Estimating arsenic biotransfer factors from feed to chicken: a viable approach to animal feed risk assessment

Reliable biotransfer factor (BTF) data are critical for animal-feed studies in human risk assessment. A wide range of BTF values have been reported (0.0015-0.83 d/kg), expressed as the ratio of total arsenic (tAs) concentration (mu g/kg) in chicken to daily intake rate (mu g/d) of tAs in feed (tAs/tAs). Data on inorganic As (iAs) for chicken meat and tAs concentrations for feed were obtained from our related study. Using the linear regression technique developed in this study, we estimated the BTF as 0.016 d/kg, as iAs/tAs (R-2 > 0.7702) and 0.55 d/kg as tAs/tAs (R-2 = 0.9743) for whole chicken meat. From a mass-balance perspective, we suggest that tAs be used as the denominator for the BTF unit. To illustrate our feed-risk assessment method, we analyzed commercial feeds for tAs concentration (n = 79). Consumption data for the general population (n = 2479) were obtained from a Taiwanese total diet study. Bivariate Monte Carlo simulations (n = 10,000) showed that the 95(th) percentile (P95) of estimated daily intake (EDI) was 0.002 mu g/kg body weight (bw)/d as iAs (< the benchmark dose lower limit of 3.0 mu g/kg bw/d). Our results thus show that the commercial chicken feeds assessed in this study are of low health concern for the general Taiwanese population. We discuss the factors that may have affected the assessment, such as the type of animals investigated, type of feeds, feed tested, type of chemical species used for BTF estimation and statistical approach.