In Clinical and Experimental Therapeutics

Current Issue
Previous Issues
Reprint Information
Back to The Journal of Applied Research
Click here for information on how to order reprints of this article.

Risk of Diabetes for Individuals with Schizophrenia Treated with Antipsychotics


Zhongyun Zhao, PhD

Sandra L. Tunis, PhD

Danielle L. Loosbrock, MHA


Outcomes Research, Eli Lilly & Company, Indianapolis, Indiana


Eli Lilly and Company provided financial support.


This manuscript was based on a presentation presented at ISPOR 7th Annual International Meeting, May 19-22, 2002, Arlington, VA


KEY WORDS: schizophrenia, antipsychotics, diabetes


There is increasing interest in examining the possible association between the use of antipsychotics and the risk of diabetes, which has been suggested by recently published reports. This issue is important to explore given that, compared with the general population, those with schizophrenia appear to be already at higher risk for the development of diabetes. The objective of this analysis was to assess the incidence of diabetes for individuals with schizophrenia treated with antipsychotics using integrated medical and prescription claims in a large commercially insured population. The findings indicate that the incidence of diabetes was not significantly different for individuals receiving atypical compared to typical antipsychotics. However, the risperidone treatment group had a higher incidence of diabetes than the typical or the olanzapine treatment groups.


Antipsychotic Treatment

Schizophrenia is a complex illness, often presenting a variety of treatment challenges.1,2 Antipsychotic medications play a critical role in controlling acute psychotic symptoms, and in preventing relapse.3-5 Although the first generation (or typical) antipsychotics are generally effective in controlling positive symptoms of schizophrenia, they are less effective in treating negative and cognitive symptoms. Additionally, they are associated with a broad spectrum of adverse events including extrapyramidal symptoms (EPS) and tardive dykinesia.6,7

Compared with the typical agents, newer generation (or atypical) antipsychotics have shown advantages in controlling negative symptoms, and in overall tolerability (decreased incidence of EPS). The atypical agents are widely considered to represent an important pharmacological advance in the treatment of schizophrenia.8-10 

Risk of Diabetes

A number of recent reports in the psychiatric literature have suggested that the use of atypical antipsychotic medications may be associated with an increased risk of developing diabetes mellitus.11-21 The possibility of such an association is important to explore given that compared to the general population, those with schizophrenia appear to be already at increased risk for developing diabetes.22-24 For example, Dixon et al.23 reported that the rate of diabetes among individuals with schizophrenia, aged 18 to 44 years (5.6%-6.7%), was substantially higher compared to that of the general US population age 20 to 39 years (1.1%).25

Several retrospective database studies have also been conducted to examine this important issue. Currently published data come from studies having diverse goals, methodologies, patient populations and limitations. Some studies focus on the risk of diabetes during treatment with antipsychotics in schizophrenia patients, while others examine the issue among all users of antipsychotics regardless of diagnosis. Some compare specific atypical groups to the use of typical antipsychotics as a class. Others include comparisons among two or more atypical antipsychotic groups. Not surprisingly, reported results and conclusions have been varied and even inconsistent. Although several studies have provided early information regarding antipsychotic use and the risk of developing diabetes, an integrated and robust picture has yet to emerge.

Based on 4-month claims data of the Veterans Health Administration, Sernyak et al.26 reported that schizophrenia patients who received atypical neuroleptics (clozapine, olanzapine, quetiapine, or risperidone) were 9% more likely to have diabetes mellitus than those who received typical neuroleptics. For patients younger than 40, each individual atypical group had a significantly increased risk of diabetes compared with the group of all typical antipsychotics; in the analysis of all ages combined, the risperidone group was the only atypical group which did not have a significantly increased risk of diabetes. Another study by Koro et al.27 concluded that, compared to patients treated with typical antipsychotics and non-users of antipsychotics, olanzapine-treated (but not risperidone-treated patients) had a significantly increased risk of diabetes in a nested case-control analysis that used a randomly selected sub-sample.

In contrast, a large, retrospective study conducted by Buse et al.,28 using the prescription claims from a PBM database, showed that among users of antipsychotics (across all diagnoses), the typical versus atypical groups did not differ significantly with respect to level of risk. Among the atypical antipsychotic groups, only the risperidone group was associated with a significantly greater risk of diabetes than was the haloperidol group. There was no statistically significant difference in the risk of diabetes between the use of olanzapine versus risperidone. Through a managed care database analysis, Lee et al.29 also showed that regardless of patient diagnosis, the 1-year risk of diabetes was statistically indistinguishable for patients initiating therapy with typical versus atypical antipsychotics. Furthermore, no significant differences were found for patients initiating therapy with either olanzapine or risperidone versus typical antipsychotics, or for olanzapine versus risperidone.

Two recently published studies specifically compared the risk of diabetes between clozapine and typical antipsychotic treatment. Among schizophrenia patients in a Medicaid database, Lund et al.30 found no significant differences in overall incidence rates for diabetes associated with clozapine versus typical antipsychotic agents, although a higher risk of diabetes with clozapine was observed among younger patients (aged 20-34 years). Using state Medicaid and Medicare data, Wang et al.31 concluded that clozapine treatment did not increase the risk of developing diabetes.

The review of these published results underscores the diversity of study methodologies and findings. In addition, studies are needed to overcome key limitations of some of these analyses. Important examples include use of only pharmacy claims data, examination of pooled disease diagnoses, relative short periods of observation, and inadequate adjustment for other risk factors.


The intent of this study was to provide additional evidence to the growing body of empirical work regarding a possible association between use of antipsychotics and risk of diabetes. The primary objective was to assess, through an integrated medical and prescription claims in a large commercially insured population, the incidence of diabetes for individuals with schizophrenia treated with antipsychotics. The incidence rates of diabetes were compared for a) patients treated with typical versus atypical antipsychotics, as well as for b) those treated with olanzapine versus risperidone, the two most frequently prescribed atypical antipsychotics.



The data used for this analysis were extracted from the IMS Health LifeLinkT Integrated Claims Solutions. It is an employer claims database representing the healthcare experience of approximately 1.6 million employees, dependents and retirees from more than 40 indemnity and preferred provider organizations plans. All types of medical service and outpatient prescription claims are included, along with patient demographics and enrollment information. The database has a low turnover rate (approximately 8% per year) and is geographically diverse (from 21 different states).

Study Group

Patients (aged 18 to 64) initiating antipsychotic treatment with atypical or typical agents, between October 1, 1996, and December 31, 1998, and diagnosed with schizophrenia (at least one diagnosis in the inpatient setting or at least two diagnoses in the outpatient setting (ICD9 295.xx)32 were identified. Atypicals include clozapine, olanzapine, quetiapine, and risperidone, and typicals include chlorpromazine, fluphenazine, haloperidol, loxapine, mesoridazine, molindone, perphenazine, prochlorperazine, promazine, thioridazine, thiothixene, and trifluoperazine. Patients were selected if they had no prescriptions for antipsychotics during the 6-month period before initiation. Furthermore, continuous eligibility for insurance coverage for 1 year before and after initiation (24 months) was required. Patients who had filled a prescription for any antidiabetic medication or received a diagnosis for diabetes in the year before initiation were excluded.

Outcome Measure

An incidence of diabetes was defined as a patient receiving two or more diagnoses for diabetes (ICD9 250.xx)33 or receiving any antidiabetic medication (insulin preparation or sulfonylurea) during the 1-year postinitiation on an antipsychotic. Because physicians may assign one diabetes diagnosis for tests to rule out diabetes, identified cases of diabetes were required to have two diagnoses if there was no use of antidiabetic medications.

Statistical Analysis

Differences in patient demographic and clinical characteristics among treatment groups were tested using the chi-squares tests (for comparison of categorical variables) and the independent t-tests (for comparison of continuous measures). Multiple logistic regression models were used to estimate odds ratios controlling for patient demographic and clinical characteristics, and for antipsychotic treatment duration. Patient demographics included age, gender, regional differences (northeast central vs. other regions), and enrollment status (enrollee vs. dependent). Clinical characteristics included the presence of general health comorbidities (hyperlipidemia (ICD9 272.xx) and hypertension (ICD9 401.xx-405.xx), and the presence of other mental health conditions (bipolar (ICD9 296.4x-296.8x), alcohol/drug dependence/abuse (ICD9 303-305.xx 291-292.xx), and nonorganic psychosis (ICD9 296.0-296.3x 296.9x 297-298.xx 300-301.xx 311.xx)). Treatment duration was defined as the number of days treated with the antipsychotic during the year after initiation. All computations were performed using the SAS System for Windows, Version 8.2.


Patient Sample 

A total of 815 patients met inclusion/exclusion criteria and were included in the analysis. The typical antipsychotic group included 353 patients and the atypical group consisted of 462 patients. Within the atypical group, there were 258 initiated with olanzapine (mean daily dose 9.96 mg/d; range from 2.5 to 25 mg/d) and 187 initiated with risperidone (mean daily dose 3.39 mg/d; range from 0.5 to 12 mg/d). There were also a small number of patients initiated on clozapine and quetiapine (11 and 6, respectively). Because these numbers were small, these two medication groups were not separately analyzed.

Table 1 displays patient demographic and clinical characteristics. Patients were generally comparable (both for the typical versus atypical groups, and for the olanzapine versus risperidone groups). Compared with the typical group, the prevalence rates of alcohol/drug dependence or abuse in the atypical, or the olanzapine groups were significantly higher (P = .0383 and P = .0256, respectively).

Significantly more patients in the atypical, or the risperidone groups had diagnoses for nonorganic psychosis (P = .0105 and
P = .0030, respectively) when compared with the typical group. Finally, in the year after the initiation of an antipsychotic, patients in the atypical group and the olanzapine group had significantly longer treatment duration for the antipsychotic agent initiated than patients in the typical group (22 days, P = .0334 and 24 days, P = .0473, respectively).

Unadjusted Diabetes Incidence Rates

Also presented in Table 1 are the unadjusted incidence rates of diabetes in the year after initiation of antipsychotic treatment. The diabetes incidence rate for the atypical group as a whole (2.81%) was not statistically different from that for the typical group (1.42%) (P = .1786).

The olanzapine group had a diabetes incidence rate nearly equivalent to the typical group (1.55% versus 1.42%, P = .8920). However, the unadjusted diabetes incidence rate for the risperidone group (4.28%) was significantly higher than for the typical group (P = .0390). The unadjusted diabetes incidence rates for olanzapine and risperidone were not significantly significant
(P = .0795).

Adjusted Odds Ratios

Multiple logistic regression models were used to derive the adjusted odds ratios (OR) of incidence rates of diabetes. Results are reported in Table 2. Covariates that were controlled in these models included age, gender, region, enrollment status, treatment duration, and the presence of general and mental health comorbidities. After controlling for these potential confounding factors, the probability of being classified as having diabetes was not significantly different for the atypical group compared to the typical group (OR=2.352; 95% confidence interval [CI]: 0.802-6.896; P = .1192) or for the olanzapine group versus the typical group (OR=1.059; CI: 0.261-4.302; P = .9361). Risperidone-treated patients had significantly higher incidence rate of diabetes compared to those treated with typicals (OR=4.061; CI: 1.230-13.40; P = .0214). The olanzapine treatment group had a significantly lower incidence rate of diabetes (OR=0.267; CI: 0.072-0.988; P = .0479), compared to the risperidone treatment group.

Additionally, as reported in Table 2, it is interesting to note that age and preexistence of lipid abnormalities were significant predictors for developing diabetes; findings consistent with previously published results.26,28,30


The findings from this retrospective, integrated claims data analysis indicate that the incidence rate of diabetes was similar for individuals receiving treatment with atypical antipsychotics compared to typical antipsychotics. The risk of diabetes among patients treated with olanzapine did not differ significantly from patients in the typical antipsychotic group. However, the risperidone group had a significantly greater risk of diabetes compared to the typical group. Additionally, individuals receiving treatment with olanzapine compared to risperidone had a lower incidence of diabetes.

These findings are consistent with results reported in recent publications.28-31 For example, Buse et al.28 found that the risk of diabetes for the typical group was not significantly different from that of the atypical group. No significant increase in the risk of diabetes was observed for either the olanzapine or the clozapine group when compared to the haloperidol group. However, the risk of diabetes in the risperidone group was significantly higher relative to the haloperidol group, though there was no statistically significant difference in the risk of diabetes between the olanzapine and risperidone cohorts.

Our findings differ from those reported in other analyses.26,27 The disagreement may be due in part, to different analytic approaches. For example, the Sernayak study26 used only 4-month data, did not exclude patients previously diagnosed with or treated for diabetes, and did not have history of medication use (no "washout" period). Thus, some identified diabetes cases in the study could represent preexistent conditions or have been associated with previous treatments.

There have been several hypotheses presented on the possible association of the risk of diabetes and atypical antipsychotic treatment. One such hypothesis is that diabetes may be a secondary event to weight gain associated with atypical antipsychotics.34 However, weight is only one of the several important risk factors for diabetes. Other factors include family history, physical and mental comorbidities, age, and lack of physical activity.35 Although weight data were not available for this study, age and preexisting lipid abnormalities were found to be significant predictors of developing diabetes (Table 2). Others have hypothesized that atypical antipsychotics may directly impair pancreatic b-cell function. A recently published randomized, single-blind, small sample study, however, found no evidence that the treatment of healthy volunteers with olanzapine or risperidone decreased the insulin secretory response to a prolonged hyperglycemic challenge.36 Because of the important monitoring and treatment implications, research regarding the possible association between the use of atypical antipsychotics and risk of diabetes will continue.  


The interpretation of the results needs to be made in light of the limitations of this study. A number of potentially important diabetes risk factors, such as family history of diabetes and related diseases, ethnicity, obesity, and sedentary life style, were not contained in the database and thus were not adjusted in the analysis, although patient demographic and clinical characteristics and treatment duration were controlled in the analysis. Second, the definition of diabetes incidence was not based on clinical laboratory tests; instead it relied on claims data for diabetes diagnoses or filled prescription for antidiabetic medications. However, the method would be unlikely to bias against or favor either antipsychotic group. Finally, this study was based on claims in a large commercially insured database; therefore, the generalization of these findings may need to be confirmed using databases from other care settings.


Results from this large, retrospective, claims database analysis indicate that the incidence of diabetes was not significantly different during treatment with atypical compared to typical antipsychotics for patients diagnosed with schizophrenia. However, the risperidone treatment group had a higher incidence of diabetes than the typical or the olanzapine treatment groups.


1. Kane JM. Schizophrenia. N Engl J Med 334(1):34-41, 1996.

2. Rice DP, Miller LS. The Economic Burden of Schizophrenia. Conceptual and Methodological Issues, and Cost Estimates. In: Moscarelli M, Rupp A, Sartorius N, eds. Handbook of Mental Health Economics and Health Policy. Vol. Vol. I: Schizophrenia, 1996.

3. Kane JM, Marder SR. Psychopharmacologic treatment of schizophrenia. Schizophrenia Bull 19:287-302, 1993.

4. American Psychiatric Association. Practice guideline for the treatment of patients with schizophrenia. Am J Psych 154(4 Suppl):1-63, 1997.

5. Beaumont G. Antipsychotics: The future of schizophrenia treatment. Curr Med Res Opinion 16:37-42, 2000.

6. Weiden PJ, Olfson M. Cost of relapse in schizophrenia. Schizophrenia Bull 21:419-429, 1995.

7. Gerlach J. The continuing problem of extrapyramidal symptoms: Strategies for avoidance and effective treatment. J Clin Psych 60(Suppl 23):20-24, 1999.

8. Worrel JA, Marken PA, Beckman SE, Ruehter VL. Atypical antipsychotic agents: a critical review. Am J Health-System Pharm 57:238-55, 2000.

9. Sauriol L, Laporta M, Edwardes MD, et al. Meta-analysis comparing newer antipsychotic drugs for the treatment of schizophrenia: Evaluating the indirect approach. Clin Therap 23:942-956, 2001.

10. Volavka J, Czobor P, Sheitman B, et al. Clozapine, olanzapine, risperidone, and haloperidol in the treatment of patients with chronic schizophrenia and schizoaffective disorder. Am J Psych 159:255-262; also contains Editorial pages 177-179, 2002.

11. Wirshing DA, Spellberg BJ, Erhart SM, et al. Novel antipsychotics and new onset diabetes. Biol Psych 44:778-83, 1998.

12. Kamran A, Doraiswamy PM, Jane JL, et al. Severe hyperglycemia associated with high doses of clozapine. Am J Psych 1994.

13. Popli AP, Konicki PE, Jurjus GJ, et al. Clozapine and associated diabetes mellitus. J Clin Psych 58108-58111, 1997; .

14. Henderson DC, Cagliero E, Gray C, et al. Clozapine, diabetes mellitus, weight gain, and lipid abnormalities: a five year naturalistic study. Am J Psych 157975-157981, 2000.

15. Fertig MK, Brooks VG, Shelton PS, English CW. Hyperglycemia associated with olanzapine. J Clin Psych 59687-59689, 1998.

16. Goldstein LE, Sporn J, Brown S, et al. New-onset diabetes mellitus and diabetic ketoacidosis associated with olanzapine treatment. Psychosomatics 40:438-443, 1999.

17. Ober SK, Hudak R, Rusterholtz A. Hyperglycemia and olanzapine. Am J Psych 156:970, 1999.

18. Croarkin PE, Jacobs KM, Bain BK. Diabetic ketoacidosis associated with risperidone treatment? [see comments]. Psychosomatics 41:369-370, 2000.

19. Wirshing DA, Pierre JM, Eyeler J, et al. Risperidone-associated new-onset diabetes. Biol Psych 50:148-149, 2001.

20. Sobel M, Jaggers ED, Franz MA. New-onset diabetes mellitus associated with the initiation of quetiapine treatment. J Clin Psych 60:556-557, 1999.

21. Procyshyn RM, Pande S, Tse G. New-onset diabetes mellitus associated with quetiapine. Can J Psych 45:668-669, 2000.

22. Mukherjee S, Decina P, Bocola V, et al. Diabetes mellitus in schizophrenic patients. Comprehens Psych 37:68-73, 1996.

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

24. Regenold WT, Thapar RK, Marano C, et al. Increased prevalence of type 2 diabetes mellitus among psychiatric inpatients with bipolar 1 affective and schizoaffective disorders independent of psychotropic drug use. J Affective Dis 7019-7026, 2002.

25. Harris MI, KM F, Cowie CC, et al. Prevalence of diabetes, impaired fasting glucose tolerance in US adults: the Third National Health and Nutrition Examination Survey, 1988-1994. Diabetes Care 21518-21524, 1998.

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

27. Koro CE, Fedder DO, L'Italien GJ, et al. Assessment of independent effect of olanzapine and risperidone on risk of diabtes among patients with schizophrenia: population based nested case-control study. Br Med J 325:243-247, 2002.

28. Buse J, Cavazzoni P, Hornbuckle K, et al. A retrospective cohort study of diabetes mellitus and antipsychotic treatment in the United States. J Clin Epidemiol 56:164-170, 2003.

29. Lee DW, Fowler RB, Kadlubek PJ, Haberman M. No significant difference in the risk of diabates mellitus during treatment with typical versus atypical antipsychotics: Results from a large observational study. Drug Benefit Trends 14:46-52, 2002.

30. Lund BC, Perry PJ, Brooks JM, Arndt S. Clozapine use in patients with schizophrenia and the risk of diabetes, hyperlipidemia, and hypertension: A claims-based approach. Arch Gen Psych 581:172-176, 2001.

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

32. Soumerai SB, McLaughlin TJ, Ross-Degnan D, et al. Effects of limiting Medicaid drug-reimbursement benefits on the use of psychotropic agents and acute mental health services by patients with schizophrenia. N Engl J Med 331:650-655, 1994.

33. Gilmer TP, O'Connor PJ, Manning WG, Rush WA. The cost to health plans of poor glycemic control. Diabetes Care 20:1847-1853, 1997.

34. Allison DB, Mentore JL, Heo M, et al. Antipsychotic-induced weight gain: a comprehensive research synthesis. Am J Psych 156:1686-1696, 1999.

35. Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 21(suppl 1):S5-S19, 1998.

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

©2000-2013. All Rights Reserved. Veterinary Solutions LLC is an authorized retailer for The Journal of Applied Research