Despite intensified research, depression remains a mental disorder with an insufficiently understood pathophysiology. Clinically applicable biomarkers are lacking but are of highest interest, particularly for preventive and predictive measures. Here, we report combined findings on untargeted metabolomics, inflammation markers in serum, mitochondrial respiration in peripheral blood immune cells, and measures of oxidative stress, along with psychometric assessments in a female outpatient cohort (n = 19) compared to non-depressed female controls (n = 23). We used Weighted Correlation Network Analysis (WCNA) to identify metabolite clusters differentiating depressed and non-depressed individuals. Furthermore, WCNA revealed associations between biological parameters – previously associated with major depression − and the identified metabolite clusters.

Including all identified metabolites, our results showed separation between participants with depression and non-depressed controls. WCNA identified ten metabolite clusters positively correlated with depression and one negatively correlated cluster. Several clusters indicated alterations in lipids and metabolites related to oxidative stress, mitochondrial function, and inflammation, underscoring the significance of inflammatory activity and oxidative stress in depression. Subsequent pathway analysis assigned the metabolites to sphingolipid, glycerophospholipid, and glycerolipid metabolism, highlighting potential targets for further research and intervention.

This study supports the view of depression as a pathophysiological condition associated with alterations in several biological parameters and reinforces previous findings on the role of lipid metabolism in depression, alongside alterations in inflammatory activity and mitochondrial bioenergetics. The data also revealed correlations between bacterial and plant metabolites and MDD, suggesting impaired gut epithelial integrity in depression.