Documenting and explaining birthweight trends in the United States, 1989-2007

Abstract

Birthweight is one of the most important health indicators for a newborn infant. Birthweight at either the lower or higher end is associated with adverse health outcomes in later life. In recent years, birthweight distribution in the United States has shifted to the lower end. This dissertation uses US vital statistics data from 1989 to 2007 to document recent birthweight trends in the US and examines the possible causes behind the trends. Results are reported for all births and by race/ethnicity/nativity. Descriptive analysis suggests that the lowering birthweight trend is the result of the rapid increase of lower-birthweight multiple births and decreasing birthweight among singleton births. The lowering birthweight is reflected in all birthweight measures. Low-birthweight rate is rising, mean birthweight is declining, and the proportion of macrosomic infants is decreasing. While this trend is most pronounced among US-born non-Hispanic whites and least among non-Hispanic blacks, it is prevalent among all race/ethnicity/nativity groups. Regression results suggest that much of the birthweight trend can be explained by shortened gestational age but common maternal socio-demographic, health and behavioral, and health care and medical intervention factors cannot fully explain the birthweight trend. Regression decomposition concludes that both the trends in maternal factors and the changes in the effects of these factors on birthweight contribute to the birthweight trend. Trend in gestational age is the biggest contributor, contributing more than 100% to the birthweight trend, while improvement in education, reduction of smoking during pregnancy and improvement in prenatal care have slowed down the birthweight decrease. Further research needs to be done to identify factors leading to the recent birthweight trend that are not available from the vital statistics.