A randomized comparison of three prandial insulin dosing algorithms for children and adolescents with Type 1 diabetes

다국어 초록 (Multilingual Abstract)

To compare systematically the impact of two novel insulin‐dosing algorithms (the Pankowska Equation and the Food Insulin Index) with carbohydrate counting on postprandial glucose excursions following a high fat and a high protein meal.
A randomized, crossover trial at two Paediatric Diabetes centres was conducted. On each day, participants consumed a high protein or high fat meal with similar carbohydrate amounts. Insulin was delivered according to carbohydrate counting, the Pankowska Equation or the Food Insulin Index. Subjects fasted for 5 h following the test meal and physical activity was standardized. Postprandial glycaemia was measured for 300 min using continuous glucose monitoring.
33 children participated in the study. When compared to carbohydrate counting, the Pankowska Equation resulted in lower glycaemic excursion for 90–240 min after the high protein meal (p < 0.05) and lower peak glycaemic excursion (p < 0.05). The risk of hypoglycaemia was significantly lower for carbohydrate counting and the Food Insulin Index compared to the Pankowska Equation (OR 0.76 carbohydrate counting vs. the Pankowska Equation and 0.81 the Food Insulin Index vs. the Pankowska Equation). There was no significant difference in glycaemic excursions when carbohydrate counting was compared to the Food Insulin Index.
The Pankowska Equation resulted in reduced postprandial hyperglycaemia at the expense of an increase in hypoglycaemia. There were no significant differences when carbohydrate counting was compared to the Food Insulin Index. Further research is required to optimize prandial insulin dosing.



This is the first study to compare three prandial insulin dosing algorithms (carbohydrate counting, the Pankowska Equation and the Food Insulin Index).

The study found no significant difference in any outcome between carbohydrate counting and the Food Insulin Index for a high protein or a high fat meal.

While the Pankowska Equation reduced postprandial hyperglycaemia, there was a significant increase in hypoglycaemic events, suggesting some children and adolescents may need clinical modification of this algorithm to avoid hypoglycaemia.

Regardless of the algorithm used participants spent a significant amount of the postprandial period in hyperglycaemia (at least 18.6% of time following a high protein meal and 31.0% of time following a high fat meal) suggesting a need for further research into optimal insulin‐dosing algorithms for meals of different macronutrient content.