A causal relationship between sarcopenia and cognitive impairment: A Mendelian randomization study

Abstract

Objective Sarcopenia and cognitive impairment often coexist in the elderly. In this study, we investigated the causal relationship between sarcopenia-related muscle characteristics and cognitive performance. Methods We used linkage disequilibrium score regression (LDSC) and Mendelian Randomization (MR) analyses to estimate genetic correlations and causal relationships between genetically predicted sarcopenia-related muscle traits and cognitive function, as well as cognitive function-based discovery samples and replicated samples. Estimated effect sizes were derived from a fixed-effects meta-analysis. Results Our univariate genome-wide association study (GWAS) meta-analysis indicated a causal relationship between appendicular lean mass (ALM) (β = 0.049; 95% confidence interval (CI): 0.032–0.066, P < 0.001) and walking pace (β = 0.349; 95% CI: 0.210–0.487, P < 0.001) with cognitive function, where a causal relationship existed between ALM in both male and female (βALM-Male(M) = 0.060; 95% CI: 0.031–0.089, PALM-M < 0.001; βALM-Female(F) = 0.045; 95% CI: 0.020–0.069, PALM-F < 0.001) with cognitive function. Low grip strength was not causally associated with cognitive function (β = -0.045; 95% CI: -0.092 - -0.002, P = 0.062). A reverse causality GWAS meta-analysis showed a causal relationship between cognitive function and ALM (β = 0.033; 95% CI: 0.018–0.048, P < 0.001) and walking pace (β = 0.039; 95% CI: 0.033–0.051, P < 0.001), where ALM in both male and female showed a causality (βALM-M = 0.041; 95% CI: 0.019–0.063, PALM-M < 0.001; βALM-F = 0.034; 95% CI: 0.010–0.058, PALM-F = 0.005). Cognitive function was not causally related to low grip strength (β = -0.024; 95% CI: -0.073–0.025, P = 0.344). Multivariable MR1 (MVMR1) analyses showed a significant causal relationship for ALM (β = 0.077; 95% CI: 0.044–0.109, P = 0.000) and walking pace (β = 0.579; 95% CI: 0.383–0.775, P = 0.000) and cognitive function. Multivariable MR2 (MVMR2) multivariate analysis showed that ALM causality remained (β = 0.069; 95% CI: 0.033–0.106, P = 0.000), and walking pace (β = 0.589; 95% CI: 0.372–0.806, P = 0.000). Conclusions Bidirectional two-sample MR demonstrated that sarcopenia-related muscle characteristics and cognitive performance were positive causal genetic risk factors for each other, while a multivariable MR study demonstrated that low ALM and a slow walking pace were causally involved in reduced cognitive performance. This study suggests a causal relationship between sarcopenia and cognitive impairment in older adults and provide new ideas for prevention and treatment.