Contribution of county-level socioeconomic indicators to racial or ethnic differences in neonatal anthropometry in the USA: a prospective cohort study.

Introduction: Racial and ethnic differences in fetal growth and birth size in the USA have not been adequately explained by individual-level socioeconomic status (SES) factors. We explored whether differences may be partially explained by county-level indicators of SES. Methods: We linked participant zip codes from the National Institute of Child Health and Human Development Fetal Growth Studies (2009-2013; n=1614) to county-level US census data to calculate a neighbourhood deprivation index, education isolation index and two indices of segregation: racial isolation and evenness. Using causal mediation methods, we evaluated the extent to which racial/ethnic differences in neonatal anthropometrics could be eliminated in a hypothetical setting where everyone lived in counties with high resource availability and racial/ethnic integration. Results: Setting racial evenness to levels consistent with the highest diversity eliminated 79.9% of the difference in birth weight between non-Hispanic White and non-Hispanic Black and all the difference (106.3%) in birth weight between Hispanic and non-Hispanic White individuals. Setting racial evenness, racial isolation and education isolation to levels consistent with higher diversity and education was also associated with similar reductions in differences for other anthropometric measures. Conclusions: Our findings suggest that, in a hypothetical scenario where everyone lived in counties with low deprivation or segregation, race/ethnic differences in neonatal anthropometry may substantially decrease or be eliminated. Our results also highlight the importance of considering community-level and structural factors in analyses of race/ethnic health disparities.