Pretty sure the causation worked the other way, but it's hardly a repeatable experiment in my case.