Do intersectional analyses of child health inequalities in cross-national data inform public health?

Reducing health inequalities is central to the Sustainable Development Goals (SDGs) and remains a major public health challenge. Intersectionality provides a framework to examine how multiple social determinants jointly shape health outcomes. Multilevel Analyses of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) estimates these combined effects in quantitative data, but their contribution to public health knowledge cross-nationally remains unclear. This study evaluates MAIHDA’s ability to describe intersectional inequalities, generate policy-relevant insights and, identifies methodological limitations. We analyzed Multiple Indicator Cluster Surveys data(2017–2023). Outcomes included early childhood development (ECDI), underweight, stunting, wasting, and overweight in children under five. Intersectional strata were defined by child sex, household wealth, mother’s education, and ethnolinguistic marginalization. Thirteen countries met inclusion criteria, yielding 24 country-outcome models. MAIHDA estimated predicted outcomes across strata and between-strata variance. Intersectional strata explained 1–7% of outcome variance, indicating modest but meaningful discriminatory accuracy. Inequality gaps between best- and worst-off strata ranged from 5.8% in Cuba (overweight) and 8.3% in Bangladesh (wasting) to 37.8%(underweight) and 41.3%(stunting) in Yemen. Wealth was the main heterogeneity contributor, explaining 47–80% of variance in 11 settings (up to 100% with maternal education). Ethnolinguistic marginalization contributed notably in Vanuatu (54%), Laos (40%), and Nepal (26%). Democratic Republic of Congo observed inequalities exceeding additive effects, indicating interaction. MAIHDA maps intersectional inequalities, but differences in country-strata size and survey responses limit comparability. Nonetheless, MAIHDA advances beyond single-determinant approaches by clarifying how social factors jointly structure health inequalities.

Key messages

• Intersectional (MAIHDA) analyses quantify health inequalities and show how social determinants jointly structure health, including context-specific interactions.

• Child health inequalities are primarily structured by wealth, with contextual contributions from education and ethnolinguistic marginalization in developing countries.