[dkirkman]

This isn't entirely snarky. The ACCORD trial was a spectacular failure (http://www.nejm.org/doi/full/10.1056/NEJMoa0802743). Nobody really knows why (as far as I know), but since high blood glucose is known to cause problems, and lowering it by giving extra insulin does not seem to help (remember that type-2's usually have high spectacularly insulin levels to begin with), we've got to consider other approaches.

Avoiding carbs seems to both normalize glucose and lower insulin in type-2s, so it's not a dumb approach. On the other hand, a low carb diet tends to increase insulin resistance, so if it turns out that both the glycemia and insulin levels are not the cause of the adverse symptoms, but rather symptoms in themselves, going low carb may make everything much worse. (e.g. it might make the numbers we can measure look better, while worsening whatever metabolic abnormality is causing the diabetes in the first place)