The usefulness of glycemic load is based on the idea that a high glycemic index food consumed in small quantities would give the same effect as larger quantities of a low glycemic index food on blood sugar. For example, white rice is somewhat high GI, so eating 50g of white rice at one sitting would give a particular glucose curve in the blood, while 25g would give the same curve but half the height. Since the peak height is probably the most important parameter, multiplying the amount of carbohydrates in a food serving by the glycemic index gives an idea of how much effect an actual portion of food has on blood sugar level.
A study of weight loss comparing low GL to high GL diets has found no significant differences between the two, indicating that excessive attention to GL within weightloss programs is misplaced and that a range of foods with widely varying GL values can be part of a healthy diet.
Glycemic load for a single serving of a food can be calculated as the quantity (in grams) of its carbohydrate content, multiplied by its GI, and divided by 100. For example, a 100g slice serving of watermelon with a GI of 72 and a carbohydrate content of 5g (it contains a lot of water) makes the calculation 5*0.72=3.6, so the GL is 3.6. A food with a GI of 100 and a carbohydrate content of 10g has a GL of 10 (10*1=10), while a food with 100g carbohydrate and a GI of just 10 also has a GL of 10 (100*0.1=10).
Data on GI and GL comes from the University of Sydney (Human Nutrition Unit) GI database at through
|Food||Glycemic index|| Carbohydrate|
|Glycemic Load||Insulin Score|
|Baguette, white, plain (France)||~95||~50%||~48||-|
|Banana, Mean of 10 studies||~52||~20%||~10||~81|
|Carrots, Mean of 4 studies||~47||~7.5%||~3.5||-|
|Corn tortilla (Mexican)||~52||~48%||~25||-|
|Potato, Mean of 5 studies||~50||~19%||~9.3||~121|
|Rice, boiled white, mean of 12 studies||~64||~24%||~15.4||~79|