Michael Sernyak, VA Connecticut Health Care System, VISN 1 MIRECC; Doug Leslie, Northeast Program Evaluation Center, VISN 1 MIRECC; Robert Rosenheck, Northeast Program Evaluation Center, VISN 1 MIRECC

Objectives: Atypical neuroleptics such as clozapine, olanzapine, quetiapine, and risperidone have demonstrated efficacy in the treatment of schizophrenia. Their use has, in general, been associated with fewer extrapyramidal side effects and, perhaps, a decreased risk of developing tardive dyskinesia.

However, there have been reports of the development of both Type I and Type II diabetes following initiation of treatment with some of the atypical neuroleptics. To date these studies have consisted primarily of small series of patients, usually without control groups.

The objective of this study was to examine a large group of patients diagnosed with schizophrenia and treated with both typical and atypical neuroleptics to determine whether atypical prescription was associated with an increased risk of diabetes. The large number of patients with schizophrenia treated within the VA system combined with the quality of the administrative databases made this a feasible objective.

Methods: All outpatients treated within the VA with a diagnosis of schizophrenia who were prescribed a neuroleptic in the last quarter of fiscal year 1999 were included in this study. If a prescription for clozapine, olanzapine, risperidone, or quetiapine was recorded then the patient was included in the atypical group.

The administrative database was also examined for diagnoses of diabetes mellitus along with information on age race, and other demographic and relevant treatment and service utilization variables. A multiple logistic regression analysis was performed controlling for these variables to compare the frequency of the diagnosis of diabetes mellitus across age groups and different atypical neuroleptics.

Results: 30,819 veterans were studied: 12,695 received typical neuroleptics and 18,124 received atypical neuroleptics (clozapine, 935 [5.2%]; olanzapine, 8,772 [48.4%]; quetiapine, 773 [4.3%]; risperidone, 7,944 [43.8%]). Ninety-five percent of patients prescribed atypical neuroleptics were not prescribed another atypical neuroleptic during the quarter.

When patients were stratified by age, more patients were diagnosed with diabetes among those treated with atypicals when compared with those treated with typical neuroleptics for those under 40 (8.74% v. 6.21%, p=0.007), 40-49 (15.89% v. 13.93%, p=0.002), and 50-59 years old (22.73% v. 20.56%, p=0.003). There were no significant differences observed in patients from60-69 years old or those over 70.

Individual medications were also examined. Multiple logistic regression revealed that there was a significantly increased risk of being diagnosed with diabetes for patients prescribed clozapine (odds ratio [OR]=1.251, 95% confidence interval [CI]=1.070-1.462), olanzapine (OR=1.107, CI=1.038-1.180), and quetiapine (OR=1.313, CI=1.113-1.547), but not risperidone (OR=1.049, CI=0.982-1.120).

Conclusions: This study of a very large number of patients treated for schizophrenia with typical and atypical neuroleptics demonstrated an increased risk of diagnosis of diabetes associated with the prescription of atypical neuroleptics. This analysis also was able to demonstrate a differential risk associated with each of the currently available neuroleptics.

Impact: The demonstrated association of a risk between atypical prescription and diabetes suggests that this complication should be considered in both the decision to prescribe and in the subsequent monitoring of the patient.

This study also demonstrates the value of using the large patient care databases within the VA to examine pertinent clinical issues.