Researchers surprised that whole milk might be healthier than skim

My favorite milk lately.

A recently published study analyzing the ECLS-B dataset found that children who drink whole milk are less likely to be overweight than children who drink skim.  The researchers said that they were surprised by their finding, as they had expected the opposite, and several researchers poo-pooed the findings, suggesting that probably fat kids were put on low-fat milk, rather than vice-versa. 

Which surprised me, given how surprised researchers in a 2005 study were at a similar finding for children.  

I don't see what's so surprising.  A hundred years ago, a livestock fattening manual said, "Skim milk is one of the most valuable adjuncts of the farm for fattening swine" and that the results were particularly good when fed with cornmeal.  There's a prevalent view that a non-sugar cereal such as cornflakes with skim or low-fat milk is a healthy breakfast, but what if it's not?  

I've looked at milk choice in other nationally representative datasets, and I found milk choice heavily confounded by socioeconomic status.  That's what surprised me the most about this study, that the socioeconomic effects didn't overwhelm the effects of the milk.  The highest socioeconomic status groups are most likely to follow prevalent nutrition advice, but they're also likely to be much healthier, making it hard to tell whether the prevalent nutrition advice is counterproductive.