The Sixteenth Conference of the International Society for Environmental Epidemiology (ISEE): Abstracts
Rull, Rudolph P.; Ritz, Beate; Shaw, Gary M.
UCLA Center for Health Policy Research
We evaluated the effects of maternal environmental exposure to 59 agricultural pesticides on neural-tube defects (NTDs) ascertained in two birth cohorts born in California between 1987 and 1991. Maternal residential proximity within 1,000 meters of pesticide applications occurring around the month of conception was assessed using a geographic model based on linking California Pesticide Use Reports (PUR) and land-use survey maps to increase the spatial resolution of the PUR. To adjust for multiple comparisons and correlated pesticide exposures, we employed hierarchical logistic regression, assessing agents in subgroups with similar physicochemical properties and comparing the estimates to those obtained from conventional logistic models. In addition, we estimated separate exposure effects for anencephaly and spina bifida, the two main NTD subtypes, and employed a polytomous hierarchical model for these outcomes. The precision of effect estimates based on the hierarchical models substantially improved compared to conventional logistic models, as effect estimates were shrunk towards the means of subgroups. Exposure prevalence to individual pesticides was low (e.g., methomyl had the highest prevalence of exposure, with 8.6% of 731 cases and 5.6% of 940 controls), thus limiting our ability to detect effects for specific pesticides. For NTDs, we observed a moderate effect for benomyl, a fungicide identified as a developmental toxin and endocrine disruptor (hierarchical OR: 2.05; 95% CI: 0.92, 4.55). Examining NTD subtypes in a polytomous hierarchical model, we also observed associations for the following organophosphosus compounds: chlorpyrifos (OR: 1.54; 95% CI: 0.87, 2.73) with spina bifida, and glyphosate (OR: 1.55; 95% CI: 0.85, 2.85) and naled (OR: 1.95; 95% CI: 0.89, 4.28) with anencephaly.
These are the first results from a human study to suggest an increased risk for NTDs from residential proximity to specific pesticides, including agents (e.g., benomyl) classified by the US and California Environmental Protection Agencies as developmental toxins. In toxicological studies of mammals and amphibians, benomyl and its metabolites have been demonstrated to inhibit microtubule formation, which is critical to the differentiation of neural tissue. Chlorpyrifos in rat embryos has been shown to induce apoptosis and inhibit mitosis of neuroepithelial cells during neurulation. Further study is needed to improve our understanding of the mechanisms for teratogenesis resulting from exposure to these agents at low doses in the environment.
© 2004 Lippincott Williams & Wilkins, Inc.