A problem with the term “placebo” is that most people, even most researchers and clinicians don't have a good definition of it, and the definition they use changes as they are thinking about it and in different circumstances.

If you choose the wrong definition for the placebo effect, then you can't study the placebo effect properly because the definition of the effect presumes what the effect actually is.

If you choose as the definition “a treatment that does nothing”, then if a treatment does “something”, then it is not a placebo (by definition). The treatments in this study did “something”, they did not “do nothing”. The treatments in this study were not placebos (under this definition).

If you define a placebo to be “a pharmacologically inactive treatment given to a patient who is told it is pharmacologically active” and it does “something”, then again, the treatments in this study were not placebos (under this definition) because the patients were told it was pharmacologically inactive.

If you want to divide treatment modalities into placebos and non-placebos, then because those two divisions are collectively exhaustive (i.e. cover all possible treatments), then your definition of “placebo” and “non-placebo” has to be collectively exhaustive too.

Clearly for placebos to do “something”, there must be physiology that does the “something” that is observed. This “something” is not mediated through a pharmacological or surgical mechanism, if the treatment does “something”, it must be mediated through some other mechanism, but a mechanism which still couples to physiology.

The definition of “placebo” that I like is “a treatment that has positive results where the positive results are not mediated through pharmacology, surgery, or other physical effects”. This has the effect of dividing all treatments first into successful and non-successful treatments and then further dividing the successful treatments into placebo and non-placebo. If you can't characterize a treatment as either a placebo or a non-placebo, then your definitions need to be modified until you can.

What clinicians and patients want are successful treatments. A treatment based on actual pharmacological and/or physical effects that couple to physiology and improve outcomes is more likely to be successful than a treatment that does not.

I think this is a good and valuable study. It shows that people can be given placebo treatments (in this case a pill with no active ingredients), be told it is a placebo, and they still exhibit improvement in symptoms.

This shows how difficult it is to do studies of treatments without active components. This study shows that people getting better is not evidence that a treatment is not a placebo. That people in a treatment leg of homeopathy, acupuncture, Reiki, prayer, etc. get better compared to a no-treatment leg is not evidence that the treatment leg was not a placebo.

I think that this is a really good approach to administering placebos for use in clinical trials. Give everyone in the trial the same mind-body spiel about placebos. It probably will increase the fraction of people in the trial that get better, on both legs. That is a good thing. If there is coupling between a placebo effect of giving a pill, and the therapeutic effect of the actual treatment (and there probably is), then you want to activate both of them as much as possible. This approach does not compromise the effectiveness of the trial. If it is no better than placebo, that difference will show up with a great placebo response as well as with a poor placebo response.

Where it is most important to have a good and robust placebo response is for ineffective treatments, or for treatments that are only a placebo. If a CAM treatment is no better than a placebo, then it is a placebo and can only be used as a placebo, not for actual clinical treatment. If your definition of placebo is something “that works”, then you end up with acupuncture using toothpicks being equivalent to acupuncture with needles. They are equivalent because both are placebos.

In no way does this study justify giving people placebos as treatment in a clinical setting if there are other treatments known to be more effective than placebo. Maybe if there are no known effective treatments, and the practitioner and patient discuss that, and the patient still wants “something”, then maybe the practitioner could give an open label placebo after discussing it with the patient. I don't see this as that different than giving psychotherapy (which under my definition is a placebo because it does not involve pharmacology, surgery or other physical interventions). Psychotherapy works, people do get better. How and why they get better is complex and doesn't involve woo-woo magic, it involves physiology, physiology which we don't yet fully understand.

What we should be doing is trying to understand the physiology of the placebo effect, so that patients can get those benefits without being deceived by clinicians or scammed by quacks.

I have blogged about the placebo effect and its relationship with nitric oxide physiology. I see the placebo effect as the physiology that triggers the “opposite” of the “fight or flight” state. When the “fight or flight” state is triggered, physiology attempts to divert all metabolic resources to “fight or flight”. Anything that does not contribute to “fight or flight” can be turned off so the ATP saved can contribute to fight or flight. The archetypal fight or flight state is running from a bear, where to be caught by the bear is certain death. Any damage or injury short of death is infinitely better than being caught by the bear. That is why in the fight or flight state, organisms can run themselves to death. It is a “feature”. Diverting ATP away from healing during fight or flight may result in survival. Any damage can be healed later. If there is too much damage to be healed, it is no worse than being caught.

It is not that death by exhaustion is better than being caught, it isn't, they are equivalent, but a physiology that can support running itself to death will escape from many more bears than a physiology that cannot. What evolution has minimized is the sum of deaths from being caught, and from running oneself to death. To minimize the sum, the organism has to be capable of both.

http://daedalus2u.blogspo...

Turning off healing may help to escape from a bear. It is not compatible with a long and healthy life. The autonomic processes turned off during fight or flight need to be turned back on. That is what the archetypal placebo, the mother's “kiss it and make it better” does. The mother is telling her child that the child can stand down from its fight or flight state.