LAUSR

Abstract This study aimed to identify potential lipid biomarkers in women with polycystic ovary syndrome (PCOS) and determine their predictive value for PCOS. Eighteen women with PCOS and 17 healthy controls were enrolled. A multi-dimensional mass spectrometry-based shotgun lipidomics approach was employed to analyze serum lipid profiles. Shotgun lipidomics revealed that the concentrations of ceramide (Cer) and phosphatidylcholine (PC) were higher (PC: 831.6 ± 217.4 vs. 605.2 ± 164.2 μmol/l; Cer: 3,387.6 ± 829.9 vs. 2,552.2 ± 679.4 nmol/l, respectively), whereas that of lysophosphatidylcholine was lower, in PCOS women than in healthy controls (82.02 ± 39.49 vs. 133.62 ± 65.36 μmol/l, respectively). Receiver operating characteristic analysis showed that the combination of Cer (OH_N16:0/N18:0) and Cer (N22:0) had the greatest discriminatory power to differentiate between women with and without PCOS (area under the curve: 0.889, 95% confidence interval: 0.784–0.994). These results indicate that the combination of Cer (OH_N16:0/N18:0) and Cer (N22:0) may represent a novel lipid predictor of PCOS.