Plasma neurotransmitter-related metabolite alterations in adolescent major depressive disorder with and without...

Whether adolescent major depressive disorder (MDD) with psychotic features has a distinct peripheral metabolite signature remains uncertain. We compared targeted plasma neurotransmitter-related metabolites in psychotic MDD, non-psychotic MDD, and healthy controls.

We enrolled 220 adolescents (53 psychotic MDD, 67 non-psychotic MDD, 100 controls) and quantified 14 metabolites spanning tryptophan, glutamate-GABA, and catecholamine pathways using LC-MS/MS. Age- and sex-adjusted ANCOVA tested group effects on metabolites, biologically informed ratios, and pathway composites with false-discovery rate (FDR) control. Nested cross-validated classifiers evaluated out-of-sample group separability. For psychotic vs non-psychotic contrasts, we reported design sensitivity and Two One-Sided Tests (TOST) equivalence testing (smallest effect size of interest |d| < 0.30).

Versus controls, both MDD subgroups showed lower kynurenic acid and kynurenic acid/kynurenine, higher glutamine, higher glutamate/(serotonin+homovanillic acid), and lower serotonin/homovanillic acid (all FDR < 0.05). The tryptophan composite declined from controls to non-psychotic to psychotic MDD, and the GABA composite was lower in psychotic MDD versus controls. No metabolite or ratio survived FDR correction between psychotic and non-psychotic MDD. With n = 53 vs 67, 80% power corresponded to approximately d ≈ 0.52, and TOST did not support equivalence at |d| < 0.30. Classifiers distinguished pooled MDD vs controls (best AUC = 0.917; balanced accuracy = 0.861) but were near chance for psychotic vs non-psychotic MDD (best AUC = 0.592).

Targeted plasma metabolomics identified a robust case-control signature in adolescent MDD, centered on tryptophan-pathway and excitation/monoamine balance measures. Within this panel, evidence for a distinct psychotic-feature subtype signature was limited, and small subtype differences remain possible.