Skip to content

Subtypes and clinical syndromes

TL;DR — Vascular cognitive impairment is a family of syndromes rather than one disease: overt or recurrent stroke, subcortical small-vessel disease, strategic infarction, haemorrhagic disease, and hypoperfusion can each produce different trajectories (O'Brien 2015, PMID 26595643; Gorelick 2011, PMID 21778438). “Stepwise decline” is neither required nor universal; diffuse small-vessel injury commonly produces gradual progression (Sachdev 2014, PMID 24632990). Lesion location, accumulated burden, brain reserve, and copathology determine the phenotype more reliably than older labels such as “multi-infarct dementia” (Jellinger 2008, PMID 18525132). Classification should therefore name the vascular substrate, cognitive severity, and confidence of attribution.

Quantitative anchors for how common each syndrome is in published cohorts: any post-stroke NCD 53.4% (Barbay 2018, PMID 30504699); dementia after recurrent stroke 41.3% versus 7.4% after first-ever stroke with prior dementia excluded (Pendlebury 2009, PMID 19782001); MCI in 47% of SPS3 lacunar-stroke participants (Jacova 2012, PMID 23034910); mixed AD+infarcts in 38.0% of a community dementia autopsy series versus 12% “vascular dementia alone” (Schneider 2007, PMID 17568013).

Syndrome map

Syndrome Dominant substrate Typical course Diagnostic hazard
Post-stroke cognitive disorder Clinical infarct or haemorrhage acute decline, partial recovery, later change pre-stroke impairment/delirium (Pendlebury 2009, PMID 19782001)
Multi-infarct pattern Multiple cortical/subcortical infarcts cumulative or stepwise label omits location and copathology
Subcortical ischaemic VCI arteriolosclerosis, WMH, lacunes gradual executive/gait syndrome common incidental WMH (Duering 2023, PMID 37236211)
Strategic-infarct syndrome single network-critical lesion abrupt focal/multidomain deficit small lesion can be missed on CT
Hypoperfusion-related injury watershed infarction or global ischaemia acute/subacute, sometimes delayed coexisting embolism/systemic illness
Haemorrhagic VCI ICH, microbleeds, superficial siderosis event-related or cumulative CAA vs hypertensive arteriopathy
Mixed vascular–degenerative vascular plus Alzheimer/other pathology often gradual with superimposed steps forced single label (Schneider 2007, PMID 17568013)

Post-stroke syndrome

Post-stroke impairment may be detected acutely, emerge as delirium resolves, improve through recovery, or progress with recurrent vascular injury and neurodegeneration. Across cohorts, prevalence depends strongly on whether pre-stroke dementia and recurrent stroke are included (Pendlebury 2009, PMID 19782001). Hospital studies estimate 53.4% for any post-stroke NCD, 36.4% mild, and 16.5% major, but test thresholds and case mix dominate comparisons (Barbay 2018, PMID 30504699).

Phase Main explanatory candidates Assessment implication
Hyperacute lesion, hypoperfusion, delirium, aphasia avoid definitive dementia label
Early weeks recovery, depression, fatigue, medication repeat domain-sensitive testing
3–6 months residual deficit plus premorbid reserve establish stable functional baseline
Later years recurrent stroke, SVD progression, copathology reassess etiology after change

Multi-infarct and recurrent-stroke disease

The historic multi-infarct concept correctly recognized cumulative tissue injury but incompletely captures silent infarcts, diffuse microvascular damage, and strategic effects. Recurrent stroke is associated with much higher dementia prevalence: more than one-third in the Pendlebury synthesis (Pendlebury 2009, PMID 19782001). The number of infarcts alone is insufficient; volume, laterality, connectivity, and reserve matter (O'Brien 2015, PMID 26595643).

Stroke subtype also predicts cognitive trajectory independently of severity. In STROKOG's latent-class analysis of 1,149 patients across nine cohorts, membership in the lowest-performing cognitive trajectory group versus the highest was predicted by large-artery versus small-vessel stroke (relative risk ratio 2.77, 95% CI 1.32–5.83) as well as moderate/severe stroke (RRR 3.17, 95% CI 1.42–7.08), diabetes (RRR 3.78, 95% CI 2.08–6.88), age, and education (Lo 2023, PMID 37072222). Beyond the first year, recurrent stroke was one of only two factors that accelerated ongoing decline, the other being older age (Lo 2022, PMID 34775838). See post-stroke cognitive impairment.

The subtype problem in pathology: single lesions may not count

The clinical taxonomy above assumes each substrate is separately consequential. Autopsy data challenge that assumption directly. Among 1,474 Rush cohort decedents (mean ~88 years at death), 80% had some cerebrovascular pathology; of these, 37% had a single cerebrovascular pathology and 63% had mixed profiles across 32 possible combinations of three vessel diseases (atherosclerosis, arteriolosclerosis, CAA) and two tissue injuries (macroinfarcts, microinfarcts). Only the mixed-profile group declined faster than participants with no cerebrovascular pathology; the single-pathology group did not, and combinations involving both atherosclerosis and arteriolosclerosis were the worst, with distinct profiles mapping onto different cognitive domains (Lamar 2022, PMID 34601898).

This is the pathological counterpart to the clinical syndrome map: a patient labelled "subcortical ischaemic VCI" on the basis of arteriolosclerosis alone may carry a substrate that, in isolation, does not measurably accelerate decline. The syndromic labels remain useful for communication and for choosing a management pathway; they should not be assumed to identify a single sufficient cause. See mixed pathology.

Subcortical ischaemic VCI

This pattern links lacunes and confluent white-matter injury to slowed processing, executive dysfunction, apathy, gait disorder, and urinary symptoms. WMH confer HR 3.3 (2.6–4.4) for stroke, 1.9 (1.3–2.8) for dementia, and 2.0 (1.6–2.7) for death in longitudinal meta-analysis (Debette 2010, PMID 20660506). STRIVE-2 cautions that visible markers overlap with aging and neurodegeneration; whole-brain quantitative measures and longitudinal change may capture disease better than a single ordinal score (Duering 2023, PMID 37236211). In SPS3, 47% of recent lacunar-stroke participants already had MCI (Jacova 2012, PMID 23034910).

Feature Supporting interpretation Limitation
Deep lacunes perforator-territory tissue loss may be clinically silent
Periventricular/deep WMH cumulative white-matter injury nonspecific
Deep microbleeds hypertensive arteriopathy pattern mixed distributions occur
Gait–executive coupling distributed frontal-subcortical disruption musculoskeletal confounding
Gradual progression ongoing diffuse disease mimics neurodegeneration

Strategic infarction

A small lesion can produce disproportionate cognitive change if it interrupts a network hub. Commonly discussed sites include thalamic nuclei, angular gyrus, basal forebrain, caudate, and medial temporal structures, but “strategic” should be supported by timing and phenotype rather than location alone (Jellinger 2002, PMID 12417375).

The list of strategic sites has now been derived rather than asserted. Multivariate lesion-symptom mapping in 410 patients with acute ischaemic stroke, using MoCA at 3–6 months and two assumption-free analyses (voxel-level and region-of-interest support-vector regression), converged on three structures for global impairment: the left angular gyrus, the left basal ganglia, and the white matter surrounding the left basal ganglia; each cognitive domain mapped to its own overlapping cortical-subcortical network (Zhao 2018, PMID 28895445). Two features of that result matter for how "strategic infarct" should be used. First, the strongly left-lateralized global-cognition result is at least partly an artifact of testing in language: an instrument that samples verbal fluency, naming, and verbal memory will find left-hemisphere lesions strategic. Second, the authors' own conclusion — that models need thousands rather than hundreds of patients — is the honest limit; a three-structure map from 410 patients is a starting point, not a lookup table, and no lesion-location model has yet been shown to predict which patients progress to dementia.

Strategic site Reported association Caveat
left angular gyrus global MoCA impairment (Zhao 2018, PMID 28895445) overlaps language-dependent test items
left basal ganglia and surrounding white matter global MoCA impairment (Zhao 2018, PMID 28895445) includes both grey-nuclear and disconnection effects
thalamus, basal forebrain, caudate, medial temporal classic clinical descriptions (Jellinger 2002, PMID 12417375) derived from case series, not mapped multivariately

Strategic locations are not only infarcts, and not only stroke cohorts

Two extensions of the strategic-lesion idea have been mapped since, and both broaden it.

Strategic lacunes in a population sample. In 1,230 dementia-free rural Chinese adults aged ≥60 in the MIND-China MRI substudy (mean age 69.4, 58.5% women), lacunes were present in 357 and MCI in 286 (243 amnestic, 43 non-amnestic). Supratentorial, internal-capsule, putamen/pallidum and insular lacunes were associated with MCI and with amnestic MCI (adjusted ORs 1.40–3.36), while infratentorial and brainstem lacunes were associated with non-amnestic MCI (ORs 2.68–3.46); the insula, internal capsule, infratentorial and brainstem associations held independently of WMH volume and perivascular-space count (Wang 2024, PMID 38511349). Two things are notable: the strategic effect is measurable below the dementia threshold in a community sample rather than only after clinical stroke, and brainstem lacunes — never on the classical strategic list — carried the strongest non-amnestic association.

Strategic white-matter hyperintensity locations. Harmonizing individual patient data from nine ischaemic stroke cohorts through the Meta VCI Map consortium (1,568 patients, 39.9% female, mean age 67.3), WMH volume in the left anterior thalamic radiation was associated with attention/executive function and information-processing speed, and WMH volume in the forceps major with processing speed — with lasso-regression coefficients larger than that of total WMH volume, independent of age, sex, education and total infarct volume (Coenen 2024, PMID 38651756). The concept of a strategic lesion therefore extends to pre-existing, non-acute, non-infarct pathology, and a tract-specific WMH measure outperforms the whole-brain volume that trials use as their endpoint.

The thalamus is four syndromes, not one

"Thalamic infarct" is the strategic lesion most often named and the one where the vascular territory determines the phenotype most sharply. The thalamic nuclei fall into five functional classes — reticular/intralaminar (arousal, nociception), sensory, effector (motor, aspects of language), associative (high-level cognition) and limbic (mood, motivation) — supplied by four arterial territories that destroy them in different combinations.

Territory Cognitive/behavioural syndrome
Tuberothalamic impaired arousal and orientation, learning and memory, personality and executive function; superimposition of temporally unrelated information; emotional facial paresis; perseveration and apathy
Paramedian decreased arousal (particularly if bilateral), impaired learning and memory, disinhibition and personality change, loss of self-activation; thalamic "dementia" with extensive lesions
Inferolateral contralateral hemisensory loss, hemiparesis, hemiataxia, pain syndromes; executive dysfunction that is often overlooked and can cause severe long-term disability
Posterior choroidal visual-field deficits, variable sensory loss, weakness, dystonia, tremor; occasionally amnesia and language impairment

Left-sided tuberothalamic and paramedian lesions produce language deficits; right-sided lesions in the same territories produce visuospatial deficits including hemispatial neglect (Schmahmann 2003, PMID 12933968). The paramedian syndrome is the one most often misread: disinhibition, personality change and loss of self-activation without focal neurological signs are difficult to distinguish from a primary psychiatric disorder, and the inferolateral executive syndrome is routinely missed because attention goes to the sensory findings (Carrera 2006, PMID 16801643). A small subcortical infarct presenting as a behavioural change rather than a cognitive one is the clearest illustration of why the "stepwise decline" stereotype fails.

The gait-and-cognition syndrome is less general than assumed

The subcortical syndrome is usually described as executive impairment plus gait disorder plus urinary symptoms. The gait component does not travel with lesion burden as reliably as that description implies. In 200 patients with minor stroke (NIHSS ≤7) scored 0–4 on the total SVD burden score at presentation and assessed at 3 years, total SVD burden was not associated with gait impairment in the whole cohort or in the 87 lacunar-stroke patients; it was associated with worse self-reported mobility only in the 113 non-lacunar stroke patients (B = −4.61, 95% CI −8.42 to −0.79, P<0.05), and only on the subjective measure rather than the timed-up-and-go (Loos 2018, PMID 28906203). The sample is modest and the finding needs replication, but a null in exactly the subgroup where the syndrome is supposed to live is worth stating rather than smoothing.

Global or regional hypoperfusion can produce watershed infarcts and selective neuronal injury. Chronic low flow is biologically plausible in severe large-vessel stenosis or impaired autoregulation, but observational perfusion abnormalities do not prove that raising pressure improves cognition (Santisteban 2023, PMID 36129176). Treatment must distinguish a reversible haemodynamic state from established tissue loss.

One human test of the chronic-hypoperfusion hypothesis is asymptomatic carotid stenosis, where flow may be reduced without any clinical infarct. Comparing 82 patients with ≥50% asymptomatic stenosis (no prior ipsilateral or contralateral stroke or TIA) against 62 controls matched for vascular comorbidity, education, estimated intelligence, and depressive symptoms, the stenosis group had worse composite cognition (P=0.02, Cohen's d=0.43), learning/memory (d=0.42), and motor/processing speed (d=0.65), with 49.4% impaired in at least two domains (Lal 2017, PMID 28712815). Cognition tracked cerebrovascular reserve — patients with a reduced breath-holding index scored worse on composite cognition (d=0.53) and learning/memory (d=0.66) — and did not correlate with degree of stenosis, least luminal diameter, plaque area, or plaque echogenicity. This supports, but does not establish, a haemodynamic explanation. Among 786 baseline assessments from 1,000 consecutive CREST-2 participants (≥70% asymptomatic stenosis, no prior stroke, mean age 70, 58% men), cognitive Z-scores standardized against the population-based REGARDS cohort for age, race and education — and further adjusted for hypertension, diabetes, dyslipidaemia and smoking — were significantly below expected at the 50th, 75th and 95th percentiles (all P<0.0001), driven largely by Word List Recall (P<0.0001 at every percentile) and Word List Learning. Left- and right-sided stenosis gave similar scores (Lazar 2021, PMID 34433306, NCT02089217). The memory-predominant deficit and absence of laterality fit a diffuse or shared-risk mechanism better than a simple ipsilateral-hypoperfusion account. Both studies are cross-sectional; neither is decisive.

What the study cannot show is reversibility: no randomized trial has demonstrated that revascularizing an asymptomatic stenosis improves cognition. CREST-2 (NCT02089217) is the trial positioned to answer it.

The cardiac route

Hypoperfusion need not be focal or arterial. Cognitive impairment is common in heart failure: pooled across the available literature, prevalence was 41.4% for cognitive impairment and 19.8% for dementia in heart-failure patients, rising with age on meta-regression (Yap 2022, PMID 35839985). Whether heart failure accelerates decline or shares its risk factors is unsettled by prevalence data alone.

The Rotterdam Study addressed that by studying people before clinical cardiac disease. Among 3,291 participants aged 58–98 free of coronary disease, heart failure, atrial fibrillation, stroke and dementia at echocardiography, followed for 21,785 (stroke) and 19,462 (dementia) person-years, better diastolic function (higher E/A ratio) predicted lower risk of both stroke (HR 0.82, 95% CI 0.69–0.98) and dementia (HR 0.82, 0.70–0.96), while better systolic function (fractional shortening) predicted lower stroke risk only (HR 0.84, 0.72–0.98). In the MRI subset (n=577), better diastolic function was associated with fewer silent infarcts, especially lacunar (de Bruijn 2015, PMID 25632093). Subclinical diastolic dysfunction is therefore a candidate upstream exposure for lacunar disease specifically — and one that current vascular-risk models do not measure.

The embolic side of the cardiac route has been measured directly, including events occurring despite baseline anticoagulation. In Swiss-AF, 1,227 atrial-fibrillation patients (mean age 71, 26.1% women, 89.9% anticoagulated) had standardized brain MRI at baseline and 2 years. Only 28 (2.3%) had a clinical stroke or TIA, yet 68 (5.5%) had ≥1 new small non-cortical or large non-cortical/cortical infarct, of which 58 (85.3%) were clinically silent and 60 (88.2%) occurred in patients anticoagulated at baseline. Patients with new infarcts declined on the Cognitive Construct score (median change −0.12, IQR −0.22 to −0.07) while those without improved slightly (+0.07, −0.09 to +0.25), and clinically overt and silent infarcts had a similar cognitive impact; new white-matter lesions and microbleeds were not associated with cognitive decline (Kühne 2022, PMID 35171989, NCT02105844). In this cohort, covert infarcts occurred roughly twice as often as clinical stroke/TIA and carried cognitive cost; baseline anticoagulation in most affected patients does not establish adherence or estimate treatment failure. These lesions are one plausible contributor to the vascular-dementia excess in atrial fibrillation reported in epidemiology.

Lobar haemorrhage, cortical superficial siderosis, microbleeds, and white-matter injury can accompany CAA. Boston criteria v2.0 recognize cognitive impairment as a clinical presentation and combine haemorrhagic with selected non-haemorrhagic MRI markers (Charidimou 2022, PMID 35841910). Deep haemorrhagic lesions more often suggest hypertensive arteriopathy, while mixed patterns require caution.

The haemorrhagic subtype is distinguished from the others by its recurrence rate: pooled across 30 studies, CAA-related intracerebral haemorrhage recurred in 23% (95% CI 18–28%), with disseminated cortical superficial siderosis (OR 3.21, 95% CI 2.25–4.58) and convexity subarachnoid haemorrhage (OR 3.05, 95% CI 1.86–4.99) the strongest markers (Jia 2023, PMID 38020625). No other vascular-cognitive subtype carries an equivalently quantified event risk, which is why the haemorrhagic phenotype dominates antithrombotic decisions even when its cognitive contribution is modest — see CAA and red flags.

Classification for trials

Trial layer Minimum stratification
Prevention vascular exposure and baseline imaging burden
Acute stroke stroke type, site, severity, reperfusion, delirium
SVD modification lacunar phenotype, WMH/lacune burden, progression
Symptomatic cognition mild/major status, functional outcome, Alzheimer biomarker
Rehabilitation deficit profile, motor/language confounding, timing

Broad entry criteria can increase recruitment but dilute mechanism-specific effects. FINESSE argues for coherent small-vessel phenotyping and intermediate outcomes while preserving clinical endpoints (Markus 2022, PMID 35969390). Mixed vascular–Alzheimer disease is the modal community autopsy finding rather than a rare subtype: Schneider found 38.0% AD plus infarcts among dementia cases, and the Nun Study showed lacunar infarcts in deep structures raising dementia odds 20.7-fold among those with AD pathology (Schneider 2007, PMID 17568013; Snowdon 1997, PMID 9052711).

Open questions

  • If tract-specific WMH volume predicts post-stroke cognition better than total WMH volume, should trials adopt tract-based endpoints? (Coenen 2024, PMID 38651756)
  • Are brainstem and infratentorial lacunes strategic lesions for non-amnestic impairment, and why are they absent from classical strategic-site lists? (Wang 2024, PMID 38511349)
  • Why is cognitive impairment in asymptomatic carotid stenosis memory-predominant and non-lateralized in CREST-2 but haemodynamically patterned in smaller series? (Lazar 2021, PMID 34433306; Lal 2017, PMID 28712815)
  • If 85% of new brain infarcts in anticoagulated atrial fibrillation are clinically silent and carry the same cognitive cost as overt stroke, should covert infarction be an anticoagulation trial endpoint? (Kühne 2022, PMID 35171989)
  • Which lesion-network measures identify truly strategic infarcts better than anatomical labels? (O'Brien 2015, PMID 26595643)
  • Can early post-stroke trajectories distinguish recovery from progressive mixed disease? (Filler 2024, PMID 38101426)
  • What combination of SVD markers defines a trial-responsive subcortical syndrome? (Duering 2023, PMID 37236211)
  • When does measured hypoperfusion represent a modifiable mechanism rather than a correlate? (Santisteban 2023, PMID 36129176)
  • Is the left-lateralized strategic map for global post-stroke cognition biology or an artifact of language-dependent screening instruments? (Zhao 2018, PMID 28895445)
  • Does restoring cerebrovascular reserve in asymptomatic carotid stenosis improve cognition, given that impairment tracks reserve rather than plaque burden? (Lal 2017, PMID 28712815)
  • If single cerebrovascular pathologies do not accelerate decline in autopsy series, what does a single-substrate syndromic label actually identify? (Lamar 2022, PMID 34601898)
  • Why does large-artery versus small-vessel stroke subtype predict the low-performing cognitive trajectory (RRR 2.77) independently of stroke severity? (Lo 2023, PMID 37072222)
  • How often is a paramedian thalamic infarct diagnosed as a primary psychiatric disorder, given that it can present without focal neurological signs? (Carrera 2006, PMID 16801643; Schmahmann 2003, PMID 12933968)
  • Why was total SVD burden associated with mobility impairment after non-lacunar but not lacunar stroke, the reverse of the expected direction? (Loos 2018, PMID 28906203)
  • Is subclinical diastolic dysfunction a modifiable upstream cause of lacunar infarction, and should it be measured in vascular-cognitive risk models? (de Bruijn 2015, PMID 25632093)
  • Does heart failure accelerate cognitive decline, or does the 41.4% impairment prevalence reflect shared vascular risk? (Yap 2022, PMID 35839985)

References

  1. O'Brien JT, Thomas A. Vascular dementia. Lancet. 2015. PMID 26595643
  2. Gorelick PB, et al. Vascular contributions to cognitive impairment and dementia. Stroke. 2011. PMID 21778438
  3. Sachdev PS, et al. Diagnostic criteria for vascular cognitive disorders. Alzheimer Dis Assoc Disord. 2014. PMID 24632990
  4. Jellinger KA. The pathology of vascular dementia: a critical update. J Alzheimers Dis. 2008. PMID 18525132
  5. Jellinger KA. The pathology of ischemic-vascular dementia: an update. J Neurol Sci. 2002. PMID 12417375
  6. Pendlebury ST, Rothwell PM. Pre- and post-stroke dementia. Lancet Neurol. 2009. PMID 19782001
  7. Barbay M, et al. Post-stroke neurocognitive disorders. Dement Geriatr Cogn Disord. 2018. PMID 30504699
  8. Duering M, et al. STRIVE-2. Lancet Neurol. 2023. PMID 37236211
  9. Schneider JA, et al. Mixed brain pathologies account for most dementia cases. Neurology. 2007. PMID 17568013
  10. Santisteban MM, Iadecola C. Hypertension, neurovascular dysfunction, and cognitive impairment. Hypertension. 2023. PMID 36129176
  11. Charidimou A, et al. Boston criteria version 2.0 for CAA. Lancet Neurol. 2022. PMID 35841910
  12. Markus HS, et al. FINESSE framework for SVD trials. JAMA Neurol. 2022. PMID 35969390
  13. Filler J, et al. Risk factors after stroke. Lancet Healthy Longev. 2024. PMID 38101426
  14. Jacova C, et al. Cognitive impairment in lacunar strokes: SPS3. Ann Neurol. 2012;72:351-62. PMID 23034910
  15. Debette S, Markus HS. Clinical importance of white matter hyperintensities. BMJ. 2010;341:c3666. PMID 20660506
  16. Snowdon DA, et al. Brain infarction and the clinical expression of Alzheimer disease. JAMA. 1997;277:813-7. PMID 9052711
  17. Zhao L, et al. Strategic infarct location for post-stroke cognitive impairment: a multivariate lesion-symptom mapping study. J Cereb Blood Flow Metab. 2018;38:1299-1311. PMID 28895445
  18. Lal BK, et al. Asymptomatic carotid stenosis is associated with cognitive impairment. J Vasc Surg. 2017;66:1083-1092. PMID 28712815
  19. Lo JW, et al. Short-term trajectories of poststroke cognitive function: a STROKOG collaboration study. Neurology. 2023;100:e2331-e2341. PMID 37072222
  20. Lo JW, et al. Long-term cognitive decline after stroke: an individual participant data meta-analysis. Stroke. 2022;53:1318-1327. PMID 34775838
  21. Schmahmann JD. Vascular syndromes of the thalamus. Stroke. 2003;34:2264-78. PMID 12933968
  22. Carrera E, Bogousslavsky J. The thalamus and behavior: effects of anatomically distinct strokes. Neurology. 2006;66:1817-23. PMID 16801643
  23. Loos CM, et al. The relation between total cerebral small vessel disease burden and gait impairment in patients with minor stroke. Int J Stroke. 2018;13:518-524. PMID 28906203
  24. Yap NLX, et al. Prevalence and incidence of cognitive impairment and dementia in heart failure - a systematic review, meta-analysis and meta-regression. Hellenic J Cardiol. 2022;67:48-58. PMID 35839985
  25. de Bruijn RF, et al. Subclinical cardiac dysfunction increases the risk of stroke and dementia: the Rotterdam Study. Neurology. 2015;84:833-40. PMID 25632093
  26. Lamar M, et al. Complex profiles of cerebrovascular disease pathologies in the aging brain and their relationship with cognitive decline. Stroke. 2022;53:218-227. PMID 34601898
  27. Jia X, et al. Risk factors for recurrent cerebral amyloid angiopathy-related intracerebral hemorrhage. Front Neurol. 2023;14:1265693. PMID 38020625
  28. Wang J, et al. Strategic lacunes associated with mild cognitive impairment in rural Chinese older adults: a population-based study. Stroke. 2024;55:1288-1298. PMID 38511349
  29. Coenen M, et al. Strategic white matter hyperintensity locations associated with post-stroke cognitive impairment: a multicenter study in 1568 stroke patients. Int J Stroke. 2024;19:916-924. PMID 38651756
  30. Lazar RM, et al. Baseline cognitive impairment in patients with asymptomatic carotid stenosis in the CREST-2 trial. Stroke. 2021;52:3855-3863. PMID 34433306
  31. Kühne M, et al. Silent brain infarcts impact on cognitive function in atrial fibrillation. Eur Heart J. 2022;43:2127-2135. PMID 35171989