Ret 1

HR was monitored via a three‐lead electrocardiogram. Arterial blood pressure was measured non‐invasively using finger photoplethysmography (Finometer, Finapres Medical Systems). The monitoring cuff was placed around the middle finger of the right hand, with the forearm and hand supported so that the cuff was at the vertical level of the heart. Core temperature (T_c) was assessed using a commercially available rectal probe (RET‐1, Physitemp Instruments) inserted 15 cm past the sphincter muscle and connected to a thermocouple meter (TC‐2000, Sable Systems). Mean skin temperature (T_sk) from four sites (standard weightings of chest, arm, thigh, and calf, (Ramanathan, 1964)) was obtained using a wireless monitoring system (iButton^®, Maxim Integrated). Analog signals of the electrocardiogram, blood pressure waveform, and T_c were sampled at 1,000 Hz using a data acquisition unit (Powerlab 16/30, ADInstruments) and analyzed using an off‐line data analysis software (LabChart 8, ADInstruments).

Skin temperature (T_sk) was measured via skin thermocouples (Omega Engineering, Stamford, CT, USA) placed on the right anterior thigh (midway between the greater trochanter and lateral condyle), chest (midway between the axilla and areola), lateral calf (midway between the tibial condyle and malleolus), and upper arm. Mean-weighted T_sk was calculated according to Ramanathan (1964) (link): T_sk = 0.3(chest + arm) + 0.2(thigh + calf). A rectal thermister (Ret-1; Physitemp Inc., Clifton, NJ, USA) was used to measure core temperature (T_C). Mean body temperature (T_b) was calculated using T_C and T_sk(19): T_b = 0.2(T_sk) + 0.8(T_C).