Minimal-model estimates of insulin sensitivity are insensitive to errors in glucose effectiveness

Area of Science:

• Metabolic research
• Physiological modeling
• Endocrinology

Background:

• The minimal-model estimates insulin sensitivity (S(I)) and glucose effectiveness (S(G)) from glucose and insulin dynamics.
• Previous work indicated minimal-model overestimates S(G) in individuals with normal insulin secretion.

Purpose of the Study:

• To investigate if overestimation of S(G) affects the estimation of S(I) using the minimal-model.
• To determine the independence of S(I) estimation from S(G) accuracy.

Main Methods:

• Analysis of 56 glucose tolerance tests from women of varying weight and PCOS status.
• Constraining the glucose effectiveness (S(G)) parameter in the minimal-model to lower values.
• Examining the impact on the insulin sensitivity index (S(I)) estimation.

Main Results:

• Constraining S(G) led to significant increases in other model parameters (P2, P3).
• Despite substantial S(G) constraint (down to 40% of unconstrained value), S(I) decreased by only 3%.
• Insulin sensitivity index (S(I)) remained largely unaffected by S(G) value adjustments.

Conclusions:

• Minimal-model-derived insulin sensitivity (S(I)) estimation is robust and independent of potential errors in glucose effectiveness (S(G)) calculations.
• The findings support the reliability of S(I) derived from the minimal-model, even with imperfect S(G) values.