Abstract
Volatile organic compounds (VOCs) are metabolic byproducts reflecting pathophysiologic processes and have been widely utilized as non-invasive biomarkers for disease differentiation. While VOCs have been extensively studied in oncology, their potential for diagnosing other prevalent public health conditions remains underexplored. This study serves as a proof-of-concept investigation, applying VOC-based detection to diabetic kidney disease (DKD), a leading cause of chronic kidney disease, to distinguish it from other nephrotic syndromes (NS). Utilizing metal oxide semiconductors, we analyzed urinary VOC profiles in 135 participants with biopsy-confirmed diagnoses. All gas sensors in the sensor chamber were of the n-type. Our results demonstrated that hydrogen and ethanol VOCs, measured at a minimum electrical resistance threshold of 4500 mV, effectively distinguished DKD from NS and other control groups. The mean differences in electrical resistance among DKD versus normal, DM without DKD, and NS were 868.3, 145.5, and 881.2 ohms, respectively (p < 0.05). Receiver operating characteristic (ROC) analysis yielded AUC values of 1.0, 0.64, and 0.99, respectively (p < 0.05), underscoring the method's diagnostic potential. VOCs primarily composed of hydrogen and ethanol in DKD patients exhibited distinct electron-release properties when interacting with metal oxide semiconductors compared to healthy individuals and NS patients. These findings suggest that urine-based VOC analysis has potential as a non-invasive alternative for distinguishing DKD from other glomerular diseases. This study highlights the feasibility of extending VOC-based biosensor technology beyond oncology, offering a novel diagnostic framework for broader clinical applications in public health and metabolic disease monitoring.