Comparison of in vitro starch digestibility methods for predicting the glycaemic index of grain foods

BACKGROUND: In vitro starch digestibility tests are useful for the prediction of glycaemic index (GI). However, there are no internationally recognised methods and no one method has been found to be suitable for all food types. This study compared six in vitro methods, using four grain foods, including those with a varied particle size and soluble fibre content. Method variations included using chewing or mincing, mincing with or without amylase and incubation in a restricted versus non-restricted system. Hydrolysis index (HI) values, calculated from the starch digestibility curves and GI prediction equations were used to compare the in vitro results to GI. RESULTS: HI values for five of the six methods ranked all foods in the same order as the GI values. Using a GI prediction equation (predicted GI_HI) the mincing (without amylase) non-restricted method had the smallest standard error of prediction between the predicted GI_HI and GI values. This method was then validated using 14 grain foods and demonstrated a significant correlation (r = 0.93, P < 0.01) between the in vitro starch digestibility and reported GI responses. CONCLUSIONS: The non-restricted mincing method showed good potential as a new in vitro starch digestibility method for predicting GI in grain foods.