Reply to TMS Wolever et al.

Dear Editor:

Wolever et al. criticize terminology but do not question the conclusions of our work (1), that the use of the recommended approach to determine the glycemic index (GI) value for a simple food, white bread, results in highly variable individual responses, such that among a group of 63 healthy individuals, the GI value for white bread ranged from 35 to 103. Although this resulted in a mean GI value of 64, classifying it as a medium-GI food, it is important to note that only for 23 volunteers did the GI value for white bread fall within the medium-GI range (56–69). For the remaining 40 volunteers (63% of our study population), blood glucose responses to the carbohydrate in white bread were between GI values of 35 and 55 (n = 22) or 70 and 103 (n = 18) for the blood glucose response to pure glucose, thus classifying white bread as a low-GI food or high-GI food, respectively. On the basis of these data we concluded that labeling foods with a single GI value was not useful, and for some individuals could be misleading. This substantial variation in GI value observed in our study is consistent with that reported by previous researchers (2–6).

Nonetheless, Wolever et al. continue to attempt to justify the inclusion of GI values on food labels on the basis that the “margin of error is ±17% of the mean; less than the ±20% difference which is legally allowed between the measured macronutrient or fiber content of foods and the values on the Nutrition Facts Table….” For our volunteers who under the most rigorous and standardized testing conditions had a postprandial hypo- or hyperglycemic response to white bread, is it ethical to recommend that they depend on published GI values for other foods that are based on an n = 10 and hope that it applies equally to them?

Our data also challenge the accepted dogma, that “GI values represent the inherent property of the food and not the metabolic response of an individual to the food.” The significant contribution of baseline glycated hemoglobin concentrations and insulin index (both P < 0.0001) to the variability in GI value determinations clearly shows that longer-term glycemic control and insulin response, even in normoglycemic individuals, affect GI values.

Finally, as a point of clarification, their statement, “GI is widely recognized to have clinical (2–4) and public health (5) significance” is an overstatement of the current status of GI on clinical outcomes and as a component of dietary guidance. Findings of randomized controlled clinical trials are mixed, with the most recent trial (7) concluding that “using glycemic index to select specific foods may not improve cardiovascular risk factors or insulin resistance.” With regard to recommendations, although a few do include a reference to GI, many national dietary guidance documents, such as the recently released 2015–2020 Dietary Guidelines for Americans (8), and health advocacy organizations, including the American Heart Association/American College of Cardiology guidelines (9), did not include GI as a component of their documents. In addition, the current American Diabetes Association recommendations are to focus on total dietary carbohydrate, rather than GI (10). We agree that instead of focusing on the GI, following basic principles of healthy eating (emphasizing whole grains, legumes, fruit and vegetables, fish, and vegetable oils and limiting red meat, full-fat dairy products, and beverages high in added sugars) along with portion control and exercise are better ways to manage and control blood sugar. The time has come to recognize that one size does not fit all.