A longitudinal study of sleep in university freshmen: facilitating and impeding factors

Establishing healthy sleeping habits is a challenge for many college students. We determined how academic schedules influenced sleep patterns across the semester, and whether these are modulated by place of residence and class start times.

A longitudinal cohort study evaluated 638 freshmen over their first 20-week semester, comprising instructional, reading, examination, and vacation weeks. Sleep duration and timing were measured daily with sleep trackers (Oura Ring 3). Time use was reported through a smartphone app.

Six hundred and thirty-eight participants (mean [standard deviation] age, 20.3 [1.3] years; females, 51.7 per cent) provided 64 642 nights of sleep data. Bedtimes were late (mean [standard deviation], 01:53 [71.76 min]). Weekday sleep duration increased after the midterm break (β = 2.74, p < .0001), arising from delayed wake times (β = 5.81, p < .0001). Pre-midterm, students woke up after class started on 42.6 per cent of days with early 08:00 classes, likely arriving late or skipping class. This percentage increased to 52.4 per cent post-midterm. Participants living on-campus (n = 357) had later bedtime and shorter weekday sleep than those off-campus (n = 281, bedtime: +37.26 min; t629 = 6.62; p < .0001; duration: -19.03 min; t629 = 6.36; p < .0001). They reported substantial in-person social and cocurricular activities on weekday nights (60 min in the 4 h before bedtime), and compositional data analysis indicated that spending more time on such activities would further delay their bedtime. Additional screen time showed the same effect for off-campus students.

During university freshman year, when students had greater freedom to determine sleep timing, they slept more by waking up later, skipping early morning classes. However, other opportunities to lengthen sleep were not taken, highlighting the value of integrating contextual information with wearable data to understand real-world sleep behavior.