Does Number And Location Of Sites Affect Mean Skin Temperature Equation Validity During Exercise?

Mean skin temperature (Tsk) can fluctuate widely (~10 °C) due to environmental conditions and metabolic rate. Accurate Tsk measurement is necessary for studying and monitoring human body heat exchange and exertional heat strain. Equation complexity and sensor technology for estimating Tsk have evolved over time to maximize efficiency of fewer measured sites while accounting for known regional differences in skin temperature, of which the Ramanathan equation is widely used (1,649 citations). PURPOSE: To assess the agreement between the standard Hardy DuBois 12-site equation (HDB12) and five other common calculations of Tsk. METHODS: Ten males (age: 37 ± 12 y; BSA: 1.94 ± 0.15 m2: body fat %: 8 ± 2) cycled 60 min at 55% VO2peak in thermoneutral (20 °C, 61% relative humidity) and warm (30 °C, 63% relative humidity) conditions with temperature sensors (iButtons) taped to the head, trunk, arm, hand, thigh, leg, and foot. Agreement between HDB12 and the Hardy DuBois 7-site (HDB7), Burton, Ramanathan, Palms/Park, and Teichner equations were assessed with mean absolute error (MAE), mean bias, root mean squared error (RMSE), and limits of agreement (LOA) (alpha = 0.05). A valid equation with good agreement was determined as MAE, mean bias, or RMSE ≤0.5 °C or LOA ≤1.0 °C. RESULTS: Core temperature increased in thermoneutral (Δ1.1 °C) and warm (Δ1.5 °C) conditions. Compared to HDB12 (neutral: 30.3 ± 1.5 °C; hot: 33.4 ± 1.1 °C), all equations had good agreement in the warm condition and all but HDB7 in neutral.Agreement with HDB12 for five Tsk equations. - HDB7 Palms/Parks Teichner Ramanathan Burton Equation Seven sites: 0.7*forehead +0.14*forearm +0.05*hand +0.07*foot +0.13*shin +0.19*front thigh +0.35*stomach Six sites: 0.14*face +0.19*chest +0.19*lower back +0.32*front thigh +0.11*forearm +0.05*hand Six sites: .149*face +0.186*inner thigh +0.186*lower back +0.186*front thigh +0.186*chest +0.107*upper arm Four sites: 0.3*chest +0.3*upper arm +0.2*front thigh +0.2*lateral lower leg Three sites: 0.5*chest +0.36*lateral lower leg +0.14*forearm Thermoneutral Mean ± SD 29.5 ± 1.6 30.6 ± 1.9 30.7 ± 1.8 30.5 ± 1.8 30.4 ± 1.8 MAE 0.9 0.5 0.5 0.5 0.5 Mean Bias -0.8 0.3 0.4 0.2 0.1 RMSE 1.1 0.7 0.6 0.6 0.7 LOA 1.4 1.2 0.8 1.0 1.3 Warm Mean ± SD 32.9 ± 1.2 33.7 ± 1.1 33.7 ± 1.1 33.6 ± 1.1 33.4 ± 1.2 MAE 0.5 0.3 0.3 0.3 0.4 Mean Bias -0.5 0.3 0.3 0.1 0.0 RMSE 0.6 0.5 0.4 0.4 0.5 LOA 0.5 0.8 0.6 0.7 1.0 Units are °C, SD = standard deviation, * p ≤ 0.05. CONCLUSION: The assessed equations had improved agreement with HDB12 in warm compared to thermoneutral condition. The HDB7 is a consistently biased (~-0.5 to -0.8 °C) calculation of Tsk. Where a 12-site measurement is not feasible, the minimal equations, such as Ramanathan, may be valid substitutes; the other tested models with increased number of sites did not enhance validity.