Background
Over the past 20 years, legislation in many high-income countries has increasingly restricted where and when smoking can take place (
1). In the United Kingdom, legislation which prohibited tobacco (cigarette) smoking in enclosed public places and workplaces was recently voted by the Royal Society for Public Health as the biggest public health achievement of the twenty-first century (
2). This was implemented with the primary aim of reducing exposure to smoke among groups including those working in the hospitality industry (
3), and evidence indicated that it was successful in achieving these aims. However, much attention also focused on its impacts on childhood exposure to tobacco. In particular, in opposing the legislation, some advocated the ‘displacement’ hypothesis, promoting concerns that public smoking bans would have the perverse effect of displacing smoking into the home, thus increasing children’s exposure to secondhand smoke (
4).
A large body of international evidence finds that this did not occur (
5). For example, a meta-analysis of the impact of public smoking bans on children’s exposure to secondhand smoke at home revealed an overall decrease in exposure (
1). Rather than displacing smoking into the home, legislation perhaps contributed further to denormalization of smoking in the presence of children, leading to reduced secondhand smoke exposure (
6‐
9), and changing perceptions of smoking as a ‘normal’ behaviour (
10). Throughout the past two decades of increasing regulation of smoking in the UK, youth smoking uptake has declined, while anti-smoking attitudes and normative perceptions among youth have hardened (
11). In the years following legislation on smoking public places, surveys in the UK and beyond showed strong public support for further action to limit children’s exposure to tobacco (
12‐
14), with smoking in cars carrying children banned across the UK from 2015 (
15), and restrictions of smoking in school grounds and playground from 2021 (
16,
17).
During the period since the introduction of smoke-free legislation, e-cigarettes have emerged and gained traction within UK markets, with rapid growth in use primarily by adult smokers and ex-smokers (
18) from around 2011, followed by a plateauing since 2013 (
19,
20). Although e-cigarettes are not harmless (
19), there is general consensus that they are less harmful than tobacco cigarettes (
21) and may be beneficial for quitting smoking. In a UK randomised trial, e-cigarettes combined with behavioural support were almost twice as effective as a cessation tool compared to nicotine replacement therapy (NRT) with behavioural support (
22). A recent randomised control trial found that female smokers improved their vascular health within 1 month of switching from tobacco to e-cigarettes, irrespective of nicotine content (
23). Hence, e-cigarettes may offer important harm reduction potential when used by smokers as a means of quitting smoking.
However, concerns regarding the emergence of e-cigarettes have centred largely on debates regarding effects on young people, in particular via their acting as a new gateway to nicotine addiction (
19,
24), or renormalising smoking (
25‐
27). Substantial debate has centred on whether e-cigarettes should be regulated in the same ways as tobacco, including whether their use in public places should be prohibited. In 2015 for example, the Welsh Government attempted unsuccessfully to extend legislation banning use of tobacco cigarettes in public spaces to include e-cigarettes, citing concerns that e-cigarettes will renormalise smoking (
28). In the US, 22 states/territories and over 900 municipalities, include e-cigarettes within legislation prohibiting their use in public places and workplaces where smoking is banned (
29,
30). Recent findings from a cross-sectional nationally representative school survey in the US found increased exposure to e-cigarette aerosol in public places was associated with increased susceptibility to overestimating peer tobacco use (
25).
A growing body of longitudinal studies finds that young people’s use of e-cigarettes is associated with subsequent smoking (
31). However, causality remains contested, with residual confounding remaining a likely partial explanation for these trends (
32). Youth smoking rates and attitudes in favour of smoking continued to decline during the emergence of e-cigarettes (
11), while our analyses of survey data and qualitative data on primary school pupils in Wales from the present study indicate that parental vaping is associated with perceived smoking norms only where it occurs alongside smoking (
33,
34). Indeed, children whose parents used e-cigarettes were more likely to perceive these as devices adults used to stop smoking, with perceiving e-cigarettes as a means of giving up smoking associated with lower reported susceptibility to smoking (
33,
34).
While use of e-cigarettes in public places has received significant policy attention, e-cigarette use in private spaces such as homes and cars has received less attention. Our previous research indicated that while smokers with children had increasingly restricted smoking in their home, almost half of children with a parent who smoked continued to report that smoking was allowed in their home in 2014 (
15). Although it remains contested whether secondhand vaping poses health risks to children, few health risks have been identified to date (
35). In Scotland, parents of infants in disadvantaged areas emphasised the value of using e-cigarettes in the home when they lacked direct access to outside space or could not leave their homes to smoke (
36). Secondary school pupils in England reported they preferred adult’s use of e-cigarette use in the home over tobacco smoking (
37), although parents employed strategies to protect young people from e-cigarette vapour by restricting use both indoors and outdoors. Among parents who remain addicted to nicotine, but face barriers to maintaining smoke-free home environments, e-cigarettes may act as a means of limiting childhood exposure to more harmful tobacco smoke (
38,
39).
There remains scant research on primary school pupils’ perceptions of, and exposure to e-cigarettes (
7,
30), as well as how this relates to perceived smoking norms. As discussed above, debates surrounding the increased visibility of e-cigarettes have focused on the potential for them to renormalise smoking through the appearance of a smoking-like behaviour being accepted among others (
26). However, as these pupils are the first generation to be born in the years immediately following smoke-free legislation (when smoking in the population has been decreasing) alongside the parallel emergence of e-cigarettes, understanding experiences and perceptions of tobacco among this cohort of children is important in shaping contemporary policy. The aims of the current study are to examine changes over time (from 2007 to 2019 for variables relating to tobacco and from 2014 to 2019 for those related to e-cigarettes) in:
1.
children’s experimentation with tobacco and e-cigarette use, and future use intentions;
2.
children’s exposure to tobacco smoke and vaping in public places and private spaces;
3.
children’s perceived smoking norms.
Finally, we examine the relationships of children’s exposure to secondhand tobacco smoke and e-cigarette vapour with perceived smoking norms.
Discussion
Our findings indicate that childhood experimentation with tobacco among 10–11 year olds in Wales decreased from 5 to 1% from 2007 to 2019, and has almost been eliminated following a period of increasingly comprehensive tobacco control policy (
3,
4). In this time period, e-cigarettes have emerged in the UK (
3,
4,
18,
20). By 2014, ever use of e-cigarettes among pupils was comparable to ever use of tobacco in 2007. However, while pupil awareness of e-cigarettes grew rapidly from 2014 to 2019, ever e-cigarette use has neither decreased nor increased significantly. Notably, adolescent data from Wales reported elsewhere shows that while experimentation with vaping increased from 2013 to 2017, it had declined again by 2019 (
45). It is unclear whether vaping experimentation among children remained steady from 2014 to 2019 or has also risen and decreased within this period. Child intentions for future smoking, while declining from 2008 to 2014, have however remained stable since 2014.
Child reports of smoking by their parent figures declined significantly since 2007. As parental smoking represents an important mechanism in the intergenerational transmission of smoking (
46), reduced parental smoking has likely played an important role in driving aforementioned reductions in childhood experimentation. By contrast, child reports of parental e-cigarette use rose significantly from 2014 to 2019. Most children who reported that a parent vapes also reported parental smoking at both time-points. However, reports that parent figures used
only e-cigarettes increased over time, matched by a decrease in those who reported that parents
only smoked. While data are not longitudinal, this shift may indicate that a growing proportion of parents in Wales are moving fully from smoking to vaping, consistent with data from elsewhere finding that e-cigarettes have become the most commonly used smoking cessation device in Wales (
47).
By most measures, children’s exposure to secondhand smoke in homes, cars and public places declined across the time-series, and continued to decline from 2014 to 2019. There was greater consistency across measures for continued reductions of smoking in cars than in homes since 2014, which might reflect a role of 2015 legislation prohibiting smoking in cars in maintaining downward pressure on trends for smoking in cars, which had decreased rapidly to that point (
48). The main exception was the percentage of children reporting family rules restricting smoking in the home, which after increasing from 2007 to 2014, showed no further change since 2014. However, in our qualitative research (
33), children talked of not having restrictions in their home because it was something that just did not happen, and was understood to be unacceptable without a need for formal rules. Hence, questions about smoking rules in the home may be losing their validity as indicators of whether smoking happens in the home as smoking in children’s home becomes more de-normalised. Consistent with this interpretation, a post-hoc analysis informed by our qualitative research indicated that the percentage of children who reported no rule on smoking in the home, but nevertheless, reported that nobody did smoke in their home increased substantially across the time-series. Notably however, there was also only a small and non-significant reduction from 2014 to 2019 in the percentage of children reporting that someone was smoking in their home the previous day. Hence, where indicators pointed to continued significant change, this was in a direction of continued decline, but there was more mixed evidence for change in exposure to smoke in homes relative to other locations.
In contrast with trends for tobacco exposure, exposure to e-cigarettes increased in all locations. The extent to which childhood exposure to e-cigarettes represents a public health problem regardless of links to smoking, remains a source of debate. In the US, some have expressed concern surrounding the potential harm for bystanders of secondhand vapour (
45‐
47). While less harmful than secondhand tobacco smoke, some argue that it is not entirely without risk (
49,
50) . Some have argued that a precautionary approach should be taken, with parents educated about potential harms (
51) and e-cigarette users offered NRT to protect children from e-cigarette aerosols (
52). In the UK, others have highlighted the need to assess nicotine exposure through vaping in cars to inform policies that might protect children (
53). However, others argue that e-cigarettes offer opportunities to reduce childhood secondhand smoke exposure in environments where tobacco would otherwise be used (
38). In both the US and UK, approaches to restricting e-cigarettes in private spaces (i.e. households) have been found to be less formalised and ‘rigid’ than for smoking cigarettes (
23,
46,
51,
54). As most adult e-cigarette users are current or ex-smokers (
18), a tendency for increased use of e-cigarette use in homes and cars perhaps reflects concern for protecting children from the harms of tobacco smoke, with e-cigarettes potentially representing one mechanism for keeping homes and cars free from tobacco smoke. However, further longitudinal research is needed to understand whether the growth of e-cigarette use reflects parents who would otherwise smoke in those locations now vaping instead, or introduces vaping into environments in which smoking had already been eliminated.
Growing visibility of e-cigarettes has led to debates over whether e-cigarettes may renormalise smoking, due to perceived similarities with traditional cigarettes (
26,
55). Re-renormalisation concerns have been a major driver of policy efforts (
56,
57), including the failed proposal to ban vaping in public places in Wales (
28). However, similar to other recent studies among adolescents (
11,
58), our findings of the continued decline in experimentation with and exposure to tobacco, in addition to continuing declines in perceived smoking norms for smoking, does not support the idea that renormalisation is occurring. Most normative measures showed continued decline, with fewer children in 2019 perceiving that most adults smoke, or that smoking was common in children their age. However, there was some evidence of diminishing disapproval of secondhand smoke over time since 2007, with fewer children saying they mind people smoking around them (although a large majority did say this at all timepoints). It is likely that children’s objection to people smoking around them peaked around the time of the high-profile introduction of smoke-free legislation, and associated campaign work. As fewer children experience exposure to secondhand smoke in their daily lives, they are perhaps less likely to hold a strong attitude toward hypothetical exposure. Notably, while disapproval of secondhand smoke remained lowest among those exposed to tobacco smoke (mirroring qualitative research that children who are exposed to it express a strong dislike of it (
33)), recent change over time in tolerance of secondhand smoke was not observed among children of smokers, perhaps indicating that this change was driven primarily by children with relatively low exposure to secondhand smoke.
However, taken together with the tendency for the lack of increase in formal rules on smoking, greater acceptance by children of people smoking around them may reflect a tendency for more limited communication with children regarding tobacco, and a perception that smoking is not an issue which affects children anymore. In our linked qualitative research, many non-smoking parents expressed a view of smoking as an adult issue, and a preference not to discuss tobacco with their children until they began to encounter it later in adolescence (
33). Anecdotally, during survey recruitment, some schools who declined to participate also expressed a view that tobacco was not really an important issue for children in their school anymore. Notably, while children’s experimentation with tobacco and perceived norms continued to decline, there was also little change in perceived susceptibility to future smoking from 2014 to 2019. Given that the most recent adolescent data in the UK show that following a long period of decline, adolescent tobacco use (along with other substances such as cannabis) is no longer decreasing (
59), it remains important to understand children’s perceptions of tobacco and drivers of these perceptions from a young age.
There was a clear graded relationship of exposure to tobacco with all markers of perceived smoking norms. Children with higher exposure to tobacco were more likely to perceive smoking as a normal adult behaviour, something that children do, and not to mind others smoking around them. Exposure to e-cigarette use showed a similar graded relationship, where tobacco exposure was not adjusted for. However, after adjusting for the tendency of children with high exposure to e-cigarettes to also be more likely to have high exposure to tobacco, a less clear relationship of e-cigarette exposure with tobacco norms emerged. There was little evidence of a relationship of exposure to e-cigarette with perceptions of adult smoking as a normative behaviour, or children minding that people smoked around them, although consistent with (
25), an association with perceiving that more than ‘hardly any’ children smoked remained. Hence, consistent with our previous analyses which found that associations of parental vaping with smoking norms were attenuated by parental smoking (
34), these data are consistent with a conclusion that associations of exposure to e-cigarettes with smoking norms are largely explained by the tendency for children who have high exposure to e-cigarettes also to have high exposure to tobacco.
Our study benefits from using nationally representative surveys of pupils in Wales, repeated over a span of 12 years. However, limitations include a repeat cross-sectional survey design, which means that causation cannot be inferred. It also relies on self-report measures that are subject to biases which may have changed over time. Declines in parental smoking for example may indicate that parents are hiding their smoking from their children, or that children are increasingly reluctant to report that parents smoke, given increased stigmatisation. Further, in line with recent trends in social surveys, the 2019 survey achieved a substantially lower response rate than earlier rounds. While this did not impact representativeness according to measures obtained, estimates of change over time may be confounded by unmeasured differences in samples due to differential response biases. While for the vast majority of items, data completeness was very high, for some models missingness exceeded 5% and should therefore be treated with caution. Our sample consisted of Year 6 pupils (age 10–11) and findings may not generalise to older youth where smoking and e-cigarette rates tend to be higher (
11,
42). Our measurement of smoking susceptibility was limited to one questionnaire item to enable change over time with past surveys; however this differs from other existing definitions using multiple items (
60,
61). While for many items, survey design is bounded by decisions made in earlier surveys, with item consistency maintained to enable analysis of change over time, our analysis of the increasing disjuncture between reports of home smoking rules and actual smoking in the home illustrates that question meanings may change over time, posing challenges for interpreting these as evidence of change over time in the underlying construct, or in how they are interpreted in an environment of diminished smoking prevalence.
Conclusions
Notwithstanding the above caveats, this study provides evidence that within primary schools in Wales, children’s experimentation with, and exposure to, tobacco cigarettes continues to decline, representing a major success of comprehensive tobacco control measures in recent years. There is consistent evidence of declines in exposure to secondhand smoke in cars, with slightly less consistent evidence that the earlier declines to exposure in the home from 2007 to 2014 have continued post 2014. These declines appear to have largely continued alongside co-occurring increases in exposure to e-cigarettes. Ongoing surveillance of trends in childhood perceptions of tobacco and e-cigarettes remains a priority as the products and the landscape in which they are bought and sold continue to evolve. Longitudinal research is needed to better understand patterns in parental movement from tobacco cigarettes to e-cigarettes, including whether e-cigarettes are displacing tobacco smoke in homes and cars, or are being introduced to environments where tobacco smoking had already been eliminated. While these findings indicate that substantially fewer children are now exposed to tobacco, experiment with tobacco themselves, or view tobacco as a normal adult behaviour than in 2007, a sizeable minority continue to be exposed to adults’ tobacco use, and to misperceive that most adults smoke. If further declines in uptake of smoking in youth and further reductions in children’s exposure to tobacco are to be achieved, it is important not to be complacent and that childhood perceptions of, and exposure to, tobacco continue to be prioritised for public health intervention.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.