PERIOMED ORIGINAL 0521 OCTOBER 2026

ORAL MICROBIOME · NEUROINFLAMMATION · CAUSALITY

The oral microbiome–brain axis.

Current evidence and proposed mechanisms.

FIRST AUTHOR · ORAL & PERIODONTAL SCIENCESImene Belhout, DMD, MScAlgeria

SECOND AUTHOR · ORAL & PERIODONTAL SCIENCESGiorgia Pieretto, DDSItaly

SCIENTIFIC REVIEWERMartina Sansavini, DDSItaly

ABSTRACT

A compelling biological framework.
An open causal question.

The oral microbiome may interact with neurological health through inflammatory, microbial, neural and metabolic pathways. Research has examined links with Alzheimer's disease, Parkinson's disease, stroke and mood disorders.

The evidence is not all of one kind. Experimental models support biological plausibility; human studies mainly show associations; genetic analyses are inconsistent; and interventional evidence has not established neurological benefit. This review asks what each layer of evidence can—and cannot—tell us.

FOUR PROPOSED ROUTES

From an oral ecosystem to a distant organ.

These pathways may overlap. For most, direct human evidence remains limited.

Bloodstream & inflammation

Microbial products and host mediators may enter the circulation and influence vascular and inflammatory signaling. [2] [18]

Direct neural routes

Trigeminal and olfactory pathways are anatomically plausible, but direct microbial dissemination remains insufficiently demonstrated in humans. [2]

Oral–gut–brain route

Swallowed oral bacteria may alter gut microbial and inflammatory networks. Cross-sectional data cannot establish temporal direction. [1] [3]

Metabolites & immune signaling

Metabolites and microbial structures may modify systemic immune responses, but their oral-specific contribution is not yet resolved. [1] [4]

ALZHEIMER'S DISEASE · THE BEST-STUDIED CASE

One hypothesis.
Three very different evidence layers.

Strength at one level does not automatically carry over to the next.

01MECHANISTIC

Biological plausibility

Porphyromonas gingivalis, gingipains and inflammatory pathways have been studied in brain tissue and experimental models. Commercial connections and model limitations matter when interpreting the findings. [5] [17]

Supports a hypothesis

02EPIDEMIOLOGICAL

Modest associations

Recent longitudinal syntheses link periodontitis and tooth loss with later cognitive outcomes, but heterogeneity, residual confounding and reverse causation temper causal interpretation. [7] [8]

Does not prove direction

03INTERVENTIONAL

No established benefit

Atuzaginstat did not meet its co-primary endpoints in the overall GAIN population. LHP588 is under evaluation in the phase 2 SPRING trial; efficacy remains unknown. [11] [12]

The clinical test remains open

BEYOND ALZHEIMER'S

Signals across neurology and mental health.

Parkinson's disease

An oral–gut signal; temporality unresolved.

Cross-sectional metagenomic findings link oral-to-gut translocation and virulence patterns with cognitive impairment. The design cannot determine cause or consequence. [3] [16]

Stroke

Association within a shared vascular context.

Periodontitis and post-stroke cognitive impairment have been linked, but vascular, metabolic and social determinants complicate direct microbial interpretations. [14] [15]

Mood & trauma

Likely bidirectional and behaviorally entangled.

Stress, oral-care behavior, periodontal inflammation and microbial composition may interact. Cross-sectional taxa differences do not establish a microbial cause of symptoms. [4] [13]

Other conditions

Preliminary, not clinically actionable.

Autism spectrum disorder, bipolar disorder, schizophrenia and epilepsy have been discussed, but current evidence is insufficient for causal or therapeutic claims. [2]

ASSOCIATION DOES NOT MEAN CAUSATION

Five questions every claim must survive.

Shared risksCould smoking, diabetes, diet, ageing or access to care explain part of the relationship?
Reverse causationCould neurological decline itself make oral care and dental attendance more difficult?
Genetic evidenceWhy do Mendelian-randomization studies reach different conclusions across phenotypes and instruments? [9] [10]
DetectionDoes finding a microbial product in diseased tissue show that it initiated the disease?
InterventionDoes changing the oral exposure improve a meaningful neurological outcome?

PRACTICAL IMPLICATIONS

Protect oral health.
Do not promise neurological prevention.

Connected biology deserves connected research—and proportionate clinical language.

Good oral hygiene and appropriate periodontal care remain important for oral health, function and quality of life. Current evidence does not establish that these measures prevent neurodegenerative disease.

Interdisciplinary collaboration may improve care when neurological conditions make daily oral hygiene or access to treatment more difficult. It should not be presented as a basis for unvalidated neurological screening in dental settings.

Probiotics, prebiotics, dietary strategies and pathogen-targeted therapies remain investigational for cognitive outcomes.

WHAT THE NEXT STUDIES MUST DO

Test time, direction and modifiability.

Follow people over time

Measure oral health and the microbiome before neurological symptoms emerge.

Standardise outcomes

Use consistent periodontal, cognitive, clinical and biological endpoints.

Intervene rigorously

Run randomized studies powered for meaningful neurological outcomes.

Separate signal from story

Triangulate primary, genetic, mechanistic and trial evidence without recycling assumptions.

THE BOTTOM LINE