I can't read that paywalled article, but from the sounds of the abstract, no methodology is introduced to demonstrate causation. Any correlation that does not in turn show causation is spurious.