0326 Influence of Sleep and Cardiovascular Health on Cognitive Outcomes in Older Adults

Abstract

Abstract Introduction High and low sleep duration have been associated with reduced executive function (EF) and episodic memory (EM). However, causal pathways have not been identified. Cardiovascular health may contribute to the relationship between sleep and cognitive aging. Sleep disturbance has been associated with increased hypertension and heart disease in older adults, and cerebrovascular insufficiency has been consistently linked with cognitive decline. Therefore, the mediating and moderating effects of cardiovascular disease risk (CVD) on the relationship between sleep duration and longitudinal change in EF and EM were examined. Methods Statistical analyses were preregistered (osf.io/6yw7g). Average total sleep time (TST) over 2-6 nights was estimated using single-channel EEG at baseline (N=332). CVD was estimated using the abbreviated Framingham Heart Score approximately one year post-baseline. Outcome variables were linear change in EF and EM composite scores [BL1] per year up to 7 years after baseline. EF subtests measured semantic fluency, working memory, and task switching. EM subtests measured verbal associative memory, narrative episodic memory, and list learning. Covariates included age, gender, Alzheimer disease (AD) biomarkers (ratio of phosphorylated tau181 to amyloid-ß42), global cognition, and apnea-hypopnea index. Results Greater TST was linearly associated with reduced CVD (ß=-.157, p=.006). Quadratic effects of TST on EF and EM were observed (EF: ß=-.105, p=.002; EM: ß=-.203, p=.007). Lowest TST was related to greater reduction in EF; highest TST was related to greater reduction in EF and EM. Mediating effects of CVD for EF and EM were observed (ps<.05) but were non-significant when age and AD biomarkers were included (p=.10, p=.08, respectively). A non-significant trend indicated CVD moderated effects of TST on EF (ß=.10, p=.055), controlling for age and AD biomarkers. Less sleep combined with higher CVD tended to be associated with greater decline in EF. CVD did not moderate effects of TST on EM (ß=.07, p=.254). Conclusion Consistent with prior literature, CVD and TST were independently associated with cognitive decline. Although findings suggest that CVD may influence effects of TST on cognitive decline, more evidence is needed. Future analyses will examine alternative sleep measures (e.g., sleep efficiency, slow wave spectral density). Support (If Any) NIH: T32-AG000030, P01-AG03991, P01-AG026276, P30-AG066444.