BACKGROUND: There is a lot of debate about the effects of smoking on COVID-19. A recent fixed-effects meta-analysis found smoking to be associated with disease severity among hospitalized patients, but other studies report an unusually low prevalence of smoking among hospitalized patients. The purpose of this study was to expand the analysis by calculating the prevalence odds ratio (POR) of smoking among hospitalized COVID-19 patients, while the association between smoking and disease severity and mortality was examined by random-effects meta-analyses considering the highly heterogeneous study populations.

METHODS: The same studies as examined in the previous meta-analysis were analyzed (N = 22, 20 studies from China and 2 from USA). The POR relative to the expected smoking prevalence was calculated using gender and age-adjusted population smoking rates. Random-effects meta-analyses were used for all other associations.

RESULTS: A total of 7162 patients were included, with 482 being smokers. The POR was 0.24 (95%CI 0.19-0.30). Unlike the original study, the association between smoking and disease severity was not statistically significant using random-effects meta-analysis (OR 1.40, 95%CI 0.98-1.98). In agreement with the original study, no statistically significant association was found between smoking and mortality (OR 1.86, 95%CI 0.88-3.94).

CONCLUSION: An unusually low prevalence of smoking, approximately 1/4th the expected prevalence, was observed among hospitalized COVID-19 patients. Any association between smoking and COVID-19 severity cannot be generalized but should refer to the seemingly low proportion of smokers who develop severe COVID-19 that requires hospitalization. Smokers should be advised to quit due to long-term health risks, but pharmaceutical nicotine or other nicotinic cholinergic agonists should be explored as potential therapeutic options, based on a recently presented hypothesis.