The Physician and Sportsmedicine
Menubar About Us Advertiser Services CME Resource Center Personal Health Journal Home

Heat Balance Limits in Football Uniforms

How Different Uniform Ensembles Alter the Equation

Tasha J. Kulka
W. Larry Kenney, PhD



BACKGROUND: Scant evidence documents the physiological and environmental stresses for football players wearing partial or full uniforms, but such information would be useful for determining the ambient temperatures and humidities associated with uncompensable heat stress during practice and games.

OBJECTIVE: This laboratory study used a physiological approach to determine critical heat balance limits (various combinations of ambient temperature and relative humidity) for subjects exercising in typical American football uniforms.

DESIGN: Eight nonheat-acclimatized men exercised at 35% VO2max in a programmable environmental chamber. In multiple trials, either dry-bulb temperature (Tdb) was held constant and ambient water vapor pressure (Pa) was systematically increased, or Pa was held constant and Tdb was systematically increased. The critical heat balance limits were determined as the environments at which body core (esophageal) temperatures were forced out of equilibrium, reflecting uncompensated heat storage imposed by those environments.

RESULTS: Critical environmental limits are presented that define the combinations of air temperature and relative humidity above which thermal balance cannot be maintained. These zones of uncompensable heat stress result in continuously rising core temperatures. Retrospective analysis reveals that documented heatstroke deaths in football players wearing full uniforms occurred at or above these critical environments.

CONCLUSION: Heat balance limits can be used in decision making and are especially relevant for preventing heat-related illness or injuries early in the football practice season. The critical limits are expanded when shorts are substituted for football pants with pads.

The beginning of the football season, typically mid-August to early September, coincides with the hottest and most humid part of the summer in many parts of the United States. These severe environmental conditions, coupled with uniforms that include protective padding and helmets, often expose athletes to significant heat stress. However, traditional approaches to determining "safe" environmental conditions for football have either relied on guesswork or simply noted the environments in which fatalities had occurred and assumed that less extreme conditions are "safe."1 The commonly used heat index is based on perceptual rather than physiological responses; therefore, it provides little useful information about the impact of heat and humidity on exercising humans.

While no environment guarantees safety from all forms of heat-related illness, it would be helpful to know in which environments football players can dissipate metabolic heat and that gained from the environment. Heat-related illnesses that involve excessive increases in body core temperature result from an imbalance between heat production and heat loss. Thus, the protective uniform and equipment required in American football strikes a precarious balance between protection from the rigors of sport and potential harm from the thermal environment as evaporation of sweat is impeded. While documentation exists about the effects of protective clothing ensembles on heat balance during work in hot industrial environments,2,3 scant empirical research exists about adverse environmental conditions for football players in partial or full uniforms.

To determine the environmental conditions associated with uncompensable heat stress during exercise in football uniforms, we employed an experimental design involving systematically varying environmental conditions.3-5 Environmental conditions (combinations of temperature and humidity) that exceed these heat-balance limits are associated with heat storage and a continuous rise in core temperature.

Methods and Procedures

Subject selection and physical variables. The Institutional Review Board at The Pennsylvania State University approved the project. All subjects provided informed consent after receiving detailed verbal and written descriptions of the potential risks associated with participation.

Eight nonsmoking, college-aged men served as volunteers after completing a thorough medical screening examination including medical history, resting electrocardiogram, and height, weight, and percent body fat determinations (skinfold measurements at seven sites: pectoral, triceps, midaxillary, abdomen, thigh, suprailiac, subscapular). Open-circuit spirometry was used during a maximum graded exercise test on a motor-driven treadmill to determine VO2max. Average subject characteristics were: age, 22 ± 0.3 years; VO2max, 49 ± 2 mL/kg/min; height, 179 ± 2 cm [5 ft 10 in.]; weight, 85 ± 5 kg [187 lb]; and body fat, 15% ± 2%.

None of the subjects was a competitive football player or athlete in another sport at the time of the study, but all were recreationally active. Subjects were not acclimatized to heat in any structured way before participation. All experiments were conducted between August and early January, the span of a typical football season.

The physiological approach described is only minimally influenced by subject size and adiposity. However, because the environmental data generated are workload specific, it is desirable to match the subjects' VO2max with that of the population of interest. The maximal aerobic capacity of university and professional football players has been reviewed.6 Although VO2max varies by position, the measured VO2max of professional defensive linemen is 44.9 ± 5.4 mL/kg/min,7 for professional offensive linemen and tight ends it is 49.9 ± 6.6 mL/kg/min, and for collegiate defensive linemen and linebackers it is 53.2 ± 7.3 mL/kg/min.8

Measurements. During these experiments conducted in an environmentally controlled chamber, dry-bulb (Tdb) and wet-bulb (Twb) temperatures were measured with precision mercury-in-glass thermometers mounted in a box configured to American Society of Heating, Refrigeration, and Air Conditioning Engineers specifications. Subjects' core temperatures (Tc) were measured using an esophageal probe (Tes) consisting of a thermistor sealed in a pediatric feeding tube. Placement was accomplished by inserting the probe through the subject's nose to a level near the atrium of the heart. The distance from nostrils to thermistor end was equal to one quarter of the subject's height. Skin temperature (Tsk) and heart rate (HR) were also monitored for subject safety, but these data were not used in the resulting environmental determinations.

During the experiments, subjects walked on a motor-driven treadmill at an exercise intensity equal to 35% VO2max to simulate the average energy expenditure of football players. This activity level was chosen as a steady-state, time-weighted average reflecting the intermittent nature of football practice and games. Football is a game of short-burst, high-intensity activities followed by brief recovery periods lasting from 25 to 40 seconds.9 No data are available that document the mean energy expenditure of football players across an entire practice or game, but published data do document the game's intermittent nature. Craig9 determined that maximal exercise time for a single player in a professional football game was 13.5 minutes. Zapiec and Taylor10 noted that in the Canadian Football League, the active playing time ranged from 5 minutes 42 seconds to 9 minutes 48 seconds in a game lasting 2 hours and 19 minutes.

Given this dearth of energy expenditure information, our a priori choice of 35% VO2max seemed reasonable. To document that each subject was exercising at his prescribed workload, VO2 was determined using a 3-minute expired air sample taken 30 minutes into the experiment.

Clothing. Each subject repeated the experiments dressed alternately in one of two types of football uniform in randomized order. The practice uniform included helmet, undershirt, shoulder pads, jersey, and shorts. The full uniform included helmet, undershirt, shoulder pads, jersey, shorts, as well as game pants with thigh and knee pads. Data from a shorts-only scenario, which was part of another study,11 are included for comparison. The shorts-only data were collected as for this study, using 11 subjects who had similar demographic characteristics. Several subjects participated in both sets of experiments.

Experimental procedures. These have been previously described in detail.2,3 Briefly, each subject underwent a series of experiments conducted in randomized order with at least 2 days separating the respective trials. During "critical temperature" (Tcrit) experiments, ambient water vapor pressure (Pa) was held constant at 10 or 16 mm Hg, and Tdb systematically increased 1°C every 5 minutes after a 30-minute equilibration period at 28°C (82.4°F). In the "critical water vapor pressure" (Pcrit) experiments, Tdb was held constant at either 29°C (84.2°F), 32°C (89.6°F), or 36°C (96.8°F), while Pa was increased 1 mm Hg every 5 minutes after the 30-minute equilibration period at 9 mm Hg. Three to seven subjects performed trials in any given combination of heat and humidity.

Once the real-time graph of the esophageal temperature (Tes) exhibited a clear increase in slope, the subject walked an additional 10 minutes to ensure that the increase was sustained and not artifactual. Then the experiment was terminated. A typical Pcrit test is shown in figure 1.

The upward inflection point was determined visually and rounded down to the nearest 5-minute chamber conditions to account for the 1- to 2-minute lag in Tes response time (a procedure based on previous studies). Lastly, the Tdb and Twb data were used to convert Twb to Pa and relative humidity using a standard psychrometric chart.

Statistical analysis was done using paired t-tests to determine differences between clothing trials. Significance was at the 0.05 level. For each ensemble, a curve was fitted to the data using DeltaGraph 5.0 software.

TABLE 1. Results From Determination of Critical Environmental Limits for Football Practice and
Full Football Uniforms*

Uniform Type 
Air Temperature 
(Tdb, °C)
 Critical Temperature 
(Tcrit, °C)
 Vapor Pressure 
(Pa, mm Hg)
 Critical Pressure 
(Pcrit, mm Hg)
Humidity (%)

Practice uniform

29.10 ± 0.10
31.97 ± 0.05
36.23 ± 0.19
39.28 ± 0.15
36.37 ± 0.63
9.83 ± 0.08
15.33 ± 0.67

20.13 ± 0.07
19.50 ± 0.39
17.50 ± 0.93

Full uniform

29.40 ± 0.00
32.09 ± 0.05
35.73 ± 0.26
36.37 ± 0.12
34.18 ± 0.41
10.12 ± 0.07
16.05 ± 0.03

15.87 ± 0.23
15.60 ± 0.22
13.30 ± 0.36

*Data are mean ±standard error.


During the experiments, the average exercise intensity was 34.35% ± 0.6% of VO2max. The Tcrit and Pa were recorded for each Tcrit experiment, and the Tdb and Pcrit were recorded for each Pcrit experiment (table 1). The aggregate mean data points were used to construct a curve for each family of critical environmental conditions (figure 2). The critical environmental conditions for each clothing condition therefore represent the mean data as a function of ambient temperature and relative humidity. Also included for comparative purposes are previously published data from subjects wearing only shorts and a t-shirt.11 Combinations of temperature and humidity above and to the right of each curve represent situations in which heat balance would be out of equilibrium, ie, a net gain in heat storage and a resultant steady rise in Tc. The heat stress for such conditions is uncompensable: The limitations to evaporative cooling supersede the evaporative cooling necessary for thermal balance.

As expected, across all protocols the practice ensemble resulted in Tcrit and Pcrit values that were significantly higher (P<0.05) than the full-gear ensembles; that is, the practice ensemble allowed for heat balance in more severe environments.

Fox et al1 published a graph of environmental conditions under which 9 high school and collegiate football heatstroke fatalities occurred. To retrospectively examine the validity of the present data, we replotted the environmental conditions associated with those fatalities (figure 3). Eight fatalities fell within temperature limits shown in figure 1. The ninth occurred at 64°F, 100% humidity, off the graph to the left but well above our derived line. With one exception, in which the values fall on the limit line, all deaths occurred in environments of uncompensable heat stress for exercise in a full football uniform.

In addition, 3 documented heat-related deaths occurred among football players during the summer of 2001: 1 from an Indiana high school, 1 from a Florida university, and 1 from a Minnesota professional team. Using meteorologic data provided by the National Weather Service, we plotted the environmental conditions associated with the 2001 deaths on the same graph, and all 3 occurred well above the limit line for possible heat balance in full football uniforms.


This investigation sought to determine environmental limits from a heat balance perspective for exercise at 35% VO2max for two football uniform configurations. The first ensemble was a typical practice uniform ("shell") that included helmet, undershirt, shoulder pads, jersey, and shorts. The second ensemble was a full uniform consisting of helmet, undershirt, shoulder pads, jersey, shorts, and game pants with thigh, knee, and hip pads. One basis for this comparison was determination of the critical temperature (Tcrit) at two low humidities at which heat storage becomes environmentally driven for a fixed metabolism and uniform ensemble. A second similar approach was used to determine the critical ambient vapor pressure (Pcrit) at three fixed Tdbs. Therefore, based solely on thermal balance, these environmental limits were determined prospectively to prevent an excessive rise in core temperature. These limits are in contrast to "safe" limits described previously with retrospective methods based on fatality-inducing environmental conditions.1

Fox et al1 were among the first investigators to attempt to show a clear qualitative effect of exercising in a football uniform on physiological variables, including HR, pulmonary ventilation, energy cost, Tc, and water loss. Similarly, Mathews et al12 reported that a full football uniform impedes evaporative heat loss by 60% to 70% compared with less burdensome clothing ensembles and produces increases in rectal temperature 50% greater than those of subjects wearing only shorts. However, neither group empirically determined the specific environmental limitations for men playing or practicing football, nor did they set standard guidelines describing the specific environmental conditions under which there should be a cessation of practice, more frequent rest periods or a lowered intensity, a greater emphasis on hydration, and/or a switch to lighter clothing.

The VO2max of the subjects tested here was similar to those of college and professional linemen and linebackers in previous studies.7-9 Because of the surprising lack of published data on the aerobic cost of football and its intermittent pace, the mean exercise intensity experienced by a player in a 2- to 3-hour football practice or game had to be approximated. Since practice, like the game itself, often consists of short bursts of activity consisting of drills and play-running interspersed with instruction and corrections, 35% VO2max could also reasonably approximate the average energy expenditure in a football practice. Still, one possible limitation of the current study and the environmental standards generated is their reliance on this estimate. Thus, the recommendations we present should be used as a rough guideline rather than a definitive rule. Additional limitations under the experimental conditions include the lack of a radiant heat load and lack of wind, factors that might be found in sunny outdoor conditions. Environmental heat balance cutoffs for all possible football scenarios are well beyond the scope of this study.

During a given experiment with changing environmental conditions, a typical pattern of physiological response emerges as illustrated in figure 1. Tsk increases early in the experiment, reflecting the warm ambient conditions, and just before the inflection point in Tc, HR increases, exhibiting a disproportionate rate of increase as heat balance is approached. Subsequently, the rate of increase in Tc accelerates as thermal equilibrium is exceeded. In the Pcrit tests, this Tc inflection occurs because evaporative cooling is limited by the decreased vapor pressure gradient between the saturated skin (Ps,sk) and the air (Ps,sk - Pa), resulting in a positive heat storage that forces Tc to rise. In contrast, the Tc inflection occurs during the Tcrit tests because of increased dry heat gain through radiation and convection, overcoming the effect of available evaporative cooling.

A standard psychrometric chart plots ambient water vapor pressure on the ordinate and dry-bulb temperature on the abscissa. The heat balance curves on a psychrometric chart are clearly curvilinear; ie, they flatten horizontally at high Pa values and flatten vertically at high Tdb values.11 The horizontal flattening reflects a diminishing efficiency of evaporative cooling at high vapor pressures (low [Ps,sk - Pa] gradient), even when ambient temperatures become lower. The vertical flattening reflects a physiological limitation in the rate of sweating and evaporative cooling when evaporation is efficient and high rates of sweat production and loss are evident.

Because relative humidity values are more readily available and more easily understood, the information is presented here as temperature-humidity combinations (see figures 2 and 3) rather than water vapor pressures. Within the range of temperatures shown, the heat balance limit line for players wearing only shorts and t-shirts is relatively straight, for those in the football practice uniform it begins to curve at both extremes, and for those in full uniform it shows even greater curvilinear trends (see figure 2). The biophysics of heat exchange between man and the environment predict such a pattern.

Compared with shorts only and practice uniforms, the pants in the full-gear ensemble allow even less skin surface for evaporation and drive the core temperature up faster in less extreme environments. This results not only in a significant difference between the ensembles' respective curves, but also in their shapes. As football players wear more clothing in practice or competition, humidity has less relative effect on evaporative cooling. This means that the barrier of the clothing to sweat permeation becomes of greater importance than the environment's ability to accept evaporated sweat from both the clothing surface and the small surface of exposed skin.

Critical Upper Limits

The graphs presented here can help define the upper critical environmental limits for exercise at 35% VO2max in full football uniform, practice uniform, and shorts and t-shirt only. They separate environmental zones of compensable and uncompensable heat stress. The addition of clothing, padding, helmet, and other gear decreases the range of environmental conditions associated with compensable heat stress in a predictable manner. Retrospective data from football heatstroke fatalities support the limits presented here.


  1. Fox EL, Mathews DK, Kaufman WS, et al: Effects of football equipment on thermal balance and energy cost during exercise. Res Q 1966;37(3):332-339
  2. Kenney WL, Lewis DA, Hyde DE, et al: Physiologically derived critical evaporative coefficients for protective clothing ensembles. J Appl Physiol 1987;63(3):1095-1099
  3. Kenney WL, Mikita DI, Havenith G, et al: Simultaneous derivation of clothing-specific heat exchange coefficients. Med Sci Sports Exerc 1993;25(2):283-289
  4. Belding HS, Kamon E: Evaporative coefficients for prediction of safe limits in prolonged exposures to work under hot conditions. Fed Proc 1973;32(5):1598-1601
  5. Lind AR: A physiological criterion for setting thermal environmental limits for everyday work. J Appl Physiol 1963;18(1):51-56
  6. Pincivero DM, Bompa TO: A physiological review of American football. Sports Med 1997;23(4):247-260
  7. Wilmore JH, Parr RB, Haskell WL, et al: Football pros' strengths—and CV weakness—charted. Phys Sportsmed 1976;4(10):44-54
  8. Smith DP, Byrd RJ: Body composition, pulmonary function and maximal oxygen consumption of college football players. J Sports Med Phys Fitness 1976;16(4):301-308
  9. Craig AB Jr: Exposure time to injury in professional football. Res Q 1968;39(3):789-791
  10. Zapiec C, Taylor AW: Muscle fibre composition and energy utilization in CFL football players. Can J Appl Sport Sci 1979;4(2):140-142
  11. Kenney WL, Zeman MJ: Psychrometric limits and critical evaporative coefficients for unacclimated men and women. J Applied Physiol 2002;92(6):2256-2263
  12. Mathews DK, Fox EL, Tanzi D: Physiological responses during exercise and recovery in a football uniform. J Appl Physiol 1969;26(5):611-615

Ms Kulka is a doctoral candidate in physical therapy at Duke University and a former Schreyer Honors College student in the department of kinesiology at The Pennsylvania State University (Penn State) in University Park. Dr Kenney is professor of physiology and kinesiology in the department of kinesiology at the Noll Physiological Research Center at Penn State. Address correspondence to W. Larry Kenney, PhD, 102 Noll Laboratory, The Pennsylvania State University, University Park, PA 16802-6900; e-mail to [email protected].

Disclosure information: Ms Kulka and Dr Kenney disclose no significant relationship with any manufacturer of any commercial product mentioned in this article. No drug is mentioned in this article for an unlabeled use.

The authors would like to thank Michael Zeman and Jane Pierzga for their assistance with data collection and the subjects for their contributions. Additionally, we thank E. Randy Eichner, MD, at the University of Oklahoma, and Joseph Ostrowski of the National Weather Service for their important comments and contributions.