Using Predictive Modeling Technique to Assess Core Temperature Adaptations from Heart Rate, Sweat Rate, and Thermal...

Assessing the adaptation of rectal temperature (T_rec) is critical following heat acclimatization (HAz) and heat acclimation (HA) because it is associated with exercise performance and safety; however, more feasible and valid methods need to be identified. The purpose of this study was to predict adaptations in T_rec from heart rate (HR), sweat rate (SR), and thermal sensation (TS) using predictive modeling techniques. Twenty-five male endurance athletes (age, 36 ± 12 y; VO_2max, 57.5 ± 7.0 mL⋅kg^-1⋅min^-1) completed three trials consisting of 60 min running at 59.3 ± 1.7% vVO_2max in a hot environment. During trials, the highest HR and TS, SR, and T_rec at the end of trials were recorded. Following a baseline trial, participants performed HAz followed by a post-HAz trial and then completed five days HA, followed by a post-HA trial. A decision tree indicated cut-points of HR (<-13 bpm), SR (>0.3 L·h^-1), and TS (≤-0.5) to predict lower T_rec. When two or three variables met cut-points, the probability of accuracy of showing lower T_rec was 95.7%. Greater adaptations in T_rec were observed when two or three variables met cut-points (-0.71 ± 0.50 °C) compared to one (-0.13 ± 0.36 °C, p < 0.001) or zero (0.0 3 ± 0.38 °C, p < 0.001). Specificity was 0.96 when two or three variables met cut-points to predict lower T_rec. These results suggest using heart rate, sweat rate, and thermal sensation adaptations to indicate that the adaptations in T_rec is beneficial following heat adaptations, especially in field settings, as a practical and noninvasive method.