Short-term ambient VOC mixtures were associated with higher daily outpatient visits and hospital admissions for pediatric atopic dermatitis, or AD, and asthma in Seoul. The finding describes a population-level association, not evidence that the exposure caused either condition.
The data were daily outpatient-visit and hospital-admission counts for people under 19 in Seoul, linked to citywide VOC concentrations from 2015 to 2019. The study therefore followed patterns in health-service use across the city rather than estimating risk for individual children.
Toluene carried the strongest weight
Toluene had the highest weight in the study's Weighted Quantile Sum regression, or WQSR, analysis. Its reported weight was 0.88 for AD and 0.75 for asthma.
That ranking does not show that toluene alone caused either outcome. Hexane and cyclohexane also received relatively high weights: 0.67 and 0.38 for AD, and 0.55 and 0.25 for asthma, respectively.
The researchers grouped VOC concentrations into n-alkanes, branched or cyclo-alkanes, alkenes or alkynes, and aromatics. They used WQSR and Bayesian Kernel Machine Regression to examine associations involving the mixtures.
The burden figures use a scenario
The analysis also reported ecological population attributable fractions, or PAFs. These are estimates tied to a chosen lower-exposure comparison, rather than counts of cases shown to be preventable in the observed data.
Under a hypothetical scenario using exposure at the 25th percentile, the reported PAF for AD was 7.0% for G4. For asthma, the reported PAFs were 4.7% for G1, 4.0% for G2 and 6.5% for G4.
Because these estimates are ecological, they describe the city-level analysis rather than individual children's risk. The supplied analysis reports no interval estimates for the PAFs or other quantitative uncertainty measures for the reported associations.
The results stop short of proving cause
Generalized additive models also produced broadly consistent single-compound associations. That consistency does not resolve whether individual VOCs, or sources shared with other pollutants, account for the observed pattern.
The authors interpret aromatics, particularly toluene, as associated with increased pediatric AD and asthma burden in Seoul. They say the roles of other VOCs need further investigation because potential shared sources with other pollutants can complicate attribution.
The retrospective ecological design used citywide exposures and daily health-service counts, so it cannot provide individual exposure or individual risk estimates. It also cannot establish causation.
The supplied analysis does not report detailed model specifications, exposure lags, covariate adjustment, concentration distributions or quantitative uncertainty estimates. Whether similar associations would appear in other populations, locations or periods, and how individual exposure measurements would change the estimates, remains open.
The findings therefore concern the studied setting and period, while the PAFs remain scenario-based ecological estimates rather than measures of personal risk.
Publication details
The citation identifies the article as the version of record dated 20 August 2026.
The retrospective study received Yonsei University Health System institutional review board approval, and informed consent was waived.
The work was supported by a National Research Foundation of Korea grant funded by the Korean Ministry of Science and ICT and the Ministry of Education, and by a National Institute of Health research project. The authors report no competing interests.
The paper states that all data generated or analyzed during the study are included in the published article.
Paper data and sources
Original title: Short-term effects and attributable burden of ambient volatile organic compound mixtures on pediatric atopic dermatitis and asthma in Seoul.
Authors: Minsun Lee, Kyungjun Jeong, Kyungnam Kim et al.
Journal/Repository: Journal of exposure science & environmental epidemiology
Status: Peer-reviewed
First online: 2026-08-20
DOI: 10.1038/s41370-026-00965-5
Original paper · Full text