The secret to preventing Alzheimer’s disease (AD) might not lie in the brain at all, but rather in the blood flowing through our veins.

Led by scientists at the National Institute on Aging, a new proteomic study has identified multiple circulating blood proteins that directly dictate genetic risk for the condition. By tracing these peripheral biological pathways, the research has identified fresh targets for non-invasive diagnostic tools and treatment strategies.

Redefining Alzheimer's risk through blood proteomics

Most AD research tends to focus on what happens inside the brain, but growing evidence shows that biological systems outside the central nervous system may also play a role. In particular, the body’s immune system seems to drive the disease in ways we are only beginning to understand.

Understanding how peripheral biology influences AD could lead to the development of accessible diagnostic tools. However, it is difficult to tell whether circulating blood proteins actually cause the disease or are just harmless byproducts.

“Although systemic processes, such as immune and metabolic dysfunction, have been reported as AD risk factors, the ability to distinguish mechanistically relevant peripheral biological pathways from those that are merely associative has been limited,” said the authors of the new study.

Previous proteomics studies in this field have relied on small sample sizes and lacked ancestral diversity, leaving African American populations underrepresented in genetic databases.

Earlier genome-wide association studies (GWAS) have successfully flagged innate immune activation and lipid processing in brain tissues, and more recently, expanded datasets mapped genetic variants that control blood protein levels.

These studies laid the groundwork to study the plasma proteome at scale. The new study aimed to link plasma protein levels directly to genetic risk to pinpoint the exact proteins that causally drive AD risk across diverse populations. To do so, the researchers executed a large-scale, multi-ethnic proteome-wide association study.

Mapping plasma proteomics to reveal new Alzheimer's targets

The team integrated plasma protein genetic models from 7,213 European American and 1,871 African American participants and blended this data with large AD genetic datasets. They used a Mendelian randomization framework, which treats inherited genetic variants as natural experiments, allowing the team to calculate whether specific blood protein levels directly dictate disease risk rather than appearing by chance. They validated their top hits using advanced statistical cross-checks, while cross-referencing findings with physical brain tissue and spinal fluid samples.

The analysis of the European American cohort uncovered 18 candidate genes with causal ties to AD risk; 12 of these genes overlapped with known targets such as APOE and TREM2, and the remaining 6 were entirely new risk genes: CD55, LILRB5, LILRB1, AZGP1, SIRPA, and SCARA5.

Two risk genes stood out: LILRB1 appeared to lower disease risk, while SIRPA increased it.

“All three causal inference methods aligned to mechanistically implicate LILRB1 and SIRPA, both of which are involved in regulation of the immune response,” said the team.

The African American cohort analysis flagged APOE as the primary driver and found 12 suggestive risk genes. As the African American genetic database was smaller, it lacked the statistical power to secure as many firm associations.

These causal plasma proteins are heavily involved in adaptive immunity, which uses specialized white blood cells, contradicting traditional brain studies that highlight innate immunity as a driver of AD.

The protein levels tracked closely with 8-year and 20-year dementia risks, aligning with established blood biomarkers such as pTau181 and NfL.

The future of Alzheimer's treatment and proteomics limitations

Discovering that peripheral adaptive immunity drives AD could change how clinicians approach treatment, potentially leading to therapies that act outside the brain.

Two of the newly discovered proteins, CD55 and SIRPA, are already being tested in phase 1 and 2 clinical trials for cancer, which could offer an immediate opportunity to repurpose existing drugs for dementia care.

The approach used also changes how scientists hunt for drug targets. “With its focus on circulating proteins as genetically determined AD risk factors, this analytic approach may be better positioned to identify putative causal AD proteins with genetic variants that may not meet genome-wide significance in a traditional GWAS,” said the authors.

The study did have limits: statistical power was unevenly distributed across the study, and as some proteins lacked strong genetic markers, large portions of the human proteome could not be evaluated. The study also only captures lifelong genetic tendencies rather than real-time shifts, which may not perfectly mirror the dynamic changes that happen day-to-day in every patient population.

Future studies are needed to determine whether AD risk comes from the total amount of circulating proteins or if genetic variants change the physical structure of the protein, altering how it binds to cells.

Reference: Walker KA, Blew C, Duggan MR, et al. Alzheimer’s disease proteome-wide association study implicates adaptive immunity and identifies risk genes LILRB1 and SIRPA. Sci Transl Med. 2026;18(856):eadx4852. doi: 10.1126/scitranslmed.adx4852