Introduction
A 2026 study published in Nature Medicine explored whether a single blood test could estimate not just whether Alzheimer’s disease biology is present — but when symptoms are likely to begin.
Primary source: Petersen KK et al. Predicting onset of symptomatic Alzheimer’s disease with plasma p-tau217 clocks. Nature Medicine (2026).
Researchers used a plasma biomarker called %p-tau217 and built mathematical “clock models” to estimate:
- The age when a person likely became biomarker-positive
- The number of years until cognitive symptoms might develop
This represents a shift from binary testing (“positive or negative”) toward biological time staging.
What Is %p-tau217?
Tau is a protein that stabilizes neurons. In Alzheimer’s disease, tau becomes abnormally phosphorylated.
%p-tau217 measures:
The proportion of tau protein phosphorylated at position 217 relative to total tau.
It strongly correlates with:
- Amyloid PET imaging
- Tau PET imaging
- Cognitive decline
Because it can be measured from blood, it is far more accessible than brain imaging.
What Is an Alzheimer’s “Clock Model”?
Disease Clocks: Biomarkers as Time
A disease clock is a way of translating a biomarker level into an estimate of where someone is on a disease timeline—for example, “years since biomarker positivity,” rather than just “positive vs negative.”
What it does
- Anchors people to a biological event (like crossing a biomarker threshold).
- Aligns trajectories so individuals can be compared in “disease time.”
- Supports forecasting of clinical transitions (like symptom onset) at the group level.
Why it’s useful
- Reduces noise created by relying on chronological age alone.
- Helps identify who is likely to convert within a trial window (e.g., 3–5 years).
- Makes staging intuitive: “years since positivity” is easier to reason about than raw units.
Alzheimer’s example
In Alzheimer’s research, plasma %p-tau217 can be modeled longitudinally to estimate the age a person likely became biomarker-positive. That estimated “positivity age” can then be used to model the expected age of symptom onset (with meaningful error margins).
Key limitation
A clock is not a personal countdown timer. Prediction error, co-pathologies, and population differences can be large—especially when applied to individuals rather than cohorts.
Traditional risk models rely heavily on chronological age.
Clock models instead attempt to estimate biological time by:
- Identifying when a biomarker crosses a positivity threshold.
- Aligning individuals by “years since biomarker positivity.”
- Modeling when symptoms typically emerge after that biological event.
Two methods were used in the study:
- TIRA (Temporal Integration of Rate Accumulation)
- SILA (Sampled Iterative Local Approximation)
Both approaches produced similar results across independent cohorts.
How Accurate Was the Prediction?
Across cohorts:
- Median absolute error: approximately 3–4 years
- Good discrimination in ranking risk
- Consistent validation across two independent research groups
This level of precision is meaningful for clinical trials, where predicting group-level conversion windows is valuable.
It is not precise enough for personal forecasting.
Why Age Changes the Timeline
One of the most important findings:
The time from biomarker positivity to symptom onset depends strongly on age at positivity.
In general:
- Individuals who became biomarker-positive around age 60 had a longer interval before symptoms.
- Those who became positive around age 80 developed symptoms sooner.
This suggests that:
- Brain resilience declines with age
- Co-existing pathologies may accelerate symptom expression
- Clinical onset reflects system vulnerability, not pathology alone
Does This Support the Plaque Theory?
The study is consistent with modern amyloid and tau biomarker models of Alzheimer’s disease:
- %p-tau217 correlates strongly with amyloid PET and tau PET.
- Positivity thresholds align with established amyloid benchmarks.
However, the findings also reinforce that:
Symptoms emerge from the interaction between pathology and brain resilience.
Pathology is necessary — but not always sufficient — to determine timing of clinical decline.
Could This Approach Apply to ALS?
Possibly — but with important differences.
Clock modeling works best when:
- A biomarker changes predictably over a long preclinical period.
- There is a definable biological transition event.
- Disease progression follows a relatively stable trajectory once initiated.
Alzheimer’s disease fits these conditions.
ALS differs in that:
- Presymptomatic windows are shorter or less clearly defined.
- Progression is often more heterogeneous.
- Biomarkers such as neurofilament light reflect active neuronal injury rather than long buildup phases.
In ALS, models are more often used to estimate rate of decline rather than long-range onset timing.
Would Knowing an Onset Profile Change Anything?
Clinical Trials
Yes. This is the primary immediate application.
Clock models may help:
- Select individuals likely to convert within 3–5 years
- Reduce trial duration
- Increase statistical power
Clinical Practice
Current expert guidance discourages biomarker testing in cognitively unimpaired individuals outside research settings.
Clock models are not yet suitable for individual life-planning decisions.
FAQ
Is this blood test available now?
Some plasma p-tau217 assays are clinically available in certain contexts. Clock-based symptom timing prediction remains a research application.
Should I get tested if I have no symptoms?
Current expert guidance does not recommend biomarker testing for cognitively unimpaired individuals outside research studies.
Does this mean doctors can predict exactly when dementia will start?
No. The typical prediction error is around 3–4 years, which is not precise enough for exact forecasting.