BACKGROUND: Patients with sarcopenia often experience cognitive decline, affecting cortical structures, but the causal link remains unclear. We used bidirectional Mendelian randomization (MR) to explore the relationship between sarcopenia-related traits and cortical structure. METHODS: We selected genetic variables from genome-wide association study data. Three different MR methods were used: inverse-variance weighted analysis, MR-Egger regression, and the weighted median test. For significant estimates, we further conducted Cochran's Q test, MR-Egger intercept test, leave-one-out analyses, and MR-PRESSO to assess heterogeneity. RESULTS: In forward MR analysis, appendicular lean mass (ALM) decreased the thickness (TH) of lateral occipital gyrus and increased the TH of pars opercularis gyrus (β = -0.0079 mm, 95% CI: -0.0117 mm to -0.0041 mm, P <
0.0002
β = 0.0080 mm, 95% CI: 0.0042 mm to 0.0117 mm, P <
0.0002). In reverse MR analysis, a significant negative correlation was found between the TH of bankssts and ALM, while positive correlations were observed between the TH of frontal pole, rostral anterior cingulate, temporal pole, and ALM. The TH of temporal pole was positively correlated with right hand grip strength (HGS-R) (β = 0.1596 mm, 95% CI: 0.1349 mm to 0.1843 mm, P <
0.0002), and the TH of pars triangularis was positively correlated with left-hand grip strength (HGS-L) (β = 0.3251 mm, 95% CI: 0.2339 mm to 0.4163 mm, P <
0.0002). CONCLUSIONS: Sarcopenia-related traits and cortical structure have bidirectional effects, supporting the muscle-brain axis theory. This links sarcopenia to neurocognitive diseases and provides new strategies for the prevention and intervention of both sarcopenia and cognitive decline.