Discovery of potential female-specific biomarkers for major depressive disorder by LC–MS-based metabolomics

Abstract

The prevalence of major depressive disorder (MDD) is higher in females than males, emphasizing the need to identify gender-specific biomarkers to improve diagnosis accuracy. In this study, a cross-sectional investigation with 258 samples was conducted to evaluate the discriminative power of potential gender-specific biomarkers for MDD. Eighteen MDD-related differential metabolites have been identified, involving pathways of phospholipids, glycerolipids, fatty acids, sphingolipids, cholesterol, vitamin E, and heme. A potential biomarker combination consisting of palmitelaidic acid, gamma carboxyethyl hydroxychroman (gamma-CEHC), and lysoPE(16:0) was confirmed for predicting depression in women using binary logistic regression analysis. To evaluate the panel's specificity, nine generalized anxiety disorder (GAD) samples, which share highly similar clinical symptoms with MDD, were included in the validation set. The discovery and validation sets yielded an area under the receiver operating characteristic curve of 0.86 and 0.83, respectively. All nine female GAD samples were correctly predicted as non-MDD, demonstrating the panel's specificity in diagnosing female MDD. Remarkably, this composite panel achieved a 75 % prediction accuracy in female samples in both the discovery and validation sets, but it did not reach 60 % prediction accuracy in male samples in either set. Our findings highlight the importance of gender-specific molecular diagnostics in developing practical and accurate diagnostic methods for MDD.