Background:
Determinants of depressive symptoms in older adults according to diabetes status have not been widely investigated. Here we examine correlates of depressive symptoms in persons without diabetes, with prediabetes, and with diagnosed diabetes.
Methods:
We used data from visit 5 (2011-2013) of the ARIC Study. Depressive symptoms were ascertained using the 11-item CES-D, with scores ≥9 indicating probable major depression, and via self-reported diagnosis of depression in the past 2 years. Diabetes was defined based on self-reported diagnosis or use of glucose lowering medication. We defined prediabetes as HbA1c of 5.7-6.4% in persons without diabetes. We examined demographic factors, markers of health status, and metrics of health care access as correlates of depressive symptoms. Prevalence ratios (PRs) were calculated using Poisson regression with robust variance estimation.
Results:
Among 6319 participants (mean age 76 years, 59% female, and 77% white), the prevalence of depressive symptoms in persons with prediabetes (8.6%) and diabetes (14.7%) was significantly higher than in persons with no diabetes (6.9%). Correlates of depressive symptoms were similar across diabetes categories. In fully adjusted models, the strongest predictors of depressive symptoms were self-reported fair or poor health (compared to excellent health) and variables related to access to, and satisfaction with, health care. In persons with diabetes, insulin use, but not HbA1c control, was significantly associated with depressive symptoms. Measures of socioeconomic status (education, income) were significantly associated with depressive symptoms in all three groups (Table).
Conclusions:
The strongest predictors of depressive symptoms were metrics of health care access and satisfaction, suggesting that system-level factors may be important in addressing depression in diabetes and prediabetes.
Socioeconomic Status and Mortality: Contribution of health care access and psychological distress among U.S. adults with diagnosed diabetes