Abstract
An association between coronavirus disease 2019 (COVID-19) and tuberculosis (TB) has been reported, although the causal relationship between these factors remains unclear. This study aimed to investigate the causal connection between COVID-19 and TB using Mendelian randomization (MR) analysis. The UK Biobank provided summary data on COVID-19 using the integrative epidemiology unit open genome-wide association studies (GWAS) pool. GWAS data on TB were also retrieved. The relationship between COVID-19 and TB was examined using 5 methods, the major method being inverse variance weighting. Additional methods included weighted median, MR-Egger regression, simple mode, weighted mode, and Wald ratio. No significant correlation was observed between COVID-19 and TB (odds ratio = 1.022, 95% confidence interval 0.955-1.032, P = .611). Reverse MR analysis also confirmed the absence of a causal relationship between COVID-19 and TB risk (P > .05). This study used several complementary MR approaches to explore the bidirectional relationship between COVID-19 and TB and revealed no significant bidirectional relationship. However, given the limited GWAS data for these 2 conditions, caution is warranted when interpreting these results. While previous epidemiological and retrospective studies have suggested that COVID-19 may impact TB, our bidirectional MR analysis based on European population genetic data suggests no two-way causality between COVID-19 and TB in this population.