Fig. 2

The dysregulated plasma metabolome of MDD patients, and the effect of ESC treatment on plasma metabolites. a Principal component analysis (PCA) based on plasma metabolomic profiles revealed a significant difference between the plasma metabolic signatures of MDD patients and those of HCs (PERMANOVA, Euclidean distance, P = 0.001). b Volcano plot of regression coefficients for the differences in metabolite levels between MDD patients and HCs. Multivariate regression models were used to explore the associations between plasma metabolites and disease status, statistical significance was determined by q value < 0.1. Compared with HCs, the MDD group was characterized by 7 up-regulated metabolites and 37 down-regulated metabolites, these metabolites were mainly involved in organic acid and amino acid. c Heatmap showing the average abundance of 41 plasma metabolites, which were significantly changed at week 2 or week 12 after ESC intervention, as determined by linear mixed models (LMM). Heatmap row annotation showing the colored coefficients derived from LMMs (Baseline vs. W2 and Baseline vs. W12) and multivariate regression models (MDD vs HC). The coefficients have different colors of blue (down-regulated or enriched in HC) and red (up-regulated or enriched in MDD). Significant differences are indicated by * p < 0.05, ** p < 0.01, *** q < 0.1. d Longitudinal changes in 6 plasma metabolites (from c), which exhibited differential trends between the R and NR groups. The P values were obtained from LMMs applied to grouped data of R and NR