一律基準を組み入れた食品残留農薬リスク評価と残留農薬実測結果の不確実性を加味した曝露評価

Abstract

We collected and analyzed the acceptable daily intakes (ADI) of 436 agricultural chemicals (include pesticides, veterinary drugs and feed additives) selected from 793 chemicals listed in so-called “positive list” system which the Ministry of Health, Labour and Welfare (MHLW) implemented, by making reference to the MHLW, JMPR, and JECFA. Theoretical maximum daily intakes (TMDI) were calculated for the 793 chemicals from provisional maximum residue limits (MRLs) and food assumptions. We calculated the ratios of TMDI to ADI (TMDI/ADI) and analyzed contribution to TMDI/ADI from former MRLs, provisional MRLs and the uniform limit. The TMDI/ADI strongly increased due to the implementation of the provisional MRLs, so that detailed exposure assessment should be performed for many agricultural chemicals.Three types of estimated daily intakes (EDI) were also calculated from each food assumption and average residue concentrations of an agricultural chemical for each food based on inspection results. In the general methods, the EDIND50 was calculated from detected samples and non-detected samples substituted by a half of quantitative limit. The EDIND0 was calculated from the substitution of zero for non-detected samples and compared to EDIND50. In the new methods that we developed, EDI1n was calculated from lognormal distribution function of residue concentrations determined from detected ratio, quantitative limit and “typical” standard deviation. In order to determine a “typical” standard deviation, sets of average (μ) and standard deviation (σ) of the logarithmic concentration were computed for pairs of an agricultural chemical and a food which have a detected range in the inspection results. We indicated that the EDIND50 was decided mainly by the quantitative limits. Although the EDIND50 was not affected by a sample number, the EDI1n was affected by uncertainty depending on a sample number. Exposure assessment methods like EDI1n should be developed in order to utilize information effectively.