Do assets explain the relation between race/ethnicity and probable depression in U.S. adults?

Abstract

Depression is a leading cause of disability in the U.S. across all race/ethnicity groups. While non-Hispanic Black and Hispanic persons have worse physical health on most indicators than non-Hispanic White persons, the literature on the association between race/ethnicity and rates of depression is mixed. Given unequal distribution of assets across racial/ethnic groups, it is possible that social and economic differences may explain differential rates of depression across race/ethnicity groups. Using National Health and Nutrition Examination Survey (NHANES) data from 2007-2016, we constructed a nationally representative sample of 26,382 adults over 18 years old (11,072 non-Hispanic White, 5,610 non-Hispanic Black, 6,981 Hispanic, and 2,719 Other race). We measured symptoms of depression using the Patient Health Questionnaire-9 (PHQ-9), with a score of 10 or more indicating probable depression. We identified three kinds of assets: financial assets (income), physical assets (home ownership), and social assets (marital status and education). We estimated the weighted prevalence of probable depression across race/ethnicity groups, odds ratios controlling for assets, and predicted probabilities of probable depression across race/ethnicity and asset groups. Three results contribute to our understanding of the differences in probable depression rates between race/ethnicity groups: 1) Non-Hispanic Black and Hispanic persons had a higher weighted prevalence of probable depression in the U.S. than non-Hispanic White persons. In models unadjusted for assets, non-Hispanic Black and Hispanic persons had 1.3 greater odds of probable depression than non-Hispanic White persons (p<0.01). 2) We found an inverse relation between assets and probable depression across all race-ethnicity groups. Also, non-Hispanic Black and Hispanic persons had fewer assets than non-Hispanic Whites. 3) When we controlled for assets, non-Hispanic Black and Hispanic persons had 0.8 times lower odds of probable depression than non-Hispanic White persons (p<0.05). Thus, when holding assets constant, minorities had better mental health than non-Hispanic White persons in the U.S. These three findings help to reconcile findings in the literature on race/ethnicity and depression. Given vastly unequal distribution of wealth in the U.S., it is not surprising that racial minorities, who hold fewer assets, would have an overall larger prevalence of mental illness, as seen in unadjusted estimates. Once assets are taken into account, Black and Hispanic persons appear to have better mental health than non-Hispanic White persons. Assets may explain much of the relation between race/ethnicity group and depression in the U.S. Future research should consider the role of assets in protecting against mental illness.

Fig 1. Weighted prevalence of probable depression by race/ethnicity and household income in U.S. adults (N = 26,382).

The National Health and Nutrition Examination Survey (NHANES) 2007–2016. Missing values for household income (n = 1,935). Weighted to the U.S. population. Probable depression defined by Patient Health Questionnaire-9 (PHQ-9) cutoff of ≥10.

Fig 2. Weighted prevalence of probable depression by race/ethnicity and educational attainment in U.S. adults (N = 26,382).

The National Health and Nutrition Examination Survey (NHANES) 2007–2016. Missing values for education (n = 19). Weighted to the U.S. population. GED is the general education diploma. Probable depression defined by Patient Health Questionnaire-9 (PHQ-9) cutoff of ≥10.

Fig 3. Weighted prevalence of probable depression by race/ethnicity and marital status in U.S. adults (N = 26,382).

The National Health and Nutrition Examination Survey (NHANES) 2007–2016. Missing values for marital status (n = 1,386). Weighted to the U.S. population. Probable depression defined by Patient Health Questionnaire-9 (PHQ-9) cutoff of ≥10.

Fig 4. Weighted prevalence of probable depression by race/ethnicity and home ownership in U.S. adults (N = 26,382).

The National Health and Nutrition Examination Survey (NHANES) 2007–2016. Missing values for home ownership (n = 295). Weighted to the U.S. population. Probable depression defined by Patient Health Questionnaire-9 (PHQ-9) cutoff of ≥10.

Fig 5. Predicted probability of depression across race/ethnicity using unadjusted and adjusted models in U.S. adults (N = 26,382).

The National Health and Nutrition Examination Survey (NHANES) 2007–2016. Weighted to the U.S. population. Unadjusted model controlled only for race/ethnicity and NHANES cohort wave. Adjusted model controlled for race/ethnicity, age, gender, education, marital status, household income, home ownership, household size, and NHANES cohort wave. Probable depression defined by Patient Health Questionnaire-9 (PHQ-9) cutoff of ≥10.