A Heart Rate Based Algorithm to Estimate Core Temperature Responses in Elite Athletes Exercising in the Heat

Non-invasive non-obtrusive continuous and real-time monitoring of core temperature (Tc) may enhance pacing strategies, the efficacy of heat mitigation measures, and early identification of athletes at risk for heat-related disorders. The Estimated Core Temperature (ECTemp™) algorithm uses sequential heart rate (HR) values to predict Tc. We examined the validity of ECTemp™ among elite athletes exercising in the heat.101 elite athletes performed an exercise test in simulated hot and humid environmental conditions (ambient temperature: 31.6 ± 1.0°C, relative humidity: 74 ± 5%). Tc was continuously measured using a validated ingestible telemetric temperature capsule system. In addition, HR was continuously measured and used to compute the estimated core temperature (Tc-est) using the ECTemp™ algorithm.Athletes exercised for 44 ± 10 min and n = 5,025 readouts of Tc (range: 35.8-40.4°C), HR (range: 45-207 bpm), and Tc-est (range: 36.7-39.9°C) were collected. Tc-est demonstrated a small yet significant bias of 0.15 ± 0.29°C (p < 0.001) compared to Tc, with a limit of agreement of ±0.45°C and a root mean square error of 0.35 ± 0.18°C. Utilizing the ECTemp™ algorithm as a diagnostic test resulted in a fair to excellent sensitivity (73-96%) and specificity (72-93%) for Tc-est thresholds between 37.75 and 38.75°C, but a low to very-low sensitivity (50-0%) for Tc-est thresholds >39.0°C, due to a high prevalence of false-negative observations.ECTemp™ provides a valuable and representative indication of thermal strain in the low- to mid-range of Tc values observed during exercise in the heat. It may, therefore, be a useful non-invasive and non-obtrusive tool to inform athletes and coaches about the estimated core temperature during controlled hyperthermia heat acclimation protocols. However, the ECTemp™ algorithm, in its current form, should not solely be used to identify athletes at risk for heat-related disorders due to low sensitivity and high false-negative rate in the upper end of the Tc spectrum.