Misfolded proteins found in motor neurons are the cause of muscle weakness and paralysis. People with Amyotrophic Lateral Sclerosis (ALS) generally live two to four years after the onset of symptoms, creating an immediate urgency for an accurate diagnosis. A critical need exists for tests that detect early signs resulting in faster enrollment and tests for new medicines. Currently, doctors rely on neurological tests and clinical symptoms.
A new study led by Bryan Traynor Ph.D. and Sonja Scholz Ph.D. both at the National Institute of Health, finds that blood samples contain detectable signs of ALS. As published in Nature Medicine August 19, 2025, the study suggests that a protein-based blood test could diagnose detectable signs of early ALS disease well before the symptoms appear and to distinguish it from other conditions.
Proteomics
One of the systems used by the team is called proteomics. It can analyze over three thousand proteins from blood samples of ALS patients. The samples were compared to those of healthy people and individuals with other neurological conditions.
According to the study, there are thirty-three proteins associated with neurons, skeletal muscle and energy metabolism. Two of these proteins had previously been linked with ALS.
Machine Learning
Researchers are using machine learning to identify certain clinical and molecular subtypes of the disease as well as to identify survival time, disease progression, and to develop non-invasive methods for monitoring severity through voice and movement data.
98% Accuracy
Machine learning was also used to create a predictive model to identify ALS. The model included twenty features most predictive of ALS. Working with these features, the team created an ALS risk score for each sample. The new model could diagnose ALS with almost 98% accuracy.
A large number of ALS blood samples used in the study had been assembled many years prior to the onset of the patients’ symptoms with the increase in risk scores correlating to the patients’ symptoms.
Most importantly, changes that were detectable began approximately ten years prior to the onset of symptoms. The scores of presymptomatic individuals increased as symptoms became evident. Therefore, the findings indicate that hidden changes surface earlier than suspected thus offering new insight into ALS.
Dr. Traynor commented that neurologists continue to diagnose ALS using a clinical analysis similar to the methods used 150 years ago. Therefore, being able to detect ALS in a blood test many years before the symptoms appear presents substantial improvement both for research and clinical care.
