Association Between District-level Aggregate Carcinogen Emissions and Thyroid Cancer Incidence in Korea: An Ecological Study Using Bayesian Spatial Models
Related Experiment Video
Updated: Nov 4, 2026

Real-World M3-BREATHE: Toward Multimodal Mobile Monitoring of Behaviour, Respiration, and Exposures for Treatment and Health Evaluation
Published on: June 5, 2026
Taehoon Kim1,2, Sangjun Lee1,2, Kyungsik Kim1,2
1Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Korea.
Objectives:
We examined the association between district-level aggregate carcinogen emissions and diagnosed thyroid cancer incidence in Korea, accounting for co-occurring environmental exposures (air pollutants, indoor radon, and proximity to nuclear power plants [NPPs]) and area-level sociodemographic factors.
Methods:
This ecological study analyzed thyroid cancer incidence (2014-2018) across 229 Korean districts (si-gun-gu) using Bayesian hierarchical Poisson models. The primary exposure was a district-level aggregate carcinogen emission indicator (area-weighted releases of International Agency for Research on Cancer Group 1/2A/2B carcinogens). Models mutually adjusted for CO, NO₂, PM₁₀, SO₂, indoor radon, and NPP proximity, together with sociodemographic, behavioral, urbanicity, and healthcare-access covariates. Relative risks (RRs) per 1-SD increase and 95% credible intervals (CrIs) were estimated overall, by sex, and by cancer-screening stratum.
Results:
The emission indicator was associated with higher incidence overall (RR, 1.04; 95% CrI: 1.01-1.06) and in women (1.04; 1.02-1.06), but not in men (1.02; 0.99-1.06). CO was inversely associated with incidence (0.96; 0.93-0.99), whereas population density was positively associated (1.13; 1.06-1.21). NO₂, PM₁₀, SO₂, indoor radon, and NPP proximity showed no association after adjustment for the urban-rural context. The emission association remained robust with alternative zero handling and exclusion of the highest-emitting districts. It did not differ between screening strata.
Conclusions:
District-level aggregate carcinogen emissions, predominantly organic solvents by mass, were associated with diagnosed thyroid cancer incidence, most clearly in women. These hypothesis-generating findings may indicate either that as-yet-unidentified chemicals increase thyroid cancer risk or that the models incompletely captured the variables underlying the reported urbanicity-thyroid cancer association.
Related Concept Videos
Criteria for Causality: Bradford Hill Criteria - II
Mutagenicity and Carcinogenicity
Confounding in Epidemiological Studies