Dynamic plasma lipidomic alterations and diagnostic potential in major depressive disorder versus bipolar disorder: A...

Major depressive disorder (MDD) and bipolar disorder (BD) are frequently misdiagnosed due to overlapping depressive symptomatology, leading to suboptimal treatment outcomes. While lipidomics has shown promise in distinguishing these disorders, dynamic lipid alterations during treatment remain unexplored.

We conducted a prospective, paired-design lipidomic study involving 31 MDD and 30 BD patients, assessed at baseline and after 2 weeks of treatment, alongside 32 matched healthy controls (HC). Plasma lipids were profiled using liquid chromatography-mass spectrometry (LC/MS). Orthogonal partial least squares-discriminant analysis (OPLS-DA), pathway enrichment, and LASSO regression were employed to identify differential lipids and construct diagnostic panels.

Significant differences in lipid classes were observed between MDD and HC, and between BD and HC, while differences between MDD and BD were not substantial. Notable variations in the relative contents of numerous lipid species were detected across the three groups. Alternations in lipid species within the MDD and BD groups before and after treatment were highly significant. Diagnostic lipid panels were developed to differentiate MDD from HC (13 lipids), BD from HC (12 lipids), and MDD from BD (13 lipids) using Venn diagrams and Lasso regression models.

This first paired follow-up lipidomics study in MDD and BD reveals dynamic lipid alterations early in treatment and proposes highly discriminative lipid panels. These findings underscore the role of lipid dysregulation in affective disorders and suggest potential biomarkers for differential diagnosis.