The American Diabetes Association nutrition principles and recommendations (1,2) do acknowledge that a number of factors influence the glycemic response to food, including the amount of carbohydrate, type of sugar, nature of the starch, cooking and food processing, particle size, food structure, and other food components (fat and natural substances that slow digestion) as well as the fasting and preprandial glucose concentrations, severity of glucose intolerance, and the second meal or lente effect (1). The question that the task force asked was, is there evidence that chronic consumption of low-glycemic index foods will contribute to improved glycemia in people with diabetes? The concern being that if another layer of complexity (glycemic index) is to be added to food/meal planning guidelines, there should be clear evidence of benefit.
To answer this question, all studies comparing low- and high-glycemic index diets for 2 weeks or longer were reviewed. As can be seen from Tables 1 and 2, the number of studies is limited. Moreover, the design and implementation of several of these studies is subject to criticism, and in none of the studies was the effect of the diets on postprandial glucose concentrations reported.
Clearly, longer and larger studies are needed to evaluate the utility of glycemic indexing. Until such studies are available, use of low-glycemic index diets is not, in our judgment, evidenced based. Recommendations by other organizations do not change this. We do acknowledge that some individuals might benefit from low-glycemic index diets. However, a decision to use such a diet should be an individual one made in consultation with a nutrition counselor.
Endpoint . | Low GI significantly better than high GI . | No significant difference . |
---|---|---|
HbA1c | 0 | 4 [n = 40] (3, 4, 6, 7) |
Fructosamine | 3 [n = 27] (3, 5, 6) | 1 [n = 9] (7) |
Fasting plasma glucose | 0 | 3 [n = 27] (3, 5, 6) |
Endpoint . | Low GI significantly better than high GI . | No significant difference . |
---|---|---|
HbA1c | 0 | 4 [n = 40] (3, 4, 6, 7) |
Fructosamine | 3 [n = 27] (3, 5, 6) | 1 [n = 9] (7) |
Fasting plasma glucose | 0 | 3 [n = 27] (3, 5, 6) |
Data are n. Numbers in parenthesis refer to the reference list. GI, glycemic index.
Endpoint . | Low GI significantly better than high GI . | No significant difference . |
---|---|---|
HbA1c | 1 [n = 16] (9) | 5 [n = 92] (4,6,8,13,15) |
Fructosamine | 3 [n = 41] (10, 11, 14) | 3 [n = 54] (8,12,13) |
Fasting plasma glucose | 0 | 9 [n = 162] (6,8–15) |
Endpoint . | Low GI significantly better than high GI . | No significant difference . |
---|---|---|
HbA1c | 1 [n = 16] (9) | 5 [n = 92] (4,6,8,13,15) |
Fructosamine | 3 [n = 41] (10, 11, 14) | 3 [n = 54] (8,12,13) |
Fasting plasma glucose | 0 | 9 [n = 162] (6,8–15) |
Data are n. Numbers in parenthesis refer to the reference list. GI, glycemic index.
References
Address correspondence to Marion J. Franz, 6635 Limerick Dr., Minneapolis, MN 55439. E-mail: marionfranz@aol.com.