Measurements

Before the game started, the activity meter (Lifecorder GS, Suzuken Co., Ltd., Nagoya, Japan) was worn on the waist and the energy expenditure (EE) (kcal) of the physical activity during the game was measured. Using the wearable photoelectric pulse wave sensor, pulse waves during sleeping was measured at home on exercise day and on Wednesday (control day) of after 3 days too. Washout of exercise effect was 3 days. Also, the EE of control day was measured from wake-up time to bedtime, and refrained from drinking alcohol on both exercise day and control day.

Pulse wave and body movement acceleration during sleep were measured using the wearable photoelectric pulse wave sensor with built-in 3-axis acceleration sensor (APM, prototype, Suzuken Co. Ltd., Nagoya, Japan). External dimensions are W52×D46×H21 mm, mass is 46 g, pulse wave is detected with LED (peak emission wavelength 570 nm) and photodetector. Piezoresistive type was used for 3-axis acceleration sensor (dynamic range, ±3 G). Sampling frequency is 32 Hz, pulse wave can be continuously recorded for 12 hr by the lithium ion battery.

After applying complex demodulation (CDM) to the pulse wave signal (Hayano et al., 1994), pulse interval time series was calculated at 2 Hz using pulse interval demodulation which directly extracts the instantaneous pulse interval as a continuous function from the pulse wave signal (Hayano et al., 2005). Fig. 1 shows pulse interval time series, pulse wave signal and 3-axis acceleration during night sleep. The mean pulse interval (MPI) and standard deviation pulse interval (SDPI) were calculated from the pulse interval time series. In addition, as frequency analysis, amplitude of the low frequency component (LF, 0.04–0.15 Hz) corresponding to HRV, amplitude of the high frequency component (HF, 0.15–0.45 Hz) and the ratio of LF to HF (LF/HF) were calculated using CDM. All indices were calculated as calculation window for 30 min. Bedtime was defined as the time when the change of 3-axis acceleration disappeared. Wake-up time was defined as the time which the acceleration change occurred when subjects woke up.

Calculation of pulse interval time series using pulse interval demodulation and 3-axis acceleration. Measurement time is 0:35 a.m.−6:35 a.m., bed time is 1:00 a.m., wake-up time is 6:30 a.m. Upper panel: pulse interval time series. Middle panel: original pulse wave time series. Lower panel: 3-axis acceleration.