Atti del XXIII Convegno Nazionale Tabagismo e Servizio Sanitario Nazionale
Pubblicato: 2021-08-31

Smoking and COVID-19: current evidence from Italy and the world

Dipartimento di Ambiente e Salute, Istituto di Ricerche Farmacologiche “Mario Negri” IRCCS, Milano


Since the spread of the coronavirus pandemic in late 2019 (SARS-CoV-2), great effort has been made to understand the potential role of selected lifestyle habits on SARS-CoV-2 infection and the progression of COVID-19 disease. Several studies have been conducted and hundreds of articles or commentaries have been published on the relationship between cigarette smoking and the risk of incidence, severity and mortality for COVID-19. These articles frequently showed contrasting results that often-confused readers. In occasion of the World No Tobacco Day we tried to elucidate on the current scientific evidence regarding the relationship between smoking and COVID-19. It was then presented a multicentric Italian study with the aim of quantifying the association between smoking and severity and mortality for COVID-19: the COSMO-IT study [1].

Since the start of the pandemic, a number of articles published in peer-reviewed journals - or in preprint archives - have provided information on the prevalence of smokers among subjects with SARS-CoV-2 infection or among series of patients with COVID-19 disease. The prevalence of smokers was quite heterogeneous, ranging from 0% to 70% in Asia, from 0% to 23% in Europe, and from 1% to 35% in the Americas. More importantly, smoking prevalence was systematically lower than adult prevalence estimates observed in countries where the studies were conducted [2]. This suggested that smoking could have a protective role on the incidence of SARS-CoV-2 infection. It is important to mention that these investigations hid numerous limitations. The study design of all these researches is the one used in case series which cannot be used to derive any causal conclusions since they do not have a control group. In addition, most of these studies are subjected to information bias and, above all, important selection bias: many case series are indeed based on selected populations with a low proportion of smokers (e.g., elderly, health workers). Other studies are not designed to investigate smoking and in some of them this information is derived only from medical records (frequently without information on smoking habits). Finally, often (or always) patients with a more severe Covid-19 are systematically excluded from the studies. Therefore, evidence of the Association between smoking and SARS-CoV-2 incidence cannot be based on this type of studies. Instead, evidence should be built on the increasing number of cohort studies investigating the incidence of SARS-CoV-2 infection on samples based on the general population. Simons et al., in a meta-analysis -based on this type of studies, updated in March 2021 [2], showed how, compared to never smokers, current smokers have a relative risk, RR, of 0.71 (confidence interval, CI, 95%: 0.61-0.82) and ex-smokers of 1.03 (95% CI: 0.95-1.11) of being infected with SARS-CoV-2.

For what concerns COVID-19 severity and mortality, the evidence is more robust: compared to non-smokers, current smokers have more frequently an unfavourable progression of the disease [3, 4]. This is confirmed by another meta-analysis conducted and updated in 2021 by Umnuaypornlert et al.[5], who pointed out that current smokers have a greater risk of both COVID-19 severity (RR = 1.58; 95% CI: 1.16-2.15) and mortality (RR = 1.46; 95% CI: 1.18-1.79) (Figure 1). Ex-smokers also show an excess of risk of a severe progression of the disease compared to never smokers. Therefore, the current scientific evidence shows that many case-series of COVID-19 patients report a low prevalence of smokers. These studies are subjected to major limitations and should not be considered for the assessment of any causal association. We need to base our knowledge on cohort studies. Meta-analyses based on these longitudinal studies show contrasting and less striking results regarding the role of smoking on SARS-CoV-2 infection. On the other hand, there is strong evidence of an excess-risk of 30-50% on COVID-19 severity and mortality forever smokers (both current and ex-smokers) compared to non-smokers.

Few data are available from Italy. The COSMO-IT study has just completed the recruitment of about 2000 patients with a confirmed diagnosis of COVID-19 and soon the analysis of the entire sample of patients will be carried out. It is an observational, longitudinal, and multicentric Italian study (24 centers selected throughout Italy, Figure 2), conducted by the Istituto di Ricerche Farmacologiche Mario Negri in Milan in collaboration with the Istituto per lo studio, la prevenzione e la rete oncologica (ISPRO) of Florence, the Società Italiana di Tabaccologia (SITAB), the Istituto Superiore di Sanità (ISS), the Sapienza University of Rome and the Istituto Nazionale dei Tumori of Milan [1].

The COSMO-IT study aims to investigate the role of smoking and other avoidable risk factors on the prognosis of COVID-19. In particular, the project aims to quantify the Association between smoking habits and risk of severity and mortality of COVID-19, assess the role of smoking cessation on the progression of COVID-19, quantify the Association between other potential risk factors and the risk of an unfavourable progression of COVID-19, compare the smoking status in hospitalized patients with that in patients placed in home isolation in some territorial realities. The questionnaire, delivered through either a paper-and-pencil mode, or an online or an offline tool, collects information on socio-demographic characteristics, habits and lifestyles, comorbidity and past drug therapies, symptoms, treatment for COVID-19, and potential risk factors, such as: body mass index, smoking habit, electronic cigarette use, heated tobacco product use, second-hand smoke, alcohol consumption, pre-existing diseases, sleep apnea and drug use. The questionnaire also contains a specific section on hospitalization or home isolation subsequent to COVID-19 diagnosis containing information on hospitalization and treatment during hospitalization/isolation, complications during hospitalization/isolation, outcome or transfer to another facility/hospital.

Some preliminary results were presented during the World No Tobacco Day. Initial analyses were carried out on 1406 COVID-19 patients from 16 centres. In this population, 18.6% had an unfavourable progression of the disease (i.e., composite outcome: intubation, intensive care unit or death). Mortality was only available in selected centers, where 15.1% of patients died due to COVID-19.

The average age of the entire sample was 65 years. It is important to note, however, that the average age of current smokers was 58, the average age of ex-smokers was 67, and the average age of those who had never smoked was 64 years. Therefore, compared to never smokers, ex-smokers were 3 years older and current smokers were 6 years younger. These age differences are not negligible. Indeed, they are hugely relevant if we consider that in Italy, as in the rest of the world, the risk of lethality for COVID-19 cases increases significantly with increasing age: even threefold every 10 years of age [6]. Therefore, even assuming that there is no Association between smoking and COVID-19 mortality risk, the different age distribution of current smokers, ex-smokers and never smokers would result in a misleading substantial reduction in the risk of mortality for current smokers compared to never smokers and to an excess of risk for ex-smokers. In the analysis of the Association between smoking and risk of (severity and) mortality for COVID-19 it is therefore necessary to remove the effect of age by conducting multivariable analyses adjusting at least for age (and sex). Unfortunately, most of the studies available so far in the scientific literature on this topic do not provide adjusted risk estimates, not even for age [2, 5]. In all these studies, providing only univariate risk estimates (also called “rough estimates”), risks are likely to be under-estimated for current smokers and over-estimated for ex-smokers. This perspective shows how it is still necessary to provide new data well collected and well analyzed to quantify the role of smoking on COVID-19. In the COSMO-IT study, multivariate analyses, adjusted for sex, age, education and comorbidity, show that, compared to men, women are less likely to develop a severe form of COVID-19 (odds ratio, OR = 0.62; 95% CI: 0.45-0.84). In addition, the frequency of severe cases or mortality for COVID-19 increases with increasing age (p per trend < 0.001).

In conclusion, the scientific literature shows contrasting results on the role of smoking on SARS-CoV-2 infection but a clear evidence of an excess risk by 30-50% on COVID-19 severity and mortality for smokers compared to non-smokers. Preliminary data from the COSMO-IT study highlight the need to consider only age-adjusted (and sex-adjusted) risk estimates. The COSMO-IT will be among the first studies providing important results on the Association between smoking, smoking cessation, electronic cigarette use, heated tobacco product use and second-hand smoke and the risk of severity and mortality for COVID-19. Above all, this study shows that with motivation and commitment, even without funding, it is possible to conduct independent multi-centric studies that can provide a huge contribution to scientific knowledge.

Figures and tables

Figura 1.Smoking and severity/mortality for COVID-19 (mod. by Umnuaypornlert et al.) [5].

Figura 2.Italian Centers (24) involved in COSMO-IT Trial.


  1. Cattaruzza MS, Gorini G, Bosetti C, Boffi R, Lugo A, Veronese C. Covid-19 and the role of smoking: the protocol of the multicentric prospective study COSMO-IT (COvid19 and SMOking in ITaly). Acta Biomed. 2020; 91:e2020062. DOI
  2. Simons D, Shahab L, Brown J, Perski O. The association of smoking status with SARS-CoV-2 infection, hospitalisation and mortality from COVID-19: a living rapid evidence review with Bayesian meta-analyses (version 11). Qeios. 2021. DOI
  3. Gallus S, Lugo A, Gorini G. No double-edged sword and no doubt about the relation between smoking and COVID-19 severity. Eur J Intern Med. 2020; 77:33-5. DOI
  4. Patanavanich R, Glantz SA. Smoking is associated with COVID-19 progression: a meta-analysis. Nicotine Tob Res. 2020; 22:1653-6. DOI
  5. Umnuaypornlert A, Kanchanasurakit S, Lucero-Prisno DEI, Saokaew S. Smoking and risk of negative outcomes among COVID-19 patients: a systematic review and meta-analysis. Tob Induc Dis. 2021; 19:09. DOI
  6. Il Sole 24 ORE. Coronavirus in Italia, i dati e la mappa.Publisher Full Text


Silvano Gallus

Dipartimento di Ambiente e Salute, Istituto di Ricerche Farmacologiche “Mario Negri” IRCCS, Milano


© Sintex Servizi S.r.l. , 2021

  • Abstract visualizzazioni - 911 volte
  • PDF downloaded - 93 volte