Articles

Placebo is not what you think it is

If I read one more crappy article about placebos, something’s gotta give, and it’s gonna be my head or my desk. Wired magazine has a new article entitled, “Placebos Are Getting More Effective. Drugmakers Are Desperate to Know Why.” Frequent readers of skeptical and medical blogs will spot the first problem: the insanely nonsensical claim that “placebos are getting better”. This not only “begs the question,” but actually betrays a fundamental misapprehension of the concept. I’ve written several times about the nature and ethical implications of placebos, but it’s time for a serious smackdown.

In clinical studies, “placebo” refers to a “treatment” which, compared with the test treatment, is inert. If I want to test a blood pressure pill for basic efficacy, I could simply give it to a group of people and see what their pressures are before and after taking the medicine. But that wouldn’t account for any blood pressure changes that occur simply from being part of the study or by chance alone. Such effects include paying better attention to one’s health and habits, trying to please the doctors, and other less clear and tangible effects.

So, to see what the real effect of the new pill is, it can be tested against a identical-appearing dummy pill, which we call “placebo”. Neither the experimenters nor the subjects know which pill is which. Then, the change in blood pressure between the real and placebo pill groups, and perhaps a no-pill group as well, can be compared to see if and to what extent the new pill lowers blood pressure; the key point being that this allows us to see how much of the blood pressure lowering effect is due to the new pill vs. other undefined factors be they random or due to being in a study.

The placebo “effect” is just that—an effect observed because of a particular situation. It does not show that there is some special benefit to sugar pills, but that when we observe people, we can measure changes that are not always due to our intended intervention. It is not possible to create strong or weak placebos, since the placebo effect is a measure of poorly defined effects and of chance alone. It is what it is.

The author of the Wired article doesn’t get it in a very profound way. I’ll give you a few laughable examples.

The fact that taking a faux drug can powerfully improve some people’s health–the so-called placebo effect–has long been considered an embarrassment to the serious practice of pharmacology.

Really? That’s not even wrong. Clinical pharmacology research depends on placebo controls to make sense of the data. The fact that certain data is affected in placebo groups does not, for example, mean everything we understand about the cytochrome P450 system is wrong.

He asserts that more and more drug studies are crapping out due to the placebo effect:

It’s not that the old meds are getting weaker, drug developers say. It’s as if the placebo effect is somehow getting stronger.

No, it’s not like that at all. Perhaps the studies are just that well done, or maybe the drugs being developed suck, or maybe companies are studying more candidate drugs and screening for efficacy. Just about any explanation that doesn’t involve aliens is better than “placebo is getting stronger”.

He goes on to talk about how placebo has become a crisis of the industry, but I have another explanation: it’s not “placebo” that’s the problem. If drugs in testing cannot outperform placebo, then the researches have done a good job of testing the drugs honestly. If the researchers are failing to develop drugs that beat placebo and the company’s bottom line is suffering, it’s not the fault of the sugar pill. Sometimes it’s either difficult or impossible to develop an effective medication. Failure is inevitable. It’s how science works. If the CEOs don’t like it, they have to either make up the data, or find a new business model.

Posted in: Science and Medicine

Leave a Comment (79) ↓

79 thoughts on “Placebo is not what you think it is

  1. jmm says:

    “It is not possible to create strong or weak placebos, since the placebo effect is a measure of poorly defined effects”

    This seems like a profoundly unscientific statement. Since the effects are poorly defined, this means they are not currently scientifically understood, which means that we do not know at this point in time what is possible and what is not. If we did really understand the placebo, maybe we could make a statement like that. Or maybe we would be making just the opposite statement. How can you know?

    Science is the study of external world as we observe it, not just those parts of it for which we already have a good mechanistic model.

  2. Harriet Hall says:

    One could argue that some placebos do seem to be “stronger” than others. There is a hierarchy: sham surgery > injections > pills. The color of the pills, the frequency of dosage and other factors have been shown to affect the response to placebos.

  3. Peter Lipson says:

    @jmm
    I half agree with you.

    Perhaps you can measure the magnitude of a placebo effect (that is, a placebo group vs completely untreated group) and you can dissect the various components of placebo, at least as it occurs in a particular study, since it’s likely that rather than “placebo”, there exists “placebo phenomena” that are different from situation to situation.

    I think the problem is not lack of coherent scientific analysis of “the placebo” but recognition that “placebo” is an inconsistent variety of effects, measurement artifacts, etc, and not an exploitable therapy.

  4. jmm says:

    Peter, I half agree with you too ;-) I agree that the “placebo” is an inconsistent ragbag of stuff that includes measurement artifacts, and I agree that studies to dissect it are both possible and much-needed. Pending the outcome of those studies, however, I think it is premature to say that it is not an exploitable therapy. Maybe not, maybe so, maybe it is for some conditions but not others. The evidence is clear enough already that it is not exploitable for cancer, but for pain and depression I think the data so far point the other way.

  5. Peter Lipson says:

    Once again, ethical concerns aside, I still don’t think it means what you think it means. One cannot exploit a non-therapy. Once can exploit hope, one can exploit positive attention, one can exploit physical contact, but one cannot exploit artifact.

    Since all of these positive things are already a part of standard medical thinking, there’s no there there.

  6. daedalus2u says:

    The problem is that all of you are not thinking about the placebo effect or the placebo response as physiology but instead as “artifact”. If organisms exhibit a placebo effect, there must be physiology that supports that effect. That physiology is not artifact, it is physiology. The goal should not be to find out what inert treatments can invoke the placebo effect, rather what are the common physiological pathway(s) that placebos invoke and how can those physiological pathways be invoked pharmacologically.

    If you had a pharmacological agent that did maximally invoke the placebo effect and did nothing else, what properties would that treatment have? I think it would be a pretty good treatment all by itself.

  7. Arnold T Pants says:

    Hasn’t there been a study showing that the effectiveness of a placebo for pain can be reduced by giving an opioid receptor antagonist? Or something to that effect.

  8. ... says:

    Thank you for covering this. I had apprehensions when reading that article, but I couldn’t come up with a coherent rebuttal.

  9. splicer says:

    “a pharmacological agent that did maximally invoke the placebo effect”
    Which would be a drug that you would test in a good study. But against a placebo it would be no better than a placebo so it would fail in clinical testing.

  10. daedalus2u says:

    Just to add to my last comment. The goal of placebo research should be to find a pharmacological treatment that will maximally invoke the placebo effect. If that pharmacological treatment is found, then the optimum way to use it would be in addition to every actual effective treatment. Then people get the sum of the effective treatment and the maximally invoked placebo effect.

    If a treatment plus maximally invoked placebo effect doesn’t do better than the maximally invoked placebo effect all by itself, then you know the treatment is no good and is just a placebo. It the treatment plus the maximally invoked placebo effect is worse than the placebo effect alone then you know the treatment has adverse effects and is worse than placebo.

    I think that is where all the CAM treatments will end up, as being demonstrably worse than a maximally invoked placebo effect. If the placebo effect is maximally invoked, there is nothing that any CAM placebo can do. If CAM doesn’t make you feel better because your physiology already is cranking out the maximum placebo effect, why would anyone put up with the cost and BS of CAM?

    I appreciate that many of you cannot appreciate this because you are thinking of the placebo effect as an artifact caused by an intervention and not as a physiological effect caused by physiological pathways with myriad triggers.

  11. Prometheus says:

    I despise the term “placebo effect” because it is not an “effect” of the placebo. A placebo – by definition – has no effect when given to an unaware subject (e.g. if you slip it to them in their coffee).

    The so-called “placebo effect” is also not necessarily “poorly defined” – since we know most or all of the components – so much as it is “poorly differentiated”. The reasons people report “getting better” with placebo include (but are not limited to):

    [1] Natural progression/fluctuation of the condition
    [2] The sense of “doing something” (as opposed to inaction)
    [3] Wish to “please” the researcher (be a “good” patient)
    [4] Desire to get well
    [5] Interaction with the researchers/clinicians
    [6] Getting attention (not trying to get attention, but the positive feelings associated with being the center of attention)
    [7] Having someone willing to listen
    [8] Change in daily routine
    [9] Expectation of improvement
    [10] Distraction

    There are surely more that I’ve forgotten. The point being that the only thing not on the list is the placebo itself! If the placebo has an action, it would be more properly called a “drug”.

    Most of the reasons for the “placebo effect” are mental (the only one that isn’t is “natural progression/fluctuation of the condition”) and so can certainly be affected by opioid antagonists, anti-depressants and other pharmaceutical and non-pharmaceutical interventions.

    One thing that predictably decreases the “placebo effect” is to tell the subject that they’re receiving a placebo. Then, all you’re left with is “natural progression/fluctuation”.

    I prefer to call these “effects” the “control response”. This gets away from the false impression that the placebo is actually doing anything.

    In fact, the placebo is only a part of the control arm of a study. The point of the control arm is to “cancel out” all of the list above; to give a “baseline improvement rate” by having a group which is as much like the treatment group as possible, including taking a sham therapy similar enough to the real one that the subjects and researchers can’t tell them apart.

    There are actually many reasons why the “placebo effect” could be getting larger, including people having more trust that medications (or other therapeutic interventions) will help them. There are doubtless many other possible explanations. However, none of this means a thing, since the purpose of the placebo group is to provide a baseline to compare the treatment group against.

    Prometheus

  12. daedalus2u says:

    splicer, no because the placebos that are used in clinical trials do not maximally invoke the placebo effect. It is only very well done placebos that mimic the effectiveness of acupuncture. Crappy placebos don’t work as well as acupuncture, but very well done placebos with trick needles or even toothpicks do work as well.

    The acupuncture trials were not designed to study the placebo effect, they were designed to study acupuncture. Those running the trials don’t have the intellectual honesty to understand that the reason the placebo acupuncture they used worked as well as real acupuncture is because they are both placebos.

    It will take a very sophisticated analysis to demonstrate that a particular pharmacological intervention does invoke the placebo effect. I don’t think there is a sophisticated enough understanding of the physiology behind the placebo effect in the research community for appropriate research to be proposed, funded, done or published. It will be considered too “high risk” to be funded because “everyone knows” that placebos are inert and artifactual and can’t be relied upon so there is no use wasting research funding on it.

    I do have a blog post on the physiology behind the placebo effect. I have posted the link before, and the post is getting kind of old. It is still the best discussion of the physiology behind the placebo effect that I am aware of, and is consistent with everything that I have been able to read on the subject.

    http://daedalus2u.blogspot.com/2007/04/placebo-and-nocebo-effects.html

    Physiology has to work the way I describe in the post. We know that all organisms allocate resources over time and prioritize those resources according to what the organisms’ control system “thinks” is the most important task. Healing does not have as high a priority as staying alive, so physiology will prioritize staying alive over staying healthy. Staying alive and staying healthy are not the same thing. When a bear is chasing you, to be caught is to die a certain death. Any injury short of death is better than being caught by a bear. When physiology triggers “bear escape mode”, you can tolerate injuries that would be debilitating under other conditions. That extra burst of energy doesn’t come from some emergency new source. That extra energy comes from turning off all non-essential systems. Systems that are not needed while running from a bear, systems such as healing. If you are running from a bear, a few molecules of ATP spent running are “worth” more than a few molecules of ATP healing.

  13. daedalus2u says:

    Prometheus, you did leave out what I consider the most important part of the placebo effect. I have numbered it zero, because it is more important than all the others in your list.

    [0] The neurogenic control of physiology such that more resources are allocated to healing.

  14. daedalus2u says:

    Prometheus, your definition of “placebo effect” is insufficient because there are some treatments than cannot be given to unaware individuals; psychotherapy for example; unless you consider giving psychotherapy to an unconscious person to be fair test of whether psychotherapy is a placebo or not.

    I think a better definition is a treatment without pharmacological or physical effects, in other words drugs or surgery. Under my definition, psychotherapy is a placebo, a placebo that is effective and which is ethical to give.

  15. weing says:

    The price sticker on a bottle of wine is a placebo. The same wine with high price tastes better. People actually enjoy it more.

  16. Harriet Hall says:

    There are at least a couple of studies showing that a placebo response can be blocked by naloxone, a narcotic antagonist that nullifies the effect of exogenous or endogenous opioids.

  17. David Gorski says:

    Yeah, there was just one last month. I was thinking of blogging it.

  18. pmoran says:

    aesalus2: “Prometheus, you did leave out what I consider the most important part of the placebo effect. I have numbered it zero, because it is more important than all the others in your list.

    [0] The neurogenic control of physiology such that more resources are allocated to healing.”

    There is no direct evidence for this, is there? Generally speaking placebo usage has no effect on measurable facets of disease, pathology or injury. So how is it aiding “healing”?

    Medical interactions that may include a placebo do seem able to influence SYMPTOMS, and I suspect they can hasten the resolution of some types of illness. But the big deficiency in your hypothesis is that no reparative resources have ever been clearly shown to be affected.

    The problem may lie partly with the word “healing”. It has become a nebulous, even a metaphorical term, through widespread loose AM use. It has lost clear medical meaning, except for surgeons when they talk about wound healing (meaning “repair”).

  19. daedalus2u says:

    For the purpose of this discussion, I am happy to limit “healing” to wound healing and objective measures of closure.

    PM, are you denying that there is such a thing as neurogenic control of healing? Is there any evidence that there is the absence of neurogenic control of healing?

    When soothing words calm someone down and reduce blood pressure, is that effect not neurogenic? The details are unknown, but the mechanism was mediated through the ears into the CNS and the ANS.

    If we did a trial where individuals following surgery were exposed to 70 dB, 90dB and 110 dB noise would there be differences in healing rates? By what mechanism is the physiology of healing affected other than neurogenic? I think we have an intuition that the higher noise levels would result in slower healing. Trials like that would be unethical, but there was a trial that looked at the effect of construction noise on recuperation rates and found slower healing at higher noise levels.

    http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=937616

    How are the effects of noise mediated other than neurogenically? I admit that right now it is much easier to suggest neurogenic methods that will slow down healing, but that does demonstrate that to some extent healing is under neurogenic control.

  20. Diane Jacobs says:

    I tend to line up alongside daedalus2u in this discussion. The scientific definition and the medical definition of “placebo” certainly do overlap, perfectly even, from a medical point of view.

    However, there are a lot of us who are not prescribers of drugs or providers of surgery. Nor are we CAMsters. We deal with people who are “non”-medical – i.e., they do not have a pathology giving rise to their symptoms, who mostly, in fact, have pain issues. From our perspective, we teach, cajole, encourage, explain, educate, cheer on, handle, and otherwise interact with patients’ nervous systems. Placebo response is exactly what we are going for, defined as nothing more than helping an individual’s brain recover its own locus (or maybe even loci) of control.

    Development of a good neuroscience-based definition of placebo, one not conflated directly into the medical definition, therefore interests me a great deal. Any other explanation of the ability of people to get themselves better, with the help of some good accompaniment or without it, makes no sense, is anti-scientific.

    Doing work in this areas are Colloca and Benedetti among others.

    1. Benedetti results 2009

    2. Colloca results 2009

    3. Benedetti’s book 2008

    4. Neurobiological Mechanisms of the Placebo Effect 2005 open access

    It might be that the current medico-scientific definition of placebo is to some future neuroscientific definition of placebo as apples are to fruit – one kind but not every kind, not the only kind.

    Diane

  21. pmoran says:

    Daesalus2. “PM, are you denying that there is such a thing as neurogenic control of healing? ”

    The question is impossibly vague, considering the many different senses in which the word “healing” is employed, but let me say “Yes I do deny it”, so that you can point me to examples where such a process has been demonstrated and I can understand what you have in mind.

    I do agree with Diane that some illnesses have a major psychosomatic component, and the patient’s state of mind can be the main determinant of outcome, but I suggest that the mechanics of this are at a high cerebral level, at the level of human perceptions and preoccupations.

    At some low level they may involve neurotransmitter activity and even modification of neaural synaptic activity, but I would not say that this was “healing under neuroitransmitter control”, or ” under neurological control” in any medically important sense.

  22. pmoran says:

    Prometheus: “There are actually many reasons why the “placebo effect” could be getting larger, including people having more trust that medications (or other therapeutic interventions) will help them. ”

    That may be true, but it is also made clear in the article that the researchers had exceptionally high expectations of the new drugs. The subjects would have been exposed to much enthusiasm during recruitment and informed consent, offering the most likely reason for an unexpected doubling of the apparent “effect-size” of placebo in these studies..

    As you point out that doesn’t mean that all those reporting benefits actually did feel materially better. Some will be trying to give the “right” answer, bless ‘em, and others would have done so anyway.

    We need to keep things as simple as possible and I suggest that this abbreviated list encompasses all the main possibly active elements in placebo control groups and also in those who claim responses to impossible treatments..

    !. Spontaneous changes in symptoms

    2.True placebo responses (pending a better phrase covering all the non-specific influences within medical encounters — it’s everything but the placebo)

    3. Reporting biases.

    I like “control response” but it makes little sense outside of controlled trials.

  23. Diane Jacobs says:

    pmoran,

    “the patient’s state of mind can be the main determinant of outcome, but I suggest that the mechanics of this are at a high cerebral level, at the level of human perceptions and preoccupations.

    At some low level they may involve neurotransmitter activity and even modification of neaural synaptic activity, but I would not say that this was “healing under neuroitransmitter control”, or ” under neurological control” in any medically important sense.”

    With all due respect, are you suggesting that there is something about neural activity at “high cerebral levels” that is categorically different from any other “neurotransmitter activity” at any other level, that cognitive and consciousness functions do not involve neurotransmitters? Or brainstem nuclei?

    I agree that the term “healing” is something that should refer to actual tissues. What I think this is about is appropriately shaping neuroplasticity, which is ongoing anyway, for better or worse; I think it’s about helping one part of a brain (the awareness part, the cognitive parts, the parts that people use when they are awake) gain better control of the rest of the brain consciousness mechanisms, such as inappropriate arousal in insulae, S1 areas, body representations, etc.

    Diane

  24. pmoran says:

    “With all due respect, are you suggesting that there is something about neural activity at “high cerebral levels” that is categorically different from any other “neurotransmitter activity” at any other level — ”

    Well, yes. There are obviously high levels of organisation and complexity involved. Neuroplasticity itself implies a high order process.

  25. Ben Kavoussi says:

    Harriet, do you have the studies that show that a placebo response can be blocked by naloxone? I am very interested to read them, since the whole frenzy about acupuncture started when naloxone was said to block its purported effects…
    whatever they are.

  26. Hinode says:

    Diane Jacobs said: “We deal with people who are “non”-medical – i.e., they do not have a pathology giving rise to their symptoms, who mostly, in fact, have pain issues. From our perspective, we teach, cajole, encourage, explain, educate, cheer on, handle, and otherwise interact with patients’ nervous systems.”

    I wouldn’t say that this is a placebo you’re talking about. If their condition is non-medical then the cause is arguably psychological. Since it is known that encouraging and interacting with patients who have psychological problems helps, I would say this is part of the treatment and no placebo.

    pmoran: “Well, yes. There are obviously high levels of organisation and complexity involved. Neuroplasticity itself implies a high order process.”

    What do you mean by high order process? I’m no neurologist, but as far as I know neuroplasticity is everything BUT conscious.
    It’s based on sensory input. If you lose a leg, there won’t be any input coming from it. The part of the brain that used to control it will take on a similar function for a different body part. No consciousness involved.

  27. daedalus2u says:

    PM, the effect of noise on wound healing rate is a good example of what I consider to be the neurogenic control of healing. Noise triggers cells in the ear to transmit impulses to the brain which are interpreted as sound. That sound is processed and the brain responds by raising the arousal state, by releasing stress compounds like cortisol and these compounds reduce the rate of healing.

    Levels of cortisol are regulated neurogenically. In my way of thinking that makes everything that cortisol regulates also regulated neurogenically. There may be non-neurogenic control of cortisol in addition but differentiating between neurogenic and non-neurogenic control of cortisol may not be possible and neither can be ignored.

    I appreciate that methods that reduce the rate of healing are not of clinical value to health care professionals. I am thinking of it as an engineer, that if there is a control system that modulates an effect in one direction, maybe that same control system can be used to modulate the same effect in another direction.

    Psychological stress is fundamentally a neurological state. All the physical effects of stress are triggered by that neurological state; they are all regulated neurogenically. There may be non-neurogenic regulation in addition, but that is in addition and the neurogenic component can’t be ignored.

  28. Diane Jacobs says:

    Hinode,
    “I wouldn’t say that this is a placebo you’re talking about. If their condition is non-medical then the cause is arguably psychological. Since it is known that encouraging and interacting with patients who have psychological problems helps, I would say this is part of the treatment and no placebo.”

    With all due respect, I think you are unnecessarily dichotomizing, and re-dualizing, the brain. Since when are “psychological problems” not a subset of “brain” problems which involve neurotransmitters? Since when does awareness of smoother, self-obtained, cognitive, perceiving and bodily function (from a patient’s first person perspective) not amount to better brain function? Activation of brain reward pathways? Neurotransmitters and neuroplasticity adapting in more favorable (from the vantage point of the self-module living inside the brain) directions?

    I agree with you that neuroplasticity is everything BUT conscious. It can be affected, however, consciously, by the patient, given a good match between patient and helper, clear education and handling by the treater, and receptivity from the patient. I suspect that the pivotal “aha” moment involves placebo response of one sort or another, a patient noticing/feeling one or more reward pathways engaging in some novel fashion. A good practitioner will take no credit for any of that – it will have been the patient’s own brain, fixing itself.

    I’m not saying these are easy questions and I’m fully aware that beginning to unwind them can take us down lots of gnarly paths that end up at edges of philosophical chasms, that it would be easier to just stick with the dualistic approach; however, honest practitioners will realize that neuroscience is rapidly shoving us all right toward those very edges. We stand to gain nothing by not facing the fact that body/mind dualism is.. well, history. You may not have had a chance to see this yet – Sean Mackey on Pain Management, a video, about an hour 18 minutes. He shows people their own brains in real time fMRI and lets them work out how to “fix” them. Too expensive to use outside research as a biofeedback tool, but fascinating for the insights it provides all the rest of us who deal with people and their non-pathological pain issues, so that we can be more science-based.

    Diane

  29. daedalus2u says:

    What does saying a condition is “non-medical” mean? A better distinction would be to say a condition is “non-physiological”. I think that most of the commenters on this thread would agree that there are no non-physiological conditions. A non-physiological condition is one that does not exist, there being nothing related to organisms that does not involve their physiology.

    I completely agree with Diane. We can’t avoid abandoning the mind/body duality and I see no advantage in trying to avoid doing so by invoking artificial and arbitrary subdivisions of the body into brain, non-brain, higher levels, lower levels and myriad other non-well defined subdivisions which are primarily based on our ignorance of the coupling between them, rather than a knowledge that they are not coupled and constitute natural subdivisions with known and well characterized coupling characteristics.

    Our default should be that we don’t know if two physiological systems are coupled, not that every physiological system can be treated as independent of every other until otherwise proven. Independence of physiological subsystems may be a useful model (even as we know that it is wrong).

    Independence of physiological subsystems may be a useful treatment modality (even as we know it is wrong and even as we know and observe interfering side effects which derive from treating them as independent).

  30. jmm says:

    What about somebody with phantom pain in an amputated limb? Pmoran, would you call this condition “psychosomatic”? True placebo responses work via psychological effects. You cannot infer from this that the conditions that are psychologically treatable in this way have a psychological cause. Sometimes, the best treatment available for a non-psychological condition may be a psychological intervention, not because it addresses the root cause, but because it is simply the best we can currently do.

    You can’t just write off every disease for which the mechanistic cause is currently unknown as “psychosomatic”. Apart from being scientifically unjustiable (absence of evidence is not evidence of absence), it alienates patients, whose psychological response works best if they receive validation that their profound suffering is in fact real.

  31. Diane Jacobs says:

    daedalus2u,

    -> “What does saying a condition is “non-medical” mean?”

    I usually use it as a synonym for “non-pathological”, or “non-emergent”, i.e., requiring no drugs or surgery or other “medical” or medico-scientific intervention. In other words, adaptation is necessary (often requiring some interactive/”interactor” therapies) but not invasive treatment (operative/”operator” therapies).

    -> “A better distinction would be to say a condition is “non-physiological”. I think that most of the commenters on this thread would agree that there are no non-physiological conditions.”

    Agree.

    jmm,

    -> “You can’t just write off every disease for which the mechanistic cause is currently unknown as “psychosomatic”. Apart from being scientifically unjustiable (absence of evidence is not evidence of absence), it alienates patients, whose psychological response works best if they receive validation that their profound suffering is in fact real.”

    Wow. Couldn’t agree more.

    Diane

  32. Diane Jacobs says:

    I should add that I’m still not sure that “pain” should be framed as a “disease” – I think it’s OK to research its biomolecular underpinnings, role of glia, microglia, etc., take note of all the neuroscientific understanding available re: how normal brains work, so that when they don’t work very well, one can do some reasoned extrapolating. I also get that finding money to do research on a problem as common as persistent pain likely is helped by classifying it as a “disease.”

    However, I’m not sure that it would help actual patients with pain get on top of that all-important ‘locus of control’ issue. Instead it could make them feel more helpless, more out of control of their own brain, more reliant on medical intervention than they already imagine themselves to be.

    Again, I use the word “medical” to reflect “drugs&surgery.”

    Diane

  33. gdjsky01 says:

    While y’all are engaging in deciding if all meds should be scrapped and everyone given sugar pills, I don’t think that addresses *this* article and certainly not the wired one. I don’t think science based medicine will benefit from funding non-treatments (or is non-treatment a treatment? Oh I get it. :) )

    Yeah… take two sugar pills and call me in the morning. Prescription sugar pills! I think I just solved how to pay for National Health care… wait… what the price of sugar today???
    ;)

  34. Diane Jacobs says:

    You can go to pubmed and enter “placebo naloxone pain” – you will get (today) 266 articles, 13 of which are review.

  35. jmm says:

    gdjsky01on, lets go with your world for a minute. Go ahead, tell a patient that you can do nothing to help them, in fact you rather think their condition might get worse. Meantime, since you don’t know what’s wrong, it must be all in their head, so could they please vacate your office as quickly as possible, and stop bothering you with complaints about how sick they are.

  36. whamo says:

    “I’m addicted to placebo. I’d quit, but it wouldn’t matter.”

    –Steven Wright

  37. pmoran says:

    Hinode, to me: “What do you mean by high order process? I’m no neurologist, but as far as I know neuroplasticity is everything BUT conscious. It’s based on sensory input. ”

    And placebo and other beneficial non-specific responses to medical interactions aren’t?

    For neuroplasticity to occur there must be some overlaying controlling mechanism.

    Higher order? Look at what a computer can do and compare that with saying “look at what electrons can do!”.

  38. pmoran says:

    For those coming in late and misinterpreting my position, my comments referred specifically to Daesalus2 theory that placebos somehow directly mobilise “healing resources” which to him mainly means nitric oxide.

    My very point is that, in general, nothing is actually being healed by placebos. Very few, if any, pathological processes are affected thereby.

    It is all happening at the level of human perceptions and preoccupations. Only a very few “lower level” physiological effects can occur secondarily to that.

    Mind-body dualism is still a useful concept, when considering the differing functions and abilities of the two conceptual domains

    Science says that they both rely on the same substances and processes, but it is not yet clear to me how that helps us understand anything.

  39. Diane Jacobs says:

    I think some of the misunderstanding stems from the perennial problem in English of verbs and nouns impersonating one another at times.

    It seems to me that “placebo (noun) effect (noun)” is a concept. Its utility as a concept is that it can be accounted for and eliminated as non-specific, a backdrop to the real deal.

    It seems to me that “placebo (adjective) response (noun)” is more descriptive of a neurological “verb” or process occurring through time, probably as natural and non-conscious as water flowing down a hill, and subject to all sorts of different influences, much as neuroplasticity itself is.

  40. jmm says:

    pmoran, your position seems to be that effects at the level of “mind” have no meaningful effects on the “body”, and that medicine concerns itself only with the “body”. This is obviously a logically untenable position for medical conditions experienced largely at the level of mind, such as pain and depression. It is also a position incompatible with current scientific information, including the cortisol-stress type data already cited.

    BTW, a meditation teacher once described for me very nicely how meditation can help with pain. He explained that deep meditation states do nothing, at least in the short-term, for the injury, and the injury send its signal up to the brain just as before. That signal is processed just as normal, until it reaches “the part of the brain that is supposed to care”. By intervening at that high-level function, the patient becomes oblivious to the pain, which they no longer register or are troubled by.

  41. trrll says:

    One thing that predictably decreases the “placebo effect” is to tell the subject that they’re receiving a placebo. Then, all you’re left with is “natural progression/fluctuation”.

    This is a rather dogmatic statement, and hardly seems science-based. One hypothesized mechanism for the “placebo effect” is some kind of conditioned physiological response to medical procedures. Conditioned responses can be independent of conscious knowledge. You will salivate when you hear the dinner bell, even if you happen to known for a fact that dinner is not being served tonight.

    Here’s a personal anecdote: When I have blood drawn, I become pale, my skin becomes clammy, and I feel faint. This is a reproducible reaction, and dramatic enough that I’ve had medical personnel comment on it even when I didn’t mention it. On one occasion, a nurse “caught” it from me and had to leave the room because she was feeling faint herself. Now this is clearly not a physiological reaction to the slight pain of needle-stick; if I turn my head and don’t watch the blood going into the syringe, it doesn’t happen. Incidentally, it is specific to having blood drawn with a syringe: a bleeding wound won’t do it, and neither will an injection. So this is a kind of “nocebo” effect. Now I am perfectly well aware that having blood drawn can’t harm me. Indeed, having blood drawn does not consciously disturb me in the least. Nevertheless, I still have this reaction. It seems to me that if an adverse effect can be independent of my conscious knowledge, then this could also be true for a therapeutic effect.

    The point is that I don’t think that one can necessarily assume that all components of the placebo response aside from “natural progression/fluctuation” will vanish if the patient is aware that they’re receiving a placebo.

  42. daedalus2u says:

    PM, not quite, NO is the signal that triggers the myriad pathways that lead to healing. There are many thousands of pathways which must be regulated precisely in sync for healing to happen. There must be a common signal that regulates all the cells in a tissue compartment to initiate the healing process (which is complex and different in each tissue compartment). NO is a large part of that signal.

    An increase in NO is the signal to trigger healing largely because the signal to trigger stress pathways is low NO. Low NO is the signal for stress largely because cytochrome c oxidase must be disinhibited to maximize oxidative phosphorylation during stress to increase ATP production rate. The more severe the stress, the higher the need for ATP, the lower the NO level, and the harder “off” the healing pathways are turned to conserve ATP for more important things like running from a bear.

    The major healing resource is ATP. That is the major resource for doing everything. That is not something that a reserve can be built up in, it is consumed as fast as it is made. The ATP level is regulated by the NO level. They both go up and down in sync. Low ATP is the signal to turn off non-essential systems such as healing. That is how low NO turns off healing, through the ATP level. All the different degenerative diseases are characterized by low ATP. My hypothesis is that they are all good regulation around a bad setpoint. Good regulation of a low ATP level caused by a low NO level (the bad setpoint).

    This low ATP is very easy to see in the neurodegenerative diseases (via magnetic resonance spectroscopy), where all of them are characterized by neuroinflammation which lowers NO levels by producing oxidative stress, which also is the trigger for ischemic preconditioning which is protective provided it doesn’t go on for too long. The NO level not coming back up is what makes ischemic preconditioning go on for too long and leading to long term degeneration.

    Stress lowers the NO level which lowers the ATP level which turns off healing to preserve ATP for tasks such as running from a bear.

    NO is the signal, ATP is both the signal and the substrate for healing (and nearly everything else that physiology does).

    We know there is a mechanism that slows down healing under conditions of high stress. That fact is not in doubt. We know that healing is a very complex process and control of that complex process must also be a complex process.

    We know that the healing rate is retarded under multiple conditions, stress, starvation, hypoxia, diabetes, ischemia, infection, inflammation. Healing still occurs under those conditions, just more slowly. Healing under those conditions occurs at many different rates. The same healing pathways proceed at different rates. Is there a different control paradigm to control healing at a different rate for each different condition in each different tissue compartment at each different rate? Very likely not. Each pathway in the healing process has to occur in concert with all other pathways. Presumably there is a common control parameter that regulates them all “in sync” in every tissue compartment. I suggest that ATP is that common control parameter. ATP is an anion, and cell membranes are impermeable to it, so there must be a diffusible signal that regulates the ATP levels “in sync” between the multiple cells in the tissue compartment that are going through the healing process “in sync”.

    We know there are neurogenic conditions that can slow healing such as stress. If a neurogenic stressor can slow healing, then a neurogenic anti-stressor can remove that stress caused retardation and accelerate healing. That neurogenic anti-stressor can only be called a placebo because it has no pharmacologic or physical component.

  43. digaman says:

    I’m the author of the Wired article, and it’s been gratifying to see so many of the points that I made in the article repeated here. I also addressed Peter’s challenges in some depth on the other copy of this blog:

    http://scienceblogs.com/whitecoatunderground/2009/09/placebo_is_not_what_you_think.php

    Peter is obviously free to have as much contempt for my reporting and writing as he wants, but I would urge anyone who has participated in this conversation to *read the article itself* before passing judgment on it.

    Thanks!

  44. Ben Kavoussi says:

    Dear Peter,

    Great post. WNYC Radio’s series called Radiolab has also an interesting program on the placebo effect. There is discussion about a WWII medic who saw its powerful effect on pain during the Battle of Anzio. Here’s the URL:

    http://www.wnyc.org/shows/radiolab/episodes/2007/05/18

  45. digaman says:

    Yep, I talk about Henry Beecher, that medic, in the Wired article. Besides promoting the adoption of the placebo-controlled randomized clinical trial as the gold standard of medicine, Beecher also advocated the concept of informed consent in research, and helped create the guidelines for “brain death” that are still in force today. (His original placebo paper was also flawed, which I talk about in the article.) Fascinating guy.

  46. pmoran says:

    “pmoran, your position seems to be that effects at the level of “mind” have no meaningful effects on the “body”, and that medicine concerns itself only with the “body”. This is obviously a logically untenable position for medical conditions experienced largely at the level of mind, such as pain and depression. It is also a position incompatible with current scientific information, including the cortisol-stress type data already cited.”

    Did I say that? I generally try to avoid generalisations, because medicine deals with an incredibly varied range of conditions. What I mainly feel strongly about is that there are serious limitations to the effect of the mind on the body.

    That remains consistent with placebos and other psychological influences having a significant effects on symptoms and even helping to resolve some kinds of illness through action at a cerebral level.

    “BTW, a meditation teacher once described for me very nicely how meditation can help with pain. He explained that deep meditation states do nothing, at least in the short-term, for the injury, and the injury send its signal up to the brain just as before. That signal is processed just as normal, until it reaches “the part of the brain that is supposed to care”. By intervening at that high-level function, the patient becomes oblivious to the pain, which they no longer register or are troubled by.”

    I have no problem with some such processes.

  47. Peter Lipson says:

    As has been pointed out, “the part of the brain that is supposed to care” is rather irrelevant to all of the most common human afflictions, such as diabetes, heart disease, hypertension, and cancer. While people deal with illness differently, one cannot lower A1C, prevent progression of atherosclerotic plaques, or shrink a tumor with the mind.

    One may—may—be able to slightly lower blood pressure “with the mind”, but for anyone with significant hypertension (ie, real patients that i deal with every day) it’s not particularly relevant.

  48. digaman says:

    > the most common human afflictions

    Antidepressants are among the most commonly prescribed class of drugs in the US.

    http://www.msnbc.msn.com/id/32274077/ns/health-mental_health/

    The placebo response in depression, anxiety, and other “mood disorders” is considerable; just ask any drug developer who has tried to develop a new class of antidepressants in the last 15 years. Pain is another common human affliction that is subject to significant placebo response.

    As pharmaceutical companies have discovered to their great expense, the biological mechanisms and social dynamics of the placebo response are worth comprehending.

  49. daedalus2u says:

    With all due respect to the expertise of a “meditation teacher”, there is no data demonstrating the attributed mechanism of simple blockage on pain, and no experimental method by which a “meditation teacher” could determine if such a hypothesis was correct.

    With all due respect to Peter Lipson, “all of the most common human afflictions, such as diabetes, heart disease, hypertension, and cancer” are all made worse by stress (leaving out cancer for the moment (we can’t expect the placebo effect to be magic and do everything ;)). If a condition is made worse by “stress”, then there is neurogenic coupling to the physiologic pathways that regulate that condition. We may not know the details of how that coupling happens, but that there is neurogenic coupling is not in doubt.

    The brain is a large metabolic load. The brain is run on glucose. Glucose is made in the liver. There has to be coupling between the glucose demand of the brain and the capacity of the liver to supply glucose to that brain. The glucose demand of the brain has to regulate the glucose production capacity of the liver. How does that happen? We don’t yet know the details, but we know it must involve certain physiological pathways. Making more glucose in the liver requires liver growth (which depends on NO). Making more glucose in the liver requires mitochondria biogenesis (which depends on NO). Making more glucose in the liver requires angiogenesis (which requires NO).

    Neural activity produces NO, the vasodilation that is measured by fMRI BOLD technique is mediated through NO. That vasodilation correlates so well with neural activity that it is used as a surrogate marker for “functional connectivity”. Many people working in the field don’t appreciate that the fMRI BOLD signal is not a direct measure of neural activity.

    When organs are transplanted all nerves going to and from them are severed and not reconnected. Livers, hearts, kidneys, and other organs still work pretty well without a neural connection to the body they are working in. However many patients who receive a transplanted organ have that organ fail via the same mechanism as their original organ failed. Since most of those failure mechanisms are exacerbated by stress, which is fundamentally a neurogenic state, but because there are no nerves connected to the failing organ, the coupling between the nervous system and the failing organ involves non-neural coupling.

    The NO signaling required for the liver to increase liver glucogenesis capacity is local, that is the NO signaling occurs inside the liver. Since that signaling is not solely neuronally mediated (because patients who receive half a liver experience it growing larger), at least some of that NO signaling occurs via compounds carried in the blood. What might those compounds be? Some compounds that inhibit nitric oxide synthase are known to be elevated in conditions exacerbated by stress (asymmetric dimethyl arginine). NO availability is known to be decreased under conditions exacerbated by stress (hypertension, obesity, diabetes). Stress is known to decrease NO availability. Meditation and the relaxation response is known to increase NO availability and to decrease symptoms of the metabolic syndrome and increase insulin sensitivity.

    http://archinte.ama-assn.org/cgi/content/full/166/11/1218

  50. Peter Lipson says:

    daedalus, i think we need to be careful with the “made worse by stress meme”

    It depends on what condition, what “made worse” means, and what “stress” means. Most studies of “stress” and disease are not that impressive as far as the most important outcomes go.

  51. daedalus2u says:

    Sorry to be so snarky, but when you say “be careful” do you mean “ignore”?

    Here is a nice review paper on psychological stress and cardiovascular disease.

    http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=18371552

    The opening sentence states ” There is an enormous amount of literature on psychological stress and cardiovascular disease.”

    The effects of stress on health are not small, are not hard to see or measure and are all in one direction. We are not interested in what “most studies” say, only what the most reliable studies say. The most reliable studies say that stress is important.

    If you have a link to a review that says that stress is not important I would like to see it so I can evaluate how good it is.

  52. Peter Lipson says:

    interesting…thanks for the link, d

  53. sstumpf says:

    I am re-posting excerpts of a comment from another thread on Kavoussi’s article “The Golden Tooth”. I describe a few points made on placebo as it relates to CAM (Snake Oil Science, Bausell, 2007) that hopefully will contribute to this discussion about the placebo effect. Many of these points have been touched on already. Personally, I think the point made

    Placebo is translated from Latin “to please”. An early 19th century definition was “given more to please than to benefit the patient”. Seminal research was published in JAMA in 1955 by Beecher who showed 35% of patients that received a placebo across 15 clinical trials responded positively. Beecher suggested that if any “non-harmful” therapy used in a clinical trial has the potential to produce this proportion of positive results then how can any trial be conducted without a placebo control group?

    Bausell defines the placebo effect as “any genuine psychological or physiological response to an inert or irrelevant substance or procedure”. Critical to the placebo effect is that the patient and the provider believe they are administering and receiving an effective treatment.

    Bausell goes on to suggest that placebo and many CAM therapies are the same thing (his book focuses on CAM research). He writes about “natural impediments to making valid inferences” which is a chapter worth reading. He begins with the fallacy of making causal inferences based on associated findings (Research 101).

    By the way, I just commented on another Kavoussi article titled “Orientalism” about NO research in acupuncture. I have insufficient expertise to comprehend the outcomes and am hoping others will explain. My purpose was otherwise.

    Here is a URL link to the placebo effect of wine prices.

    http://www.reuters.com/article/newsOne/idUSN1443681520080114

  54. digaman says:

    > Seminal research was published in JAMA in 1955 by Beecher who showed 35% of patients that received a placebo across 15 clinical trials responded positively

    I talk about that paper in my Wired article (http://www.tinyurl.com/fauxdrug). The 35% figure, which has been widely reported, is no longer considered trustworthy by placebo researchers, who point out that regression to the mean and natural history could have accounted for an apparent placebo response in many of the studies that Beecher cited. Beecher’s paper was very influential in advocating the use of placebo controls in RCTs, but as science, it was flawed. Still worth reading. It’s too bad that there’s no open-access source online (with the appropriate caveats), considering its influence.

  55. digaman says:

    Sorry, the close parenthesis seems to have undone that link. Here:

    http://www.wired.com/medtech/drugs/magazine/17-09/ff_placebo_effect

  56. Diane Jacobs says:

    Here is a 2003 article, Placebo Analgesia and the Heart.
    I found it on this list of publications that Benedetti has been involved in.

  57. jmm says:

    Peter, this is moving a little off the narrow question of the placebo and onto more general mind-body questions, but treatment of afflictions such as diabetes, heart disease and hypertension needs also to take into account the behaviors (diet and exercise) that are primary contributors. And there is NO WAY you are going to affect those behaviors without a sophisticated understanding of the strange workings of the mind. For this, the mind-body connection is of course important, eg in terms of conditioning mental responses to problem situations and behaviors.

    I’ll give you cancer as one common disease for which it can be effective (if not always humane) to treat the body only, with relatively little consideration for the mind. As for the rest, treatment needs to consider both, and their interaction, with dualism a pretty useless model.

  58. botogol says:

    I think the thrust of the article, perhaps not well expressed, was that

    changes that occur simply from being part of the study [...] Such effects include paying better attention to one’s health and habits, trying to please the doctors, and other less clear and tangible effects.

    are getting stronger.

    which is possible, and would be interesting as it might betray a growing faith in medicine in general. Taking stuff will make me better.

  59. jmm says:

    daedalus2us, as to your point that there is “no experimental method by which a “meditation teacher” could determine if such a hypothesis was correct “, the essential epistemological problem here is in fact also found in all studies of subjective experiences such as pain. You can find correlates, whether chemical or via fMRI, that are suitable for objective analysis. But what do they correlate to? The amount of pain that a subject reports that they are subjectively feeling. Ultimately, one way or another, subjective experiences have to come in to the scientific method if they are in fact the object of study, even if only as the basis justifying the subsequent use of correlates.

    The quote I gave, to do justice to its source, was primarily intended as an eloquent expression of the subjective experience of the reduced debilitation due to pain felt by an experienced meditator. Its epistemological strength comes from the fact that most or all such individuals would concur that this description is a good match to their subjective experience (ie reproducibility).

  60. Tim Kreider says:

    I thought the Wired article was quite good. I agree that the language about “placebos getting stronger” might be a bit misleading, but the second and third pages were much clearer about the different factors that confound measurement of specific effects. We probably need a different language than “placebo effect”. The simpliest change, which Benedetti uses as the title of his book, is “placebo effects” (plural).

    I agree with the Wired author that we should study the therapeutic ritual in order to better exploit it. Surely this is what the acupuncturists et al are great at!

  61. digaman says:

    I also like Benedetti’s phrase “placebo response” to make the distinction between statistical noise in trials and the physiological effect, which is why I tried to use that phrase as much as possible in the latter half of the article.

  62. daedalus2u says:

    jmm, fair enough, understood as metaphor it might be ok. My focus is more on the actual physiology of placebos and meditation, so explanations that can’t be tied to actual physiology or actual data are not very useful to me. That is the problem with most of CAM, they have data but then they attach all this crazy non-physiological explanation to it, as in the qi of acupuncture and the life energy of reiki. Those explanations are clearly wrong.

    Regarding the placebos getting stronger, I remembered one of Beecher’s original observations that people injured in wartime needed less morphine for pain relief that did civilians injured in civilian accidents. He says it quite explicitly in this paper, that placebos are more effective at relieving pain under conditions of high stress.

    http://www.ncbi.nlm.nih.gov/pubmed/13798214

    If there is a decline in the effectiveness of placebos it may be due to a higher stress levels in present times.

  63. jmm says:

    daedalus2us, I generally agree with you re crazy non-physiological explanations, only with two buts. First, data that something worked would be sufficient reason to start implementing it: an adequate physiological explanation may follow in due course, but is not a prerequisite. Second, science too uses metaphors or models, some of which seem pretty crazy and impossible when you stand back. Quantum mechanics springs to mind here, with many physicists in the area having problems with it to. But they continue to use it not because it seems mechanistically correct, but because it works in terms of predictive power. So IF a model of qi or prana worked in terms of predictive power, then there would be no reason not to use it.

    Meantime, if you want to explore some of these alternative models seriously, it helps to come up with some kind of functional translation into physiological terms. Eg prana is said to have two main forms, prana and apana. The prana form can be translated as the set of neurophysiological patterns associated with inhaling. Apana is then the set of neurophysiological patterns associated with exhaling. Manipulation of prana is then a more sophisticated extension of instructions like “take a slow deep breath out”. In this light, claims eg about changes to blood pressure and heart rate, start to become plausible under a physiological model too.

  64. Peter Lipson says:

    Tim, I agree with you that the the Wired piece got into some interesting detail. I supposed I could have saved a lot of words had I simply written, “Headline and entire first third of the piece seems to show fundamental misunderstanding of the basic concepts…but it gets better.”

  65. digaman says:

    Peter, just FYI, the headline of the piece in the magazine was “The Placebo Problem.” I had no power over the headline on the Web, and in fact, didn’t even see it until the article was online. Journalists rarely have control over headlines — they’re in the hands of attention-seeking editors, and frankly, often strike me as over the top. They tend to obliterate a lot of careful considerations, and particularly when the subject is as complex and nuanced as this, can do damage to the article itself. Anyway, this has been a fascinating conversation. You have wonderful readers here, Peter.

  66. pmoran says:

    From the Wired piece: “It’s not that the old meds are getting weaker, drug developers say. It’s as if the placebo effect is somehow getting stronger.”

    Who said this, digaman? I would expect serious clinical investigators to be well aware of all the possible internal reasons for a high apparent placebo responsivenss in clinical trials. These would have to be ruled out before considering the much less likely explanation of a trend towards increased placebo responsiveness to druhgs.

    Examples?: choosing to test drugs on more minor illnesses that are more likely to improve spontaneously, allowing researcher enthusiams to communicate themselves to the subjects which would enhance any true placebo responsiveness as well as patient resporting biases (not wanting to disappoint etc).

    It is good that you allow for some these complexities later, but our complaints about medical journalism seem justified by this example of somewhat off-target hype over substance.

  67. daedalus2u says:

    “all models are wrong, but some are useful” George E. P. Box

    I disagree that science uses some impossible models. There are many people who can’t understand the models that science uses, that does not make them impossible. Science is hard for most to understand because our brains mostly did not evolve to understand science. Mostly our brains evolved to understand other humans, to be able to communicate with them. That is why so many people think in terms of anthropomorphic models. Their brains limit their thinking to only in terms of anthropomorphic models. Anthropomorphic models really suck for thinking about science, reality and everything that is non-human. Most people are stuck thinking about reality in anthropomorphic terms because they don’t have the capacity to think in other terms.

    There is so much interpretation associated with terms such as prana and apana that the data is impossible to differentiate from the interpretation. Most descriptions only mention the interpretation and the data from which the interpretation was derived is not mentioned. If a model can’t be mapped onto a subset of reality, then it isn’t of much use. Quantum mechanics is unambiguous. The metaphors of prana and apana are either so ambiguous as to be uninterpretable or they are in fact wrong. The data is hopelessly contaminated by the bias of the interpretation driven by the hypothesis.

    The originators of the prana hypothesis didn’t have the tools to acquire the data necessary to understand chemistry and physiology. Many of the premises they used to generate the prana hypothesis are wrong. There is no “vital energy”, the heart is not the center of the body, and is not the source of that vital energy. When they practice different breathing techniques, they are not manipulating a vital energy, they are affecting their physiology.

    To me, the prana hypothesis looks like Cargo Cult Science. There isn’t much to be learned about reality by trying to learn from the followers of a Cargo Cult. Their model of reality is so distorted by their prior expectations and their wishful thinking that they are unable to even ask the right questions. If you are creative, you can generate a metaphoric understanding of a Cargo Cult and map that Cargo Cult metaphor onto a metaphor of a western capitalist economy. There are many such metaphoric mappings. But those multiple metaphoric understandings don’t help in understanding the reality of a western economy in terms of ancestor worship.

    It is a postmodern metaphoric fantasy that any and all metaphoric understandings are as good as any other. All metaphors are wrong, so in that sense the postmodern understanding is correct. All metaphors are not useful. I don’t know how to use a Cargo Cult to learn about supply of western goods. I do know how to use a Cargo Cult to learn about the psychodynamics of societies, about division of labor, about allocation of resources between individuals and over time, and especially about how people come to “know” things, about belief, delusion, truth, and skepticism.

    I don’t know how to learn about physiology from the prana hypothesis. I think that there may be a lot to learn about psychodynamics of societies, but most of the real “data” of that is hidden. The followers of the prana hypothesis think they are doing something “real”, just like the Cargo Cult followers. They would not take kindly to understanding that essentially the only scientific value in the prana hypothesis is as an example of how people can delude themselves for so long.

  68. jmm says:

    daedalus2us, the point of looking seriously at prana is not to learn something about physiology, or anything else about the physical world. It’s to learn whether the set of practical techniques (pranayama) associated with the prana metaphor are useful for manipulating physiology in a positive way. A basic understanding of the prana metaphor is essential just to get a grasp of what techniques are out there, and what each might be expected to achieve according to the prana model. There is no question about the ability of advanced pranayama practitioners to attain extreme control of many aspects of their physiology: that is well documented. The question for SBM would be which techniques can be co-opted for untrained patients with which conditions. It is a practical, prediction-driven concern.

    Actually, even traditionally prana is not considered part of the material body. It is part of the “subtle body”, to be taken less literally.

    As for the usefulness of a “wrong” model, consider ether. Maxwell used this wrong model to derive his equations. In the end, the equations that came out were no longer dependent on the ether, and it was eventually dropped as superfluous.

    Another good example is organic chemistry, which uses lots of wrong models, eg molecular orbital theory, valence bonds etc, all of which have been disproved by quantum mechanics. They remain useful though.

  69. qetzal says:

    jmm:

    There is no question about the ability of advanced pranayama practitioners to attain extreme control of many aspects of their physiology: that is well documented.

    I’ve heard claims like that before, but not seen the documentation. Can you provide some?

  70. jmm says:

    qetzal, I’m about to leave for a trip and won’t have time for sleuthing up a lot of sources, but check out

    http://yoga84291.yuku.com/topic/2633

    and links from there about Krishnamacharya (source of a good proportion of modern yoga practices) and complete heart stopping. Whether stoppage was complete or not, there were clearly some pretty extreme physiological changes. The two papers cited would probably make a good starting point for a citation search to turn up anything more recent.

    BTW, apparently Krishnamacharya’s son asked to learn this heart stoppage technique, and was told by Krishnamacharya essentially that it wasn’t “real” yoga, but more of a party trick, necessary at the time to get people to pay more attention to hatha yoga as a powerful practice. Since it didn’t further any spiritual purpose, he did not teach it.

  71. Harriet Hall says:

    A good party trick is to hide a ball in your armpit and ask someone to monitor your pulse at the wrist while you squeeze the ball and obstruct the artery. It has fooled a lot of people, including doctors.

  72. Diane Jacobs says:

    I think it’s important to keep the memes associated with (non-medical) therapies, and passed on with them, up to date. Maintaining pre-, anti-, a-, and pseudo-scientific memes and concepts alive, by bothering to learn them in the first place, or inventing new ones, then passing them on, does nothing to help either ourselves as practitioners or our patients to understand actual physiology or how brains/nervous systems take care of body business. Learning how the brain works is more interesting and more appropriate, IMO, than trying to explain something as confusing (and kinda exotically religious-sounding, frankly) as “prana” to somebody. With each patient we treat we could instead help neuroscience become better understood. Treatment is an opportunity to teach something real, pass on current science-based information in a way that’s personally meaningful to each individual. Help the world become more scientifically literate one person at a time.

  73. daedalus2u says:

    I completely agree with Diane. If you want to understand reality, you have to start with facts and build up from there. Starting with wrong metaphors won’t get you to a correct understanding. Wrong metaphors impede getting to a correct understanding because they trigger pareidolia, a type 1 error of seeing something that isn’t there. For the most part it isn’t ignorance that impedes understanding it is believing something that isn’t true.

    Trying to understand reality using concepts such as qi and prana as the basis will not work. There is not a shred of evidence for such things. There never was a shred of evidence. The ancient investigators who adopted the hypothesis of qi and prana did so with no evidence. If there was evidence, that evidence would still exist and we would have access to it in modern times. Why should we try to understand reality in terms of concepts for which there is no evidence?

    jmm, I don’t disagree that experienced practitioners of mental disciplines can produce changes in their physiology that appear remarkable. That they can do so doesn’t make their explanations of how they are doing it using non-physiological concepts such as qi and prana correct. Their abilities do not provide support for concepts such as qi and prana. Their abilities do provide support for coupling between the CNS and the ANS and some degree of volitional control over the ANS.

    I remember reading a paper where rats were connected to a biofeedback system that stimulated their reward center, and with that setup, the researchers could train the rats to stop their hearts and die.

    We don’t see practitioners of yoga living to be 150 years old. If they did have access to voluntary control of an actual “life force”, they should be able to extend their lives indefinitely. They don’t extend their lives indefinitely. I appreciate that they have all kinds of rationalizations as to why they don’t extend their lives indefinitely.

    The paper on individuals “stopping” their heart was interesting. Everything they did was completely consistent with what is understood about physiology with no hint that something like prana is needed to explain it.

  74. gregladen says:

    Peter: “Once again, ethical concerns aside, I still don’t think it means what you think it means. One cannot exploit a non-therapy. Once can exploit hope, one can exploit positive attention, one can exploit physical contact, but one cannot exploit artifact.”

    That is the biggest problem with this conception, in my view, and why the title of this post is so apt. The outcome of using a placebo or any control is assumed to be a baseline. If carrying out the act (‘treatment’ both in broad and in medical terms) causes an effect other than the chemical or physical effect of the actual treatment itself … like, somehow the act of going to the clinic to get the treatment has a positive effect on the patient or whatever … then that is subsumed under the overall effect seen in the control sample, and that is something that could be (should be?) isolated and bottled, as it were. But most of the effect that does occur is part of the Null Model and not part of the treatment. It is what would happen anyway. It is not something you can do, bottle, isolate, repeat, exploit, or make use of.

    My take on the placebo effect.

  75. Simonw says:

    I was going to post what Gregladen said.

    Placebo effect occurs in the treatment group in most cases, that is what control means. So unless the placebo is so beneficial as to leave no room for a treatment effect, then even if the placebo effect increased it would have little effect on the trial results. I suspect what the article is showing is that many common drug treatments for depression are not very effective, but the older trial data may not have been as solid as it should have been.

    Which is what is wrong with the title of the article. The article itself was interesting.

    However I disagree that placebo can’t be bottled. The complementary medicine practitioners have been bottling it for years.

    As demonstrated in acupuncture trials the placebo effect varies with the type of placebo. It is certainly believed (I don’t know what the evidence base is) by those in the field that placebo surgery is more effective than other “sham” treatments. I guess this implies that some part of the placebo effect is based on expectation (which isn’t controversial as far as I know). Similarly there is some evidence more expensive placebos work better.

    Not all placebo effects are positive. And it wouldn’t surprise me if overcrowded, uncomfortable waiting rooms, with doctors running late and poor decor have a negative effect. Where as a nice cup of tea and a oat cookie before your one of one homeopathic consultation would make you feel a lot better. The tea and the cookie will of course have active pharmacological agents in if not the treatment.

  76. jmm says:

    Diane, daedalus2us, I think you should try reading some serious texts in the history and philosophy of science. To echo the phrasing of this post, science is not what you think it is.

    Science is defined by things such as predictive fertility and consistency. It so happens that good mechanistic models supply such things. This is the reason that we as scientists strive for them. But in general, observations often proceed models. When an observation does not fit our current model, that indicates where the model needs work. Under such circumstances, we should NOT neglect the anomolous observation because we don’t understand how it could be so.

    So the question then becomes whether a metaphor like prana have predictive fertility. From my own experience, the answer is definitely yes. I can use the metaphor to customize a practice for the way that I need to manipulate my mental and physiological state on a particular day. As a scientist, I would prefer a physiological metaphor. Hence my translations above, eg “set of neurophysiological patterns associated with an inhale”. Awkward, though, you have to admit. I think the key word is “set”. Modern physiological metaphors are highly reductive, beyond what the evidence supports. Eg, we do not move a muscle in isolation, every movement is a complex operation. Yogic metaphors tends to describe things that physiological ones would consider as a complex suite.

    As for “wrong” metaphors, I have already mentioned ones like orbital theory in chemistry. Or mind-body duality for that matter. Science, past and present, is stuffed full of them. Clearly their absence is not a prerequisite for good science.

  77. digaman says:

    > It is not something you can do, bottle, isolate, repeat, exploit, or make use of.

    As I feel like I’ve said several times by now in simultaneous versions of this thread, researchers like Fabrizio Benedetti and Tor Wager are not, primarily, running clinical trials for drug companies (as Arif Khan in my article does). They and others are studying the physiological mechanisms of response to the act of taking placebos, and ways of eliciting that response for purposes other than testing active drugs against it. Just because the more common context of the word placebo is in reference to the control groups of clinical trials doesn’t mean that the word cannot be used outside of clinical trials or that their research doesn’t exist. Long before clinical trials were invented, doctors were prescribing placebos to their patients, and using that word to describe them.

  78. bruabra says:

    Just to add to the conversation…with a hello from Brazil! To discard the placebo effects as artifacts is to miss the boat completely, not to mention a total lack of awareness of the science (you guys love this word, don’t you :)) being produced on the subject.

    Definitely, Benedetti’s book: “Palcebo effects, understanding the mechanisms in health and disease” is the best single source on he matter. I’ll bring up the topic of Parkinson’s Disease, since it’s one of the best models to study the placebo effects and physiology.

    We know that substantial improvement in Parkinsonian symptoms are seen in the placebo group of many clinical trials, even when compared with a no-treatment group (that would eliminate spontaneous remission and regression to the mean).

    It was later demonstrated that expectations do modulate motor performance, in both directions (placebo and nocebo). This is true for almost any condition, and PD is no different. This explains part of the placebo effect, but not all. Let’s go to the more interesting stuff.

    We all know that PD’s pathophysiological substrate involves the disruption of dopamine function in the neural pathways from the substantia nigra pars compacta to the striatum (putamen and caudate nucleus). In later stages, non-striatal pathways also play a role. Well, it has been showed by de la Fuente-Fernandez et al. that the administration of placebo induces dopamine release in the striatum, which, it turns out, not only is associated with motor performance, but is also involved with motivation and reward expectation.

    Therefore, if the placebo releases dopamine in an area where there is dopamine deficiency, it works just like most of the drugs we use for PD… So…Can we still call it placebo?

    Benedetti has also shown how single neurons in the subthalamic nucleus (the target for deep brain stimulation procedures in PD) also respond to the placebo administration, shifting from a pattern of bursting activity to one of non-bursting discharge, that impressively correlated with patient’s clinical responses (people who didn’t change the pattern also didn’t improve…)

    Again, if an inert pill is changing neurons firing patterns, can we still call it placebo?

Comments are closed.