The Physician and Sportsmedicine
Menubar Home Journal Personal Health Resource Center CME Advertiser Services About Us


Exercise in the Heat: A Double Threat to the Immune System?

Roy J. Shephard, MD, PhD, DPE


Various manifestations of an athlete's immune function are temporarily suppressed by a bout of prolonged and exhausting exercise (1,2). In particular, decreases in natural killer (NK) cell activity and serum immunoglobulin levels can increase susceptibility to infection, and possibly death, from viral myocarditis (3). Further, many of the immune changes seem to be essentially stress responses (4,5). These findings suggest that the threat to immunocompetence could rise in an additive or even multiplicative fashion if the competitor is exposed to extremes of heat (6-8).

Available laboratory studies support this view (9), and in situations in which a substantial increase in the risk of infection has been demonstrated, such as the Los Angeles Marathon and the Comrade's Marathon in South Africa (actually a 56K ultramarathon), it seems likely that the core body temperature of many competitors reached quite high levels. Nevertheless, it remains unclear to what extent the rise in core body temperature exacerbates the situation simply by increasing "stress," stimulates the production of various hormones and cytokines, and compounds the reactions attributable to an exercise-induced acute phase/inflammatory reaction in injured tissues.

Given the effect of prolonged competitive exercise on immune function and the effects when combined with activity in the heat, every effort should be made to minimize athletes' exposure to stressors other than exercise. Possible analytic approaches to questions posed include manipulating body temperatures, assessing and/or modifying stress levels, relating immune function to plasma concentrations of "stress" hormones, and examining the effects of changes in adhesion molecule expression on the trafficking of lymphocytes.

Manipulation of Body Temperatures During Exercise

Studies of interactions between exercise and core body temperature have used climatic chambers, water bath immersion, and radiothermy (9).

Climatic chambers and water bath immersions. Early observations from climatic chamber experiments showed that heat exposure alone had little effect on immune function and that a combination of heat and exercise caused only slightly more disturbance than exercise alone (10). However, heat exposure alone increased rectal temperature by only 0.5°C, and even with the combination of exercise and heat exposure, the final rectal temperature readings averaged only a modest 38.1°C. This may be important, since the threshold temperature for any humoral or immune response to body heating alone seems to be a body core temperature of at least 38°C.

The changes with heat alone were smaller than with exercise. Nevertheless, 2 hours of passive heating in a water bath brought core body temperatures to 39.5°C. This induced increases in levels of lymphocytes (predominantly NK cells) during heating, with a concomitant decrease 2 hours postexposure (11-13). The number of neutrophils increased both during and after heating, but monocyte percentages showed their largest increase post-heating.

Thermal clamping. The thermal and mechanical components of exercise can be distinguished by having subjects exercise on a cycle ergometer while immersed to midchest in either hot or cool water (figure 1: not shown) (14,15). Such thermal clamping substantially reduced exercise-induced increases in levels of circulating lymphocytes and granulocytes, but not the increase in levels of monocytes (14). Increments in most circulating leukocyte subsets except B cells and monocytes were reduced (figure 2: not shown) (15), as were increments of plasma epinephrine, norepinephrine, dopamine, growth hormone, and cortisol. These data thus suggest that for most leukocytes there was an additive response from exercise and heat, with parallel increments in "stress" hormone secretion.

Thermal clamping also reduced the exercise-induced secretion of the proinflammatory cytokines interleukin (IL)-6, IL-12, and tumor necrosis factor-alpha (16). Possibly, the lesser secretion of cortisol when the temperature is clamped discourages monocyte migration to the injured tissues. Alternatively, thermal clamping may augment anti-inflammatory, counter-regulatory mechanisms. Since heat exacerbates exercise-induced immune disturbances, the situation could be further compounded by other stressors that augment blood levels of "stress" hormones (17).

Assessments of stress. Ratings of perceived exertion and assessments of thermal comfort deteriorate when a given intensity of exercise is compounded by heat exposure (18). Heart rate variability measurements (19) demonstrated a significant decrease in parasympathetic nervous tone when athletes performed sustained moderate exercise in a cool environment, but the addition of heat to the regimen induced a significant increase in sympathetic nerve activity.

Inhibition of Beta-Endorphins

An exercise-induced secretion of beta-endorphins has an anxiolytic effect, countering the stress associated with vigorous physical activity. Thus, administration of naltrexone hydrochloride (an opioid antagonist) blocks the elevation of mood state that is normally associated with aerobic exercise (20). However, an effective dose of naltrexone prior to a 2-hour bout of exercise at 65% of maximal oxygen intake did not change leukocyte mobilization (21). Likewise, exercise-induced changes in NK cell counts and activity were unaffected by lumbar epidural anesthesia, which blocked the normal increase in beta-endorphin levels (22).

Such observations argue against stress per se as the main mediator of immune responses to exercise. In support of this view, data show that NK cell activity increases during the first hour of exercise, when little change in blood endorphin levels has occurred, but does not increase further in the second hour, as blood endorphin levels rise (21). Nevertheless, beta-endorphin directly affects lymphocytes, at least in vitro (23). Depending on the dose, beta-endorphin may bind to either opioid or nonopioid receptors, giving an inverted-U shaped dose-response curve.

Correlations With Stress Hormones

Sustained moderate exercise induces a release of many hormones, including those involved in a "stress" response (epinephrine, norepinephrine, and cortisol), those concerned with central neural transmission (particularly dopamine), and those involved primarily in metabolic regulation (including human growth hormone). If exercise is performed in a hot environment, levels of norepinephrine, epinephrine, and cortisol are substantially increased (24). Anatomic connections between autonomic nerves and lymph glands (25), and receptors for various hormones on lymphocytes (26), suggest that hormones may mediate both exercise- and heat-induced changes in immune function.

Injection of a single dose of epinephrine mimics many of the changes seen during exercise, though increases in neutrophil counts are not as large (27-29). A similar, though less intense, response is observed after an injection of norepinephrine (5); however, exercise produces little change in lymphocyte counts after administration of the beta-blocker propranolol (30). Such observations support catecholamine involvement, but the parallel between a single injection of a hormone and its progressive release during sustained exercise is obviously incomplete.

A second possible approach is to examine correlations between hormone levels and immune responses. During moderate endurance exercise, the changes in NK cell counts and activity correlated more closely with norepinephrine than with epinephrine levels (24). Further, both hormone levels and cellular responses were exacerbated by a combination of exercise and heat. In a second such experiment (15), the intensity of effort was increased from 50% to 65% of peak aerobic power, and, probably in consequence, the balance of norepinephrine and epinephrine concentrations was altered.

Multiple regression analysis (15) demonstrated significant partial correlations with concentrations of epinephrine (for granulocytes, total lymphocytes, CD3+, and CD8+ counts), with norepinephrine (for NK, CD4+, CD8dim, and CD 19+ counts), and with dopamine (for CD4+ and CD8+ counts). Concentrations of human growth hormone were negatively associated with CD4+, but positively associated with CD8+ numbers. Cortisol was negatively associated with total leukocytes, monocytes, CD19+ B cells, and CD4+ T-cell subsets, but positively related to granulocytes (table 1: not shown).

Although lymphocytes can secrete small amounts of various hormones (31), these quantities are unlikely to affect plasma concentrations significantly. These observations thus support the view that, with the exception of CD8dim and CD 14+ counts, the effects of exercise and core temperature on leukocyte counts are mediated through alterations in the concentrations of "stress" hormones. Given the larger magnitude of hormonal change with combined exercise and heat exposure, correspondingly larger changes in immune response may be anticipated.

Expression of Adhesion Molecules and Cell Trafficking

Unfortunately, the leukocytes in peripheral blood account for only 1% to 2% of the body's total number. Thus, it is important to assess to what degree the changes in circulating cell counts and cell characteristics represent conditions in the body as a whole, and to what degree they reflect an altered expression of lymphocyte surface adhesion molecules and, thus, trafficking of cells between the circulation and tissue reserves.

To date, human studies of exercise and the expression of adhesion molecules have been completed only under warm conditions, but the findings are promising enough to merit repetition under thermally clamped conditions. Exercise combined with heat stress mobilized cells that carry high levels of the adhesion marker CD11a and mobilization of cells appeared proportional to the initial resting densities of this molecule on the cell surface (CD56+ > CD8hi+ > CD4+) (32,33). The CD4+ population had a selective mobilization of memory cells, but the mobilization of CD8+ (cytotoxic/suppressor) cells was distributed much more uniformly between naive and memory cells.

The traditional view has been that leukocyte adhesion molecules are modified by an action of catecholamines on beta-2 adrenergic receptors (34). However, explaining the strong correlations between counts for some lymphocyte subsets and norepinephrine concentrations in such terms is hard, since the beta-2 receptors respond primarily to epinephrine rather than norepinephrine. One possible alternative is that norepinephrine may be modulating leukocyte margination by an action at sympathetic nerve endings in the lymph glands and/or spleen (35-37). Alternatively, norepinephrine may mediate an increase in cardiac contractility and thus an increase of shear forces in the vessels where the leukocytes are sequestered.

Directions for Further Research

Given that cellular and cytokine responses to exercise and other stressors must usually be examined in peripheral blood rather than in the tissues in which immune responses are occurring, the concentrations of many key humoral elements remain below the detection threshold of present immunologic assay kits. Consequently, further investigations of exercise, stress, and immune function should seek to maximize immune changes by exposure to the most severe ethically acceptable combinations of stressors an athlete may encounter, such as exercise, heat, sleep deprivation, and a negative energy balance.

Some immune changes persist after thermal clamping, and these phenomena may reflect persistent increases in circulating levels of "stress" hormones, and thus changes in cellular adhesion processes. However, there may also be effects arising from local inflammatory reactions in active muscles. It would thus be helpful to explore whether residual responses are related to measures of muscle injury such as creatine kinase concentrations, monocyte counts, and plasma nitric oxide concentrations. It would also be helpful to look at the relative production of proinflammatory T-helper 1 cytokines (IL-1, IL-2, IL-6, IL-15, and interferon-gamma), and anti-inflammatory, T-helper 2 products (IL-4, IL-5, IL-6, IL-10), with and without thermal clamping. Finally, it may be worth using recently developed techniques to examine intracytoplasmic levels of IL-1, and to correlate cytokine concentrations with monocyte counts.

The impact of other "stress" hormones on immune responses merits further exploration, and arginine/vasopressin and corticotrophin-releasing hormone concentrations should be tested, as should agents that may counter muscle injury and reduce stress levels.

An Evolving Paradigm

A combination of exercise and heat stress increases sympathoadrenal activation and causes larger changes in circulating hormones than exercise alone; however, clamping core body temperature reduces changes. Thus, a greater disruption of immune function is likely when exercise is performed in the heat and is an additional reason to minimize thermal stress.


  1. Nieman DC, Johanssen LM, Lee JW, et al: Infectious episodes in runners before and after the Los Angeles marathon. J Sports Med Phys Fitness 1990;30(3):316-328
  2. Peters-Futre EM: Vitamin C, neutrophil function, and upper respiratory tract infection risk in distance runners: the missing link? Exerc Immunol Rev 1997;3:32-52
  3. Shephard RJ, Shek PN: Exercise, immunity, and susceptibility to infection: a J-shaped relationship. Phys Sportsmed 1999;27(6):47-71
  4. Brenner I, Shek PN, Zamecnik J, et al: Stress hormones and the immunological responses to heat and exercise. Int J Sports Med 1998;19(2):130-143
  5. Pedersen BK: Exercise lmmunology. Austin, TX, R G Landes, 1997
  6. Shephard RJ: Immune changes induced by exercise in an adverse environment. Can J Physiol Pharmacol 1998;76(5):539-546
  7. Shephard RJ, Castellani JW, Shek PN: Immune deficits induced by strenuous exertion under adverse environmental conditions: manifestations and countermeasures. Crit Rev Immunol 1998;18(6):545-568
  8. Shephard RJ, Shek PN: Immune dysfunction as a factor in heat illness. Crit Rev Immunol 1999;19(4):285-302
  9. Brenner IK, Shek PN, Shephard RJ: Heat exposure and immune function: Potential contributions to the immune response. Exerc Immunol Rev 1995;1(1):49-80
  10. Brenner IK, Severs YD, Shek PN, et al: Impact of heat exposure, moderate, intermittent exercise on cytolytic cells. Eur J Appl Physiol Occup Physiol 1996;74(1-2):162-171
  11. Kappel M, Stadeager C, Tvede N, et al: Effects of in vivo hyperthermia on natural killer cell on natural killer cell activity, in vitro proliferative responses and blood mononuclear cell populations. Clin Exp Immunol 1991;84(1):175-180
  12. Kappel M, Kharazmi A, Nielsen H, et al: Modulation of the counts and functions of neutrophils and monocytes under in vivo hyperthermia conditions. Int J Hyperthermia 1994;10(2):165-173
  13. Kappel M, Tvede N, Hansen MB, et al: Cytokine production ex vivo: effect of raised body temperature. Int J Hyperthermia 1995;11(3):329-335
  14. Cross MC, Radomski MW, VanHelder WP, et al: Endurance exercise with and without a thermal clamp: effects on leukocytes and leukocyte subsets. J Appl Physiol 1996;81(2):822-829
  15. Rhind SG, Gannon GA, Shek PN, et al: Contribution of exertional hyperthermia to sympathoadrenal-mediated lymphocyte subset redistribution. J Appl Physiol 1999;87(3):1178-1185
  16. Rhind SG, Castellani JW, Brenner IK, et al: Intracellular and serum cytokine profiles following exhausting exercise and cold exposure. J Appl Physiol 2001, in press
  17. Brenner IK, Castellani JW, Gabaree C, et al: Immune changes in humans during cold exposure: effects of prior heating and exercise. J Appl Physiol 1999;87(2):699-710
  18. Brenner IK, Thomas S, Shephard RJ: Spectral analysis of heart rate variability during heat exposure and repeated exercise. Eur J Appl Physiol Occup Physiol 1997;76(2):145-156
  19. Brenner IK, Thomas S, Shephard RJ: Autonomic regulation of the circulation during exercise and heat exposure: inferences from heart rate variability. Sports Med 1998;26(2):85-99
  20. Daniel M, Martin AD, Carter J: Opiate receptor blockade by naltrexone and mood state after acute physical activity. Br J Sports Med 1992;26(2):111-115
  21. Gannon GA, Rhind SG, Suzui M, et al: Beta-endorphin and natural killer cell cytolytic activity during prolonged exercise: is there a connection? Am J Physiol 1998;275(6 pt 2):R1725-R1734
  22. Klokker M, Kjaer M, Secher NH, et al: Natural killer cell response to exercise in humans: effect of hypoxia and epidural anesthesia. J Appl Physiol 1995;78(2):709-716
  23. Prete P, Levin ER, Pedram A: The in vitro effects of endogenous opiates on natural killer cells, antigen-specific cytolytic T cells and T cell subsets. Exp Neurol 1986;92(2):349-359
  24. Brenner IK, Zamecnik J, Shek PN, et al: The impact of heat exposure and repeated exercise on circulating stress hormones. Eur J Appl Physiol Occup Physiol 1997;76(5):445-454
  25. Rabin BS, Cohen S, Ganguli R, et al: Bidirectional interactions between the central nervous system and the immune system. Crit Rev Immunol 1989;9(4):279-312
  26. Madden KS, Felton DL: Experimental basis for neural-immune interactions. Physiol Rev 1995;75(1):77-106
  27. Kappel M, Tvede N, Galbo H, et al: Evidence that the effect of physical exercise on NK activity is mediated by epinephrine. J Appl Physiol 1991;70(6):2530-2534
  28. Tonnesen E, Christensen NJ, Brinklov MM: Natural killer cell activity during cortisol and adrenaline infusion in healthy volunteers. Eur J Clin Invest 1987;17(6):497-503
  29. Tvede N, Kappel M, Klarlund K, et al: Evidence that the effect of bicycle exercise on blood mononuclear cell proliferative responses and subsets is mediated by epinephrine. Int J Sports Med 1994;15(2):100-104
  30. Ahlborg B, Ahlborg G: Exercise leukocytosis with and without beta-adrenergic blockade. Acta Med Scand 1970;187(4):241-246
  31. Shephard RJ: Physical Activity, Training and the Immune Response. Carmel, IN: Cooper Publications Group, 1997
  32. Gannon GA, Rhind S, Shek PN, et al: Mobilization of LFA-1a+/- and L-selectin+/- naive and memory, CD4+ and CD8hi T cell subsets during aerobic exercise in the heat. Int J Sports Med 2000; submitted
  33. Gannon GA, Rhind S, Shek PN, et al: Differential cell adhesion molecule expression and lymphocyte mobilization during prolonged aerobic exercise Eur J Appl Physiol Occup Physiol 2001;84(1):272-282
  34. Murray DR, Irwin M, Rearden CA, et al: Sympathetic and immune interactions during dynamic exercise: mediation via a beta-2-adrenergic-dependent mechanism. Circulation 1992;86(1):203-213
  35. Bellinger DL, Ackerman KD, Felten SY, et al: Noradrenergic sympathetic innervation of thymus, spleen and lymph nodes: Aspects of development, aging and plasticity in neural immune interaction, in Hadden JW, Masek K, Nisticò G (eds): Interactions Among Central Nervous System, Neuroendocrine, and Immune Systems. Rome, Italy, Pythagora Press, 1989, pp 35-66
  36. Bullock K, Radojcic T: Characterization of muscarinic acetyl choline and beta-adrenergic receptors on fresh and cloned immunocytes, in Hadden JW, Masek K, Nisticò G (eds): Interactions Among Central Nervous System, Neuroendocrine, and Immune Systems. Rome, Italy: Pythagora Press, 1989, pp 17-34
  37. Iversen PO, Arvesen BL, Benestad HB: No mandatory role for the spleen in the exercise-induced leucocytosis in man. Clin Sci (Colch) 1994;86(5):505-510

Dr Shephard is professor emeritus of applied physiology at the University of Toronto and currently lives in Brackendale, British Columbia. Address correspondence to Roy J. Shephard, MD, PhD, DPE, 41390 Dryden Rd, Brackendale, BC V0N 1H0, Canada; e-mail to [email protected].