# Digital monitoring of disease activity in relapsing–remitting multiple sclerosis

## TL;DR

Digital tools like smartphone tests can detect disease activity in multiple sclerosis patients, sometimes even before MRI scans show changes.

## Contribution

This study shows that digital biomarkers can detect inflammation in multiple sclerosis earlier than traditional methods.

## Key findings

- The smartphone-based symbol digit test (sSDMT) was strongly linked to MRI-detected inflammation, especially when used 6 weeks before the scan.
- The 2-minute walk test (s2MWT) also showed a negative association with inflammation, influenced equally by individual and group variations.
- Traditional clinical measures like T25FW, EDSS, cSDMT, and NHPT were also negatively associated with MRI-detected inflammation.

## Abstract

Digital monitoring shows promise for detecting disease activity in people with relapsing–remitting multiple sclerosis (PwRRMS). Here, we study associations between digital biomarkers for cognition and walking ability and radiological disease activity.

In a prospective, 1-year cohort study, PwRRMS performed the smartphone-based symbol digit modalities test (sSDMT) and 2-min walk test (s2MWT) on weekly basis. MRIs and the clinical SDMT (cSDMT), expanded disability status scale (EDSS), timed 25-foot walk (T25FW) and nine-hole peg test (NHPT) were collected at baseline and every 3 months. Associations were determined using logistic generalized estimating equations. For the digital measures, associations were also analyzed using a hybrid model and were repeated with values from 6 weeks before and after MRI.

We included 57 PwRRMS. The sSDMT was negatively associated with contrast-enhancing lesions (CELs) (ORoverall 1.80, 95% CI 1.12–2.91), predominantly caused by variation within individuals (ORwithin-subjects 4.37, 2.05–9.33), with a similar relation using sSDMT values 6 weeks prior to MRI (ORoverall: 1.92, 0.947–3.90, ORwithin-subjects: 13.7, 1.74–107). The negative association between s2MWT and CELs (ORoverall 1.20, 1.04–1.38) was caused equally by variation within and between individuals. All clinical measures were negatively associated with CELs: T25FW (ORoverall 2.23, 1.50–3.32), EDSS (ORoverall 1.49, 0.932–2.39), cSDMT (ORoverall 1.20, 1.02–1.42) and NHPT (ORoverall 1.15, 1.04–1.27).

Digital biomarkers show to be capable of measuring changes in individuals when inflammation is detectable on MRI, with the sSDMT additionally capturing changes 6 weeks prior to the MRI, suggesting that early identification of inflammation using these biomarkers may be possible.

The online version contains supplementary material available at 10.1007/s00415-026-13685-5.