On 2017 Jul 26, Sally Satel commented:

Dartmouth demographer Samir Soneji and his co-authors find that the probability of cigarette smoking at follow-up is significantly higher among all e-cigarette users than among individuals who never used a nicotine product. Based on this finding, they conclude that “strong e-cigarette regulation” by the federal, state, and local governments are needed to minimize the potential “future population-level burden of tobacco.” This conclusion is unwarranted based on the nature of their results.

The article compares the probability of smoking in the post-period conditional on e-cigarette use without smoking to the probability of smoking in the post-period conditional on neither e-cigarette use nor smoking. This is not the relevant comparison for the purpose of assessing public-health risk. The relevant comparison is between smoking behavior conditional on access to e-cigarettes and smoking behavior conditional on no access to e-cigarettes, as such a comparison incorporates both the potential gateway and deterrent/diversion effects of e-cigarette use.

Such a comparison would take into account any beneficial effects of e-cigarettes on potential smokers who choose to reduce their cigarette smoking or to limit themselves to e-cigarette use altogether, as well as on smokers in the pre-period who switch to e-cigarettes partially or fully, or successfully use e-cigarettes as a cessation aid. It is on this comparison that regulatory choices should be based.

The nascent market for e-cigarettes in the United States can make robust empirical research on the consequences of these products on tobacco use challenging. The importance of a proper analytical framework is illustrated in a recent National Bureau of Economic Research working paper by economists Mike Pesko of Weill Cornell Medical College and Janet Currie of Princeton. The economists identify an important unintended consequence of minimum legal sale age laws restricting access to e-cigarettes: smoking among underage pregnant teenagers increased by more than 2 percentage points.

With teen smoking at a new low, policymakers should be celebrating a public health success instead of seeking a new regulatory expansion. Empirically, it is certainly not clear that more vaping has any causal effect on smoking among youth, as Soneji and his co-authors imply but do not demonstrate. Moreover, the type of analyses reported in JAMA Pediatrics fails to offer a reliable basis for developing an optimal regulatory framework for e-cigarettes and other modified risk tobacco products.

-Alex Brill, Sally Satel, Stan Veuger

NBER paper: http://www.nber.org/papers/w22792

On 2017 Jul 21, Samir Soneji commented:

We thank Joel Nitzkin for his interest in our article, which systematically reviewed 9 US-based longitudinal studies that assessed e-cigarette use and cigarette smoking among >17,000 adolescents and young adults. Our research concluded that e-cigarette use among adolescents who had never tried a cigarette was associated with subsequent cigarette smoking initiation and past 30-day cigarette smoking, with similar effect size across studies of adolescents and young adults. All of the studies used multivariable analysis to adjust for other factors that might make adolescent e-cigarette users at higher risk for use of multiple substances—risk factors such as friends who smoke, sensation seeking tendencies, and use of other substances like alcohol.

Nitzkin asserted three claims about the research. First, Nitzkin claimed that the studies provided no evidence that e-cigarette use is related to consistent daily cigarette smoking. Second, he claimed that e-cigarette use was simply a marker for high-risk youth who were more likely to smoke anyway. Third, he claimed that the decline in youth cigarette smoking over time at the population level proves e-cigarette use does not increase the probability of cigarette smoking at the individual level. Empirical evidence contradicts these claims, as we describe below.

Regarding Nitkin’s first claim that e-cigarette use is not related to consistent daily cigarette smoking, few adolescents smoke on a daily basis, which makes assessment of daily smoking impractical for most longitudinal studies that have a 1-2 year timeframe. Logically, smoking initiation is a necessary requisite to daily smoking. Moreover, recent longitudinal research found that smoking initiation identifies about two-thirds of adolescents who will be daily smokers two years later, with a false positive rate of 8 percent.1 In other words, smoking initiation is about as good at predicting eventual daily smoking as screening mammography is at predicting breast cancer.² Although not perfect, smoking initiation presents a public health concern especially given the growing body of evidence that e-cigarettes are used by some youth unlikely to have ever smoked cigarettes.^1,3,4 Furthermore, a recent longitudinal study by Leventhal et al. (2016) found that more frequent e-cigarette use at baseline was associated with more frequent and heavier patterns of cigarette smoking at follow-up using data from >3000 adolescents.⁵ Thus, smoking initiation, which the studies examined, is a sensible predictor of future daily smoking, and the pattern of e-cigarette use seems to predict the pattern of eventual cigarette smoking.

Regarding Nitzkin’s second claim that e-cigarette users are just high-risk youth, the combined risk estimate represents a risk that adjusts for many risk factors, as we mentioned above, that would cause some adolescents to be at risk for using multiple substances. The fact that the adjusted estimate is very strong (odds ratio of almost 4) suggests to us that it is unlikely that one or more added covariables would completely confound the e-cigarette effect. Moreover, several studies concluded that adolescents who use e-cigarettes are medium-risk youth, not those who are necessarily destined to begin cigarette smoking anyway.^6–10 Furthermore, several longitudinal studies have reported that the association between e-cigarette use and smoking initiation was strongest among the lowest risk youth (i.e., youth who stated that they were unlikely to try smoking in the future).^9,11,12

Regarding Nitzkin’s third claim that the recent decline in youth cigarette smoking proves e-cigarette use does not lead to cigarette use, youth cigarette smoking has been declining steadily in the US for the past 20 years and predates e-cigarettes.^13,14 In other words, this steady decline in youth cigarette smoking began long before the introduction of e-cigarettes into the US in 2007 and before e-cigarette use became prevalent in youth around 2011. So the decline in youth cigarette smoking cannot be attributed to the advent of the e-cigarette.

We believe our research underlines that the potential risks of e-cigarette use are significant and should not be discounted. Tobacco control efforts, including taxation, youth smoking prevention programs, and restrictions on tobacco advertising reduce youth smoking. The nearly twenty-year decline in youth smoking demonstrates the success of these tobacco control efforts despite youth e-cigarette use. We must acknowledge and address the public health harm posed by youth e-cigarette use to prevent a new generation of nicotine-addicted adult tobacco users.

References

¹ Sargent JD, Gabrielli J, Budney A, Soneji S, Wills TA. Adolescent smoking experimentation as a predictor of daily cigarette smoking. Drug Alcohol Depend. 2017;175:55-59. doi:10.1016/j.drugalcdep.2017.01.038.

² Ferrini R, Mannino E, Ramsdell E, Hill L. Screening mammography for breast cancer: American College of Preventive Medicine practice policy statement. Am J Prev Med. 1996;12(5):340-341.

³ Barrington-Trimis JL, Urman R, Leventhal AM, et al. E-cigarettes, Cigarettes, and the Prevalence of Adolescent Tobacco Use. Pediatrics. July 2016:e20153983. doi:10.1542/peds.2015-3983.

⁴ Dutra LM, Glantz SA. E-cigarettes and National Adolescent Cigarette Use: 2004–2014. Pediatrics. January 2017:e20162450. doi:10.1542/peds.2016-2450.

⁵ Leventhal AM, Stone MD, Andrabi N, et al. Association of e-Cigarette Vaping and Progression to Heavier Patterns of Cigarette Smoking. JAMA. 2016;316(18):1918-1920. doi:10.1001/jama.2016.14649.

⁶ Wills TA, Knight R, Williams RJ, Pagano I, Sargent JD. Risk Factors for Exclusive E-Cigarette Use and Dual E-Cigarette Use and Tobacco Use in Adolescents. Pediatrics. 2015;135(1):e43-e51. doi:10.1542/peds.2014-0760.

⁷ Kristjansson AL, Mann MJ, Sigfusdottir ID. Licit and Illicit Substance Use by Adolescent E-Cigarette Users Compared with Conventional Cigarette Smokers, Dual Users, and Nonusers. J Adolesc Health Off Publ Soc Adolesc Med. 2015;57(5):562-564. doi:10.1016/j.jadohealth.2015.07.014.

⁸ Thrasher JF, Abad-Vivero EN, Barrientos-Gutíerrez I, et al. Prevalence and Correlates of E-Cigarette Perceptions and Trial Among Early Adolescents in Mexico. J Adolesc Health Off Publ Soc Adolesc Med. 2016;58(3):358-365. doi:10.1016/j.jadohealth.2015.11.008.

⁹ Barrington-Trimis JL, Urman R, Berhane K, et al. E-Cigarettes and Future Cigarette Use. Pediatrics. June 2016:e20160379. doi:10.1542/peds.2016-0379.

¹⁰ Leventhal AM, Strong DR, Sussman S, et al. Psychiatric comorbidity in adolescent electronic and conventional cigarette use. J Psychiatr Res. 2016;73:71-78. doi:10.1016/j.jpsychires.2015.11.008.

¹¹ Primack BA, Soneji S, Stoolmiller M, Fine MJ, Sargent JD. Progression to traditional cigarette smoking after electronic cigarette use among US adolescents and young adults. JAMA Pediatr. September 2015:1-7. doi:10.1001/jamapediatrics.2015.1742.

¹² Wills TA, Knight R, Sargent JD, Gibbons FX, Pagano I, Williams RJ. Longitudinal study of e-cigarette use and onset of cigarette smoking among high school students in Hawaii. Tob Control. January 2016:1-6. doi:10.1136/tobaccocontrol-2015-052705.

¹³ Johnston L, O’Malley PM, Miech R, Emerson P, Bachman J, Schulenberg J. Monitoring the Future National Survey Results on Drug Use, 1975-2015: Overview, Key Findings on Adolescent Drug Use. Ann Arbor: Institute for Social Research, The University of Michigan; 2016.

¹⁴ Office on Smoking and Health. Trends in Current Cigarette Smoking. Centers for Disease Control and Prevention http://www.cdc.gov/tobacco/data_statistics/tables/trends/cig_smoking/. Accessed July 13, 2017.

On 2017 Jul 07, Joel Nitzkin commented:

This Soneji meta-analysis dealing with e-cigarettes and subsequent teen smoking,1 summarized data from nine studies, all of which share the same flaws, leading to an incorrect conclusion. These studies did not differentiate one-time or occasional use from consistent daily use. None compared smoking at follow-up in e-cigarette experimenters with kids who had experimented with or otherwise used cigarettes at baseline. Such a comparison would have reflected rates of smoking at follow-up in these same populations, had e-cigarettes not been available. Given these flaws, the only conclusion that can reasonably be drawn from the individual studies or this meta-analysis is that teens who are inclined to experiment with products disapproved by adult leadership are more likely to use both e-cigarettes and cigarettes than kids not prone to such experimentation. Neither the individual studies nor this meta-analysis give us reason to expect that reducing access to e-cigarettes or making them unattractive to potential users would reduce the numbers of teens recruited to nicotine addiction. The question as to whether e-cigarettes recruit American teens to nicotine addiction has already been answered. In June 2017, the Centers for Disease Control (CDC) published its 6th annual report showing use of tobacco-related products by high school students, by type or product, including e-cigarettes.2 During this period, e-cigarette use has gone from 1.5% of high school students in 2011 to 16.0% in 2015 and 11.3% in 2016, with significant reductions in cigarette use almost every year and no significant change in the percentage of high school students using any tobacco-related product. The data on middle school students reflects the same pattern, with much smaller numbers. If, as alleged by Soneji et al, e-cigarettes were attracting significant numbers of teens who otherwise would not have used tobacco products, there would have been significant year to year increases in the percent of teens using tobacco-related products. This did not occur. The fact that this has occurred year after year validates the impression that the teens attracted to e-cigarettes are those who would have used cigarettes, had e-cigarettes not been available. The time has come for public health authorities to consider the possibility that e-cigarettes, while not risk free, could be promoted for prevention of smoking and smoking cessation among teens inclined to smoke, without attracting yet more teens to nicotine experimentation. Joel L. Nitzkin, MD, MPH, DPA References 1. Soneji S, Barrington-Trimis JL, Wills TA et al. Association Between Initial Use of e-Cigarettes and Subsequent Cigarette Smoking Among Adolescents and Young Adults. JAMA Pediatrics. 2017 June 26:E1-E10. doi:10.1001/jamapediatrics.2017.1488 2. Jamal A, Getrzke A, Hu SS et al. Tobacco Use Among Middle and High School Students --- United States, 2011-2016. Morbidity and Mortality Weekly Report. 2017;66(23) (June 16):597-603. https://www.cdc.gov/mmwr/volumes/66/wr/pdfs/mm6623a1.pdf

On 2017 Jul 07, Joel Nitzkin commented:

This Soneji meta-analysis dealing with e-cigarettes and subsequent teen smoking,1 summarized data from nine studies, all of which share the same flaws, leading to an incorrect conclusion. These studies did not differentiate one-time or occasional use from consistent daily use. None compared smoking at follow-up in e-cigarette experimenters with kids who had experimented with or otherwise used cigarettes at baseline. Such a comparison would have reflected rates of smoking at follow-up in these same populations, had e-cigarettes not been available. Given these flaws, the only conclusion that can reasonably be drawn from the individual studies or this meta-analysis is that teens who are inclined to experiment with products disapproved by adult leadership are more likely to use both e-cigarettes and cigarettes than kids not prone to such experimentation. Neither the individual studies nor this meta-analysis give us reason to expect that reducing access to e-cigarettes or making them unattractive to potential users would reduce the numbers of teens recruited to nicotine addiction. The question as to whether e-cigarettes recruit American teens to nicotine addiction has already been answered. In June 2017, the Centers for Disease Control (CDC) published its 6th annual report showing use of tobacco-related products by high school students, by type or product, including e-cigarettes.2 During this period, e-cigarette use has gone from 1.5% of high school students in 2011 to 16.0% in 2015 and 11.3% in 2016, with significant reductions in cigarette use almost every year and no significant change in the percentage of high school students using any tobacco-related product. The data on middle school students reflects the same pattern, with much smaller numbers. If, as alleged by Soneji et al, e-cigarettes were attracting significant numbers of teens who otherwise would not have used tobacco products, there would have been significant year to year increases in the percent of teens using tobacco-related products. This did not occur. The fact that this has occurred year after year validates the impression that the teens attracted to e-cigarettes are those who would have used cigarettes, had e-cigarettes not been available. The time has come for public health authorities to consider the possibility that e-cigarettes, while not risk free, could be promoted for prevention of smoking and smoking cessation among teens inclined to smoke, without attracting yet more teens to nicotine experimentation. Joel L. Nitzkin, MD, MPH, DPA References 1. Soneji S, Barrington-Trimis JL, Wills TA et al. Association Between Initial Use of e-Cigarettes and Subsequent Cigarette Smoking Among Adolescents and Young Adults. JAMA Pediatrics. 2017 June 26:E1-E10. doi:10.1001/jamapediatrics.2017.1488 2. Jamal A, Getrzke A, Hu SS et al. Tobacco Use Among Middle and High School Students --- United States, 2011-2016. Morbidity and Mortality Weekly Report. 2017;66(23) (June 16):597-603. https://www.cdc.gov/mmwr/volumes/66/wr/pdfs/mm6623a1.pdf

On 2017 Jul 21, Samir Soneji commented:

We thank Joel Nitzkin for his interest in our article, which systematically reviewed 9 US-based longitudinal studies that assessed e-cigarette use and cigarette smoking among >17,000 adolescents and young adults. Our research concluded that e-cigarette use among adolescents who had never tried a cigarette was associated with subsequent cigarette smoking initiation and past 30-day cigarette smoking, with similar effect size across studies of adolescents and young adults. All of the studies used multivariable analysis to adjust for other factors that might make adolescent e-cigarette users at higher risk for use of multiple substances—risk factors such as friends who smoke, sensation seeking tendencies, and use of other substances like alcohol.

Nitzkin asserted three claims about the research. First, Nitzkin claimed that the studies provided no evidence that e-cigarette use is related to consistent daily cigarette smoking. Second, he claimed that e-cigarette use was simply a marker for high-risk youth who were more likely to smoke anyway. Third, he claimed that the decline in youth cigarette smoking over time at the population level proves e-cigarette use does not increase the probability of cigarette smoking at the individual level. Empirical evidence contradicts these claims, as we describe below.

Regarding Nitkin’s first claim that e-cigarette use is not related to consistent daily cigarette smoking, few adolescents smoke on a daily basis, which makes assessment of daily smoking impractical for most longitudinal studies that have a 1-2 year timeframe. Logically, smoking initiation is a necessary requisite to daily smoking. Moreover, recent longitudinal research found that smoking initiation identifies about two-thirds of adolescents who will be daily smokers two years later, with a false positive rate of 8 percent.1 In other words, smoking initiation is about as good at predicting eventual daily smoking as screening mammography is at predicting breast cancer.² Although not perfect, smoking initiation presents a public health concern especially given the growing body of evidence that e-cigarettes are used by some youth unlikely to have ever smoked cigarettes.^1,3,4 Furthermore, a recent longitudinal study by Leventhal et al. (2016) found that more frequent e-cigarette use at baseline was associated with more frequent and heavier patterns of cigarette smoking at follow-up using data from >3000 adolescents.⁵ Thus, smoking initiation, which the studies examined, is a sensible predictor of future daily smoking, and the pattern of e-cigarette use seems to predict the pattern of eventual cigarette smoking.

Regarding Nitzkin’s second claim that e-cigarette users are just high-risk youth, the combined risk estimate represents a risk that adjusts for many risk factors, as we mentioned above, that would cause some adolescents to be at risk for using multiple substances. The fact that the adjusted estimate is very strong (odds ratio of almost 4) suggests to us that it is unlikely that one or more added covariables would completely confound the e-cigarette effect. Moreover, several studies concluded that adolescents who use e-cigarettes are medium-risk youth, not those who are necessarily destined to begin cigarette smoking anyway.^6–10 Furthermore, several longitudinal studies have reported that the association between e-cigarette use and smoking initiation was strongest among the lowest risk youth (i.e., youth who stated that they were unlikely to try smoking in the future).^9,11,12

Regarding Nitzkin’s third claim that the recent decline in youth cigarette smoking proves e-cigarette use does not lead to cigarette use, youth cigarette smoking has been declining steadily in the US for the past 20 years and predates e-cigarettes.^13,14 In other words, this steady decline in youth cigarette smoking began long before the introduction of e-cigarettes into the US in 2007 and before e-cigarette use became prevalent in youth around 2011. So the decline in youth cigarette smoking cannot be attributed to the advent of the e-cigarette.

We believe our research underlines that the potential risks of e-cigarette use are significant and should not be discounted. Tobacco control efforts, including taxation, youth smoking prevention programs, and restrictions on tobacco advertising reduce youth smoking. The nearly twenty-year decline in youth smoking demonstrates the success of these tobacco control efforts despite youth e-cigarette use. We must acknowledge and address the public health harm posed by youth e-cigarette use to prevent a new generation of nicotine-addicted adult tobacco users.

References

¹ Sargent JD, Gabrielli J, Budney A, Soneji S, Wills TA. Adolescent smoking experimentation as a predictor of daily cigarette smoking. Drug Alcohol Depend. 2017;175:55-59. doi:10.1016/j.drugalcdep.2017.01.038.

² Ferrini R, Mannino E, Ramsdell E, Hill L. Screening mammography for breast cancer: American College of Preventive Medicine practice policy statement. Am J Prev Med. 1996;12(5):340-341.

³ Barrington-Trimis JL, Urman R, Leventhal AM, et al. E-cigarettes, Cigarettes, and the Prevalence of Adolescent Tobacco Use. Pediatrics. July 2016:e20153983. doi:10.1542/peds.2015-3983.

⁴ Dutra LM, Glantz SA. E-cigarettes and National Adolescent Cigarette Use: 2004–2014. Pediatrics. January 2017:e20162450. doi:10.1542/peds.2016-2450.

⁵ Leventhal AM, Stone MD, Andrabi N, et al. Association of e-Cigarette Vaping and Progression to Heavier Patterns of Cigarette Smoking. JAMA. 2016;316(18):1918-1920. doi:10.1001/jama.2016.14649.

⁶ Wills TA, Knight R, Williams RJ, Pagano I, Sargent JD. Risk Factors for Exclusive E-Cigarette Use and Dual E-Cigarette Use and Tobacco Use in Adolescents. Pediatrics. 2015;135(1):e43-e51. doi:10.1542/peds.2014-0760.

⁷ Kristjansson AL, Mann MJ, Sigfusdottir ID. Licit and Illicit Substance Use by Adolescent E-Cigarette Users Compared with Conventional Cigarette Smokers, Dual Users, and Nonusers. J Adolesc Health Off Publ Soc Adolesc Med. 2015;57(5):562-564. doi:10.1016/j.jadohealth.2015.07.014.

⁸ Thrasher JF, Abad-Vivero EN, Barrientos-Gutíerrez I, et al. Prevalence and Correlates of E-Cigarette Perceptions and Trial Among Early Adolescents in Mexico. J Adolesc Health Off Publ Soc Adolesc Med. 2016;58(3):358-365. doi:10.1016/j.jadohealth.2015.11.008.

⁹ Barrington-Trimis JL, Urman R, Berhane K, et al. E-Cigarettes and Future Cigarette Use. Pediatrics. June 2016:e20160379. doi:10.1542/peds.2016-0379.

¹⁰ Leventhal AM, Strong DR, Sussman S, et al. Psychiatric comorbidity in adolescent electronic and conventional cigarette use. J Psychiatr Res. 2016;73:71-78. doi:10.1016/j.jpsychires.2015.11.008.

¹¹ Primack BA, Soneji S, Stoolmiller M, Fine MJ, Sargent JD. Progression to traditional cigarette smoking after electronic cigarette use among US adolescents and young adults. JAMA Pediatr. September 2015:1-7. doi:10.1001/jamapediatrics.2015.1742.

¹² Wills TA, Knight R, Sargent JD, Gibbons FX, Pagano I, Williams RJ. Longitudinal study of e-cigarette use and onset of cigarette smoking among high school students in Hawaii. Tob Control. January 2016:1-6. doi:10.1136/tobaccocontrol-2015-052705.

¹³ Johnston L, O’Malley PM, Miech R, Emerson P, Bachman J, Schulenberg J. Monitoring the Future National Survey Results on Drug Use, 1975-2015: Overview, Key Findings on Adolescent Drug Use. Ann Arbor: Institute for Social Research, The University of Michigan; 2016.

¹⁴ Office on Smoking and Health. Trends in Current Cigarette Smoking. Centers for Disease Control and Prevention http://www.cdc.gov/tobacco/data_statistics/tables/trends/cig_smoking/. Accessed July 13, 2017.

On 2017 Jul 26, Sally Satel commented:

Dartmouth demographer Samir Soneji and his co-authors find that the probability of cigarette smoking at follow-up is significantly higher among all e-cigarette users than among individuals who never used a nicotine product. Based on this finding, they conclude that “strong e-cigarette regulation” by the federal, state, and local governments are needed to minimize the potential “future population-level burden of tobacco.” This conclusion is unwarranted based on the nature of their results.

The article compares the probability of smoking in the post-period conditional on e-cigarette use without smoking to the probability of smoking in the post-period conditional on neither e-cigarette use nor smoking. This is not the relevant comparison for the purpose of assessing public-health risk. The relevant comparison is between smoking behavior conditional on access to e-cigarettes and smoking behavior conditional on no access to e-cigarettes, as such a comparison incorporates both the potential gateway and deterrent/diversion effects of e-cigarette use.

Such a comparison would take into account any beneficial effects of e-cigarettes on potential smokers who choose to reduce their cigarette smoking or to limit themselves to e-cigarette use altogether, as well as on smokers in the pre-period who switch to e-cigarettes partially or fully, or successfully use e-cigarettes as a cessation aid. It is on this comparison that regulatory choices should be based.

The nascent market for e-cigarettes in the United States can make robust empirical research on the consequences of these products on tobacco use challenging. The importance of a proper analytical framework is illustrated in a recent National Bureau of Economic Research working paper by economists Mike Pesko of Weill Cornell Medical College and Janet Currie of Princeton. The economists identify an important unintended consequence of minimum legal sale age laws restricting access to e-cigarettes: smoking among underage pregnant teenagers increased by more than 2 percentage points.

With teen smoking at a new low, policymakers should be celebrating a public health success instead of seeking a new regulatory expansion. Empirically, it is certainly not clear that more vaping has any causal effect on smoking among youth, as Soneji and his co-authors imply but do not demonstrate. Moreover, the type of analyses reported in JAMA Pediatrics fails to offer a reliable basis for developing an optimal regulatory framework for e-cigarettes and other modified risk tobacco products.

-Alex Brill, Sally Satel, Stan Veuger

NBER paper: http://www.nber.org/papers/w22792