Causal effects of gut microbiota on pneumonia: a two-sample Mendelian randomization study

Abstract

Background Emerging evidence indicates substantial alterations in gut microbiota among individuals with pneumonia. Nevertheless, it is still controversial whether the alterations of gut microbiome are the causation or result in the development of pneumonia. Consequently, a Mendelian randomization (MR) study was conducted to investigate the causal effects. Methods We used IVW-FE as the main analysis method and considered the following methods, Maximum likelihood (ML), MR‒Egger regression, and simple median, to verify the authenticity of the results. In addition, the Cochran’s Q statistic, the MR Egger intercept test, the MR-PRESSO test, the leave-one-out analysis, and the Bonferroni correction were conducted as sensitivity analysis. Results The IVW-FE method revealed that a higher abundance of family Defluviitaleaceae (OR = 0.912, 95% CI = 0.864–0.964, P = 9.99E-04) was significantly correlated with a lower risk of pneumonia after the Bonferroni correction. Additionally, 6 nominal causal relationships (4 harmful and 2 preventive features) were discovered between gut microbiota and the susceptibility to pneumonia. As for three phenotypes, 10 nominal causal effects (4 harmful and 6 preventive features), 13 nominal causal effects (9 harmful and 4 preventive features), and 8 nominal causal effects (4 harmful and 4 preventive features) were identified for bacterial pneumonia, viral pneumonia, and other pneumonia, respectively. Additionally, no heterogeneity or horizontal pleiotropy was found. Conclusions This MR research found several causal effects between specific gut microbiota and pneumonia, as well as among its three phenotypes. Novel opinions for noninvasive diagnostic biomarkers and targets for therapeutic intervention in pneumonia patients were provided.