Correlation in Dose-Response to Rapid- and Long-Acting Insulin for People with Type 1 Diabetes

Annu Int Conf IEEE Eng Med Biol Soc. 2022 Jul:2022:2240-2243. doi: 10.1109/EMBC48229.2022.9871574.

Abstract

In diabetes, it can become necessary to switch between pump- and pen-based insulin treatment. This switch involves a translation between rapid- and long-acting insulin analogues. In standard-of-care translation algorithms, a unit-to-unit conversion is applied. However, this simplification may not fit all individuals. In this paper, we investigate the correlation between dose-response to rapid- and long-acting insulin in the same individual, and compare the correlation across individuals. As a measure of dose-response, we estimate the insulin sensitivity in clinical data from 25 subjects with type 1 diabetes. For parameter estimation, we use maximum likelihood with a continuous-discrete extended Kalman filter and Bergman's minimal model. The results show a weak correlation between insulin sensitivity to rapid- and long-acting insulin across individuals. On this sparse data set, the analysis suggests that the standardized unit-to-unit translation between insulin analogues may not benefit all subjects.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Diabetes Mellitus, Type 1* / drug therapy
  • Humans
  • Insulin
  • Insulin Resistance*
  • Insulin, Long-Acting

Substances

  • Insulin
  • Insulin, Long-Acting