Risk of diabetic ketoacidosis after exposure to risperidone or olanzapine

Atypical antipsychotics have been associated with metabolic abnormalities including impaired glucose metabolism, exacerbation of existing diabetes mellitus and new-onset type 2 diabetes. Not all atypical antipsychotic agents appear to have the same propensity to cause these complications.

To assess diabetic ketoacidosis risk in patients receiving risperidone or olanzapine.

California Medicaid data were evaluated for the presence of a diabetic ketoacidosis hospital claim (9th Edition of the International Classification of Diseases code 2501x) for patients receiving an atypical antipsychotic agent between July 1997 and September 2000. Initial prescription claims were identified for risperidone, olanzapine, clozapine, quetiapine and multiple atypical medications; however, the final analysis was restricted to risperidone and olanzapine owing to sample size challenges in the clozapine and quetiapine groups. Cases were specified if a claim occurred within 45 days after antipsychotic dispensation. Potential confounding variables and duration of antipsychotic exposure were included.

Initial users of risperidone (n = 51,330; 31 diabetic ketoacidosis) and olanzapine (n = 51,302; 55 diabetic ketoacidosis) were identified between July 1997 and September 2000. The adjusted risk of diabetic ketoacidosis for olanzapine versus risperidone was 1.62 (p = 0.033). The risk of diabetic ketoacidosis was associated with a longer duration of drug exposure. A progressive and statistically significant divergence in risk was observed between the two treatment groups after the first 30 days of therapy. For risperidone patients, diabetic ketoacidosis risk stabilised after the first 90 days; for olanzapine patients, diabetic ketoacidosis risk continued to increase until 360 days (study duration). For exposures of >30 days, >90 days and >180 days, diabetic ketoacidosis risk was 1.7 (p = 0.026), 2.4 (p = 0.004) and 3.5 (p = 0.001) times greater for olanzapine than risperidone. Treatment group, age, African American race and the presence of schizophrenia or diabetes were significant predictors of diabetic ketoacidosis.

The risk of diabetic ketoacidosis appears to be greater for patients exposed to olanzapine compared with risperidone after adjusting for confounding factors. This risk appears to increase with longer duration of exposure to olanzapine.