Correlation Between Estimated Thermoregulatory Responses and Pacing in Athletes During Marathon

Performance and completion rates in marathons attract considerable attention in race planning and risk management. Previous studies have explored the relationship between several parameters, including athlete age and weather, and marathon performance. In this study, for the first time, we correlate the computational estimation of core body temperature and perspiration of athletes with speed loss and completion rate in marathon races. A feature of this method is that nonlinear thermo-physiological responses, that is, core temperature change and sweating, are followed in the time domain for ambient conditions. Our computational estimation was in good agreement with group-level core temperature rise and perspiration in typical marathon races. When the ambient conditions in previous World Athletics Championships and Olympic Games were replicated, the estimated perspiration was better correlated with the marathon speed (R
²
= 0.50, $p <; 0.05$ in men) than with the ambient temperature (R
²
= 0.37, not statistically significant), which was used in conventional studies. The correlation for female athletes was better than that for male athletes. A weak correlation was observed between the completion rate and ambient conditions, as well as the thermo-physiological response in male athletes (R
²
= 0.3), whereas a strong correlation was observed in female athletes (R
²
= 0.7). This method was applied to estimate pacing in the upcoming Olympic Games in 2021 to discuss the effect of location change on marathon performance. The findings of this study may be useful for race pacing and risk management of heat-induced illness.