Longitudinal actigraphy study on sleep patterns under reduced social restrictions in Japanese university students

Sleep deprivation and irregular sleep patterns can adversely affect physical and mental health. The COVID-19 pandemic presented a naturalistic opportunity to examine how reduced social time restrictions influence sleep behavior. This study aimed to investigate both group-level and individual-level changes in sleep patterns among Japanese university students before and during the pandemic and to explore how individual characteristics may contribute to these changes.

Twenty-two female university students wore waist-worn actigraphy devices for approximately 16 weeks in both 2019 and 2020. Objective sleep data were collected alongside questionnaire assessments of chronotype, personality traits, and subjective sleep feeling.

In total, 4,432 valid days of actigraphy data were analyzed. Compared with the pre-pandemic year, sleep timing was delayed by approximately 20 min for bed-in time and 40 min for bed-out time in 2020. Time in bed (TIB) increased by about 20 min, while total sleep time (TST) remained largely unchanged. Sleep efficiency declined, but subjective sleep feeling remained stable. Individual-level analyses revealed substantial variability: 9 of 22 participants showed significant changes in TST, with both increases and decreases observed. Increased TIB was associated with later bed-out time, shorter baseline sleep duration, and lower neuroticism. A later bed-in time was associated with reduced TST.

These findings suggest that while social time restrictions can influence sleep timing and duration, the effects vary considerably across individuals. Earlier bedtimes may be more effective than simply extending TIB in promoting adequate sleep. Furthermore, individual characteristics such as personality traits may play a role in sleep adaptation under changing social contexts. Given the diversity of responses observed, both group- and individual-level perspectives are essential for understanding sleep behavior in real-world settings.