Blood test may predict start of Alzheimer's disease symptoms

An NIH-funded team developed a model using blood p‑tau217 levels to estimate when Alzheimer’s symptoms begin, reporting a median absolute error of three to four years; the models are not yet accurate enough for individual use.

Researchers led by Dr. Suzanne Schindler at Washington University developed a model that uses blood levels of the protein p‑tau217 to estimate when Alzheimer’s symptoms might begin. The team analyzed serial p‑tau217 measurements from 603 older people and published the results in Nature Medicine on Feb. 19, 2026. The work builds on earlier studies showing p‑tau217 can identify brain amyloid plaques and tau tangles before symptoms appear. Key findings: - The model estimates the age when p‑tau217 levels first become elevated, and that estimated age correlated with the age at symptom onset with a median absolute error of three to four years. - The interval from p‑tau217 elevation to symptom onset was shorter at older ages; reported examples were about 20 years from elevation at age 60 and about 11 years from elevation at age 80. - The authors note the models are not yet accurate enough for individual clinical use, and the p‑tau217 blood test is not recommended for cognitively normal people or as a stand-alone diagnostic tool. Summary: The study suggests a single p‑tau217 blood measurement could help estimate timing until symptomatic Alzheimer’s at a group level, but current models are not suitable for individual prognoses. Further research is underway to refine the models and improve prediction accuracy.