Awareness of alcohol marketing, ownership of alcohol branded merchandise, and the association with alcohol consumption, higher-risk drinking, and drinking susceptibility in adolescents and young adults: a cross-sectional survey in the UK

Objectives To explore awareness of alcohol marketing and ownership of alcohol branded merchandise in adolescents and young adults in the UK, what factors are associated with awareness and ownership, and what association awareness and ownership have with alcohol consumption, higher-risk drinking and susceptibility. Design Online cross-sectional survey conducted during April–May 2017. Setting The UK. Participants Adolescents and young adults aged 11–19 years in the UK (n=3399). Main outcome measures Alcohol Use Disorders Identification Test–Consumption (AUDIT-C) (0–12) and indication of higher-risk consumption (>5 AUDIT-C) in current drinkers. Susceptibility to drink (yes/no) in never drinkers. Results Eighty-two per cent of respondents were aware of at least one form of alcohol marketing in the past month and 17% owned branded merchandise. χ2 tests found that awareness of marketing and ownership of branded merchandise varied within drinking variables. For example, higher awareness of alcohol marketing was associated with being a current drinker (χ2=114.04, p<0.001), higher-risk drinking (χ2=85.84, p<0.001), and perceived parental (χ2=63.06, p<0.001) and peer approval of consumption (χ2=73.08, p<0.001). Among current drinkers, multivariate regressions (controlling for demographics and covariates) found that marketing awareness and owning branded merchandise was positively associated with AUDIT-C score and higher-risk consumption. For example, current drinkers reporting medium marketing awareness were twice as likely to be higher-risk drinkers as those reporting low awareness (adjusted OR (AOR)=2.18, 95% CI 1.39 to 3.42, p<0.001). Among never drinkers, respondents who owned branded merchandise were twice as likely to be susceptible to drinking as those who did not (AOR=1.98, 95% CI 1.20 to 3.24, p<0.01). Conclusions Young people, above and below the legal purchasing age, are aware of a range of alcohol marketing and almost one in five own alcohol branded merchandise. In current drinkers, alcohol marketing awareness was associated with increased consumption and greater likelihood of higher-risk consumption. In never drinkers, ownership of branded merchandise was associated with susceptibility.

Adolescents and young adults (young people) are a focal population for alcohol research 6 because consumption at this stage of development is associated with increased consumption 7 and risk of concomitant harms in later adulthood [1,2]. Global estimates indicate that 8 consumption by young people is particularly high in Europe, where the proportion of current 9 drinkers (69.5%) is higher than the five other global regions, and the proportion of lifetime 10 abstainers is lower (15.9%) [3]. In England, it is estimated that approximately half of 11-15-11 year olds (44%) have consumed an alcoholic drink, one-in-ten have consumed in the past 12 week, and nine percent have been drunk in the past month [4]. Similar estimates are reported 13 in Scotland and Wales [5,6]. Understanding the drivers of alcohol consumption in young 14 people is important given the immediate and long-term individual, social, and economic 15 consequences associated with higher-risk drinking [7]. 16 One factor routinely cited as shaping alcohol-related attitudes and behaviours in young 17 people is marketing [8,9]. Marketing is fundamentally important to alcohol producers. It 18 represents the primary method of communicating with new and existing consumers, can 19 directly encourage sales, and can increase brand salience over competitors. Accordingly, 20 alcohol companies have used highly visible marketing for over 100 years [10], with the 21 current UK landscape characterised by a complex network of mass media marketing (e.g. 22 television), alternative marketing (e.g. sponsorship), consumer marketing (e.g. price), and 23 stakeholder marketing (e.g. retailers) [9]. The importance of marketing to the alcohol industry 24 is evidenced through their annual investment, with Diageo's global marketing expenditure 25 approximately £1.8 billion [11]. Continued consolidation in the alcohol industry has also seen 26 marketing budgets, economies of scale, and intense competition [12]. 2 Content analysis research, which focuses on marketing output as the unit of analysis, 3 consistently reports that marketing may reach and influence young people. For example , 4 marketing has been reported in media environments where young people may be exposed, 5 including sports [13], social media [14], print media [15], and on-screen [16,17]. Content 6 research has also found that marketing may appeal to young people through creative designs, 7 topical and real-world associations, and positive connotations around consumption such as 8 sociability or desirable lifestyles [18,19]. It has also been suggested that commercial 9 marketing contains ambiguous messages about lower-risk consumption [20,21]. 10 Systematic reviews of consumer research, which focus on the individual as the unit of 11 analysis, provide consistent evidence that awareness of, and participation with, marketing has 12 a causal influence on consumption, including initiation and frequency of drinking [22,23]. 13 Message interpretation research has also attempted to move the debate on from whether 14 marketing is associated with consumption and onto how this influence occurs, by identifying 15 psychological mechanisms which mediate the relationship between exposure and 16 consumption [24,25]. 17 In the UK, the influence of alcohol marketing on young people, and the appropriate 18 regulatory response, has been a topic of debate for decades [9,26]. There are, however, 19 unresolved issues which have inhibited attempts to move the debate forward. The last large-20 scale assessment of young people's awareness of alcohol marketing in the UK is a decade 21 old, was only conducted in Scotland, only sampled adolescents under the minimum purchase 22 age, only considered overall marketing awareness (not frequency), and did not consider 23 higher-risk consumption [27,28]. In this study, we explore frequency of awareness for 24 alcohol marketing and ownership of alcohol branded merchandise in young people above and 25 SnapChat, Instagram or other social media); (6) famous people in films, music videos or TV 2 or pictured in magazines; (7) sports, games, or events sponsorship; (8) special price offers; 3 and (9) competitions or prize draws. For each channel, a six-point Likert scale was used to 4 measure frequency of noticing marketing in the past month (1=Everyday -6=Not in the past 5 month; Not sure). 6 For each channel, the self-reported frequency of awareness was converted into the 7 estimated number of days that marketing had been seen in a four-week period (i.e. 'one 8 month'). For example, an answer of 'everyday' for television advertising equated to 28 9 instances of awareness over four weeks (i.e. seven days a week multiplied by four). Scores 10 across the nine channels were summed to create an aggregate score, providing an 11 approximation of the number of times that participants had noticed alcohol marketing in the 12 past month. This aggregate score could only be computed for cases where a valid answer had 13 been given for all nine channels. If a participant answered 'not sure' to any of the nine 14 channels they were coded as 'not stated' for the aggregate score. The aggregate score for the 15 valid sample was split into tertiles of low (aggregate score <16), medium  and high 16 awareness (>54). 17 18

Ownership of alcohol brand merchandise 19
Participants were asked 'Do you own any merchandise (such as clothing or drinks glasses) 20 that show an alcoholic drink brand or logo?' (Yes/No/Not sure). 21 22

Drinking status 23
Participants were asked 'Have you ever had a whole alcoholic drink? Not just a sip.' Those 24 who answered 'No' were classed as never-drinkers while those who answered 'Yes' were 25 The following demographic and confounding variables were also included in initial blocks: 3 age; gender; ethnicity; IMD quintile; resident country; educational status; working status; 4 living status; frequency of mother (female carer), father (male carer) and close friend 5 drinking; perceived parental and peer acceptability of consumption; and drink age of first 6 drink. Categorical variables with >3 categories were converted into dummy (binary) variables 7 to aid interpretation and comparison. The omitted dummy variable formed the reference 8 category. For example, marketing awareness was a categorical variable with four levels: low, 9 medium, high, and not stated. Four binary variables were computed: low awareness, medium 10 awareness, high awareness, and not stated (each coded Yes=1, No=0). By including medium, 11 high and not stated in the regression analysis, and omitting low awareness, the reference 12 category was low awareness. The regressions therefore indicate the association between level 13 of consumption and medium awareness, relative to low awareness, and high awareness 14 relative to low awareness. Reference categories for each variable are displayed in results 15 (Table 3). 16 Two hierarchical logistic regressions were conducted with higher-risk drinking 17 (AUDIT-C >5) among current drinkers and susceptibility to drink among never-drinkers as 18 the dependent variables. Marketing awareness and ownership of branded merchandise were 19 the key independent variables. Where applicable, both logistic regressions controlled for the 20 same demographic and confounding variables as the linear regression. Reference categories 21 for categorical independent variables are indicated in the results (Tables 4 and 5 noticed such adverts at least weekly. Lowest awareness was for adverts in the print press 1 (Median 0 instances per month, IQR=6), on radio (Median 0, IQR = 0) and competitions 2 (Median 0, IQR=2). For each of the channels, a fifth or more (range: 19-29%) were not sure 3 how often, if at all, they had come across alcohol marketing. Overall, 82% had noticed 4 marketing through least one channel. The median aggregate alcohol marketing score was 32 (IQR=60), equating to noticing 32 10 instances of alcohol marketing in the past month (under minimum purchase age: median=28; 11 IQR=60). When categorised into tertiles, 35% of the valid sample were classified as having 12 low awareness (<16 instances per month), 32% had medium awareness , and 34% 13 had high awareness (>54). In those under the minimum purchase age, 38% had low 14 awareness, 31% medium, and 32% high. Bivariate Chi-square tests found that higher 15 awareness was significantly associated with being male, of legal purchasing age, a current 16 drinker, a higher-risk drinker, not in education, in employment, and perceiving parents and 17 peers would consider it okay to consume (Table 2). High awareness was also associated with 18 greater frequency of mother (female carer) consumption, χ 2 (16)=38. 25

17
One factor routinely cited as shaping alcohol-related attitudes and behaviours in young 18 people is marketing [8,9]. Marketing is fundamentally important to alcohol producers. It 19 represents the primary method of communicating with new and existing consumers, can 20 directly encourage sales, and can increase brand salience over competitors. Accordingly, 21 alcohol companies have used highly visible marketing for over 100 years [10], with the current 22 UK landscape characterised by a complex network of mass media marketing (e.g. television), 23 alternative marketing (e.g. sponsorship), consumer marketing (e.g. price), and stakeholder 24 marketing (e.g. to retailers) [9]. The importance of marketing to the alcohol industry is 25 evidenced through their annual investment, with Diageo's global marketing expenditure 26 approximately £1.8 billion [11]. Continued consolidation in the alcohol industry has also seen

22
In the UK, the influence of alcohol marketing on young people has been a topic of debate 23 for decades [9,28]. These debates are further supplemented by concerns about the efficacy and 24 effectiveness of self-regulation, the predominant approach employed to control alcohol 25 marketing in the UK. This includes suggestions that self-regulation provides inadequate 7 Those who answered 'No' were classed as never-drinkers while those who answered 'Yes' 8 were classed as ever-drinkers. 9 10 Alcohol consumption and higher-risk drinking 11 Alcohol use was measured through the Alcohol Use Disorders Identification Test- 21 All other respondents (i.e. those stating 'never' for frequency of consumption) were classified 22 as non-drinkers and were not asked to complete the final two items. In current drinkers, a total 23 AUDIT-C score was computed by summing the three AUDIT-C items (0-12), with a cut-off of 24 >5 used to identify higher-risk consumption [45].

23
The findings are consistent with suggestions that alcohol marketing appears in contexts 24 which may reach young people, including those under the legal purchasing age [8,9].

23
The findings are consistent with suggestions that alcohol marketing appears in contexts 24 which may reach young people, including those under the legal purchasing age [8,9].

Other information
Funding 22 Give the source of funding and the role of the funders for the present study and, if applicable, for the original study on which the present article is based 21 *Give information separately for cases and controls in case-control studies and, if applicable, for exposed and unexposed groups in cohort and cross-sectional studies.  1  2  3  4  5  6  7  8  9  10  11  12  13  14  15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  31  32  33  34  35  36  37  38  39  40  41  42  43  44  45  46