Trends in US adolescent use of vaping and flavored solutions for marijuana consumption, 2021–2024

Abstract

Purpose:

To document recent, national trends in vaping and use of flavored vaping solutions among US adolescents who use marijuana.

Methods:

Data come from Monitoring the Future, which conducted annual, cross-sectional, nationally-representative surveys from 2021–2024 of 8^th, 10^th, and 12^th grade students.

Results:

Among adolescents who used marijuana in the past 12 months, the percentage who vaped it increased over the study period from 48% to 57% in 8^th grade (P<.05), 60% to 66% in 10^th grade (P=.07), and 58% to 67% in 12^th grade (P<.01). The percentage of adolescents who used a flavored solution when vaping marijuana switched from a minority to a majority over the study period, from 47% to 63% (P<.01) in 8^th grade, 41% to 53% (P<.01) in 10^th grade, and 36% to 50% (P<.01) in 12^th grade. The percentage of all adolescents who vaped a flavored marijuana solution in the past 12 months over the study period increased from 2% to 4% (P<.01) in 8^th grade, 5% to 6% in 10^th grade (P=.25), and 7% to 9% in 12^th grade (P<.01).

Discussion:

The percentage of US adolescents who vaped flavored marijuana solutions increased from 2021–2024, one of very few drug outcomes that increased since the pandemic onset. With this increase, adolescents who used flavors to vape marijuana shifted from a minority to a majority and vaping became more common among adolescents who used marijuana. These results point to flavored marijuana vaping solutions as an increasingly important target for research and policy on adolescent marijuana use.

INTRODUCTION

Vaping marijuana is a relatively new mode of marijuana use that differs from traditional combustible use in 3 ways. First, vaping solutions come in a wide variety of flavors other than marijuana. Fruit and and sweet flavors are currently the most popular among adoelscents, with other flavors available such as mint and spice (clove) [1]. Second, because marijuana vaping products are available at very high THC concentrations that surpass marijuana flower, vaping can produce significantly more intense intoxicating effects than smoking marijuana [2]. Third, marijuana vaping is relatively more concealable than use of a lighted marijuana cigarette [3]. These factors likely contributed to the rapid increase in US adolescent marijuana vaping from 2017 to 2019, when its 12-month prevalence at the national level more than doubled in just 2 years [4].

This study updates trends in marijuana vaping among US youth for the years 2021–2024. It first examines trends in vaping as a mode of use among adolescents who use marijuana. Given the recent, rapid growth of teen marijuana vaping [4–5], we expect a substantial portion of adolescents who use marijuana are now vaping it. The study then focuses on adolescents who vape marijuana and examines recent trends in their use of flavored solutions. Because flavors increase adolescent perceptions of vaping as fun and interesting [6–7], adolescent use of flavored marijuana solutions may be substantial and/or growing. Evidence of increasing youth use of flavored marijuana vaping products could motivate efforts to create educational and media campaigns to reverse the trend and/or serve as a basis for local and state policymakers to heighten stringency and enforcement of regulatory restrictions on flavored marijuana sales.

METHODS

Data come from Monitoring the Future’s (MTF) annual nationally representative, cross-sectional samples of US 8^th, 10^th, and 12^th grade students surveyed between February and June in the years 2021 to 2024 [8]. Personnel from the University of Michigan administered MTF surveys in classrooms, and students self-completed questionnaires during a normal class period. The University of Michigan institutional review board approved the study. Informed consent followed each school’s policy and was either passive consent, with an information letter sent to all parents/caregivers describing the survey and letting them know they could refuse their child’s participation, or active consent that required all students to have signed permission slips from parents/caregivers in order to participate in the survey.

Measures

Marijuana use in the past 12 months was assessed with the question: “The next questions are about MARIJUANA or cannabis (sometimes called pot, weed, or hashish). On how many occasions (if any) have you used marijuana (smoking, vaping, edibles) during the 12 months?” and was coded 1 for those who marked the categories “1–2 Occasions,” “3–5 Occasions,” “6–9 Occasions,” “10–19 Occasions,” “20–39 Occasions” or “40 or More;” it was coded 0 for the response “0 Occasions.” From 2021–2024, the survey text for this question was modified for randomly-selected subsamples to test potential updates. Prevalence did not differ for (a) questions that included additional instructions to “not count any use of CBD products” (half the questions in each grade, p>.50) or (b) the 1/3 of the 12^th grade sample who received an older version of this question that did not include “vaping” as a mode of marijuana use (p=.91). Analyses pooled results from all versions of the question.

Use of marijuana specifically by the mode of vaping was assessed with the question “During the last 12 months have you vaped marijuana?” Responses of “yes” were coded 1 and “no” were coded 0. Preferred flavor of marijuana solution was assessed with the question “When you vape marijuana, which type of flavor do you use most often? (Select one flavor that is the best fit.),” and the response options were “Sweet,” “Fruit,” “Sour,” and “Unflavored (tastes like cannabis).” Use of a flavored marijuana solution was coded 1 for students who selected any of the first 3 options and coded 0 for those who chose the last option of “Unflavored (tastes like cannabis).”

Demographic questions assessed race/ethnicity, sex, Census region of the country, and parental education. Race/ethnicity was assessed with the question, “How do you describe yourself (Select one or more responses).” Participants were coded as “White” if they marked the response category “White;” as “Black” if they marked the response category “Black or African American;” as “Hispanic” if they marked any of the response categories “Mexican American or Chicano,” “Cuban American,” “Puerto Rican,” or “Other Hispanic or Latino;” as “Asian” if they marked “Asian American;” as “Other race/ethnicity” if they marked “American Indian or Alaska Native,” “Native Hawaiian or Other Pacific Islander,” or “Middle Eastern;” and as “More than one race/ethnicity” if they marked more than one of the above categories. Sex was assessed with the question “What is your sex?” and coded per the response categories of “Female,” “Male,” “Other,” and “Prefer not to answer.” Census region was measured as the student’s school location in the South (Delaware, District of Columbia, Florida, Georgia, Maryland, North Carolina, South Carolina, Virginia, West Virginia, Alabama, Kentucky, Mississippi, Tennessee, Arkansas, Lousiana, Oklahoma, and Texas), Midwest (Indiana, Illinois, Michigan, Ohio, Wisconsin, Iowa, Kansas, Minnesota, Missouri, Nebraska, North Dakota, and South Dakota), Northeast (Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont, New Jersey, New York, Pennsylvania), or West (Arizona, Colorado, Idaho, New Mexico, Montana, Utah, Nevada, Wyoming, Alaska, California, Hawaii, Oregon, and Washington). Parental educational status is an important component of household socioeconomic status; in this study parent has a college education was assessed with 2 separate questions: “What is the highest level of schooling your mother completed?” and “What is the highest level of schooling your father completed?”, for which the response categories were “Completed grade school or less,” “Some high school,” “Completed high school,” “Some college,” “Completed college,” “Graduate or professional school after college,” and “Don’t know, or does not apply.” Respondents were coded 1 for parental college if they reported either their mother or father had completed college or graduate/professional school after college and 0 if neither parent had.

The analyses used the randomly-selected half of the total MTF sample who received surveys with questions about their use of flavored marijuana solutions for vaping.

All statistical analyses used Stata MP version 18.5-MP2 software with the svy: suite of commands to take into account the complex multistage sample design, including clustering of respondents in schools. All estimates were weighted so that they are nationally representative of students in school. The analyses used multiple imputation to handle missing data [9] (<6% for all variables except parental education, which had 15% missing data) and used the chained equations algorithm with 20 imputed data sets [10]. The analyses used Wald χ2 analyses to test if prevalence levels significantly differed across adjacent years, as well as Poisson regression with a log link to produce relative risk ratios and associated standard errors in bivariate tests of 4-year trends and for multivariable models.

RESULTS

Descriptive Analyses of Sample

Table 1 presents the demographic descriptives of the pooled sample for 8^th, 10^th, and 12^th grade students, and Supplemental eTable 1 presents these descriptives stratified by grade. The sample consists of 54 373 students, with 17 355 in 8^th grade, 21 132 in 10^th grade, and 15 886 in 12^th grade. In the pool of all participants, 48% were male (range by grade level of 47%-49%), 46% were female (range, 46% to 47%), 2% reported “other” for sex (range 2% to 3%), and 3% preferred not to answer the sex question (range 3% to 3%). Hispanic respondents comprised 23% of the total sample (range, 20% to 27%), 14% indicated non-Hispanic Black (range, 13% to 16%), 4% indicated non-Hispanic Asian (range, 3% to 5%), 49% indicated non-Hispanic White (range, 46% to 52%), 2% reported non-Hispanic Indigenous (either Native American, Native Hawaiian, or Other Pacific Islander, range 1% to 2%), 1% indicated non-Hispanic Middle Eastern (range 1% to 2%), and 6% reported a multi-racial identity, defined as marking 2 or more of the non-Hispanic categories (range, 6% to 7%). More than half of students reported one of their parents had a college degree (59%, range 57% to 61%), and each of the 4 years from 2021 to 2024 represents a quarter of the weighted sample. These descriptives differ little across grades (Supplemental eTable 2).

Table 1:

Sample Descriptives

	Three Grades Combined (n = 54 373a)
	Percentage	95% C.I.
Year
2021	25	(21–28)
2022	25	(22–28)
2023	25	(22–28)
2024	26	(22–29)
Race/Ethnicityb
Hispanic	23	(20–27)
Black	14	(13–16)
Asian	4	(3–5)
White	49	(46–52)
Indigenous Peoples	2	(1–2)
Middle-Eastern	1	(1–2)
Multi-racial	6	(6–7)
Sex
Male	48	(47–49)
Female	46	(46–47)
Other	02	(02–03)
Prefer not to answer	03	(03–03)
Census Region
South	38	(34–42)
Northeast	17	(14–19)
Midwest	21	(18–24)
West	24	(19–29)
Parent has college education	59	(57–61)

^aSample size is unweighted.

^bSee Supplemental eTable 1 for coding of this variable.

Note: Supplemental eTable 1 presents these results stratified by grade. All prevalence and 95% confidence intervals estimates use probability weights to take into account the complex sample design of the data.

Primary Results

Figure 1 and Supplemental eTable 2 present the grade-specific trendline for marijuana vaping in the past 12 months, which is the trendline in the middle of each graph. This trendline is juxtaposed with the trendline for past 12-month marijuana use by any mode (eg, vaping, smoking, edibles), which is the top trendline in each graph. In all grades, the distance between the 2 trendlines narrows over the study period, indicating that vaping became a more common mode of use among students who used marijuana. Also presented in the graph is the trendline for use of a flavored marijuana vaping solution, which is the bottom trendline. The distance between this trendline and the marijuana vaping trendline (the middle trendline) also narrows over the study period, indicating an increase in the percentage of student who used flavors when vaping marijuana.

Figure 1: Percentage of Students Who Used Marijuana in the Past 12 Months, by Mode of Use, Grade, and Year.

Notes: Vertical lines are 95% confidence intervals for year-specific observed values. Supplemental eTable 2 presents specific estimates and tests of statistical significance both across adjacent years and 4-year linear trends.

Figure 2 and Supplemental eTable 2 present grade-specific percentages by year in vaping as a mode of marijuana use among students who have used any marijuana in the past 12 months. This information quantifies the extent to which vaping has become an increasingly common mode for marijuana use over the study period, as indicated by the narrowing distance between the top and middle trendlines in Figure 1. Specifically, in 8^th grade, the proportion of students who used marijuana by vaping it increased from 48% to 57% (p<.05 for 4-year trend), in 10^th grade from 60% to 66% (p=.07), and in 12^th grade from 58% to 67% (p<.01 for 4-year trend).

Figure 2: Percentage of Students Who Vaped Marijuana Among Those Who Used Any Form of Marijuana in the Past 12 Months, by Grade and Year.

^aSample size is unweighted.

Notes: Vertical lines are 95% confidence intervals for the year-specific observed value. See Supplemental eTable 2 for specific estimates.

Figure 3 and Supplemental eTables 2 and 3 present grade-specific percentages by year in use of flavored marijuana solutions among students who had vaped marijuana in the past 12 months. This information quantifies the extent to which flavor use has become increasingly common from 2021–2024, as indicated by the narrowing distance between the bottom and middle trendlines in Figure 1. Specifically, the proportion of students who used flavors when vaping marijuana from 2021–2024 increased from 47% to 63% (p<.01) in 8^th grade, from 41% to 53% (p<.01) in 10^th grade, and from 36% to 50% (p<.01) in 12^th grade.

Figure 3: Percentage of Students Who Used Flavored Solutions Among Those Who Vaped Marijuana in the Past 12 Months, by Grade and Year.

^aSample size is unweighted.

Notes: Vertical lines are 95% confidence intervals for the year-specific observed value. See Supplemental eTables 2 and 3 for specific estimates.

Figure 3 and eTable 4 present additional information on prevalence levels by year for use of flavors among students who had vaped marijuana in the past 12 months. Use of flavors was highest among the youngest adolescents, with 63% of 8^th grade students using flavors among those who vaped marijuana in 2024 (see eTable 3), compared to 53% in 10^th grade and 50% in 12^th grade (P<.05 for comparison of 8^th grade with the older 2 grades). “Fruit” was the most commonly used flavor in all grades among those who vaped marijuana, at 40%, 33%, and 32% in 8^th, 10^th, and 12^th grade, respectively, in 2024 (eTable 4). “Sweet” was the second most common flavor, at about half the prevalence for fruit. Use of both fruit and sweet flavors significantly increased over the study period in all 3 grades among students who vaped in the past 12 months.

Table 2 presents tests of the trends in Figures 2 and 3 from regression models that control demographics. Model 1 focuses on the prevalence of marijuana vaping among students who used marijuana in the past 12 months. The relative risk of 1.05 (p<.01) for the Year variable indicates an average increase of 5% per year in the prevalence of vaping among past 12-month marijuana users. Black students who used marijuana were less likely to vape it than were White students, as indicated by a relative risk of 0.75 (p<.01). For Black students use of vaping among past 12-month marijuana users increased at a faster rate than non-Black students, as indicated by a positive, significant interaction term of Black and Year (not shown), but their levels of vaping were nonetheless lower than their non-Black peers by the final year of the study. Vaping among students who used marijuana in the past 12 months did not vary by Census region. Analysis restricted to the South Census region – a region with virtually no legalized marijuana markets – found a similar increase in vaping over time with a relative risk of 1.07 (95% C.I. 1.03–1.10) per year, which compares to 1.05 for the whole sample.

Table 2:

Results from Models Predicting Past 12-Month Marijuana Vaping as a Function of Demographics

	Marijuana Vaping, among Students Who Used Any Form of Marijuana		Use of Flavored Solutions, among Students Who Vaped Marijuana
	(n = 10 379a)		(n = 6 995a)
	RR	95% C. I.	RR	95% C. I.
Yearb	1.05**	(1.03 to 1.07)	1.09**	(1.06 to 1.13)
Grade
8th	0.88**	(0.81 to 0.95)	1.30**	(1.17 to 1.44)
10th	1.02	(0.98 to 1.07)	1.10*	(1.01 to 1.19)
12th (reference)
Race/Ethnicity
Hispanic	0.97	(0.92 to 1.02)	1.24**	(1.14 to 1.34)
Black	0.75**	(0.68 to 0.82)	1.18**	(1.08 to 1.30)
Asian	0.99	(0.86 to 1.14)	1.08	(0.83 to 1.40)
Indigenous Peoples	0.91	(0.79 to 1.04)	1.24*	(1.03 to 1.51)
Middle-Eastern	1.02	(0.80 to 1.29)	1.27	(0.93 to 1.74)
Multi-racial	0.96	(0.88 to 1.04)	0.96	(0.83 to 1.12)
White (reference)
Sex
Female	1.00	(0.96 to 1.05)	1.07	(0.99 to 1.16)
Other	0.92	(0.78 to 1.10)	1.06	(0.86 to 1.30)
Prefer not to answer	0.91	(0.78 to 1.07)	1.03	(0.87 to 1.21)
Male (reference)
Census Region
South (reference)
Northeast	1.00	(0.94 to 1.06)	0.93	(0.85 to 1.03)
Midwest	0.97	(0.92 to 1.03)	1.06	(0.98 to 1.15)
West	1.00	(0.94 to 1.07)	1.00	(0.90 to 1.11)
Parent has college education	0.97	(0.92 to 1.01)	0.93	(0.87 to 1.00)
Constant	0.64**	(0.59 to 0.68)	0.35**	(0.31 to 0.39)

^aSample size is unweighted.

^bCentered at year 2021

Note: Supplemental eTable 4 presents these results stratified by grade.

Model 2 of Table 2 focuses on the prevalence of flavored solutions use among students who vaped marijuana in the past 12 months. The relative risk of 1.09 (p<.01) for the year variable indicates a 9% increase per year in the use of flavors among students who vaped. Compared to White respondents, the use of flavors was 24% higher (p<.01) for Hispanic students, 18% higher for Black students (p<.01), and 24% higher for indigenous students (p<.05) collapsed across years. These differences by race/ethnicity did not trend, as indicated by nonsignificant multiplicative interactions with Year (not shown). Analysis restricted to the South Census region – a region with virtually no legalized marijuana markets – found a similar increase in use of flavored vaping solutions over time with a relative risk of 1.08 (95% C.I. 1.03–1.13) per year, which compares to the 1.09 for the whole sample.

eTable 5 presents these regressions stratified by grade.

DISCUSSION

Adolescent drug use since the pandemic has charted a historically unique path. After substantial declines in use of most drugs from 2020 to 2021—many of which were the largest ever recorded by MTF—the lowered levels persisted in the following years and in many cases dropped even further. This pattern applies to all 3 of the most commonly used substances of marijuana, alcohol, and nicotine vaping, as well as drugs with lower prevalence [11]. Consequently, trends in marijuana vaping and use of flavored marijuana solutions would be expected, a priori, to follow a similar pattern.

One of this study’s most striking findings is that the percentage of all US adolescents who vaped a flavored marijuana solution increased from 2021 to 2024, a rare case of a drug use outcome that moved against the tide. To put this finding in context, in 8^th grade no other drug significantly increased for past 12-month use among all the drugs tracked by MTF since 2021 (out of 30 monitored) [11]. In 10^th grade, the only drugs to significantly increase from 2021 to 2024 were steroids, creatine (used for body building) and snus (out of 30 drugs tracked) [11], and in 12^th grade, the only drug to significantly increase during this period was creatine (out of 39 drugs tracked) [11]. Specifically, the prevalence of flavored marijuana vaping use increased from 2% to 4% in 8^th grade, from 5% to 6% in 10^th grade, and from 7% to 9% in 12^th grade.

As a result, adolescents who used flavors to vape marijuana shifted from a minority to a majority from 2021 to 2024. Use of flavored solutions among past 12-month marijuana vapers grew from a proportion of 47% to 63% in 8^th grade, from 41% to 53% in 10^th grade, and from 36% to 50% in 12^th grade. Of note is that the youngest adolescents are most likely to use flavors when vaping marijuana, with 2024 levels significantly higher than those in 10^th grade and in 12^th grade.

With the increases in the percentage of adolescents who use flavorings, vaping continued on a path towards becoming the dominant mode of marijuana use among US adolescents. The proportion who vaped marijuana among past 12-month marijuana users increased an average of 5% every year over the study period. These proportions in 2024 were 57% in 8^th grade, 66% in 10^th grade, and 67% in 12^th grade.

Accompanying the increasing popularity of vaping as a mode of marijuana use are potential increases in THC doses delivered to adolescents and the associated health consequences. THC concentrations in vaping products can greatly exceed those available in marijuana flower, and these concentrations have been substantially increasing over time [12]. Adolescent use of marijuana products with highly concentrated THC is associated with increases in frequency of cannabis use, abuse and addiction, transition to use of other addictive drugs [13], and anxiety [14–15]. Use of high-dose THC also increases the risk for acute, adverse psychological reactions such as paranoia, panic attacks, and temporary psychotic episodes, and the number of calls for such acute reactions to US Poison Centers have increased in recent years [16]. The negative effects of THC are particularly adverse during adolescence, which is a critical period of brain maturation and development [17].

These findings draw attention to flavored marijuana solutions as a target for both research and policy. Discrete-choice experiments show that adolescents are more willing to try flavored marijuana solutions, which they view as less dangerous and with greater curiosity [7]. Largely unknown is whether policy and regulations can effectively counter the lure of flavors to adolescents. One model to consider is the Master Tobacco Settlement Agreement of 1998, which implemented a suite of policies to reduce teen cigarette smoking that included increased tobacco taxes, restrictions on cigarette advertising, substantial tobacco prevention efforts in several large states, as well as a nationwide antismoking ad campaign [18]. It is notable that following these measures past 30-day smoking among 12^th grade student decreased tenfold from 35% in 1998 to 3% in 2024 [8], and during this period cigarettes remain widely available throughout the U.S.

Limiting adolescent availability to marijuana vaping devices and flavored solution is another policy approach to reduce adolescent marijuana vaping. Historically, population-level efforts to restrict adolescents’ access to marijuana have had little success, given that the percentage of 12^th grade students who report that it would be “fairly easy” or “very easy” to obtain marijuana has been 70%+ in every year from 1975 to 2024 [8]. This high availability of marijuana may explain why adolescent use levels have not gone up as an increasing number of states have legalized its use [8], to the extent that adolescent accessibility may already be at or near a ceiling. Nonetheless, policies aimed at restricting adolescent access to marijuana vaping devices and solutions may be more effective, as these products rely on specialized manufacturing technologies that offer greater potential for tracking and interception.

The need to address adolescent use of vaping devices and flavored solutions for marijuana is a nationwide concern, and extends beyond the U.S. states that have legalized recreational marijuana use. The study results indicate that the Southern region experienced increases in adolescent marijuana vaping and the use of flavored solutions comparable to the national average. The South is a theoretically strategic focus because it has a very limited legalized marijuana market; Maryland is the only state in the region with an operational, state-sanctioned retail marijuana market, which launched in 2023 [19]. These results suggest that spread of marijuana vaping devices and solutions is taking place across the entire country, regardless of individual state marijuana laws.

Finally, it is worth noting that these trends in adolescent marijuana vaping of flavored solutions have developed slowly but steadily. Since 2021 no statistically significant increase took place in adjacent years. These results serve as a timely example of how factors that influence national adolescent drug use may not have immediate effect, and, instead, take many years to unfold. Such delayed effects can apply to new products, new technology (such as flavored vaping solutions), or new policies designed to reduce adolescent drug use.

Limitations

One limitation is that the MTF sample does not include adolescents who had dropped out of school by the time of the survey administration. The finding of similar patterns of results across 8^th, 10^th, and 12^th grade suggests that these dropouts do not account for the study’s main results, given that levels of school dropout in 8^th and 10^th grades are quite small [20].

A second limitation is that adolescent marijuana use is self-reported and thus potentially subject to self-report bias. However, any such biases would be expected to be steady over the limited, 4-year period of this study, and therefore cancel out in this study’s main focus on differences in prevalence levels across years.

Conclusions

The percentage of US adolescents who vaped marijuana and flavored marijuana solutions increased from 2021 to 2024, one of very few drug outcomes that have increased since the pandemic onset. With this increase, adolescents who used flavors to vape marijuana shifted from a minority to a majority, and the use of vaping as a mode of marijuana use increased. These results point to flavored marijuana vaping solutions as an increasingly important target for research and policy.

Supplementary Material

Supp eTable 1

Supplemental eTable 1: Sample Percentages and 95% Confidence Intervals by Grade

Supp eTable 4

Supplemental eTable 4: Regression Models for Table 2 Stratified by Grade

Supp eTable 2

Supplemental eTable 2: Detailed Estimates for Figures 1 – 3

Supp eTable3

Supplemental eTable 3: Percentage of Students Who Used Specific Flavors Among Those Who Vaped Marijuana in the Past 12 Months, by Grade and Year

Implications and Contribution.

This study examines recent trends in marijuana vaping among U.S. adolescents, particularly the use of flavored marijuana solutions. Substantial increases in adolescent use of flavored vaping solutions to consume marijuana point to the need for targeted policies, interventions, and educational campaigns to address and mitigate this growing concern.