The European Journal of Public Health, Vol. 27, No. 4, 699–704 The Author 2017. Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. doi:10.1093/eurpub/ckw263 Advance Access published on 16 February 2017 ……………………………………………………………………………………………
The effectiveness of current French health warnings displayed on alcohol advertisements and alcoholic beverages
Gloria Dossou 1, Karine Gallopel-Morvan 1, Jacques-Francois Diouf 1,2
1 EHESP School of Public Health, EA 7348 MOS, Rennes Cedex, France 2 Graduate School of Management, University of Rennes 1, UMR 6211 CREM, Rennes, France
Correspondence: Karine Gallopel-Morvan, EHESP School of Public Health, 15 Avenue du Professeur Le ´on Bernard, CS 74312 – 35043 Rennes Cedex, France, Tel: +33 (0)6 75 36 91 62, Fax: +33 (0)2 99 02 26 25, e-mail: karine.gallopel-morvan@ehesp.fr
Downloaded from https://academic.oup.com/eurpub/article-abstract/27/4/699/3002004/The-effectiveness-of-current-French-health by guest on 25 August 2017
Background: Many countries use health warnings in an attempt to regulate alcohol consumption. However, there is a lack of conclusive evidence in the research on alcohol warnings to support decision-making on effective health policies. This study explores the effectiveness of two mandatory warnings introduced in France in 1991 and 2007: the first (Alcohol abuse is harmful) is displayed on alcohol advertisements; the second (a pictogram) on bottles. Given that advertising content regulations have been implemented in some countries to reduce the attractiveness of alcohol marketing (e.g. the Evin law in France), this research also aims to explore whether such regulations can improve the effectiveness of warnings. Methods: In-depth interviews were conducted with 26 French people aged 15–29 years. The effectiveness of health warnings was assessed in terms of recall, noticeability, credibility, comprehension, responsiveness, and ability to encourage moderate drinking and abstinence during pregnancy. Participants were shown alcohol advertisements and bottles that either followed or challenged content regulations. The data were analyzed using double manual coding and NVivo software. Results: While both warnings suffered from a lack of visibility and noticeability due to their size, location, and outdatedness and because of competition from marketing design elements, the warning on the advertisement that followed content regulations was most visible. Both warnings were considered to be informationally vague, lacking in credibility and ineffective in terms of making participants feel concerned and influencing consumption habits. Conclusions: Current French warnings are ineffective and require modification. Improvements are suggested regarding the design and content of warnings to help increase their effectiveness.
Introduction
Responsible for 3.3 million deaths worldwide, alcohol consumption causes diseases (cardiovascular diseases, cancer, foetal alcohol syndrome, etc.) and social problems (road accidents, murders, etc.).1 It is the third highest risk factor2 in Europe, which has one of the highest alcohol consumption rates worldwide (11 litres of pure alcohol per person per year3). Moreover, Europe’s teenage population is particularly associated with heavy drinking patterns (binge drinking).4 The WHO has recommended various measures to regulate alcohol consumption,1 including price regulations, advertising restrictions, and warnings on products. This paper focuses on the warnings displayed on alcohol advertisements and alcoholic beverages in France. Currently, 41 countries legally require alcohol warning labels on advertisements and 31 countries stipulate that bottles or containers of alcoholic beverages must carry warnings.1 Most countries use text-only warnings (e.g. USA, South Africa, Brazil), although Thailand has introduced graphic warnings. Within Europe, 10 countries use mandatory text-only warnings (including Germany, Estonia and Sweden),2 and France has also decreed that a pictogram be displayed on bottles to deter pregnant women from consuming alcohol. Past research has explored the effectiveness of such warnings. The majority of studies have tested the impact of the text warnings introduced in the USA in 1989, highlighting the risks of alcohol consumption during pregnancy or in the workplace. More recent studies have focused on alcohol warnings in the UK and Australia.5–7 Some research has revealed that warnings are noticed and recognised by heavy drinkers regularly exposed to them8–10 and that the general public are also aware of them.11 Warnings have also been found to encourage people to talk about drink-driving and alcohol consumption during pregnancy8 and to increase the perceived risks of drinking alcohol.12,13 However, some studies have found either little impact, no change or even reverse effects of alcohol warnings on drinking intentions;6,12,14–17 young drinkers display a higher intent to drink alcohol when exposed to alcohol warnings16 and warnings may have reverse effects on heavy drinkers, leading to a reduction of negative associative thoughts.17 In addition, some studies have shown that warnings have a limited effect on the drinking behaviour of pregnant women.11 In short, the existing research on alcohol warnings has produced mixed results. While most studies have shown positive effects on cognitive responses (risk perceptions, etc.), some have questioned the ability of warnings to effectively influence drinking behaviours. This lack of conclusive evidence is prejudicial to policymakers attempting to define effective prevention programmes. This is particularly true in countries other than the USA, UK, and Australia, where almost no research on this topic has been conducted. Our first research objective is therefore to explore the effectiveness of the two current mandatory warnings in France: a text message displayed at the bottom of alcohol advertisements since 1991 (alcohol abuse is harmful) and a pictogram featured on alcohol packaging since 2007 to deter pregnant women from consumption (figure 1). We chose to focus on France for several reasons. First, as alcoholic beverages and wine production are prominent in France’s history and culture, it is an interesting context in which to test the impact of warnings. Second, alcohol is a problematic issue in this country: 49000 deaths annually due to abusive consumption.18 Moreover, figures show that in 2014, 31% of 18–75 year olds in France were episodic risk drinkers and 8% were chronic risk drinkers (14% of all 18–25 year olds).19 Third, hardly any studies have evaluated the effectiveness of the French warnings. The results of national surveys have shown a decrease in awareness of the pictogram warning on bottles: 62% of French people aged 15 and over were aware of it in 2007,20 compared with only 54% in 2015.21 Fourth, other countries considering implementing or changing alcohol warnings may be interested in the effectiveness of those used in France. Our second research objective is to explore the influence of marketing context on the effectiveness of warnings. We analyze whether different advertising content (advertisements featuring product only vs. advertisements featuring attractive and evocative elements) and alcohol bottle designs (‘traditional’ bottles vs. attractive bottles, like the limited editions regularly launched by the alcohol industry to attract consumers22) have an impact on the effectiveness of alcohol warnings. No research to date has done this. The findings might help determine how marketing content regulations aimed at reducing the attractiveness of alcohol products (which are under review in a number of countries) can also increase the effectiveness of health messages. For example, France’s Evin law23 requires advertisements, packaging and bottle designs to be strictly ‘product-oriented’ (PO): they must only present factual data and objective qualities (proof, origin, composition, etc.). Attractive advertisements and bottle designs (e.g. limited editions) with positive, evocative images and/or text associating alcohol with pleasure, success, sport, sex, etc. are banned.
Methods
We used the qualitative method recommended by Martin-Moreno et al.24 to better understand how alcohol labels are interpreted. We conducted in-depth individual interviews with 26 young people (aged 15–29 years; parental consent was obtained for minors) recruited by a market research agency to understand and explore their reactions, perceptions and comprehension of warnings (table 1: sample description). This population is an important target group for policymakers. In France, 46% of the 18–25 age group engage in binge-drinking.25 Interviews lasted about one hour and took place in 2015 in Rennes (France). The participants were first asked about the perceived benefits and risks of alcohol abuse and their motives for drinking (these results will not be presented here due to space constraints). They were then asked to recall, unaided, the content of French warnings currently displayed on bottles (pictogram) and advertisements (text warning). Following this, they were exposed to four stimuli (figure 1): two real bottles of Absolut Vodka (one PO— bottle and the 2014 Andy Warhol limited edition—LE) bearing the pictogram; and two print advertisements from the champagne Moet & Chandon (one PO advertisement and one highly ‘designoriented’—DO—advertisement) displaying the warning Alcohol abuse is harmful. Consume with moderation. Participants were asked about: (i) the warnings’ noticeability and visibility on all stimuli (they had to list the elements they noticed first on the advertisements and bottles); (ii) the warnings’ credibility; (iii) their comprehension of the warnings; (iv) their responsiveness to the warnings; and (v) the warnings’ ability to encourage moderate drinking or abstinence during pregnancy. Finally, data were collected on each participant’s sex, age, education level, occupation, and drinking profile (non-drinker, moderate drinker or heavy drinker, according to the ‘Alcohol use disorders identification test’26). Before leaving, participants received a E20 gift card for their participation and signed a consent form to allow us to use their data. The interviews were audio-recorded and transcribed using a word processor. The data were analyzed using double manual coding, performed by two researchers, as well as the NVivo software, which can uncover connections that are not possible manually.27
Results
Unaided recall of the French warnings For the majority of participants, recall and awareness were higher for the text warning on the advertisements than for the pictogram. Some respondents had been aware of the latter but had never noticed it:
I knew the bottle warning existed, but I have hardly ever seen it. (male moderate drinker, 17)
Approximately two-thirds of the participants recalled the text warning. Some knew it by heart, some were unable to remember it exactly and some knew nothing about it: We see it. We read it. I think we see it all the time, but we no longer memorise it word for word. (female non-drinker, 28)
Noticeability and visibility.
To assess warning noticeability and to identify the possible impact of marketing content on responses, participants were shown the four stimuli (PO and LE bottles, PO and DO advertisements; see figure 1) and asked to list the first five elements they noticed on each stimulus. The advertisement warning was found to be more noticeable than the pictogram. There were hardly any differences identified between the two bottles, and the pictogram was barely mentioned in the participants’ top-five lists (which included brand name, decorative elements, stamp displayed on bottles, etc.). When they were asked why the pictogram had not caught their attention, they mentioned its location (on the back of the bottle), its size (too small) and its context (surrounded by other elements hindering its visibility):
There are three labels next to one another… They are the same size but I do think there is one that trumps the other in terms of importance. (female moderate drinker, 27)
The warning on the PO advertisement was listed by about half of the participants, compared with just two in the case of the DO advertisement. The other most noticed marketing elements on the PO advertisement were the brand name, bottle stamp, text on the bottle and cork. On the DO advertisement, the most noticed elements were the bottle’s golden colour, the brand name, the gold powder, the pencil (bearing the words ‘Happy New Year’) and the black ribbon. The PO advertisement was perceived as simple with no eye-catching visual elements to distract people from the warning:
…On this [PO] ad, you get the impression that the ad is so poor and dull that the warning is the most important message, and it could even give the impression that it is a prevention message. (female heavy drinker, 25)
French warning on bottles: Since 2007, the law n°2005-102 has required the display of a mandatory pictogram on the back of bottles containing alcohol to deter pregnant women from consuming alcohol.
French warning on advertisements: This was introduced in 1991 with the Evin law, which requires that the following warning be displayed at the bottom of advertisements: “Alcohol abuse is harmful.”
“Alcohol abuse is harmful. Consume with moderation”
Figure 1 Current French warnings and the four tested stimuli
The effectiveness of current French health warnings 701
In contrast, the DO advertisement was rated more attractive and as having various artifices that delay or hinder warning-processing:
In all truth, with all these artifices on the [DO] ad, we forget about the warning. (female non-drinker, 24)
Participants blamed the warning’s lack of noticeability and visibility on outdatedness (it was introduced many years ago, so they are used to it and no longer notice it), its size (too small) and its location (at the bottom of the advertisement):
This message is displayed at the bottom of advertisements, so in the end we don’t even pay attention to it anymore. (male moderate drinker, 20)
Perceived credibility
On the subject of credibility, the participants questioned whether the warnings were an industry- or government-driven initiative. Some commented that they did not understand why alcohol advertising was still permitted if the government were the source, given the harm it can cause:
This product is, in my view, even more dangerous than certain other drugs. But the government does nothing about it for cultural and economic reasons. (female heavy drinker, 25)
The participants drew an analogy with the tobacco regulations in France, which ban all forms of advertising, and pointed out the obvious conflict of interest if the alcohol industry was the source: the industry produces very attractive advertisements and bottles while displaying very small, discreet warning messages:
They [the alcohol industry] write it as small as possible and then hope that no-one will pay attention to it. (male moderate drinker, 17)
In summary, irrespective of the source of the warnings, the participants highlighted a form of hypocrisy that potentially impairs the credibility of the warnings.
Comprehension and responsiveness The pictogram was judged easier to understand than the advertisement warning. While the pictogram was generally understood, the majority of participants lamented its lack of explicitness:
It just states that it is forbidden for pregnant women […]; how dangerous is it? (male moderate drinker, 25)
Participants were more critical of the advertisement warning, however. The majority understood its general meaning, but their comprehension of the words ‘abuse’ and ‘moderation’ was problematic:
To consume with moderation is subjective. (female nondrinker, 25) We talk about alcohol abuse, but in fact people do not really know what it means. (female moderate drinker, 17)
Opinions were divided about the warnings’ ability to make people feel concerned. The pictogram was not considered sufficient to cause concern among pregnant women (the two pregnant women in the sample confirmed this). Most participants felt, however, that the advertisement warning might make people feel concerned, depending on their drinking profile. Low to moderate drinkers were seen as more likely to respond to the message than heavy drinkers and young people:
People who drink excessively, I don’t think that the ad warning will concern them. (female non-drinker, 15)
Impact on behavioural intentions The majority of participants said that the alcohol warnings were not effective in terms of changing behaviour:
It does not stop people from deciding to drink excessively when they go out, it is not dissuasive. (female moderate drinker, 29)
Various explanations were given for this, including desensitisation: where people have been overexposed to these warnings, they no longer notice them:
Table 1 Sample profile
Participant code Gender Age Drinking profile Occupation
1. AD Male 25 Heavy drinkera Student in engineering 2. AS Female 15 Non-drinkerb Junior high student 3. AG Female 24 Moderate drinkerc Care worker (unemployed) 4. BB Male 23 Moderate drinker Student in communications 5. BLP Male 20 Heavy drinker Student in geography 6. CK Female 27 Moderate drinker IT administrative assistant 7. CP Female 25 Heavy drinker Unemployed (temporary work) 8. EM Female 29 Moderate drinker Accountant 9. ES Female 16 Moderate drinker High school student 10. FS Male 25 Moderate drinker Accountant 11. GE Female 21 Heavy drinker Student in design and applied arts 12. GL Male 17 Heavy drinker High school student 13. HF Female 27 Non-drinker Unemployed (temporary work) 14. JSI Female 17 Moderate drinker High school student 15. JST Female (pregnant) 25 Non-drinker IT middle manager 16. LCLC Male 22 Heavy drinker Student in law 17. MAB Male 17 Moderate drinker High school student 18. MG Female 24 Non-drinker Collection administrator 19. MP Male 20 Moderate drinker Student in communications 20. MA Male 21 Heavy drinker Student in law 21. MEB Male 23 Heavy drinker Worker 22. NC Female 20 Heavy drinker Student in geography 23. OG Male 17 Moderate drinker High school student 24. SF Male 19 Moderate drinker Student in psychology 25. TC Male 19 Non-drinker Unemployed 26. VL Female (pregnant) 28 Non-drinker Administrative assistant
a: Those who have three or more drinks on a day when they are drinking or who have six or more drinks on a single occasion monthly. b: Those who rarely consume alcohol (one or two drinks once or twice a year). c: Those who drink periodically but never exceed a threshold of six drinks on a single occasion.
I think the warning message has lost its original impact. (female non-drinker, 25)
The presentation and content of the warnings were also criticised. Compared with the current visual, fear-inducing tobacco warnings used in France, participants judged the content of alcohol warnings too soft. Again, they highlighted their lack of noticeability due to size, location and marketing environment:
There is nothing, we just see a little thing. Is it supposed to be dissuasive? (male heavy drinker, 21)
They also viewed the warnings as insufficient in terms of tackling abusive alcohol consumption:
I think it is a bit naive to believe that one sentence will be able to change people’s attitudes. (male heavy drinker, 22)
Another criticism was that participants felt generally overexposed to health messages and warnings in other fields (tobacco, food, road accidents, etc.), and some felt irritated by them.
Discussion
Cultural context is an important factor to consider, but almost no research on alcohol warnings has been conducted outside of the USA, UK, and Australia. In addition, the existing studies have mainly used quantitative methods whereas qualitative methods can provide a better understanding of how warnings are perceived by individuals. Furthermore, no study has yet taken into account the marketing context in which warnings are displayed. Our research attempts to fill these gaps by exploring the effectiveness of the two French warnings displayed on advertisements and bottles. As is the case with the Australian warnings,28 those used in France have been found not to be effective enough because they are poorly implemented. Based on our results, a number of recommendations can be made to policymakers to increase the effect of warnings on perceptions and behaviours. First, although the advertisement warning appears to be more noticeable than the pictogram, they both suffer from a lack of visibility due to their small size, location and outdatedness (people are so accustomed to them that they no longer notice them). The context of the warnings also decreases their visibility: marketing elements catch people’s attention at the expense of the health message (the warning is less visible on the DO than on the PO advertisement). It is, therefore, recommended that warnings be changed periodically and rotated to prevent desensitisation and that they be increased in size. French warnings also suffer from a lack of noticeability due to their location. Future research should explore the effects of warnings on the front of bottles and at the top of advertisements. The regulation of alcohol advertising content, as proposed in France (Evin law), also appears to be potentially effective in terms of increasing the noticeability of warnings. The use of eye-catching colours, typography, and signs would be appropriate ways of increasing the visibility of warnings.28,29 Future research should therefore focus on identifying the most effective designs to attract people’s attention to warning labels within the competing alcohol marketing context (using eye-tracking methods, for instance). Second, we found that the young people in the sample did not clearly understand the text warning on the advertisements. They also thought the warnings on the advertisements and the bottles were too soft, and they did not feel concerned. It should be noted that the advice to ‘Consume with moderation’ is not mandatory but has been added by alcohol companies. This message—judged vague and unclear in our study—is widely used in their responsible drinking campaigns. Advocating ‘moderation’ in this way may misinform consumers because there is no threshold level for safe alcohol consumption, and even low alcohol intake increases the risk of some cancers.30 The content of French warnings should therefore be revised, especially in advertising. Further research should be conducted to investigate the messages conveyed by warnings. Past studies have revealed that content can increase effectiveness: warnings that are framed positively (Reduce your drinking to reduce your risk of cancer)5 or which give details about specific cancer risks (Alcohol increases your risk of breast cancer)5,29 are perceived to be more believable and make drinkers feel more concerned. Warnings formulated as questions also seem to be effective as they increase individual negative outcome expectancy.31 Third, our study reveals that the warnings do not motivate our sample to change their alcohol consumption behaviour. There are three main reasons for this: the warnings are weak (too small, too soft, etc.); they are not powerful enough to make people modify their behaviour; and the participants feel bombarded by other competing health messages. To improve the effectiveness of alcohol warnings, some researchers have suggested using pictorial messages, such as those found on tobacco packs (only Thailand has so far implemented pictorial warnings on alcohol products). Collymore and McDermott32 concluded that warnings designed to disgust were the most effective at reducing intended alcohol intake. Al-hamdani and Smith33 showed that text-and-image labels about liver cancer were effective in reducing product attractiveness. However, Zahra et al.34 proposed that negative imagery should be used with caution in alcohol warnings. Indeed, there is a lively debate surrounding the use of negative emotions in warnings, 35 and more research should be conducted to explore whether pictorial alcohol messages can increase the impact on behavioural changes. In spite of the importance of our research for public health actors, the findings should be considered in light of a number of limitations. As with all self-report methods, some of the participants’ answers may be biased due to their underestimation of the risks. Furthermore, only behavioural intentions (not real) were evaluated. Extrapolating the results also proved difficult within this qualitative approach because the sample was not representative. Despite these limitations, our paper provides useful recommendations for policymakers and points to further research avenues.
Funding
This work was supported by Grant no 2013-154 (I(D)MATREC) from the French National Cancer Institute, the Inter-Ministerial Mission for Combating Drugs and Addictive Behaviours (MILDECA) and the University of Paris XIII.
Conflicts of interest: None declared.
Key points This research is the first to assess the effectiveness of France’s two mandatory alcohol warnings. Results show the current format, content, and marketing context of French warnings influence their effectiveness in terms of noticeability, comprehension, responsiveness, and behavioural impact. Some recommendations are made to policymakers to increase the effectiveness of alcohol warnings.
References
1 World Health Organization, Global Status Report on Alcohol and Health 2014. http://www.who.int/substance_abuse/publications/global alcoholreport/en/ (3 June 2016, date last accessed).
The effectiveness of current French health warnings 703
Downloaded from https://academic.oup.com/eurpub/article-abstract/27/4/699/3002004/The-effectiveness-of-current-French-health by guest on 25 August 2017
2 EUROCARE, The European Alcohol Policy Alliance, What’s on the Bottle? Eurocare Reflections on Alcohol Labelling 2014. http://www.eurocare.org/resources/policy_ issues/labelling. Accessed (3 June 2016, date last accessed).
3 World Health Organization, The European Health Report 2015, Targets and Beyond – Reaching New Frontiers in Evidence. http://www.euro.who.int/en/data-andevidence/european-health-report/european-health-report-2015/ehr2015 (3 June 2016, date last accessed).
4 Braker A, Go ¨bel K, Scheithauer H, Soellner R. Adolescent alcohol use patterns from 25 European countries. J Drug Issues 2015;45:336–50.
5 Pettigrew S, Jongenelis M, Chikritzhs T, et al. Developing cancer warning statements for alcoholic beverages. BMC Public Health 2014;14:786.
6 Scholes-Balog KE, Heerde JA, Hemphill SA. Alcohol warning labels: unlikely to affect alcohol-related beliefs and behaviours in adolescents. Aust N Z J Public Health 2012;36:5.
7 Petticrew M, Douglas N, Knai C, et al. Health information on alcoholic beverage containers: has the alcohol industry’s pledge in England to improve labelling been met? Addiction 2015;doi:10.1111/add.13094.
8 Greenfield TK, Graves KL, Kaskutas LA. Long-term effects of alcohol warning labels: Findings from a comparison of the United States and Ontario, Canada. Psychol Market 1999;16:261–82.
9 Graves K. Evaluation of the alcohol warning label: a comparison of the United States and Ontario, Canada in 1990 and 1991. J Public Policy Mark 1993;12:19–29.
10 Kaskutas LA, Greenfield TK. First effects of warning labels on alcoholic beverage containers. Drug Alcohol Depend 1992;31:1–14.
11 Thomas G, Gonneau G, Poole N, Cook J. The effectiveness of alcohol warning labels in the prevention of fetal alcohol spectrum disorder: a brief review. Int J Alcohol Drug Res 2014;3:91–103.
12 MacKinnon DP, Lapin A. Effects of alcohol warnings and advertisements: a test of the boomerang hypothesis. Psychol Market 1998;15:707–26.
13 Slater MD, Domenech MM. Alcohol warnings in TV beer advertisements. J Stud Alcohol 1995;56:361–7.
14 Stockley C. The effectiveness of strategies such as health warning labels to reduce alcohol-related harms—an Australian perspective. Int J Drug Policy 2001;12:153–66.
15 Wilkinson C, Room R. Informational and warning labels on alcohol containers, sales, places and advertisements: experience internationally and evidence on effects. Drug Alcohol Rev 2009;28:341–461.
16 Snyder LB, Blood DJ. Caution: alcohol advertising and the Surgeon General’s alcohol warnings may have adverse effects on young adults. J Appl Commun Res 1992;20:37–53.
17 Brown KG, Stautz K, Hollands GJ, et al. The cognitive and behavioural impact of alcohol promoting and alcohol warning advertisements: an experimental study. Alcohol Alcohol 2016;51:354–62.
18 Guerin S, Laplanche A, Dunant A, Hill C. Alcohol-attributable mortality in France. Eur J Public Health 2013;23:588–93.
19 Observatoire Franc¸ais des Drogues et des Toxicomanies, Synthese thematique: alcool 2015, France. http://www.ofdt.fr/produits-et-addictions/de-z/alcool (3 June 2016, date last accessed).
20 Guillemont J, Le ´on C. Alcool et grossesse: connaissances du grand public en 2007 et e ´volutions en trois ans (Alcohol during pregnancy: which knowledge from 2007 to 2010 in the French population?). Evolutions 2008;15.
21 Cogordan C, Nguyen-Thanh V, Richard J-B. Alcool et grossesse. Connaissances et perception des risques. Alcool Addictol 2016;38:181–90.
22 Examples:http://www.brandingmagazine.com/2011/11/22/moet-chandons-goldenjeroboam-personalized-limited-edition-bottles/ or http://popsop.com/2014/10/absolutlaunches-the-andy-warhol-limited-edition-bottle/ (3 June 2016, date last accessed).
23 Loi Evin. Republique Francaise. Loi n 91-32 du 10 janvier 1991 relative a ` la lutte contre le tabagisme et l’alcoolisme. JORF 1991;615–8.
24 Martin-Moreno JM, Harris ME, Breda J, et al. Enhanced labelling on alcoholic drinks: reviewing the evidence to guide alcohol policy. Eur J Public Health 2013;23:1082–7.
25 Institut National de Prevention et d’Education pour la Sante ´. Barometresante ´ INPES 2014, France. http://www.inpes.sante.fr/70000/cp/15/cp150331-barometresante-2014-alcool.asp (3 June 2016, date last accessed). 26 Babor TF, Higgins-Biddle JC, Saunders JB. Monteiro MG on behalf of World Health Organization. The Alcohol Use Disorders Identification Test, 2nd edn. World Health Organization 2001, http://apps.who.int/iris/handle/10665/67205 (3 June 2016, date last accessed).
27 Miles MB, Huberman M, Saldan ˜a J.Qualitative Data Analysis. A Methods Sourcebook, 3rd edn. Sage Publications: Los Angeles, London, New Delhi, Singapore, Washington DC, ISBN 9781452257877, 2014.
28 Coomber K, Martino F, Barbour IR, et al. Doconsumers ‘Get the facts’? A survey of alcoholwarning label recognition inAustralia. BMC Public Health 2015;15:1.
29 Miller ER, Ramsey IJ, Baratiny GY, Olver IN. Message on a bottle: are alcohol warning labels about cancer appropriate? BMC Public Health 2016;16:1.
30 Bagnardi V, Rota M, Botteri E, et al. Light alcohol drinking and cancer: a meta-analysis. Ann Oncol 2013;24:301–8.
31 Krischler M, Glock S. Alcohol warning labels formulated as questions change alcohol-related outcome expectancies: a pilot study. Addict Res Theory 2015;23: 343–9.
32 Collymore NN, McDermott MR. Evaluating the effects of six alcohol-related message frames on emotions and intentions: the neglected role of disgust. J Health Psychol 2016; 21:1907–17.
33 Al-hamdani M, Smith S. Alcohol warning label perceptions: emerging evidence for alcohol policy. Can J Public Health 2015;106:395–400.
34 Zahra D, Monk RL, Corder E. ‘If you drink alcohol, then you will get cancer’: investigating how reasoning accuracy is affected by pictorially presented graphic alcohol warnings. Alcohol Alcohol 2015;50:608–16.
35 Monarrez-Espino J, Liu B, Greiner F, et al. Systematic review of the effect of pictorial warnings on cigarette packages in smoking behavior. Am J Public Health 2014;104:11–30.
704 European Journal of Public Health
Downloaded from https://academic.oup.com/eurpub/article-abstract/27/4/699/3002004/The-effectiveness-of-current-French-health