Another interesting study destroyed by its PR. When will the madness end? I seriously doubt the two effects are "equal", either; could it be that getting a better night's sleep or two removes the effect of a bad night of sleep, whereas six months of diet can't be removed in a couple of days? Just because two elements of an n-dimensional tuple are equal does not mean you've got two equal values. This study, at least, doesn't appear to address that question at all. (Which is fine! That's just a factual statement, not a criticism. Though I do hope having spent 6 months on feeding dogs they took the ~week to see about recovery times for the insulin sensitivity, perhaps for another paper.)