Antipsychotics and diabetes: review of non-prospective data

Background Much of the existing data regarding the relationship between antipsychotic medications and diabetes mellitus are non-prospective. Aims To review critically nonprospective studies examining antipsychotic medications as risk factors for diabetes mellitus. Method Database and manual searches. Results Anecdotal reports indicate that conventional and atypical antipsychotics can cause diabetes mellitus. However, retrospective studies cannot reliably quantify this association, as they do not adequately control for confounding risk factors for diabetes, or for variation in detection and diagnosis of this illness. Most studies report a higher rate of diabetes in patients who use antipsychotic medication than in non-users, and in patients taking atypical v conventional antipsychotics. Studies assessing the relative risks of diabetes between individual atypical antipsychotics are contradictory. Conclusions Retrospective data are of limited value. Well-designed prospective studies, which account for potential confounders, are needed to investigate the true association between antipsychotic medications and diabetes. Declaration of interest P.M.H. has received lecture fees from AstraZeneca, Bristol-Myers Squibb, Eli Lilly & Co. and Janssen-Cilag.