To determine whether depression is associated with an increased risk for onset of diabetes.

In 1981, a total of 3,481 household-residing adults participated in the Epidemiologic Catchment Area (ECA) Program survey at the East Baltimore site. A follow-up of that cohort after 13 years completed 1,897 interviews, amounting to >72% of survivors. In 1981, depression was assessed with the National Institutes of Mental Health (NIMH) Diagnostic Interview Schedule and diabetes, by self-report. This prospective analysis focused on subjects at risk for onset of diabetes by removing from the analysis individuals with diabetes in 1981.

There were 89 new cases of diabetes among 1,715 individuals at risk, yielding a 13-year cumulative incidence of diabetes of 5.2%. In logistic models, major depressive disorder, but not milder forms of depression or other forms of psychiatric disorder, predicted the onset of diabetes (estimated relative risk, 2.23; 95% CI 0.90–5.55). Controlling for age, race, sex, socioeconomic status, education, use of health services, other psychiatric disorders, and body weight did not weaken the relationship.

Major depressive disorder signals increased risk for onset of type II diabetes. Limitations of the findings arise from the difficulty in determining temporal order with two chronic conditions, even when the temporal order of measurement is clear. In addition, even though control variables were introduced for the use of health services, it is possible that the treatment for depression led to an earlier diagnosis of diabetes in this sample.