[zedshaw]

> The sample of those who were studied included individuals between ages 55 and 65 who had had any kind of outpatient care in the previous three years.

So they sampled a conveniently sick group of elderly men and then "controlled" for all the things that make them elderly men and that showed that the heavy drinkers lived 2nd longest (not the longest)?

I suspect that they probably had to cook their statistics in order to support their hypothesis. You find that when they have to "control for everything", instead of finding research subjects that they can actually randomize and properly control. They probably had so many variables controlled that it was just bound to fit whatever model they came up with.