Summary: Researchers have identified a specific cluster of blood proteins that change in people who later develop multiple sclerosis (MS). Using genetic-proteomic analysis and long-term biobank samples, these protein changes appear in blood up to a decade before clinical diagnosis, opening the possibility of routine, non-invasive screening to identify high-risk individuals before irreversible neurological damage occurs.
This discovery could shift MS care toward prevention—similar to how cholesterol screening informs cardiovascular risk—by enabling early identification and intervention to delay or prevent severe disability.
Key Facts
- Prevention-first approach: Because established brain damage is difficult to repair, identifying MS risk at a preclinical stage offers the best chance to preserve neurological function.
- Large-scale proteomic screen: Researchers screened more than 2,500 plasma proteins using Mendelian Randomization, a genetic statistical framework, and found 39 proteins causally linked to MS risk. Most of these proteins are involved in immune signaling pathways.
- Longitudinal biobank validation: The team used United Kingdom Biobank samples collected from 2006–2010 to examine prediagnostic blood. Among participants who later developed MS, protein alterations were detectable on average six years before diagnosis, and in some cases over ten years prior.
- Predictive protein set: Eight proteins measured in prediagnostic samples were significantly associated with future MS diagnosis, supporting their potential as early biomarkers.
- DKKL1 — a dual-purpose marker: One protein, DKKL1, showed a protective association: higher levels were linked to lower MS risk and to milder disease among those who did develop MS, making it useful for both risk stratification and prognosis.
- Next steps: The research team, led by Dr. Adil Harroud, plans larger-scale validation and aims to combine these proteomic markers with existing tools to develop a clinically useful blood screening test.
Source: McGill University
New findings show blood protein signatures that predict multiple sclerosis years before symptoms
A research team led by Dr. Adil Harroud at The Neuro (Montreal Neurological Institute-Hospital) of McGill University used genetic-proteomic integration to search for blood-based signals that indicate future MS. Proteins, the functional molecules circulating in plasma, were analyzed for causal links to MS risk. By combining genetic instruments with protein measurements, the investigators could infer which circulating proteins are likely to influence disease susceptibility.
From a discovery set of cis-acting protein quantitative trait loci (pQTL) for 2,545 plasma proteins (n = 80,824), the team evaluated genetically predicted protein levels against MS risk using Mendelian Randomization and colocalization methods. This analysis identified 39 proteins associated with MS susceptibility. Network analysis showed these proteins cluster in immune regulatory pathways, including B- and T-cell costimulation, cytokine signaling, and pathways linked to Epstein–Barr virus biology.
To test whether these proteins appear before clinical onset, the researchers analyzed prediagnostic plasma from the UK Biobank. Among 124 individuals who later developed MS and 52,515 controls, eight proteins measured in samples taken a median of 5.9 years before diagnosis were significantly associated with subsequent MS. One protein in particular, DKKL1, consistently exhibited protective associations across genetic risk, incidence, and disease severity.
These results suggest a genetically anchored proteomic signature that can identify preclinical MS biology and potentially guide early intervention. Integrating pQTLs also improved genetic fine-mapping at implicated loci and nominated additional putative risk regions for follow-up study.
“Early intervention can delay or prevent disabling symptoms of MS, but we currently lack practical tools to identify the right people in time,” says Dr. Harroud. “A blood-based screening strategy grounded in genetic evidence could provide clinicians with a way to act before irreversible damage occurs.”
The team intends to validate these candidate proteins in larger cohorts and to explore combining them with other diagnostic measures to build a clinically useful screening test.
The study was published in Annals of Neurology on May 22, 2026.
Funding: Fonds de Recherche du Québec Santé; Bougie Family Young Investigator Award; Brain Canada Foundation; Canada Brain Research Fund. UK Biobank application number 45551.
Key Questions Answered:
A: For conditions like MS, preventing or limiting early damage is far more effective than trying to repair injured brain tissue later. Detecting disease biology years before symptoms opens a window for treatments that can preserve function and reduce long-term disability.
A: Some proteins reflect biological processes that influence both disease initiation and progression. DKKL1 is an example: its levels correlated with lower risk of developing MS and, among those who developed the disease, with a milder clinical course. Such dual associations make certain proteins especially valuable for screening and prognosis.
A: The approach mirrors cholesterol testing: measure a blood biomarker long before symptoms occur to identify elevated risk, then intervene to reduce that risk. A validated MS blood test could enable the same preventive mindset in neurology.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The original journal article was reviewed in full.
- Additional context was added by editorial staff.
About this multiple sclerosis research news
Author: Shawn Hayward
Source: McGill University
Contact: Shawn Hayward – McGill University
Image credit: Neuroscience News
Original Research: “Genetic-Proteomic Integration Identifies Predictive Plasma Proteins for Multiple Sclerosis” by Yuan Ding MSc, Dylan Hamitouche, Simon Thebault MD, PhD, Patrick Kearns MBChB, MPH, Ahmed Abdelhak MD, PhD, Adil Harroud MD. Annals of Neurology. DOI: 10.1002/ana.78256. Open access.
Abstract (summarized)
Objective: MS has a prolonged preclinical phase. The goal was to identify circulating proteins that reflect early disease biology to improve risk prediction and guide interventions.
Methods: The study used cis-pQTLs for 2,545 plasma proteins from large proteogenomic datasets. Predicted protein levels were tested for association with MS risk using Mendelian Randomization and colocalization (14,802 cases, 26,703 controls). Candidates were validated in prediagnostic UK Biobank samples (124 incident cases, 52,515 controls; median 5.9 years prediagnosis) and assessed for links to disease severity (n = 12,584).
Results: Thirty-nine proteins were associated with MS risk, forming a network enriched for immune regulatory and B-cell pathways. Among proteins measured in prediagnostic samples, eight predicted time to diagnosis; DKKL1 showed protective associations across risk, incidence, and severity. Integrating pQTLs improved locus fine-mapping and nominated additional risk loci.
Interpretation: A genetic-proteomic framework can identify causal circulating proteins and support development of early predictive biomarkers with translational potential for preclinical MS detection and intervention.