Introduction: A strong predictor of stroke is abnormal sleep duration with both short and long sleep associated with increased stroke risk.
Hypothesis: We tested the hypothesis that the stroke-sleep association differs based on race and age.
Methods: Using a hypertensive subset from the NHIS dataset (2004-2013), we assessed the association between stroke prevalence and self-reported sleep duration, stratifying for race and age. Diagnosis of hypertension and stroke was based on self-report. Sleep duration was also self-reported and categorized as short (6 or fewer hours), normal/referent (7-8 h) or long (9 or more hours per day). Race was self-reported and categorized as White or Black. Age at screening was grouped into categories of 18-34, 45-64, and those 65 years and older.
Results: Of the 403, 621 patients in the NHIS dataset (2004-2013), 50.5 % (n=203794) had a diagnosis of hypertension. The average age of the cohort was 58.1 years (95% CI=57.8-58.4), 50.2% female; 15.4 % of the cohort was Black. The prevalence of abnormally short and long sleep duration was 31.3 % and 11%, respectively. The age-standardized prevalence of stroke amongst the hypertensive population for normal, short and long sleep duration was 3.5%, 4.9%, and 7.6%, respectively. Both short and long sleep duration were significant predictors of stroke (1.17; 95% confidence interval (CI) 1.06 - 1.31 and 2.51; 95% CI 2.44-2.58, respectively). An interaction term added to the model suggested that race and age modified the relationship between sleep duration and stroke (p=0.01). Long sleep was a significant predictor of stroke in all age groups, regardless of race. Short sleep was not a significant predictor of stroke in Whites. Short sleep predicted stroke in the youngest age group of Blacks (1.95; 95% CI 1.01 - 3.74). Conversely, short sleep duration was associated with a lower risk of stroke in the oldest age group of Blacks (0.78; 95% CI 0.58- 1.05).
Conclusions: Risk of stroke is predicted by abnormal sleep duration. The novel finding is that in this hypertensive cohort, the association between stroke and abnormal sleep duration differs by race and age.