The relationship between maternal risk factors and infant mortality was examined by linking 1983 and 1984 birth and infant death certificates. The maternal covariates considered were race, nativity status (foreign vs. native born) age, parity, marital status, and education. Births in Washington, Texas and California were excluded since maternal education is not listed on birth certificate. Analysis was further restricted to only live births in single deliveries and of known birth order. Subjects included 4.4 million white mothers and 926,000 black mothers. Although a large proportion of births were excluded, infant mortality based on all births was very similar to that based on births to mothers whose education level was known; thus, generalizability of findings is applicable. Several maternal risk factor categories were combined into broad maternal risk groups. Low maternal risk was classified as a woman who was married, had attained least 13 years of education, and was either primiparas aged 20 years or older or low parity multiparas aged 20 years or older. Similarly, high maternal risk was defined as a woman who was single, had less than 12 years of education, and was either less than 20 years or high parity multiparas. Although the study evaluated five dependent variables – infant, neonatal, and postneonatal mortality, very low and low birth weight – the article concentrated on the first three outcomes.

Nativity status was the only factor that had stronger effects among black women. Black foreign-born women were less likely to be at high risk compared with black native born women (10% vs. 23%). In contrast, the difference by nativity status among white women was much narrower and in the reverse direction. 4% of births to white native-born mothers and 7% of births to white foreign-born women were at high risk. Infant mortality was 28% lower among black foreign born than native born mothers (12.4 vs.17.3 deaths per 1000 live births), while, among white infants, no difference in infant mortality by nativity status was observed. A two-fold black/white differential in infant mortality was observed. The highest infant mortality rate was associated with multiparas teenagers, regardless of race. But, multiparas teenagers represented only 9% and 3% of the black and white groups respectively and, thus, could not explain the two-fold differential among the two racial groups. Regarding the broad maternal risk groups, a twofold increase in infant mortality was observed among blacks in the high risk category compared with blacks in the low risk maternal category, while near-tripling of white infant mortality rates between the low and high maternal risk groups. Furthermore, as the level of maternal risk group increased, the differential between blacks and whites declined from twofold differential (11.6 vs. 5.7) to 40% differential (21.4 vs. 15.4).

Data from the 1983 Current Population Survey indicated that the black foreign born population was of higher socioeconomic status than the black native born population, but the reverse was observed for the white population. Also, birth certificates in the study showed that foreign-born black women attained higher level of education than black native born, which is consistent with the above findings. Thus, behavioral, cultural, and nutritional factors may account for some of the differences between infant mortality among black foreign-born and native born women. Nevertheless, biases may operate whenever comparisons using migrant populations are constructed. Migrants who left their native country may not be representative of those who remained. Migrants may be either healthier or sicker than those remaining. Also, stresses related to transit may adversely affect the health of the migrants. However, based upon 1980 census, over 60% of foreign-born women lived in the US for at least 5 years. But, stresses related to family separation could still impact the health status of foreign-born, particularly pregnant, women, regardless of time they lived in the US. However, the study did not consider that the black migrant women may have arrived from less industrialized countries and experienced stress related to family separation during pregnancy. Under this scenario, the infant mortality differential between black foreign-born and native born is even greater than shown. Although this initially appears discouraging, it is actually reassuring since this demonstrates that the US white and black infant mortality disparity is not the result of genetics. Thus, the gap can be bridged through effective interventions. Furthermore, data from the birth certificates showed the existence of differential access to prenatal care among the maternal risk groups. However, differential access to care is not sufficient to explain the high difference in infant mortality among black sand whites in the same maternal risk category. Thus, differential teenage multiparas population, differential access to prenatal care, or even genetics can not fully explain the black and white infant mortality differential. Nevertheless, if the infant mortality rate in the low risk group could be achieved by the moderate and high-risk groups, a 30% reduction in infant deaths within each race would occur.