I have just read a really good account of some studies into the placebo effect. The work was carried out at the Center for Neurobiology of Stress, University of California in Los Angeles (Clin Pharmacol Ther 86(4) 352-354).
The upshot is that they used MRI and PET scans to look at activity in brain in volunteers given a variety of treatments, some of which were placebos. I can't see anything in the article that specifically says these were double blind studies but I suspect they were (the CPT journal is a bit fussy about these sorts of things). There is indeed a cerebral response when placebos are administered and even certain aspects of brain chemistry are affected. For example PET can be used to measure dopamine receptor binding in the brain.
The placebo effect has been known, of course, for a long time but it is interesting to see that some real mechanistic based research is being conducted into the effects.
Now I will play Devil's advocate here and present the following reasoning:
A placebo can be shown to elicit a physiological response
CAMS such as homoeopathy are placebos
Homoeopathy is therefore likely to elicit a physiological response, which shows that it is an active therapy.
Sure, homoeopathy when tested against a placebo fairs no better but that's like testing aspirin against aspirin.
Is there an argument for using the placebo effect as a therapy and if not, why not?
(One more for the road)