Sports selection
Guidelines with thermal thresholds were confirmed for nine of the 21 outdoor sports of the 2020 Tokyo Olympics (athletics [24][25], beach volleyball [26], field hockey [27], football [28], golf [29], rowing [30], rugby seventh [18], tennis [31], and triathlon [32]). Most sports adopted WBGT as an indicator in their guidelines (six; in rowing, WBGT is used in combination with Ta), with an upper threshold ranging 28–32.2℃. Other indicators were Ta (three sports), Heat Stress Index (one sport) and apparent temperature (one sport, as an alternative to WBGT). In six of the above nine sports, core temperatures and the parameters necessary for the simulations were found in the literature (athletics [marathon] [33][34][35], football [36][37][38], rowing [39], rugby seventh [40], tennis [41][42][43], and triathlon[44]). Thus, we verified the guideline validity for these six sports. The overall results are presented in Table 2.
Table 2 Sport selection results
Outdoor sports in Tokyo Olympics 2020
|
Thermal safety guidelines with upper thresholds
|
References with core temperature and calculation parameters (number of references and sample size [n])
|
Selected sports
|
Guidelines identified
|
Index (upper threshold)
|
Author(s) (*other than IFs)
|
Archery
|
F
|
N/A
|
|
N/A
|
F
|
Athletics (marathon)
|
T
|
WBGT (28℃)
|
ACSM*, IIRM*
|
T (3, n=99)
|
T
|
Baseball/softball
|
F
|
N/A
|
|
N/A
|
F
|
Beach volleyball
|
T
|
WBGT (31℃)
|
FIVB
|
F
|
F
|
Canoe
|
F
|
N/A
|
|
N/A
|
F
|
Cycling (mountain bike, road, BMX)
|
F
|
N/A
|
|
N/A
|
F
|
Equestrianism
|
F
|
N/A
|
|
N/A
|
F
|
Field hockey
|
T
|
Ta (42℃) /
RH (75%)
|
FIH
|
F
|
F
|
Football
|
T
|
WBGT (32℃)
|
FIFA
|
T (3, n=38)
|
T
|
Golf
|
T
|
Ta (37℃)
|
IGF
|
F
|
F
|
Modern pentathlon
|
T
|
WBGT (28℃)
|
UIPM
|
F
|
F
|
Rowing
|
T
|
Ta (38℃) /
WBGT (32℃)
|
FISA
|
T (1, n=4)
|
T
|
Rugby seventh
|
T
|
HIS (150)
|
World Rugby
|
T (1, n=22)
|
T
|
Tennis
|
T
|
WBGT (32.2℃) / Tap (34℃)
|
ITF
|
T (3, n=25)
|
T
|
Triathlon
|
T
|
WBGT (32.2℃)
|
World Triathlon
|
T (1, n=14)
|
T
|
Sailing
|
F
|
N/A
|
|
N/A
|
F
|
Shooting
|
F
|
N/A
|
|
N/A
|
F
|
Skateboarding
|
F
|
N/A
|
|
N/A
|
F
|
Sport climbing
|
F
|
N/A
|
|
N/A
|
F
|
Surfing
|
F
|
N/A
|
|
N/A
|
F
|
3x3 basketball
|
F
|
N/A
|
|
N/A
|
F
|
Abbreviations: ACSM, American College of Sports Medicine; FIFA, International Federation of Association Football; FIH, International Hockey Federation; FIVB, International Volleyball Federation; HSI, Heat Stress Index; IGF, International Golf Federation; ITF, International Tennis Federation; IIRM, International Institute for Running Medicine; Ta, ambient temperature; UIPM, Union Internationale de Pentathlon Moderne; WBGT, Wet Bulb Globe Temperature; T, True; F, False
Validation of the thermal physiology model accuracy in simulating core temperature
We validated the accuracy of the JOS-3 in simulating core temperature in 29 experiments: 11 for general exercises in chambers [45][46][47] and 18 for six sports confirmed in the previous section, using BAA. The difference between the measured and simulated temperatures averaged over the number of measurements for each case followed a normal distribution (p=0.77). Fig. 1 shows an average bias of 0.18℃ higher for the simulated values, with 95% LOA (mean±1.96 SD) ranging from -0.74 to 0.39℃. No proportional relationship was identified (p=0.07). As all but two experiments (one of laboratory exercise and one of football) fell within a range of ±0.5℃, around -0.18℃, it is judged that the JOS-3 model can be applied to validate the upper guideline thresholds if the settings correspond to the conditions of the experiments that met the accuracy requirements. The results and parameter settings for 29 experiments are shown in the SI (Supplementary Table 2–30, Supplementary Fig. 1–51). The tendency for the simulated core temperature to be higher than the measurements might be partly due to lower sweating levels in the JOS-3 than in the measured values (Supplementary Fig. 41–51).
Validation of upper thermal thresholds
The potential hazard of the upper thermal thresholds was validated according to whether the 90th, 95th, 97.5th, 99th, and 99.7th PCTL values of predicted core temperature, corrected for the systematic error from measurements (-0.18℃) exceeded the 40℃ limit. As the trend of the results is similar for any PCTL, the results for the 99.7th PCTL are explained in the text, while the results for the other PCTL are in Supplementary Fig. 52–55. The result (Fig. 2) shows a significant difference in predicted peak core temperature from below 39℃ to above 42℃ among the six sports. For football, marathons, and triathlons, the core temperature exceeded 40℃ in all calculations except in football, suggesting that the current thresholds are potentially hazardous. This is particularly true for triathlons, where core temperatures above 41℃ were predicted.
Sensitivity analyses were performed for different core temperature limits (39.4℃, 41℃) and the duration of core temperature above 40℃ (30 min).
➣ Core temperature: When the limit for core temperature was set at 41℃, the core temperature fell below the limit in marathons and football, but still exceeded in two calculations in triathlons. When the core temperature limit was set at 39.4℃, core temperatures exceeded the limit in marathons, football, and triathlons, as well as in one out of four calculations of Rugby sevens.
➣ Duration of core temperature above 40℃: If the duration of core temperature above 40℃ is allowed up to 30 minutes [22], assuming a situation where rapid cooling of the core temperature can be carried out, then elevated core temperature over time was acceptable in football, but still unacceptable in all triathlon and marathon calculations.
Based on the above set-up, to ensure that the 90/95/97.5/99/99.7th PCTL values for core temperature did not exceed 40℃ while retaining the parameter settings (other than the environmental factors), the upper thresholds for football should be revised from WBGT 32℃ to 29–31℃ or not be revised, for marathons from WBGT 28℃ to 24–27℃, and for triathlons from WBGT 32.2℃ to 23–26℃ (Fig. 3). The changes in the parameter settings of the three sports in JOS-3 and predicted core temperatures are shown in Supplementary Table 30–34 and Supplementary Fig. 56–60.