Linking the California Pesticide Use Reporting Database with Spatial Land Use Data for Exposure Assessment

ehasl research project In regions of intense agricultural production, the potential for exposure to agricultural pesticides has become an important topic of public health concern. Exposure to agricultural chemicals has been associated with cancer, immune system disorders, adverse reproductive outcomes, developmental disorders, and neurological disease. The State of California has developed a Pesticide Use Reporting Database (CPUR), with the objective of providing complete pesticide-use data for evaluating possible associations with human illness clusters. However, the reporting unit for the database is 1 mi2, which may be too large for accurately predicting exposure at the individual residence level necessary for some epidemiological studies. We used the California Department of Water Resources (CDWR) crop map database to improve the crop location attributes of the CPUR database. We generated exposure metrics based on CPUR alone and CPUR linked to CDWR for birth residences in 1988-1994 in a childhood cancer study conducted by the California Department for Health Services. Sixty-six residences had both CPUR and CDWR data for the child's birth year. We calculated metrics predicting the lbs/mile2 of pesticide for 6 pesticides with high use in the study area: herbicides, simazine and trifluralin; insecticides, dicofol and propargite; and fumigants, methyl bromide and metam sodium. We first compared the exposure classification (exposed / not exposed) to each pesticide and evaluated agreement using a chi-squared test and Cohen's kappa. We then assessed differences in predicted lbs/mile2 of pesticide within a 500-meter buffer around each residence between the metrics, a distance used in previous studies of pesticide exposure. Four of the six pesticides, simazine, trifluralin, dicofol, and methyle bromide, indicated similar categorical assignment of exposure for each metric. Cohen's kappa values for these pesticides ranged from 0.11 to 0.66. The Wilcoxon signed-rank test indicated significant differences in estimated lbs/mile2 applied within the residential buffers between the CDWR metric and the CPUR metric for 3 of the pesticides. The medians and interquartile ranges were propargite, CDWR = 0.04 (0.00-0.57) and CPUR = 89.41 (14.47-233.69); simazine, CDWR = 0.00 (0.00-0.03) and CPUR = 8.77 (0.00-50.13); methyl bromide, CDWR = 0.00 (0.00-0.00) and CPUR = 0.00 (0.00-342.27). These pesticide exposure metrics are pending field validation but show promise in predicting potential pesticide exposure. Exposure metrics that refine the locational attributes of pesticide use data such as CPUR may reduce exposure misclassification for subjects with high or low exposure.

EHASL Research Project

Miller, RS; JR Nuckols; RB Gunier; A Hertz; MH Ward; P Reynolds. Linking the California Pesticide Use Reporting Database with Spatial Land Use Data for Exposure Assessment. International Society of Exposure Analysis 12th conference and International Society for Environmental Epidemiology 14th conference, August 11-15, 2002. Vancouver, British Columbia, Canada.

View the Project in .PNG graphic Format ( 204 KB Download )

(click here for help with .png files)