Heat exposure limits for young unacclimatized males and females at low and high humidity

Abstract

Little is known about the separate and combined influences of humidity conditions, sex, and aerobic fitness on heat tolerance in unacclimatized males and females. The purpose of the current study was to describe heat tolerance, in terms of critical WBGT (WBGT_crit), in unacclimatized young males and females in hot-dry (HD) and warm-humid (WH) environments. Eighteen subjects (9 M/9F; 21 ± 2 yr) were tested during exercise at 30% ${\dot{\text{V}}\text{O}}_{2\max }$O_2max in a controlled environmental chamber. Progressive heat stress exposures were performed with either (1) constant dry-bulb temperature (T_db) of 34 and 36°C and increasing ambient water vapor pressure (P_a) (P_crit trials; WH); or (2) constant P_a of 12 and 16mmHg and increasing T_db (T_crit trials; HD). Chamber T_db and P_a, and subject esophageal temperature (T_es), were continuously monitored throughout each trial. After a 30-min equilibration period, progressive heat stress continued until subject heat balance could no longer be maintained and a clear rise in T_es was observed. Absolute WBGT_crit and WBGT_crit adjusted to a metabolic rate of 300W (WBGT₃₀₀), and the difference between WBGT_crit and occupational exposure limits (OEL; ΔOEL) was assessed. WBGT_crit, WBGT₃₀₀, and ΔOEL were higher in WH compared to HD (p < 0.0001) for females but were the same between environments for males (p ≥ 0.21). WBGT_crit was higher in females compared to males in WH (p < 0.0001) but was similar between sexes in HD (p = 0.44). When controlling for metabolic rate, WBGT₃₀₀ and ΔOEL were higher in males compared to females in WH and HD (both p < 0.0001). When controlling for sex, ${\dot{\text{V}}\text{O}}_{2\max }$ was not associated with WBGT₃₀₀ or ΔOEL for either sex (r ≤ 0.12, p ≥ 0.49). These findings suggest that WBGT_crit is higher in females compared to males in WH, but not HD, conditions. Additionally, the WBGT_crit is lower in females, but not males, in HD compared to WH conditions.

Introduction

Among unacclimatized people, sex and aerobic fitness may influence occupational heat tolerance. On average, females have a larger surface area-to-mass ratio than males, which leads to greater heat dissipation while exercising (Yanovich et al. 2020). Furthermore, females have greater active sweat gland concentrations compared to males, although maximal output per sweat gland and overall sweat rates are lower in females (Gagnon et al. 2013). Accordingly, females have a lower whole-body sweat rate compared to males when near-maximal sweating is required (e.g., high exercise intensity and/or ambient temperature) (Morimoto et al. 1967; Gagnon et al. 2013; Yanovich et al. 2020). Additionally, females typically have a lower maximal aerobic capacity (${\dot{\text{V}}\text{O}}_{2\max }$) compared to males (Sparling and Cureton 1983; Cheuvront et al. 2005; Loe et al. 2013). Higher ${\dot{\text{V}}\text{O}}_{2\max }$ is associated with improved thermoregulatory function, including higher sweat rates and improved cutaneous vasodilation (Roberts et al. 1977; Ho et al. 1997; Thomas et al. 1999; Okazaki et al. 2002; Best et al. 2012; Alhadad et al. 2019). Because females generally have a lower ${\dot{\text{V}}\text{O}}_{2\max }$, they are at a higher percentage of their ${\dot{\text{V}}\text{O}}_{2\max }$ when working at the same absolute work intensity as males. Conversely, when working at the same percentage of ${\dot{\text{V}}\text{O}}_{2\max }$, females are typically working at a lower absolute intensity, and therefore a lower metabolic heat production (Gagnon et al. 2008), compared to males. Although sex- and aerobic-fitness-related differences in metabolic heat production and thermoregulatory responses to heat stress are well established, little is known about the separate and combined influences of sex and aerobic fitness on occupational heat tolerance limits.

A progressively intensifying heat stress protocol can be used to assess heat stress limits by finding the transition point between compensable and uncompensable heat stress. Specifically, this protocol can be used to identify the critical WBGT (WBGT_crit); that is, the WBGT above which heat balance cannot be maintained in humans for a given metabolic rate and clothing configuration (Garzón-Villalba et al. 2017). The WBGT_crit is relatively constant for acclimatized participants wearing woven clothing between 20 and 70% relative humidity (Kenney et al. 1988; Bernard et al. 2005). However, WBGT may not adequately reflect the environmental contribution to heat stress at very high and very low humidity (Budd 2008). The deviation has been demonstrated in studies using the progressive heat stress protocol at critical conditions above 70% and below 20% relative humidity (Kenney et al. 1988). Under conditions of high humidity, evaporative cooling is limited by a small water vapor pressure gradient between the wetted skin and the environment. For low humidity conditions, evaporative cooling is limited by the individual’s capacity to produce sweat, and if ambient temperature is high enough, increased dry heat gain.

The commonly accepted approach for evaluation of occupational heat stress is a criterion wet bulb globe temperature (WBGT) that is adjusted for metabolic rate as described by ACGIH^® (2021), NIOSH (2016), and ISO (2017). The criterion level is set so that most acclimatized, healthy, hydrated workers can sustain the exposure over prolonged times (Lind 1963, 1970; Dukes-Dobos and Henschel 1973; Garzón-Villalba et al. 2017). The occupational heat stress exposure assessments make a distinction between unacclimatized and acclimatized workers (Dukes-Dobos and Henschel 1973; NIOSH 2016; ISO 2017; ACGIH 2021) that recognizes the increased tolerance to heat stress that accrues with acclimatization. Also recognizing the value of acclimatization, OSHA recommends having a structured acclimatization program (OSHA 2017) and notes the absence of acclimatization in citations (Tustin et al. 2018).

The current research was undertaken to investigate the WBGT_crit in a cohort of unacclimatized young males and females working at 30% ${\dot{\text{V}}\text{O}}_{2\max }$ in low- and high-humidity environments. A secondary aim was to examine the WBGT_crit adjusted to a standard, absolute metabolic rate. Last, we aimed to determine whether there is a relation between ${\dot{\text{V}}\text{O}}_{2\max }$ and metabolic rate-adjusted WBGT_crit. We hypothesized that (1) the WBGT_crit would differ between WH and HD environments, (2) WBGT_crit would be higher in females compared to males at a relative work rate of 30% ${\dot{\text{V}}\text{O}}_{2\max }$, but not when adjusted to a standard metabolic rate, and (3) ${\dot{\text{V}}\text{O}}_{2\max }$ would not be related to metabolic rate-adjusted WBGT_crit.

Methods

Subjects

All experimental protocols were approved by the Institutional Review Board at The Pennsylvania State University and conformed to the guidelines set forth by the Declaration of Helsinki. After all aspects of the experiment were explained, oral and written informed consent was obtained.

Eighteen healthy males and females (nine males and nine females) aged 18 to 30 yr were included. All subjects were healthy, normotensive, nonsmokers, and not taking any medications that might affect the physiological variables of interest in this study. Although menstrual status influences basal body temperature, the progressive heat stress protocol used herein relies on the biophysics of heat exchange to determine the point at which uncompensable heat stress begins to occur. Because there is little change in thermoregulatory capacity at different points of the menstrual cycle (Nunneley 1978), no attempt was made to control for menstrual status or contraceptive use (Kenney and Zeman 2002; Wolf et al. 2021). None of the subjects were physically active outdoors on a regular basis. Height and weight were measured with a stadiometer (Detecto, Webb City, MO, USA) and scale accurate to ±50 g (Seca, Hamburg, Germany), respectively. ${\dot{\text{V}}\text{O}}_{2\max }$ was determined with the use of open-circuit spirometry (Parvo Medics, Salt Lake City, UT, USA) during a maximal graded exercise test (Bruce et al. 1973) performed on a motor-driven treadmill. Subjects with a ${\dot{\text{V}}\text{O}}_{2\max }$ in the upper or lower 20th percentile for their sex and age (Loe et al. 2013; Liguori and American College of Sports Medicine 2020) were excluded. During the experiments, clothing was standardized with subjects wearing thin, short-sleeved cotton tee-shirts, shorts, socks, and walking/running shoes.

The ACGIH threshold limit values (TLV^®) and ISO7342 make adjustments in the observed WBGT to account for clothing that may have a different evaporative resistance than standard woven work uniforms. Based on a study that compared seminude to work clothes (Belding and Kamon 1973), the difference in WBGT was 1.8 °C. With the addition of a tee-shirt to the seminude, the clothing adjustment value was taken as −1.0 °C.

Testing procedures

Upon arrival at the laboratory, participants provided a urine sample to ensure euhydration, defined as urine specific gravity ≤1.020 (USG; PAL-S, Atago, Bellevue, WA, USA) (Kenefick and Cheuvront 2012). Two sets of protocols (four experimental visits) were used to determine either (1) the critical P_a (P_crit) for the upward inflection of esophageal temperature (T_es) at two distinct T_db values (34 and 36 °C; warm-humid conditions, WH) or (2) the critical T_db (T_crit) at two distinct P_a values (12 and 16 mmHg; hot-dry conditions, HD). Each subject completed all four conditions on separate days, separated by at least 72 hr, in random order. During each experiment, the subjects walked continuously on a motor-driven treadmill at 30% ${\dot{\text{V}}\text{O}}_{2\max }$ until a clear rise in T_es was observed. After a 30-min equilibration period, either the T_db (T_crit tests) or the P_a (P_crit tests) in the chamber was increased in a stepwise fashion (1 °C or 1 mmHg every 5 min) while chamber T_db and T_wb and subject T_es, skin temperature (T_sk), and HR were monitored. To confirm that each subject was walking at the prescribed workload, ${\dot{\text{V}}\text{O}}_{2\max }$ was determined at a single time point during 30–40 min. Walking speeds ranged from 64 to 80 m/min for females and 75 to 91 m/min for males, while treadmill gradient ranged from 0 to 2% for males and females. With no forced air movement in the environmental chambers, air movement velocity was measured at ≤0.45 m/sec (Kenney and Zeman 2002).

Measurements

Esophageal temperature (T_es) was measured with a probe made from a thermistor sealed in a pediatric feeding tube. The probe was inserted nasally and lowered in the esophagus to the level of the left atrium, ~0.25 of the subject’s standing height.

Metabolic rate [M; Watts (W)], normalized to body surface area, was calculated from oxygen consumption (${\dot{\text{V}}\text{O}}_2$; L/min) and the respiratory exchange ratio (RER; unitless), determined using indirect calorimetry (Cramer and Jay 2019), as

$$M={\dot{\text{V}}\text{O}}_{\text{2}}\cdot \frac{\left[\left(\left(\frac{\text{RER}-0.7}{0.3}\right)\cdot 21.13\right)+\left(\left(\frac{1.0-\text{RER}}{0.3}\right)\cdot 19.62\right)\right]}{60}\cdot 1000$$ (1)

Nude body mass was measured immediately before and after each experiment and sweat rate (L·h⁻¹) was determined from the loss of nude body mass. Fluid intake was prohibited between the initial and final measurements of nude body mass. Body surface area (A_D; m⁻²) was calculated using the standard DuBois equation (DuBois 1916), and sweat rates were converted and presented as relative to body surface area (mL·m⁻²·h⁻¹).

Determination of WBGT_crit

An initial rise in T_es was observed that typically began to plateau after 30–40 min and remained at an elevated steady state as T_db or P_a was systematically increased. The critical T_db or P_a was characterized by the upward inflection of T_es from the elevated steady state, which was selected graphically from the raw data. A line was drawn between the data points, starting at the 30^th min. A second line was drawn from the point of departure from the T_es equilibrium phase slope. The T_db or P_a at 1 min before the point at which the second line departed from the first was defined as the critical T_db or P_a, respectively. Inflection points were chosen by two independent investigators blinded to the condition, group, and participant. The value included in the analysis was the average of the values determined by the two investigators. A subgroup analysis (n = 44) was performed using this method and segmental linear regression, which demonstrated excellent agreement (ICC = 0.99) between the two methods, suggesting that either is a valid approach for the determination of inflection points. An inter-rater reliability (ICC) of 0.93 for the determination of the T_es inflection point has been previously reported (Wolf et al. 2021).

Psychrometric wet bulb temperature (T_pwb) at the T_es inflection point was determined using a standard psychrometric chart for critical T_db and P_a experiments. Where necessary, T_pwb was converted to natural wet bulb temperature (T_nwb) as (Alfano et al. 2012)

$$T_{nwb}=\frac{0.16\left({\text{T}}_{\text{g}}-{\text{T}}_{\text{db}}\right)+0.8}{200}\ast \left(560-2\text{RH}-5{\text{T}}_{\text{a}}\right)-0.8+{\text{T}}_{\text{pwb}}$$ (2)

where T_g is globe temperature, which was equal to T_db for the current paper. The WBGT at the T_es inflection point (i.e., the WBGT_crit) was calculated with the equation for indoor WBGT provided in ISO7243 (ISO 2017):

$$WBGT=0.7\cdot {\text{T}}_{\text{wb}}+0.3\cdot {\text{T}}_{\text{g}}$$ (3)

where T_nwb and T_db were used for T_wb and T_g, respectively.

Adjustments to WBGT_crit for metabolic rate

The experimental design called for metabolic demands based on the percent of aerobic capacity to adjust for aerobic fitness among participants and especially between the sexes. Heat stress assessments examine the overall heat stress level, which considers the absolute metabolic rate regardless of personal factors such as sex, body weight, and aerobic fitness. To account for systematic differences in metabolic rate, WBGT_crit was adjusted using two methods.

First, a slope of WBGT_crit vs. M was used. Garzón-Villalba et al. (2017) reported a slope of −0.0201 °C per W for a linear relation between WBGT_crit and M (see that paper’s appendix). A metabolic rate of 300W (moderate metabolic rate (ACGIH 2021)) was selected as the reference metabolic rate, so that

$${\text{WBGT}}_{300}={\text{WBGT}}_{\text{crit}}+0.0201\ast (\text{M}-300)$$ (4)

An alternative method to adjust WBGT_crit was used by Garzón-Villalba et al. (2017) to explore the performance of the WBGT-based exposure limits. That study compared observed WBGT_crit values to occupational exposure limits (OEL) defined by the ACGIH TLV and NIOSH recommended exposure limit (REL). Following her approach and adjusting for clothing (seminude CAV of −1 °C), the observed exposure above the OEL (ΔOEL) is

$$\text{ΔOEL}={\text{WBGT}}_{\text{crit}}-1-\left[56.7-11.5\ast {\log }_{10}(\text{M})\right]$$ (5)

Exposure response curves

To develop an exposure-response curve, the ΔOEL data were rank ordered from the lowest to highest. The probability (p) for each observation was the rank (i, starting at 1) divided by the number of observations (n) plus 1; that is

$$p_i=\frac{\text{i}}{\text{n}+1}$$ (6)

The odds for each observation was

$${\mathit{odds}}_i=\frac{{\text{p}}_{\text{i}}}{1-{\text{p}}_{\text{i}}}$$ (7)

Then the logistic regression was estimated as

$$\mathit{ln}(\mathit{odds})=\text{a}+\text{b}\text{ΔOEL}$$ (8)

where the intercept (a) and slope (b) were determined by a linear regression. Finally, the data were plotted as p vs. ΔOEL and the descriptive line was

$$p=\frac{{\text{e}}^{(\text{a}+\text{b}\text{ΔOEL})}}{1+{\text{e}}^{(\text{a}+\text{b}\text{ΔOEL})}}$$ (9)

Statistical analysis

Student’s unpaired t tests were used to compare participant characteristics. WBGT_crit, WBGT₃₀₀, ΔOEL, M, and SR were analyzed using SAS PROC MIXED (SAS, version 9.4, SAS Institute Inc., Cary, NC, USA) two-way, repeated-measures ANOVA to evaluate group (males vs. females) and condition (HD vs. WH) effects. Hedges’ g effect sizes were calculated and reported for statistically significant pairwise comparisons (small effect = 0.2; medium effect = 0.5; large effect = 0.8). Ordinary least-squares regressions were used to examine the relation between ${\dot{\text{V}}\text{O}}_{2\max }$ and WBGT₃₀₀ and ΔOEL (GraphPad Prism, GraphPad Software Inc., San Diego, CA, USA). Data are reported as mean ± standard deviation, except in Figures 1 and 2, which are presented as box-and-whisker plots with individual data points. Significance was accepted at α = 0.05.

Figure 1.

Critical WBGT limits above which equilibrium in core temperature cannot be maintained during work at 30% ${\dot{\text{V}}\text{O}}_{2\max }$ for females (open boxes) and males (gray boxes) in warm/humid and hot/dry environments. Open circles represent individual data points. Boxes represent first and third quartiles with median values denoted by the horizontal line, and whiskers indicate minimum and maximum observations. §p < 0.05 compared to females; *p < 0.05 compared to warm/humid.

Figure 2.

Critical WBGT adjusted to a work rate of 300 W (WBGT₃₀₀; Panel A) and the difference between empirically derived critical WBGT limits at 30% ${\dot{\text{V}}\text{O}}_{2\max }$ and occupational exposure limits (ΔOEL; Panel B) for females (open boxes) and males (gray boxes) in warm/humid and hot/dry environments. Open circles represent individual data points. Boxes represent first and third quartiles with median values denoted by the horizontal line, and whiskers indicate minimum and maximum observations. § p < 0.05 compared to females; *p < 0.05 compared to warm/humid.

Results

Subject characteristics

The males and females in this study were representative of the general population in this age range with respect to anthropometric characteristics and aerobic capacity (Table 1) (Kaminsky et al. 2015). The males were taller, weighed more, and had a higher mean ${\dot{\text{V}}\text{O}}_{2\max }$ and body surface area (all p < 0.05); body surface area-to-mass ratio was higher in females compared to males (p < 0.0001).

Table 1.

Subject characteristics.

Characteristic	Females (n = 9)	Males (n = 9)
Age (yr)	20 ± 2	21 ± 3
Height (m)	1.63 ± 0.1	1.85 ± 0.1§
Weight (kg)	58 ± 8	89 ± 2§
AD (m2)	1.61 ± 0.11	2.12 ± 0.12§
AD/wt (m2/kg)	0.028 ± 0.002	0.024 ± 0.001§
${\dot{\text{V}}\text{O}}_{2\max }$ (ml·kg−1·min−1)	39 ± 4	45 ± 2§

^§p < 0.05 compared to females. A_D, body surface area; A_D/wt, body surface area-to-mass ratio; ${\dot{\text{V}}\text{O}}_{2\max }$, maximal aerobic capacity. Sex comparisons were performed using Student’s unpaired t tests.

Metabolic rate and sweat rate

Table 2 presents M and SR data for males and females in WH and HD conditions. M was lower for females compared to males in WH (p < 0.0001; g = 3.62) and HD (p < 0.0001; g = 4.67). Absolute sweat rates (L·h⁻¹) were lower for females compared to males in WH (p < 0.0001; g = 2.69) and HD (p < 0.0001; g = 2.09). Similarly, sweat rates relative to A_D (mL·m⁻²·h⁻¹) were lower for females compared to males in WH (p < 0.0001; g = 2.69) and HD (p < 0.0001; g = 2.09). There were no differences in M or SR between environmental conditions for females (M: p = 0.99; SR: p = 0.49) or males (M: p = 0.99; SR: p = 0.20).

Table 2.

Metabolic rates and sweat rates for females and males in warm-humid and hot-dry conditions (mean ± SD).

	Females	Males
Metabolic rate (W)
Warm-humid	218 ± 32	399 ± 63§
Hot-dry	221 ± 33	395 ± 41§
Sweat rate (L·h−1)
Warm-humid	0.42 ± 0.15	0.69 ± 0.11§
Hot-dry	0.40 ± 0.09	0.67 ± 0.11§
Sweat rate (mL·m−2·h−1)
Warm-humid	159 ± 53	204 ± 38§
Hot-dry	151 ± 28	188 ± 41§

^§p < 0.05 compared to females. Data were analyzed using two-way, repeated-measures ANOVA.

Critical WBGT

Figure 1 depicts the WBGT_crit for males and females in WH and HD conditions. There were significant main effects for sex (p < 0.0001) and condition (p < 0.0001) and a significant interaction (p = 0.004). The interaction was driven by females in the WH condition; that is, WBGT_crit was higher for females compared to males in WH (32.5 ±0.9 °C vs. 31.0 ±0.8; p < 0.0001; g = 1.76) but was similar between sexes in HD (31.4 ±0.6 °C vs. 30.9 ±1.0; p = 0.44). Further, the WBGT_crit was lower for females in HD compared to WH (p < 0.0001; g = 1.43), whereas there was no difference in WBGT_crit between conditions for males (p = 0.21).

WBGT₃₀₀ and ΔOEL

Figure 2 presents WBGT₃₀₀ and ΔOEL for males and females in WH and HD conditions. For WBGT₃₀₀, there were significant main effects of sex (p < 0.0001) and condition (p = 0.04), and an interaction effect (p = 0.04). Similar to the unadjusted WBGT_crit, the WBGT₃₀₀ was lower for females in HD (29.7 ± 1.0 °C) compared to WH (30.9 ± 1.0 °C; p = 0.02; g = 1.2), but was not different between conditions for males (HD: 33.3 ± 1.5, WH: 33.3 ± 1.1; p = 0.99). In contrast to the unadjusted WBGT_crit, however, WBGT₃₀₀ was higher for males compared to females in HD (p < 0.0001; g = 2.82) and WH (p < 0.0001; g = 2.28).

The ΔOEL outcomes were similar to WBGT₃₀₀, with significant main effects of sex (p < 0.0001) and condition (p = 0.02), and an interaction effect (p = 0.03). The ΔOEL was lower for females in HD (0.5 ± 1.1 °C) compared to WH (1.7 ± 1.1 °C; p = 0.01; g = 1.09), but was not different between conditions for males (HD: 3.3 ± 1.2, WH: 3.4±0.9; p = 0.99). Similar to WBGT₃₀₀, the ΔOEL was higher for males compared to females in HD (p < 0.0001; g = 2.43) and WH (p < 0.001; g = 1.69).

Associations between ${\dot{\text{V}}\text{O}}_{2\max }$ and WBGT₃₀₀ and ΔOEL

When including all participants, WBGT₃₀₀ (r = 0.56, p < 0.0001) and ΔOEL (r = 0.51, p < 0.0001) were positively related to ${\dot{\text{V}}\text{O}}_{2\max }$, suggesting that aerobic capacity may influence WBGT thresholds at a standardized metabolic rate (Figure 3). However, when controlling for sex, ${\dot{\text{V}}\text{O}}_{2\max }$ was not associated with WBGT₃₀₀ or ΔOEL for either sex.

Figure 3.

Associations between (A) maximal aerobic capacity (${\dot{\text{V}}\text{O}}_{2\max }$) and critical WBGT adjusted to a work rate of 300W (WBGT₃₀₀) and (B) ${\dot{\text{V}}\text{O}}_{2\max }$ and the difference between empirically derived critical WBGT limits and occupational exposure limits (ΔOEL). When disaggregated by sex, there was no relation between ${\dot{\text{V}}\text{O}}_{2\max }$ and either WBGT₃₀₀ or ΔOEL.

Exposure response curves

Figure 4 is the exposure-response curve for the 18 participants under both conditions. Values at the bottom-left portion of the curve suggest a greater probability of uncompensable heat stress, whereas those at the top-right portion are more likely to be compensable. Females (depicted by open circles) tended to have a greater probability of uncompensable heat stress compared to males (depicted by closed circles).

Figure 4.

Exposure-response curve describing the probability of uncompensable heat stress against the difference between empirically-derived critical WBGT limits at 30% ${\dot{\text{V}}\text{O}}_{2\max }$ and occupational exposure limits (ΔOEL). Values at the bottom-left portion of the curve suggest a greater probability of uncompensable heat stress, whereas those at the top-right portion are more likely to be compensable. Females (open circles) tended to have a greater probability of uncompensable heat stress compared to men (closed circles).

Discussion

The findings of the present study demonstrate sex differences in occupational heat tolerance limits between young, unacclimatized males and females at 30% ${\dot{\text{V}}\text{O}}_{2\max }$ steady state work. The WBGT_crit was (1) higher in females compared to males in WH, but not HD, conditions, and (2) lower in females in HD compared to WH conditions, but not males.

When working at 30% ${\dot{\text{V}}\text{O}}_{2\max }$, metabolic heat production is significantly lower in females compared to males because females typically have a lower ${\dot{\text{V}}\text{O}}_{2\max }$ than males (Sparling and Cureton 1983; Cheuvront et al. 2005), resulting in lower absolute workloads when working at the same relative intensity. Lower metabolic heat production and constrained evaporative capacity in WH conditions resulted in a higher WBGT_crit for females compared to males in those humid environments. However, the WBGT_crit was significantly lower for females in HD compared to WH, and no sex difference was apparent in HD conditions. These disparate findings between WH and HD environments suggest that females have a thermoregulatory advantage in WH environments when working at a fixed relative intensity due to (1) lower metabolic heat production, (2) dry heat gain in HD environments, and, therefore, (3) a lower requirement for evaporative heat loss in WH compared to HD environments. Because females have lower sweating rates, on average, when the requirements for sweating are near maximal (Yanovich et al. 2020), they are at a relative disadvantage in HD compared to WH environments.

The 30% ${\dot{\text{V}}\text{O}}_{2\max }$ exercise intensity was chosen for this study because it approximates the intensity of many self-paced recreational activities and is the average intensity associated with an 8-hr work day in many industrial settings (Bonjer 1962). However, heat stress assessments—and occupational safety standards and guidance—examine overall heat stress as a function of absolute metabolic rate. Because the WBGT_crit was confounded by metabolic rate (Bernard et al. 2008), it was difficult to parse out the differences due to environment and sex alone. To normalize WBGT_crit, two methods were used to account for the systematic differences in absolute metabolic rate between females and males.

Adjusting WBGT_crit to a standardized metabolic rate of 300W (WBGT₃₀₀) employed a linear relationship to proportionately increase WBGT_crit for metabolic rates above 300W and to proportionately decrease it for metabolic rates less than 300 W. In this way, individual values of WBGT_crit could be compared without the influence of metabolic rate. An alternative approach to normalizing the WBGT_crit to account for differences in metabolic rate was to compute the difference from the OEL (ΔOEL), which changes with metabolic rate. Doing this does not directly adjust the WBGT_crit, but references the WBGT_crit at lower metabolic rates (and thus higher WBGT_crit) to a higher value for OEL. The main effects and interaction of sex and condition remained in the analysis of WBGT₃₀₀ and ΔOEL. WBGT₃₀₀ was (1) lower in females compared to males in both WH and HD conditions, and (2) lower in HD compared to WH conditions for females, but not males. The lower values for WBGT₃₀₀ and ΔOEL for females in both conditions demonstrated a lower heat tolerance compared to males after adjusting for metabolic rate. The interaction effect was driven by females who had a lower WBGT₃₀₀ and ΔOEL in HD conditions, which suggested a lower capacity to dissipate heat in HD conditions than WH conditions.

These findings reflect recently published data demonstrating that when metabolic rates were matched in unacclimatized males and females during a progressive heat stress protocol at 38 ° C, WBGT_crit was lower in females compared to males (Wolf et al. 2021). By definition, sweat rates and/or evaporation of sweat are maximal at the WBGT_crit, above which heat dissipation via sweating and dry heat exchange are insufficient to compensate for the heat load (i.e., heat stress becomes uncompensable). Because maximal sweating rates are lower in females compared to males (Morimoto et al. 1967; Gagnon et al. 2013; Yanovich et al. 2020), it is unsurprising that the WBGT_crit would be lower in females when controlling for metabolic rate, as with the WBGT₃₀₀ and ΔOEL. Further, there did not appear to be any advantage to normalizing WBGT_crit using WBGT₃₀₀ vs. ΔOEL, or vice versa.

Higher ${\dot{\text{V}}\text{O}}_{2\max }$ is associated with improved thermoregulatory function, including higher sweat rates and cutaneous vasodilation (Roberts et al. 1977; Ho et al. 1997; Thomas et al. 1999; Okazaki et al. 2002; Best et al. 2012; Alhadad et al. 2019). Because of the association between ${\dot{\text{V}}\text{O}}_{2\max }$ and thermoregulatory function, we examined the relations between ${\dot{\text{V}}\text{O}}_{2\max }$ and WBGT₃₀₀ and between ${\dot{\text{V}}\text{O}}_{2\max }$ and ΔOEL in the present data. We found that there were, indeed, positive associations between ${\dot{\text{V}}\text{O}}_{2\max }$ and both WBGT₃₀₀ and ΔOEL when the two sexes were aggregated. However, when the analysis was performed with the sexes disaggregated, that association was not apparent within either sex. Within each sex, fitness reflected by ${\dot{\text{V}}\text{O}}_{2\max }$ did not predict either the WBGT₃₀₀ or ΔOEL, suggesting that sex differences in heat tolerance were independent of aerobic fitness.

Finally, an exposure-response curve for unacclimatized males and females with metabolic rates at 30% ${\dot{\text{V}}\text{O}}_{2\max }$ based on ΔOEL was created. Interestingly, the curve suggested that there was about a 5% probability of an uncompensable exposure at the OEL for this unacclimatized population, although the probability was relatively higher for females than for males. Comparing the 50% probability point at ΔOEL = 2.5 °C-WBGT with that for acclimatized participants at ΔOEL = 6 °C-WBGT (Garzón-Villalba et al. 2017) suggests that the curve shifts to the right by ~3.5 °C-WBGT due to acclimatization. This is near the reported benefit associated with acclimatization of 2.5 to 3 °C-WBGT (Dukes-Dobos and Henschel 1973; Garzón-Villalba et al. 2017).

Occupational heat stress criteria, as a function of wet bulb globe temperature (WBGT) and metabolic heat production, are set by organizations such as the ACGIH (2021), NIOSH (2016), and ISO (2017). Those criteria are made with reference to the “standard man,” considered to be a representative human with a body mass of 70 kg and body surface area of 1.8 m², assuming no significant differences between sexes. Although the overall probability of uncompensable exposure at the OEL was ~5%, the data presented herein suggest that, when matching for metabolic heat production, females have a higher probability of uncompensable heat stress at the OELs defined by the ACGIH TLV and NIOSH REL. As such, further investigation may be warranted to determine whether the currently-accepted OELs are appropriate for both sexes. Furthermore, the current investigation included only HD and WH environments; thus, it is unclear how these data may extend to hot-humid environments.

In this study we chose not to control for menstrual cycle, with the assumption that hormonal fluctuations across the cycle would not meaningfully alter the biophysics of heat exchange that dictate the point of inflection during progressive heat stress. Hormone- and menstrual cycle-mediated changes in the core temperature onset threshold for cutaneous vasodilation during passive heat stress are evident (Hirata et al. 1986; Charkoudian and Johnson 1997), although data regarding changes in sweating rates across the menstrual cycle are equivocal (Inoue et al. 2005; Garcia et al. 2006; Giersch et al. 2020). Altogether, the currently available data suggest that absolute core temperature and cutaneous vasodilation responses vary across the menstrual cycle, but sweat rate, skin temperature, and changes in core temperature during exercise heat stress are likely not appreciably altered (Giersch et al. 2020). We therefore do not expect that the environmental limits for heat balance would be significantly affected by menstrual cycle phase. However, investigations in this area are limited and further inquiry is warranted.

In summary, the current study described critical WBGT limits in unacclimatized young males and females as a function of relative and absolute work rates in WH and HD environments. During exercise at 30% ${\dot{\text{V}}\text{O}}_{2\max }$, the WBGT_crit was higher for females compared to males in WH, but not HD conditions. Further, WBGT_crit was higher in WH compared to HD for females. When correcting for metabolic rate, heat tolerance limits were still higher in WH compared to HD conditions for females, but were higher in males compared to females in both WH and HD environments. Although heat tolerance, defined as (1) the WBGT_crit at a standard metabolic rate of 300W (WBGT₃₀₀), and (2) the difference between observed WBGT_crit values and the OEL defined by the ACGIH TLV and NIOSH REL, was associated with ${\dot{\text{V}}\text{O}}_{2\max }$ when all subjects were included, there was no association between ${\dot{\text{V}}\text{O}}_{2\max }$ and heat tolerance within either sex, suggesting sex differences in heat tolerance that are independent of aerobic fitness in unacclimatized young adults. Together, these findings provide novel insight regarding sex differences in heat tolerance during work in WH and HD conditions.

Data availability statement

The data that support the findings of this study are available from the corresponding author, STW, upon reasonable request.