[Estimation of the Distribution of Dietary Exposure to Lead Using Two-Dimensional Monte Carlo Simulation: An Attempt to Apply to Total Diet Samples].

Abstract

The average dietary exposure to lead (Pb) in male and female Japanese individuals >1 year of age was estimated using 280 total diet samples representing 14 food groups from 10 areas over a two-year period. A probabilistic exposure estimation was performed using a two-dimensional Monte Carlo simulation (2D-MCS) with a Bayesian estimation that consided the uncertainty of the estimation process. The Bayesian estimation was performed using the likelihood function with cumulative distribution function between the lower and upper boundary values for no-detected values. The median dietary exposure to Pb was estimated as 5.82 μg/person/day. The 90% interval was 2.51-16.9 μg/person/day. Comparison with previously reported Pb exposure values indicates that the estimation of Pb exposure distribution using total diet samples is reasonable. The contribution to Pb exposure was highest in the order of food group 8 (light-colored vegetables, mushrooms, and seaweeds: 20.0±16.1%)>food group 1 (rice and rice products: 12.3±19.0%)>food group 10 (fish and shellfish: 10.5±13.9%). Owing to the high uncertainties of contribution ratios, it was not possible to identify dominant food groups contributing to Pb exposure. However, it was evident that the uncertainty of the estimation of Pb exposure was influenced by the uncertainty of Pb concentration than the uncertainty of food consumption rate. In particular, the effect of uncertainty from the Pb concentration of the food group 1 was 68.2%. When the margin of exposures were calculated, the estimated probabilities that a value would be <1 were 14.5% for developmental neurotoxicity to children (1-6 years old), 0.13% for blood pressure and 0.93% for kidney disease in Japanese individuals ≥1 year of age. The findings suggest that the health risk due to dietary Pb exposure is small but not negligible.