Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Dec 1.
Published in final edited form as: Prev Sci. 2024 Nov 26;25(8):1298–1309. doi: 10.1007/s11121-024-01752-0

General and Specific Risk and Protective Factors for Cigarette and Electronic Nicotine Delivery System (ENDS) Use

Max A Halvorson 1, Marina Epstein 1, Justin D Caouette 1, Sarah Danzo 2, A Karryn Satchell 1, Sabrina Oesterle 3, Margaret R Kuklinski 1
PMCID: PMC11682713  NIHMSID: NIHMS2039940  PMID: 39592560

Abstract

Electronic nicotine delivery system (ENDS or e-cigarette) use is at least as common as cigarette use among today’s young adults. However, most prevention approaches are based on risk and protective factors (RPFs) that were identified with respect to cigarette use alone. To the extent that RPFs differ for cigarette and ENDS use, tailored approaches are needed to reduce the burden of nicotine use. In the current study, we examined both shared general RPFs and substance-specific RPFs across a developmental period spanning early adolescence to young adulthood, with the goal of identifying upstream preventive intervention targets for cigarette and ENDS use.

The current study used data from the Community Youth Development Study (n = 4,407) collected at 7 time points from early adolescence (age 12) through young adulthood (age 26). Using longitudinal structural equation modeling, we examined the contributions of adolescent and young adult RPFs to young adult cigarette and ENDS use. We examined general protective factors (e.g., family bonding, peer opportunities for prosocial involvement), cigarette-specific risk (e.g., friends’ cigarette use, permissive community norms), and peer polysubstance use.

General protective factors assessed in early adolescence had an indirect association with young adult cigarette and ENDS use, mediated through later RPFs. Whereas both cigarette-specific RPFs and peer polysubstance use predicted ENDS use in young adulthood, only cigarette-specific RPFs were related to cigarette use in young adulthood.

Our findings suggest that, although addressing known RPFs holds value for preventing ENDS use, additional prevention targets should also be considered. Early prevention approaches might seek to strengthen protective factors, whereas later prevention approaches might target cigarette beliefs for cigarette use and peer polysubstance use for ENDS use.

Keywords: ENDS, Cigarette use, Risk/protective factors, Adolescents, Young adults, Prevention

Introduction

Tobacco use is the number one cause of preventable death globally, costing the United States as much as $300 billion in annual lost productivity and healthcare expenses (Centers for Disease Control and Prevention, 2023; Jha & Peto, 2014). The health impacts of cigarettes are well established; a growing literature also highlights the potential harms of newer electronic nicotine delivery system (ENDS) products. ENDS are small handheld devices that heat a nicotine “e-liquid” to create a vapor that is then inhaled. The first products were introduced in 2007 and promoted as a safer alternative to cigarettes. Later research, however, has indicated that ENDS use may pose a significant threat to health through potential impacts on lung and cardiovascular health (Livingston et al., 2022; Marques et al., 2021; Wills et al., 2021) and via the association of ENDS use with continued or increased risk of cigarette smoking (Aladeokin & Haighton, 2019; Collins et al., 2019; Glasser et al., 2019). Although ENDS are marketed as a quitting tool for combustible cigarettes, the evidence that ENDS help smokers to quit is mixed (Hanewinkel & Isensee, 2015; Kosterman et al., 2021; Patil et al., 2020), and using ENDS has been shown to increase total nicotine use (Maglia et al., 2018). A growing body of evidence also suggests that ENDS use may actually increase the likelihood of initiating cigarette use among those who have not smoked previously (Epstein et al., 2021), raising concerns that ENDS use may actually increase the number of people who smoke rather than reducing it.

Because the physical harms of nicotine use accumulate over time, prevention programs delivered in childhood, adolescence, or young adulthood have the potential to greatly reduce the harms of a lifetime of nicotine use (Tobler, 1992). Moreover, because the likelihood of quitting nicotine use is low once a pattern of regular use is established in young adulthood (Frosch et al., 2009), preventing young adult use holds promise for reducing the burden of nicotine use. Given the intertwined nature of cigarette and ENDS use, prevention programs for either cigarette smoking or ENDS use should consider risk and protective factors (RPFs) for each substance, as well as those that are shared. Understanding similarities and differences in RPFs for cigarette and ENDS use can help guide the creation of effective prevention programs. If there are unique RPFs for ENDS use, new or more targeted programming may be needed for ENDS use. However, if RPFs are mostly shared between cigarette and ENDS use, existing evidence-based smoking prevention programs may be effective for the prevention of ENDS use as well.

Risk and Protective Factors for Cigarette and ENDS Use

Cigarette use tends to be driven both by general RPFs (e.g., family functioning, bonding to prosocial peers) that predict the use of a variety of substances and other behavioral health outcomes, and by cigarette-specific RPFs (e.g., peer and community norms around cigarette use; Epstein et al., 2013; Hawkins et al., 1999; Kim & Selya, 2020; Stone et al., 2012). The social development model (SDM; Catalano & Hawkins, 1996) describes pathways of risk for substance use and has been validated in several studies (Hawkins et al., 1999, 2005). Based in life course, social learning, and differential association theories, the SDM proposes that a) opportunities for prosocial engagement with communities, families, schools, and peers interact with b) skills to be successful, and c) recognition for prosocial involvement leads to greater bonding to these contexts. Bonding and other general RPFs (e.g., family bonding, prosocial peers) have been associated with reductions in the use of a variety of substances (including cigarette use) across a number of empirical studies (Iacono et al., 2008; Stone et al., 2012). The existing literature on cigarette use also points to an especially strong cigarette-specific pathway, defined by peer, parental, and community norms about tobacco use (Bailey et al., 2011; Epstein et al., 2013; Tyas & Pederson, 1998). For example, exposure to family members’ smoking and to permissive attitudes about smoking predict nicotine use over and above general RPFs (Bailey et al., 2011; Epstein et al., 2013).

The similarities between cigarette and ENDS use (e.g., both are inhaled, both deliver nicotine, visually similar) may lead youth to generalize their attitudes toward cigarettes (i.e., cigarette-specific RPFs) to ENDS products. The common liability hypothesis asserts, for example, that shared RPFs are responsible for the use of both cigarettes and ENDS (Committee on the Review of the Health Effects of Electronic Nicotine Delivery Systems et al., 2018). Some initial empirical work has provided support for this hypothesis (Kim & Selya, 2020; Sawdey et al., 2019; Stanton et al., 2023). However, there are also significant differences between cigarettes and ENDS. These include the branding of ENDS as safe “vapor,” youth-oriented flavors, and reduced harshness of inhaling vapor relative to cigarette smoke. These different features of ENDS may mean that different RPFs are associated with ENDS. Some additional evidence suggests that one ENDS-specific RPF may be polysubstance use (Bentivegna et al., 2021; Gilbert et al., 2021). Given the variety of flavors available for purchase and the novelty of ENDS over traditional cigarettes, people who are prone to novelty and experimentation with substances may be more inclined to try ENDS.

The Developmental Course of Risk for Cigarette and ENDS Use

ENDS use is most prevalent among young people in their late teens and 20s (Cornelius et al., 2020; Cullen et al., 2019). However, risk for nicotine use more broadly unfolds over the course of development, stretching back to early adolescence (Carey et al., 2019; Kwon et al., 2018). During childhood and into early adolescence, risk factors for later substance use are not yet substance specific and are centered in the general family environment (Catalano & Hawkins, 1996). As adolescence progresses, youth are also exposed to norms and attitudes outside the home about specific drugs which influence the likelihood of their use (Hawkins et al., 1992). During later adolescence, both peer and family influences communicate norms about substance use to youth, influencing their likelihood of trying and using substances. During emerging adulthood, youth might move out of the home and begin to select new contexts (workplace, college, romantic relationships, new peer groups) that also influence risk for substance use (Arnett, 2007; Elder, 1998), presenting a new and powerful set of RPFs. During this period, romantic partner and peer use may begin to be even more influential than family use.

Although the relative importance of different contexts shifts as youth age, RPFs measured in early adolescence (e.g., peer cigarette use, family tobacco use environment) tend to show an impact on nicotine use later in development (Epstein et al., 2013; Kim & Selya, 2020). Previous research has also shown that risk factors for nicotine use (e.g., peer smoking and norms about cigarette use) show considerable continuity across developmental periods as youth who use nicotine begin to self-select nicotine-using peers (Stone et al., 2012). As ENDS is a relatively new substance for today’s young adults, little is known about whether and how early RPFs for cigarette use relate to later ENDS use. A greater understanding of the progression of general and specific risk factors for nicotine use across development can enable the identification of points of prevention along this developmental path. This is especially important for understanding today’s young adults, for whom ENDS were not available during childhood and adolescence. An added benefit of studying today’s young adults is that understanding how known RPFs predict the use of a novel substance (ENDS) can help identify potential intervention strategies for reducing the use of other novel substances that may emerge in the future. For example, if favorable cigarette attitudes in adolescence are associated with ENDS use in young adulthood, interventions implemented in adolescence that challenge social norms around cigarette use might also be effective in reducing downstream ENDS use. In contrast, if ENDS-specific RPFs (e.g., adolescent peer polysubstance use) are related to ENDS use but not cigarette use, prevention programs for drugs other than cigarettes might have crossover effects on ENDS use.

The current study

In the current study, we had three primary aims: a) to identify general RPFs that predict young adult cigarette and ENDS use, b) to identify specific RPFs that predict only one type of use, and c) to understand how childhood and adolescent RPFs relate to young adult cigarette and ENDS use. To accomplish these aims, we analyzed data from a longitudinal study of youth surveyed repeatedly from early adolescence through their 20s.

In light of findings suggesting cigarette and ENDS use are partially driven by both general risk and cigarette-specific pathways (Epstein et al., 2013; Stone et al., 2012), we examined the extent to which general RPFs (e.g., family bonding) would predict both cigarette and ENDS use, cigarette-specific RPFs (e.g., community norms around smoking) would predict cigarette use more strongly than ENDS use, and peer and partner use of other drugs would emerge as an ENDS-specific risk factor. With respect to the developmental course of risk and protection, we predicted that adolescent RPFs would predict both young adult RPFs and young adult cigarette and ENDS use (Cicchetti & Rogosch, 1999; Hawkins et al., 1992).

Method

Participants

Data were provided by 4,407 participants in the Community Youth Development Study (CYDS), a longitudinal study of individuals who grew up in 24 small towns (average population 14,646) across seven states (Colorado, Illinois, Kansas, Maine, Oregon, Utah, and Washington). Communities were matched on several demographic indices and paired within each state; one community within each pair was randomly assigned to implement Communities That Care, a community-based prevention planning and implementation strategy to reduce youth behavioral health problems (Hawkins et al., 2008). The present study did not have intervention-related hypotheses, but tested focal associations for differences by intervention condition to assure that the sample could be pooled for analyses. Since no significant differences by intervention condition were found, results were reported based on analyses of the full sample (see Analysis Plan).

The sample was drawn from all fifth graders (average age 11 years in 2004) in selected public schools (76% of eligible students). Participants were first surveyed in 2004 (average age 11 years) and resurveyed annually until Grade 10 (age 16), again in Grade 12 (age 18), and every 2 to 3 years thereafter at average ages 19, 21, 23, and 26 (in 2019). Retention was above 90% at all ages prior to age 23, and above 85% at ages 23 and 26. Twenty percent of participants self-identified as Latino/a or Hispanic. Participants selected multiple categories to identify their race: 71% selected White, 6% American Indian or Alaska Native, 4% Black or African American, 2% American, 1% Native Hawaiian or Other Pacific Islander, and 21% other (of which 77% identified as Latino/a or Hispanic); 1% did not provide an answer. Fifty percent of participants self-identified as male. In young adulthood, approximately 1% of participants identified with one of the following gender identity response options: trans male/trans man, trans female/trans woman, gender queer/gender nonconforming, or something else.

Measures

Cigarette and ENDS Use

Individuals reported their past-year cigarette and ENDS use at age 26 on a 7-point ordinal scale (0 occasions [1], 1-2 occasions [2], 3-5 occasions [3], 6-9 occasions [4], 10-19 occasions [5], 20-39 occasions [6], 40+ occasions [7]). The use of each substance was coded as 1 if participants reported using at least 3 times in the past year, 0 otherwise. We selected the threshold of 3 times in the past year to balance being inclusive with screening out experimental users.

Risk and Protective Factors

RPFs were measured during three developmental periods: early adolescence (ages 12-13), adolescence (ages 15-16), and young adulthood (ages 21 and 23). Unless otherwise noted, items were asked with respect to the past year. At each developmental period, a latent RPF score was created using a three-step process. First, a mean score was calculated across the multiple items comprising each scale at each time point. Second, these scale scores were averaged across the two time points that comprised each developmental period (e.g., age 12 and age 13). Third, developmental period scores for each RPF were entered as indicators for latent factors in the confirmatory factor analysis (CFA) models described in the Analysis Plan and in Table 1. Supplemental Table A includes item text and response options for all RPF items.

Table 1.

Risk and protective factors for cigarette and electronic nicotine delivery system (ENDS) use

Risk and protective factor domain Age Loading RPF
Early adolescent – general protective factors 12-13 1.00 Family protective factors (bonding, opportunities/rewards for involvement)
.83 Peer protective factors (bonding, opportunities/rewards for involvement)
.80 School protective factors (bonding, opportunities/rewards for involvement)
Adolescent – cigarette-specific risk 15-16 1.00 Number of best friends who smoked in past year
.55 Permissive community norms against cigarette use
.61 Permissive personal attitudes about cigarettes
Adolescent – friends’ polysubstance use 15-16 1.00 Best friends used any marijuana in past year
.75 Best friends binge drank at least weekly in past year
.64 Best friends used drugsa in past year
Young adult – general protective factors 21-23 .95 Peer protective factors (bonding, opportunities/rewards for involvement)
1.00 Romantic partner protective factors (attachment, support, prosocial behavior)
.76 Workplace protective factors (attachment, opportunities/rewards for involvement)
.64 Parent protective factors (interaction, support)
Young adult – cigarette-specific risk 21-23 1.00 Permissive community norms against cigarette use
.85 Best friends who smoked at least a pack a day in past year
.57 Partner smokes at least a pack a day in past year
Young adult – partner’s and friends’ polysubstance use 21-23 .95 Best friends used any marijuana in past year
.67 Best friends binge drank at least weekly in past year
.98 Best friends used drugsa in past year
1.00 Partner used any marijuana in past year
.79 Partner binge drank at least weekly in past year
.60 Partner used drugsa in past year
Open in a new tab

Note. Loadings are from confirmatory factor analysis model with correlations between all latent factors. Measurement model fit was acceptable, X2 (243) = 1830, p < .001; CFI = .924; TLI = .906; RMSEA = .048; SRMR = .040.

a

Drugs indicates LSD, cocaine, methamphetamines, opioids, or prescription drugs not prescribed by a doctor.

Adolescent Risk and Protective Factors.
General Protective Factors

were guided by the social development model (Catalano & Hawkins, 1996). Three domains of general protection were assessed: family (11 items), peers (12 items), and school (15 items). Each domain included items related to three subscales: bonding, opportunities for prosocial involvement, and rewards for prosocial involvement. Example items included “Do you feel very close to your mother?” (family bonding), “How many of your best friends have tried to do well in school?” (peer opportunities for prosocial involvement), and “My teachers notice when I am doing a good job and let me know about it” (school rewards for prosocial involvement). Each item was assessed using Likert scale items with four response options (NO, no, yes, YES).

Cigarette-specific Risk Factors

were chosen to capture cigarette-specific attitudes, norms, and use environment. Three factors were included: friends’ cigarette use (one item), favorable norms about youth smoking in the community (three items), and the youth’s own permissive attitudes toward cigarette use (one item). Example items included “How many of your best friends have smoked cigarettes?” (friends’ use), “How wrong would most adults (over 21) in your neighborhood think it is for kids your age to smoke cigarettes?” (favorable community norms, reverse-coded), and “How wrong do you think it is for someone your age to smoke cigarettes?” (individual permissive attitudes, reverse-coded).

Friends’ Polysubstance Use

was considered as a potential ENDS-specific risk factor, given evidence of heightened rates of polysubstance use among adolescent ENDS users (Bentivegna et al., 2021; Gilbert et al., 2021). This factor included peers’ polysubstance use (three items). Respondents reported how many of their best friends (“None,” vs. “Some,” “Most,” or “All”) used alcohol, cannabis, and any other drugs (LSD, cocaine, methamphetamines, other illegal drugs, or prescription drugs not prescribed to them including opioids). Use was assessed in the past year, and risky alcohol use was operationalized as having five or more drinks at least once a week. Because ENDS were not widely available for purchase until 2011 (when participants were 17-18 years old), ENDS-specific RPFs in adolescence were not measured in this study.

Young Adult Risk and Protective Factors.
General Protective Factors.

Four domains were assessed in young adulthood: parents, peers, romantic partners, and work context. Most individuals had a romantic partner in the past year (86%), at least one close friend (93%), and/or a workplace (83%). The parent domain (three items) included support from parents and frequency of interactions with parents. The peer domain (10 items) encompassed interaction with friends, friends’ prosocial behavior, and social support. The romantic partners domain (11 items) included partner prosocial behavior, partner support, and attachment to partner. The work domain (eight items) included opportunities for work, rewards, and attachment. Example items included “How much did you depend on your mother or your father for advice and guidance?” (parent support), “How many of your best friends have volunteered to do community service?” (peer prosocial behavior), “How much did you depend on your spouse/partner for advice and guidance?” (romantic partner support), and “I gain a sense of accomplishment from my job” (work rewards). If the individual indicated they did not have a romantic partner, a close friend, or a workplace, that domain was coded as missing.

Cigarette-specific Risk Factors.

Similar to measures in adolescence, three factors were included: friends’ cigarette use (one item), favorable norms toward youth smoking in the community (one item), and partner’s cigarette use (one item). Example items included “How many of your best friends have smoked one pack of cigarettes per day?” (friends’ use), “How wrong would most adults in your neighborhood think it is for someone your age to smoke one or more packs of cigarettes per day?” (favorable community norms) and “Did your spouse/partner regularly smoke one or more packs of cigarettes per day?” (partner’s use).

Romantic Partner’s and Friends’ Polysubstance Use.

To assess partners’ substance use, participants were asked whether their partner regularly used alcohol (five or more drinks, at least once a week), cannabis, and other drugs (LSD, cocaine, methamphetamines, other illegal drugs, or prescription drugs not prescribed to them including opioids). Friends’ polysubstance use at ages 21 and 23 was assessed using the same measures as in adolescence.

Analysis Plan

Analyses were conducted using R and the Tidyverse group of software packages (R Core Team, 2013; Wickham et al., 2019). Structural equation models were fit using lavaan (Rosseel, 2012).

Missing Data.

Survey-level nonresponse was the primary form of missing data, although response rates were generally high (see Participants section). Item-level nonresponse was very rare (< 1% within a given scale) due to data validation procedures. People who identified as racial or ethnic minorities were not any less likely to have responded to the survey at age 26. Men were more likely to be missing age 26 data than women (X2(1)= 115.71, p < .001), as were participants who had smoked cigarettes at the previous time point (X2(1) = 27.30, p < .001). Missing data were handled using full-information maximum likelihood estimation (FIML). In order to provide information about causes of missingness and properly adjust FIML estimates, we included a range of covariates (including sex) in analyses that correlate both with levels of cigarette use and with missingness.

Structural Equation Models.

We fit structural equation models to examine the relationships between adolescent RPFs, young adult RPFs, and cigarette use and ENDS use in young adulthood (see Figure 1). The WLSMV estimator was used to accommodate binary outcome variables (cigarette use, ENDS use) and fit statistics were assessed using Hu & Bentler’s (1999) criteria. We built models in order of increasing complexity, following common practices in building complex SEMs (Kline, 1998). Due to the use of the WLSMV estimator, prediction of cigarette and ENDS use at age 26 uses a probit link; probit coefficients were converted to approximate odds ratios (Gelman & Hill, 2006). First, we fit confirmatory factor analysis models to create latent factors capturing RPFs in adolescence and young adulthood (see Table 1). Second, we specified correlations and predictive relations among RPFs (e.g., early adolescent general protection predicting adolescent cigarette-specific risk). Third, we specified conceptually driven residual covariances between items. These included a) items asking about the same substance (e.g., cannabis), b) items asking about the same domain (e.g., friends), and c) demographic statuses with known correlations (e.g., race and parental education). Fourth, we added cigarette and ENDS use (as measured variables) as outcomes of all of the early adolescent, adolescent, and young adult predictors; the covariance between cigarette and ENDS use was also estimated.

Fig 1. Adolescent and young adult predictors of young adult cigarette and electronic nicotine delivery system (ENDS) use.

Fig 1

Open in a new tab

Note. OR = odds ratio, ENDS = electronic nicotine delivery system. All paths depicted with a solid arrow are statistically different from zero at p < .05. Paths depicted with a dashed arrow were entered into the model but not statistically significant at α = .05. Models controlled for gender, race, ethnicity, and parental education by estimating their impact both on early adolescent general protection and on young adult cigarette and ENDS use.

These models controlled for gender, race, ethnicity, and parental education by estimating their impact both on early adolescent general protection and on young adult cigarette and ENDS use. Marginal effects for probit coefficients (raw difference in the probability of cigarette or ENDS use) were probed for groups representing higher risk for nicotine use: male, White and non-Hispanic at high (+1 standard deviation) and mean values. We tested for intervention effects by fitting separate models for control and intervention participants and testing for differences in model estimates. None were noted, so we present results for the full sample below. Both direct and indirect effects of early adolescent and adolescent RPFs on cigarette and ENDS use via young adult RPFs were investigated.

We also investigated whether the overlap between cigarette use and ENDS use might explain results. Predictors of concurrent cigarette and ENDS use at age 26 were examined to assess whether concurrent use might be explained by different predictors than each type of use alone. To do so, fit an identical model to the primary model with concurrent use as the sole outcome.

Sensitivity Analysis.

We assessed whether the relationship between RPFs and cigarette and ENDS use might be confounded by early initiation of cigarettes or ENDS by fitting two models: a) a model with early cigarette initiation and early ENDS initiation predicting adolescent RPFs, and b) a model with early cigarette and early ENDS initiation directly predicting young adult cigarette and ENDS use. Early cigarette use was indicated if participants reported any use by ninth grade (average age 15). Age of first ENDS use was retrospectively reported on the age 26 survey, and any use by age 17 (2011) was chosen to coincide with when ENDS first became widely available for purchase and first tracked by the CDC. We also fit two additional models to ensure that our results were not confounded by the correlation between cigarette use and ENDS use: a model with cigarette use as the sole outcome, and a model with ENDS use as the sole outcome.

Results

Model Fit

The final structural equation model linked general and specific RPFs to young adult cigarette and ENDS use (see Figure 1 and Appendix Table A). Age 12-13 and age 15-16 RPFs were not directly associated with age 26 cigarette or ENDS use when controlling for age 21-23 RPFs. Estimating these additional paths in the full model resulted in multicollinearity and suppressor effects, and these paths were dropped from the final analysis. The final model fit showed acceptable fit to the data (X2 (301) = 2050, p < .001; CFI = .924; TLI = .905; RMSEA = .045; SRMR = .039).

General Substance Use RPFs and Young Adult Cigarette and ENDS Use

Age 12-13 general protection showed a serially mediated effect on age 26 cigarette use via its relationships with age 15-16 and age 21-23 cigarette-specific risk. For ENDS use, age 12-13 general protection also showed a serially mediated effect on ENDS use via age 15-16 and age 21-23 cigarette risk and a smaller but statistically significant effect through age 15-16 and age 21-23 peer and partner polysubstance use. Age 21-23 general protection was not associated with age 26 cigarette or ENDS use. See Table 2 for a summary of mediated effects.

Table 2.

Direct and indirect effects of risk and protective factors on cigarette and electronic nicotine delivery system (ENDS) use

Direct effects Indirect effects
OR [95% CI] β [95% CI] β [95% CI]
Predictors of age 26 cigarette use
 21-23 cigarette-specific risk 3.65 [2.70, 4.95] .81 [.62, 1.00]
 21-23 general substance use risk 0.94 [0.83, 1.07] −.04 [−.12, .04]
 21-23 partner’s/friends’ polysubstance use 0.81 [0.61, 1.10] −.13 [−.31, .06]
 15-16 cig risk → 21-23 cig risk → 26 cig use .74 [.53, .94]
 15-16 cig risk total indirect effect .71 [.55, .87]
 12-13 gen protect → 15-16 cig risk → 21-23 cig risk → 26 cig use −.39 [−.51, −.27]
 12-13 gen protect total indirect effect −.31 [−.35, −.26]
Predictors of age 26 ENDS use
 21-23 cigarette-specific risk 1.52 [1.12, 2.05] .26 [.07, .46]
 21-23 general substance use risk 0.92 [0.80, 1.05] −.05 [−.14, .03]
 21-23 partner’s/friends’ polysubstance use 1.44 [1.11, 1.90] .23 [.06, .40]
 15-16 cig risk → 21-23 cig risk → 26 ENDS use .23 [.06, .41]
 15-16 cig risk total indirect effect .31 [.27, .44]
 15-16 PPSU → 21-23 PPSU → 26 ENDS use .07 [.01, .14]
 15-16 PPSU total indirect effect .01 [−.09, .12]
 12-13 gen protect → 15-16 cig risk → 21-23 cig risk → 26 ENDS use −.12 [−.22, −.03]
 12-13 gen protect → 15-16 PPSU → 21-23 PPSU → 26 ENDS use −.03 [−.06, .00]
 12-13 gen protect total indirect effect −.20 [−.24, −.16]
Open in a new tab

Note. β values standardized with respect to latent variables. ENDS = electronic nicotine delivery system (e-cigarettes), Cig = cigarette, Gen = general, PPSU = peer (friends and/or partner) polysubstance use, Protect = protective factors. Bolded estimates are different from zero at p < .05. Total indirect effect indicates the sum of all paths from the specified risk or protective factor to the specified outcome.

Cigarette-specific RPFs, ENDS-specific RPFs, and Young Adult Cigarette and ENDS Use

A one standard deviation increase in age 21-23 cigarette-specific risk was associated with an approximately threefold increase in the probability of using cigarettes at age 26 (.32, 95% CI [.29, .35] vs. .91 [.83, .98]). Age 15-16 cigarette-specific risk showed a mediated effect on age 26 cigarette use via age 21-23 cigarette-specific risk. A one standard deviation increase in age 21-23 cigarette-specific risk was associated with an approximately 1.5-fold increase in the probability of using ENDS at age 26 (.30 [.27, .33] vs. .48 [.33, .63]).

Age 21-23 peer and partner polysubstance use was not related to age 26 cigarette use. However, peer and partner polysubstance use was associated with an increase in the likelihood of using ENDS at age 26 (.30 [.27, .33] vs. .34 [.30, .38]). Age 15-16 friends’ polysubstance use also showed a mediated effect on age 26 ENDS use via its relationship with age 21-23 peer and partner polysubstance use.

RPFs Related to Concurrent Cigarette and ENDS Use

Concurrent use of cigarettes and ENDS at age 26 was common, with 48% of ENDS users using cigarettes and 57% of cigarette users using ENDS. RPFs related to age 26 cigarette use were also related to age 26 concurrent cigarette and ENDS use. Age 21-23 cigarette-specific risk was the strongest predictor of concurrent cigarette and ENDS use (β = .46; OR = 1.59; p < .001), and general protection (β = −.02; p = .643) and friends’ polysubstance use (β = .10; p = .303) were not statistically significantly related to dual use.

Sensitivity Analyses

Though early cigarette and ENDS initiation were related to RPFs and to age 26 use, their addition did not change the pattern of direction or effect sizes of adolescent or young adult RPFs’ associations with cigarette and ENDS use. Models examining cigarette use as the sole outcome and with ENDS use as the sole outcome had estimates that were nearly identical to the full model (standardized coefficients within .02).

Discussion

The present study examined risk and protective factors for both cigarette and ENDS use in young adulthood, finding evidence both of commonality in predictors and differential impact of certain RPFs. Moreover, RPFs measured in adolescence were related to young adult substance use via their associations with RPFs assessed in young adulthood. Known general RPFs assessed in early adolescence were related to both cigarette and ENDS use via mediated relationships with both general and specific RPFs measured in young adulthood. Cigarette-specific RPFs predicted both cigarette and ENDS use (and concurrent use of both), though the magnitude of the relationship with cigarette use was stronger. Friends’ and partners’ use of other substances (alcohol, cannabis, and other drugs) were explored as a potential ENDS-specific risk factor, and were related to ENDS but not cigarette use.

Given the high rates of ENDS use among adolescents and young adults, as well as a clear connection to cigarette use, prevention approaches that address both cigarette and ENDS use are likely needed to minimize the public health impact of nicotine use. A variety of strategies for youth cigarette use prevention have shown evidence of effectiveness, including interventions focused on enhancing general social competence, refusal skills, family bonding, and opportunities to engage with nonsmoking peers and role models (Hawkins, 2012; Lantz et al., 2000; Thomas et al., 2015). These approaches were developed prior to the introduction of ENDS. Cigarette-specific RPFs did predict both cigarette and ENDS use in young adulthood. However, friends’ and partners’ polysubstance use emerged as a unique predictor of ENDS use in young adulthood. Though some existing interventions for nicotine use problems specifically recommend involving family members or romantic partners (Daley & Marlatt, 2006), others focus solely on the individual’s beliefs about nicotine use, motivation, or coping skills (Wray et al., 2018). Moreover, most prevention and treatment programs do not consider polysubstance use when addressing nicotine use. Our work supports the consideration of friends’ and partners’ substance use when addressing ENDS use.

Our findings also emphasize the importance of targeting adolescent RPFs due to the continuity of risk for cigarette and ENDS use across development. We found a mediated effect from early adolescent general protective factors, through adolescent cigarette RPFs, to young adult polysubstance use, and finally to ENDS use. This developmental chain can be traced as far back as the middle school years, during which bonding to important prosocial domains (schools, families, and peers) is predictive of later substance use. Prevention programs that target and enhance bonding to these domains in middle school can be effective in delaying or preventing substance use (Hawkins, 2012; Hawkins et al., 2008). During adolescence, peer and community norms about a specific substance (i.e., cigarettes) are internalized and begin to predict nicotine use. At this point, interventions that target norms and attitudes might be especially impactful. This specificity of risk continues to characterize substance use into young adulthood. Taken together, our findings suggest multiple opportunities for prevention work across development, with different targets at different developmental stages.

Our study is the first to consider the longitudinal impact of youth RPFs with more proximal young adult RPFs on young adult ENDS use. Our work complements work on ENDS risk from the Population Assessment of Tobacco and Health (PATH) study (Hyland et al., 2017). Whereas the PATH study focuses on individuals living in an urban setting, our study sampled individuals who grew up in smaller communities. Our work also comes with some limitations. Although we assessed a broad array of RPFs, we did not examine ENDS-specific RPFs (i.e., norms, harm perceptions, flavor) prior to age 26 as these items were not added to the survey until that time point. Future research should assess and examine ENDS-specific RPFs, perhaps using measures that are similar to cigarette-specific RPFs such as attitudes, norms, and use environments. Another limitation is that today’s adolescents are coming of age while ENDS are already on the market, whereas our study sample was 14-15 years old when ENDS were introduced in 2007. This may somewhat limit generalizability of our findings to future generations of children who are at risk of using ENDS. However, this limitation may also represent a strength, as our work can inform understanding of risk and prevention for new drugs that will inevitably be introduced to people who did not grow up with them. We controlled analyses for race and ethnicity; however, the current study is unable to comment on reasons for racial and ethnic disproportionalities in nicotine use. Racial disproportionalities in ENDS use that remain after adjustment for socioeconomic status and other covariates might, for example, represent differences in the social environment of use, which could include being the recipient of targeted advertisements. Future research should investigate these questions in more depth.

Conclusions

In 2022, the FDA’s National Youth Tobacco Survey estimated that more than 3 million high school students used any nicotine product in the past month (U.S. Food and Drug Administration, 2023). Shifting norms, along with policy and prevention efforts, have made substantial progress in reducing cigarette use and its associated harms. As ENDS use becomes more entrenched for youth all over the world, both proven prevention strategies for the use of a variety of substances and new, targeted approaches for reducing ENDS use will be needed to address the ill health effects of nicotine use. Future work should build and test these updated interventions and continue to develop an evidence-based theoretical framework for minimizing the harms of all types of nicotine use.

Supplementary Material

Supplement

Funding.

This work was funded by grants from the National Cancer Institute (R37CA225690) and the National Institute on Drug Abuse (R01DA015183, R01DA044522). The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding agencies.

Appendix Table A. Full structural equation model results presented in Figure 1.

Outcome Predictor β [95% CI]
12-13 general protection Male gender −.12 [−.21, −.03]
Non-white race −.21 [−.33, −.10]
Hispanic ethnicity −.08 [−.21, .06]
Parental education .21 [.17, .24]
15-16 cigarette-specific risk 12-13 general protection −.53 [−.57, −.49]
15-16 friends’ use 12-13 general protection −.42 [−.46, −.38]
21-23 cigarette-specific risk 15-16 cigarette-specific risk .91 [.73, 1.08]
15-16 friends’ polysubstance use −.21 [−.39, −.03]
21-23 general 12-13 general protection .60 [.54, .65]
15-16 cigarette-specific risk −.21 [.35, −.08]
15-16 friends’ polysubstance use .09 [−.04, .21]
21-23 partner’s/friends’ use 15-16 cigarette-specific risk .27 [.12, .42]
15-16 friends’ polysubstance use .32 [.17, .47]
26 cigarette use Male gender .05 [−.04, .15]
Non-white race −.07 [−.20, .06]
Hispanic ethnicity −.14 [−.30, .01]
21-23 cigarette-specific risk .81 [.62, 1.00]
21-23 general protection −.04 [−.12, .04]
21-23 partner’s/friends’ polysubstance use −.13 [−.31, .06]
26 ENDS use Male gender .31 [.21, .41]
Non-white race −.18 [−.32, −.04]
Hispanic ethnicity −.13 [.30, .04]
21-23 cigarette-specific risk .26 [.07, .45]
21-23 general protection −.05 [−.14, .03]
21-23 partner’s/friends’ polysubstance use .23 [.06, .40]
Open in a new tab

Note. Standardized estimates are standardized with respect to latent variables (but not manifest variables). ENDS = electronic nicotine delivery system (e-cigarettes). Bolded rows are statistically significant at p < .05.

Footnotes

Ethical approval. All research protocols involving human participants were approved by the University of Washington Institutional Review Board (protocol #00005372) and were in accordance with the ethical standards of the 1964 Helsinki declaration and its later amendments.

Informed Consent. Consent was first obtained from parents of study participants in Grade 5 or Grade 6, and as participants grew older, participant consent was given at each subsequent wave by study participants themselves.

Conflict of Interest. The authors have no conflicts of interest to declare.

References

  1. Aladeokin A, & Haighton C (2019). Is adolescent e-cigarette use associated with smoking in the United Kingdom?: A systematic review with meta-analysis. Tobacco Prevention & Cessation, 5(15). 10.18332/tpc/108553 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Arnett JJ (2007). Emerging adulthood: What is it, and what is it good for? Child Development Perspectives, 1(2), 68–73. 10.1111/j.1750-8606.2007.00016.x [DOI] [Google Scholar]
  3. Bailey JA, Hill KG, Meacham MC, Young SE, & Hawkins JD (2011). Strategies for characterizing complex phenotypes and environments: General and specific family environmental predictors of young adult tobacco dependence, alcohol use disorder, and co-occurring problems. Drug and Alcohol Dependence, 118(2–3), 444–451. 10.1016/j.drugalcdep.2011.05.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bentivegna K, Atuegwu NC, Oncken C, DiFranza JR, & Mortensen EM (2021). Electronic cigarettes associated with incident and polysubstance use among youth. Journal of Adolescent Health, 68(1), 123–129. 10.1016/j.jadohealth.2020.05.026 [DOI] [PubMed] [Google Scholar]
  5. Carey FR, Rogers SM, Cohn EA, Harrell MB, Wilkinson AV, & Perry CL (2019). Understanding susceptibility to e-cigarettes: A comprehensive model of risk factors that influence the transition from non-susceptible to susceptible among e-cigarette naïve adolescents. Addictive Behaviors, 91, 68–74. 10.1016/j.addbeh.2018.09.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Catalano RF, & Hawkins JD (1996). The social development model: A theory of antisocial behavior. In Hawkins JD (Ed.), Delinquency and crime: Current theories (pp. 149–197). Cambridge University Press. [Google Scholar]
  7. Centers for Disease Control and Prevention. (2023). Economic trends in tobacco. cdc.gov/tobacco/data_statistics/fact_sheets/economics/econ_facts/index.htm
  8. Cicchetti D, & Rogosch FA (1999). Psychopathology as risk for adolescent substance use disorders: A developmental psychopathology perspective. Journal of Clinical Child Psychology, 28(3), 355–365. 10.1207/S15374424jccp280308 [DOI] [PubMed] [Google Scholar]
  9. Collins L, Glasser AM, Abudayyeh H, Pearson JL, & Villanti AC (2019). E-cigarette marketing and communication: How e-cigarette companies market e-cigarettes and the public engages with e-cigarette information. Nicotine & Tobacco Research, 21(1), 14–24. 10.1093/ntr/ntx284 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Committee on the Review of the Health Effects of Electronic Nicotine Delivery Systems, Board on Population Health and Public Health Practice, Health and Medicine Division, & National Academies of Sciences, Engineering, and Medicine, Stratton K, Kwan LY, & Eaton DL, Eds. (2018). Public health consequences of e-cigarettes. National Academies Press. 10.17226/24952 [DOI] [PubMed] [Google Scholar]
  11. Cornelius ME, Wang TW, Jamal A, Loretan CG, & Neff LJ (2020). Tobacco product use among adults—United States, 2019. Morbidity and Mortality Weekly Report, 69(46), 1736–1742. 10.15585/mmwr.mm6946a4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Cullen KA, Gentzke AS, Sawdey MD, Chang JT, Anic GM, Wang TW, Creamer MR, Jamal A, Ambrose BK, & King BA (2019). E-cigarette use among youth in the United States, 2019. JAMA, 322(21), 2095–2103. 10.1001/jama.2019.18387 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Daley DC, & Marlatt GA (2006). Overcoming your alcohol or drug problem: Effective recovery strategies, therapist guide. Oxford University Press. 10.1093/med:psych/9780195307733.001.0001 [DOI] [Google Scholar]
  14. Elder GH (1998). The life course as developmental theory. Child Development, 69(1), 1–12. 10.1111/j.1467-8624.1998.tb06128.x [DOI] [PubMed] [Google Scholar]
  15. Epstein M, Bailey JA, Kosterman R, Rhew IC, Furlong M, Oesterle S, & McCabe SE (2021). E-cigarette use is associated with subsequent cigarette use among young adult non-smokers, over and above a range of antecedent risk factors: A propensity score analysis. Addiction, 116(5), 1224–1232. 10.1111/add.15317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Epstein M, Hill KG, Bailey JA, & Hawkins JD (2013). The effect of general and drug-specific family environments on comorbid and drug-specific problem behavior: A longitudinal examination. Developmental Psychology, 49(6), 1151–1164. 10.1037/a0029309 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Frosch ZAK, Dierker LC, Rose JS, & Waldinger RJ (2009). Smoking trajectories, health, and mortality across the adult lifespan. Addictive Behaviors, 34(8), 701–704. 10.1016/j.addbeh.2009.04.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gelman A, & Hill J (2006). Data analysis using regression and multilevel/hierarchical models. Cambridge University Press. [Google Scholar]
  19. Gilbert PA, Kava CM, & Afifi R (2021). High-school students rarely use e-cigarettes alone: A sociodemographic analysis of polysubstance use among adolescents in the United States. Nicotine & Tobacco Research, 23(3), 505–510. 10.1093/ntr/ntaa037 [DOI] [PubMed] [Google Scholar]
  20. Glasser A, Abudayyeh H, Cantrell J, & Niaura R (2019). Patterns of e-cigarette use among youth and young adults: Review of the impact of e-cigarettes on cigarette smoking. Nicotine and Tobacco Research, 21(10), 1320–1330. 10.1093/ntr/nty103 [DOI] [PubMed] [Google Scholar]
  21. Hanewinkel R, & Isensee B (2015). Risk factors for e-cigarette, conventional cigarette, and dual use in German adolescents: A cohort study. Preventive Medicine, 74, 59–62. 10.1016/j.ypmed.2015.03.006 [DOI] [PubMed] [Google Scholar]
  22. Hawkins JD (2012). Sustained decreases in risk exposure and youth problem behaviors after installation of the Communities That Care prevention system in a randomized trial. Archives of Pediatrics & Adolescent Medicine, 166(2), 141–148. 10.1001/archpediatrics.2011.183 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hawkins JD, Brown EC, Oesterle S, Arthur MW, Abbott RD, & Catalano RF (2008). Early effects of Communities That Care on targeted risks and initiation of delinquent behavior and substance use. Journal of Adolescent Health, 43(1), 15–22. 10.1016/j.jadohealth.2008.01.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hawkins JD, Catalano RF, Kosterman R, Abbott R, & Hill KG (1999). Preventing adolescent health-risk behaviors by strengthening protection during childhood. Archives of Pediatrics & Adolescent Medicine, 153(3), 226–234. 10.1001/archpedi.153.3.226 [DOI] [PubMed] [Google Scholar]
  25. Hawkins JD, Catalano RF, & Miller JY (1992). Risk and protective factors for alcohol and other drug problems in adolescence and early adulthood: Implications for substance abuse prevention. Psychological Bulletin, 112(1), 64–105. 10.1037/0033-2909.112.1.64 [DOI] [PubMed] [Google Scholar]
  26. Hawkins JD, Kosterman R, Catalano RF, Hill KG, & Abbott RD (2005). Promoting positive adult functioning through social development intervention in childhood: Long-term effects from the Seattle Social Development Project. Archives of Pediatrics & Adolescent Medicine, 159(1), 25–31. 10.1001/archpedi.159.1.25 [DOI] [PubMed] [Google Scholar]
  27. Hu L, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. 10.1080/10705519909540118 [DOI] [Google Scholar]
  28. Hyland A, Ambrose BK, Conway KP, Borek N, Lambert E, Carusi C, Taylor K, Crosse S, Fong GT, Cummings KM, Abrams D, Pierce JP, Sargent J, Messer K, Bansal-Travers M, Niaura R, Vallone D, Hammond D, Hilmi N, … Compton WM (2017). Design and methods of the Population Assessment of Tobacco and Health (PATH) Study. Tobacco Control, 26(4), 371–378. 10.1136/tobaccocontrol-2016-052934 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Iacono WG, Malone SM, & McGue M (2008). Behavioral disinhibition and the development of early-onset addiction: Common and specific influences. Annual Review of Clinical Psychology, 4(1), 325–348. 10.1146/annurev.clinpsy.4.022007.141157 [DOI] [PubMed] [Google Scholar]
  30. Jha P, & Peto R (2014). Global effects of smoking, of quitting, and of taxing tobacco. New England Journal of Medicine, 370(1), 60–68. 10.1056/NEJMra1308383 [DOI] [PubMed] [Google Scholar]
  31. Kim S, & Selya AS (2020). The relationship between electronic cigarette use and conventional cigarette smoking is largely attributable to shared risk factors. Nicotine & Tobacco Research, 22(7), 1123–1130. 10.1093/ntr/ntz157 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kline RB (1998). Structural equation modeling. Guilford. [Google Scholar]
  33. Kosterman R, Epstein M, Bailey JA, Furlong M, & Hawkins JD (2021). The role of electronic cigarette use for quitting or reducing combustible cigarette use in the 30s: Longitudinal changes and moderated relationships. Drug and Alcohol Dependence, 227, 108940. 10.1016/j.drugalcdep.2021.108940 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kwon E, Seo D-C, Lin H-C, & Chen Z (2018). Predictors of youth e-cigarette use susceptibility in a U.S. nationally representative sample. Addictive Behaviors, 82, 79–85. 10.1016/j.addbeh.2018.02.026 [DOI] [PubMed] [Google Scholar]
  35. Lantz PM, Jacobson PD, Warner KE, Wasserman J, Pollack HA, Berson J, & Ahlstrom A (2000). Investing in youth tobacco control: A review of smoking prevention and control strategies. Tobacco Control, 9(1), 47–63. 10.1136/tc.9.1.47 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Livingston JA, Chen C-H, Kwon M, & Park E (2022). Physical and mental health outcomes associated with adolescent E-cigarette use. Journal of Pediatric Nursing, 64, 1–17. 10.1016/j.pedn.2022.01.006 [DOI] [PubMed] [Google Scholar]
  37. Maglia M, Caponnetto P, Di Piazza J, La Torre D, & Polosa R (2018). Dual use of electronic cigarettes and classic cigarettes: A systematic review. Addiction Research & Theory, 26(4), 330–338. 10.1080/16066359.2017.1388372 [DOI] [Google Scholar]
  38. Marques P, Piqueras L, & Sanz M-J (2021). An updated overview of e-cigarette impact on human health. Respiratory Research, 22(1), 151. 10.1186/s12931-021-01737-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Patil S, Arakeri G, Patil S, Ali Baeshen H, Raj T, Sarode SC, Sarode GS, Awan KH, Gomez R, & Brennan PA (2020). Are electronic nicotine delivery systems (ENDs) helping cigarette smokers quit?—Current evidence. Journal of Oral Pathology & Medicine, 49(3), 181–189. 10.1111/jop.12966 [DOI] [PubMed] [Google Scholar]
  40. R Core Team. (2013). R: A language and environment for statistical computing [Computer software]. R Foundation for Statistical Computing. [Google Scholar]
  41. Rosseel Y (2012). Lavaan: An R package for structural equation modeling and more. Version 0.5–12 (BETA). Journal of Statistical Software, 48(2), 1–36. 10.18637/jss.v048.i02 [DOI] [Google Scholar]
  42. Sawdey MD, Day HR, Coleman B, Gardner LD, Johnson SE, Limpert J, Hammad HT, Goniewicz ML, Abrams DB, Stanton CA, Pearson JL, Kaufman AR, Kimmel HL, Delnevo CD, Compton WM, Bansal-Travers M, Niaura RS, Hyland A, & Ambrose BK (2019). Associations of risk factors of e-cigarette and cigarette use and susceptibility to use among baseline PATH study youth participants (2013–2014). Addictive Behaviors, 91, 51–60. 10.1016/j.addbeh.2018.11.027 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Stanton CA, Tang Z, Sharma E, Seaman E, Gardner LD, Silveira ML, Hatsukami D, Day HR, Cummings KM, Goniewicz ML, Limpert J, Everard C, Bansal-Travers M, Ambrose B, Kimmel HL, Borek N, Compton WM, Hyland AJ, & Pearson JL (2023). Predictors of e-cigarette and cigarette use trajectory classes from early adolescence to emerging adulthood across four years (2013–2017) of the PATH Study. Nicotine & Tobacco Research, 25(3), 421–429. 10.1093/ntr/ntac119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Stone AL, Becker LG, Huber AM, & Catalano RF (2012). Review of risk and protective factors of substance use and problem use in emerging adulthood. Addictive Behaviors, 37(7), 747–775. 10.1016/j.addbeh.2012.02.014 [DOI] [PubMed] [Google Scholar]
  45. Thomas RE, McLellan J, & Perera R (2015). Effectiveness of school-based smoking prevention curricula: Systematic review and meta-analysis. BMJ Open, 5(3), e006976–e006976. 10.1136/bmjopen-2014-006976 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Tobler NS (1992). Drug prevention programs can work: Research findings. Journal of Addictive Diseases, 11(3), 1–28. 10.1300/J069v11n03_01 [DOI] [PubMed] [Google Scholar]
  47. Tyas SL, & Pederson LL (1998). Psychosocial factors related to adolescent smoking: A critical review of the literature. Tobacco Control, 7(4), 409–420. 10.1136/tc.7.4.409 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. U.S. Food and Drug Administration. (2023). Results from the Annual National Youth Tobacco Survey. https://www.fda.gov/tobacco-products/youth-and-tobacco/results-annual-national-youth-tobacco-survey
  49. Wickham H, Averick M, Bryan J, Chang W, McGowan LD, François R, Grolemund G, Hayes A, Henry L, & Hester J (2019). Welcome to the Tidyverse. Journal of Open Source Software, 4(43), 1686. 10.21105/joss.01686 [DOI] [Google Scholar]
  50. Wills TA, Soneji SS, Choi K, Jaspers I, & Tam EK (2021). E-cigarette use and respiratory disorders: An integrative review of converging evidence from epidemiological and laboratory studies. European Respiratory Journal, 57(1), 1901815. 10.1183/13993003.01815-2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Wray JM, Funderburk JS, Acker JD, Wray LO, & Maisto SA (2018). A meta-analysis of brief tobacco interventions for use in integrated primary care. Nicotine & Tobacco Research, 20(12), 1418–1426. 10.1093/ntr/ntx212 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement
NIHMS2039940-supplement-Supplement.docx (35.4KB, docx)

RESOURCES