The mixed dementia evidence guide

Mixed dementia and seizures

Mixed dementia means that more than one disease or type of brain injury is contributing to a person's dementia. Common combinations include Alzheimer's disease with vascular injury or with Lewy body disease. The combination can affect both seizure risk and how an event appears.

This guide explains why mixed dementia does not have one seizure profile, how to describe a possible event, what brain recordings can show, and where the evidence remains limited.

Seagull Health Evidence reviewed

Start here

Three things to know before reading the research

01

The exact combination matters

Alzheimer's disease with vascular injury differs from Alzheimer's disease with Lewy body disease. A doctor's working diagnosis, a biomarker test, and an examination of brain tissue after death also provide different levels of certainty about which conditions are present. [1] [4]

02

Describe the event before deciding what caused it

A sudden, repeated interruption in response, speech, memory, or movement deserves its own description. The mixed dementia label cannot determine whether that event was a seizure or which condition caused it. [5]

03

Different brain-recording findings must remain separate

A recorded seizure, a brief seizure-related electrical discharge, and general slowing are different findings. One should not be used as proof of another, even when a report describes the brain recording as abnormal. [4] [8]

Understanding the combination

“Mixed dementia” means that more than one condition is affecting the brain.

A person may have one main diagnosis and evidence of another disease or brain injury. Studies define these combinations in different ways, so their groups and percentages cannot always be compared directly.

A guide to interpreting the context, not a ranking of seizure risk. [1] [2] [6] [7]
CombinationWhat changes the clinical pictureRead alongside this guide
Alzheimer's disease and vascular brain injuryMemory difficulties can coexist with stroke-related language, movement, or attention changes. The location and timing of injury matter when interpreting a new episode.Alzheimer's disease
Vascular dementia
Alzheimer's disease and Lewy body diseaseMemory impairment may coexist with fluctuations, hallucinations, disturbed sleep, or parkinsonism. These features can complicate recognition without establishing that an episode is epileptic.Alzheimer's disease
Dementia with Lewy bodies
Other or more complex combinationsSeveral pathologies may be present. Their presence does not tell us which one explains a particular symptom, EEG finding, or seizure.Compare dementia types
Evidence and methods

Ask what “mixed” means in this case. Is it a clinical working diagnosis, imaging evidence of vascular injury, biomarker-supported Alzheimer's disease alongside another syndrome, or pathology confirmed after death? That distinction determines how closely a study applies.

What studies show

Seizures have been documented in people with overlapping dementia conditions.

One clinical study counted seizures in a small Alzheimer's and Lewy body group. Larger studies examined how seizures relate to combinations of brain changes found after death. Together, they support the connection without providing one percentage for all mixed dementia.

20.7%of 29 people

New seizures in a small Alzheimer's and Lewy body group

A 2017 memory-center study reported new seizures in 20.7% of 29 people who met the clinical criteria for both probable Alzheimer's disease and probable dementia with Lewy bodies. This was the highest percentage among the study's clinical groups. [3]

The finding is important, but the group was small, selected, and reviewed through past medical records. The diagnoses were not confirmed in everyone by biomarkers or examination after death. The study also did not systematically look for seizures without outward signs. This result cannot serve as a percentage for every form of mixed dementia.

Autopsy evidence · 2025

Additional Alzheimer's pathology can matter beyond the primary diagnosis.

Zawar and colleagues studied 6,085 autopsied people with cognitive impairment, including mild cognitive impairment and dementia. Among those whose primary pathology was not Alzheimer's disease, the active-seizure group had more advanced Alzheimer's-type tau pathology than the remote-seizure and no-seizure groups. [1]

The finding connects seizure history with disease processes that a primary diagnosis can conceal. However, “active” included recent recurrent seizures and/or ongoing antiseizure treatment. It did not mean that every person had uncontrolled epilepsy.

Autopsy evidence · 2026

Different accompanying pathologies show different associations.

Ting and colleagues studied 3,498 people with autopsy-confirmed Alzheimer's disease or dementia with Lewy bodies. In Alzheimer's disease, advanced tau pathology and more severe cerebral amyloid angiopathy, amyloid in blood-vessel walls, were associated with seizures. In the Lewy body group, microscopic and larger infarcts were associated with seizures. [2]

Lewy body co-pathology in the Alzheimer's group was not independently associated with seizures. More pathology does not automatically mean a predictable, additive increase in seizure risk.

How should these findings be weighed together?

The clinical subgroup and autopsy studies use different definitions, time windows, and populations. The 2026 study counted active seizures around dementia diagnosis, excluded clinical stroke and traumatic brain injury, and included only 13 seizure cases in its Lewy body group. It also restricted the extent of tau pathology in that group.

The 2025 study did not find broad group differences in vascular or Lewy body pathology. The 2026 study identified more specific associations in differently selected groups. Both used the National Alzheimer's Coordinating Center database, so they are not wholly independent population replications.

These observational findings identify relationships worth investigating. They cannot establish which pathology caused a seizure, whether seizures accelerated the pathology, or whether treatment would change the course of mixed dementia. [1] [2]

How seizures can appear

A seizure may look like a sudden break from the person's usual pattern.

Some seizures cause shaking. Others briefly affect response, memory, language, sensation, or movement. The examples below describe possible seizure patterns. They cannot identify the cause of one person's event. [5]

A break in response

An activity stops, the person stares, or their responses change abruptly. The important comparison is with their usual level of engagement, not an expectation of normal cognition.

A brief interruption in language or recall

The person suddenly cannot continue a familiar conversation or later cannot recall an interval. Existing language and memory impairment can make the beginning and end of an episode harder to recognize.

A repeated movement or unusual sensation

A focal seizure may involve rhythmic movement, automatic actions, or a subjective experience the person cannot easily describe. An isolated jerk or unusual behavior is not sufficient to establish a seizure.

A convulsion or a period of recovery

Some seizures cause obvious stiffening and jerking. Others are noticed through a change afterward, such as confusion or reduced responsiveness. Recovery observations support assessment but do not identify the cause by themselves.

Brief changes without shaking can still matter. In Beagle's broader dementia study, 39 of 71 people whose seizure type could be identified had only nonmovement seizures. This result covers several dementia conditions, not only the mixed group. [3]

Symptoms that can look similar

Mixed dementia can create several possible explanations for one event.

Changing alertness, fainting, sleep events, stroke symptoms, illness, and medication effects can resemble parts of a seizure. A clear description helps clinicians compare the possibilities and recognize when urgent assessment is needed.

Fluctuations and fainting

When Lewy body disease contributes, fluctuations can include marked changes in attention and responsiveness. The DLB consensus report notes that transient unresponsiveness can be difficult to distinguish from syncope, or fainting. That overlap also complicates seizure recognition. [6]

Sleep-related events

Dream enactment, disrupted sleep, and daytime sleepiness may be part of a Lewy body presentation. A nighttime movement is not automatically a seizure. The relationship to sleep, the event sequence, and any simultaneous recording matter. [6]

New vascular symptoms or an old injury

Sudden language loss or weakness may require emergency stroke assessment. If a seizure is related to stroke, one within seven days of a new stroke is classified differently from an unprovoked seizure associated with an older injury. “Mixed dementia” does not remove that timing distinction. [7]

Illness and medication effects

Changes in medical health, metabolic disturbance, and medication can affect cognition or seizure threshold. Sedation and adverse effects can also obscure whether function has improved. Dementia care therefore needs medical review alongside event assessment. [14]

Record the event in a consistent way. Note what was usual beforehand, what changed first, how long it lasted, whether the person could respond, what they recalled, and how they recovered. Include recent illness, medication changes, sleep, and posture.

What EEG studies have found

EEG can show a seizure, a brief discharge, or general slowing.

An EEG records the brain's electrical activity. The meaning depends on the pattern, recording length, electrode coverage, and what happened during the recording. Four findings should remain separate. [5] [8]

A subtle clinical seizure
An episode with a clinical change, such as interrupted response or speech, but no prominent convulsion. A sign can be easy to miss and still be a clinical manifestation.
A seizure without an obvious outward sign
A recording meets seizure criteria without a clear accompanying clinical change. “Silent” describes the lack of an obvious sign, not proof that the activity has no effect.
Brief seizure-related discharges
Spikes or sharp waves called epileptiform discharges may indicate a tendency toward seizures. They are not, by themselves, recorded seizures or proof that a particular episode was epileptic.
Background slowing
Slower ongoing brain rhythms may reflect brain dysfunction. This is not equivalent to epileptiform discharges or seizure activity, even when an EEG report calls it abnormal.

A 20-minute EEG study shows why the distinction matters

Van der Zande and colleagues compared three groups of 41 people: Lewy body dementia with Alzheimer's-related markers in spinal fluid, Lewy body dementia without those markers, and Alzheimer's disease. Both Lewy body groups had more general slowing and changes in connected brain activity than the Alzheimer's group. The two Lewy body groups were similar on those measures. [4]

What was common

Diffuse EEG abnormalities appeared in 36 of 41 people in the combined Lewy body and Alzheimer's biomarker group. Those abnormalities included slowing and reduced background reactivity.

What was a different finding

Epileptiform abnormalities were reported in only 1 of those 41 people during the 20-minute resting recording. Neither number is a reliable estimate of how often silent seizures occur in mixed dementia.

The study mainly measured general brain rhythms. A short EEG can provide information about dementia-related changes without showing epilepsy. It may also end before occasional seizure-related activity occurs.

Limits of EEG detection

Recording time and electrode coverage create different limits.

A longer EEG gives researchers more time to capture activity that comes and goes, especially during sleep. Activity beginning deep in the brain may still produce little or no visible signal at the scalp. The type and length of recording both affect what a study can find.

01

Did the activity happen during the recording?

Occasional activity may not occur during a short EEG. In an Alzheimer's study, repeated ear EEG recordings found substantial differences from one session to another, and much of the activity appeared at night. This study highlights the importance of recording time and sleep, but it did not include a defined mixed dementia group. [11]

02

Was the activity visible to the scalp electrodes?

In 27 people being evaluated for epilepsy surgery, scalp EEG showed only 33% of seizures in which awareness remained intact and 33% of seizures without outward signs that were recorded at the same time inside the brain. This epilepsy study explains a scalp-recording limit. It does not provide a mixed dementia miss rate. [9]

03

Was the pattern interpreted correctly?

Background slowing and altered rhythms can accompany mixed pathology. A useful report needs to identify whether it found slowing, epileptiform discharges, a seizure, or an uncertain pattern. “Abnormal” alone does not answer the seizure question. [4] [8]

Deep activity may not appear on scalp EEG. In two selected people with Alzheimer's disease, electrodes placed closer to a deep memory area recorded seizures and brief discharges that did not appear on scalp EEG. This shows a possible detection limit in dementia. It does not show how often this occurs in mixed dementia or support invasive monitoring for everyone. [10]

What can a negative EEG tell us?

It means that the reported abnormality was not found during that recording. The meaning depends on whether a familiar event occurred, whether the person slept, the signal quality, the electrode coverage, and the wider clinical history.

A negative result can leave the cause unresolved. The clinical team can decide whether another type or length of recording would be useful.

Research insights

New detection methods still need testing in mixed dementia.

Three approaches may help researchers observe activity for longer or find signals that are difficult to see. Each addresses a different problem, and none has established a screening pathway for mixed dementia.

Tested in Alzheimer's disease | Not yet tested in mixed dementia

Small ear electrodes can support repeated longer recordings

Read the Musaeus study ↗

Musaeus and colleagues studied 24 people with Alzheimer's disease and 15 people without dementia using electrodes in custom earpieces. The Alzheimer's group had more frequent brief seizure-related discharges. Repeated recordings found activity that varied from one session to another. These were brief discharges, not proof that most participants had seizures without outward signs. [11]

Discharges also appeared in nearly half the comparison group during the first recording. Discomfort caused some people to leave the study. The researchers excluded people with larger strokes, several small strokes, or other brain conditions. Those exclusions are especially important when considering mixed dementia.

Seagull's perspective: Repeated recordings may provide a better sample of activity over time. Mixed dementia studies must show whether the recordings remain comfortable, reliable, and useful when vascular or Lewy body disease is also present.

Tested in temporal lobe epilepsy | Not yet tested in mixed dementia

AI may help identify signs of deep activity in scalp EEG

Read the Abou Jaoude study ↗

The HEAnet study trained a computer system using scalp recordings taken at the same time as electrodes closer to a deep memory area. It identified patterns linked with seizure-related activity that experts found difficult to see. The study included people with temporal lobe epilepsy and people without epilepsy, not a mixed dementia group. [12]

The number of missed signals and false detections changed depending on how sensitive the system was set. Finding signs of a deep discharge is also different from finding every seizure.

Seagull's perspective: This approach may find more information in a scalp EEG. Mixed dementia testing must account for general slowing, changed brain rhythms, movement, and other recording artifacts. A system that distinguishes dementia types has not automatically been shown to detect seizures.

Tested in epilepsy | Not yet tested in mixed dementia

A sensor beneath the scalp can record for months

Read the UMPIRE study ↗

The first UMPIRE study placed a recording system beneath the scalp in 26 adults with epilepsy. Among eight people who had seizures visible on regular scalp EEG, all 25 of those seizures also appeared in the beneath-scalp recording. Longer recordings provided information beyond the participants' seizure diaries. [13]

The study did not show whether the device finds every deep seizure that regular scalp EEG cannot see. It was not designed to test benefit in dementia, and the procedure and ongoing device use add practical burdens.

Seagull's perspective: Longer observation may reveal events that are rare or unrecognized. Dementia-specific studies must show that the information changes decisions and improves care enough to justify the burden.

How Seagull evaluates a technology claim

  • What type of activity did it detect?
  • Who was included in the study?
  • How was the result confirmed?
  • How many events were missed or incorrectly identified?
  • How burdensome was the testing?
  • Did using the technology improve care?

Inclusion in this guide is research commentary. It is not a product endorsement or a statement that the technology is available for mixed dementia care.

Why this matters for care

A clear event history can connect care across specialties and settings.

People with mixed dementia may receive care through memory clinics, primary care, stroke services, sleep medicine, epilepsy services, and residential care. Each team needs a clear account of what happened and how it differed from the person's usual pattern.

Keep the event visible across settings

A care partner may see the beginning of an episode. Staff may observe the recovery, while a clinician only receives a report of “increased confusion.” Preserving the sequence, timing, and usual pattern helps every team assess the same event.

Keep treatment goals distinct

Treating established epilepsy aims to reduce seizures and their consequences. Treating isolated discharges to improve cognition is a different research question. The EAN dementia guideline emphasizes individualized seizure management, tolerability, and medication interactions; its drug-selection statement relied on indirect evidence. [14]

Does finding and treating this activity change dementia outcomes?

That is a major clinical question, not an established benefit for mixed dementia. In the 34-participant LEV-AD randomized crossover trial, levetiracetam did not improve the primary cognitive outcome overall. Exploratory analyses found improvements on selected tasks in participants with epileptiform activity. The trial involved Alzheimer's disease, not a defined mixed-dementia population. [15]

These findings support better-targeted trials. Current evidence has not shown that every brief EEG discharge needs treatment or that seizure medication slows mixed dementia. Detecting activity, improving symptoms, controlling seizures, and changing the course of dementia are separate outcomes.

What research needs to answer next

Four questions could improve mixed dementia seizure care.

The evidence supports taking seizures seriously. The next studies need to define the disease combinations, measure what current testing misses, and show whether better detection improves daily life.

  1. Which combinations carry which seizure burden?

    Studies need clearly defined combinations, enough people in each group, and consistent periods of observation. A clinical diagnosis, biomarker evidence, and brain changes confirmed after death should be reported separately.

  2. How much activity does each recording method miss?

    Studies comparing recording methods should count missed outward events, brief discharges, and seizures without outward signs separately. Recording length, sleep, electrode coverage, and interpretation all affect the result.

  3. Which findings affect everyday function?

    Research should link electrical findings with measured attention, memory, sleep, safety, and quality of life, while considering the effects of the underlying diseases and medications.

  4. Who benefits from acting on a finding?

    Trials should test whether a detection-led approach improves meaningful outcomes in defined mixed populations. False-positive diagnoses, medication effects, device burden, and care-partner workload belong in that assessment.

Sources and editorial approach

See the studies behind every claim.

Seagull Health prepared this focused review using recent dementia, seizure, and EEG research. The evidence was reviewed on . This is an educational review rather than a clinical guideline or complete systematic review.

How Seagull selects sources

The core dementia-seizure collection was reviewed alongside targeted primary studies and professional guidance. Every source below was published within the past ten years. The guide identifies whether overlap was based on symptoms, biomarkers, or brain changes confirmed after death.

How Seagull presents the evidence

Clinical group studies show how often events were recorded. Brain-tissue studies identify relationships among diseases, and EEG studies measure specific electrical findings. Guidance and early technology studies answer different questions. The guide places each limitation beside the finding and labels Seagull's perspective separately.

  1. Zawar et al. (2025). The association of seizure control with neuropathology in dementia. Brain. In the Evidence Map →

    Autopsy study of 6,085 cognitively impaired participants. Links seizure-status categories with primary and contributing pathology; does not measure EEG burden or establish treatment effects.

  2. Ting et al. (2026). Neuropathologic Correlates of Seizures in Patients With Alzheimer Disease and Dementia With Lewy Bodies. Neurology. In the Evidence Map →

    Autopsy-confirmed Alzheimer's and Lewy body groups. Informs pathology associations around diagnosis; small Lewy body seizure subgroup and selected pathology criteria limit generalization.

  3. Beagle et al. (2017). Relative Incidence of Seizures and Myoclonus in Alzheimer's Disease, Dementia with Lewy Bodies, and Frontotemporal Dementia. Journal of Alzheimer's Disease. In the Evidence Map →

    Retrospective clinical cohort with a 29-person combined DLB-AD subgroup. Source of the 20.7% subgroup finding; not a general mixed-dementia prevalence estimate.

  4. van der Zande et al. (2018). EEG Characteristics of Dementia With Lewy Bodies, Alzheimer's Disease and Mixed Pathology. Frontiers in Aging Neuroscience. In the Evidence Map →

    Three groups of 41, with Alzheimer's co-pathology defined using cerebrospinal-fluid biomarkers. Twenty-minute EEG study primarily informs background patterns, not silent-seizure burden.

  5. Beniczky et al. (2025). Updated classification of epileptic seizures: Position paper of the International League Against Epilepsy. Epilepsia. In the Evidence Map →

    International consensus terminology. Informs clinical manifestations and the distinction between awareness, responsiveness, and observable signs; not mixed-dementia frequency data.

  6. McKeith et al. (2017). Diagnosis and management of dementia with Lewy bodies: Fourth consensus report of the DLB Consortium. Neurology. In the Evidence Map →

    Consensus report. Informs the overlap with fluctuations, unresponsiveness, syncope, and sleep features when Lewy body disease contributes.

  7. Holtkamp et al. (2017). European Stroke Organisation guidelines for the management of post-stroke seizures and epilepsy. European Stroke Journal. In the Evidence Map →

    Stroke guidance. Informs the distinction between acute symptomatic and later unprovoked seizures; not a mixed-dementia treatment trial.

  8. Hirsch et al. (2021). American Clinical Neurophysiology Society's Standardized Critical Care EEG Terminology: 2021 Version. Journal of Clinical Neurophysiology. In the Evidence Map →

    Consensus definitions separating background activity, discharges, and electrographic or electroclinical seizures. Developed for critical care, not a dementia-screening recommendation.

  9. Casale et al. (2022; online 2020). The Sensitivity of Scalp EEG at Detecting Seizures: A Simultaneous Scalp and Stereo EEG Study. Journal of Clinical Neurophysiology. In the Evidence Map →

    Simultaneous recordings in 27 people with drug-resistant epilepsy undergoing surgical evaluation. Demonstrates scalp detection limits; its percentages cannot be transferred to mixed dementia.

  10. Lam et al. (2017). Silent hippocampal seizures and spikes identified by foramen ovale electrodes in Alzheimer's disease. Nature Medicine. In the Evidence Map →

    Two selected Alzheimer's cases with invasive recordings. Demonstrates deep activity not apparent on scalp EEG; not prevalence or routine-screening evidence.

  11. Musaeus et al. (2023). Detection of subclinical epileptiform discharges in Alzheimer's disease using long-term outpatient EEG monitoring. Neurobiology of Disease. In the Evidence Map →

    Exploratory ear-EEG study in 24 Alzheimer's participants and 15 controls. Limited applicability to mixed disease, with relevant vascular exclusions. Device-company funding and author relationships were disclosed.

  12. Abou Jaoude et al. (2022). Noninvasive Detection of Hippocampal Epileptiform Activity on Scalp Electroencephalogram. JAMA Neurology. In the Evidence Map →

    Machine-learning study in temporal lobe epilepsy and controls. Informs detection of deep epileptiform activity; does not validate mixed-dementia screening or clinical benefit.

  13. Halliday et al. (2025). The UMPIRE study: A first-in-human multicenter trial of bilateral subscalp monitoring for epileptic seizure detection. Epilepsia. In the Evidence Map →

    Device study in 26 adults with epilepsy. Informs subscalp recording feasibility, with scalp EEG as a comparator. Industry participation is disclosed; dementia-specific benefit was not tested.

  14. Frederiksen et al. (2020). A European Academy of Neurology guideline on medical management issues in dementia. European Journal of Neurology. In the Evidence Map →

    Guideline covering medical care and epilepsy in dementia. Its antiseizure drug-selection statement is good-practice guidance based on indirect evidence, not a mixed-dementia comparative trial.

  15. Vossel et al. (2021). Effect of Levetiracetam on Cognition in Patients With Alzheimer Disease With and Without Epileptiform Activity: A Randomized Clinical Trial. JAMA Neurology. In the Evidence Map →

    Small randomized crossover trial in Alzheimer's disease. Primary cognitive outcome was negative; exploratory findings in the epileptiform subgroup support further investigation, not established mixed-dementia benefit.