Perceived Racial Discrimination as a Predictor of Health Behaviors: the Moderating Role of Gender

Abstract

Perceived racial discrimination (PRD) has been implicated in undermining the mental and physical health of racial/ethnic minorities. Researchers have begun to explore the indirect role of health behaviors as one factor in helping to explain this relationship. The goal of the present study was to examine the relationship between PRD and a wide range of health behaviors using a prospective, longitudinal design and to explore the role of gender in moderating these relationships. Using data from the Maryland Adolescent Development in Context Study, we examined the relationship between adolescent PRD (accumulated across ages 14-21) and health behaviors (i.e., diet, substance use, exercise) at age 30 in a sample of middle-class black men and women. Using structural equation modeling, results revealed that more cumulative PRD during adolescence was associated with less healthy eating, more substance use (among men), and more exercise (among women) in young adulthood. Implications of these findings for understanding the role of health behaviors in explaining the link between PRD and health outcomes are considered.

Fig. 1

Structural equation model predicting age 30 diet. Gender was coded as female = 0 and male = 1; Path coefficients are unstandardized; standard errors are shown in parentheses; dashed lines reflect paths that were non-significant and thus removed from the model (*p < .05, **p < .001)

Fig. 2

Structural equation model predicting age 30 exercise. Gender was coded as female = 0 and male = 1; Path coefficients are unstandardized, and standard errors are shown in parentheses; dashed lines reflect paths that were non-significant and thus, removed from the model (*p < .05, **p < .001)

Fig. 3

Structural equation model predicting age 30 substance use. Gender was coded as female = 0 and male = 1; Path coefficients are unstandardized; standard errors are shown in parentheses; dashed lines reflect paths that were non-significant and thus removed from the model (^p < .10, *p < .05, **p < .001)