Does habitual cannabis use trigger psychosis, or does a tendency toward psychosis predict increased cannabis use?

“[Our] results suggest that it is more likely than not that continued cannabis use after onset of psychosis is causally associated with increased risk of relapse of psychosis, resulting in psychiatric hospitalization,” reported the authors of a study published online Sept. 28 in JAMA Psychiatry.

It is well established that cannabis use after first-episode psychosis is associated with poor outcomes; what is not so well established is why. The three most prominent theories so far are that cannabis use and psychosis share inherent genetic and/or environmental vulnerabilities, impending psychosis impels increased cannabis use, and conversely, cannabis use has a causal effect on psychosis (Lancet Psychiatry. 2016:3[5]:401) and (Lancet Psychiatry. 2015;3[5]401-2).


After conducting a 10-year, prospective cohort investigation of 220 people who presented to psychiatric services at numerous sites in London between April 2002 and July 2013, Tabea Schoeler and her associates in the current study found that the odds of a relapse of psychosis during periods of cannabis use was increased by 1.13, compared with periods of no cannabis use (95% confidence interval, 1.03-1.24) (JAMA Psychiatry. 2016 Sep 28. doi: 10.1001/jamapsychiatry.2016.2427). Further, those who used cannabis continually after first-episode psychosis were found to have a 59.1% higher risk of relapsing psychosis in that 10-year period, compared with a 28.5% increased risk in those who quit using cannabis entirely (P less than .001).

The only true way to measure causality in this area would be to administer cannabis to participants in a randomized, clinically controlled trial, thus challenging the ethics of clinical study. Instead, Ms. Schoeler of the Institute of Psychiatry, Psychology, and Neuroscience at King’s College London and her associates – including Sir Robin M. Murray, MD, a psychiatrist who has been a leading voice in favor of the view that cannabis use and psychotic episodes are often linked – used patient self-reports to assess changes in cannabis use over time. This allowed the investigators to more clearly differentiate between confounders that cannot be readily observed, and that are not likely to change over time, such as any shared genetic or environmental factors, and confounders that can be observed and that do have the potential to change, such as the change in cannabis use status over time, including nonuser vs. user status, and the pattern and frequency of use after onset of illness.

The data were collected in two separate assessments, one was conducted near the time of illness onset, and again at least 2 years after the first episode of psychosis. If interviews were not possible to complete in person, they were conducted over the phone. These data were combined with patient clinical records across the study period. Frequency of cannabis use was measured based on the Cannabis Experience Questionnaire. A patient was considered to have relapsed if he or she was admitted to an inpatient psychiatric facility for exacerbated symptoms of psychosis within 2 years of the first episode of psychosis.

As for the study population, 59.1% of whom were men with an average age of 29 years, there were no significant differences as to whether they had been diagnosed with either affective or nonaffective psychosis. However, there were significant differences between the risk of relapse and the frequency of cannabis use 2 years after onset of illness. In the nonuser group, 28.5% of patients relapsed. In the intermittent user group, 36% relapsed. In the continuous use group, 59.1% relapsed (P less than .001).

Differences also were found across the three groups in medication adherence rates. Among the nonusers, 47.7% reportedly complied with treatment. The rate was 32% among intermittent users and 20.5% among habitual users (P = .02). Illicit use of other drugs was also higher in the continuous cannabis use group: 27.3% vs. 4% in the cannabis nonuse group and 12% in the intermittent use group (P less than .001).

Also noteworthy was the age of psychosis onset across the groups. While the average age of onset across the study was 28.62 years, for habitual users the average age of illness onset was 25.44 years, compared with 29.52 in nonusers and 28.79 in intermittent users (P = .02).

These results should be viewed in light of their retrospective and self-reported nature – neither of which take into account either premorbid cannabis use or age of illness onset as either predictors or moderators – but the data suggest a dose-response relationship between cannabis use and psychosis, according to the investigators. If so, the implications of the data could reach beyond patient outcomes.

“Because cannabis use is a potentially modifiable risk factor that has an adverse influence on the risk of relapse of psychosis and hospitalization in a given individual, with limited efficacy of existing interventions, these results underscore the importance of developing novel intervention strategies and demand urgent attention from clinicians and health care policy makers,” the authors wrote.

Whether the study conclusively demonstrates that it is cannabis use – and not genetic liability or reverse causation – that drives relapse of psychosis is still not clear. By the investigators’ own admission, the effect of discontinuing cannabis use on patient outcomes has not been tested, leaving the question of a definite causal relationship between continual cannabis use and relapse of psychosis still open for debate.

Ms. Schoeler and the other investigators – with the exception of Dr. Murray – reported no conflicts of interest. Dr. Murray reported that he has received honoraria from Janssen, Sunovion, Otsuka, and Lundbeck. The study was funded in large part by the U.K.’s National Institute for Health Research and by King’s College London.

On Twitter @whitneymcknight