Monte Carlo analysis of a new model-based method for insulin sensitivity testing.

Area of Science:

• Metabolic disorders
• Clinical diagnostics
• Biomedical engineering

Background:

• Insulin resistance (IR) is a significant risk factor for type 2 diabetes and cardiovascular disease.
• Current IR assessment methods like Euglycaemic Clamp and IVGTT are too intensive for routine clinical use.
• Existing low-resolution methods such as HOMA and fasting glucose/insulin lack diagnostic precision.

Purpose of the Study:

• To develop a novel, clinically applicable test for high-resolution assessment of insulin sensitivity.
• To design a protocol that is simple, cost-effective, repeatable, and has a short duration (<1 hour).
• To validate the accuracy and repeatability of the new test against the gold-standard Euglycaemic Clamp.

Main Methods:

• A model-based measure of insulin sensitivity was utilized, showing high correlation with the clamp method.
• A novel test protocol was designed incorporating physiological dosing and a short duration.
• Monte Carlo analysis on a virtual clamp cohort (N=146) was performed to assess accuracy and repeatability.

Main Results:

• The novel test demonstrated high repeatability with a coefficient of variation (CV) of 4.5%, comparable to the clamp (CV=3.3%).
• The test showed significantly better resolution than HOMA (CV=10.0%).
• A high correlation (r=0.98) was observed between the test's measure of insulin sensitivity and the clamp-derived value.

Conclusions:

• The proposed test protocol offers a simple, cost-effective, and highly repeatable method for assessing insulin sensitivity.
• This novel test provides high-resolution data, correlating strongly with the gold-standard clamp.
• The developed protocol is suitable for earlier diagnosis and improved monitoring of interventions for insulin resistance.