Do biological, sociodemographic, and behavioral characteristics explain racial/ethnic disparities in preterm births?

Many studies find racial/ethnic disparities in a diverse set of birth outcomes. However few empirical studies have examined the existence and possible explanations for racial/ethnic disparities in preterm births using a diverse set of racial/ethnic categories and a nationally representative sample of births. This research fills that gap. Using data from the US Early Childhood Longitudinal Study - Birth Cohort (ECLS-B), this research first explores the distribution of biological, sociodemographic, and behavioral characteristics of mothers and infants based on seven categories of maternal race/ethnicity. Next, multivariable logistic regression models are estimated in a nested manner to test for possible explanations for racial/ethnic disparities in preterm births. Lastly, race-stratified models are estimated to better elucidate the mechanism leading to racial/ethnic disparities in preterm births. Results from the chi-square tests of significance for racial/ethnic differences indicate that all variables used in this analysis, except for infant's gender, differ significantly based on maternal race/ethnicity. Results from the full multivariable logistic regression model finds that the only racial/ethnic disparity found in preterm births is observed for infants born to Native American mothers compared to non-Hispanic white mothers, once all variables are controlled for in the model. Race-stratified models indicate that maternal health complications and prenatal care adequacy offer the most potential in explaining remaining racial/ethnic disparities in preterm births. Results from this research support the need to increase access to appropriate and timely prenatal care for women of all races/ethnicities in an effort to reduce racial/ethnic disparities in preterm births.