Abstract MP085: Cognitive Impairment Mediates the Impact of Short Sleep Duration on Mortality in Individuals with Cardiovascular or Cerebrovascular Disease
Background: Cardiovascular disease (CVD) and cerebrovascular disease (CBV) have been associated with short sleep duration and mortality. Furthermore, short sleep duration has been associated with impaired cognition. Most studies have been limited by using self-report measures and treating sleep duration as a sole, independent predictor, thus, its role in predicting mortality is still not well-established. Hypothesis: We hypothesized that 1) short sleep duration increases the impact of CVD and CBV on mortality and 2) cognitive impairment mediates the association of short sleep duration with mortality in those with CVD or CBV. Methods: We addressed this question in the Penn State Adult Cohort, a random, general population sample of 1,741 men and women (48.7 ± 13.5 years) who were studied in the sleep laboratory and followed-up for 16.7 ± 4.6 years. CVD was defined by a history of heart disease, including hypertension or diabetes, and CBV by a history of stroke. Polysomnographic (PSG) total sleep time was classified as normal (≥ 6 hours) and short (< 6 hours) sleep duration based on the median of the cohort. All individuals underwent a comprehensive neuropsychological evaluation, including Symbol Digit Modalities Test, Trail Making Test, Benton Visual Retention Test, Thurstone Word Fluency Test, and Mini-Mental State Examination. We tested the interaction between CVD, CBV and PSG sleep duration on mortality using Cox proportional hazard models controlling for multiple potential confounders. Results: The hazard ratios (95%CI) of mortality associated with CVD and CBV were 0.9 (0.6-1.3) and 1.3 (0.5-3.1) for individuals with normal sleep duration and 1.8 (1.3-2.5) and 2.4 (1.3-4.4) for individuals with short sleep duration (P-interaction < .05). In individuals with CVD or CBV, short sleep duration was associated with impaired processing speed, executive attention, and short-term memory (all Ps < .05). Cognitive impairment significantly mediated the impact of short sleep duration on mortality in those with CVD or CBV [proportion of mediation effects were 6.5% (1.4%-18.6%), 4.5% (0.4%-14.2%), and 6.2% (1.0%-18.4%) for processing speed, executive attention and short-term memory, respectively]. Conclusions: The risk of mortality associated with CVD and CBV is significantly increased in those with short sleep duration. Although cognitive impairment significantly mediated this association, its modest effect suggests that future studies should examine other underlying mechanisms linking short sleep duration with mortality in individuals with CVD or CBV.