Second of all, the numbers aren't disaggregated, but in the course of researching an article I wrote for Salon I did a little research on the eating habits of the American public. The people most likely to follow the low-fat diet are also the people least likely to be overweight: the affluent. The poor, who ignore such messages and eat an extraordinary amount of high-fat cheap meat in the form of fast food and such, are the most likely to be overweight. They are also the largest factor in the obesity epidemic; their children are extraordinarily more likely to be overweight or obese than the children of the middle class or affluent. So I find it hard to see how the obesity epidemic could be connected to the change in health recommendations.

I do see how it could be connected to the revolution in portion size, which happened around the same time. Pepsi did a very interesting study; they gave some test families unlimited supplies of Pepsi to see how much they would drink in a week. What they found was that no matter how much Pepsi was placed in the house, all of it would be drunk before the next refill. That led to the increase in soft drink size, starting with the 20 oz bottle and the 2-liter. Portions everywhere else started getting larger as well. The actual cost of the food being sold at most convenience outlets is far lower than the costs for land and such; they can offer attractive quantity discounts to customers (only 25 cents more to supersize!) and still make a nicer profit, because most of their cost is fixed; the extra 25 cents is pretty much gravy.

And who responds best to the larger portions, the bigger bottles? The poor, who see such purchases as thrifty, and are unaware of or ignore health messages.