Latest Issue

The drugs don’t work

A new assault by a leading psychologist on Prozac-style antidepressants claims they are worse than useless. Try telling that to the many people who believe they are life-saving

By John Cornwell   October 2009

I once saw my friend Stuart banging his head on the floor, weeping, threatening to kill himself. He was severely depressed, incapable of working, sleeping or eating. Yet within three weeks of being prescribed a course of a Prozac-type antidepressant drug, he was beginning to cope again and was back at work. As he got better he made this shocking remark: “If I was offered a choice between cancer and depression, I’d take the cancer.”

Stuart swears by his antidepressant medication, which he took for two years. “It saved my life,” he says. But the effectiveness of such drugs, known as SSRIs (of which more later), has been challenged for two decades. The latest anti-antidepressant blast is one of the most meticulously researched. In The Emperor’s New Drugs (Bodley Head) Irving Kirsch, a professor of psychology at the University of Hull, claims that SSRIs are worse than useless.

Depression is a devastating illness, causing great suffering in the afflicted and anxiety to their nearest and dearest: it can hit at any age. One in three of us will suffer a form of depression, from mild to severe, at some time in our lives. According to the EU’s public health committee, it is set to become the single most frequently suffered illness in Europe by 2020.

Before the advent of Prozac-type drugs, Stuart would have been hospitalised as a danger to himself, perhaps given electro-convulsive therapy, dosed with barbiturates or a crude sedative medication. Those were the bad old days of huge county asylums, when 50 per cent of NHS hospital beds were taken up by the mentally ill. By the 1970s, these institutions were being closed down, made possible by new types of drugs and the provision of talk therapies such as cognitive behavioural therapy (CBT). While patients suffering acute psychotic illnesses are still hospitalised in the psychiatric wings of general hospitals, depressed patients are usually treated as outpatients. So who could doubt the benefits of a class of well-tried antidepressants, which have been dispensed for two decades and account for 20m prescriptions a year in Britain? Their claimed success rate, based on clinical trials and practitioner reporting, is 70 per cent.

According to David Nutt of Imperial College, who heads one of the largest departments of psychopharmacology in Britain: “Antidepressants work in clinical practice, and everybody knows they work.”

But Kirsch has set out to challenge the antidepressant establishment. He claims that the SSRIs only appear to be effective because patients believe that they are going to work. “It’s placebo effect,” says Kirsch. In other words, the SSRIs are no better than blood-letting, snake oil, or any of the nostrums peddled before the era of modern medicine. Of course we all use placebos in everyday life, from kissing a child’s bruise better to taking a glass of brandy for a cold. I know a woman who saves red Smarties in a jar and takes one when she has a headache; she swears that it works even though she’s conscious that it’s placebo. Yet one would not want to build a health strategy for depression on such flimsy efficacy.

And Kirsch has another problem. “SSRIs have a range of insidious and hazardous side effects,” he points out. Clinicians, scientists and pharmaceutical manufacturers do not dispute this. But they think that Kirsch’s conclusions are wrong and could put a great many patients at risk. Kirsch, though, has no problem with airing their “dirty little secret.”

A lean and relaxed 67-year-old, originally from New York, Kirsch has never suffered from depression. He commutes between research programmes into placebo phenomena at centres in Boston, Florence and Hull. He used to practise as a clinical psychologist and so has a placid, thoughtful manner when interviewed. But for a decade he’s been conducting with colleagues what statisticians call “meta-analyses,” collating thousands of clinical trials to produce mega-statistical mother-of-all trials, most recently in February 2008.

Exploiting America’s Freedom of Information Act over the years, Kirsch claims to have extracted negative trial information kept from physicians and the public. These include data showing that in some trials, SSRIs did worse than placebo. Among the possible side effects of these “useless” drugs are sexual dysfunction, insomnia, long-term weight gain, diarrhoea, nausea, drowsiness, skin reactions, nervousness, anorexia and sweating—though these may not affect everyone who takes them. To this list Nice, the independent body which produces guidelines on the use of medicines in the NHS, recently added birth defects in the babies of women who were on the medication while pregnant. Among teenagers, the SSRI Seroxat has been linked with a tendency to think about suicide, as well as self-harm.

Kirsch is no lone voice in the wilderness. Peter Breggin, an American psychiatrist, had been criticising SSRIs since the early 1990s. He wrote Talking Back to Prozac (1995) to repudiate psychiatrist Peter Kramer’s Listening to Prozac (1993)—a bestseller which claimed that Prozac made patients “better than well.” Breggin complained that biological psychiatry, the practice of treating mental illness as a chemical imbalance, had been pushing out talk therapy for 50 years.

He was accused of being fanatical and unqualified. One pharmaceutical company made much of the fact that he flunked chemistry at school. Then there is David Healy of Cardiff University, a soft-spoken Irish psychiatrist who campaigns against the prescription of Seroxat for children and teenagers because of its link with suicide attempts. Healy was turned down for a job as a result, he claims, of his anti-Seroxat stand. Kirsch says he has been warned that his name on academic articles could result in a loss of pharmaceutical company funding for university departments. But he aims to bang the nails in the coffin of SSRIs with the “incontrovertible facts” in his book, the subtitle of which is Exploding the Antidepressant Myth.

Kirsch’s claims were first tentatively made in 2003, causing a furore in the psychiatric profession. Now he believes he has the killer fact. “The key to placebo effect in clinical trials is expectation. That’s why it is absolutely crucial during a trial that the clients have no inkling as to whether or not they are getting the active drug rather than a placebo pill.”

Research shows that if patients believe they are taking the real drug they are more confident of improving and so improve even if they are actually on the placebo. Conversely, if they suspect they are taking the placebo, their expectancy of improvement declines and so does their improvement. “The strict controls of trials are frequently broken in the case of antidepressants,” says Kirsch, “because the patient gets side effects from the real drug so the blindness of the trial is broken.”

In a clinical SSRI drug trial a number of patients with similar levels of depression are chosen. Half are selected at random to be given the active drug; others are given the placebo. The pills look identical. They are administered by a nurse or doctor who does not know which is which—hence the trial is “blinded.” At the end of the trial, interviews are conducted by another specialist who is also unaware of who got what, so the trial is twice blinded. Both real drug and placebo groups are then compared with a group that receives no treatment.

“The drug companies constantly single out a small comparative advantage of the real drug over placebo,” says Kirsch. “But patients on the real drug get the side effects. This means that the strict protocol of the trial has been broken. When you exclude that factor, it’s clear from the statistics that the drug does no better than placebo across all trials.”

In these trials improvement is measured by the Hamilton scale, on a range of points numbering one to 51. Kirsch reveals that the SSRIs on average score only 1.8 points above placebo. “It’s insignificant. You get a six-point improvement if the patient gets into a better sleep pattern,” says Kirsch. According to Nice, an improvement of at least three points on is necessary for the difference between a drug and a placebo to be even considered significant. Kirsch also notes that the more severe the side effects, the better the reported improvement in mood.

But Kirsch is contradicting two decades of favourable published clinical trials, and dozens of national and international drug approvals. The sale of antidepressant drugs is a $19bn-a-year global industry. He is also up against hundreds of thousands of GPs and psychiatrists who have seen their patients get better on SSRIs and who are alarmed that people will suddenly stop taking them, with serious withdrawal effects, or demand alternative therapies.

The US Food and Drug Administration and National Institutes of Health, as well as Nice, have repeatedly endorsed the efficacy of SSRIs. Above all, Kirsch is flying in the face of the gargantuan international Society for Neuroscience. I attended its 2007 annual conference in San Diego, California, for which 23,000 brain researchers were registered. Over five days some 10,000 research poster sites were displayed in relays in the exhibition halls, a significant proportion of them endorsing the neuroscience that underpins SSRI drugs. It is no secret that academic brain research is substantially supported by big pharma. But is it really possible that the mighty SSRI emperor is stark naked?

Kirsch is attempting to demolish an entrenched medico-scientific dogma, which states that you, your happiness and misery, joy and sorrow, vices and virtues, are no more than the state of your brain molecules. A few simple technical fundamentals and a bit of history help us to grasp the depth of his heresy.

“Behind every crooked thought lies a crooked molecule.” The man speaking to me was Ray Fuller, the inventor of Prozac. We were standing outside a courthouse in Louisville, Kentucky in October 1994, I was there reporting a historic liability suit against Eli Lilly, manufacturer of Prozac.

The plaintiffs were the survivors and the relatives of the victims of a shooting spree in the Louisville Courier-Journal printing plant. Eight people died and 12 were seriously injured when Joe Wesbecker went on the rampage with an AK-47, before shooting himself in front of the work manager’s office. The 47-year-old had been on Prozac for just two weeks after being diagnosed as depressed with workplace stress. Eli Lilly won the case, but later there were allegations that the company had made a secret payment to the plaintiffs, alleged by one lawyer to be “mind-boggling.”

Eli Lilly had all along denied that Prozac could have been the cause of Wesbecker‘s rampage, and Fuller was called as an expert witness to defend the drug he had created. On the steps of the courthouse he gave me an unscheduled tutorial.

Our brains, he explained, are composed of a hundred billion neurons. The prodigious electrical and chemical communications between them are essential to our thought processes and consciousness. People get depressed when serotonin, a natural chemical called a neurotransmitter, is depleted in the minuscule gaps between neurons, causing a lowering of communication thus a plummeting of emotional mood.

Prozac boosts serotonin levels and so relieves depression. It works by switching off receptor sites that normally reabsorb serotonin into the reservoirs of the cells for later reuse. The brain thus remains flooded with serotonin, increasing communication between nerve cells and so lifting mood. Hence the drug is called a selective serotonin reuptake inhibitor or SSRI. There are more than 50 neurotransmitters including serotonin in the brain. Prozac is supposed to work only on serotonin.

But not everyone agrees that depression is caused by a chemical imbalance in the brain. Scientists such as Steven Rose of the Open University, Gerald Edelman of the Neurosciences Institute in California, and neurologists António Damásio and Oliver Sacks have warned against reducing the mind to its molecules instead of taking a more holistic view of mental health.

A crucial psychopharmacological fact about the multi-billion dollar SSRI industry, however, is that no scientist, as Fuller conceded, has ever managed to measure the serotonin levels in the synaptic clefts of the brain of a living human being. It might be possible to attempt such measurements in rats, but you can’t ask a rat whether it feels depressed.

Kirsch does not think that depression is a chemical imbalance. He argues that trials “are often flawed or buried when they contradict the chemical imbalance theory.” A major weakness in the system is that “most national drug approval bodies rely on data volunteered by the pharmaceutical companies. It’s as if the fox has been hired to guard the hen house. So adverse data are often suppressed at the source.”

In advance of its publication, reaction to Kirsch’s book has been swift and condemnatory. Martin Baggaley, medical director of the South London and Maudsley NHS Trust, insists that most psychiatrists are convinced of the efficacy of SSRIs. He told me that the undermining of confidence in such drugs is dangerous: “It will lead to increased suffering and mortality through suicide.” Baggaley blames the media, and psychologists “who have a vested interest in constantly attacking antidepressants.” By that he means psychologists who make their living out of talk therapy. He says: “Depression is a major cause of suicide among young people.”

David Nutt, of the psychopharmacology department at Imperial College, worries that “public confidence in what are highly effective and safe drugs is in danger of being undermined.” Kirsch’s book was embargoed until its publication in early September and, at the time of going to press, Nutt had not yet had a chance to read it. But basing his view on Kirsch’s journal articles, Nutt complains about the trials in his meta-analysis: “These are not the best to work on.” He also says that Kirsch has no experience in conducting trials.

A more nuanced reaction comes from Peter Jones, professor of psychiatry at Cambridge University. Jones believes that the power of placebo deserves closer attention. And he adds that concerns about suicidal teenagers are overplayed: “While I am no apologist for big pharma, the fact is that in clinical trials of teenagers there has not been one suicide recorded.” Jones also notes that in trials the more “mild” the depression, the less the effect of SSRIs. As the depression increases in severity, the effect of the drug is more marked. Kirsch counters that he addresses this objection by showing that as the dosages increase with severity, so do side-effects, which lead patients to realise that they are taking the real drug.

As I went back and forth between Kirsch and the psychiatric and psychopharmacological experts, it struck me that there may be no final answer to these statistical points. But the crucial question is the direction of overall treatment of depression in Britain. On the one hand, the pharmaceutical companies need to keep developing new drugs as older ones pass out of patent; on the other, the NHS needs to find cost-effective and efficacious treatments for an enormous range of depressive illnesses in severity and in kind.

Jones believes that the key to best practice, based on the experience of the past two decades, is the patient’s own involvement in the choice of therapy. He is enthusiastic for the new nationwide scheme known as Improving Access to Psychotherapy (IAP), which is now well advanced in his area, East Anglia, where some 750 ancillary workers are in training to guide people to the most suitable therapy. Jones claims that 10 per cent of the population in any one year suffers from mild forms of recurring depression and anxiety, leading to sick leave. The evidence is that CBT is the best strategy for this group. When that fails antidepressants may be called for.

Among the GPs I have spoken to, one in three said they were unsure about whether to prescribe antidepressants except in the most serious of cases. One doctor who has worked in deprived areas, told me: “Prozac came in with a lot of media hype and people got to see it as a quick fix. So they come in demanding them, even for their children.

But adverse reports have now confused the issue. Most GPs don’t have the time to engage in talk therapy, and the availability of CBT is patchy. The IAP approach mentioned by Jones is welcomed, but there are questions about the provision of qualified help beyond the initial counselling.

Then there is my friend Michael, a recovering alcoholic. Michael was severely depressed following detox, which is normal, and his doctor prescribed an SSRI. “It’s daft to say they don’t work,” he tells me. “I don’t know, and I don’t care, whether the drug does what it’s scientifically meant to do, or not. All I know is that I feel that it has done me some good and I’m going to go on taking it. I’d rather trust my GP, frankly, than some guy who has written a book.”

We want to hear what you think about this article. Submit a letter to

More From Prospect