Electronic Cigarette: A Longitudinal Study of Regular Vapers

Transcription

1 Nicotine & Tobacco Research, 2018, doi: /ntr/ntx132 Original investigation Received January 19, 2017; Editorial Decision May 26, 2017; Accepted June 6, 2017 Original investigation Electronic Cigarette: A Longitudinal Study of Regular Vapers Jean-François Etter PhD Institute of Global Health, Faculty of Medicine, University of Geneva, Switzerland Corresponding Author: Jean-François Etter, PhD, Institute of Global Health, University of Geneva, CMU, case postale, CH-1211 Geneva 4, Switzerland. Telephone: ; Jean-Francois.Etter@unige.ch Abstract Introduction: It is unclear how vaping behavior changes over time in regular vapers, and what occurs when vapers relapse to smoking or when they stop vaping. We assessed change in vaping and smoking behaviors over 12 months in regular vapers. Methods: A longitudinal study of 3868 regular vapers enrolled on the Internet in ( baseline ), followed after one (n = 1631, 42%), three (n = 1337, 35%), six (n = 1148, 30%), and 12 months (n = 893, 23%). Results: Participants had been vaping for a median of five months at baseline. Most (77%) were former smokers, who had not smoked for a median of three months at baseline. Over 12 months, enjoyment gradually became the most frequently cited reason to vape (93%), and vaping to deal with craving for tobacco gradually decreased (from 87% to 56%). In exclusive vapers (ex-smokers), nicotine concentration in e-liquids decreased over time (from 12 to 9 mg/ml), but puffs/day remained stable (200 puffs/day). After 12 months, 9% of 687 former smokers relapsed to smoking and 28% of 64 daily smokers (dual users) stopped smoking. After 12 months, when participants stopped vaping, they tended to relapse to smoking (+18% daily smokers among those who stopped vaping versus -2% in permanent vapers, p <.001). When ex-smokers relapsed to smoking, they tended to stop vaping. Conclusions: After 12 months, enjoyment and relapse prevention were the most important reasons to vape. Rates of relapse to smoking were low in former smokers and quit rates were high in current smokers. Stopping vaping was associated with relapsing to smoking. Implications: Even in established vapers, vaping behavior and reasons to vape change over time. This should be taken into account by clinicians, manufacturers and regulators. Results from this non-representative sample can help generate hypotheses that can later be tested in representative samples of vapers. Introduction It is necessary to understand how and why e-cigarettes are used over time and how using them influences smoking behavior. This study used an online longitudinal survey of regular vapers to assess change over time in vaping and smoking behaviors. Results from online samples cannot be generalized to all vapers, but patterns may emerge that can help generate hypotheses. The behavior of vapers may change over time because their reasons to vape change, because they acquire more experience with the devices, because their need for nicotine varies, because dual users (users of both combustible and e-cigarettes) stop smoking, because ex-smokers relapse to smoking, or because new products appear. There are many longitudinal studies of vapers, most of which assessed the effects of vaping on subsequent smoking, but according to a recent comprehensive review, these studies produced inconclusive results. 1 In contrast, there is relatively little information from longitudinal studies on changes over time in patterns of e-cigarette use, nicotine concentrations in e-liquids, nicotine intake, nicotine dependence, reasons to vape, effects on tobacco withdrawal The Author(s) Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please journals.permissions@oup.com. 912

2 Nicotine & Tobacco Research, 2018, Vol. 20, No symptoms, motivation to quit, etc. 1 4 In addition, because their methods and sampling differ, these longitudinal studies disagree on almost every outcome Thus, additional data are necessary to document vaping behavior over time. 11 Understanding change over time in the behavior of e-cigarette users is important to inform consumers, clinicians, manufacturers, retailers, and policy makers. Thus, the aim of this study was to assess change over time in regular vapers, focusing on patterns of use of combustible and e-cigarettes, withdrawal symptoms, reasons for vaping, and effects of vaping cessation. Methods Data Collection The study builds on our previous studies of vapers. 2,12 15 We posted a questionnaire on our smoking cessation website Stop-Tabac.ch, 16,17 in English and French, between October 2012 and January We contacted discussion forums and websites informing about e-cigarettes or selling them, and asked them to publish links to the survey ( Participants were >18 years, and only current users of e-cigarettes were eligible. Participants who indicated an address were contacted after one, three, six and 12 months and asked to answer a follow-up questionnaire. The study was approved by the ethics committee of the Geneva University Hospitals. Participants were nor compensated. Measures Baseline For the first assessment ( baseline ), the questionnaire had three consecutive pages, and there was attrition between successive pages. Page one (n = 3868 participants) assessed: - Prior and current use of e-cigarettes: Are you currently using the electronic cigarette? We included only those who answered: Yes, I use it every day and Yes, occasionally (not daily). - Smoking status; number of combustible cigarettes per day before using e-cigarettes; smoked even one puff in past 7 days; days smoking in past 31 days. In current smokers: cigarettes per day; minutes to first cigarette of the day; addiction to tobacco assessed with one item from the Cigarette Dependence Scale: Please rate your addiction to tobacco cigarettes on a scale of 0 100: I am not addicted to tobacco cigarettes at all = 0, I am extremely addicted to tobacco cigarettes = 100 ; 18 intention to quit smoking; confidence in ability to quit if they tried; likelihood of being quit in 1 month; whether currently trying to quit or reduce smoking. In ex-smokers: date when they quit smoking. - Demographics; height and weight; country; household income (below/above average); has a diploma that gives access to University; depression assessed using the Whooley 2-item screening test; 19 hazardous drinking, assessed using the Alcohol Use Disorders Identification test (AUDIT-C, a score 4 for men and 3 for women indicates hazardous drinking); 20 cannabis use ( During the previous 12 months, how often did you take cannabis? ); use of nicotine medications. We also recorded IP addresses in order to delete duplicate records. Page two (n = 2114 participants, 55%) assessed: - Duration of current episode of e-cigarette use; nicotine content of e-liquids; time to first e-cigarette of the day; puffs per day; monthly spending on e-cigarettes and refills (prices in currencies other than USD were converted to USD). - Reasons to vape: enjoy, to deal with craving, to quit smoking or avoid relapse, to deal with tobacco withdrawal symptoms, to deal with places where they cannot smoke, unable to stop vaping, addicted to e-cigarette. - Use tank systems or pre-filled cartridges: Do you use cartridges that are already filled (ready to use), or a tank system, or do you add drops of liquid to your electronic cigarette ( drip )?. - Equipment: What equipment do you currently use most? A disposable e-cigarette (non-rechargeable); a commercial e-cigarette kit (with rechargeable battery); a modular system (I use my own combination of separate devices: batteries, atomizers, etc.). Page three (1941 participants, 50%) assessed: - Tobacco withdrawal symptoms: Minnesota Withdrawal Scale 21 plus mood swings, cough, sore throat, want to hold a cigarette between their fingers. - Perceived addiction to e-cigarettes (0 100 scale); perceived ability to stop vaping; two items of the Mood and Physical Symptoms Scale (MPSS) modified to assess the frequency and intensity of urges to vape. 22 Follow-up The follow-up questionnaire indicated that participants should answer even if they were not using e-cigarettes anymore. We used the same questionnaire for the four follow-up surveys, which covered: - Current use of e-cigarettes; duration of current episode of use; date of last use; minutes to first puff of the day; nicotine content; puffs per day; cost per month. - Reasons to vape (same as baseline). - Perceived difficulty to stop vaping; addiction to e-cigarette (scale of 0 100); strength and frequency of urges to vape (MPSS). - Smoking behavior (same as baseline). - Tobacco withdrawal symptoms (same as baseline) and body weight. Analyses To assess the effect of attrition, we compared the 1924 participants who took part only in the baseline survey with the 589 participants who took part in all five surveys, and also with the 677 participants who answered the first two pages of the baseline survey and also answered the fifth survey (at 12 months), using t tests for means and chi-square tests for proportions (Table 1). Then, we present data for all the participants in each of the five surveys (Table 2), comparing each follow-up sample with the baseline sample using paired t tests for means and Wilcoxon signed ranks tests for proportions. We used repeated-measures ANOVA to test time trends for continuous variables. In baseline daily smokers, we used two criteria to define smoking cessation at follow-up: self-classification as an ex-smoker plus not smoking even a puff of tobacco in the past 7 days. In baseline ex-smokers defined by two criteria (self-classified as ex-smokers and no puff of tobacco in the past 7 days), we used two criteria to define relapse at follow-up: self-classification as a smoker or smoked one or more puffs of tobacco in the past 7 days. The characteristics of the successive samples differ for two reasons: (1) many baseline participants did not take part at follow-up and this self-selection process was not random, and (2) change over time occurred within the group who took part in all surveys. To

3 914 Nicotine & Tobacco Research, 2018, Vol. 20, No. 8 Table 1. Characteristics of Current E-cigarette Users Recruited via Internet Between All participants at baseline Took part at baseline only Took part in all 5 surveys Answered first two pages at baseline + answered at 12m N Age, years, mean *** 45*** Sex, male, % Household income above average, % * 43** Has a diploma that gives access to University, % Depression: Whooley screening test positive, % *** 35*** AUDIT-C positive: hazardous drinking, % * 52* Used cannabis in past 12 months, % *** 15** Former smokers; the remainder=current smokers, % *** 90*** Overweight, BMI 25, % *** 54*** Currently use a nicotine medication, % *** 6*** Daily vapers; the remainder = non-daily vapers, % *** 98*** Duration of e-cigarette use, days, mean Use tanks; the remainder use pre-filled cartridges, % *** 97*** Use a modular system (own combination of batteries and *** 65*** atomizers), % Monthly spending on e-cig, USD, mean *** 69*** Volume of e-liquid per month, ml, mean *** 77*** Former smokers, N Days since they quit smoking, mean Cigarettes per day before quitting, mean *** 25*** Minutes to first cigarette of the day before they quit * smoking, mean Addiction to cigarettes before they quit smoking, scale *** 92*** 0 100, mean Daily smokers (dual users), N Cigarettes per day before started using e-cigs, mean Tobacco cigarettes per day now, mean Minutes to first cigarette of the day, mean Addiction to cigarettes, scale 0 100, mean Currently trying to quit smoking, % ** 52 Currently trying to reduce number of cig. per day, % Has decided to quit smoking in next 30 days, % * Comparison with participants who answered at baseline only: *p <.05; **p <.01; ***p <.001 assess change over time separately from the effects of attrition, we focused on the cohort of 368 participants who were exclusive vapers (i.e. ex-smokers) at all five assessments (Table 3). On each of the five surveys, these participants declared that they vaped daily or occasionally, classified themselves as ex-smokers and had not smoked even a puff during the previous 7 days (Table 3). We do not report results for the cohort of participants who were dual users at all five time points, because of the small size of this cohort (n = 19). We used the same statistical tests in Table 3 as in Table 2. Then, we examined what occurred after daily vapers stopped vaping (Table 4). For this analysis, we compared change over time in the 53 baseline daily vapers who stopped vaping after 12 months versus change in the 751 baseline daily vapers who were still vaping daily after 12 months. To assess change over time within each group, we used paired-samples t tests for means and Wilcoxon signedrank tests for proportions. To compare change over time between these two groups, we computed change scores for means (i.e. score at follow-up minus score at baseline), and used t tests to compare groups on these change scores. For proportions, we also computed change scores and compared groups using chi-square tests on these change scores. We also compared the baseline characteristics of the 53 vapers who later stopped vaping with the baseline characteristics of the 751 permanent vapers, using t tests and chi-square tests. Finally, we examined what occurred after ex-smokers relapsed to smoking (Table 5). For this analysis, we compared the 46 vapers who were ex-smokers at baseline but relapsed to smoking after 12 months with the 624 baseline ex-smokers who were still not smoking after 12 months (they classified themselves as ex-smokers and had not smoked even a puff in the 7 days before each survey). We used the same statistical tests in Table 5 as in Table 4. We used Bonferroni corrections for all analyses, which implies that results were considered significant only if p <.001. For each analysis, we used pairwise deletion of missing values and did not use imputation methods for missing values. Results Participants The first page of the baseline survey was answered by 3868 current users of e-cigarettes, the second page by 2114 people and the third page by 1941 people. An address was provided by 2687 participants (69% of 3868). Of these, 1631 answered the follow-up survey after one month, (42% of 3868 or 61% of those who provided an address), 1337 after three months (34% of 3868), 1148 after 6 months (30%), and 893 after 12 months (23%); 589 participants (15%) took part in all five surveys, and 1924 persons

4 Nicotine & Tobacco Research, 2018, Vol. 20, No Table 2. Comparison of All Participants at Five Time Points First assessment 1 month 3 months 6 months 12 months F statistic for time trend p value for time trend N 3868 a Currently use e-cig daily or occasionally, % *** 96*** 93*** 90*** 29.4 <.001 Currently use e-cigarettes daily, % 89 92*** 91*** 88*** 86*** 29.4 <.001 Currently smoke daily or occasionally, % 22 14** 15 ns 13 ns 13 ns Currently smoke daily, % 14 7** 7 ns 8 ns 7 ns Smoked 1 puff of tobacco in past 7 days, % 32 18*** 18 ns 15** 15** Days using tobacco in past 31 days, mean 9 3*** 3*** 3*** 3*** Body weight, kg, mean 77 79* 80*** 80*** 80*** In dual users: daily smokers who vape, N Cigarettes per day, mean 13 10*** 11 ns 14 ns 14 ns Minutes to first cigarette of the day, mean 45 75* 50 ns 30 ns 26 ns Addiction to cigarettes, scale 0 100, mean 79 73*** 69*** 73** 72** Duration of most recent quit attempt, days ns 153 ns 116 ns 241 ns Has decided to quit in next 30 days, % 30 16** 14* 14* 13 ns Confidence in ability to quit if they tried, % ns 42 ns 44 ns 44 ns Likelihood of being quit in 1 month, % ns 32* 26 ns 28 ns In exclusive vapers (former smokers), N Days since they quit smoking, mean ns 382*** 478*** 635*** <.001 Withdrawal symptoms, slight to severe, % Desire or craving to smoke 46 28*** 21*** 18*** 16*** 57.2 <.001 Anxious, nervous 39 29*** 25*** 23*** 23*** 24.8 <.001 Angry, irritable, frustrated 33 24*** 22*** 20*** 20*** 11.0 <.001 Depressed mood, sad 30 28*** 23*** 22*** 23*** 11.3 <.001 Difficulty concentrating 33 24*** 22*** 20*** 18*** 15.1 <.001 Insomnia, sleep problems, awakening 43 39*** 35*** 33*** 34*** 12.7 <.001 Increased appetite, hungry, weight gain ns 43 ns 40 ns 44 ns Restless, impatient 35 27*** 23*** 23*** 22*** 16.3 <.001 Mood swings 30 26*** 22*** 21*** 21*** 6.0 <.001 Coughing 25 19*** 15*** 16*** 13*** 6.5 <.001 Sore throat 19 16*** 14*** 12*** 10*** 5.6 <.001 Want to hold a cig. between my fingers 34 18*** 13*** 11*** 9*** 63.6 <.001 In former e-cigarette users, N Days since they last used e-cigarette, mean <.001 In current e-cigarette users, N 3868 a Duration of e-cig use, days *** 348*** 436*** 561*** <.001 Minutes to first e-cig puff of the day, mean 46 50*** 47*** 94** 45 ns Puffs per day on e-cig, mean ns 222* 220 ns 229 ns Use nicotine-containing e-liquid, % ns 93** 93*** 91*** Nicotine content in e-liquid, mg/ml, mean 12 12*** 11*** 10*** 10*** <.001 Monthly spending for e-cigs, USD, mean 60 53*** 52*** 50*** 50*** 12.1 <.001 Stopping using the e-cig very difficult (%) 27 22*** 22*** 23*** 23*** 7.1 <.001 Addiction to e-cigs, scale, mean 59 54*** 54*** 53*** 54*** 26.5 <.001 Frequency of urge to vape today, a lot of 30 23*** 23*** 20*** 22*** 11.4 <.001 the time to all the time, % Strength of urges to vape, strong, % 28 20*** 19*** 15*** 20*** 29.3 <.001 Reasons to vape, somewhat + very true, % Because I enjoy it 82 88*** 90*** 91*** 93*** 14.9 <.001 To deal with my craving for tobacco 87 77*** 70*** 65*** 56*** <.001 To quit smoking or avoid relapsing 95 93*** 90*** 90*** 88*** 35.9 <.001 To deal with tobacco withdrawal 87 76*** 69*** 61*** 57*** 52.7 <.001 symptoms To deal with non-smoking situations/ 46 35*** 36*** 34*** 31*** 15.4 <.001 places Because is unable to stop using e-cigarette 30 25*** 26* 26*** 29** Because I am addicted to the e-cigarette ns 42* 42* 47 ns a At baseline, n = 3868 for page 1 of the questionnaire, n = 2114 for page 2 and n = 1914 for page 3 Comparison with baseline: *p <.05, **p <.01, ***p <.001

5 916 Nicotine & Tobacco Research, 2018, Vol. 20, No. 8 (50%) took part in the baseline survey but not in any of the subsequent surveys. At baseline, the distribution of respondents by country was: France (39%), US (18%), Switzerland (12%), UK (4%), and other countries (27%). Most participants were former smokers (77%) who had already stopped smoking for a median of three months (1 st and 3 rd quartiles: 1 month and 11 months). Most participants (89%) were using e-cigarettes daily, they had already been vaping for a median of five months at baseline (1 st and 3 rd quartiles: 1 month and 12 months), most (96%) used tank systems, 58% used modular systems and 4% used pre-filled cartridges. Half of the participants were overweight (BMI 25), 43% screened positive for depression (Whooley test), 53% for hazardous drinking (AUDIT-C), and 19% had used cannabis in the past year. Most current smokers (i.e. dual users) were currently trying to quit or reduce smoking (Table 1). Attrition We compared the 1924 persons who answered only once (i.e. in the first survey) with the 589 persons who took part in all five surveys, and also with the 677 persons who took part in the first two pages of the first survey and who also answered the fifth survey (at 12 months) (Table 1). At baseline, the two samples of participants in the follow-up surveys were older than the 1924 non-participants and they were less likely to be depressed, but more likely to be daily vapers and former smokers and to be overweight. The participants at follow-up spent more money than the 1924 non-participants on e-cigarettes, they used more e-liquid and were more likely to use modular systems (i.e. their own combination of batteries and atomizers) (Table 1). Among former smokers, participants retrospectively reported a higher level of addiction to cigarettes before they quit smoking, compared with non-participants. Use of e-cigarettes and Combustible Cigarettes at Follow-up Most participants were still vaping daily after one, three, six and 12 months, and the proportion of daily vapers decreased only slightly over time (Table 2). Among daily smokers at baseline who answered the follow-up surveys, 15% (19/128) had quit smoking after one month (two criteria: self-reported non-smoker plus no puff of tobacco in past 7 days), 24% (21/86) after three months, 25% (17/69) after 6 months, and 28% (18/64) after 12 months. Among baseline former smokers who answered the follow-up surveys, 7% (88/1214) had relapsed to smoking after 1 month (two criteria: self-reported smoker or 1+ puff of tobacco in past 7 days), 9% (94/1032) after three months, 8% (74/889) after 6 months, and 9% (63/687) after 12 months. Of the 474 baseline ex-smokers who answered all five surveys, 389 (82%) remained continuously off cigarettes at 1, 3, 6, and 12 months. Of the 481 recent ex-smokers who had quit smoking 31 days before baseline and answered the follow-up surveys, 14% (27/196) had relapsed to smoking after 1 month, 22% (35/158) after 3 months, 15% (20/130) after 6 months, and 20% (23/113) after 12 months. Of the 73 baseline recent ex-smokers who had quit smoking 31 days before baseline and answered all five surveys, 53 (73%) remained continuously off cigarettes at 1, 3, 6, and 12 months. Cohort of 368 participants who were exclusive vapers (i.e. ex-smokers) at all time points Vaping Behavior Among the 368 participants in this cohort, we observed a decrease between the successive surveys for the nicotine concentration in e-liquids, the amount of money spent monthly on e-cigarettes, the perceived addiction to e-cigarettes, the frequency and strength of urges to vape, and the perceived difficulty to stop vaping. In this group, the number of puffs per day on e-cigarettes and the time to the first e-cigarette puff of the day remained stable over time (Table 3). Reasons to Use E-cigs In these 368 permanent vapers and ex-smokers, endorsement of vaping for enjoyment increased over time and was the most frequent reason to vape at 12 months. Vaping to quit smoking or to avoid relapse were the next most frequent reasons to vape at 12 months. The following reasons to vape were endorsed less and less frequently between the successive surveys: vaping to deal with craving for cigarettes and with other tobacco withdrawal symptoms, vaping to quit smoking or to avoid relapsing, and vaping to deal with smoke-free places. A minority of vapers endorsed feeling unable to stop vaping, and answers to this questions remained stable over time, as did the perception of being addicted to e-cigarettes (endorsed by half the participants). Tobacco Withdrawal Symptoms In these 368 participants, all tobacco withdrawal symptoms decreased over time, except appetite and body weight, which remained stable. Comparison of 53 Daily Vapers Who Stopped Vaping vs. 751 Daily Vapers Who Continued to Vape Daily at Follow-up At baseline, the 53 daily vapers who would later stop vaping (after 12 months) were less likely to use nicotine-containing e-liquids than the 751 daily vapers who would later still vape daily (after 12 months), they used lower concentration of nicotine in e-liquids, they inhaled fewer e-cigarette puffs/day, and they waited longer to take their first puff on their e-cigarette in the morning (Table 4). There was no between-group difference for the models of e-cigarettes used at baseline. There were more women among the 53 vapers who later stopped vaping than among the 751 permanent vapers (60% vs. 37%, X 2 = 11.2, p = 0.001). In the 53 vapers who stopped vaping after 12 months, craving to smoke decreased between baseline and 12 months, but there was no statistically significant change for the other tobacco withdrawal symptoms. In these 53 participants, the proportion of daily smokers increased from 8% to 26% between baseline and 12-month followup (p = 0.03, non-significant after Bonferroni correction), and in smokers, cigarette consumption increased non-significantly from 4 to 14 cigarettes per day between baseline and 12 months, (p = 0.11), but remained below the 21 cigarettes/day these participants smoked before they started to vape, as reported retrospectively at baseline (14 vs. 21 cig./day, t = 5.2, p <.001 from paired t test). The frequency and intensity of urges to vape decreased over time in this group. These 53 participants had stopped vaping for a median of 8 months before the 12-month survey. In contrast, in the 751 permanent daily vapers, all tobacco withdrawal symptoms except appetite/hunger decreased between baseline and 12 months. Urges to vape also decreased over time in this group. The proportion of smokers did not change significantly in this group, but the proportion reporting any tobacco use in the past 7 days decreased (from 16% at baseline to 10% after 12 months, p <.001). In people who were current smokers at baseline and at 12 months (i.e. dual users), the number of cigarettes per day, the motivation to quit smoking and the perceived ability to quit smoking remained stable over time (Table 4).

6 Nicotine & Tobacco Research, 2018, Vol. 20, No Table 3. Cohort of 368 Participants Who Were Exclusive Vapers (Ex-smokers) at All 5 Surveys First assessment 1 month 3 months 6 months 12 months F stat. for time trend p value for time trend Currently use e-cigarettes daily, % ns 100 ns 98 ns 98* Days since they quit smoking, mean *** 408*** 486*** 681*** <.001 Duration of e-cig use, days, mean ns 381*** 447*** 613*** <.001 Minutes to first e-cig puff of the day, mean ns 36 ns 36* 40* Puffs per day on e-cig, mean ns 263** 250 ns 259 ns Use nicotine-containing e-liquid, % ns 95 ns 92** 91** Nicotine content in e-liquid, mg/ml, mean 12 11*** 11*** 10*** 9*** 80.6 <.001 Monthly spending for e-cigs, USD, mean 76 69* 60*** 57*** 53*** 9.4 <.001 Stopping using the e-cig would be very difficult ns 24** 25** 22*** 8.6 <.001 impossible, % Addiction to e-cigs, scale, mean 66 60*** 58*** 58*** 55*** 18.7 <.001 Frequency of urge to vape today, a lot of the time 32 25*** 23** 22*** 21*** 7.7 <.001 to all the time, % Urges to vape, strong % 25 22*** 22*** 17*** 18*** 16.6 <.001 Reasons to use e-cigs, among current users, somewhat + very true, % Because I enjoy it 90 95*** 95*** 96*** 97*** 13.2 <.001 To deal with my craving for tobacco 86 70*** 62*** 56*** 48*** 88.2 <.001 To quit smoking or avoid relapsing 99 94*** 89*** 91*** 87*** 29.6 <.001 To deal with tobacco withdrawal symptoms 87 73*** 66*** 56*** 48*** 79.7 <.001 To deal with non-smoking situations/places 39 32*** 26*** 28*** 25*** 15.1 <.001 Because in spite of my efforts, I am unable to 36 25** 24* 28 ns 28* stop using the electronic cigarette Because I am addicted to the e-cigarette 50 47* 46* 48 ns 46* Withdrawal symptoms, slight + mild + moderate + severe, % Desire or craving to smoke 42 23*** 19*** 16*** 14*** 46.5 <.001 Anxious, nervous 40 25*** 25*** 20*** 21*** 24.5 <.001 Angry, irritable, frustrated 30 22*** 21*** 21*** 20*** 9.0 <.001 Depressed mood, sad 30 24** 22*** 19*** 20*** 10.2 <.001 Difficulty concentrating 31 20*** 23*** 20*** 18*** 14.4 <.001 Insomnia, sleep problems, awakening 40 34** 35** 31*** 34*** 6.9 <.001 Increased appetite, hungry, weight gain ns 40 ns 40 ns 39 ns Restless, impatient 38 22*** 22*** 22*** 23*** 19.0 <.001 Mood swings 31 24** 21*** 23*** 22*** 5.6 <.001 Coughing 23 17** 13*** 17** 13*** 6.3 <.001 Sore throat 20 12** 13* 11*** 11*** 6.2 <.001 Want to hold a cig. between my fingers 31 13*** 10*** 9*** 8*** 56.5 <.001 Body weight, kg, mean 82 82* 82** 82* 82** Comparison with baseline: *p <.05; **p <.01; ***p <.001; ns: p >.05 The comparison of change over time between the 53 daily vapers who stopped vaping after 12 months with the 751 permanent daily vapers showed between-group differences for change scores (i.e. scores at 12 months versus scores at baseline) for smoking status (daily, occasionally, non-smoker) and for urges to vape. That is, even though the within-group change in smoking status was not statistically significant after Bonferroni correction, the between-group difference in change over time in smoking behavior was statistically significant (+18% daily smokers among those who stopped vaping versus 2% daily smokers in permanent vapers, p <.001, Table 4). Comparison of 46 Vapers Who Relapsed to Smoking at Follow-up With 624 Permanent Ex-smokers There were more women (63%) among the 46 vapers who were exsmokers at baseline but later relapsed to smoking after 12 months ( relapsers ) than among the 624 vapers who were ex-smokers at baseline and were still not smoking after 12 months ( permanent ex-smokers, 37% of women, X 2 = 12.5, p <.001). At baseline, the relapsers were more likely than the permanent ex-smokers to have used tobacco in the previous 31 days (Table 5). Of the 46 relapsers, 35% smoked daily after 12 months and the rest smoked occasionally. In the 46 relapsers, the proportion of daily vapers decreased from 96% to 65% between baseline and 12 months, although non-significantly after Bonferroni correction (p = 0.014). In relapsers, cigarette consumption at 12-month followup (6 cig./day) was lower than before they started to vape (22 cig./ day, t = 5.4, p = <.001 from paired t test). These 46 participants had relapsed on average 5.5 months before the 12-month survey. In the 624 permanent ex-smokers, the proportion of daily vapers decreased slightly between baseline and 12 months (from 99% to 91%), monthly spending on e-cigarettes and nicotine concentration in e-liquids also decreased, but puffs/day on e-cigarettes remained stable. All tobacco withdrawal symptoms except appetite decreased, but body weight increased by 1 kg over 12 months in this group. Perceived addiction to e-cigarettes and urges to vape decreased after 12 months, vaping for enjoyment increased and all the other reasons to vape decreased in this group (Table 5).

7 918 Nicotine & Tobacco Research, 2018, Vol. 20, No. 8 Table 4. Change Over Time in Two Cohorts: Daily Vapers who were Stilly Vaping Daily After 12 Months Versus Daily Vapers who had Stopped Vaping After 12 Months Daily vapers still vaping daily at 12 m Within-group change over time, p Daily vapers who stopped vaping at 12 m Within-group change over time, p Baseline 12 months Baseline 12 months Between-group difference in change over time, p Between-group difference at baseline, p N Currently smoke daily or occasionally, % < Currently smoke daily, % < In former smokers, N Days since they quit smoking, mean < < In current smokers, N Cigarettes per day, mean Minutes to first cigarette of the day, mean Addiction to cigarettes, scale 0 100, mean Has decided to quit smoking in next 30 days, % Confidence in ability to quit smoking if they tried, % Likelihood of being quit in 1 month, % Vaping behavior Duration of e-cig use, days, mean < <.001 < Minutes to first e-cig puff of the day <.001 Puffs per day on e-cig, mean Use nicotine-containing e-liquid, % <.001 Nicotine content in e-liquid, mg/ml, mean < <.001 Monthly spending for e-cigs, USD, mean < Stopping using the e-cig would be very difficult + impossible (%) < Addiction to e-cig, scale, mean < Frequency of urge to vape today, a lot of < <.001 < the time to all the time, % Strength of urges to vape, strong, % < <.001 < Reasons, somewhat true + very true, % Because I enjoy it < To deal with my craving for tobacco < To quit smoking or avoid relapsing < To deal with tobacco withdrawal symptoms < To deal with non-smoking situations < Because is unable to stop using the e-cig < Because is addicted to the e-cigarette Withdrawal symptoms, slight + mild + moderate + severe, % Desire or craving to smoke < < Anxious, nervous < Angry, irritable, frustrated <

8 Nicotine & Tobacco Research, 2018, Vol. 20, No Table 4. Continued Between-group difference at baseline, p Between-group difference in change over time, p Daily vapers who stopped vaping at 12 m Within-group change over time, p Daily vapers still vaping daily at 12 m Baseline 12 months Baseline 12 months Within-group change over time, p Depressed mood, sad < Difficulty concentrating < Insomnia, sleep problems, awakening < Increased appetite, hungry, weight gain Restless, impatient < Mood swings < Want to hold a cig. between my fingers < Body weight, kg, mean < The comparison of change over time between the 46 relapsers and the 624 permanent ex-smokers showed a significant betweengroup differences for change scores (i.e. score at 12 months versus scores at baseline) for vaping (daily, occasionally, stopped). That is, even though the within-group change in vaping behavior among relapsers was not statistically significant after Bonferroni correction, the between-group difference in change over time in vaping behavior was statistically significant ( 31% daily vapers among relapsers versus 8% among those who did not relapse, p <.001 for betweengroup difference in change over time, Table 5). Discussion E-cigarette Use Over Time During the course of 12 months, patterns of e-cigarette use were relatively stable in this sample of regular vapers enrolled on the Internet. In former smokers, as the duration of smoking abstinence increased, the intensity of craving to smoke logically decreased, and reasons for using e-cigs changed correspondingly: e-cigarettes were used less and less to deal with urges to smoke and with other tobacco withdrawal symptoms and more and more for enjoyment. Thus, long-term vaping appears to be largely a matter of pleasure and relapse prevention. This should be taken into account by clinicians, manufacturers, retailers, public health experts and policy makers. Permanent vapers gradually decreased the nicotine concentration in their e-liquids, as in a previous report. 4 Whether this reflects an intention to reduce nicotine intake, or a reduced physiological need for nicotine, or change in the materials (e.g. subohm vaping) deserves further scrutiny (subohm vaping means using coils with resistances <1 Ohm, which produces a dense aerosol and requires liquids with low levels of nicotine). Smoking Cessation and Relapse Over 12 months, we observed low rates (9% of 687) of relapse to smoking in former smokers and high rates (28% of 64) of smoking cessation among current smokers (dual users). As a comparison, prior studies indicate that one third of former smokers who have quit smoking for one year eventually relapse to smoking during the following years, 23 compared with 9% in this sample. The relapse rate was also low (20% after 12 months) in recent quitters who had quit <1 month before baseline, whereas previous research indicates that 70 90% of recent quitters relapse to smoking within 6 12 months. 24 The majority of ex-smokers remained continuously off cigarettes during the course of the study. While these observational data do not enable us to establish causal links between vaping and smoking, our results do not contradict reviews which conclude that even though the quality of the current evidence is low, e-cigarettes may help smokers stop smoking. 1,25 Dual Users Dual users (vapers who also smoked) reduced their cigarette consumption by half (from 20 to 10 cig./day) after they started to vape, but this effect occurred before they answered our first survey. Those who continued to use both products did not further decrease their cigarette consumption during the course of the study, and they maintained their habit of smoking their first cigarette of the day soon after waking up, possibly because combustible cigarettes are better than e-cigarettes at boosting nicotine levels in the morning, when these levels are depleted. In dual users, smoking cessation rates were

9 920 Nicotine & Tobacco Research, 2018, Vol. 20, No. 8 Table 5. Change Over Time in Two Cohorts: Permanent Vapers and Ex-smokers Versus Baseline Vapers and Ex-smokers who Relapsed to Smoking After 12 Months Vapers + ex-smokers at baseline and at 12 m Within-group change over time p value Vapers+ ex-smokers who relapsed at 12 m Within-group change over time p value Between-group diff. in change over time p value Between-group difference at baseline p value N Currently use e-cigarettes daily or occas., % < <.001 Currently use e-cigarettes daily, % < Days using tobacco in past 31 days, mean 2 0 < <.001 <.001 Duration of current or most recent episode of e-cigarette use, days, mean < < Minutes to first e-cig puff of the day, mean Puffs per day on e-cigarette, mean Use nicotine-containing e-liquid, % < Nicotine content in e-liquid, mg/ml, mean 12 9 < Monthly spending for e-cigs, USD, mean < Stopping using the e-cig would be very difficult < impossible, % Addiction to e-cigs, scale, mean < Frequency of urge to vape today, a lot of the < time to all the time, % Strength of urges to vape, strong, % < Among users, somewhat + very true, %. I vape... Because I enjoy it To deal with my craving for tobacco < To quit smoking or avoid relapsing < To deal with tobacco withdrawal symptoms < To deal with non-smoking situations or < places Unable to stop using the e-cigarette Because I am addicted to the e-cigarette Withdrawal symptoms, slight + mild + moderate + severe, % Desire or craving to smoke < Anxious, nervous < Angry, irritable, frustrated < Depressed mood, sad < Difficulty concentrating < Insomnia, sleep problems, awakening < Increased appetite, hungry, weight gain Restless, impatient < Mood swings < Coughing < Sore throat 20 9 < Want to hold a cig. between my fingers 31 9 < Body weight, kg, mean <

10 Nicotine & Tobacco Research, 2018, Vol. 20, No high and motivation to quit smoking remained stable over the course of 12 months. What Happened When Vapers Stopped Vaping? When they stopped vaping, ex-smokers tended to relapse to smoking, and dual users tended to increase their cigarette consumption, but they still smoked fewer cigarettes per day than before they first started to vape. This risk of relapsing to smoking after vaping cessation should be kept in mind by vapers, even by those who have stopped smoking for several months. Exploring why women were more likely than men to stop vaping at follow-up would require a detailed analysis that is beyond the scope of this paper. What Happened When Ex-smokers Relapsed to Smoking? When vapers who were ex-smokers relapsed to smoking, they also tended to stop vaping. When ex-smokers relapsed to smoking, their cigarette consumption remained lower than before they first started to vape, which suggests that vaping during several months did not increase the severity of their nicotine addiction. We observed higher rates of relapse to smoking in women than in men, as in previous research. 26 Body Weight and Appetite In participants who were both permanent vapers and permanent exsmokers, appetite remained stable and body weight increased only slightly over 12 months, either because any post-smoking cessation weight gain took place during the first months after they quit smoking, i.e. before our baseline survey, or because of a direct effect of vaping on appetite and weight. Further studies should test whether vaping is associated with appetite and weight gain in ex-smokers, a possible effect of both nicotine and food flavors. 27 in this study were not particularly intensive users. Participants started vaping five months before our baseline survey, thus, any effects that occurred during the first months of vaping were not captured by this study design. In addition, response rates at follow-up were low, and participants at follow-up differed from the initial sample. Also, vapers may not be able to reliably report the number of puffs per day. Finally, technology progresses rapidly, and our results may not apply to future e-cigarette models. Thus, our results may have limited generalizability and should be interpreted with caution. Conclusions In regular vapers, over the course of 12 months, patterns of e-cigarette use were relatively stable. Enjoyment and relapse prevention gradually became the most important reasons to vape. Rates of relapse to smoking were low in former smokers and rates of smoking cessation were relatively high in current smokers (dual users). When ex-smokers stopped vaping, they tended to relapse to smoking, and when ex-smokers relapsed to smoking, they tended to stop vaping. These results, obtained from a non-representative sample, may help generate hypotheses that can later be tested in representative samples and experimental studies. Funding Swiss Tobacco Prevention Fund (Swiss Federal Office of Public Health), grant The Tobacco Prevention Fund had no role in the design and conduct of the study. JFE s salary is paid by the University of Geneva. Declaration of interests None declared. Characteristics of Participants We observed high rates of hazardous drinking, cannabis use, overweight and depression in this sample. The co-occurrence of tobacco smoking, alcohol abuse, use of illicit substances, obesity, depression and other psychiatric conditions is well documented in other populations, and a high prevalence of depression among vapers has been described. 34 Nicotine may have a specific function for some vapers (management of mood and psychiatric symptoms), and the need of these vapers for long-term nicotine use at sufficient dosage should be taken into account by users, manufacturers, retailers, clinicians, and regulators. Study Strengths and Limitations Strengths include originality, new insights, a large, international sample, and the longitudinal study design with several time points. However, this study was conducted in a self-selected sample of visitors of a smoking cessation website and of websites dedicated to e-cigarettes. We previously found that vapers enrolled on e-cigarette websites differed from those enrolled on neutral websites on several accounts, mainly smoking status and daily use of e-cigarettes. 12 We probably oversampled long-term users because they had more time to enroll than short-term users. Compared with representative samples of vapers, we oversampled former smokers and users of tank models However, patterns of e-cigarette use in this study were comparable to patterns observed in previous studies of vapers, 12,13,39 42 which suggests that participants Acknowledgments Vincent Baujard, from the HON Foundation, Geneva, Switzerland ( ch) developed the software for data collection. References 1. Malas M, van der Tempel J, Schwartz R, et al. Electronic Cigarettes for Smoking Cessation: A Systematic Review. Nicotine Tob Res. 2016;18(10): Etter JF, Bullen C. A longitudinal study of electronic cigarette users. Addict Behav. 2014;39(2): Zhuang YL, Cummins SE, Y Sun J, Zhu SH. Long-term e-cigarette use and smoking cessation: a longitudinal study with US population. Tob Control. 2016;25(suppl 1):i90 i Etter JF. A longitudinal study of cotinine in long-term daily users of e-cigarettes. Drug Alcohol Depend. 2016;160: Pacifici R, Pichini S, Graziano S, Pellegrini M, Massaro G, Beatrice F. Successful Nicotine Intake in Medical Assisted Use of E-Cigarettes: A Pilot Study. Int J Environ Res Public Health. 2015;12(7): Berg CJ, Barr DB, Stratton E, Escoffery C, Kegler M. Attitudes toward E-Cigarettes, Reasons for Initiating E-Cigarette Use, and Changes in Smoking Behavior after Initiation: A Pilot Longitudinal Study of Regular Cigarette Smokers. Open J Prev Med. 2014;4(10): Manzoli L, Flacco ME, Ferrante M, et al. Cohort study of electronic cigarette use: effectiveness and safety at 24 months. Tob Control. 2017;26(3): van Staden SR, Groenewald M, Engelbrecht R, Becker PJ, Hazelhurst LT. Carboxyhaemoglobin levels, health and lifestyle perceptions in smokers

11 922 Nicotine & Tobacco Research, 2018, Vol. 20, No. 8 converting from tobacco cigarettes to electronic cigarettes. S Afr Med J. 2013;103(11): Adriaens K, Van Gucht D, Declerck P, Baeyens F. Effectiveness of the electronic cigarette: An eight-week Flemish study with six-month follow-up on smoking reduction, craving and experienced benefits and complaints. Int J Environ Res Public Health. 2014;11(11): McRobbie H, Phillips A, Goniewicz ML, et al. Effects of Switching to Electronic Cigarettes with and without Concurrent Smoking on Exposure to Nicotine, Carbon Monoxide, and Acrolein. Cancer Prev Res (Phila). 2015;8(9): Kasza KA, Ambrose BK, Conway KP, et al. Tobacco-Product Use by Adults and Youths in the United States in 2013 and N Engl J Med. 2017;376(4): Etter JF, Bullen C. Electronic cigarette: users profile, utilization, satisfaction and perceived efficacy. Addiction. 2011;106(11): Etter JF. Electronic cigarettes: a survey of users. BMC Public Health. 2010;10: Etter J, Bullen C. Cotinine levels in users of electronic cigarettes. Eur Respir J. 2011;38(5): Douptcheva N, Etter JF, Gmel G. Use of electronic cigarettes among young Swiss men. J Epidemiol Community Health. 2013;67(12): Etter JF, Hughes JR. A comparison of the psychometric properties of three cigarette withdrawal scales. Addiction. 2006;101(3): Etter JF. Comparing the validity of the Cigarette Dependence Scale and the Fagerström Test for Nicotine Dependence. Drug Alcohol Depend. 2008;95(1-2): Etter JF, Le Houezec J, Perneger TV. A self-administered questionnaire to measure dependence on cigarettes: the cigarette dependence scale. Neuropsychopharmacology. 2003;28(2): Whooley MA, Avins AL, Miranda J, Browner WS. Case-finding instruments for depression. Two questions are as good as many. J Gen Intern Med. 1997;12(7): Bush K, Kivlahan DR, McDonell MB, Fihn SD, Bradley KA. The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol Use Disorders Identification Test. Arch Intern Med. 1998;158(16): Etter JF, Ussher M, Hughes JR. A test of proposed new tobacco withdrawal symptoms. Addiction. 2012;108(1): West R, Hajek P. Evaluation of the mood and physical symptoms scale (MPSS) to assess cigarette withdrawal. Psychopharmacology (Berl). 2004;177(1-2): Etter JF, Stapleton JA. Nicotine replacement therapy for long-term smoking cessation: a meta-analysis. Tob Control. 2006;15(4): Hughes JR, Keely J, Naud S. Shape of the relapse curve and long-term abstinence among untreated smokers. Addiction. 2004;99(1): Hartmann-Boyce J, McRobbie H, Bullen C, Begh R, Stead LF, Hajek P. Electronic cigarettes for smoking cessation. Cochrane Database Syst Rev. 2016;9:CD Perkins KA. Smoking cessation in women. Special considerations. CNS Drugs. 2001;15(5): Glover M, Breier BH, Bauld L. Could Vaping be a New Weapon in the Battle of the Bulge? [published online ahead of print October 25, 2016]. Nicotine Tob Res doi: /ntr/ntw Weitzman ER, Chen YY. The co-occurrence of smoking and drinking among young adults in college: national survey results from the United States. Drug Alcohol Depend. 2005;80(3): Agrawal A, Budney AJ, Lynskey MT. The co-occurring use and misuse of cannabis and tobacco: a review. Addiction. 2012;107(7): Healton CG, Vallone D, McCausland KL, Xiao H, Green MP. Smoking, obesity, and their co-occurrence in the United States: cross sectional analysis. BMJ. 2006;333(7557): Farrell M, Howes S, Bebbington P, et al. Nicotine, alcohol and drug dependence and psychiatric comorbidity. Results of a national household survey. Br J Psychiatry. 2001;179: Grant BF, Hasin DS, Chou SP, Stinson FS, Dawson DA. Nicotine dependence and psychiatric disorders in the United States: results from the national epidemiologic survey on alcohol and related conditions. Arch Gen Psychiatry. 2004;61(11): Lasser K, Boyd JW, Woolhandler S, Himmelstein DU, McCormick D, Bor DH. Smoking and mental illness: A population-based prevalence study. JAMA. 2000;284(20): Cummins SE, Zhu SH, Tedeschi GJ, Gamst AC, Myers MG. Use of e-cigarettes by individuals with mental health conditions. Tob Control. 2014;23:iii48 iii King BA, Alam S, Promoff G, Arrazola R, Dube SR. Awareness and everuse of electronic cigarettes among U.S. adults, Nicotine Tob Res. 2013;15(9): Adkison SE, O Connor RJ, Bansal-Travers M, et al. Electronic nicotine delivery systems: international tobacco control four-country survey. Am J Prev Med. 2013;44(3): Dockrell M, Morrison R, Bauld L, McNeill A. E-cigarettes: prevalence and attitudes in Great Britain. Nicotine Tob Res. 2013;15(10): Carroll Chapman SL, Wu LT. E-cigarette prevalence and correlates of use among adolescents versus adults: a review and comparison. J Psychiatr Res. 2014;54: Foulds J, Veldheer S, Berg A. Electronic cigarettes (e-cigs): views of aficionados and clinical/public health perspectives. Int J Clin Pract. 2011;65(10): Etter JF, Bullen C. Saliva cotinine levels in users of electronic cigarettes. Eur Respir J. 2011;38(5): Dawkins L, Turner J, Roberts A, Soar K. Vaping profiles and preferences: an online survey of electronic cigarette users. Addiction. 2013;108(6): Farsalinos KE, Romagna G, Tsiapras D, Kyrzopoulos S, Voudris V. Evaluating nicotine levels selection and patterns of electronic cigarette use in a group of vapers who had achieved complete substitution of smoking. Subst Abuse. 2013;7: