Smokeless tobacco has long been proposed as a low-risk alternative to cigarettes. Recently, the U.S. Food and Drug Administration (FDA) concluded that completely switching from smoking cigarettes to using certain snus products lowers health risks . As an implication, a specific brand of moist snuff (hereafter: “snus,” the Swedish-language word for the category) may now be sold in the U.S. with a modified risk claim that compares snus use to smoking: “Using General Snus instead of cigarettes puts you at a lower risk of mouth cancer, heart disease, lung cancer, stroke, emphysema, and chronic bronchitis.”
The public tends to exaggerate risks from using snus compared to risks from smoking cigarettes. Wackowski et al.  found that 74.6% of the U.S. smokers in their sample perceived snus to be as harmful or more harmful than smoking cigarettes. Even in Norway, where the number of daily snus users exceeded the number of daily smokers in 2017, the perceptions among lay people does not appear to reflect the different risk profiles of snus and cigarette smoking. In a study using survey data from Norwegians aged 16–79 in the years 2003–2018, the harmfulness of daily snus use was rated only somewhat lower than daily smoking. These perceptions did not appear to have changed during the last 16 years . As for specific diseases, another study found that the majority of Norwegian smokers estimated that the risk for cardiovascular disease, oral and stomach cancer was equal or higher for snus users .
One way of communicating the risk of tobacco to users and potential users is through Health Warning Labels (HWL) on the product itself. HWLs are short statements about health risks, or pictures illustrating severe consequences associated with using a product, that are applied on products to communicate. HWLs are a part of the tobacco control measures described in the World Health Organizations Framework Convention on Tobacco Control , which was developed to reduce the harms from tobacco. Cigarette packages are subject to more restrictive regulations compared to smokeless tobacco products (SLT), (e.g., to carry larger, graphic HWLs), which may reduce initiation and increase smoking cessation more than textual HWLs [6, 7].
Tobacco regulations for SLT products typically include textual warnings . An HWL regarding risk of cancer, “This tobacco product severely damages your health and is addictive. Causes cancer” , was removed from snus products sold within the European Union (EU) in 2003 (snus sale was only allowed in Sweden at that time), and replaced with a more general warning: “This tobacco product can damage your health and is addictive” . The HWL was modified slightly in 2016, when the modal verb “can” was removed : “This tobacco product damages your health and is addictive.” This new statement was expected to strengthen the risk message, but the effect may have been minor .
Given the reduced health risks from SLT compared to cigarettes and their potential as a harm reduced alternative to smoking, it has been argued that product information should reflect SLTs risks in comparison to smoking . Alternatives to cigarettes, such as snus, may be perceived as more favorable if the HWLs concern health risks relative to the risks of smoking, because cigarettes are a product that most people know is very harmful . It is therefore interesting from a harm reduction perspective to understand how such comparative HWLs affect risk perception and behavior [13,14,15].
To our knowledge, only a few studies have exposed participants to comparative HWLs on SLT products or models of such products. Note that messages comparing risks are typically referred to as “health warning labels” because of the shared format with standard HWLs, although the information strictly speaking concerns comparative information. In an online experiment, Canadian smokers exposed to the HWL “Using ST is less harmful than smoking cigarettes” more often reported perceptions of risk that reflected the large difference in risks between SLT and smoking than those who were exposed to regular HWLs . The youths who were exposed to the comparative HWL also had a higher likelihood of reporting future use of SLT, and a higher willingness to try SLT as a cessation aid. The comparative HWL was designed for experimental purposes and is not applied on SLT packages in Canada.
In a study sponsored by the snus manufacturer Swedish Match, participants were exposed to one of four current US HWLs, and two proposed comparative risk HWLs: “No tobacco product is safe, but this product presents lower risks to health than cigarettes”, and one substituting the word “lower” with “substantially lower” . These two were compared to an HWL stating that snus is “not a safe alternative to smoking”. Smokers who saw the two comparative HWLs perceived daily use of snus as less harmful and reported that they were more likely to use and buy snus. For people who had tried or never used snus, seeing “substantially lower risk” was associated with lower risk perception of snus and the reporting of higher likelihood of buying snus. Another study assessed the effects of comparative risk information on actual snus use, although as a one-time provision of more extensive information instead of exposure to a brief HWL . In this randomized trial, nicotine lozenge, snus, or snus combined with comparative risk information were tested as a means for smoking cessation, and cessation rates were found to be similar for all groups, with under 1.5% of the group participants staying abstinent for the whole year. The amount of snus used during cessation did not differ between the snus-only and the snus + comparative risk information group.
The studies above indicate that comparative HWLs can decrease the perception of risk from STLs relative to cigarettes. However, the studies represent only a limited number of contexts and only a few types of comparative health warnings, all of which are very general statements. The present study aimed to conceptually replicate the finding that general comparative health warnings can affect risk perceptions in a country with high prevalence of snus use, and to extend the research by investigating the impact of specific information about the relative risk. Specific versus general information have been found to be a relevant dimension in persuasion an advertisement research [19, 20].
In the present case, a general statement about less risk could mean slightly less risk, whereas a quantitative statement can better convey the magnitude of the difference in risk. In addition to the above, we were interested in whether a general non-comparative HWL, such as the one implemented in the EU, can distort the relative perceptions of risk from snus and smoking. That is, in comparison to a general statement regarding risk on snus packages, no information regarding risk may give perceptions that better reflect the large differences in risks between snus and smoking. A more detailed specification of our hypotheses follows.
Hypotheses 1a and b (H1a and H1b)
The current non-comparative EU HWL states that snus “damages your health and is addictive,” a statement that is likely to produce an overestimation of the risks from snus use. We hypothesized that seeing a snus product with an HWL based on the current EU HWL would increase risk perception of snus use compared to seeing a control condition with no HWL (H1a), and that the relative perception of risk from snus use versus smoking would become lower in favor of snus use after exposure to the EU HWL compared to the control condition (H1b). Note that we chose to remove the “[…] and is addictive” statement from the EU HWL to make it similar to the below comparative risk HWLs.
Hypotheses 2a and b (H2a and H2b)
Our first comparative HWL is a general claim that “Snus is less damaging to your health than smoking”, which we expected to adjust perceptions to reflect the differential health risks of snus and cigarettes more than would exposure to the EU-based HWL. Specifically, when compared to the EU-based HWL, this general comparative risk HWL (General CR) was expected to decrease perception of risk from snus (H2a), and to increase the difference between perceptions of risk from snus and smoking, such that the relative ratings are more in favor of snus use (H2b). Note that an increase in the (absolute) difference between snus use and smoking is reflected in a decrease in our relative score (snus risk minus smoking risk), lower scores indicating relatively less risk for snus use compared with smoking.
Hypotheses 3a and b (H3a and H3b)
The second comparative HWL specifies that “Snus is 90% less damaging to your health than smoking.” This was based on a conservative estimate reported in a study by Levy et al. . As hypothesized for the General CR condition, we expected risk perception of snus to decrease more in this percentage comparative risk (percentage CR) HWL condition than for the EU-based HWL (H3a), and that the difference in comparative risk would change more in favor of snus use in the percentage CR condition than in the EU-based HWL condition (H3b).
Hypotheses 4a and b (H4a and H4b)
As the public tend to have unrealistically high estimates of risks from snus (3), the 90% difference in the percentage CR would likely be larger than most people would expect, and may, therefore, have a greater impact on perceptions than the general CR. Thus, the Percentage CR was expected to decrease risk ratings of snus more than the general CR (H4a), and to produce a stronger change in the relative perception of risk (H4b). For the sake of completeness, we decided a priori to include the analyses of Hypothesis 4b, but this hypothesis was not preregistered.
As intentions tend to be hard to affect  and our main focus was risk perception, we explored the effect of the HWLs on intentions to use tobacco and did not include these measures in the hypotheses or in the power calculation.