Abstract
Although previous studies have assessed the effect of sleep traits on osteoarthritis (OA) risk, the association between the complex interplay of multiple sleep patterns and OA risk remains uncertain. We included participants who were free of OA at baseline based the UK Biobank. We evaluated the associations of five sleep behaviors with the risk of OA using Cox proportional hazard regression models. To explore the metabolic profile of sleep patterns, we regressed the sleep score on 167 standardized metabolites using ten iterations of LASSO model with ten-fold cross-validation. Restricted cubic splines (RCS) with four knots were used in the fully adjusted model to explore the potential non-linear association of sleep score and metabolic profile with OA risk. We discovered that individuals with poor sleep patterns experienced a notably higher incidence of OA (HR, 1.23, 95% CI, 1.18 to 1.28, P = 2.69 × 10(-23)). Furthermore, the risk of hand OA specifically was 1.29 times higher among those with poor sleep patterns compared to those with healthy sleep patterns (HR, 1.29, 95% CI, 1.12 to 1.49, P = 4.23 × 10(-4)). Individuals belonging to the highest quintile of metabolic score exhibited a 1.14-fold elevated risk of OA compared to those in the lowest quintile (HR, 1.14; 95% CI, 1.08 to 1.20; P = 3.09 × 10(-6)). Our findings have important public health implications as we provide an objective and more comprehensive evaluation of sleep patterns, and novel insights into the mechanisms linking sleep patterns and OA through metabolic profile.