Predicting arsenic relative bioavailability in contaminated soils using meta analysis and relative...

Area of Science:

• Environmental Science
• Toxicology
• Geochemistry

Background:

• In vitro assays estimate arsenic (As) bioaccessibility and relative bioavailability (RBA) for risk assessment.
• Existing data are often analyzed in isolation, limiting predictive power for specific arsenic sources.
• Meta-analysis offers a potential approach to link As sources with bioaccessibility and RBA.

Purpose of the Study:

• To predict As RBA using established in vivo/in vitro correlations from existing literature.
• To evaluate the impact of different As sources on As RBA, independent of geographic location.
• To assess the influence of various in vitro methodologies on predicted As RBA.

Main Methods:

• Compiled data from 351 soils and 514 bioaccessibility values from published literature.
• Classified soils based on arsenic source and utilized published regression models to predict As RBA.
• Standardized in vitro data (SBRC intestinal, IVG, PBET) to SBRC gastric phase values using linear regression.

Main Results:

• Predicted mean As RBA varied significantly among different As sources and in vitro methods.
• Standardization of in vitro data revealed a clear order of predicted As RBA: CCA posts ≥ herbicide/pesticide > mining/smelting > gossan soils.
• The 95th percentiles for predicted As RBA were 78.0% (CCA posts), 78.4% (herbicide/pesticide), 67.0% (mining/smelting), and 23.7% (gossan soils).

Conclusions:

• Arsenic source is a significant determinant of As RBA, even when accounting for in vitro assay variations.
• Standardizing in vitro data enhances the reliability of As RBA predictions for site-specific risk assessments.
• This meta-analysis approach provides valuable insights into source-specific arsenic exposure potentials.