Genetically-predicted effects of lifestyle factors on frailty: Evidence from Mendelian randomization study

Min Xin,Wenxin Wang,Ming Zhou,Yuhan Geng,Hao Liu,Wenxi Luo,Gong Zi Zhang,Liping Huang

DOI: https://doi.org/10.1016/j.archger.2024.105662

Abstract:Objective: To evaluate the causal relationships between genetically predicted lifestyle factors and frailty using Mendelian randomization(MR). Methods: We extracted summary data from genome-wide association studies conducted among individuals of European ancestry, examining lifestyle factors such as smoking, alcohol consumption, physical activity, and sedentary behaviors. The outcomes were assessed using Fried Frailty Score (FFS) and Frailty Index (FI). We conducted 2-sample single-variable Mendelian randomization (SVMR) and multivariable Mendelian randomization (MVMR) to simultaneously assess the independent causal effects were primarily estimated using inverse variance weighted methods. Multiple sensitivity and validation analyzes were used. Results: The IVW analyzes indicated that smoking increased frailty risk (FFS: β = 0.107, 95 % CI = 0.057 to 0.156, P < 0.001; FI: β = 0.899, 95 % CI = 0.016 to 0.191, P = 0.020.), this effect was amplified in the MVMR analysis after adjusting for alcohol consumption. Strenuous sports or other exercise(SSOE) reduced frailty risk (FFS: β = -0.473, 95 % CI = -0.646 to -0.299, P < 0.001; FI: β = -0.423, 95 % CI = -0.692 to -0.154, P = 0.002). Vigorous and moderate-to-vigorous physical activities were significantly related to lower FFS, although no effects were observed on FI. Increased television watching was linked to higher frailty incidence (FFS: β = 0.227, 95 % CI = 0.197 to 0.257, P < 0.001; FI: β = 0.297, 95 % CI = 0.249 to 0.346, P < 0.001), the impact remained persistent in MVMR adjusting for driving and computer use. Conclusion: This study suggests that modifications in smoking, alcohol consumption, and physical activity may help prevent or manage frailty.