Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Sep 1.
Published in final edited form as: Addict Behav. 2023 Apr 7;144:107713. doi: 10.1016/j.addbeh.2023.107713

National Youth Tobacco Surveys (2014–2019) Show Increasing Beliefs in the Harm and Relative Addiction of E-Cigarettes but Decreasing Associations Between Those Beliefs and E-Cigarette Use

Haijing Ma 1, Eric Gaudiello 2, Paschal Sheeran 2, Nora Sanzo 1, Erin L Sutfin 3, Seth M Noar 1,4,*
PMCID: PMC10330167  NIHMSID: NIHMS1893045  PMID: 37086609

Abstract

Background:

To evaluate the potential efficacy of increasing harm and relative addiction beliefs in discouraging e-cigarette use, we examined how adolescents’ beliefs about e-cigarettes have changed over 6 years and how the predictive validity of these beliefs has changed over time.

Methods:

Using data from the 2014–2019 National Youth Tobacco Survey (NYTS) (grades 6–12; N=117,472), we evaluated the association between adolescents’ beliefs about the harm and relative addiction of e-cigarettes and current e-cigarette use, as well as susceptibility to use. Logistic regressions and pairwise contrasts were used to analyze changes in these beliefs and determine how well these beliefs predict ever use, current use, and susceptibility to use over time.

Results:

E-cigarette harm and relative addiction beliefs tended to increase over time. In most years, these beliefs were negatively associated with e-cigarette use, including ever use, current use, and susceptibility to use. Interactions between these beliefs were also observed in some years such that harm belief better predicted use when e-cigarettes were also perceived as more addictive. Survey year also interacted with health harm and relative addiction beliefs such that the predictive validity of these beliefs for e-cigarette use decreased over time.

Conclusions:

Beliefs about e-cigarette harm and relative addiction have increased over time and predict use of, and susceptibility to, e-cigarettes among US adolescents. However, the predictive validity of these beliefs has decreased over time. Future research should explore the reasons for the decreased predictive validity of health beliefs in e-cigarette use and identify constructs that predict adolescent e-cigarette use over and above general harm and relative addiction beliefs.

Keywords: E-cigarette use, beliefs, adolescents, NYTS, harm, relative addiction

1. INTRODUCTION

Use of electronic cigarettes (e-cigarettes) among adolescents has become an epidemic in the United States (U.S. Department of Health and Human Services, 2018). According to the National Youth Tobacco Survey (NYTS), 1.4% of middle school students and 4.7% of high school students reported ever having used e-cigarettes in 2011 (Gentzke et al., 2019). By 2019, e-cigarette usage reached a record high, with 11% of middle school students and 28% of high school students having used e-cigarettes in the past 30 days (FDA, 2021).

E-cigarette use poses significant health risks for adolescents. E-cigarettes contain organic volatile compounds and metals that may impair users’ brain development and cause respiratory diseases (National Academies of Sciences, Engineering, and Medicine, 2018). The liquid that is heated to produce an aerosol contains nicotine that can cause addiction (U.S. Department of Health and Human Services. 2016). Moreover, e-cigarette use during adolescence is a precursor to use of other tobacco products (Bold et al., 2018) and predicts the amount and frequency of future use of other tobacco products (Khouja et al., 2021). Increasing e-cigarette use among youth threatens decades-long achievements in tobacco prevention and control.

Public health campaigns seeking to discourage e-cigarette use often attempt to increase adolescents’ beliefs about the harmfulness and addictiveness of e-cigarettes (FDA, 2020a, 2020b). For instance, the FDA began focusing on e-cigarette prevention in its national youth tobacco prevention campaign, “The Real Cost” in 2018. Since then, this campaign has sought to educate adolescents about the risks of e-cigarette use, including addiction and other health harms (FDA, 2021). In the same year, the Truth Initiative, a national public health organization, launched a campaign, “Safer ≠ Safe,” to correct possible misperceptions about e-cigarette risks.

Theoretical grounds for targeting e-cigarette beliefs to discourage use can be derived from the Health Belief Model and the Reasoned Action Approach which highlight the important role beliefs play in health behaviors (Sangalang et al., 2019). Key beliefs underpinning a focal behavior can be targeted in campaigns to promote behavior change (Barker et al., 2019; Brennan et al., 2012). Studies have shown that beliefs about e-cigarettes’ harm and addiction are negatively associated with adolescents’ e-cigarette use (Dobbs et al., 2017; Pfeiffer et al., 2020) and that many adolescents have low perceived harm and addiction of e-cigarettes (Gorukanti et al., 2017; Kong et al., 2015). However, it is unclear how adolescents’ beliefs about the harm and addiction of e-cigarettes have changed over time (Moustafa et al., 2021) and whether the role of health beliefs in discouraging use has changed.

Moreover, while studies have shown that both health harms and addiction messages independently discourage e-cigarette use and vaping (Noar et al., 2022), meta-analytic findings have shown a synergistic relationship between different elements of risk appraisals on intentions and behavior (Sheeran et al., 2014). Key theories also predict a synergistic effect among risk perceptions. For instance, the Extended Parallel Process Model (Witte, 1992) proposes that when both perceived severity and susceptibility are high, behavioral intentions should be higher than when only perceived severity or susceptibility are high. Thus, a synergistic effect between harm and relative addiction beliefs could be observed such that less e-cigarette susceptibility and use are observed when both harm and relative addiction appraisals are high. Examining this hypothesis would more effectively inform adolescent e-cigarette prevention messages and campaigns.

To evaluate the potential efficacy of increasing harm and relative addiction beliefs (i.e., the addictiveness of e-cigarettes vs. cigarettes) in discouraging e-cigarette use and to inform public health campaigns, the current study examined the strength of harm and relative addiction beliefs and their predictive validity for e-cigarette use among U.S. middle and high school students over six years (2014–2019). Specifically, this work addresses three research questions:

  1. How have e-cigarette harm and relative addiction beliefs changed over time?

  2. Do harm and relative addiction beliefs predict e-cigarette use and susceptibility to use?

  3. Has the predictive validity of harm and relative addiction beliefs changed over time?

2. METHODS

This study used data from the National Youth Tobacco Survey (NYTS), a nationally representative annual survey conducted by the CDC that assesses tobacco use in middle and high school students. We merged data from 2014, the first year to include measures related to e-cigarette use, to 2019, the most recent available dataset (N=117,472). In each year of the survey, there was a fairly even split on sex and representation across grades 6 – 12 (see Table S1 and S2).

2.1. Measures

2.1.1. Harm belief

Belief about the harmfulness of e-cigarettes was measured by “How much do you think people harm themselves when they use e-cigarettes some days but not every day?”, with 1=No harm; 2=Little harm; 3=Some harm; 4=A lot of harm.

2.1.2. Relative addiction belief1

The comparative addiction of e-cigarettes with cigarettes was assessed with “Do you believe that e-cigarettes are [less addictive, equally addictive, or more addictive] than cigarettes?”, scored 1, 2, and 3, respectively. This item was not included in the 2015 administration of the NYTS.

2.1.3. Ever use and current use

Ever use was measured by “Have you ever used an electronic cigarette or e-cigarette, even once or twice?” Current (past 30-day) use was measured with “During the past 30 days, on how many days did you use electronic cigarettes or e-cigarettes?”2 Prior to 2019, the current use measure had 7 response options (1=0 days; 7=All 30 days) whereas in 2019, the response became open-ended; participants had to specify a number between 0 and 30. This change led to a substantial increase in nonresponse for the item (67%), compared to ≤ 2% in each prior year. Given this rate of missing responses, it appears that most non-users did not realize they should enter a value of 0 and instead skipped the item. Consequently, we considered missing responses for 2019 to be non-users. Consistent with prior research (Wang et al., 2019), responses ≥ 1 were recoded to index current use.

2.1.4. Susceptibility to use e-cigarettes3

Susceptibility was assessed by three items on a 4-point Likert scale (1=Definitely yes; 2=Probably yes, 3=Probably not; 4=Definitely not): (1) “Have you ever been curious about using an e-cigarette?”; (2) “Do you think that you will try an electronic cigarette or e-cigarette soon?”, and (3) “If one of your best friends were to offer you an electronic cigarette or e-cigarette, would you use it?” Youth were considered susceptible if they did not answer “Definitely not” to all three items.

2.1.5. Demographics

Sex and grade-level were included as demographic control variables. Sex was female and male. Grade-level included all grades between 6th and 12th, and one combined option to indicate other or a lack of assigned grade-level.

2.2. Analytic plan

Ordinal logistic regression was used to examine whether e-cigarette harm and relative addiction beliefs changed over time. For harm, we reported the model implied mean rating; for relative addiction, we reported the model implied percentage of participants believing that e-cigarettes are more, equally, or less addictive than cigarettes. We regressed categorical dummy-coded variables for survey year, grade level, and sex on each outcome. Pairwise contrasts tested year-by-year changes in harm perception and relative addiction4.

Logistic regression was used to test for the predictive validity of harm and relative addiction beliefs on e-cigarette use: ever use, current use, and susceptibility to use. We regressed each outcome on the full set of nested terms in the three-way interaction between harm belief, relative addiction belief, and survey year (2014 as the reference category). Grade-level and sex were included as control variables. Pairwise contrasts tested changes in associations across years5. To simplify the interpretation of model coefficients, harm and relative addiction beliefs were centered; harm belief was mean-centered while relative addiction belief was contrast-centered so that zero reflected the belief that e-cigarettes were equally as harmful as cigarettes. All analyses were unweighted because of the complexities of incorporating survey weights into models that pool data across multiple years (Briesacher et al., 2012; McGee et al., 2020).

3. RESULTS

3.1. Change in harm and relative addiction beliefs over time

3.1.1. Harm belief

E-cigarette harm belief increased between 2014 and 2019. The odds that students reported higher harm belief increased in 2015 (OR=1.16, 95% CI [1.12, 1.21]) compared to 2014. In comparison to the prior year, harm belief continued to increase in 2016 (OR=1.32, 95% CI [1.27, 1.37]) and 2017 (OR=1.12, 95% CI [1.08, 1.17]) but decreased in 2018 (OR=0.95, 95% CI [0.91, 0.98]). In 2019, however, harm belief again increased (OR=1.39, 95% CI [1.34, 1.44]). The model implied mean harm rating of 2.95 in 2019 indicates that participants believed that periodic e-cigarette use causes “some harm” and represents a 17 % increase compared to the mean harm rating of 2.52 observed in 2014 (OR=2.27, 95% CI [2.19, 2.35]).

3.1.2. Relative addiction belief

E-cigarette relative addiction belief increased except in 2016–2017. The odds that students believed e-cigarettes were more addictive (relative to cigarettes) increased in 2016 compared to 2014 (OR=1.18, 95% CI [1.13, 1.23]). Relative addiction ratings increased again between 2017 and 2018 (OR=1.19, 95% CI [1.13, 1.25]). This trend continued in 2019 (OR=2.63, 95% CI [2.51, 2.75]). By 2019, only 18% of adolescents reported e-cigarettes were less addictive than cigarettes, whereas 35% and 26% reported they were equally or more addictive, respectively6 (See Table 1 and Figure 1).

Table 1.

Change in harm and relative addiction beliefs for each year compared to previous years

OR
 95% CI
 Z
 p
Harm belief
2015 vs. 2014  1.16  [1.12, 1.21]  8.01  <.001
2016 vs. 2015  1.32  [1.27, 1.37]  14.39  <.001
2017 vs. 2016  1.12  [1.08, 1.17]  6.16  <.001
2018 vs. 2017  0.95  [0.91, 0.98]  −2.81  .005
2019 vs. 2018  1.39  [1.34, 1.44]  17.71  <.001
Relative addiction belief
2016 vs. 2014  1.18  [1.13, 1.23]  7.11  <.001
2017 vs. 2016  1.00  [0.95, 1.04]  −0.20  .841
2018 vs. 2017  1.19  [1.13, 1.25]  6.86  <.001
2019 vs. 2018  2.63  [2.51, 2.75]  41.01  <.001
Open in a new tab
Figure 1.

Figure 1.

Open in a new tab

Model implied aggregate values for harm belief and relative addiction belief over time. The top panel shows the trend in the model implied mean for harm belief each year. The bottom panel shows the model implied probability for relative addiction belief each year.

3.2. Predictive validity of harm and relative addiction beliefs

3.2.1. Ever use

Harm and relative addiction beliefs both independently predicted ever use each year. Beginning in 2014, the odds that students reported ever having used e-cigarettes decreased as harm belief increased (OR=0.46, 95% CI [0.43, 0.48]). Over the next two years, harm belief was similarly predictive of ever use. Between 2017 (OR=0.43, 95% CI [0.40, 0.46]) and 2018 (OR=0.47, 95% CI [0.44, 0.49]), however, the association decreased (OR=1.08, 95% CI [1.00, 1.17]). Harm belief continued to become less predictive in 2019 (OR=0.54, 95% CI [0.52, 0.57]), with the magnitude of the association decreasing by 14.2% compared to 2014 (OR=1.16, 95% CI [1.09, 1.25]).

In 2014, the odds that students reported ever use decreased when they believed e-cigarettes were more addictive than cigarettes (OR=0.68, 95% CI [0.63, 0.75]). In each year following 2016, the predictiveness of relative addiction belief descriptively weakened; however, the year-by-year contrasts were non-significant. By 2019 (OR=0.76, 95% CI [0.73, 0.81]), the association between relative addiction and ever use decreased by 10.6% compared to 2014 (OR=1.14, 95% CI [1.02, 1.27]) (see Table 2, Table 3, and Figure 2).

Table 2.

Parameter estimates for harm belief, relative addiction belief, and the interaction term by year

Harm belief
 Relative addiction belief
 Harm x Addiction
 R2 Tjur
OR 95% CI Z p OR 95% CI Z p OR 95% CI Z p
Ever use 0.17
 2014 0.46 [0.43, 0.48] −28.38 < .001 0.68 [0.63, 0.75] −8.43 < .001 0.82 [0.76, 0.89] −4.62 < .001
 2016 0.45 [0.43, 0.47] −29.56 < .001 0.68 [0.63, 0.73] −10.38 < .001 0.84 [0.78, 0.91] −4.20 < .001
 2017 0.43 [0.40, 0.46] −27.71 < .001 0.71 [0.65, 0.77] −8.63 < .001 0.83 [0.76, 0.91] −4.10 < .001
 2018 0.47 [0.44, 0.49] −29.71 < .001 0.74 [0.69, 0.79] −9.18 < .001 0.94 [0.88, 1.01] −1.62 .106
 2019 0.54 [0.52, 0.57] −25.40 < .001 0.76 [0.73, 0.81] −10.16 < .001 0.96 [0.90, 1.01] −1.52 .128
Current use 0.12
 2014 0.39 [0.37, 0.43] −24.51 < .001 0.75 [0.66, 0.86] −4.28 < .001 0.78 [0.70, 0.88] −4.23 < .001
 2016 0.38 [0.35, 0.42] −23.08 < .001 0.72 [0.63, 0.81] −5.30 < .001 0.82 [0.73, 0.93] −3.11 .002
 2017 0.44 [0.40, 0.47] −19.52 < .001 0.85 [0.76, 0.96] −2.63 .008 0.96 [0.85, 1.09] −0.64 .525
 2018 0.45 [0.42, 0.48] −25.84 < .001 0.76 [0.70, 0.83] −6.44 < .001 0.86 [0.79, 0.94] −3.34 .001
 2019 0.55 [0.52, 0.58] −22.19 < .001 0.77 [0.72, 0.82] −8.69 < .001 0.94 [0.88, 1.01] −1.79 .073
Susceptibility to use 0.11
 2014 0.51 [0.49, 0.53] −28.29 < .001 0.81 [0.76, 0.87] −5.87 < .001 0.87 [0.81, 0.94] −3.84 < .001
 2016 0.48 [0.46, 0.50] −28.75 < .001 0.74 [0.69, 0.78] −9.36 < .001 0.89 [0.82, 0.96] −3.15 .002
 2017 0.48 [0.45, 0.50] −26.52 < .001 0.75 [0.70, 0.80] −8.08 < .001 0.91 [0.84, 0.99] −2.22 .026
 2018 0.46 [0.44, 0.48] −28.41 < .001 0.81 [0.76, 0.87] −6.25 < .001 0.94 [0.87, 1.01] −1.70 .090
 2019 0.53 [0.50, 0.56] −19.78 < .001 0.97 [0.91, 1.04] −0.79 .430 0.97 [0.89, 1.04] −0.90 .370
Open in a new tab

Note. R2 Tjur indicates the explanatory power of harm belief, relative addiction belief, and their interaction together with the control variable (i.e., sex and grade level) on the outcome.

R2 Tjur approaching 1 means high explanatory power.

Table 3.

Change in parameter estimates for harm belief, relative addiction belief, and the interaction term for each year compared to the previous year.

Harm belief
 Relative addiction belief
 Harm x Relative addiction
OR 95 °A CI Z p OR 95 % CI p OR 95 % CI p
Ever use
 2016 vs 2014 0.99 [0.91, 1.06] −0.37 0.713 1.00 [0.89, 1.12] −0.05 0.958 1.03 [0.91, 1.15] 0.43 0.664
 2017 vs 2016 0.96 [0.88, 1.03] −1.12 0.262 1.04 [0.93, 1.16] 0.69 0.489 0.98 [0.87, 1.11] −0.27 0.788
 2018 vs 2017 1.08 [1.00, 1.17] 2.02 0.043 1.05 [0.94, 1.16] 0.85 0.394 1.14 [1.01, 1.27] 2.18 0.030
 2019 vs 2018 1.16 [1.09, 1.25] 4.34 < 0.001 1.03 [0.95, 1.12] 0.79 0.431 1.01 [0.92, 1.11] 0.30 0.761
Current use
 2016 vs 2014 0.97 [0.87, 1.08] −0.56 0.578 0.95 [0.80, 1.14] −0.54 0.591 1.05 [0.89, 1.24] 0.61 0.542
 2017 vs 2016 1.14 [1.02, 1.28] 2.22 0.026 1.19 [1.01, 1.41] 2.02 0.044 1.17 [0.98, 1.38] 1.74 0.081
 2018 vs 2017 1.03 [0.93, 1.14] 0.51 0.609 0.89 [0.77, 1.03] −1.52 0.128 0.90 [0.77, 1.04] −1.44 0.151
 2019 vs 2018 1.23 [1.14, 1.33] 5.10 < 0.001 1.01 [0.91, 1.11] 0.13 0.894 1.09 [0.98, 1.22] 1.60 0.109
Susceptibility to use
 2016 vs 2014 0.94 [0.88, 1.01] −1.71 0.087 0.91 [0.82, 1.00] −2.03 0.043 1.02 [0.92, 1.13] 0.33 0.740
 2017 vs 2016 0.99 [0.92, 1.07] −0.26 0.796 1.02 [0.93, 1.12] 0.41 0.681 1.03 [0.92, 1.15] 0.49 0.621
 2018 vs 2017 0.97 [0.90, 1.04] −0.87 0.386 1.08 [0.98, 1.19] 1.62 0.106 1.03 [0.92, 1.15] 0.46 0.649
 2019 vs 2018 1.15 [1.06, 1.25] 3.35 0.001 1.20 [1.09, 1.32] 3.76 < 0.001 1.03 [0.93, 1.15] 0.56 0.577
Open in a new tab
Figure 2.

Figure 2.

Open in a new tab

Odds ratios as a function of year for each predictor and outcome. The y-axis is on a log scale to accurately represent the distance between odds ratios.

The main effects for harm and relative addiction belief were qualified by a two-way interaction (see Figure S1). Considering the two-way interaction for 2014 (OR=0.82, 95% CI [0.76, 0.89]), the association between harm belief and ever use was stronger for adolescents who perceived e-cigarettes as more addictive than cigarettes (OR=0.36, 95% CI [0.31, 0.41]) compared to those who saw e-cigarettes as equally addictive (OR=0.44, 95% CI [0.41, 0.46]) or less addictive (OR=0.53, 95% CI [0.49, 0.57]). That is, harm belief positively interacted with relative addiction belief such that harm belief became more predictive of ever use as belief in the relative addictiveness of e-cigarettes increased. Similar interactions were observed in both 2016 and 2017. The interaction effect became smaller in 2018 compared to the prior year (OR=1.14, 95% CI [1.01, 1.27]) and was non-significant in 2019. Overall, harm and relative addiction beliefs, and the interaction (together with the control variables), explained modest variance in ever use each year (R2 Tjur=0.17; see Tjur, 2009)7.

3.2.2. Current use

In 2014, the odds of being a current user decreased as harm belief increased (OR=0.39, 95% CI [0.37, 0.43]) and as relative addiction belief increased (OR=0.75, 95% CI [0.66, 0.86]). Harm (OR=0.38, 95% CI [0.35, 0.42]) and relative addiction belief (OR=0.72, 95% CI [0.63, 0.81]) were similarly predictive of current use in 2016, but both became less predictive in 2017. The association between current use and harm belief (OR=0.44, 95% CI [0.40, 0.47]) decreased by 12.4% and decreased by 16.1% for relative addiction (OR=0.85, 95% CI [0.76, 0.96]) (OR=1.19, 95% CI [1.01, 1.41]). In 2018, harm belief was similarly predictive of current use (OR=0.45), while relative addiction belief was not, and it did not meaningfully vary in 2019 (see Table 2). However, harm belief continued to become less predictive in 2019 (OR=1.23, 95% CI [1.14, 1.33]). The 2019 effect (OR=0.55, 95% CI [0.52, 0.58]) represented an 28.5% decrease in the association between harm belief and current use compared to 2014 (OR=1.46, 95% CI [1.32, 1.61]) (see Table 2, Table 3, and Figure 2).

These main effects in each survey year were qualified by a two-way interaction (see Figure S1). In 2014, harm belief was more strongly related to current use for adolescents who reported e-cigarettes were more addictive than cigarettes (OR=0.29, 95% CI [0.24, 0.35]) compared to those who endorsed equal addictiveness (OR=0.37, 95% CI [0.34, 0.41]) or less addictiveness (OR=0.48, 95% CI [0.43, 0.52]). The magnitude was similar in 2016. The harm x relative addiction interaction effect was not observed in 2017 (OR=0.96, 95% CI [0.85, 1.09]), but it returned in 2018 (OR=0.86, 95% CI [0.79, 0.94]). In 2019, the interaction again decreased in magnitude (OR=0.94, 95% CI [0.88, 1.01]). The variance explained in current use each year by harm and relative addiction beliefs, the interaction, and the control variables was modest (R2 Tjur=0.12).

3.2.3. Susceptibility to use e-cigarettes

Susceptibility to use decreased as harm belief increased (OR=0.51, 95% CI [0.49, 0.53]) and as relative addiction belief increased (OR=0.81, 95% CI [0.76, 0.87]) in 2014. Harm and relative addiction beliefs better predicted susceptibility the following year (2016), with the association for harm belief increasing by 6.1% (OR=0.94, 95% CI [0.88, 1.01]) and by 10.3% for relative addiction belief (OR=0.91, 95% CI [0.82, 1.00]). Over the next two years, harm belief was similarly predictive compared to 2016. Relative addiction belief had similar predictive validity in 2017 (OR=0.75, 95% CI [0.70, 0.80]) as in 2016, but the relationship weakened by 7.5% between 2017 and 2018 (OR=0.81; OR=1.08, 95% CI [0.98, 1.19]).

In 2019, harm and relative addiction beliefs became less predictive of susceptibility to use e-cigarettes. Specifically, the association between harm belief (OR=0.53, 95% CI [0.50, 0.56]) and susceptibility decreased by 13.2% (OR=1.15, 95% CI [1.06, 1.25]) compared to 2018, whereas it decreased by 16.5% for relative addiction belief (OR=1.20, 95% CI [1.09, 1.32]). This decrease meant that relative addiction belief were no longer associated with susceptibility to e-cigarette use in 2019 (OR=0.97, 95% CI [0.91, 1.04]) (see Table 2, Table 3, and Figure 2).

The main effects of harm and relative addiction beliefs in survey years were qualified by two-way interactions (see Figure S1). In 2014, harm belief better predicted susceptibility when e-cigarettes were seen as more addictive than cigarettes (OR=0.43, 95% CI [0.39, 0.48]) as compared to equally addictive (OR=0.50, 95% CI [0.47, 0.52]) or less addictive (OR=0.57, 95% CI [0.53, 0.61]). Similar trends were observed every year until 2019, with none of the year-on-year decreases in magnitude being statistically significant. No interaction was noted in 2019 (OR=0.97, 95% CI [0.89, 1.04]), indicating that the predictiveness of harm belief no longer depended on relative addiction belief (and vice versa). The variance explained in the outcome by harm and relative addiction beliefs, the interaction, and control variables was modest (R2 Tjur=0.11).

4. DISCUSSION

E-cigarette use among U.S. adolescents is noticeably high (Cooper et al., 2022). Public health campaigns often target adolescents’ perceptions of the harmfulness and addictiveness of vaping to discourage e-cigarette use (FDA, 2020a, 2020b), but the efficacy of this strategy needs testing to inform future campaigns. To inform public health campaigns and policy regulations, the current study investigated how adolescents’ beliefs about the harmfulness and relative addictiveness of e-cigarettes changed between 2014 and 2019 and whether the predictive validity of these beliefs has changed over time.

Our study showed that e-cigarette harm and relative addiction beliefs increased over time. Adolescents’ beliefs about the harmfulness of e-cigarettes increased by 17% in 2019 when compared to 2014, with the average belief being that they cause “some harm.” In 2019, only 18% of adolescents reported perceiving e-cigarettes as less addictive than cigarettes (compared to 30% in 2014) and 26% reported e-cigarettes as being more addictive (compared to 6% in 2014). These findings could seem heartening from a public health perspective as they demonstrate that adolescents’ perceptions of the harmfulness and relative addiction of e-cigarettes increased from 2014 to 2019. These increases may be due to a myriad of factors, including public health campaigns such as the FDA’s The Real Cost e-cigarette prevention campaign (Walker et al., 2022), news coverage of the e-cigarette or vaping product use associated lung injury (EVALI) outbreak (Morgan et al., 2021), the changing marketing and landscape of e-cigarette and vaping products, and other factors.

Harm and relative addiction beliefs were both negatively associated with use and susceptibility to use of e-cigarettes among U.S. adolescents in each year of data collection. These findings corroborate results from previous, smaller-scale studies (Dobbs et al., 2017; Pfeiffer et al., 2020). Our study also explored the synergistic effect of risk appraisals among harm and relative addiction beliefs, based on theoretical predictions and meta-analytic findings concerning risk appraisals (Sheeran et al., 2014; Witte, 1992). We observed significant interactions between harm and relative addiction beliefs on ever, current, and susceptibility to use from 2014 to 2017 such that harm belief became more predictive as relative addiction belief of e-cigarettes increased (and vice versa). Put differently, this finding suggests a synergistic relationship between harm and relative addiction beliefs such that the combination of both beliefs is associated with lower levels of current- and ever-use and susceptibility to use compared to either belief on its own (see also Zeller, 2019).

Despite increasing health beliefs over time, the study revealed a marked decrease in harm and relative addiction beliefs’ capacity to predict adolescents’ susceptibility to and use of e-cigarettes. Adolescents’ beliefs about the harmfulness of e-cigarettes tended to become less predictive of usage and susceptibility to use beginning in 2017. Similar declines in predictive validity were observed for relative addiction beliefs and the interaction between harm and relative addiction beliefs. These findings indicate that harm and relative addiction beliefs have become more weakly related to e-cigarette outcomes over time. The implication is that contemporary efforts to target these general beliefs may not be as effective in discouraging e-cigarette use as they would have been during 2014–2017. This trend may warrant concern as the decrease in the predictive validity of harm and relative addiction beliefs was observed even though rates of ever- and current use increased from 2014 to 2019.

Potential explanations of the decline in the predictive validity of health beliefs warrant discussion. The observed decline could be due to different types of measurement error such as a ceiling effect where variability in beliefs was not accurately captured because of the truncated scale. However, it is unlikely that this is the case based on the mean harm belief and the percentage of the sample reporting e-cigarettes as more addictive than cigarettes; there was room for change in health beliefs. Another possible explanation is that, with increasing exposure to information communicating the health risks of e-cigarettes, adolescents may have become habituated to this risk information in ways that blunt further changes in these beliefs (Anderson et al., 2016). It is also likely that factors other than health beliefs became more influential for adolescent e-cigarette use, for instance, perceived social norms concerning e-cigarette use (Gorukanti et al., 2017; Kong et al., 2015), positive outcome expectancies about e-cigarette use (e.g., reducing stress, providing satisfaction; Copeland et al., 2017; Fadus et al., 2019), and the ever changing marketing and e-cigarette product landscape (Kong et al., 2015). That is, the influence of concerns about harm and relative addiction on e-cigarette outcomes may be less influential as social influences and experiences using e-cigarettes become more influential. This conclusion is also supported by the fact that the predictive validity of these beliefs became much lower as overall usage increased.

The study is subject to several limitations. First, changes in NYTS survey items and survey administration could have affected the comparability of the NYTS from year to year. The change in the data collection mode (from paper and pencil in 2014–2018 to online in 2019) could have influenced responses, and the lack of data on relative addiction belief in 2015 precluded analysis of that year. Second, as some studies (Gentzke et al., 2019) have suggested, the inclusion of JUUL as an e-cigarette in the 2019 NYTS questionnaire could have affected the number of users that were not reported in previous years. Third, it is also the case that harm and relative addiction beliefs were measured by single items in the NYTS, and tests using reliable, specific, multi-item scales would offer valuable corroboration. Fourth, given the complexity of incorporating survey weights in ways that made the analyses comparable across multiple years of the NYTS, we did not incorporate survey weights in our analyses. Finally, a causal relationship between e-cigarette harm and relative addiction beliefs and use cannot be claimed due to the cross-sectional study design. Experimental evidence is needed to establish a causal relationship between health beliefs and behavior (see Noar et al., 2020, 2022; Pokhrel et al., 2019, for experimental studies of the impact of risk messages on intentions to use e-cigarettes, and Sheeran et al., 2014, for a review of experimental tests of threatening health information).

5. CONCLUSION

The present study reveals a paradox concerning adolescents’ beliefs about the harm and relative addiction of e-cigarettes. On one hand, adolescents’ beliefs that e-cigarettes are harmful and addictive have increased over time, which may be indicative of the success of public health campaigns in communicating the health risks of e-cigarette use. On the other hand, these beliefs became less predictive of susceptibility to, and use of, e-cigarettes over time, suggesting that the motivational force of these beliefs has declined. These findings imply that increasing harm and relative addiction beliefs about e-cigarettes could help to discourage adolescent e-cigarette use. However, the findings also indicate that promoting general harm and addiction consequences of e-cigarette use may be insufficient, or at least increasingly less impactful, in attempts to reduce e-cigarette use. Beliefs about other consequences (e.g., social, experiential, and affective consequences) and peer influence processes may outweigh health beliefs in adolescents’ contemporary decisions about e-cigarette experimentation and use (Conner et al., 2019; Fadus et al., 2019). It also may well be that more specific addiction and harm beliefs may better predict e-cigarette use as compared to the general measures examined here (Rohde et al., 2021; Sangalang et al., 2019). Further observational research is thus needed to identify the key predictors of adolescent e-cigarette use over and above general harm and relative addiction beliefs, and more intervention research is needed to specify how best to target these predictors in order to effectively combat e-cigarette use among youth.

Supplementary Material

1

Role of Funding Sources

This project was supported by grant number R01DA049155 from the National Institute on Drug Abuse and the FDA Center for Tobacco Products (CTP). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the Food and Drug Administration.

Footnotes

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

All authors have approved the final manuscript.

Endnotes

1

Students were also able to indicate that either (a) “I had never heard of e-cigarettes” or (b) “I don’t know enough about these products.” Except where explicitly noted, we treat such responses as missing for analytic purposes.

2

From 2014 to 2018, all participants saw the current use question, whereas for 2019 and 2020, only participants who indicated they had ever used an e-cigarette saw the current use question.

3

The 2019 survey meant that susceptibility items were only presented to participants who indicated that they had never used e-cigarettes, unlike prior years where all students were given the items. Therefore, analyses involving susceptibility only include those participants who reported never having used e-cigarettes. This is consistent with approaches used in related research (Wang et al., 2019; Margolis et al., 2016).

4

Because the relative addiction measure was omitted from the 2015 administration of the NYTS, 2014 was compared to 2016 for this variable; all other comparisons were annual.

5

Because the 2015 NYTS survey did not include a relative addiction belief measure, that year’s data was not used in these analyses.

6

This increased belief in the relative addiction of e-cigarettes appears to be partly driven by a reduction in the proportion of adolescents that reported not being knowledgeable enough about e-cigarettes to answer the question; whereas about one-third of adolescents reported insufficient awareness each year from 2014 to 2018, this proportion fell to 21% in 2019.

7

R2 Tjur is a standard measure of explanatory power bounded between 0 and 1, with R2 Tjur approaching 1 indicating there is large explanatory power.

References

  1. Ambrose BK, Rostron BL, Johnson SE, Portnoy DB, Apelberg BJ, Kaufman AR, & Choiniere CJ (2014). Perceptions of the relative harm of cigarettes and e-cigarettes among U.S. youth. American Journal of Preventive Medicine, 47(2), S53–S60. 10.1016/j.amepre.2014.04.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson BB, Vance A, Kirwan CB, Jenkins JL, & Eargle D (2016). From warning to wallpaper: Why the brain habituates to security warnings and what can be done about it. Journal of Management Information Systems, 33(3), 713–743. 10.1080/07421222.2016.1243947 [DOI] [Google Scholar]
  3. Barker JO, Kelley DE, Noar SM, Reboussin BA, Cornacchione Ross J, & Sutfin EL (2019). E-cigarette outcome expectancies among nationally representative samples of adolescents and young adults. Substance Use & Misuse, 54(12), 1970–1979. 10.1080/10826084.2019.1624773 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bold KW, Kong G, Camenga DR, Simon P, Cavallo DA, Morean ME, & Krishnan-Sarin S (2018). Trajectories of e-cigarette and conventional cigarette use among youth. Pediatrics, 141(1), e20171832. 10.1542/peds.2017-1832 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Brennan E, Momjian A, Jeong M, Naugle D, Parvanta S, & Hornik RC (2012). Mass media campaigns to reduce smoking among youth and young adults: Documenting potential campaign targets and reviewing the evidence from previous campaigns. Penn’s Center of Excellence in Cancer Communication Research, Annenberg School for Communication, University of Pennsylvania
  6. Briesacher B, Tjia J, Doubeni C, Chen Y, & Rao S (2012). Methodological issues in using multiple years of the Medicare current beneficiary survey. Medicare & Medicaid Research Review, 2(1), E1–E19. 10.5600/mmrr.002.01.A04 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Conner M, Grogan S, Simms‐Ellis R, Scholtens K, Sykes‐Muskett B, Cowap L, … & Siddiqi K (2019). Patterns and predictors of e‐cigarette, cigarette and dual use uptake in UK adolescents: Evidence from a 24‐month prospective study. Addiction, 114(11), 2048–2055. 10.1111/add.14723 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cooper M, Park-Lee E, Ren C, Cornelius M, Jamal A, & Cullen KA (2022). Notes from the field: E-cigarette use among middle and high school students—United States, 2022. Morbidity and Mortality Weekly Report, 71(40), 1283–1285. https://www.cdc.gov/mmwr/volumes/71/wr/mm7140a3.htm [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Copeland AL, Peltier MR, & Waldo K (2017). Perceived risk and benefits of e-cigarette use among college students. Addictive Behaviors, 71, 31–37. 10.1016/j.addbeh.2017.02.005 [DOI] [PubMed] [Google Scholar]
  10. Dobbs PD, Hammig B, & Henry LJ (2017). E-cigarette use among U.S. adolescents: Perceptions of relative addiction and harm. Health Education Journal, 76(3), 293–301. 10.1177/0017896916671762 [DOI] [Google Scholar]
  11. Fadus MC, Smith TT, & Squeglia LM (2019). The rise of e-cigarettes, pod mod devices, and JUUL among youth: Factors influencing use, health implications, and downstream effects. Drug and Alcohol Dependence, 201, 85–93. 10.1016/j.drugalcdep.2019.04.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Gentzke AS, Creamer M, Cullen KA, Ambrose BK, Willis G, Jamal A, & King BA (2019). Vital signs: Tobacco product use among middle and high school students — United States, 2011–2018. Morbidity and Mortality Weekly Report, 68(6), 157–164. 10.15585/mmwr.mm6806e1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Jun J, & Kim JK (2021). Do state regulations on e-cigarettes have impacts on the e-cigarette prevalence? Tobacco Control, 30(2), 221–226. 10.1136/tobaccocontrol-2019-055287 [DOI] [PubMed] [Google Scholar]
  14. Khouja JN, Suddell SF, Peters SE, Taylor AE, & Munafò MR (2021). Is e-cigarette use in non-smoking young adults associated with later smoking? A systematic review and meta-analysis. Tobacco Control, 30(1), 8–15. 10.1136/tobaccocontrol-2019-055433 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Kong G, Morean ME, Cavallo DA, Camenga DR, & Krishnan-Sarin S (2015). Reasons for electronic cigarette experimentation and discontinuation among adolescents and young adults. Nicotine & Tobacco Research, 17(7), 847–854. 10.1093/ntr/ntu257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Margolis KA, Nguyen AB, Slavit WI, & King BA (2016). E-cigarette curiosity among U.S. middle and high school students: Findings from the 2014 national youth tobacco survey. Preventive Medicine, 89, 1–6. 10.1016/j.ypmed.2016.05.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. McGee BT, Higgins MK, Phillips V, & Butler J (2020). Prescription drug spending and hospital use among Medicare beneficiaries with heart failure. Research in Social and Administrative Pharmacy, 16(10), 1452–1458. 10.1016/j.sapharm.2019.12.019 [DOI] [PubMed] [Google Scholar]
  18. Morgan JC, Silver N, & Cappella JN (2021). How did beliefs and perceptions about e-cigarettes change after national news coverage of the EVALI outbreak? PLoS One, 16(4), e0250908. 10.1371/journal.pone.0250908 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Moustafa AF, Rodriguez D, Mazur A, & Audrain-McGovern J (2021). Adolescent perceptions of E-cigarette use and vaping behavior before and after the EVALI outbreak. Preventive Medicine, 145, 106419. 10.1016/j.ypmed.2021.106419 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. National Academies of Sciences, Engineering, and Medicine. (2018). Public health consequences of e-cigarettes Washington, DC: The National Academies Press; [PubMed] [Google Scholar]
  21. Noar SM, Gottfredson NC, Kieu T, Rohde JA, Hall MG, Ma H, … & Brewer NT(2022). Impact of vaping prevention advertisements on US adolescents: A randomized clinical trial. JAMA Network Open, 5(10), e2236370–e2236370. 10.1001/jamanetworkopen.2022.36370 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Noar SM, Rohde JA, Prentice-Dunn H, Kresovich A, Hall MG, & Brewer NT (2020). Evaluating the actual and perceived effectiveness of e-cigarette prevention advertisements among adolescents. Addictive Behaviors, 109, 106473. 10.1016/j.addbeh.2020.106473 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Pfeiffer JA, Tompkins LK, Hart JL, Kesh A, Groom A, Vu THT, … & Walker KL (2020). Relationship between population characteristics, e-cigarette and tobacco-related perceptions, and likelihood of ever using e-cigarettes. Tobacco Prevention & Cessation, 6(20), published online, 10.18332/tpc/117477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Pokhrel P, Herzog TA, Fagan P, Unger JB, & Stacy AW (2019). E-cigarette advertising exposure, explicit and implicit harm perceptions, and e-cigarette use susceptibility among nonsmoking young adults. Nicotine and Tobacco Research, 21(1), 127–131. 10.1093/ntr/nty030 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Rohde JA, Vereen RN, & Noar SM (2021). Adolescents and young adults who vape or are susceptible to vaping: Characteristics, product preferences, and beliefs. Substance Use & Misuse, 56(11), 1607–1615. 10.1080/10826084.2021.1942052 [DOI] [PubMed] [Google Scholar]
  26. Sangalang A, Volinsky AC, Liu J, Yang Q, Lee SJ, Gibson LA, & Hornik RC (2019). Identifying potential campaign themes to prevent youth initiation of e-cigarettes. American Journal of Preventive Medicine, 56(2), S65–S75. 10.1016/j.amepre.2018.07.039 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sheeran P, Harris P, & Epton T (2014). Does heightening risk appraisals change people’s intentions and behavior? A meta-analytic review of the experimental evidence. Psychological Bulletin, 140(2), 511–543. 10.1037/a0033065 [DOI] [PubMed] [Google Scholar]
  28. Tjur T (2009). Coefficients of determination in logistic regression models — A new proposal: The coefficient of discrimination. The American Statistician, 63(4), 366–372. 10.1198/tast.2009.08210 [DOI] [Google Scholar]
  29. U.S. Department of Health and Human Services (2016). E-cigarette use among youth and young adults: A report of the Surgeon General U.S. Department of Health and Human Services, CDC. Atlanta, GA. [Google Scholar]
  30. U.S. Department of Health and Human Services (2018, December 18). Surgeon general releases advisory on e-cigarette epidemic among youth https://www.hhs.gov/about/news/2018/12/18/surgeon-general-releases-advisory-e-cigarette-epidemic-among-youth.html.
  31. U.S. Food and Drug Administration (2020a, April 30). Think e-cigs can’t harm teens’ health? https://www.fda.gov/tobacco-products/ctp-newsroom/fdas-comprehensive-plan-tobacco-and-nicotine-regulation
  32. U.S. Food and Drug Administration (2020b, September 17). Vaporizers, e-cigarettes, and other electronic nicotine delivery systems (ENDS) https://www.fda.gov/tobacco-products/products-ingredients-components/vaporizers-e-cigarettes-and-other-electronic-nicotine-delivery-systems-ends
  33. U.S. Food and Drug Administration (2021, February 26). The real cost campaign https://www.fda.gov/tobacco-products/public-health-education/real-cost-campaign
  34. Walker MW, Navarro M, Roditis M, & Dineva AN (2022). Adolescent risk perceptions of ENDS use: Room for change in tobacco education. Preventive Medicine Reports, 26, 101719. 10.1016/j.pmedr.2022.101719 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Wang TW, Gentzke AS, Creamer MR, Cullen KA, Holder-Hayes E, Sawdey MD, Anic GM, Portnoy DB, Hu S, Homa DM, Jamal A, & Neff LJ (2019). Tobacco product use and associated factors among middle and high school students — United States, 2019. MMWR. Surveillance Summaries, 68(12), 1–22. 10.15585/mmwr.ss6812a1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wang W, & Huang Y (2021). Countering the “harmless e-Cigarette” myth: The interplay of message format, message sidedness, and prior experience with e-cigarette use in misinformation correction. Science Communication, 43(2), 170–198. 10.1177/1075547020974384 [DOI] [Google Scholar]
  37. Witte K (1992). Putting the fear back into fear appeals: The extended parallel process model. Communications Monographs, 59(4), 329–349. 10.1080/03637759209376276 [DOI] [Google Scholar]
  38. Yang Q, Herbert N, Yang S, Alber J, Ophir Y, & Cappella JN (2020). The role of information avoidance in managing uncertainty from conflicting recommendations about electronic cigarettes. Communication Monographs, 1–23. 10.1080/03637751.2020.1809685 [DOI] [PMC free article] [PubMed]
  39. Zeller M (2019). Evolving “the real cost” campaign to address the rising epidemic of youth e-cigarette use. American Journal of Preventive Medicine, 56(2), S76–S78. 10.1016/j.amepre.2018.09.005 [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

1
NIHMS1893045-supplement-1.pdf (101KB, pdf)

RESOURCES