BACKGROUND
For ideal glycemic control, the pump user should have a meal planning approach that is as precise and flexible as the pump. Counting carbohydrate is simple and works, but is not a perfect system. Many researches indicate that not all carbohydrates create an equal response when it comes to their effect on blood glucose levels. For a better match between the glucose and insulin profiles, the glycemic index as along with counting carbohydrate might be considered. Therefore, we investigated whether the same amount of carbohydrates with different glycemic indices might require different insulin doses. METHODS: Five type 1 diabetics, using portable external pumps, whose basal rates were correctly set to maintain their blood glucose levels with in the target range under 12 hours fasting conditions, were enrolled. 50 grams of 4 carbohydrate containing foods, with different glycemic indices, were administered for 4 consecutive days to diabetic patients in an overnight fasting state. The test foods were rice, apple, milk and orange juice, for which the glycemic indices were 83, 54, 39 and 97, respectively. The insulin requirement for each food was determined so that the blood glucose level reached the target range four hours after eating. RESULTS: The glycemic indices for each food/rice ratio were significantly correlated with the insulin requirement (r = 0.586, P < 0.01). CONCLUSION: The meal-related insulin dose should be changed according to the glycemic index of the meal. Therefore both amount and source of carbohydrate determine the glucose and insulin responses of type 1 diabetic subjects