Share this article on:

Adaptation of the hypothalamopituitary adrenal axis to chronic exercise stress in humans


Medicine & Science in Sports & Exercise: August 1996 - Volume 28 - Issue 8 - p 1015-1019
Basic Sciences: Original Investigations

Repeated acute or chronic exposure to a particular stress results in adaptation whereby the hypothalamopituitary adrenal (HPA) axis becomes less responsive to subsequent or continued exposure to that particular stress. To investigate the adaptive changes that occur in the HPA axis in response to chronic stress in humans, we studied the effect of chronic exercise stress on basal activity of the HPA axis in six highly trained male ultramarathon athletes and six healthy male controls matched for body mass index.

After 3-5 d of abstention from intense physical activity, the subjects were admitted to a metabolic study ward at 1600 h. Peripheral blood was sampled initially at 0300 h, at 20-min intervals from 0400 to 0900 h, hourly from 0900 to 1200 h, and then every 2 h from 1200 to 1600 h. A 24-h urine collection was completed during the admission. Peripheral blood adrenocorticotropic hormone(ACTH) was measured by radioimmunoassay. Plasma and urinary cortisol were measured by enzyme-linked immunosorbent assay (ELISA).

Plasma ACTH and cortisol levels showed the expected diurnal change in athletes and control subjects (P = 0.00001). However, the early morning ACTH and cortisol surge occurred earlier in the athletes than in the controls (P = 0.026). Plasma ACTH levels were significantly higher in the athletes than in the control subjects (P = 0.0026). There was, however, no significant overall difference in plasma cortisol levels between the athletes and the control subjects, and urinary excretion of free cortisol was similar in the two groups.

These data show that intense physical training leads to adaptive changes in basal HPA function, including a phase shift and increased pituitary ACTH secretion, but also blunting of the adrenal cortisol response.

Department of Endocrinology, Christchurch Hospital, Christchurch, NEW ZEALAND

Submitted for publication September 1995.

Accepted for publication March 1996.

We are grateful to the endocrine special test nurses, the staff of the endocrine and steroid laboratories, and Dr. John Hellemans of Sportsmed for their assistance.

This study was supported by the New Zealand Health Research Council and Canterbury Area Health Board.

Address for correspondence: Dr. G. A. Wittert, Department of Medicine, Royal Adelaide Hospital, North Terrace, Adelaide, SA 5005, Australia

Acute stress results in activation of the hypothalamopituitary adrenal (HPA) axis. Repeated acute or chronic exposure to a particular stress in animal models results in a process of adaptation, and the HPA axis becomes less responsive to subsequent or continued exposure to that stress(13,17). This adaptation to continuous or repeated exposure to stressful stimuli is stressor-specific(17). In animal models of repeated acute or chronic stress, several changes in HPA function have been described; however, the changes that occur in the HPA axis in humans in response to chronic stress have been less well studied.

Acute intense exercise above the anaerobic threshold is an example of a stress that activates the HPA axis (22,31). Physical training, which can be considered a form of chronic stress, is associated with adrenal hypertrophy in laboratory animals(25) and reduced responsiveness of the HPA axis to a subsequent acute exercise challenge (21). In humans, adaptation of the HPA axis to chronic exercise stress also occurs, since intense exercise training results in an attenuated response of the HPA axis to the same absolute workload (4,22). The mechanism by which this occurs is not known, and the documented changes in baseline function of the HPA axis in athletes have been variable. In amenorrheic, but not eumenorrheic, endurance athletes, Loucks et al. (21) demonstrated hypercortisolemia, but peripheral plasma adrenocorticotropic hormone (ACTH) levels were not increased. In male endurance athletes, increased basal plasma ACTH with normal cortisol levels(16,23), increased plasma ACTH and plasma cortisol levels (22), and normal ACTH levels(4) have all been described. Many of these findings, however, are based on either isolated blood samples or poorly standardized times of sampling.

The aim of this study was to determine the effect of chronic exercise stress on basal activity of the HPA axis in highly trained male ultramarathon athletes.

Back to Top | Article Outline



Six intensively trained male ultramarathon runners (mean age 32 ± 4.6 yr, range 21-53) competing in the 1-d event of the New Zealand coast-to-coast marathon volunteered for the study. Competitors in this event cycle 60 km, then follow with 23 km of mountain running, a further cycle of 18 km, and 67 km of kayaking before a final cycle of 70 km. Participation in this event is limited to those who have previously completed a similar course over 2 d, in under 15 h.

Six males (mean age 21 ± 0.4 yr, range 20-22) who regularly undertook moderate noncompetitive exercise were studied as controls. The control subjects were matched with the athletes for body mass index (mean 22.7± 0.5 kg·m-3, range 20.7-24.1, and mean 22.3 ± 0.6 kg·m-3, range 21-24.3, in the athletes and controls, respectively).

All subjects had normal physical examinations and liver and renal function tests. All were nonsmokers and on no medication, and all were asked to abstain from alcohol and vigorous physical exercise for at least 24 h before the study period.

Written informed consent was obtained from all subjects, and the study was approved by the Canterbury Area Health Board Ethics Committee.

Back to Top | Article Outline


The ultramarathoners were studied 3-5 d after completion of the race but before recommencement of physical training. Both groups of subjects were admitted to the metabolic study ward at 1600 h, and a 24-h urine collection was commenced and completed before the subjects left the metabolic study ward at 1600 h the following day to measure urinary free cortisol. At 2000 h, an 18-gauge cannula was inserted into a forearm vein, and the lights turned out. The subjects maintained a recumbent posture throughout, except during ablution(from 0900 to 0920 h). Breakfast was provided at 0805 h and lunch at 1205 h. Fluids (water only) were not restricted. Blood samples were drawn at 0300 h, then at 20-min intervals from 0400 to 0900 h, hourly from 0900 to 1200 h, and every 2 h from 1200 to 1600 h for the measurement of cortisol and ACTH.

Back to Top | Article Outline

Hormone Assays

All samples from one subject were analyzed in the same assay run. Detection limits are expressed at the 95% confidence interval. Plasma ACTH was measured by radioimmunoassay (10). The detection limit of the assay in 0.5-ml plasma extracts was 3 ng·l-1, and the intra-assay coefficient of variation was 7.5%. Plasma and urinary cortisol were measured by enzyme-linked immunosorbent assay (ELISA)(19,20). The detection limit of the assay was 55 nmol·l-1, and the intra-assay coefficient of variation was 6.6%.

Back to Top | Article Outline


Results are expressed as means ± standard errors. The data were analyzed by two-way analysis of variance with repeated measures, employing program P2V of the BMDP statistics package (9), with one grouping factor (ultramarathon athlete or control) and the hormone concentration at the various times as a repeated measure. The polynomial orthogonal components of the time factor were individually tested for significance.

Back to Top | Article Outline


The subjects all slept, and there was no overall difference in sleep disturbance in the athletes compared to the control subjects. The mean(±SEM) plasma concentrations of cortisol and ACTH in the athletes and control subjects are shown in Figure 1. Plasma cortisol levels showed the expected diurnal variation in both groups (main effect for time, P = 0.00001). Although there was no significant overall difference in plasma cortisol levels between the groups (P = 0.94), the early morning increase and peak plasma cortisol levels occurred earlier in the athletes than in the control subjects (time × group interaction,P = 0.026). Urinary free cortisol excretion was similar in the two groups (Fig. 2).

The diurnal variations in plasma ACTH levels were concordant with those of cortisol in the two groups. However, unlike cortisol, plasma ACTH levels were significantly higher in the athletes than in the control subjects (main effect for group, P = 0.0026).

Back to Top | Article Outline


In highly trained male ultramarathon runners, compared to control subjects, we demonstrated a phase shift in the early morning increase in activity of the HPA axis and elevated peripheral plasma ACTH levels, without increase in peripheral plasma levels of cortisol or in 24-h urinary cortisol excretion. It is unlikely that we are seeing persistent activation of the HPA axis due to the effects of the marathon, as indices of HPA axis activity are reported to return to baseline within 24 h of a marathon (6).

To our knowledge a phase shift in activation of the HPA axis in athletes compared to controls has not been reported previously. The group of athletes studied, all amateur ultramarathoners, regularly arose early (0400-0500 h) to train before going to work, compared to the control subjects, who arose between 0630 h and 0800 h. Light clearly has a major synchronizing effect on human circadian rhythms (11). While it is possible that the phase shift is due to the differences in light exposure in the two groups, nonphotic cues also influence circadian rhythms in humans. For example, human rhythms in continuous darkness may be entrained by a rigorous schedule of bedtimes, mealtimes, and performance tasks (1). Furthermore, it has recently been shown that exercise itself can result in phase shifts of the human circadian clock (28).

In this study we chose to measure ACTH primarily during the diurnal activation of the HPA axis in order to determine the relationships between ACTH and cortisol at that time. Plasma ACTH levels were higher in the athletes than in the controls at all times during the entire 13-h period of sampling, including the late afternoon. Although we do not know whether these relationships would be maintained at the nadir of the circadian rhythm, the mean plasma cortisol levels, at least, are likely to be similar between the two groups since 24-h urinary cortisol levels did not differ.

While persistently elevated ACTH levels in the athletes indicate activation of the HPA axis, neither plasma nor urinary free cortisol levels were significantly different between the groups. Since exercise has not been shown to have any significant effect on cortisol-binding globulin (CBG)(23,29) or on the clearance rates of cortisol(29), it seems most likely that there is a reduced adrenal sensitivity to ACTH. This has previously been suggested in trained humans (18,25). Furthermore, it has recently been shown that cortisol responses to exercise in amenorrheic runners are blunted at the adrenal level (8).

Elevated basal plasma ACTH levels without increased cortisol levels have previously been reported in ultramarathon athletes compared to controls, although these observations were based on single prerace plasma samples(25). Luger et al. (22) reported elevated basal plasma ACTH and cortisol levels in highly trained subjects compared to controls. However, this observation was based on two plasma samples 15 min apart in the evening, and 24-h urinary free cortisol data were not available. In contrast to our findings, Loucks et al.(21) found no difference in plasma ACTH levels or ACTH pulsatility over 24 h in either eumenorrheic or amenorrheic women compared to sedentary controls. Overall cortisol secretory dynamics and 24-h urinary excretion in eumenorrheic athletes were similar to those of control subjects. On the other hand, amenorrheic athletes had persistent elevation of plasma cortisol throughout the 24-h sampling period and increased 24-h urinary cortisol excretion, suggesting a closer association of hypercortisolism with amenorrhea than with exercise itself. Similar findings of hypercortisolism in amenorrheic but not eumenorrheic athletes have been reported by others(8). However, increased cortisol production in eumenorrheic as well as amenorrheic runners has also been reported(29). In addition to menstrual status, increased basal activation of the HPA axis may correlate better with type of exercise and duration of training (5,8,14) rather than physical conditioning as such, suggesting that other factors may be important in the persistent activation of the HPA axis noted in some athletes. Recently, the promoter region of the corticotropin-releasing hormone gene was demonstrated to contain an estrogen response element, and sexual dimorphism in the regulation of the HPA axis has been described (26). Therefore, sexual dimorphism in the adaptation to chronic stress may also occur.

Hypercortisolemia may in fact reflect failure of adaptation of the HPA axis to exercise, and HPA axis dysfunction has been shown to be a feature of overtraining rather than intensive training in itself(2). Compulsive running has been suggested to be an analogue of anorexia nervosa (32), and some highly trained runners, as well as anorectic and depressed patients, may represent subjects in whom different stressors (excessive running, perceived environmental changes, weight loss, or emotional stress) may exceed a putative threshold, leading to sustained hypercortisolism(22).

The occurrence of adaptive responses of the HPA axis to training in humans is also suggested by studies showing that the response of the HPA axis to acute exercise to the same absolute workload is attenuated by physical training (4,18,22). Luger et al.(22) have shown convincingly that this is the consequence of a decreased ˙VO2max and have suggested that the altered responsiveness to acute exercise stress is the consequence of changes in afferent input to the HPA axis. The mechanism by which changes in baseline activity of the HPA axis occur in response to training is not known. Possible mechanisms include reduced adrenal responsiveness(7,18,25) and decreased sensitivity to the negative glucocorticoid feedback signal (14). The results of the current study suggest reduced adrenal responsiveness, since decreased sensitivity to glucocorticoid negative feedback would be expected to result in increased levels of plasma cortisol as well as ACTH.

The reason why reduced adrenal responsiveness to ACTH should occur in response to chronic exercise but not other states of ACTH excess is not clear. However, a number of factors, including corticotropin-releasing hormone(27), vasoactive intestinal polypeptide(3), and splanchnic nerve activity(12), have been shown to affect adrenal responsiveness to ACTH. Splanchnicectomy reduces the adrenal sensitivity to exogenous and endogenous ACTH (12). The splanchnic nerve carries both sympathetic and parasympathetic innervation to the adrenal gland. Norepinephrine has been found in nerve fibers and terminals within the adrenal cortex (24) and has been shown to stimulate corticosteroid secretion (30). Trained athletes have been shown to have lower plasma noradrenaline levels than those who become overtrained (15). Therefore, training-induced alterations in splanchnic nerve activity may account for reduced responsiveness of the adrenal to ACTH. An alternative explanation for the disparity between the elevated ACTH levels but normal cortisol levels in the athletes might be that the ACTH has reduced bioactivity. It is not known whether excessive exercise modifies the processing of the propiomelanocortin(POMC) message. In this study we measured ACTH by radioimmunoassay; however, we used an immunoradiometric assay to confirm our findings in a separate group of athletes (16).

Taken together these findings suggest that the normal adaptive response of the HPA axis to chronic exercise stress includes enhanced ACTH secretion but attenuated adrenal response, possibly to protect from the deleterious effects that may occur from sustained hypercortisolemia. Maladaptive responses rather than exercise per se may be responsible for the hypercortisolemia that occurs in such syndromes as the amenorrhea of exercise, hypogonadism in males, and the overtraining syndrome, which have similarities to the pathophysiological changes of anorexia and depression.

Figure 1-Mean (±SEM) plasma concentrations of cortisol and ACTH in the athletes and control subjects.

Figure 1-Mean (±SEM) plasma concentrations of cortisol and ACTH in the athletes and control subjects.

Figure 2-Similarity of urinary free cortisol excretion of the two groups.

Figure 2-Similarity of urinary free cortisol excretion of the two groups.

Back to Top | Article Outline


1. Aschoff, J. Circadian rhythms: general features and endocrinological aspects. In: Endocrine Rhythms, D. T. Krieger(Ed.), New York: Raven, 1979, pp. 1-61.
2. Barron, J. L., T. D. Noakes, W. Levy, C. Smith, and R. P. Millar. Hypothalamic dysfunction in overrtained athletes. J. Clin. Endocrinol. Metab. 60:803-806, 1985.
3. Bloom, S. R., A. V. Edwards, and C. T. Jones. Adrenal responses to vasoactive intestinal peptide in conscious hypohysectomized calves. J. Physiol. (Lond.) 391:441-450, 1987.
4. Bouno, M. J., J. E. Yeager, and A. A. Sucec. Effect of aerobic training on the plasma ACTH response to exercise. J. Appl. Physiol. 63:2499-2501, 1987.
5. Carli, G., G. Martelli, A. Viti, L. Baldi, M. Bonifazi, and C. Lupo di Prisco. The effect of swimming training on hormone levels in girls. J. Sports Med. Phys. Fitness 23:45-51, 1983.
6. Cook, N. J., A. Ng, G. F. Read, B. Harris, and D. Riad-Fahmy. Salivary cortisols for monitoring adrenal activity during marathon runs. Horm. Metab. Res. 25:18-23, 1987.
7. De Souza, M. J., M. S. Maguire, M. S. Maresh, W. J. Kraemer, K. R. Rubin, and A. B. Loucks. Adrenal activation and the response to exercise in eumenorrheic and amenorrheic runners. J. Appl. Physiol. 70:2378-2387, 1991.
8. Ding, J. H., C. B. Sheckter, B. L. Drinkwater, M. R. Soules, and W. J. Bremner. High serum cortisol levels in exercise-associated amenorrhea. Ann. Intern. Med. 108:530-534, 1988.
9. Dixon, W. J. BMDP Statistical Software. Berkeley, CA: University of California Press, 1985.
10. Donald, R. A. Radiommunoassay of corticotrophin (ACTH). In: Handbook of Radioimmunoassay, G. Abrahams (Ed.). New York: Marcel Dekker, 1977, p. 319.
11. Duk, D. J., S. Beersma, S. Daan, and A. J. Lewy. Bright morning light advances the human circadian system without affecting NREM sleep homeostasis. Am. J. Physiol. 256:R106-R111, 1989.
12. Edwards, A. V., and C. T. Jones. The effect of splanchnic nerve section on the sensitivity of the adrenal cortex to adrenocorticotropin in the calf. J. Physiol. (Lond.) 390:23-31, 1987.
13. Hashimoto, K., S. Suemaru, T. Takao, M. Suguwara, S. Makino, and Z. Ota. Corticotropin-releasing hormone and pituitary-adrenocortical responses in chronically stressed rats. Regul. Pept. 23:117-126, 1988.
14. Heuser, I. J., H. J. Wark, J. Keul, and F. Holsboer. Hypothalamic-pituitary adrenal axis function in elderly endurance athletes.J. Clin. Endocrinol. Metab. 73:458-488, 1991.
15. Hooper, S. L., L. T. Mackinnon, R. D. Gordon, and A. W. Bachmann. Hormonal responses of elite swimmers to overtraining. Med. Sci. Sports Exerc. 14:202-206, 1993.
16. Inder, W. J., J. Hellemans, M. J. Ellis, M. J. Evans, J. H. Livesey, and R. A. Donald. Elevated basal adrenocorticotropin and evidence for increased opioid central tone in highly trained male athletes.J. Clin. Endocrinol. Metab. 80:244-248, 1995.
17. Kant, G. J., T. Eggleston, L. Landman-Roberts, C. C. Kenion, G. C. Driver, and J. L. Meyerhoff. Habituation to repeated stress is stressor specific. Pharmacol. Biochem. Behav. 22:631-634, 1985.
18. Lehmann, M., K. Knizia, U. Gastmann, et al. Influence of 6-week, 6 days per week, training on pituitary function in recreational athletes. Br. J. Sports Med. 27:186-192, 1993.
19. Lewis, J. G. and P. A. Èlder. An enzyme-linked immunosorbent assay (ELISA) for plasma cortisol. J. Steroid Biochem. 22:673-676, 1985.
20. Lewis, J. G., L. Manley, J. C. Townsend, and P. A. Elder. An enzyme-linked immunosorbent assay (ELISA) for urinary free cortisol.Clin. Chim. Acta 159:205-209, 1986.
21. Loucks, A. B., J. F. Mortola, L. Girton, and S. S. Yen. Alterations in the hypothalamo-pituitary ovarian and hypothalamo-pituitary adrenal axes in athletic women. J. Clin. Endocrinol. Metab. 68:402-411, 1989.
22. Luger, A., P. A. Deuster, S. B. Kyle, et al. Acute hypothalamo-pituitary adrenal responses to the stress of treadmill exercise. Physiological adaptations to physical training. N. Engl. J. Med. 316:1309-1315, 1987.
23. Pestell, R. G., D. M. Hurley, and R. Van Dongen. Biochemical and hormonal changes during a 1000 km ultramarathon. Clin. Exp. Pharmacol. Physiol. 16:353-361, 1989.
24. Robinson, P. M., R. A. Perry, K. J. Hardy, J. P. Coghlan, and B. A. Scoggins. The innervation of the adrenal cortex in the sheep. J. Anat. 124:117-129, 1977.
25. Tharp, G. D., and R. J. Buck. Adrenal adaptation to chronic exercise. J. Appl. Physiol. 37:720-722, 1974.
26. Vamvakopoulos, N. C., and G. P. Chroussos. Evidence for direct oestrogenic regulation of corticotropin-releasing hormone gene expression. Potential implications for the sexual dimorphism of the stress response and immune/inflammatory reaction. J. Clin. Invest. 92:1896-1902, 1993.
27. Van Oers, J. W., J. P. Hinson, R. Binnekade, and F. J. Tilders. Physiological role of corticotropin-releasing factor in the control of adrenocorticotropin-mediated corticosterone release from the rat adrenal gland. Endocrinology 130:182-288, 1992.
28. Van Reeth, O., J. Sturis, M. M. Byrne, et al. Nocturnal exercise phase delays circadian rhythms of melatonin and thyrotropin secretion in normal men. Am. J. Physiol. 29:E964-E974, 1994.
29. Villanueva, A. L., C. Schlosser, B. Hopper, J. H. Liu, D. I. Hoffman, and R. W. Rebar. Increased cortisol production in women runners. J. Clin. Endocrinol. Metab. 63:133-136, 1986.
30. Walker, S. W., E. R. T. Lightly, C. Clyne, B. C. Williams, and I. M. Bird. Adrenergic and cholinergic regulation of cortisol secretion from the zona fasciculata/reticularis of bovine adrenal cortex.Endocr. Res. 17:237-265, 1991.
31. Wittert, G. A., D. E. Stewart, M. P. Graves, et al. Plasma corticotropin-releasing factor and vasopressin responses to exercise in normal man. Clin. Endocrinol. 35:311-317, 1992.
32. Yates, A., K. Leehey, and C. M. Shisslak. Running-an analogue of anorexia? N. Engl. J. Med. 308:251-255, 1983.


©1996The American College of Sports Medicine