Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 1.
Published in final edited form as: J Adolesc Health. 2020 Nov 3;68(4):801–807. doi: 10.1016/j.jadohealth.2020.09.033

Adolescent e-cigarette onset and escalation: Associations with internalizing and externalizing symptoms

Janet Audrain-McGovern a, Daniel Rodriguez b, Shannon Testa a, Emily Alexander a, Stephen Pianin a
PMCID: PMC8012221  NIHMSID: NIHMS1643817  PMID: 33158686

Abstract

Purpose:

We sought to evaluate if internalizing symptoms (i.e., anxiety, depression) and/or externalizing symptoms (i.e., impulsivity, sensation-seeking, and substance use) were risk factors for the onset of 30-day e-cigarette use and escalation in the number of days used across the following 30 months among adolescents.

Methods:

Adolescents (N = 1808) from public high schools outside of Philadelphia, PA completed in-classroom surveys at wave 1 (fall 2016, beginning of 9th grade) and at 6-month intervals for the following 30 months (spring 2019, end of 11th grade).

Results:

A two-part Latent Growth Curve Model of e-cigarette use revealed significant positive associations between externalizing factors, past 30-day e-cigarette use, and the number of days of e-cigarette use only at wave 1 (p values < .05). Cigarette smoking was associated with a slowing in the rate of onset of 30-day e-cigarette use across the 30-month follow-up (β = −0.24, z = −2.41, p = 0.02). Depression was associated with an increased rate of escalation in the number of days of e-cigarette use across the following 30-months (β = 0.01, z = 2.52, p = 0.01), while anxiety was associated with a decrease (β = −0.72, z = −2.36, p = 0.02).

Conclusions:

The findings highlight two groups of adolescents at risk for e-cigarette use: adolescents whose e-cigarette use reflects a higher-risk profile with early e-cigarette onset, and adolescents who have a lower-risk profile, at least initially, with later e-cigarette use onset. The timing and content of prevention efforts should be tailored to each group.

Keywords: adolescents, e-cigarettes, depression, anxiety, substance use

E-cigarettes are the most prevalent tobacco product used by adolescents. In 2019, 20% of adolescents reported using an e-cigarette in the past 30 days [1]. The growth in e-cigarette use has outpaced our understanding of the factors that foster regular use. Behavioral health issues broadly categorized as internalizing symptoms and externalizing symptoms have a well-established link with adolescent cigarette smoking uptake [2]. Whether e-cigarettes share a similar risk factor profile as combustible cigarettes has been a subject of much debate [3–5].

Internalizing symptoms reflect an inward reaction to distress whereas externalizing symptoms are characterized by an outward expression of distress [6]. Adolescent studies have documented that internalizing symptoms, such as anxiety and depression, increase the odds of initiating e-cigarette use up to 60% [7–9]. Externalizing symptoms, such as substance use and impulsivity, increase the odds of initiating e-cigarette use almost three-fold [7,8]. Substances associated with adolescent e-cigarette initiation include alcohol, marijuana, and combustible cigarettes [3,10–12]. Sensation-seeking, an externalizing trait, is the tendency to seek out novel and rewarding experiences [13,14]. Sensation-seeking has also been associated with an increased likelihood of ever using an e-cigarette among adolescents [15].

Longitudinal evidence is urgently needed to identify specific internalizing and externalizing symptoms that place adolescents at risk for progression in e-cigarette use beyond initiation. Studies to date have relied on brief composite scales of internalizing and externalizing symptoms [7,8], which may not provide the symptom resolution necessary to inform and target prevention efforts. In addition, determining whether these emotional and behavioral health issues are risk factors for greater use beyond initiation is critical to weighing their importance in e-cigarette prevention efforts.

This prospective cohort survey of adolescents in high school sought to determine whether adolescents who reported greater anxiety, depression, impulsivity, sensation-seeking, and comorbid substance use had a greater tendency to progress to current (past 30-day) e-cigarette use and to escalate in the number of days per month e-cigarettes were used across the 30 month follow-up. We anticipated that adolescents who became current and more frequent e-cigarette users would have few of these traditional risk factors for smoking uptake. Evidence documenting the role of these psychological and behavioral health symptoms in e-cigarette uptake among adolescents will help inform the content of e-cigarette prevention efforts as well as identify at-risk groups.

METHODS

Participants and Procedures

Participants were adolescents in the 9th grade taking part in a longitudinal cohort study of the relationship among combustible cigarette smoking, e-cigarette use, and other tobacco use. Participants were enrolled in one of four public high schools in suburban Philadelphia, PA. The schools were selected such that our sample would be demographically representative of adolescents nationwide (sex, race, ethnicity, annual household income). The cohort participants were drawn from 2198 students identified through class rosters at the start of 9th grade. Adolescents were ineligible to participate if they had a severe learning disability or if they did not speak fluent English. Based on the selection criteria, a total of 2017 of the 2198 (92%) students were eligible to participate.

Parents were mailed a study information letter (active information) with a telephone number to call to obtain answers to any questions and to decline consent for their adolescent to participate (passive consent). Of the 2017 eligible adolescents, 17 (1%) had a parent who actively declined their adolescent’s participation. Adolescents with parental consent were approached to provide their written assent for study participation. Adolescents who were absent on the assent/baseline survey days (n = 124, 6%) and adolescents who did not provide assent (n = 41, 2%) due to lack of interest were not enrolled in the cohort. Thus, 1835 of the 2000 adolescents with consent (92%) provided their assent to participate and completed a 40-minute paper and pencil survey. This baseline, or wave 1, survey was completed on-site during compulsory classes in the fall of 2016.

Adolescents completed five paper and pencil follow-up surveys at 6-month intervals with 92% completing a survey at wave 2 (N = 1687, spring 2017), 90% completing a survey at wave 3 (N = 1658, fall 2017), 89% completing a survey at wave 4 (N = 1643, spring 2018), 87% completing a survey at wave 5 (N = 1601, fall 2018), and 84% completing a survey at wave 6 (N = 1538, spring 2019). Adolescents were assigned a unique ID at the baseline survey. A cover page with the adolescent’s name connected to their ID on the follow-up survey was removed at the time of survey receipt. The participants included in this study are adolescents who had complete data on the study variables at baseline (N = 1808). The Institutional Review Board of the University of Pennsylvania and the administration of each of the four high schools approved the study. Data analyses were conducted in March of 2020.

Measures

E-cigarette Use.

The survey included an introduction explaining what e-cigarettes are, and the types of products or devices that are labeled as e-cigarettes. Images of different e-cigarette devices were provided to facilitate clarity [16,17]. From baseline (wave 1) to wave 3, these images included e-cigarettes, e-hookah, vape pens and mods. Images of USB-style pod vaporizers were added at wave 4. Excluding using an e-cigarette device for vaping marijuana, adolescents were asked: “Have you ever used an e-cigarette like the ones pictured above, even 1 or 2 times?” Adolescents who reported ever use of an e-cigarette were prompted to answer subsequent questions assessing lifetime frequency of e-cigarette use and time since last e-cigarette use. Adolescents who reported using an e-cigarette in the past 30 days were then asked on how many days in the past 30 days they used e-cigarettes [18–20]. As is standard, current use was defined as using an e-cigarette on at least one day in the past 30 days [18,21]. E-cigarette use was measured in all six waves.

Covariates.

Demographic characteristics such as sex, race, and ethnicity were assessed at baseline using self-report items. These demographic variables were included in the model to characterize the sample.

Internalizing symptoms.

Anxiety was measured with the six-item Profile of Mood States anxiety subscale [22–24]. Adolescents were asked to report the intensity of six symptoms (i.e., tense, on edge, uneasy, restless, nervous, anxious) over the past week using a Likert scale (0 = not at all to 4 = extremely). The Centers for Epidemiology Studies of Depression (CES-D) assessed depression symptoms over the past week [25]. The 20-item Likert style scale (0 = rarely or none of the time to 3 = most of the time) is frequently used to assess depressive symptoms among adolescents in epidemiological studies [9,26]. All internalizing symptoms were measured at baseline.

Externalizing symptoms.

The 18-item Current Symptoms Scale – Self Report Form measured ADHD symptoms, with nine items measuring the hyperactivity-impulsivity dimension [27]. Participants rated how frequently they experienced each symptom during the past six months from 0 (never or rarely) to 3 (very often). Sensation seeking was measured with the 8-item Brief Sensation Seeking Scale [28]. Adolescents were asked to indicate how much they agreed or disagreed with each of the eight statements (0 = strongly disagree to 4 = strongly agree). Combustible cigarette smoking, marijuana use, and alcohol use were assessed by asking adolescents if they had used these substances in the past 6 months [18,19]. All externalizing symptoms were measured at baseline.

Statistical Analyses.

A two-part Latent Growth Curve Model (LGCM) evaluated the association among depression, anxiety, impulsivity, sensation-seeking, and substance use at baseline, onset of 30-day e-cigarette use, and escalation in the number of days e-cigarettes were used in the past 30 days across the subsequent 30 months [29,30]. The first part of the model examined whether an adolescent used an e-cigarette in the past 30 days at baseline or began using e-cigarettes at least once in the past 30 days across follow-up (binary outcome, yes or no). The second part of the model examined the frequency of e-cigarette use at baseline and escalation in the number of days of use at follow-up among adolescents who reported using an e-cigarette in the past 30 days (continuous outcome, number of days used in the past 30 days, natural log transformed).

Increases in e-cigarette use were modeled across six repeated measures of e-cigarette use with the two-part model, measured six months apart, with latent variables (factors) representing baseline e-cigarette use (level factors) and change from baseline (trend factors) for the binary and continuous parts. Translating the relation between the observed use variables (repeated measures of e-cigarette use) and the underlying latent variable (factor), a positive trend value indicates the effect of each unit increase in time on use. Per convention, the relation between the factors and observed values were set a-priori. For linear growth across the six waves, the factor loadings were set to 0 (for baseline), 1, 2, 3, 4, and 5, with each unit increase referring to six months in time. The trend value indicated the speed and direction with which the propensity of use changed. Although the growth curve models assume positive change with time, change can be negative or positive. A positive value indicates an increase in propensity to use over time whereas a negative slope value indicates a decrease in propensity to use over time.

A measurement model without covariates was first estimated to test growth form (e.g., linear or quadratic). Bayesian Information Criteria (BIC) supported a linear versus quadratic growth form 6524.31 versus 6550.63, respectively. By default, the level (baseline or intercept) mean was set at zero for the binary part of the model. The linear trend mean value was −0.03, p <0.001. For the continuous part, the mean level (intercept) value was 0.01, p = 0.96, and the mean for the linear trend was 0.26, p <0.001. The two linear trend variances were constrained to equal zero to permit the model to solve, which is common with two-part LGCM. As for the intercept variances, for the binary level (intercept), the variance was 10.98, p <0.001. For the continuous level, the intercept variance was 0.75, p <0.001. Given the variance constraints, there was only one covariance; the two intercept factors were positively related, covariance = 2.25, p <0.001.

The measurement model was followed by the model including internalizing variables, externalizing variables and covariates. Based on Bayesian Information Criteria (6271.28), the final model proved to be the best fitting model. Means and standard deviations were used to describe continuous variables. Frequency distributions and proportions were used to describe categorical variables. Model parameter estimation for the two-part growth model employed maximum likelihood estimation with robust standard errors [31].

Mplus 8.2 software was used for the analyses. To account for missing data, Mplus uses a Full Information Maximum Likelihood (FIML) estimating procedure. The FIML procedure employs the Expectation Maximization (EM) algorithm, which assumes data are missing at random for the continuous part of the model [31]. For the binary part, individuals with missing data are treated as missing on both the binary and continuous model part.

A dummy school variable was created and included in the two part LGCM to test for clustering given that there were too few cluster units (schools) for a two-level analysis in Mplus [32]. There were no significant effects of school on the binary use level factor (p=0.57), binary trend factor (p=0.88), continuous level factor (p=0.61), or the continuous trend factor (p=0.46).

RESULTS

The sample characteristics are presented in Table 1. Approximately 73% of the sample was white and 79% non-Hispanic (79%). At baseline, 5.3% (n = 96) of the sample reported past 30-day e-cigarettes use. The average number of days in the past 30 days that e-cigarettes were used was 4 days (SD = 2.0). At wave 6, 30 months later, over 12% of adolescents (n = 172) reported past 30-day e-cigarette use. The average number of days in the past 30 days that e-cigarettes were used was 9.27 days (SD = 1.93).

Table 1.

Characteristics of Study Sample (N = 1808).

Categorical Variables Level N %
Sex Male 907 50.2
Female 901 49.8
Race White 1308 72.3
Black 265 14.7
Other 235 13.0
Ethnicity Hispanic 371 20.5
Non-Hispanic 1437 79.5
Cigarette Smoking Did not smoke in the past 6 months 1718 95.0
Smoked in the past 6 month 90 5.00
Marijuana Use Did not use in the past 6 months 1506 83.3
Used in the past 6 month 302 16.7
Alcohol Use Did not use in the past 6 months 1606 88.8
Used in the past 6 month 202 11.2
Continuous Variables Mean SD
Depression 18.29 9.39
Anxiety 6.27 5.87
Sensation-seeking 13.95 7.13
Impulsivity 5.56 4.63
Open in a new tab

Association between Internalizing and Externalizing Variables and the Onset of 30-day E-cigarette Use.

Effects are presented as non-standardized path coefficients, standard errors, z-statistic values, and p-values, with interpretations based on standardized coefficients (standard deviation change). The non-standardized path coefficients, standard errors, z-values, and probabilities are presented in Table 2. Figure 1 depicts the model with standardized path coefficients for significant paths only.

Table 2.

The Impact of Internalizing and Externalizing Symptomology on Adolescent 30-day E-cigarette Use. a

Binary e-cigarette level (past 30-day use at baseline)
 Binary e-cigarette trend (onset of 30-day use at follow-up)
beta SE Z-stat P-value beta SE Z-stat P-value
Sex 0.08 0.30 0.27 0.78 −0.04 0.08 −0.50 0.62
Black −0.77 0.51 −1.50 0.13 −0.08 0.14 −0.59 0.56
Other −0.03 0.42 −0.08 0.94 −0.17 0.12 −1.51 0.13
Ethnicity −0.71 0.35 −2.03 0.04 0.23 0.10 2.36 0.02
Anxiety −0.04 0.03 −1.50 0.13 0.01 0.01 1.53 0.13
Depression 0.01 0.02 0.46 0.64 0.01 0.01 0.89 0.38
Impulsivity 0.04 0.04 0.92 0.36 −0.01 0.01 −0.49 0.62
Sensation-seeking 0.06 0.03 2.20 0.03 0.01 0.01 1.36 0.17
Alcohol 1.87 0.37 5.10 <0.001 −0.07 0.10 −0.71 0.48
Marijuana 2.84 0.41 6.85 <0.001 −0.20 0.11 −1.85 0.07
Smoking 2.73 0.51 5.38 <0.001 −0.46 0.14 −3.32 <0.001
Continuous e-cigarette level (number of days used/past 30 days at baseline)
 Continuous e-cigarette trend (escalation in days used at follow-up)
beta SE Z-stat P-value beta SE Z-stat P-value
Sex −0.35 0.18 −1.90 0.06 −0.02 0.05 −0.38 0.71
Black −0.14 0.25 −0.58 0.56 −0.08 0.08 −0.99 0.32
Other 0.10 0.27 0.37 0.71 −0.02 0.09 −0.28 0.78
Ethnicity −0.28 0.21 −1.33 0.18 0.14 0.06 2.39 0.02
Anxiety 0.03 0.02 1.27 0.21 −0.01 0.01 −2.00 0.05
Depression −0.02 0.01 −1.64 0.10 0.01 0.00 1.99 0.05
Impulsivity 0.01 0.02 0.55 0.59 0.01 0.01 0.87 0.39
Sensation-seeking 0.01 0.02 0.56 0.58 0.00 0.00 0.32 0.75
Alcohol 0.43 0.21 2.09 0.04 −0.07 0.05 −1.20 0.23
Marijuana 0.67 0.22 3.07 <0.001 −0.01 0.06 −0.18 0.86
Smoking 0.61 0.24 2.62 0.01 −0.08 0.06 −1.19 0.23
Open in a new tab
a

Note: Sex (0 = male, 1 = female), Race (Black versus White; Other versus White), Ethnicity (Non-Hispanic, 0 = no, 1 = yes), alcohol use past 6 months (0 = no, 1 = yes), cigarette smoking past 6 months (0 = no, 1 = yes), and marijuana use past 6 months (0 = no, 1 = yes)

Figure 1. Two-part latent growth curve model of the effects of internalizing and externalizing symptoms on adolescent e-cigarette onset and escalation.

Figure 1.

Open in a new tab

Note: Binary level factor represents past 30-day e-cigarette use at baseline. Binary trend factor represents onset of past 30-day e-cigarette use across follow-up. Continuous level factor represents number of days of e-cigarette use in the past 30-days at baseline. Continuous trend factor represents escalation in the number of days of e-cigarette use in the past 30-days across follow-up.

Baseline 30-day e-cigarette use.

Externalizing variables had a significant and positive effect on the binary e-cigarette level factor, which represented current (past 30-day) e-cigarette use at baseline. Higher sensation-seeking scores (β = 0.06, z = 2.20, p = 0.03) were associated with having used an e-cigarette in the past 30 days at baseline. Compared to no use within the past six months, use of alcohol (β = 1.87, z = 5.10, p < 0.001), marijuana (β = 2.84, z = 6.85, p < 0.001), and cigarettes (β = 2.73, z = 5.38, p < 0.001) in the past six months were associated with having used an e-cigarette in the past 30 days at baseline. Using standardized coefficients, these externalizing variables were associated with a 0.13, 0.58, 0.89, and 0.85 standard deviation increase in the propensity for past 30-day e-cigarette use at baseline, respectively.

Ethnicity had a significant effect on the binary e-cigarette level factor. Non-Hispanic adolescents versus Hispanic adolescents had a lower tendency for e-cigarette use within the past 30-days at baseline (β = −0.71, z = −2.03, p = 0.04, standardized β = −0.22). Impulsivity, depression and anxiety did not have significant effects on the binary e-cigarette level factor.

Onset of 30-day e-cigarette use across follow-up.

Cigarette smoking had a significant negative effect on the binary e-cigarette trend factor (β = −0.46, z = −3.32, p < 0.001), indicating that cigarette smoking at baseline was associated with a slowing in the rate of past 30-day e-cigarette use onset across the 30-month follow-up. Using a standardized coefficient, cigarette smoking at baseline (versus no cigarette smoking) was associated with a −2.18 standard deviation decrease in the rate of past 30-day e-cigarette use onset across the following 30 months. Baseline measures of impulsivity, sensation-seeking, marijuana use, alcohol use, anxiety, and depression did not have a significant effect on the binary e-cigarette trend factor, suggesting no significant role in the rate of new 30-day e-cigarette use onset across follow-up.

Ethnicity had a significant and positive effect on the binary e-cigarette trend factor. Non-Hispanic adolescents versus Hispanic adolescents had a more rapid rate of 30-day e-cigarette onset across follow-up (β = 0.23, z = 2.36, p = 0.02, standardized β = 1.11).

Association between Internalizing and Externalizing Variables and E-cigarette Escalation.

Number of days of e-cigarette use at baseline.

Cigarette smoking (β = 0.61, z = 2.62, p = 0.01), marijuana use (β = 0.67, z = 3.07, p < 0.001), and alcohol use (β = 0.43, z = 2.09, p = 0.04, standardized β = 0.51) had significant positive effects on the continuous e-cigarette level factor. Use of these substances within the past six months at baseline (versus no use in the past 6 months) was positively associated with the number of days of e-cigarette use at baseline. Using a standardized coefficient, cigarette smoking, marijuana use, and alcohol use were associated with a 0.73, 0.80, and 0.51 standard deviation increase in the number of days of e-cigarette use at baseline, respectively.

Anxiety, depression, impulsivity, and sensation-seeking were not associated with the number of days of e-cigarette use at baseline. Sex had a marginally significant negative effect on the continuous e-cigarette level factor (β = −0.35, z = −1.90, p = 0.057, standardized β = −0.41), which indicated that females used e-cigarettes on fewer days at baseline compared to males.

Number of days of e-cigarette use across follow-up.

Only depression (β = 0.01, z = 1.99, p = 0.047), anxiety (β = −0.01, z = −2.00, p = 0.046), and ethnicity (β = 0.14, z = 2.39, p = 0.02, standardized β = 1.56) had significant effects on the continuous e-cigarette trend factor. Higher levels of depression and Non-Hispanic ethnicity were positively associated with the rate of escalation in the number of days of e-cigarette use across follow-up. A standard deviation increase in depression at baseline resulted in a 0.70 standard deviation increase in the rate of e-cigarette escalation across the 30-month follow-up. Non-Hispanic adolescents had a more rapid escalation in the number of days of e-cigarette use across follow-up than Hispanic adolescents. In contrast, greater anxiety was negatively associated with the rate of escalation in the number of days of e-cigarette use across follow-up. A standard deviation increase in anxiety at baseline resulted in a 0.72 standard deviation decrease in the rate of e-cigarette escalation across the 30-month follow-up.

DISCUSSION

This prospective study of adolescents offers the first evidence of associations among specific internalizing and externalizing symptoms, and the onset of and escalation in current e-cigarette use. Adolescents who were current e-cigarette users at age 14, and who used e-cigarettes on a greater number of days, had multiple externalizing risk factors compared to adolescents who were not current e-cigarette users. In contrast, these externalizing variables and the internalizing variables – anxiety and depression – were not associated with subsequent onset of current e-cigarette use. These findings suggest that there are two potential groups of adolescents at risk for e-cigarette use: adolescents with earlier onset e-cigarette use in the context of a higher-risk behavioral profile, and adolescents who have a lower-risk profile, at least initially, with later onset of e-cigarette use. These findings begin to unite two disparate views of adolescent e-cigarette onset and suggest that e-cigarette prevention efforts may need to target two distinct groups of adolescents.

Some argue that adolescents who use e-cigarettes are simply the same high-risk adolescents who are characterized by sensation-seeking, cigarette smoking, and comorbid substance use [5]. The present findings support this notion, in part, as 30-day e-cigarette use at age 14 is one marker among several in a concerning adolescent behavioral profile. Earlier onset of substance use is characteristic of a behavioral profile linked to poor health outcomes, including dependence [33,34]. Cross-sectional studies have shown that ever and current use of alcohol or marijuana are associated with greater susceptibility to e-cigarette use [35] as well as current e-cigarette use [10,11]. Our findings suggest that positive associations between comorbid substance use and e-cigarette use are evident among adolescents with earlier onset e-cigarette use.

This subgroup of adolescents likely includes dual users of electronic and combustible cigarettes who have greater emotional and behavioral health issues relative to those adolescents who only use either combustible cigarette or e-cigarettes [3,12,36,37]. Indeed, adolescents who smoked cigarettes at baseline were already more likely to currently use e-cigarettes at baseline, and thus the effect of cigarette smoking on becoming a current e-cigarette user faded across the subsequent 30 months. Adolescents who smoke and use e-cigarettes tend to smoke more frequently [38], be nicotine dependent [39], and require assistance in becoming nicotine free.

Proponents of the risks associated with adolescent e-cigarette use have shown that e-cigarettes attract adolescents with a lower psychological and behavioral risk profile [3,4]. Consistent with these observations, the externalizing variables and the internalizing variables measured at baseline were not associated with subsequent onset of 30-day e-cigarette use after age 14. However, these findings do not exclude the possibility that changes in these symptoms influenced later onset of e-cigarette use. Sensation-seeking, a trait variable, doubles the odds of susceptibility to e-cigarette use among youth [35,40] and increases the likelihood of ever use by 45% [15]. Similar to the present findings, sensation-seeking does not appear to play a significant role in transitions to current e-cigarette use among older youth and young adults [15,41].

The present study is the first to show that adolescents with higher levels of anxiety or higher levels of depression mid-adolescence do not have an increased risk of progressing to current e-cigarette use. These findings are consistent with two longitudinal studies of young adults where neither depression nor anxiety symptoms predicted subsequent 30-day e-cigarette use one to two years later [41,42]. However, adolescents with higher levels of depression evidenced a more rapid escalation in the number of days of e-cigarette use across time. Our previous research documented a similar relationship between depression symptoms and combustible cigarette smoking among young adults [26]. Adolescents who have higher levels of depression may be more sensitive to the rewarding properties of e-cigarettes with nicotine and/or flavoring [43]. Given that persistent e-cigarette use can exacerbate depression symptoms [9], targeting e-cigarette prevention interventions to adolescents with elevated depression symptoms has the potential to interrupt these bi-directional influences. In contrast, adolescents with higher levels of anxiety evidenced a slower rate of escalation in the number of days of e-cigarette use across time. As the first study to examine the effects of anxiety on adolescent e-cigarette uptake, we can speculate that nicotine’s anxiogenic effects may lessen e-cigarette use among anxious youth [44]. The findings also emphasize that internalizing symptomatology may vary in its relationship to adolescent e-cigarette use.

As the first study to examine the longitudinal relationship among specific internalizing and externalizing symptoms and e-cigarette use onset and escalation, this study has strengths as well as limitations. The sample was demographically diverse and adolescents were measured during a developmentally vulnerable period for tobacco use. We also used repeated measures of e-cigarette use, analyzing the data in a longitudinal fashion across six time points spanning three years.

It is important to note that our measures of anxiety, depression, and sensation-seeking were more detailed than previous studies [7,8], yet the self-report measures of anxiety and depression were not diagnostic assessments. In addition, other externalizing symptoms such as conduct or behavioral problems were not assessed. While internalizing and externalizing symptoms at baseline were not associated with e-cigarette onset, it is possible that increases in these symptoms across time affected subsequent e-cigarette onset. This raises the possibility that adolescents who had later e-cigarette onset had a greater risk status at the time of 30-day e-cigarette use onset. Our statistical approach permitted an assessment of e-cigarette onset and escalation in one model, but including each internalizing and externalizing variable as time-varying rendered our statistical model too complex to solve. In order to fully understand the apparent low risk profile and the role of internalizing and externalizing symptoms to later e-cigarette onset, future research will need to examine these symptoms across time. As the present study was focused on revealing relationships between internalizing and externalizing symptoms, other factors that may explain later e-cigarette onset and escalation in the putative lower risk group (e.g., product features, exposure to advertising) were not included in the model. These variables have been the focus of our as well as others’ investigations.

CONCLUSIONS

In this cohort study, the findings highlight two potential groups of adolescents at risk for e-cigarette use: adolescents whose e-cigarette use reflects a higher-risk profile with early onset of e-cigarette use; and adolescents who have a lower-risk profile, at least initially, with later onset of e-cigarette use. The timing and content of e-cigarette prevention efforts will need to be tailored to each group. Adolescents with later onset will need to be better characterized to inform prevention efforts.

SUMMARY STATEMENT.

These findings begin to unite two disparate views of adolescent e-cigarette use by providing the first evidence for two groups of adolescents at risk for e-cigarette use. The timing and content of e-cigarette prevention efforts will need to be tailored to each group.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

LITERATURE CITED

  • 1.Johnston LD, Miech RA, O’Malley PM, Bachman JG, Schulenberg JE, Patrick ME. Monitoring the Future national survey results on drug use, 1975–2019: Overview, key findings on adolescent drug use. Ann Arbor: Institute for Social Research, University of Michigan, 2020. [Google Scholar]
  • 2.Upadhyaya HP, Deas D, Brady KT, Kruesi M. Cigarette smoking and psychiatric comorbidity in children and adolescents. J Am Acad Child Adolesc Psychiatry. 2002;41(11):1294–305. [DOI] [PubMed] [Google Scholar]
  • 3.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):e43–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Glantz SA, Bareham DW. E-cigarettes: use, effects on smoking, risks, and policy implications. Annu Rev Public Health. 2018;39:215–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Abrams DB, Glasser AM, Pearson JL, Villanti AC, Collins LK, Niaura RS. Harm minimization and tobacco control: reframing societal views of nicotine use to rapidly save lives. Annu Rev Public Health. 2018;39:193–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Cosgrove VE, Rhee SH, Gelhorn HL, et al. Structure and etiology of co-occurring internalizing and externalizing disorders in adolescents. J Abnorm Child Psychol. 2011;39(1):109–23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Green VR, Conway KP, Silveira ML, et al. Mental health problems and onset of tobacco use among 12- to 24-year-olds in the PATH Study. J Am Acad Child Adolesc Psychiatry. 2018;57(12):944–54 e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Riehm KE, Young AS, Feder KA, et al. Mental health problems and initiation of e-cigarette and combustible cigarette use. Pediatrics. 2019;144(1). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Lechner WV, Janssen T, Kahler CW, Audrain-McGovern J, Leventhal AM. Bi-directional associations of electronic and combustible cigarette use onset patterns with depressive symptoms in adolescents. Prev Med. 2017;96:73–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Miech RA, O’Malley PM, Johnston LD, Patrick ME. E-cigarettes and the drug use patterns of adolescents. Nicotine Tob Res. 2016;18(5):654–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Richter L, Pugh BS, Smith PH, Ball SA. The co-occurrence of nicotine and other substance use and addiction among youth and adults in the United States: implications for research, practice, and policy. Am J Drug Alcohol Abuse. 2017;43(2):132–45. [DOI] [PubMed] [Google Scholar]
  • 12.Azagba S E-cigarette use, dual use of e-cigarettes and tobacco cigarettes, and frequency of cannabis use among high school students. Addict Behav. 2018;79:166–70. [DOI] [PubMed] [Google Scholar]
  • 13.Zuckerman M, Neeb M. Sensation seeking and psychopathology. Psychiatry Res. 1979;1(3):255–64. [DOI] [PubMed] [Google Scholar]
  • 14.Roberti JW. A review of behavioral and biological correlates of sensation seeking. J Res Pers. 2004;38(3):256–79. [Google Scholar]
  • 15.Case KR, Harrell MB, Perez A, et al. The relationships between sensation seeking and a spectrum of e-cigarette use behaviors: cross-sectional and longitudinal analyses specific to Texas adolescents. Addict Behav. 2017;73:151–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Conway KP, Green VR, Kasza KA, et al. Co-occurrence of tobacco product use, substance use, and mental health problems among youth: findings from wave 1 (2013–2014) of the population assessment of tobacco and health (PATH) study. Addict Behav. 2018;76:208–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Audrain-McGovern J, Rodriguez D, Pianin S, Alexander E. Initial e-cigarette flavoring and nicotine exposure and e-cigarette uptake among adolescents. Drug Alcohol Depend. 2019;202:149–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Wang TW, Gentzke A, Sharapova S, Cullen KA, Ambrose BK, Jamal A. Tobacco product use among middle and high school students — United States, 2011–2017. MMWR Morb Mortal Wkly Rep. 2018;67(22):629–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Audrain-McGovern J, Stone MD, Barrington-Trimis J, Unger JB, Leventhal AM. Adolescent e-cigarette, hookah, and conventional cigarette use and subsequent marijuana use. Pediatrics. 2018;142(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Cullen KA, Gentzke AS, Sawdey MD, et al. e-Cigarette use among youth in the United States, 2019. JAMA. 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Jamal A, Gentzke A, Hu SS, et al. Tobacco use among middle and high school students — United States, 2011–2016. MMWR Morb Mortal Wkly Rep. 2017;66(23):597–603. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Baker F, Denniston M, Zabora J, Polland A, Dudley WN. A POMS short form for cancer patients: psychometric and structural evaluation. Psychooncology. 2002;11(4):273–81. [DOI] [PubMed] [Google Scholar]
  • 23.McNair DM, Lorr M, Droppleman LF. Profile of mood states. San Diego, CA: Educational and Industrial Testing Service; 1992. [Google Scholar]
  • 24.Shacham S A shortened version of the Profile of Mood States. J Pers Assess. 1983;47(3):305–6. [DOI] [PubMed] [Google Scholar]
  • 25.Radloff LS. The use of the Center for Epidemiologic Studies Depression Scale in adolescents and young adults. J Youth Adolesc. 1991;20(2):149–66. [DOI] [PubMed] [Google Scholar]
  • 26.Audrain-McGovern J, Rodriguez D, Kassel JD. Adolescent smoking and depression: evidence for self-medication and peer smoking mediation. Addiction. 2009;104(10):1743–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Barkley RA. Attention-deficit hyperactivity disorder: a handbook for diagnosis and treatment. 2nd ed. New York, NY: Guilford Press; 1998. [Google Scholar]
  • 28.Hoyle RH, Stephenson MT, Palmgreen P, Lorch EP, Donohew RL. Reliability and validity of a brief measure of sensation seeking. Pers Individ Dif. 2002;32(3):401–14. [Google Scholar]
  • 29.Audrain-McGovern J, Rodriguez D, Leventhal AM, Cuevas J, Rodgers K, Sass J. Where is the pleasure in that? Low hedonic capacity predicts smoking onset and escalation. Nicotine Tob Res. 2012;14(10):1187–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Petras H, Nieuwbeerta P, Piquero AR. Participation and frequency during criminal careers across the life span*. Criminology. 2010;48(2):607–37. [Google Scholar]
  • 31.Muthén LK, Muthén BO. Mplus user’s guide. 8th ed. Los Angeles, CA: Muthén & Muthén; 2017. [Google Scholar]
  • 32.Thomas SL, Heck RH. Analysis of large-scale secondary data in higher education research: potential perils associated with complex sampling designs. Res High Educ. 2001;42(5):517–40. [Google Scholar]
  • 33.Jordan CJ, Andersen SL. Sensitive periods of substance abuse: early risk for the transition to dependence. Dev Cogn Neurosci. 2017;25:29–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.McCabe SE, West BT, McCabe VV. Associations between early onset of e-cigarette use and cigarette smoking and other substance use among US adolescents: a national study. Nicotine Tob Res. 2018;20(8):923–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Sawdey MD, Day HR, Coleman B, et al. Associations of risk factors of e-cigarette and cigarette use and susceptibility to use among baseline PATH study youth participants (2013–2014). Addict Behav. 2019;91:51–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Leventhal AM, Strong DR, Sussman S, et al. Psychiatric comorbidity in adolescent electronic and conventional cigarette use. J Psychiatr Res. 2016;73:71–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Curran KA, Burk T, Pitt PD, Middleman AB. Trends and substance use associations with e-cigarette use in US adolescents. Clin Pediatr (Phila). 2018;57(10):1191–8. [DOI] [PubMed] [Google Scholar]
  • 38.Demissie Z, Everett Jones S, Clayton HB, King BA. Adolescent risk behaviors and use of electronic vapor products and cigarettes. Pediatrics. 2017;139(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Vogel EA, Cho J, McConnell RS, Barrington-Trimis JL, Leventhal AM. Prevalence of electronic cigarette dependence among youth and its association with future use. JAMA Netw Open. 2020;3(2):e1921513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Carey FR, Rogers SM, Cohn EA, Harrell MB, Wilkinson AV, Perry CL. Understanding susceptibility to e-cigarettes: a comprehensive model of risk factors that influence the transition from non-susceptible to susceptible among e-cigarette naive adolescents. Addict Behav. 2019;91:68–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Spindle TR, Hiler MM, Cooke ME, Eissenberg T, Kendler KS, Dick DM. Electronic cigarette use and uptake of cigarette smoking: a longitudinal examination of U.S. college students. Addict Behav. 2017;67:66–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Bierhoff J, Haardorfer R, Windle M, Berg CJ. Psychological risk factors for alcohol, cannabis, and various tobacco use among young adults: a longitudinal analysis. Subst Use Misuse. 2019;54(8):1365–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Audrain-McGovern J, Strasser AA, Wileyto EP. The impact of flavoring on the rewarding and reinforcing value of e-cigarettes with nicotine among young adult smokers. Drug Alcohol Depend. 2016;166:263–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Hudson R, Green M, Wright DJ, et al. Adolescent nicotine induces depressive and anxiogenic effects through ERK 1–2 and Akt-GSK-3 pathways and neuronal dysregulation in the nucleus accumbens. Addict Biol. 2020:e12891. [DOI] [PubMed] [Google Scholar]

RESOURCES