Background: Congenital heart disease (CHD) is the most common birth defect group and a significant contributor to neonatal and infant death. CHD with single ventricle anatomy, including hypoplastic left heart syndrome (HLHS), tricuspid atresia (TA), and various double-inlet ventricle (DIV) malformations, is the most complex with the highest mortality. Prenatal risk factors associated with HLHS have been studied, but such data for DIV are lacking.
Methods: We analyzed DIV cases and nonmalformed controls in the National Birth Defects Prevention Study, a case-control, multicenter population-based study of birth defects. Random forest analysis identified potential predictor variables for DIV, which were included in multivariable models to estimate effect magnitude and directionality.
Results: Random forest analysis identified pre-pregnancy diabetes, history of maternal insulin use, maternal total lipid intake, paternal race, and intake of several foods and nutrients as potential predictors of DIV. Logistic regression confirmed pre-pregnancy diabetes, maternal insulin use, and paternal race as risk factors for having a child with DIV. Additionally, higher maternal total fat intake was associated with a reduced risk.
Conclusions: Maternal pre-pregnancy diabetes and history of insulin use were associated with an increased risk of having an infant with DIV, while maternal lipid intake had an inverse association. These novel data provide multiple metabolic pathways for investigation to identify better the developmental etiologies of DIV and suggest that public health interventions targeting diabetes prevention and management in women of childbearing age could reduce CHD risk.
Keywords: case-control studies; congenital heart disease; double-inlet ventricle.
© 2019 Wiley Periodicals, Inc.