Diabetes in pregnancy and cesarean delivery.

Abstract

To evaluate the effect of diabetes during pregnancy on cesarean delivery and to determine whether the association between diabetes during pregnancy and cesarean delivery is mediated by birth weight. South Carolina 1993 birth certificates were matched through a unique identifier with infant and maternal hospital discharge records for the same year, yielding a total study population of 42,071 singleton births. Adjusted odds ratios (ORs) and 95% CIs were determined for the association between diabetes in pregnancy and cesarean delivery through multiple logistic regression, controlling for maternal age, race, education, number of prenatal care visits, length of gestation, birth weight, and a number of medical indications. Of the study population, 0.7% were pregnancies complicated by preexisting diabetes, 2.9% were pregnancies complicated by gestational diabetes, and 23.4% were cesarean deliveries. After controlling for confounders, including birth weight, cesarean delivery was strongly associated with both preexisting diabetes (OR [95% CI] 6.20 [4.47-8.61]) and gestational diabetes (1.71 [1.41-2.07]). The estimates remained essentially unchanged without birth weight in the model, and were substantially higher in analyses restricted to deliveries without common medical indications for cesarean delivery. Both preexisting and gestational diabetes increase the risk for cesarean delivery, independent of the effect of birth weight. The association is markedly greater among women without other medical indications for cesarean delivery. The increased risk of cesarean delivery for women with diabetes is mediated through other factors, which may include practice patterns and physician referrals to high-risk care.