ABSTRACT
Observational studies have found a relationship between directly measured short leukocyte telomere length (LTL) and severe coronavirus disease 19 (COVID-19). We investigated the causal association between genetically predicted LTL and COVID-19 susceptibility or severity. A previous genome-wide association study (GWAS) of 78,592 European-ancestry participants identified single nucleotidepolymorphisms (SNPs) that can be utilized to genetically predict LTL. Summary-level data for COVID-19 outcomes were analyzed from the COVID-19 Host Genetics Initiative. A two-sample bidirectional Mendelian randomization (MR) study was designed to evaluate these causal relationships. Using an inverse-weighted MR analysis and population-based controls, genetically predicted LTL did not reveal any significant association with COVID-19 susceptibility (odds ratio (OR): 0.94; 95% CI: 0.85-1.04; p = 0.202) or severity (OR: 0.85; 95% CI: 0.70-1.03; p = 0.099). Similar results were found for five other definitions of cases/controls and/or COVID-19 outcomes. Six additional MR methods and sensitivity analyses were conducted after removing variants with potential horizontal pleiotropy and including variants at a liberal significance level, which produced similar results. Using SNPs identified for the prediction of LTL from another GWAS study, we found a non-significant association for COVID-19 susceptibility or severity with narrower CIs toward the null hypothesis. No proof of genetically predicted COVID-19 phenotypes remained causally associated with genetically predicted LTL, and the null association was consistent with a lack of significant genetic correlation. Genetic evidence does not support shorter LTL as a causal risk factor for COVID-19 susceptibility or severity.