The relationship between depression and diabetes mellitus: findings from the Hertfordshire Cohort Study

To assess the relationship between depression scores and diabetes, glucose and insulin in a cross-sectional population-based study.

One thousand, five hundred and seventy-nine men and 1418 women from the Hertfordshire Cohort Study were assessed for diabetes. Plasma glucose and insulin concentrations were measured at 0, 30 and 120 min during a standard 75-g oral glucose tolerance test. Depressive and anxiety symptoms were measured using the Hospital Anxiety and Depression Scale (HADS).

Overall, 431 (14.6%) were diagnosed with diabetes [232 men (14.9%) and 199 women (14.3%)]. One hundred and eight (47%) men and 74 (37%) women had known diabetes. The remainder were previously undiagnosed. Fifty-nine (3.7%) men and 65 (4.6%) women had possible depression (HAD-D scores 8-10) and 17 (1.1%) men and 20 (1.4%) women had probable depression (HAD-D scores > or = 11). Probable depression was associated with an adjusted odds ratio for diabetes of 3.89 [95% confidence interval (CI) 1.28-11.88] in men and 1.51 (95% CI 0.47-4.84) in women. In men without previously diagnosed diabetes, fasting insulin (P = 0.035), 2-h glucose concentrations (P = 0.028) and insulin resistance (P = 0.032) were significantly associated with HAD-D scores. With the exception of 2-h glucose concentrations (P = 0.034), the associations were not significant in women.

These data support the hypothesis that depression may increase the risk for diabetes. The relationship between depression score and metabolic variables extends across the whole population and is not confined to those with either diagnosed depression or diabetes. This relationship should lead clinicians to consider screening for diabetes in those with depression and vice versa.