Old Vs New Glycemic Index

If you seem to remember a glycemic index that had some foods with values over 100, you aren’t hallucinating. For quite awhile there were two coexisting glycemic indexes. The other one was based on white bread, rather than glucose. So white bread had a glycemic index of 100, and there were quite a few foods that were higher than 100, including potatoes, rice cakes, and some breads and cereals. On that scale, glucose had a value of 143.

The white bread index is falling out of favor, making glucose the standard. This makes intuitive sense, since 1) it is, after all, glucose in the blood that is being measured, 2) any glucose ingested goes to the blood extremely quickly, and 3) glucose is the same the world over, whereas “bread” is quite variable.

However, I think we may have lost something with the change to glucose, and that is a wider range of index values. The common way to deal with foods tested against bread is to multiply the number by .7. This reduces the range of the values. As I point out in Is the Glycemic Index Useful?, the vast majority of foods tested (which is a small proportion of total foods) have glycemic index numbers in the range of 40 to 70.

Not only do you have a narrow range of glycemic index foods, but each food has a lot of variability. There is variability between people (and even in one person from day to day), between labs where the testing is done, between cooking processes, between varieties of fruits, grains, etc, and between recipes for the various baked goods. Since each food has so much variability, you don’t want them stuffed into such a narrow range on the index. It may make it less likely that a study will get significant results, even if there is a true difference between foods. This would especially be true for studies which divide people into low, medium, and high glycemic diets. Almost by definition you aren’t going to have many people eating low glycemic diets unless they are also low carb.

Glycemic Index Food Lists