Disease-modifying therapies in managing disability worsening in adult relapsing multiple sclerosis: a population-based longitudinal study using the MSBase registry

Background Disease-modifying therapies (DMTs) reduce inflammatory activity in relapsing multiple sclerosis (MS), yet their long-term effects on disability worsening, and how these vary by age at treatment initiation in routine clinical practice, remain uncertain. We examined age-dependent associations of DMTs with disability progression across clinically meaningful Expanded Disability Status Scale (EDSS) states, including secondary progressive MS (SPMS), while accounting for latent heterogeneity in progression risk. Methods We analysed longitudinal data from the international MSBase registry, including adults with relapse-onset MS (symptom onset ≥ 18 years), ≥ 12 months of follow-up, and ≥ 2 EDSS scores recorded remote from relapse. Disability progression was defined as transitions between five EDSS-based states: nil/minimal (0–1.5), mild (2.0–2.5), moderate (3.0–3.5), gait disability (≥ 4.0), and SPMS. A continuous-time random-effects multistate model jointly estimated worsening and improvement transitions. Associations of low-efficacy and high-efficacy DMTs versus no treatment were assessed with age-by-treatment interactions. Age-stratified treatment effects were summarised as 5-year percentage inhibition of disability progression (%IDP). Random-effects captured unobserved between-person heterogeneity and identified latent progression subgroups. Results Among 15,425 participants (70.9% female), median follow-up was 9.4 years (IQR 4.7–15.6). Overall, 87.9% received DMTs, with 48.7% initiating high-efficacy and 45.9% initiating low-efficacy therapy. Both DMT classes were associated with more favourable disability trajectories across most transitions, including increased likelihood of improvement. Treatment effects were strongly age-dependent, with greatest benefit in younger adults (18–30 years), attenuation through midlife (30–50 years), and weaker or inconclusive effects after age 50, particularly for transitions to SPMS. In 5-year %IDP analyses, inhibitory effects declined monotonically with age. High-efficacy therapy conferred greater inhibition of disability progression than low-efficacy therapy, with evidence of benefit persisting into older age. Two latent subgroups were identified: a rapid-worsening group (76.1%) and a stable/improving group (23.9%), differing markedly in progression hazards and recovery potential. Conclusions In this large real-world cohort, DMTs were associated with sustained but age-dependent reductions in disability progression. Substantial heterogeneity in progression risk supports personalised treatment strategies, improved risk stratification, and enriched clinical trial design.