Tuesday December 1 2009
The risk of psychosis is linked to the strength of cannabis
Smokers of the strong ‘skunk’ variety of cannabis are seven times more likely to experience psychosis, according to the Daily Mail.
The news comes from research comparing 280 people being treated for new psychosis with 174 healthy people. It found that the same proportion of people in each group had used cannabis, but skunk users were seven times more likely to have psychosis than users of less potent varieties.
The results of this research add further weight to the growing body of evidence on the dangers of cannabis use. However, while the study had a number of strengths there are also some issues that should be considered, such as the unusually high rates of cannabis use and unemployment in both groups. Ideally, this study should be followed by research that follows healthy cannabis users over time to see if they develop mental health issues.
Where did the story come from?
This research was carried out by Dr Marta Di Forti and colleagues from the Institute of Psychiatry, King’s College, London. The study was funded by the Maudsley Charitable Fund, and a grant from the National Institute of Health Research in the UK. The study was published in the peer-reviewed medical journal, The British Journal of Psychiatry.
Several other newspapers covered this research, most reporting that there is an increased risk of psychotic symptoms due to exposure to tetrahydrocannabinol (THC), the psychoactive ingredient of cannabis. Some say that other less potent forms of cannabis, such as resin, also contain substantial quantities of another chemical called cannabidiol (CBD), which researchers think might counteract the psychotic side effects of THC.
What kind of research was this?
This was a case-control study comparing the cannabis use of people with a first episode of psychosis (cases) with that of matched healthy people (controls). They specifically looked at whether people who developed psychosis were more likely to have used cannabis of high or low potency.
The control group was carefully matched to the cases on the basis of age, gender, ethnicity, educational qualifications and employment status. While the controls were not matched to cases in terms of cannabis use, the researchers found that around the same proportion of each group had used cannabis at some time in the past. Those that reported using cannabis had started at a similar age.
What did the research involve?
The researchers collected information on cannabis use from the case group, which comprised of 280 people presenting to the South London and Maudsley NHS Foundation Trust with a first episode of psychosis. They also obtained information from 174 healthy people (the control group), and recruited them through internet and newspaper advertisements, and leafleting at train stations, shops and job centres. Cannabis was not mentioned in these adverts.
The researchers used a Psychosis Screening Questionnaire to exclude anyone with a current psychotic disorder or a previous diagnosis of psychotic illness. The participants were then asked about their use of illicit drugs. Those who reported ever using cannabis were interviewed using the Cannabis Experience Questionnaire. This asks detailed questions about lifetime patterns of cannabis and stimulant use, including age at first use, frequency and duration of use, and the specific type of cannabis used.
The researchers report that the ‘skunk’ form of cannabis contains between 12% and 18% THC (the active ingredient) and less than 1.5% cannabidiol, a substance that is thought to be protective. In contrast, cannabis resin (hash) has an average THC concentration of 3.4% but a similar proportion of cannabidiol.
As with all case control studies, it is important to ensure that the cases and controls are as closely matched as possible for all features that could conceivably influence the results. In this study, the researchers asked about age, gender, ethnicity, educational qualifications and employment status of the cases. They then made appropriate adjustments for these in their analyses. Questioning the groups revealed that:
- Cases and controls were similar in that they were young (average age of 25 and 27 years respectively),
- Cases and controls were mostly men (72% and 65% respectively),
- Cases and controls had a high proportion of unemployment (58% and 43%).
- Cases and controls had a high proportion of participants identifying as Black Caribbean (19% and 21%) or Black African ethnicities (16% and 22%). Forty-two per cent of cases and 44% of controls identified themselves as white.
Although the study reported to have assessed the use of other stimulants, it is unclear which specific drugs were asked about or whether alcohol use was assessed.
Among the 340 potential patients with first episodes of psychosis, 60 (17.6%) refused to participate.
What were the basic results?
The researchers say that their most striking finding is that patients with a first episode of psychosis preferred to use high-potency cannabis preparations, such as skunk, rather than the less potent cannabis resin.
A similar proportion of cases and controls reported having used cannabis at some point in the past (56.9% of cases and 62.5% of controls). Users reported starting cannabis at a similar age, mostly before 17 years.
After the researchers made statistical adjustments, those in the cases group were more likely to be current daily users (OR 6.4, 95% CI 3.2 to 28.6), and to have smoked cannabis for more than five years (OR 2.1, 95% CI 0.9 to 8.4). Of those who used cannabis in the cases group, 78% used skunk, compared with 37% of the control group (OR 6.8, 95% CI 2.6 to 25.4). After adjustments, the chance of developing a psychosis when using skunk was reported as increasing almost seven-fold (OR 6.8, 95% CI 2.6 to 25.4) compared with using less potent varieties.
How did the researchers interpret the results?
The researchers say that their findings “are consistent with the hypothesis that THC is the active ingredient increasing risk of psychosis”. They say this has important public health implications, given the increased availability and use of high-potency cannabis, which contains high levels of THC.
The researchers say they are not surprised by the similarly high rates of cannabis use in both groups. They say that other research has shown that 40% of adolescents aged 15–16 years in the UK have used cannabis at some point.
This study has strengths and limitations. Among its strengths are the researchers’ careful adjustments to account for the influence of age, gender, ethnicity, other stimulant use, level of education achieved and employment status (all factors known to influence the incidence rates of psychosis). This study is also large for its kind. The size of the effect seen was also large and statistically significant.
However, because it is a case control study, it has the limitation of being unable to prove causation, i.e. that cannabis use causes psychosis. The researchers mention some further limitations:
- The proportion of controls who had ever used cannabis (62%) is higher than the national average, so it is possible that the recruitment strategy oversampled users of cannabis. However, it is likely that this would have actually reduced the strength of any association observed.
- It is possible that the recruitment of controls was biased towards selecting mild cannabis users, excluding heavy users who are perhaps more likely to use skunk. This sort of recruitment bias may have explained some of the differences between the groups. However, the researchers argue that this bias was not substantial enough to account for the large differences seen.
- The answers given in the questionnaire were not confirmed by objective measures of cannabis use, such as urine, blood or hair samples. This could have added to the reliability of the study, particularly the assertion that there is a dose-response effect (the suggestion that higher doses of THC lead to a greater risk of psychosis).
Overall, this study has been well-conducted. As a case control study, it may lead to further studies of this important topic. If the 40% prevalence of cannabis use among young people is correct, it may be possible to launch a cohort study that follows a representative group of users over time to assess the development of psychosis or other mental health conditions.