Does Marijuana Cause Psychosis?

New research looks beyond the association between cannabis and schizophrenia.

Posted Jan 06, 2016

My home state of Vermont, like many others, is seriously considering legalizing marijuana for recreational use.  The arguments for and against the idea are based on economics, philosophy, and science.  The latter is where this post is targeted, specifically the widely held belief that cannabis is associated with few serious health consequences.  Many medical and substance use professionals have been puzzled as to how this view can persist despite so much evidence to the contrary.  While one reason may have to do with people simply believing what they would like to believe,  it’s also possible that those of us in the profession cause people to tune out if we overstate the risks.  Here, then, is an attempt to sort through what we actually know about one of the scarier potential risks of cannabis use, namely the link between marijuana and psychosis and psychotic disorders such as schizophrenia.

A recent review on the subject was published in the highly respected journal, Biological Psychiatry.  According to this article, there have now been 10 research studies that have examined the long-term association between cannabis and psychotic disorders.  None of the studies were perfect and all of them looked at the question a little bit differently, which is one reason why it is so easy for people on either side of the issue to cherry pick studies that support one’s prior opinions.  Seven of the 10 studies found a statistically significant link between cannabis use and later psychotic symptoms and disorders.  Putting all 10 studies together, a 40% increase risk of psychosis was found comparing people who had used cannabis to people who have never used it.  These odds tend to increase in studies that examine people who use cannabis more heavily.

Yet, as people rightly point out, association is not causation, and the authors of this study try to go into more depth as to the strength of evidence that cannabis is actually playing a causal role in this link.  Drilling down, enough of these studies have done a reasonably good job of statistically controlling for baseline factors to rule out the “reverse causation” hypothesis: namely that early psychotic symptoms lead to cannabis use rather than the other way around.  A “residual confound” problem, meaning that what looks like causation is actually the product of other unmeasured variables that lead both to cannabis use and psychosis (like genetically based predispositions), remains a possibility that might lead to an overestimation of the true risk.  However, the authors also note that other problems with these studies could lead to an underestimation of the actual risk, such as the fact that more impaired people tend to be more likely to drop out of long-term studies. 

There are also some other lines of evidence that point to a causal role of cannabis in the development of psychosis.

  • The well-known fact that acute intoxication of cannabis may produce transient psychotic symptoms.
  • An increased risk of psychosis among cannabis users who carry specific risk genes. (Ironically, however, the most famous Dunedin study that everyone quotes on the COMT gene has not been replicated).
  • Increasing evidence that the more potent marijuana that is available now may be associated with additional risk.
  • The finding that the link between cannabis and psychosis is not equal for all age groups but may be stronger for younger users.

Putting all of this together, the authors concluded that “there is a strong body of epidemiologic evidence to support the view that regular or heavy cannabis use increases the risk of developing psychotic disorders….”

While it seems to be more and more difficult to ignore these data, this increased risk needs to be understood in context.  If the overall rate of schizophrenia in the population is about 1%, then these data suggest that the risk of psychosis, at least among more casual users, is still fairly low, like 1.5%.  Given the potential of millions of new cannabis users, however, these numbers translate into many thousands of individuals with what can be quite disabling psychotic symptoms at a time when mental health and substance abuse treatment centers are already stretched extremely thin.

As these discussions continue, it will be important for people on all sides of the issue to avoid both glossing over inconvenient scientific findings about the risks of cannabis and apocalyptic prophecies about the end of civil society.  The real data that the rest of the country needs is now just starting to trickle in from states like Colorado and Washington.  Like the first studies on global warming, the information might appear somewhat haphazard, inconsistent and easy to manipulate.  With time, however, a true signal will likely emerge and I, for one, am willing to wait.

(Note: This post was adapted from a recent column in Pediatric News)

@copyright by David Rettew, MD

David Rettew is author of Child Temperament: New Thinking About the Boundary Between Traits and Illness and a child psychiatrist in the psychiatry and pediatrics departments at the University of Vermont College of Medicine.

Follow him at @Pedipsych and like PediPsych on Facebook.