Association of baseline plasma NMDA receptor–related amino acid ratios with remission following antidepressant treatment in major depressive disorder

BACKGROUND: The contribution of glutamatergic and neuroplasticity-related mechanisms to antidepressant efficacy beyond monoaminergic modulation is increasingly recognized. Amino acids related to N-methyl-d-aspartate receptor (NMDAR) function have been implicated in the pathophysiology of depression; however, their relevance to antidepressant treatment outcomes remains unclear. METHODS: In this exploratory pooled analysis, we integrated data from a prospective observational cohort study (REBELLION Study) and an open-label randomized controlled trial (GUNDAM Study). Outpatients with major depressive disorder (MDD) who initiated antidepressant monotherapy were included. Baseline plasma ratios of NMDAR-related amino acids (d-serine/l-serine, l-serine/glycine, and glutamine/glutamate) were examined in relation to remission at week 6 (17-item Hamilton Depression Rating Scale score ≤ 7) using multivariable logistic regression. Exploratory interaction analyses assessed effect modification by antidepressant pharmacological class (mirtazapine vs. non-mirtazapine). RESULTS: Among 91 participants, higher baseline d-serine/l-serine (adjusted odds ratio [aOR] per 10% of the median = 1.18, 95% confidence interval [CI] 1.01-1.38) and glutamine/glutamate (aOR per 10% of the median = 1.06, 95% CI 1.00-1.11) ratios were associated with remission at week 6. Interaction analyses suggested possible antidepressant-class differences, with d-serine/l-serine associated with remission in the non-mirtazapine group and glutamine/glutamate in the mirtazapine group. CONCLUSIONS: Baseline plasma NMDAR-related amino acid ratios were associated with remission following short-term antidepressant treatment in patients with MDD. These exploratory findings suggest that peripheral amino acid profiles may contribute to interindividual variability in antidepressant outcomes. However, they should be interpreted as hypothesis-generating rather than as evidence of biomarker utility, and require replication in independent and longitudinal cohorts.