Bioavailability of arsenic in soil: pilot study results and design considerations.

Bioavailability of arsenic (As) from ingested soil is estimated in a two-period experimental study involving 11 subjects/period. In the first period, a 7-day mass-balance study measured As in food/beverages, urine, and stool to estimate bioavailability of As in food and beverages. Food/beverage As bioavailability (percentage ingested that is not in stool samples) is estimated as 91.0% with a 95% confidence interval given by (84.1%, 97.9%). In the second 7-day study period, subjects were placed on an As suppression diet. In the evening of day 2, each subject ingested a capsule containing 0.63 g of soil, including approximately 111.7 µg of soil-As. The bioavailability estimate of As from food and beverage ingestion during the first 2 days of the second period was 89.7%. Bioavailability of soil-As was estimated over the 5-day period following capsule ingestion, accounting for estimated bioavailability of food/beverage As. Assuming analytic recovery rates of As from combined soil and food/beverage samples are equal, soil-As bioavailability is estimated as 48.7% (95% CI [36.2%, 61.3%]). Relative to bioavailability of As from food/beverage sources, soil-As is estimated to be 54.3% (95% CI [40.3%, 68.4%]) as bioavailable.