cannabis-psychosisWith the recent study from King’s College London linking “skunk” to diagnosis of psychotic disorders, I thought it would be a good time to examine the link between cannabis and psychosis in detail.

I have previously written about cannabis and psychosis while talking gangs and drugs, but we didn’t look at the evidence base.

As I was writing this post, I realised that I’ve also waded into the fields of statistics and research methodology. Hopefully, this will provide some clarity the next time a newspaper starts talking about odds and risk in healthcare.

First, let’s get some definitions on the table.

Psychosis, in essence, is the inability to distinguish what is real and what is not. It is most often talked about it terms of schizophrenia. You can read more detail about it here.

Cannabis is a group of flowering plants native to Central and South Asia. To get biological, there are three main species – Cannabis sativa, Cannabis indica, and Cannabis ruderalis. However, the cross-breeding produces all kinds of different sub-species.

Sativa is the most prevalent and the one used for hemp production. Skunk is a hybrid of sativa and indica and contains two-three times higher concentrations of tetrahydrocannabinol (THC), the main psychoactive chemical in cannabis. UK police seizures indicate that skunk is now the prevalent form of cannabis in the UK.

Cannabis and psychosis: the evidence
People have been looking at cannabis in relation to its mind-altering effects ever since they discovered it could be ingested, which is why a number of ancient cultures used it in religious rituals.

However, scientific research didn’t kick off until the 20th century (before which we used to give morphine and cocaine to babies). The key thing about research is this: you have to know what the researchers were looking for. Research can only answer the specific questions asked – remember that as we take a wander through the history of cannabis and psychosis research:

The most famous study of cannabis and psychosis is known as the Swedish conscript study, published in 1987. Firstly, it was HUGE – 45,570 military recruits followed up for fifteen years. Cohort studies don’t get much better than that.

What’s a cohort study, you cry? It’s when a group of people are followed up over a long time to see what happens to them. It is only observational not interventional – you find what’s there and don’t try to influence the outcome. It is a good research method for looking at causal association – the question of “does x increase the risk for y?”. It’s the method that was used to look at links beteween smoking and lung cancer. However, in defiance of the Latin phrase, “after it therefore because of it” is not always true.

What they found was interesting – “high consumers” of cannabis were six times more likely to develop schizophrenia than non-users. And by “high consumers”, they meant people who had smoked cannabis more than fifty times. Ever. That could be only once a week for a year.

Two points also worth noting:
One – psychosis does not equal schizophrenia. However, this was 1987 and “schizophrenia” was used fairly liberally. We might now recognise schizoaffective disorder, drug-induced psychosis and mania separately.
Two – this was self-reported. Therefore, it’s likely that people underestimated their usage. All told, this is pretty good evidence for a link between cannabis and psychosis.

To reinforce these findings, a Lancet 2007 meta-analysis – i.e. looking at the whole body of research – explored the assocation between cannabis and psychosis. Note here that we’re including studies which didn’t look at cause and effect, but looked at where x and y appear together (though they only included longitudinal studies). The study found that there was a significant association between cannabis and all psychosis (not just schizophrenia) which went beyond the transient intoxication effects. It also concluded there is a dose response – like smoking and lung cancer, the more you smoke cannabis, the more likely you are to develop psychosis.

Skunk and psychosis
Which brings us to the King’s College study. First question – what were their questions? The full title of the study is “Proportion of patients in south London with first-episode psychosis attributable to use of high potency cannabis: a case-control study”.

So, we’re not looking at schizophrenia here but “first episode psychosis”. This is relatively new term arising from the movement for early intervention in psychosis, which is the idea that we prevent long-term disability by throwing in a lot of support at the beginning of the illness. First episode psychosis can be just a one-off episode, or can be the start of schizophrenia, schizoaffective disorder, or bipolar affective disorder.

You will also have caught the word “attributable”, so it seems we’re thinking causation again. However, the methods used can only conclude an association. The introduction goes on to say that the researchers hypothesise that the frequency of use and the potency of cannabis increases risk – i.e. a dose response. They think we should focus more of types of cannabis used, using alcohol as a comparison – it’s not enough to know someone drinks twice a week, but how much they drink per session and what they drink (wine, beer, spirits) matters to the outcome.

“Case control” is a type of study where people with a problem – in this case, first episode psychosis – are compared directly to individuals without that problem but otherwise similar. This helps reduce confounders, factors that could potentially influence the outcome, like gender, ethnicity or socioeconomic status.

Their findings support an increased association between users of skunk and psychosis, and a stronger association for daily use. Media reporting has focussed on this figure of 24% new cases of psychosis related to skunk. The study has calculated this as a population attributable fraction. This means that if you eliminated skunk from the population of South London, you would prevent almost a quarter of first episode psychosis cases in this area. However, this would be unlikely to be the same for Cardiff or the Scottish Highlands.

Conclusions about cannabis and psychosis
With the research evidence we have to date, we can conclude:
– Cannabis and psychosis are strongly associated
– Cannabis has a dose-related association with psychosis – i.e. the more you smoke and the more potent the cannabis, the more like you are to have psychosis
– If we eliminated all cannabis, we would significantly reduce psychosis

This is obviously just the tip of the iceberg. We have not considered how cannabis affects people who already have a diagnosis of psychosis. We also haven’t looked into what influence other drugs might have on the illness.

As my role with Freudian Script is to look at fictional portrayals of mental health, we don’t see a lot of cannabis and psychosis. The amotivational and appetite-inducing side effects are well-covered, but the “stoner movie” subgenre focussses on the comedy not calamity associated with cannabis use.

What are your impressions of cannabis and psychosis? What does this research add to the public debate around drugs? Should Hollywood make more films about psychotic stoners? Let me know in the comments!


  • This reminds me of the experiences of Jon Snow on skunk which I’d read about recently. I do wonder what makes some people have “bad trips” while others see and experience pleasant things while in the psychotic state. Is the psychotic experience linked to the memories and experiences of the individual? Their emotions maybe? If these are factors influencing the experience of the psychosis then you could apply your fictional character’s personality and backstory to estimate how they’d experience drugs like skunk and cannabis.

    • The incidents of psychosis are documented outside the “intoxication” effects – i.e. what people may experience while high. Psychosis in this context refers to the disease state. Though you make an interesting point, because both intoxication and psychotic experiences are highly individual.

    • Fernando

      It is not surprising that someone get an anxiety/panic attack just by going through a MRI scan. If you add the effect on any psychoactive substance this claustrophobic/anxiety effect can be altered too. about this particular case you can read here a comment published in the Guardian from Val Curran, professor of pharmacology at University College London that can bring some clarity about Jon Snow experience in the MRI scan

      “Jon Snow’s very negative experience was therefore understandable even though it was uncommon. The vast majority of our volunteers completed the brain scanning session and all the other tests, and many of them said that they enjoyed the experience on cannabis more than on placebo. Jon himself carried out all the tasks outside the scanner and apologised profusely that he couldn’t tolerate being inside the scanner.”

      • Which is why we don’t base our knowledge of drug effects on one person performing a task on television.

        Anxiety is a common side effect of cannabis use, though more after the intoxication effects have worn off. However, enduring psychosis is very different from anxiety.

        Thanks for commenting.

  • Not a bad assessment. May I just make two points?

    You’ve walked straight into the misuse of the term ‘skunk’. There is only one way to use it in a scientific context and that is accurately!! It’s the name of one strain of cannabis, not particularly potent, developing about 9% THC if expertly grown.

    Secondly, The 24% figure is a statistical trick. Commonly the total of ‘population attributable fractions’ (PAF).for all risk factors add up to more than 100%. It’s a hypothetical estimate and relies on an assumption of causality which the authors of the study themselves say is not shown.

    The actual evidence of cannabis use correlating with psychosis is that the risk is extremely low:

    There is by no means a consensus amongst leading psychiatrists that ‘cannabis psychosis’ even exists:

    • Hi Peter – thanks for commenting.

      I agree that accurate usage of the word “skunk” refers to this particular strain. However, the common usage of the term refers to high-potency cannabis. It’s unfortunate that the King’s College study uses the terms interchangeably. Most usage studies don’t biologically test what people are smoking but rely on self-report, which again confuses the issue.

      To accurately test the enduring effects of “grass” v “skunk”, you would need to perform an experimental test while using two known varieties – however, I doubt you’d get ethical approval for it!

      PAF isn’t a very useful statistic, because it has little real world application – you cannot eliminate cannabis or skunk from the South London population to see what would happen in ten years. The media are using it inappropriately, as it is also a localised stat.

      Psychiatrists – and scientists, for that matter – have yet to agree as a body on anything. That’s why we have research, and meta-analysis is the strongest research method we have. Proving causation is extremely difficult, unless you have a direct experiment in animal models or humans – i.e. give cannabis, control all other factors, see what happens.

      However, the association link is too strong to ignore. The stress-vulnerability model works much better here – people who are genetically/biologically predisposed to developing psychosis receive a trigger in the form of cannabis and go on to develop psychosis.

      Discussing the realities of what we know about cannabis and psychosis, as far as we are able to know it, is an important part of the legalisation debate. Alcohol and tobacco are legal, and they have known, serious health complications related to dose-response. Cocaine and ecstasy, for example, have far fewer health problems associated with them. Legality and health effects are not necessarily associated. 😉

  • Being followed and under constant surveillance(EU report about ECHELON’s economic and private espionage dated 07/11/01 etc.) by a counterproductive(Al Capone, anyone?), corrupt(Air America, ARTICHOKE, BLUEBIRD, CHATTER, COINTELPRO, Dark Heart, Iran-Contra, Fast and Furious, Gunrunner, HSBC, Midnight Climax, MKULTRA, N987SA, etc.) and pharmacologically fascistoid(human rights, anyone?) police state which enforces a historically falsified prohibition policy, that effectively only prevents any control of access, age and quality for the multi-billion dollar corruption pyramid of a tax-free Mafia monopoly, should be eliminated first as a factor influencing the feelings and thoughts of Cannabis consumers, or you have no clean set and setting for any really scientific research. I think that is the most important, but constantly overlooked factor in this whole discussion. I know that my first time consuming Cannabis in Amsterdam was radically different from my previous experiences at home and definitely more beautiful, benign and way more relaxing(I couldn’t even believe that it’s possible, before experiencing it myself…). I had the same highest grade buds from the same strains before and I remember how I had a drastically different mindset during my whole stay, especially when I saw Dutch police officers. But I also consumed some Cannabis extracts there, which were way more potent than those I had used before and I didn’t become paranoid from any of them, not even temporarily. If you don’t have to fear that some fascistoid sadist won’t complicate or destroy your existence for a victimless, so-called “crime”, that is simply deciding for yourself what enters your own adult body and modifies your own free mind in a so-called democratic state, it just creates a better and radically different set and setting for your whole experience. I can guarantee that and know that this factor is a common issue among most Cannabis consumers who are criminalized, discriminated and stigmatized due to unjust laws. This constant threat to your individual life design can create a lot of latent and manifested aggression, that would never build up and unfold without these unfair threats to your own free self-determination as an adult, especially when you realize, that these police officers could catch real criminals, like child rapists(Bunga Bunga, IROJ etc.), corrupt politicians or other real criminals with involuntary victims, like stay-behind state terrorists(Gladio, ÖWSGV, Propaganda Due, P-26, PNAC, SDRA8, Satanique) in the mean time, instead of wasting taxes. So in addition to the repitition of a counterproductive and historically falsified policy that creates untaxed profits for violently competing criminal cartels, the creation of a cutting agents and contaminations problem and the lack of any access, age and quality control, there can’t be any real scientific research about this substance, without eliminating this methodological problem first. Plus, a lot of schizophrenic people calm themselves effectively with Cannabis(especially with CBD-rich Indica strains or hash) and can enjoy life again, so often it just became their medicine of choice, as a lot of neuroleptics have strong side effects, which often outweigh their therapeutic potential. Another very similiar methodological issue is created, when scientists assume people are in voluntary treatment for Cannabis consumption, when in reality they’re forced into this treatment by the state to reduce their sentence. Naturally most people will choose treatment over jail time then, as the traumas induced by a prison sentence can damage you permanently and change your persona in a much more threatening way than Cannabis probably ever could, especially when you’re already a vulnerable individual. Also, when you effectively claim addiction and/or insanity, it can reduce your sentence, too.

    • Thank you for your comments. You make several interesting points about the cultural context in which cannabis is consumed and how that influences the manifestation of paranoia and illness.

      We cannot eliminate society from clinical research, because otherwise research becomes pointless. What is the value of research if it only applies to an artificial population that could not possibly exist in reality? Also, the Swedish conscript study took place in the 1970s and 1980s, which was a very different cultural context to the present. The research findings are consistent across time periods and national cultures.

      When I talk about psychosis, I am specifically talking about the enduring disease states – not the transient paranoia that many experience during intoxication and immediately after.

      Whether cannabis can in fact be useful for people experiencing psychosis is interesting. The current research is inconclusive (see this helpful Cochrane Review: My own clinical experience has been that any improvement in sleep is traded off with rebound anxiety, increased paranoia and increased risk of relapse.

      Treatment for cannabis consumption is an odd one, because there is little evidence for physical dependence but quite a lot for psychological dependence. I don’t think we have mandatory treatment here in the UK, so I’m not sure of the impact.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.