In addition to the methodological errors cited in the article, the study in question was published in 2014, which means it included subjects undergoing ‘treatment’ even earlier, which I expect was a time period well-before kids started presenting with this issue as a consequence (at least in part) of the social contagion that made these matters even more prevalent. To the extent that social contagion was a significant factor in inducing kids with other underlying mental health problems to undergo these ‘treatments’, especially those in their early teens, the results in their cases are likely to make such an approach look even less justifiable.