Trends and differentials in higher-birthweight infants at 28–31 weeks of gestation, by race and Hispanic origin, United States, 1990–2002

Birth certificate gestational age data based on the date of the mother's last menstrual period (LMP) are considered problematic. Of particular concern are birthweight distributions for infants reported on the birth certificate as having been delivered at 28–31 weeks' gestation; these distributions have been shown to be distinctly bimodal. The 'second curve' of the birthweight distribution at 28–31 weeks includes implausible birthweight/gestational age combinations and, thus, has been hypothesised to represent erroneous gestational ages due to misidentification of the date of LMP. It has been suggested that such 'misclassification' has declined in recent years and that this change can affect trends in preterm birth rates (<37 weeks' gestation), particularly rates among non-Hispanic black infants. This present study used primarily simple and multivariable analyses to review trends and differentials in birthweight distributions at 28–31 weeks by race and Hispanic origin of the mother. It aggregated data for the years 1990–92 and 2000–02 from the US vital statistics Natality files.
Over the decade, the percentage of births in the second curve declined for all births and for each racial and Hispanic origin group studied. The largest decline was observed for non-Hispanic blacks; the smallest for Hispanic births. Later initiation of prenatal care, younger maternal age, lower educational attainment, higher birth order and vaginal and singleton delivery were positively associated with a larger second curve, suggesting misclassification of gestational age. Declines in the second curve over the study period were suggested to contribute significantly to the observed decrease in overall preterm birth rates for non-Hispanic black births. Further analysis is needed to estimate the influence of reporting error on preterm birth rates by race and Hispanic origin.