INTRODUCTION: Few longitudinal studies assessed whether sleep disturbances are associated with dementia risk.

METHODS: Sleep disturbances were assessed in three population-based studies (H70 study and Kungsholmen Project [Sweden]; Cardiovascular Risk Factors, Aging and Dementia study [Finland]). Late-life baseline analyses (3-10 years follow-up) used all three studies (N = 1446). Baseline ages ≈ 70 years (Cardiovascular Risk Factors, Aging and Dementia, H70), and ≈84 years (Kungsholmen Project). Midlife baseline (age ≈ 50 years) analyses used Cardiovascular Risk Factors, Aging and Dementia (21 and 32 years follow-up) (N = 1407).

RESULTS: Midlife insomnia (fully adjusted hazard ratio = 1.24, 95% confidence interval = 1.02-1.50) and late-life terminal insomnia (fully adjusted odds ratio = 1.94, 95% confidence interval = 1.08-3.49) were associated with a higher dementia risk. Late-life long sleep duration (>9 hours) was also associated with an increased dementia risk (adjusted odds ratio = 3.98, 95% confidence interval = 1.87-8.48).

DISCUSSION: Midlife insomnia and late-life terminal insomnia or long sleep duration were associated with a higher late-life dementia risk.