Introductory paragraphs

Functional vs anatomical view of the body core and shell

Uniformity of arterial and core organ temperatures

Rapid dynamics of core temperature

Core temperature uniformity in the heart failure patient

Skeletal muscle as a thermal compartment

Thermal dynamics of skeletal muscle: a store of " cold"

Thermal dynamics with exercise in heart failure patients

The cutaneous vasculature as the body's heat exchanger

The three zones of thermoregulation

Thermoregulatory control of the cutaneous vasculature

Dilator potential of the cutaneous vasculature

The neutral zone and "normal" body temperature

The zone of hypothermia

The zone of hyperthermia

The changes in the cardiovascular system that develop in chronic heart failure obviously affect thermoregulation. A constant, normal body temperature is the consequence of a regulated balance between heat production and heat loss. Both sides of this balance are altered in chronic heart failure. These alterations are likely to differ with different degrees of severity, different etiologies, and particular regime of treatment. But the literature includes few studies that directly adress questions related to thermoregulation in heart failure patients. This review has been prepared according to the assumption that what might be useful to readers who have the perspective that comes with clinical and or laboratory experience with the whole spectrum of heart failure would be a review of the fundamentals of thermoregulation with a focus on questions of how the mechanisms might be influenced in heart failure.

The discussion begins with the question of how the pattern of distribution of the limited cardiac output of a heart failure patient among the major vasculatures might alter body temperatures and the dynamics of response to altered thermal conditions and exercise. Then it focuses on the effects on the cutaneous vasculature in its role as a heat exchanger. Finally, overall responses to thermal stress are considered in terms of three zones of thermoregulation: clear disadvantages of heart failure in cold stress; likely narrowing of the range of " neutral" environments; and, in hyperthermic conditions, the possible consequences of unleashing the dilator potential of the cutaneous vasculature.

The high ratio of blood flow to mass and to metabolism in the lungs, brain, heart, kidneys and splanchnic organs makes them, altogether, a core of nearly uniform temperature represented by temperature in the pulmonary artery, Tpa. Although these ratios change in heart failure, the changes are not so great as to disturb the basic pattern of core temperature uniformity. Reduced blood flow in the periphery, however, may alter the relationship between Tpa and temperatures in the various sites used for non-invasive monitoring of what is called generically core temperature, Tc.

Usually we think of the body core temperature in an anatomical context. We see the interior of the head and trunk as a central core protected from the thermal influence of the environment and, therefore, at fairly uniform temperature, Tc. The various anatomical sites from which practical measures of Tc are obtained, e.g., the rectum or oral cavity, are chosen because they are thought to be deep enough to be within the core. The superficial regions and the limbs are a "shell" that functions to insulate the core. In this shell, temperatures are intermediate between Tc and the temperature of the skin, Tsk.

Another way to look at core temperature, more useful for considering the changes that may occur in chronic heart failure, is according to the functional organization of the organ vasculatures illustrated in the block diagram in Figure 1. In this view, the optimal measure of Tc is the temperature of arterial blood, Tart, and the non-invasive sites of measurement are evaluated according to how closely they follow Tart (1).

To begin with, temperature in the pulmonary artery is, for all practical purposes, identical to temperature in the major systemic conduit arteries. Of course, the cardiac output is vast in relation to the heat production of the pulmonary parenchyma. It is also vast in relation to the rate of respiratory heat loss -- several kilocalories of heat per hour extracted from hundreds of liters of cardiac output per hour would result in a temperature decrement of only hundredths of a degree -- but, the fact is that the water vapor in expired air evaporates from surfaces of the airways and nasopharynx. Consequently, the effect of the evaporative cooling is felt in the bronchial, not the pulmonary, vasculature (2).

The cardiac output is also vast relative to any heat taken up from the metabolism of the left ventricle. Therefore, blood passes through the aorta at a temperature nearly identical (within .01° C) to pulmonary artery temperature, Tpa (3). Furthermore, we can consider the temperature in the conduit arteries that supply the organs as uniform. The heat exchange across the walls of these vessels is insignificant because blood passes through them at high linear flow velocity (4). Also, the temperature gradient across the vessel walls is small. The exceptions to these rules are the distal arteries of the limbs.

What happens to blood temperature in the transit through the organs? A convective steady state is possible in which all the heat produced by an organ is carried away by the blood. Venous blood leaves the organ warmed by the metabolic heat removed from the organ. The temperature increment relative to arterial blood depends upon the ratio of organ metabolism to blood flow. The expected increment can be crudely estimated as numerically equal, in degrees Centigrade, to the fractional extraction of oxygen.

In reality, in the organ vasculatures in which flow rates are in the order of hundreds of ml/min per 100g of organ mass, (cerebral, coronary, splanchnic, renal) heat exchange is indeed dominated by this convective mode of heat transfer. Only a few tenths of a degree increment between venous and arterial blood temperature suffices to remove organ metabolic heat as rapidly as it is produced (5-7).

Also, the extensive network within organs of arterial distribution vessels with diameters greater than 1 mm minimizes countercurrent trapping of heat. It is vessels in the hundred micron diameter range in which significant heat transfer across vessels first occurs. Consequently, the intra-organ arterial distribution vessels deliver blood to the microvessels at very nearly the temperature of arterial blood, i.e. Tpa, and thermal equilibration occurs within the local microcirculations (8). The effectiveness of this convective process is seen in the near-uniformity of temperatures recorded within, for example, the brain of the rheusus monkey (Figure 2). Even closer uniformity is to be expected in the splanchnic and renal vasculatures in resting conditions in which fractional oxygen extraction is low.

Another consequence of organ flow rates in the order of hundreds of ml/min per 100g of organ mass is rapid thermal dynamics. Organ temperatures follow changes in Tpa with little lag. The time constant can be crudely estimated, in units of minutes, as the reciprocal of the flow in units of ml/min per 100g of organ mass. For example, the time constant of the lag of response of brain temperature relative to Tpa would be estimated as 2 min, taking brain blood flow as 50 ml/min per 100g of organ mass (9).

These factors behind uniform core temperature and rapid thermal dynamics in healthy people are altered in application to the heart failure patient but not so much so as to disturb the basic pattern within the arterial distribution system and core organs. Despite reductions in flow:metabolism and flow:mass ratios, the upper limit in oxygen extraction limits the maximal a-v temperature difference across organs. Also, the anatomy of the arterial and venous distribution systems within organs is not altered. The vascular changes likely to alter thermal dynamics and distribution of body heat and temperatures in heart failure are those that occur in skeletal muscle and skin, discussed below as components of the body shell.

How temperatures in the non-invasive sites used to sample Tc compare with Tpa has not been studied with specific reference to heart failure. Changes are likely. For example, tympanic temperature is influenced by temperature of the skin of the head (10), probably through heat exchange between superficial venous drainage and the small arterial twigs that supply the external auditory meatus; a process that might be altered through peripheral vasoconstriction in chronic heart failure. Also, temperatures in the rectum may respond to changes in Tc even more slowly than in healthy subjects if blood flow to this region is reduced.

In terms of mass, three components are important in the shell: fat, bone, and skeletal muscle. The skin, obviously part of the shell, is insignificant in terms of mass but all-important in control of heat loss from the body. The role of the cutaneous vasculature in thermal dynamics is discussed separately, below.

Fat and bone do not appear in the diagram of the organ vasculatures in Figure 1 because their share of the cardiac output is in the 5% labeled "other." Nonetheless, fat has an important thermal role, of course, because the subcutaneous component of total body fat functions as an insulating layer. The large mass and correspondingly high heat capacity of skeleton , combined with low blood flow must make it a slow-responding heat sink or source for the core organs in dynamic thermal situations. Little is known of how the heat content of these thermal compartments changes during long-term discrepancies between body heat production and heat loss. To the extent that the cachexia of chronic heart failure includes depletion of the subcutaneous fat layer, there will be a corresponding loss in ability to tolerate cold environments.

From an organ system view of the cardiovascular system and of heat distribution within the body, resting skeletal muscle belongs to the shell.

In resting skeletal muscle, flow is extremely low, relative to mass; near a mere 2 ml/min per 100g of muscle (11). Correspondingly low linear flow velocities in the arterial system allow greater heat transfer across the walls of conduit and distribution vessels. In superficial muscle and muscle of the distal limbs, heat transfer is less dominated by convection. Heat can move by conduction down gradients between intramuscular and surface temperature.

Overall, this organ system receives a large fraction, near 20%, of the cardiac output of a healthy resting person (11) and generates a large fraction of the total body heat production. If it did function as a single thermal compartment in thermal equilibrium through convective removal of heat, muscle venous blood would be nearly 1° C warmer than the arterial supply since oxygen extraction by resting skeletal muscle is high, 70% or more (11). But, because heat can leave skeletal muscle by routes other than venous blood, venous blood from resting muscle may return to the central circulation considerably cooler than arterial blood; Aulick, et al., recorded temperatures in the femoral vein over 1° C below temperature in the femoral artery in young men (12)

Skeletal muscle is a highly variable thermal compartment. The low blood flow:mass ratio at rest allows appreciable non-convective heat loss. But, when muscle contracts in exercise, blood flow increases enormously up to maximal levels in the order of hundreds of ml/min per 100g of muscle. During maintained exercise, the active muscles become part of the thermal core, in the sense that the high flow rates result in dominance of convective heat transfer via the vascular system -- temperature in active muscle will stabilize a fraction of a degree higher than Tpa within several minutes after onset of exercise. Evidence of this can be seen in the relationship between muscle temperature and esophageal temperature obtained over a range of exercise levels by Saltin and Hermansen (13).

This thermal variability of skeletal muscle causes complex thermal dynamics in healthy people and in patients in chronic heart failure. When muscle is first activated, the mass of tissue may be at temperatures well below Tpa. Functionally, this amounts to a reservoir of "cold" that is tapped when muscle blood flow increases. The blood that perfuses muscle during the immediate increase in blood flow that occurs with onset of rhythmic contractions (14) equilibrates with these low temperatures, contributing a greatly increased fraction of cooled blood to the total venous return.

This phenomenon is revealed in the immediate drop in Tpa (or esophageal temperature, Tes, which closely follows Tpa) seen at the onset of exercise in healthy people, particularly at lower levels of exercise (15,16). As exercise continues, the increase in metabolic heat production rapidly warms the active muscles above Tpa and the venous return from the active muscles causes rapid rise in Tpa (15).

When exercise stops, blood flow falls rapidly. The warmed muscles lose heat thereafter at the slow rate expected from the low flow:mass ratio in inactive muscle (16). During this long recovery, muscle temperatures can exceed Tpa; consequently, the venous blood draining these regions tends to keep Tpa elevated relative to pre-exercise levels.

With their limited exercise potential, people in heart failure are generally not capable of sustained exercise that would substantially increase Tpa. Nonetheless, whatever activation of skeletal muscle they can accomplish will result in a core temperature dynamic. With their reduced muscle mass and bare minimum of flow in inactive muscle, the muscles of the extremities can be expected to be cooler than muscles of healthy people in identical environments that are neutral to cool.

Indeed, Shellock et al., in a study illustrated in Figure 3, observed lower muscle temperatures in heart failure patients, compared to a control group of healthy individuals of similar age (17). These patients were able to achieve rates of oxygen consumption two or three times their resting levels during treadmill exercise. With onset of exercise, core temperature, measured as Tpa, fell (typically 0.5° C). Even at these low levels of exercise, the comparison group exhibited increases in Tpa. Muscle temperature rose in the patient group, but did not exceed Tpa, whereas the healthy subjects exhibited the expected increment of muscle temperature above Tpa.

In summary, a bout of exercise that would warm a healthy person somewhat and leave the person at an elevated core temperature for an hour or more afterwards would result in a fall in core temperature in a heart failure patient during exercise followed by a slow recovery to his or her original core temperature. Another situation in which activation of muscle must influence Tc in heart failure patients is shivering -- discussed below in the section on hypothermia

The concern in care of the heart failure patient for effects on skin blood flow in the range between nearly zero (extreme vasoconstriction in defense against hypothermia) to a few hundred ml/min, (release of sympathetic vasoconstrictor tone) is not the cardiac stress but the interference with thermoregulation. Under conditions of direct heating of the skin or of systemic hyperthermia, cutaneous resistance vessels dilate to extremes that have the potential of major cardiac stress. The following discussion of the cutaneous vasculature begins with its role as a heat exchanger in thermoregulation.

The fraction of the cardiac output destined for the skin leaves the central circulation at core temperature, i.e., Tpa. It returns equilibrated with skin temperature, Tsk. It mixes with the rest of the venous return which, in part, has been warmed relative to Tpa by the heat removed from the organs and may include a contribution from skeletal muscle that is below Tpa. The temperature of the mix is Tpa. As long as Tsk is lower than Tpa, the skin’s share of the cardiac output removes heat from the body. At a certain specific level of skin blood flow for a given difference between Tpa and Tsk, heat removal rate will balance metabolic rate (minus the heat lost evaporatively in expired air).

In a convective heat exchanger, heat removed is proportional to the product of the temperature difference and the flow. A smaller difference between Tpa and Tsk requires greater skin blood flow for the same heat transfer rate in inverse proportion. If the body is in thermal balance and some disturbance increases skin blood flow, the rate of heat transfer to the skin surface will increase, tending to lower Tpa (because of increased rate of removal of heat from the core) and elevate Tsk (because of increased delivery of heat to the surface that results in increased gradient between Tsk and environmental temperatures). Vice versa, a disturbance that decreases skin blood flow, such as a baroreceptor-mediated vasoconstriction of cutaneous vessels, results in trends toward higher Tpa and lower Tsk.

Three zones of thermoregulation in humans are defined by capabilities of the cutaneous vasculature.

Below a certain Tsk, vasoconstriction is insufficient to compensate for the increased Tpa:Tsk difference in cooler environments. Above a certain Tsk, the release of cutaneous vasoconstriction yields insufficient blood flow increase to compensate for the decreased Tpa:Tsk difference. Between these Tsk limits, approximately a range of 33 - 35° C, modulation of skin blood flow suffices to maintain thermal balance, provided that metabolic rate remains near resting levels. This is the "neutral zone" of human thermoregulation (18,19).

If environmental conditions push Tsk below the lower limit, heat loss exceeds heat production despite maximal cutaneous vasoconstriction. Core temperature falls. The only physiological defense in this zone of hypothermia is increase of metabolism. Shivering is activated. This is the "zone of metabolic regulation" in which at least some degree of hypothermia occurs -- for present purposes this zone will be referred to simply as the "zone of hypothermia".

If environmental conditions push Tsk beyond the upper limit, or if exercise elevates heat production, control of skin blood flow enters another realm. In this zone, variously termed the "zone of evaporative regulation" (20) or "zone of sudomotor and vasomotor regulation" (21) elsewhere, but here simply the "zone of hyperthermia", sweating is elicited through activation of the cholinergic sympathetic fibers that innervate the sweat glands. This powerful mechanism for dissipation of heat at the body surface is accompanied by activation of the specialized dilator system in the cutaneous vasculature (22,23), discussed below after a brief mention of the thermoregulatory reflex control scheme.

A similar regulatory scheme works in all three zones of thermoregulation and, as far as is known, applies to thermoregulatory control of skin blood flow as well as shivering and sweating. These effectors are driven in response to thermal inputs derived from the skin as well as from the core.

The reflex response to core temperature is, generally, the familiar sort of proportional control -- sweating and skin blood flow increase in proportion to Tc increases, shivering increases in proportion to decline in Tc. Skin temperature makes an important contribution; for a given Tc, higher Tsk increases sweating and skin blood flow, lower Tsk increases shivering.

In animals, thermosensitive units that respond to core temperature and elicit regulatory responses have been found in the spinal cord as well as the pre-optic anterior hypothalamus (24). No direct evidence of thermosensitive units in the central nervous system of humans has ever been obtained, but if we assume that the central thermosensitive sites are well-perfused, i.e., with flow:mass and flow:metabolism ratios similar to those in the brain, it follows that they will follow arterial blood temperature and will be accurately reflected in Tpa.

When supine, healthy, young men were made hyperthermic by means of a suit perfused by hot water that brought skin temperature to over 39° C, the active vasodilator response in the cutaneous vasculature increased in proportion to the steady increase in Tc (25). With Tc increases over 2° C, the associated increase in cardiac output, which at least somewhat underestimates the total increase in skin blood flow, was over 7 liters/min. Right atrial mean pressure fell 5 mm Hg (26). In these supine healthy young subjects, blood pressure was maintained despite the large reductions in systemic vascular resistance.

Extreme cutaneous vasodilation also occurs with localized heating of the skin. For example, if the forearm is sprayed with 42° C water, blood flow begins to increase immediately in the heated region of skin and, after about 20 min, reaches the extreme high levels that have been observed in the reflex initiation of active cutaneous vasodilation (27). The maximum vascular response to local heating occurs with elevation of local skin temperature to 42° C. The specific rates of blood flow attained are an individual characteristic of a person’s skin that can be reproduced in experiments separated by days or months (28). The response seems to decline with age -- a group of men of average age 71 exhibited greatly reduced maximal forearm blood flow in response to local heating, compared to a cohort of young men (29).

The cutaneous vascular response to local heating and the reflex response to high Tc are mediated by two distinct mechanisms. The dilator response to local heating is locally mediated, totally independent of activity in sympathetic nerves. The mechanisms are equally potent; either local heating or high Tc with neutral local skin temperature elicits the same maximal rate of blood flow in the skin. The same maximum is achieved when these individual influences act together, i.e., when local heating and high Tc are combined (27). Evidently, they are capable, individually or in combination, of reducing tone in the smooth muscle of cutaneous resistance vessels to a physiological minimum.

These influences are subject to modulation in reflex responses to disturbances of blood pressure. For example, blood flow in the heated forearm of a supine normothermic person falls when negative pressure is applied to the lower body to simulate the effects on blood volume distribution of upright posture (30); this stimulus also results in reduction in forearm blood flow elevated through active cutaneous vasodilation (31) in a supine hyperthermic person. Using selective blocking drugs, Johnson and colleagues have addressed questions such as whether this capability to offset the dilator influence is achieved through modulation of activity in the vasodilator fibers as opposed to superimposition of vasoconstrictor influence via the adrenergic innervation of cutaneous resistance vessels; see, for example reference (32).

The "neutral zone", in which body heat production of a resting person is balanced with heat loss solely by regulation of skin blood flow, is narrow in healthy adults. By adjusting our behavior and clothing, we manage to be comfortable in environments ranging from arctic cold to desert heat, but this accommodation to a wide range of thermal stress depends upon achieving skin temperatures within a narrow range. In classical studies with humans in whole-body calorimeters, that range was found to be approximately from 33 to 35° C, in terms of a weighted mean of temperatures over the body surface (33).

It is in this zone of neutrality that we expect to find "normal" body temperature. We would not decide a temperature was normal based on measurements obtained while a person was exercising, sweating, or shivering. We recognize that the equilibrium Tc maintained within the neutral zone defines normal core temperature.

But, this implicit recognition of necessary conditions for normal Tc may not include appreciation that the particular level of equilibrium body temperature depends upon a particular level of skin blood flow. Anything that alters skin blood flow tends to alter the equilibrium Tc in a given environment. A simple change of posture alters Tc a few tenths of a degree in a normal person (34), the result of the reflex adjustment elicited by the shift in blood volume. All those influences on blood pressure and blood volume that elicit altered sympathetic tone in cutaneous vessels can be expected to influence Tc (18).

Obviously, the specific level of Tsk must make a difference in Tc in the neutral zone as well. However, esophageal temperature (Tes) changed only 0.1° C when Tsk was switched from 33 to 35° C in a recent study (19). Surprisingly, the direction of Tes change was opposite to the direction of Tsk change -- according to this finding, Tc falls slightly when Tsk rises, within the neutral zone range. Apparently, the reflex influence of Tsk on skin blood flow outweighs the effect of the altered Tc:Tsk gradient in this "feedforward" control scheme.

These examples of thermally and non-thermally induced changes in Tc are no great cause for excitement such as alarm about possible misdiagnosis of a fever. However, they occur in the setting of one functional regulatory system interacting with another -- tendencies for large changes in Tc initiated in perturbances of the cardiovascular system are countered by the powerful thermoregulatory influence of Tc on skin blood flow. The potential for change in Tc is much larger.

Specifically, consider the observation of the small change in Tc (0.1° C) that occurs when Tsk is driven from one boundary of the neutral zone (33° C) to the other (35° C). It is the regulatory response to the Tsk change that makes the actual change in Tc so small (18,19). If no compensatory adjustment in skin blood flow had occurred, the expected change in Tc would have been equal to the change in Tsk, two degrees (because the level of heat production supports a particular Tc:Tsk gradient for a particular level of skin blood flow). We can extrapolate the observation to predict that Tc should fall nearly 2° C if Tsk were kept fixed at 33° and blood flow through the skin were elevated to the level associated with 35° C.

In short, depriving a person of normal control of the cutaneous vasculature has the potential consequence of major changes in Tc even in what are normally considered thermally neutral conditions. The drugs that are given to heart failure patients to alter their peripheral resistance must alter equilibrium Tc level and must have the effect of narrowing their neutral zone.

What are the effects on Tc regulation in neutral conditions of ACE inhibitors, of nitrate vasodilators, of alpha blockers? To the extent that they interfere with the control of skin blood flow, they must alter Tc. The only support for this in the literature that could be found was the lower average of control temperature in the group of heart failure patients studied by Shellock, et al (17). Perhaps people in heart failure learn through experience that their neutral zones are narrower and behave accordingly. But, perhaps their tendency to have lower Tc has been overlooked.

Notwithstanding the measures heart failure patients can take to protect their body temperature, the emphasis in their treatment of interfering with peripheral vasoconstriction must have the side effect of predisposition to hypothermia, particularly in consideration of the reduced level of heat production in advanced cases. On physiological principles, they must be at risk for debilitating hypothermia in environments that are merely annoyingly cool for healthy people. Once they become hypothermic enough to shiver, their shivering inefficiency, discussed below, must cause even more rapid cooling.

No direct experimental observations of consequences of deliberate exposure of heart failure patients to systemic hypothermia were found in the literature search performed for this review. Houdas, et al., cite records of increased mortality in winter (35) in ischemic heart disease patients in failure.

A separate matter, not part of the physiology of whole-body thermoregulation, is the effects of sudden cooling, i.e. stresses similar to that of the cold pressor test in which a region of skin is cooled suddenly to ice water temperatures. The physiological consequences in patients with coronary atherosclerosis include increased arterial pressure and coronary vasoconstriction (36,37). Despite their already elevated level of sympathetic nervous activity, patients in chronic heart failure respond to the cold pressor test with increased blood pressure and heart rate (38,39)

Even in a healthy person, shivering is a wasteful physiological effector. The obligatory increase in blood flow to skeletal muscle and the shaking movements increase heat loss. Despite the increased heat production, Tc is likely to continue falling.

In heart failure, onset of shivering is an obvious disadvantage simply because of the additional cardiac output necessary to supply the activated muscles. Besides that, shivering would immediately exacerbate the hypothermia. The observation by Shellock, et al., of Tc decline with onset of exercise in heart failure, cited above (17), logically extrapolates to this expectation.

When Tsk rises above the upper limit of the neutral zone, the increase in skin blood flow that occurs is insufficient to compensate for the reduced Tpa:Tsk gradient. Core temperature rises and active cutaneous vasodilation is elicited along with sweating. Besides the decline in peripheral vascular resistance, increased flow in the cutaneous vasculature increases blood volume in the cutaneous venous system which tends to cause reduction of right atrial pressure (26).

Effects of hyperthermia include a direct effect on heart rate. Temperature of the cardiac pacemaker cells increases along with Tc. Jose observed heart rate increases near 7 beats per min per ° C in subjects treated with pharmacological blocking agents to eliminate influence of the autonomic nervous system (40).

One hears anecdotes of increased admissions and mortality in heart failure patients during hot weather. Houdas, et al., cite reports of increased mortality in ischemic heart failure patients in heat waves (35), but, otherwise, no citations of mortality statistics related to hot weather were found.

Focussing on the cutaneous vasodilation, one might expect hyperthermia to be beneficial in heart failure patients -- a physiological way of reducing afterload and central venous pressure. On the other hand, the potential for reduction in peripheral resistance may outstrip the cardiac capability to maintain blood pressure.

Whether either or both of these cutaneous dilator influences is or are attenuated or altered by the drugs used to manipulate peripheral vascular resistance in heart failure patients is unknown. The experimental observation that reflexes of the sympathetic nervous system can, to some extent, override these dilator influences allows the hypothesis that a generalized vasoconstrictor influence on the periphery associated with the subnormal cardiac output in chronic heart failure would reduce the cutaneous dilator effect of a given level of systemic hyperthermia or high local skin temperature. Perhaps the innate potential of the locally-mediated dilator mechanism is attenuated in chronic heart failure, in analogy to the decline in dilator potential of local heat that has been observed in aged men (29).

Numerous studies hint at changes in regulation of peripheral vascular tone that might occur, e.g. altered NO release from the endothelium (41-43), and interaction between prostaglandins and ACE inhibitors (43,44). Whether these have implications for the control of the cutaneous circulation is not clear because the blood flow responses recorded cannot be resolved into contributions from skin versus muscle. None of these studies included specific attention to altered effects of local or systemic hyperthermia.

Another unanswered question about possible effects of ACE inhibitors rises from the possible role of the renin-angiotensin system in the normal response to systemic hyperthermia. Kosunen, et al., reported that hyperthermia activated the renin-angiotensin system in healthy young adults (45). The consequences of blocking this response in a heart failure patient through treatment with ACE inhibitors is unknown. But this unanswered question compounds another unanswered question. We do not know the physiological significance of the increased angiotensin II levels that develop during the normal response to hyperthermia For example, Escorrou, et al., were unable to discern differences in the splanchnic vasoconstrictor response in hyperthermic young men after renin release was reduced by administration of propanolol (46).

Whole-body responses to combined local and systemic heating have been recorded in heart failure patients during exposure to sauna or hot water baths (47). Tei, et al., reported increases in cardiac output and heart rate with no increase in systolic pressure in their heart failure patients in these conditions (60° C air or 41° C water). Effects on preload and diastolic pressure differed; both declined in the sauna exposure, but not in the hot bath in which right atrial and pulmonary capillary wedge pressures increased and diastolic pressure was not significantly changed. Evidently, the hydrostatic consequences of water immersion prevented the displacement of blood volume to the cutaneous venous system that must have been the cause of reduced preload in the exposure to hot air.

The authors interpreted these findings as revealing that dilation of the cutaneous vasculture is indeed a physiological means of reducing afterload with desireable consequences, possibly even potential for practical therapy. Furthermore, besides the physical findings, patients reported feelings of well-being after the experimental sessions and said they slept better. The investigators expressed no concern about the heart rate increases (20-25 beats/min) that developed by the end of the period of hyperthermia.

One clear difference between this approach to afterload reduction via increased conductance in a single vascular bed as opposed to the generalized conductance increase that presumably occurs with nitrate vasodilators is that flow must not increase in the renal and splanchnic vasculatures. In fact, if these responded in the same way as in hyperthermic healthy young adults (11,48), splanchnic and renal blood blood flow must have fallen in the heart failure patients of Tei, et al. Even with the effect on vascular resistance confined to the skin, it seems to this reviewer that the cardiac output increase in the hyperthermic patients was wasted on an organ system that did not need it.

Physiological changes that accompany chronic heart failure as well as the treatment regime influence thermoregulation in the three zones of thermal stress.

The upper boundary of the zone of hypothermia must extend to include environments that, at worst, are annoyingly chilly for people with normal cardiovascular systems but cause shivering in heart failure patients. The cardiac stress associated with the onset of shivering is likely to be compounded by increased thermal stress through enhanced rate of decline of core temperature and a consequent stimulus for still more shivering.

In the neutral zone, characterized by the range of skin temperatures over which core temperature is maintained by control of skin blood flow, long-term physiological adaptations in the control of systemic vascular resistance must alter the normal body temperature. Pharmacological interventions have the potential to deprive the patient of the ability to maintain thermal balance; elimination of cutaneous vasoconstriction could result in a 2° C decline in Tc in what would normally be a neutral environment.

In the zone of hyperthermia, in which warm environments cause elevated skin temperatures (hyperthermia due to endogenous heat production, i.e., exercise, is not discussed in this review), Tc increases as a consequence of elevated Tsk. Thermal steady states are possible in which the combined stimuli of elevated Tsk and Tc elicit reflex sweating and cutaneous vasodilation adequate to achieve thermal balance. The vasodilator component of this response is also undoubtedly influenced by the physiological changes and pharmacological interventions associated with heart failure and its treatment. Nonetheless, a vasodilator response can be elicited that has been shown experimentally to result in improvement of cardiac output similar to that observed with pharmacological vasodilator treatment but possibly with the difference that the increase in cardiac output is entirely diverted to the cutaneous vasculature.

Drs. Wayne Levy and Daniel Fishbein of the Division of Cardiology and Drs. Loring Rowell and Margaret Savage of the Department of Physiology, all at the University of Washington School of Medicine, gave valuable information, counsel and criticism in aid of preparation of this review.