Studies On Dietary Exposure Assessment With Small Sample Size Based Bootstrap Method

In developing the dietary exposure assessment work pollutant data is necessary data material. We can only get the low and random pollutant sampling date because of the limits of workload, funds and testing time. How to make the reasonable utilization of existing data is a pivotal point, considering low random pollutant data. The essay mainly discussed method of exposure assessment in the case of small sample data of pollutants and the overall unknown distribution, explore bootstrap method on parameter estimation especially in the point assessment in combination with contaminants data in the drinking water of residents of some river of China, to improve statistical support for accuracy and the references for assessment work.This paper select arsenic sample date as the research object. Pollutants overall mean, standard errors and confidence interval was estimated with the Bootstrap method. The overall mean is0.7860μg/L, and the95%confidence interval is [0.5788,1.0391] when simulation times was2000. Bootstrap method is reasonable and effective for the parameter estimation in the case of the finite sample for exposure assessment compared with traditional parameter estimation analysis method.In addition, exposure assessment of arsenic through drinking water was conduced. The arsenic average exposure levels is0.02992,0.031702,0.026603,0.025724and0.046235μg/kg bw/d. Exposure levels rule is:children> women> men, rural residents> city residents.The MOS of region urban residents and rural residents through drinking water way separately were0.29,0.31; The MOS of Men, women and children separately were0.26,0.25,0.46. The MOS of all the people were less than1, so the risk of non cancer toxicity is acceptable by means of drinking water in the region. the cancer risk rate is3.2×10-6higher than the10-6suggested by EPA.