Work in the heat is associated with many challenges. Elevations in metabolic heat production due to the physical demands of the job increase the need for heat loss to ensure thermal homeostasis. The environmental heat load experienced during work in hot environments further increases the need for heat loss from the body, especially when work is performed in the presence of powerful radiant heat sources such as the sun (when work is performed outside) and mechanized job site equipment (i.e., vehicles, tools, power sources, etc.). Moreover, elevations in humidity and the insulation from protective work uniforms can further impair the body’s ability to dissipate heat (8,22,23). Ultimately, the thermoregulatory challenges experienced during work can lead to dangerous levels of hyperthermia, which place the worker at an increased risk of heat and/or work-related injury, even in temperate environments (10).
The need to protect the occupational worker from the potential consequences of prolonged and/or severe heat stress has led to the development of multiple heat exposure guidelines that are widely recommended for use in industry. The American Conference of Governmental and Industrial Hygiene (ACGIH) published guidelines known as the threshold limit values (TLV) for heat stress consist of work–rest (WR) allocations that consider environmental conditions, expressed as wet-bulb globe temperature (WBGT), as well as estimated work intensity (i.e., metabolic rate); two key factors that influence heat exchange and thereby thermoregulation, as noted earlier. To limit the level of thermal strain experienced, the combinations of WR allocations and environmental conditions are designed to allow workers to achieve heat balance such that the rate of environmental and/or metabolic heat gain is matched by the rate of total heat loss from the body, therefore making body core temperature stable. In particular, it is assumed that the core temperature of workers will not exceed the predefined threshold of 38.0°C that may place workers at greater risk of a heat-related injury if left unchecked (1,10,19). Furthermore, the WBGT values outlined in the TLV are adjusted based on the type of clothing worn during work to account for any effect of clothing insulation on heat dissipation and therefore thermoregulation.
A major advantage of the TLV is that they consider both environmental (i.e., temperature, humidity, and solar load) and physiological (i.e., work rate) factors as well as external factors (i.e., clothing insulation) in their prescription of WR ratios, making them more encompassing in comparison with standard guidelines when assessing and managing the risk of heat strain in workers. For this reason, in addition to their simplicity and relative ease of use, the TLV guidelines have become arguably the most widely recommended heat exposure guidelines for use by industry (including mining and electrical utilities). However, to the best of our knowledge, no study has directly assessed the influence of the work exposure limits outlined in the TLV on core temperature responses during work and the associated changes in whole-body heat content. Although it is assumed that the core temperature responses during work under the TLV will be consistent, a key challenge in the interpretation of these responses presents itself in previous studies, which have observed a disparity between the measured increase in core temperature (typically esophageal or rectal temperature) and the actual amount of heat stored within the body (determined via direct calorimetry) (13). In line with this, multiple studies demonstrate markedly different changes in body heat content despite similar core temperature responses between groups (i.e., young and older adults as well as individuals with type II diabetes) (15,18,26) as well as within a group of young adults wearing different industrial protective clothing ensembles exercising in hot conditions (23).
Although the TLV exposure limits for work in hot environments are based on measures of body core temperature, the evaluation of these guidelines using thermometry in conjunction with direct calorimetry (a method that allows for the precise quantification of whole-body heat exchange and therefore the corresponding change in body heat content ) can allow for a better interpretation and understanding of the changes in heat distribution within the body in context of the observed core temperature responses. Therefore, the purpose of this study was to evaluate heat strain as determined by thermometry (i.e., rectal and skin temperatures) as well as the change in body heat content via direct calorimetry during work performed under the guidelines of the ACGIH TLV. In particular, the body temperature (rectal and skin temperatures) responses as well as the change in body heat content (direct calorimetry) were evaluated during four work protocols of varying WR allocations performed at various environmental heat loads (as defined by WBGT). These combinations of environmental conditions and WR allocations were chosen to reflect the specifications outlined in the TLV (1,19) and are representative of the ambient conditions and physical demands typically experienced in industries such as mining (11,16) and electrical utilities (4,20,21).
The current study was approved by the University of Ottawa Health Sciences and Science Research Ethics Board and is in accordance with the Declaration of Helsinki. Written and informed consent was obtained from all volunteers before their involvement in the study.
Nine healthy males took part in this study. All participants were nonsmokers, physically active (≥3 d·wk−1 of structured physical activity, ≥30 min in duration), and free from cardiovascular and metabolic disease. The physical characteristics (mean ± SD) of the participants were as follows: age, 21 ± 3 yr; height, 1.79 ± 0.06 m; body mass, 78.7 ± 12.6 kg; body fat percentage, 14.8% ± 7.7% and peak oxygen consumption (V˙O2peak), 59.2 ± 13.8 mL·kg−1·min−1.
All participants completed one screening session and four experimental trials. On the day of each trial, participants arrived at the laboratory after consuming a small breakfast (i.e., dry toast and juice) that did not include tea or coffee. Participants were also asked not to drink alcohol or perform exercise for 24 h before experimentation. Furthermore, to ensure participants reported to the laboratory well hydrated, they were instructed to drink 500 mL of water the night before as well as the morning of each session. During the screening session, body height, mass, density, and fat percentage, as well as VO2peak were determined. Body height was measured using an eye-level physician stadiometer (model 2391; Detecto, Webb City, MO), whereas body mass was measured using a high-performance digital weighing terminal (model CBU150X; Mettler Toledo Inc., Mississauga, ON, Canada). Body density was measured using the hydrostatic weighing technique and used to calculate body fat percentage (25). V˙O2peak was assessed during a maximal incremental protocol performed on a treadmill. Throughout the protocol, oxygen uptake was determined by measuring expired oxygen (O2) and carbon dioxide (CO2) concentrations (AMETEK model S-3A/1 and CD 3A; Applied Electrochemistry, Pittsburgh, PA). V˙O2peak was taken as the highest average rate of oxygen uptake measured over 30 s.
Upon arrival to the laboratory on the days of the experimental trials, participants voided their bladder and inserted a general-purpose temperature probe for the measurement of rectal temperature. Thereafter, a measurement of nude body mass was obtained. Participants were instrumented with a heart rate monitor before they entered the calorimeter chamber regulated to the WBGT for that day’s trial (see next section). Participants were then instrumented with skin temperature sensors after which they remained seated in a semirecumbent position. Thereafter, the participant performed a 120-min work protocol (semirecumbent cycling) at a fixed rate of metabolic heat production (i.e., the energy from metabolism liberated as heat—calculated as metabolic rate minus external work) of ∼360 W (considered the onset of a “moderate” work demand according to ACGIH [1,15]), a work intensity often observed during many tasks performed in the mining (16) and electrical utilities industries (21). Each work protocol consisted of different WR ratios performed at different environmental conditions (expressed as WGBT), based on the TLV (1,15). The first protocol consisted of 120 min of continuous cycling at a WBGT of 28.0°C (dry-bulb temperature, ∼41.0°C; relative humidity, ∼19.5%) (CON[28.0°C]). The remaining three protocols were composed of intermittent work bouts performed at various WR ratios and WBGT. One intermittent protocol was performed at a WR of 3:1 (15-min cycling + 5-min recovery × 6 cycles) and under a WBGT of 29.0°C (dry-bulb temperature, ∼43.0°C; relative humidity, ∼17.5%) (WR3:1[29.0°C]). The next was performed at a WR of 1:1 (15-min cycling + 15-min recovery × 4 cycles) and under a WBGT of 30.0°C (dry-bulb temperature, ∼46.0°C; relative humidity, ∼13.5%) (WR1:1[30.0°C]). The final protocol was performed under a WR of 1:3 (15-min cycling + 45-min recovery × 2 cycles) at a WBGT of 31.5°C (dry-bulb temperature, ∼46.5°C; relative humidity, ∼17.5%) (WR1:3[31.5°C]). Therefore, a total of 120, 90, 60, and 30 min of work were performed in CON[28.0°C], WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C], respectively. Because the WR allocations stipulated by the TLV assume that work and recovery are performed under similar conditions, participants remained seated on the cycle ergometer within the calorimeter for all recovery periods. Each work protocol was performed in a partially counterbalanced order for each participant (due to the fact that the number of participants and experimental trials did not allow for full counterbalancing) and separated by at least 48 h. During each trial, participants wore a 100% cotton work uniform (typically used in the mining (16) and electrical utilities industries ) consisting of a pair of work pants and a long-sleeved undergarment top as well as a helmet and a pair of gloves. No corrections to the WGBT values are required with this work uniform (1,19).
Rectal temperature was measured using a general-purpose thermocouple temperature probe (Mon-a-therm General Purpose Temperature Probe; Mallinckrodt Medical Inc., St. Louis, MO) inserted ∼12 cm past the anal sphincter. Skin temperature was measured at four locations (i.e., upper back, chest, thigh, and calf) over the left side of the body using a 0.3-mm-diameter T-type (copper/constantan) thermocouples (Concept Engineering, Old Saybrook, CT) affixed to the skin with surgical tape. Mean skin temperature was subsequently calculated using the following weights: 30% upper back, 30% chest, 20% thigh, and 20% calf. Temperature data were collected using an HP Agilent data acquisition module (model 3497A; Agilent Technologies Canada Inc., Mississauga, ON, Canada) at a rate of one sample every 15 s and simultaneously displayed and recorded in spreadsheet format on a personal computer with LabVIEW software (Version 7.0, National Instruments, Austin, TX).
The modified Snellen direct air calorimeter was used to determine the rates of evaporative heat loss and dry heat exchange (the sum of radiant, convective, and conductive heat exchange) (24). Inflow and outflow measurements of absolute humidity and air temperature were collected at 8-s intervals. Absolute humidity was measured using high precision dew point hygrometry (model 373H; RH Systems, Albuquerque, NM), whereas air temperature was measured using high precision resistance temperature detectors (Black Stack model 1560, Hart Electronics, UT). A known heat source placed in the effluent air stream was used to determine air mass flow through the calorimeter. Absolute humidity, air temperature, and air mass flow were displayed on a personal computer and recorded in real time using LabVIEW software (version 7.0; National Instruments). The rate of evaporative heat loss was calculated using the calorimeter outflow–inflow difference in absolute humidity, multiplied by the air mass flow (kg air·s−1) and the latent heat of vaporization of sweat (2426 J·g sweat−1). Similarly, the rate of dry heat loss was calculated using the calorimeter outflow–inflow difference in air temperature, multiplied by the air mass flow and specific heat capacity of air (1005 J·kg air−1·°C−1).
Metabolic energy expenditure was quantified via indirect calorimetry. Oxygen (O2) and carbon dioxide (CO2) concentrations of expired gas samples drawn from a 6-L fluted mixing box located within the calorimeter were determined using electrochemical gas analyzers (AMETEK model S-3A/1 and CD 3A, Applied Electrochemistry). Expired air was recycled back into the calorimeter chamber to account for respiratory dry and evaporative heat loss. Before each calorimetry session, a gas mixture of 17% O2, 4% CO2, and balance nitrogen was used to calibrate the gas analyzers, and a 3-L syringe was used to calibrate the turbine ventilometer. Metabolic heat production during cycling was subsequently calculated as metabolic rate minus external work (13).
Heart rate (HR) was recorded continuously and stored every 15 s using a Polar coded WearLink and transmitter, a Polar RS400 interface, and a Polar Trainer 5 software (Polar Electro, Kempele, Finland). Thermal sensation was determined using the ASHRAE seven-point scale ranging from 0 (neutral) to 7 (very, very hot), whereas the rate of perceived exertion (RPE) was measured using the Borg 14-point scale ranging from 6 (no exertion) to 20 (maximal exertion) (3).
Minute averages were calculated for rectal and mean skin temperature and reported at baseline, at the 15-min time point of each protocol (i.e., the 15-min point of the 2-h exercise in CON[28.0°C] and at the end of the first exercise bout of WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C]), as well as at the end of the 2-h experimental protocol. Furthermore, the rate of change in rectal temperature was calculated for each 5-min interval. On the basis of the average increase in rectal temperature over the last hour of each work protocol, the projected rectal temperature of each participant was determined for an extended 4-h period, which is consistent with typical industry work duration (pre- and postlunch work periods) (16). Mean body temperature was calculated from measurements of core and mean skin temperatures using a core-to-skin ratio of 0.9 to 0.1 (6). The change in body heat content for each exercise bout and recovery period was determined by calculating the temporal summation of whole-body heat loss and metabolic heat production as determined by direct calorimetry and indirect calorimetry, respectively. The summation of body heat storage during each exercise bout and recovery period was calculated to determine the total change in body heat content for each exercise/recovery cycle. The cumulative change in body heat content during exercise and recovery as well as the entire work protocol was evaluated as the sum of the change in body heat content calculated for each exercise, recovery, and exercise/recovery cycle, respectively. In addition, the change in mean body temperature was calculated from the change in body heat content determined via direct calorimetry using the following equation:
where ΔTb is the change in mean body temperature (°C), ΔHb is the change in body heat content (kJ), mb is the participant body mass (kg), and Cb is the average specific heat capacity of the body (i.e., 3.47 kJ·kg−1·°C−1). Finally, minute averages were computed for HR and reported at baseline, at the 15-min time point of the work protocol, and at the end of exercise (i.e., the 120-min time point of CON[28.0°C] and the end of the final exercise bout in WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C]) along with thermal sensation and RPE.
The body temperature responses (i.e., rectal, skin, and mean body), HR, and perceived strain (i.e., thermal sensation and RPE) were evaluated using a two-way ANOVA with the factor of time (three levels: baseline, 15 min [i.e., the 15-min point of the 2-h exercise in CON[28.0°C] and the end of the first exercise bout in the other work protocols], and end of protocol and the factor of work protocol (four levels: CON[28.0°C], WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C]). In addition, the rate of change in rectal temperature was evaluated within each work protocol to determine whether heat balance was achieved (defined as a rate of change in rectal temperature of 0°C·min−1). In CON[28.0°C], WR1:1[30.0°C], and WR1:3[31.5°C], this was evaluated every 15 min, whereas it was assessed at the end of each exercise bout (15 min) and recovery period (5 min) in WR3:1[29.0°C]. χ2 analysis was used to detect differences in the proportion of workers who achieved rectal temperatures above and below the TLV upper limit core temperature threshold of 38.0°C at the end of the 120-min work protocol as well as for the projected 4-h period. A one-way ANOVA was used to evaluate the change in body heat content during exercise and recovery as well as the cumulative change in body heat content for the entire protocol (four levels: CON[28.0°C], WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C]). When a main effect was observed, post hoc comparisons were conducted using two-tailed paired-samples t-tests. Finally, the change in mean body temperature calculated via thermometry and calorimetry were compared for each work protocol using two-tailed paired-samples t-tests. The level of significance for all analyses was set at P ≤ 0.05. All statistical analyses were completed using the software package SPSS 22.0 for Windows (IBM, Armonk, NY). Values are presented as mean ± 95% confidence interval unless otherwise indicated.
Rectal temperature (Table 1, interaction of work protocol and time, P = 0.05) increased from baseline to the 15-min point (i.e., end of the first exercise period) of each work protocol (all P values < 0.01) and increased further from the 15-min point to the end of the work protocol (i.e., 120 min) (all P values < 0.01). No differences in rectal temperature were observed between conditions at baseline (all P values ≥ 0.61) or at the 15-min time point of each work protocol (all P values ≥ 0.24). Nevertheless, rectal temperature was greater at the end of the work protocol (i.e., 120 min) in CON[28.0°C] in comparison with WR1:1[30.0°C] (P = 0.02) and WR1:3[31.5°C] (P = 0.02) as well as in WR3:1[29.0°C] relative to WR1:3[31.5°C] (P = 0.04). Results for mean skin temperature (Table 1) demonstrated no interaction of time and protocol (P = 0.37), nor a main effect of time (P = 0.42) or work protocol (P = 0.56) (Table 1).
When the average rate of change in rectal temperature was evaluated (Fig. 1), it was determined that rectal temperature was increasing throughout CON[28.0°C] (all P values ≤ 0.01), in which exercise was performed continuously, as well as at the end of each 15-min exercise bout in WR3:1[29.0°C] (all P values ≤ 0.02), WR1:1[30.0°C] (all P values < 0.01), and WR1:3[31.5°C] (both P < 0.01). Furthermore, a positive rate of change in rectal temperature was observed in the first 15-min recovery period of WR1:1[30.0°C] (P = 0.05) and the initial 30 min of the 45-min first recovery period of WR1:3[31.5°C] (both P ≤ 0.03). In contrast, the rate of change in rectal temperature was 0°C·min−1 throughout the remaining recovery periods in all work protocols (all P values ≥ 0.13).
The majority of participants did not exceed the TLV upper limit core temperature threshold of 38.0°C in WR1:1[30.0°C] (proportion below threshold: 100%; χ2, P = not applicable) and WR1:3[31.5°C] (proportion below threshold: 88%; χ2, P = 0.03) but not CON[28.0°C] or WR3:1[29.0°C] (proportion below threshold: both 67%; χ2, both P = 0.32) (Fig. 2). However, when evaluated for the projected 4-h period, the proportion of workers who achieved rectal temperatures above and below the TLV threshold were statistically similar (proportion below threshold: 25%, 33%, 25%, and 75% in CON[28.0°C], WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C], respectively; χ2, all P values ≥ 0.10).
In parallel to rectal temperature, mean body temperature (interaction of work protocol and time, P = 0.05), as determined via thermometry, increased from baseline rest to the 15-min time point of the work protocol (all P values < 0.01) and from the 15-min time point to the end of the protocol (all P values < 0.01). Moreover, mean body temperature was elevated in CON[28.0°C] (37.63°C ± 0.25°C) relative to WR1:1[30.0°C] (37.38°C ± 0.12°C; P = 0.04) and WR1:3[31.5°C] (37.30 ± 0.18; P = 0.04) at the end of the work protocol (i.e., 120 min).
There was a main effect of work protocol on the change in body heat content during the first 15 min (i.e., the first exercise bout of WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C]) of each work protocol (P > 0.01). In particular, a greater change in body heat content was observed in WR1:1[30.0°C] (124 ± 15 kJ) and WR1:3[31.5°C] (141 ± 18 kJ) in comparison with both CON[28.0°C] (94 ± 25 kJ; both P ≤ 0.05) and WR3:1[29.0°C] (98 ± 15 kJ; both P ≤ 0.03).
The cumulative change in body heat content for exercise and recovery as well as the total change in body heat content for the entire protocol is depicted in Fig. 3. A main effect of work protocol was found for the cumulative change in body heat content during exercise (P = 0.02) and recovery (P = 0.01). In particular, there was a greater change in body heat content during exercise in WR3:1[29.0°C] relative to CON[28.0°C] (P = 0.05) and WR1:3[31.5°C] (P = 0.01) as well as in WR1:1[30.0°C] compared with WR1:3[31.5°C] (P = 0.01). During recovery, a greater amount of heat was stored in WR1:3[31.5°C] in comparison with both WR3:1[29.0°C] (P < 0.01) and WR1:1[30.0°C] (P < 0.01) in which there was a reduction in whole-body heat content. However, when the total change in body heat content was evaluated between work protocols, no differences were observed (main effect of work protocol, P = 0.70).
When the relative changes in mean body temperature calculated via thermometry and direct calorimetry (Fig. 4) were compared, similar changes in mean body temperature were observed in CON[28.0°C] (P = 0.32). In contrast, the relative change in mean body temperature was greater when calculated from calorimetry relative to thermometry in WR3:1[29.0°C] (P = 0.03), WR1:1[30.0°C] (P = 0.02), and WR1:3[31.5°C] (P < 0.01).
HR and perceived strain indices.
HR, thermal sensation, and RPE for each work protocol are presented in Table 2. No differences in HR were observed between work protocols at baseline (P = 0.71), at the 15-min point (P = 0.92), or at the end of exercise (P = 0.94) between work protocols. At baseline rest, thermal sensation was similar between work protocols (P = 0.55). Moreover, thermal sensation was similar between work protocols at the end of baseline rest (P = 0.55), whereas both thermal sensation and RPE were similar at the 15-min point of the work protocol (both P ≥ 0.31). The end-of-exercise responses of both measurements were also similar (both P ≥ 0.32).
The principal finding of the current study is that moderate-intensity work in hot environments performed within the TLV guidelines led to variable core temperature responses between both participants and conditions. Although the average rectal temperatures observed were below 38.0°C, heat balance (as defined by the rate of change in rectal temperature) was not achieved in any of the 120-min work protocols. Therefore, core temperature would have continued to climb if work was continued, potentially leading to dangerous levels of hyperthermia. In addition, differences in the average core and mean body temperature responses were observed between work protocols such that the incorporation of longer rest periods (due to higher ambient temperatures) was associated with smaller increases in rectal temperature relative to work protocols with shorter rest periods. However, these findings were not paralleled by differences in the change in body heat content between work protocols as assessed by direct calorimetry. Altogether, the current study demonstrates that the TLV did not result in stable core temperature at or below 38.0°C, which would be required to prevent dangerous increases in core temperature during extended work periods, and therefore, refinement of the current guidelines is warranted.
The primary purpose of the TLV is to prevent core temperature from exceeding predefined thresholds during work in the heat by prescribing conditions in which heat balance (as defined by stable core temperature responses) can ostensibly be achieved. At first glance, it would seem that the TLV guidelines succeed in their purpose as mean rectal temperature responses did not exceed ∼37.9°C (Table 1); however, heat balance was not observed during exercise in any work protocol (Fig. 1). Although participants were in heat balance during the recovery periods in WR3:1[29.0°C], WR1:1[30.0°C], and WR1:3[31.5°C], it is important to note that a consistent increase in rectal temperature was noted in CON[28.0°C], in which exercise was performed in a continuous manner. Given that heat balance is typically observed within 30–60 min of the onset of continuous exercise (under compensable conditions) (13) and that exercise in CON[28.0°C] was performed under the lowest WBGT, it is unlikely that rectal temperature would reach stable values under any of the conditions used and would have continued to climb if work was extended past 120 min. In fact, although the TLV guidelines were successful in preventing rectal temperature from exceeding 38.0°C in WR1:1[30.0°C] and WR1:3[31.5°C] (but not CON[28.0°C] and WR3:1[29.0°C]) in the majority of participants, it was estimated that these guidelines would not prevent core temperature from exceeding this threshold in any of the conditions used if work was extended beyond 2 h (Fig. 2).
In the current study, it was projected that rectal temperature would have reached ∼40.0°C (considered a medical diagnosis of exertional heat stroke ) in one participant after 4 h of work under the conditions used in CON[28.0°C] (Fig. 2). Importantly, this response is similar to our recent findings of core temperatures reaching and exceeding 39.5°C in two workers (of a total of 32 tested) during ∼4.5 h of electrical utilities work (21), an industry in which the TLV guidelines are recommended for use. In the aforementioned study, we suggested that the high levels of thermal strain experienced by electrical utility workers may have resulted from stipulated requirements of the TLV not being met (i.e., WR allocations were variable in nature, workers often reported to work dehydrated, and clothing insulation was excessive during work with energized power lines) (21). However, the current study, in which work was performed in a regimented manner by euhydrated and lightly clothed participants, suggests that the TLV may not be adequately protecting all workers, even if used as intended.
In addition to the potential that the TLV may not prevent the core temperature of all workers from exceeding the upper limit threshold during work under the conditions used (Fig. 2), inconsistent mean body temperature responses were also observed. In particular, rectal temperature was greater in CON[28.0°C] (∼37.9°C) in comparison with WR1:1[30.0°C] (∼37.6°C) and WR1:3[31.5°C] (∼37.5°C) as well as in WR3:1[29.0°C] (∼37.8°C) relative to WR1:3[31.5°C] at the end of the 120-min work protocol. Importantly, this is in contrast to the results gleaned via direct calorimetry, which demonstrate similar changes in whole-body tissue heat content between work protocols (Fig. 3). In fact, when the calculated mean body temperature (considered a proxy measure of body heat storage ) responses were compared, thermometry was shown to underestimate the change in mean body temperature relative to direct calorimetry in all protocols except CON[28.0°C] (Fig. 4).
The observed disparities in the increase in core temperature and the change in whole-body heat content are likely the result of a heterogeneous distribution of heat within the body’s “core” tissues (i.e., organs, muscles) (13,27). The distribution of heat within the body is influenced by the thermoregulatory and cardiovascular changes associated with exercise and recovery as well as the performance of multiple work bouts (7,13,14,27). Upon the cessation of exercise, there is an abrupt, centrally mediated suppression in the heat loss responses (7,14). This response is associated with altered baroreflex activity, which also contributes to systemic vasodilation and a pooling of blood in the extremities (13). Presumably, the higher end-protocol core temperatures observed in conditions with longer exercise time and shorter recovery periods (albeit lower environmental temperatures) were likely due to sustained convective heat transfer from the muscles to the core associated with the maintenance of muscle blood flow during exercise. In contrast, the incorporation of longer recovery periods would have led to reduced muscle-to-core heat flux given the aforementioned postexercise reductions in muscle blood flow and a pooling of blood in the limbs (7).
Although it could be argued that the discrepant core temperature responses observed between work protocols are due to the total amount of exercise performed per se and not associated with postexercise alterations in blood flow and thereby heat transfer, ample evidence, albeit indirect, exists in support of the latter mechanism. For instance, it is well established that skeletal muscle can store considerable amounts of heat during exercise (up to ∼3°C increase in active muscle temperature) (9). During prolonged continuous exercise (e.g., CON[28.0°C]), it is presumed that an equilibrium in heat flux between the body’s compartments (i.e., muscle to the core, core to the skin, etc.) would occur, resulting in steady state muscle and core body temperatures. Indeed, Kenny et al. (17) recently reported ∼2.4°C and ∼0.4°C increases in vastus lateralis (active) and rectal temperature, respectively, during the first 30 min of a 60-min exercise bout (metabolic heat production, ∼450 W; dry-bulb temperature, 30°C) that were stable for the subsequent 30 min of exercise thereafter. By contrast, is a study that employed 30-min intermittent exercise bouts (interspersed with 15-min recovery periods) performed under comparable conditions (metabolic heat production, ∼500 W; dry-bulb temperature, 30°C) (12). Although vastus lateralis temperature rose ∼2.4°C during the first bout (12) (similar to Kenny et al. ), by the end of the second and third bouts, it had increased to an additional temperature of 0.2°C and 0.3°C from baseline values, respectively (a significant increase was noted between the first and second bouts) (12). Altogether, these studies support the proposed influence of the thermoregulatory and cardiovascular changes that occur between exercise and recovery (7,13,14) on the dynamic transfer of heat within the body and may help to explain the disparate core temperatures and changes in body heat content observed in the present study.
The scientific evidence clearly indicates that the development of heat-related illness and injury is driven by changes in the local temperature of thermosensitive tissues (i.e., brain, gastrointestinal tract, etc.) (14). The current findings suggest that the TLV may not adequately prevent marked increases in the temperature of these tissues and that refinement of current exposure limits is warranted. However, an important consideration for future guidelines is the method by which core temperature (for the use of defining exposure thresholds) is assessed, as the observed response can vary markedly depending on the measured tissue bed (e.g., esophageal, rectal, and tympanic) (5,27). For example, a recent study reported peak esophageal temperatures that were ∼0.50°C higher during continuous exercise relative to 5-min intermittent bouts interspersed with 5-min recovery periods (both with 60-min total exercise and recovery time each), despite no difference in peak rectal temperature or the cumulative change in body heat content between conditions (metabolic heat production, ∼650 W; dry-bulb temperature, 35°C) (5). Noteworthy, these results are also consistent with the current findings of a discrepancy between rectal temperature and the change in body heat content, which, as discussed, likely stem from differences in compartmental heat distribution (13,27).
Although excess heat stored in the periphery (i.e., in the musculature) does not directly influence the temperature of thermosensitive tissues, it still contributes to the risk of heat-related illness and injury if it is transferred to the core, a situation that may occur during large changes in blood flow such as in situations in which the worker is required to perform continuous work. In this way, the combined usage of classic thermometric measures as well as method of assessing changes in whole-body heat content (e.g., direct or partitional calorimetry) would allow for a better interpretation of the observed core temperature responses in the context of changes in the level of whole-body heat content. This all-encompassing method of assessing hyperthermia represents a valuable tool that should be used to direct the development of new exposure limits and/or the refinement of current guidelines, especially given that marked increases in mean body temperature, and therefore the level of whole-body hyperthermia, may go undetected by measurements of core temperature alone.
In summary, we show that heat balance was not achieved during moderate-intensity work performed in accordance with the TLV. It is therefore likely that the upper limit core temperature of 38.0°C would be surpassed in some workers during extended work shifts (i.e., ≥4 h) performed under these guidelines. In addition, we report inconsistent core temperature responses such that greater increases in rectal temperature were observed in protocols with longer exercise periods (and therefore reduced recovery times), albeit lower ambient temperatures. However, these responses were not paralleled by differences in the cumulative change in body heat content (via direct calorimetry) between work protocols. Foremost, the current study indicates that the TLV do not adequately protect workers from potentially dangerous increases in hyperthermia. Moreover, our findings suggest that the manner in which work is performed influences the distribution of heat within the body and thereby the measured increase in core temperature. Given that the heat exposure guidelines are typically based on the latter, future work concentrated on the development of new approaches and/or refinement of current guidelines to better protect the occupational worker during work in the heat should consider the influence of the manner in which work is performed (i.e., imposed recovery periods, work in intensity, etc.) on the dynamic transfer of heat within the body.
The authors would like to thank all of the participants who took part in the study as well as Dr. Jill Stapleton for her invaluable contribution to data collection. The authors would also like to thank Mr. Michael Sabino of Can-Trol Environmental Systems Limited (Markham, ON, Canada) for his support. All experiments took place at the Human and Environmental Physiology Research Unit of the University of Ottawa.
This research was in part supported by the Electrical Power Research Institute, the Ontario Ministry of Labour, and the Natural Sciences and Engineering Research Council of Canada (All funds held by G. P. Kenny) Dr. Glen P. Kenny is supported by a University of Ottawa Research Chair Award. Mr. Robert D. Meade and Mr. Martin P. Poirier are supported by the Natural Sciences and Engineering Research Council of Canada Alexander Graham Bell Graduate Scholarships (CGS-M and CGS-D, respectively). The provision of financial support does not in any way infer or imply endorsement of the research findings by either agency.
The authors declare no conflict of interest.
The results of the present study do not constitute endorsement by the American College of Sports Medicine or other funding agencies.