I think we have two problems here. One is, as you say, due to the poor definition of GET. It is likely that GET was applied with far more sensitivity in the PACE trial than most people experience it through ME centres and the like. The other is that CBT and GET would benefit a number of healthy people that I know, as it would a wide variety of people with different illnesses. My guess is that people with ME would show less improvement than a random sample of healthy people, simply because I do not believe that these therapies are relevant to ME, and can actually do harm. The problem with measuring the effect though is that many measurements are designed at measuring the ill, and healthy folk tend to clump at the top end. Effects on the 6 minute walking test would be interesting though. Would healthy folk show a greater improvement after a year of walking each day than people with ME? Is that an obvious question? Is it obviously an obvious question?