We read with interest the recent report by Nashed et al that described changes in the microbiota caused by viral infections, which in turn affected the severity of the clinical disease.1–4 Interactions between viral infections and microbiota complicate their relationship with host immunity and clinical diseases. However, it is unclear whether those associations are causal. Furthermore, the interactions between the microbiota and infections, as well as the outcomes of those interactions, have only been studied to a limited extent.5–8 We conducted an in-depth statistical analysis based on two-sample bidirectional Mendelian randomisation (MR) to broaden our understanding and provide a comprehensive assessment of the causal relationship between 9 common infections, 140 microbes and 38 clinical diseases.

As instrumental variables for MR analysis, independent genetic variants with genome-wide suggestive significance (p<5e-06) in each exposed dataset from public genome-wide association studies were chosen (online supplemental table 1). The MR causality test was primarily performed with the inverse variance weighted method (figure 1A). Pleiotropy was assessed using the MR Egger regression, and heterogeneity was examined using Cochran’s Q-test. Subsequently, a rigorous filtering procedure was performed, specifically removing MR results based on fewer than three single-nucleotide polymorphisms (SNPs) and MR results that …