Overview
- Washington University researchers report in Nature Medicine that plasma p‑tau217 models estimate the age symptoms will begin with a margin of roughly three to four years.
- The models were trained on longitudinal data from 603 adults in the Knight ADRC and ADNI cohorts and showed consistent performance across multiple assays, including PrecivityAD2 and an FDA‑cleared p‑tau217 test.
- Timing varied by age, with elevation at about 60 years predicting symptoms roughly 20 years later versus about 11 years when elevation occurred around age 80.
- The team released all model‑development code and a public web tool to enable transparency, external validation, and refinement.
- Researchers emphasize near‑term value for selecting prevention‑trial participants likely to develop symptoms within a study window, and current guidance does not recommend testing cognitively unimpaired people outside research.