By Michael Slezak
A study linking daily cannabis use to suicide got a lot of coverage earlier this week, but, Michael Slezak says, correlation does not mean causation
“Cannabis catastrophic for young brains” screamed the headline on an Australian medical news website this week.
The article, and others like it, were reporting on a study linking teenage cannabis use with school dropouts, addiction and suicide, published in the The Lancet Psychiatry.
Echoing the research findings, the articles declared that if teenagers smoke cannabis daily, it makes them seven times more likely to commit suicide compared with non-users. Indeed, “there is no safe level of use”, most reported.
They also urged caution to legislators around the world that are gingerly taking steps towards weakening prohibition of cannabis, lest young people get easier access to it.
Gone to pot
So does smoking pot cause suicide? The Lancet authors say it probably does.
Their study combined data from three previous longitudinal studies which together tracked cannabis use in more than 3000 people in Australia and New Zealand over many years. The authors looked for associations between the frequency of cannabis use before the age of 17 and outcomes, such as high school completion, until the people reached the age of 30. They found that those who smoked cannabis daily before they were 17 had lower odds of finishing high school and getting a degree than people who had never used cannabis. Larger increased odds were associated with cannabis dependence later in life, trying other illicit drugs and suicide attempts.
But longitudinal studies only show correlation, not causation. The difficulty is that people take drugs for a reason, and that reason could be what’s causing the outcome. In the case of school dropout, suicide and daily pot smoking, it is not hard to imagine what else could be going on in someone’s life to engender these behaviours.
The authors tried to rule out 53 such possibilities, things like pre-existing behavioural disorders, depression or parental divorce. But because a study like this can’t be randomised, there are likely to be other possible cofounders. “There is still the possibility that adolescents who use marijuana have other aspects of their lives that make them more susceptible to poor outcomes during adulthood,” says Simon Denny from the University of Auckland in New Zealand.
“It is likely that significant proportion of the users have pre-existing problems and seek cannabis as a way out,” says David Nutt from Imperial College London.
One thing that is often a sign of causality in non-randomised studies is a dose-response relationship. And they found that here: the more cannabis that teenagers smoked, the more likely they were to drop out of school or commit suicide.
But dose-response relationships can be confounded too. It’s easy to imagine that while a relatively happy, well-adjusted teenager might never smoke cannabis, or only occasionally, a very troubled teen might start using it daily, and there could be a gradual rise in use between the two extremes.
Thomas Lumley, a biostatistician from the University of Auckland says there are many such alternative possibilities that are consistent with the data. For example, the possibility that students with more ambition or more interest in intellectual activities are less likely to become heavy cannabis users and are also more likely to do well in school, he says.
“It’s not that this sort of explanation is more plausible than a biochemical effect, it’s that the analyses are not able to distinguish one from the other: the research provides no additional evidence over and above the prior plausibility of the explanations,” he says.
Assuming cannabis does cause these undesirable outcomes, the authors hint – and many commentators explicitly say – that weakening prohibition of cannabis will be harmful. In a comment piece published to coincide with the study in the same journal, Merete Nordentoft from the University of Copenhagen in Denmark says that decriminalisation and legalisation will probably decrease prices, increase use and “lead to more young people having difficulties with school completion”.
Max Abbott of the Auckland University of Technology disagrees that liberalisation will make things worse, despite being convinced that cannabis probably does cause negative outcomes. “In my view [the law] has failed to reduce cannabis use, needlessly criminalised tens of thousands of New Zealanders, wasted police and court resources and fuelled organised crime. I believe the adverse health and social costs of current legislation and attempts to enforce it are likely to far outweigh any direct costs of this type stemming directly from cannabis use,” he says.
“The authors should not have drawn these erroneous policy implications from this research,” says Alex Wodak from St Vincent’s hospital in Sydney, Australia. He points out that in Australia, 92 per cent of people say it is easy to buy cannabis, replicating similar findings elsewhere. “How could reform make cannabis more available?” he asks. “There have been about a dozen papers where cannabis prevalence before and after liberalisation have been compared. Cannabis prevalence has not increased after liberalisation.”
What’s more, he says a legal regulated market can limit access for young people. “I don’t recall Al Capone insisting that he would never sell alcohol to young people,” he says.