Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Association between electronic nicotine delivery systems and electronic non-nicotine delivery systems with initiation of tobacco use in individuals aged < 20 years. A systematic review and meta-analysis

  • Sze Lin Yoong ,

    Roles Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing

    syoong@swin.edu.au

    Affiliations School of Health Sciences, Swinburne University of Technology, Hawthorn, Victoria, Australia, School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia

  • Alix Hall,

    Roles Data curation, Formal analysis, Investigation, Methodology, Supervision, Validation, Writing – review & editing

    Affiliations School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia, Priority Research Centre for Heath Behaviour, University of Newcastle, Callaghan, NSW, Australia, Hunter New England Population Health, Wallsend, NSW, Australia

  • Heidi Turon,

    Roles Data curation, Project administration, Writing – review & editing

    Affiliations School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia, Priority Research Centre for Heath Behaviour, University of Newcastle, Callaghan, NSW, Australia, Hunter New England Population Health, Wallsend, NSW, Australia

  • Emily Stockings,

    Roles Data curation, Investigation, Methodology, Writing – review & editing

    Affiliation National Drug and Alcohol Research Centre, University of New South Wales, Randwick, NSW, Australia

  • Alecia Leonard,

    Roles Data curation, Investigation, Writing – review & editing

    Affiliations School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia, Priority Research Centre for Heath Behaviour, University of Newcastle, Callaghan, NSW, Australia

  • Alice Grady,

    Roles Data curation, Methodology, Writing – review & editing

    Affiliations School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia, Priority Research Centre for Heath Behaviour, University of Newcastle, Callaghan, NSW, Australia, Hunter New England Population Health, Wallsend, NSW, Australia

  • Flora Tzelepis,

    Roles Data curation, Methodology, Writing – review & editing

    Affiliations School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia, Priority Research Centre for Heath Behaviour, University of Newcastle, Callaghan, NSW, Australia, Hunter New England Population Health, Wallsend, NSW, Australia

  • John Wiggers,

    Roles Investigation, Methodology, Resources, Writing – review & editing

    Affiliations School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia, Priority Research Centre for Heath Behaviour, University of Newcastle, Callaghan, NSW, Australia, Hunter New England Population Health, Wallsend, NSW, Australia

  • Hebe Gouda,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliation No Tobacco Unit, Department of Health Promotion, World Health Organization, Geneva, Switzerland

  • Ranti Fayokun,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliation No Tobacco Unit, Department of Health Promotion, World Health Organization, Geneva, Switzerland

  • Alison Commar,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliation No Tobacco Unit, Department of Health Promotion, World Health Organization, Geneva, Switzerland

  • Vinayak M. Prasad,

    Roles Conceptualization, Resources, Writing – review & editing

    Affiliation No Tobacco Unit, Department of Health Promotion, World Health Organization, Geneva, Switzerland

  • Luke Wolfenden

    Roles Conceptualization, Funding acquisition, Investigation, Methodology, Resources, Supervision, Writing – original draft

    Affiliations School of Medicine and Public Health, University of Newcastle, Callaghan, NSW, Australia, Hunter Medical Research Institute, New Lambton Heights, NSW, Australia, Priority Research Centre for Heath Behaviour, University of Newcastle, Callaghan, NSW, Australia, Hunter New England Population Health, Wallsend, NSW, Australia

Abstract

Background

This systematic review described the association between electronic nicotine delivery systems and electronic non-nicotine delivery systems (ENDS/ENNDS) use among non-smoking children and adolescents aged <20 years with subsequent tobacco use.

Methods

We searched five electronic databases and the grey literature up to end of September 2020. Prospective longitudinal studies that described the association between ENDS/ENNDS use, and subsequent tobacco use in those aged < 20 years who were non-smokers at baseline were included. The Joanna Briggs Institute Critical Appraisal Checklist was used to assess risk of bias. Data were extracted by two reviewers and pooled using a random-effects meta-analysis. We generated unadjusted and adjusted risk ratios (ARRs) describing associations between ENDS/ENNDS and tobacco use.

Findings

A total of 36 publications met the eligibility criteria, of which 25 were included in the systematic review (23 in the meta-analysis) after exclusion of overlapping studies. Sixteen studies had high to moderate risk of bias. Ever users of ENDS/ENNDS had over three times the risk of ever cigarette use (ARR 3·01 (95% CI: 2·37, 3·82; p<0·001, I2: 82·3%), and current cigarette use had over two times the risk (ARR 2·56 (95% CI: 1·61, 4·07; p<0·001, I2: 77·3%) at follow up. Among current ENDS/ENNDS users, there was a significant association with ever (ARR 2·63 (95% CI: 1·94, 3·57; p<0·001, I2: 21·2%)), but not current cigarette use (ARR 1·88 (95% CI: 0·34, 10·30; p = 0·47, I2: 0%)) at follow up. For other tobacco use, ARR ranged between 1·55 (95% CI 1·07, 2·23) and 8·32 (95% CI: 1·20, 57·04) for waterpipe and pipes, respectively. Additionally, two studies examined the use of ENNDS (non-nicotine devices) and found a pooled adjusted RR of 2·56 (95% CI: 0·47, 13·94, p = 0.035).

Conclusion

There is an urgent need for policies that regulate the availability, accessibility, and marketing of ENDS/ENNDS to children and adolescents. Governments should also consider adopting policies to prevent ENDS/ENNDS uptake and use in children and adolescents, up to and including a ban for this group.

Introduction

Electronic Nicotine Delivery Systems (ENDS) and Electronic Non-Nicotine Delivery Systems (ENNDS) are systems that use devices to heat liquids to create aerosols that are inhaled by users. These are most commonly in the form of an ‘e-cigarette’, but come in other forms (e.g ‘e-pipe’, ‘e-shisha’, ‘e-cigars’). [1] These systems typically contain flavourings, propylene glycol, glycerine and, for ENDS–nicotine. ENDS/ENNDS were first introduced into markets in the 2000s and have been promoted aggressively by manufacturers as “reduced harm products” or “alternatives” to conventional cigarettes. [2] The use of ENDS/ENNDS among children and adolescents however is increasing in some countries, especially among those who had never used tobacco, [3] indicating that such products are not solely used or targeted at adults. [4] In many developed countries, including Canada and the United States (US), ENDS/ENNDS use far surpasses the rates of tobacco use among adolescents in high school. [57]

Of concern is an increasing body of evidence suggesting ENDS/ENNDS use may accrue a range of health risks for different age groups. [8, 9] Constituents of e-liquids, such as propylene glycol and glycerine form toxic aldehydes when heated, of which the long-term effects of exposure remains unknown. [10] ENDS/ENNDS use can also impact on the respiratory system and is associated with adverse effects on the developing brain. [10] A recent position statement by the European Association of Preventive Cardiology reported that e-cigarettes may have negative effects on cardiovascular health for both adolescents and adults. [11] There is a rapidly developing empirical evidence describing a longitudinal association between ENDS/ENNDS and cigarette use among young people.

The first review of three prospective cohort studies in those <20 years in 2016 commissioned by the World Health Organization (WHO), reported that non-smoking e-cigarette users had twice the odds of being a conventional cigarette user at follow-up. [3] Since then, there have been several systematic reviews including at time of conducting our review, the most recent by Khouja and colleagues. [1216] The review by Khouja included 17 studies with individuals aged <30 years, published up to November 2018. The majority of studies were conducted in the US and found a significant adjusted association between ENDS/ENNDS use among non-smokers at baseline and later cigarette use (OR: 2·92 (95% CI 2·30, 3·71). Since this review a number of longitudinal studies have been published from a broader range of countries. [17, 18] An updated systematic review to reflect the contemporary evidence is warranted, as more countries are enacting or planning to enact policy or programs to deter ENDS/ENNDS use in young people globally. [19]

Therefore, this review synthesised findings from studies assessing the longitudinal association between ENDS and/or ENNDS use and later cigarette (primary outcome) and other tobacco product initiation (secondary outcome) among children and adolescents aged < 20 years, who were never smokers at baseline. Additionally, it sought to describe the longitudinal association of ENNDS and flavoured ENDS/ENNDS and subsequent tobacco use.

Methods

Search strategy and selection criteria

This systematic review and meta-analysis is undertaken consistent with guidance by Joanna Briggs Institute (JBI) [20] and reported in accordance with Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines. [21] It was prospectively registered in the PROSPERO database (CRD42020199485).

Studies were included if they were prospective longitudinal studies assessing the relationship between ENDS and/or ENNDS use at baseline and initiation of cigarette and other tobacco products at follow-up, among children and adolescents aged less than 20 years who were non-tobacco users at baseline. Case control, cross-sectional and retrospective studies were excluded to capture only studies with the lowest risk of bias for assessing association. [22] There were no restrictions on the year of publication, length or location of the study, peer review status, or language of publication.

We conducted an electronic search of the following databases: Medline, Web of Science, CINAHL, Embase and Wiley Cochrane Library using search terms for the following ‘electronic nicotine delivery systems (ENDS) electronic non-nicotine delivery systems (ENNDS), AND ‘prospective studies’ AND ‘children and adolescents’ (see S1 Appendix for search strategy) on the September 2020. The reference lists of all relevant reviews and eligible papers were also screened. We undertook a grey literature search based on guidance from previous reviews, [23] which included searching OpenGrey (a grey literature database) and Google and Google scholar to identify relevant studies using the following terms ‘electronic cigarette’, ‘e-cigarette’, ‘electronic nicotine delivery systems (ENDS), ‘electronic non-nicotine delivery systems (ENNDS)’, ‘e-hookah’ and ‘juul’. The first 500 titles of each search were sorted by relevance were assessed by one reviewer in October 2020 (SLY).

An information specialist used EndNote version X9.2 software (Thomson Reuters, PA, U.S.) to filter duplicate studies. Title and abstract screening were undertaken using Covidence software [24] by two reviewers, and discrepancies resolved by consensus (SLY, AH). Full text was obtained and assessed for eligibility in accordance with the criteria described above by two reviewers (AL, ES). All conflicts were resolved by discussion and included a third reviewer (SLY), where necessary.

All data were extracted by a first reviewer (AG, FT, SLY or HT) and double checked by a second reviewer not involved in the original extraction of the study (SLY, HT or AL). Discrepancies were highlighted and checked by a third reviewer (AH). The following information was extracted: participant characteristics, study design, country, data collection modality and measure, sampling frame and recruitment, proportion and number of ENDS/ENNDS users separately where reported, tobacco users as well as non-users at each time point, relevant measures of association between ENDS and ENNDS users and future cigarette and other tobacco product initiation (e.g. risk ratios, odds ratios), estimates of variance and covariates adjusted for, follow-up time points, type of tobacco products assessed and flavours.

The JBI Critical Appraisal Checklist for prevalence studies was used to assess the quality of each study by two reviewers (AG, HT, FT, AL). [25] Discrepancies were checked by a third reviewer (SLY). The tool consists of nine items examining the following: sample representativeness, sampling methods, adequacy of sample size, participant and setting descriptions, coverage of sample, objectivity and reliability of measures, appropriateness of statistical analysis, confounding factors identified and accounted for, and objective classification of subpopulations (Yes; No; Unclear; and N/A). An additional tenth criterion relating to participant retention was added to allow for assessment of attrition bias. Two reviewers also assessed four supplementary criteria detailed in the Bradford-Hill criteria relevant to establishing causality between exposure and outcome. [26] (see S1 Table)

Data analysis

All analyses were undertaken using Stata version 14.2. [27] Effect estimates (extracted or converted to Risk Ratios (RRs)) of the association between ENDS/ENNDS use at baseline and initiation of cigarette or other tobacco use at follow up were combined using the DerSimonian and Laird random effects method. [28]

The primary outcome variable was ever and current cigarette smoking. For ever cigarette smoking, this included lifetime ever use. For current cigarette use, this included use in the past 30 days, frequent and daily cigarette use. The exposure variable was ever and current ENDS/ENNDS use. For ever use of ENDS and/or ENNDS, this was defined as lifetime ever use. For current use of ENDS and/or ENNDS, this included use in the past 30 days, recent use and self-defined current use.

A p-value of 0·05 was used to determine a statistically significant association. Where it was not appropriate to undertake a meta-analysis (due to heterogeneity or small number of studies), study findings were narratively described.

For studies that did not report the unadjusted RRs, these were calculated using the data extracted from the original study or converted from an odds ratio (OR). In instances where studies reported an adjusted odds ratio (AOR) rather than an adjusted RR, these were also converted to an RR. The formula from the Cochrane Handbook (Section 15.4.4.4) [29] was used to convert ORs to RRs. The ACR was calculated on a per study basis as the risk of later smoking among controls, whereby the control was defined as no ENDS/ENNDS use at baseline. In instances where a study did not provide sufficient data to calculate a study-specific ACR, the average ACR from other studies was used.

Where multiple follow-up points were available, the furthest time from baseline was included. Additionally, when a study reported a slight variation for the same outcome, using overlapping datasets, the outcome most closely aligned with the aims was chosen. Where multiple effect estimates exist controlling for different confounders, we included the ones that controlled for demographics that had evidence of association with tobacco uptake (sex, age, socioeconomic status and susceptibility to tobacco use), where available.

A number of planned subgroup analyses were undertaken [12] including: country (grouped into US, United Kingdom (UK) and other), study quality (<7 and 7 or more on the Joanna Briggs scale), and Bradford-Hill’s causal inference score (> = 3 and <3). Additionally, we also undertook a subgroup analysis by length of follow up (≤ 12 months and >12 months) and publication year (≤ 2018 and >2018). We planned to undertake sensitivity analysis by funding source (e.g. industry/non-industry), however no industry funded studies were included in the meta-analysis.

Heterogeneity of study effect estimates were evaluated using the I2 statistic. A funnel plot and the Duval and Tweedie trim-and-fill method was used to examine possible publication bias and provide an estimate of the bias-adjusted pooled estimates. [3032]

Results

Of the 1,668 studies included after removal of duplicates, 452 articles underwent full text screen, of which 35 were included (see Fig 1). Of those, 10 were excluded from the final analysis as there was overlap of data with other studies included in this review. A total of 25 studies were included in the review, of which 23 were included in any meta-analyses (Fig 1). [17, 18, 3353]

thumbnail
Fig 1. PRISMA flowchart outlining study inclusion and exclusion.

https://doi.org/10.1371/journal.pone.0256044.g001

The studies were conducted in the US (n = 13), Germany (n = 3), UK (n = 2), Scotland (n = 1), Canada (n = 1), Finland (n = 1), Mexico (n = 1), Taiwan (n = 1), Netherlands (n = 1) and Romania (n = 1) with data collection occurring from 2013–2016 at baseline (see Table 1). Sample sizes ranged from 164 to 17,318 and participants were aged between 11 to 26 years (as studies were eligible for inclusion if they had a mean age of <20). The follow-up period was between six to 24 months, and all studies used self-reported measures to assess cigarette (and/or tobacco) use at follow up. Overall, 21 studies assessed cigarette smoking only as an outcome, [17, 18, 3441, 4347, 4954] three assessed cigarettes and other tobacco [33, 42, 48] and one assessed other tobacco only. [55] All studies referred to ENDS/ENNDS as e-cigarettes. Two studies specifically assessed the use of non-nicotine e-cigarettes [40, 51] while one study compared flavoured and non-flavoured e-cigarettes. [54]

thumbnail
Table 1. Characteristics of study included in the review.

https://doi.org/10.1371/journal.pone.0256044.t001

Sixteen studies had high to moderate risk of bias (defined as meeting less than 7 of the 11 risk of bias criteria), [33, 34, 3641, 44, 4651, 53] while nine had a low risk (defined as meeting 7 or more criteria). [17, 18, 35, 42, 43, 45, 52, 54, 55] (See Fig 2). Key methodological issues identified in the studies were the sampling frame was not appropriately representative of the target population (n = 6 studies rated as high risk [37, 38, 42, 44, 50, 51] and 4 studies rated as unclear [46, 48, 49, 53]), lack of use of valid methods to identify the condition (n = 23 used self-reported measures without established psychometrics and were rated unclear [17, 18, 3336, 3841, 4355]), and lack of information regarding whether the response rate was adequate or appropriately managed (n = 8 studies rated as high risk [18, 36, 39, 40, 43, 44, 50, 55] and 22 studies rated as unclear [17, 3335, 37, 38, 41, 4649, 51, 53, 54]). All 25 studies were rated as low risk on the criteria for appropriate statistical analysis and 18 were also rated as low risk for adequate sample size.[17, 18, 3437, 3944, 47, 5052, 54, 55] For the Bradford-Hill criteria, 13 studies met ≥ three of the four criteria. [3335, 37, 38, 41, 42, 45, 47, 49, 51, 54, 55]. All studies rated low risk for temporality, and most were also rated low risk for specificity (n = 24). Only three studies were rated low risk for the dose responsivity criteria, [34, 38, 47]. The majority of studies met the criteria for specificity and all met the criteria for temporality. All studies except one included in this review reported a positive association, with 13 reporting an adjusted odds ratio of > 4.

Seventeen studies assessed the association between ever ENDS/ENNDS use and subsequent ever cigarette use. [17, 18, 33, 3538, 4144, 46, 48, 49, 51, 53] The adjusted RRs ranged from 1·39 (95% CI: 1·01, 1·91) to 12·86 (95% CI: 3·59, 46·05); with a pooled RR of 3·01 (95% CI: 2·37, 3·82, p<0·001; I2 = 82·3%, p <0·001) (see Fig 3). Most studies adjusted for covariates including sex and age or grade (n = 15), with the majority (n = 14) also adjusting for additional variables including susceptibility to smoking, influence by friends and family, psychological constructs and status, and exposure to advertising.

thumbnail
Fig 3. Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and subsequent ever cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.g003

Six studies assessed the association between ever ENDS/ENNDS use at baseline and subsequent current cigarette use at follow-up. [34, 35, 40, 44, 46, 50] The adjusted RRs ranged from 1·40 (95% CI: 1·22, 1·60) to 3·53 (95% CI: 1·98, 6·30); with a pooled RR of 2·56 (95% CI: 1·61, 4·07, p<0·001; I2 = 77·3%, p = 0·001) (see Fig 4).

thumbnail
Fig 4. Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and subsequent current cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.g004

Four studies assessed the association between current ENDS/ENNDS use at baseline and subsequent ever cigarette use at follow-up. [39, 45, 50, 52] The adjusted RRs ranged from 2·21 (95% CI: 1·74, 2·80) and 4·78 (95% CI: 1·91, 11·96) with a pooled RR of 2·63 (95% CI: 1·94, 3·57, p<0·001; I2 = 21·2%, p >0·05) (see Fig 5).

thumbnail
Fig 5. Forest plot of adjusted risk ratios assessing the association between current e-cigarette use at baseline and subsequent ever cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.g005

Two studies assessed association between current ENDS/ENNDS use at baseline and subsequent current cigarette use at follow-up. [47, 50] The adjusted RRs were 1·16 (95% CI: 0·11, 12·36) and 3·15 (95% CI: 0·27, 36·48), with a pooled RR of 1·88 (95% CI: 0·34, 10 ·30, p = 0·467; I2 = 0%, p >0·05) (see Fig 6).

thumbnail
Fig 6. Forest plot of adjusted risk ratios assessing the association between current e-cigarette use at baseline and subsequent current cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.g006

S3 Table lists the four studies that assessed the association between ENDS/ENNDS use at baseline and subsequent use of other tobacco products including hookah, cigar, pipe, and other tobacco products at follow-up, where significant associations were reported.

Only two studies [40, 51] assessed the association between ENNDS use at baseline and subsequent cigarette use (current or ever) at follow-up. The pooled adjusted RR of 2.56 (95% CI: 0·47, 13·94, I2 = 77·5%, p = 0.277;) (see Fig 7). No study reported on association between ENNDS use with subsequent use of other tobacco products.

thumbnail
Fig 7. Forest plot of adjusted risk ratios assessing the association between ever ENNDS use at baseline and subsequent current or ever cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.g007

The unadjusted estimates are available as supplementary materials (S2 and S3 Tables, S1S3 Figs).

One study [54] reported no difference in uptake of ENDS/ENNDS use at follow up between flavoured vs unflavoured e-cigarette use at baseline (RR: 0·24 (95% CI 0·05, 1·0) when controlling for sex, age, state, school type, migration background, parent’s qualifications, socioeconomic status (SES), multiple personality traits, and consumption of five substances.

The adjusted RRs were similar by geographic location, year of publication, and length of follow up (see S4A–S4C Fig). There were some differences in effect sizes by study quality, with higher quality studies reporting lower adjusted RRs (risk of bias ≥7 (higher quality): 2·16 (95% CI: 1·47, 3·16, p<0·001; I2 = 85·0%, p <0·001) compared to lower quality studies (risk of bias scores <7: 3·57 (95% CI: 2·69, 4·73, p<0·001; I2 = 76·9%, p < ·001)) see S4D Fig. Studies that scored > = 3 on the Bradford-Hill criteria for causal inference had higher adjusted RRs of 4·47 (95% CI: 3·28, 6·09, p<0·001; I2 = 65·0%, p = 0·006) relative to studies that scored <3: 2·21 (95% CI: 1·80, 2·70, p<0·001; I2 = 64·1%, p = 0·004) (see S4E Fig).

The adjusted RRs for baseline ever ENDS/ENNDS use and current cigarette use at follow-up were similar by geographic location, year of publication, length of follow up, study quality, and score for Bradford-Hill causal inference (S5A–S5E Fig).

We did not undertake subgroup analysis examining other associations due to the small number of studies included in the main meta-analyses (four or less).

For ever ENDS/ENNDS use at baseline and ever cigarette use at follow-up the adjusted results, three studies were estimated as missing due to funnel plot asymmetry. Results from the trim-and-fill analysis found that the bias-adjusted pooled RR was 2·75 (95% CI: 2·16, 3·49), which was only slightly lower than the adjusted pooled RR from the primary analysis (see Fig 8).

thumbnail
Fig 8. Funnel plot illustrating results from trim-and-fill analysis of adjusted log RRs for outcome ever e-cigarette use at baseline and ever cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.g008

For ever ENDS/ENNDS use at baseline and current cigarette use at follow-up, the adjusted results two studies were estimated as being missing due to funnel plot asymmetry. Results from the trim-and-filled analysis found that the bias-adjusted pooled RR was 2·21 (95% CI: 1·55, 3·17), which was slightly lower than the original estimate (see Fig 9).

thumbnail
Fig 9. Funnel plot illustrating results from trim-and-fill analysis of adjusted log RRs for outcome ever e-cigarette use at baseline and current cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.g009

Discussion

This review supports evidence of a longitudinal association between ENDS/ENNDS use at baseline and subsequent tobacco use in those aged <20 years. Studies included in the meta-analysis found a significant positive adjusted association between ever ENDS/ENNDS and current cigarette use (2·56 (95% CI: 1·61, 4·07) at follow-up among children and adolescents aged <20 years. A positive association was also found between current e-cigarette use and current cigarette use at follow-up (RR: 1·88 (95% CI: 0·34, 10·30)), and ENNDS use at baseline and later cigarette use (RR: 2·56 (95% CI: 0·47, 13·94)). Despite the relatively large effect size, evidence of these associations was not statistically significant potentially due to the small number of studies included, and thus require further exploration in prospective studies.

Our findings are similar, albeit slightly weaker, to those reported by Khouja et al, [12] where a significant association between e-cigarette use among non-smokers and later tobacco smoking was found. The similarity may, in part, be due to the inclusion of many of the same studies. However, our review included more recently published studies, included a broader representation of study locations outside of the US (13/25 studies), focused entirely on children and adolescents (whereas the review by Khouja et al. included those up until the age of 30) [12], and excluded case-control studies that are at risk of increased bias. Consequently, this study has improved both the robustness, precision of aggregate analysis and international applicability of findings from prior reviews.

In our exploratory subgroup analysis, we found that higher quality studies had small estimates than lower quality studies. The impact of different methodological biases have been explored in a recent review examining the association between e-cigarette use and initiation of conventional cigarette use. [15] This review described potential bias relating to attrition, where studies that reported on findings from complete case analyses found larger effect sizes than when imputed data was included. Additionally, studies that adjusted for a more comprehensive list of known confounders also reported smaller estimates, compared to those that adjusted for fewer confounders. [15, 56] Future studies need to better consider and address such methodological differences to provide better estimates of the association between e-cigarette use and conventional cigarette uptake. All but one of the studies included in this review reported a positive association (RR>1) between ENDS/ENNDS use and future cigarette use among children and adolescents. The only industry-funded study that met the eligibility criteria for this review was excluded from the meta-analysis due to overlap of data with other studies. The authors of this study undertook various sensitivity analysis adjusting for multiple confounders. [56, 57] Whilst the authors concluded that adjustment for various confounders including propensity to smoke reduced the strength of the association, all adjusted odds ratios were larger than one, consistent with findings from non-industry sponsored studies.

Our review found evidence of a consistent positive association between ENDS/ENNDS use and cigarette smoking across a large number of studies internationally. This provides strong evidence to support the causal relationship between ever ENDS/ENNDS and ever smoking for this age group. These findings are of concern as other cross-sectional studies have reported that children and adolescents who use ENDS and/or ENNDS have different psychological profiles to current smokers, and would have otherwise have been at low risk of smoking. [5861] As such, there is an urgent need for governments internationally to take action to regulate the availability and marketing of ENDS/ENNDS products to children and adolescents.

Further, the US Surgeon General’s Report concluded that ENDS/ENNDS were unsafe for use among children and adolescents due to a range of health-related adverse effects. [62] The use of ENDS/ENNDS may also contribute to increased burden of tobacco-related harms on individuals and communities. [63] In part due to such an association, modelling weighing the potential health benefits (e.g. cessation among established smokers) and harms associated with e-cigarettes found, overall, that ENDS/ENNDS use would yield a net harm and lead to 1,510,000 years life lost in the US. [63] This modelling is based on results from a single clinical trial of ENDS/ENNDS provided as part of medically-supervised cessation benefits Such findings are consistent with later reviews of randomised trials assessing the use of ENDS/ENNDS, [64, 65] however presents an overestimation of benefit when used as consumer products in the general population. As presented in a synthesis of observational studies, there are no apparent population-level increase in cessation when using e-cigarettes as a consumer product [65].

Given such considerations, a report by WHO provides a range of policy options including a ban on their sale; product taxation; and preventing the use of ENDS/ENNDS indoors and in areas to prevent use in in children and adolescents but also uptake in adults more broadly [66, 67]. These are supported by recommendations and policy statements nationally and internationally.[67, 68] Given the susceptibility of children and adolescents to marketing and the appeal of flavouring, governments should restrict all forms of promotion and marketing to children and adolescents and ban all characterising flavours. [68, 69] A number of recently published studies have also reported promising findings regarding the impact of local retail regulations, [70] and the prohibition of the sale of flavoured products on ENDS/ENNDS use in youth, [71] however, rigorous evaluation of the impacts of comprehensive policy approaches is warranted. Early findings from two studies suggest mixed findings between ENNDS and cigarette smoking. Whilst, still inconclusive, precautionary principles should be in place when considering the regulation for all forms of e-cigarettes, including those that do and do not contain nicotine.

There were few studies that measured association between current ENDS/ENNDS and current cigarette use. Further studies are needed to establish whether current ENDS/ ENNDS result in current cigarette given this Similarly, there were few studies assessing the impact of non-nicotine and flavoured tobacco products, and as such any conclusions need to be interpreted in light of this. Most studies were conducted in high-income countries. Consequently, the study results may be limited in their generalisability. The data from included studies may also be subject to social desirability and other reporting biases due to the self-report nature of the data collection methods. There was high heterogeneity in the meta-analysis, unexplained by the subgroup analysis, indicating that the reasons for the variation remains unknown. The trim and fill funnel plots suggest there may be some publication bias, but the bias-adjusted estimates were similar to those calculated from the main analysis. Finally, despite efforts to select outcomes that controlled for pre-specified confounders, restricting outcomes that controlled for these confounders only was not always possible. Consequently, there were differences between studies in terms of the characteristics that were controlled for, which may contribute to the high level of heterogeneity.

Nonetheless, the findings provide consistent evidence from observational studies of an association between ENDS/ENNDS use among non-smoking children and adolescents, and subsequent tobacco use, in particular cigarettes. Government regulation and implementation to prevent use of ENDS/ENNDS among youth however varies considerably globally. [69, 72] The experience of global efforts to combat the use of conventional cigarettes and other tobacco products suggests that such efforts are inadequate to sufficiently avert the projected harms, if the current trajectory continues. There is a need for countries internationally to prioritise the adoption and implementation of comprehensive measures as outlined in the WHO Framework Convention on Tobacco Control to prevent uptake of ENDS/ENNDS and regulates availability in children and adolescents, up to imposing a ban, to prevent uptake of ENDS/ENNDS for this group.

Supporting information

S1 Fig. Forest plot of unadjusted risk ratios assessing the association between ever e-cigarette use at baseline and ever cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.s002

(DOCX)

S2 Fig. Forest plot of unadjusted risk ratios assessing the association between ever e-cigarette use at baseline and current cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.s003

(DOCX)

S3 Fig. Forest plot of unadjusted risk ratios assessing the association between current e-cigarette use at baseline and ever cigarette use at follow-up.

https://doi.org/10.1371/journal.pone.0256044.s004

(DOCX)

S4 Fig.

(A) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and ever tobacco use at follow-up by country. (B) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and ever tobacco use at follow-up by year of publication. (C) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and ever cigarette use at follow-up by length of follow-up. (D) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and ever cigarette use at follow-up by overall risk of bias score. (E) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and ever cigarette use at follow-up by Bradford Hill’s criteria for causal inference.

https://doi.org/10.1371/journal.pone.0256044.s005

(DOCX)

S5 Fig.

(A) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and current tobacco use at follow-up by country. (B) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and current tobacco use at follow-up by year of publication. (C) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and current tobacco use at follow-up by length of follow-up. (D) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and current tobacco use at follow-up by overall risk of bias score. (E) Forest plot of adjusted risk ratios assessing the association between ever e-cigarette use at baseline and current tobacco use at follow-up by risk of bias score for causal inference.

https://doi.org/10.1371/journal.pone.0256044.s006

(DOCX)

S1 Table. Additional Bradford-Hill causal inference criteria.

https://doi.org/10.1371/journal.pone.0256044.s007

(DOCX)

S2 Table. Unadjusted and adjusted risk ratios for association between ENDS/ENNDS and cigarette use.

https://doi.org/10.1371/journal.pone.0256044.s008

(DOCX)

S3 Table. Unadjusted and adjusted risk ratios for association between ENDS/ENNDS and other tobacco products.

https://doi.org/10.1371/journal.pone.0256044.s009

(DOCX)

References

  1. 1. U.S. Food & Drug Administration. Vaporizers, E-Cigarettes, and other Electronic Nicotine Delivery Systems (ENDS) 2020. Available from: https://www.fda.gov/tobacco-products/products-ingredients-components/vaporizers-e-cigarettes-and-other-electronic-nicotine-delivery-systems-ends.
  2. 2. Ramamurthi D, Gall PA, Ayoub N, Jackler RK. Leading-Brand Advertisement of Quitting Smoking Benefits for E-Cigarettes. Am J Public Health. 2016;106(11):2057–63. Epub 2016/09/15. pmid:27631743.
  3. 3. Yoong SL, Tzelepis F, Wiggers J, Oldmeadow C, Chai LK, Paul C, et al. Prevalence of smoking-proxy electronic inhaling system (SEIS) use and its association with tobacco initiation in youths: a systematic review. World Health Organization, 2016.
  4. 4. Yoong SL, Stockings E, Chai LK, Tzelepis F, Wiggers J, Oldmeadow C, et al. Prevalence of electronic nicotine delivery systems (ENDS) use among youth globally: a systematic review and meta-analysis of country level data. Australian and New Zealand Journal of Public Health. 2018;42(3):303–8. pmid:29528527
  5. 5. Baker HM, Kowitt SD, Meernik C, Heck C, Martin J, Goldstein AO, et al. Youth source of acquisition for E-Cigarettes. Preventive Medicine Reports. 2019;16:101011. pmid:31890469
  6. 6. Government of Canada. Canadian Tobacco, Alcohol and Drugs Survey (CTADS): summary of results for 2017 2019 [5th November 2020]. Available from: https://www.canada.ca/en/health-canada/services/canadian-tobacco-alcohol-drugs-survey/2017-summary.html.
  7. 7. Gentzke AS, Wang TW, Jamal A, Park-Lee E, Ren C, Cullen KA, et al. Tobacco Product Use Among Middle and High School Students—United States, 2020. Morbidity and Mortality Weekly Report. 2020;69(50):1881. pmid:33332300
  8. 8. Gotts JE, Jordt S-E, McConnell R, Tarran R. What are the respiratory effects of e-cigarettes? BMJ. 2019;366:l5275. pmid:31570493
  9. 9. Sapru S, Vardhan M, Li Q, Guo Y, Li X, Saxena D. E-cigarettes use in the United States: reasons for use, perceptions, and effects on health. BMC Public Health. 2020;20(1):1518. pmid:33032554
  10. 10. Fadus MC, Smith TT, Squeglia LM. The rise of e-cigarettes, pod mod devices, and JUUL among youth: Factors influencing use, health implications, and downstream effects. Drug and Alcohol Dependence. 2019;201:85–93. Epub 2019/05/23. pmid:31200279.
  11. 11. Kavousi M, Pisinger C, Barthelemy J-C, Smedt DD, Koskinas K, Marques-Vidal P, et al. Electronic cigarettes and health with special focus on cardiovascular effects: position paper of the European Association of Preventive Cardiology (EAPC). European Journal of Preventive Cardiology. 2020:2047487320941993. pmid:32726563.
  12. 12. Khouja JN, Suddell SF, Peters SE, Taylor AE, Munafo MR. Is e-cigarette use in non-smoking young adults associated with later smoking? A systematic review and meta-analysis. Tobacco Control. 2020;30:8–15. pmid:32156694
  13. 13. Glasser A, Abudayyeh H, Cantrell J, Niaura R. Patterns of E-Cigarette Use Among Youth and Young Adults: Review of the Impact of E-Cigarettes on Cigarette Smoking. Nicotine & Tobacco Research. 2018;21(10):1320–30. pmid:29788314
  14. 14. Aladeokin A, Haighton C. Is adolescent e-cigarette use associated with smoking in the United Kingdom?: A systematic review with meta-analysis. Tobacco Prevention & Cessation. 2019;5:15. Epub 2020/05/16. pmid:32411879; PubMed Central PMCID: PMC7205081.
  15. 15. Chan GCK, Stjepanović D, Lim C, Sun T, Shanmuga Anandan A, Connor JP, et al. Gateway or common liability? A systematic review and meta-analysis of studies of adolescent e-cigarette use and future smoking initiation. Addiction. 116(4):743–56. pmid:32888234
  16. 16. Baenziger O, Ford L, Yazidjoglou A, Joshy G, Banks E. E-cigarette use and combustible tobacco cigarette smoking uptake among non-smokers, including relapse in former smokers: umbrella review, systematic review and meta-analysis. MedRxiv [Preprint]. 2020:Forthcoming. pmid:32511587
  17. 17. Chien YN, Gao W, Sanna M, Chen PL, Chen YH, Glantz S, et al. Electronic Cigarette Use and Smoking Initiation in Taiwan: Evidence from the First Prospective Study in Asia. International Journal of Environmental Research and Public Health. 2019;16(7). Epub 2019/04/03. pmid:30935027; PubMed Central PMCID: PMC6480595.
  18. 18. Hansen J, Janssen J, Morgenstern M, Hanewinkel R. E-Cigarette Use and Later Use of Conventional Cigarettes—Results of a Prospective Observational Study over 2 Years. Pneumologie (Stuttgart, Germany). 2020;74(1):39–45. Epub 2019/11/23. pmid:31756736.
  19. 19. Bhalerao A, Sivandzade F, Archie SR, Cucullo L. Public Health Policies on E-Cigarettes. Current Cardiology Reports. 2019;21(10):111. pmid:31463564
  20. 20. Joanna Briggs Institute. 5.3.1 Indications for systematic reviews of prevalence and incidence: Joanna Briggs Institute; 2020 [cited 2020]. Available from: https://wiki.joannabriggs.org/display/MANUAL/5.3.1+Indications+for+systematic+reviews+of+prevalence+and+incidence.
  21. 21. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA. 2000;283(15):2008–12. Epub 2000/05/02. pmid:10789670.
  22. 22. Murad MH, Asi N, Alsawas M, Alahdab F. New evidence pyramid. Evid Based Med. 2016;21(4):125–7. Epub 2016/06/23. pmid:27339128.
  23. 23. Godin K, Stapleton J, Kirkpatrick SI, Hanning RM, Leatherdale ST. Applying systematic review search methods to the grey literature: a case study examining guidelines for school-based breakfast programs in Canada. Systematic Reviews. 2015;4(1):138. pmid:26494010
  24. 24. Covidence systematic review software. Melbourne, Australia: Veritas Health Innovation.
  25. 25. Aromataris E, Munn Z, (Editors). Joanna Briggs Institute Reviewer’s Manual: The Joanna Briggs Institute; 2017. Available from: https://reviewersmanual.joannabriggs.org/
  26. 26. Fedak K, Bernal A, Capshaw Z, Gross S. Applying the Bradford Hill criteria in the 21st century: how data integration has changed causal inference in molecular epidemiology. Emerg Themes Epidemiol. 2015;12(14).
  27. 27. StataCorp. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC; 2019.
  28. 28. Tufanaru C, Munn Z, Stephenson M, Aromataris E. Fixed or random effects meta-analysis? Common methodological issues in systematic reviews of effectiveness. JBI Evidence Implementation. 2015;13(3):196–207. pmid:26355603.
  29. 29. Schünemann HJ, Vist GE, Higgins JP, Santesso N, Deeks JJ, Glasziou P, et al. Chapter 15: Interpreting results and drawing conclusions. 2020. In: Cochrane Handbook for Systematic Reviews of Interventions version 61 (updated September 2020) [Internet]. Cochrane. Available from: www.training.cochrane.org/handbook.
  30. 30. Lin L, Chu H. Quantifying publication bias in meta-analysis. Biometrics. 2018;74(3):785–94. pmid:29141096
  31. 31. Mavridis D, Salanti G. How to assess publication bias: funnel plot, trim-and-fill method and selection models. Evidence Based Mental Health. 2014;17(1):30-. pmid:24477535
  32. 32. Borenstein M, Hedges LV, Higgins JPT, Rothstein HR. Introduction to Meta‐Analysis. New York: John Wiley & Sons; 2009.
  33. 33. Barrington-Trimis JL, Urman R, Berhane K, Unger JB, Cruz TB, Pentz MA, et al. E-Cigarettes and Future Cigarette Use. Pediatrics. 2016;138(1):e20160379. pmid:27296866
  34. 34. Barrington-Trimis JL, Kong G, Leventhal AM, Liu F, Mayer M, Cruz TB, et al. E-cigarette Use and Subsequent Smoking Frequency Among Adolescents. Pediatrics. 2018;142(6):e20180486. pmid:30397165
  35. 35. Berry KM, Fetterman JL, Benjamin EJ, Bhatnagar A, Barrington-Trimis JL, Leventhal AM, et al. Association of Electronic Cigarette Use With Subsequent Initiation of Tobacco Cigarettes in US Youths. JAMA Network Open. 2019;2(2):e187794–e. pmid:30707232
  36. 36. Best C, Haseen F, Currie D, Ozakinci G, MacKintosh AM, Stead M, et al. Relationship between trying an electronic cigarette and subsequent cigarette experimentation in Scottish adolescents: a cohort study. Tobacco Control. 2018;27(4):373–8. pmid:28735273
  37. 37. Conner M, Grogan S, Simms-Ellis R, Flett K, Sykes-Muskett B, Cowap L, et al. Do electronic cigarettes increase cigarette smoking in UK adolescents? Evidence from a 12-month prospective study. Tobacco Control. 2018;27(4):365–72. pmid:28818839
  38. 38. East K, Hitchman SC, Bakolis I, Williams S, Cheeseman H, Arnott D, et al. The Association Between Smoking and Electronic Cigarette Use in a Cohort of Young People. The Journal of Adolescent Health. 2018;62(5):539–47. Epub 2018/03/04. pmid:29499983; PubMed Central PMCID: PMC5938086.
  39. 39. Hammond D, Reid JL, Cole AG, Leatherdale ST. Electronic cigarette use and smoking initiation among youth: a longitudinal cohort study. CMAJ: Canadian Medical Association Journal. 2017;189(43):E1328–e36. Epub 2017/11/01. pmid:29084759; PubMed Central PMCID: PMC5662449.
  40. 40. Kinnunen JM, Ollila H, Minkkinen J, Lindfors PL, Timberlake DS, Rimpelä AH. Nicotine matters in predicting subsequent smoking after e-cigarette experimentation: A longitudinal study among Finnish adolescents. Drug and Alcohol Dependence. 2019;201:182–7. Epub 2019/06/27. pmid:31238240.
  41. 41. Kong G, Mayer ME, Barrington-Trimis JL, McConnell R, Leventhal AM, Krishnan-Sarin S. Longitudinal associations between use and co-use of cigars and cigarettes: A pooled analysis of three adolescent cohorts. Drug and Alcohol Dependence. 2019;201:45–8. Epub 2019/06/11. pmid:31181436; PubMed Central PMCID: PMC6612437.
  42. 42. Leventhal AM, Strong DR, Kirkpatrick MG, Unger JB, Sussman S, Riggs NR, et al. Association of Electronic Cigarette Use With Initiation of Combustible Tobacco Product Smoking in Early Adolescence. JAMA. 2015;314(7):700–7. pmid:26284721
  43. 43. Loukas A, Marti CN, Cooper M, Pasch KE, Perry CL. Exclusive e-cigarette use predicts cigarette initiation among college students. Addictive Behaviors. 2018;76:343–7. Epub 2017/09/12. pmid:28892771; PubMed Central PMCID: PMC5614895.
  44. 44. Lozano P, Barrientos-Gutierrez I, Arillo-Santillan E, Morello P, Mejia R, Sargent JD, et al. A longitudinal study of electronic cigarette use and onset of conventional cigarette smoking and marijuana use among Mexican adolescents. Drug and Alcohol Dependence. 2017;180:427–30. Epub 2017/10/11. pmid:28988005; PubMed Central PMCID: PMC5771440.
  45. 45. Miech R, Patrick ME, O’Malley PM, Johnston LD. E-cigarette use as a predictor of cigarette smoking: results from a 1-year follow-up of a national sample of 12th grade students. Tobacco Control. 2017;26(e2):e106–e11. pmid:28167683
  46. 46. Morgenstern M, Nies A, Goecke M, Hanewinkel R. E-Cigarettes and the Use of Conventional Cigarettes. Dtsch Arztebl Int. 2018;115(14):243–8. pmid:29716689.
  47. 47. Osibogun O, Bursac Z, Maziak W. E-Cigarette Use and Regular Cigarette Smoking Among Youth: Population Assessment of Tobacco and Health Study (2013–2016). American Journal of Preventive Medicine. 2020;58(5):657–65. Epub 2020/03/10. pmid:32147371; PubMed Central PMCID: PMC7174087.
  48. 48. Pénzes M, Foley KL, Nădășan V, Paulik E, Ábrám Z, Urbán R. Bidirectional associations of e-cigarette, conventional cigarette and waterpipe experimentation among adolescents: A cross-lagged model. Addictive Behaviors. 2018;80:59–64. Epub 2018/01/23. pmid:29355818; PubMed Central PMCID: PMC5807159.
  49. 49. Primack BA, Soneji S, Stoolmiller M, Fine MJ, Sargent JD. Progression to Traditional Cigarette Smoking After Electronic Cigarette Use Among US Adolescents and Young Adults. JAMA pediatrics. 2015;169(11):1018–23. Epub 2015/09/09. pmid:26348249; PubMed Central PMCID: PMC4800740.
  50. 50. Spindle TR, Hiler MM, Cooke ME, Eissenberg T, Kendler KS, Dick DM. Electronic cigarette use and uptake of cigarette smoking: A longitudinal examination of U.S. college students. Addictive Behaviors 2017;67:66–72. Epub 2016/12/31. pmid:28038364; PubMed Central PMCID: PMC5250543.
  51. 51. Treur JL, Rozema AD, Mathijssen JJP, van Oers H, Vink JM. E-cigarette and waterpipe use in two adolescent cohorts: cross-sectional and longitudinal associations with conventional cigarette smoking. European Journal of Epidemiology. 2018;33(3):323–34. Epub 2017/12/21. pmid:29260431; PubMed Central PMCID: PMC5889768.
  52. 52. Watkins SL, Glantz SA, Chaffee BW. Association of Noncigarette Tobacco Product Use With Future Cigarette Smoking Among Youth in the Population Assessment of Tobacco and Health (PATH) Study, 2013–2015. JAMA Pediatrics. 2018;172(2):181–7. Epub 2018/01/04. pmid:29297010; PubMed Central PMCID: PMC5801043.
  53. 53. Wills TA, Knight R, Sargent JD, Gibbons FX, Pagano I, Williams RJ. Longitudinal study of e-cigarette use and onset of cigarette smoking among high school students in Hawaii. Tobacco Control. 2017;26(1):34–9. pmid:26811353
  54. 54. Friedman AS, Xu S. Associations of Flavored e-Cigarette Uptake With Subsequent Smoking Initiation and Cessation. JAMA Network Open. 2020;3(6):e203826–e. pmid:32501490
  55. 55. Hansen J, Hanewinkel R, Morgenstern M. Electronic cigarette advertising and teen smoking initiation. Addictive Behaviors. 2020;103:106243. Epub 2019/12/20. pmid:31855726.
  56. 56. Lee PN, Coombs KJ, Afolalu EF. Considerations related to vaping as a possible gateway into cigarette smoking: an analytical review [version 3; peer review: 2 approved]. F1000Research. 2019;7:1915. Epub 2019/08/01. pmid:31354936; PubMed Central PMCID: PMC6652100.
  57. 57. Lee P, Fry J. Investigating gateway effects using the PATH study. F1000Research. 2019;8:264. Epub 2020/01/21. pmid:31956397; PubMed Central PMCID: PMC6950312.
  58. 58. Bowe AK, Doyle F, Stanistreet D, O’Connell E, Durcan M, Major E, et al. E-Cigarette-Only and Dual Use among Adolescents in Ireland: Emerging Behaviours with Different Risk Profiles. International Journal of Environmental Research and Public Health. 2021;18(1). Epub 2021/01/21. pmid:33466304; PubMed Central PMCID: PMC7795664.
  59. 59. Creamer MR, Dutra LM, Sharapova SR, Gentzke AS, Delucchi KL, Smith RA, et al. Effects of e-cigarette use on cigarette smoking among U.S. youth, 2004–2018. Prev Med. 2021;142:106316. Epub 2020/12/05. pmid:33272598; PubMed Central PMCID: PMC7796895.
  60. 60. Dutra LM, Glantz SA. E-cigarettes and National Adolescent Cigarette Use: 2004–2014. Pediatrics. 2017;139(2). Epub 2017/01/25. pmid:28115540; PubMed Central PMCID: PMC5260150.
  61. 61. Barrington-Trimis JL, Berhane K, Unger JB, Cruz TB, Huh J, Leventhal AM, et al. Psychosocial Factors Associated With Adolescent Electronic Cigarette and Cigarette Use. Pediatrics. 2015;136(2):308–17. Epub 2015/07/29. pmid:26216326; PubMed Central PMCID: PMC4516947.
  62. 62. U.S. Department of Health and Human Services. E-Cigarette Use Among Youth and Young Adults. A Report of the Surgeon General. Atlanta, USA: Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 2016.
  63. 63. Soneji SS, Sung H-Y, Primack BA, Pierce JP, Sargent JD. Quantifying population-level health benefits and harms of e-cigarette use in the United States. PLOS ONE. 2018;13(3):e0193328. pmid:29538396
  64. 64. Hartmann-Boyce J, McRobbie H, Lindson N, Bullen C, Begh R, Theodoulou A, et al. Electronic cigarettes for smoking cessation. Cochrane Database of Systematic Reviews. 2021. pmid:33913154
  65. 65. Wang RJ, Bhadriraju S, Glantz SA. E-Cigarette Use and Adult Cigarette Smoking Cessation: A Meta-Analysis. Am J Public Health. 2021 Feb;111(2):230–246. Epub 2020 Dec 22. pmid:33351653; PMCID: PMC7811087.
  66. 66. World Health Organization (WHO). WHO Report on the Global Tobacco Epidemic. Geneva: WHO, 2019.
  67. 67. WHO Study Group on Tobacco Product Regulation. Report on the scientific basis of tobacco product regulation: seventh report of a WHO study group. Geneva: WHO, 2019.
  68. 68. McDonald CF, Jones S, Beckert L, Bonevski B, Buchanan T, Bozier J, et al. Electronic cigarettes: A position statement from the Thoracic Society of Australia and New Zealand. Respirology. 2020;25(10):1082–9. pmid:32713105
  69. 69. Ferkol TW, Farber HJ, La Grutta S, Leone FT, Marshall HM, Neptune E, et al. Electronic cigarette use in youths: a position statement of the Forum of International Respiratory Societies. The European Respiratory Journal. 2018;51(5). Epub 2018/06/01. pmid:29848575.
  70. 70. Astor RL, Urman R, Barrington-Trimis JL, Berhane K, Steinberg J, Cousineau M, et al. Tobacco Retail Licensing and Youth Product Use. Pediatrics. 2019;143(2):e20173536. pmid:30617237
  71. 71. Kingsley M, Setodji CM, Pane JD, Shadel WG, Song G, Robertson J, et al. Short-Term Impact of a Flavored Tobacco Restriction: Changes in Youth Tobacco Use in a Massachusetts Community. American Journal of Preventive Medicine. 2019;57(6):741–8. pmid:31668668
  72. 72. Kennedy RD, Awopegba A, De León E, Cohen JE. Global approaches to regulating electronic cigarettes. Tobacco Control. 2017;26(4):440–5. pmid:27903958