Agree with this observation. This might simply be a burn-off effect. Consider a simple model. Fact one: Obesity is disproportionately concentrated in the less-e

Agree with this observation. This might simply be a burn-off effect. Consider a simple model. Fact one: Obesity is disproportionately concentrated in the less-educated. Fact two: Urbanization is disproportionately concentrated in the highly-educated. Well, what we would see over time is faster rising obesity in rural areas than urban areas. But that's simply a manifestation of the sorting effect. The underlying populations (the educated vs. uneducated) have not fundamentally changed, just relocated. Without considering this we might be spuriously fooled into misunderstanding the causality of obesity. Under this model, rural living itself has absolutely nothing to do with rising obesity. It's merely serving as a proxy for a more important variable.