Determination of cadmium relative bioavailability in contaminated soils and its prediction using in vitro methodologies.

In this study, cadmium (Cd) relative bioavailability in contaminated (n = 5) and spiked (n = 2) soils was assessed using an in vivo mouse model following administration of feed containing soil or Cd acetate (reference material) over a 15 day exposure period. Cadmium relative bioavailability varied depending on whether the accumulation of Cd in the kidneys, liver, or kidney plus liver was used for relative bioavailability calculations. When kidney plus liver Cd concentrations were used, Cd relative bioavailability ranged from 10.1 to 92.1%. Cadmium relative bioavailability was higher (14.4-115.2%) when kidney Cd concentrations were used, whereas lower values (7.2-76.5%) were derived when liver Cd concentrations were employed in calculations. Following in vivo studies, four in vitro methodologies (SBRC, IVG, PBET, and DIN), encompassing both gastric and intestinal phases, were assessed for their ability to predict Cd relative bioavailability. Pearson correlations demonstrated a strong linear relationship between Cd relative bioavailability and Cd bioaccessibility (0.62-0.91), however, stronger in vivo-in vitro relationships were observed when Cd relative bioavailability was calculated using kidney plus liver Cd concentrations. Whereas all in vitro assays could predict Cd relative bioavailability with varying degrees of confidence (r(2) = 0.348-0.835), large y intercepts were calculated for a number of in vitro assays which is undesirable for in vivo-in vitro predictive models. However, determination of Cd bioaccessibility using the intestinal phase of the PBET assay resulted in a small y intercept (5.14; slope =1.091) and the best estimate of in vivo Cd relative bioavailability (r(2) = 0.835).