Abstract
Purpose
Youth who engage with online tobacco marketing may be more susceptible to tobacco use than unengaged youth. This study examines online engagement with tobacco marketing and its association with tobacco use patterns.
Methods
Cross-sectional analysis of youth aged 12–17 years who participated in Wave 1 of the Population Assessment of Tobacco and Health (PATH) Study (N=13,651). Engagement with tobacco marketing was based on ten survey items including signing up for email alerts about tobacco products in the past six months. Logistic regression was used to examine association of online engagement with tobacco marketing and susceptibility to use of any tobacco product among never-tobacco users, ever having tried tobacco, and past 30-day tobacco use.
Results
An estimated 2.94 million US youth (12%) engaged with ≥ 1 forms of online tobacco marketing. Compared to no engagement, the odds of susceptibility to the use of any tobacco product among never-tobacco users was independently associated with the level of online engagement: adjusted odds ratio (AOR) =1.48 (95% confidence interval [CI] 1.24–1.76) for 1 form of engagement and AOR=2.37 (95% CI 1.53–3.68) for ≥2 forms of engagement. The odds of ever having tried tobacco were also independently associated with level of online engagement: AOR=1.33 (95% CI 1.11–1.60) for 1 form of engagement and AOR=1.54 (95% CI 1.16–2.03) for ≥2 forms of engagement. The level of online engagement was not independently associated with past 30-day tobacco use.
Conclusion
Online engagement with tobacco marketing may represent an important risk factor for onset of tobacco use in youth.
INTRODUCTION
Tobacco advertising expenditure on the Internet including tobacco company websites grew more than thirty-fold, from $0.7 million dollars in 1999 to $23.1 million dollars in 2013 [1,2]. In addition to marketing on tobacco company websites, tobacco brand and product promotions abound on social media platforms such as Facebook, Twitter, and YouTube [3–6]. Online advertising affords new opportunities to reach potential and current tobacco users and to offer product discounts in a largely unregulated environment [7–9]. Online marketing may be even more effective than traditional marketing in promoting tobacco use among youth because it provides consumers greater opportunities for engagement and interaction with pro-tobacco content [10–12].
We do not yet know the extent to which youth—both those who currently use tobacco and those who have never used tobacco—engage with online tobacco marketing. Current youth tobacco users may seek online venues to purchase tobacco products and bypass age-verification measures [13]. Youth who have never used tobacco may engage with tobacco marketing while online, and this engagement may increase their susceptibility to tobacco use that may lead to experimentation with tobacco products. Although public education efforts aim to disrupt these attitudinal changes among youth, such efforts may be less effective against online forms of marketing. Additionally, the voluntary MSA between tobacco companies and state governments that restricted tobacco product marketing was developed for traditional products (mainly cigarettes) and traditional media channels (e.g., print media), and was implemented well before the proliferation of online marketing. Thus, quantifying the scope of youth exposure to online marketing and its relation with tobacco use intention and behavior can provide evidence for the development and implementation of future regulations [14].
This analysis examines this research gap with data from a large, nationally representative population-based study that assesses online engagement and use of multiple tobacco products. It is hypothesized that, among youth, greater levels of online engagement will be associated with greater susceptibility to tobacco product use among never-tobacco users and higher likelihood of ever having tried tobacco, past 30-day use of tobacco, after accounting for socio-demographic and behavioral risk factors for tobacco use and exposure to marketing in traditional venues.
METHODS
Data
Data are from Wave 1 of the Population Assessment of Tobacco and Health (PATH) Study conducted from September 12, 2013 to December 15, 2014 [15]. The PATH Study is a nationally-representative, longitudinal cohort study of 45,971 adults and youth in the US, ages 12 years and older. The National Institutes of Health, through the National Institute on Drug Abuse, is partnering with the Food and Drug Administration’s Center for Tobacco Products to conduct the PATH Study under a contract with Westat. The PATH Study used Audio-Computer Assisted Self-Interviews (ACASI) available in English and Spanish to collect information on tobacco-use patterns and associated health behaviors. This analysis draws from the 13,651 Youth Interviews (all participants were ages 12 to 17 years). Parent Interviews (n = 13,589) were conducted with one parent of nearly every youth participant. Recruitment employed address-based, area-probability sampling, using an in-person household screener to select youths and adults. Adult tobacco users, young adults ages 18 to 24 and African Americans were oversampled relative to population proportions. The weighting procedures adjusted for oversampling and nonresponse; combined with the use of a probability sample, the weighted data allow the estimates produced by the PATH Study to be representative of the non-institutionalized, civilian US population. The weighted response rate for the household screener was 54.0%. Among households that were screened, the overall weighted response rate was 78.4% for the Youth Interview. Further details regarding the PATH Study design and methods are published by Hyland et al [16]. and on the PATH Study’s website [17]. Westat’s Institutional Review Board approved the study design and protocol and the Office of Management and Budget approved the data collection.
Missing data on age, sex, race, Hispanic ethnicity were logically assigned from household screener data, as described in the PATH Study Restricted Use File User Guide [17].
Outcomes
Three tobacco-related outcomes were examined: 1) susceptibility to tobacco use among never-tobacco users, 2) ever having tried any tobacco product among all respondents, and 3) past 30-day tobacco use among ever tobacco users. Products of interest included: cigarettes, electronic cigarettes (e-cigarettes), cigars (traditional, cigarillos, and filtered), pipes, hookah (water pipe), snus pouches, other smokeless tobacco, dissolvable tobacco, bidis, and kreteks. First, never-tobacco users were considered susceptible to tobacco use if they responded “definitely yes”, “probably yes”, or “probably no” to one of the following questions for one or more tobacco products: 1) “If one of your friends offered you a (cigarette, e-cigarette, etc.), would you try it?”, 2) “Do you think you will smoke a (cigarette, e-cigarette, etc.) sometime in the next year?”, and 3) “Have you ever been curious about smoking/using a (cigarette, e-cigarette, etc.)?” [18,19]. Second, respondents were considered to have ever tried tobacco if they responded affirmatively to queries on ever use of one or more tobacco products (e.g., “Have you ever tried cigarette smoking, even one or two puffs?”). Finally, respondents were considered to be past 30-day tobacco users if they responded ‘earlier today’, ‘not today but sometime during the past 7 days’, or ‘not during the past 7 days but sometime during the past 30 days’ to use of one or more tobacco products within the past 30 days (e.g., “When was the last time you smoked a cigarette, even one or two puffs?”).
Online Engagement and Covariates
The primary variable of interest was the level of online engagement with tobacco marketing, which equaled the sum of affirmative answers to 10 items that assessed online engagement (Table 2 and Appendix Table 1; e.g., “In the past 6 months, have you ever signed up for email alerts about tobacco products, read articles online about tobacco products, or watched a video online about tobacco products?”). The level of online engagement was categorized into sum scores: 0 items, 1 item, and 2 or more items. Online engagement scores of 2 or higher were collapsed because only 0.8% of respondents reported engagement with three or more items.
Table 2.
Prevalence of Engagement to Online Tobacco Marketing
| Form of Engagement1 | Count | Wgt. Prev, % (95% CI) |
|---|---|---|
| Signed up for e-mail alerts about tobacco products, past 6 months | 605 | 4.6 (4.2, 5.0) |
|
| ||
| Used a smart phone to scan a QR code for a tobacco product | 395 | 2.9 (2.6, 3.2) |
|
| ||
| Visited at least one tobacco brand website | 326 | 2.3 (2.1, 2.5) |
|
| ||
| Received a discount coupon for any tobacco product online | 321 | 2.2 (2.1, 2.4) |
|
| ||
| Liked or followed at least one tobacco brand on social networking site | 218 | 1.5 (1.3, 1.7) |
|
| ||
| Played online game related to a tobacco brand2 | 161 | 1.1 (0.9, 1.4) |
|
| ||
| Sent link about a tobacco brand on social networking site | 126 | 0.8 (0.7, 1.0) |
|
| ||
| Received information from tobacco companies online | 110 | 0.8 (0.6, 0.9) |
|
| ||
| Scanned a QR code for a tobacco product that took respondent to a tobacco company website | 46 | 0.3 (0.2, 0.5) |
|
| ||
| Registered on at least one tobacco brand website | 38 | 0.2 (0.2, 0.3) |
|
| ||
| Online Engagement Score 0 | 12,024 | 88.2 (87.5, 88.8) |
| Online Engagement Score 1 | 1,218 | 8.9 (8.3, 9.6) |
| Online Engagement Score 2 or More | 409 | 2.9 (2.5, 3.3) |
See Appendix Table 1 for exact wording of survey questions
Brands include Marlboro, Newport, Camel, American Spirit, and Copenhagen
Receptivity to tobacco marketing was also assessed through traditional media channels. Commercial vendors collected a pool of ad images during the year prior to the Wave 1 PATH Study from television, magazine, and newspaper ads, as well as mailer campaigns. Each respondent was shown 20 randomly selected images of tobacco ads from a pool of 689 images (or 459 images if respondent interviewed after January 2014), stratified by four product types: cigarettes, e-cigarettes, cigars, and smokeless tobacco. Respondents were asked if they recalled seeing each tobacco ad and if so, whether they liked it. They were also asked if they had a favorite tobacco ad. Following the approach of Pierce et al. [20], who operationally defined receptivity as first remembering and then demonstrating an affective response to marketing, receptivity to tobacco marketing was categorized into four categories according to the level of response: none (saw 0 ads and liked 0 ads and had 0 favorite ads), low (saw ≥ 1 ad and liked 0 ads and had 0 favorite ads), moderate (saw ≥1 ad and either liked ≥1 ad or had ≥1 favorite ads, but not both), and high (saw ≥1 ad and liked ≥1 ad and had ≥1 favorite ads).
Other covariates included age, sex, race/ethnicity, and parental education level. Mental health status was assessed by internalizing and externalizing problems based on the Global Appraisal of Individual Needs-Short Screener [21]. The level of internalizing problems was categorized based on the sum score of 4 items: 0–1 items (low), 2–3 items (moderate), and 4 items (high). Similarly, the level of externalizing problems was categorized based on the sum score of 5 items: 0–1 (low), 2–3 items (moderate), and 4–5 items (high). Sensation seeking was assessed as the mean of three items modified from the Brief Sensation Seeking Scale measured on a 5-point Likert scale (e.g., “I like to do frightening things”) and then categorized into terciles [22,23]. See Appendix Table 2 for a list of survey items for the level of internalizing problems, externalizing problems, and sensation seeking.
Respondents’ overall use of the internet and social networking behavior including frequency of internet use, frequency of social networking account use, and whether respondents used a smart phone were also assessed. Additional covariates included parent report of youths’ school performance during interview with parent; youths’ weekly income from a job, family, or allowance; and any reported exposure to smoking in the youths’ home, in a car, at school, or outdoors. Finally, other substance use was assessed including past 30-day binge alcohol drinking (≥5 alcoholic drinks in a day for males and ≥4 alcoholic drinks in a day for females), past-year marijuana use, and past-year non-prescription drug use (e.g., cocaine and unprescribed methylphenidate [Ritalin™]). The level of other substance use was categorized into 0, 1, and 2 or more other substances.
Analyses
Throughout the analyses, the balanced repeated replication weights were utilized with Fay’s correction (shrinkage factor set at 0.3). First, the prevalence of forms of engagement with online tobacco marketing was estimated. Confidence intervals were reported using the incomplete beta function [24]. Second, the weighted prevalence of each of the tobacco-related outcomes by the level of online engagement to tobacco marketing was assessed by Pearson’s chi-squared test statistic with second-order Rao-Scott corrections [25].
Third, a multivariable weighted logistic regression model was fit among the N=10,246 respondents who never used tobacco with susceptibility to tobacco as the dependent variable and level of online engagement as the primary variable of interest, while controlling for other covariates (described above). Next, similarly specified multivariable weighted logistic regression models were fit for the following outcomes: 1) having ever tried tobacco (among N=13,115 respondents with known tobacco use status) and 2) past 30-day tobacco use (among N=2869 ever users). For the regression models, multiple imputation was performed to account for missing data in the ten individual items comprising the level of online engagement and other covariates (e.g., parental education). The multiple imputation method utilized assumed that missing data were missing at random [26]. The percentage of missing data for the ten individual items comprising the level of online engagement and other covariates ranged from 0.0% to 4.0%; overall 1508 records (11.5%) contained missing information on at least one of these variables. Five multiply-imputed data sets were generated, the weighted logistic regression models described above were fit, and the parameter estimates accounting for imputation uncertainty were combined.
RESULTS
Study Population
Socio-demographic characteristics of the PATH Study youth sample are described in Table 1. The sample was nearly equally split among 12–14 and 15–17 year olds (51.3% versus 48.7%, respectively). Respondents were 47.5% non-Hispanic White, 28.4% Hispanic, 13.2% non-Hispanic Black and 51.2% male. A majority of respondents accessed the internet several times a day (61.2%), used their social networking account at least daily (32.7% several times a day and 27.3% once per day), and used a smart phone (69.6%).
Table 1.
Characteristics of Wave 1 PATH Study Youth Sample, Overall and by Tobacco Outcome
| N | Unweighted. Prev., % | Weighted Prev., % | Susceptible to Any Tobacco Use Among Never-Tobacco Users 2, Weighted Prev., % (95% CI) | Ever Tried Tobacco Among All Respondents3, Weighted Prev., % (95% CI) | Past 30-Day Tobacco Use Among Ever-Tobacco Users4, Weighted Prev., % (95% CI) | |
|---|---|---|---|---|---|---|
| Total | 13,651 | 100 | 100 | 43.8 (42.8, 44.8) | 21.8 (21.1, 22.6) | 42.5 (40.6, 44.5) |
|
| ||||||
| Age Group (Yrs.) | ||||||
| 12–14 | 6,998 | 51.3 | 50.4 | 38.4 (37.1, 39.7) | 10.7 (9.9, 11.5) | 28.7 (25.1, 32.3) |
| 15–17 | 6,653 | 48.7 | 49.6 | 51.1 (49.5, 52.6) | 32.8 (31.6, 34.0) | 46.8 (44.5, 49.0) |
|
| ||||||
| Sex | ||||||
| Female | 6,657 | 48.8 | 48.7 | 44.1 (42.7, 45.6) | 20.3 (19.2, 21.3) | 40.5 (37.6, 43.4) |
| Male | 6,994 | 51.2 | 51.3 | 43.5 (42.1, 44.9) | 23.3 (22.2, 24.4) | 44.3 (41.6, 47.0) |
|
| ||||||
| Race/Ethnicity | ||||||
| Non-Hispanic White | 6,478 | 47.5 | 53.5 | 41.6 (40.2, 43.0) | 23.4 (22.3, 24.5) | 45.5 (42.8, 48.2) |
| Hispanic | 3,880 | 28.4 | 22.0 | 48.4 (45.7, 51.1) | 18.8 (16.8, 20.8) | 42.1 (36.0, 48.2) |
| Non-Hispanic Black | 1,801 | 13.2 | 13.4 | 46.4 (32.9, 59.9) | 21.5 (11.1, 32.0) | 53.6 (25.8, 81.4) |
| Multiple races | 767 | 5.6 | 4.2 | 37.4 (32.1, 42.7) | 8.2 (5.4, 11.0) | 35.4 (18.3, 52.5) |
| Asian/Pacific Islander | 394 | 2.9 | 4.6 | 50.6 (46.2, 55.1) | 30.5 (27.0, 34.0) | 37.6 (30.8, 44.3) |
| American Indian or Alaska Native | 70 | 1.9 | 0.4 | 41.1 (34.3, 47.9) | 15.2 (10.3, 20.0) | 40.2 (21.9, 58.4) |
|
| ||||||
| Parental Education | ||||||
| At Least Some College | 8,148 | 60.1 | 63.9 | 44.3 (43.0, 45.6) | 19.8 (18.9, 20.8) | 42.2 (39.5, 44.8) |
| High School Graduate | 2,570 | 19.0 | 18.1 | 42.2 (39.9, 44.6) | 25.7 (23.9, 27.5) | 44.7 (40.5, 48.9) |
| Less than High School | 2,834 | 20.9 | 18.0 | 43.5 (41.3, 45.7) | 24.8 (23.1, 26.6) | 41.5 (37.3, 45.6) |
|
| ||||||
| Weekly Income | ||||||
| None | 4,432 | 32.9 | 33.3 | 38.2 (36.6, 39.9) | 15.2 (14.0, 16.3) | 33.3 (29.3, 37.2) |
| $1–$20 | 6,196 | 46.0 | 45.3 | 45.0 (43.6, 46.5) | 18.5 (17.5, 19.6) | 42.0 (38.8, 45.1) |
| $21–$50 | 1,406 | 10.4 | 10.5 | 54.0 (50.7, 57.4) | 33.8 (31.2, 36.5) | 43.9 (39.0, 48.9) |
| $51+ | 1,429 | 10.6 | 10.9 | 51.1 (47.5, 54.7) | 43.9 (41.1, 46.6) | 52.1 (47.8, 56.4) |
|
| ||||||
| School Performance | ||||||
| Mostly As | 3,358 | 24.9 | 26.7 | 37.2 (35.4, 39.0) | 10.7 (9.6, 11.8) | 33.1 (27.8, 38.4) |
| As and Bs | 4,639 | 34.4 | 34.1 | 44.2 (42.6, 45.9) | 19.0 (17.7, 20.2) | 38.0 (34.4, 41.6) |
| Mostly Bs | 1,184 | 8.8 | 8.8 | 46.3 (42.8, 49.7) | 23.0 (20.4, 25.6) | 42.6 (35.9, 49.3) |
| Bs and Cs | 2,591 | 19.2 | 18.4 | 49.4 (47.0, 51.8) | 29.6 (27.7, 31.5) | 43.8 (39.9, 47.8) |
| Mostly Cs to Mostly Fs | 1,719 | 12.7 | 12.0 | 50.5 (47.3, 53.7) | 41.0 (38.5, 43.5) | 52.3 (48.1, 56.4) |
|
| ||||||
| Internet Access | ||||||
| Several Times/Day | 8,325 | 61.2 | 62.2 | 48.8 (47.5, 50.1) | 23.9 (22.9, 24.9) | 40.9 (38.5, 43.3) |
| Once/Day | 2,005 | 14.7 | 15.0 | 35.1 (32.6, 37.5) | 17.1 (15.4, 18.9) | 40.3 (34.6, 46.1) |
| 3–5 Days/Week | 1,308 | 9.6 | 9.3 | 40.3 (37.2, 43.4) | 21.3 (18.8, 23.7) | 49.4 (42.8, 55.9) |
| 1–2 Days/Week | 551 | 4.1 | 3.9 | 30.8 (26.5, 35.2) | 13.0 (10.1, 16.0) | 41.4 (29.3, 53.5) |
| Less Than Once/Week | 1,415 | 10.4 | 9.6 | 36.5 (33.6, 39.5) | 19.4 (17.1, 21.7) | 51.5 (44.4, 58.5) |
|
| ||||||
| Social Networking Account Use | ||||||
| Several Times A Day | 4,452 | 32.7 | 33.1 | 53.5 (51.7, 55.4) | 30.3 (28.8, 31.7) | 46.0 (43.1, 48.9) |
| Daily | 3,725 | 27.3 | 27.3 | 46.5 (44.6, 48.4) | 21.9 (20.5, 23.3) | 40.0 (36.4, 43.7) |
| Weekly | 1,593 | 11.7 | 11.7 | 43.6 (40.8, 46.4) | 18.0 (16.0, 20.0) | 39.4 (33.2, 45.6) |
| Monthly Or Less Often | 1,143 | 8.4 | 8.4 | 36.4 (33.2, 39.7) | 13.9 (11.7, 16.1) | 28.7 (20.9, 36.4) |
| No Social Networking Account | 2,709 | 19.9 | 19.6 | 30.6 (28.6, 32.5) | 12.8 (11.4, 14.1) | 43.4 (37.4, 49.5) |
|
| ||||||
| Use Smart Phone | ||||||
| Yes | 9,470 | 69.6 | 69.8 | 48.0 (46.7, 49.2) | 25.0 (24.1, 25.9) | 43.0 (40.8, 45.2) |
| No | 4,145 | 30.4 | 30.2 | 35.5 (33.8, 37.2) | 14.3 (13.1, 15.4) | 40.6 (36.1, 45.1) |
|
| ||||||
| Sensation Seeking1 | ||||||
| Low | 4,563 | 33.5 | 33.3 | 25.5 (24.1, 26.9) | 11.9 (10.9, 12.9) | 35.7 (31.1, 40.2) |
| Moderate | 5,228 | 38.3 | 38.4 | 48.0 (46.4, 49.6) | 18.6 (17.5, 19.7) | 38.0 (34.6, 41.4) |
| High | 3,845 | 28.2 | 28.2 | 66.7 (64.7, 68.7) | 37.5 (35.9, 39.2) | 47.8 (45.0, 50.7) |
|
| ||||||
| Internalizing Problems | ||||||
| Low | 6,379 | 47.9 | 48.1 | 32.8 (31.5, 34.2) | 16.0 (15.1, 17.0) | 40.9 (37.5, 44.3) |
| Moderate | 3,890 | 29.2 | 29.0 | 51.2 (49.3, 53.0) | 21.3 (19.9, 22.6) | 40.9 (37.2, 44.5) |
| High | 3,043 | 22.9 | 22.9 | 62.5 (60.2, 64.7) | 34.4 (32.6, 36.2) | 44.9 (41.6, 48.1) |
|
| ||||||
| Externalizing Problems | ||||||
| Low | 6,602 | 50.0 | 49.7 | 32.5 (31.2, 33.8) | 15.4 (14.4, 16.3) | 40.8 (37.5, 44.2) |
| Moderate | 5,351 | 40.5 | 41.0 | 55.2 (53.6, 56.9) | 24.9 (23.6, 26.1) | 40.6 (37.7, 43.5) |
| High | 1,245 | 9.4 | 9.3 | 67.9 (64.2, 71.5) | 43.3 (40.4, 46.3) | 50.4 (45.8, 55.0) |
|
| ||||||
| Exposure to Smoking | ||||||
| No | 7,841 | 59.1 | 59.6 | 38.6 (37.4, 39.8) | 12.0 (11.2, 12.8) | 27.6 (24.4, 30.8) |
| Yes | 5,420 | 40.9 | 40.4 | 55.2 (53.4, 56.9) | 36.1 (34.7, 37.5) | 49.5 (47.1, 51.9) |
|
| ||||||
| Other Substance Use5 | ||||||
| 0 | 11,891 | 89.2 | 89.2 | 42.6 (41.6, 43.6) | 13.9 (13.2, 14.6) | 28.7 (26.3, 31.2) |
| 1 | 1,194 | 9.0 | 8.9 | 84.2 (79.5, 89.0) | 81.2 (78.8, 83.6) | 54.6 (51.2, 58.0) |
| 2 or More | 248 | 1.9 | 1.9 | 96.3 (79.0, 99.5) | 93.6 (90.5, 96.7) | 79.0 (73.5, 84.5) |
Terciles of mean sensation seeking score not evenly distributed because of heaping in the score.
N=10,246
N=13,115
N=2869
Other substance use includes past 30-day binge alcohol drinking, past-year marijuana use, and past-year non-prescription drug use.
Note: Prev= Prevalence; CI=Confidence Interval. Tobacco products included: cigarettes, electronic cigarettes (e-cigarettes), cigars (traditional, cigarillos, and filtered), pipes, hookah (water pipe), snus pouches, other smokeless tobacco, dissolvable tobacco, bidis, and kreteks.
Prevalence of Engagement with Online Tobacco Marketing
Among US youth, 88.2% had not engaged with any form of online tobacco marketing, 8.9% had engaged with one form, and 2.9% had engaged 2 or more forms (Table 2). Common forms of online engagement included: signing up for email alerts about tobacco products in the past 6 months (4.6%), using a smart phone to scan a quick response (QR) code for a tobacco product (2.9%), visiting at least one tobacco website (2.3%), receiving a discount coupon for any tobacco product online (2.2%), and liking or following a tobacco brand on a social networking site (1.5%). Although the total percent of youth engaged in online tobacco marketing was relatively small, it represents approximately 2.9 million US youth who reported some interaction with online tobacco marketing.
Prevalence of Tobacco Use and Its Association with Online Engagement
Among US youth, 43.8% of never-tobacco users were susceptible to at least one tobacco product use. Additionally, 21.8% of all youth had ever tried a tobacco product and 42.5% of ever-tobacco users were past 30-day tobacco product users (Table 3). More than one in every two respondents who had ever tried any tobacco product had tried multiple (≥2) tobacco products (56.4%=12.3% / 21.8%).
Table 3.
Prevalence of Susceptibility, Ever Having Tried Tobacco, and Past 30-day Tobacco Use by Level of Engagement to Online Tobacco Marketing1 (Weighted Percent and 95% Confidence Interval)
| Outcome | Overall | Online Engagement Score 0 | Online Engagement Score 1 | Online Engagement Score 2 or More |
|---|---|---|---|---|
| Susceptible to Use of Any Tobacco Product Among Never-Tobacco Users2 | 43.8 (42.8,44.8) | 41.7 (40.6,42.7) | 60.7 (57.2,64.2) | 79.5 (73.5,85.5) |
|
| ||||
| Ever Tried Any Tobacco Product Among All Respondents3 | 21.8 (21.1,22.6) | 19.7 (19.0,20.5) | 31.8 (29.0,34.6) | 54.1 (48.8,59.3) |
| Cigarette | 13.9 (13.3,14.6) | 12.5 (11.8,13.1) | 19.9 (17.5,22.4) | 39.2 (34.1,44.3) |
| E-Cigarette | 11.1 (10.5,11.7) | 9.7 (9.1,10.3) | 18.3 (16.0,20.6) | 30.3 (25.5,35.1) |
| Cigar | 7.7 (7.2,8.2) | 6.7 (6.2,7.2) | 11.6 (9.7,13.5) | 26.0 (21.5,30.5) |
| Hookah | 7.7 (7.3,8.2) | 6.8 (6.3,7.3) | 12.2 (10.3,14.2) | 21.6 (17.3,26.0) |
| Smokeless | 4.9 (4.5,5.3) | 4.3 (3.9,4.7) | 7.5 (5.9,9.1) | 17.3 (13.3,21.3) |
| Multiple Products | 12.3 (11.7,12.9) | 10.9 (10.3,11.5) | 18.4 (16.1,20.8) | 35.4 (30.3,40.4) |
|
| ||||
| Past 30-Day Use of Any Tobacco Product Among Ever-Tobacco Users4 | 42.5 (40.6,44.5) | 40.9 (38.7,43.1) | 43.4 (37.9,48.9) | 58.6 (51.6,65.7) |
| Cigarette | 21.9 (20.3,23.5) | 20.5 (18.8,22.3) | 23.1 (18.6,27.7) | 35.4 (28.8,42.0) |
| E-Cigarette | 15.0 (13.6,16.4) | 14.5 (13.0,16.1) | 14.3 (10.5,18.1) | 21.3 (15.6,27.0) |
| Cigar | 12.2 (10.9,13.4) | 11.3 (9.9,12.7) | 12.0 (8.6,15.4) | 22.1 (16.4,27.8) |
| Hookah | 7.9 (6.9,8.9) | 7.3 (6.2,8.4) | 9.1 (6.1,12.2) | 12.5 (7.9,17.0) |
| Smokeless | 7.4 (6.3,8.4) | 6.8 (5.6,7.9) | 7.1 (4.3,9.9) | 14.7 (9.6,19.8) |
| Multiple Products | 17.3 (15.7,18.8) | 16.0 (14.4,17.7) | 17.7 (13.5,22.0) | 30.4 (23.7,37.1) |
Five most prevalent tobacco products shown
N=10,246
N=13,115
N=2869
Online engagement was associated with each of the tobacco outcomes. Among never-tobacco users, the prevalence of susceptibility to any tobacco product was higher across increasing levels of online engagement: 41.7%, 60.7% and 79.5% for scores of 0, 1, and 2 or more, respectively. The prevalence of ever having tried tobacco and past 30-day tobacco use also was higher across increasing levels of online engagement; this association held for each class of tobacco product. For example, the prevalence of ever having tried e-cigarettes among all respondents increased from 9.7% to 18.3% to 30.3% as the score increased from 0 to 1 to 2 or more. The prevalence of past 30-day e-cigarette use among ever-tobacco users was approximately equal for online engagement scores 0 and 1 (14.5% and 14.3%, respectively) and increased to 21.3% for a score of 2 or more.
Multivariable Analyses
Adjusting for socio-demographic and behavioral characteristics, higher levels of online engagement were associated with higher odds of susceptibility to tobacco use and ever having tried tobacco (Table 4). The odds of susceptibility to tobacco use were 1.48 times higher (95% confidence interval [CI] 1.24 to 1.76) for respondents with an online engagement score of 1 and 2.37 times higher (95% CI 1.53 to 3.68) for respondents with a score of 2 or more, compared to respondents with a score of 0. The odds of ever having tried tobacco were 1.33 times higher (95% CI 1.11 to 1.60) for respondents with an online engagement score of 1 and 1.54 times higher (95% CI 1.16 to 2.03) for respondents with a score of 2 or more compared to respondents with a score of 0. No significant independent associations were observed between online engagement and past 30-day tobacco use.
Table 4.
Multivariable Logistic Regression Results
| Susceptible to Any Tobacco Use Among Never-Tobacco Users | Ever Tried Tobacco Among All Respondents | Past 30-Day Tobacco Use Among Ever-Tobacco Users | ||||
|---|---|---|---|---|---|---|
| Adj. OR | 95% CI | Adj. OR | 95% CI | Adj. OR | 95% CI | |
| Tobacco Marketing | ||||||
| Online Engagement Score (Ref: 0) | ||||||
| 1 | 1.48 | (1.24,1.76) | 1.33 | (1.11,1.60) | 1.04 | (0.79,1.37) |
| 2 or More | 2.37 | (1.53,3.68) | 1.54 | (1.16,2.03) | 1.30 | (0.94,1.80) |
| Marketing Receptivity (Ref: None) | ||||||
| Low | 1.35 | (1.22,1.49) | 0.97 | (0.85,1.11) | 0.76 | (0.61,0.95) |
| Moderate | 3.22 | (2.57,4.03) | 1.70 | (1.35,2.13) | 1.95 | (1.45,2.62) |
| High | 2.33 | (1.92,2.84) | 2.45 | (2.07,2.89) | 3.38 | (2.70,4.22) |
|
| ||||||
| Internet Use | ||||||
| Internet Access (Ref: Less than Once/Week) | ||||||
| 1–2 Days/Week | 0.71 | (0.54,0.93) | 0.56 | (0.38,0.83) | 0.46 | (0.26,0.80) |
| 3–5 Days/Week | 0.91 | (0.73,1.13) | 0.78 | (0.60,1.02) | 0.71 | (0.49,1.04) |
| Once/Day | 0.83 | (0.68,1.00) | 0.75 | (0.59,0.97) | 0.56 | (0.39,0.82) |
| Several Times/Day | 0.99 | (0.83,1.17) | 0.70 | (0.56,0.87) | 0.46 | (0.33,0.65) |
| Social Networking Account Use (Ref: No Account) | ||||||
| Monthly or Less Often | 1.01 | (0.84,1.22) | 1.05 | (0.80,1.38) | 0.74 | (0.49,1.12) |
| Weekly | 1.23 | (1.04,1.46) | 1.10 | (0.88,1.39) | 0.99 | (0.70,1.40) |
| Daily | 1.32 | (1.14,1.53) | 1.20 | (0.98,1.46) | 1.06 | (0.79,1.43) |
| Several Times a Day | 1.49 | (1.28,1.73) | 1.46 | (1.19,1.78) | 1.42 | (1.05,1.91) |
| Use Smart Phone (Ref: No) | 0.91 | (0.82,1.01) | 0.88 | (0.76,1.01) | 0.90 | (0.72,1.13) |
|
| ||||||
| Individual Risk Factors | ||||||
| 15–17 Yrs Old (Ref: 12–14 Yrs Old) | 1.26 | (1.15,1.39) | 2.18 | (1.92,2.46) | 2.56 | (2.09,3.14) |
| Male (Ref: Female) | 1.07 | (0.97,1.18) | 1.17 | (1.04,1.32) | 1.21 | (1.02,1.45) |
| Race/Ethnicity (Ref: Non-Hispanic White) | ||||||
| Non-Hispanic Black | 1.33 | (1.15,1.54) | 0.58 | (0.48,0.71) | 0.64 | (0.49,0.84) |
| American Indian/Alaska Native | 1.33 | (0.71,2.49) | 0.76 | (0.36,1.58) | 1.17 | (0.39,3.50) |
| Asian/Pacific Islander | 1.04 | (0.81,1.36) | 0.40 | (0.24,0.66) | 0.37 | (0.17,0.80) |
| Multiple Races | 1.28 | (1.04,1.58) | 1.16 | (0.91,1.49) | 0.74 | (0.53,1.03) |
| Hispanic | 1.44 | (1.28,1.62) | 0.84 | (0.73,0.97) | 0.66 | (0.53,0.82) |
| Sensation Seeking (Ref: Low) | ||||||
| Moderate | 1.96 | (1.76,2.18) | 1.27 | (1.09,1.47) | 1.30 | (1.03,1.63) |
| High | 3.15 | (2.76,3.59) | 2.03 | (1.73,2.38) | 1.96 | (1.54,2.49) |
| Internalizing Problems (Ref: Low) | ||||||
| Moderate | 1.36 | (1.20,1.53) | 0.91 | (0.78,1.07) | 0.85 | (0.67,1.08) |
| High | 1.53 | (1.33,1.75) | 1.23 | (1.04,1.46) | 1.20 | (0.93,1.55) |
| Externalizing Problems (Ref: Low) | ||||||
| Moderate | 1.34 | (1.17,1.53) | 1.05 | (0.88,1.24) | 0.74 | (0.57,0.96) |
| High | 1.89 | (1.64,2.17) | 0.95 | (0.78,1.14) | 0.59 | (0.44,0.78) |
| Weekly Income (Ref: None) | ||||||
| $1-$20 | 1.06 | (0.96,1.18) | 0.94 | (0.81,1.08) | 1.17 | (0.94,1.46) |
| $21-$50 | 1.40 | (1.17,1.66) | 1.70 | (1.41,2.06) | 1.63 | (1.21,2.19) |
| $51+ | 1.17 | (0.97,1.41) | 1.97 | (1.63,2.37) | 2.08 | (1.60,2.72) |
| School Performance (Ref: Mostly As) | ||||||
| As and Bs | 1.22 | (1.08,1.37) | 1.71 | (1.44,2.03) | 1.79 | (1.36,2.37) |
| Mostly Bs | 1.31 | (1.09,1.57) | 1.95 | (1.54,2.45) | 2.15 | (1.50,3.09) |
| Bs and Cs | 1.39 | (1.20,1.61) | 2.48 | (2.04,3.01) | 2.28 | (1.68,3.07) |
| Mostly Cs to Mostly Fs | 1.52 | (1.27,1.82) | 3.76 | (3.07,4.62) | 3.87 | (2.88,5.20) |
| Other Substance Use (Ref: 0) | ||||||
| 1 | 3.42 | (2.33,5.01) | 12.70 | (10.52,15.35) | 7.95 | (6.55,9.65) |
| 2 or More | 9.98 | (1.19,84.00) | 27.36 | (16.02,46.73) | 23.32 | (15.94,34.14) |
|
| ||||||
| Other Risk Factors | ||||||
| Parental Education (Ref: At Least Some College) | ||||||
| High School Graduate | 0.82 | (0.72,0.93) | 1.19 | (1.03,1.39) | 1.11 | (0.89,1.39) |
| Less than High School | 0.89 | (0.78,1.01) | 1.20 | (1.03,1.40) | 1.02 | (0.81,1.29) |
| Exposure to Smoking (Ref: No) | 1.46 | (1.32,1.61) | 2.28 | (2.02,2.57) | 2.73 | (2.24,3.33) |
Note: Adj=Adjusted; OR=Odds Ratio; CI=Confidence Interval; Yrs=Years; Ref=Reference; —=Not Applicable. Each model simultaneously adjusted for all covariates listed in first column.
In addition to the level of online engagement, higher levels of receptivity to traditional tobacco marketing channels were independently associated with greater odds of susceptibility, ever having tried tobacco, and past 30-day tobacco use. For example, the odds of susceptibility increased from 1.35 times higher (95% CI, 1.22 to 1.49) for low marketing receptivity, 3.22 times higher (95% CI, 2.57 to 4.03) for moderate marketing receptivity, and 2.33 times higher (95% CI, 1.92 to 2.84) for high marketing receptivity, compared to respondents with no marketing receptivity. The odds of susceptibility were higher for adolescents with higher levels of sensation seeking, internalizing disorders, and externalizing disorders. For example, the odds of susceptibility increased from 1.36 times higher (95% CI, 1.20 to 1.53) for respondents with a moderate level of internalizing disorders and 1.53 times higher (95% CI, 1.33 to 1.75) for respondents with a high level of internalizing disorders compared to respondents with a low level of internalizing disorders. Finally, the odds of susceptibility, ever having tried tobacco, and past 30-day tobacco use were also higher for older adolescents, as well as adolescents exposed to smoking, performed at lower levels in school, and used other substances. For example, the odds of past 30-day tobacco use increased from 7.95 times higher (95% CI, 6.55 to 9.65) for use of one other substance to 23.32 times higher (95% CI, 15.94 to 34.14) for use of two or more other substances, compared to respondents with no other substance use.
DISCUSSION
Three central findings are reported in this cross-sectional analysis of engagement with online tobacco marketing in a nationally representative sample of youth. First, 12%, or approximately 2.9 million youth, engaged with at least one form of online tobacco marketing. Second, higher levels of online engagement were associated with greater susceptibility to tobacco use among never-tobacco users and ever having tried tobacco. Third, higher levels of receptivity to tobacco marketing in traditional media venues were also associated with these tobacco-related outcomes, independent of online engagement.
Adolescents and young adults who are susceptible to tobacco use are, indeed, more likely to initiate use than their non-susceptible counterparts [18,19,27,28]. For example, a six-year longitudinal study of 1,574 never cigarette-smoking adolescents (aged 12–15 years old at baseline) found that the sensitivity and positive predictive value of the cigarette-specific susceptibility index equaled 78.9% and 19.0%, respectively, for smoking ≥100 cigarettes in respondents’ lifetime [19]. The odds of hookah smoking initiation were 2.52 times higher (95% CI, 1.39 to 4.60) among college freshmen susceptible to hookah smoking compared to their non-susceptible classmates in a four-year longitudinal study [27]. Finally, the odds of e-cigarette use initiation were 4.27 time higher among middle and high school students susceptible to e-cigarette use compared to their non-susceptible classmates in a one-year longitudinal study [28].
Our finding of the strong association between online engagement and susceptibility links this activity to the earliest stages of tobacco product use and adds to a well-established body of research on the effect of traditional tobacco advertising and promotion [29–31]. Youth who have never used tobacco and who enter e-cigarette brand websites, for example, can see what others write about their experiences with products on message boards, as well as interact with the website through its games, videos, and contests. Social networking sites can influence youth to become part of online peer networks around specific tobacco products. This stimulation and opportunity to socialize can magnify the effectiveness of online tobacco marketing compared to traditional marketing in reaching susceptible new users, changing perceived norms, and altering risk perceptions associated with tobacco products [12]. In addition, youth who are susceptible to tobacco use may engage with online tobacco marketing to learn more about specific products, as well as seek pleasure and reassurance from tobacco advertising [32,33].
Our findings also strengthen long-standing concern about youth exposure to tobacco advertising on interactive and participatory websites that emphasize user-generated content [5,6,34–36]. For example, the proportion of middle and high school students who reported exposure to pro-tobacco messages on the internet increased from 22% in 2000 to 33% in 2004 [37]. Our study considers engagement with—rather than exposure to—online tobacco marketing and concludes approximately 2.9 million adolescents (12%) engaged with such marketing in 2013–2014. Our finding represents a public health concern because experiment-based studies find online engagement increases advertising effectiveness [12]. Thus, close monitoring of online tobacco marketing and youth engagement trends over time is warranted because of its potential to influence pro-tobacco attitudes among youth.
An additional issue is the interplay between traditional and online marketing. Tobacco ad images in magazines, for example, now often refer viewers to online venues. Our findings extend and support earlier studies conducted before the internet era that conclude receptivity to traditional tobacco marketing increased the risk of susceptibility to cigarette smoking [20,29,38]. Notably, online tobacco marketing seems to be capturing different populations of youth compared to traditional venues (as suggested by the relatively low correlation between the two exposures). Future research may seek to better understand how youth encounter online venues.
Several important limitations are noted. First, the temporal order of engagement with online tobacco marketing and tobacco use cannot be established given the cross-sectional nature of the study. Based on future waves of the PATH Study, the longitudinal association between online engagement among non-tobacco using youth and their risk of tobacco use—accounting for known psychosocial and behavioral risk factors—can be better determined. Alternatively, youth who are already susceptible to tobacco use or used tobacco in the past 30 days may be more likely to subsequently engage with online marketing. Second, our analysis relies upon respondents’ self-report of online engagement and tobacco use, both of which may be subject to recall bias. Third, the frequency and recency of engagement with online tobacco marketing cannot be determined. Fourth, although engagement among youth was studied, engagement among young adults may be equally important because the older group serves as aspirational role models to the younger group [39]. Fifth, online engagement may have been underestimated, as youth may receive information about and discuss tobacco products on social media platforms that were not studied (e.g., Snapchat™ and Instagram™). Sixth, we assessed engagement with online tobacco marketing overall and not for specific tobacco products (e.g., e-cigarettes). Thus, we cannot ascertain, for example, if engagement with online e-cigarette marketing is cross-sectionally associated with susceptibility to e-cigarette use. Finally, the predictive validity of the susceptibility index found among younger adolescents may not extend to older adolescents.
In conclusion, a substantial number of youth engage with online tobacco marketing. Online engagement with tobacco marketing may represent an important risk factor for youth tobacco use that has important regulatory implications because youth who engage with online tobacco marketing may be more susceptible to tobacco use than unengaged youth.
IMPLICATIONS & CONTRIBUTIONS.
Findings from this nationally representative sample suggest that youth who engage with online tobacco marketing may be more susceptible to tobacco use than unengaged youth. Continued monitoring of online engagement may help identify those at risk for future tobacco use.
Acknowledgments
This manuscript is supported with Federal funds from the National Institute on Drug Abuse, National Institutes of Health, and the Food and Drug Administration, Department of Health and Human Services, under a contract to Westat (Contract No. HHSN271201100027C). The views and opinions expressed in this manuscript are those of the authors only and do not necessarily represent the views, official policy or position of the U.S. Department of Health and Human Services or any of its affiliated institutions or agencies. The corresponding author, Dr. Soneji, affirms that he listed as authors everyone who contributed significantly to the work.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Federal Trade Commission. Federal Trade Commission Cigarette Report for 2013. 2016. [Google Scholar]
- 2.Federal Trade Commission. Federal Trade Commission Smokeless Tobacco Report for 2013. 2016. [Google Scholar]
- 3.Richardson A, Vallone DM. YouTube: a promotional vehicle for little cigars and cigarillos? Tob Control. 2014;23:21–6. doi: 10.1136/tobaccocontrol-2012-050562. [DOI] [PubMed] [Google Scholar]
- 4.Elkin L, Thomson G, Wilson N. Connecting world youth with tobacco brands: YouTube and the internet policy vacuum on Web 2. 0. Tob Control. 2010;19:361–6. doi: 10.1136/tc.2010.035949. [DOI] [PubMed] [Google Scholar]
- 5.Freeman B. New media and tobacco control. Tob Control. 2012;21:139–44. doi: 10.1136/tobaccocontrol-2011-050193. [DOI] [PubMed] [Google Scholar]
- 6.Ribisl KM, Jo C. Tobacco control is losing ground in the Web 2. 0 era: invited commentary. Tob Control. 2012;21:145–6. doi: 10.1136/tobaccocontrol-2011-050360. [DOI] [PubMed] [Google Scholar]
- 7.Freeman B, Chapman S. Open source marketing: Camel cigarette brand marketing in the “Web 2. 0” world. Tob Control. 2009;18:212–7. doi: 10.1136/tc.2008.027375. [DOI] [PubMed] [Google Scholar]
- 8.Richardson A, Ganz O, Vallone D. Tobacco on the web: surveillance and characterisation of online tobacco and e-cigarette advertising. Tob Control. 2014 doi: 10.1136/tobaccocontrol-2013-051246. tobaccocontrol-2013-051246. [DOI] [PubMed] [Google Scholar]
- 9.Tessman GK, Caraballo RS, Corey CG, et al. Exposure to Tobacco Coupons Among U.S. Middle and High School Students. Am J Prev Med. 2014;47:S61–8. doi: 10.1016/j.amepre.2014.05.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ribisl KM. The potential of the internet as a medium to encourage and discourage youth tobacco use. Tob Control. 2003;12:i48–59. doi: 10.1136/tc.12.suppl_1.i48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lewis MJ, Yulis SG, Delnevo C, et al. Tobacco Industry Direct Marketing after the Master Settlement Agreement. Health Promot Pract. 2004;5:75S–83S. doi: 10.1177/1524839904264596. [DOI] [PubMed] [Google Scholar]
- 12.Calder BJ, Malthouse EC, Schaedel U. An Experimental Study of the Relationship between Online Engagement and Advertising Effectiveness. J Interact Mark. 2009;23:321–31. [Google Scholar]
- 13.Williams RS, Derrick J, Ribisl KM. Electronic cigarette sales to minors via the internet. JAMA Pediatr. 2015;169:e1563. doi: 10.1001/jamapediatrics.2015.63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Andrews JC, Choiniere CJ, Portnoy DB. Opportunities for Consumer Research from the Food and Drug Administration’s Center for Tobacco Products. J Public Policy Mark. 2015;34:119–30. [Google Scholar]
- 15.United States Department of Health and Human Services. National Institutes of Health. National Institute on Drug Abuse, United States Department of Health and Human Services. Food and Drug Administration. Center for Tobacco Products. Population Assessment of Tobacco and Health (PATH) Study. United States: Restricted-Use Files 2016. [Google Scholar]
- 16.Hyland A, Ambrose BK, Conway KP, et al. Design and methods of the Population Assessment of Tobacco and Health (PATH) Study. Tob Control. 2016 doi: 10.1136/tobaccocontrol-2016-052934. tobaccocontrol-2016-052934. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.United States Department of Health and Human Services. Population Assessment of Tobacco and Health (PATH) Study 2013–2016 [United States] Restricted-Use Files User Guide. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor]; 2015. [Google Scholar]
- 18.Pierce JP, Choi WS, Gilpin EA, et al. Validation of susceptibility as a predictor of which adolescents take up smoking in the United States. Health Psychol Off J Div Health Psychol Am Psychol Assoc. 1996;15:355–61. doi: 10.1037//0278-6133.15.5.355. [DOI] [PubMed] [Google Scholar]
- 19.Strong DR, Hartman SJ, Nodora J, et al. Predictive Validity of the Expanded Susceptibility to Smoke Index. Nicotine Tob Res. 2015;17:862–9. doi: 10.1093/ntr/ntu254. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pierce J, Won C, Elizabeth G, et al. Tobacco industry promotion of cigarettes and adolescent smoking. JAMA. 1998;279:511–5. doi: 10.1001/jama.279.7.511. [DOI] [PubMed] [Google Scholar]
- 21.Dennis ML, Chan Y-F, Funk RR. Development and validation of the GAIN Short Screener (GSS) for internalizing, externalizing and substance use disorders and crime/violence problems among adolescents and adults. Am J Addict Am Acad Psychiatr Alcohol Addict. 2006;15(Suppl 1):80–91. doi: 10.1080/10550490601006055. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hoyle RH, Stephenson MT, Palmgreen P, et al. Reliability and validity of a brief measure of sensation seeking. Personal Individ Differ. 2002;32:401–14. [Google Scholar]
- 23.Zuckerman M. Sensation Seeking and Risk. Washington, DC, US: American Psychological Association; 2007. [Google Scholar]
- 24.Korn E, Graubard B. Confidence Intervals For Proportions With Small Expected Number of Positive Counts Estimated From Survey Data. Surv Methodol. n.d;23:193–201. [Google Scholar]
- 25.Rao JNK, Scott AJ. On Chi-Squared Tests for Multiway Contingency Tables with Cell Proportions Estimated from Survey Data. Ann Stat. 1984;12:46–60. [Google Scholar]
- 26.Honaker J, King G, Blackwell M. Amelia II: A Program for Missing Data. J Stat Softw. 2011;45:1–47. [Google Scholar]
- 27.Lipkus IM, Reboussin BA, Wolfson M, et al. Assessing and Predicting Susceptibility to Waterpipe Tobacco Use Among College Students. Nicotine Tob Res. 2015;17:1120–5. doi: 10.1093/ntr/ntu336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Bold Krysten, Kong Grace, Cavallo Dana, et al. E-cigarette Susceptibility as a Predictor of Youth Intiation of E-Cigarettes. Nicotine Tob Res. doi: 10.1093/ntr/ntw393. In Press. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Evans N, Farkas A, Gilpin E, et al. Influence of Tobacco Marketing and Exposure to Smokers on Adolescent Susceptibility to Smoking. J Natl Cancer Inst. 1995;87:1538–45. doi: 10.1093/jnci/87.20.1538. [DOI] [PubMed] [Google Scholar]
- 30.Pierce JP, Messer K, James LE, et al. Camel No. 9 cigarette-marketing campaign targeted young teenage girls. Pediatrics. 2010;125:619–26. doi: 10.1542/peds.2009-0607. [DOI] [PubMed] [Google Scholar]
- 31.National Center for Chronic Disease Prevention and Health Promotion (US) Office on Smoking and Health. Preventing Tobacco Use Among Youth and Young Adults: A Report of the Surgeon General. Atlanta (GA): Centers for Disease Control and Prevention (US); 2012. [PubMed] [Google Scholar]
- 32.Aitken PP, Eadie DR, Hastings GB, et al. Predisposing effects of cigarette advertising on children’s intentions to smoke when older. Br J Addict. 1991;86:383–90. doi: 10.1111/j.1360-0443.1991.tb03415.x. [DOI] [PubMed] [Google Scholar]
- 33.Portnoy DB, Wu CC, Tworek C, et al. Youth curiosity about cigarettes, smokeless tobacco, and cigars: prevalence and associations with advertising. Am J Prev Med. 2014;47:S76–86. doi: 10.1016/j.amepre.2014.04.012. [DOI] [PubMed] [Google Scholar]
- 34.American Legacy Foundation. Vaporized: E-Cigarettes, Advertising, and Youth. Washington DC: 2014. [Google Scholar]
- 35.Duke JC, Lee YO, Kim AE, et al. Exposure to Electronic Cigarette Television Advertisements Among Youth and Young Adults. Pediatrics. 2014;134:e29–36. doi: 10.1542/peds.2014-0269. [DOI] [PubMed] [Google Scholar]
- 36.Harkin Senators Tom, Rockefeller John, Blumenthal Richard, Makey Edward, Brown Sherrod, Reed Jack, Boxer Barbara, Merkley Jeff. Congressional Staff of Senator Richard Durbin; Representative Henry Waxman; Representative Frank Pallone. Gateway to Addiction? A Survey of Popular Electronic Cigarette Manufacturers and Targted Marketing to Youth. US Congress; 2014. [Google Scholar]
- 37.Duke JC, Allen JA, Pederson LL, et al. Reported exposure to pro-tobacco messages in the media: trends among youth in the United States, 2000–2004. Am J Health Promot. 2009;23:195–202. doi: 10.4278/ajhp.071130126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sargent JD, Dalton M, Beach M. Exposure to Cigarette Promotions and Smoking Uptake in Adolescents: Evidence of a Dose-Response Relation. Tob Control. 2000;9:163–8. doi: 10.1136/tc.9.2.163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Kastenbaum R, Derbin V, Sabatini P, et al. “The Ages of Me”: Toward Personal and Interpersonal Definitions of Functional Aging. Int J Aging Hum Dev. 1972;3:197–211. [Google Scholar]