Abstract
BACKGROUND: Data on the association between free-living daytime sunlight exposure and incident dementia are scarce. The objective is to evaluate whether the time spent in outdoor light is related to the dementia risk and to investigate whether the optimal duration varies with clinical parameters. METHODS: Data were from a prospective cohort of 362,094 UK Biobank participants. A questionnaire survey was conducted to investigate how many hours the participants spent outdoors on typical summer and winter days. A restricted cubic spline (RCS) was performed to explore the potential nonlinear relationship between sunlight exposure and the risk of dementia. We used multivariate Cox proportional hazard regression models to estimate the hazard ratios (HRs) for the associations between sunlight exposure and dementia outcomes, with the change points as a reference. RESULTS: After a median follow-up of 9.0 years, 4149 (1.15%) individuals were diagnosed with dementia. RCS showed a J-shaped relationship between time spent in outdoor light and the dementia risk, with the lowest risk at three change points (1.5 h/day on average, 2 h/day in summer, and 1 h/day in winter). Cox hazard regression models showed a marked increase in risk at low exposure (HR=1.287, 95%CI 1.094-1.515) but a relatively slow increase at higher exposure (HR=1.070, 95%CI 1.031-1.10). Results are more pronounced among participants over 60 years old, females, and those with exactly 7 h of sleep every night. CONCLUSIONS: Sunlight exposure had a J-shaped association with dementia risk. Giving detailed guidance on sunlight exposure can effectively prevent dementia.