Sleep Duration and Long-Term Mortality After Stroke: A Nationwide Analysis

Sara Hassani,Bruce Ovbiagele,Daniela Markovic,Amytis Towfighi
DOI: https://doi.org/10.1101/2023.12.11.23299835
2024-11-01
Abstract:Background and Objectives-Sleep duration is a key marker of ideal cardiovascular health in the American Heart Association's new Life's Essential 8 construct, but studies on its relation to stroke are scarce. In this cross-sectional study, we assessed the association between sleep duration and mortality among self-reported stroke survivors in a nationally representative U.S. sample. Methods-Cross-sectional data (2005–2014) from the National Health and Nutrition Examination Surveys database in patients aged ≥18 (n=30,295) with a self-reported history of stroke (n=1,045) were linked to the 2015 National Death Index to determine the association between nightly sleep duration and mortality. Relationships between sleep duration (short: <7 hours, normal: 7-8 hours, long: >8 hours) and demographic characteristics were assessed. Relationships between sleep duration and mortality (all-cause and cardiovascular causes) were evaluated adjusting for demographic and clinical variables, with multivariable Cox regressions. Results-Among stroke survivors, prevalence of short sleep duration increased with younger age while prevalence of long sleep duration increased with older age (p<0.001). There were no significant sex differences in sleep duration. Prevalence of short sleep duration was higher among Hispanic and Non-Hispanic Black individuals compared to White individuals (p<0.0001). Stroke survivors with long sleep duration (vs. normal) were more likely to be older, less educated, and overweight, and those with short sleep duration (vs. normal) were more likely to be younger, non-White, and obese (age p<0.001, race/ethnicity p<0.001, education p=0.02, BMI p=0.001). Long sleep duration was associated with higher all-cause mortality in the unadjusted model (HR 1.92, 1.48-2.51, p<0.0001). After adjusting for demographic variables, this effect was attenuated (HR 1.35, 1.04-1.75, p=0.027). Once comorbidities were added to the model, the association was no longer significant. There were no significant associations between short sleep duration and all-cause mortality, and no significant associations between sleep duration and cardiovascular mortality. Sensitivity analyses further verified the reliability of these conclusions. Conclusion-There is no independent association between sleep duration and all-cause or cardiovascular mortality after stroke.
Cardiovascular Medicine
What problem does this paper attempt to address?