From the Editor

It’s legal. Are people using more? And has cannabis use disorder become more common?

This week, we look at a new paper considering cannabis legalization and use. The authors draw on American data where legalization is increasingly found across different states though not as extensively as in Canada; to date, 11 US states have legalized recreational cannabis, with 33 (and D.C.) having legalized medical marijuana.

In a new JAMA Psychiatry paper, Magdalena Cerdá and her co-authors use the National Survey on Drug Use, a major survey involving more than half a million participants, considering marijuana use, frequent use, and cannabis use disorder. What effect does legalization have? They find that cannabis use disorder is more common in adolescents after legalization and for adults who are 26 years of age and older, use, frequent use, and substance use are all up.


We consider the paper. We also discuss the commentary that accompanies it. Finally, with an eye closer to home, we ask: are the findings relevant here in Canada?



“Association Between Recreational Marijuana Legalization in the United States and Changes in Marijuana Use and Cannabis Use Disorder From 2008 to 2016”

Magdalena Cerdá, Christine Mauro, Ava Hamilton,  Natalie S. Levy, Julián Santaella-Tenorio, Deborah Hasin, Melanie M. Wall, Katherine M. Keyes, Silvia S. Martins

JAMA Psychiatry, 13 November 2019  Online First

Marijuana is one of the world’s most widely used psychoactive substances. Approximately 183 million people, or 3.8% of the world’s population, used marijuana in 2015. Although occasional marijuana use is not associated with substantial problems, long-term, heavy use is linked to psychological and physical health concerns, lower educational attainment, decline in social class, unemployment, and motor vehicle crashes.

In the United States, 33 states and the District of Columbia have enacted medical marijuana legalization (MML), and 11 states and the District of Columbia have enacted recreational marijuana legalization (RML). Reducing racial/ethnic inequalities in marijuana-related arrests and convictions has been a key aim of legalization. At the same time, questions have arisen about the impact of legalization on marijuana use.10 Prior studies did not find any association between MML and adolescent marijuana use; however, they did find increases in adult past-month use and cannabis use disorder (CUD) after MML. Findings on MML may not apply to RML, because RML could have a broader impact through changes in pricing, access, acceptability, and advertising.

Little is known about the association of RML with marijuana use and how it differs from that of MML with marijuana use…

In this study, we addressed these gaps by comparing past-month marijuana use, frequent marijuana use, and CUD before and after RML enactment among adolescents, young adults, and adults, using data from 2008 to 2016. To distinguish the association of RML enactment with new marijuana users from that with problematic use among existing users, we examined changes in frequent marijuana use and CUD among the overall sample as well as among users. States that did not enact RML, regardless of whether they had enacted MML, were included in the analyses to control for nationwide trends in marijuana use.

magdalena_cerda2Magdalena Cerdá

Here’s what they did:

  • The authors used data from the National Survey on Drug Use and Health (NSDUH) which collects data on more than “17  500 adolescents (aged 12-17 years), 17  500 young adults (aged 18-25 years), and 18 800 adults (aged ≥26 years) on a yearly basis.”
  • “In-person audio, computer-assisted self-interviews were conducted by professional field interviewers; response rates ranged from 55% to 75%, depending on the year.”
  • The exposure variable: the state-level enactment of RML in Colorado, Washington, Alaska, and Oregon.
  • Statistical analyses were done. “We calculated the prevalence of past-month marijuana use, past-month frequent marijuana use, and CUD for each year from 2008 to 2016, providing estimates of trends in marijuana use before any RML enactment (ie, 2008-2011) and after RML enactment (ie, 2013-2016 or 2015-2016, depending on the state of residence).” The authors also did a series of multilevel logic regression models.

Here’s what they found:

  • The study included 505  796 respondents. (!!)
  • Demographics: 51.51% were females; 77.24% participants were 26 years or older; 65.43% were white.
  • Teens. “Among the 12- to 17-year-old respondents, the prevalence of past-month marijuana use and past-month frequent use following state RML enactment did not change in the overall sample or among users. However, after RML enactment, past-year CUD prevalence increased slightly among all 12- to 17-year-old respondents… This increase was 25% higher than that for participants in the same age group in states with no RML enactment. Among the past-year users, CUD increased from 22.80% to 27.20%…”
  • Young adults. No change. (!)
  • Adults. “Among respondents aged 26 years or older, past-month marijuana use after RML enactment increased from 5.65% to 7.10%… Furthermore, past-month frequent use increased from 2.13% to 2.62%… and past-year CUD increased from 0.90% to 1.23%… Among users, past-month frequent use and past-year CUD did not increase…”

In the present study, RML enactment during 2012 to 2015 was followed by a small increase in adolescent CUD and by an increase in past-month marijuana use, past-month frequent marijuana use, and past-year CUD among adults 26 years or older.

A few thoughts:

  1. This is a good paper.
  1. The data set is impressive. It’s not just that there were so many participants, but the data was broken down by age.
  1. The big finding? There are two. First, adolescent cannabis use disorder was up. Second, use, frequent use, and cannabis use disorder were all up among adults 26 years and older.
  1. But how do we interpret the result about adolescents? The authors did a sensitivity analysis for adolescents without conclusive results. (They note: “The extent to which such confounders exist is unclear because our difference-in-difference design accounted for unmeasured time-invariant sources of confounding and also adjusted for measured time-varying individual- and state-level demographic characteristics. However, the small E-values warrant a conservative interpretation of the increase in CUD among participants aged 12 to 17 years.”)

In a podcast interview with Dr. John Torous, the paper’s first author suggests that we should take the finding for adolescents with “a grain of salt.” (You can find the full interview here:

  1. The data is good but we should recognize its limitations. Remember that of the four states with legalization, Oregon and Alaska had less than two years of complete data to analyze. It’s early in understanding of how legalization may or may not affect use.
  1. The accompanying commentary is lucid.


“Where Is Cannabis Legalization Leading?”

Richard A. Grucza and Andrew D. Plunk

JAMA Psychiatry, 13 November 2019  Online First

The results will provide fodder for legalization advocates and opponents alike, but taken as a whole, they suggest a more nuanced view. Legalization does not seem to have thus far resulted in large increases in the prevalence of past-year cannabis use among adolescents and young adults. Increases in cannabis use among adults of legal age were observed, but these should not be surprising because at least part of the rationale for creating regulated commercial markets for cannabis is to provide access for adults who choose to use it.

gruczaRichard A. Grucza

So opens a commentary paper by Grucza (Washington University School of Medicine) and Plunk (Eastern Virginia Medical School).

They see a public policy angle to the study:

More concerning from a public health perspective, the authors document statistically significant increases in the prevalence of frequent use, defined as more than 20 days in the past month, and in past-year CUD among adults 26 years and older. These increases may be substantial, reaching as high as 25% to 35%. Critics might argue that the DSM-IV CUD construct as operationalized casts too broad of a net; that is, it seems unlikely that everybody who meets the criteria for CUD in a general population epidemiologic survey (more than 12% of people who use cannabis) is in need of treatment. However, nearly 1 in 4 individuals who meet this threshold report not being able to quit or cut down on use despite wanting to do so.

They push further, drawing a parallel with the end of prohibition.

[C]annabis legalization is similar to the end of the prohibition of alcohol. The comparison between commercial legalization of these drugs is one that public health and policy reform advocates should keep in mind moving forward; that is, while there is fairly wide public acceptance of cannabis use (as with alcohol), it will nonetheless have real health consequences for many users. Prohibition may protect individuals from health consequences, but it is becoming increasingly untenable as a policy option for cannabis. Thus, instead of seeking to prohibit a relatively common adult behavior, reducing harms associated with problematic use should be seen as the preferred public health strategy.

A few thoughts:

  1. This is well argued commentary.
  1. They are right about the data – it can be interpreted both ways. There was change, but not incredible change.
  1. And they are right about the public health perspective.
  1. But all this discussion of the American experience begs a simple question: what are the implications for our own experiment in Canada, first with legalization for medical use and, more recently, with recreational cannabis? Canadian data suggests that use and cannabis use disorder were both higher in this country than in the United States (prior to legalization). So, are new American findings applicable to Canada? Does legalization (medical and recreational) lead to a more permissive society in which cannabis is used and misused more? Or are Americans catching up to Canadian numbers – and we are unlikely to see a post-legalization spike here?

We draw on some Canadian data. ODUHS gives an extensive look at substance use among Ontario students. With decades of survey results, the trend is clear: substance use is collectively down (to the year 2017). It’s too early to consider than in the context of legalization, but as societal views of cannabis have changed, there has been no surge in use – at least with this demographic group in one province. (See here for a full report:—osduhs/drug-use-among-ontario-students-1977-2017—detailed-findings-from-the-osduhs.pdf?la=en&hash=2B434CDAAD485834497E3B43F2264BDEB255F29F.)

  1. What are the clinical implications of this major study and its commentary? For us Canadian clinicians, the simple lesson may be this: adolescents are at risk of greater misuse, and we should spend time exploring their attitudes towards marijuana in our clinics and EDs.


Reading of the Week. Every week I pick articles and papers from the world of Psychiatry.