Infant mortality of 1980-1982 birth cohorts are analyzed in terms of prior background variables (mother's education, race/ethnicity -- Anglo, Black, and Hispanic, rural-urban residence, and mother's marital status -- legitimacy of baby) and proximate variables (trimester of first prenatal care, sex of baby, length of pregnancy -- gestational age, mother's age, birthweight, plurality, birth order, prior fetal or infant deaths, and hospital birth) according to characteristics of death (age of baby at death and cause of death). Cause of death is classified variously as underlying cause, total mentions of a single cause, and multiple-cause combinations. The unique data set is obtained by linkage of birth-infant death merged records produced by the State of Florida with multiple-cause coded records of death provided by NCHS. The files include 414,832 births and an estimated 5,400-5,700 infant deaths, of which about forty percent are for blacks. Analysis uses cross-classificatory and both logic and hazards procedures. The findings should (a) cast light on the relative importance of different classifications of cause of death for infant mortality, (b) identify the most significant prior and proximate variables that contribute to infant mortality in a cause-of-death framework, (c) measure direct effects on the variables, (d) point to vital data limitations in need of further attention, and (e) inform those who make and enforce health policies of effects on infant mortality that are amenable to possible modifications.
| Eberstein, I W; Nam, C B; Hummer, R A (1990) Infant mortality by cause of death: main and interaction effects. Demography 27:413-30 |
| Nam, C B; Eberstein, I W; Deeb, L C (1989) Sudden infant death syndrome as a socially determined cause of death. Soc Biol 36:1-8 |
