Changes in sleep dimensions, cognitive transitions, and incident dementia: a two-cohort longitudinal study

Abstract Study Objectives This study investigated the associations between changes in sleep dimensions, cognitive transition, and incident dementia. Methods Using data from the UK Biobank (UKB) and the China Health and Retirement Longitudinal Study (CHARLS), we systematically investigated longitudinal changes in eight distinct sleep dimensions, including sleep duration, chronotype, and napping, etc. Cognitive transitions were assessed through changes in standardized cognitive test scores in the UKB and categorized cognitive states (normal cognition, mild cognitive impairment [MCI], and probable dementia) in CHARLS. We used generalized linear models for cognitive scores, logistic regression for cognitive status transitions, and Cox models for dementia risk associated with changes in sleep dimensions. Results A total of 8994 and 14 720 participants were involved in the change-to-change analyses and change-to-dementia analyses, respectively. Compared to individuals who maintained their original sleep dimensions, those who changed their sleep duration to the optimal range (7–8 h/day) or shifted chronotype to morningness exhibited higher overall cognitive scores (β = 0.15, p = .037; β = 0.23, p = .011). Conversely, transitioning to non-optimal sleep duration (OR = 1.07, p = .034) or declining overall sleep quality (OR = 1.06, p = .006) increased the risk of cognitive decline from normal baseline. Napping cessation increased the risk of MCI progression to dementia (OR = 1.16, p = .001). Transitioning to non-optimal sleep duration (HR = 1.82, p = .005) and discontinuing napping (HR = 2.13, p = .015) were associated with a higher incident of all-cause dementia. Conclusions Maintaining optimal or optimizing sleep duration, preserving napping habits, and transitioning to a morning chronotype are essential for dementia prevention.