A New Model to Estimate Bolus Insulin Need

Abstract

Background:

Patients with type 1 diabetes usually use the carbohydrate (CHO) counting method to establish their bolus insulin need. However, most still struggle with blood glucose control. We believe that for good control their insulin requirements should strive to mimic the insulin secretion of those without diabetes. The objective here is to develop, from first principles, a better model than the CHO counting model for calculating the bolus insulin need of subjects without diabetes. Such a model may also provide better blood glucose control for patients with type 1 diabetes. This will be investigated in a future article.

Methods:

Equations for metabolism of CHO and for insulin response were derived from first principles. Clinical trials were used to verify these models. The final results—namely, the new bolus insulin requirement equations—were verified using clinical trials by other researchers (Wolever and co-workers). Their methods used are described in their articles given in the list of references.

Results:

The postprandial insulin secretion relationships resulted in an average Pearson R ² of 0.807 (for the new method) versus the old method's R ² of 0.562 (CHO counting).

Conclusions:

The newly derived equation provides a better approximation than the CHO counting method of insulin secretion due to metabolized blood glucose energy from ingested carbohydrates for those without diabetes. We believe that insulin dosage requirements for a patient with type 1 diabetes should mimic the insulin secretion of those without diabetes. If this is true, it means that the new equation should also estimate bolus insulin need for a patient with type 1 diabetes more accurately than before. This will be investigated in a future article.