Abstract
Introduction Individual- and neighborhood-level social determinants of health (SDOH) have been assessed separately in pregnancy, but their relationship to one another remains uncertain. We investigated the intersectionality of three neighborhood-level SDOH measures with three individual-level SDOH measures. This was done to examine the concomitant experiences of multiple SDOH in pregnancy. Methods A secondary analysis of data from the Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-To-Be. We assessed three neighborhood-level SDOH measures using geocoded participant home addresses in the first trimester at the census-tract level: (1) high socioeconomic disadvantage (in tertiles) by the 2015 Area Deprivation Index, (2) inadequate food access by the USDA Food Access Research Atlas, and (3) low walkability by the EPA National Walkability Score. We assessed three individual-level SDOH measures: low household income, lower educational attainment, and Medicaid insurance. We examined the combinations of these three neighborhood SDOH and three individual SDOH measures by graphical visualization and using statistical tests to assess overall differences in the distribution of these measures. Results Of 9588 nulliparous individuals, adverse neighborhood-level SDOH [high socioeconomic disadvantage (28%), inadequate food access (24%), and low walkability (66%)] and adverse individual-level SDOH [low household income (19%), lower educational attainment (23%), and Medicaid insurance (33%)] were common in early pregnancy. Six percent of individuals lived in a community with all three adverse neighborhood-level SDOH measures. Of those living in a community with at least two neighborhood-level SDOH measures, 23% lived in areas with inadequate food access and low walkability, 19% with high socioeconomic disadvantage and low walkability, and 1% with high socioeconomic disadvantage and inadequate food access. Overall, 23% lived in a community with no adverse neighborhood-level SDOH, and among this group, 88% had no adverse individual-level SDOH. There were significant differences in adverse individual-level SDOH based on whether individuals lived in a community with all three adverse neighborhood-level measures [low household income (39%), lower educational attainment (44%), Medicaid (55%)], any two measures [low household income (22%), lower educational attainment (27%), Medicaid (37%)], or only one measure [low household income (14%), lower educational attainment (17%), Medicaid (27%)] (p < 0.001 for all). Conclusion Among nulliparous individuals in early pregnancy, the frequency of adverse individual-level SDOH was generally higher when they lived in communities with more adverse neighborhood-level SDOH. Future approaches that identify and classify the multifaceted and multilevel nature of structural determinants as they relate to pregnancy outcomes are needed.