TL;DR
  • Blood tests measuring p-tau217 can now identify the amyloid pathology behind Alzheimer's disease from a routine blood draw. The FDA cleared the first one in 2025 and cleared Roche's single-biomarker Elecsys pTau217 test on August 24, 2026, bringing the total to four cleared tests.
  • These tests are cleared for adults 55 and older who already have symptoms or complaints of cognitive decline. They are not screening tests, and no blood test diagnoses Alzheimer's on its own.
  • APOE4 is the strongest common genetic risk factor, carried by roughly one in four people. It raises risk and lowers the age of onset, but it is not a diagnosis or a verdict.
  • Combining the two is where the new science is. A September 2026 Lancet Neurology analysis of about 8,500 people found that once p-tau217 is elevated, APOE4 carriers develop symptoms in about 3 to 4 years versus 5 to 6 years for others.
  • Medicare pays for biomarker testing only after symptoms appear. And federal genetic-privacy law (GINA) does not cover long-term care or life insurance, which is why families should settle coverage questions before pursuing predictive genetic testing.

What Alzheimer's Blood Tests Actually Measure

The tests in clinical use do not measure memory, mood, or "how Mom is doing." They measure proteins that leak into the bloodstream when Alzheimer's pathology is present in the brain.

The headline protein is p-tau217, a phosphorylated form of tau that tracks closely with both amyloid plaques and tau tangles, the two hallmarks of Alzheimer's disease. Some tests report p-tau217 on its own; others report it as a ratio against beta-amyloid 1-42. The Alzheimer's Association's 2025 clinical practice guideline reviewed 49 observational studies and 31 blood-based tests measuring p-tau217, p-tau181, p-tau231, and amyloid ratios, and deliberately declined to endorse any brand, calling rankings premature (Alzheimer's Association, AAIC 2025).

What matters to families is the practical shift. The first cleared blood test was reviewed as an alternative to amyloid PET brain scans, which the FDA described as costly, time-consuming, and involving radiation exposure (U.S. Food and Drug Administration). A blood draw at a familiar lab is a very different experience for an 80-year-old than a spinal tap or a trip to a regional imaging center.

Which Tests Are FDA-Cleared, and Who They Are For

Blood-based Alzheimer's tests referenced in this post. Cleared means the FDA reviewed the test for marketing; it does not mean the test is covered by insurance.
TestWhat it measuresStatusIntended population
Lumipulse G pTau217/β-Amyloid 1-42 Plasma Ratio (Fujirebio) Ratio of p-tau217 to beta-amyloid 1-42 First FDA-cleared blood test used in diagnosing Alzheimer's, cleared via 510(k) with Breakthrough Device designation Adults 55 and older with signs and symptoms of cognitive decline, in a specialized care setting
Elecsys pTau217 plasma test (Roche, developed with Eli Lilly) p-tau217 alone, reported as positive, intermediate, or negative FDA-cleared August 24, 2026; described as the first single-biomarker test supporting rule-in and rule-out in both primary and specialty care Individuals 55 and older with signs, symptoms, or complaints of cognitive decline
Elecsys ApoE4 immunoassay (Roche) ApoE4 protein in plasma; reports carrier or non-carrier CE-marked in Europe, not FDA-cleared; showed 100% concordance with APOE4 genotyping in a 607-participant study Adults with signs and symptoms of cognitive impairment

As of the Elecsys clearance, four blood-based biomarker tests for Alzheimer's disease are FDA-cleared (Alzheimer's Association). Two points repeat across every official statement, and they are the ones families should hold onto. First, none of these tests is a stand-alone diagnostic. Second, none is authorized as a screening test for people without symptoms (FDA).

How Accurate Are They?

In the multi-center study supporting the first clearance, 499 plasma samples from cognitively impaired adults were compared against amyloid PET or cerebrospinal fluid results. Of those with a positive blood result, 91.7% had amyloid plaques confirmed; of those with a negative result, 97.3% had a negative PET or CSF result. Fewer than 20% of results were indeterminate (FDA).

The Alzheimer's Association guideline sets the bar a clinician should apply. A test used for triage needs at least 90% sensitivity and 75% specificity, where a negative result rules out Alzheimer's pathology with high probability and a positive result should be confirmed by PET or CSF testing. A test used as a full substitute for PET or CSF needs at least 90% sensitivity and 90% specificity. The panel noted that many commercially available tests do not meet those thresholds (Alzheimer's Association).

The question to ask in the exam room "Is this test being used to rule Alzheimer's out, or to confirm it?" Those are different jobs with different accuracy requirements, and the answer determines whether a positive result should be followed by a PET scan or a spinal tap. The guideline is also explicit that a blood test should not be ordered before a comprehensive clinical evaluation, and that results must always be read in clinical context.

The APOE4 Gene: What It Does and Does Not Mean

Everyone carries two copies of the APOE gene, one from each parent, in some combination of the APOE2, APOE3, and APOE4 variants. APOE3 is the most common. APOE2 is mildly protective. APOE4 is the strongest common genetic risk factor for late-onset Alzheimer's disease, and each additional copy raises risk and lowers the average age at onset.

About 25% of people carry one copy of APOE4, and roughly 2% to 3% carry two (BrightFocus Foundation). Among people with Alzheimer's disease, an estimated 40% to 60% carry the variant (Roche Diagnostics).

Forward-looking research following more than 10,000 people produced the estimates below. They are considerably less alarming than the numbers families often encounter online, which frequently come from backward-looking autopsy studies that overstate risk.

Estimated risk of cognitive impairment by age 85, by number of APOE4 copiesEstimated risk ranges: no APOE4 copies 10 to 15 percent, one copy 20 to 25 percent, two copies 30 to 55 percent.0%15%30%45%60%Estimated risk by age 85No APOE4 copiesabout 75% of people10–15%One APOE4 copyabout 25% of people20–25%Two APOE4 copiesabout 2-3% of people30–55%
Estimated risk of developing mild cognitive impairment or disabling cognitive impairment due to Alzheimer's disease by age 85, based on prospective follow-up of more than 10,000 people. Data: BrightFocus Foundation, Understanding Your APOE Status. Population frequencies from the same source.

Read the top row again, because it is the one that gets skipped: even with no APOE4 copies, estimated risk is 10% to 15%. And most people with one copy never develop Alzheimer's dementia. APOE4 is a risk factor, not a diagnosis and not a fate.

Adult son sitting at a kitchen table with his mother, reviewing a printed medical results page together

Putting Biology and Genetics Together: The Timing Question

The most consequential 2026 finding is not about whether someone will develop Alzheimer's, but when. A pooled analysis of prospective cohorts covering roughly 8,500 participants across diverse racial and ethnic groups, published in The Lancet Neurology on September 9, 2026, found that p-tau217 alone was not robust enough to predict the timing of symptom onset, but that adding APOE status sharpened it considerably. The projections held across backgrounds.

Years from elevated p-tau217 to the start of cognitive symptomsAPOE4 carriers develop symptoms about three to four years after p-tau217 becomes elevated; people with other APOE variants about five to six years.01234567Years after p-tau217 becomes elevatedCarries one or more APOE4 copies3–4 yrsOther APOE variants5–6 yrs
Time from elevated plasma p-tau217 to the onset of cognitive symptoms, by APOE genotype. Data: Xu et al., "Plasma phosphorylated tau 217 concentrations, APOE genotype, and timing of cognitive impairment," The Lancet Neurology (2026), as reported by Medical Xpress.

For clinicians, that gap is a treatment-window question: elevated p-tau217 in an APOE4 carrier suggests roughly three years before symptoms emerge, which researchers describe as a candidate window for preventive therapy. For families, it is a planning question, and the planning window is the same three to six years.

Where Cognitive Markers Fit

Blood biomarkers describe biology. Cognitive markers describe function, and function is what determines whether someone can still manage medications, drive, or live alone. Standard neuropsychological testing remains the backbone, and newer approaches, including speech and language analysis and multimodal digital assessments, are being studied as earlier and less burdensome ways to detect change (SpeechDx / Alzheimer's Drug Discovery Foundation).

The practical framing families can use: biology tells you what is present, cognitive testing tells you what it is costing, and the long-term care question lives almost entirely in the second column. Insurance benefit triggers, home-care decisions, and safety judgments are built on function, not on a lab value.

Should Someone Without Symptoms Get Tested?

The current answer from the researchers doing this work is generally no. Asymptomatic people are not eligible for the approved anti-amyloid antibody drugs, and Columbia's Richard Mayeux, a senior author on the Lancet Neurology analysis, put it plainly: "I don't recommend that people get these tests right now if they're asymptomatic." Trials are underway to test whether those drugs help asymptomatic people with elevated p-tau217, and that evidence may change the answer.

APOE status does carry one concrete treatment implication today. Carriers, especially people with two copies, have a meaningfully higher risk of amyloid-related imaging abnormalities (ARIA), the brain swelling and microbleeds that can occur during treatment with drugs such as Leqembi and Kisunla. ARIA is usually detectable on MRI and often symptomless, but in the original studies roughly 3 in 1,000 people developed a brain hemorrhage resulting in permanent disability or death (BrightFocus Foundation). That is why APOE status increasingly informs whether and how aggressively these therapies are used.

Neurologist speaking with an adult daughter and her mother across a consultation desk, with a cognitive screening worksheet on the table

Cost, Medicare, and the Coverage Gap

The cost story is genuinely better than the PET era. The blood tests can reduce the need for PET scans that run upward of $10,000, and in one Michigan clinic's largely low-income population patients faced roughly $50 to $100 out of pocket for biomarker testing (Association of Health Care Journalists).

The gap is timing. Medicare pays for biomarker blood testing only after a patient is already showing outward symptoms of cognitive decline; used before symptoms appear, the same test is considered screening and is not covered. The Alzheimer's Screening and Prevention Act of 2025 (the ASAP Act, H.R. 6130), introduced November 19, 2025 by Rep. Vern Buchanan and Rep. Paul Tonko, would amend the Social Security Act to create Medicare coverage for FDA-cleared blood-based early-detection tests furnished on or after January 1, 2028 (Congress.gov). It remains a bill, not law, and should be treated that way in any family's planning.

The Planning Trap: Genetic Results and Long-Term Care Insurance

This is the part of the conversation that gets missed, and it is the part most relevant to long-term care planning.

Most people believe the Genetic Information Nondiscrimination Act of 2008 protects them broadly. It does not. GINA prohibits health insurers from using genetic information for eligibility, coverage, underwriting, or premiums, and prohibits employers from requiring or requesting it. But GINA does not apply to life insurance or long-term care coverage (National Human Genome Research Institute). Long-term care, life, and disability insurers sit in that gap and may consider genetic information in underwriting (The Scientist).

The sequencing implication is straightforward. If long-term care insurance (LTC or LTCI) is part of the plan, the underwriting conversation generally belongs before predictive genetic testing, not after. A blood biomarker test ordered as part of a symptomatic medical workup is a different situation, since medical records already reflect the cognitive complaint. Either way, the decision deserves a conversation with an independent advisor or an elder law attorney before results exist, because a result cannot be un-known. For background on how long-term care is actually paid for, our sister resource FundingDependency.com walks through the funding paths in plain language.

This is educational information, not legal, insurance, or medical advice. State laws vary, and some states add protections beyond GINA.

What Families Should Actually Do This Month

  • Start with the evaluation, not the lab slip. If you are seeing changes, ask for a comprehensive cognitive evaluation. Guidelines say the blood test comes after that, not instead of it.
  • Write down what you have observed, with dates. Missed bills, repeated questions, a wrong turn on a familiar route. Function is what clinicians and insurers act on.
  • Ask what the test is being used for and what happens with each result: positive, negative, or indeterminate.
  • Handle long-term care insurance questions first if predictive genetic testing is on the table, given the GINA gap.
  • Get the legal documents in place early. Durable power of attorney, healthcare surrogate, and HIPAA authorizations are easier to complete while capacity is unquestioned.
  • Use the diagnostic window as a planning window. Three to six years is enough time to choose care settings deliberately rather than under crisis pressure.

The real advance here is not certainty. It is lead time. A result that arrives three years before the first missed bill gives a family something that was almost never available before: the chance to make the big decisions calmly, together, while the person they are made for can still help make them.

Frequently Asked Questions

Can a blood test diagnose Alzheimer's disease by itself?

No. The FDA-cleared blood tests are cleared to aid diagnosis, not to replace it. Both the FDA and the Alzheimer's Association state that results must be interpreted alongside a full clinical evaluation, and that no blood test should be ordered before that evaluation happens.

Who are Alzheimer's blood tests currently intended for?

Adults roughly 55 and older who already have signs, symptoms, or complaints of cognitive decline. The cleared tests are not authorized as screening tests for people without symptoms.

Does carrying one copy of APOE4 mean you will get Alzheimer's?

No. About one in four people carry a copy of APOE4, and most never develop Alzheimer's dementia. Estimated risk of cognitive impairment by age 85 runs about 10 to 15 percent with no copies, 20 to 25 percent with one copy, and 30 to 55 percent with two copies.

Does Medicare pay for an Alzheimer's blood test?

Medicare currently pays for biomarker blood testing only after a person is showing outward symptoms of cognitive decline. Testing before symptoms appear is treated as screening and is not covered. The proposed ASAP Act (H.R. 6130) would create a Medicare coverage pathway for FDA-cleared blood-based screening tests furnished on or after January 1, 2028.

Can long-term care insurers use genetic test results?

Federal law is narrower than most families assume. GINA covers health insurance and employment; it does not apply to life insurance or long-term care coverage. That is why many advisors suggest settling long-term care insurance questions before pursuing predictive genetic testing.

What is the difference between a blood biomarker and a cognitive marker?

A blood biomarker such as p-tau217 reflects underlying brain pathology. Cognitive markers, including memory testing, speech and language patterns, and digital assessments, measure function. Clinicians increasingly read them together: biology tells you what is present, function tells you what it is costing.

Topics in This Post

Sources for this post (all accessed September 2026):