Selected environmental exposures and risk of neural tube defects

With a birth prevalence of 1 in 1000, neural tube defects (NTD)s contribute considerably to morbidity and healthcare costs. Known genetic and environmental (non-inherited) risk factors for NTDs account for a small portion of risk, suggesting unidentified risk factors. In animal studies, maternal alcohol and pesticide exposures, independently, led to excess neural cell death, resulting in too few cells for neural tube closure. Human studies report no association between alcohol exposure and NTDs, but small to moderate positive associations for pesticide exposure. Such human etiologic studies of NTDs require a large base population, but frequently include only live births. Exclusion of cases by pregnancy outcomes may create ascertainment and response bias, complicating interpretation of findings.

Using data from the National Birth Defects Prevention Study (NBDPS) and the Iowa Registry for Congenital and Inherited Disorders (IRCID), the independent effects of maternal periconceptional (1 month prior through 2 months postconception) alcohol and occupational pesticide exposure on the development of NTDs were examined, and differences in Iowa NTD cases were characterized by pregnancy outcome.

Maternal reports of alcohol exposure were obtained for 1223 NTD case infants and 6807 control infants. Adjusted odds ratios, estimated using multivariate logistic regression, were near unity for NTDs by any maternal alcohol exposure, binge episode(s), and type(s) of alcohol consumed. Occupational pesticide exposure was assigned by industrial hygienists for mothers of 502 case and 2950 control infants. Adjusted odds ratios for any exposure and cumulative exposure to any pesticide, insecticides only, and insecticides + herbicides + fungicides were near unity for NTDs. Insecticide + herbicide exposure was positively associated with spina bifida. Among the 279 Iowa NTD case infants ascertained by the IRCID, 167 live births and 112 were other pregnancy outcomes (fetal deaths and elective terminations), which increased in proportion over time. Selected infant and maternal characteristics of live births and other pregnancy outcomes were similar. NBDPS eligibility varied significantly by pregnancy outcome, but participation rates did not. NTD case mothers were similar to Iowa NBDPS control mothers.

Efforts were made to improve upon prior etiologic studies of these exposures and NTDs, including increased sample size and improved exposure specificity. Some exposure strata (e.g., herbicides only) and outcome strata (e.g., other rare subtypes) were limited by small numbers. All results may have been affected by response and ascertainment bias. Future studies should aim to use similarly detailed exposure classification methods, increase sample size in less prevalent NTD subtypes, and improve ascertainment of fetal deaths.