Grouped Pearson Product Moment Correlationsa Between Physiologic and Perceptual Responses During Exercise in a 33°C Environment

The rate of Tre rise was different among all conditions (CON = 0.026 ± 0.008, PART = 0.034 ± 0.006, FULL = 0.042 ± 0.010) (F2,27 = 10.69, P < .001). Using the experimental RTre values and the mean observed initial Tre (36.99 for CON, 36.98 for PART, and 36.89 for FULL), we calculated the estimated time to reach the limit of 40°C. We found a noteworthy reduction in the estimated time to reach a Tre of 40°C in the FULL versus the PART condition and in the PART versus the CON condition (Table 3).

At first, the scaled score for initial and final time point ESQ responses was compiled for each condition. Initial responses were not different between the CON and the PART conditions (t9 = 0.768, P = .46), between the PART and FULL conditions (t9 = −1.048, P = .32), or between the CON and FULL conditions (t9 = 0.293, P = .78). Within each condition, the scaled score increased (from the initial to the final time point) for the CON (t9 = −2.371, P = .04), PART (t9 = −4.583, P = .001), and FULL (t9 = −5.103, P = .001) conditions. The scaled score final time point responses were greater in the PART than in the CON (t9 = −2.363, P = .04) condition and were greater in the FULL than in the CON (t9 = −2.569, P = .03) condition. However, we found no difference between the PART and FULL conditions.

Finally, to examine which questions were most sensitive to the perceived differences while wearing a football Bas eball Uniforms , individual t tests were performed. In 3 of 16 statements, the final time point mean responses were lower in the CON than in the PART or FULL conditions (Table 4): “I feel dizzy” (t9 = −2.905, P = .02 and t9 = −2.571, P = .03, respectively); “I feel faint” (t9 = −2.739, P = .02 and t9 = −2.806, P = .02, respectively); and “I feel warm” (t9 = −4.431, P = .002 and t9 = −5.461, P < .001, respectively) (Figure 3). The final time point mean values were lower in the CON than in the FULL condition only for 4 additional statements: “I have a headache” (t9 = −3.498, P = .007); “It is hard to breathe” (t9 = −2.882, P = .02); “I feel sick/nauseated” (t9 = −2.535, P = .03); and “I feel irritable” (t9 = −2.449, P = .04). Final time point responses to the statement “I feel weak” were greater only in the PART condition (compared with the CON [t9 = −4.714, P = .001] condition).

Modified Environmental Symptoms Questionnaire responses (mean ± SD) before and immediately after exercise, n = 10. A, “I feel dizzy;” B, “I feel faint;” C, “I feel warm.” a Indicates different from control uniform condition final time point value (P < .05).

Humidi ty between the skin and uniform for the PART and FULL conditions was similarly near maximum at the end of exercise (90.41% ± 4.43% and 89.03% ± 3.57%, respectively; t9 = −0.953, P = .37). The uniform humidity data for the PART condition, the FULL condition, and the 2 uniform conditions combined were then plotted against each participant's THM rating matched at specific times. The resulting Pearson product moment correlation coefficients for THM versus uniform humidity in the PART condition (n = 10), the FULL condition (n = 10), and combined uniform conditions (n = 20) were r = 0.475 (P = .002), r = 0.756 (P < .001), and r = 0.585 (P < .001), respectively.

During the 30 experiments (10 per uniform condition), 19 ended before completion of 60 minutes of treadmill walking because the participants were physically exhausted. In these 19 trials, no participant reached the predetermined Tre limit of 40°C or displayed signs or symptoms of heat illness. Although the mean exposure time was lower in the FULL condition than in the PART condition, this value was not different because of the variability and range of exposure times in each trial. We attributed these differences in exposure time to the cardiovascular physical fitness and heat tolerance levels of participants.

Bec ause of the differences in heat exposure time and the limitations of the Tsk readings, the RTre may be a more descriptive variable of the participant's response to exercise in the heat while wearing the different uniforms. Although no final Tre differences were seen between the PART and FULL conditions, the RTre values supported the hypothesis that the addition of shoulder pads and a helmet inhibits body heat loss. Based on the RTre for each uniform, prolonged exercise resulted in a Tre of 40°C in a shorter time in the FULL condition than in the PART condition.

T he similar THM ratings in the PART and FULL conditions are contrary to the research of Cotter et al,30 who reported that the head was the most sensitive part of the body and strongly influenced the perception of total-body THM. In our study, we found no change in THM when a helmet was added. This dissimilarity was not surprising given the environmental conditions and metabolic requirements of our protocol. The extreme drive for cooling via skin blood flow may be reflected by near-maximal and similar Tsk being measured throughout the protocol. Gagge et al31 reported that Tsk must differ by at least 2°C to be perceived as being different. The Tsk in our study ranged from 33.0 ± 1.0°C to 34.5 ± 1.1°C in the CON condition, from 33.5 ± 1.0°C to 35.0 ± 1.0°C in the PART condition, and from 33.4 ± 0.9°C to 35.6 ± 1.0°C in the FULL condition. Therefore, a between-conditions difference in THM ratings at milestone time points was not hypothetically anticipated.

The thermoregulatory theory of fatigue32 postulates that a critical Tre (approximately 40°C) results in voluntary exhaustion. In our study, exercise time was reduced in the PART and FULL conditions compared with the CON condition. However, THM responses were not different among conditions at exhaustion. Before reaching the Tre threshold and concurrent with exercise cessation, a participant's THM was almost maximally rated in the PART and FULL conditions and remained close to maximum until the end of exercise, despite a rising Tre.

At the end of exercise, no correlation existed between THM and either Tre or Tsk (ie, maximal perceptual scores were perceived before participants reached the critical Tre). The lack of agreement between the perceptual ratings and the physiologic responses to cessation of exercise may have been due to the high humidity inside the uniform, which decreased each participant's ability to dissipate heat, because both skin and air within the microenvironment were virtually saturated. Increased humidity within the uniform may have increased THM at the skin, resulting in elevated perceptual ratings before increased heat storage and elevated Tre. A second plausible explanation for the inability of THM to predict exercise cessation involves the difference between peripheral and deep-tissue thermosensors. Thermosensors located near the skin surface are responsible for behavioral modifications,33 which play a large role in perception of heat. Deep-tissue thermosensors are different because their input results in the body's autonomic responses to high internal body temperatures.33 Afferent impulses from both areas are relayed to the hypothalamus, are integrated, and influence efferent responses.33,34 However, in the PART and FULL conditions, the skin was insulated, except where the Tsk readings were taken. This may have resulted in high activation of skin thermosensors, which superseded the input from deep-tissue thermosensors and may have resulted in the elevated perceptual ratings that were observed before reaching a critical Tre.

The THM and MPN were positively correlated at both the post-RBL and final time points. This result may reflect the opening of specific ion channels that are associated with thermal nociception (ie, thermal pain that occurs under high levels of THM).35 Theoretically, it would be useful to have a THM scale similar to MPN, which contains a point beyond the maximal number rating. The participant then could rate perceived pain that is due to THM. We recommend combining the MPN and THM ratings and evaluating the resulting scale in future studies.

As environmental humidity increases, the ability of the body to dissipate heat via evaporation decreases as a result of the increased partial pressure of water in the air surrounding the body; this, in turn, results in greater heat storage. Therefore, an increase in environmental humidity should result in an increased THM rating. Indeed, increased microenvironmental humidity between the uniform and skin (PART and FULL conditions) correlated positively with THM. Compared with the CON condition, the FULL and PART conditions also contained an increased microenvironmental humidity between the skin and uniform. Thus, increased THM due to uniform humidity, regardless of environmental humidity, can reflect limited heat loss and, in turn, can contribute to decreased performance.

The only significant ESQ perceptual difference appeared in final time point ESQ responses. Interestingly, 2 of the 4 ratings that were higher in the FULL than in the CON condition but were not higher in the PART than in the CON condition were related to specific equipment items. When the helmet was worn (FULL condition), responses to “I have a headache” were greater relative to the CON condition. In addition, when the shoulder pads were worn, applying pressure to the chest and upper back (FULL condition), responses to “It is hard to breathe” were greater relative to the CON condition.

F inally, based on past football heat studies,7–10,14 we expected to observe a more pronounced perceptual difference among conditions, but this did not occur. However, the correlations among different perceptual scales and between the perceptual scales and physiologic responses yielded novel information. The RPE was positively correlated with TST, THM, and MPN after RBL and at exhaustion/completio n. These correlations indicated that all measured perceptions may be integrated and may influence RPE ratings. Finally, although the correlation between final RPE and final Tre that was expected based on past studies7–10,14 was not present, a positive correlation between RTre and final RPE was observed.

Al though American football uniforms, including shoulder pads and helmets, protect players from impact, they reduce heat dissipation, increasing the risk of heat illness secondary to dehydration and hyperthermia. Our study confirmed that the addition of a uniform with or without pads increases RTre, Tsk, and RPE at a given workload and decreases the amount of exercise an individual can safely perform. These findings should be considered by coaches who design practices, particularly if players are not heat acclimatized or if they are unfit in the early season, when most heat illnesses occur. Increased strain was not reflected by any of the perceptual ratings. This may be an important risk factor for hyperthermia or heat stroke, because football athletes commonly exercise 2 to 3 hours during practice sessions. Our data indicated that vigilance on the part of the coaching and athletic training staffs is imperative, because a football athlete may not be aware that he is experiencing life-threatening hyperthermia.