Abstract
Background:
Little is known about factors that influence marijuana vaping among young people. We examined cigarette, e-cigarette and marijuana use experiences, social network characteristics and exposure to direct e-cigarette marketing as predictors of marijuana vaping initiation and escalation.
Methods:
One-year prospective data were collected between 2017 and 2019 from 2,327 young adults (Mean age = 21.2; SD = 2.1; 54% women) attending 2-year and 4-year colleges in Hawaii.
Results:
Among participants who were never marijuana users at baseline, being a dual user of cigarette and e-cigarette at baseline was the strongest predictor of marijuana vaping initiation, followed by baseline cigarette-only and e-cigarette-only use. Higher prevalence of regular marijuana users in one’s social networks, but not e-cigarette users or cigarette smokers, significantly predicted marijuana vaping initiation a year later. Among baseline current e-cigarette users and lifetime marijuana users, higher presence in social networks of individuals who frequented vape shops at baseline was a significant predictor of increased current marijuana vaping at one-year follow-up. In addition, greater presence in social networks of marijuana users and individuals more involved with vape shops predicted higher marijuana vaping.
Conclusions:
Dual use of cigarette and e-cigarette and greater presence in social networks of marijuana users and people who frequent vape shops appear to be robust predictors of marijuana vaping onset and escalation among young adults. In addition to promoting e-cigarette use prevention/cessation, efforts to control marijuana vaping may need to consider promoting smoking prevention/cessation and the effects of increasing prevalence of marijuana use.
Keywords: Marijuana vaping, e-cigarette, cigarette, young adults, social networks
1. Introduction
A majority of the recent cases of electronic cigarette (e-cigarette) or vaping product associated lung injury (EVALI) in the United States have been linked with consumption of e-liquids containing vitamin E-acetate-laced tetrahydrocannabinol (THC), the primary psychoactive constituent of marijuana (Perrine et al., 2019). Between 2017 and 2019, the prevalence rate of past-30-day marijuana vaping increased from 5% to 14% among U.S. 12th graders (Miech et al., 2019). Among young adult current e-cigarette users, the past-year prevalence of marijuana vaping is likely to be 18% (Trivers et al., 2018). Young adults (18–25 year olds) have been identified as a group most at risk for THC vaping (Navon et al., 2019). Currently, the factors that influence young adults’ marijuana vaping behavior are not well understood, which limits our knowledge of how to develop effective policies and interventions to control marijuana vaping and related health risks.
The social-ecological theory (Bronfenbrenner, 1977) provides a framework that may help understand the factors that work on multiple levels to influence a health risk behavior such as marijuana vaping. The theory’s basic postulate is that individual level factors influence a behavior within a broader interpersonal level, which in turn operates within a much broader community-societal domain as represented by physical, economic, cultural, and policy environments (Golden & Earp, 2012). Taking the social ecological framework as a guiding model, we examine the relative effects of intrapersonal, interpersonal, and community-societal-level risk factors on marijuana vaping progression among young adults.
The variables we examine within the intrapersonal domain include demographics (e.g., age, gender, ethnicity), cigarette and e-cigarette use history and characteristics, and sensation seeking tendencies, which is known to be a strong predictor of substance use and abuse (Stacy, Newcomb, & Bentler, 1993; Simon, Stacy, Sussman, & Dent, 1994). A previous study longitudinal study on predictors of marijuana vaping among young adults (Cassidy et al., 2018) found lifetime e-cigarette use to be a predictor of marijuana use initiation at a later time-point. Lifetime cigarette smoking, however, did not prospectively predict marijuana vaping. Current cigarette and e-cigarette use were not examined the study.
Research, in general, has linked e-cigarette use with marijuana use initiation and increase in future marijuana use frequency (Audrain-McGovern et al., 2018; Chadi et al., 2019; Dai et al., 2018). But most of this research has not distinguished between vaping marijuana and marijuana use through other means (e.g., smoking). Additionally, past research has rarely examined the prospective effects of current e-cigarette use on marijuana vaping after accounting for the effects of current cigarette smoking and dual use of e-cigarette and cigarette. High positive correlation has been consistently reported between combustible tobacco use, notably cigarette smoking, and marijuana use (Agrawal et al., 2012; Ramo et al., 2012). A majority of regular marijuana users also smoke cigarettes (Schauer et al., 2015; Schauer and Peters, 2018). Cigarette smoking is known to act as a “gateway” to marijuana use among adolescents (Agrawal et al., 2012; Kandel et al., 1992). Hence, there is clearly a need to understand how exclusive e-cigarette or cigarette use and dual use of cigarette and e-cigarette impact marijuana vaping behavior over time.
To our knowledge, no study has examined e-cigarette use characteristics as predictors of the increase in marijuana vaping at a later time-point. Nicotine content and e-cigarette device characteristics may be of particular interest because of their implications for e-cigarette regulations. Past research suggests that nicotine and THC may reinforce each other’s effects by acting synergistically in the neural reward pathways (Rabin and George, 2015), raising the possibility that nicotine use via e-cigarettes may promote marijuana vaping over time. Evidence suggests that among young adult co-users of marijuana and tobacco, sequential use (e.g., smoking cigarette after consuming marijuana) and co-administration (i.e., consuming both products at the same time; e.g., joint, blunt) are highly prevalent (e.g., Tucker et al., 2019). Although Tucker et al. (2019) did not find co-administration of marijuana and nicotine via e-cigarette to be a common practice in their sample, sequential use and/or co-administration nicotine and marijuana is plausible and deserve attention.
Furthermore, there are several different types of e-cigarette devices which may be used to vape marijuana, including closed and open systems. Marijuana is vaped by heating THC-containing juices and oils, dry herbs, or concentrates in cartridges and pods. Anecdotal evidence suggests that prefilled “hash oil” cartridges and pods may be obtained to fit “Cigalike” and pod-based devices, respectively. In addition, “hash oil” may be vaped directly or mixed with propylene glycol or vegetable glycerine, the solvents commonly used in regular e-liquid.
We examined social network characteristics as representatives of the interpersonal domain and experiences with direct marketing via digital means (e.g., e-mail, social media) as representatives of the community-societal domain. Social network characteristics, mainly peer use, has been strongly associated with substance use initiation and maintenance among youths and young adults (Hoffman et al., 2006; Andrews et al., 2002). The previous longitudinal study (Cassidy et al., 2018) provided a strong evidence showing that peer marijuana vaping is a predictor of marijuana vaping among young adults. The current study expands the assessment of social networks beyond peers. In addition to assessing e-cigarette, cigarette, and marijuana use characteristics of the network members, we assess the network members’ links with vape shops.
At present knowledge is lacking regarding how social network links with vape shops influence marijuana vaping among youths and young adults. Also, currently there is limited understanding as to whether young people’s exposure to vape shops’ or e-cigarette manufacturers’ attempts at direct marketing via social media and/or other means (e.g., e-mail, text messages) has any effect on their marijuana vaping behavior, beyond the influence of immediate social networks. Recent research shows that marketing of e-cigarette devices and liquids on social media is widespread (Allem et al., 2017; Allem et al., 2019; Laestadius et al., 2019). Furthermore, there is evidence suggesting that e-cigarette companies and vape shops employ social media and other channels for direct marketing among potential consumers (Chu et al., 2015; Dai and Hao, 2017). Given the purported link between the recent outbreak of EVALI and THC-containing e-liquids bought off the street (Moritz et al., 2019), examining the role of vape shops in influencing marijuana vaping may have important policy implications related to e-cigarette regulations.
Thus, to address the current gaps in the literature, this study examined one-year prospective predictors of marijuana vaping initiation and escalation in an ethnically diverse sample of young adults. Specifically, we examined demographics, e-cigarette use and cigarette smoking history and characteristics, social network characteristics, and exposure to direct marketing as predictors of marijuana vaping initiation among young adults who had never used marijuana, via vaping or otherwise, in their lifetime at baseline. Next, we tested the effects of the predictors on the levels of current marijuana vaping at follow-up, among baseline lifetime marijuana users and current e-cigarette users, adjusting for the baseline level of current marijuana vaping and baseline measures of demographic and other covariates. Among current e-cigarette users at baseline, along with other predictors of interest, we examined e-cigarette use characteristics, including device and nicotine concentration, as predictors of future marijuana vaping.
The current research is based on a sample from Hawaii. A brief background to marijuana-related policies in Hawaii is as follows. Hawaii was the first U.S. state to legalize medical marijuana use in 2000 (Act 228; State of Hawaii Department of Health, 2020).But access to dispensaries had been limited for years after the passage of Act 228. Another law passed in 2015 (Act 241) allowed for the dispensation of medical marijuana products. Currently, there are eight state-approved dispensaries across Hawaiian Islands (State of Hawaii Department of Health, 2020). Recreational marijuana is not legal in the state, although since 2019, the punishment for possessing small amounts (i.e., up to three grams) of marijuana has been drastically reduced (Rodriguez, 2019).
2. Materials and Methods
2.1. Procedures
Participants were recruited from two four-year universities and four two-year (community) colleges on Oahu, Hawaii. All institutions belonged to the same University system. E-mail addresses were obtained for all 18–25 year olds enrolled across campuses. Invitations to participate in the screener survey were sent to randomly selected e-mail addresses. The link to the screener survey was accompanied by an invitation text which described the study in generic terms, as a study on marketing and young adult health behavior. The screener survey asked questions about age, sex, tobacco, alcohol, and dietary behaviors. The screener survey also asked students for their contact information, including phone number.
Students were given on average two weeks of time to respond to the screener survey and provided up to three reminders. Approximately 60% of those who were sent the invitation completed the screener survey. This response rate is similar to or higher than rates reported for other studies (Loukas et al., 2019; Sutfin et al., 2015) recruiting young adults from colleges through e-mail invitations. However, those who responded to the e-mail invitations were predominantly women who had never smoked a cigarette. In order to obtain a sample that was more gender balanced as well as included proportions of cigarette smokers and experimenters comparable to or higher than the national distribution of smokers in the 18–25 years age group, we supplemented the e-mail recruitment with classroom-based recruitment. We randomly selected on average 40 classes from each participating campus and presented the study in those classrooms. Students approached in the classrooms completed the paper-and-pencil version of the screener survey. The average response rate across classrooms was 80%.
Across the e-mail and classroom methods, total 3664 were screened to be eligible to participate in the study and were invited to participate in the study. Participants provided written consent to participate in the longitudinal study and were sent a unique link to the baseline survey via e-mail. A total of 2622 participants completed the baseline survey. The baseline participants were contacted 6 and 12 months later to complete six-month and one-year follow-up surveys. The one-year follow-rate was 89%: 2327 participants completed the one-year follow-up survey. Both the baseline and follow-up surveys were programmed on Inquisit (2015).
2.2. Measures
2.2.1. Demographics:
Participants self-reported age, ethnicity, sex, and parental income. Ethnicity was determined based on two items. To determine ethnicity, participants were asked “What is ethnic background?” and were provided with a list of ethnicities common in Hawaii and the U.S. The question was asked in two different ways. The first question asked participants to refer to the list and “check all that apply.” The second question asked participants to choose the ethnic background that they identify with most. The response to the second question was utilized to assign mixed-ethnicity individuals into a particular ethnic category.
2.2.2. Social network characteristics.
Social network characteristics were assessed using the egocentric method (Burt, 1984). Participants were asked to nominate up to 5 individuals who they spend the most time with or talk to most often. Next, they were asked a number of questions on each of the individuals they nominated. Five such questions were relevant to the current investigation: cigarette smoking (e.g., “Does this person smoke cigarettes?” Response options: “No, not at all,” “Yes, sometimes,” and “Yes, regularly”); e-cigarette use; marijuana use; whether the person frequents vape shops (“Does this person go to vape shops?” Response options: “No, not at all,” “Yes, sometimes,” and “Yes, regularly”); and whether the person knows someone who works at a vape shop. The prevalence of a characteristic of interest was determined by summing up responses across the nominations.
2.2.3. Direct marketing.
To assess exposure to e-cigarette direct marketing, participants were asked, “Has an e-cigarette company, including a vape shop, ever sent you information through…”, and were provided with the following options: “The mail,” “e-mail,” “text messages,” and “social media (e.g., Facebook, Instagram).” Exposure through social media was dealt with separately in analyses.
2.2.4. Cigarette smoking.
Lifetime cigarette smoking was assessed with a single item: “How many cigarettes have you smoked in your entire life?” (“None, I have never smoked a cigarette,” “1–100,” “More than 100 cigarettes”). For analysis, the variable was dichotomized into “ever use” (1) and “never use” (0). Current cigarette smoking was assessed in terms of past-30-day smoking: “During the last 30 days (1 month), on how many days did you smoke a cigarette?” (8-point scale: “0 days,” “1–2 days,” “3–5 days,”…, “All days”). Participants were also asked: “How do you describe your current cigarette smoking behavior?” (“I don’t smoke,” “I smoke sometimes,” “I smoke regularly”).
2.2.5. E-cigarette use, nicotine concentration, dependence, and device characteristics.
Lifetime e-cigarette use was measured with a single item: “Have you ever used an electronic cigarette (e-cigarette) or a similar vaping device, even if it was just a puff?” (Yes, No). Current e-cigarette use was assessed in terms of past-30-day e-cigarette use: “During the last 30 days (1 month), on how many days did you use an electronic cigarette (e-cigarette) or a similar vaping device?” (8-point scale: “0 days,” “1–2 days,” “3–5 days,”…, “All days”). In addition, we asked participants: “How often, if at all, do you currently use an e-cigarette?”(Response options: “Daily”, “Less than daily, but at least once a week”, “Less than weekly, but at least once a month”, “Less than weekly, but at least once a month”, “Less than monthly”, “Not at all”) (Pearson et al., 2018). Current e-cigarette users self-reported their preferred nicotine concentration by selecting from a list of options or by writing in the provided space. E-cigarette dependence was assessed with a validated 10-item scale (Foulds et al., 2015). E-cigarette users’ preference for each type of e-cigarette device was assessed by providing participants with example images of the devices and their corresponding popular names. They were then asked how often they used a particular device: “I have never used this kind,” “I used to but not anymore,” “Rarely,” “Sometimes,” and “Always.” Devices assessed included “Cigalike,” “Ego-style tank systems,” “Mods,” and “JUUl or similar pod-based devices.” The pod-based devices were not assessed in baseline.
2.2.6. Marijuana use.
Lifetime marijuana use and lifetime marijuana vaping were separately assessed using language adapted from the question that assessed lifetime e-cigarette use. Similarly, past 30-day cigarette smoking questions were adapted to assess current marijuana use and marijuana vaping.
2.3. Data analysis
Data were analyzed using the SAS software, Version 9.4 of the SAS System for Windows (2017).
2.3.1. Analysis among lifetime never marijuana users at baseline.
The first set (Set 1) of analysis involved examining baseline predictors of marijuana vaping initiation at one-year follow-up among baseline never marijuana users, by running multiple logistic regression models. Baseline never marijuana users included individuals who had never used marijuana in their lifetime at baseline, by vaping or otherwise. Initiation of marijuana vaping was operationalized as lifetime use reported at six-month follow-up or one-year follow-up.
2.3.2. Analysis among lifetime marijuana users at baseline.
The second set (Set 2) of analysis involved examining baseline predictors of the frequency of past-30-day marijuana vaping at one-year follow-up among lifetime marijuana users at baseline, by running zero-inflated negative binomial regression models. The models adjusted for the effects of past-30-day marijuana vaping frequency at baseline. Lifetime marijuana users at baseline included those who had used marijuana at least once in their lifetime at baseline, by vaping or otherwise.
2.3.3. Analysis among current e-cigarette users at baseline.
The third set (Set 3) of analysis involved examining the baseline predictors of past-30-day marijuana vaping frequency at one-year follow-up among current e-cigarette users at baseline, by running zero-inflated negative binomial regression models. The models accounted for the effects of past-30-day marijuana vaping frequency at baseline.
For each set of analyses, predictors were tested in the following manner:
2.3.4. Intrapersonal-level predictors:
For all sets (1, 2, and 3) the following demographic variables were tested simultaneously as predictors of marijuana vaping initiation one-year later, along with baseline current cigarette and e-cigarette use and where appropriate, marijuana vaping characteristics: age, gender, ethnicity, family/household income, college type, and sensation seeking. For sets 1 and 2, baseline cigarette and e-cigarette use characteristics were tested in terms of current e-cigarette-only, cigarette-only, and dual use (all dummy-coded with reference to current non-use of either or both products). Set 3 included baseline past-30-day marijuana vaping frequency and past-30-day cigarette and e-cigarette use frequencies, along with baseline measures for nicotine concentration, e-cigarette device type, and e-cigarette dependence. Because use JUUL or similar pod-based devices were not assessed in baseline, we were unable to test the prospective effects of pod-based devices on marijuana vaping. However, use of pod-based devices was assessed at one-year follow-up. Hence, the association between use of pod-based devices and marijuana vaping was examined cross-sectionally.
2.3.5. Interpersonal-level predictors:
For all sets (1, 2, and 3), we tested each social network variable separately as a predictor, adjusting for all intrapersonal variables relevant to each set of analyses discussed above. The social network variables included presence of cigarette, e-cigarette, and marijuana users in participants’ social networks and number of individuals in participants’ social networks who visited vape shops or knew someone who worked at a vape shop.
2.3.6. Community-societal-level predictors:
For all sets (1, 2, and 3), we tested each direct marketing variable separately as a predictor, adjusting for all intrapersonal and interpersonal variables relevant to each set of analyses discussed above. The direct marketing variables included exposure to direct marketing via e-mail, mail, and text messaging, and exposure to direct marketing via social media (i.e., Facebook, Instagram, & Twitter).
3. Results
3.1. Participants
Table 1 shows the baseline demographic and tobacco product (cigarette smoking, e-cigarette) and marijuana use characteristics of the entire sample as well as the following subgroups: lifetime never marijuana users, lifetime marijuana users, and current e-cigarette users. Note that baseline lifetime never and ever marijuana users were distinct, non-overlaping subsamples and included individuals who had never and ever used marijuana, by vaping or any other means. Never and ever marijuana users included 11% and 34% current e-cigarette users at baseline, respectively. Conversely, current e-cigarette users at baseline included 75% lifetime marijuana users. Between baseline and one-year follow-up among baseline never marijuana users, 4.2% (n = 44) reported marijuana vaping initiation and 17% (n = 185) reported marijuana use initiation in general. In addition, among lifetime marijuana users at baseline, past-30-day marijuana vaping frequency (i.e., non-use vs use), increased from 7% (n = 88) at baseline to 9% (n = 112) at one-year follow-up. The mean level of past-30-day marijuana vaping frequency also increased by 2% [i.e., mean = 0.10 (SD = 1.3), p < 0.01)].
Table 1.
Characteristics of the baseline sample
| All (N = 2622) | Lifetime marijuana non-users (n = 1162) | Lifetime marijuana users (n = 1434) | Current e-cigarette users (n = 812) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Mean (SD) | % | Mean (SD) | % | Mean (SD) | % | Mean (SD) | % | ||
| Age | 21.2 (2.2) | 20.8 (2.1) | 21.5 (2.2) | 21.2 (2.2) | |||||
| Gender | |||||||||
| Men | 46 | 51 | 41 | 53 | |||||
| Women | 54 | 49 | 59 | 47 | |||||
| Ethnicity | |||||||||
| White | 25 | 18 | 30 | 23 | |||||
| Asian | 26 | 32 | 21 | 23 | |||||
| Filipino | 18 | 21 | 15 | 19 | |||||
| NHPI | 21 | 20 | 22 | 26 | |||||
| Other | 11 | 9 | 12 | 9 | |||||
| Family Income | |||||||||
| 0–$39 K | 23 | 25 | 21 | 22 | |||||
| $40 K – $79 K | 33 | 35 | 32 | 35 | |||||
| $80 K – $119 K | 26 | 25 | 25 | 24 | |||||
| $120 K – $159 K | 10 | 9 | 10 | 10 | |||||
| Over $160 K | 9 | 6 | 11 | 10 | |||||
| Cigarette smoking | |||||||||
| Never | 47 | 75 | 26 | 19 | |||||
| Former/Experimenter | 31 | 16 | 43 | 37 | |||||
| Current | 22 | 9 | 31 | 44 | |||||
| E-cigarette use | |||||||||
| Never | 35 | 61 | 13 | ||||||
| Experimenter | 34 | 28 | 53 | ||||||
| Current | 31 | 11 | 34 | ||||||
| Lifetime marijuana use | |||||||||
| Has not used | 54 | 0 | 25 | ||||||
| Has used but not vaped | 38 | 72 | 25 | ||||||
| Has vaped | 16 | 28 | 50 | ||||||
| Past-30-day marijuana use | |||||||||
| Has used but not vaped | 20 | 36 | 25 | ||||||
| Has vaped | 7 | 13 | 23 | ||||||
| Presence in social networksa | |||||||||
| Cigarette smokers | 1.9 (1.1) | 1.6 (0.9) | 2.2 (1.1) | 2.4 (1.2) | |||||
| E-cigarette users | 1.9 (1.1) | 1.6 (1.0) | 2.0 (1.2) | 2.6 (1.3) | |||||
| Marijuana users | 2.3 (1.3) | 1.7 (1.0) | 2.8 (1.3) | 2.9 (1.3) | |||||
| Those who visit vape shops often | 1.7 (1.1) | 1.6 (0.9) | 1.8 (1.1) | 2.2 (1.2) | |||||
| Those who Know someone who works at a vape shop | 2.3 (1.4) | 2.2 (1.4) | 2.4 (1.4) | 2.5 (1.5) | |||||
| Ever exposed to direct marketing | |||||||||
| Mail, e-mail, text messages | 25.5 | 9.3 | 20.3 | 30.2 | |||||
| Social media | 9.2 | 5.4 | 12.1 | 15.3 | |||||
Notes. SD = Standard Deviation; NHPI: Native Hawaiian/Other Pacific Islader;
Range = 1–5
We compared the baseline sample (N = 2622) with the longitudinal sample (N = 2327) on the key variables considered in the current study and found no statistically significant differences between the two samples. This was consistent across lifetime marijuana never users at baseline (longitudinal n = 1057), lifetime marijuana users at baseline (longitudinal n = 1249), and current e-cigarette users at baseline (longitudinal n = 686)
3.2. One-year predictors of marijuana vaping initiation among baseline never marijuana users
Table 2 shows the results of the regression analysis examining the predictors of marijuana vaping initiation one year later. Demographic variables were not associated with marijuana vaping initiation. Adjusting for demographic variables, higher sensation seeking at baseline predicted higher likelihood of marijuana vaping initiation one year later. Adjusting for sensation seeking and demographic variables, relative to non-use, current e-cigarette-only use, cigarette-only use, and dual use at baseline predicted higher likelihood of marijuana vaping initiation one year later. For example, being a dual user at baseline was associated with 13 times higher likelihood of initiating marijuana vaping one year later. Being a cigarette-only smoker and e-cigarette-only user increased the odds of marijuana vaping by approximately 6 and 5 times, respectively.
Table 2.
Predictors of marijuana vaping initiation at one-year follow-up among baseline marijuana naïve participants (n = 1162)
| Marijuana vaping at 1-year follow-up (Dependent variable) | ||
|---|---|---|
| Baseline predictors (independent variables) | Odds Ratio (95% Confidence Interval) | |
| Demographics | ||
| Age | 0.89 (0.76, 1.04) | |
| Gender | Men | 1 |
| Women | 0.72 (0.36,1.43) | |
| Ethnicity | ||
| White | 1 | |
| Filipino | 0.62 (0.20, 1.93) | |
| NHPI | 1.02 (0.36,2.87) | |
| Asian | 1.41 (0.55,3.60) | |
| Other | 1.28 (0.35,4.63) | |
| College type | ||
| 4-year | 1 | |
| 2-year | 0.93 (0.45, 1.91) | |
| Family income | 1.04 (0.89,1.20) | |
| Sensation seeking | 1.06 (1.01, 1.12)*** | |
| Baseline current tobacco product use | ||
| No use | 1 | |
| E-cigarette only use | 4.63 (1.98, 10.8)*** | |
| Cigarette only use | 5.93 (1.79,19.7)** | |
| Dual use | 13.4 (5.37,33.3)*** | |
| Baseline social network characteristics | ||
| No. of e-cigarette users | 1.12 (0.94, 1.57) | |
| No. of cigarette smokers | 1.01 (0.76, 1.36) | |
| No. of marijuana users | 1.32 (1.03,1.67)* | |
| No. of individuals who frequent vape shops | 1.06 (0.74,1.51) | |
| No. of individuals who know someone who works at a vape shop | 1.09 (0.89, 1.34) | |
| Direct marketing | ||
| Exposure through mail, e-mail, text messaging | 1.08 (0.52,2.24) | |
| Exposure through social media | 1.20 (0.58,2.47) |
p < 0.05;
p < 0.01;
p < 0.001
Among interpersonal variables, after adjusting for all intrapersonal variables, only the prevalence of marijuana users in one’s social network was found to be a significant predictor of marijuana vaping initiation at follow-up. Having one additional regular marijuana user in one’s network increased the likelihood of initiating marijuana vaping by 32%. Exposure to direct e-cigarette marketing at baseline was not found to be associated with increased likelihood of marijuana vaping initiation.
3.3. One-year predictors of current marijuana vaping among baseline lifetime marijuana users
Table 3 shows the results of the analysis examining the associations between potential baseline predictors of current marijuana vaping frequency (i.e., past-30-day use frequency) one year later among lifetime marijuana users at baseline, adjusting for baseline past-30-day marijuana vaping frequency. Among intrapersonal variables, current cigarette-only, e-cigarette-only, and dual use status at baseline were significantly associated with increased frequency of past-30-day marijuana vaping at one-year follow-up. Relative to being a current tobacco product non-user at baseline, being a cigarette-only smoker, an e-cigarette-only user, or a dual user was associated with 3.56, 3.29, or 4.22 (i.e., exponentials of Table 3 parameter estimates) times increase in the frequency of past-30-day marijuana vaping one-year later. Adjusting for baseline marijuana vaping and intrapersonal variables, we found that greater presence in social networks of cigarette smokers, marijuana users, and individuals who frequented vape shops at baseline were significant predictors of increased current marijuana vaping at one-year follow-up. Greater presence of e-cigarette users in social networks at baseline did not predict higher current marijuana vaping at follow-up. We did not find associations between exposure to direct e-cigarette marketing at baseline and marijuana vaping at follow-up.
Table 3.
Predictors of past-30-day marijuana vaping at one-year follow-up among baseline ever users of marijuana (n = 1249)
| Parameter estimate (Standard Error) | 95% Confidence Interval | ||
|---|---|---|---|
| Baseline predictors (independent variables) | |||
| Demographics | |||
| Age | −0.02 (0.04) | −0.10 – 0.07 | |
| Female Gender | 0.04 (0.18) | −0.31 – 0.39 | |
| Ethnicity | |||
| Asian | −0.15 (0.24) | −0.62 – 0.33 | |
| Filipino | −0.56 (0.28)* | −1.11 – −0.008 | |
| NHPI | −0.14 (0.25) | −0.63 – 0.35 | |
| Other | 0.20 (0.27) | −0.33 – 0.72 | |
| Two vs. four year college | −0.007 (0.18) | −0.37 – 0.35 | |
| Family income | 0.02 (0.04) | −0.06 – 0.09 | |
| Sensation seeking | 0.008 (0.01) | −0.02 – 0.03 | |
| Past-30-day marijuana vaping | 0.19 (0.07)** | 0.04 – 0.33 | |
| Current tobacco product use | |||
| E-cigarette only | 1.19 (0.23)*** | 0.75 – 1.65 | |
| Cigarette only | 1.27 (0.32)*** | 0.65 – 1.89 | |
| Dual use | 1.44 (0.26)*** | 0.94 – 1.95 | |
| Baseline social network characteristics | |||
| No. of e-cigarette users | 0.06 (0.04) | −0.01 – 0.12 | |
| No. of cigarette smokers | 0.07 (0.03)* | 0.001 – 0.15 | |
| No. of marijuana users | 0.06 (0.03)* | 0.002 – 0.12 | |
| No. of individuals who frequent vape shops | 0.08 (0.03)** | 0.02 – 0.14 | |
| No. of individuals who know someone who works at a vape shop | 0.04 (0.03) | −0.01 – 0.09 | |
| Direct marketing | |||
| Exposure through mail, e-mail, text messaging | 0.11 (0.14) | −0.17 – 0.38 | |
| Exposure through social media | 0.21 (0.15) | −0.08 – 0.51 |
p < 0.05;
p < 0.01;
p < 0.001
3.4. One-year predictors of increased marijuana vaping among baseline current e-cigarette users
Among baseline current e-cigarette users, after adjusting for frequency of baseline marijuana vaping, we found that higher baseline past-30-day cigarette smoking frequency, but not past-30-day e-cigarette use frequency, was significantly associated with increased past-30-day marijuana vaping frequency at one-year follow-up. One unit increase in baseline past-30-day cigarette smoking frequency was associated with 1.09 times, or 9%, increase in past-30-day marijuana vaping frequency one year later. After adjusting for demographics and baseline past-30-day marijuana vaping, cigarette smoking, and e-cigarette use frequencies, none of the e-cigarette use characteristics, including device type, nicotine concentration, and e-cigarette dependence, was associated with marijuana vaping at one-year follow-up. This applied to the cross-sectional analyses on the one-year follow-up data as well, in which we examined pod-based devices. After adjusting for the baseline intrapersonal variables and the e-cigarette use characteristics, only the following baseline social network variables were significantly associated with follow-up marijuana vaping frequency: greater presence in social networks of individuals who frequented vape shops and who knew someone who worked at vape-shops. Exposure to direct e-cigarette marketing at baseline was not associated with follow-up current marijuana vaping behavior.
4. Discussion
This is one of the first studies to examine the prospective predictors of marijuana vaping among young adults across several domains and in subgroups defined by baseline marijuana use and e-cigarette use behavior. The investigation resulted in findings which are likely to enhance our understanding of the etiology of marijuana vaping and possibly have implications for marijuana vaping prevention among young people and development of regulations on e-cigarettes.
As far as effects of intrapersonal predictors are concerned, we found limited effects of demographic variables, including age and sex, on marijuana vaping initiation and escalation. This was somewhat surprising given that previous studies have identified younger age as a potential risk factor for marijuana vaping initiation (Daniulaityte et al., 2017; Navon et al., 2019). We found e-cigarette, cigarette, and dual use of both, to be significant predictors of marijuana vaping initiation among baseline marijuana never users, and increase in marijuana vaping among lifetime marijuana users at baseline. Dual use, in particular, appears to have the strongest effect on marijuana vaping initiation and escalation, followed by cigarette smoking, and then e-cigarette use. These findings are mostly consistent with past studies, including the previous longitudinal study on marijuana vaping initiation among young adults (Cassidy et al., 2019) and a prospective study with adolescents (Audrian-McGovern et al., 2018). The current findings build on the previous studies’ findings by showing that current cigarette smoking and dual use may have stronger, additive effects on marijuana vaping initiation and increase vis-à-vis current e-cigarette use status. Among baseline e-cigarette users, we did not find e-cigarette use characteristics such as dependence and preferred device type on increased marijuana vaping one year later. This may indicate that nicotine and marijuana vaping characteristics may be independent among young adults.
Among interpersonal predictors, we found that greater presence of marijuana users in social networks, but not the presence of cigarette smokers or e-cigarette users, was significantly associated with marijuana vaping initiation. This is consistent with the previous longitudinal study’s (Cassidy et al., 2018) finding that individuals who are more likely to have peers who initiate marijuana vaping are more likely to initiate marijuana vaping themselves. The findings regarding the lack of associations between the greater presence of cigarette smokers and e-cigarette users in one’s social network and marijuana use initiation are unique to the current study. These findings indicate that the influence of marijuana users in one’s social network is more likely to result in the uptake of marijuana vaping than that of e-cigarette users or cigarette smokers. Surprisingly, we did not find any association between links with vape shops and marijuana vaping initiation. Future studies may need to explore this lack of association in more detail.
We found that greater presence of cigarette smokers in social networks, along with greater presence of marijuana users, and presence of individuals with links to vape shops, were predictive of increase in marijuana vaping among baseline lifetime marijuana users. These findings may suggest that different types of interpersonal mechanisms may be at work in the etiology of marijuana vaping initiation versus maintenance or escalation. Among marijuana never users, greater affiliation with marijuana users may play a singularly important role in influencing marijuana vaping onset. Whereas being part of a network that actively engages in cigarette smoking, marijuana use, and vape shop cultures may be important in sustaining and escalating marijuana vaping behavior. Clearly, more research is needed to understand the roles of vape shops in influencing marijuana vaping behavior. The current research provides some evidence suggesting that vape shops may play a role in promoting marijuana vaping. This may in turn have regulatory significance and deserves attention from regulatory bodies.
Among baseline e-cigarette users, after adjusting for the effects of baseline marijuana vaping, we found that higher past-30-day cigarette use frequency at baseline was predictive of higher past-30-day marijuana vaping frequency one year later. The same kind of association was not found between baseline past-30-day e-cigarette use frequency and past-30-day marijuana vaping frequency one year later. These findings indicate that among current e-cigarette users, those who smoked cigarettes more frequently (i.e. heavier dual users) are more likely to vape marijuana one year later, whereas higher frequency of e-cigarette use in itself may not be associated with higher marijuana vaping. Further research is needed to test whether cigarette smoking potentiates the effects of marijuana vaping more than e-cigarette use. In general, it appears that efforts to counter marijuana vaping may need to consider targeting different types of social influence and the role of cigarette smoking.
We did not find any prospective effect of exposure to direct e-cigarette marketing on marijuana vaping initiation or escalation. It is likely that the content of direct marketing is circumspect and does not promote marijuana vaping. Research, however, is needed to better understand the content of direct e-cigarette marketing and whether or not it incorporates messages on marijuana vaping. There are limitations to this research. Our participants were young adults; hence, the findings may not completely generalize to adolescents. The rate of marijuana vaping onset in the current sample was relatively low; because of which some of the confidence intervals of our findings are large. In addition, our assessments of marijuana vaping were not detailed enough to elicit different modes of marijuana vaping such as heating wax or oil versus dabbing or mixing concentrates in regular e-liquid. Because we did not assess the preference for pod-based devices at baseline, we were not able to prospectively test the associations between the preference for pod-based devices and marijuana vaping. We did not assess or test marijuana marketing as a predictor of marijuana vaping. Our measure of direct marketing was limited to e-cigarettes. Lastly, we did not assess vaping of marijuana for medical purposes. Because this research is based on a sample from Hawaii, where only medical marijuana is legal, assessment of marijuana vaping for medical marijuana might have enhanced the understanding the current findings.
5. Conclusions
Despite the limitations, this study provided new information on the predictors of marijuana vaping initiation and increase among young adults. Among young adults who have never used marijuana, being an exclusive e-cigarette or cigarette user or a dual user of cigarette and e-cigarette significantly increases the risk of marijuana vaping initiation. In addition, interacting closely with marijuana users increases the risk of marijuana vaping initiation. Among lifetime marijuana users, cigarette smoking, e-cigarette use, and dual use are likely to be independent predictors of increased marijuana vaping over time. Among baseline lifetime marijuana users and current e-cigarette users, higher cigarette smoking frequency, greater presence of marijuana users in social networks, and greater presence of individuals in social networks who interact with vape shops, appear to be important predictors of marijuana vaping. The current findings imply that, in addition to targeting e-cigarette use behavior, prevention efforts designed to curb marijuana vaping should continue to target cigarette smoking and marijuana use behaviors. Future research and practice may benefit from gaining a better understanding regarding the social influence mechanisms that promote marijuana vaping initiation versus escalation.
Table 4.
Predictors of marijuana vaping at one-year follow-up among baseline current e-cigarette users (n = 686)
| Parameter Estimate (Standard Error) | 95% Confidence Interval | ||
|---|---|---|---|
| Demographics | |||
| Age | −0.05 (0.04) | −0.12 – 0.03 | |
| Female gender | 0.02 (0.16) | −0.29 – 0.34 | |
| Ethnicity | |||
| Asian | −0.14 (0.21) | −0.56 – 0.28 | |
| Filipino | −0.50 (0.23)* | −0.95 – −0.05 | |
| NHPI | −0.16 (0.22) | −0.58 – 0.25 | |
| Other | 0.30 (0.22) | −0.13 – 0.72 | |
| Two vs. four year college | 0.19 (0.16) | −0.12 – 0.51 | |
| Family income | −0.02 (0.03) | −0.08 – 0.05 | |
| Sensation seeking | −0.004 (0.01) | −0.03 – 0.02 | |
| Past-30-day marijuana vaping frequency at baseline | 0.14 (0.04)*** | 0.06 – 0.22 | |
| Baseline tobacco product and marijuana use frequencies | |||
| Past-30-day cigarette use | 0.09 (0.04)** | 0.02 – 0.17 | |
| Past-30-day e-cigarette use | 0.02 (0.03) | −0.04 – 0.09 | |
| Baseline e-cigarette use characteristics | |||
| Nicotine concentration | 0.12 (0.07) | −0.02 – 0.25 | |
| E-cigarette use dependence | 0.02 (0.02) | −0.03 – 0.06 | |
| Device type | |||
| Cigalike | −0.04 (0.07) | −0.17 – 0.09 | |
| Ego-style tank systems | 0.05 (0.07) | −0.09 – 0.20 | |
| Mods | −0.04 (0.07) | −0.17 – 0.09 | |
| Baseline social network characteristics | |||
| No. of e-cigarette users | 0.04 (0.03) | −0.02 – 0.10 | |
| No. of cigarette smokers | 0.02 (0.03) | −0.04 – 0.08 | |
| No. of marijuana users | 0.08 (0.07) | −0.05 – 0.21 | |
| No. of individuals who frequent vape shops | 0.05 (0.02)* | 0.0014 – 0.01 | |
| No. of individuals who know someone who works at a vape shop | 0.05 (0.02)* | 0.008 – 0.10 | |
| Direct marketing | |||
| Exposure through mail, e-mail, text messaging | −0.07 (0.11) | −0.28 – 0.15 | |
| Exposure through social media | −0.07 (0.13) | −0.32 – 0.18 |
p < 0.05;
p < 0.01;
p < 0.001
Highlights.
Study addressed lack of knowledge about predictors of marijuana vaping among young adults
Dual use of cigarette and e-cigarette is a strong predictor of marijuana vaping onset and escalation
Greater presence of marijuana users in social networks predicts marijuana vaping onset and escalation
Current e-cigarette users who have friends/family that frequent vape shops or know someone who works at a vapeshop are susceptible to increased future marijuana vaping
Funding Source
This work was supported by the National Institutes of Health [grant numbers R01CA202277, R01CA228905].
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of interest
No conflict declared.
References
- Agrawal A, Budney AJ, Lynskey MT, 2012. The co-occurring use and misuse of cannabis and tobacco: a review. Addiction 107:1221–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Allem JP, Ferrara E, Uppu SP, Cruz TB, Unger JB, 2017. E-Cigarette Surveillance With Social Media Data: Social Bots, Emerging Topics, and Trends. JMIR Public Health Surveill 3:e98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Allem JP, Majmundar A, Dharmapuri L, Cruz TB, Unger JB, 2019. E-liquid-related posts to Twitter in 2018: Thematic analysis. Addict Behav Rep 10:100196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Andrews JA, Tildesley E, Hops H, Li F, 2002. The influence of peers on young adult substance use. Health Psychol 21:349–357. [DOI] [PubMed] [Google Scholar]
- Audrain-McGovern J, Stone MD, Barrington-Trimis J, Unger JB, Leventhal AM, 2018. Adolescent E-Cigarette, Hookah, and Conventional Cigarette Use and Subsequent Marijuana Use. Pediatrics 142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bronfenbrenner U, 1977. Toward an experimental ecology of human development. Am Psychol 32: 513–531 [Google Scholar]
- Burt RS, 1984. Network items and the general social survey. Social Networks 6:293–339. [Google Scholar]
- Chadi N, Schroeder R, Jensen JW, Levy S, 2019. Association Between Electronic Cigarette Use and Marijuana Use Among Adolescents and Young Adults: A Systematic Review and Meta-analysis. JAMA Pediatr:e192574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chu KH, Sidhu AK, Valente TW, 2015. Electronic Cigarette Marketing Online: a Multi Site, Multi-Product Comparison. JMIR Public Health Surveill 1:e11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dai H, Catley D, Richter KP, Goggin K, Ellerbeck EF, 2018. Electronic Cigarettes and Future Marijuana Use: A Longitudinal Study. Pediatrics 141. [DOI] [PubMed] [Google Scholar]
- Dai H, Hao J, 2017. Direct Marketing Promotion and Electronic Cigarette Use Among US Adults, National Adult Tobacco Survey, 2013–2014. Prev Chronic Dis 14:E84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Daniulaityte R, Lamy FR, Barratt M, Nahhas RW, Martins SS, Boyer EW, Sheth A, Carlson RG, 2017. Characterizing marijuana concentrate users: A web-based survey. Drug Alcohol Depend 178:399–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Foulds J, Veldheer S, Yingst J, Hrabovsky S, Wilson SJ, Nichols TT, Eissenberg T, 2015. Development of a questionnaire for assessing dependence on electronic cigarettes among a large sample of ex-smoking E-cigarette users. Nicotine Tob Res 17:186–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hoffman BR, Sussman S, Unger JB, Valente TW, 2006. Peer influences on adolescent cigarette smoking: a theoretical review of the literature. Subst Use Misuse 41:103–55. [DOI] [PubMed] [Google Scholar]
- Golden SD, Earp JA, 2012. Social ecological approaches to individuals and their contexts: twenty years of health education & behavior health promotion interventions. Health Educ Behav 39:364–72. [DOI] [PubMed] [Google Scholar]
- Inquisit, 4 ed., 2015. Millisecond, Inc., Seattle, WA. [Google Scholar]
- Kandel DB, Yamaguchi K, Chen K, 1992. Stages of progression in drug involvement from adolescence to adulthood: further evidence for the gateway theory. J Stud Alcohol 53:447–57. [DOI] [PubMed] [Google Scholar]
- Laestadius LI, Wahl MM, Pokhrel P, Cho YI, 2019. From Apple to Werewolf: A content analysis of marketing for e-liquids on Instagram. Addict Behav 91:119–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Loukas A, Marti CN, Perry CL, 2019. Trajectories of Tobacco and Nicotine Use Across Young Adulthood, Texas, 2014–2017. Am J Public Health 109:465–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Miech RA, Patrick ME, O’Malley PM, Johnston LD, Bachman JG, 2019. Trends in Reported Marijuana Vaping Among US Adolescents, 2017–2019. Jama. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Moritz ED, Zapata LB, Lekiachvili A, Glidden E, Annor FB, Werner AK, Ussery EN, Hughes MM, Kimball A, et al. , 2019. Update: Characteristics of Patients in a National Outbreak of E-cigarette, or Vaping, Product Use-Associated Lung Injuries United States, October 2019. MMWR Morb Mortal Wkly Rep 68:985–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Navon L, Jones CM, Ghinai I, King BA, Briss PA, Hacker KA, Layden JE, 2019. Risk Factors for E-Cigarette, or Vaping, Product Use-Associated Lung Injury (EVALI) Among Adults Who Use E-Cigarette, or Vaping, Products - Illinois, July-October 2019. MMWR Morb Mortal Wkly Rep 68:1034–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pearson JL, Hitchman SC, Brose LS, Bauld L, Glasser AM, Villanti AC, McNeill A, Abrams DB, Cohen JE, 2018. Recommended core items to assess e-cigarette use in population-based surveys. Tob Control 27:341–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Perrine CG, Pickens CM, Boehmer TK, King BA, Jones CM, DeSisto CL, Duca LM, Lekiachvili A, Kenemer B, et al. , 2019. Characteristics of a Multistate Outbreak of Lung Injury Associated with E-cigarette Use, or Vaping - United States, 2019. MMWR Morb Mortal Wkly Rep 68:860–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rabin RA, George TP, 2015. A review of co-morbid tobacco and cannabis use disorders: possible mechanisms to explain high rates of co-use. Am J Addict 24:105–16. [DOI] [PubMed] [Google Scholar]
- Ramo DE, Liu H, Prochaska JJ, 2012. Tobacco and marijuana use among adolescents and young adults: a systematic review of their co-use. Clin Psychol Rev 32:105–21. [DOI] [PMC free article] [PubMed] [Google Scholar]; SAS System for Windows, 9.4 ed., 2017. SAS Institute Inc., Cary, NC. [Google Scholar]
- Rodriguez A, 2019. Hawaii decriminalizes small amounts of marijuana for January 2020. Retrieved on April 19, 2020 from https://www.usatoday.com/story/news/nation/2019/07/10/hawaii-decriminalizes-marijuana-january-2020/1690727001/
- Schauer GL, Berg CJ, Kegler MC, Donovan DM, Windle M, 2015. Assessing the overlap between tobacco and marijuana: Trends in patterns of co-use of tobacco and marijuana in adults from 2003–2012. Addict Behav 49:26–32. [DOI] [PubMed] [Google Scholar]
- Schauer GL, Peters EN, 2018. Correlates and trends in youth co-use of marijuana and tobacco in the United States, 2005–2014. Drug Alcohol Depend 185:238–44. [DOI] [PubMed] [Google Scholar]
- Simon TR, Stacy AW, Sussman S, Dent C, 1994. Sensation seeking and drug use among high risk latino and anglo adolescents. Pers Individ Dif 17: 665–672. [Google Scholar]
- Stacy A D. Newcomb M, and Bentler P, 1993. Cognitive motivations and sensation seeking as long-term predictors of drinking problems. J Soc Clinic Psychol 12:1–24. [Google Scholar]
- State of Hawaii Department of Health, 2020. Medical cannabis dispensary program. Retrieved on April 19, 2020 from https://health.hawaii.gov/medicalcannabisdispensary/important-documents/
- Sutfin EL, Reboussin BA, Debinski B, Wagoner KG, Spangler J, Wolfson M, 2015. The Impact of Trying Electronic Cigarettes on Cigarette Smoking by College Students: A Prospective Analysis. Am J Public Health 105:e83–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Trivers KF, Phillips E, Gentzke AS, Tynan MA, Neff LJ, 2018. Prevalence of Cannabis Use in Electronic Cigarettes Among US Youth. JAMA Pediatr 172:1097–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tucker JS, Pedersen ER, Seelam R, Dunbar MS, Shih RA, D’Amico EJ, 2019. Types of cannabis and tobacco/nicotine co-use and associated outcomes in young adulthood. Psychol Addict Behav 33:401–11. [DOI] [PMC free article] [PubMed] [Google Scholar]