Secondary Logo

Journal Logo

Physiological Responses to Overdressing and Exercise-Heat Stress in Trained Runners


Medicine & Science in Sports & Exercise: June 2018 - Volume 50 - Issue 6 - p 1285–1296
doi: 10.1249/MSS.0000000000001550

Heat acclimation is the best strategy to improve performance in a hot environment. Many athletes seeking the benefits of heat acclimation lack access to a hot environment for exercise and, thus, rely on overdressing to simulate environmental heat stress. It is currently unknown whether this approach produces the requisite thermoregulatory strain necessary for heat acclimation in trained men and women.

Purpose To compare physiological and cellular responses to exercise in a hot environment (HOT; 40°C, 30% RH) with minimal clothing (clo = 0.87) and in a temperate environment (CLO; 15°C, 50% RH) with overdressing (clo = 1.89) in both men and women.

Methods HR, rectal temperature (Tre), mean skin temperature (Tsk), sweating rate (SR), and extracellular heat shock protein (eHSP)72 were measured in 13 (7 males, 6 females) well-trained runners (V˙O2max: 58.7 ± 10.7 mL·kg−1·min−1) in response to ~60 min of treadmill running at 50%–60% V˙O2max in HOT and CLO.

Results Tre increased in both conditions, but the increase was greater in HOT (ΔTre HOT: 2.6°C ± 0.1°C; CLO 2.0°C ± 0.1°C; P = 0.0003). SR was also higher in HOT (1.41 ± 0.1 L h−1; CLO: 1.16 ± 0.1 L·h−1; P = 0.0001). eHSP72 increased in HOT (% change: 59% ± 11%; P = 0.03) but not in CLO (6% ± 2%; P = 0.31). Mean Tsk and HR were not different between HOT and CLO in men but were higher in HOT for women.

Conclusion These data support the idea that overdressing during exercise in a temperate environment may produce the high Tre, Tsk, HR, and SR necessary for adaptation, but these responses do not match those in hot, dry environments. It is possible that greater exercise stimulus, warmer environment, or more clothing may be required to allow for a similar level of acclimation.

1University of Oregon Department of Human Physiology, Eugene, OR; and

2United States Army Research Institute of Environmental Medicine, Natick, MA

Address for correspondence: Christopher T. Minson, Ph.D., Department of Human Physiology, University of Oregon, 122C Esslinger Hall, Eugene, OR 97403; E-mail:

Submitted for publication May 2017.

Accepted for publication December 2017.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Web site (

Heat acclimation requires moderate-intensity exercise in a hot environment (~40°C) be performed repeatedly over 5–14 d to improve heat tolerance and performance in the heat (1). This potential performance benefit is appealing for elite athletes seeking a competitive advantage in hot weather competition, but many athletes live and train in areas of the world with more temperate climates and do not have access to the resources (i.e., environmental chamber) necessary for exercise–heat acclimation. Promising alternate heat acclimation strategies for athletes include postexercise hot tub (2) or sauna bathing (3); however, overdressing in temperate conditions may be the simplest, most cost-effective model while also being least likely to disrupt their training program.

Training in “sweat clothing” or encapsulating military garments has been investigated for its utility in improving heat tolerance in healthy young men in sport (4) and occupational settings (5). Gross physiological responses have been measured in physically fit men comparing a warm/humid environment (35°C, 60% RH) with minimal clothing and a temperate (19°C, 50% RH) environment with sweat clothing (6), and more recently, HR, sweating rate, and temperature responses have been examined in trained male cyclists performing an 80-min bout of outdoor cycling while wearing “winter clothing” (7). No study has yet examined overdressing in well-trained men and women and compared physiological, perceptual, and cellular responses to standard exercise–heat acclimation conditions. In addition, making this comparison with the hot/dry environmental conditions used in heat acclimation protocols where performance benefits are observed (8) allows for further examination of the potential for overdressing to mimic heat acclimation toward performance benefits. Although overdressing appears to create a similar heat tolerance benefit as warm/humid heat acclimation, it is currently unknown how effective this may be in well-trained endurance athletes, who are capable of producing extremely high heat loads for prolonged periods of time (9,10) and who rely heavily on evaporative heat loss and a favorable core-to-skin gradient for this work to remain compensable (11). By examining the acute responses in well-trained athletes between traditional exercise–heat acclimation and overdressing during exercise in a temperate environment, insight may be gained to design appropriate clothing ensembles and work rates to achieve the requisite stimuli for heat acclimation while avoiding risk of exertional heat injury in endurance athletes. In advance of hot-weather Olympic Games and World Cup competitions, the potential utility of overdressing in well-trained athletes requires further study. Whether overdressing can create a local environment sufficient to raise body core temperature consistent with environmental heat stress protocols that induce heat acclimation is not known, and currently there are very limited data on which to build recommendations for overdressing strategies in athletes.

In addition, overdressing has only been studied in men (4–6), and the higher body surface area (BSA) to mass ratios (12) and potentially reduced sudomotor function (13) of women as compared with men may alter the response to overdressing, which relies primarily upon limiting evaporative heat loss to drive increased core and skin temperatures. Although these differences between men and women are minimized when workload is equated to body mass or BSA, the lower absolute workload and smaller body mass of women (reduced metabolic heat production creating a smaller increase in temperature within clothing) coupled with a lower sweating response (smaller increase in relative humidity and dew point within clothing) may not be adequate to create a thermally stressful microenvironment within the clothing.

The stimuli for heat acclimation are multifaceted and combine thermoregulatory, cellular, and perceptual stress. Optimal heat acclimation requires repeated bouts of elevated core temperature >38.5°C (14), elevated skin temperature (15), and a sweating response resulting in high skin wettedness (16,17). In addition, recent research has highlighted the cellular responses required for heat acclimation in animals (18) and humans (19). As more is being learned about the cellular and systemic stimuli required for heat acclimation, examining the physiological responses as well as cellular stress signals may provide insight into the efficacy of overdressing to produce heat acclimation and associated benefits. Extracellular heat shock protein 72 (eHSP72) has been found to increase acutely with elevations in body core temperature (20) as observed during exercise in the heat (21,22) and is thought to be involved in an immune response to exercise and hyperthermia (23). Although not universally noted (24), acute bouts of exercise-heat stress result in transient increases in eHSP72 (22,25), whereas heat acclimation results in increases in intracellular HSP72 (26) and decreases in eHSP72 (22). The specific mechanisms by which intracellular and extracellular HSP72 contribute to the acclimation response have not yet been elucidated; however, when HSP production is pharmacologically blocked, heat acclimation responses are severely blunted (19), suggesting HSP72 signaling may be an important acute response for adaptation to heat.

The purpose of this study was to examine the potential for insulative clothing worn in a controlled temperate environment to simulate physiological and cellular stress similar to chamber conditions used in heat acclimation protocols in well-trained male and female endurance athletes. A secondary purpose was to examine heat balance variables to develop guidance on the optimal heat stimulus for overdressing for use in heat acclimation protocols in well-trained endurance athletes. It was hypothesized that overdressing during exercise would cause similar increases in core temperature, skin temperature, HR, and sweating rate in both men and women. In addition, it was hypothesized that eHSP72 would increase in both conditions pre- to postexercise. The results of this study could lead to the validation of a low-cost, practical model for athletes seeking the performance benefits of exercise–heat acclimation while living and training in a temperate climate.

Back to Top | Article Outline



Thirteen (7 males, 6 females) well-trained runners participated in this study. All subjects provided oral and written informed consent before participation in the study. All experimental procedures were approved by the Institutional Review Board at the University of Oregon. Subjects were young (24 ± 6), healthy (no medications, no chronic diseases, and no history of heat illness), and not heat acclimated (not having trained in weather >25°C in 3 months or undergone a heat acclimation protocol in >6 months). Although all well-trained athletes exhibit some degree of heat acclimation, this testing took place in Eugene, Oregon, between January and June, a period in which monthly average high temperature falls between 8°C and 22°C. To qualify for participation, subjects had to perform aerobic exercise >7 h·wk−1 and exceed the 85% for maximal oxygen uptake (V˙O2max) for age and sex. A summary of physical characteristics is listed in Table 1.

All subjects underwent V˙O2max testing before study participation. Briefly, they reported to the laboratory having fasted for at least 2 h and having refrained from exercise for at least 12 h. Height and body mass were measured, and the subject was then allowed to warm up on a treadmill for 10–15 min at a self-selected speed. After warm-up, subjects were instrumented with an HR monitor and Hans Rudolph two-way valve and mouthpiece (Shawnee, KS) to measure V˙O2 via open circuit spirometry (ParvoMedics, Sandy, UT). The graded exercise test was performed at 7 mph (for men) or 6 mph (for women) and 0% grade for the first 2 min. The grade was increased by 2% every 2 min until volitional fatigue. V˙O2max was determined as the peak V˙O2 over a 15-s increment during the last completed stage of the test. The experimental sessions were scheduled at least 48 h after the V˙O2max testing day.

Back to Top | Article Outline

Experimental sessions

All subjects participated in two experimental sessions separated by at least 7 d. The hot trial (HOT) took place in an environmental chamber set to 40°C, 30% RH environment (wet bulb temperature [Twb] = 25.3°C, dew point temperature [Tdp] = 19.1°C) with 1 m·s−1 wind speed, and subjects were minimally clothed. The overdressing trial (CLO) took place in the same environmental chamber set to 15°C, 50% RH (Twb = 9.73°C, Tdp = 4.7°C), with 1 m·s−1 wind speed, and subjects were clothed in multiple layers as described in “Clothing.” In addition to the ensembles described, female subjects also wore the same sports bra for both HOT and CLO trials, and all subjects were instructed to wear the same footwear for both trials. Trials took place at the same time of day for each subject to minimize circadian influence on core temperature, and female subjects were tested during the early follicular phase (or placebo phase for oral contraceptive users). In addition, preexercise fluid intake and recent exercise sessions were held constant between the two testing days, and trial order was randomized and counterbalanced.

Subjects arrived at the laboratory having fasted for at least 2 h and having refrained from exercise for at least 12 h. All subjects were given 5 mL·kg−1 of fluid before commencing the study. After drinking the fluid, a small venous blood sample was drawn for analysis of eHSP72. Subjects were then instrumented with a flexible thermistor (YSI, Yellow Springs, OH) self-inserted 8–10 cm beyond the anal sphincter for measurement of rectal temperature (Tre). HR was measured using a chest strap (Polar Electro OY, New York, NY) and copper-constantan thermocouples were placed at four sites (Left chest, arm, thigh, and calf) for measurement of skin temperature (Tsk).

Dry seminude weight (wearing only dry running shorts and a dry sports bra for women) was measured before and after heating (Sartorius AG, Goettingen, Germany) to the nearest 0.005 kg for calculation of sweat losses. After measurement of body weight, subjects were then instructed to dress in the appropriate clothing ensemble. All subjects wore identical clothing in the same manner (fully zipped and tucked in, sleeves and ankles cinched). To quantify the microenvironment within the clothing in CLO, a portable temperature and relative humidity sensor was threaded into the suit above the base layer of clothing to the level of the xiphoid process and secured in place.

After dressing, subjects entered the environmental chamber and stood on the treadmill for baseline measurements. After all measurements were recorded, they began running on the treadmill at 1% grade and a speed to elicit 50%–60% V˙O2max (mean for all trials = 55% ± 2%). The approximate speed was precalculated using prediction equations from the American College of Sports Medicine Guidelines for Exercise Testing and Prescription and validated with metabolic feedback (ParvoMedics) in the first 3 min of exercise. The speed/grade was held constant for up to 60 min. The goal of using this range of relative intensity was to match common heat acclimation strategies in athletes (8) using submaximal workloads expressed relative to V˙O2max, rather than equating heat production and imposing higher-intensity workloads that athletes might struggle to complete in thermally stressful conditions. Although recent studies suggest matching for workload relative to body mass to match core temperature rise between individuals (27,28), the repeated-measures design allowed for comparison within an individual with workload relative to body mass held constant.

Tre, HR, and Tsk were measured continuously throughout exercise. Mean Tsk was calculated as previously described (29). In addition, metabolic measurements (V˙O2, V˙CO2, RER) were made for the first 3 min and then in 3-min increments every 10 min (minutes 7–10, 17–20, etc.), and perceptual ratings were assessed every 5 min. Perceptual scales included whole-body RPE (6–20 scale), upper and lower body rating of thermal sensation (6–20 scale), and upper and lower body skin wettedness (0–12 scale). Subjects were instructed on standard use of the perceptual scales before their first trial day. Suit temperature and RH were recorded every 5 min during exercise, and Twb within the suit was calculated for each subject throughout exercise. After completing exercise, subjects were seated in a phlebotomy chair, and a second blood sample was taken immediately. After sampling, subjects removed all clothing and instrumentation and then toweled off thoroughly before changing back into the dry running shorts and dry sports bra (women) for assessment of postexercise seminude mass. In addition, all clothing was placed in a plastic bag and weighed before testing and immediately after subjects removed it at the end of testing to account for sweat trapped in clothing.

Back to Top | Article Outline


The precise environmental conditions and clothing worn in the temperate environment were determined using the Heat Strain Decision Aid in collaboration with the United States Army Research Institute of Environmental Medicine based on metabolic rate and insulative properties of the garments to most closely simulate exercise in a hot environment (30). The primary goal was to closely match within-subject core temperature responses between the two conditions. Clothing was selected based on the modeling results of the Heat Strain Decision Aid (31) to approximate the thermal properties necessary to match thermal strain imposed by exercise in the heat, incorporating preferred garments of well-trained athletes (i.e., commercial technical clothing designed for sport). The “overdressing” ensemble was pilot tested during exercise in 15°C with 50% humidity and 1 m·s−1 airflow and compared with matched intensity exercise in the same environment in a singlet and shorts in one individual. The additional clothing led to a 1.7°C greater ΔTre (3.2°C vs 1.5°C) and an HR, on average, 22 bpm higher during exercise (141 vs 119 bpm), suggesting the clothing created substantial thermal strain beyond exercise in a temperate environment.

Insulative and evaporative properties of the clothing were tested following established standards of practice (ASTM) using a Measurement Technologies North West, Newton 20 zone articulated manikin located in an environmental chamber. Total thermal insulation (clo) and evaporative resistance (im) were measured at three wind speeds: 0.5, 1.5, and 2.4 m·s−1 with ambient conditions of Ta = 20°C and 50% RH for the clo tests and Ta = 35°C and 40% RH for the im tests. Three replications were completed at each wind speed for assessment of clo and for assessment of im with each clothing configuration (18 total replications for each ensemble). In HOT, subjects wore a sleeveless running singlet and lightweight running shorts (clo = 0.61; im/clo = 0.89). In CLO, subjects wore two midweight wicking long sleeve T-shirts, long-sleeve shirt and coat with heat-retaining lining, waterproof rain jacket, fleece hat and mittens on their upper body, and fitted short running tights, long midweight running tights, long loose-fit track pants, waterproof rain pants, socks, and running shoes on their lower body (clo = 1.89; im/clo = 0.16). The clo and im values were additionally used along with V˙O2, Tsk, sweat losses, environmental conditions, and workload to calculate heat balance in each environment using well-established heat balance equations (32), with the addition of metabolic heat production calculated relative to body mass (27,28). Metabolic heat production (expressed in W·kg−1) was additionally analyzed relative to ΔTre (°C·h−1) to examine workloads necessary for optimal heat stimulus. We defined an optimal increase as 2°C, which would allow for individuals to reach ~38.5°C–39°C by the end of exercise, a Tre range high enough to allow for adaptation with repeated exposures (14) but still safely below values associated with heat injury and below temperatures typically achieved by high-level athletes without consequence in competition (33).

Back to Top | Article Outline

Heat balance

Mean body temperature (Tb) was calculated using 0.8 and 0.2 as weighting coefficients for Tre and mean weighted Tsk, respectively (34). All heat balance components represent the integrated average of required input measurements (Ta, vapor pressure, Tsk, Tre, etc.) over the entire bout of exercise. Heat storage was calculated using thermometry rather than calorimetry because of calorimetry’s inability to precisely measure evaporative sweat losses (Esk) and thus solve the equation for heat storage. Instead, the solution to Esk was derived after applying a correction to thermometric heat storage (34). Any remaining error compared with calorimetry was assumed equal between trials. Heat production and dry heat exchange were calculated for each subject using classic partitional calorimetry. Esk was solved by subtracting heat storage from dry heat gain/loss; rate was calculated based on exercise duration. Sweat evaporative efficiency was calculated from the ratio of Esk to measured sweat rate, expressed as a percentage. All equations used are described in detail elsewhere (32).

Back to Top | Article Outline

Plasma HSP72 measurement

Venous blood samples were drawn from the antecubital fossa into EDTA tubes and immediately placed on ice. Samples were centrifuged at 4°C, 2750 RPM for 10 min, and plasma was aliquoted into cryovials and stored at −80°C until analysis. eHSP72 was analyzed using a commercially available ELISA kit (High-Sensitivity HSP 70 kit; ENZO Life Sciences, Farmingdale, NY). Recommended human plasma dilution (1:5) and a standard curve ranging from 0.2 to 12.5 ng·mL−1 (r2 = 0.999) were used. The ELISA sensitivity is reported to be 90 pg·mL−1, and the intra-assay coefficient of variation was 4.6%.

Back to Top | Article Outline


Tre, mean Tsk, HR, and perceptual scales were compared over time using a two-way repeated-measures ANOVA, and significant F values were examined using Tukey post hoc analysis. Sweating rate, mean V˙O2, ΔTre, and change in eHSP72 (ΔHSP72 pre- to postexercise) were compared between HOT and CLO trials using paired-sample t-tests. Differences between male and female responses were examined using unpaired t-tests, and Twb within the suit was compared between men and women using a two-way ANOVA with Tukey post hoc analysis where appropriate. To examine thresholds for metabolic heat production associated with an adequate increase in Tre in HOT and CLO (defined as 2°C increase from resting Tre), linear regression analysis was performed using metabolic heat production expressed in watts per kilogram and ΔTre (°C·h−1) in each environment and then statistically compared. The regression equations were then solved for a 2°C Tre increase to compare the workload (W·kg−1) necessary to achieve the requisite heat storage for adaptations.

The primary variable of interest was Tre during exercise. Using conventional β = 0.20 and α = 0.05, a sample size of five subjects was determined to be adequate to detect differences >0.25°C (normal circadian variation in Tre). Additional subjects were recruited so that each sex group, when examined independently, remained adequately powered. All statistical analyses were performed using Sigmaplot 11.0 except slope comparison of linear regression analyses, which were performed using GraphPad Prism 6.

Back to Top | Article Outline


All subjects completed both trials, but one subject stopped at 40 min and four subjects stopped at 50 min of exercise. Of the five subjects who stopped early, four were due to attainment of Tre > 39.9°C (two males in CLO, one male and one female in HOT) and one due to volitional fatigue (female in HOT). In all cases, the premature stopping occurred in the first trial, and we were able to match the second trial so that an equivalent exercise stimulus was present in both trials. With the exception of the subject experiencing volitional fatigue, these subjects had the highest heat production relative to body mass. Because all but one subject exercised for at least 50 min, Figures 1 and 2 display the first 50 min of data collection, as the final 10-min mean values of the 60-min session are skewed by the remaining individuals with the lowest work rates and core temperatures.





Temperature and HR responses to exercise are presented in Figure 1. Tre increased throughout both HOT and CLO trials but was significantly higher in HOT compared with CLO beginning at 25 min. HR and mean Tsk values were not significantly different between trials.

Changes in core temperature (ΔTre HOT: 2.6°C ± 0.1°C; CLO: 2.0°C ± 0.1°C; P = 0.0003) and sweating rate were higher in HOT (1.41 ± 0.10 L·h−1; 0.81 ± 0.14 L·m−2·h−1) compared with CLO (1.16 ± 0.11 L·h−1; 0.67 ± 0.16 L·m−2·h−1; P = 0.0001). No differences were observed for mean V˙O2 (HOT: 33.4 ± 0.4 mL·kg−1·min−1, 57% ± 2% V˙O2max; CLO: 30.8 ± 0.5 mL·kg−1·min−1, 53% ± 1% V˙O2max; P = 0.42), final RPE (HOT: 15 ± 1; CLO: 14 ± 1; P = 0.76), ratings of thermal sensation (upper body HOT: 19 ± 1 CLO: 19 ± 1; P = 0.52; lower body HOT: 19 ± 1 CLO: 18 ± 1 P = 0.38), or skin wettedness (upper body HOT: 11 ± 1 CLO: 11 ± 1; P = 0.84; lower body HOT: 10 ± 1 CLO: 10 ± 1; P = 0.31) between trials.

Back to Top | Article Outline

Sex differences in responses

Workload, calculated by metabolic heat production from V˙O2 during exercise, was higher in men than women (Table 2) when expressed relative to BSA in both HOT and CLO. When expressed relative to body mass, workload was not statistically different between men and women, although it tended to be higher in men. Change in core temperature relative to metabolic heat production per kilogram body mass (W·kg−1) was not different between men and women in HOT but was significantly lower in women as compared with men in CLO (see Table 2). Similarly, heat storage (W·m−2) was not significantly different between men and women in HOT but was lower in women in CLO (Table 2). In women, Tre was lower in CLO as compared with HOT beginning at 25 min, and mean Tsk was significantly lower in CLO beginning at 35 min (Fig. 2). In men, Tre was lower in CLO beginning at 30 min, but mean Tsk was not significantly different at any time point between HOT and CLO. HR responses were not different between trials in men; however, a main effect of trial (CLO < HOT) was observed in women.

Sweating rate was lower in women as compared with men in both HOT (women: 1.10 ± 0.12 L·h−1; men: 1.67 ± 0.07 L·h−1; P = 0.003) and CLO (women: 0.91 ± 0.14 L·h−1; men: 1.36 ± 0.12 L·h−1; P = 0.03). However, sweat trapped in clothing was significantly higher in CLO for men (0.62 ± 0.11 L) compared with women (0.30 ± 0.04 L), making the effective sweat losses more similar (men: 0.68 ± 0.05 L; women: 0.55 ± 0.07 L; P = 0.14). In both environments, women tended to have a higher evaporative efficiency (Table 2). In CLO, the Twb within the clothing was significantly lower in women versus men beginning at 5 min (see Fig. 3). The change in core temperature (ΔTre) from 0 to 50 min was more closely matched between men and women in HOT (men: 2.27°C ± 0.06°C; women: 2.03°C ± 0.1°C; difference of 0.24°C) as compared with CLO (men: 1.87°C ± 0.08°C; women: 1.37°C ± 0.13°C; difference of 0.5°C), despite equivalent workloads between trials for each individual subject. The difference in ΔTre between trials (HOT-CLO) was 0.35°C ± 0.10°C in men and 0.75°C ± 0.09°C in women (P = 0.058). Individual changes in ΔTre in HOT versus CLO as well as ΔTre in both environments plotted relative to workload (W·kg−1) are displayed in Figure 4.









Back to Top | Article Outline

Linear regression analysis

When examining the effects of metabolic heat production in watts per kilogram (x-axis) on ΔTre (y-axis) using linear regression analysis, slopes were significantly nonzero in both HOT (P = 0.0064, r = 0.71) and CLO (P = 0.0003, r = 0.84). Importantly, the slopes of the two lines were not significantly different (P = 0.74), but the y-intercept was significantly higher in HOT (P = 0.028), consistent with clothing–environment interactions despite our rational attempt to model equivalent stress (32). The requisite heat production (W·kg−1) needed to achieve ΔTre = 2.0°C was 10.3 W·kg−1 in HOT and 11.9 W·kg−1 in CLO, respectively. The 1.6 W·kg−1 difference equates, importantly, to the requirement of ~1.3 METs greater exercise intensity when overdressing.

Back to Top | Article Outline

Heat shock protein 72

Blood was obtained and analyzed from both trials for 10 of 13 subjects. One subject did not have a detectable level of HSP72 in any sample, bringing the reported n = 9 (4 males and 5 females). eHSP72 increased significantly pre- to postexercise in HOT (preexercise: 0.56 ± 0.23 ng·mL−1; postexercise: 0.67 ± 0.22 ng·mL−1; 59% ± 34% increase; P = 0.03), but not in CLO (preexercise: 0.62 ± 0.25 ng·mL−1; postexercise: 0.64 ± 0.25 ng·mL−1; 6% ± 7% increase; P = 0.31). Changes were modest (<50% change) in most subjects, and baseline (preexercise) eHSP72 levels were not different between trials. Individual magnitude of response was variable, ranging from a >300% increase to a 25% decrease (see Figure, Supplemental Digital Content 1, Which displays the individual eHSP72 responses to HOT and CLO,

Back to Top | Article Outline


This study examined physiological, perceptual, and cellular responses in men and women during exercise with overdressing as compared with exercise in a hot/dry environment. The key finding of this investigation was that CLO was not equivalent to HOT for creating the heat stress desired for inducing heat acclimation. More specifically, we found that 1) peak Tre and ΔTre were lower in CLO compared with HOT, whereas mean Tsk, HR, and perception of effort were similar between HOT and CLO; 2) measures of physiological heat strain were more closely matched between CLO versus HOT in men as compared with women; 3) metabolic heat production of 11.9 W·kg−1 was associated with an optimal Tre increase in CLO; and 4) eHSP72 increased in HOT but not CLO; however, changes were modest in most subjects. These findings suggest that overdressing is a practical model for athletes seeking the benefits of exercise–heat acclimation while living and training in a temperate climate; however, exercise intensity must be carefully considered as the acute stimulus of exercise with overdressing is not as great as the acute stimulus of traditional exercise–heat acclimation.

Heat gain during exercise is primarily dependent on exercise intensity (metabolic heat production) but can also be affected by environment (temperature, humidity, wind, and solar radiation) and an individual’s ability to dissipate heat through evaporation and other means, which is related to both environment and clothing. Although environment can have a large effect on the balance between heat gain and loss during exercise, thermal insulation and evaporative resistance properties of clothing can also dramatically alter heat dissipation in otherwise favorable environmental conditions. Furthermore, although the clo value used in this study was lower than in previous work investigating “sweat clothing” during training in team sport athletes (4), the ensemble was selected based on athlete preference and verified with biophysical modeling using the U.S. Army’s Heat Strain Decision Aid (30), which accounted for a higher metabolic heat production in highly fit athletes and predicted a greater rise in Tre than observed in the study. It is plausible that this overprediction was due to evaporative cooling being higher than anticipated due to a phenomenon known as heat piping (35), wherein sweat evaporates from the skin beneath vapor-impermeable garments then collects on a cool inner surface of clothing, rewets the skin, and undergoes a second phase change (reevaporates) at the warm skin surface. The end result is greater evaporative cooling than predicted based on measured sweat losses. It is possible that greater insulation (higher clo value) would have resulted in closer matching of Tre between HOT and CLO; however, it is also possible that heat piping would continue to allow for greater cooling despite the added insulation. In previous work investigating training in sweat clothing (clo = 2.4) to improve heat tolerance (6), lower Tre responses were similarly observed, despite additional insulation, compared with training in a hot, humid environment. In addition, both groups saw a similar decrease in end-exercise Tre over the 10 d of acclimation (6), suggesting that thermoregulatory benefits (enhanced sweating response) of heat acclimation are still apparent in overdressing. Importantly, heat tolerance did improve in subjects with sweat clothing, despite the smaller rise in Tre in this previous study (6), suggesting that heat acclimation can still occur.

The necessary thermal stimuli for heat acclimation include elevated core temperature (14), elevated Tsk (15), and high skin wettedness (sweating response) (16,17). Tre increased throughout both HOT and CLO trials; however, Tre was significantly higher after 25 min in HOT, suggesting the clothing was not as thermally stressful as the hot/dry environment. Tre ≥ 38.5°C has been examined as a critical threshold in isothermic heat acclimation protocols (14), with the total time above 38.5°C being viewed as an important stimulus for heat acclimation. All subjects surpassed this threshold in HOT, but only 10/13 attained Tre ≥38.5°C in CLO. Time above 38.5°C averaged 20 ± 2 min in HOT and 8 ± 3 min in CLO. Mean Tsk and sweating rates were high in both conditions, but whole-body sweat losses were again higher in HOT compared with CLO, driven by the higher evaporative requirement (Ereq) for heat balance (36). Lower sweating rates in CLO may be due to a combination of lower Ereq, heat piping, and possibly hidromeiosis, a decline in sweat output when skin is very wet. In addition to thermal stimuli, the training benefits of heat acclimation may be influenced in part by higher HR during exercise (37), creating an increased training stimulus in hot compared with cool environments for an equivalent workload. HR was not different between conditions during exercise, peaking at >80% age-predicted max in all subjects. The combination of these factors suggests that heat acclimation through overdressing may not be as thermally stressful as traditional exercise–heat acclimation, but the key acute physiological responses (elevated Tre, Tsk, and sweating) and training load (elevated HR) may still be adequate for adaptation, similar to recent heat acclimation protocols in athletes (38,39). Contrary to this, eHSP72 did not increase in CLO, suggesting that cellular adaptation may require greater stress (more clothing, longer duration, higher ambient temperature, or higher exercise intensity) compared with exercise in the heat. Future research examining a full heat acclimation protocol in this population is needed.

When male and female responses were separated, differences in physiological responses to overdressing (CLO) appeared to emerge between sexes. The outfits and fit were matched between men and women except sports bras in women; however, difference in Tre between CLO and HOT occurred earlier into exercise in women, and on average, the peak Tre for females did not surpass 38.5°C in CLO (Fig. 2). This blunted rise in Tre was accompanied by a lower Twb within the clothing throughout exercise, which likely resulted in the observed lower mean Tsk and HR for women in CLO. Women also experienced a lower sweating rate in CLO compared with HOT, likely driven by a lower Tre and related to the lower Twb within the clothing. The difference between sexes in Tre response is largely mediated by differences in work rate (absolute V˙O2), with possible contributions of a lower body mass and BSA rather than sex per se (40). Indeed, the women in this study had a lower absolute workload (secondary to a lower V˙O2max), and body mass, and higher BSA/mass ratios compared with male subjects. Recent research has highlighted the importance of expressing workload relative to body mass rather than BSA to equate core temperature rise during exercise in the heat (27,28). This research was done exclusively in male subjects in hot environments with large differences in body mass. In agreement with previous findings, the lower relative (W·kg−1) workload (Table 2) resulted in a slightly lower rise in Tb in women as compared with men in HOT (ΔTb 0.23°C lower in women; Table 2). However, this difference was magnified in CLO (ΔTb 0.71°C lower in women) despite equivalent workloads between environments for each individual. When viewing the rise in core temperature relative to the metabolic heat production normalized for body mass (W·kg−1·°C−1), near-perfect matching occurred in HOT, again agreeing with previous research equating core temperature rise during exercise in the heat when workload is expressed relative to body mass. By contrast, women in CLO had a significantly lower rise in Tre than men relative to metabolic heat production (W·kg−1), suggesting that workload relative to body mass could not entirely account for the differences in Tre observed in CLO (Table 2 and Fig. 4). Sex differences in thermoregulation related to sweating response (41,42) may also contribute in CLO, which relies on metabolic heat production and sweat to create a warm, humid microenvironment within the clothing. Therefore, the difference in response between HOT and CLO in women may be due to a lower workload coupled with a lower reliance on sweating.

Differences in sweating capacity between men and women become apparent at high fixed heat production rates (13,43). In HOT, men were likely able to offset a greater absolute metabolic heat production with a greater sweating capacity, resulting in similar heat gain between men and women. By contrast, in CLO, men maintained a higher absolute heat production, but the advantage of this sweating capacity was limited by overdressing, whereas in women, the lower Ereq due to net dry heat loss rather than gain (Table 2) may have been below the threshold at which sweating capacity becomes limited in women, thus resulting in less of an imbalance between Ereq and evaporative heat loss and lower heat gain. This difference appears to be advantageous for women in warm–humid environments where evaporative heat loss is limited, but not in hot–dry environments (42). As the microenvironment within the suit mimicked warm–humid conditions (Fig. 3), this difference may be the primary driving factor behind the lower Tre in women during CLO as compared with HOT. In addition, the higher sweating rate for men in CLO was largely trapped within the clothing (contributing to lower evaporative efficiency, Table 2), thus not providing cooling benefit and creating conditions more favorable for heat storage. In CLO, the combination of a lower Twb within the clothing, a lower sweating rate, a lower mean Tsk, and a lower HR for women compared with men indicates that overdressing may not be an equally viable strategy in women, or in individuals exercising at a lower work rate. However, it is possible that more clothing or a higher workload for women in CLO would equalize these responses and elicit a similar heat strain stimulus in women. Future research should equate workload in watts per kilogram during overdressing protocols in well-trained men and women, to examine whether this potential sex difference is driven primarily by workload relative to body mass, or whether the lower sweating rates in women require greater compensation via workload or clothing ensemble. The challenge with equating heat production relative to body mass, based on our analysis of workload for “optimal Tre stimulus,” is that the heat production necessary for a 2°C increase in Tre (11.9 W·kg−1) represents 54% ± 3% of V˙O2max for the men in our study and 68% ± 4% of V˙O2max for the women. Given that V˙O2max is reduced during heat stress, women at a similar fitness level to those in this study would be unlikely to maintain this workload for 1 h whereas the men could complete work in this range with relative ease. By contrast, the heat production required for a 2°C increase in HOT (10.3 W·kg−1) represented 47% ± 2% and 59% ± 3% of V˙O2max for the men and women in this study, respectively, a range that is more feasible during exercise/heat stress.

Heat shock proteins, both intracellular and extracellular, have been observed to increase in response to exercise, acute heat stress, and heat acclimation. Although the role of HSPs in the acclimatory response and in cross-tolerance is still being elucidated (44), it is clear that acclimation is diminished if HSP production is inhibited (19). The magnitude of response appears varied, but may be related in part to fitness level, as the stressor may be less novel in well-trained individuals. The increase in eHSP72 observed in HOT was a similar magnitude as observed in previous research (21) during exercise-heat stress, but baseline values were lower in the current study. This may be due to well-trained endurance athletes exhibiting a partially heat acclimated profile by virtue of regular exercise-induced elevations in core temperature (45). Indeed, heat acclimation has been shown to lower baseline eHSP72 levels closer to those observed in the present study (21). Baseline eHSP72 levels in CLO were similar to HOT but did not increase after exercise-heat stress. This may be due to a lower peak Tre in CLO, although when individual responses were examined compared with peak Tre, Δ Tre, and fitness level, no clear patterns emerged. It is unclear whether the magnitude of change in eHSP72 observed in HOT is physiologically important, as previous studies have observed much larger increases in eHSP72 (25,46). Again, this may be due to the training status of the subject population, and the fact that exercise training confers some degree of heat acclimation (45). Although eight of nine subjects in HOT experienced an increase in eHSP72 from pre- to postexercise, few had large increases (n = 2; 97%–317% increase), whereas most experienced very small changes (n = 6; 8%–35% increase) or did not change (n = 1). By contrast, in CLO, five subjects experienced small increases (3%–37% increase), one stayed the same and four subjects had small decreases (3%–26% decrease). The variation in responses did not appear to be related to fitness level, ΔTre, peak Tre, relative exercise intensity, sweat losses, or order of study condition.

The physiological responses to overdressing observed in the present study suggest that the heat strain stimulus may be adequate to induce acclimation if repeated over time (5–14 d), but the lower change in Tre and more modest cellular responses in CLO suggest that the time course could be delayed or acclimation may be less robust as compared with classic hot/dry exercise–heat acclimation protocols. However, in previous work by Dawson et al. (6), similar improvements in heat tolerance were observed between overdressing during exercise and exercise in warm–humid conditions despite lower Tre during training in the overdressing group. The acclimation response to repeated overdressing has not been observed in women, and the greater difference in HR, mean Tsk, and Tre may combine to blunt the acclimatory response if repeated overdressing during exercise is used as an acclimation strategy. These findings have bearing for athletes seeking a competitive advantage in hot weather but may also affect athletes hoping to achieve the heat acclimation performance benefits that some have observed in other environments, including cool (3,8,38) and hypoxic (47,48) conditions. Although much debate remains regarding this potential performance benefit (49,50), the possibility of a competitive edge remains appealing for those at the highest levels of sport.

In summary, overdressing during exercise induces elevated Tre, Tsk, HR, and sweating responses; however, Tre increase is lower compared with exercise in a hot environment (traditional exercise–heat acclimation conditions). Overdressing appears to be a more favorable strategy in men compared with women, although it is possible that women (or smaller, less fit individuals) may be able compensate for this difference through a higher relative workload or additional clothing. In addition, the increase in eHSP72 observed in HOT was blunted or absent in CLO, suggesting that the acclimation response with repeated exposures could be blunted as well. These findings complement previous studies on overdressing in trained men and indicate that both workload (metabolic heat production) and clothing ensemble (heat retention) are important considerations in implementing overdressing as a heat acclimation strategy in trained men and women.

The authors thank Julio Gonzalez, Vienna Brunt, and Alex Woldt for their technical assistance. Funding was provided by the Kenneth and Kenda Singer Endowment. The authors do not have any conflicts of interest, and all results are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by the American College of Sports Medicine.

Back to Top | Article Outline


1. Racinais S, Alonso JM, Coutts AJ, Flouris AD, Girard O, González-Alonso J, et al. Consensus recommendations on training and competing in the heat. Br J Sports Med. 2015;49(18):1164–73.
2. Zurawlew MJ, Walsh NP, Fortes MB, Potter C, Walsh NP. Post-exercise hot water immersion induces heat acclimation and improves endurance exercise performance in the heat. Scand J Med Sci Sports. 2016;26:745–54.
3. Scoon GS, Hopkins WG, Mayhew S, Cotter JD. Effect of post-exercise sauna bathing on the endurance performance of competitive male runners. J Sci Med Sport. 2007;10(4):259–62.
4. Dawson B. Exercise training in sweat clothing in cool conditions to improve heat tolerance. Sports Med. 1994;17(4):233–44.
5. McLellan TM, Aoyagi Y. Heat strain in protective clothing following hot-wet or hot-dry heat acclimation. Can J Appl Physiol. 1996;21(2):90–108.
6. Dawson B, Pyke FS, Morton AR. Improvements in heat tolerance induced by interval running training in the heat and in sweat clothing in cool conditions. J Sports Sci. 1989;7(3):189–203.
7. Stevens CJ, Plews DJ, Laursen PB, Kittel AB, Taylor L. Acute physiological and perceptual responses to wearing additional clothing while cycling outdoors in a temperate environment: a practical method to increase the heat load. Temperature. 2017;4(4):414–9.
8. Lorenzo S, Halliwill JR, Sawka MN, Minson CT. Heat acclimation improves exercise performance. J Appl Physiol. 2010;109(4):1140–1147.
9. Fox RH. Heat stress and athletics. Ergonomics. 1960;3(4):307–13.
10. Maron MB, Wagner JA, Horvath SM. Thermoregulatory responses during competitive marathon running. J Appl Physiol. 1977;42(6):909–14.
11. Kenefick RW, Cheuvront SN, Sawka MN. Thermoregulatory function during the marathon. Sports Med. 2007;37(4–5):312–5.
12. Havenith G. Human surface to mass ratio and body core temperature in exercise heat stress—a concept revisited. J Therm Biol. 2001;26(4–5):387–93.
13. Gagnon D, Kenny GP. Sex modulates whole-body sudomotor thermosensitivity during exercise. J Physiol. 2011;589(24):6205–17.
14. Fox RH, Goldsmith R, Kidd DJ, Lewis HE. Acclimatization to heat in man by controlled elevation of body temperature. J Physiol. 1963;166(3):530–47.
15. Regan JM, MacFarlane DJ, Taylor NA. An evaluation of the role of skin temperature during heat adaptation. Acta Physiol Scand. 1996;158(4):365–75.
16. Buono MJ, Numan TR, Claros RM, Brodine SK, Kolkhorst FW. Is active sweating during heat acclimation required for improvements in peripheral sweat gland function? Am J Physiol Regul Integr Comp Physiol. 2009;297(4):R1082–5.
17. Gonzalez RR, Berglund LG, Gagge AP. Indices of thermoregulatory strain for moderate exercise in the heat. J Appl Physiol. 1978;44(6):889–99.
18. Horowitz M, Assadi H. Heat acclimation-mediated cross-tolerance in cardioprotection: do HSP70 and HIF-1alpha play a role? Ann N Y Acad Sci. 2010;1188:199–206.
19. Kuennen M, Gillum T, Dokladny K, Bedrick E, Schneider S, Moseley P. Thermotolerance and heat acclimation may share a common mechanism in humans. Am J Physiol Regul Integr Comp Physiol. 2011;301(2):R524–33.
20. Sandström ME, Madden LA, Taylor L, Siegler JC, Lovell RJ, Midgley A, et al. Variation in basal heat shock protein 70 is correlated to core temperature in human subjects. Amino Acids. 2009;37:279–84.
21. Magalhães F, de C, Amorim FT, Passos RL, Fonseca MA, Oliveira KP, Lima MR, et al. Heat and exercise acclimation increases intracellular levels of Hsp72 and inhibits exercise-induced increase in intracellular and plasma Hsp72 in humans. Cell Stress Chaperones. 2010;15(6):885–95.
22. Marshall HC, Ferguson RA, Nimmo MA. Human resting extracellular heat shock protein 72 concentration decreases during the initial adaptation to exercise in a hot, humid environment. Cell Stress Chaperones. 2006;11(2):129–34.
23. Whitham M, Fortes MB. Heat shock protein 72: release and biological significance during exercise. Front Biosci. 2008;13:1328–39.
24. Yamada PM, Amorim FT, Moseley P, Robergs R, Schneider SM. Effect of heat acclimation on heat shock protein 72 and interleukin-10 in humans. J Appl Physiol. 2007;103(4):1196–204.
25. Garrett AT, Goosens NG, Rehrer NJ, Patterson MJ, Harrison J, Sammut I, et al. Short-term heat acclimation is effective and may be enhanced rather than impaired by dehydration. Am J Hum Biol. 2014;26:311–20.
26. McClung JP, Hasday JD, He J, Montain SJ, Cheuvront SN, Sawka MN, et al. Exercise–heat acclimation in humans alters baseline levels and ex vivo heat inducibility of HSP72 and HSP90 in peripheral blood mononuclear cells. Am J Physiol Regul Integr Comp Physiol. 2008;294(1):R185–91.
27. Cramer MN, Jay O. Selecting the correct exercise intensity for unbiased comparisons of thermoregulatory responses between groups of different mass and surface area. J Appl Physiol. 2014;116(9):1123–32.
28. Jay O, Bain AR, Deren TM, Sacheli M, Cramer MN. Large differences in peak oxygen uptake do not independently alter changes in core temperature and sweating during exercise. Am J Physiol Regul Integr Comp Physiol. 2011;301(3):R832–41.
29. Ramanathan NL. A new weighting system for mean surface temperature of the human body. J Appl Physiol. 1964;19:531–3.
30. O’Brien C, Blanchard LA, Cadarette BS, Endrusick TL, Xu X, Berglund LG, Sawka MN, Hoyt RW. Methods of evaluating protective clothing relative to heat and cold stress: thermal manikin, biomedical modeling, and human testing. J Occup Environ Hyg. 2011;8(10):588–99.
31. Potter AW, Blanchard LA, Friedl KE, Cadarette BS, Hoyt RW. Mathematical prediction of core body temperature from environment, activity, and clothing: the Heat Strain Decision Aid (HSDA). J Therm Biol. 2017;64:78–85.
32. Gagge AP, Gonzalez RR. Mechanisms of heat exchange: biophysics and physiology. In: Comprehensive Physiology. John Wiley & Sons, Inc: Hoboken (NJ); 2011.
33. Ely BR, Ely MR, Cheuvront SN, Kenefick RW, Degroot DW, Montain SJ. Evidence against a 40 degrees C core temperature threshold for fatigue in humans. J Appl Physiol. 2009;107(5):1519–25.
34. Snellen JW. Mean body temperature and the control of thermal sweating. Acta Physiol Pharmacol Neerl. 1966;14:99–174.
35. Havenith G, Richards MG, Wang X, Bröde P, Candas V, den Hartog E, et al. Apparent latent heat of evaporation from clothing: attenuation and “heat pipe” effects. J Appl Physiol. 2008;104(1):142–9.
36. Gagnon D, Jay O, Kenny GP. The evaporative requirement for heat balance determines whole-body sweat rate during exercise under conditions permitting full evaporation. J Physiol. 2013;591(11):2925–35.
37. Keiser S, Flück D, Hüppin F, Stravs A, Hilty MP, Lundby C. Heat training increases exercise capacity in hot but not in temperate conditions: a mechanistic counter-balanced cross-over study. Am J Physiol Heart Circ Physiol. 2015;309(5):H750–61.
38. Neal RA, Massey HC, Tipton MJ, Young JS, Corbett J. Effect of permissive dehydration on induction and decay of heat acclimation, and temperate exercise performance. Front Physiol. 2016;7:564.
39. Gibson OR, Mee JA, Tuttle JA, Taylor L, Watt PW, Maxwell NS. Isothermic and fixed intensity heat acclimation methods induce similar heat adaptation following short and long-term timescales. J Therm Biol. 2015;49–50:55–65.
40. Havenith G, Coenen JM, Kistemaker L, Kenney WL. Relevance of individual characteristics for human heat stress response is dependent on exercise intensity and climate type. Eur J Appl Physiol. 1998;77:231–41.
41. Kenney WL. A review of comparative responses of men and women to heat stress. Environ Res. 1985;37(1):1–11.
42. Shapiro Y, Pandolf KB, Avellini BA, Pimental NA, Goldman RF. Physiological responses of men and women to humid and dry heat. J Appl Physiol. 1980;49(1):1–8.
43. Gagnon D, Kenny GP. Sex differences in thermoeffector responses during exercise at fixed requirements for heat loss. J Appl Physiol (1985). 2012;113(5):746–57.
44. Amorim FT, Fonseca IT, Machado-Moreira CA, Magalhães F, de C. Insights into the role of heat shock protein 72 to whole-body heat acclimation in humans. Temperature (Austin). 2015;2(4):499–505.
45. Gisolfi CV, Cohen JS. Relationships among training, heat acclimation, and heat tolerance in men and women: the controversy revisited. Med Sci Sports Exerc. 1979;11(1):56–9.
46. Gibson OR, Dennis A, Parfitt T, Taylor L, Watt PW, Maxwell NS. Extracellular Hsp72 concentration relates to a minimum endogenous criteria during acute exercise–heat exposure. Cell Stress Chaperones. 2014;19(3):389–400.
47. Heled Y, Peled A, Yanovich R, Shargal E, Pilz-Burstein R, Epstein Y, et al. Heat acclimation and performance in hypoxic conditions. Aviat Space Environ Med. 2012;83(7):649–53.
48. White AC, Salgado RM, Astorino TA, Loeppky JA, Schneider SM, McCormick JJ, et al. The effect of 10 days of heat acclimation on exercise performance in acute hypobaric hypoxia (4350 m). Temperature (Austin). 2016;3(1):176–85.
49. Minson CT, Cotter JD. CrossTalk proposal: heat acclimatization does improve performance in a cool condition. J Physiol. 2016;594(2):241–3.
50. Nybo L, Lundby C. CrossTalk opposing view: heat acclimatization does not improve exercise performance in a cool condition. J Physiol. 2016;594(2):245–7.


Supplemental Digital Content

Back to Top | Article Outline
© 2018 American College of Sports Medicine