Experiences of Discrimination and Alcohol Involvement Among Young Adults at the Intersection of Race/Ethnicity and Gender

• Published in: Journal of racial and ethnic health disparities 2024-10
• Main Authors: Lopez-Vergara, Hector Ismael, Rozum, William, Charvis, Jodi M Sutherland, Iacoi, Sydney, Vergara-Lopez, Chrystal, Stein, L A R
• Format: Article
• Language: English
• ISSN: 2197-3792; 2196-8837; 2196-8837
• DOI: 10.1007/s40615-024-02191-x

Although discrimination is an important social determinant of alcohol involvement, there is a dearth of research testing these associations across race/ethnicity and gender. This is an important research gap given that experiences of discrimination and therefore links with alcohol involvement may vary as a function race/ethnicity and gender intersectional identities. We tested for measurement invariance in discrimination and alcohol involvement and examined group differences in means and covariances. The sample consisted of n = 1187 young adults (ages 18-26; n = 193 Black women, n = 209 Latina women, n = 186 White women, n = 198 Black men, n = 203 Latino men, and n = 198 White men). We found evidence for differential item functioning for discrimination and alcohol involvement that violated assumptions needed to make manifest between-group comparisons. To model the source and degree of differential item functioning, we used partial measurement invariance and dropped a discrimination item that did not reliably overlap with the latent factor for White women. After accounting for differential item functioning, Black women and men reported the highest discrimination, followed by Latinx women and men, and then White women and men. White women reported the most alcohol involvement, followed by White men, Latina women, Latino men, Black men, and Black women. Discrimination and alcohol involvement were positively associated for all groups except White women, though effect sizes varied with Black men exhibiting the largest effect. An intersectionally valid understanding of discrimination and alcohol involvement may necessitate statistical approaches that can test for (and model) differential item functioning prior to making between-group quantitative comparisons.