Factors affecting the estimated probabilistic acute dietary exposure to captan from apple consumption

The effect of the number of pesticide residue values below the LOQ/LOD of analytical methods, the variability of residues in individual fruits, mass of fruit units and the number of bootstrap iterations was studied on the probabilistically estimated acute exposure of consumers. The 4720 daily apple consumption data and the results of 1239 apple sample analyses for captan residues, performed within the Hungarian monitoring programme between 2005 and 2011, were used in this study as model matrix. Up to about 95th percentile exposure (µg/(kg bw·day)), simply multiplying each residue in composite samples with each consumption value gave similar estimates to those obtained with the complex procedure taking also into account the mass of and residues in individual fruits. However, the exposure above the 95th percentile calculated with the complex procedure gradually increased with increasing percentile level compared to the simple procedure. Including the high number of non-detects reduced the estimated exposure, which was the highest when only the residues measured in treated fruits were taken into account. The number of bootstrap iterations between 100 and 10,000 did not significantly affect the calculated exposure. The 99.99th percentile exposure amounted to 17.9% of the acute reference dose of 300 µg/(kg bw·day) for women of childbearing age.