Abstract
Aims Growing evidence suggests an association between the use of sedative-hypnotic medications and risk of dementia. The aim of this study is to examine this association using a meta-analysis approach. Methods MEDLINE (PubMed) and Scopus were systematically searched for studies published in English only. The quality of studies was evaluated using the Newcastle-Ottawa scale, and an overall odds ratio was pooled using a random-effects model. Results A total of 35 articles were included in the analysis. Pooled odds ratios (ORs) for dementia from all records were (OR; 1.33, 95% CI 1.19-1.49) for benzodiazepine (BZD) combined use (Subgroup-1), (OR: 1.46, 95% CI 1.23-1.73) for short-acting BZD use (Subgroup-2), (OR: 1.72, 95% CI 1.48-1.99) for long-acting BZD use (Subgroup-3), (OR: 1.13, 95% CI 0.97-1.32) for BZDs without specification of duration of action (Subgroup-4), (OR: 1.64, 95% CI 1.13-2.38) for the combined BZDs and Z-drugs, (OR: 1.43, 95% CI 1.17-1.74) for Z-drugs only, (OR: 1.14, 95% CI 0.88-1.46) for antidepressant use, (OR: 0.97, 95% CI 0.68-1.39) for antipsychotic use and (OR: 0.98, 95% CI 0.85-1.13) for anticonvulsant use. When sensitivity analysis was performed, association between overall use of BZDs and short-acting BZDs with the increased risk of dementia disappeared after exclusion of studies that were not adjusted for age covariate (OR: 1.2, 95% CI 1.0-1.44) and (OR: 1.22, 95% CI 0.75-2.01), respectively. Adjustment for protopathic bias by introduction of a lag period showed no evidence of increased risk of dementia with the use of BZDs (Subgroup-1) (OR: 1.14, 95% CI 0.82-1.58), Z-drugs (OR: 1.29, 95% CI 0.78-2.13), and combined BZDs and Z-drugs (OR: 1.51, 95% CI 0.91-2.53). Combined use of BZDs and Z-drugs showed more positive association when only studies of non-user design were analysed (OR: 2.75, 95% CI 2.23-3.39). Conclusions All the investigated sedative-hypnotics showed no association with increased risk of dementia except for BZDs. However, the observed association with BZDs did not persist after exclusion of studies with potential reverse causation and confounding by indication. Therefore, this association needs to be assessed carefully in future research.