Risk assessment for furan contamination through the food chain in Belgian children

Authors:
Publisher website
Google Scholar
Abstract
Young, old, pregnant and immuno-compromised persons are of great concern for risk assessors as they represent the sub-populations most at risk. The present paper focuses on risk assessment linked to furan exposure in children. Only the Belgian population was considered because individual contamination and consumption data that are required for accurate risk assessment were available for Belgian children only. Two risk assessment approaches, the so-called deterministic and probabilistic, were applied and the results were compared for the estimation of daily intake. A significant difference between the average Estimated Daily Intake (EDI) was underlined between the deterministic (419 ng kg−1 body weight (bw) day−1) and the probabilistic (583 ng kg−1 bw day−1) approaches, which results from the mathematical treatment of the null consumption and contamination data. The risk was characterised by two ways: (1) the classical approach by comparison of the EDI to a reference dose (RfDchronic-oral) and (2) the most recent approach, namely the Margin of Exposure (MoE) approach. Both reached similar conclusions: the risk level is not of a major concern, but is neither negligible. In the first approach, only 2.7 or 6.6% (respectively in the deterministic and in the probabilistic way) of the studied population presented an EDI above the RfDchronic-oral. In the second approach, the percentage of children displaying a MoE above 10,000 and below 100 is 3–0% and 20–0.01% in the deterministic and probabilistic modes, respectively. In addition, children were compared to adults and significant differences between the contamination patterns were highlighted. While major contamination was linked to coffee consumption in adults (55%), no item predominantly contributed to the contamination in children. The most important were soups (19%), dairy products (17%), pasta and rice (11%), fruit and potatoes (9% each).

This publication has 34 references indexed in Scilit: