Epidemic of youth nicotine addiction? What does the National Youth Tobacco Survey reveal about high school e-cigarette use in the USA? (Preprint)

Background: In November 2018 the US Food and Drug Administration announced restrictions on e-cigarette manufacturers in response to a perceived epidemic of e-cigarette use among high school students. The stimulus was headline figures from the 2018 National Youth Tobacco Survey (NYTS). We analyse e-cigarette use and dependence in the NYTS in relation to lifetime history of use of tobacco products. Design and setting: Nationally representative survey of high school students in 2017 and 2018. Participants: 10,186 students in 2017 and 10,991 in 2018. Measurements: Any use of e-cigarettes in past 30 days, frequent e-cigarette use ( ≥ 20 of past 30 days) and indicators of dependence (craving in past 30 days; use within 30 minutes of waking and days used in lifetime) were analysed in relation to lifetime tobacco product use history, ranging from never use through to lifetime smoking of >100 cigarettes. Findings: Past-30-day e-cigarette use increased by 78% from 11.7% in 2017 to 20.8% in 2018. In both years, use was strongly associated with lifetime tobacco use history: it was seen in 8.4% of never tobacco users in 2018, in 29.0% of those who had tried a non-combustible, but never a combustible, product (OR 4.4 (CI 2.8-7.2) by comparison with never tobacco users), and in 71.0 % of those who had smoked more than 100 cigarettes in their lifetime (OR 26.8, CI 17.7-40.5) Frequent use occurred in 0.1% of never tobacco users in 2017 and 1.0% in 2018. Findings from 2014 and 2015 showed that first product tried was overwhelmingly cigarettes among those with a substantial lifetime cigarette history. Among past-30-day e-cigarette users who had never tried tobacco products in 2018, 3.8% reported craving, 3.1% reported wanting to use within 30 minutes of waking, and 61.8% said they had used e-cigarettes Conclusions: Data from the NYTS do not support claims of a new epidemic of nicotine addiction stemming from use of e-cigarettes, nor concerns that declines in youth tobacco addiction stand to be reversed after years of progress. Among current e-cigarette users who had never tried tobacco products, responses consistently pointed to minimal dependence.

e-cigarette use in 2018 among US high school students (1). Compared with 2017, a 78% increase in current (past-30-days) e-cigarette use was noted (1). At the same time, the FDA put out a press statement quoting HHS Secretary Alex Azar: "T hese new data show that America faces an epidemic of youth e-cigarette use, which threatens to engulf a new generation in nicotine addiction."(2) A statement from (then current, now former) FDA Commissioner Scott Gottlieb, issued on the same day, said: "T he data show that kids using e-cigarettes are going to be more likely to try combustible cigarettes later.
T his is a large pool of future risk. … T he data from this nationally representative survey….
show astonishing increases in kids' use of e-cigarettes and other ENDS , reversing years of favorable trends in our nation's fight to prevent youth addiction to tobacco products.
T hese data shock my conscience"(3). Gottlieb announced a series of regulatory initiatives to address the perceived problem of youth e-cigarette use. T he FDA moves were based on their analysis of preliminary data from the 2018 NYT S, received in September 2018. T heir publicly revealed analysis did not consider how e-cigarette use in high school students was related to patterns of use of combustible tobacco products, especially cigarettes.
In February 2019, a further report was published (4), giving more detailed tables, and giving data on frequent use (defined as on 20 or more days in the past 30 days). "During 2017-2018, frequent e-cigarette use increased significantly by 38.5% among current ecigarette users (from 20.0% to 27.7%); no significant change in frequent use was observed for other tobacco products". Commissioner Gottlieb issued a further statement on the same day, in which he continued to raise serious concerns: "T hese data are a sobering reminder of the initial results we shared last fall indicating a rampant rise of youth e-cigarette use, which has prompted the FDA to take a series of escalating regulatory actions ….. …. I will not allow a generation of children to become addicted to nicotine through e-cigarettes." "T his is an exploding epidemic. We'll continue to put the full scope of our regulatory tools against this mounting public health crisis.." (5) In the UK use of e-cigarettes by adolescents has not attracted such concern.
Commentators have noted that e-cigarette use is strongly associated with cigarette smoking in children and rare amongst never smokers (6)(7)(8). T he NYT S public use data file for the 2018 survey became available in March 2019. In this paper we analyse ecigarette use and indicators of tobacco dependence in the NYT S in relation to lifetime history of other tobacco products. We employ data from the 2014 and 2015 NYT S to examine which product was tried first in tobacco users' careers. We also report on time trends in prevalence of e-cigarette and combustible tobacco use in the USA. and ever-use (ever trying) are ascertained.
We derived a measure of lifetime tobacco product use history as follows: Respondents were asked "About how many cigarettes have you smoked in your entire life?" with 7 response options going from "I have never smoked cigarettes, not even one or two puffs" through "100 or more cigarettes (5 or more packs)". We added further categories for those who reported ever-use of e-cigarettes but no other tobacco product; a noncombustible tobacco product, but never any combustible; and use of a combustible product (cigars, pipes, bidi, hookah) but never cigarettes. T hus there were 10 response categories in all for lifetime tobacco use history, ranging from never use of any tobacco product through to smoking of more than 100 cigarettes.

Statistical analysis
We used the complex survey analysis module in SPSS 24 to adjust for the sampling design of the survey and to generate estimates (±95% confidence intervals) applicable to the US population. We examined the univariate association between current use of ecigarettes and our measure of lifetime tobacco product use history. We employed a logistic regression analysis to derive estimates of the odds ratios (±95% CIs) for current e-cigarette use. We conducted similar analyses to examine use of e-cigarettes for [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] days and 20 or more days within the past 30 days. We examined responses to other questions (e.g. "During the past 30 days, have you had a strong craving or felt like you really needed to use a tobacco product of any kind?"; and "How soon after you wake up do you want to use a tobacco product?" to provide an indication of dependence. We were comprised by "I do not want to use tobacco" (60.4%) and "I rarely want to use tobacco" (28.1%). 12.7% said they had used e-cigarettes on only one day in their life, and a further 49.1% had used them on between 1 and 10 days. T his pattern of responding contrasted markedly with that seen in e-cigarette users with a lifetime history of smoking more than 100 cigarettes: 74.5% of these reported experiencing craving; 51.4% wanted to use within 30 minutes of waking, and only 8.8% said '"I do not want to use tobacco" and 6.4% that they rarely wanted to use. T he pattern of responses to these dependence indicators by lifetime tobacco use history was similar in 2017 (T able 2, top).
Which product used first by tobacco use history T able 3 shows, by lifetime tobacco use history, which product respondents reported using first in their tobacco career. T he more cigarettes students reported having smoked in their lifetime, the higher the chances were that cigarettes were the first product used: in 2014, just over three quarters (75.6%) of current e-cigarette users who had smoked more than 100 cigarettes named cigarettes as the first tobacco product they had used, while just 2.2% had used e-cigarettes first. Among those who had smoked a few puffs but never a whole cigarette in their lifetime, 49.3% reported the cigarette as first product used and 18.8% had first used e-cigarettes. In 2015, 87.7% of current e-cigarette smokers who had smoked more than 100 cigarettes said cigarettes were the first product they had tried, while 7.6% identified e-cigarettes as the first product tried. Among those who had had just a few puffs of a cigarette in their lifetime, 53.1% had tried cigarettes first, and 25.4% had tried e-cigarettes first.
T rends in use of cigarettes, combustibles and e-cigarettes in NYT S 1999-2018 T here has been a continuing decline in current cigarette smoking in high school students, from 28.5% in 1999 to 8.1% in 2018 (see Figure). have used e-cigarettes in the past 30 days than students who had never tried any tobacco product. Use of e-cigarettes on 20 or more days in the past month was seen in only 1.0% of those who had never tried any tobacco product in 2018.
It is notoriously problematic to draw inferences about direction of causality from crosssectional data. In principle, the strong and graded association observed between likelihood of using e-cigarettes in the past 30 days and lifetime history of use of tobacco products could point to an effect of using e-cigarettes on subsequent uptake and use of cigarettes and other combustible products. T his appears to be the view adopted by the FDA. While it may well be the case that in some individual instances initial trying of an e- cigarette led on to trying and using cigarettes, the data strongly suggest that this is not the dominant pattern observed at the level of the whole population. Among high school students we found that, for the great majority of those with any substantial cigarette smoking history, cigarettes were the first tobacco product tried, prior to any use of ecigarettes. Clearly, for these students their use of cigarettes and the development of characteristic nicotine dependence must be attributed to cigarettes as the uptake product, rather than to e-cigarettes. Similarly, the observed rapid decline in trying combustible products and in the prevalence of cigarette smoking since 1999 has not yet given any sign of being reversed through the upsurge of e-cigarette use since 2011 (10). At the population level, therefore, the NYT S fails to give evidence of e-cigarettes acting as a gateway to smoking in adolescents. In these circumstances, there is plausibility to the suggestion that e-cigarettes are likely to reduce the disease burden in the US by helping adult smokers to quit (11).
We found little evidence of substantial nicotine addiction attributable to the use of ecigarettes. Among all students who were past-30-day-cigarette users but had never tried tobacco products, responses consistently pointed to minimal dependence with only 3.8% reporting any craving for tobacco products, and 3.1% reporting wanting to use within 30 minutes of waking. Over 60% reporting using a cigarette on 10 or fewer days in their lifetime. Only 3.4% were classified as frequent users of e-cigarettes on 20 or more days in the past month. T his contrasted markedly with students who had smoked more than 100 cigarettes (so meeting the US definition for regular cigarette smoking), where 74.5% reported craving, 51.4% wanted to use within 30 minutes of waking, 52.4% used e-cigarettes on 20 or more days in the past month, and 64.0% had used e-cigs on more than 100 days in their lifetime. T his group had mostly started their tobacco careers with cigarettes, and their pattern of dependence typifies that attributable to cigarette use.
In 2014, the FDA deemed e-cigarettes to be a tobacco product and so brought them within their regulatory competence. It appears that high school students in the US disagree with the FDA's classificatory scheme, as close to half of past-30-day e-cigarette users who had never tried any other tobacco products denied using a tobacco product in the past month. At the very least, this suggests their self-concept is not that of being a tobacco product user. T his perception of e-cigarettes as something different and separate from tobacco could serve to reduce their chances of progression to using conventional tobacco products.
In summary, data from the NYT S survey do not provide support for claims of a new epidemic of nicotine addiction stemming from use of e-cigarettes, nor for concerns that declines in youth tobacco addiction stand to be reversed after years of progress.
Symptoms of dependence were rare among students who use e-cigarettes without having used any other tobacco products.
T his paper is not intended as a challenge to the current direction of FDA policy on the regulation of e-cigarettes. T hat would be presumptuous on our part. Rather, we have sought to examine the evidence brought forward to support new regulatory initiatives.
We find a gaping chasm between the vision of an epidemic of e-cigarette use threatening to engulf a new generation in nicotine addiction and the reality of the evidence contained in the NYT S. As patterns of youth nicotine and tobacco use continue to evolve, careful surveillance of survey findings will remain of critical importance.

Funding
Cancer Research UK funded JB's salary (C1417/A22962). T he funders had no final role in the study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. All researchers listed as authors are independent from the funders and all final decisions about the research were