Predicting When Alzheimer’s Symptoms Will Emerge

By Bio-IT World News Staff 

February 24, 2026 | 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 more have published the results of a study (Nature Medicine, DOI: 10.1038/s41591-026-04206-y) that determined a blood-based biomarker model that estimates not only the likelihood of Alzheimer’s disease but also when symptoms are likely to emerge.  

The study centers on the ratio of phosphorylated to non-phosphorylated plasma tau at position 217 (%p-tau217). The researchers chose this biomarker because “%p-tau217 captures the co-evolution of amyloid and tau pathologies across the AD continuum.” In other words, the plasma measures of phosphorylated tau and %p-tau217 have high associations with PET scans looking at amyloid plaque buildup and tau tangle buildup, as well as brain volume and cognition scores.  

Using longitudinal samples from approximately 900 participants in the Knight Alzheimer’s Disease Research Center and the Alzheimer’s Disease Neuroimaging Initiative, the research team constructed a mathematical “clock” model that tracks disease progression from the point of biomarker positivity. Two computational approaches—temporal integration of rate accumulation and sampled iterative local approximation—were applied to estimate when individuals transitioned to %p-tau217 positivity and to forecast symptom onset. Across cohorts and modeling strategies, predictions were broadly consistent, particularly within mid-range biomarker values. 

Alzheimer’s pathology can accumulate for 10 to 20 years before cognitive symptoms manifest, complicating clinical trial design and therapeutic timing. Having a way of estimating the onset of symptoms offers a strategic advantage in patient stratification. Commercial platforms are already positioned in this space, including C2N Diagnostics’ PrecivityAD2, Janssen and Quanterix’s LucentAD, Quanterix’s Simoa ALZpath, and Fujirebio’s Lumipulse assays—all of which incorporate phosphorylated tau measurements. 

The model is not yet suitable for individual prognostication, given a median error range of three to five years and limited demographic diversity in the study populations. However, as FNIH expands its AD Biosignature Project, the integration of additional biomarkers, such as eMTBR-tau243 or cerebrovascular disease indicators, may further refine predictive precision. 

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