That’s what I don’t get and would love someone to post full text of the study. They claim this is a “modifiable” risk. I find that claim dubious and would love to see how they support that, and why this isn’t simply descriptive/correlational.