2) The review included 25 trials on exercise training. In comparison, a BMJ review from last year found only 1. The authors were thus extremely lenient in their inclusion criteria.
We scanned these papers, and most were of low quality and published in obscure journals.