Objective sleep characteristics and incident osteoarthritis risk: A prospective analysis of the UK biobank cohort.

Abstract

Objective: Sleep is a modifiable lifestyle factor potentially related to osteoarthritis (OA), but previous studies have mainly relied on self-reported sleep duration or quality. Using accelerometer-derived data, we examined associations of multidimensional objective sleep characteristics with incident OA.

Method: We included 61,904 UK Biobank participants who were free of OA at baseline and had accelerometer data. Six objective sleep metrics across five dimensions were derived: sleep quantity (average sleep duration), sleep quality (sleep efficiency), sleep regularity (sleep regularity index and sleep duration variability), sleep timing (sleep midpoint), and circadian rhythm strength (relative amplitude). Cox models examined associations with incident OA. Restricted cubic splines (RCS) and interaction analyses were conducted.

Results: Longer average sleep duration (HR 0.93, 95% CI 0.91-0.95), higher sleep efficiency (HR 0.95, 95% CI 0.92-0.97), and stronger relative amplitude (HR 0.93, 95% CI 0.90-0.95) were associated with lower OA risk, whereas delayed sleep midpoint was associated with higher risk (HR 1.08, 95% CI 1.05-1.11; all FDR < 0.001). RCS analyses indicated significant non-linear associations for these four metrics except sleep efficiency. Stronger circadian rhythmicity may attenuate the OA risk associated with delayed sleep midpoint. Associations were broadly consistent for weight-bearing joint OA (hip and knee), but not non-weight-bearing joint (hand) OA.

Conclusions: Delayed sleep timing and weaker circadian rhythm strength were associated with higher incident OA risk, with particularly higher risk observed when delayed sleep timing was accompanied by weaker circadian rhythmicity, extending previous evidence on sleep duration and quality by highlighting sleep timing and circadian rhythmicity.