An Alzheimer’s blood test can now spot amyloid in the brain from a single draw. Four of them are FDA cleared in the United States. All four are cleared only for adults who already have memory or thinking problems (FDA). A positive result means amyloid is probably there. It does not mean you have dementia, because plenty of clear-thinking older people carry amyloid too (Meta-analysis).
What the Alzheimer’s Blood Test Measures
These tests read two proteins floating in your blood. The first is a tagged form of a brain protein called tau, and labs measure the version known as p-tau217. The second is amyloid beta, the protein that clumps into plaques between brain cells.
In Alzheimer’s disease, p-tau217 in the blood climbs as those plaques build up. A high reading tracks what a brain scan would show (FDA).
Fujirebio’s Lumipulse was cleared first, on 16 May 2025 (FDA). It reports a ratio between p-tau217 and amyloid beta, and the FDA says the two halves are not meant to be read on their own.
Roche’s Elecsys p-tau181 test followed in October 2025 (FDA). It is built for family doctors, and it works by ruling Alzheimer’s out. A negative result lines up well with a clean brain scan. A positive one often does not, so the patient goes on for more testing.
C2N’s PrecivityAD2 was cleared in August 2026 (FDA). It weighs the protein fragments on a mass spectrometer and returns a score from 0 to 100.
Roche’s Elecsys p-tau217 test was cleared in the same month (FDA). It uses two cutoffs, one to rule amyloid out and one to rule it in. The clearance summary reports how it did in family practice and in specialist clinics.
Clearance is a lower bar than it sounds. The FDA read the data, agreed the test matches older methods, and let the maker sell it. That is permission to market. How useful the test turns out to be for you is a separate question.
The advance is still real. Until now, confirming amyloid meant a PET brain scan or a lumbar puncture, the spinal tap that draws fluid from the lower back. A blood draw can do that first look instead.
Cleared for Symptoms, Not Screening
Read the fine print on any of the four and the same line shows up. The test is for adults who already have memory or thinking symptoms and are being evaluated for them.
The FDA’s wording for the first test is blunt. The test “is not intended as a screening or stand-alone diagnostic test”. Results “must be interpreted in conjunction with other patient clinical information”. A positive result “does not establish a diagnosis of Alzheimer’s disease” (FDA).
The age floors differ a little. Both Roche tests start at 55 (FDA), and PrecivityAD2 starts at 40 (FDA). For the Fujirebio test, the FDA’s press release says 55 while its own technical review says 50 (FDA).
The symptom line exists because that is where the numbers came from. In the first clearance study, about half of the 499 samples came from people who did have amyloid (FDA). That mix is normal in a memory clinic and unusual in the wider public. When a condition is uncommon in the group you test, a larger share of the positive results are wrong.
Researchers ran a p-tau217 test in 2,916 adults with normal thinking (Study). About 1 in 3 of them had amyloid, which is high for people with no symptoms. Even so, roughly 21 in every 100 positive results were false alarms. In the same paper’s recruitment model, finding 100 people with amyloid meant blood-testing 677.
The Alzheimer’s Association reached the same conclusion in 2025, in its first guideline on these blood tests. It covers people with measurable problems in thinking. It says plainly that it does not cover screening, non-specialist settings, or people with no concerns about their thinking (Guideline). It also says the blood test should follow a full clinical evaluation, never lead it.
Amyloid Is Not a Diagnosis
A positive result tells you one thing: amyloid is probably there. It does not tell you that you have dementia, or that amyloid is why you lost the name of a film. It says nothing about when symptoms might arrive, or whether they will arrive at all.
Plenty of people carry amyloid and think perfectly well. One meta-analysis pooled the raw records of 2,914 people with normal thinking (Meta-analysis). About 10 in every 100 were amyloid positive at age 50. At 70 it was about 23 in 100, and at 90 about 44 in 100. A later analysis of more than 19,000 people put the figure near 24 in every 100 (Pooled analysis). When the cutoffs came from the data instead, the share ran about 10 in every 100 higher.
Those figures come mostly from research volunteers, so a study that sampled a whole county is a useful check (Study). It found the same climb. Fewer than 3 in every 100 people in their fifties were amyloid positive while thinking clearly. By their eighties it was about 41 in 100.
None of that makes a positive result harmless. Researchers followed people with normal thinking for about three and a half years (Study). Among those with amyloid but no tau spread, roughly 9 in every 100 developed mild memory problems or dementia. Among people with no amyloid it was about 3 in 100. Where tau had also spread through the brain, about half declined.
Over a lifetime the gap widens. One long study in Minnesota followed people who carry the APOE e4 gene variant. Among those with a positive scan at 65, estimated remaining lifetime dementia risk ran from roughly 44 to 74 in every 100 (Study). Where a person sat in that range depended on sex and on how much amyloid there was. The other end of the range matters just as much, because a sizeable share never developed dementia at all. In the same population, age drove the ten-year risk while the amount of amyloid drove the lifetime risk (Study).
So a test result gives you a probability about the years ahead. A diagnosis is a judgement about how you are now. It still comes from a clinician who has looked at your thinking, your history and anything else that could explain the symptoms.
The Gray Zone in the Middle
The first cleared test was measured against 499 samples (FDA). About 9 in every 10 positive results were confirmed by a brain scan or spinal fluid, and about 97 in every 100 negative results were confirmed too. Those numbers come from people who already had symptoms.
Fewer than 20 in every 100 got no answer at all, and that group rarely makes the headlines. Their result came back between the two cutoffs. In the clearance study, 98 people landed there, and 49 of them had amyloid while 49 did not (FDA). Those are the odds of a coin toss.
The band is wide on every test. Roche’s p-tau217 test puts about 20 in every 100 results there, in family practice and specialist care alike (FDA). The uncertain band on PrecivityAD2 covers about 17 in 100 (FDA). An Italian memory clinic saw about 23 in every 100 patients land in the gray zone (Study). More than half of those patients did turn out to have Alzheimer’s biology, so a middling result is a reason to keep looking.
In practice it means a referral: a PET scan, a lumbar puncture, or the same blood test again months later.
Two ordinary things also move the number. Weak kidneys push p-tau217 up. One study looked only at people with no amyloid (Study). About 1 in 14 with healthy kidneys still tested positive. Among those with reduced kidney function it was about 3 in 10. About half of the small group with the poorest function tested positive.
Body weight may push the number the other way. Heavier people had lower readings on all four tests in one study that measured p-tau217 (Study). That was a single snapshot. The same people showed a faster rise over time, so a low reading still leaves the question open.
Each cutoff also belongs to the test that set it, so a number from one lab does not transfer to another.
Why the Result Matters Now
Two drugs now strip amyloid out of the brain: lecanemab and donanemab. Both are antibodies given by drip, and an amyloid answer is what decides whether they are on the table at all.
In donanemab’s main trial, 1,736 people were split by coin flip into drug or dummy infusions for 18 months (Trial). Among those with less tau on their scans, the treated group declined about a third less on the trial’s rating scales. Across the whole trial the gap was smaller. The treated group still declined, only more slowly.
A 2026 Cochrane review pooled the trials of this drug class in people with mild memory problems or mild dementia. At 18 months it found little to no difference in how severe the dementia was, and it rated that evidence low quality (Cochrane review).
Brain swelling and small brain bleeds happen often enough that patients need repeat MRI scans during treatment. On lecanemab, about 13 in every 100 had brain swelling, against about 2 in 100 on dummy infusions (Safety analysis). Small bleeds or iron deposits showed up in about 17 in 100, against about 9 in 100. On donanemab, about 24 in every 100 had brain swelling (Trial). Three treated participants died of causes the investigators linked to the drug, as did one person on dummy infusions.
The APOE e4 gene raises that risk sharply. A later analysis of the donanemab trials counted how many copies each person carried (Analysis). Brain swelling hit about 15 in every 100 with no copy, 24 in 100 with one copy, and 42 in 100 with two.
Few patients qualify in the first place. In one Swedish memory clinic, 26 of 410 patients met the criteria for lecanemab (Study). That is about 1 in 16 of the clinic, and only about 38 in every 100 of the amyloid-positive patients. Under European rules, which exclude people carrying two copies of APOE e4, it fell to 14. A US population study found 19 people eligible out of 237 with amyloid-positive memory problems (Study).
Cost narrows the door further. Lecanemab launched at a list price near $26,500 a year, and researchers estimated scans, infusions and monitoring at roughly $7,330 more per patient (Analysis). They put the yearly bill for a patient without supplemental cover near $6,600.
A positive blood test is most useful when the person can actually get the treatment it points to.
Sold as a Longevity Panel
You can already buy this test yourself. A large US lab launched a direct-to-consumer amyloid blood test in July 2023, with no doctor’s order needed (Viewpoint). Others now bundle p-tau217 into wellness and longevity panels.
That is the one use no cleared test supports, and three things tend to go wrong.
Start with the arithmetic. In a low-risk group, a bigger share of the positives are wrong (Study). Then there is the middle band. Roughly 1 in 5 results come back there, with no clinician attached to say what happens next.
The aftermath is the part where the evidence is kinder than you might expect. Researchers told 1,167 older adults with no thinking problems that their amyloid scan was positive (Study). Over the short term, depression, anxiety and suicidal thoughts did not rise. Worry about Alzheimer’s did. But all of that ran inside a study, with teaching beforehand and trained counselling after. The authors said the result may not carry over when the news is delivered any other way. A result that appears in an app arrives without any of that support.
The Alzheimer’s Association says much the same. It cautions that consumer tests “can lead to confusion, misinterpretation, or unwarranted anxiety among patients” (Guideline).
One more cost rarely gets priced in. Federal rules stop health insurers from using your genes or your history against you. They do not reach long-term-care, life or disability policies (Viewpoint). If an application asks and you hide a result you already know about, that can count as fraud and void the policy.
None of this is a judgement on anyone who bought one. The same pattern shows up across the panel market, in biological age tests and in organ age scores. The problem is testing with no clinical question behind it.
Frequently Asked Questions
Can I Get an Alzheimer’s Blood Test With No Symptoms?
You can buy one, but no cleared test is meant for that. All four FDA-cleared tests are for adults already being evaluated for memory or thinking problems, and the wording rules out screening (FDA). The Alzheimer’s Association guideline also leaves out people who have no concerns about their thinking (Guideline).
Does a Positive Result Mean I Will Get Dementia?
No. It means amyloid is probably present, which raises your risk without settling anything. Take people with normal thinking, a positive amyloid scan and no spreading tau. About 9 in every 100 developed mild memory problems or dementia over roughly three and a half years (Study). Lifetime risk is higher, and it may shift with your age, sex, genes and amyloid level (Study).
Is the Blood Test as Accurate as a PET Scan?
It comes close in the right patients, but it is an aid rather than a replacement. In the first cleared test’s study, about 9 in 10 positives and about 97 in 100 negatives matched a scan or spinal fluid (FDA). The Alzheimer’s Association sets a higher accuracy bar for replacing a scan than for sorting who needs one (Guideline).
What Does a Borderline Result Mean?
It means the test did not answer the question. In the first clearance study, the middle band split almost exactly in half (FDA). Your clinician will usually order a PET scan or a lumbar puncture, or repeat the blood test later (Study).
Can a Test Result Affect My Insurance?
It may affect some policies. Federal protections cover health insurance, not long-term-care, life or disability cover (Viewpoint). Doctors writing in JAMA Neurology warn that hiding a known result on an application could be treated as fraud.
Key Takeaways
- Cleared for symptoms only. All four tests are for adults already being evaluated for memory and thinking decline, and none is cleared for screening (FDA).
- Amyloid is not dementia. About 10 in every 100 clear-thinking 50-year-olds carry amyloid, rising to about 44 in 100 at age 90 (Meta-analysis).
- Roughly 1 in 5 results settle nothing. Fewer than 20 in every 100 samples in the first clearance study came back in the uncertain middle (FDA).
- Few people qualify for treatment. In one memory clinic, about 1 in 16 patients met the criteria for lecanemab (Study).
- The drugs carry real risk. Brain swelling affected about 13 in every 100 on lecanemab and about 24 in 100 on donanemab (Safety analysis).
- Skip the self-ordered version. Consumer tests can cause confusion and unwarranted anxiety, the Alzheimer’s Association cautions (Guideline).
What to Do About Your Memory
If your memory worries you, or worries someone who loves you, the first move is an appointment rather than a purchase.
A proper workup starts with the boring things, and the boring things are treatable. Standard guidance asks for thyroid tests, vitamin B12, a blood count and a metabolic panel in nearly every patient assessed for these symptoms (Guideline). It also names sleep apnea, low mood, heavy drinking and medicines that dull thinking as possible contributors. Checking for sleep apnea and heavy drinking is called out directly.
Those checks turn up less than you might hope. A 2003 review of 39 studies found a potentially reversible cause in about 9 in every 100 dementia cases (Meta-analysis). Well under 1 in 100 actually reversed. More often these problems sit on top of something else and make it worse, and treating them still helps.
A blood test belongs inside that visit, ordered by someone who will read it next to everything else they know about you. Ordered on its own, it is a number with no question behind it.
The test is a real advance, and it answers a narrower question than the headlines suggest. That is worth knowing before you order one.
This article is for educational purposes and is not medical advice. Talk to a qualified clinician before changing your health regimen.
Leave a comment