Validity of the CORE wearable sensor during constant-load cycling exercise in the heat

Heat exposure is an increasing threat to human health and accurately monitoring core body temperature is vital to avoiding heat-related illnesses during exercise and outdoor occupations. Recently, a wearable device (CORE^TM) was developed to non-invasively estimate core body temperature (T_C). The purpose of this study was to determine the validity of the CORE sensor and identify subcutaneous fat as a potential contributor to the device's bias. Twenty-four healthy males (n = 13) and females (n = 11) completed 60 min of cycling exercise at 60 % of their maximum aerobic power output (166 ± 46.5 W) in the heat (35.9 °C & 20.7 % relative humidity). During exercise, T_C and skin temperature (T_SK) measures from the CORE sensor were compared to rectal probe and skin surface probe, respectively. Midaxillary subcutaneous fat thickness (MSFT) at the site of the CORE sensor was measured via ultrasonography. Throughout the entire exercise bout, the CORE displayed a non-statistically significant overestimation of T_C (0.15 ± 0.43 °C, p = 0.057). However, CORE significantly overestimated T_C during the third (0.31 ± 0.43 °C, p = 0.002) and fourth (0.36 ± 0.45 °C, p = 0.004) quarters of exercise. CORE also significantly overestimated T_SK (1.28 ± 0.96 °C, p < 0.0001) throughout the entire exercise bout. MSFT was positively correlated with the bias in T_C (overestimation by CORE) during the third (r = 0.438, p = 0.032) and fourth (r = 0.482, p = 0.017) quarters of exercise. In conclusion, compared to rectal temperature, the CORE sensor overestimated T_C during constant-load exercise in a hot environment. Overestimated T_SK and local subcutaneous fat may be potential contributors to bias in the CORE sensor's T_C estimation.