This document outlines the methodology behind the calculations of Dietary Glycaemic Index (GI) and Glycaemic Load (GL) as used in the WISP program. Discussions about the validity of calculating these factors for the diets of individuals, based on tables of GI values for individual foods can be found elsewhere in this archive. This document also explains why there are GI/GL options in WISP's recipe calculation system. WISP contains a nutrient databank of approximately 5300 food items. For each of these, a slot is provided for the GI value (range 0-100), where a food containing no carbohydrate would have a zero GI and pure glucose would have a GI of 100. Where possible, values of GI have been taken from published tables (e.g., those from the University of Sydney and other sources). In other cases values have been derived from like foods with similar carbohydrate content.

Work done originally by Flint et al (and later by Wolever et al. at the University of Toronto) proposed that GI tables, when properly used, could form the basis of an evaluation of dietary GI and a mean daily GL. On the basis of the formulae presented, WISP was modified to allow these calculations to be carried out. As there are wide variations in GI values of foods, these results should be taken as an approximation at best and particular care should be employed in the accurate selection of food items.

Wolever also concluded that a previous assertation that protein and fat had a confounding effect in mixed meals was incorrect and that 90% of the glucose and insulin response correlated with the glycaemic load.

The standard formula for calculating Dietary GI is

∑(GI of each food x CHO amount in the quantity of that food eaten)/Mean daily CHO intake

and for Glycaemic Load

∑(GI of each food x CHO amount in the quantity of that food eaten)/100

The sum covers those foods in the diet with a carbohydrate content. WISP calculates these values and displays them as part of the UNA screen.

Whilst this process works for foods which have a GI value, WISP also allows users to incorporate the analyses of recipes into dietary nutritional determinations. The analysis results of recipes are saved as 'custom' foods in the main nutrient databank. These foods can then be used in intake analyses. It was therefore necessary to identify a GI value for a recipe. This is done in an identical manner to the dietary formula above, with the exception that the divisor is in this case the total carbohydrate in the serving of the recipe.

As the tables do contain missing GI values for some foods, a decision was taken to evaluate the carbohydrate proportion in the recipe for foods where no GI was available. If this proportion exceeds a threshold, an alarm is shown on screen (the yellow warning triangle in WISP's recipe entry/edit and Scratchpad functions). The current theshold is set at 20%.

Following the calculation of the GI of the recipe, there were requests to provide the GL as well. As the GL is not used in WISP's intake analysis process and is dependent on the portion size chosen, it is provided as an on-screen result only. The calculation of the recipe GL is simply the GI value x the recipe total carbohydrate/100. It is provided for those users wishing to have the figure, though the scientific relevance of it may be debatable. It can, at least, be used for relative comparison between recipes, perhaps to see which had a lower/higher GL and to categorise them as low, medium or high GL.

Thomas MS Wolever, Ming Yang, Xiao Yi Zeng, Fiona Atkinson and Janette C Brand-Miller
Food glycemic index, as given in Glycemic Index tables, is a significant determinant of glycemic responses elicited by composite breakfast meals
Amer J Clin Nutr 2006; 83, 1306-1312

Flint A, Møller BK, Raben A, Pedersen D, Tetens I, Holst JJ, Astrup A.
The use of glycaemic index tables to predict glycaemic index of composite breakfast meals.
Brit J Nutr 2004; 91:979-989.