I wanted to have another look at this paper in light of the recent CPET study that reported no difference.
The point of the 2 day CPET is to show a difference between the Vo2 and workload on day 2 of the test. That decrease is meant to represent the impact of PEM.
However, this paper does not address the change in CPET results between the days at all as far as I can see. The paper focuses on the differences in Vo2 and workload, but not on the differences in the changes between test 1 and 2. I feel like I have to be missing something because that makes the results completely uninteresting... The vast majority of the significant differences between healthy controls and pwME in test 2 were also present in test 1. So all this shows is that the controls were somewhat fitter than pwME, which is not surprising and is just a reflection of poor control selection. I honestly can't understand why they wouldn't have commented on the differences between the responses to the CPETs.