TITLE: Thermoregulation during physical exercise: reflexions on exercise-induced hyperthermia

Transcription

1 TITLE: Thermoregulation during physical exercise: reflexions on exercise-induced hyperthermia AUTHORS: M. Curé*, L. Bourdon*, B. Melin * and A. buguet** INSTITUTIONS: *Unité de bioénergétique et environnement et ** Unité de physiologie de la vigilance, Centre de recherches du Service de santé des armées, B.P. 87, LA TRONCHE CEDEX Current title: exercise-induced hyperthermia Key-words: thermoregulation - exercise - hyperthermia - heat stroke ADDRESS FOR CORRESPONDENCE: Médecin en Chef Michel Curé Direction centrale du Service de santé des armées Bureau Recherche 14, rue Saint-Dominique armées Summary The temperature of the body is neither homogeneous nor constant. It is controlled by the hypothalamus, which functions as if comparing a temperature-related error signal, resulting from the integration of central and peripheral receptors, to a referential temperature. The amplitude of the resulting signal conditions the intensity of thermoregulatory responses. During physical exercise, elevated metabolic heat production may lead to hyperthermia (temperature of the body being greater than the referential). The magnitude of this hyperthermia is influenced by exercise intensity (% VO2max), ambient conditions and subject s physical training. Exercise interferes also with thermolysis effectors. In recent years, the existence of a specific brain cooling system has been hypothesized, through a redistribution of the cooled venous blood from the face towards intracanial sinuses and afferent arteries. Fever (temperature of the body equal to the referential) may participate in the development of exercise-induced hyperthermia. This pathological increase in body temperature is related to the release of exogenous (bacteria) or endogenous (interleukines 1 and 6, tumor necrosis factor) pyrogenic products in the blood stream. The intestinal flora may play a role in this process. Therefore, fever may allow the development of heat stroke by decreasing the error signal. Body temperature, which represents an instant image of the thermal state, is not controlled through a stable homeostatic system. It varies throughout the nychthemeron, following a circadian cycle under the influence of physiological and pathological events, of which exercise represents one of the most prominent. However, body temperature is regulated through a specific homeostasis. This is demonstrated, for example, by the reproducibility of exercise-induced thermal reactions in an individual and the difficulty to modify these reactions by exposing the

2 subject to cold or heat. Thermal responses are subject to significant individual variations which must be considered when analyzing exercise-induced hyperthermia. We shall describe the thermal balance of the body, resulting from metabolic heat production and complex heat exchanges with the environment. Thermal regulation depends on a mysterious mechanism with a diencephalic integration level. The cooling of the cephalic structures stimulates this hypothalamic center and induces thermoregulatory responses. The participation of the immune system in the physiological control of body temperature has recently been reported, presenting new concepts in the analysis of exercise-induced hyperthermia. The great impact of pharmacology on thermoregulation will not be treated here, although it may modify fundamental physiological reactions and influence heat hazards. THERMAL BALANCE AND THERMOREGULATION The regulation of body temperature implies that heat gain (thermogenesis, external heat load) and heat loss (thermolysis through convection, radiation and evaporation) are equal. The equation of the thermal balance of the body is presented as follows: Hm ± R ± C ± K - E = S Where: Hm = metabolic heat production, R = heat exchanges through radiation, C = heat exchanges through convection, K = heat exchanges through conduction, which are negligible in the air, E = heat exchanges through evaporation, S = heat storage or heat debt. (Conventionally, the period sign placed above the symbols represents heat flow, i.e., the quantity of heat exchanged by time unit). Body temperature remains stable when heat storage S = 0. Then: Hm = ± R ± C ± K - E The thermoregulatory system can be characterized by its basic components: thermoreceptors, integration center, effectors, and the regulated physiological variable. The regulated physiological variable is not well defined. It is not identified to a local temperature, nor mean body temperature. It may be represented by a flexible and adaptable temperature-related signal, following an unidentified strategy. However, to simplify matters, the regulated variable is represented by core temperature, i.e., the temperature of vital organs such as the head and trunk. Thermophysiologists usually consider the organism as being composed of two compartments: one stable core compartment and one variable peripheral compartment. This

3 consideration, which does not take into account the numerous thermoreceptors located out of the nervous system and the skin, has the advantage of representing satisfactorily thermolytic and thermogenetic responses. Skin, brain and central organ temperatures are integrated in hypothalamic centers by comparing the afferent signal made of a complex association of core (90 % of the resulting signal) and peripheral temperatures. The response of the effectors is also considered as proportional to core temperature, provided that the skin temperature remains constant. Any skin temperature variation will induce a nonlinear reversion in this relationship. The triggering of the system depends on an error signal (load error), which integrates inversely core and peripheral temperatures. In other words, the thermoregulatory system is not stimulated if core temperature increases while peripheral temperature decreases simultaneously. The reverse is also true. Therefore, each core temperature corresponds to a given temperature through which the system remains balanced. For example vasoconstriction occurring with exercise onset may add to the initial passive rise in core temperature by inducing a decrease in skin temperature. Each thermolytic effector may be controlled individually, despite the similarity of the signal produced by hypothalamic thermoregulatory neurons. For example, the onset of sweating is preceded by a release of the vasomotor tone of superficial veins and resistive skin vessels (Fox et al., 1963; Wenger and Roberts, 1980). Thermoregulatory centers appear to recognize as a reference (set point) a physical state related to core temperature. Above or below the set point, a thermoregulatory signal reaches the integrators which control given effectors, forcing them to respond. This notion of a set point represents a valuable mathematical concept in the search for a model; it is still valid and has not yet been replaced by another theory. Most scientists agree to consider that the set point can be modified acutely and temporarily through the action of endogenous pyrogenes, leading to fever (see below). THERMOREGULATION DURING PHYSICAL EXERCISE Metabolic heat production may rise considerably during exercise. Endurance athletes can produce more than 1,200 watts of exceeding heat for hours, without any side effect. The organism has great abilities for heat dissipation, which are triggered by small increases in core temperature. Body temperature augments during exercise. This increase was first described as being proportional to the exercise power (NIELSEN, 1938), which led abusively to the concept of a shift in the set point during exercise. However, when in a steady state, the increase in temperature depends essentially on relative exercise power (i.e., % VO2max) and thermal environmental conditions (DAVIES et al., 1976). This agrees with the hypothesis that the error signal which is necessary to trigger thermolytic processes is due to a passive increase in temperature, reflecting the proportional nature of the control system, which makes a shift in the set point unnecessary (many papers, often contradictory, are reviewed in GISOLFI and WENGER, 1984). A

4 modification of the set point would in fact implicate a uniform shift in the triggering thresholds of all thermolytic responses, which is not the case during exercise (GISOLFI and WENGER, 1984). At a given exercise power, physical training reduces the level of core temperature reached at the steady state plateau, due to an increase in the sensitivity of the sweating mechanism. In this case, as observed during heat acclimation, a downward shift in the set point may occur (NADEL et al., 1974; BAUM et al., 1976; STOLWIJK et al., 1977). A variation of the set point has also been invoked in the circadian temperature variations (STEPHENSON et al., 1984). Considering that exercise does not change the set point, diverse reflexes related to exercise may interfere with the function of thermoregulatory effectors and change their response threshold without any thermoregulatory purpose. An example is given by cardiovascular adaptations and non-thermal influences on sweat glands. Variations in blood pressure, plasma osmolality, water compartment volumes, stimulation of mechanoreceptors or chemosensitive muscular fibers, etc. are susceptible to change thermal responses in a way which is not always reproducible. A transient stimulation of sweating can be observed at the beginning of lower limb exercise (STOLWIJK and NADEL, 1973), probably because of sympathetic stimulation of sweat glands (SATO and SATO, 1981). Skin blood flow is lower when exercising in the heat (ROWELL, 1977, 1983), which may be to maintain the necessary muscle perfusion to cover metabolic needs. Maximum skin blood flow depends on posture, being at its lowest value in the upright position (BRENGELMANN et al., 1977) and lower when sitting than laying down (NADEL et al., 1979). This reflects a certain level of skin vasoconstriction which depends mainly on blood pressure control. In order to lengthen exercise duration, the organism uses a compromise strategy between the thermoregulatory reactions and the muscular metabolic needs. In this condition, the sweatingtemperature relationship remains unchanged and posture changes do not influence the set point (GISOLFI and WENGER, 1984). The exercise progressive dehydration due to thermolytic sweat loss enhances the body temperature increase (GREENLEAF, 1979; MELIN et al., 1990), because of the decrease in plasma volume and the increase in extracellular osmolality. Skin vasodilation and sweating then begin at a higher core temperature (JOHNSON and PARK, 1981; ROBERTS and WENGER, 1980), i.e., the slope of the temperature response curve of the effectors is lowered (FORTNEY et al., 1981). The hyperosmolarity may imply a shift in the set point, provided that the sweating and vasomotor thresholds be uniformly modified (GISOLFI and WENGER, 1984). The extrahypothalamic deep thermoreceptors (HAMMEL and SHARP, 1971) may also modify the response threshold of certain effectors. Working muscle heat production can also be dissipated locally, partly through a passive conductive transfer towards the periphery, but primly through blood convective heat exchanges towards the skin by the communicating veins. Most of the heat produced is thus dissipated without any transfer to the body core. When this mechanism is not sufficient, extra heat is stored in the core and muscles, and the major thermolytic mechanisms (increase in skin blood flow, sweating) are set off. Heat is transferred from the core to the skin, where it is dissipated by sweat evaporation, and radiation and convection to the environment when possible. The quality of the

5 atmosphere, principally regarding humidity, is then of great importance to assure an efficient thermolysis. SELECTIVE CEREBRAL COOLING The concept of selective cerebral cooling has been developed by CAPUTA and CABANAC (1978). These authors attributed a specific thermolytic function to certain elements of the cephalic venous system, although specific anatomic structures described in other mammals, such as the carotid rete mirabilis, were not found in humans. BUGUET et al. (1976) had already stressed the role of countercurrent heat exchanges at the level of the face and neck vessels. The cephalic extremity is fundamental in thermoregulation. The brain, which represents only 2 % of the body weight, is responsible for 20 % of the energy expenditure in a resting subject (120 g of glucose, representing 480 kcal per day). The cerebral tissue is most sensitive to heat, as witnessed by the early behavioral dysfunction in heat stroke. Therefore, the existence of such a developed brain as in man can only be conceived with a proper cooling system (FALK, 1990). The face is more sensitive to heat than the rest of the body (NADEL et al., 1972), with an early vasodilation and sweating, while thermolytic processes are later inhibited in case of dehydration (CAPUTA and CABANAC, 1988). The cooling of the face alone in a hyperthermic subject increases his performance (HIRATA et al., 1987). The anatomy of the cephalic venous system is complex and its function not yet well understood. A system of emissary veins and microvessels on the calvarium surface, especially in the parietal area, seems to play a role in head cooling in hyperthermic states. The direction of the blood stream in these vessels differs with the thermal state of the body. Hyperthermia induces the cooled venous blood to flow inside the skull. The situation is reversed in normothermia or hypothermia (CAPUTA et al., 1978; CABANAC and BRINNEL, 1985). In the hyperthermic subject, the beneficial effect of this cooling mechanism can be deduced from the observed reduction in tympanic temperature (BRINNEL, 1986, 1987, 1989), thought to relate closely to the decrease in brain temperature. This phenomenon is enhanced in the upright position, which lowers intracranial pressure. As stated by IWABUCHI et al. (1986), this could account for the high sensitivity of new-born to hyperthermia. The compression of the angular ophthalmic veins reduces the temperature gradient between the tympanic and trunk temperatures (NAGASAKA et al., 1989). This specific heat exchange mechanism is completed by countercurrent heat exchanges between the carotid blood and the cooled venous blood from the face and nasal mucosa, as postulated by BUGUET et al. (1976). A difference between core temperature measured by a rectal or an esophageal probe and brain temperature may therefore be observed. This may account for the importance of the hyperthermia developed by some marathon runners (MARON et al., 1977) and the absence of any pathological consequence in these athletes. This may also explain the complete recovery in some heat stroked patients.

6 FEVER AND EXERCISE-INDUCED HYPERTHERMIA Regardless of the theory concerning the central control system of body temperature, the concepts of fever and hyperthermia can be differentiated on the basis of the thermoregulatory reactions involved in both cases (SNELL and ATKINS, 1968). Normothermia is defined as the coincidence of the actual body temperature and its reference at the control system level; hypothermia or hyperthermia create a signal of error which triggers the appropriate regulatory responses of the system. Fever is the consequence of an upward shift of the referential system and is accompanied by increased metabolic heat production and reduced heat loss. Shivering preceding the febrile peak and sweating during the defervescence are signs of the translation from a febrile state to a temporary hyperthermic state. The effects of exercise and pyrogenic processes are partly additional, as indicated by the time course of temperature in a subject incubating an infectious disease (personal unpublished data). The febrile state derives from the interaction of exogenous pyrogenic substances (endotoxines, muramyl-dipeptide) and endogenous pyrogenes produced by the stimulated macrophage (interleukines 1 and 6, IL-1, IL-6; tumor necrosis factor, TNF). These substances act on the thermoregulatory centers through prostaglandines (see BERNHEIM, 1986, and STITT, 1986 for a review). The increase in body temperature stimulates the organism defense mechanisms (see KLUGER, 1991, and SIMON, 1991 for a review). Hyperthermia mobilizes white blood cells, especially neutrophiles, stimulates the cytokine production (interferon, interleukines, etc.), facilitates the activation of T cells and lowers plasma iron. These changes are observed during exercise and may prepare the organism to resist to potent aggression. Many peptides which are released during exercise, such as IL-1, interferon or TNF (VITI et al., 1985; CANNON et al., 1986; SHECHTMAN et al., 1988), have pyrogenic properties and may therefore participate in the determination of exercise-induced hyperthermia. In 1983, CANNON and KLUGER showed that human blood sampled during exercise increased body temperature when administered to rats. Also, even when thermoregulatory adjustments are efficient, the hyperthermic effect may be prolonged for hours after cessation of the exercise (HAIGHT and KEATINGE, 1973; CONN et al., 1990). This observation may be related to the fact that pyrogenic peptides released during exercise return slowly to normal values (CANNON et al., 1986). This component of hyperthermia, which fits with the nosologic classification of fever because of the upward shift of the set point, would be added to the hyperthermic effects of the increased metabolic activity of the muscle at work. The origin of endogenous pyrogenes during exercise, i.e., the process leading to the release of cytokines, remains unknown. Following a triathlon, BOSENBERG et al. (1988) demonstrated an increase in plasma lipopolysaccharides (pyrogenic endotoxines released by Gram (-) bacteria). The circadian variations of body temperature, considered by STEPHENSON et al. (1984) as resulting from set point changes, may well be related to cyclic variations of pyrogenic substances such as PGE2 (SCALES and KLUGER, 1987). Such a hypothesis is supported by the following

7 recently published data. The treatment with salicylate (SCALES and KLUGER, 1987) or per os antibiotics (KLUGER et al., 1990) reduces diurnal and nocturnal temperature in healthy animals. In monkeys, GATHIRAM et al. (1987; 1988) described a stress-induced endotoxemia, which was prevented by antibiotic (per os kanamycine) or corticoid administration. Intestinal antibiotics are known to improve survival time in experimental heat stroke in dogs (BYNUM et al., 1979). The development of endotoxine resistance in rats is accompanied by a better tolerance to heat stress, during which Gram (-) intestinal flora has been shown to increase (DUBOSE et al., 1983). In heat stroke patients, there is an augmentation of blood lipopolysaccharides (GRABER et al., 1971). Therefore, knowing that hyperthermia improves microbial growth and activity, one may hypothesize that the intestinal flora may modulate the increase in body temperature during exercise, with differences related to the physiological or physiopathological condition of the individual. Any latent or patent inflammatory state, interfering strongly with body temperature regulation, may induce fever and lead to the development of heat stroke. CONCLUSION The exercise-induced increase in body temperature is a regulated process which is proportional to the relative power of the exercise. This mechanism integrates the thermal sensory signal processing with exercise-dependent central factors and humoral endogenous factors participating to the defense of the organism. A regulated stress hyperthermia (LONG et al., 1990), may precede the exercise-induced body temperature increase (RENBOURN, 1960), which adds to the complexity of the system. In this context, the continuum between exercise-induced hyperthermia and heat stroke may be related to a disturbance in the regulation of temperature control, with undiscernible borders between physiology and physiopathology. Therefore, the development of heat stroke does not require any muscle anomaly, which may however trigger it and impair its evolution.