Adults with hard to treat epilepsy who used cannabis almost every day had more than twice as many convulsive seizures a year as those who did not use it, and they reached the point of needing specialist inpatient monitoring thirteen years earlier.
The finding comes from a study published in Epilepsia on 28 September 2026, led by researchers at Washington University in St Louis using patient records from the University of Wisconsin School of Medicine and Public Health. It arrives while cannabis is still widely sold on the promise that it helps with sleep, anxiety and, in some marketing, seizures themselves.
What the researchers looked at
The team reviewed the records of 64 adults with drug resistant focal epilepsy who were admitted to an epilepsy monitoring unit between 2019 and 2024. Drug resistant means seizures that two or more antiseizure medicines have failed to control. In the unit, doctors watch seizures as they happen so they can adjust medication or assess whether surgery is an option.
Twenty two of the 64 had documented cannabis use on 20 or more days a month. Their average age was 31 and roughly a quarter were women. The 42 with no documented use averaged 40 years old and just over half were women.
The gap between the groups was wide. Convulsive seizures, the kind that spread across both sides of the brain, affected 91 per cent of the frequent users against 57 per cent of the nonusers. Frequent users had more than twice as many of those seizures a year. They were also on more antiseizure medicines and were more likely to have temporal lobe epilepsy.
Five years against eighteen
The clearest number in the study is how long it took each group to end up in specialist care. Frequent cannabis users were admitted to the monitoring unit an average of five years after their first seizure. For nonusers the average was 18.1 years. The difference was statistically significant.
That gap matters because admission to a monitoring unit usually marks the point at which standard treatment has stopped working. Reaching it in five years rather than eighteen suggests the condition hardened against medication far faster in the group using cannabis most days.
“We did not anticipate how striking the results would be,” senior author Aaron Struck told Healio.
Which came first
The obvious objection is that people with worse epilepsy might turn to cannabis for relief, which would explain the association without cannabis doing any harm. The records point the other way. Among the frequent users whose timelines were documented, cannabis use began before the seizures did.
The profile of that group was also unusual. They were mostly men in their twenties with none of the risk factors that normally sit behind severe epilepsy, such as the structural brain damage that typically shows up in patients whose seizures have resisted treatment for years.
Struck has said he started asking the question after seeing patients with severe seizures and no apparent reason for them.
Cannabis is not cannabidiol
Much of the public confusion here comes from treating one plant as one medicine. A purified cannabidiol extract has been approved for use alongside other medicines in a small number of rare epilepsy syndromes, including Dravet and Lennox-Gastaut, after randomised trials in children and adolescents. That approval covers a regulated, measured, single compound.
The cannabis people buy is a different product. Its main psychoactive ingredient is tetrahydrocannabinol, which can alter how readily neurons fire and may make seizures more likely rather than less. The same research group has published two case reports of men with focal epilepsy whose cannabis use coincided with breakthrough seizures and poor control, including a suggestion that it blunted the effect of implanted neurostimulation.
A regulatory approval for one compound in one group of patients is routinely stretched into a general claim that cannabis treats epilepsy. The new data give patients a reason to doubt that claim.
What the authors will and will not say
The researchers are careful about what 64 records can prove. The study is retrospective, the sample is small, and cannabis use was taken from what patients reported and what clinicians wrote down. An association is not causation, and the team says so plainly. A follow up study is planned to work out how much use matters and whether stopping helps.
First author Santiago Philibert-Rosas said the honest answer to a patient asking whether cannabis is affecting their seizures is that doctors do not yet know.
Struck drew a practical conclusion anyway. If a patient is using cannabis for anxiety or sleep, he said, “I’d prefer to treat those problems with drugs we understand” rather than something largely unregulated whose effect on the brain is unclear.
The question nobody is asking patients
The practical lesson is small and immediate. Cannabis use is rarely covered in a busy neurology appointment, and patients who use it for sleep or anxiety often do not think to mention it. Asking the question costs nothing and may change what a clinician recommends.
The wider lesson is harder. Daily and near daily cannabis use in the United States now outnumbers comparable daily drinking, and the products on sale have been normalised faster than anyone has measured what they do. This study looked at one condition, in one monitoring unit, over five years, and found a group of young men whose epilepsy ran out of treatment options in a third of the usual time.
That is the pattern worth watching. Where cannabis has been sold as low risk and quietly medicinal, the research keeps arriving afterwards, and it keeps pointing in the same direction.
(Source: WRD News)