Quantitative measures of recent and lifetime agricultural pesticide use are associated with increased pesticide concentrations in house dust

Objective: Elevated pesticide concentrations have been found in dust from homes with residents who use agricultural pesticides, but few studies have compared these concentrations to quantitative measures of their use. We evaluated household pesticide dust concentrations in relation to quantitative, active ingredient-specific metrics of agricultural pesticide use in the Biomarkers of Exposure and Effect in Agriculture Study.

Methods: Participants provided vacuum dust samples (2013-2018) and information regarding recent (last 12 months) and lifetime pesticide use. Thirty-two pesticide analytes were measured in 295 dust samples from 213 participants; 54 had repeated measurements (median = 96 days between visits). We used mixed-effects quantile regression models to estimate relative differences in pesticide concentrations for recent and lifetime agricultural use (number of days, intensity-weighted days), recent home/garden use (yes/no), and household characteristics. Only household characteristics were examined for dacthal because of no use information. We calculated intraclass correlation coefficients (ICCs) to evaluate temporal variability. We report only descriptive statistics for pesticides with detection rates <25 %.

Results: For currently used pesticides, quantitative measures of recent agricultural use were associated with significantly increased household pesticide dust concentrations for malathion, metolachlor, acetochlor, cyfluthrin, and atrazine (p-trends < 0.001), but not permethrin. Similarly, quantitative measures of lifetime use were associated with increased concentrations of malathion, metolachlor, carbaryl, diazinon, and atrazine (p-trends < 0.001), but not permethrin, chlorpyrifos, or chlorothalonil. For banned pesticides, ever agricultural use was associated with elevated chlordane and heptachlor concentrations and non-significantly elevated dieldrin concentrations, but not lindane, p,p-DDD, p,p-DDE, or p,p-DDT. Recent home/garden use predicted increased malathion, carbaryl, and cyfluthrin concentrations. ICCs (range = 0.57-0.90) suggested moderate to high correlation over 3-6 months. Detection rates were <25 % for alachlor, butylate, EPTC, metribuzin, simazine, carbofuran, coumaphos, as well as for three banned pesticides (cyanazine, aldrin, endosulfan).

Conclusions: Household pesticide dust concentrations were strongly associated with the frequency of agricultural pesticide use.