ABSTRACT
BACKGROUND@#As one of the most common endocrinal disorders for women at childbearing age, the diagnostic criteria of polycystic ovary syndrome (PCOS) have been defined differently among different international health organizations. Phenotypic heterogeneity of PCOS also brings about difficulties for its diagnosis and management assessment. Therefore, more efficient biomarkers representing the progression of PCOS are expected to be integrated into the monitoring of management process using metabolomic approaches.@*METHODS@#In this prospective randomized controlled trial, 117 PCOS patients were enrolled from December 2016 to September 2017. Classical diagnostic parameters, blood glucose, and metabolome were measured in these patients before and at 2 months and 3 months of different medical interventions. The receiver operating characteristic (ROC) curves were built based on multivariate statistical analysis using data at baseline and 3 months' management, and combinational biomarkers with appreciable sensitivity and specificity were selected, which then validated with data collected at 2 months.@*RESULTS@#A set of metabolites including glutamic acid, aspartic acid, 1-methylnicotinamide, acetylcarnitine, glycerophosphocholine, and oleamide were filtered out with high performance in representing the improvement through 3-month management of PCOS with high sensitivity and specificity in ROC analysis and validation with other two groups showed an appreciable area under the curve over 0.96.@*CONCLUSIONS@#The six metabolites were representative of the remission of PCOS through medical intervention, making them a set of potential biomarkers for assessing the outcome of PCOS management.@*TRIAL REGISTRATION@#ClinicalTrials.gov, NCT03264638.