[x0054]

The original article is using a sample group which is just much too small to make any reasonable statistical correlation, and in it's conclusion is making claims about causation without even being able to prove correlation.