The new nutrition recommendations that the American Diabetes Association issued last week budged the premier American diabetes association into the 21th Century. It gave its limited -- and probably reluctant -- stamp of approval to low-carb diets.

But the ADA moved little or no distance further in its tenuous support for the glycemic index that it initially gave in the previous nutrition recommendations. It issued the previous recommendations in September 2006. The ADA concluded then that for people with diabetes " the use of glycemic index and load may provide a modest additional benefit over that observed when total carbohydrate is considered alone."

This year the ADA came to exactly the same lukewarm conclusion.

The position statement bases its reservations mostly on "potential methodological problems" that a 2006 review in the British Journal of Nutrition noted. But that article used food frequency questionnaires, which are notoriously inaccurate.

Nevertheless, the ADA acknowledes that several randomized clinical trials reported that low-glycemic index diets reduce blood glucose levels among people with diabetes. But some other clinical trials were not able to confirm it, the position statement says.

The other basis of the ADA's half-hearted endorsement of the glycemic index is that "responses to specific carbohydrate-containing food is a concern." What are those responses?

"The glycemic index does not address other metabolic issues related to...low glycemic index sweeteners, particularly fructose." This phrase comes from the article the position statement cites. It's a 2004 article in Obesity Research, "Carbohydrates and Increases in Obesity: Does the Type of Carbohydrate Make a Difference?"

The lead author of that article is Judith Wylie-Rosett, who happens to be the co-chair of the writing panel for the ADA's nutrition recommendations.

Good point. It's one that I addressed here recently in my article on "The Trouble with Fructose."

While fructose does have the lowest glycemic index of any sugar, high-fructose diets lead us to secrete more insulin, which in turn leads to more insulin resistance. That's because fructose goes to our liver instead of directly into our bloodstream like other sugars. Because the glycemic index does not take the effects of fructose into account, this is indeed a limitation that we need to consider.

But, proponents of the glycemic index have never said to forget everything else we know about diet. "The glycemic index should not be your only criterion when selecting what to eat," I wrote on my main glycemic index page. "The total amount of carbohydrate, the amount and type of fat, and the fiber and salt content are also important dietary considerations."

These seem to be minor limitations to the glycemic index compared with its benefits. In fact, the ADA's position statement on its nutrition recommendations recognize one huge advantage of a low-glycemic diet.