Short-term sleep deprivation and human thermoregulatory function during thermal challenges

Relatively short periods of inadequate sleep provoke physiological and psychological perturbations, typically leading to functional impairments and degradation in performance. It is commonly accepted that sleep deprivation also disturbs thermal homeostasis, plausibly enhancing susceptibility to cold- and heat-related illnesses. Herein, we summarize the current state of human-based evidence on the impact of short-term (i.e., ≤ 4 nights) sleep deprivation on autonomic and behavioural thermoeffectors during acute exposure to low and high ambient temperatures. The purpose of this brief narrative review is to highlight knowledge gaps in the area and stimulate future research to investigate whether sleep deprivation constitutes a predisposing factor for the development of thermal injuries.

Considering the essential role of sleep in the regulation of the milieu intérieur, its potential restorative function (see Frank & Benington, 2006), and the close interrelationship of sleep with the thermoregulatory system (see Krauchi & Deboer, 2010), it is reasonable to assume that sleep deficit might also disturb thermal homeostasis, enhancing susceptibility to cold-[i.e., (non-)freezing cold injury and hypothermia] and heat-related (i.e., heat exhaustion and heatstroke) illnesses. Such a notion seems to be supported by anecdotal fieldbased observations (Pugh, 1966;Rav-Acha et al., 2004;Young et al., 1998); and thus, sleep deprivation has conventionally been suggested to be a predisposing factor to the development of thermal injury (Armstrong et al., 2007;Casa et al., 2012;Castellani et al., 2006;Rintamaki, 2000). Nonetheless, the findings derived from original experimental investigations, wherein the confounding influence of other non-thermal stressors (e.g., physical exertion, caloric deficit) on thermoregulation has been eliminated, appear to be ambiguous.
The purpose of this brief narrative review, therefore, is to summarize the available evidence on the effects of sleep loss on autonomic and behavioural thermoeffectors during acute exposure to low and high ambient temperatures. Our literature search was limited to healthy human populations and to studies evaluating resting and exercise thermoregulation after relatively short periods (i.e., ≤4 nights) of experimentally induced sleep deprivation.

Thermoregulation in response to cold
During cold stress, thermal homeostasis is preserved via a sympathetically mediated increase in peripheral vasomotor tone, attenuating heat loss to the surroundings, and the activation of muscle shivering and, to a small extent, of non-shivering thermogenesis.
Yet, during sustained cold exposure, protracted peripheral vasoconstriction can lead to local cold injury, especially in the acral skin regions (i.e., finger, toe, nose). Also, when the rate of endogenous heat production fails to offset the rate of body heat loss, hypothermia can eventuate.
Military-based studies have demonstrated that periods of sustained operations, during which TSD or PSD is commonly encountered, may compromise thermoregulation to cold, enhancing the risk of hypothermia. Thus, a substantial long-term (61 days) sustained operation blunted shivering thermogenesis and cutaneous vasoconstriction, and accelerated body core cooling during a 4 h cold (10 • C) air exposure (Young et al., 1998). In line with this, Castellani et al. (2003) found that

New Findings
• What is the topic of this review?  (Macdonald et al., 1984), as well as fatigue engendered by prolonged and exhaustive exercise (Castellani et al., 1999(Castellani et al., , 2001, appear to impair the reflex vasoconstrictor response to cold, whereas hypoglycaemia can inhibit shivering thermogenesis (Gale et al., 1981;Passias et al., 1996), thereby potentiating the reduction in body T c . Lastly, morphological alterations are also caused by such multistressor regimens; reductions in body lean and fat mass often ensue (Young et al., 1998), degrading the insulative capacity (Cannon & Keatinge, 1960).
The distinct effects of short-term TSD on thermoeffector responses to whole-body cold exposure have been examined in eight independent studies, of which only two have indicated that the cold-defence effectors are modified by TSD (Table 1). Firstly, Landis et al. (1998) noted that, in sleep-deprived women, the reduction in body T c was slightly accelerated during brief intervals of mild skin cooling [skin temperature (T sk ) was deliberately fluctuated and clamped from 38 to 32 • C]. However, the mechanisms underlying this response are unclear; the cold-induced changes in non-glabrous (forearm) skin blood flow were unaltered by TSD, whereas the endogenous production of heat was not monitored. Secondly, Savourey and Bittel (1994) The studies are reported in chronological order. Note that the responses describe differences from the respective control conditions (i.e., normal nocturnal sleep). Abbreviations: AP, arterial pressure; CIVD, coldinduced vasodilatation; HR, heart rate; N, total number of participants;Ṁ, metabolic heat production; n.a., information not available; SkBF, skin blood flow; T c , body core temperature; TSD, total sleep deprivation; T sk , skin temperature; ♂, male participants; ♀, female participants; response to cold and hastened the threshold onset for shivering, which apparently compensated for the increased rate of heat loss, hence obviating a potentiation of hypothermia after TSD.
Contrary to the aforementioned two studies, the others have failed to detect any adverse effects of TSD on thermoregulation to cold, during both resting conditions (Caine-Bish et al., 2004;Costa et al., 2010;Fiorica et al., 1968;Oliver et al., 2015) and moderateintensity exercise (Kolka et al., 1984). Of interest is the work of Oliver et al. (2015) and Costa et al. (2010), who, by using a within-subject design, found that 53 h of TSD combined with or without negative energy balance neither perturbed thermoeffector capacities (Oliver et al., 2015) nor modulated the immune reactions (the suppression of circulating lymphocytes, neutrophil degranulation and saliva secretory immunoglobulin A) evoked by acute cold (0 • C air) stress (Costa et al., 2010). Lastly, the rewarming response after a period of whole-body cooling does not appear to be disturbed by TSD (Esmat et al., 2012;Oliver et al., 2015).
Sleep deprivation has also been regarded as a factor predisposing to the development of local cold injury (Rintamaki, 2000). Indeed, 29 h of TSD augmented finger vasoconstriction instigated by a 30 min direct localized cooling (5 • C water) and delayed the spontaneous rewarming of the extremity after this cooling (Sauvet et al., 2012).
The response was attributable to a TSD-dependent microvascular endothelial dysfunction (for review, see Cherubini et al., 2021), characterized mainly by increased concentrations of endothelin 1 (a potent vasoconstrictor) and perhaps by reduced concentrations of nitric oxide (a potent vasodilator) (Sauvet et al., 2010(Sauvet et al., , 2012(Sauvet et al., , 2017. The lack of sleep, however, did not modify the incidence or the magnitude (i.e., onset and amplitude) of finger cold-induced vasodilatation (aka CIVD), which typically intervenes during cooling of the hands and/or feet (Keramidas et al., 2019) and, presumably, serves a cryoprotective function against cold injury (Wilson & Goldman, 1970).
Collectively, the prevailing evidence suggest that, during wholebody mild cold stress, short-term TSD does not exert any prominent influence on the function of autonomic thermoeffectors; therefore, further work is required to establish whether inadequate sleep per se potentiates the risk of hypothermia. Sleep deprivation, however, might represent a contributing factor for the development of freezing and non-freezing cold injury, given that, on the basis of a single study (Sauvet et al., 2012), the acral skin constrictor responsiveness to localized cooling is aggravated by TSD.

Thermoregulation in response to heat
During heat stress, thermal homeostasis is maintained through peripheral vasodilatation and sweating, which respectively facilitate the dry and evaporative heat transfer from the body to the environment.
However, when the capacity to dissipate heat is limited (e.g., in hot and humid conditions or while wearing personal protective clothing and equipment), especially during exercise, the resultant sustained increases in body heat storage augment the T c elevation, which in turn degrades performance and can eventually lead to the development of heat-related illnesses, such as heat exhaustion and heatstroke.
It is generally accepted that sustained periods of wakefulness disturb exercise thermoregulation and thus increase the risk of exertional heat illness (Armstrong et al., 2007;Casa et al., 2012).
Seminal work by Sawka et al. (1984) showed that, in aerobically fit men, TSD (33 h) suppresses dry and evaporative heat loss during submaximal whole-body exercise performed in thermally compensable conditions (28 • C, 30% relative humidity). This early observation was corroborated by two later studies that evaluated the impact of TSD, with a duration similar to that in the study by Sawka et al. (1984), on exercise thermoregulation during exposure to hot (35 • C) ambient conditions (i.e., in thermally uncompensable state).
Namely, Kolka and Stephenson (1988) demonstrated that the regional (forearm) dry heat exchange was impaired; furthermore, Dewasmes et al. (1993), who clamped the local (chest and thigh) T sk at 35.5 • C, found that the secretion of sweat in these regions was reduced. Ergo these studies demonstrated that TSD modulates peripheral heat loss during exercise provocations. Nevertheless, the origin of this thermoregulatory response remains unsettled. It has been attributed either to a peripheral desensitization, judging by the reduction in the gain of heat-loss effectors (Kolka & Stephenson, 1988;Sawka et al., 1984) and/or to a centrally mediated delay in the activation threshold for heat dissipation (Dewasmes et al., 1993;Kolka & Stephenson, 1988).
However, and regardless of the exact mechanism, it is noteworthy that, apart from work by Sawka et al. (1984), wherein a tendency for a slightly greater increase in body T c (∼0.2 • C) was noted after TSD, the blunted responsiveness of the monitored regional (nonglabrous) heat-dissipation pathways did not lead to significant heat gains, and therefore, the exercise-induced elevations of internal body temperature were not augmented by TSD (Dewasmes et al., 1993;Kolka & Stephenson, 1988). Considering the large inter-regional variation in cutaneous vasomotion (Caldwell et al., 2014(Caldwell et al., , 2016 and sweat production (Cramer et al., 2012;Machado-Moreira & Taylor, 2017), it is not known whether the glabrous skin areas (e.g., forehead) responded in a different manner from the non-glabrous sites after TSD.
The results from the aforementioned exercise-based investigations appear to concur, in part, with those of studies that examined the influence of TSD on thermoregulation during passive heat stress (Table 2). For instance, during a brief period of mild (38 • C) whole-body skin heating, the magnitude of forearm vasodilatation was attenuated in sleep-deprived women, but the rates of change in sweating and body T c remained unaltered (Landis et al., 1998). Moreover, Fujita et al. (2003) noted that, during a 60 min whole-body exposure to 30 • C air and while the lower legs were immersed in 42 • C water, the total sweat loss (indicated by the changes in body weight) was impaired by TSD; however, the increases in regional dry heat loss (reflected by an increase in back and forearm skin blood flow) and body T c were potentiated and diminished, respectively. In addition, after 36 h of TSD, the body T c elevation during iterative exposures to severe humidity) was dampened, despite a reduction in the whole-body sweat loss (Cernych et al., 2021).
The negative influence of TSD on autonomic thermoeffectors during exercise is not a ubiquitous finding, however (Table 2). Relf et al. (2018) used a closed-loop exercise protocol, during which nine young women ran, before and after a sleepless night, for 30 min at fixed (clamped) metabolic heat-production rates (10 W kg −1 ) at 39 • C (41% relative humidity). Notably, neither sweating nor body T c responses were influenced by TSD, but the prevalence of the self-reported symptoms associated with heat exhaustion, such as nausea, lightheadedness and confusion, was aggravated (see next section). Likewise, Muginshtein-Simkovitch et al. (2015) found in well-trained men that, although the perceived thermal discomfort was compounded, the exercise-induced thermal strain at 40 • C (40% relative humidity) was not amplified by 24 h of TSD. These authors, however, argued that the magnitude of the TSD response might be dependent on the individual chronotype; that is, people with an evening chronotype are more vulnerable than others.
Nevertheless, this assumption emanated from a very small sample size (three evening chronotypes vs. eight intermediate chronotypes) and needs to be justified by further experiments. Lastly, there is a sole indication (Hom et al., 2012) that one night of TSD did not disturb the thermo-adaptive modifications evoked by repeated heat exposures; adjustments that were described by enhanced sweating rates and lower increases in T c during exercise.
With regard to the impact of PSD, relatively short periods of sleep restriction (1-4 nights; 4-5 h sleep per night) ostensibly do not impose a potent risk for thermoregulatory dysfunction during heat stress (Cernych et al., 2021;Muginshtein-Simkovitch et al., 2015;Tokizawa et al., 2015). Of note, however, is the work by Tokizawa et al. (2015) showing that, although a night of PSD (4 h of sleep) did not perturb temperature regulation during morning exercise, it exacerbated the exercise-induced elevation in body T c during successive exercise bouts performed during the same afternoon (ambient conditions: 35 • C, 40% relative humidity). Whether this delayed response was attributable to the extended period of wakefulness and/or to the accumulated amounts of fatigue provoked by the preceded morning exercise remains unclear. Arguably, it could also be associated with a PSD-induced circadian dysregulation emerging in the afternoon trials; however, it should be noted that, although T c and T sk exhibit a distinct cyclic variation over a 24 h period (Aldemir et al., 2000;Aoki et al., 2001;Kräuchi & Wirz-Justice, 1994;Refinetti & Menaker, 1992), the sweating response to exercise in the heat does not appear to describe a circadian rhythmicity, at least in waking periods (Ravanelli & Jay, 2020). Nevertheless, it is noteworthy that the enhanced degree of thermal strain noted in the afternoon trials was not counteracted or at least mitigated by a 30 min post-lunch nap that occurred between the morning and afternoon sessions.
Altogether, TSD might modify the efficiency of heat-dissipation effectors during compensable and uncompensable heat stress. There is no compelling evidence, however, that TSD potentiates hyperthermia during resting conditions or exercise. Also, ≤4 days of PSD do not seem to hinder the ability to thermoregulate in the heat.

SLEEP DEPRIVATION AND BEHAVIOURAL THERMOEFFECTORS
Thermoregulatory behaviour describes the deliberate actions, either simple or more complex, that aim to create a thermally comfortable (micro)environment. Examples of thermobehavioural means include, inter alia, changes in body position or clothes and adjustments of exercise intensity. These behavioural counteractions constitute the most powerful response to defend body T c from environmental thermal challenges (Hardy, 1971;Schlader & Vargas, 2019), and their utilization seems to prevent or at least postpone the recruitment of the energetically costly autonomic thermoeffectors (i.e., sweating and shivering) (Schlader et al., 2018). To our knowledge, only one study in humans has attempted to evaluate directly (i.e., in experimental conditions of behavioural freedom) whether the desire for and the magnitude of thermobehaviour are modulated by sustained periods of wakefulness. Namely, 50 h of TSD did not modify the selfpaced exercise intensity and duration nor the thermal and perceptual (affective valence) strain while walking in 0 • C air (Kolka et al., 1984).
The conscious behavioural interventions are motivated by adequate changes in thermal perception invoked by T sk and/or T c displacements (Gagge et al., 1967;Satinoff, 1996). Therefore, valuable insights into the impact of sleep deprivation on behavioural thermoregulation may be gained by the assessment of the self-reported discriminative (i.e., thermal sensation) and, especially, hedonic [i.e., thermal (dis)comfort] perceptions. Considering the recurrently described influence of sleep deficit on cognition (Killgore, 2010), it might be assumed that the thermoperceptual responsiveness to thermal stimuli would also be affected; however, the results on whole-body thermoperception after a period of sleep loss appear to be inconsistent (Tables 1 and 2). Thus, 53 h of TSD exacerbated thermal discomfort during the initial 1 h of a 4 h exposure to 0 • C air; a response that was independent of any TSD-driven variation in the whole-body thermal state of subjects (Oliver et al., 2015). Such a thermoperceptual sensitization, however, is not supported by other cold-relevant studies (Costa et al., 2010;Esmat et al., 2012;Landis et al., 1998). Conflicting findings have also been obtained during whole-body heat stress. For instance, after a sleepless night, Muginshtein-Simkovitch et al. (2015) found that the subjective ratings for thermal discomfort and of effort perception were enhanced during moderate exercise performed at 40 • C, whereas Relf et al. (2018) did not observe any modifications of the perceived thermal pleasantness while exercising at 39 • C (in both studies, the relative humidity was ∼40%). Nevertheless, in the latter study, the prevalence of self-reported symptoms associated with heat exhaustion (e.g., nausea, lightheadedness, confusion) was increased after TSD.
Furthermore, after a night of PSD, the sensations of fatigue and hotness, but not of thermal discomfort, were aggravated during submaximal exercise in thermally uncompensable conditions (35 • C, 40% relative humidity; Tokizawa et al., 2015). Of interest was that a 30 min nap interval, although failing to attenuate the thermal and cardiovascular strain encountered during exercise after TSD, did alleviate the PSD-evoked sensations (i.e., subjects felt less hot; Tokizawa et al., 2015). This nap-related thermoperceptual desensitization might, in fact, compromise behavioural thermoregulatory adjustments, presumably imposing a greater risk of heat exhaustion (cf. Moore et al., 2015). Lastly, there are indications that sleep deficit produces a hyperalgesic response to regional, noxious thermal stimuli; that is, the heat and cold pain thresholds might be accelerated after TSD (Kundermann et al., 2004). A 29 h period of sleep deprivation tended to augment the cold-induced pain during a 30 min hand immersion in 5 • C water; a response that might have been determined by the enhanced degrees of finger vasoconstriction after TSD (Sauvet et al., 2012).
Collectively, information pertaining to the effects of sleep deprivation on human behavioural thermoregulation is lacking.
Further work is also needed to elucidate whether the magnitude of thermoreceptive and thermonociceptive adaptations potentially induced by sleep deprivation would exert an influence on the decisionmaking to thermoregulate behaviourally.

CONCLUDING REMARKS
This brief review demonstrates that, in humans, a paucity of data exists concerning the distinct effects of relatively short periods (≤4 nights) of sleep deprivation on the function of autonomic and behavioural thermoeffectors. A few studies have provided some evidence that inadequate sleep might disturb thermal homeostasis ( Figure 1); however, additional research is required to establish whether sleep loss constitutes a risk factor either causal or contributing to the development of thermal injuries. Accordingly, future investigations should evaluate this notion in a larger cohort, involving both male and female participants, use counterbalanced cross-over designs and, preferably, use ecologically valid protocols with respect to the sleepdeprivation treatments and the thermal provocations. In this context, the ability to thermoregulate after periods of sustained wakefulness needs to be assessed specifically under uncompensable thermal loads; for instance, during whole-body cold-water immersion (i.e., moderate to severe cold stress) and during prolonged and demanding work performed in hot and humid conditions while wearing protective clothing ensembles. Whether or to what extent sleep deprivation causes circadian dysregulation of thermoeffector capacity should also be determined. In this regard, the levels of melatonin, which modify skin circulation (Aoki et al., 2006(Aoki et al., , 2008 and are probably linked with the diurnal fluctuations in cutaneous vasomotor tone (Krauchi & Deboer, 2010), are enhanced by sleep deprivation (Salín-Pascual et al., 1988;Zeitzer et al., 2007); nevertheless, the contribution of the sleep deprivation-induced changes in melatonin to human thermoregulation while in thermal extremes is not known. Studies should also seek to examine, in a well-controlled manner, the interaction of sleep deficit and other non-thermal stressors [e.g., (mal)nutrition, dehydration, physical exertion, altitude], conditions that are often encountered simultaneously in real-life settings, on temperature regulation. Lastly, the influence of behavioural countermeasures, commonly used to minimize the functional impairments caused by sleep deprivation