Tracking the night: measuring age and sex patterns in sleep duration using wearable technology

Sleep duration is a key component of overall sleep health, but prior population-level studies characterizing this have relied on brief self-report questions (often one item) or used different objective devices within the same study. We examined the normal variation of sleep duration in an adult population using a single consumer-grade wearable device with a unified algorithm.

Retrospective cohort study conducted in the United States. Data were analyzed from 274 128 US-based adults aged 20 to 69 who used a Samsung Galaxy Watch between February 2023 and April 2023; participants were included if they had ≥20 valid weekdays and ≥8 valid weekend days of data. Sleep duration was the primary outcome, defined as the longest continuous nighttime sleep period between 6:00 pm and 6:00 am averaged over a 3-month period. Sleep duration and weekday-weekend variability were examined across age groups and by sex using descriptive statistics and independent t-tests.

Overall, average sleep duration was 7.57 h, with a 10th-90th percentile range of 6.5-8.9 h. Sleep duration was shortest in the 40-49-year-old group (7.54 h) and longest in the 60-69-year-old group (7.75 h; p < .001). Overall, 23.0 per cent of adults slept <7 h, more commonly among those aged 40-49 (25.1 per cent) and 50-59 (24.7 per cent). Across all age groups, weekend sleep was longer than weekday sleep by an average of 28 min, with the largest gap in the 40-49-year-old group (34 min) and the smallest in the 60-69 year-old group (20 min). Women consistently slept longer than men (+18 min on average) and exhibited greater between-subject variability in total sleep duration (SD = 1.61 h for women vs. 1.54 h for men).

This study demonstrates considerable variability in objectively measured sleep duration across adulthood, spanning a broad range and differing by age groups and sex. These findings provide reference distributions that may inform clinical expectations and public health messaging regarding sleep duration. Statement of Significance This study provides large-scale objective reference distributions of sleep duration among US-based adults using a single consumer-grade wearable device and unified sleep-scoring algorithm. The findings show substantial variability in sleep duration across adulthood, with the shortest sleep and greatest weekday-weekend differences observed in middle-aged adults, and consistently longer sleep among women than men. These results may help contextualize individual sleep duration estimates from consumer wearables and inform clinical and public health messaging about adult sleep duration.