Plasma p-Tau217 Testing

Peer-reviewed human evidence is now substantial. A JAMA 2026 pooled analysis (Buckley et al.) harmonized six longitudinal cohorts across North America, Japan and Australia (n=2,684 cognitively unimpaired adults, median follow-up 5.4 years, maximum 13.5 years, 478 progression events). Each 1-SD increase in baseline plasma p-tau217 carried a hazard ratio of 1.38 (95% CI 1.30-1.46) for progression to cognitive impairment, and 1.32 (1.24-1.41) after adjustment including amyloid-PET Centiloids. Critically, the study reports ABSOLUTE rather than relative risk: 24% (95% CI 20-28%) five-year risk at high p-tau217 (1.1-2.4 SD) and 38% (95% CI 33-43%) at very high (>2.5 SD). Elevated p-tau217 also predicted faster cognitive decline, independently of and synergistically with amyloid-PET burden. The important caveat is that these were selected research and clinical-trial cohorts; absolute risk figures will differ in unselected primary-care populations, and the authors call for validation there before individual prognostic use. UPDATE (August 16, 2026): PEER-REVIEWED. Nature Aging 2026 (Bellomo, Vermunt, del Campo et al.) reports a large multicentre plasma proteomics study identifying blood biomarkers that discriminate Alzheimer's disease, dementia with Lewy bodies and frontotemporal dementia, and that support molecular staging within disease. This matters for how p-tau217 is positioned: most blood-based neurodegeneration work to date, including p-tau217, addresses the AD-versus-not question, whereas differential diagnosis across syndromes is the harder and more clinically consequential problem - and it is where a single-analyte AD-specific marker is least informative. Assay standardisation across platforms and prospective validation in unselected memory-clinic populations remain outstanding, and cohort-based discrimination performance typically overstates real-world performance. This is not yet a deployable panel and does not displace p-tau217.

Evidence Scale