RESUMO
BACKGROUND: Given the need for disease-modifying therapies for dementia, drug repurposing appears to be a promising approach, at least as a risk reduction treatment. Preclinical studies suggest that antidepressants-in particular selective serotonin reuptake inhibitors-have beneficial effects on dementia-related biomarkers and functional outcomes, although clinical data are inconclusive. The present case-control study aimed to evaluate the effects of antidepressant drug classes and individual compounds with different treatment durations on the risk of developing dementia. METHODS: Analyses are based on data from the German Disease Analyzer database (owned and maintained by IQVIA) and included 62,317 subjects with an incident dementia diagnosis (ICD-10: F01, F03, G30, F06.7) and controls matched by age, sex, and physician between January 2013 and December 2017. Logistic regression analyses adjusting for health insurance status and comorbid diseases associated with dementia or antidepressant use were performed to investigate the association between dementia incidence and treatment with 4 major antidepressant drug classes and 14 of the most frequently prescribed individual substances. RESULTS: In 17 of 18 comparisons, long-term treatment (≥ 2 years) with any antidepressant was associated with a lower incidence of dementia than short-term treatment. Tricyclic and herbal antidepressants were associated with a decrease in dementia incidence, especially with long-term treatment. The lowest risks for dementia on an individual substance basis were identified for long-term treatment with escitalopram (odds ratio [OR] = 0.66; 95% CI, 0.50-0.89) and Hypericum perforatum (OR = 0.6; 95% CI, 0.51-0.70). CONCLUSIONS: Long-term treatment with tricyclic antidepressants, Hypericum perforatum, or escitalopram may be associated with reduced incidence of dementia. If antidepressant therapy is well tolerated, continuation-even if depressive symptoms have resolved-may be considered even beyond the purpose of relapse prevention. Future combined analyses of multinational registries and long-term clinical trials are needed to substantiate these findings.