The Impact of Alcohol Marketing on Youth Drinking Behaviour: A Two-stage Cohort Study
- 1Institute for Social Marketing, Division of Marketing, University of Stirling, Stirlingshire, UK
- 2Open University Business School, The Open University, Walton Hall, Milton Keynes MK7 6AA, UK
- *Corresponding author: Email: r.gordon{at}open.ac.uk
- Received February 22, 2010.
- Revision received July 20, 2010.
- Accepted July 23, 2010.
Abstract
Aim: To examine whether awareness of, and involvement with alcohol marketing at age 13 is predictive of initiation of drinking, frequency of drinking and units of alcohol consumed at age 15. Methods: A two-stage cohort study, involving a questionnaire survey, combining interview and self-completion, was administered in respondents’ homes. Respondents were drawn from secondary schools in three adjoining local authority areas in the West of Scotland, UK. From a baseline sample of 920 teenagers (aged 12–14, mean age 13), in 2006, a cohort of 552 was followed up 2 years later (aged 14–16, mean age 15). Data were gathered on multiple forms of alcohol marketing and measures of drinking initiation, frequency and consumption. Results: At follow-up, logistic regression demonstrated that, after controlling for confounding variables, involvement with alcohol marketing at baseline was predictive of both uptake of drinking and increased frequency of drinking. Awareness of marketing at baseline was also associated with an increased frequency of drinking at follow-up. Conclusions: Our findings demonstrate an association between involvement with, and awareness of, alcohol marketing and drinking uptake or increased drinking frequency, and we consider whether the current regulatory environment affords youth sufficient protection from alcohol marketing.
INTRODUCTION
That many adolescents have used or do you use alcohol is beyond equivocacy. In most countries within the European Union (EU), for instance, more than 70% of youth (15–16 years) admit to drinking alcohol within the previous year, and over 50% within the past month. Further, in the UK, levels of youth binge drinking and past-year and past-month drunkenness are considerably higher than in the rest of the EU (Hibell et al., 2009). So too are the levels of consumption, which have almost doubled between 1990 and 2007 in England (Fuller, 2009). It is these hazardous youth drinking behaviours that represent a major public health concern, given the possible injurious consequences (HM Government, 2007), including poor educational performance, risky sexual behaviour and teenage pregnancy (Newbury-Birch et al., 2009; OECD, 2009), crime and disorder (Hibell et al., 2009; Home Office, 2004) and a range of physical and psychological harms (HES, 2007; Scottish Government, 2010). Additionally, using alcohol at an earlier age is a predictor of future dependency (Bonomo et al., 2004; Newbury-Birch et al., 2009).
Many protective and risk factors have been identified for youth drinking uptake and behaviour. Alcohol marketing has been suggested as one of these risk factors (Babor et al., 2003), with recent systematic reviews appearing to support this assertion (Anderson et al., 2009; Smith and Foxcroft, 2009). This has led to some within the public health field calling for a complete ban on alcohol marketing, arguing that it is pervasive and linked with youth drinking initiation, consumption levels and continued drinking (Anderson, 2009; BMA, 2009; Godlee, 2009). A recent meta-review, however, conducted in the UK on behalf of the Department for Children, Schools and Families, does not even consider marketing among the many risk factors identified (Newbury-Birch et al., 2009). Although this seems an important omission, there is a paucity of research exploring the relationship between alcohol marketing and youth drinking behaviour in Europe, and in the UK an absence of longitudinal research—a more powerful design that allows greater confidence when exploring potentially causal links (Gunter et al., 2009; Prime Minister's Strategy Unit, 2004). Highlighting this point, Anderson et al.'s (2009) systematic review of the existing longitudinal research found that 10 of the 13 studies identified were from the USA, one was from New Zealand and only two from Europe; in Belgium and Germany. The European Commission department concerned with health, DG SANCO, also acknowledged the lack of European studies, and in response to this has recently funded a multi-country EU study called the ‘Amphora Project’ (European Commission, 2009) as well as the aforementioned German study (Hanewinkel and Sargent, 2009).
To address this gap in the literature we present findings from a UK cohort study. Funded as part of the National Preventive Research Initiative (NPRI), the study examines the cumulative impact of alcohol marketing communications on youth drinking during the period when most adolescents start experimenting with alcohol, from age 13 to 15 (Black et al., 2009). In addition, and unlike most research in this area, we also examined non-traditional alcohol marketing channels such as new media, sponsorship and e-marketing.
METHODS
Design
Data come from two waves of a cohort study called Assessing the Cumulative Impact of Alcohol Marketing on Youth Drinking. The baseline was conducted from October 2006 to March 2007 and the follow-up was conducted 2 years later, from October 2008 to March 2009. The study design was adapted from research on tobacco marketing in the UK (MacFadyen et al., 2001). Questionnaire development was informed by extensive formative research and pre-testing (Gordon et al., 2010). Cross sectional data from baseline are reported elsewhere (Gordon et al., in press).
Setting and sample
The study was conducted within three local authority areas in the West of Scotland. The baseline sample was recruited via an information pack (containing an information sheet, parental and respondent consent forms and offering a small gift token for participation) sent to the homes of all second year pupils (12–14 years, mean age 13) attending state secondary schools in each local authority area. In one local authority area, the invitations were mailed to respondents’ homes. In two local authorities, schools were asked to distribute the packs to pupils to take home. Of the 920 baseline respondents, a cohort of 552 were followed up 2 years later when in fourth year (14–16 years, mean age 15).
Baseline characteristics of sample
The achieved cohort sample was evenly distributed by gender, 50% male (n = 275) and 50% female (n = 277) (Table 1a). Social grade, classified using the National Readership social grading system, was based on the occupation of the chief income earner. Approximately two-fifths (41%, n = 224) were ABC1 (middle class) and approximately three-fifths (59%, n = 326) were C2DE (working class) (Wilmshurst and MacKay, 1999), which is largely consistent with national census data (45.6% ABC1, 54.4% C2DE; GROS, 2001). Sample ethnicity was predominantly white 94% (n = 515), with 3% (n = 19) identifying themselves as Asian, 1% mixed race (n = 7), 1% black (n = 6), <1% Chinese (n = 1) and <1% other (n = 1). For religious identification, most of the sample were Christian 65% (n = 354) or had no religiosity 31% (n = 169), with 3% Muslim (n = 19) and 1% other (n = 5).
Baseline characteristics and description of dependent and independent variables included in logistic and multiple regression models
Baseline characteristics of cohort vs drop-out sample
Compared with the cohort that was successfully followed up, the drop-out sample had a higher proportion of girls (50% girls in follow up sample, 57% girls in sample lost to attrition, P < 0.05) and a higher proportion of middle class (ABC1) respondents (41% ABC1 in follow up sample, 55% ABC1 in sample lost to attrition, P < 0.001). The cohort did not differ from the drop-out sample in terms of drinking status, age, ethnicity or religion.
Data collection
Fieldwork comprised face-to-face interviews conducted in-home, by professional interviewers, with an accompanying self-completion questionnaire to gather sensitive data on drinking behaviour. Respondent confidentiality and anonymity of personal data were assured. Parental permission and participant consent were obtained prior to interview at each wave. Numbered show cards were used throughout the interviewer administered the questionnaire to maximize privacy and enable respondents to answer freely without fear of conveying their answers to others who may be present in the room. Participants sealed their self-completion questionnaire in an envelope before handing it to the interviewer. Ethical approval was obtained from the University of Stirling Ethics Committee and interviewers adhered to the Market Research Society Code of Conduct (MRS, 2005).
Measures
The measures employed in this study were based upon a review of the youth drinking literature. Owing to constraints in terms of space within the questionnaire and the requirement to include a number of measures assessing marketing awareness and involvement, some measures, such as parental control, were not included in the study.
Demographic and school-related control variables
Data were recorded on age, gender, social grade (based upon occupation of chief income earner), ethnicity and religion. Liking of school was rated on a five-point scale, from ‘dislike school a lot’ (1) to ‘like school a lot’ (5). Rating of own school work in relation to others in their year was also rated on a five-point scale, from ‘a lot worse’ (1) to ‘a lot better’ (5).
Drinking-related control variables
Drinking among parents, siblings and friends was assessed using four self-completion items. Participants were asked whether their mother (father) drinks alcohol nowadays, with four responses to each item: yes, no, not sure and I do not have/don't see my mother (father). Those who indicated they had brothers or sisters were asked whether any of their brothers or sisters drink alcohol: yes, no and don't know. Participants were asked to indicate, on a five-point scale, how many of their friends drink alcohol once a week: all of them, most of them, about half of them, a few of them, none of them and not sure. Perceptions of others’ views on them trying alcohol was assessed using three self-completion items, which asked whether their brother(s) or sister(s), parents or closest friends would consider it ok or not ok for them to ‘try drinking alcohol to see what it's like’. Response categories were: ok, not ok and don't know, which were combined following principal components analysis at baseline and are reported elsewhere (Gordon et al., in press).
Drinking behaviour
Drinking status was assessed by asking the question: ‘Have you ever had a proper alcoholic drink—a whole drink, not just a sip?’ Those answering affirmatively were classified as drinkers, and those who had not done so as non-drinkers.
Uptake of drinking was based on changes in drinking status between waves and was coded as (1) for baseline non-drinkers who were drinkers at follow-up and coded as (0) for baseline non-drinkers who remained non-drinkers at follow-up.
Number of alcoholic units consumed last time respondents had an alcoholic drink was calculated by estimating the amount, in millilitres, of each type of alcoholic drink consumed and the ABV of each drink, based on responses to the following: brand or name of drink(s) consumed, type(s) of alcohol consumed (e.g. beer, wine, vodka), drinking vessel(s) used (recorded using a visual) and the amount of each drink consumed (more than one full bottle/can/glass, one full bottle/can/glass, 3/4, 1/2, 1/4 or less than 1/4 of a bottle/can/glass).
Frequency of drinking was recorded by asking respondents how often they usually had an alcoholic drink (daily, twice per week, weekly, fortnightly, monthly, only a few times per year or I never drink alcohol now).
Alcohol marketing
Alcohol marketing awareness was assessed for 15 types of marketing identified from formative research (Gordon et al., 2010). Participants were shown a series of 15 cards with examples of different forms of alcohol marketing (Table 2a) and were asked to indicate whether or not they had come across alcohol being marketed in each of these ways. The number of channels through which participants had noticed marketing was calculated by counting the number of positive responses for each of the 15 channels listed in Table 2a.
Adolescents’ awareness of and involvement in alcohol marketing at baseline
Involvement in alcohol marketing was assessed by showing participants eight cards with examples of different forms of alcohol promotional activities and asking them to indicate whether or not they had participated in each. The amount of alcohol marketing participated in was calculated by counting the number of positive responses for each of the eight forms listed in Table 2b.
Liking of alcohol advertising was measured on a five-point scale, from ‘dislike alcohol adverts a lot’ (1) to ‘like alcohol adverts a lot’ (5).
Statistical analysis
The analyses looked at four outcome variables—uptake of drinking, uptake of fortnightly drinking, uptake of monthly drinking and units of alcohol consumed at follow-up. For each of these outcomes, four models were run to separately examine their potential association with amount of alcohol marketing aware of, amount of alcohol marketing involved in, the number of brands recalled and appreciation of alcohol advertising, all measured at baseline. Table 1 shows the dependent and independent variables used within a series of logistic regression models that were run to examine the association between baseline characteristics and uptake of drinking (models 1–4) and uptake of frequent drinking (models 5–12). Table 1 also shows the multiple regression models (models 13–16) that were run to examine the association between baseline characteristics and the amount of alcohol consumed at follow-up. In the logistic and multiple regression models, independent variables were entered in blocks, using forward likelihood ratio, with the marketing variable of interest entered in the final block to examine the potential contribution after important confounding variables had been considered.
Among those who were non-drinkers at baseline (n = 350), logistic regression was used to examine which baseline characteristics were associated with uptake of drinking by follow-up. Uptake of drinking was used as the dependent variable (1 = started drinking by follow-up, 0 = remained non-drinker at follow-up). In each of the logistic regression analyses, several potentially confounding variables were controlled for and entered in the following blocks: (1) drinking among siblings, friends, mother and father; (2) gender, age, social grade, ethnic group and religion; (3) perceptions of siblings’, parents’, friends’ views on whether it was ok to try drinking; (4) liking of school, perception of own school work relative to others in their year, liking of adverts in general. Four separate models were run. In each case, the control variable in the final block was varied as follows: the number of alcohol-marketing channels' respondents were aware of (Model 1); involvement with alcohol marketing (Model 2); number of brands recalled (Model 3); appreciation of alcohol advertising (Model 4).
Logistic regression was also used to examine the frequency of drinking. This was examined at two levels: (1) uptake of fortnightly drinking and (2) uptake of monthly drinking (among those who, at baseline, did not drink at all or drank less than fortnightly or monthly, respectively). Owing to the small sample size, these analyses also included uptake of at least fortnightly (or monthly) drinking among baseline non-drinkers rather than just increased frequency among existing drinkers. Independent variables were again entered in blocks to control for potentially confounding variables. The first block controlled for baseline drinking status and subsequent blocks controlled for the same variables included in the analysis of uptake of drinking.
Among those who were drinkers at follow-up (n = 342), multiple regression was used to examine the relationship between baseline characteristics and units consumed at follow-up (see Table 1, Models 13–16). Blocks 1 and 2 controlled for baseline drinking status and baseline units consumed, respectively, and subsequent blocks controlled for the same variables included in the analysis of uptake of drinking. Four separate models were run, again varying the control variable in the final block; number of forms of alcohol marketing aware of at baseline (Model 13); number of forms of alcohol marketing involved in at baseline (Model 14); number of brands recalled at baseline (Model 15); baseline appreciation of alcohol advertising (Model 16).
In the logistic and multiple regression models, cases were excluded if they had missing data on one or more of the variables being assessed in the model. The number of excluded cases in any analyses ranged from 16 to 32, representing a very small portion of the eligible sample in each (5–6%).
RESULTS
Alcohol drinking behaviour
At follow-up, 62% (n = 342) reported having tried an alcoholic drink. This is lower than the prevalence (81%) from national survey data (Black et al., 2009). Mean age for consumption of first alcoholic drink was 13.4 years (SD = 1.44) and mean number of units consumed for last drink at follow-up was 7.12 (SD = 7.37).
Awareness of alcohol marketing
At baseline, there was very high awareness of alcohol marketing, with 97% aware of at least one form of alcohol marketing. The sample was aware of, on average, five marketing channels (see Table 2a). Awareness was measured across a range of channels, including advertisements and promotions, sports-related marketing, electronic communications and arts-related marketing, with awareness highest for TV/cinema advertising (77%), branded clothing (67%), sports sponsorship (63%), price promotions (59%) and signs or posters in-store (58%).
Involvement with alcohol marketing
At baseline, more than half (56%) had participated in at least one form of alcohol marketing. The most common types of alcohol-marketing respondents that were involved with were ownership of alcohol branded clothing (45%) and free branded gifts (11%) (see Table 2b).
Association between alcohol marketing and initiation of drinking
Among the 350 who were non-drinkers at baseline, 47% (n = 165) started drinking between baseline and follow-up. Logistic regression demonstrated a significant association between the amount of alcohol marketing that non-drinkers were involved in at baseline and their uptake of drinking at follow-up (Model 2; see Table 3). Involvement with alcohol marketing at baseline increased their chance/risk of initiation of drinking at follow-up (adjusted OR = 1.31, P < 0.05). Other factors associated with uptake of drinking were having siblings who drink (adjusted OR = 1.97, P < 0.05 compared with having non-drinking siblings) and holding more positive perceptions that others consider it ok for them to drink (adjusted OR = 1.19, P < 0.01). Uptake of drinking was less likely among non-white ethnic groups (adjusted OR = 0.1, P < 0.01). A further logistic regression, Model 4, indicated that initiation of drinking was also more likely among those with greater appreciation of alcohol advertising at baseline (adjusted OR = 1.272, 95% CI 1.005–1.610, P < 0.05). After controlling for confounders, no association was found between uptake of drinking and baseline awareness of alcohol marketing (Model 1) or number of brands recalled at baseline (Model 3).
Logistic regression of association between amount of involvement in alcohol marketing and uptake of drinking (dependent variable = uptake of drinking (1 = initiated drinking, 0 = remained non-drinker) and base: all in cohort who were non-drinkers at baseline)
Association between alcohol marketing and frequency of drinking
Among the 513 who were non-drinkers or drank less often than fortnightly at baseline, 15% (n = 78) had taken up more frequent drinking (at least fortnightly) at follow-up. Logistic regression found that uptake of fortnightly drinking was more likely among those with a higher involvement with alcohol marketing at baseline (adjusted OR = 1.43, P < 0.05; Model 6, see Table 4). Other factors associated with becoming a fortnightly drinker were being a drinker at baseline (adjusted OR = 1.85, P < 0.05) and holding more positive perceptions that others consider it acceptable for them to drink (adjusted OR = 1.18, P < 0.05). Those who indicated a religious affiliation were less likely than those with no religious affiliation to become fortnightly drinkers (adjusted OR = 0.57, P < 0.05). Further, logistic regressions (Model 5) indicated that uptake of fortnightly drinking was also more likely among those with greater awareness of alcohol marketing at baseline (adjusted OR = 1.11, 95% CI 1.005–1.234, P < 0.05; Model 5) and those with greater appreciation of alcohol marketing at baseline (adjusted OR = 1.295, 95% CI 1.002–1.674, P < 0.05; Model 8). After controlling for confounders, no association was found between uptake of fortnightly drinking at follow-up and number of brands recalled at baseline (Model 7).
Logistic regression of association between amount of involvement in alcohol marketing at baseline and drinking becoming more frequent (drinking at least fortnightly) at follow-up (dependent variable = whether had become fortnightly drinker (or more frequent) at follow-up (1 = yes became fortnightly (or more frequent) drinker, 0 = remained non-drinker or less than fortnightly drinker) and base: all in cohort who were non-drinkers at baseline or drank less often than fortnightly)
Among the 498 who were non-drinkers or drank less often than monthly at baseline, 23% (n = 115) had taken up more frequent drinking (at least monthly) at follow-up. As shown in Table 5, uptake of monthly drinking was more likely among those with a higher involvement with alcohol marketing at baseline (adjusted OR = 1.33, P < 0.05; see Table 5, Model 10). Becoming a monthly drinker was also associated with believing that others consider it acceptable for them to try drinking (adjusted OR = 1.25, P < 0.001), having siblings who drink (adjusted OR = 2.06, P < 0.01) and having a mum who drinks (adjusted OR = 1.88, P < 0.05). Those indicating a religious affiliation were less likely than those with no religious affiliation to take up monthly drinking (adjusted OR = 0.58, P < 0.05). After controlling for confounders, no association was found between uptake of monthly drinking at follow-up and baseline awareness of alcohol marketing (Model 9), number of brands recalled at baseline (Model 11) or baseline appreciation of alcohol advertising (Model 12).
Logistic regression of association between amount of involvement in alcohol marketing at baseline and drinking becoming more frequent (drinking at least monthly) at follow-up (dependent variable = whether had become monthly drinker (or more frequent) at follow-up (1 = yes became monthly (or more frequent) drinker, 0 = remained non-drinker or less than monthly drinker) and base: all in cohort who were non-drinkers at baseline or drank less often than monthly)
Alcohol marketing and units of alcohol consumed last time had a drink
Multiple regression analysis, controlling for demographics, baseline drinking status, amount consumed at baseline and other drinking related variables found no association between units consumed at follow-up and baseline measures of awareness or involvement in alcohol marketing, number of brands recalled or appreciation of alcohol advertising (Models 13–16).
DISCUSSION
The findings show a small but significant association between awareness of and involvement with alcohol marketing, and youth drinking behaviour, even after controlling for important confounding variables. They also show a small but significant association between appreciation of alcohol advertising and youth drinking behaviour. Marketing is of course only one of a number of variables that can influence youth drinking with other factors such as family drinking and peer influence also significant, often to a greater degree. However, our findings from the UK are consistent with previous research and add further weight to there being an association between alcohol marketing and youth drinking behaviour (Anderson et al., 2009; Meier et al., 2008), with higher awareness of alcohol marketing at baseline, predicting increased frequency of drinking at follow-up. This dose–response relationship is also consistent with that found with awareness of tobacco marketing and tobacco consumption among young people (Davis et al., 2008).
Unlike most previous research, we examined the influence of alcohol marketing across a wide range of marketing channels, including new media, sponsorship and e-marketing, helping to demonstrate the extent, nature and reach of contemporary alcohol marketing in the UK. Indeed, at baseline, young people were aware of an average of five alcohol marketing channels. Previous research has found associations between channels such as TV, print advertising and in-store promotion (Ellickson et al., 2005; Snyder et al., 2006; Stacey et al., 2004) and youth drinking behaviours. Although the sample size in the current study does not allow sufficient power to detect the effect of individual marketing channels on drinking behaviour, some channels are clearly more prominent than others. Almost two-thirds of youth (63%) were aware of sports sponsorship and 45% owned alcohol-branded clothing, which is most likely due to ownership of football shirts from the two major football teams in the area, which are sponsored by a beer brand (Gordon et al., in press). This is a relationship which is concerning given the appeal of sport to young people (Stainback, 1997).
At baseline, 12% of the cohort was aware of alcohol marketing on social networking sites, and 7% accessed alcohol marketing through this channel. Interestingly, although not reported in the results at follow-up awareness of (34%) and involvement with (18%) social networking sites increased markedly, which is testament to the growth of new media as a marketing tool. This is disconcerting, if not entirely surprising, given that a recent report by Ofcom found that approximately half of 11–17 year olds have a social networking profile (Ofcom, 2008). Furthermore, the level of awareness of (23%) and involvement with (6%) alcohol-branded mobile phone/computer screensavers at baseline illustrates the reach of alcohol marketing across a range of communication channels. The opportunity to enjoy 24 h connectivity through mobile web browsing restricts the ability to monitor new media use and control the level of exposure to content such as alcohol marketing. Given that technological advancement in new media develops at such pace regulation tends to lag behind, which gives rise to concerns over the impact alcohol marketing in new media has on young people (BMA, 2009).
Our findings point to the need for additional research on the impact of new media, and other less researched forms of alcohol marketing such as sponsorship (House of Commons Health Committee, 2010), to help assess the cumulative effect of all alcohol marketing on youth drinking (Hastings et al., 2005). Our measures of awareness and involvement in alcohol marketing were based on self-report measures modelled on successful approaches used in the tobacco marketing field (MacKintosh et al., 2008) and reflect young people's recall of the different forms of alcohol marketing that they may have been exposed to. The approach provides a valuable insight into the extent of awareness of and involvement with different forms of alcohol marketing. However, given that we did not assess volume of exposure, e.g. number of hours exposed to TV advertising (Jernigan et al., 2007; Chung et al., 2010), then further research exploring the level of exposure to alcohol marketing and association with youth drinking in the UK would also be welcome. Finally, cohort studies tracking young people through to adulthood would also help provide information on the longer term effects of alcohol marketing once adulthood is reached.
The study is not without limitations. Four main issues limit the generalizability of the findings: (i) the study location, which was confined to the West of Scotland; (ii) the small, albeit significant, effect size of alcohol marketing on drinking behaviour featuring fairly wide confidence intervals; (iii) the relatively small cohort sample; (iv) loss of respondents due to attrition. However, there are reasons to suggest that each of these potential limiting factors have not had a significant effect on the study findings; (i) despite the study location, awareness of alcohol marketing in conventional and electronic media, and sports sponsorship, is unlikely to differ significantly across the UK; (ii) although the alcohol industry criticises research finding only a small causal effect between marketing and drinking behaviour on the grounds that it does not consider other factors influencing alcohol behaviour (ICAP, 2003), we did examine and control for multiple predictor variables within the analyses; (iii) despite a relatively small sample our findings are consistent with previous longitudinal research from outside the UK; (iv) while the baseline gender and social grade characteristics of the cohort differed from those lost to attrition, these characteristics were controlled for in each analysis and there were no differences, between those followed up and those lost to attrition in terms of drinking status, age, ethnicity or religion.
Our findings and the existing evidence base have important implications for the regulation of alcohol marketing in the UK, and indeed elsewhere. The current co-regulatory system employed in the UK appears to provide inadequate protection for youth from exposure to alcohol marketing. Co-regulation is incontrovertibly preferable to self-regulation, which has been found to be ineffective for smoking (Saloojee and Hammond, 2001) and gambling (McMillen and Toms, 1998), but it is reliant upon industry co-operation. In the face of research demonstrating the impact of alcohol marketing on youth, and accepting that the primary imperative for alcohol companies is to increase profit and market share, regulators must act accordingly.
Policy options available include more stringent regulation controlling the content and level of exposure to alcohol marketing across all channels, including new media, sponsorship and e-marketing. The ‘Loi Evin’ in France, for instance, is an example of more robust regulation which restricts alcohol marketing content and exposure, particularly with regards to sports sponsorship (Rigaud and Craplet, 2004). However, such a framework would require explicit guidance on what is allowed, rather than merely stating what is forbidden, in order to avoid ambiguity. Another option is a complete ban on some forms of alcohol marketing (Anderson, 2009; Gilmore, 2009), although even here it would be imprudent to ignore lessons from the tobacco field. Restrictions in some forms of tobacco marketing only have a marginal impact on behaviour as tobacco companies simply reallocate marketing spend to unregulated channels (Davis et al., 2008). What is clear is that the evidence and current focus on alcohol marketing as a public health concern suggests that the time for a considered policy response is now (House of Commons Health Committee, 2010).
Funding
This research was funded by a grant from the National Prevention Research Initiative [The National Preventive Research Initiative (NPRI) is an initiative to support high-quality research aimed at identifying effective approaches to reduce risk factors and influence health behaviour in order to positively impact upon the incidence of new cases of major preventable diseases. The initiative is supported by a consortium of major research bodies and charities, including the Economic and Social Research Council; Medical Research Council; British Heart Foundation; Cancer Research UK; Department of Health; Diabetes UK; Food Standards Agency; Research and Development Office for the Northern Ireland Health and Social Services; Chief Scientist Office, Scottish Executive Health Department; Wales Office of Research & Development; World Cancer Research Fund.], grant number G0501282.
Conflict of interest statement. None declared.
Acknowledgements
We thank the respondents and acknowledge the support of the field-force team and the staff at all participating local authorities and schools; and Professor Gerard Hastings and Douglas Eadie at the Institute for Social Marketing, University of Stirling and Dr Fiona Harris at the Open University for helping with the design and development of this project.
- © The Author 2010. Published by Oxford University Press on behalf of the Medical Council on Alcohol. All rights reserved
