New antipsychotic medications such as olanzapine and risperidone often can effectively control schizophrenic symptoms without the extrapyramidal side effects of traditional agents. However, these newer medications are associated with metabolic changes, such as weight gain, dyslipidemia, and even cardiomyopathy, which could cause significant morbidity. Koro and colleagues studied the risk of diabetes in patients taking olanzapine and risperidone.

The authors used data from the General Practice Research Database, which involved more than 400 practices in England and Wales with more than 30 million patient-years of observations. Case patients were defined as persons being treated for schizophrenia who developed diabetes within three months of the start of the study period. All new cases were verified by a search of medical records for any diagnosis of diabetes or prescription for insulin or an oral hypoglycemic agent. For each case patient, they matched six control patients by age, sex, and length of observation. All control patients were diagnosed with or being treated for schizophrenia but had no diagnosis of diabetes or record of diabetic therapy.

Of 19,637 patients being treated for schizophrenia, 451 patients with new cases of diabetes were matched with 2,696 control patients. The study population consisted of equal numbers of men and women, and the average age of participants was 51 years. Olanzapine was prescribed for 970 patients (4.9 percent) and risperidone for 1,683 patients (8.6 percent). During the follow-up period (mean, 5.2 years), 451 patients developed diabetes. The incidence of diabetes in all patients with schizophrenia who were treated with antipsychotics was 4.4 per 1,000 person-years, but it was significantly related to the type of therapy.

In patients treated with olanzapine, the rate was 10 per 1,000 person-years, and in patients treated with risperidone it was 5.4 per 1,000 person-years. Compared with no antipsychotic treatment, the adjusted odds ratio for diabetes was 5.8 in the olanzapine group and 2.2 in the risperidone group. Compared with conventional antipsychotic therapy, the adjusted odds ratio for diabetes was 4.2 for olanzapine and 1.6 for risperidone. The risks were statistically significant for olanzapine therapy but not for risperidone.

The authors discuss several mechanisms that could explain the metabolic changes that occur during olanzapine therapy. Patients gained an average of 4 kg (8 lb, 12 oz) with olanzapine and 2 kg (4 lb, 6 oz) with risperidone. The authors conclude that olanzapine has a significant and consistent association with diabetes, and they caution physicians to be alert for this development in patients undergoing long-term treatment with this agent.