Blood Tests Could Predict Alzheimer’s Years Before Symptoms Appear

0 Share Newsweek is a Trust Project member See more of our trusted coverage when you search. Prefer Newsweek on Google to see more of our trusted coverage when you search. A blood biomarker increasingly used to identify Alzheimer's disease may be less effective than expected at predicting who will develop the condition before symptoms appear, according to a new study.

Researchers found that P-tau 217, one of the most prominent blood-based biomarkers for Alzheimer's, had only moderate predictive power in people who had no cognitive symptoms. In contrast, a biomarker that detects amyloid beta protein misfolding appeared to identify future Alzheimer's risk much earlier in the disease process.

The study analyzed blood samples from the ESTHER study, a large long-term investigation of the health of older adults, to compare how well different biomarkers could predict eventual Alzheimer's dementia in people who were symptom-free at the time of testing.

The findings suggest that different biomarkers may be useful at different stages of the disease. While P-tau 217 showed "outstanding diagnostic effectiveness" once Alzheimer's had reached the clinical stage, its predictive performance was significantly weaker before symptoms emerged.

The findings arrive as disease-modifying anti-amyloid therapies are becoming available for some patients with Alzheimer's disease. These treatments can slow disease progression, but they are intended to be used during the early symptomatic stages, including mild cognitive impairment and mild Alzheimer's dementia.

The therapies can also cause serious side effects, including amyloid-related imaging abnormalities (ARIA), which involve brain swelling and microbleeds. As a result, researchers say identifying the right patients as early as possible has become an increasingly important goal.

Researchers reported that P-tau 217 achieved an area under the curve (AUC) score of 0.67 during the preclinical stage of the disease. By comparison, a biomarker measuring amyloid beta protein misfolding achieved an AUC of 0.79 in cognitively normal individuals.

When researchers combined the protein misfolding marker with demographic information, genetic factors and other blood biomarkers, the resulting biomarker panel reached an AUC of 0.87, allowing highly accurate prediction of a future Alzheimer's diagnosis long before symptoms appeared.

Prof. Dr. Klaus Gerwert, director of the Center for Protein Diagnostics (ProDi) at Ruhr University Bochum, told Newsweek the findings reflect the order in which disease-related changes occur in the brain.

"Because Alzheimer's starts with misfolding of a beta, this induces plaque formation and also phosphorylation of tau 217. Therefore, tau 217 is an excellent but late biomarker indicating the symptomatic stage. Misfolding indicates the initial molecular changes," he said.

According to the researchers, the results suggest that protein misfolding represents the earliest measurable blood-based marker of Alzheimer's disease. While P-tau 217 remains highly effective once the disease has clinically manifested, Gerwert and colleagues argue that it may not be the best tool for identifying people at increased risk years before symptoms begin.

Dr. Jennifer Timmons, a longevity physician and founder and medical director of Timmons Wellness, told Newsweek that predicting who will eventually develop Alzheimer's remains challenging.

She noted that Alzheimer's is influenced by multiple factors, making prediction more complex than any single biomarker can capture.

One complication, according to Timmons, is that Alzheimer's is “not caused by just one factor. It's determined by vascular health, genetics, inflammation, and metabolic health, so it's just more nuanced and complicated. All of those factors play into whether you develop Alzheimer's or not.”

However, experts say several hurdles remain before such testing could become routine.

"We need more evidence in more diverse populations, and definitely in people who do not have cognitive symptoms," Timmons said. "We need more evidence of how accurate these new blood tests are and how the results should be interpreted in people with no cognitive symptoms."

She added that physicians would also need clear guidelines on who should be tested, as well as confirmatory testing, specialist counseling and established treatment pathways.

Gerwert said regulatory approval remains one of the biggest obstacles.

"We need the certification of FDA and EMA. To overcome this hurdle as a small start-up is very cost-intensive. We do think it will take only few more years," he said.

Timmons also cautioned that identifying people as high risk years before symptoms develop carries potential downsides.

"Learning you are at high risk for Alzheimer's can create anxiety, uncertainty, and fear for patients," she said.

She emphasized that neither genetic tests nor current blood tests provide a definitive diagnosis and that patients should understand the limitations before undergoing screening.

Trares, K., Duman, D., Beyer, L., et al. (2026). Predictive value of plasma p-tau217 for Alzheimer's dementia compared to other blood-based biomarkers. EMBO Molecular Medicine , 10.1038/s44321-026-00483-9. Advance online publication. https://doi.org/10.1038/s44321-026-00483-9

Contact Newsweek editors on this story: Kara Dolman and Gray R. Thomas

Indexed and credited by AIPROPX. Originating outlet: Newsweek. Open at source →