Heat stress and dehydration in adapting for performance: Good, bad, both, or neither?

Heat and dehydration are strongly inter-connected

Heat stress elevates body tissue temperatures, which stimulate sweating and cutaneous vasodilatation to increase heat dissipation if the environment permits. Warmer environments add to heat stress by reducing (i) internal gradients for convective, conductive and mass flow heat transfer from the core to the skin, and (ii) external gradients for heat transfer from the skin via convection and radiation (conduction is usually negligible). Humid environments add to heat stress by reducing the vapor pressure gradient from the skin to the environment, which reduces the rate at which sweat can evaporate. Evaporation is already the dominant means of heat loss at usual training- or competitive-exercise intensities, so hot or humid environments exacerbate the dehydration that normally accompanies exercise. Dehydration by sweating leaves a smaller volume of more concentrated body fluid, including of the blood plasma, i.e., a hyperosmotic hypovolemic hypohydration; more so in trained and acclimated athletes because of their higher rates of work and concomitant heat production, sweating requirement and capacity for sodium reabsorption.^6,7

Not only does heat stress cause hypohydration, but heat and hypohydration each incur wide-ranging physiological and psychophysical strain independently, several examples of which are illustrated in Figure 1. In many respects their acute effects are synergistic – a notable exception being that hyperosmotic hypovolemia can attenuate vasodilation and sweating,^8,9 thereby further increasing heat strain. Thus, each can exacerbate the other. The acute effects of heat strain and hypohydration need to be considered because they could stimulate or impair adaptations. Many effects of heat stress and hypohydration are mediated at least partly via increased tissue temperatures and cardiovascular strain.^10-12 Thus, the physiological effects in the left side of Figure 1 tend to drive those on the right. Cardiac output and muscle perfusion can be compromised by the combination of prolonged or intense endurance exercise in warm conditions, in an upright posture, and hypohydrated, at least in laboratory conditions.^13-15 Heat strain and hypohydration - even at mild levels in a lab environment (i.e., low air velocity) - can each reduce central venous pressure and stroke volume, and increase glycogenolysis^16,17 oxidative stress^18,19 and several neuro-endocrine responses.^20-22 Given that some of these stimuli can elicit adaptation, as explained below, it seems plausible – but speculative - that hypohydration could therefore potentiate some heat-induced adaptations irrespective of whether it exerts independent effects. It must be acknowledged however that individual stressors may provide interference (e.g., potentially with heat and hypoxia on plasma volume) or an excessive net stress²³ and thus act to attenuate adaptive responses. Heat and hypohydration could conceivably interfere hypothalamically (discussed below) but otherwise would seem to have mainly synergistic effects at least systemically.

Some differential effects of heat stress and hypohydration are evident (Table 1). For example, oxidative stress may be more attributable to hypohydration than heat during exercise.¹⁸ Whether this stress would help, hinder or have no effect on adaptation is, to our knowledge, unknown, and might depend on the athlete's age, sex, use of exogenous antioxidants^24,25 and balance between exogenous heat stress and exercise per se.¹⁹ Hypohydration is also more important than heat in driving the fluid regulatory hormones, although both hot-dry and warm-wet heat stress potentiate the hypohydration-induced increases in aldosterone during exercise performed at low intensity.²² In contrast, high core temperature per se reduces cerebral perfusion, due in part to effects of hyperventilation-induced hypocapnia and higher cerebrovascular reactivity to CO₂ in exercise.^26,27 Heat-induced hypocapnia does not seem to be exacerbated by hypohydration, whether at mild levels in exercise (2.5% BM)²⁸ or moderate levels at rest (5%).²⁹ In turn, hypohydration at 2% BM does not measurably exacerbate heat-induced reductions in cerebral perfusion during passive heat stress,³⁰ but 3% BM exacerbates orthostatically-induced reductions in perfusion when normothermic upon standing, independently of blood pressure.^31,32 Given that prolonged endurance exercise itself causes marked reduction in orthostatic tolerance,^33-35 hypohydration might be considered to have a minor role relative to those of exercise and heat in impaired cerebrovascular perfusion during or immediately following exercise, including the risk of syncope.

Strenuous exercise or exercise in the heat can increase the permeability of tight junctions of the gut^36-39 and blood brain barrier.^40,41 At least in the case of the blood brain barrier, hypohydration may play a larger role than heat,⁴⁰ which might account for the lack of an observable effect of exercise in the heat in some studies.^42,43 Physiologically-relevant levels of heating increase the permeability of epithelial cell tight junctions of the gut in rats,⁴⁴ and in gut and kidney cell cultures.^45,46 Heat stress also stimulates the important signaling molecule, mTOR,^47,48 as well as glucose sensitivity at least in aged muscle, via HSP72 induction.⁴⁹ Contrary to this, hyperosmotic hypohydration may oppose both of these effects⁵⁰ via volume regulatory signaling mechanisms and/or oxidative stress. It is therefore apparent that heat and hypohydration each elicit a wide range of physiological effects as a normal part of exercise or additional stress; many of these effects are synergistic although some are not.

The adaptive responses to exercise or heat acclimation/acclimatization programs are likely facilitated by a combination of muscular activity, heat, orthostasis, and dehydration. The highly interconnected nature of these stressors makes it problematic to attribute adaptations primarily to an individual stressor or a defined combination. Rarely are the effects of each stressor delineated and their individual contributions accounted for. The left side of Table 1 highlights the acute contributions of heat, dehydration, and orthostasis to overall physiological and psychological strain under exertional stress, and the subsequent functional outcomes. The right side summarizes the state of knowledge regarding the importance of these stressors to adaptation, namely reducing the strain to each of the systems, and improving functional outcomes. Because of the interconnected acute effects of these stressors and a lack of research delineating their contributions in acclimation, their separate and potentially combined roles in the acclimated or adapted phenotype must be scrutinized; hence the purpose, scope and progression of this review.

Why might heat stress be important for adaptation?

Given the many acute effects of heat stress (e.g., Fig. 1), it is unsurprising that repeated exposure to exogenous heat stress drives adaptations at intracellular, tissue, organ and systemic levels that collectively lessen physiological and perceptual strain² and improve exercise capacity, at least in the heat. Several such adaptations are qualitatively similar to those arising from aerobic training, which is also unsurprising given the large endogenous heat stress of training⁵¹ and the cross tolerance provided by some adaptations (described below). For example, acclimation-induced lowering of sympathetic activation^52,53 and body tissue temperatures could both contribute to glycogen sparing in exercise.^54,55 The exogenous heat stress provides large additional physiological and heat tolerance benefits for untrained and moderately-trained individuals and modest but important gains for well-trained individuals.^56,57 The warmer periphery provides several benefits over endogenous heat stress alone; it increases tissue temperatures throughout the body, and signaling therein,^58,59 increases central cardiovascular and fluid regulatory strain, functionally adapts heat loss effectors in the skin,^60,61 and improves familiarity.

In a wider context, short-term (8 weeks) hot water immersion water has been shown to improve cardiovascular and cerebrovascular health in young healthy individuals,^62,63 and may provide similar benefits for diseased populations.^64-67 Regular bouts of passive heat stress (sauna bathing) have also been associated with increased longevity among the general population in Finland,⁶⁸ and of improved clinical outcomes in cardiovascular-diseased cohorts in Japan⁶⁹ - albeit with orthostatic and perhaps fluid regulatory stressors in both instances. The changes in physiological function with heat acclimation are addressed below.

Thermoregulatory

Heat acclimation widens the core temperature band available for exercise by reducing the resting core temperature at the time of day in which the acclimation bouts are performed.^70,71 This benefit is more important in uncompensable heat stress,⁵² and can be negated mostly by prior activity (e.g., warm up) but also by hypohydration.⁷² The upper core temperature at which exhaustion occurs in uncompensable heat stress seems not to be raised by short-term acclimation but is higher in aerobically-trained individuals.⁷³ Skin blood flow and sweating adapt over a similarly rapid time course, to concomitantly increase heat loss power due to both local and central adaptations.^51,60,74 The effectors activate at a lower core temperature (T_c), have higher sensitivity (relative to T_c), and attain higher maxima (sweating only⁶⁰), even in highly-trained athletes.

Cardiovascular

Adaptations occur in all components of the cardiovascular system. Peripheral heating increases endothelial shear stress by virtue of increased blood flow, and can thereby improve flow-mediated dilation⁷⁵ and vessel calibre.⁷⁶ A similarly biphasic adaptation seems to occur in the myocardium, initially showing increased contractility and stress tolerance, then increased metabolic efficiency (in rats^58,77). Microvascular function (using the cutaneous circulation as a model) is improved following prolonged passive heat acclimation at least partly by way of greater nitric oxide (NO) bioavailability.⁷⁸ Orthostatic tolerance also develops rapidly, in conjunction with a rapid increase in blood volume via expansion of the plasma volume (i.e., hypervolemia). Several factors point to the hypervolemic response to heat acclimation as being a key mediator of improved cardiovascular and thermoregulatory function; e.g., its extent correlates with the acclimation-induced reduction in heart rate and the increases in skin blood flow and sweat rate during exercise in the heat.⁷⁹ The lower thermal and cardiovascular strain is likely to underlie the other benefits at least in part, such as glycogen sparing and lower perceived exertion. Improved exercise tolerance arises from some unknown combination of these adaptations.

A sustained reduction in central venous pressure (CVP) during and/or after exercise has emerged as a key mediator of adaptive hypervolemia, hence the separated inclusion of orthostatic stress in Table 1. Figure 2 is an extension of a model developed by Convertino⁷⁹ to illustrate how exercise may lead to hypervolemia via reduced CVP, especially when supplemented with heat stress. Convertino's model is extended to incorporate subsequent findings and theories. First, experiments from the John B Pierce laboratory confirmed the findings of Convertino⁸⁰ on the importance of reduced CVP, and the roles of albumin synthesis and sodium retention.^81-83 Upright compared with supine posture during and following exercise substantially increases the aldosterone response and modestly suppresses the atrial natriuretic peptide response, leading to an expansion of plasma volume within one day.⁸³ Second, a role for erythropoietin release in response to reductions in central venous pressure (independent of hypoxia) has been proposed.⁸⁴ Third, additional benefits of expanded fluid volumes are incorporated.^85,86 The possible roles of hyperosmolality on ADH, leading to hypervolemia, were also proposed by Convertino et al.⁸⁰ as non-thermal factors.

Endurance athletes already have markedly expanded red cell and plasma volumes,⁸⁷ which limits the stroke volume-related benefit of further expansion for exercise in temperate conditions.^88,89 Nonetheless, athletes show further volume expansion from heat acclimation, along with reduced heart rates in submaximal exercise and increased maximal aerobic power, anaerobic threshold, and time trial performance in hot laboratory conditions (Table 2).^90,91 Individual differences in the hypervolemic response to heat acclimation are pronounced.⁹² The extent of the increase in exercising, but perhaps not resting, volume correlates with ergogenic variability between individuals.⁹³ On average at least, heat acclimation does not appear to improve the defense of plasma volume during exercise in the heat.⁹⁴ As yet it is unclear whether the key putative adaptation is an expansion of PV as measured in the resting state^79,95 or the exercising state only.⁹³

Muscle and epithelia

Heating skeletal muscle either in the absence of exercise or after exercise can stimulate HSP72 and mTOR,⁴⁷ thereby protecting against disuse atrophy and enhancing muscle regrowth in rats⁹⁶ and producing hypertrophy in humans.⁹⁷ Heating of muscle might therefore seem attractive for strength athletes, but it should be noted that (i) heat per se is much less effective than exercise itself in conditioning muscle against mechanical overload,⁹⁸ (ii) when applied before exercise, heat actually preconditions muscle to become less perturbed⁹⁸ and thus less responsive to a short-term overload,⁹⁹ and (iii) HSP72 is stimulated by intense exercise as well. Incorporating heat into a resistance-training program is therefore not straight forward, even from a conditioning perspective. However, heat acclimation is effective for aerobically-demanding exercise in the heat, and is broadly consistent with the phenotypic responses obtained from aerobic training.

Heat shock proteins have many intracellular roles, one of which is to protect against multiple stressors,^45,100 including otherwise-lethal heat stress. The induction of HSP72 in various tissues shows a heat-dose dependency, with thermal intensity being more important than its duration.^59,101 In regard to acclimation for performance, it is also noteworthy that cellular stress during sporting competition is characterized by the intensity of multiple stressors rather than the more prolonged, lower-grade stress that characterizes many occupational circumstances. Heat acclimation includes increased HIF-1α in rat hearts - which also appears to upregulate renal mRNA erythropoietin, among other targets - and along with HSP72, strongly protects the myocardium against ischemia/reperfusion injury.^102,103 Thus, generalized cellular stress protection, including thermotolerance and ischemia/reperfusion tolerance,^102,104 may be as important for athletes as simply being heat acclimated/acclimatized. In contrast, heat acclimation of a systemic nature may suffice for the less intense demands on workers, coaches and other athletes (e.g., bowls or archery), and these can be achieved as much by the volume as the intensity of heat stress.^105,106

The muscle energetics of heat acclimation have been eloquently described by Horowitz and colleagues based on extensive research on soleus and cardiac muscle in animal models (reviewed in ref. 107). Long-term heat acclimation is characterized by metabolic efficiency, particularly in the heart, achieved without sacrificing autonomic function, and appears to be modulated by sustained low plasma thyroxine concentrations (relative to pre-acclimation and the short-term acclimated phenotype). The relevance of this research for untrained and endurance-trained humans has yet to be determined. For instance, the 30 day heat acclimation used in the rodents has yet to be replicated in humans, much less in endurance-trained humans. As such we are unaware of any human characterization of many features of the long-term heat acclimated rat; such as sustained low plasma thyroxine contributing to altered calcium handling mechanics, and increased myocardial work efficiency of the heart.

Horowitz and colleagues^77,108,109 have also identified the response of the heat acclimated phenotype to novel superimposed hypohydration. Sustained and severe hypohydration (10% BM) abolished the beneficial thermoregulatory characteristics of the heat acclimated rodent,^108,109 and disrupted the gene profile developed by long-term heat acclimation.⁷⁷ While such findings provide valuable mechanistic insight, the extent to which such effects apply to self-autonomous, exercise-conditioned humans is yet to be determined, especially given the self- vs. externally-imposed control of hydration status, its severity, and its novelty.

But does supplemental heat improve adaptation for performance per se?

Several studies have examined whether heat acclimation enhances cardiovascular adaptations and performance in temperate conditions (12–22°C), as summarised in Table 2 and in recent reviews.^110-113 The potential mechanisms are summarised above and in Table 1 of Corbett et al..¹¹⁰ Heat acclimation was shown 40 y ago to increase V̇O₂max in temperate conditions, although early studies used relatively untrained participants^114,115 or uncontrolled designs.^114,116 To our knowledge, Morrison et al.⁹² were the first to address the ergogenic effects in highly-trained athletes, i.e., in an already strongly-stress-adapted phenotype, in a controlled-design study. The matched variable during the heat and control training regimes was the athlete's perception of exertion, which has high ecological validity. Those researchers observed a large variability in both resting hypervolemia and cycling time trial performance outcomes, with little relation between them. Scoon et al.⁹⁵ and then Creasy et al.¹¹⁷ also used crossover experimental designs, in which endurance athletes undertook sauna bathing to volitional tolerance following most training bouts for 3 weeks. The intention was to maximize cardiovascular and thermal strain while already warmed and vasodilated by the training bout. Scoon et al.⁹⁵ observed expanded plasma volume (Evans Blue dye), which correlated closely with a modest but meaningful improvement in running performance, whereas Creasy et al.¹¹⁷ found no such benefits on hypervolemia (CO dilution) or performance in highly-trained rowers when tested 2, 5 and 9 d following heat acclimation. The reason for the different outcomes from these studies is unknown, but it is noted that running for ∼15 min and rowing for <7 min place different demands on the cardiovascular system in regard to orthostatic stress and oxygen delivery to exercising muscle. Both cohorts were clearly orthostatically stressed in the heat bouts, but drinking was discouraged in Scoon et al.,⁹⁵ whereas participants in Creasy et al.¹¹⁷ drank more fluid than they lost in the sauna. Heat stress after training was also employed by Zurawlew et al.¹¹⁸ but using hot-water immersion and a between-subjects design. They observed an increase in 5-km time trial performance in hot (33˚C), but not temperate (18˚C) conditions. While hot water immersion may provide a simple and practical means for athletes to acclimatize, hydrostatic pressure effects of water could conceivably impact hematological adaptations, and the most efficacious method is undetermined (see below).

Lorenzo et al.⁶⁰ have addressed the ergogenic issue comprehensively in cohorts of well-trained cyclists, with extensive performance and physiological assessments in both warm and temperate conditions. The aerobic measures were improved 5–10% on average in both environments, in conjunction with a typical magnitude of expansion in plasma volume (∼4.5%, at rest) and a substantial increase in maximal cardiac output in both environments. Similar to other studies, the standard deviation between individuals approximated the mean treatment effect, for both physiological and performance variables. Those researchers used matched work rate between heat acclimation and control training conditions, in contrast to the approach used by Morrison et al.,⁹² also in cyclists. On the other hand, Neal et al.¹¹⁹ administered isothermic short-term heat acclimation to trained cyclists and triathletes and found that temperate performance (peak power output; PPO) tended to increase by ∼2% (P = 0.06). A two-week heat acclimatization of 9 trained cyclists improved hot-weather time-trial performance back to temperate-performance levels, which correlated with the heat-induced hypervolemia,¹²⁰ but performance in the temperate environment was not improved beyond that of the control group.¹²¹ Interestingly, plasma expanded by >10% in both groups. Thus, it remains unclear whether heat acclimation confers cross-tolerance to improve performance in less stressful conditions, and what features of the adapted phenotype convey this (if any) benefit.

The ergogenic effect of heat acclimation via water immersion is similarly unresolved because the data are sparse and conflicting. Bradford et al.¹²² had swimmers undertake 6 hours of mixed-intensity training (subjectively matched, to volitional tolerance) in hot (33˚C) vs. temperate (28˚C) water in a crossover fashion, and demonstrated clearly trivial effects for swimming performance in both water temperatures, and unclear effects for terrestrial performance. The posture, hydrostatic squeeze and lack of heat strain from swimming in such water were all suggested to account for the lack of any apparent physiological adaptations. Conversely, Hue et al.¹²³ had 6 competitive swimmers undertake 14 swims (∼60 km) in the heat (air 30˚C and water 35˚C), or at altitude (air 4˚C and water 27˚C). The 400-m performance times in temperate water were unchanged 10 d afterward but were improved by “10%” in the tropical group at 30 d afterward.

The ergogenic potential of heat for sports involving variable exercise intensities is also unresolved. Buchheit et al.^124,125 and have found that field team sport athletes (Football/soccer and Australian Football League) also seem to gain functionally-important physiological and performance adaptations from heat acclimation. Those studies have strong ecological validity in their conditioning, testing in lab and field settings, and the calibre of athlete tested. Unfortunately the lack of a control group precludes the subtraction of any potential camp effects or early-season training effects in regard to the role of heat as a conditioning stimulus. Chen et al.¹²⁶ had elite racket sport athletes undergo short-term heat acclimation using matched incremental exercise to that of a control group, and found a trivial increase in time-to-exhaustion in the heat but not in temperate conditions. Finally, while the combination of hypoxia and heat may act synergistically to maximise hematologic adaptations and improve performance in both endurance and field sport competition, the scant findings are inconclusive.^115,125

Methods of acclimation: Disparate heat acclimation regimes have been used with athletes and have produced meaningful physiological and performance outcomes (Table 2). In fact, so many variables differ between the relatively few studies on athletes that it is impossible to identify an optimal acclimation regime. Presumably the essential components are: achieving prolonged (60-100 min) warming of the skin during or following exercise, perhaps on consecutive days,¹²⁷ in which core temperature is raised (1–2°C), and remaining upright (i.e., a combination of all component stressors). It is however important to realize that individual differences to acclimation are larger than the differences between protocols (Table 2). Thus, any one optimal or most efficacious heat acclimation regime seems unlikely to exist for use across athletes, with little evidence to indicate one method of sustaining thermal strain throughout the regime is more effective than another.^128,129

Adaptation to heat stress attenuates performance decrements in similarly stressful environments, secondary to reduced physiological and psychological strain. Important unresolved issues include the magnitude of these effects, particularly in well-trained athletes,^112,113 and its usefulness in cross-tolerance to different stressors and environments, such as cold and altitude.^130-134 Only one-third of studies (6/19) on this topic have formally compared control against HA effects (Table 2), and measurement error is seldom taken into consideration in the design, execution and interpretation of studies (Table 2,¹³⁵); both of which make it even more difficult to establish true effects. Some behavioral responses should be reported throughout the conditioning regime (e.g., dietary behavior before, during and after individual sessions, particularly sodium, carbohydrate and protein intake) to appropriately assess the role of other stressors (e.g., dehydration), or their contribution to individual differences. The right side of Table 1 illustrates the current knowledge (or lack of) regarding how individual stressors contribute to adaptation. While heat appears to be the primary stimulus in many cases (e.g., in reducing thermal, cardiovascular, and psychological strain), it is likely influenced by synergistic or antagonistic effects of concurrent stressors.

Why consider dehydration?

It would be valuable to know whether the hypohydration incurred during bouts of training or acclimation is harmful, helpful or of no consequence for adaptation,¹³⁶ at the mild-to-moderate levels (typically <3%) that develop volitionally and become self-limiting during such training and acclimatization. Astoundingly, the roles of mild dehydration in adaptation appear to be unknown despite it being a typical stressor for athletes during fitness training, especially in summer.^57,137 Consideration is not given to the resting situation, because healthy individuals with access to water become thirsty and drink before body fluid deficits are incurred.^138,139 Similarly, severe hypohydration (>5%) is not considered here for reasons described below and because it is (i) self-limiting, (ii) unpleasant and thus distracting, (iii) therefore unnatural or at least unusual in a non-competitive setting, (iv) possibly unhealthy, and (v) may not stimulate further net compensatory responses especially relative to catabolic effects.

Adolph and colleagues undertook extensive research on dehydration and acclimation more than 75 y ago,¹⁴⁰ but they focused more on whether humans could acclimate to dehydration, i.e., in reducing the volume of fluid intake required to offset BM losses during work in the heat. Water requirements were not lessened to any functional extent by repeated dehydration, which is logical and important for physical activity in a hot environment because heat balance is governed by the requirement for evaporation. Other studies have looked at various physiological and genomic impacts of hypohydration before and after heat acclimation.^{22,72,73,77,108,109,141,142} Notwithstanding the important contributions of such studies, again, they were not designed to address the issue of a possible role of ecologically-valid dehydration in adaptation. The extent of hypohydration imposed in such studies (5–6% in non-athlete humans and 10% in rats) is also larger than would reasonably occur in training, acclimation or acclimatization, and the participants were not permitted any behavioral regulation other than avoidance of further dehydration by volitional cessation of exercise.

Before considering the potential relevance of volitional hypohydration in adaptation, its acute effects on physiology and aerobic performance must be critiqued briefly for three reasons. First, adaptation is the cumulative outcome from additive bouts of stress, so knowledge of the acute effects in the self-regulated setting of training or acclimation is important (e.g., see Figures 1 and 2, and left side of Table 1). Second, if quality conditioning is that which maximizes its work volume (absolute intensity and/or duration), then volitional dehydration might be ill-advised because even these magnitudes are advocated as impairing work tolerance and promoting fatigue.^143-146 Third, if ecologically-valid extents of hypohydration have little or no effect on thermal or cardiovascular strain when training outdoors in temperate or warm conditions (see below), then hypohydration would presumably also fail to provide a stimulus for adaptation in training or in heat acclimatizing exercise where airflow is high (e.g., cycling, running, rowing). Hypohydration would still be relevant for heat acclimation though.

Does dehydration affect physiological strain and tolerance? Of relevance to the heat acclimation context, hypohydration has been advised against in a general sense because it can increase T_c at rest and during exercise, and reduce the maximal T_c tolerated^57,147 thus potentially truncating the volume of exposure. But, the reduced tolerance to hyperthermia was demonstrated in the setting of a hot lab environment with an imposed prior hypohydration of 8% BM. In free-living field settings, fit individuals will voluntarily exercise intensely at high T_c even at these rarely-encountered magnitudes of hypohydration.^148,149 In fact, even during the artificial setting of externally-imposed hypohydration at a more typical magnitude (2–2.5% BM) while encapsulated in protective clothing in a research laboratory, thermal tolerance remains high in fit individuals.^73,86 While some aerobic training takes place indoors (e.g., in northern hemisphere winters), the thermal strain caused by being indoors with limited airflow would be at least as important as any impact of volitional hypohydration.^150,151 On the other hand, if being indoors is for the purpose of heat acclimating, then the purpose is not to maximize work volume, but to promote cardiovascular and thermal strain. So, effects such as impaired mood, increased glycogenolysis, reduced CVP, increased heart rate, and increased tissue temperatures become relevant - some beneficially and others detrimentally. Further, if hypohydration confers greater strain than when euhydrated, then its' presence may provide a time efficient alternative for conditioning (i.e., same strain in less time). Several recent studies using ecologically-valid designs, along with quantitative reviews (meta-analysis) have shown that hypohydration of 2–3% has little or no measurable effects on physiological strain, and no effect on psychophysical strain or performance.^152-156 In contrast, if participants are deprived of fluid before or during exercise against their behavioral drive, are not given opportunity to familiarize to that stressor¹⁵⁷ or are tested in lower airflow environments, then physiological, psychophysical strain and tolerance of exercise are all affected substantially (reviewed in ref. 158). In summary, volitional dehydration might be relevant in heat acclimation, but less so in outdoor training or acclimatization because its effects are smaller than is typically conveyed in the literature.

Why might dehydration enhance adaptation? One reason is by exacerbating strain (Fig. 1) and enhancing compensatory adaptations (e.g., ECFV; Fig. 2). Another possible candidate for adaptation to dehydration transients would be in fluid regulation, since most physiological systems adapt and at multiple levels in response to repeated stress. Fluid regulatory adaptations could involve afferent, central or efferent structural or functional components.^159,160 Renal concentrating ability is markedly (40–50%) increased during short-term (3-d) sustained hypohydration, and is inhibited during short-term over-drinking in humans.^161,162 While these renal adaptations would theoretically improve fluid balance during the often-obligatory dehydrative stress of athletic competition, they seem unrealistic because they were demonstrated using a sustained, high magnitude of hypohydration, which might interfere with muscle metabolic control, protein synthesis and hypothalamic adaptations to heat acclimation,^50,77,163 not to mention cognitive and psychosocial effects. Secondly, any reduction in renal excretion of water during sporting competition would have only small effects on fluid balance in view of the dominance of sweating in dehydration during competitive-intensity exercise.

Horowitz and colleagues^164,165 highlighted the ability of animals (desert spiny mouse) to acclimate to prolonged dehydration, namely to better maintain plasma volume in the face of acute dehydration. Such defense was attributed to decreased permeability of the vascular capillary bed, and thus less ultrafiltration. Whether such effects would occur in humans is unknown, and may be even less likely for the self-limited magnitudes and durations of hypohydration involved with training and heat acclimatization.

Plasma volume expansion is reported as being observed most commonly after 5 d of exercising upright in the heat, while “properly hydrated.”¹⁶⁶ Such appraisals acknowledge the value of exercise, heat and orthostasis¹⁶⁶ while precluding that of dehydration. As mentioned above, if dehydration independently, and in combination with heat and exercise, reduces CVP, even in a temperate environment,¹⁶⁷ and increases fluid regulatory, thermal, and cardiovascular strain, then it seems reasonable to suggest that its addition could augment the stimulus for adaptation.

Relative to untrained individuals, athletes have altered neuroendocrine control (ADH vs. plasma osmolality) and are less sensitive to thermal and cardiovascular effects of hypohydration when exercising with high airflow in both temperate^160,168 and warm environments,¹⁵⁵ but not when exercising with compromised airflow in the heat.^72,169 Athletes perceive thirst as sensitively as untrained individuals, and voluntarily rehydrate to a similar extent – at least during cycling in lab trials¹⁶⁸ - so they do not appear to be more predisposed to an insidious progression of dehydration during training or acclimation (Fig. 3). Athletes dehydrate more quickly in outdoor training,^168,170 due to higher endogenous heat production, but less markedly so during indoor training and heat acclimation bouts because of the imposed (exogenous) heat exposure (and thus similar sweat rates attained). Even indoors, their higher T_c tolerance, work rates and sweat rates will incur more hypohydration. However, drinking behavior increases as heat acclimation progresses, apparently in relation to the renin-angiotensin-aldosterone (RAAS) pathway more than ADH.¹⁷⁰ Hyperosmolality in exercise may contribute to a rise in anti-diuretic hormone (ADH) and to fluid expansion with repeated exposures but the increase in osmolality is at least partly attributable to exercise per se.⁸⁰ The modest exercise intensities used in prolonged heat acclimation bouts involve little increase in osmolality or ADH.¹⁷⁰ In conclusion, some fluid regulatory differences exist for athletes relative to untrained individuals, but not in their behavioral drive to limit dehydration. Although renal concentrating ability can be improved by hypohydration, this adaptation is not warranted. Adaptations to dehydration in training or acclimation should be studied using athletes during heat acclimation, for reasons discussed above.

Is there evidence for enhanced or attenuated adaptations from volitional dehydration?

The available studies do not reveal whether voluntary dehydration enhances or impairs adaptation during heat acclimation because the requisite information is not always available (Table 1), and the studies are too disparate in other respects (Table 2). If an effect exists it is presumably not dramatic because heat acclimation regimes using almost no replacement or full replacement have achieved similar hypervolemic and performance outcomes. Fleming and James¹⁵⁷ have recently demonstrated that familiarisation can occur within 4 exposures to hypohydration and exercising heat stress, such that its ergolytic effect became non-significant, as mentioned above. It is unclear how much of this recovery in performance was physiological, but participants still showed higher heart rates, RPE and T_c in exercise with hypohydration in that low airflow environment, so the habituation was presumably psychological rather than physiological.

We studied the separated role of dehydration in heat acclimation, using 9 fit male volunteers (V̇O_{2 peak} 60 ±7 mL/kg/min).¹⁷¹ Participants were acclimated on two occasions, once with no dehydration and once with minimal rehydration (0.1 L; achieving ∼2% hypohydration), in crossover fashion, for 7.5 h (90-min/d for 5 d) using controlled hyperthermia (T_c = 38.5°C). Importantly, core (rectal) temperature was clamped to prevent the additional thermal strain that would otherwise ensue in the calm lab conditions, so any difference in outcome was delineated from any such thermal effect. Participants were given no verbal or written expectation as to which direction the effects might be, if any. The acclimation-induced expansion in plasma volume tended to be larger across acclimation with dehydration than euhydration (by 4.5%; 95%CI: −1 to 10%; P = 0.06; Fig. 4A), and be correlated with a rise in the aldosterone response across acclimation. Similarly, the change in body mass from pre to post acclimation was significantly (albeit trivially) larger with dehydration than euhydration (by 0.8 kg; 95%CI: 0.1 to 1.5; P = 0.03; unpublished results; see Fig. 4B), potentially reflecting the higher total body water content subsequent to greater fluid retention. The acclimation-induced reduction in end-exercise heart rate during a standardized heat stress test was also larger across the dehydration acclimation regime (by 11 b/min: −1 to 22; P = 0.05; Fig. 4C). But the differences between hydration regimes were unclear for most endocrine, cardiovascular, psychophysical and ergogenic outcomes (e.g., Fig. 4D). Thus, dehydration to ∼2% BM did not impair heat adaptation in fit males, and may have enhanced some aspects of short-term heat acclimation.^119,171 However, in view of the distinct lack of research on the role of hydration when conditioning the cardiovascular, thermoregulatory and fluid regulatory systems, and the inconsistent findings from the two studies, it is clearly not possible to suggest whether dehydration during stress is beneficial, counterproductive, or neither.

Caveats with dehydration: Research from the lab of Dr Hiroshi Nose has demonstrated the importance of replenishing protein and carbohydrate soon after a bout of exercise, especially in older individuals.^172-174 Partial rehydration with carbohydrate and amino acid-containing fluids (1.8 g amino acids/kg BM, in 3.2 mL water/kg) increased plasma albumin content and plasma volume restoration following a single bout of interval exercise, in old and young men.¹⁷³ Post-exercise supplementation, when applied across 8 wk of aerobic training in older men, enabled expansion of plasma albumin content and plasma volume.¹⁷⁴ This hypervolemia was somewhat defended during exercise, in conjunction with less cardiovascular strain and enhanced thermoeffector responses, compared to responses obtained from an equivalent fluid volume of placebo replenishment following each training bout. Similar results were obtained with young men training for 5 d in warm conditions.¹⁷² Given those findings and existing knowledge that older individuals are slow to perceive hypohydration and subsequently rehydrate,¹⁷⁵ early nourishment following exercise seems warranted, especially in older individuals and regardless of the magnitude of hypohydration caused by exercise. Since mild hypohydration can also impair mood^176,177 and cognition¹⁷⁸ at rest, full rehydration early following exercise may be merited. This actually remains unresolved because of the multiple problems of validity with such studies (see ref. 158), and because of opposing data on effects of early rehydration (see below).

The adaptive stimulus of a given exercise bout, or heat exposure, may be also be determined by the time course and extent of rehydration. Plasma volume expansion is consistently evident in response to acute and repeated exercise bouts, particularly utilizing 8 × 4 min at ∼85% peak V̇O₂ with 5-min active recovery between repetitions (John B. Pierce Laboratory). In these and similar studies characterizing post-exercise plasma volume expansion, fluid is not made available during the exercise, and rehydration begins after one hour^179,180 or at least 2 hours^{82,83,181,182} following the exercise. Yet fluid retention is consistently stimulated, producing a rebound hypervolemic response. It appears that fluid regulatory hormonal responses (particularly plasma aldosterone) is apparent with prolonged maintenance of hypohydration, up to at least 6 hours post exercise,¹⁸³ and may be attenuated on provision of food and fluid.¹⁸³ Similarly, Costill et al.¹⁸⁴ found an impaired accrual of plasma volume if drinking carbohydrate and electrolyte beverage relative to water only, early after each of 5 daily bouts of dehydrating to 3%. Further, the reduction in CVP following prolonged dehydrating (∼3% BM) exercise is evident even after eating and drinking to satiety 1 hour after exercise cessation,¹⁶⁷ and thus the timing of rehydration (and nutrient replenishment) may provide a window for manipulating an adaptive stimulus. As some functional hypohydration is likely following exercise and particularly exogenous heat stress (see Fig. 3 in ref. 158), individual differences in adaptive responses (particularly hematological) may therefore be influenced by rehydration regimes following the conditioning stimulus. Further research is warranted to determine the role of hydration before, during, and following conditioning sessions on adaptive responses.

Individual differences

The effects of dehydration do not impact all individuals equally. The degree to which dehydration will stimulate behavioral (thirst), and physiological responses (e.g., ADH secretion) differs due to genetic variation in the threshold and sensitivity of the osmoregulatory systems.¹⁸⁵ This variability makes some individuals reluctant to drink during prolonged exercise,¹⁸⁶ and can be further influenced by their fitness (neuroendocrine response;¹⁶⁸) or acclimatization to heat (quantity of fluid consumed;¹⁷⁰). It therefore seems likely that if permissive dehydration were beneficial during heat acclimation, or ergogenic in its own right, the degree of dehydration that is beneficial would likely differ between individuals. As indices of hydration used for prescribing drinking are confounded by the same variability,¹⁸⁷ it would appear that self-regulated rehydration is as appropriate as a prescribed drinking regime for most individuals.

Sex effects on chronic adaptations to heat stress are conspicuously under researched,¹⁸⁸ whereas those on chronic adaptations to hydration stress are still relatively unknown. It appears that females may require more heat acclimation to achieve the same attenuation in cardiovascular and thermoregulatory strain as males,¹⁸⁹ and any differences in adaptive responses may be partly explained by body composition.¹⁹⁰ The female heart also appears to be less sensitive to heat acclimation-induced HSP induction, possibly due to an inhibitory role of estrogen on HSP transcription and expression (in rats;¹⁹¹). Sex differences therefore seem likely to contribute to variability in adaptive responses to heat, but this requires further research.

Several other issues remain unresolved or unexamined, including the separate and interactive roles of typically-encountered magnitudes of heat and dehydration in regard to (i) intracellular responses in vivo, especially on oxidative stress, energy metabolism and cellular tolerance^18,19,50; (ii) adaptation, especially in young versus older adults, and in regard to anti-inflammatory interventions^192,193; (iii) red cell volume, (iv) the large individual differences in cardiovascular and functional impacts of heat acclimation, and (iv) short vs. long term adaptations, especially for endurance athletes. Functional effects should be assessed under psychologically and physically valid conditions; if representing training, these should ideally include blinding and strong airflow, respectively, whereas if representing heat acclimation, indoor environments or encapsulation, then airflow is unwarranted.

Conclusions and perspectives

Voluntary dehydration is an inherent part of exercise, with athletes typically drinking only half of their fluid loss (as approximated from mass loss, with its limitations). It remains unclear whether ad libitum drinking optimizes performance in competition, partly because laboratory-based research has limited validity in addressing that issue. Nonetheless, a far greater number of exercise bouts are performed in training, the major purpose of which is to adapt multiple systems to improve fitness. It is therefore remarkable that almost no research has been undertaken to determine whether dehydration enhances, impairs or does not substantively affect these adaptations. Dehydration is increased by exogenous heat stress, such as heat acclimation or acclimatization. Dehydration exacerbates the magnitude of strain in several physiological systems, and can increase thermal strain by attenuating the heat loss effectors. While heat is almost certainly the stress of major benefit in driving adaptations, more research is needed to delineate the roles of heat and dehydration. The effectiveness of heat acclimation for enhancing adaptations and performance remains unclear for team sport and for endurance athletes despite a surge in studies on this topic. Whether team sport or endurance athletes should drink ad libitum, or more avidly during aerobic training and heat acclimation is also not known, but rehydrating in conjunction with amino acids, carbohydrates, and sodium after training seems valuable, especially in older athletes. It is clear that prolonged orthostasis (during or following exercise or heat exposure) facilitates a more beneficial hormonal profile (for enhanced fluid regulation), however the time course of rehydration (in recovery), and its possible potentiating role¹⁹⁴ in prolonging cardiovascular and fluid regulatory strain remain unclear. As is stands, the effectiveness of heat acclimation as a strategy to enhance adaptation for performance in a cool environment is unclear, as is the role of volitional dehydration within such heat acclimation or within normal training for endurance performance in hot, warm or cool environments.