Smoking rates in the United States have declined substantially since the release of Smoking and Health: Report of the Advisory Committee to the Surgeon General of the Public Health Service in 1964, when the prevalence of current cigarette smoking was around 42 percent. Recent estimates reveal that since 1964, tobacco control in the United States has led to 8 million fewer premature deaths and has extended the mean life span at age 40 by about 2 years (Holford et al., 2014). However, tobacco use continues to have major public health implications; while the prevalence of current cigarette smoking among U.S. adults has declined to around 18 percent (Schiller et al., 2014), more than 42 million American adults still smoke (HHS, 2014).
The Family Smoking Prevention and Tobacco Control Act of 2009 (hereafter referred to as the Tobacco Control Act) amended the Federal Food, Drug, and Cosmetic Act, granting the Food and Drug Administration (FDA) broad authorities over tobacco products. The Tobacco Control Act directed FDA to, among other things, issue regulations to restrict cigarette and smokeless tobacco retail sales to youth and to restrict tobacco product advertising and marketing to youth. The act, however, prohibits FDA from taking several specific steps, including establishing a minimum age of sale
of tobacco products to persons over 18 years of age.1 On the other hand, the Tobacco Control Act directed FDA to convene a panel of experts to conduct a study on “the public health implications of raising the minimum age to purchase tobacco products” and to submit a report to Congress on the issue.
In August 2013 FDA contracted with the Institute of Medicine (IOM) to convene a committee to:
- Examine existing literature on tobacco use initiation, and
- Use modeling and other methods, as appropriate, to predict the likely public health outcomes of raising the minimum age for purchase of tobacco products to 21 years and 25 years.
The resulting IOM Committee on the Public Health Implications of Raising the Minimum Age for Purchasing Tobacco Products, assembled to address these issues, was composed of experts in public health law, the epidemiology of tobacco use and tobacco risks, adolescent and young adult development, risk behaviors and perceptions, public health policy and practice, and public policy modeling.
During a discussion at the first public meeting of the committee, a representative of the Center for Tobacco Products of FDA urged the committee to include in its analysis the impact of raising the minimum age of legal access to tobacco products (MLA) to 19 years of age. The public health impacts examined in this report include tobacco initiation, prevalence, morbidity, and mortality. The committee uses the term “tobacco product” to mean any product covered by FDA regulatory authority, although most of the literature and the modeling focus on cigarettes. The committee did not consider the economic impact of raising the MLA, nor did it compare the effects of raising the MLA with other youth-oriented tobacco control policies.
The Tobacco Control Act refers to both minimum age for purchase2 and minimum age for sale.3 The committee focused on the implications of raising the MLA in the context of the body of youth access laws and enforcement policies currently in place across the country. These laws and policies vary considerably, not only in the scope of conduct that is prohib-
1 Family Smoking Prevention and Tobacco Control Act of 2009, Public Law 111-31 § 906. 111th Cong. (June 22, 2009).
2 Id. § 104.
3 Id. § 906.
ited but also in the prescribed penalties for violations. What they all have in common, however, is a focus on curtailing retail access to tobacco products by underage persons, with little, if any, emphasis on punishing the underage users of tobacco products. The committee’s charge requests conclusions regarding the public health implications of raising the MLA without any recommendations regarding whether the MLA should be raised.
Brain development continues until about age 25. While the development of some cognitive abilities is achieved by age 16, the parts of the brain most responsible for decision making, impulse control, sensation seeking, future perspective taking, and peer susceptibility and conformity continue to develop and change through young adulthood. Adolescent brains are uniquely vulnerable to the effects of nicotine and nicotine addiction. Adolescent and young adult developmental trajectories may be altered by social and environmental contextual influences, including normative developmental transitions into and out of school or work or changes in living arrangements or relationships.
According to the most recent results from an annual survey of adolescents in grades 8, 10, and 12, American teens are smoking less than ever before (Johnston et al., 2014b). Cigarette smoking in this age group peaked in 1996–1997 before beginning a fairly steady and substantial decline that continued through the mid-2000s. This decline in adolescent smoking has continued since then, but at a slower rate (HHS, 2014). Data from 2012 show that 34.1 percent of Americans between 21 and 25 were current cigarette users, making that the age group with the highest prevalence of cigarette smoking (SAMHSA, 2013). While almost 90 percent of people who have ever smoked daily first tried a cigarette before 19 years of age, the fact that nearly all others who ever smoked daily tried their first cigarette before the age of 26 should not be overlooked (see Table 2-8 in Chapter 2). Additionally, only 54 percent of daily smokers are smoking daily before age 18, but 85 percent are doing so by age 21 and 94 percent before age 25. These data strongly suggest that if someone is not a regular tobacco user by 25 years of age, it is highly unlikely they will become one.
Although most states currently set the minimum age of legal access to tobacco at 18, four states set it at 19, and New York City and several other localities around the country have raised the MLA to 21. All 50 states and
the District of Columbia prohibit commercial transfers to underage persons, while 48 states and the District of Columbia also prohibit noncommercial transfers (e.g., giving, exchanging, bartering, furnishing, or otherwise distributing tobacco). Based on random, unannounced compliance inspections of tobacco retailers, the national average rate of tobacco sales to underage individuals (i.e., noncompliance) in 2013 was 9.6 percent.
Active enforcement of tobacco minimum age restrictions, including meaningful penalties for violations, increases retailer compliance and decreases the availability of retail tobacco to underage persons. However, it is difficult to know precisely how much increasing retailer compliance reduces the availability of retail tobacco to underage persons or how much the decreased retail availability of tobacco affects underage tobacco use because of the continued availability of tobacco from noncommercial sources. Underage users rely primarily on “social sources” (friends and relatives) to get tobacco, and there is little evidence that underage individuals are obtaining tobacco from the illegal commercial market. Bans on the noncommercial distribution of tobacco by friends, proxy purchasers, and other social sources are not well-enforced.
Through an iterative and consensus-driven process, the committee considered how these age-related effects would translate into potential changes in the rates of initiation across different age segments through adolescence and young adulthood for each of the three policy options (raising the MLA to 19, 21, or 25 years of age). The committee assigned ordered, categorical labels to its estimates as small, medium, or large. The committee attached numeric ranges to each of the magnitude estimate descriptors for use in the modeling. The committee used increments of 5 percent, ranging from 5 to 30 percent, to quantify the range of possible changes in initiation rates for use in the models. The committee has more confidence in its estimates pertaining to raising the MLA to 19 or 21 than in its estimates pertaining to raising the MLA to 25 because of the greater level of extrapolation needed for estimating change and also other factors that appear with increased age.
Conclusion 7-1: Increasing the minimum age of legal access to tobacco products will likely prevent or delay initiation of tobacco use by adolescents and young adults.
The definition of “initiation” used in this report, including in the modeling, is having smoked 100 cigarettes. This definition is based on data obtained from the National Health Interview Survey. Smoking at least 100 cigarettes in one’s lifetime goes beyond occasional trying or “experimenta-
tion.” To achieve the benchmark of 100 cigarettes, one must have access to cigarettes over a period of time and have developed symptoms of dependence and stronger motives for use beyond perceived peer or social group pressure (Dierker and Mermelstein, 2010).
A critical component in the development of dependence and continued tobacco use is the reinforcing effects of nicotine. Adolescent brains have a heightened sensitivity to the rewarding effects of nicotine, and this sensitivity diminishes with age (Adriani et al., 2006; Jamner et al., 2003). Thus, the probability that a user escalates to dependence after the first few trials is likely to decrease the further one moves away from adolescence.
Changes in the initiation of tobacco use would not necessarily be linear with increases in the MLA or be equal for all segments of underage individuals. Changing the MLA has an indirect effect of helping to change norms about the acceptability of tobacco use, but this effect may take time to build. In addition, the norms about acceptability of tobacco use are also likely to vary by age, with greater perceived unacceptability for those the farther away from the MLA. If the MLA increases to 21, the social unacceptability of smoking will be greater for a 16-year-old than for a 20-year-old.
Given the assumption that changes in the MLA could have differential effects on adolescents at different ages, the committee considered possible changes in initiation rates for three age divisions: (1) adolescents under age 15; (2) adolescents between the ages of 15 and 17; and (3) individuals at age 18 for estimates with an MLA of 19, or individuals at ages 18 to 20 or 21 to 24 for an MLA of 21 or 25, respectively. These age groupings reflect not just differences in years from the MLA but also several important developmental transitions that play a role in tobacco use.
Conclusion 7-2: Although changes in the minimum age of legal access to tobacco products will directly pertain to individuals who are age 18 or older, the largest proportionate reduction in the initiation of tobacco use will likely occur among adolescents 15 to 17 years old.
Conclusion 7-3: The impact on initiation of tobacco use of raising the minimum age of legal access to tobacco products (MLA) to 21 will likely be substantially higher than raising it to 19, but the added effect of raising the MLA beyond age 21 to age 25 will likely be considerably smaller.
Many adolescents under age 15 are not yet in high school or of driving age. Adolescents under age 15 are less likely to have coworkers or members
of their peer networks who are over the MLA (with the likelihood decreasing as the MLA increases). Thus, social network sources and mobility are most restricted for adolescents under age 15. For adolescents under 15 years of age, raising the MLA from 18 to 19 may have only a modest impact on reducing social sources, given the small difference in age. Increasing the MLA to 21, however, would provide a greater distancing of social sources. Although 19-year-olds may still be in high schools and thus potentially influence those under 15, it is far less likely that 21-year-olds are in the same social networks. On the other hand, increasing the MLA from 21 to 25 will not be likely to achieve many additional notable reductions in social sources for those under 15 beyond what is achieved with an MLA of 21.
Although social sources play a central role in establishing adolescent tobacco use patterns, other factors that contribute to early adolescent tobacco use (for those who initiate before age 15) may limit the reductions that would be achieved with increases in the MLA. Adolescents who reach a level of 100 cigarettes before 15 may be those who are most susceptible to the reinforcing effects of nicotine, who have higher levels of psychological or substance use comorbidities, who have a combination of problem behaviors (of which tobacco use is one manifestation), and who have social networks within which tobacco and other substances are more readily available, regardless of age. Thus, the committee also expects that there may be limits to how much changes in the MLA will affect this subset of adolescents. Considering the balance of these factors, the committee estimates that for adolescents under age 15 reductions in initiation will be small for an MLA of 19 and medium for an MLA of 21 and an MLA of 25.
The committee expects that the greatest gains in reducing tobacco use will be achieved for adolescents between the ages of 15 and 17. Negative consequences for tobacco use, through parental or school controls, are still relevant, and changes in the MLA are likely to increase these negative consequences as social norms adjust. Adolescents in this age group are still most likely to get tobacco through social sources (committee analysis of Arrazola et al., 2014; Johnston et al., 2014a). Between the ages of 15 and 17 adolescent mobility increases with driving privileges. Social networks and potential social sources of tobacco start to increase as some adolescents take on formal, part-time jobs with coworkers who may be over the MLA. Changing the MLA to 19 may not change social sources substantially for these adolescents, but the committee expects that raising the MLA to 21 will substantially impact initiation. Raising the MLA to 25 may provide only a modest additional reduction in initiation over that achieved with an MLA of 21, given that changes to social network sources may not be substantially different.
Balancing these factors, the committee estimates that the reduction in initiation in this age group will likely exceed that seen in adolescents less than
15 years of age for all policy options. Furthermore, the committee estimates that the higher the MLA, the greater the effect on initiation rates will be.
By age 18, many adolescents graduate from high school and have numerous life transitions, including entering higher education, exposure to more adults in the workforce, leaving home, and significant changes in social networks. Patterns of initiation to date also show a tailing off of initiation by age 18 (committee analysis of Johnston et al., 2014a). Given that the social networks of 18-year-olds overlap more with 19-year-olds, the committee expects a small reduction in initiation for 18-year-olds for an MLA of 19. The committee expects similar effects on initiation rates for 19- and 20-year-olds as for 18-year-olds with an MLA of 21 or 25. This expectation of increased effect is due primarily to the increased social distancing expected when the MLA is raised to 21 or 25, but it also takes into account the benefit of the additional maturing of executive functions among young adults, the decreased sensitivity to the rewarding properties of nicotine, the additional social norms proscribing tobacco use, and tobacco’s decreased social value and the decreased motives for use as individuals enter the workforce or parenthood.
Changes in initiation for young adults in the 21–24 age group were considered only for the case of raising the MLA to 25. Even under the current MLA of 18, the probability of initiation at these ages is substantially lower than for adolescents and younger adults. However, current patterns of tobacco marketing suggest that young adults are increasingly targeted in tobacco promotions (Ling and Glantz, 2002), and tobacco promotions are frequently linked with bar settings and alcohol consumption, which may also keep this age group susceptible to initiation (Ling and Glantz, 2002). In addition, the committee considered that there may be more lax enforcement for an MLA of 25. Considering the balance of factors, the committee expects that some reduction in initiation will still occur with an MLA of 25 but that this reduction will be small.
Conclusion 7-4: Based on the modeling, raising the minimum age of legal access to tobacco products, particularly to age 21 or 25, will likely lead to substantial reductions in smoking prevalence.
Two tobacco simulation models commissioned by the committee, SimSmoke and the Cancer Intervention and Surveillance Modeling Net-
work (CISNET) smoking population model, suggest significant reductions in smoking prevalence from 2015 to 2100 in the United States, even under a status quo scenario with regard to the MLA; these declines reflect ongoing benefits from prior tobacco control policies. The models predict that raising the MLA would lead to considerable additional reductions in smoking prevalence based on the committee’s conclusions about the likely reductions in smoking initiation described above. Specifically, both models estimate that raising the MLA will lead to approximately a 3 percent decrease in smoking prevalence for an MLA of 19, a 12 percent decrease for an MLA of 21, and a 16 percent decrease for an MLA of 25 above and beyond the decrease predicted in the status quo scenario.
Given the likelihood that raising the MLA would decrease the rates of initiation of tobacco use by adolescents and young adults, it follows that tobacco-related disease and death would also decrease, generally in proportion to the decrease in tobacco use.
Conclusion 8-1: Based on the modeling, raising the minimum age of legal access to tobacco products will likely lead to substantial reductions in smoking-related mortality.
Conclusion 8-2: Based on a review of the literature, raising the minimum age of legal access to tobacco products (MLA) will likely immediately improve the health of adolescents and young adults by reducing the number of those with smoking-caused diminished health status. As the initial birth cohorts affected by the policy change age into adulthood, the benefits of the reductions of the intermediate and long-term adverse health effects will also begin to manifest. Raising the MLA will also likely reduce the prevalence of other tobacco products and exposure to secondhand smoke, further reducing tobacco-caused adverse health effects, both immediately and over time.
Adolescents and adults most commonly use tobacco in the form of cigarettes, and the adverse health effects of cigarettes are best documented among all the various forms of tobacco use. Cigarette smoking is causally associated with a broad spectrum of adverse health effects that begin soon after the onset of regular smoking and significantly diminish the health status of the smoker compared to nonsmokers. Cigarette smoking causes many adverse health effects with an intermediate latency, such as subclinical atherosclerosis, impaired lung development and function, diabetes, periodontitis, exacerbation of asthma, subclinical organ injury, and adverse sur-
gical outcomes. Cigarette smoking is also causally associated with a broad spectrum of long-latency adverse health effects, such as chronic obstructive pulmonary disease, coronary heart disease, and numerous cancers, that cause suffering, impaired quality of life, and premature death. Results from both models suggest that reductions in smoking-related mortality following an increase in the MLA will be large but will not be observed for at least 30 years after the increased MLA takes effect. For example, if the MLA were raised now to age 21 nationwide, modeling suggests that for the cohort of people born between 2000 and 2019 there would be approximately 10 percent fewer lifetime premature deaths, lung cancer deaths, and years of life lost (YLL) from cigarette smoking. Given the status quo projections, this translates to approximately 249,000 fewer premature deaths, 45,000 fewer deaths from lung cancer, and 4.2 million fewer YLL.4
Smoking combustible tobacco products other than cigarettes, such as pipes and cigars, is causally associated with a broad spectrum of adverse health effects. The impact of raising the MLA on morbidity and mortality from these products would depend on the risk profile of each product and the degree to which that product is used in the population over time. Raising the MLA can also be expected to lessen exposure to secondhand smoke from cigarettes and other combustible tobacco products. Secondhand smoke exposure is causally associated with a number of adverse health effects.
Conclusion 8-3: Based on a review of the literature and on the modeling, an increase in the minimum age of legal access to tobacco products will likely improve maternal, fetal, and infant outcomes by reducing the likelihood of maternal and paternal smoking.
Maternal smoking during pregnancy and secondhand smoke exposure during infancy are causally associated with many adverse health outcomes. Such exposures not only leave exposed infants prone to various short- and long-term health risks but can also result in death. The SimSmoke model projected the effects of raising the MLA on the incidence of select maternal–child outcomes. Relative to the status quo, if the MLA were raised now to age 21 nationwide, modeling projects that by 2100 there would be an estimated 286,000 fewer pre-term births, 438,000 fewer cases of low birth
4 All absolute differences, including the numbers of premature deaths, lung cancer deaths, and YLL, are relative to underlying status quo projections. These status quo projections estimate decreases in smoking prevalence and thus smoking-attributable morbidity and mortality. As such, the committee encourages the reader to focus on the percentage reduction rather than on the absolute numerical estimates.
weight, and roughly 4,000 fewer sudden infant death syndrome (SIDS) cases among mothers age 15 to 49.5
The Tobacco Control Act sets a “floor” of 18 on the MLA, while allowing states and localities to raise the age. Unless Congress acts to raise the age on a national basis or delegates authority to FDA to do so, one might expect a patchwork of different MLAs in different states and localities, as existed for alcohol for many decades, rather than a uniform MLA across all of the 51 jurisdictions. The simulations described in Chapters 7 and 8 model a situation in which increases in the MLA would be adopted and implemented on a nationwide basis. In the absence of a national MLA, the public health impact of raising the MLA for tobacco would be dependent, first and foremost, upon the degree to which local and state governments take up this policy. To the extent that states choose not to raise the MLA, the effects estimated in Chapters 7 and 8 are not likely to be realized.
The strength and efficacy of existing state and local tobacco control programs vary significantly, reflecting differences in the number and intensity of tobacco control activities and the resources allocated to support them. The modeling essentially aggregates each state’s tobacco control activities, whether they are strong or weak. To the extent that policy makers in individual states want to derive state-based estimates from the findings of a national modeling exercise, they will have to take into account whether the existing levels of tobacco control activity in their states are comparable to the “average” state. If they are much weaker or stronger, extrapolation from the modeling used in this report may not be suitable.
The committee expects social sources, especially proxy purchases, to remain the primary sources of tobacco for underage persons, and it has been realistic about the high level of continuing availability to underage adolescents and young adults who are in the workforce or in college environments. Our estimates in this respect are predicated on relatively conservative assumptions. Although access to social sources could be reduced significantly if the laws prohibiting transfers to underage persons were aggressively enforced, the committee does not expect such a radical change in enforcement policy in the foreseeable future, especially under a higher MLA, because of likely public resistance. However, if a state or locality ramped up the threat of detection and punishment against social sources,
5 All absolute differences, including the numbers of cases of pre-term births, low birth weight, and SIDS, are relative to underlying status quo projections. These status quo projections predict that there will be decreases in smoking prevalence, and thus smoking-attributable morbidity and mortality.
the impact on adolescent and young adult consumption could be greater than the committee has projected.
Concerns about adolescent vulnerability to addiction and immaturity of judgment support an underage access restriction, but they do not resolve the policy question about the specific age at which the line should be drawn. The argument against raising the MLA above 18 is predicated on the assumption that adolescents older than 17 are mature enough to make their own decisions about what is in their best interests. However, evidence suggests that capacities related to mature judgment, especially in emotionally charged situations or in situations in which peer influence plays a role, are still developing into the early 20s. Many young people in their late teens and early 20s may also still be at elevated risk, developmentally speaking, to becoming addicted to nicotine. A balance needs to be struck between the personal interest of young adults in making their own choices and society’s legitimate concerns about protecting the public health and discouraging young people from making decisions they may later regret (IOM, 2007; IOM and NRC, 2004). Although some line is required, 18 is not the only developmentally plausible place to draw it. Every state sets the legal age for certain activities higher or lower for different policy purposes, and state legislators will likely continue to draw the line in different places in different policy contexts (Bonnie and Scott, 2013; Hamilton, 2010; Steinberg, 2012).
The committee assumes that the MLA will be increased for all tobacco products, including electronic nicotine delivery systems (ENDS), and that the intensity of enforcement will be the same for all products. The committee sees no reason to believe that the effects of the legal norm and its enforcement on retailer compliance, retail availability, or access to social sources would differ materially for ENDS as compared with other tobacco products. Given the evidence that adolescents who currently initiate tobacco use with ENDS rather than with conventional tobacco products are younger (Wills et al., 2014), the main effect of raising the MLA for ENDS will likely be to reduce the number of adolescents and young adults who initiate tobacco use with ENDS. However, recent trends suggest that ENDS initiation is already increasing and is likely to increase even if the MLA is raised. Increased initiation of ENDS use may reduce initiation of cigarette use because some adolescents and young adults who otherwise would have initiated cigarette users will become ENDS users instead. It may also delay initiation of cigarette use for others, including some proportion who would not have otherwise used traditional cigarettes. Presumably FDA and state policy makers will take these possibilities into account in setting the MLA and will carefully monitor the promotion and use of ENDS, especially by adolescents and young adults.
Although the full benefits of preventing initiation of tobacco use will take decades to accrue, some direct health benefits, including those from
reduced secondhand smoke exposure, will be immediate. Perhaps the greatest uncertainty in the committee’s assessment is the currently unpredictable effects of the marketing and use of ENDS and other novel tobacco products. However, in the absence of transformative changes in the tobacco market, social norms and attitudes, or the epidemiology of tobacco use, the committee is reasonably confident that raising the MLA will reduce tobacco initiation, particularly among adolescents 15 to 17 years of age, will improve health across the life span, and will save lives.
Adriani, W., V. Deroche-Gamonet, M. Le Moal, G. Laviola, and P. V. Piazza. 2006. Preexposure during or following adolescence differently affects nicotine-rewarding properties in adult rats. Psychopharmacology 184(3–4):382–390.
Arrazola, R. A., N. M. Kuiper, and S. R. Dube. 2014. Patterns of current use of tobacco products among U.S. high school students for 2000–2012—Findings from the National Youth Tobacco Survey. Journal of Adolescent Health 54(1):54–60.
Bonnie, R. J., and E. S. Scott. 2013. The teenage brain: Adolescent brain research and the law. Current Directions in Psychological Science 22(2):158–161.
Dierker, L., and R. Mermelstein. 2010. Early emerging nicotine-dependence symptoms: A signal of propensity for chronic smoking behavior in adolescents. Journal of Pediatrics 156(5):818–822.
Hamilton, V. E. 2010. Immature citizens and the state. Brigham Young University Law Review 1055(4):1055–1148.
HHS (Department of Health and Human Services). 2014. The health consequences of smoking—50 years of progress: A report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.
Holford, T. R., R. Meza, K. E. Warner, C. Meernik, J. Jeon, S. H. Moolgavkar, and D. T. Levy. 2014. Tobacco control and the reduction in smoking-related premature deaths in the United States, 1964–2012. JAMA 311(2):164–171.
IOM (Institute of Medicine). 2007. Ending the tobacco problem: A blueprint for the nation. Edited by R. J. Bonnie. Washington, DC: The National Academies Press.
IOM and NRC (Institute of Medicine and National Research Council). 2004. Reducing underage drinking: A collective responsibility. Edited by R. J. Bonnie. Washington, DC: The National Academies Press.
Jamner, L. D., C. K. Whalen, S. E. Loughlin, R. Mermelstein, J. Audrain-McGovern, S. Krishnan-Sarin, J. K. Worden, and F. M. Leslie. 2003. Tobacco use across the formative years: A road map to developmental vulnerabilities. Nicotine & Tobacco Research 5(Suppl 1):S71–S87.
Johnston, L. D., P. M. O’Malley, J. G. Bachman, J. E. Schulenberg, and R. A. Miech. 2014a. Monitoring the Future: National survey results on drug use, 1975–2013: Volume 1, Secondary school students. Ann Arbor: Institute for Social Research, University of Michigan.
Johnston, L. D., P. M. O’Malley, R. A. Miech, J. G. Bachman, and J. E. Schulenberg. 2014b. Monitoring the Future: National survey results on drug use, 1975–2013: Overview, key findings on adolescent drug use. Ann Arbor: Institute for Social Research, University of Michigan.
Ling, P. M., and S. A. Glantz. 2002. Why and how the tobacco industry sells cigarettes to young adults: Evidence from industry documents. American Journal of Public Health 92(6):908–916.
SAMHSA (Substance Abuse and Mental Health Services Administration). 2013. Results from the 2012 National Survey on Drug Use and Health: Summary of national findings. Rockville, MD: Substance Abuse and Mental Health Services Administration; Center for Behavioral Statistics and Quality; Department of Population Surveys.
Schiller, J. S., B. W. Ward, and G. Freeman. 2014. Early release of selected estimates based on data from the 2013 National Health Interview Survey. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention and National Center for Health Statistics. http://www.cdc.gov/nchs/data/nhis/earlyrelease/earlyrelease201406.pdf (accessed February 22, 2015).
Steinberg, L. 2012. Should the science of adolescent brain development inform public policy? Issues in Science and Technology 28(3):67–78.
Surgeon General’s Advisory Committee on Smoking and Health. 1964. Smoking and health: Report of the Advisory Committee to the Surgeon General of the Public Health Service. Washington, DC: U.S. Government Printing Office.
Wills, T. A., R. Knight, R. J. Williams, I. Pagano, and J. D. Sargent. 2014. Risk factors for exclusive e-cigarette use and dual e-cigarette use and tobacco use in adolescents. Pediatrics 135(1):e43–e51.