By Clinical Research News Staff 

March 26, 2026 | In a study published in Nature Medicine (DOI: https://doi.org/10.1038/s41591-026-04206-y), researchers from the Foundation for the National Institutes of Health (FNIH), Washington University in St. Louis, the University of California San Francisco, the University of Wisconsin-Madison, and other collaborators found that a plasma biomarker—the ratio of phosphorylated to non-phosphorylated plasma tau at position 217, written %p-tau217—can be used to construct a “clock model” that can predict when Alzheimer’s disease symptoms appear.  The clock model holds strong potential in clinical research settings. 

Amyloid plaques can accumulate 10-20 years before cognitive decline becomes evident, with tau pathology emerging later and tracking with symptom severity. Identifying individuals in this preclinical window has become a central objective for therapeutic development. 

To that end, researchers selected %p-tau217 as a biomarker because it was shown to correlate strongly with amyloid and tau PET imaging, brain atrophy, and cognitive performance. The researchers used %p-tau217 measurements from two longitudinal cohorts from the Knight Alzheimer’s Disease Research Center and the Alzheimer’s Disease Neuroimaging Initiative, totaling roughly 900 participants followed for five to seven years. 

The team also used two mathematical modeling approaches—temporal integration of rate accumulation and sampled iterative local approximation—to estimate when individuals crossed a biomarker positivity threshold and to project forward to expected symptom onset. 

Although predictions were not reliable at upper and lower levels of %p-tau217, the models were largely concordant across methods and cohorts. Notably, demographic variables such as APOE4 status, sex, and education exerted minimal influence on projected timelines. More striking was an age-dependent effect: individuals who became %p-tau217 positive at age 60 were estimated to develop symptoms approximately 14 years later, compared with 6.2 years for those positive at age 80. 

For Alessio Travaglia, director of translational science and neuroscience at the FNIH, this discovery was unexpected. “Certainly it's going to help the design of clinical trials moving forward clinical trials moving forward, because I could set a very specific enrollment criteria. I could have a specific age as my inclusion criteria.” The clock model could allow sponsors to recruit appropriate participants to get useful data.  

The authors caution that the model is not suitable for individual decision-making. With a median error of three to five years and limited racial diversity in the underlying cohorts, generalizability remains constrained. Ongoing expansion efforts, including FNIH’s Alzheimer's disease BioSignature Project, aim to validate and refine the approach in broader populations. 

To read the full story written by Allison Proffitt, visit Diagnostics World News.