Here is something that I am embarrassed I didn’t catch before – the strong correlation between height and obesity in the NHANES data.
Even with a linear relationship the correlation is strong
With a second order polynomial it becomes striking
What makes this particularly interesting is not only how tight the correlation is but that the two are currently negatively correlated in the cross sectional data but I suspect in the past they were positively correlated.
So one very naive interpretation is that you have a single driving force but that it manifests in some as height and in others as increased body mass. I hope to have more to say on this later this week, schedule permitting.