STUDY OBJECTIVE: The use of benzodiazepines and the development of dementia is controversial, with studies indicating that benzodiazepines could be either a protective factor or a risk factor for dementia, or no association may exist between the two. Our objective was to identify whether such an association exists. DESIGN: Systematic review and meta-analysis of 12 prospective and retrospective cohort studies and case-control studies. PARTICIPANTS: A total of 981,133 (in the systematic review) and 980,860 (in the meta-analysis) adults or elderly individuals. MEASUREMENTS AND MAIN RESULTS: A search of the PubMed, LILACS, and Cochrane Central Register of Controlled Trials databases, as well as a manual search of the reference lists of the included publications and reviews, was performed. We included studies that reported the incidence of dementia and in ever users of benzodiazepines. Data were analyzed by using a random effects model in R software. Quality of the evidence was assessed with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) certainty ratings system. The results of the main meta-analysis suggest that benzodiazepines can be a risk factor for developing dementia (odds ratio 1.38, 95% confidence interval 1.07-1.77; I2 = 98%; 95% prediction interval 0.58-3.25; very low certainty). CONCLUSION: Our results suggest an association between the use of benzodiazepines and the development of dementia. However, the current evidence lacks the power to infer differences between the effects of Alzheimer's disease and vascular dementias, long-acting and short-acting benzodiazepines, and various exposure loads (duration and dose). Future long-term prospective cohort studies are necessary, with adequate adjustments for confounding variables, strategies to minimize reverse causality, reporting of subgroups aimed at greater homogeneity of findings, adequate statistical power to identify high-magnitude effects, and defined daily dose analyses for dose-response gradient.
STUDY OBJECTIVE: The use of benzodiazepines and the development of dementia is controversial, with studies indicating that benzodiazepines could be either a protective factor or a risk factor for dementia, or no association may exist between the two. Our objective was to identify whether such an association exists. DESIGN: Systematic review and meta-analysis of 12 prospective and retrospective cohort studies and case-control studies. PARTICIPANTS: A total of 981,133 (in the systematic review) and 980,860 (in the meta-analysis) adults or elderly individuals. MEASUREMENTS AND MAIN RESULTS: A search of the PubMed, LILACS, and Cochrane Central Register of Controlled Trials databases, as well as a manual search of the reference lists of the included publications and reviews, was performed. We included studies that reported the incidence of dementia and in ever users of benzodiazepines. Data were analyzed by using a random effects model in R software. Quality of the evidence was assessed with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) certainty ratings system. The results of the main meta-analysis suggest that benzodiazepines can be a risk factor for developing dementia (odds ratio 1.38, 95% confidence interval 1.07-1.77; I2 = 98%; 95% prediction interval 0.58-3.25; very low certainty). CONCLUSION: Our results suggest an association between the use of benzodiazepines and the development of dementia. However, the current evidence lacks the power to infer differences between the effects of Alzheimer's disease and vascular dementias, long-acting and short-acting benzodiazepines, and various exposure loads (duration and dose). Future long-term prospective cohort studies are necessary, with adequate adjustments for confounding variables, strategies to minimize reverse causality, reporting of subgroups aimed at greater homogeneity of findings, adequate statistical power to identify high-magnitude effects, and defined daily dose analyses for dose-response gradient.
Authors: Javier Santabárbara; Darren M Lipnicki; Beatriz Olaya; Beatriz Villagrasa; Patricia Gracia-García; Juan Bueno-Notivol; Antonio Lobo; Raúl López-Antón Journal: J Clin Med Date: 2020-05-06 Impact factor: 4.241