The Legal Cannabis Reduces Prescription Codeine

States with legal recreational cannabis see a reduction in demand for prescription codeine, according to a new study led by researchers at the University of Pittsburgh and Cornell University.

Codeine is an opioid medication that binds to the opioid receptors in the brain and spinal cord. Like other opioids, it can cause physical and psychological dependence.

The authors concluded: “We find that RCLs [Recreational Cannabis Laws] lead to a reduction in codeine dispensed at retail pharmacies. Among prescription opioids, codeine is particularly likely to be used non-medically. Thus, the finding that RCLs appear to reduce codeine dispensing is potentially promising from a public health perspective.”

Indeed, people swapping out opioids for cannabis is excellent news. But how soon should we be celebrating? This study wasn’t a randomized control trial. It used a two-way fixed-effects regression.

So what’s the verdict? Does legal cannabis reduce prescription codeine?

Cannabis & Codeine: The Study’s Methodology

Cannabis & Codeine: The Study's Methodology

A two-way fixed-effects regression is a statistical method that researchers use to analyze the relationship between two variables where both variables have fixed effects. “Fixed effects” means that the variables are not randomly assigned but are inherent characteristics of the studied units.

In this example, the relationship between recreational cannabis laws and prescription codeine. This method is popular for controlling for unobserved heterogeneity in panel data studies. However, there are several downsides to using a two-way fixed-effects regression in a study:

  1. Complexity: Two-way fixed-effects regressions can be more complex to estimate and interpret than other regression models, such as a one-way fixed-effects or a random-effects model.
  2. Assumptions: Two-way fixed-effects regressions rely on certain assumptions about the underlying data, such as the independence of the errors and normality. The regression results may be biased or unreliable if researchers do not meet these assumptions.
  3. Limited inference: Researchers typically use two-way fixed-effects regressions to estimate within-unit associations and make inferences about the population of units. It may not be possible to generalize the findings to a larger population or to make causal claims.
  4. Power: Two-way fixed-effects regression is computationally intensive. Thus it may require a large sample size to have enough power to detect statistical significance. This could be a problem with limited datasets (but is irrelevant to this codeine and cannabis study).
  5. Data availability: Two-way fixed-effects regression requires panel data, which is a type of data that contains observations on multiple units over time. This type of data may not be available for all research questions.

The researchers analyzed data from the Drug Enforcement Administration’s Automation of Reports and Consolidation Orders System (ARCOS).

Details of the Study

Legal Cannabis Reduces Prescription Codeine

If this study’s results are accurate, this is certainly news worth celebrating. The researchers found that in states where cannabis was legal:

  • A reduction of 26% in the pharmacy-based distribution of codeine and as much as a 37% reduction after recreational cannabis laws have been in effect for four years. 
  • Minimal impact on the distribution of other opioids such as oxycodone, hydrocodone and morphine in any setting.
  • Minimal impact on codeine distribution by hospitals that often have less permissive policies than pharmacies.

“This finding is particularly meaningful,” said senior author Coleman Drake, Ph.D., in a press release. “Among prescription opioids, codeine misuse is especially high. Our findings suggest recreational cannabis use may be a substitute for codeine misuse.”

“Increasing legal access to cannabis may shift some consumers away from opioids and towards cannabis,” said Johanna Catherine Maclean, Ph.D., George Mason University. “While all substances have some risks, cannabis use is arguably less harmful to health than the non-medical use of prescription opioids.”

Indeed, whether the study is describing reality or confirming its bias, the conclusion should always be: legalize it.