Role of depressive symptoms in explaining socioeconomic status disparities in dietary quality and central adiposity among US adults: a structural equation modeling approach

Abstract

Background: The link between socioeconomic status (SES), depression, dietary quality, and central adiposity remains unclear.

Objective: Pathways linking SES to dietary quality and central adiposity through depressive symptoms were examined across sex-ethnicity groups.

Design: Extensive data on US adults aged 30-64 y from the Healthy Aging in Neighborhoods of Diversity across the Life Span (HANDLS) study were used in multiple linear logistic regression models and structural equation models to test pathway associations. Measures included Center for Epidemiologic Studies-Depression (CES-D) scores, 2005 Healthy Eating Index (HEI) values, and dual-energy X-ray absorptiometry. Sample sizes for most analyses ranged between 1789 for anthropometric outcomes and 1227 for trunk fat outcomes.

Results: The CES-D score was associated with lower HEI scores in all sex-ethnicity groups, except in African American men, and with higher waist-to-hip ratios (WHRs) among African American women. A CES-D score > or =16 was positively associated with waist circumference (WC) and with trunk fat among white women and men, respectively. SES was positively related to central adiposity among African American men (central obesity and WC) and African American women (central obesity and percentage trunk fat) but was inversely related to central adiposity among white women. Among whites only, the total positive effect of SES on HEI was significantly mediated by CES-D score. Among white women, the total inverse effect of SES on WC and WHR was significantly explained by the CES-D score and HEI, whereas the CES-D score was positively associated with WHR among African American women, independently of SES.

Conclusion: Future mental health interventions targeted at reducing SES disparities in dietary quality and central adiposity may have different effects across sex-ethnicity groups.

FIGURE 1

Findings from the use of structural equation models: mechanisms explaining socioeconomic status (SES) disparities in central adiposity through Center for Epidemiologic Studies–Depression (CES-D) score and the Healthy Eating Index (HEI) with waist circumference (WC; A), waist-to-hip ratio (WHR; B), and trunk fat as a percentage of body fat (C) as final outcomes. Age groups, marital status, and smoking status were treated as exogenous variables, which affected all other endogenous variables. One of the exogenous variables was dropped from each model with SES, CES-D, HEI, or central adiposity as outcomes depending on their significance to gain 1 df. Numbers on the arrows represent path coefficients ± SE. Fit statistics and sample sizes are presented for each stratum-specific model. Multigroup analysis was conducted to assess significant differences between groups for each path coefficient.

FIGURE 1

Findings from the use of structural equation models: mechanisms explaining socioeconomic status (SES) disparities in central adiposity through Center for Epidemiologic Studies–Depression (CES-D) score and the Healthy Eating Index (HEI) with waist circumference (WC; A), waist-to-hip ratio (WHR; B), and trunk fat as a percentage of body fat (C) as final outcomes. Age groups, marital status, and smoking status were treated as exogenous variables, which affected all other endogenous variables. One of the exogenous variables was dropped from each model with SES, CES-D, HEI, or central adiposity as outcomes depending on their significance to gain 1 df. Numbers on the arrows represent path coefficients ± SE. Fit statistics and sample sizes are presented for each stratum-specific model. Multigroup analysis was conducted to assess significant differences between groups for each path coefficient.