Plasma p-tau217 Blood Test Predicts 5- and 10-Year Risk of Cognitive Decline
A blood test for p-tau217, combined with age, sex, and APOE ε4, estimates a cognitively healthy person's absolute risk of mild cognitive impairment.
Summary
Alzheimer disease changes in the brain can begin years before memory problems appear, yet some people with these changes never become symptomatic. Researchers from Mayo Clinic and the Alzheimer's Disease Neuroimaging Initiative followed 2,178 cognitively healthy adults aged 60 to 90 to see how a blood marker called plasma p-tau217 relates to later decline. They combined the marker with age, sex, and APOE ε4 gene status to estimate actual risk over 5 and 10 years. For a 75-year-old, going from a low to a high p-tau217 level raised 5-year risk of mild cognitive impairment from 7% to 17% and 10-year risk from 15% to 35%. Age mattered a great deal too: at a high p-tau217 level, 10-year risk was 17% at age 65 but 66% at age 85. The findings suggest a single blood test can give individualized risk estimates, useful for prognosis and for selecting people for prevention trials.
Detailed Summary
Alzheimer disease pathology can be detected years before symptoms appear, but not everyone with a positive biomarker goes on to develop cognitive problems. For people worried about their brain health, and for clinicians counseling them, the key question is not just whether a biomarker is positive but what the actual chance of decline is over a useful time horizon.
This cohort study drew on two sources: the population-based Mayo Clinic Study of Aging (1,644 participants in Olmsted County, Minnesota) and the multicenter Alzheimer's Disease Neuroimaging Initiative (534 participants). All were cognitively unimpaired adults aged 60 to 90 with baseline plasma p-tau217 measured on the Fujirebio Lumipulse platform, APOE genotyping, and longitudinal clinical follow-up. The investigators used continuous-time multistate models to estimate transitions from normal cognition to mild cognitive impairment (MCI) and dementia.
Higher p-tau217 was strongly associated with progression. Comparing the 90th with the 10th percentile (0.324 vs 0.072 pg/mL), hazards of clinical progression were roughly 2.6 to 3.0 times higher across sex and APOE ε4 groups in the Mayo cohort. In ADNI the estimates ranged from about 2.1 to 4.0. At age 75, 5-year MCI risk rose from 7% to 17%, and 10-year risk from 15% to 35%. Age had a powerful effect: at 0.324 pg/mL, 5-year MCI risk rose from 6% at age 65 to 37% at age 85, and 10-year risk from 17% to 66%. Results were similar using the p-tau217/Aβ42 ratio.
The authors conclude that plasma p-tau217, combined with age, sex, and APOE ε4 status, enables individualized absolute risk estimates for cognitively unimpaired people. This could support prognosis discussions and enrich preclinical Alzheimer trials with those most likely to progress.
Several caveats apply. Both cohorts are research populations that may not reflect broader, more diverse groups. A single baseline measurement was used. This summary relies on the abstract alone, so details on modeling, missing data, and subgroup performance are unavailable.
Key Findings
- A high versus low p-tau217 level (0.324 vs 0.072 pg/mL) was linked to about 2.6 to 3.0 times greater hazard of progression in the Mayo cohort.
- At age 75, 10-year MCI risk rose from 15% to 35% going from low to high plasma p-tau217.
- At high p-tau217, 10-year MCI risk was 17% at age 65 versus 66% at age 85.
- Combining p-tau217 with age, sex, and APOE ε4 allows individualized absolute risk estimates for cognitively healthy adults.
- The p-tau217/Aβ42 ratio performed similarly to p-tau217 alone.
Methodology
Cohort study of cognitively unimpaired adults aged 60 to 90 from the Mayo Clinic Study of Aging (n=1,644) and ADNI (n=534), with baseline Lumipulse plasma p-tau217, APOE genotyping, and long-term follow-up. Continuous-time multistate models estimated transitions from normal cognition to MCI and dementia, yielding 5- and 10-year absolute risks.
Study Limitations
This summary is based on the abstract only, so detailed methods and subgroup analyses could not be reviewed. The cohorts are research populations that may limit generalizability. Risk estimates rely on a single baseline measurement on one assay platform.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe:
