Hidden ‘cognition-first’ subtype of multiple sclerosis identified

Online cognitive tests have revealed a distinct subtype of multiple sclerosis (MS), offering a faster way to flag those who have been undetected.

A study published in Nature Communications by Imperial College London in partnership with King’s College London, Swansea University and the UK MS Register, analysed data from people with MS who completed a self-administered cognitive test called Cognitron. Using these data and a clinic-based comparator, the team identified a ‘cognition-first’ MS subtype.

People with MS can experience cognitive problems that aren’t always obvious in a clinic visit. This short, self-administered online test which could be completed at home on people’s own devices revealed this subtype of MS which has significant cognitive impairment and minimal motor disability.

Professor Adam Hampshire, Professor of Cognitive and Computational Neuroscience at King’s College London and Visiting Professor at Imperial College London said, “Classic approaches to cognitive assessment are difficult to deploy in this sort of context due to the cost of supervised testing. By using a state of the art fully automated assessment approach we were able to test thousands of people with MS quickly, sensitively and at little cost. These sorts of numbers make it possible to characterise symptoms and impairments in MS in more detail than has been previously achieved.”

Measuring cognitive impairments

The programme involved members of the UK MS Register taking part in a three-stage observational study.

• Stage 1 (3,066 members) established feasibility using 22 online tasks from home and identified those most sensitive to MS-related changes.
• Stage 2 (2,696 members) built on this and replicated findings in an independent sample and validated a 12-task Cognitron-MS battery selected.
• Stage 3 finally evaluated the test against standard in person assessments to ensure the platform’s comparative validity and clinical utility.

Cognitive impairment affects 40–70% of people with MS, commonly impacting information-processing speed, working memory and sustained attention. Despite this, cognition is not routinely assessed in clinical practice and is absent from current clinical phenotype definitions. Brief at-home checks provide a practical way to bring cognition into routine MS reviews and prioritise full neuropsychological assessments for those who need it most.

Data-driven clustering of cognitive performance and motor patient-reported outcomes revealed a prevalent ‘cognition-first’ subtype with marked cognitive deficits and little visible motor disability. The authors report that this disability is often under-recognised in routine pathways that prioritise motor signs.

Dr Annalaura Lerede, first author, said: “Our findings represent a significant step forward in our understanding of MS. We found examples of people with this unrecognised subtype across all ages, no matter how long they’d had the disease. This reflects the risk of underestimating disability severity in this group throughout patients’ disease courses.

“Assessing cognitive impairment is not a standard part of clinical practice. This research provides a viable and easily implemented means of testing cognitive function at home and could help vast numbers of people better understand their condition and direct in person assessments to the people that need them the most.”

Dr Annalaura Lerede conducted this research as part of her UKRI Centre for Doctoral Training in AI for Healthcare PhD at Imperial College London and continued it during an IoPPN-funded postdoctoral position at King’s College London, in partnership with the UK MS Register.

The research was also conducted with clinicians at Imperial College Healthcare NHS Trust and therefore is close to care. Working within NHS services enables the team to design and test pathway changes – where, when, and for whom remote cognitive checks on personal devices add value – within live clinical workflows under appropriate governance. This embedded clinical–academic–research partnership enables pragmatic evaluations conducted at pace, supporting effective and timely adoption into routine care pathways.

Paper: Lerede A, Moura A, Giunchiglia V, et al. Large-scale online assessment uncovers a distinct Multiple Sclerosis subtype with selective cognitive impairment. Nature Communications (Published 3 September 2025). DOI: 10.1038/s41467-025-62156-4.