Cannabis and psychosis? A critique of the Lancet's meta-analysis
In late July of 2007, a media storm erupted then rapidly spun out of control - over a report issued by the Lancet. What we heard in these torrential blasts was that science had finally discovered a causal link between long-term cannabis consumption and psychosis. This critique will explore three fundamental questions regarding the scientific viability of the Lancet's meta-analysis. Two of the questions raised are philosophical in nature. The other goes to methodology. The Lancet's conclusion: "we have described a consistent association between cannabis use and psychotic symptoms" was drawn from a curious approach towards criteria and data. (The reasons for and wisdom of will be punctured and dissected in this essay.) The questions which I'll ask and answer follow: Are the definitions of the psychotic symptoms - which the researchers used to draw their alleged link between cannabis consumption and psychosis logically coherent? (2) Should psychiatrists have been given the final say to decide which ideas human beings are medically or otherwise permitted to entertain and express? Why is the relative frequency - the clustering - of the psychotic symptoms, said to have been reported by the subjects, not presented in any of the studies? (This is a fatal methodological flaw, common to this line of research.) The Lancet report is a meta-analysis, containing and evaluating 11 "independent" studies, purportedly linking cannabis consumption with either psychotic symptoms or psychotic disorders. From what I've learned in reading these studies is that the researchers sought to "forge the link" with tools made from questionable, though "universally accepted criteria" - criteria which psychiatrists invented to define psychotic symptoms. But, if the causal link cannot be forged due to logical weaknesses or contradictions in the definitions of these symptoms, then the link remains at best undemonstrated. [1] Listed below are the symptoms most commonly cited in the 11 studies. They were copied and pasted directly in their authors own words from the CHDS study. This study was one of the 11 featured prominently in the Lancet report. the symptoms "hearing voices that other people do not hear" "the idea that someone else can control your thoughts" "other people being aware of your private thoughts" "having thoughts that are not your own" "having ideas and beliefs that others do not share" "the idea that something is seriously wrong with your body" "never feeling close to another person" "the idea that something is wrong with your mind" "feeling other people cannot be trusted" "feeling that you are watched or talked about by others"
I.
For the sake of concision, I'll concentrate on and analyze one of the symptoms from the list - no doubt its weakest, i.e. "having ideas and beliefs that others do not share". Let's logically un-pack what "having ideas and beliefs that others do not share" means as a psychotic symptom. What's a shared idea? It's an idea that is entertained by at least two people. It's also an idea that was once - by definition - unshared. Under the above criterion, every new idea is a psychotic symptom for at least one person - its originator. A novel idea begins its "life" as unshared - because it's a product of an individual human mind. Unless innate ideas exist, all ideas are initially unshared and - according to psychiatry - psychotic symptoms. Logically - under this disease model of unshared ideas, there would have to be a transmission of the disease from the originator of the idea to the ideas's second adopter, then to the third adopter and ad infinitum; thus, infecting any person who held it. The only way to block the emergence of this pathological chain-reaction would be that the acceptence of the idea by the second adopter (ipso facto/presto chango) transformed the symptom of the disease - a novel idea - into a non-symptom. A knowledge-destroying paradox is produced when shared ideas are elevated over unshared ones - much less than when having unshared ideas is turned into a symptom of a lurid pathology. If all novel ideas are objectively classifiable as diseased, then the status of human knowledge - itself - is problematic. The first psychiatrist who entertained the idea that having ideas and beliefs that others do not share is a psychotic symptom had that idea as his psychotic symptom by his own definition - because no-one else shared or could have shared the idea. This idea was - trivially - his own. Since this psychiatric criterion for psychosis rests on a psychotic symptom, then "psychiatric knowledge" cannot reliably picture how things stand in the world, or more to the point, it cannot prescribe which ideas should not be held in human minds. [3] Psychiatrists demand to be taken as the only epistemic authorities around. They lay down what counts as adequate and inadequate ideas. Yet, no-one demands that they follow their own rules. It's more coherent to define a psychotic symptom as "having a belief that psychiatrists have conclusively decided which ideas are adequate and which are not", than defining it as "having ideas and beliefs that others do not share" - because the former position does not yield the knowledge-destroying paradox. In the last 400 years, many of the ideas and beliefs which contributed to the expansion of human knowledge through science were not shared, nor could they have been shared. Consistently applying the above criterion to Isaac Newton would diagnosis him as suffering a psychotic symptom (or an episode) when he entertained the idea of universal gravitation. In the Age of Reason, the commonly held belief was that God had organized nature into a finely-tuned machine, consisting of interlocking parts, and that these parts, contiguously and sequentially, interacted with each other - like the spindles and gears in a watch. A bizzare idea occurred to Newton, namely that heavenly bodies, separated by cosmic distances, communicated with each other instantaneously and through no detectable medium. Newton imagined these bodies "acting at a distance" on each other - an idiosyncratic idea. He put something occult into the machine, and science has been profiting, if not reeling, from his once unshared idea (universal gravitation) ever since. If psychotics are the ones responsible for generating the conceptual novelty that our civilization produces, then what does this say about the status of human knowledge in our civilization? The intellectual progress of our species could not have come about had unshared ideas been diagnosed as "the symptoms of pathology" and their bearers banished or "treated". Being a crude metaphysical realist, I'd like to point out that an idea is true independently of whether it is shared or by how many others share it. An idea - because and only because - it is shared does not guaranty that idea a lock on truth (or even health). Yet, we've allowed psychiatrists to dream that they've solved ancient philosophical puzzles - puzzles over which they exert no legitimate jurisdiction. Perhaps, someone should awaken them from their dream. Sir Karl R. Popper developed a model of the human mind, based on how genes operate. Genes work - in part - by blindly varying and selectively retaining subcomponents of DNA, giving rise to the mechanism of biological novelty known as mutation. To Popper, the human mind is a hyper-active dynamo generating bundles of ideas - the majority of which are never shared. Because many of these ideas are inchoate and unintelligible, they could not begin to be shared. The ones grasped and retained by a thinker do not necessarily become commonly held as well. Since all human minds grind out massive quantities of unshared ideas, then all humans are but one self-report away from being labeled as suffering from psychotic symptoms. How do these psychiatrists know that there's a detectable difference between the minds of 99% of people who are never judged as having psychotic symptoms and the minds of approximately 1% who are? The only logical conclusion is that all people entertain unshared ideas (i.e. psychotic experiences) - though they go almost entirely unreported. Therefore, it's redundant for psychiatrists to pronounce that a minute fraction of those who smoke pot will verbally express psychotic symptoms - because all humans generate ideas which are not shared, and these are declared in one of the criteria to be psychotic experiences. More to the point: if every member of the human population has the trait of generating unshared ideas, researchers demonstrate nothing by singling out a small proportion of the members of the marijuana-smoking population and concluding that they uniquely possess the trait. By way of contrast, should a geneticist advance a theory, running that every human being is genetically defective, then attempt to single out the sub-Saharan human population as uniquely defective, he would have - emptily - concluded that some human population was defective. Instantly and unlike the cases in which the conclusions of psychiatrists are tolerated by educated people, this geneticist would be unmasked as a dolt and a racist. Whether or not Popper's model accurately captures how the human mind functions is beside the point. All human minds evolve unshared ideas, whether through dreams, (mis)perceptions, flights of the imagination, theorizing about nature as Newton did, etc. Predictably, Popper's gene-based model of the human mind was his novel idea - yet another instance of a psychotic symptom (or episode).
II.
In this research, there's a category error, regarding how the various types of symptoms are compared (or better - illicitly compared). Having ideas and beliefs that others do not share and hearing voices that other people do not hear are quite distinct - phenomenally. Suppose that as a long-term effect of smoking pot, x has a belief that others do not share and y, as a long-term effect of smoking pot, hears a voice that other people do not hear. Yet, both x and y report their symptoms to a psychiatrist and each are - duly - labeled as experiencing a psychotic symptom. Beliefs lack a beginning and an ending, an "authentic duration" - unlike hearing a real or even an imaginary voice does. When a new belief takes hold of me, I cannot know precisely when it appears and when it vanishes. If I have a shared or an unshared belief, how many instances of it do I have? Even when I do not consciously entertain this belief, the belief seems to be operating somehow "behind the scenes" - since I'd never (privately nor publicly) repudiated it. Though the above distinction may sound like hair-splitting, it's relevant. In the longitudinal studies, referred to in the meta-analysis, there are tables, plotting incidences of entirely unidentified psychotic symptoms on a yearly basis. One symptom, claimed by a subject in a year, makes that subject a sufferer of psychotic symptoms for that year. By definition, an unshared belief generates numerous "symptom-incidences". But, how many of these "symptom-incidences" does an unshared belief produce in 12 months? one? twelve? infinity? Arguably - because the adoption of an unshared belief generates numerous symptom-incidences, a person who smoked pot for five years then reported the adoption of an unshared belief over the course of a year is more psychotic than a person who smoked pot for the last thirty years and admitted to "hearing" one imaginary voice once in a year and only in that year. [5, 6] [7] Further - immediately below, there is a deliberate equivocation, invented by the researchers, confounding the supposed short-term and long-term pharmacological consequences of cannabis. This equivocation, then, is used to leverage the conclusion that long-term cannabis consumption causes psychotic symptoms and disorders.
"A dose-response effect was observed in all studies that examined the relation to increasing cannabis exposure. Only three studies 26,29,31 examined psychotic disorders as an outcome;"
The authors of the Lancet confidently weigh in with their conclusion that these "Findings were consistent with a dose-response effect" and "A dose-response effect was observed". For any dose-response effect claim to hold, the effect must succeed the cause immediately - not years or decades down the line. What these authors have not-too-cleverly done is to redefine what a "dose-response effect" means. A dose-response effect has to do with the acute reactions triggered by a drug. It has nothing to do with a long-term, cumulative "outcome" associated with the use of that drug. A dose-response - by definition - cannot operate over an extended span of time. The authors need to take a step back, come to grips with this long and short term duality and decide what sort of research they intend to conduct: either evaluate the short-term responses of a drug on its consumers or follow the long-term linkage between a drug and its consequences, but not do both simultaneously. To present a chronic outcome of pot consumption as a dose-response effect is to conflate the distinction between the short-term and the long-term, and - in the process, commit the fallacy of equivocation. [8] For whatever reason, the authors of these studies have chosen not to represent - numerically, proportionally or otherwise - a break-down of the symptoms self-reported by the subjects. As such, this omission should be seen as marking the utter pointlessness of this line of research. There are no charts, plotting the distributions of the most to the least frequently reported symptoms. Hence, we cannot know either how or even that these symptoms cluster. Precisely, what we would like to know is this: which symptoms were the most frequently reported and which were the least frequently reported? Though there may be many possible explanations for the non-appearance of these data, three automatically spring to mind: The data do not exist. The data do exist, but were never drawn up to represent the clustering. (3) The authors are - for their own reasons - sealing this information off from public inspection. The danger with defining a symptom as elastically as "having ideas and beliefs that others do not share" is that (see above) every member of the human species has these experiences. It might have been instructive to know if 99.9% of those pot-smokers who self-reported psychotic symptoms claimed "having ideas and beliefs that others do not share" and 0.1% claimed any other symptom from off that list or vice versa or in some combination with the other symptoms thereof. But, the data are sorely missing from every study. The relative frequency of this or that (from the list above) pot-induced "psychotic symptom" is unknowable - as there is no presentation of them. The lack of depicting such critical data makes these studies 11 little black boxes. [9] [10] [11] Yet again, we are left with a stale diktat. We must meekly accede to whatever psychiatrists say comprise adequate and inadequate ideas and that the chronic consumption of pot (though rarely) causes inadequate ones, yet we remain in the dark not knowing how the pot-caused inadequate ideas are distributed - assuming that they are distributed! Had these researchers the capacity for honesty, they would be forced to concede that there is no reliable way to falsify this theory - which grafted together a set of logically weak (or self-contradictory) criteria and a lack of inspectible data. No conclusion flowing from the criteria and the "data", contained in these studies, can be taken on other than a religious basis.
III.
There's no civilization-ruining problem - when a fraction of the population smokes pot, then a few go on to utter "in confidence" to psychiatrists - words to the effect: "Whoa, old shrink-dude, sometimes I believe I can juggle the planets. HaHaHa!". This type of a remark may be the expected, yet highly nuanced, verbal behavior engaged in by those - who are members of a disliked sub-culture, not the permanent "dysregulation" of a person's nervous system. Those who engage in practices, generally frowned on, have been known to protest their assigned (or chosen) roles in the culture by employing language to provoke (or frustrate) those in authority - especially those authority figures who seek to keep them in that role or who intend to further marginalize them by classifying them as mentally ill. (Language-use involves intentional, rule-bound decision-making skills. An undesired speech act belongs to an entirely different category than that of the spastic discharge of a nerve terminal.) The above remark, released by our fictional pot-smoker, however "off the wall" would be a shared idea within his sub-culture, so it cannot under one of the psychiatrist's criteria be described as a psychotic symptom. In fact, psychiatrists - under their own rules - would be logically compelled to pronounce it (as a shared idea), downright healthy! The personal testimonies of pot-smokers are almost universally deemed less reliable - than those who do not smoke pot. To me, it's a bit mercenary to attribute an ad hoc truthfulness to the verbal expressions of those, belonging to such a "dishonest population" - only when they incriminate themselves by "privately confessing" to psychiatrists which "psychotic symptoms" they say they have. N.B. I struggled with the idea of analyzing a few of the other symptoms, but prudence got the better of me. Here's another gem from the list: "having thoughts that are not your own". If a person has thoughts not of his own, he has (or thinks he has) shared ideas. So, let's get this one straight: if you entertain unshared ideas, then you are psychotic, but if you entertain shared ideas, then you are psychotic. No human being could escape the diagnosis of being psychotic. My guess is that these "symptoms" do not have orthodox word-meanings. Only those indoctrinated into a cult could begin to unravel the mysteries of this esoteric arcanum. If there's one default, intellectual responsibility, it's to honestly try to understand the world - not deliberately pursue a misunderstanding of it through privately re-defined word-meanings. Through the ginning-up of criteria, playing hide and seek with data and equivocating with words, these psychiatrists and their researchers have contributed to the de-objectification of science. In conclusion, it may be less delusional to lapse into the "pot-caused psychosis", conceiving oneself to be a Napoleon, than it is to parade around as though one were a sort of an Aristotle, believing oneself to have solved more than a few of the persisting problems in philosophy. It appears that only psychiatrists and the researchers - who take up the cudgels for them - are entitled to their delusions.