Undermodeling affects minimal model indexes: insights from a two-compartment model

Area of Science:

• Metabolic Physiology
• Mathematical Modeling
• Endocrinology

Background:

• Minimal models are crucial for studying glucose metabolism using intravenous glucose tolerance tests (IVGTT).
• Previous work highlighted the limitations of the single-compartment assumption in these models.
• Re-evaluation of minimal model assumptions is needed.

Purpose of the Study:

• To assess the impact of single-compartment undermodeling on glucose metabolism indexes.
• To compare the sensitivity of classic (cold) and labeled (hot) minimal models to undermodeling.
• To examine the relationship between minimal model indexes and glucose clamp measurements.

Main Methods:

• Utilized a two-compartment model to simulate glucose kinetics.
• Performed theoretical analysis and simulation studies.
• Compared indexes derived from minimal models with those from glucose clamp techniques.

Main Results:

• Single-compartment undermodeling affects classic model indexes (SG, SI) more than labeled model indexes (SG*, SI*).
• The glucose effectiveness index (SG) is particularly sensitive to early IVGTT events and glucose pool exchange.
• Physiological interpretation of SG requires careful consideration of its local descriptive nature.

Conclusions:

• The single-compartment assumption in minimal models can lead to significant inaccuracies, especially for classic model indexes.
• Labeled minimal models demonstrate greater robustness against undermodeling.
• Researchers should exercise caution when interpreting classic minimal model indexes, particularly SG, due to potential confounding factors in early IVGTT data.