Effect of a National Tobacco Public Education Campaign on Youth’s Risk Perceptions and Beliefs About Smoking

Abstract

Purpose:

To assess the relationship between youth’s exposure to the Food and Drug Administration’s national tobacco public education campaign, The Real Cost, and changes in campaign-targeted beliefs.

Design:

Longitudinal design with baseline survey and 2 postcampaign follow-up surveys.

Sample:

Youth from 75 US media markets (N = 1680) who completed all 3 surveys and had experimented with or were susceptible to future cigarette smoking.

Measures:

Exposure was measured by self-reported frequency of ad exposure and media market–level target rating points. Agreement with 30 self-reported tobacco-related beliefs was assessed in 3 categories: (1) beliefs specifically targeted by campaign messages (campaign-targeted belief), (2) beliefs not targeted by the campaign (nontargeted beliefs), and (3) beliefs corresponding to other media campaigns on air concurrent with The Real Cost (ambiguous beliefs).

Analysis:

Descriptive analyses of aggregate changes in beliefs and logistic regressions to examine the association between campaign exposure and beliefs.

Intervention:

The Real Cost.

Results:

Agreement with campaign-targeted beliefs increased from baseline to first and second follow-ups, with a mean relative increase of 10.4% and 11.5%, respectively. Nontargeted beliefs did not change substantially. Both measures of campaign exposure were positively associated with increased odds of agreeing with 5 of 8 campaign-targeted beliefs. Exposure was not significantly associated with 12 of 14 nontargeted tobacco-related beliefs.

Discussion:

A sustained national tobacco public education campaign can change population-level perceptions of tobacco-related harms among youth.

Keywords

tobacco control mass media campaign adolescent logistic regression

Purpose

In February 2014, the Food and Drug Administration’s (FDA) Center for Tobacco Products launched The Real Cost, the first of several complementary national public education campaigns designed to prevent and reduce tobacco use among US youth aged 12 to 17. Initial campaign advertising under The Real Cost was designed to prevent the initiation of cigarette smoking among youth who have never smoked but are susceptible to smoking (susceptible nonsmokers) and to discourage further smoking among youth who have experimented with smoking in the past (experimenters).

A substantial body of research finds that media campaigns are effective at reducing youth smoking prevalence,^1,2 preventing initiation,^3
–5 and disrupting progression to established smoking.^3,6 Moreover, campaigns that are designed for and delivered to a general population of at-risk youth can be effective across gender and racial/ethnic groups.⁷ Based on the strength of this research evidence, mass media campaigns are widely considered to be a “best practice for tobacco control.”^8
–10

The Real Cost is based on the Theory of Planned Behavior, which posits that behavior change is a result of changes in beliefs that, in turn, influence attitudes toward a behavior, perceptions of associated social norms, and/or self-efficacy to engage in or refrain from a behavior.^11,12 The Centers for Disease Control and Prevention’s (CDC) evidence-based guidance for implementing effective tobacco mass media campaigns is based on this theoretical framework, and numerous studies support its value for developing and evaluating public health and education campaigns.^10,13,14

The Real Cost has aired continuously at high media levels since its launch, resulting in few unexposed youth aged 12 to 17 within the first year. More than 90% of the target audience (ie, youth experimenters and susceptible nonsmokers) reported ad awareness 8 months after campaign launch, and most surveyed youth considered advertisements to be effective based on assessments of perceived effectiveness.¹⁵ Studies indicate that perceived effectiveness of ad is predictive of attitudes and behavioral outcomes across a range of health behaviors.^16

–19 Based on CDC guidance and evidence from prior youth tobacco education campaign evaluations, The Real Cost should be influencing campaign-targeted beliefs^2,10,20 after 14 to 16 months.

The current study examined the association between temporal and geographic variation in campaign exposure and individuals’ beliefs with a focus on message content specific to The Real Cost.²¹ We examined data from the first 3 waves of a national longitudinal survey of US youth to determine whether campaign exposure influenced campaign-targeted beliefs among the target audience.

Methods

Design and Sample

Data are from a nationally representative longitudinal survey of US youth aged 11 to 16 at baseline who would age into The Real Cost’s audience of 12- to 17-year-olds after campaign launch. The study was conducted as part of The Real Cost evaluation. We used an address-based sampling frame to randomly draw households clustered in US census block groups within 75 media markets and supplemented the frame with market research databases to identify households likely to have at least 1 eligible youth (approximately 5% of households). A lead letter describing the study was mailed to each of the selected addresses. Subsequently, a field interviewer visited each address to secure an immediate interview or schedule one for a later date. In-person baseline data collection took place from November 11, 2013, through March 31, 2014. First follow-up data collection (July 24, 2014, to October 27, 2014) and second follow-up data collection (April 6, 2015, to July 4, 2015) consisted of online or in-person interviews with youth after obtaining parental permission and youth assent. Youth participants received a $20 incentive for completion of the baseline survey and $20 or $25 for completion of each follow-up. The final total sample size for the second follow-up survey was 5240 youth. The study was approved by institutional review boards at FDA and the researching institution.

The primary target audience consisted of youth experimenters and susceptible nonsmokers.²² Thus, we examined 1680 youth (32.1% of total sample) who were among the campaign’s target audience at baseline and completed the second follow-up. Experimenters (n = 399) were defined as youth who reported smoking fewer than 100 cigarettes in their lifetime at baseline. Susceptible nonsmokers (n = 1281) were defined as never smokers (not even 1 or 2 puffs) who indicated an openness to trying cigarettes at baseline. Specifically, susceptible nonsmokers did not respond “definitely not” to at least 1 of 3 items assessing future intentions to smoke: (1) “Do you think that you will try a cigarette soon?,” (2) “Do you think you will smoke a cigarette anytime during the next year?,” and (3) “If one of your best friends offered you a cigarette, would you smoke it?”²³

At baseline, 7418 sampled households were eligible to participate (ie, contained youth aged 11 to 16 years). An additional 2083 households with unknown eligibility were estimated to be eligible. Participants in 4538 households completed a survey. Thus, the unweighted household-level response rate was 47.8% and the weighted household-level response rate was 43.7% at baseline (American Association of Public Opinion Research Response Rate #3 formula). At the first follow-up, the weighted person-level response rate was 85.6%. At the second follow-up, the weighted person-level response rate was 84.0% of the first follow-up. To produce reliable estimates of target population parameters, baseline analysis weights that accounted for unequal probabilities of selection at each of 3 stages (ie, media market, census block groups, and address) were adjusted for nonresponse. Then, weights were calibrated to the Census 2010 population totals of the baseline target population with poststratification for gender and race/ethnicity. At each follow-up, weights were adjusted for nonresponse and calibrated to population totals with poststratification.

Campaign Message and Implementation

The central theme of The Real Cost is “Every cigarette costs you something.” Advertising in the first year focused on short-term health effects (eg, cosmetic effects) and loss of control (eg, nicotine addiction) associated with smoking. Advertising in the second year continued these themes and introduced new messages about the danger of the mix of toxic chemicals in cigarette smoke (Appendix A). The media purchase for The Real Cost included advertising on national television, radio, the Internet, and out-of-home displays, as well as in magazines and at movie theatres directed at youth aged 12 to 17. This study examines the effects of paid video advertisements airing on television and online (“digital”). Similar to other studies,⁹ other channels were excluded due to measurement limitations and/or low relative media weight. Nielsen’s system of target rating points (TRPs) are the standard unit of measurement for media delivery. Target rating points are calculated as the product of 2 measures: the percentage of a target population potentially exposed to advertisements (reach) and the average number of times advertisements may have been seen (frequency) over a time period. Youth in the study markets were exposed to a mean of 2511 cumulative television TRPs and 629 cumulative digital TRPs between baseline and the first follow-up. A mean of 2902 television TRPs and 904 total digital TRPs aired between the first and second follow-ups. Four campaign ads aired between baseline and the first follow-up: “Your Teeth” (aired March 17, 2014, to September 9, 2014, December 29, 2014, to June 8, 2015), “Your Skin” (February 10, 2014, to October 27, 2014, December 29, 2014, to June 8, 2015), “Bully” (February 10, 2014, to December 8, 2014), and “So Bossy” (March 17, 2014, to September 22, 2014). Your Teeth and Your Skin remained on air between the first and second follow-ups, and 2 new ads were introduced: “Stay in Control” (November 3, 2014, to June 8, 2015) and “7000” (December 15, 2014, to June 22, 2015).

Measures

Outcome variables

Tobacco-related beliefs

Thirty items measured youth’s beliefs about cigarette smoking. Three coders reviewed The Real Cost advertisements to identify belief items that captured each advertisement’s primary message. Rater agreement was high (overall κ = 0.88; individual ad κ from 0.72 to 1.00). Table 1 shows the 8 items that correspond to specific messages highlighted in advertisements (“campaign targeted”) and 14 items that do not correspond to campaign messages (“nontargeted”). Eight items were coded as “ambiguous” because these messages are contained in other national media campaigns on air concurrent with The Real Cost, specifically Truth Initiative’s youth prevention campaign truth® and CDC’s adult cessation campaign Tips From Former Smokers (see Table 1). Respondents indicated their agreement with each statement and were dichotomized by response as “agree”/”strongly agree” versus “neither”/”disagree”/”strongly disagree.”

Table 1.

Percentage of Youth Who Agree or Strongly Agree With Smoking-Related Belief Statements, by Study Wave.

Belief Statement	Baseline (%)	First Follow-up (%)	Second Follow-up (%)	Change Baseline to First Follow-up (%)	Change From Baseline to Second Follow-up (%)
Items related to advertisements
If I smoke, I will lose my teeth^a	65.1	76.7	78.7^b	11.6%	13.6%
If I smoke, I will have problems with my teeth^a	86.6	87.5	87.2	0.9%	0.6%
If I smoke, I will get wrinkles^c	58.4	72.2	78.9^b,d	13.8%	20.5%
If I smoke, I will develop skin problems^c	67.4	78.8	80.2^b	11.4%	12.8%
If I smoke, I will be controlled by smoking^e	59.5	67.9	66.2^b	8.4%	6.7%
If I smoke, I will become addicted^e	75.1	77.9	72.7	2.8%	−2.4%
If I smoke, I will inhale poisons^f	79.0	85.1	86.1^b	6.1%	7.1%
Cigarette ingredients are dangerous^f	87.4	86.0	86.9	−1.4%	−0.5%
Mean absolute change across items				6.7%	7.3%
Mean relative change across items				10.4%	11.5%
Ambiguous items
If I smoke I will: Damage my body	84.5	91.5	90.3^b	7.0	5.8
If I smoke I will: Develop cancer of the lip, mouth, tongue, or throat	79.9	84.3	82.6	4.4	2.7
If I smoke I will: Develop a smoking-related disease	80.3	81.8	84.0	1.5	3.7
Smoking can cause immediate damage to my body	57.8	54.8	57.4	−3.0	−0.4
If I smoke, I will be unable to stop when I want to	62.8	64.2	63.1	1.3	0.3
If I started to smoke occasionally, I would not become addicted	26.5	29.2	23.7	2.7	−2.8
If I smoke I will: Be unattractive	54.7	61.1	64.4^b	6.4	9.7
Cigarette ingredients are disgusting	78.4	80.5	79.9	2.1	1.5
Mean absolute change across items				2.8	2.6
Mean relative change across items				4.6	3.0
Items unrelated to advertisements
If I smoke, I will develop sexual and/or fertility problems	48.0	52.3	54.5^b	4.3	6.5
If I smoke, I will decrease my sports performance	84.1	85.3	84.2	1.2	0.1
If I smoke, I will waste money on cigarettes	91.8	89.0	89.1	−2.8	−2.7
If I smoke, I will harm others with secondhand smoke	81.6	81.5	83.1	−0.1	1.5
If I smoke, I will be a bad influence on others	80.8	78.8	78.0	−2.0	−2.8
Smoking is a way to show others you’re not afraid to take risks	14.9	13.8	9.5^b	−1.1	−5.4
Smoking cigarettes helps people relieve stress	38.9	40.6	41.8	1.7	2.9
Smoking cigarettes can help keep your weight down	15.9	17.8	16.5	1.9	0.6
If I smoke, I will lose my taste buds	55.6	62.6	67.3^b	7.0	11.7
If I smoke, I will have trouble breathing	55.7	61.2	59.6	5.5	3.9
If I smoke, I will shorten my life	84.6	84.6	86.8	0.0	2.2
If I smoke, I will get sick more often	74.2	74.0	76.0	−0.2	1.8
It is safe for me to smoke for only a year or two, as long as I quit after that	18.8	12.3	7.0^b,d	−6.5	−11.8
If I smoke, I will have bad breath	88.3	90.1	88.9	1.8	0.6
Mean absolute change across items				0.8	0.7
Mean relative change across items				0.2	−3.2

^aPrimary message of The Real Cost advertisement “Your Teeth.”

^bSignificant difference from baseline (P < .05).

^cPrimary message of The Real Cost advertisement “Your Skin.”

^dSignificant difference from the first follow-up (P < .05).

^ePrimary message of The Real Cost advertisements “Bully,” “So Bossy,” and “Stay in Control.”

^fPrimary message of The Real Cost advertisement “7000.”

Independent variables

The Real Cost media exposure

Youth recall of The Real Cost brand name was measured at follow-up by showing respondents the campaign logo and asking, “In the past 3 months, have you seen or heard the following slogan or theme [The Real Cost]?” with a response option of “yes” or “no.” To assess the frequency of ad exposure at follow-up, respondents viewed each ad once via video stream within the context of the survey. After viewing each ad, respondents reported their frequency of exposure to the ad since the last survey date (ie, “never,” “rarely,” “sometimes,” “often,” or “very often”). These procedures are a validated method of measuring ad exposure and capture encoded exposure more precisely than tests that rely on memory recall alone.²⁴ Frequency of ad exposure was measured from 0 (“never”) to 4 (“very often”). Overall campaign awareness was defined as a response of “rarely” or greater to at least 1 advertisement.

Similar to previous studies,^2,25,26 the primary exogenous measure of interest is cumulative TRPs by media market assessed for the periods between baseline and the first follow-up and between the first and second follow-ups. Two TRP measures represent (1) the total cumulative television and digital TRPs across all ads for the study period and (2) the total TRPs for advertisements corresponding to specific beliefs. For example, the TRPs for the ad Your Skin, which emphasizes the relationship between smoking and wrinkles, were used to examine changes in the item, “If I smoke, I will get wrinkles.” Cumulative total TRPs are a standard measure in media research,^2,25,26 whereas ad-specific TRPs provide more precise data linking specific ad messages with relevant beliefs.

Potential confounders

Youth risk-related variables

Models controlled for a number of factors known to influence susceptibility to tobacco use and risky behaviors.²⁷ We used the Brief Sensation Seeking Scale (BSSS-4)²⁸: (1) I would like to explore strange places; (2) I like to do frightening things; (3) I like new and exciting experiences, even if I have to break the rules; and (4) I prefer friends who are exciting and unpredictable. Responses ranged from 1 (“disagree strongly”) to 5 (“agree strongly”) and a Cronbach α = .74. Educational plans were assessed with the item “How far do you think you will go in school?” with response options from 1 (“I don’t plan to go to school anymore”) to 8 (“graduate, medical, or law school”). School environment was measured as the mean of 3 items: (1) I feel close to people at my school, (2) I am happy to be at my school, and (3) I feel like I am a part of my school. Responses ranged from 1 (“disagree strongly”) to 5 (“agree strongly”) with a Cronbach’s α = .79. School performance was assessed with the item “How well would you say you have done in school?” with response options from 1 (“much worse than average”) to 5 (“much better than average”). Parent communication was a mean of 2 items with a Cronbach α = .69: “Thinking about the adult or adults you live with, would you say you are satisfied with the way you communicate with each other” (responses from 1 [“very unsatisfied”] to 5 [“very satisfied”]) and “How close do you feel to the adult or adults you live with?” (responses from 1 [“not close at all”] to 5 [“very close”]).

Exposure to other media campaigns

We included dichotomous variables for self-reported awareness of the truth® and Tips From Former Smokers campaigns, assessed at each follow-up. Awareness of each campaign brand name was measured at follow-up by showing respondents the campaign logo and asking, “In the past 3 months, have you seen or heard the following slogan or theme [truth®/Tips From Former Smokers]?,” with a response option of “yes” (1) or “no”(2).

Demographic and other variables

Our analysis included a range of covariates similar to those used in other media studies.^2
–4,26 Baseline individual characteristics included age (continuous measure); an indicator for female (male excluded as the reference); indicators for African American, Hispanic, and other non-Hispanic race/ethnicity (white excluded as the reference); youth weekly income (continuous); the presence of a tobacco user in the household; and television use, daily hours watching television across media devices (low, medium, and high). State adult smoking prevalence (2013 Behavioral Risk Factor Surveillance System) was included (1 unit increase = 1 percentage point), as were 3 media market variables: median population size (in tens of thousands), median income (in tens of thousands of dollars), and media market education level (the proportion with bachelor’s degree or higher). Because the baseline data collection occurred after campaign launch (range: 1-48 days) for 21.8% of the sample, a postcampaign indicator variable was also included.

Analyses

Descriptive statistics summarize data on The Real Cost brand and overall campaign awareness. We used multivariable logistic regression models to estimate the odds of each campaign-targeted outcome variable as a function of each exposure measure (cumulative total TRPs, ad-specific TRPs, and self-reported frequency of exposure) using a time-series approach with a pooled sample. Models also estimated the odds of each nontargeted and ambiguous outcome as a function of cumulative total TRPs. The value of the exposure measures was zero at baseline and nonzero at the first and second follow-up. Outcomes at all 3 waves control for the individual and market-level covariates noted above. All model estimates were weighted and adjusted to account for the clustering of data on individuals (3 observations per person using Stata’s xtlogit command) and the temporal ordering of observations within each individual data cluster. In models using total TRPs, the TRPs were rescaled to yield odds ratios for the increased odds of each outcome, given a unit increase of 2000 total TRPs in each media market. This scaling unit represents the approximate size of the market-level media buy for The Real Cost in the implementation period and is consistent with CDC guidelines.¹⁰ In models with ad-specific TRPs, the TRPs were scaled to a unit increase of 1000 TRPs, which represents a doubling of ad-specific message TRPs over the same period. We found that the 2 TRP exposure variables were best characterized by a square root functional form, which adequately characterizes asymptotic diminishing returns with no negative TRP effects over the range of observed data. Specification tests (Linktest in Stata 14.0) and postestimation comparisons of model fit statistics (Akaike and Bayesian information criterion) favored the square root specification of TRPs over other forms (eg, linear and parabolic). Other alternate specifications examined the following variables in addition to those included in the final models but did not alter the results: parental education (missing for 7% of sample), 2012 state cigarette excise taxes and 2012 per capita state tobacco funding, and media market fixed effects. All analyses were conducted with Stata version 14.0 statistical software.

Results

Sample Characteristics

The majority of respondents in the unweighted sample (n = 1680) were aged 14 to 16 (61.9%; Table 2). The sample was evenly divided by gender; it was 46.4% white, non-Hispanic; 35.7% Hispanic; 7.6% black, non-Hispanic; 3.2% Asian American or Pacific Islander; and 7.2% other or multiracial. Almost half of respondents (41.9%) reported living in the same household as a tobacco user. More youth in the study sample were located in the West and fewer were located in the Midwest compared with the national geographic distribution of youth (2010 US Census). An analysis of sample attrition from baseline to second follow-up found that the unweighted samples were similar across age, gender, and household tobacco use categories. There were a few descriptive differences over time; the follow-up sample contained slightly more white respondents, fewer black respondents, fewer smokers, and more susceptible nonsmokers than the baseline sample. However, all descriptive differences were very small; the mean absolute value of difference across comparisons was 1.0%.

Table 2.

Characteristics of Study Sample.

Characteristic^a	Study Sample
Characteristic^a	N	Unweighted %	Weighted %
Age
11	159	9.5	11.4
12	209	12.4	12.7
13	272	16.2	17.0
14	333	19.8	22.0
15	367	21.9	18.8
16	340	20.2	18.1
Gender
Female	853	50.8	48.0
Male	827	49.2	52.0
Race/ethnicity
White, non-Hispanic	779	46.4	51.8
Black, non-Hispanic	127	7.6	9.2
Hispanic	600	35.7	30.2
Asian or Pacific Islander	54	3.2	2.3
Other or multiracial	120	7.2	6.6
Tobacco user in the household
Yes	649	41.9	47.4
No	900	58.1	52.6
Smoking status
Susceptible nonsmoker	1281	76.3	74.5
Experimenter	399	23.8	25.5
Census region
Northeast	190	11.3	16.7
Midwest	305	18.2	12.9
South	601	35.8	32.0
West	584	34.8	38.4

^aAll characteristics were measured at baseline.

Ad Campaign Recall

Youth recall of The Real Cost brand name was 54.1% at the first follow-up and 80.5% at the second follow-up. Overall awareness of one or more advertisements was 92.7% at the first follow-up and 97.0% at the second follow-up. At the second follow-up, overall awareness did not differ significantly by age, gender, race/ethnicity, whether or not the youth lived with a smoker, or status as experimenter or susceptible nonsmoker. Youth awareness of specific advertisements at the first follow-up was as follows: 83.3% (Your Skin), 82.8% (Your Teeth), 81.6% (Bully), and 73.4% (So Bossy). At the second follow-up, awareness was as follows: 93.5% (Your Skin), 91.6% (Your Teeth), 74.7% (Stay in Control), and 67.0% (“7000”).

Youth Changes in Agreement With Beliefs

On average, more youth agreed with campaign-targeted belief statements at follow-ups than baseline. Table 1 presents the simple descriptive of beliefs at each wave among the target audience for the longitudinal sample (n = 1680). Agreement with beliefs increased an average of 10.4% at the first follow-up and 11.5% at the second follow-up, relative to baseline levels. Among the 14 beliefs not targeted by the campaign, agreement levels increased 0.2% at the first follow-up and decreased 3.2% at the second follow-up, relative to baseline levels. Among the 8 ambiguous beliefs, average agreement levels increased 4.6% at the first follow-up and 3.0% at the second follow-up, relative to baseline.

Association Between Exposure and Change in Outcomes

Table 3 displays data from 3 regression analyses on the association between agreement with beliefs in the past year and 3 measures of campaign exposure: self-reported frequency of exposure, total TRPs, and ad-specific TRPs. Increasing levels of exposure to The Real Cost were associated with positive changes in 5 of 8 campaign-targeted beliefs from baseline to both follow-ups. For example, youth who recalled seeing Your Teeth more often (ie, 1 unit increase on a scale of 0-4) had a 22% increase in the odds of agreeing with the related belief “If I smoke, I will lose my teeth” (odds ratio [OR] = 1.22; 95% confidence interval [CI] = 1.13-1.32). For every increase of 2000 total TRPs, the odds of agreeing with the same belief increased by 41% (OR = 1.41, 95% CI = 1.15-1.73); for every increase of 1000 ad-specific TRPs, the odds of agreeing with the same belief increased by 58% (OR = 1.58, 95% CI = 1.16-2.16). Detailed model results among the study target audience of 1680 youth are presented in Supplemental Appendix Tables B1 through B4. Supplemental model results among the full sample of 5240 youth show a similar pattern of campaign effects on beliefs (Supplemental Appendix Tables B5 through B7).

Table 3.

The Real Cost Advertising Exposure’s Influence on Campaign-Targeted Beliefs Statements.

Belief Statement Related to The Real Cost	Cumulative Total TRPs		Ad-Specific TRPs		Self-Reported Frequency of Exposure
Belief Statement Related to The Real Cost	OR	95% CI	OR	95% CI	OR	95% CI
If I smoke, I will lose my teeth^a	1.41^b	1.15-1.73	1.58^c	1.16-2.16	1.22^b	1.13-1.32
If I smoke, I will have problems with my teeth^a	0.85	0.64-1.13	0.77	0.51-1.16	0.98	0.88-1.10
If I smoke, I will get wrinkles^d	2.08^b	1.73-2.50	2.82^b	2.17-3.67	1.30^b	1.21-1.40
If I smoke, I will develop skin problems^d	1.43^b	1.19-1.71	1.62^b	1.25-2.11	1.19^b	1.10-1.29
If I smoke, I will be controlled by smoking^e	1.43^b	1.16-1.75	1.47^b	1.19-1.82	1.13^c	1.05-1.22
If I smoke, I will become addicted^e	0.85	0.70-1.02	0.84	0.69-1.03	1.01	0.92-1.11
If I smoke, I will inhale poisons^f	1.58^b	1.23-2.02	1.53	1.00-2.35	1.19^g	1.04-1.35
Cigarette ingredients are dangerous^f	1.06	0.81-1.40	1.09	0.71-1.69	1.04	0.91-1.19

Abbreviations: CI, confidence interval; OR, odds ratio; TRPs, target rating points. ^aPrimary message of The Real Cost advertisement “Your Teeth.”

^b P < .001.

^c P < .01.

^dPrimary message of The Real Cost advertisement “Your Skin.”

^ePrimary message of The Real Cost advertisements “Bully,” “So Bossy,” and “Stay in Control.”

^fPrimary message of The Real Cost advertisement “7000.”

^g P < .05.

We also examined 2 models examining the effect of exposure (cumulative total TRPs) for 14 beliefs not related to The Real Cost, controlling for the same set of covariates. Cumulative total TRPs measures were unrelated to 12 of the 14 beliefs (see Supplemental Appendix Table B8 for model results for unrelated beliefs). Two items were associated with TRPs: “If I smoke, I will lose my taste buds” (total TRP: OR = 1.41; 95% CI = 1.19-1.67) and “Smoking cigarettes is a way to help keep your weight down” (total TRP: OR = 0.72; 95% CI = 0.59-0.88).

Model results were mixed for the 8 ambiguous beliefs. Cumulative total TRPs were positively associated with 2 of the 8 beliefs: “If I smoke, I will be unattractive” (total TRP: OR = 1.34; 95% CI = 1.12-1.60; messages included in The Real Cost and truth®) and “Cigarette ingredients are disgusting” (total TRP: OR = 1.26; 95% CI = 1.01-1.56; messages included in The Real Cost and truth®; see Supplemental Appendix Table B8 for model results for ambiguous beliefs).

Figure 1 presents dose–response patterns of beliefs that were associated with campaign exposure. Postestimation predicted values showed that the average relative change in beliefs for each increase of 1000 cumulative total TRPs is 9.3% from 0 to 1000 TRPs; an additional 3.2% from 1000 to 2000 TRPs; and an additional 2.2% from 2000 to 3000 TRPs.

Figure 1.

Estimated dose–response curves for effects of cumulative total TRPs on belief statements.

Discussion

This study found that exposure to The Real Cost influenced campaign-targeted beliefs about tobacco use among youth nationwide. Similar to research findings from previous smoking prevention campaigns,^29,30 positive shifts in campaign beliefs were demonstrated within the context of relatively high levels of baseline agreement with tobacco-related beliefs. The proportion of youth that hold campaign-positive beliefs at follow-up increased by 11.5% on average relative to baseline. Multivariate models show a consistent pattern of effects in which campaign-targeted beliefs are more likely to change than nontargeted beliefs as a result of campaign exposure. After accounting for the potential effects of other tobacco education advertising on air, dose–response effects are observable in models examining potential exposure measures using TRPs and self-reported frequency of exposure to advertisements. Given the widespread reach and heavy media dose of The Real Cost, models examining beliefs unrelated to campaign advertising were informative to diminish concerns regarding alternative explanations for campaign-positive findings. We found no consistent simple change or dose–response patterns between campaign exposure and beliefs not targeted by The Real Cost advertisements or systematic differences in agreement levels for campaign-targeted and nontargeted beliefs at baseline.

This study adds to the body of research demonstrating that state and national tobacco education campaigns are effective. A recent study using this nationally representative sample found that youth’s self-reported exposure to The Real Cost was associated with a reduction in smoking initiation from 2014 to 2016.³¹ The current study findings suggest that The Real Cost changed tobacco-related beliefs that may have led to these positive shifts in behavioral outcomes based on theoretical and empirical evidence.^32,33 Future research will examine whether The Real Cost exposure is associated with other behavioral effects and explore changes in campaign-related beliefs and behaviors within population subgroups and among black and Hispanic youth in particular.

This study has several limitations. First, maturation effects are not accounted for in the pre–post study design. However, research suggests that as a youth cohort ages, those youth are at greater risk of smoking susceptibility and initiation,²⁷ which would suggest increased pro-tobacco beliefs over time. The current study showed a pattern of null effects across nontargeted and ambiguous items, which provides evidence to counter this potential bias. Second, although panel conditioning could affect the results, this bias would be expected to result in similar conditioning patterns across all outcomes, which is inconsistent with observed patterns. Third, sample attrition over time may have resulted in biases among youth at the second follow-up. Overall, there were few differences between youth in the completed sample and those lost to follow-up; however, youth who were susceptible to smoking at baseline dropped out of the study at a higher rate than those who were not. These youth may represent the more vulnerable subgroup that is more likely to smoke. Survey weighting procedures account for this differential attrition. Fourth, there may be measurement error as the result of the self-reported nature of the data and the use of single items to assess the beliefs and some covariates. Fifth, although the model controls for youth’s exposure to other tobacco-related media campaigns, this might not fully account for the independent or synergistic effects of the other campaigns. Sixth, a majority of respondents completed the survey in their home in the presence of an interviewer and a parent or guardian. Although the survey was self-administered via laptop, the presence of the interviewer and parent or guardian may have influenced self-reports, particularly for sensitive questions.

Smoking has been described as a pediatric disease because most tobacco dependence begins during adolescence with cigarette experimentation.³⁴ The Real Cost focuses on harmful effects of tobacco use that are salient to adolescents, such as the cosmetic effects of tobacco use. However, the campaign’s true potential impact may occur well beyond adolescence. Youth-focused campaigns to prevent smoking and disrupt progression to established smoking may have long-term effects on a cohort’s rate of tobacco-related disease, such as cardiovascular disease, respiratory disease, and cancer.^10,27 Moreover, reductions in smoking prevalence among youth today may result in lower rates of future smoking-attributable mortality, health-care costs, and lost workplace productivity.^35
–37 Media campaigns such as The Real Cost may influence a generation of US youth to abstain from tobacco use throughout their lives and avoid its devastating health consequences.

So What?

What is already known on this topic?

A substantial body of research finds that media campaigns are effective at reducing youth smoking prevalence, preventing smoking initiation, and disrupting progression to established smoking.

What does this article add?

This study is the first in a decade to examine the effects of a youth-specific national media campaign on teens. Using a strong nonexperimental longitudinal design, we found that youth’s exposure to the Food and Drug Administration’s The Real Cost influenced their beliefs about tobacco use. This study expands the knowledge base on the measurement of dose–response relationships for media campaigns, specifically for population-level beliefs associated with subsequent change in tobacco use and initiation. Further, our analysis of beliefs that were categorized by independent coders as either related or unrelated to campaign advertisements is a unique contribution to this research area, providing a new approach to examining effects in the context of a media environment that includes multiple other public tobacco-related campaigns.

What are the implications for health promotion practice or research?

Targeted mass media campaigns can change population-level perceptions of tobacco-related harms and should continue to be promoted to youth at risk of smoking. Sustained campaigns such as The Real Cost may influence a generation of US youth to abstain from tobacco use throughout their lives and avoid its devastating health consequences. Study findings may also provide insights that are useful in the design of future public education campaigns for youth.

Supplemental Material

Supplemental Material, TheRealCost_1_31_2017_Appendix_A - Effect of a National Tobacco Public Education Campaign on Youth’s Risk Perceptions and Beliefs About Smoking

Supplemental Material, TheRealCost_1_31_2017_Appendix_A for Effect of a National Tobacco Public Education Campaign on Youth’s Risk Perceptions and Beliefs About Smoking by Jennifer C. Duke, Matthew C. Farrelly, Tesfa N. Alexander, Anna J. MacMonegle, Xiaoquan Zhao, Jane A. Allen, Janine C. Delahanty, Pamela Rao, and James Nonnemaker in American Journal of Health Promotion

Supplemental Material

Supplemental Material, TheRealCost_1_31_2017_Appendix_B - Effect of a National Tobacco Public Education Campaign on Youth’s Risk Perceptions and Beliefs About Smoking

Supplemental Material, TheRealCost_1_31_2017_Appendix_B for Effect of a National Tobacco Public Education Campaign on Youth’s Risk Perceptions and Beliefs About Smoking by Jennifer C. Duke, Matthew C. Farrelly, Tesfa N. Alexander, Anna J. MacMonegle, Xiaoquan Zhao, Jane A. Allen, Janine C. Delahanty, Pamela Rao, and James Nonnemaker in American Journal of Health Promotion

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This article was funded by the Food and Drugs Administration Center for Tobacco Products (contract HHSF223201310001B).

Supplemental Material

Supplementary material for this article is available online.

References

White

Warne

Spittal

Durkin

Purcell

Wakefield

. What impact have tobacco control policies, cigarette price and tobacco control programme funding had on Australian adolescents’ smoking? Findings over a 15-year period. Addiction. 2011;106(8):1493–1502.

Farrelly

Davis

Haviland

Messeri

Healton

. Evidence of a dose-response relationship between “truth” antismoking ads and youth smoking prevalence. Am J Public Health. 2005;95(3):425–431.

Sly

Hopkins

Trapido

Ray

. Influence of a counteradvertising media campaign on initiation of smoking: the Florida “truth” campaign. Am J Public Health. 2001;91(2):233–238.

Farrelly

Nonnemaker

Davis

Hussin

. The Influence of the National truth campaign on smoking initiation. Am J Prev Med. 2009;36(5):379–384.

Weiss

Cen

Schuster

. Longitudinal effects of pro-tobacco and anti-tobacco messages on adolescent smoking susceptibility. Nicotine Tob Res. 2006;8(3):455–465.

Siegel

Biener

. The impact of an antismoking media campaign on progression to established smoking: results of a longitudinal youth study. Am J Public Health. 2000;90(3):380–386.

Allen

Duke

Davis

Kim

Nonnemaker

Farrelly

. Using mass media campaigns to reduce youth tobacco use: a review. Am J Health Promot. 2015;30(2):e71–e82.

Institute of Medicine. Ending the Tobacco Problem: A Blueprint for the Nation. Washington, DC: The National Academies Press, Institute of Medicine of the National Academies; 2007.

National Cancer Institute. The Role of the Media in Promoting and Reducing Tobacco Use. Tobacco Control Monograph No. 19. Bethesda, MD: US Department of Health and Human Services, National Institutes of Health, National Cancer Institute; 2008.

10.

Centers for Disease Control and Prevention. Best Practices for Comprehensive Tobacco Control Programs—2014. Atlanta, GA: US 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; 2014.

11.

Fishbein

. A consideration of beliefs and their role in attitude measurement. In: Fishbein

, ed. Readings in attitude theory and measurement. New York, NY: Wiley; 1967:257–266.

12.

Ajzen

. The Theory of Planned Behavior. Organ Behav Hum Decis Process. 1991;50(2):179–211.

13.

Hersey

Niederdeppe

Evans

. The theory of “truth”: how counterindustry campaigns affect smoking behavior among teens. Health Psychol. 2005;24(1):22–31.

14.

Scheier

Grenard

Holtz

. An empirical assessment of the Above the Influence advertising campaign. J Drug Educ. 2011;41(4):431–461.

15.

Duke

Alexander

Zhao

. Youth’s awareness of and reactions to the real cost national tobacco public education campaign. PLoS One. 2015;10(12):e0144827.

16.

Davis

Nonnemaker

Duke

Farrelly

. Perceived effectiveness of cessation advertisements: the importance of audience reactions and practical implications for media campaign planning. Health Commun. 2013;28(5):461–472.

17.

Duke

Allen

Eggers

Nonnemaker

Farrelly

. Exploring differences in youth perceptions of the effectiveness of e-cigarette television advertisements. Nicotine Tob Res. 2016;18(5):1382–1386.

18.

Alvaro

Crano

Siegel

Hohman

Johnson

Nakawaki

. Adolescents’ attitudes toward antimarijuana ads, usage intentions, and actual marijuana usage. Psychol Addict Behav. 2013;27(4):1027–1035.

19.

Dillard

Weber

Vail

. The relationship between the perceived and actual effectiveness of persuasive messages: a meta-analysis with implications for formative campaign research. J Commun. 2007;57:613–631.

20.

Farrelly

Davis

Duke

Messeri

. Sustaining ‘truth’: changes in youth tobacco attitudes and smoking intentions after 3 years of a national antismoking campaign. Health Educ Res. 2009;24(1):42–48.

21.

Hornik

. Public health communication: Evidence for behavior change. London, NJ: Lawrence Erlbaum Associates; 2002.

22.

Mowery

Farrelly

Haviland

Gable

Wells

. Progression to established smoking among US youths. Am J Public Health. 2004;94(2):331–337.

23.

Pierce

Choi

Gilpin

Farkas

Merritt

. Validation of susceptibility as a predictor of which adolescents take up smoking in the United States. Health Psychol. 1996;15(5):355–361.

24.

Southwell

Barmada

Hornik

Maklan

. Can we measure encoded exposure? Validation evidence from a national campaign. J Health Commun. 2002;7(5):445–453.

25.

Wakefield

Spittal

Yong

Durkin

Borland

. Effects of mass media campaign exposure intensity and durability on quit attempts in a population-based cohort study. Health Educ Res. 2011;26(6):988–997.

26.

Farrelly

Duke

Davis

. Promotion of smoking cessation with emotional and/or graphic antismoking advertising. Am J Prev Med. 2012;43(5):475–482.

27.

US Department of Health and Human Services. Preventing Tobacco Use Among Youth and Young Adults: A Report of the Surgeon General. Atlanta, GA: US 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; 2012.

28.

Stephenson

Hoyle

Palmgreen

Slater

. Brief measures of sensation seeking for screening and large-scale surveys. Drug Alcohol Depend. 2003;72(3):279–286.

29.

Farrelly

Healton

Davis

Messeri

Hersey

Haviland

. Getting to the truth: evaluating national tobacco countermarketing campaigns. Am J Public Health. 2002;92(6):901–907.

30.

Sly

Heald

Ray

. The Florida “truth” anti-tobacco media evaluation: design, first year results, and implications for planning future state media evaluations. Tob Control. 2001;10(1):9–15.

31.

Farrelly

Duke

Nonnemaker

. Association between The Real Cost media campaign and smoking initiation among youths—United States, 2014-2016. MMWR Morb Mortal Wkly Rep. 2017;66(2):47–50.

32.

Brennan

Gibson

Momjian-Kybert

Liu

Hornik

. Online supplementary materials: promising themes for antismoking campaigns targeting youth and young adults. Tob Regul Sci. 2017;3(1):29–46.

33.

Hornik

Woolf

. Using cross-sectional surveys to plan message strategies. Soc Mark Q. 1999;5(2):34–41.

34.

Hilts

. F.D.A. head calls smoking a pediatric disease. New York Times. March 9, 1995.

35.

Centers for Disease Control and Prevention(CDC). Smoking-attributable mortality, years of potential life lost, and productivity losses—United States, 2000-2004. MMWR Morb Mortal Wkly Rep. 2008;57(45):1226–1228.

36.

Bishop

Kennedy

Simpson

Pechacek

. Annual healthcare spending attributable to cigarette smoking: an update. Am J Prev Med. 2015;48(3):326–333.

37.

Carter

Abnet

Feskanich

. Smoking and mortality—beyond established causes. N Engl J Med. 2015;372(7):631–640.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.70 MB

0.05 MB