This meta analysis needs to be replicated and revised. There are other errors as well, for instance, if you look at the body of evidence and their funnel plot, you can see they did not include studies that were positive, regrardless of the size of the studies (so, see the right lower quadrant of the funnel plot, and indeed, the right half of the funnel, it's practically empty with three studies close to the null line, and two studies included. It's almost like a foregone conclusion.