In American football, thermal homeostasis is challenged, and when exposed to both exercise and environmental heat stress, the thermoregulatory system aids in reducing thermal strain (1,6). Heat storage in the body is commonly quantified by measuring core body temperature (T). Metabolic heat production is one of the single greatest influences on the magnitude of rise in T (5,22,28), whereas behavioral modification is one method for an individual to control the balance of heat gain and loss (17). For example, a football player may take off his helmet during a break or decrease his intensity level during a drill (i.e., self-paced), which may minimize the amount of stress that is placed on the thermoregulatory system (17). However, a player may have limited control (i.e., regulated) while being required to make a certain time in a conditioning session or not be allowed to take off equipment, which could negatively effect and ultimately overwhelm their ability to maintain a T within a level that is optimal for health and safety.
An additional factor that can affect the body's ability to thermoregulate is a state of significant hypohydration. A reduction in blood volume occurs with hypohydration, and this reduction in blood volume leads to cardiovascular (CV) strain. Cardiovascular strain due to hypohydration compromises thermoregulation and leads to a rise in T (10). As a football athlete exercises in the heat, hypohydration leads to CV drift (i.e., a rise in heart rate [HR] while maintaining a given intensity) (4,10,25). This can limit CV capacity (as observed with decreases in V̇o2max) and result in a diminished ability to perform aerobic work (14). It has been repeatedly observed that hypohydration decreases exercise performance in the heat (10–13). In addition, laboratory studies have demonstrated that progressive increases in intensity (14,15) and hypohydration (16–18) will exacerbate T increases.
Despite laboratory studies identifying the effect of intensity and hypohydration on T, several field studies have failed to show the same link (9,15,19,26). Field studies that account for hydration status or intensity when measuring T are few and include varying athletic populations. In a study by Godek et al. (15), where intensity was not measured, dehydration levels experienced by subjects were <2.5%, and these values did not correlate with an increase in T. Although few studies have controlled for intensity (9,19,26), the studies that have measured percent dehydration during exercise, albeit when subjects were already in a hypohydrated state. In these cases, a direct correlation between percent dehydration and a rise in T was found (9,19,26).
Previous field studies involving football players have found no correlation between dehydration and a rise in T; however, they did not control for intensity (13). It is important to reiterate that the primary determinant of a rise in T is exercise intensity, but hydration status can potentially further modulate T responses, and therefore warrants further investigation. Therefore, the purpose of our study was to examine the influence of intensity and hydration status on T in heat-acclimatized National Collegiate Athletic Association (NCAA) Division I football players during preseason practices in the heat. The authors' hypothesis was that the metrics of exercise intensity would significantly correlate with the residual variation in T and that hydration status would have an additional significant effect.
Experimental Approach to the Problem
Although the interaction of intensity, hydration status, and T have been examined in various studies with runners (2,4,9,19,22), it is unknown if these same relationships hold true for other populations. It is plausible that because the activity profile of American football is drastically different than that of competitive running, these responses may also differ. Therefore, we used an observational study design to determine the degree of correlation among our dependent variables (hydration status and physical performance metrics) to that of core body temperature in collegiate football players during preseason training camp.
Twenty-nine male football players volunteered to participate in this study. All players were members of the same NCAA Division I, Bowl Championship Series conference football team in the Southeast United States. Players had been participating in conditioning sessions in the weeks before the start of preseason practices and therefore were considered to be heat acclimatized. Participants included all playing positions except for place kickers and punters because of the limited physical activity of these positional players during practice. Recruiting was done during the morning of each practice day. Each day included a maximum of 10 players being tested, and no subject's data were used more than once. Participants were then separated into positional groups as either linemen (L) (n = 14) or nonlinemen (NL) (n = 15). In addition, comparisons were also made between starters (S) (n = 12) and nonstarters (n = 17); however, no significant differences existed between these groups; thus, data were collapsed for presentation purposes. Demographic information for each group is presented in Table 1. This study was reviewed and approved by the University of Connecticut Institutional Review Board. Informed consent was signed by all subjects and collected by the researchers. Age range for all subjects was 18- to 22-years-old.
Inclusion criteria included being a member of the football team and participating in the first 8 days of preseason practices. Participants were excluded if they had any disease that could cause complications from taking an ingestible thermistor such as suspected obstructive disease of the gastrointestinal (GI) tract (i.e., diverticulitis and inflammatory bowel disease), exhibited or had a history of disorders or impairment of the gag reflex, had previous GI surgery, or had hypomotility of the GI tract. All participants completed a medical history questionnaire and a preseason risk assessment questionnaire before data collection and were cleared to participate by the team physician. All participants read and signed an informed consent before the start of this study, and the university's institutional review board approved the study.
Data were collected during the first 8 days of preseason football practice (August) on the university's campus (days 1 through 4) and off-campus at a training facility (days 5 through 8) in hot conditions (Table 2).
Each of the 8 days involved single practice sessions, except for day 6, which involved 2 practice sessions. Equipment worn each day varied and followed in line with the NCAA heat acclimatization guidelines. On days 1 and 2, only helmets and shorts were worn, on days 3, 4, 6 (morning session), and 8 helmets, shorts, and shells were worn, and on days 5, 6 (evening practice), and 7, full pads were worn. Each practice began with a warm-up and team stretch, was followed by “position drills”, and ended with “team drills” (Table 2). Position drills involved players being grouped based on their playing position and performing position specific activities within those groups. Team drills that occurred in a similar fashion, however, involved combining similar position groups and performing drills against one another (i.e., offensive linemen vs. defensive linemen, wide receivers vs. defensive backs, etc.). On days 3, 5, and 8, a short break (9 ± 3 minutes) was given between position drills and team drills. In addition, practice on day 5 and the evening practice on day 6 ended with a team scrimmage (25 ± 16 minutes).
Experimental Procedures and Measurements; Gastrointestinal Temperature
Ingestible thermistors (HQ, Inc., Palmetto, FL, USA) were distributed to subjects the morning of the practice session, at least 4 hours before the start of practice. Gastrointestinal temperature (Tgi) was monitored approximately every 5 minutes during practice using a handheld device (HQ, Inc.). In compliance with the normal guidelines set by the medical staff, any athlete who reached a Tgi of >103° F would be temporarily removed from practice to be cooled (by cold water dousing) to a Tgi <102.5° F. If sensors were not producing accurate readings because of not yet being in the small intestine or already being passed by the subject, that participant's data were excluded. Body mass used to calculate body mass index (BMI) was an average of 3 measurements taken the 3 consecutive weeks before the study, in addition to a measurement of the subject's height.
Intensity/Global Positioning System Device
Before each practice session, participants were provided a global positioning system (GPS) device (MinimaxX 2.5; Catapult innovations, Melbourne, Australia) (Figure 1A.). This device uses precise microtechnology sensors to analyze specific components of athletic performance such as distance covered (DC), acceleration/deceleration, and velocity (Vavg). For the purpose of analysis, V was further broken into 2 categories: velocity at a speed of 2–4 m·s−1 (V2–4) was used to indicate a moderate velocity, whereas velocity at a speed of 4–6 m·s−1 (V4–6) was used to indicate a moderate-high velocity. The GPS unit was turned on and placed inside the pocket of a protective garment. The unit was positioned at the upper back just above the shoulder blades (Figure 1B.), and the garment was worn underneath the players' regular protective equipment.
A HR monitor (Polar Electro, Inc., Lake Success, NY, USA) was placed inside the chest strap of the garments, and HR was recorded synchronously with the GPS data. Global positioning system and HR data were continuously recorded for the duration of each practice session. The GPS device recorded data at 5 Hz and stored physical components of player movements such as DC, V, and HR. After each practice session, participants returned the GPS device, and the data were uploaded to a personal computer and stored for analysis using manufacturers' software (LoganPlus 4.4.0; Catapult Innovations). For the purpose of statistical analysis, the variables used to determine intensity included average HR (HRavg), DC, and Vavg V2–4 V4–6.
Before and after each practice, subjects were weighed wearing only shorts using a calibrated scale (Tanita Corp., Tokyo, Japan) to determine percent body mass loss (%BML). In addition, subjects were asked to provide a urine sample for the determination of urine color (Ucol) and urine specific gravity (Usg). Ucol was determined using a Ucol chart (3), whereas Usg was determined using a refractometer (model A300CL; Spartan, Tokyo, Japan). For the purpose of statistical analysis, %BML, Ucol, and Usg were grouped to indicate overall hydration status.
Descriptive data (mean ± SD) were calculated for all variables. The relationship between measures of hydration status (Usg, Ucol, %BML), intensity (HRavg, DC, Vavg V2–4 V4–6), T (Tavg and Tmax), and body size (BMI and body surface area [BSA]) were analyzed with a Pearson product-moment correlation. In addition, to examine the relative contribution of exercise intensity, hydration status, and individual characteristics to the overall variability observed in Tavg and Tmax, a multivariate regression analysis was performed. Distance covered was used to calculate Vavg, but because practice times were not consistent, DC was not used in the multivariate regression analysis for intensity measures. Using an independent samples t test, each measure for Tgi, hydration measures, and intensity measures were compared between groups. Significance was set at p ≤0.05. All statistical analyses were performed using SPSS version 18.0.
There were significant correlations between Tavg and HRavg (p = 0.001) and between Tmax and HRavg (p = 0.004) (Figure 2).
Table 3 shows the cumulative results for the multivariate regression analysis used to explain the variability of Tavg and Tmax. Specific to Tavg, intensity as measured by HRavg (138 ± 9 bpm), V2–4 (4.2 ± 1.7%), V4–6 (0.6 ± 0.6%), and Vavg (0.36 ± 0.10 m·s−1) significantly accounted for 42% of the variability observed in Tavg (38.32 ± 0.34° C). Hydration as measured by %BML (−1.56 ± 0.80%) and post Usg (1.025 ± 0.006) did not significantly improve the model, but the overall variability in Tavg was still significantly explained. When adding individual characteristics, as measured by playing position (L vs. NL), BMI (32.05 ± 9.26), and BSA (2.46 ± 0.41 m2), the cumulative variability accounted for increased to 60%.
Specific to Tmax, intensity measures significantly accounted for 43% of the variability observed in Tmax (38.83 ± 0.42° C). Adding hydration measures to the model did not significantly improve the model but the overall variability in Tmax was still significantly explained. When adding individual characteristics, the cumulative variability accounted for increased to 53%.
Tavg during practice was not statistically significant (p > 0.05) between L vs. NL 38.20 ± 0.28° C vs. 38.43 ± 0.37° C, respectively. Tmax was also not significantly different (p > 0.05) between L (38.73 ± 0.33° C) and NL (38.92 ± 0.50° C).
Table 4 displays intensity measures between groups. HRavg during practice was not significantly different (p > 0.05) between L and NL. The relationship between Tavg and HRavg for L and NL and Tmax and HRavg are shown in Figures 3 and 4, respectively.
The DC during practice was lower in L than NL (2,583 ± 353 m; 3,677 ± 855 m, p < 0.001, respectively). During practice, L had a lower Vavg than NL (0.29 ± 0.04 m·s−1; 0.42 ± 0.09 m·s−1, p < 0.001, respectively). We also chose to express these variables as overall practice time spent in each of these velocity zones. Lineman spent less time at V2–4 (p < 0.001) and at V4–6 (p < 0.001) than NL. Percent of practice time spent at each velocity zone is shown in Figure 5.
Table 5 displays hydration indices between groups. Percent body mass loss, pre Usg, post Usg, pre Ucol, and post Ucol were not significantly different (p > 0.05) between L and NL (Table 6).
The purpose of this study was to determine the influence of intensity measures on T in NCAA Division I football players during preseason practices and to examine how much of an additional influence hydration status and individual characteristics had on a rise in T. To the author's knowledge, this study is the first to use a collegiate football population and observe both intensity and hydration status influences on a rise in T during preseason training camp.
The primary influence on a rise in T was exercise intensity. We were able to show that variation in the intensity of exercise during a football practice significantly correlated with the observed variation in Tavg and Tmax. It is important to note that the measures used to quantify exercise intensity in the current protocol are limited in their ability to fully depict the physical demands imposed on football players. Components such as vertical movement (i.e., a receiver jumping to catch a pass) and isometric force (i.e., a lineman blocking an opposing linemen) are not captured by the GPS device. Therefore, it is reasonable to assume that the clinical significance between exercise intensity and T may be greater than what the statistical correlation demonstrated. In previous studies, exercise intensity has been shown to be one of the single greatest influences in a rise in T (5,13,16,20,22,27,28). Our findings support these previous studies that show intensity to have the greatest influence on T. Tucker et al. (28) had 10 male subjects perform 20-km time-trial cycling bouts in both hot and cool conditions and was able to show that in a self-paced exercise in the heat, intensity is reduced to reduce heat production and therefore the rate of rise in T. Furthermore, Jay et al. (16) showed that heat production per unit body mass described changes in T irrespective of fitness status. Most of the football practices were self-paced and allowed the players to dictate their own intensity; however, as this was an observational study, outside influences such as competitive drive and coaches influences may have also played a role.
When examining potential differences among positional groups, this analysis produced a significant relationship between some, but not all, intensity variables. Despite differences in V and DC between L and NL (Table 4), no differences were observed in HR, Tavg, or Tmax (p > 0.05). We reported L to have a significantly lower Vavg, lower percentage of time in higher V zones, and less DC as shown in Table 4 and Figure 5. As a result, it would be implied that L would also have a lower T because of less overall movement and movement at slower speeds. However, based on the work done by Deren et al. (11), this could be explained by that sweating efficiency, and therefore heat dissipation, may have been compromised in the L group secondary to the morphological differences seen in L compared with NL. Similarly, in a study where V was controlled for, Marino et al. (25) showed that compared with heavier runners, those with a lower-body mass produced and stored less heat when running at the same speeds in the heat. Because there are different drills between L compared with NL, speed is not always similar between groups during practices. If speed between groups was required to be the same, we can hypothesize the heat storage for that of the L to be exacerbated because of variability in height, mass, and BSA (23). In addition, as L are required to participate in drills that require more isometric force production than NL, this would also likely increase heat storage in this group. High metabolic heat production and a rise in T is largely representative of exercise intensity and body mass and can influence the onset and rate of sweating (17,20). Finally, the difference in air velocity between L and NL may also partly explain this result. As indicated from our velocity data, NL move at faster rates throughout the practice session. As a result, evaporative heat loss is likely higher in those individuals, thereby allowing for greater cooling.
Hydration indices, when added to measures of intensity, did not significantly influence the prediction in the rise of Tavg and Tmax (p > 0.05). Decades of previous laboratory studies have shown that progressive amounts of dehydration result in an increase in T (2,21,24). Sawka et al. (24) showed a correlation between varying levels of hypohydration and T using 8 heat-acclimatized males that performed heat stress tests at a euhydrated state and at 3, 5, and 7% levels of hypohydration. This study was able to show that when intensity is controlled for, dehydration significantly contributed to the rise in T. However, it is important to note that these studies (2,21,24) included protocols in which subjects began the exercise bout in a hypohydrated state. This may largely explain why the results of the current study did not match that of previous literature, while also emphasizing the importance of starting an exercise session in a euhydrated state. Another explanation is that the current study implemented a field setting in which intensity was not regulated. An athlete may be dehydrated, and because of that, they may choose to not work at a high intensity. As a result, you may see a dehydrated individual with a less drastic rise in T compared with a hydrated individual who is able to work at a higher intensity.
Recent field studies using football players who have not measured or controlled for intensity failed to find a correlation between hydration status and T rise (7,13,15,29). For example, Yeargin et al. (29) observed DI football players during 11 days of heat acclimatization during preseason practices. The lack of association of T and dehydration was attributed to not measuring intensity, a main contributor to a rise in T, as well as to a mild degree of dehydration (−1.2%). To our knowledge, there are only 3 previous field studies that have controlled for intensity. For example, Casa et al. (9) showed that experienced heat-acclimatized runners had a greater rise in T when dehydrated during a submaximal run. However, as previously discussed, these studies all involved protocols in which subjects began the exercise bout in a hypohydrated state. Although these studies (9,19,26) showed an additional influence of dehydration on T, it is not possible to conclude that this finding was due solely to the active dehydration endured during the exercise protocol or if this finding was confounded due to the hypohydrated state at the start of the protocol. Likewise, as this study did not control for intensity, we found no additional influence on the rise in T due to hydration status.
The average %BML of −1.56% from our study was mild and not as drastic as that from previous studies (−3.6% to −4.5%) (9,19,26). Studies that have shown %BML ranging from −1.11% to −1.78% (7,13,15,18) have found no relationship with T. These studies did not measure intensity and therefore cannot determine whether the lack of correlation between dehydration and T was due to the athlete not performing at a high level of intensity or due solely to dehydration. According to the National Athletic Trainers' Association Fluid Replacement Position Statement (8), significant dehydration is classified as %BML of −3% to −5%. As shown in Table 5, our participants' %BML falls into the well hydrated category. This likely was attributed to the accessibility of fluids and the educational resources provided on hydration for these athletes.
In conclusion, exercise intensity measures in the field setting account for a significant amount of the variability seen in the rise of T. All participants were exposed to similar influences that effect heat dissipation such as wearing similar equipment, being acclimatized to the environment, and equal environmental influences with temperature and humidity. As a result, metabolic heat production was the most variable factor from a heat balance perspective. This was shown by the significant influence that intensity had on a rise in T. Hydration status was not found to be an additional predictor in the variability in a rise in T. These findings were consistent with other field-based studies that were not able to control for intensity, as well those finding minimal influence with mild levels of dehydration (<2%).
First, exercise intensity is the primary influence on metabolic heat production during exercise, and depending on how well the body can dissipate that heat will determine how high an athlete's T is going to rise. The drills and intensity level that football players perform at is highly variable and can contribute to the individuality of metabolic heat production. Therefore, how hard an athlete is pushed or how hard they push themselves through activity will be the most influential reason for rise in T. As a high T (>40° C) is one of the determining signs of an exertional heat stroke, it is essential to understand that the physical activity intensity of an athlete is crucial in the manifestation of this illness. Second, the lack of association of hydration status with Tmax and Tavg does not negate the importance of drinking fluids for athletes. Although maintaining hydration is imperative to an athlete's health and can influence performance, hydration was not shown to be as much of a contributor to a critically high T as is intensity. However, it must be emphasized that this study produced only moderate thermoregulatory stress, as exhibited by the moderate elevations in T and minimal dehydration throughout exercise.
One limitation was that the population used was a very homogeneous population as it pertains to hydration indices. They were extremely well educated in the realm of hydration and had easy access to fluids constantly through practice. Although fluid intake was not measured, it could have been the reason that there were no extreme levels of dehydration. There was also a large number of medical staff monitoring the players during practice. For those players using ingestible thermistors, the medical staff would keep a close eye on those who reached 102.5° F and would cease activity for those who reached 103° F until their T returned to safer limits. These resources for fluid intake and with access to medical staff are not consistent across various sports settings, especially at the high school level.
Given that this was an observational field study, several confounding variables that may also influence the rise in T were not able to be controlled for, namely hydration status, sleep patterns, and nutritional intake. The most important of these is hydration status in which we were unable to control prepractice hydration status nor the fluid the subjects consumed during practice. However, given the relatively minor %BML (−1.56%) of our subjects during practice, we can reasonably conclude that they did a good job of hydrating during practice. Nonetheless, variance among starting hydration status may have influenced some of the results.
Given that this is the first field study to view the relationship of intensity with T using American football players, it is imperative that studies continue to analyze the effects of intensity and hydration measures on T in this population. Although preliminary research has observed a reduced sweating efficiency in linemen (11), future studies should measure sweat rates among different positions, while measuring intensity to analyze metabolic rate and T changes. Different populations with varied access to hydration supplies or even varied levels of fitness as seen in high school athletes should be studied, as this could reveal dynamic results. Future studies monitoring an individual's intensity over time and the influence on HR, T, and sweat rate could be done to evaluate current acclimatization guidelines at the high school and collegiate settings.
1. Armstrong LE, Johnson EC, Casa DJ, Ganio MS, McDermott BP, Yamamoto LM, Lopez RM, Emmanuel H. The American football uniform: Uncompensable heat stress and hyperthermic exhaustion. J Athl Train 45: 117–127, 2010.
2. Armstrong LE, Maresh CM, Gabaree CV, Hoffman JR, Kavouras SA, Kenefick RW, Castellani JW, Ahlquist LE. Thermal and circulatory responses during exercise: Effects of hypohydration, dehydration, and water intake. J Appl Physiol (1985) 82: 2028–2035, 1997.
3. Armstrong LE, Soto JA, Hacker FT Jr, Casa DJ, Kavouras SA, Maresh CM. Urinary indices during dehydration, exercise, and rehydration. Int J Sport Nutr 8: 345–355, 1998.
4. Armstrong LE, Whittlesey MJ, Casa DJ, Elliott TA, Kavouras SA, Keith NR, Maresh CM. No effect of 5% hypohydration on running economy of competitive runners at 23 degrees C. Med Sci Sports Exerc 38: 1762–1769, 2006.
5. Bergeron MF, McKeag DB, Casa DJ, Clarkson PM, Dick RW, Eichner ER, Horswill CA, Luke AC, Mueller F, Munce TA, Roberts WO, Rowland TW. Youth football: Heat stress and injury risk. Med Sci Sports Exerc 37: 1421–1430, 2005.
6. Binkley HM, Beckett J, Casa DJ, Kleiner DM, Plummer PE. National athletic trainers' association position statement: Exertional heat illnesses. J Athl Train 37: 329–343, 2002.
7. Byrne C, Lee JK, Chew SA, Lim CL, Tan EY. Continuous thermoregulatory responses to mass-participation distance running in heat. Med Sci Sports Exerc 38: 803–810, 2006.
8. Casa DJ, Armstrong LE, Hillman SK, Montain SJ, Reiff RV, Rich BS, Roberts WO, Stone JA. National athletic trainers' association position statement: Fluid replacement for athletes. J Athl Train 35: 212–224, 2000.
9. Casa DJ, Stearns RL, Lopez RM, Ganio MS, McDermott BP, Walker Yeargin S, Yamamoto LM, Mazerolle SM, Roti MW, Armstrong LE, Maresh CM. Influence of hydration on physiological function and performance during trail running in the heat. J Athl Train 45: 147–156, 2010.
10. Coyle EF, Gonzalez-Alonso J. Cardiovascular drift during prolonged exercise: New perspectives. Exerc Sport Sci Rev 29: 88–92, 2001.
11. Deren TM, Coris EE, Bain AR, Walz SM, Jay O. Sweating is greater in NCAA football linemen independently of heat production. Med Sci Sports Exerc 44: 244–252, 2012.
12. Du Bois D, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. 1916. Nutrition 5: 303–311, 1989; discussion 312–313.
13. Fowkes Godek S, Godek JJ, Bartolozzi AR. Thermal responses in football and cross-country athletes during their respective practices in a hot environment. J Athl Train 39: 235–240, 2004.
14. Ganio MS, Wingo JE, Carroll CE, Thomas MK, Cureton KJ. Fluid ingestion attenuates the decline in VO2peak associated with cardiovascular drift. Med Sci Sports Exerc 38: 901–909, 2006.
15. Godek SF, Bartolozzi AR, Burkholder R, Sugarman E, Dorshimer G. Temperature and percentage of dehydration in professional football linemen and backs during preseason practices. J Athl Train 41: 8–14, 2006; discussion 14–17.
16. Jay O, Bain AR, Deren TM, Sacheli M, Cramer MN. Large differences in peak oxygen uptake do not independently alter changes in core temperature and sweating during exercise. Am J Physiol Regul Integr Comp Physiol 301: R832–R841, 2011.
17. Johnson EC, Ganio MS, Lee EC, Lopez RM, McDermott BP, Casa DJ, Maresh CM, Armstrong LE. Perceptual responses while wearing an american football uniform in the heat. J Athl Train 45: 107–116, 2010.
18. Laursen PB, Suriano R, Quod MJ, Lee H, Abbiss CR, Nosaka K, Martin DT, Bishop D. T and hydration status during an ironman triathlon. Br J Sports Med 40: 320–325, 2006, discussion 325.
19. Lopez R. Examining the Influence of Hydration Status on Physiological Responses and Running Speed During Trail Running in the Heat With Controlled Exercise Intensity [Doctoral Dissertation]. Mansfield, CT: University of Connecticut; 2011.
20. Marino FE, Mbambo Z, Kortekaas E, Wilson G, Lambert MI, Noakes TD, Dennis SC. Advantages of smaller body mass during distance running in warm, humid environments. Pflugers Arch 441: 359–367, 2011.
21. Montain SJ, Coyle EF. Influence of graded dehydration on hyperthermia and cardiovascular drift during exercise. J Appl Physiol (1985) 73: 1340–1350, 1992.
22. Noakes TD, Myburgh KH, du Plessis J, et al. Metabolic rate, not percent dehydration, predicts rectal temperature in marathon runners. Med Sci Sports Exerc 23: 443–449, 1991.
23. Noel MB, VanHeest JL, Zaneteas P, Rodgers CD. Body composition in division I football players. J Strength Cond Res 17: 228–237, 2003.
24. Sawka MN, Young AJ, Francesconi RP, Muza SR, Pandolf KB. Thermoregulatory and blood responses during exercise at graded hypohydration levels. J Appl Physiol (1985) 59: 1394–1401, 1985.
25. Shirreffs SM, Armstrong LE, Cheuvront SN. Fluid and electrolyte needs for preparation and recovery from training and competition. J Sports Sci 22: 57–63, 2004.
26. Stearns RL, Casa DJ, Lopez RM, McDermott BP, Ganio MS, Decher NR, Scruggs IC, West AE, Armstrong LE, Maresh CM. Influence of hydration status on pacing during trail running in the heat. J Strength Cond Res 23: 2533–2541, 2009.
27. Tatterson AJ, Hahn AG, Martin DT, Febbraio MA. Effects of heat stress on physiological responses and exercise performance in elite cyclists. J Sci Med Sport 3: 186–193, 2000.
28. Tucker R, Marle T, Lambert EV, Noakes TD. The rate of heat storage mediates an anticipatory reduction in exercise intensity during cycling at a fixed rating of perceived exertion. J Physiol 574: 905–915, 2006.
29. Yeargin SW, Casa DJ, Armstrong LE, Watson G, Judelson DA, Psathas E, Sparrow SL. Heat acclimatization and hydration status of American football players during initial summer workouts. J Strength Cond Res 20: 463–470, 2006.