Tobacco use, as a leading risk factor of death and disability due to respiratory diseases, was expected to increase the risk of SARS-CoV-2 infection and COVID-19 deaths. Earlier epidemiological studies, however, showed that smokers were underrepresented among patients hospitalized due to COVID-19. The most recent meta-analysis has confirmed these early findings, showing that current smokers had lower risk of SARS-CoV-2 infection than never smokers (Relative risk 0.71, 95% Credible interval 0.61; 0.82). A majority of these studies are based on samples of hospitalized patients, tested population or specific population groups, which might not represent the general population. In addition, data on tobacco use has been primarily collected from electronic health records or retrospectively and therefore prone to misclassification and information bias.
A message of a protective effect of tobacco use could undermine public health efforts to curb its use and reduce the perception of harm in the general population. Studies with general population samples and prospective data collection are thus urgently needed.
The aim of the study is to examine the association between tobacco use and the risk of SARS-CoV-2 infection. We will explore several forms of tobacco use (smoking, moist smokeless tobacco and e-cigarettes) and investigate whether introducing a potential collider bias by adjusting for mediating risk factors (alcohol use, physical activity and obesity) could have explained earlier results. We will use data from a prospective cohort study of nationally representative health surveys in Finland linked to SARS-CoV-2 incidence data, which is less subject to collider and recall bias than previous case-control studies.
