Do you have more info? Would be great if you explain how they do demonstrate causality. Not really sure why it is on me to refute a paywalled article that has no indication of being relevant. Please note that the OC who posted the link to the study (and to whom I was responding) specifically said "correlated".