Six months after a lacunar stroke with a clean motor recovery, a retired accountant's daughter does the talking in clinic: he abandoned the taxes halfway through, double-paid two bills, and takes all afternoon to do what used to take an hour. Ask him to recall three words and he gets all three. Ask him to draw a clock or alternate letters and numbers, and the trouble is suddenly visible.

That dissociation — memory roughly intact, executive function and processing speed failing — is the signature of vascular cognitive impairment, and it is why VCI hides from families, from three-word recall, and from clinicians who equate "dementia" with "forgetting." It is also, among all the roads to dementia, the one a stroke service is best positioned to block.

The size of the problem — before and after the stroke

The epidemiology is sobering in both directions. Pooled data across hospital-based cohorts show roughly 10% of patients are already demented at the time of a first stroke, about 10% develop new dementia soon after a first stroke, and after recurrent stroke the figure exceeds 30%.1 Milder impairment short of dementia is more common still. Two clinical translations follow: cognition deserves the same follow-up billing as motor recovery — and recurrence prevention is cognition prevention, which puts the entire secondary-prevention apparatus, from antithrombotics to the CHA₂DS₂-VASc decision, in the dementia-prevention business.

Four mechanisms, usually mixed

VCI is not one disease; it is a final common pathway with four main lesion routes, formalized in the AHA/ASA statement on vascular contributions to cognitive impairment:2

  • Multi-infarct disease — cumulative cortical and subcortical infarcts, classically stepwise in course.
  • Strategic single infarct — one lesion in a cognition-critical node: the thalamus (including the artery-of-Percheron pattern from our PCA stroke guide), caudate, angular gyrus, or hippocampal territory. One artery, a dementia-scale deficit.
  • Cerebral small vessel disease — the workhorse mechanism: lacunes, white matter hyperintensities, and microbleeds accumulating silently for years, eroding processing speed and executive function without a single "stroke day." This is the substrate patients and scanners both under-report.
  • Hemorrhagic disease — macrohemorrhage and the amyloid-related small vessel pathologies, where cognition, bleeding risk, and antithrombotic decisions tangle together.

And over age 75, the honest label is usually mixed disease: vascular pathology lowering the threshold at which concurrent Alzheimer pathology becomes symptomatic. The practical corollary cuts through the nosology: whatever the mix, the vascular component is the modifiable one.

Assessment: the right test, at the right time, minus the mimics

The right test

Because VCI's center of gravity is executive function, attention, and speed, screening tools weighted toward memory under-detect it. The Montreal Cognitive Assessment (MoCA) — with its trail-making, clock, fluency, and abstraction items — is the pragmatic screen for this population. After TIA and stroke it is more sensitive than the MMSE, which ceilings out on executive and visuospatial items; it is not a diagnosis.35 Score it with its education correction, note aphasia, neglect, and motor limits that invalidate items, and treat any screen as a screen: formal neuropsychological testing is the arbiter when stakes (work, driving, capacity) are high.

The right time

Testing in the first days after a stroke measures the acute injury, delirium, and the hospital — not the trajectory. Screen briefly before discharge if it changes disposition, but anchor the meaningful assessment at roughly three to six months post-stroke, when edema, delirium, and early fatigue have settled and what remains is what the patient will live with. Then follow it: VCI is a trajectory diagnosis, and a second data point a year later is worth more than any single score.

Minus the mimics

Three reversible impostors produce low scores and false labels. Depression is the great one — apathy, poor concentration, and give-up answers mimic executive failure, it affects roughly a third of survivors, and it is treatable: screen it deliberately with the PHQ-9 before or alongside any cognitive label (the rest of those screens live on the calculator hub). Persisting delirium after hospitalization fluctuates in a way degenerative disease does not. And uncorrected hearing and vision quietly fail testing for sensory reasons. Post-stroke fatigue rounds out the list — a patient who cannot stay alert through a clinic afternoon will not survive a MoCA either.

Before you tell a family it's vascular dementia, be able to say you have ruled out the three things that impersonate it — because those three you can fix.

Treatment and prevention: honest about drugs, aggressive about vessels

Start with the honest part: no medication is approved for VCI itself. Cholinesterase inhibitors and memantine have shown small, inconsistent cognitive-scale benefits in vascular dementia trials without convincing functional gains — reasonable to consider case-by-case, especially where mixed Alzheimer pathology is likely, but never the center of the plan.2

The center of the plan is vascular, and it has real evidence behind it:

  • Blood pressure is the headline — with two caveats from the trial everyone quotes. In SPRINT MIND, intensive systolic control (<120 mmHg vs <140) reduced mild cognitive impairment (HR 0.81) and the combined MCI-or-dementia outcome, but the primary endpoint of probable dementia was not significant (HR 0.83, p=0.10). SPRINT also excluded prior stroke and diabetes.4 Do not import a <120 target wholesale onto a stroke clinic panel. After stroke or TIA, AHA/ASA secondary prevention typically aims for <130/80 in most patients with hypertension — distinct from the deliberately permissive acute-phase targets we cover in acute BP targets.6
  • Prevent the next stroke — every recurrence steepens the cognitive slope, so the unglamorous checklist (antithrombotic fitted to mechanism, anticoagulation for AF, statins, diabetes control, smoking cessation) is cognition therapy by another name.
  • Exercise, and treat what worsens cognition — physical activity, hearing correction, sleep apnea treatment, and a hard look at anticholinergic and sedating medication lists. The lifestyle case is laid out for patients in our vascular health and brain function guide.
  • Support the life around the cognition — occupational therapy for compensatory strategies, driving assessment where executive function is in question, financial safeguards for the patient double-paying bills, and caregiver education early rather than after the crisis.

For families who want the fuller plain-language picture of how stroke and dementia intertwine, our patient-facing companion piece — stroke and dementia — is written to be handed over; this page is its clinician counterpart.

The bottom line

Look for VCI where it lives — in executive function and speed, not three-word recall — with a MoCA at three to six months, after depression, delirium, and the senses have been dealt with. Name the mechanism, because mechanism is what you treat. And be candid that the pharmacy shelf is nearly bare while the prevention shelf is full: the blood pressure cuff, the anticoagulant matched to the mechanism, the statin, and the next stroke that never happens are the most effective cognitive interventions vascular medicine owns. The accountant's taxes are a vascular outcome. Treat them like one.