Untargeted Screening of Environmental and Endogenous Metabolites and Risk of Incident COPD: A Prospective Study in Three European Populations

ObjectiveChronic obstructive pulmonary disease (COPD) is a leading cause of death in the industrialized world. Although smoking, air pollution, and occupational exposures are well established risk factors, the molecular pathways linking environmental exposures and biological susceptibility to COPD remain incompletely understood. Untargeted metabolomics offers a unique opportunity to simultaneously capture internalized environmental chemicals and endogenous metabolic perturbations. However, large prospective studies integrating broad exposomic and metabolic screening prior to COPD onset are lacking. MethodsWe conducted a nested case-control study within three European population-based cohorts (Doetinchem Cohort Study, KORA, SAPALDIA) and analyzed 1473 prospectively collected plasma samples. COPD was defined by a pre-bronchodilation FEV1/FVC ratio below 0.7 at follow-up (4-16 years after blood sample collection). We applied complementary untargeted liquid- and gas chromatography high-resolution mass spectrometry (LC- and GC-HRMS), enabling extensive coverage of endogenous metabolism and exogenous environmental contaminants, including pesticides, plastic-related chemicals, and polycyclic aromatic hydrocarbons. Controls maintained normal lung function and were matched to cases on age, sex, follow-up time, and sample collection round. We performed separate conditional logistic regression models for each metabolomic feature, and used Mummichog for prediction of biological pathways involved. The false discovery rate (FDR) was controlled using the Benjamini-Hochberg procedure. Long-term measurement reliability was evaluated using intraclass correlation coefficients (ICCs) from repeat samples in the Doetinchem Cohort Study. ResultsIn total, thousands of metabolomic features were screened, including 724 annotated exogenous compounds, 13 endogenous metabolites, and 197 features that could be derived as both. Nicotine and cotinine intensity levels were statistically significantly associated with COPD incidence at an FDR of 10%, validating the analytical and epidemiologic framework. Lower levels of butyrylcarnitine were related to COPD onset in never-smokers. Beyond smoking-related markers, lower levels of butyrylcarnitine were associated with increased COPD risk among never-smokers, implicating altered mitochondrial fatty-acid metabolism as a potential early pathway independent of tobacco exposure. Although most screened environmental contaminants, including PAHs and pesticides, were not associated with COPD at stringent significance thresholds, restricting analyses to temporally stable metabolites identified the insecticide metabolite phenyl N-methylcarbamate as a predictor. ConclusionThis large-scale, prospective untargeted metabolomics study represents one of the most comprehensive assessments to date of both environmental and endogenous metabolic predictors of COPD. Our findings demonstrate the feasibility of exposome-wide molecular screening years before disease onset, identify butyrylcarnitine as a novel metabolic predictor in never-smokers, and highlight the importance of accounting for temporal variability in metabolomic epidemiology.