Sleep disturbances and subsequent suicide risk in dementia: A nationwide cohort study.

BACKGROUND: Individuals with dementia have an elevated suicide risk. Sleep disturbances are common, but the role of depression in this association remains unclear. METHODS: This nationwide retrospective cohort study linked Taiwan's National Health Insurance Research Database with the Death Registry. Individuals aged ≥50 years with newly diagnosed dementia during 2012-2019 were followed for suicide outcomes through 2023. The index date was defined as 3 years after sleep-disturbance diagnosis or the corresponding matched date. Sleep disturbances were identified using diagnostic codes and pharmacologic records. Suicide attempts and deaths were ascertained from inpatient claims and mortality data. Cox models estimated adjusted hazard ratios (aHRs), and mediation analysis examined the potential indirect effect of depression. RESULTS: Among 85,920 individuals with dementia, including 19,982 with sleep disturbances, 2325 suicide events occurred; 645 involved individuals with sleep disturbances. Suicide incidence was higher in the sleep-disturbance group (868.3 vs 729.7 per 100,000 person-years). Sleep disturbances were associated with increased suicide risk after adjustment for sociodemographic and clinical covariates (aHR, 1.12; 95% CI, 1.02-1.23). Male sex, low urbanization, cardiometabolic comorbidities, and frequent emergency department visits were additional predictors (≥3 visits/year: aHR, 2.54; 95% CI, 2.20-2.94). The natural indirect effect through depression was not statistically significant. CONCLUSIONS: Sleep disturbances were associated with suicide risk among individuals with dementia, independent of sociodemographic and clinical covariates. This association persisted after accounting for depression, suggesting involvement of other mechanisms. Routine assessment of sleep disturbances may help identify individuals at elevated suicide risk.