OBJECTIVES. This study examines changing patterns of inequalities in US infant, neonatal, and postneonatal mortality rates between 1969 and 2001 by area deprivation and maternal education.
METHODS. A deprivation index was linked to county vital records data to derive annual infant mortality rates by deprivation quintiles from 1969 to 2000. Rates by maternal education were computed for 1986, 1991, 1996, and 2001 using national linked birth/infant death files. Log-binomial regression was used to estimate relative risks of infant mortality by deprivation and time period. Cox regression was used to model overall and birth weight–specific infant mortality risks by maternal education after adjusting for covariates. Temporal disparities were summarized by log-linear regression and inequality indices.
RESULTS. Although absolute disparities have narrowed over time, relative socioeconomic disparities in infant mortality have increased since 1985. In 1985–1989, infants in the most deprived group had, respectively, 36% and 57% higher risks of neonatal and postneonatal mortality than infants in the least deprived group. The corresponding relative risks increased to 43% and 96% in 1995–2000. The adjusted risk of infant mortality was 22% higher in 1986 for mothers with <12 years of education than for those with ≥16 years of education, with the relative risk increasing to 41% in 2001. Disparities were greatest among normal birth weight infants, with education-specific relative risks of neonatal and postneonatal mortality increasing significantly between 1986 and 2001.
CONCLUSIONS. Dramatic declines in infant mortality among all of the socioeconomic groups during 1969–2001 represent a major public health success. However, substantial socioeconomic disparities persisted in both neonatal and postneonatal mortality. Relatively larger declines in infant and postneonatal mortality among higher socioeconomic groups have contributed to the widening gap in mortality since 1985. Persistent disparities in infant mortality may reflect increasing polarization among socioeconomic groups in material and social conditions, smoking during pregnancy, and health care services.