Microbiota-mitochondria axis in neurodegenerative disorders and retinal dysfunction: mechanisms and emerging therapeutic opportunities
This narrative review does not establish causality or effect magnitude between the microbiota-mitochondria axis and neurodegenerative or retinal diseases — the evidence base remains insufficient to guide clinical practice.
| Population | Patients with neurodegenerative diseases (Alzheimer's, Parkinson's) and retinal diseases (AMD, diabetic retinopathy) or experimental models |
|---|---|
| Intervention | Strategies targeting the microbiota-mitochondria axis (probiotics, postbiotics, MitoQ, SS-31, urolithin A, nicotinamide riboside, coenzyme Q10) |
| Comparator | Not applicable (narrative review without pre-specified comparator) |
| Outcome | Neuronal mitochondrial function; Alzheimer's disease progression; Parkinson's disease progression; Age-related macular degeneration; Gut SCFA production; Efficacy of mitochondrial interventions (MitoQ, SS-31, urolithin A); Neuroinflammation (NF-κB, NLRP3, IL-6, TNF-α) |
Summary of findings
| Outcome | Effect | 95% CI | Certainty | Clinical relevance | Notes |
|---|---|---|---|---|---|
| Neuronal mitochondrial function | no pooled effect size reported; mechanistic data only from in vitro and animal models | — | Very low | — | |
| Alzheimer's disease progression | no quantitative synthesis; narrative compilation of heterogeneous observational and preclinical data | — | Very low | — | |
| Parkinson's disease progression | no pooled effect size; mechanistic associations from animal models and observational studies | — | Very low | — | |
| Age-related macular degeneration | no quantitative synthesis; mechanistic associations only | — | Very low | — | |
| Gut SCFA production | descriptive data only; Proteobacteria expansion from <10% to 20-30% in dysbiosis reported without IC 95% | — | Low | — | |
| Efficacy of mitochondrial interventions (MitoQ, SS-31, urolithin A) | no pooled RCT data for neurodegenerative endpoints; preclinical and early-phase data only | — | Very low | — | |
| Neuroinflammation (NF-κB, NLRP3, IL-6, TNF-α) | no quantitative synthesis; pathway associations from in vitro and animal models without human clinical data | — | Very low | — |
Context
Mitochondrial dysfunction and gut dysbiosis co-occur in Alzheimer's, Parkinson's, and retinal degeneration, but causal relationships in humans are not established. Narrative reviews synthesize mechanistic hypotheses but do not replace controlled clinical trials. The clinical relevance of microbial metabolites (SCFAs, LPS, TMAO) in neurodegeneration requires prospective and interventional data not provided by this study.
What the study showed
The review compiles mechanistic evidence that dysbiosis reduces protective SCFAs and elevates circulating LPS, activating inflammatory pathways (NF-κB, NLRP3) that impair neuronal and retinal mitochondrial function. No aggregated data with effect sizes, 95% CIs, or p-values from original quantitative analyses are reported. Cited studies are heterogeneous in design (in vitro, animal models, observational, isolated RCTs), precluding unified conclusions on direction or magnitude of effect in humans. No evaluated therapeutic intervention demonstrated consolidated clinical efficacy for neurodegenerative outcomes in this review.
How it was done
Narrative review with systematic database search (PubMed, Scopus, Web of Science, Google Scholar) covering articles up to January 2025. No registered protocol, PRISMA criteria not explicitly declared, and study selection did not follow a documented double-blind process. Synthesis is qualitative, without meta-analysis.
Effect magnitude
No aggregated effect size is calculated. Individual data cited (e.g., Proteobacteria expansion from <10% to 20–30% in dysbiosis) are descriptive and extracted from heterogeneous primary studies without 95% CI reported in this review.
Risk of bias
Narrative review without registered protocol, without applied risk-of-bias tool (AMSTAR-2 not used), with high risk of selection and confirmation bias. Heterogeneity of primary studies (in vitro, animal, human) precludes causal inference. Absence of quantitative synthesis and GRADE assessment limits recommendation strength.
What this study does NOT prove
This study does not prove causality between gut dysbiosis and neurodegeneration in humans, nor the efficacy of any therapeutic intervention targeting the microbiota-mitochondria axis. Described mechanisms are largely derived from animal models and in vitro studies without clinical validation.
In clinical practice
In the absence of consolidated clinical efficacy evidence, none of the described interventions (MitoQ, SS-31, urolithin A, specific probiotics) should be prescribed based on this review for prevention or treatment of Alzheimer's, Parkinson's, or retinal diseases. Clinicians should await results from RCTs with validated clinical endpoints before incorporating these strategies into practice.
Limitations
Narrative review without registered protocol, without applied risk-of-bias tool (AMSTAR-2 not used), with high risk of selection and confirmation bias. Heterogeneity of primary studies (in vitro, animal, human) precludes causal inference. Absence of quantitative synthesis and GRADE assessment limits recommendation strength.
What is still missing
Phase II/III RCTs with primary clinical endpoints (cognition, retinal progression, mortality) and validated mitochondrial biomarkers in neurodegenerative populations. Large-scale Mendelian randomization studies to establish causality between microbiome composition and neurodegeneration.
Technical appendix
Version history
- 1.0 · 2026-10-03 — Auto-generated under Evidence Standard v1.0