In plain terms

The claim being assessed: Oral microbial pathogens, principally Porphyromonas gingivalis, causally drive systemic cardiovascular and neurodegenerative disease in humans.

Where the evidence stands: Γ — Plausible mechanism. There is a believable biological story and partial support, but the causal step has not been demonstrated.

What this mechanism does: This makes the process worse or keeps it going, but something else starts it.

Strongest evidence for it: Γ, Moderate confidence, for the preregistered human vascular-and-neurodegenerative bundle. Scope-matched aggregation takes the highest subclaim whose scope matches the central claim; only the co-central human subclaims SC3 and SC4 match that scope, and both are Γ. Primary-subclaim sensitivity is Γ; weakest-link sensitivity is Β because the presence/co-occurrence subclaim is observational. Model causality and systemic physiological effects are credited at Ε in their stated scopes rather than allowed to replace human disease causality. The convergence was assessed as a set; the missing link is specific human clinical mediation, not an automatic requirement for an impossible experiment.

What decisive evidence would look like

In plain terms, researchers would change the proposed cause, confirm it changed as intended, and measure a meaningful health outcome while checking other explanations. The result should be able to support or count against the claim.

Relative explanatory fit: Alternative ranked ahead

What this review means for Oral Microbiome → Systemic Disease

Claim being assessed: Oral microbial pathogens, principally Porphyromonas gingivalis, causally drive systemic cardiovascular and neurodegenerative disease in humans.. This review asks whether the proposed factor causes the health outcome—not just whether the two appear together. For the wording assessed here, the evidence is rated Plausible mechanism (HSEF Γ; Moderate confidence). There is a believable biological story and partial support, but the causal step has not been demonstrated. That rating applies to this claim and scope; it does not mean the biological process is irrelevant in every setting.

Implications for health and treatment: This is an assessment of a causal explanation, not proof that a particular supplement, diet, test, or treatment improves health. A treatment claim needs studies that test the intervention in people and measure meaningful benefits and harms. These findings can help frame research and informed clinical questions, but they do not identify what is right for one person.

Limits for individual care: Evidence from cells, animals, associations, or a plausible mechanism may not predict what happens in people. The grade is limited to the exact claim, outcomes, and sources reviewed, and it is not a diagnosis or an individual forecast. Do not start, stop, or change care based on this page; discuss personal decisions with a qualified clinician.

Medical information notice: This review provides general health information, not medical advice. It cannot diagnose a condition or replace an individual discussion with a qualified clinician.

This assessment rests on 69 sources and 75 individually checked claims. 75 of those claims were verified against the original paper.

A low grade is not an accusation. It means the causal work has not been done yet, which is the ordinary state of most actively studied mechanisms in biology.

At a glance

Sub-claims assessed separately

ID	Sub-claim	Type	Level	Pivotal card
SC1	Severe periodontal disease co-occurs with cardiovascular events and incident dementia.	existence	Β	—
SC2	Oral P. gingivalis inoculation accelerates vascular and neural pathology in rodents.	sufficiency	Ε	Card ↗
SC3	Oral microbial mechanisms causally drive human atherosclerotic cardiovascular disease.	generalization	Γ	Card ↗
SC4	Oral microbial mechanisms causally drive human Alzheimer disease.	generalization	Γ	Card ↗
SC5	Oral inflammation contributes to systemic inflammatory tone.	mechanism	Ε	—

Visual Evidence Cards · Pivotal Subclaims

Pivotal subclaim illustrations

Dark-theme simulated-micrograph renders and concise causal evidence summaries for the pivotal subclaims that govern this review's grade. Click any card to inspect the master 2400×2400 render.

Master 2400×2400 render ↗

Master 2400×2400 render ↗

Master 2400×2400 render ↗

Figures

Explanatory figures summarizing this review. They carry no evidential weight of their own; where a figure and the review differ, the review governs.

Full assessment

From here the review is written for a technical reader: caps and thresholds, competing models, the independence basis, and the citation ledger.

Sources: 69 primary reports; claims: 75; verified: 75; unverified: 0. Verification is bounded attribution. Access: 10 partial-body and 59 complete abstract/status examinations. Literature cutoff: 2026-10-06; source examination UTC date: 2026-10-06. No complete-paper audit or blinded calibration claimed. Adversarial review: Generated by Sol 6.1; adversarially reviewed by Claude Opus 5.5 (Anthropic) on 2026-10-06. 1 correction applied (see ADVERSARIAL REVIEW AND APPROVAL). Status: Approved as revised.

AT A GLANCE

Oral inflammation has systemic effects; oral pathogens are not yet established as causal drivers of human ASCVD and Alzheimer disease.

Deciding experiment: Run a prospectively pathogen-defined clinical programme with selective Pg suppression versus matched periodontal inflammatory care, documented oral and systemic/CNS engagement, prespecified interactions and independently adjudicated clinical vascular or AD outcomes in separate disease cohorts.

Belief versus evidence: Γ, Moderate confidence, for the preregistered human vascular-and-neurodegenerative bundle. Scope-matched aggregation takes the highest subclaim whose scope matches the central claim; only the co-central human subclaims SC3 and SC4 match that scope, and both are Γ. Primary-subclaim sensitivity is Γ; weakest-link sensitivity is Β because the presence/co-occurrence subclaim is observational. Model causality and systemic physiological effects are credited at Ε in their stated scopes rather than allowed to replace human disease causality. The convergence was assessed as a set; the missing link is specific human clinical mediation, not an automatic requirement for an impossible experiment.

Best-supported rival: Systemic inflammatory burden. Compare selective pathogen suppression against equally effective nonspecific dental inflammation control, with baseline pathogen/CNS measurements and blinded clinical outcomes. This distinguishes microbial mediation from inflammatory burden and common-cause association.

COMPETING MECHANISTIC MODELS

Rank 1: Systemic inflammatory burden (Ε). Oral inflammatory burden can affect vascular physiology and interact with existing systemic disease. Randomized BP, FMD and structural vascular changes and model host perturbations support bounded effects. Null CAD FMD and PAVE hsCRP contrasts, transient early inflammation and nonspecific periodontal treatment limit universality.

Rank 2: Common-cause susceptibility (Β). Smoking, deprivation, metabolic disease, host genetics and health-care access contribute to both dental and systemic outcomes. Shared genetic architecture, observational heterogeneity and nonspecific exposure proxies favor substantial common causes. Common causes alone cannot explain all randomized physiological effects or strain-resolved model phenotypes.

Rank 3: Direct oral microbial pathogenesis (Γ). Oral pathogens directly drive human vascular and Alzheimer disease. Oral exposure, gingipain/FimA subtraction and inhibitor rescue are the strongest recipient-model evidence. Clinical microbial mediation remains unresolved; oral null/intracerebral positive results and nonspecific human cohorts restrict the claim.

Rank 4: Antimicrobial amyloid response (Γ). Amyloid can participate in an innate response to microbial exposure. The Dominy model links infection, Aβ induction and in-vitro antimicrobial activity. This does not show that human plaques are net protective or that infection is necessary for sporadic AD.

Rank 5: Secondary colonization after barrier failure (Γ). Age or neurodegeneration can permit microbial components or DNA to enter damaged tissues. Postdiagnostic detection and BBB pathway models are compatible with secondary entry. Early model pathogen perturbations can cause injury; temporal human barrier-first colonization is not established by compatibility.

The tournament does not alter the nominal level. Compatible pathways and flexible alternatives are not automatically proved by explaining observations.

FULL REVIEW

Registered claim and scope

Oral microbial pathogens, principally Porphyromonas gingivalis, causally drive systemic cardiovascular and neurodegenerative disease in humans.

Use the index central claim and disease-separated subclaims. The human vascular and Alzheimer applications are co-central; the lower of SC3 and SC4 is the scope-matched bundled gate. Animal inoculation and inflammatory-tone changes are scored separately, and cannot replace human disease causality. This co-central interpretation of scope-matched is declared before scoring.

Selection: public-interest rank 21 from the editorial ranking compiled 2026-09-23; categorical prior verdicts were visible. Private fresh review; earlier documents are preserved.

Evidence and causal interpretation

The epidemiology is substantial but heterogeneous. Korean brushing/stroke, ARIC dementia and NHANES interactions are positive, whereas Swedish dementia, ARIC eight-year cognitive change and most adjusted tooth-count comparisons are null. Periodontal disease, antibody response, tooth loss and hygiene behavior are different exposures. Positive WHICAP and Kentucky antibody results are not consistently Pg-specific (PMID32432793, PMID32727837, PMID35099831, PMID25363061, PMID25597561, PMID25522313, PMID22546352).

Human vascular interventions are an important positive part of the case. Tonetti changes FMD with an initial inflammatory worsening; a hypertensive trial lowers ambulatory BP; the 2025 trial changes two-year carotid structure. Recent-STEMI FMD is positive but a stable-CAD trial is null. These results establish bounded systemic effects while leaving clinical event mediation and pathogen identity unresolved (PMID17329698, PMID31504461, PMID40827724, PMID31810741, PMID29873864).

PREMIERS directly tests hard clinical events and cannot be omitted as if no such randomized study existed. Intensive versus standard active care was not superior in 280 stroke/TIA patients; the small event count leaves broad uncertainty. PAVE is a feasibility pilot with considerable community-care crossover, not a definitive event-efficacy test. Positive post-randomization subgroup analyses do not preserve randomization (PMID37548008, PMID19186958).

Human genetics supplies both positive susceptibility associations and negative MR analyses. The JACC UK Biobank paper uses a PRS, explicitly avoids MR because of instrument weakness, and acknowledges pleiotropy. Null host-liability instruments do not exclude acquired microbial effects. Shared genetic architecture supports alternatives without proving that a shared variant acts through oral infection (PMID39290820, PMID19214202, PMID25466412, PMID35052354, PMID38648391, PMID41916952).

Rodent vascular experiments support genuine conditional effects: FimA or gingipain subtraction, TLR2 necessity and protective NOD2 activation alter vascular pathology. Differences in bacterial burden, host lipid context and forced exposure leave important mediation questions. Pathogens can aggravate induced infarction as well as atheroma; these are different causal phenotypes (PMID15123526, PMID25663345, PMID24002697, PMID24324141, PMID40859845).

Dominy supplies oral infection and inhibitor contrasts, but its human PCR signals include nondemented controls and its small CSF pilot lacks a matched healthy CSF arm. Oral inhibitor benefit is coupled to lower bacterial burden; a separate direct-gingipain injection establishes toxicity after the barrier is bypassed. Gad67 counts, not an audited total-neuron survival assay, underpin one protective result. Other models show oral-null/intracerebral-positive Aβ and less pathology after coinoculation, mapping the limits rather than cancelling the positive model results (PMID30746447, PMID35328748, PMID39255392).

The 2026 clock paper has a six-month exposure protocol, antibiotic conditioning and cognitive/pathology effects. Its AKT and Bmal1 manipulations occur in cultured glia, so they do not establish in-vivo clock-mediated necessity. The twelve-week dual-pathogen study is a different experiment. This is a concrete correction to the broad Consensus synthesis (PMID42271186, PMID39255392).

The supplementary GAIN registry check records an overall efficacy null and sponsor-claimed pathogen-positive subgroup benefits. It is linked separately in registered-trial-check.md, outside the 69 peer-reviewed-paper count and positive threshold evidence. Consensus did not supply this trial full text. Its value here was targeted body leads and gap structure; original checks prevented overinterpretation of PRS and conflation of exposure durations.

HYPOTHESIS STATE EVALUATION FRAMEWORK (HSEF v5.3)

Subclaim matrix and aggregation

Preregistered aggregation: scope-matched. Γ, Moderate confidence, for the preregistered human vascular-and-neurodegenerative bundle. Scope-matched aggregation takes the highest subclaim whose scope matches the central claim; only the co-central human subclaims SC3 and SC4 match that scope, and both are Γ. Primary-subclaim sensitivity is Γ; weakest-link sensitivity is Β because the presence/co-occurrence subclaim is observational. Model causality and systemic physiological effects are credited at Ε in their stated scopes rather than allowed to replace human disease causality. The convergence was assessed as a set; the missing link is specific human clinical mediation, not an automatic requirement for an impossible experiment. Sensitivities: weakest-link Β; primary-subclaim Γ; scope-matched Γ.

Decisive-perturbation inventory and causal tally

[1] Ligature plus patient plaque and oral F. nucleatum accelerated atheroma in high-cholesterol atheroprone mice; hepatic lipid synthesis was linked to glycolytic signaling. Forced exposure, ApoE/diet context and in-vitro pathway inhibition do not establish pathogen-specific human ASCVD mediation.

[2] Ligature/plaque experiments, gnotobiotic mice, host knockouts and adhesion mutants linked oral pathobionts to B2-cell mobilization and worse experimental infarction; the human arm was associational. This tests aggravation after induced MI, not spontaneous human coronary initiation or proof that Pg alone is necessary.

[4] Periodontitis-linked genetic instruments and a 101-person hypertensive RCT supported a periodontal-treatment effect on two-month ambulatory BP, with cytokine and T-cell changes. Short-term BP/endothelial effects are narrower than atherosclerotic events; host-instrument pleiotropy and nonspecific treatment do not isolate Pg.

[7] Four-week Pg/S. gordonii oral lavage in ApoE-deficient mice produced bone loss, hepatic inflammatory expression and impaired aortic acetylcholine relaxation without detected caecal-community differences. Mixed-pathogen exposure and atheroprone genotype do not isolate a human Pg pathway; endothelial function is not a hard clinical event.

[10] Two-year randomized subantimicrobial doxycycline lowered hs-CRP and MMP-9 in 128 postmenopausal periodontal-maintenance participants, without an overall lipid effect. Systemic doxycycline is a host-modulating drug, so the effect cannot be assigned to elimination of oral microbes; no CAD event benefit was tested.

[11] Pg challenge increased early atheroma, macrophages and local innate markers in ApoE-knockout mice; prior immunization prevented this local response without elevated systemic mediators. Artificial challenge and prophylactic immunization demonstrate model capacity, not spontaneous human disease or an obligatory systemic-CRP pathway.

[12] A 52-person randomized study found periodontal treatment prevented worsening of platelet activation over three months, with differences in turnover and collagen responses. The treated arm was largely unchanged rather than normalized; platelet assays do not demonstrate prevention of MI or stroke.

[16] Cyclic di-AMP prophylaxis reduced periodontal bone loss, lipids and atheroma in Pg-challenged, ligature-treated high-fat ApoE-knockout mice. A broad STING/trained-immunity adjuvant is not pathogen-specific rescue or proof of necessary human Pg infection.

[17] Human gingipain/tau associations accompanied mouse oral-infection brain bacterial/amyloid findings and inhibitor protection in neurotoxicity paradigms. Postmortem detection cannot establish antecedent oral-to-brain transit in humans; the routes, outcomes and controls of distinct mouse experiments must remain separate. Original sections show PCR positivity also in five of six nondemented controls, no healthy CSF comparison in the ten-AD-person pilot, direct hippocampal gingipain injection in one toxicity experiment, and oral inhibitor effects coupled to reduced pathogen burden. Gad67-positive interneuron counts are not a complete unbiased neuron-survival audit.

[20] Two-month gavage of periodontal-patient salivary communities worsened cognition, amyloid and inflammation in APP/PS1 transgenic mice compared with healthy saliva. Crude transfer does not identify a necessary Pg strain, separate gut/community/inflammation mediation, or establish endogenous human AD.

[21] Gingival exposure to Pg or its EVs impaired mouse memory, unlike oral gavage; labeled-vesicle transport to the ipsilateral trigeminal ganglion was reduced by neurectomy. Label movement is not intact replicating-organism transit; neuronal markers are not stereological loss, and route-specific inoculation bounds generalization.

[22] Repeated tail-vein bacteremia increased rat BBB leakage; cellular Cav-1 reduction and Mfsd2a overexpression reversed permeability responses. Forced bloodstream delivery and cell-level rescue do not establish an oral initiating route or rescue of human dementia.

[23] Pg-induced periodontitis impaired cognition in wild-type mice and worsened APP/PS1 decline, with IFN-beta, glial IFITM3 and amyloid changes. The abstract does not report IFITM3-specific loss/rescue; correlated pathway activation does not isolate the mediator or establish human initiation.

[24] Ligature periodontitis increased insoluble amyloid in 5xFAD mice, but glial and cytokine effects differed from wild type, including reduced plaque-associated microglia. State-specific responses oppose a simple uniform activation account; amyloid/glial changes do not demonstrate cognitive rescue or a Pg-specific cause.

[26] Eight-week oral OMV dosing impaired aged-mouse memory and increased barrier, inflammatory and phospho-tau abnormalities; MCC950 attenuated cellular inflammasome/conditioned-medium effects. The NLRP3 inhibition was in vitro, not a demonstrated in-vivo cognitive rescue; forced vesicle dose and substrate assays do not establish human AD.

[27] Daily intraperitoneal Pg LPS worsened memory, synaptic proteins and tau phosphorylation in APP knock-in mice without changing amyloid; cellular GSK3/TNF interventions supported a pathway. A systemic endotoxin regimen is not oral infection, and cell-level inhibition is not organismal mediator-specific rescue.

[32] Oral infection in ApoE-deficient mice yielded Pg brain DNA at 12/24 weeks and neuronal complement-fragment deposition in a subset. Bacterial DNA and complement staining do not prove viable CNS replication, generalized neuronal death or spontaneous human initiation.

[34] A month of oral Pg administration impaired mouse cognition with gut-composition, glymphatic, inflammatory and neuronal changes. Co-occurring routes and pathways were not separately rescued in the abstract; alpha diversity was unchanged and an induced model is not human mediation.

[39] Oral Pg increased phenylephrine vasoconstriction in healthy and ApoE-deficient mice, while acetylcholine and nitroprusside relaxation remained similar. The receptor-specific response is not uniform endothelial failure and does not demonstrate human clinical-event causality.

[40] Subcutaneous Pg after femoral injury increased neointima and MCP1 in wild-type mice; TLR2 deficiency prevented this response despite antibody induction. Arterial injury and forced systemic exposure establish a recipient-model dependency, not spontaneous oral-pathogen-mediated human ASCVD.

[41] Oral wild-type Pg increased plaque, bone loss and inflammatory signals in hyperlipidemic Apoe-shl mice relative to gingipain-null and FimA-deficient strains. Virulence mutations also change dissemination and burden; NLRP3 expression was measured without a reported NLRP3 necessity manipulation in the abstract.

[42] Oral wild-type Pg but not FimA-mutant Pg accelerated periodontal and atherosclerotic pathology in ApoE-null mice; immunization prevented the phenotype. Both strains were detected by PCR in blood and aorta; coupled changes in periodontal disease and virulence do not isolate invasion as the sole mediator.

[43] Pg impaired endothelial integrity and vasorelaxation; Sirt3 activation reversed mitochondrial and vascular defects in cells and orally exposed mice. Honokiol activation is not equivalent to a selective genetic rescue, and vascular physiology does not establish human clinical event prevention.

[44] Wild-type Pg caused endothelial junction loss and leakage in human cells and systemically exposed zebrafish, unlike gingipain-deficient bacteria. Cell culture and forced systemic zebrafish infection do not reproduce endogenous human oral-to-vascular mediation.

[45] NOD2 deficiency worsened oral-Pg inflammation, cholesterol and atheroma in ApoE mice; MDP stimulation reduced these phenotypes. NOD2 signaling was protective here; systemic MDP also affected weight and metabolic inflammation. The linked correction concerns an author name, not a new study.

[46] In 135 otherwise healthy periodontal patients, randomized intensive treatment improved FMD and reduced carotid intima-media thickness at 24 months; blood pressure and PWV did not substantially differ. The primary outcome was vascular structure, not MI, stroke or death; the PubMed abstract confidence interval is malformed and is not reproduced.

[48] Randomized intensive treatment in 120 periodontal patients transiently worsened FMD and inflammation at 24 hours but improved FMD at 60 and 180 days; no cardiovascular events occurred. Healthy participants and a surrogate endpoint limit clinical ASCVD inference; the parent paper has a correction to tPA variability reporting and a figure day label.

[54] Oral Pg altered oral and gut communities, barrier measures and AD-like pathology; 3-hydroxykynurenine exposure promoted neuronal apoptosis in model systems. The abstract does not report necessity blockade of the entire oral-to-brain pathway; indirect metabolic mediation remains distinct from direct CNS infection.

[65] Six months of oral Pg after antibiotic conditioning caused poorer water-maze performance, neuronal pathology and altered hippocampal rhythms; cultured-glial AKT activation and Bmal1 knockdown interrogated the clock pathway. Original Methods/Results specify twice-weekly exposure for 24 weeks after kanamycin, matched oral handling, and cell-line rather than in-vivo pathway rescue. Cognitive pathology is present; in-vivo clock-mediated necessity remains untested in the examined sections.

[66] Twelve-week oral Pg and T. denticola exposures produced differing brain pathology in female mice; coinoculation was less pathogenic than each alone. Pathogen combination and host context alter outcomes; AD-like histology is not human AD, and a stronger universal synergistic prediction failed in this context.

[67] In APP-TgCRND8 mice, chronic oral infection caused bone loss and gliosis without accelerating Aβ; live intracerebral Pg increased Aβ after eight weeks, unlike heat-killed bacteria. Direct brain injection bypasses the proposed oral route; peripheral and central results provide a route boundary rather than a general oral-infection proof.

[68] A randomized 48-person recent-STEMI trial improved six-month FMD after periodontal treatment; inflammatory markers and cardiovascular events did not significantly differ. The small trial's primary endpoint was endothelial physiology; it cannot establish prevention of recurrent clinical events or Pg-specific mediation.

Tally: 32 discriminating contrasts. Null and negative contrasts are discriminators, not positive threshold crossings. Reports and replications are not independent causal lines. Shared cohorts, experimental platforms, transgenic lineages and repeated outcomes do not constitute fully independent causal lines.

Cap audit, precedence and Β/Γ/Δ boundary

The microbiome rubric explicitly credits transfer for recipient-model subclaims while universal cap 5 restricts inference to endogenous human biology. SC2 is therefore Ε, not demoted for being experimentally inoculated, while SC3/SC4 require their own human evidence. Clinical physiology is a real treatment effect; cap 4 still blocks converting a marker or structural change into disease-outcome proof. Eradication trials, natural experiments and human genetic convergence remain valid upper-ladder routes. The assembled human case is not literally recapped as cell culture: it stops at unresolved microbial mediation.

Gamma hard-rule audit: Human surrogate improvements do not by themselves demonstrate clinical event mediation. Forced oral or intracerebral infection does not establish natural human oral-to-brain causality. Host genetic liability and polygenic scores are not equivalent to selective pathogen perturbation. Body sections and contrary controls are considered together; no blanket absence-of-RCT cap is imposed.

SC1 cap: “Association-only microbiome work caps at Β.” Prospective cohorts support co-occurrence, with substantial endpoint and population heterogeneity. Swedish dementia, ARIC cognitive decline, adjusted tooth-count and selected cardiovascular comparisons are null or mixed. Association is established as a recurring observation, not a universal risk relation or a microbial causal pathway.

SC1 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. Prospective cohorts support co-occurrence, with substantial endpoint and population heterogeneity. Swedish dementia, ARIC cognitive decline, adjusted tooth-count and selected cardiovascular comparisons are null or mixed. Association is established as a recurring observation, not a universal risk relation or a microbial causal pathway.

SC2 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: yes. Defined oral exposures, bacterial virulence subtraction and host-pathway perturbations produce vascular and neural phenotypes in particular rodent systems. Oral versus intracerebral infection and single versus dual inoculation map important failures. Recipient-model causality is credited, but a single conserved organism-specific mediator across both organ systems is not independently isolated.

SC3 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. Periodontal treatment causally changes blood pressure or vascular function in some populations, and a two-year randomized trial changes carotid structure. These are positive human effects. However, Pg-specific mediation of clinical atherosclerotic events is not established: PREMIERS did not show superiority on hard events, PAVE was a feasibility pilot with crossover, and genetic liability studies do not provide a specific infection instrument.

SC4 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: no. Human periodontal associations and low-biomass pathogen detection do not establish an antecedent oral cause of AD. Rodent toxicity/rescue is route- and host-dependent. Genetic analyses are null but imperfect infection instruments. The supplementary GAIN sponsor summary reports no overall efficacy and a positive pathogen-defined subgroup whose effect sizes and multiplicity could not be independently checked; it does not carry a positive threshold.

SC5 field-valid definition: “ - Gnotobiotic or transfer experiment showing phenotype movement, at minimum - Even without defined-consortium precision or robust rescue, a transfer paradigm in vivo is the minimum that crosses this threshold

Per Cap precedence above, a gnotobiotic or forced-transfer experiment crosses Δ for the recipient-animal subclaim it was actually run in, while universal cap 5 (seeding or forced-transfer paradigms cap at Γ) still blocks the inference from that result to endogenous human biology — score the two scopes separately rather than choosing between the rules.” Positive full-scope result: yes. Randomized periodontal interventions show context-dependent systemic physiological responses, including endothelial function and blood pressure, together with inflammatory changes. Acute treatment can initially increase inflammation; coronary FMD and PAVE hsCRP nulls map limits. This bounded contribution is not organism-specific proof of human ASCVD or AD causation.

Upward threshold ladder

SC1 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. Association-only presence is not promoted by more correlated cohorts; a separately registered causal subclaim requires a selective perturbation.

SC2 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: yes.

SC2 Ε: “Requires context-specific causal effect with defined limits. A transferable phenotype exists, but microbial identity or generalizability remains incomplete.” Satisfied: yes.

SC2 Ζ: “Requires: - Defined consortium or strain-level causal evidence - Rescue logic - Replication across at least two relevant models or cohorts” Satisfied: no. Cross-facility matched-colonization strain subtraction and complementation, with blinded neuronal/vascular function and live organism versus product tracing, would isolate the mediator and test failed contexts.

SC3 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. A clinical-outcome trial or rigorous natural experiment with selective microbial engagement and explicit mediation, sufficient events, and matched dental inflammatory care would cross the human gate.

SC4 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: no. A prospectively pathogen-defined AD study with validated CNS infection measurement and selective therapeutic engagement, clinical benefit, multiplicity-controlled interaction and independent replication would cross the gate.

SC5 Δ: “Requires a transfer paradigm or causal perturbation in vivo.” Satisfied: yes.

SC5 Ε: “Requires context-specific causal effect with defined limits. A transferable phenotype exists, but microbial identity or generalizability remains incomplete.” Satisfied: yes.

SC5 Ζ: “Requires: - Defined consortium or strain-level causal evidence - Rescue logic - Replication across at least two relevant models or cohorts” Satisfied: no. Mechanistically selective oral inflammatory perturbation, physiological rescue, timing and independent clinical contexts would distinguish shared systemic drug effects from oral mediation.

Upper ladder: Eradication with hard outcomes, strain-level complemented rescue and orthogonal human causal instruments were considered. Defined virulence controls establish bounded model capability. Human treatment studies do not selectively isolate microbial mediation, and the hard-event randomized contrast was nonsuperior with limited precision. Thus higher levels remain reachable without an all-purpose RCT demand.

Independence: Reports are grouped by PREMIERS, ARIC, NHANES III and WHICAP. UK Biobank imaging and PRS analyses share participants, while MR papers reuse summary instruments. Strain subtraction, host perturbation and human randomization exclude different errors, but no assembled set supplies a specific human clinical microbial mediator. Rodent replication across host genotypes is not automatically a second independent causal line.

Bacterial factor subtraction [42, 41, 17]: Counters nonspecific inoculation alone, but virulence changes also change colonization and burden.

Host pathway perturbation [40, 45, 43]: Counters a purely passive bacterial association in recipient models; pathways have independent systemic effects.

Randomized human periodontal treatment [48, 4, 46, 60, 69]: Counters baseline dental-care confounding for physiological treatment effects; lacks specific pathogen mediation of clinical disease.

Falsification and pitfalls

Expected oral amyloid acceleration failed in a tested host. PMID35328748 oral infection caused dental disease and gliosis without Aβ acceleration, while intracerebral live Pg was positive. Additional downgrade: False.

Uniform treatment response is contradicted. PMID29873864 CAD FMD and PMID19186958 ITT hsCRP were null; PMID17329698 showed early deterioration followed by improvement. Additional downgrade: False.

Hard vascular superiority was not demonstrated. PMID37548008 is a nonsuperior active-comparator phase II result with few events, not an equivalence proof. Additional downgrade: False.

Association and surrogate outcome substituted for clinical microbial mediation. SC3/SC4; separated in outcome-specific cards.

PCR or antigen presence treated as viable initiating infection. Dominy controls and low-biomass attribution.

Reports counted as independent studies. Named cohort groups and genetic-source overlap.

Outcome-defined successful scope. Host, route, baseline disease and exposure are fixed before results; failures remain included.

Adversarial steelman: above, then below

The strongest Δ case combines multiple randomized vascular effects, cIMT structural improvement, genetic information and pathogen rescue models. This clearly defeats the view that oral disease has only local effects. Yet a positive vascular response does not isolate Pg or show clinical ASCVD events, and it does not rescue the separate AD gate.

The strongest Β case points to confounding, indirect oral measures, weak MR instruments and low-biomass contamination. It understates actual bacterial virulence and host-pathway perturbations and randomized human physiological effects; the overall mechanistic case exceeds association alone.

Reject Δ for the bundled human claim on missing clinical microbial mediation; reject Β for the whole mechanistic claim because recipient and human physiological causal effects are real. Retain Β for SC1, Ε for the bounded successes and Γ for the human gates.

Final score, causal direction and M/S/D

Γ, Moderate confidence, for the preregistered human vascular-and-neurodegenerative bundle. Scope-matched aggregation takes the highest subclaim whose scope matches the central claim; only the co-central human subclaims SC3 and SC4 match that scope, and both are Γ. Primary-subclaim sensitivity is Γ; weakest-link sensitivity is Β because the presence/co-occurrence subclaim is observational. Model causality and systemic physiological effects are credited at Ε in their stated scopes rather than allowed to replace human disease causality. The convergence was assessed as a set; the missing link is specific human clinical mediation, not an automatic requirement for an impossible experiment.

Steps 1–2 do not describe all data: recipient-model subtraction and human randomization show causal effects beyond co-occurrence, and reverse causation is not the only direction. Step 3 lacks human initiation plus prevention plus rate limitation; step 4 lacks universal requirement. Step 5 fits bounded amplification of vascular or neural pathology and systemic inflammatory physiology. This is a conditional role inference, not a claim that human Pg-mediated AD has been proved.

Momentum: Mechanistic only. The 2025 Sirt3 endothelial model and 2026 PI3K/AKT-clock work interrogate model pathways; the two-year cIMT trial adds human physiology. No verified currently running pathogen-selective clinical-outcome study crossing both human gates is asserted.

Stability: Stable. No named pending result was verified that is likely to move the co-central human score imminently. A new target-engaged clinical result would matter, but a hypothetical trigger does not set Rising.

Divergence: Contested; no single implied-level arithmetic. These dimensions do not change level or confidence.

PROPOSED DECISIVE EXPERIMENTS

Run a prospectively pathogen-defined clinical programme with selective Pg suppression versus matched periodontal inflammatory care, documented oral and systemic/CNS engagement, prespecified interactions and independently adjudicated clinical vascular or AD outcomes in separate disease cohorts.

Readout: Clinical event or cognitive/functional trajectory with organism load, viability-sensitive measures and mediator timing. Controls: Matched dental inflammation improvement, placebo/active comparator, non-Pg pathogen comparison, blinded outcomes, contamination/spike-in controls and multiplicity plan. Supporting: Selective sustained engagement changes the registered clinical trajectory beyond matched inflammation control and replicates in the prespecified pathogen-defined population. Falsifying: Adequate engagement with sufficiently precise clinical nulls, or equal benefit from nonspecific inflammation control without pathogen mediation.

RESEARCH TIMELINE

KNOWLEDGE GAPS

Separate direct microbial invasion from systemic inflammatory, gut-metabolic and barrier routes; inhibitor benefit coupled to lower load cannot identify one mediator.

Determine whether viable CNS organisms precede AD pathology with sensitive contamination-controlled assays; postmortem DNA is insufficient.

Resolve clinical event benefit after periodontal treatment with adequate event counts; positive FMD or cIMT should motivate, not substitute for, that study.

Verify GAIN subgroup definitions, effect estimates, multiplicity, interactions and safety from complete primary results before pathogen-defined benefit is used causally.

Test complemented bacterial mutants at matched colonization across independent facilities, both sexes and prespecified host backgrounds.

Model protective immune signaling and acute treatment-related inflammatory spikes rather than treating all activation as harmful.

METHODS AND CONSENSUS CONTRIBUTION

Consensus was used through the signed-in browser for discovery, counterevidence, research-gap and timeline suggestions, and attempted checks of specific indexed full-text passages. Returned Methods/Results passages were reconciled with original-paper passages where available; failed retrievals, source substitutions and unverified aspects are recorded in the case reconciliation. Successful passage checks are distinguished from provider summaries. Provider screening counts, citation counts and meters were not used as primary counts or HSEF scores. This source log records access depth; it does not claim complete-paper or supplement review. The monthly API allowance was conserved: zero Consensus API calls in this batch.

REVIEWER NOTES

69 primary papers, ten partial original-body checks and one supplementary primary sponsor registry summary. No complete-paper/supplement audit or blinded calibration. Secondary evidence commentary PMID33772118 and the duplicate CCDC91 conference report were excluded from the primary count. Linked corrections to PMID17329698 and PMID24324141 remain attached; no malformed abstract interval is reproduced. Consensus queries: two Pro, zero new Deep, zero API.

Search leads, reviews, protocols and correction notices do not pad the primary count. Downloaded XML does not earn examination credit. Linked notices remain attached; unexamined correction details limit inference. This is a coordinator source/claim consistency assessment, not independent blinded calibration.

ADVERSARIAL REVIEW AND APPROVAL

Authoring model: Sol 6.1. Adversarial reviewer: Claude Opus 5.5 (Anthropic). Date: 2026-10-06.

Scope of checks. PMIDs verified against PubMed: 69/69 resolve to the stated title, first author, year and journal (no retractions or expressions of concern found). Load-bearing claims spot-checked against PubMed abstracts: 27 (12 in a first pass, including Dominy PMID30746447, UK Biobank PRS PMID39290820, PREMIERS PMID37548008, PAVE PMID19186958, PMID35328748, PMID39255392, PMID35099831 and PMID25363061; 15 in a second pass covering evidence cards [4], [10], [12], [40], [41], [42], [45], [46], [48], [68] and timeline rows 2004, 2007, 2009, 2013, 2014, 2018, 2020 and 2026). All were accurate. Scoring logic checked: hard caps (SC1 association cap at Β; universal cap 5 blocking transfer-to-human inference for SC3/SC4), Β/Γ/Δ boundary, declared aggregation rule against the framework text, consistency of subclaim levels with their rationales, competing-model levels against the subclaim matrix, and the timeline years against the cited records.

Rejected findings (checked, no change). - The pre-scoring declaration in "Registered claim and scope" (and scorecard aggregation_rationale, hypothesis.conjunctive_note) glosses the rule as "the lower of SC3 and SC4". It is preserved verbatim as the preregistration record; because SC3 = SC4 the gloss has no effect on the letter, and only the post-scoring derivation was corrected. - Could scope-matched return Ε via SC5? No: SC5's scope is inflammatory/endothelial outcomes, not human ASCVD or AD causation. Raising the letter would substitute surrogate physiology for the central claim, which the framework's cap 4 forbids. - Timeline 2020 for PMID31810741 (PubMed e-pub December 2019; issue 2020) and reference year 2014 for PMID24002697 (e-pub 2013) are within the ±1 print/e-pub convention. - The GAIN/atuzaginstat registry summary is correctly kept outside the 69-paper count and carries no positive threshold. - Γ for SC3 is defensible: the randomized human vascular effects (FMD, ambulatory BP, cIMT) are real but surrogate. PREMIERS was not superior on hard events, and no trial isolates Pg-specific mediation. A Δ promotion would need a clinical-outcome perturbation.

Residual limitations. Spot-checks used PubMed abstracts plus the case's own partial-body passages, not complete papers or supplements. The GAIN subgroup claims remain unverifiable from primary data. Consensus was not queried by the reviewer, under the batch instruction.

Approval: Following adversarial review and the corrections listed above, the revised document is approved. — Claude Opus 5.5 (Anthropic), 2026-10-06.

REFERENCES

1. Zhou LJ, Lin WZ, Meng XQ et al. Periodontitis exacerbates atherosclerosis through Fusobacterium nucleatum-promoted hepatic glycolysis and lipogenesis. Cardiovascular research. 2023. PMID 36943793. DOI: 10.1093/cvr/cvad045. Access: abstract_metadata. Study: PMID36943793.

2. Chen BY, Zhu H, Li YL et al. Oral Pathobionts Aggravate Myocardial Infarction Through Mobilization of B2 Cells. Circulation. 2025. PMID 40859845. DOI: 10.1161/CIRCULATIONAHA.125.074837. Access: abstract_metadata. Study: PMID40859845.

3. Chang Y, Woo HG, Lee JS et al. Better oral hygiene is associated with lower risk of stroke. Journal of periodontology. 2021. PMID 32432793. DOI: 10.1002/JPER.20-0053. Access: abstract_metadata. Study: PMID32432793.

4. Czesnikiewicz-Guzik M, Osmenda G, Siedlinski M et al. Causal association between periodontitis and hypertension: evidence from Mendelian randomization and a randomized controlled trial of non-surgical periodontal therapy. European heart journal. 2019. PMID 31504461. DOI: 10.1093/eurheartj/ehz646. Access: full_text_partial. Study: PMID31504461.

5. Johansson CS, Ravald N, Pagonis C et al. Periodontitis in patients with coronary artery disease: an 8-year follow-up. Journal of periodontology. 2014. PMID 23725030. DOI: 10.1902/jop.2013.120730. Access: abstract_metadata. Study: PMID23725030.

6. Sen S, Redd K, Trivedi T et al. Periodontal Disease, Atrial Fibrillation and Stroke. American heart journal. 2021. PMID 33503409. DOI: 10.1016/j.ahj.2021.01.009. Access: abstract_metadata. Study: ARIC.

7. Parvaneh M, Witting PK, Ku J et al. Periodontitis induces endothelial dysfunction in mice. Scientific reports. 2021. PMID 34294791. DOI: 10.1038/s41598-021-94418-8. Access: abstract_metadata. Study: PMID34294791.

8. Tiensripojamarn N, Lertpimonchai A, Tavedhikul K et al. Periodontitis is associated with cardiovascular diseases: A 13-year study. Journal of clinical periodontology. 2021. PMID 33386631. DOI: 10.1111/jcpe.13418. Access: abstract_metadata. Study: PMID33386631.

9. Akhi R, Lavrinienko A, Hakula M et al. Oral Microbiota Linking Humoral Response, Periodontitis and Atherosclerosis. Journal of clinical periodontology. 2025. PMID 40796367. DOI: 10.1111/jcpe.70001. Access: abstract_metadata. Study: PMID40796367.

10. Payne JB, Golub LM, Stoner JA et al. The effect of subantimicrobial-dose-doxycycline periodontal therapy on serum biomarkers of systemic inflammation: a randomized, double-masked, placebo-controlled clinical trial. Journal of the American Dental Association (1939). 2011. PMID 21357860. DOI: 10.14219/jada.archive.2011.0165. Access: abstract_metadata. Study: PMID21357860.

11. Miyamoto T, Yumoto H, Takahashi Y et al. Pathogen-accelerated atherosclerosis occurs early after exposure and can be prevented via immunization. Infection and immunity. 2006. PMID 16428788. DOI: 10.1128/IAI.74.2.1376-1380.2006. Access: abstract_metadata. Study: PMID16428788.

12. Laky M, Anscheringer I, Wolschner L et al. Periodontal treatment limits platelet activation in patients with periodontitis-a controlled-randomized intervention trial. Journal of clinical periodontology. 2018. PMID 29972709. DOI: 10.1111/jcpe.12980. Access: abstract_metadata. Study: PMID29972709.

13. Xu F, Lu B et al. Prospective association of periodontal disease with cardiovascular and all-cause mortality: NHANES III follow-up study. Atherosclerosis. 2011. PMID 21831372. DOI: 10.1016/j.atherosclerosis.2011.07.091. Access: abstract_metadata. Study: NHANES-III.

14. Winning L, Patterson CC, Cullen KM et al. The association between subgingival periodontal pathogens and systemic inflammation. Journal of clinical periodontology. 2015. PMID 26309048. DOI: 10.1111/jcpe.12450. Access: abstract_metadata. Study: PMID26309048.

15. Ma C, Wu M, Gao J et al. Periodontitis and stroke: A Mendelian randomization study. Brain and behavior. 2023. PMID 36621868. DOI: 10.1002/brb3.2888. Access: abstract_metadata. Study: PMID36621868.

16. Wu Q, Li Z, Zhang Y et al. Cyclic di-AMP Rescues Porphyromonas gingivalis-Aggravated Atherosclerosis. Journal of dental research. 2023. PMID 37029659. DOI: 10.1177/00220345231162344. Access: abstract_metadata. Study: PMID37029659.

17. Dominy SS, Lynch C, Ermini F et al. Porphyromonas gingivalis in Alzheimer's disease brains: Evidence for disease causation and treatment with small-molecule inhibitors. Science advances. 2019. PMID 30746447. DOI: 10.1126/sciadv.aau3333. Access: full_text_partial. Study: PMID30746447.

18. Ide M, Harris M, Stevens A et al. Periodontitis and Cognitive Decline in Alzheimer's Disease. PloS one. 2016. PMID 26963387. DOI: 10.1371/journal.pone.0151081. Access: abstract_metadata. Study: PMID26963387.

19. Rubinstein T, Brickman AM, Cheng B et al. Periodontitis and brain magnetic resonance imaging markers of Alzheimer's disease and cognitive aging. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2024. PMID 38278517. DOI: 10.1002/alz.13683. Access: abstract_metadata. Study: WHICAP.

20. Lu J, Zhang S, Huang Y et al. Periodontitis-related salivary microbiota aggravates Alzheimer's disease via gut-brain axis crosstalk. Gut microbes. 2022. PMID 36175166. DOI: 10.1080/19490976.2022.2126272. Access: abstract_metadata. Study: PMID36175166.

21. Ma X, Shin YJ, Yoo JW et al. Extracellular vesicles derived from Porphyromonas gingivalis induce trigeminal nerve-mediated cognitive impairment. Journal of advanced research. 2023. PMID 36796586. DOI: 10.1016/j.jare.2023.02.006. Access: abstract_metadata. Study: PMID36796586.

22. Lei S, Li J, Yu J et al. Porphyromonas gingivalis bacteremia increases the permeability of the blood-brain barrier via the Mfsd2a/Caveolin-1 mediated transcytosis pathway. International journal of oral science. 2023. PMID 36631446. DOI: 10.1038/s41368-022-00215-y. Access: abstract_metadata. Study: PMID36631446.

23. Kong L, Li J, Gao L et al. Periodontitis-induced neuroinflammation triggers IFITM3-Aβ axis to cause alzheimer's disease-like pathology and cognitive decline. Alzheimer's research & therapy. 2025. PMID 40684176. DOI: 10.1186/s13195-025-01818-3. Access: abstract_metadata. Study: PMID40684176.

24. Kantarci A, Tognoni CM, Yaghmoor W et al. Microglial response to experimental periodontitis in a murine model of Alzheimer's disease. Scientific reports. 2020. PMID 33122702. DOI: 10.1038/s41598-020-75517-4. Access: abstract_metadata. Study: PMID33122702.

25. Beydoun MA, Beydoun HA, Hossain S et al. Clinical and Bacterial Markers of Periodontitis and Their Association with Incident All-Cause and Alzheimer's Disease Dementia in a Large National Survey. Journal of Alzheimer's disease : JAD. 2020. PMID 32280099. DOI: 10.3233/JAD-200064. Access: abstract_metadata. Study: NHANES-III.

26. Gong T, Chen Q, Mao H et al. Outer membrane vesicles of Porphyromonas gingivalis trigger NLRP3 inflammasome and induce neuroinflammation, tau phosphorylation, and memory dysfunction in mice. Frontiers in cellular and infection microbiology. 2022. PMID 36017373. DOI: 10.3389/fcimb.2022.925435. Access: abstract_metadata. Study: PMID36017373.

27. Jiang M, Zhang X, Yan X et al. GSK3β is involved in promoting Alzheimer's disease pathologies following chronic systemic exposure to Porphyromonas gingivalis lipopolysaccharide in amyloid precursor proteinNL-F/NL-F knock-in mice. Brain, behavior, and immunity. 2021. PMID 34391814. DOI: 10.1016/j.bbi.2021.08.213. Access: abstract_metadata. Study: PMID34391814.

28. Cirstea MS, Kliger D, MacLellan AD et al. The Oral and Fecal Microbiota in a Canadian Cohort of Alzheimer's Disease. Journal of Alzheimer's disease : JAD. 2022. PMID 35275538. DOI: 10.3233/JAD-215520. Access: abstract_metadata. Study: PMID35275538.

29. Rolim Tde S, Fabri GM, Nitrini R et al. Evaluation of patients with Alzheimer's disease before and after dental treatment. Arquivos de neuro-psiquiatria. 2014. PMID 25517641. DOI: 10.1590/0004-282X20140140. Access: abstract_metadata. Study: PMID25517641.

30. Kamer AR, Pirraglia E, Tsui W et al. Periodontal disease associates with higher brain amyloid load in normal elderly. Neurobiology of aging. 2015. PMID 25491073. DOI: 10.1016/j.neurobiolaging.2014.10.038. Access: abstract_metadata. Study: PMID25491073.

31. Bahar B, Kanagasingam S, Tambuwala MM et al. Porphyromonas gingivalis (W83) Infection Induces Alzheimer's Disease-Like Pathophysiology in Obese and Diabetic Mice. Journal of Alzheimer's disease : JAD. 2021. PMID 34151813. DOI: 10.3233/JAD-210465. Access: abstract_metadata. Study: PMID34151813.

32. Poole S, Singhrao SK, Chukkapalli S et al. Active invasion of Porphyromonas gingivalis and infection-induced complement activation in ApoE-/- mice brains. Journal of Alzheimer's disease : JAD. 2015. PMID 25061055. DOI: 10.3233/JAD-140315. Access: abstract_metadata. Study: PMID25061055.

33. Ward A, Park H, Cheng B et al. Cerebrospinal fluid microbiome revisited: no evidence of resident bacteria in archived samples. Microbiology spectrum. 2025. PMID 40960941. DOI: 10.1128/spectrum.01308-25. Access: abstract_metadata. Study: PMID40960941.

34. Chi L, Cheng X, Lin L et al. Porphyromonas gingivalis-Induced Cognitive Impairment Is Associated With Gut Dysbiosis, Neuroinflammation, and Glymphatic Dysfunction. Frontiers in cellular and infection microbiology. 2021. PMID 34926316. DOI: 10.3389/fcimb.2021.755925. Access: abstract_metadata. Study: PMID34926316.

35. Schwahn C, Frenzel S, Holtfreter B et al. Effect of periodontal treatment on preclinical Alzheimer's disease-Results of a trial emulation approach. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2022. PMID 34050719. DOI: 10.1002/alz.12378. Access: abstract_metadata. Study: PMID34050719.

36. Sparks Stein P, Steffen MJ, Smith C et al. Serum antibodies to periodontal pathogens are a risk factor for Alzheimer's disease. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2012. PMID 22546352. DOI: 10.1016/j.jalz.2011.04.006. Access: abstract_metadata. Study: PMID22546352.

37. Morozumi T, Yashima A, Gomi K et al. Increased systemic levels of inflammatory mediators following one-stage full-mouth scaling and root planing. Journal of periodontal research. 2018. PMID 29603738. DOI: 10.1111/jre.12543. Access: abstract_metadata. Study: PMID29603738.

38. Ghizoni JS, Taveira LA, Garlet GP et al. Increased levels of Porphyromonas gingivalis are associated with ischemic and hemorrhagic cerebrovascular disease in humans: an in vivo study. Journal of applied oral science : revista FOB. 2012. PMID 22437687. DOI: 10.1590/s1678-77572012000100019. Access: abstract_metadata. Study: PMID22437687.

39. Pereira RB, Vasquez EC, Stefanon I et al. Oral P. gingivalis infection alters the vascular reactivity in healthy and spontaneously atherosclerotic mice. Lipids in health and disease. 2011. PMID 21586133. DOI: 10.1186/1476-511X-10-80. Access: abstract_metadata. Study: PMID21586133.

40. Kobayashi N, Suzuki J, Ogawa M et al. Porphyromonas gingivalis promotes neointimal formation after arterial injury through toll-like receptor 2 signaling. Heart and vessels. 2014. PMID 24002697. DOI: 10.1007/s00380-013-0405-3. Access: abstract_metadata. Study: PMID24002697.

41. Yamaguchi Y, Kurita-Ochiai T, Kobayashi R et al. Activation of the NLRP3 inflammasome in Porphyromonas gingivalis-accelerated atherosclerosis. Pathogens and disease. 2015. PMID 25663345. DOI: 10.1093/femspd/ftv011. Access: abstract_metadata. Study: PMID25663345.

42. Gibson FC, Hong C, Chou HH et al. Innate immune recognition of invasive bacteria accelerates atherosclerosis in apolipoprotein E-deficient mice. Circulation. 2004. PMID 15123526. DOI: 10.1161/01.CIR.0000129769.17895.F0. Access: abstract_metadata. Study: PMID15123526.

43. Xu S, Zheng C, Huang J et al. Porphyromonas gingivalis Induces Endothelial Dysfunction Through Sirt3-Dependent CypD Acetylation. Journal of periodontal research. 2025. PMID 40344434. DOI: 10.1111/jre.13416. Access: abstract_metadata. Study: PMID40344434.

44. Farrugia C, Stafford GP, Potempa J et al. Mechanisms of vascular damage by systemic dissemination of the oral pathogen Porphyromonas gingivalis. The FEBS journal. 2021. PMID 32681704. DOI: 10.1111/febs.15486. Access: abstract_metadata. Study: PMID32681704.

45. Yuan H, Zelkha S, Burkatovskaya M et al. Pivotal role of NOD2 in inflammatory processes affecting atherosclerosis and periodontal bone loss. Proceedings of the National Academy of Sciences of the United States of America. 2013. PMID 24324141. DOI: 10.1073/pnas.1320862110. Access: abstract_metadata. Study: PMID24324141.

46. Orlandi M, Masi S, Lucenteforte E et al. Periodontitis treatment and progression of carotid intima-media thickness: a randomized trial. European heart journal. 2025. PMID 40827724. DOI: 10.1093/eurheartj/ehaf555. Access: abstract_metadata. Study: PMID40827724.

47. Pikel K, Logue L, Verkuilen H et al. Determinants of post-stroke cognitive impairment in patients with periodontal disease. Journal of stroke and cerebrovascular diseases : the official journal of National Stroke Association. 2025. PMID 40273963. DOI: 10.1016/j.jstrokecerebrovasdis.2025.108327. Access: abstract_metadata. Study: PREMIERS.

48. Tonetti MS, D'Aiuto F, Nibali L et al. Treatment of periodontitis and endothelial function. The New England journal of medicine. 2007. PMID 17329698. DOI: 10.1056/NEJMoa063186. Access: full_text_partial. Study: PMID17329698.

49. Zhou M, Dong J, Zha L et al. Causal Association between Periodontal Diseases and Cardiovascular Diseases. Genes. 2021. PMID 35052354. DOI: 10.3390/genes13010013. Access: abstract_metadata. Study: PMID35052354.

50. Huang J, Deng A, Bai Y et al. Alzheimer's disease and oral manifestations: a bi-directional Mendelian randomization study. Frontiers in neurology. 2024. PMID 38817545. DOI: 10.3389/fneur.2024.1391625. Access: abstract_metadata. Study: PMID38817545.

51. Schaefer AS, Bochenek G, Jochens A et al. Genetic evidence for PLASMINOGEN as a shared genetic risk factor of coronary artery disease and periodontitis. Circulation. Cardiovascular genetics. 2015. PMID 25466412. DOI: 10.1161/CIRCGENETICS.114.000554. Access: abstract_metadata. Study: PMID25466412.

52. Hu C, Li H, Huang L et al. Periodontal disease and risk of Alzheimer's disease: A two-sample Mendelian randomization. Brain and behavior. 2024. PMID 38648391. DOI: 10.1002/brb3.3486. Access: full_text_partial. Study: PMID38648391.

53. Beydoun MA, Beydoun HA, Hedges DW et al. Infection burden, periodontal pathogens, and their interactive association with incident all-cause and Alzheimer's disease dementia in a large national survey. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2024. PMID 39115027. DOI: 10.1002/alz.14141. Access: abstract_metadata. Study: NHANES-III.

54. Zhu H, Huang C, Luo Z et al. Porphyromonas gingivalis Induces Disturbance of Kynurenine Metabolism Through the Gut-Brain Axis: Implications for Alzheimer's Disease. Journal of dental research. 2025. PMID 39905278. DOI: 10.1177/00220345241303141. Access: abstract_metadata. Study: PMID39905278.

55. Demmer RT, Norby FL, Lakshminarayan K et al. Periodontal disease and incident dementia: The Atherosclerosis Risk in Communities Study (ARIC). Neurology. 2020. PMID 32727837. DOI: 10.1212/WNL.0000000000010312. Access: abstract_metadata. Study: ARIC.

56. Holmer J, Eriksdotter M, Häbel H et al. Periodontal conditions and incident dementia: A nationwide Swedish cohort study. Journal of periodontology. 2022. PMID 35099831. DOI: 10.1002/JPER.21-0518. Access: abstract_metadata. Study: PMID35099831.

57. Naorungroj S, Schoenbach VJ, Wruck L et al. Tooth loss, periodontal disease, and cognitive decline in the Atherosclerosis Risk in Communities (ARIC) study. Community dentistry and oral epidemiology. 2015. PMID 25363061. DOI: 10.1111/cdoe.12128. Access: abstract_metadata. Study: ARIC.

58. Stewart R, Stenman U, Hakeberg M et al. Associations between oral health and risk of dementia in a 37-year follow-up study: the prospective population study of women in Gothenburg. Journal of the American Geriatrics Society. 2015. PMID 25597561. DOI: 10.1111/jgs.13194. Access: abstract_metadata. Study: PMID25597561.

59. Noble JM, Scarmeas N, Celenti RS et al. Serum IgG antibody levels to periodontal microbiota are associated with incident Alzheimer disease. PloS one. 2014. PMID 25522313. DOI: 10.1371/journal.pone.0114959. Access: abstract_metadata. Study: WHICAP.

60. Sen S, Curtis J, Hicklin D et al. Periodontal Disease Treatment After Stroke or Transient Ischemic Attack: The PREMIERS Study, a Randomized Clinical Trial. Stroke. 2023. PMID 37548008. DOI: 10.1161/STROKEAHA.122.042047. Access: full_text_partial. Study: PREMIERS.

61. Jin T, Lin J, Zhang P et al. Exploring the shared genetic architecture between periodontitis and cardiovascular disease. BDJ open. 2026. PMID 41916952. DOI: 10.1038/s41405-026-00421-4. Access: full_text_partial. Study: PMID41916952.

62. Schaefer AS, Richter GM, Groessner-Schreiber B et al. Identification of a shared genetic susceptibility locus for coronary heart disease and periodontitis. PLoS genetics. 2009. PMID 19214202. DOI: 10.1371/journal.pgen.1000378. Access: abstract_metadata. Study: PMID19214202.

63. Sanghvi MM, Ramírez J, Chadalavada S et al. The Association Between Periodontal Disease and Cardiovascular Disease: Insights From Imaging, Observational, and Genetic Data. JACC. Advances. 2024. PMID 39290820. DOI: 10.1016/j.jacadv.2024.101241. Access: full_text_partial. Study: PMID39290820.

64. Offenbacher S, Beck JD, Moss K et al. Results from the Periodontitis and Vascular Events (PAVE) Study: a pilot multicentered, randomized, controlled trial to study effects of periodontal therapy in a secondary prevention model of cardiovascular disease. Journal of periodontology. 2009. PMID 19186958. DOI: 10.1902/jop.2009.080007. Access: full_text_partial. Study: PMID19186958.

65. Huang C, Cai Q, Feng Z et al. Porphyromonas gingivalis disrupts hippocampal circadian clock via PI3K/AKT pathway, exacerbating Alzheimer-like pathology. Alzheimer's & dementia : the journal of the Alzheimer's Association. 2026. PMID 42271186. DOI: 10.1002/alz.71543. Access: full_text_partial. Study: PMID42271186.

66. Ciccotosto GD, Mohammed AI, Paolini R et al. Chronic Oral Inoculation of Porphyromonas gingivalis and Treponema denticola Induce Different Brain Pathologies in a Mouse Model of Alzheimer Disease. The Journal of infectious diseases. 2024. PMID 39255392. DOI: 10.1093/infdis/jiae286. Access: abstract_metadata. Study: PMID39255392.

67. Aravindraja C, Sakthivel R, Liu X et al. Intracerebral but Not Peripheral Infection of Live Porphyromonas gingivalis Exacerbates Alzheimer's Disease Like Amyloid Pathology in APP-TgCRND8 Mice. International journal of molecular sciences. 2022. PMID 35328748. DOI: 10.3390/ijms23063328. Access: full_text_partial. Study: PMID35328748.

68. Lobo MG, Schmidt MM, Lopes RD et al. Treating periodontal disease in patients with myocardial infarction: A randomized clinical trial. European journal of internal medicine. 2020. PMID 31810741. DOI: 10.1016/j.ejim.2019.08.012. Access: abstract_metadata. Study: PMID31810741.

69. Saffi MAL, Rabelo-Silva ER, Polanczyk CA et al. Periodontal therapy and endothelial function in coronary artery disease: A randomized controlled trial. Oral diseases. 2018. PMID 29873864. DOI: 10.1111/odi.12909. Access: abstract_metadata. Study: PMID29873864.

Source-specific evidence cards

[1] Periodontitis exacerbates atherosclerosis through Fusobacterium nucleatum-promoted hepatic glycolysis and lipogenesis.

Shows: Ligature plus patient plaque and oral F. nucleatum accelerated atheroma in high-cholesterol atheroprone mice; hepatic lipid synthesis was linked to glycolytic signaling.

Does not show: Forced exposure, ApoE/diet context and in-vitro pathway inhibition do not establish pathogen-specific human ASCVD mediation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[2] Oral Pathobionts Aggravate Myocardial Infarction Through Mobilization of B2 Cells.

Shows: Ligature/plaque experiments, gnotobiotic mice, host knockouts and adhesion mutants linked oral pathobionts to B2-cell mobilization and worse experimental infarction; the human arm was associational.

Does not show: This tests aggravation after induced MI, not spontaneous human coronary initiation or proof that Pg alone is necessary.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

[3] Better oral hygiene is associated with lower risk of stroke.

Shows: In 206602 Korean participants, frequent brushing was associated with lower incident stroke during median 10.4-year follow-up.

Does not show: Dental-care behavior, socioeconomic factors and health behaviors confound mediation; stroke subtypes overlap and the result is not a randomized pathogen intervention.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[4] Causal association between periodontitis and hypertension: evidence from Mendelian randomization and a randomized controlled trial of non-surgical periodontal therapy.

Shows: Periodontitis-linked genetic instruments and a 101-person hypertensive RCT supported a periodontal-treatment effect on two-month ambulatory BP, with cytokine and T-cell changes.

Does not show: Short-term BP/endothelial effects are narrower than atherosclerotic events; host-instrument pleiotropy and nonspecific treatment do not isolate Pg.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Methods:paragraph1, 1:Methods:paragraph2, 1:Methods:paragraph4, 1:Methods:paragraph5, 1:Methods:paragraph11. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC5.

[5] Periodontitis in patients with coronary artery disease: an 8-year follow-up.

Shows: An eight-year CAD follow-up found no significant association between baseline periodontal severity and later MI, revascularization or CAD death.

Does not show: A modest established-CAD cohort can miss effects; the null does not prove no periodontal contribution in other stages.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[6] Periodontal Disease, Atrial Fibrillation and Stroke.

Shows: ARIC periodontal severity and dental-care use were associated with incident AF; statistical mediation suggested AF might explain part of the periodontal-stroke association.

Does not show: Shared ARIC reports are not independent causal lines, and observational mediation does not identify a microbial cause.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[7] Periodontitis induces endothelial dysfunction in mice.

Shows: Four-week Pg/S. gordonii oral lavage in ApoE-deficient mice produced bone loss, hepatic inflammatory expression and impaired aortic acetylcholine relaxation without detected caecal-community differences.

Does not show: Mixed-pathogen exposure and atheroprone genotype do not isolate a human Pg pathway; endothelial function is not a hard clinical event.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC5.

[8] Periodontitis is associated with cardiovascular diseases: A 13-year study.

Shows: Among 1850 Thai adults followed 13 years, severe periodontitis predicted CHD, but not total cardiovascular events or stroke, after adjustment.

Does not show: The CHD interval was wide and endpoints differed; association cannot establish pathogen mediation or generalize to every vascular event.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[9] Oral Microbiota Linking Humoral Response, Periodontitis and Atherosclerosis.

Shows: Northern Finland cohort data associated F. alocis, periodontal pockets, phosphorylcholine antibody responses and carotid thickness.

Does not show: Cross-sectional microbial/antibody mediation is not causal mediation and is an alternative to a Pg-specific explanation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3, SC5.

[10] The effect of subantimicrobial-dose-doxycycline periodontal therapy on serum biomarkers of systemic inflammation: a randomized, double-masked, placebo-controlled clinical trial.

Shows: Two-year randomized subantimicrobial doxycycline lowered hs-CRP and MMP-9 in 128 postmenopausal periodontal-maintenance participants, without an overall lipid effect.

Does not show: Systemic doxycycline is a host-modulating drug, so the effect cannot be assigned to elimination of oral microbes; no CAD event benefit was tested.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[11] Pathogen-accelerated atherosclerosis occurs early after exposure and can be prevented via immunization.

Shows: Pg challenge increased early atheroma, macrophages and local innate markers in ApoE-knockout mice; prior immunization prevented this local response without elevated systemic mediators.

Does not show: Artificial challenge and prophylactic immunization demonstrate model capacity, not spontaneous human disease or an obligatory systemic-CRP pathway.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC5.

[12] Periodontal treatment limits platelet activation in patients with periodontitis-a controlled-randomized intervention trial.

Shows: A 52-person randomized study found periodontal treatment prevented worsening of platelet activation over three months, with differences in turnover and collagen responses.

Does not show: The treated arm was largely unchanged rather than normalized; platelet assays do not demonstrate prevention of MI or stroke.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC5.

[13] Prospective association of periodontal disease with cardiovascular and all-cause mortality: NHANES III follow-up study.

Shows: NHANES III mortality follow-up associated severe periodontitis with cardiovascular/all-cause death in men aged 30–64, but not older men or women.

Does not show: Subgroup heterogeneity, residual health-behavior confounding and shared NHANES material limit causal and independence claims.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[14] The association between subgingival periodontal pathogens and systemic inflammation.

Shows: In 518 older dentate Northern Irish men, detectable subgingival Pg was associated with higher CRP after selected adjustments.

Does not show: The cross-sectional finding cannot establish direction or exclude all obesity, smoking and host-susceptibility pathways.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC5.

[15] Periodontitis and stroke: A Mendelian randomization study.

Shows: Genetic proxies for chronic/aggressive periodontitis were null for overall ischemic stroke; a nominal chronic-periodontitis cardioembolic association contrasted with null large-artery and small-vessel results.

Does not show: Limited periodontal instruments, multiple subtype tests and host pleiotropy prevent interpreting the isolated nominal result as microbial mediation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

[16] Cyclic di-AMP Rescues Porphyromonas gingivalis-Aggravated Atherosclerosis.

Shows: Cyclic di-AMP prophylaxis reduced periodontal bone loss, lipids and atheroma in Pg-challenged, ligature-treated high-fat ApoE-knockout mice.

Does not show: A broad STING/trained-immunity adjuvant is not pathogen-specific rescue or proof of necessary human Pg infection.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC5.

[17] Porphyromonas gingivalis in Alzheimer's disease brains: Evidence for disease causation and treatment with small-molecule inhibitors.

Shows: Human gingipain/tau associations accompanied mouse oral-infection brain bacterial/amyloid findings and inhibitor protection in neurotoxicity paradigms.

Does not show: Postmortem detection cannot establish antecedent oral-to-brain transit in humans; the routes, outcomes and controls of distinct mouse experiments must remain separate. Original sections show PCR positivity also in five of six nondemented controls, no healthy CSF comparison in the ten-AD-person pilot, direct hippocampal gingipain injection in one toxicity experiment, and oral inhibitor effects coupled to reduced pathogen burden. Gad67-positive interneuron counts are not a complete unbiased neuron-survival audit.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 0:INTRODUCTION:paragraph13, 0:INTRODUCTION:paragraph14, 0:INTRODUCTION:paragraph26, 0:INTRODUCTION:paragraph27, 0:INTRODUCTION:paragraph30, 0:INTRODUCTION:paragraph31. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[18] Periodontitis and Cognitive Decline in Alzheimer's Disease.

Shows: A six-month observational cohort of 60 mild/moderate AD participants linked baseline periodontitis to faster ADAS-cog decline and increased inflammatory tone.

Does not show: This is established-AD progression, not initiation, and frailty/caregiving or disease severity can affect oral health despite baseline adjustment.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4, SC5.

[19] Periodontitis and brain magnetic resonance imaging markers of Alzheimer's disease and cognitive aging.

Shows: WHICAP oral-health data in 468 people associated periodontal/tooth and microbial-response features with brain MRI atrophy and vascular markers.

Does not show: The analysis does not show AD causality; bacterial-specific associations differed and MRI endpoints are not a randomized clinical trajectory.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[20] Periodontitis-related salivary microbiota aggravates Alzheimer's disease via gut-brain axis crosstalk.

Shows: Two-month gavage of periodontal-patient salivary communities worsened cognition, amyloid and inflammation in APP/PS1 transgenic mice compared with healthy saliva.

Does not show: Crude transfer does not identify a necessary Pg strain, separate gut/community/inflammation mediation, or establish endogenous human AD.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[21] Extracellular vesicles derived from Porphyromonas gingivalis induce trigeminal nerve-mediated cognitive impairment.

Shows: Gingival exposure to Pg or its EVs impaired mouse memory, unlike oral gavage; labeled-vesicle transport to the ipsilateral trigeminal ganglion was reduced by neurectomy.

Does not show: Label movement is not intact replicating-organism transit; neuronal markers are not stereological loss, and route-specific inoculation bounds generalization.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[22] Porphyromonas gingivalis bacteremia increases the permeability of the blood-brain barrier via the Mfsd2a/Caveolin-1 mediated transcytosis pathway.

Shows: Repeated tail-vein bacteremia increased rat BBB leakage; cellular Cav-1 reduction and Mfsd2a overexpression reversed permeability responses.

Does not show: Forced bloodstream delivery and cell-level rescue do not establish an oral initiating route or rescue of human dementia.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[23] Periodontitis-induced neuroinflammation triggers IFITM3-Aβ axis to cause alzheimer's disease-like pathology and cognitive decline.

Shows: Pg-induced periodontitis impaired cognition in wild-type mice and worsened APP/PS1 decline, with IFN-beta, glial IFITM3 and amyloid changes.

Does not show: The abstract does not report IFITM3-specific loss/rescue; correlated pathway activation does not isolate the mediator or establish human initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[24] Microglial response to experimental periodontitis in a murine model of Alzheimer's disease.

Shows: Ligature periodontitis increased insoluble amyloid in 5xFAD mice, but glial and cytokine effects differed from wild type, including reduced plaque-associated microglia.

Does not show: State-specific responses oppose a simple uniform activation account; amyloid/glial changes do not demonstrate cognitive rescue or a Pg-specific cause.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4, SC5.

[25] Clinical and Bacterial Markers of Periodontitis and Their Association with Incident All-Cause and Alzheimer's Disease Dementia in a Large National Survey.

Shows: NHANES III linked periodontal pockets and multi-pathogen IgG composites to incident dementia/AD mortality over long follow-up, particularly in older adults.

Does not show: Composite antibodies and administrative outcomes do not identify active CNS infection; multiple analyses of the same NHANES cohort are one study family.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[26] Outer membrane vesicles of Porphyromonas gingivalis trigger NLRP3 inflammasome and induce neuroinflammation, tau phosphorylation, and memory dysfunction in mice.

Shows: Eight-week oral OMV dosing impaired aged-mouse memory and increased barrier, inflammatory and phospho-tau abnormalities; MCC950 attenuated cellular inflammasome/conditioned-medium effects.

Does not show: The NLRP3 inhibition was in vitro, not a demonstrated in-vivo cognitive rescue; forced vesicle dose and substrate assays do not establish human AD.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[27] GSK3β is involved in promoting Alzheimer's disease pathologies following chronic systemic exposure to Porphyromonas gingivalis lipopolysaccharide in amyloid precursor proteinNL-F/NL-F knock-in mice.

Shows: Daily intraperitoneal Pg LPS worsened memory, synaptic proteins and tau phosphorylation in APP knock-in mice without changing amyloid; cellular GSK3/TNF interventions supported a pathway.

Does not show: A systemic endotoxin regimen is not oral infection, and cell-level inhibition is not organismal mediator-specific rescue.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[28] The Oral and Fecal Microbiota in a Canadian Cohort of Alzheimer's Disease.

Shows: A Canadian case-control cohort found altered oral microbiota and greater Pg prevalence in AD, but no significant oral or gut microbial association with cognition.

Does not show: A 45-case/54-control cross-section cannot establish direction; the lack of a cognition signature limits a simple abundance-to-progression inference.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[29] Evaluation of patients with Alzheimer's disease before and after dental treatment.

Shows: Before/after dental care in 29 mild-AD patients improved pain, periodontal indices, quality of life and reported functional impairment.

Does not show: An uncontrolled multicomponent intervention with symptomatic dental benefits cannot establish disease modification or microbial mediation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4, SC5.

[30] Periodontal disease associates with higher brain amyloid load in normal elderly.

Shows: Clinical attachment loss was associated with greater amyloid PET uptake in 38 cognitively normal older people after selected adjustment.

Does not show: Small observational amyloid imaging does not establish a causal infectious origin or clinical trajectory change.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[31] Porphyromonas gingivalis (W83) Infection Induces Alzheimer's Disease-Like Pathophysiology in Obese and Diabetic Mice.

Shows: Oral Pg exposure in obese diabetic db/db mice produced glial/tau immunoreactivity and signaling changes, but silver staining detected neither amyloid plaques nor neurofibrillary tangles.

Does not show: This is a constrained comorbidity model; transcriptional pathway activation and immunoreactivity cannot be equated with full AD or proven neuronal loss.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[32] Active invasion of Porphyromonas gingivalis and infection-induced complement activation in ApoE-/- mice brains.

Shows: Oral infection in ApoE-deficient mice yielded Pg brain DNA at 12/24 weeks and neuronal complement-fragment deposition in a subset.

Does not show: Bacterial DNA and complement staining do not prove viable CNS replication, generalized neuronal death or spontaneous human initiation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[33] Cerebrospinal fluid microbiome revisited: no evidence of resident bacteria in archived samples.

Shows: Spike-in and meningitis-positive controls worked in 176 archived CSF specimens, while ordinary samples yielded negligible bacterial reads consistent with contamination/error.

Does not show: This limits a resident-CSF microbiome claim; it does not exclude transient infection, focal brain organisms or every oral inflammatory pathway.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

[34] Porphyromonas gingivalis-Induced Cognitive Impairment Is Associated With Gut Dysbiosis, Neuroinflammation, and Glymphatic Dysfunction.

Shows: A month of oral Pg administration impaired mouse cognition with gut-composition, glymphatic, inflammatory and neuronal changes.

Does not show: Co-occurring routes and pathways were not separately rescued in the abstract; alpha diversity was unchanged and an induced model is not human mediation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[35] Effect of periodontal treatment on preclinical Alzheimer's disease-Results of a trial emulation approach.

Shows: A trial-emulation comparison of 177 treated and 409 untreated cohort members associated periodontal care with less AD-pattern brain atrophy after median 7.3 years.

Does not show: Treatment was not randomized, the groups came from different cohorts, and MRI atrophy is neither AD diagnosis nor a clinical dementia endpoint.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

[36] Serum antibodies to periodontal pathogens are a risk factor for Alzheimer's disease.

Shows: Kentucky longitudinal samples showed elevated antibodies to F. nucleatum and P. intermedia before AD conversion, after selected adjustment.

Does not show: The principal positive species were not Pg; antibody antecedence to symptoms does not establish antecedence to preclinical brain disease or causality.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[37] Increased systemic levels of inflammatory mediators following one-stage full-mouth scaling and root planing.

Shows: Full-mouth scaling in 31 patients improved periodontal/bacterial measures by six weeks but produced an acute next-day systemic inflammatory response.

Does not show: The uncontrolled before/after design and procedure-induced acute inflammation do not establish chronic pathogenic mediation or vascular-event prevention.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC5.

[38] Increased levels of Porphyromonas gingivalis are associated with ischemic and hemorrhagic cerebrovascular disease in humans: an in vivo study.

Shows: A small stroke case-control study found worse periodontal health and more Pg-contaminated sites in stroke patients.

Does not show: The title's in-vivo wording does not make this an intervention; post-stroke sampling and a very wide association interval permit reverse causation/confounding.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[39] Oral P. gingivalis infection alters the vascular reactivity in healthy and spontaneously atherosclerotic mice.

Shows: Oral Pg increased phenylephrine vasoconstriction in healthy and ApoE-deficient mice, while acetylcholine and nitroprusside relaxation remained similar.

Does not show: The receptor-specific response is not uniform endothelial failure and does not demonstrate human clinical-event causality.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

[40] Porphyromonas gingivalis promotes neointimal formation after arterial injury through toll-like receptor 2 signaling.

Shows: Subcutaneous Pg after femoral injury increased neointima and MCP1 in wild-type mice; TLR2 deficiency prevented this response despite antibody induction.

Does not show: Arterial injury and forced systemic exposure establish a recipient-model dependency, not spontaneous oral-pathogen-mediated human ASCVD.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

[41] Activation of the NLRP3 inflammasome in Porphyromonas gingivalis-accelerated atherosclerosis.

Shows: Oral wild-type Pg increased plaque, bone loss and inflammatory signals in hyperlipidemic Apoe-shl mice relative to gingipain-null and FimA-deficient strains.

Does not show: Virulence mutations also change dissemination and burden; NLRP3 expression was measured without a reported NLRP3 necessity manipulation in the abstract.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[42] Innate immune recognition of invasive bacteria accelerates atherosclerosis in apolipoprotein E-deficient mice.

Shows: Oral wild-type Pg but not FimA-mutant Pg accelerated periodontal and atherosclerotic pathology in ApoE-null mice; immunization prevented the phenotype.

Does not show: Both strains were detected by PCR in blood and aorta; coupled changes in periodontal disease and virulence do not isolate invasion as the sole mediator.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[43] Porphyromonas gingivalis Induces Endothelial Dysfunction Through Sirt3-Dependent CypD Acetylation.

Shows: Pg impaired endothelial integrity and vasorelaxation; Sirt3 activation reversed mitochondrial and vascular defects in cells and orally exposed mice.

Does not show: Honokiol activation is not equivalent to a selective genetic rescue, and vascular physiology does not establish human clinical event prevention.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

[44] Mechanisms of vascular damage by systemic dissemination of the oral pathogen Porphyromonas gingivalis.

Shows: Wild-type Pg caused endothelial junction loss and leakage in human cells and systemically exposed zebrafish, unlike gingipain-deficient bacteria.

Does not show: Cell culture and forced systemic zebrafish infection do not reproduce endogenous human oral-to-vascular mediation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2.

[45] Pivotal role of NOD2 in inflammatory processes affecting atherosclerosis and periodontal bone loss.

Shows: NOD2 deficiency worsened oral-Pg inflammation, cholesterol and atheroma in ApoE mice; MDP stimulation reduced these phenotypes.

Does not show: NOD2 signaling was protective here; systemic MDP also affected weight and metabolic inflammation. The linked correction concerns an author name, not a new study.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMIDnot supplied. Subclaims: SC2, SC5.

[46] Periodontitis treatment and progression of carotid intima-media thickness: a randomized trial.

Shows: In 135 otherwise healthy periodontal patients, randomized intensive treatment improved FMD and reduced carotid intima-media thickness at 24 months; blood pressure and PWV did not substantially differ.

Does not show: The primary outcome was vascular structure, not MI, stroke or death; the PubMed abstract confidence interval is malformed and is not reproduced.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC5.

[47] Determinants of post-stroke cognitive impairment in patients with periodontal disease.

Shows: Cross-sectional baseline analysis of 280 PREMIERS participants associated severe periodontitis with post-stroke cognitive impairment.

Does not show: This report is not a randomized cognitive-treatment result; stroke severity, education and race-related factors remain relevant, and it shares the PREMIERS trial population. Publication-type labeling was overridden to observational baseline analysis.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3, SC4.

[48] Treatment of periodontitis and endothelial function.

Shows: Randomized intensive treatment in 120 periodontal patients transiently worsened FMD and inflammation at 24 hours but improved FMD at 60 and 180 days; no cardiovascular events occurred.

Does not show: Healthy participants and a surrogate endpoint limit clinical ASCVD inference; the parent paper has a correction to tPA variability reporting and a figure day label.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: Original web Methods: enrollment and vascular assessment; linked correction NEJMx180022. No complete-paper or supplement audit claimed. ErratumIn: PMID29898373. Subclaims: SC3, SC5.

[49] Causal Association between Periodontal Diseases and Cardiovascular Diseases.

Shows: Two-sample Mendelian randomization found no supported periodontitis effect on major cardiovascular outcomes across several estimators.

Does not show: Genetic liability instruments are not pathogen-specific eradication experiments and may not capture acquired infection severity or timing.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

[50] Alzheimer's disease and oral manifestations: a bi-directional Mendelian randomization study.

Shows: Bidirectional genetic analyses did not support oral disease liability causing AD; a reverse AD-risk association was reported for oral cancer.

Does not show: This does not isolate Pg infection, and instrument validity and phenotype specificity restrict a universal absence conclusion.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

[51] Genetic evidence for PLASMINOGEN as a shared genetic risk factor of coronary artery disease and periodontitis.

Shows: Candidate-gene analysis linked aggressive periodontitis and CAD susceptibility through shared ANRIL and PLG-related loci.

Does not show: Shared genetic susceptibility can support common causes; it is not Mendelian randomization or a randomized clinical intervention despite its database trial tag. Publication-type labeling was overridden to genetic association.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[52] Periodontal disease and risk of Alzheimer's disease: A two-sample Mendelian randomization.

Shows: Periodontitis-to-AD Mendelian randomization using FinnGen and two AD datasets produced null estimates with reported sensitivity analyses.

Does not show: Host liability instruments and relaxed selection thresholds do not exclude all microbial causal effects, and summary datasets are reused across MR reports. Original Methods use p<1e-5 instruments and a small clinically ascertained AD case count in one outcome; nonsignificant Egger tests do not prove absence of pleiotropy.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:METHOD, 2:RESULTS. No complete-paper or supplement audit claimed. none_found. Subclaims: SC4.

[53] Infection burden, periodontal pathogens, and their interactive association with incident all-cause and Alzheimer's disease dementia in a large national survey.

Shows: NHANES III antibody and infection-burden models associated Pg-related interactions with incident AD over linked follow-up.

Does not show: Antibody exposure, multiple interactions and CMS outcome ascertainment do not prove CNS infection; other NHANES III reports are not independent cohorts.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[54] Porphyromonas gingivalis Induces Disturbance of Kynurenine Metabolism Through the Gut-Brain Axis: Implications for Alzheimer's Disease.

Shows: Oral Pg altered oral and gut communities, barrier measures and AD-like pathology; 3-hydroxykynurenine exposure promoted neuronal apoptosis in model systems.

Does not show: The abstract does not report necessity blockade of the entire oral-to-brain pathway; indirect metabolic mediation remains distinct from direct CNS infection.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[55] Periodontal disease and incident dementia: The Atherosclerosis Risk in Communities Study (ARIC).

Shows: Among 8275 ARIC participants followed for an average 18.4 years, severe periodontitis showed a modest association with incident dementia.

Does not show: Some tooth-loss and cognitive endpoints had intervals spanning no association; confounding and shared ARIC participants preclude an independent causal line.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[56] Periodontal conditions and incident dementia: A nationwide Swedish cohort study.

Shows: A Swedish matched cohort of 7992 deep-pocket patients and 29182 controls found similar dementia incidence over mean 7.6 years and no adjusted association.

Does not show: The exposure is a periodontitis proxy; the wide adjusted interval does not demonstrate equivalence or rule out all pathogen-defined effects.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[57] Tooth loss, periodontal disease, and cognitive decline in the Atherosclerosis Risk in Communities (ARIC) study.

Shows: An ARIC subset showed lower cognitive performance with complete tooth loss but no greater eight-year cognitive decline predicted by periodontal disease, tooth count or edentulism.

Does not show: This shares the ARIC source population and evaluates limited cognitive tests rather than biomarker-defined AD or Pg exposure.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[58] Associations between oral health and risk of dementia in a 37-year follow-up study: the prospective population study of women in Gothenburg.

Shows: In the Gothenburg women's cohort, most adjusted tooth-count comparisons over 37 years were not associated with dementia; one examination remained positive.

Does not show: Tooth count is an indirect exposure and selective positive time points do not establish periodontal microbial mediation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[59] Serum IgG antibody levels to periodontal microbiota are associated with incident Alzheimer disease.

Shows: In a 219-person WHICAP case-cohort, high A. naeslundii IgG predicted higher AD risk and E. nodatum IgG lower risk after adjustment.

Does not show: The reported positive pathogen was not Pg; baseline age differed markedly, antibody response is indirect, and other WHICAP reports reuse the cohort.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC4.

[60] Periodontal Disease Treatment After Stroke or Transient Ischemic Attack: The PREMIERS Study, a Randomized Clinical Trial.

Shows: PREMIERS randomized 280 recent stroke or TIA patients to intensive or standard periodontal care; 12-month death, MI or stroke occurred in 11 versus 17, without significant superiority.

Does not show: Both arms received active care, the phase II trial had few events and broad uncertainty; nonsuperiority is not proof of no clinical effect or pathogen irrelevance. Original Methods show an open-label masked-endpoint trial, standard-arm rescue treatment for periodontal worsening, a low observed event rate relative to planning, baseline diabetes/BMI imbalance, and no missing primary time-to-event data.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: Original web Methods: Design, Treatment, Statistical considerations; Results: Trial population and events. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

[61] Exploring the shared genetic architecture between periodontitis and cardiovascular disease.

Shows: Bidirectional and multivariable MR did not support direct periodontitis-CVD causality, while colocalization and cross-trait analyses identified shared loci and immune-metabolic pathways.

Does not show: Shared variants and pathway annotation do not establish oral infection mediation; phenotype instruments and reused summary data constrain independence.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Materials and methods:paragraph2, 1:Materials and methods:paragraph4, 1:Materials and methods:paragraph5, 1:Materials and methods:paragraph8. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[62] Identification of a shared genetic susceptibility locus for coronary heart disease and periodontitis.

Shows: Candidate-gene analysis identified shared 9p21 susceptibility for aggressive periodontitis and coronary disease, distinct from a nearby diabetes variant.

Does not show: Shared ANRIL-region association does not resolve disease direction or prove pathogen-mediated coronary disease.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[63] The Association Between Periodontal Disease and Cardiovascular Disease: Insights From Imaging, Observational, and Genetic Data.

Shows: In UK Biobank, self-reported periodontal disease was associated with cardiac remodeling and incident CAD; its polygenic score also associated with CAD.

Does not show: The polygenic-score analysis is not a valid pathogen-specific causal instrument; weak liability and pleiotropy prevent treating it as decisive Mendelian randomization. Original Discussion expressly says MR was avoided because of instrument weakness and acknowledges horizontal or indirect pleiotropy.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 2:Discussion:paragraph7, 2:Discussion:paragraph8, 2:Discussion:paragraph10. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

[64] Results from the Periodontitis and Vascular Events (PAVE) Study: a pilot multicentered, randomized, controlled trial to study effects of periodontal therapy in a secondary prevention model of cardiovascular disease.

Shows: The 303-person PAVE pilot improved periodontal measures but showed no intention-to-treat reduction in hsCRP at six months; nearly half of community controls obtained dental care.

Does not show: Post-randomization any-care and obesity subgroup comparisons do not retain the randomized contrast; this feasibility pilot was not powered for clinical cardiovascular efficacy.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: Original web Introduction/Methods: pilot aims and randomization. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC5.

[65] Porphyromonas gingivalis disrupts hippocampal circadian clock via PI3K/AKT pathway, exacerbating Alzheimer-like pathology.

Shows: Six months of oral Pg after antibiotic conditioning caused poorer water-maze performance, neuronal pathology and altered hippocampal rhythms; cultured-glial AKT activation and Bmal1 knockdown interrogated the clock pathway.

Does not show: Original Methods/Results specify twice-weekly exposure for 24 weeks after kanamycin, matched oral handling, and cell-line rather than in-vivo pathway rescue. Cognitive pathology is present; in-vivo clock-mediated necessity remains untested in the examined sections.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:METHODS:paragraph6, 1:METHODS:paragraph7, 1:METHODS:paragraph12, 1:METHODS:paragraph17, 2:RESULTS:paragraph0, 2:RESULTS:paragraph6, 2:RESULTS:paragraph8. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[66] Chronic Oral Inoculation of Porphyromonas gingivalis and Treponema denticola Induce Different Brain Pathologies in a Mouse Model of Alzheimer Disease.

Shows: Twelve-week oral Pg and T. denticola exposures produced differing brain pathology in female mice; coinoculation was less pathogenic than each alone.

Does not show: Pathogen combination and host context alter outcomes; AD-like histology is not human AD, and a stronger universal synergistic prediction failed in this context.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[67] Intracerebral but Not Peripheral Infection of Live Porphyromonas gingivalis Exacerbates Alzheimer's Disease Like Amyloid Pathology in APP-TgCRND8 Mice.

Shows: In APP-TgCRND8 mice, chronic oral infection caused bone loss and gliosis without accelerating Aβ; live intracerebral Pg increased Aβ after eight weeks, unlike heat-killed bacteria.

Does not show: Direct brain injection bypasses the proposed oral route; peripheral and central results provide a route boundary rather than a general oral-infection proof.

Access/status: Complete primary abstract and linked-status metadata examined. Selected body sections: 1:2. Results:paragraph0, 1:2. Results:paragraph5, 1:2. Results:paragraph9, 1:2. Results:paragraph10. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC4.

[68] Treating periodontal disease in patients with myocardial infarction: A randomized clinical trial.

Shows: A randomized 48-person recent-STEMI trial improved six-month FMD after periodontal treatment; inflammatory markers and cardiovascular events did not significantly differ.

Does not show: The small trial's primary endpoint was endothelial physiology; it cannot establish prevention of recurrent clinical events or Pg-specific mediation.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC5.

[69] Periodontal therapy and endothelial function in coronary artery disease: A randomized controlled trial.

Shows: In 69 stable coronary patients, periodontal treatment improved dental status but did not improve three-month FMD relative to control; some vascular inflammatory markers were maintained.

Does not show: Short follow-up and limited size constrain inference; positive adhesion-marker comparisons do not replace the null vasodilation contrast.

Access/status: Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3, SC5.

Documentary record

Every grade on this site is reproducible from its record. The files below are the posted edition exactly as assessed.

• SHA256.txt
• codex-correction-provenance.json
• consensus-addendum.md
• consensus-validation.json
• edition-approval-audit.json
• oral-systemic-disease.review.md
• oral-systemic-disease.scorecard.json
• validation-report.json
• validation-round2-2026-10-14.json
• validation.txt

Sources: 69 · load-bearing claims: 75 · verified: 75 · unverified: 0 · contradicted: 0 · schema 5.3