Glycemic Contributions to Depressive Outcomes: National Health and Nutrition Examination Surveys, 2005-2012 Open Access

Chandrasekar, Eeshwar Kaushik (2015)

Permanent URL:


To address limited national data exploring the factors that contribute to depressive outcomes among individuals with diabetes, we developed regression models to estimate the independent associations between glycemic categories (normal glycemia, prediabetes, diabetes) and depressive outcomes in the United States. We used data from 21,618 adults surveyed in the 2005-2012 National Health and Nutrition Examination Surveys, a nationally representative sample of non-institutionalized US civilians. Diabetes was classified based on self-reporting a physician diagnosis of diabetes or measured A1C ≥ 6.5% (≥48mmol/mol). Individuals without a diagnosis of diabetes, but whose AIC was 5.7-6.4% (39 to 47mmol/mol) were classified as having pre-diabetes and those with AIC < 5.7% (<39mmol/mol) were classified as having normal glycemic status. We used the PHQ-9 screening questionnaire to determine prevalence and odds of ‘Clinically Significant Depressive Symptoms' (CSDS) [PHQ-9 score ≥ 10 or antidepressant use] and ‘Major Depressive Syndrome' [PHQ-9 score ≥12] (MDS). We calculated prevalence of CSDS and MDS standardized to the 2000 US census population, and used multivariate regression models to quantify the independent associations between glycemic categories and both depressive outcomes. The age standardized prevalence of CSDS in 2005-2008 was 15.2% [14.0-16.4], 15.9% [13.4-18.7], and 26.2% [21.2-31.9] for individuals with normal glycemia, pre-diabetes, and diabetes, respectively. For MDS, the age-standardized prevalence was 1.0% [0.7-1.3], 1.5% [0.9-2.5], and 2.6% [1.3-4.9] for the three glycemic groups, respectively. There were no significant changes in either CSDS or MDS prevalence between 2005-2008 and 2009-2012 in either crude or age-standardized estimates. While having diabetes was associated with two-fold higher odds of depressive outcomes in crude models, having diabetes was independently associated with a 25% greater odds of CSDS in adjusted models. The association between diabetes and depressive symptoms suggests a need to further integrate depression screening and treatment into routine diabetes management.

Table of Contents


Research Design and Methods...3

Data Sources...3


Statistical Analysis...5



Manuscript Acknowledgments...16

Tables and Figures...17

Table 1: Selected Characteristics By Time Period, NHANES 2005-2012...17

Figure 1: Prevalence of Depressive Outcomes by Glycemic Category...18

Table 2: Adjusted Odds Ratios of CSDS...19

Table 3: Adjusted Odds Ratios for MDS...20


Table 1A: Selected Characteristics by Glycemic Status, NHANES 2005-2012...23

Figure 2: Directional Acyclic Graphs - Sociodemographic...24

Figure 3: Directional Acyclic Graphs - Full Model...25

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
Subfield / Discipline
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Partnering Agencies
Last modified

Primary PDF

Supplemental Files