Evidence From the Scientific Assessment of Electronic Cigarettes and Their Role in Tobacco Harm Reduction

ECs were introduced to the market in 2004 and came to prominence around 2010. Despite this short period, the types of devices have cycled through several generations (Figure 1) (28), in each of which many different devices are available (36). From simple closed, low-power ‘cig-a-like’ devices (generation one), they transformed to refillable long-term use (generation two) and modifiable (generation three) systems, and further to discreet closed system devices with many safety features and improved nicotine delivery (generation four; Figure 1).

Figure 1

The basic elements of ECs are a battery connected to a heating element and an e-liquid reservoir. In first generation, single-piece cig-a-like products, all parts are encased into one (usually disposable) unit. The three-piece open systems of generations two and three consist of a separate battery, e-liquid cartridge, and atomizer (vaporizes the e-liquid), and parts and components may be swapped to customize the vaping experience (e.g., by increasing voltage, tank size, and/or e-liquid strength and flavour). Fourth-generation two-piece products consist of a battery and a cartomizer (a combined e-liquid cartridge and atomizer), which is the only exchangeable part. EC emissions may differ by product characteristics and operation and user behaviour (28). Many manufacturers’ design changes have been based on feedback from users searching for products that can meet their nicotine preference with sufficient satisfaction to enable them to replace cigarettes while favouring safety and flexibility (21, 36). In most countries, cig-a-like devices account for only 15% of the EC market, except in the USA, where they are used by around 50% of vapers (37).

The most common reasons for smokers rejecting ECs relate to performance (mimicking smoking and effectiveness at lessening cravings for smoking), but other common reasons are ease of use. In a survey published by Action on Smoking and Health (ASH) (38), 7% of users stopped using ECs because of difficulties in replacing components, refilling e-liquid, or due to leaking. Wadsworth et al. (39) found that the ease of using cig-a-like ECs (confidence in nicotine dose, no refilling required, ease of availability compared with later-generation devices, etc.) made them a popular first device for vapers. By contrast, third-generation modifiable models were considered “bulky” or “scary”. However, cig-a-likes (39) and e-cigarettes in general (40) were often found to be unsatisfactory.

An unpleasant problem is so-called dry wicking, more commonly known as “dry puff”, which is caused when the wick in the atomizer is not saturated, leading to the coil becoming overheated and thermal breakdown of solvents in e-liquid (41,42,43,44,45,46,47). The dry wick effect can occur when the e-liquid runs out and the user puffs deeply or if the voltage on modifiable devices is too high for the heating system. While this phenomenon is thought to increase the emission of toxicants (42), it causes a very unpleasant acrid taste (48) that is generally viewed as sufficiently arresting to prevent harm (24). In generation four products, some manufacturers have introduced pre-set power settings to help users to vary their vaping experience while avoiding dry wicking. For example, a distiller-plate heating system, which heats the e-liquid directly and replaces the coil and wick system, has increased nicotine delivery while minimizing the risk of dry puffs (21).

Thousands of flavours are available for ECs, and even within individual flavours, many variations in formulations exist. Sensorial aspects, such as sweetness, coolness, and vapour visibility or smoothness, play important parts in product acceptability (49,50,51). A systematic review by Zare et al. (52) found that all vapers preferred flavoured ECs, particularly sweet flavours, irrespective of age group. Adult consumers identify taste and variety of flavours as important characteristics of ECs (53). Most first-time purchases of ECs and e-liquids contain fruit flavours (54, 55). However, some studies suggest that flavours could be a determinant for adolescents trying ECs and, potentially, transitioning to smoking (56) or that flavours could reinforce the reward obtained from nicotine vaping products, increasing their potential for abuse liability (57, 58). An ASH survey (59) found that fruit flavours have overtaken both tobacco flavours and mint as the preferred e-liquid, accounting for nearly one-third of flavoured e-liquids used, and those using fruits or sweet flavours were more likely than tobacco flavour users to vape in order to quit smoking (60).

In many countries, ECs are regulated (in around 100 countries) (61), but the regulations are inconsistent and generally cover marketing, labelling, ingredients, and/or taxation, leading to highly variable product standards globally. However, only six countries have no regulations beyond minimum age for purchase (61). Marketing authorizations or product notifications/marketing applications may be required before products can enter the market. Pre-market authorizations involve the regulator giving permission for marketing after review of an application, placing some of the responsibility for a product on the regulator. By contrast, with product notifications/marketing applications, while the regulator may act on information provided, responsibility for the product remains with the submitter. The European Parliament, the US Food and Drug Administration (FDA), and various other national authorities request data on ingredients and various compounds in emissions, including types, quantities, and origins. Forty-two countries have banned ECs, most frequently based on the perceived risk of youth nicotine addiction or the potential that yet unknown long-term effects of vaping might outweigh any health benefits (61,62,63,64).

In the USA, regulation of ECs was introduced in 2016 (65). All new tobacco products, including ECs despite their lack of tobacco leaves, are subject to approval via Pre-Market Tobacco Product Applications (PMTA) (66). Manufacturers that wish to be able to claim that an individual tobacco product (but not a product class) reduces risk to health compared with smoking must also make a Modified Risk Tobacco Product Application (67, 68). Each submission requires a dossier with description and formulation of the product, description of non-clinical and clinical research findings relating to the effects of the product on tobacco-related diseases and other health-related conditions, how it will be used, as well as, examples of labelling and a description of how the product would be marketed. So far, numerous Pre-Market Tobacco Product Applications have been made for ECs but no Modified Risk Tobacco Product Application submission for ECs has yet been approved by the FDA. Dossier review times are expected to run into several years, making it likely the market will not easily benefit from product improvements as the category develops. In the EU, the Tobacco Products Directive 2014/40/EU (TPD2) regulates the manufacture, sale, and marketing of tobacco products (69). It incorporates a product notification system and aims “to facilitate the smooth functioning of the internal market for tobacco and related products, taking as a base a high level of health protection”. The first factor requires an assessment of the product relative to the products already on the market, whereas the second focuses on the safety of the product itself. The TPD2 sets certain minimum safety and quality requirements, including, but not limited to, maximum nicotine concentration (20 mg/mL) and maximum volumes for cartridges, tanks, and nicotine liquid containers (2 mL), child-resistant and tamper proof features on devices and e-liquid bottles, and refilling systems that prevent leaking. These are, in addition to EU safety regulations on restriction of hazardous substances in electrical and electronic equipment products (70), the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) regulation (71), testing requirements for electrical parts, and monitoring and reporting of adverse events (69). TPD2 is expected to be revised in the mid-2020s.

In most countries, advertising and marketing of ECs is highly controlled in terms of where and what information may be displayed and how. In the UK, the Committee of Advertising Practice does not allow advertising of ECs to make medicinal claims (e.g., can help with smoking cessation) unless they have marketing authorization from the Medicines and Healthcare Products Regulatory Agency, and messages must not appeal to youth or encourage non-smokers or non-nicotine-users to use ECs (72). In the US, modified risk claims can be made only for products authorized by the FDA via the Modified Risk Tobacco Product Application process (73). A few countries, like Canada and New Zealand, allow promotion of the use of ECs as less harmful alternatives to tobacco smoking by providing balanced risk information for consumers (74, 31).

As well as top-line requirements, further detail on interpretation and compliance are provided by regulatory guidance documents, national, regional, and international technical product standards, and voluntary industry codes. Technical committees creating standardised safety and quality guidance for ECs have been set up at the International Organization for Standardization (ISO) and the European Committee for Standardization (CEN), involving experts from industry, regulators, consumers, and other relevant stakeholders. Each has published two technical standards on generating emissions for measurements, what to measure in emissions, device safety, and analytical methods for measuring the main components in e-liquids (75,76,77,78). Further guidelines are being developed on electrical safety, manufacturing, ingredients, additional emissions and e-liquid measurement methods, consumer information, and labelling. Adoption of product standards is important to consumers to inform them about the quality of ECs, and to regulators and industry to clarify methodologies and data generated for marketing applications. Specifically, a framework of testing standards for ECs has been proposed, as many methods used had simply been adapted from cigarette testing methods. This will enable clearer comparison of aerosol yields within and across different product categories, including cigarettes (79).

Compounds of interest in ECs have often been based on smoking toxicant emissions (97) and many studies use smoking as a comparator, meaning that smoke toxicants have been the focus of most assessments so far (98,99,100,101,102). Investigational approaches are, therefore, evolving in response to emerging evidence and growing understanding of EC aerosol composition.

The EU TPD2 stipulates chemical emissions testing for multiple priority compounds, including acetaldehyde, acrolein, and formaldehyde (84). The US FDA has identified 92 harmful or potentially harmful constituents (HPHCs) in addition to nicotine (103) and public comment was recently sought on the proposal to add a further 19 compounds to the list (104). However, to our knowledge, only one study has investigated the emissions of these additional HPHCs (22). In practice, it has been suggested that Pre-Market Tobacco Product Applications should report at least 32 compounds (66). ECs generally do not reach temperatures higher than 250 °C during normal use or more than 350 °C under dry wicking conditions (105). However, a huge variety of devices and e-liquids exist (106) and many different puffing conditions and analytical techniques have been used to assess them (42, 107,108,109), but, overall, results indicate significantly lower levels of toxicants in EC aerosols than in cigarette smoke (20, 22, 43, 44, 102, 110, 111).

As explained in the product stewardship section (see p. 66–67), robust toxicological approaches guided by regulation are essential to screen ingredients and complex mixtures (92, 154). In the screening phase, in silico approaches can be useful to identify hazards from known substances in e-liquids and to estimate toxicity of substances for which toxicological profiles are not well characterised (155). Zarini et al. (156) used an in silico approach to develop quantitative structure-activity relationship models from data in the literature and toxicological databases in order to prioritize e-liquid ingredients according to potential acute toxicity. They used this approach to classify 264 e-liquid ingredients and flavours according to European classification labelling and packaging criteria and recommended this method to generate information for use in a weight-of-evidence approach (156).

Substances of potential concern identified through in silico toxicology should be investigated by in vitro and in vivo methods, including toxicological assays where appropriate, to calculate thresholds of concern (157). The US FDA still recommends in vivo studies when assessing acute effects in the respiratory system (158). The European Chemicals Agency (ECHA) (159) favours a weight-of-evidence approach, reflecting changing attitudes and laws in multiple countries about moving toxicology analysis away from animal testing towards innovative high-throughput and cell-culture platforms.

In vitro studies range from traditional toxicological models adapted from smoke exposure studies (160) to three-dimensional cell-culture models that recreate organotypic tissue (161, 162). Multiple studies have investigated acute and chronic toxicological risk of ECs (163, 164), including testing for mutagenicity (91, 115, 165,166,167), cytotoxicity (potency) (91, 115, 168,169,170,171), genotoxicity (91, 115, 172,173,174), oxidative stress (175, 176), and wound healing (177).

Traditional toxicological tests yield complex results and are affected by variability found between and within cell lines, limited translatability to the in vivo assays, and a lack of benchmarks to contextualize the findings (171, 178). Additionally, results might be affected by the diversity of EC designs, as found in a critical review of toxicological in vivo and in vitro studies by Wang et al. (164). Overall, though, the weight of evidence indicates that ECs present lower risks to users and bystanders than conventional cigarettes (179).

Broader toxicological approaches, such as systems toxicology, may be applied. Systems toxicology employs data from techniques like transcriptomics or proteomics from exposed cells to investigate pathways involved in oxidative stress, inflammation, cell proliferation, or DNA damage, and may highlight previously unidentified potential risks (180,181,182,183).

Computational risk assessment methodologies have been proposed to compare cancer potencies across tobacco products (47, 184). Cancer potency can be calculated from EC chemical emissions data and enable comparisons between products by factoring consumer exposure to different products. Stephens (47) suggested that ECs only have 0.004 times the carcinogenicity of cigarette smoke whilst still being 10.7 times more carcinogenic than nicotine inhalers.

Smoking increases health risks to cardiovascular, respiratory, and other systems soon after the onset of smoking, with the risks of death, disease, and disability rising with increasing duration of use. However, these excess risks are largely reversible with smoking cessation (7). Acute measurements of circulation and lung function may improve within 3–9 months of quitting, and coronary heart disease excess risk due to smoking is halved after 1 year and completely reversed by around 15 years (2, 185). The aim of assessing clinical effects of smoking, therefore, is not only to investigate damage caused but also to assess whether changing behaviour can reduce these excess risks and/or lead to functional benefits.

Disease-associated biomarkers known as biomarkers of potential harm (BoPHs) could provide valuable information characterising the risk of tobacco products. BoPHs were defined by the Institute of Medicine as the “measurement of an effect due to exposure; these include early biological effects, alterations in morphology, structure, or function, and clinical symptoms consistent with harm; also includes ‘preclinical changes” (11). Chang et al. (228) describe in detail BoPH in the context of tobacco product assessment as well as the types of BoPHs identified during this FDA public workshop. Table 4 also summarizes results from studies evaluating biomarkers of potential harm.

Usability of BoPHs to characterize risk associated with ECs is underpinned by reversibility of smoking harm achieved through complete cessation of tobacco products. Most clinical studies compare disease-associated BoPHs in vapers against those in smokers and a cessation group used as a reference. Many BoPHs are not smoking specific or even disease specific, so contextualization against effects observed in cessation groups is crucial to evaluate validity and biological relevance.

The same shortcomings observed in BoE studies are often found in BoPH studies but are often accentuated due to the non-specific nature of the biomarkers, leading to smaller effect sizes, i.e. requiring larger sample sizes. Other important factors to consider when designing these types of studies is to ensure their time frame is long enough for assessment of biomarker reversibility, as many BoPHs may take months to reveal significant changes.

Few BoPH studies have been done in EC users, and most are cross-sectional and compare self-reported vapers with smokers. No standard panel of BoPHs has been established, but most studies focus on end points related to cancer, cardiovascular disease, the respiratory system, oxidative stress, and inflammation.

Spirometry measurements of lung function, like forced expiratory volume in 1 s (FEV₁) and forced vital capacity (FVC), have been used as surrogate end points for COPD. Most studies comparing FEV₁, FVC, or ratio of FEV₁ to FVC in healthy vapers and smokers have shown no significant differences (229,230,231). Cibella et al. (229) did, however, report significant increases of forced expiratory flow at 25–75% of FVC as well as fractional exhaled nitric oxide. D’Ruiz et al. (232) observed some significant changes for FEV₁ and FVC 5 days after smokers switched to an EC in a clinical setting, but, unexpectedly, saw no change in the cessation group.

For cancer, besides BoEs like NNAL and volatile organic compounds, studies overwhelmingly present lower levels of BoPHs in vapers than in smokers but higher than in cessation groups. Other BoPHs, such as the inflammation biomaker 8-epi-prostaglandin F_2α, have also been associated with lung cancer (233), and levels are significantly lower in exclusive vapers than in smokers (234). Similarly, a cross-sectional study by Song et al. (235) concluded that most inflammatory BoPHs and cytokine concentrations in EC users were intermediate between those of smokers and non-smokers. In contrast, levels of 8-hydroxy-2'-deoxyguanosine, a biomarker for oxidative stress (236), have been reported to be higher in EC users than in non-smokers and not different to those in smokers (222).

Improvements in vascular health markers have been reported in smokers who switched to ECs for 1 month, particularly among women (237). Various BoPHs, such as 8-isoprostane (238) and 11-dehydro-thromboxane B₂ (239), have been associated with cardiovascular disease. Differences have been reported between smokers and vapers for 11-dehydro-thromboxane B₂ (234) but not for 8-isoprostane (222). Some reports have pointed to EC use as a potential cause of endothelial dysfunction (240) and heart rate variability (241, 242). In the PATH study, the association between vaping and heart failure was investigated, and the researchers concluded that respondents suffering heart failure had higher odds of being dual users than exclusive EC users (243). A comparison of vapers and/or dual users’ odds of suffering heart failure against those in smokers or never smokers would be of interest.

Most clinical studies assessing ECs have been small and with designs that limit interpretation and generalizability of conclusions. Studies should include appropriate controls, such as a cessation group (including people using NRT), as recommended by the Institute of Medicine (11), to provide context and determine the relevance of changes in relation to the time frame of the study. Much larger and longer-term studies assessing real-world behaviours are needed to fully understand the effects of switching to ECs, as the changes to health might take much longer to manifest than those traceable by other biomarkers in the short term (244).

A survey conducted in six European countries suggested that among respondents who were aware of ECs in 2016, 58.5% perceived them to be as risky or of higher risk than cigarettes, increasing to 61.8% in 2018 (270). Only around a quarter of respondents perceived reduced risk with ECs (28.6% and 28.4% in 2016 and 2018, respectively). A similar increasing trend in perceived high risk of ECs was seen in a US survey, where 11.5% (95% CI 10.0–13.2%) of adults viewed them to be as harmful as smoking in 2012 but 36.4% (35.1–37.7%) did so in 2017 (271). The importance of risk perception amongst vapers is clear, as shown in another study where dual users who perceived ECs as being less harmful than smoking were more likely to be exclusive vapers after 1 year than those who did not perceive them as less harmful (7.5% vs 2.7%, adjusted odds ratio 2.9, 95% CI 1.7–4.8) (269). It is worth noting that, according to an analysis based on the PATH study, the proportion of US adults perceiving e-cigarettes as harmful or more than cigarettes steadily increased from 53.7% to 72.7% between 2014 and 2016 (272). Other perceptions, such as product addictiveness, social acceptability, potential to harm others, and environmental impact, may also affect product choice. The most common reasons to use ECs among adults and youth in one survey were smoking cessation and novelty, respectively (273). In another study, the two main reasons for using ECs by current and former smokers were perceptions that they could be helpful to quit and were less harmful than smoking (274). In a Europe-wide survey in 2014, the main reasons given for use of ECs were to stop or reduce smoking and being able to use them in places where smoking was banned (275), while in 2020, reducing tobacco consumption was still the main reason for using ECs followed by believing they were less harmful than cigarettes (14). In the UK, ASH (276) reported that the three main reasons for vaping were as an aid to quit smoking, remain abstinent from smoking, and save money.

One criticism of ECs is their inefficacy for sustaining exclusive vaping and propensity to facilitate dual usage (277). These views, in part at least, were based on early studies of products providing low nicotine delivery and hence low user satisfaction (278). While ECs are not marketed as smoking cessation products, a large 1-year study in the UK showed that the smoking abstinence rate among people who switched to later-generation ECs was roughly double that among those who switched completely to NRT (212).

For smokers attempting to quit smoking, higher reward from ECs could help them to completely transition away from combustible cigarettes, as frequency of use of vaping products has been positively associated with becoming a former smoker with quit durations of 2–6 years (188, 279). However, increases in nicotine exposure could raise concerns about the abuse liability of ECs.

Abuse liability is the likelihood of engaging in persistent and problematic use of a substance that will lead to undesirable consequences (280). High nicotine delivery via ECs could increase the risk of abuse liability. However, if ECs can provide a viable alternative to cigarettes, they might have a public health benefit (24, 281, 282) and some degree of abuse liability might be acceptable (283, 284). A standard methodology has been established to assess the abuse liability of pharmaceuticals (285) and is largely adaptable for tobacco products (280, 286). This type of study is likely to be most relevant in countries where high concentrations of nicotine are allowed in e-liquids, such as the USA (287). In a prospective direct comparison of ECs with smoking and NRT (nicotine gum), Stiles et al. (96) found that while their study showed ECs probably had some degree of abuse liability, it was much closer to that of NRT than to combustible cigarettes. This conclusion was reinforced at category level by the analyses of symptoms of dependence collected through national US surveys, the Population Assessment of Tobacco and Health (PATH) (288, 289) and the National Adult Tobacco Survey (290). Both identified dependence symptoms in EC users but to a substantially lesser degree than in smokers. Another retrospective study, based on the International Tobacco Control database, investigated differences in symptoms of dependence between vaping and non-vaping ex-smokers in four countries (USA, UK, Australia, and Canada) (291). Vapers were more likely to perceive themselves as very addicted to smoking, but at the same time to feel more confident about remaining abstinent from smoking and to experience fewer urges to smoke than non-vapers.

Vaping products have experienced an increase in popularity all around the World. In Europe estimated vaping prevalence increased from 1.5% in 2014 to 1.8% in 2017 (292). Ever use ranged widely from 5.5% to 56.5% and was highest in younger age groups, with studies reporting consumption from respondents ranging from 10–24 years old. In the USA a larger study with 158,626 participants (293) reported ever use of vaping products of 14.8%, 12.8%, 9.4%, 7.0%, 2.3% in the age groups 18–24, 25–34, 35–44, 45–64, ≥65 years, respectively, while regular use (use in ≥20 of the previous 30 days) was more comparable across the same age groups (1.3%, 1.3%, 1.2%, 1.0%, 0.4%, respectively). In the UK, where uptake of ECs has been high, ever use in 2019 was estimated to be 16.5%, 20.1%, 12.3%, 10.8%, 5.2% in the age groups 16–24, 25–34, 35–49, 50–59, ≥60 years and 3.3%, 9.2%, 7.3%, 7.1%, 3.0%, respectively were regular users (294). Prevalence of vaping is generally lower among women than men, as for smoking (292, 295). Most vapers smoke or have smoked in the past. Vapers without smoking history account for a very small proportion of all vapers (26, 292, 294, 296,297,298).

Youth EC initiation has taken centre stage in the EC scientific world since the US Surgeon General declared EC use among youth “an epidemic” (299). Public Health England's (PHE) annual review estimated that EC prevalence with use of once per week or less was around 5% among 11–18-year olds, but varied by age. Prevalence among 11-year olds was 1% up to 5% among 15 year olds (26). A large 2017 review of surveys involving 60,000 children aged 11–16 years in the UK suggested that the prevalence of more frequent regular use (at least weekly) was 3%, even though proportions of 7–18% had ever tried ECs (300). Most young people regularly using ECs were already smokers, and among never smokers, only 0.1–0.5% regularly used ECs. In the US, uptake seems to be higher, with 13.1% of middle-school and high-school students nationally reported to be current EC users in 2015–2017, although most users (76.7%) also used at least one other tobacco product (301). However, smoking initiation rates among youth in the US reached the lowest recorded rate of 2.29% in 2018 (302), and similar patterns have been observed across European countries where ECs are widely available (303). Friend or caregiver smoking seems to increase the likelihood of children trying ECs, as does lower socioeconomic status (304, 305). Differing perceptions of risks have been found across subgroups of young people (e.g., classified by ethnic origin or sexuality) (306), but little work has been done on social factors (307), and this area needs greater attention.

Gateway theory originates in the use of a psychoactive drug being viewed as increasing the likelihood of using further drugs. When applied to vaping, it describes changes in a biological and/or behavioural pathway seen with use of a lower-risk product, such as an EC, increasing the risk of smoking combustible cigarettes for a higher reward (308). This theory has been cited to support the banning or restriction of access to EC products and liquids containing nicotine, for instance in Australia (309), but it is not backed up by reliable evidence (29, 310, 311). Application of such policies risk denying adult smokers the opportunity to switch to ECs and might increase relapse to smoking or smoking initiation (312). Kasza et al. (313) analysed the US PATH study database from 2013 to 2016 and found that ever users of ECs aged 12–17 years were more likely to have smoked in the previous 30 days at follow-up (adjusted odds ratio 3.4, 95% CI 2.4–4.7%). However, the likelihood of smoking was increased by use of any other tobacco product, including hookahs and smokeless tobacco. EC use was similarly increased by smoking (adjusted odds ratio 2.9, 95% CI 2.1–4.0%). Thus, changing to smoking could be explained by youth experimentation and propensity to risk (314, 315). Concurrent use of nicotine products has also been reported by Auf et al. (316). However, national surveys do not suggest any gateway effect. Studies from the USA (288, 302, 317), UK, New Zealand (318) and Canada (319), have concluded that (i) smoking usually precedes vaping; (ii) regular vaping is rare amongst never smokers; (iii) not all adolescent vapers used nicotine and dependence was lower than for smokers; (iv) since vaping has been introduced, smoking rates have continued to decline. The Canadian study concluded, “when it comes to smoking cigarettes, very few smoked cigarettes in addition to vaping and fewer still believe they started smoking after they had started vaping” (319).

Consumer use behaviour studies can provide an understanding of puffing topography, potential nicotine uptake, and variability in patterns of usage between different EC types and users (37, 320). The data are crucial for understanding how consumers use products and play a key part in setting the scientific framework for assessing risk potential. EC use differs significantly from cigarette smoking, although more information is needed on topography, perceptions, and behaviours. The US FDA recommends conducting “actual use” studies in “real-world conditions” within their Modified Risk Tobacco Product Application guidelines (73).

Consumer use behaviour studies should adhere to the principles of good clinical research practice, which aim to protect the wellbeing of participants and ensure appropriate study execution and data traceability. Attempts are being made to establish independent ethics committees that can weigh risks against (participant and society) benefits of use behaviour studies conducted with consumer products. Such committees would ensure that the rights, safety, and wellbeing of trial participants prevail over the interests of researchers, science, and society by providing participants with adequate information on study products; details of the study in a scientifically sound protocol; obtaining informed consent from every participant prior to participation; ensuring that the data generated are stored, recorded, handled, and accurately interpreted, verified, and reported by implementing quality systems and procedures.

In a typical EC use behaviour study, volunteers who have given informed consent are trained in the use of the products where necessary to ensure familiarity and are provided with the test product for use at home over a fixed period. Consumption patterns at home are self-reported while periodic attendance at a study site might also be required to assess puffing behaviour with a topography device (320, 321). Questionnaires might be used to obtain feedback about the sensory experience of vaping, satisfaction with selected attributes, and overall acceptability (322). Results from typical topography studies have been summarised in Table 5.

Hammond et al. (323) concluded that ECs topography studies “show a high degree of stability in puffing behaviour within the same subject over time, but considerable variability between e-cigarettes users”. Thus, more-granular information is needed to improve the accuracy of topography data and how they relate to nicotine exposure in users’ everyday environments. To enable the collection of real-life data over longer periods of time, attempts have been made to incorporate topography devices into ECs that record various attributes and transmit data via an internet-enabled electronic device to cloud data platforms (Table 6). Further work is needed to validate the accuracy and reliability of the data generated by these connected ECs compared to traditional instruments, but they seem not to substantially affect puffing behaviour (324).

Topography data can be used to improve understanding of baseline characteristics related to EC use, which can then be used to establish representative vaping machine protocols. While standardization of regimes would be valuable to industry, academics, and regulators, no single regime is likely to represent true human behaviour or produce emissions linked closely enough to human exposure given the wide range of puffing behaviours. CORESTA Recommended Method N^o 81 lays out the essential requirements necessary to generate and collect EC aerosol for analytical testing purposes (112), but the parameters do not reflect intense use.

Rather than using maximum settings for intense regimes, parameters for puff duration, volume, frequency, profile, and number, battery charge status, coil or atomiser age, voltage setting, ventilation setting, and device orientation should be based on representative human usage data. These parameters are not all independent and improved understanding of how different combinations affect the amount of aerosol generated will be central to defining protocols for testing and regulating ECs.