We read with great interest the letter by Rossato and Di Vincenzo1 commenting on our editorial “Tobacco and COVID-19: A position from Sociedade Portuguesa de Pneumologia” published in Pulmonology in December 2020.2 We would like to thank the authors for the challenging discussion points and we acknowledge that, as in all fields of science, new data are continuously being collected and interpreted and we are still far from definitive conclusions about how smoking really impacts COVID-19.

The observations that report a low rate of smokers among COVID-19 patients are pertinent but have yet to be confirmed by good quality epidemiological studies designed to address this question. As we mentioned, a high number of studies does not report smoking status or does not distinguish never-smokers from missing data and this may pose considerable bias.2 At the present time, we cannot rule out the hypothesis that smoking may not constitute a strong risk factor for COVID-19, but we have to further specify what questions we are posing: what is the difference in risk between present smokers and having a past history? Are smokers more prone to contract SARS-Cov-2 infection in general? We do not have enough information concerning asymptomatic infection and smoking, and we know there may be a bias toward higher testing rates in smokers,3 due to higher rates of respiratory symptoms. An interesting study in Nature4 shows that there are difficulties in correctly adjusting for other covariates — depending on the covariates included, a smoking history could either be associated with a higher risk of COVID-19 (age and sex adjusted only) or lower risk (fully adjusted model). After adjusting only for demographic factors (age, sex, deprivation and ethnicity), the authors found a non-significant positive hazard ratio for current smoking (HR 1.07 (0.98–1.18)), excluding any protective effect of nicotine and suggesting that any increased risk with current smoking is likely to be small.

Rossato and Di Vincenzo also quote the new paper by Tomchaney et al.,5 reporting conflicting new results that show decreased expression of ACE2 receptors in both bronchial and alveolar epithelial cells exposed to cigarette smoke. These are puzzling data that still wait peer-review and publication. Animal studies are undoubtedly important to open new pathophysiologic hypothesis; however, we still face uncertainties as to the real role of ACE-2 receptor modulation and the risk of infection. One of the problems concerns tobacco smoke, a mixture of thousands of chemicals interacting together, and the presumptive role of nicotine. The studies by Russo et al.6 have thrown some light on how isolated nicotine may facilitate SARS-Cov-2 infection: nicotine, even at low concentrations, increases ACE-2 levels in bronchial cells. Besides, they showed that ACE-2 increase is specifically mediated by α7-nAChR, suggesting that smoking may promote cellular uptake mechanisms of SARS-CoV-2 through α7-nAChR signaling. The presence of this receptor in neuronal tissues also raises questions about the impact of smoking in COVID-19 pathophysiology in several organs, including the brain.

In the absence of well-designed studies, with large populations, any hypothesis on the effect of smoking or nicotine in the risk of COVID-19 remains unproven. More than five decades of research in vast population-based studies were needed to establish the causative effect of tobacco in several deadly diseases. Our minds should be open to change, but in the present state of knowledge, we stand with the WHO and adopt a cautious recommendation against any putative protective effect of smoking.