Can Epilepsy Be Fatal? SUDEP Explained, the Risk Factors and How to Reduce Them

A mother once told a neurologist that she had read about SUDEP on a forum at two in the morning, then spent the rest of the night sitting on the floor outside her teenage son’s bedroom, listening for his breathing. He was fine. He had been fine for months. But nobody had ever said the word to her out loud, and the silence had made it louder.

That is the strange position many families find themselves in. Epilepsy is one of the most common neurological conditions on earth, yet the question of whether it can kill you tends to be answered either with a shrug or with a headline. Neither helps you decide what to do on a Tuesday.

So here is the honest version: what actually causes death in people with epilepsy, how often it happens, who carries the highest risk, and which everyday measures have real evidence behind them. Some of it is reassuring. Some of it is a call to action. All of it is better than guessing.

Is epilepsy itself deadly, or is that a myth?

Epilepsy is not a terminal diagnosis, and most people who live with it will die of something unrelated, decades later. Yet the condition does carry a measurable shadow. The World Health Organization estimates that around 50 million people worldwide have epilepsy and that the risk of premature death is up to three times higher than in the general population. That figure is real, and it deserves a clear explanation rather than a soothing dismissal.

Three things sit behind it. The first is the underlying cause: epilepsy that follows a stroke, a brain tumor, or a serious head injury shares the mortality of those conditions. The second is the seizures themselves, which can lead to drowning, falls, burns, and car crashes. The third is a group of deaths that happen during or shortly after a seizure with no other explanation, the phenomenon known as SUDEP.

The gap between a person whose seizures are fully controlled and one who has several convulsive seizures a month is enormous. In practice, the phrase “epilepsy can be fatal” describes a spectrum, and where you sit on it depends heavily on seizure type, seizure frequency, and how consistently treatment is followed. The WHO also notes that up to 70% of people with epilepsy could become seizure-free with appropriate diagnosis and treatment, which reframes the whole question: much of the risk is not fixed.

What is the leading cause of death in epilepsy?

Ask a room of clinicians and you will hear two answers, and both are correct depending on the group you describe. Across everyone with epilepsy, deaths from the underlying cause, from cardiovascular disease, and from cancer outnumber deaths directly triggered by seizures. For people whose seizures are not controlled, the picture changes sharply. The Centers for Disease Control and Prevention describes SUDEP as the leading cause of death in people with uncontrolled seizures.

Behind SUDEP come several other seizure-related causes. Status epilepticus, a seizure that will not stop on its own, can damage the brain and body and is treated as a medical emergency. Drowning is disproportionately common, because a seizure in a bathtub, pool, or lake removes a person’s ability to keep their head above water. Falls and traffic injuries follow, along with aspiration when vomit or saliva enters the lungs during a seizure.

Suicide and self-harm are also elevated among people with epilepsy, partly because depression and anxiety are common companions to the condition and partly because of social isolation and stigma. This part is rarely mentioned in a first appointment, which is a missed opportunity, since mood is treatable and asking about it costs nothing.

The takeaway for a family is not a ranking table to memorize. It is that most of the seizure-related causes cluster around the same root: seizures that keep happening, especially the convulsive kind.

What is SUDEP, exactly?

SUDEP stands for sudden unexpected death in epilepsy. The definition is deliberately narrow. It refers to a person with epilepsy who dies suddenly, often but not always after a seizure, when no injury, no drowning, no status epilepticus, and no other cause is found, even after an autopsy. If a cause is found, it is not SUDEP.

Because it is a diagnosis of exclusion, researchers grade cases by how thoroughly other explanations were ruled out. “Definite” SUDEP means an autopsy was performed and found nothing else. “Probable” SUDEP fits the pattern but lacks an autopsy. “Possible” SUDEP means a competing cause could not be excluded. This grading matters because it explains why estimates vary between studies and why the true number is likely higher than official records suggest. A death certificate that says “natural causes” may hide a SUDEP case.

The typical scene, as the Cleveland Clinic and CDC describe it, is quiet rather than dramatic. The person is usually found in bed, frequently lying face down, with signs that a tonic-clonic seizure occurred shortly before. There is rarely a witness. That is part of what makes SUDEP so hard to study and so frightening to families who hear about it for the first time.

The word “unexpected” carries weight. SUDEP is not a slow decline. It is a sudden event in someone who was, minutes earlier, sleeping normally.

How common is SUDEP? Putting the numbers in perspective

Numbers are where fear either shrinks or grows, so they should be exact. The CDC estimates that roughly 1 in 1,000 adults with epilepsy die from SUDEP each year. In children the annual rate is lower, at about 1 in 4,500. The American Academy of Neurology and American Epilepsy Society guideline, published through the NIH’s PubMed Central, arrived at closely matching figures: about 0.22 per 1,000 children and 1.2 per 1,000 adults per year.

Group	Approximate annual SUDEP rate	Source
Children with epilepsy	About 1 in 4,500	CDC; AAN/AES guideline
Adults with epilepsy (all)	About 1 in 1,000	CDC; AAN/AES guideline
Adults with frequent tonic-clonic seizures	Substantially higher (see risk factors below)	AAN/AES guideline

One in 1,000 per year sounds small in a single year and larger over a lifetime, which is why both perspectives are needed. For a person who becomes seizure-free, the yearly risk falls toward the background risk of the general population. For someone with three or more convulsive seizures a year, it rises many times over. The average hides both ends.

There is one more reason the number is useful: it sets a scale for weighing safety measures. A precaution that reduces a 1-in-1,000 risk by half is worthwhile. A precaution that reshapes a family’s entire life around a risk already near zero is not, and clinicians increasingly say so.

Why does SUDEP happen? What the evidence actually shows

The honest answer is that the mechanism is not fully understood, but the outline has become much clearer over the past decade. Studies of people who died while being monitored in hospital epilepsy units, cited by the Cleveland Clinic and the CDC, show a fairly consistent sequence. A generalized tonic-clonic seizure ends. Breathing becomes shallow or stops. The heart rhythm slows or becomes erratic. Within minutes, both breathing and heartbeat cease. The brain’s electrical activity, meanwhile, goes almost silent after the seizure, a state sometimes called postictal suppression.

Breathing appears to be the weak link. The brainstem regions that drive automatic breathing and that respond to rising carbon dioxide seem to be temporarily switched off after a convulsive seizure. Lying face down into a pillow may worsen this by limiting airflow and trapping exhaled air. Some researchers describe SUDEP as a failure of the body’s alarm system: the signals that would normally make a sleeping person gasp, turn, or wake do not fire.

Heart rhythm problems play a role in some cases, and there is interest in shared genetic factors between epilepsy and certain inherited heart conditions. Whether these are the primary cause or a downstream effect of oxygen deprivation is still debated.

What this means practically is that anything reducing the number of tonic-clonic seizures, and anything ensuring someone is present to reposition and stimulate a person afterward, targets the two points in the chain where intervention appears possible.

Can epilepsy cause death in sleep?

Yes, and this is the pattern that unsettles people most. The majority of SUDEP deaths occur at night or during sleep, and the person is typically found in bed. The CDC and Cleveland Clinic both note that many are discovered lying face down.

Why sleep? Several threads come together. Seizures themselves are more likely during certain sleep stages and in the transition between sleep and waking for many people. Nobody is watching, so the simple interventions that a bystander would perform in daytime, such as turning the person on their side or calling their name, do not happen. And the breathing suppression that follows a convulsive seizure is more dangerous when a face is pressed into soft bedding.

It is worth separating two ideas that often blur together. Having seizures during sleep is common and, on its own, not a death sentence. The risk concentrates in nighttime seizures that are tonic-clonic, unsupervised, and frequent. A person whose only seizures are brief focal episodes while asleep is in a very different position from someone having a convulsive seizure alone every few weeks.

This is why the AAN/AES guideline singled out nocturnal supervision as a factor associated with lower risk, and why questions about sleeping arrangements, roommates, and monitoring belong in a routine epilepsy appointment rather than a crisis.

Who is at risk of SUDEP?

Risk is not evenly spread. The AAN/AES guideline reviewed the evidence and found one factor dwarfing all others: the frequency of generalized tonic-clonic seizures. Other factors carry weaker or more mixed evidence, but they still shape how clinicians think about an individual.

Risk factor	What the evidence shows
Three or more tonic-clonic seizures per year	Roughly 15-fold higher risk compared with people who have none (AAN/AES guideline)
Not being seizure-free for the past 1 to 5 years	Associated with higher risk; seizure freedom is protective
Nighttime seizures without supervision	Associated with higher risk; presence of another person is associated with lower risk
Missed or irregular medicine	Frequently identified in SUDEP cases; increases seizure frequency
Long duration of epilepsy, onset in childhood	Modest association in several studies
Intellectual disability, alcohol or substance misuse	Reported associations, evidence less consistent

Notice what is missing from the table: sex, ethnicity, and age are weak or inconsistent predictors compared with seizure frequency. That should redirect attention. The most useful question is not “what kind of person am I” but “how many convulsive seizures am I having, and are any of them happening alone at night.”

There is a hopeful corollary. Because the dominant risk factor is modifiable, a change in treatment that turns monthly convulsive seizures into none does more to lower SUDEP risk than any device or bedtime ritual.

Does the type of seizure matter?

Enormously. Epilepsy is not one condition but a family of them, and the seizure types within that family carry very different risks. Focal seizures with retained awareness, brief absence seizures, and myoclonic jerks are disruptive and deserve treatment, but they are not the seizures found again and again in SUDEP investigations.

Generalized tonic-clonic seizures are. These are the convulsive events most people picture: loss of consciousness, stiffening, rhythmic jerking, often lasting one to three minutes, followed by a period of deep unresponsiveness. The AAN/AES guideline’s headline finding was that people having three or more of these a year faced about 15 times the SUDEP risk of people having none. Even one or two a year raised the risk compared with zero.

The reason traces back to mechanism. A tonic-clonic seizure places heavy demand on the heart and lungs, floods the body with stress hormones, and is followed by that dangerous window of suppressed brain activity and shallow breathing. Smaller seizures do not produce the same cascade.

This distinction has a practical edge. When a clinician asks how your seizures have been, “about the same” is not enough. Counting convulsive seizures specifically, and noting whether they clustered at night, gives the care team the single most decision-relevant piece of information about your risk. A seizure diary that separates types is worth more than a general impression.

Status epilepticus: the other emergency

SUDEP takes most of the headlines, but a second seizure-related emergency is more common and, crucially, more visible. Status epilepticus describes a seizure that does not stop on its own or a series of seizures without recovery in between. The Mayo Clinic and NHS both use a simple threshold for the public: a convulsive seizure lasting more than five minutes should be treated as an emergency.

Why five minutes? Most tonic-clonic seizures end within one to three minutes. Once a seizure passes the five-minute mark, the odds that it will stop by itself fall and the odds of complications rise. Prolonged convulsions strain the heart, drop oxygen levels, and can raise body temperature and damage muscle. Left untreated, status epilepticus can cause lasting brain injury or death.

Treatment exists and works best when started early. Emergency teams use rescue medicines that calm overactive brain circuits, and hospital care can escalate from there. Some people with a history of prolonged seizures are prescribed a rescue medicine for family members to give at home, with a plan written by their neurologist; when and how to use it is a decision for the prescribing clinician, not a blog post.

The message for bystanders is unambiguous: note the time when a seizure starts, and if it passes five minutes, call emergency services. Waiting to see if it will pass is the wrong instinct here.

Drowning, falls and the everyday risks that are easy to forget

Not every epilepsy-related death is mysterious. Some are painfully ordinary, and because they are ordinary they are among the most preventable.

Water is the clearest example. A seizure in a bathtub can be fatal in a few inches of water, since the person cannot lift their head. The NHS, Mayo Clinic, and Cleveland Clinic all recommend showering rather than bathing when alone, swimming only with someone who knows about your epilepsy and can pull you out, and avoiding open water without a companion.

Heights and heat come next. Falls from ladders, stairs, or bicycles during a seizure cause head injuries and fractures. Cooking is a source of burns, which is why some people use back burners, microwave ovens, or kettles with automatic shut-off. None of this requires living in a padded room; it requires thinking through the two or three moments in a normal day where a sudden loss of awareness would be most dangerous.

Driving deserves its own sentence. Rules vary by country and by US state, but they share a principle: a documented seizure-free period is required before driving, precisely because a seizure at the wheel endangers others as well as the driver. Following those rules is not bureaucracy; it is one of the few epilepsy safety measures with a clear public-health rationale.

These risks fall away almost entirely when seizures are controlled, which brings the discussion back to treatment.

How to reduce the risk: seizure control comes first

If SUDEP risk tracks the number of convulsive seizures, then the most powerful safety measure is not a gadget or a sleeping position. It is achieving the lowest seizure count possible, and the evidence supports that view.

Anti-seizure medicines work by steadying the electrical excitability of brain cells, through various mechanisms such as modulating the flow of sodium or calcium into neurons or enhancing the brain’s natural inhibitory signaling. Finding the right one is often a process rather than a single decision. Many people need a period of weeks to months to reach a stable dose and judge the effect, and some need more than one attempt. That timeline is normal, not a sign of failure.

Missed doses are a recurring finding in SUDEP investigations, and the reason is straightforward: skipping medicine invites breakthrough seizures, and breakthrough seizures are often tonic-clonic. Pill organizers, phone reminders, and a blister-pack pharmacy service are unglamorous tools with a real safety payoff. Any change to how medicine is taken, including stopping because you feel well, should go through the prescribing clinician.

When two well-chosen medicines have failed to control seizures, the condition is described as drug-resistant, and the NIH and Cleveland Clinic recommend referral to a specialist epilepsy program. Options then can include surgery to remove the seizure focus, implanted stimulation devices, or dietary therapy. Whether any of these is right for a particular person is a decision for the treating team, but the question should at least be asked, because continuing uncontrolled seizures is itself a risk.

Sleep, supervision and seizure-alert devices: what is proven and what is not

Once seizure control is being addressed, the second lever is the nighttime environment, and here the evidence is more mixed, so it deserves honest sorting.

Supervision has the strongest support. The AAN/AES guideline found that having another person in the bedroom, or a system of regular nighttime checks, was associated with lower SUDEP risk. The logic is simple: a witness can turn a person on their side, clear the airway, speak to them, and call for help. Sharing a room with a partner, sibling, or roommate is the low-tech version of this.

Seizure-detection devices come next. Wearable sensors that pick up the rhythmic movement of a tonic-clonic seizure and alert a caregiver have shown reasonable accuracy for convulsive seizures in studies, and some have regulatory clearance. Two caveats matter. First, they detect the seizure; a person still has to respond, so a device without a responder is an alarm nobody hears. Second, no device has yet been proven in a trial to prevent SUDEP. Detection is the plausible step toward prevention, not the same thing.

Anti-suffocation pillows, which are designed to allow airflow if a face is pressed into them, have no strong evidence either way. They are unlikely to cause harm and are a reasonable choice for someone who has convulsive seizures in bed.

The pattern here is worth stating plainly: measures that bring a person to the bedside have evidence; measures that replace a person do not yet.

Can people with epilepsy live alone?

Many do, safely and happily, and the answer should never be a blanket no. Independence is a legitimate health outcome in its own right, and the question deserves a risk-based answer rather than a protective reflex.

For someone whose seizures are controlled, or whose seizures are non-convulsive and predictable, living alone carries little additional risk compared with anyone else. The NHS encourages people with epilepsy to lead full, independent lives and to focus safety planning on specific hazards rather than on restriction.

The calculation changes for a person having frequent tonic-clonic seizures, particularly at night, because unwitnessed nocturnal convulsive seizures are the core SUDEP scenario. That does not automatically mean living with family. It might mean a roommate who knows what to do, a seizure-detection wearable connected to someone nearby, a daily check-in call, or a landlord and neighbor who are aware. It might mean a temporary arrangement while medicine is being adjusted, revisited once seizures settle.

Some practical habits reduce risk in any living arrangement: showering rather than bathing, keeping bathroom doors unlocked, using cooking appliances that switch off automatically, and telling at least one nearby person about your epilepsy and how to help.

The best conversation about living alone happens with the neurologist, with a seizure diary on the table, rather than around a family dinner where fear tends to win the argument.

When to see a doctor, and when to call emergency services

Epilepsy care runs on two clocks. The slow clock is routine follow-up. The fast clock is the emergency. Knowing which one you are on matters more than almost anything else in this article.

Book an appointment soon, rather than waiting for the next scheduled review, if seizures have become more frequent or more severe, if a seizure type has changed, if you have had a tonic-clonic seizure after a long seizure-free period, if you are missing doses because of side effects or cost, or if you are planning pregnancy or a major change in sleep or work patterns. New or worsening low mood also belongs on this list.

Call emergency services immediately, based on guidance from the Mayo Clinic and NHS, if a seizure lasts longer than five minutes, if breathing or consciousness does not return once the seizure stops, if a second seizure begins before the person has recovered, if the person is injured, in water, pregnant, or has diabetes, or if a seizure occurs in someone who has never had one before. Difficulty breathing after a seizure is a red flag in its own right, not something to sleep off.

Bystanders can help in the meantime by cushioning the head, moving hard objects away, not restraining the person or putting anything in their mouth, and turning them onto their side as soon as the jerking stops. Timing the seizure and staying until the person is fully awake are the two most useful things a witness can do.

Why talking about SUDEP with your care team makes it less frightening, not more

For years many clinicians avoided the word, reasoning that a rare risk did not justify the anxiety. The AAN/AES guideline moved firmly the other way, recommending that clinicians tell people with epilepsy and their families about SUDEP and their individual level of risk. Surveys behind that shift found what the mother in our opening story already knew: people would rather hear it from a clinician than from a forum at two in the morning.

The conversation does several things at once. It explains why medicine adherence is not a nagging point but a safety measure. It gives a reason to pursue better seizure control instead of tolerating “only” a few convulsive seizures a year. It opens the door to practical questions about sleeping arrangements and devices. And for the many people at low risk, it offers something rarely given in medicine: explicit, evidence-based reassurance.

Bring specifics to that appointment. How many tonic-clonic seizures in the past year, and when. Whether any happened during sleep and whether anyone was present. Whether doses have been missed and why. Whether the current medicine was the first, second, or third tried. Those facts turn a vague worry into a plan.

Epilepsy can be fatal. For most people it will not be, and the difference between those two statements is largely made of decisions that are still available to make.

Frequently asked questions

What is the leading cause of death in epilepsy?

Among people with uncontrolled seizures, SUDEP is the leading cause of death according to the CDC. Across everyone with epilepsy, deaths from the underlying cause of the epilepsy, heart disease, and cancer are more common. Other seizure-related causes include status epilepticus, drowning, falls, and traffic injuries. Depression and suicide are also elevated in epilepsy and are treatable, so mood belongs in the conversation.

Can people with epilepsy live alone?

Many people with epilepsy live alone safely, especially when seizures are controlled or non-convulsive. The decision should be based on individual risk: frequent nighttime tonic-clonic seizures without anyone present are the main SUDEP scenario, so people in that situation may benefit from a roommate, a detection device linked to a responder, or regular check-ins while treatment is optimized. Discuss the arrangement with your neurologist using your seizure diary.

Can epilepsy cause death in sleep?

Yes. Most SUDEP deaths happen during sleep, and the person is often found in bed, frequently face down, after a tonic-clonic seizure. The suspected mechanism is suppressed breathing and heart rhythm after the seizure, made worse by an obstructed airway and the absence of anyone to help. Nighttime seizures that are brief and non-convulsive carry far lower risk than frequent, unsupervised convulsive ones.

Who is at risk of SUDEP?

The strongest risk factor is frequent generalized tonic-clonic seizures; three or more a year raises risk roughly 15-fold according to the AAN/AES guideline. Not being seizure-free, having seizures at night without supervision, and missing medicine doses also increase risk. Long-standing epilepsy that began in childhood shows a modest association. Age, sex, and ethnicity are weak predictors compared with seizure frequency.

How common is SUDEP?

The CDC estimates about 1 in 1,000 adults with epilepsy die from SUDEP each year, and about 1 in 4,500 children. The AAN/AES guideline found similar rates, around 1.2 per 1,000 adults and 0.22 per 1,000 children annually. These averages hide wide variation: people who are seizure-free approach general population risk, while those with frequent convulsive seizures face substantially higher odds.

What causes SUDEP?

The exact cause is not fully known, but monitored cases show a pattern: a tonic-clonic seizure ends, breathing becomes shallow or stops, heart rhythm slows or becomes erratic, and brain activity is deeply suppressed. Failure of automatic breathing drive appears central, possibly worsened by lying face down. Heart rhythm problems and shared genetic factors with certain cardiac conditions may contribute in some cases.

Can SUDEP be prevented?

Risk can be lowered, although no measure eliminates it. The best-supported steps are reducing tonic-clonic seizures through consistent, well-managed treatment, avoiding missed doses, and having someone present or checking at night. Seizure-detection devices can alert a responder but have not been proven to prevent SUDEP in trials. If two medicines have failed, asking about specialist evaluation for other treatments is also a risk-lowering step.

How long does a seizure have to last to be dangerous?

Most tonic-clonic seizures end within one to three minutes. Guidance from the Mayo Clinic and NHS treats a convulsive seizure lasting more than five minutes as an emergency, because the chance it will stop on its own falls and the risk of injury to the brain and body rises. Also call emergency services if a second seizure begins before recovery or breathing does not return to normal.

Do seizure alarms and anti-suffocation pillows work?

Wearable seizure-detection devices can accurately detect convulsive seizures and alert a caregiver, but they only help if someone can respond, and none has been shown in a trial to prevent SUDEP. Anti-suffocation pillows have little evidence either way; they are unlikely to cause harm and may be reasonable for people with convulsive seizures in bed. Human presence has the strongest evidence of any nighttime measure.

Should I be worried about SUDEP if my seizures are controlled?

For people who have been seizure-free, especially free of tonic-clonic seizures, SUDEP risk falls toward that of the general population. The AAN/AES guideline identifies seizure freedom as protective. The most useful action is staying on treatment as prescribed and reporting any breakthrough seizure promptly. Ask your clinician to explain your individual risk; for many people, that conversation is genuinely reassuring.

References

• Sudden Unexpected Death in Epilepsy (SUDEP): Centers for Disease Control and Prevention
• Practice guideline summary: Sudden unexpected death in epilepsy incidence rates and risk factors (AAN/AES): NIH PubMed Central
• Epilepsy: World Health Organization fact sheet
• Living with epilepsy: NHS

This article is for general information only and is not a substitute for professional medical advice. Please consult a qualified doctor about your individual situation.