Age-specific alterations in gut microbiota–plasma metabolome networks in first-episode, drug-naïve major depressive disorder: adolescents versus adults
Adolescents and adults with first-episode, drug-naïve MDD display distinct, non-overlapping gut dysbiosis and plasma metabolic perturbation profiles, with more extensive microbial remodeling in adolescents and predominant energy dysfunction and excitotoxicity in adults.
| Population | First-episode, drug-naïve MDD patients: 28 adolescents and 34 adults; age- and sex-matched healthy controls: 22 adolescents and 21 adults |
|---|---|
| Exposure | Gut microbiota profiling (16S rRNA gene sequencing) and untargeted plasma metabolomics integrated with microbiome-metabolome network analysis |
| Comparator | Age- and sex-matched healthy controls within each age group |
| Outcome | Gut microbiota alpha diversity (Chao1 index); Gut microbiota taxonomic composition (Firmicutes, Bacteroidetes, butyrate-producing genera); Plasma metabolome profile in adolescent MDD (neurodevelopmental and oxidative pathways); Plasma metabolome profile in adult MDD (energy dysfunction and excitotoxicity); Microbiota-metabolome interactome in adult MDD (correlations with pro-inflammatory sphingolipids and polyamines); MGB axis signature differentiation between adolescent and adult MDD (direct age-group comparison) |
Summary of findings
| Outcome | Effect | 95% CI | Certainty | Clinical relevance | Notes |
|---|---|---|---|---|---|
| Gut microbiota alpha diversity (Chao1 index) | p < 0.05 (direction: reduced in TMDD vs controls); in the effect size or 95% CI reported | — | Low | — | 1 studies |
| Gut microbiota taxonomic composition (Firmicutes, Bacteroidetes, butyrate-producing genera) | p-values not fully reported; in the OR/RR/SMD or 95% CI reported for individual taxa | — | Low | — | 1 studies |
| Plasma metabolome profile in adolescent MDD (neurodevelopmental and oxidative pathways) | Qualitative enrichment analysis; in the effect size or 95% CI reported | — | Low | — | 1 studies |
| Plasma metabolome profile in adult MDD (energy dysfunction and excitotoxicity) | Qualitative enrichment analysis; in the effect size or 95% CI reported | — | Low | — | 1 studies |
| Microbiota-metabolome interactome in adult MDD (correlations with pro-inflammatory sphingolipids and polyamines) | Correlation network analysis; in the standardized effect size or 95% CI reported | — | Low | — | 1 studies |
| MGB axis signature differentiation between adolescent and adult MDD (direct age-group comparison) | Qualitative; in the formal between-group statistical comparison with effect size or 95% CI reported | — | Low | — | 1 studies |
Context
MDD affects distinct age groups with heterogeneous pathophysiology, yet age-specific signatures of the microbiota-gut-brain axis in drug-naïve patients remain poorly characterized. Identifying differential patterns between adolescents and adults is relevant for developing targeted biomarkers and interventions. This study takes an integrative multi-omics approach to map these differences in an exploratory manner.
What the study showed
Adolescent MDD (TMDD) showed more extensive microbial remodeling, including reduced Chao1 index (p < 0.05), Bifidobacterium loss, and enrichment of Bacteroidetes, Gemmatimonadetes, and Planctomycetes, alongside shared depletion of Firmicutes and butyrate-producing genera (Anaerobutyricum, Blautia). Adult MDD (AMDD) presented a metabolic profile dominated by TCA cycle dysfunction, purine metabolism, and alanine/aspartate/glutamate pathways. The AMDD interactome showed attenuation of protective lipid correlations and strengthening of pro-inflammatory sphingolipid and polyamine links, while the adolescent interactome remained relatively stable. The authors do not report standardized effect sizes (OR, RR, SMD) or confidence intervals for most association findings.
How it was done
Cross-sectional case-control observational study; total n = 105 (62 MDD patients, 43 controls). Fecal samples underwent 16S rRNA gene sequencing; plasma was analyzed by untargeted metabolomics. Integrated bioinformatics and network analysis were applied to map microbiota-metabolome interactions. No longitudinal follow-up; single cross-sectional assessment.
Effect magnitude
The study reports statistical significance for Chao1 index (p < 0.05 in adolescents) and taxonomic compositional differences, but does not provide standardized effect sizes with 95% CI for central outcomes, limiting quantification of the true magnitude of observed effects.
Risk of bias
Small sample (28 adolescent MDD, 34 adult MDD) with insufficient statistical power for high-dimensional multi-omics analyses — high risk of uncorrected false positives. Cross-sectional design precludes causal inference; cause versus consequence of depression cannot be distinguished. No formal risk-of-bias tool applied (RoB 2 or ROBINS-I not used); dietary, socioeconomic, and lifestyle confounders were not systematically controlled. 16S rRNA sequencing has limited taxonomic resolution at species level. Untargeted metabolomics generates hypotheses, not confirmations.
What this study does NOT prove
This study does not prove causality between microbiota or metabolome alterations and MDD development in any age group. It does not demonstrate that microbiota interventions will produce clinical benefit, nor that identified profiles are MDD-specific rather than non-specific correlates of disease state.
In clinical practice
This study does not support any immediate change in clinical practice. Findings are exploratory and generated in a small, unvalidated sample. Clinicians should await replication in larger independent cohorts before considering microbiota-directed interventions based on these data.
Limitations
Small sample (28 adolescent MDD, 34 adult MDD) with insufficient statistical power for high-dimensional multi-omics analyses — high risk of uncorrected false positives. Cross-sectional design precludes causal inference; cause versus consequence of depression cannot be distinguished. No formal risk-of-bias tool applied (RoB 2 or ROBINS-I not used); dietary, socioeconomic, and lifestyle confounders were not systematically controlled. 16S rRNA sequencing has limited taxonomic resolution at species level. Untargeted metabolomics generates hypotheses, not confirmations.
What is still missing
Longitudinal studies with larger samples and rigorous control of confounders (diet, exercise, sleep) are needed to validate identified signatures and test whether microbiota-directed interventions differentially alter depressive outcomes by age group.
Technical appendix
Version history
- 1.0 · 2026-10-14 — Auto-generated under Evidence Standard v1.0