The Scientific Basis for the Regulation of Flavors in Tobacco Products

Abstract

Effective tobacco policies are important for reducing the harm of tobacco use and can have a broad impact at the population level. This review provides an overview of how clinical science can inform tobacco policies with a focus on policies related to flavored tobacco products, using menthol cigarettes as an illustrative example. Specifically, this review summarizes the role of flavors in tobacco use and the history of regulation of flavored tobacco products by the US Food and Drug Administration (FDA), provides an overview of clinical research methods used to contribute to the scientific evidence to inform FDA tobacco policies, discusses key findings related to menthol tobacco products using these methods, and proposes future directions for clinical research. As the tobacco marketplace continues to evolve with new products and flavor chemicals, ongoing clinical science will be essential for establishing evidence-based policies to protect public health and reduce tobacco-related health disparities.

1. Introduction

Tobacco use is the leading cause of preventable disease and death in the United States (USDHHS 2012, 2014). Currently, 46 million US adults (18.7%) ages 18 years and older use tobacco products regularly (e.g., some days or every day). Cigarettes are the most commonly used tobacco product (11.5%), followed by e-cigarettes (4.5%), cigars (3.5%), smokeless tobacco (2.1%), and pipes (0.9%) (Cornelius et al. 2023). Among US adolescents, e-cigarettes are the most commonly used nicotine/tobacco product (14.1%), followed by cigars (2.8%); cigarettes (2.0%); and other tobacco products such as smokeless tobacco, hookah tobacco, and nicotine pouches (<2% each) (Park-Lee et al. 2022). More than 90% of adults who currently smoke cigarettes first started smoking when they were younger than 18 years, often paving the way for lifelong dependence on tobacco (USDHHS 2014). Furthermore, each year more than 480,000 deaths in the United States are attributed to cigarette smoking, including secondhand smoke exposure (USDHHS 2014). Thus, crucial public health goals to reduce tobacco use and prevent tobacco-related death and disease include preventing young individuals from initiating use of tobacco products and helping individuals who use tobacco products to quit.

Effective tobacco policies are an important part of reducing the harm of tobacco use and can have broad impacts at the population level. In the United States, the Food and Drug Administration (FDA) has the authority to regulate tobacco products for the protection of public health at the federal level, and states and localities can also implement their own tobacco regulations. Scientific evidence on factors that influence initiation of tobacco use, continued use, the development of dependence, and cessation is important in informing tobacco policies and developing tobacco product standards. In this article, we provide an overview of how clinical science can inform tobacco regulatory policies with a focus on policies related to flavored tobacco products, using menthol cigarettes as an illustrative example. We focus on flavors in tobacco products because flavors are critical in influencing initiation, experimentation, continued use, and maintenance of addiction to tobacco products. Specifically, we review the history of regulation of flavored tobacco products by the FDA, discuss the role of menthol flavor in tobacco use, and provide an overview of clinical research methods used to contribute to the scientific evidence on the potential effects of FDA tobacco policies. We discuss key findings related to menthol tobacco products obtained through these clinical research methods and propose future directions for clinical research.

1.1. Federal Regulation of Flavors in Tobacco Products

The 2009 Family Smoking Prevention and Tobacco Control Act (referred to as the Tobacco Control Act) gave the FDA (2009) the authority to regulate the manufacturing, marketing, sale, and distribution of cigarettes, roll-your-own tobacco, and smokeless tobacco. The FDA has used this authority to create standards for constituents in tobacco products, including prohibiting “characterizing flavors” other than menthol in cigarettes, since 2009. Specifically, the FDA (2009) ruling prohibited

“a cigarette or any of its component parts (including the tobacco, filter, or paper) from containing as a constituent (including a smoke constituent) or additive, an artificial or natural flavor (other than tobacco or menthol) or an herb or spice, including strawberry, grape, orange, clove, cinnamon, pineapple, vanilla, coconut, licorice, cocoa, chocolate, cherry, or coffee, that is a characterizing flavor of the tobacco product or tobacco smoke.”

The FDA (2022c) has specified several factors that are relevant in determining whether a tobacco product has a characterizing flavor, including:

“1) The presence and amount of artificial or natural flavor additives, compounds, constituents, or ingredients, or any other flavoring ingredient in a tobacco product, including its components or parts; 2) The multisensory experience (i.e., taste, aroma, and cooling or burning sensations in the mouth and throat) of a flavor during use of a tobacco product, including its components or parts; 3) Flavor representations (including descriptors), either explicit or implicit, in or on the labeling (including packaging) or advertising of tobacco products; and 4) Any other means that impart flavor or represent that the tobacco product has a characterizing flavor.”

1.2. How Menthol-Flavored Tobacco Products Impact Public Health

As noted in Section 1.1, menthol was exempt from the 2009 regulation that restricted other characterizing flavors in cigarettes (FDA 2009). Menthol is considered a characterizing flavor in cigarettes because it is a flavoring compound that produces a minty taste and cooling sensation. A great deal of research has investigated the role of menthol flavor in cigarettes, using both preclinical and clinical studies (for a complete review of the scientific evidence supporting the policy to restrict menthol flavor in cigarettes, see FDA 2022c). The findings indicate that menthol has direct pharmacological and reinforcing effects and serves as a conditioned cue that develops over time (Wickham 2020). Specifically, the cooling sensation of menthol can reduce irritation, coughing, and aversiveness of nicotine and of smoking, thereby making it easier to smoke and inhale more deeply (Kreslake et al. 2008, Wise et al. 2012). The beneficial sensory effects of menthol together with the reinforcing effects of nicotine may increase the likelihood of continued experimentation, progression to regular use, and development of nicotine dependence. For instance, the rewarding effects of nicotine result from nicotine binding with nicotinic acetylcholine receptors to increase dopamine release in brain reward circuits (Picciotto & Kenny 2021).With continued use, nicotine receptors are upregulated, and the number of active nicotinic receptors increases (Cosgrove et al. 2009). Discomfort from nicotine withdrawal happens when these high levels of nicotinic receptors are unoccupied as a result of low levels of nicotine in the brain, thus prompting continued use to alleviate nicotine withdrawal (Dani & Heinemann 1996, Picciotto & Kenny 2021). Interestingly, research indicates that menthol smoking is associated with greater upregulation of nicotinic receptors compared with nonmenthol smoking (Brody et al. 2013), consistent with evidence that individuals who smoke menthol cigarettes have greater nicotine dependence (Villanti et al. 2017).

Additionally, menthol flavors may become conditioned cues, and aspects of the smell, flavor, sensory effects, or packaging may elicit cravings and desire for continued use (Wickham 2020). For example, tobacco packaging is an important way for the tobacco industry to market its products and communicate with current and potential consumers (NCI 2008). Focus groups with adolescents and young adults found that seeing flavors on tobacco product packages is appealing and is a reason for purchasing or having the intent to purchase the product (Kong et al. 2017). The tobacco industry uses words, colors, and images on tobacco packages to create a “sensation transference” (Lempert & Glantz 2017), which means that the sensation featured on the packaging is attributed to the product itself. In the case of menthol cigarettes, the green color and packaging are prominent and associated with positive sensations, including “fresh” and “smooth” (Wackowski et al. 2018).

Research indicates that the effects of menthol increase the appeal of cigarettes; make it easier to smoke; and are associated with greater rates of cigarette initiation, greater nicotine dependence, and lower rates of successful quitting (FDA 2013, Tob. Prod. Sci. Advis. Comm. 2011, Yerger & McCandless 2011). Although people who smoke menthol cigarettes try to quit at similar or higher rates compared with those who smoke nonmenthol cigarettes (Levy et al. 2011a), they are less likely to quit successfully (Delnevo et al. 2011, Foulds et al. 2010, Levy et al. 2011a, Smith et al. 2014). Thus, menthol cigarette use remains a critical public health issue. Furthermore, while there has been an overall decline in cigarette smoking prevalence in recent years, menthol cigarette use has not declined at the same rate. Indeed, the proportion of people who smoke menthol cigarettes has increased (SAMHSA 2011), and national sales data indicate that the menthol cigarette market share increased by nearly 10 percentage points from 2000 to 2018 (Delnevo et al. 2020b). Continued menthol cigarette smoking suggests that these individuals may be helped less by existing tobacco control policies (Tauras et al. 2010) and remain highly vulnerable to the negative health consequences of smoking (USDHHS 2014). Thus, in spring 2022, the FDA (2022c) announced a proposed rule to prohibit menthol as a characterizing flavor in cigarettes, although this ruling has not yet been finalized. Clinical research plays a valuable role in understanding the potential effects of this policy to further support the FDA’s regulatory efforts. Specific research methods and key findings are described further in Section 2.

1.3. Considering Tobacco Policy Impacts on Priority Populations

It is critical that tobacco policies advance health equity and reduce tobacco use among individuals and communities who face disproportionate negative health consequences due to tobacco use. Notably, menthol cigarettes are used at disproportionately high rates among specific priority populations, including youth and young adults, pregnant individuals, low-income individuals, individuals who identify as LGBTQIA+, and certain racially/ethnically minoritized groups such as Black individuals (Delnevo et al. 2020a; USDHHS 2014, 2020). For example, national estimates indicate that the prevalence of menthol cigarette use among those who currently smoke is two to three times higher among individuals who identify as Black compared with any other racial group, with more than 85% of Black people who smoke preferring menthol cigarettes (Caraballo & Asman 2011, Villanti et al. 2016). Furthermore, menthol cigarette smoking is associated with reduced cessation outcomes, particularly among Black individuals (e.g., Delnevo et al. 2011, FDA 2013, Gundersen et al. 2009, Keeler et al. 2017, Smith et al. 2020). Thus, it is important to evaluate the impact of a menthol ban in cigarettes on these specific populations. Doing so includes examining the potential consequences of the policies as well as considering what additional supports may be needed when policies are enacted to provide effective solutions to reduce cigarette smoking and promote health equity.

1.4. Other Flavors and Other Tobacco Products

As the FDA plans to prohibit menthol flavor in cigarettes, questions remain about the regulation of other flavors and other noncigarette tobacco products for which the evidence to inform policy is more limited. For example, synthetic coolants (e.g., WS-3, WS-5, WS-23) are a category of flavoring compounds that have cooling properties without an accompanying minty aroma and taste; thus, they might circumvent flavor regulations if taste and aroma are considered essential to the FDA’s current definition of a characterizing flavor. The tobacco industry has introduced these coolants into a variety of tobacco products, including cigarettes ( Jabba et al. 2023a, Page et al. 2023), e-cigarette liquids ( Jabba et al. 2022), and nicotine pouches ( Jabba et al. 2023b), to potentially avoid menthol bans and/or to enhance the appeal and flavor of the product. For instance, new “ice” e-liquids combine these synthetic coolants with sweet (e.g., fruit/candy) flavors that are also appealing to youth (Davis et al. 2021, Leventhal et al. 2023). Thus, replacement of menthol in cigarettes (or other combustible products) with synthetic coolants could undermine the public health benefit of a menthol cigarette ban, and some cigarette companies have started selling “nonmenthol” cigarettes containing synthetic coolants in response to a menthol cigarette ban already in place in California ( Jabba et al. 2023a, Jewett & Baumgaertner 2023).

Furthermore, research is needed to inform the regulation of flavors in other noncigarette tobacco products. The FDA (2016) gained the authority to regulate all other tobacco products, including e-cigarettes, cigars, hookah tobacco, and pipe tobacco, following the Deeming Rule in 2016. Currently, the FDA (2020) has prioritized enforcement on restricting flavors other than menthol and tobacco flavors (e.g., prohibiting fruit, sweet, mint, and other flavors) in cartridge-based e-cigarettes, but flavors remain available in many types of e-cigarette devices on the market, and an important question remains about the role of menthol flavor in e-cigarettes. For example, in the context of a menthol cigarette ban, adults who smoke menthol cigarettes may turn to e-cigarettes that come with menthol flavors. This shift may be beneficial for public health if people are using less harmful products and are able to quit smoking (Gades et al. 2022). However, it is also important to consider the impact of flavored e-cigarettes on youth or those who do not smoke, since appealing flavors are often cited as a reason for youth use of e-cigarettes (Park-Lee et al. 2022). Additionally, the FDA (2022a) has proposed a plan to prohibit characterizing flavors in cigars. Cigar products, which include premium cigars, large manufactured cigars, cigarillos, and little filtered cigars, come in a variety of appealing flavors and tend to be used at higher rates by priority populations including young adults and Black individuals (NASEM 2022). Thus, clinical science will also be essential in providing scientific evidence to inform regulatory policies about these other tobacco products.

2. Research Methods to Inform Tobacco Regulation

Clinical research plays a key role in advancing policies, including tobacco policies. In making regulatory decisions, the primary goal of the FDA is to regulate tobacco products for the protection of public health. In the case of the proposed menthol cigarette ban, this means that the FDA (2022c) considers the scientific evidence and evaluates how the regulatory action will influence the risks and benefits at the population level, considering the impact on both those who currently use and do not use tobacco products. Considerations also include the impact on those who might switch between tobacco products or start using additional tobacco products, those who previously used tobacco products and quit, and those who do not currently use products but might initiate or reinitiate tobacco use (FDA 2022c).

To support evidence-based policies, the FDA (2022b) has defined several priority areas, seeking scientific information on addiction, appeal, behavior, toxicity, impact analysis, health effects, marketing influences, and communications related to tobacco products or their constituents (Table 1). In this section, we outline various human clinical research methods that have been valuable for providing scientific evidence about the potential effects of a policy restricting menthol cigarettes, including survey methods, product choice and marketplace studies, tobacco administration paradigms, cue-exposure research, clinical trials, and simulation models. Note that these clinical research methods are not mutually exclusive, and studies may use multiple research methods together to address research questions to inform regulatory policies. We also acknowledge that there are other research methods that do not involve human subjects, such as toxicology studies and preclinical animal models that address important areas of the FDA research priorities. While these methods are not within the scope of this review, comprehensive reviews on toxicology and animal models are available (Bagdas et al. 2022, Gordon et al. 2022).

Table 1.

U.S. Food and Drug Administration Center for Tobacco Products Research Priority Areas

Research Priority	Definition
Product Composition and Design	Understanding the chemical constituents in tobacco products and the methods of measuring them across product with diverse characteristics.
Toxicity	Understanding how tobacco products and changes to tobacco product characteristics affect their potential to cause morbidity and mortality in users and nonusers through secondary exposure.
Addiction	Understanding the effect of tobacco product characteristics on addiction and abuse liability across populations.
Health Effects	Understanding the short- and long-term health effects of tobacco products (excluding conventional cigarettes) with an emphasis on longitudinal data.
Behavior	Understanding the knowledge, attitudes, perceptions, and behaviors related to tobacco product use and the impact of tobacco product characteristics on behaviors across populations.
Communications	Understanding how to effectively communicate to the public regarding nicotine and the health effects of tobacco products through media campaigns and digital media.
Marketing Influences	Understanding the impact of marketing on susceptibility to and initiation of using tobacco products (both classes of products and products within classes) and transitions between experimentation, initiation, regular use, product switching, dual use, and cessation-related behaviors among different populations.
Impact Analysis	Understanding the potential or actual impact of FDA regulatory actions.

FDA 2022b. FDA Tobacco Product Research Priorities. URL: https://www.fda.gov/tobacco-products/research/research-priorities#:~:text=CTP%20encourages%20research%20studies%20to,mental%20health%20conditions%20and%2For

2.1. Survey Studies

Survey studies have been applied to tobacco regulatory research in many ways, including assessing the potential effects of policies prior to implementation, assessing the actual impact of policies using pre-post survey designs, and assessing emerging trends and changes in the marketplace via novel methods using social media analysis. We outline the various methods and key findings of several survey study designs below, including social media analysis.

2.1.1. Self-reported behavioral intentions and perceptions: method overview.

Survey research can be used to assess behavioral intentions to hypothetical policy scenarios, such as how adults who smoke menthol cigarettes think they would respond if menthol cigarettes were no longer available. These surveys typically include possible response options such as the perceived likelihood of quitting smoking or switching to another tobacco product. These methods have many advantages, including being able to collect behavioral data from a specific population (e.g., adults who smoke menthol cigarettes, young adults who use both cigarettes and e-cigarettes; Pacek et al. 2019). Survey items assessing possible responses to hypothetical tobacco policies can be included in population-level surveys (Zatonski et al. 2018), online surveys (Pearson et al. 2012, ´ Wackowski et al. 2015), and regional convenience samples to study a specific population of interest (e.g., residents in public housing; Dearfield et al. 2022). Surveys are also dynamic assessment tools that can be readily modified to reflect changes in the market (e.g., assessing the likelihood of switching to new or emerging tobacco products). The major limitation of this study design is that responses are hypothetical, and behavioral intentions are not perfect correlates of actual behaviors.

2.1.2. Self-reported behavioral intentions and perceptions: key findings.

Across studies, the most common response to a hypothetical menthol cigarette ban is reported plans to quit smoking among both young adults (Rose et al. 2019, Wackowski et al. 2014) and older adults (D’Silva et al. 2015, Dearfield et al. 2022, Zatonski et al. 2018) who currently smoke menthol cigarettes, ´ followed by switching to another product such as e-cigarettes (D’Silva et al. 2015,Wackowski et al. 2015), nonmenthol cigarettes (Wackowski et al. 2014, 2015; Yang et al. 2022), or cigars (Yang et al. 2022). Notably, support for a policy prohibiting menthol cigarettes and plans to quit smoking are most strongly endorsed by Black young adults (Rose et al. 2019, Wackowski et al. 2014) and older adults (Yang et al. 2022) who currently smoke menthol cigarettes.

2.1.3. Pre-post policy surveys: method overview.

Other types of survey research can overcome the limitations of hypothetical behavioral intentions by conducting pre-post surveys to assess the policy impact in locations where tobacco policies are proposed and enacted. These methods can address key questions, including questions about the impact of flavored tobacco policy restrictions on tobacco sales or tobacco use behavior. These studies involve repeated surveys over time and use interrupted time series analyses or pre-post ban assessments to examine changes in smoking behavior (e.g., initiation, quitting, or switching to another product) and tobacco sales after the policy is implemented (e.g., Chaiton et al. 2020b, 2021a). Advantages of these methods include quantifying the effect of a policy after enactment and assessing direct behavioral outcomes. One challenge with this method concerns survey timing. For example, policy enactment (i.e., when it is introduced as law) sometimes precedes policy enforcement (i.e., when compliance is mandated) if there is a grace period to allow retailers to become compliant. Thus, behavioral outcomes may lag policy changes, so it is crucial to consider the appropriate timing to capture the outcome of interest. Also, it can be challenging to parse out the specific effect of a flavor ban policy from other societal impacts that happen concurrently; for instance, other factors related to tobacco products (e.g., changes in product costs, new emerging products, changes in purchase age) or society at large (e.g., changing conditions around the COVID-19 pandemic) may simultaneously influence tobacco use.

2.1.4. Pre-post policy surveys: key findings.

Retail Sales.

Survey studies have examined retail sales in regions and countries where menthol cigarette ban policies have been enacted [e.g., in Canada (Chaiton et al. 2020b) and US states such as Massachusetts (Asare et al. 2022)]. For example, after a menthol cigarette ban in Canada, studies showed significant declines in menthol cigarette sales (Brown et al. 2022; Chaiton et al. 2020b, 2021b) and total cigarette sales overall (Chaiton et al. 2020b, 2021b) as well as no increase in illicit menthol cigarettes sold (Stoklosa 2019), suggesting the effectiveness of the menthol ban policy. However, some studies have shown unintended effects of menthol ban policies, including substitution with nonmenthol cigarettes or purchasing from First Nations reserves where regulations were not in place, suggesting the importance of assessing substitution and unregulated jurisdictions to understand the overall policy impact (Carpenter & Nguyen 2021). Survey studies also found that tobacco companies promoted alternative products as replacements for menthol cigarettes, including using similar color or brand descriptors, or adding flavor capsules to mimic menthol cigarettes and circumvent the flavor restrictions in Canada (Borland et al. 2019). Thus, it is important to consider assessing related products when surveying the impact of tobacco policies.

Smoking Behavior.

Survey studies have also evaluated smoking behavior following menthol cigarette restrictions examining both short-term outcomes (e.g., 1 month later; Chaiton et al. 2018) and long-term outcomes (e.g., 1 and 2 years later; Chaiton et al. 2020a, 2021a) following the prohibition of menthol cigarettes in Canada. Findings indicate that actual behavior immediately following the ban differed from anticipated behavior. Most adults (60%) who smoked menthol cigarettes anticipated that they would continue smoking and switch to nonmenthol cigarettes, and only 15% anticipated that they would quit smoking; however, more people actually quit smoking after the policy was implemented (29%) (Chaiton et al. 2018). Follow-up surveys conducted 1 year (Chaiton et al. 2020a) and 2 years (Chaiton et al. 2021a) after the menthol cigarette ban was implemented indicated that 63% of adults who previously smoked menthol cigarettes attempted to quit smoking, and they were more than twice as likely to successfully quit smoking after the policy, compared with those who smoked nonmenthol cigarettes.

Related studies in the United States have assessed the impact of a flavor ban. Specifically, after the Tobacco Control Act restricted all flavors (e.g., fruit, candy) from cigarettes (excluding menthol) in 2009, population survey studies among youth observed lower rates of cigarette use overall (Courtemanche et al. 2017, Rossheim et al. 2020). However, rates of menthol cigarette use and use of other flavored tobacco products (e.g., cigars) increased significantly, suggesting that the overall effect of this policy on youth tobacco use was likely diminished due to the continued availability of other flavored tobacco products and menthol cigarettes (Courtemanche et al. 2017). Similar findings were noted in studies of specific jurisdictions where policies were implemented in the United States, such as comprehensive flavored tobacco product bans restricting sales of menthol cigarettes and all other flavored tobacco products (except those with tobacco flavor) in San Francisco in 2019 (Yang et al. 2020) and in Minnesota in 2018 (Olson et al. 2022). Survey results indicated lower rates of tobacco product use among youth (Olson et al. 2022) and young adults (ages 18–34; Yang et al. 2020) after these flavor ban policies. However, many survey respondents reported continued access to flavored tobacco products, including from in-person retailers, neighboring jurisdictions without restrictions, and online purchases. Overall, studies examining policy enforcement and identifying ways to increase retailer compliance (e.g., retailer education prior to enforcement; Vyas et al. 2021) are therefore essential to maximize the policy impact.

2.1.5. Social media analysis: method overview.

Social media is widely used by individuals of all ages; 90% of the US population used social media in 2022 (Dixon 2023). The pervasiveness of social media makes it an optimal source for marketing. Social media is widely used by the tobacco industry to enhance product visibility, circumvent tobacco marketing restrictions on traditional media, and reach many people quickly at a low cost (Freeman 2012). Additionally, social media provides a unique opportunity to understand individuals’ perceptions and use behaviors related to tobacco products, including policies and social environmental factors that contextualize these perceptions and behaviors (Ayers et al. 2014). The conversations on social media are public and happen in real time, organically, and at a fast speed, so they can serve as a first line of information to provide instantaneous insights into tobacco products before traditional survey methods can be used. Furthermore, accumulating evidence indicates that exposure to tobacco content on social media among youth is associated with lower risk perceptions and greater tobacco use (Donaldson et al. 2022, Lee et al. 2023a, Zheng et al. 2021). Thus, it is critical for public health officials and policy makers to be aware of trends in product innovations and marketing as well as novel, alternative use behaviors that may undermine policies (Allem et al. 2017b, Ayers et al. 2014). Limitations of social media analysis methods include the potential for false information through bots and sampling bias. For instance, sampling bias may occur if individuals who have extreme views (overly positive, overly negative) are more vocal on social media. However, social media evidence can reveal emerging trends, and these data can inform traditional survey studies.

2.1.6. Social media analysis: key findings.

Social media analysis has been used to provide scientific evidence on flavored tobacco products to inform regulations. For example, studies of social media ads indicate that appealing flavors are one of the most common marketing themes (Lee et al. 2023b, Liang et al. 2015, McCausland et al. 2019), and analyses of discussions and comments on popular social media platforms (e.g., Twitter/X, Reddit, Instagram) indicate that the availability of fruit and sweet flavors is part of the appeal of diverse tobacco products such as e-cigarettes, hookah tobacco, and cigarillos (Allem et al. 2017a, 2018; Czaplicki et al. 2020; Lu et al. 2020; Wang et al. 2015). Social media analysis can also be used to understand how the tobacco industry responds to policies as well as what public reactions to these policies are. For instance, an analysis of Instagram posts of nicotine/tobacco vendor accounts indicated that the nicotine/tobacco industry responded in a variety of ways to bypass the FDA’s restriction of sales of certain flavored cartridge-based e-cigarettes (Kostygina et al. 2022). These industry responses include featuring off-brand substitutes to add flavors, promoting new flavored product technologies to create a new device that could be used to vape flavors, offering do-it-yourself tutorials on making one’s own flavored e-liquid, and providing instructions on international delivery to allow access to flavored products that may no longer be sold in the United States (Kostygina et al. 2022).

Additionally, studies examining posts on X (formerly Twitter) have been used to understand the public response to the proposed ban on menthol cigarettes (Allem et al. 2023, Zhou et al. 2023). These studies observed themes related to health equity and social justice in posts about menthol cigarette bans. Specifically, positive posts tended to emphasize banning menthol cigarettes as a positive social justice move that would benefit the Black community and improve public health, while negative posts tended to relate to racial discrimination (e.g., the policy discriminating against Black people who smoke) and how alternative tobacco/nicotine products that come in menthol flavors many diminish the impact of the menthol cigarette ban (Zhou et al. 2023). These findings highlight areas that could be addressed in public health campaigns to respond to potential concerns that the public may have with the proposed policy. In summary, the ability to obtain real-time marketing information through social media analyses of novel tobacco products, the industry’s reaction to policies, and public reactions to tobacco products and policies is crucial in serving as an early warning system to inform tobacco regulatory science.

2.1.7. Conclusions and future directions

Survey studies and social media analysis are valuable clinical research tools for understanding the potential and actual impact of tobacco policies. Notably, policies restricting the sale of menthol cigarettes have been enacted in several places worldwide (e.g., Brazil, European Union, United Kingdom, Turkey, specific US states/cities), and other places (e.g., New Zealand, the United States) are considering national sales restrictions. Thus, additional survey research is needed in these areas to assess the short- and long-term impact of these policies, including identifying potential unintended outcomes and loopholes to evade polices that will be important for considering how to maximize the policy impact to benefit public health. Furthermore, purposeful survey sampling is needed across multiple priority populations to understand the potential impact of a menthol cigarette ban among those who use menthol cigarettes at high rates.

2.2. Product Choice and Marketplace Designs

Several methods used in tobacco regulatory science are based on economic principles such as the influence of price, product substitution, and demand elasticity, and these methods are used to understand how policies influence demand for tobacco products (Tidey et al. 2016). These methods can be used to examine demand across different subgroups (e.g., young adults versus older adults, those who currently use a product versus those who do not) to anticipate the potential effects of regulatory policies on these subgroups. We outline various methods and key findings of product choice and marketplace designs below.

2.2.1. Purchase tasks and the Experimental Tobacco Marketplace: method overview.

Behavioral economic paradigms, including purchase tasks (like the Cigarette Purchase Task) and the Experimental Tobacco Marketplace (ETM), have been used to answer questions about the regulation of flavors in tobacco products (Bickel et al. 2018, Tidey et al. 2016). In purchase tasks, individuals are provided with a series of options and asked to indicate how much of a specified tobacco product they would use at a range of escalating hypothetical prices (for a review and recommendations for standardization of procedures, see Reed et al. 2020). The resulting data yield several parameters, with demand for the product being measured as the amount used or purchased across the range of prices. Demand for various products can be compared to understand the degree to which one product might substitute for another, such as comparing demand for one’s own cigarette with demand for an alternative tobacco product.

In contrast to purchase tasks, which evaluate one product at a time, the ETM seeks to evaluate behavioral choices under more real-world circumstances, in which people can choose among many tobacco products simultaneously and in different combinations (Bickel et al. 2018). The impact of different regulatory policies on tobacco purchasing is simulated by systematically varying the products that are available for purchase or the price at which products are available within each marketplace. To extend these paradigms from solely a hypothetical choice to more real-world behavior, some study designs provide participants with the option to actually receive a product that they purchased from the marketplace (e.g., Kotlyar et al. 2022).

2.2.2. Purchase tasks and the Experimental Tobacco Marketplace: key findings.

Studies using hypothetical purchase tasks or ETM designs that examined the potential effects of a menthol cigarette ban found that many adults who smoke menthol cigarettes would likely switch to nonmenthol cigarettes (Denlinger-Apte et al. 2021, Kotlyar et al. 2022). However, findings also indicate demand for other combustible tobacco products such as menthol little cigars and cigarillos (Denlinger-Apte et al. 2021) or menthol e-cigarettes (Kotlyar et al. 2022) as a substitute for menthol cigarettes in the context of a menthol cigarette ban.

The ETM method can also be used to evaluate potential effects of regulatory policies and tobacco industry responses to regulatory policy. For instance, one response from the tobacco industry to the proposed menthol cigarette ban has been to adapt packaging to convey associations to menthol, even when menthol specifically is removed, as seen in Canada after this policy was implemented (Borland et al. 2019). In this regard, an ETM study in an online sample of 1,197 individuals who smoked menthol cigarettes demonstrated high demand for nonmenthol replacement products labeled “green,” which could promote continued smoking if these products were available on the market and undermine the public health benefit of a menthol ban (Guillory et al. 2020).

2.2.3. Discrete choice experiments: method overview.

Discrete choice experiments (DCEs) are another research method for investigating the potential effects of policies prior to implementation. Based on consumer choice theory (Lancaster 1966), DCEs evaluate consumers’ stated preference for goods or services, in this case a tobacco product, as a function of a set of different attributes. For example, to assess the impact of flavors on product preference, a DCE study could evaluate consumer preference for cigarettes versus e-cigarettes based on flavors and other attributes such as price, nicotine concentrations, device type, warning labels, or health effects (Buckell et al. 2019, 2023; Pesko et al. 2016). Using an experimental design, consumers are asked to choose between two or more products, with levels of the different attributes varying across several choice trials. Preferences are determined by analyzing how different levels of the attributes affect the choices. This information can then be used to estimate the effects of different policies on tobacco product choices, such as banning menthol from all tobacco products versus only from cigarettes. DCE methods enable the examination of features of tobacco products that may not yet exist or for which inadequate data are available from surveys. Similarly, the resulting preference data can be used to estimate the effects of policies that have yet to be implemented. Like other hypothetical assessments, a major limitation of this study design is that participant responses on DCE methods may not truly represent their actual behaviors.

2.2.4. Discrete choice experiments: key findings.

DCEs have compared how flavor attributes influence choices for cigarettes and e-cigarettes in the context of flavor bans. For instance, a DCE study of 2,031 adults who smoke or recently quit smoking examined preference for cigarettes and e-cigarettes under six potential flavor policies, while controlling for other attributes of both products (i.e., price, level of nicotine, and years of life lost) (Buckell et al. 2019). Participants were asked which product they would choose (cigarettes or e-cigarettes) in various scenarios (e.g., menthol available in cigarettes and all flavors in e-cigarettes, banning menthol in cigarettes and allowing all flavors in e-cigarettes, a ban on all flavors in both products) to examine the potential for switching from cigarettes to e-cigarettes, a potentially less harmful product. The scenario when menthol was banned in cigarettes and flavors were allowed in e-cigarettes was associated with the lowest preference for choosing cigarettes. Banning all flavors in cigarettes and e-cigarettes was associated with a small increase in cigarette choice; a decrease in e-cigarette choice; and the largest increase in choosing to opt out of selecting a tobacco product, which might indicate quitting.

Another study examined whether policy-relevant attributes of cigarettes and e-cigarettes might motivate adults with little interest in quitting smoking to switch to e-cigarettes (Buckell et al. 2023). Participants chose between their usual cigarettes and two e-cigarettes (pod and disposable) that varied in flavor (tobacco, menthol/mint, fruit, sweet) as well as in other attributes (e.g., nicotine concentration, health effects, and price). A latent class analysis of choices yielded two classes, termed non-switchers (68% of the sample) and switchers (32%). The non-switchers had strong preferences for their cigarettes, and the study estimated that this group would be less likely to choose e-cigarettes in the context of a menthol cigarette ban. Both groups disliked nontobacco e-cigarette flavors compared with their own cigarettes, suggesting that flavor availability in e-cigarettes may be unlikely to promote switching among adults not interested in quitting smoking. Likewise, a study in young adults (ages 18–22) who smoke and had tried e-cigarettes also identified two latent subgroups: those who prefer cigarettes and those who prefer vaping (Buckell & Sindelar 2019). Consistent with the findings in adults not interested in quitting, flavor availability in e-cigarettes did not influence cigarette choice for those who preferred cigarettes. Among those who preferred vaping, sweet and fruit flavors were valued. Both groups also valued healthier e-cigarettes and lower price. Thus, flavor availability may influence smoking behavior differently by subgroup, and increasing the price of cigarettes relative to e-cigarettes could have an impact on smoking reductions. Although the FDA cannot regulate the cost of tobacco products, states and municipalities can do so through taxation.

2.2.5. Conclusions and future directions.

Consistent with survey research and assessments of behavioral intentions, these product choice and marketplace studies suggest that a regulation prohibiting menthol in cigarettes is likely to reduce cigarette use among some individuals who smoke menthol cigarettes. However, other individuals continue to select nonmenthol cigarettes as alternatives, especially in scenarios that capitalize on the use of cigarette packaging with colors or designs associated with menthol cigarettes. In addition, DCE experiments suggest that the availability of flavors in e-cigarettes is a preferred attribute for some (e.g., young adults who prefer vaping) but not others (e.g., those who prefer cigarettes, those who are not interested in quitting). As a result, policies related to banning menthol in cigarettes and policies that might ban flavors in e-cigarettes may have different effects on these subgroups and need to be considered in estimating the public health impact. These purchase tasks have also been adapted to other tobacco products, including e-cigarettes (Cassidy et al. 2020) and little cigars (Mead-Morse et al. 2022), thereby opening avenues for research on the role of flavors in these products as well. Other important areas for further research include using these methods to delineate the anticipated effects of specific regulatory policies in subgroups of individuals, including those who smoke menthol cigarettes at disproportionately high rates, and assessing the impact of novel menthol cigarette replacements or other cooling products on choice behavior.

2.3. Controlled Laboratory Studies

Human laboratory paradigms can be used to evaluate responses to actual product use or responses to cues, such as the sight or smell of tobacco products. The resulting information on the effects of product constituents, such as menthol and other flavors (e.g., sweet flavors) on appeal and abuse liability, can inform product standards and policies to reduce the harm of tobacco products.

2.3.1. Tobacco/nicotine administration studies: method overview.

Controlled tobacco administration studies have been used to examine the independent effects of menthol flavor and the interactions with nicotine. Self-administration methods have been used with standardized research cigarettes (i.e., Spectrum cigarettes available from the National Institute on Drug Abuse), available in both menthol and nonmenthol and in several nicotine concentrations. Other self-administration methods include intravenous (IV) nicotine delivery to isolate the effects of menthol on nicotine reward that are independent of sensory interactions between menthol and nicotine in the mouth and throat (Valentine et al. 2018). Self-administration methods often include random assignment to conditions, and participants sample and rate their experience puffing the product in the laboratory under controlled conditions. Ratings may include taste (e.g., fruity, minty, sweet, bitter), chemosensory effects (e.g., coolness, irritation, harshness, throat hit), and rewarding effects (e.g., liking, wanting, satisfaction) (Carter et al. 2009). Additional methods, such as puff topography, can be used to capture information about puff duration and intensity. These taste and chemosensory studies may be followed by choice periods in which the participants choose between products previously sampled and use them ad libitum to examine how flavors influence use.

Limitations of these administration paradigms are that they typically have small sample sizes and are intensive, often requiring multiple days of in-person sessions, although recent adaptations allow participants to complete experiments from home (Peasley-Miklus et al. 2023). Additionally, following guidelines for the administration of drugs to human participants (NIDA 2019), these administration studies typically require that participants have current experience with tobacco products, limiting their utility for evaluating appeal in other subgroups such as new or former users. Additionally, in studies of menthol in cigarettes, random assignment to menthol conditions is less typical, with participants sampling menthol and nonmenthol cigarettes based on their preexisting preferences. Human laboratory tobacco administration research has not studied the effects of varied menthol concentrations in cigarettes. However, a method to amend Spectrum cigarettes to deliver different levels of menthol has been developed ( Jackson et al. 2023). In contrast, e-cigarette liquids can be purchased or made to specifications to systematically vary concentrations of both flavor and nicotine. Thus, e-cigarette laboratory studies are in a unique position to isolate the sensory effects of menthol and other flavors in combination with nicotine and independently of other irritants in tobacco smoke (Rosbrook & Green 2016).

2.3.2. Tobacco/nicotine administration studies: key findings.

Self-administration studies have examined the effects of menthol versus nonmenthol cigarettes at varying nicotine concentrations to understand the interactive effects on perceptions of how much “liking,” “satisfaction,” “nicotine,” or “flavor” was experienced and how “strong” the cigarette was. These studies have generally, but not always (Perkins et al. 2018), found that individuals who smoke menthol cigarettes report similar positive subjective effects across a range of different nicotine concentrations (Karelitz & Perkins 2021, Perkins et al. 2017, Rubenstein et al. 2022). For example, one study that tested six nicotine concentrations, ranging from 0.4 to 17.4 mg/g, found that those who smoked menthol cigarettes reported relatively consistent positive subjective effects across all but the highest nicotine concentration, while liking and satisfaction were low for nonmenthol cigarettes at low nicotine concentrations and increased with greater nicotine concentrations (Karelitz & Perkins 2021). These findings suggest that menthol may enhance the reinforcing effects of cigarettes even at low concentrations of nicotine. However, these effects may be specific to inhaled menthol and nicotine from cigarettes. For instance, IV nicotine administration paradigms have been used along with e-cigarette flavor delivery to test interactions of menthol and nicotine that are independent of the nicotine effects experienced in the mouth and throat when nicotine is inhaled or administered orally (MacLean et al. 2021, Valentine et al. 2018). The findings indicated that those who prefer menthol (versus nonmenthol) cigarettes had more blunted positive effects of IV nicotine across a range of menthol concentrations (MacLean et al. 2021), so inhaled menthol did not modify the positive subjective effects of IV nicotine (Valentine et al. 2018). Additionally, menthol flavor reduced cravings and urges to smoke, irrespective of nicotine concentration, and this effect was specific to those who prefer menthol cigarettes but not nonmenthol (Valentine et al. 2018), supporting the notion that menthol flavor acts as a conditioned cue that can alleviate urges to smoke.

To isolate the effects of menthol in cigarettes, other tobacco administration studies have used crossover designs in which participants smoke menthol and nonmenthol cigarettes at separate 2-week intervals and then provide subjective ratings of their experience smoking the designated cigarettes in a laboratory session (Strasser et al. 2013, Watson et al. 2017). The findings indicate that for adults who usually smoke menthol cigarettes, switching from menthol to nonmenthol resulted in more negative ratings of taste, mildness, aftertaste, and smoke smell (Strasser et al. 2013). Similarly, a separate crossover study that included adults who preferred menthol or nonmenthol cigarettes indicated that smoking the nonpreferred cigarette flavor was less enjoyable, less satisfying, and more irritating, and this effect was more pronounced for those who preferred menthol cigarettes (Watson et al. 2017). These studies suggest that, in the context of a menthol ban, the taste and experience of smoking nonmenthol cigarettes will be less appealing for those who prefer menthol.

To inform a product standard about menthol concentrations in tobacco products, e-cigarette administration studies have sought to examine how various menthol concentrations influence user experience (Rosbrook & Green 2016). In a youth sample, menthol flavors improved the taste of e-liquids even at a barely perceptible concentration (0.5%), and the high concentration of menthol (3.5%) increased liking/wanting of the product more than no menthol (Krishnan-Sarin et al. 2017). Additionally, there was a trend toward a nicotine–menthol interaction in which liking for the high concentration of nicotine (12 mg/mL) was increased by high menthol compared with low menthol (0.5%) or no menthol. Similar results have been found in adults (Rosbrook & Green 2016). These data suggest that menthol may contribute to the use of tobacco products with higher nicotine concentrations by altering sensory experiences, providing coolness, reducing the harshness of nicotine, and increasing liking/wanting.

2.3.3. Cue-exposure paradigms: method overview.

Given that tobacco and nicotine administration paradigms are limited to studying effects in those who have experience using the product, other cue-exposure methods are useful to study appeal in nonusers. For instance, studies in children who have never used tobacco products have measured liking and interest in using cigarettes and e-cigarettes following exposure to advertisements featuring flavored versus nonflavored products (Vasiljevic et al. 2016). Biomarkers of brain responses to these cues also provide an innovative measure of risk that is not subject to reporting biases (Falk et al. 2010, Garrison et al. 2018). These cue-exposure paradigms can also be used to study how flavors in packaging or other marketing materials contribute to appeal and intentions to use tobacco products among those who currently smoke (Shi et al. 2023).

2.3.4. Cue-exposure paradigms: key findings.

Cue-exposure studies have evaluated how representations of flavors on packaging and in advertisements may contribute to the appeal of tobacco products. For example, briefly viewing menthol cigarette packaging prior to smoking cues induced greater smoking cue reactivity, as measured with functional magnetic resonance imaging, than viewing nonmenthol packaging among participants who typically smoked menthol cigarettes (N = 42) compared with those who smoked nonmenthol cigarettes (N = 33) (Shi et al. 2023). Among young adults who did not smoke cigarettes (N = 26), however, there was no difference in brain activation to menthol versus nonmenthol cigarette packaging. These data suggest that the impact of menthol advertisements and packaging differs depending on the population, such that menthol cigarette packaging may be most important for those already smoking menthol cigarettes. In contrast, advertisements for sweet flavors in tobacco products could have a more general effect, given that most people have prior experience with sweet flavors in foods and beverages. For example, advertisements for sweet-flavored e-cigarette liquids elicited greater brain activation and higher ratings of liking and intent to try in comparison to tobacco-flavored e-cigarettes among young adults, and these advertisements interfered with visual attention to and memory of health warnings about the tobacco product (Garrison et al. 2018).

2.3.5. Conclusions and future directions.

Laboratory studies employing tobacco/nicotine administration and cue-exposure paradigms provide valuable information about the role of menthol flavor or other (e.g., sweet) flavors in appeal, liking, and subjective experience. Several methodological issues and areas of inquiry deserve attention in future studies. For example, current paradigms using e-cigarettes to study flavors vary in the number of puffs taken during directed self-administration, ranging from 1 to 10 puffs. Although fewer puffs increase feasibility, a single puff may be inadequate to model user experiences because the effects of nicotine may require greater exposure and because perceptions of a flavor, such as menthol, change with repeat exposure (Dessirier et al. 2001). In addition, it would be helpful to establish consensus about dimensions on which flavors should be rated, with assessments of liking, sweetness, coolness/coldness, harshness/irritation, and smoothness considered as core measures. Tobacco/nicotine administration and cue-exposure studies also need to consider how prior tobacco product experience (e.g., e-cigarettes, cigarettes, menthol, nonmenthol) may moderate the effects of interest.

2.4. Clinical Trial Designs

Clinical trial designs, such as those traditionally employed to study behavioral and pharmacological treatments, have been applied to tobacco regulatory science research to evaluate key policy impact questions, including how restricting menthol cigarettes influences smoking behavior, nicotine dependence, and quitting among adults who currently smoke menthol cigarettes. Other key questions relate to the potential of other flavored products such as e-cigarettes to be an alternative to smoking combustible cigarettes, including whether individuals who smoke menthol cigarettes will switch to flavored e-cigarettes, and what the impact of switching is on dependence, interest in quitting, and health.

2.4.1. Clinical trial designs: method overview

In general, clinical trials enroll adults who currently smoke cigarettes who may or may not be planning on quitting smoking in the near term but are interested in switching, or willing to switch, to other products. These studies typically evaluate the impact of switching participants from usual-brand combustible cigarettes to other tobacco products (e.g., nonmenthol cigarettes, other flavored products like menthol e-cigarettes). The primary outcomes often include the use of the tobacco product (e.g., cigarettes, e-cigarettes) measured through self-reports and/or structured assessments of daily product use (e.g., time-line follow-back interviews; Sobell & Sobell 1992). Counts of spent filters and unused cigarettes or used e-cigarette pods/cartridges are other methods for quantifying use (Bold et al. 2020). In addition, biomarkers of recent smoking or other tobacco use are often measured to confirm and quantify changes in smoking (e.g., carbon monoxide, plasma nicotine, urine cotinine, and/or nicotine levels) (Bold et al. 2020, Kotlyar et al. 2022).

The benefits of clinical trial designs include evaluating effects in more naturalistic conditions to complement and extend the methods in more tightly controlled self-administration studies that have shorter exposures. Clinical trial methods also overcome limitations in other designs that assess hypothetical choices, since clinical trials assess actual behavior and product use. However, limitations include the possibility of contamination or problems with study adherence, since trials are conducted in the real world and people may still have access to tobacco products other than what has been assigned for the study. For instance, studies that assess the potential impact of regulations that prohibit menthol cigarettes may assign people to use nonmenthol cigarettes instead of their usual menthol cigarettes. To encourage adherence and to monitor use of non-assigned cigarettes, studies may give free tobacco products based on the assigned condition, collect measures of used and unused products to confirm use, and examine biomarkers such as urine menthol glucuronide (e.g., Bold et al. 2020).

2.4.2. Clinical trial designs: key findings.

Several product-switching clinical study designs have been used. For example, a nonrandomized trial recruited 29 adults who smoked menthol cigarettes and switched all participants to nonmenthol cigarettes to model the effects of a menthol cigarette ban (Bold et al. 2020). Findings indicated significant reductions in cigarettes per day and nicotine dependence as well as significant increases in quitting motivation and confidence. Notably, race was a significant moderator of smoking outcomes, such that the largest reductions in cigarettes per day were observed among Black individuals. A separate randomized study specifically recruited Black individuals who smoked menthol cigarettes and who were interested in quitting (Kotlyar et al. 2021a). Participants were randomized to continue smoking menthol cigarettes (N = 60) or switch to nonmenthol cigarettes (N = 62) for 1 month prior to their chosen quit date. During this 1-month period, those who switched to nonmenthol cigarettes smoked fewer cigarettes per day, had significantly lower withdrawal symptom severity, and had higher perceived effectiveness at quitting. After the quit date, those in the nonmenthol cigarette group had longer delayed times to lapse or relapse compared with the menthol cigarette group, although the differences in time to lapse or relapse were not statistically significant (Kotlyar et al. 2021b).

Other studies have used a switching trial design to model regulatory scenarios in cigarettes, e-cigarettes, or both products to understand the impact of restricting available e-cigarette flavors on smoking behavior and cessation among adults who smoke. For instance, one study examined the effects of switching adults who smoke menthol cigarettes to either menthol-flavored (N = 39) or tobacco-flavored (N = 25) e-cigarettes (Nollen et al. 2023). Both groups had substantial reductions in cigarettes per day and high ratings of e-cigarette liking and acceptability. Another study used a clinical trial design with an experimental marketplace simulation in which 47 adults who smoked menthol cigarettes were randomized to conditions of (a) no menthol ban, (b) a menthol ban for cigarettes but not e-cigarettes, or (c) a menthol ban for both cigarettes and e-cigarettes (total ban), and participants could choose the product they wanted to use for 6 weeks (Kotlyar et al. 2022). More participants selected the e-cigarette in the condition with a menthol ban for cigarettes but not e-cigarettes (69%), compared with 38% no ban and 40% total ban conditions. Although there were no statistically significant differences in cigarette smoking, exploratory analyses indicated that those who used menthol e-cigarettes in these scenarios (versus no e-cigarette use) smoked fewer cigarettes, suggesting that menthol e-cigarettes may be a substitute for adults who smoke menthol cigarettes, especially in the context of a menthol cigarette ban.

2.4.3. Conclusions and future directions.

Clinical trial data from these switching study designs are valuable for assessing behavior outcomes across a range of possible regulatory scenarios. Studies to date have had small sample sizes that were likely not sufficiently powered to detect statistically significant differences, and further large-scale clinical trials are needed and are currently underway. Additional research with clinical trial designs could address important research questions such as investigating the impact of novel nonmenthol cooling products or the interaction of both flavor and nicotine concentration on real-world smoking behavior and health outcomes to inform policy decisions.

2.5. Simulation Models

Simulation or population-level modeling studies are used to estimate the potential impact of a policy before it goes into effect using advanced statistical models. Simulation studies have examined national-level (Issabakhsh et al. 2024, Levy et al. 2011b) and regional-level (Li et al. 2021) effects of policies in the United States, as well as internationally (e.g., in Singapore; Zeng et al. 2022).

2.5.1. Simulation models: method overview.

Simulation models use current rates of smoking prevalence and initiation, as well as associated outcomes of disease and death and economic impact, to project estimates of the future effects on tobacco use outcomes (e.g., rates of initiation, continued use, cessation) given a status quo scenario (no change in policy) in contrast to a new policy (e.g., prohibiting menthol flavor in cigarettes). Benefits of simulation models include the ability to estimate predictions on a population level before policies go into effect. Simulation models can also consider the effect of these scenarios among specific priority populations, for example, by sex or race/ethnicity. Simulation models have inherent limitations in estimating future projections because they are based on current evidence and are contingent on assuming accurate input values, such as rates of initiation and cessation. Evidence from other clinical research, as described in this review, can be used to inform the input values to yield more accurate estimates of the impact of future policies on the population.

2.5.2. Simulation models: key findings.

A seminal simulation study quantified the potential impact of a policy prohibiting menthol flavor in cigarettes in the United States, based on various scenarios (Levy et al. 2011b). In the most conservative scenario (i.e., a 10% reduction in initiation and a 10% increase in cessation), the menthol ban would result in more than one million fewer individuals smoking 10 years after the policy implementation. Additionally, the simulation model estimated that up to 633,252 smoking-attributable deaths would be averted over 40 years, onethird of them among Black individuals, a population that has disproportionately high rates of menthol cigarette use and associated tobacco-related health disparities (Levy et al. 2011b). More recent simulation studies have shown similar effects and indicate that a menthol cigarette ban would lower rates of initiation and continued cigarette smoking and reduce life-years lost and premature deaths (Issabakhsh et al. 2024, Levy et al. 2023), particularly among Black individuals (Issabakhsh et al. 2024). Other outcomes that have been examined include rates of smoking-related disease and economic costs. The findings indicate that a menthol cigarette ban would substantially lower the rates of smoking-related disease, including rates of myocardial infarction and stroke (Li et al. 2021), and reduce health care costs per person by an average of $1,836, or $1.62 billion among all adults who smoke, over a 20-year period (Li et al. 2021). Simulation models have also been expanded to estimate the effects of a menthol ban in cigarettes and cigars (Levy et al. 2023) and to estimate effects of more or less conservative flavor restrictions across international regions (e.g., a full ban versus a partial ban that would exclude menthol and clove cigarettes; Zeng et al. 2022).

2.5.3. Conclusions and future directions.

Simulation models are valuable for estimating the impact of potential policies on population-level changes in behavior and health outcomes. Simulation models are also dynamic and can be readily applied as changes in the tobacco product marketplace rapidly evolve, including changes related to tobacco industry responses to federal policies. Future research using simulation models could address important questions, including examining how novel nonmenthol cooling products influence behavior and health outcomes in various policy scenarios, the impact of flavor regulations for other tobacco products on smoking behavior and smoking-related health outcomes, the impacts of potential policies across specific priority populations, and which policy scenarios help promote health equity and reduce tobacco-related health disparities.

2.6. Integration

Overall, these clinical research findings provide evidence supporting the FDA’s proposed policy to prohibit menthol in cigarettes. These studies indicate that flavors in general are appealing in tobacco products and that menthol specifically increases the reward and reinforcement from cigarettes, as shown in tobacco/nicotine administration studies. Thus, removing menthol from cigarettes should reduce rates of cigarette initiation and continued use and make cigarettes less rewarding for those who currently smoke, thus supporting quitting. Studies of behavioral intentions show that some adults who smoke menthol cigarettes anticipate that they would switch to a nonmenthol cigarette or alternative tobacco product in the context of a menthol cigarette ban, although survey studies examining the effect after a menthol ban policy was implemented indicate that a greater proportion of people quit smoking than anticipated. Furthermore, studies that intentionally switch adults from menthol to nonmenthol cigarettes show reductions in smoking and nicotine dependence. However, findings indicate that the magnitude of the effect of a menthol cigarette ban on tobacco use and public health may be influenced by the availability of other flavored tobacco products, as noted in several research designs.

Thus, findings converge across these methods to support the policy of removing menthol as a characterizing flavor from cigarettes to reduce the appeal, use, and addictive potential of cigarettes. However, there are areas of conflict, specifically (a) in considering the impact of flavor availability for other products, such as e-cigarettes, in the context of a menthol cigarette ban and (b) in balancing concerns of appeal and initiation among nonusers alongside the goals of reducing cigarette smoking and related harm. Moreover, the impact of other flavored tobacco products in the context of a menthol cigarette ban may be influenced by several factors related to the population being studied (e.g., age, race/ethnicity), as well as other individual factors (e.g., nicotine dependence, nicotine metabolism, prior experience with alternative products). These are important areas for further research that can directly inform FDA regulatory actions.

3. Summary and Areas for Future Research

The FDA has taken historic initiatives to ban characterizing flavors from cigarettes (in 2009) and restrict flavors in cartridge-based e-cigarettes to only tobacco and menthol flavors (in 2020), and it has announced a proposed standard to prohibit menthol flavor in cigarettes and characterizing flavors including menthol in cigars (in 2022), although this ruling has not yet been finalized. These policy initiatives have been informed by the results of tobacco regulatory research, such as those described in this review. This review highlights the benefits and limitations of these research methods and emphasizes the complexities in studying the potential effects of tobacco policies. Future studies should consider mixed methods in evaluating flavors in tobacco products to inform tobacco regulatory science to overcome the weaknesses of each method alone with the strengths of the other methods. Notably, given the growing diversity of tobacco products and the changing tobacco marketplace, it is important to consider the impact of other product availability in the context of a policy change, as a policy change does not happen in isolation.

Additionally, it is important to examine and consider the tobacco industry’s innovations in response to policies, and more comprehensive regulation may be needed. Specifically, the effect of synthetic coolants on nicotine and tobacco use is a pressing issue for tobacco regulatory research. Other natural and synthetic ingredients are available and may be used as alternative cooling agents to substitute for menthol. Synthetic chemicals are especially concerning since they have distinct cooling effects but lack distinguishing characterizing flavors or aromas, so it is unclear whether they are included in the current FDA guidelines banning characterizing flavors. Thus, to maximize the public health benefit, it may be important to consider policies that extend beyond menthol and characterizing flavors alone and also consider the sensory effects of flavor chemicals.

Other critical areas for future research remain. For instance, understanding how to encourage successful smoking cessation among those who smoke menthol cigarettes, especially in the context of a menthol cigarette ban, is an area of high priority for clinical research. Maximizing cigarette cessation rather than switching to nonmenthol cigarettes, especially among those who use menthol cigarettes at disproportionate rates, will be essential for reducing individual harm and improving public health. Achieving these goals will require supplementing policy enforcement with public health education, greater access to cessation resources, and community input to identify strategies most needed to help people quit. Relatedly, it is essential to consider other individual- and system-level factors that are associated with tobacco product use, including targeted marketing by the tobacco industry. Ongoing research is needed to identify the ways the tobacco industry continues to evolve and target consumers in the context of new flavor policies. Research is needed to identify effective public health messaging to promote smoking cessation and effective harm reduction. While there is evidence that switching to noncombustible products (e.g., e-cigarettes) may promote cigarette abstinence among adults who smoke (Lindson et al. 2024), little research has examined the role of specific flavors in these noncombustible products in supporting cessation (Hung et al. 2021). Lastly, more research is needed to understand the intersection of potential policies across products, such as whether policies should be implemented concurrently or sequentially to maximize the public health impact.

There is substantial scientific evidence to support the proposed FDA policy prohibiting menthol flavor in cigarettes for the protection of public health. Furthermore, comprehensive policies that prohibit cooling flavor additives in cigarettes should be considered to close possible loopholes that are being exploited by the tobacco industry to circumvent menthol cigarette regulations. As the tobacco marketplace continues to evolve with new products and flavor chemicals, clinical scientists will play an essential role in advancing research to inform tobacco policy.

SUMMARY POINTS.

1. Clinical science is essential to informing evidence-based policies to protect public health and reduce tobacco-related health disparities.

2. The US Food and Drug Administration (FDA) has already enforced regulations including banning characterizing flavors (other than menthol) from cigarettes (since 2009) and prioritizing enforcement of the removal of flavors other than tobacco and menthol in cartridge-based e-cigarettes (since 2020), and it has announced a proposed standard to prohibit menthol flavor in cigarettes and characterizing flavors including menthol in cigars (in 2022).

3. Menthol cigarette smoking is disproportionately high among specific priority populations, including youth and young adults, certain racially/ethnically minoritized groups such as Black individuals, pregnant individuals, low-income individuals, and individuals who identify as LGBTQIA+.

4. A range of clinical research designs, including survey methods, product choice and marketplace studies, self-administration and cue-exposure studies, clinical trials, and simulation models, have provided substantial scientific evidence to support the proposed FDA policy prohibiting menthol flavor in cigarettes for the protection of public health.

5. The tobacco industry continues to evolve by creating new products, including cigarettes with cooling flavor additives, to circumvent menthol cigarette regulations, and studies need to consider the impact of existing and emerging flavors in other tobacco products that may undermine the intended flavor ban.

FUTURE ISSUES.

1. More information is needed to understand the impacts of potential policies across specific priority populations, including which policy scenarios help promote health equity and reduce tobacco-related health disparities.

2. Studies are needed to examine policy effects and unintended consequences to understand gaps in enforcement and loopholes that may undermine the policy effects.

3. Research is needed to understand the impact of flavors in other noncigarette tobacco products and the impact of novel, nonmenthol cooling products on behavior and health outcomes, especially in the context of a menthol cigarette ban.

4. Research is needed to understand how to encourage successful smoking cessation among those who smoke menthol cigarettes, especially in the context of a menthol cigarette ban.

5. Future studies should consider mixed methods in evaluating the effects of flavors in tobacco products to overcome the weaknesses of each method alone with strengths of the other methods.