Another gem from McArdle (at the top of a lengthy, fact-unfilled post about obesity):
But of course, as I pointed out elsewhere, while being rural is correlated with being fatter, it's also correlated with being healthier (though that advantage may be eroding). It's impossible to tease out the countervailing effects, so which should we do?
Ooga booga, so do nothing? If only someone could develop a mathematical study of such issues, (let's call it, statistics), and then develop a form of this study that works to discriminate between various contributing factors (let's call it, multiple regression analysis), and then apply this technique to producing an actual study (e.g., "Dietary intake, exercise, obesity, and noncommunicable disease in rural and urban populations of three Pacific Island communities," or "Obesity and Health Status in Rural, Urban, and Suburban Southern Women"). I'm such a dreamer.