Optimizing messages and images for e-cigarette warnings

Abstract

Background.

The US Food and Drug Administration (FDA) requires e-cigarettes to have a single addiction warning, but many other health harms are associated with vaping and warnings grow stale over time. We aimed to develop new warning messages and images to discourage e-cigarette use.

Methods.

Participants were 1,629 US adults who vaped or smoked. We randomized them to evaluate 7 of 28 messages on newly developed warning themes (metals exposure, DNA mutation, cardiovascular problems, chemical exposure, lung damage, impaired immunity, addiction), and the current FDA-required warning (total of 8 messages). Then, participants evaluated images of hazards (e.g., metal), internal harms (e.g., organ damage), or people experiencing harms.

Results.

Regarding intended effects, new warning themes all discouraged vaping more than the current FDA-required warning (all p<.001), led to greater negative affect (all p<.001), and led to more anticipated social interactions (all p<.001). The most discouraging warnings were about toxic metals exposure. Regarding unintended effects, the new themes led to more stigma against people who vape (6 of 7 themes, p<.001) and led to a greater likelihood of thinking vaping is more harmful than smoking (all 7 themes, p<.001), though unintended effects were smaller than intended effects. Images of harms (internal or people experiencing) discouraged vaping more than images of hazards (all p<.001).

Discussion.

Vaping warning policies should communicate a broader range of hazards and harms, beyond addiction, to potentially increase awareness of health harms. Images of internal harm or people experiencing harms may be particularly effective at discouraging vaping.

Electronic cigarettes (e-cigarettes, also called vapes) are popular in the US, where an estimated 2.1 million youth in 2021 and 9.1 million adults in 2020 currently vaped [1,2]. Vaping exposes individuals to nicotine and other toxicants that lead to harms including addiction, respiratory damage, and cardiovascular problems, among others [3,4]. Identified health hazards associated with vaping include exposure to harmful chemicals (e.g., formaldehyde), heavy metals (e.g., lead), ultrafine particles, and carcinogenic volatile organic compounds [5–8] Also, emerging identified health harms include respiratory symptoms, hypertension, arterial stiffness, DNA damage, and inflammation [9–12].

Reducing all tobacco product use is an important public health goal [13]. E-cigarettes are not safe for people who do not smoke but may present lower risk than smoking, and providers may support people who smoke switching to e-cigarettes for patients who have been unsuccessful in quitting using other methods [14]. Adults who smoke cigarettes who are able to switch to vaping fully will not have exposure to many carcinogens produced during combustion [15–17]. However, e-cigarettes are a possible gateway to initiate cigarette smoking among people who do not smoke and have potential to thwart people who smoke from trying proven cessation tools [5,18,19]. Although long-term population effects of vaping are unknown, it is crucial to communicate growing evidence of vaping risks to discourage use. Ideally, e-cigarette warnings would reduce vaping among people who do not smoke without undermining motivations to quit smoking combustible cigarettes (as an unintended consequence) among people who smoke [20].

Warnings are an evidence-based policy to convey health risks induced by tobacco products and increase public awareness [21]. At least 40 countries require health warnings on e-cigarette packaging [22]. For example, the US Food and Drug Administration (FDA) now requires e-cigarette packages to have the standard text-only warning about nicotine addiction (“WARNING: This product contains nicotine. Nicotine is an addictive chemical”) and can require additional warnings. Previously, we found the current FDA nicotine addiction warning is likely to motivate people to quit vaping and discourage smoking, but extending warning themes to include other health hazards and harms is a clear next step to enhance the impact of text-only warnings [20]. As FDA’s single warning can “grow stale” or lose its effectiveness to discourage use over time [23,24], having newly developed warning messages based on present-day evidence makes warnings rotation possible that could help maximize the public health benefit of warnings [21].

In a 2021 report on the global tobacco epidemic, the World Health Organization emphasizes that “strong graphic health warnings should be mandated for all e-cigarette products, in line with overall tobacco control strategies to deter use by young people” [25]. However, only eight countries mandate the appearance of large graphic health warnings on e-cigarette packaging [25]. Research on the most promising kinds of images to include in warnings could inform the creation of new graphic e-cigarette warning policies.

Pictorial warnings (text-based messages paired with images) have consistently been shown to discourage tobacco use, including vaping [20,26–28]. Paired images provide visual evidence for the text that can increase the impact of text-only tobacco warnings through negative feeling and thinking about tobacco products [29,30]. Pictorial warnings showing internal harms (damage inside the body) or a person experiencing harms (individuals suffering from disease) often discourage smoking cigarettes more than symbolic (abstract) representations [26,30–32]. Visual representations of hazards (e.g., chemicals) on e-cigarette warnings have also led to greater discouragement of vaping [20]. No studies to our knowledge have investigated how the impact of these promising image types – hazards, internal harms, or a person experiencing harms – differ among novel warning themes for e-cigarettes.

To better inform e-cigarette warning policies for packaging and advertising, it is important to develop evidence-based e-cigarette warning statements and images that resonate with individuals who vape, smoke, or both. Therefore, in this experiment, we aimed to identify new warning themes, individual warning messages, and images for pictorial warnings to discourage vaping. We hypothesized that people would perceive novel warning themes as more effective for discouraging vaping than the FDA-required warning. We also explored whether novel warning themes influence known mechanisms for impact from the Tobacco Warnings Model [30], according to which warnings lead to negative emotional reactions (affect) and social interactions to reduce use. We also explored the impact of warnings on the potential unintended consequences of increased stigma and incorrect harm perceptions (i.e., that vaping is more harmful than smoking) [33].

Methods

Participants

Participants were a national convenience sample of 1,629 US adults recruited through the Qualtrics survey panel. Participants were people who vape (defined as currently vaping every day or some days or ever vaping) but did not smoke, people who smoke (defined as having smoked at least 100 cigarettes and now smoking every day or some days) who did not use e-cigarettes, or people who vape and smoke (defined as a current e-cigarette use and current smoking). For experiments, convenience samples typically yield the same pattern of experimental impact as population-based samples [34]. Participants completed two experiments in Spring 2021 that were embedded in a single online survey.

Procedures and measures: Message experiment

The first experiment sought to identify the most promising e-cigarette warning themes and messages (Figure 1).

Figure 1.

Study design

Stimuli.

We identified warning themes based on relevant scientific literature on the hazards and harms associated with vaping (Figure 2) [5–8,10–12,25,35,36]. After reviewing the strength of existing evidence and feedback from an external advisory board, we selected seven themes: metals exposure, DNA mutation, cardiovascular problems, chemical exposure, lung damage, impaired immunity, and addiction. For each theme, we created four warning messages (28 messages). We added the FDA-required addiction warning as an eighth theme (control). We divided the messages into message sets with one message from each theme (eight messages per set) so participants could review a set. The sets were similar in word and character counts and reading levels.

Figure 2.

Messages by warning theme

Procedures.

We randomized participants to one of the four warning message sets and evaluated each of the eight messages in the set. Each warning message appeared on its own at the top of a new screen in a random order, followed by survey items for perceived message effectiveness, negative affect, social interactions, stigma, and perceived harm items described below.

Measures.

The main outcome was perceived message effectiveness (PME), assessed with an item from the UNC PME scale: “How much does this message discourage you from wanting to vape” This item has the strongest psychometric properties of the three items in the scale [37]. Secondary outcomes were negative affect, a negative emotional response that explains warning impact, assessed with a validated item [38,39], “How much does this message make you feel afraid” and stigma assessed with an adapted item for internalized bias [40], “How much will this message make people look down on someone who vapes.” The 5-point response scale for these measures ranged from “not at all” (coded as 1) to “a great deal” (5). Another secondary outcome that explains warning impact was anticipated social interactions [30], “How likely are you to talk about this message with others in the next week” with a 5-point response scale that ranged from “not at all likely” (1) to “extremely likely” (5). The last secondary outcome was perceived harm, “Does this message make you think vaping is less harmful, about the same, or more harmful than smoking cigarettes.” We recoded participants who incorrectly responded there was higher perceived harm (1; i.e., vaping was perceived as “more harmful” than smoking cigarettes) vs. not (0; i.e., responses of “less harmful,” “about the same,” or “not sure”), adapted from the Population Assessment of Tobacco and Health study [41].

Procedures and measures: Image experiment

The second experiment sought to identify the most promising images for each theme.

Stimuli.

We identified images from commercial stock photography, specialized medical and government open access sites, and created images by photographing or compositing existing images. For each of the eight warning themes, we selected images to illustrate image type: hazard (e.g., metals), internal harm (e.g., diseased organ), and a person experiencing harms (e.g., suffering from symptoms). All images of a person experiencing harms included a person’s face, while internal harms only showed body parts or organs. If an image type was not applicable (e.g., no hazard other than vaping in general), we included additional images for the internal or a person experiencing harm categories (see Figure 3 for images by panel).

Figure 3.

Image types by warning theme

Note. Images categories with two options are designated as image “A” and image “B” in analyses.

Procedures.

Participants examined two to four images for each warning theme in a panel; we divided the themes into two panels to reduce burden on participants. Images appeared in a random order with a randomly selected message that the participant previously saw from that theme. Participants evaluated the images for each theme as described below.

Measures.

Participants selected the image in each theme that “best represents this message” and “most discourages you from wanting to vape.” They then evaluated each of the 2–4 images for that warning theme on PME, “How much does this image discourage you from wanting to vape,” with a 5-point response scale that ranged from “not at all” (1) to “a great deal” (5).

Data Analysis

We examined descriptive statistics for warning themes, messages, and image outcomes. To identify the most promising warning message themes, we estimated linear mixed models for each outcome (PME, negative affect, social interactions, and stigma) with warning theme as the predictor. In each model, the FDA warning theme (control) was the reference group. We included a 3-level categorical main effect of smoking status subgroup in an exploratory mixed model along with warning theme to explore whether PME differed among participants who vape, smoke, or both (dual use). To examine the impact of message theme on higher perceived harm, we estimated a logistic mixed model with warning theme as the predictor (and FDA warning as the reference group). To identify promising images, we conducted linear mixed models with image type as the predictor and PME as the outcome. Internal harm images were the reference group in these models.

Results

Participants’ mean age was 45 years (Table 1). Most were white (87%), and 32% were college graduates. Participants were people who smoke (43%), people who vape and smoke (40%), and people who vape (15%).

Table 1.

Participant demographics (n = 1,629)

	n (%)
Age (M = 45.70, SD = 15.43)
Young adults (18–25)	142 (9%)
Adults (26 or older)	1487 (91%)
Gender
Female	903 (55%)
Male	716 (44%)
Nonbinary or other response	10 (<1%)
Transgender
No	1579 (97%)
Yes	42 (3%)
Race
White	1430 (88%)
Black or African American	133 (8%)
Other	130 (8%)
Hispanic Ethnicity
No	1509 (93%)
Yes	118 (7%)
Education
Less than high school	81 (5%)
High school diploma	428 (26%)
Some college	417 (26%)
Associate’s degree	181 (11%)
Bachelor’s degree	280 (17%)
Graduate or professional degree	242 (15%)
Tobacco use status
Currently smoke	721 (43%)
Currently vape and smoke	661 (41%)
Currently vape only	182 (11%)
Ever vaped only	65 (4%)
Current use of other tobacco products
Traditional cigars	276 (17%)
Cigarillos, filtered cigars, or little cigars	441 (27%)
Pipe filled with tobacco	134 (8%)
Hookah	226 (14%)
Smokeless tobacco	158 (10%)
Income (annual)
Less than $10,000	158 (10%)
$10,000 to $14,999	122 (8%)
$15,000 to $24,999	202 (12%)
$25,000 to $34,999	218 (13%)
$35,000 to $49,999	193 (12%)
$50,000 to $74,999	228 (14%)
$75,000 to $99,999	180 (11%)
$100,000 to $149,999	224 (14%)
$150,000 or more	103 (7%)

Warning Messages

Intended effects.

The perceived message effectiveness (PME), to discourage vaping, of the warning message themes ranged from a mean of 2.68 (SD=1.43) for the FDA warning to 3.41 (SD=1.35) for the novel metal exposure theme. Other novel warning themes discouraged vaping with mean ratings of 3.32 (SD=1.39) for DNA mutation, 3.29 (SD=1.35) for cardiovascular problems, 3.26 (SD=1.37) for chemical exposure, 3.21 (SD=1.36) for lung damage, 3.18 (SD=1.40) for impaired immunity, and 2.88 (SD=1.42) for new messages about addiction. The new warning message themes all discouraged vaping more than the FDA-required warning (all p<.001, Table 2). We observed a similar pattern of results for PME of warning themes across tobacco use status subgroups (Appendix A). Participants who only smoked cigarettes saw the warnings as more discouraging overall compared to people who only vaped (F [2, 1626] = 30.20, p < .001). People who vaped and smoked’ average PME ratings did not differ from respondents who only vaped.

Table 2.

Impact of warning message theme (n=1,629)

	Intended effects			Unintended effects
Warning message theme	Perceived message effectiveness M (SD)	Negative affect M (SD)	Social interaction M (SD)	Stigma M (SD)	Higher perceived harm %	OR (95% CI)
Metals exposure	3.41 (1.35)**	3.21 (1.36)**	2.70 (1.44)**	2.87 (1.37)**	43%	6.61 (5.26–8.31)
DNA mutation	3.32 (1.39)**	3.14 (1.37)**	2.68 (1.46)**	2.86 (1.36)**	39%	4.29 (3.42–5.38)
Cardiovascular problems	3.29 (1.35)**	3.12 (1.36)**	2.62 (1.44)**	2.79 (1.36)**	35%	3.11 (2.48–3.90)
Chemical exposure	3.26 (1.37)**	3.06 (1.36)**	2.61 (1.43)**	2.79 (1.36)**	35%	2.96 (2.36–3.71)
Lung damage	3.21 (1.36)**	3.01 (1.38)**	2.57 (1.45)**	2.78 (1.35)**	32%	2.31 (1.84–2.89)
Impaired immunity	3.18 (1.40)**	2.98 (1.38)**	2.57 (1.43)**	2.74 (1.36)**	33%	2.51 (2.00–3.15)
Addiction	2.88 (1.42)**	2.66 (1.41)**	2.42 (1.41)**	2.59 (1.33)	28%	1.47 (1.17–1.84)
FDA-required warning (ref group)	2.68 (1.43)	2.52 (1.39)	2.31 (1.40)	2.57 (1.32)	24%

Note. % for higher perceived harm reflects percent of participants who incorrectly responded that vaping was “more harmful” than smoking cigarettes. The overall F-tests were PME: F(7, 11,397)=159.73; negative affect: F(7, 11,398)=176.76; social interactions: F(7, 11,398)=58.41; stigma F(7, 11,396)=38.75; harms F(7, 11,391)=50.2; all p<.001.

^*p<.05,

^**p<.001

The two most discouraging messages, descriptively, from the metal exposure theme with novel hazards and harms were: “E-cigarette vapor may contain metals that are inhaled deep into the lungs,” and “Vaping can put lead, nickel, and chromium in your lungs” among messages that were similarly discouraging (Appendix A). Another highly discouraging message from the DNA mutation theme was: “E-cigarette vapor can have formaldehyde, a chemical that causes DNA damage.” The most discouraging message from the chemical exposure theme was: “E-cigarette vapor can contain acrolein, which can cause irreversible lung damage.” New warning messages about addiction (e.g., withdrawal, cravings) were descriptively the least discouraging among the novel messages. Within each theme, most novel messages were rated as similarly high without statistical differences for discouraging vaping (Appendix B).

The pattern of results for warning message themes was similar for negative affect and anticipated social interactions (Table 2). The new warning themes led to more negative affect than the FDA-required warning (all p<.001). The new themes also led to more anticipated social interactions than the FDA-required warning (all p<.001).

Unintended effects.

Participants rated the new warning themes (except addiction) as more stigmatizing for those who vape, compared to the FDA-mandated warning (all p<.001). The new warning themes also led to a greater likelihood of incorrectly believing vaping was more harmful than smoking (higher perceived harm), as compared to the FDA-mandated warning: ranging from metals exposure (43% vs. 24%, p=.01) to our new addiction theme (28% vs. 24%). Data on the impact of individual messages on negative affect, social interactions, stigma, and perceived harm appear in Appendix B.

Warning Images

Images showing internal harms (e.g., diseased organs) or a person experiencing the harms were perceived as the most discouraging image for the warning message themes (all p<.001, except for lung damage). Data on which image participants thought most discouraged them followed the same pattern of results as the mean discouragement ratings of the images (Table 3). Internal harm images were selected as most discouraging for 4 of 8 warning message themes: metals exposure (52% selected showing metals in the lungs), chemical exposure (58% selected an x-ray with lung spots), cardiovascular damage (54% selected heart surgery), and the FDA-required warning (50% selected a brain). Images of a person experiencing the harms were most discouraging for 3 of 8 warning message themes: DNA mutation (70% selected a cancer patient), impaired immunity (71% selected a patient with oxygen mask), and addiction (42% selected a person with a fishhook in their lip). Participants selected internal harm and a person experiencing harm images as equally discouraging for one theme: lung damage (52% selected damaged lungs outside the body vs, 48% selected a person with oxygen mask). Participants selection for which image best represented the message followed the same pattern.

Table 3.

Evaluations of message image type (n=1,629)

Warning theme and image type	Perceived message effectiveness M (SD)	Most discourages % selected	Best represents % selected
Metals exposure
Hazard	2.92 (1.44)**	14	13
Internal harm (ref group)	3.68 (1.24)	52	58
Person experiencing harm	3.42 (1.28)**	33	29
DNA mutation
Hazard A	2.71 (1.41)**	8	12
Hazard B	2.86 (1.33)**	7	10
Internal harm (ref group)	3.07 (1.36)	15	37
Person experiencing harm	3.87 (1.28) **	70	41
Cardiovascular problems
Internal harm A	3.35 (1.29)**	13	28
Internal harm B (ref group)	3.77 (1.27)	54	25
Person experiencing harm	3.70 (1.24)	33	47
Chemical exposure
Hazard A	2.81 (1.44)**	9	13
Hazard B	3.05 (1.36)**	10	19
Internal harm (ref group)	3.90 (1.18)	58	40
Person experiencing harm	3.64 (1.25)**	24	28
Lung damage
Internal harm (ref group)	3.53 (1.31)	52	48
Person experiencing harm	3.51 (1.26)	48	52
Impaired immunity
Internal harm (ref group)	3.31 (1.28)	29	44
Person experiencing harm	3.82 (1.20) **	71	56
Addiction
Hazard	2.80 (1.37)**	11	11
Internal harm (ref group)	3.12 (1.36)	21	20
Person experiencing harm A	3.19 (1.36)	26	49
Person experiencing harm B	3.39 (1.41) **	42	21
FDA-required warning
Hazard	2.84 (1.41)**	23	22
Internal harm (ref group)	3.23 (1.36)	50	33
Person experiencing harm	3.04 (1.34)**	27	45

Note. Bolded text shows the highest PME in each theme. The overall F-tests were metals exposure: F(2, 1636)=178.30, p<.001; DNA mutation: F(3, 2451)= 251.04, p<.001; cardiovascular problems: F(2, 1610)=75.89, p<.001; chemical exposure: F(3, 2416)=275.06, p<.001; lung damage: F(1, 819)=.10; p=.76; impaired immunity: F(1, 806)=169.77, p<.001; addiction: F(3, 2454)=61.63, p<.001; FDA-required warning: F(2, 1609)=40.13, p<.001.

^**p<.001

Discussion

Warnings are a promising communication strategy for educating the public about the health hazards and harms from vaping. E-cigarettes typically contain nicotine and chemicals that are known to damage respiratory and cardiovascular systems, and induce other harms [3,4,8,42,43]. While the FDA-required warning informs people about the risks of nicotine as an addictive chemical, there are no required warnings about other known toxins and health harms from e-cigarettes that may be more effective.

It is important to explore new e-cigarette warning themes and rotate them to 1) communicate critical risk information and 2) to decrease the likelihood that warnings grow stale or wear out over time [24]. We developed e-cigarette warnings for seven novel themes: metals exposure, DNA mutation, cardiovascular problems, chemical exposure, lung damage, impaired immunity, and life disruptions from addiction. We found them all to discourage vaping more than the current FDA-required warning, along with leading to greater negative affect and anticipated social interactions – two key mechanisms underlying the impact of warnings in the Tobacco Warnings Model [30]. Novel warning themes discouraged vaping more than the FDA-required warning among those who vape, smoke, or both (dual use), a promising finding for efforts to reduce all tobacco use. Consistent with main effects of tobacco warnings and campaigns where people who currently use tobacco have lower intended reactions (e.g., PME) than people who do not use tobacco [44], our warnings were more discouraging among people who smoke but did not vape than those who only vape and could be beneficial as we encourage people who smoke to quit smoking cigarettes and then to quit vaping. Future studies could examine the impact of vaping warnings on tobacco use behavior, including by subgroup of tobacco use (e.g., people who only vape versus people who both vape and smoke cigarettes) and test these warnings among adolescents.

Our findings support e-cigarette warning messages that warn consumers about hazards and health harms beyond nicotine addiction. Within each of our novel themes, most of the messages were rated highly and did not differ from one another. Promising warning messages communicated about risks of metals exposure (“E-cigarette vapor may contain metals that are inhaled deep into the lungs”), DNA mutations (“E-cigarette vapor can have formaldehyde, a chemical that causes DNA damage”), or chemical exposure (“E-cigarette vapor can contain acrolein, which can cause irreversible lung damage”), among others. Based on descriptive ratings, warning messages about addiction withdrawal (“Vaping withdrawal can make you moody and unfocused”) were less promising. Our previous work suggested hazard warnings do not need to include harms to effectively discourage vaping [20]; however, the inclusion of harms does not reduce the perceived effectiveness of the warnings.

Our findings align with other warning research showing novel themes about e-cigarette hazards and harms (e.g., chemical exposure, cancer) may discourage vaping more than nicotine addiction messages [45–47]. This growing body of evidence suggests we have a large pool of viable messages about hazard, harms, or both to be used for e-cigarette warnings and public education campaigns [48,49]. These messages could be rotated over time to maximize their communication and public health impact.

Implementing novel warning themes may have potential unintended effects that should be considered. Our findings indicate that vaping warnings about novel hazards and harms may also somewhat increase stigma toward people who vape, compared to the FDA-required warning. Stigma may undermine our public health goals to reduce tobacco use. Smoking-related stigma may encourage more quit attempts, but it is associated with less successful quitting [33]. Additionally, warnings about novel hazards and harms may confuse the public about the relative risks of vaping (vs. smoking). Most public health experts agree that while not safe, vaping is less harmful than smoking combustible cigarettes. However, most US adults have misperceptions that vaping is as harmful or more harmful than smoking cigarettes [50,51]. It is important to note that we tested e-cigarette warnings alone and not in the context of cigarette warnings, which could have heightened the impact of e-cigarette warnings on this outcome. Future research should examine potential unintended behavioral effects of vaping warnings, such as increased smoking, as well as the impact of e-cigarette warnings on inaccurate risk perceptions, in the context of strong pictorial cigarette pack warnings.

Warning images showing damage to the body (internal) from vaping or experiencing symptoms or the harm were consistently the most promising images to discourage vaping. These images performed better than images of hazards (e.g., showing metals alone). Showing the harm done is important when communicating about health risks, as images are perceived as visual evidence [52–54]. As in other e-cigarette and tobacco pictorial warnings research [55–58], two types of images best represent warning messages and discourage vaping – graphic internal harms and external suffering from harms. Showing internal damage to the body through graphic medical illustrations, x-rays, or images of an open body during surgery is a promising way to visualize vaping warnings [59]. Showing an external view of someone suffering from or experiencing symptoms of a harm is also promising for vaping warning messages [55]. In all instances of our external images, we included the face of the person as well as their human form, known to increase visual processing resources [60,61]. While the future of pictorial warnings on tobacco products in the US remains unclear, images of damage inside the body or individuals suffering can illustrate harms in media campaigns via print, broadcast, and on social media to discourage vaping.

Showing hazards alone, without internal or external harm, is not recommended to discourage vaping. It did not matter if the hazards were shown concretely (e.g., metal scraps, jars of formaldehyde) or as symbolic representations (e.g., e-cigarette with fishhook), images of hazards were consistently the least discouraging and representative of the warning message. Notably, as with our other image selection research [62], we found a simple task asking participants to pick the most discouraging image to be an efficient and effective method determining promising images.

The study’s use of a convenience sample of individuals who use e-cigarettes, smoke, or both potentially limits the generalizability of findings to other populations. However, the comparative ratings may hold similarly in other samples, as we have found that convenience samples and probability-based samples often yield similar patterns of experimental findings [34]. The findings are also limited to the text for themes and images used. Different images representing the categories of hazards, internal harm, or a person experiencing harms may yield different results. Finally, given that this study focused on message development and refinement, we did not evaluate the impact of messages and images on behavioral outcomes.

Conclusion

E-cigarette warnings can be an important tool to discourage vaping among people who currently vape, smoke, or use both products. Our study suggests novel warnings should contain text-based information about harms and hazards, including metals exposure, DNA mutation, cardiovascular problems, chemical exposure, lung damage, impaired immunity, and life disruptions from addiction. Pairing text warnings with images showing internal harms (e.g., diseased organs) or someone experiencing harms (e.g., suffering symptoms) will likely discourage vaping more than showing the hazards (e.g., chemicals). These findings can inform decision making about e-cigarette labeling and packaging.

What is already known on this topic

Warnings are effective for reducing cigarette smoking and other unhealthy behaviors. However, the impact of e-cigarette warnings remains uncertain.

What this study adds

In a national study of US adults, e-cigarette warnings were effective in discouraging vaping and led to greater negative affect and anticipated social interactions. Unintended effects included greater stigma against people who vape and thinking vaping is more harmful than smoking, though intended effects were larger than unintended effects. Images of harms more effectively discouraged vaping than did images of hazards.

How this study might affect research, practice or policy

The US currently requires one warning which will become stale and ineffective. This study identifies the following topics as promising for new warnings: metals exposure, DNA mutation, cardiovascular problems, chemical exposure, lung damage, impaired immunity, and new messages about addiction. This studies also identifies promising image types for pictorial warnings: showing internal harms or people experiencing harms.