Comparative safety of antipsychotics: another look at the risk of diabetes

The association between the use of antipsychotics and diabetes mellitus (DM) is still unclear, as depicted by several conflicting reports. Our study aims to assess the risk of DM in new users of antipsychotics.

Our nested case-control study used the Quebec Health Insurance Board databases. People in the source cohort were DM-free and had initiated an antipsychotic treatment. Subjects were cohort members who initiated an antidiabetic or had a diagnosis of DM during their follow-up period. Three variables were used to assess antipsychotic exposure: the antipsychotic used (any typical, clozapine, olanzapine, quetiapine, risperidone, or more than 1 drug); the number of 30-day periods of use; and antipsychotic use at index date (current or past). A paired multivariate logistic regression model was used to calculate adjusted odds ratios.

Among the 88 467 people included in the cohort, 6109 subjects with DM were identified and were matched to 61 090 control subjects. New users of quetiapine were less likely to develop DM than new users of typical antipsychotics (OR, 0.89; 95% CI 0.81 to 0.99). The risk of DM was not statistically different across the atypical antipsychotics. A longer exposure to any antipsychotic (for each 30-day period, OR 1.009; 95% CI 1.006 to 1.011) and current use of antipsychotics (OR 1.26; 95% CI 1.17 to 1.36) were associated with DM.

These results suggest that metabolic parameters of people exposed to antipsychotics should be monitored, irrespective of the drug taken, among the drugs available at the time of analysis.