Often, a strong expectation that a person will do well on a particular regime is enough to cause that person to do well on that particular regime. And I'm talking about measurable, physical changes.

I'd like to see some proper triple-blind studies involving people who haven't strongly invested themselves in the outcome.