Epilepsy can return after years, even decades, of being seizure-free, and this happens more often than most people expect. In one cohort study of patients who had been seizure-free for at least two years and then tapered off medication, roughly seven in ten experienced a seizure recurrence within five years of stopping treatment. The risk is not the same for everyone, though, and the factors that influence it range from the type of epilepsy and its underlying cause to what a brain scan or EEG looks like at the time medication is withdrawn.
How Often Seizures Return After Remission
The likelihood of relapse depends heavily on whether you stay on medication or stop it. An American Academy of Neurology practice advisory found that among adults who had been seizure-free for two years, those who tapered off their antiseizure medications had about a 15% recurrence rate over two to five years, compared with roughly 7% among those who stayed on treatment.1PubMed. Antiseizure Medication Withdrawal in Seizure-Free Patients: Practice Advisory Update Summary Those numbers may sound modest, but they reflect a carefully selected population of people whose doctors felt comfortable trying a taper. In less carefully screened groups, the figures climb sharply. A Chinese cohort study tracked 423 patients who withdrew from medication after at least two seizure-free years and found a cumulative recurrence rate of about 46% within the first year and nearly 72% beyond five years, with the majority of relapses clustered in the first twelve months after stopping drugs.2PubMed Central. Relapse After Drug Withdrawal in Patients with Epilepsy After Two Years of Seizure-Free
Even for people who remain on medication, seizure freedom is not always permanent. Some epilepsy syndromes carry a lifelong tendency toward relapse, and new seizures can surface years later due to changes in brain structure, new health conditions, or shifts in medication effectiveness. The point is that remission in epilepsy is real and common, but it does not always mean the condition has been cured.
What Makes a Relapse More Likely
Certain patterns in a person’s epilepsy history strongly predict whether seizures will eventually return. The risk factors most consistently identified across studies include the underlying cause of the seizures, the type of epilepsy syndrome, the number of medications needed before remission was achieved, and what the EEG looks like around the time drugs are withdrawn.
Structural brain abnormalities stand out as one of the strongest predictors. A study examining the causes of epilepsy found that hippocampal sclerosis, malformations of cortical development, and combinations of the two were associated with particularly low rates of long-term seizure freedom. Among patients with hippocampal sclerosis alone, only about 11% remained seizure-free over the long term; with dual pathology, that dropped to 3%.3PubMed. Is the underlying cause of epilepsy a major prognostic factor for recurrence? In other words, the more structurally abnormal the brain tissue driving the seizures, the harder it is for remission to hold.
In children, a study of antiepileptic drug withdrawal found that about 62% of recurrences happened within the first six months of stopping medication. The two factors that independently predicted relapse were an abnormal first EEG and a history of needing more than one drug to achieve seizure control.4PubMed Central. Risk of recurrence after discontinuation of antiepileptic drug therapy in children with epilepsy Needing multiple medications to get seizures under control in the first place is a signal that the underlying epilepsy is more resistant and more likely to reassert itself.
Juvenile myoclonic epilepsy (JME), one of the most common generalized epilepsy syndromes, has its own set of red flags. A large meta-analysis of individual patient data identified nine variables that independently predicted drug resistance and relapse in JME, including psychiatric comorbidities, having all three seizure types (myoclonic jerks, absence seizures, and generalized tonic-clonic seizures), focal abnormalities on EEG, a history of febrile seizures, and a family history of epilepsy.5PubMed Central. Individualised prediction of drug resistance and seizure recurrence after medication withdrawal in people with juvenile myoclonic epilepsy JME is often described in textbooks as “well-controlled on medication,” but the relapse rates after withdrawal are high, and many neurologists recommend lifelong treatment for this syndrome.
The Timing Problem With Medication Withdrawal
For anyone who has been seizure-free for a few years, the natural question is whether they still need medication. The evidence shows this is a genuinely difficult decision. The AAN practice advisory noted that for children, there was probably no meaningful difference in recurrence risk between tapering after two years versus four years of seizure freedom.1PubMed. Antiseizure Medication Withdrawal in Seizure-Free Patients: Practice Advisory Update Summary This suggests that for pediatric epilepsy, waiting an extra two years on medication may not buy additional protection once you finally stop.
The peak danger zone is the first year after stopping. In the cohort study tracking 423 patients, nearly half of all recurrences happened during that first year.2PubMed Central. Relapse After Drug Withdrawal in Patients with Epilepsy After Two Years of Seizure-Free After about three years off medication, the cumulative rate plateaued, meaning that if you make it through the first couple of years without a seizure, your risk of late relapse drops considerably. This pattern holds in pediatric studies as well, where the bulk of recurrences cluster in the first six months.4PubMed Central. Risk of recurrence after discontinuation of antiepileptic drug therapy in children with epilepsy
This early clustering of relapses has practical implications. It means the decision to taper medication is not a one-way door. If you pass through the high-risk window without incident, the odds increasingly favor staying seizure-free. But it also means the months immediately after withdrawal require extra vigilance, and many neurologists recommend slower tapers, more frequent follow-up, and lifestyle precautions like avoiding sleep deprivation during this period.
Relapse After Epilepsy Surgery
Surgery is often considered when medication fails, and for many people it produces lasting seizure freedom. But even after successful surgery, late relapses do occur. A study of 159 patients who had been seizure-free for at least five years after surgery found an annual relapse rate of about 4% per year between years five and ten. Despite this ongoing trickle of relapses, nearly 90% of the group remained in remission at the end of follow-up.6PubMed. Prognosis after late relapse following epilepsy surgery
The pattern of recurrence after surgery differs from medication withdrawal. A longitudinal study found that about 82% of post-surgical seizure recurrences were “early,” meaning they happened within two years. Among those who relapsed early, the vast majority continued to have seizures despite optimized medical treatment. Late recurrences, happening after two years, accounted for only about 18% of cases, and of those, only about a quarter went on to have ongoing seizures.7PubMed Central. Predictors of Postoperative Seizure Recurrence This is encouraging: a late relapse after surgery is less likely to herald a return to uncontrolled epilepsy than an early one.
A multicenter study looking at relapse after surgical remission found that 55 patients relapsed after achieving a two-year seizure-free period. The relapse rates were comparable between medial temporal resections and neocortical resections, but the single factor that predicted relapse in the medial temporal group was how long it had taken to achieve remission in the first place.8PubMed. Predicting long-term seizure outcome after resective epilepsy surgery A slow road to initial remission signals that the epileptic network may be more resilient and harder to dismantle permanently.
What EEG Can and Cannot Tell You
An EEG that still shows epileptiform discharges (the electrical spikes associated with seizure-prone brain tissue) is one of the most consistent red flags for relapse. The AAN advisory specifically noted that an epileptiform EEG in pediatric patients increases the risk of seizure recurrence when medications are tapered.1PubMed. Antiseizure Medication Withdrawal in Seizure-Free Patients: Practice Advisory Update Summary But the standard in-clinic EEG, which typically records for 20 to 40 minutes, misses a lot.
Longer recordings appear to be more useful. A study of adults with learning disabilities who had been seizure-free for at least three years (with a median of ten years) found that all six patients whose ambulatory EEG showed epileptiform discharges before medication withdrawal went on to relapse. Among the nine without discharges on ambulatory EEG, five still relapsed, bringing the overall relapse rate in this small sample to about 74%.9PubMed Central. Prolonged epileptic discharges predict seizure recurrence in JME The presence of discharges was a near-perfect predictor of relapse, but their absence did not guarantee safety. This distinction matters: a clean EEG is reassuring but not a guarantee, while an abnormal one should raise serious caution.
In JME specifically, researchers found that the maximum length of epileptic discharges on prolonged ambulatory EEG was a strong predictor of recurrence, with a cutoff of about 2.7 seconds discriminating between relapsers and non-relapsers with very high accuracy. The same measurements taken from a standard short recording lacked both sensitivity and specificity.9PubMed Central. Prolonged epileptic discharges predict seizure recurrence in JME This highlights a frustrating gap in routine practice: the standard EEG that most clinics offer may simply not capture enough information to guide this decision well.
When a “Relapse” Is Actually Something Else
Not every episode that looks like a seizure is one. Syncope (fainting), psychogenic non-epileptic seizures (PNES), panic attacks, movement disorders, and certain cardiac arrhythmias can all mimic epileptic seizures convincingly. A review of the differential diagnosis literature found that misdiagnosing non-epileptic events as epilepsy is a well-recognized problem that can delay correct treatment and expose people to unnecessary antiseizure medications with their own side effects.10Epileptic Disorders. How to distinguish seizures from non-epileptic manifestations
This is relevant for anyone in long-term remission who has a suspicious episode. If you have been seizure-free for years and then experience something that feels like a seizure, the appropriate response is investigation, not automatic resumption of medication. A thorough workup that includes video-EEG monitoring can distinguish an epileptic seizure from a mimic. PNES in particular are common in people with a history of epilepsy, and some people have both genuine epileptic seizures and PNES, which complicates management. Jumping to the conclusion that your epilepsy has returned without proper evaluation risks restarting medications you may not need.
Can Treatment Work Again After a Relapse?
This is one of the questions people fear most: if seizures come back, will the medication that worked before still work? The answer is usually yes, but not always. A long-term population-based study of childhood-onset epilepsy followed patients who relapsed after stopping medication. Of those who restarted treatment, about three-quarters eventually regained seizure control, but the timeline could be long. Two patients in the study did not achieve a five-year remission until ten to nineteen years after restarting medication. The remaining patients who restarted drugs never achieved that level of control, and factors like an identifiable structural cause and localization-related epilepsy were associated with failure to regain control.11PubMed. Prognosis of seizure recurrence after stopping antiepileptic drugs in seizure-free patients
For JME, the picture is somewhat more optimistic. Among 116 people who restarted treatment after a recurrence and were followed for at least two years, about 78% regained seizure freedom for at least twelve months.12eClinicalMedicine. Prediction of treatment outcomes in juvenile myoclonic epilepsy This is reassuring, but the roughly one in four who did not regain full control represents a real cost of attempting withdrawal. These numbers are worth weighing carefully before deciding to taper, especially in syndromes known for high relapse rates.
Lifestyle Triggers and the Biology of Vulnerability
Even when someone’s epilepsy has been quiet for years, certain conditions can unmask a lingering vulnerability. Sleep deprivation is one of the most well-documented triggers. Research has consistently shown that losing sleep increases the risk of seizures, to the point that clinicians have historically used sleep deprivation as a diagnostic tool to provoke epileptiform activity on EEG. The mechanism involves a decrease in GABA-mediated inhibition, the same neurotransmitter system that is disrupted in epilepsy itself.13PubMed Central. Sleep deprivation: a risk for epileptic seizures
Alcohol withdrawal, high fever, extreme physical or emotional stress, and certain medications (including some antidepressants and antibiotics that lower the seizure threshold) can also provoke seizures in someone with a latent predisposition. For a person in remission, these triggers may never cause a problem. But for someone whose brain has a lower seizure threshold, especially if they have tapered or stopped their medication, these factors can tip the balance. The underlying concept, supported by laboratory research, is that seizures are not a sudden switch but reflect a gradual loss of the brain’s ability to absorb destabilizing electrical activity. When that resilience is already diminished, a single bad night of sleep or a bout of heavy drinking may be enough to push the network past its threshold.14PubMed Central. Loss of neuronal network resilience precedes seizures and determines the ictogenic nature of interictal synaptic perturbations
The Practical Fallout of a Relapse
Beyond the medical implications, a seizure relapse creates immediate disruptions in daily life. In most jurisdictions, a new seizure triggers a mandatory period during which you cannot legally drive. The duration varies by state and country, ranging from three months to a year or more. A small survey of people affected by this found that about 85% reported their quality of life became worse due to driving restrictions, and about two-thirds said their sense of independence was much worse. Roughly 31% had lost a job because of seizures, though none attributed job loss directly to the inability to drive.15American Epilepsy Society. Impact of Driving Restriction on People with Epilepsy (IDRoPS)
For someone who has been living as though their epilepsy is behind them, possibly for years, these consequences hit harder than they might for someone who has been managing active seizures all along. Losing driving privileges, disclosing a seizure to an employer, recalculating insurance, and restarting daily medication are all concrete disruptions. These practical stakes make the decision about whether to attempt a medication taper more fraught than the recurrence statistics alone would suggest.
New Tools for Predicting Who Will Relapse
Researchers are working on better ways to identify which patients can safely stop medication and which are likely to relapse. One approach uses structural brain imaging. A study found that patients who went on to relapse after drug withdrawal had measurable differences in cortical thickness, surface area, and cortical volume in specific brain regions at the time of withdrawal compared with those who stayed seizure-free.16PubMed. Microstructural features of the cerebral cortex: Implications for predicting epilepsy relapse after drug withdrawal These differences were detectable on high-resolution MRI scans that are already used in clinical practice, raising the possibility that imaging could eventually help personalize the withdrawal decision.
Machine learning models are also being tested. One study combined brain imaging features with clinical factors in a support vector machine model and achieved moderate predictive accuracy for seizure recurrence after a first unprovoked seizure, though the performance was not dramatically better than clinical judgment alone.17PubMed Central. Structural brain imaging biomarkers for predicting seizure recurrence after a first unprovoked seizure The most informative imaging features were asymmetries between the two brain hemispheres in gray matter volume and regional curvature patterns. These tools are not ready for routine clinical use, but they represent a move toward making the recurrence question less of a coin flip and more of an informed estimate.
Outside of the clinic, wearable technology is opening up new possibilities for long-term monitoring. Devices that measure movement, heart rate, skin conductance, and muscle electrical activity are being developed to detect seizures in real time and, eventually, to forecast periods of elevated risk. Longer-term implantable EEG systems have also demonstrated the ability to track seizure activity over months, providing an objective record that goes far beyond what a clinic visit can capture.18PubMed Central. Seizure Diaries and Forecasting With Wearables: Epilepsy Monitoring Outside the Clinic For someone considering medication withdrawal, having continuous monitoring during the high-risk taper period could catch the earliest signs of trouble before a full-blown seizure occurs.