Abstract
Background:
Ineffective cannabis regulatory frameworks such as prohibition have sparked interest in alternative solutions to reduce individual and societal harms. While it has been suggested that the recreational legalization process has yielded early successes, the relatively recent implementation of the novel policies has provided a modest time frame for a truly thorough establishment and assessment of key population-level indicators. The following systematic review focuses on identifying the downstream public health sequelae of cannabis legalization policies, including parameters such as cannabis consumption rates, hospitalization rates, vehicular accidents and fatalities, criminal activity, and suicidal behaviors, as well as other substance use trends.
Methods:
An exhaustive search of the MEDLINE and Google Scholar databases were performed to identify high-quality (1) longitudinal studies, which (2) compared key public health outcomes between regions which had and had not implemented recreational cannabis legalization (RML) policies, (3) using distinct databases and/or time frames. Thirty-two original research articles were retained for review.
Results:
Adult past-month cannabis consumption (26+ years) seems to have significantly increased following RML, whereas young adult (18–26 years) and adolescent (12–17 years) populations do not show a significant rise in past-month cannabis use. RML shows preliminary trends in increasing service use (such as hospitalizations, emergency department visits, or poisonings) or vehicular traffic fatalities. Preliminary evidence suggests that RML is related to potential increases in serious/violent crimes, and heterogeneous effects on suicidal behaviors. While the research does not illustrate that RML is linked to changing consumptions patterns of cigarette, stimulant, or opioid use, alcohol use may be on the rise, and opioid prescribing patterns are shown to be significantly correlated with RML.
Conclusion:
The current data supports the notion that RML is correlated with altered cannabis consumption in adults, potentially increased criminal activity, and a decline in opioid quantities and prescriptions provided to patients. Future work should address additional knowledge gaps for vulnerable populations, such as individuals with mental health problems or persons consuming cannabis frequently/at higher THC doses. The effects of varying legalization models should also be evaluated for their potentially differing impacts on population-level outcomes.
1. Introduction
It has been suggested that international cannabis prohibition mandates have failed to achieve key goals such as harm reduction, increased prevention and treatment, and have instead generated negative consequences, including increased contributions toward global disease burden over time (1, 2) and exacerbating social inequalities through disproportionate impacts on people of color (3, 4). As such, ineffective regulatory frameworks such as cannabis prohibition have sparked interest in alternative solutions to reduce individual and societal harms (5). Recently, countries such as Uruguay, Canada, and certain states of the United States have enacted recreational marijuana laws (RML). While the overarching frameworks vary between locations, certain RML laws propose to enhance the protection of vulnerable populations; strengthen health education programs; provide access to quality-controlled cannabis; and enable the close monitoring of public health outcomes through these new regulatory frameworks (6). Despite these beneficial aspirations, the enactment of cannabis legalization policies remains hotly questioned. Several thought leaders have denoted an opposition against the hasty implementation of legalization policies, warning against the escalation of use and related harms among the most vulnerable populations, such as youth (7), an increase in driving under the influence, or increased risk of using other drugs, including harder drugs (8). Despite having collected close to a decade of research evidence, we have yet to determine unequivocal findings to support either side of the discussion relating to recreational cannabis legalization laws.
Regarding the impacts of RML on general cannabis consumption, studies conducted in several states across the US have found discrepant results. Initial evidence in adult populations have found increases in cannabis use over time (9–11), decreases of use (12), or even a lack of change altogether (13). Youth populations also demonstrate varying effects, with evidence for overall exacerbated use (14), diminished use (15), or show no impact (16). Of importance, the largest source of data collected on consumption metrics relate to past month cannabis use, few have investigated frequent use, and sparse have examined trends in cannabis use disorder. As marijuana consumption trends may vary over time, using outcome metrics such as past month marijuana use may not provide an accurate reflection of true individual consumption trends over time. This may entail an over or under-estimation in the number of individuals at highest risk of adverse health consequences associated with cannabis use.
Beyond simple consumption patterns, several other population parameters have been monitored over time to determine the impacts of RML. Seminal work developed by Lake and colleagues highlighted the use of 28 indicators to monitor RML effects, including public safety measures such as vehicle injuries/fatalities and crimes; other substance use and overdose trends; and hospitalizations related to cannabis use (17). Research has suggested potential surges in vehicular fatalities and crimes–specifically, increases in crimes such as burglary, larceny, violent assaults, and so forth (18, 19). Preliminary evidence points to potential increases in healthcare service use related to cannabis (20–22). However, these initial assumptions seem to be skewed by an overrepresentation of increases in specific states, such as Colorado. Other substance use, such as alcohol, tobacco, or illicit drugs use, has seen trends of increases (23), decreases (11), and no changes (24).
The discrepant results in the current literature can be partially attributed to the methodology and sampling used in the research studies. Most are performed in a single location, thus omitting trends over the same time course in a comparator location. Thus, such studies may highlight changes that are not necessarily related to legalization per se, but may actually reflect other unspecific factors, such as the perception of harms, for instance. Other studies have used a comparator location but have collected data only post-legalization. These are both critical methodological aspects to consider, as certain locations may be already experiencing upticks in cannabis consumption prior to legalization, thus post-legalization patterns across regions should be interpreted with caution. Certain studies collect only a single datapoint prior to RML implementation, or only a single datapoint post RML implementation, providing little information on the trends already occurring prior to RML implementation, as well as little information into long-term effects if studying a short post-RML period. Considering the limitations of studies using these methodological strategies, it would be beneficial to update the current state of the knowledge of RML impacts on population health metrics using longitudinal comparative studies.
As such, the following systematic review seeks to shed light on the clouded debate of the impacts of RML on key public health metrics. Importantly, we aim to perform a systematic review of studies which will provide a high level of insight: research articles which follow RML and non-RML states, with a baseline assessment of public health trends prior to RML implementation. This systematic review will focus on key metrics outlined by Lake et al. (17) to examine if RML implementation affects youth/young adult/adult cannabis consumption, service use, vehicular crashes/fatalities, crimes (unrelated to cannabis possession), and other drug use. The evidence provided in this review will help provide recommendations for future cannabis legalization policy research.
2. Methods
2.1. Search strategy
The search strategy was completed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) standards (25). Potential articles were discovered through an exhaustive search of the MEDLINE database and Google Scholar for studies expanding from January 1, 2012–which corresponds to the year where recreational cannabis was legalized for the first time in Colorado and Washington–until February 1, 2022. The following terms were employed to direct our search for research articles: (“marijuana/marihuana,” “cannabis,” “illicit), the independent factor (“legalization,” “recreational”) and the outcomes of interest (“use,” “consumption,” “hospital*”, “traffic,” “crime,” “alcohol,” “stimulant,” “opioid,” “nicotine”). Cross-referencing of previous systematic reviews on the topic was also performed.
2.2. Eligibility criteria
Longitudinal observational studies were retained for the purposes of this systematic review. Specifically, we retained studies that: (i) had a baseline assessment (pre) prior to the implementation of recreational cannabis legalization, and a subsequent assessment (post) at least 6 months after the implementation of RML.; and (ii) which also longitudinally assessed at least one comparator location (control) which did not undergo RML. Of note, in some article, the same subjects were investigated over time, while in others, multiple measures were acquired over time in different samples of persons living within a state.
2.2.1. Exclusion criteria
In addition to the above-referenced criteria, studies were excluded if they evaluated medical cannabis legalization. We omitted studies which focused on solely on the impact of recreational cannabis legalization on arrests for possession of cannabis. Studies in languages other than English were also excluded. We did not retain studies that lacked a comparison group, or studies that did not have at least one pre-legalization evaluation and one post-legalization evaluation. There was an important number of publications which utilized overlapping databases and/or time points to study the effects of RML. As such, for all overlapping research initiatives, M.A. and S.P. identified the studies used for primary analyses purposes, which provided the latest data, included the highest number of participants and/or the highest number of states, and provided the longest follow-up period. Any overlapping studies which investigated single locations were retained for secondary analysis purposes when these studies reported data on specific outcomes that had not been reported in the primary analyses (example: specific effects in particular locations). The final decision on the inclusion and exclusion of studies was determined by consensus between M.A. and S.P.
2.3. Data extraction and quality assessment
The following information was extracted by two independent authors (M.A. and I.Z.): (1) type of population studied (including sample size (if available), average age, sex ratio); (2) RML locations studied; (3) non-RML comparator locations; (4) years assessed; (5) data source; (6) outcome measures analyzed; (7) confounding factors controlled for/considered in the analyses; (8) statistical models used in the analyses; (9) overarching results.
A quality assessment was then performed on the retained articles using an adapted version of the Newcastle - Ottawa Quality Assessment Scale for cohort studies (26). Briefly, studies were rated on strength of sampling selection, comparability, outcome, and follow-up time. As per the tool, studies were rated using a 3-point system (0–2 points), and accumulated scores on the 7 rated items qualified them as having either weak (0–4), moderate (5–9) or strong (10–14) reporting strength.
3. Results
3.1. Study selection
Out of 3645 studies identified in the database search, 125 articles underwent full-text screening, whereby 93 articles were excluded, predominantly because they were not longitudinal in nature (26), did not include a comparator group (19), or the database and time frame used in the analysis overlapped with another study retained for review (20). Thirty-two unique articles stemming from this database search were included in the primary analyses (Figure 1 PRISMA flow chart).
Figure 1
3.2. Study characteristics
Studies investigating the effects of cannabis legalization on health outcomes are summarized in Table 1. Most studies included in the review investigated multiple locations which had or had not implemented RML. While most studies accounted for a large number of covariates, including age, sex, socioeconomic status, prior education, and prescription drug monitoring programs, it should be noted that a few publications failed to account for more than a few basic confounders.
Table 1
| Study | Year | Population | Location (s) studied | Comparator location (s) | Years assessed | Data source | Outcome (s) of interest | Ascertainment of outcome (s) of interest | Brief main findings of the impacts of RML |
|---|---|---|---|---|---|---|---|---|---|
| Alcocer | 2020 | All persons in participating states | CO | 32 non-RML states | 1999–2017 | WONDER | Opioid mortality | Opioid overdose mortality rate per 100,000 in population | No significant difference |
| Alley | 2020 | 18–26-year-old college students | All RML states | All non-RML states | 2008–2018 | NCHA-II | Other drug use | Self-reported use | Decreased odds of binge drinking, no significant difference for other drug use |
| Bae | 2019 | 18–26-year-old college students | 7 RML states | 41 non-RML states | 2008–2018 | NCHA-II | Cannabis use | Self-reported past-month use (any), frequent use (>20 uses in last month) | Increased odds of cannabis use |
| Benedetti | 2021 | Adult drivers | All RML states | All non-RML states | 2013-2017 | TSCI | Driving under the influence of cannabis | Self-reported past-year driving within 1 h of marijuana use | No significant difference |
| Bhave | 2020 | All persons in participating states | CO, OR, NV, WA | Synthetic control | 2012–2017 | Retail scanner data from A.C. Nielsen | Nicotine use | Weekly cigarette sales in packs | Increased odds of nicotine use |
| Cerdà | 2020 | Persons 12+ in participating states | CO, WA, AK, OR | All non-RML states | 2008–2016 | NSDUH | Frequent use in the past month, past-year CUD overall | Self-reported past-month use, frequent use (>20 days or more of use in the past month), past-year prevalence of CUD (instrument that assessed symptoms corresponding to DSM-IV criteria) | Group 12–17 years: No increase in frequent use; increased past-year CUD prevalence Group 18–26 years: No difference for any outcome Group 26+ years: Increase in both outcomes |
| Cerdà | 2017 | High school students | CO, WA | 45 non-RML states | 2010–2015 | MTF | Cannabis use | Self-reported past-month use (any) | Increased odds of cannabis use in youth in Washington (grades 8–10) |
| Chan | 2020 | All persons in participating states | All RML states | Non-RML states | 1999–2017 | NVSS | Opioid mortality | Means of opioid mortality rates (per 100,000 population) | Decreased odds of opioid mortality |
| Coley | 2021 | Youth | 6 RML states | Non-RML states | 2015 and 2017 | YRBS | Cannabis use, other drug use | Past-month marijuana, alcohol, cigarette, e-cigarette use (number of times) | No significant difference for cannabis use, small increased odds of cigarette use |
| Delling | 2019 | Inpatients in participating states | CO | NY, OK | 2010–2014 | HCUP | Service use | Total number of hospitalizations, length of inpatient stay, healthcare costs, hospitalization related to multi vehicle collisions | Increased odds of cannabis-related service use |
| Doucette | 2021 | All persons in participating states | CO, WA | Synthetic control | 2000–2018 | NCHS | Suicide rate | Annual, state-level deaths by suicide | Increased odds of death by suicide in Washington, no significant difference in Colorado |
| Drake | 2021 | All persons in participating states | CA, ME, MA, NV | Non-RML states | 2011–2017 | HCUP | Service use | Log opioid-related ED visit rates per 100,000 population in states | No significant difference in service use |
| Kamer | 2020 | All persons in participating states | CO, WA, OR, and AK | 20 non-RML states | 2008–2018 | FARS | Traffic fatality rates | Traffic fatality rates | Increased odds of traffic fatalities |
| Kerr | 2017 | 18–26-year-old college students | OR | 6 non-RML states | 2014 and 2016 | HMS | Cannabis use, other drug use | Self-reported past-month use (any) of marijuana, cigarette and frequency of heavy alcohol | Increased odds of cannabis use only in recent heavy alcohol users |
| Kim | 2021 | All persons in participating states | AK, CA, CO, DC, MA, ME, NV, OR, WA | Non-RML states | 2004–2017 | NSDUH | Cannabis use, alcohol use | Self-reported past month marijuana use (any), past month alcohol use (any) | Increased odds of cannabis use only in adults |
| Kropp Lopez | 2020 | All persons in participating states | CO | UT, MD | 2007–2017 | DEA ARCOS | Opioid prescriptions | Prescription opioid distribution for OUD treatment (oral morphine milligram equivalents) | Significantly increased oral MME |
| Lane | 2019 | All persons in participating states | CO, WA, OR | AL, AR, FL, GA, IN, IA, KY, MI, MS, MO, ND, NC, SC, SD, TN, TX, VA, WV, WI | 2009–2016 | WONDER | Traffic fatalities | Monthly traffic fatalities rates per million residents | Decreased odds of traffic fatalities only in Washington |
| Lopez | 2021 | Medicaid enrollees in participating states | AK, CA, CO, DC, MA, ME, NV, OR, WA | Non-RML states | 2013–2017 | Medicare Part D Prescription Drug Event database | Opioid prescriptions by orthopedic surgeons | Annual aggregate daily doses of all opioid medications (excluding buprenorphine) prescribed by orthopedic surgeons in each US state (and DC) | No association between RML and opioid prescriptions |
| Lu | 2018 | All persons in participating states | CO, WA | Non-RML states | 1999–2016 | FBI's UCR | Crimes | Monthly crime rates: violent, property, aggravated assault, auto theft, burglary, larceny, and robbery rates | No significant difference for violent crimes, only short-term increase in property crimes in Colorado |
| Lu | 2020 | All Medicaid enrollees in participating states | All RML states | Non-RML states | 2005–2019 | Consumer Expenditure Interview Survey | Alcohol use | Alcohol expenditures | Increase in alcohol use |
| Masonbrink | 2021 | Youth | CA, CO, DC, MA, WA | AZ, CT, DE, FL, IL, IN, MD, MN, MO, NJ, NY, OH, PA, UT | 2008–2019 | Inpatient Essentials database | Service use | Annual incidence of a hospitalization with a cannabis-related diagnosis (i.e., cannabis-related hospitalization) | Increased odds of service use |
| Matthay | 2021 | All Medicare enrollees in participating states | All RML states | All non-RML states | 2003–2017 | Clinformatics Data Mart; Optum Inc | Self-harm, Crimes | Claims for self-harm and assault injuries based on International Classification of Diseases codes | Increased odds of self-harm injury in males <21 years old, increased odds of physical assault for males and females <21 years old |
| McMichael | 2020 | Patients at outpatient pharmacies in participating states | All RML states | MCL and no marijuana law states | 2011–2018 | Symphony Health's IDV® dataset | Opioid prescriptions | (1) the total number of MMEs prescribed by each provider, (2) the total days' supply prescribed by each provider, (3) the number of unique patients to whom each provider prescribed opioids, (4) the percentage of a provider's patients receiving any opioids, and (5) whether a provider prescribed any opioids. | Significantly decreased oral MME |
| Mennis | 2021a | All young adults in participating states | All RML states | All non-RML states | 2008–2017 | SAMHSA (TEDS-A), NSDUH | Service use | Treatment admissions for cannabis (number of young adult's treatment admissions/young adult population), | Decreased odds of service use |
| Mennis | 2021b | Youth | CO, WA | Non-RML states | 2008–2017 | SAMHSA TEDS-A | Service use | Mean observed treatment admissions rates (per 10,000 population) | Increased odds of service use in Colorado |
| Miller | 2017 | 18–26-year-old college students | WA | National average | 2005–2015 | NCHA, NSDUH | Cannabis use, other drug use | Self-reported past-month use | Increased odds of cannabis use, no significant difference for other drugs |
| Shi | 2020 | All persons in participating states | All RML states | Non-RML states | 2010–2017 | USNPDS | Service use | Cannabis exposures reported to the US National Poison Data System | Increased odds of service use (unintentional exposures and exposures without medical consequences) |
| Shi | 2019 | All Medicaid enrollees in participating states | CO, WA, AK, DC, OR, CA, MA, ME, NV | HI, MI, MT, NM, RI, VT | 2010–2017 | Medicaid State Drug Utilization Data | Opioid prescriptions | (1) Number of opioid prescriptions, (2) Total doses of opioid prescriptions (in quantity of MME) | Significantly decreased oral MME |
| Veligati | 2020 | All persons in participating states | All RML states | All non-RML states | 1990–2016 | NIAAA, AEDS, Tax Burden on Tobacco | Other drug use | Per capita consumption of alcohol and cigarettes as measured by state tax receipts | No significant difference |
| Wallace | 2020 | 18–26-year-old college students | CO | National Average | 2011–2015 | NCHA | Cannabis use | Self-reported 30-day use of cannabis | Increased odds of cannabis use |
| Weinberger | 2022 | All persons in participating states | All RML states | Non-RML states | 2004–2017 | NSDUH | Cannabis use, nicotine use | Self-reported past-month cannabis use (any) | Decreased odds of cannabis use in youth, increased odds of cannabis use in adults, decreased odds of nicotine use in youth |
| Wen | 2021 | Patients with employer- sponsored health insurance | All RML states | Non-RML states | 2009–2015 | Truven Health MarketScan Commercial Claims and Encounters Database | Opioid prescriptions | Monthly MME per enrollee | Significant decreased of oral MME |
Overview of longitudinal studies investigating the impact of cannabis legalization public health outcomes used for primary analyses.
RML, Recreational Marijuana Legalization; WONDER, Wide-ranging ONline Data for Epidemiologic Research; NCHA, National College Health Assessment; TSCI, Traffic Safety Culture Index; MTF, Monitoring the Future; NVSS, National Vital Statistics System; YRBS, Youth Risk Behavior Surveillance; HCUP, Healthcare Utilization Project; NCHS, National Center for Health Statistics; FARS, Fatality Analysis Reporting System; HMS, Healthy Minds Study; NSDUH, National Survey on Drug Use and Health; DEA ARCOS, Drug Enforcement Administration Automation of Reports and Consolidated Orders System; FBI's UCR, Federal Bureau of Investigation Uniform Crime Reporting; SAMHSA TEDS-A, Substance Abuse and Mental Health Services Administration Treatment Episode Data Set; USNPDS, US National Poison Data System; NIAAA, National Institute on Alcohol Abuse and Alcoholism; AEDS, Alcohol Epidemiologic Data System; MME, milligram morphine equivalent.
3.3. Study quality and reporting strength
As described in Table 2, the range of scores from the extracted articles varied greatly (between 5 and 12), with an average overall strong quality score of 10.25. As a quality score of 10 and above is considered as methodologically robust, 25 studies of the total 32 were deemed as good-quality evidence to accurately depict the relationships between RML and the selected population-level outcomes. Overall, the selected samples were representative of the targeted population; the nature of the studies yielded large sample sizes including thousands of persons. The intervention and comparator locations were clearly defined and represented a large number of states. The outcomes included relatively objective observable outcomes, mostly from government-mandated databases. A wide range in the years assessed was noted, whereby studies had a follow-up period of 4–18 years post-baseline assessment.
Table 2
| First Author | Year | Representativeness | Sample Size | Non-respondents | Ascertainment | Comparability | Assess outcome | Statistical test | Follow-up time frame | Score |
|---|---|---|---|---|---|---|---|---|---|---|
| Alcocer | 2020 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| Alley | 2020 | 1 | 1 | 0 | 2 | 2 | 2 | 1 | 2 | 11 |
| Bae | 2019 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| Benedetti | 2021 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 1 | 11 |
| Bhave | 2020 | 1 | 1 | 0 | 2 | 2 | 2 | 1 | 1 | 10 |
| Cerdà | 2020 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| Cerdà | 2017 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 1 | 11 |
| Chan | 2020 | 1 | 1 | 1 | 2 | 2 | 1 | 1 | 2 | 10 |
| Coley | 2021 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 0 | 10 |
| Delling | 2019 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 1 | 11 |
| Doucette | 2021 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| Drake | 2021 | 1 | 1 | 1 | 2 | 1 | 0 | 1 | 1 | 8 |
| Kamer | 2020 | 1 | 1 | 0 | 2 | 1 | 2 | 1 | 2 | 10 |
| Kerr | 2017 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 0 | 10 |
| Kim | 2021 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| Kropp Lopez | 2020 | 1 | 1 | 1 | 1 | 0 | 2 | 0 | 2 | 8 |
| Lane | 2019 | 1 | 1 | 0 | 2 | 0 | 2 | 1 | 1 | 8 |
| Lopez | 2021 | 1 | 1 | 0 | 1 | 2 | 2 | 1 | 1 | 9 |
| Lu | 2018 | 1 | 1 | 1 | 2 | 0 | 2 | 1 | 2 | 10 |
| Lu | 2020 | 1 | 1 | 1 | 2 | 1 | 2 | 1 | 2 | 11 |
| Masonbrink | 2021 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| Matthay | 2021 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| McMichael | 2020 | 1 | 1 | 1 | 2 | 1 | 2 | 1 | 1 | 10 |
| Mennis | 2021b | 1 | 1 | 1 | 1 | 1 | 2 | 1 | 1 | 9 |
| Mennis | 2021a | 1 | 1 | 1 | 2 | 1 | 2 | 1 | 1 | 10 |
| Miller | 2017 | 1 | 1 | 0 | 2 | 2 | 2 | 1 | 2 | 11 |
| Shi | 2020 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 1 | 11 |
| Shi | 2019 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 1 | 11 |
| Veligati | 2020 | 1 | 1 | 1 | 2 | 2 | 2 | 1 | 2 | 12 |
| Wallace | 2020 | 1 | 1 | 0 | 1 | 0 | 1 | 1 | 0 | 5 |
| Weinberger | 2022 | 1 | 1 | 1 | 2 | 1 | 2 | 1 | 2 | 11 |
| Wen | 2021 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 8 |
Bias assessment of longitudinal studies investigating the impact of cannabis legalization public health outcomes used for primary analyses.
3.4. Results of study outcomes
3.4.1. Adult consumption
One research article examined the impact of RML on adult (26 years+) past-month cannabis use (11). Using the National Survey on Drug Use and Health (NSDUH) across 11 states. The authors determined that RML was associated with an increase in past-month adult consumption. One study evaluated past-month frequent use, as well as past-year CUD prevalence in adults using the NSDUH across 4 RML locations: Colorado, Washington, Alaska, and Oregon (10). The data yielded a significant increase in past-month frequent use (from 2.13 to 2.62%), as well as an increase in past-year CUD (from 0.90 to 1.23%).
3.4.2. Young adult (18–26) consumption
Three publications assessed the impact of recreational marijuana legalization on young adult past-month cannabis use in multiple RML states (9, 11, 27). While two studies demonstrate a lack of effect of RML, Bae and Kerr (9) found that college students in states with legalized recreational cannabis use had an increased prevalence of past-month use [adjusted Odds Rations (OR) of 1.23]. Three additional studies were retained for secondary analysis purposes and investigated the effects of RML specifically in in Colorado, Oregon, and Washington (28–30). In all three cases, RML was linked to increased past-month cannabis use in young adults. Regarding frequent use, Bae and Kerr (9) has found an increased adjusted Odds Rations (OR) of 1.18, whereas Cerdà et al. (10) failed to find evidence of increased past-month frequent use, or past-year prevalence of CUD, among young adults.
3.4.3. Youth (12–17)
Three primary articles investigated past-month cannabis use in adolescents (1, 31, 32). While Kim et al. (11) found a decrease in past month use, Coley et al. (32) did not find evidence for an increase or decrease in use, and Cerdà et al. (31) only found increases in past-month use of eighth and tenth graders. One additional study was retained for secondary analysis purposes and demonstrated that RML was associated with heightened past-month use in Alaskan youth (17). Cerdà et al. (10) did not report increases in past-month frequent use, however, did denote an increase in past-year CUD prevalence in youth (OR 1.25; 95%, confidence interval (CI) 1.01–1.55).
3.4.4. Healthcare-related service use
Four articles studied RML effects and service use, including cannabis-related hospitalizations, emergency department visits and reported cannabis exposures (33–36). Three studies denoted heightened service use in association with RML status, whereas Mennis et al. (37) found a decrease in cannabis-related treatments admissions in young adults in seven legalized states.
3.4.5. Multi-vehicle collisions, traffic fatalities or driving under the influence
Four studies assessed traffic-related accidents, injuries, and driving while intoxicated according to recreational legalization status. Delling et al. (33) extracted multi-vehicle collision data from the Healthcare Cost and Utilization Project database for the state of Colorado and found a significant impact of RML. These findings were echoed by Kamer et al. (38), who used Fatality Analysis Reporting System (FARS) data to demonstrate a link between a doubling in traffic fatality rates and RML in Colorado, Washington, Oregon, and Alaska. Lane et al. (39) utilized the Centers for Disease Control and Prevention Wide-ranging ONline Data for Epidemiologic Research (CDC WONDER) database to show that Washington state experienced an increase in traffic fatalities, whereas Colorado and Oregon did not. While no information on location is provided, Benedetti et al. (40) extracted data from the Traffic Safety Culture Index (TSCI) and found no effect of RML status on driving while under the influence.
3.4.6. Crime
Two studies evaluated the effects of RML on crimes excluding arrests for marijuana possession. Lu et al. (18) reviewed data extracted from the Federal Bureau of Investigation's Uniform Crime Reporting Program for the states of Colorado and Washington. Between 1999 and 2016, the authors concluded that violent crimes did not significantly increase in either state due to RML, however certain property crimes rates were significantly heightened post-legalization. In Colorado, larceny seemed to drive property crime rate increases, whereas in Washington, rates of burglaries and aggravated assaults were predominantly affected. As well, Matthay et al. (41) used Clinformatics data to determine that RML status was associated was linked significant increases in assaults of persons younger than 21 years of age.
3.4.7. Alcohol use
A series of seven articles investigated the association between recreational cannabis legalization and alcohol use. Three studies provide evidence for an increase in alcohol consumption in RML states, as reported by the Consumer Expenditure Interview Survey (42), HCUP data (33), and the ACHA-National College Health Assessment II (NCHA-II) (23), across RML states. Curiously, 3 studies failed to show an association between legalization and alcohol use (24, 27, 32). One research article demonstrated a decrease in alcohol use following legalization in Colorado across 11 US RML states (11).
3.4.8. Cigarette use
Six studies were retained to evaluate cigarette consumption in response to RML implementation in the US. Most studies failed to find an effect of recreational legalization on tobacco use, as per data derived from the Healthy Minds Study database in Oregon youth/young adults (27), the NCHA-II in several RML states (23), the Tax Burden on Tobacco data (24), and the Youth Risk Behavior Surveys (YRBS) in six RML locations (32). Nonetheless, longitudinal data has also found evidence of potential increases in cigarette use in Colorado, Oregon, Nevada, and Washington (43), as well as decreases in use in almost 10 states (15).
3.4.9. Opioid metrics
3.4.9.1. Opioid use
One original research article evaluated the impact of RML on self-reported opioid use. Alley et al. (23) collected responses to self-reported past-month opioid use from the NCHA-II from over 800 000 college students and determined that legalization status was not associated with opioid consumption in young adults (23).
3.4.9.2. Opioid-related service use
Two research studies assessed the effects of RML on opioid-related service use. Drake et al. (44) examined the opioid-related emergency department visit rates per 100,000 population in California, Maine, Massachusetts, and Nevada using the Healthcare Cost and Utilization Project (HCUP) database. While they found initial decreases in opioid-related ED visits in RML states, the effects were abolished by the end of the study period (44). Mennis et al. (37) explored the impact of RML in adolescents and young adults (12–24 years of age) in Washington and Colorado compared to non-RML states regarding opioid-related treatment admission from the Substance Abuse and Mental Health Services Administration (SAMHSA) Treatment Episode Dataset–Admissions database. The first difference-in-difference analyses determined that RML was not linked to treatment admissions. However, when analyzed separately, Colorado yielded a significant increase in opioid-related treatment admissions, while Washington demonstrated a significant decrease in opioid-related treatment admissions (37). In sum, the current data do not provide sufficient evidence to support the notion that RML is correlated with alterations in opioid-related service use.
3.4.9.3. Opioid prescriptions
Five studies assessed the impact of RML on opioid prescriptions. McMichael et al. (45) collected information from over 1 billion individual prescriptions derived from the Symphony Health's IDV® (Integrated Dataverse) dataset of patients of outpatient pharmacies in all RML and non-RML states. Using difference-in-difference analyses, the authors determined that recreational cannabis legalization corresponded with a significant decrease in the quantity of opioids (in morphine milligram equivalents or MMEs) prescribed to patients (45). This significant finding was echoed throughout three of the other four studies which investigated MMEs as their main outcome of interest. Wen et al. (46) retrieved MME data from patients with employer-sponsored health insurance between RML and non-RML states and found a significant 13% reduction in monthly MMEs in RML state patients (46). Shi et al. (47) extracted MME doses for Medicaid patients of RML vs. non RML states using the Medicaid State Drug Utilization Data and yielded total MME dose reductions for Schedule 3 opioids by 30% in RML states (47). Lopez et al. (48) used an indirect measure to investigate opioid use, through prescription opioid distribution numbers for opioid use disorder treatment (in MME equivalents) and found a reduction in this outcome in Colorado and Maryland–but not for the state of Utah. Only one study failed to establish a significant association between recreational cannabis legislation and MMEs (48). Shi et al. (47) also determined that the average number of opioid prescriptions written by physicians declined by 32% with the implementation of recreational cannabis legalization policies. In sum, the current data supports the notion that RML is correlated with a change in opioid prescription practices, including a reduction of average MMEs prescribed to patients, as well as the number of prescriptions provided to patients.
3.4.9.4. Opioid-related deaths
One study observed trends in opioid-related deaths pre and post recreational cannabis legalization. Alcocer et al. (49) investigated a wide temporal range (1999–2017) in Colorado and extracted data from the CDC's WONDER database and found no evidence of RML effects on opioid overdose mortality rates per 100,000 population when compared to a synthetic control model (i.e., pooled data from multiple donor states to provide an accurate comparator) (49).
3.4.10. Suicide
Two studies evaluated the effects of RML on deaths by suicide. Matthay et al. (41) examined claims for self-harm injuries based on International Classification of Diseases codes from all RML states using the Clinformatics Data Mart. The analyses yielded a significant association between heightened rates of self-injury for your males in states that had legalized recreational cannabis (41). Doucette et al. (50) performed a more restricted analysis on data derived from Washington State and Colorado and found heterogeneous effects of RML. Specifically, Washington state youth and young adults demonstrated a link between deaths by suicide and RML status, whereas Colorado residents did not (50).
4. Discussion
The following review aimed to evaluate the evidence linking population-level health metrics with the implementation of recreational legalization policies. Through a literature review, we identified 32 studies which investigated key metrics, such as cannabis consumption, healthcare-related service use, crime, traffic crashes/fatalities, suicidal behaviors, and other drug use. Due to our stringent methodological criteria, all included studies in the review were performed in the United States of America. Overall, the evidence illustrates a lack of effect of RML on adolescent and young adult populations, and a possible increase in service use, vehicle related crashes and fatalities, and alcohol consumption. The data has not signaled an increase in nicotine use; however, it does correlate with a decrease in opioid prescriptions. It is also important to highlight the dearth of research with controlled designs related to the impact of recreational legalization of marijuana on criminality (excluding drug possession-related crimes), as well as deaths due to opioid overdoses or suicide.
To date, the evidence suggests moderate increases in past-month cannabis use in adult populations and no increase in adolescents or young adults (11). These data illustrate two central points. First, the lack of clearly detrimental effects seen in adolescence and early adulthood years is important considering that one of main concerns that was raised prior to RML was that such policy change could contribute to the development of ancillary impairments caused by increased cannabis use during early periods of brain maturation (51, 52). Second, the observation of increased consumption in adults is based on one single study which met the aforementioned methodological criteria. There is therefore a need to replicate these results in future research. As well, the results yielded from current studies refer to past-month use, which is an outcome that cannot differentiate between adult populations that are occasionally experimenting with cannabis from populations that are transitioning from occasional use to heavy use or cannabis use disorder. Early work by Montgomery et al. (53) has discerned potential increases in newly onset cannabis use in the adult population following RML, but not the underage population, suggesting heightened experimenting among adults who may not have otherwise tried cannabis, however these findings should be replicated before deemed as conclusive (53).
The data included in this review which evaluated frequent past-month cannabis use, and past-year CUD prevalence, across the age groups, was mainly extracted from a single study, and thus caution must be exerted when interpreting the findings. Nonetheless, the preliminary evidence points to increased frequent use, and CUD prevalence in the adult population. This evidence could potentially indicate a heightened rate of transition from occasional use to problematic use; acute monitoring of the situation is warranted in future studies. Among the research articles that did not meet the inclusion criteria (i.e., not comparative and/or longitudinal studies), the collected evidence is heterogeneous however does point to a potential increase as well [for a review, see (54)]. In the young adult population, the authors found no evidence of increased frequent use or problematic use, which may suggest limited enduring effects in this age group. Interestingly, in youth, authors have failed to establish altered frequent past-month consumption, however early evidence from the comparative longitudinal study by Cerdà et al. (10) highlights a heightened prevalence of past-year CUD. As cannabis potency has been continuously increasing over the last decades (55) is posited to be associated with increased adverse health outcomes, and as heavy cannabis use is associated with more harm to psychological and physical health than occasional use (56). Comparative and longitudinal studies on this issue are required in the future to evaluate the enduring impact of recreational cannabis legalization on youth marijuana consumption and health.
With respect to post-legalization trends of motor vehicles crashes, the evidence is mixed, however indicates potential increases. Specifically, the studies encompassing early adopter regions, such as Colorado, Washington, Oregon, and Alaska have shown increases in traffic crashes/fatalities (33, 38). Other studies including a vaster range of states show more divergent effects (39, 40). It may be posited that the differing modalities of RML may be associated with differential effects, or that states which legalized recreational cannabis at a later time point learned from the experiences of states which had legalized recreational cannabis earlier on (57). Nonetheless, when drawing upon the evidence generated by non-comparative or non-longitudinal studies, patterns of increases also emerge (58–63). Careful surveillance of this key metric in future research is recommended to fully grasp the weight and extent of the impact of RML.
Service use trends more readily demonstrate increases, and were predominantly related to cannabis-related hospitalizations, however divergent trends were noted among youth, with one study demonstrating increases (34), and a second study yielding decreases in hospitalizations (35). Reasons for hospitalizations may vary substantially from one individual to another, so future studies will need to disentangle these differences. Otherwise, similar populations, locations, and timeframes were utilized to study this outcome, and it is difficult to determine at this time why opposing trends surfaced from the data.
Prior to the legalization of recreational cannabis use, certain assumptions were formulated about the anticipated impact of these types of policies on the use of other substances. On the one hand, it was hypothesized that the legalization of cannabis could lead to an increase in the co- use of other substances, presumably through a mechanism of cross-sensitization (64). Others proposed, on the contrary, that by legalizing cannabis, consumers would be less exposed to organized crime to obtain the substance, thus potentially discouraging additional access of other substances through this illicit point of contact (65). Finally, other authors, inspired by the theory of self-medication, postulate that by making cannabis more accessible, consumers could substitute their consumption of other substances by turning to cannabis (66). According to our review, we observe, in the case of tobacco, an absence of change in consumption, whereas in the case of alcohol, 3 out of 7 studies have shown an increase in the consumption of this substance. A lack of studies of non-prescription opioids does not allow for any concluding remarks to be made at this time. The reasons why we denote opposing effects of legalization on tobacco and alcohol are difficult to ascertain. In the future, research should focus on alcohol consumption, which remains, to this day, one of the substances with the highest social, economic and health impacts (67, 68).
One of the most robust associations observed in this systematic review is the correlation between RML and prescription opioids. Specifically, of the five studies which investigated the effects of RML on opioid prescription patterns, none reported a significant increase; only one reported a lack of effect; and the remaining five studies reported decreases in MMEs and number of prescriptions. These RML data parallel and align with previous medical marijuana legalization data, which report decreases in the number of opioid prescriptions provided to patients; the number of prescriptions filled by patients; the number of prescriptions discontinued early by patients; MMEs prescribed to patients, etc. (48, 69, 70). Most research articles included on this topic were evaluated as having high-quality evidence. As such, the evidence is sufficient to establish a potentially beneficial association between recreational marijuana legislation and prescription opioid patterns. Influencing prescription practices and restricting access to opioids are two public health strategies which have already been implemented by the Center for Disease Control (CDC) to contain the opioid epidemic (71); one may speculate that the reduction in prescriptions denoted in the current review may be accounted for by these strategies (72). However, the comparative nature of the articles retained in this review suggest that RML states find greater reductions in opioid prescriptions compared to non-RML locations, indicating that RML status may be contributing to a synergistic effect and amplifying these efforts. Beyond this general observation, future research should clarify the nature of this relationship. For instance, it remains to be determined if there are subgroups of healthcare practitioners or organizational services (i.e., surgeons, emergency medicine physicians, family physicians; hospital, community services, etc.) that are more strongly changing their prescription habits, and if there are subcategories of patients who are targeted by these declining practices (i.e., cancer patients, patients undergoing surgeries requiring pain management care, patients with chronic pain, etc.). Likewise, it remains to be determined if the changes in opioid prescription are directly related or not to the providing of alternatives to patients (i.e., medical cannabis prescriptions). Finally, it must be noted that a causational relationship has not yet been established between RML status and opioid prescription patterns.
While a changes in opioid prescription was observed, no effects of cannabis legalization were observed on opioid-related deaths. One can hypothesize that downstream cascading effects may require lengthier follow-up periods to capture differences. Alternatively, the lack of effect on mortality despite the decrease in opioid prescriptions could be explained by the fact that most opioid-related deaths are due to the consumption of particularly powerful opioids (i.e., fentanyl) procured outside of clinical settings. Nonetheless, it is important to note that the lack of effect on opioid-related deaths is based on a single comparative and longitudinal study. Data derived from research which was not selected as a part of this review show diverging patterns, exhibiting patterns of increases, lack of effects, and decreases (73, 74). Finally, it is worth mentioning that most studies on opioid-related outcomes failed to account for significant confounders such as policies related to the delivery of overdose healthcare services, and access to overdose treatment, including naloxone and buprenorphine, which may directly impact opioid-related outcomes (75).
It is important to know that few studies corresponding to the above-reference inclusion and exclusion criteria investigated criminal activity (outside drug-relate possession crimes); evidence from non-comparative or non-longitudinal nature are conflicting, positing increases and decreases in crimes such as violent crime, property crime and sexual assaults (65, 76, 77). Similarly, only two longitudinal and comparative studies investigated the impact of RML on suicide, and none evaluated cannabis potency. Regarding the potency of cannabis, it has been steadily increasing decades before the enactment of any cannabis regulations, transitioning from an approximate 2% of delta-9-tetrahydrocannabinol (Δ9-THC) in 1970, to close to 15% in 2016 (55). Stronger potency of Δ9-THC content in cannabis products is vital to monitor, as it is most likely the main component responsible for the psychological, cognitive and health harms of cannabis (78, 79). We found no comparative and longitudinal study that has evaluated Δ9-THC potency changes before, and subsequent to, RML implementation. To fully assess the consequences of recreational cannabis regulations on public health, it will be relevant to assess this outcome in the future.
Despite the narrow inclusion criteria of this review, it is relevant to compare current findings with population-level health data derived from other adult (recreational) regulatory frameworks, such as the ones in Uruguay and Canada. The Uruguayan experience of recreational legalization has yielded preliminary results which are largely in accordance with the present review, with noted potential increases in the prevalence of adult use, a lack of effect on use in youth, a lack of effect on other drug use, an increase in service use such as hospital visits for intoxication, as well as an increase in serious crimes such as homicides and traffic fatality rates (80–83). Despite the recent recreational legalization in Canada, several publications have yielded crucial insights to the impacts of RML on population health. Echoing most findings from the present publication, the primary evidence suggests adult consumption is on the rise, however CUD prevalence remains stable (12, 84); RML is associated with mixed, yet potentially minimal impacts on consumption in youth (85), and may be linked with possible increases in service use, such as emergency department visits or unintentional cannabis intoxications (86–88). Vast efforts are still ongoing across both nations to better grasp the implications of recreational legalization on public health outcomes.
5. Limitations of the current systematic review
The strengths of the studies collected in this systematic review include their longitudinal study design, which captures important temporal variations of outcomes; and that all studies included one or more comparator locations, which controls for diverging trends occurring outside of cannabis legislation policies. Despite these strengths, a few limitations should also be noted. First, there is a lack of longitudinal comparative studies to investigate key populational health outcomes, and stronger efforts in elucidating these outcomes are required to allow for informed policy decisions. For instance, no controlled study specifically examined the effects of RML on cannabis-related mental health outcomes. Second, the implementation of the comparator criteria entailed the exclusion of all studies derived from Canadian settings. In Canada, cannabis has been legalized across all provinces, thus makes it impossible to carry out studies with comparators locations. It is possible that the trends observed in the United States may not be representative of the Canadian experience of legalization, as there are notable differences in legalization modalities between countries. For example, the Canadian experience of recreational legalization is more standardized across regions than the US experience, is overall more restrictive in terms of licensing, home growing and possession, but more liberal in terms of age of consumption, location of consumption, and limits for driving under the influence (89). Nonetheless, the data extracted from Canadian settings seems to largely parallel findings from the United States, except for consumption data, demonstrating a mixed effect RML on daily cannabis use, whereas US data suggests a likelier increase in daily use (10, 84), and service use, with once again mixed effects in Canada, and more suggestive increasing patterns in the US (90, 91).
There are several confounding factors which are infrequently controlled for, though should be accounted for when analyzing the implications of RML on public health outcomes. For example, the proliferation of marketing strategies of edibles (92) mounted alongside several reports of non-compliant tactics, especially regarding youth (93, 94) may be contributing to an uptick in adverse public health outcomes, such as pediatric exposures to cannabis or emergency department visits (95, 96). Another notable confounder in the landscape of recreational legalization is its potential “spillover effect” to neighboring non-legalized states (19, 97). This could potentially give rise to the under-estimation of effects between legalization states, especially if these neighboring states are included as comparator locations in the analyses. In addition, several studies included in the current review did not differentiate between the legalization and the delayed enactment of recreation cannabis policies; this is a crucial variable to consider in future research, as prior data has already shown a correlation between the number of outlets opened and the prevalence of consumption (98, 99). Several studies analyzed data timepoints less than one-year post-enactment–thus limiting the ability to identify patterns which either require a lengthier time to be detected or identifying patterns which do not endure in time. Finally, the legalization of medical marijuana has rendered it more difficult to scrutinize the consequences of recreational legalization, as the evidence has shown medical legalization influences public attitudes, opinions, and behaviors (100, 101).
6. Conclusion
Considering the entirety of the collected evidence, RML is preliminarily associated with increases in adult consumption of cannabis–but not youth consumption; however, little data from controlled studies is available on frequent/problematic cannabis use. RML is also linked to potential increases in service use, as well as traffic crashes and fatalities. Due to the lack of evidence, we could not determine any patterns associated to crimes and suicide. A potential increase in alcohol use has been observed, while no differences were observed in the case of nicotine. Interestingly, the data demonstrated a reduction of opioid prescriptions in RML states compared to non-RML states. We cannot determine if this effect yields an overall benefit or risk to mortality or morbidity of at-risk populations and thus should be a key focus for future research. Another gap in the field is the lack of controlled studies on the potential impact of RML on mental health outcomes. Finally, further research is clearly needed on the differences in RML policies.
Statements
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
SP and MA conceptualized the systematic review and wrote the manuscript. MA and IZ performed the search and extracted the information. AD and IZ provided critical comments. All authors approved the final version of the manuscript.
Acknowledgments
AD is holder of a Junior II Investigator Award from the Fonds de Recherche du Québec en Santé. SP is holder of the Eli Lilly Canada Chair in Schizophrenia Research.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1.
HallW. The future of the international drug control system and national drug prohibitions. Addiction. (2018) 113:1210–23. 10.1111/add.13941
2.
SmythBPChristieGIG. Opposition to cannabis legalization on public health grounds. Lancet Reg Health. (2021) 9:100142. 10.1016/j.lanwpc.2021.100142
3.
AdinoffBReimanA. Implementing social justice in the transition from illicit to legal cannabis. Am J Drug Alcohol Abuse. (2019) 45:673–88. 10.1080/00952990.2019.1674862
4.
HallW. The costs and benefits of cannabis control policies. Dialog Clin Neurosci. (2020) 22:281–7. 10.31887/DCNS.2020.22.3/whall
5.
RehmJFischerB. Cannabis legalization with strict regulation, the overall superior policy option for public health. Clin Pharmacol Ther. (2015) 97:541–4. 10.1002/cpt.93
6.
Health Canada. (2019). The Final Report of the Task Force on Cannabis Legalization and Regulation. Available online at: https://www.canada.ca/content/dam/hc-sc/healthy-canadians/migration/task-force-marijuana-groupe-etude/framework-cadre/alt/framework-cadre-eng.pdf (accessed July 06, 2022).
7.
KalantH. A critique of cannabis legalization proposals in Canada. Int J Drug Policy. (2016) 34:5–10. 10.1016/j.drugpo.2016.05.002
8.
WittchenH-U. Estimating harful effects of cannabis and use for policy makers: shifting from one mistake to the next?Addiction. (2010) 105:1334–5. 10.1111/j.1360-0443.2010.02937.x
9.
BaeHKerrDCR. Marijuana use trends among college students in states with and without legalization of recreational use: initial and longer-term changes from 2008 to 2018. Addiction. (2020) 115:1115–24. 10.1111/add.14939
10.
CerdáMMauroCHamiltonALevyNSSantaella-TenorioJHasinDet al. Association between Recreational Marijuana Legalization in the United States and Changes in Marijuana Use and Cannabis Use Disorder from 2008 to 2016. JAMA Psychiatry. (2020) 77:165–71. 10.1001/jamapsychiatry.2019.3254
11.
KimJHWeinbergerAHZhuJBarrington-TrimisJWykaKGoodwinRD. Impact of state-level cannabis legalization on poly use of alcohol and cannabis in the United States, 2004–2017. Drug Alcohol Depend. (2021) 218. 10.1016/j.drugalcdep.2020.108364
12.
TurnaJBelisarioKBalodisIvan AmeringenMBusseJMacKillopJ. Cannabis use and misuse in the year following recreational cannabis legalization in Canada: A longitudinal observational cohort study of community adults in Ontario. Drug Alcohol Depend. (2021) 225:108781. 10.1016/j.drugalcdep.2021.108781
13.
DoranNStrongDMyersMGCorreaJBTullyL. Post-legalization changes in marijuana use in a sample of young California adults. Addict Behav. (2021) 115:106782. 10.1016/j.addbeh.2020.106782
14.
LeeMHKim-GodwinYSHurH. Adolescents' marijuana use following recreational marijuana legalization in Alaska and Hawaii. Asia Pac J Publ Health. (2021) 34:65–71. 10.1177/10105395211044917
15.
WeinbergerAHWykaKKimJHSmartRMangoldMSchanzerEet al. A difference-in-difference approach to examining the impact of cannabis legalization on disparities in the use of cigarettes and cannabis in the United States, 2004–17. Addiction. (2022) 117:1768–77. 10.1111/add.15795
16.
AndersonDMSabiaJJ. Notice of retraction and replacement. anderson et al. association of marijuana legalization with marijuana use among US high school students, 1993-2019. JAMA Netw Open. (2021) 4:e2124638. 10.1001/jamanetworkopen.2022.1473
17.
LakeSKerrTWerbDHaines-SaahRFischerBThomasGet al. Guidelines for public health and safety metrics to evaluate the potential harms and benefits of cannabis regulation in Canada. Drug Alcohol Rev. (2019) 38:606–21. 10.1111/dar.12971
18.
LuRWillitsDStohrMKMakinDSnyderJLovrichNet al. The Cannabis Effect on Crime: Time-Series Analysis of Crime in Colorado and Washington State. Justice Q. (2019) 38:565–95. 10.1080/07418825.2019.1666903
19.
WuGBoatengFDLangX. The spillover effect of recreational marijuana legalization on crime: evidence from neighboring states of Colorado and Washington State. J Drug Issues. (2020) 50:392–409. 10.1177/0022042620921359
20.
HallWLynskeyM. Assessing the public health impacts of legalizing recreational cannabis use: the US experience. World Psychiatry. (2020) 19:179–86. 10.1002/wps.20735
21.
RobertsBA. Legalized cannabis in Colorado emergency departments: a cautionary review of negative health and safety effects. West J Emerg Med. (2019) 20:557–72. 10.5811/westjem.2019.4.39935
22.
WangGSHoyteCRooseveltGHeardK. The continued impact of marijuana legalization on unintentional pediatric exposures in Colorado. Clin Pediatr. (2018) 58:114–6. 10.1177/0009922818805206
23.
AlleyZMKerrDCRBaeH. Trends in college students' alcohol, nicotine, prescription opioid and other drug use after recreational marijuana legalization: 2008–2018. Addict Behav. (2020) 102. 10.1016/j.addbeh.2019.106212
24.
VeligatiSHowdeshellSBeeler-StinnSLingamDAllenPCChenLSet al. Changes in alcohol and cigarette consumption in response to medical and recreational cannabis legalization: evidence from U.S. state tax receipt data. Int J Drug Policy. (2020) 75:102585. 10.1016/j.drugpo.2019.10.011
25.
LiberatiAAltmanDGTetzlaffJMulrowCGøtzschePCIoannidisJPAet al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ. (2009) 339:b2700. 10.1136/bmj.b2700
26.
WellsGSheaBO'ConnellDPetersonJWelchVLososMet al. The Newcastle-Ottawa Scale (NOS) for Assessing the Quality If Nonrandomized Studies in Meta-Analyses. Available online at: http://www.ohri.ca/programs/clinical_epidemiology/oxford.htm (accessed July 1, 2022).
27.
KerrDCRBaeHPhibbsSKernAC. Changes in undergraduates' marijuana, heavy alcohol and cigarette use following legalization of recreational marijuana use in Oregon. Addiction. (2017) 112:1992–2001. 10.1111/add.13906
28.
KerrDCRBaeHKovaAL. Oregon recreational marijuana legalization: changes in undergraduates'marijuana use rates from 2008 to 2016. Psychol Addict Behav. (2018) 32:670–8. 10.1037/adb0000385
29.
MillerAMRosenmanRCowanBW. Recreational marijuana legalization and college student use: Early evidence. SSM Popul Health. (2017) 3:649–57. 10.1016/j.ssmph.2017.08.001
30.
WallaceGTParnesJEPrinceMAConnerBTRiggsNRGeorgeMWet al. Associations between marijuana use patterns and recreational legislation changes in a large Colorado college student sample. Addict Res Theory. (2020) 28:211–21. 10.1080/16066359.2019.1622003
31.
CerdáMWallMFengTKeyesKMSarvetASchulenbergJet al. Association of state recreational marijuana laws with adolescent marijuana use. JAMA Pediatr. (2017) 171:142–9. 10.1001/jamapediatrics.2016.3624
32.
ColeyRLKruzikCGhianiMCareyNHawkinsSSBaumCF. Recreational marijuana legalization and adolescent use of marijuana, tobacco, and alcohol. J Adolesc Health. (2021) 69:41–9. 10.1016/j.jadohealth.2020.10.019
33.
DellingFNVittinghoffEDewlandTAPletcherMJOlginJENahGet al. Does cannabis legalisation change healthcare utilisation? A population-based study using the healthcare cost and utilisation project in Colorado, USA. BMJ Open. (2019) 9:e027432. 10.1136/bmjopen-2018-027432
34.
MasonbrinkARRichardsonTHallMCatleyDWilsonK. Trends in adolescent cannabis-related hospitalizations by state legalization laws, 2008–2019. J Adolesc Health. (2021) 69:999–1005. 10.1016/j.jadohealth.2021.07.028
35.
MennisJStahlerGJMcKeonTP. Young adult cannabis use disorder treatment admissions declined as past month cannabis use increased in the U.S.: an analysis of states by year, 2008–2017. Addict Behav. (2021) 123:107049. 10.1016/j.addbeh.2021.107049
36.
ShiYLiangD. The association between recreational cannabis commercialization and cannabis exposures reported to the US National Poison Data System. Addiction. (2020) 115:1890–9. 10.1111/add.15019
37.
MennisJStahlerGJMasonMJ. Treatment admissions for opioids, cocaine, and methamphetamines among adolescents and emerging adults after legalization of recreational marijuana. J Subst Abuse Treat. (2021) 122:108228. 10.1016/j.jsat.2020.108228
38.
KamerRWarshafskySKamerG. Change in traffic fatality rates in the first 4 states to legalize recreational marijuana. JAMA Intern Med. (2020) 180:1116–8. 10.1001/jamainternmed.2020.1769
39.
LaneTJHallW. Traffic fatalities within US states that have legalized recreational cannabis sales and their neighbours. Addiction. (2019) 114:847–56. 10.1111/add.14536
40.
BenedettiMHLiLNeurothLMHumphriesKDBrooks-RussellAZhuM. Self-reported driving after marijuana use in association with medical and recreational marijuana policies. Int J Drug Policy. (2021) 92:102944. 10.1016/j.drugpo.2020.102944
41.
MatthayECKiangMElserHSchmidtLHumphreysK. Evaluation of State Cannabis laws and rates of self-harm and assault. JAMA Network Open. (2021) 4:e211955. 10.1001/jamanetworkopen.2021.1955
42.
LuT. Marijuana legalization and household spending on food and alcohol. Health Econ. (2021) 30:1684–96. 10.1002/hec.4266
43.
BhaveAMurthiBPS. A Study of the Effects of Legalization of Recreational Marijuana on Sales of Cigarettes. SSRN Working Paper. (2020). 10.2139/ssrn.3508422
44.
DrakeCWenJHindeJWenH. Recreational cannabis laws and opioid-related emergency department visit rates. Health Econ. (2021) 30:2595–605. 10.1002/hec.4377
45.
McMichaelBJvan HornRLViscusiWK. The impact of cannabis access laws on opioid prescribing. J Health Econ. (2020) 69:102273. 10.1016/j.jhealeco.2019.102273
46.
WenJWenHButlerJSTalbertJC. The impact of medical and recreational marijuana laws on opioid prescribing in employer-sponsored health insurance. Health Econ. (2021) 30:989–1000. 10.1002/hec.4237
47.
ShiYLiangDBaoYAnRWallaceMSGrantI. Recreational marijuana legalization and prescription opioids received by Medicaid enrollees. Drug Alcohol Depend. (2019) 194:13–9. 10.1016/j.drugalcdep.2018.09.016
48.
LopezAKKNicholsSDChungDYKaufmanDEMcCallKLPiperBJ. Prescription Opioid Distribution after the Legalization of Recreational Marijuana in Colorado. Int J Environ Res Public Health. (2020) 17:3251. 10.3390/ijerph17093251
49.
AlcocerJJ. Exploring the effect of Colorado's recreational marijuana policy on opioid overdose rates. Public Health. (2020) 185:8–14. 10.1016/j.puhe.2020.04.007
50.
DoucetteMLBorrupKTLapidusGWhitehillJMMcCourtADCrifasiCK. Effect of Washington State and Colorado's cannabis legalization on death by suicides. Prev Med. (2021) 148:106548. 10.1016/j.ypmed.2021.106548
51.
GreenKMDohertyEEEnsmingerME. Long-term consequences of adolescent cannabis use: examining intermediary processes. Am J Drug Alcohol Abuse. (2017) 43:567–75. 10.1080/00952990.2016.1258706
52.
LisdahlKMWrightNEMedina-KirchnerCMapleKEShollenbargerS. Considering cannabis: the effects of regular cannabis use on neurocognition in adolescents and young adults. Curr Addict Rep. (2014) 1:144–56. 10.1007/s40429-014-0019-6
53.
MontgomeryBWRobertsMHMargerisonCEAnthonyJC. Estimating the effects of legalizing recreational cannabis on newly incident cannabis use. PLoS ONE. (2022) 17:e0271720. 10.1371/journal.pone.0271720
54.
HasinDSShmulewitzDSarvetAL. Time trends in US cannabis use and cannabis use disorders overall and by sociodemographic subgroups: a narrative review and new findings. Am J Drug Alcohol Abuse. (2019) 45:623–43. 10.1080/00952990.2019.1569668
55.
ElSohlyMAMehmedicZFosterSGonCChandraSChurchJC. Changes in cannabis potency over the last 2 Decades (1995–2014): analysis of current data in the United States. Biol Psychiatry. (2016) 79:613–9. 10.1016/j.biopsych.2016.01.004
56.
HallWDegenhardtL. Adverse health effects of non-medical cannabis use. Lancet. (2009) 374:1383–91. 10.1016/S0140-6736(09)61037-0
57.
FarmerCMMonfortSSWoodsAN. Changes in traffic crash rates after legalization of marijuana: results by crash severity. J Stud Alcohol Drugs. (2022) 83:494–501. 10.15288/jsad.2022.83.494
58.
HansenBMillerKWeberC. Early evidence on recreational marijuana legalization and traffic fatalities. Econ Inq. (2020) 58:547–68. 10.1111/ecin.12751
59.
InstituteHLD. Recreational marijuana and collision claim frequencies. HLDI Bulletin. (2018) 35:1–14. Available online at: https://www.iihs.org/media/f5fb46ff-d4b7-47b5-9c2c-951f2a30e0a6/8W5rpg/HLDI%20Research/Bulletins/hldi_bulletin_37-20.pdf (accessed July 10, 2022).
60.
Monfort SS,. Effect of Recreational Marijuana Sales on Police-reported Crashes in Colorado, Oregon, Washington. Insurance Institute for Highway Safety. (2018). Available online at: https://www.iihs.org/api/datastoredocument/bibliography/2173#:~:text=Colorado%2C%20Washington%2C%20and%20Oregon%20experienced,they%20not%20legalized%20recreational%20marijuana. (accessed July 12, 2022).
61.
VoglerJ. State Marijuana Policies and Vehicle Fatalities. SSRN Working Paper. (2017). p. 3013701. 10.2139/ssrn.3013701
62.
WindleSBEisenbergMJReynierPCabausselJThombsBDGradRet al. Association between legalization of recreational cannabis and fatal motor vehicle collisions in the United States: an ecologic study. CMAJ. (2021) 9:E233–41. 10.9778/cmajo.20200155
63.
WindleSBSochaPNazif-MunozJIHarperSNandiA. The impact of cannabis decriminalization and legalization on road safety outcomes: a systematic review. Am J Prev Med. (2022) 63:1037–52. 10.1016/j.amepre.2022.07.012
64.
MelbergHOJonesAMBretteville-JensenAL. Is cannabis a gateway to hard drugs?. Empir Econ. (2010) 38:583–603. 10.1007/s00181-009-0280-z
65.
DragoneDPraroloGVaninPZanellaG. Crime and the legalization of recreational marijuana. J Econ Behav Org. (2019) 159:488–501. 10.1016/j.jebo.2018.02.005
66.
KvammeSLPedersenMMThomsenKRThylstrupB. Exploring the use of cannabis as a substitute for prescription drugs in a convenience sample. Harm Reduct J. (2021) 18:7210.1186/s12954-021-00520-5
67.
MossHB. The impact of alcohol on society: A brief overview. Soc Work Public Health. (2013) 28:175–7. 10.1080/19371918.2013.758987
68.
ParkSHKimDJ. Global and regional impacts of alcohol use on public health: Emphasis on alcohol policies. Clin Molecul Hepatol. (2020) 26:652–61. 10.3350/cmh.2020.0160
69.
BradfordACBradfordWDAbrahamAAdamsGB. Association between US state medical cannabis laws and opioid prescribing in the medicare part d population. JAMA Intern Med. (2018) 178:667–72. 10.1001/jamainternmed.2018.0266
70.
FlexonJLStolzenbergLD'AlessioSJ. The effect of cannabis laws on opioid use. Int J Drug Policy. (2019) 74:152–9. 10.1016/j.drugpo.2019.09.013
71.
DowellDHaegerichTMChouR. CDC guideline for prescribing opioids for chronic pain—United States, 2016. JAMA. (2016) 315:1624–45. 10.1001/jama.2016.1464
72.
BohnertASBGuyGPLosbyJL. Opioid prescribing in the united states before and after the centers for disease control and prevention's 2016 opioid guideline. Ann Intern Med. (2018) 169:367–75. 10.7326/M18-1243
73.
BleyerABarnesB. Opioid death rate acceleration in jurisdictions legalizing marijuana use. JAMA Intern Med. (2018) 178:1280–1. 10.1001/jamainternmed.2018.3888
74.
ChanNWBurkhardtJFlyrM. The effects of recreational marijuana legalization and dispensing on opioid mortality. Econ Inq. (2020) 58:589-606. 10.1111/ecin.12819
75.
HuTSnider-AdlerMNijmehLPyleA. Buprenorphine/naloxone induction in a Canadian emergency department with rapid access to community-based addictions providers. CJEM. (2019) 21:492–8. 10.1017/cem.2019.24
76.
BrinkmanJMok-LammeD. Not in my backyard? Not so fast. The effect of marijuana legalization on neighborhood crime. Reg Sci Urban Econ. (2019) 78:103460. 10.1016/j.regsciurbeco.2019.103460
77.
ThackerJMartinMECristyYRabideauDShivelyMKlingR. Exploring the neighborhood-level impact of retail marijuana outlets on crime in Washington State. J Quant Criminol. (2021). 10.1007/s10940-021-09534-5
78.
AshtonCH. Pharmacology and effects of cannabis: a brief review. Br J Psychiatr. (2001) 178:101–6. 10.1192/bjp.178.2.101
79.
D'SouzaDCPerryEMacDougallLAmmermanYCooperTWuYTet al. The psychotomimetic effects of intravenous delta-9-tetrahydrocannabinol in healthy individuals: implications for psychosis. Neuropsychopharmacology. (2004) 29:1558–72. 10.1038/sj.npp.1300496
80.
BaudeanM. Five Years of Cannabis Regulation What Can We Learn From the Uruguayan Experience? IN: The Routledge Handbook of Post-Prohibition Cannabis Research. New York, NY. (2021).
81.
LaqueurHRivera-AguirreAShevACastillo-CarnigliaARudolphKERamirezJet al. The impact of cannabis legalization in Uruguay on adolescent cannabis use. Int J Drug Policy. (2020) 80:102748. 10.1016/j.drugpo.2020.102748
82.
Nazif-MunozJIOulhoteYOuimetMC. The association between legalization of cannabis use and traffic deaths in Uruguay. Addiction. (2020) 115:1697–706. 10.1111/add.14994
83.
Rivera-AguirreACastillo-CarnigliaALaqueurHSRudolphKEMartinsSSRamírezJet al. Does recreational cannabis legalization change cannabis use patterns? Evidence from secondary school students in Uruguay. Addiction. (2022) 117:2866–77. 10.1111/add.15913
84.
RotermannM. What has changed since cannabis was legalized?Health Rep. (2020) 31:11–20. 10.25318/82-003-x202000200002-eng
85.
Rubin-KahanaDSCrépaultJFMathesonJLe FollB. The impact of cannabis legalization for recreational purposes on youth: a narrative review of the Canadian experience. Front Psychiatry. (2022) 2148:984485. 10.3389/fpsyt.2022.984485
86.
CohenNGalvis BlancoLDavisAKahaneAMathewMSchuhSet al. Pediatric cannabis intoxication trends in the pre and post-legalization era. Clin Toxicol. (2022) 60:53–8. 10.1080/15563650.2021.1939881
87.
O'BrienMRogersPSmithE. A chart review of emergency department visits following implementation of the cannabis act in Canada. Can J Med. (2022) 4:13–21. 10.33844/cjm.2022.6016
88.
YeungMEMWeaverCGHartmannRHaines-SaahRLangE. Emergency department pediatric visits in alberta for cannabis after legalization. Pediatrics. (2021) 148:e2020045922. 10.1542/peds.2020-045922
89.
LancioneSWadeKWindleSBFilionKBThombsBDEisenbergMJ. Non-medical cannabis in North America: an overview of regulatory approaches. Public Health. (2020) 178:7–14. 10.1016/j.puhe.2019.08.018
90.
AugerNLuuTMAyoubABilodeau-BertrandMLoELowN. Cannabis-related hospitalizations among youth in canada before and after cannabis legalization. J Addict Med. (2021) 15:245–247. 10.1097/ADM.0000000000000747
91.
BaranieckiRPanchalPMalhotraDDAliferisAZiaZ. Acute cannabis intoxication in the emergency department: the effect of legalization. BMC Emerg Med. (2021) 21:32. 10.1186/s12873-021-00428-0
92.
MacCounRJMelloMM. Half-baked — the retail promotion of marijuana edibles. N Engl J Med. (2015) 372:989–91. 10.1056/NEJMp1416014
93.
SheikhanNYPintoAMNowakDAAbolhassaniFLefebvrePDuhMSet al. Compliance with cannabis act regulations regarding online promotion among Canadian commercial cannabis-licensed firms. JAMA Network Open. (2021) 4:e2116551–e2116551. 10.1001/jamanetworkopen.2021.16551
94.
ShiYPaculaRL. Assessment of recreational cannabis dispensaries' compliance with underage access and marketing restrictions in California. JAMA Pediatr. (2021) 175:1178–80. 10.1001/jamapediatrics.2021.2508
95.
BergerE. Legal marijuana and pediatric exposure: pot edibles implicated in spike in child emergency department visits. Ann Emerg Med. (2014) 64:A19–21. 10.1016/j.annemergmed.2014.08.010
96.
WangGSle LaitMCDeakyneSJBronsteinACBajajLRooseveltG. Unintentional pediatric exposures to marijuana in Colorado, 2009-2015. JAMA Pediatr. (2016) 170:e160971–e160971. 10.1001/jamapediatrics.2016.0971
97.
HaoZCowanBW. The cross-border spillover effects of recreation marijuana legalization. Econ Inq. (2020) 58:642–66. 10.1111/ecin.12764
98.
EversonEMDilleyJAMaherJEMackCE. Post-legalization opening of retail cannabis stores and adult cannabis use in Washington State, 2009–2016. Am J Public Health. (2019) 109:1294–301. 10.2105/AJPH.2019.305191
99.
PedersenERFirthCLRodriguezAShihRASeelamRKrausLet al. Examining associations between licensed and unlicensed outlet density and cannabis outcomes from preopening to postopening of recreational cannabis outlets. Am J Addict. (2021) 30:122–30. 10.1111/ajad.13132
100.
CerdáMWallMKeyesKMGaleaSHasinD. Medical marijuana laws in 50 states: Investigating the relationship between state legalization of medical marijuana and marijuana use, abuse and dependence. Drug Alcohol Depend. (2012) 120:22–7. 10.1016/j.drugalcdep.2011.06.011
101.
SznitmanSRLewisN. Examining effects of medical cannabis narratives on beliefs, attitudes, and intentions related to recreational cannabis: a web-based randomized experiment. Drug Alcohol Depend. (2018) 185:219–25. 10.1016/j.drugalcdep.2017.11.028
Summary
Keywords
cannabis, legalization, recreational, review, longitudinal
Citation
Athanassiou M, Dumais A, Zouaoui I and Potvin S (2023) The clouded debate: A systematic review of comparative longitudinal studies examining the impact of recreational cannabis legalization on key public health outcomes. Front. Psychiatry 13:1060656. doi: 10.3389/fpsyt.2022.1060656
Received
03 October 2022
Accepted
20 December 2022
Published
11 January 2023
Volume
13 - 2022
Edited by
Matthew S. Ellis, Washington University in St. Louis, United States
Reviewed by
Ruben David Baler, National Institutes of Health (NIH), United States; Omayma Alshaarawy, College of Human Medicine, Michigan State University, United States
Updates
Copyright
© 2023 Athanassiou, Dumais, Zouaoui and Potvin.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Stéphane Potvin ✉ stephane.potvin@umontreal.ca
This article was submitted to Addictive Disorders, a section of the journal Frontiers in Psychiatry
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.