Benzodiazepine use, incident dementia and Alzheimer's disease, and progression to severe dementia: A nationwide Korean cohort study

BackgroundThe association between benzodiazepine (BZD) use and dementia risk remains controversial owing to conflicting observational findings and incomplete adjustment for confounders.ObjectiveTo examine the association between BZD use and incident dementia, including Alzheimer's disease (AD), and to determine whether BZD use accelerates progression to severe dementia, using nationwide claims data with detailed comorbidity adjustment.MethodsWe conducted a matched cohort study using Korean National Health Insurance Service data, comparing 132,805 BZD users with 132,805 age- and sex-matched controls. Incident dementia (AD [F00], vascular dementia [F01], and other subtypes) was defined by ≥3 diagnostic claims within one year, and severe cases by national special-registration codes. Hazard ratios (HRs) were estimated using Cox models adjusted for the Charlson Comorbidity Index (CCI) and clinical covariates.ResultsDementia incidence was 382.4/100,000 person-years in BZD users versus 280.7 in controls. BZD users had higher multimorbidity (CCI ≥7: 42.5% versus 26.5%) and psychiatric comorbidity (e.g., anxiety disorders: 42.6% versus 2.7%). After full adjustment, the HR for overall dementia attenuated from 1.43 to 1.15 (95% CI, 1.09-1.22), with similar attenuation for AD and severe dementia. The risk of progression from dementia to severe dementia was not increased among BZD users (adjusted HR, 0.98; 95% CI, 0.84-1.14), with no dose- or duration-response observed.ConclusionsBZD use showed only a modest association with incident dementia, largely attributable to the higher multimorbidity burden among users, and was not associated with accelerated progression to severe dementia.