| This increasing placebo effect over time has been discussed for a number of years now. I'm familiar with it in the context of antidepressant efficacy, which has shown decreased effects over time relative to placebos for the reasons outlined in the piece. There was supposed to be a big study of it jointly between the FDA and pharm companies, but the results were never published. I have this fuzzy memory that the FDA arranged something where the pharm companies would basically open their data vaults under the condition that nothing would be publicly released. The speculation was, at least with antidepressants, that when antidepressants were first gaining traction, that people were unfamiliar with them and skeptical of the idea that a pill would decrease depression, or would do so without lots of side effects etc. As time went on, the theory goes, people developed this schema that "pill equals feeling more happy, less depressed" and so the placebo effect increased. So the increase in placebo efficacy had more to do with changing culturally mediated expectations (marketing?) than anything else. Or at least that was the theory last time I read about it in any depth. This stuff is fascinating to me for all sorts of reasons really and I believe it's really understudied. There's some analogous debates in the psychotherapy literature, for what it's worth. The problem there is that therapy has been shown to work in all sorts of ways, but different types of therapy don't really have differential efficacy, espcially when you control for publication bias etc. But designing a "behavioral placebo" is really difficult, in part because it's so transparent, but also because at some point a behavioral control starts to look like a therapy. You can use drug placebos as controls, but that's not really the same. My guess is the implications of placebo effects extend far beyond medicine into all sorts of other things, into psychosocial phenomena etc. E.g., what about a placebo for quality of living increases? Relationship satisfaction? Etc. |