I set out to write an article about cannabis use and mental health. Yet, going through the scientific research and journal articles, something hit me and it hit me HARD.
As someone that worked as a behavioural scientist with just under a decade’s worth of academic research under the belt, I was struck by the fundamental issues and limitations that seem to be inherent — nay, ingrained — in cannabis research when it comes to the social and behavioural sciences.
Not only did this lead me down a rabbit hole of problems and issues found within the cannabis literature, it also really made me question the validity of a large proportion of the reported results.
Medical marijuana and recreational cannabis, as well as the legalisation thereof, is a hot topic at the moment. Increasingly, governments and states worldwide are entering into discussions about the ratification of cannabis laws.
At the same time, more and more people of all age groups, socioeconomic strata and cultural backgrounds are smoking pot for a variety of reasons, ranging from mitigating pain to enjoying it as an alternative to alcohol. Moreover, despite years and years of warnings from governments and scientists alike, most do so without developing a dependence or suffering any debilitating side-effects.
Yet, scientific study after scientific study claims that there is a definitive link between cannabis dependency and mental disorders. Any search of any academic database brings up paper after paper touting the dangers of cannabis. This 2007 study even definitively concluded that:
“A range of psychopathological conditions, including depressive, … and … anxiety disorders as well as the degree of their comorbidity are significantly associated with incident [cannabis use] CU and progression to [cannabis use dependence] CUD, even when controlling for externalising disorders.” (note the words externalising disorders – not factors)…
And, despite the odd article vaguely referring to cannabis as not being THAT harmful to your mental health (mostly coming from the crowd doing research on the synthesised versions), the overarching message is clear: One should assume that the use of cannabis will lead to dependency and all the problematic patterns that are associated with it and that, therefore, all use is harmful use.
But is this really so?
The way in which cannabis research is conducted, especially in relation to mental health, simply sucks. No way to sugar coat it and no way around it. There are major and fundamental limitations when it comes to cannabis research and these issues are so hardcore that it makes me severely doubt the validity of the majority of findings coming out of it.
The issues range from the classification systems used to categorise users, to methodological and statistical reporting issues to the types of sample populations and recruitment methods generally employed within studies.
Further considerations – albeit mostly political yet completely related – are that, for the most part, cannabis is still a restricted substance, thus making the research itself illegal and therefore limited. Similarly, the studies that are conducted, tend to be funded by government organisations, pharmaceutical companies and/or universities (the latter being dependent upon the aforementioned for grants), all of whom have a vested interest in keeping the status quo – that of a monetised industrial-pharmaceutical complex making billions each year off of mental health disorders.
For simplicity sake, I have divided the problems surrounding cannabis research into two groups and as such, will discuss them separately. The first issue is the scientific limitations, with the second being political issues (which will be discussed in a later post).
Let’s look at the first of these – scientific issues and limitations – in more detail.
When looking into the cannabis literature, one finds that cannabis users are predominantly classified on the basis of dependency and/or frequency of use. Diagnoses are based on either the World Health Organization’s International Statistical Classification of Diseases and Related Health Problems (ICD-10) classification or the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM-V).
In order to be diagnosed with cannabis dependence, all that is required is that the participant meets three out of six criteria for the ICD-10 or three out of seven criteria for the DSM-IV-R. Yes, you read that right, only THREE out of six or seven criteria. Not only that, the classification systems themselves only look at a very specific set of “symptoms” basically ignoring overall patterns of behaviour or physical well-being.
So, first off… even though these classification systems are used interchangeably when categorising research participants into experimental groups, they are NOT THE SAME and can (and often does) lead to differing diagnoses amongst users.
Therefore, while one user in one study may be diagnosed as cannabis dependent, another participant in another study with similar usage patterns, may not. Diagnoses tend to sometimes be so arbitrary that in some studies, research participants were diagnosed as being dependent even though cannabis was used less than monthly or even as little once or twice in their life. I don’t know about you, but that makes me question the validity of user classifications.
However, it gets better. These classification systems assume that there is always some type of injury – physical and/or mental – associated with use, thus also assuming that dependent users suffer more harm than nondependent users.
What researchers and these diagnostic manuals do is to focus on physical symptoms and behaviours surrounding cannabis use instead of assessing the actual harm caused by cannabis use itself. This is an important distinction because what this means is that even though the user may present with a very specific and very narrow set of symptoms, these may not actually be problematic to the user him-/herself.
The methodological issues surrounding cannabis use in the context of psychological and social research are perfectly expressed by Macleod et al. in this quote from their 2004 review paper:
‘…although some measures were similar across studies, no two studies measured either illicit drug exposure or psychosocial outcome in the same way. Additionally, potential confounding factors were inconsistently assessed across studies . . .’
Simply put, they couldn’t complete their meta-analysis exploring the relationship between cannabis and psychosocial sequelae because there was no standardised methodology across any of the studies. None. Let that shit sink in…
Apparently, within cannabis research, there is so little consensus amongst scientist as to the standardisation of basic concepts, that no two studies are similar enough to be able to make a direct comparison. That type of uniqueness is fine in the more creative pursuits, but it is not ok in science.
For example… cannabis users are sometimes categorised on the basis of the frequency of use. Some studies classify frequency of use in terms of “heavy users”, “regular users”, and “non-users”. Other times, researchers prefer time-related frequencies such as “daily”, “weekly”, “less than weekly” and “not at all”.
Then, within these categories, the frequency is arbitrarily assigned to “heavy users” in one study being someone that smokes more than x number of joints per week whilst in another “daily use” is defined as (surprisingly) daily use – whether daily use is 1 joint or 10 bongs.
This brings us to another point. Confounding factors (things that have an influence on the ability to completely standardise something) are rarely controlled for. Things like the potency of the cannabis being used by participants are rarely if ever, considered nor is the actual manner of consumption (smoked/ vaporised/ consumed/ topical application).
One cannot help but notice that this, once again, results in the arbitrary classification of users.
Similarly, even though there are several common risk factors found in both substance abuse and mental illness populations, common risk factors such as childhood abuse, socioeconomic backgrounds, hereditary factors and life-stressors are also rarely considered or accounted for. Accounting for these types of variables is experimental methods 101 and not doing so, is just terrible science…
Another issue is that of the way in which the results are reported. A LOT of results are being reported without effect sizes. Now, for people that are not statistically inclined, an effect size tells you how “real” your results are. In any type of testing, if you test enough people, you will find an effect.
Think of it as tossing a coin. Even though statistically speaking, tossing a coin should result in a 50/50 spread of heads or tails, if you toss a coin enough times, eventually, you will have more heads than tails or vice versa. And so too it often goes in the behavioural sciences – you carry on testing until you find an effect because having a publishable paper is what brings in the money.
Now, whether results are because of an actual effect or because of the aforementioned testing methodology, is impossible to tell when effect sizes aren’t reported. With the absence of effect sizes, the reader really has no way of knowing whether what is reported as being statistically significant, actually is.
And speaking of statistically significant results. There is a huge difference between something being statistically significant and something being clinically significant.
Within the cannabis literature, very rarely do researchers discuss the real-life implications of their findings and, often, what the statistics espouse as statistically significant differences between “users” and “non-users” or “heavy users” and “infrequent users” does not translate to clinical deficits, social incompetence or an inability to function adequately on a day-to-day basis.
Approach & Sampling
Another issue that stands out in the cannabis literature is the way in which the entire thing is being approached. At the moment cannabis research in the context of the behavioural and social sciences focus mainly on the identification of cannabis-related risk factors. This is done in one of two ways: general population studies which tend to compare users to non-users; and cannabis user studies comparing “dependent/ frequent users” with “light/ infrequent users”.
For general populations studies, sampling is done by recruiting university students and/or household surveys and/or school surveys. For user studies, participants are recruited from users seeking treatment and/or treatment centres themselves and/or referrals by health care providers.
Can you see where I am going with this?
None of these samples takes into account that people who differ in the way in which they use cannabis most likely also differ in other ways. A high school student’s outlook, environment, social norms and level of maturity is vastly different to that of a suburban housewife. Comparing a weed smoking teenager to an abstinent mother is, quite frankly, ludicrous.
Similarly, when considering the second type of study and its sampling methods, users looking for, or already undergoing treatment, will be more likely to already be dependent and experience adverse user-related effects. Definitely more so than users that are not seeking treatment.
It also skews the bigger picture by not considering the much larger (and often hidden) population of cannabis users who experience no difficulties or adverse effects – or at least not to the extent that they seek treatment.
Yet, despite these issues, discrepancies and limitations associated with the scientific cannabis literature, we are expected to accept policies and legislation based on information that cannot be deemed 100% accurate. We are supposed to blindly believe that the sciences have all the answers and that it knows best. And don’t forget that, according to this same science, most pot smokers will become delinquent, anxiety-ridden, suicidal psychotics…
Luckily though, with the recent increase in the normalisation of cannabis use across a large and variable cross-section of the population, people’s experiences are, for the most part, telling a different story. And, as is attested to by the increasing number of individuals merrily lighting up whilst leading happy, normal and productive lives, this stereotypical profile of cannabis users as dropouts, hippies and stoners, just does not hold water anymore.
By and large, researchers tend to focus on mental illness as either being the major contributor to, or the main consequence of, cannabis use/dependency while largely ignoring other factors such as environment, life stressors, family history or cultural background.
Although these studies provide us with some insight into the cannabis dependence and mental illness, albeit in very specific circumstances, the whole story has by far has not been told, and nor is it being told now.
As long as participants are viewed only as cannabis users instead of as multi-dimensional individuals whom just happen to use cannabis, the validity of the research findings in relation to the use of cannabis and its effects on mental health, should and must be questioned.
I would love to hear what you think? Have you had any negative mental health experiences with cannabis? Or have you successfully used cannabis to alleviate things like anxiety or depression?
Lieze Boshoff is a Cannabis Content & Copywriter that mixes her knowledge of the cannabis industry, psychology and marketing strategies to help cannapreneurs & cannabusinesses establish their brand’s authority and grow their market share.
Want to see how she can help you? Contact her here for a free consultation.