SAN ANTONIO — December 4, 2024 — Southwest Research Institute (SwRI) collaborated with the Environmental Protection Agency (EPA) to characterize the chemical makeup of 81 common household items. Researchers also evaluated the potential risk to users.
Exposure to chemicals can cause negative health effects, according to the Centers for Disease Control and Prevention. Building on previous research to identify chemicals in consumer goods, SwRI and EPA also analyzed how samples of rubber, plastic, clothing, upholstery and fabric responded to environmental factors, such as a hot car or being worn.
The study, published in the Environmental Science & Technology journal, examines four years of data captured with advanced chromatography, suspect screening, non-targeted analysis and the SwRI-developed machine-learning method Highlight™. Instead of screening a sample against individual known compounds, this method allows scientists to identify, characterize and evaluate a large library of chemicals through suspect-screening analysis. The method identified 88,795 unique chemical features and 1,883 compound groups from 13 analytical batches.
"Highlight leverages machine learning algorithms for rapid pattern matching, which accelerated the workflow," said William Watson, a research engineer in SwRI's Intelligent Systems Division and the study's lead author.
Another aim of the study was to advance the field of exposomics, which explores how a lifetime of chemical exposure from the environment, diet, lifestyle and other sources impacts human health. Characterizing chemicals in household items and common sources of exposure may help with future biomonitoring efforts.
"Consumer products don't just consist of one chemical. Think of it as a mixed bag of related chemicals," said Dr. Kristin Favela, a staff scientist in SwRI's Chemistry and Chemical Engineering Division. "We wanted to determine if chemicals in the samples were 'emittable' or 'extractable' to understand the magnitude and likelihood of human chemical exposure."
SwRI exposed samples of clothing, upholstery, fabrics, rubber and plastics to two different heat settings and solvent strengths. The researchers wanted to determine if the test samples would emit chemical vapors that might be inhaled in an indoor environment, like a hot car, or if worn. The study also explored whether chemicals could be extracted to better understand real-world exposure risks, such as when a child chews on a household item.
"Along with helping to advance our understanding of the risk chemical exposure poses to the public, this study also demonstrates our capability to use machine learning and Highlight findings to retrospectively analyze and understand older datasets," said Watson.
Using EPA's Toxicity Forecasting program, ToxCast, the team performed an additional analysis and interpretation to predict risk based on human exposure and available biological activity data. Among the 88 confirmed chemicals that were both extractable and emittable, 66 had available ToxCast data, and a majority of the ToxCast in vitro assay data (92%, an average of 441 assays per chemical) indicated no activity dependent on concentration. No data was available for the other 22 chemicals. However at a higher concentration, synthetic antioxidant BKF, which can be used to stabilize plastics and rubbers, did show adverse effects when exposure reached 42.3 mg/kg/day. The research may help advance a screening model that can predict emission activity for household items.
The title of the paper is "Discerning Emittable from Extractable Chemicals Identified in Consumer Products by Non-Targeted GCxGC-TOFMS," and it appears in Environmental Science & Technology journal and can be accessed at: https://pubs.acs.org/doi/10.1021/acs.est.4c07903 .
