Discrimination of Bipolar Disorder and Schizophrenia Patients Based on LC-HRMS Lipidomics

<b>Background/Objectives:</b> Schizophrenia (SCH) and bipolar disorder (BD) share overlapping symptoms and genetic factors, making differential diagnosis challenging and often leading to misdiagnosis. This study aimed to identify potential lipid biomarkers of serum capable of distinguishing BD from SCH. <b>Methods:</b> Lipid profiles of serum from 30 SCH and 31 BD patients were analyzed in triplicates using liquid chromatography-high-resolution mass spectrometry (LC-HRMS). Chemometric analysis was applied, including class and gender identifiers. Orthogonal partial least squares (OPLS) models with 1000 cross-validations were used to validate feature subsets. <b>Results:</b> The chemometric analysis included the most relevant metabolites in the comparison between all samples of SCH and BD patients, identifying five key biomarkers (LPC 16:0, SM 33:1, SM 32:1, compound C30H58O3, and PC 30:0) with VIP scores > 1 for distinguishing BD from SCH. Gender-specific models revealed five biomarkers in males (SM 32:1, SM 33:1, PC 32:1, PC 30:0, and FA 16:1) and two in females (LPC 16:0 and C30H58O3). These biomarkers primarily belonged to glycerophospholipids (GPs) and sphingophospholipids (SPs). <b>Conclusions:</b> Comparative lipid profiling between SCH and BD, including gender-specific subgroups, enabled identification of potential diagnosis-specific biomarkers. Elevated levels of GPs and SPs in SCH patients suggest lipid metabolism differences that may support improved diagnostic accuracy and personalized treatment strategies.