Quick answer: Functional seizures are real and involuntary. They are not epilepsy, and they are not performed on purpose. Wales measured them in national health records. The United States has no equivalent national count. US evidence comes from epilepsy clinics, one Ohio county study, and single health-system records. After a US diagnosis, care is usually a neurologist, sometimes an epilepsy monitoring unit, then a search for therapy that insurance may not cover. Call 911 for a first seizure, an injury, trouble breathing, or an episode that is not the person’s usual pattern. A familiar episode can stay with a written plan.
Why these numbers are not American numbers
Swapping the word Wales for the name of an American state would make the study false. The researchers used Welsh National Health Service records, which cover most people who live in Wales and use one health system. The United States does not have that kind of record. Care is split across private insurance, Medicare, Medicaid, veterans’ care, and people with no steady coverage. There is no national register of functional seizures, and no federal census of the condition.
You may still hear the older name psychogenic nonepileptic seizures, or PNES. The American Academy of Neurology guideline uses functional seizures. Some clinicians say dissociative seizures. Those names refer to the same events. A normal EEG during the event does not mean the person chose it.
What the Welsh study actually found
This is United Kingdom research. FND Connect’s plain-English report of the Welsh work is a UK report. The counts in it are not US rates.
Researchers searched anonymized Welsh health records from 2003 through 2023 and identified about 2,800 people whose records indicated functional seizures. For 2023 they estimated about 90 recorded cases for every 100,000 people living in Wales. When they checked the method against people already known to have the diagnosis, it found only about half of them, because the diagnosis is often missing, vague, or still written down as epilepsy. The 90-per-100,000 figure is a recorded minimum for Wales, not a complete count, and not a count of Americans.
About 30% of the people identified in Wales also had epilepsy. That is why a single emergency instruction is not enough for everyone in that study. The plan has to say what each kind of episode usually looks like.
Hospital use often falls after an epilepsy diagnosis. In the Welsh records, emergency visits and admissions often stayed high after a functional-seizure diagnosis. The report reads that as missing follow-up, not as people choosing the emergency department. For 2023, estimated National Health Service hospital costs were about 3,000 British pounds per person with functional seizures, compared with about 700 British pounds for a similar person without seizures. Those are pounds spent by the Welsh health service. They are not US hospital charges, and they should not be converted into dollars and treated as an American price.
The same Welsh analysis found about twice the risk of death during the study period, compared with similar people who did not have seizures. That is a group finding over many years. It is not a prediction for any one person, in Wales or in the United States. The report does not give a US death rate, and this page will not invent one.
Open the UK Wales report (PDF)
What the United States has instead of a national count
There is no Wales-style national count. A 2015 review from the Jefferson Comprehensive Epilepsy Center in Philadelphia (Asadi-Pooya and colleagues, Epilepsy & Behavior) said the epidemiology is scarce, the studies do not use one diagnostic standard, and most reports come from specialty centers. Those center reports cannot be applied to the whole country.
That review repeated an incidence range of 1.4 to 4.9 new cases per 100,000 people per year. The range mixes Iceland, Scotland, and one US county. It is not a United States incidence. This page does not use it as one.
The US study inside that range was done in Hamilton County, Ohio. Szaflarski and colleagues (Neurology, 2000) looked back at 1995 through 1998 and found a mean incidence of 3.03 per 100,000 per year. People had to be referred to a specialist center and diagnosed with video-EEG. The investigators wrote that this likely missed people diagnosed without that test, or diagnosed somewhere else. It is one county, from the late 1990s. It is not a national rate, and it is not a current rate.
The figure people often quote as “2 to 33 per 100,000” is not a head count either. Benbadis and Hauser (Seizure, 2000) calculated it. They stacked assumptions: how common epilepsy is, what share is hard to treat, what share of those patients reach an epilepsy center, and what share of center patients have functional seizures. Change any assumption and the result moves. It was a reasonable warning that the condition is not rare. It was not a census.
A later calculation (Asadi-Pooya, International Journal of Neuroscience, 2023) offered a US prevalence of 108.5 per 100,000 for 2019, with a confidence interval from 39.2 to 177.8, and an incidence of 3.1 per 100,000 per year (interval 1.1 to 5.1). The interval on prevalence is wider than the entire Welsh recorded rate. The author built it from earlier studies, not from a national file of Americans. Some clinical summaries now repeat 108.5 as if the country had been counted. This page does not use 108.5, or 3.1, as the US rate.
The largest US clinical file in open view is not a state count either. At Vanderbilt University Medical Center, Goleva and colleagues (JAMA Network Open, 2020) used billing codes and chart language to flag 3,341 adults among 2,346,808 adult records, which is 0.14% of that hospital system’s file. A review of 50 flagged charts found a documented, clinically established, or probable diagnosis in 35. The other charts were mostly “possible,” and one did not support the label. One academic medical center in Nashville, searched by an algorithm, is not Tennessee and not the United States. A July 2026 correction changed association statistics in that paper. Those odds ratios are not repeated here, because this page has not rechecked the corrected figures. The paper’s own abstract also attaches a “74% women” figure to the full 3,341. In the results, that percentage belongs to a smaller medical-home subgroup. It is not published here.
Specialty clinics do see this often, which is a different statement from “this percent of Americans have it.” The 2015 Philadelphia review reported that about 5% to 10% of outpatients in epilepsy clinics, and about 20% to 40% of inpatients on epilepsy monitoring units, are found to have PNES. Those are shares of people already sent to seizure centers. They are not the share of the US population. This page does not turn those clinic shares into a national total. The people in those clinics were already referred for seizures.
An estimate, not a census
FND Connect has not counted Americans with functional seizures. Nobody has. The figures below are arithmetic on other people’s studies, applied to the Census Bureau’s resident population on July 1, 2025: 341,784,857. They are functional seizures only. They are not a count of everyone with FND. Functional weakness, tremor, walking problems, and sensory symptoms are not in the Ohio study, and this page will not invent a national total for all of FND from a seizure study.
New diagnoses in a year, from the Ohio study. Hamilton County recorded a mean of 3.03 new video-EEG diagnoses per 100,000 people per year in 1995–1998. If that same rate applied to the whole country at the 2025 population, the sum is about 10,400 new diagnoses a year (3.03 × 341,784,857 ÷ 100,000). That is an illustration. The Ohio investigators wrote that they missed people who were diagnosed without video-EEG, or diagnosed somewhere else. The country is also larger and older than that county was in the late 1990s. Treat about 10,000 a year as a floor for new specialist diagnoses, not as the number of Americans who had a first functional seizure last year, and not as the number living with the condition.
People living with functional seizures, from a published model. The 2023 calculation put US prevalence at 108.5 per 100,000, with an interval from 39.2 to 177.8. That interval is the author’s uncertainty, not a count of states. Applied to the same 2025 population, the midpoint is about 371,000 people and the interval runs from about 134,000 to about 608,000. We did not rebuild the model. We multiplied the published rates by a newer population. A range that wide is a warning. It is not a head count, and it should not be quoted as “371,000 Americans have functional seizures” without the interval and the method.
Clinic percentages are left out of both sums on purpose. Five to ten percent of an epilepsy clinic is not five to ten percent of the United States.
US papers also do not agree on one national share of people who have both functional seizures and epilepsy. Clinic series are cited anywhere from about 10% to about 22%. That is not the Welsh 30%, and neither set of figures is a count of everyone in the United States. What matters for the person in front of you is whether they have both, not which paper’s percentage you remember.
What US care usually looks like after the diagnosis
There is no single American pathway. A typical route starts with a primary care doctor, an emergency department, or a neurologist. The diagnosis itself is clinical. The December 2025 American Academy of Neurology guideline, the first AAN guideline on a functional neurological disorder, says clinicians should take a detailed history, talk to witnesses, look at smartphone video, and consider video-EEG when a typical episode can be captured. Video-EEG on an epilepsy monitoring unit is the most reliable test when it is available. Not every hospital has a unit, and not every plan approves the admission.
The same guideline says to look for co-occurring epilepsy and for psychiatric conditions, and to explain the diagnosis in a way that is clear and respectful. Functional seizures are not a finding of “nothing wrong.” Antiseizure medicines, including benzodiazepine rescue medicines, should not be prescribed for functional seizures unless the person also has epilepsy or another reason for that drug. The National Institute of Neurological Disorders and Stroke says the same about antiseizure medicines and functional seizures.
The treatment the guideline supports is psychological care aimed at the seizures, not another epilepsy drug. Functional-seizure-specific cognitive behavioral therapy is the approach with the most trials. The guideline also names neurobehavioral therapy, paradoxical therapy, and motivational interviewing combined with psychotherapy. These are specific treatments. They are not a suggestion to “relax,” and they are not a statement that the seizures are voluntary. When walking, speech, or daily function is also affected, rehabilitation can sit alongside that therapy. Whether a plan pays for any of it is a separate question.
The guideline’s lead author, Benjamin Tolchin of Yale, has said that most people with functional seizures still do not receive targeted, evidence-based treatment. The practical reasons are ordinary American health-care problems, not a finding that treatment does not exist. Mental health coverage is often a separate benefit, with prior authorization, visit caps, and a directory of therapists who have never treated functional seizures. A neurologist can make the diagnosis and have no in-network clinician to refer to. Medicaid, Medicare, and private plans do not follow one rulebook, and the rule changes with the state and the contract. A denial from one plan is not a medical conclusion. Ask what was denied, whether the plan has an appeal, and whether neurology, therapy, and rehabilitation were billed as separate requests.
Programs this site has actually checked are listed on Finding FND care. That list is not a complete map of the United States.
911, or staying with a written plan
A written plan is for episodes the person already knows. It should say what a usual episode looks like, how long it tends to last, what helps, and what not to do. People nearby protect the head, move hard objects, loosen tight clothing at the neck if that is easy, time the episode, and stay. They do not restrain the person and they do not put anything in the mouth. They do not give epilepsy rescue medicine unless the plan, written with a clinician, says this person also has epilepsy and this particular episode needs that medicine.
Call 911 for a first seizure, an injury, trouble breathing, or an episode that is not the person’s usual pattern. “Usual” has to be defined in the plan. A longer spell, a new movement, a fall with a head strike, or not waking the way they usually do is a reason to call, even if functional seizures are already diagnosed. People who have both epilepsy and functional seizures need the plan to say which features belong to which kind of episode. If you are not sure, call 911.
Do not drive while episodes still take awareness, vision, or control of movement, and ask the clinician whether your state motor vehicle agency has to be told. Rules differ by state. Driving with FND explains how to ask. A mental-health crisis is a separate call: 988, by phone or text. Use 911 when the body emergency above is the problem.
What US readers can still take from the Welsh study
Leave the Welsh rates in Wales. The lessons that travel are about care, not about how many Americans have the condition.
- Functional seizures are real, involuntary, and often under-recorded. A missing code is not a mild symptom.
- A diagnosis without an explanation and a next step leaves people in emergency care. That pattern showed up in a whole country’s records. US clinics describe the same gap, without a national number attached.
- Some people have epilepsy as well. The emergency plan has to allow for both. Do not copy the Welsh 30% onto an American clinic.
- A written plan is part of treatment. It tells friends, schools, and emergency crews when to stay and when to call.
- A higher death rate in a research group is a reason to take follow-up seriously. It is not a forecast for one person.
The proposed Welsh pathway, with a named navigator and specialist teams across the country, is a plan for Wales. It is not a description of what an American insurance card provides.
This is general information for people in the United States. It is not a diagnosis, a coverage decision, or personal medical advice. If symptoms are new, sudden, severe, or unlike the usual pattern, call 911.
Sources
- FND Connect, Functional seizures in Wales: plain-English report of the Welsh records research (PDF, July 2026). UK report. Source of the Welsh counts, the coding gap, the hospital-cost comparison in pounds, and the group mortality finding.
- Health and Care Research Wales, Functional/Dissociative Seizures in Wales. The funded Welsh project (lead Dr William Pickrell). The project page describes the aim. It does not itself publish the 2,800 or the 90-per-100,000 figures.
- Asadi-Pooya AA et al. Epidemiology of psychogenic nonepileptic seizures. Epilepsy & Behavior. 2015;46:60-65. Jefferson Comprehensive Epilepsy Center, Philadelphia. Scarcity of population data; 5% to 10% of epilepsy-clinic outpatients and 20% to 40% of epilepsy-monitoring-unit inpatients in specialty series.
- Szaflarski JP, Ficker DM, Cahill WT, Privitera MD. Four-year incidence of psychogenic nonepileptic seizures in adults in Hamilton County, OH. Neurology. 2000;55(10):1561-1563. Mean incidence 3.03 per 100,000 per year, 1995–1998, one county, video-EEG at a referral center.
- Benbadis SR, Hauser WA. An estimate of the prevalence of psychogenic non-epileptic seizures. Seizure. 2000;9:280-281. The 2 to 33 per 100,000 range is a calculation from assumptions, not a head count.
- Asadi-Pooya AA. Incidence and prevalence of psychogenic nonepileptic seizures (functional seizures): a systematic review and an analytical study. International Journal of Neuroscience. 2023;133(6). Source of the modeled US prevalence, 108.5 per 100,000 (interval 39.2 to 177.8). Scaled here to the 2025 Census population. Not a national census.
- U.S. Census Bureau, QuickFacts, United States. Resident population estimate for July 1, 2025 (Vintage 2025): 341,784,857. Used only as the multiplier.
- Goleva SB et al. Epidemiology of functional seizures among adults treated at a university hospital. JAMA Network Open. 2020;3(12):e2027920. Vanderbilt records only. Association statistics were corrected in 2026 and are not repeated here.
- Tolchin B et al. Management of functional seizures practice guideline executive summary. Neurology. 2026;106:e214466. Published online December 10, 2025. Endorsed by the American Epilepsy Society and the Functional Neurological Disorder Society. AAN guideline page. AAN news release, including the statement that most people do not receive targeted treatment.
- National Institute of Neurological Disorders and Stroke. Functional neurologic disorder. Antiseizure medicines do not treat functional seizures.