So smoking skunk cause psychosis, but milder cannabis doesn't?

The Mail on Sunday has shouted that
‘ cannabis TRIPLES psychosis risk’ and that
skunk is to blame for ‘1 in 4 of all new
serious mental disorders’. Is this what the
study they cite shows? Well, no, but it’s
really interesting research which could
advance our understanding of the
relationship between cannabis and
psychosis.
The research compares a group of first-
episode psychosis patients in South
London with a group from the general
population in the same area. Not only did
the authors ask participants about the
frequency of their cannabis use, but also
about the type of cannabis they used–
specifically whether it was skunk, or hash.
They found that those with psychosis were
much more likely to have used skunk every
day, than to have never used cannabis.
Conversely, people who smoked hash every
day were no more likely to have psychosis
than people who never tried cannabis.
But does this mean cannabis triples
psychosis risk? The headline reported in
the Mail came from a ‘population
attributable fraction’ that the paper
calculated: the number of cases of
psychosis that would be prevented if all
skunk use was removed, assuming that
cannabis causes psychosis. The authors of
the study estimated this to be 24%.
However, the authors clearly point out that
they cannot be sure the association seen in
their study is causal.
There’s lots of consistent evidence
associating cannabis use with psychosis
and schizophrenia, but as the saying goes,
correlation isn’t causation. For all sorts of
reasons, you can’t randomly assign one
group of teenagers to use cannabis, and
another not to. This means you have to
observe what people choose to do, and the
people who choose to smoke cannabis
might be different in a variety of other
ways, which could be the cause of the
increase in psychosis risk. Although you
can control for these in analyses, you can
never be sure you’ve adequately adjusted
for them.
This study uses a case-control design. This
type of study can lead to bias if the control
population is not adequately selected. In
this article, the controls were selected from
people who lived in the same area of
London. However, the authors noted some
difference between the cases and controls,
in particular those with psychosis were
more likely to be male, more likely to be
white Caucasian, and more likely to be
heavy smokers. Psychosis is more common
in men, and more common in ethnic
minorities (in particular those of migrant
status). The authors accounted for these
differences and adjusted for other
potential confounders, but there could be
others that weren’t measured.
Another drawback of case control studies is
that the exposure measure (cannabis use)
has to be assessed retrospectively. This can
lead to biased findings if having psychosis
might affect likelihood to over- or under-
report former drug use. Given that an
association was found with skunk but not
hash, perhaps this is unlikely to be a
problem, but again the authors can’t be
sure.
For all the caveats though, this study is
really important. It’s the first of its kind to
try and separate out different cannabis
potencies in this way. Past cannabis
research has mainly focused on the effects
of THC – which has been shown in
randomized trials to induce transient
psychotic experiences. However, recent
evidence has hinted that cannabidiol
(CBD), another cannabinoid, might be
protective against psychosis. Skunk
commonly has higher levels of THC than
hash (government reports suggest 15% in
skunk and 5% in hash), but skunk often
only contains traces of CBD, while hash
tends to have roughly equivalent CBD and
THC.
Skunk and hash use were self-reported in
this study, so we have to assume that
participants knew what they were
smoking, and can only guess at the true
levels of THC and CBD in the cannabis
being used. Ideally, samples would have
been taken and the precise levels of these
cannabinoids assessed, but this isn’t
practical in studies of this size and design.
If it is the case that skunk use is causing an
increased risk of psychosis, does this mean
that it’s responsible for one in four of all
serious mental disorders, as the Mail
claims? Ignoring that this paper was
specifically about psychosis, not ‘serious
mental disorders’ more generally, perhaps
in certain high risk populations the impact
of cannabis could be this high, for example
in those with a genetic predisposition to
psychosis, but it’s a big leap to conclude
that from one study.
This new research is an important step on
the road to understanding the nature of
the association between cannabis and
psychosis, but once again media
exaggeration or misrepresentation of the
findings could risk the message being
ignored by the people most likely to
benefit from it. Just in the same way that a
pint of beer of an evening is likely to have
a different health impact to a pint of
vodka, the same could be true for skunk
compared to hash. I look forward to more
research investigating this.
Update: When I wrote this, only the Mail
had covered the research. However, the
Guardian also ran with the headline
‘skunk triples risk of serious psychotic
episode’, and the Telegraph with ‘super
strong cannabis responsible for a quarter
of new psychosis cases’. So apologies to the
Mail for singling them out!