What is the glycemic index, and is it useful?

The idea is simple. Two foods with equal carbohydrate don’t move through digestion at one speed. White bread breaks apart fast in the mouth and gut; a legume in its hull takes that road far slower. The glycemic index tries to turn that speed gap into one ranking.

Measurement starts in a lab. Volunteers eat a portion with digestible carbohydrate held constant; blood samples follow at set intervals for two hours. Another day, the same person eats the reference food, most often pure glucose or white bread. Put the areas under the two curves in ratio and the food gets a rank number. The ranking splits three ways: low, medium, high.

A rank number alone doesn’t describe the plate. Glycemic load adds the portion to that same ranking; it weighs the index with the carbohydrate in that portion. Watermelon is a good example. It sits high in the ranking, but one slice carries little carbohydrate. High index, low load.

What shakes the ranking starts at the stove, fridge and plate. The more a food is handled, the further it drifts from the version on the measuring table.

• Cooking: pasta cooked long and pasta kept firm don’t share a rank.
• Ripeness: a greenish banana and a speckled banana behave differently.
• Grinding: the same grain behaves like another food once it becomes fine flour.
• Company: the same carbohydrate moves slower alongside fat and protein.
• Personal difference: even in one person, the same bread doesn’t answer the same each day.

That last item has been measured. In a repeat study with over sixty healthy volunteers, the same white bread’s index swung about a fifth from day to day in one person. Between people the spread was wider still. So what looks like one number on a list is really a wide range.

Still, the measure has a real use. Comparing two foods from one family, it gives a rough idea: a whole-kernel grain and a product from its flour split apart on this scale. The gap between whole-wheat flour and white flour, though, isn’t as wide as people think. What draws the line isn’t the whole-grain label, it’s whether the kernel was ground. Current clinical guidelines call the index only one face of carbohydrate quality, not the sole decision tool. Reviews underline another detail: the low end of the ranking is mostly high-fiber foods. So it’s hard to tell how much of the difference is index, how much fiber. A separate piece takes on the division of labor between amount and type.