The Paradox of Perceived Discrimination in Later Life: Evidence from a Nationally Representative Study in Brazil
Abstract
Background: Perceived discrimination is a social determinant of health associated with poorer physical and mental outcomes, reflecting structural inequalities that manifest intersectionally through age, gender, race, and social class. This study analyzed variations in perceived discrimination across age groups among Brazilian adults and older adults (≥50 years), using data from the Brazilian Longitudinal Study of Aging (ELSI-Brazil, 2019–2021). Methods: This is a cross-sectional, population-based study with probabilistic sampling and national representativeness. Perceived discrimination was measured using five items from the Everyday Discrimination Scale (EDS). Given the high proportion of zero counts (77.2%) and data overdispersion, a Zero-Inflated Negative Binomial (ZINB) model with robust variance was applied to estimate Incidence Rate Ratios (IRR), accounting for the complex survey design. Results:The Wald F test indicated significant differences across age groups (p=0.012), with the highest mean observed among participants aged 50–59 years (1.16; 95% CI: 0.87–1.45) and the lowest among those aged 80 years or older (0.69; 95% CI: 0.45–0.93). The ZINB model confirmed the adequacy of the approach (alpha: p<0.001; zero-inflation term: p<0.001). In the count process, age was a significant predictor: individuals aged 80 years or older had a 20% lower expected discrimination rate (IRR = 0.80; 95% CI: 0.67–0.97) compared to those aged 50–59 years. Conclusions: These findings support the "Paradox of Perceived Discrimination", indicating that reports of unfair treatment tend to decline with advancing age despite increased objective vulnerability. This trend potentially reflects psychosocial self-protection mechanisms and internalized age-related beliefs, suggesting the need for intersectional approaches in health policies.
Related articles
Related articles are currently not available for this article.