On Monday, this study - Persistant Cannibis Users Show Neuropsychological Decline from Childhood to Midlife - was published in the Proceedings of the National Academy of Sciences (PNAS). It's a straightforward title, a straightforward study, and the resulting news coverage (a Google News search found more than 600 stories) was pretty much universally to the point.
To give you a few examples:
Pot Smoking May Leavener Mark on Teen Brains, from Amanda Gardner at CNN.
Teenage Marijuana Use May Permanently Reduce IQ, from Naomi Seck at Agence France-Press (shown here in Australia's Cosmos magazine)
Early Marijuana Use Linked to I.Q. Loss, from Benedict Carey at The New York Times.
Clear and consistent, right? So why do I want to complicate this picture?. It's really because one of my top ten (okay, make it top three) questions when I see a study like this is: What's the mechanism? And if you take a closer look at this particular study you realize there's nothing to answer that. Rather there's a note from the researchers - "findings are suggestive of a neurotoxic effect of cannabis on the adolescent brain" - which recommends further study into this very issue.
So without a cause-and-effect experiment, how did these researchers find a link between teenage marijuana use and a decline in cognitive processes? As it turns out, they analyzed data from a cohort study from New Zealand, which followed 1037 individuals from the small town of Dunedin from their births in 1972 or 1973 until they reached the age of 38. (A cohort is a group of people with common characteristics and such a study is usually longitudinal, following them for a number of years and analyzing for risk factors.)
According to the PNAS paper: "Cannabis use was ascertained in interviews at ages 18, 21, 26, 32, and 38 y(ears). Neuropsychological testing was conducted at age 13 y, before initiation of cannabis use, and again at age 38 y, after a pattern of persistent cannabis use had developed." What the researchers found was a pattern of lowered IQ at the age of 38 for those who had checked yes on marijuana use as adolescents onward. Those who checked yes to starting later in life didn't show such an effect, even when educational background was factored in.
In other words, they found an association - and granted, a pretty interesting one - that raised the question of whether chemical compounds in cannibis have a damaging effect on the developing brain. But it's essential to remember that longitudal cohort studies are not always reliable - partly because (surprise) people don't always tell the truth on surveys. And partly because researchers don't always ask every pertinant question. In other words, yet again, they should be approached by journalists with a sense of realistic caution.
As an excellent story by Maia Szalavitz at Time, "Does Weekly Marijuana Use by Teens Really Cause a Drop in IQ?" points out, only a small percentage of people in the study were confirmed as heavy rather than recreational marijuana users. And the study did not investigate other well-known factors that contribute to cognition problems, such as head injuries or certain illnesses.
Science blogger, Dean Burnett at The Guardian, raised the issue even more directly. In a post titled, Cannabis Reduces IQ (and appreciation of context). He notes: "It's important to keep in mind the scientific context of drug use. Many drugs can be damanging, but, then, the majority of recreational drugs were introduced for their medical applications. It's not always a clearcut case of 'drugs=bad thing', there are numerous other variables to consider."
To be fair, most stories, like this one by Tanya Lewis at Wired, did note the study limitations. You just have to read a way down to find the qualifications. And as the Lewis story also emphasises, one reason for that is that even researchers critical praised the careful analysis and the real logic of the work. Still, my own preference would to have given those cautionary notes a little more emphasis in the reporting. It would allow all of us - researchers and journalists alike - to look smarter if the next cohort looks different or if the eventual cause-and-effect work suggests that we didn't quite hit the target.