COVID-19 and smoking: Considerations after two years

Although both virulence and pathogenicity of SARS-CoV-2 have considerably declined over time, the number of coronavirus disease 2019 (COVID-19) cases and the consequent clinical burden remain substantial worldwide, thus persuading the World Health Organization (WHO) to reiterate the status of public health emergency of international concern for this pathology.

One of the pivotal aspects in the managed care of COVID-19 encompasses the identification of subjects at enhanced risk of developing severe and/or critical illness, since an early and targeted treatment may contribute to lower the clinical, social, and healthcare burden of this life-threatening condition. A recent meta-analysis published by Vardavas and colleagues [1], including 88 cohort studies with both gender- and age-adjusted information from nearly 7 million European participants, has corroborated the straightforward concept that the most predictive factors of unfavorable disease progression are male sex and obesity, along with a kaleidoscope of comorbidities such as chronic obstructive pulmonary disease (COPD), cardiovascular disease, heart failure, cancer, diabetes, renal and liver impairment, dementia, and immunosuppression. Although this conclusion is certainly not surprising and supports the findings of many other similar publications converging to identify these and other possible genetic, clinical, and environmental factors as key determinants of unfavorable prognosis, there is still an open debate about the role played by additional conditions, one of the most disputed of which is undoubtedly smoking.

In the first early meta-analysis that we published in this journal in 2020 [2], which included only 5 studies and 1399 Chinese COVID-19 patients, we failed to find a significant association between active (cigarette) smoking and the risk of developing severe COVID-19 illness (pooled odds ratio [OR], 1.69; 95%CI, 0.41–6.92; p = 0.254). Notably, after the publication of our meta-analysis, Simons et al. also performed a living rapid evidence review with Bayesian meta-analyses (233 studies included) [3]. They found that current smokers had a lower risk of SARS-CoV-2 infection compared with never smokers (relative risk [RR], 0.74; 95% CI, 0.58–0.93], also evidencing that the data linking current smoking with hospitalization (RR, 1.06; 0.82–1.35), higher disease severity (RR, 1.25; 0.85–1.93), and mortality (RR, 1.22; 95%CI, 0.78–1.94) were inconclusive. Another large cohort study published by Chen et al. [4], reporting data on ∼1.2 million adult participants followed from February 1 to August 31, 2020, revealed that current smokers had an over 30% lower risk of COVID-19-related death (hazard ratio (HR), 0.57; 95%CI, 0.49–0.67). These results, especially those of our former meta-analysis, were fiercely contested by a number of subsequent letters, even in different journals, attempting to evidence some flaws in calculations and conclusions. Although stirring up further controversies or directly replying is out of the scope of this article, and while we are not willing here to specifically address some justified – along with others clearly biased – considerations, we have been offered the opportunity to provide a two-year recap of the many controversies emerged around the role played by smoking in influencing the clinical course of SARS-CoV-2 infection.

The best starting point is reporting the outcome of recent (large) studies and meta-analyses. The most recent meta-analysis that we could find about this topic has been published by Gallus and colleagues (including 320 publications) [5]. Notably, the authors found a positive relationship between current smoke and disease severity (OR, 1.35; 95%CI, 1.22–1.48) and mortality (OR, 1.32; 95%CI 1.20–1.45), but not with hospitalizations (1.08; 95%CI 0.98–1.19), which implicitly underlines that the link between current smoking and clinical severity is not so straightforward. However, the correlation was seemingly clearer with former smoking since a positive relationship was found with all three endpoints (OR between 1.16–1.46). Similar conclusions emerged from another meta-analysis by Patanavanich et al. [6], where the authors could estimate a marginally significant association between current smoking and COVID-19-related mortality (OR, 1.26; 95%CI, 1.01–1.58), which, however, turned out to be no longer significant in people living in high-income countries (OR 1.14; 95%CI, 0.91–1.43). As concerns post- (or long) COVID, Tsampasian et al. highlighted a relatively modest association between higher risk of this condition and smoking status (OR, 1.10; 95%CI, 1.07–1.13) [7]. In another meta-analysis where the authors explored the role of potential predictors of venous thromboembolism in patients with COVID-19 [8], smoking history was paradoxically associated with a lower risk of thrombotic events (OR, 0.71; 95%CI, 0.54–0.93). In one of the largest epidemiological studies ever published so far on this matter, which included over 28 million French inhabitants [9], Semenzato et al. failed to find a significant association between smoking and COVID-19-related hospitalization (OR, 1.03; 95%CI, 0.91–1.16) or in-hospital death (OR, 1.01; 95%CI, 0.75–1.37). This is not really surprising as the existence of a negative association between current smoking and risk of severe COVID-19 illness was confirmed in other epidemiological investigations such as that published by Galanti et al. (RR for hospital admission, 0.60; 95%CI, 0.47–0.76; RR for intensive care, 0.43; 95%CI, 0.21–0.89) [10]. Even in the large study published by Piasecki et al. [11], former but not current smoking emerged as a predictive factor for severe outcomes in patients hospitalized with COVID-19. Importantly, using serum cotinine levels (i.e., a more validated tool for establishing current smoking status than self-reporting use), Tomaselli et al. identified a lower rate of positive SARS-CoV-2 serology among current smokers compared to never/former smokers [12]. The abovenoted results affirm that a simple conclusion that current smoking increases the risk of clinical deterioration in COVID-19 is largely unsupported according to the currently available scientific evidence.

The challenging biological relationship between cigarette smoke and COVID-19 has also been addressed in some interesting works. For example, Aliee et al. reported that the expression of angiotensin-converting enzyme 2 (ACE2; the main SARS-CoV-2 host receptor) in the nasal mucosa does not differ between current and never smokers, thus highlighting that enhanced cell penetration of SARS-CoV-2 is unlikely in the former class of subjects [13]. In another experimental study, Tomchaney et al. found that mice exposed to cigarette smoke displayed reduced levels of ACE2 in bronchial and alveolar epithelial cells, while such treatment was also effective in lowering SARS-CoV-2 replication in Calu-3 cells [14]. In support of these findings, another work published by Porter et al. showed that cigarette smoke extracts may be modestly effective in enhancing the expression of full-length ACE2 on primary human differentiated bronchial epithelial cells [15], but does not significantly modify airway cell infection by SARS-CoV-2. Interestingly, Dormoy et al. postulated that nicotinic acetylcholine receptors (nAChRs) might act as SARS-CoV-2 co-receptors [16]; nicotine may hence act as an inhibitor of cell entry by competing with SARS-CoV-2 for cell surface receptors. This hypothesis has been addressed in an experimental study, where the authors administered nicotine in drinking water to hACE2 mice, observing a lower probability of SARS-CoV-2 RNA expression and pathology in their brain [17]. On the other hand, in another study, nicotine was found to have no significant cytoprotective activity against infection of Vero E6 cells challenged with SARS-CoV-2 [18]. Although the number of randomized trials that have assessed the role of nicotine replacement therapy (NRT) in modulating the risk of severe COVID-19 illness remains negligible, the findings of the study published by Piasecki et al. deserve to be mentioned [11]. Briefly, the authors found that the mortality rate of COVID-19 smokers prescribed with NRT was nearly half than those who did not use NRT (i.e., 4.5 vs. 7.7%; OR, 0.64; 95%CI, 0.50–0.82). Importantly, Muralidharan et al. studied the acute impact of short exposure to cigarette smoke extract and/or alcohol on SARS-CoV-2 infection of ciliated human bronchial epithelial cells of healthy subjects [19], and found that viral titers were only increased in cells treated with both compounds but not with cigarette smoke extract or alcohol alone, thus emphasizing that cigarette smoke may only play a subsidiary role in the pathogenesis of COVID-19. Accordingly, the biological evidence that cigarette smoke alone would enhance the risk of infection and/or unfavorable disease progression remains weak, at best.

Overall, one important aspect that must be highlighted, is that clinical studies and meta-analyses published in different periods should not be directly compared since the epidemiology and clinical burden of COVID-19 have both considerably changed over time due to the emergence of new viral variants, widespread vaccination, earlier and more accurate diagnosis, improved therapeutic management and even human habits. For example, Sarich et al. conducted an interesting meta-analysis about smoking behaviors before and after the onset of the COVID-19 pandemic [20], reporting that not only was the rate of people smoking during the pandemic lower than before (i.e., around 13% less), but 21% of smokers admitted to having decreased the use of tobacco, while 27% reported an increase. The human population has also changed throughout these past 3 years, whereby most of the nearly 7 million deaths caused by COVID-19 involved older people [21]. Thus, a straightforward translation of findings between different publications is profoundly misleading.

The synergy between cigarette smoke and pre-existing morbid conditions is another aspect that needs to be highlighted. Irrespective of the long-term harms directly related to cigarettes and second-hand smoke, such as cancer and cardiovascular disease, smoking also generates a number of acute unfavorable consequences on health, thus aggravating an already compromised status. Thus, at this point, it is certainly challenging to dissect the acute effects on SARS-CoV-2 infection of smoking per se from its negative impact on other co-morbidities or conditions known to enhance the risk of severe COVID-19 illness. For example, He et al. reported that smoking had a strong interaction with the number of comorbidities in influencing COVID-19 severity [22].

In conclusion, we believe that providing a biased view based on preconceived notions does not help science and medicine. Smoking remains the leading cause of preventable morbidity and mortality worldwide, but struggling to find a robust relationship between current cigarette smoking and the risk of developing SARS-CoV-2 infection and/or severe COVID-19 illness is an almost illusory and even counterproductive enterprise, more or less like negating the side effects of the vaccines. Providing preconceived medical theories and unsubstantiated scientific assumptions can only pave the way for the spread of negative beliefs and behaviors. This is exactly what we should avoid. Transparent communication is the key to delivering a trustworthy message. The unfavorable consequences of cigarette smoking are paramount and indisputable. A recent retrospective analysis has estimated that active smoking and exposure to secondhand smoke have caused nearly 21 million premature (and otherwise preventable) deaths in the US during the past 50 years [23]. It has also been recently demonstrated that active smoking may enhance the risk of developing adverse events after COVID-19 vaccination [24]. The evidence that quit smoking is perhaps one of the best things that one could ever do for his/her health is certainly inalienable and should foster the development of anti-smoking campaigns, especially among the youth, who may then experience the largest burden of long-term unfavorable health consequences. Nonetheless, we also emphasize here that the potential role of current smoking in influencing the clinical course of COVID-19 remains controversial and may even be confounded by synergistic effects, e.g., long-term smokers may actually develop a kaleidoscope of conditions - namely COPD and cardiovascular disease - that per se influence the prognosis of SARS-CoV-2 infection.

Declaration of Competing Interest

All authors have no actual or potential conflict of interest including any financial, personal or other relationships with other people or organizations within three years of beginning the submitted work that could inappropriately influence, or be perceived to influence, their work.