Cannabis use can lead to relapse in psychosis, partially because patients stop taking medication


A lot of research has been conducted on cannabis and psychosis. Previous studies reviewed here on the Mental Elf have shown that smoking high potency ‘skunk’ appears to increase the risk of psychosis (Kennedy, 2015) and that cannabis use increases the likelihood of having a psychotic episode in those who are deemed to be high risk (Walsh, 2015).

There is some disagreement in the literature about causality (which came first: the spiff or the psychosis). There is also uncertainty around extent to which cannabis use can induce psychosis that would not have already happened: you are more likely to get psychosis from cannabis use if you are already high risk for psychosis (Henquet et al., 2015). There is also considerable interaction with other risk factors such as sexual abuse (Konings et al., 2012).

What is less controversial is that cannabis is commonly used in those with first episode psychosis: 34% are currently using (Myles et al., 2016). There are also numerous studies showing a negative impact on cannabis use in those with first episode psychosis, such as higher rates of relapse and hospitalisation and poorer functioning. There has however been less research on why this is the case. Previous research has shown that those who are using cannabis might be less adherent to medication; this study set out to see if adherence to medication plays a role in the impact of cannabis use on psychosis.

Those with a first episode of psychosis are more likely to use cannabis.

Those with a first episode of psychosis are more likely to use cannabis.


245 patients with first episode psychosis (non-organic or affective) were recruited from south London, ages 18-65. Of these:

  • 40% had never used cannabis,
  • 20% previously used,
  • 15% were ‘intermittent’ users
  • 26% continued to use cannabis frequently (at least once a month for the two year period).

They were followed up for two years after their initial psychotic episode with medical records screened for details on relapse, and a structured questionnaire to collect information about cannabis use. The Life Chart Schedule was used to group participants as ‘non-adherent’ if they were non-compliant 67-100% of the time.


As expected, continuing to use cannabis and poor medication compliance were both linked independently to poorer outcomes: more likely to relapse, sooner, more often and to need more intense care over the two year period. Longer relapses were linked with cannabis use but not medication.

The authors used structural equation modelling (statistics which make my brain hurt!) to show that adherence was partially responsible for the impact of cannabis use on:

  • Increased risk of relapse
  • Shorter time to relapse
  • More relapses
  • Greater care intensity needed

Overall between 20% and 36% of the effect of cannabis on these outcomes were due to poor medication compliance.

The authors were aware that the relationship might work another way: those who stopped taking medication might be more likely to use cannabis, perhaps as self-medication. However there was no evidence to support this pathway.

Both cannabis use and poor adherence to medication increased the risk of being re-admitted to hospital.

Both cannabis use and poor adherence to medication increased the risk of being re-admitted to hospital.


Previous research has shown that cannabis use is linked to poor outcomes and higher risk of relapse in psychosis.

This study adds that part of this is due to poor medication compliance in those who use cannabis. It explains about 30% of this relationship, so a number of other factors account for poorer outcomes for those with psychosis who use cannabis.

Strengths and limitations

This study is well conducted with a robust methodology in particular a longitudinal design, and measuring real-world clinical outcomes in terms of re-admission details from medical records. A number of potential confounding factors of ethnicity, care intensity for initial episode and other drug use are also controlled for. After reading it I feel confident that poor medication compliance is certainly a factor in poorer outcomes for cannabis users.

However some questions remain. As the authors acknowledge, their study showed this was a factor, but not why it was: the mechanisms underlying this are not known. One possible factor not considered is the role of self-medication: those with psychosis sometimes use cannabis to cope with symptoms such as depression (Green et al, 2004).

The authors acknowledge a limitation that those under 18 were excluded, but they don’t acknowledge that research has shown those who use cannabis tend to have a first episode of psychosis on average 2.7 years earlier than those who don’t use cannabis (Large et al., 2011). This is a potential confound: cannabis users might be more likely to be excluded from this study due to a greater likelihood of first episode under the age of 18.

The authors claim that those using cannabis were no less likely to take part in the study, but they cannot test this as consent was not given. 33% of those invited refused to take part, and I wonder if those who refuse to take part might be more likely to be cannabis users as research shows substance use predicts overall disengagement with services. I have certainly found from my ongoing study of cannabis and psychosis that many with historical cannabis use take part, but few who are currently using opt-in despite us knowing it is commonly used in our patients. In addition, those who do not opt in might be less compliant with medication: If you do not think you have psychosis and therefore do not feel you need medication, why would you take part in a study on medication adherence in psychosis?

It is possible that those who were disengaging from interventions such as medication were less likely to take part in the study.

It is possible that those who were disengaging from interventions such as medication were less likely to take part in the study.

Implications for practice

The authors suggest a need to develop interventions to reduce cannabis use and also increase medication adherence in those with psychosis. An accompanying commentary by Engh and Bramness (2017) discusses possible medication to reduce the urges to use cannabis.

Certainly this study demonstrates the importance of addressing cannabis use in those with first episode psychosis, and considering that they might be more likely to stop medication and relapse as a result. It is worth considering here previous research on the potential of CBT for those who refuse antipsychotic medication (Tomlin & Badenoch, 2014).

This study showed that cannabis use leads to poor mediation compliance, but not why this is the case. Therefore at present implications are fairly general. If further research shows specifically why this is the case then more specific interventions can be developed. I suspect that self-medication might play a role here, and this lends itself well to psychological therapies, for example providing relaxation strategies for anxiety and behavioural activation for depression to reduce the need for cannabis use as an ‘alternative medication’. Alternatively CBT or Motivational Interviewing could be used to test a client’s possibly inaccurate belief that cannabis use improves their mental wellbeing.

We need to consider the impact of cannabis use on medication compliance, but we need to know more about why this is a problem in the first place.

We need to consider the impact of cannabis use on medication compliance, but we also need to know more about why this is a problem in the first place.

Conflicts of interest

I am currently writing up two papers on psychological mechanisms of the relationship between cannabis use and paranoia which was partially funded by Research Capability Funding by the National Institute for Health Research.


Primary paper

Schoeler T, Petros N, Di Forti M, Klamerus E, Foglia E, Murray R, Bhattacharyya S. (2017) Poor medication adherence and risk of relapse associated with continued cannabis use in patients with first-episode psychosis: a prospective analysis. The Lancet Psychiatry DOI:

Other references

Engh JA, Bramness JG. (2017) Psychosis relapse, medication non-adherence, and cannabis. The Lancet Psychiatry 2017 4(8) 578-579. DOI:

Green B, Young R & Kavanagh D. (2005) Cannabis use and misuse prevalence among people with psychosis. British Journal of Psychiatry 187, 306-313.

Henquet C, Krabbendam L, Spauwen J. et al (2005) Prospective cohort study of cannabis use, predisposition for psychosis, and psychotic symptoms in young people. British Medical Journal, 330, 11-14.

Kennedy E. (2015) High potency cannabis and the risk of psychosis. The Mental Elf, March 24 2015.

Konings M, Stefanis N, Kuepper R, et al (2012) Replication in two independent population-based samples that childhood maltreatment and cannabis use synergistically impact on psychosis risk. Psychological Medicine, 42, 149-159. [PubMed abstract]

Large M, Sharma S, Compton MT et al. (2011). Cannabis use and earlier onset of psychosis: a systematic meta-analysis (PDF). Archives of General Psychiatry 68 555-561.

Myles H, Myles H & Large, M. (2016) Cannabis use in first episode psychosis: Meta-analysis of prevalence, and the time course of initiation and continued use. Australian and New Zealand Psychiatry 50(3) 208-219. [Abstract]

Pilot study suggests that CBT may be a viable alternative to antipsychotics for people with schizophrenia, or does it?

Alcohol use confounds the relationship between cannabis use and conversion to psychosis

Photo credits

Share on Facebook Tweet this on Twitter Share on LinkedIn Share on Google+