Effects of advertising features on smokers’ and non-smokers’ perceptions of a reduced nicotine cigarette modified risk tobacco product

Abstract

Introduction:

Research is needed to determine the impact of marketing on perceptions and use of reduced nicotine content (RNC) cigarettes, particularly as U.S. regulators have permitted the sale of a RNC cigarette modified risk tobacco product (MRTP) that seeks further authorization to advertise using modified risk claims. This study examined the effects of two advertising elements (product name and disclaimer content) on perceptions of a RNC cigarette MRTP.

Methods:

Adult participants (n = 807, 28.7% smokers, 58.2% male, 74.2% Non-Latinx White) completed an online MTurk survey. Participants were randomized to view one of six RNC cigarette advertisements, using a 2 × 3 between-subject factorial design to manipulate product name (‘Moonlight’ vs. ‘Moonrise’) and disclaimer content (industry-proposed: “Nicotine is addictive. Less nicotine does NOT mean a safer cigarette” vs. focused: “Less nicotine does NOT mean a safer cigarette” vs. no content), then completed recall and product perception questionnaires.

Results:

All participants who viewed the industry-proposed disclaimer (vs. no content) perceived greater addiction risk (p’s < .05). Non-smokers who viewed this disclaimer also perceived greater health risks and held fewer false beliefs (p’s < .05). Smokers who viewed Moonlight (vs. Moonrise) ads perceived lower health risks (p < .05).

Conclusions:

Disclaimer content may effectively inform consumers about addiction risk of a new RNC cigarette MRTP, and further inform non-smokers about health risks. This element, however, had little effect on perceived health risks among smokers, among whom the Moonlight product name was associated with health risk misperceptions similar to the banned “light” descriptor.

INTRODUCTION

The 2009 Family Smoking Prevention and Tobacco Control Act authorized the U.S. Food and Drug Administration (FDA) to regulate tobacco products, including nicotine content and marketing, provided that its actions demonstrate empirical public health benefits.¹ Accumulating evidence^2–6 supports reducing the nicotine content in combustible cigarettes to non-addictive levels as a strategy for decreasing tobacco use and disease at the population level.^7,8 However, research has not yet explored if potentially misleading marketing could dampen the public health impact of a reduced nicotine product standard.

The tobacco industry has long exploited marketing to alter consumer perceptions and maintain product sales.^9–11 Previously labeled “light” cigarettes were promoted implicitly through their labeling, advertising, and light-colored packaging as safer alternatives to other commercial cigarettes,^9,12 despite no evidence that these products actually reduced harm.^13–15 Though a reduced nicotine product standard would likely produce population benefits,¹⁶ reduced nicotine content (RNC) cigarettes similarly are not less harmful than conventional cigarettes. Yet research suggests that consumers perceive them as safer,^17–19 which is a concern for ensuring that a RNC standard benefits public health, as smokers may be less willing to quit or may increase their use of perceived “safer” cigarettes, while non-smokers may be more likely to initiate use. How much misleading marketing could further exacerbate these concerns is unknown.

While the FDA deliberates regulating nicotine content, the tobacco product manufacturer 22^nd Century received FDA permission to sell menthol and non-menthol versions of its lowest RNC cigarette brand in December 2019,²⁰ and currently awaits further authorization to advertise these products with modified risk claims.²¹ Despite FDA prohibition of misleading descriptors (e.g., “light,” “low-tar”) in cigarette marketing, 22^nd Century chose “Moonlight” as the RNC cigarette brand name and uses other marketing conventions reminiscent of “light” cigarettes (e.g., white packages).²² Advertisements proposed in the product’s marketing authorization application also include a voluntary disclaimer stating, “Nicotine is addictive. Less nicotine does NOT mean a safer cigarette. All cigarettes can cause disease and death” to warn consumers of product risks, although this content appears in small font size, is not emphasized, and is less relevant than when used on other tobacco product warnings because RNC cigarettes contain little nicotine. Examining how these advertising features affect consumer perceptions may anticipate how marketing could more broadly affect perceptions of RNC cigarettes in the event of a reduced nicotine product standard.

This study examined how the product name and disclaimer content in a RNC cigarette advertisement may influence adults’ product perceptions. We hypothesized that, consistent with prior research observing that descriptors like “light”, “mild”, and “natural” contribute to misunderstanding tobacco product risks,^23–25 the “Moonlight” name would be associated with lower risk perceptions, more favorable attitudes, and greater intentions to try the product compared to a mock brand name of similar length and meaning but without the banned descriptor (“Moonrise”). We further hypothesized that compared to a concise or “focused” disclaimer lacking the uninformative “Nicotine is addictive” statement (“Less nicotine does NOT mean a safer cigarette. All cigarettes can cause disease and death.”), the industry-proposed voluntary disclaimer and no disclaimer would be associated with lower risk perceptions, more positive attitudes toward, and greater intentions to use the product.

METHODS

Participants and study design

We recruited a national convenience sample through the Amazon Mechanical Turk (MTurk) crowdsourcing platform^26–28 to complete a one-time online survey. The survey utilized a 2 × 3 between-subjects design to manipulate two advertisement elements: (1) product name (2-level factor: Moonlight vs. Moonrise), and (2) disclaimer content (3-level factor: no content vs. industry-proposed content vs. focused content). Those eligible were ≥ 18 years of age, registered on MTurk with a U.S. location, had a Human Intelligence Task (HIT) approval rating ≥ 99%, and had completed ≥ 1000 HITs.

The study was described on MTurk as a “survey about health behaviors and product marketing” to discourage dishonest reporting of tobacco use behaviors and perceptions. Eligible participants followed a weblink to the informed consent document, which stated that the study used deception and would debrief and compensate participants only after they completed the full survey. Those who consented to participate completed an unpaid screener survey to determine smoking status based on self-reported past 30-day cigarette use; we categorized those who reported smoking on 0 days as “non-smokers” and ≥ 1 day as “smokers.” To conceal the study’s purpose from participants, the screener assessed past 30-day use of various other health behaviors: e.g., wearing seatbelts, using marijuana, brushing teeth.

Participants next completed questionnaires assessing demographic information, tobacco use history, and risk perceptions of conventional cigarettes. They were randomized to view one of six advertisement conditions (Figure 1) for 30 seconds,²⁹ then completed recall, risk perceptions, attitudes, and use intentions questionnaires regarding the advertised product. Afterward, participants received a $1.50 incentive/compensation via MTurk and were debriefed regarding the study’s purpose. The University of Pennsylvania Institutional Review Board approved all procedures.

Figure 1.

Study advertisement conditions. Clockwise from top left shows: 1) Moonlight product name, industry-proposed disclaimer; 2) Moonlight product name, focused disclaimer; 3) Moonlight product name, no disclaimer; 4) Moonrise product name, industry-proposed disclaimer; 5) Moonrise product name, focused disclaimer; 6) Moonrise product name, no disclaimer.

Advertisement conditions

We modified an advertisement included with 22^nd Century’s application to market their product as reduced risk,²¹ selected specifically because it did not overtly target consumers by demographic (e.g., age) or cigarette preference (e.g., menthol).

Product name.

Advertisements contained either the industry-proposed Moonlight product name or Moonrise.

Disclaimer content.

Advertisements contained one of three disclaimer types: (1) industry-proposed content: “Nicotine is addictive. Less nicotine does NOT mean a safer cigarette. All cigarettes can cause disease and death,” (2) focused content: “Less nicotine does NOT mean a safer cigarette. All cigarettes can cause disease and death,” and (3) no content. The focused content excluded the “nicotine is addictive” portion of the industry-proposed content to directly and concisely inform smokers of the product’s harms, as research indicates that many smokers are confused about nicotine’s role in the health consequences of smoking (i.e., many know nicotine is addictive but also think nicotine is the cancer-causing agent in cigarettes), particularly regarding RNC cigarettes.^30,31 The Surgeon General’s warning “Cigarette Smoke Contains Carbon Monoxide” was standardized across all advertisements.

Outcome measures

Recall

To assess engagement with key features of the advertisement, similar to our previous work examining recall of advertising^29,32 and warning label³³ content, we used four forced-response items corresponding to recall of the product’s name, the content of the disclaimer, the percentage of reduced nicotine the product offered, and the content of the Surgeon General’s warning. Responses for the name, Surgeon General’s warning, and nicotine reduction percentage recall items contained both correct and bogus (e.g., Sunlight) options; responses for the disclaimer content item included each of the assigned conditions and “I don’t know.”

Risk beliefs

A scale previously used to measure RNC cigarette risk beliefs (α’s > 0.82)^17,29,34,35 asked participants to compare the advertised cigarettes to “regular” cigarettes on eight statements (e.g., lower nicotine, lower tar, less addictive) using a five-point responses (1 = “definitely untrue”, 5 = “definitely true”), later scored as dichotomous response categories (0 = “correct” and 1 = “incorrect”; unsure responses were scored as incorrect) according to factual correctness,³⁶ and summed to create an overall false beliefs measure.^17,29

Perceived risks

Perceived health risks were assessed using the mean of a 7-point Likert scale (1 = “very low”, 7 = “very high”) on which participants indicated their risk of developing six health conditions¹⁸ (lung cancer, heart disease, stroke, emphysema, bronchitis, other cancers) after using “regular” cigarettes (if non-smokers), their preferred brand (if smokers), and the advertised cigarettes (α’s > 0.9). Because the disclaimer content condition manipulated addiction risk messaging, we also included an item assessing perceived risk of addiction, examined as a separate outcome.

Attitudes

Attitudes were assessed using the mean of an 8-item, 7-point semantic differential scale^37,38 that asked, “If the advertised cigarettes became commercially available in the next 30 days, which of the words below would best describe your use of this product?” Items were a) bad/good, b) unenjoyable/enjoyable, c) unpleasant/pleasant, d) foolish/wise, e) difficult/easy, f) more/less harmful, g) not under/under my control, and h) less/more healthy. Higher scores indicate more favorable attitudes (α’s > 0.83).

Intentions

Two exploratory items assessed participants’ intentions to use the advertised cigarettes if 1) given the opportunity to purchase them and 2) if given them for free for one month. Responses were: a) purchase/smoke only these cigarettes, b) purchase my preferred brand of cigarettes, c) purchase an alternate nicotine product, d) try to quit smoking, or e) none of the above. I do not use or am not interested in using cigarettes. Due to low endorsement of responses c-e, responses were subsequently dichotomized as either choosing to purchase/smoke only the advertised cigarettes vs. all other options.

Other measures

Demographic and tobacco use history information.

Measures were assessed using items from the PhenX Toolkit (https://www.phenxtoolkit.org/collections/trrc). Smokers also reported their preferred brand cigarette characteristics (e.g., name, primary pack color, mentholation) and completed the Fagerström Test for Cigarette Dependence (FTCD).³⁹

Attention checks.

To identify inattentive participants, per the recommendations of Strickland & Stoops,⁴⁰ we included three attention checks: 1) selection of a specific color from a list of four choices; 2) selection of misspelled word from a list of four choices; and 3) consistently stating their state across two item formats (selection from a list of all U.S. states/territories vs. providing open-ended text).

Data quality.

Consistent with our previous work,⁴¹ we asked participants if (a) they took their time in completing the survey; (b) their data should be retained; and (c) they experienced any computer problems during the survey. To reduce likelihood of bot responding,⁴² we included a CAPTCHA (“Completely Automated Public Turing test to tell Computers and Humans Apart”) task at the survey end; participants unable to pass this element did not receive a completion code nor compensation.

Analytic plan

Analyses were conducted using IBM SPSS Statistics v26. Chi-square independence and unpaired t- tests or one-way analysis of variance (ANOVA) compared categorical and continuous variables, respectively, by smoking status and advertisement condition.

Main and interactive effects of product name and disclaimer content were analyzed using: binary logistic regression for recall and intention outcomes; two-way ANOVA for false belief and attitude outcomes; and two-way analysis of covariance (ANCOVA) for perceived health risk items (analyses controlled for perceived health risks of regular cigarettes for non-smokers and risks of own brand for smokers). All analyses were stratified by smoking status.

A priori power analyses conducted using G*Power v3.1.9.2. based on our prior work^29,34 determined that n = 300 would detect a small-medium product name * disclaimer content interactive effect on a continuous perception outcome (partial η² = 0.05–0.07;) with β = 95%, using a two-tailed significance test with α = 0.05. Anticipating discarding 10–30% of participant data due to data quality issues,^42,43 we overrecruited participants with the goal of obtaining 300 per smoking group to balance randomization to each advertisement condition.

RESULTS

Sample characteristics

The full survey was completed by 997 people after excluding individuals who were not compensated or took the eligibility screener multiple times. Given recent concerns about MTurk data quality,⁴³ we engaged in conservative data cleaning procedures, excluding data from 190 individuals for the following, potentially overlapping reasons: failed attention checks, inappropriately endorsed data validity questions, reported conflicting tobacco use behaviors between screener and final survey, listed as their preferred cigarette brand anything other than a verifiable domestic cigarette brand, or provided inappropriate responses to two open-ended prompts (e.g., “GOOD”, “nice”). Excluded individuals reported younger age [t (331.5) = 2.9, p = .005] and were more likely to report: being male [χ² (1) = 4.7, p = .030], being smokers [χ² (1) = 209.2, p < .001], being Non-Latinx Whites [χ² (4) = 92.8, p < .001], and having a bachelor’s degree or higher [χ² (1) = 50.5, p < .001].

The analytic sample (n = 807) was, on average, 38.0 years old (SD = 11.3; range = 18 – 74) and predominately comprised of males (58.2%), Non-Latinx Whites (74.2%), and individuals with a bachelor’s degree or greater (60.1%). Participants took, on average, 13.8 minutes (SD = 8.3; range = 2.2 – 57.8) to complete the survey, with the majority (98.0%) completing it using a desktop or laptop computer. Smokers were more likely than non-smokers to be male (p = .042) and marginally more likely to be Latinx White (p = .059); see Table 1. Smokers took longer to complete the survey, likely due to completing additional measures (e.g., more tobacco use history questions), and endorsed greater past 30-day non-cigarette tobacco product use than non-smokers (p’s < .001).

Table 1.

Sample (N = 807) characteristics presented as M (SD) or n (%).

	Smokers (n = 232)	Non-smokers (n = 575)	Statistics
Demographic information
Age	37.4 (11.2)	38.2 (11.3)	t (805) = −0.9
Gender*
Male	148 (63.8%)	322 (56.0%)	χ2 (1) = 4.1
Female	84 (36.2%)	253 (44.0%)
Ethnicity/Race
Non-Latinx White	176 (75.9%)	423 (73.6%)	χ2 (4) = 9.1
Non-Latinx Black	21 (9.1%)	54 (9.4%)
Latinx White	17 (7.3%)	26 (4.5%)
Latinx Black	1 (0.4%)	0 (0.0%)
Other	17 (7.3%)	72 (12.5%)
Education
Some college or less	103 (44.4%)	219 (38.1%)	χ2 (1) = 2.7
Bachelor’s degree or higher	129 (55.6%)	356 (61.9%)
Tobacco use history
# of days smoked in past 30***	20.0 (11.5)	0 (0.0)	t (231.0) = 26.5
Cigarettes per day***	9.1 (8.1)	1.2 (4.3)	t (319.6) = 13.7
Nicotine dependence (FTCD)	3.0 (2.1)	---
% reporting past 30-day use of
Cigarettes***	232 (100%)	0 (0.0%)	χ2 (1) = 582.0
E-cigarettes/vaping devices***	122 (67.4%)	53 (33.1%)	χ2 (1) = 39.9
Cigars/little cigars***	80 (43.0%)	16 (6.7%)	χ2 (1) = 83.1
Time spent on survey (min)***	18.1 (9.8)	12.0 (6.8)	t (325.42) = 8.7

Note.

^*indicates significant difference by smoking status at the p < .05 level,

^***= p < .001.

There were no differences in sample characteristics by advertisement condition among non-smokers (p’s > .2). Among smokers, only age differed by condition, F (5, 226) = 2.7, p = .021: those randomized to the Moonlight focused condition were younger than those in the Moonrise industry-proposed condition (p = .01).

Main outcomes

Recall

Among the full sample, recall was high for product name (86.1%) and nicotine reduction percentage (87.5%), moderate for the Surgeon General’s warning (68.2%), and low for disclaimer content (49.9%). As shown in Table 2, smokers were more likely than non-smokers to recall the product name [χ² (1) = 4.3, p = .039] but less likely to recall disclaimer content [χ² (1) = 4.6, p = .031]. Recall of the Surgeon General’s warning and nicotine reduction percentage did not differ by smoking status [χ² (1)’s = 1.8 and 2.0, p’s = .175 and .161, respectively].

Table 2.

Rates of correct recall (%’s) of advertisement elements by study conditions.

			Study conditions
			Moonlight			Moonrise
Smoking status	Advertising elements	Overall	Industry	Focused	None	Industry	Focused	None
Smokers	Brand name	90.1	90.9	97.0	90.0	88.6	90.0	85.7
	Disclaimer content	44.0	42.4	48.5	40.0	43.2	45.0	45.2
	SG warning	64.7	51.5	75.8	65.0	68.2	57.5	69.0
	% nicotine reduction	84.9	93.9	84.8	85.0	84.1	80.0	83.3
Non-smokers	Brand name	84.5	90.3	90.4	84.7	79.2	78.4	84.5
	Disclaimer content	52.3	57.0	53.2	57.1	41.7	45.4	59.8
	SG warning	69.6	67.7	77.7	63.3	74.0	62.9	72.2
	% nicotine reduction	88.5	88.2	92.6	88.8	86.5	85.6	89.7

Likelihood of correct recall of each advertising element did not differ by brand name, disclaimer content, or their interaction regardless of smoking status [χ² (5)’s = 0.6 – 10.2, p’s = .070 – .987] (Table 2).

False beliefs

As shown in Table 3, on average, smokers held more false beliefs about the advertised cigarettes compared to non-smokers, t (805) = 5.6, p < .001. Among non-smokers, there was a significant effect of disclaimer on false beliefs, F (2, 569) = 6.6, p = .001: compared to those who viewed advertisements with no content [M (SE) = 4.1 (0.1)], those who viewed ads with the industry-proposed [M (SE) = 3.5 (0.1), p = .001] and focused [M (SE) = 3.7 (0.1), p = .033] content had fewer false beliefs about the advertised cigarettes. Neither brand name nor its interaction with disclaimer content affected non-smokers’ false beliefs, F’s = 1.7 and 0.6, p’s = .193 and .565, respectively.

Table 3.

Outcome measures (%’s and means and standard errors) by study conditions.

			Study conditions
			Moonlight			Moonrise
Smoking status	Outcomes	Overall	Industry-proposed	Focused	None	Industry-proposed	Focused	None
Smokers	False beliefsc	4.5 (0.1)	4.6 (0.3)	5.2 (0.3)	4.7 (0.3)	4.5 (0.3)	4.2 (0.3)	4.5 (0.3)
	Perceived health risksa	4.8 (0.1)	4.7 (0.1)	4.7 (0.1)	4.6 (0.1)	5.0 (0.1)	4.9 (0.1)	4.9 (0.1)
	Perceived addiction riskb	4.9 (0.1)	5.1 (0.2)	4.8 (0.2)	4.4 (0.2)	5.4 (0.2)	5.0 (0.2)	5.4 (0.2)
	Attitudes	4.3 (0.1)	4.0 (0.2)	4.6 (0.2)	4.2 (0.2)	4.3 (0.2)	4.2 (0.2)	4.4 (0.2)
	Intentions to purchase	33.2%	18.2%	27.3%	37.5%	40.9%	37.5%	33.3%
	Intentions to use if provided free-of-charge	49.1%	39.4%	39.4%	52.5%	54.5%	50.0%	54.8%
Non-smokers	False beliefsb	3.8 (0.1)	3.3 (0.2)	3.7 (0.2)	4.0 (0.2)	3.7 (0.2)	3.7 (0.2)	4.2 (0.2)
	Perceived health risksb	5.6 (0.1)	5.8 (0.1)	5.6 (0.1)	5.5 (0.1)	5.7 (0.1)	5.6 (0.1)	5.4 (0.1)
	Perceived addiction riskb	5.5 (0.1)	6.0 (0.3)	5.3 (0.3)	5.2 (0.1)	5.6 (0.1)	5.5 (0.1)	5.4 (0.1)
	Attitudes	2.7 (0.1)	2.6 (0.1)	2.6 (0.1)	2.7 (0.1)	2.8 (0.1)	2.6 (0.1)	2.7 (0.1)
	Intentions to purchase	2.6%	2.2%	1.1%	5.1%	3.1%	1.0%	3.1%
	Intentions to use if provided free-of-charge	3.5%	2.2%	1.1%	7.1%	4.2%	2.1%	4.1%

Note. Superscript a indicates brand name effect; b = disclaimer effect; c = interactive effect.

Among smokers, there were no main effects of brand name and disclaimer content, F’s = 1.5 and 1.4, p’s = .219 and .259, respectively, but there was a significant disclaimer and brand name interaction, F (2, 226) = 10.2, p = .029, such that those who saw the Moonlight ad with the focused disclaimer [M (SE) = 5.2 (0.3)] held more false beliefs about the advertised cigarettes than those who saw the Moonrise focused disclaimer [M (SE) = 4.2 (0.3), p = .008] and the Moonlight industry-proposed disclaimer [M (SE) = 4.1 (0.3), p = .018] advertisements.

Perceived risks

Smokers (vs. non-smokers) perceived lower health risks, t (383.1) = −7.8, p < .001, and lower addiction risk, t (394.9) = −4.7, p < .001, from using the advertised product (Table 3). Among non-smokers, disclaimer content affected both perceived health risks and addiction risk, F (2, 568) = 9.0 and 5.8, p’s < .001 and .003, respectively. Non-smokers who viewed advertisements with industry-proposed content (vs. no content) perceived greater health risks [M (SE) = 5.8 (0.1) vs. 5.5 (0.1), p < .001] and addiction risk [M (SE) = 5.8 (0.1) vs. 5.3 (0.1), p = .015]; those who viewed industry-proposed (vs. focused) content also perceived greater addiction risk [M (SE) = 5.4 (0.1), p = .001]. There were no product name or interactive effects on perceived health or addiction risks (F’s < .01 – 2.7, p’s = .066 – 935).

Among smokers, disclaimer content had no effect on perceived health risks, F (2, 225) = 0.2, p = .730, but did significantly affect perceived addiction risk, F (2, 225) = 3.8, p = .023: smokers who viewed the industry-proposed disclaimer (vs. no content) perceived greater addiction risk [M (SE) = 5.2 (0.2) vs. 4.6 (0.2), p = .019]. Product name significantly affected perceived overall health risks, F (1, 225) = 4.1, p = .016), but had no effect on perceived addiction risk, F (1, 225) = 2.1, p = .144. Those who saw advertisements featuring the Moonlight vs. Moonrise name perceived lower risks from using the advertised product [M (SE) = 4.7 (0.1) vs. 4.9 (0.1)]. There was no interactive effect on overall perceived health or addiction risks, F’s (2, 225) ≤ 0.1, p’s = .998 and 954, respectively.

Attitudes

Mean attitudes toward using the advertised product were more favorable among smokers vs. non-smokers, t (460) = 19.2, p < .001 (Table 3). There were no main or interactive effects of product name and disclaimer content on mean attitudes toward the advertised cigarettes regardless of smoking status, F’s < 0.1 – 2.4, p’s = .095 – .927.

Intentions

As expected, smokers were more likely than non-smokers to endorse intentions to purchase the advertised cigarettes and intentions to use the advertised cigarettes if provided free-of-charge, χ² (1)’s = 153.1 and 248.9, respectively, p’s < .001 (Table 3). Product name, disclaimer content, and their interaction had no effect on intention items regardless of smoking status [χ² (5) = 2.5 – 6.7, p’s = .241 – .583].

DISCUSSION

This study examined the effects of product name and disclaimer content on perceptions of a novel RNC cigarette. We found robust effects of disclaimer content among non-smokers: those who viewed the industry-proposed content (“Nicotine is addictive. Less nicotine does NOT mean a safer cigarette”) perceived greater health and addiction risks of, and held fewer false beliefs about, the advertised product than those who saw no disclaimer. Compared to the focused content (“Less nicotine does NOT mean a safer cigarette”), the industry-proposed content better informed non-smokers of addiction risk but did not differ on perceived health risks or false beliefs. The industry-proposed content (vs. no content) also increased smokers’ perceived addiction risk, likely because of its mention of addiction, but did not affect perceived health risks. While there was no product name effect among non-smokers, smokers who saw Moonlight-branded (vs. Moonrise-branded) advertisements perceived lower health risks of the advertised cigarettes. Neither advertising element appeared to influence attitudes or intentions toward using the RNC cigarettes.

Findings suggest that 22^nd Century’s proposed statement aimed at informing consumers of product risks may be effective for non-smokers. However, among smokers, the industry-proposed disclaimer generally performed similarly to no disclaimer, suggesting limited utility: it sufficiently informs smokers of addictiveness risks but not general health harms, which present a larger concern when using combustible cigarettes. These findings are consistent with studies demonstrating that disclaimers do not offset harm misperceptions that accompany descriptive terms like “organic” and “natural” on tobacco products,^23,24 and with work suggesting that smokers do not look at, or engage with, textual warning content included in product marketing.⁴⁴ Because disclaimer recall was worse among smokers than non-smokers, these findings may similarly reflect a lack of engagement with this content, although this should be confirmed by future research. Disentangling whether smokers simply do not attend vs. attend to but do not process this content may help determine future strategies for communicating risk information. For example, attention might be enhanced by graphic (vs. textual) warnings, while information processing might be improved by briefer, more effective warning statements like “Smoking kills.”

Findings also suggest that the Moonlight name – like the prohibited ‘light’ descriptor and other available descriptors (“natural”, “additive-free”) – misleads smokers about product health risks. This finding is important because product name was the only advertisement element that smokers better recalled than non-smokers, suggesting smokers are more likely to engage with branding than risk information (e.g., Surgeon General’s warning). This is striking given that the advertisement did not prominently feature the product name (Figure 1) and was viewed for 30 seconds only. If authorized, the Moonlight product will be marketed for current smokers only to “reduce their cigarette consumption”—vague wording that does not specify the magnitude of behavioral reduction necessary to meaningfully decrease smoking-related health risks.^45,46 Further, because higher nicotine content combustible tobacco products are commercially available, Moonlight’s availability will not replicate the conditions or benefits of a mandated product standard, and thus could pose increased risks to smokers who engage in dual use of these products. However, we observed no product name effects on perceptions for non-smokers, suggesting the name is not misleading to new/former users.

Finally, we found no differences between advertisement conditions on attitudes or intentions toward using the RNC cigarette brand regardless of smoking status, suggesting that brand name and disclaimer content had little impact on product appeal, or at least may require additional exposure before eliciting changes in these outcomes. These data may help to ameliorate concerns that the manipulated features of RNC advertisements may encourage non-smokers to initiate use of the product.

Taken together, these findings broadly reflect the importance of descriptors and disclaimers on perceptions of emerging tobacco products, which apply to other regulatory environments outside the U.S. as they confirm prior research demonstrating that including disclaimers in tobacco product advertising does not offset misperceptions associated with other marketing attributes (i.e., branding).^23,24 Still, these findings should be interpreted considering some limitations. First, data were self-reported by an MTurk convenience sample and may not generalize to the larger population, although studies indicate that performance on experimental tasks is similar between online panels and laboratory-based studies.^26–28 Related, because we categorized smoking status based only on reported past 30-day cigarette use, it is possible that our non-smokers were not entirely comprised of true non-smokers; further research is warranted to explore differences in reactions to RNC cigarette advertising between never and former smokers. Second, because data were collected during the severe acute respiratory syndrome coronavirus (SARS-CoV-2) pandemic, it is possible that self-reported behaviors and perceptions may not represent behaviors under normal circumstances. Third, we manipulated two features of a single advertisement from 22^nd Century’s marketing application. Many complex factors are involved in shaping consumer perceptions of tobacco products⁴⁷ and other marketing materials emphasizing different advertising elements may produce different effects on product perceptions, attitudes, and behavioral intentions. Finally, the relative font size and placement of the disclaimer may have minimized engagement with its content; however, these aspects of the product advertisement are consistent with those proposed in the product’s marketing application.

In conclusion, study findings suggest that two advertising features differentially impact smokers’ and non-smokers’ perceptions of a new RNC cigarette product: the product names tested had little effect among non-smokers, but smokers viewing advertisements with a name utilizing a prohibited misleading descriptor perceived lower health risks from using the product. Non-smokers who viewed an industry-proposed disclaimer perceived greater health and addiction risk from using the product, but only smokers’ perceptions of addiction risk were affected by this element. Findings indicate that including a banned descriptor in the naming of a RNC cigarette product is likely to mislead smokers about health risks. Further, the industry-proposed disclaimer may be useful for informing non-smokers of product risks if a RNC standard is implemented, but it is not informative for smokers—the intended consumers—if sold within the current tobacco market as a modified risk tobacco product.

What this paper adds.

• Implementing a reduced nicotine product standard for combustible cigarettes in the U.S. may reduce tobacco-related morbidity and mortality at the population level; however, it is unknown whether product marketing may dampen these anticipated benefits.

• This study provides evidence that the industry-proposed voluntary disclaimer in advertising for a new reduced nicotine content (RNC) cigarette brand may increase consumers’ perceived risk of addiction and improve non-smokers’ health risk perceptions, but this element appears unlikely to address other perceived health risks among current smokers—the intended consumers of this product.

• Further, smokers perceived lower health risks from using RNC cigarettes advertised with the name “Moonlight,” illustrating the potential harm of allowing variations of banned misleading descriptors.