Abstract
Environ Int. 2026 Jul 28;215:110442. doi: 10.1016/j.envint.2026.110442. Online ahead of print.
ABSTRACT
BACKGROUND: Disinfection by-products, including those arising from secondary reactions between chlorinated reagents and natural organic matter such as trihalomethanes (THMs), have been associated with several adverse health outcomes, such as bladder and colorectal cancer.
AIMS: We aim to identify metabolic signatures of THM exposure in a high exposure setting (swimming pool), validate these markers in an external dataset with low exposure contrast (residential tap water), and characterise the multi-omic responses to THM exposures.
MATERIALS AND METHODS: In PISCINA-II study (discovery data), n = 58 volunteers swam in a chlorinated pool for 40 min. Metabolomics and proteomics in serum and THM concentrations in exhaled breath were measured before and after swimming. Metabolome-wide association study (MWAS) using multivariate normal models was used to identify signatures of exposure to individual THM components. Multivariate analysis coupling stability-calibrated consensus clustering and sparse partial least square models selected sets of features jointly affected by exposures. Stability-calibrated conditional independence networks integrated selected metabolic features and immune proteomic markers.Selected features were finally annotated and further validated in the MCC-Spain study with n = 293 cancer-free individuals.
RESULTS: MWAS identified 188 features significantly associated with at least one THM component. Multivariate analysis selected 21 metabolic clusters, containing 72 individual features, of which 12 were tentatively identified. Of these, L-kynurenine, Lyso-PE (20:4), piperine and cortisone were also identified in the MCC-Spain study despite much lower exposure contrasts. We visualised the complex correlations across molecular features jointly associated with THM exposures.
CONCLUSIONS: This study provides detailed characterisation of molecular responses to THM exposure from different routes and at different levels. Future studies are needed to understand the health implications of the involved metabolites.
PMID:42561590 | DOI:10.1016/j.envint.2026.110442
UK DRI Authors
