Epilepsy and Dementia: Studying the Hidden Role of Seizures

The brain is coursing with electricity. When that electricity misfires, that abnormal electrical activity is best known for triggering epileptic seizures. Now, scientists are uncovering signs that this same disruptive activity could also be playing a role in dementia.

Neurologist and neuroscientist Dr. Dona Kim Murphey shares how seizures and dementia may be more closely connected than many realize. Murphey is the chief science officer at Synthesys Brain Health, a neurodiagnostics technology company, and the founder of PrognosUs, a startup focused on the early detection and management of neurological conditions.

In this conversation with Being Patient founder Deborah Kan, Murphey explains the bidirectional relationship between epilepsy and dementia. She highlights why seizures — especially subtle, focal ones that can go unnoticed — may accelerate cognitive decline in Alzheimer’s and related dementias. She also discusses the importance of recognizing these symptoms early, educating families and caregivers, and considering EEG testing as part of the diagnostic toolkit.

Being Patient: Are seizures a symptom of dementia? Or do seizures cause dementia? What do we know exactly?

Dr. Dona Murphey: We think that seizures, or epilepsy — a disorder of seizures — can actually herald, complicate, or be driven by what is happening in dementia. Specifically, with respect to Alzheimer’s dementia, that’s where we have the greatest evidence for this. Late-onset epilepsy, which is a seizure disorder that develops as we age, rises in parallel with increased incidence of Alzheimer’s disease.

We see this bidirectional relationship between epilepsy and dementia such that when you have epilepsy, diagnosed when you were a child, in middle age, or later in life but without any cognitive symptoms at that time, that actually puts you at about two times the risk for dementia. The reciprocal is also true: if you have dementia, as that dementia progresses, you also have about double the risk of having epilepsy concomitantly.

“We see this bidirectional relationship between epilepsy and dementia such that when you have epilepsy… that actually puts you at about two times the risk for dementia. The reciprocal is also true: if you have dementia… you also have about double the risk of having epilepsy concomitantly.”

Being Patient: If you have a seizure, are you defined as an epileptic? What actually defines epilepsy?

Murphey: We typically think of it as an unprovoked seizure, but it depends on some circumstances. In a person who presents with a clinical history consistent with a potential seizure, we’re going to do diagnostic workups that often include imaging of the brain, like an MRI, and an EEG, which measures the brain’s electrical activity.

If the person has some underlying structural abnormality in the brain and they have their first seizure consistent with where that abnormality is, we may call that epilepsy even without a second event.

If somebody comes in with a history of something like alcohol withdrawal or head trauma, which can provoke a seizure, we will not call that epilepsy. If there’s no abnormality on brain imaging, we don’t call that epilepsy necessarily.

Being Patient: Do seizures caused by dementia or Alzheimer’s look different than they do otherwise?

Murphey: The scientific literature suggests that about 30 to 80 percent of the seizures in people with dementia are of the type where awareness is impaired. They are typically focal, coming from one area of the brain, and present as you describe: the person is in the middle of a task and suddenly stops. They may have a blank stare, be briefly unresponsive for about one to two minutes, and may have stereotyped behaviors like lip smacking, chewing, or picking with their fingers. You might not notice this, especially as the dementia advances, because of their cognitive changes. It might not be very distinct from how they are ordinarily.

It’s something to look out for and to mention to the doctor. Most people diagnosing dementia are not having these conversations, which makes it more challenging because it accelerates the cognitive decline that’s happening at the same time.

Being Patient: Which part of the brain are you talking about, and how does it impact that part?

Murphey: It’s the middle and underneath part of the brain called the temporal lobe. That gets preferentially involved in the kinds of seizures associated with dementia. We think this is related to protein deposition in dementia. Those proteins are deposited in these areas of the brain — amyloid and tau.

In animal models, this causes hyperexcitability in the hippocampus, the memory-forming part of the brain in that mesial temporal lobe, or inner underbelly of the brain. That hyperexcitability prevents normal memory consolidation in sleep, and then it goes on to affect potentially other parts of the brain.

When you have a focal seizure, if your brain takes care of it, it doesn’t go anywhere from there. But if it doesn’t, it can spread from that area to involve other areas and potentially the entire brain, which is the most dangerous circumstance because that can compromise your breathing and your heart rate.

“Most people diagnosing dementia are not having these conversations, which makes it more challenging because it accelerates the cognitive decline that’s happening at the same time.”

Being Patient: What can you see in an EEG? Can it determine if someone’s at risk, or if they’ve had seizures before?

Murphey: It can be a bit of both. If somebody comes in with a clinical history suspicious for a seizure, especially this kind of seizure — they blank out for a minute — with an existing diagnosis of dementia, you want to have a high level of suspicion and should order an EEG at that point. Even if you had a short study that didn’t capture sleep, it’s very possible that you didn’t capture an abnormality. If you capture sleep and sleep-wake transitions, you can be more confident. The longer the study, the more likely you are to capture an abnormality.

EEGs will capture epileptiform activity, and sometimes that’s not correlated with seizures. A lot of studies are ordered for memory loss or cognitive decline — about half of ours are. That does not mean the person has dementia. It may mean they have just epilepsy. Epilepsy alone, especially if uncontrolled, can cause cognitive decline. Looking for abnormal discharges is useful, because those do correlate with cognitive decline, and seizures do as well.

Being Patient: How do people prepare for this? If you have a loved one with dementia, what do you look out for? Most people aren’t participating in sleep studies, especially if they have a diagnosis. So what do you do?

Murphey: We have to educate all family members. At the beginning of the disease, people who have the disease may be able to recognize that they’re “losing time.” Patients need to be educated, and families, if they observe this behavior, need to have a high level of suspicion.

If you see that kind of behavior, you don’t want them to be engaged in activities where if they lost consciousness or had altered consciousness for a short period of time, they’d be in danger or they’d put other people in danger.

People have set their houses on fire because they’re in the middle of cooking and they have one of these episodes. You can drown if you are swimming, or if you’re even bathing and you’re not supervised. Operating heavy machinery, driving — these things can all be very dangerous.

Being Patient: How much do we know about seizures causing dementia, and how much higher is the risk?

Murphey: This is two different things. One is the seizures themselves, which are a more prolonged, at least 10 seconds of evolving electrically abnormal activity in the brain. Seizures themselves can be clinically apparent, or might not be. They might be totally not apparent to the person themselves or to the family members around them, or this thing they call subclinical abnormal electrical activity, or epileptiform activity.

In either case, we’re seeing a correlation between those things and cognitive decline. It’s really important that people are aware that either of those things can increase the risk. It is about two to three times the risk when you have those kinds of abnormalities. Specifically, that was for late-onset epilepsy, but you do have increased risk with this subclinical epileptiform activity and with seizures in general.

Being Patient: Do we know the relationship between plaque in the brain and risk of seizures in dementia?

Murphey: There is animal and clinical evidence showing subclinical epileptiform activity years before dementia develops, which is part of the evidence that suggests that maybe it plays a role in causing dementia. You also see increased epilepsy with dementia progression.

It’s important for families and patients to be educated about the increased likelihood they have of this other disorder — that, whether it’s causing dementia, it’s making the dementia worse. It’s worth considering treating that condition, because it is treatable.

Unfortunately, not with the same kinds of outcomes as for epilepsy in general. The epilepsy associated with dementia has about a 25 percent responsiveness to medications, as opposed to epilepsy without dementia that you might have had from early life that has about a 65 percent response to medications. The outcomes are not as great, but a 25 percent response is still worthwhile if that means you’re going to slow down the progression of the disease.

Being Patient: What anti-seizure medications are best for epilepsy in dementia?

Murphey: The one that appears best is levetiracetam, or Keppra. Unfortunately, when I talk about the balance you have to strike between side effects and the desired effects of the medication, it’s that levetiracetam is well known to produce behavioral problems, disturbances, and sometimes that’s already an issue with our loved ones with dementia. Outbursts, anger, agitation. We have to be careful with that.

Being Patient: I want to be clear on what to tell our neurologist. Many won’t run EEGs. What do we do?

Murphey: A lot of what we do with our caregivers in terms of supporting them is equipping them with the language to be able to advocate best for their loved ones or with their loved ones. I think it’s important for caregivers or patients early on in the disease to know how to speak the doctor’s language. To say, my loved one has these events that last one to two minutes, and they blank out. To describe it in the way that I’ve described — that’s how we characterize them as epileptologists or neurologists.

Being Patient: Tell me about where the research is heading. What more do we need to know in order to really get better treatment and better recognition about epilepsy and dementia?

Murphey: As far as the research, what is really exciting is some of the EEG studies that we’re doing are quantitatively analyzing the activity in the brain of people who have dementia. What we’re measuring in the brain is electrical activity. We look at that in terms of the waves of the brain. Those waves have a certain frequency, and they have composite frequencies — different frequencies contained within your brain while you’re sleeping, while you’re awake, while you’re engaged or not in activities. Those things, when we measure them, are distinct distributions between those frequencies in people who have dementia, Alzheimer’s dementia, and people who don’t.

Now, there is no difference between people who have dementia and this epileptiform activity versus people who have dementia and who don’t have the epileptiform activity. Those background frequencies are indistinguishable between those two populations. But between people who have dementia and people who don’t, they’re distinguishable.

Being Patient: Could that be a way to recognize risk presymptomatically?

Murphey: Potentially. We’re thinking about blood biomarkers or other kinds of minimally invasive or noninvasive testing. Absolutely, people are working on that.

Being Patient: For the ApoE4 population, would it be worth monitoring EEG to understand risk?

Murphey: Yeah, absolutely. I think the drug companies who are developing drugs for dementia, it would really behoove them to look at incorporating EEG early on, to see if they can stratify patients, if they can characterize them in a way that allows them then to say, how does my drug work on these subpopulations of people that have these kinds of EEG abnormalities or not. Whether they are subclinical epileptiform activity or seizures, or there are these differences in the quantitative analyses of the frequencies in those brainwave studies. That would be amazing to do. People are working on it, but it hasn’t been adopted broadly. There is more work to be done.

The other exciting horizon, as we think beyond amyloid pathology, is about abnormal inflammation, neuroinflammation, and abnormal electrical activity in the brain. I think we’re going to have a much richer idea of the various things that are happening and the various points of intervention.