Drinking water contribution to aggregate perchlorate intake of reproductive-age women in the United States estimated by dietary intake simulation and analysis of urinary excretion data

Estimates of perchlorate intake by the US population can be derived from either urinary excretion data or through simulation of dietary intake. Estimates from surveys of urinary excretion (NHANES) are subject to substantial uncertainty owing to the small numbers of subjects for which data are currently available. In addition, current excretion estimates are derived from spot'' urine samples and include a component of short-term (intra-day) variability that may give biased estimates of the variability in average daily intakes. Previous dietary estimates have generally not included any contribution from drinking water, owing to a lack of data related to perchlorate concentrations in water supplies. In this paper, we derive simulation (Monte Carlo) estimates of dietary perchlorate intake distributions for reproductive-age women, which include explicit contributions from drinking water, and compare them to estimates based on urinary excretion. Perchlorate concentrations in water were estimated based on measurements from the US Environmental Protection Agency's UCMR1 database, and from other regional studies of perchlorate contamination. We find that including the drinking water contributions in the dietary simulations yields increases in the population's geometric mean perchlorate intake of 3-8 percent, with a conservative maximum of about 24 percent, compared to intakes estimated based on food intake alone. The intake distributions estimated from dietary and water consumption were found to be very similar to estimates based on creatinine-adjusted perchlorate excretion data from the NHANES, except for having lower population variability. When the dietary simulation data were adjusted to include a contribution from short-term variability similar to that in the spot'' urine samples, the variability in the NHANES and diet-derived estimates were found to be very similar. Our analyses indicate that a reasonable upper-bound estimate for the 95th percentile perchlorate intake among women of reproductive age in the US is on the order of 1.5 x 10(-4) mg/kg/day. Journal of Exposure Science and Environmental Epidemiology (2010) 20, 288-297; doi:10.1038/jes.2009.50; published online 23 September 2009