Tuesday, July 31, 2007

cannabis

Now that I've calmed down a little bit re. last
Friday's cannabis story and the way it was reported in
the media, can I just try to un-muddy the waters by
clarifying that:

1) the relevant paper in the Lancet did not contain
any new data. It was a meta-study, i.e. the authors
did nothing but review existing studies of cases where
mental illness coincides with a history of cannabis
use. Considering this, I thought that The Lancet was
guilty of fuelling hype when they flagged this paper
up with not one but two accompanying commentaries (and
probably a press release!).

2) if for a second we choose to believe the claim
(still based on a rather small number of cases, with
no systematic epidemiology!) that people who have used
cannabis are a bit more likely to develop
schizophrenia than people who haven't, the serious
newspapers (like most disappointingly, The Guardian)
which reported this study uncritically, should have
clarified that this observation does not prove a
causal connection.

3) ... and even if there was a causal connection, any
serious report should have clarified that the
connection can go two different directions. Unlike in
smoking / lung disease, where the cause-effect route
goes one way, mental illness changes the way people
behave. Thus, somebody who is harbouring a disposition
to develop schizophrenia (or depression, or any other
mental illness) may be more likely to try and even
enjoy mind-altering drugs.

Personally I think that if there is a causal
connection behind these stats, it is much more likely
to go in that direction, in that disposition to mental
illness causes drug use, not the other way round.
That is of course a matter of opinion. But the claim
that there is a proven causal connection as reported
in the media, that is just bad science.

I've also sent a letter to the Guardian on this matter, which got published
here.

No comments: