People with psychosis have an elevated risk of depression and diabetes but no large-scale study has characterized their relationship. We aimed to assess this association and to evaluate possible explanatory factors.


Analysis of cross-sectional data from a national sample of 1155 people with psychosis who gave a fasting blood sample and could be tested for diabetes mellitus. The association between current diabetes mellitus and current depression was estimated using logistic regression, adjusted for age, sex and current psychotropic medication.


A diagnosis of depression was significantly associated with diabetes (OR = 2.16, P = .048) and diabetes medication (OR = 2.50, P = .050) in people with schizophrenia but no other psychosis subtype. Adjustment for cognitive processing speed and current residence (especially psychiatric hospitalization) attenuated that association to nonsignificance. Diabetes and diabetes medication were not significantly associated with antidepressant or mood stabilizer medication.


Clinicians should be aware that people with schizophrenia and diabetes have twice the rate of current depression, and that comorbid diabetes and depression is associated with cognitive impairment and hospitalization. Efforts to disentangle the causal pathways between diabetes, depression, and cognition in people with schizophrenia may be complicated by multiple indications in people with psychosis for the prescription of depression medication, and their lack of association with diabetes mellitus.