Discussion

Substantial variation was observed in the prevalence of ever and current use of waterpipe and e-cigarettes across EU MS, with current use remaining low despite higher levels of ever use. Our study shows contrasting temporal patterns in waterpipe and e-cigarette use (2012–2023): waterpipe use declined from 2017, with a temporary peak in 2020, whereas e-cigarette use increased substantially across most population groups throughout the study period. Trends in ever and current waterpipe use differed primarily by age group, with more pronounced variation across cohorts over time. By contrast, trends by gender and financial circumstance were less distinct, generally following similar trajectories over time, differing mainly in magnitude rather than direction. For e-cigarettes, increases in ever and current use were more pronounced among women, who showed the largest relative increases over time.

The observed levels of use among young people exceed estimates in EU-wide analyses of the general adult population for both products,16, 17, 18 highlighting the need for targeted prevention strategies, given the long-term risk of nicotine dependence with early exposure.8 However, prevention approaches may require different strategies for experimentation versus more sustained patterns of use. The lower prevalence of current use compared with ever use, suggests that experimentation remains common among young people, consistent with previous EU analysis.17 Curiosity and flavour appeal are commonly reported motivations for experimentation with both waterpipe and e-cigarettes,25,26 whereas regular e-cigarette use is more frequently associated with smoking reduction or cessation attempts.25 A qualitative study also suggests greater social acceptability of experimental e-cigarette use, whereas more regular use is less accepted,27 and waterpipe use remains strongly embedded in social contexts, which may limit progression to habitual use. A concerning pattern is co-use or transition to smoking,14,16 which has been linked to underestimation of nicotine dependence and the perception that occasional use is non-harmful.28 The role of e-cigarettes and other novel nicotine products in tobacco harm reduction remains the subject of ongoing debated; however, the present study was not designed to evaluate this framework. Rather, it provides descriptive evidence on population-level trends in ever and current e-cigarette use to inform tobacco control policy. These findings underscore the importance of distinguishing determinants of experimentation from those associated with regular use.

Furthermore, to assess whether our findings were sensitive to how ever and current use were defined, we explored alternative indicators, distinguishing between ever regular use, ever use (limited to experimentation), redefining current use to include those using less than monthly, and incorporating daily use as a marker of intensity. These additional analyses yielded broadly consistent temporal patterns, suggesting that our conclusions about changing experimentation and recent use among young people are unlikely to be materially affected by the specific operational definitions of ever and current use.

The concurrent decline in waterpipe use and increase in e-cigarette use among this age group mirrors patterns observed in the wider adult EU population.2,29 E-cigarettes may be more accessible at the population level, with fewer logistical and financial barriers compared with waterpipe use, which is often restricted to hospitality venues and requires greater coordination and costs. Evidence from a German cohort suggests that waterpipe use may precede later e-cigarette initiation among adolescents,30 which may be consistent with overlap in patterns of product ever use. Patterns across Northern European countries in this study are compatible with this hypothesis, with Estonia, Latvia, and Lithuania showing co-occurring high prevalence of both products. However, this represents an ecological inference based on country-level co-occurrence; these ecological patterns do not provide evidence of individual-level transitions between products, and no causal relationship or directional substitution can be inferred. Given that waterpipe use typically occurs in social settings, this may reflect contexts in which experimentation with nicotine-containing products co-occurs. These patterns may also be consistent with shared liability to nicotine use behaviours, whereby common genetic, psychological, and social factors predispose individuals to experiment with multiple products rather than reflecting directional transitions between them.31 Evidence from an ecological study of tobacco control policies in Europe and Canada found that policies targeting social norms around smoking were most strongly associated with declines in adolescent smoking,32 highlighting the importance of addressing social context in prevention strategies.

We observed aged-related heterogeneity in waterpipe use, with successive waves showing higher lifetime exposure in older cohorts, whereas exposure to e-cigarettes increased across all age groups. These patterns suggest age- and cohort-related differences in alternative tobacco and nicotine-containing product use among young people, although no directional substitution between products can be inferred. Nonetheless, the combination of declining waterpipe use and rising e-cigarette use is consistent with population-level changes in patterns of product use over time, whereby youth nicotine experimentation may be shifting from an analogue, venue-based social activity (waterpipe) towards a more digital, highly individualised, and, in some context, more affordable activity (e-cigarettes), in line with evolving product design and marketing strategies. Our findings therefore are compatible with the interpretation that waterpipes and e-cigarettes may play overlapping roles as youth experimentation products at the population level, without implying that one product has directly replaced the other or that they are equivalent in terms of health risk.

Financial circumstances shaped prevalence of use differently for each product. Young people reporting financial difficulties showed greater increases in waterpipe use in 2020, despite COVID-19-related social restrictions, which would have been expected to reduce use. Shifts in waterpipe use from public venues to homes during COVID-19 restrictions may have coincided with increased opportunities for regular use, especially in households with existing waterpipe users.26 However, this finding should be interpreted with caution given the methodological deviations affecting that wave.

For e-cigarettes, exposure increased among those with minimal financial difficulties. Individuals with greater spending capacity may be more likely to experiment, while e-cigarettes may serve as a lower-cost alternative for reducing cigarette consumption among some individuals with fewer resources.25 Economic considerations may thus influence ever use among price-sensitive young people, as demonstrated previously, where higher cigarette prices are linked to increased e-cigarette use.33 The absence of harmonised e-cigarette taxation across the EU creates price differentials between MS and facilitates cross-border purchases, and could be addressed by including e-cigarettes within the EU Tobacco Taxation Directive.34

A prominent finding was the nearly eightfold increase in current e-cigarette use among women. While earlier literature has found higher prevalence among men,10 the observed patterns may reflect staggered adoption across genders. This warrants attention, as it suggests that women represent an increasingly affected group, consistent with potential influences such as gendered digital marketing practices. However, it is important to note that e-cigarette use may reflect heterogeneous behavioural contexts, including experimentation among people who have never smoked as well as use among individuals attempting to reduce or quit smoking. User-led lifestyle branding may circumvent existing marketing restrictions under the Tobacco Products Directive (TPD),35 and is enabled by regulatory fragmentation across the Unfair Commercial Practices Directive,36 the Digital Services Act,37 and the Audiovisual Media Services Directive,38 none of which provides comprehensive coverage of user-based promotion of nicotine-containing products. Furthermore, the TPD’s failure to regulate the visual device design, unlike traditional cigarettes or flavour appeal, leaves a critical gap that may be associated with regular usage. This surge in regular e-cigarette use, alongside patterns observed for waterpipes, is consistent with population-level continuity in nicotine use behaviours rather than demonstrating individual-level initiation or transition between products in this group. Consequently, gender-specific trajectories of use and associated health outcomes must be monitored, with targeted policy interventions considered to address the appeal and promotion of e-cigarettes among young women.

To our knowledge, this is the first study to examine trends in waterpipe and e-cigarette use among young people in the EU overall and by sociodemographic characteristics. This approach addresses a previously underexplored research question by examining these products in parallel. Stratification by sociodemographic factors enabled identification of distinct trends between population subgroups to inform targeted interventions and research.

As a repeated cross-sectional design, the study does not allow inference on individual-level transitions between products, and all observed patterns reflect population-level prevalence and co-occurrence rather than behavioural trajectories. Accordingly, it cannot provide evidence that rising e-cigarette use has directly caused a decline in waterpipe use. The categorisation of smoking status (current, former, never) does not fully capture smoking uptake, intensity, duration, or patterns of multiple product use. In the absence of more detailed smoking histories, these patterns cannot distinguish between initiation, dual or poly-use, or substitution, and should therefore be interpreted as population-level prevalence trends rather than evidence of individual-level transition pathways. In addition, the findings reflect the 27 current EU Member States, and therefore may not be generalisable to countries with different regulatory approaches.

Additionally, small sample sizes in subgroup analyses for current use and in certain countries introduced uncertainty, reflected in wider CIs. Readers should therefore interpret these subgroup estimates with caution. This was compounded by methodological deviations in 2020, which may have affected comparability with other waves. We conducted sensitivity analyses excluding the 2020 wave, presented in the Supplementary Material (Supplementary Tables S9 and S10). Without the 2020 wave, no spike in current waterpipe use was observed. Given that this spike may reflect methodological factors and/or the unique circumstances of the COVID-19 pandemic, it is uncertain whether it should be considered in the interpretation of longer-term trends. Nonetheless, the main trends reported are based on comparisons between 2012/2017 and 2023, when the methodology was consistent.

Observations with missing data were excluded (4.8% of the analytical sample). Missingness was assumed to be broadly consistent with a Missing at Random mechanism, although this cannot be formally verified. While any deviation from this assumption could introduce bias, the relatively small proportion of missing data suggests that any resulting bias is likely to be limited, though it remains a study limitation. Moreover, weighting was applied to the analyses, ensuring that the results are representative of young people in the EU.

Self-reported measures of product use are widely used and well validated in epidemiological research, particularly in confidential survey settings such as those used in this study.39,40 While some degree of reporting error and misclassification cannot be excluded, this is unlikely to be systematic across countries or sociodemographic groups or to substantially affect the observed associations. Additionally, variations in question phrasing, structure, and response categories across waves may have introduced some reporting variability, potentially affecting comparability between waves.41 However, these differences were conceptually minor, as key measures such as ever use consistently captured whether an individual had ever tried or used the product. As such, any resulting variability is likely small and unlikely to materially influence the observed trends.

In conclusion, there was an overall decrease in waterpipe use and an increase in e-cigarette use between 2012 and 2023. Fewer young people are now exposed to waterpipe in the EU, whereas more are exposed to e-cigarettes. These findings suggest that, although waterpipes and e-cigarettes differ in composition, risk profile, and role in adult harm-reduction frameworks, they may play overlapping roles as vehicles for nicotine experimentation and regular use among young people at the population level. They also suggest the need for gender- and price-targeted strategies to reduce nicotine use among young people in the EU and support the goal of a tobacco-free generation. Ongoing surveillance is needed to capture evolving patterns of nicotine use and changes in the distribution of risk across subgroups. Insights into patterns of use, such as social settings, reasons for initiation, and continuation, would provide valuable context for profiling user groups and tailoring interventions.