Protein-based blood test detects early signs of ALS

September 16, 2025

Protein-based blood test detects early signs of ALS

At a Glance

Researchers have found proteins in blood that accurately detect amyotrophic lateral sclerosis (ALS) long before symptoms emerge.
The discovery may offer the first definitive ALS diagnostic test and a potentially promising way to track disease progression in clinical trials.

Doctor holding blood sample in test tube.

Results of an NIH-supported study suggest it is possible to detect ALS via a simple blood test and distinguish it from other clinically relevant conditions, even before symptoms arise.

Billion Photos / Shutterstock

In amyotrophic lateral sclerosis, or ALS, misfolded proteins in motor neurons cause progressive muscle weakness and paralysis. People with ALS often die within 2 to 4 years after symptoms start. Many doctors struggle to confirm the diagnosis before symptoms get severe. There’s a critical need for a test that could detect early signs. Such a test would allow earlier treatments and speedier enrollment in clinical trials for testing new medicines.

To diagnose ALS, doctors now rely on clinical symptoms and neurological tests. A new study led by Dr. Bryan J. Traynor at NIH’s National Institute on Aging and Dr. Sonja W. Scholz of the National Institute of Neurological Disorders and Stroke shows that detectable signs of the disease can be found in blood samples. This suggests that a protein-based blood test could diagnose the disease long before symptoms appear. Their findings appeared in Nature Medicine on August 19, 2025.

The researchers used an approach called proteomics to analyze more than 3,000 proteins found in blood samples from ALS patients. They compared the protein data to samples from healthy people and those with other neurological conditions.

The study team found 33 proteins that distinguish ALS from other neurological conditions. Only two of those proteins had been linked to ALS before. Further study indicated that the 33 proteins are involved in the normal functioning of skeletal muscle, neurons, and energy metabolism.

Next, the researchers used machine learning to develop a predictive model for identifying ALS. The resulting model included 20 features that were most predictive of ALS. From these features, they developed an ALS risk score that could be calculated for each sample. This model could diagnose ALS with more than 98% accuracy.

Many of the blood samples from ALS patients were collected years before their symptoms began. Risk scores for these samples were correlated with the time to symptom onset. Scores increased as people with ALS got closer to showing symptoms.

The findings suggest it’s possible to detect ALS in a simple blood test and distinguish it from other clinically relevant conditions. The results also hint that detectable changes related to ALS begin up to 10 years before symptoms. This may enable pre-symptomatic patients to enroll in clinical trials. The findings also show that hidden changes underlying ALS—most unknown before—arise much earlier than had been suspected. Besides their clinical implications, these discoveries offer new insight into this devastating condition.

“Today, neurologists still diagnose ALS by a clinical examination much the same way it was done 150 years ago,” Traynor says. “Detecting the disease in a blood test years before symptoms show up substantially moves the ball down the field for both clinical care and for research efforts to develop effective treatments.”

—by Kendall K. Morgan, Ph.D.

References

. Chia R, Moaddel R, Kwan JY, Rasheed M, Ruffo P, Landeck N, Reho P, Vasta R, Calvo A, Moglia C, Canosa A, Manera U, Snyder A, Saez-Atienzar S, Grassano M, Brunetti M, Casale F, Ray A, Arvind K, Comertpay B, Zhu M, Gibbs JR; American Genome Center; Alba C, Dawson TM, Rosenthal LS, Hall AJ, Pantelyat AY, Narendra DP, Ehrlich DJ, Walker KA, Kosa P, Bielekova B, Egan JM, Candia J, Tanaka T, Ferrucci L, Dalgard CL, Scholz SW, Chiò A, Traynor BJ. Nat Med. 2025 Aug 19. doi: 10.1038/s41591-025-03890-6. Epub ahead of print. PMID: 40830661.

Funding

NIH’s National Institute on Aging (NIA), National Institute of Neurological Disorders and Stroke (NINDS), National Institute of Allergy and Infectious Diseases (NIAID); Merck Sharp & Dohme Corporation; Centers for Disease Control and Prevention; Muscular Dystrophy Association; Microsoft Research; Packard Center for ALS Research at Johns Hopkins; ALS Association; Cerevel Therapeutics (now part of AbbVie, Inc.); Italian Ministry of Health; Progetti di Rilevante Interesse Nazionale program of the Ministry of Education; Horizon 2020 program; Horizon Europe program.

��ѿ��ý