Anti-depressants make me sad

Anti-depressants have been around since the 1950’s, but it wasn’t until SSRIs (Selective Serotonin Reuptake Inhibitors), the first being Prozac, came on the market in the 1980’s that things really took off. The safer profile meant that primary care physicians could start prescribing, and Prozac itself became a blockbuster drug (over $US1 billion in annual sales) and a household name. Others followed and by 2005, anti-depressants were the most prescribed drugs in the USA. For most patients, however, they are no better than placebo. That doesn’t mean that they don’t’ work, they just don’t work any better than placebo.

The studies

The actual effect (here) of these drugs over placebo drugs is not that great: about 2 points on the 62 point Hamilton Depression Scale – a difference not deemed clinically important.

Further, this more recent meta-analysis showed that the effect of anti-depressants over placebo was negligible for those with mild, moderate or severe depression – it was only significant for those with very severe depression.

This study showed that selective (biased) reporting of SSRI trials led to favourable studies being over-reported, and in those published, the results were reported with a bias towards a positive result.

This study further explores this bias. Instead of reviewing all of the published trials, like we normally do, they looked at all of the trials that were submitted to the FDA (as part of the requirement for getting the drugs approved, accessed via Freedom of Information Act). They found that of the trials that were not published, 22 out of 23 were negative (showed the drug to be ineffective), and that 11 of the published trials that were negative (according to the FDA) were published in a way that made them look positive. If you summarised the published trials, you came up with 94% of the trials being positive, making anti-depressants look very good. If you summarised all of the trials that were done (including the unpublished trials) you only get 51% positive, and remember, some of those positive ones are really negative. Our decisions should be based on the totality of evidence. It appears that we may be basing many of our decisions on a biased selection of the evidence.

But even the Cochrane reviews (of published studies) show very small effects for anti-depressant use (primary care, the elderly, post-natal depression) and their review of anti-depressants versus active (therefore better) placebos also showed minor differences, and concluded: “This suggests that the effects of antidepressants may generally be overestimated and their placebo effects may be underestimated”. In fact, the Cochrane reviews of exercise and St John’s wort for depression showed similar effectiveness to anti-depressants, but with fewer side effects. They even show that music therapy, or relaxation therapy are effective treatments.

The downside

The use of SSRIs was extended to adolescents, but reports of an increase in the rate of suicide and suicide ideation in children and adolescents taking SSRIs put a dampener on that practice. SSRIs have also been linked to birth defects (here). This is beyond their standard side effects, which are too numerous to list here.

The punch line

The biological mechanism for most anti-depressants (SSRIs, TCAs, MAOIs) is based on the ‘serotonin hypothesis’ that low levels of serotonin make you sad. However, this hypothesis is based on the fact that these drugs, which happen to increase the serotonin levels around the neurons, are anti-depressants, making for fairly circular reasoning. How then do we explain why newer anti-depressants, called SSREs (Selective Serotonin Reuptake Enhancers), which have the exact opposite effect of SSRIs, seem to work just as well? My simple explanation is that they both work by the placebo effect, and that the serotonin hypothesis is simplistic and wrong. Don’t worry, I am sure that some clever doctors out there are working out a contorted biological explanation of this paradox.

The bottom line

Anti-depressants are regularly used as punching bags for those who want to highlight the role of Big Pharma, conflicts of interests amongst psychiatrists, problems with advisory bodies, medicalisation, over-treatment, over-diagnosis, and the social construction of 'diseases'. My point is simple, and I think it is the issue at the core of all of the problems mentioned: these drugs either don't work or barely work at all, beyond the placebo effect.


Addit 11 Sept 2012: See this link for a review on antidepressants, balancing harm vs good on an evolutionary basis.