The Alzheimer's Blood Test, Explained for Families

A blood draw can now help answer a question that used to require a PET scan or a spinal tap. Here is what these tests can tell you — and the four words that matter most: not a screening test.

What these tests measure

They detect proteins in the blood that reflect the amyloid and tau changes happening in the brain.

For most of the history of Alzheimer's diagnosis, confirming the disease meant a specialized PET scan costing thousands of dollars or a lumbar puncture — a needle into the spinal canal. Neither is something a primary care doctor can arrange easily, and both are why so many families never got a definitive answer at all.

Blood biomarker tests changed that. The most useful marker is a form of tau protein called p-tau217, which tracks closely with the amyloid plaque burden in the brain. Some tests combine it with the ratio of two amyloid fragments, amyloid beta 42 and 40.

What the test tells you is whether the biological hallmarks of Alzheimer's disease are likely present. It does not tell you how impaired someone is, what stage they are in, or how fast they will decline.

Which tests are actually FDA-cleared

Two — and the distinction between a cleared test and a lab-developed one matters more than the marketing suggests.

• Fujirebio Lumipulse G pTau 217/β-Amyloid 1-42 Plasma Ratio — cleared in May 2025, the first blood test the FDA cleared for Alzheimer's. Its intended use is adults aged 50 and older who are already showing signs and symptoms of cognitive decline and are being evaluated in a specialized care setting. The label states plainly that it is not a screening or stand-alone diagnostic test. Results come back positive, negative or indeterminate.
• Roche Elecsys pTau181 — cleared in October 2025, and the first blood test cleared for use in primary care. It is designed to help rule Alzheimer's-related amyloid pathology out in patients 55 and older who have symptoms, with a negative predictive value of about 98%. In plain terms: it is very good at telling a family doctor that Alzheimer's is unlikely, which spares a great many people a specialist referral and a long wait.

Everything else you may have seen advertised — including Quest's AD-Detect panels and C2N's PrecivityAD2 — are laboratory-developed tests. They may be excellent, and PrecivityAD2 in particular performs well in studies, but they have not been through FDA clearance and are not held to the same published standard.

How well do they work?

Well enough to substantially change what a family doctor can do — and the newest evidence is striking.

At the Alzheimer's Association International Conference in London in July 2026, researchers led by Dr Sebastian Palmqvist of Lund University reported on more than 1,300 patients and 165 physicians in Sweden. Using the PrecivityAD2 blood test, primary care doctors diagnosed Alzheimer's correctly 93% of the time, up from 65% without the result — close to the roughly 94% achieved by specialists. The result changed the diagnosis or the care plan for more than half of the patients.

Two honest caveats. These findings were presented at a conference and have not yet been peer-reviewed. And the study used one specific test in one health system; results do not automatically transfer to every product on the market.

The guideline that tells you which tests to trust

In 2025 the Alzheimer's Association set accuracy thresholds — and warned that many tests being sold do not meet them.

The Association's first clinical practice guideline on blood-based biomarkers, published in July 2025, is the most useful thing a family can know about in this area.

• A test used for triage — to help rule Alzheimer's out — should have at least 90% sensitivity and at least 75% specificity, with positive results confirmed by PET or spinal fluid testing.
• A test used to replace PET or spinal fluid testing should reach at least 90% on both measures.
• The guideline warns explicitly that many commercially available blood tests do not meet these thresholds and have not been validated to the same standard or across diverse populations.
• It applies only to people with objective cognitive impairment being seen in specialized care, and it says a blood test should not be ordered before a comprehensive clinical evaluation.

That gives you a question worth asking out loud: which specific test are you using, and does it meet the confirmatory threshold or the triage one?

Why this is not a screening test

A positive result in someone without symptoms does not mean they have Alzheimer's, and no treatment follows from it.

This is the part that gets lost. Every cleared test is intended for people who already have cognitive symptoms. None is approved for screening healthy people.

The reason is mathematical. When a condition is uncommon in the group being tested, even a good test produces a large share of false positives. Critics of early direct-to-consumer testing pointed out that among cognitively healthy people with a family history, only a minority of positive results would turn out to be true. A frightening result in a well person can mean months of distress, and it can distract from the far more common and often treatable causes of memory trouble — depression, thyroid disease, sleep apnoea, B12 deficiency, medication side effects.

Amyloid in the brain also does not equal dementia. Some people carry substantial amyloid and never develop symptoms in their lifetime.

What it costs and whether Medicare pays

Medicare pays once symptoms are present. It does not pay to screen someone who is well.

Coverage today is limited to people already showing symptoms of cognitive decline. Screening asymptomatic people is not covered, and the bipartisan ASAP Act introduced in March 2026 would create a Medicare pathway for it — which itself confirms that the pathway does not currently exist.

Out-of-pocket prices vary widely by test and by region. Reported figures range from roughly $50 to $100 at some health systems up to $500 to $1,000 for the FDA-cleared Lumipulse test. Ask what the specific test will cost before it is drawn, and ask whether financial assistance is available — many health systems have it and do not advertise it.

How to use this at your next appointment

The test is a lever for getting a stalled evaluation moving, not a substitute for one.

• If you have been told to wait and see, ask directly: is a blood biomarker test appropriate here, and if not, why not?
• Ask which test the practice uses and whether it is FDA-cleared or laboratory-developed.
• Bring a dated record of specific changes in everyday function. These tests are for people with symptoms, so documenting the symptoms is what makes the test available to you.
• Understand what each result opens. A negative result on a rule-out test is genuinely reassuring and redirects the search toward other causes. A positive result usually needs confirmation and, importantly, it is what makes someone eligible to be considered for the disease-modifying drugs.
• Do not order a test for yourself online in place of seeing a doctor. Even the best test needs a clinician to interpret it against everything else going on.