Comparison of residents' pesticide exposure with predictions obtained using the UK regulatory exposure assessment approach

The UK regulatory methods currently used for estimating residents' potential pesticide exposure were assessed to determine whether they provide sufficiently conservative estimates. A non-random sample of 149 residents living within 100 m of fields where pesticides were sprayed provided first morning void urine samples one and/or two days after spraying. Using farmers' spray information, regulatory exposure assessment (REA) models were applied to estimate potential pesticide intake among residents, with a toxicokinetic (TK) model used to estimate urinary biomarker concentrations in the mornings of the two days following the spray. These were compared with actual measured urinary biomarker concentrations obtained following the spray applications. The study focused on five pesticides (cypermethrin, penconazole, captan, chlorpyrifos and chlormequat). All measured cypermethrin urinary biomarker levels were lower than the REA-predicted concentrations. Over 98% and 97% of the measured urinary biomarker concentrations for penconazole and captan respectively were lower than the REA-predicted exposures. Although a number of the chlorpyrifos and chlormequat spray-related urinary biomarker concentrations were greater than the predictions, investigation of the background urinary biomarker concentrations suggests these were not significantly different from the levels expected had no pesticide spraying occurred. The majority of measured concentrations being well below the REA-predicted concentrations indicate that, in these cases, the REA is sufficiently conservative.