Home / Resources / Clinical Gems / International Textbook of Diabetes Mellitus, 4th Ed., Excerpt #86: Measuring Insulin Action In Vivo

International Textbook of Diabetes Mellitus, 4th Ed., Excerpt #86: Measuring Insulin Action In Vivo

Aug 15, 2017

The model

It was a challenge to develop a computer model that could account for plasma dynamics in a compact but accurate package. The model was the simplest representation we could devise, which was based upon known physiology, and which could accurately describe moment-by-moment plasma dynamics. The model continues to thrive in that it continues to be the basis for a large number of clinical investigations (∼50 per year) as well as a robust literature related to its mathematical and computer characteristics [43,60,61,79 – 82,86 – 96].

Practical application of the minimal model

In contrast to the glucose clamp procedure, neither online mChapter15Table15.1easurements nor external control of infusions are necessary to perform the minimal model procedure. Temporary intravenous lines are placed in antecubital veins in a relaxed subject. After basal sampling, glucose is injected and frequent sampling is begun. To facilitate the computer’s ability to estimate metabolic parameters (cf. Table 15.1) insulin is injected at 20 min and sampling continued until 3 h [92,93]. Samples are spun and plasma is stored frozen for later analysis of glucose and insulin. It is straightforward for two experimenters to perform the FSIGT on one subject.

Yields from the minimal model

By computer analysis using the program MINMOD Millennium [97], a set of values is calculated for each individual subjected to the FSIGT test (Table 15.1). Insulin sensitivity from the model is expressed as parameter SI. The latter parameter from the minimal model is fundamentally analogous to the previously mentioned SIP(clamp) estimated from the euglycemic clamp [60]. In fact, the parameters are strongly correlated with each other [60 – 62,98 – 100]. The correlation is not perfect, however. This may be due to inherent errors in measurement of each (errors in plasma measurements, sampling times, etc.). Also, it may be due to the fact that the minimal model SI is estimated from a dynamic test, whereas the analogous clamp parameter is from a steady-state test [82]. It is apparent that these very different tests — the FSIGT versus the clamp — may not be measuring the exact same things, but what they are reflecting are highly correlated. The clamp emphasizes steady-state effects of insulin on glucose uptake and production often at a single insulin level. The minimal model also measures the factors that determine the rates of distribution of insulin as well as glucose among different compartments of the organism, and factors that determine how rapidly insulin can stimulate glucose metabolism by accessing tissues (whether by changing blood flow distribution or endothelial transport) and by mobilizing transporters, activating enzymes, changing expression, and so on. The relative importance of insulin resistance to pathogenesis disease, whether measured by a steady-state or a dynamic approach, is of great interest.

Additional important information is gleaned by using the minimal model approach. One important parameter, also obtained from the euglycemic clamp (by the ratio of the increment in plasma insulin to the insulin infusion rate) is an estimate of the metabolic clearance rate of the hormone. As shown in Figure 15.6, insulin is injected into the patient 20 min after the glucose. The rate of decline in the insulin level after this injection is proportional to the rate of insulin clearance. The metabolic clearance rate of insulin is estimated by the following equation:

Chapt15Part5where MCRI represents the metabolic clearance rate of insulin, Dose reflects the amount of insulin injected at t = 20 min during the FSIGT (in mUkg−1), Ins(t) is the plasma insulin concentration at time t, and Ins(0) is the insulin concentration at t = 0 (=basal insulin, in mU L−1 ).

However, this estimate is imperfect, because the time course of plasma insulin is also modified based upon the change in the endogenous release of the hormone. In our laboratory, we are developing new methods for providing a more accurate estimate of insulin clearance, which utilizes the C-peptide measurements made following the insulin injection at 20 min after glucose.

Click here to view all Chapter 15 references.