Waterpipe Tobacco Warnings: An Experimental Study Among a Nationally Representative Sample of US Young Adults Free

Abstract

Background

Waterpipe tobacco (WT) smoking by young adults remains high and misperceptions are common. Product warnings can increase knowledge of harms and reduce use. The goal of this study was to test warning statements, including the FDA-required nicotine warning (prior to implementation), on young adults’ thinking about harms of and discouragement from WT smoking.

Methods

We conducted a between-subjects experiment in a nationally representative telephone survey of 1152 young adults aged 18–29. Participants were randomly assigned to hear one of five warning statements and reported how much, on a 4-point scale, the warning made them think about the harms and discouraged them from WT smoking.

Results

The sample was 36.8% female, 57.8% white, 20.2% Black, 24.1% Hispanic, with a mean age of 23.2 (SE = 0.25). Under half (43.5%) had ever smoked WT. There were significant differences among the statements on both thinking about harms (p < .0001) and discouragement (p < .0001). The FDA-required “nicotine” warning led to the lowest thinking about harms (M = 2.85, SE = 0.08) and was the least discouraging (M = 2.86, SE = 0.08), while the “100 cigarettes” warning resulted in the greatest thinking about harms (M = 3.62, SE = 0.05) and was the most discouraging (M = 3.56, SE = 0.06).

Conclusions

The nicotine warning resulted in the lowest levels of thinking about harms and discouragement from WT smoking, suggesting limited impact. However, a warning focused on comparing smoke inhalation from WT smoking to cigarettes seems promising. Warnings should cover a broad range of WT health effects, and possibly comparisons to cigarettes. Findings also have implications for the content of international waterpipe warnings.

Implications

This study indicates that the nicotine warning is the least effective at making young adults think about the harms of and discouraging WT smoking. The FDA and other countries should consider requiring warnings to cover a broader range of health harms, misperceptions, and possibly comparisons to cigarettes.

Waterpipe tobacco (WT) smoking by young adults remains high. The 2018 Monitoring the Future study indicated the annual prevalence of WT smoking among those aged 19–28 was 13.3%.¹ While rates of combustible cigarette smoking have decreased, a recent systematic review found increases in past 30-day trends of WT smoking, with 2%–3% increases between 2011 and 2016, from longitudinal studies of US young adults.²

WT smokers are exposed to high levels of toxicants, including carcinogenic polycyclic aromatic hydrocarbons, carbon monoxide, and heavy metals.^3–7 Exposure to these carcinogens and toxicants is often in higher levels than in cigarette smoke. This is likely because a single WT session may last 30 min or more and expose users to between 39 and 172 times the volume of smoke relative to consuming a single cigarette.^3,7–10 WT smoking results in many of the same health harms as cigarette smoking, including lung, oral, and esophageal cancers; respiratory illness; chronic obstructive pulmonary disease; cardiovascular disease; and low birth-weight.^11–15 WT smoking is also associated with unique health risks such as carbon monoxide toxicity, due to the burning charcoal,¹⁶ and spread of communicable diseases.¹⁷ Even infrequent WT smoking is associated with health harms, such as lung-related abnormalities¹⁸; a single WT smoking session can lead to acute adverse increases in arterial stiffness, a precursor to pathogenesis and progression of high blood pressure, among young adults.¹⁹

Although WT smoking carries similar health risks as cigarette smoking, many consumers erroneously believe it is less harmful and less addictive than cigarette smoking,^20–22 and these misperceptions are positively associated with use.^20–25 Young adults who believe WT smoking is less harmful than cigarette smoking were 2.7 times more likely to be current waterpipe smokers than those who perceived WT smoking to be as harmful or more harmful than cigarette smoking.²⁴ WT smoking involves the passage of air, heated by burning charcoal, first through tobacco and then through water, prior to inhalation. This feature may be, at least partially, responsible for reduced harm perceptions, as young adults believe the water filters out the toxins, including nicotine, making it less harmful and less addictive.²³ Young adults also incorrectly believe that health harms are unlikely due to infrequent use.²⁶

Communication approaches, including product warnings, are needed to communicate harm to consumers. There is now consistent, compelling evidence that cigarette pack warnings are a practical, effective, and low-cost way to communicate harms and reduce smoking behaviors.^27,28 Additionally, emerging research suggests that warnings can deter youth from cigarette smoking.^28–31

The design of warnings, including the warning statement itself, is influential for cognitive processing and discouraging use. The US Food and Drug Administration (FDA) requires a single text warning on waterpipe packaging and advertising about the addictive nature of nicotine (This product contains nicotine. Nicotine is an addictive chemical.), which began in mid-2018. WT is the only combustible tobacco product with a single warning, and it is focused exclusively on nicotine. Other warning statements that address health harms and comparative risks (e.g., comparison to cigarette smoking) may be promising to correct misperceptions of harm among young adults. Young adults are concerned about inhaling chemicals,³² however, many youth are unaware these chemicals are in WT or erroneously believe that water in the bowl filters out toxic chemicals.³³

Few studies on the impact of WT warnings have been conducted to date. Using a discrete choice experiment, Salloum et al. found that WT warnings displayed on a café menu decreased product appeal.³⁴ A sample of college students who had ever smoked WT rated how much each of the six text warning statements and four pictorial warnings placed on different parts of the waterpipe device motivated them to think about reducing or stopping WT smoking. Results indicated that messages focused on harm to children were rated as most motivating.³⁵ Finally, a recent mixed-methods approach with four studies of young adults (three convenience samples and one random sample of first year college students) found the FDA nicotine warning for WT did not resonate with young adult focus group participants and scored low on measures of perceived effectiveness. Instead, warning focused on WT harm were perceived as more effective³⁶.

The goal of the current study was to understand the potential effectiveness of differing WT warnings by testing five warning statements, including the required nicotine warning, on young adults’ thinking about the harms of and discouragement from WT smoking in a nationally representative survey of young adults in the US.

Methods

Participants

Data were collected from a nationally representative sample of young adults. From August 2016 to May 2017 (prior to implementation of the required warning), the Carolina Survey Research Laboratory (CSRL) conducted phone surveys for a nationally representative probability sample of adults 18 years of age and older, living in the US with random digit dialing. Interviewers conducted the survey in English or Spanish and young adults were compensated $40 for their participation. The nationally representative probability sample for the adult phone survey was recruited using two independent and non-overlapping random digit dialing frames (both landline and cell phones), which resulted in 96% coverage of all US adult households. Cellular numbers were oversampled to increase the participation of young adults. CSRL oversampled in counties with higher prevalence of cigarette smokers and low-income respondents. Additionally, CRSL used a respondent driven sampling approach to further enhance recruitment of cigarette smokers and gay, lesbian, and bisexual individuals (GLB). Respondents who self-identified as GLB were always asked to refer GLB friends; otherwise, a participant was randomly assigned to refer friends who used tobacco products or friends who identified as GLB. Additional details about the sampling procedure can be found in Agans et al.³⁷ Interviewers obtained verbal consent from the young adults. The response rate for the adult survey was 38.7%, calculated using the American Association for Public Opinion Research Formula #4. This study focuses on the subset of 1152 young adults aged 18–29. The Institutional Review Board at the University of North Carolina reviewed and approved study procedures (NO. 13–2779).

Stimuli

Our development of statements to test for this study was driven by our desire to address a wide-range of targets, including misperceptions, comparisons to cigarettes, and long-term health harms. We developed four new warnings and utilized a revised version of the FDA-mandated nicotine warning, for a total of five warnings.

The first warning compared the volume of smoke inhaled during a single WT smoking session based on several lab-based studies comparing volume of smoke inhalation^3,7–10: Warning: One hour of hookah smoking is about the same amount of smoke as 100 cigarettes.

The second warning focused on the increased amount of carbon monoxide exposure, relative to cigarettes, likely due to the burning charcoal^38,39: Warning: Hookah smoke contains more carbon monoxide than cigarette smoke.

The third warning targeted the widely-held, but incorrect, misperception that the water in the hookah filters out the toxic chemicals, rendering it less harmful²³: Warning: The water in a hookah does not filter out the toxic chemicals.

The fourth warning focused on cancer^12,13: Warning: Hookah smoking causes cancer.

The fifth warning was adapted from the FDA-mandated warning required for WT: Warning: Hookah smoke contains nicotine. Nicotine is an addictive chemical.

Procedure

After items assessing tobacco use behaviors, young adults were randomized to hear one of five warnings, with the prompt, “Imagine seeing this hookah warning…” After hearing the warning, two outcomes measures were assessed (described below).

Measures

Demographics

We assessed age, sex, race (white, Black, or other), ethnicity (Not Hispanic or Hispanic, Latino, or Spanish origin), and sexual orientation.

Cognitive Elaboration

Amount of issue-relevant thinking, or cognitive elaboration, was assessed with the item, “How much does this warning make you think about the harmful effects of smoking tobacco in a hookah?” Cognitive elaboration is a critical component of the Elaboration Likelihood Model and is predictive of impact for cigarette warnings.⁴⁰ Response options were “not at all” (coded as 1), “a little” (2), “somewhat” (3), or “a lot” (4) for both items.

Discouragement

Discouragement is one of three items assessing Perceived Message Effectiveness (PME). PME is also predictive of message impact for tobacco control message.⁴¹ Due to space limitations, we were only able to include the discouragement item, “How much does this warning discourage you from wanting to smoke tobacco in a hookah?” Response options were “not at all” (coded as 1), “a little” (2), “somewhat” (3), or “a lot” (4) for both items.

WT Smoking

Ever use status was determined with the item “Have you ever smoked tobacco in a hookah, even one or two puffs?” Participants who responded “no” were classified as never users. Those who responded “yes” were classified as ever users.

Exposure to WT Smoking

All participants were asked if they have ever seen the package that WT comes in. Responses were “yes” or “no.”

Analyses

We calculated descriptive statistics to describe overall sample demographic characteristics and prevalence of WT use. We estimated weighted percentages using PROC SURVEYFREQ and weighted means using PROC SURVEYMEANS in SAS V9.4 to account for sampling design features. These characteristics were then compared across the five warning conditions using Rao-Scott Chi-Square tests for categorical variables and F-tests for continuous measures. Linear regression models were fit for the outcomes of perceived harm and discouragement with warning condition included as a main effect. Overall tests of differences between warning conditions were conducted using F-tests with t-tests for pairwise comparisons. After examining main effects of warning condition, an interaction between warning condition and ever use of waterpipe tobacco was included. Models were estimated using PROC SURVEYREG in SAS V9.4 to account for the sampling design features. All tests were two-sided with a .05 significance level.

Results

Young adults were 42.5% female, 55.4% White, 21.4% African American, and 23.6% Hispanic (see Table 1). Fewer than half of participants had mothers with a college education (38.7%) and 5.7% reported being GLB. Participants had a mean age of 21.1 y (SD = 2.4). Almost half (45.4%) of young adults were ever users of waterpipe tobacco and 36.8% reported exposure to a WT package. The five randomized warning groups did not differ on age, sex, race, mother’s education, sexual orientation, or ever use of WT (p > .05).

Impact on Cognitive Elaboration About the Harmful Effects of WT Smoking

There was a main effect of the warnings on cognitive elaboration about the harmful effects of WT smoking, F(4, 891) = 19.23, p < .0001. The required “nicotine” warning statement led to the least amount of thinking (M = 2.85, SE = 0.08) and less thinking about the harmful effects than all other warnings (p’s < .01). The “100 cigarettes” warning led to significantly more thinking about harmful effects (M = 3.62, SE = 0.05) than any of the other warnings, except the “carbon monoxide” warning. The “carbon monoxide” warning led to the second greatest amount of thinking (M = 3.48, SE = 0.06), and was greater than the “cancer” warning (M = 3.28, SE = 0.07), the “filter out the toxins” warning (M = 3.19, SE = 0.07), and the “nicotine” warning. The “cancer” warning led to greater thinking about the harms than did the “nicotine” warning but was not significantly different from the “filter out the toxins” warning (see Table 2).

Thinking about the harmful effects of WT smoking varied between ever waterpipe smokers and never waterpipe smokers for some warning comparisons. For never WT smokers, the “nicotine” warning (M = 2.98, SE = 0.10) was associated with less thinking about the harmful effects than the other four warnings (p-values < .05). Among ever WT smokers, the “nicotine warning” (M = 2.69, SE = 0.10) resulted in less thinking about the risks than the “carbon monoxide” (M = 3.31, SE = 0.09) and the “100 cigarettes” warning (M = 3.55, SE = 0.08). The “carbon monoxide” and the “100 cigarettes” warnings resulted in greater thinking about risks than either the “cancer” (M = 2.97, SE = 0.10) or “filters out the toxins” (M = 2.84, SE = 0.12) warnings, but were not different from each other.

Impact on Discouraging WT Smoking

There was also a main effect of the warnings on discouraging WT use F(4, 891) = 14.13, p < .0001. The required “nicotine” warning statement was rated as the least discouraging (M = 2.86, SE = 0.08) and was significantly lower than all other warnings (p’s < .01) (see Table 3). The “100 cigarettes” warning led to significantly more discouragement (M = 3.56, SE = 0.06) than the “cancer” warning (M = 3.23, SE = 0.08) and the “filter out the toxins” warning (M = 3.25, SE = 0.08), but did not differ from the “carbon monoxide warning” (M = 3.41, SE = 0.06). The “carbon monoxide,” “cancer,” and “filter out the toxins” warnings did not differ from one another on discouragement and there were no differences by waterpipe smoking status for discouragement, F(4, 891) = 1.96, p = .10.

Discussion

Among young adults, the FDA-required nicotine warning resulted in the lowest levels of thinking about risks and discouragement from WT smoking. These findings indicate the current warning is unlikely to have the desired effect of discouraging WT smoking. While the introduction of this novel warning may initially have some impact, additional optimized warnings are needed for WT warnings to have an enduring impact.

Our study uncovered two potentially promising targets for WT warnings. First, warnings comparing aspects of WT to cigarettes seem promising. The statements comparing the volume of smoke and carbon monoxide exposure to cigarettes were both rated highly. However, research is needed to ensure an accurate comparison without inadvertently making one product seem less dangerous than another.⁴² Moumen et al.,³⁶ also found that comparative messages were rated highly on perceived effectiveness. With careful development, comparison warnings may have the unique ability to extend the effects of tobacco control efforts that have reduced cigarette smoking to other products like WT. Comparison warnings may also help to dispel the widespread myth that WT smoking is much less harmful than cigarette smoking.

Second, warnings beyond nicotine seem more impactful than nicotine warnings alone. This finding has important regulatory implications in the US and worldwide. The World Health Organization Framework Convention on Tobacco Control (FCTC) requires all signatory countries to adopt and implement tobacco product warning labels (Article 11).⁴³ Although the number of warnings is not specified, the FCTC does require multiple warnings and recommends that parties consider establishing two or more sets of warnings for rotation. Article 11 also specifies that warnings should cover a range of harmful health effects, including addiction.⁴³ However, our findings suggest that warnings focused on nicotine addiction may be less impactful than those focused on other health effects. This finding is consistent with two recent studies of text messages for e-cigarettes, that also found nicotine and addiction warnings were less discouraging than warnings focused on health harms.^44,45

In the US, the FDA should consider requiring additional WT warnings to cover a broad range of health effects. WT is the only combustible tobacco product with just the nicotine warning. Under federal law, cigarettes will be required to implement 11 pictorial warnings as of October 16, 2021, and cigar products are required to rotate six text warnings that cover a broad range of health effects, pending ongoing litigation.^46,47 Despite WT smoking being linked to several serious health outcomes, the FDA only required the nicotine warning for WT. WT smoking includes toxic exposure to far more chemicals than nicotine alone, as well as having unique source of carcinogens from the charcoal during use. Warnings that highlight the breath of health effects are likely needed to elicit the affective and cognitive reactions that are necessary to change behavior.⁴⁰

A limitation of this study is the mode used for data collection. Participants heard, rather than read, WT warnings. However, our study provides an important first step to determine the most effective statements to begin to develop effective warnings. This study was also limited in the number of warning statements that could be tested, giving us hints of what may work but not permitting firm conclusions about effective warning themes for WT smoking. Future studies should address a broader range of health effects, both acute and chronic, as well as other misperceptions that may influence use, in order to understand effective statements and more broadly, effective themes for WT warnings.

Future work should replicate these effects with warnings printed on WT packaging. In this study, over a third of young adults reported exposure to the packaging that WT comes in at least one time. These results highlight the importance of package warnings, which are already required by law in the US and are an important part of the FCTC. Studies of warning implementation in hookah cafes are also needed to address the challenges of exposure that is unique to WT smoking. While consumers often see product warnings for other tobacco products each time the product is used, this may not be the case for many WT users who primarily smoke in cafés and lounges.⁴⁸ Studies that place warnings in a variety of locations (e.g., packages, waterpipes) and realistic use settings (e.g., where use may last over an hour) are needed to truly understand their impact on behavior. In a laboratory study, Maziak et al. found that exhaled carbon monoxide, a marker of consumption, was lower in a condition where participants were exposed to a pictorial warning label on the waterpipe than in a no warning control condition.⁴⁹ This is promising and yet much more work remains to be done to build the science of WT warnings.

In conclusion, our study provides a test of five warnings for WT smoking, and demonstrates the varying receptivity of these different statements among young adults. Our results highlight the need for the FDA to consider expanding the required WT warning beyond the singular focus on addiction, and provide a basis for much additional research on WT warnings. Given the serious health consequences of use and the common misperceptions that WT smoking is safer than cigarette smoking, urgent action by the FDA is needed.

Funding

Research reported in this manuscript was supported by grant number P50CA180907 from the National Cancer Institute and FDA Center for Tobacco Products (CTP). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Food and Drug Administration.

Declaration of Interests

SMN has served as a paid expert witness in government litigation against tobacco companies. The remaining authors do not have any declaration of interests to disclose.

References