Abstract
Nicotine vaping has risen sharply among emerging adults, emphasizing the need to clarify psychosocial factors underlying regular and problematic use. This study examined whether emotion dysregulation (ED) is associated with vaping among university students and whether descriptive norms (perceived peer use) and injunctive norms (perceived peer approval) account for indirect associations linking ED and vaping. Canadian undergraduates (N = 723; Mage = 19.28, SD = 2.08; 78.1% female; 37.8% non-Caucasian) completed an online survey assessing vaping, perceived norms, and emotion dysregulation. ED was significantly associated with both past-30-day and problematic vaping. Students consistently overestimated peer vaping and approval. Descriptive norms for “typical students” showed significant indirect effects linking ED with both outcomes, whereas norms for same-gender peers, close friends, and all injunctive norms did not. Prevention efforts may benefit from combining emotion regulation skill-building with interventions targeting misperceived vaping norms among university students.
Introduction
Decades of research on combustible cigarette smoking have established a substantial evidence base documenting its health consequences and multifaceted biopsychosocial determinants (DeAtley et al., 2020; Kang & Choi, 2025). This foundation has shaped public understanding, attitudes, and behaviors surrounding tobacco use (Cummings & Proctor, 2014; U.S. Department of Health and Human Services, 2014; World Health Organization, 2024). However, nicotine consumption patterns have shifted markedly with the rise of electronic cigarettes (e-cigarettes), battery-powered devices that aerosolize liquid solutions for inhalation. Initially marketed as reduced-harm tools to support adult smoking cessation, e-cigarettes have rapidly gained popularity among individuals with no prior history of combustible cigarette use, particularly adolescents and emerging adults, who now represent the demographic with the highest prevalence of use (Jackson et al., 2024; Statistics Canada, 2023; U.S. Department Health Service, Office of the Surgeon General, 2016). This trend has prompted concern that e-cigarettes may constitute a new route to nicotine initiation (Chen et al., 2023; Jackson et al., 2024; Martinelli et al., 2023; O’Brien et al., 2021) and may contribute to broader patterns of polysubstance involvement, given evidence of frequent co-use with alcohol and cannabis (Czoli et al., 2023; Zuckermann et al., 2019).
Mounting evidence now links e-cigarette use (vaping) to a range of adverse health outcomes, including respiratory and cardiovascular dysfunction, oxidative stress, and systemic inflammation (Izquierdo-Condoy et al., 2024). Risks are compounded by the proliferation of largely unregulated devices and e-liquids with variable, insufficiently characterized toxicological profiles (Marques et al., 2021; Ren et al., 2022; Sachdeva et al., 2023) as well as emerging findings that e-cigarette dependence levels may approximate those associated with combustible cigarettes (Kundu et al., 2025). These concerns are especially salient during emerging adulthood, a developmental period marked by ongoing neurobiological maturation, elevated reward sensitivity, and heightened exposure to academic, social, and occupational demands (Arnett, 2005; Meehan et al., 2024; Varma et al., 2021). In Canada, where vaping among adults aged 20−24 more than doubled between 2013 and 2022, most recent users have never smoked combustible cigarettes, suggesting that vaping may now serve as a primary pathway to nicotine initiation (Statistics Canada, 2013, 2023). Within university contexts, the intersection of autonomy, peer influence, and institutional pressures may further intensify vulnerability to use (Hiler et al., 2020; Nabil et al., 2024; Oh et al., 2021). Therefore, in the absence of robust regulatory oversight and longitudinal evidence, understanding the modifiable psychosocial mechanisms shaping vaping among emerging adults remains essential for guiding prevention and intervention efforts.
Emotion Dysregulation and Vaping
Over the recent years, accumulating research has begun to link vaping to various mental health issues, such as heightened anxiety, depression, psychological distress, and impulsivity, particularly among adolescents and young adults (Becker et al., 2021; Ibrahim et al., 2025; Kava et al., 2024). However, one especially promising, yet under-examined transdiagnostic construct is emotion dysregulation (ED). ED refers to difficulties in understanding, accepting, and flexibly managing emotional experiences, including challenges in modulating emotional responses, controlling impulsive reactions, and employing adaptive strategies, such that emotional states interfere with effective, goal-directed behavior (Gratz & Roemer, 2004, 2008; Gross, 1998). Notably, systematic reviews and meta-analyses demonstrate a robust association between ED and substance use across alcohol, tobacco, cannabis, and other substances (Chrétien et al., 2025; González-Roz et al., 2024; Mansueto et al., 2024; Stellern et al., 2023).
The link between ED and substance use can be further understood through the Self-Medication Hypothesis (SMH), which proposes that individuals who struggle to regulate their emotions may use substances to alleviate underlying emotional or psychological distress. Such individuals may be particularly vulnerable due to heightened reactivity to negative affect and reduced capacity to regulate these states effectively (Hall & Queener, 2007; Khantzian, 1997). Consistent with this framework, emerging evidence demonstrates a reliable association between ED and vaping-related cognitions and behaviors. Individuals with greater ED report higher vaping dependence, stronger positive and negative affect-related expectancies (McLeish et al., 2024; Zvolensky et al., 2021), and greater perceived barriers to quitting (Manning et al., 2020). Additionally, research indicates that negative urgency and related emotional vulnerabilities increase susceptibility to vaping indirectly through heightened emotion-regulation difficulties (Reff & Baschnagel, 2021) and predict stronger beliefs that vaping alleviates negative affect, beliefs that subsequently increase the likelihood of vaping among university students (Brockenberry et al., 2022). Despite these findings, this work represents only an initial step in clarifying the role of ED in vaping behavior, and little is known about the mechanisms that drive this relationship.
Perceived Social Norms and Vaping
One potential mechanism linking emotion dysregulation to vaping is the influence of perceived social norms. Perceived social norms can be defined as cognitive-affective representations in which individuals’ beliefs about the attitudes and behaviors of important others, including both perceptions of what others do (descriptive norms) and what others approve or disapprove of (injunctive norms; Berkowitz, 2004; Bergquist & Ekelund, 2025; Borsari & Carey, 2003; Cialdini et al., 1990, 1991). According to Social Norms Theory, individuals frequently misperceive the prevalence and acceptability of peers’ substance use, such that overestimations of descriptive norms and injunctive norms can promote alignment with these perceived norms, often exerting a stronger influence on behavior than objective patterns of use (Berkowitz, 2004, 2005, Neighbors et al., 2007, 2008, 2016b; Perkins & Berkowitz, 1986). These misperceptions may be especially consequential for health-risk behaviors, where perceived permissiveness may function as a cognitive justification for use. Drawing on Social Cognitive Theory, such perceived norms may be understood as social cues that shape outcome expectancies – that is, anticipated emotional, social, or functional consequences of engaging in vaping – which in turn, serve as proximal regulators of behavior (Bandura, 1986, 1989, 2004).
Indeed, a substantial body of research shows that perceived social norms are among the strongest predictors of substance use in young people, with robust evidence in the alcohol literature (Duckworth et al., 2025; Neighbors et al., 2007) and growing support for similar effects in combustible and electronic cigarette use. Recent studies have found that greater perceived nicotine use in friends and family (i.e., descriptive norms) and also greater perceived social approval (i.e., injunctive norms) was related to increased vaping (East et al., 2019; Lozano et al., 2019; Oh et al., 2022; Romm et al., 2022). Moreover, leveraging individuals’ tendency to conform to perceived norms, researchers have increasingly implemented norms-based interventions aimed at correcting cognitive misperceptions of peer substance use, with evidence overwhelmingly concentrated in alcohol interventions (Berkowitz, 2005; Neighbors, Lewis, et al., 2016) and a smaller but growing literature extending this approach to other substance-related behaviors (Lemmel & Morina, 2024; Pischke et al., 2021).
Perceived social norms function as proximal subjective cognitive appraisals, with individuals responding not to others’ actual behaviors but to their own interpretations, evaluations, and comparisons of those behaviors and the social pressures they believe exist – appraisals that subsequently elicit emotional responses and shape behavior (Fishbein & Yzer, 2003; Packard & Schultz, 2023; Rimal & Real, 2003, 2005). Moreover, beyond signaling social approval or providing behavioral heuristics, perceived social norms may also function as emotional regulators; when individuals experience negative emotional states, they often look to others’ behavior for reassurance and guidance (Bergquist & Ekelund, 2025). Difficulty regulating emotional states may therefore influence how social information is interpreted, including perceptions of descriptive and injunctive vaping norms. Individuals with greater emotion dysregulation may be more likely to perceive vaping as common or socially acceptable, particularly when such perceptions align with anticipated affective relief or distress reduction (Bandura, 1986, 1989, 1998, 2004; Hall & Queener, 2007; Khantzian, 1997). In turn, perceiving vaping as prevalent and acceptable may be associated with greater likelihood of use.
Emotion dysregulation can also be conceptualized as a relatively trait-like vulnerability involving enduring difficulties identifying, tolerating, and modulating emotional states across situations, while still exhibiting state-dependent fluctuations in response to contextual demands (Gratz & Roemer, 2004, 2008; Olatunji et al., 2024; Wang et al., 2024). Consistent with this view, substance use research typically frames emotion dysregulation as an antecedent vulnerability that biases downstream cognitive and behavioral processes, including coping motives, affect-regulation expectancies, and related reinforcement mechanisms (Aurora & Klanecky, 2016; Paulus et al., 2021; Zvolensky et al., 2021). In contrast, perceived social norms are inherently cue-dependent, contextually malleable, and readily shaped by situational feedback and informational exposure (Buckner et al., 2019; Gelfand et al., 2024; Liu et al., 2020; Pilatti et al., 2021; Prince & Carey, 2010).
Additionally, examining perceived norms across different referent groups is important because their influence partly depends on the proximity and social relevance of the reference group (Borsari & Carey, 2003; Collins & Spelman, 2013; Cox & Bates, 2011; Lac & Donaldson, 2018; Patrick et al., 2012). Perceived substance use and approval among close friends may be particularly influential because these peers are more personally salient and embedded in everyday contexts of use (Collins & Spelman, 2013; Cox & Bates, 2011; Lac & Donaldson, 2018). In contrast, perceptions of broader groups (e.g., typical students) may rely more on generalized assumptions or misperceptions and less on direct interpersonal experience (Borsari & Carey, 2003; Cox & Bates, 2011). Considering multiple referent groups may therefore clarify when and for whom perceived norms are most strongly associated with substance use outcomes.
Collectively, this pattern of evidence supports conceptualizing perceived norms as a plausible social-cognitive mechanism through which emotion dysregulation may be indirectly associated with vaping behavior. Yet, to our knowledge, no empirical studies have tested this mediating pathway for vaping or other substances.
Characterizing vaping during the critical developmental period of emerging adulthood requires attention to both how often a person has vaped recently and whether the person's use has become compulsive or difficult to control. Past-30-day frequency captures current involvement and progression toward more regular use, a pattern linked to persistent vaping, greater nicotine dependence, and respiratory symptoms (Cho et al., 2025; Creamer et al., 2019; Han et al., 2024; Tackett et al., 2024). By contrast, problematic vaping reflects dependence-related features such as craving, withdrawal, loss of control, and difficulty abstaining – features that mark a more harmful level of involvement and are associated with greater dependence severity, shorter abstinence during quit attempts, more difficulty quitting, and broader physical and emotional burden (Cristol et al., 2024; Merreighn et al., 2025; Parks et al., 2025).
The Current Study
Thus, the present study seeks to address key gaps in the literature by examining the association between emotion dysregulation and vaping among university students and, to our knowledge, represents the first investigation to assess indirect associations through perceived social norms. Identifying such pathways may inform the development of targeted prevention and cessation interventions, which are particularly needed given the rising prevalence of vaping among emerging adults. We hypothesized that (1) higher levels of emotion dysregulation would be positively associated with both past-30-day vaping frequency and problematic vaping; (2) emotion dysregulation would be associated with elevated perceived descriptive and injunctive norms for distal reference groups (e.g., same-sex peers, typical students), but not for proximal groups (e.g., close friends); and (3) heightened descriptive and injunctive norms would mediate the relationship between emotion dysregulation and both personal and problematic vaping.
Methods
Study Participants and Design
This cross-sectional study included 723 undergraduate students (M age = 19.28 years, SD = 2.08) recruited from five Canadian universities: Toronto Metropolitan University, York University, Mount Saint Vincent University, the University of Manitoba, and the University of Calgary. Participants were between 18 and 29 years of age, fitting the definition of emerging adulthood (Arnett et al., 2014). We conducted a secondary analysis of an existing dataset (original study details available upon request), focusing only on participants who reported using an electronic nicotine product in the past 12 months. Individuals outside the 18–29 age range or who had not used an e-cigarette in the past year were excluded from the analyses. Regarding biological sex assigned at birth, 18.8% of participants were male, and 81.2% were female. In terms of gender identity, 18.0% identified as man, 78.1% as woman, 1.0% as transgender man, 0.3% as agender, 0.6% as gender-questioning, 1.2% as non-binary, 0.6% as genderqueer, and 0.3% as an ‘other’ identity. The sample was ethnically diverse, with participants identifying as Middle Eastern, North African, or Central Asian (5.8%), East Asian, South-East Asian, or Pacific Islander (15.1%), Hispanic or Latino (3.9%), Caucasian or White (62.2%), Black (6.1%), Aboriginal (4.7%), South Asian (8.2%), and other backgrounds (3.3%).
Procedure
Informed consent was obtained from all participants prior to data collection. Recruitment occurred through psychology student research pools at five universities across Canada (York University, Toronto Metropolitan University, University of Calgary, University of Manitoba, Mount Saint Vincent University), and data were collected as part of a larger online survey examining addiction and mental health comorbidities from 2021 to 2022. Participants completed a 30-min online questionnaire assessing demographic characteristics, substance and behavioral addictions, as well as a range of mental health and psychological variables. Student participants received course credit as compensation. The study was approved by the Research Ethics Board at each participating institution.
Measures
Demographic Characteristics
Participants reported basic demographic information including age (in years), biological sex assigned at birth (“Male” or “Female”), and current gender identity (selected from options or self-described). They also provided information on their racial/ethnic background (choosing from a list of categories or “other” with a fill-in option).
Electronic Nicotine Product (ENP) Use
Vaping was assessed with questions about ever use and recent use of ENPs. All participants first answered a screening question: “Have you used any electronic nicotine products in the past 12 months? (ENPs include e-cigarettes, vape pens, personal vaporizers, e-cigars, e-pipes, e-hookahs, JUUL, and hookah pens).” Only those who answered “Yes” (indicating past-year use) continued to subsequent questions. Recent use frequency was measured by asking: “In the past 30 days, on how many days did you use ENPs?” with response options ranging from 0 to 30 days. This past-month usage count served as our measure of frequency of ENP use.
Problematic Vaping
To gauge problematic or addictive patterns of vaping, we adapted the Screener for Substance and Behavioral Addictions (SSBA; Schluter et al., 2018) for vaping. The SSBA is a brief screening tool originally designed for various substance and behavior addictions. Participants indicated how often in the past 12 months they experienced each of four symptoms while using e-cigarettes: (1) “I used it too much,” (2) “Once I started, I couldn’t stop,” (3) “I felt I had to do it in order to function,” and (4) “I continued to do it, even though it caused problems.” Responses were on a 5-point Likert scale from 0 (“None of the time”) to 4 (“All of the time”), with additional options “I didn’t do this at all” (treated as 0) and “Don’t know/Prefer not to say” (treated as missing). We summed the four items to create a problematic vaping score ranging from 0 to 16, with higher scores indicating greater problematic use. In prior validation studies, the SSBA has demonstrated good internal consistency and evidence of construct validity across substance-related behaviors (Hodgins et al., 2022; Schluter et al., 2018). In our sample, the SSBA adapted for vaping showed acceptable internal consistency (Cronbach’s α = 0.79).
Perceived Descriptive Norms of Vaping
Perceived descriptive norms were measured with three parallel items asking participants to estimate peer vaping. Specifically, participants were asked: “In the past 30 days, on how many days do you think the following types of students at your university used electronic nicotine products?” The reference groups for these estimates were: (1) “a typical student at your university,” (2) “a student of the same gender as you at your university,” and (3) “one of your close friends.” For each reference group, participants provided a numeric estimate between 0 and 30 days. These items index the participant’s perception of peer vaping frequency for distal vs. proximal peers. Higher values reflect a belief that peers use e-cigarettes more frequently.
Perceived Injunctive Norms of Vaping
Perceived injunctive norms were assessed with a similar format, focusing on perceived approval of vaping. Participants were asked: “How much do you think the following types of students at your university approve of using electronic nicotine products?” The same three reference groups were considered: (1) a typical student, (2) a student of the same gender, and (3) a close friend. Responses were given on a 5-point scale from 1 (“Strongly disapprove”) to 5 (“Strongly approve”), with higher scores indicating a perception that the group is more approving of vaping. These three items provide insight into the student’s belief about the social acceptability of vaping among peers.
Emotion Dysregulation
Emotion dysregulation was measured using the Difficulties in Emotion Regulation Scale – Short Form (DERS-SF; Gratz & Roemer, 2004; Kaufman et al., 2016). The DERS-SF is an 18-item self-report questionnaire derived from the original 36-item DERS, capturing six domains of emotion regulation difficulties: nonacceptance of emotions, difficulty engaging in goal-directed behavior when distressed, impulse control difficulties, lack of emotional awareness, limited access to emotion regulation strategies, and lack of emotional clarity. Participants rated each item on a 5-point frequency scale ranging from 1 (“Almost never true of me”) to 5 (“Almost always true of me”). Example items include “When I’m upset, I feel ashamed of myself for feeling that way” and “I have difficulty making sense of my feelings.” We used the total DERS-SF score as an index of overall emotion dysregulation, with higher scores representing greater difficulty in emotion regulation. In prior validation studies, the DERS-SF has demonstrated good to excellent internal consistency and strong convergent and predictive validity across various samples (Kaufman et al., 2016; Victor & Klonsky, 2016). In the present sample, the DERS-SF demonstrated good internal consistency (Cronbach’s α = 0.86).
Data Analytic Overview
Participants who failed two or more attention checks, indicated that they had not responded honestly, completed less than 98% of the survey, completed the survey in under 5 min, or provided no data on key indicator variables were excluded prior to analysis. Based on these criteria, 30 participants were removed. Vaping-specific items were administered only to participants who reported electronic nicotine product use within the past 12 months; thus, missing responses to vaping items among non-users reflected planned missingness due to survey branching logic. Among participants eligible to complete these measures, item-level missingness was minimal (<1% across study variables). Missing data were not imputed at the item or scale level, and scale scores were treated as missing if one or more items were incomplete. Analyses were conducted using complete-case procedures (listwise deletion) within each model.
All analyses were conducted using IBM SPSS Statistics (Version 28). Prior to hypothesis testing, we examined the distributions of all key variables (e.g., emotion dysregulation scores, vaping frequency, perceived norms) for any deviations from normality. Skewness and kurtosis values were within acceptable ranges for psychometric data, indicating that no transformations were necessary (Weston & Gore, 2006). We first computed descriptive statistics (frequencies, means, standard deviations) for the main variables and Pearson bivariate correlations to assess initial associations among ED, vaping outcomes, and perceived norms. Next, to evaluate the presence of normative misperception, we performed paired-sample t-tests comparing participants’ self-reported 30-day vaping to their perceived 30-day use of a typical student, same-gender student, and close friend. Effect sizes for mean differences are reported as Cohen’s d.
Given the cross-sectional design, the mediation analyses examine indirect associations consistent with the proposed model but do not permit causal or temporal inference. For the primary mediation analyses, we specified six separate models using the PROCESS macro for SPSS (Model 4) (Hayes, 2022). Each model tested whether perceived norms mediated the association between ED (as the independent variable, centered for analysis) and a vaping outcome (dependent variable), with bootstrapped confidence intervals for indirect effects. We examined mediation for both outcomes (problematic vaping and past-30-day vaping frequency) across each of the three reference groups for norms (typical student, same-gender student, close friend). In these models, ED was entered as the predictor (X), the perceived norm (either descriptive or injunctive for a given reference group) as the mediator (M), and the vaping outcome as the criterion (Y). We controlled for sex (female vs. male) in all models given the predominance of female students in our sample, and because gender can relate to both emotion regulation and substance use behaviors. Each model used 5,000 bias-corrected bootstrap samples to estimate indirect effects; a mediation effect was considered statistically significant if the 95% confidence interval for the indirect effect did not include zero (Preacher & Hayes, 2008). For clarity, we present simplified path diagrams for each reference group in the Results (Figures 1−6 correspond to each model). Path Analysis Model of Associations Between Emotion Dysregulation and Problematic Vaping for Typical Student Norms. Note. Mediation model illustrating the interplay between emotional dysregulation and problematic vaping, mediated by descriptive and injunctive norms within the typical student group (T). Values presented on paths include unstandardized coefficients (b) and standardized coefficients (β) with standard errors (SE) and 95% confidence intervals (CI). R2 values indicate variance explained. Solid bolded lines indicate statistically significant associations, while dashed lines denote non-significant associations Path Analysis Model of Associations Between Emotion Dysregulation and Vaping Frequency for Typical Student Norms. Note. Mediation model illustrating the interplay between emotional dysregulation and vaping frequency, mediated by descriptive and injunctive norms within the typical student group (T). Values presented on paths include unstandardized coefficients (b) and standardized coefficients (β) with standard errors (SE) and 95% confidence intervals (CI). R2 values indicate variance explained. Solid bolded lines indicate statistically significant associations, while dashed lines denote non-significant associations Path Analysis Model of Associations Between Emotion Dysregulation and Problematic Vaping for Same-Sex Student Norms. Note. Mediation model illustrating the relationship between emotional dysregulation and problematic vaping, mediated by descriptive and injunctive norms within the same-sex student group (SS). Values presented on paths include unstandardized coefficients (b) and standardized coefficients (β) with standard errors (SE) and 95% confidence intervals (CI). R2 values indicate variance explained. Solid bolded lines indicate statistically significant associations, while dashed lines denote non-significant associations Path Analysis Model of Associations Between Emotion Dysregulation and Vaping Frequency for Same-Sex Student Norms. Note. Mediation model illustrating the interplay between emotion dysregulation and vaping frequency, mediated by descriptive and injunctive norms within the typical student group (SS). Values presented on paths include unstandardized coefficients (b) and standardized coefficients (β) with standard errors (SE) and 95% confidence intervals (CI). R2 values indicate variance explained. Solid bolded lines indicate statistically significant associations, while dashed lines denote non-significant associations Path Analysis Model of Associations Between Emotion Dysregulation and Problematic Vaping for Close Friends’ Norms. Note. Mediation model illustrating the relationship between emotional dysregulation and problematic vaping, mediated by descriptive and injunctive norms within the close friends’ group (CF). Values presented on paths include unstandardized coefficients (b) and standardized coefficients (β) with standard errors (SE) and 95% confidence intervals (CI). R2 values indicate variance explained. Solid bolded lines indicate statistically significant associations, while dashed lines denote non-significant associations Path Analysis Model of Associations Between Emotion Dysregulation and Vaping Frequency for Close Friends’ Norms. Note. Mediation model illustrating the interplay between emotion dysregulation and vaping frequency, mediated by descriptive and injunctive norms within the close friends’ group (CF). Values presented on paths include unstandardized coefficients (b) and standardized coefficients (β) with standard errors (SE) and 95% confidence intervals (CI). R2 values indicate variance explained. Solid bolded lines indicate statistically significant associations, while dashed lines denote non-significant associations
Results
Descriptive Statistics, Bivariate Correlations, and Preliminary Analyses
Descriptive Statistics and Bivariate Correlations
Note. M, Mean; SD, Standard deviation; T, Typical student; SS, Same-sex student; CF, Close friend.
*p < .05. **p < .01.
Paired-samples t-tests were conducted to compare participants’ self-reported past-30-day vaping with their perceptions of typical student, same-sex peer, and close friend use. Participants reported significantly lower vaping (M = 9.84, SD = 11.78) than they attributed to a typical student (M = 17.61, SD = 9.12), t (717) = 15.01, p < .001, d = 0.56. A similar pattern emerged for perceived same-sex peer use (M = 16.30, SD = 8.99), t (715) = 12.61, p < .001, d = 0.47, and perceived close friend use (M = 14.86, SD = 11.54), t (715) = 11.46, p < .001, d = 0.42.
Mediation Analysis: Path Model
Before conducting the mediation analyses, all assumptions of multiple regression were evaluated and met. Bootstrapped mediation models were then estimated using the SPSS PROCESS macro to test whether perceived descriptive and injunctive norms mediated the association between ED and two vaping outcomes: problematic vaping and past-30-day vaping frequency. Separate models were tested for each reference group: typical students, same-sex students, and close friends.
Models 1 and 2: Typical Students
These models examined whether perceived norms for typical students mediated the relationship between emotion dysregulation and (a) problematic vaping and (b) past-30-day vaping frequency. There was a significant indirect effect of ED on problematic vaping through descriptive norms for typical students (b = 0.0046, SE = 0.0024, 95% CI [0.0008, 0.0102]). In contrast, injunctive norms for this reference group did not mediate the association (b = –0.0001, SE = 0.0017, 95% CI [–0.0037, 0.0034]). The direct effect of ED on problematic vaping remained significant (b = 0.0636, SE = 0.0140, 95% CI [0.0362, 0.0910]). Figure 1 presents the full mediation model.
A similar pattern emerged for frequency of vaping. ED showed a significant indirect effect through descriptive norms for typical students (b = 0.0127, SE = 0.0062, 95% CI [0.0030, 0.0266]), whereas the indirect effect through injunctive norms was not significant (b = –0.0013, SE = 0.0037, 95% CI [–0.0097, 0.0054]). The direct effect of ED on vaping frequency was also significant (b = 0.0724, SE = 0.0311, 95% CI [0.0113, 0.1335]). Figure 2 illustrates the model pathways.
Although the indirect effects for descriptive norms of typical students were statistically significant, their magnitude was small, indicating modest indirect associations between emotion dysregulation and vaping outcomes.
Models 3 and 4: Same-Sex Students
For problematic vaping, there was no significant indirect effect of ED through either descriptive norms (b = 0.0035, SE = 0.0023, 95% CI [0.0000, 0.0087]) or injunctive norms (b = 0.0017, SE = 0.0018, 95% CI [–0.0017, 0.0055]) within the same-sex student reference group. However, the direct effect of ED on problematic vaping remained statistically significant (b = 0.0628, SE = 0.0139, 95% CI [0.0355, 0.0901]). These results are summarized in Figure 3.
A similar pattern emerged for past-30-day vaping frequency. No significant indirect effects were found through descriptive norms (b = 0.0086, SE = 0.0053, 95% CI [–0.0001, 0.0203]) or injunctive norms (b = 0.0041, SE = 0.0039, 95% CI [–0.0033, 0.0126]) within the same-sex student reference group. The direct effect of ED on vaping frequency remained significant (b = 0.0708, SE = 0.0310, 95% CI [0.0099, 0.1317]). Mediation pathways for these models are presented in Figure 4.
Models 5 and 6: Close Friends
For problematic vaping, no significant indirect effects of ED through either descriptive norms (b = 0.0019, SE = 0.0053, 95% CI [–0.0085, 0.0126]) or injunctive norms (b = 0.0011, SE = 0.0019, 95% CI [–0.0024, 0.0051]) were observed within the close-friend reference group. However, the direct effect of ED on problematic vaping remained significant (b = 0.0650, SE = 0.0126, 95% CI [0.0402, 0.0898]). The model pathways are presented in Figure 5.
A similar pattern emerged for past-30-day vaping frequency. No significant indirect effect was found through descriptive norms (b = 0.0055, SE = 0.0155, 95% CI [–0.0259, 0.0356]) or injunctive norms (b = 0.0008, SE = 0.0020, 95% CI [–0.0028, 0.0053]). Nevertheless, the direct effect of ED on vaping frequency remained significant (b = 0.0772, SE = 0.0271, 95% CI [0.0240, 0.1304]). A summary of this mediation model is provided in Figure 6.
Discussion
The findings of the present study reveal several significant and novel insights into the psychosocial correlates of vaping among university students. First, consistent with our initial hypothesis, emotion dysregulation (ED) was positively associated with both past-30-day vaping frequency and problematic vaping. Although research directly examining this relationship remains limited, our findings align with emerging evidence demonstrating links between ED and recent vaping (Tran & Morrell, 2025), e-cigarette dependence (Manning et al., 2020), and current vaping (Brockenberry et al., 2022). University students with higher ED have been shown to report greater vaping dependence and stronger affect-related expectancies, both positive and negative (McLeish et al., 2024), and emotional vulnerabilities predict stronger beliefs that vaping alleviates negative affect, which in turn increases likelihood of use (Brockenberry et al., 2022). These findings mirror patterns observed in the broader tobacco literature, where individuals who smoke cigarettes frequently endorse perceived coping-related benefits, including beliefs that smoking helps manage psychological distress and mitigates negative emotions (Johnson et al., 2008; Piper et al., 2004), consistent with the self-medication hypothesis (Hall & Queener, 2007; Khantzian, 1997).
Consistent with our second hypothesis, ED was positively associated with perceived descriptive norms for typical students, as well as with perceived injunctive norms for both typical and same-sex students, but not for close friends. These findings suggest that higher levels of ED may heighten students’ susceptibility to perceived peer norms. Although no prior research has directly examined this relationship in the context of substance use, related work has shown that other psychological vulnerabilities, such as depressive symptoms, are associated with stronger perceived descriptive and injunctive alcohol norms among university students (Linden & Lau-Barraco, 2013). An additional noteworthy finding was that across all three reference groups (typical students, same-sex students, and close friends), participants consistently overestimated others’ vaping relative to their own. This pattern aligns with previous research documenting similar normative overestimations among university students (Nabil et al., 2024; Oh et al., 2022) and youth populations more broadly (East et al., 2019).
With respect to our final mediation hypothesis, results provided only partial support. When typical students served as the reference group, perceived descriptive norms significantly mediated the associations between ED and both problematic vaping and past-30-day vaping frequency. However, neither descriptive nor injunctive norms mediated these relationships when same-sex students or close friends were the referents. It is important to note that the observed associations and indirect effects were small in magnitude, indicating that emotion dysregulation and perceived social norms account for a relatively modest proportion of variance in vaping outcomes. Such effect sizes may be broadly consistent with social-cognitive models, including Social Cognitive Theory, which emphasize reciprocal determinism and the dynamic interplay between individual factors, social influences, and environmental contexts in shaping health behaviors (Bandura, 1986, 1998, 2004). Although no prior studies have examined this pathway in the context of vaping, related work shows that psychological vulnerabilities such as social anxiety and depressive symptoms predict alcohol use partly through perceived norms, with descriptive norms consistently emerging as the strongest mediator among university students (Ham & Hope, 2006) and emerging adults (Lau-Barraco et al., 2016). The stronger role of descriptive norms in our findings aligns with evidence that they more directly capture perceived behavioral frequency or quantity and thus exert stronger influences on substance use (Borsari & Carey, 2003; Neighbors et al., 2007, 2008). In contrast, injunctive norms reflect perceived approval and tend to show meaningful effects only for highly proximal referents, such as close friends (Borsari & Carey, 2003; Lac & Donaldson, 2018; Neighbors et al., 2008). This distinction may also explain our finding that descriptive norms were directly associated with both problematic vaping and frequency across all reference groups, whereas injunctive norms showed a direct association with problematic use only when close friends were the referent.
An additional explanation for these patterns may reflect differences between distal and proximal reference groups in the formation of normative perceptions. Perceptions of “typical students” often rely on generalized beliefs about campus behavior rather than direct observation, making them particularly susceptible to normative misperceptions and cognitive biases (Bessenyei & Yakovenko, 2023; Collins & Spelman, 2013; Lac & Donaldson, 2018; Patrick et al., 2012). Individuals with greater emotion dysregulation may be especially likely to rely on these broader social perceptions when interpreting peer behavior, which could explain why ED predicted norms for typical and same-sex students but not for close friends. In contrast, perceptions of close friends’ behavior are typically informed by direct interaction and therefore tend to be more accurate, leaving less opportunity for emotional vulnerabilities to shape them. Consistent with this interpretation, descriptive norms for typical students – but not those for more proximal groups – showed significant indirect effects linking ED with vaping outcomes, suggesting that emotionally vulnerable students may be particularly influenced by perceptions of broader campus behavior rather than by the behavior of their immediate peer network.
Implications
The present findings carry several important implications for university student health and substance-use prevention. First, the observed associations between emotion dysregulation and both vaping frequency and vaping-related problems suggest that vaping may function, for some students, as a maladaptive emotion-regulation strategy, consistent with the self-medication hypothesis (Hall & Queener, 2007; Khantzian, 1997). These findings suggest that strengthening adaptive emotion-regulation skills may represent a valuable target for prevention and intervention efforts among university students. Cognitive-behavioural and transdiagnostic interventions that target emotion regulation, such as the Unified Protocol, as well as mindfulness-based interventions, have demonstrated efficacy in improving emotional functioning across a variety of psychological difficulties (Easdale-Cheele et al., 2026; Farchione et al., 2024; Saccaro et al., 2024). Integrating emotion-regulation skills training into campus counselling services, brief psychoeducational workshops, or first-year transition programming may therefore represent a promising strategy for reducing coping-motivated vaping among students who experience difficulties managing emotions adaptively.
Second, the finding that students misperceive peer vaping behavior highlights the importance of addressing normative misperceptions within university substance-use prevention initiatives. Normative education approaches that provide accurate, university-specific information about peer behavior may help correct students’ overestimations of vaping prevalence and peer approval (Larimer et al., 2023; Saxton et al., 2021). Such strategies may be delivered through campus-wide campaigns, digital media, or orientation programming. Finally, the present findings may also inform vaping cessation interventions for emerging adults. For example, a recent randomized clinical trial demonstrated that a tailored, interactive text-messaging intervention providing coping strategies and behavioral support significantly increased vaping abstinence among adolescents (Graham et al., 2024). Incorporating components that explicitly address emotion regulation difficulties and correct misperceptions about peer vaping may further strengthen such programs by targeting both psychosocial mechanisms associated with vaping.
Limitations and Future Research
Although this study offers valuable insights into the indirect associations linking emotion dysregulation, perceived norms, and vaping, several limitations should be noted. First, reliance on self-reported data may introduce recall and reporting biases, as participants may underreport or overreport their vaping. Future research may benefit from incorporating more objective tracking methods, such as daily journals or ecological momentary assessment, to improve measurement accuracy. Second, problematic vaping was assessed using self-report measures rather than clinical diagnostic tools. Although informative, these measures lack the precision of standardized clinical assessments, and no validated diagnostic screening tool for problematic vaping currently exists. The development and validation of such tools would enable future studies to incorporate standardized clinical assessments or diagnostic interviews.
Third, the sample consisted primarily of female, Caucasian/white Canadian undergraduates, which may limit generalizability to other genders, ethnic groups, and populations. More diverse and representative samples are needed to enhance applicability of findings across demographic groups. Fourth, the cross-sectional design restricts causal interpretation. Although mediation analyses were used to estimate indirect associations consistent with the proposed theoretical model, the design does not allow the establishment of temporal precedence or causal inference (Hayes, 2022). Accordingly, the observed indirect effects should be interpreted as statistical associations consistent with the model rather than evidence of causal pathways. Longitudinal research would allow stronger inferences regarding the temporal relationships among emotion dysregulation, perceived norms, and vaping behavior. Additionally, the observed effect sizes were small, indicating that emotion dysregulation and perceived norms likely represent modest, incremental correlates rather than primary determinants of vaping outcomes. Another limitation is that the study did not examine vaping initiation, as vaping-related measures were administered only to participants who reported vaping within the past 12 months. As a result, analyses focused on correlates of vaping among individuals who had already initiated use. This is noteworthy given that emerging adulthood is a developmental period characterized by experimentation and substance-use onset, during which psychosocial factors such as peer influence, stress, and anxiety may play important roles (Kinouani et al., 2025). Future research should therefore examine whether emotion dysregulation and perceived peer norms similarly relate to vaping initiation among emerging adults who have not previously used electronic nicotine products. Finally, this study used only the overall DERS-SF score rather than its subscales. Although this approach simplified the analytic models, future research could examine specific facets of emotion dysregulation to determine how distinct regulatory processes uniquely relate to vaping and perceived norms.
Conclusions
In summary, this study provides novel evidence that emotion dysregulation is associated with university students’ vaping, both directly and through its relationship with perceived social norms. Descriptive norms for typical students showed significant indirect associations with vaping frequency and problematic use. Moreover, students consistently overestimated peer vaping and approval. These findings suggest that students with greater emotion-regulation difficulties may be particularly vulnerable to perceived social influences surrounding vaping. By clarifying how emotional vulnerabilities intersect with perceived norms, this study contributes to a growing evidence base for prevention strategies that integrate emotion-regulation skill-building with norms-based interventions. Such approaches may help reduce vaping-related harms and support healthier developmental trajectories during emerging adulthood.
Supplemental Material
Supplemental Material - Emotion Dysregulation and E-Cigarette Use Among University Students: The Mediating Role of Perceived Social Norms
Supplemental Material for Emotion Dysregulation and E-Cigarette Use Among University Students: The Mediating Role of Perceived Social Norms by Arshia Soleimankhan, Jonathan David, Hyoun S. Kim, Sarah Dermody, David C Hodgins, N. Will Shead, Achala H. Rodrigo, and Matthew T. Keough in Emerging Adulthood
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Transparency and Openness Statement
The raw data contained in this manuscript are not openly available due to privacy restrictions set forth by the institutional ethics board, but can be obtained upon reasonable request from the corresponding author following the completion of a privacy and fair use agreement. The analysis code/syntax used for the analyses are not openly available for download, but can also be obtained from the corresponding author upon reasonable request by the journal. All the materials used in this study, including study measures, may be openly available for download, however restrictions may be imposed by the governing institutions and authors that own such measures. The study did not include a pre-registration plan for data collection or analysis.
Supplemental Material
Supplemental material for this article is available online.
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
