Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 May 1.
Published in final edited form as: Drug Alcohol Depend. 2018 Jan 31;186:257–263. doi: 10.1016/j.drugalcdep.2017.11.027

A longitudinal study of risk perceptions and e-cigarette initiation among college students: Interactions with smoking status

Maria Cooper 1, Alexandra Loukas 2, Kathleen R Case 1, C Nathan Marti 2, Cheryl L Perry 1
PMCID: PMC5911205  NIHMSID: NIHMS938609  PMID: 29626778

Abstract

Background

Recent data suggest that lower perceived risks of e-cigarettes are associated with e-cigarette use in young adults; however, the temporality of this relationship is not well-understood. We explore how perceptions of harmfulness and addictiveness of e-cigarettes influence e-cigarette initiation, and specifically whether this association varies by cigarette smoking status, in a longitudinal study of tobacco use on college campuses.

Methods

Data are from a 5-wave 24-college study in Texas. Only students who reported never using e-cigarettes at wave 1 were included (n=2,565). Multilevel discrete-time hazard models, accounting for school clustering, were used. The dependent variable, ever e-cigarette use, was assessed at each wave. Both time-varying (e-cigarette perceptions of harmfulness and addictiveness, age, use of cigarettes, use of other tobacco products, and use of other substances) and time-invariant demographic covariates were included. Two-way interactions between each e-cigarette perception variable and current conventional cigarette use were tested to determine if the hypothesized relationship differed among smokers and non-smokers.

Results

21% of all never e-cigarette users at baseline had initiated e-cigarette ever use by wave 5. Significant two-way interactions qualified the relationship between risk perceptions and e-cigarette initiation. Specifically, perceptions of a lower degree of harmfulness (OR=1.13, p=.047) and addictiveness (OR=1.34, p<.001) of e-cigarettes predicted initiation among non-smokers, but not among current smokers.

Conclusion

Perceiving a lower degree of risk of e-cigarettes contributes to subsequent e-cigarette initiation among non-smokers, but not among current smokers. Findings have implications for prevention campaigns focusing on the potential harm of e-cigarettes for non-smoking college students.

Keywords: electronic cigarettes, tobacco use, cigarette smoking, college students, risk perceptions

1. Introduction

Among adults in the United States (U.S.), young adults report the highest rates of electronic cigarette (“e-cigarette”) use; in fact, in 2013–2014, those aged 18–24 accounted for nearly one-quarter of adult e-cigarette users (Hu et al., 2016). The recent rise in prevalence of e-cigarette use among young adults has raised concerns about how these vulnerable populations perceive risks related to these products. While e-cigarettes are likely less harmful than combustible tobacco products, research on their health risks suggests that e-cigarettes are not harmless (U.S. Department of Health and Human Services, 2016). Risks of e-cigarette use to young adults include: increased susceptibility to nicotine addiction, as certain brain structures are still maturing during this period(Giedd and Rapoport, 2010; Somerville and Casey, 2010), exposure to e-cigarette aerosol, which contains carbonyl compounds and other chemicals known to be harmful to health (U.S. Department of Health and Human Services, 2016), and increased likelihood of transitioning to combustible tobacco products (Spindle et al., 2017). Several health behavior theories, such as the Theory of Planned Behavior and the Health Belief Model, emphasize the role of factors such as knowledge, attitudes and beliefs in influencing subsequent individual behaviors (Champion and Skinner, 2008; Montano and Kasprzyk, 2008). These theories highlight an individual’s beliefs about potential harm as a central component in predicting future behavior. The perception of e-cigarettes as less harmful than conventional cigarettes has been found to be particularly prevalent among young adults (Choi and Forster, 2013). In rankings of harmfulness of the full spectrum of tobacco products, e-cigarettes are consistently found to be rated as least harmful (Berg et al., 2015b; Roditis et al., 2016; Wackowski and Delnevo, 2016). Thus, it is imperative to understand risk perceptions as an important predictor for e-cigarette use in the young adult population.

While it is a well-established phenomenon that estimated risks, including perceptions about potential harms and addiction, are related to substance use, including conventional cigarettes (Murphy-Hoefer et al., 2004; Weinstein et al., 2005) and more recently, e-cigarettes (Amrock et al., 2015; Chaffee et al., 2015; Saddleson et al., 2015; Tan and Bigman, 2014), the temporality of this relationship is not as well-understood. Literature examining the associations between perceptions of harm and addictiveness and tobacco product use over time is limited. Studies have suggested that harm perceptions influence later conventional cigarette use in youth (Krosnick et al., 2006; Song et al., 2009), and lower relative perceived harm of e-cigarettes compared to cigarettes is predictive of subsequent e-cigarette use among current and former smokers aged 18 to over 55 (Brose et al., 2015). However, longitudinal studies examining alternative tobacco product use in college students or young adults are scarce, and a review of the limited literature reveals contrary findings (Berg et al., 2015a; Choi and Forster, 2014). For example, a study of college students examining longitudinal predictors of cigarette use and alternative tobacco use, separately, found that predictors of alternative tobacco use (a variable that included cigar products, hookah, smokeless tobacco and e-cigarettes), at one-year follow-up, included male gender and baseline smoking, alcohol, and alternative tobacco use. Lower perceived harm of alternative tobacco products at baseline only approached significance in predicting subsequent use (Berg et al., 2015a). However, in another prospective study of young adults, Choi and Forster (2014) found that lower harm perceptions of e-cigarettes relative to cigarettes predicted e-cigarette use at one-year follow up, but beliefs that e-cigarettes were less addictive than cigarettes did not (Choi and Forster, 2014). To our knowledge, Choi and Forster (2014) is the only prior study to date examining perceptions of e-cigarette addictiveness prospectively among young adults. Additional research is warranted to clarify whether perceptions of harm and addiction are indeed prospective risk factors for e-cigarette initiation among college students.

While the majority of the existing literature that examines the association between risk perceptions of e-cigarettes and e-cigarette use considers conventional cigarette use as a potential confounder in the relationship, few have considered cigarette smoking as a moderator. One exception is the study by Choi and Forster (2014), which found that harm perceptions assessed in 2010/2011 influenced subsequent e-cigarette use in 2011/2012 equally among young adult smokers and non-smokers (Choi and Forster, 2014). The degree to which the relationship between risk perceptions and subsequent e-cigarette use varies by smoking status warrants additional exploration, as it is likely that young adults who currently smoke cigarettes may not evaluate the risks associated with e-cigarettes in the same way as those who do not smoke.

The majority of previous research on tobacco risk perceptions utilizes measures of comparative risk (Berg et al., 2015b; Choi and Forster, 2014; Pearson et al., 2012; Smith et al., 2007; Tan and Bigman, 2014; Wackowski and Delnevo, 2016), whereas few studies measure the absolute risk of tobacco products. There is an emerging consensus that measures of comparative risk may mask quantifiable differences between perceptions of various tobacco products when they are compared to conventional cigarettes (Kaufman et al., 2016). Recent studies have underlined the importance of using metrics of absolute risk where tobacco product perceptions are measured in isolation and can yield more nuanced information (Popova and Ling, 2013; Wackowski et al., 2016). In addition, many youth and young adults are initiating tobacco use exclusively with e-cigarettes (Centers for Disease Control and Prevention, 2016; Trinidad et al., 2017). As such, it is important to determine how they perceive the risks of e-cigarettes alone, not just in comparison to cigarettes.

The current study aims to contribute to our limited understanding regarding how risk perceptions influence e-cigarette initiation. Specifically, we address two research questions: do perceptions of 1) harmfulness of e-cigarettes to health, and 2) addictiveness of e-cigarettes, predict future e-cigarette use up to two years later in a longitudinal study of tobacco use on college campuses? Further, we aim to test whether these relationships vary by current conventional cigarette smoking status. This manuscript extends previous work by Choi and Forster (2014) by: presenting more recent data collected after the nationwide surge in e-cigarette use in 2014; utilizing a larger sample; and implementing measures of absolute risk perceptions of e-cigarettes rather than comparative risk perceptions.

2. Materials and methods

2.1 Study design

This study consists of a longitudinal analysis of data collected from waves 1 – 5 from the “Marketing and Promotions across Colleges in Texas project (Project M-PACT).” Project M-PACT is a rapid response surveillance study (O’Connor et al., 2009), which utilizes a design intended to identify, characterize and monitor new and traditional tobacco products by collecting survey data at 6-month intervals among college students in the five counties surrounding the four largest cities in Texas (Austin, Dallas/Fort Worth, Houston, San Antonio).

2.2 Procedure

Eligible students attending 24 colleges were recruited to participate in the online survey via email invitation, which included a link to an eligibility survey. There were two eligibility criteria for participating in Project M-PACT. First, participants were required to be full- or part-time degree-seeking undergraduate students attending a four-year college or a vocational/technical program at a two-year college. Secondly, participants were required to be 18–29 years old. Overall, 13,714 students were eligible to participate in the study, and of these, 40% (n = 5,482) provided consent and completed the survey. Additional information regarding participant recruitment and incentives is reported elsewhere (Loukas et al., 2016).

The wave 1 survey was administered between November 2014 and February 2015 to 5,482 students. Subsequent survey waves were administered every six months, with retention rates ranging from 77.9–81.1% of the 5,482 students. The university’s institutional review board (IRB) approved this study’s protcol and procedures.

2.3 Participants

Participants were students attending one of 24 two- and four-year colleges. The current analysis was restricted to students aged 18–25 —a commonly-used age range for young adulthood in the tobacco control literature (U.S. Department of Health and Human Services, 2012) — who had complete data on at least one follow-up wave and had never used e-cigarettes at baseline (n=2,565). Retention rates of these 2,565 students were 89.3% at wave 2 (n=2,290), 88.2% at wave 3 (n=2,263), 90.9% at wave 4 (n=2,332), and 88.0% at wave 5 (n=2,258).

2.4 Measures

Study measures were reviewed by nine tobacco control experts. In addition, all final items were modified after an iterative process of cognitive interviewing with 25 young adults not part of the present study (Hinds et al., 2016).

2.4.1 E-cigarette initiation

The event of interest was e-cigarette initiation. At each of the five waves, participants were asked “Have you ever used an ENDS product, (i.e., e-cigarette, vape pen, or e-hookah) as intended (i.e., with nicotine cartridges and/or e-liquid/e-juice), even one or two puffs?”

2.4.2 Perceived harm of e-cigarettes

To study the prospective association between harm perceptions at the previous wave and e-cigarette initiation, the primary independent variable of interest was perceived harm of e-cigarettes. At each wave, participants were asked “How harmful are ENDS products to health?” Four numbered response options ranged from 1 to 4, and the extreme values were labeled: “1 - not at all harmful” and “4 - extremely harmful.” Responses were reverse coded to facilitate interpretation.

2.4.3 Perceived addictiveness of e-cigarettes

To study the prospective association between perceived addictiveness at the previous wave and e-cigarette initiation, the primary independent variable of interest was perceived addictiveness of e-cigarettes. At each wave, participants were asked “How addictive are ENDS products?” Three response options were: “1 - not at all addictive,” “2 – somewhat addictive” and “3 - very addictive.” Again, responses were reverse coded to facilitate interpretation.

2.4.4 Covariates

Four socio-demographic variables were included in the statistical models; sex (0=female/1=male), race/ethnicity, age in years and type of college attended (0=two-year/1=four-year). The race/ethnicity variable included five categories: White, non-Hispanic, Hispanic/Latino, African American, Asian, and Other.

The following covariates were also included in the model to control for effects of other behaviors associated with e-cigarette use (Saddleson et al., 2015). Dummy variables were created for: 1) past 30-day use of conventional cigarettes (0=no cigarette use/1=cigarette use); 2) past 30-day use of other tobacco products, including cigar products, hookah and smokeless tobacco (0=used no other tobacco/1=used one other tobacco product and 0=used no other tobacco/1=used two or more other tobacco products); and 3) current use of other substances, including past 30-day marijuana use and past 14-day binge drinking, which was defined as 5 or more drinks of alcohol in one day (0=used no other substances/1=used one other substance and 0=used no other substances/1=used two other substances).

2.4.5 Attrition analyses

A series of univariable logistic regression analyses examined differences in wave 1 study variables between students who were included in the present study (n=2,565) and those who were lost to attrition (n=189). Study variables were regressed on a dichotomous variable coded for attrition status (0=attrited/1=not attrited). We examined attrition for students included in the survival analyses compared to those who were excluded due to not completing at least one follow up wave. Results revealed no statistically significant differences for the following variables at baseline: sex, age, type of college attended, cigarette smoking, use of other tobacco products, use of one other substance, perceptions of e-cigarette harmfulness to health or perceptions of e-cigarette addictiveness. There were significant group differences in race/ethnicity where those reporting Asian race were more likely to have completed at least one follow up wave compared to those who were not of Asian race (96.1% vs. 92.3%, p=.001), and those reporting using two other substances were less likely to have completed at least one follow up wave compared to those who did not (85.7% vs. 93.4%, p=.003).

2.5 Statistical analysis

Descriptive statistics for the variables of interest at baseline were calculated. Statistical tests (t-tests for continuous variables and chi-square tests for categorical variables) compared participants who initiated use of e-cigarettes during any of the study periods to those who did not. Multilevel discrete-time hazard models (Singer and Willett, 2003) were used to test the research questions examining perceptions of 1) harmfulness of e-cigarettes to health and 2) addictiveness of e-cigarettes in future e-cigarette initiation. Discrete-time models are applied when the exact time of the event (i.e., e-cigarette initiation) cannot be identified. In the present study, e-cigarette initiation occurred within a six-month interval between two of the study waves; thus, models contained four discrete time periods, which were the four six-month periods in between the five survey assessments. The model was fit to a person-period data set that contained a dummy variable for each of the four six-month periods and did not contain an intercept.

Estimated hazards, i.e., the probability that an individual will initiate e-cigarette use while that individual is at risk, were calculated for each time interval (Rabe-Hesketh and Skrondal, 2012). To estimate the effects of covariates on the hazard, both time-invariant (race/ethnicity, sex, type of college) and time-varying predictors (perceptions of harm and addictiveness, age, current use of cigarettes, current use of other tobacco products, and current use of other substances) were included as fixed effects. The discrete-time hazard model is implemented by coding time periods in a no-intercept logistic regression model and thus does not have a hazard function typical of most survival models (e.g., Cox proportional hazard models). Singer and Willett (2003) recommend the odds transformation which compares the magnitude of the probability that the event will occur and the probability that it will not occur; as such, model covariate effects are reported as odds ratios:

Estimatedodds=hazard1-hazard

Both variables measuring perceptions of harm and addictiveness were lagged one time period to obtain the appropriate temporal ordering based on our study questions. Other time-varying covariates were not lagged one time period as our study purpose was not to determine if tobacco and other substance use predicted later e-cigarette initiation, but rather to control for the effects of concomitant tobacco and substance use, which may co-occur within the same time period as e-cigarette initiation. Respondents were nested within their baseline college or university and the time parameters for each study period were treated as random effects.

A single model with both independent variables of interest related to e-cigarette perceptions yielded multicollinearity as Variance Inflation Factors (VIFs) for both variables were greater than 10 (Neter et al., 1990). As such, each perception variable was entered into separate models resulting in four models testing the two research questions. Variables were entered in two steps. For model 1, survey period, harm perceptions and all other covariates were entered. For model 2, the interaction term between harm perceptions and current smoking was added to the previous model (model 1) to determine if the relationship varied by current smoking status. This two-way interaction was calculated as the product of harm perceptions and current cigarette smoking (0=no, 1=yes). Similarly, for model 3, survey period, perceptions of addictiveness and all other covariates were entered, and model 4 added a two-way interaction between perceptions of addictiveness and current smoking to the previous model (model 3). Statistically significant interactions were probed (Aiken et al., 1991) using p<.05 as the threshold for significance to determine the nature of the interaction. Interaction estimates, which are ratios of odds ratios, are reported as exp(β), consistent with recommendations from Norton et al. (2004).

We elected to not impute missing data due to the fact that conventional multiple imputation software does not accommodate event timing and censoring (Allison, 2010). For each model, missing data for the variables of interest ranged from 1.70% to 1.75%, and as such, the sample size for each model varied minimally. This is well below the threshold of 10% of missing data to warrant imputation. The inclusion of a variety of covariates also ameliorates missing data bias and represents an acceptable solution to non-response bias (Gelman and Hill, 2007). All analyses were conducted using STATA 14.0 (College Station, TX).

3. Results

3.1 Descriptive statistics

Between waves 2 and 5, 518 of the 2,565 young adults initiated e-cigarette use. Descriptive statistics for the independent variables and other covariates are presented in Table 1.

Table 1.

Descriptive Statistics at Baseline (Wave 1) – (n=2,565)

All participants (n=2,565) Never E-cigarette Users (n=2,047) E-cigarette Initiators (n=518)
Age
Mean (95% CI)		 20.0
(20.0 – 20.1)		 20.0
(19.9 – 20.1)		 20.2*
(20.1 – 20.4)
Sex
% female (95% CI)		 66.4%
(64.6% – 68.2%)		 67.7%
(65.6% – 69.7%)		 61.6%**
(57.3% – 65.7%)
Race
% (95% CI)
 Non-Hispanic White 33.9%
(32.1% – 35.7%)		 34.6%
(32.66% – 36.7%)		 31.1%
(27.2% – 35.2%)
 Hispanic/Latino 27.9%
(26.2% – 29.6%)		 26.5%
(24.6% – 28.4%)		 33.4%**
(29.5% – 37.6%)
 African American 8.5%
(7.5% – 9.6%)		 8.7%
(7.5% – 10.0%)		 7.7%
(5.7% – 10.4%)
 Asian 22.3%
(20.8% – 24.0%)		 23.0%
(21.2% – 24.8%)		 19.9%
(16.7% – 23.6%)
 Other 7.4%
(6.5% – 8.5%)		 7.3%
(6.2% – 8.5%)		 7.9%
(5.9% – 10.6%)
School Level
% 4-year (95% CI)		 93.5%
(92.4% – 94.3%)		 93.5%
(92.3% – 94.4%)		 93.4%
(90.9% – 95.3%)
Perceived Harmfulness to Health of E- cigarettesa
Mean (95% CI)		 2.8
(2.8 – 2.9)		 2.9
(2.8 – 2.9)		 2.8*
(2.7 – 2.8)
Perceived Addictiveness of E- cigarettesb
Mean (95% CI)		 2.3
(2.2 – 2.3)		 2.3
(2.3 – 2.3)		 2.2***
(2.1 – 2.2)
Current Cigarette Use
% yes (95% CI)		 5.1%
(4.3% – 6.0%)		 2.5%
(1.9% – 3.3%)		 15.4%***
(12.6% – 18.8%)
Other Tobacco Product Usec
% yes (95% CI)
 1 product 6.0%
(5.1% – 7.0%)		 3.9%
(3.1% – 4.8%)		 14.3%***
(11.5% – 17.6%)
 2+ products 1.3%
(0.9% – 1.8%)		 0.6%
(0.3% – 1.0%)		 4.1***
(2.7% – 6.1%)
Other Substance Used
% yes (95% CI)
 1 substance 16.5%
(15.1% – 18.0%)		 12.5%
(11.1% – 14.0%)		 32.4%***
(28.5% – 36.6%)
 2 substances 3.5%
(2.9% – 4.3%)		 2.1%
(1.6% – 2.8%)		 9.1%***
(6.9% – 11.9%)
Open in a new tab

Note: Chi –square tests (categorical variables) and t-tests (continuous variables) compared never e-cigarette users to e-cigarette initiators

***

p<.001

**

p <.01

*

p<.05

a

Measured on a scale from 1-Not at all harmful to 4-Extremely harmful

b

Measured on a scale from 1-Not at all addictive to 3-Very addictive

c

Includes past-30 day use of: cigar products, hookah and smokeless tobacco

d

Includes past-30 day use of marijuana and past 14-day binge drinking

3.2 Life table

The distribution of e-cigarette initiation over the study period is shown in a life table (Table 2). By the end of wave 5, 21% of all e-cigarette-naïve young adults had initiated e-cigarette use. The hazard presented for each time period represents the probability of e-cigarette initiation in that period, given the participant had not previously initiated e-cigarette use. Survival estimates (i.e., the probability that a participant does not experience the event) are also presented in Table 2.

Table 2.

Life Table

Interval Time Beginning Total Number of E-cigarette Users Survival (95% CI) Cumulative Failure (95% CI) Hazard (95% CI)
Period 1: Wave 1 – Wave 2 2565 197 .92 (.91 – .93) .08 (.07 – .09) .08 (.07 – .09)
Period 2: Wave 2 – Wave 3 2298 130 .87 (.86 – .88) .13 (.12 – .14) .06 (.05 – .07)
Period 3: Wave 3 – Wave 4 2112 129 .82 (.80 – .83) .18 (.17 – .20) .06 (.05 – .07)
Period 4: Wave 4 – Wave 5 1858 62 .79 (.77 – .81) .21 (.12 – .23) .03 (.03 – .04)
Open in a new tab

3.3 Discrete-time survival analysis

3.3.1 Perceived harm

In model 1, lower perceived harm was not statistically significantly associated with e-cigarette initiation in the subsequent time period. Current cigarette use, other tobacco use and other substance use were statistically significantly associated with increasing odds of e-cigarette initiation. In model 2, a statistically significant interaction effect was found between current conventional cigarette use and perceived harm. Probing the interaction using the methods outlined by Aiken and West (1991) indicated that the effect varied by smoking status. Specifically, lower perceived harm of e-cigarettes was associated with greater odds of initiation among non-smokers (exp(β)=1.13, p=.047), but there was no statistically significant association among current smokers (exp(β)=0.77, p=.062).

3.3.2 Perceived addictiveness

In model 3, lower perceived addictiveness was associated with increased odds of e-cigarette initiation in the subsequent time period (OR=1.26; p=.003). Again, current cigarette use, other tobacco use and other substance use were statistically significantly associated with increasing odds of e-cigarette initiation. In model 4, a statistically significant interaction effect was found between current conventional cigarette use and lower perceived addictiveness, indicating that the effect varied by smoking status. Specifically, lower perceived addictiveness of e-cigarettes was associated with greater odds of initiation among non-smokers (exp(β)=1.34, p<.001), but there was no statistically significant association among current smokers (exp(β)=0.90, p=.553).

4. Discussion

This study found that 21% of e-cigarette-naïve young adults initiated e-cigarette use during young adulthood, highlighting that e-cigarette initiation is not limited to adolescents. Further, this study provides evidence for the effect of low perceptions of harmfulness to health and addictiveness on e-cigarette initiation among a sample of college students. After testing interactions with current smoking status, moderating effects were found for both perceived harmfulness to health and perceived addictiveness. These results revealed that lower perceptions on both variables predicted e-cigarette initiation among non-smoking college students, but not among college students currently smoking cigarettes. Potential explanations for the differential findings for non-smokers versus smokers include 1) the role of nicotine dependence and 2) the use of e-cigarettes for smoking cessation.

One possibility for the lack of impact of perceived harm of e-cigarettes for cigarette users is that current smokers likely already exhibit symptoms of nicotine dependence (Moran et al., 2004). While nicotine dependence is a complex construct and is not necessarily directly related to risk perceptions, it may outweigh other risk factors, including risk perceptions in the initiation of other nicotine products such as e-cigarettes. Conversely, for non-smokers, consistent with the theoretical frameworks provided by the Health Belief Model and Theory of Planned Behavior, lower risk perceptions may be important factors in the initiation of e-cigarette use. Recent research conducted by Loukas et al. (2016) supports the premise that nicotine dependence is associated with polytobacco use among college students. Specifically, researchers found that one symptom of cigarette dependence (i.e., ever needing a cigarette) was significantly associated with increased odds of current polytobacco use (Loukas et al., 2016). While the study collapsed other alternative tobacco use (cigars, hookah, e-cigarettes), e-cigarettes were among the most prevalent alternative tobacco products used. In addition, the present study’s findings are supported by previous research that found that the association between other psychosocial risk factors and e-cigarette use differed by current smoking status among Texas college students (Case et al., 2017). Namely, higher sensation seeking was associated with increased odds of e-cigarette use for non-current cigarette smokers but not for current smokers (Case et al., 2017). Ultimately, more research is needed to determine the moderating effect of cigarette smoking status on the association between risk factors (e.g., harm perceptions, perceptions of addictiveness) and e-cigarette use. Specifically, future research should examine whether symptoms of nicotine dependence modify the association between harm perceptions and e-cigarette use in young adults.

The use of e-cigarettes as cessation devices offers another potential explanation for the differential findings for the associations between harm perceptions and perceived addictiveness and e-cigarette initiation by smoking status. While the effectiveness of e-cigarettes as tools for smoking cessation is hotly debated (Kalkhoran and Glantz, 2016), perceptions of e-cigarettes as effective smoking cessation devices are frequently cited as reasons for using e-cigarettes among youth and young adults (Camenga et al., 2016; Kong et al., 2014). In a recent study that explored factors associated with using e-cigarettes to quit smoking, researchers found that 41.8% of current youth and young adult smokers reported ever using e-cigarettes to quit smoking (Camenga et al., 2016). Importantly, harm perceptions of e-cigarettes were not associated with using e-cigarettes as cessation devices after controlling for demographic and tobacco use behaviors. Ultimately, our results may be consistent with the Camenga et al. (2016) study that found that harm perceptions are not an important factor in the use of e-cigarettes as cessation devices for current smokers. It is important to note, however, that we did not ask whether e-cigarettes were used for the purposes of smoking cessation. Future research should further explore this hypothesis by examining the specific associations between perceptions of harm and addictiveness and the initiation of e-cigarette use for cessation purposes.

The results from the current study differ from previous longitudinal research that examined the relationship between risk perceptions and e-cigarette use among young adults. While Berg et al. (2015) found no significant association between perceived relative harm of alternative tobacco products and subsequent use, Choi and Forster (2014) reported an association between lower perceived relative harm of e-cigarettes compared to cigarettes and future use of e-cigarettes, but none for perceived relative addictiveness and no evidence of a moderating effect for current smokers in either relationship. The current study’s findings may be explained by several factors. First, unlike the previously cited studies, the current study measures young adults’ perceptions of harmfulness to health and addictiveness on an absolute scale, rather than comparing them to conventional cigarettes. These measurement differences may have accounted for the lack of a positive, statistically significant relationship between e-cigarette perceptions and initiation among smokers. In the current study, absolute perceived risk may not have been an important predictor of e-cigarette initiation among smokers, because regardless of their perceived level of absolute risk, it is plausible that smokers still viewed e-cigarettes as less harmful relative to cigarettes. Second, the statistically significant interactions found in the current study could be the result of a larger sample size (n=2,565) compared to Choi and Forster (2014)’s sample (n=1,379), resulting in more power to detect such effects.

Until recently, e-cigarettes were not regulated by the federal government, and manufacturers were not required to list harmful or potentially harmful product constituents, which may contribute to perceptions about their safety. The Food and Drug Administration (FDA)’s final “deeming rule,” published in August of 2016, requires tobacco manufacturers to report such ingredients. In addition, the rule also requires e-cigarette manufacturers to include pre-determined health warnings—including: “WARNING: This product contains nicotine. Nicotine is an addictive chemical.”—on e-cigarette packaging and advertisements (Food and Drug Administration, 2016). Future longitudinal research is needed to ascertain how these new requirements may affect e-cigarette risk perceptions over time, especially among young populations who have not yet initiated use, for whom these perceptions are a significant risk factor.

The FDA’s deeming rule also prohibits false or misleading statements about e-cigarettes on product labeling or in ads (Food and Drug Administration, 2016). Our findings on the prospective association of risk perceptions on later e-cigarette uptake provide scientific support for the continued need for this regulation. Risk perceptions may be affected by environmental influences such as marketing and advertising. E-cigarettes are advertised on social media platforms such as Facebook and Twitter, both of which have a large number of young adult users. In 2015, 88% of 18–21 years old had seen an e-cigarette advertisement via at least one of the following media: TV, radio, print, online and retail (The Truth Initiative, 2015). Further, the vast majority of e-cigarette retail websites either implies or includes overt health claims, including references that e-cigarettes are healthier products or assist with smoking cessation (Grana and Ling, 2014).

Our results also have implications for tobacco prevention and messaging, which should include e-cigarettes. Study findings emphasize the importance of prevention programs for college students, as approximately one-fifth of our e-cigarette-naïve students used e-cigarettes for the first time during young adulthood. These results support the need to implement tobacco prevention programs that include alternative tobacco products such as e-cigarettes, not only during adolescence, but also during young adulthood, targeting young populations before they begin using these products. Additionally, health communication campaigns should communicate the risks of e-cigarettes to non-smokers in this age range, a demographic that is not insignificant. According to the 2015 National Health Interview Survey, among e-cigarette users aged 18–24, 40% had never smoked cigarettes (Centers for Disease Control and Prevention, 2016).

This study has strengths and limitations which are inherent in its longitudinal design. Conclusions about the temporality of relationships between risk factors and initiation of tobacco products are limited in cross-sectional studies, which currently represent the majority of the literature on risk perceptions and tobacco use. The longitudinal nature of our data allows us to better understand the temporality of such relationships. However, we must caution against interpreting the results too broadly. As the study design was not experimental, results should not be interpreted as causal. In addition, results cannot be applied to observations that were left censored (i.e., those who had already initiated e-cigarette use at wave 1 and were removed from the analysis). We cannot determine the role of e-cigarette perceptions on their initiation, and these students who were earlier adopters may have different risk profiles than the students who were included in the study. Finally, attrition is a form of bias present in longitudinal research. Attrition analyses revealed that students reporting Asian race were less likely to have attrited and therefore are overrepresented in the study, while those who reported recent marijuana use and binge drinking were more likely to have attrited and were underrepresented in the study.

5. Conclusions

In this sample of e-cigarette-naïve college students, perceptions about the harmfulness and addictiveness of e-cigarettes influenced e-cigarette initiation over a two-year period among non-smokers of cigarettes, but not among current smokers. Findings emphasize the need for e-cigarette prevention programs and health communication messages targeted to young adult non-smokers.

Table 3.

The Effect of Low Perceived Harm and Addictiveness on E-cigarette Initiation (Multilevel Discrete-time Hazard Models)

Perceived Harm
Model 1		 Perceived Harm with Interaction
Model 2		 Perceived Addictiveness
Model 3		 Perceived Addictiveness with Interaction
Model 4
Covariate Parameter Estimate Odds Ratio (95% CI) Parameter Estimate Odds Ratio (95% CI) Parameter Estimate Odds Ratio (95% CI) Parameter Estimate Odds Ratio (95% CI)
Period 1 −3.54 0.03*** (0.01 – 0.10) −3.68 0.03*** (0.01 – 0.09) −3.81 0.02*** (0.01 – 0.08) −3.89 0.02*** (0.01 – 0.07)
Period 2 −3.85 0.02*** (0.01 – 0.08) −4.00 0.02*** (0.01 – 0.07) −4.12 0.02*** (0.00 – 0.06) −4.21 0.01*** (0.00 – 0.05)
Period 3 −3.78 0.02*** (0.01 – 0.08) −3.93 0.02*** (0.01 – 0.07) −4.05 0.02*** (0.00 – 0.06) −4.13 0.02*** (0.00 – 0.06)
Period 4 −4.52 0.01*** (0.00 – 0.04) −4.67 0.01*** (0.00 – 0.04) −4.79 0.01*** (0.00 – 0.03) −4.87 0.01*** (0.00 – 0.03)
Perceived harm (T-1) 0.06 1.06 (0.95 – 1.18) .12 1.13* (1.00 – 1.27) -- -- -- --
Perceived addictiveness (T-1) -- -- -- -- 0.23 1.26** (1.08 – 1.46) 0.30 1.34*** (1.14 – 1.58)
Current Cigarette Use x Perceived harm (T-1) -- -- −0.37 0.69* (0.51 –0.92) -- -- -- --
Current Cigarette Use x Perceived Addictiveness (T-1) -- -- -- -- -- -- −0.41 0.67* (0.45 –0.99)
Age 0.02 1.02 (0.97 – 1.08) 0.02 1.02 (0.97 – 1.08) 0.02 1.02 (0.97 – 1.08) 0.02 1.02 (0.97 – 1.08)
Sex −0.12 0.89 (0.73 – 1.09) −0.11 0.89 (0.73 – 1.09) −0.12 0.89 (0.73 – 1.08) −0.13 0.88 (0.72 –1.07)
Race/ethnicity
 Non-Hispanic White (ref) -- 1.00 -- 1.00 -- 1.00 -- 1.00
 Hispanic/Latino 0.34 1.41** (1.10 – 1.79) 0.34 1.40** (1.10 – 1.79) 0.31 1.36 (1.07 – 1.74) 0.31 1.36* (1.07 – 1.73)
 African- American 0.03 1.03 (0.69 – 1.53) 0.04 1.04 (0.70 – 1.55) 0.00 1.00 (0.67 – 1.49) 0.00 1.00 (0.67 – 1.49)
 Asian 0.08 1.08 (0.82 – 1.42) 0.10 1.10 (0.84 – 1.45) 0.05 1.05 (0.80 – 1.39) 0.05 1.05 (0.80 – 1.39)
 Other 0.31 1.36 (0.94 – 1.97 0.32 1.38 (0.95 – 2.00) 0.30 1.35 (0.93 – 1.95) 0.30 1.36 (0.93 – 1.97)
School level −0.05 .95 (0.64 – 1.42) −0.06 0.95 (0.63 – 1.41) −0.06 0.95 (0.63 – 1.41) −0.07 0.94 (0.63 – 1.40)
Current Cigarette Use 1.05 2.84*** (2.09 – 3.86) 1.85 6.39*** (3.21 – 12.73) 1.03 2.80*** (2.06 – 3.80) 1.79 5.99*** (2.70 – 13.26)
Other Tobacco
 None (ref) -- 1.00 -- 1.00 -- 1.00 -- 1.00
 1 product 1.08 2.93*** (2.21 – 3.89) 1.09 2.98*** (2.25 – 3.96) 1.07 2.92*** (2.20 – 3.87) 1.06 2.90*** (2.18 – 3.84)
 2+ products 1.30 3.69*** (2.12 – 6.44) 1.40 4.04*** (2.31 – 7.05) 1.30 3.65*** (2.09 – 6.38) 1.33 3.80*** (2.18 – 6.61)
Other Substances
 None (ref) -- 1.00 -- 1.00 -- 1.00 -- 1.00
 1 substance 0.92 2.53*** (2.03 – 3.15) 0.93 2.53*** (2.03 – 3.15) 0.91 2.49*** (2.00 – 3.11) 0.91 2.50*** (2.00 – 3.11)
 2 substances 1.21 3.35*** (2.38 – 4.73) 1.21 3.35*** (2.37 – 4.71) 1.19 3.27*** (2.32 – 4.61) 1.19 3.31*** (2.35 – 4.66)
Open in a new tab
*

p<.05

**

p<.01

***

p<.001

Bolded indicates statistical significance, p<.05

Highlights.

  • We examine prospective risk factors of e-cigarette use in college students

  • Among never users at wave 1, 21% initiated e-cigarette use over the two year period

  • Lower risk perceptions predicted e-cigarette use for non-smoking college students

  • Results support need to highlight risks of e-cigarette use in non-smoking students

Acknowledgments

Role of Funding Source

This work was supported by grant number [1 P50 CA180906-01] from the National Cancer Institute at the National Institutes of Health and the Food and Drug Administration, Center for Tobacco Products (CTP). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Food and Drug Administration.

Footnotes

Contributors

All authors contributed to the study’s concept and design. MC conducted literature searches and provided summaries of previous research studies. MC and CNM conducted the statistical analysis. MC, AL, KRC and CNM wrote the first draft of the manuscript and all authors contributed to and have approved the final manuscript.

Conflict of Interests

No conflict declared.

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 citable 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.

References

  1. Aiken LS, West SG, Reno RR. Multiple regression: Testing and interpreting interactions. Newbury Park; Sage: 1991. [Google Scholar]
  2. Allison PD. Survival Analysis. In: Hancock GR, Mueller RO, Stapleton LM, editors. The reviewer’s guide to quantitative methods in the social sciences. Routledge; New York: 2010. pp. 413–432. [Google Scholar]
  3. Amrock SM, Zakhar J, Zhou S, Weitzman M. Perception of e-cigarettes’ harm and its correlation with use among U.S. adolescents. Nicotine Tob Res. 2015;17:330–336. doi: 10.1093/ntr/ntu156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Berg CJ, Romero DR, Pulvers K. Perceived harm of tobacco products and individual schemas of a smoker in relation to change in tobacco product use over one year among young adults. Subst Use Misuse. 2015a;50:90–98. doi: 10.3109/10826084.2014.958858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Berg CJ, Stratton E, Schauer GL, Lewis M, Wang Y, Windle M, Kegler M. Perceived harm, addictiveness, and social acceptability of tobacco products and marijuana among young adults: Marijuana, hookah, and electronic cigarettes win. Subst Use Misuse. 2015b;50:79–89. doi: 10.3109/10826084.2014.958857. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brose LS, Brown J, Hitchman SC, McNeill A. Perceived relative harm of electronic cigarettes over time and impact on subsequent use. A survey with 1-year and 2-year follow-ups. Drug Alcohol Depend. 2015;157:106–111. doi: 10.1016/j.drugalcdep.2015.10.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Camenga DR, Kong G, Cavallo DA, Krishnan-Sarin S. Current and former smokers’ use of electronic cigarettes for quitting smoking: An exploratory study of adolescents and young adults. Nicotine Tob Res. 2017;19:1531–1535. doi: 10.1093/ntr/ntw248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Case K, Loukas A, Harrell M, Wilkinson A, Springer A, Pérez A, Creamer M, Perry CL. The association between sensation seeking and e-cigarette use in Texas young adults: A cross-sectional study. J Am Coll Health. 2017;65:277–285. doi: 10.1080/07448481.2017.1282487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Centers for Disease Control and Prevention. QuickStats: Cigarette smoking status among current adult e-cigarette users, by age group — National Health Interview Survey, United States, 2015. MMWR Morb Mortal Wkly Rep. 2016:65. doi: 10.15585/mmwr.mm6542a7. [DOI] [PubMed] [Google Scholar]
  10. Chaffee BW, Gansky SA, Halpern-Felsher B, Couch ET, Essex G, Walsh MM. Conditional risk assessment of adolescents’ electronic cigarette perceptions. Am J Health Behav. 2015;39:421–432. doi: 10.5993/AJHB.39.3.14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Champion VL, Skinner CS. In: The Health Belief Model. In Health Behavior and Health Education: Theories, Research, and Practice. Glanz K, Rimer BK, Viswanath K, editors. San Francisco, CA: Jossey Bass; 2008. [Google Scholar]
  12. Choi K, Forster J. Characteristics associated with awareness, perceptions, and use of electronic nicotine delivery systems among young US Midwestern adults. Am J Public Health. 2013;103:556–561. doi: 10.2105/AJPH.2012.300947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Choi K, Forster JL. Beliefs and experimentation with electronic cigarettes: A prospective analysis among young adults. Am J Prev Med. 2014;46:175–178. doi: 10.1016/j.amepre.2013.10.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Food and Drug Administration. Deeming tobacco products to be subject to the Federal Food, Drug, and Cosmetic Act, as amended by the Family Smoking Prevention and Tobacco Control Act; restrictions on the sale and distribution of tobacco products and required warning statements for tobacco products. Federal Registrar 81 FR. 2016:28973–29106. [PubMed] [Google Scholar]
  15. Gelman A, Hill J. Data analysis using regression and multilevelhierarchical models. Cambridge University Press; New York, NY, USA: 2007. [Google Scholar]
  16. Giedd JN, Rapoport JL. Structural MRI of pediatric brain development: What have we learned and where are we going? Neuron. 2010;67:728–734. doi: 10.1016/j.neuron.2010.08.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Grana RA, Ling PM. “Smoking revolution”: A content analysis of electronic cigarette retail websites. Am J Prev Med. 2014;46:395–403. doi: 10.1016/j.amepre.2013.12.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hinds JT, Loukas A, Chow S, Pasch KE, Harrell MB, Perry CL, Delnevo C, Wackowski OA. Using cognitive interviewing to better assess young adult e-cigarette use. Nicotine Tob Res. 2016;18:1998–2005. doi: 10.1093/ntr/ntw096. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hu SS, Neff L, Agaku IT, Cox S, Day HR, Holder-Hayes E, King BA. Tobacco product use among adults — United States, 2013–2014. MMWR Morb Mortal Wkly Rep. 2016:685–691. doi: 10.15585/mmwr.mm6527a1. [DOI] [PubMed] [Google Scholar]
  20. Kalkhoran S, Glantz SA. E-cigarettes and smoking cessation in real-world and clinical settings: A systematic review and meta-analysis. Lancet Resp Med. 2016;4:116–128. doi: 10.1016/S2213-2600(15)00521-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Kaufman AR, Suls JM, Klein WM. Communicating tobacco product harm: Compared to what? Addict Behav. 2016;52:123–125. doi: 10.1016/j.addbeh.2015.06.039. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kong G, Morean ME, Cavallo DA, Camenga DR, Krishnan-Sarin S. Reasons for electronic cigarette experimentation and discontinuation among adolescents and young adults. Nicotine Tob Res. 2014;17:847–854. doi: 10.1093/ntr/ntu257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Krosnick JA, Chang L, Sherman SJ, Chassin L, Presson C. The effects of beliefs about the health consequences of cigarette smoking on smoking onset. J Commun. 2006;56:S18–S37. [Google Scholar]
  24. Loukas A, Chow S, Pasch KE, Li X, Hinds JT, Iii, Marti CN, Harrell MB, Creamer MR, Perry CL. College students’ polytobacco use, cigarette cessation, and dependence. Am J Health Behav. 2016;40:514–522. doi: 10.5993/AJHB.40.4.13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Montano DE, Kasprzyk D. Theory of reasoned action, theory of planned behavior, and the integrated behavioral model. In: Glanz K, Rimer BK, Viswanath K, editors. Health behavior and health education: Theory, research and practice. 4. Jossey-Bass; San Francisco: 2008. [Google Scholar]
  26. Moran S, Wechsler H, Rigotti NA. Social smoking among US college students. Pediatrics. 2004;114:1028–1034. doi: 10.1542/peds.2003-0558-L. [DOI] [PubMed] [Google Scholar]
  27. Murphy-Hoefer R, Alder S, Higbee C. Perceptions about cigarette smoking and risks among college students. Nicotine Tob Res. 2004;6:371–374. doi: 10.1080/14622200412331320770. [DOI] [PubMed] [Google Scholar]
  28. Neter J, Wasserman W, Kutner MH. Applied linear models, regression, analysis of variance and experimental designs. RD Irwin; Boston, USA: 1990. [Google Scholar]
  29. Norton EC, Wang H, Ai C. Computing interaction effects and standard errors in logit and probit models. Stata J. 2004;4:154–167. [Google Scholar]
  30. O’Connor RJ, Cummings KM, Rees VW, Connolly GN, Norton KJ, Sweanor D, Parascandola M, Hatsukami DK, Shields PG. Surveillance methods for identifying, characterizing, and monitoring tobacco products: Potential reduced exposure products as an example. Cancer Epidem Biomar. 2009;18:3334–3348. doi: 10.1158/1055-9965.EPI-09-0429. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Pearson JL, Richardson A, Niaura RS, Vallone DM, Abrams DB. e-Cigarette awareness, use, and harm perceptions in US adults. Am J Public Health. 2012;102:1758–1766. doi: 10.2105/AJPH.2011.300526. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Popova L, Ling PM. Perceptions of relative risk of snus and cigarettes among US smokers. Am J Public Health. 2013;103:e21–23. doi: 10.2105/AJPH.2013.301547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Rabe-Hesketh S, Skrondal A. Multilevel and longitudinal modeling using stata. StataCorp LP; 2012. [Google Scholar]
  34. Roditis M, Delucchi K, Cash D, Halpern-Felsher B. Adolescents’ perceptions of health risks, social risks, and benefits differ across tobacco products. J Adolesc Health. 2016;58:558–566. doi: 10.1016/j.jadohealth.2016.01.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Saddleson ML, Kozlowski LT, Giovino GA, Hawk LW, Murphy JM, MacLean MG, Goniewicz ML, Homish GG, Wrotniak BH, Mahoney MC. Risky behaviors, e-cigarette use and susceptibility of use among college students. Drug Alcohol Depend. 2015;149:25–30. doi: 10.1016/j.drugalcdep.2015.01.001. [DOI] [PubMed] [Google Scholar]
  36. Singer JD, Willett JB. Applied longitudinal data analysis: Modeling change and event occurrence. Oxford university press; 2003. [Google Scholar]
  37. Smith SY, Curbow B, Stillman FA. Harm perception of nicotine products in college freshmen. Nicotine Tob Res. 2007;9:977–982. doi: 10.1080/14622200701540796. [DOI] [PubMed] [Google Scholar]
  38. Somerville LH, Casey BJ. Developmental neurobiology of cognitive control and motivational systems. Curr Opin Neurobiol. 2010;20:236–241. doi: 10.1016/j.conb.2010.01.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Song AV, Morrell HE, Cornell JL, Ramos ME, Biehl M, Kropp RY, Halpern-Felsher BL. Perceptions of smoking-related risks and benefits as predictors of adolescent smoking initiation. Am J Public Health. 2009;99:487–492. doi: 10.2105/AJPH.2008.137679. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Spindle TR, Hiler MM, Cooke ME, Eissenberg T, Kendler KS, Dick DM. Electronic cigarette use and uptake of cigarette smoking: A longitudinal examination of US college students. Addict Behav. 2017;67:66–72. doi: 10.1016/j.addbeh.2016.12.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Tan AS, Bigman CA. E-cigarette awareness and perceived harmfulness: Prevalence and associations with smoking-cessation outcomes. Am J Prev Med. 2014;47:141–149. doi: 10.1016/j.amepre.2014.02.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. The Truth Initiative. Vaporized: Youth and young adult exposure to e-cigarette marketing. 2015 Accessed at: https://truthinitiative.org/research/vaporized-majority-youth-exposed-e-cigarette-advertising.
  43. Trinidad DR, Pierce JP, Sargent JD, White MM, Strong DR, Portnoy DB, Green VR, Stanton CA, Choi K, Bansal-Travers M. Susceptibility to tobacco product use among youth in wave 1 of the Population Assessment of Tobacco and Health (PATH) study. Prev Med. 2017;101:8–14. doi: 10.1016/j.ypmed.2017.05.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. U.S. Department of Health and Human Services. Preventing tobacco use among youth and young adults: A report of the surgeon general. Centers for Disease Control and Prevention, Office on Smoking and Health; Atlanta, GA: 2012. [PubMed] [Google Scholar]
  45. U.S. Department of Health and Human Services. A report to the surgeon general. U.S. Department of Health and Human Services, Public Health Service, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health; Atlanta, GA: 2016. E-cigarette Use among youth and young adults. [Google Scholar]
  46. Wackowski OA, Bover Manderski MT, Delnevo CD. Comparison of direct and indirect measures of e-cigarette risk perceptions. Tob Regul Sci. 2016;2:38–43. doi: 10.18001/TRS.2.1.4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wackowski OA, Delnevo CD. Young adults’ risk perceptions of various tobacco products relative to cigarettes: Results from the National Young Adult Health Survey. Health Educ Behav. 2016;43:328–336. doi: 10.1177/1090198115599988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Weinstein ND, Marcus SE, Moser RP. Smokers’ unrealistic optimism about their risk. Tob Control. 2005;14:55–59. doi: 10.1136/tc.2004.008375. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES