Metabolic profiling identifies phospholipids as potential serum biomarkers for schizophrenia

•51 phospholipids are significantly different between SCZ patients and controls.•These significant phospholipids include PCs, LPCs, PEs, LPEs and SMs.•Extensive disturbances of phospholipids may be involved in the development of SCZ.•A panel of 6 metabolites could discriminate SCZ patients from healthy controls. Schizophrenia (SCZ) is a multifactorial psychiatric disorder. However, the molecular pathogenesis of SCZ remains largely unknown, and no reliable diagnostic test is currently available. Phospholipid metabolism is known to be disturbed during disease processes of SCZ. In this study, we used an untargeted liquid chromatography-mass spectrometry (LC-MS)-based metabolic profiling approach to measure lipid metabolites in serum samples from 119 SCZ patients and 109 healthy controls, to identify potential lipid biomarkers for the discrimination between SCZ patients and healthy controls. 51 lipid metabolites were identified to be significant for discriminating SCZ patients from healthy controls, including phosphatidylcholines (PCs), lysophosphatidylcholines (LPCs), phosphatidylethanolamines (PEs), lysophosphatidylethanolamines (LPEs) and sphingomyelins (SMs). Compared to healthy controls, most PCs and LPCs, as well as all PEs in patients were decreased, while most LPEs and all SMs were increased. A panel of six lipid metabolites could effectively discriminate SCZ patients from healthy controls with an area under the receiver-operating characteristic curve of 0.991 in the training samples and 0.980 in the test samples. These findings suggest that extensive disturbances of phospholipids may be involved in the development of SCZ. This LC-MS-based metabolic profiling approach shows potential for the identification of putative serum biomarkers for the diagnosis of SCZ.