Bidirectional associations between sleep and physical activity investigated using large-scale objective monitoring data

Background: The extent to which people routinely co-attain recommended sleep and physical activity levels, as well as bidirectional associations between both health behaviours, are poorly understood at the global level. This study aimed to describe the routine co-attainment of adequate sleep and daily step count thresholds and investigate non-linear associations between objective sleep and daily step count in a large multi-national sample of objective health monitoring data. Methods: Data were collected from 70,963 users of two consumer-available health devices—an under-mattress sleep sensor and wrist-worn health tracker—between January 2020 and September 2023. Generalised additive models were used to investigate potentially non-linear, bidirectional, exposure-response relationships between sleep parameters (sleep duration, sleep efficiency, and sleep onset latency) and step count at the next-day/night level. Subgroup analyses were undertaken to investigate age-related differences in all effects. Results: We show that only 12.9% of people achieve the recommended sleep duration of 7-9hrs/night and >8,000 steps/day, with 16.5% having short sleep (<7hrs/night) and sedentary lives (<5,000 steps/day). Approximately 6hrs sleep equates to the greatest next-day step count (e.g., +339 steps vs 8 hrs/night), and sleep efficiency positively predicts next-day step count in a dose-dependent manner (25th vs 75th percentile: +282 steps/day). Sleep appears largely unaffected by previous-day step count. Effects are similar across age groups but decline in magnitude when adjusted for ‘awake duration’. Conclusions: Our findings provide insight into the bidirectional relationship between sleep-activity globally and highlight the need to ensure sleep and activity health recommendations are mutually attainable.