p-Tau217: Is the Alzheimer's Blood Test like a looking glass portending the future?

For those of us at increased genetic risk for Alzheimer’s disease - I know the dilemma! For years, I was one of those sitting on the fence about imaging and testing for AD pathology! During my visit to RetainMed in Boca Raton on July 1st - I decided it was time to get a baseline and see if my extensive AD prevention protocol is making a difference and protecting my brain. While the verdict is still out - I’ve been doing a deeper dive into the topic in preparation for what I’m told when I have my follow-up visit next week!

Is Alzheimer’s pathology actually developing in my brain and if so, how much time do I have before it affects my cognition - and - more importantly - is there anything else I can do about it (considering my already extensive prevention protocol).

Until very recently, we haven’t had a good way to answer that question.

Now a simple blood biomarker called p-tau217 has changed the conversation.

We’ve already learned that p-tau217 can identify Alzheimer’s pathology with remarkable accuracy.

But research published in JAMA in July 2026 suggests something much more consequential:

p-tau217 may help predict the likelihood that a cognitively healthy person will develop cognitive impairment years into the future.

And that raises an important question: If p-tau217 can predict our future, can we actually still change it?

Increasingly, the answer appears to be yes.

What Exactly Is p-Tau217?

Tau is a protein found primarily inside neurons, where one of its normal jobs is helping stabilize the cellular structures called microtubules.

In Alzheimer’s disease, tau becomes abnormally phosphorylated.

One particular form - tau phosphorylated at amino acid 217, or p-tau217 - levels rise in the blood in association with Alzheimer’s pathology, particularly amyloid accumulation.

Importantly, p-tau217 can begin increasing years before someone develops noticeable cognitive symptoms.

That makes it very different from a cognitive test.

A cognitive test asks:

How well is my brain functioning today?

p-Tau217 potentially asks:

Is the Alzheimer’s disease process already underway?

And now researchers are asking whether it can tell us something about where that process is headed.

Nearly 2,700 Cognitively Healthy People

The new study pooled data from 2,684 cognitively unimpaired older adults enrolled in six longitudinal studies in North America, Australia and Japan.

Their median age was about 70.

Researchers measured baseline plasma p-tau217 and then followed participants for a median of 5.4 years, with some followed for as long as 13.5 years.

During that time, 478 participants progressed to cognitive impairment.

The relationship with p-tau217 was striking.

For every one-standard-deviation increase in p-tau217, the risk of progressing to cognitive impairment increased by approximately 38%.

To explain Standard Deviation:

• Average p-tau217 = 0.20 pg/mL

• Standard deviation = 0.10 pg/mL

• Someone at 0.20 is around the average.

• Someone at 0.30 is 1 SD above the average.

• Someone at 0.40 is 2 SD above the average.

But the absolute numbers are even more interesting.

What Happened Over Five Years?

Researchers divided participants into four p-tau217 categories.

Their estimated five-year risks of developing cognitive impairment were:

Low p-tau217: 12%

Intermediate: 15%

High: 24%

Very high: 38%

At ten years the modeled estimates became:

Low: 40%

Intermediate: 45%

High: 62%

Very high: 78%

That last number certainly gets your attention.

Someone in the very-high p-tau217 category had an estimated 78% probability of developing cognitive impairment within ten years.

There is an important caveat: only about 5% of participants had more than ten years of follow-up, so the researchers specifically caution that the ten-year estimates are considerably less certain than the five-year numbers.

Still, this represents an extraordinary change in what an Alzheimer’s blood test might eventually be able to tell us.

It Predicted Cognitive Decline Even Beyond Amyloid

One finding I found particularly interesting was that p-tau217 wasn’t simply acting as a cheap substitute for an amyloid PET scan.

The researchers adjusted their analysis for amyloid PET burden.

p-tau217 still independently predicted progression to cognitive impairment.

Higher p-tau217 was also associated with faster decline on longitudinal cognitive testing.

In other words, this blood marker appears to contain information about disease trajectory that isn’t captured simply by knowing how much amyloid someone has.

That makes p-tau217 potentially much more than a diagnostic test.

It could become a prognostic biomarker.

There’s Now an Alzheimer’s “Clock,” Too

Another 2026 study took this concept even further.

Researchers at Washington University followed more than 600 older adults with serial blood measurements and developed a model based on the age at which p-tau217 becomes abnormal.

They found that this could help estimate when Alzheimer’s symptoms were likely to begin.

The predicted onset was generally within approximately three to four years of the actual age of symptom onset.

We’re clearly not at the point where your doctor can look at your blood test and say:

“You will develop Alzheimer’s at age 78.”

But we’re moving remarkably closer to something that would have sounded like science fiction not very long ago.

But Here’s Where This Gets Really Interesting

A biomarker that predicts disease is useful.

A biomarker that predicts disease and can be changed could be transformative.

And we already have evidence that p-tau217 isn’t simply a fixed number marching inevitably upward.

Treatments can move it.

That distinction is enormously important.

Our APOE genotype is fixed. Our age is certainly fixed! But p-tau217 is a dynamic biomarker.

And one of the most surprising demonstrations of that has come from a drug that wasn’t originally developed as an Alzheimer’s drug at all.

Enter Obicetrapib

Obicetrapib is a CETP inhibitor (written about last week) developed primarily to treat cardiovascular disease by dramatically changing lipoprotein metabolism.

But investigators analyzing Alzheimer’s biomarkers in the Phase 3 BROADWAY cardiovascular trial found something unexpected.

The effect appeared particularly pronounced among APOE4 carriers.

In the APOE4/4 subgroup, p-tau217 fell approximately 7.8% over 12 months in participants receiving obicetrapib while increasing approximately 12.7% in the placebo group.

That produced a placebo-adjusted difference of approximately:

20.5%

And p-tau217 wasn’t the only Alzheimer’s-related biomarker moving in a favorable direction.

Changes were also reported in Aβ42/40, the p-tau217/Aβ ratio, GFAP and NfL.

That doesn’t prove Obicetrapib prevents Alzheimer’s.

It doesn’t prove that reducing p-tau217 by 20% reduces dementia risk by 20%.

And the APOE4/4 subgroup was small. But it does prove something extremely important for how we think about this biomarker:

p-tau217 can change.

So Here’s the Billion-Dollar Question

Put these two discoveries together.

The new JAMA study tells us:

Higher p-tau217 → substantially greater future risk of cognitive impairment.

The Obicetrapib data tell us:

An intervention can lower or stabilize p-tau217.

Which leads to the question that Alzheimer’s prevention research now urgently needs to answer:

If we lower p-tau217 early enough, do we lower the person’s future risk of cognitive decline?

We don’t know yet.

But that is an entirely different question from the one Alzheimer’s researchers were asking even a few years ago.

Neurology Is Starting to Look Like Cardiology

Cardiology doesn’t wait until someone has a heart attack before worrying about cardiovascular disease.

We measure LDL.

We measure ApoB.

We measure blood pressure.

We measure coronary calcium.

Then we identify risk and attempt to change the trajectory before the clinical event occurs.

Imagine doing something similar with Alzheimer’s disease!

At some point we may routinely measure:

APOE genotype → underlying genetic risk

Aβ42/40 → amyloid biology

p-tau217 → Alzheimer’s disease activity and trajectory

GFAP → astrocyte activation

NfL → neuronal injury

And instead of waiting for someone’s memory to fail, we may intervene while that person is still completely cognitively normal.

Then we measure again.

Did p-tau217 rise?

Did it stabilize?

Did it fall?

That would represent a fundamental change in Alzheimer’s medicine.

One Test Probably Won’t Be Enough

This also changes how I think about having a p-tau217 test performed.

A single measurement is useful.

But a baseline and capturing this marker over the years offers much more insight.

Suppose someone has p-tau217 measured at age 65, again at 66 and again at 67.

There are three very different possibilities:

0.12 → 0.12 → 0.11

versus

0.12 → 0.18 → 0.27

versus

0.27 → 0.22 → 0.17 following an intervention

The starting number / baseline is important to know. And the direction over time will likely matter just as much.

We don’t yet have validated clinical rules allowing us to interpret those trajectories that simply. Different p-tau217 assays also produce different numerical values, so numbers cannot necessarily be compared across laboratories.

But longitudinal monitoring is where I suspect these biomarkers are ultimately heading.

What This Means for APOE4 Carriers

For APOE4 carriers, I think this development is particularly significant.

Knowing that we carry APOE4 tells us that our lifetime probability of developing Alzheimer’s is elevated.

It does not tell us whether Alzheimer’s pathology is developing in our brain today.

That’s an enormous distinction.

A 55-year-old APOE4 carrier with normal Alzheimer’s biomarkers is in a fundamentally different biological situation from a 55-year-old APOE4 carrier whose p-tau217 is already rising.

Genetics tells us about risk.

Biomarkers may increasingly tell us about disease activity.

And perhaps, eventually, repeated biomarkers will tell us whether our prevention strategies are actually working.

We’re Not Quite There Yet

I don’t want to get ahead of the science.

The new p-tau217 risk estimates apply to populations, not with certainty to individuals.

The ten-year numbers are based on relatively sparse long-term follow-up.

We don’t yet have universally standardized thresholds across all p-tau217 assays.

And most importantly:

We have not yet proven that lowering plasma p-tau217 in an asymptomatic person prevents Alzheimer’s dementia.

That’s the missing piece.

For most of my life, Alzheimer’s disease has been diagnosed by what it eventually did to someone’s memory.

Now we can increasingly see its biology in a vial of blood years before that happens.

And perhaps the most encouraging development of all is that the number in that vial isn’t necessarily written in stone.

We can measure p-tau217.

It can predict future cognitive risk.

We’ve demonstrated that it can change.

Now we need to find out whether changing it can change our future.

For those of us carrying APOE4, I can’t think of many more important questions for Alzheimer’s prevention research to answer.