Intersectional vulnerability in the relationship between discrimination and inflammatory gene expression

Abstract

Addressing social disparities in health and well-being requires understanding how the effects of discrimination become biologically embedded, and how embedding processes might vary across different demographic contexts. Emerging research suggests that a threat-related gene expression response may contribute to social disparities in health. We tested a contextual vulnerability model of discrimination embedding using an empirical intersectionality (interaction discovery) analysis of pro-inflammatory gene expression in a national sample of non-institutionalized, English-speaking adults with RNA biomarker data (n = 543). At the time of data collection, the average age of participants was 55 years (SD = 13.26) and approximately half identified as female (50.46%). Most participants identified as White (∼73%) and had some college experience (∼60%). Results showed significant variation in the strength of association between daily discrimination and inflammatory gene expression by race and sex (b = -0.022; 95% CI:-0.038,-0.005, p = .009) with the estimated marginal association larger for racially-minoritized males (b = 0.007; 95% CI:-0.003,0.017, p = .163), compared to White males (b = -0.006; 95% CI:-0.013,0.001, p = .076). This study indicates that the link between daily discrimination and inflammatory gene expression may vary by sociodemographic characteristics. To improve initiatives and policies aimed at ameliorating disparities within populations, greater attention is needed to understand how interlocking systems of inequalities contribute to physiological health.

Keywords: CTRA; Discrimination; Gene expression; Inflammation.

Fig. 1

Prediction of Inflammatory Gene Expression by Discrimination x Racially-minoritized x Sex. Note. The slopes derive from the ordinary least squares regression predicting the inflammatory gene expression composite score, choosing predictors using bidirectional elimination. Error bars derive from 95% confidence intervals.