Predictors of type 2 diabetes in a nationally representative sample of adults with psychosis

D.L. Foley, A.J. Mackinnon, Vera Morgan, Gerald Watts, J.J. Mcgrath, D.J. Castle, Anna Waterreus, C.A. Galletly

Research output: Contribution to journalArticlepeer-review

28 Citations (Web of Science)


Antipsychotic drugs such as clozapine and olanzapine are associated with an increased risk for type 2 diabetes, but relatively little is known about the relationship between risk factors for type 2 diabetes established in the general population and type 2 diabetes in people with psychosis. We estimated the prevalence of established risk factors and their association with type 2 diabetes in a nationally representative sample of people with an ICD-10 psychosis (N=1642) who gave a fasting blood sample (N=1155). Logistic regression was used to summarize associations adjusted for age and sex. In this sample, whose mean duration of psychosis was 14.7 years, 12.1% (13.1% of women and 11.5% of men) had type 2 diabetes at age 18-64 years based on current fasting blood glucose levels or treatment with a hypoglycaemic drug. Risk was greatly increased in young adults compared with the general population and peaked in middle age. Risk factors in the general population were common in people with psychosis and strongly associated with type 2 diabetes in those people. Treatment with clozapine was associated with an increased risk and treatment with olanzapine with a decreased risk for type 2 diabetes. The development of diabetes or pre-diabetes may therefore influence the likelihood of treatment with olanzapine over time. The strongest predictors of type 2 diabetes in a multivariate model were a body mass index of at least 40 and treated hypercholesterolemia, followed by a body mass index between 35 and 39.9, a family history of diabetes and treated hypertension. There was minimal to no confounding of the association between type 2 diabetes and current clozapine or olanzapine treatment, but neither association remained significant after adjustment for other predictors. Longitudinal relationships among predictors are likely to be complex, and previous antipsychotic drug treatment may at least partly explain risks associated with severe obesity, dyslipidemia and hypertension. A focus on weight loss is warranted in people with psychosis, but prevention strategies for type 2 diabetes should be broadened to include those with emerging dyslipidemia, hypertension and a family history of diabetes. © 2014 World Psychiatric Association.
Original languageEnglish
Pages (from-to)176-183
JournalWorld Psychiatry
Issue number2
Publication statusPublished - 2014


Dive into the research topics of 'Predictors of type 2 diabetes in a nationally representative sample of adults with psychosis'. Together they form a unique fingerprint.

Cite this