Associations of Risk Factors of E-cigarette and Cigarette Use and Susceptibility to Use Among Baseline PATH Study Youth Participants (2013-2014)

Abstract

Introduction:

Improved understanding of the relationships between ever use and susceptibility to e-cigarettes and cigarettes and risk factors of tobacco use can inform efforts to prevent youth tobacco use.

Methods:

Multiple logistic regression analysis was conducted using data from youth (ages 12-17 years) who had ever heard of e-cigarettes at baseline of the PATH Study (n=12,460) to evaluate associations between ever use of e-cigarettes, cigarettes, or susceptibility to either or both products, and known risk factors for tobacco use, adjusting for age, race/ethnicity, and gender.

Results:

Those susceptible to using e-cigarettes, cigarettes, or both had increased odds of risk factors for tobacco use compared to committed never users. Similarities and differences were observed among ever e-cigarette, cigarette, and dual users. Dual users had higher odds of nearly all risk factors (aOR range=1.6-6.8) compared to e-cigarette only users. Cigarette only smokers had higher odds of other tobacco use (aOR range=1.5-2.3), marijuana use (aOR=1.9, 95%CI=1.4-2.5), a high GAIN Substance Use score (aOR=1.9, 95%CI=1.1-3.4), low academic achievement (aOR range=1.6-3.4), and exposure to others smoking (aOR range=1.8-2.1) than e-cigarette only users. No group differences were observed for GAIN externalizing, sensation seeking, or household use of non-cigarette tobacco. The majority of past 30-day e-cigarette users had smoked cigarettes.

Conclusions:

We found similarities and differences among ever e-cigarette only, cigarette only, and dual users for risk factors of tobacco use. Understanding e-cigarette and cigarette users may inform youth tobacco use prevention efforts and inform future analyses assessing probability of progression to current tobacco use.

BACKGROUND

In the United States (U.S.), the prevalence of e-cigarette use among youth dramatically increased from 2011 to 2015 and then somewhat declined in 2016. Despite the drop, e-cigarettes remain the most commonly used tobacco product among youth.^1,2 Concurrently, youth cigarette smoking declined to historic lows.^1,2 The trend of increased e-cigarette use could indicate that youth at risk for tobacco use are replacing cigarettes with e-cigarettes;³ however, it is also possible that youth who otherwise might have never smoked conventional cigarettes are trying e-cigarettes.^4–7 A 2018 National Academies of Science, Engineering and Medicine Report summarized three hypotheses seeking to explain the impacts of youth e-cigarette use on cigarette smoking:⁸ 1) the diversion hypothesis suggests that high-risk youth with a proclivity for risk-taking behavior who might otherwise smoke cigarettes now use e-cigarettes instead; 2) the common liability hypothesis suggests that positive associations between e-cigarette use and cigarette smoking are due to shared risk factors; and 3) the catalyst hypothesis suggests that low-risk youth, who would otherwise not be susceptible to cigarette smoking, are drawn to e-cigarette use, which in turn increases proclivity for cigarette smoking.

One approach to examining whether e-cigarettes are attracting youth who might otherwise not use cigarettes is to explore risk factors traditionally associated with cigarette smoking among youth e-cigarette users.⁹ A study of youth susceptible to e-cigarettes suggests that psychological problems, rebelliousness, other substance use, and household exposure to smoking are determinants of e-cigarette use susceptibility.¹⁰ Additionally, studies have documented both similarities and differences in risk factors associated with e-cigarette and cigarette use.^4,11 E-cigarette use and cigarette smoking may be associated with similar psychosocial factors (e.g. use among friends, use among people at home, friends’ attitudes)^4,11; however, both e-cigarette and cigarette single product users may have fewer risk factors compared to those who report using both products (i.e., dual users).^4,11,12 For example, a study of 9^th and 10^th graders in Hawaii found e-cigarette only users were distinct from non-users, but similar to cigarette only smokers, although e-cigarette users consistently had lower risk status compared to dual users.^6,11

Using youth data from the Population Assessment of Tobacco and Health (PATH) Study, this study describes the distribution of traditional risk factors for cigarette smoking (e.g., substance use, academic achievement, household tobacco exposure), comparing similarities and differences among seven mutually exclusive groups based on cigarette smoking, e-cigarette use, and susceptibility status. Improved understanding of the relationships between ever use and susceptibility and risk factors for e-cigarette and cigarette use can inform efforts to prevent youth tobacco use. Lastly, to further describe youth e-cigarette users, we assessed lifetime history, frequency of e-cigarette use, and reasons for use among the subset of ever youth users who reported past 30-day e-cigarette use at study baseline.

METHODS

The current analysis utilizes data from 12,460 youth (ages 12-17 years) who had ever heard of e-cigarettes out of a total of 13,651 PATH Study Wave 1 (September 2013-December 2014) youth participants. Detailed sampling and study methodology have been published elsewhere.^13,14 The PATH Study is a nationally representative household-based longitudinal study sponsored by the National Institute of Health’s National Institute of Drug Abuse and the Food and Drug Administration’s Center for Tobacco Products and implemented by Westat. The PATH Study utilizes Audio-Computer Assisted Self-Interviews (ACASI) available in English and Spanish to collect information on tobacco use patterns and associated behaviors on youth and parent use of tobacco products, and provides images, text descriptors, and brand examples to aid in product recognition. The weighted response rate for the household screener was 54%, and among households completing the screener, the youth weighted response rate in Wave 1 was 78.4%. Weighting procedures further adjust for nonresponse. Parents/guardians provided consent and youth assented to study participation. Westat’s Institutional Review Board approved the study design and protocol and the Office of Management and Budget approved the data collection.

Measures

E-cigarette susceptibility and use

Youth respondents were asked whether they had “ever seen or heard of an electronic cigarette or an e-cigarette before this study.” Those who had heard of e-cigarettes were asked if they had ever used e-cigarettes, even one or two times. Susceptibility to future use was assessed among never users who reported having heard of the product using a three-item enhanced susceptibility scale:^15,16 “Have you ever been curious about using e-cigarettes?” (response options: very curious; somewhat curious; a little curious; not at all curious); “Do you think you will try an e-cigarette soon?” and “If one of your best friends were to offer you an e-cigarette, would you use it?” (response options for both: definitely yes; probably yes; probably not; definitely not). Responses signifying the strongest rejection (not at all curious/definitely not) on all three measures were considered committed never users; respondents with a combination of missing information and responses other than the strongest rejection of the three susceptibility responses were classified as susceptible. Respondents with a combination of the strongest rejection plus missing information were classified as missing.

Cigarette smoking susceptibility and use

All respondents were asked whether they had ever smoked a cigarette, even one or two times. Susceptibility to future cigarette smoking was assessed among never smokers utilizing three items. Two of the items were identical to those asked for e-cigarettes: “Have you ever been curious about using cigarettes?” and “If one of your best friends were to offer you a cigarette, would you use it?”¹⁶ The third question for cigarettes was: “Do you think you will smoke a cigarette in the next year? (response options: definitely yes; probably yes; probably not; definitely not).¹⁶ Responses signifying other than the strongest rejection (not at all curious/definitely not) on all three measures were considered susceptible to future cigarette smoking; missing responses were handled similarly to e-cigarette susceptibility.

Definitions of e-cigarette and cigarette susceptibility and ever use

Based on the criteria above, respondents were classified into one of seven mutually exclusive groups: 1) not susceptible to and have never tried e-cigarettes or cigarettes (committed never users), 2) susceptible to e-cigarettes, but not cigarettes (susceptible to e-cigarettes only), 3) susceptible to cigarettes, but not e-cigarettes (susceptible to cigarettes only), 4) susceptible to both e-cigarettes and cigarettes (susceptible to e-cigarettes and cigarettes), 5) ever e-cigarette users who have never tried cigarettes (e-cigarette only users), 6) ever cigarette smokers who have never tried e-cigarettes (cigarette only s), and 7) dual e-cigarette and cigarette ever users/smokers (dual users). Other tobacco use was not taken into consideration in creating these groups.

Past 30-day e-cigarette and cigarette use

To further characterize e-cigarette use among youth at baseline, patterns of e-cigarette use among those who had ever heard of e-cigarettes and reported use in the past 30-days were explored. Detailed patterns of e-cigarette use including duration of use, frequency of use, use of flavors, and past 30-day use of other tobacco products (traditional cigars, cigarillos, filtered cigars, snus pouches, smokeless, hookah, pipe, dissolvables, bidis, or kreteks) were compared among past 30-day exclusive e-cigarette only users, past 30-day e-cigarette and cigarette users (who may also have used other products), and past 30 day e-cigarette and at least one other (non-cigarette) tobacco product users. Additionally, past 30-day e-cigarette users who reported having used a lifetime equivalent of at least one disposable e-cigarette or cartridge and had used an e-cigarette in the past 30 days were asked about their reasons for use, stratified by smoking status.

Known risk factors for tobacco use

Covariates for analysis were selected based on known risk factors for youth tobacco use.^8,17 Demographic variables included age, gender, and race/ethnicity. Missing data on demographics and education were imputed as described in the PATH Study Restricted Use Files User Guide.¹⁴ Exposure to tobacco use was explored by assessing cigarette smoking and tobacco use among household members, as well as any exposure to others smoking within the past seven days. Ever use of marijuana and alcohol were assessed. Severity of substance abuse symptoms, internalizing, and externalizing disorders were assessed using subscales of the Global Appraisal of Individual Needs Short Screener (GAIN-SS).¹⁸ Problems experienced within the past year were tallied and dichotomized by severity, with four or more problems reported in the past year categorized as a high substance use, internalizing, and externalizing disorder score. Sensation-seeking was assessed via three modified items from the Brief Sensation Seeking Scale: 1) “I like to do frightening things”, 2) “I like new and exciting experiences even if I have to break the rules”, and 3) “I prefer friends who are exciting and unpredictable”.¹⁹ Response options ranged from strongly disagree (0) to strongly agree (4) and were summed to create an overall score (range: 0-12), which was dichotomized at the upper quartile for analysis (score ≥ 6 indicated high sensation seeking). The Brief Sensation Seeking Scale was found to be internally consistent in the PATH Study (Cronbach’s α=0.76).²⁰ Parent-reported past-year academic achievement was categorized into four grade levels: mostly A’s or A’s and B’s; mostly B’s or B’s and C’s; mostly C’s or C’s and D’s; or lower.

Statistical Analysis

Analyses were conducted using SAS 9.4, utilizing replicate weights and balanced repeated replication to account for the PATH Study’s complex survey design. Descriptive statistics were used to explore the distribution of known risk factors by susceptibility and ever use of cigarettes and e-cigarettes. Multinomial logistic regression, adjusted for age, race, ethnicity, and gender, was used to evaluate associations between known risk factors for tobacco use and susceptibility to or ever use of e-cigarettes and cigarettes. In the first model, those susceptible to e-cigarettes only, cigarettes only, or both products were compared to committed never users. In the second model, committed never users, cigarette only smokers, and dual users were compared to e-cigarette only users. Lastly, characteristics and patterns of e-cigarette use in the past 30-days, in addition to reasons for use, were examined by current tobacco use status.

RESULTS

Among all youth at PATH Study Wave 1, 91.4% (n=12,460) had heard of e-cigarettes (all percentages are weighted; all n’s are unweighted). Of these, 81.4% were non-users of e-cigarettes or cigarettes, with 46.0% committed never users (n=5,701), 7.2% susceptible to e-cigarettes (n=894), 10.1% susceptible to cigarettes (n=1,263), and 18.1% susceptible to both e-cigarettes and cigarettes (n=2,276). Additionally, 18.5% of youth were ever users of either product, with 3.9% e-cigarette only users (n=488), 6.9% cigarette only smokers (n=875), and 7.7% dual users (n=963). Detailed information on the distribution of demographic, behavioral, and psychosocial characteristics is presented in Supplemental Table 1.

Risk factors and susceptibility to e-cigarettes, cigarettes, or both products

Committed never users had the lowest percentage of youth reporting all risk factors (Supplemental Table 1). Compared to committed never users, those susceptible to e-cigarettes only, cigarettes only, or both products had significantly higher odds of each tobacco use risk factor assessed after controlling for age, race and gender (Table 1). In many instances, the magnitude of the observed associations was similar between those susceptible to e-cigarettes only and cigarettes only compared to committed never users, and larger for those susceptible to both products compared to those susceptible to one (Table 1). For example, compared to committed never users, those susceptible to either e-cigarettes or cigarettes had approximately two and a half times the odds of ever trying alcohol (aOR e-cigarettes = 2.5, 95% CI= 2.0-3.1) (aOR cigarettes = 2.6, 95% CI= 2.2-3.1), whereas those susceptible to both had approximately four times the odds of ever trying alcohol (aOR= 4.1, 95% CI= 3.4-4.9) compared with committed never users.

Table 1.

Multinomial logistic regression of demographic and psychosocial correlates related to tobacco use by susceptibility to cigarettes and e-cigarettes–PATH Study Wave 1 Data (2013-2014)¹

	Non-Users (N=10,134)
	Committed Never Users: Cigarettes and E-cigarettes2 n=5,701		Susceptible to E-cigarettes3 n=894		Susceptible to Cigarettes3 n=1,263		Susceptible to E-cigarettes + Cigarettes3 n=2,276
	n (%) [95% CI]		n (%) [95% CI]	aOR [95% CI]3	n (%) [95% CI]	aOR [95% CI]3	n (%) [95% CI]	aOR [95% CI]3
Other Tobacco Use
Ever non-cigarette combustible tobacco use4	117 (2.1%) [1.8-2.6]	REF	69 (8.7%) [6.8-11.1]	3.7 (2.7-5.2)**	72 (6.5%) [5.1-8.2]	2.9 (2.1-4.1)**	160 (7.5%) [6.3-8.8]	3.1 (2.5-3.9)**
Ever smokeless tobacco use5 (excluding E-cigarettes)	47 (0.9%) [0.6-1.2]	REF	—¶	1.6 (0.7-3.3)	15 (1.4%) [0.8-2.2]	1.8 (1.0-3.1)*	62 (3.1%) [2.3-4.0]	3.4 (2.3-5.1)**
Other Substance Use
Ever use of marijuana6	152 (2.8%) [2.3-3.3]	REF	73 (9.0%) [7.3-11.2]	3.0 (2.2-4.0)**	85 (7.3%) [5.8-9.2]	2.6 (1.9-3.5)**	212 (10.1%) [8.9-11.5]	3.3 (2.6-4.3)**
Ever alcohol use7	437 (8.4%) [7.5-9.3]	REF	150 (19.2%) [16.5-22.4]	2.5 (2.0-3.1)**	212 (18.9%) [16.7-21.2]	2.6 (2.2-3.1)**	572 (28.1%) [25.9-30.4]	4.1 (3.4-4.9)**
Psychosocial Factors
High GAIN Substance Use Scale score8	13 (0.2%) [0.1-0.4]	REF	—¶	3.2 (1.0-10.5)	—¶	2.9 (1.0-8.0)*	23 (1.1%) [0.7-1.6]	4.3 (2.0-8.9)**
High GAIN Internalizing Scale score9	794 (15.4%) [14.2-16.7]	REF	177 (21.7%) [18.7-24.9]	1.5 (1.3-1.9)**	252 (22.7%) [20.0-25.6]	1.5 (1.3-1.9)**	673 (33.6%) [31.3-35.9]	2.8 (2.4-3.2)**
High GAIN Externalizing Scale score10	1,186 (23.3%) [22.0-24.6]	REF	299 (37.0%) [33.7-40.4]	2.0 (1.7-2.3)**	414 (37.6%) [34.7-40.7]	2.0 (1.7-2.3)**	957 (48.9%) [46.7-51.1]	3.2 (2.9-3.6)**
High Sensation Seeking score11	1,523 (29.2%) [27.8-30.8]	REF	366 (45.1%) [41.3-49.0]	1.9 (1.6-2.3)**	516 (44.5%) [41.4-47.6]	2.0 (1.7-2.2)**	1,179 (57.7%) [55.5-59.9]	3.2 (2.8-3.6)**
Academic Achievement12
Mostly B’s or B’s and C’s compared to Mostly A’s or A’s and B’s	1,304 (23.7%) [22.4-25.1]	REF	242 (28.8%) [25.8-32.0]	1.3 (1.1-1.5)**	306 (25.3%) [22.7-28.0]	1.1 (0.9-1.4)	603 (28.8%) [26.5-31.3]	1.4 (1.2-1.6)**
Mostly C’s or C’s and D’s compared to Mostly A’s or A’s and B’s	379 (6.6%) [5.9-7.5]	REF	80 (9.2%) [7.4-11.4]	1.5 (1.1-2.0)*	107 (8.7%) [7.2-10.5]	1.4 (1.1-1.8)*	224 (10.1%) [8.9-11.5]	1.8 (1.5-2.2)**
Mostly D’s or D’s and F’s or mostly F’s compared to Mostly A’s or A’s and B’s	71 (1.2%) [0.9-1.7]	REF	16 (1.8%) [1.1-3.0]	1.5 (0.8-2.8)	23 (1.8%) [1.2-2.8]	1.6 (1.0-2.6)	49 (2.3%) [1.7-3.0]	2.0 (1.3-3.1)*
Household Tobacco Use
Household member smokes cigarettes (in addition to other forms of tobacco)13 compared to no household use of any form of tobacco	1,297 (22.9%) [21.1-24.9]	REF	239 (27.8%) [24.9-31.0]	1.4 (1.2-1.6)**	319 (25.8%) [23.1-28.8]	1.3 (1.1-1.6)*	596 (27.5%) [24.7-30.5]	1.4 (1.2-1.6)**
Household member uses some other form of tobacco (without cigarettes)13 compared to no household use of any form of tobacco	317 (6.3%) [5.5-7.2]	REF	62 (8.0%) [6.3-10.1]	1.5 (1.1-1.9)*	85 (7.6%) [5.9-9.6]	1.4 (1.0-1.9)	181 (9.1%) [7.7-10.7]	1.7 (1.3-2.1)**
Any exposure to others smoking within the past 7 days14	1,608 (30.2%) [28.7-31.8]	REF	349 (43.8%) [40.2-47.6]	1.8 (1.5-2.1)**	447 (37.5%) [34.6-40.6]	1.4 (1.2-1.6)**	966 (47.6%) [45.3-49.9]	2.1 (1.9-2.4)**

Estimates in bold are statistically significantly different, * indicates p<0.05 and ** indicates p<0.001;

^¶Estimate not presented because relative standard error ≥30% or denominator <50.

¹Covariates included in the multinomial logistic models include age, treated as a continuous variable; gender, treated dichotomously (male or female); race, categorized as white race alone, black race alone, Asian race alone, or other race, including multiracial; and Hispanic ethnicity, categorized as Hispanic or non-Hispanic.

²A total of 353 committed never users of cigarettes and e-cigarettes had missing covariates (i.e., age, gender, race, or ethnicity) and were dropped from the model.

³A total of 70 susceptible to e-cigarettes, 103 susceptible to cigarettes, and 205 susceptible to both cigarettes and e-cigarette users had missing covariates and were dropped from the model. Susceptibility to products other than e-cigarettes or cigarettes was not assessed.

⁴Youth who reported ever smoking a traditional cigar, cigarillo, or filtered cigar, even one or two times; ever smoking a pipe filled with tobacco, even one or two puffs; ever smoking tobacco in a hookah, even one or two puffs; or having tried a bidi or kretek, even one or two times were classified as ever non-cigarette combustible product users. A combination of missing data and “no’s” to any of the products included in this variable were counted as missing (n=10).

⁵Youth who reported ever having used snus pouches, loose snus, moist snuff, dip, spit or chewing tobacco, or dissolvable tobacco products (such as Ariva, Stonewall or Camel Orbs, Sticks or Strips), even one or two times were classified as ever smokeless tobacco product users. A combination of missing data and “no’s” to any of the products included in this variable (i.e., YS1002_01, YS1002_02, YU1003, and YD1002) were counted as missing (n=89).

⁶Youth were asked “Have you ever used marijuana, hash, THC, grass, pot or weed? (Yes/No)”, unless they had previously reported having ever smoked part or all of a cigar, cigarillo, or filtered cigar with marijuana in it. Affirmative responses to either question classified as an ever marijuana user (n=35 missing).

⁷All youth were asked “Have you ever used alcohol at all, including sips of someone’s drink or your own drink?” (Yes/No). Those that answered affirmatively were asked “About how old were you when you had your first alcoholic drink, other than small tastes or sips?” Those indicating that “I have never had an alcoholic drink other than small tastes or sips” were reclassified as never users for the purpose of this analysis. Those who did not know or refused to answer: 1) if they had ever used alcohol at all, or 2) the age when they first had an alcoholic drink were excluded from this variable (n=49 missing).

⁸The GAIN-SS Substance Use subscale consists of seven items: (1) used alcohol or other drugs weekly or more often; (2) spent a lot of time getting alcohol or other drugs; (3) spent a lot of time using or recovering from alcohol or other drugs; (4) kept using alcohol or other drugs, even though it was causing social problems, leading to fights, or getting into trouble with other people; (5) use of alcohol or other drugs caused reduced involvement in activities at work, school, home, or social events; (6) withdrawal problems; and (7) use of alcohol or other drugs to stop being sick or avoid withdrawal problems. Youth were asked to report whether they had experienced each problem: within the past month; 2-12 months ago; over a year ago; never; or don’t know/refused. Problems experienced within the past year were tallied and dichotomized by severity, with four or more problems reported in the past year categorized as a high score (n=253 missing; includes those missing responses for any of the scale items).

⁹The GAIN-SS Internalizing subscale consists of four items, asking about the last time you had significant problems with: (1) feeling very trapped, lonely, sad, blue, depressed, or hopeless about the future; (2) sleep trouble, such as bad dreams, sleeping restlessly, or falling asleep during the day; (3) feeling very anxious, nervous, tense, scared, panicked, or like something bad was going to happen; and (4) becoming very distressed and upset when something reminded you of the past. Youth were asked to report whether they had experienced each problem: within the past month; 2-12 months ago; over a year ago; never; or don’t know/refused. Problems experienced within the past year were tallied and dichotomized by severity, with four problems reported in the past year categorized as a high score (n=206 missing; includes those missing responses for any of the scale items).

¹⁰The GAIN-SS Externalizing subscale consists of seven items, asking about the last time you did the following two or more times: (1) had a hard time paying attention at school, work, or home; (2) had a hard time listening to instructions at school, work, or home; (3) lied or conned to get things you wanted or to avoid having to do something; (4) were a bully or threatened other people; (5) started physical fights with other people; (6) felt restless or the need to run around or climb on things; and (7) gave answers before the other person finished asking the question. Youth were asked to report whether they had experienced each problem: within the past month; 2-12 months ago; over a year ago; never; or don’t know/refused. Problems experienced within the past year were tallied and dichotomized by severity, with four or more problems reported in the past year categorized as a high score (n=372 missing; includes those missing responses for any of the scale items).

¹¹Youth were asked to indicate whether they agreed or disagreed with the following three measures regarding sensation seeking: (1) I like to do frightening things; (2) I like new and exciting experiences, even if I have to break the rules; and (3) I prefer friends who are exciting and unpredictable. Response options for each item were strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree. Responses were scored according to strength of agreement (4, 3, 2, 1, 0), and then summed to create an overall score. Based on the overall distribution of scores, the approximate upper quartile (75%) of six or higher was selected as the cut-off for high levels of sensation seeking. Scores were then dichotomized to indicate high (≥6) versus low (<6) levels of sensation seeking (n=152 missing; includes those missing responses for any of the scale items).

¹²Academic achievement: n=83 missing (including n=38 where school was ungraded). This item was reported by parents in the parent interview.

¹³All youth were asked “Does anyone who lives with you now do any of the following? Choose all that apply: 1) Smoke cigarettes; 2) Use smokeless tobacco, such as chewing tobacco, snuff, dip, or snus; 3) Smoke cigars, cigarillos, or filtered cigars; 4) Use any other form of tobacco; 5) No one who lives with me now uses any form of tobacco; 6) Don’t know/refused. Responses were categorized as no household use of any form of tobacco (option 5), household member smokes cigarettes (in addition to other forms of tobacco) (option 1), or household member uses some other form of tobacco (without cigarettes) (options 2, 3, or 4, and not 1) (n=66 missing).

¹⁴All youth were asked “During the past seven days, about how many hours were you around others who were smoking [whether or not you were smoking yourself]? Include time in your home, in a car, at school, or outdoors.” All responding with a value of >0 hours were classified as exposed to others smoking in the past seven days (n=251 missing).

Risk factors and ever use of e-cigarettes, cigarettes, or both products

Compared to e-cigarette only users, committed never users had lower odds of each tobacco use risk factor after controlling for age, race, ethnicity and gender (Table 2). Between the ever use groups, differences and similarities were observed and, for some factors, a risk gradient emerged. For example, compared to ever e-cigarette only users, ever cigarette only smokers had twice the odds of ever using marijuana (aOR= 1.9, 95% CI=1.4-2.5); the association was even stronger for dual users (aOR=5.5, 95% CI=4.0-7.5). Similar results were seen for the GAIN Substance Use scale; compared to ever e-cigarette only users, ever cigarette only smokers (aOR= 1.9, 95% CI=1.1-3.4) and dual users (aOR= 5.7, 95% CI=3.5-9.3) had increased odds of having high scores.

Table 2.

Multinomial logistic regression of demographic and psychosocial correlates related to tobacco use by ever use of cigarettes and e-cigarettes–PATH Study Wave 1 Data (2013-2014)¹

			Ever Product Users (n=2,326)
	Committed Never Users: Cigarettes and E-cigarettes2 (n=5,701)		Ever E-cigarette Only Users (no Cigarette Smoking)3 (n=488)		Ever Cigarette Only Smoking (No E-cigarette Use)3 (n=875)		Ever Dual E-cigarette and Cigarette Use3 (n=963)
	n (%) [95% CI]	aOR [95% CI]	n (%) [95% CI]		n (%) [95% CI])	aOR [95% CI]	n (%) [95% CI])	aOR [95% CI]
Other Tobacco Use
Ever non-cigarette combustible tobacco use4	117 (2.1%) [1.8-2.6]	0.06 (0.05-0.08)**	148 (32.9%) [28.1-38.0]	REF	340 (41.8%) [37.9-45.7]	1.5 (1.1-2.0)*	670 (72.6%) [69.4-75.6]	5.4 (4.0-7.2)**
Ever smokeless tobacco use5 (excluding E-cigarettes)	47 (0.9%) [0.6-1.2]	0.10 (0.06-0.15)**	39 (9.3%) [6.8-12.6]	REF	135 (16.7%) [14.3-19.4]	2.3 (1.5-3.3)**	310 (34.5%) [30.5-38.8]	4.9 (3.2-7.5)**
Other Substance Use
Ever use of marijuana6	152 (2.8%) [2.3-3.3]	0.08 (0.06-0.11)**	142 (32.0%) [27.9-36.5]	REF	394 (47.6%) [43.3-52.0]	1.9 (1.4-2.5)**	656 (71.4%) [67.9-74.7]	5.5 (4.0-7.5)**
Ever alcohol use7	437 (8.4%) [7.5-9.3]	0.10 (0.08-0.12)**	233 (52.7%) [48.0-57.4]	REF	462 (56.6%) [52.7-60.5]	1.0 (0.8-1.4)	643 (71.3%) [68.2-74.3]	1.8 (1.4-2.4)**
Psychosocial Factors
High GAIN Substance Use Scale score8	13 (0.2%) [0.1-0.4]	0.06 (0.02-0.12)**	20 (4.1%) [2.6-6.4]	REF	59 (7.8%) [5.9-10.1]	1.9 (1.1-3.4)*	168 (18.9%) [16.1-22.0]	5.7 (3.5-9.3)**
High GAIN Internalizing Scale score9	794 (15.4%) [14.2-16.7]	0.39 (0.31-0.49)**	141 (30.9%) [26.8-35.3]	REF	315 (39.0%) [35.2-43.0]	1.3 (1.0-1.7)	375 (41.6%) [38.2-45.0]	1.6 (1.2-2.0)**
High GAIN Externalizing Scale score10	1,186 (23.3%) [22.0-24.6]	0.32 (0.26-0.40)**	212 (47.7%) [42.6-52.8	REF	373 (47.0%) [43.6-50.5]	1.0 (0.8-1.2)	457 (51.7%) [48.4-55.0]	1.2 (1.0-1.6)
High Sensation Seeking score11	1,523 (29.2%) [27.8-30.8]	0.24 (0.19-0.30)**	296 (66.0%) [60.9-70.6]	REF	484 (60.1%) [55.9-64.2]	0.8 (0.6-1.0)	637 (70.8%) [67.7-73.6]	1.2 (0.9-1.5)
Academic Achievement12
Mostly B’s or B’s and C’s compared to Mostly A’s or A’s and B’s	1,304 (23.7%) [22.4-25.1]	0.6 (0.5-0.8)**	153 (31.9%) [27.6-36.6]	REF	280 (35.6%) [32.2-39.2]	1.6 (1.2-2.1)*	326 (36.8%) [33.5-40.4]	2.1 (1.5-2.8)**
Mostly C’s or C’s and D’s compared to Mostly A’s or A’s and B’s	379 (6.6%) [5.9-7.5]	0.5 (0.3-0.7)**	55 (11.8%) [9.1-15.1]	REF	145 (17.9%) [15.4-20.7]	2.4 (1.6-3.5)**	191 (20.9%) [17.7-24.4]	3.4 (2.2-5.1)**
Mostly D’s or D’s and F’s or mostly F’s compared to Mostly A’s or A’s and B’s	71 (1.2%) [0.9-1.7]	0.5 (0.2-1.0)*	—¶	REF	44 (4.9%) [3.6-6.6]	3.4 [1.5-7.5]*	76 (8.2%) [6.6-10.1]	6.8 [3.2-14.1]**
Household Tobacco Use
Household member smokes cigarettes (in addition to other forms of tobacco)13 compared to no household use of any form of tobacco	1,297 (22.9%) [21,1-24.9]	0.5 (0.4-0.7)**	154 (31.3%) [26.5-36.4]	REF	412 (48.9%) [44.6-53.1]	2.1 (1.7-2.8)**	508 (53.8%) [49.8-57.8]	2.8 (2.0-3.8)**
Household member uses some other form of tobacco (without cigarettes)13 compared to no household use of any form of tobacco	317 (6.3%) [5.5-7.2]	0.4 (0.3-0.5)**	54 (12.8%) [9.6-16.8]	REF	82 (10.4%) [8.5-12.6]	1.1 (0.7-1.6)	75 (8.4%) [6.8-10.4]	0.9 (0.6-1.4)
Any exposure to others smoking within the past 7 days14	1,608 (30.2%) [28.7-31.8]	0.4 (0.3-0.4)**	253 (54.3%) [49.7-58.9]	REF	551 (68.1%) [64.8-71.3]	1.8 (1.4-2.2)**	739 (82.8%) [79.9-85.4]	4.2 (3.2-5.4)**

Estimates in bold are statistically significantly different, * indicates p<0.05 and ** indicates p<0.001;

^¶Estimate not presented because relative standard error ≥30% or denominator <50.

¹Covariates included in the multinomial logistic models include age, treated as a continuous variable; gender, treated dichotomously (male or female); race, categorized as white race alone, black race alone, Asian race alone, or other race, including multiracial; and Hispanic ethnicity, categorized as Hispanic or non-Hispanic.

²A total of 353 committed never users of cigarettes and e-cigarettes had missing covariates (i.e., age, gender, race, or ethnicity) and were dropped from the model.

³A total of 33 ever e-cigarette, no cigarette users, 50 ever cigarette, no e-cigarette users, and 45 ever cigarette and e-cigarette users had missing covariates and were dropped from the model. Use of other tobacco products was not assessed in these groups.

⁴Youth who reported ever smoking a traditional cigar, cigarillo, or filtered cigar, even one or two times; ever smoking a pipe filled with tobacco, even one or two puffs; ever smoking tobacco in a hookah, even one or two puffs; or having tried a bidi or kretek, even one or two times were classified as ever non-cigarette combustible product users. A combination of missing data and “no’s” to any of the products included in this variable were counted as missing (n=10).

⁵Youth who reported ever having used snus pouches, loose snus, moist snuff, dip, spit or chewing tobacco, or dissolvable tobacco products (such as Ariva, Stonewall or Camel Orbs, Sticks or Strips), even one or two times were classified as ever smokeless tobacco product users. A combination of missing data and “no’s” to any of the products included in this variable (i.e., YS1002_01, YS1002_02, YU1003, and YD1002) were counted as missing (n=89).

⁶Youth were asked “Have you ever used marijuana, hash, THC, grass, pot or weed? (Yes/No)”, unless they had previously reported having ever smoked part or all of a cigar, cigarillo, or filtered cigar with marijuana in it. Affirmative responses to either question classified as an ever marijuana user (n=35 missing).

⁷All youth were asked “Have you ever used alcohol at all, including sips of someone’s drink or your own drink?” (Yes/No). Those that answered affirmatively were asked “About how old were you when you had your first alcoholic drink, other than small tastes or sips?” Those indicating that “I have never had an alcoholic drink other than small tastes or sips” were reclassified as never users for the purpose of this analysis. Those who did not know or refused to answer: 1) if they had ever used alcohol at all, or 2) the age when they first had an alcoholic drink were excluded from this variable (n=49 missing).

⁸The GAIN-SS Substance Use subscale consists of seven items: (1) used alcohol or other drugs weekly or more often; (2) spent a lot of time getting alcohol or other drugs; (3) spent a lot of time using or recovering from alcohol or other drugs; (4) kept using alcohol or other drugs, even though it was causing social problems, leading to fights, or getting into trouble with other people; (5) use of alcohol or other drugs caused reduced involvement in activities at work, school, home, or social events; (6) withdrawal problems; and (7) use of alcohol or other drugs to stop being sick or avoid withdrawal problems. Youth were asked to report whether they had experienced each problem: within the past month; 2–12 months ago; over a year ago; never; or don’t know/refused. Problems experienced within the past year were tallied and dichotomized by severity, with four or more problems reported in the past year categorized as a high score (n=253 missing; includes those missing responses for any of the scale items).

⁹The GAIN-SS Internalizing subscale consists of four items, asking about the last time you had significant problems with: (1) feeling very trapped, lonely, sad, blue, depressed, or hopeless about the future; (2) sleep trouble, such as bad dreams, sleeping restlessly, or falling asleep during the day; (3) feeling very anxious, nervous, tense, scared, panicked, or like something bad was going to happen; and (4) becoming very distressed and upset when something reminded you of the past. Youth were asked to report whether they had experienced each problem: within the past month; 2–12 months ago; over a year ago; never; or don’t know/refused. Problems experienced within the past year were tallied and dichotomized by severity, with four problems reported in the past year categorized as a high score (n=206 missing; includes those missing responses for any of the scale items).

¹⁰The GAIN-SS Externalizing subscale consists of seven items, asking about the last time you did the following two or more times: (1) had a hard time paying attention at school, work, or home; (2) had a hard time listening to instructions at school, work, or home; (3) lied or conned to get things you wanted or to avoid having to do something; (4) were a bully or threatened other people; (5) started physical fights with other people; (6) felt restless or the need to run around or climb on things; and (7) gave answers before the other person finished asking the question. Youth were asked to report whether they had experienced each problem: within the past month; 2–12 months ago; over a year ago; never; or don’t know/refused. Problems experienced within the past year were tallied and dichotomized by severity, with four or more problems reported in the past year categorized as a high score (n=372 missing; includes those missing responses for any of the scale items).

¹¹Youth were asked to indicate whether they agreed or disagreed with the following three measures regarding sensation seeking: (1) I like to do frightening things; (2) I like new and exciting experiences, even if I have to break the rules; and (3) I prefer friends who are exciting and unpredictable. Response options for each item were strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree. Responses were scored according to strength of agreement (4, 3, 2, 1, 0), and then summed to create an overall score. Based on the overall distribution of scores, the approximate upper quartile (75%) of six or higher was selected as the cut-off for high levels of sensation seeking. Scores were then dichotomized to indicate high (≥6) versus low (<6) levels of sensation seeking (n=152 missing; includes those missing responses for any of the scale items).

¹²Academic achievement: n=83 missing (including n=38 where school was ungraded). This item was reported by parents in the parent interview.

¹³All youth were asked “Does anyone who lives with you now do any of the following? Choose all that apply: 1) Smoke cigarettes; 2) Use smokeless tobacco, such as chewing tobacco, snuff, dip, or snus; 3) Smoke cigars, cigarillos, or filtered cigars; 4) Use any other form of tobacco; 5) No one who lives with me now uses any form of tobacco; 6) Don’t know/refused. Responses were categorized as no household use of any form of tobacco (option 5), household member smokes cigarettes (in addition to other forms of tobacco) (option 1), or household member uses some other form of tobacco (without cigarettes) (options 2, 3, or 4, and not 1) (n=66 missing).

¹⁴All youth were asked “During the past seven days, about how many hours were you around others who were smoking [whether or not you were smoking yourself]? Include time in your home, in a car, at school, or outdoors.” All responding with a value of >0 hours were classified as exposed to others smoking in the past seven days (n=251 missing).

Similarities were seen between ever e-cigarette only and cigarette only users on GAIN Internalizing scale scores and alcohol use, although dual users had increased odds of both risk factors, a finding that remained after controlling for age, race, ethnicity and gender (Table 2). Specifically, for GAIN Internalizing, no difference was seen between ever cigarette only smokers and e-cigarette users, but dual users had slightly higher odds of high GAIN Internalizing scores (aOR=1.6, 95% CI=1.2-2.0) than ever e-cigarette users. Similar results were seen between ever cigarette only smokers and e-cigarette users, but dual users had higher odds of alcohol use than ever e-cigarette only users (aOR=1.8, 95% CI=1.4-2.4). No differences were seen between ever e-cigarette only, cigarette only, and dual users for low academic achievement, high GAIN Externalizing scores, and high sensation seeking scores, which persisted after adjusting for age, race, ethnicity and gender.

Patterns of e-cigarette use in past 30 days and reasons for use by smoking status

Table 3 presents information for past 30-day e-cigarette use overall and by other tobacco use status (past 30-day: e-cigarette only use, e-cigarette and cigarette use [may have used other products], and e-cigarette and other [non-cigarette] tobacco product use). Among the 3.0% of past 30-day youth e-cigarette users (n=398), 37.0% used e-cigarettes only in the past 30 days, 45.1% used e-cigarettes and cigarettes, and 17.8% used e-cigarettes and at least one non-cigarette tobacco product. Frequency of past 30-day e-cigarette use was generally low, with 73.4% having used on five or fewer days. There were no differences in frequency of e-cigarette use by other tobacco use status. Among past 30-day e-cigarette and cigarette users, 40.8% smoked cigarettes 1-5 days and 39.6% smoked 20 or more days in past month. Furthermore, about half (50.6%) of past 30-day e-cigarette users (including users of additional products) started in the last year, but when stratified by other tobacco use status, 62.2% e-cigarette only users started in the last year compared to only 39.2% of dual e-cigarette and cigarette users. After stratifying by smoking status, top reasons for e-cigarette use were similar among dual users and e-cigarette only users (Supplemental Table 2). However, significantly more dual users than e-cigarette only users endorsed reasons for e-cigarette use including: e-cigarettes are less harmful, e-cigarettes smell less than cigarettes, e-cigarettes can be used in places where smoking is not allowed, and e-cigarettes feel like smoking a regular cigarette.

Table 3.

Characterization of duration, intensity, and frequency of e-cigarette use among youth past 30-day e-cigarette users, overall and by current tobacco use status–PATH Study Wave 1 Data (2013-2014)

	Overall1 n=398 (3.0% of full sample)	Past 30-Day E-Cigarette Only Users2 n= 141 (37.0%)	Past 30-Day E-Cigarette Users + Cigarette Smokers3 n=190 (45.1%)	Past 30-Day E-Cigarette + Other (Non-Cigarette) Tobacco Users4 n=67 (17.8%)
	n (%) [95% CI]	n (%) [95% CI]	n (%) [95% CI]	n (%) [95% CI]
Current Age
12-14 years	73 (16.6%) [13.2-20.6]	30 (18.9%) [13.3-26.3]	31 (15.2%) [10.5-21.4]	12 (15.4%) [8.5-26.3]
15-17 years	325 (83.4%) [79.4-86.8]	111 (81.1%) [73.7-86.7]	159 (84.8%) [78.6-89.5]	55 (84.6%) [73.7-91.5]
Cumulative History of E-cigarette Use5
1 puff, never a whole e-cigarette	121 (31.1%) [26.4-36.2]	57 (40.3%) [30.9-50.5]	43 (23.9%) [18.2-30.7]	21 (30.1%) [20.4-42.1]
1-10 e-cigarette(s)	208 (52.7%) [47.1-58.3]	68 (50.0%) [39.6-60.4]	108 (56.8%) [49.7-63.7]	32 (47.8%) [34.8-61.1]
11+ e-cigarettes	67 (16.2%) [12.7-20.5]	15 (9.7%) [5.8-15.7]	39 (19.3%) [14.0-26.0]	13 (22.1%) [13.8-33.5]
Time Since First E-cigarette Use6
<1 year	193 (50.6%) [45.3-55.9]	84 (62.2%) [53.4-70.2]	73 (39.2%) [32.4-46.5]	36 (55.6%) [43.1-67.3]
1-<2 years	143 (34.4%) [29.7-39.4]	41 (28.3%) [20.8-37.3]	79 (39.9%) [32.7-47.6]	23 (33.2%) [22.3-46.1]
≥2 years	61 (15.0%) [11.8-18.8]	15 (9.5%) [5.7-15.5]	38 (20.9%) [15.2-28.0]	¶
Frequency of Past 30-Day E-cigarette Use7
1 day	121 (30.8%) [26.1-36.0]	56 (40.4%) [32.1-49.2]	47 (25.2%) [19.3-32.2]	18 (25.4%) [16.0-37.9]
2-5 days	165 (42.6%) [37.0-48.3]	53 (38.2%) [30.1-46.9]	84 (46.0%) [38.8-53.5]	28 (42.8%) [30.4-56.1]
6-20 days	74 (18.5%) [14.8-22.9]	24 (17.2%) [11.6-24.8]	36 (17.9%) [13.1-24.0]	14 (22.7%) [13.5-35.5]
21-30 days	33 (8.1%) [5.4-11.9]	¶	22 (10.8%) [7.0-16.4]	¶
Owns an E-cigarette8	125 (31.6%) [27.2-36.5]	34 (23.6%) [17.3-31.2]	66 (34.0%) [27.6-41.0]	25 (42.5%) [30.8-55.1]
Cigarette Smoking Status
Never smoker	90 (23.6%) [19.4-28.3]	67 (46.7%) [38.6-55.1]	--	23 (35.1%) [23.1-49.2]
Ever smoked a cigarette, but not within past month	118 (31.3%) [26.2-36.9]	74 (53.3%) [44.9-61.4]	--	44 (64.9%) [50.8-76.9]
Current Smoker9	190 (45.1%) [40.2-50.1]	--	190 (100.0%)	--
Past 30-day smoker, smoked 1-5 days during past month	--	--	73 (40.8%) [32.4-49.8]	--
Past 30-day smoker, smoked 6-19 days during past month	--	--	38 (19.6%) [14.5-25.9]	--
Past 30-day smoker, smoked >=20 days during past month	--	--	74 (39.6%) [31.6-48.3]	--
Use of Flavored E-cigarettes
First e-cigarette tried was flavored10	340 (86.6%) [82.7-89.8]	127 (90.3%) [83.9-94.4]	153 (82.0%) [75.9-86.8]	60 (90.7%) [81.1-95.7]
E-cigarette currently/most recently used was flavored10	341 (86.2%) [81.3-90.0]	125 (88.2%) [80.4-93.1]	156 (83.2%) [76.0-88.6]	60 (89.7%) [78.5-95.4]

^¶Estimate not presented because relative standard error ≥30% or denominator <50; -- not applicable.

¹n=20 who responded they had used in the past 30 days, but later reported using 0 out of the past 30 days are excluded from this table.

²Any past 30-day e-cigarette user who reported also using cigarettes, any type of cigar, snus, smokeless, hookah, pipe, dissolvables, bidis or kreteks in the past 30 days was excluded from this column. One past 30-day e-cigarette user refused to answer whether they had smoked a cigarette in the past 30 days and one reported that they didn’t know if they had used a cigarillo in the past 30 days. These two respondents are currently included in this column. (i.e., treating don’t know or refused as no’s).

³Any past 30-day e-cigarette users who also reported past 30-day cigarette smoking and past 30-day use of any type of cigar, snus, smokeless, hookah, pipe, dissolvables, bidis or kreteks were included in this column.

⁴Any past 30-day e-cigarette user who reported NO past 30-day use of cigarettes, but past-30-day use of any type of cigar, snus, smokeless, hookah, pipe, dissolvables, bidis or kreteks were included in this column.

⁵Cumulative history of e-cigarette use: n=2 missing/don’t know/refused.

⁶Time since first use of an e-cigarette was calculated by subtracting YE1006 (“How old were you when you first tried an e-cigarette, even one or two times?”) from current age. N=1 reported don’t know to age started using.

⁷Frequency of past 30-day e-cigarette use: n=2 refused to answer, n=2 replied “don’t know” and n=1 was coded as “improbable response removed.”

⁸“Do you own your own e-cigarette?” (YE1090): n=3 missing (refused to answer).

⁹Current smoking status (past 30-days): n=5 missing smoking status (n=2 don’t know, n=2 refused, and n=1 improbable response removed), as these youth did not give an answer for the number of days smoked in the past 30 days.

¹⁰“Flavored to taste like menthol, mint, clove, spice, candy, fruit, chocolate, alcohol (such as wine or cognac), or other sweets?” (YE1108 and YE1130): n=1 missing first e-cigarette was flavored, 0 missing e-cigarette used most recently was flavored.

DISCUSSION

The public health consequences of e-cigarette use in the U.S. rests in large part on how youth are using the product and the subsequent effect on long-term tobacco use behaviors. In the current analysis, we observed differences and similarities for known tobacco use risk factors across mutually exclusive groups of susceptibility and ever use of e-cigarettes and cigarettes, and describe patterns of use among current e-cigarette users. Those susceptible to both products reported more and stronger associations with traditional risk factors than those susceptible to one product. Those susceptible to a single product, however, still showed increased risk status compared to committed never users. In general, committed never users had the lowest odds of any of the studied tobacco use risk factors, while selected risk factors (ever other tobacco/marijuana/alcohol use, low academic achievement, and exposure to tobacco use), were significantly associated with single product use (e-cigarette or cigarette), consistent with previously reported research.^11,12 Additionally, ever dual users had higher odds of nearly all risk factors compared to ever e-cigarette only use. The majority of past 30-day e-cigarette users had smoked cigarettes and had used e-cigarettes five days or less in the past 30 days.

Our findings inform discussions about hypotheses related to the impact of e-cigarette use on cigarette smoking among youth. The overlap in known risk factors among ever e-cigarette and cigarette users (such as ever marijuana and other tobacco use) suggest these single product users are similar on some risk factors, which could support the common liability theory. Additionally, similarities in scores on the psychosocial scales (GAIN Internalizing, GAIN Externalizing, and Sensation seeking) could support the notion that youth who participate in high-risk behaviors and may otherwise have smoked cigarettes may be using e-cigarettes instead (diversion hypothesis). Nonetheless, some differences were observed between ever e-cigarette and cigarette users on other risk factors, which may suggest that e-cigarettes could be attracting youth with lower risk status, possibly supporting the catalyst hypothesis. However, the cross-sectional nature of the data limits interpretations of temporality and causality.

Previous studies have assessed similarities and differences in risk factors associated with ever e-cigarette and cigarette use, and have highlighted the importance of internalizing, externalizing, and substance use behaviors among youth tobacco users. However, these studies are limited by smaller, regional samples.^4,11,12 In our study, consistent with prior research, we found higher internalizing scores among dual users, but similar scores among ever e-cigarette and cigarette users.¹² We saw no differences in externalizing scores by user group.¹¹ Lastly, sensation seeking had an ordered pattern, with ever dual users having the greatest percentage of high sensation-seeking scores, followed by ever cigarette users and e-cigarette users, consistent with previous results.¹¹

Ever dual users and cigarette smokers had higher odds of high GAIN Substance Use scores than ever e-cigarette users, with the largest percentage of high scores seen among ever dual users, which is a common theme in previous research.^11,12 There was a similar pattern observed for marijuana use, although findings in this area have been mixed in previous studies.^11,12 Our finding of ever dual users having the highest percentage of alcohol use, with similar use in cigarette and e-cigarette single product users, has been previously reported.^11,12 Ever cigarette only smokers and dual users had higher odds of being exposed to a household member who smokes compared to ever e-cigarette only users, as previously reported.⁴

In our study, about 74% of past 30-day e-cigarette users used e-cigarettes fewer than five days, similar to data from 2014 National Youth Tobacco Survey for e-cigarette users (middle school: 71.8%; high school 61.6%) and cigarette smokers (middle school: 61.1%; high school 49.3%).²¹ Among all past 30-day e-cigarette users, over 80% had used less than 10 times ever, indicating that at baseline many youth users may have started in the past year. However, past 30-day e-cigarette users who also used other tobacco products appeared to be more consistent e-cigarette users. Past 30-day dual users reported a greater cumulative history of e-cigarette use (i.e., higher number used), a longer time since first using e-cigarettes, and more frequent e-cigarette use. Additionally, about half of past 30-day dual users also smoked cigarettes on more than 15 days. Similar reasons for e-cigarette use were observed when stratified by smoking status; however, dual users were more likely to report items related to health effects or avoidance of detection of e-cigarette use.

This study has several limitations. First, this study did not address all factors associated with youth tobacco use (e.g. product availability in home, marketing exposure), some possibly unique to e-cigarette use. Subsequent waves of the PATH Study may capture such information. Second, tobacco use, including e-cigarette and cigarette use, is self-reported, which may be subject to biases (i.e., response or self-report bias); however, such measures have been shown to be valid and reliable among youth and adult cigarette smokers.^23,24 Third, the associations and significance tests were not adjusted for multiple comparisons. Based on the number of models conducted, a Bonferroni corrected p-value of p=0.05 would be p=0.004. Nonetheless, most odds ratios were significant at the p=0.001. Fourth, Wave 1 of the PATH Study may not have captured the full range of electronic nicotine devices, which could result in underreporting of e-cigarette use. Lastly, susceptibility to and ever use of tobacco products other than e-cigarettes and cigarettes were not considered when defining the mutually exclusive groups.

Conclusion

This manuscript adds to the literature by describing the distribution of tobacco use risk factors among youth susceptible to or ever users of e-cigarettes or cigarettes using a nationally-representative sample. In this study, committed never users had the lowest odds of any risk factor, followed by those susceptible to single product use, then those susceptible to both products. We found similarities and differences among ever e-cigarette users, cigarette smokers, and dual users in terms of known risk factors. Compared to e-cigarette only users, dual users had higher odds of other tobacco, alcohol, and marijuana use, internalizing problems, low academic achievement, and exposure to others’ tobacco use. Ever cigarette smokers had higher odds of other tobacco use, marijuana use, high GAIN substance scores, lower academic achievement, and exposure to others smoking than e-cigarette only users, but the increased odds were lower than those observed for dual users. No differences were observed in the ever use groups for GAIN externalizing, sensation seeking, or household use of other tobacco. This analysis may inform future studies examining transitions between products and assessing the influence of tobacco use risk factors on the probability of progression to current tobacco use.