The causal association of smoking, alcohol intake, and coffee intake with the risk of bacterial pneumonia: A Mendelian randomization study.

BACKGROUND

At present, the association of smoking, alcohol intake, and coffee intake with the risk of bacterial pneumonia (BP) remains controversial. In this study, we used a 2-sample Mendelian randomization (MR) analysis to estimate the association of smoking, alcohol intake, and coffee intake with the risk of BP.

METHODS

We extracted genetic variants associated with smoking initiation and cigarettes per day from the Genome-Wide Association Study and Sequencing Consortium of Alcohol and Nicotine Use database (944,625 individuals). We also extracted genetic variants associated with past tobacco smoking, alcohol intake frequency, and coffee intake from the UK Biobank database (1,316,166 individuals). BP outcomes were chosen from the FinnGen genome-wide association studies (GWAS) database (7987 patients and 188,868 controls). The inverse variance-weighted method was used primarily to calculate odds ratios (OR) and 95% confidence intervals (CI). Sensitivity analysis using different approaches such as weighted median, MR Egger, and MR pleiotropy residual sum and outlier (MR-PRESSO) have been implemented, as well as leave-one-out analysis to identify pleiotropy.

RESULTS

The 2-sample MR analysis supported the causal association of genetically predicted cigarettes per day (OR: 1.23, 95% CI: [1.08-1.39], P < .01] and smoking initiation (OR: 1.22, 95% CI: [1.03-1.44], P = .02) with the risk of BP, but not past tobacco smoking, alcohol intake frequency, and coffee intake. Heterogeneity (P > .05) and pleiotropy (P > .05) tests provided confirmatory evidence for the validity of our MR estimates.

CONCLUSION

Our findings provide relevant evidence for a favorable causal association of genetically predicted smoking initiation and cigarettes per day with BP risk. However, there may not be a causal association between past tobacco smoking, alcohol intake, and coffee intake with increased BP incidence rates.