To estimate the association between depressive symptoms and Type 2 diabetes, as well as previously undetected diabetes, in a large population-based sample in Germany and to determine associated variables.
We used baseline data on 4595 participants (age 45-75 years, 50.2% women) from the German Heinz Nixdorf Recall study, a population-based, prospective cohort study which started in 2000. Diabetes mellitus was assessed by self report (physician diagnosis or medication), undiagnosed diabetes based on blood glucose levels. Depressive symptoms were assessed using the Center for Epidemiologic Studies Depression (CES-D) scale short form (cut-off >or= 15 points). We fitted multiple logistic regression models.
The prevalence of diagnosed and previously undetected diabetes was 9.3% (95% confidence interval 8.2-11.6) and 7.6% (6.6-8.8) in men and 6.0% (5.1-7.1) and 3.2% (2.5-4.0) in women, respectively. Compared with non-diabetic women, the prevalence of depressive symptoms was not significantly different in diabetic women (age-adjusted odds ratio, 95% confidence interval 1.48; 0.98-2.24) and women with undiagnosed diabetes (0.67; 0.33-1.36). In men, the prevalence of depressive symptoms tended to be lower in diabetic than in non-diabetic subjects (0.62; 0.35-1.09), but the depressive symptoms were significantly less frequent in men with undiagnosed diabetes (0.30; 0.13-0.70). The pattern remained after further adjustment. Significant associations with depressive symptoms were found for co-morbidities and living without a partner in both women and in men, and for body mass index and activity level in women only.
After adjustment for relevant covariates, the association between depressive symptoms and Type 2 diabetes was heterogenous in our population-based study. In subjects with undiagnosed diabetes, however, depressive symptoms were less frequent in men. Co-morbidities and psychosocial conditions are strongly associated with depressive symptoms.