The vascular dementia evidence guide
Vascular dementia develops when strokes or other blood-vessel problems injure the brain and affect thinking and daily life. Seizures are documented, but the risk and the way they appear depend on the type, location, and timing of the brain injury.
This guide explains what studies have found, how a possible seizure can differ from a new stroke or the return of an old stroke symptom, what brain recordings can show, and what researchers still need to learn.
Start here
A large stroke, damage from small blood vessels, and a combination of vascular injury with Alzheimer's disease can affect the brain differently. The diagnosis alone does not provide one seizure-risk estimate for every person. [3]
A seizure within days of a new stroke differs from one linked with an older injury. The important timeline begins with the stroke or other brain injury, not with the date the dementia was diagnosed. [4]
Stroke studies show that longer EEG monitoring can detect more electrical activity than a recording limited to the first hour. This does not mean that any test finds every seizure, and stroke studies do not show how often hidden activity occurs in established vascular dementia. [8] [9]
How often seizures are reported
Two studies directly examined vascular dementia. One measured how many people had epilepsy when they were assessed. The other measured new seizure or epilepsy diagnoses over time. Each percentage answers a different question.
A 2024 clinic study reported epilepsy requiring medication in 8.9% of its vascular dementia group. The likely range around this estimate was wide, from 1.4% to 16.4%, because the group was small. This was a specialty-clinic result rather than a population survey or a study of seizures without outward signs. [1]
In a UK primary-care study, 55 of 4,205 people with vascular dementia and no earlier seizure diagnosis later received a seizure or epilepsy diagnosis. The rate was 7.5 per 1,000 person-years, compared with 0.8 in people without dementia. This is a rate over time, not 7.5% of the group. [2]
The association is more than plausible. Epilepsy is present in clinical vascular dementia populations, and new seizure diagnoses occur more often than in dementia-free comparators. That makes the issue relevant to memory services, primary care, stroke follow-up, rehabilitation, and residential care.
Neither study systematically measured every brief event or every seizure without visible signs. Clinical records and medication-based definitions answer a different question from prolonged brain monitoring. Their estimates should not be averaged or turned into one person's forecast.
The Imfeld study is a foundational exception to this guide's preference for recent evidence: it provides a large, direct vascular dementia cohort and a dementia-free comparison. It examined UK records from 1998 to 2008 in adults aged 65 or older.
Its limits matter. Dementia subtype and seizure coding could be misclassified, and cognitive severity was unavailable. It establishes a historical population-level association, not today's exact prevalence or an EEG-defined burden. The newer clinic study leads the page; this older study supplies a different, still important kind of evidence. [2]
Why timing matters
The same phrase, “seizures after stroke,” can describe different clinical situations. The standard timing distinction refers to the stroke, not to the onset of memory problems.
| Timing | What the category means | Why it matters |
|---|---|---|
| Within seven days of a stroke | A seizure occurs in close relation to the new brain injury. Doctors call this an acute symptomatic seizure. | One such event does not automatically mean the person has lasting epilepsy. |
| More than seven days after a stroke | A seizure linked with the older injury can indicate post-stroke epilepsy. | The chance of another seizure may be high enough to diagnose epilepsy after one event. Clinicians still need to check for other immediate causes. |
Stroke studies link seizure risk with the size and severity of the injury and whether it involves the brain's outer layer, called the cortex. An imaging report of small-vessel changes provides different information from a documented stroke involving the cortex. [4]
Seagull's perspective: The words “vascular dementia” begin the clinical question. They cannot provide a seizure risk score by themselves.
Post-stroke dementia includes dementia of different causes identified after a stroke. Vascular cognitive impairment can also develop without a recognized clinical stroke. Coexisting neurodegenerative disease complicates attribution. [3]
This is why an acute stroke EEG study cannot establish how often silent seizures occur across all people living with vascular dementia.
Keep three histories together: vascular injuries, cognitive change, and seizure-like episodes. Their order and overlap are more informative than a diagnosis label alone.
How seizures can appear
A focal seizure begins in one area or network of the brain. It can affect movement, language, sensation, awareness, or response. The examples below describe possible seizure patterns. They cannot identify the cause of one person's event. [5]
The person stops an activity, stares, or responds differently for a brief interval. What they can recall afterward and whether they could respond during it are separate observations.
Speech or understanding may be interrupted. Existing aphasia, a language difficulty after brain injury, can make a new change difficult to identify.
Rhythmic movement of a hand or face, a repeated automatic action, or a sudden unusual sensation may occur. The sequence matters more than a single isolated sign.
A focal seizure can spread and cause bilateral stiffening and jerking. Not every seizure follows that course; absence of shaking should not end the inquiry. [5]
Example of a useful event description
“Usually answers short questions despite difficulty finding words. At breakfast, stopped answering and repeatedly moved the right hand for about a minute. Was much less responsive afterward. Last known at the usual baseline was 8:10.”
This note records the person's usual ability, the time the change began, the affected side, the duration, and the recovery. Sudden neurological changes need prompt assessment even when a seizure seems possible.
Symptoms that can look similar
Vascular brain injury may already affect attention, thinking speed, language, or movement. A useful account explains what changed from the person's usual function and how suddenly it happened. [3]
Symptoms that improve are not automatically harmless. A transient ischemic attack, or TIA, can be a warning of stroke. A dementia or epilepsy diagnosis does not remove that possibility. American Stroke Association guidance
Earlier weakness or language difficulty can return during illness or other physical stress. Doctors call this post-stroke recrudescence. A study of 153 people linked these episodes with infection, low blood pressure, low sodium, and other factors. Clinicians still had to assess for a new stroke and seizure. [6]
Transient neurological episodes also occur with syncope, migraine, and toxic or metabolic disturbances. These alternatives appear alongside seizures in clinical stroke-mimic research. Their presence is a reason to broaden assessment, not to dismiss an event. [14]
Dementia can make a brief interruption hard to report. An observer may miss the beginning or see only the recovery. The EAN dementia guideline highlights how cognitive impairment and limited caregiver recognition can conceal seizures. [12]
Record the change in plain language. “Less responsive than usual, started suddenly, lasted two minutes” carries more information than “another dementia spell.” Repeated episodes are important to report, and a clinician must still consider several possible causes.
What EEG studies have found
An EEG records the brain's electrical activity. A report should explain whether it found a seizure, a brief seizure-related discharge, general slowing, or another uncertain pattern.
These distinctions draw on standardized EEG terminology. They prevent a study reporting discharges, slowing, or a combined category from being presented as a study of silent seizures. [7]
A 2026 study compared the first 60 minutes with at least 12 hours of the same EEG in 283 selected patients after acute ischemic stroke. Monitoring began within seven days. Longer recording identified more activity. [8]
| Finding | First 60 minutes | At least 12 hours |
|---|---|---|
| Brief epileptiform discharges | 3% | 11% |
| Electrographic seizures | 0.7% | 4% |
The scope: selected acute stroke patients, not vascular dementia screening or a randomized trial. Added detection does not establish that treating every extra finding improves outcomes. [8]
Limits of EEG detection
Two questions matter: did the activity occur during the recording, and was it visible to the electrodes?
The stroke comparison above found more activity over at least 12 hours than during the first hour. A short recording may end before occasional activity occurs. [8]
In 27 people being evaluated for epilepsy surgery, electrodes placed inside the brain recorded 172 seizures while scalp EEG showed 100. Detection differed by seizure type. This epilepsy study explains a scalp-recording limit. It does not support routine invasive monitoring in dementia. [9]
A recorded seizure, a brief discharge, and a slowing pattern answer different questions. A report that calls all three “silent seizures” misrepresents the evidence. [7] [10]
Recording time and signal visibility are separate limits. Longer EEG provides more time to capture activity that comes and goes. It cannot guarantee that activity beginning deep in the brain will appear clearly at the scalp. A negative result can leave the cause of an event unresolved.
A detection percentage depends on who was studied, which events counted, recording duration, electrode coverage, and the reference used to establish a seizure. A sensitivity estimate from presurgical epilepsy patients cannot be transferred to vascular dementia or to every ambulatory EEG system. [9]
For this subtype, the responsible question is specific: which clinically relevant events remain undetected by which method, in which vascular dementia population? The studies cited here do not settle that question.
Why this matters for care
European dementia guidance focuses on reducing seizures and their consequences while protecting quality of life. Medication-related sleepiness, changes in thinking, and drug interactions need particular attention. The treatment advice is based on general epilepsy evidence rather than a vascular dementia treatment trial. [12]
A new event history deserves its own explanation. Ask whether a reported change was sustained, intermittent, witnessed, or linked to a specific injury or illness.
A clear account of usual language, movement, and responsiveness helps communicate what is new. A history of dementia must not obscure an acute change.
A family member, therapist, or care worker may witness a brief event that never happens during a clinic visit. Consistent descriptions can connect observations that otherwise remain isolated.
The studies on this page do not establish that screening everyone with vascular dementia, or treating every isolated discharge, slows cognitive decline. That boundary does not diminish the importance of recognizing and treating diagnosed seizures. Medication decisions require an individualized clinical assessment; this guide does not recommend starting, stopping, or changing treatment.
Research insights
Each technology addresses a different problem. Some predict future epilepsy after stroke. Others make EEG available sooner. Neither approach has yet established a screening pathway for vascular dementia.
The SeLECT-EEG study included 1,105 people who survived a stroke caused by a blocked blood vessel and had an early EEG. Seizure-related activity and slowing in one brain region were linked with later epilepsy. Adding EEG information improved prediction compared with the existing clinical model. [10]
What it means: This study predicted future epilepsy after stroke. It did not detect a current seizure without outward signs. Slowing may help predict later risk even though slowing itself is not a seizure. The model was developed for a specific stroke group and is not a vascular dementia calculator.
Seagull's perspective: Future tools should clearly name the people and clinical question they are designed for. A model tested after one kind of stroke needs separate study before it is used in long-term vascular cognitive impairment.
2020 / Rapid access to EEG
DECIDE evaluated rapid-response EEG in five intensive-care units. It enrolled 181 patients, with complete clinical and EEG data for 164. The typical wait was five minutes with the rapid system compared with 239 minutes for conventional EEG. [11]
What it means: This study involved patients with suspected seizures without shaking. It was not a dementia screening trial. Faster access did not show that the system found every seizure or improved thinking. The publication disclosed roles connected with the manufacturer.
Seagull's perspective: Easier access is worth studying. Any dementia use would still need evidence about electrode coverage, false alarms, comfort, expert review, and changes in patient care.
2026 / Vascular brain health
The 2026 American Heart Association statement describes age-related cognitive impairment as a process involving interacting vascular and other mechanisms. It calls for deeper understanding of injury and repair while maintaining the importance of managing modifiable vascular risks. [13]
Seagull's research question: where does abnormal electrical activity fit into that interaction? Studies should establish whether it is a marker of injury, an additional contributor to impairment, or both, and whether its treatment changes outcomes. This is a research direction, not a conclusion of the AHA statement.
How Seagull evaluates a technology claim
Inclusion in this guide is research commentary. It is not a product endorsement or a statement that the technology is available for vascular dementia care.
What research needs to answer next
Clinical evidence shows that seizures belong in the vascular dementia conversation. The next studies need to define who is affected, what current testing misses, and whether better detection improves care.
Studies should distinguish cortical stroke, small-vessel injury, hemorrhage, and mixed pathology, with clear dementia definitions and disease stage. An inclusive label should not hide clinically important differences.
How often are actual seizures recorded in people living with established vascular dementia? Count seizures separately from discharges and slowing, and explain how recording duration and coverage affect the result.
Can longer or easier-to-use recordings answer meaningful questions in homes and care facilities? Feasibility, missed events, false positives, and the burden on patients and caregivers belong in the same evaluation.
Does identifying and appropriately treating previously unrecognized activity improve safety, cognition, communication, or daily function? Trials should separate seizure control from claims about changing dementia progression.
Sources and evidence notes
Vascular dementia studies provide the burden estimates. Stroke and epilepsy studies help explain timing, event patterns, and EEG limits. Clinical guidance provides definitions and care context. Each note explains how directly the source applies.
A large clinical-record study can support an association without capturing unrecognized seizures. A within-recording comparison can show added EEG yield without proving better patient outcomes. A guideline's good-practice statement is not the same as trial evidence.
This guide is a focused evidence synthesis, not a systematic review or a formal certainty grading. Sources from the preceding ten years are prioritized. The 2013 vascular dementia cohort is retained as an explicitly identified foundational exception. Seagull's interpretation and research priorities are labeled separately from study findings.
Muroni A, et al. Point prevalence of epilepsy in dementia: A “real-world” estimate. Epileptic Disorders. 2024;26:209–214. DOI: 10.1002/epd2.20199.
Direct vascular dementia evidence. Clinic estimate, not population prevalence. PubMed abstract reviewed; the full paper and subgroup denominator were not available for this review.
Imfeld P, et al. Seizures in patients with Alzheimer's disease or vascular dementia: A population-based nested case-control analysis. Epilepsia. 2013;54:700–707.
Foundational direct evidence. Full paper reviewed. Historical UK primary-care cohort; recorded diagnoses, not systematic EEG monitoring. Dementia subtype, seizure coding, and unavailable cognitive severity limit interpretation.
El Husseini N, et al. Cognitive impairment after ischemic and hemorrhagic stroke. AHA/ASA scientific statement. Stroke. 2023;54:e272–e291.
Definitions and cognitive context. Distinguishes vascular cognitive impairment, vascular dementia, and post-stroke dementia. Not a seizure-prevalence study.
Holtkamp M, et al. European Stroke Organisation guidelines for the management of post-stroke seizures and epilepsy. European Stroke Journal. 2017;2:103–115.
Stroke guidance. Supports timing categories and injury context. Treatment recommendations were based on very-low-certainty evidence.
Beniczky S, et al. Updated classification of epileptic seizures: Position paper of the International League Against Epilepsy. Epilepsia. 2025;66:1804–1823.
Seizure classification. Supports descriptions of clinical manifestations, not vascular dementia-specific frequencies.
Topcuoglu MA, et al. Recrudescence of deficits after stroke: Clinical and imaging phenotype, triggers, and risk factors. JAMA Neurology. 2017;74:1048–1055.
Recognition context. Retrospective stroke cohort. Its diagnostic criteria required exclusion of alternative explanations; it is not a home-triage rule.
Hirsch LJ, et al. American Clinical Neurophysiology Society's standardized critical care EEG terminology: 2021 version. Journal of Clinical Neurophysiology. 2021;38:1–29.
EEG terminology. Distinguishes seizure patterns from other abnormalities.
Schubert KM, et al. Continuous versus short EEG after ischemic stroke. Annals of Neurology. 2026;100:400–415.
Added detection, selected stroke cohort. Retrospective comparison; shares centers and participants with the SeLECT-EEG research program.
Casale MJ, et al. The sensitivity of scalp EEG at detecting seizures: A simultaneous scalp and stereo EEG study. Journal of Clinical Neurophysiology. 2022;39:78–84.
Indirect detection evidence. Drug-resistant epilepsy surgery candidates, not a vascular dementia cohort.
Schubert KM, et al. The role of electroencephalography in predicting post-stroke seizures and an updated prognostic model (SeLECT-EEG). Annals of Neurology. 2025;98:814–825.
Prognostic research. Selected ischemic stroke cohorts; not a silent-seizure detector or vascular dementia risk calculator.
Vespa PM, et al. Evaluating the clinical impact of rapid response electroencephalography: The DECIDE multicenter prospective observational clinical study. Critical Care Medicine. 2020;48:1249–1257.
Technology and access. ICU workflow study with disclosed manufacturer relationships. Not a dementia outcomes trial or a product endorsement.
Frederiksen KS, et al. A European Academy of Neurology guideline on medical management issues in dementia. European Journal of Neurology. 2020;27:1805–1820.
Dementia care guidance. The epilepsy section relies on indirect evidence and clinical good practice. It does not establish treatment benefit for isolated EEG discharges.
Sorond FA, et al. Vascular contributions to cognitive impairment and brain health: Clinical update and mechanistic considerations. AHA scientific statement. Stroke. 2026;57:e254–e270.
Broader vascular context. Supports an integrated brain-health perspective, not an EEG screening or seizure-treatment recommendation.
Jun-O'Connell AH, et al. Recrudescence of old stroke deficits among transient neurological attacks. The Neurohospitalist. 2019;9:183–189.
Competing explanations. Clinical study of transient neurological presentations, including stroke mimics. Not a vascular dementia cohort.