Correlation Between Wrist Activity Monitor and Electrophysiological Measures of Sleep in a Simulated Shiftwork Environment for Younger and Older Subjects

Although several studies have examined the correlation between nocturnal PSG and activity measurement, validation studies of actigraphically measured sleep in shiftworking populations have not been reported. This study investigates the correlation between sleep recorded using EEG and actigraphic techniques during a simulated 12-hour shift rotation. Thirty-two subjects were allocated to groups according to age. Group (1) included sixteen subjects mean (+/- s.d.) age of 21.2 +/- 2.7 years, and Group (2) included sixteen subjects mean (+/- s.d.) age of 43.9 +/- 6.8 years. An adaptation night was followed by two 12-hour day shifts (7 am-7 pm), 24 hours off and then two 12-hour night shifts (7 pm-7 am). For the entire study subjects wore an activity monitor, and while in bed, sleep was recorded using polysomnography; both techniques were collected in 30-second epochs. A high epoch for epoch agreement between wrist activity monitoring and EEG measures of sleep was recorded for daytime and nighttime sleep periods (80-90%). There was a high correlation between EEG and actigraphically recorded sleep duration in young (0.98-0.77) and older (0.78-0.96) subjects for all sleep periods. Sleep efficiency correlations were extremely variable for both the young (0.72-0.15) and older (-0.18-0.58) subjects for daytime and nighttime sleep periods. Taken together these results suggest that wrist activity monitoring is a valid measure of sleep/wake activity and sleep duration, in a simulated shiftwork environment. However, some caution should be used for more specific measures, such as sleep efficiency particularly in older subjects.