Background: Large-for-gestational-age (LGA) or macrosomic infants are associated with adverse maternal and neonatal outcomes. It is unclear if these associations are stronger using customised birthweight centiles. We compared outcomes between term infants defined macrosomic by birthweight >4000 g (Macro4000) or LGA by population centiles (LGApop) with those defined LGA by customised centiles (LGAcust).
Methods: This is a prospective cohort study of 2668 term nulliparous women recruited into the Screening for Pregnancy Endpoints (SCOPE) study centres in Auckland, New Zealand and Adelaide, Australia. Maternal (caesarean delivery, postpartum haemorrhage) and infant (severe neonatal morbidity/mortality and admission to neonatal intensive care) outcomes in Macro4000 and LGA groups were compared with appropriate-for-gestational-age infants by customised centiles using logistic regression.
Results: Customised centiles defined fewer infants as LGA (10.3% LGAcust, 14.8% Macro4000, 11.2% LGApop). However customised centiles showed stronger association with adverse outcomes. Pre-labour and intrapartum caesarean section were increased twofold in LGAcust pregnancies, including those that were not Macro4000 or LGApop. Postpartum haemorrhage was increased twofold in mothers of LGAcust infants only when infants were also LGApop. Severe neonatal morbidity/mortality or admission to neonatal intensive care was increased twofold in LGAcust who were also either Macro4000 or LGApop. Importantly 52.3% of Macro4000 and 25.5% of LGApop infants were AGAcust and not at increased risk of most adverse maternal or neonatal outcomes.
Conclusions: The use of customised centiles are more strongly associated with adverse birth outcomes and its use should be considered in the definition of LGA.