American football uniforms elicit thermoregulatory failure during a heat tolerance test

ABSTRACT

The Israeli Defense Force’s heat tolerance test (HTT) helps clinicians make return-to-activity decisions following exertional heatstroke. Participants fail the test and are “heat intolerant” if rectal temperature (T_REC) or heart rate (HR) exceed 38.5°C or 150 bpm, respectively. Ideally, tests assessing athlete heat tolerance would incorporate sport-specific factors (e.g., protective equipment). Because few clothes are worn during a HTT, its ability to assess American football players’ heat tolerance may be limited. We hypothesized wearing an American football uniform (PADS) during a HTT would lead to more classifications of heat intolerance. In this randomized, counterbalanced, crossover study, 10 men without recent exertional heat illness (age: 23 ± 3 y; mass: 78.5 ± 10.3 kg; height: 179.6 ± 7.6 cm) completed a standard HTT (CONTROL) or an HTT with PADS donned. T_REC and HR were monitored continuously for 2 hours or until T_REC reached 39.5°C. We noted when HTT failure criteria occurred. All participants failed the HTT in PADS (n = 2, T_REC >38.5°C; n = 8, HR >150 bpm); 5 failed in CONTROL (n = 1, T_REC >38.5°C; n = 4, HR >150 bpm). Participants completed more of the HTT before failure in CONTROL than PADS (61.7 ± 23.5 min vs. 43.4 ± 14.2 min; t₉ = 1.9, P =.04). The HTT cannot be made more sport-specific by simply donning PADS because PADS impaired thermoregulatory ability and produced more false positive HTT results. Consequently, the HTT should not be the sole determinant of an American football players return-to-activity following heat illness. New methods of testing heat tolerance in American football players are needed since the existing HTT is not sport specific.ABBREVIATIONS: EHS: exertional heatstroke; HR: heart rate; HTT: The Israeli Defense Force’s heat tolerance test; PADS: full American football uniform consisting of a helmet; shoulder, knee, thigh, hip and tailbone pads; a jersey top; undergarments; and half-length pants; PHT: probability of heat tolerance; RMANOVA: repeated measures analysis of variance; RPE: rating of perceived exertion; RTP: return to play; TCR: thermal-circulatory ratio; T_REC: rectal temperature; VO_2max: maximal oxygen consumption

KEYWORDS:

• Exertional heatstroke
• sporting equipment
• thermoregulation

Introduction

Exertional heatstroke (EHS) is one of three leading causes of sudden death in sport and physical activity, and is diagnosed when body core temperature exceeds 40.5°C (105°F) and the individual experiences central nervous system dysfunction[1]. EHS has killed 61 American football players between 1996 and 2018 [2] but is 100% survivable if properly recognized and treated[3]. The risk of morbidity, mortality, and organ damage increases the longer body temperature stays above 40.5°C[4], making it imperative that EHS diagnosis and treatment occurs quickly. Moreover, adherence to a clinician-guided return-to-play (RTP) protocol optimizes EHS recovery and ensures that thermal tolerance has been achieved [5,6].

One popular, functional test used to identify whether an individual can achieve thermal balance (i.e., heat gain offset by heat loss) following a heat injury is the Israeli Defense Force’s heat tolerance test (HTT)[7]. The HTT consists of an individual performing low-intensity exercise for 2 hours in the heat (40°C, 40% humidity) in standard workout apparel (i.e., socks, undergarments, shorts, shoes and sports bra [if female]). Based on current test criteria for heat tolerance, individuals who are heat tolerant will have their rectal temperature (T_REC) and heart rate (HR) plateau beneath the threshold values of 38.5°C (101.3°F) and 150 bpm, respectively [7,8]. Heat intolerance, and hence failure of the HTT, occurs if either T_REC or HR exceed these thresholds. Clinicians can also use this data to calculate the probability of heat tolerance (PHT) which is useful in reducing errors associated with manual HTT result interpretation[8]. The PHT ranges from 0 to 1 and classifies individuals into 1 of 3 categories. Values less than 0.50 indicate an individual is “heat intolerant;” values between 0.50 and 0.89 indicate “borderline heat tolerant;” and 0.90 to 1.0 indicates the individual is “heat tolerant.”

Functional tests are a necessary part of returning physically-active individuals to normal levels of activity following EHS. Ideally, tests used to help clinicians make determinations about an individual’s ability to thermoregulate should include sport-specific equipment and demands since both affect the body’s ability to thermoregulate [9,10]. Unfortunately, the standard HTT was designed for a military application and is not sport-specific to American football nor have any prior authors have attempted to modify the parameters of the HTT to make it more sport-specific to American football. Nevertheless, clinicians have utilized the HTT to determine the readiness of American football players to return to competition following diagnosed exertional heatstroke[11]. However, this may place American football athletes at risk of future heat injury since the HTT does not factor in the higher metabolic loads experienced by athletes wearing equipment [10,12]. Since American football players have the highest risk of developing exertional heat illnesses of all athletes[13], it is crucial to understand whether existing thermoregulatory tests can be made more sport specific to American football. Therefore, we questioned whether the HTT could be made more sport-specific by participants simply wearing a full American football uniform (PADS) during the test. We suspected PADS would invalidate the HTT since uncompensable heat stress occurs when PADS are worn during exercise in the heat [9,10,14].

The purpose of our study was two-fold. First, we determined if wearing PADS would lead to more classifications of heat intolerance during the HTT. Second, we examined the PHT with and without PADS[8]. We hypothesized participants would have higher T_REC and HR when wearing PADS during a HTT and that more participants would meet the heat intolerant criteria, as defined by the PHT, while wearing PADS.

Materials and methods

A repeated-measures crossover design guided this counterbalanced, laboratory study. We estimated sample size a priori using G*Power (v.3.1; Kiel, Germany) using the following parameters: alpha = 0.05, beta = 0.9, effect size = 0.25°C (T_REC)[14], nonsphericity correction = 0.4, and a correlation among repeated measures of 0.8. Based on these parameters, we estimated 8 subjects were needed to identify statistical differences between conditions.

A convenience sample of 11 physically-active and healthy men volunteered for this study. None of our volunteers were current, competitive American football players. One individual discontinued testing due to compliance and scheduling difficulty. Ten participants finished our study Table 1. Individuals were not permitted to participate if they self-reported: (1) a diagnosed cardiac, respiratory, musculoskeletal, gastrointestinal, or neurological condition; (2) a history of serious heat illness within the 6 months prior to the study; (3) an injury that prevented them from exercising; or (4) a sedentary lifestyle (exercising fewer than 3 times a week for 30 minutes)[15]. Heat acclimatization status was not assessed prior to participation and likely varied between participants given testing occurred from the Spring to Fall in the Midwestern United States. All procedures were approved by Central Michigan University’s institutional review board, and individuals provided written consent prior to participation.

Table 1. Subject demographics and descriptive information

	CONTROL	PADS
Demographics
	Age (y)	23 ± 3
	Height (cm)	179.6 ± 7.6
	Body Mass Index (kg/m^2)	24.2 ± 2.6
	Body Density (g·cc^−[^5])	1.08 ± 0.01
	Body Fat (%)	9 ± 2
	Body Surface Area (m^2)	1.97 ± 0.17
	VO2max (mL/kg/min)	51.4 ± 4.7
Hydration Measurements
	Pre-exercise Usg	1.006 ± 0.005	1.005 ± 0.004
	Body Mass Pre-exercise (kg)	78.7 ± 10.4	78.3 ± 11.5
	Body Mass Post-exercise (kg)	77.4 ± 10.4	77.3 ± 11.1
	Fluid Ingested (L)	0.69 ± 0.24	1.15 ± 0.87
	Sweat Rate (L/h)	0.96 ± 0.14	1.24 ± 0.28 ^a
	Post-testing Hypohydration (%)	2.4 ± 0.2	2.8 ± 0.7 ^b
Environmental Conditions
	Ambient Temperature (°C)	39.5 ± 0.9	39.6 ± 0.9
	Relative Humidity (%)	42.4 ± 1.1	41.8 ± 1.8

Data are means ± SD, n = 10. PADS = full American football uniform. Usg = urine specific gravity. ^a = PADS > CONTROL (t₉ = 5.6, P < 0.001). ^b = PADS > CONTROL (t₉ = 1.9, P = 0.047).

Participants reported for 3 days of testing separated by at least 48 hours; actual time between HTT testing days was 7 ± 2 days (median and interquartile range). We instructed participants to avoid tobacco, exercise, caffeine, and alcohol 24 hours prior to testing. They were also instructed to drink sufficient water before testing and at least 0.5 liters of water within the final hour before testing. Participants were instructed to sleep for at least seven hours the night before each HTT and to fast two hours before testing[16].

On day 1 (fitness assessment), participants voided their bladders completely, and urine specific gravity was assessed with a refractometer (SUR-Ne refractometer, Atago USA Inc., Bellevue, WA). Participants were rescheduled for another testing day at least 24 hours later if their urine specific gravity indicated they were too hypohydrated (i.e., >1.020)[17]. Participants drew a random number from a container, indicating the testing order they would receive on days 2 and 3 of testing (i.e., PADS then CONTROL or vice versa).

Participants were weighed nude to the nearest hundredth of a kilogram on a scale (Defender #5000, Ohaus Corp, Parsippany, NJ). We instructed participants to put on standard workout apparel minus a t-shirt (i.e., undergarments, shorts, socks, shoes) and we measured skinfold thicknesses at the quadriceps, abdomen, and chest in triplicate[18]. Participants put on their t-shirt and completed a maximum exercise test on a treadmill to determine maximal oxygen consumption (VO_2max; Vyntus metabolic cart, CareFusion, Yorba Linda, CA). Briefly, this test consisted of participants running on a treadmill at a speed of 8 kph (5 mph). After 3 minutes the speed increased to 10 kph (6.2 mph). After another 3 minutes, the speed increased to 12 kph (7.5 mph). Speed was then left unchanged but treadmill grade increased by 1% per minute until either oxygen consumption plateaued, the participant quit, or a rating of perceived exertion (RPE) of 20 was reported[19]. RPE was reported every 2 minutes. Following maximum exercise testing, they were weighed and excused.

On days 2 and 3 (HTT testing), participants’ hydration status was assessed per day 1 procedures. Participants self-inserted a rectal thermistor 15 cm from the anal sphincter [20] (YSI 4600 Precision Thermometer with model 401 probe, Advanced Industrial Systems, Prospect, KY) and donned a HR monitor (Polar Electro, Lake Success, NY). For CONTROL, participants dressed in standard workout apparel minus a t-shirt (i.e. undergarments, shorts, socks, shoes)[16]. For PADS, in addition to the standard workout apparel, participants wore pants, t-shirt, shoulder pads, helmet, and padding over the tailbone, hips, thighs, and knees (total weight = 5.7 kg). For an exact description of PADS, we direct the reader to our prior work[21].

After dressing, participants entered an environmental chamber (Espec North America, Inc, Hudsonville, MI) set to 40°C and 40% relative humidity though actual environmental chamber conditions varied slightly from these settings (39.5 ± 0.9°C, 42.4 ± 1.1% humidity; Kestrel Heat Stress Tracker #4400; Nielsen-Kellerman, Boothwyn, PA). We recorded T_REC and HR and they immediately began the HTT. Participants walked on a treadmill at a speed of 5 km/h (3.1 mph) with a 2% incline for 120 minutes, or until T_REC reached 39.5°C (safety threshold). We allowed participants to continue past the traditional termination criteria for T_REC (38.5°C [101.3°F]) and HR (150 bpm) as a small pilot to determine if new, higher termination thresholds could be identified for when PADS were donned.

T_REC and HR were monitored continuously and recorded every 5 minutes. We recorded if and when T_REC or HR exceeded 38.5°C or 150 bpm, respectively. Participants were permitted to consume room-temperature water ad libitum throughout the duration of the HTT. Water bottles were replaced halfway through the HTT or if the participant consumed all the water. We measured water volume before and after exercise to ensure valid sweat rate estimates[22]. Following the HTT, participants exited the environmental chamber, dried themselves completely, removed the HR monitor and rectal thermistor, and were weighed nude a second time. Participants then got dressed and were excused.

Data were assessed for normality via Shapiro-Wilk tests and we calculated means and standard deviations. Since the duration of the HTT differed between and within participants, we statistically analyzed only the times where T_REC and HR data were common to all participants. We used two-way (equipment condition x time) repeated measures analyses of variance (RMANOVA) to determine if differences in T_REC or HR existed between conditions over time. If we observed a significant interaction, Tukey-Kramer post-hoc tests identified differences between conditions within each time point.

We calculated the rate of HR and T_REC increase for each half of the total exercise duration in each condition to aid in the identification of any plateauing. Since many participants did not complete the 2-hour HTT in PADS, each person’s completion time was divided in half to allow comparisons between participants. We analyzed this data with a 2 × 2 RMANOVA (equipment condition x half of testing) and Tukey-Kramer post-hoc tests. We also used the T_REC and HR data and linear modeling to predict when HTT failure would occur in both conditions.

Finally, we used dependent t-tests to determine whether differences between conditions existed for exercise duration before reaching HTT failure criteria, fluid consumption, PHT values, and sweat rates. The PHT was calculated by inputting T_REC and HR into an algorithm found at https://phtheller.shinyapps.io/HTTest/[8]. This algorithm was created by other authors who analyzed HTT data from 175 individuals who experienced a heat illness. These authors utilized complex mathematical modeling techniques to maximize sensitivity, specificity, and accuracy of the PHT[8]. Significance was accepted when P ≤ 0.05 (Number Cruncher Statistical Software v.2007, Kaysville, UT).

Results

Participants were euhydrated before testing and self-reported compliance with all pre-testing instructions Table 1. Participants failed the HTT faster when they wore PADS than CONTROL (t₉ = 1.9, P = 0.04; Table 2). Moreover, more participants failed the HTT in PADS than CONTROL, most because HR exceeded 150 bpm Table 2. We stopped the HTT for 8 participants in PADS because their T_REC reached 39.5°C (100.7 ± 11.7 min). None of the CONTROL participants reached this safety cutoff.

Table 2. Heat tolerance test results

	CONTROL	PADS
HTT Failure Data
Number of subjects who failed (n)	5 of 10	10 of 10
Failed because HR >150 bpm (n)	4 of 10	8 of 10
Failed because Trec >38.5°C (n)	1 of 10	2 of 10
Time to failure for those subjects who could not complete the HTT (min)	61.7 ± 23.5 ^a	43.4 ± 14.2
Probability of Heat Tolerance Data
Probability of Heat Tolerance	0.36 ± 0.47 ^a	0 ± 0
Subjects classified as “heat intolerant” (n)	6 of 10	10 of 10
Subjects classified as “borderline” (n)	1 of 10	0 of 10
Subjects classified as “heat tolerant” (n)	3 of 10	0 of 10
TREC and HR Data
First half of HTT, Rate of TREC Increase (°C/min)	0.016 ± 0.005 ^a, b	0.025 ± 0.006 ^b
Second half of HTT, Rate of TREC Increase (°C/min)	0.006 ± 0.003 ^a	0.021 ± 0.005
First half of HTT, Rate of HR Increase (bpm/min)	0.72 ± 0.32 ^a, b	1.34 ± 0.35 ^b
Second half of HTT, Rate of HR Increase (bpm/min)	0.26 ± 0.12 ^a	0.48 ± 0.18

Data are means ± SD, n = 10. HTT = heat tolerance test. ^a = CONTROL different from PADS (P < 0.05). ^b = Different from second half of testing within each condition (P < 0.05).

T_REC differed between equipment conditions over time (F_16,144 = 146.8, P < 0.001; Figure 1). T_REC was higher in PADS than CONTROL from minute 30 to the end of testing (P < 0.05). Participants’ T_REC increased at different rates during the HTT (F_1,9 = 41.8, P < 0.001; Table 2). T_REC increased faster with PADS donned than CONTROL uniform in both halves of the HTT (P < 0.05, Table 2). Similarly, T_REC increased faster during the first half than second half within each equipment condition (P < 0.05). Linear modeling of the T_REC data indicated the time to HTT failure threshold (x) would occur in 59 minutes with PADS (y = 0.0262x + 36.955, R² = 0.99). Comparatively, linear modeling predicted CONTROL would finish the HTT before T_REC exceeded the 38.5°C failure threshold (y = 0.0105x + 37.215, R² = 0.93).

Figure 1. T_REC (a) and HR (b) responses with (PADS) and without (CONTROL) American football equipment donned during a HTT. X-axis error bars for the final PADS measurements indicate the SD for the exercise duration before T_REC was 39.5°C. Time 0 indicates the start of the HTT. Dashed horizontal lines indicate values indicative of a failed HTT. Data are means ± SD (n = 10). ^a = PADS > Control (P < 0.05)

HR also differed between equipment conditions over time (F_16,144 = 35.5, P < 0.001;

Figure 1). HR was higher in PADS than CONTROL from minute 20 to the end of testing (P < 0.05). Like T_REC, the rate of HR increase occurred differently between conditions over time (F_1,9 = 23.7, P < 0.001; Table 2). The rate of HR increase occurred faster with PADS than CONTROL in both halves of the HTT (P < 0.05). The rate of HR increase slowed during the second half of the HTT within both conditions (P < 0.05). Linear modeling of the HR data indicated participants would fail the HTT in 49 minutes with PADS (y = 0.925x + 104.04, R² = 0.92). Comparatively, CONTROL finished the HTT before exceeding the HR failure threshold of 150 bpm (y = 0.3301x + 109.8, R² = 0.85).

Sweat rates (t₉ = 5.6, P < 0.001) and PHT values (t₉ = 2.4, P = 0.02) were different between PADS than CONTROL Tables 1 and 2. While participants ingested 39% more fluid with PADS donned, this was not statistically different than CONTROL (t₉ = 1.8, P = 0.051, Table 1). Average PHT values indicated participants were heat intolerant (i.e., PHT<0.5) in both equipment conditions.

Discussion

The HTT was originally designed by the Israeli Defense Force to help individuals return to active military duty following exertional heat illness [7,23]. However, we used it as our testing protocol because of its popularity, validity, wide-spread use after an exertional heat illness [11,24] and prior application to heat acclimate subjects before determining the effect of clothing on thermoregulation[25]. To our knowledge, we are the first to attempt an equipment-based modification to the HTT and observed PADS resulted in 100% HTT failure and more classifications of “heat intolerance” in healthy men without a recent history of heat illness. Our results suggest PADS should not be worn during the HTT protocol proposed by the Israeli Defense Force[7]. Others [26] have also had difficulty making the HTT more sport-specific. Roberts et al [26]. altered the HTT by having an endurance athlete run continuously in a climatic chamber to simulate a race. Unfortunately, continuous running did not allow plateauing of T_REC or HR before age-predicted maximum, and was not recommended as a way to make the HTT more sport specific to endurance runners.

PADS resulted in HTT failure because they significantly impaired thermoregulation during exercise. Even though the HTT is considered low-intensity exercise, the weight of the PADS (~5.7 kg) increased metabolic heat production and contributed to incompensable heat stress [27,28]. Mathews et al [28]. observed that PADS increased heat production and oxygen consumption by 25 kcal/m²/hr and 0.15 L/min more than shorts, respectively, during 30 minutes of exercise. Moreover, much of PADS consist of impermeable materials which impair sweat evaporation and increase cardiovascular strain [9,10,28]. For example, Mathews et al [28]. observed HR plateaued in shorts but continued to increase without a plateau in PADS. Similarly, we observed no plateaus in HR or T_REC above 150 bpm and 38.5°C, respectively, indicating uncompensable heat stress in PADS during the HTT. Consequently, PADS increased cardiovascular strain and caused an inability to dissipate heat during exercise.

Our results agree with other authors [29] who suggested the HTT should not be the only functional test or clinical test used when making RTP decisions. First, the HTT validity is questionable since 50% of our participants failed the HTT in CONTROL despite having no recent history of heat illness and PHT values indicated the majority of CONTROL participants were heat intolerant though the PHT did successfully identify heat intolerance with PADS (0 false positives or negatives). In patients with a history of heat illness, the HTT performed much better with sensitivity, specificity, and diagnostic accuracy of 66.7%, 77.7%, and 77.2%, respectively[24]. However, others have questioned the validity of the HTT because there is some inherent subjectivity that HR and T_REC plateau, and sensitivity and specificity change depending on criteria used to define failure [16,24]. Our results indicate further testing and validation of the PHT is required before it can be recommended to clinicians as a way to improve objectivity of HTT results. Second, physiological characteristics and thermoregulation can be influenced by the presence, or absence, of training and acclimatization [29,30]. Johnson et al [30]. observed a decrease in HR and gastrointestinal temperature and an increase in sweat rate and overall exercise duration after a 15-day exercise and heat exposure protocol. Consequently, heat tolerance should not be viewed as a fixed characteristic that athletes either “have” or “have not.” Finally, heat tolerance and thermoregulatory ability may be affected by training status, lean body mass, and adipose tissue mass [9,31,32]. Our participants had varying levels of physical fitness, as indicated by their VO_2max data but many anthropometrics (e.g., height, weight, BMI, VO_2max) were comparable to heat tolerant military subjects[33]. However, different levels of physical fitness may have contributed to the number of false positive results observed in CONTROL since VO_2max is inversely correlated with max body core temperature[33]. Therefore, as athletes work to return to their pre-injury levels of fitness and physical characteristics, there may be concomitant changes in their thermoregulatory ability, and thus, HTT results[33].

The results of our small pilot project to determine if new higher termination criteria could be identified with PADS donned was unsuccessful since no higher HTT failure thresholds could be identified. Previously, a rise in T_REC of 0.0075°C/min (0.45°C total) over the 2^nd hour of the HTT was found to be more objective in determining whether participants’ data plateaued with 100% sensitivity and specificity[34]. Unfortunately, the authors could not identify a similar rise using HR data[34]. Similarly, thermal-circulatory ratios (TCR) <0.279°C/bpm have been used to identify heat intolerance during the HTT with excellent specificity and sensitivity[35]. In our study, participants’ rise in T_REC well exceeded this 0.0075°C/min rate in both halves of testing and calculated TCR values dropped below 0.279°C/bpm after 35 minutes with PADS. This and our linear modeling further support the unlikelihood of finding new higher thresholds using the standard HTT protocol. Consequently, any future investigation of new termination thresholds with PADS must modify other characteristics of the test to identify thermal balance (e.g., ambient temperature/humidity, exercise intensity).

We acknowledge three limitations of our study. First, our participants were not American football players. They were also lighter and had less body fat than those American football players most at risk of EHS[32]. Since HTT results can be influenced by body fat and fitness levels[33], we cannot assume that this study’s findings translate to similar HTT results for an American football player. Second, we did not control for participant acclimatization, which is a factor in heat tolerance[1]. However, the HTT protocol does not specify that individuals should be acclimatized before testing [7] though the HTT protocol was primarily based on data from acclimatized subjects [8,23]. Third, we only tested one uniform type in an environmental setting with exercise parameters (e.g., continuous v. intermittent exercise; low intensity v. high intensity, respectively) that vastly differ from the recommendations for safe American football activity[1]. Future research should examine how various uniform types affect thermoregulation in a variety of thermal environments since different equipment ensembles, environments, and exercise intensities affect thermoregulation [9,10,36,37].

In summary, PADS significantly influenced HR and T_REC and the results of a HTT. While the HTT may be a helpful tool for clinicians, it should not be the sole basis of return-to-play decisions[29]. Moreover, the HTT should not be used as a medical test. Rather, the HTT can be used as a framework for studying heat tolerance with various parameters potentially adjusted based on the population being studied. Similarly, HTT results should only be considered a reflection of an athlete’s current abilities during a functional test rather than an indication or absence of pathological disability[24]. EHS return-to-play criteria should continue to include a battery of clinical tests, physician clearance, and individualized gradual progression of exercise and equipment over a period of multiple weeks[38]. Developing a new HTT specific to American football is complicated [29] and is not as easy as simply having subjects don PADS. Instead, future research should attempt to develop a sport-specific HTT protocol by altering other parameters of the test (e.g., treadmill speed and incline) and having athletes wear PADS in environmental conditions safe for American football participation[1].

Ethics approval

Central Michigan University Institutional Review Board approved this study.

Consent to participate

All participant signed an informed consent approved by the Institutional Review Board of Central Michigan University prior to testing.

Acknowledgments

We thank Michael Szymanski, MS, ATC and Courtney Zickmund, AT, ATC for assisting with data collection. We also thank Central Michigan University’s Office of Research and Graduate Studies, Herbert H. and Grace A. Dow College of Health Professions, and Honors Program for funding this project.

Disclosure statement

No potential conflict of interest was reported by the authors.

Availability of data and material

These data has not been published elsewhere nor is any data set available publicly at this time.

Additional information

Funding

This project was not funded by any external agency. It was funded internally as indicated in the Acknowledgment section.