European Journal of Applied Physiology

, Volume 113, Issue 9, pp 2353–2360

Three nights of sleep deprivation does not alter thermal strain during exercise in the heat


    • Extremes Research GroupCollege of Health and Behavioural Sciences, Bangor University
  • Adam D. Harper Smith
    • Extremes Research GroupCollege of Health and Behavioural Sciences, Bangor University
  • Umberto Di Felice
    • Extremes Research GroupCollege of Health and Behavioural Sciences, Bangor University
    • Department of Biomedical Sciences and TechnologiesUniversity of L’Aquila
  • Neil P. Walsh
    • Extremes Research GroupCollege of Health and Behavioural Sciences, Bangor University
Original Article

DOI: 10.1007/s00421-013-2671-2

Cite this article as:
Moore, J.P., Harper Smith, A.D., Di Felice, U. et al. Eur J Appl Physiol (2013) 113: 2353. doi:10.1007/s00421-013-2671-2



Individuals exposed to total sleep deprivation may experience an increased risk of impaired thermoregulation and physiological strain during prolonged physical activity in the heat. However, little is known of the impact of more relevant partial sleep deprivation (PSD). This randomized counterbalanced study investigated the effect of PSD on thermal strain during an exercise-heat stress.


Ten healthy individuals performed two stress tests (45 min running, 70 % \({\dot{V}\text{O}}_{2\hbox{max} }\) 33 °C, 40 % RH). Each trial followed three nights of controlled sleep: normal [479 (SD 2) min sleep night−1; Norm] and PSD [116 (SD 4) min sleep night−1]. Energy balance and hydration state were controlled throughout the trials. Rectal temperatures (Tre), mean skin temperature (\(\bar{T}_{\text{sk}}\)), heart rate (HR), RPE, and thermal sensations (TS) were measured at regular intervals during each heat stress trial.


There was a significant main effect of time (P < 0.05) for all of these variables. However, no differences (P > 0.05) were observed between PSD and Norm, respectively, for Tre [39.0 (0.5) vs. 39.1 (0.5)  °C], \(\bar{T}_{\text{sk}}\), [36.1 (0.6) vs. 36.0 (0.7)  °C] and HR [181 (13) vs. 182 (13) beats min−1)] at the end of exercise-heat stress. There were no differences (P > 0.05) in \(\bar{T}_{\text{sk}}\), PSI, RPE, TS and whole-body sweat rate between PSD versus Norm.


Since greater physiological strain during exercise-heat stress did not follow three nights of PSD, it appears that sleep loss may have minimal impact upon thermal strain during exercise in the heat, at least as evaluated within this experiment.


Exertional heat stressBody core temperatureSleep deprivationThermal strain


Physically stressful military, occupational and recreational tasks performed in hot environments are frequently associated with thermal strain. Furthermore, under such extreme conditions, there is an increased risk of an individual developing exertional heat stroke (EHS), primarily due to a failure of heat dissipation (Sawka et al. 2011). Although multiple risk factors exist for EHS, limited sleep, a common feature of military and other operations, is frequently cited in reviews and position statements as an important predisposing factor (e.g., Armstrong et al. 2007; Casa et al. 2012; Epstein and Roberts 2011). This is despite relatively limited knowledge of how sleep deprivation influences physiological responses during strenuous physical activity in the heat.

Laboratory studies of thermoregulatory effects during exercise-heat stress following one sleepless night (i.e., total sleep deprivation, TSD) report subtle impairment of autonomic (sweating and vasodilatation) temperature regulation (Dewasmes et al. 1993; Kolka and Stephenson 1988; Sawka et al. 1984). The mechanism for modification of sweating and vasodilatation during exercise-heat stress following one night of TSD is unclear. It could be due to altered central nervous system function or changes in peripheral input. However, in our view, an important observation from all these studies is that not one found a significantly greater elevation of core temperature (Tc) during exercise-heat stress following TSD. Thus, even though there may be a change in localized responses there is no evidence of any effect on whole-body thermoregulation. This is frequently overlooked in reviews and position statements relating to the role of sleep loss and deprivation in heat stress and exercise Thus, based on our assessment of the literature, the impact of sleep deprivation on exercise thermoregulation remains far from certain.

In contrast to one night of continuous sleep deprivation, partial sleep deprivation (PSD) i.e., sleep restriction lasting several nights is more representative of interrupted sleep that occurs because of disturbances associated with time zone transitions (Reilly et al. 1997), prolonged military continuous operations (Olsen et al. 2010), or changes in working patterns associated with shift work (Akerstedt 2003). To our knowledge, no investigation exists that has evaluated the impact of PSD on whole-body thermoregulation and thermal strain during exercise in the heat. Therefore, the absence of information relating to possible outcomes accompanying PSD, taken together with an apparent lack of increased Tc following TSD, indicates to us that the impact of sleep deprivation on thermoregulation in the heat has not yet been resolved.

The purpose of this study was to evaluate the effect of PSD on thermal strain during steady-state exercise in a moderately hot environment. Sleep was restricted to 6 h over three nights. Care was taken to isolate PSD from other factors that are reported to modify thermoregulation, such as circadian rhythm (Stephenson et al. 1984), hydration state (Montain et al. 1995), and chronic exertional fatigue, and negative energy balance (Young et al. 1998). Evaluation of physiological strain was by measurement and interpretation of the following parameters: rectal temperature (Tre), mean skin temperature (\(\bar{T}_{\text{sk}}\)), heart rate (HR), and body mass loss due to sweating (Δmsw). Based upon our uncertainty surrounding the effects of one night of TSD, we were unsure how three nights of PSD, which is more practically relevant, might affect physiological strain in unacclimatized males exercising in the heat.



Twelve participants, recruited from the student community, originally volunteered for the study. All the volunteers provided their written informed consent to procedures approved by the local Research Ethics Committee. To eliminate the confounding influences of gender on physiological responses to heat stress, only non-acclimatized, non-smoking males, free from any known cardiovascular, metabolic, and respiratory diseases were considered. One participant withdrew from the study due to time commitments, another was excluded once it became apparent that he might have become heat acclimatized while on vacation before entering the study. The physical characteristics of ten participants who completed the study are as follows (mean ± SD): age = 20.5 ± 1.6 years, height = 1.79 ± 0.09 m, body mass = 71.2 ± 8.5 kg, body fat = 14.2 ± 3.0 %, body surface area = 1.89 ± 0.2 m2, \({\dot{V}\text{O}}_{2\hbox{max} }\) = 63.1 ± 4.1 ml kg−1 min−1. Participants visited the laboratory on four separate occasions, consisting of preliminary measurements, a familiarization session, and two experimental trials. Data were collected during winter and spring months (December–April) to minimize the effects of natural heat acclimatization. Participants were instructed to refrain from strenuous exercise for 24 h, and to avoid the consumption of caffeine and alcohol for 72 h, before each visit to the laboratory.

Preliminary measurements

Body height and nude body mass (NBM) were measured using a precision stadiometer and balance (Seca 705, Hamburg, Germany), respectively. Body surface area (AD) was subsequently calculated from the measurements of body height and mass (DuBois and DuBois 1916). Body composition was determined by whole-body dual X-ray absorptiometry (DXA; Hologic QDR1500, Bedford MA, USA). Maximal oxygen uptake \(\left( {{\dot{V}\text{O}}_{2\hbox{max} } } \right)\) was determined by automated indirect calorimetry (Metalyzer 3B, Cortex Biophysik, Leipzig, Germany) during a progressive incremental running protocol (Oliver et al. 2007) on a motorized treadmill (h/p/cosmos Mercury 4.0, Nussdorf-Traunstein, Germany). The treadmill speeds that elicited 50 and 70 % \({\dot{V}\text{O}}_{2\hbox{max} }\) at a 1 % gradient were extrapolated, these were verified by a subsequent bout of treadmill exercise.


Seven days after their first visit, participants reported to the laboratory for a familiarization session that included 45 min of steady-state sub-maximal exercise, running on a treadmill in a climate chamber (Design Environmental Ltd, Ebbw Vale, UK) at 33 °C and 40 % RH and air speed of 2.3 m s−1. During this session, the participants were familiarized with the 15-point Borg scale for perceived effort (RPE) (Borg 1970) and a 13-point scale for the measurement of thermal sensation (Hollies 1977).

Experimental procedures

Participants completed two randomized and counterbalanced exercise-heat stress tests separated by a minimum of 14 days. Prior to each test, participants performed a treatment that involved them remaining in a laboratory suite under supervision for a period encompassing three nights (Fig. 1). For PSD, sleep was restricted to six 1-h sleep periods, available at the same time every night (i. e, 23:45 and 04:45 h). In the case of the normal sleep treatment (Norm), three 8-h sleep periods were available, each one between 23:45 and 07:45 h.
Fig. 1

A schematic representation of the experiment, from Start on day 1 to Finish on Day 4. Note that the diagram only shows scheduled sleep periods for the partial sleep deprivation (PSD) trial. For details about the sleep periods for the normal sleep (Norm) trial and a description of activities pursued by participants when not engaged in procedures specific to the experiment, please refer to the text

On the first day of a trial, participants reported to the laboratory suite at 08:00 h, following a night of normal sleep. When not engaged in procedures specific to the experiment, participants undertook educational activities, read, watched videos, played games and conversed with other participants, research staff or visitors to the laboratory. Participant compliance was verified by continuous monitoring by research staff. To control for the effect of energy and fluid intake on the responses to exercise-heat stress, participants were provided with standardized meals designed to meet individual daily requirements (2,818 ± 193 kcal day−1) (Cunningham 1980). Meal timing was consistent between trials, starting in the evening prior to the first day of a trial. Fluid intake equivalent to 40 ml kg−1 NBM day−1 was provided in equal volumes every 3 h. Hydration status was monitored via urine specific gravity (USG) measured at the same time each morning (Armstrong et al. 1994). Urine samples were collected into universal containers and immediately analyzed for USG using a handheld refractometer (Atago Uricon-Ne, NSG Precision Cells, Farmingdale, NY).

Nocturnal sleep (apparent inactivity) was monitored by accelerometry (GTIM, ActiGraph LLC, Florida, USA). Physical activity was monitored via electronic pedometers (Digi-walker SW-200, Yamax, Tokyo, Japan). Furthermore, participants undertook a daily bout of low intensity exercise (walking 50 % \({\dot{V}\text{O}}_{2\hbox{max} }\), 60 min day−1), performed at 20 °C and 50 % RH and constant air speed. The exercise bout, included to reproduce regular activity on military duty, took place at the same time of the afternoon. On each occasion, participants were allowed to ingest water ad libitum from the allotted amount of their daily dose.

On the fourth morning, following instrumentation and measurement of pre-test body mass, participants entered a climate chamber at 33 °C, 40 % RH, and controlled air speed (2.3 m s−1). They rested for 45 min seated upright, before completing 45-min steady-state exercise running on a treadmill at 70 % of \({\dot{V}\text{O}}_{2\hbox{max} }\). The exercise-heat stress test was performed at the same time, between 1100 and 1300 h, on the fourth day of each trial. Participants wore standardized clothing i.e., running shorts, socks and shoes. Each individual was monitored continuously throughout stabilization and exercise-heat stress. Water was prescribed at a dose of 3 ml kg−1 NBM during stabilization and 5 ml kg−1 NBM during exercise-heat stress. If Tre reached 39.9 °C or anyone exhibited any other signs or symptoms of EHI, he would have been removed from the chamber immediately, even if exercise was incomplete. After exiting the chamber, post-test body mass was measured and the participant was given water to replace sweat loss during the exercise stress.


Tre, an index of body core temperature, was measured using a soft thermistor probe (Grant REC, Shepreth, UK) inserted approximately 12 cm above the anal sphincter. Contact skin temperature was measured at four points (on the chest at a point midway between the acromion process and the nipple, the anterior mid-bicep, the anterior mid-thigh, and the medial calf) using wireless high-resolution Thermocron iButtons (DS1921H-F5, iButton®, Maxim Integrated Products USA) fixed to the skin using surgical tape. Previously, we have reported that this technique is valid for human skin temperature measurement (Smith et al. 2010). Temperature data were registered using a portable data logger (Grant SQ2020) at a rate of one sample every second, simultaneously displayed, and recorded manually every minute. Subsequently, the 4-point weighting equation of Ramanathan (1964) was used to calculate \(\bar{T}_{\text{sk}}\).

Heart rate (HR) was monitored continuously using a coded transmitter (T-31 Polar Electro Oy, Kempele, Finland) and recorded every 5 min. A physiological heat strain index (PSI) was calculated as follows:
$${\text{PSI}} = 5(T_{\text{ret}} - T_{\text{re0}} ) \times (39.5 - T_{\text{re0}} )^{ - 1} + \, 5({\text{HR}}_{t} - {\text{HR}}_{0} ) \times ({\text{HR}}_{\hbox{max} } - {\text{HR}}_{0} )^{ - 1}$$
where Tre0 and HR0 are initial measurements, Tret and HRt are measured at any time during exercise and HRmax was the value achieved during the maximal exercise test (Tikusis et al. 2002).
Participants were also asked to rate RPE and TS at 5 min intervals during the exercise-heat stress test. Whole-body sweat losses during exercise were estimated as follows:
$$\left[ {\Updelta {\text{NBM}} + {\text{FI}} - {\text{UV }}} \right]/{\text{time}}\,\left( {\text{h}} \right),$$
where ΔNBM is the difference in NBM pre–post exercise, FI is fluid intake and UV is urine volume output produced between measurements. Post-exercise NBM was recorded following complete drying of the subject to remove surface sweat. The contribution of non-sweat losses (i.e., respiratory losses or metabolic exchange) was deemed constant and equal for all subjects, due to the fixed exercise intensity.

Statistical analyses

Using conventional alpha (0.05) and beta (0.20) assumptions, we estimated that ten participants would provide sufficient power to detect a meaningful minimum condition difference in Tre at the end of exercise-heat stress (two-tailed). The estimate was based upon a minimum important difference of 0.5 °C, which is half of the difference between the upper clinical thresholds for mild hyperthermia (38.5 °C) and moderate hyperthermia (heat exhaustion) (39.5 °C) (Taylor et al. 2008). Variability in the measurement of Tre was estimated at 0.25 °C (Consolazio et al. 1963). All statistical calculations were performed using SPSS version 15.0 (SPSS inc., Chicago, USA). The level of significance for all analyses was set at an alpha level of 0.05. Comparisons between values for sleep duration, physical activity, NBM, USG, and whole-body sweat rate were investigated using Student’s paired t tests. The 95 % confidence interval for the mean difference in Tre at the end of exercise-heat stress between conditions was also calculated. Two-factor (i.e., sleep condition and time), repeated measures ANOVA were performed to evaluate differences in Tre, \(\bar{T}_{\text{sk}} ,\) HR, PSI, RPE and thermal sensation (TS) within and between trials, using a sampling period of 5 min. For data where Mauchly’s sphericity assumption was violated, the Greenhouse-Geisser correction factor to the degrees of freedom was applied. Data are expressed throughout as means and standard deviation (SD), unless otherwise indicated.


By design, total sleep time per night (accelerometry) was significantly shorter during PSD compared to Norm (116 (4) vs. 479 (2) min sleep night−1, P < 0.001). Physical activity (pedometry) was not significantly different (Norm = 14,430 (3,355) steps vs. PSD = 15,760 (2,298) steps; P = 0.31). There was no significant difference in USG (Norm = 1.011 (0.01) vs. PSD = 1.016 (0.14) g mL−1; P = 0.21) measured before entering the environmental chamber on the day of an exercise-heat stress test. The total volume of water ingested during stabilization and exercise-heat stress in the environmental chamber was not different between trials (Norm = 0.59 (0.09) L vs. PSD = 0.56 (0.06) L; P = 0.20).

Physiological responses during stabilization

Mean values for Tre, \(\bar{T}_{\text{sk}}\) and HR recorded at the start and end of stabilization are presented in Table 1. During stabilization, there were no significant differences in Tre (P = 0.86), \(\bar{T}_{\text{sk}}\) (P = 0.47) or HR (P = 0.317) for PSD vs. Norm. Furthermore, Tre did not change with time (P = 0.61) during stabilization. However, there was a main effect of time for \(\bar{T}_{\text{sk}}\) (P < 0.001) and HR (P < 0.001) during seated rested in the climate chamber. There was no significant interaction between trial and time for Tre (P = 0.80), \(\bar{T}_{\text{sk}}\) (P = 0.54) or HR (P = 0.99).
Table 1

Physiological measurements at the start and at the end of 45-min stabilization (seated rest) in a climate controlled chamber (33 °C and 40 % RH)








Tre (°C)

37.09 (0.32)

37.13 (0.23)

37.11 (0.43)

37.14 (0.26)

\(\bar{T}_{\text{sk}}\) (°C)

31.78 (0.75)

35.14 (0.38)

32.06 (0.38)

35.14 (0.34)

HR (beats min−1)

80 (15)

86 (12)

85 (19)

87 (15)

Ten healthy males were studied in the middle of the day following three nights of normal sleep (8 h available sleep per night sleep; Norm) and three nights of partial sleep deprivation (2 h available sleep per night, PSD). Data are means and standard deviations

Physiological and perceptual responses during exercise-heat stress

The Tre after 45 min of moderate exercise-heat stress in the environmental chamber in both Norm and PSD conditions is shown for each individual in Fig. 2. For both conditions, at the end of exercise-heat stress all participants were either mildly or moderately hyperthermic. However, the condition difference in Tre was equal to or greater the minimum important difference in only three out of ten participants. Furthermore, PSD had no consistent effect on Tre i.e., values were higher in five participants and lower in five participants compared with Norm. Consequently, no significant differences were found for Tre values (Norm = 39.10 (0.45)  °C; PSD = 39.05 (0.46)  °C; 95 % CI of the difference = −0.25 to 0.35 °C, P = 0.71).
Fig. 2

Individual differences in rectal temperature between Normal and PSD sleep trials after steady-state exercise (running for 45 min at 70 % \({\dot{V}\text{O}}_{2\hbox{max} }\)) in a climate controlled chamber (33 °C and 40 % RH). Mean data and 95 % CI for the difference in the means of ten healthy males are also presented

Figure 3 shows the mean responses of Tre, \(\bar{T}_{\text{sk}}\) and HR calculated at 5-min intervals during the time course of the exercise-heat stress test. A significant main effect of time was found for each of Tre (P < 0.001), \(\bar{T}_{\text{sk}}\). (P < 0.001), and HR (P < 0.01). However, there was no significant effect of trial for any of these physiological variables (Tre, P = 0.52; \(\bar{T}_{\text{sk}}\), P = 0.52; HR, P = 0.19). There was no significant interaction (P > 0.05) between trial and time for any of these variables.
Fig. 3

Rectal temperature a, mean skin temperature b and heart rate c recorded at 5-min intervals during steady-state exercise (running for 45 min at 70 % \({\dot{V}\text{O}}_{2\hbox{max} }\)) in a climate controlled chamber (33 °C and 40 % RH). Ten healthy males were studied in the middle of the day following three nights of normal sleep (8 h available sleep per night sleep; Norm) and three nights of partial sleep deprivation (2 h available sleep per night, PSD) Data are means and standard deviations

As expected, time had a significant effect (P < 0.001) on PSI during exercise-heat stress. PSI increased from 3.5 (0.5) to 8.4 (1.3) in the Norm trial and from 3.5 (0.5) to 8.5 (1.2) in the PSD trial. However, there was no significant difference (P = 0.68) between trials, and no significant interaction (P = 0.28) between time and trial.

Whole-body sweating rate during exercise-heat stress was not different between trials (Norm = 1.30 (0.41) L h−1 vs. PSD = 1.26 (0.42) L h−1; P = 0.77).

Neither RPE (P = 0.22) nor TS (P = 0.11) during exercise-heat stress was significantly different between trials, although there was a main effect for time (RPE, P = 0.04; TS, P = 0.02, Fig. 4). There was no significant interaction (P > 0.05) between trial and time for RPE or TS.
Fig. 4

Measures of thermal sensation a and perceived exertion (RPE) b recorded at 5-min intervals during steady-state exercise (running for 45 min at 70 % \({\dot{V}\text{O}}_{2\hbox{max} }\)) in a climate controlled chamber (33 °C and 40 % RH). Ten healthy males were studied in the middle of the day following three nights of normal sleep (8 h available sleep per night sleep; Norm) and three nights of partial sleep deprivation (2 h available sleep per night, PSD). Data are means and standard deviations


This study evaluated the impact of sleep deprivation on whole-body thermoregulation and physiological heat strain during moderate exercise-heat stress. Previous studies have investigated the effects of one night without sleep, but ours is the first to investigate partial sleep deprivation. PSD may be more practically relevant in military, occupational, and recreational settings. With control of circadian rhythm, energy balance, hydration state, and daily physical activity, we observed that sleep restricted to around 6 h in total over three nights had no impact upon Tre, \(\bar{T}_{\text{sk}} ,\) HR, PSI, WBSR, RPE, and TS during exercise-heat stress in unacclimatized males. These findings indicate that PSD does not alter physiological strain during exercise-heat stress, at least when sleep loss is isolated from other potentially detrimental factors. This conclusion challenges the notion that sleep deprivation, in this case PSD, is associated with impaired whole-body thermoregulation and increased thermal strain during exercise in the heat. In addition, this new information about PSD is important given that many reviews and position statements report that sleep loss might predispose an individual to EHS (Armstrong et al. 2007).

Athletes, military personnel and those in physically demanding occupations often combine periods of prolonged wakefulness with vigorous physical activity in a hot environment. Furthermore, in addition to sleep deprivation, dehydration, negative energy balance and exertional fatigue are also possible. Each of these has been identified as an individual factor that predisposes humans to accelerated hyperthermia and EHS (Casa et al. 2012). Experimental TSD is a useful and widely used paradigm for studying the effects of sleep loss, and thermoregulatory control during exercise-heat stress has been investigated after one sleepless night (Dewasmes et al. 1993; Kolka and Stephenson 1988; Sawka et al. 1984). Small changes (shifts in threshold and or sensitivity) in local sweating and cutaneous vasodilatation have been observed, possibly due to altered central nervous function. This has lead to a speculation that the thermoregulatory effector responses for a given mean body temperature may be impaired following sleep loss (Sawka et al. 2011).

In contrast to TSD, experimental PSD is more representative of occupational sleep loss and deprivation and, to our knowledge, there has been no laboratory study of thermoregulation during exercise-heat stress following PSD. Therefore, we selected a sleeping pattern specifically to simulate periods of wakefulness interspersed with short sleep bouts typical of military training for operations lasting several days or more (Olsen et al. 2010). A similar sleep paradigm has already been used by studies that investigate thermoregulation in military personnel exposed to multifactorial stress (Castellani et al. 2003). In the present study, however, we carefully controlled acclimation status, circadian rhythm, and hydration state, all of which are factors reported to alter the control of thermoregulatory responses (Sawka et al. 2011).

In the present study, participants sat quietly for 45 min in an environmental chamber (33 °C and 40 % RH) to allow for thermal equilibration prior to exercise-heat stress. This period of stabilization was accompanied by a significant elevation of \(\bar{T}_{\text{sk}}\), but not Tre. No significant differences between conditions were found in either variable such that exercise-stress tests commenced at similar levels of Tre and \(\bar{T}_{\text{sk}}\) in both trials. Exercise-heat stress consisted of 45 min of treadmill running at 70 % \({\dot{V}\text{O}}_{ 2max}\) during which we were unable to detect any statistically significant difference in responses of Tre, \(\bar{T}_{\text{sk}}\) and HR. Furthermore, we were only able to detect the minimum important change in Tre in 30 % of participants and this effect was inconsistent i.e., an increase in Tre in one participant and a decrease in Tre in two. Finally, we did not observe any differences in the ratings of perceived exertion or thermal sensation between PSD and Norm trials. Taken together this physiological and perceptual data suggests that sleep deprivation, at least in the form of PSD, has minimal impact upon thermal strain during moderate exercise-heat stress.

Several differences between our study and previous investigations require further discussion. Notwithstanding variation across studies in relation to the time of day, heat stress test (mode, intensity, duration and ambient temperature), as well as the season of the year during which a study was conducted, the most obvious difference is the nature of sleep deprivation i.e., total vs. partial. It could be argued that, even though sleep was limited to around 6 h in three and a half days, the PSD paradigm investigated here was not as strenuous as continuous sleep deprivation used in previous studies night (Dewasmes et al. 1993; Kolka and Stephenson 1988; Sawka et al. 1984). Therefore, perhaps it is only to be expected that no difference in thermal strain be observed between PSD vs. Norm trials in this study. Nevertheless, PSD is a realistic everyday situation and, in our view, worthy of laboratory investigation. Furthermore, even though there are reports of impaired thermoregulatory responses following continuous sleep deprivation, this does not necessarily translate into increased physiological strain during exercise-heat stress. In fact, none of the previously cited TSD studies found any differences in core (esophageal, Tes) temperature, \(\bar{T}_{\text{sk}} ,\) or HR during exercise-heat stress following one night without sleep TSD. This is overlooked in review articles and position statements citing that sleep loss is a risk factor for EHS. The present study corroborates the past findings for measurements of thermal strain, but here we used a more relevant paradigm of sleep loss experienced during prolonged military continuous operations, shift work, and time zone transitions.

In summary, in this controlled experiment, we found no evidence of any increase in thermal strain during moderate exercise-heat stress following PSD in healthy young males. This leads us to conclude that impaired whole-body thermoregulation during exercise is not a feature of three nights of PSD. Furthermore, although the literature frequently suggests that an interaction of sleep loss and accelerated hyperthermia may predispose to EHS, we suggest that the contribution of sleep deprivation, and PSD in particular, is worthy of reconsideration. This is the case, at least, when the effects of other factors known to influence thermoregulation during physical activity can be limited.


The Army Recruiting and Training Division supported the work. The authors would like to acknowledge the volunteers who participated in the study. The authors would also like to thank Katharine Richardson for her assistance with data collection, and Kevin Williams for his expert technical assistance.

Conflict of interest

The authors have no conflicts of interest to declare.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013