Most of us know someone who worries about their memory as the years add up.
A simple blood draw now promises an early answer in just a few hours. The trouble is that the answer is not always the right one.
Researchers at University Medicine Halle in Germany report a two-step method that may cut the false positives shadowing the newest Alzheimer's blood tests.
Their findings were published in October 2026 in Nature Communications.
Also read: These common vegetables contain surprisingly high amounts of potassium
The test they are trying to improve measures a protein fragment called p-tau217. In Alzheimer's disease, the brain builds up abnormal tau and amyloid, and tiny amounts leak into the blood.
For a few years now, p-tau217 has looked like one of the best signals that this process is underway.
Why p-tau217 can light up for other reasons
The problem is that p-tau217 is not born only in the brain. Tau also sits in peripheral nerves and in muscle, in a longer form the researchers call "big tau."
When muscle or nerve tissue is damaged, that peripheral tau can leak into the blood and nudge the p-tau217 reading up, even in people who do not have Alzheimer's.
Also read: Taking blood pressure pills? This common mistake could cost you the benefits
The team tested 315 people from neurological centers in Halle, Mannheim and Milan. The group included 101 patients with Alzheimer's disease, 139 with amyotrophic lateral sclerosis (ALS, a motor neuron disease that destroys muscle), and 75 controls.
A conventional blood test flagged high p-tau217 in both the Alzheimer's and the ALS groups. Reading those results in a clinic, a doctor could easily send an ALS patient down an Alzheimer's workup by mistake.
"Conventional blood tests cannot reliably distinguish between these different forms of tau," Professor Markus Otto, who leads neurology at University Medicine Halle, said in a statement according to Medical Xpress.
"Consequently, elevated p-tau217 levels may incorrectly indicate Alzheimer's disease when they are actually associated with a muscle or nerve disorder."
Also read: Your tap water may contain twice as many forever chemicals as reported
How the extra antibody step changes the answer
The fix is to add one step before the measurement. The researchers use a specific antibody to pull out only the brain-derived tau from the blood sample. The muscle and nerve forms are left behind. Then the p-tau217 level is measured on what remains.
With that filtering step in place, high p-tau217 showed up only in the Alzheimer's patients. The ALS patients and the controls looked normal. In other words, the signal that was leading tests astray was removed before the test was run.
The authors are careful not to oversell the result. It is a single study, and they call for standardization and validation in larger, independent groups before any clinic switches over.
What this could mean for patients in the US
Blood-based Alzheimer's testing is moving fast in US clinics. In May 2025, the FDA cleared the first blood test to help identify amyloid pathology in symptomatic adults aged 50 and older, Fujirebio's Lumipulse p-tau217 and beta-amyloid plasma ratio test.
Also read: What the color of your stool can reveal about your liver and gut health
More patients are being screened each year, and false positives can trigger weeks of anxiety and expensive follow-up imaging.
A method that filters out peripheral tau would not replace the current workflow. It would be an added quality check for patients whose results sit in the borderline range, or who also have a muscle or nerve condition that might muddy the picture.
If you have had or expect to have a p-tau217 blood test, nothing changes today. The antibody step is still a research tool.
Anyone with questions about a borderline or surprising result should bring it back to the ordering clinician, who can weigh it against memory testing and imaging before any diagnosis is settled.
Also read: How eating spinach every day may affect your blood pressure and muscle strength
This article is made and published by Ida-Marie Palm Varbæk, who may have used AI in the preparation.