0392 Insomnia with Short Sleep Duration Is Associated with Heart Disease and Stroke: Evidence from the UK Biobank Cohort

Abstract Introduction Evidence on the association of insomnia short sleep duration phenotype (ISSD) with heart disease and cerebrovascular disease (CBVD) is limited and based on cohorts using polysomnography. The aim of this study is to assess the risk of ISSD with incident heart disease and CBVD including stroke in the large UK biobank cohort based on self-reported insomnia and objectively measured habitual sleep duration by accelerometer. Methods We identified 96219 participants from the UK Biobank who underwent a 1-week objective sleep measure during 2013-2016. Short sleep duration was defined as ≤ 7.3 hours/night (i.e., median), after excluding participants who slept ≤ 3 hours/night or ≥11 hours/night. Self-reported insomnia symptoms obtained ± 1-year within the objective sleep measure, which was available from 4156 participants, was used to define insomnia status. Participants reported “Usually” having insomnia symptoms (vs. Never/Rarely/Sometimes) were categorized as having insomnia. Four sleep phenotypes based on the presence of short sleep duration (Yes/No) and/or insomnia symptoms (Yes/No) were created. Major confounding factors, including age, sex, BMI, smoking, and alcohol intake, were extracted from the study visit closest to the objective sleep measure. To evaluate the associations between the sleep phenotypes with incident heart and CBVD, multivariable-adjusted Cox proportional hazards models were used. The effective sample sizes for analysis on heart and cerebrovascular diseases were 3630 and 4064, respectively. Results Compared to normal sleepers with normal sleep duration, only ISSD phenotype was associated with significant risk for heart disease (HR=1.77, 95%CI=1.25-2.49, P=0.001) whereas neither the insomnia normal sleep duration group nor the normal sleepers short sleep duration group were associated with increased risk for heart disease. Similarly, compared to normal sleepers with normal sleep duration, only the ISSD group was associated with increased risk for CBVD including stroke (HR=2.18, 95%CI=1.21-3.92, P=0.009) whereas there was no association between the other two sleep groups and CBVD. Conclusion These data from the large UK biobank database using an ecologically friendly method to assess habitual objective sleep duration suggest that it is the combination of insomnia plus short sleep duration (ISSD), that increase significantly the risk for heart disease and stroke. Support (if any)