0855 Sleep and Circadian Rhythms in Community-Dwelling Older Adults: Differences Between Mild Cognitive Impairment and Normal Cognition

Abstract Introduction Sleep and circadian rhythms have long been proposed as early meaningful markers of cognitive decline, yet it is unclear whether patterns differ between older adults with mild cognitive impairment (MCI) and normal cognition. We therefore investigated differences in subjective sleep measures and objective sleep and circadian parameters between community-dwelling older adults with and without MCI. Methods We analyzed baseline data from 195 participants enrolled in two clinical trials (NCT03959202; NCT05064007). Subjective sleep encompassed insomnia symptoms measured by the Insomnia Severity Index (ISI) and overall sleep quality measured using the Pittsburgh Sleep Quality Index (PSQI). Objective sleep included two nights of in-home EEG (Sleep Profiler™), which measured time spent in Stage 3 Non–Rapid Eye Movement sleep (NREM3) and the number of autonomic arousals during sleep. Additionally, circadian parameters derived from ≥3 nights of actigraphy included interdaily stability (IS), reflecting rest–activity patterns over several days, and intradaily variability (IV), capturing the degree of fragmentation within each day. Montreal Cognitive Assessment scores of 18–25 indicated MCI. A multivariate analysis of covariance (MANCOVA) tested group differences across all sleep and circadian measures adjusting for age, sex, and comorbidities. Significant findings were followed by FDR-corrected ANCOVAs. Results Participants were 69.7±6.7 years old, 78.9% female, 41.5% had MCI, had 2.0±1.5 comorbidities, and 46.7% reported insomnia symptoms (ISI>8). Cognitive status was significantly associated with the combined set of six sleep and circadian measures (Wilks’ λ=0.729, F(6,68)=4.21, p=0.001, partial η²=0.271). Follow-up ANCOVAs showed that individuals with MCI had higher IS, suggesting more rigid day-to-day activity rhythms (p< 0.001, pFDR=0.002, partial η²=0.162). MCI showed reduced NREM3 time, but this was not significant after FDR correction (partial η² = .06, p = .042, qFDR = 0.088). No other measures remained significant after correction. Conclusion Our results suggest that individuals with MCI show more consistent daily circadian rhythms. However, the elevated stability may signal increased behavioral rigidity, potentially reflecting reduced lifestyle variability. Future longitudinal work is needed to determine whether these circadian features represent early behavioral changes that are linked to cognitive decline. Support (if any) The National Institute of Nursing Research (R00 NR016484) and the National Institute on Aging (R21AG078917) supported this study.