Influence of sleep and cardiovascular health on cognitive trajectories in older adults.

Abstract

Age-related changes in sleep have been associated with cognitive decline, yet causal pathways have not been identified. Evidence suggests reduced cardiovascular health may be a consequence of poor sleep and a precursor to cognitive decline. This observational cohort study used path analyses to determine whether cardiovascular disease risk mediated or moderated effects of sleep on yearly longitudinal change in cognition, estimated with linear growth models. Total sleep time (TST), sleep efficiency (SE), and relative spectral power of slow wave activity (SWA; 1-4 Hz) and slow oscillations (SO; 0.5-1 Hz), were measured with single-channel home EEG. Cardiovascular disease risk (CVR) was estimated as 10-year Framingham Risk Score 1-year post-sleep. Outcomes were yearly change in executive function (EF), episodic memory (EM), and processing speed (PS) over 2-5 years post-sleep. 342 participants (mean age 73.5 +/- 5.6 years, 51 % female) were included. Shorter TST was linearly associated with increased CVR across all models (βs = -0.18(0.058) - -0.19(0.059), ps< 0.002). TST was indirectly associated with EF and PS decline through CVR, such that associations between short TST and cognitive decline were partially due to higher CVR. All other mediating and moderating effects were nonsignificant after multiple comparisons. Indirect associations between short sleep duration and greater decline in executive function and processing speed were found through higher CVR, suggesting a potential mechanism by which sleep leads to cognitive decline. Findings support the prioritization of adequate sleep duration to preserve both cardiovascular and cognitive health in later life.