What happened
- Researchers at Lund University (Sweden), with Nagoya University (Japan) and colleagues in Bologna (Italy), published a study on amyotrophic lateral sclerosis (ALS) in EBioMedicine. The release is dated October 20, 2026.
- They found that high levels of neurofilament light chain (NfL), a marker of nerve-cell damage, are linked to shorter survival, and that the link is stronger in patients with abnormal beta amyloid, a marker of Alzheimer’s.
- They used Japanese cohorts with 264 ALS patients and 164 controls with other neurological diseases. About 14% of ALS patients showed beta-amyloid changes.
- Machine-learning models combining NfL with other biomarkers improved ALS identification, including in borderline cases. Validation used 153 patients from Bologna.
Why it matters
- If 14% of ALS patients carry Alzheimer-type changes, the prognosis of one in seven changes without the diagnosis changing. Two diseases may overlap under a single label.
- For neurologists and hospitals in the region, which depend on low-cost tests, blood or cerebrospinal-fluid biomarkers are the realistic way to guide prognosis.
- The consequence the release does not develop: the study does not show that treating the amyloid component improves survival. Today it serves to estimate prognosis, not to change treatment.
The number
14%. Share of ALS patients in the Japanese cohorts who showed beta-amyloid changes associated with Alzheimer’s.
Context
ALS is a fast-progressing neurodegenerative disease with no cure. Biomarkers like NfL allow damage to be measured without waiting for new symptoms. The authors come from groups with cohorts in three countries, but report no Latin American patients. For clinical data analysis, see Illumina’s multiomic genomics platform.
What’s next
- No clinical trials announced from this finding.
- Pending: confirming the link in larger and more diverse cohorts.
- No announced dates. This note is not medical advice.
Bottom line
Models combining biomarkers now appear in neurology as they do in other areas of health. The known risk is cohort bias, which a study of pediatric triage with language models already showed.
Sources
Edited by Rodrigo Cornejo. How we select and verify: who writes these notes.




