Hua Jin, MD

Since their introduction, the atypical antipsychotics clozapine (Clozaril), risperidone (Risperdal), olanzapine (Zyprexa), quetiapine (Seroquel) and ziprasidone (Geodon) have been widely prescribed for the management of patients with schizophrenia and other psychotic disorders or severe behavioral disturbances. This is, in part, due to emerging evidence suggesting beneficial effects on positive symptoms, negative symptoms and cognition in schizophrenia, combined with the lower propensity of the atypical antipsychotics for extrapyramidal symptoms or tardive dyskinesia than conventional antipsychotics. However, there have been case reports, retrospective studies and recently published epidemiological data suggesting that certain of these newer agents may be associated with metabolic abnormalities, including significant weight gain, hypertriglyceridemia, and new-onset type 2 diabetes mellitus (DM) or diabetic ketoacidosis (DKA). 

As the proportion of patients with psychosis and other disorders using atypical antipsychotics continues to grow, these potential metabolic adverse effects are a focus of interest, not only because of the resultant medical complications--including mortality--but also the greater cost of care exacted for the evaluation and treatment of metabolic complications and their sequelae. The concern about medication-related DM is heightened in patients with schizophrenia due to the twofold greater prevalence of type 2 DM in this group compared to the general population (Dixon et al., 2000). Moreover, the possibility that DKA may arise at times as the first presentation of DM is alarming, as this is a serious acute medical condition with significant morbidity and mortality. 

Since the first atypical antipsychotic was introduced in 1987, numerous single cases or case series have reported the potential association between atypical agents and new-onset DM or exacerbation of pre-existing DM. Early last year, we reported data based on an analysis of 45 published cases (clozapine [20 cases], olanzapine [19 cases], quetiapine [three cases] and risperidone [three cases]) with sufficient documentation to make an association between the use of an atypical antipsychotic and the development of new-onset DM (Jin et al., 2002). The mean age was 40.3 years (range=16 to 56), 87% were male, and 47% were African-American. At time of diagnosis, 63% had blood glucose values greater than 500 mg/dL, and 50% manifested no weight gain at time of presentation with DM. It should be noted that at baseline, 84% were >5% over ideal body weight, and 42% of these new-onset diabetes cases initially presented with DKA. The mean duration of atypical antipsychotic exposure prior to the development of DM or DKA was 19 weeks (range=two weeks to 124 weeks), with 14% developing DM or DKA within a month after starting the atypical antipsychotic.

Over the past two years, Elizabeth Koller, M.D., of the U.S. Food and Drug Administration, and colleagues have used data from the FDA MedWatch surveillance program pooled with published cases and abstracts to examine the association between atypical antipsychotic therapy and DM, hyperglycemia or DKA. The data for clozapine and olanzapine have been published (Koller and Doraiswamy, 2002; Koller et al., 2001), while preliminary findings for risperidone were presented at a meeting (Koller et al., 2002). Koller and colleagues identified 753 cases having DM, of which 384 cases were treated with clozapine, 237 with olanzapine and 132 with risperidone. Furthermore, among the new-onset DM patients, DKA was reported in 80 of both the olanzapine and clozapine cases and in 36 risperidone cases. Of particular concern were the reported deaths: 25 for clozapine, 15 for olanzapine and five for risperidone. Among patients who developed hyperglycemia or DM on clozapine or olanzapine, more than 60% of cases were noted within six months after starting atypical antipsychotic therapy. Although the number of individuals exposed to risperidone is significantly greater than those exposed to olanzapine, the retrospective and unsystematic nature of case reporting does not permit calculation of relative risk between agents.

Since 2001, several large retrospective studies have examined the relative risk of DM associated with different antipsychotics, primarily in patients with schizophrenia. Comparing 552 patients with schizophrenia on clozapine and 2,461 on typical agents in the Iowa Medicaid database, researchers found a relative risk of developing DM of 2.5 only among those on clozapine ages 20 to 34, but not for the cohort as a whole (Lund et al., 2001). A case control study of older diabetics (mean age=63.6 ±18.3) and controls (mean age=61.9 ±17.5) examined the odds of developing DM related to clozapine exposure (Wang et al., 2002). In this group of primarily nonpsychotic patients (only 40.3% had a psychotic disorder), neither clozapine dose nor duration of exposure (up to 176 days) was associated with increased odds of developing DM.

In 2002, three large comparative database studies were published examining DM prevalence and risk among several atypical agents and conventional antipsychotics. One analyzed health plan claims databases comprising 2.5 million individuals to report on the odds of developing DM during 12 months of exposure in psychotic patients (diagnosed by 

 codes) taking risperidone, olanzapine, clozapine, and high-potency and low-potency typical antipsychotics, compared to untreated patients (Gianfrancesco et al., 2002). Only risperidone did not demonstrate significantly increased odds for DM compared to untreated patients (odds ratio [OR]=0.88), while olanzapine (OR=3.10), clozapine (OR=7.44), high-potency typicals (OR=2.13) and low-potency typicals (OR=3.46) had significantly greater risk. 

The relationship of treatment with typical or atypical antipsychotics to the diagnosis of DM identified in patients diagnosed with schizophrenia (using ICD-9 codes) was examined in the national U.S. Department of Veterans Affairs databases (Sernyak et al., 2002). Of 38,632 patients studied, 22,648 received atypical antipsychotics, of which 94% received the same atypical during the four-month observation period of the study, with only 8.9% also receiving a prescription for a typical antipsychotic. Overall, patients on atypical antipsychotics were 9% more likely to have DM than those who received conventional antipsychotics (

=0.002). The odds ratio for developing DM was significantly increased for patients of all ages taking clozapine, olanzapine and quetiapine, but not risperidone, relative to conventional agents. It is worth noting, however, that among those under age 40, exposure to any of the atypicals increased odds of developing DM compared to those on typicals. 

A nested case-control study of 19,637 patients with schizophrenia without diabetes prior to the study period was performed using the General Practice Research Database in the United Kingdom (Koro et al., 2002). Patients taking olanzapine had significantly increased risk of developing DM compared to non-users of antipsychotics and those taking conventional antipsychotics; those on conventional antipsychotics had increased risk compared to non-users; and those on risperidone had a non-significant increased risk of developing DM compared to non-users of antipsychotics and those taking conventional antipsychotics. 

Conflicting results have been reported in prospective studies related to glucose dysregulation and atypical antipsychotics, depending upon the population studied. A placebo-controlled study evaluating the effect of olanzapine (10 mg/day) and risperidone (4 mg/day) to placebo on insulin response in normal subjects after 15 days to 17 days found an approximate 25% increase in insulin response to a prolonged hyperglycemic clamp in the olanzapine and risperidone groups (Sowell et al., 2002). There was also an approximate 18% decrease in the insulin sensitivity index in both treatment groups. These changes correlated significantly (r=0.5576, 

=0.019) with change in body mass index (BMI). In contrast, another study of oral glucose tolerance testing in age- and BMI-matched patients with schizophrenia receiving clozapine, olanzapine, risperidone and conventional antipsychotics and untreated healthy control subjects found that patients taking either clozapine or olanzapine had significant glucose elevations compared to patients taking conventional antipsychotics and healthy control subjects (Newcomer et al., 2002). Risperidone-treated patients in this study had elevations in fasting and postload glucose levels only in comparison with healthy control subjects and no differences in glucose levels compared to patients receiving conventional antipsychotics.

Despite some limitations on case series analysis and certain large sample retrospective studies, the published data do provide a sense of the relative effects of atypical antipsychotics on glucose regulation and other metabolic changes in psychotic patients. Given the difficulty in studying medication-naive patients with schizophrenia, the extent of glucose dysregulation and other metabolic changes attributable to schizophrenia itself may never be accurately known. However, numerous studies have suggested that pharmacological treatment may be associated with an increased risk of glucose dysregulation and other metabolic abnormalities that are not entirely explained by the diagnosis of schizophrenia itself. 

Clinicians are thus confronted with two major practice issues. First, whether there are any significant differences among different antipsychotics that might be associated with these metabolic effects, particularly glucose dysregulation. Second, if the use of atypical agents does have a relationship with glucose intolerance, what potential risk factors are relevant to the establishment of monitoring protocols. 

Based on results of case series analysis and studies reported, patients with African-American ethnicity, first-degree family history of DM and baseline obesity appear to be at increased risk for the development of glucose dysregulation during atypical antipsychotic therapy. Almost all age groups have the potential risk to develop new-onset diabetes associated with the usage of atypical antipsychotics, although the fact that one-fourth of the reported cases were under age 35 is quite concerning. Weight gain during treatment does not appear to be a necessary factor associated with new-onset diabetes since 50% of the cases in our review who developed new-onset DM in our study manifested no weight gain; however, over 80% of patients were overweight before starting atypical agent therapy (Jin et al., 2002), suggesting that baseline weight is a significant risk factor for DM or DKA in patients taking atypical antipsychotics. 

Although the large fraction of reported cases with DKA may illustrate a reporting bias toward interesting or extreme clinical findings, these cases obviously give a sense of the potential risk of therapy of atypical agents, a liability that must be managed by monitoring to adequately balance the risk-benefit ratio of treatment. The presence of DKA with type 2 DM is typically a marker of severe metabolic stress, with the development of ketosis occurring in those individuals with impaired insulin secretion due to decreased pancreatic insulin reserve. The absence of significant physical illness among these patients with DKA (e.g., infection) suggests that the atypical agent itself, by some unknown mechanism, may be a metabolic stressor in susceptible individuals. To our best knowledge, no research data have suggested the potential predictors of DKA in patients who developed new-onset DM during the treatment of atypical antipsychotics. 

Nevertheless, clinicians should be aware of the relatively brief time course to the development of new-onset diabetes during atypical antipsychotic treatment in some patients, that the absence of significant weight gain is not uncommon with new-onset DM, and that the first presentation may indeed be diabetic ketoacidosis, a potentially life-threatening condition. Although there are no consensus guidelines on monitoring of serum glucose in those receiving atypical antipsychotics, recommendations in the Table are based upon the fact that diabetes is overrepresented in patients with schizophrenia and that this risk may be magnified with the use of atypical agents. 

Dixon L, Weiden P, Delahanty J et al. (2000), Prevalence and correlates of diabetes in national schizophrenia samples. Schizophr Bull 26(4):903-912.
 

Gianfrancesco FD, Grogg AL, Mahmoud RA et al. (2002), Differential effects of risperidone, olanzapine, clozapine, and conventional antipsychotics on type 2 diabetes: findings from a large health plan database. J Clin Psychiatry 63(10):920-930.
 

Jin H, Meyer JM, Jeste DV (2002), Phenomenology of and risk factors for new-onset diabetes mellitus and diabetic ketoacidosis associated with atypical antipsychotics: an analysis of 45 published cases. Ann Clin Psychiatry 14(1):59-64.
 

Koller EA, Doraiswamy PM (2002), Olanzapine-associated diabetes mellitus. Pharmacotherapy 22(7):841-852.
 

Koller EA, Doraiswamy PM, Cross JT (2002), Risperidone-associated diabetes. P1-83. Presented at the 84th Annual Meeting of the Endocrine Society. San Francisco; June 19-22.
 

Koller E, Schneider B, Bennett K, Dubitsky G (2001), Clozapine-associated diabetes. Am J Med 111(9):716-723.
 

Koro CE, Fedder DO, L'Italien GJ et al. (2002), Assessment of independent effect of olanzapine and risperidone on risk of diabetes among patients with schizophrenia: population based nested case-control study. BMJ 325(7358):243.
 

Lund BC, Perry PJ, Brooks JM, Arndt S (2001), Clozapine use in patients with schizophrenia and the risk of diabetes, hyperlipidemia, and hypertension: a claims-based approach. Arch Gen Psychiatry 58(12):1172-1176.
 

Newcomer JW, Haupt DW, Fucetola R et al. (2002), Abnormalities in glucose regulation during antipsychotic treatment of schizophrenia. Arch Gen Psychiatry 59(4):337-345.
 

Sernyak MJ, Leslie DL, Alarcon RD et al. (2002), Association of diabetes mellitus with use of atypical neuroleptics in the treatment of schizophrenia. Am J Psychiatry 159(4):561-566.
 

Sowell MO, Mukhopadhyay N, Cavazzoni P et al. (2002), Hyperglycemic clamp assessment of insulin secretory responses in normal subjects treated with olanzapine, risperidone, or placebo. J Clin Endocrinol Metab 87(6):2918-2923.
 

Wang PS, Glynn RJ, Ganz DA et al. (2002), Clozapine use and risk of diabetes mellitus. J Clin Psychopharmacol 22(3):236-243.

Articles

The atypical antipsychotics have become the treatment of choice for patients with psychotic and other behavioral disorders. However, case reports, retrospective studies and epidemiological data suggest that these medications may be associated with new-onset type 2 diabetes and diabetic ketoacidosis.

Hua Jin, MD

Articles

Glucose Dysregulation