By "carrot", I meant "wombat".
And literally the topic of this thread is an article discussing how small sample sizes, bad methodology, and statistical massaging/manipulation (depending on where you want to situate people on the stupid/evil continuum) mean that nutritional studies can't actually be trusted. The evidence that we have is all over the map.