Blood test for tau protein shows promise as Alzheimer's screening tool

A simple blood test might soon become the first line of defense against Alzheimer's disease. Researchers working with data from the Alzheimer's Disease Neuroimaging Initiative have found that measuring a protein fragment called p-tau217 in plasma can reliably detect the presence of tau tangles in the brain—the hallmark pathology of Alzheimer's—without requiring expensive PET imaging scans.

The study examined 419 participants across three diagnostic categories: 229 who were cognitively normal, 149 with mild cognitive impairment, and 41 with Alzheimer's disease dementia. All had their blood drawn for analysis of five biomarkers—p-tau217, amyloid-beta 42, amyloid-beta 40, neurofilament light chain, and glial fibrillary acidic protein—and all underwent tau PET imaging to establish ground truth. The researchers then tested how well the blood markers could predict which brains actually harbored tau pathology.

P-tau217 emerged as the strongest single predictor. Across the entire cohort, it achieved an area under the curve of 0.671, meaning it correctly distinguished tau-positive from tau-negative cases about 67 percent of the time—better than any other blood marker tested. But the real story lay in the variation by diagnosis. In people already diagnosed with Alzheimer's dementia, p-tau217 performed far more reliably, reaching an AUC of 0.785. In cognitively normal individuals and those with mild cognitive impairment, accuracy dropped considerably, a reminder that biomarkers behave differently depending on where someone sits on the disease spectrum.

One of the study's most practical contributions was the introduction of a two-cutoff strategy rather than a single threshold. In clinical practice, blood tests often produce ambiguous middle-zone results that leave doctors uncertain whether to recommend further imaging. By establishing both a lower and upper cutoff for p-tau217, the researchers substantially reduced these gray-zone classifications. This matters because it streamlines decision-making: results either clearly suggest tau pathology is present, clearly suggest it is absent, or fall into a narrower band requiring confirmatory imaging. The approach could accelerate recruitment into clinical trials and make screening more efficient in real-world settings.

The findings come from a harmonized dataset using a standardized tau-PET threshold derived independently across multiple sites, lending them credibility. The work was supported by the National Institute on Aging, the National Institute of Biomedical Imaging and Bioengineering, and the Canadian Institutes of Health Research, among other funders. Yet the researchers are careful not to overstate their results. Overall accuracy remained moderate, and performance varied significantly by diagnostic group. They acknowledge that further validation across different populations—different ages, ethnicities, and disease stages—is essential before p-tau217 can be confidently deployed as a first-line screening tool in diverse clinical settings.

What the study does establish is a proof of concept: a blood test can approximate what a brain imaging scan reveals about tau pathology, at least in some populations. If that finding holds up in larger, more diverse cohorts, it could transform how Alzheimer's screening works. Instead of starting with expensive PET scans, clinicians could begin with a simple blood draw, reserving imaging for those whose results warrant it. For a disease that affects millions and for which early detection increasingly matters, that shift could mean the difference between catching pathology early and missing the window for intervention.