The interaction of prenatal solvent exposures with genetic polymorphisms in solvent-metabolizing enzymes: Evaluation of risk among infants with congenital heart defects

Abstract

Birth defects are a major public health problem, among which cardiovascular malformations (CVM) carry a disproportionate burden of mortality and disability. Identifying preventable risk factors for CVM is crucial, yet few risk factors are known. One class of environmental agent associated with increased risk of CVM is organic solvents, used widely in industry and in household products. Genetic polymorphisms in solvent-metabolizing enzymes have been recently discovered: some individuals carry null mutations in glutathione-S-tranferase (GST) mu and theta genes, which alter the metabolism of solvents. To test the hypothesis that polymorphisms in GSTmu and GSTtheta modify the risk of CVM among infants whose mothers reported solvent exposures during pregnancy, stored blood spots from Maryland's Newborn Screening Program were obtained for 328 cases and 480 controls of the Baltimore-Washington Infant Study, a population-based study of risk factors. DNA extracted from the dried blood spots was analyzed by polymerase chain reaction coupled with HPLC to detect homozygous null mutations. Logistic regression analysis was used to adjust case-control odds ratios for confounders while stratifying on effect modifiers. Black infants with the genotype mu+/theta+ (both genes present) whose mothers reported paint exposures were at 7.6 times higher risk (95% confidence interval 1.2-47.1) than were unexposed black infants with the same genotype to have pulmonic valve stenosis; the odds ratio was 6.0 (1.0-37.3) after adjustment for alcohol consumption. There was no significant solvent-genotype interaction among white infants with this heart defect. Similar results were observed among black mu+/theta+ infants with atrial septal defect: the odds ratio was 15.2 (3.0-78.2) for any solvent exposure and 22.7 (3.7-140.9) for paint exposures, but whites were not at increased risk. However, white infants were at increased risk of developing aortic valve stenosis and coarctation of the aorta if their mothers reported exposure to "primary" solvents, a more pure form of exposure (odds ratio = 3.2 (95% CI 1.2-8.4)), but the risk was independent of the GST genotype. Although based on heterogeneous exposures and small numbers of exposed subjects, these results suggest that variations in GST genes mediate susceptibility of certain heart defects to the effects of prenatal solvent exposures.