0890 Multiple Sleep Measures and Cognition by Race/Ethnicity and Age

Abstract

Abstract Introduction Disturbed sleep may affect cognition directly (e.g., through sleep fragmentation) and indirectly through effects on related health outcomes (e.g., cardiovascular health). However, investigations of sleep and cognition among diverse racial/ethnic samples and women are sparse. Methods We analyzed longitudinal associations among 50,000 Sister Study participants (enrollment: 2003-2009) who reported on short (< 7 hours) and long sleep duration (>9 hours), sleep debt (≥2-hour difference between longest and shortest sleep duration during the week), frequent napping (≥3 times/week), and insomnia symptoms. Participants completed an eight-item Dementia Screening Interview during two follow-up periods (2012-2014; 2017-2019). Poor cognition was quantified as a change in cognitive function indicative of impairment in ≥2 items. Adjusting for sociodemographic and health behavior characteristics, we used log-binomial regression to estimate risk ratios (RRs) and 95% confidence intervals (CIs) for associations between sleep dimensions and cognition, by race/ethnicity and age. Results Mean age was 55±8.9 years, 83% self-identified as Non-Hispanic [NH]-White, 9% NH-Black, and 5% Hispanic/Latina. Compared to NH-White women, NH-Black and Hispanic/Latina women reported higher prevalence of all poor sleep dimensions but lower prevalence of poor cognition. Participants with short and long sleep duration, sleep dept, napping, and insomnia symptoms had a higher prevalence of poor cognition (RR=1.09 [95% CI: 1.03-1.16]; RR=1.19 [1.06-1.35]; RR=1.18 [1.11-1.25]; RR=1.27 [1.18-1.39]; RR=1.33 [1.26-1.41], respectively). Estimates for short sleep with poor cognition were higher among minoritized racial/ethnic groups (RRNH-Black=1.15 [0.91-1.45]; RRHispanic/Latina=1.48 [1.09-2.03]; RRNH-White=1.08 [1.01-1.15]), although interactions terms were not statistically significant (p-value= 0.37 and 0.07, respectively). Patterns were similar for insomnia. Sleep debt and insomnia symptoms estimates with poor cognition were higher among younger than older groups (sleep debt: RRAge 26-55 years=1.20 [1.11-1.31]; RRAge 56-76 years =1.14 [1.05-1.25]; insomnia: RRAge 26-55 years=1.39 [1.27-1.51]; RRAge 56-76 years =1.27 [1.18-1.38]). Interaction terms were non-significant (p-value= 0.96 and 0.26, respectively). Conclusion Poor sleep health was associated with poorer cognition, and associations may be stronger among minoritized and younger women. Future studies using a life course perspective among large, more diverse populations can elucidate how early-life and later-life factors (e.g., psychosocial experiences; environmental conditions) accumulate to impact sleep health and contribute to differential cognitive risk in adulthood. Support (if any)