Association Between Psychological Distress and Incident Dementia in a Population-Based Cohort in Finland

IMPORTANCE: Symptoms of psychological distress have shown association with subsequent dementia, but the nature of association remains unclear. OBJECTIVE: To examine the association of psychological distress with etiological risk of dementia and incidence of dementia in presence of competing risk of death. DESIGN, SETTING, AND PARTICIPANTS: This cohort study consisted of population-based cross-sectional National FINRISK Study surveys collected in 1972, 1977, 1982, 1987, 1992, 1997, 2002, and 2007 in Finland with register-based follow-up; and the cohort was linked to Finnish Health Register data for dementia and mortality for each participant until December 31, 2017. Participants included individuals without dementia who had complete exposure data. Data were analyzed from May 2019 to April 2022. EXPOSURES: Self-reported symptoms of psychological distress: stress (more than other people), depressive mood, exhaustion, and nervousness (often, sometimes, never). MAIN OUTCOMES AND MEASURES: Incident all-cause dementia, ascertained through linkage to national health registers. Poisson cause-specific hazard model (emphasizing etiological risk) and Fine–Gray subdistribution hazard model (emphasizing effect on incidence) considering dementia and death without dementia as competing risks. Covariates of age, sex, baseline year, follow-up time, educational level, body mass index, smoking, diabetes, systolic blood pressure, cholesterol, and physical activity. Sensitivity analysis was performed to reduce reverse causation bias by excluding individuals with follow-up less than 10 years. RESULTS: Among 67 688 participants (34 968 [51.7%] women; age range, 25 to 74 years; mean [SD] age, 45.4 years), 7935 received a diagnosis of dementia over a mean follow-up of 25.4 years (range, 10 to 45 years). Psychological distress was significantly associated with all-cause dementia in a multivariable Poisson model, with incidence rate ratios from 1.17 (95% CI, 1.08-1.26) for exhaustion to 1.24 (95% CI, 1.11-1.38) for stress, and remained significant in sensitivity analyses. A Fine–Gray model showed significant associations (with hazard ratios from 1.08 [95% CI, 1.01-1.17] for exhaustion to 1.12 [95% CI, 1.00-1.25] for stress) for symptoms other than depressive mood (hazard ratio, 1.08 [95% CI, 0.98-1.20]). All the symptoms showed significant associations with competing risk of death in both models. CONCLUSIONS AND RELEVANCE: In this cohort study, psychological distress symptoms were significantly associated with increased risk of all-cause dementia in the model emphasizing etiological risk. Associations with real incidence of dementia were diminished by the competing risk of death.