But the fat these drugs stimulate is, in fact, subcutaneous fat. And if subcutaneous fat is really healthy fat, then having more of the fat cells stimulated by these drugs would be a good thing, even if we weighed a bit more as a result.

Young women tend to have a lot of subcutaneous fat. As we age, we lose a lot of this fat, and but we gain fat in our midsection, being transformed from pear shapes into the more male pattern apple shapes. Asian Indians with diabetes often have stick-like thin arms and legs and large bellies. Perhaps the problem here is as much that they don't have enough subcutaneous fat as that they have too much fat around their viscera.

And people with fat-wasting diseases, such as the wasting that often results from some anti-AIDS drugs, often become insulin resistant, an effect that used to be considered paradoxical because insulin resistance was associated with obesity. But perhaps we need enough subcutaneous fat to balance the visceral fat in order to be healthy.

The real question, of course, is how we can change these patterns. When we cut calories, are we losing beneficial subcutaneous fat or visceral fat? I suspect it's the latter. Years ago, whenever I'd go on a diet, I'd notice that my stomach shrank pretty quickly, but it took a very long time to get the fat off my hands or legs. One book says that with a 5 or 10% weight loss, you lose 30% of your visceral fat.

Is there any kind of diet that would encourage the loss of visceral fat instead of subcutaneous fat more than any other type of diet? Does exercise reduce visceral fat more than subcutaneous fat? What can we do to increase subcutaneous fat other than taking a glitazone drug?

There are many unanswered questions here, but this research gives scientists a clue about where to look for exciting new answers.