Study casts doubt on link between cannabis, teen IQ drop
SYDNEY (Reuters) - A landmark study suggesting a link between cannabis use and a drop in teenage IQ may not have gone far enough in its research, with any falls in IQ more likely due to lower socioeconomic status than marijuana, according to a Norwegian study.
The latest work, which appears in the journal PNAS, Proceedings of the National Academy of Sciences of the United States of America, also suggests that different policy steps might be needed in that case.
"My study essentially shows that the methods used and analyses presented in the original research are insufficient to rule out other explanations (for lower IQ)," said Ole Rogeberg, an economist at the Frisch Centre for Economics Research in Oslo, to Reuters.
The Dunedin Multi-disciplinary Health and Development Study is an ongoing report produced by New Zealand's University of Otago, monitoring 1,037 New Zealand children born between April 1972 and March 1973. The study followed them for 40 years.
The participants were periodically tested for IQ and other indices including drug taking, and in 2012 clinical psychologist Madeline Meier produced a study saying there was a link between teenage cannabis use and a lower IQ.
Researchers in the Meier study compared the IQ trends of people who never smoked cannabis with four groups of those who did: people who smoked, people who scored as dependent in a follow-up survey, those who scored as dependent twice and those who scored as dependent three times.
The study found IQ declines increasing "linearly" with cannabis use, Rogeberg wrote in PNAS.
The crucial assumption in the Meier study is that cannabis use is the only relevant difference between the groups tested, he said. His use of a simulation model showed that it may be premature to draw a causal inference between marijuana use and falling IQ scores.
For one thing, other writing about the Dunedin group on which Meier's study is based suggest that early cannabis use is more common for people with poor self-control, previous conduct problems, and high scores on risk factors linked to low family socioeconomic status, he wrote. Continued...