Jan 23 2009

Some Follow Up On Migraines

Published by under Uncategorized
Comments: 117

Yesterday I wrote about a new Cochrane review of acupuncture and migraines. The most significant result of the review was that published studies show no difference between “real” acupuncture and “sham” acupuncture. This is the part of trials that can be blinded – well, single-blinded at least (the subject does not know which is real vs sham, but the acupuncturist does). This blinded data is negative, indicating that it does not matter where you stick the needles, which is a strong indication that the underlying philosophy of acupuncture is false. Further, now that acupuncture studies are being done with placebo acupuncture – opaque sheaths and dull needles that do not penetrate the skin, allowing for double-blind studies – these are also coming out negative.

Therefore, I think it is reasonable to conclude from the literature that it does not matter where you stick acupuncture needles, or even if you stick them. Therefore, any observed benefit must be either: 1) an artifact of the study design (i.e. placebo effects); or 2) a result of other things that are happening in association with the acupuncture treatment.

Several points were raised in the comments I want to address in some detail. Further, there are many questions about migraines, and since migraines make up a significant portion of my clinical practice I can clear up some of the confusion as well.


First some background on migraines.   Migraines are a specific kind of primary headache, meaning that the headache is the disorder and is not secondary to an underlying problem, like a brain tumor. Migraine are variable in their presentation, and so are defined as having some of a list of symptoms. These include unilateral (one-sided), throbbing, can be triggered by various things, and associated with nausea, vomiting, and sensitivity to light, sound, and strong odors. Migraines can also occur with an aura, which is a neurological symptom occuring before or during the headache, most commonly visual changes like dark spots (scotoma) or flashing lights (visual fortifications).

Although there is no objective laboratory diagnostic test for migraines – the standard headache workup will be negative – the diagnosis can be reliably made in many cases based upon a very typical clinical presentation. Of course, the edges are blurry, and some patients overlap with tension or other types of headache. Despite the lack of an objective test, there are documented physiological changes that occur during migraines. There is a vasocontrictive phase which is associated with  a wave of decreased electrical activity (called spreading depression) across the cortex. This is followed by a vasodilation phase. It was originally thought that these vascular changes were what caused the migraine, but now it seems as if this is just one consequence of the deeper physiological mechanisms of migraine.

The current synthesis is that a migraine is an attack, not unlike a seizure, in the trigeminovascular reflex which is mediated in the brainstem. The trigeminal nucleus, which mediates sensation to the face and part of the scalp, becomes hypersensitive, causing a cascade of effects. There is still much we do not understand about migraine physiology, but I am just giving a superficial overview of what we do know.

One more interesting bit – there is a genetic condition known as familial hemiplegic migraine which results in frequent severe migraines associated with stroke-like neurological deficits. The mutation is in a gene for a calcium channel, which is evidence that calcium regulation may play a role in migraine in some patients.

Another aspect of migraine that is important to understand is that it is probably not one discrete entity but a collection of physiological entities that all lead to a final common pathway of trigeminovascular activation. Therefore, different patients with migraine may have significant physiological differences in terms of their triggers and which treatments they are likely to respond to.

The good news is that migraines are very treatable. There are a number of strategies which have to be individualized to the patient. The first consideration is migraine frequency. If a patient is having one migraine every 3-4 months than preventive treatments are not likely to be worth it. You can focus on acute treatments (called abortive therapy, because they are aimed at aborting the migraine in progress).  If, however, someone is having 3-5 migraines per week then abortive therapy has to be minimized and the focus should be on preventive therapy.

Prevention takes several forms. The first preventive therapy to try is always removal of triggers, because they are the safest, easiest, and most effective. The most common trigger is caffeine – which can trigger a migraine when used, or when one is withdrawing from the caffeine. Caffeine withdrawal can cause headaches in non-migraineurs as well, but that is distinct from a caffeine-withdrawal triggered migraine. Either way – remove caffeine, wait a couple of months, and then reassess migraine frequency. Other triggers include bright light, strong odors, lack of sleep (very important and often overlooked) and various kinds of foods.  Therefore step one is removal of triggers and lifestyle changes.

If these measures are not sufficient and headaches remain frequent, then there are a number of supplements and medications that can be effective. Vitamin B2, magnesium, and Coenzyme Q10 have all been shown to be effective for migraine prophylaxis. These can be used as initial prophylaxis for milder cases, in patients who want to avoid medications, or as adjuctive therapy with a prescription medication. The medications that have shown to be effective include verapamil, some antidepressants like amitryptyline, and anti-seizure medications like topiramate.

Although controversial for a while, recent large controlled trials support the use of botulinum injections (Botox) for migraine prevention.  There are several possible mechanisms for this, including the obvious muscle relaxation. But also, the botulinum is known to be taken up by the nerve endings and is transported to the cell body. Therefore it may act in the trigeminal nucleus to inhibit activity and increase the threshold for a migraine attack.

At this point the only way to know which treatment will be most effective for which patient is trial and error. I do anticipate the day, however, when a genetic profile might greatly enhance our ability to predict response.

Abortive therapy involves adequate hydration, treating nausea if present, and treating the migraine attack itself (not just the pain). The two components of the attack that are targetted are inflammation (usually with NSAIDS or aspirin-like drugs), and changes in serotonin. Triptans (like sumatriptan, brand-name Imitrex) target serotonin receptors. For some patients rest is often a critical component as well.

Acupuncture for migraine

You will notice I did not add acupuncture to the list, for the reasons I stated above. The research clearly indicates, in my opinion, that there is no mechanism and no added value to acupuncture needles – whatever you do with them. In response to my post yesterday, Sonic wrote:

Dr. N- I understand what you are saying about the unblinded and sham acupunture. But-

“In the four trials in which acupuncture was compared to a proven prophylactic drug treatment, patients receiving acupuncture tended to report more improvement and fewer side effects.”

If that statement is true, then the ‘rituals’ of acupunture work better than the ‘proven prophylactic drug treatment.’
So wouldn’t it be wise to do what works? That is the actual question, right? If a ritual is a better pain reliever than a chemical- why not use the ritual? (I’ll bet there is less liver damage from the ritual…)

The “liver damage” comment refers to one drug, valproic acid, that can cause liver toxicity. This is an older drug and, although it is one of only two drugs actually FDA approved for migraine prevention, it is rarely used for this reason.  I personally never use it. There are many newer drugs with no organ toxicity to use.

Sonic’s assumption is that, of the two possibilities for a recorded effect that I mentioned above, artifact vs non-specific effects, it is the non-specific effects that are important. I disagree with this. Although we cannot say for sure, the fact that the difference between acupuncture and no acupuncture is unblinded opens the door wide for artifacts. Typically, the people doing these studies are acupuncture proponents. We know from reviews of the literature that if proponents fund studies (let alone actually carry them out) they are much more likely to be favorable. It is probable that subjects who would volunteer for such a study are also open to acupuncture, and may have strong beliefs in it. Therefore most of the people involved in these studies want them to work, and therefore I put very little faith in any unblinded results.

Only tight protocols are of any value in such situations, and the tight protocols are negative. The simplest explanation for that is that acupuncture does not work.

Second – the degree to which the effect is real but due to non-specific effects also does not justify using acupuncture. These non-specific effects include relaxation, palpation of the muscles prior to needle insertion, positive therapeutic attention, expectation, conditioning, and the temporary effect of introducing a novel component into therapy (all proven to have a measurable effect). It is unscientific to conclude that a specific intervention (acupuncture in this case) works because of non-specific effects.

In my reviews of these studies the control group getting standard medical treatment is inadequate as a control. Typically they give a single drug, often underdosed or for too short a period, and not comprehensive migraine treatment as I outlined above. This leaves the door wide open for non-medical interventions that are known to work and which can ride along with the acupuncture to give the appearance of an effect.

What this means is that we should eliminate acupuncture entirely from the equation. Any benefit from the ritual of acupuncture can be derived by using the non-specific elements stripped away from the pseudoscience of acupuncture itself. This already is done – with massage, relaxation therapy, biofeedback, stress reduction, or just plain exercise – all of which improve migraines.


This brings me back to the original reason for my post yesterday – it is highly deceptive to conclude from this recent review that “acupuncture works.” That is a lie, or a gross misinterpration of the science. Rather, we can conclude from these reviews the exact oppostive – that acupuncture does not work. We need to finally set aside the needles (which, btw, have their own complications) and all the pseudoscientific claims that come with them. They are a distraction, they are insidious in that they instill unscientific and bizarre beliefs in the public that will cause mischief downstream, and they are a scientific dead-end.


117 responses so far

117 Responses to “Some Follow Up On Migraines”

  1. medmonkeyon 23 Jan 2009 at 11:14 am

    Thanks for clearing that up, Dr. N!

  2. calinthaluson 23 Jan 2009 at 11:14 am

    Thank you for the post Dr. N. I’ve had migraines since I was a teenager, and have never gotten any real medical attention for them. I only get them on average once every couple of months. I get non-specific headaches more often. However, growing up poor and raising a family left my health care far behind. I tried cleaning my act up over the last couple of years, and did attempt to research migraines online. However, as you can imagine, the results from google aren’t exactly clear.

    Normally, I can feel them coming on and if I take ibuprofen (800mg) and lay down in a dark room with a cold rag on my forehead for about an hour it will fade. If I wait too long, nothing at all seems to help for hours.

    Anyway, my daughter has them now as well and I am grateful for the information you have posted here. It was mentioned somewhere else that caffeine and nicotine can have an effect. I quit smoking back in June and haven’t seen an improvement…although it’s still early. I’m afraid it would be more difficult to quit caffeine than smoking…but maybe I can get my daughter to quit with the Mountain Dew while she’s young.

  3. daedalus2uon 23 Jan 2009 at 12:02 pm

    That is interesting about calcium involvement in migraines. Calcium regulates nitric oxide synthase via activation of calmodulin.

    Nicotine is not the only pharmacologically active chemical in tobacco smoke, carbon monoxide is pharmacologically active too. It activates some of the same pathways that NO does (but with much lower efficiency so that ppm levels of CO mimic ppb levels of NO). The NO-mimetic effects of CO may counter some of the adverse effects of nicotine and the tars in tobacco smoke (not a reason to smoke).

    The use of B2, magnesium and coenzyme Q10 to reduce migraine is interesting. Magnesium is used to treat preeclampsia, which is more common in migraineurs. Preeclampsia and migraine are associated with hyperhomocysteinemia and preeclampsia is also associated with elevated asymmetric dimethyl arginine (an endogenous nitric oxide synthase inhibitor). Both migraine and preeclampsia are associated with elevated cardiovascular disease. I suspect that there would be elevated asymmetric dimethyl arginine in migraineurs too and that low NO is a final common pathway in both.

  4. MBoazon 23 Jan 2009 at 12:18 pm

    That all sounds good, but my dental hygienist’s brother’s old roommate gets acupuncture for his migraines, and he says it totally works.

  5. TheBlackCaton 23 Jan 2009 at 12:26 pm

    My mother tried acupuncture for migraines for a while. It was completely and totally ineffective and she ended up quitting.

  6. Thenewyorkdolleyon 23 Jan 2009 at 12:35 pm

    “Yeah, and you can’t PROVE that it isn’t the acupuncture that helps your dental hygienist’s brother’s old roommate, no matter what you do. So take that science!”

    This mirrors an arguement I recently had with my girlfriend about acupuncture and cancer, in which she flat out stated that it’s impossible to prove that acupuncture doesn’t cause spontaneous remission – even if a (hypothetical) study showed that 2% of both the experiment and control groups experienced it…


  7. HHCon 23 Jan 2009 at 12:50 pm

    I know a Chicagoan, martial artist and Hubbard street Dancer who was in significant lower back pain. He strongly believed the acupuncture treatments he purchased healed him. Now I know it was more likely the non-specific effects that worked.

  8. Karl Withakayon 23 Jan 2009 at 12:56 pm

    How consistant is the placebo effect for an individual? Can you select against placebo? Is it possible to produce an experimental group that is inherantly less succeptable to the placebo effect?

    In other words, is it possible to run a subject group through a series of placebos and filter out those that show a response such that you are left with a group that is significantly less likely to show a placebo effect, or is the underlying psychology too complex since the placebo effect in due in part to the individual’s opinion of the individual therapy being tested?

  9. [...] – The Guardian and the Telegraph have done better jobs with their headlines UPDATES (23rd Jan) – Further discussion by Steven Novella where he touches some of the ethical issues and the non-specific effects of acupuncture: It is [...]

  10. daedalus2uon 23 Jan 2009 at 1:34 pm

    Karl, I have a blog post on what I think is the physiology behind the placebo effect. I see it as the normal allocation of ATP and other resources by physiology. When you are under stress, in a “fight or flight” state, healing has a lower priority than maintaining an available supply of ATP to your muscles. Healing gets turned off so that the ATP that healing would consume can be used for something more important, such as running from a bear.


    Of course physiology is very complicated, and there are a zillion pathways that consume ATP, and physiology allocates ATP among those zillions of pathways. How does physiology do that? Very well; astonishingly well; so well that people have posited a mythical principle (homeostasis) that maintains ATP levels constant. There is no such principle as homeostasis, but some myths are difficult to displace.

    Under my hypothesis of the placebo effect (which is not uniquely mine), once physiology has fully allocated ATP and other resources to healing, there is nothing more that any placebo can do.

    Healing is a very complex process; far more complex than we have the ability to regulate artificially. The only way we are going to improve healing is by invoking the normal healing pathways and facilitating their proper function.

  11. DevilsAdvocateon 23 Jan 2009 at 3:06 pm

    Though a licensed clinician in one field, I’m a layman in the medical issues typically addressed here, and this post is precisely why I read this blog. Thanks so very much for the basic info on migraines.

  12. Steven Novellaon 23 Jan 2009 at 3:43 pm

    Karl – that is an interesting question. The pharmaceutical industry is very interested in doing just that – screening out placebo responders. They argue that the placebo effect dilutes out the effect of the drug, making it harder to detect and requiring larger studies with more patients.

    The scientific community has not bought this argument, though. It seems suspiciously like an attempt to rig the numbers. However, if it were done properly – subjects would be screened for being a placebo responder, then re-randomized to either treatment or placebo, it could make sense.

    All the consequences of this have to be sorted out, though. Would this really be selecting for people who are more likely to tell if they are on placebo vs drug, therefore making blinding more difficult? I don’t know. The statisticians have to sort all this out before we change the process of RCTs.

  13. Watcheron 23 Jan 2009 at 3:50 pm

    [quote] That all sounds good, but my dental hygienist’s brother’s old roommate gets acupuncture for his migraines, and he says it totally works. [/quote]

    Wow, I hope your kidding when you type that …

    I mean, after the data shows that it doesn’t work when put to the most rigorous of scientific tests, your willing to take the word of a friends cousin twice removed who saw a movie with you once that once accidentally used your grandfathers toothbrush that was left in a motel 8 in nassau, bahamas back in ’84? That’s just not rational …

  14. Medvetenskapon 23 Jan 2009 at 4:23 pm

    a note on the studies having control groups getting standard pharmacological treatment.

    Two of these studies contain information revealing that a relevant proportion of participants withdrew their informed consent after being allocated to the pharmacological treatment group (control group). And during the study additional patients dropped out of the control group.

    It is therefore obvious that the patients recruited to these studies had a preference for acupuncture (due to some factor introducing bias in the recruitment process), hoping to be allocated to the acupuncture group.

    This clearly put in question the validity of the effect comparison between acupuncture and standard treament. Therefore, in my opinion, one cannot draw the conclusion that acupuncture, sham or real, mitigate migraine better than conventional pharmacological treament.

  15. daedalus2uon 23 Jan 2009 at 4:39 pm

    I don’t think it will be possible to screen out a class of people as “placebo responders”. I think that anyone and everyone can/will respond to placebo, but their responses are heterogeneous. Different placebo responses will occur for different disorders treated with different placebo techniques for different people at different times. The same person could have a different placebo effect at a different time. The effect depends on their immediate physiological state which is variable and how that couples to their psychological state (which is variable too) by what ever placebo technique is used. CAM providers are quite ingenious at coming up with ever more placebo techniques.

    A technique that I think will work is to use something as a “placebo” that pharmacologically invokes the “placebo effect”. I appreciate that this is controversial, but the technique I am using to raise NO levels should (according to my understanding of the placebo effect) invoke the placebo effect by raising NO levels. Once that is done, there is nothing more that any placebo can do. If both legs of the trial had their placebo effect maximally triggered by raising their NO level, the therapeutic effect should be easier to see due to less scatter due to the placebo effect.

    Where this might be really useful is in looking at CAM techniques where now what the trials are looking at is the effectiveness of one placebo technique vs. another; acupuncture vs. sham acupuncture. I think this is why sham acupuncture sometimes works better than “real” acupuncture. Sham acupuncture can be a better placebo (because it doesn’t break the skin) than “real” acupuncture. If you added a “pharmacological placebo” (something which invokes the placebo effect pharmacologically), lesser placebos (everything else) would be shown to be less effective.

  16. Karl Withakayon 23 Jan 2009 at 5:47 pm


    Are there people who are, for lack of a better term, “reverse placebo responders”?

    These would be people who are less likely to notice the actual, real effects of a drug or therapy due to either being less observant/ self aware or having the preconceived notion that nothing will help them.

    If such people do exist, trying to filter out “placebo responders” could inadvertently include a disproportionate number of “reverse placebo responders” unless you retested your filtered group for response to real, effective treatment/therapy.

  17. colluvialon 23 Jan 2009 at 6:00 pm

    Since we recognize that the placebo effect brings some degree of symptom relief to some people, why not figure out ways of enhancing it? For patients who are not responding to other treatments, wouldn’t it be useful to have other ways of providing relief?

    I am NOT saying we should mislead patients by implying that the placebo treatment is having some physical effect. Instead, be honest and treat it as a kind of psychotherapy. Have licensed practitioners administer it under the supervision of an MD.

    Wouldn’t this also be a way to pull the rug out from under CAM practitioners? We could offer placebo treatments under the watchful eye of an MD (have the best of both worlds!) and maybe even be able to develop more successful rituals than treating patients like voodoo dolls or sucking out their blood. And, of course, all the while being completely honest about what’s going on.

  18. Karl Withakayon 23 Jan 2009 at 6:12 pm

    RE: colluvial’s comment,

    Have there been any studies done on the long term viability of placebo effect as therapy? I wonder if under colluvial’s conditions, the placebo benefit would eventually taper off and symptoms would revert to the mean.

  19. daedalus2uon 23 Jan 2009 at 7:07 pm

    Karl, there is a “nocebo” effect, which occurs when a patient is believes that a treatment will cause harm. Pain probably triggers some nocebo effect all by itself. A poor bedside manner will too. The nocebo effect is the opposite of the placebo effect. It is a triggering of the “fight or flight” state and the diversion of ATP supplies away from healing and held in reserve for immediate consumption.

    There is a gigantic problem with bias in selecting the patient population for study. Each patient is unique, and each will have a different course of the disease. Selecting the patients who one thinks will do well is a sure way to bias the trial. That is why the best trials have good randomization; that is an actual random division between the different treatments is used.

    An obvious source of bias in acupuncture trials is that only people who “believe” in acupuncture will even consider it as treatment and consent to a trial which includes it. I never would. Because I believe acupuncture to only be a placebo, I would be quite annoyed if any health care provider suggested it, and that would make any other treatment they suggest suspect too (in my view) and would invoke a nocebo effect in me.

    RE: colluvial’s comment, I consider psychotherapy to actually be a placebo treatment; that is it is a treatment mediated through communication and not through an active surgical or pharmacological technique. I say that as someone who has had (and has very strongly benefitted from) 20+ years of psychotherapy.

  20. cwfongon 23 Jan 2009 at 9:03 pm

    daedalus: Does that mean you suspect the therapist lied to you for 20+ years, but in an entirely trustworthy manner?

  21. Tressaon 23 Jan 2009 at 9:34 pm

    As a life long (so far anyway) migraine sufferer I am always on the look out for information. Thank you!

  22. daedalus2uon 23 Jan 2009 at 9:55 pm

    cwfong, Not at all. Psychotherapy does “work”. I am very pleased with the results. I consider it to be a “placebo” effect simply because it is mediated through communication and not via surgery or pharmacology.

    Therapists lying to their patients doesn’t work. That is bad therapy. If you ever do find your therapist lying to you, that is a sign that they are a bad therapist and you should get another.

    For there to be a placebo effect, there has to be physiology that produces the placebo effect. Understanding how to invoke the physiology behind the placebo effect in the right place at the right time and in the right way would add a great deal to the effectiveness of medicine.

    CAM is not going to figure that out.

  23. cwfongon 23 Jan 2009 at 10:08 pm

    The essence of the placebo effect is its successful application of false information. Thus to view psychotherapy as a placebo treatment is to believe that either the therapist was deceptive or that therapy generates a form of self-deception.
    Both of which of course are sometimes true, but arguably have not always been the case, and have rarely been its purpose.

  24. daedalus2uon 23 Jan 2009 at 11:12 pm

    I disagree with your definition of the placebo effect. Deception is neither necessary nor sufficient for a treatment to be a placebo.

    As I use it, the term means a physiological effect (usually healing) which is the consequence of a treatment that does not involve physiologically active pharmacology or physiologically active surgery.

    The archetypal placebo is the mother’s “kiss it and make it better”. If a mother kisses a boo-boo and the child feels better, there is no deception involved. The child actually does feel better precisely as the mother said the child would.

    If the mother said “I have a pharmacologically active kiss that will provide pain relief”, she would be lying.

    People do get better after using CAM treatments that are placebos. If the CAM provider simply says “this will make you feel better”, and the patient does get better, then there is no deception, and the placebo effect is successful. If the CAM provider says “this homeopathic agent will provide a pharmaceutical effect”, then the CAM provider is lying. Similarly if an acupuncturist says “these needles will activate your chi”, the acupuncturist is lying because there is no such thing as chi. If the acupuncturist says “these needles are a placebo but you will feel better”, and the patient does feel better, there has been no deception.

  25. colluvialon 24 Jan 2009 at 12:03 am

    While it has typically been the manner in which the placebo effect is reported, is it necessary that the patient be convinced that a physical effect is being produced when it actually isn’t? I don’t see how it’s essential that the patient be mislead. I think daedalus2u makes a good point about psychotherapy. There is no physical or pharmocological agent, but still there is improvement.

    It takes a brain to experience pain or suffering. It does not seem like a stretch to imagine that it could be alleviated by manipulating the way the brain processes the experience.

  26. cwfongon 24 Jan 2009 at 12:10 am

    Of course there is deception involved when the mother assures a child that the kiss will make it better – she knows that it’s not the kiss but what she claims will be the effect that causes the child to think that whetever he feels next is better.
    In any case the example is irrelevant because you have simply added your own definition to the word while completely ignoring the accepted meaning of placebo.

    And to quote one source: Placebos by providing false information change a person’s expectations, and through this, the top down management by their brain upon their body.

  27. HHCon 24 Jan 2009 at 12:30 am

    I referred to Barron’s Dictionary of Medical Terms for the words, placebo effect. The definition is “change, usually beneficial, occuring after a substance (a placebo) is taken that is not the result of any property of that substance but usually reflects the faith or expectations that the person has in the substance”.
    Psychotherapy would have placebo effects if your therapist gave you a sugar pill or a glass of distilled water and you consumed them, then proceeded with the counseling session.

  28. cwfongon 24 Jan 2009 at 12:53 am

    Not unless he ascribed some therapeutically beneficial effects to be expected from either the pill or the water.

  29. daedalus2uon 24 Jan 2009 at 10:11 am

    In the context of this discussion we are including placebo treatments that do not involve ingesting something, treatments such as acupuncture and sham acupuncture.

    If you want to include consuming something as part of a placebo, then consuming nothing pretty obviously can’t have a pharmacological effect and any treatment where there is no consumption of anything must be a placebo (unless there is effective manipulation of the body, as in surgery).

    A placebo can change a person’s expectation and change their top-down management of their body. But false information is not necessary to accomplish that. I think that conscious expectations are not necessary either.

    Over at SBM. Harriet once mentioned that she obtained a placebo effect in herself. (I am doing this from memory so I may have confabulated some of the details). Her report was that she was having bad allergies, but didn’t have any antihistamines or other allergy-type medicine, so she took a NSAID in the hope that it would have a placebo effect, and it did. Where was the deception? She knew that what she was taking wouldn’t have a pharmacological effect, she knew the only therapeutic effect would be mediated through the placebo effect, and she got better demonstrating that the placebo effect worked and that deception was not a necessary component to it.

    If the mother doesn’t know any physiology, but is just going by her experience as a mother (i.e. that uber-powerful mommy sense) that kissing a boo-boo does relieve pain in the child, how is she being deceptive? You may not like her explanation, her explanation may not correspond to any reality that you subscribe to, but if she believes it, she is not being deceptive. Since there is nothing pharmacological going on, any resolution of pain can only be from a placebo effect.

  30. cwfongon 24 Jan 2009 at 11:40 am

    If what you recall is correct, Harriet knew she was fooling her system simply because something worked that otherwise wouldn’t have unless she presented it as workable. Expectations don’t have to be conscious – your instincts are motivated by unconscious expectations all the time.
    Your mother may not have been intentionally deceptive, but we both know she was wrong about the kiss itself making pain go away. I’d put this in the category of old wives tales, which are a form of deceptive wisdom, to coin a phrase.
    But your continuing to add your own definition to the word placebo doesn’t make everyone else’s wrong. If false information isn’t a part of the process than it isn’t proper to use the placebo label.
    But I’m done trying to change your mind on this – take it up with Dr. N – see if he agrees with you.

  31. Fifion 24 Jan 2009 at 1:26 pm

    cwfong and daedelus – Regarding the child being kissed by it’s mother… Pain behavior is learned, generally when we’re a kid. A child learns what behavior gets rewarded (kisses, a treat to “feel better”, hugs and attention) or punished (being shunned, hit or otherwise rejected) and then proceeds to behave in a way that will be rewarded or at least not punished. We learn to ignore or magnify our pain, to resist or embrace or merely to accept and move on but very few of us just learn to experience pain as it as it is without cultural/psychological baggage that contextualizes and influences our experience.

  32. HHCon 24 Jan 2009 at 2:26 pm

    Nonsteroid anti-inflammatory drugs are not placebos. Continual use could cause stomach irritation or intestinal bleeding. The good doctor, HH, would have had to consume a couple of Jelly Bellys to get a placebo effect.

  33. daedalus2uon 24 Jan 2009 at 2:37 pm

    cwfong, you are unnecessarily complicating the concept of placebo. A treatment may work by either pharmacology or by the placebo effect. The individuals involved may not know which is which. That is the whole point of randomized double-blind placebo trials. The clinician and the patient both don’t know if the placebo is being given or the drug (which may or may not have pharmacologically effective properties).

    You are completely missing the point. The mother isn’t wrong about the pain going away, it does go away. She isn’t suggesting a mechanism, so she can’t be wrong about the mechanism. The child doesn’t have a conceptualization of a mechanism either, so the child can’t be deceived about what the mechanism is.

    I am not prepared to say what should or should not happen regarding the healing properties of a mother’s kiss and pathologize a child’s resolution of pain as being due to a mother’s deception. Physiology is really complicated. There is a great deal that we don’t know about it, and plausible mechanisms may be incorrect. Tying a placebo effect to deception unnecessarily complicates it. Limiting the term “placebo” to cases involving deception then requires another term for placebo-like effects in the absence of deception.

    One of the original observations of the placebo effect was in soldiers wounded in war. Soldiers with severe injuries required much less morphine for pain relief than did civilians with injuries of the same level of severity. The reduced need for morphine for pain relief was attributed to a “placebo effect” due to the soldier’s relief at surviving the battle. They were injured but alive and now out of the battle. Their physiology could stand down from the “fight or flight” state. Civilians were never in that state when they were injured, so being removed from battle didn’t give them a placebo effect. There was no need to be deceptive about the dose of morphine being given. Wounded soldiers required less then wounded civilians. Not because soldiers were more pain tolerant, but because they were in a situation where they didn’t need the signal of pain, so their physiology turned it down. False information didn’t play a role in the soldiers’ pain relief.

    There was a surgical technique to divert blood from the mammary artery to the heart in an attempt to reduce angina. It worked. It turns out it worked just as well as a placebo operation that just opened up the chest without rerouting the vasculature. The surgeons doing the operation were not being deceptive, there was plausible physiology behind what they were doing, it just turns out that the placebo effects of the operation were sufficiently stronger than the physiological effects of rerouting the vessels that the surgery wasn’t needed.

    My definition of placebo is essentially equivalent to what HHC gave, but expanded to include placebo treatments such as acupuncture. If a treatment works, and doesn’t have a physiologically effective pharmacological or surgical component, then it worked via the placebo effect.

  34. Fifion 24 Jan 2009 at 2:43 pm

    daedelus – You’re ignoring the fact that the child may not actually be in pain and may be manipulating a parent to get attention. Or that the mother may actually be increasingly the experience of pain by rewarding exaggerated expressions of pain (that the child may even be faking since they’re not actually experiencing pain). There are very potent psychosocial aspects to pain expression and experience.

  35. Mostly Harmlesson 24 Jan 2009 at 2:43 pm

    My experience with headaches and Acupuncture was actually critical to my development as a skeptic, here’s my story if anyone is interested.

    During grad school I started getting cluster headaches on a daily basis. I would get a headache every few hours for 20-45 minutes at a time, and it was the most excruciating pain I’ve ever had in my life. It felt like my head was tearing itself apart, and all I could do while suffering the attack was to pace around to distract myself. I went to the school clinic to get diagnosed, and they believed the whole thing was triggered by a bad sinus infection I had just gotten over and they started me on a whole suite of medications in an attempt to get them under control. I started out taking an oral steroid and Tylenol, which did help for a few months, limiting the headaches to 1 or 2 a day, but after about 4 months, the headaches were back to every few hours. Thankfully, my professors were all very accommodating during this whole experience.

    The doctors then dropped the steroid, and gave me a selection of pain killers to take at the onset of a headache, and my neurologist proscribed Depakote after a CAT scan revealed “Nothing Remarkable about my brain” (Still one of my favorite phrases of the whole ordeal) in the hope that it would mitigate the headaches. Eventually, I was down to 2-3 a day, usually when I relaxed or in the middle of the night (what a wonderful thing to wake up to).

    Anyway, 10 months went by from the onset of the headaches, and I decided to give acupuncture a try (this was still during my alt-med days, not knowing any better). I didn’t stop the depakote, but I went to the acupuncture clinic for a couple of weeks and started to see improvement to the point that the headaches finally stopped all together. With my doctor’s permission, I stopped the depakote, and was headache free for about 8 months. At this point I was convinced the Acupuncture had cured me. To drive the point home even further, the headaches started to come back after that 8 month window, I went to the acupuncturist, and they went away again.

    A few years later I start getting into the skeptical community and start reading about all the research on acupuncture, and I have to admit it was really hard to reconcile what I had experienced with what I was reading. I kept thinking that there had to be something wrong with the literature, or rationalizing away the evidence. I’ve eventually had to accept that despite my experience with acupuncture, I don’t know why my headaches went away. It could have been regression to the mean, spontaneous remission, the depakote finally working, anything. But I’ve accepted that I have to treat my own experience with acupuncture the same way I treat any anecdote about homeopathy, ufo’s etc.. I’ve come to trust the controlled trials over what I experienced myself.

    It has been enlightening, if a bit humbling, to have to admit to my friends and family that my prior praise of acupuncture was misplaced. I have found, however, that admitting how my own experience blinded me in that situation, and accepting that I don’t know why I got better gives me an in when discussing any pseudoscience they happen to hold dear. It helps me stress that sometimes not knowing the answer is better than accepting an answer on poor evidence.

  36. daedalus2uon 24 Jan 2009 at 3:11 pm

    Fifi, I am not ignoring that possibility, I am only looking at a subset of the mother-child interactions where the child is in pain and a mother’s kiss it and make it better does exactly that.

    If there is no actual pain reduction, then there is no placebo effect. If the child is faking it, then there is no placebo effect, just as if a patient is faking it, then there is no therapeutic effect of a drug effective for the condition that is being faked.

    I am not at all discounting psychosocial aspects of pain, just trying to define the placebo effect in a way that is most useful and which encompasses all that is relevant.

  37. DevilsAdvocateon 24 Jan 2009 at 4:32 pm

    Um, as the father of nine kids and a frequent boo-boo kisser, these are usually administered immediately after said boo-boo and the pain is going away very soon anyway because most of these boo-boos are quite minor. It may be a correlation error to assign value to the placebic kiss for a minor pain that was already lessening upon administration of the boo boo kiss. These sorts of minor pain go away by themselves.

    Also, when a child falls and skins a knee, the distress is as much due to the fear attending a fall off a bike or whatever caused the boo boo as it is the actual pain from the abrasion or other minor injury. The boo boo kiss may calm and quiet the child, but this isn’t necessarily the result of pain relief. It may be just fear relief.

  38. trrllon 24 Jan 2009 at 4:46 pm

    I still remember from childhood the miraculous analgesic effect of a band-aid (potential research topic: do cartoon characters on the band-aid improve pain relief?)

    I’ve read quite a few clinical trials that included a placebo run-in, with responders dropped from the actual trial, but I don’t know if there is any evidence that this reduces the rate of responding to placebo. It may just be something that is done on a “Well, it can’t hurt” basis. None of the studies that I’ve seen that did this had particularly low rates of placebo responding. Has anybody seen a study that compared a placebo run-in to no run-in?

  39. cwfongon 24 Jan 2009 at 4:48 pm

    daedalus writes:
    “You are completely missing the point. The mother isn’t wrong about the pain going away, it does go away. She isn’t suggesting a mechanism, so she can’t be wrong about the mechanism. The child doesn’t have a conceptualization of a mechanism either, so the child can’t be deceived about what the mechanism is.”

    But you have completely misunderstood the dynamics involved. The mother tells the child a kiss will alleviate pain. She carries out the act. The child, assuming it wasn’t faking, feels the pain abated.

    Nobody has to lie, nobody has to be deceived for misinformation to nevertheless be effective. Because kisses don’t alleviate pain absent the suggestion that they will. The mother knows from experience that there will be a mechanism here that works, She can be completely wrong about why it works or that the kiss is essential to the process. A hug or a rub or even a pinch would work if she believed it and assured the child it would. The touchless laying on of hands is the best example of this.
    And you of all people should know that pain is a sensation controlled as much or more from the part that receives it as by the part that generates it.
    And to say that since the child has no conceptualization about the mechanism, it therefor can’t be deceived about how it works, is just ridiculous. Deception doesn’t have to be intentional and it doesn’t have to be “inserted” b the process. It can be activated by habituation, by trust, false beliefs, ad infinitum.
    We are all fooled all the time by believing in a particular combination of circumstances as the predictable causes of a series of prior experiences.
    But remember that it was you who began this by stating: “The archetypal placebo is the mother’s “kiss it and make it better”. If a mother kisses a boo-boo and the child feels better, there is no deception involved.”
    If so, you shouldn’t have called it a placebo, as without an ultimately deceptive manipulation of the body’s conscious or unconscious expectations, there is no placebo involved.

    Tell me where that definition is wrong. No wait, tell me exactly how you define the term.

  40. daedalus2uon 24 Jan 2009 at 4:54 pm

    DA, those reassurances from a parent that the injury is minor and not something to be distressed about is part of what makes the pain and distress go away. If a parent did the opposite and went into hysterics about a minor boo-boo, the pain and distress would escalate (or would go away more slowly). Pain getting better and pain getting worse due to non-pharmacological interventions are two sides of the same effect, the placebo effect and the nocebo effect.

  41. daedalus2uon 24 Jan 2009 at 5:09 pm

    Quoting HHC’s comment above from Barron’s Dictionary of Medical Terms.

    The definition is “change, usually beneficial, occurring after a substance (a placebo) is taken that is not the result of any property of that substance but usually reflects the faith or expectations that the person has in the substance”.

    My definition of placebo effect is a physiological change produced by a treatment where the treatment does not have an effect mediated through chemical (i.e. pharmacological) effects or mechanical (i.e. surgery) effects.

    Often expectations can be manipulated by deception, but that is not a necessary aspect of a placebo effect. The only necessary aspects are real physiological changes in the absence of a stimulus that could cause that via pharmacological or surgical means.

    Could you give me your definition of placebo effect so I can say how I disagree with it?

  42. cwfongon 24 Jan 2009 at 5:20 pm

    Here’s one of the nocebo effect you claim some knowledge of:
    “In the strictest sense, a nocebo response occurs when a drug-trial’s subject’s symptoms are worsened by the administration of an inert, sham,[1] or dummy (simulator) treatment, called a placebo.”

    Then see this, also from Wikipedia:
    Placebo effect: Expectations
    The physiological effect of a placebo depends upon its suggested or anticipated action. A placebo described as a muscle relaxant will cause muscle relaxation and if the opposite, muscle tension.[28] A placebo presented as a stimulant will have this effect on heart rhythm, and blood pressure, but when administered as an depressant, the opposite effect.[29]
    Related to this power of expectation is the person’s belief that the treatment that they are taking is real: in both those taking real drugs and those taking placebos, those people that believe they are taking the real treatment (whether they in fact are or not) show a stronger effect, and vise versa, those that think they are taking the placebo (whether they are or not) a lesser one. This effect has been found in the success of therapies as diverse as nicotine replacement therapy,[30] the transplantation of human embryonic neurons into the brains of those with advanced Parkinson’s disease.[31] and acupuncture analgesia after dental treatment.[32] There is also an expectation effect upon whether people think the treatment will work: those sceptical about acupuncture analgesia show less placebo effect than those confident that it is a highly effective therapy.[33]
    Another factor is whether the treatment is given overtly or covertly and so able to effect a person’s expectations. Pain killing and anxiety reducing drugs that are infused secretly without an individual’s knowledge are less effective (in the latter case no more so than saline) than when a patient knows they are receiving them. Likewise, the effects of stimulation from implanted electrodes in the brains of those with advanced Parkinson’s disease are greater when they are aware they are receiving this stimulation.[34]

  43. Fifion 24 Jan 2009 at 5:28 pm

    daedelus – You’re assuming the pain actually does go away. I’m with DA here, this is a very poor example because there’s the assumption the child is actually in pain (when they may not be, they may just be surprised that they were airborn and now on their ass) and you’re assuming that the kiss removed the pain (which it may not have existed and if it did exist may actually persist after the kiss). A lot of the time when kids take a spill they’re more surprised and discombobulated than they are hurt (how the parent responds contributes greatly to how the child learns to manage pain and suprise/shock).

  44. cwfongon 24 Jan 2009 at 5:36 pm

    And although I posted this Wikipedia definition earlier, it bears repeating:

    “The placebo effect is the medical phenomena in which a person’s beliefs about an inert substance or a sham therapy results in that treatment having the expected consequences of those beliefs upon health. The placebo effect can also be an additional boost for a real therapy or drug beyond that warranted solely by its actual physiological action.
    The scientific basis of the placebo effect is uncertain (that is not to say it is unscientific nor an unproven phenomena). One aspect is linked to the meaning with which people experience their treatment, usually, their expectations, both conditioned and verbal, about its efficacy. These changes in expectation are linked to changes in the higher cerebral cortical areas of the brain. Another aspect is that these areas of the brain have an evolved role in regulating in a top down manner the body’s physiology, including its homeostasis, and its response to illness (through the sympathetic and parasympathetic systems). While this control is not voluntary, it is open to be shaped by what a person feels and believes. The exact way is not yet understood. Placebos by providing false information change a person’s expectations, and through this, the top down management by their brain upon their body.
    Historians of medicine believe that placebo effects were responsible for the efficacy of most medicine until the twentieth century. Placebos are also widely used in contemporary healthcare.[3]
    The placebo effect places doctors and researchers in ethical dilemmas in regard to their honesty with patients when using sham and inert control treatments in clinical research trials.”

    Sham, inert, expectations, ethical dilemmas, honesty, are the operative considerations. I don’t see any of that in your definition except honesty, which you use in the affirmative, when placebo effect is clearly distinguished by its absence.

  45. Fifion 24 Jan 2009 at 5:37 pm

    daedalus – Actually people are quite often explicitly told that a treatment doesn’t work but they believe it does anyway so they request it. (It’s quite common in chronic pain situations.) It’s still considered a placebo effect in my experience.

  46. cwfongon 24 Jan 2009 at 6:00 pm

    Self-deception is not in daedalus’ lexicon of deceptive medicaments, since it’s a motivating factor for the patient rather than the doctor.
    That the doctor may know of this beforehand does not make him, in deadalus’ view, complicit in the deception.

  47. daedalus2uon 24 Jan 2009 at 6:07 pm

    The definition you give does not cover placebo effects correctly.

    In the case of nausea, a pharmacologically inert material said to make nausea worse actually makes it better and the same inert material said to make nausea better actually makes it worse.


    The inert material had effects the opposite of what the subjects were told to expect.

    This study included instrumental measures of gastric motility, so it isn’t that the subjective experience was manipulated, gastric motility was affected as well.

    So is the material said to make nausea worse a placebo because it makes it better? Or is it a nocebo because it is said to make nausea worse?

    Or is it the definition of placebo that requires placebo effects to correspond to expectations that is wrong. I choose to use a definition of placebo that is simpler and which corresponds to reality.

  48. Fifion 24 Jan 2009 at 6:10 pm

    Well I guess one could name it the “self delusion effect” but that doesn’t make much sense and is kind of beside the point since someone may intellectually know that something isn’t effective but still respond none the less (what we think, what we feel and what we believe, and how we respond, aren’t always aligned – they’re often not). To me this is still an aspect of the placebo effect and to call it something else is ignoring part of what can happen with a placebo effect, which in turn potentially gets in the way of understanding what’s going on.

  49. daedalus2uon 24 Jan 2009 at 6:30 pm

    Fifi, I agree that when physiological changes align with expectations and when they do not align with expectation that both circumstances are due to the placebo effect.

    As I discuss in my blog post on placebos, I think that difference relates to the physiology behind the placebo effect. Much of the ANS is regulated by NO. Stress is a low NO state. The “fight or flight” state is a low NO state. Coming down from a “fight or flight” state requires sufficient NO to overcome the hysteresis of superoxide generation during “fight or flight” to switch physiology from the low NO state of “fight or flight” to the high NO state of rest. I see the physiology of the placebo effect as the neurogenic generation of NO so as to switch physiology to a high NO state.

    Much of the enteric nervous system is nitrergic. It is NO that causes the smooth muscle relaxation that activates gut motility. Being told a substance will relieve nausea would (I expect) raise NO levels due to a reduction in the state of stress, the individual would be in less of a state of “fight or flight”. Similarly, being told a substance will make nausea worse would activate the fight or flight state.

    Activating the fight or flight state is something that people do all the time. That is what athletes do to “psych themselves up” before a competition; that is what soldiers do before going into battle. Activating the fight or flight state is easy because there are zillions of pathways that make superoxide and many mechanisms to generate it. That superoxide pulls the NO level down and activates the fight or flight state. There are far fewer mechanisms to generate NO, and one of the most important sources of NO (basal NO) is very easy to perturb. That is the subject of my research. If the basal level of NO is perturbed, then it is much more difficult for physiology to “stand down” from the fight or flight state.

  50. DevilsAdvocateon 24 Jan 2009 at 6:34 pm

    Daedalus: “DA, those reassurances from a parent that the injury is minor and not something to be distressed about is part of what makes the pain and distress go away.”

    And until one can establish how much of a part, and establish how much was upset over fear instead of actual pain, any conclusion as to (a) what made the pain go away, and (b) why it went away, is pure assumption. The pain was going to go away whether a boo boo kiss was administered or not.

    Daedalus: “If a parent did the opposite and went into hysterics about a minor boo-boo, the pain and distress would escalate (or would go away more slowly).”

    Distress, certainly, would escalate, but you have no idea if the actual pain would increase. This too is a blind assumption.

    Daedulus: “Pain getting better and pain getting worse due to non-pharmacological interventions are two sides of the same effect, the placebo effect and the nocebo effect.”

    In the model you’ve set up – boo boo kisses in response to minor boo boos – the pain is going away and going away quickly no matter what the caregiver does. That is neither placebo nor nocebo. That is the natural course of the body’s own reparative qualities.

  51. daedalus2uon 24 Jan 2009 at 6:50 pm

    “That is the natural course of the body’s own reparative qualities.”

    DA, that is exactly my point. All the placebo effect does is tweak the body’s own reparative qualities. If you are in a high stress fight or flight situation, your body needs to keep ATP supplies at the ready to be used for things like running from a bear. That ATP can’t be used for healing. Turn off the fight or flight situation and that ATP supply can be used for healing.

    The placebo effect isn’t based on anything pharmacological; it can only be based on the body’s own reparative qualities. It constitutes a change in the regulation of physiology, a diversion of ATP from fight or flight to healing.

    A parent telling the child that all is well and they can stand down from their fight or flight state is the essence of the placebo effect. Once the child learns how to evaluate situations, the parent’s input isn’t needed and the child can invoke the standing down from the fight or flight state by her/himself.

    The NO/ATP connection is the major first order effect of the placebo effect. There are higher order effects that are more subtle and more complicated.

  52. Fifion 24 Jan 2009 at 6:54 pm

    daedelus – It seems to me that your NO unifying theory is blinding you a bit here to many other possibilities. (Everything looks like a nail to a man with a hammer, etc…)

  53. cwfongon 24 Jan 2009 at 6:57 pm

    daedalus, I note than when there’s a question of motivation involved, you revert to an explanation of the physiological changes without much concern for the niceties of their psychological impetus. Whether or not fight or flight is involved with the placebo effect says nothing about what prompted that response and even more, says nothing about how your contemporaries have defined placebo.

    If the definition of an effect doesn’t fit a particular reality, then you should define that reality by using a different definition – not by changing the meaning of the word that wrongly defined the situation to begin with. This would seem to be a form of semantic fallacy.

    Otherwise you might as well see if you can adjust the reality to fit what your contemporaries have agreed the word was originally meant to define.

  54. pecon 24 Jan 2009 at 8:16 pm

    After about a minute of searching I found the abstract of a meta analysis, which says: “For the primary outcome of short-term relief of chronic pain, the meta-analyses showed that acupuncture is significantly more effective than sham treatment”


    So you are not being accurate or scientific when you say there is no evidence that acupuncture can be effective.

    I’m sure I could find many more like this, and can’t imagine how you have never seen any positive evidence for acupuncture.

  55. cwfongon 24 Jan 2009 at 8:19 pm

    Your mention of self-delusion made me look up something I read long ago on that subject:

    “When we talk about self-delusion and wonder how it is possible to ignore the fact that we are lying to ourselves, and not be immediately appalled at the prospect, we are assuming that we all have a yearning for absolute truth and nothing less. But we don’t.

    We have a yearning for feeling good about our prospects, and if the truth could be that our prospects are hopeless, for example, we can expect no satisfaction from accepting that as a fact. Any satisfaction would more likely come from an appreciation of such a truth as an abstract concept, and a contentment with our ability to utilize that concept.

    But since our brains are prone to seeking truth as a matter of probability rather than certainty, and can seldom determine any such truth to an absolute degree, we always have the benefit of doubting the accuracy of any troubling assessment, unless the consequences of entertaining such doubts could be imminent and severe.

    And without a mind-set that appreciates the contemplation of truth for its own sake, we may often see truth in the abstract as antipathetic to achieving a satisfactory state of mind – or at least hostile to the prospects of enjoying our existence for the short term.”

    I personally feel that hope adds probability points to the assessment process, but that’s just me.

  56. daedalus2uon 24 Jan 2009 at 8:20 pm

    My definition of placebo is fully consistent with the definition given in Barron’s Dictionary of Medical Terms. It is not consistent with the definition in Wikipedia, or with other wrong definitions. It is not consistent with definitions that attempt to constrain it in non-physiologic ways.

    The placebo effect is a complex concept. Many people do not understand it and have a wrong idea of what it is and what it means. There is not general agreement on what the definition of placebo is among researchers. There are some senior researchers that I am working with who I am pretty sure have exactly the same definition of the placebo effect as I do, and see it in exactly the same physiological terms (as being invoked by high NO).

    Physiologic effects can be produced counter to expectations via non-pharmacologic means. I call those placebo effects. What do you call them?

    In the example I cited above, where the same inert material gave both positive and negative effects, both of which were the opposite of expectations. What do you call that?

    To me it doesn’t make sense to have different names for the same concept. That is highly frowned upon in medical terminology. Each concept is supposed to have only a single name, so that there is no ambiguity in meaning.

    Fifi, I am very much aware that having a hammer makes problems look like nails. That quote is (currently) on the first page of my blog. Not having a screwdriver can make screws look like nails too. I see my NO hypothesis like a screwdriver. With it I am able to differentiate the screws which can be dealt with via NO from the nails, rivets and bolts that require other tools. Having a hammer making problems look like nails is a type 1 error, a false positive. That is more of an NT theory of mind type problem (as I discuss in detail on that blog).

    I am focused on physiology. Psychology affects physiology and is ultimately caused by physiology. I am trying to focus on the physiology without getting caught up in the psychology because psychology is much more complicated. In considering the placebo effect, my conceptualization is transparent to psychology. That is I treat psychology as a ‘black box” and don’t try to look inside it. What ever effects psychology is having, until they have physiological effects I am not going to consider them.

    Psychology is much more individual than is physiology. Individuals can be conditioned to do just about anything. They can be conditioned to respond in just about any way to just about anything. Other people can study those things; that is not something I want to look at in the near term.

  57. daedalus2uon 24 Jan 2009 at 8:29 pm

    Fifi, I am always ready to consider other possibilities. If you can suggest another mechanism I would be happy to consider it.

  58. cwfongon 24 Jan 2009 at 9:04 pm

    daedalus: You have pretty well confirmed my diagnosis here. The placebo effect is the physiological response to that particular psychological “cause” which most will agree demonstrates how deception or misinformation can, under particular circumstances, mimic the cause that the response was initially designed for. (Using “designed” metaphorically.)

    You can choose not to look into a psychological cause, but that’s not the same as arguing that others can’t see or haven’t seen what you chose not to.

  59. daedalus2uon 24 Jan 2009 at 9:17 pm

    cwfong, It can mimic the cause, or it can mimic the opposite of the cause. Which is it?

    Which effect will this disinformation have? To mimic the cause, or to mimic the opposite of the cause?

    If the “disinformation” can have an effect of either sign (positive or negative), what utility does having the term “disinformation” have in the definition?

    By what mechanism(s) does this psychological cause have its effect(s)? And of what sign is the effect, positive or negative?

  60. cwfongon 24 Jan 2009 at 10:29 pm

    It mimics the cause, or the response wouldn’t approximate what the particular mechanism was designed to react to or “expect.” There would be no point in mimicking the opposite of what the mechanism was fashioned to expect before taking the action that the process was intended to cause. What the placebo might accomplish through the deception, among numerous other things, would be an avoidance of side effects from the actual drug or process being replaced by the sham process.

    As I quoted from the Wikipedia source, the scientific basis of the placebo effect is uncertain. And that definition also had about as much as I can tell you here about the expectation tinkering mechanism.

    The utility of using “disinformation” in my explanation was to affirm the essential nature of the deception – deception can otherwise be a particular lack of information, concealment of destructive force, etc., etc. Placebos can of course be used with other than good intentions, such as by CAM practitioners, whose practices can amount to criminal negligence.

    I confess I don’t understand what you are otherwise asking about positive or negative signage.

  61. cwfongon 24 Jan 2009 at 10:36 pm

    You’ve probably already read this, but if not, you may find its information about the pain mechanism helpful:

  62. weingon 24 Jan 2009 at 10:43 pm

    Many years ago in Eastern Europe I witnessed an iliac crest bone marrow biopsy being done on a child to assess response to chemotherapy for leukemia. There was no analgesic or local anesthetic given, nor was there any acupuncture. The child was held in such a way so that he could not see what was being done to him. He did not cry out or even wince. Could I conclude that no treatment provided adequate pain relief?

  63. cwfongon 25 Jan 2009 at 12:11 am

    Could you repeat the question?

  64. cwfongon 25 Jan 2009 at 12:35 am

    Or maybe they just told him earlier that if he made any noise, they’d kill him. But that if he was good kid, he could share in the profits.

  65. cwfongon 25 Jan 2009 at 1:04 am

    daedalus: I read the citation here, http://www.psychosomaticmedicine.org/cgi/content/full/68/3/478
    and I see where your questions about negative effect come from.
    But this citation also gives the explanation and it’s consistent with my necessarily brief description of the process given earlier.
    If you are also asking if any results are other than expected by the experimenters, of course there are.

  66. weingon 25 Jan 2009 at 8:43 am


    That’s one possibility. I was thinking along the lines of stoicism, distraction, or nerve damage from the leukemia itself or the chemotherapy. I know when I had a root canal done without any anesthesia, I couldn’t help clenching my fists and tightening up.

  67. daedalus2uon 25 Jan 2009 at 11:40 am

    cwfong, I am not asking for an explanation of the effect, I am asking what do you call the effect that they observed? You can’t call it a “placebo effect” under your definition because the result was the opposite of what was expected. The treatment expected to worsen nausea actually made it better.

    I call it a placebo effect because it is a physiological effect from a non-pharmacologically active treatment. If you want to call it any kind of “placebo effect”, you have to change your definition so that it fits the described case.

    The descriptive definition of an effect should not have the mechanism called out in the definition. I will use a contrived analogy, but it does fit. The analogy is “broken bone” = “placebo effect”.

    The first clinician with patients with fractured bones, observes that all of them occurred by the patient slipping on a banana peel, so he defines “broken bone” as “fractured bone due to slipping on a banana peel”.

    A second clinician has patients with fractured bones due to falling off a cliff. She wants to call them “broken bones”, but the first clinician says “no, the definition of broken bones is only for fractured bones from slipping on a banana peel, use another term”. So she calls them “croken bones”.

    A third clinician has patients with fractured bones due to falling in a ditch. They are not “broken bones” or “croken bones”, so he uses the term “droken bones”.

    A fourth clinician has a patient who fractured a bone due to a fall caused by an electric shock. They are not “broken bones”, or “croken bones”, or “droken bones”; she needs a new term so she uses “eroken bones”.

    A fifth clinician has a patient who fractured a bone while trying to do a flip. This is not a “broken bone”, or “croken bone”, or “droken bone”, or “eroken bone”, he needs a new term, “froken bone”.

    At a conference, while discussing the different treatments for all these different problems, a sixth clinician says “these are all really the same, a fractured bone due to mechanical trauma, the details of how it happened are not important in how to understand it or treat it. Let’s take the fracture mechanism out of the definition and just call any fractured bone a broken bone. Then it will also cover bones fractured due to mechanisms we haven’t observed yet.”

    I understand your desire to include deception in your definition of “placebo effect”, but that forces you to impute deception when ever there is a placebo effect. In Harriet’s case, your imputation of deception is extremely contrived. Harriet knew that what she was taking was non-pharmacologically active for the effect she wanted. She wasn’t deceiving herself, what was she deceiving? Her liver? Livers don’t have sufficient cognitive power to be deceived.

    In the case of the mother comforting her child, none of the parties are aware of any deception, the only “deception” is imputed by you because the outcome (resolution of pain) doesn’t match the reality that you think should have happened (non-resolution of pain absent pharmacologically active agents). Mother-child interactions are sufficiently complex and insufficiently understood that to impute a single mechanism (such as deception) to the results of a comforting action by the mother is very likely wrong. That there is a physiological effect of the resolution of stress and pain is not in question. That there was no pharmacologically active compound used by the mother is not in question either. To me, that is sufficient to label the effect a “placebo effect”.

  68. HHCon 25 Jan 2009 at 11:55 am

    weing, OI!

  69. Fifion 25 Jan 2009 at 12:50 pm

    daedelus – “In the case of the mother comforting her child, none of the parties are aware of any deception, the only “deception” is imputed by you because the outcome (resolution of pain) doesn’t match the reality that you think should have happened (non-resolution of pain absent pharmacologically active agents).”

    How do you know none of the parties are aware of the deception? That’s a huge assumption to make and you keep making all kinds of assumptions like this that aren’t based in any evidence (just a desire for this example to fit your NO hypothesis). Children are very adept at consciously manipulating their parents – you’ll often see a kid look to see what effect his crying or tantrum is having. The parent usually doesn’t really believe that a kiss actually takes away injury or physical suffering, otherwise they’d be getting people to kiss their own booboos. Parents consciously deceive their children all the time – in fact childhood is a time when parents often encourage things they know to be fantasy and magical thinking because it’s a useful way to manipulate their children (“be good or Santa won’t bring presents, he can see everything you can do” for example)

  70. daedalus2uon 25 Jan 2009 at 2:54 pm

    Fifi, I don’t doubt that there are mother-child interactions that do involved deception. But there are cases that do not involve deception. The definition needs to fit those cases too. Cases I am most thinking about are with very young children, infants before they have the cognitive capacity to be deceived. The case that trrll mentioned, where a Band-Aid had miraculous analgesic powers.

    For you to impute that all mother-child interactions where every mother’s non-pharmacological intervention that causes physiological effects necessarily involves deception is to go far beyond what I have stated.

    The act of comforting a child will cause physiological changes; reductions in stress hormones, reductions in heart rate, blood pressure, changes in respiration. What “deception” does that involve? There is deception when the parent says “it will be all right” when the parent knows it won’t be. There is not deception when the parent says “it will be all right” when the parent knows it actually will be all right. I don’t think it is useful to have a definition of the physiological changes that occur in the child that is contingent on the truth value of the statements that the parent is making. One could call it “comforting” when the statements are true and “placebo comforting” when the statements are false. What is it when the parent believes it will be all right and it turns out not to be all right? (“false comforting”?) Or if the parent believes it will not be all right but it does turn out all right? (“false placebo comforting”?)

    The physiological changes that occur in the child are (very likely) the same in all of these instances if the child doesn’t know if it will be all right or not. That is the physiology that I am interested in, not the details of the truth value of what is communicated to the child by what medium. There is a lot more communicated between the mother and child than the “facts” of whether or not it will be all right. The major thing communicated is that the mother is present and is aware of the child’s difficulties and is prepared to deal with them as the child’s mother. That is really all the child needs to know, everything else is a relatively unimportant detail. Much of that is communicated somatically, by the child being held, by the child perceiving the mother’s state of agitation and stress, by the child smelling the mother’s stress pheromones, by a whole host of other things that are mostly not understood. Much of the body language is much more difficult to be deceptive about than is spoken language.

    If the mother is holding the child and the child is comforted; the somatic message that mother is present, is aware of the child’s needs and is prepared to deal with them as the child’s mother may trump all other messages.

  71. cwfongon 25 Jan 2009 at 3:03 pm

    daedalus writes: “She wasn’t deceiving herself, what was she deceiving? Her liver? Livers don’t have sufficient cognitive power to be deceived.”
    That’s exactly where you are wrong, and probably why you are having trouble with this whole concept. There’s a calculative aspect to all our bodily functions – they aren’t just in the brain, but they assist the brain, as do the brain’s computational functions work with them and off of them.
    These functional areas can be fed signals that tell them one substance is on the way or has arrived that is in fact another substance fashioned so the apparatus doesn’t know the difference – until it’s either too late to react defensively or it has acted as it was expected to do and the substance has been naturally eliminated. Or a combination of the above responses. (I add that as I detect an “either or” aspect of you analytical approach that is often enabler of the self-deceptive process that has much to do with this phenomena.)

    But this aspect of the function falls now in your area of expertise, not mine. I’m more of a cognitive apparatus guy, and that apparatus has functional elementsi all over the place. I’m a bit surprised you didn’t know that.

  72. Fifion 25 Jan 2009 at 3:10 pm

    daedalus – You’re still trying to make reality fit into your NO hypothesis and not really owning up to making assertions that aren’t based in evidence. Kisses also don’t work to get rid of pain with babies that are fussy because they’ve got a diaper uncomfortably full of poo, who want to be fed or who are having teething pain. And how can a kid “not understand” but be excited and rewarded enough by cartoon characters on a band aid for it to make a difference?!? That simply makes no sense to assert!

    Um, young children’s cognitive capacities are very easily deceived – I don’t know where you get the idea that they’re not! Why do you think that young children or infants can’t be deceived?

    You’re also making incorrect assertions about what I said. Seriously, you’re trying to shape reality to support your hypothesis by dismissing anything that doesn’t support it!

  73. Watcheron 25 Jan 2009 at 3:12 pm


    “I’m sure I could find many more like this, and can’t imagine how you have never seen any positive evidence for acupuncture.”

    It’s not that it’s never been seen, it’s that the scientific methods as a basis for these meta-analysis are bogus. Every acupuncture “experiment” to date has been non-blinded. Practicing acupuncturists perform their needling on willing, non-screened, patients. There is no control, there is only, “How many felt better after completing the session.”

    This new study introduces a double blind aspect to the study of acupuncture and it’s effectiveness in treating lower back pain. What they found was that by double blinding the study, there is no statistical difference between the sham and acupunctured groups.

  74. Watcheron 25 Jan 2009 at 3:23 pm

    “I’m sure I could find many more like this, and can’t imagine how you have never seen any positive evidence for acupuncture.”

    Like I stated before, the positive evidence was based off of bad experimental practices that every PhD candidate knows not to neglect. You need to have a double-blinded study for it to be a rigorous test!

    I would suggest you go back and read the blog post the day before this one came out. It may answer some of your “problems” with his post. If you dont want to though, here’s the beef in tortilla. Acupuncturists want to disregard the new study, the one that is experimentally rigorous, in favor of the old ones because they were positive. Let me rephrase this, they want to use the old studies because they fit their ideals and ends, not because its good science.

  75. daedalus2uon 25 Jan 2009 at 4:24 pm

    cwfong, what kind of deception was involved in the nausea study? People were told it would make their nausea worse and they were deceived because it actually made it better? If they were told the opposite, that it would make their nausea better it actually did the opposite, it made it worse.

    Do you have actual data that these functional areas are being fed wrong signals or is that simply an ad hoc assumption you are using to generate epicycles to modify your “deception” idea of placebos? Epicycles as were used in the geocentric model of the solar system to make everything fit?

    We go from being told something is pharmacologically active when it is not and producing the same effect as the pharmacological agent. That is a clear case of deception at the cognitive level producing the expected effect.

    Harriet knew she was taking something not pharmacologically active but it had the desired effect. You now say that the deception occurred at a functional level between the cognitive level and the effect level.

    In the nausea case I brought up, the substance had the opposite effect. At what level does the deception occur in this case?

    Since many of those signals are mediated by NO, an inappropriately high NO level could be construed as a deceptive signal, but (I think) that is a poor way to look at it. Deception connotes that there is a “correct” signal and that via deception an incorrect signal was sent and/or received. Since we don’t know (and have no way of knowing) what the “correct” signal “should be”, imputing that the signal is deceptive is to make unwarranted assumptions. What theoretical basis is there for a placebo effect only corresponding to expectations, other than the assertion that it does? The data says otherwise (in some cases).

    Fifi, what was the deception re the band aid? I have no doubt that the child did get excited. What does getting excited have to do with deception about the analgesic properties of band aids? Band aids don’t have analgesic properties other than via the placebo effect. Band aids with cartoon characters don’t have analgesic properties either. Unless there is a claim that cartoon characters do improve the analgesic properties of band aids there isn’t a deception that they do.

    I am not trying to assert that each and every case of a mother comforting her child is due to changes in NO physiology. You seem to be saying that none are. A mother’s kiss does not relieve every instance of an infant crying. It does relieve many of them. That a mother’s kiss does not relieve the crying due to a poopy diaper is not evidence that a mother’s kiss does not relieve crying due to a boo-boo.

  76. Fifion 25 Jan 2009 at 4:40 pm

    daedelus – “Cases I am most thinking about are with very young children, infants before they have the cognitive capacity to be deceived. The case that trrll mentioned, where a Band-Aid had miraculous analgesic powers.”

  77. cwfongon 25 Jan 2009 at 5:53 pm

    daedalus, you are a lot like the horse that refuses to drink the water that it asked for help in finding to begin with. Deception as it involves the placebo effect is essentially misinformation. Depending on the particular problem to be addressed, it can be inserted at any end of it, in the middle, or be endemic to the target mechanism and activated accordingly. They write books about this stuff and I’m not going to try to write such a book here.
    I’d consider addressing the nausea case you brought up, but then i expect you’d find another example that you would hope I couldn’t deal with. Hope being the entry point to the authentication of our delusions.

    And as I believe I said before, things don’t always go as planned. That’s why we keep experimenting.

    Let me add something else here. You said you avoid looking into the the black box of psychology. it would be simplistic to conclude that you are simply afraid of what you would find there. I suspect that you were born with or have acquired a life strategy that doesn’t involve the use of deceptive practices, and worse (or maybe better), is not fully equipped to spot or defend against the subtle natures of some such. That does not mean that you are precluded from learning how others do so. You don’t have to become like them in the process.

  78. HHCon 25 Jan 2009 at 5:55 pm

    We know that HH, Harriet Hall was hassled by a headache. But how huge a headache? A NSAID was said to work on this headache of unknown type. This self-diagnosed condition was cured by an over- the- counter drug. I hesitate to ask. but what was the method of consumption? Were any fluids imbibed which naturally would make Harriet happy? I do not believe this to be a proper placebo clinical trial. But, Harriet’s headache hassles should be brief.

  79. daedalus2uon 25 Jan 2009 at 7:00 pm

    cwfong, you are not understanding my point. I am not talking about mechanisms for the placebo effect, simply what the definition of the placebo effect is. You want to bring in a mechanism of deception into the definition of the placebo effect. I have brought up cases of a physiologic effect mediated through a non-pharmacologic mechanism that does not fit your definition because your definition is over specified.

    We haven’t even gotten to mechanisms yet. I am not trying to address mechanisms in the definition. I want the definition of the placebo effect to be independent of the mechanism(s) behind the placebo effect because we don’t know what they are, or how many there are, or if they are the same for different individuals, or if they are the same for different placebo treatments.

    You want to include “deception” in the definition, but are prepared to expand what deception means from communication from an outsider, to implied communication, to internal thoughts, to cryptic signals of the autonomic nervous system. With that expanded definition of “deception”, you can rationalize anything.

    I appreciate that the placebo effect is complicated. Over specifying the definition to limit the definition to only certain types of placebos is the wrong way to try and study the physiology of the generic placebo effect. It is like the broken bone analogy that I used. You can treat each separate case of a non-pharmacological agent having a physiological effect, as a separate case but then each case has to be treated uniquely. Then you only have a collection of unique anecdotes, not a phenomenon that can be studied.

    Treating broken bones from slipping on a banana peel separately from broken bones from falling off a cliff, separately from broken bones from falling in a ditch, separately from broken bones from falling after an electrical shock and separately from broken bones from doing a flip can be done. It is much more difficult to do research and understand broken bone therapeutics if each type of broken bone is treated differently.

    I agree that experiments don’t always go as expected. When our experiments don’t generate the results we expect, we have to modify our expectations, not the results of the experiment. I have modified my concept of the placebo effect to include the nausea experiment where the placebo effect was the opposite of what was expected both by the experimenters and by the subjects. My definition of the placebo effect is consistent with every example that I am aware of. It is also consistent with the generally accepted definition in Barons. Yours isn’t. You have to modify your concept of “deception” to make it fit.

    I do have Asperger’s, so being deceptive is difficult for me. It is not something I like to do and so I try to avoid it as much as possible. Being honest with myself is extremely important in doing the work that I do. If I am not intellectually honest, the equipment I design and build doesn’t work. This is a characteristic feature of people with Asperger’s. I go into my hypothesis of why that is on my blog re the theory of mind vs. the theory of reality.

    I am not afraid of the complexity of psychology. I know it is orders of magnitude more complicated than physiology which is orders of magnitude more complicated than most people appreciate. Psychology is certainly involved in many aspects of the placebo effect for many individuals. There also has to be physiology behind the placebo effect. Ultimately is it physiology that generates the physiological effects that the placebo effect is mediating, be that pain reduction, increased healing, allergy relief, or what ever. I want to understand the physiology of the placebo effect without invoking psychology; because once psychology is invoked the problem becomes intractably complex and difficult.

    Invoking psychology doesn’t change the need to understand the physiology of the placebo effect, it simply adds a great deal of complexity to it by making the psychology of the individuals involved in the placebo effect important in any (and every) study of it. Never is there enough data on the individuals involved in any published study to be able to understand their psychology. To require understanding the psychology of the individuals involved before trying to understand their responses to the placebo effect is to make such studies impossibly difficult.

  80. cwfongon 25 Jan 2009 at 7:41 pm

    I will have to answer by stating categorically that you cannot satisfactorily understand the placebo effect without invoking psychology. It’s a manipulative process where the physiological functions are being affected. The manipulation is psychological. No-one should object to your interest in working on the physiological aspects alone, but they can’t help but object when you deny there are important psychological aspects, and I particularly object to the assertion that the focus should not be on deception.
    I nave in no way expanded on the accepted definition of deception. I have merely pointed out the numerous, and yes deceptive, ways that it can be applied to what is essentially a deceptive process.
    Psychology is not impossibly difficult. Although some such as evolutionary psychologists seem to find it impossible to get right.
    But it’s a science, or can be, nevertheless. You are scientist. Give some respect to those scientists who may be able to supplement your own work in this field.

  81. daedalus2uon 25 Jan 2009 at 9:27 pm

    cwfong, you are really not understanding what I am saying and where I am coming from.

    Deception can be an important part of the placebo effect. Has it been demonstrated to be the only important thing of the placebo effect? Is it the only thing that is important in the placebo effect? Is it the only part of the placebo effect? You seem to think so, so much that any aspect of the placebo effect that does not involve deception you consider so unimportant and so diametrically opposed to what the placebo effect is as to be not worth being considered within the placebo concept.

    I understand the role that deception plays in the placebo concept. You do not understand that there is a role for anything else. You do not understand that there is a role for non-deception in the placebo effect even when I show you research that cannot be explained by your deception concept; research that disproves your hypothesis that the only important concept in the placebo effect is deception.

    I will state categorically that you cannot understand the placebo effect without understanding how it can function in the absence of deception. That you don’t want to consider and understand any aspects of the placebo effect that don’t involve deception is fine. Don’t pretend that I am wrong because I can look at it with a broader perspective than you can.

    Science and physiology and psychology are big concepts, big enough for me to want to look at a small part of physiology without simultaneously looking at larger parts. I am really making quite good progress at understanding the placebo effect without invoking psychology. I don’t dispute that psychology is important.

    All organisms have “top down” control of physiology. All organisms with nervous systems have “top down” neurogenic control of physiology. The placebo effect didn’t suddenly “poof” into existence with humans. The neurogenic and physiological mechanisms behind the placebo effect occur in other organisms too. I think that some of them occur in just about every organism. Do we need to consider the “psychology” of every organism that placebo-like effects occur in?

  82. cwfongon 25 Jan 2009 at 10:38 pm

    If there is no deception involved then you can say it is similar to the placebo effect except what is being used to effect the results can’t be called a placebo. If you can’t use a placebo to cause the effect, then, while it still may be a deceptive process, there was no placebo. But correspondingly, if there was no deception, you won’t be correct in using the term placebo either.

    Placebo-like effects may occur in other organisms by mistake, and the mistake could conceivably be caused by a form of deception – such as from ingesting the wrong kind of mushroom and getting a protein fix nonetheless. Because there is a deceptive process when a plant’s evolvement is motivated by masking their nature from predators, etc.

    And organisms do develop manipulative strategies, which, in a metaphorical sense, are part of their “psychology.”

    The placebo effect in a way did suddenly spring into existence with humans because in essence it’s the label we gave to the recognition that such a mechanism existed in nature.

  83. sonicon 25 Jan 2009 at 10:47 pm

    Unfortunately what I am reading here is what I generally read about the placebo effect. (Clearly little agreement and less understanding)
    I hope you can understand why I would opt for fewer days of unbearable pain- regardless of the philosophical objections. If the acupuncture provides fewer days of unbearable pain for someone-as the studies suggest they do- I would hope that any sane person would allow them and in fact encourage them to get the treatment that works for them. (Is anyone suggesting that the people in the studies did not experience fewer days of unbearable pain?)

  84. cwfongon 25 Jan 2009 at 11:28 pm

    Acupuncture’s effect has been shown as comparable to a placebo effect, but since acupuncture has not been used as a furtive substitute for some other process, it doesn’t qualify as a placebo.

    Nothing’s stopping you from using it if you find that it has adequately assisted the process of fooling yourself in an advantageous fashion.
    Even if it may have delayed some more necessary treatment, hell, we all can appreciate living in the moment.

  85. sonicon 26 Jan 2009 at 3:49 am

    I admit that if I can fool myself into good health (which apparently I can) I am not embarrassed by that fact.
    To the contrary I would want to get better at it. (The skill of ‘fooling myself into good health’ seems to be one I would like to master.)

  86. cwfongon 26 Jan 2009 at 4:07 am

    You have already fooled yourself into equating good health with temporary pain relief.

  87. Fifion 26 Jan 2009 at 9:26 am

    cwfong – “You have already fooled yourself into equating good health with temporary pain relief.”

    Exactly! People often believe that the pain is the injury or disease when the pain is a symptom of the underlying problem (to be entirely simplistic, pain is the body’s way of saying “stop doing that NOW” or “take care of this NOW” and quite often follows discomfort and minor pains that have been ignored…which is one reason many chronic pain patients can tend to be more prevalent amongst certain personality types). Temporary pain relief (or permanent pain relief) of a psychosomatic nature tends to indicate that the problem itself may very well be psychosomatic in nature. In reality, the temporary pain relief is only a bandaid that covers the deeper psychosocial issues that are manifesting themselves as psychosomatic pain.

    Conversely – and very relevant to the “wellness” industy – minor discomfort or general dissatisfaction gets sold as being “unwell” and we’re assured we should feel fantastic ALL the time (a bit like a fairytale happy ending, it’s not surprising that people who “have it all” but who aren’t happy or content are attracted to woo!…it’s also why teaching these kinds of people how to have a bit of gratitude and appreciation – how to find meaning in their life – can change their life dramatically).

    That one would need to “fool” oneself into good health and fool oneself (and others by a ritual performance of pain) that one is in bad health – tends to indicate that the pain may be serving a purely emotional/psychological need (aka psychosomatic pain). If the problem is an emotional/psychological one, the actual root of the issue is not only being ignored by all the ritual pain performances (that serve a purpose other than “healing”) but it’s also being exploited by the CAM practitioner. Ultimately, the CAM practitioner is engaging in a neurotic relationship with their client and encouraging it for profit (though most of this is likely to just be an unconscious neurotic dance between the CAM practitioner and their client).

  88. RickKon 26 Jan 2009 at 12:59 pm

    So the basic question is – if placebos work for many patients, should we fund/support placebo treatment?

    Should health insurance cover:
    - sugar pills (the “pirin” tablets in “The Birdcage”)
    - energy field manipulation (Reiki)
    - homeopathic remedys
    - accupuncture
    - mediums
    - culturally-appropriate faith healers (Peter Popoff, witch doctors, group prayer, etc.)
    - crystals & pyramids
    - etc.

    Once something has been proven (as accupuncture has) to be a placebo remedy, should we continue funding research into it?

    It would certainly be a nice way to classify medice versus “alternative therapies”. Those with actual physiological effect would be medicine, those without would be “alternative”.

    Medical students would continue to concentrate on courses like biology and chemistry, while students of alternative therapies would take courses in influencing, public speaking, and the performing arts. Imagine, the Juilliard School of Complementary and Alternative Medicine.

    This could also spawn some truly useful research into methods of enhancing the placebo effect. For example, does computerized mood lighting and carefully selected background music enhance the naturopathic experience?

    Maybe that’s our way out of this recession, to support the economy by building the “meditainment” industry. It seems like a winning idea, as there is apparently vast demand, but the barriers to entry are quite low.

    Food for thought.

  89. Fifion 26 Jan 2009 at 1:39 pm

    RickK – My feeling is that we should understand what component of the treatment is actually the therapeutic component and then use that. I see no reason to validate or promote the non-therapeutic aspects since it will merely lead to more confusion.

    There’s also these fields of medicine called “psychiatry” and “psychotherapy” (not to mention social work and all the other “helping” professions)….

    If people really only need a hug then that’s what they should get (along, perhaps, with some assistance learning how to get hugs and the support they need from their friends and family rather than having to pay someone to hug them). Of course, sometimes what’s really needed is some psychotherapy so that the person can learn how to get their emotional/relationship/nurturing needs met in a healthy way. Conversely, maybe doctors need more training in how to provide comfort for patients or this needs to be properly integrated/re-integrated into the contemporary practice of medicine (rather than it being a matter of being a people person/empathetic by nature or not).

    And, please don’t suggest SCAM get more involved in the entertainment industry – they’re already smearing their vaseline covered hands all over the place and using it as a platform for propaganda and to promote their pseudoscience. Why do you think Scientology chases movie stars so hard?

  90. Watcheron 26 Jan 2009 at 8:57 pm


    And how long before the insurance companies start ONLY providing minimum coverage with the cheaper homeopathy route? I mean, 30% is better than nothing right? It’s no 95%, but hey what do you expect.

    It’s a slippery slope my friend. There is nothing right about homeopathy and the lies it peddles as science. At best, they offer foot remedies. At worst, you get the AIDS deniers and their snake oil. There is no one or the other, only both or neither. As soon as placebos are given as true remedies, it opens the door to the worst kind of scam artists and predators bent only on making money off of the ignorant.

  91. Watcheron 26 Jan 2009 at 8:58 pm

    “Why do you think Scientology chases movie stars so hard?”

    Because even if they havn’t gotten anything right to date, they still have a kick ass marketing firm backing them up! :D

  92. RickKon 26 Jan 2009 at 9:17 pm

    For the record, the prior post was typed with tongue firmly planted in cheek.

    That said, I do want to explore this just a bit.

    Fifi said: “maybe doctors need more training in how to provide comfort for patients or this needs to be properly integrated/re-integrated into the contemporary practice of medicine”

    So, let’s say we have a doctor who provides a valid medical treatment for chronic pain with a 50% success rate – 50% of the patients report being cured.

    Now, if the doctor learns to smile and speak slowly and look concerned while giving the treatment, let’s say 55% of patients report being cured. That’s a valid improvement in bedside manner and considered good medicine.

    Finally, let’s say the doctor puts on a red silk brocade coat and a blue silk hat, lights 3 sticks of incense, and waves his hands around the patients and calls it Ancient Laotian Energy Juggling. And let’s say 65% of patients report being cured.

    Is it medicine or is it CAM?

  93. wertyson 26 Jan 2009 at 10:29 pm

    Myofascial pain in the trapezius and neck muscles can be a triggering factor fpr headaches, or cause referred pain. If an acupuncturist puts a needle in the trigger point they may inadvertantly perform a useful ‘dry needling’ treatment which can relieve headaches due to these soft tissue triggers.

    Therefore it appears as though miraculous cures are possible which have no basis in the placebo effect.

    As for botulinum toxin, interestingly it is of definitely no benefit for ‘tension headaches’ which was the one headache type it would seem a no-brainer to be effective in. The data for migraine is much stronger, though not totally accepted yet. Figuring out the mechanism for this somewhat unexpected result will give very useful insights into both migraine and ‘tension headache’.

  94. sonicon 27 Jan 2009 at 5:07 am

    and what if we found that different people responded to different types of ‘ritual’?
    I’ve often wondered what would happen if a drug company tested the new drug by having a doctor give out the placebos and a guy with unkempt hair, vomit on his shirt, and urine on his pants saying, “dude this is the good sh#T!!!” giving out the drug to be tested.
    I’m guessing we would find the value of ‘ritual’ to be greater than thought.

    I don’t take pain medication because I know the difference between good health and temporary relief.

  95. Fifion 27 Jan 2009 at 9:34 am

    sonic – You’re lucky not to have chronic pain, it’s much easier physically and psychologically to support transitory pain. It’s also quite difficult for most people to imagine (and have empathy for) the chronic pain of others – which is why so many people have a stoic “just get over it” attitude while having no real concept or empathy for the person in pain.

    Most minor recurring pain is transitory and self resolving – or easy to prevent or resolve with regular exercise in the case of transitory back pain. There’s absolutely nothing wrong with availing oneself of termporary relief for temporary pain – even though some people have been encultured to think being stoic is somehow morally superior, more “manly” or ultimately somehow more desirable and constructive (it’s actually the kind of pain behavior and psychology that can be causitive of the kind of RSIs that result in chronic pain so can be a destructive response to pain over the long-term). In fact, more recent research is showing that early intervention in pain caused by traumatic injuries with pain medication is highly preventative of chronic pain.

    Being healthy and “feeling well/good” aren’t always the same thing – though this is what SCAM often is selling and why so many people believe it “works” (often with a condition that was really self limiting and not particularly severe to begin with).

  96. daedalus2uon 27 Jan 2009 at 11:45 am

    Here are a number of scenarios. Which ones are due to the placebo effect and which ones are not.

    A. A patient is injected with 1 mL of saline and experiences pain relief.

    B. A patient is injected with 0.001 mL of saline and experiences pain relief.

    C. A patient is injected with 1 molecule of saline and experiences pain relief. (I appreciate that saline doesn’t come in molecules, consider this to be the smallest non-zero quantity of saline possible)

    D. A patient is stuck with a needle attached to a syringe filled with saline but nothing is injected and experiences pain relief.

    E. A patient is stuck with a needle attached to a syringe that is empty, nothing is injected and experiences pain relief.

    F. A patient is stuck with a needle that is not attached to a syringe and experiences pain relief.

    G. A patient is stuck with an acupuncture needle not attached to a syringe and experiences pain relief.

    H. A patient is stuck with a sham acupuncture needle not attached to a syringe and experiences pain relief.

    Which of the cases (if any) are mediated through the placebo effect? To me, they are all mediated through the placebo effect.

  97. HHCon 27 Jan 2009 at 12:19 pm

    daedalus2u, Nice post. Your NO theory applies to exertion headaches located in the occipital area. Patients strain while lifting or hold their breath during the physical exertion.

  98. weingon 27 Jan 2009 at 12:49 pm

    Does the patient experience pain relief when the needle is withdrawn?

  99. daedalus2uon 27 Jan 2009 at 1:14 pm

    For the thought experiment I am contemplating, consider that the “pain” measurement is taken some time after the cessation of treatment; 10 minutes or so; a time long enough for stuff to happen and for the transient to have died down somewhat. I am more concerned with how these different scenarios do, or do not, meet peoples’ definitions of the placebo effect.

  100. cwfongon 27 Jan 2009 at 1:59 pm

    You have left out any mention of whether the practitioner is knowingly witholdng information from the patient in the process. It’s the extent of that information and his purpose in withholding it that makes the difference.

    Did your therapist, for example, advise you to be honest with yourself without pointing out that parts of your brain withhold information from the other parts for the good of the whole shebang? Making the effort to be honest with yourself factually impossible, although the atempt itself can effect a certain symptomatic relief.

  101. daedalus2uon 27 Jan 2009 at 2:30 pm

    cwfong, in the context of the thought experiment I propose, the clinician administering the shots and needle sticks is completely honest with the patient about exactly what is being done.

    As is always the case, we don’t know precisely what is going on in the mind of either the clinician or the patient, only what has been done.

    If you want to, you can consider separate cases where the clinician is being truthful and where the clinician is silent and where the clinician is lying (if this changes whether you think it is a placebo or not).

    You can also consider separate cases where the patient has different mind sets (if this changes whether you think it is a placebo or not).

  102. cwfongon 27 Jan 2009 at 3:07 pm

    Ultimately, the only way to know if an actual placebo is the cause of the effect is to know the intention of the practitioner, and in turn be able to believe he’s giving you a truthful rundown of his part in the process.

    Remember, the phrase is “placebo effect” and that would seem to imply that a placebo be the cause of the effect. So you should be concerned with the definition of placebo as cause rather than having the effect be the definitive aspect of its cause – since similar effects cab have diverse causes and similar causes can have diverse effects, all things never in fact being equal.

  103. daedalus2uon 27 Jan 2009 at 4:33 pm

    cwfong, so your position is that whether scenarios A-H are mediated through a placebo effect depends on the detailed mental state of the person administering it?

    So if the person administering scenarios A-H has the mental state that makes the effect not be a placebo effect, what do we call it?

    What if the needle sticks were administered mechanically? If a mechanical device were used for the administration? If the administration mechanism didn’t have a mental state how does that influence whether it is a placebo or not or if the effect is mediated through the placebo effect or not?

  104. cwfongon 27 Jan 2009 at 4:35 pm

    There’s no good way to simplify an explanation of this process, but I’ll try one more example:
    Hope influences expectations. In a product testing scenario, you can give a patient misinformation by giving no information as to the nature of his particular medication under circumstances where he is also told there’s a fifty fifty chance it’s active rather than inert, and that if active, there’s hope on your part that it will be beneficial.

    Hope that he’s getting the active version fuels the expectations that lead to the physiological effect that you can then examine – one purpose of course being to see if the active version causes efects that are dissimilar to the inert one. And thus separating the results caused by expectations alone from those caused by the actual chemical makeup of the test product.
    So then someone will ask, where’s the deception, since the doctor didn’t lie to the patient? But in a way the doctor did “lie” to the patient by not explaining that he was counting on knowledge that under these circumstances, the patient will often lie to himself.
    It’s the form of lie similar to what we all recognize as suspension of disbelief. The patient decides to :believe that the “pill” he got was the active version because he wants it to be and uses that hope to persuade himself that it is.
    It’s then left for you to examine when and how expectations alone can cause physiological change – if that is one of your purposes for the experiment – and if it wasn’t, it should have been.
    But it won’t be if you didn’t understand that expectations are the key to both csuse snd effect, and how those expectations are subject to manipulation.

  105. cwfongon 27 Jan 2009 at 4:58 pm

    As to those other questions you asked in the interim, you really should be able to answer them yourself based on what I said about where the cause lies. Such as asking yourself how a machine would administer something if it wasn’t dependent on the intentions of the living being that somewhere in the process had to have programmed it.
    Of course if the machine had evolved to the extent that it could conduct it’s own experiments and /or make medical decisions on it’s own, then it could decide to use the same procedure that its human makers might have, and call or label it the same thing.
    But since you have failed to understand that all life’s motivational strategies are produced by calculative mechanisms, you won’t entertain the possibility that a machine could ever carry out such strategies on its own.

  106. daedalus2uon 27 Jan 2009 at 6:20 pm

    I already said that to me, each scenario occurred via the placebo effect. It is by the placebo effect when the clinician is being honest, when the clinician is being deceitful, when the clinician is a machine and has no mental state. You are saying that to determine if pain relief occurred via the placebo effect we need to know the detailed psychological state of everyone involved, and that if a machine is involved we need to know if the machine is self-aware or how it was programmed? I could imagine a machine using a random or non-deterministic method for determining injection parameters for example volume of saline.

    OK, but if it didn’t occur via the placebo effect, by what effect did it occur by? I am not asking for the mechanism, just a name to call it.

    Do we simply list them as “pain relief mechanism A”, “pain relief mechanism B”, “pain relief mechanism C”, and so on?

  107. cwfongon 27 Jan 2009 at 7:36 pm

    List them anyway it makes sense to you, because you persist in calling them placebo effects whether there was an actual placebo involved or not. By my definition a placebo effect can only come from a placebo or placebo like cause. A placebo is a substitute for a cause that appears to be the same thing but isn’t. It’s not even placebo-like unless the substitution involves deliberate changes in some particulars that the persons involved were not aware of.
    How many ways do I need to say this? I don’t know, but I’m not that interested in finding out.
    I expect or suspect that a person who doesn’t understand that even babies can react differently to deception (think peekaboo), or that there are probably no games that we all play that aren’t based on use of deceptive skill-sets, will not understand the placebo process – which is really run as a sophisticated form of game. How is it a game? If you have to ask, you wont understand the answer.

    In your mind, you have set up questions that you seem to think require answers that disprove my hypothesis. But it’s not just my hypothesis. All you have demonstrated is that you haven’t made a study of the literature on the subject sufficient to ask anything close to the right questions.

  108. daedalus2uon 27 Jan 2009 at 8:39 pm

    If I understand what you are saying, then each of the scenarios I mention could definitely be due to a placebo effect or could definitely be not due to a placebo effect depending on circumstances not mentioned in the scenarios as I outlined them?

    I don’t understand your hypothesis about the placebo effect well enough to even attempt to disprove it. I don’t even understand your definition of the placebo effect.

  109. HHCon 27 Jan 2009 at 9:30 pm

    I think there is confusion in thinking about the interaction between experimental type and the placebo effects. There is the double blind, experimenter and subject don’t know which condition they are assigned. There is blind, subject does not know which condition he/she is in. There is deaf, subject does not hear experimenter. There is dumb, subject reports to experiment drunk.

  110. cwfongon 27 Jan 2009 at 10:00 pm

    Each scenario you mentioned could be, depending on what the patient was told about the procedure that in some fashion appeared to have fooled his brain or his body into reacting to a cause that was in fact other than they were induced to expect would be involved.

    Although in G, it would really be hard to do unless you blindfolded the pstient so he had some room to doubt the nture of the needle.

    And in H, a sham acupuncture needle would likely need to be something that wasn’t actually a needle at all – since it’s my undersanding that the acupuncture’s art is supposedly in the placement, and not a property of the needle itself.

  111. cwfongon 27 Jan 2009 at 10:06 pm

    Also I agree that if the doctor even appeared to be drunk, expectations could be affected even if the pills were real.

  112. daedalus2uon 27 Jan 2009 at 10:49 pm

    A sham acupuncture needle is (as discribed by Dr N above) a dull needle spring loaded into a hollow case, so when the needle is pushed against a site, the needle retracts into the case without puncturing the skin.

    There are some studies that show that sham acupuncture works better than real acupuncture. My interpretation is that the real acupuncture causes a slight injury and so activates the fight or flight system. Sham acupuncture by not puncturing the skin is a better placebo than real acupuncture.

  113. cwfongon 27 Jan 2009 at 11:21 pm

    I suppose that would also be because real acupuncture is not a placebo for itself.

    And to be a bit more serious, the “real” acupuncture is nevertheless an inherently deceptive procedure to the extent it can depend on the ignorance of one or both of the participants. So the sham acupuncture becomes a deceptive way to expose the deceit endemic to the entire practice. Realistically it’s meaningless to designate one such procedure as a placebo for an almost equally bogus other.

  114. cwfongon 28 Jan 2009 at 4:13 am

    I probably should leave well enough alone, but I doubt that the effect of the acupuncture needle has anything to do with the the fight-or-flight response (also, according to Wikipedia, called the fright, fight or flight response, hyperarousal or the acute stress response). The feeling of the prick (reminds me of an old Bob Hope joke about vaccination) is expected by the patient and stress or anxiety has little to do with the response. More likely the insertion of the needle triggers the expectations already implanted in the patient’s mind that the discomfort alleviation process has begun.

    Also I doubt that the sham acupuncture would “work” well at all if there hadn’t been previous exposure to the “real” version. That would have been where the expectations were first learned. And if the experimenters tell the test subjects that the newer version will work even better because it’s more subtle, etc., viola, it works better.
    But I’m not sure that in fact there are any significant differences between the real fakery and the substitute version.

  115. cwfongon 29 Jan 2009 at 7:18 pm

    Here’s an article that says what I was trying to illustrate much better than I could have either said or illustrated:

  116. aquademiaon 02 Feb 2009 at 12:38 am

    That’s quite interesting about Botox and migraine prevention. Is there any evidence to support a similar beneficial effect for tension headaches? It would make sense, physiologically, and I was wondering if any randomized control studies have been conducted.

  117. bitterhopon 12 Feb 2009 at 1:39 am

    In medical school, one of my instructors admitted during a lecture that, when he provided medical care he also “cheer-led” his patients so as to take advantage of the placebo effect. Presuming you are providing evidenced-based treatment, is there any ethical reason not to do this?

Trackback URI | Comments RSS

Leave a Reply

You must be logged in to post a comment.