It has been known for several years that a strong bidirectional relationship exists between diabetes and depression. Observational studies show that depression usually predates diabetes and that the presence of depression increases the risk of future diabetes; but diabetes also increases the risk of subsequent depression. However, most studies were done in Europe and the USA, and the coexistence of other mental disorders as an explanatory factor had not been systematically investigated. A new study in Diabetologia by Peter de Jonge and his international colleagues has used the framework of the World Mental Health surveys to examine the relation between a wide range of mental health disorders and diabetes diagnosis. Their data support the focus on depression as an independent risk factor for diabetes, but also suggest that this focus should be extended to impulse control disorders.