Skip to content
BioSignal
Condition

Epilepsy & Seizures

A seizure is an event; epilepsy is a disease — and one does not automatically mean the other

Last reviewed
July 2026
Version
1.0
Review cadence
Annually

What it is

A seizure is an event: a burst of abnormally synchronised electrical activity in the brain, producing anything from a few seconds of blank staring to a full convulsion, depending on where it starts and how far it spreads. Epilepsy is a disease: an enduring predisposition to generate unprovoked seizures. The distinction is not pedantry — it is the difference between something that happened to your brain and something about your brain. A seizure that occurs during severe alcohol withdrawal, a dangerously low blood sugar, an acute head injury or a sodium collapse is a provoked seizure: the brain responded to an extreme circumstance more or less as any brain would, and the diagnosis is the circumstance, not epilepsy. The International League Against Epilepsy defines epilepsy practically as two unprovoked seizures more than 24 hours apart, or one unprovoked seizure where the recurrence risk is comparable to that after two (roughly 60% or more over ten years), or a diagnosed epilepsy syndrome. Seizures themselves are also not one thing: they are classified by where they begin — focal (starting in one region), generalised (engaging both sides from the outset), or unknown onset — and by whether awareness is retained. Absence seizures, brief lapses that are often mistaken for daydreaming, are generalised. And not every episode that looks like a seizure is one: psychogenic nonepileptic events are real, distressing, involuntary episodes that arise from a different mechanism entirely and do not respond to antiseizure medication — misdiagnosing them as epilepsy leads to years of ineffective drugs.

Why it matters

Classification is not filing. Where a seizure begins determines what it looks like, what may be causing it, which investigations help and which treatments are appropriate — and getting it wrong sends care in the wrong direction for years. The other reason precision matters is that the popular picture of a seizure is a convulsion on the floor, so the seizures that do not look like that get missed: brief absences dismissed as inattention, focal seizures with retained awareness described as odd feelings, nocturnal events noticed by nobody. Meanwhile the first-aid folklore that surrounds the convulsion people do recognise is actively harmful — the instruction to put something in the mouth causes injuries and cannot prevent the thing it claims to. Most seizures stop on their own within a couple of minutes and need protection rather than intervention; a small proportion do not stop, and that is a genuine emergency measured in minutes. This page exists to separate those two situations, and to do it without either alarming people about ordinary events or reassuring them past the point where help is needed.

The evidence

What BioSignal knows about treating this

Guideline-anchored clinical contextLast reviewed July 2026 · reviewed annually · 11 references

What works for Epilepsy & Seizures

BioSignal’s clinical summary, most important first.

  1. Antiseizure medication — the mainstay; the choice depends on seizure type, sex, age, comorbidity and pregnancy plans, and is a specialist decision rather than a ranking
  2. Treating the provoking cause, where the seizure was provoked — because that is the diagnosis, and antiseizure medication is often not the answer
  3. Trigger management — sleep, alcohol moderation, and above all not missing doses
  4. Epilepsy surgery — for selected drug-resistant focal epilepsy, and under-used relative to its evidence
  5. Vagus nerve stimulation and other neurostimulation — for selected drug-resistant epilepsy
  6. Ketogenic and modified Atkins diets — established in selected, usually paediatric, drug-resistant epilepsy under specialist supervision; not a general dietary claim
  7. Emergency treatment of prolonged seizures — benzodiazepines are first-line, given by trained people or by an agreed rescue-medication plan

Signal Records relevant to this condition

Interventions and contributing factors — some of these records describe a cause rather than a cure. The rating shown is BioSignal’s confidence in that Signal Record, not a claim about how well it treats this condition. Open any of them for the full evidence.

Start Here

New to this? Read these first

  1. FoundationSleepWhy sleep is a non-negotiable pillar that touches every other system.
  2. Signal RecordBenzodiazepinesEffective, appropriate, and genuinely risky — and which of those dominates depends on why, how much, and how long
  3. Body SystemBrainCognition, mood, memory, and neurological health.
  4. ConditionStrokeA brain attack — and largely preventable
Typical Journey

How this usually unfolds

  1. Recognize risk factors
  2. Get diagnosed
  3. Track key biomarkers
  4. Lifestyle first
  5. Evidence-based treatment
  6. Long-term monitoring

An orientation to how this topic is typically approached — not medical advice.

Who is at risk

  • A previous brain insult — stroke, significant head injury, brain infection or neurosurgery
  • Structural brain abnormality on imaging
  • Family history and genetic epilepsy syndromes
  • Febrile seizures in childhood (a modest association; most children who have them never develop epilepsy)
  • Brain tumour
  • Neurodevelopmental conditions
  • Older age — stroke and neurodegeneration make new-onset epilepsy common later in life
  • Sleep deprivation, alcohol, and abrupt withdrawal of antiseizure medication or regular benzodiazepines — triggers in someone already predisposed, and provoking causes in someone who is not

How it's diagnosed

Epilepsy is diagnosed clinically, and the single most valuable investigation is not a machine — it is a description. What happened before, during and after; whether awareness was lost; how long it lasted; what someone watching actually saw. An eyewitness account, or a phone video, frequently settles what an EEG cannot, which is why bringing one to the appointment is genuinely useful advice. Tests support that history rather than replace it. An EEG can show epileptiform activity and helps classify the epilepsy and estimate recurrence risk — but a normal routine EEG does not exclude epilepsy, because a routine recording samples a short window between seizures and most people with epilepsy have normal brain activity most of the time. Treating a normal EEG as a negative result is one of the commonest misunderstandings in this area. MRI looks for a structural cause and matters most in focal epilepsy and new-onset seizures in adults. Blood tests, ECG and toxicology look for provoked causes and for the imitators — a faint from a cardiac arrhythmia can produce a few convulsive jerks and be mistaken for a seizure, which is a mistake that runs in the dangerous direction. Prolonged video-EEG is what distinguishes epileptic seizures from psychogenic nonepileptic events when the two are genuinely in question.

  • A detailed history — the account of the event, and an eyewitness description or video, outweigh most tests
  • EEG — supports diagnosis, classification and recurrence risk; a normal routine EEG does NOT rule out epilepsy
  • MRI brain — for a structural cause, particularly in focal or adult-onset seizures
  • Blood glucose, sodium, calcium, and toxicology — to identify a provoked seizure
  • ECG — to catch the cardiac causes of collapse that imitate seizures
  • Prolonged video-EEG — the test that separates epileptic seizures from psychogenic nonepileptic events
Most important

Key biomarkers

Biomarker pages for this condition are on the roadmap.

Day to day

Lifestyle

  • Take the medication consistently — missed doses are among the most common reasons a controlled epilepsy stops being controlled
  • Never stop an antiseizure medication abruptly; withdrawal of antiseizure drugs, and of regular benzodiazepines, can provoke seizures including prolonged ones
  • Protect sleep — deprivation is a well-recognised trigger, which is why shift patterns and all-nighters matter more here than as general advice
  • Be careful with alcohol, and particularly with stopping heavy drinking suddenly, which provokes seizures in its own right
  • Tell any prescriber you have epilepsy — interactions run both ways, and several antiseizure drugs reduce the effectiveness of hormonal contraception
  • If you could become pregnant, raise it BEFORE conceiving — this is the single highest-value conversation in epilepsy care, and it is one that needs planning rather than reaction
  • Driving is restricted after a seizure, but the rules differ by country and state — check your own licensing authority, because this page cannot tell you your law
  • Consider safety in ordinary settings — baths and swimming are where seizures most often turn fatal, and a shower or supervised swimming is a reasonable adaptation rather than a life restriction
Explore

Explore this condition across BioSignal

Related Foundations

SleepBrain & Cognitive Health

Related Signal Records

BenzodiazepinesAlcohol

Related body systems

Brain

Related conditions

StrokeMigraineAnxiety DisordersDepressionInsomniaAlzheimer's Disease

Frequently asked questions

Should you put something in the mouth of someone having a seizure, so they don't swallow their tongue?

No. Never. This is the most persistent and most harmful piece of first-aid folklore there is, and both halves of it are wrong. You cannot swallow your tongue — it is anchored to the floor of the mouth and is anatomically incapable of being swallowed. And putting an object, or fingers, into the mouth of someone having a convulsion causes real injuries: broken teeth, lacerations, a bitten rescuer, and objects that themselves become a choking hazard. Do not do it. Do not restrain them either — you cannot stop the seizure by holding someone down, and you can cause fractures and dislocations trying. What actually helps is unglamorous: note the time, ease them to the ground if they are falling, put something soft under the head, move hard or sharp objects out of the way, loosen anything tight at the neck, and once the convulsion stops, roll them onto their side to keep the airway clear. Stay with them until they are properly back to themselves, and be calm about the confusion and drowsiness afterwards, which is normal and can last a while. Call for emergency help if the seizure lasts around five minutes or more, if another follows without recovery in between, if they are injured, are having trouble breathing, are in water, are pregnant, do not regain consciousness, or if it is their first seizure.

Does having a seizure mean I have epilepsy?

No — and this is the distinction the whole page turns on. Around one person in ten will have a seizure at some point, while epilepsy affects far fewer, because most single seizures do not mean the brain has an enduring tendency to produce them. A provoked seizure has a cause that would push most brains over the threshold: severe alcohol withdrawal, a very low blood sugar, an acute head injury, a serious infection, a sodium or calcium collapse, certain drugs and overdoses. In that situation the diagnosis is the provoking cause, treating it is the priority, and long-term antiseizure medication is often not the answer at all. Epilepsy means unprovoked seizures with an enduring predisposition to more — formally, two unprovoked seizures more than 24 hours apart, or one unprovoked seizure when the risk of another is as high as it would be after two (about 60% or more over ten years), or a recognised epilepsy syndrome. That definition exists precisely so a single event is not treated as a life sentence, and so a person whose risk genuinely is that high is not made to wait for a second seizure to be taken seriously.

My EEG was normal — does that mean I don't have epilepsy?

No, and this misunderstanding causes real harm in both directions. A routine EEG records for a short period, usually while you are awake and between seizures. Epilepsy is a tendency to have seizures, not a permanent abnormality visible at all times, so most people with epilepsy have a normal brain recording most of the time. A normal routine EEG therefore does not exclude epilepsy — it simply did not catch anything during the window it was watching. What an EEG is genuinely good for is the other direction: when it does show epileptiform activity, that supports the diagnosis, helps classify the seizure type, and meaningfully raises the estimated risk of another seizure — it is one of the factors that shifts the decision about starting medication after a first event. If the diagnosis remains uncertain, longer or sleep-deprived recordings, or prolonged video-EEG, catch more. The rule of thumb worth carrying is that in epilepsy, an abnormal EEG is informative and a normal one is not reassuring — the diagnosis lives in the history.

I've had one seizure. Do I have to start medication?

Not necessarily, and this is a genuine decision rather than an automatic one. The American Academy of Neurology's guideline sets out the trade-off honestly. After a first unprovoked seizure, the risk of another is highest in the first two years and sits somewhere between 21% and 45% — a wide range, because it depends on you. Four things push you toward the higher end: a prior brain insult such as a stroke or significant head injury, an EEG showing epileptiform abnormalities, a meaningful abnormality on brain imaging, and a seizure that happened during sleep. Starting medication immediately, rather than waiting for a second seizure, is likely to reduce recurrence within those first two years. But — and this is the part that usually goes unsaid — it may not improve quality of life, and beyond about three years it is unlikely to improve the long-term outlook as measured by sustained seizure remission. So the drug buys a reduction in short-term recurrence rather than a better ultimate outcome, and what that is worth depends on your circumstances: what a seizure would cost you at work, whether you drive, how you feel about the side effects. That is a conversation, not a formula, and it is one worth having properly.

When is a seizure an emergency?

Most seizures are not, and that is worth saying first: a typical convulsive seizure stops on its own within about one to two minutes, and what follows — confusion, drowsiness, sore muscles, a bitten tongue — is expected rather than alarming. The emergency is when it does not stop. The International League Against Epilepsy set the operational threshold for convulsive status epilepticus at five minutes, which is the point beyond which a seizure is unlikely to stop by itself and treatment should begin; the ILAE was explicit that the underlying evidence is incomplete and that this is a best estimate rather than a precise biological boundary. Call emergency services if a convulsive seizure lasts around five minutes or longer, if a second seizure begins without recovery in between, if breathing is difficult or the person stays blue, if they do not regain consciousness, if they are injured, if the seizure happened in water, if the person is pregnant, if it is their first seizure, or if you suspect an overdose, poisoning, serious infection, stroke or a metabolic cause. Benzodiazepines are the first-line emergency treatment and work best given early — which is the reason for the clock rather than a reason to panic at two minutes.

I have epilepsy and I want to get pregnant — what do I need to know?

That this is the conversation to have BEFORE you conceive, not after — and that it involves a real trade-off rather than a simple answer. The prospective EURAP data, covering more than 7,000 pregnancies, put numbers on it: major congenital malformations occurred in 10.3% of pregnancies exposed to valproate, compared with 5.5% for carbamazepine, 2.9% for lamotrigine and 2.8% for levetiracetam, and the risk rose with dose for several drugs including valproate. Fetal valproate exposure has also been associated with lower cognitive outcomes in childhood. That is why valproate is avoided wherever possible in anyone who could become pregnant, and why regulators have restricted it. Here is the uncomfortable part that honest coverage has to include: valproate is also, for generalised epilepsy, more effective than the main alternative — in the SANAD II trial, levetiracetam was neither clinically effective nor cost-effective compared with valproate, and the authors framed their results as informing precisely this discussion about the benefit and harm of avoiding it. So this is not a choice between a good drug and a bad one; it is a decision about two real risks, one to the pregnancy and one from uncontrolled seizures, which are themselves dangerous. Do not stop or change medication on your own on the strength of this page. Several antiseizure drugs also reduce the effectiveness of hormonal contraception, which is a separate and immediate reason to have the conversation early.

Can I drive?

Not immediately after a seizure — and beyond that, BioSignal cannot tell you, because the answer is legal rather than medical and it genuinely differs depending on where you live. Every jurisdiction restricts driving after a seizure, but the required seizure-free period, the rules for provoked versus unprovoked events, the treatment of seizures that occur only during sleep, the position on medication changes, and whether the reporting duty falls on you or your doctor all vary by country and, in the United States, by state. Anyone who gives you a single confident number for this is telling you about somewhere, not necessarily about you. The two things that are universally true: check with your own licensing authority or your neurologist, who will know your local rule; and do not simply stop taking medication in order to shorten the wait, because that is one of the most reliable ways to provoke the seizure that restarts the clock.

What are non-epileptic seizures — is it 'all in the mind'?

No, and that framing is both wrong and harmful. Psychogenic nonepileptic seizures are episodes that look like epileptic seizures but do not arise from the abnormal electrical discharge that defines epilepsy. They are not faked, not deliberate, and not under the person's control — they are involuntary events with a different mechanism, frequently linked to psychological trauma or distress, and they cause genuine disability. Two things make them matter here. First, they are common enough to be a leading reason for referral to specialist centres, and they are frequently misdiagnosed as epilepsy — which means years of antiseizure medication that cannot work, with all its side effects, while the treatment that could help is never started. Prolonged video-EEG, which records the brain during an actual event, is what settles the question. Second, the two can coexist in the same person, which is exactly why this is a diagnosis for a specialist rather than a conclusion to reach from a description. The right response to this diagnosis is not embarrassment or disbelief; it is a referral to people who treat it, because it is treatable.

Evidence summary

A seizure is an event and epilepsy is a disease characterised by an enduring predisposition to unprovoked seizures — operationally defined by the ILAE as two unprovoked seizures more than 24 hours apart, one unprovoked seizure with a recurrence risk comparable to that after two (approximately 60% over ten years), or a diagnosed epilepsy syndrome. Provoked seizures (alcohol withdrawal, hypoglycaemia, acute brain insult, metabolic derangement, drugs) are not epilepsy, and their management is the management of the provoking cause. Seizures are classified by onset — focal, generalised, or unknown — under the ILAE operational classification (2017, updated 2025), and classification drives investigation and treatment. Diagnosis is clinical: the history and an eyewitness or video account outweigh testing, and a normal routine EEG does not exclude epilepsy because it samples a short interictal window; an epileptiform EEG is informative in the positive direction and is one of four factors (with prior brain insult, imaging abnormality and nocturnal onset) that raise recurrence risk after a first seizure. The AAN guideline quantifies that first-seizure decision honestly: recurrence risk is greatest in the first two years at 21-45%, immediate antiseizure medication is likely to reduce recurrence within that window but may not improve quality of life, and beyond three years is unlikely to improve sustained remission — a reduction in short-term events rather than a better long-term outcome. Drug selection is a specialist decision that this page does not make, but the central tension is documented: prospective EURAP data across 7,355 pregnancies found major congenital malformations in 10.3% with valproate versus 5.5% carbamazepine, 2.9% lamotrigine and 2.8% levetiracetam, dose-dependent for several agents, and fetal valproate exposure is associated with poorer childhood cognitive outcomes — while SANAD II found levetiracetam neither clinically effective nor cost-effective compared with valproate for generalised epilepsy, so avoiding valproate carries its own cost that must be weighed rather than assumed away. Convulsive status epilepticus is operationally defined at five minutes (ILAE t1), with the ILAE noting the evidence is incomplete and the threshold a best estimate; benzodiazepines are first-line emergency treatment. First aid is protective, not interventional: nothing in the mouth, no restraint — the tongue cannot be swallowed, and both actions cause injury. Psychogenic nonepileptic events are involuntary, are not epilepsy, do not respond to antiseizure medication, and are distinguished by prolonged video-EEG. SUDEP is a recognised risk that guidelines address. This page covers seizures and epilepsy; syncope, stroke, migraine, panic and tremor are separate, and febrile seizures in children are not covered here.

References & sources

  • Fisher RS, Acevedo C, Arzimanoglou A, et al. ILAE official report: a practical clinical definition of epilepsy. Epilepsia 2014;55(4):475-482 (PMID 24730690; DOI 10.1111/epi.12550)
  • Fisher RS, Cross JH, French JA, et al. Operational classification of seizure types by the International League Against Epilepsy: position paper of the ILAE Commission for Classification and Terminology. Epilepsia 2017;58(4):522-530 (PMID 28276060; DOI 10.1111/epi.13670)
  • Beniczky S, Trinka E, Wirrell E, et al. Updated classification of epileptic seizures: position paper of the International League Against Epilepsy. Epilepsia 2025;66(6):1804-1823 (PMID 40264351; DOI 10.1111/epi.18338)
  • Trinka E, Cock H, Hesdorffer D, et al. A definition and classification of status epilepticus — report of the ILAE Task Force on Classification of Status Epilepticus. Epilepsia 2015;56(10):1515-1523 (PMID 26336950; DOI 10.1111/epi.13121)
  • Krumholz A, Wiebe S, Gronseth GS, et al. Evidence-based guideline: management of an unprovoked first seizure in adults. Neurology 2015;84(16):1705-1713 (PMID 25901057; DOI 10.1212/WNL.0000000000001487)
  • Glauser T, Shinnar S, Gloss D, et al. Evidence-based guideline: treatment of convulsive status epilepticus in children and adults — report of the Guideline Committee of the American Epilepsy Society. Epilepsy Curr 2016;16(1):48-61 (PMID 26900382; DOI 10.5698/1535-7597-16.1.48)
  • Marson A, Burnside G, Appleton R, et al. (SANAD II). The SANAD II study of the effectiveness and cost-effectiveness of levetiracetam, zonisamide, or lamotrigine for newly diagnosed focal epilepsy: an open-label, non-inferiority, multicentre, phase 4, randomised controlled trial. Lancet 2021;397(10282):1363-1374 (PMID 33838757; DOI 10.1016/S0140-6736(21)00247-6)
  • Marson A, Burnside G, Appleton R, et al. (SANAD II). The SANAD II study of the effectiveness and cost-effectiveness of valproate versus levetiracetam for newly diagnosed generalised and unclassifiable epilepsy: an open-label, non-inferiority, multicentre, phase 4, randomised controlled trial. Lancet 2021;397(10282):1375-1386 (PMID 33838758; DOI 10.1016/S0140-6736(21)00246-4)
  • Tomson T, Battino D, Bonizzoni E, et al. (EURAP Study Group). Comparative risk of major congenital malformations with eight different antiepileptic drugs: a prospective cohort study of the EURAP registry. Lancet Neurol 2018;17(6):530-538 (PMID 29680205; DOI 10.1016/S1474-4422(18)30107-8)
  • Meador KJ, Baker GA, Browning N, et al. (NEAD Study Group). Fetal antiepileptic drug exposure and cognitive outcomes at age 6 years: a prospective observational study. Lancet Neurol 2013;12(3):244-252 (PMID 23352199; DOI 10.1016/S1474-4422(12)70323-X)
  • Harden C, Tomson T, Gloss D, et al. Practice guideline summary: sudden unexpected death in epilepsy incidence rates and risk factors. Neurology 2017;88(17):1674-1680 (PMID 28438841; DOI 10.1212/WNL.0000000000003685)

Educational information — not medical advice

Condition pages orient you across the evidence; they don't diagnose or treat. Diagnosis and management belong with a qualified clinician. See our Medical Disclaimer.

Is this condition page clear, accurate, and useful? Your feedback shapes what we review next.

Give feedback