Prevalence and Correlates of Past 30-day Dual-Vaping of Nicotine and Cannabis among Adolescents in Five New England States

Abstract

Significance:

Adolescent vaping behavior includes vaping of multiple substances, including both nicotine and cannabis (dual-vaping). This study describes the prevalence and the sociodemographic correlates of past 30-day dual-vaping.

Methods:

We recruited adolescents ages 13-17 from five New England states (Massachusetts, Connecticut, Rhode Island, Vermont, New Hampshire) through the Prodege online survey panel from April 2021 to August 2022. Dual-vaping was defined as vaping both nicotine and cannabis (THC and/or CBD) in the past 30-days. We analyzed the prevalence of sole-nicotine, sole-cannabis, and dual-vaping of nicotine and cannabis and used multinomial logistic regression to examine associations between sociodemographic factors and sole- and dual-vaping of nicotine and cannabis.

Results:

The analytic sample included 2013 observations from 1858 participants (mean age 15.1 years, 46.2% female, 74.1% White, 82.2% heterosexual). Among these observations, 5.6% reported past 30-day sole-nicotine vaping, 5.5% reported sole-cannabis vaping, and 7.3% had dual-vaped. Correlates for higher odds of past 30-day dual-vaping included total social media sites used and household tobacco use, in contrast with sole-cannabis vaping, which included older age and self-reported depression (all p’s <0.05).

Discussion:

Adolescent past 30-day dual-vaping of nicotine and cannabis was more prevalent than past 30-day sole-vaping of either nicotine or cannabis alone. Future studies should continue to collect detailed data on the type of substances, besides nicotine, that adolescents are vaping.

INTRODUCTION

Background

The 2022 National Youth Tobacco Survey found that 14.1% of U.S. high school students reported past 30-day e-cigarette use (vaping), and the 2022 Monitoring the Future (MTF) survey found that 10.3% of 10^th graders and 14.8% of 12^th graders reported vaping cannabis in the past 30-days.^1–3 Dual-use of nicotine/tobacco and cannabis is associated with worse overall outcomes in terms of severity of nicotine dependence, tobacco cessation outcomes, psychopathology, and greater health risks.^1–6 One study found that dual-vaping of nicotine and cannabis was associated with greater likelihood of other illicit substance use.⁷

Thus, efforts to prevent and reduce adolescent vaping must address both nicotine and cannabis. Understanding the correlates associated with vaping multiple substances may help with designing enhanced prevention strategies for those already using or at risk of vaping single substances. To better understand adolescent vaping behaviors, research must incorporate other substances besides nicotine, including sole- and dual-vaping of cannabis.^7–11 We are specifically defining “dual-vaping” as past 30-day vaping nicotine (nicotine containing e-liquid) and vaping cannabis (e.g., using e-cigarettes or vaping devices that contain or use cannabis, defined as THC and/or CBD). This definition of dual-vaping includes simultaneous and concurrent vaping of cannabis and nicotine.

Research on the prevalence of dual-vaping among adolescents is scant, with one study using 2017 MTF data finding 7.5% of the sample of high school students in the U.S. reported past-30-day dual-vaping of cannabis and nicotine.¹² The 2022 MTF survey found that 14.2% of 10^th graders and 20.7% of 12^th graders had vaped nicotine in the past 30-days, and 10.3% of 10^th graders and 14.8% of 12^th graders vaped cannabis in the past 30-days,¹³ but does not report the dual-vaping prevalence.

Adult-use cannabis was legalized in Massachusetts in 2016.¹⁴ Since then, neighboring states Connecticut, Vermont, and Rhode Island have also enacted policies to legalize adult-use cannabis, and retail sales in these states began at the end of 2022.^15–17 This is especially concerning, given the fact that cannabis e-liquids tend to be of extremely high potency and are harmful to the lungs.^18,19 In 2019, the increase in e-cigarette or vaping product use-associated lung injury (EVALI) caused by vaping illicit THC oils highlighted the immediate risks of adolescents vaping cannabis with modified vaping cartridges.²⁰ Nicotine and cannabinoids tend to be used together to amplify the intoxicating effects and because both substances can be inhaled.^21–23 This dual-use of both nicotine and cannabis during adolescence in particular, versus adulthood, is associated with psychosocial risk factors and disrupting adolescent brain development.^24,25 It is important to examine, from a public health perspective, adolescent vaping of multiple substances, including cannabis, to provide approaches that are tailored to address the needs of adolescents at greater risk for vaping. Thus, there is a growing need to understand the predictors and correlates associated with adolescent dual-vaping of nicotine and cannabis.

The purpose of this study was to identify prevalence and characteristics of adolescents who report past 30-day vaping of both cannabis and nicotine in five New England states: Massachusetts (MA), (Connecticut (CT), Vermont (VT), New Hampshire (NH), and Rhode Island (RI). Approximately half of the study population was recruited from Massachusetts, as this analysis was part of a larger study aimed at assessing tobacco use behaviors following the Massachusetts 2019 Act on restriction of sales of flavored tobacco products. The study’s objectives were to: 1) Describe the prevalence of adolescent dual-vaping during a 17-month period (April 2021 to August 2022); and 2) Identify correlates associated with dual-vaping among adolescents. Our study is novel in that it focuses explicitly on the use of cannabis via vaping and examines the epidemiology and correlates associated with adolescents engaging in different types of sole- and dual-vaping.

METHODS

Study Design and Sampling Frame

The study design consisted of monthly repeated cross-sectional online surveys administered to participants who were part of an existing online survey panel with Prodege from April 2021 to August 2022. Prodege is an online market research company with access to existing panel members who are available to take part in surveys. The study population included a convenience sample of adolescents aged 13-17 in five New England states: Massachusetts, Rhode Island, Vermont, Connecticut, and New Hampshire. Surveys made available on panel members’ survey dashboard to current Prodege panel members, including anyone over the age of 13, with potential email reminders to take the survey. Adults were asked if they had a child aged 13 to 17 in their household and were prompted to give their phone to the child if so.^26–30 Parents were not required to provide consent to the survey, as parental consent was waived by the Institutional Review Board (IRB). All adolescent participants were shown an assent form before beginning the survey, providing all study information. Participants had to agree to the form to proceed with the survey. Participants were permitted to complete the survey in subsequent waves; however, subsequent repeated observations (n=155) were removed from the final analytic sample (N=1858). The University of Pennsylvania’s IRB approved the study protocol., other m

Variables and Measures

Dual-vaping. The main outcome of interest was past 30-day dual-vaping, operationalized as a categorical variable. In the survey, participants were asked: “Have you used any of the following vape products in the past 30 days?” with the possible responses of “Nicotine,” “Marijuana, marijuana concentrates, marijuana waxes, THC, or hash oils,” “CBD,” “Other,” and “Don’t know.” THC and CBD were asked about separately in the measure, as THC is the psychoactive and CBD is the non-psychoactive ingredient in cannabis, and there are many THC-only and CBD-only products. THC is the substance that is associated with divergent effects on dependency and other brain effects compared with CBD.³¹ “Dual-vaping” was defined as past 30-day use of both vaping nicotine and cannabis (“Marijuana, marijuana concentrates, marijuana waxes, THC, or hash oils” and/or “CBD”). “Other” responses were manually inspected and recoded to the appropriate category.

Vaping was operationalized as follows:

Y = 0 : did not vape either nicotine or cannabis in the past 30-days / “Don’t know”

Y = 1 : vaped only nicotine in the past 30-days

Y = 2 : vaped only cannabis in the past 30-days

Y = 3 : vaped both nicotine and cannabis in the past 30-days

Substances vaped the most: We asked participants: “Which of the following substances do you vape most often?” with the answer options of: “Nicotine,” “Marijuana, marijuana concentrates, marijuana waxes, THC, or hash oils,” “CBD,” “Other,” and “Don’t know.” This question was asked to participants who reported past 30-day dual-vaping throughout the entire 17-month study period (n=140).

Substances used to initiate vaping: We asked participants: “Which of the following did you use first when you started to vape? I used an e-cigarette or vape product with…” with the answer options of: “Only nicotine,” “Only THC,” “Only CBD,” “Both THC and CBD,” “Marijuana but don’t know if it was THC or CBD,” “Both nicotine and marijuana at the same time,” and “I don’t remember/I don’t know.” This question was included in the survey from September 2021 to August 2022, and asked to participants who reported past 30-day dual-vaping (n=74).

Covariates: Covariates collected at the individual level included gender (female, male, nonbinary/other), sexual orientation (heterosexual, not heterosexual), parental education (Less than high school, High school or more), type of school (public, not public), self-reported health status³² (less than good, good or better), self-reported depression based on 1 item on persistent feelings of sadness or hopelessness³³ (yes, no), paid employment in the past 12-months (yes, no), participation in organized activities in the past week (yes, no), household tobacco/e-cigarette use (yes, no), and total number of social media sites used (continuous, range of 0-13). We used continuous age rather than school grade for the analysis because the data were collected online rather than in a school setting.^34,35 We looked at race and ethnicity as two separate constructs (and two separate variables).³⁶ Race was recoded to a combined single variable with five mutually-exclusive categories: White, Black, Asian, Other, and Multiple. Ethnicity was recoded to a binary variable of Hispanic/Latina/x/o origin, or not Hispanic/Latina/x/o origin (mutually exclusive).^37,38 We also collected data on social and behavioral factors around life changes due to the COVID-19 pandemic, using the CRISIS measure.³⁹ We operationalized the “COVID-19 Impact Score” by totaling the “Yes”/(1) responses for the seven COVID-19 questions from the CRISIS measure for each participant, creating a summary score/continuous variable with a range of 0-7 to represent the impact of COVID-19 on participants, with a higher score representing experiencing more restrictions due to COVID-19. These covariates were included in our study due to prior work highlighting the association between gender, sexual identity, socioeconomic factors, paid employment, mental health, and peer influence on adolescent vaping.^40–46

Data Analyses

For objective 1, we utilized the analytic sample of total participant observations and conducted descriptive analyses of the prevalence of past 30-day vaping of nicotine and cannabis. We also conducted descriptive analyses of two survey questions regarding which substances they vaped the most and substances used to initiate vaping.

For objective 2, we used multinomial logistic regression for our categorical vaping outcome using the svyVGAM package in R.⁴⁷,⁴⁸ The adjusted model included age, sex, race, ethnicity, sexual orientation, and our covariates of interest. Significance was assessed at the alpha = 0.05 level (two-tailed).

RESULTS

Participant Demographics

Our five-state sample included participants with an average age of 15.1 years old (SD: 1.4), and 46.2% identified as female, 49.1% as male, and 4.6% as non-binary (Table 1). The majority were heterosexual (82.2%), white (74.1%), non-Hispanic (84.8%), and attended a public school (76.5%).

Table 1:

Demographic information of unique participants of study population (N=1858)

	n (%)
State
MA	900 (48.4)
CT	518 (27.9)
NH	215 (11.6)
RI	146 (7.9)
VT	79 (4.3)
Age (mean, sd)	15.1 (1.4)
Gender
Female	858 (46.2)
Male	913 (49.1)
Nonbinary/other	87 (4.7)
Sexual Orientation
Heterosexual	1527 (82.2)
Not Heterosexual	331 (17.8)
Race
White	1377 (74.1)
Black	141 (7.6)
Asian	94 (5.1)
Other	96 (5.2)
Multiple	150 (8.1)
Ethnicity
Hispanic/Latina/x/o	282 (15.2)
Not-Hispanic/Latina/x/o	1576 (84.8)
Parental Education
Less than high school	418 (22.5)
High school graduate or more	1440 (77.5)
School Type
Public	1422 (76.5)
Not public	436 (23.5)
Employed for Pay	632 (34.0)
Self-rated Health
Less than good	134 (7.2)
Good or better	1724 (92.8)
Self-reported Depression	519 (27.9)
Household tobacco use	665 (35.8)
Organized Activities	895 (48.2)
Total Number of Social Media Sites Ever Used (mean, sd)	4.4 (2.1)

Note: We include Latinx as a label to describe gender-expansive people of Latin American origin or descent and as a gender-inclusive term that seeks to increase the visibility of gender-expansive people.

Characterizing the Dual-Vaping Prevalence

Prevalence of past 30-day vaping was 18.6% (Table 2); 5.6% reported vaping solely nicotine, 5.6% reported vaping solely cannabis (THC and/or CBD), and 7.5% reported past 30-day dual-vaping. Of the participants that reported past 30-day vaping, over one-third (38.8%) reported dual-vaping. Less than 1% (n=14) reported “Don’t know” for which substances they vaped in the past 30-days.

Table 2:

Prevalence of types of vaping for N=1858 participants

	N (%)
Past 30-day any Vaping
Nicotine	245 (13.2)
Cannabis (THC and/or CBD)	240 (12.9)
Don’t know	14 (0.8%)
Past 30-day Sole-Vaping
Nicotine	105 (5.7%)
Cannabis (THC and/or CBD)	100 (5.4%)
Past 30-day Dual-Vaping
Nicotine & Cannabis (THC and/or CBD)	140 (7.5%)

Substances Vaped the Most

Among participant observations that reported past 30-day dual-vaping (n=140), over half reported vaping nicotine most often (59.3%), over one-third reported vaping THC most often (36.4%), and 4.3% reported vaping CBD most often.

Substances Used to Initiate Vaping

Of the participants observations that reported past 30-day dual-vaping (n=70) from September 2021 onward, about half (54.3%) reported vaping only nicotine first, almost one-third (28.6%) reported that they initiated vaping only with some form of cannabis (THC and/or CBD) first, and 7.1% reported initiating vaping with both nicotine and THC at the same time.

Correlates of Vaping Status

Below we present the correlates of each vaping outcome, compared to no vaping. See Supplemental Tables 1 and 2 for comparisons of the vaping outcomes with sole nicotine vaping and sole cannabis vaping.

Past 30-day Sole-Nicotine Vaping vs. No Vaping

Having a parent who did not complete high school (Relative Risk Ratio (RRR): 1.78, 95% Confidence Interval (CI): 1.13-2.78), household tobacco use (RRR: 2.35, CI: 1.52-3.65) and higher number total of social media sites used (RRR: 1.19, CI: 1.08-1.30) were significantly associated with higher odds of sole-vaping of nicotine compared to no vaping (Table 3). Identifying as Hispanic/Latina/x/o (RRR: 0.40, CI: 0.20-0.80) was associated with lower odds of sole-vaping of nicotine compared to no vaping.

Table 3:

Regression Results of Adjusted Multinomial Logistic Model for Categorical Vaping Outcome (N=1858 participants; referent outcome is No Vaping), RRR (95% Confidence Interval), p-value

	Past 30-day Sole Nicotine Vaping	Past 30-day Sole Cannabis Vaping	Past 30-day Dual-Vaping
Time (wave)	1.05 (1.01-1.10), 0.018	0.99 (0.95-1.04), 0.727	1.05 (1.01-1.10), 0.013
State
MA (ref)
CT	1.34 (0.84-2.14). 0.220	0.92 (0.56-1.53), 0.762	1.20 (0.73-1.97), 0.469
NH	0.88 (0.47-1.68), 0.709	0.65 (0.30-1.42), 0.282	0.62 (0.31-1.24), 0.172
RI	0.44 (0.16-1.27), 0.129	1.22 (0.57-2.60), 0.615	1.83 (1.00-3.33), 0.048
VT	0.70 (0.21-2.33), 0.564	1.16 (0.38-3.52), 0.791	2.31 (1.12-4.76), 0.023
Age	1.07 (0.92-1.25), 0.404	1.59 (1.32-1.92), <0.001	1.40 (1.19-1.64), <0.001
Gender
Female (ref)
Male	0.96 (0.62-1.50), 0.873	1.28 (0.82-1.98), 0.279	0.88 (0.58-1.34)
Nonbinary/other	0.28 (0.06-1.42), 0.125	0.23 (0.05-0.99), 0.049	0.32 (0.12-0.82), 0.018
Sexual Orientation
Not Heterosexual	0.91 (0.45-1.84), 0.800	1.53 (0.84-2.76), 0.161	1.49 (0.84-2.64), 0.170
Race
White (ref)
Black	0.31 (0.09-1.00), 0.050	1.27 (0.59-2.73), 0.534	0.45 (0.19-1.07), 0.072
Asian	1.43 (0.65-3.13), 0.371	0.51 (0.11-2.25)	N/A*
Other	0.90 (0.33-2.46), 0.837	0.66 (0.20-2.15), 0.494	0.42 (0.14-1.33), 0.143
Multiple	0.52 (0.20-1.35), 0.178	1.46 (0.76-2.83), 0.256	0.54 (0.24-1.20), 0.131
Ethnicity
Hispanic/Latina/x/o	0.40 (0.20-0.80), 0.009	1.06 (0.58-1.94), 0.850	0.40 (0.20-0.78), 0.008
Parental Education
Less than High School	1.78 (1.13-2.78), 0.012	1.45 (0.91-2.32), 0.122	1.16 (0.75-1.80), 0.492
School Type
Public	1.50 (0.87-2.58), 0.148	1.57 (0.91-2.70), 0.103	1.29 (0.79-2.10), 0.307
Employed for Pay
Yes	0.98 (0.62-1.55), 0.926	1.92 (1.21-3.02), 0.005	1.98 (1.33-2.94), 0.001
Self-rated Health
Good or better	1.32 (0.52-3.36), 0.557	0.88 (0.42-1.82)	0.74 (0.38-1.45)
Self-reported Depression
Yes	1.26 (0.80-2.00), 0.324	1.87 (1.15-3.06), 0.012	1.18 (0.74-1.86), 0.487
Household tobacco use
Yes	2.35 (1.52-3.65), <0.001	3.13 (2.05-4.77), <0.001	5.48 (3.60-8.34), <0.001
Organized Activities
Yes	0.68 (0.44-1.05), 0.081	0.87 (0.55-1.38), 0.557	0.78 (0.53-1.15), 0.214
Total Social Media Sites	1.19 (1.08-1.30), <0.001	1.08 (0.97-1.21), 0.165	1.37 (1.25-1.50), <0.001
COVID-19 Impact Score	1.09 (0.95-1.26), 0.230	1.11 (0.95-1.30), 0.169	1.23 (1.07-1.41), 0.003

Note:

^*N/A due to no observations in this cell

Past 30-day Sole-Cannabis Vaping vs. No Vaping

Age (RRR: 1.59, CI: 1.32-1.92), paid employment (RRR: 1.92, CI: 1.21-3.02), self-reported depression (RRR: 1.87, CI: 1.15-3.06), and household tobacco use (RRR: 3.13, CI: 2.05-4.77) were significantly associated with higher odds of sole-vaping of cannabis compared to no vaping. Identifying as non-binary gender (RRR: 0.23, CI: 0.05-0.99) (vs. female) was associated with lower odds of sole-vaping of cannabis compared to no vaping.

Past 30-day Dual-Vaping vs. No Vaping

Older age (RRR: 1.40, CI: 1.19-1.64), paid employment (RRR: 1.98, CI: 1.33-2.94), household tobacco use (RRR: 5.48, CI: 3.60-8.34), higher number total of social media sites used (RRR: 1.37, CI: 1.25-1.50), and higher COVID-19 impact score (RRR: 1.23, CI: 1.07-1.41) were associated with significantly higher odds of dual-vaping compared to no vaping. Living in VT (RRR: 2.31, CI: 1.12-4.76) and RI (RRR: 1.83, CI: 1.00-3.33) (vs. MA) was significantly associated with higher odds of dual-vaping compared to no vaping. Identifying as non-binary gender (RRR: 0.32, CI: 0.12-0.82) (vs. female), and Hispanic/Latina/x/o ethnicity (RRR: 0.40, CI: 0.20-0.78) (vs. non-Hispanic) were associated with lower odds of dual-vaping compared to no vaping.

Past 30-day Dual-Vaping vs. Sole Nicotine Vaping

Older age (RRR: 1.31, CI: 1.06-1.62), living in RI (RRR: 4.12, CI: 1.26-13.47) (vs. MA), paid employment (RRR: 2.03, CI: 1.14-3.59), household tobacco use (RRR: 2.33, CI: 1.31-4.15), and higher number of total social media sites used (RRR: 1.16, CI: 1.03-1.30) were associated with significantly higher odds of dual-vaping compared to sole nicotine vaping (Supplemental Table 1).

Past 30-day Dual-Vaping vs. Sole Cannabis Vaping

Household tobacco use (RRR: 1.75, CI: 1.01-3.05), higher number of total social media sites used (RRR: 1.27, CI: 1.11-1.45), and higher COVID-19 impact score (RRR: 1.11, CI: 0.82-1.33) were associated with significantly higher odds of dual-vaping compared to sole cannabis vaping. Older age (RRR: 0.88, CI: 0.69-1.11), identifying as multiple races (RRR: 0.37, CI: 0.14-0.97) (vs. White), and identifying as Hispanic/Latina/x/o (RRR: 0.37, CI: 0.16-0.88) were associated with significantly lower odds of dual-vaping compared to sole cannabis vaping (Supplemental Table 2).

DISCUSSION

Our findings show that over one-third of a sample of New England adolescents who reported past 30-day vaping had dual-vaped nicotine and cannabis, which is consistent with other studies that have measured co-use of nicotine and cannabis.^49,50 Of our sample who had dual-vaped, over 40% reported vaping cannabis the most (of nicotine and cannabis), further highlighting this increased popularity of vaping as a mode of administration of cannabis among adolescents.^{11,43,51–53} This is of particular concern, given the high concentrations of THC now found in cannabis e-liquids,^18,19,24 and the potential harms of cannabis to the adolescent developing brain.^{54, 55} While nicotine and cannabis have separate cognitive impacts on brain development, their interaction may cause further cognitive deficits.⁵⁶ Cannabis use has also been found to be associated with greater nicotine addiction among adolescents.⁵⁷

Studies tend to show that adolescents begin with tobacco and nicotine use, then move to cannabis use.^58,59 Yet, we found that almost one-third of our sample who had dual-vaped initiated vaping with cannabis first prior to nicotine vaping. Future research on dual-vaping behavior should attempt to monitor which substances adolescents initiate vaping with and then inform the focusing of early prevention interventions on specific substances to prevent future multiple substance use.¹¹

Different correlates were associated with each of the behaviors of sole-nicotine vaping, sole-cannabis vaping, and dual-vaping. We found that participating in paid employment, having household tobacco use, and using more social media sites were correlates associated with dual-vaping, supporting previous research on the importance of these correlates with risk of multiple substance use.^{42,44,60–62} This may be due to peers offering these products to them at their workplace, seeing colleagues using them at work, or observing a sibling or parent dual-vaping at home.^44,62 These findings are consistent with a recent study in Pennsylvania that found peer vaping and positive e-cigarette expectations as associated risk factors of adolescent dual-vaping compared with no use.⁵² Similarly, a recent study in Massachusetts found that adolescent dual-use of cannabis and e-cigarettes in the past 30-days was associated with participation in paid employment and household tobacco and cannabis use.⁴²

Having a higher COVID-19 impact score was associated with dual-vaping but not sole-vaping behaviors, suggesting that those experiencing the most social restrictions during the COVID-19 pandemic were the ones using more types of substances, possibly as an outlet for the stress of the pandemic. This is consistent with prior research that has found that although overall prevalence of vaping decreased during the COVID-19 pandemic, those who were vaping were vaping more frequently and heavily.⁶³ This finding also highlights the relevance of mental health and dual-vaping, as adolescents may be vaping multiple substances to cope with stress and negative emotions.⁶⁴ We found that participants who identified as non-binary gender (vs. female) had higher odds of reporting dual-vaping and sole-cannabis vaping. This may be due to the gender-related minority stressors that these individuals experience,⁶⁵ and prior research has found higher levels of substance use, including cannabis vaping, among adolescents who identify as a sexual or gender minority.^43,66–70

Self-rated depression and having a parent who did not complete a bachelor’s degree were associated only with sole-cannabis vaping. Emotional problems and socioeconomic background are early individual and environmental risk factors of adolescent substance use,⁷¹ and those adolescents who are only vaping cannabis may be more heavily influenced by these risk factors and health disparities. Correlates related to lower economic status, including lower parental education and reporting working for pay, tended to be associated with the vaping outcomes, similar to prior work around adolescent e-cigarette use disparities.^40,72 These findings provide justification for inclusion of more detailed economic measures in the assessment of tobacco and cannabis use during adolescence, as such measures are largely missing from large surveillance surveys. Future interventions to prevent vaping among adolescents from lower socioeconomic backgrounds may be needed.

Our study utilized a new measure to ask more specifically what substances adolescents are vaping with their e-cigarette devices. It should become a standard for all surveys to ask more granularly what substances people are putting into their vaping devices when measuring use, to more fully capture not just nicotine vaping, but also cannabis and dual-vaping.⁷³ Cannabis and nicotine are very different substances with different health and behavioral consequences,^8,55,74 and should be distinguished when characterizing vaping use and behaviors. To ensure accurate data capture, it should be standard in adolescent vaping screening and surveillance tools to include non-nicotine substances such as THC and/or CBD. Vaping of other non-nicotine substances, such as melatonin, have also become more popular among adolescents,⁷⁵ and the health claims of these non-nicotine vapes may reduce harm perceptions of vaping among adolescents.

Limitations and Future Directions

Our operationalization of dual-vaping was unable to capture when exactly participants were vaping nicotine versus cannabis in the past 30-days, and whether participants were concurrently vaping both nicotine and cannabis within the same vaping session. When asking about the substance participants vaped first, we did not ask for age of initiation for each substance, leaving room for recall bias. To enhance efficiency of data collection, we used an opt-in online panel that was not representative of adolescents in the five states. Our approach is not dissimilar from other studies that have used non-representative samples to assess novel substance use behaviors among adolescents.⁵²,⁷⁶ Another limitation is that this study recruited adolescent participants through their parents; therefore, certain adolescents may be less likely to participate, for example, those who did not want their parents to know about their vaping behaviors. Our study focused on the U.S. geographic region of New England where state policies have restricted the sales of flavored tobacco products and followed the legislative trend of legalizing adult-use cannabis. Our findings would not be generalizable to adolescents in every state.

Conclusion

Our findings demonstrate that adolescents in the New England region who vape are more likely to report past 30-day dual-vaping than sole-vaping either nicotine or cannabis alone, making dual-vaping the single largest vaping behavior category. Capturing what adolescents are vaping means asking what substances adolescents are vaping in their devices, and to specifically include non-nicotine substances such as THC, CBD, or non-psychoactive substances.⁷⁵ Future research should implement more granular measures of vaping and examine the differing correlates of sole- vs. dual-vaping. Our study addresses a key gap in adequate survey measures for assessing adolescent vaping of multiple substances and may help in characterizing and targeting future surveillance efforts for adolescent vaping.⁷³

Highlights.

• Vaping includes not just nicotine but also cannabis

• This study identified prevalence and correlates of dual-vaping of nicotine/cannabis

• Prevalence of past 30-day dual-vaping was more prevalent than sole-vaping

• Correlates differed for sole-vaping and dual-vaping of nicotine and cannabis.