What Alzheimer’s Blood Tests Can’t Tell You About Your Dementia Risk

A simple blood test can now hint at whether the proteins characteristic of Alzheimer’s disease are accumulating in the brain, often years before any signs of cognitive decline. But for someone who feels perfectly fine, the result may raise more questions than it answers.

The U.S. Food and Drug Administration has cleared four blood tests to detect signs of changes associated with the disease. Two, Elecsys pTau217 and PrecivityAD2, were cleared in August. While these tests are meant for people who already have memory or thinking problems, clinicians say people without symptoms are increasingly seeking them out, either through their doctors or commercial labs.

“Two years ago we very rarely saw asymptomatic individuals getting tested,” says Suzanne Schindler, a neurologist at Washington University School of Medicine involved in validating C2N’s PrecivityAD2 test. “But that has really changed.” Some people have watched a parent or loved one develop the disease. “This is their biggest fear,” Schindler says.

The trouble is that a positive result can’t reliably tell someone whether—or when—they’ll develop symptoms. And while these tests can detect signs of disease years before memory problems begin, no approved treatments exist for people without symptoms. For someone who feels fine, a positive result can bring anxiety, more tests, or concerns about future eligibility for life or long-term care insurance without necessarily changing their medical care.

It’s why current guidelines advise against testing people without cognitive symptoms.

As these tests become more widely available, the question is no longer just how early Alzheimer’s can be detected, but how much that information can actually tell us about our future health.

Alzheimer’s disease is marked by the buildup of two kinds of debris in the brain: amyloid plaques, sticky clumps of a protein fragment called beta-amyloid that collect outside brain cells, and tangles of a protein called tau that form inside them. These changes are thought to begin a decade or more before symptoms.

(Inside the the Quest to Deliver Alzheimer's Drugs Directly to the Brain.)

Both new tests measure p-tau217, a form of tau that increases in the blood as amyloid accumulates in the brain. It is considered one of the most accurate blood markers of amyloid studied so far. Elecsys pTau217, produced by Roche, uses antibodies to measure p-tau217. PrecivityAD2, produced by C2N, uses a mass spectrometer to measure several forms of beta-amyloid and tau to indicate a likelihood of amyloid presence.

But detecting these biological changes doesn’t tell you what they mean for someone’s future health.

“It depends on how you define Alzheimer’s, and there is debate about that,” Schindler says. In 2024, an Alzheimer’s Association workgroup redefined the disease by biology alone, meaning someone with no symptoms but abnormal biomarkers could be diagnosed. But in practice, the group doesn’t recommend testing asymptomatic people outside of research, and others argued many in this group might never develop symptoms. These brain changes are more common with age, and some people die with signs of Alzheimer’s in the brain, without having been symptomatic during life, says Richard Mayeux, a neurologist at Columbia University. A positive result “is not a death sentence, but it is a sign there’s something going on,” Mayeux says.

Interpreting the results can be complicated for other reasons, too. Kidney disease, for example, can falsely elevate blood test results, says Jonathan Graff-Radford, a neurologist at Mayo Clinic. And for people who are experiencing symptoms, there is a possibility they might be attributed to depression, medication, or other illnesses that are not necessarily Alzheimer’s, says Mayeux.

These uncertainties matter most for people without cognitive problems. Current clinical guidelines don’t recommend testing this group, in part because there isn’t enough data to know how accurately the tests perform in people without symptoms, Graff-Radford says.

(It may be possible to detect Alzheimer’s risk sooner—as early as your 20s.)

Still, some people arrive at the clinic with a result and then don’t know what to do, Mayeux says. Often, neither do their doctors. “We’re not sure how to manage these individuals,” Schindler says. “There aren’t clear recommendations on that.”

To complicate matters, some results aren’t clearly positive or negative, “and that’s actually by design,” Schindler says. When she and other experts set performance standards for these tests, they decided having more than 20 percent of people in this gray area “is not that informative.” Some uncertainty is unavoidable because the biological changes associated with Alzheimer’s develop gradually. “It’s not like you’re negative one day, and then you suddenly become positive,” Schindler says.

If a gray-area result won’t change any medical decisions, Schindler says it may make sense to repeat the test in a year or two, since people in that range are more likely to test positive eventually. Otherwise, if the person might qualify for treatment, she orders an amyloid PET scan.

Even when results are clearly positive, predicting what happens next is more complicated. Higher p-tau217 levels are associated with a greater risk of developing cognitive impairment. A July study in JAMA followed more than 2,600 cognitively healthy adults around age 70. Among those with high p-tau217 levels, an estimated 24 percent developed cognitive impairment within five years, compared with 38 percent of those with very high levels. Even in that highest-risk group, most did not develop cognitive impairment during those five years. Over 10 years, the risks were much higher (up to 78 percent), but the data are less reliable because few people were followed for that time period. “It’s predictive, but it’s not a guarantee,” Graff-Radford says.

Such findings can guide further research, but it’s “really premature to use them on an individual level,” Schindler says.

Age also plays a significant role. In that same study, a high p-tau217 level at 65 years old meant a 17 percent chance of cognitive impairment within five years. But at age 85, it was almost 40 percent.

Genes may also shape when symptoms emerge. In a September study of more than 8,500 people, Mayeux’s team found that once p-tau217 rose, people carrying APOE4, a genetic risk factor for Alzheimer’s, tended to develop cognitive problems sooner than those without it. But how much genes add to these predictions is less clear. The JAMA study found that carrying APOE4 didn’t change people’s odds of developing symptoms once researchers accounted for p-tau217 levels.

Even as researchers get better at predicting who might develop symptoms, knowing someone is at risk doesn’t necessarily mean doctors can prevent the disease.

“I always feel that screening tests should be linked to a possible treatment,” says Mayeux, pointing to mammograms and cholesterol checks as examples. For healthy people, Alzheimer’s blood testing doesn’t yet change care and risks unnecessary anxiety, extra tests, or inappropriate treatment.

The two anti-amyloid drugs on the market, lecanemab and donanemab, are approved only for people with mild cognitive impairment or mild dementia. They slow decline but “they don’t cure it, they don’t restore your memory,” Mayeux says. “Right now, I can’t imagine putting somebody on a lifelong monoclonal antibody,” who has no symptoms, he says.

(Why this new Alzheimer’s drug is eliciting both optimism and caution.)

Researchers are now testing whether treating people before symptoms appear could make a difference. Two clinical trials, AHEAD 3-45 and TRAILBLAZER-ALZ 3, are currently testing whether these drugs can delay decline in healthy people with early Alzheimer’s-related brain changes, but no symptoms. Graff-Radford says some results could be published in 2027, potentially changing how doctors test and treat people without symptoms.

Mayeux compares the possibility to taking statins to lower the risk of heart disease. “It’s not going to prevent you from having a heart attack, you still might,” he says. But it can lower the odds. If anti-amyloid drugs could do the same for Alzheimer’s, “then that’s a different ballgame.” Schindler expects testing would then become routine.

For now, doctors fall back on advice that’s good for almost everyone. “We recommend that people exercise, remain cognitively engaged, wear hearing aids if needed, sleep well, decrease their alcohol intake,” Schindler says, adding that patients typically don’t want to hear this. “Patients are looking for some kind of treatment.” In some cases, she says, enrolling in a research study might be an option.

Ultimately, Graff-Radford hopes the tests will get better at predicting who will actually develop symptoms. “Because that’s truly what people care about,” he says, “it’s not what’s necessarily happening in the brain,” but whether they’ll get sick.