Predicted glycemic index and glycemic index of rice varieties grown in Taiwan

Two cooked brown rice and six white rice varieties were selected for assessing the variations in predicted glycemic index (pGI) determined by using in vitro starch digestion and the glycemic index (GI) determined in vivo. Marked varietal differences in apparent amylose content, dietary fiber content, pGI, and GI were observed. Most of the tested rice samples were classified as medium-GI foods. The varieties Khazar and Taikeng 9 were categorized as high-GI foods when bread was used as the reference. But brown and white rice samples of TRGC9152 and Taichung Sen 17 fell into the low-GI category when glucose was used as the reference. A significant correlation coefficient (r = 0.946) was found between pGI and GI of rice samples by using bread as the reference with a regression equation of GI = 28.778 + 0.717 × pGI (R2 = 0.8951, P ≤ 0.001). Overall, the in vitro pGI measurement is a rapid and useful method to predict the GI of cooked rice samples.