You are here

Article Text

Article menu

Download PDFPDF

Research paper
Impact of a negative emotional antitobacco mass media campaign on French smokers: a longitudinal study
  1. Romain Guignard1,
  2. Karine Gallopel-Morvan2,
  3. Ute Mons3,4,
  4. Karin Hummel5,
  5. Viêt Nguyen-Thanh1
  1. 1 Department of Prevention and Health Promotion, Santé Publique France, The National Public Health Agency, Saint-Maurice, France
  2. 2 EA 7348 MOS, EHESP School of Public Health, Rennes, France
  3. 3 Cancer Prevention Unit, German Cancer Research Center (DKFZ), Heidelberg, Germany
  4. 4 WHO Collaborating Centre for Tobacco Control, German Cancer Research Center (DKFZ), Heidelberg, Germany
  5. 5 Department of Health Promotion, Maastricht University (CAPHRI), Maastricht, The Netherlands
  1. Correspondence to Romain Guignard, Department of Prevention and Health Promotion, Santé Publique France, The National Public Health Agency, Saint-Maurice 94 415, France; Romain.guignard{at}santepubliquefrance.fr

Abstract

Background Mass media campaigns to encourage smoking cessation have been shown to be effective in a context of comprehensive tobacco control programme. The effectiveness of antismoking ads that evoke negative emotions remains unclear, in particular in countries with high smoking prevalence and among smokers with low perceived susceptibility, low self-efficacy or who are not users of smoking cessation services.

Objective To evaluate short-term and long-term effects of a 1-month French national highly emotional media campaign, with a focus on these specific targets.

Design A 6-month longitudinal survey by Internet. A sample of 3000 smokers were interviewed before the media campaign (T0). They were contacted again just after (T1) and 6 months after the campaign (T2).

Outcomes Perceived susceptibility to the risks of smoking, self-efficacy to quit smoking, use of smoking cessation services (quitline and website) and 7-day quitting.

Methods The analysis was carried out on 2241 individuals who answered at T1 and T2. Multiple logistic regressions were computed to test the association between the change in each outcome at T1 and T2 and the level of exposure based on self-reported recall.

Results Self-reported recall was associated with an increase in perceived susceptibility and with use of cessation services. Campaign recall was also associated with higher 7-day quitting immediately after the campaign (OR=1.8 (1.0 to 3.2), P<0.05).

Conclusions Fear-appeal mass media campaigns can be effective in encouraging cessation among smokers in a country with high smoking prevalence (France), but should be accompanied by convincing self-efficacy messages.

  • media
  • social marketing
  • cessation
  • socioeconomic status
  • public policy

https://doi.org/10.1136/tobaccocontrol-2017-053936

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Tobacco use kills >7 million people every year worldwide.1 The WHO Framework Convention on Tobacco Control (FCTC) is a response to this pandemic, providing various support tools to develop effective tobacco control policies.2 One of these tools are mass media campaigns as proposed in Article 12 which calls for the use of all available communication tools to promote and strengthen public awareness of tobacco control issues. Generally, mass media campaigns to encourage smoking cessation have been shown to be effective in a context of comprehensive tobacco control programme.3–7 However, the effectiveness of antismoking ads that evoke negative emotions by featuring shocking imagery or personal testimonials remains unclear.8–11

    The idea of such fear appeals is to associate a practice (smoking) with its negative consequences (death, disease and so on) to generate a negative emotion (fear, disgust, guilt, anger and so on), to eventually increase risk perception and to motivate people to quit.

    For instance, the national US campaign launched by the Centers for Disease Control and Prevention  called ‘Tips From Former Smokers’ that depicted smoking-related suffering in real people was effective in increasing population-level quit attempts,12 and the personal testimonials campaign (‘TRU’) delivered in North Carolina about negative consequences of tobacco on the smokers’ family decreased smoking prevalence among youths.13 Regarding the use of shocking images, the original versions of the ‘Sponge’ Australian advertisement (1983) and of the Australian ‘Artery’ campaign (1997) were associated respectively with a 3.4%14 and a 1.4% decline in smoking prevalence.15 The remake of these campaigns in 2007 and 2008 also achieved high levels of recognition and generated thoughts of quitting.16 Studies that compared negative emotional antitobacco campaigns with other ad contents have highlighted that emotional ads are more effective,17 18 especially with regards to quitting behaviour.19 20

    Even if favourable effects have been demonstrated, negative emotional ads remain unclear for several reasons.

    First, some studies have found little or no effect of such ads on people’s reactions8 21–23: Richardson et al for example, highlighted that at the same level of media exposure, positive campaigns in England were more effective at increasing quitline calls compared with negative emotional ones.24 This might be explained by cognitive dissonance in smokers exposed to fear appeals—they may resist the message by minimising the threat, denying the risk, rejecting the problem, or avoiding the message,25 particularly when they feel unable to stop smoking (low self-efficacy)8 21 or when they do not feel concerned by the risk (low perceived susceptibility).26 The effects of emotional negative ads on different targets have been explored in the literature, but results are unclear and to our knowledge, smokers with low self-efficacy and low perceived susceptibility have never been tested. Furthermore, two reviews showed that the relationship between campaign effectiveness and gender, age or education was not systematic27 and that the impact of mass media campaigns on equity was inconsistent.28 Another review by Durkin et al concluded that negative health effects messages using testimonials or graphic depictions may be effective across different socioeconomic groups and sometimes more in low socioeconomic groups, compared with how-to-quit messages, while the effectiveness of different message types across age and gender revealed no consistent patterns.4 More largely, a recent meta-analysis on a large scope of public health campaigns showed that fear appeal messages were more effective in women than in men.29

    Second, there are methodological concerns, as most studies analysed short-term effects rather than long-term effects, used cross-sectional rather than longitudinal design or were conducted in an experimental laboratory setting with low generalisability to a real-world context.22

    Third, most research comes from countries with Anglo-Saxon cultures or with generally low smoking prevalence rate. However, research on the effectiveness of antitobacco media campaigns showed differences among countries that may be explained by cultural differences.30 Study results on shocking warning labels on tobacco packs revealed that they had less impact on behavioural intentions among French adolescents than among US ones.31 In addition, France is also a specific country as tobacco prevalence is very high (34.5% of the French population were smokers in 2016,32 compared with 17%, for example, in the UK33) and the context is not favourable to the use of fear appeals since a part of French health promotion practitioners are opposed to them.34 35 Thus, there is a need to explore the effects of negative emotional antitobacco media campaigns in countries with different cultures, higher smoking rates and strong reactions against these tactics, as reactance might be higher.

    The present study examines the impact of a French national campaign featuring fictional testimonials and highly emotional content (guilt, fear), in a 6-month longitudinal study aiming at identifying the relationship between smokers’ exposure to the campaign and its impact on perceptions, the use of smoking cessation services and quitting behaviour. This study particularly focuses on the reactions of daily smokers with low perceived susceptibility, low self-efficacy or who are not users of smoking cessation services, and also investigates socioeconomic differences.

    Methods

    Description of the ‘farewell message’ French antitobacco media campaign

    A new 1-month French national campaign was launched on 26 September 2014 by the National Institute for Health Promotion and Health Education (INPES) within the framework of the National Smoking Reduction Program.36 The campaign was called ‘farewell message’ and consisted of TV ads (two 30 s videos) and radio messages (two 30 s messages) that were shown on the largest television channels, on the Internet and on radio stations in France (versions of the ads are available at http://inpes.santepubliquefrance.fr/30000/actus2014/035-tabac-risques.asp). A sad goodbye message of a woman to her two daughters, and of a man to his wife, respectively, was heard in a voice-over. This voice-over was accompanied by a video showing hospital employees cleaning the room where the smoker had just died. At the end of the ads, two written messages were inserted consecutively: a negative one (‘one in every two smokers will die from tobacco’) and a positive one (‘To quit smoking, call the 39 89–0.15 cents of Euros per minute- or go to tabac-info-service.fr’). The positive message was inserted because the fear-appeal literature has underlined that negative emotional campaigns must be combined with instructions and efficacy statements that rise the target’s ability to perform the recommended response. The radio messages were the same as the TV ones, including the same negative and self-efficacy statements at the end. Among 25–49 years old (the core target group), the ‘gross rating points’ (GRPs) of this campaign were 968 for TV and 516 for radio. The GRPs measure the total amount of the advertising exposures produced by a media schedule during a specific period of time.37 A post-test of the ads conducted among 1014 French people revealed that a majority felt guilt and fear when exposed to the ads, without any difference between smokers and non-smokers (67% agreed to say that ads made them feel guilty; 52% said that ads provoked fear38).

    Study design and sample

    A 6-month longitudinal survey was implemented from September 2014 to March 2015 to assess the effectiveness of the campaign. A sample of 3000 daily smokers aged 15–85 years was recruited in an access panel by a French market research company (BVA) to be interviewed via Internet. The screening questions were: ‘do you smoke, even occasionally?’ and ‘how many cigarettes do you smoke on average, including roll-your-own tobacco?’39 40 Only smokers who reported that they smoked on average at least one cigarette per day were included. The sample was selected according to quotas on gender, age, occupation, size of urban unit and region, reflecting the 2010 smokers’ population structure observed in a French national representative survey.40 41 Respondents were first interviewed before the media campaign (T0). All of them were contacted again just after the end of the campaign (T1) and 6 months after the campaign (T2) (figure 1).

    Figure 1
    Figure 1

    Participants and schedule of the longitudinal survey.

    Of the 3000 smokers interviewed at T0, 2241 (74.7%) could be followed-up at T1 and were considered in the present analyses, of those 1846 also answered at T2 (61.5%). The proportion of participants who were lost to follow-up was higher in women (T1: 28.2% vs 22.8% in men, P=0.001; T2: 41.5% vs 35.9%, respectively, P=0.002). The attrition rate decreased with age (T1: 37.5% in 15–24 years, 25.1% in 25–49 years, and 18.5% in 50–85 years, P<0.001; T2: 54.8%, 38.8% and 28.1%, respectively, P<0.001). Attrition was not different according to the level of education, nor according to level of addiction (Heaviness of Smoking Index: HSI),42 quit attempt in the last 30 days, and intention to quit.

    The design of this research has been approved by the French commission on data privacy and public liberties (CNIL).

    Measures

    The aim of the longitudinal survey was to assess the association between exposure to the campaign and different smokers’ reactions.

    Exposure

    Exposure to the campaign was measured with self-reported campaign recall. It was recorded at T1 by showing consecutively both TV spots and by asking for each one: ‘Here is a spot broadcast on television, Internet and radio. Please indicate whether you remember having seen or heard it’. Respondents reporting having seen or heard one or both spots were regarded as having recalled the campaign.

    Perceived susceptibility, self-efficacy and use of smoking services

    Different smokers’ reactions were measured: the perceived susceptibility to the harms of smoking, the perceived ability to quit (self-efficacy), the use of smoking cessation services and quitting behaviour.

    Perceived susceptibility was assessed by asking ‘How worried are you that smoking will damage your health in the future?’ with different response options (0: not at all/rather not=low perceived susceptibility; 1: rather/very =  high perceived susceptibility).43

    Self-efficacy was assessed with the question: ‘If you decided to quit smoking, how confident are you that you would succeed?’ (0: not at all/rather not=low self-efficacy; 1: rather/very =  high self-efficacy).44

    To measure the use of smoking cessation services, people were asked if they had called the tobacco quitline Tabac Info Service (TIS) (39 89) or if they had visited the website ‘Tabac-info-service.fr’ in the last 30 days.

    All these outcomes were collected at each survey wave (T0, T1 and T2). Perceived susceptibility and self-efficacy were skipped for respondents who reported they did not smoke at T1 or T2.

    To measure quitting behaviour, respondents who reported having quit smoking for at least 7 days at T1 and T2 were considered as quitters.

    Media utilisation

    For each respondent, the frequency of TV viewing (‘every day’, ‘almost every day’, ‘one or two times a week’, ‘less often’ and ‘never’) was collected for each time slot (before 9 am, 9–12 am, 12 am–1.30 pm, 1.30–3 pm, 3–5 pm, 5–7 pm, 7–8.45 pm, 8.45–10.30 pm and after 10.30 pm). A global frequency of TV viewing was assessed by considering the highest frequency recorded. It was recoded in three categories: ‘every day’, ‘almost every day’ and ‘less often’.

    Level of nicotine addiction

    The level of nicotine addiction was assessed at T0 with the HSI, recoded into three categories, from the lowest level of nicotine addiction to the highest: 0–1, 2–3 and 4–6.42

    Sociodemographic variables

    Sociodemographic variables were included in the questionnaire at T0: gender, age, education, income and employment status. The level of education was recoded into three categories according to the highest diploma obtained by the respondent (less than high school, high school completed and college graduate). The household income was collected, calculated per consumption unit (CU) and recoded according to the terciles of its distribution. CUs are used to compare households of different sizes and compositions by assigning a coefficient to each member of the household: 1 CU for the first adult in the household, 0.5 CU for other people aged ≥14 years and 0.3 CU for children under 14 years (Organisation for Economic Co-operation and Development scale, French National Institute for Statistics and Economic Studies).45 Categories for employment status were: employed, unemployed (looking for a job), retired, student or inactive.

    Statistical analyses

    Poststratification weights were computed at each wave to reflect the smokers’ population structure in terms of gender, age, region, size of urban unit and occupation.

    The first set of analyses were carried out on subsamples of respondents at T1 with low perceived susceptibility, low self-efficacy or who did not use the quitline or cessation website at baseline (n=525, n=1024, n=2191 and n=2143, respectively).

    For each subsample, bivariate analyses were performed between the change in outcome at T1 and T2 (from low to high perceived susceptibility, from low to high self-efficacy and so on) and campaign recall. Pearson’s χ2 tests were used for comparisons between respondents who recalled the campaign and others.

    The associations between campaign recall and the change in each of the outcomes at T1 and T2 were assessed using multiple logistic regression models adjusted for gender, age, employment status, education, income, HSI at baseline and the frequency of TV viewing. In this way, the models could be used to assess transitions observed in T1 and T2 compared with T0, according to the level of exposure.

    Second, in order to identify the associations between the level of exposure and quitting, multiple logistic regression models were computed on the whole sample, with adjustment for the same confounders as well as quit attempts in the last 30 days reported at baseline.

    Finally, interactions with gender, age, education and income were added in separate logistic regressions. ORs for stratified analyses are shown in the text when the interactions were significant.

    Results

    Sample description

    At T0, a majority of respondents had high perceived susceptibility (75.7%) and about half of respondents had high self-efficacy (54.2%). Very few respondents reported using TIS quitline (2.4%) or website (4.4%) in the last 30 days. At T1, 48.3% of respondents reported that they watched TV ‘every day’, 37.8% ‘almost every day’ and 13.9% ‘less often’; 80.6% reported having recalled the media campaign. The sociodemographic characteristics and HSI at baseline for each subsample are shown in table 1.

    View this table:
    Table 1

    Description of the subsamples at baseline: sociodemographic characteristics and heaviness of smoking index

    Results on specific subsamples

    Change in perceived susceptibility among smokers whose perceived susceptibility was low at baseline

    Among respondents at T1 whose perceived susceptibility to health risks of smoking was low at baseline (n=525), the ones who recalled the campaign were more likely to report an increased perceived susceptibility just after the campaign (43.2% vs 27.0% among the ones who did not recall the campaign, P=0.003), but not 6 months later. After adjustment for potential confounders, campaign recall was associated with increased perceived susceptibility at T1 (adjusted OR (aOR)=1.8; 95% CI 1.1 to 3.0), but not at T2 (table 2).

    View this table:
    Table 2

    Change from low to high perceived susceptibility at T1 and T2 among smokers with low perceived susceptibility at T0, according to campaign recall: weighted percentages, aORs and CIs (n=493)

    No significant interaction was found with gender, age, education or income.

    Change in self-efficacy among smokers whose self-efficacy was low at baseline

    Among respondents with low self-efficacy at baseline (n=1024), there was no association between campaign recall and self-efficacy at T1, but those who recalled the campaign were less likely to report high self-efficacy at T2 (26.1% vs 37.3% among the ones who did not recall the campaign, P=0.009). After adjustment for sociodemographic characteristics, HSI and frequency of TV viewing, campaign recall was no more significantly associated with change in self-efficacy 6 months after the campaign (aOR=0.7; 95% CI 0.5 to 1.1) (table 3).

    View this table:
    Table 3

    Change from low to high self-efficacy at T1 and T2 among smokers with low self-efficacy at T0, according to campaign recall: weighted percentages, aORs and CIs (n=981)

    No interaction was found with gender, age, education or income.

    Change in the use of cessation help among smokers who were not users at baseline

    About 6 months after the campaign (T2), the proportion of respondents who used TIS in the last 30 days was higher among the ones who recalled the campaign: 3.0% of them used TIS quitline and 4.7% used TIS website, compared with 0.8% (P=0.011) and 2.1% (P=0.023) of the ones who did not recall the campaign, respectively. These associations were significant when adjusting for confounders (aOR=3.1; 95% CI 1.2 to 8.3; aOR=2.2; 95% CI 1.1 to 4.6, respectively). There was no significant difference according to campaign recall at T1 (table 4).

    View this table:
    Table 4

    Use of Tabac Info Service quitline and website at T1 and T2 among smokers who did not use it at T0, according to campaign recall: weighted percentages, aORs and CIs (n=2191 for quitline; n=2143 for website)

    There was no interaction with gender, age, education or income concerning the use of these cessation services.

    Quitting behaviour

    The proportion of quitters observed just after the campaign was higher among respondents who recalled the campaign (7.3% vs 4.4%, P=0.034). At T2, there was no significant difference on quitting behaviour. In multivariate analyses, respondents who recalled the campaign were more likely to have quit smoking for at least 7 days just after the campaign (aOR=1.8; 95% CI 1.0 to 3.2), but not 6 months later (table 5).

    View this table:
    Table 5

    7 day quitting at T1 and T2 according to campaign recall: weighted percentages, aORs and CIs (n=2241)

    There was a significant interaction between quitting at T2 and age group (P=0.044). Respondents aged ≥50 years who recalled the campaign were more likely to have quit 6 months after the campaign (aOR=4.0; 95% CI 1.2 to 13.1; P=0.024) while no significant association was found in other age groups.

    Discussion

    Studies on negative emotional antitobacco media campaigns have mostly been conducted in Anglo-Saxon countries. However, researchers have underlined that prevention messages should be analysed in their specific cultural and environmental contexts related to consumption habits, social influence and so on.46 Indeed, cultural patterns are particularly important as they influence people’s comprehension and perceptions of risks displayed in prevention messages.47

    Another limitation of previous research is that the effectiveness of emotional negative ads on daily smokers with low perceived susceptibility, low self-efficacy or who are not users of smoking cessation services has rarely been studied.

    The present study aimed at evaluating the short-term and long-term impact of a negative emotional antitobacco mass media campaign using a longitudinal sample of French smokers. The recall of the campaign was associated with quitting just after the campaign in the whole sample and with quitting 6 months later among smokers aged ≥50 years. Results showed that among the ones with low perceived susceptibility at baseline, exposure to the campaign was associated with increased perceived susceptibility just after the campaign. The use of website and phone-based cessation services increased 6 months after the campaign among smokers who recalled the campaign. These results support conclusions of previous research that showed that negative health effects messages that use graphic images, personal testimonials or emotional contents are effective in changing smokers’ attitudes or beliefs and in increasing quitting behaviour.4 17 18 Our research underlined that these positive effects can also occur on daily smokers with low perceived susceptibility and low use of cessation services.

    Conversely, among smokers whose self-efficacy was low at baseline, the ones who recalled the campaign were less likely to have an increased self-efficacy at 6 months. This can be explained by the fact that such a target group needs even more information on how to quit smoking. This result is in line with the literature on fear appeal messages. A meta-analysis supported the hypothesis that fear appeals which lack efficacy statements produce weaker effects than the ones that include efficacy statements.29 Thus, with regards to antismoking campaigns, the effects of such messages are reinforced if smokers perceive quitting as an effective way to avoid the threat (« efficacy ») and if they believe they are able to quit smoking (« self-efficacy »). It must be noted that, in the present study, the subsample of smokers with low self-efficacy had a higher level of dependency than the rest of the population. Thus, referral to cessation help programme at the end of the ads might not have been strong enough to persuade low self-efficacy and highly addicted smokers of their ability to quit. After adjustment for confounding factors, the association between campaign recall and change in self-efficacy was no longer significant.

    The absence of an association in the whole sample between exposure to the campaign and quitting at 6 months is consistent with previous research that showed that mass media campaigns often only have short-term effects (at 2 or 3 months) on prevalence or quit attempts, which plead for repeated cycles of broadcasting of high-intensity campaigns (at least 1000 GRP per quarter).5 37 However, the present campaign could have had an effect on long-term quitting among the oldest smokers, perhaps because of its content that emphasised the end of life. A review published in 2013 showed that the relationship between the effectiveness of mass media campaigns and age was not consistent.27 The analysis of interactions did not show any other difference according to gender, education or income. Finally, mass media campaigns need to be part of comprehensive tobacco control programme to have a sustainable effect on smoking behaviour.27

    The present study has several strengths: it relies on a longitudinal sample of smokers, which enabled us to assess the individual changes in perceptions or behaviours according to the level of exposure. Survey waves were specifically designed to capture short-term and long-term effects.

    This study also has some limitations. For budgetary reasons, it was not possible to draw on a probabilistic sample and the sample size was too small to focus on specific populations such as smokers of low socioeconomic status or smokers aged 25–49 years who were the primary campaign target group. Data are self-reported and the risk of social desirability bias cannot be excluded. We also cannot rule out reverse causation arising from a higher susceptibility to anti-smoking messages among those who contemplate quitting. Post-stratification weighting reduces to some extent, but does not eliminate, non-response biases.48 Despite multiple adjustments in the measure of associations with campaign recall, remaining confounding biases cannot be excluded. Finally, the respondents were highly educated compared with the general population of smokers: 72% were high school graduate or higher versus 38% in a representative sample of smokers (French Health Barometer 2014).49 50

    Further research is needed concerning the impact of negative emotional media campaigns on smokers with low perceived susceptibility to the harms of smoking or low self-efficacy to quit, using random surveys with larger sample sizes. Studies in non-Anglo-Saxon countries should be encouraged as cultural and social context may influence the effectiveness of such campaigns.

    Conclusion

    Exposure to an antismoking negative emotional media campaign was associated with quitting behaviour, with the use of smoking cessation services and with increased perceived susceptibility among smokers distant from the harms of smoking, without any significant difference according to gender, education or income. This supports the use of fear-appeal messages to promote smoking cessation, including in countries with high levels of smoking prevalence or among smokers with low perceived susceptibility. However, negative effects on self-efficacy have also been suggested. To prevent these adverse reactions to such messages, previous research recommends juxtaposing negative messages with strong reassuring messages.

    What this paper adds

    • Mass media campaigns are effective to encourage smoking cessation in a context of comprehensive tobacco control programme.

    • The effectiveness of negative emotional ads, such as fear-appeal ads, remains unclear, in particular on reluctant targets.

    • Fear-appeal mass media campaigns can be effective in encouraging cessation among smokers with low perceived susceptibility and in countries with high smoking prevalence, but self-efficacy messages should be strong enough to persuade highly dependent smokers who do not feel able to quit.

    Acknowledgments

    The authors thank Pierre Arwidson (Santé publique France) for his help in the conception stage of the study, Jean-Baptiste Richard and Raphaël Andler (Santé publique France) for their methodological advises and Bérengère Gall (BVA Institute) for the coordination of survey field and exploratory analyses.

    References

    1. ↵
      World Health Organization. WHO report on the global tobacco epidemic, 2017: monitoring tobacco use and prevention policies. Geneva, Switzerland: WHO, 2017.
    2. ↵
      World Health Organization. WHO framework convention on tobacco control. Geneva, Switzerland: WHO, 2003.
    3. ↵
      2. Allen JA ,
      3. Duke JC ,
      4. Davis KC , et al
      . Using mass media campaigns to reduce youth tobacco use: a review. Am J Health Promot 2015;30:e71–e82.doi:10.4278/ajhp.130510-LIT-237
      OpenUrlCrossRefPubMed
    4. ↵
      2. Durkin S ,
      3. Brennan E ,
      4. Wakefield M
      . Mass media campaigns to promote smoking cessation among adults: an integrative review. Tob Control 2012;21:127–38.doi:10.1136/tobaccocontrol-2011-050345
      OpenUrlAbstract/FREE Full Text
    5. ↵
      2. Wakefield MA ,
      3. Durkin S ,
      4. Spittal MJ , et al
      . Impact of tobacco control policies and mass media campaigns on monthly adult smoking prevalence. Am J Public Health 2008;98:1443–50.doi:10.2105/AJPH.2007.128991
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      2. Wakefield MA ,
      3. Loken B ,
      4. Hornik RC
      . Use of mass media campaigns to change health behaviour. Lancet 2010;376:1261–71.doi:10.1016/S0140-6736(10)60809-4
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      2. Siegel M ,
      3. Biener L
      . The impact of an antismoking media campaign on progression to established smoking: results of a longitudinal youth study. Am J Public Health 2000;90:380–6.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      2. Ruiter RA ,
      3. Kessels LT ,
      4. Peters GJ , et al
      . Sixty years of fear appeal research: current state of the evidence. Int J Psychol 2014;49:63–70.doi:10.1002/ijop.12042
      OpenUrlCrossRefPubMed
    9. ↵
      2. Marteau TM ,
      3. Hall S
      . EU’s anti-smoking stance needs to be more than frightening. BMJ 2001;323:635.doi:10.1136/bmj.323.7313.635
      OpenUrlFREE Full Text
    10. ↵
      2. Hastings G ,
      3. MacFadyen L
      . Controversies in tobacco control: the limitations of fear messages. Tob Control 2002;11:73–5.
      OpenUrlAbstract/FREE Full Text
    11. ↵
      2. Biener L ,
      3. Taylor TM
      . The continuing importance of emotion in tobacco control media campaigns: a response to Hastings and MacFadyen. Tob Control 2002;11:75–7.doi:10.1136/tc.11.1.75
      OpenUrlFREE Full Text
    12. ↵
      2. McAfee T ,
      3. Davis KC ,
      4. Alexander RL , et al
      . Effect of the first federally funded US antismoking national media campaign. Lancet 2013;382:2003–11.doi:10.1016/S0140-6736(13)61686-4
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      2. Kandra KL ,
      3. McCullough A ,
      4. Summerlin-Long S , et al
      . The evaluation of North Carolina’s state-sponsored youth tobacco prevention media campaign. Health Educ Res 2013;28:1–14.doi:10.1093/her/cys085
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      2. Pierce JP ,
      3. Macaskill P ,
      4. Hill D
      . Long-term effectiveness of mass media led antismoking campaigns in Australia. Am J Public Health 1990;80:565–9.doi:10.2105/AJPH.80.5.565
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      2. Tan N ,
      3. Wakefield M ,
      4. Freeman J
      . Changes associated with the national tobacco campaign: results of the second follow-up survey. Australia’s national tobacco campaign: evaluation report. Canberra: Commonwealth Department of Health and Aged Care, Commonwealth of Australia, 2000:21–76.
    16. ↵
      2. Cotter T ,
      3. Perez D ,
      4. Dunlop S , et al
      . The case for recycling and adapting anti-tobacco mass media campaigns. Tob Control 2010;19:514–7.doi:10.1136/tc.2009.035022
      OpenUrlAbstract/FREE Full Text
    17. ↵
      2. Biener L ,
      3. McCallum-Keeler G ,
      4. Nyman AL
      . Adults' response to Massachusetts anti-tobacco television advertisements: impact of viewer and advertisement characteristics. Tob Control 2000;9:401–7.doi:10.1136/tc.9.4.401
      OpenUrlAbstract/FREE Full Text
    18. ↵
      2. Davis KC ,
      3. Nonnemaker JM ,
      4. Farrelly MC , et al
      . Exploring differences in smokers' perceptions of the effectiveness of cessation media messages. Tob Control 2011;20:26–33.doi:10.1136/tc.2009.035568
      OpenUrlAbstract/FREE Full Text
    19. ↵
      2. Farrelly MC ,
      3. Duke JC ,
      4. Davis KC , et al
      . Promotion of smoking cessation with emotional and/or graphic antismoking advertising. Am J Prev Med 2012;43:475–82.doi:10.1016/j.amepre.2012.07.023
      OpenUrlCrossRefPubMed
    20. ↵
      2. Durkin SJ ,
      3. Biener L ,
      4. Wakefield MA
      . Effects of different types of antismoking ads on reducing disparities in smoking cessation among socioeconomic subgroups. Am J Public Health 2009;99:2217–23.doi:10.2105/AJPH.2009.161638
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      2. Peters GJ ,
      3. Ruiter RA ,
      4. Kok G
      . Threatening communication: a critical re-analysis and a revised meta-analytic test of fear appeal theory. Health Psychol Rev 2013;7(Suppl 1):S8–31.doi:10.1080/17437199.2012.703527
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      2. Hastings G ,
      3. Stead M ,
      4. Webb J
      . Fear appeals in social marketing: strategic and ethical reasons for concern. Psychology and Marketing 2004;21:961–86.doi:10.1002/mar.20043
      OpenUrlCrossRefWeb of Science
    23. ↵
      2. Halkjelsvik T ,
      3. Lund KE ,
      4. Kraft P , et al
      . Fear appeals in advanced tobacco control environments: the impact of a national mass media campaign in Norway. Health Educ Res 2013;28:888–97.doi:10.1093/her/cyt064
      OpenUrlCrossRefPubMed
    24. ↵
      2. Richardson S ,
      3. Langley T ,
      4. Szatkowski L , et al
      . How does the emotive content of televised anti-smoking mass media campaigns influence monthly calls to the NHS Stop Smoking helpline in England? Prev Med 2014;69:43–8.doi:10.1016/j.ypmed.2014.08.030
      OpenUrlCrossRefPubMed
    25. ↵
      2. Kessels LT ,
      3. Ruiter RA ,
      4. Wouters L , et al
      . Neuroscientific evidence for defensive avoidance of fear appeals. Int J Psychol 2014;49:80–8.doi:10.1002/ijop.12036
      OpenUrlCrossRefPubMed
    26. ↵
      2. Peretti-Watel P ,
      3. Constance J ,
      4. Guilbert P , et al
      . Smoking too few cigarettes to be at risk? Smokers' perceptions of risk and risk denial, a French survey. Tob Control 2007;16:351–6.doi:10.1136/tc.2007.020362
      OpenUrlAbstract/FREE Full Text
    27. ↵
      2. Bala MM ,
      3. Strzeszynski L ,
      4. Topor-Madry R , et al
      . Mass media interventions for smoking cessation in adults. Cochrane Database Syst Rev 2013;6:Cd004704.doi:10.1002/14651858.CD004704.pub3
      OpenUrlPubMed
    28. ↵
      2. Brown T ,
      3. Platt S ,
      4. Amos A
      . Equity impact of population-level interventions and policies to reduce smoking in adults: a systematic review. Drug Alcohol Depend 2014;138:7–16.doi:10.1016/j.drugalcdep.2014.03.001
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      2. Tannenbaum MB ,
      3. Hepler J ,
      4. Zimmerman RS , et al
      . Appealing to fear: a meta-analysis of fear appeal effectiveness and theories. Psychol Bull 2015;141:1178–204.doi:10.1037/a0039729
      OpenUrlCrossRefPubMed
    30. ↵
      2. Laroche M ,
      3. Toffoli R ,
      4. Zhang Q , et al
      . A cross-cultural study of the persuasive effect of fear appeal messages in cigarette advertising: China and Canada. Int J Advert 2001;20:297–317.doi:10.1080/02650487.2001.11104895
      OpenUrl
    31. ↵
      2. Andrews JC ,
      3. Netemeyer RG ,
      4. Burton S , et al
      . Effects of plain package branding and graphic health warnings on adolescent smokers in the USA, Spain and France. Tob Control 2016;25:e120–e126.doi:10.1136/tobaccocontrol-2015-052583
      OpenUrlAbstract/FREE Full Text
    32. ↵
      2. Pasquereau A ,
      3. Gautier A ,
      4. Andler R , et al
      . [Tobacco and e-cigarette in France: levels of consumption according to the preliminary results from the 2016 Health Barometer]. Bulletin épidémiologique hebdomadaire 2017;12:214–22.
      OpenUrl
    33. ↵
      Office for National Statistics. Adult smoking habits in the UK: 2016. United Kingdom: Office for National Statistics, 2017.
    34. ↵
      2. Lecorps P
      . [Educating through fear, an anti-tobacco campaign]. Sante Publique 2002;14:285–6.
      OpenUrlPubMed
    35. ↵
      2. Ferron C
      . [Influenza vaccination: was it relevant to use fear?]. Sante Publique 2010;22:249–52.
      OpenUrlPubMed
    36. ↵
      Ministère des affaires sociales, de la santé et des droits des femmes. Programme national de réduction du tabagisme 2014-2019. Paris: Ministère des affaires sociales, de la santé et des droits des femmes, 2015.
    37. ↵
      2. Wakefield MA ,
      3. Spittal MJ ,
      4. Yong HH , et al
      . Effects of mass media campaign exposure intensity and durability on quit attempts in a population-based cohort study. Health Educ Res 2011;26:988–97.doi:10.1093/her/cyr054
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      BVA. Post-test of the 2014 smoking prevention campaign "Farewell message", 2014.
    39. ↵
      2. Pasquereau A ,
      3. Guignard R ,
      4. Andler R , et al
      . Electronic cigarettes, quit attempts and smoking cessation: a 6-month follow-up. Addiction 2017;112:1620–8.doi:10.1111/add.13869
      OpenUrlCrossRefPubMed
    40. ↵
      2. Guignard R ,
      3. Wilquin JL ,
      4. Richard JB , et al
      . Tobacco smoking surveillance: is quota sampling an efficient tool for monitoring national trends? A comparison with a random cross-sectional survey. PLoS One 2013;8:e78372.doi:10.1371/journal.pone.0078372
    41. ↵
      2. Husky MM ,
      3. Guignard R ,
      4. Beck F , et al
      . Risk behaviors, suicidal ideation and suicide attempts in a nationally representative French sample. J Affect Disord 2013;151:1059–65.doi:10.1016/j.jad.2013.08.035
      OpenUrl
    42. ↵
      2. Heatherton TF ,
      3. Kozlowski LT ,
      4. Frecker RC , et al
      . Measuring the heaviness of smoking: using self-reported time to the first cigarette of the day and number of cigarettes smoked per day. Br J Addict 1989;84:791–800.doi:10.1111/j.1360-0443.1989.tb03059.x
      OpenUrlCrossRefPubMedWeb of Science
    43. ↵
      2. Yong HH ,
      3. Borland R ,
      4. Hyland A , et al
      . How does a failed quit attempt among regular smokers affect their cigarette consumption? Findings from the International Tobacco Control four-country survey (ITC-4). Nicotine Tob Res 2008;10:897–905.doi:10.1080/14622200802023841
      OpenUrlCrossRefPubMedWeb of Science
    44. ↵
      2. Borland R ,
      3. Yong HH ,
      4. Balmford J , et al
      . Motivational factors predict quit attempts but not maintenance of smoking cessation: findings from the International Tobacco Control Four country project. Nicotine Tob Res 2010;12 Suppl(Suppl 1):S4–S11.doi:10.1093/ntr/ntq050
      OpenUrlCrossRefPubMedWeb of Science
    45. ↵
      National Institute of Statistics and Economic Studies (INSEE). Consumption unit, 2016. https://www.insee.fr/en/metadonnees/definition/c1802.
    46. ↵
      2. Purmehdi M ,
      3. Legoux R ,
      4. Carrillat F , et al
      . The effectiveness of warning labels for consumers: a meta-analytic investigation into their underlying process and contingencies. Journal of Public Policy & Marketing 2017;36:36–53.doi:10.1509/jppm.14.047
      OpenUrl
    47. ↵
      2. Smith-Jackson T
      . Culture and warnings. Wogalter MS , ed. Handbook of warnings. Mahwah, New Jersey: Lawrence Erlbaum Associates, 2006:363–71.
    48. ↵
      2. Baker R ,
      3. Brick JM ,
      4. Bates NA , et al
      . Report of the AAPOR task force on non-probability sampling. American Association for Public Opinion Research (AAPOR), 2013.
    49. ↵
      2. Andler R ,
      3. Guignard R ,
      4. Wilquin JL , et al
      . Electronic cigarette use in France in 2014. Int J Public Health 2016;61:159–65.doi:10.1007/s00038-015-0773-9
      OpenUrl
    50. ↵
      2. Richard J-B ,
      3. Andler R ,
      4. Gautier A , et al
      . Effects of using an overlapping dual-frame design on estimates of health behaviors: a French general population telephone survey. Journal of Survey Statistics and Methodology 2017;5:254–74.
      OpenUrl

    Footnotes

    • Contributors RG and VN-T: conceived the study, supervised the design of the questionnaires and data collection. RG: carried out the statistical analyses and wrote the first draft of the manuscript (methods, results and discussion). KG-M: managed the literature searches and wrote the first draft of the introduction. KGM, UM, KH and VN-T: advised on the design of the analyses and substantially contributed to the interpretation of the results. All authors: revised drafts critically for important intellectual content, and reviewed and approved the final manuscript.

    • Competing interests None declared.

    • Ethics approval French Commission on Data Privacy and Public Liberties (Commission nationale de l’informatique et des libertés - CNIL).

    • Provenance and peer review Not commissioned; externally peer reviewed.