Mapping the physiological landscape of body movements during nocturnal sleep and wakefulness and their cardiovascular correlates with a wearable multi-sensor array.

Spontaneous motor activity is a physiological feature of sleep and is enhanced in sleep-related movement disorders. We aimed to develop a measurement and analysis approach to movements during nocturnal sleep and wakefulness, as well as their cardiovascular correlates, entirely based on a wearable multi-sensor array, and to test its feasibility and internal consistency. Twelve healthy participants slept overnight at home wearing an array of seven wearable sensors: five accelerometers, a photoplethysmograph, and an electrocardiograph. Sleep-wake states were determined from wrist actigraphy and sleep diaries. We developed an algorithm for detecting movements based on body segment acceleration and validated it against visual annotations of accelerometer tracings. The F1 scores ranged from 90.6% to 94.4% for different segments. Our algorithm indicated that movements during sleep were significantly fewer than during wakefulness (21.4 ± 1.5 vs. 90.4 ± 6.2 per hour, p < 0.001), with relatively more segmental (42.8% vs. 15.0%) and lower-body (15.3% vs. 5.7%) movements and relatively fewer global movements (23.7% vs. 66.1%). Heart rate started to increase and pulse wave amplitude started to decrease at, or up to 6 s before, movement onset, depending on movement type and wake-sleep state, consistent with cardiac activation and peripheral vasoconstriction due to central autonomic commands. The magnitude of these cardiovascular responses positively correlated with intensity particularly of global movements (p < 0.001). These results demonstrate the feasibility and internal consistency of a new approach, entirely based on wearables sensors, to map the physiological landscape of body movements and their cardiovascular correlates during nocturnal sleep and wakefulness.