- Email: [email protected]
Media alcohol advertising with drinking behaviors among young adolescents in Taiwan
Accepted Manuscript Title: Media alcohol advertising with drinking behaviors among young adolescents in Taiwan Authors: Chuan-Yu Chen, Hsueh-Yu Huang, Fang-Yi Tseng, Yu-Chan Chiu, Wei J. Chen PII: DOI: Reference:
S0376-8716(17)30232-6 http://dx.doi.org/doi:10.1016/j.drugalcdep.2017.03.041 DAD 6448
To appear in:
Drug and Alcohol Dependence
Received date: Revised date: Accepted date:
8-9-2016 31-3-2017 31-3-2017
Please cite this article as: Chen, Chuan-Yu, Huang, Hsueh-Yu, Tseng, FangYi, Chiu, Yu-Chan, Chen, Wei J., Media alcohol advertising with drinking behaviors among young adolescents in Taiwan.Drug and Alcohol Dependence http://dx.doi.org/10.1016/j.drugalcdep.2017.03.041 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 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.
Media alcohol advertising with drinking behaviors among young adolescents in Taiwan
Chuan-Yu Chena,b,c*, Hsueh-Yu Huanga,c, Fang-Yi Tsenga, Yu-Chan Chiud, Wei J. Chene
a
Institute of Public Health, National Yang-Ming University, Taipei, Taiwan
b
Center of Neuropsychiatric Center, National Health Research Institutes, Miaoli, Taiwan
c
Children and Family Research Center, National Taiwan University, Taipei, Taiwan
d
Department of Bio-industry Communication and Development, National Taiwan University, Taipei, Taiwan
e
Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan
Correspondence: Chuan-Yu Chen, PhD, Professor National Yang-Ming University Institute of Public Health No.155, Sec.2, Linong Street, Taipei, 112 Taiwan Medical Building II, Rm 210 Email: [email protected] Tel: 886-02-2826-7344 Fax: 886-02-2822-4850
Highlight
TV, in-store display, and website are common channels for alcohol ads exposure.
Exposure to TV alcohol ads may increase drinking initiation in early adolescence.
Exposure to ads on the website and radio may elevate occasional drinking. 1
ABSTRACT Objectives: To investigate potential effects of alcohol ads in six major marketing channels on drinking behaviors among young adolescents in Taiwan. Methods: The data were derived from the Alcohol-Related Experiences among Children study. The baseline sample was comprised of 1926 seventh-eighth graders from 11 public middle schools in Taipei in 2010; follow-up was conducted one year later (follow-up rate=97%). Information concerning individual sociodemographics, family characteristics, exposure to media portrayals of drinking and alcohol ads on major marketing channels, and drinking experience was collected through web-based self-administered questionnaires. Complex survey analyses were used to evaluate the association estimates, with stratification by prior drinking experiences in childhood. Results: Television, in-store displays, and websites are the three most common marketing channels for young adolescents to report past-month alcohol advertising exposure. With statistical adjustment for potential confounders and six market channels, exposure to alcohol ads on television was associated with subsequent increased drinking initiation (adjusted odds ratio [aOR]=2.62; 95% CI=1.14–6.02). For those who have initiated alcohol use in childhood, the exposure to ads on the web (aOR=1.50; 95% CI=1.04–2.15) and radio (aOR=2.58; 95% CI=1.60–4.15) may elevate subsequent risk of occasional drinking. Exposure to media drinking portrayals was not related to subsequent drinking behaviors in this sample. Conclusions: Our results demonstrated that the effects of alcohol advertising on drinking behaviors in early adolescence may differ by marketing channels. Preventive strategies targeting underage drinking should consider restraining marketing channels (e.g., websites and radio) from certain advertising content and placement.
Keywords: adolescent, alcohol advertising, marketing channel 1. Introduction Alcohol consumption and harmful drinking have become a norm in young people, particularly in high-income countries (World Health Organization, 2014). More than 4 in 5 European students aged 15–16 2
years had drunk alcohol on at least one occasion in their lifetimes, and roughly 30% of 8th, 10th, and 12 graders in the US had ever been drunk (Hibell et al., 2012; Johnston et al., 2015). In Taiwan, school surveys with national representativeness found the lifetime prevalence rate of alcohol drinking among 8th graders has increased from 56% in 2006 to 65% in 2012 (Health Promotion Administration, 2012). Early-onset alcohol drinking is associated with an increased risk of health problems and developmental impairment (Masten et al., 2008; Windle et al., 2008). Data from the World Health Organization Global Burden of Disease study indicated that alcohol is the leading risk factor for death and disability-adjusted life years among people aged 15–24 years (Mokdad et al., 2016). To provide targets for preventing early-onset alcohol initiation and continued alcohol drinking, it is imperative to identify modifiable factors shaping the transition of alcohol experiences in the first two decades of one’s life (Masten et al., 2008; Windle et al., 2008). With their rising penetration into daily lives of youngsters, the role of media on shaping youngsters’ drinking behaviors has been gradually recognized (Brown and Witherspoon, 2002). The media are hypothesized to influence the drinking behaviors through several processes, among which alcohol advertising is especially notable given its expansion and diversity (Anderson et al., 2009; Casswell, 2004; Engels et al., 2009; Jernigan et al., 2017; Robinson et al., 1998; Sargent et al., 2006). Globally, the top nine alcohol beverage companies have spent at least €2.5 billion on traditional media advertising in 2004 (Anderson, 2007). The advertising industry self-regulates the content of alcohol advertising, ostensibly to protect adolescents against the exposure to alcohol advertising. Nevertheless, the guidelines concerning the use of content appealing to minors are often violated (Babor et al., 2013), and young people are greatly exposed to alcohol marketing (Jernigan, 2010). Alcohol companies spent more than $8 billion on alcohol advertising on television between 2001 and 2009 in the US, and exposure to alcohol advertising on television among underage youth increased by 71% during that same period (The Center on Alcohol Marketing and Youth, 2010). The weekly exposure to retail-ad was estimated at two thirds for middle school students in California (Hurtz et al., 2007). In Britain, a report shows that 60% of young people were exposed to alcohol advertising on a daily basis through television and 25% through magazines (Atkinson et al., 2011). 3
Although variation exists in measure and coverage of marketing channel (e.g., media platform-specific or composite measure), in assessment timeframe of exposure (e.g., past-year or past-month), and in outcome measure (e.g., alcohol initiation or binge drinking), evidence from longitudinal studies has shown that exposure to alcohol advertising in traditional media platforms, including television (Robinson et al., 1998; Ross et al., 2014), radio (Collins et al., 2007), movies/films (Sargent et al., 2006), magazines (Ellickson et al., 2005), and outdoors outlets (e.g., billboards or convenience stores)(Pasch et al., 2007), is a significant predictor for underage behavior. Recently, with easier access to internet and mobile phone in young population, attention has also been directed to emerging media outlets (Atkinson et al., 2011; Lobstein et al., 2017). To illustrate, a cross-sectional study in Australia reported that exposure to alcohol advertising on the internet is associated with a 36% increase in the chance of past-month drinking behavior (Jones and Magee, 2011). Another study, using the average score to reflect advertisement exposure in the past year through television, newspapers or magazines, outdoors billboards, alcohol portraits in TV/movies, and the discussion on the internet, found that one point increase in alcohol ads exposure elevates alcohol initiation by 68% and persistence by 44% among old adolescents in Taiwan (Chang et al., 2014). The information conveyed in alcohol advertising often generates positive expectancies toward alcoholic beverages (Chen et al., 2016; Fleming et al., 2004; Ross et al., 2014); several studies also showed that higher exposure to alcohol advertising is associated with increased drinking intention or favorable attitudes among children and adolescents (Collins et al., 2007; Grube and Wallack, 1994; Unger et al., 2003). These pro-alcohol cognitive, intentional, and attitude attributes are improtant precursors for adolescents’ subsequent drinking behaviors and problems. A recent longitudinal study utilized masked ad images (i.e., showing advertising images with the logo removed) to measure contact frequency and brand recall for alcohol advertisements among sixth graders in Germany. The study concluded that 35% of total effects on the
connection
between
alcohol
adverting
exposure
and
alcohol
initiation
is
due
to
positive attitudes toward alcohol (Morgenstern et al., 2011). To date, research on the connection between alcohol advertising and underage drinking behaviors suffers from two important limitations. First, earlier studies often examined advertising effects either with a focus on traditional marketing channels (e.g., television or movie) or with a pooled measure (Chang et al., 4
2014; Snyder et al., 2006). The prospective evidence pertaining to certain emerging marketing channels awaits accumulation (McClure et al., 2016). Second, with few exception (Chang et al., 2014), existing research on the longitudinal associations linking ads exposure and alcohol drinking has been mostly limited in the high-income countries with mature alcohol markets. There is a need to examine advertising effects on young populations in the emerging alcohol markets wherein alcohol companies have gradually increased their marketing expenditures over the past decade. To address the abovementioned research gaps, we turned to a longitudinal study that prospectively followed up a cohort of seventh graders (aged 13–14 years old) to study prevalence of alcohol advertising exposure through media platform in young adolescents in Taiwan. Given that the risk factors for drinking problem evolution are stage-dependent (e.g., initiation and continued drinking)(Anthony and Chen, 1999; Chang et al., 2014; Ellickson et al., 2005; Gordon et al., 2010; McClure et al., 2016), the present study further assessed the relationship linking each media platform with subsequent alcohol initiation and occasional drinking behaviors through early adolescence. 2. Methods 2.1. Study Design and Population The data were derived from the Alcohol-Related Experiences among Children (AREC) study (Lee et al., 2015). Taiwan has nine years of compulsory education: six years of elementary school and three years of middle school. Utilizing the multistage sampling procedure, in late 2009 we first randomly selected 11 of the 73 public middle schools in Taipei (which more than 90% of the city’s middle school students attend); all seventh graders in the sampled schools were eligible to participate in the study. Upon receiving positive consent from both guardians (e.g., parents) and the students themselves, the students were asked to fill out the self-administered web-based questionnaire during regular class hours (baseline response rate=55%) while in-class teachers were absent. The baseline sample comprised 1926 seventh and eighth graders (aged 13-14 years) in 2010; the second assessment was conducted in the ninth grade, with an average follow-up interval of 13 months (n=1870; follow-up rate=97%). All study procedures were approved by the institutional review board of the National Health Research Institutes (EC0980605). The present study excluded seven students with missingness in the variables of major interest, resulting in an analytic sample 5
of 1863. 2.2. Measures The questionnaire mostly took students less than 45 minutes to complete. The questionnaire included individual sociodemographic and pubertal characteristics, family structure and socioeconomic status, family and peer drinking behaviors, exposure to alcohol advertising through six types of media, and individual drinking behaviors. 2.2.1 Outcome variables. At both waves of assessment, alcohol-drinking experiences were assessed by a series of questions (Lee et al., 2015). The first one was “Not including a sip or alcohol in food, have you ever drunk any alcoholic beverages in your lifetime?” If the response was positive, additional questions were asked, including drinking experience in the 12 months preceding the survey. In the present study, we defined “incident drinkers” as those who responded as (i) having never drank at the first assessment and (ii) having drunk on at least one occasion in the past 12 months at the follow-up assessment. For the youngsters who reported ever drank alcohol at baseline survey, we used the second quartile of past-year drinking occasions in the 9th grade as the cut-off point (see Table 1) to define “occasional drinking”—having drunk on four or more occasions in the 12 months preceding the follow-up assessment. 2.2.2 Independent variables. Exposure to alcohol advertising at the baseline was assessed by the question “In the past month, have you ever seen any alcohol advertising or promotional messages in the following marketing channel?” This was asked for nine types of marketing channels: television, movies, newspapers, magazines, radio, stores (e.g., convenience stores and supermarkets), billboards/posters/flyers, internet (e.g., websites), and e-mail; if positive, the frequency of exposure was further queried using four categories: 1–7 days, 8–14 days, 15–29 days, and 30 or more days. In the present analyses, we regrouped the frequency variable of alcohol advertising exposure into 0 days, 1–14 days, and 15 or more days. Considering what youngsters had daily access to and the type of media alcohol industries tend to use in Taiwan (Center for Media Literacy, 2008), we decided to focus on the six major types of marketing channels for alcohol advertising: television, newspapers, magazines, radio, Internet, and in-store displays. 2.2.3 Covariates. Baseline variables thought to be associated with alcohol advertising exposure and alcohol drinking in Taiwan were taken into account as important confounding factors (Lee et al., 2015; Ting 6
et al., 2015). These included gender, living with parents (living with one or none vs. living with both), parental education (both/either high school or below vs. both college or above), monthly allowance (NTD 0– 499 vs. 500 or above; 1 USD≈31 NTD), parental drinking (none vs. either one drinking), best friend’s drinking (none vs. at least one), and future intent to drink (definitely/probably not vs. definitely/probably yes). Other than alcohol ads, youths’ media exposure to alcohol-related images was assessed by the question (Engels et al., 2009):“In the past month, on how many days have you seen drinking portrayals on television, in a movie, or on the Internet?” 2.3. Statistical analyses Since the samples were drawn via multistage probability sampling procedures, the present study used standard survey analysis to take into account sampling parameters (e.g., sampling weights to compensate for variation in sample-selection probability and within school non-independence). We first examined the distribution of individual sociodemographics, family and peer drinking, and past-month exposure to alcohol advertising with stratification by baseline lifetime drinking experience. Next, we adopted survey logistic regression to estimate the association linking media alcohol advertising with incident drinking and continued occasional drinking. In Models 1 and 2, the market channel was entered one at a time: we first included individual sociodemographic characteristics and family and peer drinking in Model 1, followed by the addition of the exposure to media portrayals of drinking and the intent to drink in the seventh grade (Collins et al., 2007; Grube and Wallack, 1994). Finally, we entered all market channels simultaneously in Model 3. All statistical analyses were carried out by SAS 9.3 survey procedures.
3. Results Approximately 45% of the surveyed students had drank alcoholic beverages at least once at baseline, 70% had at least one drinking parent, and 79% had recent exposure through the media to portrayals of drinking (Table 1). Alcohol-experienced young adolescents, when compared with their alcohol-naïve peers, were more likely to be male, not live with both parents, have a higher monthly allowance, have drinking parents, have at least one drinking best friend, have exposure to drinking portrayals on television, in movies, and on the internet, as well as have positive future drinking intent (p<0.05). Among the alcohol naïve at 7
baseline, 17% have initiated alcohol drinking by 9th grade. An estimated 46% of the alcohol-experienced had drank on at least one occasion in the year preceding the ninth grade survey , and the second quartile of cumulative drinking occasion was three. Among the six major marketing channels in Taiwan, television was the most common one through which seventh graders are exposed to alcohol advertising (>88%), followed by in-store displays and websites (see Figure1). In general, the proportion of past-month exposure to alcohol advertisements at baseline survey for the alcohol experienced (AE) was slightly higher than their alcohol-naïve (AN) counterparts. Drinking experience-related differences in exposure to alcohol advertising were also manifested in higher exposure frequency (i.e., 15 days or above), particularly through in-store displays, websites, and magazines (P<0.001). Table 2 presents the prospective association linking exposure to alcohol advertising with incident alcohol drinking. With statistical adjustment for sociodemographic covariates (see Model 1), television was the only significant market channel linked with increased drinking initiation (1–14 days: aOR=3.20; 95% CI=1.54–6.63; 15 or more days: aOR=2.35; 95% CI=1.22–4.55). However, after taking exposure to media portrayals of drinking and drinking intention into account (Model 2), the exposure to ads in stores (e.g., 15 or more days: aOR=1.47; 95% CI=1.07–2.03) and on newspaper (e.g., 1-14 days: aOR=0.77; 95% CI=0.63-0.94) became significant. Finally, with all marketing channels added into the model (see Model 3), the baseline exposure to alcohol ads on television was linked with increased drinking initiation (e.g., 1–14 days: aOR=2.62; 95% CI=1.14–6.02), whereas for newspapers the risk was lowered (e.g., 1–14 days: aOR=0.65; 95% CI=0.49–0.87). Having at least one drinking best friend (aOR=1.70) or having higher monthly allowance (aOR=1.56) was found to slightly increase the risk of alcohol initiation. For the alcohol-experienced 7th graders, having exposure to alcohol advertising through websites or radio slightly elevated the risk of occasional drinking in the subsequent year (see Table 3, Model 1). The risk estimates for the website (1–14 days: aOR=1.50; 95% CI=1.04-2.15) and the radio (15 or more days: aOR=2.58; 95% CI=1.60–4.15) did not have appreciable changes once the exposure to media portrayals of drinking, drinking intention, and all market channels were added into model (see Model 3). In addition, the intent to drink at baseline (aOR=2.28; 95% CI=1.72–3.02) appeared to the strongest predictor. 8
4. Discussion The present study is one of the first few to examine the prospective relationship linking media alcohol advertising with alcohol initiation and continuation among young adolescents in emerging markets. We found that television, stores, and websites were the three most common media platforms through which youngsters recalled alcohol ads exposure; alcohol-experienced youngsters reported a higher frequency of exposure to alcohol ads. Independent of gender, living with parents, parental education, parental drinking, best friend’s drinking behavior, monthly allowance, drinking intention, and exposure to media drinking portrayals at baseline, having exposure to alcohol ads on television were associated with increased risk of alcohol initiation, whereas exposure to alcohol ads on websites or radio were associated with subsequent elevated risk of occasional drinking. The estimated rates of past-month alcohol ad exposure through television, magazines, and websites in our young population were slightly lower than those in the United Kingdom (i.e., 97%, 80%, and 84%, respectively)(Atkinson et al., 2011). However, our estimates of exposure to alcohol ads through magazines, websites, and in-store displays were considerably higher than the estimates in Australia (71%, 53%, and 75%, respectively)(Jones and Magee, 2011). In addition to true cross-country variation, the observed differences in exposure to alcohol ads might be partially explained in the context of population characteristics (e.g., 12–18 years old vs. 13–14 years old), survey period, and assessment timeframe (e.g., lifetime vs. past month). Although alcohol companies claimed to have strengthened their self-regulatory standards, our data revealed that exposure to both alcohol advertising on major marketing channels and media portrayals of drinking appeared to be very common (or to be the norm) for young adolescents in Taiwan. Of particular importance are higher alcohol advertising exposure rates among those seventh graders who have drank alcohol in their childhood, regardless of media outlets (Anderson et al., 2009). These findings may be the results of several pathways: (i) the alcohol-experienced children were more likely to notice or be attracted to alcohol advertising; (ii) the alcohol-experienced children were more likely to recognize or remember alcohol advertising; and (iii) the alcohol-experienced children were more likely to become chronically exposed to alcohol advertising to develop brand allegiance (Anderson et al., 2009; Macklin, 1994; McClure et al., 2013). This observation implicated that the cognitive and affective 9
processing of alcohol advertising is probably not identical across subgroups defined by drinking experiences and reinforced the possibility that advertising effects are not static and may change as drinking experience evolves (Jernigan et al., 2017; McClure et al., 2013). Consistent with prior research highlighting the importance of alcohol advertising on underage drinking behaviors (Anderson et al., 2009; Chang et al., 2014; Collins et al., 2007; Grube and Wallack, 1994), our results suggested that the association may vary by media platform. Although alcohol industries integrated their commercial communication across marketing channels, the association estimates from multivariate models yield some insights for marketing channel-differential impact over the course of alcohol drinking behaviors (Ellickson et al., 2005). In Taiwan, television viewing and convenience stores visits are a common part of daily life. In the case of television, although there has been time restriction on alcohol advertising (e.g., 9:00 p.m.-6:00 a.m.) to protect young people from exposure, the exposure was still high because nearly 60% of children watched television after 9 p.m. (Chen et al., 2016; Ross et al., 2013). The frequent exposure to alcoholic beverage images may normalize drinking behaviors and elevate positive alcohol expectancies, consequently predisposing some youngsters to alcohol initiation (Chen et al., 2016; Fleming et al., 2004; McClure et al., 2013). We also noted the risk of alcohol initiation was not in a dose-response relationship with the days of televised advertisement exposure. Future research is needed to elaborate time, contextual, and advertisement variables, such as exposure timing (e.g., weekday/weekend), recessive advertisement in television programs (e.g., idol dramas), and advertising content/brands, to delineate the non-linearity relationship (Padon et al., 2016; Ross et al., 2015). For the alcohol-experienced adolescents, alcohol advertising on marketing channels such as radio and websites can provide information pertaining to brand, price, and point of sale, which may facilitate the development of drinker identity and increase drinking occasions (McClure et al., 2013). Additionally, some of the popular music aired involved elements associated with alcohol brand and drinking context (Primack et al., 2012), which is particularly common during summer recess and holiday seasons in Taiwan. Although exposure to newspaper alcohol advertising was linked with a significant increase in alcohol initiation by crude association analyses, multivariate analyses unexpectedly found an inverse association. This observation might be partly due to three reasons: first, those who read newspapers may represent a 10
subsample of adolescents who generally endorsed a lower risk of alcohol initation through early adolescence (e.g., coming from families with higher socioeconomic status)(Center for Media Literacy, 2008). Second, alcohol-related news often features negative or disastrous events (e.g., sexual assaults, family violence, and drunk driving). Third, it is possible that the newspaper advertising is more likely to be price than lifestyle, which may not attractive or appealing to youth (Padon et al., 2016). Another interesting finding is that the salient effects of drinking intent only appeared in continued drinking. A series of post-hoc Sobel tests analyses suggested that “drinking intent” probably played a mediating role in the process linking website and radio alcohol advertising with drinking on four or more occasions. Although the intent to drink and exposure to alcohol advertising were measured at the same time, it is possible that certain advertising information delivered through radio or websites may foster greater marketing receptivity and reinforce the desire to drink (Collins et al., 2007; Fleming et al., 2004; McClure et al., 2013; Morgenstern et al., 2011). 4.1. Limitations and Strengths This study has four main limitations. First, the measure of exposure to alcohol advertising was limited to only the month preceding the survey, which may not be representative of a youth’s cumulative or past-year experience; however, asking young respondents to recall such exposure over a longer time frame is less likely to produce satisfactory validity. To reflect a longer period of cumulative exposure, some objective methods have been implicated (e.g., adstock or the Nielsen data)(Chen et al., 2016; Morgenstern et al., 2011; Ross et al., 2014). Second, because our data were collected at two time points, the causal inference regarding the connection between alcohol advertising and drinking behaviors may be affected if reciprocal causations operated in between. Nevertheless, we have carried out the analyses with stratification by prior drinking experience to reduce possible bias. Future studies with data from multiple time points with a larger number of alcohol-naïve children through early adolescence could help delineate intertwined relationships. Third, the relatively small sample size may pose a threat to estimation precision, which is particularly true when the incident drinking is uncommon. Finally, because the concept and definition of “drink (amount)” is not widely used in the study population, it is not possible to assess problematic alcohol consumption through the measure of “drink” (e.g., binge drinking). The approach that assessed frequency of alcohol consumption by “cumulative drinking occasions” and defined occasional drinkers on the basis of the sample-oriented quartile 11
may warrant further attention in measure validation. Similarly, our sample represents early adolescents in a metropolitan city in Taiwan, the generalizability to other societies may be restrained. Despite these limitations, our study has several strengths. First, our assessment of marketing channels is more comprehensive than that in most published reports. The exposure-frequency measure for alcohol advertising on each market channel also provides more detailed information than the binary responses commonly used. In addition, the prospective nature of the present study may build better temporality. 4.2.
Conclusions
This investigation, to our knowledge, is one of the few to report the prospective relationships linking a rather comprehensive list of alcohol marketing channels with drinking behaviors among young adolescents in emerging markets. Although current regulations establish standards for the broadcasting of alcohol advertising messages in Taiwan, these standards do not cover all media (e.g., social media) and the regulations are loosely enforced in certain media (e.g., pirate radio). For the internet, age verification screen should be placed before alcohol advertisement is broadcasting in the gateway websites (such as Yahoo and MSN) or on video sites (such as YouTube). Finally, given how international alcohol industries have extended alcohol sales into emerging markets, future work is warranted to monitor the content, form, placement, and intensity of alcohol advertising on traditional (e.g., television and radio) and emerging (e.g., websites and social media) marketing channels, with the aid of certain youth exposure indicators (e.g., gross rating points and over exposure)(Jernigan, 2011; Morgenstern et al., 2011).
Author Disclosures
Author Contributions CYC was responsible for the development of the research theme, the analysis, and the drafting of the manuscript. HYH assisted the analyses and data interpretation. FYT helped with the field work and data analyses. YCC instructed result interpretation and revised the draft. WJC assisted study design and draft revision. All authors have critically reviewed the content and approved the final version submitted for 12
publication.
Role of Funding Source Nothing declared.
Conflict of Interest The authors declare no conflict of interest with respect to the conduct or content of this work.
Acknowledgement This work was supported by grants from the Ministry of Science and Technology (95-2314-B-400-009-MY3, 98-2314-B-010-038-MY3, 101-2628-B-010-004-MY3, and 104-2314-B-010 -008 -MY3 to C. Y. Chen) the National Health Research Institutes (MDPP04-014), and a grant from the Ministry of Education of Taiwan ROC, Aim for the Top University Plan to CYC. The funding agencies had no role in the study design; in the collection, analysis, or interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
13
References Anderson, P., 2007. The Impact Of Alcohol Advertising: ELSA project report on the evidence to strengthen regulation to protect young people. National Foundation for Alcohol Prevention. Utrecht. Anderson, P., de Bruijn, A., Angus, K., Gordon, R., Hastings, G., 2009. Impact of alcohol advertising and media exposure on adolescent alcohol use: A systematic review of longitudinal studies. Alcohol Alcohol. 44, 229-243. Anthony, J.C., Chen, C.Y., 1999. Epidemiology of drug dependence. In: Textbook of Substance Abuse Treatment. American Psychiatric Press Inc, Washington DC, pp.55-72. Atkinson, A., Elliot, G., Ellis, M., Sumnall, H., 2011. Young People, Alcohol and the Media. Joseph Rowntree Foundation, New York. Babor, T.F., Xuan, Z., Damon, D., Noel, J., 2013. An empirical evaluation of the US Beer Institute's self-regulation code governing the content of beer advertising. Am. J. Public Health 103, e45-51. Brown, J.D., Witherspoon, E.M., 2002. The mass media and American adolescents’ health. J. Adolesc. Health 31, 153-170. Casswell, S., 2004. Alcohol brands in young peoples' everyday lives: New developments in marketing. Alcohol Alcohol. 39, 471-476. Center for Media Literacy, 2008. 2008 National Survey on Adolescent Media Use. Taiwan. Chang, F.C., Lee, C.M., Chen, P.H., Chiu, C.H., Miao, N.F., Pan, Y.C., Huang, T.F., Lee, S.C., 2014. Using media exposure to predict the initiation and persistence of youth alcohol use in Taiwan. Int. J. Drug Policy 25, 386-392. Chen, Y.Y., Chiu, Y.C., Ting, T.T., Liao, J., Chen, W.J., Chen, C.Y., 2016. Television viewing and alcohol advertising with alcohol expectancies among school-aged children in Taiwan. Drug Alcohol Depend. 162, 219-226. Collins, R.L., Ellickson, P.L., McCaffrey, D., Hambarsoomians, K., 2007. Early adolescent exposure to alcohol advertising and its relationship to underage drinking. J. Adolesc. Health 40, 527-534. Ellickson, P.L., Collins, R.L., Hambarsoomians, K., McCaffrey, D.F., 2005. Does alcohol advertising 14
promote adolescent drinking? Results from a longitudinal assessment. Addiction 100, 235-246. Engels, R.C., Hermans, R., van Baaren, R.B., Hollenstein, T., Bot, S.M., 2009. Alcohol portrayal on television affects actual drinking behaviour. Alcohol Alcohol. 44, 244-249. Fleming, K., Thorson, E., Atkin, C., 2004. Alcohol advertising exposure and perceptions: Links with alcohol expectancies and intentions to drink or drinking in underaged youth and young adults. J. Health Commun. 9, 3-29. Gordon, R., MacKintosh, A. M., Moodie, C., 2010. The impact of alcohol marketing on youth drinking behaviour: A two-stage cohort study. Alcohol Alcohol. 45, 470–480. Grube, J.W., Wallack, L., 1994. Television beer advertising and drinking knowledge, beliefs, and intentions among schoolchildren. Am. J. Public Health 84, 254-259. Health Promotion Administration, 2012. Global School-based Student Health Survey (GSHS), Ministry of Health and Welfare: Taiwan. Hibell, B., Guttormsson, U., Ahlström, S., Balakireva, O., Bjarnason, T., Kokkevi, A., Kraus, L., 2012. The 2011 ESPAD Report. Substance use among students in 36 European countries. The Swedish Council for information on Alcohol and Other Drugs, Stockholm, Sweden. Hurtz, S.Q., Henriksen, L., Wang, Y., Feighery, E.C., Fortmann, S.P., 2007. The relationship between exposure to alcohol advertising in stores, owning alcohol promotional items, and adolescent alcohol use. Alcohol Alcohol. 42, 143-149. Jernigan, D.H., 2010. The extent of global alcohol marketing and its impact on youth. Contemp. Drug Probl. 37, 57-89. Jernigan, D.H., 2011. Framing a public health debate over alcohol advertising: The Center on Alcohol Marketing and Youth 2002-2008. J. Public Health Policy 32:165-179. Jernigan, D., Noel, J., Landon, J., Thornton, N., Lobstein, T., 2017. Alcohol marketing and youth alcohol consumption: A systematic review of longitudinal studies published since 2008. Addiction 112, Suppl. 1,7-20. Johnston, L., O'Malley, P., Miech, R., Bachman, J., Schulenberg, J., 2015. Monitoring the Future National Survey Results on Drug Use: 1975-2014. Overview, Key Findings on Adolescent Drug Use. Institute 15
for Social Research, The University of Michigan. Ann Arbor. Jones, S.C., Magee, C.A., 2011. Exposure to alcohol advertising and alcohol consumption among Australian adolescents. Alcohol Alcohol. 46, 630-637. Lee, I.C., Ting, T.T., Chen, D.R., Tseng, F.Y., Chen, W.J., Chen, C.Y., 2015. Peers and social network on alcohol drinking through early adolescence in Taiwan. Drug Alcohol Depend. 153, 50-58. Lobstein, T., Landon, J., Thornton, N., Jernigan, D., 2017. The commercial use of digital media to market alcohol products: A narrative review. Addiction 112 Suppl. 1:21-27 Macklin, M.C., 1994. The effects of an advertising retrieval cue on young children's memory and brand evaluations. Psychol. Market. 11, 291-311. Masten, A.S., Faden, V.B., Zucker, R.A., Spear, L.P., 2008. Underage drinking: A developmental framework. Pediatrics 121, Suppl. 4, S235-251. McClure, A.C., Tanski, S.E., Li, Z., Jackson, K.,Morgenstern, M., Li, Z., Sargent, J.D., 2016. Internet alcohol marketing and underage alcohol use. Pediatrics 137: e20152149. McClure, A.C., Stoolmiller, M., Tanski, S.E., Engels, R.C., Sargent, J.D., 2013. Alcohol marketing receptivity, marketing-specific cognitions, and underage binge drinking. Alcohol. Clin. Exp. Res. 37, Suppl. 1, E404-413. Mokdad, A.H., Forouzanfar, M.H., Daoud, F., Mokdad, A.A., El Bcheraoui, C., Moradi-Lakeh, M., Kyu, H.H., Barber, R.M., Wagner, J., Cercy, K., Kravitz, H., Coggeshall, M., Chew, A., O'Rourke, K.F., Steiner, C., Tuffaha, M., Charara, R., Al-Ghamdi, E.A., Adi, Y., Afifi, R.A., Alahmadi, H., AlBuhairan, F., Allen, N., AlMazroa, M., Al-Nehmi, A.A., AlRayess, Z., Arora, M., Azzopardi, P., Barroso, C., Basulaiman, M., Bhutta, Z.A., Bonell, C., Breinbauer, C., Degenhardt, L., Denno, D., Fang, J., Fatusi, A., Feigl, A.B., Kakuma, R., Karam, N., Kennedy, E., Khoja, T.A., Maalouf, F., Obermeyer, C.M., Mattoo, A., McGovern, T., Memish, Z.A., Mensah, G.A., Patel, V., Petroni, S., Reavley, N., Zertuche, D.R., Saeedi, M., Santelli, J., Sawyer, S.M., Ssewamala, F., Taiwo, K., Tantawy, M., Viner, R.M., Waldfogel, J., Zuñiga, M.P., Naghavi, M., Wang, H., Vos, T., Lopez, A.D., Al Rabeeah, A.A., Patton, G.C., Murray, C.J., 2016. Global burden of diseases, injuries, and risk factors for young people's health during 1990-2013: A systematic analysis for the Global Burden of 16
Disease Study 2013. Lancet 387, 2383-2401. Morgenstern, M., Isensee, B., Sargent, J. D., Hanewinkel, R., 2011. Attitudes as mediators of the longitudinal association between alcohol advertising and youth drinking. Arch. Pediatr. Adolesc. Med. 165, 610–616. Padon, A.A., Rimal, R.N., DeJong, W., Siegel, M., Jernigan, D., 2016. Assessing youth-appealing content in alcohol advertisements: application of a content appealing to youth (CAY) index. Health Commun. Dec. 16, 1-10. Pasch, K.E., Komro, K.A., Perry, C.L., Hearst, M.O., Farbakhsh, K., 2007. Outdoor alcohol advertising near schools: What does it advertise and how is it related to intentions and use of alcohol among young adolescents? J. Stud. Alcohol Drugs 68:587–596. Primack, B.A., Nuzzo, E., Rice, K.R., Sargent, J.D., 2012. Alcohol brand appearances in US popular music. Addiction 107, 557-566. Robinson, T.N., Chen, H.L., Killen, J.D., 1998. Television and music video exposure and risk of adolescent alcohol use. Pediatrics 102, e54. Ross, C.S., de Bruijn, A., Jernigan, D.H., 2013. Do time watersheds for alcohol advertising reduce youth exposure? J. Public Affairs 13,123-129. Ross, C.S., Maple, E., Siegel, M, DeJong, W., Naimi, T.S., Ostroff, J., Padon, A.A., Borzekowski, D.L., Jernigan, D.H., 2014. The relationship between brand-specific alcohol advertising on television and brand-specific consumption among underage youth. Alcohol. Clin. Exp. Res. 38, 2234-2242. Ross, C.S., Maple, E., Siegel, M., DeJong, W., Naimi, T.S., Padon, A.A., Borzekowski, D.L., Jernigan, D.H., 2015. The relationship between population-level exposure to alcohol advertising on television and brand-specific consumption among underage youth in the US. Alcohol Alcohol. 50, 358-364 Sargent, J.D., Wills, T.A., Stoolmiller, M., Gibson, J., Gibbons, F.X., 2006. Alcohol use in motion pictures and its relation with early-onset teen drinking. J. Stud. Alcohol Drugs 67, 54-65. Snyder, L.B., Milici, F.F., Slater, M., Sun, H., Strizhakova, Y., 2006. Effects of alcohol advertising exposure on drinking among youth. Arch. Pediatr. Adolesc. Med. 160, 18-24. The Center on Alcohol Marketing and Youth, 2010. Youth Exposure to Alcohol Advertising on Television, 17
2001-2009. John Hopkins University. USA. Unger, J.B., Schuster, D., Zogg, J., Dent, C.W., Stacy, A.W., 2003. Alcohol advertising exposure and adolescent alcohol use: A comparison of exposure measures. Addict. Res. Theory 11, 177-193. Windle, M., Spear, L.P., Fuligni, A.J., Angold, A., Brown, J.D., Pine, D., Smith, G.T., Giedd, J., Dahl, R.E., 2008. Transitions into underage and problem drinking: Developmental processes and mechanisms between 10 and 15 years of age. Pediatrics 121, S273-S289. World Health Organization, 2014. Global Status Report on Alcohol and Health 2014. World Health Organization Press, Switzerland.
18
Figure Legends
Fig. 1. The percentage of marketing channel-specific past-month exposure to alcohol advertising, by baseline drinking experience (n=1863)
Note: AN: Alcohol naïve; AE: Alcohol experience *P<0.05, **P<0.01, ***P<0.00
19
Table 1 Selected characteristics of participating students at baseline (n=1863). Variablesa,b,c Lifetime drinking status at seventh grade Overall Alcohol naïve Alcohol experienced (n=1863) (n=1050) (n=813) n %wt n %wt n %wt * Gender Male 885 47.32 490 46.09 395 48.9 Female 978 52.68 560 53.91 418 51.1 *** Living with parents Either one or none 319 16.42 136 12.32 183 21.71 Both 1543 83.53 914 87.68 629 78.17 Parental education Both/either below high school 967 50.6 544 50.27 423 51.01 Both college or above 833 46.2 477 47.16 356 44.96 d *** Monthly allowance (NTD ) 0–499 854 45.97 544 52.00 310 38.19 500 or above 966 51.73 480 45.52 486 59.75 *** Parental drinking None 507 27.08 366 34.77 141 17.16 Either or both 1299 69.88 653 62.33 646 79.61 *** Best friend’s drinking None 1524 82.29 958 91.42 566 70.51 At least one 339 17.71 92 8.58 247 29.49 *** Recent exposure to media drinking portrayals None 403 21.47 270 25.51 133 16.26 1–14 days 1124 60.55 640 61.34 484 59.53 15 or more days 336 17.98 140 13.16 196 24.21 *** Intent to drink Definitely/probably no 1324 71.23 951 90.62 373 46.24 Definitely/probably yes 539 28.77 99 9.38 440 53.76 *** Past-year alcohol drinking at 9th grade None 1310 70.66 871 83.18 439 54.5 Yes 553 29.35 179 16.82 374 45.5 *** Occasions of drinking among past-year drinkers at 9th grade Mean (SE) 3.78 0.52 2.67 0.25 4.31 0.63 Min–Max 1 200 1 30 1 200 Q1, Q2, Q3 (1, 2, 4) (1, 2, 3) (1, 3, 5) Note. *P<0.05, **P<0.01, ***P<0.001 a
Variables were assessed at seventh grade (baseline) if not otherwise specified; Columns may not add up to 100% due to missing data; c All variables were estimated and examined by complex survey analyses; d NTD: New Taiwan Dollars, 1 USD = 31 NTD in 2010. b
20
Table 2 Association estimates between alcohol advertisements exposure and alcohol initiationa (n=1050) Crude Model 1 Model 2
Variables
OR
(95% CI)
aOR (95% CI)
aOR
Model 3
(95% CI) aOR
(95% CI)
Alcohol advertisement exposure at 7th gradeb Television (Ref=None) 1–14 days
3.16
(1.56–6.42)**
3.20
(1.54-6.63)**
2.75
(1.39-5.45)**
2.62
(1.14-6.02)*
15 or more days
2.42
(1.22–4.81)*
2.35
(1.22-4.55)*
2.06
(1.07-3.95)*
1.73
(0.68-4.39)
1–14 days
1.41
(0.94–2.14)
1.36
(0.91-2.02)
1.24
(0.81-1.92)
1.09
(0.57-2.12)
15 or more days
1.69
(1.26–2.25)***
1.55
(1.17-2.06)
1.47
(1.07-2.03)*
1.41
(0.86-2.31)
1.28
(0.92–1.79)
1.23
(0.88-1.70)
1.13
(0.79-1.63)
1.09
(0.65-1.81)
Stores (Ref=None)
Websites (Ref=None) 1–14 days
*
1.46
(1.01–2.13)
1.38
(0.97-1.97)
1.31
(0.91-1.89)
1.40
(0.80-2.47)
1–14 days
0.84
(0.69–1.01)
0.84
(0.69-1.02)
0.77
(0.63-0.94)**
0.65
(0.49-0.87)**
15 or more days
1.04
(0.85–1.27)
1.03
(0.86-1.23)
1.03
(0.86-1.22)
0.89
(0.66-1.21)
1–14 days
1.39
(0.78–2.50)
1.35
(0.80-2.28)
1.27
(0.77-2.11)
1.31
(0.69-2.50)
15 or more days
0.97
(0.64–1.47)
0.96
(0.65-1.42)
0.96
(0.62-1.47)
0.91
(0.54-1.53)
1–14 days
0.94
(0.61–1.46)
0.97
(0.65-1.47)
0.99
(0.65-1.50)
0.91
(0.59-1.40)
15 or more days
0.80
(0.50–1.26)
0.78
(0.52-1.16)
0.74
(0.47-1.16)
0.67
(0.37-1.20)
0.96
(0.73–1.27)
0.97
(0.78-1.20)
0.96
(0.39-2.36)
15 or more days Newspapers (Ref=None)
Magazines (Ref=None)
Radio (Ref=None)
Gender (Ref=Female) Male
Living with parents (Ref=Either one or none) Both
0.87
(0.36–2.12)
Parental education (Ref=Both or either one below high school) Both college or above
0.84
(0.61–1.16)
0.92
(0.72-1.19)
0.93
(0.71–1.21)
0.86
(0.66-1.13)
1.81
(1.19–2.76)**
1.70
(1.13-2.56)*
(1.28–2.09)***
1.56
(1.16-2.09)**
Parental drinking (Ref=None) Either or both Best friend’s drinking (Ref-None) At least one
Monthly allowance (Ref=0–499) (NTD)c 500 or above
1.64
Exposure to media drinking portrayals (Ref=None) 1–14 days
1.72
(1.22–2.42)**
1.36
(1.00-1.85)
15 or more days
1.49
(0.87–2.57)
1.18
(0.57-2.43)
1.52
(0.72-3.20)
Intent to drink at 7th grade (Ref=Definitely/probably no) Definitely/probably yes
1.59
(0.85–2.97)
Note. *P<0.05, **P<0.01, ***P<0.001, cOR: crude odds ratio, aOR: adjusted odds ratio, CI: confidence interval
21
a
Alcohol drinking was assessed at 9th grade follow-up Past-month interval, the baseline assessment c NTD: New Taiwan Dollars, 1USD= 31 NTD in 2010 Model 1 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, and best friend’s drinking; the market channel was entered one at a time. Model 2 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, best friend’s drinking, exposure to media drinking portrayals and the intent to drink at seventh grade; the market channel was entered one at a time. Model 3 adjusted for all listed variables. b
22
Table 3 Association estimates between alcohol advertisements exposure and occasional alcohol drinking (n=813).
Continued occasional drinkinga Crude Model 1 cOR (95% CI) aOR (95% CI)
Variables
Model 2 aOR (95% CI)
Model 3 aOR (95% CI)
Alcohol advertisement exposure at 7th gradeb Television (Ref=None) 1–14 days
1.32
(0.83–2.21)
1.15
(0.71-1.85)
1.13
(0.67-1.90)
1.00
(0.58-1.74)
15 or more days
0.97
(0.61–1.55)
0.84
(0.52-1.36)
0.89
(0.52-1.53)
0.90
(0.49-1.65)
1–14 days
0.95
(0.63–1.43)
0.92
(0.60-1.39)
0.85
(0.55-1.32)
0.77
(0.48-1.24)
15 or more days
1.28
(0.84–1.94)
1.04
(0.67-1.61)
1.10
(0.69-1.76)
0.80
(0.48-1.36)
1.80
(1.30–
1.61
(1.17-2.23)**
1.60
(1.15-2.24)**
1.50
(1.04-2.15)*
1.09
(0.76-1.57)
1.11
(0.77-1.60)
1.15
(0.74-1.77)
0.92
(0.69-1.24)
0.88
(0.66-1.18)
0.75
(0.55-1.04)
0.97
(0.61-1.55)
Stores (Ref=None)
Websites (Ref=None) 1–14 days
2.49)*** 15 or more days
1.21
(0.85–1.74)
0.91
(0.68–1.22)
Newspapers (Ref=None) 1–14 days
1.61
(1.09–2.37)
*
1.42
(0.95-2.10)
1.55
(1.04-2.31)
1–14 days
1.45
(1.09–1.92)*
1.29
(0.97-1.72)
1.31
(0.98-1.76)
1.16
(0.84-1.60)
15 or more days
1.16
(0.73–1.86)
1.06
(0.67-1.70)
1.13
(0.71-1.79)
1.04
(0.63-1.72)
1–14 days
1.12
(0.79–1.57)
1.23
(0.88-1.73)
1.11
(0.79-1.55)
1.16
(0.82-1.64)
15 or more days
2.50
(1.62–
2.42
(1.55-3.79)***
2.90
(1.85-4.53)***
2.58
(1.60-4.15)***
1.02
(0.78-1.33)
0.99
(0.71-1.36)
15 or more days
*
Magazines (Ref=None)
Radio (Ref=None)
3.87)*** Gender (Ref=Female) Male
0.84
(0.64–1.09)
Living with parents (Ref=Either one or none) Both
1.12
(0.81–1.54)
Parental education (Ref=Both or either one below high school) Both college or above Parental
0.62
(0.32–1.22)
0.83
(0.63-1.08)
2.18
(0.98–4.87)
1.78
(1.25-2.55)**
(0.89–1.59)
1.09
(0.80-1.47)
1.51
(1.14-1.99)**
drinking
(Ref=None) Either or both
Best friend’s drinking (Ref=None) At least one
1.19
Monthly allowance (Ref=0–499) (NTD) 500 or above
1.16
c
(0.45–2.94)
Exposure to media drinking portrayals (Ref=None) b 23
1–14 days
1.60
(1.09–2.34)*
1.26
(0.83-1.91)
15 or more days
1.26
(0.82–1.95)
1.05
(0.64-1.73)
2.28
(1.72-3.02)***
Intent to drink at 7th grade (Ref=definitely no or probably no) Definitely/probably yes
1.67
(1.28– 2.19)***
Note. *P<0.05, **P<0.01, ***P<0.001, cRR: crude relative risk, aRR: adjusted relative risk, CI: confidence interval a 0–3 occasions as reference group; alcohol drinking was assessed at 9 th grade follow-up; b Past-month interval, the baseline assessment; c NTD: New Taiwan Dollars, 1USD= 31 NTD in 2010; Model 1 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, and best friend’s drinking; the market channel was entered one at a time. Model 2 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, best friend’s drinking, exposure to media drinking portrayals at seventh grade, and intent to drink at seventh grade; the market channel was entered one at a time. Model 3 adjusted for all listed variables.
24
S0376-8716(17)30232-6 http://dx.doi.org/doi:10.1016/j.drugalcdep.2017.03.041 DAD 6448
To appear in:
Drug and Alcohol Dependence
Received date: Revised date: Accepted date:
8-9-2016 31-3-2017 31-3-2017
Please cite this article as: Chen, Chuan-Yu, Huang, Hsueh-Yu, Tseng, FangYi, Chiu, Yu-Chan, Chen, Wei J., Media alcohol advertising with drinking behaviors among young adolescents in Taiwan.Drug and Alcohol Dependence http://dx.doi.org/10.1016/j.drugalcdep.2017.03.041 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 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.
Media alcohol advertising with drinking behaviors among young adolescents in Taiwan
Chuan-Yu Chena,b,c*, Hsueh-Yu Huanga,c, Fang-Yi Tsenga, Yu-Chan Chiud, Wei J. Chene
a
Institute of Public Health, National Yang-Ming University, Taipei, Taiwan
b
Center of Neuropsychiatric Center, National Health Research Institutes, Miaoli, Taiwan
c
Children and Family Research Center, National Taiwan University, Taipei, Taiwan
d
Department of Bio-industry Communication and Development, National Taiwan University, Taipei, Taiwan
e
Institute of Epidemiology and Preventive Medicine, College of Public Health, National Taiwan University, Taipei, Taiwan
Correspondence: Chuan-Yu Chen, PhD, Professor National Yang-Ming University Institute of Public Health No.155, Sec.2, Linong Street, Taipei, 112 Taiwan Medical Building II, Rm 210 Email: [email protected] Tel: 886-02-2826-7344 Fax: 886-02-2822-4850
Highlight
TV, in-store display, and website are common channels for alcohol ads exposure.
Exposure to TV alcohol ads may increase drinking initiation in early adolescence.
Exposure to ads on the website and radio may elevate occasional drinking. 1
ABSTRACT Objectives: To investigate potential effects of alcohol ads in six major marketing channels on drinking behaviors among young adolescents in Taiwan. Methods: The data were derived from the Alcohol-Related Experiences among Children study. The baseline sample was comprised of 1926 seventh-eighth graders from 11 public middle schools in Taipei in 2010; follow-up was conducted one year later (follow-up rate=97%). Information concerning individual sociodemographics, family characteristics, exposure to media portrayals of drinking and alcohol ads on major marketing channels, and drinking experience was collected through web-based self-administered questionnaires. Complex survey analyses were used to evaluate the association estimates, with stratification by prior drinking experiences in childhood. Results: Television, in-store displays, and websites are the three most common marketing channels for young adolescents to report past-month alcohol advertising exposure. With statistical adjustment for potential confounders and six market channels, exposure to alcohol ads on television was associated with subsequent increased drinking initiation (adjusted odds ratio [aOR]=2.62; 95% CI=1.14–6.02). For those who have initiated alcohol use in childhood, the exposure to ads on the web (aOR=1.50; 95% CI=1.04–2.15) and radio (aOR=2.58; 95% CI=1.60–4.15) may elevate subsequent risk of occasional drinking. Exposure to media drinking portrayals was not related to subsequent drinking behaviors in this sample. Conclusions: Our results demonstrated that the effects of alcohol advertising on drinking behaviors in early adolescence may differ by marketing channels. Preventive strategies targeting underage drinking should consider restraining marketing channels (e.g., websites and radio) from certain advertising content and placement.
Keywords: adolescent, alcohol advertising, marketing channel 1. Introduction Alcohol consumption and harmful drinking have become a norm in young people, particularly in high-income countries (World Health Organization, 2014). More than 4 in 5 European students aged 15–16 2
years had drunk alcohol on at least one occasion in their lifetimes, and roughly 30% of 8th, 10th, and 12 graders in the US had ever been drunk (Hibell et al., 2012; Johnston et al., 2015). In Taiwan, school surveys with national representativeness found the lifetime prevalence rate of alcohol drinking among 8th graders has increased from 56% in 2006 to 65% in 2012 (Health Promotion Administration, 2012). Early-onset alcohol drinking is associated with an increased risk of health problems and developmental impairment (Masten et al., 2008; Windle et al., 2008). Data from the World Health Organization Global Burden of Disease study indicated that alcohol is the leading risk factor for death and disability-adjusted life years among people aged 15–24 years (Mokdad et al., 2016). To provide targets for preventing early-onset alcohol initiation and continued alcohol drinking, it is imperative to identify modifiable factors shaping the transition of alcohol experiences in the first two decades of one’s life (Masten et al., 2008; Windle et al., 2008). With their rising penetration into daily lives of youngsters, the role of media on shaping youngsters’ drinking behaviors has been gradually recognized (Brown and Witherspoon, 2002). The media are hypothesized to influence the drinking behaviors through several processes, among which alcohol advertising is especially notable given its expansion and diversity (Anderson et al., 2009; Casswell, 2004; Engels et al., 2009; Jernigan et al., 2017; Robinson et al., 1998; Sargent et al., 2006). Globally, the top nine alcohol beverage companies have spent at least €2.5 billion on traditional media advertising in 2004 (Anderson, 2007). The advertising industry self-regulates the content of alcohol advertising, ostensibly to protect adolescents against the exposure to alcohol advertising. Nevertheless, the guidelines concerning the use of content appealing to minors are often violated (Babor et al., 2013), and young people are greatly exposed to alcohol marketing (Jernigan, 2010). Alcohol companies spent more than $8 billion on alcohol advertising on television between 2001 and 2009 in the US, and exposure to alcohol advertising on television among underage youth increased by 71% during that same period (The Center on Alcohol Marketing and Youth, 2010). The weekly exposure to retail-ad was estimated at two thirds for middle school students in California (Hurtz et al., 2007). In Britain, a report shows that 60% of young people were exposed to alcohol advertising on a daily basis through television and 25% through magazines (Atkinson et al., 2011). 3
Although variation exists in measure and coverage of marketing channel (e.g., media platform-specific or composite measure), in assessment timeframe of exposure (e.g., past-year or past-month), and in outcome measure (e.g., alcohol initiation or binge drinking), evidence from longitudinal studies has shown that exposure to alcohol advertising in traditional media platforms, including television (Robinson et al., 1998; Ross et al., 2014), radio (Collins et al., 2007), movies/films (Sargent et al., 2006), magazines (Ellickson et al., 2005), and outdoors outlets (e.g., billboards or convenience stores)(Pasch et al., 2007), is a significant predictor for underage behavior. Recently, with easier access to internet and mobile phone in young population, attention has also been directed to emerging media outlets (Atkinson et al., 2011; Lobstein et al., 2017). To illustrate, a cross-sectional study in Australia reported that exposure to alcohol advertising on the internet is associated with a 36% increase in the chance of past-month drinking behavior (Jones and Magee, 2011). Another study, using the average score to reflect advertisement exposure in the past year through television, newspapers or magazines, outdoors billboards, alcohol portraits in TV/movies, and the discussion on the internet, found that one point increase in alcohol ads exposure elevates alcohol initiation by 68% and persistence by 44% among old adolescents in Taiwan (Chang et al., 2014). The information conveyed in alcohol advertising often generates positive expectancies toward alcoholic beverages (Chen et al., 2016; Fleming et al., 2004; Ross et al., 2014); several studies also showed that higher exposure to alcohol advertising is associated with increased drinking intention or favorable attitudes among children and adolescents (Collins et al., 2007; Grube and Wallack, 1994; Unger et al., 2003). These pro-alcohol cognitive, intentional, and attitude attributes are improtant precursors for adolescents’ subsequent drinking behaviors and problems. A recent longitudinal study utilized masked ad images (i.e., showing advertising images with the logo removed) to measure contact frequency and brand recall for alcohol advertisements among sixth graders in Germany. The study concluded that 35% of total effects on the
connection
between
alcohol
adverting
exposure
and
alcohol
initiation
is
due
to
positive attitudes toward alcohol (Morgenstern et al., 2011). To date, research on the connection between alcohol advertising and underage drinking behaviors suffers from two important limitations. First, earlier studies often examined advertising effects either with a focus on traditional marketing channels (e.g., television or movie) or with a pooled measure (Chang et al., 4
2014; Snyder et al., 2006). The prospective evidence pertaining to certain emerging marketing channels awaits accumulation (McClure et al., 2016). Second, with few exception (Chang et al., 2014), existing research on the longitudinal associations linking ads exposure and alcohol drinking has been mostly limited in the high-income countries with mature alcohol markets. There is a need to examine advertising effects on young populations in the emerging alcohol markets wherein alcohol companies have gradually increased their marketing expenditures over the past decade. To address the abovementioned research gaps, we turned to a longitudinal study that prospectively followed up a cohort of seventh graders (aged 13–14 years old) to study prevalence of alcohol advertising exposure through media platform in young adolescents in Taiwan. Given that the risk factors for drinking problem evolution are stage-dependent (e.g., initiation and continued drinking)(Anthony and Chen, 1999; Chang et al., 2014; Ellickson et al., 2005; Gordon et al., 2010; McClure et al., 2016), the present study further assessed the relationship linking each media platform with subsequent alcohol initiation and occasional drinking behaviors through early adolescence. 2. Methods 2.1. Study Design and Population The data were derived from the Alcohol-Related Experiences among Children (AREC) study (Lee et al., 2015). Taiwan has nine years of compulsory education: six years of elementary school and three years of middle school. Utilizing the multistage sampling procedure, in late 2009 we first randomly selected 11 of the 73 public middle schools in Taipei (which more than 90% of the city’s middle school students attend); all seventh graders in the sampled schools were eligible to participate in the study. Upon receiving positive consent from both guardians (e.g., parents) and the students themselves, the students were asked to fill out the self-administered web-based questionnaire during regular class hours (baseline response rate=55%) while in-class teachers were absent. The baseline sample comprised 1926 seventh and eighth graders (aged 13-14 years) in 2010; the second assessment was conducted in the ninth grade, with an average follow-up interval of 13 months (n=1870; follow-up rate=97%). All study procedures were approved by the institutional review board of the National Health Research Institutes (EC0980605). The present study excluded seven students with missingness in the variables of major interest, resulting in an analytic sample 5
of 1863. 2.2. Measures The questionnaire mostly took students less than 45 minutes to complete. The questionnaire included individual sociodemographic and pubertal characteristics, family structure and socioeconomic status, family and peer drinking behaviors, exposure to alcohol advertising through six types of media, and individual drinking behaviors. 2.2.1 Outcome variables. At both waves of assessment, alcohol-drinking experiences were assessed by a series of questions (Lee et al., 2015). The first one was “Not including a sip or alcohol in food, have you ever drunk any alcoholic beverages in your lifetime?” If the response was positive, additional questions were asked, including drinking experience in the 12 months preceding the survey. In the present study, we defined “incident drinkers” as those who responded as (i) having never drank at the first assessment and (ii) having drunk on at least one occasion in the past 12 months at the follow-up assessment. For the youngsters who reported ever drank alcohol at baseline survey, we used the second quartile of past-year drinking occasions in the 9th grade as the cut-off point (see Table 1) to define “occasional drinking”—having drunk on four or more occasions in the 12 months preceding the follow-up assessment. 2.2.2 Independent variables. Exposure to alcohol advertising at the baseline was assessed by the question “In the past month, have you ever seen any alcohol advertising or promotional messages in the following marketing channel?” This was asked for nine types of marketing channels: television, movies, newspapers, magazines, radio, stores (e.g., convenience stores and supermarkets), billboards/posters/flyers, internet (e.g., websites), and e-mail; if positive, the frequency of exposure was further queried using four categories: 1–7 days, 8–14 days, 15–29 days, and 30 or more days. In the present analyses, we regrouped the frequency variable of alcohol advertising exposure into 0 days, 1–14 days, and 15 or more days. Considering what youngsters had daily access to and the type of media alcohol industries tend to use in Taiwan (Center for Media Literacy, 2008), we decided to focus on the six major types of marketing channels for alcohol advertising: television, newspapers, magazines, radio, Internet, and in-store displays. 2.2.3 Covariates. Baseline variables thought to be associated with alcohol advertising exposure and alcohol drinking in Taiwan were taken into account as important confounding factors (Lee et al., 2015; Ting 6
et al., 2015). These included gender, living with parents (living with one or none vs. living with both), parental education (both/either high school or below vs. both college or above), monthly allowance (NTD 0– 499 vs. 500 or above; 1 USD≈31 NTD), parental drinking (none vs. either one drinking), best friend’s drinking (none vs. at least one), and future intent to drink (definitely/probably not vs. definitely/probably yes). Other than alcohol ads, youths’ media exposure to alcohol-related images was assessed by the question (Engels et al., 2009):“In the past month, on how many days have you seen drinking portrayals on television, in a movie, or on the Internet?” 2.3. Statistical analyses Since the samples were drawn via multistage probability sampling procedures, the present study used standard survey analysis to take into account sampling parameters (e.g., sampling weights to compensate for variation in sample-selection probability and within school non-independence). We first examined the distribution of individual sociodemographics, family and peer drinking, and past-month exposure to alcohol advertising with stratification by baseline lifetime drinking experience. Next, we adopted survey logistic regression to estimate the association linking media alcohol advertising with incident drinking and continued occasional drinking. In Models 1 and 2, the market channel was entered one at a time: we first included individual sociodemographic characteristics and family and peer drinking in Model 1, followed by the addition of the exposure to media portrayals of drinking and the intent to drink in the seventh grade (Collins et al., 2007; Grube and Wallack, 1994). Finally, we entered all market channels simultaneously in Model 3. All statistical analyses were carried out by SAS 9.3 survey procedures.
3. Results Approximately 45% of the surveyed students had drank alcoholic beverages at least once at baseline, 70% had at least one drinking parent, and 79% had recent exposure through the media to portrayals of drinking (Table 1). Alcohol-experienced young adolescents, when compared with their alcohol-naïve peers, were more likely to be male, not live with both parents, have a higher monthly allowance, have drinking parents, have at least one drinking best friend, have exposure to drinking portrayals on television, in movies, and on the internet, as well as have positive future drinking intent (p<0.05). Among the alcohol naïve at 7
baseline, 17% have initiated alcohol drinking by 9th grade. An estimated 46% of the alcohol-experienced had drank on at least one occasion in the year preceding the ninth grade survey , and the second quartile of cumulative drinking occasion was three. Among the six major marketing channels in Taiwan, television was the most common one through which seventh graders are exposed to alcohol advertising (>88%), followed by in-store displays and websites (see Figure1). In general, the proportion of past-month exposure to alcohol advertisements at baseline survey for the alcohol experienced (AE) was slightly higher than their alcohol-naïve (AN) counterparts. Drinking experience-related differences in exposure to alcohol advertising were also manifested in higher exposure frequency (i.e., 15 days or above), particularly through in-store displays, websites, and magazines (P<0.001). Table 2 presents the prospective association linking exposure to alcohol advertising with incident alcohol drinking. With statistical adjustment for sociodemographic covariates (see Model 1), television was the only significant market channel linked with increased drinking initiation (1–14 days: aOR=3.20; 95% CI=1.54–6.63; 15 or more days: aOR=2.35; 95% CI=1.22–4.55). However, after taking exposure to media portrayals of drinking and drinking intention into account (Model 2), the exposure to ads in stores (e.g., 15 or more days: aOR=1.47; 95% CI=1.07–2.03) and on newspaper (e.g., 1-14 days: aOR=0.77; 95% CI=0.63-0.94) became significant. Finally, with all marketing channels added into the model (see Model 3), the baseline exposure to alcohol ads on television was linked with increased drinking initiation (e.g., 1–14 days: aOR=2.62; 95% CI=1.14–6.02), whereas for newspapers the risk was lowered (e.g., 1–14 days: aOR=0.65; 95% CI=0.49–0.87). Having at least one drinking best friend (aOR=1.70) or having higher monthly allowance (aOR=1.56) was found to slightly increase the risk of alcohol initiation. For the alcohol-experienced 7th graders, having exposure to alcohol advertising through websites or radio slightly elevated the risk of occasional drinking in the subsequent year (see Table 3, Model 1). The risk estimates for the website (1–14 days: aOR=1.50; 95% CI=1.04-2.15) and the radio (15 or more days: aOR=2.58; 95% CI=1.60–4.15) did not have appreciable changes once the exposure to media portrayals of drinking, drinking intention, and all market channels were added into model (see Model 3). In addition, the intent to drink at baseline (aOR=2.28; 95% CI=1.72–3.02) appeared to the strongest predictor. 8
4. Discussion The present study is one of the first few to examine the prospective relationship linking media alcohol advertising with alcohol initiation and continuation among young adolescents in emerging markets. We found that television, stores, and websites were the three most common media platforms through which youngsters recalled alcohol ads exposure; alcohol-experienced youngsters reported a higher frequency of exposure to alcohol ads. Independent of gender, living with parents, parental education, parental drinking, best friend’s drinking behavior, monthly allowance, drinking intention, and exposure to media drinking portrayals at baseline, having exposure to alcohol ads on television were associated with increased risk of alcohol initiation, whereas exposure to alcohol ads on websites or radio were associated with subsequent elevated risk of occasional drinking. The estimated rates of past-month alcohol ad exposure through television, magazines, and websites in our young population were slightly lower than those in the United Kingdom (i.e., 97%, 80%, and 84%, respectively)(Atkinson et al., 2011). However, our estimates of exposure to alcohol ads through magazines, websites, and in-store displays were considerably higher than the estimates in Australia (71%, 53%, and 75%, respectively)(Jones and Magee, 2011). In addition to true cross-country variation, the observed differences in exposure to alcohol ads might be partially explained in the context of population characteristics (e.g., 12–18 years old vs. 13–14 years old), survey period, and assessment timeframe (e.g., lifetime vs. past month). Although alcohol companies claimed to have strengthened their self-regulatory standards, our data revealed that exposure to both alcohol advertising on major marketing channels and media portrayals of drinking appeared to be very common (or to be the norm) for young adolescents in Taiwan. Of particular importance are higher alcohol advertising exposure rates among those seventh graders who have drank alcohol in their childhood, regardless of media outlets (Anderson et al., 2009). These findings may be the results of several pathways: (i) the alcohol-experienced children were more likely to notice or be attracted to alcohol advertising; (ii) the alcohol-experienced children were more likely to recognize or remember alcohol advertising; and (iii) the alcohol-experienced children were more likely to become chronically exposed to alcohol advertising to develop brand allegiance (Anderson et al., 2009; Macklin, 1994; McClure et al., 2013). This observation implicated that the cognitive and affective 9
processing of alcohol advertising is probably not identical across subgroups defined by drinking experiences and reinforced the possibility that advertising effects are not static and may change as drinking experience evolves (Jernigan et al., 2017; McClure et al., 2013). Consistent with prior research highlighting the importance of alcohol advertising on underage drinking behaviors (Anderson et al., 2009; Chang et al., 2014; Collins et al., 2007; Grube and Wallack, 1994), our results suggested that the association may vary by media platform. Although alcohol industries integrated their commercial communication across marketing channels, the association estimates from multivariate models yield some insights for marketing channel-differential impact over the course of alcohol drinking behaviors (Ellickson et al., 2005). In Taiwan, television viewing and convenience stores visits are a common part of daily life. In the case of television, although there has been time restriction on alcohol advertising (e.g., 9:00 p.m.-6:00 a.m.) to protect young people from exposure, the exposure was still high because nearly 60% of children watched television after 9 p.m. (Chen et al., 2016; Ross et al., 2013). The frequent exposure to alcoholic beverage images may normalize drinking behaviors and elevate positive alcohol expectancies, consequently predisposing some youngsters to alcohol initiation (Chen et al., 2016; Fleming et al., 2004; McClure et al., 2013). We also noted the risk of alcohol initiation was not in a dose-response relationship with the days of televised advertisement exposure. Future research is needed to elaborate time, contextual, and advertisement variables, such as exposure timing (e.g., weekday/weekend), recessive advertisement in television programs (e.g., idol dramas), and advertising content/brands, to delineate the non-linearity relationship (Padon et al., 2016; Ross et al., 2015). For the alcohol-experienced adolescents, alcohol advertising on marketing channels such as radio and websites can provide information pertaining to brand, price, and point of sale, which may facilitate the development of drinker identity and increase drinking occasions (McClure et al., 2013). Additionally, some of the popular music aired involved elements associated with alcohol brand and drinking context (Primack et al., 2012), which is particularly common during summer recess and holiday seasons in Taiwan. Although exposure to newspaper alcohol advertising was linked with a significant increase in alcohol initiation by crude association analyses, multivariate analyses unexpectedly found an inverse association. This observation might be partly due to three reasons: first, those who read newspapers may represent a 10
subsample of adolescents who generally endorsed a lower risk of alcohol initation through early adolescence (e.g., coming from families with higher socioeconomic status)(Center for Media Literacy, 2008). Second, alcohol-related news often features negative or disastrous events (e.g., sexual assaults, family violence, and drunk driving). Third, it is possible that the newspaper advertising is more likely to be price than lifestyle, which may not attractive or appealing to youth (Padon et al., 2016). Another interesting finding is that the salient effects of drinking intent only appeared in continued drinking. A series of post-hoc Sobel tests analyses suggested that “drinking intent” probably played a mediating role in the process linking website and radio alcohol advertising with drinking on four or more occasions. Although the intent to drink and exposure to alcohol advertising were measured at the same time, it is possible that certain advertising information delivered through radio or websites may foster greater marketing receptivity and reinforce the desire to drink (Collins et al., 2007; Fleming et al., 2004; McClure et al., 2013; Morgenstern et al., 2011). 4.1. Limitations and Strengths This study has four main limitations. First, the measure of exposure to alcohol advertising was limited to only the month preceding the survey, which may not be representative of a youth’s cumulative or past-year experience; however, asking young respondents to recall such exposure over a longer time frame is less likely to produce satisfactory validity. To reflect a longer period of cumulative exposure, some objective methods have been implicated (e.g., adstock or the Nielsen data)(Chen et al., 2016; Morgenstern et al., 2011; Ross et al., 2014). Second, because our data were collected at two time points, the causal inference regarding the connection between alcohol advertising and drinking behaviors may be affected if reciprocal causations operated in between. Nevertheless, we have carried out the analyses with stratification by prior drinking experience to reduce possible bias. Future studies with data from multiple time points with a larger number of alcohol-naïve children through early adolescence could help delineate intertwined relationships. Third, the relatively small sample size may pose a threat to estimation precision, which is particularly true when the incident drinking is uncommon. Finally, because the concept and definition of “drink (amount)” is not widely used in the study population, it is not possible to assess problematic alcohol consumption through the measure of “drink” (e.g., binge drinking). The approach that assessed frequency of alcohol consumption by “cumulative drinking occasions” and defined occasional drinkers on the basis of the sample-oriented quartile 11
may warrant further attention in measure validation. Similarly, our sample represents early adolescents in a metropolitan city in Taiwan, the generalizability to other societies may be restrained. Despite these limitations, our study has several strengths. First, our assessment of marketing channels is more comprehensive than that in most published reports. The exposure-frequency measure for alcohol advertising on each market channel also provides more detailed information than the binary responses commonly used. In addition, the prospective nature of the present study may build better temporality. 4.2.
Conclusions
This investigation, to our knowledge, is one of the few to report the prospective relationships linking a rather comprehensive list of alcohol marketing channels with drinking behaviors among young adolescents in emerging markets. Although current regulations establish standards for the broadcasting of alcohol advertising messages in Taiwan, these standards do not cover all media (e.g., social media) and the regulations are loosely enforced in certain media (e.g., pirate radio). For the internet, age verification screen should be placed before alcohol advertisement is broadcasting in the gateway websites (such as Yahoo and MSN) or on video sites (such as YouTube). Finally, given how international alcohol industries have extended alcohol sales into emerging markets, future work is warranted to monitor the content, form, placement, and intensity of alcohol advertising on traditional (e.g., television and radio) and emerging (e.g., websites and social media) marketing channels, with the aid of certain youth exposure indicators (e.g., gross rating points and over exposure)(Jernigan, 2011; Morgenstern et al., 2011).
Author Disclosures
Author Contributions CYC was responsible for the development of the research theme, the analysis, and the drafting of the manuscript. HYH assisted the analyses and data interpretation. FYT helped with the field work and data analyses. YCC instructed result interpretation and revised the draft. WJC assisted study design and draft revision. All authors have critically reviewed the content and approved the final version submitted for 12
publication.
Role of Funding Source Nothing declared.
Conflict of Interest The authors declare no conflict of interest with respect to the conduct or content of this work.
Acknowledgement This work was supported by grants from the Ministry of Science and Technology (95-2314-B-400-009-MY3, 98-2314-B-010-038-MY3, 101-2628-B-010-004-MY3, and 104-2314-B-010 -008 -MY3 to C. Y. Chen) the National Health Research Institutes (MDPP04-014), and a grant from the Ministry of Education of Taiwan ROC, Aim for the Top University Plan to CYC. The funding agencies had no role in the study design; in the collection, analysis, or interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
13
References Anderson, P., 2007. The Impact Of Alcohol Advertising: ELSA project report on the evidence to strengthen regulation to protect young people. National Foundation for Alcohol Prevention. Utrecht. Anderson, P., de Bruijn, A., Angus, K., Gordon, R., Hastings, G., 2009. Impact of alcohol advertising and media exposure on adolescent alcohol use: A systematic review of longitudinal studies. Alcohol Alcohol. 44, 229-243. Anthony, J.C., Chen, C.Y., 1999. Epidemiology of drug dependence. In: Textbook of Substance Abuse Treatment. American Psychiatric Press Inc, Washington DC, pp.55-72. Atkinson, A., Elliot, G., Ellis, M., Sumnall, H., 2011. Young People, Alcohol and the Media. Joseph Rowntree Foundation, New York. Babor, T.F., Xuan, Z., Damon, D., Noel, J., 2013. An empirical evaluation of the US Beer Institute's self-regulation code governing the content of beer advertising. Am. J. Public Health 103, e45-51. Brown, J.D., Witherspoon, E.M., 2002. The mass media and American adolescents’ health. J. Adolesc. Health 31, 153-170. Casswell, S., 2004. Alcohol brands in young peoples' everyday lives: New developments in marketing. Alcohol Alcohol. 39, 471-476. Center for Media Literacy, 2008. 2008 National Survey on Adolescent Media Use. Taiwan. Chang, F.C., Lee, C.M., Chen, P.H., Chiu, C.H., Miao, N.F., Pan, Y.C., Huang, T.F., Lee, S.C., 2014. Using media exposure to predict the initiation and persistence of youth alcohol use in Taiwan. Int. J. Drug Policy 25, 386-392. Chen, Y.Y., Chiu, Y.C., Ting, T.T., Liao, J., Chen, W.J., Chen, C.Y., 2016. Television viewing and alcohol advertising with alcohol expectancies among school-aged children in Taiwan. Drug Alcohol Depend. 162, 219-226. Collins, R.L., Ellickson, P.L., McCaffrey, D., Hambarsoomians, K., 2007. Early adolescent exposure to alcohol advertising and its relationship to underage drinking. J. Adolesc. Health 40, 527-534. Ellickson, P.L., Collins, R.L., Hambarsoomians, K., McCaffrey, D.F., 2005. Does alcohol advertising 14
promote adolescent drinking? Results from a longitudinal assessment. Addiction 100, 235-246. Engels, R.C., Hermans, R., van Baaren, R.B., Hollenstein, T., Bot, S.M., 2009. Alcohol portrayal on television affects actual drinking behaviour. Alcohol Alcohol. 44, 244-249. Fleming, K., Thorson, E., Atkin, C., 2004. Alcohol advertising exposure and perceptions: Links with alcohol expectancies and intentions to drink or drinking in underaged youth and young adults. J. Health Commun. 9, 3-29. Gordon, R., MacKintosh, A. M., Moodie, C., 2010. The impact of alcohol marketing on youth drinking behaviour: A two-stage cohort study. Alcohol Alcohol. 45, 470–480. Grube, J.W., Wallack, L., 1994. Television beer advertising and drinking knowledge, beliefs, and intentions among schoolchildren. Am. J. Public Health 84, 254-259. Health Promotion Administration, 2012. Global School-based Student Health Survey (GSHS), Ministry of Health and Welfare: Taiwan. Hibell, B., Guttormsson, U., Ahlström, S., Balakireva, O., Bjarnason, T., Kokkevi, A., Kraus, L., 2012. The 2011 ESPAD Report. Substance use among students in 36 European countries. The Swedish Council for information on Alcohol and Other Drugs, Stockholm, Sweden. Hurtz, S.Q., Henriksen, L., Wang, Y., Feighery, E.C., Fortmann, S.P., 2007. The relationship between exposure to alcohol advertising in stores, owning alcohol promotional items, and adolescent alcohol use. Alcohol Alcohol. 42, 143-149. Jernigan, D.H., 2010. The extent of global alcohol marketing and its impact on youth. Contemp. Drug Probl. 37, 57-89. Jernigan, D.H., 2011. Framing a public health debate over alcohol advertising: The Center on Alcohol Marketing and Youth 2002-2008. J. Public Health Policy 32:165-179. Jernigan, D., Noel, J., Landon, J., Thornton, N., Lobstein, T., 2017. Alcohol marketing and youth alcohol consumption: A systematic review of longitudinal studies published since 2008. Addiction 112, Suppl. 1,7-20. Johnston, L., O'Malley, P., Miech, R., Bachman, J., Schulenberg, J., 2015. Monitoring the Future National Survey Results on Drug Use: 1975-2014. Overview, Key Findings on Adolescent Drug Use. Institute 15
for Social Research, The University of Michigan. Ann Arbor. Jones, S.C., Magee, C.A., 2011. Exposure to alcohol advertising and alcohol consumption among Australian adolescents. Alcohol Alcohol. 46, 630-637. Lee, I.C., Ting, T.T., Chen, D.R., Tseng, F.Y., Chen, W.J., Chen, C.Y., 2015. Peers and social network on alcohol drinking through early adolescence in Taiwan. Drug Alcohol Depend. 153, 50-58. Lobstein, T., Landon, J., Thornton, N., Jernigan, D., 2017. The commercial use of digital media to market alcohol products: A narrative review. Addiction 112 Suppl. 1:21-27 Macklin, M.C., 1994. The effects of an advertising retrieval cue on young children's memory and brand evaluations. Psychol. Market. 11, 291-311. Masten, A.S., Faden, V.B., Zucker, R.A., Spear, L.P., 2008. Underage drinking: A developmental framework. Pediatrics 121, Suppl. 4, S235-251. McClure, A.C., Tanski, S.E., Li, Z., Jackson, K.,Morgenstern, M., Li, Z., Sargent, J.D., 2016. Internet alcohol marketing and underage alcohol use. Pediatrics 137: e20152149. McClure, A.C., Stoolmiller, M., Tanski, S.E., Engels, R.C., Sargent, J.D., 2013. Alcohol marketing receptivity, marketing-specific cognitions, and underage binge drinking. Alcohol. Clin. Exp. Res. 37, Suppl. 1, E404-413. Mokdad, A.H., Forouzanfar, M.H., Daoud, F., Mokdad, A.A., El Bcheraoui, C., Moradi-Lakeh, M., Kyu, H.H., Barber, R.M., Wagner, J., Cercy, K., Kravitz, H., Coggeshall, M., Chew, A., O'Rourke, K.F., Steiner, C., Tuffaha, M., Charara, R., Al-Ghamdi, E.A., Adi, Y., Afifi, R.A., Alahmadi, H., AlBuhairan, F., Allen, N., AlMazroa, M., Al-Nehmi, A.A., AlRayess, Z., Arora, M., Azzopardi, P., Barroso, C., Basulaiman, M., Bhutta, Z.A., Bonell, C., Breinbauer, C., Degenhardt, L., Denno, D., Fang, J., Fatusi, A., Feigl, A.B., Kakuma, R., Karam, N., Kennedy, E., Khoja, T.A., Maalouf, F., Obermeyer, C.M., Mattoo, A., McGovern, T., Memish, Z.A., Mensah, G.A., Patel, V., Petroni, S., Reavley, N., Zertuche, D.R., Saeedi, M., Santelli, J., Sawyer, S.M., Ssewamala, F., Taiwo, K., Tantawy, M., Viner, R.M., Waldfogel, J., Zuñiga, M.P., Naghavi, M., Wang, H., Vos, T., Lopez, A.D., Al Rabeeah, A.A., Patton, G.C., Murray, C.J., 2016. Global burden of diseases, injuries, and risk factors for young people's health during 1990-2013: A systematic analysis for the Global Burden of 16
Disease Study 2013. Lancet 387, 2383-2401. Morgenstern, M., Isensee, B., Sargent, J. D., Hanewinkel, R., 2011. Attitudes as mediators of the longitudinal association between alcohol advertising and youth drinking. Arch. Pediatr. Adolesc. Med. 165, 610–616. Padon, A.A., Rimal, R.N., DeJong, W., Siegel, M., Jernigan, D., 2016. Assessing youth-appealing content in alcohol advertisements: application of a content appealing to youth (CAY) index. Health Commun. Dec. 16, 1-10. Pasch, K.E., Komro, K.A., Perry, C.L., Hearst, M.O., Farbakhsh, K., 2007. Outdoor alcohol advertising near schools: What does it advertise and how is it related to intentions and use of alcohol among young adolescents? J. Stud. Alcohol Drugs 68:587–596. Primack, B.A., Nuzzo, E., Rice, K.R., Sargent, J.D., 2012. Alcohol brand appearances in US popular music. Addiction 107, 557-566. Robinson, T.N., Chen, H.L., Killen, J.D., 1998. Television and music video exposure and risk of adolescent alcohol use. Pediatrics 102, e54. Ross, C.S., de Bruijn, A., Jernigan, D.H., 2013. Do time watersheds for alcohol advertising reduce youth exposure? J. Public Affairs 13,123-129. Ross, C.S., Maple, E., Siegel, M, DeJong, W., Naimi, T.S., Ostroff, J., Padon, A.A., Borzekowski, D.L., Jernigan, D.H., 2014. The relationship between brand-specific alcohol advertising on television and brand-specific consumption among underage youth. Alcohol. Clin. Exp. Res. 38, 2234-2242. Ross, C.S., Maple, E., Siegel, M., DeJong, W., Naimi, T.S., Padon, A.A., Borzekowski, D.L., Jernigan, D.H., 2015. The relationship between population-level exposure to alcohol advertising on television and brand-specific consumption among underage youth in the US. Alcohol Alcohol. 50, 358-364 Sargent, J.D., Wills, T.A., Stoolmiller, M., Gibson, J., Gibbons, F.X., 2006. Alcohol use in motion pictures and its relation with early-onset teen drinking. J. Stud. Alcohol Drugs 67, 54-65. Snyder, L.B., Milici, F.F., Slater, M., Sun, H., Strizhakova, Y., 2006. Effects of alcohol advertising exposure on drinking among youth. Arch. Pediatr. Adolesc. Med. 160, 18-24. The Center on Alcohol Marketing and Youth, 2010. Youth Exposure to Alcohol Advertising on Television, 17
2001-2009. John Hopkins University. USA. Unger, J.B., Schuster, D., Zogg, J., Dent, C.W., Stacy, A.W., 2003. Alcohol advertising exposure and adolescent alcohol use: A comparison of exposure measures. Addict. Res. Theory 11, 177-193. Windle, M., Spear, L.P., Fuligni, A.J., Angold, A., Brown, J.D., Pine, D., Smith, G.T., Giedd, J., Dahl, R.E., 2008. Transitions into underage and problem drinking: Developmental processes and mechanisms between 10 and 15 years of age. Pediatrics 121, S273-S289. World Health Organization, 2014. Global Status Report on Alcohol and Health 2014. World Health Organization Press, Switzerland.
18
Figure Legends
Fig. 1. The percentage of marketing channel-specific past-month exposure to alcohol advertising, by baseline drinking experience (n=1863)
Note: AN: Alcohol naïve; AE: Alcohol experience *P<0.05, **P<0.01, ***P<0.00
19
Table 1 Selected characteristics of participating students at baseline (n=1863). Variablesa,b,c Lifetime drinking status at seventh grade Overall Alcohol naïve Alcohol experienced (n=1863) (n=1050) (n=813) n %wt n %wt n %wt * Gender Male 885 47.32 490 46.09 395 48.9 Female 978 52.68 560 53.91 418 51.1 *** Living with parents Either one or none 319 16.42 136 12.32 183 21.71 Both 1543 83.53 914 87.68 629 78.17 Parental education Both/either below high school 967 50.6 544 50.27 423 51.01 Both college or above 833 46.2 477 47.16 356 44.96 d *** Monthly allowance (NTD ) 0–499 854 45.97 544 52.00 310 38.19 500 or above 966 51.73 480 45.52 486 59.75 *** Parental drinking None 507 27.08 366 34.77 141 17.16 Either or both 1299 69.88 653 62.33 646 79.61 *** Best friend’s drinking None 1524 82.29 958 91.42 566 70.51 At least one 339 17.71 92 8.58 247 29.49 *** Recent exposure to media drinking portrayals None 403 21.47 270 25.51 133 16.26 1–14 days 1124 60.55 640 61.34 484 59.53 15 or more days 336 17.98 140 13.16 196 24.21 *** Intent to drink Definitely/probably no 1324 71.23 951 90.62 373 46.24 Definitely/probably yes 539 28.77 99 9.38 440 53.76 *** Past-year alcohol drinking at 9th grade None 1310 70.66 871 83.18 439 54.5 Yes 553 29.35 179 16.82 374 45.5 *** Occasions of drinking among past-year drinkers at 9th grade Mean (SE) 3.78 0.52 2.67 0.25 4.31 0.63 Min–Max 1 200 1 30 1 200 Q1, Q2, Q3 (1, 2, 4) (1, 2, 3) (1, 3, 5) Note. *P<0.05, **P<0.01, ***P<0.001 a
Variables were assessed at seventh grade (baseline) if not otherwise specified; Columns may not add up to 100% due to missing data; c All variables were estimated and examined by complex survey analyses; d NTD: New Taiwan Dollars, 1 USD = 31 NTD in 2010. b
20
Table 2 Association estimates between alcohol advertisements exposure and alcohol initiationa (n=1050) Crude Model 1 Model 2
Variables
OR
(95% CI)
aOR (95% CI)
aOR
Model 3
(95% CI) aOR
(95% CI)
Alcohol advertisement exposure at 7th gradeb Television (Ref=None) 1–14 days
3.16
(1.56–6.42)**
3.20
(1.54-6.63)**
2.75
(1.39-5.45)**
2.62
(1.14-6.02)*
15 or more days
2.42
(1.22–4.81)*
2.35
(1.22-4.55)*
2.06
(1.07-3.95)*
1.73
(0.68-4.39)
1–14 days
1.41
(0.94–2.14)
1.36
(0.91-2.02)
1.24
(0.81-1.92)
1.09
(0.57-2.12)
15 or more days
1.69
(1.26–2.25)***
1.55
(1.17-2.06)
1.47
(1.07-2.03)*
1.41
(0.86-2.31)
1.28
(0.92–1.79)
1.23
(0.88-1.70)
1.13
(0.79-1.63)
1.09
(0.65-1.81)
Stores (Ref=None)
Websites (Ref=None) 1–14 days
*
1.46
(1.01–2.13)
1.38
(0.97-1.97)
1.31
(0.91-1.89)
1.40
(0.80-2.47)
1–14 days
0.84
(0.69–1.01)
0.84
(0.69-1.02)
0.77
(0.63-0.94)**
0.65
(0.49-0.87)**
15 or more days
1.04
(0.85–1.27)
1.03
(0.86-1.23)
1.03
(0.86-1.22)
0.89
(0.66-1.21)
1–14 days
1.39
(0.78–2.50)
1.35
(0.80-2.28)
1.27
(0.77-2.11)
1.31
(0.69-2.50)
15 or more days
0.97
(0.64–1.47)
0.96
(0.65-1.42)
0.96
(0.62-1.47)
0.91
(0.54-1.53)
1–14 days
0.94
(0.61–1.46)
0.97
(0.65-1.47)
0.99
(0.65-1.50)
0.91
(0.59-1.40)
15 or more days
0.80
(0.50–1.26)
0.78
(0.52-1.16)
0.74
(0.47-1.16)
0.67
(0.37-1.20)
0.96
(0.73–1.27)
0.97
(0.78-1.20)
0.96
(0.39-2.36)
15 or more days Newspapers (Ref=None)
Magazines (Ref=None)
Radio (Ref=None)
Gender (Ref=Female) Male
Living with parents (Ref=Either one or none) Both
0.87
(0.36–2.12)
Parental education (Ref=Both or either one below high school) Both college or above
0.84
(0.61–1.16)
0.92
(0.72-1.19)
0.93
(0.71–1.21)
0.86
(0.66-1.13)
1.81
(1.19–2.76)**
1.70
(1.13-2.56)*
(1.28–2.09)***
1.56
(1.16-2.09)**
Parental drinking (Ref=None) Either or both Best friend’s drinking (Ref-None) At least one
Monthly allowance (Ref=0–499) (NTD)c 500 or above
1.64
Exposure to media drinking portrayals (Ref=None) 1–14 days
1.72
(1.22–2.42)**
1.36
(1.00-1.85)
15 or more days
1.49
(0.87–2.57)
1.18
(0.57-2.43)
1.52
(0.72-3.20)
Intent to drink at 7th grade (Ref=Definitely/probably no) Definitely/probably yes
1.59
(0.85–2.97)
Note. *P<0.05, **P<0.01, ***P<0.001, cOR: crude odds ratio, aOR: adjusted odds ratio, CI: confidence interval
21
a
Alcohol drinking was assessed at 9th grade follow-up Past-month interval, the baseline assessment c NTD: New Taiwan Dollars, 1USD= 31 NTD in 2010 Model 1 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, and best friend’s drinking; the market channel was entered one at a time. Model 2 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, best friend’s drinking, exposure to media drinking portrayals and the intent to drink at seventh grade; the market channel was entered one at a time. Model 3 adjusted for all listed variables. b
22
Table 3 Association estimates between alcohol advertisements exposure and occasional alcohol drinking (n=813).
Continued occasional drinkinga Crude Model 1 cOR (95% CI) aOR (95% CI)
Variables
Model 2 aOR (95% CI)
Model 3 aOR (95% CI)
Alcohol advertisement exposure at 7th gradeb Television (Ref=None) 1–14 days
1.32
(0.83–2.21)
1.15
(0.71-1.85)
1.13
(0.67-1.90)
1.00
(0.58-1.74)
15 or more days
0.97
(0.61–1.55)
0.84
(0.52-1.36)
0.89
(0.52-1.53)
0.90
(0.49-1.65)
1–14 days
0.95
(0.63–1.43)
0.92
(0.60-1.39)
0.85
(0.55-1.32)
0.77
(0.48-1.24)
15 or more days
1.28
(0.84–1.94)
1.04
(0.67-1.61)
1.10
(0.69-1.76)
0.80
(0.48-1.36)
1.80
(1.30–
1.61
(1.17-2.23)**
1.60
(1.15-2.24)**
1.50
(1.04-2.15)*
1.09
(0.76-1.57)
1.11
(0.77-1.60)
1.15
(0.74-1.77)
0.92
(0.69-1.24)
0.88
(0.66-1.18)
0.75
(0.55-1.04)
0.97
(0.61-1.55)
Stores (Ref=None)
Websites (Ref=None) 1–14 days
2.49)*** 15 or more days
1.21
(0.85–1.74)
0.91
(0.68–1.22)
Newspapers (Ref=None) 1–14 days
1.61
(1.09–2.37)
*
1.42
(0.95-2.10)
1.55
(1.04-2.31)
1–14 days
1.45
(1.09–1.92)*
1.29
(0.97-1.72)
1.31
(0.98-1.76)
1.16
(0.84-1.60)
15 or more days
1.16
(0.73–1.86)
1.06
(0.67-1.70)
1.13
(0.71-1.79)
1.04
(0.63-1.72)
1–14 days
1.12
(0.79–1.57)
1.23
(0.88-1.73)
1.11
(0.79-1.55)
1.16
(0.82-1.64)
15 or more days
2.50
(1.62–
2.42
(1.55-3.79)***
2.90
(1.85-4.53)***
2.58
(1.60-4.15)***
1.02
(0.78-1.33)
0.99
(0.71-1.36)
15 or more days
*
Magazines (Ref=None)
Radio (Ref=None)
3.87)*** Gender (Ref=Female) Male
0.84
(0.64–1.09)
Living with parents (Ref=Either one or none) Both
1.12
(0.81–1.54)
Parental education (Ref=Both or either one below high school) Both college or above Parental
0.62
(0.32–1.22)
0.83
(0.63-1.08)
2.18
(0.98–4.87)
1.78
(1.25-2.55)**
(0.89–1.59)
1.09
(0.80-1.47)
1.51
(1.14-1.99)**
drinking
(Ref=None) Either or both
Best friend’s drinking (Ref=None) At least one
1.19
Monthly allowance (Ref=0–499) (NTD) 500 or above
1.16
c
(0.45–2.94)
Exposure to media drinking portrayals (Ref=None) b 23
1–14 days
1.60
(1.09–2.34)*
1.26
(0.83-1.91)
15 or more days
1.26
(0.82–1.95)
1.05
(0.64-1.73)
2.28
(1.72-3.02)***
Intent to drink at 7th grade (Ref=definitely no or probably no) Definitely/probably yes
1.67
(1.28– 2.19)***
Note. *P<0.05, **P<0.01, ***P<0.001, cRR: crude relative risk, aRR: adjusted relative risk, CI: confidence interval a 0–3 occasions as reference group; alcohol drinking was assessed at 9 th grade follow-up; b Past-month interval, the baseline assessment; c NTD: New Taiwan Dollars, 1USD= 31 NTD in 2010; Model 1 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, and best friend’s drinking; the market channel was entered one at a time. Model 2 adjusted for gender, living with parents, parental education, and monthly allowance, parental drinking, best friend’s drinking, exposure to media drinking portrayals at seventh grade, and intent to drink at seventh grade; the market channel was entered one at a time. Model 3 adjusted for all listed variables.
24