Skip to navigationSkip to content
A row of vials of blood that will be analyzed in a lab setting.
Reuters/Michaela Rehle
The future of dementia diagnosis—but not for a few years at least.
VITAL FLUIDS

There are a lot of promising blood tests for Alzheimer’s—here’s how to keep track

By Katherine Ellen Foley

The race for the blood test to detect early Alzheimer’s disease is on.

Keep in mind, “race” is a bit of a misnomer—unless we’re thinking of the longest ultramarathon ever. There won’t be a clear winner this year, or even in the next few years. This is in part because of the nature of the disease itself: Alzheimer’s takes decades to develop before a person starts showing symptoms like confusion and memory loss. Finding a blood test that can diagnose the condition early will take as long to discover as the disease itself takes to manifest—if not longer.

That said, there’s been a lot of news around the idea of a blood test for Alzheimer’s. Most recently, today (Aug. 1) researchers from Washington University at St. Louis published evidence for a blood plasma test that seems to predict amyloid in the brain with over 90% accuracy (when accounting for participants’ genetics—more on that later).

Earlier this year, a separate team from WUSTL published work (paywall) backing up a blood test for another type of brain damage. And in 2018, a research team based at the National Center for Geriatrics and Gerontology in Japan published results (paywall) of a blood test that could tell the difference between healthy individuals, those with mild cognitive impairment (often the first sign of dementia), and Alzheimer’s disease.

Don’t be fooled by headlines: None of these blood tests are ready for clinical use yet, nor will they be for at least a few years. But if you’re following this space, here’s how to keep track of all the developments across the globe, and why they’re important.

Why do we need a blood test for Alzheimer’s? Don’t we already have ways of diagnosing the disease?

Technically, yes—but the timing is off.

Alois Alzheimer first discovered his namesake dementia by conducting an autopsy of a woman who died of the disease in the early 1900s. Her brain was shriveled and filled with clumps of proteins—which scientists later identified as amyloid beta and tau.

For much of the 20th century, Alzheimer’s could only be diagnosed by looking at the brain post-mortem—which wasn’t particularly useful for the patient in question. It wasn’t until the early 2000s that researchers finally perfected tools like PET scans, which can identify amyloid beta in the brain, and lumbar punctures to identify amyloid and tau in the fluid that surrounds the brain.

There are a couple of issues with these tests. From a practical standpoint, they’re expensive (PET scans can be thousands of dollars) and invasive (spinal taps involve a large needle put into the spine—although they should be painless).

Because of these constraints, doctors don’t perform these procedures unless they’re trying to confirm a diagnosis of Alzheimer’s disease (or rule out another form of dementia). At that point, it may be too late to provide treatment: The amyloid beta and tau have often been in the brain for up to 20 years, building slowly over time. Part of the reason drugs for Alzheimer’s disease have failed again and again is because they’ve been administered too late.

A blood test could skirt these issues. First, it’d be easier to administer, and could theoretically be a part of routine blood work. It could also detect the disease earlier—meaning that patients would actually have time to enroll in clinical trials or, when drugs are finally available, start taking them sooner to actually slow down or halt the disease.

Why don’t we have a blood test already? 

It’s been a challenge for scientists to figure out how proteins in the blood may or may not reflect changes in the brain. In the paper that came out today, researchers looked at a ratio of two versions of amyloid, called amyloid-42 and amyloid-40, in the blood of 158 participants. Years of research have shown that generally, the lower the ratio of these two proteins, the more likely a person is to have amyloid in their brains. But finding the precise threshold that indicates amyloid’s presence in the brain has been difficult, especially because the variations in blood amyloid are so small.

While this research was ongoing, other scientific groups have been exploring using alternative biomarkers. The group at Japan’s National Center for Geriatrics and Gerontology, for example, has been looking at ratios of still other amyloid proteins in the blood as an indication of amyloid in the brain.

Amyloid isn’t the only possible signal of Alzheimer’s, either. Earlier this year, a separate group of researchers at WUSTL published work showing that the levels of a protein called neurofilament light chain in the blood could predict brain damage. This means it could be detecting Alzheimer’s—but it could also be detecting Parkinson’s disease, fronto-temporal dementia, or any other kind of dementia. Without knowing which type of dementia a person has, it’s impossible to treat.

So what are the updates that this new paper offers? 

In the paper that came out today, researchers were able to use blood tests to predict someone’s amyloid status. Of the 158 participants in this clinical trial, 148 were cognitively normal, and 100 of those individuals didn’t have any amyloid showing up on PET scans from when their blood was taken. Blood tests accurately reflected their PET results 88% of the time.

When researchers took into account and whether or not they had APOE-4, the test was even more accurate. Because people with this mutation on the APOE gene are already at a higher risk of developing Alzheimer’s, lower levels of amyloid-42/amyloid-40 in the blood indicate an even higher risk of developing amyloid in the brain. The team found that they could predict someone’s current chances of having amyloid beta in the brain even better, with 94% accuracy.

When will I be able to take one of these tests in my doctor’s office? 

Unfortunately, not for a while.

The work published today was a followup to a study published in 2017 that used a similar method, but was only conducted on 41 participants.

In order to develop a test that definitively diagnoses Alzheimer’s, researchers will need to include even more participants—in the thousands—to ensure its validity. They’ll also have to keep one of these studies going long enough to see if some participants actually go on to develop the disease—which will be expensive.

Fortunately, many groups have financial backing to bring these tests to market. Akinori Nakamura, a neuroscientist with Japan’s National Center for Geriatrics and Gerontology, is working with the public Japan-based company Shimadzu Corp to bring this test to market, per the Associated Press. And the most recent research out of WUSTL has a home, too: Randall Bateman, the neuroscientist who led the work, has patented and licensed the test to C2N Diagnostics, a US-based company he co-founded.