Genetically Predicted Levels of Serum Metabolites and Risk of Sarcopenia: A Mendelian Randomization Study

Abstract

Metabolites’ connection to sarcopenia through inflammation and mitochondrial dysfunction is presumed, but their impact remains unclear due to limitations in conventional observational studies caused by confounding bias and reverse causation. We conducted a Mendelian randomization (MR) analysis to elucidate the association of serum metabolites with sarcopenia and its related traits, i.e., appendicular lean mass and grip strength. Genetic instruments to proxy the serum metabolites were extracted from the most comprehensive genome-wide association study on the topic published so far. The corresponding summary statistics for the associations of genetic instruments with outcomes were calculated from the UK Biobank (n = 324,976 participants). The primary analyses were assessed by an inverse-variance weighted (IVW) method. The weighted median and MR-PRESSO methods were used as sensitive analyses. Fourteen genetically predicted serum metabolites were associated with the risk of sarcopenia (PIVW < 0.05). Two metabolites showed the overlapped association with sarcopenia and its related traits, which were isovalerylcarnitine (sarcopenia: odds ratio [OR] = 4.00, 95% confidence interval [CI] = 1.11~14.52, PIVW = 0.034; appendicular lean mass: β = −0.45 kg, 95% CI = −0.81~−0.09, PIVW = 0.015; grip strength: β = −1.51 kg, 95% CI = −2.31~−0.71, PIVW = 2.19 × 10−4) and docosapentaenoate (sarcopenia: OR = 0.16, 95% CI = 0.03~0.83, PIVW = 0.029; appendicular lean mass: β = −0.45 kg, 95% CI = 0.08~0.81, PIVW = 0.016). Twenty-seven metabolites were suggestive associated with appendicular lean mass or grip strength. This MR study provided evidence for the potential effects of metabolites on sarcopenia.