Sleep, metabolites, and global cognition: a mediation analysis of plasma metabolic profiles in the West China Health and Aging Cohort

BACKGROUND: Growing evidence connects sleep disorders with cognitive impairment in older adults, though underlying biological mechanisms remain unclear. We investigated how plasma metabolic profiles mediate the relationship between sleep patterns and global cognitive function. METHODS: In a cross-sectional analysis of 3888 participants aged ≥60 years from the West China Health and Aging Cohort, we analyzed associations between 221 metabolites with both sleep characteristics and cognitive outcomes (assessed via Mini-Mental State Examination), evaluated mediating roles of individual metabolites and composite metabolite scores, and assessed joint associations of genetic susceptibility, sleep patterns, and metabolic profiles with cognitive function. Analyses were adjusted for demographic, lifestyle, comorbidity, and medication factors. RESULTS: We identified 93 metabolites associated with sleep characteristics and 26 linked to cognitive outcomes. Four individual metabolites, including ketoleucine, dodecanoic acid, ribonic acid, and ortho-hydroxyphenylacetic acid, mediated 1.67%-3.25% of the sleep-cognition associations. Composite metabolite scores, both unrestricted and those restricted to branched-chain amino acid (BCAA) pathways, demonstrated stronger mediation effects, with proportions reaching up to 13.6%. Participants with concurrent exposure to poor sleep, high genetic risk, and adverse metabolic profiles showed significantly worse cognitive outcomes compared to those without these risk factors, with effect sizes exceeding those in single or dual exposure groups. CONCLUSIONS: Plasma metabolic signatures, particularly those involving BCAA metabolism, may serve as biological intermediaries linking poor sleep to worse cognitive function in older adults and could help identify populations at an elevated risk of cognitive impairment. Longitudinal and experimental studies are required to validate these associations and develop metabolic-based strategies for cognitive preservation.