Results: the role of e-cigarettes in the renormalisation of smoking prior to full implementation of the Tobacco Products Directive regulations

Some text in this chapter has been reproduced from Hallingberg et al.10 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.

In addition, some text in this chapter has been reproduced from Brown et al.105 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.

This chapter aims to provide an understanding of the context into which TPD regulations were implemented, by addressing the following research questions:

• Have trajectories in young people’s ever and current tobacco smoking been significantly interrupted (positively or negatively) by growing prevalence of e-cigarettes?
• How do young people perceive risks and social norms surrounding e-cigarettes?

The chapter does so by bringing together quantitative analyses of trends in young people’s tobacco use before and after the emergence of e-cigarettes in the UK, with qualitative interview data from young people in seven schools exploring young people’s perceptions of e-cigarettes and smoking. For change in secular trends, to strengthen or challenge conclusions regarding causal inference, we examine plausibility and specificity. Specificity is assessed by examining whether change in trend is specific to tobacco or occurs for other substance use behaviours whose trajectories would not plausibly be affected by growth in e-cigarette use (e.g. alcohol and cannabis use). Plausibility is assessed (1) by testing whether or not trends in young people’s attitudes towards smoking as an acceptable thing for young people their age to do changed during emergence of e-cigarettes and (2) through qualitative interviews with young people regarding processes through which e-cigarettes might renormalise smoking.

Quantitative analysis

Trajectories of young people’s smoking prevalence from 1998 to 2015

The analytical sample for smoking prevalence from 1998 to 2015 included 242,855 and 243,111 young people for ever and regular smoking, respectively. Table 3 presents percentages (and 95% CIs) for each outcome from 1998 to 2015, overall and for each country. In all countries, there was a consistent downwards trend over time in both outcomes. In 1998, most young people (around three in five) aged 13–15 years had tried smoking. Over the course of the time series, this declined to become a minority behaviour, with approximately one in five young people reporting having tried smoking in 2015. In 1998, approximately one in five young people reported being regular smokers across countries. Regular smoking fell to between 4% and 6% in each country by the last measurement point. For both ever and regular smoking, there was some evidence of decline beyond the ‘intervention’ point (i.e. from 2011 to 2015 when e-cigarette use was becoming more prevalent in the UK).

Table 4 shows odds ratios (ORs) and 95% CIs from segmented regression analyses for ever and regular smoking (for the whole sample and for gender and year group subgroups). In our a priori-specified models, which assumed a linear decline in smoking from 1998 to 2010, there was evidence of negative change in both level and post-slope terms after the intervention point for ever smoking. This change was consistent with a conclusion that ever smoking declined faster during the period when e-cigarettes were increasing in prevalence in the UK than it had been up to that point. However, visual inspection provided some evidence of non-linearity during the baseline period. Addition of a quadratic term confirmed that non-linearity was significant. Accounting for non-linearity led to a changed conclusion of no significant change in trend for ever smoking. For regular smoking, there was no evidence of change in trend in linear models, which is consistent with a conclusion of continued decline at the same rate as the secular trend. However, after adjusting for non-linearity of the baseline trend, a marginally significant (p = 0.03) slowing in rate of decline for regular smoking after the ‘intervention’ point was found. In both cases, significant reduction in level at the intervention point is evident. In a range of sensitivity analyses, ORs for change in trend for ever tobacco smoking ranged from 0.98 to 1.04, and were mostly non-significant, hence, providing little evidence of change in trend for ever use. ORs for change in trend for ever smoking post 2010 ranged from 0.99 to 1.11, with most ORs supporting a conclusion of a significant positive change in trend (see Report Supplementary Material 1, Table 1).

In subgroup analyses, the slowing of decline for regular smoking was limited to groups for whom rates had declined most rapidly up to 2010 (i.e. girls and 13-year-olds). Additional subgroup analyses by free school meal entitlement (limited to Scotland and England) indicated a marginally significant increase in rate of decline of ever smoking for young people with a higher SES (high SES: OR 0.95, 95% CI 0.91 to 1.00, p = 0.043; low SES: OR 1.01, 95% CI 0.91 to 1.12, p = 0.919). For both groups, change in trend for regular smoking was similar to the whole sample, falling short of significance in both instances (high SES: OR 1.03, 95% CI 0.88 to 1.21, p = 0.0.561; low SES: OR 1.03, 95% CI 0.94 to 1.11, p = 0.734). Country-specific analyses according to differing measures of SES are presented in Report Supplementary Material 1 data, but should be treated with caution given the small number of data points available for individual countries other than England (for additional subgroup analyses see Report Supplementary Material 1, Tables 2–6).

Specificity of change in trend: trajectories in ever alcohol and cannabis

To evaluate whether or not changes in trend were specific to smoking, we examined time, level and post-slope terms for the same period for ever alcohol and cannabis use. Table 5 presents percentages (and 95% CIs) of young people who reported having tried alcohol and cannabis by year and country. At first measurement in 1998, most young people reported that they had tried alcohol, declining in each country to approximately half by the end of the time series. Approximately one-quarter of young people in Scotland and England, and one-third of young people in Wales, reported having tried cannabis in 1998, declining to around 1 in 10 by the end of the time series.

As indicated in Table 6, in models assuming a linear baseline trend, there was evidence of a slight acceleration of the decline in alcohol use and no change in trend for cannabis use. However, as for tobacco use, adjusting for non-linearity of the secular trend altered trend estimates. For both alcohol and cannabis use, there was a significant positive change in trend, with the decline in both slowing after the ‘intervention’ point. This was consistent across subgroups. Hence, the slight disruption in trend for regular smoking was reflected, to a larger extent, in other substances for which e-cigarettes would not be expected to cause a change in trend. Both were measures of ‘ever’ use, whose counterpart measure for tobacco did not show evidence of change in trend, hence, suggesting that any slowing in decline for smoking was lagging behind change in trend for other substances. In a range of sensitivity analyses, ORs for change in trend for alcohol ranged from 1.05 to 1.16 and for cannabis ORs ranged from 1.05 to 1.24. The ORs were significant in all analyses, consistently indicating a significant positive change in trend (i.e. slowing in decline) for ever alcohol and cannabis use (see Report Supplementary Material, Table 1).

Plausibility of change occurring via renormalisation

Trajectories of young people’s smoking attitudes from 1999 to 2015

Table 7 presents percentages (and 95% CIs) of young people reporting that trying smoking or smoking weekly is OK for someone their age, by year and country. When first asked in England in 1999, a large majority (70%) of young people reported that it was OK for someone their age to try smoking. However, by the end of the time series in England in 2014, this percentage had approximately halved. This question was first included in Scotland in 2006, at which point slightly more than a half of young people reported that it was OK to try smoking, falling to one-quarter by 2015. The percentage of young people reporting that it was OK to try smoking appeared to continue to decline beyond 2010. In England only, from 2003, young people were also asked whether or not it was OK for someone their age to smoke weekly, with approval of regular smoking substantially lower than for experimentation (36% vs. 64%). Approval of regular smoking declined across the time series, with approximately one in seven young people reporting that it is OK for someone their age to smoke weekly by 2014. As for approval of trying smoking, this decline continued beyond the ‘intervention’ point.

Table 8 presents findings from segmented regression analyses of change in level and trend for smoking attitudes following emergence of e-cigarettes. In all models, there was a significant decline over time in the percentage of young people reporting that smoking was OK. There was some inconsistency in size and direction of change in level across models and subgroups, with these changes mostly non-significant. However, across both linear and quadratic models, and most subgroups, the rate of decline in the percentage of young people saying that trying smoking is OK and weekly smoking is OK increased post 2010. Hence, data are consistent with a conclusion that during the period of growing prevalence of e-cigarettes the percentage of young people considering smoking to be OK continued to decline and, indeed, declined at a faster rate than the baseline trend. In a range of sensitivity analyses, ORs for change in trend for attitudes towards smoking post 2010 ranged from 0.83 to 0.88, indicating a significant acceleration in decline in all cases (see Report Supplementary Material, Table 1).

Further subgroup analyses by free school meal entitlement indicated that change in trend for both higher and lower SES groups was of similar magnitude for saying that trying smoking is OK (OR 0.84, 95% CI 0.81 to 0.88, p < 0.001 for high SES; OR 0.85, 95% CI 0.77 to 0.95, p = 0.003). The increase in rate of decline for smoking attitudes for saying that weekly smoking is OK was significant for young people not entitled to free school meals only (OR 0.80, 95% CI 0.72 to 0.88; p < 0.001), with a smaller and non-significant increase in decline for young people entitled to free school meals (OR 0.96, 95% CI 0.76 to 1.21; p = 0.742).

Views of young people on smoking and vaping norms

To further explore the plausibility of observed changes in trend for smoking occurring due to renormalisation, this section presents qualitative analyses from interviews with 76 young people (boys, n = 37; girls, n = 39) within seven schools in England (n = 2) Scotland (n = 2) and Wales (n = 3), collected from March to June 2017, around the end of the transitional period in the implementation of TPD regulations. Data presented here have been previously published in Brown et al.105 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text. Data are discussed under the following four themes:

• Perceived similarities and differences between e-cigarettes and tobacco.
• Perceived prevalence of vaping and impacts on smoking norms.
• Supply, device characteristics and vape marketing.
• Attitudes towards nicotine in vapes.

Country [i.e. Wales (W), Scotland (S) and England (E)] and school code, group number and gender are indicated in brackets.

Perceived similarities and differences between e-cigarettes and tobacco cigarettes

The term ‘e-cigarette’ was widely rejected by participants, with ‘vapes’ and ‘vaping’ generally preferred. Participants most commonly described ‘tank’- or ‘pen’-style devices as used by their age group, rather than ‘cig-a-like’ models. There was general awareness of how vapes functioned, including the need to charge them and the production of vapour rather than smoke. When asked why young people vaped, reasons included for fun with peer group, for flavours, and to show off to peers both in real life and online through posting tricks on social media. These, predominantly social, reasons contrasted with perceived reasons for adult use, described primarily as to stop smoking:

Like, you don’t see people our age going round with the ones that look like cigarettes ... Maybe adults would, the ones trying to get off smoking.

Several commonalities between vapes and tobacco cigarettes were described, such as potential addictiveness, nicotine content and potential to be harmful. Stated differences were more extensive, with non-nicotine e-liquid options and range of flavours most commonly referenced and vaping defined as different from smoking. Developments in technology, including styles of device and range of flavours, appeared to enhance differentiation from tobacco cigarettes.

In discussing perceived risks of vaping, there was much variation in responses within groups, with peers often disagreeing over risks or stating something that others were unfamiliar with leading to discussion over the validity of claims. Although smoking harms are taught in school from an early age, this was not the case for vaping, with no young people reporting that their school had covered e-cigarettes and frequent reference to getting information from social media. There was also notable variation in awareness of the legal age for vaping in the UK. Types of risks identified also often differed from risks of smoking, with health risks, such as cancers and respiratory conditions, less likely to be mentioned than for smoking. For vaping, risks commonly focused on anecdotes of operating or charging problems, such as devices ‘blowing up’, with young people frequently citing stories from social media and unknown toxicity of the chemicals in liquids. Although clearly viewed as less harmful than smoking by most young people, notions of ‘harming differently’ were evident:

Some people say they have even more ... they claim to be less harmful but actually they’re more harmful in fact, apparently.

Many young people expressed caution that harms may yet be discovered in the future, as had happened with smoking:

It’s just with e-cigarettes, you are not sure what’s bad about them. Like, you know some of them are nicotine. So it’s just, what health impacts? Smoking, you know that you can get lung cancer, but e-cigarettes there isn’t the same health issues.

Perceptions of the prevalence of vaping and norms for smoking and vaping

Although advised that they did not have to discuss their own experiences with e-cigarettes, many young people did discuss trying vaping. This discussion often occurred at group level, with all young people in a group having tried e-cigarettes or none having done so, reflecting shared norms and behaviours within groups of friends. Being a regular vaper was rarely discussed but emerged in two male groups, where members reported regularly vaping non-nicotine liquids to do tricks. This patterning by group was not as evident for smoking, with fewer young people stating that they had tried smoking and just a few individuals stating that they were regular smokers. Most participants agreed that young people’s vaping had increased in the last few years, and most indicated that they saw more young people vaping than using tobacco cigarettes now. Vaping was largely described by participants as casual, or occasional, and strongly socially motivated, for example occurring with peers at parties where a device is brought by one person and shared. However, many participants also suggested that vaping was a fad that may have peaked:

Not too long ago there was like a whole new craze about it because everybody was thinking it would be a safer alternative to smoking and ever since then the boom of it has started to die down and we haven’t really seen much of anymore.

Regular use (i.e. daily or weekly) of e-cigarettes by young people was described as rare, adopted as either a means to perform tricks or as something primarily done by smokers. Habitual users were described as more likely to either be dual users of tobacco and vapes or as members of smoking peer groups. It was also frequently stated that this group of users was more likely to be engaged in alcohol and cannabis use, with regular vaping perceived as a component of a broader clustering of risk behaviours. Regular smokers were described in negative terms as the ‘kind of disruptive ones’ (E1, 3, F) or as the ‘people who think they’re hard’ (S1, 2, F), and were seen to be less academically engaged:

I don’t think it’s the more intelligent people that smoke.

Some participants also suggested that regular smokers they were likely to use cannabis as part of their identity as smokers:

... a lot of them will just smoke anything they can get their hands on.

Increased prevalence of casual vaping, including popularisation of tricks among peer groups and on social media, was described as making experimental use acceptable, as well as reducing concerns over associated risks. ‘Having a go’ on a vape as a social activity was met with little disapproval, with reasons for experimental use cited as (1) because peers were doing so, (2) for flavours (including swapping flavours) and (3) for doing ‘tricks’. However, many participants differentiated between acceptance of casual vaping and disapproval of regular use unless as an aid to stop smoking. Regular use was otherwise pointless and, in some cases, ‘not cool’:

If you’re just doing it to be sociable and pretend, and say, ‘Yeah I’m vaping’, and just blowing out, you’re just going to be bullied, not like bullied constantly, but you’re going to be taken the mick out of you.

An exception was regular vaping to perform tricks:

I think most people if you’re going to vape you’ve got to do tricks with it in our year because if you’re not doing tricks you’re not cool.

Throughout the data, there was little variation by gender for most themes; however, an area where variation occurred was consensus in most groups that boys were more likely to be involved in trick performance. A small number of non-smoking boys discussed this and its sharing through social media, with those who did so confident of little censure from peers:

You see some people sharing videos on Facebook and they’re having a big puff of this e-cigarette and they’re making rings with it and it looks cool, it’s almost like a new kind of trend.

Perceived peer and family disapproval of regular smoking or vaping was strong. When discussing expected peer reactions to people becoming smokers, responses included the expectation of respect for their choice [e.g. ‘... if you want to do them just do them’ (E1, 4, B)] and the, more common, expectation of negative responses, usually from concern over addiction:

I know my friendship group would say like, what are you doing? Just stop it now while you’ve got the chance.

Reactions of disgust were also mentioned:

I think with my friends it would repel them.

Potential risk of social loss due to exclusion from peer groups was often described as a consequence of becoming a smoker, but less evident for vaping. Shared norms within groups of friends were evident and tended to manifest in relative acceptance of vaping across all in the group or shared rejection, with some groups keen to identify themselves as not the ‘types’ to do that. Most group discussions suggested that smoking by young people was driven by addiction or a deliberate adoption of risk:

... because they think it [tobacco] makes them look cooler because it can harm you.

In several groups, it was suggested that, rather than making smoking acceptable, the increased availability of vaping actually made smoking seem more socially unacceptable by comparison, as there is now a less harmful alternative:

I think with the introduction of e-cigarettes, I mean tobacco was already considered pretty dangerous but I feel like people are being more sceptical about tobacco use and stuff, and e-cigarettes ... I think it kind of makes us think that tobacco is more dangerous than it was before.

All groups stated that parents would have a stronger negative reaction to smoking than vaping, although parents reacting badly to the latter was also reported by many and, again, tended to be consistent in groups. Some groups suggested, however, that parents may be less anxious about vaping than tobacco smoking:

I think they would prefer vaping to smoking, I would say, if it had to be one.

Reasons for strong parental disapproval of smoking included risks being so well established that ignorance of harm could not be a defence, and decreased peer influence to smoke due to lower prevalence and acceptability in young people:

I don’t think they’d see any excuse for tobacco cigarettes because there’s no pressure to have them, there’s no influence.

Some participants, particularly participants in groups that were keen to identify themselves as not being the type to vape, further expressed concern that parents perceived a greater alignment of drug use and tobacco and, therefore, were more likely to fear that their child was involved in other drug use, such as cannabis, if their child was smoking tobacco than if they were vaping:

I think that after that they will just think worse of us, like they think we’d be doing drugs and stuff like that because of that one thing [tobacco].

Device characteristics, supply and marketing

There was general agreement within and across groups that vapes were easily obtained, and more accessible than tobacco cigarettes. Young people predominantly discussed availability through online sales or informal school supply chains. Within-group discussion was common, with participants sharing stories of who in or around school may be the one selling supplies. Although some participants cited young people in school selling tobacco cigarettes, this was less common, with tobacco largely expected to come from older people supplying it, including peers/siblings and proxy purchasing. The dominance of this informal supply line meant that most young people had seen tobacco cigarettes outside the packaging only, with brand awareness low. Although most participants were aware of specialist shops in their area or the nearest larger town, few participants knew these shops by name or saw them as a likely source of supply, with informal and online routes seen as less restricted. Several participants cited that vaping had come into school first through the ‘populars’ and had spread as a casual and socially driven practice, including sharing flavours and performing tricks. This contrasted with the more deviant status of smoking – and smokers. Flavours were highly significant, both for taste and social gains of swapping, and lively within-group discussion was common with regard to who had tried which flavours and which were preferred:

I just liked the different flavours. ’Cos my friend had jam donut, another friend had gummy bear flavour and Heisenberg which is a minty flavour which is quite nice.

Flavours further differentiated younger, casual users from both adult vapers and from tobacco smokers:

Because they’re always shown with all the flavours and I think if you were trying to quit smoking you wouldn’t be bothered by all the flavours.

It was common for participants to discuss vaping devices as accessories, with appeal associated with the style and cost of equipment:

I think what makes it more appealing, there are some designs on them. So people are just like, ‘Oh that’s a cool design’. And different flavours and they’re comparing with their friends. Oh look at yours, yours is red and mines like camouflage.

For the few participants who reported vaping for tricks, device type was also important, including use of ‘bigger boxes’:

I think our age group normally go for them because you can make a fat cloud out of them.

This sense of vaping showing off financial status and being started by more popular young people contrasted with comments on the typology of tobacco users, commonly defined negatively, including as being the ‘scruffy’/‘naughty’/‘disruptive’ young people who were already likely to reject school rules and behavioural norms.

Attitudes towards nicotine in vapes

In initial discussions about similarities and differences between vapes and tobacco cigarettes, there was widely varying understanding of the presence or absence of nicotine in e-cigarette liquids. Some participants were unsure or unaware that nicotine could be present and were surprised when others highlighted this, stating that they had thoughts liquids were just:

... syrupy stuff in different flavours.

When asked to describe reasons why young people smoke, addiction was frequently cited, but this was not the case in relation to vaping, where social reasons, such as fitting in and showing off where dominant. However, when asked to identify potential risks of vaping, addiction was cited by at least one participant in most groups. Although some participants stated that they would avoid nicotine-based liquids because of fear of addiction, for many participants who described experimental use, nicotine content was either an unknown or less important than flavour:

I don’t know. It’s not really something you ask if its nicotine or no nicotine. You more just ask them what kind of flavour is it

This perhaps reflects the context of use, where vapes used by young people commonly belonged to someone else and were passed around for people to try rather than being something actively sought. Where vapes were tried once with little intention of using again, concerns over nicotine were secondary to social factors driving experimentation:

Int: Yeah, so just once?

E1, 3, F, P: Yeah.

Int: And do you know if that had nicotine or no nicotine in it?

E1, 3, F, P: I have no idea [laughs]

Chapter summary

Ever use of tobacco continued to decline in the period during which e-cigarette use was growing, with little evidence of change in rate of decline. There was a marginal slowing in decline for regular use. However, this was reflected to a greater extent across other substances, for which slowing in declines in ever use were occurring, by contrast to counterpart measures of ever tobacco use. Our analysis of quantitative and qualitative data did not identify plausible mechanisms via which e-cigarettes might have caused slowing in decline of regular smoking. Young people’s attitudes towards whether or not its OK for young people to smoke continued to harden against smoking during this time. In qualitative interviews, e-cigarettes were viewed by young people as distinct from tobacco use, with widespread rejection of the term ‘e-cigarette’ in favour of terms such as ‘vapes’. Growing experimentation with e-cigarettes was widely approved of, with young people discussing use of e-cigarettes in particular social contexts, such as at parties and imitating tricks, but regular use outside this context was not. Although findings provide limited support for the notion that e-cigarettes renormalise smoking, the findings highlight some concerns with potential normalisation of some forms of e-cigarette use and potentially nicotine, with many young people reportedly unaware of whether or not devices they had used contained nicotine. Young people emphasised informal supply chains and a perception that e-cigarettes were obtained via peers or the internet rather than through retailers, with social media channels also seen as a key source of marketing exposure. Hence, although data prior to implementation of the TPD regulations indicated that regular use remained limited and concentrated primarily among smokers, preventing young people’s use of e-cigarettes remains a priority.