GI Testing

Glycemic Index Testing

The GI Foundation SA has done hundreds, probably more than a thousand, Glycemic Index (GI) tests over the past 20 years. In 2010 an international protocol for GI-testing was developed by the GI-testing scientists of the world, including ourselves, and compiled as the ISO26642 document. Along with many other labs, the ISO26642 is the standard the GI Foundation SA complies with, on a voluntarily basis. The ISO document is comprehensive and detailed, but the most important aspects of the GI concept and the actual GI testing, are described here below.

We use the GI as a measure, on a scale from 1 to 100, of how fast a carbohydrate containing food is digested and absorbed in the body. Glucose is assigned a numerical value of 100, and as it is absorbed almost immediately giving a sharp, quick rise in blood glucose levels. All other carbohydrate foods are compared to glucose as the reference food. In some cases white bread is used as the reference food, but the GI-values are calibrated by a factor to bring it on par with glucose reference values.

Testing for the GI of a given food is done on real people, eating real food and taking actual blood glucose readings. A GI test team typically consists of at least ten people to meet that ISO26624 standard. Only healthy individuals are allowed in a test team. Typically a team would be made up of some younger and some older individuals, as well as those that exercise and those who do not exercise. In order for all GI testing throughout the world to be consistent, the portion size has been standardized to 50g glycemic carbohydrate in the food. This means that the portion of food tested must contain exactly 50g of carbohydrate that will affect blood glucose levels after ingestion. For example, should we wish to test apples, each tester would have to eat about three small apples weighing exactly 380g in total, which contains 50g glycemic carbohdydrate.

Before consuming the test food, blood glucose levels are taken to determine the starting point. The food is then eaten within 12 – 15 minutes. Thereafter, blood glucose readings are taken every 15 minutes for 2 hours. All the readings are plotted on a graph where blood glucose concentrations are plotted against time elapsed. The area under the curve (AUC) of the test food is compared to the AUC of the glucose test average of the specific tester, and this gives us the glycemic index. The calculation of the AUC is done in accordance with the ISO26642 standard.

This means that the numerical value obtained is a RATIO value (test food over glucose, the reference food), and that is why it is relatively constant in ALL individuals. This is because one is comparing the test food, to glucose in the same individual.

However, sometimes it is impossible to test the GI of a food e.g. chutney or peppadews, since testing chutney would imply that each subject would have to eat about one cup of chutney and testing peppadews would imply that they would have to eat one and a half bottles of peppadews each! Other products that contain very little carbohydrates need to be tested to establish their effect on blood glucose levels. This may mean that the test subjects would have to drink about 1 litre of flavoured mineral water! It is neither fair nor ethical to expect subjects to test these products at all or in these large quantities!

Sometimes it is therefore necessary to opt for the second choice, which is to do a glycemic response (GR) test on about double the regular serving size, especially in lower carbohydrate products or to opt for the third choice, which is to do a theoretical calculation. If either of these methods of GI or GR determination has been used, it will always be stated on the packaging of the food or in the specific GI list on this website. Theoretical calculations are also frequently used to calculate the effect a mixed meal will have on blood glucose levels.

Some foods have not been tested for their GI and therefore we have “estimated” some of them, using GI values we have and doing a theoretical calculation. We always state in the GI lists when we have done so, just to give you an indication of what the GI value could possibly be.