Skip to content

Cerebral small-vessel disease

TL;DR — Cerebral small-vessel disease (SVD) is the dominant substrate of insidious vascular cognitive impairment, but it is a heterogeneous set of arteriolar, capillary, and venular processes rather than an MRI synonym (Wardlaw 2013, PMID 23867200; Duering 2023, PMID 37236211). Lacunes, WMH, microbleeds, enlarged perivascular spaces, recent small subcortical infarcts, and atrophy are standardized visible markers; microinfarcts, endothelial dysfunction, and blood–brain barrier leakage are incompletely visible (Duering 2023, PMID 37236211). Cognition is usually affected through distributed disconnection and network inefficiency, often alongside gait and neuropsychiatric change (Desmond 2004, PMID 15537510; Clancy 2021, PMID 33539776). Marker burden is probabilistic, not diagnostic, and frequently coexists with Alzheimer pathology.

Pathological families

Family Vessel process Typical clues
Arteriolosclerosis wall thickening, lipohyalinosis, lumen narrowing hypertension, deep lacunes/WMH
CAA amyloid-β in cortical/leptomeningeal vessels lobar microbleeds, siderosis (Charidimou 2022, PMID 35841910)
Inflammatory/immune vessel inflammation or CAA-related inflammation subacute symptoms, edema/enhancement
Genetic arteriopathy NOTCH3, HTRA1, COL4A1/2 and others early onset/family pattern (Choi 2015, PMID 25692103)
Venous/collagen processes altered drainage/barrier structure selected inherited or systemic phenotypes

STRIVE-visible features

Marker Working imaging definition Cognitive relevance Main caveat
Recent small subcortical infarct acute perforator-territory lesion focal injury plus network disruption clinical lacunar syndrome may mismatch
Lacune CSF-like cavity from prior small infarct/haemorrhage accumulated tissue loss distinguish perivascular space
WMH T2/FLAIR hyperintensity processing speed/executive association common and nonspecific
Microbleed small susceptibility focus arteriopathy and haemorrhage risk sequence-dependent detection
Cortical superficial siderosis surface haemosiderin strongly supports CAA pattern focal trauma/SAH alternatives
Perivascular spaces fluid spaces along penetrating vessels candidate clearance/vascular marker rating variability
Brain atrophy reduced brain volume integrates cumulative injury not etiologically specific

STRIVE created the shared vocabulary in 2013; STRIVE-2 updated acquisition, quantitative markers, and links to neurodegeneration in 2023 (Wardlaw 2013, PMID 23867200; Duering 2023, PMID 37236211).

How SVD affects cognition

  1. Repeated small infarcts destroy tissue.
  2. Diffuse white-matter damage disconnects cortical–subcortical networks.
  3. Endothelial and BBB dysfunction alter the tissue environment.
  4. Impaired autoregulation increases vulnerability to pressure fluctuations.
  5. Reduced reserve amplifies the effect of Alzheimer and other pathology.

The resulting group phenotype favors slowed processing and executive dysfunction, but domain patterns overlap broadly with neurodegenerative disease (Desmond 2004, PMID 15537510). A systematic review of 23 prospective studies/11,486 cognitively healthy older adults linked sporadic SVD to decline in global cognition (effect −0.10, 95% CI −0.14 to −0.05), executive function (−0.18, −0.24 to −0.11), memory (−0.12, −0.17 to −0.07), and attention (−0.17, −0.23 to −0.11); WMH associations were larger than those for microbleeds (Jansma 2024, PMID 38415688). Debette and Markus meta-analysed 22 longitudinal studies: WMH associated with stroke (HR 3.3, 2.6–4.4), dementia (HR 1.9, 1.3–2.8), and death (HR 2.0, 1.6–2.7) (Debette 2010, PMID 20660506). In LADIS (n=639; 3-year follow-up), 90 participants developed dementia and 147 cognitive impairment without dementia; baseline diabetes independently predicted decline after adjustment for age, education, white-matter-change severity, and temporal atrophy (Verdelho 2010, PMID 20625169).

Wardlaw et al. frame SVD as a microvessel disorder whose visible MRI lesions are the late, incomplete sample of endothelial dysfunction, BBB leakage, impaired vasodilation, stiffening, and impaired interstitial drainage (Wardlaw 2019, PMID 31097385; Pantoni 2010, PMID 20610345).

Does SVD progression cause dementia, or accompany it?

The strongest observational case for causality comes from RUN DMC, which followed 503 people with sporadic SVD and no dementia for a median 13.2 years, with dementia status available for 498 (99.0%). Dementia occurred in 108 (21.5%) — 38 Alzheimer, 34 vascular, 26 mixed. Independently associated with all-cause and vascular dementia were baseline WMH volume (HR 1.31 per SD, 95% CI 1.02–1.67), the presence of DWI-positive lesions (HR 2.03, 95% CI 1.01–4.04), and peak width of skeletonized mean diffusivity (PSMD; HR 1.24 per SD, 95% CI 1.02–1.51). Crucially, WMH progression itself predicted incident dementia (HR 1.76 per SD, 95% CI 1.18–2.63), establishing temporal ordering that a cross-sectional association cannot (Jacob 2023, PMID 37073486).

Genetic evidence agrees on WMH and disagrees on almost everything else. Two-sample Mendelian randomization using AD GWASs with up to 75,024 cases found genetically determined WMH burden associated with Alzheimer disease (OR 1.43, 95% CI 1.10–1.86, after accounting for pulse pressure) and with an AD-meta phenotype (OR 1.19, 95% CI 1.06–1.34) — yet in the same paper's longitudinal individual-level analysis of 10,699 incident all-cause dementia cases, genetically determined WMH showed no association (HR 1.02, 95% CI 1.00–1.04; P=0.07), and blood-pressure traits appeared spuriously protective against AD through shared instruments (Sargurupremraj 2024, PMID 38776079). Illness-death modelling excluded selective survival as the explanation. A genetic instrument that is causal in the summary-statistics frame and null in the individual-level frame is an unresolved methodological contradiction, not a settled finding, and it directly limits how confidently WMH can be used as a trial endpoint (see OQ-7).

Where genetics converges is on which risk factors to target. Across GIGASTROKE lacunar stroke (n=6,811) and imaging GWASs (WMH n=55,291; fractional anisotropy n=36,460; mean diffusivity n=36,012), genetically predicted systolic and diastolic blood pressure, diabetes liability, obesity, smoking, and triglycerides (and the triglyceride/HDL ratio) were detrimental, while HDL and moderate-to-vigorous physical activity were protective — and LDL cholesterol was not (Koohi 2025, PMID 39661645). Drug-target proxies pointed specifically to calcium-channel blockade for imaging markers, and to CETP inhibition, lipoprotein-lipase enhancement, and GIP-receptor obesity drugs for lacunar stroke. A separate MR study using WMH (n=18,381), microbleed, perivascular-space, and lacunar-stroke GWASs found systolic BP raised basal-ganglia perivascular spaces (OR 1.05, 95% CI 1.04–1.07), hippocampal perivascular spaces (OR 1.04, 95% CI 1.02–1.05), and lacunar stroke (OR 1.41, 95% CI 1.30–1.54); genetically proxied calcium-channel blockade lowered WMH volume (β −0.287, 95% CI −0.408 to −0.165) and basal-ganglia perivascular spaces (OR 0.85, 95% CI 0.81–0.89) (Ma 2024, PMID 38818733). Two independent genetic analyses nominating calcium-channel blockers is the same signal that Syst-Eur produced in a randomized trial three decades earlier (see prevention) — a convergence that has still not been tested prospectively.

The markers that do not behave as expected

Enlarged perivascular spaces. Harmonized rating across five population cohorts in the UNIVRSE consortium (3,575 people aged 63–73) found total ePVS counts unrelated to MMSE (mean difference per unit 0.001, 95% CI −0.007 to 0.008) and to a general fluid cognitive factor (0.002, 95% CI −0.001 to 0.006), unchanged by adjustment for cardiovascular risk factors or other MRI markers, and unchanged in region-specific analyses (Hilal 2018, PMID 30068634). A STRIVE marker with no detectable cross-sectional cognitive correlate in the general population should not be summed into a total-burden score as if it carried the same weight as a lacune. Except that longitudinally it does behave like one. In 414 community-dwelling adults aged 72–92 assessed biennially for 8 years with 3T MRI, severe PVS pathology (top quartile counts) in both basal ganglia and centrum semiovale independently predicted incident dementia over 8 years (OR 2.91, 95% CI 1.43–5.95, p=0.003) and greater 4-year global cognitive decline, after adjustment for WMH volume, microbleed count and lacune count; centrum-semiovale PVS alone also predicted decline, while basal-ganglia PVS behaved differently — a dissociation the authors read as different underlying pathologies (Paradise 2021, PMID 33504642).

A third study splits the difference and points at the measurement. In 120 patients with lacunar stroke and confluent WMH scanned annually for 3 years and tested annually for 5, PVS volume predicted nothing — not SVD marker progression, not conversion to dementia, whether measured at baseline or as 3-year change — while DTI-ALPS declined over time and predicted dementia strongly (HR 0.328, 95% CI 0.183–0.588, P<0.001), surviving adjustment for median mean diffusivity (HR 0.290, 0.139–0.602). But the association between change in DTI-ALPS and dementia lost significance once PSMD and median MD were entered, which the authors take to mean part of the ALPS signal is conventional diffusion pathology wearing a glymphatic label (Hong 2024, PMID 38465597).

The three results are compatible under one reading: counted perivascular spaces carry prognostic information at the extremes and over long horizons (Paradise), carry none cross-sectionally at population average burden (Hilal), and carry none as a segmented volume in an already-severe SVD cohort (Hong) — while the diffusion index that is supposed to measure their function outperforms all of them and may not be measuring them at all. What has never been done is a head-to-head of PVS count, PVS volume and DTI-ALPS in one cohort against a pathological reference.

Cortical microinfarcts. Neuropathology finds them in roughly a third of SVD brains, yet almost none are visible on structural MRI. The RUN DMC-Intense study scanned 54 people with SVD monthly for a median 39.5 weeks (472 DWI scans): chronic cortical microinfarcts were present in 35% (95% CI 24–49%) at baseline, and 21 acute microinfarcts appeared in 7 participants (cumulative incidence 13%, 95% CI 6–24%) — every one of which disappeared on follow-up MRI (ter Telgte 2020, PMID 32065609). The lesions accumulating in pathology series are therefore largely invisible to any cross-sectional scan, which means the imaging-visible SVD burden systematically undercounts ischaemic injury by an unknown factor. How large that undercount is can be bounded from pathology. A systematic review of 32 neuropathological studies covering 10,515 people found microinfarcts in a weighted average of 62% of vascular dementia, 43% of Alzheimer disease, 33% of mixed Alzheimer-plus-cerebrovascular dementia, and 24% of non-demented older individuals, frequently multiple per brain, distributed across all regions but concentrated in cortex and watershed territories, and ranging from 50 μm to a few millimetres — the upper end of which is within reach of high-resolution MRI (Brundel 2012, PMID 22234334). A lesion present in nearly two-thirds of vascular dementia brains and largely absent from the images used to diagnose it is the single clearest example of the gap between the pathological and the radiological definitions of this condition. Microinfarcts have heterogeneous causes spanning small- and large-vessel disease, atrial fibrillation, and heart failure, and mechanisms including vascular remodelling, inflammation, BBB leakage, venule congestion, hypoperfusion, and microembolism; their effects extend beyond the lesion core to perilesional and global cortical atrophy (Huang 2024, PMID 37470314).

How large is the WMH–cognition effect, and is it domain-specific?

Meta-analysis across 23 cross-sectional (n=8,685) and 14 longitudinal (n=7,731) studies, restricted to those matching or adjusting for age, sex and education, found WMH presence associated with deficits in every examined domain, all small: general intelligence (Fisher z −0.10, 95% CI −0.19 to −0.04), memory (−0.08, −0.13 to −0.06), processing speed (−0.11, −0.17 to −0.07), attention and executive function (−0.11, −0.16 to −0.07), and perception/construction (−0.15, −0.21 to −0.07). WMH progression carried effects roughly three times larger than WMH presence — general intelligence −0.31 (−0.50 to −0.02) and attention/executive function −0.32 (−0.34 to −0.28) (Kloppenborg 2014, PMID 24814849).

Two things follow. First, the "executive/processing-speed signature" is a difference of degree, not of kind: the effect on memory and perception is nearly as large, which is why the profile cannot partition pathology (see assessment). Second, change matters more than level — consistent with the MarkVCID finding that WMH growth has a cognitive correlate while regression does not (see biomarkers) — which is an argument for progression-based rather than threshold-based enrolment in trials.

Why marker burden explains so little variance: reserve

The lesion–symptom correlation in SVD is persistently modest, and part of the residual is host factors rather than measurement error. In 615 LADIS participants evaluated with MRI and neuropsychology at 3 years and re-interviewed at up to 7 years, higher educational and occupational attainment predicted slower decline in processing speed, executive function and memory independently of WMH volume, and moderated the effect of WMH on processing speed and memory; education attenuated the relationship between WMH volume and 7-year cognitive status, and both proxies predicted sustained functional independence and lower mortality at 7 years (Jokinen 2016, PMID 27951523). This has a direct trial consequence that is rarely acted on: if reserve moderates the marker–outcome slope, then a trial that balances arms on WMH volume but not on education has balanced the wrong variable.

Progression, atrophy and disability

WMH growth is not an isolated white-matter process. In 99 patients with symptomatic small-vessel disease followed ≥1 year with voxel-wise rate-of-change mapping, the rate of WMH expansion and the rate of grey-matter atrophy correlated at Pearson r = −0.69 (P<1×10⁻⁷), with significant annual grey-matter and whole-brain loss. WMH growth was regionally heterogeneous, fastest within long association fasciculi, and faster WMH progression was associated with higher cortical grey-matter atrophy rates in medial-frontal, orbitofrontal, parietal and occipital regions (family-wise-error corrected P<0.05); grey-matter loss, not white-matter loss, was the main contributor to whole-brain atrophy (Lambert 2016, PMID 26936939). The implication the authors draw — that slowing WMH growth might protect against secondary grey-matter atrophy — is a mechanistic hypothesis, not a demonstrated effect.

Functionally, the disability consequence is large and dose-dependent. In LADIS, 639 non-disabled older outpatients graded mild, moderate or severe on the Fazekas scale were followed a mean 2.42 years for transition from independence to disability (instrumental ADL ≥2) or death. Annual rates were 10.5%, 15.1% and 29.5% across the three severity strata (log-rank P<0.001), with severe versus mild giving an adjusted HR 2.36 (95% CI 1.65–3.81) — independent of baseline atrophy and infarct count, and only slightly attenuated by incident stroke and dementia (Inzitari 2009, PMID 19581317). Other independent predictors were age, atrial fibrillation and complaint of gait disturbance. This is the strongest available evidence that white-matter change matters for daily life rather than only for test scores, and it does not run through dementia.

Perivascular spaces and the clearance hypothesis

Perivascular spaces are the STRIVE marker whose biology is least settled. The comprehensive review positions them as passageways around arterioles, capillaries and venules along which interstitial fluid and waste move — particularly during sleep — and relates their MRI morphology to cognition, vascular risk factors, lesions, sleep patterns and cerebral haemodynamics, while being explicit that much remains unknown and controversial (Wardlaw 2020, PMID 32094487).

Longitudinal data give the marker a temporal ordering it otherwise lacks. In 289 SVD participants (median age 67.0, 42.9% women) with 220 rescanned after a median 3.4 years, a latent-change-score model found baseline DTI-ALPS predicted change in basal-ganglia perivascular-space volume (β = −0.09, P=0.030) but not the reverse (β = −0.08, P=0.110) — a directional asymmetry consistent with impaired perivascular fluid movement preceding space enlargement. Baseline BG-PVS volume predicted growth in WMH volume (β=0.33) and lacune count (β=0.28); free-water fraction predicted WMH, lacunes and brain-volume change (β=−0.45) and decline in MMSE (β=−0.17), cognitive index (β=−0.25) and processing speed (β=−0.29); DTI-ALPS predicted WMH and brain-volume change and cognitive decline in the opposite direction (all p-corrected ≤0.002) (Li 2024, PMID 38626373). This ordering is consistent with clearance-related measures preceding lesion accumulation, but an observational latent-change model does not prove that sequence. DTI-ALPS is a diffusion index, not a measurement of flow, and calling it "glymphatic function" outruns what it measures.

Genetics beyond monogenic disease

WMH burden is highly heritable and polygenic. Multiethnic GWAS meta-analysis of 21,079 dementia- and stroke-free individuals across 29 cohorts (17,936 European, 1,943 African, 795 Hispanic, 405 Asian ancestry) confirmed the known chr17q25 locus (P=2.7×10⁻¹⁹) and identified novel loci at chr10q24 (P=1.6×10⁻⁹) and chr2p21 (P=4.4×10⁻⁸) in European samples, plus chr1q22 (P=2.0×10⁻⁸) and chr2p16 (P=1.5×10⁻⁸) in the combined multiethnic analysis. The implicated genes span Alzheimer disease (chr2p21, chr10q24), intracerebral haemorrhage (chr1q22), neuroinflammatory disease (chr2p21) and glioma (chr10q24, chr2p16) — pointing to inflammatory and glial-proliferative pathways alongside the assumed ischaemic ones (Verhaaren 2015, PMID 25663218). The non-European samples were too small to power ancestry-specific discovery, which remains the main gap.

Sample size has since transformed the picture. A GWAS of three SVD MRI markers in UK Biobank and other sources — WMH (n=42,310), fractional anisotropy (n=17,663) and mean diffusivity (n=17,467) — identified 31 loci, 21 of them previously unreported, and a transcriptome-wide association study implicated 66 genes across the three traits (Persyn 2020, PMID 32358547). The jump from 5 loci in 21,079 people to 31 in 42,310 indicates the trait is highly polygenic and that discovery was power-limited rather than saturated; it also means locus-counting comparisons between papers of different eras carry no information about biology.

Beyond cognition

Manifestation Candidate relationship Evidence boundary
Gait slowing/falls white-matter network disruption association varies by gait measure (Blumen 2023, PMID 35797006)
Apathy/depression frontal-subcortical circuits WMH–apathy OR 1.41 (1.05–1.89); delirium OR 2.9 (1.12–7.55) (Clancy 2021, PMID 33539776)
Urinary symptoms distributed subcortical disease nonspecific in older adults
Stroke lacunar infarction and ICH mechanism differs by arteriopathy
Delirium vulnerability reduced reserve under-standardized

Total-burden scores

Simple scores count major MRI features and improve communication, but equal weighting implies equivalence that biology does not guarantee. A single new lacune, extensive WMH progression, and numerous lobar microbleeds have different mechanisms and risks. STRIVE-2 therefore encourages feature-specific and quantitative measures where feasible (Duering 2023, PMID 37236211).

Approach Advantage Limitation
Fazekas WMH grade rapid and widely understood ceiling effects, ordinal
Feature count easy total-burden summary equal weights
Lesion segmentation continuous volume scanner/algorithm dependence
Diffusion metrics detects normal-appearing white matter injury protocol harmonization
Network analysis mechanistic link to cognition computational and validation burden
Longitudinal change measures progression registration and attrition bias

Equal weighting is theoretically indefensible and empirically works anyway. In 1,152 Framingham Heart Study participants over 55 free of dementia, stroke and other neurological disease at MRI, a simple additive score counting microbleeds, covert infarcts, extensive WMH, high-burden perivascular spaces and cortical superficial siderosis identified 211 (18%) with a score ≥2; over a median 7.4 years those participants had HR 1.67 (95% CI 1.05–2.66) for all-cause dementia adjusted for the Framingham Stroke Risk Profile and HR 1.76 (1.10–2.81) adjusted for vascular risk factors individually, with Harrell c-statistics of 0.82–0.83 — statistically indistinguishable from the FSRP itself (Pinheiro 2025, PMID 40953349). Two readings compete. Either the unweighted score is capturing a genuine common pathway that makes feature-specific weighting unnecessary, or it performs at FSRP level because it is largely a proxy for the same vascular risk burden, in which case an MRI adds prediction that a questionnaire already provides. The study cannot distinguish these, and it did not validate externally.

Computer-aided extraction is expanding, but a systematic review found heterogeneity in validation and reporting that limits transportability (Jiang 2022, PMID 35914668).

Risk factors and prevention

Hypertension is the leading modifiable exposure; diabetes, smoking, dyslipidaemia, sleep and activity patterns may contribute through overlapping vascular routes (Gorelick 2011, PMID 21778438). The ESO covert-SVD guideline recommends risk-factor management while noting limited direct randomized evidence for many cognitive endpoints (Wardlaw 2021, PMID 34414301).

Antiplatelet treatment should not be inferred from WMH alone: preventing atherothrombotic events and treating covert SVD are different indications. Similarly, very intensive blood-pressure lowering may have net vascular benefit in selected populations but must be individualized where orthostatic symptoms or frailty threaten perfusion and falls (Hughes 2020, PMID 32427305; Bhanu 2021, PMID 34752479).

Trial design

FINESSE proposes mechanistically coherent populations, harmonized imaging endpoints, and clinically meaningful outcomes for SVD trials (Markus 2022, PMID 35969390).

The mechanistic-target problem is illustrated by MINERVA, a phase-II experimental-medicine trial in which 44 people with moderate-to-severe SVD took minocycline or placebo for 3 months against co-primary outcomes of microglial signal (¹¹C-PK11195 PET) and BBB permeability (dynamic contrast-enhanced MRI). Both abnormalities were confirmed present in SVD, and minocycline changed neither (microglial binding RR 1.01, 95% CI 0.98–1.04; BBB permeability RR 0.97, 95% CI 0.91–1.03), nor serum inflammatory markers (Brown 2024, PMID 38629936). The trial's own conclusion — that whether these mechanisms are disease-causing remains unclear — is the honest position: a null experimental-medicine trial with a drug that worked in rodents leaves the neuroinflammation hypothesis untested rather than refuted, but it also demonstrates that advanced imaging endpoints are feasible enough to falsify mechanistic claims within 3 months.

Design choice Preferred practice
Entry phenotype specify symptomatic lacunar stroke, covert SVD, or cognitive syndrome
Imaging common STRIVE definitions and central QC
Endpoint pair imaging progression with cognition/function
Duration sufficient for measurable progression
Copathology measure or stratify Alzheimer biomarkers where feasible
Safety stroke, ICH, falls, syncope, renal events

Open questions

  • Which quantitative MRI marker is a valid surrogate for cognitive benefit? (Duering 2023, PMID 37236211)
  • Are enlarged perivascular spaces causal, compensatory, or merely correlated with injury? (Duering 2023, PMID 37236211)
  • Can monogenic SVD provide efficient proof-of-mechanism for sporadic disease? (Whittaker 2022, PMID 35699195)
  • Which blood-pressure strategy best balances SVD prevention against orthostatic harm? (Wardlaw 2021, PMID 34414301)
  • Why is genetically determined WMH causal for dementia in two-sample Mendelian randomization but null in individual-level longitudinal analysis of the same construct? (Sargurupremraj 2024, PMID 38776079)
  • Should enlarged perivascular spaces be weighted in total-burden scores at all, given no cognitive association across five harmonized population cohorts? (Hilal 2018, PMID 30068634)
  • By how much does the disappearance of acute cortical microinfarcts on MRI cause imaging-based SVD burden to undercount ischaemic injury? (ter Telgte 2020, PMID 32065609; Huang 2024, PMID 37470314)
  • Do the two independent genetic nominations of calcium-channel blockade for SVD imaging markers justify a dedicated randomized trial? (Koohi 2025, PMID 39661645; Ma 2024, PMID 38818733)
  • Do perivascular-space counts predict dementia (OR 2.91 over 8 years) while perivascular-space volumes do not, and if so which is the biologically meaningful measurement? (Paradise 2021, PMID 33504642; Hong 2024, PMID 38465597; Hilal 2018, PMID 30068634)
  • Does an unweighted multimarker SVD score add dementia prediction beyond a vascular risk questionnaire, or merely restate it? (Pinheiro 2025, PMID 40953349)
  • If microinfarcts are present in 62% of vascular dementia brains and largely invisible on MRI, can any imaging-based severity scale be calibrated? (Brundel 2012, PMID 22234334; ter Telgte 2020, PMID 32065609)
  • Should SVD trials stratify on education and occupation, given that reserve moderates the WMH–cognition slope independently of lesion volume? (Jokinen 2016, PMID 27951523)
  • Are microglial activation and BBB leakage in SVD causal or epiphenomenal, given that suppressing neither was achieved by minocycline? (Brown 2024, PMID 38629936)
  • If WMH progression has a threefold larger cognitive effect than WMH presence, should trials enrol on progression rather than on burden? (Kloppenborg 2014, PMID 24814849)
  • Does slowing WMH growth prevent the correlated grey-matter atrophy (r = −0.69), or are both downstream of a shared process? (Lambert 2016, PMID 26936939)
  • Why does severe leukoaraiosis triple the annual rate of transition to disability (29.5% vs 10.5%) largely independently of stroke and dementia? (Inzitari 2009, PMID 19581317)
  • Does impaired perivascular clearance precede perivascular-space enlargement, as the directional latent-change result suggests, or is DTI-ALPS measuring something else? (Li 2024, PMID 38626373; Wardlaw 2020, PMID 32094487)
  • Do the inflammatory and glial-proliferative loci at chr2p21 and chr10q24 indicate a non-ischaemic route to white-matter injury? (Verhaaren 2015, PMID 25663218)

References

  1. Wardlaw JM, et al. Neuroimaging standards for research into small vessel disease. Lancet Neurol. 2013. PMID 23867200
  2. Duering M, et al. Neuroimaging standards for research into small vessel disease—advances since 2013. Lancet Neurol. 2023. PMID 37236211
  3. Charidimou A, et al. Boston criteria version 2.0 for cerebral amyloid angiopathy. Lancet Neurol. 2022. PMID 35841910
  4. Choi JC. Genetics of cerebral small vessel disease. J Stroke. 2015. PMID 25692103
  5. Desmond DW. The neuropsychology of vascular cognitive impairment. J Neurol Sci. 2004. PMID 15537510
  6. Clancy U, et al. Neuropsychiatric symptoms associated with cerebral small vessel disease. Lancet Psychiatry. 2021. PMID 33539776
  7. Jansma A, et al. Sporadic cerebral small vessel disease and cognitive decline in healthy older adults. J Cereb Blood Flow Metab. 2024. PMID 38415688
  8. Debette S, Markus HS. The clinical importance of white matter hyperintensities on brain MRI: systematic review and meta-analysis. BMJ. 2010;341:c3666. PMID 20660506
  9. Verdelho A, et al. Differential impact of cerebral white matter changes, diabetes, hypertension and stroke on cognitive performance among non-disabled elderly: the LADIS study. J Neurol Neurosurg Psychiatry. 2010;81:1325-30. PMID 20625169
  10. Wardlaw JM, et al. Small vessel disease: mechanisms and clinical implications. Lancet Neurol. 2019;18:684-696. PMID 31097385
  11. Pantoni L. Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol. 2010;9:689-701. PMID 20610345
  12. Jacob MA, et al. Cerebral small vessel disease progression and the risk of dementia: a 14-year follow-up study. Am J Psychiatry. 2023;180:508-518. PMID 37073486
  13. Sargurupremraj M, et al. Genetic complexities of cerebral small vessel disease, blood pressure, and dementia. JAMA Netw Open. 2024;7:e2412824. PMID 38776079
  14. Koohi F, et al. Association of vascular risk factors and drug targets with cerebral small vessel disease: a Mendelian randomization study. Stroke. 2025;56:e37-e47. PMID 39661645
  15. Ma Y, et al. Blood pressure, antihypertensive drug targets and cerebral small vessel disease: a Mendelian randomization study. J Cereb Blood Flow Metab. 2024. PMID 38818733
  16. Hilal S, et al. Enlarged perivascular spaces and cognition: a meta-analysis of 5 population-based studies. Neurology. 2018;91:e832-e842. PMID 30068634
  17. Paradise M, et al. Association of dilated perivascular spaces with cognitive decline and incident dementia. Neurology. 2021;96:e1501-e1511. PMID 33504642
  18. Hong H, et al. Relationship of perivascular space markers with incident dementia in cerebral small vessel disease. Stroke. 2024;55:1032-1040. PMID 38465597
  19. ter Telgte A, et al. Contribution of acute infarcts to cerebral small vessel disease progression. Ann Neurol. 2020;87:264-274. PMID 32065609
  20. Huang J, et al. Cerebral microinfarcts revisited: detection, causes, and clinical relevance. Int J Stroke. 2024;19:7-15. PMID 37470314
  21. Brundel M, et al. Cerebral microinfarcts: a systematic review of neuropathological studies. J Cereb Blood Flow Metab. 2012;32:425-36. PMID 22234334
  22. Kloppenborg RP, et al. Presence and progression of white matter hyperintensities and cognition: a meta-analysis. Neurology. 2014;82:2127-38. PMID 24814849
  23. Lambert C, et al. Longitudinal patterns of leukoaraiosis and brain atrophy in symptomatic small vessel disease. Brain. 2016;139:1136-51. PMID 26936939
  24. Inzitari D, et al. Changes in white matter as determinant of global functional decline in older independent outpatients: LADIS. BMJ. 2009;339:b2477. PMID 19581317
  25. Jokinen H, et al. Cognitive reserve moderates long-term cognitive and functional outcome in cerebral small vessel disease. J Neurol Neurosurg Psychiatry. 2016;87:1296-1302. PMID 27951523
  26. Wardlaw JM, et al. Perivascular spaces in the brain: anatomy, physiology and pathology. Nat Rev Neurol. 2020;16:137-153. PMID 32094487
  27. Li Y, et al. Perivascular fluid dynamics and small vessel disease progression: a latent change score analysis. Alzheimers Dement. 2024. PMID 38626373
  28. Verhaaren BF, et al. Multiethnic genome-wide association study of cerebral white matter hyperintensities on MRI. Circ Cardiovasc Genet. 2015;8:398-409. PMID 25663218
  29. Persyn E, et al. Genome-wide association study of MRI markers of cerebral small vessel disease in 42,310 participants. Nat Commun. 2020;11:2175. PMID 32358547
  30. Blumen HM, et al. Gait in cerebral small vessel disease, pre-dementia, and dementia. Int J Stroke. 2023. PMID 35797006
  31. Pinheiro A, et al. Multimarker cerebral small vessel disease score and risk of incident dementia in the Framingham Heart Study. Neurology. 2025;105:e214113. PMID 40953349
  32. Jiang J, et al. Computer-aided extraction of select MRI markers of cerebral small vessel disease. Neuroimage. 2022. PMID 35914668
  33. Gorelick PB, et al. Vascular contributions to cognitive impairment and dementia. Stroke. 2011. PMID 21778438
  34. Wardlaw JM, et al. ESO guideline on covert cerebral small vessel disease. Eur Stroke J. 2021;6:CXI-CLXII. PMID 34414301
  35. Hughes D, et al. Association of blood pressure lowering with incident dementia or cognitive impairment. JAMA. 2020. PMID 32427305
  36. Bhanu C, et al. Drug-induced orthostatic hypotension: a systematic review and meta-analysis. PLoS Med. 2021;18:e1003821. PMID 34752479
  37. Markus HS, et al. Framework for Clinical Trials in Cerebral Small Vessel Disease (FINESSE). JAMA Neurol. 2022. PMID 35969390
  38. Brown RB, et al. Minocycline in cerebral small vessel disease: MINERVA. J Neurol Neurosurg Psychiatry. 2024. PMID 38629936
  39. Whittaker E, et al. Cerebral phenotypes associated with monogenic cerebral small-vessel disease. J Am Heart Assoc. 2022. PMID 35699195