Joanna Moncrieff,
Senior Lecturer UCL,

I have renamed my talk Drug Treatment in Modern Psychiatry, The History of a Delusion, because it occurred to me that the standard beliefs about modern drug treatments in psychiatry are similar to delusions. They are fixed and probably false, and based on a distorted reading of the evidence.

Almost every person, who sees a psychiatrist at the moment, is prescribed at least one sort of psychotropic drug. All the drugs that are commonly prescribed were introduced into psychiatry from the 1950’s onwards. Some people talk of this period in the 1950’s and the introduction of this new range of drug, as a revolution, the psycho-pharmacological revolution. According to this view, at last psychiatry had found new and effective remedies that helped transform it into a modern medical speciality. What I want to suggest, is that things were actually quite different. The way psychiatry is practised today is not very different from the way it was practised 50 or 100 years ago. What is different however, is the way that psychiatry views itself, and particularly its drug treatments. What is different is psychiatry’s representation of its activities.

My first slide (table 1) shows a list of drugs that were in use prior to the 1950’s.

When I started looking into the history of drug treatment of psychiatry, it was quickly obvious that drugs have been used extensively since the modern Institution of Psychiatry gets going in the early 19^th Century. Most drugs that were in use prior to the 1950’s were various forms of sedatives. In the 19^th Century, Opium and Bromides were used. In the 1920’s and 1930’s various Barbiturates were synthesised and these became probably the most popular drugs in the first half of the 20^th Century. Paraldehyde was also used extensively, and from the 1930’s some stimulants started to be used.

In my research with patient case notes from the Maudsley and Bethlem Hospital from the 1930’s and 1940’s it was obvious that drugs were used extensively. All patients, and this includes both inpatients and outpatients, were on at least one drug and most were on several. However, drugs were rarely documented or discussed in the actual notes. You had to look at the medication card to find evidence that drugs were used at all. Similarly there was almost no mention of drug treatments in textbooks of the time. Maybe a line would refer to the fact that sedatives might occasionally be useful in such and such a condition. And there were very few articles published on drugs in the research journals of the time. In fact the only ones we found concerned drugs for the treatment of epilepsy.

Joel Braslow, a Psychiatrist and Medical Historian, finds something very similar in his study of a large state Psychiatric Hospital in the United States. He also found evidence of extensive use of drugs. He suggested that drug treatment was paid little attention because drugs were regarded as crude and blunt instruments that could be used to achieve sedation. They were effectively regarded as chemical restraints. They were not regarded as specific treatments and were not thought to have any effect on the disease process.

In contrast, you will see on the slide that I have listed Thyroxine and sex hormones. For a brief period in the 1940’s there was a theory that schizophrenia was caused by thyroid dysfunction, and thyroid hormone was prescribed quite commonly to people with schizophrenia. The decision to start thyroid hormone was commonly noted in the case notes after a ward round. There was also a theory that schizophrenia was caused by sex hormone abnormalities, and the research papers recommended that sex hormones could be administered as a treatment for schizophrenia. However, in my research I did not find any evidence of them actually being prescribed.

In contrast to drug treatments, great attention was given to the various physical treatment procedures that were tried out on psychiatric patients, particularly in the middle decade of the 20^th Century (slide/table 2).

Now psychiatry has a long history of subjecting patients to bizarre, dramatic and usually painful and barbaric procedures. Sometimes these were intended merely to restrain patients. Sometimes these procedures were thought to cure a specific condition. However, the middle decades of the 20^th century were a period of great enthusiasm for physical treatments. New treatment procedures caught on widely and they became the focus of asylum life. Some of the procedures came to be accepted as specific treatments for specific conditions.

I am going to talk a bit about insulin coma therapy and shock treatment. The slide shows an advertisement for Insulin, for use in Insulin Coma Therapy. Insulin was isolated in 1922, and the idea of Insulin Coma Therapy was introduced into psychiatry in 1933, by an Austrian Psychiatrist called Manfred Sackell. He had previously been using it to treat Morphine addiction. He claimed that it had very good results in the treatment of schizophrenia, and it was always regarded as a specific treatment for schizophrenia. A German Psychiatrist writing in the 1940’s claims that the introduction of Insulin Therapy was "the decisive step from a purely symptomatic to a curative view of therapy", this Psychiatrist also describes how within a few years, every psychiatric clinic in Germany was doing Insulin Coma Therapy. It had a very high mortality rate. It may have had a tranquillising effect on patients by inducing brain damage through the prolonged deprivation of the brain cells of glucose, as suggested in a new book by journalist Robert Whitaker, called Mad in America. It was also a very dramatic procedure, with patients being put into this long coma, and then re-awoken quite suddenly by the injection of glucose. An SHO in Psychiatry in the 1950’s suggested that its effect was actually a placebo effect, and a result of the drama of the whole procedure. A randomised trial was then conducted from which it was concluded that insulin coma therapy was not effective.

The important point that I want to stress is that Insulin Coma Therapy was always regarded as a specific treatment for schizophrenia, and was probably the first thing in common use, that was.

The 1930’s were also period when shock therapy was introduced. This slide shows a picture of a man having an un-modified convulsion induced by Metrazole. Subsequently electricity was used to induce the fits, in a procedure known as ECT. Shock Treatment came to be associated with the treatment for depression and manic depression, although at first it was suggested as a treatment for schizophrenia and, in fact in reality, it was administered to the vast majority of patients in Institutions, when it was at its most popular.

So by the 1950s psychiatry had two specific treatments: insulin coma therapy for schizophrenia and ECT for mood disorders.

This was the context into which the new range of drugs was introduced in the 1950’s. Psychiatric Hospitals had become places of feverish activity focused around various quite dramatic physical procedures, particularly Insulin Coma Therapy and ECT and Chemical Shock Therapy. With the advent of Insulin Coma Therapy, there had begun to be the idea that there were specific treatments for specific disorders. The next slide (table 3) shows the rise in the use of Chlorpromazine in French Hospitals after its introduction in 1952.

Its use started sweeping across the United States from its introduction in 1954, fuelled by a huge promotion campaign by Smith, Klein & French, who marketed it as Thorazine. At first the new tranquillisers were regarded as just that, a new form of tranquilliser. The effect of taking these drugs was explicitly compared to having a lobotomy, and they were thought to induce "chemical lobotomies". However, by the 1960’s the view of what these drugs were up to, had fundamentally changed. It was now being suggested that they were in fact, specific treatments for schizophrenia or psychosis. Textbooks of this period start to say that the major tranquillisers have some impact on the fundamental biological disorder that underlies psychosis. Hence, they were named anti-psychotics. It was this view that led to the Dopamine hypothesis that suggests that schizophrenia is caused by over-activity of Dopamine or Dopamine receptors, and the pharmacology of drug treatments now basically sets the agenda for biological research into psychiatric disorders. Researchers are still trying to amend and contort the Dopamine hypothesis, despite the fact that its obvious and glaring problems have been pointed out at least 30 years ago.

In contrast to the old drug treatments, the new drugs were given great attention and a rapturous welcome. Their use was recorded and discussed in the patient case notes and letters. Numerous journal articles were published on them; about 80-90% of articles in 1960 and 1965 concerned new drug treatments. And whole sections and chapters of textbooks were now devoted to them.

So within 10 years, we rapidly go from a situation where drug therapy was regarded as so mundane, that it was not even mentioned to one in which drug treatment is seen as the centre piece of treatment and the stimulus to understanding the fundamental nature of mental disorder.

I am going to take the evidence on antidepressants as an example. These slides just show you how the use of anti-depressants has escalated over the last 10 years or so, since the early 1990’s. This coincides with the introduction of the SSRI’s, a new and therefore, expensive class of anti-depressants.

The first study of Imipramine with depressed patients was set up in 1955, and the results were published in 1958. It was Roland Kuhn who conducted this study who really formulated the idea of an anti-depressant drug.

I started looking at anti-depressant research, because I felt intuitively and logically that the idea that a drug can cure depression is a very strange one. There are thousands and thousands of trials of anti-depressant drugs, so it is very difficult for any one person to have a real over-view and a very thorough knowledge of all this research. However, it seems to me, and there are quite a few other people saying this in the scientific literature at the moment, that there are various problems with anti-depressant trials, which probably mean that the effects of anti-depressants have been over-estimated (table 4).

One of these is the problem of unblinding. This is the idea that in supposedly double blind trials, patients and investigators will get to know whether they are on the active medication or the inert placebo, because of the physiological effects of being on active medication, including side effects. Therefore the trials are not truly double blind, and the outcome may be influenced by the expectations of all involved. I think there are also problems about the validity of the measurement instruments that are used. For example it has been suggested that the Hamilton Rating Scale for Depression merely reflects the effects of anti-depressant drugs, rather than actually rating the state of depression itself. In trials where the people are on long term treatment, and some people are randomised to have their treatment withdrawn and put on placebo, I think you often get some sort of discontinuation effect, which can be mistaken for signs of relapse or worsening of the condition. There are also problems of a publication bias, and methods of analysis. It has been shown, for example, in a recent meta-analysis by P. Bollini and colleagues, (British Journal of Psychiatry 1999) that analysing only the people who complete treatment, gives a biased result in favour of the anti-depressant treatment. So I believe that all these problems raise serious doubts about whether anti-depressants work at all, i.e. whether they really have any alleviating effect on depression.

I looked at a small group of trials that had used active placebos, placebos containing some sort of active substance that was used in order to try and mimic some of the side effects of the drugs used, and I found that most of these found only very small and mostly non-significant differences between the anti-depressant and the active placebo.

However, I was most surprised when I looked at some of the other trials that had used ordinary inert placebos. Some of the largest and best known of these trials are actually also negative, although they are not generally acknowledged as such. For example, the famous MRC Trial that was conducted in this country does not find any significant difference between Imipramine and placebo on its main categorical outcome. In addition the next slide (table 5) shows that there was virtually no difference between imipramine and placebo on the symptom scale which was used.

Also a very large trial in America conducted by the NIMH also basically found no difference between Imipramine and placebo. Of course neither of these trials was reported quite like that. In the MRC trial, they focus on another way of categorising the patients, which produces a small and just about significant difference. It is also notable that they don’t present overall results of their symptom scale, presumably because this showed no difference. In the NIMH study, they report in detail, minor differences between Imipramine and Chlorpromazine, and they report that three weeks into the seven week trial, there seem to be a small difference in favour of Imipramine over placebo on some items, however, they fail to mention that basically at the end of the trial, there was no difference between the two drugs. To be fair to the authors, in their conclusions they do acknowledge that the differences they found between Imipramine and placebo were small and of doubtful clinical relevance.

This is just the evidence about antidepressants. A professor of psychiatry once said to me that the evidence base for antidepressants was, he believed, the most secure in psychiatry. If that was suspect then everything in psychiatry was. And that is pretty much how I think it is! I think you can look at the evidence on most drug treatments in psychiatry and find that it is very much less convincing than it is usually portrayed.

Prior to the 1950’s drugs were understood according to the properties of the drugs themselves. They were regarded therefore, as producing pretty crude effects, generally either sedative or stimulant. They were not thought to mimic natural biological states. Many drugs were used for their sedative and restraining effects, and stimulants were used, apparently to try and pep up or stimulate depressed patients.

After the 1950’s, in the era of the anti-biotic, and also after the isolation of various hormones, drugs came to be understood in a different way. Psychiatric drugs came to be seen as working by influencing the disease process itself. This suited the needs of psychiatry and also fitted in with the views that had been held about some of the physical treatments that had become popular, especially Insulin Coma Therapy. This was a much more sophisticated and glamorous view of drug treatment, and for people working in a very biologically orientated framework, it seemed to present a more optimistic view of psychiatric intervention. However, I would argue that it is a totally false perception of the way that drugs work.

It was obviously in the interests of the Pharmaceutical Industry to be able to promote drug treatments as specific treatments, rather than simply as symptomatic and restraining devices.

The way that modern psychiatric drugs are regarded has helped to present psychiatry as an activity akin to medicine. Throughout the 20^th Century, Psychiatrists were engaged in a campaign to improve their status within medicine and versus other disciplines.

This is the argument that psychiatry acts as a covert or disguised form of social control. I have found in other research I have done on the Mental Health Act, that often the Government in the UK has been more enthusiastic about a medical and biological approach to psychiatry than the Psychiatric Profession itself. A medical view of psychiatry is appealing to Governments because it presents a wonderfully simple view of what are in fact very complex problems. By doing this, however, it prevents Society from formulating an appropriate and helpful response to madness and mental distress.

I will finish this talk by just reading out to you a quote by Joel Braslow, which I think sums up the misrepresentation of psychiatric drug treatments.

"The advent of antipsychotic drugs has allowed me to therapeutically discipline my patients and to simultaneously reaffirm the biological nature of their disorder. Like our predecessors we are locked within an epistemological framework that cannot easily distinguish between control and cure"

What he is suggesting, and I would agree with, is that we, as psychiatrists and as a society, are still using drugs to sedate, restrain and sometimes to punish people; but we are pretending that we are not.

April 2002.

Much of this talk is based on material published in the following papers:

J. Moncrieff. (1999) "An investigation into the precedents of modern drug treatment in psychiatry" History of Psychiatry, volume 10, 475-490

J. Moncrieff (2001) "Are antdepressants over-rated?" Journal of Nervous and Mental Disease, vol 189, 288-295.