Economist Article : Anti-depressants, hope from a pill

Back

Antidepressants

Hope from a pill

Feb 28th 2008 NEW YORK
From The Economist print edition

Disagreements over whether drugs to combat depression are worth taking

ANTIDEPRESSANTS have long been the source of controversy. Amphetamines were widely used as an antidote to neurotic depression into the 1960s, until such “pep pills” came to be seen as doing more harm than good. Similar worries are now engulfing today's antidepressants, like Prozac and Paxil, which are among the most widely prescribed drugs in the world. Two new studies have stirred things up: one warning that antidepressants do not help most people very much, and the other gushing that they are a marvellously cheap way to save lives.

Most antidepression pills prescribed today are selective serotonin reuptake inhibitors (SSRIs), a type of drug that boosts the amount of serotonin hanging around in the brain. Serotonin is a brain chemical closely associated with mood. Boosting its level this way might therefore improve a person's mood. Earlier versions were less effective than modern pills, which have fewer side effects and are less toxic in overdose. The use of SSRIs worldwide has shot up from below 3 billion doses in 1995 to over 10 billion in 2004.

Is all the pill popping doing any good? There is recent evidence that it can lead young people to act on suicidal thoughts, prompting America's Food and Drug Administration (FDA) to insist on warnings. SSRIs have generally been seen as a way to ease depression in adults without killing them. Derek Summerfield argued in the Journal of the Royal Society of Medicine that, although there is no epidemic of depression, “the case for an epidemic of antidepressant prescribing is now cast iron.”

A study published in this week's Public Library of Science (PLoS) Medicine, an open-access scientific journal, raises doubts about dispensing such drugs so freely. Irving Kirsch, of the University of Hull, and his colleagues scrutinised the clinical trials for several new antidepressants, taking care to include those never published (but which, by law, have to be reported to the FDA). They found that SSRIs did not help the vast majority of depressed people much more than placebos did. The net benefits over placebos did not usually reach the level considered big enough to be of clinical significance by Britain's National Institute for Health and Clinical Excellence (NICE).

This study points to two factors that bedevil proponents of SSRIs: publication bias and the power of placebos. Dr Kirsch believes published data “give an exaggerated view of a drug's benefit.” People with very severe depression did see benefits above the NICE threshold, but even that was not a ringing endorsement. Dr Kirsch explains that this was not because SSRIs worked much better in the very seriously depressed, but rather that the effectiveness of placebos dropped off sharply in such people, making the drug look better.

There are two sorts of criticism of the idea that SSRIs are mostly a waste of money. One comes from those who say the study itself is rubbish. David Nutt, of the University of Bristol, says failed drug trials often remain unpublished because their design is shoddy or their results uninteresting. He criticises the PLoS paper as a “mishmash of quality trials and lousy trials leading to a false criticism of these drugs”, which he maintains do help those with depression even if their effectiveness falls below the NICE's “arbitrary” threshold. Dr Nutt thinks it is misleading to compare these drugs with placebos, since what matters is that they work when compared with some alternatives, such as “talk therapy”, for which he believes there is even less evidence of effectiveness.

Good-news potions

Yet there is reason to think that unpublished studies do reveal some important and rather unflattering details about antidepressants. A well-designed study published in January in the New England Journal of Medicine looked at a larger group of antidepressants and concluded there was indeed “a bias toward the publication of positive results”: 94% of the published trials were positive, whereas only about half of the unpublished ones were.

That bolsters the PLoS paper, but there is another sort of critique that challenges its conclusions. A recent study published by America's National Bureau of Economic Research (NBER) analysed data from 26 countries over several decades to determine what effect SSRIs have had on suicides. Its authors argue that antidepressants are in fact “a very cost-effective means for saving lives”.

Countries with both high and low initial rates of antidepressant use saw similar trends in suicides until SSRIs were introduced. Jens Ludwig, of the University of Chicago, argues that countries that took to the new drugs saw a relative decline in suicides. After controlling for many variables, his NBER team reaches the cheerful conclusion that an increase in sales of one pill per person per year (about a 12% increase over the level in 2000) leads to a decline in suicide mortality of about 5%.

So are SSRIs to be shunned or saluted? The controversy will rage on, but Erick Turner, of the Portland VA Medical Centre in Oregon, suggests a third way. As one of the authors of the New England Journal of Medicine paper, he says the study confirms that most antidepressants do not work as well as published reports claim. That suggests many people, especially children, should be more careful about using them. But that does not mean they are pointless. He speculates that, if other therapies fail, “maybe all you need is a minor or mediocre effect in order to reduce suicides overall.” That may be particularly true for those who are closest to the edge of darkness.

Back