Study Objective. To explore the clinical characteristics of hyperglycemia in patients treated with risperidone.

Design. Pharmacovigilance survey of spontaneously reported adverse events in risperidone-treated patients, with reports of haloperidol-associated hyperglycemia used as a control.

Setting. Government-affiliated drug evaluation center.

Intervention. The Food and Drug Administration MedWatch surveillance program was queried (risperidone, 1993-February 2002; haloperidol, late 1970s-February 2002) and results pooled with published cases.

Measurements and Main Results. We identified 131 reports of risperidone-associated hyperglycemia in addition to seven reports of patients with hyperglycemia who received combined risperidone-haloperidol therapy and six reports of acidosis that occurred in the absence of hyperglycemia. We found 13 reports of haloperidol-associated hyperglycemia and 11 reports of acidosis without hyperglycemia. Of the reports of risperidone-associated hyperglycemia (monotherapy), 78 patients had newly diagnosed hyperglycemia, 46 had exacerbated preexisting diabetes, and 7 could not be classified. Mean ± SD age was 39.8 ± 17.4 years (range 8–96 yrs). Patients with new-onset diabetes (mean ± SD age 34.8 ± 15.7 yrs) were younger than those with preexisting diabetes (mean ± SD age 48.8 ± 17.5 yrs). The overall male:female ratio was 1.5. In most patients, hyperglycemia appeared within 3 months of the start of risperidone therapy. Severity of disease ranged from mild glucose intolerance to diabetic ketoacidosis or hyperosmolar coma. Twenty-six patients with acidosis or ketosis were reported. Four patients died.

Conclusion. Atypical antipsychotic treatment may unmask or precipitate hyperglycemia. Although such cases attributed to clozapine or olanzapine are more numerous than those associated with risperidone, the number for risperidone-associated hyperglycemia is relatively higher than that observed with the conventional neuroleptic haloperidol.