The glycemic index: methodological aspects related to the interpretation of health effects and to regulatory labeling.

The glycemic index (GI) is an experimental system that classifies carbohydrates (CHO) and CHO-containing foods according to their blood glucose-raising potential. It is based on the glycemic response following the ingestion of a test food containing a defined amount of available CHO relative to that of an equi-carbohydrate portion of either white bread or glucose. The concept has been extended to mixed meals and whole diets where the GI of the meal/diet is expressed as the weighted average of the GI of each food, based on the percentage of the total mealldiet CHO provided by each food. Over the last few decades, a substantial number of epidemiological and interventional studies have reported beneficial associationsleffects of lower GI diets across a wide spectrum of pathophysiological conditions, including diabetes, cardiovascular disease, obesity, and certain forms of cancer. This has prompted proponents of the GI to recommend its use for dietary planning and labeling purposes. However, the currently recommended GI methodology is not well standardized and has several flaws, which brings into question the strength of evidence attributed to the health effects of low-GI diets. This review focuses exclusively on the methodological aspects of the GI, how they might impact the interpretation of data related to the purported health benefits of low GI diets, and the considerations for the use of the GI in food labeling. In addition, alternative systems for classifying the glycemic effects of CHO-containing foods are briefly discussed.