Circulating proteomic landscape of lung function

Background Large-scale genetic and epigenetic studies have identified numerous genes linked to lung function. However, proteomics, which can offer more direct insights into pathophysiological processes, remains underexplored. We aimed to identify circulating proteins related to lung function. Methods In 20 823 adults (71% European, 15% African and 15% Asian ancestries) across five cohorts we investigated spirometry parameters (forced expiratory volume in 1 s (FEV 1 ), forced vital capacity (FVC) and FEV 1 /FVC) in relation to abundance in circulation of 4693 proteins assessed using the SomaScan platform. Study-level associations were determined using robust linear regression, adjusting for confounders including age, sex, height, weight and smoking. Results were then meta-analysed using inverse variance weighting. Results In this multi-ancestry population, 1055 proteins were significantly associated with a lung function trait after Bonferroni correction (p<1×10 −5 ). The 473 enriched pathways identified include those involving inflammation and organismal injury. Protein–protein networks indicate potential orchestrators of lung function, including STAT3 and EGFR. Associations with 411 proteins were validated in the UK Biobank using the Olink 3K platform (560 overlapping proteins). 179 proteins identified were related to COPD in our data. While most associated proteins are likely biomarkers of impaired lung function, Mendelian randomisation provides preliminary evidence suggesting potential causality for 34 proteins. Our findings include known biomarkers of lung diseases, including COPD. Notably, 89% of associated proteins have not been previously implicated in lung function. Conclusion This comprehensive investigation identified novel protein–lung function associations that could improve understanding of lung disease pathogenesis, aid in the discovery of circulating biomarkers and accelerate the development of new management strategies for respiratory conditions.