^aOffice of Research and Development, United States Environmental Protection Agency, 1200 Pennsylvania Ave, Washington, District of Columbia 20460, USA

Correspondence: Matthew Lorber, Office of Research and Development, United States Environmental Protection Agency, NCEA (8623N), 1200 Pennsylvania Ave, NW, Washington, DC 20460, USA. Tel.: +703 347 8535; Fax: +703 347 8692; E-mail: Lorber.matthew@epa.gov

Received 31 October 2007; Accepted 22 January 2008; Published online 16 April 2008.

Abstract

A simple two-compartment first-order pharmacokinetic model that predicts concentrations of perchlorate in blood and urine was constructed and validated. The model was validated using data from a high-dose experiment in humans where doses and resulting concentrations of perchlorate in blood and urine were well documented. Specifically, data were available for individuals who had been dosed at 0.5, 0.1, and 0.02 mg/kg/day for 14 consecutive days, significantly higher than the average background dose, which is estimated to be less than 0.0001 mg/kg/day. The average measured urine concentration in the high-dose regime during the experiment was 15.4 mg/l compared with an average prediction of 17.3 mg/l. In the medium-dose regime, the average measured was 3.0 mg/l compared with 4.1 mg/l predicted, and in the low-dose regime, the average measured was 0.53 mg/l compared with 0.68 mg/l predicted. For blood, the analogous results include 0.51 mg/l measured compared with 0.54 mg/l predicted in the high-dose regime and 0.12 mg/l measured versus 0.11 mg/l predicted in the medium-dose regime. The model was then used to study background exposures to perchlorate. A national sampling of perchlorate in urine showed a median concentration of 0.0035 mg/l, and this was used to back-calculate a dose of 0.000064 mg/kg/day. This finding was independently verified with the modeling structure of this study, as use of that back-calculated dose of 0.000064 mg/kg/day resulted in predictions of urine concentration with an average virtually identical at 0.0033 mg/l. An examination of literature data on the possible pathways of exposure suggests that the consumption of foods, rather than ingestion of water, dominates background exposures. Daily variation in urine concentration was studied with the model, and it was found that concentrations in the morning hours were lower than concentrations in the afternoon and evening hours, corresponding to the time when most exposure was assumed to occur.