Racial inequity in preterm delivery among college-educated women: The role of racism.

BACKGROUND

Non-Hispanic Black (NHB) women face a 50% increased risk of delivering preterm compared to non-Hispanic White (NHW) women in the United States. Sociodemographic and pregnancy risk factors do not fully explain this inequity. This inequity exists even among women with a college education, although recent empirical analysis on racial inequities in preterm delivery (PTD) among college-educated women is lacking. Furthermore, the contribution of preconception risk factors to the racial inequity in PTD has not been examined.

OBJECTIVES

To determine whether: (i) there is a NHB-NHW inequity in PTD among college-educated women; (ii) the prevalence of known, measured sociodemographic, pregnancy, and preconception PTD risk factors differs between NHB and NHW college-educated women; (iii) equalising the distribution of risk factors between college-educated NHB and NHW women reduces or eliminates the racial inequity in PTD.

METHODS

We analysed US natality data from 2015 to 2016 among women with a college degree or higher (n = 2 326 512). We calculated frequencies of sociodemographic, pregnancy, and preconception risk factors among all women and separately by race/ethnicity. We used modified Poisson regression models to estimate the association between race/ethnicity and PTD controlling for known, measured sociodemographic, pregnancy, and preconception factors.

RESULTS

The largest percentage point differences in risk factors between NHW and NHB women were observed for marital status, trimester of care initiation, body mass index, and birth interval. Among college-educated women, the unadjusted risk of PTD for NHB women was 1.77 (95% CI 1.74, 1.79) times the risk for NHW women. After controlling for sociodemographic, pregnancy, and preconception factors, this attenuated to RR 1.47 (95% CI 1.45, 1.49).

CONCLUSIONS

A racial inequity in PTD persists among college-educated women. Racism contributes to the NHB-NHW inequity in PTD, in part, through its influence on known sociodemographic, pregnancy, and preconception risk factors for PTD and, in part, through unmeasured pathways.