Machine Learning Study Finds Children Exposed to Toxic Air Particularly Vulnerable to Asthma

Air toxics are atmospheric contaminants that pose significant risks to both human health and the environment. A study in Science of the Total Environment used machine learning (ML) to examine air toxics associated with asthma symptoms.

The study, conducted in Spokane, Washington, examined 269 elementary school students to identify specific combinations of 109 air toxics associated with asthma symptoms.

Investigators employed data sourced from Spokane Public Schools, detailing the frequency of asthma symptoms exhibited by the participating children. The Environmental Protection Agency’s Air Toxics Screening Assessment and National Air Toxics Assessment served as invaluable resources to gauge exposure to air toxics. The researchers delineated 3 exposure periods—the most recent year (2019), last 3 years (2017-2019), and last 5 years (2014-2019)—to analyze the data.

Central to the study’s methodology was the utilization of the Data-driven ExposurE Profile (DEEP) extraction method, a ML-based approach tailored to uncovering complex associations between environmental exposures and health outcomes. Through DEEP, the researchers identified 25 combinations of air toxics significantly linked to asthma symptoms across the specified exposure periods.

Three combinations—1,1,1-trichloroethane, 2-nitropropane, and 2,4,6-trichlorophenol—emerged as consistently associated with asthma symptoms across all exposure periods. Additionally, four air toxics—1,1,1-trichloroethane, 1,1,2,2-tetrachloroethane, BIS (2-ethylhexyl) phthalate (DEHP), and 2,4-dinitrophenol—were identified as significant only when occurring in combination with other toxics, underscoring the nuanced interplay of pollutants often overlooked by traditional statistical methods.

Moreover, the application of DEEP uncovered a vulnerable subset of children disproportionately exposed to 13 of the 25 significant air toxic combinations. These children exhibited a higher prevalence of asthma symptoms compared with their counterparts, signaling the urgent need for targeted interventions to mitigate exposure and alleviate respiratory distress.

“By providing evidence on air toxic combinations associated with childhood asthma, our findings may contribute to the regulation of these toxics to improve children‘s respiratory health,” investigators concluded.