A GIS-Based Model of Historical Pesticide Exposure on the Risk of Breast Cancer

A geographic information system (GIS)-based model of historical pesticide exposure, set in California’s Central Valley, the most agriculturally productive region in the United States, was used to evaluate risk of postmenopausal breast cancer from hormone-related pesticides in a pilot case-control study. Lifetime residential and occupational histories were geocoded and linked to the California Pesticide Usage Reporting System data combined with land use surveys to determine exposure to specific pesticides. Cases (N=155) were women diagnosed with breast cancer in 2007-8 from the California Cancer Registry and controls (N=150) were recruited from tax assessor and Medicare listings in the same geographic area from 2001-2011.

Compared to unexposed women, those exposed at both residence and workplace had an increased risk of breast cancer from exposure to chlorpyrifos, an estrogen receptor agonist (OR = 1.83, 95% CI: 1.02, 3.28). We found null associations for 1,3-dichloropropene and diazinon, two pesticides without known endocrine-disrupting effects that were included to evaluate the specificity of our model.

Although we observed only a modest increase in breast cancer risk for a selected group of estrogen-mimicking organochlorine insecticides (OR = 1.40, 95% CI: 0.89, 2.19), we found differential effects among women aged 55-64 compared to women aged 65-74 years. Younger women had an increased risk of breast cancer (OR = 1.65, 95% CI: 0.66, 4.15) compared to older women (OR = 1.10, 95% CI: 0.47, 2.59), and although this finding was not statistically significant, it is suggestive of an exposure window that warrants further study. Women exposed to exogenous hormone-like chemicals during critical periods of breast development may be important for breast cancer risk later in life.

 The GIS-based methods presented here provide a more comprehensive exposure assessment using historical data than has been done previously. Methods used in other epidemiologic studies of breast cancer risk and exposure to pesticides typically aggregate pesticides of varying toxicological properties. When we considered exposure to all pesticides combined, we also observed no increase in breast cancer risk, suggesting that less specific methods of measuring pesticide exposure are biased towards the null.