INTRODUCTION

In the past decade, young adults have been leading adopters of e-cigarettes, which are electronic devices that aerosolize a nicotine-containing liquid that is inhaled. In 2018, the percentage of adults aged 18–24 years who had ever used an e-cigarette was 26% compared to 21%, 11%, and 5% among adults aged 25–44, 45–64, and >65 years, respectively¹. More recent pod-style devices appeal to young people because of their easy concealment, flavors, novel technology, and perceived safety relative to cigarettes². Concerns about e-cigarette use by young people include negative impacts on the developing brain, exposure to toxicants in e-cigarette aerosol, and increased risk of using combustible cigarettes and other substances².

The transition from adolescence to early adulthood is an important developmental stage that can include major life changes such as leaving a parental home, identity development, social group formation, and behavioral experimentation. For young adults who attend college, college’s unique social environment may promote substance use, including e-cigarette use, through social network influence³ and (mis)perceived norms of substance use among peers⁴. Indeed, college students report that friends influence their e-cigarette initiation⁵, report commonly using e-cigarettes with others and at social occasions⁶, and perceive high acceptability of e-cigarette use among peers^5,7,8.

The use of multiple substances is common among young adults⁹, and the college environment might contribute to use. While past 30-day use of e-cigarettes is higher among young adults who do not attend versus attend college (15% vs 10%, respectively), among young adults who use e-cigarettes, the prevalence of multiple substance use is higher among college young adults⁷. Risks of using multiple substances include worsened side effects (e.g. blackouts), heightened risk of overuse and overdose, impaired judgment, and risky behaviors¹⁰. Long-term risks include addiction, related health consequences from all substances used, and worsened mental illness^11,12. Multiple substance use might also encourage heavier use. For example, dual use of e-cigarettes and combustible cigarettes is associated with heavier use of both substances among college students¹³.

Several individual factors are associated with the use of multiple substances among college students, including age and identifying as White, male, sexual minority, and belonging to a college fraternity or sorority¹⁴. Previous studies conducted among young adults have found a greater likelihood of other substance use, including binge drinking and cannabis use, among those who use e-cigarettes compared to those who do not¹⁵. For example, adolescents and young adults who use e-cigarettes have higher odds of using cannabis¹⁶, and studies among youth have found an association between cannabis use and initiation of e-cigarette use¹⁷. E-cigarette use with and without other substances has not been well explored among college students.

Substance use among college students, including e-cigarette use, is associated with poor mental health (e.g. psychiatric disorders, high stress, depression)¹⁸. Many potential explanations exist for this relationship, including self-medication¹⁹ and nicotine’s negative impact on the developing brain in ways that undermine mental health, such as oxidative stress and emotional dysregulation²⁰.

Although education level has a strong and consistently positive association with health outcomes, including substance use and mental health²¹, college also presents unique risks for substance use and mental distress. The prevalence of several mental health conditions among college students, including depression, anxiety, and suicidality, has been increasing²². As a result, college students are a priority population for both substance use prevention and mental health intervention.

Understanding the relationships between mental health distress and e-cigarette use is critical, given their increasing prevalence and potential intersections. The common use of e-cigarettes and other substances among young people makes it important to differentiate between patterns of e-cigarette use with and without the use of other substances and their associations with mental health. This study aims to: 1) describe the prevalence of use of both e-cigarettes and other substances (ECIG+ use); and 2) assess the relationship between mental health and ECIG+ use among college students.

METHODS

Design and sample

This secondary analysis used data from the US American College Health Association’s National College Health Assessment (ACHA-NCHA IIc), a national survey of undergraduate and graduate students collected from 138 colleges who self-selected to participate in Fall 2018 and Spring 2019²³. Only institutions that sampled all students or used random sampling techniques were included in the ACHA-NCHA public-use database. Reliability and validity analyses have demonstrated strong consistency over survey periods. We restricted our analysis to undergraduate students aged 18–24 years with complete data on our measures (Figure 1; n=55654). No Institutional Review Board approval was needed since the dataset is publicly available (with ACHA approval) and de-identified.

Figure 1

Study population flow chart, National College Health Assessment IIc, United States, Fall 2018–Spring 2019

RESULTS

Majorities of the sample identified as cisgender women (68%), straight (80%), and White (59%) and attended four-year (91%) and public (66%) institutions (Table 1). Ninety percent of students reported having an A or B cumulative grade average. The mean psychological distress score was 5.6 (SD=2.9; range 0–11). About one-third (31%) of students had received a mental health diagnosis/treatment in the past 12 months. All bivariate associations between our individual characteristics and ECIG use type were statistically significant after applying the Bonferroni correction (Supplementary file Table S1).

Table 2 reports the prevalence of past 30-day substance use in our sample. Alcohol was the most used substance (58%), followed by cannabis (23%) and e-cigarettes (15%). Among those who reported past 30-day e-cigarette use, nearly all (95%) reported also using at least one other substance; alcohol appeared in several use patterns, and the most common unique substance use combination was e-cigarette, alcohol, and cannabis use (26%). Only 2% of these students used e-cigarettes with other tobacco products only. Figure 2 displays the top substance use combinations among students who used e-cigarettes and other substances based on the ten combinations with prevalence ≥1%. Circle size is proportional to the number of respondents reporting the use of that substance, and overlapping circles indicate substance use combinations.

Figure 2

Top substance use combinations among students who used e-cigarettes and other substances, National College Health Assessment IIc, United States, Fall 2018–Spring 2019 (N=7987)

Table 3 displays the results from our multinomial logistic regression models. In our fully specified model, greater psychological distress was associated with higher odds of ECIG+ use versus no substance use (AOR=1.1; 95% CI: 1.1–1.2) and higher odds of no e-cigarette use but other substance use (AOR=1.1; 95% CI: 1.1–1.2). The association between psychological distress and sole e-cigarette use was not significant. Receiving a mental health diagnosis/treatment in the past 12 months was associated with higher odds of sole e-cigarette use (AOR=1.6; 95% CI: 1.3–2.0), ECIG+ use (AOR=1.5; 95% CI: 1.4–1.6), and no e-cigarette use but other substance use (AOR=1.1; 95% CI: 1.1– 1.2) compared to no substance use. Our findings for mental health were the same in our model accounting for only student and mental health characteristics.

Several student characteristics were significantly associated with ECIG use type, including gender identity, sexual orientation, race and ethnicity, self-rated health, year in school, cumulative grade average, social fraternity or sorority membership, and current residence. For example, compared to White students, students of all other racial and ethnic identities had lower odds of ECIG+ use versus no substance use (AORs range: 0.51–0.86). The associations between student characteristics and ECIG use type were generally similar across the three models, although some findings changed in statistical significance when accounting for mental health and college characteristics.

Attending a college with a campus size of ≥20000 versus <2500 was associated with lower odds of sole e-cigarette use versus no substance use (AOR=0.58; 95% CI: 0.37–0.91). In contrast, the odds of no e-cigarette use but other substance use were higher among students on campuses of ≥20000 (AOR=1.11; 95% CI: 0.81–1.52). Students who attended a public versus private, non-religious institution had lower odds of no e-cigarette use but other substance use versus no substance use (AOR=0.68; 95% CI: 0.51–0.91). No college characteristics were significantly associated with ECIG+ use.

DISCUSSION

This study investigated the prevalence and correlates of ECIG+ use among college students, a priority population for both substance use prevention and mental health support. Notable strengths include the use of a large national sample, our focus on multiple substances, and the inclusion of mental health variables, all of which may enhance understanding of college students’ substance use. Our analysis identified key patterns that help us better understand e-cigarette use among college students. While psychological distress was not associated with sole ECIG use, both psychological distress and past-year mental health diagnosis/treatment were positively associated with ECIG+ use versus no substance use. This latter finding is consistent with growing evidence of the association between e-cigarette use and mental health outcomes^18,24.

Consistent with a previous study on adolescents²⁵, we found that sole e-cigarette use was rare among college students and that nearly all students who used e-cigarettes also reported past 30-day use of another substance. Among other substances reported, alcohol, cannabis, and tobacco were the most common. Very few students who used e-cigarettes used only other nicotine products. Further, substance-use combinations, including cannabis (59%), were more common than combinations including tobacco (39%). This finding relates to the increasing popularity of cannabis; according to data from the Monitoring the Future panel survey, between 2017 and 2019, the past 30-day prevalence of cannabis use increased from 5% to 14% among college students²⁶.

Previous studies of college students have identified unique behavioral patterns of multiple substance use²⁷, suggesting that a variety of interventions may be necessary to reduce substance use among college students. For example, illicit drugs accounted for a relatively small share of substance use; however, use that includes illicit drugs may be especially dangerous and important to address. Identifying students who have particularly high-risk substance use and delivering tailored interventions should remain a part of the repertoire of campus responses.

Interventions incorporating technology (e.g. text messages, phone calls, online resources) show promise in improving mental health outcomes among college students; many of these interventions use evidence-based approaches such as cognitive behavioral therapy²⁸. Prior studies have emphasized the importance of integrated care for mental health and substance use disorders²⁹, so we recommend that both mental health symptoms and e-cigarette use be considered when designing treatment interventions for college students.

Interventions in the policy and regulatory environment may offer more impact than individual-level interventions. For example, campus-wide bans on tobacco products and vaping, campus or city restrictions on alcohol outlet density, and local or state regulation of commercial cannabis may reduce college students’ substance use. Given that <40% of campuses have adopted an e-cigarette-free policy and only 20% meet the American College Health Association recommendations for tobacco-free policies³⁰, there is room for additional action on campus to deter substance use.

At present, the evidence-based interventions (EBIs) for use with college students are better developed for alcohol than other substances. Fewer EBIs have been developed for college students to prevent e-cigarette, cannabis, and other substance use. Thus, we see an opportunity to adapt and test existing tools, such as the Truth Initiative’s ‘This is Quitting’ program designed for teenagers and young adults³¹ to reduce substance use and improve mental health outcomes among college students.

Our analysis also suggests that there are certain demographic groups that warrant focused attention. Consistent with prior studies^13,25, we found that cisgender males were more likely to report sole e-cigarette use and ECIG+ use versus no substance use than cisgender female students. A research report published by the National Institute on Drug Abuse describes several reasons why males may be more likely to use tobacco products, such as greater activation of reward pathways, greater responsiveness to environmental cues, and cultural factors³². Regardless of mechanisms, high-use prevalence suggests the need to tailor substance use prevention interventions for males.

Racial and ethnic minority students had consistently lower odds of ECIG+ use versus no substance use than their White peers. This finding may reflect a previously documented pattern of later age of substance-use initiation among these groups³³. If so, prevention efforts in college settings may be an important strategy to address lifetime substance use disparities. There may also be untapped factors that could be leveraged in the college context, such as racial and ethnic community affiliation for health promotion.

Our sensitivity analyses suggest that higher ECIG+ use versus no substance use among bisexual females and males may be driving our findings of sexual orientation differences. This is consistent with prior studies showing a higher prevalence of substance use among bisexual individuals when compared to both heterosexual peers and other sexual minority groups³⁴. In a recent study characterizing substance use disparities, bisexual women, in particular, were at an elevated risk for multiple substance use behaviors³⁴, suggesting that this group may need focused and culturally appropriate prevention interventions.

Interestingly, no college characteristics were associated with ECIG+ use versus no substance use. However, students at public institutions had lower odds of no e-cigarette use but other substance use than students at private religious institutions. Larger campus size was also associated with lower odds of sole e-cigarette use and higher odds of no e-cigarette but other substance use. Differences in the social context and/or student body composition might help explain this result, though additional research is needed to fully understand these differences.

The high prevalence of other substance use among students who used e-cigarettes warrants an in-depth investigation into the motivations and consequences of this pattern of use. Potential motivations might include a desire to enhance positive subjective effects, physiological triggers for additional use, experimentation, and desire to reduce the use of one substance³⁵. Further, shared devices may facilitate e-cigarette and cannabis use, but less is known about shared delivery mechanisms for illicit drugs other than cannabis.

Emerging research to identify physiological impacts, such as the toxicant load of dual-use (e.g. e-cigarette and combustible tobacco or alcohol), is important. Understanding the most appealing substance-use combinations, their motivations, and their health impacts could enhance the impact of prevention efforts for e-cigarettes by recognizing their frequent use of other substances. Continued exploration of the relationship and directionality between mental health and ECIG+ use is also critical to developing supportive and effective interventions.

CONCLUSIONS

This study provides insights about ECIG+ use, among a large sample of college students coming from 138 institutions that may inform future research and practice. Nearly all college students who reported using e-cigarettes also reported using at least one other substance. Therefore, intervention efforts to reduce e-cigarette use should consider the role of these other substances. It is also important to consider poor mental health, which was associated with higher odds of ECIG+ use versus no substance use. While additional research is needed to clarify the direction of this relationship, developing interventions that can both address substance use and enhance mental health, and that are tailored to groups that experience the highest rates of ECIG+ use, may have the greatest impact on college student health.