Cigarette smoking and SARS-CoV-2 infection: multivariable regression and Mendelian randomization analyses in the...

Evidence of whether smoking affects the risk for SARS-CoV-2 infection is mixed. We aimed to clarify the inconsistencies in previous findings and whether the association is potentially causal using different Mendelian randomization (MR) methods.

We examined associations between smoking traits (initiation, cessation, and intensity) and SARS-CoV-2 infection in multivariable logistic regression, and one-sample and two-sample MR analyses. The study included n = 47,506 female and n = 28,229 male study subjects from the Norwegian Mother, Father, and Child Cohort Study (MoBa) with questionnaire data and genotype information. SARS-CoV-2 infection status was obtained by data linkage to the national health registry MSIS.

We found no clear evidence of an association between smoking initiation and SARS-CoV-2 infection (multivariable regression: OR 1.08, 95% CI 0.96 to 1.20, one-sample multivariable MR analysis: OR 1.02, 95% CI 0.96 to 1.09, two-sample MR: OR 1.10 (95% CI 1.06 to 1.13). Also, we found no clear evidence of an association with smoking intensity (multivariable regression: OR 0.78, 95% CI 0.62 to 0.96, one-sample multivariable MR: OR 1.04, 95% CI 0.76 to 1.42, two-sample MR: OR 1.01, 95% CI 0.95 to 1.09, per 1 SD increase in number of cigarettes per day). Nor was there any association with smoking cessation. These findings did not change after accounting for educational level, BMI or risk-taking behavior in multivariable MR analyses.

We did not find robust evidence of causal associations between smoking and SARS-CoV-2 infection. Our investigation of potential violations to MR assumptions highlights the limitations of this approach to examine infection risk associated with smoking.