Comparison of two different frailty metrics and associations with COVID-19: A bidirectional Mendelian randomization study

Background: Several traditional observational studies suggested a strong association between frailty with coronavirus disease 2019 (COVID-19). However, whether the observed association reflects causality remained unclear. We employed a bidirectional Mendelian randomization (MR) study to investigate the causal relationship of frailty, measured by the Frailty Index and Fried Frailty Score, with COVID-19. Methods We extracted summary genome-wide association statistics for the Frailty Index (N = 164,610), Fried Frailty Score (N = 386,565), COVID-19 (Ncase = 159,840, Ncontrol = 2,782,977), hospitalized COVID-19 (Ncase = 44,986, Ncontrol = 2,356,386) and severe COVID-19 (Ncase = 18,152, Ncontrol = 1,145,546). Independent single nucleotide polymorphisms at genome-wide significance for each phenotype were taken as instruments. The random-effects inverse‐variance weighted method was applied as the primary method, followed by various sensitivity and validation analyses. Results No causal effect of Frailty Index between COVID-19 was observed. Genetically predicted Fried Frailty Score was significantly associated with increased risk of COVID-19 hospitalization (odds ratio [OR] 2.06, 95% confidence interval [CI] 1.45–2.91, P < 0.0001), and suggestively associated with higher risk of COVID-19 susceptibility (OR = 1.19, 95% CI 1.01–1.39; P = 0.035) and COVID-19 severity (OR = 2.10, 95% CI 1.10–4.01; P = 0.025). Sensitivity and validation analyses also received broadly concordant results. There is no insignificant association for reverse causation. Conclusion Our study demonstrated that Fried Frailty Score could increase the risk of COVID-19. Future development should focus on long-term mutual influence between frailty and COVID-19 to alleviate the complications of diseases.