Harmful substances in soil and dust may enter into human body through direct inhalation, digestion, and dermal contact causing adverse health effects. Infants and toddlers are less tolerant to toxic substances than adults, the toxic substances will pose higher risk to the children. Soil and dust ingestion thus becomes a critical exposure factor to determine the risk to children. The objective of this study is to establish a soil exposure database for infants and toddlers living near soil contaminated sites. Soils with low pH and high clay content in both contaminated and control sites at northern, middle, and southern parts of Taiwan are collected. Soil, dust, food and children’s feces are sampled to estimate the local mouthing information and quantify soil/dust ingestion for infants and toddlers. Eighteen children are also randomly selected from database of Year 2011 Project to perform duplicate sampling/examination during 2012 winter for seasonal comparison.Results showed that the collected soils during 2012 project with textures classified as loamy sand to sandy loam in the contaminated and control sites. The Ti and Si contents in soils ranged within 1.97-10.0 and 105-384 g/kg, respectively. The Ti and Si contents in dusts ranged within 3,020-11,200 and 542-152,000 mg/kg, respectively. The Ti and Si contents in feces ranged within 112-4,230 and 82.6-2,620 mg/kg, respectively. The Ti and Si contents in foods ranged within 2.09-25.2 and 89.8-466 g/kg, respectively. For the total 84 children in the 2011 and 2012 projects, the median hand-to-mouth and nondietary object-to mouth frequencies were 9.05 and 9.86 contacts/h, respectively. The median hand-to-mouth and nondietary object-to mouth contact time were 0.35 and 0.35 min/h. The median hand-to-moth and object-to-mouth frequencies and duration were negatively correlated significantly. The nondietary object (excluding pacifier)-to-mouth frequencies and medium times for children using pacifier were significantly greater than those without using pacifier. Based on the duplicate results of the 18 children from 2011 project, the soil digestiondetermined by Ti during 2012 winter was significantly greater than that from 2011 summer. The difference in soil digestion determined by Si for these two seasons, however, was insignificant. The soil digestion for children living in sites with middle –to-basic pH loamy sand was significantly greater than those with low pH, highly clayey soil based on Ti but statistically insignificant based on Si. Using tracer element-based mass-balance approach, the mean and medium children (0-5 years old) soil ingestion in 2011 project were estimated at 2,510 and 1,690 mg/day based on Ti content and at 379 and 12.3 mg/day based on Si content, respectively. Inputting the data from 2011 project into the SHEDS-soil/dust model, the mean and medium children (0-2 years old) soil ingestion were 16 and 7 mg/day, respectively. Summarizing the data of 2011 and 2012 projects, it was elucidated that the feces amount, food, nutrient supplement, gastrointestinal transit time, the falling of teeth, and the falling of painting could influence the estimation of soil/dust ingestion. We suggest that thelocal parameters/distributions used to input into human exposure and dose simulation models should be increased, such as hand-to-mouth and nondietary object to mouth contact frequency for children between 2 and 6 years old, ratio of indoor/outdoor residence time, dust loading of indoor floor, and soil-skin adherence. The obtained soil/dust digestion can be compared to those from biomarker mass balance method. The local nondietary digestion factor and soil/dust digestion can thus be established.

contaminated site、soil/dust ingestion、tracer element-based mass-balance、Stochastic Human Exposure and Dose Simulation model