Risk of Treatment-Emergent Diabetes Mellitus in Patients Receiving Antipsychotics

Leslie L Citrome; Richard IG Holt; Woodie M Zachry; Jerry D Clewell; Paul A Orth; Jamie L Karagianis; Vicki Poole Hoffmann


The Annals of Pharmacotherapy. 2007;41(10):1593-1603. 

In This Article

Abstract and Introduction

Background: Type 2 diabetes mellitus has been reported during antipsychotic treatment.
Objective: To quantify the potential risk of treatment-emergent diabetes mellitus among patients receiving antipsychotic medications.
Methods: The MEDLINE and Psychinfo databases were searched using the key words antipsychotic (including individual drug names), diabetes, risk, and incidence for all English-language articles published between 1966 and 2005. Risk calculations were performed using data obtained from pharmacoepidemiologic studies that met the following criteria: (1) cohort design, (2) determination of preexisting diabetes, (3) inclusion of antipsychotic monotherapy as an exposure variable, and (4) comparison with exposure to first-generation antipsychotics. Studies meeting these criteria were used to calculate incidence, attributable risk between agents, and number needed to harm.
Results: A total of 25 observational pharmacoepidemiologic studies were found comparing antipsychotics on the outcome of diabetes mellitus. Sufficient information was provided in 15 of the reports to be able to estimate attributable risk. Attributable risk for individual second-generation antipsychotics relative to first-generation antipsychotics ranged from 53 more to 46 fewer new cases of diabetes per 1000 patients. Little observable difference was noted between the individual second-generation antipsychotics versus first-generation antipsychotics on this outcome. However, few of the studies controlled for body weight, race or ethnicity, or the presence of diabetogenic medications. None adjusted for familial history of diabetes, levels of physical activity, or diet, as this information is not usually available in the databases used in harmacoepidemiologic studies.
Conclusions: Based on the published pharmacoepidemiologic reports reviewed, the avoidance of diabetes as an outcome cannot be predictably achieved with precision by choice of a second-versus a first-generation antipsychotic. Risk management for new-onset diabetes requires the assessment of established risk factors such as family history, advancing age, non-white ethnicity, diet, central obesity, and level of physical activity.

Second-generation (atypical) antipsychotics have become widely used in the treatment of schizophrenia and bipolar mania. The principal safety feature distinguishing them from the older neuroleptics has been their lower propensity to produce extrapyramidal adverse events.[1] However, a therapeutic controversy exists about the potential for metabolic adverse events during treatment with second-generation antipsychotics, in particular, treatment-emergent type 2 diabetes mellitus.

A higher than expected prevalence of diabetes had been noted among patients with schizophrenia well before the advent of effective pharmacologic treatment.[2] However, the results of some studies have led to the hypothesis that treatment may play a role in the development of diabetes, perhaps with the occurrence of weight gain in patients with an underlying degree of insulin resistance.[3,4] Prospective controlled studies have demonstrated the occurrence of glucose abnormalities among patients receiving some atypical antipsychotics. For example, in patients with treatment-resistant schizophrenia who were randomized to receive clozapine, olanzapine, risperidone, and haloperidol, an increase in mean glucose blood levels over time reached statistical significance in the clozapine (p < 0.01) and haloperidol (p < 0.03) groups after 8 weeks and in the olanzapine (p < 0.02) group after 14 weeks.[5] Fourteen of the 101 patients developed abnormal glucose levels (>125 mg/dL) at some point during randomized treatment: 6 received clozapine, 4 received olanzapine, 3 were given risperidone, and 1 received haloperidol.

Diabetes mellitus may take years to manifest itself in vulnerable individuals. In the absence of data from long-term prospective trials, a number of retrospective pharmacoepidemiologic studies have been performed examining the potential association between exposure to antipsychotics and the occurrence of diabetes mellitus. In general, the findings suggest that there may be an increased risk of diabetes during treatment with antipsychotics compared with the general population and that exposure to second-generation antipsychotics may pose a higher risk than exposure to first-generation (typical) antipsychotics.[4] An influential consensus panel report concluded that there might be differences among the various second-generation antipsychotics in terms of their association with diabetes mellitus, with clozapine and olanzapine evidencing risk, and with discrepant data for risperidone and quetiapine.[6] However, a Food and Drug Administration review of the data concluded that there was insufficient evidence to differentiate risks of diabetes among the second-generation antipsychotics, resulting in a label change with similar wording for all.[7]

We examined safety in terms of a patient's risk for developing diabetes mellitus in the course of treatment with antipsychotics. Incidence of new-onset diabetes mellitus, attributable risk between antipsychotic agents, and number needed to harm (NNH) were calculated using data obtained from published pharmacoepidemiologic studies.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: