Application potential of skin temperature for sleep-wake classification

Abstract

It’s well known that skin temperature regularly fluctuates during the sleeping period, but few studies quantitatively analyzed its changed during sleep-awake transformation. In laboratories where the bedroom environment was strictly regulated, 14 subjects (7 males and 7 females) slept for 4 days periodically during 2-month period, and their electroencephalogram (EEG), electrooculogram (EOG) and electrocardiogram (ECG) were record by polysomnography (PSG), to determine the time of falling asleep and waking up. In addition, skin temperatures at seven points (forehead, chest, forearm, hand, thigh, calf and foot) were continuously measured by pyrobutton throughout the whole period. For the purpose of seeking the appropriate judgment index in sleep-wake classification, significant change (p < 0.5), large effect size (Cohen's d greater than 0.8), and relatively small individual difference (C.V < 0.5) were proposed as the selection criteria. The analysis results showed that, change value of skin temperature at single point is not recommended, even if it met partial criteria. The skin temperature gradient has potential to be used as a quantitative index for sleep-wake classification, because it could satisfy the three criteria at the same time. Skin temperature gradients between limbs and chest gradually increased before falling asleep and decreased after waking up. At the moment of falling asleep, the average skin temperature gradients between hand and chest were 1.55 °C in males and 1.84 °C in females, respectively, while between foot and chest were 1.05 °C in males and 1.58 °C in females, respectively. Although the skin temperature gradient at the stage of waking up showed a similar law to that of falling asleep, it needs to be further explored since there are more complex situations.