In Fall 2020, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has Europe firmly in its grip again, and Switzerland, like its neighboring countries, is experiencing a strong second wave. Because the virus is now spreading again, it is important to provide better protection for people at elevated risk of developing severe courses of coronavirus disease 2019 (COVID-19). However, this can only be done if such risk groups are better defined and identified.

Over the last months, the evidence base has become increasingly strong that obesity, alongside with other relevant risk factors, is substantially and independently associated with adverse outcomes of COVID-19 ( Lockhart & O’Rahilly, 2020). A systematic review and pooled analysis of 75 studies recently confirmed that individuals with obesity were more at risk for COVID‐19 positive, for hospitalization, for ICU admission, and for mortality ( Popkin et al., 2020). In the largest analyzed cohort study to date significantly increased risks of COVID-19-related death were seen for obesity grade 3 (body mass index (BMI) ≥40.0kg/m2) and grade 2 (BMI 35.0–39.9kg/m2), but not for obesity grade 1 (BMI 30.0–34.9kg/m2) and overweight (BMI 25.0–29.9kg/m2) ( Williamson et al., 2020).

Beyond that, it has been shown that the association between obesity and COVID-19 is independent of diabetes and other diseases frequently associated with obesity and the metabolic syndrome, meaning that also so-called “healthy” obese people might be at elevated risk of severe COVID-19 ( Beretta, 2020; Tartof et al., 2020). Switzerland seems to be no exception: For example, obesity was a significant part of the high burden of comorbidities in a sample of patients hospitalized with COVID-19 at the cantonal hospital Aarau (Switzerland) in Spring 2020 ( Gregoriano et al., 2020). Thus, two pandemics are currently intertwining ( Lockhart & O’Rahilly, 2020), the communicable one of COVID-19 and the non-communicable one of obesity, both affecting significant parts of the population in Switzerland.

However, for Switzerland, due to the lack of synthesis studies, it is currently not clear how many people – in relative and absolute numbers – are actually affected by obesity at all, and thus have an increased risk of a severe COVID-19 course. There are a handful of individual representative population-based studies conducted between 2012 and 2017, but there are no overall estimates. Consequently, we aim to provide a first overall estimate of how many people in Switzerland are currently affected by grade 2 and 3 obesity. Such estimations may help to better estimate the relevance and size of this group at elevated risk, which could be incorporated into strategies to protect risk groups during the still unfolding COVID-19 pandemic.

We accessed five representative national population-based studies which were conducted between 2012 and 2017 and which include data on height and weight of adult men and women in Switzerland ( Table 1). All five studies are based on representative samples and provide official sample weightings which allow to correct for non-participation. For all data sets we included only adults aged 18 years and older.

The Swiss Health Survey (SHS) is a representative survey of the health status, health behavior and use of health services of the Swiss population and has been conducted every five years since 1992. In the present study, the waves 2012 (SHS12, n=21,597 participants) and 2017 (SHS17, n=21,597 participants) were included. Both survey years were analyzed separately. A detailed description of the data collection, the recruitment procedure, the participation rate and the strategy of sample weighting was published elsewhere ( Bundesamt für Statistik, 2014; Storni et al., n.d.). The SHS data were provided by the Federal Statistical Office (FSO). The Swiss Household Panel (SHP) is a representative longitudinal study of the social changes and living conditions of the Swiss population and has been conducted annually since 1999 with three samples (1999, 2004 and 2013). This study used cross-sectional data from the third sample from 2013 (n=5,040 participants). A detailed description of the data collection, the recruitment process, the participation rate and the strategy of sample weighting was published elsewhere ( Voorpostel et al., 2020). The SHP data are available from the Swiss Centre of Expertise in the Social Sciences (FORS) website. menuCH is the first representative national nutrition survey of Switzerland conducted between 2014 and 2015 (n=2,057 participants). A detailed description of the data collection, weighting strategy, the recruitment process and the participation rate is published elsewhere ( Bochud et al., 2017; Pasquier et al., 2017). The data were provided by the Federal Food Safety and Veterinary Office (FSVO). The Swiss Survey on Income and Living Conditions (SILC) is a representative study of income and living conditions of Swiss households. Households are surveyed over several years and new households are added each year. This study uses cross-sectional data from 2017 (n=12,980 participants). A detailed description of the data collection, the recruitment procedure, the participation rate and the strategy of sample weighting was published elsewhere ( Schweiz. Bundesamt für Statistik BFS, 2014). The SILC data were provided by the Federal Statistical Office (FSO). Since all five data sets are publicly accessible for research purposes and are fully anonymized, no ethics permission was required for the present study.

With the exception of menuCH (where body height and body weight were also measured) the data sets contain exclusively self-reported height and weight information. Based on height and weight we calculated BMI (kg/m2). Very few people with a BMI <14.0 or >60.0 kg/m2 were excluded. Thereafter, BMI was categorized according to the WHO groups for underweight (BMI <18.5 kg/m2), normal-weight (18.5–24.9 kg/m2), pre-obesity (25.0–29.9 kg/m2), obesity grade 1 (30.0–34.9 kg/m2), obesity grade 2 (35.0–39.9 kg/m2), and obesity grade 3 (≥40.0 kg/m2) ( World Health Organization (WHO), 2020).

Our overall estimation shows that in Switzerland in 2012–2017, 31.72 % (95% CI 31.07-32.37) of the adult population were affected by overweight (BMI 25.0–29.9kg/m2) ( Table 2). Of the 11.20% who were obese (BMI ≥30.0kg/m2), 8.86% (95% CI 8.30-9.42) suffered from grade 1 obesity (BMI 30.0–34.9 kg/m2), 1.76% (95% CI 1.50-2.02) from grade 2 obesity (BMI 35.0–39.9 kg/m2), and 0.58% (95% CI 0.50-0.66) from grade severe 3 obesity (BMI>=40.0 kg/m2). Converted into estimated absolute population numbers, this corresponds to a total of approximately n=154,515 people who suffer from grade 2 or 3 obesity (n=116,216 and n=38,298, respectively). Overall, an estimated total of n=2,834,088 or 42.92% of the adult population in Switzerland are overweight or obese (BMI>=25.0 kg/m2).

Table 3 shows the convergence across the individually estimated relative frequencies from each of the five analyzed surveys. The proportion for grade 2 obesity ranged between 1.5% and 2.2%, and for grade 3 obesity between 0.5% and 0.7%. The highest relative frequencies for all three grades of obesity were calculated for the menuCH data.

The stratification by sex shows that men were more affected by overweight and grade 1 obesity ( Figure 1 and Table 4. However, for grade 2 and 3 obesity, there is a weak tendency towards women being slightly more affected. Although there is an increase in the proportion in all overweight and obesity (sub-)categories with increasing age, there are still considerable proportions of younger adults (aged 18–35 years) who suffer from obesity grade 2 (1.04%, 95%CI 0.43-1.65) or grade 3 (0.44%, 95%CI 0.11-0.77).

We show that the proportion of people in Switzerland who suffer from grade 2 or 3 obesity and thus an increased risk of severe COVID-19 course seems small in terms of percentages (2.34%), but when converted into absolute numbers it translates into a considerable number of human lives. At least 150,000 adults in Switzerland suffer from grade 2 or grade 3 obesity and are thus at elevated risk for adverse outcomes of COVID-19. This also includes many younger people in Switzerland.

It goes without saying that the causes of severe COVID-19 are only fragmentarily understood at this time ( Pairo-Castineira et al., 2020) and that obesity frequently coincides with other risk factors for severe COVID-19 such as male sex, older age, diabetes, elevated blood pressure, and other comorbidities. However, recent studies suggest, that the association between obesity and severe COVID-19 is independent from these other cofactors, and experts debate diverse pathways and mechanisms which link both pandemics. It should be mentioned that an increased risk for disease severity among grade 3 obese people is already well documented for influenza, especially in past pandemics such as the Swine flu of 2009 ( Morgan et al., 2010).

Among the possible explanations that link obesity with severe COVID-19 disease, the following pathways are currently mentioned in the literature: First, there are mechanical reasons which can worsen the prognosis in hospitals, as obesity is known to be associated with reduced lung volume/function, restricted airflow, and poor response to mechanical ventilation ( Caci et al., 2020; Wadman, 2020). Second, obesity can be related to other complications, such as renal failure, cardiovascular dysfunction, hypertension, blood clotting, and vascular damage, which in turn can further influence negative outcomes of COVID-19 ( Caci et al., 2020; Frühbeck et al., 2020). Third, the biological and physiological pathways probably include obesity-driven chronic low-grade inflammation and metabolically dysregulated immune response to infection, which may drive organ injury in the development of severe COVID-19 and impair viral clearance ( Caci et al., 2020; Chiappetta et al., 2020; Frühbeck et al., 2020; Huizinga et al., 2020; Smith et al., 2020; Wadman, 2020). However, other factors are currently proposed as well, such as links to fat embolism ( Cinti et al., 2020) or growth hormone insufficiency ( Lubrano et al., 2020). Finally, behavioral and social aspects could also play a role: People suffering from obesity may delay seeking medical care due to fear of being stigmatized which could increase their likelihood of severe COVID-19 ( Wadman, 2020). However, this would still have to be verified for Switzerland on the basis of hospitalization datas around the lockdown in spring 2020.

By October 2020, the Federal Office of Public Health declares that only people with grade 3 obesity (BMI>=40.0kg/m2, meaning weighting at least 126kg for an average tall man of 178cm or at least 111kg for an average tall women of 166cm) have an increased risk of severe COVID-19 ( Bundesamt für Gesundheit (BAG), 2020). However, this rather narrow definition of an obesity-related risk group probably needs to be extended based on recent evidence being published: In a large sample from Paris, all three grades of obesity doubled mortality in patients hospitalized with Covid‐19 ( Czernichow et al., 2020), and in another large UK sample there was an upward linear trend in the likelihood of COVID-19 hospitalization with increasing BMI, that was evident already in overweight people (BMI 25.0–29.9kg/m2) ( Hamer et al., 2020). The same pattern was documented in New York, where an increased risk of dying from COVID-19 was not only found in obese but also in overweight people with a BMI 25.0–29.9kg/m2 ( Nakeshbandi et al., 2020). In our studies of large samples of conscripts for the Swiss Armed Forces, i.e. data with a high coverage for young and mostly healthy men, we show that, on the one hand, inflammatory blood parameters (CRP, leukocytes, neutrophils and basophils) ( Staub et al., 2018) and, on the other hand, endurance performance (VO2max) ( Gassmann et al., 2020) do not suddenly increase or decrease only with obesity (BMI ≥30.0kg/m2), but gradually increase with increasing BMI, even within the overweight BMI range (25.0–29.9kg/m2).

In sum, based on international studies, there is increasing evidence that grade 1 obesity and overweight are also associated with adverse COVID-19 outcomes. If this is confirmed, then significantly more people in Switzerland (approximately 585,042 with grade 1 obesity and 2,094,531 with overweight) would be exposed to this increased risk. However, as long as most studies virtually dichotomize BMI (obese vs. non-obese) and do not model it continuously and even non-linearly, it will remain unclear whether many more people with lighter forms of obesity and overweight are also at elevated risk. It would also be important in this context to go beyond BMI and also investigate other anthropometric proxies for body fat distribution, as it has recently been shown that patients with visceral adiposity or high intramuscular fat deposition have a particularly higher risk for critical COVID-19 illness and should be monitored more carefully when hospitalized ( Yang et al., 2020).

Our study comes with limitations: First, the BMI is not an ideal measure of body composition because it cannot distinguish between muscle and fat in terms of weight. Nevertheless, at a population level, BMI is still strongly correlated with body fat ( Keys et al., 1972; Kit et al., 2014). Second, in four of the five studies included, information on height and weight was self-reported. In Switzerland, too, it has already been confirmed that men in particular report being taller than they are (especially with increasing age) and women, on the other hand, assess themselves less heavy ( Faeh et al., 2009; Faeh et al., 2008; Vinci et al., 2019a). In both cases this would mean rather under- than overestimation of overweight and obesity levels in our analysis. In the case of the menuCH data, we observed that the overall proportions for the BMI categories according to WHO differ only marginally when comparing the measured with the self-reported BMI values. We did therefore not adjust our analysis for this potential bias. Third, population-based health surveys can be subjected to a healthy and higher education participation bias, meaning that participants in such surveys are generally slightly healthier and better educated than the general population ( Bopp et al., 2014; Volken, 2013). The last two listed limitations, measured anthropometrics and participant bias might be among the reasons why in menuCH the proportions of overweight and obese participants was slightly higher than in the other four included surveys. Apart from the comparably small sample size and the relatively low participation rate (38%), one of menuCH's special features is that among the overweight (7.2%), and especially among the obese participants (14.1%), there are comparatively many people who have stated that they are on a weight-loss diet ( Vinci et al., 2019b). This might be related to the fact that there was increased interest in a nutrition survey among these participants, which led to a slight increase in the participation rate, especially among obese people. However, the applied sample weights do not adjust for this participation bias.

Because synthesis studies have been lacking until now, it was unclear how large the group of adults is in Switzerland, which has an increased risk of a severe COVID-19 course due to obesity. We have shown in this publication: This affects at least 150,000 adult people in Switzerland with grade 2 or 3 obesity. However, these numbers are 3.8 to 13.6 times higher if grade 1 obesity and overweight people are also included in this risk group, for which there are arguments arising in the latest literature. Relative and absolute numbers provide relevant but different perspectives. We think that especially in a pandemic, in which communication via numbers is so dominant, it is important not to lose the relation to the absolute numbers, especially when it concerns human lives.

The health emergency caused by the COVID-19 pandemic currently diverts attention from the prevention and care of non-communicable chronic diseases, such as obesity, to communicable diseases ( Dicker et al., 2020). It is important to focus increasingly on the interlinks between the two pandemics, because obviously in this case a communicable and a non-communicable pandemic are linked. Professional societies such as European Association for the Study of Obesity (EASO) are currently developing position statements and guidelines on the manifold challenges for obese people during this COVID-19 pandemic ( Frühbeck et al., 2020).

We think that two fields of action are particularly important: A) Current NCD prevention programs aimed at preventing weight gain in the general population should be continued or even intensified, despite the current focus on infectious diseases. B) The numerous people who also suffer from obesity in Switzerland should be given more attention and support. This applies not only to obese people who are infected with the virus, but to this numerous group in general, as stigmatization, lockdowns, stress, anxiety, or isolation have many consequences for health behaviors and well-being on various levels, especially in such vulnerable groups with preconditions ( Frühbeck et al., 2020). A particular challenge for such public health programs at various levels is certainly that overweight and obese people are not a homogenous group, as we have shown in our previous work ( Vinci et al., 2019b), and specific subgroups might need specifically tailored approaches. Regionally and socially differentiated public health strategies are very important in identifying these risk groups, especially if one has to react quickly and with limited resources. However, the effort might be worth it: If the current scientific state of knowledge becomes more and more solid and obesity really does play that major role in severe COVID-19 courses, then every kilo of body weight that is not gained in lockdowns or that is lost counts ( Beretta, 2020). It might even be doubly worth the effort, as obesity is known to be a risk factor for other factors such as cardiovascular diseases or diabetes, which are themselves (independently?) associated with increased risk of severe COVID-19 courses.