Physiological mechanisms and observed health impacts related to the glycaemic index: some observations.

30 Suppl 3:S72-8.

This paper discusses common misconceptions about the glycaemic index (GI). The rate of carbohydrate digestion is only one of the many determinants of GI and, hence, in vitro methods cannot reliably predict the GI of a food. GI has been criticized as being extremely variable between subjects and not applying in mixed meals. In fact, however, GI controls for differences in glycaemic response between individuals and apparent differences are due to within-individual, variation which does not detract from the ability of GI to predict average glycaemic responses. The conclusion that GI does not apply in mixed meals is based on flawed methodology; recent studies show that nearly 90% of the variation in glycaemic response of realistic mixed meals can be explained by differences in carbohydrate content and GI. The health benefits of low GI foods tend to be ascribed to their slow rate of digestion and reduced post-prandial insulin responses. While the former is important, it is not the only mechanism involved, since low GI sugars may have some of the same effects as slowly absorbed starch. The role of insulin is questioned because, unlike GI, the insulinaemic index (II) of foods may vary in different subjects and, hence, may not be a valid measure. It is not clear that high insulin exacerbates obesity; indeed, there is good evidence that hyperinsulinaemia is associated with reduced weight gain. The public health message about GI is usually to avoid refined foods, eat more fruits and vegetables instead of concentrated juice, and have more pasta and less potato. However, with respect to GI, all of these recommendations are wrong and will ultimately cause confusion and rejection of the concept. An alternative, more scientifically accurate message is suggested.International Journal of Obesity (2006) 30, S72-S78. doi:10.1038/sj.ijo.0803496.