A century of exercise physiology: concepts that ignited the study of human thermoregulation. Part 3: Heat and cold tolerance during exercise

In this third installment of our four-part historical series, we evaluate contributions that shaped our understanding of heat and cold stress during occupational and athletic pursuits. Our first topic concerns how we tolerate, and sometimes fail to tolerate, exercise-heat stress. By 1900, physical activity with clothing- and climate-induced evaporative impediments led to an extraordinarily high incidence of heat stroke within the military. Fortunately, deep-body temperatures > 40 °C were not always fatal. Thirty years later, water immersion and patient treatments mimicking sweat evaporation were found to be effective, with the adage of cool first, transport later being adopted. We gradually acquired an understanding of thermoeffector function during heat storage, and learned about challenges to other regulatory mechanisms. In our second topic, we explore cold tolerance and intolerance. By the 1930s, hypothermia was known to reduce cutaneous circulation, particularly at the extremities, conserving body heat. Cold-induced vasodilatation hindered heat conservation, but it was protective. Increased metabolic heat production followed, driven by shivering and non-shivering thermogenesis, even during exercise and work. Physical endurance and shivering could both be compromised by hypoglycaemia. Later, treatments for hypothermia and cold injuries were refined, and the thermal after-drop was explained. In our final topic, we critique the numerous indices developed in attempts to numerically rate hot and cold stresses. The criteria for an effective thermal stress index were established by the 1930s. However, few indices satisfied those requirements, either then or now, and the surviving indices, including the unvalidated Wet-Bulb Globe-Thermometer index, do not fully predict thermal strain.

Human cold habituation: Physiology, timeline, and modifiers

Habituation is an adaptation seen in many organisms, defined by a reduction in the response to repeated stimuli. Evolutionarily, habituation is thought to benefit the organism by allowing conservation of metabolic resources otherwise spent on sub-lethal provocations including repeated cold exposure. Hypermetabolic and/or insulative adaptations may occur after prolonged and severe cold exposures, resulting in enhanced cold defense mechanisms such as increased thermogenesis and peripheral vasoconstriction, respectively. Habituation occurs prior to these adaptations in response to short duration mild cold exposures, and, perhaps counterintuitively, elicits a reduction in cold defense mechanisms demonstrated through higher skin temperatures, attenuated shivering, and reduced cold sensations. These habituated responses likely serve to preserve peripheral tissue temperature and conserve energy during non-life threatening cold stress. The purpose of this review is to define habituation in general terms, present evidence for the response in non-human species, and provide an up-to-date, critical examination of past studies and the potential physiological mechanisms underlying human cold habituation. Our aim is to stimulate interest in this area of study and promote further experiments to understand this physiological adaptation.

Skin-surface cooling elicits peripheral and visceral vasoconstriction in humans

Skin-surface cooling elicits a pronounced systemic pressor response, which has previously been reported to be associated with peripheral vasoconstriction and may not fully account for the decrease in systemic vascular conductance. To test the hypothesis that whole body skin-surface cooling would also induce renal and splanchnic vasoconstriction, 14 supine subjects performed 26 skin-surface cooling trials (15-18 degrees C water perfused through a tube-lined suit for 20 min). Oral and mean skin temperature, heart rate, stroke volume (Doppler ultrasound), mean arterial blood pressure (MAP), cutaneous blood velocity (laser-Doppler), and mean blood velocity of the brachial, celiac, renal, and superior mesenteric arteries (Doppler ultrasound) were measured during normothermia and skin-surface cooling. Cardiac output (heart rate x stroke volume) and indexes of vascular conductance (flux or blood velocity/MAP) were calculated. Skin-surface cooling increased MAP (n = 26; 78 +/- 5 to 88 +/- 5 mmHg; mean +/- SD) and decreased mean skin temperature (n = 26; 33.7 +/- 0.7 to 27.5 +/- 1.2 degrees C) and cutaneous (n = 12; 0.93 +/- 0.68 to 0.36 +/- 0.20 flux/mmHg), brachial (n = 10; 32 +/- 15 to 20 +/- 12), celiac (n = 8; 85 +/- 22 to 73 +/- 22 cm.s(-1).mmHg(-1)), superior mesenteric (n = 8; 55 +/- 16 to 48 +/- 10 cm.s(-1).mmHg(-1)), and renal (n = 8; 74 +/- 26 to 64 +/- 20 cm.s(-1).mmHg(-1); all P < 0.05) vascular conductance, without altering oral temperature, cardiac output, heart rate, or stroke volume. These data identify decreases in vascular conductance of skin and of brachial, celiac, superior mesenteric, and renal arteries. Thus it appears that vasoconstriction in both peripheral and visceral arteries contributes importantly to the pressor response produced during skin-surface cooling in humans.

Active cutaneous vasodilation in resting humans during mild heat stress

The role of skin temperature in reflex control of the active cutaneous vasodilator system was examined in six subjects during mild graded heat stress imposed by perfusing water at 34, 36, 38, and 40°C through a tube-lined garment. Skin sympathetic nerve activity (SSNA) was recorded from the peroneal nerve with microneurography. While monitoring esophageal, mean skin, and local skin temperatures, we recorded skin blood flow at bretylium-treated and untreated skin sites by using laser-Doppler velocimetry and local sweat rate by using capacitance hygrometry on the dorsal foot. Cutaneous vascular conductance (CVC) was calculated by dividing skin blood flow by mean arterial pressure. Mild heat stress increased mean skin temperature by 0.2 or 0.3°C every stage, but esophageal and local skin temperature did not change during the first three stages. CVC at the bretylium tosylate-treated site (CVCBT) and sweat expulsion number increased at 38 and 40°C compared with 34°C ( P < 0.05); however, CVC at the untreated site did not change. SSNA increased at 40°C ( P < 0.05, different from 34°C). However, SSNA burst amplitude increased ( P < 0.05), whereas SSNA burst duration decreased ( P < 0.05), at the same time as we observed the increase in CVCBTand sweat expulsion number. These data support the hypothesis that the active vasodilator system is activated by changes in mean skin temperature, even at normal core temperature, and illustrate the intricate competition between active vasodilator and the vasoconstrictor system for control of skin blood flow during mild heat stress.

Control of cutaneous vascular conductance and sweating during recovery from dynamic exercise in humans

The purpose of the study was to examine the effect of 1) passive (assisted pedaling), 2) active (loadless pedaling), and 3) inactive (motionless) recovery modes on mean arterial pressure (MAP), skin blood flow (SkBF), and sweating during recovery after 15 min of dynamic exercise. It was hypothesized that an active recovery mode would be most effective in attenuating the fall in MAP, SkBF, and sweating during exercise recovery. Six male subjects performed 15 min of cycle ergometer exercise at 70% of their predetermined peak oxygen consumption followed by 15 min of 1) active, 2) passive, or 3) inactive recovery. Mean skin temperature (T̄sk), esophageal temperature (Tes), SkBF, sweating, cardiac output (CO), stroke volume (SV), heart rate (HR), total peripheral resistance (TPR), and MAP were recorded at baseline, end exercise, and 2, 5, 8, 12, and 15 min postexercise. Cutaneous vascular conductance (CVC) was calculated as the ratio of laser-Doppler blood flow to MAP. In the active and passive recovery modes, CVC, sweat rate, MAP, CO, and SV remained elevated over inactive values ( P < 0.05). The passive mode was equally as effective as the active mode in maintaining CO, SV, MAP, CVC, and sweat rate above inactive recovery. Sweat rate was different among all modes after 8 min of recovery ( P < 0.05). TPR during active recovery remained significantly lower than during recovery in the passive and inactive modes ( P < 0.05). No differences in either Tesor T̄skwere observed among conditions. Given that MAP was higher during passive and active recovery modes than during inactive recovery suggests differences in CVC may be due to differences in baroreceptor unloading and not factors attributed to central command. However, differences in sweat rate may be influenced by factors such as central command and mechanoreceptor stimulation.

Active recovery attenuates the fall in sweat rate but not cutaneous vascular conductance after supine exercise

The purpose of this study was to identify whether baroreceptor unloading was responsible for less efficient heat loss responses (i.e., skin blood flow and sweat rate) previously reported during inactive compared with active recovery after upright cycle exercise (Carter R III, Wilson TE, Watenpaugh DE, Smith ML, and Crandall CG. J Appl Physiol 93: 1918-1929, 2002). Eight healthy adults performed two 15-min bouts of supine cycle exercise followed by inactive or active (no-load pedaling) supine recovery. Core temperature (Tcore), mean skin temperature (Tsk), heart rate, mean arterial blood pressure (MAP), thoracic impedance, central venous pressure ( n = 4), cutaneous vascular conductance (CVC; laser-Doppler flux/MAP expressed as percentage of maximal vasodilation), and sweat rate were measured throughout exercise and during 5 min of recovery. Exercise bouts were similar in power output, heart rate, Tcore, and Tsk. Baroreceptor loading and thermal status were similar during trials because MAP (90 ± 4, 88 ± 4 mmHg), thoracic impedance (29 ± 1, 28 ± 2 Ω), central venous pressure (5 ± 1, 4 ± 1 mmHg), Tcore(37.5 ± 0.1, 37.5 ± 0.1°C), and Tsk(34.1 ± 0.3, 34.2 ± 0.2°C) were not significantly different at 3 min of recovery between active and inactive recoveries, respectively; all P > 0.05. At 3 min of recovery, chest CVC was not significantly different between active (25 ± 6% of maximum) and inactive (28 ± 6% of maximum; P > 0.05) recovery. In contrast, at this time point, chest sweat rate was higher during active (0.45 ± 0.16 mg·cm-2·min-1) compared with inactive (0.34 ± 0.19 mg·cm-2·min-1; P < 0.05) recovery. After exercise CVC and sweat rate are differentially controlled, with CVC being primarily influenced by baroreceptor loading status while sweat rate is influenced by other factors.