Cannabis use among adolescents and emerging adults who use e-cigarettes: Findings from an online, national U.S. Sample

Abstract

Concurrent use of e-cigarettes and cannabis among adolescents and emerging adults is a growing public health concern. More research is needed describing cannabis use among adolescents and emerging adults who vape. The objective of this study was to characterize cannabis use among young people who had ever used e-cigarettes (age 14–20), particularly their use of blunts and liquid cannabis vape (LCV) products. Using cross-sectional data from a national online survey, we describe their patterns of cannabis use, detail their use of flavored cannabis and tobacco products, and estimate associations of demographic factors and other current substance use behaviors with levels of blunt and LCV use. Of the 2253 respondents in the sample, 1379 (61%) reported some form of cannabis use in the past 30 days, among whom 80% used flavored cannabis (including edibles). Significant associations with current cannabis use were observed on several demographic measures, with current cannabis blunt use more frequent among participants not in school, non-Hispanic Blacks, multiracial respondents, and those whose incomes do not meet their expenses. Other than income, demographic characteristics were generally not associated with LCV use frequency. Use of other substances was associated with more frequent use of both blunts and LCV in the past 30 days, and enrollment in college or the military seems somewhat protective for emerging adults. These findings suggest a need for tailored prevention efforts among high-risk adolescents and emerging adults, potential regulation of added flavors in commercialized cannabis products, and stronger enforcement of retail restrictions for individuals under age 21 more broadly.

1.1. Introduction:

Cannabis use among adolescents and emerging adults is a major public health concern. In 2019, 37% of high school students reported lifetime cannabis use, and 22% reported use in the past 30 days.^1,2 While minors (individuals under age 21) do not have access to legal cannabis in any state outside of a controlled medical program, minors in states with legalized medical or recreational cannabis report greater access to cannabis products.^3,4 The potential harms of cannabis use for adolescents are well established,^5–9 specifically the potential for long-term effects of regular use on adolescent brain development.⁸ Negative effects of cannabis use among adolescents include difficulty with cognitive functions like learning, thinking, and problem solving, in addition to memory loss, difficulty maintaining attention, and impaired coordination.¹⁰ Moreover, adolescents engaging in heavy cannabis use have increased risks for mental health problems, including anxiety, depression, substance use disorder, psychosis, and schizophrenia.^11,12

Cannabis and tobacco products are growing increasingly intertwined from regulatory and public health perspectives, with increasing concurrent use (i.e., use of both substances in the same time frame), including using tobacco products to deliver cannabis in both combustible and aerosolized forms.^4,13–18 While reasons for cannabis and tobacco use among this population may differ, there are likely some shared predisposing factors. For example, added flavors, particularly sweet flavors, may motivate youth initiation and continued use of tobacco and e-cigarettes.^19–23 Individuals using cannabis are exposed to added flavors through flavored wrapping papers for joints, flavored cigars for blunts, and flavor enhanced cannabis leaf or flower. Limited data exist on the prevalence of flavored cannabis use, with one smaller study suggesting flavor use similar to some tobacco products.²⁴

The combined use of cannabis and tobacco in blunt form is a major factor in the high prevalence of cigar use among adolescents,²⁵ especially among African American youth.^26–28 Blunts are particularly attractive to adolescents and emerging adults who use cannabis as the cigars used for blunts are inexpensive and widely available, specific brands have shell perforations and other design elements making them ideal for blunt rolling, blunts are also often used publicly by celebrities, and blunt smoking is often seen as a social activity.^29–34

As the cannabis industry has expanded, new products similar in design and function to e-cigarettes have become broadly available and use among this population is increasing.^14–16 In 2019, past 30-day prevalence of cannabis vaping was reported by 14% of 12^th graders, and past 30-day prevalence for all high school students increased significantly over one year.³⁵ New commercial “e-cannabis” products allow for non-combustible means of consuming cannabis liquid, oil, and wax (hereafter termed “Liquid Cannabis Vape” or LCV), often with added flavors. LCV products, like e-cigarettes, may also be perceived as safer and more discreet than combustible cannabis.³⁶ LCV products are not harmless; cannabis vaping produces greater physiological and psychological effects compared with traditional smoking.³⁷ Moreover, in 2019 illicit LCV products containing vitamin E acetate were implicated in an outbreak of severe lung injury, raising concerns about potential health effects of cannabis vaping.³⁸

This study characterizes patterns of cannabis use via blunts and LCV among adolescents and emerging adults who use e-cigarettes. It uses cross-sectional data from a national (U.S.) online survey of individuals aged 14–20 years who report repeated use of e-cigarettes to estimate relative risk of more frequent blunt and LCV use.

2.1. Methods:

This study used a cross-sectional national online survey of U.S. adolescents (14–17 years) and emerging adults (18–20 years) who reported ever using e-cigarettes at least 3 times. A threshold of ≥3 times was set as an inclusion criterion to help assure the study population included participants familiar with e-cigarettes from repeated use.³⁹ Participants were recruited from existing, actively managed market research panels aggregated by Qualtrics. Those potentially meeting predetermined eligibility criteria were provided a description of the study’s goals, expectations, and voluntary nature and then given the option to participate. Of 8860 who completed screener questionnaires, 2712 participants met eligibility criteria (e-cigarette use ≥3 times and age 14–20), and 2253 completed the survey. Surveys were completed between March 18, 2021, and April 25, 2021. The median survey completion time was 9 minutes. Participants were offered incentives that varied by panel but typically consisted of points redeemable for awards. Approval for this research was granted by the Internal Review Board at the University of California San Francisco.

2.1.1. Data

The survey included adaptive questioning measures related to the use of e-cigarettes and other tobacco products, cannabis, and alcohol, in addition to flavored commercial product use and sociodemographic characteristics. State and personal identifiers were not collected.

2.1.2. Independent Variables

Demographic variables included constructed categorical indicators for age and school enrollment status (age 14–17 and enrolled in school; age 18 or older and enrolled in high school; age 18 or older and enrolled in college, training program, or active-duty military; any respondent not enrolled school, college, or military), gender (female, male, non-binary or refused), race/ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, Asian or Pacific Islander, Multiracial, or some other race), and income sufficiency (“Live comfortably”; “Meet needs with a little left over”; “Just meet basic expenses”; “Do not meet basic expenses”). Dichotomous indicators were also constructed for total lifetime use of e-cigarettes (<100 times; ≥100 times).

To assess lifetime substance ever-use we used the question, “Please select which of the following products you have ever used,” to construct dichotomous indicators of use for the various substance use categories. Similarly, to assess past 30-day substance use we used the question “Of the products you just selected, please indicate how many days you used that product in the past 30 days.” We also used these measures to construct dichotomous indicators of use (no use in past 30 days; used 1 or more days in past 30 days) for the various substances used. The substance use categories included in the analysis were cannabis (joints, blunts, waterpipes, dry leaf vapes, liquid/oil/wax vapes/dabs, edibles), e-cigarettes (any type of e-cigarette, including stick-like disposable, cigalike disposable, reusable pod, mod, rebuildable, and pen-like refillable devices), smokeless tobacco products (chewing tobacco, moist snuff, and snus), combustible tobacco products (cigarettes, tobacco only cigars, and hookah/waterpipes), and alcohol (beer, wine, and liquor).

The survey included an adaptive format where respondents who reported past 30-day use of a given product were then asked follow-up questions on the use of flavors in those products. For consistency in assessing the use of flavored products we transformed more detailed flavored product reports into a more general form (i.e., any flavored product use, no flavor or unflavored product, or I don’t know).

2.1.3. Dependent Variables:

The survey included measures capturing cannabis use by modality. To capture blunt use, the question “In the past 30 days, have you used marijuana (cannabis) smoked as a blunt (a cigar with marijuana inside)” was used. To capture LCV use, the question “In the past 30 days, have you used any marijuana (cannabis) oil, wax, or liquid to vape or dab” was used. Outcomes for this analysis included constructed measures for past 30-day frequency of cannabis use by modality (cannabis smoked as a blunt in the past 30 days and LCV used in past 30 days). We began by creating a nominal variable where frequency of product use in the past 30 days was grouped into five categories for analysis, “no [blunt/LCV] use ever”, “no [blunt/LCV] used in the past 30 days”, “used [blunts/LCV] 1–5 days”, “used [blunts/LCV] 6–19 days”, and “used [blunts/LCV] 20 or more days”. This approach for presenting prevalence of past 30-day cannabis use as nominal variable follows the approach used by the CDC to present tobacco surveillance data from the National Youth and Tobacco Survey, with the addition of including a category for those who have ‘never used’ blunts or LCV for a more complete quantitative analysis.⁴⁰

2.1.4. Statistical Approach:

This study followed an exploratory approach to describe cannabis use patterns. First, descriptive statistics were produced for past 30-day cannabis use, and for demographic variables. Univariate analyses were performed to describe the sample. Group differences on cannabis use for discrete variables were identified using chi-square tests. Descriptive statistics were also produced for flavor use across products. Finally, multinomial logistic regressions were used to estimate relative risk ratios of the constructed categories for cannabis use frequency in the past 30 days via blunts and LCV, while adjusting for school enrollment status by age group, other demographic characteristics, and co-use of other substances. Only 1 observation was missing across the co-variables used in multivariable analysis; thus, models were fitted with listwise deletion. The Benjamini-Hochberg Procedure (BHP) for controlling the false discovery rate was used.⁴¹ An adjusted p value of .05 or less was considered statistically significant on all two-tailed tests. All data management and statistical analyses were performed using Stata 17.1 (StataCorp). Reporting is consistent with the CHERRIES checklist for cross-sectional studies.⁴²

3.1. Results:

Of the 2253 respondents in the sample,1379 (61%) reported using cannabis in the past 30 days (Table 1). Most of the sample (73%) were emerging adults aged 18–20 years. The largest share of respondents was age 18 and over and enrolled in college, a training program, or the military (36%), followed by adolescents enrolled in middle school or high school (27%), and respondents 18 and over and in high school (16%). Most respondents were female (65%), while 32% were male, and 3% were non-binary or refused the question. The majority of respondents were also non-Hispanic Whites (57%), while 19% were Hispanic, 9% were non-Hispanic Blacks, and Asians/Pacific Islanders, multiracial respondents, and those of some other race were each roughly 5%. Most reported their personal income was sufficient, allowing them to live comfortably (41%) or meeting their basic needs with a little left over (30%).

Table 1.

Characteristics of the Study Population by Current (Past 30-Day) Cannabis Use Status

	All Respondents	No Cannabis Used in Past 30 days	Used Cannabis in Past 30 days	Adj. p - Value 1
	(N = 2253)	(n = 874)	(n = 1379)
DEMOGRAPHIC CHARACTERISTICS
Age group category				<.001
Adolescents (age 14 –17 years)	27.5%	31.9%	24.7%
Emerging adults (age 18–20 years)	72.5%	68.1%	75.3%
Age & School enrollment				0.02
Age 14–17, enrolled in school	26.6%	30.6%	24.0%
Age 18 or older and enrolled in high school	16.1%	15.1%	16.7%
Age 18 or older and enrolled in college, training program, or active-duty military	36.3%	34.9%	37.1%
Any respondent not enrolled school, college, or military	21.1%	19.4%	22.2%
Gender				0.63
Female	65.1%	66.3%	64.4%
Male	31.5%	30.9%	31.9%
Non-binary or refused	3.4%	2.9%	3.7%
Race/Ethnicity				0.04
White (non-Hispanic)	56.9%	56.8%	56.9%
Hispanic or Latinx	18.8%	20.7%	17.6%
Black or African American (non-Hispanic)	8.9%	7.8%	9.6%
Asian or Pacific Islander (non-Hispanic)	5.3%	6.2%	4.8%
Bi-racial or Multi-racial (non-Hispanic)	4.8%	3.2%	5.7%
Some other race2 (non-Hispanic)	5.3%	5.4%	5.3%
Personal income				0.17
Live comfortably	42.6%	44.6%	41.3%
Meet needs with a little left over	29.5%	30.3%	28.9%
Just meet basic expenses	20.6%	18.9%	21.6%
Do not meet basic expenses	7.4%	6.2%	8.2%
SUBSTANCE USE BEHAVIORS
Used any cannabis product ever
No	14.6%	37.5%	0.0%
Yes	85.4%	62.5%	100.0%
Used cannabis more than 100 times ever				<.001
No	59.3%	84.9%	43.0%
Yes	40.8%	15.1%	57.0%
Used any cannabis product in past 30 days
No	38.8%	100.0%	0.0%
Yes	61.2%	0.0%	100.0%
Used e-cigarettes more than 100 times ever				<.001
No	48.3%	57.9%	42.1%
Yes	51.7%	42.1%	57.9%
Used e-cigarettes in past 30 days				<.001
No	26.8%	45.5%	14.9%
Yes	73.2%	54.5%	85.1%
Used smokeless tobacco in past 30 days				<.001
No	84.4%	90.9%	80.4%
Yes	15.6%	9.2%	19.7%
Used combustible cigarettes in past 30 days				<.001
No	74.1%	85.1%	67.1%
Yes	25.9%	14.9%	32.9%
Used cigars, cigarillos, or small cigars3 in past 30 days				<.001
No	84.0%	93.8%	77.8%
Yes	16.0%	6.2%	22.2%
Used hookah, shisha, or tobacco water pipe in past 30 days				<.001
No	87.4%	96.3%	81.8%
Yes	12.6%	3.7%	18.2%
Used alcohol (like beer, wine, or liquor) in past 30 days				<.001
No	39.0%	60.4%	25.5%
Yes	61.0%	39.6%	74.6%

¹The Benjamini-Hochberg Procedure was used to control the false discovery rate via multiple comparisons, adjusted p-values are presented here

²Some other race included Native American, Native Hawaiian, or Pacific Islander; Middle Eastern, or “Other”

³Cigar use category was ‘tobacco only (excludes blunts)’

Cannabis use in our sample was more common among emerging adults, non-Hispanic Blacks, multiracial (2 or more non-Hispanic) respondents, and those whose incomes do not meet their expenses (Table 1). In terms of other substance use behaviors, self-reported use of multiple substances in the past 30 days was common: those using cannabis in the past 30 days were more likely than those not using cannabis to report using e-cigarettes, other forms of tobacco, and alcohol in the past 30 days.

Cannabis use in multiple modalities was common in this sample (Table 2). In total, 61% of respondents reported cannabis use in the past 30 days. Of these, 75% reported smoking cannabis in joints, spliffs, or a small pipe, 62% reported smoking blunts, and 70% reported using liquid cannabis vape (LCV). Respondents using cannabis in the past 30 days were also likely to use e-cigarettes (85%) and alcohol (75%). Among those using other forms of tobacco, the prevalence of cannabis use was 85% among those using cigars, 89% among those using hookah/waterpipe, and above 70% among those using the other tobacco products examined.

Table 2.

Past 30-Day Substance Use Behaviors by Product Type

	Past 30-day Substance Use 1
	Full sample	Cannabis	E-cigarettes	Smokeless tobacco	Cigarettes	Cigars	Hookah/Water pipe	Alcohol
	(N = 2253)	(n = 1379)	(n = 1649)	(n = 351)	(n = 584)	(n = 360)	(n = 283)	(n = 1374)
CANNABIS USE BEHAVIORS
Cannabis use of any kind	61.2%	100.0%	71.1%	77.2%	77.7%	85.0%	88.7%	74.8%
Cannabis smoked as a joint, spliff, or in a small pipe	46.1%	75.3%	54.8%	62.7%	65.8%	74.7%	77.7%	58.7%
Cannabis smoked as a blunt (marijuana filled cigar)	38.2%	62.4%	46.3%	58.1%	56.5%	73.3%	75.6%	48.4%
Cannabis smoked from a bong, water pipe, or “moke”	39.9%	65.2%	48.2%	58.1%	62.2%	71.7%	74.6%	51.4%
Cannabis flower, bud, or leaf from a dry leaf vaporizer	22.5%	36.7%	27.6%	47.6%	36.1%	49.4%	58.0%	29.8%
Cannabis oil, wax, or liquid to vape or dab	43.1%	70.4%	51.1%	59.0%	57.5%	69.4%	71.4%	54.9%
Cannabis in an edible cookie, candy, food, or drink	34.7%	56.8%	41.4%	54.4%	49.1%	58.6%	67.8%	45.6%
TOBACCO AND ALCOHOL USE BEHAVIORS
E-cigarettes or vapes	73.2%	85.1%	100.0%	95.4%	87.8%	91.1%	94.7%	85.7%
Smokeless tobacco products (chew, dip, snus, etc.)	15.6%	19.7%	20.3%	100.0%	33.2%	43.3%	47.7%	19.6%
Cigarettes	25.9%	32.9%	31.1%	55.3%	100.0%	67.5%	53.4%	32.5%
Cigars, cigarillos, or small cigars2	15.9%	22.2%	19.9%	44.4%	41.6%	100.0%	49.5%	20.3%
Hookah, shisha, or tobacco water pipe	12.6%	18.2%	16.3%	38.5%	25.9%	38.9%	100.0%	16.1%
Alcohol (like beer, wine, or liquor)	60.9%	74.5%	71.4%	76.6%	76.4%	77.5%	78.1%	100.0%

¹Non-mutually exclusive denominator categories

²Cigar use category was ‘tobacco only (excludes blunts)’

Across cannabis and tobacco products, flavored product use was common among those reporting past 30-day use (Table 3). In total, nearly 80% of those using cannabis indicated consuming any flavored cannabis product (including edibles). Other sources of flavored products included flavored wrapping paper for joints or spliffs, flavored cigars for blunts, or flavored cannabis liquid, oil, or wax for vaping or dabbing. Among all those reporting past 30-day cannabis use, 36% used flavored blunts and 36% used flavored LCV. However, 57% of those reporting past 30-day blunts use, used flavored cigars, and 51% of those who report past 30-day LCV use, used flavored cannabis liquid, oil, or wax for vaping. Flavored product use was also high across the various tobacco products used by those reporting cannabis use: for example, 92% of those using e-cigarettes and cannabis reporting any flavored e-cigarette use.

Table 3.

Current Use of Flavored Products Among Cannabis Users

	N	Any flavored product	No flavored product	I don’t know
FLAVORED CANNABIS PRODUCT USE
Any use of flavored cannabis products
All past 30-day cannabis users, edibles included	1379	79.6%
Past 30-day cannabis users, edibles excluded	1339	64.7%
Flavored cannabis flower, bud, or leaf
All past 30-day cannabis users	1379	24.4%	64.7%	10.9%
Dry leaf2 cannabis users only	1222	27.6%	60.2%	12.3%
Flavored wrapping paper used for joints or spliffs (excluding blunts)
All past 30-day cannabis users	1379	19.7%	73.0%	7.3%
Joint and spliff users only	1038	26.1%	64.2%	9.7%
Flavored cannabis blunt use
All past 30-day cannabis users	1379	35.8%	56.5%	7.7%
Blunt users only	860	57.4%	30.2%	12.3%
Flavored cannabis oil, wax, or liquid to vape or dab
All past 30-day cannabis users	1379	36.0%	58.5%	5.5%
Vaped or dabbed cannabis users only	970	51.1%	41.0%	7.8%
Use of flavors in tobacco products among cannabis users 1
Flavored e-cigarettes or vapes	1171	91.9%	5.2%	2.9%
Flavored moist snuff, chewing tobacco, or snus	142	76.8%	14.8%	8.5%
Flavored cigarettes (menthol)	454	60.4%	36.1%	3.5%
Flavored cigars, cigarillos, or small cigars (excluding blunts)	306	73.9%	16.3%	9.8%
Flavored hookah, shisha, or tobacco water pipe	250	77.6%	7.6%	14.8%

¹Flavored product use shown for each tobacco product among past 30-day dual-users of cannabis and that specific tobacco product

²Dry leaf users include those reporting past 30-day use of joint, spliff, pipe, blunt, waterpipe, or dry leaf vaporizer

Race/ethnicity, school or program enrollment, and recent substance use were correlated with past 30-day blunt use frequency, as modeled with multinomial logistic regression (Table 4). Relative to having tried but not used blunts in the past 30 days, never using blunts was less likely among emerging adults out of high school than among adolescents in middle or high school. Asian/Pacific Islander (1.87[1.17;3.00]) and Hispanic (1.50[1.12;2.02]) respondents were more likely to have never used blunts compared to non-Hispanic White respondents. Respondents who used e-cigarettes and alcohol in the past 30 days were less likely to have never used blunts.

Table 4.

Multinomial Logistic Regression Estimating Relative Risk Ratios of Current Cannabis Blunt Use Frequency¹

Cannabis Blunt Use in Past 30 Days	No blunt use ever RRR (95% CI)	No blunt use in past 30 days	Used blunts 1–5 days RRR (95% CI)	Used blunts 6–19 days RRR (95% CI)	Used blunts 20 or more days RRR (95% CI)
Age & School Enrollment
Age 14–17, enrolled in school	reference	reference	reference	reference	reference
Age 18–20, enrolled in high school	0.85 [0.60–1.20]	reference	0.77 [0.49–1.19]	1.41 [0.82–2.44]	0.91 [0.55–1.51]
Age 18–20, enrolled in college, training program, or military	0.61 [0.46–0.80]***	reference	0.73 [0.52–1.02]*	0.66 [0.41–1.06]	0.55 [0.36–0.84]*
Any respondent not enrolled in school, college, or military	0.57 [0.41–0.80]**	reference	0.79 [0.53–1.19]	1.36 [0.81–2.27]	1.21 [0.77–1.81]
Gender
Female	reference	reference	reference	reference	reference
Male	0.92 [0.72–1.18]	reference	1.15 [0.86–1.53]	1.03 [0.70–1.50]	1.14 [0.81–1.59]
Non-binary or refused	0.99 [0.55–1.76]	reference	0.86 [0.40–1.84]	0.70 [0.24–1.98]	0.62 [0.24–1.60]
Race/Ethnicity
Non-Hispanic White	reference	reference	reference	reference	reference
Hispanic or Latinx	1.50 [1.12–2.02]*	reference	1.52 [1.05–2.20]	2.08 [1.32–3.30]**	1.88 [1.22–2.89]*
Black or African American (non-Hispanic)	1.27 [0.83–1.94]	reference	1.78 [1.07–2.98]	2.60 [1.40–4.81]**	3.47 [2.04–5.92]***
Asian or Pacific Islander (non-Hispanic)	1.87 [1.17–3.00]*	reference	0.91 [0.46–1.78]	1.06 [0.41–2.73]	1.85 [0.90–3.80]
Bi-racial or Multiracial (non-Hispanic)	1.07 [0.59–1.95]	reference	2.53 [1.38–4.65]**	3.20 [1.51–6.81]**	3.54 [1.79–6.99]***
Some other race2 (non-Hispanic)	1.27 [0.76–2.11]	reference	1.05 [0.55–2.00]	1.79 [0.87–3.68]	1.41 [0.70–2.84]
Personal income
“Live comfortably”	reference	reference	reference	reference	reference
“Meet needs with a little left over”	0.97 [0.75–1.25]	reference	0.84 [0.62–1.15]	1.06 [0.69–1.62]	1.14 [0.78–1.68]
“Just meet basic expenses”	1.28 [0.94–1.73]	reference	1.18 [0.82–1.71]	1.93 [1.23–3.06]*	2.36 [1.56–3.58]***
“Do not meet basic expenses”	1.04 [0.63–1.69]	reference	1.40 [0.80–2.43]	2.25 [1.18–4.30]*	3.70 [2.11–6.47]***
Recent Substance Use Behaviors
None in past 30-days	reference	reference	reference	reference	reference
Used e-cigarettes in past 30 days	0.66 [0.51–0.85]**	reference	2.44 [1.67–3.58]***	2.17 [1.29–3.64]**	2.56 [1.61–4.05]***
Used smokeless tobacco in past 30 days	1.51 [1.04–2.20]	reference	1.23 [0.82–1.83]	1.14 [0.69–1.86]	0.99 [0.62–1.56]
Used combustible cigarettes in past 30 days	1.05 [0.77–1.41]	reference	1.28 [0.93–1.78]	1.40 [0.93–2.11]	1.29 [0.89–1.87]
Used cigars, cigarillos, or small cigars in past 30 days3	1.07 [0.68–1.68]	reference	2.91 [1.90–4.45]***	2.54 [1.52–4.25]***	4.55 [2.88–7.17]***
Used hookah, shisha, or tobacco water pipe in past 30 days	1.40 [0.83–2.34]	reference	3.00 [1.84–4.89]***	5.60 [3.28–9.57]***	4.04 [2.41–6.77]***
Used alcohol (like beer, wine, or liquor) in past 30 days	0.64 [0.51–0.81]***	reference	1.64 [1.21–2.22]**	2.52 [1.65–3.86]***	2.13 [1.48–3.06]***

¹N = 2252

²Race/ethnicity included Hispanic or Latinx, Non-Hispanic Black or African American, and Some other race. “Some other race” included Native American, Native Hawaiian, or Pacific Islander; Middle Eastern, or “Other Non-White”

³Cigar use category was ‘tobacco only (excludes blunts)’

⁴The Benjamini-Hochberg Procedure was used to control the false discovery rate via multiple comparisons; adjusted p-values are presented here:

^*= p<.05

^**= p<.01

^***= p<.001

Abbreviations: CI = confidence interval; RRR = relative risk ratio

Relative to having tried but not used blunts in the past 30 days, emerging adults enrolled in college, a training program, or the military were less likely to have used blunts 1–5 days compared to adolescents in middle or high school (0.73[0.52;1.02]). Multiracial respondents were more likely to use blunts (2.53[1.38;4.65]) 1–5 days compared to non-Hispanic White respondents. Respondents using e-cigarettes, tobacco only cigars, hookah/water pipes, and alcohol were also more likely to use blunts 1–5 days than 0 days compared to those not using these products.

Multiracial (3.20[1.51;6.81]); non-Hispanic Black (2.60[1.40;4.81]); and Hispanic (2.08[1.32;3.30]) respondents were more likely to have used blunts 6–19 days compared to non-Hispanic White respondents. Greater likelihood of using blunts 6–19 days was also observed for those indicating their income was insufficient. Respondents using e-cigarettes, tobacco only cigars, hookah/water pipes, and alcohol were also more likely to use blunts 6–19 days.

Emerging adults enrolled in college, a training program, or the military had lower risk of using blunts ≥20 days compared to adolescents in middle and high school. Multiracial (3.54[1.79;6.99]) and non-Hispanic Black (3.47[2.04;5.92]) respondents had the greatest risk of using blunts ≥20 days compared to non-Hispanic White respondents. Those whose income was insufficient and respondents using e-cigarettes, tobacco only cigars, hookah/water pipes, and alcohol were also more likely to use blunts ≥20 days.

School or program enrollment and recent substance use, but not gender or race/ethnicity, were correlated with past 30-day Liquid Cannabis Vape (LCV) use frequency, as modeled with multinomial logistic regression (Table 5). Relative to having tried but not used LCV in the past 30 days, those who reported using e-cigarettes had statistically lower likelihood of never using LCV (0.58[0.45;0.76]), while those using smokeless tobacco had greater likelihood (1.73[1.16;2.59]) of never using LCV.

Table 5.

Multinomial Logistic Regression Estimating Relative Risk Ratios of Current Liquid Cannabis Vape (LCV) Use Frequency¹

Liquid Cannabis Vape (LCV) Use in Past 30 Days	No LCV use ever RRR (95% CI)	No LCV use in past 30 days	Used LCV 1–5 days RRR (95% CI)	Used LCV 6–19 days RRR (95% CI)	Used LCV 20 or more days RRR (95% CI)
Age & School Enrollment
Age 14–17, enrolled in school	reference	reference	reference	reference	reference
Age 18–20, enrolled in high school	0.92 [0.64–1.33]	reference	0.86 [0.55–1.33]	1.15 [0.67–1.96]	1.15 [0.73–1.80]
Age 18–20, enrolled in college, training program, or military	0.79 [0.59–1.06]	reference	0.64 [0.45–0.90]*	0.61 [0.39–0.95]	0.61 [0.42–0.88]*
Any respondent not enrolled in school, college, or military	0.73 [0.52–1.03]	reference	0.53 [0.34–0.80]**	1.05 [0.65–1.71]	0.91 [0.60–1.38]
Gender
Female	reference	reference	reference	reference	reference
Male	0.94 [0.73–1.21]	reference	0.92 [0.68;1.25]	0.96 [0.66;1.39]	1.33 [0.98;1.80]
Non-binary or refused	1.08 [0.57–2.06]	reference	0.94 [0.14;2.12]	2.10 [0.91;4.83]	1.22 [0.54;2.78]
Race/Ethnicity
Non-Hispanic White	reference	reference	reference	reference	reference
Hispanic or Latinx	1.46 [1.03–1.91]	reference	1.16 [0.80–1.70]	1.01 [0.64–1.60]	1.16 [0.78–1.72]
Black or African American (non-Hispanic)	1.44 [0.95–2.18]	reference	1.16 [0.69–1.93]	0.69 [0.35–1.35]	0.93 [0.54–1.61]
Asian or Pacific Islander (non-Hispanic)	1.29 [0.80–2.09]	reference	0.87 [0.47–1.64]	0.83 [0.37–1.84]	0.80 [0.40–1.59]
Bi-racial or Multiracial (non-Hispanic)	1.46 [0.79–2.67]	reference	1.74 [0.88–3.44]	2.33 [1.12–4.86]	2.10 [1.07–4.12]
Some other race2 (non-Hispanic)	1.88 [1.08–3.25]	reference	1.55 [0.81–2.98]	0.52 [0.20–1.39]	1.27 [0.64–2.50]
Personal income
“Live comfortably”	reference	reference	reference	reference	reference
“Meet needs with a little left over”	0.96 [0.73–1.25]	reference	1.03 [0.75–1.42]	1.12 [0.74–1.68]	0.97 [0.69–1.36]
“Just meet basic expenses”	1.13 [0.83–1.55]	reference	1.10 [0.75–1.61]	2.08 [1.35–3.21]**	1.45 [0.99–2.13]
“Do not meet basic expenses”	0.96 [0.58–1.59]	reference	1.49 [0.85–2.60]	1.61 [0.82–3.15]	2.08 [1.21–3.58]*
Recent Substance Use Behaviors
None in past 30-days	reference	reference	reference	reference	reference
Used e-cigarettes in past 30 days	0.58 [0.45–0.76]***	reference	1.68 [1.15–2.44]*	1.59 [0.99–2.56]	1.76 [1.19–2.59]**
Used smokeless tobacco in past 30 days	1.73 [1.16–2.59]*	reference	1.56 [1.02–2.38]	1.45 [0.88–2.38]	1.00 [0.64–1.58]
Used combustible cigarettes in past 30 days	1.00 [0.73–1.37]	reference	1.10 [0.78–1.55]	1.03 [0.69–1.56]	1.24 [0.88–1.76]
Used cigars, cigarillos, or small cigars in past 30 days3	0.75 [0.48–1.17]	reference	1.60 [1.03–2.48]	2.40 [1.48–3.91]***	1.90 [1.22–2.93]**
Used hookah, shisha, or tobacco water pipe in past 30 days	1.82 [1.08–3.06]	reference	3.22 [1.92–5.40]***	3.87 [2.19–6.84]***	3.27 [1.93–5.54]***
Used alcohol (like beer, wine, or liquor) in past 30 days	0.84 [0.65–1.07]	reference	2.58 [1.88–3.54]***	2.66 [1.80–3.94]***	2.46 [1.78–3.40]***

¹N = 2252

²Race/ethnicity included Hispanic or Latinx, Non-Hispanic Black or African American, and Some other race. “Some other race” included Native American, Native Hawaiian, or Pacific Islander; Middle Eastern, or “Other Non-White”

³Cigar use category was ‘tobacco only (excludes blunts)’

⁴The Benjamini-Hochberg Procedure was used to control the false discovery rate via multiple comparisons; adjusted p-values are presented here:

^*= p<.05

^**= p<.01

^***= p<.001

Abbreviations: CI = confidence interval; RRR = relative risk ratio

Relative to having tried but not used LCV in the past 30 days, those not enrolled in school (0.53[0.34;0.80]) and emerging adults enrolled in college, a training program, or the military (0.64[0.45;0.90]) were less likely to use LCV 1–5 days compared to adolescents enrolled in school. Those using hookah/waterpipes and alcohol had greater likelihood of using LCV 1–5 days.

Greater likelihood of using LCV 6–19 days was observed among those whose income was ‘just enough’ to meet daily expenses (2.08[1.35;3.21]). Greater likelihood of using LCV 6–19 days was also observed for those using tobacco only cigars, hookah/waterpipe, and alcohol.

Lower risk of using LCV ≥20 days was observed for emerging adults enrolled in college, a training program, or the military compared to adolescents enrolled in school. Respondents who reported that their income does not meet daily expenses (2.08 [1.21;3.57]) were more likely to use LCV ≥20 days than 0 days, as were those using e-cigarettes, tobacco only cigars, hookah/waterpipes, and alcohol.

4.1. Discussion:

This study gives greater insight into the behaviors of adolescents and emerging adults who use e-cigarettes as they relate to cannabis use generally, and specific insights into the use of added flavors in various cannabis products and the demographic factors associated with more frequent use of blunts and LCV. This study adds to a growing body of literature exploring the intersection of tobacco and cannabis use,^43–48 demonstrating higher rates of multiple substance use among this high-risk population than generally observed in studies prior to Covid-19,^48,49 and offering important new findings for policy makers and prevention professionals.

In particular, this study allows for detailing cannabis and other substance use (i.e., tobacco and alcohol use) behaviors among adolescents and emerging adults already engaged in e-cigarette use, behaviors that could place them at elevated risk for experiencing the negative physiological and sociological effects of long-term drug and alcohol use.^50,51 Based on these results, not only is cannabis use widespread among adolescents and emerging adults who use e-cigarettes, but use of flavored cannabis products among those using cannabis approaches or exceeds flavored tobacco among those who report using various tobacco products. Cannabis consumption co-administered with tobacco (e.g., blunts) or by aerosolizing devices resembling e-cigarettes (e.g., LCV) is also common and represents a potential area for better coordination between tobacco and cannabis control policies and regulations to prevent use among minors.

Cannabis continues to be the most commonly used controlled substance by adolescents and emerging adults.⁵² As this study affirms, despite age restrictions, individuals under age 21 are consuming cannabis through a variety of modalities, including more traditional combustible methods like blunts, and via a growing number of new commercial ‘e-cannabis’ products like LCV. Previous research details how commercialization of cannabis directly drives greater use among this population by increasing access to products via diversion,^53,54 and indirectly through marketing and messaging framing cannabis use as safe, medically necessary, or normatively acceptable for recreation.⁵⁵ This was seen in restrictive Covid-19 policies, when in-person classes for adolescents and emerging adults were suspended, but in-person medical and recreational cannabis sales continued.^56,57 It is possible some adolescents and emerging adults have access to cannabis-based products through their state’s medical cannabis program.⁵⁸ The present study did not ask about medical cannabis use. However, the American Academy of Pediatrics stands in opposition to cannabis use in patients under age 21, save for ‘life-limiting or seriously debilitating conditions.’⁵⁹

Flavors are a plausible factor contributing to the use of both cannabis and tobacco. We found similarly high levels of flavored product use across both cannabis and tobacco products, demonstrating the appeal of flavors. It seems plausible that many of the reasons youth previously reported for using flavored tobacco hold for flavored cannabis, given the high flavored product use among those using cannabis and the similarity in delivery. Flavor is also a major driver in cigar use initiation⁶⁰ and a way to enhance or improve the taste of cannabis when used in blunts.^31,61 Our research shows commercialized cannabis products are easily accessed by minors despite retail restrictions. The cannabis industry may also be targeting this population with campaigns on social media marketing flavored ‘aromatherapy’ products, while minimizing the fact they also contain cannabis.⁶²

Co-administration of cannabis and tobacco via blunts in this study is strongly associated with use of tobacco only cigars, reinforcing findings indicating blunt use is a driver of cigar use among this population.²⁵ What is unclear is whether one precedes the other and acts as a catalyst for use, given use of cigarettes or other tobacco products can enhance or prolong the high associated with cannabis use.^13,29,30 Other studies have also suggested misconceptions about the risks of blunt use and its status as a form of both cannabis and tobacco.^63–67 The study does not assess risk perceptions of blunt use relative to other forms of cannabis use. However, we found greater LCV use than blunt use, raising the question of whether adolescents and emerging adults who use e-cigarettes because of perceived safety relative to combustible cigarettes, are also using LCV as a ‘safer’ alternative to smoking cannabis.

Here, LCV use is second only to any form of smoked cannabis use (i.e., joints, spliffs, or small pipes), and higher than blunt use. As observed elsewhere, those who reported more frequent LCV use also reported greater polysubstance use,⁶⁸ however we did not see similar differences associated with sex/gender. Given the associations here between LCV use and age/school enrollment status, regulators should consider tighter retail restrictions on LCV products. At the same time, appropriate substance use prevention programming is necessary to address increasing LCV use.

4.1.1. Limitations:

The limitations come from the cross-sectional, market panel survey design. Online research panels are used widely, including in tobacco control research⁶⁹ and among adolescents and emerging adults.^70,71 While panel members may differ from the general U.S. population,⁷² participants reflect diversity in income, race/ethnicity, and geography. While restricting the sample to include only those using e-cigarettes reduces generalizability to the population at large, it remains meaningful to consider substance use among adolescents and emerging adults already engaged in tobacco use. The questionnaire did not ask about medical cannabis program enrollment, obtaining cannabis from a dispensary, or intended medical use. State of residence was not recorded.

5.1. Conclusion:

This study characterizes cannabis use in a national sample of adolescents and emerging adults who report ever using e-cigarettes, describing use of flavored cannabis and tobacco products, and estimating the risk of past 30-day cannabis use via blunts and LCV at higher frequency. We found high levels of cannabis, tobacco, and alcohol use, raising concerns about product availability to adolescents and emerging adults. Flavored cannabis and tobacco use was high across all products, indicating flavors are widely accessible and a potential driver in co-use. We found greater risk of blunt use at higher frequencies for minority youth and those experiencing income insufficiency, factors associated with psychosocial distress, and found increased odds when using tobacco and alcohol. Regulators must focus on interrupting access to these substances by minors and better regulation of flavors in cannabis products.

Abbreviations:

LCV

liquid cannabis vape