Understanding events and EEG
Some seizures cause shaking. Others cause a brief change in response, speech, memory, sensation, or movement. An EEG may also record seizure-related activity when no outward change is seen. Knowing exactly what happened and what the EEG recorded makes the result easier to understand.
Start with the finding
An EEG uses electrodes to record the brain's electrical activity. The phrase “abnormal EEG” does not tell us whether a seizure occurred. A useful report identifies the type of activity that was found. [1] [2]
A pause, stare, jerk, or language change is an observation. The person's history, witness accounts, examination, and selected tests help a clinician decide what caused it. Epilepsy can be diagnosed even when an EEG is normal because the EEG is only one part of the assessment. [3]
Myoclonus means a brief, shock-like jerk and can have several causes. The DLB guide explains why sudden jerks and seizures must be counted separately.
Understanding a negative recording
Longer EEG recordings provide more time to capture activity that comes and goes, including during sleep. Activity beginning in a small or deep brain area may still produce little or no visible signal at the scalp.
Activity that comes and goes may not occur during a short EEG. Sleep and repeated sessions can change what is found. In an Alzheimer's ear EEG study, the discharges recorded in one person varied substantially between sessions. [6]
Activity beginning in a small or deep brain area may not produce a clear scalp pattern. Studies recording from the scalp and inside the brain at the same time have shown this limit, including in selected people with Alzheimer's disease. [4] [5]
Muscle activity, electrode problems, normal variants, and abnormal background rhythms can complicate review. Definitions and reviewer agreement matter alongside recording length. [2] [7]
Alzheimer's study | Five selected people
Devulder and colleagues placed electrodes closer to deep temporal brain areas. At least 95% of the brief discharges recorded by those electrodes did not appear on scalp EEG. Two focal seizures in one person also did not appear on the scalp recording. [4]
What the number means: The 95% figure describes brief discharges in a small, selected study. It is not a general estimate of seizures missed by longer EEG across dementia.
Epilepsy study | 27 people evaluated for surgery
Casale and colleagues recorded from the scalp and inside the brain at the same time. Scalp EEG showed 33% of seizures in which awareness remained intact and 33% of seizures without outward signs. It showed 97% of focal seizures that affected awareness. [5]
What the number means: The study demonstrates a scalp-recording limit in a selected epilepsy group. It does not measure EEG performance in dementia.
It means that this activity was not identified under the recording conditions. Interpretation depends on the clinical history, whether a typical episode was captured, sleep, duration, signal quality, and electrode coverage.
NICE advises against using EEG to exclude epilepsy. Additional recording may be considered when uncertainty remains, but this is not a recommendation to screen every person with dementia or to use invasive electrodes routinely. [3]
A negative result does not prove that hidden seizures are occurring, either. Both overconfidence in a negative test and automatic assumptions of invisible seizures can mislead.
Reading the research accurately
A study may report the percentage of people with a finding under its recording conditions. To measure how reliably a test detects activity, researchers need another dependable method to show which activity was truly present. Comparing a longer wearable recording with a short routine EEG changes both the equipment and the amount of time recorded.
During the first recording, Musaeus's Alzheimer's ear EEG study found brief discharges in 18 of 24 people with Alzheimer's disease and 7 of 15 people without dementia. The difference in those proportions was not statistically clear, although discharges occurred more often in the Alzheimer's group. [6]
Research insights
The Research insights page explains what each study found, how it might apply to dementia, and what remains unknown. Each approach addresses a different recording problem.
Repeated ear-EEG in a selected Alzheimer's population. [6]
Read the home-recording insight →Computer-assisted discharge detection in epilepsy research. [8]
Read the signal-detection insight →Subscalp recording in an epilepsy study. [9]
Read the longer-term monitoring insight →Keep the purpose clear. These insights evaluate research rather than recommend a device for one person. A method can detect more activity before researchers know whether using it improves care.
From detection to benefit
Treating diagnosed epilepsy and treating brief discharges to improve thinking are different questions. In a trial of 34 people with Alzheimer's disease, levetiracetam did not improve the main thinking measure across the full group. Additional analyses found improvement on selected tasks among people with seizure-related electrical activity. The result supports more research. It has not established that every EEG discharge should be treated or that the findings apply across dementia types. [10]
Sources and scope
The sources below distinguish dementia research, related epilepsy research, and clinical terminology. The five subtype guides provide the wider disease-specific evidence.
Beniczky et al. (2025). Updated classification of epileptic seizures: Position paper of the International League Against Epilepsy.
Consensus terminology for seizures and clinical manifestations; not a dementia-specific prevalence study.
Hirsch et al. (2021). American Clinical Neurophysiology Society's Standardized Critical Care EEG Terminology: 2021 Version.
Critical-care EEG definitions help distinguish seizure patterns from other abnormalities. They do not determine an individual's diagnosis in isolation.
NICE (updated 2025). Epilepsies in children, young people and adults: Diagnosis and assessment of epilepsy.
General epilepsy guidance, including the limits of a negative EEG. Not a recommendation for universal dementia screening.
Devulder et al. (2025; published online 2024). Epileptic activity on foramen ovale electrodes is associated with sleep and tau pathology in Alzheimer's disease.
Five selected Alzheimer's cases with deep and scalp recordings. Demonstrates scalp-invisible activity, not its population prevalence.
Casale et al. (2022; published online 2020). The Sensitivity of Scalp EEG at Detecting Seizures: A Simultaneous Scalp and Stereo EEG Study.
Simultaneous recordings in 27 people undergoing epilepsy surgical evaluation. Detection percentages must not be transferred directly to dementia.
Musaeus et al. (2023). Detection of subclinical epileptiform discharges in Alzheimer's disease using long-term outpatient EEG monitoring.
Ear-EEG study of 24 Alzheimer's patients and 15 controls. Discharge detection, not silent-seizure prevalence; company support and author relationships disclosed.
Szabo et al. (2026). Refining Detection of Subclinical Epileptiform Activity in Alzheimer's Disease: A Case-Control Study and Call for a Consensus.
Overnight case-control study, 30 participants per group. Highlights the influence of recording and interpretation methods.
Abou Jaoude et al. (2022). Noninvasive Detection of Hippocampal Epileptiform Activity on Scalp Electroencephalogram.
Computer-assisted detection of deep-activity signatures in temporal lobe epilepsy research, not clinical validation in dementia.
Halliday et al. (2025). The UMPIRE study: A first-in-human multicenter trial of bilateral subscalp monitoring for epileptic seizure detection.
First-in-human epilepsy study with 26 implanted adults. Scalp-visible comparison seizures are not a reference for all hidden activity. Industry involvement disclosed.
Vossel et al. (2021). Effect of Levetiracetam on Cognition in Patients With Alzheimer Disease With and Without Epileptiform Activity: A Randomized Clinical Trial.
Thirty-four-participant crossover trial. The overall primary outcome was negative; findings in an epileptiform subgroup were exploratory.