I've read that there are several SNPs associated with increased risk of clinical or morbid obesity. I was wondering if there is any evidence that these are under positive selection. Would you expect them to be so?

I guess I'm interested to know if they would previously offered a selective advantage and only recently, say in the post-war period in the West morbid obesity become a problem due to differences in the diet.

Related question: How many, and how severe, are known single gene polymorphisms for obesity?