AbstractBackgroundThe mechanisms linking short sleep duration to increased dementia risk remain unclear. Sleep duration has a powerful influence on blood pressure (BP), with short sleep durations increasing hypertension risk. Since high BP is a dementia risk factor, short sleep may increase risk of cognitive impairment via adverse effects on vascular health. We assessed whether the association of sleep duration with cognition and vascular brain injury was moderated by hypertensive status.Method684 dementia‐free participants (mean age 62±9 years; 53.1% female) from the community‐based Framingham Heart Study Offspring and Omni 1 cohort completed assessments of cognition, office BP, and self‐reported habitual sleep duration; 637 also completed brain MRI. Hypertensive status was defined as a combination of systolic BP ≥ 140mmHg and/or diastolic BP ≥90mmHg or use of anti‐hypertensive medication. Outcomes included the Trail Making Test B minus A, Visual Reproductions, Similarities, and white matter hyperintensity volume (WMHV) expressed as a percentage of intracranial volume. Linear regressions were performed between total sleep time and cognitive domains adjusted for demographic factors, APOE e4 positivity, use of sleeping pills, and BMI. An interaction term for hypertensive status was included in regression models.ResultShort sleep durations of ≤5 hours/night occurred in 13% of participants and 35% were hypertensive. There were no direct associations between sleep duration and any of the outcomes. However, there was a significant interaction between sleep duration and hypertensive status when predicting the outcomes of executive function/processing speed (Trail Making) and WMHV. When data were stratified by hypertensive status, shorter sleep duration was associated with poorer executive function/processing speed scores in the hypertensive group (β±SE = 0.04±0.02 per hour increase in sleep duration, p = 0.02), but not in the normotensive group (β±SE = ‐0.01±0.01, p = 0.22). Similarly, shorter sleep duration was associated with higher WMHV burden in the hypertensive group (β±SE = ‐0.13±0.06, per hour decrease in sleep duration, p = 0.03), but not in the normotensive group (β±SE = 0.01±0.04, p = 0.760).ConclusionThe combination of short sleep and hypertension is associated with worse cognitive performance and brain injury. Further work will examine this interaction more completely.