A Blood Test Reads Alzheimer's Risk Years Early

A single blood draw, taken from someone who feels perfectly sharp and may not show any obvious symptoms may be able to estimate how likely that person is to face Alzheimer’s-related memory problems years down the road. That is what researchers at the Mass General Brigham Neuroscience Institute reported in a large new study, published in JAMA and presented at the Alzheimer’s Association International Conference.

The marker at the center of the work is p-tau217, a form of the tau protein that circulates in the blood. In a cognitively healthy older adult, its level can help gauge the likelihood of developing cognitive impairment over the next five to ten years, which is where the framing of a warning up to a decade early comes from.

The finding rests on a deep pool of data. Researchers combined six long-running studies from North America, Japan, and Australia, some observational and some clinical trials, covering 2,684 adults who showed no cognitive impairment at the outset. Some participants had been tracked since 2004; the most recent follow-ups came in 2025. Every participant started with a baseline p-tau217 measurement and a brain scan, then came back year after year for reassessment. By the end, roughly 478 had crossed into cognitive impairment.

The gap between high and low readings was stark. Among the people who began with the highest p-tau217 levels, the estimated risk of cognitive impairment was 38 percent over five years and 78 percent over ten. The researchers hold that ten-year number more loosely, since far fewer people had been followed that long and the data behind it is thinner.

On its own, the marker forecast risk without leaning on amyloid-beta plaques from a PET scan or genetic risk factors such as APOE4. That independence matters in practice, because PET imaging is costly and hard to access, while a blood test is neither.

The test estimates risk. It does not diagnose Alzheimer’s, cure it, or work as a screening test for the general public. No treatment yet changes the course of the disease for people flagged as high-risk, and the researchers are not suggesting that healthy people go ask for the test today. “Today, our medical advice would remain the same regardless of test results: exercise regularly, maintain a healthy diet, and prioritize sleep and overall wellness,” said Dr. Reisa Sperling, the study’s senior author.

The value of knowing, for now, is on the research side. Prevention research runs on exactly this kind of information. To test whether a therapy can slow or stop Alzheimer’s before it starts, scientists need to enroll people who are actually on that path, and to know roughly how soon it may arrive. A marker that ranks risk by degree and by timeline gives those trials a way to find the right participants. And if the treatments now under study prove out, the same blood test could help identify who stands to benefit and reach them early, while there is more of the brain left to protect.

Sperling described the destination in familiar terms. “Our long-term goal is to get us to where cholesterol testing is in predicting your risk of a heart attack,” she said.

The comparison is a reassuring one. A high cholesterol number works as a prompt: it tells a person and their doctor where to pay attention and what to try, well before any crisis. The hope for p-tau217 is a version of that for the brain, a way to move the whole conversation earlier, into the years when there is the most time to act.