BACKGROUND: Associations between depression and dementia could express a causal relationship, reverse causality or be explained by health-related factors. This study explores the association of depression and indicators of depression severity with subsequent risk of dementia while ensuring temporality and adjusting for important health-related factors.

METHOD: 595,828 men from the Danish Conscription Database born in 1939-59 with register-based information on lifetime depression and covariates at age 55 years were followed in nationwide registers to identify dementia cases until 2016. Associations were analyzed using Cox proportional Hazard regression models with adjustment for intelligence, education level, body mass index, and comorbidities.

RESULTS: The dementia incidence per 1000 person-years was 1.2 cases for men without prior depression and 2.1 and 3.6 cases for men who had depression identified by antidepressants and hospitalization, respectively. Compared to no prior depression, depression identified by antidepressant medication was associated with 1.94 times ([95% confidence interval (CI) 1.81;2.07]) higher hazard of dementia and depression identified by hospitalization with depression was associated with 2.18 [95% CI: 1.95;2.45] higher hazard of dementia. Long-term course of depression identified by antidepressant prescriptions (>20 prescriptions), was associated with 40% (95% CI: 1.23;1.59) higher hazard of dementia compared to having ≤10 prescriptions.

LIMITATIONS: This study is restricted to men and dementia cases until age 57-77 years.

CONCLUSION: Men with depression before late midlife are subject to a higher risk of dementia later in life. Clinicians should be aware of dementia symptoms in patients with a long history of depression to initiate early treatment.