Regarding their theory that this might explain the mixed results for programs designed to involve exercise and that some people might need some sort of therapy to persuad them to exercise first: somebody pointed out to me today that he is basically what GET-based CBT and GET are.