Serum GFAP and NfL augment a metabolomics-driven strategy for long-term prediction of multiple sclerosis progression

Why this research matters

For many people living with multiple sclerosis (MS), the most pressing question is not just what the disease is doing right now, but how quickly it will worsen in the future. Today, doctors rely mainly on symptoms and brain scans to judge progression, which often reveal changes only after damage has already occurred. This study explores whether a simple blood test, reading the body’s chemical “fingerprints,” can forecast who is most likely to experience a faster decline, long before it becomes obvious in daily life.

Looking for warning signs in the blood

MS is traditionally split into two broad forms: relapsing-remitting MS, marked by flare-ups and partial recovery, and secondary progressive MS, in which disability steadily worsens. Yet this neat divide does not reflect reality for many patients. Some people quietly worsen without clear relapses, while others remain relatively stable for years. The researchers set out to move beyond labels and instead search for biological signs in blood that track the true underlying disease process, whether or not a person has entered the “progressive” category. To do this, they focused on two types of signals: small molecules involved in energy and fat metabolism, and two protein fragments that leak from damaged nerve and support cells in the brain.

Reading the body’s chemical fingerprints

The team analysed stored blood samples from a carefully followed Swiss MS cohort, along with an independent group from Oxford, UK. Using a technique called nuclear magnetic resonance (NMR), they measured many small molecules at once, capturing a broad “metabolomic” profile. They paid special attention to fats carried in the blood (lipoproteins), simple sugars like glucose, and several amino acids that help fuel brain cells. At the same time, they measured two proteins already known to reflect damage in the nervous system: glial fibrillary acidic protein (GFAP), released when support cells in the brain are injured, and neurofilament light chain (NfL), shed when nerve fibers are harmed. By combining these data with long-term clinical records, they asked which patterns in the first available blood sample could predict who would later worsen.

Spotting stage and future risk from one sample

The metabolomic patterns alone were able to distinguish people with relapsing-remitting MS from those with secondary progressive MS, confirming earlier work that the blood chemistry of these groups differs in a consistent way. More importantly, the baseline metabolic profile also carried information about the future: individuals who later transitioned to secondary progressive disease tended to show lower levels of certain blood fats and distinct shifts in key amino acids and glucose. When these selected metabolites were analysed together in a multivariable model, they correctly classified most patients as future “progressors” or “non-progressors,” regardless of their official MS stage at the time. In other words, the blood chemistry captured a sense of where the disease was heading, not just where it had been.

Stronger predictions by combining signals

While the metabolic signatures were informative on their own, the predictions became even more powerful when they were integrated with GFAP and NfL. Higher baseline GFAP levels, in particular, were strongly linked to later disability worsening that was not explained by obvious relapses. When the researchers built models that blended five key metabolites with GFAP or NfL, the ability to separate future progressors from non-progressors improved markedly, correctly identifying the vast majority of patients in the Swiss cohort. Similar metabolic shifts were observed in the independent Oxford cohort, supporting the idea that the findings are not limited to a single clinic or patient selection.

What this could mean for people with MS

Taken together, this work suggests that a carefully designed blood test, combining broad metabolic readouts with a small set of nerve-damage proteins, could offer a practical tool to gauge an individual’s risk of MS progression well before major disability occurs. Rather than waiting for walking ability or other functions to visibly decline, clinicians could use this information to refine monitoring, adjust treatment intensity, or enrol patients in trials aimed at slowing progression. Although larger studies are still needed, the study points toward a future in which managing MS becomes less about reacting to setbacks and more about anticipating and preventing them.

Citation: Kacerova, T., Willemse, E., Oechtering, J. et al. Serum GFAP and NfL augment a metabolomics-driven strategy for long-term prediction of multiple sclerosis progression. Commun Med 6, 182 (2026). https://doi.org/10.1038/s43856-026-01453-5

Keywords: multiple sclerosis progression, blood biomarkers, metabolomics, GFAP and neurofilament, precision neurology