BackgroundMyasthenia gravis (MG) is an autoimmune disorder affecting the neuromuscular junction. Despite the potential benefits of higher physical activity and lower sedentary behavior in MG patients, evidence from observational studies for the effect of physical activity on the risk of MG is limited and inconclusive.MethodsWe employed linkage disequilibrium score (LDSC) regression, two-sample Mendelian randomization (MR), and its multivariable extension analyses (MVMR) to assess the relationship between leisure screen time (LST), moderate-to-vigorous intensity physical activity during leisure time (MVPA) and the risk of MG using genome-wide association studies (GWAS) summary datasets. MR analyses were performed using the inverse-variance-weighted (IVW), weighted-median, and MR-Egger regression. Sensitivity analyses were further performed using alternative instruments to test the robustness of our findings.ResultsWe found evidence of genetic overlap between LST (rg = 0.113, P = 0.023) and MG, as well as between MVPA (rg=-0.220, P = 0.0001) and MG, using LDSC method. The results of the MR suggested an association between genetic liability to LST and increased risk of MG (IVW OR = 1.609, 95% CI = 1.153 to 2.244; P = 0.005). This association was particularly notable for late-onset MG (IVW OR = 1.698, 95% CI = 1.145 to 2.518; P = 0.008), but not for early-onset MG. Consistent findings were obtained in the MVMR analysis using BMI as covariate (IVW OR = 1.593, 95% CI 1.167 to 2.173, P = 0.003). However, the MR analysis does not support a substantial causal effect of MVPA on the risk of MG.ConclusionOur findings support a causal effect of sedentary behavior as measured by LST on MG, indicating that lack of exercise may play a role in the development of MG. Longitudinal and interventional studies of this association are warranted.