An Alzheimer’s Blood Test Held Up Well Against the Spinal-Fluid Standard

CATEGORY: BRAIN & MENTAL HEALTH

Until recently, confirming Alzheimer’s disease meant a spinal tap or a specialized brain scan, and most people being evaluated for memory problems never receive either. A simple blood test correctly identified whether Alzheimer’s was present or absent in about 85% to 91% of 1,767 patients at clinics in Sweden, Spain and Italy.¹ The same group had earlier reported a similar blood test at about 90% accuracy. Doctors making that call the traditional way, with an exam, memory testing and a brain scan, reached 61% in primary care and 73% among dementia specialists.²

A Tau Fragment That Leaks Out of the Brain

What the test measures is p-tau217, a protein fragment in the blood. It is a form of tau, one of the two proteins that build up abnormally in the brains of people with Alzheimer’s. As tau begins tangling inside brain cells, small amounts of the fragment slip into the bloodstream. That is smoke escaping under a door: a signal of what is going on inside, without anyone opening the room. The new study used a fully automated, commercially available version that ordinary hospital laboratories can run.¹

Judged Against Spinal Fluid in 1,767 Patients

Every participant had come to a doctor with memory or thinking complaints. About 1,200 were seen at specialist memory clinics and 548 at ordinary primary care practices. Each gave a blood sample and also had a spinal-fluid test, the established laboratory standard for detecting Alzheimer’s changes, and the blood result was scored against it using cutoff values fixed in advance. Roughly half of them turned out to have Alzheimer’s changes.¹

A Gray Zone Lifted Accuracy to 92% to 94%

In the specialist clinics, blood and spinal fluid agreed 89% to 91% of the time. In primary care it was 85%. Accuracy slipped to 83% among people aged 80 and older. Using two cutoffs instead of one, with a clearly negative zone, a clearly positive zone and a gray zone between them, lifted accuracy to 92% to 94%. The catch is that 12% to 17% of people landed in the gray zone and would need further testing.¹ A mass-spectrometry version of the test matched the automated one in specialist clinics, did better in primary care, and was unaffected by age.¹

What a Positive Result Does Not Settle

The test has been studied only in people who already had symptoms, so nothing here speaks to people without them. A positive result says Alzheimer’s-type changes are likely present. It does not say how far those changes explain a person’s symptoms. Thyroid problems, depression and medication effects can all cloud memory, and some people carry Alzheimer’s changes for years without ever developing dementia. Roughly one result in ten was wrong in one direction or the other, which is why the authors describe the test as a way to streamline diagnosis rather than replace clinical judgment. American regulators cleared the first blood test of this type for clinical use in 2025.

Independent work lines up. A 2025 review pooling 30 studies found blood p-tau217 correctly flagged about 82% of people with Alzheimer’s-type protein changes and correctly cleared about 86% of those without, with accuracy strongest in people who already had memory complaints.³ A Spanish study of 493 people, using the same automated platform, reported 92% and 91% on those two measures against spinal-fluid results.⁴

Key Takeaways

• An automated p-tau217 blood test identified Alzheimer’s changes with 85% to 91% accuracy in 1,767 people with memory complaints, checked against the spinal-fluid standard.
• An earlier study put a similar blood test near 90% accuracy, against 61% for primary care doctors and 73% for specialists relying on exams, memory tests and brain scans.
• Accuracy fell somewhat after age 80, about one result in ten was wrong, and the test has been studied only in people who already had symptoms.

References

1. Palmqvist S, Warmenhoven N, Anastasi F, et al. Plasma phospho-tau217 for Alzheimer’s disease diagnosis in primary and secondary care using a fully automated platform. Nature Medicine. 2025;31(6):2036–2043. link
2. Palmqvist S, Tideman P, Mattsson-Carlgren N, et al. Blood Biomarkers to Detect Alzheimer Disease in Primary Care and Secondary Care. JAMA. 2024;332(15):1245–1257. link
3. Khalafi M, Dartora WJ, McIntire LBJ, et al. Diagnostic accuracy of phosphorylated tau217 in detecting Alzheimer’s disease pathology among cognitively impaired and unimpaired: a systematic review and meta-analysis. Alzheimer’s & Dementia. 2025;21(2):e14458. link.
4. Martínez-Dubarbie F, Guerra-Ruiz A, López-García S, et al. Diagnostic performance of plasma p-tau217 in a memory clinic cohort using the Lumipulse automated platform. Alzheimer’s Research & Therapy. 2025;17(1):68. link.