Remote Exercise for MS: Feasible, Cognitive Gains Unproven

Remote exercise programs can help people with multiple sclerosis (MS) stay active at home. What they cannot yet promise is faster thinking. The current evidence supports the delivery system: people with MS can use internet-based coaching, home exercise, and digital activity programs safely and acceptably. Evidence that these programs improve cognitive processing speed—the ability to take in and use information quickly—has not caught up.

That distinction matters. MS affects the brain's white matter, and slower processing speed can disrupt work, driving, conversation, and daily planning long before a person looks physically disabled. Exercise has well-established value for physical function and general health. But treating a cognitive symptom requires evidence on cognition itself, not an assumption that stronger legs automatically mean faster mental processing.

A remote exercise platform is no longer hypothetical

A 2025 scoping review by Adnan et al. mapped 54 published home e-based exercise studies in MS, conducted from 2008 through 2023 and enrolling 2,359 people. Programs included live videoconferencing, apps, e-coaching, and hybrid formats. They used aerobic exercise, resistance work, yoga, or combined training.

This is enough research to settle one practical question: remote delivery is possible. A separate 12-week multicenter randomized trial by Hartung et al. tested internet-based exercise and physical-activity promotion in people with MS. Participants received tailored home prescriptions and video-chat coaching. The program was feasible, safe, and acceptable, and it increased physical activity. Cognitive processing speed was not an outcome, so the study establishes the route to an intervention, not a cognitive effect.

That is still useful. Travel, fatigue, heat sensitivity, mobility limits, and uneven access to rehabilitation can all make conventional exercise programs hard to sustain in MS. Remote options can reduce those barriers. A working delivery system gives researchers a realistic way to run larger cognitive-endpoint trials across multiple sites.

The cognitive claim runs into a direct null result

The most relevant caution comes from MS itself. In secondary analyses of a randomized controlled trial, Riemenschneider et al. found that exercise therapy in early MS improved physical function but did not significantly improve cognitive outcomes. This does not mean exercise lacks value for people with MS. It means physical improvement cannot substitute for a measured cognitive benefit.

The result also limits a familiar storyline: exercise increases fitness, fitness supports the brain, therefore exercise should improve cognition. Biology does not always honor tidy arrows. Cognitive outcomes depend on the population studied, baseline impairment, the exercise dose, the comparison group, the test used, and whether the intervention engages cognition directly.

Evidence outside MS does not rescue that claim. A one-year randomized trial by Tarumi et al. tested aerobic exercise in cognitively normal older adults and described the evidence for preventing or slowing age-related cognitive decline as inconsistent. A year of supervised aerobic exercise is a more intensive test than a short home program. If cognitive effects remain inconsistent under those conditions, a 12-week internet program in MS needs direct evidence before anyone presents it as a treatment for processing speed.

Why the MS evidence cannot yet give a dose prescription

The Adnan review identified several field-wide problems that block a clean answer. Studies used inconsistent exercise-intensity prescriptions and reporting, so researchers cannot determine whether participants received comparable doses. Adherence measures also varied, making it hard to separate a weak intervention from weak participation. Without objectively verified exercise exposure—such as reliable step counts or heart-rate data—a dose-response relationship remains speculative.

Cognitive measurement poses another problem. The reviewed studies used heterogeneous cognitive outcomes, preventing a pooled analysis focused on processing speed. Most were small feasibility studies rather than adequately powered randomized trials, and follow-up periods were usually short. Even if a program improves a test score during treatment, the literature cannot yet say whether the gain persists after coaching stops.

Remote exercise for MS has established feasibility. Cognitive efficacy, especially for processing speed, remains unestablished.

This is not a semantic distinction. A program can be popular, accessible, and physically helpful while still failing to change the cognitive outcome a patient most wants to improve.

Pairing movement with cognitive training has a better rationale

Studies in adjacent populations suggest that directly training cognition may matter. In a randomized clinical trial of 72 community-dwelling adults with amnestic mild cognitive impairment, mean age 72.4 years, Fairchild et al. compared combined aerobic and resistance exercise plus cognitive training with stretching plus cognitive training. The combined program improved cognitive function. The population was not MS, and both groups received cognitive training, so this study does not show that exercise alone improves cognition or that the same program will work in MS. It does show that a multimodal design can produce cognitive effects in a group with established cognitive vulnerability.

Another randomized study examined computerized cognitive training with or without aerobic exercise in older adults ages 65 to 85. The analysis included 41 participants assigned to cognitive training alone, 41 assigned to cognitive training plus aerobic exercise, and 42 active controls. After eight weeks, Boa Sorte Silva et al. reported reduced within-person variability in cognitive performance after cognitive and exercise training. Such variability can be an early marker of cognitive decline. It is a sensitive performance measure, not proof that participants preserved real-world function, and it does not establish an MS treatment. Still, it points toward a more plausible design than exercise as a lone cognitive lever.

Personal response may matter, but the genetics are preliminary for MS

A six-month aerobic-exercise randomized trial in women with early-stage breast cancer found that BDNF methylation and the BDNF rs6265 Val66Met genotype significantly moderated cognitive response. The finding offers a concrete explanation for why average exercise effects can hide responders and nonresponders.

It does not establish that BDNF testing can identify exercise responders in MS. Breast cancer survivors may have chemotherapy-related neurotoxicity, hormonal changes, and different inflammatory biology. BDNF is biologically plausible here, but plausibility is not a clinical recommendation. An MS trial would need to measure the same genetic and epigenetic factors prospectively, then test whether they predict a prespecified cognitive outcome.

Diet belongs in the measurement plan

Diet may also affect cognitive outcomes enough to complicate exercise trials. A 2025 systematic review and meta-analysis by Guo et al. included 19 randomized trials and 12,119 healthy adults. Dietary interventions produced a small positive effect on overall cognitive function versus control (standardized mean difference 0.14; 95% confidence interval, 0.08 to 0.20; P<0.00001). The result was statistically significant, but the effect size was modest, and the population was healthy adults rather than people with MS.

That evidence does not justify selling a particular diet as a cognitive treatment for MS. It does support measuring baseline dietary patterns in future trials. If diet differs substantially between exercise groups, it can blur the interpretation of a cognitive result. A study designed around cognition should account for lifestyle inputs that may influence cognition.

What a useful next trial would test

The model for rigor exists. The EXERT study used a Phase 3, multicenter randomized design to compare aerobic exercise with stretching-flexibility in adults with mild cognitive impairment and Alzheimer's biomarkers. Its population differs from MS, and it cannot answer the MS question. Its design does show the standard needed when cognition is the endpoint rather than a hopeful side effect.

An MS trial should make cognitive processing speed a primary endpoint, define and report exercise intensity consistently, verify adherence objectively, use a credible comparator, and follow participants beyond the active intervention. It should also test a combined exercise-and-cognitive-training program, record dietary patterns, and prespecify whether BDNF-related markers modify response.

For now, people with MS can reasonably use remote exercise to build activity and physical function, with appropriate clinical guidance. They should not buy a program on the promise that it will restore processing speed. The evidence supports the platform. The cognitive prescription still needs to be written.