Sorry, I can't provide you with any good hypothesis. I'm just looking at what is known assuming that if something is of statistical significance without an obvious cause, that there is probably something going on that we don't see. Yet, correlation is no proof auf causation.