Accurate diagnosis of dementia has fundamentally changed in the past few years, moving from guesswork based on cognitive tests alone to a blood-based science that can identify the specific proteins destroying a patient’s brain. The primary biomarkers now needed are phosphorylated tau variants—especially p-tau217 and p-tau181—which can detect Alzheimer’s disease pathology with 88% accuracy from a simple blood draw. Beyond Alzheimer’s, alpha-synuclein in the cerebrospinal fluid identifies Lewy body dementia, TDP-43 markers pinpoint frontotemporal dementia, and emerging plasma biomarkers like neurofilament light chain and GFAP reveal whether neurodegeneration is occurring at all.
A 62-year-old man experiencing memory lapses and movement problems might have received a diagnosis of “possible dementia” five years ago and never known whether he had Alzheimer’s, Lewy bodies, or vascular disease—today, a blood test can tell him definitively within days. These biomarkers work because they detect the actual pathological proteins accumulating in the brain before symptoms become severe. This shift from clinical guessing to biological certainty has opened new doors: patients can now be identified at much earlier stages, enrolled in disease-modifying treatments, and monitored for decline with measurable biological markers rather than subjective cognitive tests. The 2024 diagnostic framework from the National Institute on Aging has officially embraced biomarkers as central to diagnosis, meaning the tools are no longer relegated to research centers but are moving into community clinics and primary-care offices.
Table of Contents
- What Makes Phosphorylated Tau the Gold Standard for Alzheimer’s?
- How Alpha-Synuclein Biomarkers Identify Lewy Body Dementia
- Distinguishing Frontotemporal Dementia Subtypes With Extracellular Vesicle Markers
- From Blood Tests to Clinic: How Biomarkers Enter Clinical Practice
- The Real-World Risks and Adverse Events of Biomarker-Guided Treatment
- Vascular Dementia and Emerging White Matter Biomarkers
- Non-Specific Neurodegeneration Markers That Work Across All Dementia Types
What Makes Phosphorylated Tau the Gold Standard for Alzheimer’s?
Phosphorylated tau—particularly p-tau217—has emerged as the most reliable biomarker for Alzheimer’s disease, with a sensitivity of 88.1% and specificity of 88.7%, meaning it correctly identifies Alzheimer’s pathology in roughly 9 out of 10 patients who have it and correctly rules it out in 9 out of 10 who don’t. The key advantage of p-tau217 over older markers like p-tau181 is its magnitude of change: in Alzheimer’s patients, p-tau217 rises 3.3 to 3.9 times higher than in other dementia types, creating a much sharper biological signal. Plasma p-tau217 performs almost identically to cerebrospinal fluid (CSF) p-tau217—achieving 82% sensitivity for amyloid detection and 83% for tau pathology in blood—which means patients no longer need a lumbar puncture to get an accurate diagnosis.
The Roche Elecsys platform received authorization in 2025 specifically for p-tau181 as a primary-care test to rule out amyloid pathology, marking an important milestone: automated equipment now exists to run these tests in regular laboratory settings rather than only in specialized centers. For practitioners, the combination of amyloid-beta 42 with p-tau181 (the Aβ42/p-tau181 ratio) offers 90% sensitivity for detecting Alzheimer’s with an AUC of 0.93—meaning the test’s ability to discriminate between disease and health is nearly perfect. The practical implication is that a family doctor seeing a patient with early cognitive complaints can order a blood test, get results in days, and know with high confidence whether Alzheimer’s pathology is present, rather than scheduling multiple appointments and repeat cognitive testing over months.
How Alpha-Synuclein Biomarkers Identify Lewy Body Dementia
Lewy body dementia accounts for about 5% of all dementia cases but is notoriously underdiagnosed because its symptoms—hallucinations, movement problems, and fluctuating cognition—overlap with Parkinson’s disease, Alzheimer’s disease, and depression. The cerebrospinal fluid alpha-synuclein seed amplification assay (αSyn-SAA) changes this picture by detecting actual pathological alpha-synuclein accumulation: in patients with clinically diagnosed Lewy body dementia, the assay is positive in 68% of cases, with excellent sensitivity and specificity compared to Alzheimer’s disease. Newer blood-based alpha-synuclein biomarkers—extracted from extracellular vesicles and measured via seeding assays—are now becoming available and promise to offer the same accuracy without requiring a cerebrospinal fluid sample.
A critical clinical reality that alpha-synuclein testing has exposed is that roughly 70% of Lewy body dementia patients also harbor Alzheimer’s pathology—amyloid and tau accumulation alongside the alpha-synuclein. This means a patient diagnosed with Lewy body dementia alone may actually be suffering from two simultaneous neurodegenerative diseases, and a comprehensive biomarker panel using both plasma p-tau and plasma alpha-synuclein reveals this dual pathology. The limitation here is important: neither plasma nor CSF biomarkers can tell you the exact distribution of pathology in the brain—only that it’s present—so imaging may still be needed to understand which brain regions are most affected and why a patient’s particular symptoms emerged.
Distinguishing Frontotemporal Dementia Subtypes With Extracellular Vesicle Markers
Frontotemporal dementia comes in multiple forms—behavioral variant FTD, progressive supranuclear palsy (PSP), and primary progressive aphasia being the most common—and misdiagnosis is frequent because behavioral changes and language problems can also indicate depression, psychiatric illness, or even Alzheimer’s disease. Extracellular vesicle tau and TDP-43 biomarkers now provide biological signals that distinguish these entities: the ratio of 3R tau to 4R tau in extracellular vesicles differs between progressive supranuclear palsy and behavioral variant FTD, and this distinction was validated in a cohort of 704 patients including 37 with confirmed genetic mutations and 31 with neuropathological verification—meaning it’s been tested against the gold standard of brain autopsy. Elevated extracellular vesicle TDP-43 specifically marks ALS and FTD-TDP variants, providing a way to identify patients whose primary pathology is TDP-43 protein rather than tau or amyloid. A real-world example: a 55-year-old woman presents with increasing apathy, poor judgment, and difficulty starting tasks—symptoms that could be depression or early Alzheimer’s disease.
Standard cognitive testing is nonspecific. EV tau and TDP-43 biomarkers reveal TDP-43 elevation without significant tau pathology, pointing the clinician toward behavioral variant FTD rather than Alzheimer’s disease. This diagnosis shift is critical because behavioral FTD often benefits from behavioral interventions and specific medications (SSRIs, rather than antiamyloid treatments), whereas standard Alzheimer’s therapies are unlikely to help and may carry unnecessary risk. The limitation to acknowledge: extracellular vesicle biomarkers are still primarily research tools and not yet widely available in clinical practice outside specialized centers, though that is changing rapidly in 2025-2026.
From Blood Tests to Clinic: How Biomarkers Enter Clinical Practice
The conversion of biomarker research into clinical reality has accelerated due to fully automated platforms like Lumipulse, which can run high-throughput phosphorylated tau assays in standard laboratory equipment. This matters because a community hospital or outpatient clinic in a rural area no longer needs a specialized neurochemistry lab; they can run the test using equipment they already own, with technician training rather than requiring a PhD-level researcher. The Alzheimer’s Association released its first Clinical Practice Guideline for Blood-Based Biomarkers at AAIC 2025, recommending specialist use of blood biomarker testing in people with cognitive impairment, which formalized what many centers were already doing informally.
The practical choice between plasma and CSF biomarkers now favors plasma in most settings: plasma p-tau217 offers comparable accuracy to CSF, can be drawn in any clinic, and carries no risk of infection or neurological complication (unlike lumbar puncture). A patient can have blood drawn at their primary-care office and results returned in 1-2 weeks, rather than being referred to a neurology center for a lumbar puncture and waiting 3-4 weeks for results. However, CSF testing may still be preferred when multiple biomarkers are needed simultaneously (full phosphorylated tau panel, neurofilament, GFAP) or in research settings where the absolute highest sensitivity is needed, because CSF concentrations are higher and the biological signal is stronger. For most diagnostic situations in clinical practice, plasma has become the first-line approach.
The Real-World Risks and Adverse Events of Biomarker-Guided Treatment
Identifying amyloid or tau pathology via biomarkers is only the first step; the second is treating it. Two monoclonal antibodies that remove amyloid from the brain—lecanemab (Leqembi, FDA approved 2023, EMA authorized November 2024) and donanemab (FDA approved July 2024)—have shown modest benefits in early symptomatic patients but come with significant risks that biomarker testing has made visible. In clinical trials and real-world use, lecanemab caused amyloid-related imaging abnormalities (ARIA) in 10% of patients—fluid accumulation (ARIA-E) or microhemorrhages (ARIA-H) in the brain visible on MRI—and infusion reactions in 22.5% of patients, requiring medical supervision during each infusion. The slowed decline achieved by lecanemab is real but modest: 27% slowing of cognitive decline over 18 months in Phase III trials (CLARITY AD), which translates to roughly a 2-3 month delay in progression rather than a reversal or halt of disease.
Donanemab showed similarly modest benefits in Phase III (TRAILBLAZER-ALZ 2). A critical limitation is that biomarkers identify pathology in cognitively normal individuals too—amyloid and tau can accumulate in the brain a decade or more before symptoms appear—but the treatments only work in people already experiencing cognitive impairment. This creates a gap: biomarker testing can identify at-risk individuals, but current anti-amyloid drugs don’t prevent symptom onset, only slow progression after cognition is already affected. Additionally, lecanemab and donanemab are limited in their use; lecanemab in Europe is restricted to people who either don’t carry the APOE4 genetic risk factor or are APOE4 heterozygotes (one copy), excluding the highest-risk patients from treatment.
Vascular Dementia and Emerging White Matter Biomarkers
Vascular dementia—caused by reduced blood flow to brain tissue due to small vessel disease—accounts for 15-20% of dementia cases but has historically lacked specific biomarkers, relying instead on imaging and clinical judgment. Recent research, crystallized at a 2025 workshop on vascular dementia, has identified both structural MRI markers and novel plasma biomarkers. White matter hyperintensities (WMHs) on MRI are the gold standard imaging sign of cerebral small vessel disease and are associated with vascular dementia risk, but they’re present in many cognitively normal older adults and don’t predict individual progression reliably.
Diffusion MRI (dMRI) can now quantify early tissue changes before WMHs become visible, offering a way to detect small vessel disease at earlier stages. New plasma biomarkers—oligodendrocyte myelin glycoprotein (OMG) and neuronal pentraxin receptor (NPTXR)—have been shown to correlate with white matter hyperintensity volume in the brain, suggesting they may reflect the underlying small vessel pathology. These markers are still primarily research tools but offer promise for identifying vascular dementia in primary-care settings without requiring MRI.
Non-Specific Neurodegeneration Markers That Work Across All Dementia Types
Neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP) are biomarkers that don’t identify a specific disease but instead reflect the degree of neurodegeneration and neuroinflammation occurring in the brain. Elevated NfL predicts clinical decline in Alzheimer’s disease and distinguishes true neurodegenerative conditions from mimics like depression or normal cognitive aging, meaning a high NfL level in a patient with cognitive complaints suggests something is actually wrong in the brain, even if other biomarkers haven’t clarified what.
GFAP reflects astrogliosis and neuroinflammation in early Alzheimer’s disease and can be elevated before amyloid or tau biomarkers reach diagnostic thresholds, potentially identifying at-risk individuals even earlier than current protocols allow. A combined plasma biomarker panel using p-tau217, p-tau231, p-tau181, NfL, GFAP, and amyloid-beta variants now provides clinicians with early detection of pathology, risk assessment for progression, differential diagnosis capability across dementia types, and longitudinal monitoring tools to track whether treatments are actually slowing decline. Plasma p-tau217 for distinguishing Alzheimer’s from non-Alzheimer’s dementias achieves 94% sensitivity and 91% specificity (AUC 0.96), matching the diagnostic accuracy of cerebrospinal fluid but with far less invasiveness.
- —
