Human Chemical Exposure from Background Emissions in the United States and the Implication for Quantifying Risks from Marginal Emission Increase

Abstract

The linear dose–response relationship has long been assumed in assessments of health risk from an incremental chemical emission relative to background emissions. In this study, we systematically examine the relevancy of such an assumption with real-world data. We used the reported emission data, as background emissions, from the 2017 U.S. National Emission Inventory for 95 organic chemicals to estimate the central tendencies of exposures of the general U.S. population. Previously published nonlinear dose–response relationships for chemicals were used to estimate health risk from exposure. We also explored and identified four intervals of exposure in which the nonlinear dose–response relationship may be linearly approximated with fixed slopes. Predicted rates of exposure to these 95 chemicals are all within the lowest of the four intervals and associated with low health risk. The health risk may be overestimated if a slope on the dose–response relationship extrapolated from toxicological assays based on high response rates is used for a marginal increase in emission not substantially higher than background emissions. To improve the confidence of human health risk estimates for chemicals, future efforts should focus on deriving a more accurate dose–response relationship at lower response rates and interface it with exposure assessments.

Exposure Factors vs. Bioaccessibility in the Soil-and-Dust Ingestion Pathway: A Comparative Assessment of Uncertainties Using MC2D Simulations in an Arsenic Exposure Scenario

“…Inclusion of the bioaccessibility parameter in the risk equations can significantly reduce its magnitude. However, there are other sources of uncertainty that can significantly affect the indeterminacy of risk, such as those associated with the toxicity value, environmental background exposure (Barrio-Parra et al 2018;Li and Li 2021;Stanfield et al 2021), or, as already mentioned, those associated with the exposure parameters (Mingot et al 2011). Regarding the latter, the uncertainty produced by the use of default values can be reduced through their characterisation by local studies (e.g.…”