化学污染物慢性膳食暴露评估全概率模型构建研究

发布时间：2018-11-06 08:06

【摘要】：目的提高食品安全风险评估精度,构建化学污染物慢性膳食暴露评估全概率模型。方法利用我国膳食调查、污染物监测数据以及相应的人口学资料构建膳食暴露评估全概率模型。通过在消费量和污染物总体中进行蒙特卡洛(Monte Carlo)抽样,匹配相乘后得到暴露量的概率分布;构建贝塔二项正态分布(Betabinomial and normal,BBN)模型,将横断面调查获得的短期暴露量近似"拉伸"为长期(慢性)暴露量;通过Monte Carlo模拟和自助法(Bootstrap)对人群膳食暴露量进行变异度和不确定度分析。结果以江苏省居民铅膳食暴露评估为例,构建了化学污染物慢性膳食暴露全概率评估模型。全概率模型和半概率模型比较显示:两种模型计算的暴露量均值接近,但全概率模型计算结果的变异度大于半概率模型,表现为其低端百分位数小于半概率模型,高端百分位数大于半概率模型。结论化学污染物慢性膳食暴露全概率模型评估结果较半概率模型保守。
[Abstract]:Objective to improve the accuracy of food safety risk assessment and to establish a full probability model for the assessment of chronic dietary exposure of chemical pollutants. Methods A full probability model of dietary exposure assessment was established by using dietary survey, pollutant monitoring data and corresponding demographic data in China. The probability distribution of exposure is obtained by Monte Carlo (Monte Carlo) sampling in consumption and pollutant population. The model of beta binomial normal distribution (Betabinomial and normal,BBN) was constructed, and the short-term exposure obtained by cross-sectional survey was similar to "stretching" as long term (chronic) exposure. The variation and uncertainty of dietary exposure were analyzed by Monte Carlo simulation and self-help (Bootstrap). Results A total probability assessment model of chronic dietary exposure of chemical pollutants was established by taking the assessment of dietary exposure of Jiangsu residents as an example. The comparison between the full probability model and the half probability model shows that the mean value of the exposure calculated by the two models is close, but the variation of the calculated result of the full probability model is greater than that of the half probability model, which shows that the low end percentile of the model is smaller than the half probability model. The high-end percentile is larger than the half probability model. Conclusion the total probability model of chronic dietary exposure of chemical pollutants is more conservative than the semi-probability model.
【作者单位】：东南大学公共卫生学院流行病与卫生统计学系;
【基金】：国家自然科学基金资助项目(81172769)
【分类号】：R155

【参考文献】