Exploratory Longitudinal Pilot Study of Proteomic and Metabolomic Profiling Suggests a Hypothesized...

Major Depressive Disorder (MDD) with atypical features, briefly named Atypical Depression (AFD), exhibits affective instability and metabolic-immune abnormalities, yet objective biomarkers remain lacking. This exploratory longitudinal pilot study aimed to evaluate whether plasma proteomic and metabolomic profiling could nominate candidate immune-metabolic signatures associated with AFD.

This longitudinal pilot study included five female AFD patients assessed at baseline, 3 months, and 1 year, and five healthy female controls assessed at baseline. Plasma samples were collected at predefined time points for SomaScan®-based proteomic profiling and UHPLC-QTOF-MS-based untargeted metabolomic profiling. Ensemble Feature Selection (EFS), co-expression network analysis (WGCNA), and multi-omics integration (O2PLS, correlation) were applied to identify immuno-metabolic signatures associated with affective instability.

EFS nominated candidate proteins (e.g., IFN-λ1, PIAS4, CD63) and metabolites (e.g., glutamic acid, uric acid, stearic acid) involved in immune response and oxidative stress as prioritized features in AFD. WGCNA clustered 19 proteomic modules, with at least eight enriched for immune/inflammatory pathways. The salmon module, associated with innate immunity, showed a nominal inverse Pearson correlation with stearic acid (r = -0.55, p = 0.01). Cross-omics analysis further suggested shared proteomic-metabolomic covariance, and focused analysis showed an inverse PIAS4-stearic acid pattern in the joint latent space.

The findings suggest that AFD may involve dynamic immune-metabolic alterations. The longitudinal design indicates that these molecular changes may evolve over time rather than represent fixed abnormalities, potentially reflecting ongoing interactions between inflammatory processes and metabolic regulation. Notably, the exploratory inverse relationship between stearic acid and a PIAS4-centered immune module points to a potential link between lipid metabolism and innate immune activity. Based on these multi-omics observations and prior literature, it was hypothesized that immune-lipid interactions may contribute to mood instability in AFD.

The exploratory longitudinal pilot study nominates PIAS4-centered immuno-metabolic signals as candidate molecular features of AFD. The findings support the potential utility of longitudinal multi-omics for mapping dynamic biomarker networks in affective disorders. However, larger, independent, medication-informed, and demographically diverse cohorts are required to validate these candidate signatures and clarify their relevance to diagnosis, treatment response, and disease-course monitoring.