Gut-Derived Metabolites and Cognitive Health: Narrative Review on SCFAs and TMAO
This narrative review does not support causal inference between SCFAs/TMAO and cognitive outcomes; observational associations exist but human evidence remains preliminary and heterogeneous.
| Outcome | Grade | Direction | Effect | Studies |
|---|---|---|---|---|
| Fecal SCFA levels and Parkinson's disease severity | C | ▲ Favorable | associação inversa qualitativa; sem tamanho de efeito reportado com IC | 3 |
| SCFA profiles and amyloid status/cognitive decline in AD | C | ▲ Favorable | associação qualitativa; sem IC 95% | 2 |
| TMAO in human cerebrospinal fluid | C | — Neutral | detecção qualitativa de TMAO no LCR; sem tamanho de efeito | 2 |
| Plasma TMAO and mild cognitive impairment | C | ▼ Unfavorable | associação positiva qualitativa TMAO↑ com CCL; sem OR/RR com IC | 1 |
| TMAO and Alzheimer's disease biomarkers | C | ▼ Unfavorable | associação positiva qualitativa; sem IC 95% | 2 |
| Blood-brain barrier integrity (experimental models) | D | ▲ Favorable | efeito protetor de SCFAs em modelos in vitro/animal; sem dados humanos quantitativos | — |
| Causal inference SCFA/TMAO → cognition in humans | D | — Insufficient | nenhum RCT disponível; zero dados causais | — |
Context
SCFAs and TMAO are microbial metabolites with mechanistic plausibility for modulating brain function via the gut-brain axis. Most available human studies are observational and cross-sectional, precluding causal conclusions. A narrative synthesis is warranted to map gaps before interventional trials are designed.
What the study showed
Observational studies report that reduced fecal SCFA levels associate with greater clinical severity in Parkinson's disease (Unger, Tan, Chen) and that fecal SCFA profiles vary by amyloid status and cognitive decline trajectory across the AD continuum (Kuehn 2025). Elevated plasma TMAO was associated with mild cognitive impairment in elderly individuals at high cardiovascular risk (Buawangpong) and with AD biomarkers and neuroimaging features (Vogt, Yaqub). No study provides pooled standardized effect sizes with 95% CIs; data are qualitative and heterogeneous. Human causality has not been established.
How it was done
Non-systematic narrative review; PRISMA not applied; no formal search strategy, explicit inclusion criteria, or risk-of-bias assessment. Integrates human cross-sectional and cohort observational studies, animal models, and mechanistic data. Total number of included studies is not declared.
Effect magnitude
No pooled effect size with 95% CI was calculated or reported; the review is qualitative and does not permit effect magnitude quantification.
Limitations
Narrative review without PRISMA, without formal risk-of-bias assessment (RoB 2, ROBINS-I, or AMSTAR-2 not applied), subject to selection and publication bias. All cited human studies are observational/cross-sectional, precluding causal inference. Uncontrolled confounders include diet, medications, renal function, disease stage, and comorbidities. SCFAs measured in feces rather than plasma or CSF limits systemic inference.
In clinical practice
Clinicians should not alter clinical practice based on this review; no interventional recommendation is supported by the presented data. Monitor future RCTs with SCFAs or TMAO modulators in cognitively at-risk populations before any clinical application.
What is still missing
Longitudinal studies and RCTs are needed to test causality between SCFA/TMAO profiles and cognitive outcomes. Standardization of metabolite measurement methods (fecal vs. plasma vs. CSF) is a prerequisite for comparability.