Prozac nation: do antidepressants work?

Research published last month in online journal Public Library of Science Medicine startled the media and the mental health sector. The study, a meta-analysis of published and unpublished trials into the effectiveness of modern antidepressants, such as Prozac, found they outperformed a placebo only in people with the most severe depression.

Even then the researchers concluded that they outperformed a placebo only because people with the most severe depression were less responsive to the placebo effect, rather than more responsive to the drugs.

The drug companies have challenged the findings, arguing they go against a wider body of science, but for some the study is just the latest confirmation that antidepressants should not be used to tackle society’s wider problems.

More than 31 million prescriptions for antidepressants were written in England last year, according to government figures, up 6% on 2006.

The reasons are not hard to understand. The Liberal Democrats surveyed 100 NHS trusts earlier this year and found the average longest wait for psychotherapy or counselling was seven months. Six trusts had longest waits of more than two years. Doctors are prescribing antidepressants because they feel there are few other options.

In January doubts emerged that the antidepressants prescribed by the doctors may not be as effective as they had been led to believe. Research published in the New England Journal of Medicine revealed that selective publishing of the results of antidepressant trials have given professionals and the public a skewed perspective.

The research compared all the antidepressant trials registered with the US Food and Drug Administration with those that made it to publication in a medical journal. Of the 74 registered antidepressant studies analysed, 38 gave a positive outcome. All but one were published. The remaining 36 studies returned either negative or questionable results for the drugs. Of these, 33 were not published or were published in a way that suggested their findings had been positive.

The result of this selective reporting is significant. While in reality 51% of the antidepressant trials had a positive outcome, any researcher, doctor or member of the public studying the published research would get the impression that 94% of the trials had been positive.

Inappropriate prescribing

The researchers did not investigate the reasons for non-publication, which may be based on a failure to submit manuscripts or decisions by journal editors not to publish. And they point out that each drug was still shown to be superior to placebo.

But they add: “The true magnitude of each drug’s superiority to placebo was less than a diligent literature review would indicate. By altering the apparent risk-benefit ratio of drugs, selective publication can lead doctors to make inappropriate prescribing decisions that may not be in the best interest of their patients and, thus, the public health.”

Tim Kendall, psychiatrist and deputy director of the Royal College of Psychiatrists’ research and training unit, found similar evidence of selective reporting when he studied the use of antidepressants for children in 2004. When this was countered and all the research made available to a group of professionals assembled for the study they decided they would prescribe only one of the six drugs tested. Previously they had been happy to prescribe them all.

Risk-benefit calculations are not just an academic exercise, as antidepressants can have potentially severe side effects. The National Institute for Health and Clinical Excellence (Nice) now recommends only one antidepressant for children after the others were linked to suicidal behaviour. Even that should be offered in serious cases only and used alongside talking treatments, the guidelines say.

The RCP says most people get only mild, if any, side-effects but different drugs can cause symptoms with varying degrees of seriousness. One commenter on Community Care‘s Mad World weblog described how an antidepressant given to him when he was a teenager “made me impotent when I began to take it and despite the fact I was only on the medication for four months my sexual functioning has never returned to date, almost 12 years later”.

Kendall says antidepressants should not be used for people with mild or moderate depression, pointing to the positive results from cognitive behavioural therapy and exercise.

But he argues that for more severe cases doctors need to look at combining antidepressants and psychological therapy. “If a drug treatment works it’s working within two to three weeks. If you’ve got someone with severe depression and you’ve got a way you can help quickly then we will use it. There are some decent trials which do show that Prozac is better than placebo in severe cases.”

Amplified placebo effect

But some commentators have a more fundamental problem with antidepressants. Joanna Moncrieff, psychiatrist and senior lecturer at University College London, says the overall difference between antidepressants and placebo is “very small” and could be due to the drugs’ sedative effect rather than any effectiveness in tackling depression. She also suggests there could be an “amplified placebo effect” with antidepressants because people know they are taking an active drug due to the side-effects.

One problem many opponents have with antidepressants is the explanation used by the drugs industry for how they work. Serotonin specific re-uptake inhibitors (SSRIs), the type of antidepressants introduced in the 1990s which include Prozac, ensure serotonin, which has been linked with mood, stays in the brain longer.

Moncrieff argues that evidence on the role of serotonin is “contradictory and inconclusive” but that has not stopped the pharmaceutical industry promoting the “chemical imbalance theory”. The theory is that depression is caused by a serotonin imbalance in the brain, which antidepressants can correct. But it has never been proved.

US researchers Jeffrey Lacasse and Jonathan Leo argued in the Public Library of Science in November 2005 that the claimed efficacy of SSRIs was cited as indirect support for the serotonin hypothesis. They compared this to saying that, because aspirin cures headaches, headaches must be due to low levels of aspirin in the brain.

Moncrieff goes further than this, suggesting that depression itself is a “pretty meaningless label that is applied to individuals with a huge array of different problems”.

She says: “It is better to try to work out what the problem is than to come up with a pseudo-medical label for it. When you have identified the problem, you can work out the best solution – perhaps marital therapy for relationship problems, social activities for loneliness, taking action at work for work-related problems.

“In my clinical experience, even severe depression is an emotional reaction to a situation that someone finds difficult or intolerable and the key is to identify what this situation is and how it can be changed.”

But Kendall is adamant: “Saying depression doesn’t exist doesn’t help. People still say they are miserable, have no positive feelings and want to kill themselves.

“We should have a broad approach to helping people with what we might loosely call depression and you probably do need to make your approach to each person individually but you’re still going to at times resort to medication and psychological treatments.”

Related article

Essential information on mental health

Contact the author

Simeon Brody

This article appears in the 13 March issue under the headline “Don’t bin the pills just yet”