Abstract
With the acceleration of global population aging, sarcopenia (SA) and cognitive impairment have become major public health concerns. Observational studies suggest associations between them, but such findings may be biased by confounding and reverse causation. Previous Mendelian randomization (MR) studies have yielded inconsistent results. Therefore, this study applied a bidirectional MR design to clarify their causal relationships. We obtained genetic instrumental variables for SA-related traits, including hand grip strength, lean mass, and usual walking pace, from the UK Biobank. Data on cognitive performance were derived from a large genome-wide association studies meta-analysis published in Nature Genetics in 2018 (n = 257,841), with replication performed in an independent cohort from the IEU Open genome-wide association studies database (n = 22,593). Bidirectional MR analyses were primarily conducted using the inverse variance weighted method, supplemented by MR Egger, weighted median, and weighted mode approaches to validate robustness. Heterogeneity and pleiotropy were evaluated by Cochran Q and MR Egger intercept tests, while leave-one-out and Steiger directionality tests assessed robustness. Inverse variance weighted analysis demonstrated significant positive causal effects of genetically predicted right hand grip strength (odds ratio [OR] = 1.112, 95% confidence interval [CI]: 1.032-1.198, P = .005), right arm fat-free mass (OR = 1.060, 95% CI: 1.021-1.100, P = .002), left arm fat-free mass (OR = 1.046, 95% CI: 1.008-1.085, P = .017), right leg fat-free mass (OR = 1.060, 95% CI: 1.023-1.098, P = .001), left leg fat-free mass (OR = 1.050, 95% CI: 1.014-1.087, P = .006), whole body fat-free mass (OR = 1.081, 95% CI: 1.045-1.119, P < .001), and walking pace (OR = 1.414, 95% CI: 1.172-1.705, P < .001) on cognitive performance. In the reverse MR analysis, no significant causal effects of cognitive performance on SA-related traits were observed, except for walking pace. Genetic prediction of SA-related traits may indicate that cognitive impairment is a potential pathogenic factor. Monitoring the muscle health of older adults and preventing the onset or progression of SA could potentially slow cognitive decline. Furthermore, bidirectional MR results suggest a strong causal relationship between walking pace and cognitive performance.