Abstract
Vahratian et al. (1) conclude that the recently observed decline in preterm birth among non-Hispanic African Americans in North Carolina is an artifact related to temporal reductions in the misclassification of full-term births (as preterm births). We are concerned that the bimodal birth weight distribution problem identified by Vahratian et al. (1) in 1989 was observed at 28–31 weeks' gestation but not at 32–36 weeks. This is important, because most of the decline in preterm birth among non-Hispanic African Americans in North Carolina and the United States (2–4) was observed at 32–36 weeks' gestation (from 14.4 percent in 1989 to 12.9 percent in 1999 (1)) rather than at 28–31 weeks' gestation (from 2.2 percent in 1989 to 1.9 percent in 1999 (1)).