UCLA

Pesticides are a group of chemicals that are used to control harmful insects, weeds, fungi, and other forms of plant or animal life. Individuals may be exposed to pesticides from residues in food or home/garden use or in occupational settings, as well as spray drift from agricultural applications on the fields. Exposure to pesticides can cause a series of chronic adverse effects in humans including reproductive toxicity, neurotoxicity, and carcinogenicity.

In a large population-based sample, we examined whether prenatal exposure to agricultural pesticides contributes to risk of preterm birth or term low birthweight in California. Subjects were children born between 1998 and 2010 had been randomly selected from CA birth records as controls to study autism or childhood cancers. We employed a Geographic Information System (GIS)-based system to estimate residential exposures to agriculturally applied pesticides within 2km of birth addresses for 17 individual pesticides and 3 chemical classes. First or second trimester exposures to selected individual pesticides (e.g., glyphosates OR=1.05, 95% CI: 1.02-1.08; paraquat OR=1.07, 95% CI: 1.03-1.11; imidacloprid OR=1.06, 95% CI: 1.03-1.10) and the classes of organophosphates, carbamates, and pyrethroids were associated with small increases (3-7%) in risk for preterm birth, with stronger estimated effects in females. We did not find enough evidence to support associations between exposures to pesticides and term low birthweight. Future studies that assess and integrate maternal pregnancy exposures at both workplaces and residences are needed.

Record-linkage studies of residential proximity to agricultural pesticides and childhood cancers often only use maternal address at birth or address at cancer diagnosis to assess exposures in early childhood, possibly leading to exposure misclassification. We examined patterns of and identify factors that may predict residential mobility in early childhood, and assessed the impact of mobility on early childhood exposure measures. We found that older age at diagnosis, younger maternal age, lower maternal education, non-Hispanic ethnic background, and other proxies for lower socioeconomic status were predictors of higher residential mobility. There was moderate to strong agreement (kappa=0.7-0.8) between the first year of life exposures assessed at birth address and diagnosis address or LexisNexis addresses, but agreement decreased as the distance increased. These findings suggest that birth residence or diagnosis residence should be used with caution when estimating environmental exposures in early life of children. Future research should consider factors that help correct for the exposure misclassification introduced by residential mobility.

To extend the previous study, we further assessed the effect estimates for early life exposure to pesticide based on registry-acquired addresses, and compared the effect estimates with those using reconstructed address history based on LexisNexis records, using childhood brain tumors as an example. Additionally, we also examined whether selection bias may occur when restricting our study population to children with available LexisNexis records. In the subset of children with all three types of addresses, we found no associations between children’s lifetime exposures to agricultural pesticides of interest and brain tumors. Among these children, effect estimates were generally attenuated when exposures were estimated based on LexisNexis addresses. Interestingly, among the children whose LexisNexis addresses were not available, we observed the largest increased risks for brain tumor and several pesticides based on birth and diagnosis addresses alone. We concluded that using a single source of address to assess children’s early life environmental exposure to pesticides does not account for residential mobility and may cause exposure misclassification that leads to bias in effect estimates. Future research should also be cautious about the potential for selection bias introduced by the limitations of database containing public records.