Sleep duration, genetic susceptibility, and Alzheimer's disease: a longitudinal UK Biobank-based study

Abstract

Abstract Background Alzheimer's disease (AD) is the most frequently occurring type of dementia. Concurrently, inadequate sleep has been recognized as a public health epidemic. Notably, genetic and environmental factors are now considered contributors to AD progression. Objective To assess the association between sleep duration, genetic susceptibility, and AD. Methods and results Based on 483,507 participants from the UK Biobank (UKB) with an average follow-up of 11.3 years, there was a non-linear relationship between AD incidence and sleep duration (P for non-linear < 0.001) by restricted cubic splines (RCS). Sleep duration was categorized into short sleep duration (< 6 h/night), normal sleep duration (6–9 h/night), and long sleep duration (> 9 h/night). No statistically significant interaction was identified between sleep duration and the AD-GRS (Alzheimer's disease genetic risk score, P for interaction = 0.45) using Cox proportional risk model. Compared with the participants who had a low AD-GRS and normal sleep duration, there was associated with a higher risk of AD in participants with a low AD-GRS and long sleep duration (HR = 3.4806; 95% CI 2.0011–6.054, p < 0.001), participants with an intermediate AD-GRS and long sleep duration (HR = 2.0485; 95% CI 1.3491–3.1105, p < 0.001), participants with a high AD-GRS and normal sleep duration (HR = 1.9272; 95% CI 1.5361–2.4176, p < 0.001), and participants with a high AD-GRS and long sleep duration (HR = 5.4548; 95% CI 3.1367–9.4863, p < 0.001).In addition, there was no causal association between AD and sleep duration using Two Sample Mendelian randomization (MR). Conclusion In the UKB population, though there was no causal association between AD and sleep duration analyzed using Two Sample MR, long sleep duration (> 9 h/night) was significantly associated with a higher risk of AD, regardless of high, intermediate or low AD-GRS. Prolonged sleep duration may be one of the clinical predictors of a higher risk of AD.