Article Text
PDF
Statistics from Altmetric.com
Introduction
Reductions in adolescent cigarette use have been offset by increased use of alternative tobacco products.1 2 In 2017, 6.4% of adolescents in Massachusetts reported current cigarette use, while 20.1% reported electronic vapour product use (referred to as e-cigarettes).2 Furthermore, nationwide, over two-thirds of adolescent e-cigarette users in the USA report using flavoured products.3 While the Family Smoking Prevention and Tobacco Control Act (Tobacco Control Act) in 2009 afforded the Food and Drug Administration the authority to regulate tobacco products, it did not pre-empt state or municipal governments to enact stricter policies than federal laws seeking to stem tobacco use.4 There has since been an increasing enactment of state and local laws, such as cigarette excise taxes and smoke-free legislation.5
New, progressive local policies seek to curb adolescent tobacco use by reducing access and exposure. Flavoured tobacco product restrictions limit sales of flavoured tobacco products other than cigarettes to specific retail locations or flavours, excluding menthol,6 which provide stricter coverage than the Tobacco Control Act. Tobacco 21 policies prohibit the sale of all tobacco products to persons under age 21. Such policies have spread since Needham, Massachusetts became the first municipality to introduce a tobacco 21 policy in 2005,7 culminating in the passage of federal tobacco 21 legislation in 2019.8 Smoke-free legislation prohibiting e-cigarettes expand smoke-free laws in restaurants and workplaces that originated before e-cigarettes were marketed.
There have been few evaluations of emerging local laws on adolescent tobacco use.9–11 Using data from representative cross sections of Massachusetts high school students from 2011 to 2017, we exploited this variability to evaluate three such laws.12 Our aims were to examine the associations between local flavoured tobacco product restrictions, tobacco 21 policies and smoke-free laws prohibiting e-cigarettes with adolescent cigarette and e-cigarette use, and to assess whether policy effects varied by age.
Methods
Participants
We acquired data from the Massachusetts Youth Health Survey (YHS)—high school, which assesses the health and well-being of 9–12 grade students in public high schools.13 The anonymous survey is administered in randomly selected high schools in odd-numbered years. From 2011-2017, 72-81% of sampled high schools agreed to participate with an overall student response of 62-69%.14
We analysed four sweeps of the YHS as questions on e-cigarettes were first introduced in 2011. Of the 10 668 adolescents who completed the 2011–2017 high school YHS surveys, we excluded those with missing information on race/ethnicity, sex, age, or tobacco use, and students younger than age 14. Final analytic samples included 9988 adolescents for cigarette use and 10 168 for e-cigarette use.
Tobacco use
Adolescents reported lifetime and prior month use of cigarettes, which we combined into a count variable of days smoked in the past month (0–30). Adolescents also reported use of a variety of tobacco products in the past month: those who endorsed ‘electronic cigarettes or e-cigarettes, such as Ruyan or NJOY’ (2011–2015 surveys) or ‘used an electronic vapour product’ (2017 survey) were coded as currently using e-cigarettes (yes/no).
Respondent characteristics
Adolescents reported their age (years), sex (male, female), race and ethnicity (White, Black, Hispanic, Asian, Multiracial/Native American). The YHS provided the county of each adolescents’ school, the smallest geographic identifier available, collapsing the smallest counties.
Tobacco-control policies
We generated three county-level tobacco-control policy measures in each year of the survey. To construct the proportion of each county exposed to a policy (0–1), we divided the total number of residents in each covered municipality within the county by the total county population based on the 2010 Census.15 Flavoured tobacco product restrictions limiting the sale of flavoured tobacco products other than cigarettes to adult-only retail tobacco stores16 were acquired from the Massachusetts Association of Health Boards’ tobacco maps.17 Since only the year of implementation was available, the policy was considered effective in the subsequent sweep of the YHS. Tobacco 21 policies prohibiting the sale of all tobacco products to persons under the age of 21 years were acquired from the Preventing Tobacco Addiction Foundation.18 Smoke-free legislation prohibiting e-cigarettes in 100% smoke-free restaurants was acquired from the American Nonsmokers’ Rights Foundation.19 As the YHS was administered in the spring, the proportion of tobacco 21 and smoke-free coverage was based on whether legislation was effective by March. We linked year-specific county-level coverage for each policy to each adolescent.
Analysis
We conducted difference-in-differences models20 to link changes in tobacco-control policies to changes in prior month tobacco use within and between counties, using zero-inflated negative binomial regression to assess days of cigarette use and logistic regression to assess use of e-cigarettes. Models included the three policy variables, respondent characteristics, and county and year fixed effects. We then added interactions between each policy and adolescent age, evaluated by an adjusted Wald test, to assess whether policy effects varied across age groups. We conducted analyses using svy commands in Stata V.15.1 to accommodate the complex survey design, in which students were clustered within schools, and weights to account for school-level non-response, student non-response and poststratification adjustment.14
Results
Adolescents averaged 16.1 years of age; 49% were female; 67% White, 15% Hispanic, 9% Black and 5% Asian. The prevalence of prior month cigarette use decreased from 13.0% to 6.6% from 2011 to 2017, while e-cigarette use increased from 3.9% to 19.0%. During this period, there was increasing coverage of all three tobacco-control policies across Massachusetts counties. Flavoured tobacco product restrictions and smoke-free laws prohibiting e-cigarettes increased from 0% in 2011 to an average of 35.9% and 57.9%, respectively, in 2017. Tobacco 21 policies increased from 0.6% of coverage in 2011 to 55.5% by 2017.
Results from difference-in-differences models (table 1) show that increasing implementation of flavoured tobacco product restrictions was associated not with a reduction in the likelihood of cigarette use, but with a decrease in the level of cigarette use among users (Coefficient −1.56; 95% CI −2.54 to −0.58). A significant interaction (p=0.03) revealed reductions in levels of use among 14 and 18 year olds. In contrast, we found no significant overall associations between tobacco 21 policies or smoke-free laws prohibiting e-cigarettes with cigarette use. However, a significant age interaction with tobacco 21 policies (p=0.01) suggests a reduction in cigarette use among 18 year olds, which was the only subgroup directly targeted by this policy shift.
Effects from difference-in-differences models of tobacco-control policies and current cigarette and e-cigarette use
Flavoured tobacco product restrictions also were associated with a reduction in adolescent e-cigarette use (Coefficient −0.87; 95% CI −1.68 to −0.06) (table 1). There were no significant associations between tobacco 21 policies or smoke-free laws prohibiting e-cigarettes with e-cigarette use, nor were there significant age interactions in relation to e-cigarette use (all p>0.05).
The correlations between the three policies ranged from 0.58 to 0.79. We re-estimated models with one policy at a time across both outcomes and findings were robust (results not shown).
Discussion
We found that the implementation of flavoured tobacco product restrictions in Massachusetts was associated with reductions in current cigarette and e-cigarette use, with the largest decreases among 14 and 18 year olds. Massachusetts became the first state to prohibit the sale of all flavoured tobacco products, including flavoured e-cigarettes, in November 2019.6 Our findings suggest that local legislation can reduce adolescent tobacco use and municipalities should enact stricter tobacco-control policies when not pre-empted by state law.4
Among adolescent users of tobacco products, over 70% report using flavoured products.3 Since the Tobacco Control Act only banned flavoured cigarettes, municipalities have restricted access to other flavoured tobacco products by limiting points of sale or flavours.6 Local bans have driven reductions in flavoured tobacco product availability and sales.9–11 21 22 Evaluations of such restrictions have found declines in adolescents ever trying flavoured tobacco products and use of any tobacco products9 11 as well as decreases in adolescent use of e-cigarettes.10 Our results corroborate these findings, and expand their generalisability by examining the introduction of these policies in municipalities across Massachusetts over a period of active policy change.
In contrast, we found limited evidence that tobacco 21 policies or smoke-free laws prohibiting e-cigarettes were associated with adolescent cigarette or e-cigarette use. The first tobacco 21 law in Needham, Massachusetts in 2005 decreased adolescent cigarette use,23 a pattern replicated in larger, more representative studies.24 25 However, a New York City evaluation found a more muted effect.26 Our results showing effects only on 18-year-olds provides evidence that the policy was effective for reducing use among its primary target age. E-cigarette vapour has been shown to contain lower nicotine and carcinogens than cigarette smoke, but less is known about health risks.27 Municipalities and states have been enacting 100% smoke-free laws that include e-cigarettes to protect non-users,19 but there are limited evaluations of their impact on tobacco use. Agaku et al found that adolescents in states with smoke-free laws prohibiting cigarettes and e-cigarettes perceived e-cigarettes as more harmful than youth in states without such laws.28 More research is needed on effects of local-level tobacco-control policies on attitudes, access, and use of other products and across youth of various ages, as well as whether suites of policies12 are needed to create sustained change.
There are a number of limitations to note. Although we included county fixed effects in our models to control for local unmeasured factors and year fixed effects to account for state and federal policy changes, there may have been other tobacco-control activities that occurred over the study period that we did not capture. The YHS samples adolescents attending public high schools, excluding youth attending private or religious schools or not attending school. The survey is cross-sectional, and our analyses cannot deduce causal effects. The lowest geographical identifier in the dataset is county of each school, and municipal-level policies were aggregated to the county level. As such, we could not assess whether policies were in place within the actual municipality of each student’s residence. This imprecision suggests that our findings may underestimate the true effect of tobacco restrictions on adolescent tobacco use. The YHS question on e-cigarettes asks about any use, not the level of use, and does not include current products commonly used by adolescents, such as Juuls, and thus may underestimate use.
In sum, we found that adolescents decreased their use of cigarettes and e-cigarettes in response to the local implementation of flavoured tobacco product restrictions, while 18 year olds decreased their cigarette use in response to tobacco 21 restrictions. Our results provide further support for evidence accumulating across US municipalities of local policies reducing adolescent access and exposure to tobacco products,9–11 as well as the importance of comprehensive, multilevel efforts to decrease tobacco use. States that pre-empt local communities from enacting stricter tobacco-control policies are missing opportunities to reduce adolescent tobacco use.
What this paper adds
Many US states, including Massachusetts, do not pre-empt municipalities from enacting stricter tobacco-control policies than state or federal laws. Although new, progressive local policies seek to curb adolescent tobacco use by reducing access and exposure, there have been few evaluations of emerging local laws on adolescent tobacco use.
We found that adolescents in Massachusetts counties with increasing implementation of flavoured tobacco product restrictions reported reductions in current e-cigarette use and levels of cigarette use, while only 18 year olds decreased cigarette use in response to tobacco 21 implementation. In contrast, we found no evidence of associations between smoke-free laws prohibiting e-cigarettes with use of either tobacco product.
States that pre-empt local communities from enacting stricter tobacco-control policies, particularly flavoured tobacco product restrictions, are missing opportunities to reduce adolescent tobacco use.
Ethics statements
Patient consent for publication
References
Footnotes
Correction notice In the 'Participants' section of the 'Methods', final sentence has been updated to 'From 2011-2017, 72-81% of sampled high schools agreed to participate with an overall student response of 62-69%'.
Contributors SSH conceptualised and designed the study, participated in data collection and interpretation of the findings, and drafted the initial manuscript. CK and MO'B participated in data collection, analysis, and reviewed and revised the manuscript. RLC conceptualised and designed the study, participated in the interpretation of the findings, and reviewed and revised the manuscript. All authors approved the final manuscript as submitted.
Funding This work was supported in part by a grant from the American Lung Association (PP- 625245) (PI: S.S.H.).
Disclaimer The funder had no role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; and in the decision to submit the article for publication.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.