Metabolite Fusion between Breath and Blood Enables More In-Depth Understanding of the Endogenous Metabolome

Blood is a widely used sample type in metabolomics but often loses volatile compounds during analysis. In contrast, exhaled breath offers a noninvasive and complementary matrix that retains these volatiles. However, the accuracy of metabolite identification in breath remains a key challenge. To address this, we developed and integrated three novel strategies to enhance the characterization of the human metabolome: (1) a controlled exercise protocol was applied to capture biologically relevant metabolic changes, (2) water-gas partition coefficients and a self-built breath metabolomics database were incorporated to enhance the identification accuracy of breath metabolites, and (3) fusion of breath and blood metabolites was conducted to expand metabolite coverage and validate the reliability of breath metabolite identification. Using high-resolution tandem mass spectrometry, we conducted an untargeted metabolomics analysis of breath and blood samples collected during exercise. A total of 66 metabolites were uniquely identified in breath, 59 were unique to blood, and only 4 were shared between the two. Fusion of breath and blood data expanded the coverage of exercise-associated metabolic pathways and revealed breath-specific markers of exercise-induced metabolic changes. This study presents an accurate and integrative strategy for breath metabolite discovery, advancing our understanding of in vivo metabolism and offering promising biomarkers for smart wearable devices.