The glycemic index is a measure of how high the average eater's blood sugar rises after eating a particular food item. Many fad diets call for avoiding foods with a high glycemic index.
 How it's calculated
The glycemic index of a particular foodstuff is calculated by making test subjects eat an amount of the foodstuff that contains 50 grams of total carbohydrate. The subjects' blood glucose is then measured at regular intervals over the next two hours, and the glycemic index is derived from the size of the response curve.
Pure glucose is defined as having a glycemic index of exactly 100. Other food items are scaled in comparison to glucose; an item which resulted in half as high a blood sugar response as eating pure glucose, for example, would be assigned a glycemic index of 50.
 Problems with using the glycemic index
Different subjects can respond to the same food item differently. The same subject can respond to the same food item differently at different times of day. Slight variations in the makeup of a particular food item can cause marked differences in the measured glycemic index; one batch of potatoes can be measured to have a moderately high G.I., while another batch of the same variety of potatoes can be measured to have a very high G.I..
 Portion size
The glycemic index isn't scaled for a reasonable portion size. As stated above, the subjects must eat an amount of the foodstuff that contains 50 grams of total carbohydrate. For table sugar, this is about 2 ounces; for carrots, this is a pound and a half, and nobody wants to eat a pound and a half of carrots except Bugs Bunny.
This means that foods with a low carbohydrate concentration will, ironically, often have a high glycemic index. Carrots have a G.I. of 71, which is about the same as white bread. This can lead otherwise rational individuals to steer clear of foods that would have little to no impact on their blood sugar if eaten in a normal portion size.
A more sane quantity, called the glycemic load, is scaled for portion size. However, glycemic load data are more difficult to come by than glycemic index data, which is why you see so many more glycemic index fad diets than glycemic load fad diets.
 It's for Type 2 diabetics
The glycemic index and glycemic load were designed to reduce the "glucose spiking" that can cause problems for people with Type 2 diabetes. They were not designed as weight loss tools, and the evidence linking low-GI or low-GL foods with successful dieting practices is spotty at best and nonexistent at worst.
- The University of Sydney's Glycemic Index database.
- Harvard medical school's list of 100 foods by G.I. and Glycemic Load
- ↑ Freeman, Janine (September 2005). "The Glycemic Index debate: Does the type of carbohydrate really matter?". Diabetes Forecast. Archived from the original on February 14, 2007. http://web.archive.org/web/20070214135006/http://www.diabetes.org/glycemic-index.jsp.
- ↑ http://www.straightdope.com/columns/read/2326/is-the-glycemic-index-the-new-key-to-healthy-eating