The paper by Hodge et al. (1) published in the November 2004 issue of Diabetes Care aptly contrasts the potential benefits of moderately high-carbohydrate diets with a low glycemic index (GI) versus diets that have a lower glycemic load (GL) by virtue of a low carbohydrate content. In their prospective analysis of a cohort of ∼36,000 adults followed for 4 years, Hodge et. al found that higher-carbohydrate diets were associated with a lower risk of development of type 2 diabetes. However, the type of carbohydrate was equally important: low-GI carbohydrates reduced the risk, while high-GI carbohydrates increased the risk. Thus, low GI and low GL are not equivalent and produce different clinical outcomes.
Because this issue may be confusing to some readers, it is important to clarify the difference between GI and GL. Both the quality and quantity of carbohydrate determines an individual’s glycemic response to a food or meal (2). By definition, the GI compares equal quantities of available carbohydrate in foods and provides a measure of carbohydrate quality. Available carbohydrate can be calculated by summing the quantity of available sugars, starch, oligosaccharides, and maltodextrins. As defined (3), the GL is the product of a food’s GI and its total available carbohydrate content: glycemic load = [GI × carbohydrate (g)]/100.
Therefore, the GL provides a summary measure of the relative glycemic impact of a “typical” serving of the food. Foods with a GL ≤10 have been classified as low GL, and those with a value ≥20 as high GL (4). In healthy individuals, stepwise increases in GL have been shown to predict stepwise elevations in postprandial blood glucose and/or insulin levels (5).
It can be seen from the equation that either a low-GI/high-carbohydrate food or a high-GI/low-carbohydrate food can have the same GL. However, while the effects on postprandial glycemia may be similar, there is evidence that the two approaches will have very different metabolic effects, including differences in β-cell function (6), triglyceride concentrations (7), free fatty acid levels (7), and effects on satiety (8). Hence, the distinction has important implications for the prevention and management of diabetes and cardiovascular disease. Our concern is that the use of the GL or “glycemic response” in isolation may lead to the habitual consumption of lower-carbohydrate diets.
The simplest way to consume a moderately high-carbohydrate, but low-GI diet is to follow the new 2005 Dietary Guidelines for Americans (9) and to incorporate the recommendations of the World Health Organization/Food and Agriculture Organization (10); that is, the GI should be used to compare foods of similar composition within food groups. By choosing the lower-GI options within a food category (breads, breakfast cereals, etc.), an individual automatically chooses those with a lower GL. Because most fruit and vegetables, other than potatoes, are not major contributors to carbohydrate intake, their GI should not be the basis for restriction.
The important message for clinicians, nutritionists, and food industry professionals is that the evidence, as it stands, suggests that for preventing type 2 diabetes, we ought to encourage low-GI carbohydrate foods but not those that simply have low “net carbs,” low GL, or produce a low glycemic response.
References
T.M.S.W. is president of the Board of Directors of, holds stock in, and has received grant/research support from G.I. Testing. T.M.S.W. also holds stock in G.I. Laboratories.
