Blood protein shape changes may enable early blood test detection of Alzheimer's

Diagnosing Alzheimer's disease currently requires either a PET brain scan that costs between $3,000 and $6,000 and is rarely covered by insurance, or a lumbar puncture to extract cerebrospinal fluid, an invasive procedure that many patients and physicians prefer to avoid. A study published in PNAS in March 2025 describes a possible alternative: three blood proteins whose three-dimensional shapes change in a measurable pattern that tracks Alzheimer's progression, detectable years before serious cognitive symptoms appear.

The research team at the University of Warwick used a technique called ion mobility mass spectrometry to analyze the structural conformations of proteins in blood plasma from 146 participants, including people at different stages of Alzheimer's and healthy controls of similar age. The technique can distinguish between different three-dimensional shapes of the same protein molecule, not just its presence or absence, which is what makes it useful for tracking subtle early-stage disease changes.

Which proteins were identified and what changes they show

The three proteins identified are clusterin, fibrinogen gamma chain, and alpha-2-macroglobulin. All three are already known to have connections to neurological health through prior research. Clusterin, for example, is a molecular chaperone that helps prevent protein misfolding in cells and has been associated with amyloid-beta clearance in the brain in previous studies. The Warwick team found that clusterin's structural conformation in blood plasma shifted measurably between cognitively healthy participants, those with mild cognitive impairment, and those with confirmed Alzheimer's dementia.

Fibrinogen gamma chain is involved in blood clotting but also circulates in plasma in forms that interact with inflammatory signaling pathways. Its shape profile in the study was distinct in Alzheimer's patients compared to controls, with a more compact conformation that the researchers linked to altered interactions with complement proteins in the blood. Alpha-2-macroglobulin is a broad-spectrum protease inhibitor and showed a conformational change that was most pronounced in participants with early-stage disease, suggesting it may be particularly useful as a pre-symptomatic marker.

Why protein shape matters more than protein quantity

Most existing blood-based Alzheimer's biomarker research focuses on measuring the concentration of specific proteins, particularly phosphorylated tau (p-tau217) and amyloid-beta 42 to 40 ratios, both of which are already in clinical use in some memory clinics. Those concentration-based tests are useful, but they measure how much of a protein is present rather than what shape it is in. A protein can be present in normal quantities while operating abnormally if its structure has changed, which is the biological reality this new research is trying to capture.

Protein misfolding is, in fact, central to Alzheimer's pathology. The amyloid plaques that accumulate in Alzheimer's brains are made of beta-amyloid protein that has folded into a sheet-like conformation rather than its normal shape, triggering aggregation. Tau neurofibrillary tangles form through a similar structural change in the tau protein inside neurons. The Warwick team's hypothesis is that the conformational changes they detected in blood proteins may reflect systemic changes in protein handling that happen throughout the body during Alzheimer's progression, not just inside the brain.

How accurate the blood protein test was in this study

Using the conformational data from all three proteins together as a composite marker, the Warwick team's model distinguished Alzheimer's patients from healthy controls with an area under the receiver operating curve of 0.89, which indicates strong but not definitive discrimination. The model performed less well at distinguishing early-stage Alzheimer's from mild cognitive impairment caused by other conditions, achieving an AUC of 0.74 for that comparison. An AUC of 1.0 would represent perfect classification, and clinical diagnostic tests generally aim for above 0.85.

The early-stage discrimination problem is the one that matters most clinically, because patients and physicians want to know whether mild memory symptoms are early Alzheimer's or something else before significant neurodegeneration has occurred. The composite protein test's 0.74 performance at that stage is better than chance and better than clinical symptom assessment alone, but not sufficient on its own for a reliable diagnostic. The Warwick team suggested combining the conformational markers with existing concentration-based tests like p-tau217 could improve accuracy at the early stage, though that combination has not been formally tested yet.

How this differs from existing blood biomarker tests already in use

Phosphorylated tau p-tau217 blood tests, developed by groups including Lilly and ALZpath, have been validated in large cohort studies and received CE marking in Europe in 2023 for clinical use in memory clinics. The US FDA granted Breakthrough Device designation for p-tau217-based tests in 2024. These existing tests work by measuring concentration ratios of specific tau fragments in blood plasma. They are sensitive for detecting established Alzheimer's pathology but have lower accuracy for pre-symptomatic detection, particularly in individuals who have brain amyloid accumulation but no detectable tau pathology yet.

The protein conformation approach is earlier in its validation timeline and uses more expensive, specialized equipment that is not widely available in clinical laboratories. Ion mobility mass spectrometry instruments cost between $200,000 and $500,000 and require trained operators, which limits near-term clinical deployment. The Warwick team acknowledged in their paper that translating the technique to a scalable, point-of-care format would require significant engineering work. Their immediate goal is validation in a larger independent cohort, which they plan to conduct in collaboration with the UK Biobank, targeting a sample of 500 participants across disease stages.

What early detection would actually change for patients

The clinical value of early Alzheimer's detection has increased substantially since lecanemab, sold as Leqembi, received full FDA approval in July 2023. Lecanemab slows cognitive decline by approximately 27 percent over 18 months in patients with early-stage Alzheimer's who have confirmed amyloid pathology, but it only works if administered before significant neurodegeneration occurs. The treatment window is narrow, and identifying candidates early enough to benefit from anti-amyloid therapy requires detection well before patients develop obvious dementia symptoms.

A routine blood test that flags early protein conformational changes could enable broader screening in primary care settings, routing high-risk individuals to confirmatory testing with PET scans or cerebrospinal fluid analysis before proceeding to treatment decisions. That screening model would reduce the diagnostic burden on expensive neuroimaging equipment and allow lecanemab and subsequent anti-amyloid drugs to reach more patients within their effective treatment window. The Warwick team's validation study in UK Biobank participants is expected to complete enrollment by the end of 2025, with results planned for submission in mid-2026.