SciCurious has written a review of an interesting paper suggesting a correlation between obesity and city vs. non-city life. As usual, the review by Sci is excellent, but I have a comment or two to add.
Having read the review and then the paper, I had to ask if it might be possible to conclude based on the data presentation that "race" (and thus "genetics") underlies the observed effect. This is because of this graph:
The results as depicted here divides the population into black vs white, making it appear that skin color is a major factor. The paper does not make that specific argument…