- Email: [email protected]
Medical marijuana laws and adolescent use of marijuana and other substances: Alcohol, cigarettes, prescription drugs, and other illicit drugs
Drug and Alcohol Dependence 183 (2018) 62–68
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Full length article
Medical marijuana laws and adolescent use of marijuana and other substances: Alcohol, cigarettes, prescription drugs, and other illicit drugs
T
⁎
Magdalena Cerdáa, , Aaron L. Sarvetb,c, Melanie Wallb,c,d, Tianshu Fengc, Katherine M. Keyesb,e, Sandro Galeaf, Deborah S. Hasinb,c,e a
Department of Emergency Medicine, University of California, Davis, 2315 Stockton Blvd., 95817 Sacramento, CA, United States Department of Psychiatry, College of Physicians and Surgeons, Columbia University, New York, NY, United States c New York State Psychiatric Institute, New York, NY, United States d Department of Biostatistics, Mailman School of Public Health, Columbia University, New York, NY, United States e Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, United States f Boston School of Public Health, Boston University, Boston, MA, United States b
A R T I C L E I N F O
A B S T R A C T
Keywords: Medical marijuana Marijuana legalization Substance use Adolescents
Background: Historical shifts have taken place in the last twenty years in marijuana policy. The impact of medical marijuana laws (MML) on use of substances other than marijuana is not well understood. We examined the relationship between state MML and use of marijuana, cigarettes, illicit drugs, nonmedical use of prescription opioids, amphetamines, and tranquilizers, as well as binge drinking. Methods: Pre-post MML difference-in-difference analyses were performed on a nationally representative sample of adolescents in 48 contiguous U.S. states. Participants were 1,179,372 U.S. 8th, 10th, and 12th graders in the national Monitoring the Future annual surveys conducted in 1991–2015. Measurements were any self-reported past-30-day use of marijuana, cigarettes, non-medical use of opioids, amphetamines and tranquilizers, other illicit substances, and any past-two-week binge drinking (5+ drinks per occasion). Results: Among 8th graders, the prevalence of marijuana, binge drinking, cigarette use, non-medical use of opioids, amphetamines and tranquilizers, and any non-marijuana illicit drug use decreased after MML enactment (0.2–2.4% decrease; p-values: < 0.0001–0.0293). Among 10th graders, the prevalence of substance use did not change after MML enactment (p-values: 0.177–0.938). Among 12th graders, non-medical prescription opioid and cigarette use increased after MML enactment (0.9–2.7% increase; p-values: < 0.0001–0.0026). Conclusions: MML enactment is associated with decreases in marijuana and other drugs in early adolescence in those states. Mechanisms that explain the increase in non-medical prescription opioid and cigarette use among 12th graders following MML enactment deserve further study.
1. Introduction Since 1996, 29 states and Washington, D.C. (as of November 2017) enacted legislation permitting the medical use of marijuana. While multiple studies examined the impact of medical marijuana legalization (MML) on adolescent marijuana use (Harper et al., 2012; Hasin et al., 2015; Wall et al., 2011), less is known about MML effects on adolescent use of other substances. If marijuana and other substances are complementary, then increased marijuana use (e.g., through changes in availability and/or price) should increase alcohol and other substance use. This could occur if combined use produces a synergistic psychoactive effect or if marijuana serves as a gateway to other drugs (Kandel and Kandel, 2015; Kandel et al., 1992; Moore, 2010).
⁎
Alternatively, if marijuana and other substances are substitutes, then increased marijuana use should decrease other substance use, which could occur if the substances have similar psychoactive properties and marijuana becomes more accessible. These relationships are usually investigated when a policy directly affects the substance it targets. Virtually all evidence indicates no effect of state MML on adolescent marijuana use (Anderson et al., 2013; Choo et al., 2014; Harper et al., 2012; Hasin et al., 2015; Lynne-Landsman et al., 2013; Schuermeyer et al., 2014; Wall et al., 2011; Wen et al., 2015). However, Colorado studies show that a 2009 federal policy change (reduction in likelihood of prosecution for medical use in MML states; Ogden, 2009) was associated with increased use among adolescents (Salomonsen-Sautel et al., 2014; Schuermeyer et al., 2014). This suggests possible heterogeneous
Corresponding author. E-mail addresses: [email protected], [email protected] (M. Cerdá).
https://doi.org/10.1016/j.drugalcdep.2017.10.021 Received 19 July 2017; Received in revised form 13 October 2017; Accepted 14 October 2017 Available online 07 December 2017 0376-8716/ © 2017 Elsevier B.V. All rights reserved.
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
MTF employs a complex survey design; from a sample of randomly selected geographic units, eligible schools are selected to participate with probability proportional to school size. Non-responding schools are replaced with a similarly-sized school from the same geographic area; schools are asked to participate in the study for two consecutive years. Between 1991 and 2015, 1,179,372 middle and high school students participated in the Monitoring the Future study (423,899 8th graders; 386,596 10th graders; and 368,877 12th graders). Student response rates ranged from 79 to 91% (mean = 86.5%), with the majority of non-response due to absence. Less than 1.7% of students refused to participate in the survey (Miech et al., 2015b). While most drug use was assessed on all questionnaires in MTF, nonmedical prescription opioid use was not, out of a concern that 8th and 10th graders were over-reporting use of prescription opioids, due to difficulty with the definitions. Of the total sample, 82.9% of 8th and 10th graders were randomly assigned to forms that included questions about non-medical prescription opioid use. Analysis of the effects of medical marijuana laws on each substance use behavior excluded participants with missing data for that outcome. The final analytic sample included the following percentages of participants for each substance: binge drinking, 92.3% (N = 1,088,923); cigarette use, 97.3% (N = 1,147,963); non-medical prescription opioid use among those randomized to a form querying use, 96.6% (N = 944,546); non-medical prescription amphetamine use, 97.1% (N = 1,144,618); non-medical prescription tranquilizer use, 97.0% (N = 1,144,029); any illicit drug other than marijuana, 94.3% (N = 1,116,951); and marijuana use, 96.7% (N = 1,140,768). Data were collected from students during normal class periods via paper-and-pencil questionnaires. Participation was anonymous for 8th and 10th graders, and confidential for 12th graders, as some identifying information was collected for follow-up purposes. All study procedures are annually approved by University of Michigan’s Institutional Review Board (Johnston et al., 2014).
MML effects across periods, warranting exploration of cross-substance effects. Further, prior to enactment, MML states have higher rates of marijuana use than non-MML states (Hasin et al., 2015; Keyes et al., 2016; Martins et al., 2016; Wall et al., 2016; Wall et al., 2011), which, according to the gateway hypothesis, could contribute to later initiation of other substances. Studies of abused substances as substitutes or complements mainly focus on alcohol and marijuana. Some found substitution after change in alcohol prices (Cameron and Williams, 2001; Chaloupka and Laixuthai, 1997) and in the legal drinking age (Crost and Guerrero, 2012; DiNardo and Lemieux, 2001), while others showed complementarity after change in alcohol price (Pacula, 1998; Saffer and Chaloupka, 1999; Williams et al., 2004) and legal drinking age (Yoruk and Yoruk, 2011b). The two studies that examined MML and adolescent use of alcohol and other non-marijuana substances also did not agree. One study (Pacula et al., 2013) did not show MML effects overall on alcohol use, but found a positive relationship between home cultivation provisions and alcohol use and between dispensary provisions and alcohol treatment admissions. This study’s inconsistent results were only found at the extremes of alcohol severity. Further, each dataset included different states and age groups, making comparisons difficult. In the second study (Wen et al., 2015), MML did not affect alcohol or other substance use (cocaine, heroin, non-medical use of prescription opioids) among those aged 12–20. However, this study included only eight years and a small number of MML states. Thus, the evidence is inconsistent, warranting further study. Some studies show a relationship between MML enactment and adult decreases in opioid-related harm (Bachhuber et al., 2014; Kim et al., 2016; Pacula et al., 2015), which is potentially explained if marijuana provides a substitute for opioids to treat chronic pain, ameliorate opioid withdrawal symptoms, or assist in recovery from opioid dependence (Lynch and Ware, 2015; Scavone et al., 2013). However, no prior studies examined MML effects on non-medical prescription opioid use across the full range of MML states or addressed MML effects on other types of prescription drugs. We know of no study that examined pre-post MML differences in adolescent binge drinking and use of other non-marijuana substances in all 48 contiguous U.S. states with data that pre-dated the first MML. Our study addressed two questions. First, were participants generally at higher risk for use of marijuana, cigarettes, non-medical use of prescription drugs, illicit drugs, or binge drinking in states that ever passed a MML by 2015 than in other states? This question extends our prior work by examining whether states that ever passed MML are generally at higher risk for use of a wide spectrum of substances. Second, did states that enacted MML exhibit greater change in the prevalence of marijuana use, cigarette use, non-medical prescription drug use, illicit drug use or binge drinking following MML enactment than states that never enacted MML? While results for marijuana were previously reported (Hasin et al., 2015), we include them here because of the additional year of data (2015) available and to evaluate whether MML impacts marijuana and other substances the same way (complementarity) or in opposite ways (substitution). As a sensitivity analysis, we also examined potential differences in the effect of MML enactment on substance use following the 2009 change in federal prosecution policy in MML states (Ogden, 2009).
2.2. Measures The outcome variables were: participant reports of any marijuana use, any binge drinking (five or more drinks in a row), any cigarette use, prescription opioid, amphetamine and tranquilizer use without a prescription or doctor’s permission, and any other illicit substance use. The most recent timeframe, use within the last 30 days (or in the case of binge drinking, in the past two weeks) was used for best recall (Bachman et al., 2015; Miech et al., 2015b). Given the interest in change in non-medical prescription opioid use following MML enactment (Miech et al., 2016; Yoruk and Yoruk, 2011a), prescription drugs including opioids (e.g., morphine, codeine, Vicodin, etc.), amphetamines (e.g., Adderall, Ritalin, etc.), and tranquilizers (e.g., Librium, Valium, Xanax, etc.) were examined separately. Illicit drugs included LSD and other hallucinogens, crack and other cocaine, heroin, amphetamines, tranquilizers, and, among 12th graders only, nonmedical use of sedatives and narcotics other than heroin (Johnston et al., 2017). The exposure variable was state-level medical marijuana law, i.e., whether the respondent lived in a state with a MML when surveyed. The year that MML was passed was determined by review of publicly available state policies conducted by a team of legal scholars, policy analysts and economists from the RAND corporation (Pacula et al., 2014). Dates of MML enactment are included in Appendix S1, and range from 1996 to 2014, providing for 1–19 years of post-legalization followup time, depending on the state. Individual-level covariates included grade, age, sex, race/ethnicity (black, Hispanic, white, Asian, or other), and socioeconomic status (highest parental education as reported by the student: less than high school, high school graduate or equivalent, some post-secondary education, or a 4-year college degree or higher). School-level covariates
2. Methods 2.1. Study design and participants Monitoring the Future (MTF) is an annual, nationally representative, cross-sectional survey of students attending public and private schools in the 48 contiguous U.S. states (Bachman et al., 2015; Miech et al., 2016). Consistent design methodology from its inception allows robust examination of historical trends. Since 1991, students were sampled from 8th, 10th and 12th grades. 63
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
Finally, taking advantage of having five years of data prior to the first state enactment of MML, we implemented a test of trend in the years prior to MML passage to validate the parallel paths assumption (Angrist and Krueger, 1999) (i.e., that trends across time were similar in nonMML and MML states before MML passage). We fit an interaction between time-invariant MML status and all components of the secular trend spline (5 degrees of freedom). SAS code for the primary analyses is provided in the Appendix S2. We used SAS Proc Glimmix (version 9.4). Since the relationship between MML enactment and substance use differed by grade in our data (p-values: < 0.0001–0.006), our models included a multiplicative interaction term between grade and the medical marijuana law variables. Estimated model parameters were used to compute covariate-adjusted estimates of the prevalence of use of these substances within states that ever and never passed a medical marijuana law, and the prevalence of use of these substances before and after enactment of a medical marijuana law. Adjusted prevalence rates for each grade were estimated for each year and were standardized to the grade-specific distributions of individual- and school-level covariates derived from the entire MTF sample (1991–2015) and to the distributions of state-level variables derived from the US population. All models adjusted for the following state-level covariates: % male, % white, % aged 10–24 years, and% older than 25 years without a high school education. Models estimating the prevalence of binge drinking also adjusted for per-gallon state excise taxes on packaged beer. Models estimating the prevalence of cigarette smoking also adjusted for per-pack state excise taxes on cigarettes but not per-gallon state beer excise taxes. See Appendix S3, Table S2 for descriptive statistics of these characteristics in states that did and did not enact MML. Multiple imputation was employed (Proc MI, Proc MI Analyze, SAS version 9.4) to account for missing covariate data for age (2.98%), sex (4.15%), ethnicity (4.27%), and parental education (8.05%); no other covariates had missing data. MTF data were not available for every grade and year combination for some states; the state-level random effects in the multi-level model allow for smoothing of estimates across missing grade-years.
included number of students per grade, whether the school was public or private, and whether the school was in a metropolitan statistical area (United States Census Bureau, 2014). State-level covariates included percentages of each state’s population that was male, white, aged 10–24 years, and older than 25 years without a high school degree. These data were taken from US Censuses conducted in years 1990, 2000, and 2010, which were assigned to participants in survey years 1991–5, 1996–2005, and 2006–15, respectively. Any apparent associations between MML enactment and marijuana use, binge drinking, cigarette use, non-medical prescription drug use, and other illicit substance use could be partly explained by concurrent changes in the prices of substances other than marijuana. To address this concern, we adjusted for a measure of per-gallon state excise taxes on packaged beer (Beer Institute, 2016), per-pack state excise taxes on cigarettes (Orzechowski and Walker, 2015), and yearly, state-specific, purity-adjusted heroin prices. Beer tax data were taken from the 2016 Brewers Almanac and measured alcohol taxes in 1991–2015. Cigarette tax data were taken from the Tax Burden on Tobacco, Volume 49 and measured cigarette taxes in 1991–2014. State cigarette tax levels in 2014 were carried forward to 2015, as data were not available for that year. Yearly state-specific time series of per gram prices of 34% pure heroin at the retail level (≤2 g per sale) were predicted using a random effects estimator that controls for year, expected pure grams, city/state, and region (Caulkins et al., 2004; Kilmer et al., 2014). Heroin price data from the System to Retrieve Information from Drug Evidence was only available for years 1991–2013, which would have restricted our study years; hence, these data were only used in a sensitivity analysis as described below. 2.3. Statistical analyses We used difference-in-difference (DiD) models to examine whether the magnitude of change between MTF surveys differed in adolescents by state MML status (Angrist and Pischke, 2009; Imbens and Wooldridge, 2009). These models assume that trends in non-MML states reflect what would have happened in MML states if they had not passed MML, thus providing information about changes due to MML passage. By focusing on within-state change, DiD tests “difference out” fixed, unmeasured factors that do not vary within states and thus adjust for pre-existing, stable, between-group differences (Hunt and Miles, 2015) that might bias results. We also standardized groups across time to the overall weighted distribution of model covariates, to account for timevarying measured covariates that could confound our associations of interest. We employed multi-level logistic difference-in-difference models of individuals nested within states with two primary predictor variables: a dichotomous time-invariant MML indicator of whether a participant lived in a state that had ever passed an MML by 2015 (regardless of the year in which the participant was measured), and a time-varying MML variable that indicated if a participant lived in a state that had passed an MML in the year surveyed. The time-invariant indicator provides an estimate of the pre-law differences in log odds of substance use between individuals in states that did and did not pass an MML by 2015. The time-varying indicator provides an estimate of the average difference in log odds of substance use in the years prior to versus the years following MML enactment, controlling for contemporaneous changes in non-MML states. A positive or negative estimate of the time-varying indicator indicates larger change (increase or decrease, respectively) in substance use in the MML states post-passage compared to change in use in nonMML states during the same period, net of overall historical trends. Individual and state-level covariates were included as controls in the model. State-level random intercepts and random effects were included to account for the clustering of individuals by state. The non-linear secular trends of substance use over time were modeled using a piecewise cubic spline parameterized the same way for all outcomes. A separate logistic regression model was fit for each substance use outcome.
2.4. Sensitivity analyses To assess whether the effect of MML differed after federal policy towards MML states changed in 2009, additional analyses were run incorporating another time-varying variable indicating whether the MML was in place from 2009 onward or not. To fully address the concern about potential confounding by concurrent changes in the prices of alcohol, tobacco and illicit drugs, we also conducted a sensitivity analysis in which we adjusted all models for yearly per-gallon state excise taxes on packaged beer, per-pack state excise taxes on cigarettes, and yearly, state-specific, purity-adjusted heroin prices. 3. Results First, we examined whether adolescents were generally at higher risk for substance use in states that ever enacted MML by 2015. Among 10th and 12th graders, marijuana use in the prior 30 days was more prevalent before MML was enacted in states that passed MML (Table 1: difference in proportions, MML vs. non-MML states: 10th graders: 2.9%; 12th graders: 4.8%; p-values: 0.0017–0.0278). In contrast, among 10th graders, cigarette use, non-medical prescription amphetamine use and non-medical prescription tranquilizer use were less prevalent in MML states (difference in proportions: −0.5 to −2.4%; p-values: 0.0047–0.0371) before MML was enacted. Among 12th graders, nonmedical prescription tranquilizer use (difference in proportions: −0.4%; p-value: 0.0347) was also less prevalent in MML states (Table 1) before MML was enacted. Second, we examined whether adolescents in states that had 64
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
Table 1 Adolescent use of drugs and alcohol (past 30 days) in the 48 contiguous states between 1991 and 2015: pre-law differences between states that did and did not legalize medical marijuana. Adjusted prevalence
Adjusted odds ratio (95% CI)
Medical marijuana law passed by 2015 %
Medical marijuana law not passed by 2015 %
OR
95% CI
P value
Marijuana
8th grade 10th grade 12th grade
7.9 17.3 21.9
6.7 14.4 17.1
1.18 1.24 1.36
(0.98,1.44) (1.02,1.51) (1.12,1.65)
0.0862 0.0278 0.0017
Binge drinking2
8th grade 10th grade 12th grade
9.2 18.3 25.8
10.0 19.7 25.7
0.91 0.91 1.01
(0.78,1.06) (0.78,1.06) (0.87,1.17)
0.2330 0.2299 0.9162
Cigarettes3
8th grade 10th grade 12th grade
10.8 16.6 22.7
12.1 19.0 25.1
0.87 0.85 0.87
(0.75,1.02) (0.73,0.99) (0.75,1.02)
0.0888 0.0371 0.0828
NM Prescription opioids
8th grade 10th grade 12th grade
2.1 2.8 2.2
1.9 3.1 2.5
1.10 0.90 0.86
(0.93,1.29) (0.77,1.06) (0.73,1.01)
0.2668 0.2264 0.0731
NM Prescription amphetamines
8th grade 10th grade 12th grade
2.6 3.4 3.3
2.8 4.1 3.7
0.93 0.84 0.89
(0.79,1.08) (0.72,0.98) (0.76,1.04)
0.3444 0.0260 0.1355
NM Prescription Tranquilizers
8th grade 10th grade 12th grade
1.0 1.5 1.7
1.1 2.0 2.1
0.90 0.75 0.81
(0.73,1.11) (0.61,0.91) (0.66,0.99)
0.3353 0.0047 0.0347
Any illicit drug (ex. Marijuana)
8th grade 10th grade 12th grade
4.4 6.2 8.2
4.5 6.7 8.6
0.98 0.92 0.94
(0.85,1.13) (0.80,1.06) (0.82,1.08)
0.7419 0.2718 0.3860
1
These analyses adjusted for individual, school and state-level covariates. Individual-level covariates included grade, age, sex, race/ethnicity (black, Hispanic, white, Asian, or other), and socioeconomic status (indicated by highest parental education as reported by the student: less than high school, high school graduate or equivalent, some post-secondary education, or a 4-year college degree or higher). School-level covariates included number of students per grade, whether the school was public or private, and whether the school was in a metropolitan statistical area. State-level covariates included percentages of each state’s population that was male, white, aged 10–24 years, and older than 25 years without a high school degree. 2 Models estimating the prevalence of binge drinking also adjusted for per-gallon state excise taxes on packaged beer up to 2015. 3 Models estimating the prevalence of cigarette smoking also adjusted for per-pack state excise taxes on cigarettes up to 2015. Table 2 Adolescent use of drugs and alcohol (past 30 days) before and after enactment of medical marijuana in the 21 contiguous states that passed medical marijuana laws up to 2015 Adjusted prevalence
Adjusted odds ratio (95% CI)
Before law passed%
After law passed%
OR
95% CI
P value
Marijuana
8th grade 10th grade 12th grade
7.9 17.3 21.9
5.8 17.3 21.3
0.72 1.01 0.96
(0.62,0.84) (0.87,1.16) (0.83,1.11)
< 0.0001 0.9375 0.6204
Binge drinking2
8th grade 10th grade 12th grade
9.2 18.3 25.8
6.8 18.6 27.1
0.72 1.02 1.07
(0.65,0.79) (0.93,1.12) (0.98,1.17)
< 0.0001 0.6396 0.1365
Cigarettes3
8th grade 10th grade 12th grade
10.8 16.6 22.7
8.2 16.4 25.4
0.74 0.98 1.17
(0.66,0.82) (0.89,1.09) (1.06,1.29)
< 0.0001 0.7711 0.0026
NM Prescription Opioids
8th grade 10th grade 12th grade
2.1 2.8 2.2
0.9 2.5 3.1
0.43 0.89 1.42
(0.36,0.52) (0.75,1.05) (1.21,1.66)
< 0.0001 0.1664 < 0.0001
NM Prescription Amphetamines
8th grade 10th grade 12th grade
2.6 3.4 3.3
1.9 3.5 3.4
0.71 1.01 1.02
(0.63,0.81) (0.91,1.13) (0.92,1.14)
< 0.0001 0.8069 0.6889
NM Prescription Tranquilizers
8th grade 10th grade 12th grade
1 1.5 1.7
0.8 1.6 1.8
0.83 1.08 1.08
(0.71,0.98) (0.93,1.25) (0.94,1.24)
0.0293 0.3228 0.2920
Any illicit drug (ex. marijuana)
8th grade 10th grade 12th grade
4.4 6.2 8.2
3.4 6.1 8.7
0.77 0.97 1.07
(0.69,0.86) (0.87,1.07) (0.97,1.18)
< 0.0001 0.5518 0.1567
1
These analyses adjusted for individual, school and state-level covariates. Individual-level covariates included grade, age, sex, race/ethnicity (black, Hispanic, white, Asian, or other), and socioeconomic status (indicated by highest parental education as reported by the student: less than high school, high school graduate or equivalent, some post-secondary education, or a 4-year college degree or higher). School-level covariates included number of students per grade, whether the school was public or private, and whether the school was in a metropolitan statistical area. State-level covariates included percentages of each state’s population that was male, white, aged 10–24 years, and older than 25 years without a high school degree. 2 Models estimating the prevalence of binge drinking also adjusted for per-gallon state excise taxes on packaged beer up to 2015. 3 Models estimating the prevalence of cigarette smoking also adjusted for per-pack state excise taxes on cigarettes, but not per-gallon state beer excise taxes up to 2015.
65
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
harm of marijuana and other substance use than older adolescents. In contrast, although marijuana use did not increase among 12th graders following enactment of MML laws, a post-MML increase in cigarette and non-medical prescription opioid use did occur. Because MML enactment was unrelated to marijuana use in this age group, it is difficult to interpret the findings related to cigarettes and non-prescription opioid use as evidence of a “substitution” effect. However, as shown in this study and in prior work (Hasin et al., 2015), states that passed MML already had higher rates of marijuana use in this age group pre-legalization. Legalizing marijuana and the higher pre-MML levels of marijuana use could have led to decreased perceived harmfulness and increased receptivity to the use of other, related and widely-available substances post-MML. However, we cannot rule out other, unobserved confounders that might explain these findings, including changes in the price of marijuana and prescription opioids. The specific increase in non-medical use of prescription opioids among 12th graders following MML is of particular concern and deserves further study. Prior studies have found that opioid-related harm decreases after MML (Bachhuber et al., 2014; Kim et al., 2016; Powell et al., 2015), while we found that among late adolescents, non-medical prescription opioid use increased following MML enactment. Given the adverse consequences associated with non-medical prescription opioid abuse (Imtiaz et al., 2014; Meyer et al., 2014), the trends in nonmedical prescription opioid use among adolescents following MML need to be monitored as more years of post-MML data become available. In particular, future studies should consider whether these findings persist net of other co-occurring changes that likely influenced the higher increase in non-medical prescription opioid use within MML states compared to non-MML states. Our study offers a number of strengths. First, it provides the first look at pre-post MML differences in non-marijuana substance use in the full range of contiguous U.S. states that enacted MML by 2015 with data predating the enactment of any MML. Second, it specifically compares the impact that MML had on non-medical use of different types of prescription drugs. Third, it presents specific trends in different types of substance use pre- and post-MML enactment and pre- and post-change in federal policy towards MML states for a nationally representative sample of early, middle, and late adolescents. Study findings should be considered in light of the following limitations. First, substance use was self-reported, contributing to potential response bias related to marijuana legalization; however, the data collection method used by MTF has been validated (Bachman et al., 1990; Miech et al., 2015a; Miech et al., 2016; Patrick and Schulenberg, 2010; Schulenberg et al., 1994; Staff et al., 2010), and the assessments were conducted in confidential settings. Second, we can only generalize to the school-attending population of adolescents, who constitute the large majority of U.S. adolescents (Miech et al., 2016). Third, we only examine MML enactment – future studies should examine the contribution of specific features of MMLs, such as provisions for dispensaries, home cultivation, and patient registry requirements (Salomonsen-Sautel et al., 2014; Schuermeyer et al., 2014). Fourth, we do not examine the role that potential mechanisms can play in explaining the relationships between MML enactment and substance use. Fifth, we assumed that passage of MML in one state did not affect the behaviors of individuals in nearby, non-MML states. Sixth, while we accounted for policies that could have affected cigarette, alcohol and heroin prices, there may have been other co-occurring policy changes that could have confounded the relationships of interest. Future studies could examine frequency of use of the different substances as an additional outcome and whether the relationships shown here are modified by gender, race or socioeconomic status. Past-month use of marijuana, alcohol, cigarettes, non-medical use of prescription opioids, tranquilizers and amphetamines, illicit drugs, and binge drinking decreased among 8th graders following MML enactment; no post-MML change was found among 10th graders, and a postMML increase in cigarette use and non-medical prescription opioid use
enacted MML were at higher risk of substance use after enactment of the law compared to before enactment of the law, controlling for contemporaneous trends in non-MML states. Among 8th graders, the prevalence of marijuana use, binge drinking, cigarette use, non-medical prescription opioid use, non-medical prescription amphetamine use, non-medical prescription tranquilizer use, and any non-marijuana illicit drug use decreased after MML enactment (Table 2: post- vs. pre-MML enactment difference in proportions: −0.2 to −2.4%; p-values: < 0.0001–0.0293). Among 10th graders, the prevalence of substance use did not change significantly after MML enactment (p-values: 0.1664–0.9375). Among 12th graders, non-medical prescription opioid use and cigarette use increased after MML enactment (post- vs. preMML enactment difference in proportions: 0.9–2.7%; p-values: < 0.0001–0.0026). No other prevalence of substance use changed significantly after MML enactment (p-values: 0.148–0.692). Tests of parallel paths (prior to passage) indicated no significant differences for all substances. 3.1. Sensitivity analyses Tests of differential changes in use due to MML after changes in federal policy towards MML states in 2009 indicated no new findings that were not already present looking at MML effects across the whole period, suggesting there was no differential change after 2008 over and above the change detected following MML enactment (Appendix S4, Table S3a). After adjusting for state-level yearly beer and cigarette tax levels, and heroin price levels in models of all substances, estimates of pre-law differences between MML and non-MML states were attenuated and no longer significant (Appendix S4, Table S3b). However, estimates of the effects of MML enactment on substance use within MML states did not change with the inclusion of these additional controls (Appendix S4, Table S3c). 4. Discussion We presented national evidence on the impact that enactment of medical marijuana laws had on marijuana use, binge drinking, cigarette use, non-medical prescription drug use, and illicit drug use among 8th, 10th, and 12th graders. The impact of MML enactment on substance use differed by grade. Among 8th graders, marijuana use, binge drinking, cigarette use, non-medical prescription opioid, amphetamine, and tranquilizer use, and other illicit drug use decreased following MML enactment. Among 10th graders, MML was unrelated to changes in use of any substance. Among 12th graders, non-medical prescription opioid use and cigarette use increased following MML enactment, although use of marijuana and other substances did not change. Our age-specific findings contrast with earlier studies that did not differentiate between early and late adolescents and failed to find relationships between MML enactment and use of various substances other than marijuana. The parallel decrease among 8th graders in marijuana use and binge drinking, cigarette, non-medical prescription drug and other illicit drug use following MML enactment suggests that changes associated with MML enactment may have had a complementary impact across substances in the youngest adolescents. Public messaging about the harms associated with adolescent use of marijuana may increase following MML enactment, and eighth graders may be more sensitive than others to public risk messages about substance use, particularly marijuana use. Prior work supports this hypothesis, as it showed that approximately one third of the decrease in marijuana use among 8th graders after MML enactment was explained by increased perceived harmfulness of marijuana in this group (Keyes et al., 2016). Further, parents of young adolescents may be more likely to monitor substance use following MML enactment. Prior studies show that parental influence is highest in early adolescence, waning as adolescents age (Guo et al., 2002; Latendresse et al., 2008; Scalici and Schulz, 2014). Thus, younger adolescents may be more receptive to parental messaging about the 66
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
was found among 12th graders. Our findings suggest that MML enactment may result in complementary decreases in marijuana and other drugs in early adolescence and in an increase in non-medical prescription opioid and cigarette use in late adolescence. Future studies should examine these questions with additional years of post-MML data and investigate the mechanisms that explain the impact that medical marijuana laws have on substances other than marijuana.
econometric evidence. East. Econ. J. 23, 253–276. Choo, E.K., Benz, M., Zaller, N., Warren, O., Rising, K.L., McConnell, K.J., 2014. The impact of state medical marijuana legislation on adolescent marijuana use. J. Adolesc. Health 55, 160–166. Crost, B., Guerrero, S., 2012. The effect of alcohol availability on marijuana use: evidence from the minimum legal drinking age. J. Health Econ. 31, 112–121. DiNardo, J., Lemieux, T., 2001. Alcohol, marijuana, and American youth: the unintended consequences of government regulation. J. Health Econ. 20, 991–1010. Guo, J., Hill, K.G., Hawkins, J.D., Catalano, R.F., Abbott, R.D., 2002. A developmental analysis of sociodemographic, family, and peer effects on adolescent illicit drug initiation. J. Am. Acad. Child Adolesc. Psychiatry 41, 838–845. Harper, S., Strumpf, E.C., Kaufman, J.S., 2012. Do medical marijuana laws increase marijuana use?: Replication study and extension. Ann. Epidemiol. 22, 207–212. Hasin, D.S., Wall, M., Keyes, K.M., Cerda, M., Schulenberg, J., O'Malley, P.M., Galea, S., Pacula, R., Feng, T., 2015. Medical marijuana laws and adolescent marijuana use in the USA from 1991 to 2014: Results from annual, repeated cross-sectional surveys. Lancet Psychiatry 2, 601–608. Hunt, P.E., Miles, J., 2015. The impact of legalizing and regulating weed: issues with study design and emerging findings in the USA. Curr. Top. Behav. Neurosci. 34, 173–198. Imbens, G.W., Wooldridge, J.M., 2009. Recent developments in the econometrics of program evaluation. J. Econ. Lit. 47, 5–86. Imtiaz, S., Shield, K.D., Fischer, B., Rehm, J., 2014. Harms of prescription opioid use in the United States. Subst. Abuse Treat. Prev. Policy 9, 43. Johnston, L.D., O'Malley, P.O., Miech, R.A., Bachman, J.G., Schulenberg, J.E., 2014. Monitoring the Future National Results on Drug Use: 1975–2013: Overview, Key Findings on Adolescent Drug Use Institute for Social Research. The University of Michigan, Ann Arbor. Johnston, L.D., O'Malley, P.M., Miech, R.A., Bachman, J.G., Schulenberg, J.E., 2017. Monitoring the Future National Survey Results on Drug Use, 1975–2016: Overview, Key Findings on Adolescent Drug Use Institute for Social Research. The University of Michigan, Ann Arbor, MI. Kandel, D., Kandel, E., 2015. The Gateway Hypothesis of substance abuse Developmental, biological and societal perspectives. Acta Paediatr. 104, 130–137. Kandel, D.B., 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, 447–457. Keyes, K.M., Wall, M., Cerda, M., Schulenberg, J., O'Malley, P.M., Galea, S., Feng, T., Hasin, D.S., 2016. How does state marijuana policy affect US youth? Medical marijuana laws, marijuana use and perceived harmfulness: 1991–2014. Addiction 111, 2187–2195. Kilmer, B., Everingham, S., Caulkins, J., Midgette, G., Reuter, P., Pacula, R., Burns, R., Han, B., Lundberg, R., 2014. What America’s users spend on illicit drugs: 2000–2010. In: RAND Report RR-534-ONDCP. Santa Monica, CA, February 2014. Kim, J., Santaella-Tenorio, J., Mauro, C., Wrobel, J., Cerdá, M., Keyes, K., Hasin, D.S., 2016. State medical marijuana laws and the prevalence of opioids detected among fatally injured drivers. Am. J. Public Health 106, 2032–2037. Latendresse, S.J., Rose, R.J., Viken, R.J., Pulkkinen, L., Kaprio, J., Dick, D.M., 2008. Parenting mechanisms in links between parents' and adolescents' alcohol use behaviors. Alcohol Clin. Exp. Res. 32, 322–330. Lynch, M.E., Ware, M.A., 2015. Cannabinoids for the treatment of chronic non-cancer pain: an updated systematic review of randomized controlled trials. J. Neuroimmune Pharmacol. 10, 293–301. Lynne-Landsman, S.D., Livingston, M.D., Wagenaar, A.C., 2013. Effects of state medical marijuana laws on adolescent marijuana use. Am. J. Public Health 103, 1500–1506. Martins, S.S., Mauro, C.M., Santaella-Tenorio, J., Kim, J.H., Cerda, M., Keyes, K.M., Hasin, D.S., Galea, S., Wall, M., 2016. State-level medical marijuana laws, marijuana use and perceived availability of marijuana among the general U.S. population. Drug Alcohol Depend. 169, 26–32. Meyer, R., Patel, A.M., Rattana, S.K., Quock, T.P., Mody, S.H., 2014. Prescription opioid abuse: a literature review of the clinical and economic burden in the United States. Popul. Health Manage. 17, 372–387. Miech, R.A., Johnston, L.D., O'Malley, P.M., Bachman, J.G., Schulenberg, J., 2015a. Monitoring the Future National Survey Results on Drug Use, 1975–2013: Volume I: Secondary School Students. Institute for Social Research, The University of Michigan, Ann Arbor. Miech, R.A., Johnston, L.D., O'Malley, P.M., Bachman, J.G., Schulenberg, J., 2015b. Monitoring the Future National Survey Results on Drug Use, 1975–2014: Volume I, Secondary School Students. Institute for Social Research, The University of Michigan, Ann Arbor. Miech, R.A., Johnston, L.D., O'Malley, P.M., Bachman, J.G., Schulenberg, J.E., 2016. Monitoring the Future National Survey Results on Drug Use, 1975–2015: Volume I, Secondary School Students. Institute for Social Research, The University of Michigan, Ann Arbor, MI. Moore, S.C., 2010. Substitution and complementarity in the face of alcohol-specific policy interventions. Alcohol Alcohol. 45, 403–408. Ogden, D., 2009. Memorandum for Selected United States Attorneys on Investigations and Prosecutions in States Authorizing the Medical Use of Marijuana. The United States Department of Justice Archives Accessed at: https://www.justice.gov/ archives/opa/blog/memorandum-selected-united-state-attorneys-investigations-andprosecutions-states. Orzechowski, Walker, 2015. In: The Tax Burden on Tobacco Volume 49, 1970–2014. Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Pacula, R., Powell, D., Heaton, P., Sevigny, E., 2013. Assessing the Effects of Medical Marijuana Laws on Marijuana and Alcohol Use: The Devil Is in the Details NBER
Contributors Cerdá designed the study, interpreted results, and wrote the manuscript; Wall designed the analytic plan; Sarvet, Feng and Wall conducted analyses and wrote the methods and results sections of the manuscript; Keyes and Galea participated in study design and provided input into interpretation of the study results and manuscript drafts; and Hasin provided critical input into study design, design of the analytic plan, interpretation of results, and reviewed manuscript drafts. All authors approved of the final manuscript. Role of funding source Nothing declared. Conflict of interest The authors have no conflicts of interest to declare. Acknowledgements This work was supported by the National Institutes of Health (R01DA034244, R01DA040924, K01DA030449, K01AA021511, T32DA031099) and by the New York State Psychiatric Institute. Funders had no role in study design, data collection, analysis, writing of the manuscript, or decision to submit to this journal. We thank Dr. Rosalie Pacula, Dr. Patrick O’Malley, and Dr. John Schulenberg for their invaluable feedback on this manuscript. This study was funded by R01DA034244 (Hasin), R01DA040924 (Cerdá), K01DA030449 (Cerdá), K01AA021511 (Keyes), T32DA031099 (Hasin) and by the New York State Psychiatric Institute (Hasin, Wall). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.drugalcdep.2017.10.021. References Anderson, D., Hansen, B., Rees, D., 2013. Medical marijuana laws, traffic fatalities, and alcohol consumption. J. Law Econ. 56, 333–369. Angrist, J.D., Krueger, A.B., 1999. Empirical strategies in labor economics. In: In: Ashenfelter, O., Card, D. (Eds.), Handbook of Labor Economics Vol. 3A. Elsevier, Amsterdam, The Netherlands, pp. 1277–1366. Angrist, J.D., Pischke, J.S., 2009. Mostly Harmless Econometrics: An Empiricist's Companion. Princeton University Press, Princeton. Bachhuber, M.A., Saloner, B., Cunningham, C.O., Barry, C.L., 2014. Medical cannabis laws and opioid analgesic overdose mortality in the United States, 1999–2010. JAMA Intern. Med. 174, 1668–1673. Bachman, J.G., Johnston, L.D., O'Malley, P.M., 1990. Explaining the recent decline in cocaine use among young adults: further evidence that perceived risks and disapproval lead to reduced drug use. J. Health Soc. Behav. 31, 173–184. Bachman, J.G., Johnston, L.D., O'Malley, P.M., Schulenberg, J.E., Miech, R.A., 2015. The Monitoring the Future Project After Four Decades: Design and Procedures (Monitoring the Future Occasional Paper No. 82). Institute for Social Research, The University of Michigan, Ann Arbor MI. Institute, Beer, 2016. Brewers Almanac. Beer Institute, Washington, D.C. Cameron, L., Williams, J., 2001. Cannabis, alcohol and cigarettes: substitutes or complements? Econ. Rec. 77, 19–34. Caulkins, J., Pacula, R., Arkes, J., Reuter, P., Paddock, S., Iguchi, M., Riley, J., 2004. The Price and Purity of Illicit Drugs: 1981 Through the Second Quarter of 2003. Office of National Drug Control Policy, Washington D.C. Chaloupka, F.J., Laixuthai, A., 1997. Do youths substitute alcohol and marijuana?: Some
67
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
Colorado compared to non-medical marijuana states: 2003–11. Drug Alcohol Depend. 140, 145–155. Schulenberg, J., Bachman, J.G., O'Malley, P.M., Johnston, L.D., 1994. High school educational success and subsequent substance use: a panel analysis following adolescents into young adulthood. J. Health Soc. Behav. 35, 45–62. Staff, J., Schulenberg, J.E., Maslowsky, J., Bachman, J.G., O'Malley, P.M., Maggs, J.L., Johnston, L.D., 2010. Substance use changes and social role transitions: proximal developmental effects on ongoing trajectories from late adolescence through early adulthood. Dev. Psychopathol. 22, 917–932. United States Census Bureau, 2014. Metropolitan and Micropolitan Statistical Areas Main. (https://www.census.gov/population/metro/. Accessed on April 15, 2014). Wall, M.M., Poh, E., Cerda, M., Keyes, K.M., Galea, S., Hasin, D.S., 2011. Adolescent marijuana use from 2002 to 2008 Higher in states with medical marijuana laws, cause still unclear. Ann. Epidemiol. 21, 714–716. Wall, M.M., Mauro, C., Hasin, D.S., Keyes, K.M., Cerda, M., Martins, S.S., Feng, T., 2016. Prevalence of marijuana use does not differentially increase among youth after states pass medical marijuana laws: commentary on and reanalysis of US National Survey on Drug Use in Households data 2002–2011. Int. J. Drug Policy 29, 9–13. Wen, H., Hockenberry, J.M., Cummings, J.R., 2015. The effect of medical marijuana laws on adolescent and adult use of marijuana, alcohol, and other substances. J. Health Econ. 42, 64–80. Williams, J., Liccardo Pacula, R., Chaloupka, F.J., Wechsler, H., 2004. Alcohol and marijuana use among college students: economic complements or substitutes? Health Econ. 13, 825–843. Yoruk, B.K., Yoruk, C.E., 2011a. The impact of minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use: evidence from a regression discontinuity design using exact date of birth. J. Health Econ. 30, 740–752. Yoruk, B.K., Yoruk, C.E., 2011b. The impact of minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use: evidence from a regression discontinuity design using exact date of birth. J. Health Econ. 30, 740–752.
Working Paper. National Bureau of Economic Research, Cambridge, MA. Pacula, R.L., Bousetad, A.E., Hunt, P., 2014. Words can be deceiving: a review of variation among legally effective medical marijuana laws in the United States. J. Drug Policy Anal. 7, 1–9. Pacula, R., Powell, D., Heaton, P., Sevigny, E., 2015. Assessing the effects of medical marijuana laws on marijuana use: the devil is in the details. J. Policy Anal. Manage. 34, 7–31. Pacula, R.L., 1998. Does increasing the beer tax reduce marijuana consumption? J. Health Econ 17, 557–585. Patrick, M.E., Schulenberg, J.E., 2010. Alcohol use and heavy episodic drinking prevalence and predictors among national samples of American eighth- and tenth-grade students. J. Stud. Alcohol Drugs 71, 41–45. Powell, D., Pacula, R.L., Jacobson, M., 2015. Do Medical Marijuana Laws Reduce Addiction and Deaths Related to Pain Killers RAND Working Paper. Rand Corporation, Santa Monica, CA. Saffer, H., Chaloupka, F., 1999. Demographic differentials in the demand for alcohol and illicit drugs. In: Chaloupka, F., Grossman, M., Bickel, W., Saffer, H. (Eds.), The Economic Analysis of Substance Use and Abuse: An Integration of Econometric and Behavioral Economic Research. University of Chicago Press. Salomonsen-Sautel, S., Min, S.J., Sakai, J.T., Thurstone, C., Hopfer, C., 2014. Trends in fatal motor vehicle crashes before and after marijuana commercialization in Colorado. Drug Alcohol Depend. 140, 137–144. Scalici, F., Schulz, P.J., 2014. Influence of perceived parent and peer endorsement on adolescent smoking intentions: parents have more say, but their influence wanes as kids get older. PLoS One 9, e101275. Scavone, J.L., Sterling, R.C., Van Bockstaele, E.J., 2013. Cannabinoid and opioid interactions: implications for opiate dependence and withdrawal. Neuroscience 248, 637–654. Schuermeyer, J., Salomonsen-Sautel, S., Price, R.K., Balan, S., Thurstone, C., Min, S.J., Sakai, J.T., 2014. Temporal trends in marijuana attitudes, availability and use in
68
Contents lists available at ScienceDirect
Drug and Alcohol Dependence journal homepage: www.elsevier.com/locate/drugalcdep
Full length article
Medical marijuana laws and adolescent use of marijuana and other substances: Alcohol, cigarettes, prescription drugs, and other illicit drugs
T
⁎
Magdalena Cerdáa, , Aaron L. Sarvetb,c, Melanie Wallb,c,d, Tianshu Fengc, Katherine M. Keyesb,e, Sandro Galeaf, Deborah S. Hasinb,c,e a
Department of Emergency Medicine, University of California, Davis, 2315 Stockton Blvd., 95817 Sacramento, CA, United States Department of Psychiatry, College of Physicians and Surgeons, Columbia University, New York, NY, United States c New York State Psychiatric Institute, New York, NY, United States d Department of Biostatistics, Mailman School of Public Health, Columbia University, New York, NY, United States e Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, NY, United States f Boston School of Public Health, Boston University, Boston, MA, United States b
A R T I C L E I N F O
A B S T R A C T
Keywords: Medical marijuana Marijuana legalization Substance use Adolescents
Background: Historical shifts have taken place in the last twenty years in marijuana policy. The impact of medical marijuana laws (MML) on use of substances other than marijuana is not well understood. We examined the relationship between state MML and use of marijuana, cigarettes, illicit drugs, nonmedical use of prescription opioids, amphetamines, and tranquilizers, as well as binge drinking. Methods: Pre-post MML difference-in-difference analyses were performed on a nationally representative sample of adolescents in 48 contiguous U.S. states. Participants were 1,179,372 U.S. 8th, 10th, and 12th graders in the national Monitoring the Future annual surveys conducted in 1991–2015. Measurements were any self-reported past-30-day use of marijuana, cigarettes, non-medical use of opioids, amphetamines and tranquilizers, other illicit substances, and any past-two-week binge drinking (5+ drinks per occasion). Results: Among 8th graders, the prevalence of marijuana, binge drinking, cigarette use, non-medical use of opioids, amphetamines and tranquilizers, and any non-marijuana illicit drug use decreased after MML enactment (0.2–2.4% decrease; p-values: < 0.0001–0.0293). Among 10th graders, the prevalence of substance use did not change after MML enactment (p-values: 0.177–0.938). Among 12th graders, non-medical prescription opioid and cigarette use increased after MML enactment (0.9–2.7% increase; p-values: < 0.0001–0.0026). Conclusions: MML enactment is associated with decreases in marijuana and other drugs in early adolescence in those states. Mechanisms that explain the increase in non-medical prescription opioid and cigarette use among 12th graders following MML enactment deserve further study.
1. Introduction Since 1996, 29 states and Washington, D.C. (as of November 2017) enacted legislation permitting the medical use of marijuana. While multiple studies examined the impact of medical marijuana legalization (MML) on adolescent marijuana use (Harper et al., 2012; Hasin et al., 2015; Wall et al., 2011), less is known about MML effects on adolescent use of other substances. If marijuana and other substances are complementary, then increased marijuana use (e.g., through changes in availability and/or price) should increase alcohol and other substance use. This could occur if combined use produces a synergistic psychoactive effect or if marijuana serves as a gateway to other drugs (Kandel and Kandel, 2015; Kandel et al., 1992; Moore, 2010).
⁎
Alternatively, if marijuana and other substances are substitutes, then increased marijuana use should decrease other substance use, which could occur if the substances have similar psychoactive properties and marijuana becomes more accessible. These relationships are usually investigated when a policy directly affects the substance it targets. Virtually all evidence indicates no effect of state MML on adolescent marijuana use (Anderson et al., 2013; Choo et al., 2014; Harper et al., 2012; Hasin et al., 2015; Lynne-Landsman et al., 2013; Schuermeyer et al., 2014; Wall et al., 2011; Wen et al., 2015). However, Colorado studies show that a 2009 federal policy change (reduction in likelihood of prosecution for medical use in MML states; Ogden, 2009) was associated with increased use among adolescents (Salomonsen-Sautel et al., 2014; Schuermeyer et al., 2014). This suggests possible heterogeneous
Corresponding author. E-mail addresses: [email protected], [email protected] (M. Cerdá).
https://doi.org/10.1016/j.drugalcdep.2017.10.021 Received 19 July 2017; Received in revised form 13 October 2017; Accepted 14 October 2017 Available online 07 December 2017 0376-8716/ © 2017 Elsevier B.V. All rights reserved.
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
MTF employs a complex survey design; from a sample of randomly selected geographic units, eligible schools are selected to participate with probability proportional to school size. Non-responding schools are replaced with a similarly-sized school from the same geographic area; schools are asked to participate in the study for two consecutive years. Between 1991 and 2015, 1,179,372 middle and high school students participated in the Monitoring the Future study (423,899 8th graders; 386,596 10th graders; and 368,877 12th graders). Student response rates ranged from 79 to 91% (mean = 86.5%), with the majority of non-response due to absence. Less than 1.7% of students refused to participate in the survey (Miech et al., 2015b). While most drug use was assessed on all questionnaires in MTF, nonmedical prescription opioid use was not, out of a concern that 8th and 10th graders were over-reporting use of prescription opioids, due to difficulty with the definitions. Of the total sample, 82.9% of 8th and 10th graders were randomly assigned to forms that included questions about non-medical prescription opioid use. Analysis of the effects of medical marijuana laws on each substance use behavior excluded participants with missing data for that outcome. The final analytic sample included the following percentages of participants for each substance: binge drinking, 92.3% (N = 1,088,923); cigarette use, 97.3% (N = 1,147,963); non-medical prescription opioid use among those randomized to a form querying use, 96.6% (N = 944,546); non-medical prescription amphetamine use, 97.1% (N = 1,144,618); non-medical prescription tranquilizer use, 97.0% (N = 1,144,029); any illicit drug other than marijuana, 94.3% (N = 1,116,951); and marijuana use, 96.7% (N = 1,140,768). Data were collected from students during normal class periods via paper-and-pencil questionnaires. Participation was anonymous for 8th and 10th graders, and confidential for 12th graders, as some identifying information was collected for follow-up purposes. All study procedures are annually approved by University of Michigan’s Institutional Review Board (Johnston et al., 2014).
MML effects across periods, warranting exploration of cross-substance effects. Further, prior to enactment, MML states have higher rates of marijuana use than non-MML states (Hasin et al., 2015; Keyes et al., 2016; Martins et al., 2016; Wall et al., 2016; Wall et al., 2011), which, according to the gateway hypothesis, could contribute to later initiation of other substances. Studies of abused substances as substitutes or complements mainly focus on alcohol and marijuana. Some found substitution after change in alcohol prices (Cameron and Williams, 2001; Chaloupka and Laixuthai, 1997) and in the legal drinking age (Crost and Guerrero, 2012; DiNardo and Lemieux, 2001), while others showed complementarity after change in alcohol price (Pacula, 1998; Saffer and Chaloupka, 1999; Williams et al., 2004) and legal drinking age (Yoruk and Yoruk, 2011b). The two studies that examined MML and adolescent use of alcohol and other non-marijuana substances also did not agree. One study (Pacula et al., 2013) did not show MML effects overall on alcohol use, but found a positive relationship between home cultivation provisions and alcohol use and between dispensary provisions and alcohol treatment admissions. This study’s inconsistent results were only found at the extremes of alcohol severity. Further, each dataset included different states and age groups, making comparisons difficult. In the second study (Wen et al., 2015), MML did not affect alcohol or other substance use (cocaine, heroin, non-medical use of prescription opioids) among those aged 12–20. However, this study included only eight years and a small number of MML states. Thus, the evidence is inconsistent, warranting further study. Some studies show a relationship between MML enactment and adult decreases in opioid-related harm (Bachhuber et al., 2014; Kim et al., 2016; Pacula et al., 2015), which is potentially explained if marijuana provides a substitute for opioids to treat chronic pain, ameliorate opioid withdrawal symptoms, or assist in recovery from opioid dependence (Lynch and Ware, 2015; Scavone et al., 2013). However, no prior studies examined MML effects on non-medical prescription opioid use across the full range of MML states or addressed MML effects on other types of prescription drugs. We know of no study that examined pre-post MML differences in adolescent binge drinking and use of other non-marijuana substances in all 48 contiguous U.S. states with data that pre-dated the first MML. Our study addressed two questions. First, were participants generally at higher risk for use of marijuana, cigarettes, non-medical use of prescription drugs, illicit drugs, or binge drinking in states that ever passed a MML by 2015 than in other states? This question extends our prior work by examining whether states that ever passed MML are generally at higher risk for use of a wide spectrum of substances. Second, did states that enacted MML exhibit greater change in the prevalence of marijuana use, cigarette use, non-medical prescription drug use, illicit drug use or binge drinking following MML enactment than states that never enacted MML? While results for marijuana were previously reported (Hasin et al., 2015), we include them here because of the additional year of data (2015) available and to evaluate whether MML impacts marijuana and other substances the same way (complementarity) or in opposite ways (substitution). As a sensitivity analysis, we also examined potential differences in the effect of MML enactment on substance use following the 2009 change in federal prosecution policy in MML states (Ogden, 2009).
2.2. Measures The outcome variables were: participant reports of any marijuana use, any binge drinking (five or more drinks in a row), any cigarette use, prescription opioid, amphetamine and tranquilizer use without a prescription or doctor’s permission, and any other illicit substance use. The most recent timeframe, use within the last 30 days (or in the case of binge drinking, in the past two weeks) was used for best recall (Bachman et al., 2015; Miech et al., 2015b). Given the interest in change in non-medical prescription opioid use following MML enactment (Miech et al., 2016; Yoruk and Yoruk, 2011a), prescription drugs including opioids (e.g., morphine, codeine, Vicodin, etc.), amphetamines (e.g., Adderall, Ritalin, etc.), and tranquilizers (e.g., Librium, Valium, Xanax, etc.) were examined separately. Illicit drugs included LSD and other hallucinogens, crack and other cocaine, heroin, amphetamines, tranquilizers, and, among 12th graders only, nonmedical use of sedatives and narcotics other than heroin (Johnston et al., 2017). The exposure variable was state-level medical marijuana law, i.e., whether the respondent lived in a state with a MML when surveyed. The year that MML was passed was determined by review of publicly available state policies conducted by a team of legal scholars, policy analysts and economists from the RAND corporation (Pacula et al., 2014). Dates of MML enactment are included in Appendix S1, and range from 1996 to 2014, providing for 1–19 years of post-legalization followup time, depending on the state. Individual-level covariates included grade, age, sex, race/ethnicity (black, Hispanic, white, Asian, or other), and socioeconomic status (highest parental education as reported by the student: less than high school, high school graduate or equivalent, some post-secondary education, or a 4-year college degree or higher). School-level covariates
2. Methods 2.1. Study design and participants Monitoring the Future (MTF) is an annual, nationally representative, cross-sectional survey of students attending public and private schools in the 48 contiguous U.S. states (Bachman et al., 2015; Miech et al., 2016). Consistent design methodology from its inception allows robust examination of historical trends. Since 1991, students were sampled from 8th, 10th and 12th grades. 63
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
Finally, taking advantage of having five years of data prior to the first state enactment of MML, we implemented a test of trend in the years prior to MML passage to validate the parallel paths assumption (Angrist and Krueger, 1999) (i.e., that trends across time were similar in nonMML and MML states before MML passage). We fit an interaction between time-invariant MML status and all components of the secular trend spline (5 degrees of freedom). SAS code for the primary analyses is provided in the Appendix S2. We used SAS Proc Glimmix (version 9.4). Since the relationship between MML enactment and substance use differed by grade in our data (p-values: < 0.0001–0.006), our models included a multiplicative interaction term between grade and the medical marijuana law variables. Estimated model parameters were used to compute covariate-adjusted estimates of the prevalence of use of these substances within states that ever and never passed a medical marijuana law, and the prevalence of use of these substances before and after enactment of a medical marijuana law. Adjusted prevalence rates for each grade were estimated for each year and were standardized to the grade-specific distributions of individual- and school-level covariates derived from the entire MTF sample (1991–2015) and to the distributions of state-level variables derived from the US population. All models adjusted for the following state-level covariates: % male, % white, % aged 10–24 years, and% older than 25 years without a high school education. Models estimating the prevalence of binge drinking also adjusted for per-gallon state excise taxes on packaged beer. Models estimating the prevalence of cigarette smoking also adjusted for per-pack state excise taxes on cigarettes but not per-gallon state beer excise taxes. See Appendix S3, Table S2 for descriptive statistics of these characteristics in states that did and did not enact MML. Multiple imputation was employed (Proc MI, Proc MI Analyze, SAS version 9.4) to account for missing covariate data for age (2.98%), sex (4.15%), ethnicity (4.27%), and parental education (8.05%); no other covariates had missing data. MTF data were not available for every grade and year combination for some states; the state-level random effects in the multi-level model allow for smoothing of estimates across missing grade-years.
included number of students per grade, whether the school was public or private, and whether the school was in a metropolitan statistical area (United States Census Bureau, 2014). State-level covariates included percentages of each state’s population that was male, white, aged 10–24 years, and older than 25 years without a high school degree. These data were taken from US Censuses conducted in years 1990, 2000, and 2010, which were assigned to participants in survey years 1991–5, 1996–2005, and 2006–15, respectively. Any apparent associations between MML enactment and marijuana use, binge drinking, cigarette use, non-medical prescription drug use, and other illicit substance use could be partly explained by concurrent changes in the prices of substances other than marijuana. To address this concern, we adjusted for a measure of per-gallon state excise taxes on packaged beer (Beer Institute, 2016), per-pack state excise taxes on cigarettes (Orzechowski and Walker, 2015), and yearly, state-specific, purity-adjusted heroin prices. Beer tax data were taken from the 2016 Brewers Almanac and measured alcohol taxes in 1991–2015. Cigarette tax data were taken from the Tax Burden on Tobacco, Volume 49 and measured cigarette taxes in 1991–2014. State cigarette tax levels in 2014 were carried forward to 2015, as data were not available for that year. Yearly state-specific time series of per gram prices of 34% pure heroin at the retail level (≤2 g per sale) were predicted using a random effects estimator that controls for year, expected pure grams, city/state, and region (Caulkins et al., 2004; Kilmer et al., 2014). Heroin price data from the System to Retrieve Information from Drug Evidence was only available for years 1991–2013, which would have restricted our study years; hence, these data were only used in a sensitivity analysis as described below. 2.3. Statistical analyses We used difference-in-difference (DiD) models to examine whether the magnitude of change between MTF surveys differed in adolescents by state MML status (Angrist and Pischke, 2009; Imbens and Wooldridge, 2009). These models assume that trends in non-MML states reflect what would have happened in MML states if they had not passed MML, thus providing information about changes due to MML passage. By focusing on within-state change, DiD tests “difference out” fixed, unmeasured factors that do not vary within states and thus adjust for pre-existing, stable, between-group differences (Hunt and Miles, 2015) that might bias results. We also standardized groups across time to the overall weighted distribution of model covariates, to account for timevarying measured covariates that could confound our associations of interest. We employed multi-level logistic difference-in-difference models of individuals nested within states with two primary predictor variables: a dichotomous time-invariant MML indicator of whether a participant lived in a state that had ever passed an MML by 2015 (regardless of the year in which the participant was measured), and a time-varying MML variable that indicated if a participant lived in a state that had passed an MML in the year surveyed. The time-invariant indicator provides an estimate of the pre-law differences in log odds of substance use between individuals in states that did and did not pass an MML by 2015. The time-varying indicator provides an estimate of the average difference in log odds of substance use in the years prior to versus the years following MML enactment, controlling for contemporaneous changes in non-MML states. A positive or negative estimate of the time-varying indicator indicates larger change (increase or decrease, respectively) in substance use in the MML states post-passage compared to change in use in nonMML states during the same period, net of overall historical trends. Individual and state-level covariates were included as controls in the model. State-level random intercepts and random effects were included to account for the clustering of individuals by state. The non-linear secular trends of substance use over time were modeled using a piecewise cubic spline parameterized the same way for all outcomes. A separate logistic regression model was fit for each substance use outcome.
2.4. Sensitivity analyses To assess whether the effect of MML differed after federal policy towards MML states changed in 2009, additional analyses were run incorporating another time-varying variable indicating whether the MML was in place from 2009 onward or not. To fully address the concern about potential confounding by concurrent changes in the prices of alcohol, tobacco and illicit drugs, we also conducted a sensitivity analysis in which we adjusted all models for yearly per-gallon state excise taxes on packaged beer, per-pack state excise taxes on cigarettes, and yearly, state-specific, purity-adjusted heroin prices. 3. Results First, we examined whether adolescents were generally at higher risk for substance use in states that ever enacted MML by 2015. Among 10th and 12th graders, marijuana use in the prior 30 days was more prevalent before MML was enacted in states that passed MML (Table 1: difference in proportions, MML vs. non-MML states: 10th graders: 2.9%; 12th graders: 4.8%; p-values: 0.0017–0.0278). In contrast, among 10th graders, cigarette use, non-medical prescription amphetamine use and non-medical prescription tranquilizer use were less prevalent in MML states (difference in proportions: −0.5 to −2.4%; p-values: 0.0047–0.0371) before MML was enacted. Among 12th graders, nonmedical prescription tranquilizer use (difference in proportions: −0.4%; p-value: 0.0347) was also less prevalent in MML states (Table 1) before MML was enacted. Second, we examined whether adolescents in states that had 64
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
Table 1 Adolescent use of drugs and alcohol (past 30 days) in the 48 contiguous states between 1991 and 2015: pre-law differences between states that did and did not legalize medical marijuana. Adjusted prevalence
Adjusted odds ratio (95% CI)
Medical marijuana law passed by 2015 %
Medical marijuana law not passed by 2015 %
OR
95% CI
P value
Marijuana
8th grade 10th grade 12th grade
7.9 17.3 21.9
6.7 14.4 17.1
1.18 1.24 1.36
(0.98,1.44) (1.02,1.51) (1.12,1.65)
0.0862 0.0278 0.0017
Binge drinking2
8th grade 10th grade 12th grade
9.2 18.3 25.8
10.0 19.7 25.7
0.91 0.91 1.01
(0.78,1.06) (0.78,1.06) (0.87,1.17)
0.2330 0.2299 0.9162
Cigarettes3
8th grade 10th grade 12th grade
10.8 16.6 22.7
12.1 19.0 25.1
0.87 0.85 0.87
(0.75,1.02) (0.73,0.99) (0.75,1.02)
0.0888 0.0371 0.0828
NM Prescription opioids
8th grade 10th grade 12th grade
2.1 2.8 2.2
1.9 3.1 2.5
1.10 0.90 0.86
(0.93,1.29) (0.77,1.06) (0.73,1.01)
0.2668 0.2264 0.0731
NM Prescription amphetamines
8th grade 10th grade 12th grade
2.6 3.4 3.3
2.8 4.1 3.7
0.93 0.84 0.89
(0.79,1.08) (0.72,0.98) (0.76,1.04)
0.3444 0.0260 0.1355
NM Prescription Tranquilizers
8th grade 10th grade 12th grade
1.0 1.5 1.7
1.1 2.0 2.1
0.90 0.75 0.81
(0.73,1.11) (0.61,0.91) (0.66,0.99)
0.3353 0.0047 0.0347
Any illicit drug (ex. Marijuana)
8th grade 10th grade 12th grade
4.4 6.2 8.2
4.5 6.7 8.6
0.98 0.92 0.94
(0.85,1.13) (0.80,1.06) (0.82,1.08)
0.7419 0.2718 0.3860
1
These analyses adjusted for individual, school and state-level covariates. Individual-level covariates included grade, age, sex, race/ethnicity (black, Hispanic, white, Asian, or other), and socioeconomic status (indicated by highest parental education as reported by the student: less than high school, high school graduate or equivalent, some post-secondary education, or a 4-year college degree or higher). School-level covariates included number of students per grade, whether the school was public or private, and whether the school was in a metropolitan statistical area. State-level covariates included percentages of each state’s population that was male, white, aged 10–24 years, and older than 25 years without a high school degree. 2 Models estimating the prevalence of binge drinking also adjusted for per-gallon state excise taxes on packaged beer up to 2015. 3 Models estimating the prevalence of cigarette smoking also adjusted for per-pack state excise taxes on cigarettes up to 2015. Table 2 Adolescent use of drugs and alcohol (past 30 days) before and after enactment of medical marijuana in the 21 contiguous states that passed medical marijuana laws up to 2015 Adjusted prevalence
Adjusted odds ratio (95% CI)
Before law passed%
After law passed%
OR
95% CI
P value
Marijuana
8th grade 10th grade 12th grade
7.9 17.3 21.9
5.8 17.3 21.3
0.72 1.01 0.96
(0.62,0.84) (0.87,1.16) (0.83,1.11)
< 0.0001 0.9375 0.6204
Binge drinking2
8th grade 10th grade 12th grade
9.2 18.3 25.8
6.8 18.6 27.1
0.72 1.02 1.07
(0.65,0.79) (0.93,1.12) (0.98,1.17)
< 0.0001 0.6396 0.1365
Cigarettes3
8th grade 10th grade 12th grade
10.8 16.6 22.7
8.2 16.4 25.4
0.74 0.98 1.17
(0.66,0.82) (0.89,1.09) (1.06,1.29)
< 0.0001 0.7711 0.0026
NM Prescription Opioids
8th grade 10th grade 12th grade
2.1 2.8 2.2
0.9 2.5 3.1
0.43 0.89 1.42
(0.36,0.52) (0.75,1.05) (1.21,1.66)
< 0.0001 0.1664 < 0.0001
NM Prescription Amphetamines
8th grade 10th grade 12th grade
2.6 3.4 3.3
1.9 3.5 3.4
0.71 1.01 1.02
(0.63,0.81) (0.91,1.13) (0.92,1.14)
< 0.0001 0.8069 0.6889
NM Prescription Tranquilizers
8th grade 10th grade 12th grade
1 1.5 1.7
0.8 1.6 1.8
0.83 1.08 1.08
(0.71,0.98) (0.93,1.25) (0.94,1.24)
0.0293 0.3228 0.2920
Any illicit drug (ex. marijuana)
8th grade 10th grade 12th grade
4.4 6.2 8.2
3.4 6.1 8.7
0.77 0.97 1.07
(0.69,0.86) (0.87,1.07) (0.97,1.18)
< 0.0001 0.5518 0.1567
1
These analyses adjusted for individual, school and state-level covariates. Individual-level covariates included grade, age, sex, race/ethnicity (black, Hispanic, white, Asian, or other), and socioeconomic status (indicated by highest parental education as reported by the student: less than high school, high school graduate or equivalent, some post-secondary education, or a 4-year college degree or higher). School-level covariates included number of students per grade, whether the school was public or private, and whether the school was in a metropolitan statistical area. State-level covariates included percentages of each state’s population that was male, white, aged 10–24 years, and older than 25 years without a high school degree. 2 Models estimating the prevalence of binge drinking also adjusted for per-gallon state excise taxes on packaged beer up to 2015. 3 Models estimating the prevalence of cigarette smoking also adjusted for per-pack state excise taxes on cigarettes, but not per-gallon state beer excise taxes up to 2015.
65
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
harm of marijuana and other substance use than older adolescents. In contrast, although marijuana use did not increase among 12th graders following enactment of MML laws, a post-MML increase in cigarette and non-medical prescription opioid use did occur. Because MML enactment was unrelated to marijuana use in this age group, it is difficult to interpret the findings related to cigarettes and non-prescription opioid use as evidence of a “substitution” effect. However, as shown in this study and in prior work (Hasin et al., 2015), states that passed MML already had higher rates of marijuana use in this age group pre-legalization. Legalizing marijuana and the higher pre-MML levels of marijuana use could have led to decreased perceived harmfulness and increased receptivity to the use of other, related and widely-available substances post-MML. However, we cannot rule out other, unobserved confounders that might explain these findings, including changes in the price of marijuana and prescription opioids. The specific increase in non-medical use of prescription opioids among 12th graders following MML is of particular concern and deserves further study. Prior studies have found that opioid-related harm decreases after MML (Bachhuber et al., 2014; Kim et al., 2016; Powell et al., 2015), while we found that among late adolescents, non-medical prescription opioid use increased following MML enactment. Given the adverse consequences associated with non-medical prescription opioid abuse (Imtiaz et al., 2014; Meyer et al., 2014), the trends in nonmedical prescription opioid use among adolescents following MML need to be monitored as more years of post-MML data become available. In particular, future studies should consider whether these findings persist net of other co-occurring changes that likely influenced the higher increase in non-medical prescription opioid use within MML states compared to non-MML states. Our study offers a number of strengths. First, it provides the first look at pre-post MML differences in non-marijuana substance use in the full range of contiguous U.S. states that enacted MML by 2015 with data predating the enactment of any MML. Second, it specifically compares the impact that MML had on non-medical use of different types of prescription drugs. Third, it presents specific trends in different types of substance use pre- and post-MML enactment and pre- and post-change in federal policy towards MML states for a nationally representative sample of early, middle, and late adolescents. Study findings should be considered in light of the following limitations. First, substance use was self-reported, contributing to potential response bias related to marijuana legalization; however, the data collection method used by MTF has been validated (Bachman et al., 1990; Miech et al., 2015a; Miech et al., 2016; Patrick and Schulenberg, 2010; Schulenberg et al., 1994; Staff et al., 2010), and the assessments were conducted in confidential settings. Second, we can only generalize to the school-attending population of adolescents, who constitute the large majority of U.S. adolescents (Miech et al., 2016). Third, we only examine MML enactment – future studies should examine the contribution of specific features of MMLs, such as provisions for dispensaries, home cultivation, and patient registry requirements (Salomonsen-Sautel et al., 2014; Schuermeyer et al., 2014). Fourth, we do not examine the role that potential mechanisms can play in explaining the relationships between MML enactment and substance use. Fifth, we assumed that passage of MML in one state did not affect the behaviors of individuals in nearby, non-MML states. Sixth, while we accounted for policies that could have affected cigarette, alcohol and heroin prices, there may have been other co-occurring policy changes that could have confounded the relationships of interest. Future studies could examine frequency of use of the different substances as an additional outcome and whether the relationships shown here are modified by gender, race or socioeconomic status. Past-month use of marijuana, alcohol, cigarettes, non-medical use of prescription opioids, tranquilizers and amphetamines, illicit drugs, and binge drinking decreased among 8th graders following MML enactment; no post-MML change was found among 10th graders, and a postMML increase in cigarette use and non-medical prescription opioid use
enacted MML were at higher risk of substance use after enactment of the law compared to before enactment of the law, controlling for contemporaneous trends in non-MML states. Among 8th graders, the prevalence of marijuana use, binge drinking, cigarette use, non-medical prescription opioid use, non-medical prescription amphetamine use, non-medical prescription tranquilizer use, and any non-marijuana illicit drug use decreased after MML enactment (Table 2: post- vs. pre-MML enactment difference in proportions: −0.2 to −2.4%; p-values: < 0.0001–0.0293). Among 10th graders, the prevalence of substance use did not change significantly after MML enactment (p-values: 0.1664–0.9375). Among 12th graders, non-medical prescription opioid use and cigarette use increased after MML enactment (post- vs. preMML enactment difference in proportions: 0.9–2.7%; p-values: < 0.0001–0.0026). No other prevalence of substance use changed significantly after MML enactment (p-values: 0.148–0.692). Tests of parallel paths (prior to passage) indicated no significant differences for all substances. 3.1. Sensitivity analyses Tests of differential changes in use due to MML after changes in federal policy towards MML states in 2009 indicated no new findings that were not already present looking at MML effects across the whole period, suggesting there was no differential change after 2008 over and above the change detected following MML enactment (Appendix S4, Table S3a). After adjusting for state-level yearly beer and cigarette tax levels, and heroin price levels in models of all substances, estimates of pre-law differences between MML and non-MML states were attenuated and no longer significant (Appendix S4, Table S3b). However, estimates of the effects of MML enactment on substance use within MML states did not change with the inclusion of these additional controls (Appendix S4, Table S3c). 4. Discussion We presented national evidence on the impact that enactment of medical marijuana laws had on marijuana use, binge drinking, cigarette use, non-medical prescription drug use, and illicit drug use among 8th, 10th, and 12th graders. The impact of MML enactment on substance use differed by grade. Among 8th graders, marijuana use, binge drinking, cigarette use, non-medical prescription opioid, amphetamine, and tranquilizer use, and other illicit drug use decreased following MML enactment. Among 10th graders, MML was unrelated to changes in use of any substance. Among 12th graders, non-medical prescription opioid use and cigarette use increased following MML enactment, although use of marijuana and other substances did not change. Our age-specific findings contrast with earlier studies that did not differentiate between early and late adolescents and failed to find relationships between MML enactment and use of various substances other than marijuana. The parallel decrease among 8th graders in marijuana use and binge drinking, cigarette, non-medical prescription drug and other illicit drug use following MML enactment suggests that changes associated with MML enactment may have had a complementary impact across substances in the youngest adolescents. Public messaging about the harms associated with adolescent use of marijuana may increase following MML enactment, and eighth graders may be more sensitive than others to public risk messages about substance use, particularly marijuana use. Prior work supports this hypothesis, as it showed that approximately one third of the decrease in marijuana use among 8th graders after MML enactment was explained by increased perceived harmfulness of marijuana in this group (Keyes et al., 2016). Further, parents of young adolescents may be more likely to monitor substance use following MML enactment. Prior studies show that parental influence is highest in early adolescence, waning as adolescents age (Guo et al., 2002; Latendresse et al., 2008; Scalici and Schulz, 2014). Thus, younger adolescents may be more receptive to parental messaging about the 66
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
was found among 12th graders. Our findings suggest that MML enactment may result in complementary decreases in marijuana and other drugs in early adolescence and in an increase in non-medical prescription opioid and cigarette use in late adolescence. Future studies should examine these questions with additional years of post-MML data and investigate the mechanisms that explain the impact that medical marijuana laws have on substances other than marijuana.
econometric evidence. East. Econ. J. 23, 253–276. Choo, E.K., Benz, M., Zaller, N., Warren, O., Rising, K.L., McConnell, K.J., 2014. The impact of state medical marijuana legislation on adolescent marijuana use. J. Adolesc. Health 55, 160–166. Crost, B., Guerrero, S., 2012. The effect of alcohol availability on marijuana use: evidence from the minimum legal drinking age. J. Health Econ. 31, 112–121. DiNardo, J., Lemieux, T., 2001. Alcohol, marijuana, and American youth: the unintended consequences of government regulation. J. Health Econ. 20, 991–1010. Guo, J., Hill, K.G., Hawkins, J.D., Catalano, R.F., Abbott, R.D., 2002. A developmental analysis of sociodemographic, family, and peer effects on adolescent illicit drug initiation. J. Am. Acad. Child Adolesc. Psychiatry 41, 838–845. Harper, S., Strumpf, E.C., Kaufman, J.S., 2012. Do medical marijuana laws increase marijuana use?: Replication study and extension. Ann. Epidemiol. 22, 207–212. Hasin, D.S., Wall, M., Keyes, K.M., Cerda, M., Schulenberg, J., O'Malley, P.M., Galea, S., Pacula, R., Feng, T., 2015. Medical marijuana laws and adolescent marijuana use in the USA from 1991 to 2014: Results from annual, repeated cross-sectional surveys. Lancet Psychiatry 2, 601–608. Hunt, P.E., Miles, J., 2015. The impact of legalizing and regulating weed: issues with study design and emerging findings in the USA. Curr. Top. Behav. Neurosci. 34, 173–198. Imbens, G.W., Wooldridge, J.M., 2009. Recent developments in the econometrics of program evaluation. J. Econ. Lit. 47, 5–86. Imtiaz, S., Shield, K.D., Fischer, B., Rehm, J., 2014. Harms of prescription opioid use in the United States. Subst. Abuse Treat. Prev. Policy 9, 43. Johnston, L.D., O'Malley, P.O., Miech, R.A., Bachman, J.G., Schulenberg, J.E., 2014. Monitoring the Future National Results on Drug Use: 1975–2013: Overview, Key Findings on Adolescent Drug Use Institute for Social Research. The University of Michigan, Ann Arbor. Johnston, L.D., O'Malley, P.M., Miech, R.A., Bachman, J.G., Schulenberg, J.E., 2017. Monitoring the Future National Survey Results on Drug Use, 1975–2016: Overview, Key Findings on Adolescent Drug Use Institute for Social Research. The University of Michigan, Ann Arbor, MI. Kandel, D., Kandel, E., 2015. The Gateway Hypothesis of substance abuse Developmental, biological and societal perspectives. Acta Paediatr. 104, 130–137. Kandel, D.B., 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, 447–457. Keyes, K.M., Wall, M., Cerda, M., Schulenberg, J., O'Malley, P.M., Galea, S., Feng, T., Hasin, D.S., 2016. How does state marijuana policy affect US youth? Medical marijuana laws, marijuana use and perceived harmfulness: 1991–2014. Addiction 111, 2187–2195. Kilmer, B., Everingham, S., Caulkins, J., Midgette, G., Reuter, P., Pacula, R., Burns, R., Han, B., Lundberg, R., 2014. What America’s users spend on illicit drugs: 2000–2010. In: RAND Report RR-534-ONDCP. Santa Monica, CA, February 2014. Kim, J., Santaella-Tenorio, J., Mauro, C., Wrobel, J., Cerdá, M., Keyes, K., Hasin, D.S., 2016. State medical marijuana laws and the prevalence of opioids detected among fatally injured drivers. Am. J. Public Health 106, 2032–2037. Latendresse, S.J., Rose, R.J., Viken, R.J., Pulkkinen, L., Kaprio, J., Dick, D.M., 2008. Parenting mechanisms in links between parents' and adolescents' alcohol use behaviors. Alcohol Clin. Exp. Res. 32, 322–330. Lynch, M.E., Ware, M.A., 2015. Cannabinoids for the treatment of chronic non-cancer pain: an updated systematic review of randomized controlled trials. J. Neuroimmune Pharmacol. 10, 293–301. Lynne-Landsman, S.D., Livingston, M.D., Wagenaar, A.C., 2013. Effects of state medical marijuana laws on adolescent marijuana use. Am. J. Public Health 103, 1500–1506. Martins, S.S., Mauro, C.M., Santaella-Tenorio, J., Kim, J.H., Cerda, M., Keyes, K.M., Hasin, D.S., Galea, S., Wall, M., 2016. State-level medical marijuana laws, marijuana use and perceived availability of marijuana among the general U.S. population. Drug Alcohol Depend. 169, 26–32. Meyer, R., Patel, A.M., Rattana, S.K., Quock, T.P., Mody, S.H., 2014. Prescription opioid abuse: a literature review of the clinical and economic burden in the United States. Popul. Health Manage. 17, 372–387. Miech, R.A., Johnston, L.D., O'Malley, P.M., Bachman, J.G., Schulenberg, J., 2015a. Monitoring the Future National Survey Results on Drug Use, 1975–2013: Volume I: Secondary School Students. Institute for Social Research, The University of Michigan, Ann Arbor. Miech, R.A., Johnston, L.D., O'Malley, P.M., Bachman, J.G., Schulenberg, J., 2015b. Monitoring the Future National Survey Results on Drug Use, 1975–2014: Volume I, Secondary School Students. Institute for Social Research, The University of Michigan, Ann Arbor. Miech, R.A., Johnston, L.D., O'Malley, P.M., Bachman, J.G., Schulenberg, J.E., 2016. Monitoring the Future National Survey Results on Drug Use, 1975–2015: Volume I, Secondary School Students. Institute for Social Research, The University of Michigan, Ann Arbor, MI. Moore, S.C., 2010. Substitution and complementarity in the face of alcohol-specific policy interventions. Alcohol Alcohol. 45, 403–408. Ogden, D., 2009. Memorandum for Selected United States Attorneys on Investigations and Prosecutions in States Authorizing the Medical Use of Marijuana. The United States Department of Justice Archives Accessed at: https://www.justice.gov/ archives/opa/blog/memorandum-selected-united-state-attorneys-investigations-andprosecutions-states. Orzechowski, Walker, 2015. In: The Tax Burden on Tobacco Volume 49, 1970–2014. Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Pacula, R., Powell, D., Heaton, P., Sevigny, E., 2013. Assessing the Effects of Medical Marijuana Laws on Marijuana and Alcohol Use: The Devil Is in the Details NBER
Contributors Cerdá designed the study, interpreted results, and wrote the manuscript; Wall designed the analytic plan; Sarvet, Feng and Wall conducted analyses and wrote the methods and results sections of the manuscript; Keyes and Galea participated in study design and provided input into interpretation of the study results and manuscript drafts; and Hasin provided critical input into study design, design of the analytic plan, interpretation of results, and reviewed manuscript drafts. All authors approved of the final manuscript. Role of funding source Nothing declared. Conflict of interest The authors have no conflicts of interest to declare. Acknowledgements This work was supported by the National Institutes of Health (R01DA034244, R01DA040924, K01DA030449, K01AA021511, T32DA031099) and by the New York State Psychiatric Institute. Funders had no role in study design, data collection, analysis, writing of the manuscript, or decision to submit to this journal. We thank Dr. Rosalie Pacula, Dr. Patrick O’Malley, and Dr. John Schulenberg for their invaluable feedback on this manuscript. This study was funded by R01DA034244 (Hasin), R01DA040924 (Cerdá), K01DA030449 (Cerdá), K01AA021511 (Keyes), T32DA031099 (Hasin) and by the New York State Psychiatric Institute (Hasin, Wall). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.drugalcdep.2017.10.021. References Anderson, D., Hansen, B., Rees, D., 2013. Medical marijuana laws, traffic fatalities, and alcohol consumption. J. Law Econ. 56, 333–369. Angrist, J.D., Krueger, A.B., 1999. Empirical strategies in labor economics. In: In: Ashenfelter, O., Card, D. (Eds.), Handbook of Labor Economics Vol. 3A. Elsevier, Amsterdam, The Netherlands, pp. 1277–1366. Angrist, J.D., Pischke, J.S., 2009. Mostly Harmless Econometrics: An Empiricist's Companion. Princeton University Press, Princeton. Bachhuber, M.A., Saloner, B., Cunningham, C.O., Barry, C.L., 2014. Medical cannabis laws and opioid analgesic overdose mortality in the United States, 1999–2010. JAMA Intern. Med. 174, 1668–1673. Bachman, J.G., Johnston, L.D., O'Malley, P.M., 1990. Explaining the recent decline in cocaine use among young adults: further evidence that perceived risks and disapproval lead to reduced drug use. J. Health Soc. Behav. 31, 173–184. Bachman, J.G., Johnston, L.D., O'Malley, P.M., Schulenberg, J.E., Miech, R.A., 2015. The Monitoring the Future Project After Four Decades: Design and Procedures (Monitoring the Future Occasional Paper No. 82). Institute for Social Research, The University of Michigan, Ann Arbor MI. Institute, Beer, 2016. Brewers Almanac. Beer Institute, Washington, D.C. Cameron, L., Williams, J., 2001. Cannabis, alcohol and cigarettes: substitutes or complements? Econ. Rec. 77, 19–34. Caulkins, J., Pacula, R., Arkes, J., Reuter, P., Paddock, S., Iguchi, M., Riley, J., 2004. The Price and Purity of Illicit Drugs: 1981 Through the Second Quarter of 2003. Office of National Drug Control Policy, Washington D.C. Chaloupka, F.J., Laixuthai, A., 1997. Do youths substitute alcohol and marijuana?: Some
67
Drug and Alcohol Dependence 183 (2018) 62–68
M. Cerdá et al.
Colorado compared to non-medical marijuana states: 2003–11. Drug Alcohol Depend. 140, 145–155. Schulenberg, J., Bachman, J.G., O'Malley, P.M., Johnston, L.D., 1994. High school educational success and subsequent substance use: a panel analysis following adolescents into young adulthood. J. Health Soc. Behav. 35, 45–62. Staff, J., Schulenberg, J.E., Maslowsky, J., Bachman, J.G., O'Malley, P.M., Maggs, J.L., Johnston, L.D., 2010. Substance use changes and social role transitions: proximal developmental effects on ongoing trajectories from late adolescence through early adulthood. Dev. Psychopathol. 22, 917–932. United States Census Bureau, 2014. Metropolitan and Micropolitan Statistical Areas Main. (https://www.census.gov/population/metro/. Accessed on April 15, 2014). Wall, M.M., Poh, E., Cerda, M., Keyes, K.M., Galea, S., Hasin, D.S., 2011. Adolescent marijuana use from 2002 to 2008 Higher in states with medical marijuana laws, cause still unclear. Ann. Epidemiol. 21, 714–716. Wall, M.M., Mauro, C., Hasin, D.S., Keyes, K.M., Cerda, M., Martins, S.S., Feng, T., 2016. Prevalence of marijuana use does not differentially increase among youth after states pass medical marijuana laws: commentary on and reanalysis of US National Survey on Drug Use in Households data 2002–2011. Int. J. Drug Policy 29, 9–13. Wen, H., Hockenberry, J.M., Cummings, J.R., 2015. The effect of medical marijuana laws on adolescent and adult use of marijuana, alcohol, and other substances. J. Health Econ. 42, 64–80. Williams, J., Liccardo Pacula, R., Chaloupka, F.J., Wechsler, H., 2004. Alcohol and marijuana use among college students: economic complements or substitutes? Health Econ. 13, 825–843. Yoruk, B.K., Yoruk, C.E., 2011a. The impact of minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use: evidence from a regression discontinuity design using exact date of birth. J. Health Econ. 30, 740–752. Yoruk, B.K., Yoruk, C.E., 2011b. The impact of minimum legal drinking age laws on alcohol consumption, smoking, and marijuana use: evidence from a regression discontinuity design using exact date of birth. J. Health Econ. 30, 740–752.
Working Paper. National Bureau of Economic Research, Cambridge, MA. Pacula, R.L., Bousetad, A.E., Hunt, P., 2014. Words can be deceiving: a review of variation among legally effective medical marijuana laws in the United States. J. Drug Policy Anal. 7, 1–9. Pacula, R., Powell, D., Heaton, P., Sevigny, E., 2015. Assessing the effects of medical marijuana laws on marijuana use: the devil is in the details. J. Policy Anal. Manage. 34, 7–31. Pacula, R.L., 1998. Does increasing the beer tax reduce marijuana consumption? J. Health Econ 17, 557–585. Patrick, M.E., Schulenberg, J.E., 2010. Alcohol use and heavy episodic drinking prevalence and predictors among national samples of American eighth- and tenth-grade students. J. Stud. Alcohol Drugs 71, 41–45. Powell, D., Pacula, R.L., Jacobson, M., 2015. Do Medical Marijuana Laws Reduce Addiction and Deaths Related to Pain Killers RAND Working Paper. Rand Corporation, Santa Monica, CA. Saffer, H., Chaloupka, F., 1999. Demographic differentials in the demand for alcohol and illicit drugs. In: Chaloupka, F., Grossman, M., Bickel, W., Saffer, H. (Eds.), The Economic Analysis of Substance Use and Abuse: An Integration of Econometric and Behavioral Economic Research. University of Chicago Press. Salomonsen-Sautel, S., Min, S.J., Sakai, J.T., Thurstone, C., Hopfer, C., 2014. Trends in fatal motor vehicle crashes before and after marijuana commercialization in Colorado. Drug Alcohol Depend. 140, 137–144. Scalici, F., Schulz, P.J., 2014. Influence of perceived parent and peer endorsement on adolescent smoking intentions: parents have more say, but their influence wanes as kids get older. PLoS One 9, e101275. Scavone, J.L., Sterling, R.C., Van Bockstaele, E.J., 2013. Cannabinoid and opioid interactions: implications for opiate dependence and withdrawal. Neuroscience 248, 637–654. Schuermeyer, J., Salomonsen-Sautel, S., Price, R.K., Balan, S., Thurstone, C., Min, S.J., Sakai, J.T., 2014. Temporal trends in marijuana attitudes, availability and use in
68