Physical Activity, Exercise, and Rehabilitation in Myasthenia Gravis: A Systematic Review and Narrative Synthesis

Myasthenia gravis causes fluctuating, fatigable weakness, creating uncertainty about physical activity and exercise. This systematic review examined habitual physical activity and sedentary behaviour, barriers and facilitators to exercise, structured exercise, respiratory muscle training, and physical rehabilitation in autoimmune myasthenia gravis. PubMed, Scopus, and Web of Science Core Collection were searched from inception to 30 July 2026, without database date or language filters; backward citation searching was completed on 1 August 2026. Primary human studies published in English were eligible if they reported a relevant measure of physical activity, a defined physical intervention, or barriers, facilitators, or professional advice as an explicit primary focus. We included randomized trials, non-randomized intervention studies, observational and qualitative studies, and case series of at least two participants. Multiple reports from the same cohort or trial were linked before synthesis. Risk of bias was assessed with tools appropriate to each design, and certainty was rated by outcome domain. Of 2,767 database records imported into EndNote (Clarivate, Philadelphia, PA), 1,946 remained after deduplication. Across database and citation searching, 66 reports were assessed in full, and 47 reports representing 38 study units (whole studies or distinct study components) were included. In several observational datasets, greater activity was associated with less fatigue or lower disease burden; however, most analyses were cross-sectional and vulnerable to confounding and reverse causality. Intervention studies were small and heterogeneous. Some secondary outcomes and results measured after treatment favoured the intervention for respiratory muscle performance, functional capacity, strength, or patient-reported outcomes, and some measures improved within groups; however, several controlled primary outcomes were null. Evidence from perioperative care, recovery after crisis or hospitalization, and paediatric populations addressed distinct clinical contexts and was not generalized to stable community-dwelling adults. Some selected cohorts achieved high completion or attendance, but reporting of adverse events was incomplete, and certainty about safety was very low. Individually adapted activity and exercise appeared feasible for many selected participants with clinically stable, predominantly mild-to-moderate disease. However, efficacy remains very uncertain, and safety has not been established across the broader myasthenia gravis population.