Can I test my risk of Alzheimer's with a blood test (p-tau217)?

In people with memory complaints, the p-tau217 blood test produces impressive numbers. General practitioners correctly diagnose Alzheimer's in roughly 61% of cases, and memory clinic specialists reach 73%. With the p-tau217 test, accuracy rose in both groups to 88-92%. That is a significant jump, even for specialists.

The test measures a specific protein fragment in your blood that is linked to the protein accumulations in the brain that are characteristic of Alzheimer's. How accurate? In studies that compared the blood test with a brain scan or a lumbar puncture, the test scored an accuracy of 0.963 to 0.974 on a scale of 0 to 1. That is comparable to what can be measured via a lumbar puncture, a test that is more painful and more expensive.

Yet an important caveat applies. The 2025 Alzheimer's Association guideline states that a blood test may only replace a cerebrospinal fluid test or PET scan if it achieves at least 90% sensitivity and specificity. Many commercially available tests do not yet reach that threshold. And even a test that does reach it can never replace a full clinical evaluation by a physician.

For people without complaints, or as a self-test via the internet, this research is not intended. The studies were conducted in people who presented with cognitive complaints, not in healthy individuals who want to test preventively. There is also a practical distinction: p-tau217 performs better than other comparable markers in distinguishing Alzheimer's from other brain diseases such as frontotemporal dementia. That distinction matters, because a mix-up in diagnoses has real consequences for treatment.

In short: the blood test has a solid scientific foundation as a tool within specialised memory care, but which specific test your doctor uses, and whether it meets the quality threshold, is best discussed with your neurologist or memory clinic.