Can a Blood Test Detect Alzheimer’s? What the New Tests Actually Tell Us
The idea of identifying Alzheimer’s disease through a blood draw once sounded distant. It is now part of clinical practice, but the headline needs an important second sentence: a blood test does not, by itself, tell the whole story of a person’s memory.
In May 2025, the U.S. Food and Drug Administration cleared the first blood test intended to aid in diagnosing Alzheimer’s disease. The Lumipulse G pTau217/β-amyloid 1-42 Plasma Ratio measures two blood markers whose relationship helps predict whether amyloid plaques are present in the brain. The intended patients are adults 55 or older who have signs and symptoms of cognitive decline, evaluated in a specialized care setting. In August 2026, the FDA also cleared Roche’s Elecsys Phospho-Tau (217P) Plasma test, another blood-based aid to assessing amyloid pathology.
Why does this matter? Until recently, confirming Alzheimer’s-related brain changes often meant an amyloid PET scan or a cerebrospinal fluid test obtained through lumbar puncture. Those methods remain important, but they can be expensive, harder to access, or more burdensome. A well-validated blood test may help a clinician decide what further evaluation is needed and may allow some patients to avoid an unnecessary scan.
The first test’s FDA review compared results from 499 cognitively impaired adults with amyloid PET or cerebrospinal fluid findings. Among those with a positive blood result, 91.7% had evidence of amyloid pathology on the comparison test; among those with a negative result, 97.3% had a negative comparison result. These figures describe that study population, not a guarantee for any individual. Some results were indeterminate, and false positives and false negatives remain possible. In 2026, an FDA recall record also documented a correction involving specified lots of the Lumipulse test because of potentially inaccurate classifications. Laboratories and clinicians must follow applicable notices and interpret any result in context.
Amyloid pathology is one piece of a clinical puzzle. Memory changes can have other causes, and the presence of amyloid does not automatically explain every symptom. A clinician still needs a history, examination, assessment of daily function, medication review, and other tests when appropriate. The FDA explicitly says the Lumipulse test is not for general screening or stand-alone diagnosis.
That distinction matters for community outreach. A brief cognitive screen at a Neuro-Reach workshop can start a useful conversation and encourage follow-up. It cannot determine whether someone has Alzheimer’s disease, and a volunteer should never present it as such. Better tools in the clinic are exciting precisely because they can make the next step after a concern more informed,
not because they eliminate the need to listen to the person and family.
Comments