Available carbohydrates have a crucial role in human subjects' diet, providing the majority of energy metabolites. In 1980, the glycaemic index (GI) concept was developed as a tool to compare foods in their ability to provide glucose to the blood circulation after ingestion and absorption in individuals. Studies have shown a relationship between GI and non-communicable metabolic diseases such as type 2 diabetes(1). However, measuring the glycaemic response in vivo is time-consuming, expensive and requires the participation of human volunteers(Reference Jenkins2). The aim of the study was to investigate the relationship between GI and macronutrient composition using statistical methods, and to test the hypothesis that GI can be predicted from composition data without the need for human volunteers. The relationship between GI and macronutrient composition was investigated in twenty foods from the cereal and legume groups using multiple regression analysis methods. The results indicated that starch and protein were the only macronutrient that correlated significantly with GI values (Pearson=0.523 P<0.05, and Pearson's r=−0.513, P<0.05, respectively). A model was established correlating GI to protein and starch content (expressed in g/50 g available carbohydrate portion) as GI=54.201−2.304 (protein)+0.731 (starch). The equation closely predicted GI values of cereal based foods, but was not suitable for legume-based foods (see table below). Future work will concentrate of confirming the predictions using in vivo GI measures.
Predicted GI values calculated regression equation compared with published GI values(Reference Aston, Jackson and Monsheimer3).
