In recent years, the use of e-cigarettes, also known as “vapes”, has emerged as a global public health priority. The significant growth in e-cigarette use is particularly troubling in adolescent populations, especially among youth who have never used tobacco cigarettes (Aljandaleh et al., 2020; Gallus et al., 2021; Health, 2016; Kennedy et al., 2017; Yoong et al., 2021). In Australia, the most recent (2019) nationally representative survey reported that 10% of 14–17-year olds had ever used an e-cigarette (Welfare, 2020). Most recently, data from a survey of 4204 adolescents aged 14–17 years old showed this number has increased to 26% (Gardner et al., 2023). This is a significant concern due to the unknown long-term health effects of e-cigarettes (Banks et al., 2023). Given the potential for harm and the rapidly growing trajectories of use in youth, there is an urgent need to establish a robust body of evidence to inform prevention efforts within Australia’s unique legal (Alcohol & Drug Foundation, 2023) and social context (Jongenelis, 2023).

Peers play a significant role in shaping adolescent risk behaviour (Forman-Alberti, 2015). Social influence theory explains this relationship as occurring due to the function of social comparison, whereby individuals determine whether their own behaviour is appropriate through continuous comparison with others’ (Maxwell, 2002). Additionally, peer selection, whereby an individual selects a peer group based on their behaviour (e.g. using e-cigarettes), is another key process contributing to e-cigarette use behaviour among social networks (Valente et al., 2023). Adolescents in particular look to their peers to shape normative beliefs, and interpret information related to risky behaviour (Maxwell, 2002). Much adolescent e-cigarette research currently focuses on demographic correlates (e.g. age, gender, ethnicity) of use (Bold et al., 2017; Camenga et al., 2014; Carroll Chapman & Wu, 2014; Donaldson et al., 2021; Gardner et al., 2023; Hrywna et al., 2020; Krishnan-Sarin et al., 2015; McCabe et al., 2020); however, less is known about modifiable peer factors, such as perceived peer vaping behaviour and bullying victimisation and/or perpetration, which could inform future prevention efforts.

Peer substance use, both actual and perceived, has been shown to be associated with increased adolescent substance use both internationally (Dishion & Loeber, 1985; Er et al., 2019; Simons-Morton & Farhat, 2010; Watts et al., 2023) and in Australia (Gazis et al., 2009; White et al., 2013). Emerging international research, predominantly from the USA, indicates this association holds true for adolescent e-cigarette use (Cavazos-Rehg et al., 2021; Durkin et al., 2021; Patanavanich et al., 2021; Valente et al., 2023; Wang et al., 2022). It is however important to note that all but one study (Valente et al., 2023), measured perceived, rather than actual peer e-cigarette use. For example, in a survey of 562 adolescents in the USA, the perceived proportion of friends who use e-cigarettes directly correlated with a higher likelihood of having ever used an e-cigarette (Durkin et al., 2021). In Australia, a cross-sectional study of 15–30-year olds found a positive association between number of friends who use e-cigarettes and increased e-cigarette use (Pettigrew et al., 2023); however, younger adolescents (aged 15–18) only made up a minority of the sample. Alarmingly, e-cigarette use now commonly occurs in many contexts, including in schools (Jongenelis & Robinson, 2023; Pettigrew et al., 2023). Further research among Australian adolescents is needed to specifically examine the association between perceived peer use and rates of e-cigarette use, as well as other important peer- and school-related factors, such as bullying.

The association between bullying perpetration and substance use (including alcohol, marijuana and cigarette use) is well established (Carlyle & Steinman, 2007; Gaete et al., 2017; Kelly et al., 2015; Lambe & Craig, 2017; Picoito et al., 2019; Stone & Carlisle, 2017; Thomas et al., 2017). To our knowledge, only one US study has examined the association between bullying perpetration and adolescent e-cigarette use, finding nicotine e-cigarette use to be associated with both in-person and online bullying perpetration (Boccio & Leal, 2022). Despite a less expansive literature base, research has identified an association between bullying victimisation and increased substance use (Kelly et al., 2015; Lee et al., 2022; Moore et al., 2017; Tharp-Taylor et al., 2009). Emerging research suggests this relationship holds firm when applied to e-cigarette use, although none has been conducted in Australia (Azagba et al., 2020; Boccio & Leal, 2022; Doxbeck, 2020; Ihongbe et al., 2021; Liu et al., 2023; Mereish et al., 2023; Tabaac et al., 2021). Only one study has examined the relationship between bully-victim status (i.e. an individual who has been both a victim and perpetrator of bullying) and e-cigarette use, finding they were strongly associated (Boccio & Leal, 2022). To our knowledge, this study will be the first to examine the relationship between bullying (victimisation and/or perpetration) and e-cigarette use in a large Australian sample.

Emerging research suggests that a better understanding of the relationship between peer factors and e-cigarette use in adolescents is needed, as peer-related behaviours have the potential to be modified through preventative intervention. Using cross-sectional data collected as part of the Health4Life Study (Teesson et al., 2020), a school-based cluster randomised controlled trial (RCT), this study aims to fill that gap by examining the associations between peer-related risk factors and e-cigarette use in a large sample of Australian adolescents. Specifically, this paper will address the following research questions:

  1. 1.

    Is perceived peer e-cigarette use associated with adolescent e-cigarette use?

  2. 2.

    Is bullying (victimisation and/or perpetration) associated with adolescent e-cigarette use?

Methods

Participants

This study used cross-sectional data collected from participants who completed the 36-month post-baseline assessment of the Health4Life study, a cluster RCT conducted in 71 secondary schools in Australia from 2019-2022.

Despite the longitudinal nature of the cohort, data on e-cigarette use were only collected during the final follow-up period between July-December 2022 (n = 4204; 63% of baseline sample; see Table 1). Further details, including recruitment and consent procedures, can be found elsewhere (Champion et al., 2023; Teesson et al., 2020).

Table 1 Participant characteristics stratified by “ever used e-cigarette”
Full size table

Measures

E-Cigarette Use

Participants were provided with a definition of e-cigarettes, as well as a list of additional names they are called (e.g. vapes, Juul, pods). Ever use of e-cigarettes was measured using a single item, “Have you ever used an e-cigarette, even one or two puffs?”, with response options “yes” or “no”.

Bullying

Bullying perpetration, victimisation and bully-victim status were determined by using two questions from the Olweus Bully/Victim Questionnaire (Olewus, 1993), which has demonstrated good psychometric properties (Kelly et al., 2015; Kyriakides et al., 2006). Participants responded “yes” or “no” to the items “have you ever been bullied?” and “have you ever bullied others?” A bully-victim variable was created to represent those who reported both bullying and being bullied by others.

Perceived Proportion of Friends Who Use e-Cigarettes

Perceived peer e-cigarette use was measured using a single item, “About what proportion of your friends and acquaintances have used e-cigarettes (even one or two puffs)?” This was measured via a 6-point Likert scale, with responses coded 0- “none”, 1- “less than half”, 2- “about half”, 3- “more than half”, 4- “all or almost all”.

Sociodemographic Covariates

Gender identity was measured using an item extracted from surveys administered by ACON (a LGBTQIA+ affirmative organisation). Responses included “male”, “female”, “non-binary/gender fluid”, “other identity” and “prefer not to say”.

School socio-educational advantage was measured using the Index of Socio-Educational Advantage (ICSEA). Each school in Australia is allocated a relative position (scaled to a normal distribution), which captures features of the school student population.

School geographical remoteness (major city or regional) was used as a substitute for participant geographic remoteness, as many students were unaware of their current postcode.

Statistical Analysis

Separate multilevel logistic regression models investigated the associations between ever using an e-cigarette and each peer-related factor: bullying perpetration, bullying victimisation, bully-victim status and perceived peer use. To account for the clustered nature of the data, we included a random effect for school. To account for any potential effect of the Health4Life intervention on student outcomes, we controlled for intervention status in all models. To test the robustness of the associations between our outcomes, we ran the same models controlling for sociodemographic characteristics that may have been related to using e-cigarettes: age, gender, state (New South Wales, Queensland, Western Australia), school type (catholic, independent, public), school geographic remoteness (major city vs regional) and relative socio-economic status. We report both the adjusted and unadjusted odds ratios for the associations between each peer risk factor and lifetime e-cigarette use. All models were conducted in R (version 4.3.1) (RStudio Team, 2021) using the lme4 (Bates et al., 2015) package.

Results

Sample Characteristics

Sample characteristics are reported in Table 1. A total of 4204 participants were included in these analyses (Mage = 15.70, SD = 0.60; female = 47.1%). Twenty-six percent of the sample reported having ever used an e-cigarette. In comparison to those who reported never using an e-cigarette, those who had used an e-cigarette demonstrated higher rates of bullying perpetration, bullying victimisation, being a bully-victim and perceived that a higher proportion of their friends had used e-cigarettes. E-cigarette use by socio-demographic characteristics (gender, state, school type, remoteness and relative socio-economic status) for this sample have been reported elsewhere (Gardner et al., 2023).

Peer Correlates of e-Cigarette Use

Results of the multilevel logistic regressions are reported in Table 2. Higher perceived peer e-cigarette use was associated with greater odds of ever having used an e-cigarette (OR = 2.59, 95% CI = 2.42, 2.77, p < .001). Bullying victimisation (OR = 1.26, 95% CI = 1.08, 1.46, p = .004), bullying perpetration (OR = 3.00, 95% CI = 2.45, 3.66, p < .001) and being a bully-victim (OR = 2.58, 95% CI = 2.06, 3.24, p < .001) were associated with increased odds of ever having used an e-cigarette.

Table 2 Summary of separate mixed effects logistic regression models investigating the association between ever having used an e-cigarette and perceived peer e-cigarette use, bullying victimisation, bullying perpetration and bully-victim status
Full size table

Discussion

This is the first paper to examine the relationships between peer factors and e-cigarette use in a large sample of Australian adolescents. These findings build on international research that has demonstrated associations between adolescent e-cigarette use and both perceived peer use (Barrington-Trimis et al., 2015; Cavazos-Rehg et al., 2021; Durkin et al., 2021; Patanavanich et al., 2021; Valente et al., 2023; Wang et al., 2022) and bullying involvement (Azagba et al., 2020; Boccio & Leal, 2022; Doxbeck, 2020; Ihongbe et al., 2021; Liu et al., 2023; Mereish et al., 2023; Tabaac et al., 2021). Our findings indicate that there is a robust cross-sectional association between ever using an e-cigarette and bullying perpetration, bullying victimisation, bully-victim status and perceived peer use, even after adjusting for key sociodemographic characteristics.

The finding that those who have ever used an e-cigarette, in comparison to non-users, were 2.59 times more likely to perceive their peers as using e-cigarettes, suggests that peer influence, peer selection and social context potentially play a role in adolescent e-cigarette use. These results are consistent with international literature that emphasises the role of social networks in e-cigarette uptake and continuation in adolescents (Pichel et al., 2022). E-cigarette use behaviour may be more normalised or socially accepted in some peer groups, and adolescents may select or be influenced by peers who share similar beliefs (Valente et al., 2023). An association between perceived peer use and e-cigarette use presents an opportunity to challenge misconceptions through normative education (e.g. 76% of the sample perceived at least half of their peers to have used an e-cigarette, yet only 26% reported ever trying an e-cigarette), and aligns with research that has found adolescents tend to overestimate their peers’ substance use (Henneberger et al., 2019).

An implication of these results is the possibility of leveraging the influence of peers within prevention programs. There is good evidence for the use of peer-led and social influence principles to prevent and reduce substance use, including e-cigarette use (Veenstra & Laninga-Wijnen, 2022; Wyman et al., 2021). These projects incorporate established behaviour modification techniques (challenging social norms, increasing knowledge and refusal skills, improving health literacy) alongside a peer-to-peer education structure, capitalising on the strong theoretical basis of social diffusion theory (Wyman et al., 2021). Peer-led and social influence programs provide the opportunity to challenge normative beliefs about vaping (Sanchez et al., 2019). It may also be important within these interventions to identify particular clusters of peer groups that are at-risk and offer additional supports alongside the peer-led intervention.

Interestingly, perpetrators of bullying were three times more likely than non-perpetrators to have used an e-cigarette, while bullying victims were more likely than non-victims to have used an e-cigarette. While these associations individually are consistent with existing literature (Azagba et al., 2020; Boccio & Leal, 2022; Doxbeck, 2020; Ihongbe et al., 2021; Liu et al., 2023; Mereish et al., 2023; Tabaac et al., 2021), it is noteworthy to examine the possible explanations for the difference in odds for perpetrators and victims. Bullying perpetrators often experience difficulties in emotional regulation, externalising behaviour and impulsivity, and are more likely to associate with delinquent peers and seek social dominance. These are all factors associated with both bullying perpetration and substance use (Arcadepani et al., 2021; Cho, 2018; Luk et al., 2016; Reijntjes et al., 2013; Zucker et al., 2011). This interrelationship does not appear as strong in bullying victims, despite research positing that victims may turn to substance use as a form of maladaptive coping (Doherty et al., 2012; Lee et al., 2022), and experience anxiety and depression at higher rates (Balluerka et al., 2023), our results suggest that the individual and peer-community contexts of perpetrators are more closely interrelated with e-cigarette use than victims. Importantly, bully-victims were also more likely than those who reported no bullying involvement to have ever used an e-cigarette. Bully-victims are a particularly vulnerable group and experience greater rejection by peers, as well as significant adjustment issues (Boccio & Leal, 2022). Future research could investigate the motivations of bullying perpetrators, bullying victims and bully-victims in relation to using e-cigarettes in order to develop and implement targeted preventive strategies.

The association between all three subtypes of bullying involvement highlights the importance of strengthening anti-bullying measures at an individual level through education about respect, individualised support for both victims and perpetrators, and at a school/community level through review of school anti-bullying policy, group programs and education. Additionally, the present findings suggest that prevention programs that target shared underlying risk factors, such as externalising problems (e.g. the Good Behaviour Game) (Kellam et al., 2011) or personality traits (Kelly et al., 2020), may be worth further exploration as strategies to prevent both bullying and substance use, including e-cigarette use.

This study has several limitations. Although data for this study were derived from students attending secondary schools across three Australian states, it is not nationally representative (Showalter & Mullet, 2017). It does however use the largest, most recent sample of Australian adolescents to examine peer correlates of e-cigarette use and aligns with nationally representative data in another research (Kelly et al., 2016). Secondly, the data are cross-sectional, as e-cigarette data was only collected at the 36-month follow-up timepoint, precluding longitudinal examinations. Further research is required to examine the directionality of associations between peer correlates and adolescent e-cigarette use, and to establish causality longitudinally. Additionally, the bullying questionnaire items were not specifically concerned with e-cigarette use or related peer pressure, so future research may wish to examine effects of more specific questions. Our measure of peer e-cigarette use was limited to one item that assessed perceived peer use and we did not collect data to enable analysis of larger friendship group and/or whole school networks to better understand the peer processes occurring and how these relate to e-cigarette use.

Importantly, future research should also seek to examine the associations between actual peer use and adolescent e-cigarette use. One potential mechanism for this is through social network analysis, whereby an index individual nominates peers closest to them, allowing for peer networks to be identified (Valente et al., 2023). Thus, self-reported actual use can be compared with perceived use to identify inconsistencies. These analyses may also disentangle peer selection and peer influence to evaluate their individual impacts, as research suggests they exert different effects at each developmental stage (Mundt et al., 2012). Additional investigation into bullying and underlying factors that may increase vulnerability towards e-cigarette use is also an important area for future research. While this paper focuses on two important peer correlates, future studies should also longitudinally consider other important factors which may be related to e-cigarette use, such as peer pressure, resistance skills, personality traits, school policies, home environment and parenting practices, in order to advance understanding and inform development of preventive interventions. Notwithstanding these limitations, this study significantly adds to the urgent need for research in the space of e-cigarette use among young people.

Conclusions

This cross-sectional analysis revealed important findings regarding the associations between peer factors and adolescent e-cigarette use. Our results suggest that perceived peer use and bullying may be important targets for intervention efforts to prevent e-cigarette use, such as peer-led programs and normative education about peer vaping use. However, further longitudinal research is required to determine causality and to investigate other mechanisms of peer influence.