Critical appraisal of the European Union Scientific Committee on Health, Environmental and Emerging Risks (SCHEER) Preliminary Opinion on electronic cigarettes

Not Quitting

The assumption that people who smoke want to quit is drawn from self-reports in surveys, but it constitutes a vague wish. On the other hand, when asked directly about quitting activities, a staggeringly high number of EU adults who smoke have no intention to quit [13].

Furthermore, for those that wish to quit, cigarette quit rates are very low, from 3%–12%, and relapse rates are very high, from 75%–80% in the first six months and 30%– 40% even after one year of abstinence [see studies cited in 14]. Quitting is not a single event but a dynamic process, and relapse is a common component of this process. While international guidelines place great emphasis on relapse prevention, very little can be offered to help smokers who have relapsed [15]. This challenge calls for innovative and effective alternatives for relapse prevention, including the use of ENDS.

Reduced Emissions Compared to Cigarettes

For those who do not make the effort to quit, and for those who cannot quit or remain abstinent, ENDS may provide an alternative to cigarettes. It is generally accepted that ENDS have substantially lower and fewer toxic emissions than cigarettes. The National Academies of Science, Engineering, and Medicine state “There is substantial evidence that except for nicotine, under typical conditions of use, exposure to potentially toxic substances from e-cigarettes is significantly lower compared with combustible tobacco cigarettes” [9, p. 6]. For example, emissions testing [16] found that ENDS toxicant emissions are 9 to 450 times lower than cigarette smoke. Furthermore, many toxic substances in cigarettes are not emitted by ENDS, as clearly illustrated in the research [17] that did not detect 61 of 79 compounds present in tobacco smoke (tally of Table 1). These findings are corroborated in the review by the European Respiratory Society [18] and other reviews [19, 20]. Stephens [21] calculated the cancer potency of ENDS emissions to have 0.004 of the relative lifetime cancer risk of tobacco smoke.

While major assumptions and guesswork are required to translate reductions in emissions into an estimate of actual health risk, it is impossible to believe that the orders-of-magnitude differences do not represent enormously lower risk. The Opinion never once compares the emission levels of ENDS vapour with cigarette smoke.

The UK regulatory organization, Committee on Toxicity of Chemicals in Food, Consumer Products and the Environment conducted an extensive and systematic review on the potential toxicological risks of ENDS [22]. Their conclusion:

“The use of E(N)NDS products, produced according to appropriate manufacturing standards and used as recommended, as a replacement for CC [cigarette] smoking, is likely to be associated with a reduction in overall risk of adverse health effects, although the magnitude of the decrease will depend on the effect in question.” (p. 28)

Biomarker Evidence

Except for one study on nicotine exposure, the Opinion does not present biomarker studies. Emissions testing is product testing, while biomarker studies provide evidence on actual human exposure to toxins. This is a serious shortcoming as research finds that exposure levels for a number of toxins are similar between ENDS users and non-smokers. Biomarker data is relevant evidence for demonstrating reductions in toxicant exposures from ENDS compared to cigarettes. The following studies represent some of the many biomarker studies that observe a substantial reduction in levels of toxins in the bodies of the participants who substituted ENDS for cigarette smoking.

The Opinion considered metal exposures, but did not cite a study [23] that reviewed blood lead (N=1899) and urinary cadmium, barium, and antimony (N=1302) test data in the 2015-2016 US National Health and Nutrition Examination Survey (NHANES). There were no significant differences in the levels of exposure to metals between participants who had never used ENDS and participants who were current or former ENDS users. The researchers conclude that ENDS are not a source of exposure to these heavy metals.

In a biomarker study funded by the US National Cancer Institute [24] 28 ENDS users who had quit smoking for a minimum of 2 months were given urine tests. The participants had significantly lower biomarkers of exposure compared to cigarette smokers. Their 1-HOP levels (exposure to polycyclic aromatic hydrocarbons) were similar to non-smokers. The study found significantly lower levels of metabolites in ENDS users compared to cigarette users. These were

* 1-hydroxypyrene/1-HOP (a marker for polycyclic aromatic hydrocarbons [PAH])

* 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol and its glucuronides/total NNAL (a marker for nicotine toxin)

* 3-hydroxypropylmercapturic acid/3-HPMA (a marker for acrolein)

* 2-hydroxypropylmercapturic acid/2-HPMA (a marker for propylene oxide)

* 3-hydroxy-1-methylpropylmercapturic acid/HMPMA (a marker for crotonaldehyde)

* S-phenylmercapturic acid/ SPMA (a marker for benzene).

The researchers conclude that

“levels of a suite of urinary toxicant and carcinogen metabolites were significantly lower in e-cigarette users than in cigarette smokers. These results suggest that e-cigarette use may be safer than cigarette smoking, at least with respect to the compounds studied here, which represent typical carcinogens and toxicants believed to be involved in causing cancer in cigarette smokers.” (p. 708).

A clinical study [25] examined the effects of four weeks of ENDS substitution with 33 participants who smoked. The mean 3-HPMA levels (acrolein) decreased 79% for exclusive ENDS users and 60% for dual ENDS and cigarette users. Carbon monoxide levels decreased by 80% in exclusive ENDS users and 52% in dual users. Reduction of carbon monoxide levels to within normal limits has been reported soon after switching from conventional cigarette smoking to exclusive ENDS use in other studies [26-28].

A before and after study [29] tested 20 Polish adult cigarette users for biomarkers of exposure after 2 weeks, with half of the participants substituting ENDS and half continuing to smoke. Significant reductions in exposure levels were detected for many toxicants in the ENDS users, listed in Table 2 below.

The substantial and significant reduction in 1,3-butadiene is particularly noteworthy as it is the greatest source of cancer risk in tobacco smoke [30]. Goniewicz et al. conclude that “e-cigarettes may effectively reduce exposure to toxic and carcinogenic substances among smokers who switched to these products” (p. 165).

A 4-week observational study was conducted with 40 adult cigarette users who added or substituted ENDS use [31]. Biomarker levels for exposure to NNAL, benzene, and acrylonitrile were significantly reduced in all participants. Participants reporting exclusive ENDS use for at least two weeks exhibited additional significant reductions in metabolite levels of ethylene oxide and acrylamide. Exclusive ENDS users had reductions in acrolein levels bringing them into the range of non-smokers.

While significant reductions in biomarkers of exposure are not evidence of an absence of risk, these studies (and industry studies not cited) demonstrate that exposures to toxicants are substantially and significantly lower for ENDS users compared to those who smoke cigarettes. People who smoke can substantially reduce their exposure to known toxicants by substituting ENDS for cigarettes, even when it is not complete substitution.

Individual and Population Health Benefits

We are not aware of any studies showing a negative impact on the health of individuals who substitute ENDS for cigarette smoking. Although concern has been raised about the long-term health impacts of ENDS, a handful of short-term and long-term clinical trials have demonstrated health benefits from ENDS substitution. It is regrettable that there are so few clinical trials on ENDS health effects outside of cessation, although smoking cessation has known health benefits. Complete ENDS substitution for smoking may well prove over time to substantially lessen the risks of continued smoking.

In a 12-month assessment of the impact of ENDS use on blood pressure (BP) in 89 patients with hypertension, consistent and clinically significant improvement in systolic and diastolic BP as well as in BP control was observed in those who switched to regular ENDS use [32]. These findings are in agreement with the 8.8 mmHg reduction in systolic BP at 12-month in a prospective randomized control trial looking at the effect of smoking cessation by using ENDS in subjects with high BP at baseline [33].

A randomized controlled trial (N=114) [34] demonstrated that four weeks of ENDS substitution for smoking resulted in significant improvements in flow-mediated dilation and decreases in vascular stiffness compared to the cigarette user arm, indicating a reduced risk for cardiovascular disease.

The respiratory health effects of ENDS have been addressed in two recent review articles and their conclusions are conflicting [35, 36]. Both these reviews cite in vitro and in vivo studies, but findings from human studies support the view that ENDS use shows no evidence of health harms and even health improvements in patients with chronic obstructive pulmonary disease (COPD) or asthma.

In a small study of daily ENDS users who had never smoked, no noticeable changes in health outcomes were detected over the 3.5 year observation period [37]. Daily usage of ENDS caused no significant changes in measures of lung function, respiratory symptoms and lung inflammation and no significant structural abnormalities could be identified on high-resolution computerised tomography (HRCT) of the lungs.

A five year follow-up clinical study (assessments at 12, 24, 48, and 60 months) of patients with COPD compared 19 patients who completely or partially substituted ENDS for smoking to 20 controls who smoked [38]. Switching to ENDS contributed to a rapid improvement in cardiorespiratory health, with better quality of life, improved exercise tolerance, and an approximately 50% reduction in COPD exacerbations. Of note, these health gains were consistent throughout the 5-year follow-up.

A study [39] evaluated exclusive ENDS users with asthma who had stopped smoking. In one section of the study, a web survey (N=382), 91.6% self-reported no worsening of symptoms from ENDS use. The second part of the research conducted clinical testing of 10 ENDS users with asthma at baseline, 3 months, and 6 months. The participants experienced a significant increase in asthma symptom control and improvements in the Asthma Quality of Life Questionnaire (AQLQ) scores over the course of the study.

In a 2-year prospective clinical study, ENDS use consistently improved objective and subjective asthma outcomes [40]. ENDS use was well tolerated, and exposure to e-liquid aerosol in this vulnerable population did not trigger any asthma attacks.

A positive impact of ENDS substitution on population health has been also predicted by population-based models. Prochaska and Benowitz, leading expert researchers on tobacco, state

“While e-cigarettes may have adverse effects on respiratory health and possibly other diseases, the harm is generally accepted to be much less than that of cigarette smoking. Thus, if smokers were to switch completely to e-cigarettes, then smoking-related disease is predicted to decrease substantially. Population-based models of the impact of e-cigarette use predict an overall health benefit” [41, pp. 17-18].

The WHO Study Group on Tobacco Product Regulation concurs, “The available evidence indicates a possible positive effect of ENDS on population health, particularly if appropriate ENDS regulation is enacted to maximize their benefits and minimize their risks” [10, p. 60]. The potential benefits of ENDS substitution for cigarette smoking on individual and population health certainly merit its inclusion in the final report.

Common Liabilities, an Alternative Hypothesis

The Terms of Reference specified an examination of a gateway effect. The gateway hypothesis is not the only explanation for the observed correlation between youth ENDS and cigarette use [9, 10]. The common liabilities theory posits that a “common latent propensity to risky behaviour” leads to concurrent cigarette and ENDS use [10, p. 57]. A review by the European Respiratory Society states that shared risk factors are “likely alternative explanations supported by the literature” [18, p. 14]. Hammond et al. [42] in a one year longitudinal cohort study (Canada, N=19,130) conclude that “it is highly plausible that ‘common factors’ account for a substantial proportion of increased cigarette-smoking initiation among e-cigarette users” (p. E1135). The common liabilities hypothesis that psychosocial factors explain poly-substance use by youth should be evaluated as an alternative explanation to a gateway effect.

In a recent editorial in the American Journal of Public Health [43], the gateway theory has been labelled as a “common deception.” Review teams are more sanguine, but unconvinced of a gateway hypothesis. A narrative review by [44] on youth ENDS use states that there is no strong evidence supporting the gateway hypothesis. A systematic review and meta-analysis [45] on adolescent ENDS use and smoking initiation reached the conclusion that “it is not clear how much of the relationship is causal (gateway effect) or is due to common liability…there is much less conclusive evidence for a gateway effect” (p. 13).

While no definitive explanation has been reached on why youth ENDS ever-users have a higher prevalence rate of smoking than those who never used ENDS, almost all of the studies come from US data. Considering this, we think it is important to note the evidence from a major survey of French youth conducted with compulsory participation for all 17 year olds [46]. In the 2017 survey (n=17,862 ever smokers) an analysis showed that youth who ever-used ENDS were less likely to be daily cigarette users (relative risk (RR)=0.62, CI 0.60-0.64) than ever-smokers who had never tried ENDS. Youth who tried ENDS first before ever-smoking were less likely to be daily cigarette users (RR=0.76, CI 0.66 – 0.89). Chyderiotis et al. observe that these findings are “in contradiction with the gateway hypothesis” (p. 5). Of course, no one study can prove or disprove a gateway effect, but this cross-sectional study is one of the very few studies available investigating a gateway effect among youth in an EU member state.

Renormalization

The Opinion alludes to a “resurgence of cigarette smoking” suggesting a renormalization of smoking. This is not happening, as clearly shown by the accelerated decline in smoking prevalence in EU member states, some of which have the highest prevalence of ENDS users. Changes in the trends of smoking prevalence are the best indicator of the absence or presence of renormalization [47]. Based on WHO data, between 2016 and 2018, 24 of 27 EU member states experienced declines in the prevalence of cigarette use for the population 15 years old and older. Seven member states had cigarette prevalence declines of 6% or better and three member states had declines over 10% during the 2-year period. See Table 3 below.

ENDS Prevalence Trends

The highest ENDS prevalences are between 4.1% and 5.7% in eight member states while the prevalence is under 2% in 13 member states. Nor are EU prevalences “increasingly rising” as stated in the Opinion, and in fact they have been relatively stable from 2017 to 2019. During this period three member states had no increase in the prevalence and seven member states had an increase of 0.2 percentage points or less. Only two member states had an increase in the prevalence of 1.0% or higher, the Netherlands with 1.0% and Portugal with 1.3%. See Table 4 below.

The Working Group cannot legitimately imply that ENDS may cause a resurgence of cigarette use because cigarette use had declined, often substantially, in almost every EU member state. Consumer data contradicts the statement that the prevalence of ENDS use among EU adults is rising quickly. In many EU member states ENDS usage prevalence has been relatively stable with only a very small rise, and in some member states no rise, between 2017 and 2019.

For data on the prevalence of youth use, the ESPAD report 2020, the European School Survey Project on Alcohol and Other Drugs, is due out November 12, 2020. Its findings should be included in the final report. Some prevalence data in the Opinion are from 2015 and earlier and so have a limited value for reporting current prevalences. Much of the prevalence data in the Opinion are for ever-use, a measurement that grossly over-represents the actual number of ENDS users, as discussed below.

Frequency of Use

The Opinion frequently cites ever-use data as if it were evidence of usage prevalence. Ever-use is a problematic measurement that captures a substantial number of one-time triers and substantially overestimates actual usage. Indeed, it does not even track it, given that within a cohort ever-use will always increase over time even if usage prevalence drops.

Ever-use data on youth users is unreliable for exposure assessment “as ever use can include using an e-cigarette once across the lifetime, the extent of increased nicotine exposure as a result of ever e-cigarette use is unclear” [48, p. 616]. Data from the Global Youth Tobacco Survey (GYTS) shows that 27% to 55% (varying by country) of EU youth who ever tried ENDS did so on only one occasion.

For Italian youth (15-19 years old), the ESPAD Italy 2017 survey found that over 70% of youth ever-users had used ENDS only one to nine times [50, supplementary material].

Ever-use measurement counts a substantial number of youth ENDS experimenters who do not go on to become regular users [51]. Similarly, among youth aged 13 to 17 in Norway in a four year longitudinal (2015-2019) qualitative study (50 semi-structured group and 175 individual interviews), it was often the case that as youth became older, ENDS use became viewed as a childish practice that they discarded [52].

Current youth use, defined as any use in the past 30 days, includes a substantial number of youths who use ENDS on only one or two days, including those who only tried ENDS once in their lives but it happened to be during that month. See Table 6 below.

Ever-use data for EU adults is also problematic for exposure assessment and estimating prevalence. The 2016 European Regulatory Science on Tobacco (EUREST-PLUS ITC, N=1178) found that among adult ever-users in Germany, Greece, Hungary, Poland, Romania, Spain, 38% had used ENDS 1-2 times and 21% had used ENDS 3-10 times. Furthermore 85% of ever-users were no longer using ENDS [53]. Current use of ENDS defined as any-past-30-day use is not a good indicator of sustained use as infrequent users of five days or less a month “includes many individuals who can be expected to discontinue use within 1 year” [54, p. e92].

The measurement of ever-use for youth and adults results in a misrepresentation of prevalence because many are simply experimenting on a very limited basis and do not continue use. The measurement of any-past-30-day use, the classification for a current user, implies a regular pattern of use. This representation is belied by the finding that 42% - 61% of EU youth defined as current users are infrequent users, only once or twice a month. This measurement may also be capturing recent triers.

Non-Nicotine Use

The Opinion makes no reference to use of non-nicotine ENDS, an omission of important data on nicotine exposure. In the EU, evidence shows that a substantial number of youth and adults do not use nicotine liquids.

Many EU youth ENDS ever-users report that they used non-nicotine products. In Finland: 52% of boys and 48% of girls used only non-nicotine ENDS, and nicotine use declined from 2013 to 2019 [55]. In France: 42.2% of ever-smokers and 92.9% of non-smokers used only non-nicotine ENDS [56]. In Italy: 72.0% overall, 40.3% used only non-nicotine and 31.7% used both nicotine and non-nicotine [49]. In Sweden: 38% used only non-nicotine ENDS [57]. A two year longitudinal cohort study in Finland (N=3474) found that exclusive use of non-nicotine ENDS by adolescents was not associated with a higher probability of becoming a daily smoker compared to never ENDS users [58].

Non-nicotine ENDS use also appears to be common among youth classified as “current” (recent) ENDS users. A US survey (N=1589, 15-17 years old) reporting past-30-day use found 29% used non-nicotine liquids and 39% reported using both nicotine and non-nicotine liquids [59].

Three studies indicate that many EU adults ENDS users are using non-nicotine liquids. A 2016 survey of French young adults (19-22 years old) recent ENDS users found 61 of 98 used only non-nicotine ENDS and an additional 19 reported using both [60]. A 2016 face-to-face interview project with 600 adult daily ENDS users in Barcelona, Spain reported that 33.7% of users quitting smoking and 43.6% of users reducing cigarette use did not use nicotine liquids [61]. In a 2016 online survey of current ENDS users in Poland (N=1142), 9.8% started ENDS use because they could use non-nicotine liquids [62].

Nicotine use during pregnancy is a known risk factor for adverse neonatal outcomes. A 2015 survey [63] of women in two US states having given birth that year found that 35.2% of women who used ENDS in the last three months of pregnancy (even once) used non-nicotine ENDS. During the period of three months before pregnancy to 2–6 months after delivery, 41.4% of ever-users used only non-nicotine ENDS.

Exposure assessments for ENDS use are complex. Approximately 75% of the Opinion’s statements on exposures are based on ever-use data. This data should either be excluded in the final report, or flagged to be as viewed with caution, because most of those included in these counts either used ENDS only a very small number of times or exclusively used non-nicotine ENDS. Infrequent use should, at a minimum, be noted in the exposure assessments. Reporting the prevalence of non-nicotine use is critical to exposure assessments because of its implications for the risks of nicotine addiction, pregnancy outcomes, and for the cardiovascular effects of ENDS use (see section following).