Aim: A previous large case-control study had documented association between large for gestational age birth weight and stillbirth. We sought to replicate this novel finding.Methods: Retrospective cohort with nested case-control analyses. The data source was a large Canadian perinatal database. Stillbirth was defined as no signs of life at delivery ≥23-weeks gestation. Small and large for gestational age at birth weights were defined using standard Canadian, American, and ultrasound norms.Results: The study population was comprised of 693,186 live births and 3275 stillbirths. Using population norms, large for gestational age (LGA) at birth was not associated with stillbirth in adjusted analysis (OR 0.57, 95% CI [0.45, 0.71]). A statistically significant interaction between pre-existing diabetes and LGA was observed (OR 2.48, 95% CI [1.20, 5.15]). Further case-control analysis found if ultrasound norms were used and controls were selected by date of delivery then LGA was associated with stillbirth (OR 1.91, 95% CI [1.64-2.22]). However, if controls were matched by gestational age at birth then the association was no longer apparent (OR 0.98, 95% CI [0.85-1.13]).Conclusions: Large for gestational age at birth is not associated with stillbirth in the general population. Previously reported association between LGA and stillbirth was likely an artifact of the sampling strategy. LGA may increase the risk of stillbirth in pregnancies complicated by pre-existing and gestational diabetes.
Keywords: Diabetes in pregnancy; large for gestational age; macrosomia; stillbirth.