Researchers found sits at the center of this dementia and brain health question.

Researchers have identified a groundbreaking combination of four blood proteins that can predict dementia risk with approximately 90% accuracy—sometimes more than 15 years before symptoms appear. These markers, known as GFAP, NEFL, GDF15, and LTBP2, work together to identify who is most likely to develop dementia in the coming years, giving both patients and clinicians a precious window for early intervention. This breakthrough comes from large-scale analysis of the UK Biobank and other international cohorts, representing one of the most significant advances in dementia detection in recent years.

The ability to predict dementia so far in advance changes everything about how we approach brain health. Instead of waiting for cognitive decline to become noticeable, people can now learn their risk status through a simple blood test and take action while their brain is still largely protected. For families with a history of dementia, for aging adults concerned about their memory, and for healthcare providers trying to identify high-risk patients early, these biomarkers offer clarity and hope. This article explains what these four blood markers are, how they work together to predict dementia across all types (Alzheimer’s, vascular, and others), what the 90% accuracy actually means in practical terms, and what the next steps are for people who learn they’re at higher risk.

What Are the Four Blood Markers That Predict Dementia?
How Accurate Is 90% Prediction, and What Does It Actually Mean?
How Do These Blood Markers Differ Across Dementia Types?
How Soon Can You Get a Blood Test for These Markers?
What’s the Relationship Between These Blood Markers and Actual Brain Changes?
What Can You Do If Your Blood Markers Indicate Higher Dementia Risk?
The Future of Dementia Prediction and Prevention
Conclusion

What Are the Four Blood Markers That Predict Dementia?

The four proteins identified as the strongest predictors of dementia are GFAP, NEFL, GDF15, and LTBP2—each playing a distinct role in how the brain ages and degenerates. GFAP (glial fibrillary acidic protein) is the single strongest predictor; people with higher GFAP levels are more than twice as likely to develop dementia compared to those with low levels. This protein is produced by astrocytes, support cells in the brain that activate when nerve cells are damaged or stressed. When GFAP is elevated, it signals that your brain is already experiencing inflammation or injury at the cellular level, even if you feel completely fine. NEFL (neurofilament light chain) is the second-strongest marker and specifically indicates nerve fiber damage and degeneration.

Think of NEFL as a warning light that neurons themselves are breaking down; high levels suggest your brain’s wiring is already showing signs of wear. Together, GFAP and NEFL provide complementary information—one about brain inflammation, the other about direct neuronal damage—which is why combining them improves prediction accuracy significantly. GDF15 and LTBP2 are both involved in inflammation, cell growth and death, and the way cells respond to stress—all processes tied to increased dementia risk. Unlike GFAP and NEFL, which directly reflect what’s happening in the brain tissue, GDF15 and LTBP2 are part of your body’s systemic inflammatory response. When all four proteins are elevated together, they paint a complete picture: your brain is inflamed, your neurons are degenerating, and your body’s stress response systems are activated.

What Are the Four Blood Markers That Predict Dementia?

How Accurate Is 90% Prediction, and What Does It Actually Mean?

When researchers say these four markers predict dementia with 90% accuracy, they mean that when you combine the protein levels with conventional risk factors—age, sex, education level, and genetic history (APOE status)—the model correctly identifies who will develop dementia 90% of the time. This is exceptionally high compared to other predictive tools in medicine. However, there’s an important distinction: 90% accuracy in a research study on a UK Biobank cohort doesn’t automatically translate to 90% accuracy for every individual person. The 15-year prediction window is equally important: researchers can “quite reliably predict dementia 15 years before the diagnosis of the disease,” according to lead researcher Jianfeng Feng.

This is transformative because it means you’re not getting a general risk assessment—you’re getting something much more specific. If your blood markers indicate high dementia risk, there’s a 15-year window during which early interventions (lifestyle changes, cognitive training, management of cardiovascular risk factors) might actually prevent or delay the disease. However, there’s a limitation to keep in mind: these markers have been validated across large research cohorts, but they’re not yet routine clinical tests available in every doctor’s office. The blood tests exist and are being used in research settings, but broader clinical implementation is still underway. Also, prediction models perform better in research cohorts than in real-world populations; once these tests become widely available, real-world accuracy may be slightly lower, especially in populations different from the UK Biobank sample.

How Do These Blood Markers Differ Across Dementia Types?

The remarkable aspect of these four markers is that they predict dementia risk across all the major types: Alzheimer’s disease (which accounts for 70% of all dementia cases), vascular dementia (20% of cases), and other types of dementia. This is different from some other biomarkers that are highly specific to Alzheimer’s but don’t capture risk for other dementia subtypes. GFAP, NEFL, GDF15, and LTBP2 appear to reflect general brain aging and neurodegeneration processes common to multiple dementia pathologies. For example, someone with elevated GFAP and NEFL might develop Alzheimer’s disease specifically, with tangles of tau and amyloid in the brain. Someone else with the same blood marker profile might develop vascular dementia, where small strokes accumulate and damage brain tissue.

The blood markers themselves don’t tell you which type you’ll develop—they tell you that something is going wrong in your brain tissue. Additional testing, like genetic screening for the APOE4 gene or advanced imaging, can help narrow down which type of dementia you’re at higher risk for. This universality across dementia types is why these markers are potentially so valuable in clinical practice. You don’t need separate tests for different dementia pathways; one blood test captures risk across the board. It also means that prevention and early intervention strategies, which increasingly focus on cardiovascular health, cognitive stimulation, and social engagement, are appropriate across dementia types.

How Do These Blood Markers Differ Across Dementia Types?

How Soon Can You Get a Blood Test for These Markers?

These blood tests aren’t yet available at most standard doctor’s offices or commercial labs like LabCorp or Quest Diagnostics, though that’s changing rapidly. Currently, they’re used primarily in research settings and specialized dementia clinics affiliated with major medical centers. If you’re interested in testing, your options depend on where you live and what research studies are happening in your area. Several medical centers have begun offering these tests as part of cognitive aging screening programs.

If you have memory concerns or a strong family history of dementia, talking to your primary care doctor about access to these biomarkers is reasonable—they may be able to refer you to a dementia specialist or memory clinic that offers testing. Some memory clinics also participate in research studies where participants get tested at no cost. Within the next few years, these tests are expected to become more widely available, especially as more insurance companies recognize their value in identifying people who could benefit from early interventions. The transition from research tool to clinical reality is underway, but we’re not quite at the point where any primary care doctor can order them as easily as a standard blood sugar test. In the meantime, if testing interests you, don’t hesitate to ask your doctor about participation in research studies or referral to a specialized memory clinic.

What’s the Relationship Between These Blood Markers and Actual Brain Changes?

The four markers predict dementia because they’re reflecting actual pathological changes happening in your brain tissue. GFAP elevation indicates that astrocytes—the brain’s support cells—are activated, which happens when neurons are damaged or inflamed. NEFL in the bloodstream comes from broken down nerve fibers, so elevated levels literally mean neurons are degenerating. GDF15 and LTBP2 are produced as part of the inflammatory cascade that accompanies neurodegeneration. This is different from some biomarkers that are merely correlated with disease risk.

These markers aren’t just statistical associations; they reflect biological processes directly involved in how the brain ages and fails. This is why they’re such strong predictors—they’re measuring the actual problem, not just factors that happen to be associated with the problem. One important caveat: elevated blood markers don’t guarantee you’ll definitely develop dementia, even with 90% accuracy. The 10% of people who don’t develop dementia despite elevated markers might have better resilience, more cognitive reserve, or other protective factors that prevent clinical symptoms from emerging. This is why these tests are best understood as a risk indicator that should prompt conversation with a doctor about prevention strategies, not as a definitive diagnosis or guarantee of future disease.

What's the Relationship Between These Blood Markers and Actual Brain Changes?

What Can You Do If Your Blood Markers Indicate Higher Dementia Risk?

If testing becomes available and your results suggest elevated dementia risk, the most important step is to focus on modifiable risk factors. Blood marker levels are largely fixed—you can’t lower GFAP or NEFL through an intervention—but you can still substantially reduce dementia risk through lifestyle changes.

This is where having an early warning matters most: with 15 years to work with, you have time to build cognitive reserve and cardiovascular protection. The most evidence-backed interventions include maintaining cardiovascular health (controlling blood pressure, managing cholesterol, not smoking), cognitive engagement (learning new skills, challenging your brain), physical exercise (aim for 150 minutes of moderate activity weekly), quality sleep (7-9 hours nightly), managing depression and isolation, and following a Mediterranean or DASH diet rich in plants and low in processed foods. A person with elevated blood markers who implements these changes might still develop cognitive decline, but they’re likely to develop it later and more slowly than someone who doesn’t take action.

The Future of Dementia Prediction and Prevention

These blood markers represent a shift in how medicine approaches dementia—from waiting for symptoms to appear, to predicting and preventing disease years in advance. As testing becomes more widespread and clinicians gain experience using these markers to guide prevention, we’re likely to see improved outcomes simply because people are intervening earlier. The concept of “precision dementia prevention” is becoming real: instead of recommending the same prevention strategies to everyone over 65, doctors can identify who actually needs intensive intervention.

The research pipeline continues to advance. Scientists are investigating additional blood protein combinations, exploring whether cognitive training or medications might be most effective for people with specific biomarker profiles, and working to make these tests faster and cheaper so they can reach global populations. Within the next 5-10 years, blood-based biomarkers for dementia prediction should be as routine as cholesterol screening is today. For anyone concerned about dementia risk, knowing that these tools are emerging offers both clarity and hope—the clarity of early warning, and the hope of a long window for prevention.

Conclusion

Researchers have identified a four-protein combination—GFAP, NEFL, GDF15, and LTBP2—that predicts dementia with approximately 90% accuracy, sometimes 15 years before symptoms appear. This breakthrough offers an unprecedented opportunity to identify people at risk while they still have time to intervene through lifestyle changes, cardiovascular management, cognitive engagement, and other prevention strategies. The accuracy of these markers, combined with their ability to predict risk across all major dementia types, makes them among the most significant advances in dementia detection in recent decades.

If you’re concerned about dementia risk, especially with a family history of the disease, ask your doctor about access to these blood biomarkers. As testing becomes more widely available over the coming years, it will likely become a standard part of cognitive aging screening. Even if testing isn’t yet available in your area, the science is clear: the lifestyle changes that lower dementia risk—exercise, cognitive engagement, cardiovascular health, quality sleep, social connection—benefit your brain regardless of what your blood markers show. The window is open to prevent or delay dementia; these new blood tests help identify who should take action now.