I have no idea, but a statistical correlation does not justify the blanket characterization made by the article, where "risk-taker" is essentially held as a synonym for "privileged".
Can you show an example of damage done by this sort of "defamation"?