More on Northern vs. Southern Europeans study

In my cursory reading of the paper from this previous post, I missed this, which I later found in a news story from the Texas A&M Anthropology in the News website:

They looked at SNPs associated with rheumatoid arthritis and found that, when they corrected for ancestry, several of the genes that were previously believed to be good candidates for being involved in the disease were no longer candidates at all. They also corrected for ancestry in a data set looking at lactose intolerance. "When we did not control for differences in population structure, we got a lot of false associations," Seldin explained.