A $7 Seizure Drug Might Prevent Alzheimer's. But You'd Need to Start Taking It 20 Years Early.

A decades-old seizure medication costing as little as $7 a month can stop Alzheimer's toxic brain plaques from forming. Northwestern University scientists published the finding in Science Translational Medicine on February 11, 2026. The drug is levetiracetam, and it's been sitting on pharmacy shelves since 1999.

Here's the catch: it only works if you take it long before Alzheimer's starts destroying your brain. Possibly 20 years before.

The $384 Billion Problem Nobody's Solving at the Root

Alzheimer's costs the United States $384 billion a year in health and long-term care. The newest treatments — lecanemab at $26,500 per year and donanemab at $32,000 per year, plus thousands more for brain scans — clear existing amyloid plaques after they've already formed.

That's like mopping the floor while the tap's still running.

Jeffrey Savas, a neurologist at Northwestern's Feinberg School of Medicine, and his team asked a different question: what if you could stop the tap?

How a Seizure Drug Rewires the Alzheimer's Pathway

The team studied genetically engineered mice, lab-grown human neurons, and brain tissue from deceased people with Down syndrome. They found that amyloid-beta 42 — the most toxic protein fragment in Alzheimer's — builds up inside synaptic vesicles. These are the tiny packets neurons use to send signals to each other.

Levetiracetam binds to a protein called SV2A that controls how neurons recycle these vesicles. By slowing that recycling slightly, the drug keeps another protein — amyloid precursor protein, or APP — on the cell's surface longer.

Why does that matter? When APP gets pulled inside the cell through certain pathways, it's more likely to get chopped into amyloid-beta 42. Keep APP on the surface, and the toxic fragment never gets made.

"In our 30s, 40s and 50s, our brains are generally able to steer proteins away from harmful pathways," Savas said. "As we age, that protective ability gradually weakens."

The distinction matters. This isn't age being a disease. It's age weakening a defense that most brains maintain naturally for decades. In Alzheimer's brains, too many neurons lose that defense at once.

The Down Syndrome Connection

More than 95% of people with Down syndrome develop Alzheimer's by around age 40. They carry an extra copy of the gene involved in amyloid production on their triplicated chromosome.

The Northwestern team studied brain tissue from people with Down syndrome who died in their 20s and 30s — before Alzheimer's symptoms appeared but after the earliest molecular changes had begun. These brains showed the same buildup of presynaptic proteins the team had found in their mouse models.

That's what researchers call the "paradoxical stage" of Alzheimer's: before synapses are lost and dementia sets in, presynaptic proteins accumulate. If you could intervene during that window, you might prevent the cascade entirely.

"Conceivably, if you started giving these patients levetiracetam in their teenage years, it could actually have a preventative therapeutic benefit," Savas said.

Real Patients, Real Data, Cautious Results

Because levetiracetam's already widely prescribed for seizures, the team didn't need to wait for a clinical trial to get human data. They pulled records from the National Alzheimer's Coordinating Center and compared Alzheimer's patients who'd taken levetiracetam with those on other anti-seizure drugs or nothing.

The levetiracetam group showed a longer time between cognitive decline diagnosis and death.

Savas is careful about overselling it. "Although the magnitude of change was small — on the scale of a few years — this analysis supports the positive effect of levetiracetam to slow the progression of Alzheimer's pathology," he said.

A few years doesn't sound like much. But for a disease where current treatments cost $26,500-$32,000 annually and slow decline by roughly 27%, a $7 generic extending the timeline at all is worth paying attention to.

The 20-Year Problem

Here's where prevention gets complicated. Levetiracetam won't help anyone who already has dementia. By that point, too much brain tissue is gone.

Savas estimates high-risk individuals would need to start taking it up to 20 years before current FDA-approved blood tests could detect even mildly elevated amyloid levels. That's an almost impossible clinical ask. You'd be prescribing a daily medication to healthy people in their 30s or 40s based on genetic risk alone.

That's why the team plans to start with populations where the risk is certain: people with genetic forms of Alzheimer's and individuals with Down syndrome. If it works there, the conversation about broader prevention opens up.

What This Actually Changes

The drug itself isn't perfect. Savas admits levetiracetam breaks down quickly in the body. His team is already developing a longer-lasting version that better targets the specific mechanism they discovered.

But the bigger story isn't one drug. It's the biology.

Every major Alzheimer's treatment on the market attacks the problem after plaques have formed. Lecanemab and donanemab are antibodies that clear existing amyloid. They're expensive. They require IV infusions and regular brain scans. And they slow decline — they don't stop it.

What Savas's team found is where the plaques come from and how to prevent them from being made. That's a different approach entirely. Even if levetiracetam itself isn't the final answer, the mechanism it revealed — APP trafficking through synaptic vesicle recycling — is a new drug target that nobody was looking at before.

"Our new results uncovered new biology while also opening doors for new drug targets," Savas said.

The Alzheimer's field has spent decades and billions chasing plaques after they form. The answer might involve stopping them from forming at all. And the first clue came from a generic drug that costs less than a streaming subscription.