In The News: Single Test Could Predict Alzheimer's Symptoms Before They Even Begin

In The News: Single Test Could Predict Alzheimer's Symptoms Before They Even Begin

By Lydia Partick

Today's "In The News" examines a report concerning a new publication in Nature Medicine February 19, 2026, that a single blood plasma result from p-tau217 level couples with specific analysis by a mathematical model (now available in a web-based application) can not only make a diagnosis but can predict the time to onset of cognitive symptoms in people with no symptoms. Lead author, Kellen Petersen PhD, a mathematician, analyzed data from the Washington University Alzheimer's Disease Research Center (ADRC) and from the Alzheimer's Disease Neuroimaging Initiative (ADNI).

ADNI was established in 2004. It is a multicenter project (63 sites), which collects data from cognitively normal individuals, those with mild cognitive impairment (MCI), and those with mild AD dementia. Information about MRI, spinal fluid, and clinical status are stored. Amyloid PET imaging was added to the ADNI protocol in 2005, and tau PET was added in 2015. ADNI has been administered by the USC Epstein Family Alzheimer's Therapeutic Research Institute (ATRI) since 2015. 

The p-tau217 blood test is a biomarker: that is, it is a way for doctors to make a diagnosis of Alzheimer's Disease (AD) without directly imaging the brain to see the plaques and tangles pathology characteristic of the condition. Before biomarkers, like PET scans and blood tests, the only way to definitively know that someone had AD was by directly looking at brain tissue. The p-tau217 test is FDA approved for people over the age of 55 who have symptoms of dementia. However, scientists now believe that the test (which correlates with the presence of plaques and tangles) may be predictive of preclinical AD (the brain changes have started, but cognitive function is normal).

This study has gone further. The collaboration between mathematician and physician has led to what may be a method for not only predicting AD in asymptomatic people, but predicting when the symptoms are going to begin. 

Dr. Petersen indicated that the clock model could enhance clinical trial efficiencies by predicting which study participants were more likely to become symptomatic within certain timeframes. More importantly, it is hoped that as the model is refined and date from other serum biomarkers is studied, this model could be useful for individual patient care. 

The authors shared all computer code used for the clock model by creating a web-based application so that others can collaborate, and an improved refined model can be more rapidly made available. 

More research is needed before this is incorporated into standard practice for care, although it would be very valuable for clinicians to help them help their patients (and families) to plan for the future when the model is verified and more precise. Similarly, this could be enormously helpful for research into interventions created to slow or halt the progression of cognitive decline. 

The authors of this study, Dr. Kellen and Dr. Suzanne Schindler, are to be congratulated for a creative and collaborative effort.