# HSEF v5.3 review: Vascular Hypothesis of Dementia

> Sources: 69 primary reports; claims: 74; verified: 74; unverified: 0. Verification is bounded attribution. Access: 7 partial-body and 62 complete abstract/status examinations. Literature cutoff: 2026-10-07; source examination UTC date: 2026-10-07. No complete-paper audit or blinded calibration claimed.
> **Adversarial review:** Generated by Sol 6.1; adversarially reviewed by Claude Opus 5.5 (Anthropic) on 2026-10-14. 2 corrections applied (see ADVERSARIAL REVIEW AND APPROVAL). Status: **Approved as revised.**

## AT A GLANCE

**Vascular injury can cause cognitive impairment and amplify Alzheimer disease, but vascular initiation of its amyloid/tau cascade is not established.**

| Level | Confidence | Causal role | Momentum | Stability | Divergence |
|---|---|---|---|---|---|
| Γ — Mechanistically plausible, partial system support | Moderate | reinforcing / stabilizing factor | Mechanistic only | Stable | Contested |

**Deciding experiment:** Enroll baseline vascular-risk and molecularly amyloid/tau-negative strata, then selectively correct the implicated endothelial/pericyte lesion with verified brain engagement. Combine with orthogonal protein-directed comparison and a non-APP-initiated endogenous model using mediator reinstatement. Follow new biomarker conversion, adjudicated disease and cognitive trajectory, with independent replication.

**Belief versus evidence:** Preregistered scope-matched SC3 gives Γ. Vascular cognitive injury reaches Ζ through convergent endogenous perturbations, inherited vasculopathies and randomized human vascular prevention. Those successes do not establish the distinguishing vascular-first AD amyloid/tau sequence. Protein-first injury, rescue without amyloid change, and opposite tau responses remain discriminators.

**Best-supported rival:** Parallel independent pathologies. Perturb vascular and protein processes independently before onset. Parallel injury predicts cognitive improvement without prevention of protein accumulation; vascular-first predicts prevention of new amyloid/tau after engaged vascular protection.

## COMPETING MECHANISTIC MODELS

**Rank 1: Parallel independent pathologies (Ε).** Vascular lesions and proteinopathies can injure cognition independently and jointly. Mixed autopsies, APOE4 BBB prediction independent of measured proteins, PTN rescue despite flow deficits and Debio-025 benefit without Aβ change support conditional causal contributions. Autopsy and independent prediction alone do not prove exclusivity; GLUT1 can modify amyloid clearance.

**Rank 2: Proteinopathy-first vascular amplification (Ε).** Proteins can impair vessels, which amplify established disease. Aβ human-slice constriction and APP knock-in vascular abnormalities, tau-peptide endothelial injury and amyloid-model vascular rescue establish this route. Applied proteins are artificial and vascular injuries can independently impair cognition; it is not every disease initiating pathway.

**Rank 3: Amyloid-driven vascular injury (Ε).** Amyloid vasculopathy produces cerebrovascular dysfunction and injury. Hereditary APP CAA trajectories and model vascular amyloid/hypoperfusion interactions support this defined route. CAA is not all sporadic AD and many vascular injuries do not require amyloid.

**Rank 4: Two-Hit vascular initiation (Γ).** Vascular dysfunction starts the AD amyloid/tau cascade and dementia. Early GLUT1/pericyte lesions and APP-model accelerated deposition provide a coherent mechanism. The upstream APP lesion, absence of tangles in pure VaD, opposite tau response and retracted landmark reports limit causal primacy.

**Rank 5: Shared systemic inflammatory driver (Γ).** Systemic exposures produce both vascular and protein injury. Hepatic APOE and inflammatory/vascular biomarker relationships provide plausible common routes. Pleiotropy and observational paths do not isolate the shared initiating mediator.

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

Cerebrovascular dysfunction, including blood–brain barrier breakdown, pericyte loss and chronic hypoperfusion, initiates the Alzheimer amyloid and tau cascades and thereby dementia.

Use the index Two-Hit proposition. Vascular injury causing cognitive impairment is distinct from vascular initiation of AD-specific pathology. SC3 gates the distinguishing claim; universal dementia extrapolation is separately reported.

Selection: public-interest rank 29 from the editorial ranking compiled 2026-09-23; categorical prior verdicts were visible. Private fresh review; earlier documents are preserved.

### Evidence and causal interpretation

Vascular risk predicts dementia most clearly at midlife, but prediction is not a causal instrument. CAIDE, Whitehall and ARIC provide long intervals and substantial age dependence; population-attributable fractions assume preventability. Autopsy mixtures show infarcts lower the AD-lesion burden needed for clinical expression, while ACT and Rush identify several concurrent paths. Atherosclerosis can lack an AD-protein association. These favor conditional and parallel contributions rather than one universal cascade (PMID16914401, PMID29901708, PMID40455489, PMID9052711, PMID17568013, PMID17879383, PMID19521084).

Original adult pericyte-ablation Results use dual-promoter CreER/iDTR and toxin controls. Severe vascular failure precedes NeuN/neurite loss and cognitive impairment. Physiological PTN infusion preserves neurons and cognition despite persistent flow/BBB abnormalities; PTN knockdown alone does not kill neurons, but increases vulnerability to other injury. This establishes a multi-hit vascular/trophic route, without testing AD amyloid or tau onset (PMID31235908).

Original GLUT1 Results separate endothelial from astrocytic deletion for BBB breakdown, and demonstrate increased late Aβ/NeuN loss in Slc2a1-deficient APP mice. Regional GLUT1 or LRP1 re-expression reduces Aβ. Vascular structure changes precede overt neuronal loss, but APP is present from development, and the authors explicitly allow developmental synergy. This is endogenous vascular contribution and clearance modulation, not established spontaneous vascular-first AD (PMID25730668).

Original Montagne human Results show APOE4 BBB damage and a pericyte marker predicting later cognition independently of measured amyloid/tau. The longitudinal subset is 146, with imaging/CSF denominators varying; cognitively normal APOE4 participants already have some regional amyloid. This is not a protein-free pre-onset intervention. Nation and earlier imaging support barrier relevance, while DELCODE and differing water/albumin permeability measures demonstrate assay and disease-stage dependence (PMID32376954, PMID30643288, PMID25611508, PMID39085945, PMID34041783).

Original Debio-025 Results begin treatment at 10–12 months after APOE4/5xFAD injury. Thirty days improve barrier integrity and prevent further neuronal/cognitive decline without changing Aβ. Improvement concerns the additional APOE4 deficit, and does not restore all 5xFAD performance to wild type. Astrocytic APOE4 removal can also protect vessels despite increased CAA. These decouple vascular cognitive benefit from amyloid reduction and constrain the universal cascade (PMID35291561, PMID36922879).

Original Nortley human/rat slice experiments show applied Aβ constricts capillaries, with scrambled controls. APP knock-in cortex has pericyte-localized constriction, whereas plaque-free cerebellum does not. CNP reverses acute constriction, while ROS/endothelin blockers prevent further narrowing without immediate baseline restoration. Human biopsy association uses few patients despite many measured vessels, and the roughly 50% flow estimate is modeled. Tau peptides and injected Aβ add reverse protein-to-vessel routes (PMID31221773, PMID38420777, PMID38475929).

Hypoperfusion effects are not uniform: APP mice can accumulate Aβ and aged 3xTg/wild-type mice increase selected phospho-tau without increasing Aβ42. Original Gheni Results instead place BCAS two months after tau seeding and find reduced tau accumulation, more white-matter injury and limited behavioral synergy. CTSD induction/colocalization suggests a clearance mediator without demonstrating necessity. NACC remote-stroke associations have selection and covariate limits; they do not show stroke prevents AD (PMID19646974, PMID27050297, PMID39621511).

Original cis-tau Results induce endogenous cis P-tau after BCAS and show antibody/DAPK1 subtraction prevents injury and executive dysfunction. Pure human VaD brains lack overt NFTs. Young/aged htau treatment is a separate tauopathy experiment; aged rescue occurs without reducing NFT-like pathology. Combining these models cannot manufacture a single demonstrated BCAS-to-classic-AD initiating sequence (PMID34078745).

The 2025 rural-China cluster trial randomizes 33995 adults with uncontrolled hypertension in 326 villages and reduces 48-month incident all-cause dementia. It tests provider-led blood-pressure care, not a selective BBB/pericyte intervention or amyloid/tau onset. SPRINT reduces MCI/composite outcomes but probable dementia is nonsignificant in both primary and extended reports; extended ascertainment is incomplete. PROGRESS benefit is tied especially to recurrent-stroke injury. Independent trial evidence establishes clinically relevant vascular contribution without proving AD-specific primacy (PMID40258956, PMID30688979, PMID39819096, PMID12742805).

HYVET and preDIVA primary dementia contrasts are null, INFINITY changes MRI lesions without gait/cognitive benefit, and FINGER bundles vascular monitoring with diet, exercise and training. Syst-Eur follow-up is the same initially randomized cohort after controls were offered therapy. Clinical prevention heterogeneity limits universal treatment claims and cannot be resolved by counting each follow-up as a new trial (PMID18614402, PMID27474376, PMID31607143, PMID25771249, PMID9802273, PMID12374512).

NOTCH3 and TREX1 provide independent vascular genetic routes to SVD/cognitive injury with variable penetrance. APP/PS1 families and hereditary CAA can have presymptomatic perfusion changes, but their initiating mutations already exist. EPAD statistical mediation and vascular single-nucleus risk-gene expression are mechanistic leads rather than selective causal instruments. New ANGPT2 and LRP1 interventions modify genetically initiated AD models (PMID32732295, PMID27604306, PMID11402113, PMID9539411, PMID35360926, PMID39288341, PMID35165441, PMID41529684, PMID41052971).

Two pericyte landmarks are excluded: Sagare 2013 PMID24336108 and Montagne 2018 PMID29400711 have publisher retractions after image/data concerns. Their notices are examined and preserved, and neither report contributes positive evidence or corpus count. Remaining reports retain their own status and shared-platform limits. Consensus broad output incorrectly labels human vascular-only classic AD initiation Moderate; the cited observational paper does not establish that. Its targeted original passages support specific PTN/GLUT1/Debio endpoints, independently confirmed here; the China trial was unavailable to it. SciSpace leads resolve to five selected primaries and one duplicated preprint, without padding the 69-report corpus.

## HYPOTHESIS STATE EVALUATION FRAMEWORK (HSEF v5.3)

<!-- consensus-consistency-2026-10-09 -->
### Consensus provenance reconciliation — 9 October 2026

Case-specific Consensus discovery records were located and linked to this final edition. This supplements the review's provenance; it leaves the original assessed claim, grades and source counts intact. Provider summaries and full-text badges do not establish examination of original primary passages by this reviewer. Source access and incorporation retain the limits stated in the original methods and any dated research update.

Recorded evidence artifacts:

- `deliverables/hsef_v53_popularity_ranks_26_30_2026-10-07/two-hit-vascular-dementia/consensus-leads-search-log.json`
- `deliverables/hsef_v53_popularity_ranks_26_30_2026-10-07/two-hit-vascular-dementia/consensus-leads.json`
- `deliverables/hsef_v53_popularity_ranks_26_30_2026-10-07/two-hit-vascular-dementia/consensus/broad-browser-snapshot.txt`
- `deliverables/hsef_v53_popularity_ranks_26_30_2026-10-07/two-hit-vascular-dementia/consensus/query-ledger.json`
- `deliverables/hsef_v53_popularity_ranks_26_30_2026-10-07/two-hit-vascular-dementia/consensus/reconciliation.md`

The located artifacts do not state a reliable source-comparison count. No count is inferred.

No earlier adversarial approval has been verified for this edition at the supplement date. This provenance supplement has no separate approval and is not an independent or blinded adjudication. No score or confidence change is made by this reconciliation.
<!-- end-consensus-consistency -->


### Subclaim matrix and aggregation

| Subclaim | Scope | Level | Decision |
|---|---|---|---|
| SC1 Vascular risk and cerebrovascular lesions predict dementia. | baseline midlife exposures and late-life incident dementia | Β | Midlife risks, imaging and mixed autopsy lesions predict dementia, with substantial age, region and assay dependence. The descriptive association claim is Β; intervention evidence is assigned to SC2 rather than silently changing SC1. |
| SC2 Cerebrovascular injury causes cognitive impairment. | defined infarction, small-vessel disease and vascular interventions | Ζ | Adult pericyte ablation with PTN rescue, endogenous ischemia/cis-tau subtraction, rare genetic vasculopathies and randomized blood-pressure treatment converge on vascular cognitive injury across different failure structures. China incident dementia and PROGRESS stroke-linked cognitive benefit add human functional evidence. SPRINT primary dementia, preDIVA and INFINITY cognitive nulls limit universal therapeutic efficacy but do not negate the assembled causal set. Ζ applies to defined vascular injury, not an AD-specific initiating cascade or universal Η trajectory. |
| SC3 Vascular breakdown initiates AD-specific amyloid and tau cascades. | prespecified AD vascular and proteinopathy settings | Γ | GLUT1 and hypoperfusion accelerate pathology in genetically initiated APP models; vascular stress can induce endogenous cis P-tau and cognitive injury without AD tangles. Human BBB prediction is independent of measured protein markers, not a demonstrated path from BBB to proteins. APP/tau can instead injure vessels. No selective pre-onset human or non-protein-initiated natural set establishes the specific two-hit AD sequence. Mechanistic support warrants Γ. |
| SC4 The Two-Hit sequence generalizes to late-life dementia. | late-life dementias and mixed pathologies | Γ | Mixed dementia includes independent infarcts, amyloid/tau, Lewy pathology and variable inherited vascular routes. Multiple clinical prevention contrasts are null or endpoint-dependent; hypoperfusion has opposite protein effects across models. A universal sequence cannot be inferred from its conditional amplification successes. |

Preregistered aggregation: scope-matched. Preregistered scope-matched SC3 gives Γ. Vascular cognitive injury reaches Ζ through convergent endogenous perturbations, inherited vasculopathies and randomized human vascular prevention. Those successes do not establish the distinguishing vascular-first AD amyloid/tau sequence. Protein-first injury, rescue without amyloid change, and opposite tau responses remain discriminators. Sensitivities: weakest-link Β; primary-subclaim Γ; scope-matched Γ.

### Decisive-perturbation inventory and causal tally

[1] Pdgfrb-deficient adult mice develop vascular dysfunction before general neuronal injury. Engineered pericyte loss does not establish AD-specific proteinopathy onset; linked correction details unexamined.

[2] APOE4 activates pericyte CypA-MMP9 and genetic or pharmacological pathway subtraction protects BBB. APOE affects multiple compartments; BBB protection is not spontaneous human AD prevention; correction details unexamined.

[4] Inducible pericyte ablation causes flow/BBB deficits and neuronal/cognitive loss; physiological PTN infusion preserves neurons and cognition despite persistent circulatory defects. Severe acute dual-promoter ablation is not gradual human disease; amyloid/tau initiation is not tested; PTN depletion alone does not kill neurons.

[5] Developmental pericyte coverage determines BBB permeability before astrocyte development. Developmental barrier formation is not adult AD etiology or clinical trajectory.

[11] At 10–12 months Debio-025 improves APOE4 BBB and prevents further neuronal loss and cognitive decline without changing Aβ pathology. Treatment modifies the APOE4 increment rather than restoring all 5xFAD deficits to wild type; APP mutations precede the vascular insult.

[12] Endothelial GLUT1 loss causes early BBB failure; combined Slc2a1 deficiency/APP increases later Aβ and NeuN loss; GLUT1/LRP1 re-expression reduces Aβ. APP overexpression is present from development; late acceleration and regional clearance rescue are not initiation of spontaneous human AD or whole-animal cognitive reversal.

[13] Applied Aβ constricts human/rat capillaries; APP knock-in mice show endogenous plaque-region pericyte constriction; CNP reverses acute constriction. Small human biopsy and mouse samples; thousands of vessel measurements are not independent patients; modeled flow reduction is not measured human flow.

[14] AD-related tau peptides cause endothelial calcium and vasodilatory dysfunction. Peptide exposure demonstrates a possible protein-to-vessel route, not its prevalence in spontaneous AD.

[15] Aβ exposure injures pericytes and endothelial capillary structure. Exogenous challenge establishes capacity, not ordinary endogenous initiating direction.

[16] Early astrocytic APOE4 removal in 5xFAD reduces vascular injury despite increased vascular amyloid. Genetic APP proteinopathy persists and cell-specific APOE effects are not a vessel-only intervention.

[17] Anti-Ly6G relieves neutrophil capillary stalls and rapidly improves flow and memory in APP/PS1 and 5xFAD mice. Acute functional rescue occurs after engineered proteinopathy and is not prevention of AD onset.

[19] Microinfarcts produce sex-dependent cognitive injury while decreasing plaques without changing soluble Aβ in APP/PS1 mice. Surgical infarcts and strain/sex differences constrain generalization; lower plaques are not lower overall injury.

[20] BCAS performed two months after tau seeding reduces accumulated tau while worsening white matter; NACC remote stroke associates with less AD pathology. Seeded tau is imposed first; CTSD colocalization/upregulation is a candidate mediator without necessity subtraction; autopsy associations retain survival and diagnosis biases.

[21] Hypoperfusion in APP23 exacerbates amyloid/cognitive abnormalities and galantamine ameliorates phenotypes. APP is genetically upstream; galantamine has nonvascular neuronal actions.

[22] Unilateral occlusion raises selected phospho-tau epitopes in aged wild-type/3xTg mice without changing Aβ42. Age/genotype-specific phosphorylation is not a complete AD amyloid/tau initiating cascade.

[23] BCAS in APP-transgenic mice increases intracellular and fibrillar Aβ deposits. Transgenic APP is upstream, and accelerated deposition does not identify sporadic onset.

[24] BCAS augments vascular amyloid and microinfarcts in CAA mice; human postmortem observations agree. Amyloid vasculopathy is pre-existing; terminal association and hypoperfusion amplification are not vascular-first AD initiation.

[25] Chronic hypoperfusion produces white-matter injury and selective spatial working-memory deficits. Excluding mice with neuronal perikaryal injury conditions analysis on an outcome; impairment is not necessarily AD dementia.

[26] BCAS induces endogenous cis P-tau; cis antibody and DAPK1 loss preserve white matter and executive function; separate htau mice respond to treatment. Human VaD brains lack overt NFTs; htau AD-like tangles arise in a separate model, so BCAS-to-classic-AD initiation is not demonstrated.

[28] Astrocytic APOE4 deletion ameliorates BBB impairment relative to matched APOE3 designs. APOE has nonvascular effects; barrier rescue does not establish that amyloid/tau initiation depends on barrier failure.

[29] Liver-only APOE4 causes vascular/synaptic/cognitive abnormalities and exacerbates APP-model amyloid; APOE3 plasma improves aged phenotypes. Systemic pleiotropy and APP crossing distinguish vascular contribution from AD-specific primacy.

[31] SPRINT intensive blood-pressure treatment reduces MCI and combined cognitive endpoints. Primary probable-dementia contrast is nonsignificant; early cardiovascular stopping and exclusion of diabetes/stroke limit scope; no AD biomarker-onset endpoint.

[32] SPRINT extended follow-up retains MCI/composite benefit. Probable dementia remains nonsignificant; only 59% of eligible participants were ascertained; same randomized cohort and unexamined correction do not add independent replication.

[33] Rural-China cluster randomization of 33995 adults in 326 villages lowers 48-month incident all-cause dementia with provider-led blood-pressure treatment. Bundled care is not a selective BBB/pericyte intervention; all-cause dementia benefit does not demonstrate prevention of AD-specific amyloid/tau; abstract/status depth only.

[34] HYVET-COG randomized antihypertensive treatment in adults over 80 and measured incident dementia. Short follow-up primary dementia contrast is nonsignificant; meta-analysis pooling is not a positive primary trial result.

[35] Syst-Eur randomized antihypertensive therapy yields fewer clinical dementia cases over two years. Only 32 events and borderline significance; clinical classification does not establish biomarker-defined AD prevention.

[36] Syst-Eur extended follow-up reports continued dementia incidence difference including clinical AD cases. Control participants were offered active treatment after blinded phase; same initial cohort, not an independent randomized trial or proof of molecular onset prevention.

[37] INFINITY lowers white-matter hyperintensity accumulation with intensive ambulatory blood-pressure control. Gait and cognition contrasts are null despite MRI effects; biomarkers cannot stand for preserved clinical trajectory.

[38] preDIVA vascular-care cluster trial shows no overall incident dementia or disability reduction. Low baseline risk and strong usual care reduce intervention contrast; null bundled care is not selective pathway falsification.

[39] PROGRESS reduces cognitive decline with perindopril-based therapy after stroke/TIA; dementia benefit concentrates with recurrent stroke. Overall dementia contrast is null and benefit without recurrent stroke is not established; this supports stroke-related injury prevention rather than obligatory AD cascade interruption.

[40] TREAT-SVD crossover compares cerebrovascular reactivity under three drugs; CADASIL shows differential physiological effects. No drug difference in sporadic SVD; four-week physiology is not functional or dementia prevention; small analyzed subgroups.

[41] FINGER multidomain intervention improves a neuropsychological composite over two years. Diet, exercise and cognitive training accompany vascular monitoring; benefit cannot be assigned to vascular correction or AD onset prevention.

[56] NOTCH3 variants produce heterogeneous SVD/vascular phenotypes in population carriers. Many imaged carriers remain normal into late life; penetrance and selection prevent universal dementia inference.

[57] NOTCH3 domain location stratifies CADASIL stroke/SVD risk across clinical and population samples. Genetic risk strata are not intervention outcomes; phenotype/penetrance varies and AD amyloid/tau onset is not tested.

[58] TREX1 families develop systemic retinal/cerebral vasculopathy with cognitive decline in a substantial subset. Rare familial systemic disease is a distinct vascular cause, not universal AD dementia or a vessel-only mediator instrument.

[67] LRP1-targeted transport therapy clears amyloid and improves cognition in AD models. Treatment after existing pathology tests clearance and delivery, not whether vascular failure initiated AD.

[68] Endothelial ANGPT2 loss reduces amyloid, while overexpression worsens BBB/inflammatory/memory phenotypes in 5xFAD. Genetic APP/PS1 pathology is upstream; expression/phenotype rescue does not demonstrate ordinary human vascular-first onset.

Tally: 37 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

SC2 endogenous injury and rescue plus human clinical trajectories form a convergent causal set; associative legs do not re-cap the set at Β. SC3 separately asks AD-specific initiation and remains Γ because genetically imposed proteins, injury-related tau and independent BBB prediction do not establish it. Model contribution is preserved rather than transferred to broad etiology.

Gamma hard-rule audit: In vitro toxicity or barrier networks do not establish human AD onset. Exogenous Aβ/tau challenge or APP overexpression does not establish spontaneous vascular-first etiology. Biomarker, MRI and fitted mediation changes are not incident AD-specific proteinopathy prevention. Presymptomatic inherited APP/PS1 perfusion change is not demonstrated pre-proteinopathy change.

SC1 cap: “Pure association evidence — however large, however replicated — cannot exceed Β.” Midlife risks, imaging and mixed autopsy lesions predict dementia, with substantial age, region and assay dependence. The descriptive association claim is Β; intervention evidence is assigned to SC2 rather than silently changing SC1.

SC3 cap: “Maximum if mechanism is biologically plausible, supported by in vitro or simplified animal paradigms, but no disease-relevant necessity is shown in vivo.” GLUT1 and hypoperfusion accelerate pathology in genetically initiated APP models; vascular stress can induce endogenous cis P-tau and cognitive injury without AD tangles. Human BBB prediction is independent of measured protein markers, not a demonstrated path from BBB to proteins. APP/tau can instead injure vessels. No selective pre-onset human or non-protein-initiated natural set establishes the specific two-hit AD sequence. Mechanistic support warrants Γ.

SC4 cap: “Maximum if mechanism is biologically plausible, supported by in vitro or simplified animal paradigms, but no disease-relevant necessity is shown in vivo.” Mixed dementia includes independent infarcts, amyloid/tau, Lewy pathology and variable inherited vascular routes. Multiple clinical prevention contrasts are null or endpoint-dependent; hypoperfusion has opposite protein effects across models. A universal sequence cannot be inferred from its conditional amplification successes.

SC1 field-valid definition: “
- Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo
- Selective pathway inhibition that changes onset/progression while preserving broader viability
- Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. Midlife risks, imaging and mixed autopsy lesions predict dementia, with substantial age, region and assay dependence. The descriptive association claim is Β; intervention evidence is assigned to SC2 rather than silently changing SC1.

SC2 field-valid definition: “
- Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo
- Selective pathway inhibition that changes onset/progression while preserving broader viability
- Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: yes. Adult pericyte ablation with PTN rescue, endogenous ischemia/cis-tau subtraction, rare genetic vasculopathies and randomized blood-pressure treatment converge on vascular cognitive injury across different failure structures. China incident dementia and PROGRESS stroke-linked cognitive benefit add human functional evidence. SPRINT primary dementia, preDIVA and INFINITY cognitive nulls limit universal therapeutic efficacy but do not negate the assembled causal set. Ζ applies to defined vascular injury, not an AD-specific initiating cascade or universal Η trajectory.

SC3 field-valid definition: “
- Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo
- Selective pathway inhibition that changes onset/progression while preserving broader viability
- Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. GLUT1 and hypoperfusion accelerate pathology in genetically initiated APP models; vascular stress can induce endogenous cis P-tau and cognitive injury without AD tangles. Human BBB prediction is independent of measured protein markers, not a demonstrated path from BBB to proteins. APP/tau can instead injure vessels. No selective pre-onset human or non-protein-initiated natural set establishes the specific two-hit AD sequence. Mechanistic support warrants Γ.

SC4 field-valid definition: “
- Endogenous genetic deletion, suppression, or degron depletion of the proposed mediator in vivo
- Selective pathway inhibition that changes onset/progression while preserving broader viability
- Vagotomy-type natural experiment, only if confounding is rigorously controlled AND effect is disease-initiation relevant rather than merely associated” Positive full-scope result: no. Mixed dementia includes independent infarcts, amyloid/tau, Lewy pathology and variable inherited vascular routes. Multiple clinical prevention contrasts are null or endpoint-dependent; hypoperfusion has opposite protein effects across models. A universal sequence cannot be inferred from its conditional amplification successes.

### Upward threshold ladder

SC1 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Keep prediction as descriptive and test intervention in a distinct causal subclaim.

SC2 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: yes. 

SC2 Ε: “Requires causal support in a defined context, but with clear restriction by subtype, disease stage, model, or tissue.

Use Ε when there is a real causal contribution, but generalization to the entire disease class fails.” Satisfied: yes. 

SC2 Ζ: “Requires:
- Multiple independent causal lines in vivo
- Replication across at least two relevant systems
- Narrowing of major alternative models” Satisfied: yes. 

SC2 Η: “Requires:
- Strong convergent causal evidence across systems
- Intervention-sensitive change in trajectory
- No major unresolved contradiction about whether the mechanism is upstream versus downstream” Satisfied: no. Reconcile late-life prevention nulls with adequate exposure and baseline risk, demonstrate sustained functional trajectories and replicate mechanistically matched vascular restoration.

SC3 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Before proteinopathy, selectively preserve BBB/pericyte/flow function in a prospectively defined non-APP-initiated setting and show prevention of new amyloid/tau plus cognitive disease, with reverse-direction protein intervention and mediator reinstatement controls.

SC4 Δ: “Requires at least one genuine causal perturbation in a disease-relevant in vivo system, or equivalent human natural experiment that materially supports the mechanism.” Satisfied: no. Map baseline vascular/protein/genetic strata prospectively and establish the ordering in independent disease-separated populations rather than pooled late-life syndromes.

Upper ladder: Randomized China, PROGRESS, SPRINT and other clinical endpoints and NOTCH3/TREX1 natural experiments are live upper-level evidence for vascular injury. No unethical randomized induction of human vascular damage is required. Existing convergent instruments reach SC2 Ζ but neither incident all-cause dementia nor familial SVD tests AD-specific proteinopathy initiation. Molecular sequencing and selective engagement are the remaining scope-specific requirements.

Independence: Distinct adult cell ablation, ischemic mediator subtraction, clinical randomization and rare vasculopathy routes converge on defined vascular cognitive injury. Several pericyte/GLUT1/APOE4 reports share a laboratory platform and possible assay vulnerabilities. SPRINT and Syst-Eur follow-ups, ARIC and Whitehall reports are not independent cohorts. These lines do not converge on vascular initiation of human AD-specific proteins. Report-specific retractions remove two landmarks from the positive corpus; other reports are not invalidated by affiliation alone.

Inducible cell loss and neurotroph rescue [4]: Adult cell ablation and PTN rescue distinguish cell-derived neuronal support from persistent flow loss.

Ischemic mediator subtraction [26]: Endogenous cis-tau/DAPK1 subtraction after BCAS tests injury mediation through a different molecular instrument.

Randomized clinical vascular prevention [33, 39]: Human incident dementia/cognitive endpoints test intervention-sensitive trajectories rather than animal marker expression.

Inherited non-AD vasculopathies [58, 56]: Independent familial vascular routes exclude APP overexpression as a required cause of vascular cognitive injury, while retaining penetrance and pleiotropy limits.

### Falsification and pitfalls

The nominal causal ordering can run in reverse or fail to generate proteins. Aβ/tau injure vessels, BCAS after tau seeding lowers accumulation and pure VaD cis-tau brains lack overt NFTs. Additional downgrade: False.

Engaged vascular benefit can decouple from amyloid. Debio-025 neuronal/cognitive protection occurs without Aβ change, and PTN preserves neurons despite persistent circulatory defects. Additional downgrade: False.

Retracted pericyte studies inflate convergence. PMID24336108 and PMID29400711 excluded with publisher notices.

All-cause dementia prevention becomes AD molecular prevention. China trial and stroke-related PROGRESS outcomes.

Thousands of vessels counted as independent humans. Nortley human biopsy and mouse sampling.

Statistical mediation establishes temporal causality. EPAD and VEGF path models.

Presymptomatic inherited disease means vascular-first. APP/PS1 and hereditary CAA.

Provider Moderate initiation label becomes framework score. Consensus broad table unsupported human initiation claim rejected.

Partial body retrieval called whole-paper validation. Seven selected Results bodies; clinical papers abstract/status only.

### Adversarial steelman: above, then below

The strongest higher-level case combines early endothelial GLUT1 deficiency, clearance rescue, APOE4 barrier prediction and incident dementia prevention. These establish more than passive vascular association, and SC2 receives Ζ. APP already precedes the model vascular insult, independent human BBB prediction does not lead to proteins, and clinical endpoints are not molecular onset, so the same set does not cross SC3 Δ.

A Β-only account misses adult pericyte loss/rescue, endogenous ischemic mediator subtraction, human slice perturbation and randomized prevention. The AD-initiation chain has credible mechanistic support even though primacy is unestablished.

Γ is retained for the distinguishing AD-specific claim, with Ζ for defined vascular cognitive injury. Neither vascular injury as passive response nor universal vascular-first AD fits the full evidence.

### Final score, causal direction and M/S/D

Preregistered scope-matched SC3 gives Γ. Vascular cognitive injury reaches Ζ through convergent endogenous perturbations, inherited vasculopathies and randomized human vascular prevention. Those successes do not establish the distinguishing vascular-first AD amyloid/tau sequence. Protein-first injury, rescue without amyloid change, and opposite tau responses remain discriminators.

In AD the vascular pathway can amplify proteinopathy and cognitive injury; removing it can improve function without removing proteins. Defined vascular injury can independently initiate non-AD cognitive disease. No common rate-limiting AD initiating step is demonstrated.

Momentum: Mechanistic only. 2025 incident dementia prevention and 2026 endothelial ANGPT2 model perturbations refine clinical contribution and molecular amplification. No currently recruiting decisive onset trial is asserted.

Stability: Stable. No verified named pending result is identified as likely to change SC3 imminently; pre-onset selective molecular prevention would do so.

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

## PROPOSED DECISIVE EXPERIMENTS

Enroll baseline vascular-risk and molecularly amyloid/tau-negative strata, then selectively correct the implicated endothelial/pericyte lesion with verified brain engagement. Combine with orthogonal protein-directed comparison and a non-APP-initiated endogenous model using mediator reinstatement. Follow new biomarker conversion, adjudicated disease and cognitive trajectory, with independent replication.

Readout: Incident amyloid/tau conversion, neuronal/white-matter integrity, dementia subtype and sustained cognition. Controls: Baseline genotype/proteins, age, stroke, vascular lesion, BBB assay specificity, general hemodynamics, bundled care, mortality and attrition. Supporting: Engaged vascular preservation prevents new AD-specific proteins and disease, and matched lesion reinstatement reverses that protection in predicted strata. Falsifying: Adequate early vascular correction preserves cognition without changing protein onset, or protein-directed correction prevents vascular injury instead.

## RESEARCH TIMELINE

| Year | Milestone | Primary source | Interpretation |
|---|---|---|---|
| 1997 | Infarcts and AD clinical expression | [PMID9052711](https://pubmed.ncbi.nlm.nih.gov/9052711/) | Autopsy interaction, not molecular initiation. |
| 1998 | Syst-Eur randomized prevention | [PMID9802273](https://pubmed.ncbi.nlm.nih.gov/9802273/) | Small clinical event count. |
| 2003 | PROGRESS stroke-linked cognitive prevention | [PMID12742805](https://pubmed.ncbi.nlm.nih.gov/12742805/) | Overall dementia contrast null. |
| 2007 | Mixed pathologies in Rush/ACT | [PMID17568013](https://pubmed.ncbi.nlm.nih.gov/17568013/) | Independent injury contributions. |
| 2010 | Pericytes and BBB biology | [PMID21040844](https://pubmed.ncbi.nlm.nih.gov/21040844/) | Engineered lesion, linked correction. |
| 2012 | APOE4-CypA barrier route | [PMID22622580](https://pubmed.ncbi.nlm.nih.gov/22622580/) | Genetic model contribution. |
| 2013/2024 | Sagare pericyte report and retraction | [PMID38570509](https://pubmed.ncbi.nlm.nih.gov/38570509/) | Excluded original PMID24336108. |
| 2015 | GLUT1/endothelial clearance rescue | [PMID25730668](https://pubmed.ncbi.nlm.nih.gov/25730668/) | APP already upstream. |
| 2016 | preDIVA null prevention | [PMID27474376](https://pubmed.ncbi.nlm.nih.gov/27474376/) | Bundled vascular care. |
| 2018/2024 | White-matter pericyte report and retraction | [PMID38580816](https://pubmed.ncbi.nlm.nih.gov/38580816/) | Excluded original PMID29400711. |
| 2019 | Pericyte ablation/PTN rescue | [PMID31235908](https://pubmed.ncbi.nlm.nih.gov/31235908/) | Flow and neuronal outcomes decouple. |
| 2019 | Aβ capillary constriction | [PMID31221773](https://pubmed.ncbi.nlm.nih.gov/31221773/) | Protein-to-vessel direction. |
| 2019 | SPRINT cognitive outcomes | [PMID30688979](https://pubmed.ncbi.nlm.nih.gov/30688979/) | MCI benefit; primary dementia nonsignificant. |
| 2020 | APOE4 BBB longitudinal prediction | [PMID32376954](https://pubmed.ncbi.nlm.nih.gov/32376954/) | Independence from proteins is not primacy. |
| 2021 | Cis-tau ischemic mediation | [PMID34078745](https://pubmed.ncbi.nlm.nih.gov/34078745/) | VaD lacks NFTs; htau model separate. |
| 2021 | Advanced APOE4 vascular rescue | [PMID35291561](https://pubmed.ncbi.nlm.nih.gov/35291561/) | Aβ unchanged; index PMID appears 2022. |
| 2024 | EPAD path-model mediation | [PMID39288341](https://pubmed.ncbi.nlm.nih.gov/39288341/) | Fitted order not causal proof. |
| 2025 | Hypoperfusion reduces seeded tau | [PMID39621511](https://pubmed.ncbi.nlm.nih.gov/39621511/) | Online 2024; journal 2025. |
| 2025 | Rural-China incident dementia prevention | [PMID40258956](https://pubmed.ncbi.nlm.nih.gov/40258956/) | All-cause clinical endpoint. |
| 2026 | Endothelial ANGPT2 perturbation | [PMID41529684](https://pubmed.ncbi.nlm.nih.gov/41529684/) | Genetically initiated 5xFAD contribution. |

## KNOWLEDGE GAPS

Distinguish clinical presymptomatic from pre-proteinopathy timing.

Test AD-specific conversion after engaged vascular prevention, not merely all-cause dementia.

Replicate pericyte/GLUT1 pathways on independent platforms after report-specific retractions.

Separate pericyte trophic support from flow and barrier failure with factorial rescues.

Reconcile increased cis/phospho-tau with reduced seeded tau under hypoperfusion.

Map brain region, sex, genotype and baseline vascular burden prospectively.

Validate BBB assay identity and keep retinal proxies separate.

Resolve clinical prevention heterogeneity, adequate exposure and selective mediation.

## 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 reports examined: seven selected partial bodies and 62 complete abstract/status examinations. No whole-paper or supplement audit. A one-sentence preDIVA follow-up abstract without evaluable outcomes is excluded. Report-specific retractions are excluded. Scope-matched SC3 Γ; defined vascular injury SC2 Ζ. Two Consensus Pro browser queries, zero API/Deep; one SciSpace discovery query with five matched primaries and deduplicated Nortley preprint. Carried controls are not fresh blind calibration.

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:** 2026-10-14.
This edition is the revised copy produced by the 2026-10-09 Consensus-consistency audit; it was reviewed
as it now stands, and the only difference from the 2026-10-07 edition is the added provenance
reconciliation block, which introduced no error.

**Scope of checks.** REFERENCE PMIDs verified against PubMed metadata: 69/69 resolve to the stated
title, first author, journal and year; DOIs cross-checked for all entries that carry one. Retraction
status re-checked for every reference: none of the 69 carries a retraction, retraction-of-publication or
expression-of-concern flag, and the two retracted pericyte landmarks the review excludes
(PMID 24336108, PMID 29400711) are confirmed by PubMed to carry the Retracted Publication type and are
correctly absent from the reference list and the corpus count. Load-bearing claims spot-checked against
PubMed abstracts: 12 (the rural-China cluster trial's randomisation and outcome; SPRINT's nonsignificant
probable-dementia contrast alongside significant MCI and composite results; the APOE4 BBB prediction
independent of amyloid and tau; cyclophilin-A suppression improving barrier integrity without changing
amyloid; endothelial versus astrocytic GLUT1 deletion; the pericyte-ablation/pleiotrophin rescue;
cis P-tau without tangles in vascular cognitive impairment; hypoperfusion reducing tau accumulation;
amyloid-oligomer capillary constriction; astrocytic APOE4 removal; the ANGPT2 and LRP1 interventions).
Scoring logic re-derived against `references/hsef-framework.md` and the neurodegeneration rubric: ladder
letter and canonical descriptor, cap precedence, the pre-registered scope-matched aggregation and its
reported alternatives, the causal-role derivation, and the momentum artifact.

**Findings.**

| ID | Severity | Location | Original → corrected | Evidence |
|---|---|---|---|---|
| 1 | Major | REFERENCES #31 (PMID 30688979) | ", Williamson JD, Pajewski NM et al." → "SPRINT MIND Investigators for the SPRINT Research Group, Williamson JD, Pajewski NM et al." | Known reference-builder defect: a consortium first author is dropped and leaves a stray leading comma. The JAMA 2019;321(6):553-561 byline is "SPRINT MIND Investigators for the SPRINT Research Group"; PubMed records it in the CollectiveName field, which the builder does not read. |
| 2 | Minor | REFERENCES #45 (PMID 9052711) | "DOI: . Access: abstract_metadata." → "Access: abstract_metadata." | Empty DOI rendering defect. PubMed holds no DOI for this 1997 JAMA record, so the empty field is removed rather than filled with an unverified string. |

**No level, confidence, role, momentum, stability or divergence value was changed.** Final score remains
Γ — Mechanistically plausible, partial system support; confidence Moderate; role reinforcing /
stabilizing factor; SC2 remains Ζ and is reported separately at its own scope.

**Rejected findings (checked, no change).**

1. *Scope-matched aggregation.* The rule is "the highest subclaim whose scope matches the central claim's
   scope; narrower subclaims reported separately". The central claim is vascular initiation of the AD
   amyloid/tau cascades and thereby dementia. SC2 (Ζ) is about vascular injury causing cognitive
   impairment, which is a different scope, so it correctly does not set the headline; the scope-matched
   subclaims are SC3 and SC4, both Γ. The derivation wording says this explicitly rather than treating Ζ
   as the headline. Reported alternatives (weakest-link Β, primary-subclaim Γ) are correct. No change.
2. *`v5.claim_form` warning.* The validator warns that `entry_form` is "phenomenon". The registered claim
   states initiation plus its consequence ("initiates the ... cascades and thereby dementia"), which is
   the rejected-finding convention for this warning. No change.
3. *Validator lay-summary failure.* The v5.4 framework copies in this package add a mandatory
   `v5.lay_summary` check. This review is v5.3 and the version stored with the case (validator v2.3)
   returns ok; the failure is a validator-version artifact. No lay summary was added.
4. *Momentum.* "Mechanistic only" names its artifact (the 2025 incident-dementia prevention trial and the
   2026 endothelial ANGPT2 model perturbation) and correctly declines "Decisive work underway", since no
   recruiting pre-onset selective-engagement trial is identified. No change.
5. *Reference title mismatches at #11 (PMID 35291561) and #56 (PMID 32732295).* PubMed's title field
   drops the italicised gene names (APOE4, NOTCH3); the review's titles are the published ones. Rejected
   per the italicised-gene-name convention.
6. *Retracted-landmark handling.* Checked and correct: both retracted reports are named, their notices
   recorded, and neither contributes positive evidence or corpus count.

**Residual limitations.** Verification is bounded attribution: 7 of 69 reports were examined at
partial-body depth and 62 at abstract/status depth, so claims resting on unexamined Methods detail
(for example the "146 longitudinal subset", the "59% ascertained" figure and the Debio-025 dosing window)
were checked for consistency with the abstract but could not be re-derived from the full text here.

**Approval:** Following adversarial review and the corrections listed above, the revised document is approved. — Claude Opus 5.5 (Anthropic), 2026-10-14.

## REFERENCES

1. Bell RD, Winkler EA, Sagare AP et al. Pericytes control key neurovascular functions and neuronal phenotype in the adult brain and during brain aging. Neuron. 2010. [PMID 21040844](https://pubmed.ncbi.nlm.nih.gov/21040844/). DOI: 10.1016/j.neuron.2010.09.043. Access: abstract_metadata. Study: pericyte-Pdgfrb-platform.

2. Bell RD, Winkler EA, Singh I et al. Apolipoprotein E controls cerebrovascular integrity via cyclophilin A. Nature. 2012. [PMID 22622580](https://pubmed.ncbi.nlm.nih.gov/22622580/). DOI: 10.1038/nature11087. Access: abstract_metadata. Study: PMID22622580.

3. Kisler K, Nelson AR, Rege SV et al. Pericyte degeneration leads to neurovascular uncoupling and limits oxygen supply to brain. Nature neuroscience. 2017. [PMID 28135240](https://pubmed.ncbi.nlm.nih.gov/28135240/). DOI: 10.1038/nn.4489. Access: abstract_metadata. Study: pericyte-Pdgfrb-platform.

4. Nikolakopoulou AM, Montagne A, Kisler K et al. Pericyte loss leads to circulatory failure and pleiotrophin depletion causing neuron loss. Nature neuroscience. 2019. [PMID 31235908](https://pubmed.ncbi.nlm.nih.gov/31235908/). DOI: 10.1038/s41593-019-0434-z. Access: full_text_partial. Study: PMID31235908.

5. Daneman R, Zhou L, Kebede AA et al. Pericytes are required for blood-brain barrier integrity during embryogenesis. Nature. 2010. [PMID 20944625](https://pubmed.ncbi.nlm.nih.gov/20944625/). DOI: 10.1038/nature09513. Access: abstract_metadata. Study: PMID20944625.

6. Montagne A, Nation DA, Sagare AP et al. APOE4 leads to blood-brain barrier dysfunction predicting cognitive decline. Nature. 2020. [PMID 32376954](https://pubmed.ncbi.nlm.nih.gov/32376954/). DOI: 10.1038/s41586-020-2247-3. Access: full_text_partial. Study: PMID32376954.

7. Nation DA, Sweeney MD, Montagne A et al. Blood-brain barrier breakdown is an early biomarker of human cognitive dysfunction. Nature medicine. 2019. [PMID 30643288](https://pubmed.ncbi.nlm.nih.gov/30643288/). DOI: 10.1038/s41591-018-0297-y. Access: abstract_metadata. Study: PMID30643288.

8. Montagne A, Barnes SR, Sweeney MD et al. Blood-brain barrier breakdown in the aging human hippocampus. Neuron. 2015. [PMID 25611508](https://pubmed.ncbi.nlm.nih.gov/25611508/). DOI: 10.1016/j.neuron.2014.12.032. Access: abstract_metadata. Study: PMID25611508.

9. Halliday MR, Rege SV, Ma Q et al. Accelerated pericyte degeneration and blood-brain barrier breakdown in apolipoprotein E4 carriers with Alzheimer's disease. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism. 2016. [PMID 25757756](https://pubmed.ncbi.nlm.nih.gov/25757756/). DOI: 10.1038/jcbfm.2015.44. Access: abstract_metadata. Study: PMID25757756.

10. Blanchard JW, Bula M, Davila-Velderrain J et al. Reconstruction of the human blood-brain barrier in vitro reveals a pathogenic mechanism of APOE4 in pericytes. Nature medicine. 2020. [PMID 32514169](https://pubmed.ncbi.nlm.nih.gov/32514169/). DOI: 10.1038/s41591-020-0886-4. Access: abstract_metadata. Study: PMID32514169.

11. Montagne A, Nikolakopoulou AM, Huuskonen MT et al. APOE4 accelerates advanced-stage vascular and neurodegenerative disorder in old Alzheimer's mice via cyclophilin A independently of amyloid-β. Nature aging. 2021. [PMID 35291561](https://pubmed.ncbi.nlm.nih.gov/35291561/). DOI: 10.1038/s43587-021-00073-z. Access: full_text_partial. Study: PMID35291561.

12. Winkler EA, Nishida Y, Sagare AP et al. GLUT1 reductions exacerbate Alzheimer's disease vasculo-neuronal dysfunction and degeneration. Nature neuroscience. 2015. [PMID 25730668](https://pubmed.ncbi.nlm.nih.gov/25730668/). DOI: 10.1038/nn.3966. Access: full_text_partial. Study: PMID25730668.

13. Nortley R, Korte N, Izquierdo P et al. Amyloid β oligomers constrict human capillaries in Alzheimer's disease via signaling to pericytes. Science (New York, N.Y.). 2019. [PMID 31221773](https://pubmed.ncbi.nlm.nih.gov/31221773/). DOI: 10.1126/science.aav9518. Access: full_text_partial. Study: PMID31221773.

14. Mughal A, Sackheim AM, Koide M et al. Pathogenic soluble tau peptide disrupts endothelial calcium signaling and vasodilation in the brain microvasculature. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism. 2024. [PMID 38420777](https://pubmed.ncbi.nlm.nih.gov/38420777/). DOI: 10.1177/0271678X241235790. Access: abstract_metadata. Study: PMID38420777.

15. Chen S, Guo D, Zhu Y et al. Amyloid β oligomer induces cerebral vasculopathy via pericyte-mediated endothelial dysfunction. Alzheimer's research & therapy. 2024. [PMID 38475929](https://pubmed.ncbi.nlm.nih.gov/38475929/). DOI: 10.1186/s13195-024-01423-w. Access: abstract_metadata. Study: PMID38475929.

16. Xiong M, Wang C, Gratuze M et al. Astrocytic APOE4 removal confers cerebrovascular protection despite increased cerebral amyloid angiopathy. Molecular neurodegeneration. 2023. [PMID 36922879](https://pubmed.ncbi.nlm.nih.gov/36922879/). DOI: 10.1186/s13024-023-00610-x. Access: abstract_metadata. Study: PMID36922879.

17. Cruz Hernández JC, Bracko O, Kersbergen CJ et al. Neutrophil adhesion in brain capillaries reduces cortical blood flow and impairs memory function in Alzheimer's disease mouse models. Nature neuroscience. 2019. [PMID 30742116](https://pubmed.ncbi.nlm.nih.gov/30742116/). DOI: 10.1038/s41593-018-0329-4. Access: abstract_metadata. Study: PMID30742116.

18. Fernandez-Klett F, Brandt L, Fernández-Zapata C et al. Denser brain capillary network with preserved pericytes in Alzheimer's disease. Brain pathology (Zurich, Switzerland). 2020. [PMID 32876357](https://pubmed.ncbi.nlm.nih.gov/32876357/). DOI: 10.1111/bpa.12897. Access: abstract_metadata. Study: PMID32876357.

19. Lecordier S, Pons V, Rivest S et al. Multifocal Cerebral Microinfarcts Modulate Early Alzheimer's Disease Pathology in a Sex-Dependent Manner. Frontiers in immunology. 2021. [PMID 35173711](https://pubmed.ncbi.nlm.nih.gov/35173711/). DOI: 10.3389/fimmu.2021.813536. Access: abstract_metadata. Study: PMID35173711.

20. Gheni G, Shinohara M, Masuda-Suzukake M et al. Cerebral hypoperfusion reduces tau accumulation. Annals of clinical and translational neurology. 2025. [PMID 39621511](https://pubmed.ncbi.nlm.nih.gov/39621511/). DOI: 10.1002/acn3.52247. Access: full_text_partial. Study: PMID39621511.

21. Zhai Y, Yamashita T, Nakano Y et al. Chronic Cerebral Hypoperfusion Accelerates Alzheimer's Disease Pathology with Cerebrovascular Remodeling in a Novel Mouse Model. Journal of Alzheimer's disease : JAD. 2016. [PMID 27314529](https://pubmed.ncbi.nlm.nih.gov/27314529/). DOI: 10.3233/JAD-160345. Access: abstract_metadata. Study: PMID27314529.

22. Qiu L, Ng G, Tan EK et al. Chronic cerebral hypoperfusion enhances Tau hyperphosphorylation and reduces autophagy in Alzheimer's disease mice. Scientific reports. 2016. [PMID 27050297](https://pubmed.ncbi.nlm.nih.gov/27050297/). DOI: 10.1038/srep23964. Access: abstract_metadata. Study: PMID27050297.

23. Kitaguchi H, Tomimoto H, Ihara M et al. Chronic cerebral hypoperfusion accelerates amyloid beta deposition in APPSwInd transgenic mice. Brain research. 2009. [PMID 19646974](https://pubmed.ncbi.nlm.nih.gov/19646974/). DOI: 10.1016/j.brainres.2009.07.078. Access: abstract_metadata. Study: PMID19646974.

24. Okamoto Y, Yamamoto T, Kalaria RN et al. Cerebral hypoperfusion accelerates cerebral amyloid angiopathy and promotes cortical microinfarcts. Acta neuropathologica. 2012. [PMID 22170742](https://pubmed.ncbi.nlm.nih.gov/22170742/). DOI: 10.1007/s00401-011-0925-9. Access: abstract_metadata. Study: PMID22170742.

25. Coltman R, Spain A, Tsenkina Y et al. Selective white matter pathology induces a specific impairment in spatial working memory. Neurobiology of aging. 2011. [PMID 20961660](https://pubmed.ncbi.nlm.nih.gov/20961660/). DOI: 10.1016/j.neurobiolaging.2010.09.005. Access: abstract_metadata. Study: PMID20961660.

26. Qiu C, Albayram O, Kondo A et al. Cis P-tau underlies vascular contribution to cognitive impairment and dementia and can be effectively targeted by immunotherapy in mice. Science translational medicine. 2021. [PMID 34078745](https://pubmed.ncbi.nlm.nih.gov/34078745/). DOI: 10.1126/scitranslmed.aaz7615. Access: full_text_partial. Study: cis-tau-BCAS-htau-platform.

27. Laing KK, Simoes S, Baena-Caldas GP et al. Cerebrovascular disease promotes tau pathology in Alzheimer's disease. Brain communications. 2020. [PMID 33215083](https://pubmed.ncbi.nlm.nih.gov/33215083/). DOI: 10.1093/braincomms/fcaa132. Access: abstract_metadata. Study: PMID33215083.

28. Jackson RJ, Meltzer JC, Nguyen H et al. APOE4 derived from astrocytes leads to blood-brain barrier impairment. Brain : a journal of neurology. 2022. [PMID 34957486](https://pubmed.ncbi.nlm.nih.gov/34957486/). DOI: 10.1093/brain/awab478. Access: abstract_metadata. Study: PMID34957486.

29. Liu CC, Zhao J, Fu Y et al. Peripheral apoE4 enhances Alzheimer's pathology and impairs cognition by compromising cerebrovascular function. Nature neuroscience. 2022. [PMID 35915180](https://pubmed.ncbi.nlm.nih.gov/35915180/). DOI: 10.1038/s41593-022-01127-0. Access: abstract_metadata. Study: PMID35915180.

30. Yang AC, Vest RT, Kern F et al. A human brain vascular atlas reveals diverse mediators of Alzheimer's risk. Nature. 2022. [PMID 35165441](https://pubmed.ncbi.nlm.nih.gov/35165441/). DOI: 10.1038/s41586-021-04369-3. Access: abstract_metadata. Study: PMID35165441.

31. SPRINT MIND Investigators for the SPRINT Research Group, Williamson JD, Pajewski NM et al. Effect of Intensive vs Standard Blood Pressure Control on Probable Dementia: A Randomized Clinical Trial. JAMA. 2019. [PMID 30688979](https://pubmed.ncbi.nlm.nih.gov/30688979/). DOI: 10.1001/jama.2018.21442. Access: abstract_metadata. Study: SPRINT-cognitive-cohort.

32. Reboussin DM, Gaussoin SA, Pajewski NM et al. Long-Term Effect of Intensive vs Standard Blood Pressure Control on Mild Cognitive Impairment and Probable Dementia in SPRINT. Neurology. 2025. [PMID 39819096](https://pubmed.ncbi.nlm.nih.gov/39819096/). DOI: 10.1212/WNL.0000000000213334. Access: abstract_metadata. Study: SPRINT-cognitive-cohort.

33. He J, Zhao C, Zhong S et al. Blood pressure reduction and all-cause dementia in people with uncontrolled hypertension: an open-label, blinded-endpoint, cluster-randomized trial. Nature medicine. 2025. [PMID 40258956](https://pubmed.ncbi.nlm.nih.gov/40258956/). DOI: 10.1038/s41591-025-03616-8. Access: abstract_metadata. Study: PMID40258956.

34. Peters R, Beckett N, Forette F et al. Incident dementia and blood pressure lowering in the Hypertension in the Very Elderly Trial cognitive function assessment (HYVET-COG): a double-blind, placebo controlled trial. The Lancet. Neurology. 2008. [PMID 18614402](https://pubmed.ncbi.nlm.nih.gov/18614402/). DOI: 10.1016/S1474-4422(08)70143-1. Access: abstract_metadata. Study: PMID18614402.

35. Forette F, Seux ML, Staessen JA et al. Prevention of dementia in randomised double-blind placebo-controlled Systolic Hypertension in Europe (Syst-Eur) trial. Lancet (London, England). 1998. [PMID 9802273](https://pubmed.ncbi.nlm.nih.gov/9802273/). DOI: 10.1016/s0140-6736(98)03086-4. Access: abstract_metadata. Study: Syst-Eur-cohort.

36. Forette F, Seux ML, Staessen JA et al. The prevention of dementia with antihypertensive treatment: new evidence from the Systolic Hypertension in Europe (Syst-Eur) study. Archives of internal medicine. 2002. [PMID 12374512](https://pubmed.ncbi.nlm.nih.gov/12374512/). DOI: 10.1001/archinte.162.18.2046. Access: abstract_metadata. Study: Syst-Eur-cohort.

37. White WB, Wakefield DB, Moscufo N et al. Effects of Intensive Versus Standard Ambulatory Blood Pressure Control on Cerebrovascular Outcomes in Older People (INFINITY). Circulation. 2019. [PMID 31607143](https://pubmed.ncbi.nlm.nih.gov/31607143/). DOI: 10.1161/CIRCULATIONAHA.119.041603. Access: abstract_metadata. Study: PMID31607143.

38. Moll van Charante EP, Richard E, Eurelings LS et al. Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): a cluster-randomised controlled trial. Lancet (London, England). 2016. [PMID 27474376](https://pubmed.ncbi.nlm.nih.gov/27474376/). DOI: 10.1016/S0140-6736(16)30950-3. Access: abstract_metadata. Study: PMID27474376.

39. Tzourio C, Anderson C, Chapman N et al. Effects of blood pressure lowering with perindopril and indapamide therapy on dementia and cognitive decline in patients with cerebrovascular disease. Archives of internal medicine. 2003. [PMID 12742805](https://pubmed.ncbi.nlm.nih.gov/12742805/). DOI: 10.1001/archinte.163.9.1069. Access: abstract_metadata. Study: PMID12742805.

40. Kopczak A, Stringer MS, van den Brink H et al. Effect of blood pressure-lowering agents on microvascular function in people with small vessel diseases (TREAT-SVDs): a multicentre, open-label, randomised, crossover trial. The Lancet. Neurology. 2023. [PMID 37863608](https://pubmed.ncbi.nlm.nih.gov/37863608/). DOI: 10.1016/S1474-4422(23)00293-4. Access: abstract_metadata. Study: PMID37863608.

41. Ngandu T, Lehtisalo J, Solomon A et al. A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial. Lancet (London, England). 2015. [PMID 25771249](https://pubmed.ncbi.nlm.nih.gov/25771249/). DOI: 10.1016/S0140-6736(15)60461-5. Access: abstract_metadata. Study: PMID25771249.

42. Gottesman RF, Wu A, Coresh J et al. Associations of Vascular Risk and Amyloid Burden with Subsequent Dementia. Annals of neurology. 2022. [PMID 35732594](https://pubmed.ncbi.nlm.nih.gov/35732594/). DOI: 10.1002/ana.26447. Access: abstract_metadata. Study: ARIC-PET-overlap.

43. Gottesman RF, Schneider AL, Zhou Y et al. Association Between Midlife Vascular Risk Factors and Estimated Brain Amyloid Deposition. JAMA. 2017. [PMID 28399252](https://pubmed.ncbi.nlm.nih.gov/28399252/). DOI: 10.1001/jama.2017.3090. Access: abstract_metadata. Study: ARIC-PET-overlap.

44. Egle M, Deal JA, Walker KA et al. Association between retinal microvascular abnormalities and late-life brain amyloid-β deposition: the ARIC-PET study. Alzheimer's research & therapy. 2024. [PMID 38711107](https://pubmed.ncbi.nlm.nih.gov/38711107/). DOI: 10.1186/s13195-024-01461-4. Access: abstract_metadata. Study: ARIC-PET-overlap.

45. Snowdon DA, Greiner LH, Mortimer JA et al. Brain infarction and the clinical expression of Alzheimer disease. The Nun Study. JAMA. 1997. [PMID 9052711](https://pubmed.ncbi.nlm.nih.gov/9052711/). Access: abstract_metadata. Study: PMID9052711.

46. Schneider JA, Arvanitakis Z, Bang W et al. Mixed brain pathologies account for most dementia cases in community-dwelling older persons. Neurology. 2007. [PMID 17568013](https://pubmed.ncbi.nlm.nih.gov/17568013/). DOI: 10.1212/01.wnl.0000271090.28148.24. Access: abstract_metadata. Study: PMID17568013.

47. Sonnen JA, Larson EB, Crane PK et al. Pathological correlates of dementia in a longitudinal, population-based sample of aging. Annals of neurology. 2007. [PMID 17879383](https://pubmed.ncbi.nlm.nih.gov/17879383/). DOI: 10.1002/ana.21208. Access: abstract_metadata. Study: PMID17879383.

48. Luoto TM, Haikonen S, Haapasalo H et al. Large vessel cerebral atherosclerosis is not in direct association with neuropathological lesions of Alzheimer's disease. European neurology. 2009. [PMID 19521084](https://pubmed.ncbi.nlm.nih.gov/19521084/). DOI: 10.1159/000222779. Access: abstract_metadata. Study: PMID19521084.

49. Lorenzini L, Maranzano A, Ingala S et al. Association of Vascular Risk Factors and Cerebrovascular Pathology With Alzheimer Disease Pathologic Changes in Individuals Without Dementia. Neurology. 2024. [PMID 39288341](https://pubmed.ncbi.nlm.nih.gov/39288341/). DOI: 10.1212/WNL.0000000000209801. Access: abstract_metadata. Study: PMID39288341.

50. Kivipelto M, Ngandu T, Laatikainen T et al. Risk score for the prediction of dementia risk in 20 years among middle aged people: a longitudinal, population-based study. The Lancet. Neurology. 2006. [PMID 16914401](https://pubmed.ncbi.nlm.nih.gov/16914401/). DOI: 10.1016/S1474-4422(06)70537-3. Access: abstract_metadata. Study: PMID16914401.

51. Abell JG, Kivimäki M, Dugravot A et al. Association between systolic blood pressure and dementia in the Whitehall II cohort study: role of age, duration, and threshold used to define hypertension. European heart journal. 2018. [PMID 29901708](https://pubmed.ncbi.nlm.nih.gov/29901708/). DOI: 10.1093/eurheartj/ehy288. Access: abstract_metadata. Study: Whitehall-II.

52. Allan CL, Zsoldos E, Filippini N et al. Lifetime hypertension as a predictor of brain structure in older adults: cohort study with a 28-year follow-up. The British journal of psychiatry : the journal of mental science. 2015. [PMID 25497301](https://pubmed.ncbi.nlm.nih.gov/25497301/). DOI: 10.1192/bjp.bp.114.153536. Access: abstract_metadata. Study: Whitehall-II.

53. Smith JR, Pike JR, Gottesman RF et al. Contribution of Modifiable Midlife and Late-Life Vascular Risk Factors to Incident Dementia. JAMA neurology. 2025. [PMID 40455489](https://pubmed.ncbi.nlm.nih.gov/40455489/). DOI: 10.1001/jamaneurol.2025.1495. Access: abstract_metadata. Study: PMID40455489.

54. Pålhaugen L, Sudre CH, Tecelao S et al. Brain amyloid and vascular risk are related to distinct white matter hyperintensity patterns. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism. 2021. [PMID 32955960](https://pubmed.ncbi.nlm.nih.gov/32955960/). DOI: 10.1177/0271678X20957604. Access: abstract_metadata. Study: PMID32955960.

55. Littau JL, Velilla L, Hase Y et al. Evidence of beta amyloid independent small vessel disease in familial Alzheimer's disease. Brain pathology (Zurich, Switzerland). 2022. [PMID 35695802](https://pubmed.ncbi.nlm.nih.gov/35695802/). DOI: 10.1111/bpa.13097. Access: abstract_metadata. Study: PMID35695802.

56. Rutten JW, Hack RJ, Duering M et al. Broad phenotype of cysteine-altering NOTCH3 variants in UK Biobank: CADASIL to nonpenetrance. Neurology. 2020. [PMID 32732295](https://pubmed.ncbi.nlm.nih.gov/32732295/). DOI: 10.1212/WNL.0000000000010525. Access: abstract_metadata. Study: PMID32732295.

57. Hack RJ, Gravesteijn G, Cerfontaine MN et al. Three-tiered EGFr domain risk stratification for individualized NOTCH3-small vessel disease prediction. Brain : a journal of neurology. 2023. [PMID 36535904](https://pubmed.ncbi.nlm.nih.gov/36535904/). DOI: 10.1093/brain/awac486. Access: abstract_metadata. Study: PMID36535904.

58. Stam AH, Kothari PH, Shaikh A et al. Retinal vasculopathy with cerebral leukoencephalopathy and systemic manifestations. Brain : a journal of neurology. 2016. [PMID 27604306](https://pubmed.ncbi.nlm.nih.gov/27604306/). DOI: 10.1093/brain/aww217. Access: abstract_metadata. Study: PMID27604306.

59. Kelleher J, Dickinson A, Cain S et al. Patient-Specific iPSC Model of a Genetic Vascular Dementia Syndrome Reveals Failure of Mural Cells to Stabilize Capillary Structures. Stem cell reports. 2019. [PMID 31680059](https://pubmed.ncbi.nlm.nih.gov/31680059/). DOI: 10.1016/j.stemcr.2019.10.004. Access: abstract_metadata. Study: PMID31680059.

60. Shao W, Oliveira DV, Naia L et al. Impairment of hippocampal gamma oscillations, mitochondria and neurovascular function in CADASIL. Brain : a journal of neurology. 2026. [PMID 41622715](https://pubmed.ncbi.nlm.nih.gov/41622715/). DOI: 10.1093/brain/awag033. Access: abstract_metadata. Study: PMID41622715.

61. Shi H, Koronyo Y, Rentsendorj A et al. Identification of early pericyte loss and vascular amyloidosis in Alzheimer's disease retina. Acta neuropathologica. 2020. [PMID 32043162](https://pubmed.ncbi.nlm.nih.gov/32043162/). DOI: 10.1007/s00401-020-02134-w. Access: abstract_metadata. Study: PMID32043162.

62. Lin Z, Sur S, Liu P et al. Blood-Brain Barrier Breakdown in Relationship to Alzheimer and Vascular Disease. Annals of neurology. 2021. [PMID 34041783](https://pubmed.ncbi.nlm.nih.gov/34041783/). DOI: 10.1002/ana.26134. Access: abstract_metadata. Study: PMID34041783.

63. Edwards NC, Lao P, Alshikho MJ et al. Alzheimer Disease, Vascular Disease, and Blood-Brain Barrier Permeability Biomarkers in Middle-Aged Adults. Neurology. 2025. [PMID 41052375](https://pubmed.ncbi.nlm.nih.gov/41052375/). DOI: 10.1212/WNL.0000000000214220. Access: abstract_metadata. Study: PMID41052375.

64. Preis L, Villringer K, Brosseron F et al. Assessing blood-brain barrier dysfunction and its association with Alzheimer's pathology, cognitive impairment and neuroinflammation. Alzheimer's research & therapy. 2024. [PMID 39085945](https://pubmed.ncbi.nlm.nih.gov/39085945/). DOI: 10.1186/s13195-024-01529-1. Access: abstract_metadata. Study: PMID39085945.

65. Johnson KA, Lopera F, Jones K et al. Presenilin-1-associated abnormalities in regional cerebral perfusion. Neurology. 2001. [PMID 11402113](https://pubmed.ncbi.nlm.nih.gov/11402113/). DOI: 10.1212/wnl.56.11.1545. Access: abstract_metadata. Study: PMID11402113.

66. Julin P, Almkvist O, Basun H et al. Brain volumes and regional cerebral blood flow in carriers of the Swedish Alzheimer amyloid protein mutation. Alzheimer disease and associated disorders. 1998. [PMID 9539411](https://pubmed.ncbi.nlm.nih.gov/9539411/). DOI: 10.1097/00002093-199803000-00008. Access: abstract_metadata. Study: PMID9539411.

67. Chen J, Xiang P, Duro-Castano A et al. Rapid amyloid-β clearance and cognitive recovery through multivalent modulation of blood-brain barrier transport. Signal transduction and targeted therapy. 2025. [PMID 41052971](https://pubmed.ncbi.nlm.nih.gov/41052971/). DOI: 10.1038/s41392-025-02426-1. Access: abstract_metadata. Study: PMID41052971.

68. Lee E, Kim S, Zhu CL et al. Angiopoietin-2 aggravates Alzheimer's disease by promoting blood-brain barrier dysfunction and neuroinflammation. Cell reports. 2026. [PMID 41529684](https://pubmed.ncbi.nlm.nih.gov/41529684/). DOI: 10.1016/j.celrep.2025.116621. Access: abstract_metadata. Study: PMID41529684.

69. van Dijk SE, van der Grond J, Lak J et al. Longitudinal Progression of Magnetic Resonance Imaging Markers and Cognition in Dutch-Type Hereditary Cerebral Amyloid Angiopathy. Stroke. 2022. [PMID 35360926](https://pubmed.ncbi.nlm.nih.gov/35360926/). DOI: 10.1161/STROKEAHA.121.035826. Access: abstract_metadata. Study: PMID35360926.

## Source-specific evidence cards

### [1] Pericytes control key neurovascular functions and neuronal phenotype in the adult brain and during brain aging.

**Shows:** Pdgfrb-deficient adult mice develop vascular dysfunction before general neuronal injury.

**Does not show:** Engineered pericyte loss does not establish AD-specific proteinopathy onset; linked correction details unexamined.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID37797582. Subclaims: SC2, SC3.

### [2] Apolipoprotein E controls cerebrovascular integrity via cyclophilin A.

**Shows:** APOE4 activates pericyte CypA-MMP9 and genetic or pharmacological pathway subtraction protects BBB.

**Does not show:** APOE affects multiple compartments; BBB protection is not spontaneous human AD prevention; correction details unexamined.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID37142785. Subclaims: SC2, SC3.

### [3] Pericyte degeneration leads to neurovascular uncoupling and limits oxygen supply to brain.

**Shows:** Pericyte-deficient mice lose neurovascular coupling and develop oxygen/metabolic stress.

**Does not show:** Shared Pdgfrb platform and physiology do not independently prove amyloid/tau initiation.

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

### [4] Pericyte loss leads to circulatory failure and pleiotrophin depletion causing neuron loss.

**Shows:** Inducible pericyte ablation causes flow/BBB deficits and neuronal/cognitive loss; physiological PTN infusion preserves neurons and cognition despite persistent circulatory defects.

**Does not show:** Severe acute dual-promoter ablation is not gradual human disease; amyloid/tau initiation is not tested; PTN depletion alone does not kill neurons.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: 1:RESULTS:paragraph2, 1:RESULTS:paragraph3, 1:RESULTS:paragraph4, 1:RESULTS:paragraph10, 1:RESULTS:paragraph12, 1:RESULTS:paragraph13. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

### [5] Pericytes are required for blood-brain barrier integrity during embryogenesis.

**Shows:** Developmental pericyte coverage determines BBB permeability before astrocyte development.

**Does not show:** Developmental barrier formation is not adult AD etiology or clinical trajectory.

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

### [6] APOE4 leads to blood-brain barrier dysfunction predicting cognitive decline.

**Shows:** Human APOE4-associated BBB disruption and baseline pericyte-injury marker predict subsequent cognitive decline independently of measured amyloid/tau.

**Does not show:** No vascular manipulation; cognitively normal APOE4 participants can already have amyloid; baseline marker prediction is not molecular primacy.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: 0:Introductory Paragraph:paragraph1, 0:Introductory Paragraph:paragraph3, 0:Introductory Paragraph:paragraph5, 0:Introductory Paragraph:paragraph6, 0:Introductory Paragraph:paragraph9. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC3.

### [7] Blood-brain barrier breakdown is an early biomarker of human cognitive dysfunction.

**Shows:** Early cognitive dysfunction associates with hippocampal capillary damage irrespective of measured amyloid/tau biomarkers.

**Does not show:** Cross-sectional independence does not establish a cascade from vascular injury to proteins.

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

### [8] Blood-brain barrier breakdown in the aging human hippocampus.

**Shows:** Hippocampal BBB permeability increases with aging and mild cognitive impairment.

**Does not show:** Imaging association lacks a selective counterfactual or AD-specific initiating event.

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

### [9] Accelerated pericyte degeneration and blood-brain barrier breakdown in apolipoprotein E4 carriers with Alzheimer's disease.

**Shows:** APOE4 postmortem brain has pericyte loss and CypA pathway activation.

**Does not show:** Terminal tissue cannot order the initiating lesions; genotype is pleiotropic.

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

### [10] Reconstruction of the human blood-brain barrier in vitro reveals a pathogenic mechanism of APOE4 in pericytes.

**Shows:** Human 3D BBB APOE4 models show NFAT-associated cerebral amyloid angiopathy with inhibitor sensitivity.

**Does not show:** Culture and accompanying models do not establish BBB failure preceding all proteinopathy; correction details unexamined.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID33514950. Subclaims: SC3.

### [11] APOE4 accelerates advanced-stage vascular and neurodegenerative disorder in old Alzheimer's mice via cyclophilin A independently of amyloid-β.

**Shows:** At 10–12 months Debio-025 improves APOE4 BBB and prevents further neuronal loss and cognitive decline without changing Aβ pathology.

**Does not show:** Treatment modifies the APOE4 increment rather than restoring all 5xFAD deficits to wild type; APP mutations precede the vascular insult.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: 0:Results:paragraph11, 0:Results:paragraph12, 0:Results:paragraph15, 0:Results:paragraph16, 0:Results:paragraph18. No complete-paper or supplement audit claimed. ErratumIn: PMID37117809. Subclaims: SC2, SC3.

### [12] GLUT1 reductions exacerbate Alzheimer's disease vasculo-neuronal dysfunction and degeneration.

**Shows:** Endothelial GLUT1 loss causes early BBB failure; combined Slc2a1 deficiency/APP increases later Aβ and NeuN loss; GLUT1/LRP1 re-expression reduces Aβ.

**Does not show:** APP overexpression is present from development; late acceleration and regional clearance rescue are not initiation of spontaneous human AD or whole-animal cognitive reversal.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: 1:Results:paragraph9, 1:Results:paragraph11, 1:Results:paragraph12, 1:Results:paragraph18, 1:Results:paragraph22, 1:Results:paragraph23. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

### [13] Amyloid β oligomers constrict human capillaries in Alzheimer's disease via signaling to pericytes.

**Shows:** Applied Aβ constricts human/rat capillaries; APP knock-in mice show endogenous plaque-region pericyte constriction; CNP reverses acute constriction.

**Does not show:** Small human biopsy and mouse samples; thousands of vessel measurements are not independent patients; modeled flow reduction is not measured human flow.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: 0:Amyloid β constricts human capillaries at pericytes, 2:Pericytes constrict capillaries in human cognitive decline patients with Aβ deposition, 4:Pericytes constrict capillaries in vivo in AD mice, 6:Reversal of Aβ-evoked capillary constriction. No complete-paper or supplement audit claimed. none_found. Subclaims: SC3.

### [14] Pathogenic soluble tau peptide disrupts endothelial calcium signaling and vasodilation in the brain microvasculature.

**Shows:** AD-related tau peptides cause endothelial calcium and vasodilatory dysfunction.

**Does not show:** Peptide exposure demonstrates a possible protein-to-vessel route, not its prevalence in spontaneous AD.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. UpdateOf: PMID37609200. Subclaims: SC3.

### [15] Amyloid β oligomer induces cerebral vasculopathy via pericyte-mediated endothelial dysfunction.

**Shows:** Aβ exposure injures pericytes and endothelial capillary structure.

**Does not show:** Exogenous challenge establishes capacity, not ordinary endogenous initiating direction.

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

### [16] Astrocytic APOE4 removal confers cerebrovascular protection despite increased cerebral amyloid angiopathy.

**Shows:** Early astrocytic APOE4 removal in 5xFAD reduces vascular injury despite increased vascular amyloid.

**Does not show:** Genetic APP proteinopathy persists and cell-specific APOE effects are not a vessel-only intervention.

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

### [17] Neutrophil adhesion in brain capillaries reduces cortical blood flow and impairs memory function in Alzheimer's disease mouse models.

**Shows:** Anti-Ly6G relieves neutrophil capillary stalls and rapidly improves flow and memory in APP/PS1 and 5xFAD mice.

**Does not show:** Acute functional rescue occurs after engineered proteinopathy and is not prevention of AD onset.

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

### [18] Denser brain capillary network with preserved pericytes in Alzheimer's disease.

**Shows:** Two human frontal-cortex cohorts show preserved pericytes and increased capillary density in AD.

**Does not show:** Regional postmortem counterexample does not rule out hippocampal or functional vascular injury.

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

### [19] Multifocal Cerebral Microinfarcts Modulate Early Alzheimer's Disease Pathology in a Sex-Dependent Manner.

**Shows:** Microinfarcts produce sex-dependent cognitive injury while decreasing plaques without changing soluble Aβ in APP/PS1 mice.

**Does not show:** Surgical infarcts and strain/sex differences constrain generalization; lower plaques are not lower overall injury.

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

### [20] Cerebral hypoperfusion reduces tau accumulation.

**Shows:** BCAS performed two months after tau seeding reduces accumulated tau while worsening white matter; NACC remote stroke associates with less AD pathology.

**Does not show:** Seeded tau is imposed first; CTSD colocalization/upregulation is a candidate mediator without necessity subtraction; autopsy associations retain survival and diagnosis biases.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: 2:Results:paragraph0, 2:Results:paragraph11, 2:Results:paragraph13, 2:Results:paragraph14, 2:Results:paragraph18. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

### [21] Chronic Cerebral Hypoperfusion Accelerates Alzheimer's Disease Pathology with Cerebrovascular Remodeling in a Novel Mouse Model.

**Shows:** Hypoperfusion in APP23 exacerbates amyloid/cognitive abnormalities and galantamine ameliorates phenotypes.

**Does not show:** APP is genetically upstream; galantamine has nonvascular neuronal actions.

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

### [22] Chronic cerebral hypoperfusion enhances Tau hyperphosphorylation and reduces autophagy in Alzheimer's disease mice.

**Shows:** Unilateral occlusion raises selected phospho-tau epitopes in aged wild-type/3xTg mice without changing Aβ42.

**Does not show:** Age/genotype-specific phosphorylation is not a complete AD amyloid/tau initiating cascade.

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

### [23] Chronic cerebral hypoperfusion accelerates amyloid beta deposition in APPSwInd transgenic mice.

**Shows:** BCAS in APP-transgenic mice increases intracellular and fibrillar Aβ deposits.

**Does not show:** Transgenic APP is upstream, and accelerated deposition does not identify sporadic onset.

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

### [24] Cerebral hypoperfusion accelerates cerebral amyloid angiopathy and promotes cortical microinfarcts.

**Shows:** BCAS augments vascular amyloid and microinfarcts in CAA mice; human postmortem observations agree.

**Does not show:** Amyloid vasculopathy is pre-existing; terminal association and hypoperfusion amplification are not vascular-first AD initiation.

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

### [25] Selective white matter pathology induces a specific impairment in spatial working memory.

**Shows:** Chronic hypoperfusion produces white-matter injury and selective spatial working-memory deficits.

**Does not show:** Excluding mice with neuronal perikaryal injury conditions analysis on an outcome; impairment is not necessarily AD dementia.

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

### [26] Cis P-tau underlies vascular contribution to cognitive impairment and dementia and can be effectively targeted by immunotherapy in mice.

**Shows:** BCAS induces endogenous cis P-tau; cis antibody and DAPK1 loss preserve white matter and executive function; separate htau mice respond to treatment.

**Does not show:** Human VaD brains lack overt NFTs; htau AD-like tangles arise in a separate model, so BCAS-to-classic-AD initiation is not demonstrated.

**Access/status:** Complete primary abstract and linked-status metadata examined. Selected body sections: 1:RESULTS:paragraph0, 1:RESULTS:paragraph1, 1:RESULTS:paragraph2, 1:RESULTS:paragraph6, 1:RESULTS:paragraph7, 1:RESULTS:paragraph8. No complete-paper or supplement audit claimed. none_found. Subclaims: SC2, SC3.

### [27] Cerebrovascular disease promotes tau pathology in Alzheimer's disease.

**Shows:** White-matter vascular pathology associates with total tau/Braak measures; aged MCAO produces phospho-tau.

**Does not show:** Mixed observational/ischemic evidence does not show onset of the specific amyloid/tau cascade or exclude injury-related tau release.

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

### [28] APOE4 derived from astrocytes leads to blood-brain barrier impairment.

**Shows:** Astrocytic APOE4 deletion ameliorates BBB impairment relative to matched APOE3 designs.

**Does not show:** APOE has nonvascular effects; barrier rescue does not establish that amyloid/tau initiation depends on barrier failure.

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

### [29] Peripheral apoE4 enhances Alzheimer's pathology and impairs cognition by compromising cerebrovascular function.

**Shows:** Liver-only APOE4 causes vascular/synaptic/cognitive abnormalities and exacerbates APP-model amyloid; APOE3 plasma improves aged phenotypes.

**Does not show:** Systemic pleiotropy and APP crossing distinguish vascular contribution from AD-specific primacy.

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

### [30] A human brain vascular atlas reveals diverse mediators of Alzheimer's risk.

**Shows:** Single-nucleus human vascular atlas identifies AD-associated cellular changes and risk-gene expression.

**Does not show:** Risk-gene expression is not a causal vascular instrument; terminal transcriptomes cannot order events.

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

### [31] Effect of Intensive vs Standard Blood Pressure Control on Probable Dementia: A Randomized Clinical Trial.

**Shows:** SPRINT intensive blood-pressure treatment reduces MCI and combined cognitive endpoints.

**Does not show:** Primary probable-dementia contrast is nonsignificant; early cardiovascular stopping and exclusion of diabetes/stroke limit scope; no AD biomarker-onset endpoint.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [32] Long-Term Effect of Intensive vs Standard Blood Pressure Control on Mild Cognitive Impairment and Probable Dementia in SPRINT.

**Shows:** SPRINT extended follow-up retains MCI/composite benefit.

**Does not show:** Probable dementia remains nonsignificant; only 59% of eligible participants were ascertained; same randomized cohort and unexamined correction do not add independent replication.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID40315397. Subclaims: SC1, SC2.

### [33] Blood pressure reduction and all-cause dementia in people with uncontrolled hypertension: an open-label, blinded-endpoint, cluster-randomized trial.

**Shows:** Rural-China cluster randomization of 33995 adults in 326 villages lowers 48-month incident all-cause dementia with provider-led blood-pressure treatment.

**Does not show:** Bundled care is not a selective BBB/pericyte intervention; all-cause dementia benefit does not demonstrate prevention of AD-specific amyloid/tau; abstract/status depth only.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2, SC3.

### [34] Incident dementia and blood pressure lowering in the Hypertension in the Very Elderly Trial cognitive function assessment (HYVET-COG): a double-blind, placebo controlled trial.

**Shows:** HYVET-COG randomized antihypertensive treatment in adults over 80 and measured incident dementia.

**Does not show:** Short follow-up primary dementia contrast is nonsignificant; meta-analysis pooling is not a positive primary trial result.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [35] Prevention of dementia in randomised double-blind placebo-controlled Systolic Hypertension in Europe (Syst-Eur) trial.

**Shows:** Syst-Eur randomized antihypertensive therapy yields fewer clinical dementia cases over two years.

**Does not show:** Only 32 events and borderline significance; clinical classification does not establish biomarker-defined AD prevention.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [36] The prevention of dementia with antihypertensive treatment: new evidence from the Systolic Hypertension in Europe (Syst-Eur) study.

**Shows:** Syst-Eur extended follow-up reports continued dementia incidence difference including clinical AD cases.

**Does not show:** Control participants were offered active treatment after blinded phase; same initial cohort, not an independent randomized trial or proof of molecular onset prevention.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMIDnot supplied. Subclaims: SC1, SC2.

### [37] Effects of Intensive Versus Standard Ambulatory Blood Pressure Control on Cerebrovascular Outcomes in Older People (INFINITY).

**Shows:** INFINITY lowers white-matter hyperintensity accumulation with intensive ambulatory blood-pressure control.

**Does not show:** Gait and cognition contrasts are null despite MRI effects; biomarkers cannot stand for preserved clinical trajectory.

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

### [38] Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): a cluster-randomised controlled trial.

**Shows:** preDIVA vascular-care cluster trial shows no overall incident dementia or disability reduction.

**Does not show:** Low baseline risk and strong usual care reduce intervention contrast; null bundled care is not selective pathway falsification.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [39] Effects of blood pressure lowering with perindopril and indapamide therapy on dementia and cognitive decline in patients with cerebrovascular disease.

**Shows:** PROGRESS reduces cognitive decline with perindopril-based therapy after stroke/TIA; dementia benefit concentrates with recurrent stroke.

**Does not show:** Overall dementia contrast is null and benefit without recurrent stroke is not established; this supports stroke-related injury prevention rather than obligatory AD cascade interruption.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [40] Effect of blood pressure-lowering agents on microvascular function in people with small vessel diseases (TREAT-SVDs): a multicentre, open-label, randomised, crossover trial.

**Shows:** TREAT-SVD crossover compares cerebrovascular reactivity under three drugs; CADASIL shows differential physiological effects.

**Does not show:** No drug difference in sporadic SVD; four-week physiology is not functional or dementia prevention; small analyzed subgroups.

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

### [41] A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): a randomised controlled trial.

**Shows:** FINGER multidomain intervention improves a neuropsychological composite over two years.

**Does not show:** Diet, exercise and cognitive training accompany vascular monitoring; benefit cannot be assigned to vascular correction or AD onset prevention.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [42] Associations of Vascular Risk and Amyloid Burden with Subsequent Dementia.

**Shows:** ARIC hypertension and amyloid independently predict dementia without detected multiplicative synergy.

**Does not show:** Absence of multiplicative interaction is not absence of additive interaction; observational hazards cannot order the pathology.

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

### [43] Association Between Midlife Vascular Risk Factors and Estimated Brain Amyloid Deposition.

**Shows:** Multiple midlife vascular risks associate with later brain amyloid while late-life risks do not.

**Does not show:** Long interval supports prediction but not an unconfounded vascular-to-amyloid initiating mechanism; same ARIC platform.

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

### [44] Association between retinal microvascular abnormalities and late-life brain amyloid-β deposition: the ARIC-PET study.

**Shows:** ARIC retinal features at midlife do not predict amyloid; late-life continuous SUVr association differs from binary positivity.

**Does not show:** Retinal proxies and endpoint definitions constrain inference; same cohort is not independent replication.

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

### [45] Brain infarction and the clinical expression of Alzheimer disease. The Nun Study.

**Shows:** Nun Study infarcts lower the burden of AD lesions associated with clinical dementia.

**Does not show:** Small autopsy sample and wide intervals; interaction in clinical expression does not imply infarcts initiated plaques/tangles.

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

### [46] Mixed brain pathologies account for most dementia cases in community-dwelling older persons.

**Shows:** Rush autopsies show mixed pathologies increase probability of clinical dementia.

**Does not show:** Postmortem mixture supports additive injury but cannot establish chronological initiation.

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

### [47] Pathological correlates of dementia in a longitudinal, population-based sample of aging.

**Shows:** ACT autopsies identify independent contributions of microinfarcts, Braak pathology and Lewy bodies to dementia.

**Does not show:** Selected autopsy fraction and weighting retain possible selection/confounding; independent association is not random intervention.

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

### [48] Large vessel cerebral atherosclerosis is not in direct association with neuropathological lesions of Alzheimer's disease.

**Shows:** Large-artery atherosclerosis in 466 autopsies does not associate with AD plaques/tangles.

**Does not show:** Large-vessel pathology differs from BBB microvascular dysfunction; negative subtype result does not exclude all vascular contributions.

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

### [49] Association of Vascular Risk Factors and Cerebrovascular Pathology With Alzheimer Disease Pathologic Changes in Individuals Without Dementia.

**Shows:** EPAD structural-equation analysis fits vascular risk/SVD/amyloid/tau/cognitive pathways in nondemented participants.

**Does not show:** Statistical mediation and fitted ordering do not establish causality, temporal precedence, or absence of alternative models.

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

### [50] Risk score for the prediction of dementia risk in 20 years among middle aged people: a longitudinal, population-based study.

**Shows:** Midlife CAIDE factors predict late-life dementia over about 20 years.

**Does not show:** Predictive discrimination is not causal mediation or a validated preventable fraction.

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

### [51] Association between systolic blood pressure and dementia in the Whitehall II cohort study: role of age, duration, and threshold used to define hypertension.

**Shows:** Whitehall II blood pressure at age 50 predicts dementia, unlike measurements at 60 or 70.

**Does not show:** Age-dependent association retains reverse causation and selection; CVD-free association is not a selective vascular instrument.

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

### [52] Lifetime hypertension as a predictor of brain structure in older adults: cohort study with a 28-year follow-up.

**Shows:** Longitudinal midlife blood pressure relates to later white-matter lesions.

**Does not show:** Hippocampal/global atrophy associations are not established; tissue injury proxy is not molecular AD initiation.

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

### [53] Contribution of Modifiable Midlife and Late-Life Vascular Risk Factors to Incident Dementia.

**Shows:** ARIC estimates substantial vascular-risk-associated dementia fractions before age 80 and smaller fractions at older ages.

**Does not show:** Population-attributable fractions assume causal and intervention relationships; they are not demonstrated preventable percentages.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. none_found. Subclaims: SC1, SC2.

### [54] Brain amyloid and vascular risk are related to distinct white matter hyperintensity patterns.

**Shows:** Amyloid-related and vascular-risk-related white-matter patterns differ topographically.

**Does not show:** Cross-sectional topography cannot order cause and effect or validate a universal two-hit sequence.

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

### [55] Evidence of beta amyloid independent small vessel disease in familial Alzheimer's disease.

**Shows:** PSEN1 familial AD has SVD changes that are not explained solely by CAA and can be covert on MRI.

**Does not show:** Known PSEN1 lesion precedes the vascular phenotype; familial tissue does not establish sporadic vascular initiation.

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

### [56] Broad phenotype of cysteine-altering NOTCH3 variants in UK Biobank: CADASIL to nonpenetrance.

**Shows:** NOTCH3 variants produce heterogeneous SVD/vascular phenotypes in population carriers.

**Does not show:** Many imaged carriers remain normal into late life; penetrance and selection prevent universal dementia inference.

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

### [57] Three-tiered EGFr domain risk stratification for individualized NOTCH3-small vessel disease prediction.

**Shows:** NOTCH3 domain location stratifies CADASIL stroke/SVD risk across clinical and population samples.

**Does not show:** Genetic risk strata are not intervention outcomes; phenotype/penetrance varies and AD amyloid/tau onset is not tested.

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

### [58] Retinal vasculopathy with cerebral leukoencephalopathy and systemic manifestations.

**Shows:** TREX1 families develop systemic retinal/cerebral vasculopathy with cognitive decline in a substantial subset.

**Does not show:** Rare familial systemic disease is a distinct vascular cause, not universal AD dementia or a vessel-only mediator instrument.

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

### [59] Patient-Specific iPSC Model of a Genetic Vascular Dementia Syndrome Reveals Failure of Mural Cells to Stabilize Capillary Structures.

**Shows:** CADASIL patient-derived mural cells destabilize capillary networks with VEGF-sensitive rescue.

**Does not show:** In vitro networks do not establish organismal cognitive trajectory or specific endogenous human rescue.

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

### [60] Impairment of hippocampal gamma oscillations, mitochondria and neurovascular function in CADASIL.

**Shows:** NOTCH3 R182C mice and human tissue show vascular-smooth-muscle metabolic and neuronal activity abnormalities.

**Does not show:** Mechanism-associated physiology without selective rescue or AD-specific protein onset cannot close the nominal chain.

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

### [61] Identification of early pericyte loss and vascular amyloidosis in Alzheimer's disease retina.

**Shows:** AD retinal tissue shows pericyte loss related to amyloid.

**Does not show:** Retinal pathology is not interchangeable with brain BBB function or chronological AD origin.

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

### [62] Blood-Brain Barrier Breakdown in Relationship to Alzheimer and Vascular Disease.

**Shows:** Water-permeability BBB measures associate with AD biomarkers while albumin permeability tracks vascular risks.

**Does not show:** Assays measure different transport/barrier properties; association cannot assign a common causal mechanism.

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

### [63] Alzheimer Disease, Vascular Disease, and Blood-Brain Barrier Permeability Biomarkers in Middle-Aged Adults.

**Shows:** Midlife vascular/VEGF biomarker paths associate with AD-related measures.

**Does not show:** Cross-sectional path analysis cannot validate the article's proposed causal ordering.

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

### [64] Assessing blood-brain barrier dysfunction and its association with Alzheimer's pathology, cognitive impairment and neuroinflammation.

**Shows:** DELCODE BBB permeability increases in dementia but not MCI; sPDGFRβ has no overall diagnostic/APOE difference.

**Does not show:** Small sample and divergent permeability/injury assays; age/inflammation associations and null amyloid/tau-positivity relations limit a general initiating claim.

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

### [65] Presenilin-1-associated abnormalities in regional cerebral perfusion.

**Shows:** Presymptomatic PS1 mutation carriers have cerebral perfusion abnormalities.

**Does not show:** Presymptomatic means before clinical symptoms, not before molecular proteinopathy; genotype is already an initiating lesion.

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

### [66] Brain volumes and regional cerebral blood flow in carriers of the Swedish Alzheimer amyloid protein mutation.

**Shows:** Small APP Swedish family shows flow changes before marked atrophy in some carriers.

**Does not show:** Five carriers with youngest participants normal; no demonstration that vascular change precedes amyloid production.

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

### [67] Rapid amyloid-β clearance and cognitive recovery through multivalent modulation of blood-brain barrier transport.

**Shows:** LRP1-targeted transport therapy clears amyloid and improves cognition in AD models.

**Does not show:** Treatment after existing pathology tests clearance and delivery, not whether vascular failure initiated AD.

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

### [68] Angiopoietin-2 aggravates Alzheimer's disease by promoting blood-brain barrier dysfunction and neuroinflammation.

**Shows:** Endothelial ANGPT2 loss reduces amyloid, while overexpression worsens BBB/inflammatory/memory phenotypes in 5xFAD.

**Does not show:** Genetic APP/PS1 pathology is upstream; expression/phenotype rescue does not demonstrate ordinary human vascular-first onset.

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

### [69] Longitudinal Progression of Magnetic Resonance Imaging Markers and Cognition in Dutch-Type Hereditary Cerebral Amyloid Angiopathy.

**Shows:** Longitudinal hereditary CAA shows progressive vascular-reactivity change before clinical symptoms.

**Does not show:** Mutant APP is upstream; no parallel CBF/cognitive change in presymptomatic carriers, and clinical timing does not equal pre-proteinopathy timing.

**Access/status:** Complete primary abstract and linked-status metadata examined. No complete-paper or supplement audit claimed. ErratumIn: PMID35605042. Subclaims: SC1, SC3.