About the SA Blog Network



Critical views of science in the news
Cross-Check Home

Psychiatrists, Instead of Being Embarrassed by Placebo Effect, Should Embrace It, Author Says

Email   PrintPrint

Walter Brown, a professor of psychiatry at Brown and Tufts, first caught my attention in the mid-1990s when I was researching my December 1996 Scientific American article “Why Freud Isn’t Dead,” on lack of progress in psychiatry. My research persuaded me that the placebo effect (which I have written about here and here) accounts for most if not all of the benefits of psychotherapy and drug treatments for depression. Brown provoked a furor among his colleagues by proposing that psychiatrists prescribe placebo pills for mildly and moderately depressed patients, a topic that he revisited in a 1998 article for Scientific American. He has delved even further into the implications of the placebo effect on psychiatry and other fields of medicine in his incisive new book The Placebo Effect in Clinical Practice (Oxford University Press), which I highly recommend. I recently interviewed Brown:

Horgan: What have we learned about the placebo effect since Henry Beecher’s landmark 1955 work The Powerful Placebo?

Brown: Beecher’s paper put the placebo effect on the map, and his general proposition that the placebo effect is ubiquitous has withstood the test of time.  He also proposed, among other things, that there is a constant placebo response across conditions. 35.2 % improve with placebo was the figure he came up with. But since then, thanks to thousands of placebo-controlled clinical trials, we have learned that some conditions are far more placebo responsive than others.  Even among pain syndromes, which for the most part show robust placebo effects, there are differences: for example, post operative pain appears more placebo responsive then migraine headaches. About 40% of mildly to moderately depressed patients improve with placebo as opposed to only 10-20% of those with obsessive compulsive disorder. Irritable bowel syndrome is highly placebo responsive–about 40% improve with placebo–whereas only about 20% of people with chronic fatigue syndrome get better with placebo.  (I cover these issues in my book’s chapter on variations).

Since Beecher, laboratory based and clinical studies have identified some of the mechanisms behind the placebo effect. Expectation is the most widely studied. Both rigorously designed and controlled laboratory studies and clinical studies as well show that what one anticipates from a treatment has a profound impact on what one does experience. In the past decade a number of studies have shown that when people get placebo but believe that they’re receiving a medication they undergo some of the same brain changes that occur with the active medicine. (I go into the details in the chapter on expectation). The effect of expectation on response to placebo and other treatments seems to rest on a fundamental psychobiologic process (whatever that means). In the past 60 years we’ve also learned about the role of conditioning in the placebo effect and–a special interest of mine–the role of certain elements of the treatment situation and doctor-patient relationship in bringing about placebo effects.

Horgan: Haven’t clinical trials eliminated concerns that many modern medical treatments, when they work, are harnessing the placebo effect?

Brown: Although drugs need to demonstrate efficacy in controlled clinical trials in order to get FDA approval, many widely used treatments are not subjected to clinical trials, including psychotherapies, surgical procedures and all the alternative treatments. And once a drug gets FDA approval it can be used in a so-called “off-label” manner for any condition including those not studied in the clinical trials that led to approval. Most drugs are frequently used “off-label” for conditions in which their efficacy has not been carefully studied or studied at all. Even placebo-controlled trials are no guarantee that a drug that looks good is not deriving its benefit from the placebo effect; double blind clinical trials are not truly double blind. Even though they are designed to eliminate the bias that comes from knowledge about whether drug or placebo is on offer, almost invariably the investigators conducting the trial know, because of side effects, who’s getting what.

Placebo effects continue to be mistaken for treatment effects with troubling frequency. As just one example, vertebroplasty–injecting cement into a fractured vertebra–was widely used as a treatment for vertebral fracture from the early 1990s through the first decade of this century until a controlled trial showed that a sham (placebo) procedure (nothing injected) was equally effective in  reducing pain and disability.  (I discuss this particular study in the first chapter)

Horgan: Do you worry that raising the awareness of patients about the placebo will undermine patients’ trust in modern medicine?

Brown: It may cause people to wonder if their improvement is “just” a placebo effect and if the treatment they’re getting is not “really” working. But I believe that most folks trust their own doctors–if not doctors in general–and will believe what their doctors tell them about a treatment’s inherent effectiveness.

Horgan: Why do you focus in your book so much on psychiatry?

Brown: My original concept for the book was to look at the placebo effect in mental health alone. But as I started to do the research and write it I decided to look at the placebo effect more generally and go beyond psychiatric illness to medicine in general. Some of the focus on psychiatry derives from the original impetus for the book. I also focus on psychiatry because my own research on the placebo effect has been in depressive illness, and the condition in which the placebo effect has been most studied is probably depression. A good bit of what we have learned about the placebo effect in depression sheds light on the placebo effect in other conditions. It’s also the case that a number of psychiatric conditions are highly placebo-responsive. Also psychotherapy has a lot in common with placebo treatment; the relationship between the two is a matter of controversy and I wanted to tackle that issue. And finally, although I believe that the placebo effect is pertinent to all illnesses and treatments, as a psychiatrist my expertise and interests lie primarily in psychiatry.

Horgan: What are the implications of the placebo effect for psychiatry? Given the side effects of many psychiatric drugs, should psychiatrists prescribe placebo treatments more often for mental disorders?

Brown: Given the high rates of improvement with placebo–close to the rates with drugs–in some psychiatric conditions such as mild to moderate depression and panic disorder–and the side effects and expense of drugs, I think it does make sense for psychiatrists to prescribe  placebo treatments in some circumstances. The placebo could be a pure placebo—i.e., a sugar pill–or a nontoxic alternative therapy given to promote a placebo effect.  I go into the details of how to go about this and the ethical and clinical implications in the last chapter. Of equal importance, psychiatrists like all health professionals should apply what is known about mobilizing the placebo to enhance the benefit of all treatments.

Horgan: Do you agree with Jerome Frank [a prominent investigator of the efficacy of psychotherapy] that psychotherapists, like shamans and faith healers, are just harnessing the placebo effect?

Brown: I wouldn’t say “just”; the placebo effect can be pretty powerful and harnessing it is not a trivial intervention. But I do agree with Jerome Frank that psychotherapists, shamans and faith healers accomplish what they do by providing the common factors of treatment–the presence of a healing authority, a healing ritual, expectation of recovery, etc.–that are also found with placebo treatment, that promote a placebo effect, and that are probably the active ingredients of all the psychotherapies.

Horgan: Have your writings about the placebo effect in psychiatry gotten you in trouble with other psychiatrists?

Brown: In 1994 the journal Neuropsychopharmacology published a paper in which I proposed that in some circumstances depression should be treated with placebo. The paper was followed by invited commentaries from three psychiatrists, two psychologists and one internist. All but the internist freaked out over the idea–it was irresponsible, unethical, dangerous, etc. I would guess that some psychiatrists will object to what I say about the commonalities between psychotherapy and placebo, and others will object to and have objected to my position on the similarity in outcome between placebo and drugs for mild to moderate depression.  Oh well.



About the Author: Every week, hockey-playing science writer John Horgan takes a puckish, provocative look at breaking science. A teacher at Stevens Institute of Technology, Horgan is the author of four books, including The End of Science (Addison Wesley, 1996) and The End of War (McSweeney's, 2012). Follow on Twitter @Horganism.

The views expressed are those of the author and are not necessarily those of Scientific American.

Rights & Permissions

Comments 10 Comments

Add Comment
  1. 1. tmonk 8:50 am 03/12/2013

    Neither you nor the author seem to have the slightest idea of what major depression is.Most scientists find at present that the best model is an impaired regulation of the default network (of which there is no mention). This is based on an attempt to try to apply rigorous scientific experimentation-something also not mentioned.
    I am not sure why I even bother to reply any more to lay literature except to help prevent perhaps one person to take the conversation transcribed seriously.If you want to do any public service, I suggest passing on puckish and pursuing a modicum of elbow grease and interviewing someone who is a scientist, rather than one promoting selling a book.
    One second thought-the suggestions here are rather appalling.I suggest you look at the clip of Richard Feynman on his 1 min opinion on research in the social sciences- hopeful and sobering,and not self serving.

    Link to this
  2. 2. stanleyh 11:57 am 03/12/2013

    The general concept of placebo sometimes obscure rather than enlighten the complex interactions between mind and body. “placebo” puts under the same umbrella effects as diverse as:
    - the common biases, flaws and limitations in clinical trials and clinical practice that tends to exaggerate the real physiological improvements that took place (research is done by fallible humans, not gods),
    - the average natural course of many illnesses is improvement with time (with or without fake pills),
    - real physiological positive effects of treatment that are mediated by psychoneuroimmunology, psychology, expectations, conditioning, and other brain/body influences (given the known anatomy of the nervous system, including the brain, and its interaction with so many organs and hormonal systems, it would be astonishing if the mind was not able to influence many diseases),
    - physiological effects of treatment mediated indirectly by the mind (motivation, behavior), by having somebody take better care of their body through improving nutrition/rest/sleep/activity/…

    Knowing the aggregate placebo importance (like the often quoted 30% placebo improvement) of those independent factors is borderline useless, and at worst very misleading.

    Finally, it should be clear that deception/lying (even by deliberate omission rather than straight lie) towards patients is both immoral and dangerous. If someone does not think he/she is able to reasonably raise people expectations of wellness without resorting to lying (or without realizing the long-term dangers of such lying), maybe he/she should not be a doctor. For what is worth, research shows that explaining people they are given a placebo does not necessarily impact the positive results.

    Link to this
  3. 3. rshoff 6:15 pm 03/12/2013

    The placebo effect is not real from the perspective of treating underlying conditions. When doctors see patients improve with placebo treatment, they are simply seeing a change in a patients attitude and how that in turns changes the way the patient describe his/her symptoms. Medicine has never, or rarely, empirically assessed the actual medical condition. It has focused on managing symptoms. Therefore, the placebo effect helps trick patients into reducing their complaints. It does not resolve the underlying condition. In fact, it pushes any acknowledgement of an underlying condition right under the rug. Prescribing placebos should be illegal. Placebos have not been approved by the FDA for particular medical conditions the way pharmaceuticals have. Granted, drug companies are pushers of a sort. However, those drugs save millions of peoples lives around the world every hour of every day. Many more millions are provided with a better life because of real drugs, not sugar pills. Next time you have strep throat, pop a sugar pill. I’m sure you’ll feel better in a few days, at least until the rheumatic fever sets in….

    Link to this
  4. 4. IslandGardener 6:17 pm 03/12/2013

    The commentators don’t seem to have read the article carefully enough.

    For example, tmonk says ‘Neither you nor the author seem to have the slightest idea of what major depression is’ but the article says things like ‘Given the high rates of improvement with placebo–close to the rates with drugs–in some psychiatric conditions such as mild to moderate depression and panic disorder–and the side effects and expense of drugs, I think it does make sense for psychiatrists to prescribe placebo treatments in some circumstances’.
    Not major depression then.

    stanleyh says ‘Knowing the aggregate placebo importance (like the often quoted 30% placebo improvement) of those independent factors is borderline useless, and at worst very misleading’ – but why?
    The article says ‘the placebo effect can be pretty powerful and harnessing it is not a trivial intervention’.

    This to me is the essence of it – the ‘placebo effect’ is in fact our own powerful self-healing abilities, and if anybody can help us to release it, then we can get better, physically or mentally. We may need other things as well, but this amazing self-healing ability is something we need to study and make better us of it.

    Link to this
  5. 5. rshoff 6:22 pm 03/12/2013

    “what one anticipates from a treatment has a profound impact on what one does experience.”

    Interesting choice of words: ‘experience’ What exactly does one ‘experience’?

    “In the past decade a number of studies have shown that when people get placebo but believe that they’re receiving a medication they undergo some of the same brain changes that occur with the active medicine.”

    Simply put, I don’t believe it. I will admit that there may be an overlap between placebo and a pharmacological prescription, however, that overlap is limited to patients ‘experience’ of symptoms, not the progression of the illness.

    Link to this
  6. 6. rshoff 6:30 pm 03/12/2013

    Attempting to self-heal with placebo is at the expense of identifying and treating an underlying condition. For example, depression (and anxiety) kills and a truly depressed person is at risk of dying if given a placebo when there is a true underlying condition that the brain already cannot ‘self heal’ that goes untreated. If it could self heal, it would. Even psychiatrists understand that our consciousness is a very small part of brain function (I prefer to think of it as happenstance). If we were able to will ourselves to heal, we would be able to will ourselves dead. And that we cannot do. There are many things we cannot will, and to imply that tricking our conscious selves will lead to a healing experience is extremely unkind and unfair to the many people who are suffering and dying at no fault or weakness of their own.

    Link to this
  7. 7. stanleyh 12:15 am 03/13/2013

    To IslandGardener: I am sorry if my comment was not clear. I consider “placebo” research as very important, and I also agree that harnessing its power is important.

    My criticism was narrow and focused on:
    - aggregating in the same number or the same concept things like bias in self-report and things like psychoneuroimmunology effects (each are important, but the aggregate is ambiguous and meaningless).
    - the suggestion that prescribing placebo pills without full disclosure or any kind of deception about treatment is ever a good idea.

    Link to this
  8. 8. quincybones 2:04 am 03/13/2013

    As a dentist who has used conscious IV sedation for anxious patients for many years I combine sedative and anxiolytic drugs with local anesthesia. However,I also routinely use positive suggestion in order to maximize the placebo response. I have no problem with this, as the benefits include much lower drug dosages,increased patient safety,and faster postoperative recovery. For optimal outcomes, trust building is also achieved via an initial “iatrosedative” interview. Here,fears are discussed,along with the procedure to be used. Results confirm this approach,with minimal,or even spectacular results, where no drugs have been injected after the IV cannula was placed. Therapeutic use of placebos is all too often underutilized in clinical practice.
    (Published references are available on request)

    Link to this
  9. 9. shimagyoh 2:45 pm 03/19/2013

    Medical ethics demand that at least two indispensable conditions must be satisfied before a placebo is given: First, thorough history must be taken to ensure that resolvable social problems are not the cause of symptoms, e.g. marital or work-place problems. Second, thorough examination and investigations must be done to exclude some underlying organic disease where strongly suspected. Then the most difficult part, the practitioner has to decide whether it is right to play the confidence trick, whether the end justifies the means. When you consider that some 70% of patients vising a doctor’s general outpatient clinic have only psychosomatic problems and the majority would not take kindly to scientific reasoning, in my view, it does!

    Link to this
  10. 10. rshoff 12:45 pm 03/25/2013

    @shimagyoh, I would tend to agree with your idea, but with two exceptions.

    1 – A third party does not have a right to mislead you. The end does not justify the means because no one has the right to take that power over another person without full disclosure and their permission. Which would defeat the intent of a placebo treatment.

    2 – Once a condition is determined to be psychosomatic it will always be labeled psychosomatic. No one will ever take a second look to try and diagnose the patient once labeled.

    So, instead of a physician prescribing a placebo treatment, they need to be honest with their patient. Perhaps by saying, “I don’t know”. “I don’t know of a treatment for these symptoms”. “I’m sorry”. “I’m sorry I can’t help you feel better”. Then the patient may be open to other general suggestions. Such as lifestyle changes to compensate. Nutrition, PT, Exercise, etc, etc, etc. Heck, at that point, meditation may even be suggested as a lifestyle change that could, for some unknown reason, help them feel better.

    Link to this

Add a Comment
You must sign in or register as a member to submit a comment.

More from Scientific American

Scientific American MIND iPad

Give a Gift & Get a Gift - Free!

Give a 1 year subscription as low as $14.99

Subscribe Now >>


Email this Article