PubMedDate not supplied
Linked birth and death records provided the population for a study of trends in low birth weight (LBW) rates in Baltimore between 1972 and 1977 and of the effect of changes in the characteristics of the childbearing population on these trends. The impact of shifts in the birth weight distribution on neonatal mortality rates was also investigated. Trends were analyzed for unstandardized LBW rates as well as for rates standardized on the distributions of maternal age, education, gravidity, prior pregnancy losses, and marital status.Between 1972 and 1977, the 1,500 and 2,000 gm rates rose significantly by approximately 1 infant per 1,000 live births per year among whites and 2 infants per 1,000 live births among nonwhites. Despite declines in rates for most weights, the effect of these increases was a rise in neonatal mortality rates for both races, but especially for nonwhites.The population of women delivering in Baltimore in 1977 became slightly older, slightly more educated, and of higher gravidity than in 1972, but these changes had little impact on yearly fluctuations in LBW rates. In contrast, increases in births to unmarried women and to women with at least one prior pregnancy loss were related to rising LBW rates. For both races, standardization on marital status and prior pregnancy losses diminishes the increase in the LBW rate over the study period, especially when standardization is performed simultaneously for both variables. These findings hold within maternal age, education, and gravidity groups. However, the LBW rates for nonwhite teenage mothers and for nonwhite women with 12 years of less education increased significantly over the study period, regardless of standardization.