Prospect magazine has an interesting article covering psychologist Irving Kirsch’s widely publicised meta-analyses that have questioned whether Prozac-style SSRI antidepressants are any better than placebo.
Kirsch has become well known for requesting unpublished trial data via the US Freedom of Information Act and pooling it with the published evidence. The conclusion of his latest re-analysis was that there was little difference between sugar pills and SSRIs in the treatment of depression.
This has kicked up all sorts of merry hell, not least because the media reported (and the Prospect article implies) that ‘antidepressants don’t work’ which is clearly false. They do work, but the debate is over how much of the effect is due to placebo.
It’s not quite as simple as it seems of course, as not everyone agrees with Kirsch’s methods and, as noted in an insightful 2008 paper, his argument is based on the assumption that people who respond to antidepressants also respond to placebo in a similar way, when we know there are individual variations in both.
Kirsch apparently has a book coming out shortly which is likely to restart the debate and it’s likely to be heated.
There are some hints of this in the article where several prominent psychiatric scientists give variations on the “don’t criticise the evidence, you’re harming children!” argument. In fact, head of the NHS trust where my research institution is based apparently blames ‘the media, and psychologists’ “who have a vested interest in constantly attacking antidepressants”. Yes, we’ve reached that level already.
We went through a very similar process when concerns over whether SSRIs increased suicidal thinking in adolescents were raised. Lots of similar mud-flinging ensued.
Interestingly, a meta-analysis of suicide attempts and suicidal thoughts in 372 trials just published in the British Medical Journal found that overall SSRIs had no effect on risk of self-harm, and that when the data was divided by age, there was a slight increase in thoughts and attempt in people younger than 25 and a slight decrease in adults aged over 65 (the comments on the article are also worth reading).
It’s probably worth saying that even in young people self-harm when taking antidepressants is very rare, but the fact that the drugs had no overall protective effect except in older people should give us pause for thought.
But getting people to focus on the evidence when they’re wound up is like getting people to focus on the fire exits during a strip show. We all accept the importance of doing so but few can quite manage it when the time comes.