Cannabis and mental health

Is there a link between cannabis, psychosis and schizophrenia?

Towards the end of January 2004, the media began to report stories about the links between cannabis and mental disorders such as schizophrenia. The allegations are not new - for hundreds of years there have been claims that cannabis causes insanity. Although even more recent studies looking critically at this have generally found little or no evidence of a causal link, there is certainly evidence that cannabis makes the symptoms of schizophrenia worse in some people.

Before interpreting reports on this subject it is important to consider the difference between causation and correlation - a difference that is often ignored by the media and indeed some "experts" who often confuse the two.

Statistics tend to show that the phenomena of correlation between cannabis and, for instance, schizophrenia exists - that is that on average, a person who uses cannabis is more likely to be schizophrenic than one who does not. What they do generally not show is causation - that the use of cannabis causes an otherwise "normal" person to become schizophrenic.

This is a vital difference to understand when weighing up the risks of cannabis usage.

There have been many theories suggested and investigated as to why correlation may exist in the absence of causation. These include suggestions that people already suffering from psychoses may be more likely to try cannabis - a reversal of the causation theory. People already suffering from such disorders, diagnosed or not, may actually use cannabis as a medication to ward off the unpleasant effects of the illness. People already diagnosed as having psychiatric problems may find cannabis helpful in dealing with the side effects of the medication that is given to them. Another theory is that there is some other factor(s) that makes both schizophrenic disorders and cannabis usage more likely to occur. This may be down to demographics, environment, personality types or some completely unrelated factor. Several studies have found that cannabis use follows the onset of mental disorder symptoms, which adds to the refutation of cannabis causing such disorders.

As a result of this lack of causation evidence, many reputable bodies have disregarded, or found it unlikely, that cannabis causes schizophrenia or other mental disorders per se. A few examples:

The Advisory Council on the Misuse of Drugs reported that "The other main concern about the chronic use of cannabis is whether it can lead to mental illness (especially schizophrenia). ... no clear causal link has been demonstrated." (from 'The classification of cannabis under the Misuse of Drugs Act 1971')

The House of Lords Science and Technology Committee wrote that "...we do not believe that it can cause schizophrenia in a previously well user with no predisposition to develop the disease" (from 'Cannabis: The Scientific And Medical Evidence' )

The Royal College of Psychiatrists has been quoted as saying that "there is little evidence that cannabis use can precipitate schizophrenia or other mental illness in those not already predisposed to it" (quoted in the Lord's Hansard text of 14th January 2004, specifically here).

However, a couple of weeks before the reclassification of cannabis was due to take place, the media was reporting expert's opinions in a way which suggests there is a direct causation link after all, citing various studies that purport to show this link, and drawing conclusions that for instance cannabis can cause psychoses such as schizophrenia. These reports are being used both as a health warning, and also in an effort to argue against the reclassification of cannabis and to reduce the hope of future legalisation.

Three of the studies most often mentioned were published in the British Medical Journal at the end of 2002. These are:

  • Patton et al. "Cannabis use and mental health in young people: cohort study"
    BMJ. 2002 Nov 23;325(7374):1183-4.

    This study categorised school-going adolescents in terms of cannabis use - whether they used cannabis never, less than weekly, at least weekly and daily use. Throughout the period of the study, they were tested to assess depression and anxiety. After using various mostly undescribed statistical adjustments, the sole resulting statistically significant conclusion was that "Frequent cannabis use in teenage girls predicts later depression and anxiety, with daily users carrying the highest risk". Interestingly, when examining the data, frequent cannabis use by men appears to reduce the chances of depression and anxiety by over 50%. This figure was however not statistically significant in the study, and no mention was made of it. Nonetheless it may be an interesting avenue to follow up.

    They also suggest an explanation other than cannabis causing mental disorders in their discussion saying:
    "Psychosocial mechanisms - for example, the adoption of a counter­cultural lifestyle possibly underlie the association. Social consequences of frequent use include educational failure, dropout, unemployment, and crime - all factors that may lead to higher rates of mental disorders.". It could therefore be argued that as it is largely the prohibition of cannabis that keeps an association with crime and expulsion / sacking are sometimes a consequence of being caught with an illegal substance, that if this theory is correct, not only does cannabis not directly cause anxiety and depression, but the prohibition of cannabis actually adds to it.


  • Arseneault et al. "Cannabis use in adolescence and risk for adult psychosis: longitudinal prospective study"
    BMJ. 2002 Nov 23; 325(7374): 1212-3.

    This study divides a population of adolescents up into groups that have never used cannabis, that have used cannabis by the age of 15 and those that have used cannabis by the age of 18. It then attempts to measure whether or not they have either schizophreniform disorder or depressive symptoms. The authors note no major effect on depressive symptoms. They do however conclude 'Using cannabis in adolescence increases the likelihood of experiencing symptoms of schizophrenia in adulthood' and suggest that "people who used cannabis by age 15 were four times as likely to have a diagnosis of schizophreniform disorder at age 26 than controls". There was however no statistically significant increase in likelihood of schizophreniform disorder in those who only started using cannabis by age 18.

    They went on to attempt to compensate for other drug use, and the same results were found. These adolescents had previously been testing for psychotic symptoms at the age of 11. Using this information they adjusted the original figures to compensate for childhood psychotic symptoms. At this point, the increase in likelihood of schizophreniform disorder for those who had used cannabis by the age of 15, whilst it remained on average higher than the controls, was no longer statistically significant. Critics of this study have claimed that this control for psychotic symptoms at 11 may not have much meaning, as the symptoms (or absence of symptoms) may be entirely different by the age of 15 or 18. "Full blown schizophrenia" for instance, typically only becomes apparent between the ages of 18 and 25.

  • Zammit et al. "Self reported cannabis use as a risk factor for schizophrenia in Swedish conscripts of 1969: historical cohort study" BMJ. 2002 Nov 23; 325(7374): 1199.

    This study looks at a population of 50,087 Swedish men, conscripted to take part in military training, almost all of which were aged 18-20. It measures reported level of use of cannabis use admitted to prior to conscription, and examines the proportion of each category of user who end up being diagnosed as schizophrenic. They conclude that "Cannabis use is associated with an increased risk of developing schizophrenia, consistent with a causal relation." A previous study had used this same population and produced similar conclusions, but had been largely discredited as anything other than correlation, as amongst other reasons, under further investigation it turned out when a subset of the population was taken, half of them had also used amphetamines, which have been known to precipitate schizophrenia. There were also several other non-drug related common factors relating to all the schizophrenic population studies.

    This report tries to remove the some of these issues and look only at people who had admitted to using cannabis and no other (illegal) drugs prior to conscription. They also try to adjust the results for external factors they consider significant. The outcome is that for most of the categories of frequency of use there are no statistically significant differences between cannabis users and non-users in terms of developing schizophrenia. Indeed using cannabis between 11 and 50 times prior to conscription appeared to reduce the chances of developing schizophrenia somewhat - although this too was not a statistically significant nature. However for people who used cannabis more than 50 times there appears to be a very statistically significant increase in the chances of being diagnosed schizophrenic.

    Critics of this study claim that the data is not complete enough to draw conclusions about causation, for example there is no data on what drugs were used by the conscripts after they were enrolled, and many of them were not diagnosed schizophrenic until years after. We do not have data as to any psychiatric disorders prior to cannabis use. As a side note, another (unmentioned) result this study seems to show, when discussing other factors that may influence becoming schizophrenic, is that on average the people who had used cannabis prior to conscription had a higher IQ than those who hadn't!

Any criticisms above should not be taken as an instruction to ignore the studies results - certainly those results that they actually described had rather than those that were sensationalised in the media. They certainly do not prove that there is no link between cannabis and mental disorders. However one may wish to use caution when interpreting the results. Research on illegal drugs, as used in the real world, is notoriously difficult for numerous reasons. Generic criticisms of some of the above studies and more include the fact that they all relied on self-reporting of cannabis (and other drug) use. There are several reasons a person may lie about their drug use, not least due to any self-perceived stigma surrounding their use, or simply having a bad memory! In each of the studies above the researchers have wisely tried to compensate for possible non-cannabis related factors such as use of other drugs or previous psychotic symptoms - however these are (often undocumented) statistical corrections and can only at best approximately reflect the influence they might have in the real world. It is impossible to adjust for absolutely everything that might influence a person's life, so an arbitrary decision has to be as to which events or circumstances to ignore, with the result there may well be further unknown confounding factors. Also importantly is the lack of scientific control of exactly what substance is used. It is widely known that cannabis is often impure - whilst a substance remains illegal it is impossible for quality control to take place - for instance soapbar is often mixed with various unsavoury substances. Some of these substances have been known to be, for example carcinogenic, so it is not impossible to envisage that some might increase the odds of psychosis. Likewise "drug" is often substituted for "illegal drug", meaning that alcohol and tobacco use is often not taken into account. The latter is particularly important as the majority of users in this country use tobacco in their cannabis spliffs.

Other people have performed epidemiological research, which look at the distribution and rate of increase of schizophrenia and other mental disorders as time goes past. Some studies have compared the level of schizophrenia in a society with the level of cannabis use. Rather than looking at individuals, they use the logical premise that if cannabis use causes a disorder, the more cannabis is used by a population the more cases of that particular disorder we will see. However this is rarely found to be the case. Data from a report "Cannabis and psychosis. Is there epidemiological evidence for an association?" (The British Journal of Psychiatry 157: 25-33 (1990)) appeared to show this is not the case. Likewise, the recreational use of cannabis became most popular in Britain around the beginning of the 1960s. Use has almost constantly risen sharply since then, yet when examining rates of schizophrenia between 1952 and 1986, Der et al. found that "there has been a substantial decrease, beginning in the mid-1960s, in the incidence of schizophrenia" ("Is schizophrenia disappearing?", Lancet. 1990 Mar 3; 335(8688): 513-6). More recently in 2003, Degenhardt et al. looked specifically at this issue in a report "Testing hypotheses about the relationship between cannabis use and psychosis" (Drug Alcohol Depend. 2003 Jul 20;71(1):37-48) and concluded that using their data "Cannabis use does not appear to be causally related to the incidence of schizophrenia". However they did find that it is possible that "its use may precipitate disorders in persons who are vulnerable to developing psychosis and worsen the course of the disorder among those who have already developed it".

During the run-up to reclassification, the media appeared to be reporting stories regarding cannabis causing schizophrenia. However, from the above it seems apparent that some caution should be taken before accepting this as the proven truth. No mechanism of such a cause has been discovered, and the sensationalist way in which it is reported, when contrasted with the epidemiological research mentioned above, makes them seem more like scare stories. However, it is best to err on the safe side - as uncertain as the connection between cannabis and schizophrenia may be in contrast to what has been reported, most reports suggest a correlation between schizophrenia and cannabis use, and it has not been definitively proven that there is no causal connection. The development of schizophrenia or another mental disorder, however infrequent, ruins people's lives, so it would be wise to keep researching this matter in the hope that a definitive answer can be found, and moreover whatever information is found as to the causes of such illnesses used to reduce the incidence of schizophrenia and other psychoses as much as possible.

Consequences for the legalisation argument

If there was to be a connection, causal or otherwise, between cannabis and schizophrenia for example, the evidence seems to show that it occurs very much predominantly in those people who use cannabis very frequently at a very early age. Constant daily use of cannabis in the early teens seems to be a good predictor of increased chances of later mental problems, irrespective of whether one causes the other. Common sense would suggest that it is not a good idea for young people still developing and maturing both physically and mentally to frequently use any drug at this time for a number of reasons. We should seek to minimise this frequent use by children as much as possible in our country's drug policy. The prohibition of cannabis, whether it is in class B or class C, does nothing to address this problem, and indeed exacerbates it. Under prohibition, the widespread use of cannabis is entirely uncontrollable. It is impossible to regulate the massive black market in which the supply of cannabis takes place, often in the hands of organised criminals out to make a profit, whatever the cost. This being the case, a dealer may well sell cannabis to anyone who can pay, irrespective of his or her age. In order to establish the necessary age controls that could prevent young children from using cannabis and putting themselves at risk, similar to those surrounding alcohol and cigarettes, cannabis must be legalised and securely regulated. No matter what one's perspective on psychosis connections is, cannabis use, like almost everything else in the world, not entirely harmless and it is important that we bring it under control - a situation which prohibition entirely prevents. It may seem a big step, but the status-quo undoubtedly adds harm to the use of cannabis.

In a similar vein, legalising and regulating cannabis would give us some control over the substance actually sold as cannabis. As mentioned above, much cannabis is impure and mixed with various possibly harmful substances. Whether by accident, or more usually to increase profits, this adds a further level of danger to its use. Adverse effects people may get when using cannabis, may actually be caused or exacerbated by whatever unknown substances are contained within the "cannabis" the black market has supplied them. Finally, even in its pure form, not all cannabis is the same. If frequent cannabis use is found to cause problems, then it is likely that these problems will exert a dose-response relationship - that is the stronger the cannabis, the bigger the risks. The unregulated market currently provides cannabis with a THC content ("strength") typically from 2% - 20% often without differentiation. If legal, cannabis could be graded by strength. In a similar way to the recommended alcohol - a drug which is known to cause alcohol psychosis under certain circumstances - unit limit advice given by the Government, guidelines could be researched and provided as to what a sensible level of use might be for an average person, and education could be provided on the risks of irresponsible use. All these harm-reducing and potentially life-saving matters are impossible under prohibition, under which no control can be had over cannabis use whatsoever. Even, or rather especially, if cannabis is proven to cause any form of psychosis, schizophrenia or other mental disorders, the only appropriate response would be to legalise the industry and regulate the use which history demonstrates is impossible to prevent and is becoming ever-more prevalent.

Further reading

BMJ articles
BMJ editorial comment 2002 (£)

Further reading
Causal association between cannabis and psychosis - examination of the evidence.
Early adolescent cannabis exposure and positive and negative dimensions of psychosis.
Effects of cannabis and psychosis vulnerability in daily life: an experience sampling test study.

UKCIA Risks Section

Does Cannabis Make you mad? UKCIA blog analysis