Is food type important for in vitro post ingestion bioaccessibility models of polychlorinated biphenyls sorbed to soil?

Soils are sorbents for many organic compounds and children consume relatively large amounts of soil. To improve the estimated health risks from this exposure pathway, we examined the role of co-ingested foods in determining the post-ingestion bioaccessibility (mobilization) of 18 polychlorinated biphenyls (PCBs) sorbed to 10 characterized soils. The bioaccessibility test system (DIN 19738, 2004) was an in vitro, 3-compartment, digestive tract containing salts, protein, and bile. Each soil was fortified with PCBs, then, digestive fluids appropriate to each compartment, were added sequentially. Next, digestive fluid and soil were seperated and PCB concentrations in both media were measured. This complete test system was then reduced to assess contributions of individual endogenous digestive fluid constituents (water, salts, pancreatin, bile, and mucin) and representative foods: protein (bovine serum albumin (BSA)), sugar (glucose), and fat (oleic acid). Then, the influence of increasing concentrations of BSA, glucose, and oleic acid was evaluated (individually) complete test systems. In a subset of the samples, solid phase microextraction (SPME) was used to measure freely dissolved PCBs. Across all treatments, percent soil organic carbon was the most influential bioaccessibility determinant, accounting for ≥87% of the explained variation. When evaluated individually, pancreatin, mucin, BSA, bile, and oleic acid each effectively increased PCB bioaccessibility and reduced freely dissolved PCB concentrations. This suggests competitive sorption of PCBs by organic constituents of the digestive constituents. Without sink material, intra-PCB mobilization differences were observed as freely dissolved PCB concentrations inversely correlated (p < 0.05) with their respective log K_ow's. When added to the complete digestive fluid, increasing oleic acid mass increased PCB bioaccessibility (p < 0.05), while adding more BSA or glucose did not (p > 0.05). This indicates that fat intake may be the sole consideration needed when modeling dietary contributions to bioaccessibility of soil sorbed PCBs.