Posts

 

 

Are you 18 years and older, live/work in the greater Columbus, OH area, and identify as a member of the LGBTQ+ community or serve the LGBTQ+ community in your work? The Ohio State University’s Center for HOPES and PS LGBTQ Equity Lab, along with Franklin County Public Health, and Columbus Public Health is conducting a study to investigate the needs and assess of the LGBTQ+ community in greater Columbus, OH. Eligible participants will receive a $50 gift card for their insight. 

Follow the link to find out if you are eligible: 

Associations Between Current Tobacco Use Status, Region, and Tobacco Messaging Perceptions: A Secondary Analysis

Author: Ashley Meadows; Academic Advisor: Dr. Ashley Felix, PhD; Committee Member: Dr. Joanne Patterson, PhD

ABSTRACT

Objective

We assessed how current tobacco use status and region lived in affect the perceptions surrounding tobacco educational messages for those currently not using tobacco. Analysis of messaging perceptions allows for a better understanding of how message type influences current non-users’ tobacco usage status. Our goal was to better understand the messaging that is currently being used to discover if current tobacco education prevention messages motivate current non-users to start smoking cigarettes or e-cigarettes, as well as to better understand the importance of residence in driving tobacco status and how this plays a role in their perspective on the effectiveness of tobacco education prevention messages.

Methods

Young adults aged 18-35 years living in the United States during the year 2022 were analyzed, with the influence of sociodemographic characteristics assessed.  These individuals were divided into two analysis groups for tobacco usage status; current user (n=443) and not current user (n=2,363) as well as into four analysis groups for regional location; Northeast (n=471), West (n=606), South (n=1,129), and Midwest (n=600). For inclusion in the study, participants had to report their current tobacco status, report regional location, and participate in questions regarding whether messages they viewed would motivate non-users to initiate cigarette smoking or vaping. Participants with missing or incomplete demographic data were excluded from the model (N=43). Utilizing multivariate logistic regression, we examined interactions between current tobacco status and region, for their association with effects perceptions on vaping and cigarette smoking initiation on current non-users.

Results

Of the population within this study sample, 84.21% were not a current user and 40.24% were currently living in the southern region of the United States. We found that young adults who reported currently using cigarettes and/or vapes have a 1.29 (95% CI: 1.02 – 1.63) and 1.89 (95% CI: 1.36 – 2.63) greater odds of agreeing that messages will motivate non-users to start vaping or cigarette smoking respectively compared to those not currently using tobacco products. We found that after accounting for covariates in adjusted logistic regression models, the region lived in was again no longer significant with the dependent variable.

Conclusion

Effects perceptions surrounding tobacco educational messages regarding non-users potential smoking initiation differed by tobacco usage status. Individuals married/divorced as well as cisgender men may benefit from targeted intervention strategies to persuade non-users more effectively from tobacco initiation. While we did not observe regional differences in our study, future studies would need to evaluate if separating regions into individual states would produce similar results to previous tobacco research findings.

Keywords: Effects Perceptions; Smoking Status; Region; Young adults; Tobacco; United States. 

INTRODUCTION

Tobacco use is the leading cause of preventable premature disease and death and is responsible for about 480,000 deaths in the United States each year (Prochaska & Benowitz, 2016). Although smoking prevalence has declined in the U.S., declines have been slower for young adults (Cantrell et al., 2018; Villanti et al., 2010). This is not surprising, as young adulthood, spanning from 18-35 years of age, is a crucial developmental period when individuals begin to experiment and regularly use tobacco products (Messeri et al., 2019). In a longitudinal analysis of PATH study data, 25.9% of young adults who reported never using nicotine/tobacco products, initiated use at follow-up, with 10% reporting new combustible tobacco use (Cooper et al., 2022). Rates of polytobacco use (use of multiple nicotine and tobacco products) and poly substance use (use of tobacco and other substances) have also increased (Collins et al., 2017). Critically, after the 1998 Master Settlement Agreement restricted companies from marketing tobacco to adolescents, Big Tobacco directed their marketing efforts to young adults (Hammond, 2005; Ling & Glantz, 2002; Hafez, 2005; Sepe et al., 2002). Young adulthood also coincides with the legal age of purchasing tobacco products (Ribisl & Mills, 2019); together, these factors may contribute to prosocial tobacco norms seen in this age group (Ribisl & Mills, 2019).

Electronic cigarettes, better known as nicotine vapes, have gained popularity among young adults and now surpass the use of traditional cigarettes (Sood et al., 2018). The uptick in usage is due, in part, to the public belief that, compared to cigarette smoking, vaping nicotine is less harmful to one’s own health and to the health of others (Romijnders et al., 2018). While smoking cessation is a commonly cited reason for vaping among middle-aged and older adults, young adults are more likely to start vaping due to peer influence and curiosity  (Pepper et al., 2016; Zhu et al., 2014); however, product characteristics, including flavors (e.g., fruit or mint vs tobacco) (Pang et al., 2023), nicotine concentration and form (e.g., nicotine salts vs free-based nicotine (Benowitz, 2022), and device type (e.g. pod vs. cig-a-like) (Lee et al., 2022) are key factors in sustaining nicotine dependence (Sargent et al., 2022). Given the abuse liability of nicotine vaping among young people, the FDA has begun regulating the sale and distribution of e-cigarettes, including required warning statements (US Food and Drug Administration, 2016) and flavor bans (Tam et al., 2023). Studies suggest that young adults are interested in quitting vaping (Cuccia et al., 2021) because they are concerned about the negative impacts on vaping long-term on physical health and well-being (Cuccia et al., 2021). While e-cigarettes are less likely than traditional cigarettes to cause lung damage; significant harms, including elevated blood pressure, respiratory damage, and brain development damages are known health harms (Noar et al., 2022; Bhatnagar et al., 2014; Chen, Bullen, & Dirks, 2017 National Academies of Sciences Engineering and Medicine, 2018; U.S. Department of Health and Human Services, 2016).

Effective tobacco prevention media campaigns can significantly decrease tobacco initiation, cessation, and prevalence of tobacco use. Media campaign messages achieve this because they are communicated more quickly and effectively to wide audiences with a relatively low cost per person, and findings show that there is an association between high exposure to anti-tobacco campaigns and increased knowledge of smoking harms to one’s health (Bafunno et al., 2020). Growing literature suggests that analyzing how effective an anti-tobacco advertisement is would be to analyze the ad through the lens of effects perceptions. This effects perceptions analysis helps to better predict the impact the ad itself has on this change of behavior (Baig et al., 2021; Brennan et al., 2014; Rohde et al., 2021). Effects perceptions are judgements made by the participant about a message campaign ad’s protentional to change the continuation of smoking or cease to smoke due to the message. Examples include the extent to which a message would make one believe that e-cigarettes or cigarette smoking is harmful, or the extent to which a message would motivate one to not vape or cigarette smoke (e.g., “This ad would motivate me to stop smoking”) (Noar et al., 2020).

To attempt to prevent current non-users from ever picking up smoking at all, it is important to obtain perspectives on those young adults who currently do use tobacco to discover what forms of messaging they believe could have likely prevented them from picking up any form of tobacco product to begin with (Pang et al., 2023). Current tobacco using individuals provide critical insight on the motivations for continued smoking despite knowledge of health risks. This specific targeting of current tobacco users can help better formulate the necessary strategies to reduce the number of current non-users from tobacco initiation.

Research shows that certain people have a greater exposure to tobacco and tobacco culture because of their geographic location and demographics. These regional policy variations impact young adults and should be considered in addressing cessation interventions. Up to date U.S. census regional data shows that current cigarette smoking is highest in the Midwest and South and lowest in the West. In the Midwest, 14 of every 100 adults are currently cigarette smoking, while nearly 9 of every 100 adults who live in the West are currently cigarette smoking (CDC, 2023). Current research is still in progress to determine which regions are most prevalent to e-cigarette usage; it is not yet clear the true disparities regions in the U.S. have for e-cigarette use. However, what is known is that according to the CDC, many of the same states that are in the Midwest and Southern regions are reporting adult e-cigarette use to fall within the top 25% of all U.S. states to have highest prevalence rates (Truth Initiative, 2023). These regional disparities that the United States has is due to a combination of factors, including powerful tobacco industry interference in legislation, and prevention of local jurisdictions from adopting strong tobacco prevention measures (Truth Initiative, 2023). Research shows that the tobacco prevention and cessation programs provided are one of the primary ways of educating young adults about the dangers of beginning smoking (American Lung Association, 2023).

We evaluated whether current smoking status was associated with young adults’ perceptions of tobacco education prevention messages, including whether exposure to messages about the shared health risks of smoking and vaping may influence smoking/vaping initiation among non-users (i.e., effects perceptions). We also assessed whether region of residence moderated the association between smoking status and effects perceptions.

 

METHODS

STUDY DESIGN

Data came from a parent study designed to test the effectiveness of health messages about vaping and smoking on message and risk perceptions among young adults (age 18-35 years).  Individuals were eligible to participate if they were 18-35 years old and reported living in the United States. We oversampled for tobacco use and self-reported minoritized sexual orientation and gender identity. The sample (N=2,849) was recruited into an online survey experiment via Prolific, an online platform for online subject recruitment for research (https://prolific.com), in fall 2022. Participants were paid $3.50 for participation via Prolific. The parent study was approved by the Institutional Review Board at our University.  For inclusion in this secondary data analysis, participants had to report their current tobacco status, (whether they were currently cigarette smoking or vaping) and respond to questions about whether the message that they viewed would motivate a non-user to start vaping or cigarette smoking (N = 2,806) in a self-reported online questionnaire.

MEASURES

Outcome variables

Two separate items assessed effects perceptions, or the perceived impact of exposure to the messages on smoking and vaping behaviors among non-users, see supplemental figures 2-4 for examples of messaging participants were shown. Participants were asked to respond to the following items, “These messages will motivate non-users to start vaping” and “These messages will motivate non-users to start smoking cigarettes”. Both items were scored using a 5-point Likert-type scale (1= strongly disagree, 5= strongly agree). Responses for each item were recoded such that participants reporting “neutral (3)” to “strongly agree (5)” were coded as believing that exposure to the messages would motivate non-users to start vaping/smoking vs. those who reported “strongly disagree (1)” and “disagree (2)” were coded as believing that messages would not motivate non-users to start vaping/smoking (coded 0). Interview questions with response options are provided in Supplemental Figure S1.

 

Independent variables

Current tobacco user status was assessed via the two-question method. First, participants were asked if they had “Ever smoked a cigarette EVEN one puff?”. Participants who responded “Yes” were then presented with a follow-up item: “Do you currently smoke cigarettes every day, some days, or not at all?”. Those who reported currently smoking every day or some days were recoded as a “current cigarette smoker” while those reported never, or former use were recoded as a “not current cigarette smoker”. Similarly, we asked participants if they had “Ever vaped nicotine, EVEN once?”. Participants who responded “Yes” were then presented with a follow-up item: “Do you currently vape nicotine every day, some days, or not at all?”. Those who reported currently vaping every day or some days were recoded as a “current nicotine vaper” while those reported never, or former use were recoded as a “not current nicotine vaper”. To retain an adequate sample size of young adults currently engaged in nicotine and tobacco use, we combined items into a binary variable such that participants coded as a current cigarette smoker, or a current nicotine vaper were recoded as “current tobacco user” while those identified as not a current smoker or vaper were recoded as “not a current tobacco user”.

Region was assessed with a single question, “What U.S. state, district, or territory do you currently reside in?” Responses were combined into a four tiered variable per U.S. census division of land definitions (U.S. Department of Commerce, 2023). Responses were coded into “Northeast” consisting of 9 U.S. states, (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont, New Jersey, New York, Pennsylvania; coded 0), “West” consisting of 13 U.S. states (Arizona, Colorado, Idaho, New Mexico, Montana, Utah, Nevada, Wyoming, Alaska, California, Hawaii, Oregon, Washington; coded 1), “South” consisting of 16 U.S. states (Delaware, Washington DC, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, West Virginia, Alabama, Kentucky, Mississippi, Tennessee, Arkansas, Louisiana, Oklahoma, Texas ; coded 2), and “Midwest” consisting of 12 U.S. states (Indiana, Illinois, Michigan, Ohio, Wisconsin, Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, South Dakota; coded 3).

Sociodemographic Covariates

Race and ethnicity were assessed with a single question in which participants were asked to select all responses that applied to them: “What racial and/or ethnic groups do you identity with?” Responses were recoded into two variables. Ethnicity was coded such that participants responding “Hispanic/Latinx/Latino/Latina” were coded as Hispanic/Latinx. Race was coded such that participants reporting any Race (including one or more Races) besides “non-Hispanic White/Caucasian” were coded as non-Hispanic BIPOC (Black, Indigenous, and People of Color).

Age of participant was assessed with a single open-ended question, “How old are you? (years)”.  Responses were recoded into two variables “18-24” (coded 0) and “25-35” (coded 1).

Gender identity was assessed using the two-step method in which participants were asked to report their sex-assigned-at-birth (male, female, intersex) and current gender identity (Female, Male, Transgender female/Transgender woman, Transgender male/Transgender man, Non-binary, Gender queer, Agender). Responses were recoded into a categorical variable such that participants were assigned as “Cisgender Male” (coded 0), “Cisgender Female” (coded 1), and “Transgender/Non-Binary+” (coded 2).

Income was assessed with a single item, “What is your total yearly income from all sources?” Responses were recoded into a categorical variable, where income categories were defined as “$50,000+” (coded 0), “$20,000-49,999” (coded 1), “<$20,000” (coded 2).

The community in which a participant lived was assessed with one item, “What type of community do you live in?” Response options included, “Large city” (coded 0), “Suburb near a large city” (coded 1), “Small city or town” (coded 2), and “Rural area” (coded 3).

Marital status was assessed with a single question, “What is your marital status?” Responses were recoded into a categorical variable such that participants were assigned as “Single/Divorced” (coded 0), “Partnered, cohabitating” (coded 1), “Partnered not cohabitating” (coded 2), and “Married” (coded 3).

Education level was assessed with a single question, “What is your highest grade completed?” Responses were recoded into a binary variable such that participants were assigned as “Less than high school/high school/certificate” (coded 0), and “Some college and above” (coded 1).

STATISTICAL ANALYSES

We first applied descriptive statistics to examine the sociodemographic characteristics of the sample by current tobacco status and region. We applied theory- and data-driven (Rothman et al., 2008) methods for identifying potential confounders. After identifying potential covariates and confounders based on significance in univariable models, we used multivariate logistic regressions to model the unadjusted association between baseline risk factors and (1) effects perceptions on vaping and (2) effects perceptions on cigarette smoking. Potential confounders and covariates were entered into multivariable logistic regressions to estimate adjusted effects.

We tested for an interaction between tobacco status and region to assess whether the association between tobacco status and effects perceptions specific to current non-users of cigarettes and vapes differed by region. We verified these interactions through sensitivity analysis (Tables 7-9), as well as constructed goodness of fit models to ensure the logistic regression models were adequately fit (Tables 10 and 11). As there were no significant interaction effects, we present main effects models only. An alpha of p<.05 was used to indicate statistical significance in final models.

 

RESULTS

SAMPLE CHARACTERISTICS

Tables 1 and 2 present sociodemographic characteristics of the sample based on current tobacco use and region lived. Participants overall were mostly white (63.44%), non-Hispanic (86.60%), cisgender female (48.36%), and had some college education or above (77.76%). Most of the participants within the sample were not a current tobacco user (84.21%) and were currently living in the southern region of the United States (40.24%).

PERCEIVED EFFECT OF MESSAGE EXPOSURE ON NON-USERS’ MOTIVATION TO BEGIN VAPING

Unadjusted logistic regression models indicated that our independent variable of interest, current tobacco status, was significantly associated with respondents’ perceptions that exposure to messages will motivate non-users to start vaping (Table 1). Among young adults in our sample, more current tobacco users agreed that messages will motivate non-users to begin vaping (21.58%) than current non-users (26.19%). Five sociodemographic covariates (age, gender, income, community type, marital status) were significantly associated with current tobacco usage status and, therefore, included in the final multivariable model.

However, unadjusted logistic regression models for our second independent variable of interest, region, indicated no statistically significant association between respondents’ perceptions that exposure to messages would motivate non-users to begin vaping (Table 2). Among young adults in our sample, more people living in the Midwest agreed that the messages would motivate non-users to begin vaping (25.17%) than those in the Northeast (22.72%), West (20.96%), and South (21.35%). Five sociodemographic covariates (race, ethnicity, income, community type, education) were significantly associated with region and, therefore, included in the final multivariable model.

After accounting for covariates in adjusted logistic regression models, there was no statistically significant association between region and dependent variable.

After accounting for covariates in adjusted logistic regression models, race, ethnicity, age, gender, income, community type, and education were no longer significantly associated with perceptions that messages would motivate non-users to begin vaping.

Final models indicated that young adults who reported currently using tobacco products had a 1.29 (95% CI: 1.02 – 1.63) greater relative odds of agreeing that exposure to messages will motivate non-users to start vaping than those not currently using tobacco products (Table 5).

Compared to those who reported their marital status as being single/divorced, those who were partnered and not cohabitating reported lower relative odds of agreeing that exposure to messages will motivate non-users to start vaping (OR: 0.71; 95% CI: 0.52 – 0.99).

PERCEIVED EFFECT OF MESSAGE EXPOSURE ON NON-USERS’ MOTIVATION TO BEGIN CIGARETTE SMOKING

Unadjusted logistic regression models indicated that our independent variable of interest, current tobacco status, was significantly associated with respondents’ perceptions that exposure to messages will motivate non-users to begin cigarette smoking (Table 1). Among young adults in our sample, more current users agreed that messages will motivate non-users to begin cigarette smoking (12.19%) than current non-users (6.26%). Two sociodemographic covariates (age, community type) and four potential confounders (race, gender, income, marital status) were significantly associated with perceptions around non-users’ motivation to begin cigarette smoking, and so were included in the final multivariable model.

However, unadjusted logistic regression models for our second independent variable of interest, region indicated no statistically significant association between respondents’ perceptions that exposure to messages would motivate non-users to begin cigarette smoking (Table 2). Among young adults in our sample, more people living in the Midwest agreed that messages were motivating nonsmokers to begin cigarette smoking (8.33%) than those in the northeast (8.07%), West (6.77%), and South (6.47%). Three sociodemographic covariates (ethnicity, community type, education) and four potential confounders (race, gender, income, marital status) were significantly associated with respondents’ perceptions that exposure to messages would motivate non-users’ to begin cigarette smoking, and so were included in the final multivariable model.

After accounting for covariates and potential confounders in adjusted logistic regression models, region was again no longer significantly significant with the dependent variable (Table 6).

After accounting for covariates and potential confounders in adjusted logistic regression models, race, ethnicity, age, income, community type, and education were no longer significantly associated with perceptions that messages would motivate non-user’s to begin cigarette smoking.

Final models indicated that young adults who reported currently using tobacco products have a 1.89 (95% CI: 1.36 – 2.63) greater relative odds of agreeing that messages will motivate non-users to start cigarette smoking than those not currently using tobacco products (Table 6).

Compared to those who identified as cisgender male, those who identified as cisgender female (OR: 0.66; 95% CI: 0.49-0.90) or transgender/NB+ (OR: 0.41; 95% CI: 0.20-0.83) reported lower relative odds of agreeing that messages will motivate non-users to start cigarette smoking.

Compared to those who reported their marital status as being single/divorced, those who were partnered and not cohabitating reported lower relative odds of agreeing that messages will motivate non-users to start cigarette smoking (OR: 0.52; 95% CI: 0.29 – 0.95).

DISCUSSION

In this secondary analysis of an experimental study testing the effectiveness of tobacco education prevention messages, we examined associations between participant characteristics and risk perceptions. To our knowledge, this is the first paper to examine how exposure to tobacco education messages is associated with young adults’ perceptions of the potential effect of exposure to these messages on smoking initiation among current non-users. In our multivariable models, we observed that current users perceived the messages as likely to motivate non-users to begin cigarette smoking or vaping. Conversely, those who were partnered and not cohabitating were less likely to perceive messages as likely to motivate non-users to initiate cigarette smoking or vaping; while those who identified as cisgender female or transgender/nonbinary were less likely to perceive messages as likely to motivate non-users to begin vaping. Contrary to our initial hypothesis, we found no significant regional differences in effects perceptions pertaining to vaping or smoking initiation among non-users. When controlling for potential confounders, other participant characteristics, including race, ethnicity, age, income, community type, and education were not associated with effects perceptions around vaping and smoking initiation among non-users.

We observed that current tobacco users perceived the messages as likely to motivate non-users to begin smoking or vaping. This belief may be related to how smoking initiation begins in this age group. Literature shows that the most likely time in one’s life to begin experimentation with nicotine products is young adulthood (Messeri et al., 2019); as such, this group may be more attentive to the potential for messaging to influence smoking and vaping uptake during young adulthood. Additionally, social pressure from peers and friends is an critical determinant of smoking and vaping initiation among young people (Littlecott et al., 2023). Accordingly, current users may believe that anti-tobacco messages are not strong enough to deter non-users from nicotine and tobacco use if other outside forces are more prominent (Kelsh et al., 2023; Tenny et al., 2024).

We also observed that those who were partnered and not cohabitating perceived the messages as less likely to motivate non-users to begin smoking or vaping. This is in contrary to literature surrounding current smoking status within the United States (CDC, 2023) as literature around tobacco usage and cohabitation indicates a higher prevalence of tobacco usage.  Young adults who are not cohabiting with their partners are more likely to engage in dating behaviors, including participating in social settings where co-use of tobacco and other substances is present, including bar and parties (Trotter, 2002). In these spaces, the social pressure to use tobacco is high for young adults and tobacco use may be considered normed behavior (Littlecott et al., 2023). This however is in contrast of what we observed. Potentially, this could be due in part of stigma around the smell of cigarettes and vapes, if one partner is not smoking, they could motivate the smoking partner to quit to continue to be with them (Wadsworth et al., 2016).  Additionally, having a partner could provide support measures in other way to deal with stressful situations that would otherwise cause them to initiate smoking.  Furthermore, having a partner as support when attempting to quit smoking is a powerful quit technique that many smoking individuals express as beneficial on their quit journey (Branstetter et al., 2012). This demographic of young adults may believe that the effects of messages are enough to keep non-users away from initiation in combination with the support their partner provides to them.

We observed that those who identified as cisgender female or transgender/nonbinary had lower relative odds of agreeing that messages would motivate non-users to initiate vaping, co compared to cisgender males. Current census data supports that men are more likely to be currently vaping than females (CDC, 2023). This observation may be due to other coping mechanisms that females or transgender/nonbinary individuals have found for stress related factors in everyday life (Pericot-Valverde et al., 2021). Additionally, this demographic has higher rates of experiencing sexism and misogyny than their cisgender counterparts (Arayasirikul et al., 2022; Verbeek et al., 2020; White Hughto et al., 2015). Vaping may be seen as more stigmatizing within female or transgender/nonbinary population as social circles may be more disapproving of seeing them smoke (Wadsworth et al., 2016). This demographic may believe that their surroundings as well as messaging may be enough to keep non-users away from vaping initiation.

Initially, we hypothesized that those living in the south would be more likely to endorse that messages would motivate smoking or vaping initiation; however, we observed no statistically significant associations between outcomes and region. Our hypotheses were based on regional variations in smoking patterns and with regional variations in smoking policy. Our findings might reflect a lack of data regarding how long the participant lived in the current region, if they were born there, and where close family and friends live. Additionally, how we defined region is broad based upon CDC census data. However, dividing region this way is heterogeneous and includes states that have less restrictive tobacco laws (e.g. Alabama) and states/territories with more restrictive laws (e.g. District of Columbia). This lack of granular data may be contributing to missing important patterns by region that our study was unable to show. Future studies would need to further evaluate if dividing regional data into separate states would produce results like previous research findings in regard to various policies and restrictions within each state.

Due to larger sample size, our study was able to have a current tobacco users comparison group that included both cigarette and e-cigarette users to assess the unique perceptions held by this high-risk population. However, the parent study of this sample oversampled LGBTQ+ individuals, a group who have been historically targeted by Big Tobacco to begin tobacco smoking, which may further explain the higher rates of perceptions around effects messages in our study. Additionally, our study was a cross-sectional study, which obscures bidirectional associations and cannot assess changes in effects messaging perceptions over the life course. This is especially important as tobacco use can begin and end across several months in the year. Prospective analyses documenting changes in correlation across young adults’ lifetimes would help to better understand the effect of these factors on effects perceptions around tobacco use with current non-users, as well as identify critical timepoints for intervention.

REFERENCES

American Lung Association. (2022). Tobacco 21 Laws. American Lung Association. https://www.lung.org/policy-advocacy/tobacco/prevention/tobacco-21-laws

American Lung Association. (2023). Tobacco Prevention and Cessation Funding. American Lung Association. https://www.lung.org/research/sotc/state-grades/state-rankings/tobacco-prevention-funding

Arayasirikul, S., Turner, C., Trujillo, D., Sicro, S. L., Scheer, S., McFarland, W., & Wilson, E. C. (2022). A global cautionary tale: Discrimination and violence against trans women worsen despite investments in public resources and improvements in health insurance access and utilization of health care. International Journal for Equity in Health, 21(1), 32. https://doi.org/10.1186/s12939-022-01632-5

Bafunno, D., Catino, A., Lamorgese, V., Del Bene, G., Longo, V., Montrone, M., Pesola, F., Pizzutilo, P., Cassiano, S., Mastrandrea, A., Ricci, D., Petrillo, P., Varesano, N., Zacheo, A., & Galetta, D. (2020). Impact of tobacco control interventions on smoking initiation, cessation, and prevalence: A systematic review. Journal of Thoracic Disease, 12(7), 3844–3856. https://doi.org/10.21037/jtd.2020.02.23

Baig, S. A., Noar, S. M., Gottfredson, N. C., Lazard, A. J., Ribisl, K. M., & Brewer, N. T. (2021). Message perceptions and effects perceptions as proxies for behavioral impact in the context of anti-smoking messages. Preventive Medicine Reports, 23, 101434. https://doi.org/10.1016/j.pmedr.2021.101434

Barnes, A. J., Bono, R. S., Lester, R. C., Eissenberg, T. E., & Cobb, C. O. (2017). Effect of Flavors and Modified Risk Messages on E-cigarette Abuse Liability. Tobacco Regulatory Science, 3(4), 374–387. https://doi.org/10.18001/TRS.3.4.1

Barrington-Trimis, J. L., Braymiller, J. L., Unger, J. B., McConnell, R., Stokes, A., Leventhal, A. M., Sargent, J. D., Samet, J. M., & Goodwin, R. D. (2020). Trends in the Age of Cigarette Smoking Initiation Among Young Adults in the US From 2002 to 2018. JAMA Network Open, 3(10), e2019022. https://doi.org/10.1001/jamanetworkopen.2020.19022

Bascom, E. (2022). US states with the highest smoking rates in 2022. Healio. https://www.healio.com/news/primary-care/20220906/us-states-with-the-highest-smoking-rates-in-2022

Benowitz, N. L. (2022). The Central Role of pH in the Clinical Pharmacology of Nicotine: Implications for Abuse Liability, Cigarette Harm Reduction and FDA Regulation. Clinical Pharmacology & Therapeutics, 111(5), 1004–1006. https://doi.org/10.1002/cpt.2555

Boakye, E., Osuji, N., Erhabor, J., Obisesan, O., Osei, A. D., Mirbolouk, M., Stokes, A. C., Dzaye, O., El Shahawy, O., Hirsch, G. A., Benjamin, E. J., DeFilippis, A. P., Robertson, R. M., Bhatnagar, A., & Blaha, M. J. (2022). Assessment of Patterns in e-Cigarette Use Among Adults in the US, 2017-2020. JAMA Network Open, 5(7), e2223266. https://doi.org/10.1001/jamanetworkopen.2022.23266

Branstetter, S. A., Blosnich, J., Dino, G., Nolan, J., & Horn, K. (2012). Gender differences in cigarette smoking, social correlates and cessation among adolescents. Addictive Behaviors, 37(6), 739–742. https://doi.org/10.1016/j.addbeh.2012.02.007

Brennan, E., Durkin, S. J., Wakefield, M. A., & Kashima, Y. (2014). Assessing the effectiveness of antismoking television advertisements: Do audience ratings of perceived effectiveness predict changes in quitting intentions and smoking behaviours? Tobacco Control, 23(5), 412–418. https://doi.org/10.1136/tobaccocontrol-2012-050949

CDC. (2023). Current Cigarette Smoking Among Adults in the United States. Centers for Disease Control and Prevention. https://www.cdc.gov/tobacco/data_statistics/fact_sheets/adult_data/cig_smoking/index.htm

Collins, L. K., Villanti, A. C., Pearson, J. L., Glasser, A. M., Johnson, A. L., Niaura, R. S., & Abrams, D. B. (2017). Frequency of Youth E-Cigarette, Tobacco, and Poly-Use in the United States, 2015: Update to Villanti et al., “Frequency of Youth E-Cigarette and Tobacco Use Patterns in the United States: Measurement Precision Is Critical to Inform Public Health.” Nicotine & Tobacco Research, 19(10), 1253–1254. https://doi.org/10.1093/ntr/ntx073

Cooper, M., Day, H. R., Ren, C., Oniyide, O., Corey, C. G., Ambrose, B. K., Michael Cummings, K., Sargent, J., Niaura, R., Pierce, J. P., Kaufman, A., Choi, K., Goniewicz, M. L., Stanton, C. A., Villanti, A., Kasza, K., Bansal-Travers, M., Silveira, M. L., Kimmel, H. L., … Hyland, A. J. (2022). Correlates of tobacco product initiation among youth and young adults between waves 1–4 of the population assessment of tobacco and Health (PATH) study (2013–2018). Addictive Behaviors, 134, 107396. https://doi.org/10.1016/j.addbeh.2022.107396

Cornelius, M., Loretan, C. G., Jamal, A., Davis Lynn, B., Mayer, M., alcantara, I., & Neff, L. J. (2023). Tobacco Product Use Among Adults – United States, 2021. CDC, 72(18), 9.

Dai, H., & Leventhal, A. M. (2019). Prevalence of e-Cigarette Use Among Adults in the United States, 2014-2018. JAMA, 322(18), 1824. https://doi.org/10.1001/jama.2019.15331

Erhabor, J., Boakye, E., Obisesan, O., Osei, A. D., Tasdighi, E., Mirbolouk, H., DeFilippis, A. P., Stokes, A. C., Hirsch, G. A., Benjamin, E. J., Rodriguez, C. J., El Shahawy, O., Robertson, R. M., Bhatnagar, A., & Blaha, M. J. (2023). E-Cigarette Use Among US Adults in the 2021 Behavioral Risk Factor Surveillance System Survey. JAMA Network Open, 6(11), e2340859. https://doi.org/10.1001/jamanetworkopen.2023.40859

Gulsen, A., Uslu, B., & Department of Chest Diseases, Yedikule Chest Diseases and Thoracic Surgery Ed. & Research Hospital, Istanbul, Turkey. (2019). Health hazards and complications associated with electronic cigarettes: A review. Turkish Thoracic Journal. https://doi.org/10.5152/TurkThoracJ.2019.180203

Holford, T. R., McKay, L., Jeon, J., Tam, J., Cao, P., Fleischer, N. L., Levy, D. T., & Meza, R. (2023). Smoking Histories by State in the U.S. American Journal of Preventive Medicine, 64(4), S42–S52. https://doi.org/10.1016/j.amepre.2022.08.018

Hrywna, M., Kong, A. Y., Ackerman, C., Giovenco, D. P., Spillane, T. E., Lee, J. G. L., Hudson, S. V., & Delnevo, C. D. (2023). Assessing the Effectiveness of Tobacco 21 Laws to Reduce Underage Access to Tobacco: Protocol for a Repeated Multi-Site Study. Methods and Protocols, 6(2), 27. https://doi.org/10.3390/mps6020027

Kasza, K. A., Edwards, K. C., Kimmel, H. L., Anesetti-Rothermel, A., Cummings, K. M., Niaura, R. S., Sharma, A., Ellis, E. M., Jackson, R., Blanco, C., Silveira, M. L., Hatsukami, D. K., & Hyland, A. (2021). Association of e-Cigarette Use With Discontinuation of Cigarette Smoking Among Adult Smokers Who Were Initially Never Planning to Quit. JAMA Network Open, 4(12), e2140880. https://doi.org/10.1001/jamanetworkopen.2021.40880

Kelsh, S., Ottney, A., Young, M., Kelly, M., Larson, R., & Sohn, M. (2023). Young Adults’ Electronic Cigarette Use and Perceptions of Risk. Tobacco Use Insights, 16, 1179173X2311613. https://doi.org/10.1177/1179173X231161313

Krosnick, J. A., Malhotra, N., Mo, C. H., Bruera, E. F., Chang, L., Pasek, J., & Thomas, R. K. (2017). Perceptions of health risks of cigarette smoking: A new measure reveals widespread misunderstanding. PLOS ONE, 12(8), e0182063. https://doi.org/10.1371/journal.pone.0182063

Lee, Y. H., Liu, Z., Fatori, D., Bauermeister, J. R., Luh, R. A., Clark, C. R., Bauermeister, S., Brunoni, A. R., & Smoller, J. W. (2022). Association of Everyday Discrimination With Depressive Symptoms and Suicidal Ideation During the COVID-19 Pandemic in the All of Us Research Program. JAMA Psychiatry, 79(9), 898. https://doi.org/10.1001/jamapsychiatry.2022.1973

Littlecott, H. J., Moore, G. F., Evans, R. E., Melendez-Torres, G. J., McCann, M., Reed, H., Mann, M., Dobbie, F., Jennings, S., Donaldson, C., & Hawkins, J. (2023). Perceptions of friendship, peers and influence on adolescent smoking according to tobacco control context: A systematic review and meta-ethnography of qualitative research. BMC Public Health, 23(1), 424. https://doi.org/10.1186/s12889-022-14727-z

Lovett, I. (2010). Californians are Smoking Less and Less. The New York Times. https://www.nytimes.com/2010/12/26/us/26smoking.html

Martinasek, M., Tamulevicius, N., Gibson-Young, L., McDaniel, J., Moss, S. J., Pfeffer, I., & Lipski, B. (2021). Predictors of Vaping Behavior Change in Young Adults Using the Transtheoretical Model: A Multi-Country Study. Tobacco Use Insights, 14, 1179173X2098867. https://doi.org/10.1177/1179173X20988672

Messeri, P., Cantrell, J., Mowery, P., Bennett, M., Hair, E., & Vallone, D. (2019). Examining differences in cigarette smoking prevalence among young adults across national surveillance surveys. PLOS ONE, 14(12), e0225312. https://doi.org/10.1371/journal.pone.0225312

National Center for Health Statistics. (2023). Current Electronic Cigarette Use Among Adults Aged 18 and Over: United States, 2021. CDC. https://www.cdc.gov/nchs/products/databriefs/db475.htm

Noar, S. M., Gottfredson, N. C., Kieu, T., Rohde, J. A., Hall, M. G., Ma, H., Fendinger, N. J., & Brewer, N. T. (2022). Impact of Vaping Prevention Advertisements on US Adolescents: A Randomized Clinical Trial. JAMA Network Open, 5(10), e2236370. https://doi.org/10.1001/jamanetworkopen.2022.36370

Obisesan, O. H., Osei, A. D., Uddin, S. M. I., Dzaye, O., Mirbolouk, M., Stokes, A., & Blaha, M. J. (2020). Trends in e-Cigarette Use in Adults in the United States, 2016-2018. JAMA Internal Medicine, 180(10), 1394. https://doi.org/10.1001/jamainternmed.2020.2817

Pang, Q., Wang, L., Yao, J., Yuen, K. F., Su, M., & Fang, M. (2023). Smoking cessation policy and treatments derived from the protective motivation of smokers: A study on graphic health warning labels. Frontiers in Psychology, 14, 1205321. https://doi.org/10.3389/fpsyg.2023.1205321

Pericot-Valverde, I., Heo, M., Litwin, A. H., Niu, J., & Gaalema, D. E. (2021). Modeling the effect of stress on vaping behavior among young adults: A randomized cross-over pilot study. Drug and Alcohol Dependence, 225, 108798. https://doi.org/10.1016/j.drugalcdep.2021.108798

PMI Science. (2021). Relative risk: Providing critical context for a better understanding. Philip Morris International. https://www.pmiscience.com/en/news-events/scientific-update-magazine/relative-risk/

Ribisl, K. M., & Mills, S. D. (2019). Explaining the Rapid Adoption of Tobacco 21 Policies in the United States. American Journal of Public Health, 109(11), 1483–1485. https://doi.org/10.2105/AJPH.2019.305269

Rohde, J. A., Noar, S. M., Prentice-Dunn, H., Kresovich, A., & Hall, M. G. (2021). Comparison of Message and Effects Perceptions for The Real Cost E-Cigarette Prevention Ads. Health Communication, 36(10), 1222–1230. https://doi.org/10.1080/10410236.2020.1749353

Rothman, K. J., Lash, T. L., & Greenland, S. (2008). Modern Epidemiology (Third Edition). Lippincott Williams & Wilkins.

Sargent, J. D., Stoolmiller, M., Dai, H., Barrington-Trimis, J. L., McConnell, R., Audrain-McGovern, J., & Leventhal, A. M. (2022). First E-Cigarette Flavor and Device Type Used: Associations With Vaping Persistence, Frequency, and Dependence in Young Adults. Nicotine & Tobacco Research, 24(3), 380–387. https://doi.org/10.1093/ntr/ntab172

Snell, W., & Goetz, S. (n.d.). Overview of Kentucky’s Tobacco Economy. AEC. https://www2.ca.uky.edu/agcomm/pubs/aec/aec83/aec83.pdf

Tenny, S., Brannan, J. M., & Brannan, G. D. (2024). Qualitative Study. In StatPearls. StatPearls Publishing. http://www.ncbi.nlm.nih.gov/books/NBK470395/

Trotter, L. (2002). Socially cued smoking in bars, nightclubs, and gaming venues: A case for introducing smoke-free policies. Tobacco Control, 11(4), 300–304. https://doi.org/10.1136/tc.11.4.300

Truth Initiative. (2023). Tobacco Nation: A Call to Eliminate Geographic Smoking Disparities in the U.S. Truth Initiative Inspiring Lives Free From Smoking, Vaping & Nicotine. https://truthinitiative.org/tobacconation

U.S. Department of Commerce. (2023). Census Regions and Divisions of the United States. Geography Division. https://www2.census.gov/geo/pdfs/maps-data/maps/reference/us_regdiv.pdf

U.S. Department of Health and Human Services. (2021). The Health Consequences of Smoking—50 Years of Progress: A report of the Surgeon General. CDC. https://www.hhs.gov/sites/default/files/consequences-smoking-exec-summary.pdf

Verbeek, M. J. A., Hommes, M. A., Stutterheim, S. E., Van Lankveld, J. J. D. M., & Bos, A. E. R. (2020). Experiences with stigmatization among transgender individuals after transition: A qualitative study in the Netherlands. International Journal of Transgender Health, 21(2), 220–233. https://doi.org/10.1080/26895269.2020.1750529

Wackowski, O., Hammond, D., O’Connor, R., Strasser, A., & Delnevo, C. (2016). Smokers’ and E-Cigarette Users’ Perceptions about E-Cigarette Warning Statements. International Journal of Environmental Research and Public Health, 13(7), 655. https://doi.org/10.3390/ijerph13070655

Wadsworth, E., Neale, J., McNeill, A., & Hitchman, S. (2016). How and Why Do Smokers Start Using E-Cigarettes? Qualitative Study of Vapers in London, UK. International Journal of Environmental Research and Public Health, 13(7), 661. https://doi.org/10.3390/ijerph13070661

White Hughto, J. M., Reisner, S. L., & Pachankis, J. E. (2015). Transgender stigma and health: A critical review of stigma determinants, mechanisms, and interventions. Social Science & Medicine, 147, 222–231. https://doi.org/10.1016/j.socscimed.2015.11.010

WHO. (2024). Current Smoking of any Tobacco Product. The Global Health Observatory. https://www.who.int/data/gho/indicator-metadata-registry/imr-details/346

Yang, B., Owusu, D., & Popova, L. (2019). Testing messages about comparative risk of electronic cigarettes and combusted cigarettes. Tobacco Control, 28(4), 440–448. https://doi.org/10.1136/tobaccocontrol-2018-054404

 

 

TABLES AND FIGURES

Table 1. Sociodemographic characteristics according to current tobacco use status among young adults living in the US (N=2,806).
	Overall	Not Current User	Current User	p
Sample size	N = 2,806	n = 2,363 (84.21%)	n = 443 (15.79%)
Race				0.24
White	1780 (63.44%)	1510 (63.90%)	270 (60.95%)
BIPOC	1026 (36.56%)	853 (36.10%)	173 (39.05%)
Hispanix/Latinx				0.48
Yes	376 (13.40%)	312 (13.20%)	64 (14.45%)
No	2430 (86.60%)	2051 (86.80%)	379 (85.55%)
Age				0.03
18-24	920 (32.79%)	795 (33.64%)	125 (28.22%)
25-35	1886 (67.21%)	1568 (66.36%)	318 (71.78%)
Gender				<0.001
Cisgender Male	1191 (42.44%)	967 (40.92%)	224 (50.56%)
Cisgender Female	1357 (48.36%)	1168 (49.43%)	189 (42.66%)
Transgender/NB+	258 (9.19%)	228 (9.65%)	30 (6.77%)
Income				<0.001
≥$50,000	746 (26.59%)	598 (25.31%)	148 (33.41%)
$20,000-49,999	1003 (35.74%)	827 (35.00%)	176 (39.73%)
<$20,000	1057 (37.67%)	938 (39.70%)	119 (26.86%)
Community Type				<0.001
Large City	833 (32.81%)	672 (28.44%)	161 (36.34%)
Suburb near large city	1049 (41.32%)	910 (38.51%)	139 (31.38%)
Small city, town	657 (25.88%)	571 (24.16%)	86 (19.41%)
Rural area	267 (10.52%)	210 (8.89%)	57 (12.87%)
Marital Status				0.002
Single/Divorced	1442 (51.39%)	1238 (52.39%)	204 (46.05%)
Partnered, cohabiting	581 (20.71%)	466 (19.72%)	115 (25.96%)
Partnered, not cohabiting	306 (10.91%)	268 (11.34%)	38 (8.58%)
Married	477 (17.00%)	391 (16.55%)	86 (19.41%)
Region				0.11
Northeast	471 (16.79%)	383 (16.21%)	88 (19.86%)
West	606 (21.60%)	526 (22.26%)	80 (18.06%)
South	1129 (40.24%)	950 (40.20%)	179 (40.41%)
Midwest	600 (21.38%)	504 (21.33%)	96 (21.67%)
Education				0.49
<High school/HS/Certificate	624 (22.24%)	531 (22.47%)	93 (20.99%)
Some College and above	2182 (77.76%)	1832 (77.53%)	350 (79.01%)
Non-users Vaping motivation				0.03
Non-motivating	2,180 (77.69%)	1,853 (78.42%)	327 (73.81%)
Motivating	626 (22.31%)	510 (21.58%)	116 (26.19%)
Non-users Cigarette Motivation				<0.001
Non-motivating	2,604 (92.80%)	2215 (93.74%)	389 (87.81%)
Motivating	202 (7.20%)	148 (6.26%)	54 (12.19%)

 

 

Table 2. Sociodemographic characteristics according to region lived in among young adults living in the US (N=2,806).
	Overall	Northeast	West	South	Midwest	p
Sample size	N = 2,806	n = 471 (16.79%)	n= 606 (21.60%)	n= 1,129 (40.24%)	n= 600 (21.38%)
Smoking status						0.11
Not Current User	2,363 (84.21%)	383 (81.32%)	526 (86.80%)	950 (84.15%)	504 (84.00%)
Current User	443 (15.79%)	88 (18.68%)	80 (13.20%)	179 (15.85%)	96 (16.00%)
Race						<0.001
White	1780 (63.44%)	303 (64.33%)	322 (53.14%)	709 (62.80%)	446 (74.33%)
BIPOC	1026 (36.56%)	168 (35.67%)	284 (46.86%)	420 (37.20%)	154 (25.67%)
Hispanix/Latinx						<0.001
Yes	376 (13.40%)	60 (12.74%)	118 (19.47%)	149 (13.20%)	49 (8.17%)
No	2430 (86.60%)	411 (57.26%)	488 (80.53%)	980 (86.80%)	551(91.83%)
Age						0.70
18-24	920 (32.79%)	147 (31.21%)	199 (32.84%)	383 (33.92%)	191 (31.83%)
25-35	1886 (67.21%)	324 (68.79%)	407 (67.16%)	746 (66.08%)	409 (68.17%)
Gender						0.27
Cisgender Male	1191 (42.44%)	194 (41.19%)	282 (46.53%)	474 (41.98%)	241 (40.17%)
Cisgender Female	1357 (48.36%)	230 (48.83%)	266 (43.89%)	554 (49.07%)	307 (51.17%)
Transgender/NB+	258 (9.19%)	47 (9.98%)	58 (9.57%)	101 (8.95%)	52 (8.67%)
Income						0.03
≥$50,000	746 (26.59%)	149 (31.63%)	179 (29.54%)	273 (24.18%)	145 (24.17%)
$20,000-49,999	1003 (35.74%)	157 (33.33%)	202 (33.33%)	419 (37.11%)	225 (37.50%)
<$20,000	1057 (37.67%)	165 (35.03%)	225 (37.13%)	437 (38.71%)	230 (38.33%)
Community Type						<0.001
Large City	833 (32.81%)	154 (32.70%)	223 (36.80%)	292 (25.86%)	164 (27.33%)
Suburb near large city	1049 (41.32%)	162 (34.39%)	239 (39.44%)	418 (37.02%)	230 (38.33%)
Small city, town	657 (25.88%)	107 (22.72%)	116 (19.14%)	276 (24.45%)	158 (26.33%)
Rural area	267 (10.52%)	48 (10.19%)	28 (4.62%)	143 (12.67%)	48 (8.00%)
Marital Status						0.24
Single/Divorced	1442 (51.39%)	235 (49.89%)	338 (55.78%)	582 (51.55%)	287 (47.83%)
Partnered, cohabit	581 (20.71%)	102 (21.66%)	111 (18.32%)	229 (20.28%)	139 (23.17%)
Partnered, not cohabiting	306 (10.91%)	60 (12.74%)	62 (10.23%)	121 (10.72%)	63 (10.50%)
Married	477 (17.00%)	74 (15.71%)	95 (15.68%)	197 (17.45%)	111 (18.50%)
Education						0.01
<High school/HS/Certificate	624 (22.24%)	97 (20.59%)	112 (18.48%)	257 (22.76%)	158 (26.33%)
Some College and above	2182 (77.76%)	374 (79.41%)	494 (81.52%)	872 (77.24%)	442 (73.67%)
Non-users Vaping motivation						0.25
Non-motivating	2,180 (77.69%)	364 (77.28%)	479 (79.04%)	888 (78.65%)	449 (74.83%)
Motivating	626 (22.31%)	107 (22.72%)	127 (20.96%)	241 (21.35%)	151 (25.17%)
Non-users Cigarette Motivation						0.43
Non-motivating	2,604 (92.80%)	433 (91.93%)	565 (93.23%)	1056 (93.53%)	550 (91.67%)
Motivating	202 (7.20%)	38 (8.07%)	41 (6.77%)	73 (6.47%)	50 (8.33%)

 

 

	Table 3.  Sociodemographic characteristics according to effects perceptions around non-users’ motivation to begin vaping among young adults living in the US (N=2,806).
		Full sample	Believe that messages will not motivate non-users to start vaping	Believe that messages will motivate non-users to start vaping		p
	Sample size	N = 2,806	n = 2,180 (77.69%)	n = 626 (22.31%)
	Smoking status					0.03
	Not Current User	2,363 (84.21%)	1,853(85.00%)	510 (81.47%)
	Current User	443 (15.79%)	327 (15.00%)	116 (18.53%)
	Region					0.25
	Northeast	471 (16.79%)	364 (16.70%)	107 (17.09%)
	West	606 (21.60%)	479 (21.97%)	127 (20.29%)
	South	1129 (40.24%)	888 (40.73%)	241 (38.50%)
	Midwest	600 (21.38%)	449 (20.60%)	151 (24.12%)
	Race					0.57
	White	1780 (63.44%)	1389 (63.72%)	391 (62.46%)
	BIPOC	1026 (36.56%)	791 (36.28%)	235 (37.54%)
	Hispanix/Latinx					0.78
	Yes	376 (13.40%)	290 (13.30%)	86 (13.74%)
	No	2430 (86.60%)	1890 (86.70%)	540 (86.26%)
	Age					0.51
	18-24	920 (32.79%)	708 (32.48%)	212 (33.87%)
	25-35	1886 (67.21%)	1472 (67.52%)	414 (66.13%)
	Gender					0.85
	Cisgender Male	1191 (42.44%)	931 (42.71%)	260 (41.53%)
	Cisgender Female	1357 (48.36%)	1051 (48.21%)	306 (48.88%)
	Transgender/NB+	258 (9.19%)	198 (9.08%)	60 (9.58%)
	Income					0.82
	≥$50,000	746 (26.59%)	580 (26.61%)	166 (26.52%)
	$20,000-49,999	1003 (35.74%)	785 (36.01%)	218 (34.82%)
	<$20,000		815 (37.39%)	242 (38.66%)
	Community Type					0.38
	Large City	833 (32.81%)	636 (32.28%)	197 (34.62%)
	Suburb near large city	1049 (41.32%)	809 (41.07%)	240 (42.18%)
	Small city, town	657 (25.88%)	525 (26.65%)	132 (23.20%)
	Rural area	267 (10.52%)	210 (10.66%)	57 (10.02%)
	Marital Status					0.19
	Single/Divorced	1442 (51.39%)	1116 (51.19%)	326 (52.08%)
	Partnered, cohabiting	581 (20.71%)	450 (20.64%)	131 (20.93%)
	Partnered, not cohabiting	306 (10.91%)	252 (11.56%)	54 (8.63%)
	Married	477 (17.00%)	362 (16.61%)	115 (18.37%)
	Education					0.08
	<High school/HS/Certificate	624 (22.24%)	501 (22.98%)	123 (19.65%)
	Some College and above	2182 (77.76%)	1679 (77.02%)	503 (80.35%)
Table 4.  Sociodemographic characteristics according to effects perceptions around non-users’ motivation to begin cigarette smoking among young adults living in the US (N=2,806).
		Full sample	Messages WILL NOT motivate non-users to start cigarette smoking		Messages WILL motivate non-users to start cigarette smoking		P
Sample size		N = 2,806	n = 2,604 (92.80%)		n = 202 (7.20%)
Smoking status							<0.001
Not Current User		2363 (84.21%)	2215 (85.06%)		148 (73.27%)
Current User		443 (15.79%)	389 (14.94%)		54 (26.73%)
Region							0.43
Northeast		471 (16.79%)	433 (16.63%)		38 (18.81%)
West		606 (21.60%)	565 (21.70%)		41 (20.30%)
South		1129 (40.24%)	1056 (40.55%)		73 (36.14%)
Midwest		600 (21.38%)	550 (21.12%)		50 (24.75%)
Race							0.01
White		1780 (63.44%)	1670 (64.13%)		110 (54.46%)
BIPOC		1026 (36.56%)	934 (35.87%)		92 (45.54%)
Hispanix/Latinx							0.99
Yes		376 (13.40%)	349 (13.40%)		27 (13.37%)
No		2430 (86.60%)	2255 (86.60%)		175 (86.63%)
Age							0.42
18-24		920 (32.79%)	859 (32.99%)		61 (30.20%)
25-35		1886 (67.21%)	1745 (67.01%)		141 (69.80%)
Gender							<0.001
Cisgender Male		1191 (42.44%)	1076 (41.32%)		115 (56.93%)
Cisgender Female		1357 (48.36%)	1279 (49.12%)		78 (38.61%)
Transgender/NB+		258 (9.19%)	249 (9.56%)		9 (4.46%)
Income							0.01
≥$50,000		746 (26.59%)	675 (25.92%)		71 (35.15%)
$20,000-49,999		1003 (35.74%)	935 (35.91%)		68 (33.66%)
<$20,000		1057 (37.67%)	994 (38.17%)		63 (31.19%)
Community Type							0.15
Large City		833 (32.81%)	759 (32.24%)		74 (40.00%)
Suburb near large city		1049 (41.32%)	978 (41.55%)		71 (38.38%)
Small city, town		657 (25.88%)	617 (26.21%)		40 (21.62%)
Rural area		267 (10.52%)	250 (10.62%)		17 (9.19%)
Marital Status							0.04
Single/Divorced		1442 (51.39%)	1321 (50.73%)		121 (59.90%)
Partnered, cohabiting		581 (20.71%)	545 (20.93%)		36 (17.82%)
Partnered, not cohabiting		306 (10.91%)	293 (11.25%)		13 (6.44%)
Married		477 (17.00%)	445 (17.09%)		32 (15.84%)
Education							0.87
<High school/HS/Certificate		624 (22.24%)	580 (22.27%)		44 (21.78%)
Some College and above		2182 (77.76%)	2024 (77.73%)		158 (78.22%)

 

Table 5. Binary Logistic regression model of the association between current smoking status/region and effects perception around non-users motivation to begin vaping among young adults living in the US (N=2,806)
	AOR	95% CI
Smoking status
Not Current User	Ref	Ref
Current User	1.29*	(1.02 – 1.93)
Region
Northeast	Ref	Ref
West	0.88	(0.66 – 1.18)
South	0.92	(0.71 – 1.20)
Midwest	1.16	(0.87 – 1.55)
Black, Indigenous, or other person of color	1.06	(0.87 – 1.30)
Hispanic/Latinx	1.03	(0.78 – 1.36)
Age 25-35	0.90	(0.73 – 1.10)
Gender
Cisgender Male	Ref	Ref
Cisgender Female	1.04	(0.86 – 1.26)
Transgender/NB+	1.08	(0.78 – 1.51)
Income
≥$50,000	Ref	Ref
$20,000-49,999	1.04	(0.82 – 1.32)
<$20,000	1.17	(0.90 – 1.50)
Community Type
Large City	Ref	Ref
Suburb near large city	0.97	(0.78 – 1.20)
Small city, town	0.81	(0.63 – 1.05)
Rural area	0.90	(0.64 – 1.27)
Marital Status
Single/Divorced	Ref	Ref
Partnered, cohabiting	1.01	(0.79 – 1.28)
Partnered, not cohabiting	0.71*	(0.52 – 0.99)
Married	1.15	(0.88 – 1.50)
Some College and above	1.24	(0.99 – 1.56)

(Note: * p<0.05)

 

 

 

Table 6. Binary Logistic regression model of the association between current smoking status/region and effects perception around non-users motivation to begin cigarette smoking among young adults living in the US (N=2,806)
	AOR	95% CI
Smoking status
Not Current User	Ref	Ref
Current User	1.89***	(1.36 – 2.63)
Region
Northeast	Ref	Ref
West	0.80	(0.50 – 1.27)
South	0.82	(0.54 – 1.25)
Midwest	1.13	(0.72 – 1.78)
Black, Indigenous, or other person of color	1.33	(0.96 – 1.83)
Hispanic/Latinx	0.86	(0.55 – 1.36)
Age 25-35	0.97	(0.68 – 1.38)
Gender
Cisgender Male	Ref	Ref
Cisgender Female	0.66**	(0.49 – 0.90)
Transgender/NB+	0.41**	(0.20 – 0.83)
Income
≥$50,000	Ref	Ref
$20,000-49,999	0.72	(0.51 – 1.03)
<$20,000	0.70	(0.47 – 1.03)
Community Type
Large City	Ref	Ref
Suburb near large city	0.79	(0.26 – 1.11)
Small city, town	0.79	(0.52 – 1.19)
Rural area	0.83	(0.47 – 1.47)
Marital Status
Single/Divorced	Ref	Ref
Partnered, cohabiting	0.73	(0.49 – 1.11)
Partnered, not cohabiting	0.52*	(0.29 – 0.95)
Married	0.70	(0.45 – 1.08)
Some College and above	0.96	(0.67 -1.39)

(Note: * p<0.05 **p<0.01 ***p<0.001)

 

 

Appendix A: Literature Review

U.S. CURRENT SMOKING PREVALENCE

Cigarette smoking remains within the United States as the leading cause of preventable death. Cigarette smoking accounts for more than 480,000 deaths each year, which is about 1 in 5 deaths in the U.S. (U.S. Department of Health and Human Services, 2021). Within the United States, “current” cigarette smoking status is defined as people who reported smoking at least 100 cigarettes during their lifetime and who smoke either daily or occasionally at the time of survey (WHO, 2024). In 2021, an estimated 28.3 million U.S. adults were currently smoking, which is approximately 12 of every 100 adults (Cornelius et al., 2023). As of now, current cigarette smoking is higher among men than women. Approximately 13% of adult men are cigarette smoking while approximately 10% of adult women are smoking (Cornelius et al., 2023). Current smoking status around the United States also varies by race/ethnicity, education level, income, U.S. census region data, marital status, sexual orientation, and by age (Cornelius et al., 2023).  According to research, current cigarette smoking is highest among people aged 25-44 years of age and is lowest among people aged 18-24 years (CDC, 2023). However, this decrease in young adult’s cigarette smoking may be misleading or overly optimistic as young adults are shown to be transitioning to smoking other forms of tobacco products, such as e-cigarettes (Martinasek et al., 2021).

Continuous efforts within the United States to reduce the initiation of cigarette smoking, especially everyday cigarette smoking, remains a primary concern. One key target group that the U.S. is attempting to stop or never begin the use cigarettes is the young adult population (Barrington-Trimis et al., 2020). According to research, 87% of adults who had ever smoked cigarettes daily had tried their first cigarette by the age of 18 (Hrywna et al., 2023; U.S. Department of Health and Human Services, 2021). In March 2015, a report from the National Academy of Medicine showed that the idea “Tobacco 21” could prevent 223,000 deaths from smoking and provided insightful scientific information to help begin advocacy to increase the age of tobacco sales from 18 to 21 (American Lung Association, 2022). The passage of this law was a monumental moment in public health. A substantial proportion of beginning smokers are currently young adults, this is a shift from what was previously shown as adolescence. Being able to visibly see that shifting the law to 21 has made a significant impact of the initiation of cigarette smoking (Barrington-Trimis et al., 2020), giving public health educators more time to explain the risks of cigarette smoking to the population and prevent more individuals from smoking initiation. However, the extent to which Tobacco 21 will continue to reduce youth access to tobacco products depends on consistent implementation, as well as preventing older individuals from selling tobacco products to young adults below age 21 (Hrywna et al., 2023).

 

U.S. CURRENT E-CIGARETTE PREVALENCE

E-cigarettes (better known as vapes) are the second most commonly used tobacco product among U.S. adults (Erhabor et al., 2023). A continuous trend being observed across studies has shown an increase in the proportion of individuals indicating that they use e-cigarettes daily.  Current e-cigarette use is based on a response of “every day” or “some days” to the question, “Do you now use e-cigarettes or other electronic vaping products every day, some days, or not at all?” This survey question can be asked of adults who had ever tried an e-cigarette, even one time (National Center for Health Statistics, 2023). These recent findings suggest a potential shift from the experimental usage of e-cigarettes to an established use (Boakye et al., 2022; Dai & Leventhal, 2019). Nationally, 4.5% of adults aged 18 and over are current e-cigarette users.  Recent smaller surveys have indicated that the national number of individuals smoking e-cigarettes is expected to increase in upcoming years (Dai & Leventhal, 2019). Current e-cigarette usage within the United States varies by race/ethnicity, income, and age. White non-Hispanic adults are more likely to be current e-cigarette users than Asian non-Hispanic, Black, and Hispanic adults (National Center for Health Statistics, 2023). In 2021, people aged 18-24 were found to be most likely to use e-cigarettes among all adults (National Center for Health Statistics, 2023). Research shows 11 out of 100 young adults aged 18-24 are currently using e-cigarettes.

While e-cigarettes are seen in some age group populations to serve as a smoking cessation aid, for young adults who have had no prior exposure to combustible cigarettes, e-cigarette usage remains a public health concern (Kasza et al., 2021; Obisesan et al., 2020). The recent passage of the Tobacco 21 law also applies to e-cigarettes. The benefits that cigarette smoking reduction has seen in recent years provides a hopeful ideation that e-cigarette usage will follow a decrease in initiation by young adults. However, the tobacco 21 law holds the same as with cigarette smoking. The extent to which tobacco 21 can cause reduction in e-cigarette usage greatly depends on consistent implementation (Hrywna et al., 2023). Young adults aged 18-35 will still need to be provided educational materials as for the dangers e-cigarettes pose to one’s health, and prevention interventions to keep young adults away from e-cigarettes (Gulsen et al., 2019).

 

ANTI-TOBACCO AD CAMPAIGNS

Tobacco educational advertisements are marketing strategies created through public health officials to highlight the importance of tobacco cessation. Tobacco educational advertisements are broken up into various categories depending on what messages surrounding tobacco are trying to be displayed. One form of tobacco educational advertisements is absolute risk messaging. This strategy for messaging provokes the tobacco user to end all forms of tobacco products, for example, “end all smoking of cigarettes and e-cigarettes” (Krosnick et al., 2017). Relative risk messages, however, provoke the tobacco user to switch completely to vaping only instead of smoking cigarettes or smoking both vapes and cigarettes. An example of relative risk messaging would be “protect your health, switch completely from cigarette smoking to vaping only” (Yang et al., 2019).

There have been mixed findings throughout the literature about the overall effectiveness of absolute risk messages vs. relative risk messages (Barnes et al., 2017; Wackowski et al., 2016; Yang et al., 2019). However, there are advantages to both messaging types depending on what prevention methods are to be targeted. Absolute risk messaging in theory is the best way to minimize a current smoker’s risk of smoking related disease as they would be informed to quit tobacco and all forms of nicotine. However, absolute risk messaging relies heavily on alternative therapies that could cause the smoker to quit all forms of nicotine, which for many current smokers is not an option that they are willing to do (PMI Science, 2021). Contrarily, relative risk helps provide informative comparisons between cigarettes and other nicotine products. Relative risk messages can help a current smoker understand the reduced risk of completely switching to another less harmful nicotine product. While this type of messaging has a rate reduction or harm reduction positive message in play, there are disadvantages of utilizing relative risk messages as well (PMI Science, 2021). Relative risk messages can also give smokers a false sense that switching to another form of nicotine will then have low to no health risks involved, which is not supported by research around other nicotine products. Additionally, switching products may still find a current smoker unable to quit the new nicotine product (PMI Science, 2021). While reducing harm from the original cigarettes, it is still not in a current smoker’s best interest to then smoke the new nicotine product continuously for the lifetime duration.

In previous years, the research around how these campaigns were affecting tobacco cessation were done through message perceptions. Message perceptions are judgements about whether a message promotes further processing that leads to persuasion (Noar et al., 2022).  Examples of message perceptions are, “This ad is attention-grabbing” or, “This ad is informative”. However, in recent studies there has been evidence to shift away from analyzing through message perceptions and to analyze tobacco educational messages through effects perceptions (Noar et al., 2022). Effects perceptions are judgements about a message’s potential to change antecedents of behavior. Examples of effects perceptions are, “This ad would motivate me not to vape” or, “This ad motivates me not to smoke” (Noar et al., 2022). Through literature it has been found that message perceptions are not as effective as effects perceptions on understanding the entire thought process of participants when viewing tobacco educational messages. Additionally, message perceptions were found to have greater measurement error than effects perceptions scales due to evidence about how message perceptions measure typically contain “noise” when participants make message ratings (Noar et al., 2022). This finding has also been shown to have similar findings in other studies; other studies have shown that only effects perceptions predicted actual effectiveness outcomes after adjusting for message perceptions, while message perceptions exhibited counterintuitive results after adjusting for effects perceptions (Baig et al., 2021; Brennan et al., 2014; Rohde et al., 2021).

 

REGIONAL VARIATIONS IN SMOKING

According to U.S. census data from the CDC, the United States is separated into four regions, the South, Midwest, West, and Northeast. Within the U.S. current smoking varies by region lived in. Current cigarette smoking in 2021 was found to be highest in the Midwest and the South and lowest in the West (CDC, 2023). For example, West Virginia falls within the Southern region of the United States and in 2022 had the highest smoking rate in the U.S. at 23.8% (Bascom, 2022) and Kentucky, in the Southern region as well is not far behind, where 23.6% of adults are currently smoking. Utah resides in the Western region of the United States and in 2022 had the lowest smoking rate, with less than 10% of the state’s population currently smoking (Bascom, 2022) and California, also in the west, has 7.6% of adults currently smoking.

There are clear smoking disparities among regions lived in for the U.S. population. Research shows that having protections in place such as smoke-free laws, higher tobacco taxes, and the tobacco 21 law all help to lower smoking initiation and smoking rates (Truth Initiative, 2023). States that have significantly fewer laws and protections in place are shown to have higher rates of smoking and smoking initiation. There are various reasons as to why regionality may be playing a role in smoking and initiation; social and cultural differences across the country may contribute to different attitudes towards tobacco use and passing laws surrounding tobacco control policies (Holford et al., 2023). For example, individuals living in a state where tobacco is more commonplace may not understand the harms around tobacco smoking or may face greater peer pressure from friends and family members to begin smoking at a younger age. Individuals living in states that have lower smoking rates may face criticism for beginning smoking, may not have access to outdoor smoking areas, or live in a place where it is not legal to smoke indoors. Additionally, states have varying agendas as to why laws are passed around tobacco control. For example, in Kentucky tobacco has historically been a very important part of the state’s agricultural economy and culture. Census of Agriculture revealed that tobacco accounted for more than 40% of the net cash return from agricultural sales in Kentucky (Snell & Goetz, n.d.). Comparatively to California, a state that has one of the lowest tobacco rates, shows that an increase on taxes for tobacco products, combined with anti-smoking media campaigns has helped the state remain low in smoking initiation and rates. From this, the state has seen lung cancer rates drop more than three times as fast than in the rest of the country, saving the state an estimated $86 billion in health care costs (Lovett, 2010). These driving differences in monetary state impact can be another reason as to why tobacco initiation and rates are varied across regions within the United States.

Appendix B: Supplemental figures

SENSITIVITY ANALYSIS

Table 7. Binary Logistic regression model of the association between ever smoker vs. never smoker status/region and effects perception around non-users motivation to begin vaping among young adults living in the US (N=2,806)
	AOR	95% CI
Smoking status
Never Smoker	Ref	Ref
Ever Smoker	1.04	(0.86 – 1.25)
Region
Northeast	Ref	Ref
West	0.87	(0.65 – 1.17)
South	0.92	(0.71 – 1.20)
Midwest	1.16	(0.87 – 1.54)
Black, Indigenous, or other person of color	1.08	(0.89 – 1.31)
Hispanic/Latinx	1.03	(0.78 – 1.36)
Age 25-35	0.90	(0.73 – 1.10)
Gender
Cisgender Male	Ref	Ref
Cisgender Female	1.03	(0.89 – 1.30)
Transgender/NB+	1.07	(0.78 – 1.51)
Income
≥$50,000	Ref	Ref
$20,000-49,999	1.03	(0.82 – 1.32)
<$20,000	1.15	(0.90 – 1.50)
Community Type
Large City	Ref	Ref
Suburb near large city	0.96	(0.78 – 1.20)
Small city, town	0.81	(0.63 – 1.05)
Rural area	0.91	(0.64 – 1.27)
Marital Status
Single/Divorced	Ref	Ref
Partnered, cohabiting	1.01	(0.79 – 1.28)
Partnered, not cohabiting	0.71*	(0.52 – 0.99)
Married	1.15	(0.89 – 1.51)
Some College and above	1.24	(0.99 – 1.56)

(Note: * p<0.05 **p<0.01 ***p<0.001)

 

 

 

Table 8. Binary Logistic regression model of the association between ever smoker vs. never smoker status/region and effects perceptions strongly disagreeing vs. everyone else around non-users motivation to begin vaping among young adults living in the US (N=2,806)
	AOR	95% CI
Smoking status
Never Smoker	Ref	Ref
Ever Smoker	1.02	(0.86 – 1.20)
Region
Northeast	Ref	Ref
West	0.95	(0.74 – 1.22)
South	0.92	(0.74 – 1.14)
Midwest	1.03	(0.80 – 1.32)
Black, Indigenous, or other person of color	1.22*	(1.04 – 1.44)
Hispanic/Latinx	1.03	(0.78 – 1.36)
Age 25-35	0.91	(0.77 – 1.09)
Gender
Cisgender Male	Ref	Ref
Cisgender Female	0.96	(0.81 – 1.13)
Transgender/NB+	1.18	(0.89 – 1.57)
Income
≥$50,000	Ref	Ref
$20,000-49,999	1.14	(0.82 – 1.32)
<$20,000	1.26*	(1.01 – 1.57)
Community Type
Large City	Ref	Ref
Suburb near large city	0.96	(0.78 – 1.20)
Small city, town	0.81	(0.63 – 1.05)
Rural area	0.95	(0.64 – 1.27)
Marital Status
Single/Divorced	Ref	Ref
Partnered, cohabiting	0.98	(0.79 – 1.28)
Partnered, not cohabiting	0.81	(0.62 – 1.04)
Married	1.09	(0.89 – 1.38)
Some College and above	1.37**	(1.13 – 1.66)

 

(Note: * p<0.05 **p<0.01 ***p<0.001)

 

 

 

Table 9. Binary Logistic regression model of the association between ever vs. never smoking status/region and effects perceptions strongly disagreeing vs. everyone else around non-users motivation to begin cigarette smoking among young adults living in the US (N=2,806)
	AOR	95% CI
Smoking status
Never Smoker	Ref	Ref
Ever Smoker	1.14	(0.93 – 1.41)
Region
Northeast	Ref	Ref
West	0.92	(0.74 – 1.22)
South	0.93	(0.74 – 1.14)
Midwest	0.98	(0.80 – 1.32)
Black, Indigenous, or other person of color	1.34**	(1.08 – 1.66)
Hispanic/Latinx	1.03	(0.78 – 1.36)
Age 25-35	0.89	(0.71 – 1.12)
Gender
Cisgender Male	Ref	Ref
Cisgender Female	0.75**	(0.61 – 0.93)
Transgender/NB+	0.69	(0.47 – 1.03)
Income
≥$50,000	Ref	Ref
$20,000-49,999	0.80	(0.62 – 1.02)
<$20,000	0.83	(0.63 – 1.09)
Community Type
Large City	Ref	Ref
Suburb near large city	0.92	(0.78 – 1.20)
Small city, town	0.73*	(0.55 – 0.98)
Rural area	0.93	(0.64 – 1.37)
Marital Status
Single/Divorced	Ref	Ref
Partnered, cohabiting	0.86	(0.65 – 1.14)
Partnered, not cohabiting	0.67*	(0.47 – 0.97)
Married	1.03	(0.77 – 1.37)
Some College and above	1.16	(0.90 – 1.49)

(Note: * p<0.05 **p<0.01 ***p<0.001)

 

 

GOODNESS OF FIT

 

Table 10. Hosmer-Lemeshow GOF test after logistic model of vaping motivation
Number of Observation	Number of groups	Hosmer-Lemeshow chi2(8)	Prob > chi2
2,806	10	9.31	0.3170

Since the p-value is 0.3170 we can conclude that the model provides adequate fit to the data.

 

 

Table 11. Hosmer-Lemeshow GOF test after logistic model of cigarette smoking motivation
Number of Observation	Number of groups	Hosmer-Lemeshow chi2(8)	Prob > chi2
2,806	10	3.50	0.8992

Since the p-value is 0.8992 we can conclude that the model provides adequate fit to the data.

 

 

Figure 1. Variables for Analysis from Parent Study

Variables for Analysis	Question	Response Options
Current User	“Do you currently smoke cigarettes every day, some days, or not at all?” AND “Do you currently use e-cigarettes or electronic nicotine vapes every day, some days, or not at all? “	Everyday some days Not at all
Region	“What US state, district, or territory do you currently reside in?”	Drop down menu to select from of all U.S. States, districts, and territories.   OR selection choice: “I do not live in the United States”
Non-users Vaping Motivation	“Please rate each statement from 1(strongly disagree) to 5(strongly agree)”	These messages will motivate NON-USERS to start vaping
Non-users Cigarette Motivation	“Please rate each statement from 1(strongly disagree) to 5(strongly agree)”	These messages will motivate NON-USERS to start smoking cigarettes.
Education	“What is your highest grade completed?”	11th grade High school diploma or GED Technical school (Vo Tech, Career Certificate, etc) Some college (not graduated) Associate’s Degree Bachelor’s Degree Master’s Degree Doctoral Degree or other terminal Professional Degree (e.g. MD, JD)
Gender	“What is your gender?”	Female Male Transgender female/Transgender woman Transgender male/Transgender man Non-binary Gender queer Agender I use different words to describe my gender
Marital Status	“What is your marital status?”	Single, never married Dating or partnered, and living together Dating or partnered, but not living together Married, and living together Married, but not living together Divorced Widowed
Hispanix/Latinx	“What racial and/or ethnic groups do you identify with? (Check all that apply)”	Asian Black/African American Hispanic/Latinx/Latino/Latina Middle Eastern Native American or Alaskan Native Pacific Islander White/Caucasian I use other words to describe my race and ethnicity
Community Type	“What type of Community do you live in?”	Large City Suburb near a large city Small city or town Rural area
Age	“How old are you?”	Text entry
Income	“What is your total yearly income from all sources? Please report your individual income, not including the income of other people you might live with.”	Less than $20,000 $20,000 – $35,000 $35,001 – $50,000 $50,001 – $75,000 Greater than $75,000
Race	“What racial and/or ethnic groups do you identify with? (Check all that apply)”	Asian Black/African American Hispanic/Latinx/Latino/Latina Middle Eastern Native American or Alaskan Native Pacific Islander White/Caucasian I use other words to describe my race and ethnicity

 

 

 

Mental Health and Risk Behaviors among Lesbian, Gay, Bisexual, Transgender, and Queer Adults Living in the United States, by State Policy Inclusivity per the Human Rights Campaign: A Secondary Analysis of the 2022 Behavioral Risk Factor Surveillance System Data

Author: Elle Elson

Committee Members: Dr. Amy Ferketich, PhD, MAS, MA & Dr. Joanne Patterson, PhD, MPH, MSW

Abstract

Background: Recent decades have seen the emergence of social and legal protections for lesbian, gay, bisexual, transgender, and queer (LGBTQ+) individuals in the United States (US), yet a rapid increase in anti-LGBTQ+ legislation in 2022 highlights ongoing challenges. Discriminatory policies and societal stigma contribute to health disparities, particularly among LGBTQ+ young adults, with population-based implications for mental health, substance abuse, and access to health care.

Methods: The 2022 Behavioral Risk Factor Surveillance System (BRFSS) and 2022 Human Rights Campaign State Equality Index (SEI) were used to examine the relationships between state-level LGBTQ+ inclusivity (inclusive or restrictive) and mental health and risk behaviors (self-reported overall health, poor mental health days, depression diagnosis, cigarette use, e-cigarette use, binge drinking, heavy drinking, dual substance use of binge/heavy drinking and any nicotine use) among LGBTQ+ adults. Bivariate survey-weighted associations and multiple logistic regression models analyzed their disparate relationships.

Results: Restrictive states had greater relative odds of LGBTQ+ individuals not having health insurance (Adjusted Odds Ratio [AOR] 2.01, 95% Confidence Interval [CI] 1.61-2.53, p < 0.05), current smoking (AOR 1.22, 95% CI 1.02-1.45, p < 0.05), current use of e-cigarettes (AOR 1.22, 95% CI 1.03-1.45, p < 0.05), use of any type of nicotine (AOR 1.28, 95% CI 1.11-1.47, p < 0.05), and dual nicotine use (AOR 1.33, 95% CI 1.02-1.73, p < 0.05), compared to inclusive states.

Conclusion: LGBTQ+ adults living in restrictive policy states have increased odds of nicotine use and worse access to health care. State-level policy makers should consider their LGBTQ+ residents when proposing restrictive or discriminatory LGBTQ+ legislation and the negative health consequences for their constituents. Expanding access to low-cost, LGBTQ+ affirming health care with knowledge of the communities’ needs is critical for improving health outcomes in this marginalized population.

Current Study: The objective of this study was to examine how state-level LGBTQ+ policies impact mental health outcomes and risk behaviors among LGBTQ+ adults, comparing states with restrictive or inclusive LGBTQ+ policies and legislation. We hypothesized that states with more restrictive LGBTQ+ policies would evidence worse mental health outcomes and increased risk behaviors among LGBTQ+ adults. These relationships were analyzed in a sample of 17,988 LGBTQ+ adults who participated in the 2022 BRFSS. Our health indicators included prevalence of substance use, mental health diagnoses, access to health insurance, and overall health outcomes. We compared groups based on where they lived (restrictive vs. inclusive states with respect to state-level LGBTQ+ policies).

Methods

Data Source

The Behavioral Risk Factor Surveillance System (BRFSS) is a CDC-assisted health-data collection project conducted in partnership with state health departments to compromise all 50 states and the District of Columbia, Puerto Rico, Guam, and the US Virgin Islands (Behavioral Risk Factor Surveillance System 2022 Summary Data Quality Report, 2023). Data is collected from non-institutionalized adults (18 years old and greater) via landline telephone or cellular telephone interviews. In 2022, BRFSS had an overall response rate of 45.0% (BRFSS 2022 Summary Data Quality Report, 2023). The population of interest for this study consists of the states that included the optional module 26, which assessed participants sexual orientation and gender identity (SOGI) to determine LGBTQ+ status. A participant’s LGBTQ+ status was created from the 2022 BRFSS SOGI responses. For this categorization, participants were selected from the following variables of sexual orientation and gender minority from the 2022 BRFSS module. To determine sexual orientation, respondents were asked “Which of the following best represents how you think of yourself?”, with response options of “Gay; Straight, that is, not gay; Bisexual; Something else; I don’t know; Refused” (Center for Disease Control and Prevention [CDC] LLCP 2022: Codebook Report, 2022). For this analysis, sexual minority is defined as participants who answered “gay, bisexual, or something else” for sexual orientation. To determine gender minority status, respondents were asked, “Do you consider yourself to be transgender?”, with response options of, “Yes, Transgender, male-to-female; Yes, Transgender, female-to-male; Yes, gender nonconforming; No; Don’t know; Refused” (CDC, 2022). For this analysis, gender minority is defined as participants who answered “transgender male-to-female, female-to-male, or gender nonconforming” on the gender identity question. Henceforth, a participant was determined to be LGBTQ+ if they were a sexual minority and/or gender minority (SGM).

The 2022 Human Rights Campaign (HRC) State Equality Index (SEI) is a holistic assessment of statewide laws, policies, and court decisions that affect LGBTQ+ equality (Human Rights Campaign, 2022). Researchers for the SEI assessment included staff attorneys, pro bono attorneys, and law fellows who assessed the SEI for 52 markers of inclusivity (Human Rights Campaign, 2022). The SEI analyzed five categories of policies: Parenting laws and policies, Relationship recognition and religious refusal laws, Non-discrimination laws and policies, Hate crimes and criminal justice laws, and Youth-related laws and policies (Human Rights Campaign, 2022). States were then assigned one of four levels of LGBTQ+ equality: 1. High priority to achieve basic equality (lowest rating); 2. Building equality; 3. Solidifying equality; and 4. Working towards innovative equality (Human Rights Campaign, 2022). States were dichotomized into Inclusive and Restrictive states for statistical analysis, based on the statistical model from existing literature (White et al., 2023). The 31 included states are as follows: Alaska, Colorado, Connecticut, Delaware, Georgia, Hawaii, Illinois, Indiana, Iowa, Kansas, Louisiana, Maryland, Massachusetts, Michigan, Minnesota, Missouri, Montana, Nevada, New Mexico, North Carolina, North Dakota, Ohio, Pennsylvania, Rhode Island, Texas, Utah, Vermont, Virginia, Washington, West Virginia, and Wisconsin.

LGBTQ+ inclusive states consisted of: Alaska, Colorado, Connecticut, Delaware, Hawaii, Illinois, Iowa, Maryland, Massachusetts, Minnesota, Nevada, New Mexico, Pennsylvania, Rhode Island, Utah, Vermont, Virginia, Washington, West Virginia, Wisconsin (n=20).

LGBTQ+ Restrictive states consisted of: Georgia, Indiana, Kansas, Louisiana, Michigan, Missouri, Montana, North Carolina, North Dakota, Ohio, Texas (n=11). The final analytical sample (n=31 states; N=17,988 adults) was determined by what States had Sexual Orientation and Gender Identity (SOGI) variables from an optional module from the BRFSS. States without SOGI variables were excluded from analysis.

Measures

Independent variable

State-level policy: After SGM participants were flagged in the BRFSS population, they were categorized under inclusive or restrictive state-level LGBTQ+ policy from the HRC SEI.

Dependent variables

Health insurance: A participant’s health insurance status was determined from the question, “What is the current primary source of your health insurance?”, with response options of, “A plan purchased through an employer or union; A private nongovernmental plan that you or another family members buys on your own; Medicare; Medigap; Medicaid; Military related; Indian Health Services; State sponsored health plan; Other government program; No coverage of any type; Don’t know; Refused” (CDC, 2022). Participants were coded with having health insurance for all response options, except “no coverage of any type,” “don’t know,” or “refused.”

General health: A participant’s self-reported general health status was determined from the question, “Would you say that in general your health is–“, with response options of, “Excellent, Very Good, Good, Fair, Poor” (CDC, 2022). For analysis, general health was categorized as “good or greater” and “fair or less.”

Mental health: A participant’s mental health status was determined from the question, “Now thinking about your mental health, which includes stress, depression, and problems with emotions, for how many days during the past 30 days was your mental health not good?”, with a response of the numbers of days, between 0 to 30 (CDC, 2022). For analysis, mental health was categorized into 0 to 13 poor mental health days or 14+ poor mental health days (CDC, 2004).

Depressive disorder: A participant’s depressive disorder status was determined from the question, “Has a medical provider ever told you had a depressive disorder (including depression, major depression, dysthymia, or minor depression?” with response options of, “Yes, No, don’t know, Refused” (CDC, 2022). For analysis, presence of a depressive disorder was binary coded with yes or no.

Smoking status: A participant’s current cigarette smoking status was determined from a two-step question. First, participants were asked, “Have you smoked at least 100 cigarettes in your entire lifetime?” with response options of, “Yes, No, don’t know, Refused” (CDC, 2022). Second, they responded to, “Do you now smoke cigarettes –” with response options of, “Every day, some days, not at all, don’t know, Refused” (CDC, 2022). For analysis, current cigarette smoking status was binary coded with yes as those who responded with “every day” or “some days” on the second question, and “not at all” from the second question or “no” from the first question, as not currently smoking cigarettes.

E-cigarettes: A participant’s current e-cigarette use was determined from the question, “Would you say you have never used e-cigarettes or other electronic vaping products in your entire life or now use them every day, use them some days, or used them in the past but do not currently use them at all?” with response options of, “Never used e-cigarettes in your entire life, Use them every day, Use them some days, Not at all, Don’t know, Refused” (CDC, 2022). For analysis, current e-cigarette use was binary coded with no as “never used e-cigarettes” or “not at all” and yes as “use them every day” or “use them some days.”

Any nicotine: A participant’s current use of any nicotine was a created variable, with a yes for either cigarette and/or e-cigarette use.

Dual nicotine: A participant’s status of current dual nicotine use was a created variable, with a yes for both cigarette and e-cigarette use.

Binge drinking: A participant’s status for binge drinking was a two-level question, with a positive response to drinking in the past 30 days and the appropriate rate of binge drinking based on participant’s sex (5 for men or 4 for women) on the question of, “Considering all types of alcoholic beverages, how many times during the past 30 days did you have [5 for men, 4 for women] or more drinking on an occasion?” (CDC, 2022). Any alcohol use was asked, “During the past 30 days, how many days per week or per month did you have at least one drink of any alcoholic beverage?”, with response options of, “Numbers of days per week, Numbers of days in past 30 days, not at all, don’t know, Refused” (CDC, 2022).

Heavy drinking: A participant’s status for heavy drinking was a three-level question, with a positive response to drinking in the past 30 days, and calculated amounts for heavy drinking based on participant’s sex from the maximum and average drinking questions (men having more than 14 drinks per week and women having more than seven drinks per week) (CDC, 2022). Maximum drinking was measured with, “During the past 30 days, what is the largest number of drinks you had on any occasion?” and average drinking was measured with, “During the past 30 days, on the days when you drank, about how many drinks did you drink on average?” (CDC, 2022).

Polysubstance use (binge drinking and any nicotine): A participant’s polysubstance use was defined as both binge drinking in the past 30 days and currently using any type of nicotine (cigarette or e-cigarette).

Polysubstance use (heavy drinking and any nicotine): A participant’s polysubstance use was defined as both heavy drinking in the past 30 days and currently using any type of nicotine (cigarette or e-cigarette).

Control variables

The control variables for analytical consideration were identified from previous research that informed this study question (White et al., 2023). The six categorical demographic variables were age, Race/Ethnicity, education, employment, income, and sex.

Age: Upon logistic regression analysis, age was split into two levels, 18-34 years old (young adults) and 35+ years old. With the greatest disparity in substance use compared to heterosexual peers, this variable was divided with young adults (18-34 years) and adults aged 35 years and older.

Race/Ethnicity: A participant’s race and ethnicity were BRFSS calculated variables from multiple race and ethnicity questions that are detailed in the “Calculated Variables in the 2022 BRFSS Data File” report (CDC, 2023). For logistic regression analysis, race/ethnicity were categorized into six levels: White, non-Hispanic; Black, non-Hispanic; American Indian and Alaska Native (AIAN) & Native Hawaiian and Other Pacific Islander (NHOPI); Asian, non-Hispanic; Multiracial, non-Hispanic; and Hispanic.

Education: A participant’s education level was calculated from the question, “What is the highest grade or year of school you completed?” with response options of “Never attended school or only kindergarten, Grades 1 through 8 (elementary), Grades 9 through 11 (some high school), Grade 12 or GED (High school graduate), College 1 year to 3 years (some college or technical school, College 4 years or more (College graduate), Refused” (CDC, 2022). For logistic regression analysis, education was collapsed into two levels, high school/GED or less and attended or graduated college/technical school.

Employment and income: Employment and income variables were not used in logistic regression analysis due to extensive missingness (10% missing for employment and 20% missing for income). As such, socio-economic status was supplemented by education level, depicted above.

Sex: A participant’s sex was not a reliable measure in this subpopulation as currently asked in BRFSS, “Are you male or female?” with response options of, “Male, Female, Non-binary (insert sex at birth module or terminate phone call, Don’t know/not sure (insert sex at birth module or terminate phone call, Refused (insert sex at birth module or terminate phone call)” (CDC, 2022). Appropriate examples of measuring sex and gender are detailed in “Measuring Sex, Gender, Identity, and Sexual Orientation” by The National Academies Press (2022).

Statistical Analyses

Survey-weighted descriptive statistics were used to categorize the analytic sample by state-level LGBTQ+ inclusivity. Survey-weighted and design adjusted chi-square analyses were performed to measure bivariate associations between state-level inclusivity status and each outcome variable (mental health and risk behaviors). Bivariate survey-weighted and design adjusted logistic regression models were performed to compare the intersectionality of risk among this population, without adjusting for demographic variables. Survey-weighted and design adjusted logistic regression models analyzed the outcome variables by state-level LGBTQ+ inclusivity, and controlled for age, race/ethnicity, and education. Other control variables were removed from the model due to missingness (income and employment) or for lack of statistical effect (sex).

Each logistic regression model was adjusted to remove missing data of demographic (Table 1) and individual outcome variable responses, to reduce bias with listwise deletion. All logistic regression models were adjusted for all control variables (age, race/ethnicity, and education), along with one individual outcome per model (self-reported overall health, poor mental health days, depression diagnosis, cigarette use, e-cigarette use, binge drinking, heavy drinking, dual substance use of binge/heavy drinking and any nicotine use). As a result, 12 unique survey-weighted and adjusted logistic regression models had varying sample sizes to ensure maximum representation in each subpopulation from varied missingness among outcome variables. During all analyses, the subpopulation command was implemented with Stata to ensure appropriate adjustment for weighted percentages (Table 1) and statistical analyses (Table 2, Table 3) [StataBE 18 was used for all analyses]. Survey-weighting for the 2022 BRFSS data involved using a specific weighting procedure for core data and the SOGI measurement model. In addition, design adjustments are calculated based on specific weighting of their survey design to accurately represent the national population, which can be found from the “Complex Sampling Weights and Preparing 2022 BRFSS Module Data for Analysis” from the CDC (2023). An alpha of .05 indicated statistically significant differences. As a secondary analysis of publicly available, non-identifiable data, this study was exempt from Institutional Review Board approval.

 

Results

The analytic sample consisted of 17,988 self-reported SGM participants in 31 US states (Table 1). After survey-weighting, the sample was evenly distributed across LGBTQ+ inclusive (51%) and restrictive states (49%) (Table 1, Figure 1). Sex was evenly distributed and not included in the final logistic regression model due to negligible differences when controlling for demographics. Additionally, sex is not a reliable measure in the LGBTQ+ population, as transgender and gender non-conforming individuals are not accurately measured with how this variable was written in the BRFSS survey. The majority of self-reported SGM young adults, 18 to 34 years, and were the primary age of interest for health-related disparities, particularly among risk behavior outcomes. White, non-Hispanic participants represented less than 60% in both state levels, with a higher proportion of Hispanic individuals living in restrictive states (20%) (Table 1). This was a well-educated sample, with less than 13% of participants who had less than a High School education (Table 1).

Primary outcomes were analyzed with Rao-Scott chi-square tests to show the disparity of LGBTQ+ participants living in restrictive policy states. A greater prevalence was found for being uninsured (χ² <0.000), current smoking (χ² = 0.002), current use e-cigarettes (χ² = 0.026), use any type of nicotine (χ² = 0.001), and dual nicotine use of cigarettes and e-cigarettes (χ² = 0.027) (Table 2).

Unadjusted logistic regression models demonstrated statistically significant disparities in the same outcomes, not having health insurance state (Odds Ratio [OR] 2.12, 95% CI 1.72-2.62, p < 0.05), currently using cigarettes (OR 1.29, 95% CI 1.10-1.51, p < 0.05), current use of e-cigarettes (OR 1.20, 95% CI 1.02-1.41, p < 0.05), use of any type of nicotine (OR 1.26, 95% CI 1.10-1.44, p < 0.05), and dual nicotine use (OR 1.35, 95% CI 1.03-1.77, p < 0.05) (Table 3).

Adjusted logistic regressions models illustrated greater disparity of not having any health insurance in restrictive states (Adjusted Odds Ratio [AOR] 2.01, 95% Confidence Interval [CI] 1.61-2.53, p < 0.05), currently using cigarettes (AOR 1.22, 95% CI 1.02-1.45, p < 0.05), current use of e-cigarettes (AOR 1.22, 95% CI 1.03-1.45, p < 0.05), use of any type of nicotine (AOR 1.28, 95% CI 1.11-1.47, p < 0.05), and dual nicotine use (AOR 1.33, 95% CI 1.02-1.73, p < 0.05) (Table 3). All remaining outcome variables showed increased odds of poor mental and risk behavior outcomes, including worse self-reported health (AOR 1.09), having a depressive disorder (AOR 1.04), poor mental health in the last month (AOR 1.10), binge drinking in the last month (AOR 1.10), heavy drinking (AOR 1.09), and engaging in polysubstance use of nicotine and binge (AOR 1.09) or heavy drinking (AOR 1.06). However, these adjusted logistic regression models did not show statistically significant results compared to LGBTQ+ individuals in inclusive policy states.

Table 1. LGBTQ+ subpopulation demographics from 2022 Behavioral Risk Factor Surveillance System by state equality rating from 2022 Human Rights Campaign LGBTQ+ State Equality Index^{^}

		Inclusive States	Restrictive States
Variable	Weighted % (unweighted N)^§
Total Sample (17,988)		51.2 (12,218)	48.8 (5,770)
Sex††
Male		40.3 (5,065)	37.2 (2,362)
Female		59.7 (7,153)	62.8 (3,408)
Age (in years)††
18-24		30.9 (2,160)	32.3 (1,169)
25-34		28.0 (2,860)	27.3 (1,359)
35-44		15.3 (2,103)	14.7 (878)
45-54		8.8 (1,453)	9.9 (663)
55-64		8.0 (1,535)	7.2 (626)
65 and older		9.0 (2,107)	8.6 (1,075)
Race/Ethnicity
White, non-Hispanic		59.5 (8,334)	53.8 (3,979)
Black, non-Hispanic		9.3 (735)	12.4 (518)
AIAN & NHOPI#		2.4 (304)	1.3 (114)
Asian, non-Hispanic		5.4 (477)	5.0 (126)
Multiracial, non-Hispanic		5.9 (539)	4.7 (193)
Hispanic		14.6 (1,444)	20.4 (657)
Missing		(385)	(183)
Education
Less than high school, High school, or GED		40.0 (3,563)	45.4 (2,010)
Attended college/technical school		31.6 (3,224)	31.0 (1,659)
Graduated college/technical school		28.2 (5,397)	23.3 (2,084)
Missing		(34)	(17)
Employment
Employed		66.0 (7,309)	64.5 (3,266)
Not employed		25.2 (1,999)	27.5 (911)
Retired		8.9 (1,829)	8.1 (884)
Missing		(1,081)	(709)
Income
Less than $15,000		5.2 (689)	7.4 (427)
$15,000 – 24,999		8.5 (1,014)	10.5 (609)
$25,000 – 34,999		10.5 (1,236)	11.6 (709)
$35,000 – 49,999		10.7 (1,330)	11.5 (727)
$50,000 – 99,999		22.5 (2,950)	20.7 (1,283)
$100,000 – 199,999		14.8 (1,995)	11.5 (690)
$200,000 and above		4.5 (611)	3.9 (175)
Missing		(2,393)	(1,150)
^ Survey-weighting performed from CDC guidelines from 2022 BRFSS data guidelines; § Weighted % are column; †† No missing data; #American Indian, Alaska Native, Native Hawaiian, or other Pacific Islander, non-Hispanic

 

Table 2. Descriptive results of mental health and risk behaviors from 2022 Behavioral Risk Factor Surveillance System by 2022 Human Rights Campaign LGBTQ+ State Equality Index^{^}                                      

Variable	Proportion for Inclusive States	Proportion for Restrictive States	χ2
Self-reported overall health			0.112
Fair or poor health	21.8	23.9
Have any health insurance			<0.000
No	7.6	14.9
Have a depressive disorder			0.684
Yes	45.9	45.3
Poor mental health days in last month			0.258
14+ days	33.3	35.0
Current cigarette smoker			0.002
Yes	13.5	16.7
Current e-cigarette use			0.026
Yes	14.5	17.0
Any nicotine use			0.001
Yes	24.0	28.0
Dual Nicotine Use			0.027
Yes	3.8	5.1
Binge drinking in last month			0.225
Yes	22.2	23.8
Heavy drinker			0.570
Yes	8.2	8.7
Polysubstance use			0.363
Binge drinking and current nicotine user	10.4	11.3
Polysubstance use			0.687
Heavy drinking and current nicotine user	4.5	4.7
^ Survey-weighting performed from CDC guidelines from 2022 BRFSS data guidelines

 

Table 3. Logistic regression analysis of mental health and risk behaviors among LGBTQ+ adults in the 2022 Behavioral Risk Factor Surveillance System Living in Restrictive States defined by the 2022 Human Rights Campaign LGBTQ+ State Equality Index^{^}

Mental Health and Risk Behaviors	Odds Ratio (95% CI)	p-value	Adjusted Odds Ratio (95% CI) £	p-value
Self-reported overall healtha
Fair or poor health	1.13 (0.97, 1.30)	0.112	1.09 (0.94, 1.26)	0.267
Have any health insuranceb
No	2.12 (1.72, 2.62) **	<0.000	2.01 (1.61, 2.53) **	<0.000
Have a depressive disorderc
Yes	0.98 (0.87, 1.10)	0.684	1.04 (0.92, 1.64)	0.565
Poor mental health days in last monthd
14+ days	1.08 (0.95, 1.22)	0.258	1.10 (0.96, 1.25)	0.160
Current cigarette smokere
Yes	1.29 (1.10, 1.51) **	0.002	1.22 (1.02, 1.45) **	0.002
Current e-cigarette userf
Yes	1.20 (1.02, 1.41) **	0.026	1.22 (1.03, 1.45) **	0.021
Any nicotineg
Yes	1.26 (1.10, 1.44) **	0.001	1.28 (1.11, 1.47) **	0.001
Dual nicotine useh
Yes	1.35 (1.03, 1.77) **	0.027	1.33 (1.02, 1.73) **	0.033
Binge drinking in last monthi
Yes	1.10 (0.95, 1.27)	0.225	1.10 (0.95, 1.27)	0.181
Heavy drinkerj
Yes	1.06 (0.86, 1.31)	0.181	1.09 (0.88, 1.35)	0.570
Polysubstance usek
Binge drinking and current nicotine user	1.10 (0.90, 1.33)	0.364	1.09 (0.89, 1.33)	0.394
Polysubstance usel
Heavy drinking and current nicotine user	1.06 (0.81, 1.37)	0.687	1.06 (0.81, 1.37)	0.697

^ Survey-weighting performed from CDC guidelines from 2022 BRFSS data guidelines

^£ Models adjusted for Income, Age, Education, and Race/Ethnicity
** indicates p < 0.05

^aReferent (Ref) = Good or better health. ^bRef. = No health insurance. ^cRef. = No depressive disorder. ^dRef. = Zero-13 poor mental health days in last month. ^eRef. = No current cigarette use. ^fRef. = No current e-cigarette use. ^gRef. = Not a nicotine user. ^hRef. = No dual nicotine use. ⁱRef. = Not a binge drinker. ^jRef. = Not a heavy drinker. ^kRef. = No binge drinking and nicotine use. ^lRef. = No heavy drinking and nicotine use.

Figure 1. State LGBTQ+ policy type from 2022 Human Rights Campaign LGBTQ+ State Equality Index

 

Congrats to our team member, Emma Jankowski, on helping to write this paper on the sustainability of social needs resolution interventions! The full text is linked in our bio!

Happy International Women’s Day! Let’s celebrate the social, economic, cultural, and political achievements of women and continue to advocate for women’s equality.

Presenter: Joanne G. Patterson (1,2) 

Co Authors: Alysha C. Ennis (1), Emma Jankowski (1), Grace Turk (1), Ashley Meadows (1), Caitlin Miller (1), Hayley Curran (2), Sydney Galusha (1)

1. The Ohio State University College of Public Health
2. Center for Tobacco Research, The Ohio State University James-Comprehensive Cancer Center

 

Introduction:

• Lesbian, gay, bisexual, transgender, and queer young adults (LGBTQ+ YA) report high rates of cigarette smoking and nicotine vaping (1-6).
• Mass-reach anti-tobacco communications can increase public knowledge of tobacco harms and decrease use, yet they may not engage LGBTQ+ YA (7-17).
• No studies describe effective anti-tobacco message framing for SGM and non-SGM YA engaged in dual use, though understanding these nuances is important for developing inclusive anti-tobacco communications.
• We conducted a formative evaluation to inform culturally targeted (CT) anti-tobacco communications. 

 

Methods:

• We reviewed existing CT anti-tobacco campaigns before conducting in-depth focus groups of N=22 LGBTQ+ YA (18-35) ever dual users to assess best practices for message design (visuals, semantics).
• We applied findings to develop 9 CT and 9 non-targeted (NT) messages, which an expert panel (N=7) of LGBTQ+ community partners, scientists, and LGBTQ+ YA reviewed.
• Messages are being experimentally tested in a remote eye tracking study with LGBTQ+ YA.

 

Results:

• LGBTQ+ YA were skeptical of CT anti-tobacco campaigns featuring stereotypical representation of LGBTQ+ individuals and questioned the motive of cultural targeting,
  • “Are you genuinely […] advertising to me, or are you advertising to some, like, monolithic LGBT group that you think exists?” (FG34, Queer, he/him).
• Communications featuring naturally posed models and a diversity of LGBTQ+ people engaged participants more than overly posed models and oversaturated colors:
  • “If I was scrolling, [the natural ad] would actually make me stop… but the second I see those saturated blue, purple looks, I’m like “That’s an ad”, and I just scroll right past it.” (FG23, Bisexual, he/him/she/her).
• Personal stories were well received:
  • “I also like the quotes, and that it has the people’s name there. It makes it feel more personal…they’re reaching out to you with their story.” (FG21, Bisexual, they/them).
• Unclear visuals and slogans were negatively received.
• We applied findings to develop CT anti-tobacco communications featuring naturalistic LGBTQ+ models of diverse races, ethnicities, and genders. We paired harms messaging with personal stories and subtle cultural cues (e.g., “our health”, pronouns). These are shown at the bottom of this post.
• The expert panel confirmed that messages were culturally and scientifically relevant.  

Theme	Code	Definition
Ad Design
	Font	Participants discuss liking or disliking font choices/typography
	Layout	Participants discuss liking or disliking spacing, layout, or white space.
	Colors	Participants discuss liking or disliking colors
	Graphic type	Participants discuss liking or disliking the type of graphic (e.g. photograph vs. illustration/cartoon)
	Brand Identity	Discusses that the ad design matches or does not match the product being sold given what is known/presumed about a brand (e.g., of “not matching” brand identity:  Kandy Pens ad image of women/men being intimate and product not featured; “I like that they used their brand name as kind of like a play on words”).    “This ad tells me nothing about what this company is or does, or anything.”
	Creativity	Participants discuss whether an ad does or does not feel creative or clever with respect to its design (e.g., Bud Light ad where “L G B T” were highlighted).
	Aesthetically pleasing	Participants discuss whether  an ad is overall aesthetically pleasing or not
Ad Content: Imagery
	Imagery – Representative	Participants discuss feeling though the images in ad represent them/people they know (i.e., looks like me, acts like me) or feature real representation of LGBTQ people generally.
	Imagery – Liking	General like code for imagery
	Imagery – Disliking	General dislike code for imagery
	Imagery – Subtle/Overt	Participants discuss the subtlety or overtness of the LGBTQ elements within an ad
	Imagery – Pride Flags and rainbows	Discusses the liking or disliking of LGBTQ flags and colors within ads
	Imagery – Who	Participants describe liking or disliking having posed (“fake”)   vs. more natural looking (“real”) models in the ad.
Ad Content: Language
	Language – Word choice	Participants discuss liking or disliking word choice
	Language – Efficacy of absolute risk vs. self-efficacy messaging	Participants discuss liking or disliking absolute risk messages as compared to self-efficacy messages
	Language – Slang use	Participants discuss liking or disliking the use of slang in an ad (e.g. words like “slay”, “queen”)
	Language- Humor	Participants discuss liking or disliking the use of humor in advertisements
	Language- slogans or taglines	Participants discuss liking or disliking the use of slogans, taglines, or catch phrases in an advertisement (e.g., “Quitting isn’t a perfect process” or “Made with Pride”)
Ad Content: Representativeness
	Inclusivity	Participants discuss whether or not the ad is representative of LGBTQ identities   (L – G – B – T  – Q – NB)
	Stereotyping	Participants discuss feeling as though the ads represent stereotypes of the LGBTQ community, in imagery, language, content, etc.
	Intersectionality	Participants discuss whether or not ads are intersectional in terms of identities that are not within the LGBTQ umbrella such as racial identity or class status
	Authenticity	Participants discuss feeling as though ads are inauthentic/authentic; (e.g., feeling like ads have been created by those not within the LGBTQ community/ feeling as though ads have been created by those within the LGBTQ community    (authentic ads may take into consideration the feelings, wishes and traditions towards the LGBTQ community)
	Fetishization of LGBTQ community	Participants discuss ads sexualizing or fetishizing the LGBTQ community
	Target Audience	Discusses whom they believe an ad was targeted towards
	Normalization/Visibility	Participants discuss ads being used to normalize or make visible LGBTQ people and relationships. Word “representative” might be used by participants.
Context
	Brand partnerships	Participants discuss liking or disliking the inclusion of brand partnerships with LGBTQ organizations (e.g. GLADD, Rainbow Railroad)
	Ad placement	Participants discuss where they see culturally-targeted ads (e.g. social media, malls, TV)
	Outdated/Current	Discusses whether the language, content, and/or design of ad feels outdated or current (e.g., compared to the current time period/context).
	Rainbow capitalism	Participants discuss only seeing culturally-targeted ads during Pride Month, or being performative/used just to make money
	Pandering	Participants discuss feeling as though companies are trying to please the LGBTQ community by acting in a way they believe the LGBTQ community would want them to act
	Corny/Trite	Participants discuss advertisements feeling “corny” or trying too hard. (e.g. describing things as “tumblr-core,” “white woman’s instagram,” “millennial”, “mom”)
	Necessity	Participants discuss whether or not they view LGBTQ+ advertising as necessary/needed for LGBTQ community
	General feelings	Participants discuss how they feel about LGBTQ culturally targeted advertising generally; whether like, dislike, or neutral
	Personal Experience	Discusses how their personal experience influences their perception of an ad
	Favorite	Participant discusses an ad as their favorite
	Purchasing	Discusses buying and purchasing product advertised in the ad shown

 

Conclusions:

• LGBTQ+ YA were distrustful of CT communications that leveraged “stereotyped” LGBTQ+ imagery.
• As authenticity is important to LGBTQ+ YA, co-creating CT anti-tobacco communications may enhance acceptability, engagement, and effectiveness.
• Eye-tracking research will objectively assess the effect of CT (vs NT control) communications on engagement. 

 

Funding/Acknowledgements:

• Thank you to all members of the Practice and Science for LGBTQ+ Health Equity Lab for their contributions.
• This research was funded by the National Institutes of Health, National Cancer Institute and FDA Center for Tobacco Products (K99CA260718 and R00CA260718; PI: JGP), and supported by the Ohio State University Comprehensive Cancer Center and the Ohio State University College of Public Health. 

 

References:

1. Ridner S, Ma J, Walker K, et al. Cigarette smoking, ENDS use and dual use among a nationalsample of lesbians, gays and bisexuals. Tob Prev Cessat. 2019;5(December). doi:10.18332/tpc/114229
2. Delahanty J, Ganz O, Hoffman L, Guillory J, Crankshaw E, Farrelly M. Tobacco use among lesbian, gay, bisexual and transgender young adults varies by sexual and gender identity. Drug Alcohol Depend. 2019;201:161-170. doi:10.1016/j.drugalcdep.2019.04.013
3. Fallin-Bennett A, Lisha NE, Ling PM. Other Tobacco Product Use Among Sexual Minority Young Adult Bar Patrons. Am J Prev Med. 2017;53(3):327-334. doi:10.1016/j.amepre.2017.03.006
4. Nayak P, Salazar LF, Kota KK, Pechacek TF. Prevalence of use and perceptions of risk of novel and other alternative tobacco products among sexual minority adults: Results from an online national survey, 2014–2015. Prev Med. 2017;104:71-78. doi:10.1016/j.ypmed.2017.05.024
5. Osibogun O, Taleb ZB, Bahelah R, Salloum RG, Maziak W. Correlates of poly-tobacco use among youth and young adults: Findings from the Population Assessment of Tobacco and Health study, 2013–2014. Drug Alcohol Depend. 2018;187:160-164. doi:10.1016/j.drugalcdep.2018.02.024
6. Stanton CA, Bansal-Travers M, Johnson AL, et al. Longitudinal e-Cigarette and Cigarette Use Among US Youth in the PATH Study (2013–2015). JNCI J Natl Cancer Inst. 2019;111(10):1088-1096. doi:10.1093/jnci/djz006
7. Farrelly MC, Nonnemaker J, Davis KC, Hussin A. The Influence of the National truth® Campaign on Smoking Initiation. Am J Prev Med. 2009;36(5):379-384. doi:10.1016/j.amepre.2009.01.019
8. Farrelly MC, Duke JC, Nonnemaker J, et al. Association Between The Real Cost Media Campaign and Smoking Initiation Among Youths — United States, 2014–2016. MMWR Morb Mortal Wkly Rep. 2017;66(02):47-50. doi:10.15585/mmwr.mm6602a2
9. Sly D, Hopkins R, Trapido E, Ray S. Influence of a counteradvertising media campaign on initiation of smoking: the Florida “truth” campaign. Am J Public Health. 2001;91(2):233-238. doi:10.2105/AJPH.91.2.233
10. Weiss JW, Cen S, Schuster D, et al. Longitudinal effects of pro‐tobacco and anti‐tobacco messages on adolescent smoking susceptibility. Nicotine Tob Res. 2006;8(3):455-465. doi:10.1080/14622200600670454
11. Siegel M. What the FDA Gets Wrong About E-Cigarettes. Bloomberg. https://www.bloomberg.com/view/articles/2017-03-16/what-the-fda-gets-wrong-about-e-cigarettes?in_source=embedded-checkout-banner. Published March 16, 2017. Accessed July 25, 2023.
12. Calabro KS, Khalil GE, Chen M, Perry CL, Prokhorov AV. Pilot study to inform young adults about the risks of electronic cigarettes through text messaging. Addict Behav Rep. 2019;10:100224. doi:10.1016/j.abrep.2019.100224
13. U.S National Cancer Institute. A Socioecological Approach to Addressing Tobacco-Related Health Disparities | Division of Cancer Control and Population Sciences (DCCPS). A Socioecological Approach to Addressing Tobacco-Related Health Disparities; 2017. Accessed July 25, 2023. https://cancercontrol.cancer.gov/brp/tcrb/monographs/monograph-22
14. Duke JC, Farrelly MC, Alexander TN, et al. Effect of a National Tobacco Public Education Campaign on Youth’s Risk Perceptions and Beliefs About Smoking. Am J Health Promot. 2018;32(5):1248-1256. doi:10.1177/0890117117720745
15. Kranzler EC, Hornik RC. The Relationship Between Exogenous Exposure to “The Real Cost” Anti-Smoking Campaign and Campaign-Targeted Beliefs. J Health Commun. 2019;24(10):780-790. doi:10.1080/10810730.2019.1668887
16. The Real Cost E-Cigarette Prevention Campaign. Published online July 21, 2023. Accessed July 31, 2023. https://www.fda.gov/tobacco-products/real-cost-campaign/real-cost-e-cigarette-prevention-campaign#:~:text=Our%20Goal%3A%20Educate%20youth%20about,addiction%20from%20using%20e%2Dcigarettes
17. This Free Life Campaign. Published online March 11, 2022. Accessed July 31, 2023. https://www.fda.gov/tobacco-products/public-health-education-campaigns/free-life-campaign

 

Images Presented in Focus Groups:

 

Culturally Targeted Imagery for Eye-Tracking

 

Control Imagery for Eye-Tracking

** At this point, we are pre-publication. If you would like to see more images, please reach out to the Principal Investigator, Joanne Patterson (patterson.1191@osu.edu).