Sleep variability and risk of type 2 diabetes mellitus: Mediating role of social participation in a national longitudinal cohort study.

Night to night variability in sleep duration may contribute to insulin resistance through circadian misalignment, altered autonomic and hormonal stress responses, and the co-occurrence of less favorable behaviors such as lower physical activity and poorer diet. Social participation is a potentially modifiable aspect of social connectedness that may shape daily routines and stress coping. We therefore examined the association between sleep variability and incident type 2 diabetes mellitus (T2DM) and tested whether social participation mediates part of this association. We followed 8864 China Health and Retirement Longitudinal Study participants aged ≥45 years and free of T2DM in 2011. Sleep variability was the coefficient of variation of self-reported nightly hours across 2011 to 2018 waves. Social participation summed 9 community activities. Incident T2DM (2013-2020) was identified by self-report, medication use, or fasting glucose. Multivariable Cox models estimated hazard ratios; mediation was examined with bootstrap procedures. During 66,025 person-years, 1042 T2DM cases occurred. Compared with the lowest sleep-variability quartile, the highest quartile had a 78 % higher T2DM risk (hazard ratio = 1.78, 95 % confidence interval 1.62-1.97) after adjusting for demographics, socioeconomic status, lifestyle factors, body mass index, and depressive symptoms. Greater social participation correlated with lower sleep variability and might be linked to lower T2DM incidence, mediating 22.3 % of the total effect (P < .01). Night to night sleep irregularity was associated with a higher risk of T2DM, and social participation explained a modest proportion of this association. Interventions promoting regular sleep schedules and community involvement may offer complementary strategies to prevent diabetes in aging Chinese adults.