Absence of a causal link between COVID-19 and deep vein thrombosis: Insights from a bi-directional Mendelian randomisation study.

Abstract

Several large-scale observational studies have found deep vein thrombosis (DVT) to be related with coronavirus disease 2019 (COVID-19). However, whether there is a clear causal connection between the two is unknown. Our primary analytical method was the inverse variance-weighted (IVW) approach, complemented by the Mendelian randomisation-Egger (MR-Egger) and weighted median methods. We also used MR-Egger to examine the presence of pleiotropy and the Mendelian randomisation pleiotropy residual sum and outlier (MR-PRESSO) approach to analyse for heterogeneity in the data. We did not observe a direct causal relationship between COVID-19 susceptibility (odds ratio (OR) = 1.023; 95% confidence interval (CI) = 0.828-1.264, standard error (SE) = 0.108, P = 0.833), hospitalisation (OR = 1.030; 95% CI = 0.943-1.125, SE = 0.374, P = 0.720), severity (OR = 0.994; 95% CI = 0.923-1.071, SE = 0.038, P = 0.877), and DVT. The results of the reverse Mendelian randomisation (MR) for DVT and COVID-19 susceptibility exhibited heterogeneity and horizontal pleiotropy. Even after removing outliers, we detected no direct causal relationship between the two (OR = 1.015; 95% CI = 0.954-1.080, SE = 0.032, P = 0.630). Similarly, we found no direct causal relationship between DVT and COVID-19 hospitalisation (OR = 0.999; 95% CI = 0.907-1.102, SE = 0.050, P = 0.999) or severity (OR = 1.014; 95% CI = 0.893-1.153, SE = 0.065, P = 0.826). In this MR study, we identified no direct causal impact in a European population between DVT and the COVID-19 susceptibility, severity, or hospitalisation.