BACKGROUND: Drugs with anticholinergic (AC) properties are frequently prescribed in the elderly for several conditions, including depression. Previous cohort and case-control studies among individuals with depression reported a higher risk of dementia among users of AC antidepressants compared with non-AC antidepressants. However, these studies did not generate measures of absolute risk under these treatment strategies.
OBJECTIVES: To estimate the 15-year risk of dementia among AC antidepressant initiators vs. non-AC antidepressant initiators vs. antidepressant non-initiators.
METHODS: We used the UK CPRD database and emulated a target trial monthly from 2001 until 2017 with three arms: AC antidepressant initiators, non-AC antidepressant initiators, and antidepressant non-initiators. Patients aged ≥ 50 years, with a history of depression, with no history of dementia, and no antidepressant prescription within the past year were eligible and contributed to as many trials as they were eligible for. Patients were followed from baseline until the earliest of a dementia diagnosis, death, loss to follow-up, or 31 December 2017. We used pooled logistic regression to estimate the intention-to-treat effect of AC antidepressant initiation, non-AC antidepressant initiation, and no antidepressant initiation on incident dementia via absolute risks, adjusting for baseline confounders.
RESULTS: Among 1,281,714 eligible individuals, 77,365 initiated AC antidepressants, 36,984 initiated non-AC antidepressants, and 1,167,365 did not initiate any antidepressant. Over the 15 years of follow-up (median follow-up of 4.3 years), 729 individuals were diagnosed with dementia in the AC antidepressant initiator group, 417 in the non-AC antidepressant initiator group, and 16,613 in the non-initiator group. The risk of dementia (95% confidence intervals) after 15 years of follow-up was 8.1% (7.1%-8.2%) for AC-antidepressants initiators, 6.9% (5.9%-7.5%) for non-AC antidepressants initiators, and 6.0% (5.5%-6.3%) for non-initiators. The estimated 15-year risk differences for dementia were 2.1% (1.4%;2.4%) comparing AC-antidepressant initiators vs non-initiators, 0.9% (0.3%;1.9%) comparing non-AC antidepressant initiators vs non-initiators, and 1.2% (-0.2%;1.9%) comparing AC antidepressant initiators vs non-AC initiators.
CONCLUSIONS: We estimated that the excess risk associated with AC properties of antidepressants on the 15-years risk of dementia was 1.2%. Our results are compatible with a 15-years excess risk that can go from 0.2 fewer cases to 1.9 extra cases per 100 individuals.
