Anecdotal reports suggest that the first night of sleep monitoring using a wrist-actigraphy monitor may result in impaired sleep when compared to subsequent nights, due to increased levels of anxiety and awareness of being monitored. This phenomenon has been seen in sleep laboratories with polysomnographic monitoring. However, this is yet to be established for wrist actigraphy monitoring in the research literature. A total of 240 healthy adult participants (177 male, 63 female; age range, 18-35 years) had their sleep monitored using wrist actigraphy over a period of five nights of 'normal' sleep (1,200 nights of data). Sleep variables including sleep latency, wake episodes, wake after sleep onset, awakenings per hour, time in bed, total sleep time and sleep efficiency % were evaluated across all nights of sleep. Comparisons were made using repeated measures ANOVAs, mean differences, range of mean differences, intraclass correlation coefficients (ICC) and Cohen's d effect sizes. There were no significant differences between the first night and subsequent nights for any of the measured sleep variables (p < .05) and all differences were associated with trivial effect sizes (d < 0.2). ICCs ranged from 0.35 to 0.62 (low to moderate). Despite claims of impaired sleep during the first night of sleep monitoring, our results indicate that a familiarization period may not be necessary when monitoring sleep in healthy participants using wrist actigraphy. However, the response is highly individual and further research is required to assess personality traits and responses to sleep monitoring.
Keywords: PSG; accuracy; anxiety; latency; reliability.
© 2020 European Sleep Research Society.