I have several issues with that study, starting with the fact that they left out 63% of the participants for no clear reason, which is even more damning given the fact that the cohort controlled by the authors themselves. Furthermore, their positive results are based on 280 individuals, not the initial sample.