Distinguishing pathological from constitutional small for gestational age births in population-based studies

Abstract

Background Small for gestational age (SGA) can occur following a pathological process or may represent constitutionally small fetuses. However, distinguishing these processes is often difficult, especially in large studies, where the term SGA is often used as a proxy for restricted fetal growth. Since biologic variation in fetal size is largely a third trimester phenomenon, we hypothesized that the definition of SGA at term may include a sizeable proportion of constitutionally small fetuses. In contrast, since biologic variation in fetal size is not fully expressed in (early) preterm gestations, it is plausible that SGA in early preterm gestations would comprise a large proportion of growth restricted fetuses. Aim We compared mortality and morbidity rates between SGA and appropriate for gestational age (AGA) babies. Subjects A population-based study of over 19 million non-malformed, singleton births (1995–04) in the United States was performed. Gestational age (24–44 weeks) was based on a clinical estimate. SGA and AGA were defined as sex-specific birthweight < 10th and 25–74th centiles, respectively, for gestational age. All analyses were adjusted for a variety of confounding factors. Outcome measures Excess mortality risk in SGA and AGA babies. Results On an additive scale, stillbirth and neonatal mortality rates were higher at every preterm gestation among SGA than AGA births, and similar at term gestations. An inverse relationship between gestational age and excess deaths between SGA and AGA babies delivered at < 37 weeks was evident. Conclusions In early preterm gestations, the definition of SGA may well be justified as a proxy for IUGR. In contrast, SGA babies that are delivered at term are likely to be constitutionally small.