Compare dementia types
Seizures have been documented across all five dementia types in this atlas. What differs is the underlying brain disease, how an event may appear, which symptoms can look similar, and what each study measured.
How often seizures are reported
These selected findings cannot be used to rank dementia types. Some count seizures near diagnosis, some count diagnosed epilepsy in a clinic, and one comes from a small group with two dementia conditions.
| Dementia context | A directly relevant finding | How to read it |
|---|---|---|
| Alzheimer's disease | 5.7% had active seizures near diagnosis. Ting et al. (2026) found seizures in 174 of 3,040 people whose Alzheimer's diagnosis was confirmed after death. The study counted seizures within three years before or after diagnosis. [1] | The study excluded people with a clinical history of stroke or traumatic brain injury. Other studies followed people for longer and measured different outcomes.More Alzheimer's clinical evidence → |
| Dementia with Lewy bodies | 2.8% had active seizures near diagnosis. The same study found seizures in 13 of 458 people whose DLB diagnosis was confirmed after death. [1] | The number of seizure cases was small, and the study limited which brain changes could be included. Broader Lewy body disease studies may also include Parkinson disease.More DLB clinical evidence → |
| Frontotemporal dementia | 6.5% had diagnosed epilepsy in the year of FTD diagnosis. Kilpeläinen et al. (2025): 245 people with FTD; the matched-control figure was 1.8%. [2] | Registered epilepsy diagnoses, not all possible events or single seizures. Excess epilepsy was also present before FTD diagnosis. This is not an individual prediction of future dementia.More FTD clinical evidence → |
| Vascular dementia | 8.9% had epilepsy at a clinic assessment. Muroni et al. (2024) counted seizure-related conditions that required medication. The likely range around the estimate was wide, from 1.4% to 16.4%. [3] | This was a clinic result rather than a population survey. The wide range shows substantial uncertainty, and the reviewed abstract did not report how many people were in the vascular dementia group.More vascular dementia evidence → |
| Mixed dementia | 20.7% in a small combined DLB-AD group. Beagle et al. (2017) studied 29 people meeting clinical criteria for both probable DLB and probable Alzheimer's disease. [4] | A retrospective clinical subgroup, not all mixed dementia. It does not estimate the burden in Alzheimer's plus vascular disease or establish that two diseases double the risk.More mixed dementia evidence → |
Recognizing the event
A sudden change in response, speech, memory, or movement can matter in every dementia type. The person's usual pattern and the features of the dementia shape what observers notice.
Brief interruptions in memory, awareness, or conversation can be absorbed into a general impression of cognitive decline. Nonmotor seizures and convulsions are both reported.
Recognition focus: a discrete episode within the ongoing memory difficulties, including its onset, repeated sequence, and recovery.
Fluctuating attention, hallucinations, dream enactment, parkinsonism, and fainting can overlap with parts of an event. Clinical reports also describe focal seizures, convulsions, and prolonged seizure states.
Recognition focus: what distinguishes the episode from the person's usual fluctuation, sleep, and movement patterns. Myoclonus, a brief jerk, is not automatically a seizure.
Behavioral change, impaired language, apathy, and repetitive actions can complicate interpretation. A recent clinical cohort described focal seizures with impaired awareness. [5]
Recognition focus: an interruption beyond the usual behavioral or language difficulty, using familiar nonverbal responses when speech is limited.
Prior stroke deficits and new vascular symptoms can overlap with seizure presentations. The timing and location of brain injury are relevant, and new stroke-like symptoms require urgent assessment.
Recognition focus: the event's timeline alongside the vascular history. An old stroke does not explain every new episode.
Several disease-related explanations may coexist. The actual combination matters, whether Alzheimer's and vascular injury, Alzheimer's and Lewy body disease, or another overlap.
Recognition focus: preserve the whole episode rather than assigning each feature to a different diagnosis before it has been assessed.
Across all five types: Record what was usual beforehand, what changed first, whether the person could respond, how long the event lasted, and what recovery looked like. Include recent illness, sleep, and medication changes. Seagull developed these prompts to support clearer descriptions. They have not been tested as a diagnostic tool.
What EEG studies have found
General slowing, brief seizure-related electrical discharges, and actual seizures are different findings. A percentage for brief discharges cannot be presented as a rate of seizures without outward signs.
| Dementia context | What the guides document | What remains unresolved |
|---|---|---|
| Alzheimer's disease | Several prolonged-recording studies identify discharges without recognized clinical seizures. Small invasive studies demonstrate deep activity missed on scalp EEG. A 2026 controlled study also illustrates lower yields with different selection and review criteria. [6] | A generalizable burden, which findings affect function, and who benefits from treatment. More recording time does not resolve every detection problem.See detection studies → |
| Dementia with Lewy bodies | An early ear EEG study found brief discharges during the first recording in 8 of 10 people with DLB and 7 of 15 people without DLB. The difference in those proportions was not statistically clear, although the discharges occurred more often in DLB. [7] | The 80% figure describes brief discharges, not seizures without outward signs. The small groups, frequent findings in controls, limited electrode coverage, and comfort concerns limit the result. |
| Frontotemporal dementia | A recent study found brief seizure-related discharges in two of 62 tested people who had no history of clinical seizures. Only selected people received EEG rather than the full FTD group. [5] | Researchers still do not know how common this activity is, what effects it may have, or how well different recording methods perform in FTD. |
| Vascular dementia | Stroke research provides useful information about electrical abnormalities and added detection with continuous EEG. Acute stroke populations are not equivalent to vascular dementia populations. | Direct, representative vascular dementia studies separating background slowing, discharges, and seizures.See the stroke evidence and its limits → |
| Mixed dementia | A study of DLB with and without Alzheimer's-related biomarkers found general slowing in both groups. In the combined group, 36 of 41 people had broad EEG abnormalities, while only one had a seizure-related abnormality during the 20-minute recording. [8] | Researchers still need to measure actual seizures in clearly defined combinations and test new methods when the brain's usual rhythms are already changed. |
Need the terminology first? Read what EEG recordings can show, the limits of detection, and what “silent” does and does not mean.
What research needs to answer next
Distinguish clinical syndromes, confirmed pathology, age and stage, and mixed disease. A broad dementia category can conceal meaningful differences.
Count clinical seizures, epilepsy diagnoses, discharges, and electrographic seizures separately. Report sleep, duration, coverage, and interpretation criteria.
Test whether better detection changes care and improves function, safety, sleep, or quality of life. More detected abnormalities alone do not establish benefit.
Seagull's perspective: The evidence already documents a meaningful seizure burden. The gaps show researchers where better studies are needed. See how the evidence is assessed.
Sources and scope
This page summarizes the five subtype guides. Each guide provides the full study context, limitations, and research discussion. The selected sources below support the numerical examples used here.
Ting et al. (2026). Neuropathologic Correlates of Seizures in Patients With Alzheimer Disease and Dementia With Lewy Bodies. In the Evidence Map →
Autopsy-confirmed cohort; active seizures around diagnosis, not lifetime prevalence.
Kilpeläinen et al. (2025). Prevalence of Epilepsy in Frontotemporal Dementia and Timing of Dementia Diagnosis. In the Evidence Map →
Retrospective case-control study; registered epilepsy diagnoses at specified time points.
Muroni et al. (2024). Point prevalence of epilepsy in dementia: A “real-world” estimate. In the Evidence Map →
Clinical prevalence estimate. Published abstract reviewed; subgroup denominator and full text were not available for this review.
Beagle et al. (2017). Relative Incidence of Seizures and Myoclonus in Alzheimer's Disease, Dementia with Lewy Bodies, and Frontotemporal Dementia. In the Evidence Map →
Retrospective memory-center study. Its time-to-event estimates and small DLB-AD subgroup proportion are different measurements.
Hashmi et al. (2025). Co-morbid seizures in frontotemporal dementia: What do they tell us?. In the Evidence Map →
Clinical FTD cohort; detailed seizure observations but selective EEG testing.
Szabo et al. (2026). Refining Detection of Subclinical Epileptiform Activity in Alzheimer's Disease: A Case-Control Study and Call for a Consensus. In the Evidence Map →
Thirty Alzheimer's participants and 30 controls. Informs method-related differences in discharge detection; see the Alzheimer's guide for the wider recording literature.
Musaeus et al. (2023). Subclinical Epileptiform Activity in Dementia with Lewy Bodies. In the Evidence Map →
Exploratory ear-EEG study, not a validated population estimate of silent seizures.
van der Zande et al. (2018). EEG Characteristics of Dementia With Lewy Bodies, Alzheimer's Disease and Mixed Pathology. In the Evidence Map →
Short resting-state recordings in three groups of 41. Background abnormalities and epileptiform findings were separate outcomes.