham and Womens Hospital Harvard Medical School Boston, Massachusetts 02115
Address correspondence and requests for reprints to: Gail Kurr Adler, M.D., Harvard Medical School, Endocrinology/Hypertension Division, 221 Longwood Avenue, RFB 289, Boston, Massachusetts 02481.
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
It is controversial by which mechanism caffeine, a nonselective adenosine receptor antagonist, improves exercise performance. One long-standing hypothesis is that caffeine increases fat mobilization and, therefore, spares muscle glycogen. The increased concentration of muscle glycogen is hypothesized to lead to increased exercise endurance. The study by Laurent et al. (4), published in this issue of The Journal Endocrinology & Metabolism, suggests that in extremely physically fit young men with high muscle glycogen content consumption of caffeine does not alter the effect of exercise on either circulating free fatty acids or muscle glycogen stores. This study is particularly compelling because measurements of muscle glycogen were performed not by muscle biopsy, with its inherent problems of pain and small tissue samples, but by noninvasive 13C-nuclear magnetic resonance spectroscopy. This technology has the advantage of being very sensitive, reproducible, and of examining intact muscles rather than small biopsies. Although this study refutes the hypothesis that caffeine exerts its effects on exercise endurance through preservation of muscle glycogen, it lends support to an alternate hypothesis that caffeine increases exercise performance through effects on the hypothalamic-pituitary-adrenal (HPA) axis and autonomic nervous system (ANS).
In response to exercise in humans, there is release into the hypophyseal-portal system of the hypothalamic ACTH secretagogues CRH and vasopressin (5, 6). These secretagogues act on the pituitary to increase production of the precursor polypeptide pro-opiomelanocortin. Pro-opiomelanocortin undergoes posttranslational processing to produce ACTH, ß-endorphin, and other peptides. ACTH stimulates the adrenal to secrete cortisol. ß-endorphin is an endogenous opioid. The exercise-induced increase in HPA axis activity is proportional to the intensity of exercise as measured by maximal oxygen consumption irrespective of physical conditioning (7). Exercise also induces an increase in circulating epinephrine and norepinephrine. This increase in catecholamines is similarly dependent on the intensity and duration of exercise (8). The current study by Laurent et al. (4) confirms and extends previous studies by demonstrating that consumption of caffeine (6 mg/kg) 90 min before 2 h of moderate cycling results in significant increases in the exercise-induced rises in plasma epinephrine, ß-endorphin, and cortisol.
It has been suggested that this increased activation of the HPA
axis and ANS may mediate the performance-enhancing effects of caffeine,
possibly through the effects of these neuroendocrine systems on fatigue
and pain. Although there is no direct evidence supporting this
hypothesis, there is indirect evidence that these systems may play an
important role in exercise capabilities. Odagiri et
al. (9) showed, in men who had finished a triathlon,
that individuals with high fatigue had lower ß-endorphin and lower
epinephrine levels compared with individuals who completed the
triathlon but had less fatigue. In another study, running was shown to
increase ß-endorphin and CRH levels, and the extent of CRH elevation
correlated with positive mood changes (10). Administration
of naloxone, an opioid antagonist, decreased exercise performance in
one study (11). In rodents, blocking the actions of
central CRH with a CRH receptor antagonist decreased the ability of
rats to run on a treadmill (12). In humans, -adrenergic
blockade during exercise resulted in a significantly shorter time to
exhaustion (13).
These effects of caffeine to stimulate HPA axis and catecholamine responses to exercise are not unique to the exercise stimulus. Caffeine consumption significantly enhances the epinephrine, GH, and cortisol responses to hypoglycemia and has been hypothesized to have an important positive effect on the ability of insulin-dependent diabetics to recognize hypoglycemia (14). Individuals who habitually consume caffeine have been shown to have increased cortisol and catecholamine levels in response to a stressful psychosocial laboratory test as well as increased epinephrine levels during normal work hours (15, 16). In some studies, caffeine has been shown to increase neuroendocrine function in resting individuals. Thus, caffeine seems to have a stimulatory effect on both the ANS and the HPA axis at baseline and to augment the activation of these systems in response to physical and psychological stresses.
It is not unexpected that caffeine would have parallel effects on catecholamine and HPA axis function because there are multiple bidirectional interactions between ANS and HPA axis. There are noradrenergic inputs to CRH neurons of the paraventricular nucleus of the hypothalamus that are derived from medullary catecholaminergic groups (17) and from the locus ceruleus/norepinephrine center in the brainstem (18). Glucocorticoids stimulate the activity of phenylethanolamine N-methyltransferase, the enzyme that converts norepinephrine to epinephrine in the adrenal medulla (19). ACTH may have a direct impact on ANS function. In animals, an ACTH infusion, but not a cortisol infusion, increased messenger RNA levels of tyrosine hydroxylase and dopamine ß-hydroxylase in sympathetic ganglia (20). In women, an infusion of ACTH acutely increased sympathetic outflow to the muscle vascular bed (21). Furthermore, treatment of healthy subjects with prednisone, which inhibits CRH and ACTH, decreased sympathetic nerve activity and plasma norepinephrine levels (22). Finally, cortisol pretreatment inhibits the rise in epinephrine, norepinephrine, ACTH, and cortisol normally induced by hypoglycemia (23). It is an open question whether glucocorticoid treatment would impair exercise performance.
There are significant interindividual differences in the activation of the HPA axis and in the rise of circulating catecholamines in response to psychological and physical stresses. Men categorized as being either low or high ACTH responders to an exercise test continued to show the same low or high ACTH response when subjected to a psychological stress, suggesting that there are inherent differences between individuals in HPA axis reactivity (24). Gender may have an important effect on HPA axis reactivity. Women seem to release less ACTH than men in response to psychological stress (25). A multivariant analysis including 70 men and women showed significantly lower epinephrine and norepinephrine responses to hypoglycemia in women compared with men (26). Women also release less epinephrine and norepinephrine than do men in response to exercise, despite reaching the same maximal oxygen consumption (27). It is not known whether these differences in HPA axis and ANS activation mediate some of the gender differences in exercise performance.
The etiology for these differences in neuroendocrine and ANS responsiveness to stress is not known. One study of monozygotic and dizygotic twin pairs suggested a genetic influence on HPA axis function (28). Aging and environmental factors may alter HPA axis reactivity. There are also a number of physiological conditions associated with fatigue and decreased activity of the HPA axis and/or ANS, including fibromyalgia, chronic fatigue syndrome, rheumatoid arthritis, hypothyroidism, the postpartum period, and steroid withdrawal. In fibromyalgia, individuals with the lowest epinephrine response to hypoglycemia had the greatest impairment in health (29). Thus, the activity of the HPA axis and ANS may be influenced by a wide variety of genetic, environmental, and physiological factors.
Caffeine consumption improves exercise performance and increases activity of HPA axis and ANS. Whether this activation of neuroendocrine function is a marker or a mediator of the actions of caffeine on exercise performance is not known. It is intriguing to consider, however, that individual differences in exercise endurance, fatigue, and pain at baseline and in response to a variety of stresses may, in part, be mediated by activity of HPA axis and ANS. Additional studies with pharmacological manipulations of these axes may prove useful in better understanding normal human physiology, as well as the pathophysiology, of a variety of disorders characterized by fatigue or impaired exercise performance.
![]() |
Acknowledgments |
---|
![]() |
Footnotes |
---|
Received April 17, 2000.
Accepted April 17, 2000.
![]() |
References |
---|
![]() ![]() ![]() |
---|