Application of a New Non-invasive Breath Test in the Diagnosis and Screening of Gout

The cohort was recruited at the Affiliated Hospital of Jiangxi University of Chinese Medicine between December 2025 and March 2026. In this non-randomized study, consecutive eligible volunteers were enrolled to minimize selection bias. Healthy controls were recruited from routine health examination programs and were required to have normal joint mobility and sensation, normal serum urate levels, and no history of chronic renal or metabolic disease. In this single-center study, breath samples from 201 participants (99 gout patients, 102 healthy controls) were analyzed by extractive electrospray ionization mass spectrometry (EESI-MS). Machine-learning algorithms were applied to build diagnostic models and identify discriminatory metabolites. Clinical risk factors and tongue-coating features were also assessed. There are three categories of data in total: exhaled breath data, clinical risk factor data, and tongue manifestation images collected from both the experimental group and the control group.Statistical analyses were performed using SPSS 25 and R 4.3.3, and machine-learning tasks were implemented in Python 3.11 within VS Code 1.123.0. This study was supported by the National Natural Science Foundation of China (Grant No. 82560945), the Leading Talent Program of Jiangxi Provincial Department of Science and Technology (Grant No. 20243BCE51009), the Jiangxi Provincial Special Fund for Postgraduate Innovation (Grant No. YC-2025-B168), and the Postgraduate Innovation Special Fund of Jiangxi University of Chinese Medicine (Grant No. XJ-S202537).