For quite some time, social scientists have noted people living in impoverished, racially segregated neighborhoods are more likely to be obese and diabetic than people from integrated, less impoverished neighborhoods (even after adjusting for known predispositions for obesity and diabetes).
Is it the neighborhood that makes people fat and obese? Perhaps poor and segregated neighborhoods—lacking grocery stores and safe streets to exercise on, to start—doom people to obesity and diabetes? Or is it the other way: The sort of person who would be lazy and inept enough to live in a poor and segregated neighborhood lazy and inept enough to be doomed to obesity and diabetes.
A study recently published in the New England Journal of Medicine randomized a few hundred people living in poor, segregated neighborhoods in the United States to receive no housing help, a housing voucher to move and live anywhere they want, or a voucher limited to neighborhoods with low levels of poverty.
Ten years later, the people randomized with vouchers for less impoverished neighborhoods had (statistically significantly) much lower rates of obesity and diabetes as compared to those randomized to receive no voucher to move. Compared to people who received unrestricted vouchers, there was a (statistically insignificant) reduced rates of both obesity and diabetes.
It's the neighborhood.
More amazingly, if you think about this sort of voucher system as a medical treatment—like dialysis or insulin—it's pretty damn cost effective. Only about twenty people needed to move in order to prevent one case of diabetes or obesity. Makes one wish we didn't give up on the War on Poverty.