Trends in depression by glycemic status: Serial cross-sectional analyses of the National Health and Nutrition Examination Surveys, 2005-2016

Document Type

Journal Article

Publication Date



Primary care diabetes








Comorbidities; Depression; Diabetes; Epidemiology


AIMS: We examined changes in the prevalence of elevated depressive symptoms among US adults with diabetes, prediabetes, and normal glycemic status during 2005-2016. METHODS: We analyzed data from 32,676 adults in the 2005-2016 National Health and Nutrition Examination Surveys. We defined diabetes as self-reporting a physician diagnosis of diabetes or A1C ≥ 6.5% [48 mmol/mol], and prediabetes as A1C 5.7-6.4% [39-46 mmol/mol]. We used the 9-item Patient Health Questionnaire (PHQ-9) score ≥ 10 or antidepressant use to define 'clinically significant depressive symptoms' (CSDS) and PHQ-9 score ≥ 12 as 'Major Depressive Disorder' (MDD). We calculated prevalence age-standardized to the 2000 US census and used logistic-regression to compute adjusted odds of CSDS and MDD for 2005-2008, 2009-2012, and 2015-2016. We analyzed the prevalence of A1C ≥ 9.0% [75 mmol/mol], systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg, non-HDL cholesterol ≥ 130 mg/dL, and current smoking among adults with diagnosed diabetes by depressive status. RESULTS: The prevalence of CSDS increased among individuals with normal glycemic status from 15.0% (13.5-16.2) to 17.3% (16.0-18.7) (p = 0.03) over 2005-2016. The prevalence of CSDS and MDD remained stable among adults with prediabetes (~ 16% and 1%, respectively) and diabetes (~ 26% and ~3%). After controlling for glycemic, sociodemographic, economic, and self-rated health variables, we found 2-fold greater odds of CSDS among unemployed individuals and 3-fold greater odds among those with fair/poor self-rated health across all survey periods. Cardiometabolic care targets for adults with diagnosed diabetes were stable from 2005 to 2016 and similar across depressive status. CONCLUSIONS: One-fourth of adults with diabetes have comorbid CSDS; this prevalence remained stable over 2005-2016 with no change in diabetes care. At the population level, depression does not appear to impact diabetes care, but further research could explore subgroups that may be more vulnerable and could benefit from integrated care that addresses both conditions.