Farm children's exposure to herbicides: comparison of biomonitoring and questionnaire data.

BACKGROUND: Pesticide exposure has been associated with various childhood cancers. However, most studies rely on questionnaires, with few using biologic measures of dose. This study was designed to measure herbicide exposure directly in children of farm applicators, and to compare these results with exposure imputed from questionnaire information.
METHODS: Two consecutive 24-hour urine samples were collected from 92 children of Ontario farm applicators who used the herbicides 2,4-D (2,4-dichlorophenoxyacetic acid) or MCPA (4-chloro-2-methylphenoxyacetic acid) for the first time during 1996. The farm applicator completed questionnaires describing his pesticide-handling practices as well as the child's location during the various stages of handling these pesticides.
RESULTS: Approximately 30% of the children on farms using these herbicides had detectable concentrations in their urine, with maximum values of 100 microg/L for 2,4-D and 45 microg/L for MCPA. Children with higher levels were more likely to be boys and to have parents who also had higher mean urinary concentrations. The sensitivity and specificity of a simple indicator of use were 47% and 72%, respectively, for 2,4-D, and 91% and 30%, respectively, for MCPA, using the biomonitoring data as the gold standard.
CONCLUSIONS: Information on living on a farm, or on living on a farm where a specific pesticide is used, is not enough to classify children's exposures. Given this potential for misclassification, we urge incorporation of biomonitoring studies in subsets of children at least to estimate the extent of misclassification.