Chronic respiratory diseases (CRDs) are a growing global health concern. Emerging evidence implicates circulating metabolites in their development and progression, but definitive causal insights into how metabolic patterns influence disease pathways remain limited. We conducted a two-sample Mendelian randomization (MR) analysis to evaluate potential causal effects of 1400 serum metabolites on 5 major CRDs: chronic obstructive pulmonary disease (COPD), asthma, idiopathic pulmonary fibrosis (IPF), sarcoidosis, and pneumoconiosis. Each disease was analyzed separately; no composite CRD outcome was constructed. The inverse-variance weighted method was the primary approach, complemented by MR-Egger, weighted median, and MR-PRESSO for outlier detection and correction. Robustness was examined using sensitivity analyses, including Cochran Q for heterogeneity, MR-Egger intercept for directional pleiotropy, the MR-PRESSO global test, and leave-one-out analyses. MR analyses identified significant causal associations between multiple metabolites/metabolite ratios and the 5 CRDs, revealing distinct metabolic profiles for each disease. For COPD, we found 54 potentially causal metabolites (30 risk, 24 protective). Asthma showed 36 associations (13 risk, 20 protective). IPF had 39 (22 risk, 17 protective). Sarcoidosis exhibited the broadest signature with 69 associations (24 risk, 45 protective). Pneumoconiosis showed 53 (25 risk, 28 protective). Most signals were disease-specific; only a small subset overlapped across at least 2 diseases (e.g., four shared between COPD and sarcoidosis, two between pneumoconiosis and asthma), suggesting partially shared pathways. No metabolite displayed consistent associations across all 5 diseases. Circulating metabolites exhibit protective or detrimental causal effects on COPD, asthma, IPF, sarcoidosis, and pneumoconiosis. Effects are heterogeneous and largely disease-specific, with limited overlap across conditions, indicating predominantly distinct etiologic pathways and offering mechanistic insights that may inform risk stratification and target prioritization. Further validation using integrated multi-omics and experimental models is warranted to refine mechanisms and assess translational potential.
Gao et al. (Fri,) studied this question.