The new generation antidepressants are no better than placebo in mild-moderate depression according to a new analysis of published and unpublished trials that were submitted during the drugs’ approval.
The study is published in PLoS Medicine and despite the huge headlines it has generated, is not entirely surprising.
Psychologist Irving Kirsch, who led this new research, has conducted several previous studies looking at the effectiveness of SSRI antidepressant drugs and found similar results, although this is the first time that the study has factored in the severity of depression.
This study focused on the drugs fluoxetine (Prozac), venlafaxine (Effexor), nefazodone (Serzone), and paroxetine (Seroxat or Paxil) and used the US Freedom of Information Act to request data on (mainly) negative trials that haven’t been published to complement the data set from published trials.
In this new analysis, only in severe depression did these medications show a distinct improvement over placebo, and this, the authors suggest, is because of the reduced placebo effect in the severely depressed, rather than than the fact that the medication has a differential effect in those most affected by mood disorders.
It’s important to note that the study didn’t show that the drugs had no effect in mild-moderate depression. They were all associated with an improvement in depression, but this was no different from placebo (a powerful effect in itself).
It’s also important to note that this finding doesn’t apply to all antidepressant drugs, and that it doesn’t apply to the use of these four drugs in all situations. They are also commonly prescribed for anxiety disorders which weren’t investigated in this study.
However, this is another example of how drug companies’ attempts to obscure data from negative trials are coming back to haunt them.
The Times has one of the best write-ups but as usual, the PLoS article has a jargon-free summary included so you can get the findings from the source even if you’re not familiar with scientific writing.
UPDATE: An important clarification from PJ, taken from the comments:
I think that by saying “this was no different from placebo” you are being misleading. Strictly speaking it was statistically different from placebo but did not reach the NICE criteria for a clinically significant difference:
“a three-point difference in Hamilton Rating Scale of Depression (HRSD)scores or a standardized mean difference (d) of 0.50″