Abstract
Although many states’ policies and laws concerning marijuana use have recently become more permissive, little is known about the impact of recreational marijuana legalization (RML) on rates of use, risk factors associated with use, and patterns of use over time. We compared samples from two longitudinal studies focused on understanding risk and protective factors related to substance use from adolescence to young adulthood (N = 1,468). The samples were collected 10 years apart from the same neighborhoods in an urban area, and the same measures, research design, and data collection procedures were used in each study. As such, the samples are matched on many demographic variables and provide a unique opportunity to compare rates of use and other associated risk factors before and after RML in Oregon. Our results suggest increased marijuana use in a 30-day time frame among Sample 2 during the young adult years, the time at which RML went into effect in Oregon. In Sample 2, young adults had 2.12 times the odds in Sample 1 of using marijuana at age 24, and they were more likely than those in Sample 1 to report use over multiple time points in young adulthood. Overall, our results suggest that young adults after RML are more likely to use marijuana than young adults were before RML 10 years earlier. Implications for prevention and education are discussed in light of rising rates of daily and 30-day use patterns among this population.
Keywords: Marijuana legalization, substance use, emerging adulthood, recreational marijuana use
State-level legalization of recreational marijuana in the United States is spreading nationwide, yet the public health impact of legalization is still largely unknown. Research is still in the early stages of understanding the impact of these policies on behavior, focusing largely on usage rates and attitudes about marijuana. As such, recreational marijuana legalization (RML) has been associated with increased use among college students (Kerr, Bae, & Koval, 2018), yet research findings are mixed for other age groups. For example, after legalization in Washington, increased rates of use were observed in 8th and 10th grade students, but not among 12th graders (Cerdá et al., 2017). Yet other research which examined marijuana use in Washington using different data and methods of analysis found that rates among adolescent samples remained stable after legalization (Dilley et al., 2018). Regardless of rates of use, it is clear that attitudes about use are more positive among youth in states with RML, including lower perceived harm from use and more favorable attitudes about use (Fleming, Guttmannova, Cambron, Rhew, & Oesterle, 2016). .
Because research on the impact of legalization is in its early stages, many questions about RML effects remain unanswered, such as the effects of RML on patterns of use over time and long- and short-term outcomes associated with varying patterns of use. Unfortunately, an insufficient number of longitudinal samples limit opportunities to make comparisons of cohorts of young adults who came of age pre- and post-RML. Yet research on patterns of marijuana use during early adulthood, and outcomes associated with these patterns, generally suggest reason for concern if RML were to result in patterns of heavier use. Those who use marijuana frequently in early adulthood are at increased risk for a variety of mental and physical health problems, such as anxiety, depression, alcohol use, substance abuse/dependency, sexual risk behaviors, cancers, injuries, and accidents (Epstein et al., 2015; Gordon, Conley, & Gordon, 2013; Hall, 2015; Volkow, Baler, Compton, & Weiss, 2014). Persistent use is also associated with diminished cognitive functioning and motivation, which negatively affect important developmental tasks of early adulthood, such as setting and achieving educational and occupational goals. Indeed, recent research links marijuana use in young adults to alcohol use and to negative consequences of use during the first years of college (Gunn et al., 2018), such as poor achievement and lack of motivation. Chronic users are more likely than abstainers to have lower degree attainment, be unemployed in adulthood, show lower work commitment and achievement, have lower income and reduced financial stability, and greater welfare dependency (Brook, Lee, Finch, Seltzer, & Brook, 2013; Epstein et al., 2015; Lee, Brook, Finch, & Brook, 2015; Schulenberg et al., 2005).
Additional questions about RML relate to concerns about how legalization might affect marijuana use during certain developmental periods. Developmental research suggests that young adulthood is a particularly vulnerable time for substance use: usage rates of marijuana, alcohol, and other drugs increase during this period and typically peak around age 20 (Chen & Jacobsen, 2012). Although it is unclear how legalization might affect use for this age group, national data collected by Monitoring the Future reveal that young adults ages 19–28 reported increased rates of marijuana use between 2007 and 2017, a period characterized by rapid change in many states’ laws regarding the decriminalization of cannabis and legalization of medical and recreational cannabis. During this ten-year period the MTF study found a 3.4% increase in young adults’ reported use of marijuana in one’s lifetime, a 7% increase in reported use in the past 30 days, and a 2.8% increase in reported daily use (Schulenberg et al., 2017). Research looking specifically at young adult usage rates in states before and after RML went into effect is very limited, and findings are unclear. For example, a study of Colorado college students found no differences in usage rates before and after RML (Jones, Jones, & Peil, 2018). In contrast, Kerr and colleagues documented increased cannabis use overall in six of seven U.S. universities, but significantly greater post-legalization increases for Oregon students, suggesting an increase in use among Oregon college students after RML for students who reported recent heavy alcohol use (Kerr et al., 2018). It is important to note that these studies focused only on college students. At present, no research has examined changes in use after RML among young adults not enrolled in college, nor have developmental models from adolescence to adulthood been used to examine change in use over time.
Based on studies linking heavy marijuana use during early adulthood with various negative health and vocational outcomes, it could be assumed that increased usage as a function of RML would also lead to increases in adverse consequences associated with use, but no research has yet been conducted to examine this theory directly. It also could be possible that increased usage associated with RML may lead to fewer associated risks and general acceptance of use in communities. It is unclear whether increased usage rates, as a function of RML, are associated with increased patterns of risk behavior or different patterns of risk behavior.
Data across two individual longitudinal research projects provide a unique opportunity to examine marijuana use in different cohorts of youth, starting in high school and continuing through the transition to young adulthood. In the two projects, a large sample of children were recruited 10 years apart from the same neighborhoods in an urban area. The samples were matched on basic demographic variables, including sex, age, and neighborhood. The goals of the original studies were the same—to follow children through high school and to examine their development over time, with a focus on risk and protective factors that exert an impact on substance use during the young adult years. In Project 1 (Project Alliance 1; PAL 1), students were followed starting in 1995 through the young adult years. In Project 2 (Project Alliance 2; PAL 2), students were followed starting in 2005 through the young adult years. During data collection for PAL 2, RML went into effect in Oregon, where the young adults were residing. Data for the final waves of PAL 2 were collected after the initial legalization of marijuana (Measure 91 was passed in November 2014 and possession/use was legalized beginning July 1, 2015) and after the start of legal sales of recreational marijuana (October 1, 2015). This provided a unique opportunity to compare two cohorts of individuals with similar demographic backgrounds and to examine their trajectories of use over time, as well as to investigate the specific impact of RML on use patterns and co-occurring risk factors. Because the same measures were collected across these samples, and the samples were collected 10 years apart in overlapping schools/neighborhoods, they provide an ideal opportunity to compare two groups of individuals and can provide key information about the impact of RML on usage rates, patterns of usage, and associated risk factors over time.
As such, the goal of this project was to examine three research questions as they relate to RML and marijuana use in the two study samples. First, we used three time points of data that were matched for age across the two studies to examine rates of usage across the samples from high school to young adulthood. On the basis of prior research, we predicted that in comparison with PAL 1 young adults, PAL 2 young adults would show increased marijuana use as a function of RML. Second, we examined risk factors associated with use and compared these risk factors across samples. Given the limited research in this area, we had no specific hypotheses regarding risk. Last, we examined patterns of usage across the samples. Again, given the lack of research in this area, our research was exploratory in nature and would provide important information about the patterns of marijuana use before and after legalization in this matched sample.
Methods
Participants
The data used in this study were derived from two longitudinal projects, PAL 1 and PAL 2, both initially funded by the National Institute on Drug Abuse as randomized, controlled trials (total N = 1,468 across both projects). Recruitment procedures were identical for both projects and began when youth were adolescents. All procedures were approved by the Institutional Review Board at the University of Oregon. In each study, all sixth-grade students in three Title 1 middle schools located in North and Northeast Portland, OR were recruited across two years, and 80% were enrolled. In each study, students and their families were randomly assigned to receive the same family-centered intervention during the middle school years. Participants were followed annually into high school and young adulthood; intervention and control participants were retained at the same rate across each project, with 82% retention for PAL 1 and 78% retention for PAL 2 during the young adult years. In each sample, a high percentage of families qualified for free/reduced-cost lunch at time of recruitment (57% PAL 1 and 64% PAL 2). These samples were still financially at risk at T3, with an average parental household income of $33,000 per year in PAL 1 (2008) and $43,000 per year in PAL 2 (2017).
Table 1 includes the ages of participants in PAL 1 and in PAL 2. Multiple waves of data were collected in each study, but for this article we include only data that were collected at the same age across studies, which we refer to as Wave 1 (T1, early high school), Wave 2 (T2, age 20–22), and Wave 3 (T3, age 22–24). In PAL 1 these data were collected from 2000 to 2010 (N = 946); In PAL 2 these data were collected from 2009 to 2018 (N = 522). The original samples were similar in terms of race/ethnicity, with 42% (PAL 1) and 36% (PAL 2) reporting as European American and the rest of the sample reporting a range of ethnicities, primarily Latino, African American, and biracial.
Table 1.
Demographic Characteristics and Substance Use by Study
| PAL 1 | PAL 2 | |||
|---|---|---|---|---|
| Age in years at Wave 1 [M, (SD)] | 15.16 | (0.40) | 15.08 | (0.43) |
| Age in years at Wave 2 [M, (SD)] | 22.30 | (0.64) | 21.52 | (0.70) |
| Age in years at Wave 3 [M, (SD)] | 23.33 | (0.65) | 22.92 | (0.70) |
| Gender (% female) | 48.2 | 49.4 | ||
| Ethnic minority (% yes) | 53.8 | 64.0 | ||
| Wave 1 marijuana use past 30 days (% yes) | 14.1 | 14.5 | ||
| Wave 1 alcohol use past 30 days (% yes) | 19.2 | 19.8 | ||
Note. M = mean, SD = standard deviation.
Despite being college age at T3, a relatively low percentage of each sample was engaged in higher education: 25.0% of PAL 1 and 30.6% of PAL 2 were actively pursuing or had already completed a 4-year degree. RML went into effect in Oregon when the PAL 2 participants were young adults, and thus the data collected on this sample at Waves 2 and 3 were collected after legalization.
Measures
Age.
Age was calculated in months by comparing participants’ birthdates with the date the surveys were completed at T1, T2, and T3.
Gender.
Participant gender was coded as 0 = male and 1 = female.
Minority status.
Participants in both samples were asked at T1 to indicate their ethnicity. Participants were given a list of seven options and asked to check all that apply. For our study, responses were recoded into two categories: 0 = nonminority (i.e., European American) and 1 = ethnic minority.
Alcohol use, past 30 days.
Participants in both samples were asked at T1 to report the number of alcoholic drinks they had consumed in the past month. Response options ranged from none to 41 or more. For our study, responses were recoded to create a dichotomous “alcohol use in the past 30 days” score for T1 (1 = yes, 0 = no). At T2 and T3, participants were asked three questions about the frequency with which they had consumed beer, wine/wine coolers, and hard liquor in the past 3 months. Response options were never, once or twice, once a month, once every 2–3 weeks, once a week, 2–3 times a week, 4–6 times a week, once a day, and 2–3 times a day (or more). Responses to these three questions were then recorded at each wave to create a dichotomous “alcohol use in the past 30 days” score (1 = yes, 0 = no).
Marijuana use, past 30 days.
Participants in both samples were asked at T1 to report the number of times in the past month they had smoked marijuana or hashish. Response options ranged from none to 41 or more. For our study, responses were recoded to create a dichotomous “marijuana use in the past 30 days” score for T1 (1 = yes, 0 = no). At T2 and T3, participants were asked how often they had used marijuana in the past 3 months. Response options were never, once or twice, once a month, once every 2–3 weeks, once a week, 2–3 times a week, once a day, and 2–3 times a day (or more). Responses were recoded at each wave to create a dichotomous “marijuana use in the past 30 days” score (1 = yes, 0 = no), yet the original scoring was also retained to examine any differences between samples in patterns of usage.
Educational attainment.
Participants in both samples were asked at T2 and T3 to indicate the highest level of education they had completed. The original response options ranged from seventh grade or less to graduate professional training, graduate degree; these were recoded into no high school, high school or GED, and high school plus, the latter including all participants who had completed high school and at least 1 year of college or trade school.
Depression.
In PAL 1, depression was measured at T2 and T3 via the Brief Symptom Inventory (BSI; Derogatis & Melisaratos, 1983). The BSI includes six items used to assess depression symptoms, such as “during the past week, how much were you bothered by feeling blue?” Participants responded to these questions using a five-point scale with end points of not at all and very much. In PAL 2, depression was measured at T2 and T3 via the Adult Self-Report scale developed by Achenbach (Achenbach & Rescorla, 2003). The ASR includes 14 items used to assess depressive problems, such as “I am unhappy, sad, or depressed.” Participants responded to these questions using a three-point scale with anchors of not true, somewhat or sometimes true, and very true or often. Because depression was measured differently between the two samples, t-scores were calculated to be able to make cross-sample comparisons.
Antisocial behavior.
Participants in both samples were asked at T2 and T3 to report the frequency with which they had engaged in 20 antisocial behaviors (e.g., stealing, carrying weapons, engaging in IV drug use) in the past 3 months by using a four-point scale with endpoints of never and every day. Responses were dichotomized to “never” and “at least once,” and then summed to create an antisocial behavior score.
Statistical Analyses
Preliminary analyses included chi-square and grouped t-tests to examine differences between participants who provided marijuana use data at all three assessment waves and participants who had missing marijuana use data. We then evaluated the equivalency of the PAL 1 and PAL 2 studies at Wave 1, when marijuana use was illegal for both samples, on demographic characteristics and reports of marijuana and alcohol use in the past 30 days. Next, binary logistic regression models with the logit link function and reporting of odds ratios and 95% confidence intervals were used to evaluate the association between the PAL 1/PAL 2 study indicator variable and reports of any marijuana use in the past 30 days at Waves 2 and 3, when marijuana use was illegal in the PAL 1 study but legal in the PAL 2 study. Multivariate logistic regression models were used to test, separately, potential moderating effects of risk, protective, and demographic characteristics. Tests of moderation included the simultaneous entry of the moderator variable, the study indicator variable, and the two-way multiplicative term of the moderator and study indicator variable. We then restricted the analytic sample to participants who reported marijuana use in the past 30 days and used ordinal logistic regression to compare the PAL 1 and PAL 2 studies on the frequency of marijuana use. Missing data were imputed at all three waves using Bayesian estimation (Enders, 2010). The imputation model included all variables described in the Measures section and a variable indicating inclusion in PAL 1 or PAL2. Gender and minority status were shown to be related to missingness, as described below in preliminary analyses. Twenty datasets were imputed based on recommendations to impute high numbers of datasets in order to render multiple imputation that is equivalent to maximum likelihood (Graham, Olchowski, & Gilreath, 2007). Parameter estimates were averaged for the twenty datasets and standard errors combined using Rubin’s rules (Rubin, 1987).
Results
Preliminary Analysis
We carefully examined the data and patterns of attrition in the two samples prior to conducting analyses. Failure to provide marijuana use data at all three waves (33%) was not significantly associated with participant age at Wave 1, marijuana or alcohol use at Wave 1, or alcohol use at Wave 2 (all p-values > .535). However, failure to provide marijuana use data at all waves was associated with sex, χ2(1, N = 1,465) = 13.63,p < .001; minority status, χ2(1, N = 1,393) = 10.74, p = .001; and marijuana use at Wave 2, χ2(1, N = 1,195) = 7.37, p = .006. Females were more likely than males to provide data at all three waves (52% vs. 48%), minorities were more likely than nonminorities to provide data at all three waves (55% vs. 45%), and those using marijuana at Wave 2 were less likely than non–marijuana users to provide data at all three waves (38% vs. 62%).
Equivalency of Studies at Wave 1
Table 1 shows demographic characteristics and substance use at Wave 1, when marijuana use was illegal for both studies and the samples were in high school (average age 15). Results from the group comparisons show significant and small effect size differences in mean age, t(1,466) = 3.37,p = .001, and percentage of ethnic minorities, χ2(1, N = 1,468) = 14.32,p < .001. The PAL 2 sample was younger and more racially diverse than the PAL 1 sample. Age and minority status were adjusted in all subsequent logistic regression models. PAL 1 and PAL 2 did not differ on sex or self-report of any marijuana or alcohol use in the past 30 days at Wave 1.
Marijuana Use at Waves 2 and 3
Self-reported use of any marijuana in the past 30 days (see Figure 1) at Wave 2 was less for PAL 1 participants (37%) than for PAL 2 participants (48%).1 Adjusted odds ratios at Wave 2 from the logistic regression models (see Table 2) revealed that PAL 2 participants had 1.78 times the odds of reporting marijuana use than did PAL 1 participants, a significant effect. Self-reported use at Wave 3 decreased for PAL 1 participants (33%) and increased for PAL 2 participants (51%). Adjusted odds ratios at Wave 3 showed PAL 2 participants had 2.12 times the odds of reporting marijuana use than did PAL 1 participants, a significant effect.
Figure 1.
The percentage of past 30-day marijuana use for the PAL 1 and PAL 2 studies. Odds ratios (ORs) associated with the test of study differences in marijuana use adjusted for age and minority status. Unless specified as legal, marijuana use was illegal at the time of data collection. ***p < .001.
Table 2.
Logistic Regression Models Comparing Study Differences on Rates of Marijuana Use
| Model | Beta | SE | p-value | OR | 95% CI |
|---|---|---|---|---|---|
| Wave 1 marijuana use past 30 days | |||||
| Study | 0.06 | 0.17 | .786 | 1.05 | 0.75–1.45 |
| Age at Wave 1 | 0.38 | 0.19 | .052 | 1.46 | 0.99–2.16 |
| Minority status | 0.13 | 0.17 | .418 | 1.14 | 0.83–1.58 |
| Wave 2 marijuana use past 30 days | |||||
| Study | 0.58 | 0.14 | <.001 | 1.78 | 1.35–2.36 |
| Age at Wave 2 | 0.07 | 0.09 | .413 | 1.08 | 0.90–1.28 |
| Minority status | −0.51 | 0.12 | <.001 | 0.60 | 0.47–0.76 |
| Wave 3 marijuana use past 30 days | |||||
| Study | 0.75 | 0.13 | <.001 | 2.12 | 1.65–2.72 |
| Age at Wave 3 | −0.16 | 0.09 | .073 | 0.85 | 0.71–1.02 |
| Minority status | −0.55 | 0.12 | <.001 | 0.58 | 0.45–0.73 |
Note. SE = standard error, OR = odds ratio, CI = confidence interval.
The reference group for study indicator is PAL 1, and the reference group for minority status is nonminority. Age is a continuous measure reported in years.
The pattern of change in past 30-day marijuana use between Waves 2 and 3 was also examined. The pattern of change significantly differed between studies, χ2(1, N = 1,468) = 46.33, p < .001. Compared with PAL 1, PAL 2 had fewer participants who reported no use at both waves (56% vs. 40%), more participants who reported no marijuana use at Wave 2 but marijuana use at Wave 3 (8% vs. 12%), fewer participants who reported marijuana use at Wave 2 and no marijuana use at Wave 3 (11% vs. 9%), and more participants who reported marijuana use at both waves (25% vs. 39%). Thus, all patterns of use between Waves 2 and 3 show greater 30-day marijuana use when marijuana use was legal (i.e., for PAL 2) than when marijuana use was illegal (i.e., for PAL 1).
Moderation Effects
Table 3 shows descriptive statistics for the risk and protective factors by study. The risk and protective factors along with demographic characteristics were modeled as moderators of study (PAL 1 vs. PAL 2). No significant moderating effects were found for risk and protective factors or demographic characteristics at Wave 2 (all p-values > .061) or Wave 3 (all p-values > .068). Moderation results thus showed 30-day marijuana use between studies did not differ as a function of sex, minority status, age, educational attainment, depression, antisocial behavior, and past 30-day alcohol use. The results suggest that the pattern of greater marijuana use by young adults in PAL 2 did not differ as a function of key demographic characteristics or risk factors.
Table 3.
Risk and Protective Factors by Study
| PAL 1 | PAL 2 | |||
|---|---|---|---|---|
| Depression [M, (SD)] | ||||
| Wave 2 | 54.11 | (10.49) | 57.09 | (9.19) |
| Wave 3 | 53.81 | (10.42) | 57.28 | (9.32) |
| Antisocial behavior [M, (SD)] | ||||
| Wave 2 | 4.01 | (2.51) | 3.73 | (2.34) |
| Wave 3 | 4.02 | (2.48) | 3.96 | (2.52) |
| Alcohol use past 30 days (% yes) | ||||
| Wave 2 | 83.1 | 80.3 | ||
| Wave 3 | 82.7 | 84.0 | ||
| Educational attainment (%) | ||||
| Wave 2 | ||||
| High school or less | 67.3 | 62.8 | ||
| High school plus post-high school | 32.7 | 37.2 | ||
| Wave 3 | ||||
| High school or less | 79.2 | 71.3 | ||
| High school plus post-high school | 22.8 | 28.8 | ||
Note. M = Mean, SD = standard deviation. Means, standard deviations, and percentages are pooled across 20 imputed datasets.
Comparisons of Marijuana Users
Participants who reported any marijuana use in the past 30 days were compared, by study, on the frequency of use at Waves 2 and 3 (see Table 4). Frequency of use did not differ by study at Wave 2 (OR = 1.14, 95% CI = 0.79–1.65), or Wave 3 (OR = 1.39, 95% CI = 0.95–2.04). Thus, although overall rates of past 30-day marijuana use were greater for the full analytic sample pre- to post- legalization, frequency of use was remarkably similar, and the pattern of usage was nearly identical. This suggests that although the rate of usage increased for PAL 2, the pattern of usage (e.g., daily use vs. weekly use) did not vary across the samples.
Table 4.
Frequency of Marijuana Use in the Past 3 Months Among Those Reporting Any Use in the Past 30 Days and All Participants
| Current users | All participants | |||
|---|---|---|---|---|
| PAL 1 % | PAL 2 % | PAL 1 % | PAL 2 % | |
| Wave 2 | ||||
| Never | n/a | n/a | 63.4 | 52.5 |
| Once or twice | 24.4 | 26.7 | 8.9 | 12.7 |
| Once a month | 5.3 | 3.7 | 2.0 | 1.8 |
| Once every 2–3 weeks | 11.0 | 7.8 | 4.0 | 3.6 |
| Once a week | 4.9 | 6.5 | 1.8 | 3.1 |
| 2–3 times a week | 14.9 | 13.5 | 5.4 | 6.4 |
| Once a day | 15.6 | 15.7 | 5.7 | 7.5 |
| 2–3 times a day | 24.0 | 26.3 | 8.8 | 12.5 |
| Wave 3 | ||||
| Never | n/a | n/a | 66.6 | 48.7 |
| Once or twice | 22.3 | 25.2 | 7.5 | 12.9 |
| Once a month | 6.2 | 3.9 | 2.1 | 2.0 |
| Once every 2–3 weeks | 7.9 | 5.3 | 2.6 | 2.7 |
| Once a week | 8.7 | 5.8 | 2.9 | 3.0 |
| 2–3 times a week | 13.2 | 11.1 | 4.4 | 5.7 |
| Once a day | 16.2 | 14.2 | 5.4 | 7.3 |
| 2–3 times a day | 25.3 | 34.5 | 8.5 | 17.7 |
Note. Percentages are pooled across 20 imputed datasets.
Discussion
We examined marijuana use across two cohorts of youth who were demographically similar and grew up in the same area of Portland, Oregon ten years apart. These samples provide a unique opportunity to examine the impact of RML on rates of marijuana use, risk factors associated with use, and patterns of marijuana use across time as youth transition to early adulthood. And because the majority of individuals in each sample did not attend college, this study contributes significantly to the existing literature, which has primarily focused on college students. Because research on the effects of RML is just beginning, many of our research questions were exploratory. However, as predicted, the young adult participants in PAL 2 demonstrated increased rates of usage during both waves of the young adult years as a function of RML. These young adults were also more likely to report use at both waves of data collection than were young adults in PAL 1, as well as increased use at Wave 3 when they had reported no use at Wave 2. At Wave 3 (age 22–24), they were more likely to report 30-day usage than were young adults in PAL 1, 10 years earlier. Because the samples were matched on demographic variables and the same research design and procedures were used to collect the data, this is a strong test of the question of increased usage as a function of RML. Although prior research has also suggested an increase usage of marijuana at this age as a result of RML, no studies have examined patterns of use in a community, non–college sample (Kerr et al., 2018).
When we compared the samples on other variables, such as risk factors and patterns of usage, we found the samples to be quite similar. No risk factors, including depression, education, and alcohol use, moderated the higher rates of marijuana use for PAL 2 participants, suggesting these risk factors posed no substantial influence on difference in rates of marijuana use prior to and after RML. Furthermore, patterns of usage did not differ by sample. In both studies, the majority of users were either heavy users (more than once a day) or infrequent users (once or twice a month). Although the results certainly suggest an increase in daily users, they also suggest that RML did not substantially impact the overall pattern of use among young adults. In other words, increased base rates of use did not lead to increases in the percentage of daily users. In fact, the percent of young adults currently using marijuana who were using daily remained very similar across samples before and after RML. Despite the fact that young adults were more likely to report using marijuana after RML, they appear to exhibit a pattern of behavior similar to that of a comparison sample 10 years earlier. The public health impact of legalization, based on these findings, is that more young adults are using marijuana but in a similar pattern of usage to those 10 years ago, which nonetheless suggests a higher base rate of daily usage in the population.
With the increasing legalization of marijuana across the United States and internationally (e.g., Canada), it is critical that we understand the impact of these changed laws on addiction, behavior, and outcomes associated with substance use. This research is a first step in that direction, and it suggests that RML has increased marijuana use among young adults living in Oregon. Nevertheless, there are limitations to our study. First, this is a non-experimental design, and therefore increased rates of use may be due to other factors not measured in this study (e.g., unemployment rates). Second, no data are available concerning young adults at later ages; research is needed that will follow samples into the adult years to further assess the impact of increased use on behavioral outcomes. It may be that the effects of increased usage do not emerge until later in life. There may also be other associated risks, such as lack of motivation, relationship problems, and neurocognitive consequences of heavy use. For example, lower cognitive performance, altered structural brain development, and alterations in brain functioning have been associated with heavy marijuana use in adolescence (Jacobus & Tapert, 2014; Meier et al., 2012). Future research should examine these additional risk factors in the adult years in that they may develop later in life.
It is also important to note that this sample was drawn from a community in Oregon where marijuana use is common and socially acceptable. Rates of use in both samples were higher than the national average usage rates, with 14.5% and 20% of both samples reporting daily usage at Wave 2 (the national average for non–college students is 13%; Miech et al., 2018). At Wave 3, PAL 2 young adults reported a 25% daily use rate, double the national average. Furthermore, even before legalization, PAL 1 participants reported 30-day use rates above the national average (27.7% vs. a national average of approximately 15.8% in 2005; Johnston, O’Malley, Bachman, & Schulenberg, 2006). After RML, PAL 2 participants reported a 30-day use rate of 34.8%, double the national average (18%; Schulenberg et al., 2017). The rate of use among young adults in Oregon was already higher than the national rate, even without RML. This is a potentially significant context for this research, and data in communities with lower baseline levels of use should be examined to assess changes related to RML across various populations.
Last, our research has meaningful implications for prevention and intervention. Patterns of marijuana use are changing as a function of legalization, and more young adults are using later into their lives than was the case 10 years ago. This finding is consistent with recent research findings that suggest increased patterns of usage into later life (age 30) for recent cohorts of individuals (Terry-McElrath & Patrick, 2018). Marijuana use extending into the adult years has been associated with a number of risk factors, including cognitive and memory problems, mental health problems, and addiction (Auer et al., 2015; Volkow et al., 2014). It is critical that we provide young adults with educational materials and support for understanding the effects of marijuana use on behavior. One challenge is that non-college-bound young adults often lack health care or community mental health services, and as a result, this is a difficult population to reach. Public health campaigns that include information about marijuana use and associated risks should be part of the nationwide movement to legalize this substance. In addition, more research is needed to understand the long-term effects of marijuana use into the adult years.
Acknowledgments
Multiple grants to the first author funded this research: National Institute on Drug Abuse (DA018374) and National Institute on Child Health and Human Development (HD075150). Grants from the National Institute on Drug Abuse (DA07031 and DA13773) and a grant from the National Institute on Alcohol Abuse and Alcoholism (AA12702) to Thomas Dishion also supported this research.
Footnotes
This research has not been published previously or presented at any national conferences.
We also computed adjusted odds ratios at Waves 2 and 3 with observed data, and results of study differences in self-reported use of marijuana in the past 30 days were quite similar to results based on imputed data. With observed data only, at Wave 2, PAL 2 participants had 49% greater ratio of odds of reporting marijuana use than did PAL 1 participants, a significant effect (p = .024). At Wave 3, PAL 2 participants had 82% greater ratio of odds of reporting marijuana use than did PAL 1 participants, a significant effect (p <.001).
References
- Achenbach TM, & Rescorla LA (2003). Manual for the ASEBA Adult Form & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, & Families. [Google Scholar]
- Auer R, Vittinghoff E, Yaffe K, Künzi A, Kertesz SG, Levine DA, … Pletcher MJ (2015). Association between lifetime marijuana use and cognitive function in middle age: The Coronary Artery Risk Development in Young Adults (CARDIA) study. JAMA Internal Medicine, 176, 352–361. doi: 10.1001/jamainternmed.2015.7841 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Brook JS, Lee JY, Finch SJ, Seltzer N, & Brook DW (2013). Adult work commitment, financial stability, and social environment as related to trajectories of marijuana use beginning in adolescence. Substance Abuse, 34, 298–305. doi: 10.1080/08897077.2013.775092 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cerdá M, Wall M, Feng T, Keyes KM, Sarvet A, Schulenberg J, … Hasin DS (2017). Association of state recreational marijuana laws with adolescent marijuana use. Journal of the American Medical Association, 171(2), 142–149. doi: 10.1001/jamapediatrics.2016.3624 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen P, & Jacobson KC (2012). Developmental trajectories of substance use from early adolescence to young adulthood: Gender and racial/ethnic differences. Journal of Adolescent Health, 50(2), 154–163. doi: 10.1016/j.jadohealth.2011.05.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Derogatis LR, & Melisaratos N (1983). The Brief Symptom Inventory: An introductory report. Psychological Medicine, 13, 14–27. doi: 10.1017/S0033291700048017 [DOI] [PubMed] [Google Scholar]
- Dilley JA, Richardson SM, Kilmer B, Pacula RL, Segawa MB, & Cerdá M (2018). Prevalence of cannabis use in youths after legalization in Washington state. Journal of the American Medical Association. Advance online. doi: 10.1001/jamapediatrics.2018.4458 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Enders CK (2002). Applied missing data analysis. New York: Guildford Press. [Google Scholar]
- Epstein M, Hill KG, Nevell AM, Guttmannova K, Bailey JA, Abbott RD, & Hawkins JD (2015). Trajectories of marijuana use from adolescence to adulthood: Environmental and individual correlates. Developmental Psychology, 51, 1650–1663. doi: 10.1037/dev0000054 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fleming CB, Guttmannova K, Cambron C, Rhew IC, & Oesterle S (2016). Examination of the divergence in trends for adolescent marijuana use and marijuana-specific risk factors in Washington state. Journal of Adolescent Health, 59(3). Advance online. doi: 10.1016/j.jadohealth.2016.05.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gordon AJ Conley JW, & Gordon JM (2013). Medical consequences of marijuana use: A review of current literature. Current Psychiatry Report, 15, 419. doi: 10.1007/s11920-013-0419-7 [DOI] [PubMed] [Google Scholar]
- Graham JW, Olchowski AE, & Gilreath TD (2007). How many imputations are really needed? Some practical clarifications of multiple imputation theory. Prevention Science, 8(3), 206–213. doi: 10.1007/s11121-007-0070-9 [DOI] [PubMed] [Google Scholar]
- Gunn RL, Norris AL, Sokolovsky A, Micalizzi L, Merrill JE, & Barnett NP (2018). Marijuana use is associated with alcohol use and consequences across the first 2 years of college. Psychology of Addictive Behaviors, 32(8), 885–894. doi: 10.1037/adb0000416 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hall W (2015). What has research over the past two decades revealed about the adverse health effects of recreational cannabis use? Addiction, 110, 19–35. doi: 10.1111/add.12703 [DOI] [PubMed] [Google Scholar]
- Jacobus J, & Tapert SF (2014). Effects of cannabis on the adolescent brain. Current Pharmaceutical Design, 20, 2186–2193. doi: 10.2174/13816128113199990426 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jones J, Jones KN, & Peil J (2018). The impact of the legalization of recreational marijuana on college students. Addictive Behaviors, 77, 255–259. doi: org/10.1016/j.addbeh.2017.08.015 [DOI] [PubMed] [Google Scholar]
- Johnston LD, O’Malley PM, Bachman JG, & Schulenberg JE (2006). Monitoring the Future national survey results on drug use, 1975–2005: Volume II, College students and adults ages 19–45 (NIH Publication No. 06–5884). Bethesda, MD: National Institute on Drug Abuse. [Google Scholar]
- Kerr DCR, Bae H, & Koval AL (2018). Oregon recreational marijuana legalization: Changes in undergraduates’ marijuana use rates from 2008 to 2016. Psychology of Addictive Behaviors, 32(6), 670–678. doi: 10.1037/adb0000385. [DOI] [PubMed] [Google Scholar]
- Lee JY, Brook JS, Finch SJ, & Brook DW (2015). Trajectories of marijuana use from adolescence to adulthood predicting unemployment in the mid-30s. The American Journal on Addictions, 24, 452–459. doi: 10.1111/ajad.12240 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Meier MH, Caspi A, Ambler A, Harrington H, Houts R, Keefe RSE, … Moffitt TE (2012). Persistent cannabis users show neuropsychological decline from childhood to midlife. Proceedings of the National Academy of Sciences, USA, 109, E2657–E2664. doi: 10.1073/pnas.1206820109 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Miech RA, Schulenberg JE, Johnston LD, Bachman JG, O’Malley PM, & Patrick ME (2018, December). National adolescent drug trends in 2018. Monitoring the Future: Ann Arbor, MI: Retrieved from http://www.monitoringthefuture.org. [Google Scholar]
- Salas-Wright CP, Vaughn MG, Todic J, Cordova D, & Perron BE (2015). Trends in the disapproval and use of marijuana among adolescents and young adults in the United States: 2002–2013. American Journal of Drug and Alcohol Abuse, 41(5), 392–404. doi: 10.3109/00952990.2015.1049493 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schulenberg JE, Johnston LD, O’Malley PM, Bachman JG, Miech RA, & Patrick ME (2017). Monitoring the Future national survey results on drug use, 1975–2017: Volume II, College students and adults ages 19–55. Ann Arbor: Institute for Social Research, The University of Michigan. [Google Scholar]
- Schulenberg JE, Merline AC, Johnston LD, O’Malley PM, Bachman JG, & Laetz VB (2005). Trajectories of marijuana use during the transition to adulthood: The big picture based on national panel data. Journal of Drug Issues, 35, 255–280. doi: 10.1177/002204260503500203 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rubin DB (1987). Multiple imputation for nonresponse in surveys. New York: John Wiley & Sons, Inc. [Google Scholar]
- Terry-McElrath YM, & Patrick ME (2018). U.S. adolescent alcohol use by race/ethnicity: Consumption and perceived need to reduce/stop use. Journal of Ethnicity in Substance Abuse. Advance online. doi: 10.1080/15332640.2018.1433094 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Volkow ND, Baler RD, Compton WM, & Weiss SR (2014). Adverse health effects of marijuana use. New England Journal of Medicine, 370, 2219–2227. doi: 10.1056/NEJMra1402309 [DOI] [PMC free article] [PubMed] [Google Scholar]