New 'potaganda' ignites the naysayers

Among my circles it's said that if Miranda Devine opposes you, you must be on a good track. Trouble is, she writes superficially plausible 'sound-bites' very well, so her latest attack on drug law reform will convince some people who don't really know the facts.

Again Ms Devine is fuelled by a report from the National Cannabis Prevention and Information Centre (NCPIC), a body generously funded by our cash-strapped health department to demonise cannabis under the guise of science.

Their work to date has been, on the whole, almost laughable, one study attempting to link cannabis and violence through a dodgy analysis of patients at St Vincents Hospital Emergency. Never mind that one doctor who worked there said he had never once seen a patient present because of cannabis problems.


Another recent NCPIC report attempted to show that pot causes respiratory problems. As with most of their reports, they don't present new findings, just analyse old findings in a different way and package them up as fodder for the likes of Miranda Devine and to please their funding body.

In this case they referenced two studies by UCLA Medical professor Donald Tashkin which analysed the components of cannabis smoke and found it had lots of nasties in it, then extrapolated that cannabis might be more carcinogenic than tobacco smoke. The media of course jumped on these with glee and ran shock-horror headlines world-wide.

Tashkin's findings so suited the US government that they funded a much larger population study which unfortunately for them showed that long-term pot smokers had LESS cancer than the control group. OOPs.

But NCPIC appear to have ignored this study in their review, listing only the two earlier Tashkin studies.

Similarly, it would not be difficult to collect a bunch of studies that link pot and schizophrenia, especially if you ignore the really big problem for prohibitionists as explained in a much more balanced report of the NCPIC study in Time magazine:

Still, none of the data linking marijuana use and psychosis can prove causality or sufficiently explain why rates of schizophrenia have remained stable or even declined since the 1950s, while marijuana use has increased exponentially. Unlike rates of cigarette smoking and lung cancer, which rise in tandem, marijuana smoking rates in the population do not correlate with higher rates of schizophrenia.

But the real fallacy in Miranda Devine's argument is that, even if pot does cause mental illness, prohibition only worsens the problem.

This is because under prohibition, pot is easier to get than alcohol for under-age kids (those most at-risk) as a small-time peer-to-peer dealer network exists in every town and high school, so distribution becomes self-funding or even profitable. The bad elements combine it with gang culture to really set up potential criminal lifestyles.

Legalised regulation would starve these networks of reliable supply and many kids would have to either get older people to buy for them (and this could be the target of severe penalties and credible education campaigns) or grow their own (at least gardening would be healthier than bashing the private school kids and pinching their mobiles to fund a purchase, as happens now).

I wonder how Ms Devine would react if they brought back prohibition of alcohol along with all the crime, corruption and death that caused. Prohibition doesn't work, no matter how much quasi-scientific spin you produce to support it.

And it's interesting that the most enthusiastic media backers of NCPIC's work also support the Iraq war and deny climate change. Nice company you keep, NCPIC. The director, Professor Jan Copeland, must be proud of herself.

PS: This study of 600,000 patients in the UK showed that psychosis rates were not increasing, whereas if the cannabis/psychosis link was as real as some claim, a 29 percent increase would have been expected. Note, this again is a population study, dealing with real people.

PPS: Informed commentary on Crickey supports the line I take above, pointing out that the NCPIC study excluded 80% of available data for various technical reasons. One Michael R James writes:

But even if we give the benefit of the doubt to such studies or the hypothesis, I doubt anyone seriously believes the "solution" is more draconian prohibition. And good luck on counseling kids on preventive measures for a disease which less than 1% of them will suffer. If anything these studies suggest that decriminalization and regulation of the product (THC levels etc) would be the way to go. But then commonsense never applies.

Full text of his comment below:


"Cannabis and schizophrenia:

"Michael R. James writes: Re. "Why pot really is making kids sick: the new scientific line" (yesterday, item 9). The article by Sophie Cousins was too emphatic in accepting the conclusion from the newly published study. Schizophrenia (SZ) is a terrible disease particularly as it afflicts adolescents and young adults but these strenuous attempts to place blame on a single environmental factor seems misguided.

"Readers should understand that this paper has not generated new primary data but is instead a meta-study of previously published data. A meta-study attempts to extract analyses of greater statistical significance from as many published studies as can be found to meet specified eligibility criteria. Meta-studies are used when there is confusion or conflicting conclusions from individual studies, usually over a long period of time. They are often employed, or resorted to, when statistical power of individual studies is inadequate. Thus, by definition the hypotheses being tested are not easy to prove and the phenomena being examined are complex and poorly understood. Disease of mental health fit this bill to an extreme.

"Clearly any attempt to increase power by pooling data from different studies is itself susceptible to problems, perhaps none more so than in mental health research. The meta-study must attempt to establish rules that pool appropriate data and exclude incompatible data. In this case, of 443 relevant published papers examined 354 of them were excluded from the meta-study. So 80% of papers were excluded for the various reasons given by Large et al. For example only those were retained that reported time of onset of psychosis rather than time of first treatment. These exclusions were an attempt to remove potential confounding issues that may have previously hidden significant associations. But excluding 80% of data raises the risk of ascertainment bias or publication bias, notwithstanding the authors’ statement “There was no statistical evidence of publication bias.” Statisticians try to take care of such nuisances but this scientist hardly ever believes their claims.

"Personally I am not going to be persuaded by studies like this. There are too many other possibilities, for example an ascertainment bias caused by susceptible teens being over-represented in the drug user group (or starting earlier, or smoking more). There are contrary studies suggesting that cannabis use may actually ameliorate psychosocial effects in SZ.  I also find the effect described (cannabis use advances disease onset by 2.7 y) suspiciously high especially since it comes from all users, not, say, high users. The authors reported that heavy use caused a greater effect but was not statistically significantly separable from light users; there is almost the hint in the paper that single use may be enough!

"More worrying is how enthusiastically politicians and some in the medical profession will grasp onto conclusions like this. Indeed, there is the money to perform these kinds of studies because cannabis is the third most commonly used addictive drug, not because it is the third most harm inducing drug. (That honour will lie with drugs like Prozac or Stilnox and other prescribed behaviour-modifying drugs our society is overdosing itself on.) Like the hysteria over ecstasy (on average one death per year compared to 12,000 alcohol-related deaths (2006 data, all ages) or log orders lower than youth alcohol consumption or dozens of other things kids use or do) the harm from cannabis is self-evidently not easy to find or we wouldn’t need to spend so much research money searching for it.

"And as a Crikey commenter noted "about 20 people under 25 experience early onset of schizophrenia and have smoked marijuana at least once" are not compelling numbers. Total abstinence is not going to make much difference to SZ prevalence—maybe none if, as is likely, all such people will still suffer SZ a few years later -- or have much public health impact at all really.

"But even if we give the benefit of the doubt to such studies or the hypothesis, I doubt anyone seriously believes the "solution" is more draconian prohibition. And good luck on counseling kids on preventive measures for a disease which less than 1% of them will suffer. If anything these studies suggest that decriminalization and regulation of the product (THC levels etc) would be the way to go. But then commonsense never applies."