Population insulin sensitivity from sparsely sampled oral glucose tolerance tests

Highlights

• Population estimates of insulin sensitivity have been useful to characterize different populations at high risk for diabetes.

• The use surrogate measures of insulin sensitivity, such as HOMA-IR, while popular, remain controversial.

• The population estimates of S_I may be useful in the analysis of retrospective data where sparse sampling was employed.

Abstract

Objective

This work aimed to estimate population-level insulin sensitivity (S_I) from 2-hour oral glucose tolerance tests (OGTT) with less than 7 samples.

Research design and methods

The current methodology combines the OGTT mathematical model developed by Dalla Man et al., with nonlinear multilevel (NLML) statistical model to estimate population-level insulin sensitivity (S_I) from sparsely sampled datasets (3 or 4 samples per subject obtained in 120 min).

To validate our novel methodology of population S_I estimation, we simulated 50 virtual subjects. We simulated 10 observations per subject over 240 minutes. After estimating their S_I using the OGTT model, the virtual subjects were split into two groups, subjects with S_I above the average and ones with below average. Subsequently, the simulated data were analyzed using statistical software and employing a t-test. The mean estimates of population S_I for the two groups of virtual subjects and their respective 95% CI were compared to the estimates obtained with our novel NLML group S_I estimates obtained using the 3 and 4 time points per subject.

To further validate the performance of the novel NLML model, a set of 34 prediabetic and 30 diabetic subjects with T2D was used. As outlined above for the in-silico subjects, differences between the prediabetic and T2D subjects in regard to S_I was assessed using the classical two-stage approach (individual S_I estimation followed by statistical comparison of the two groups). The average estimates obtained with the classical two-stage approach were compared to the group estimated obtained with the NLML approach using 3 (0, 60, and 120 minutes) points per subject obtained in 120 minutes.

Results

Unique and identifiable individual estimates of S_I were obtained for all virtual subjects. In comparison to the subjects with above average S_I (n=25), the subjects with simulated below average S_I (n=25) exhibited significantly lower insulin sensitivity (P<0.001). Our novel NLML population model confirmed these findings (4-point OGTT: P<0.001; 3-point OGTT: P<0.001). In a similar fashion to the one outlined for the virtual subjects, the median insulin sensitivities estimated with the classical two-stage approach were different between the prediabetic (n=34) and T2D subjects (n=32, P=0.004). Using 3 points per subject, our novel NLML model confirmed these findings (P<0.001).

Conclusions

The population estimates of S_I from OGTT data is an effective tool to assess population insulin sensitivity and assess differences that may not be possible when calculating individual S_I or when less than 7 samples are available.

Introduction

Estimates of insulin sensitivity have been useful to characterize different populations at high risk for diabetes. Several methods exist to estimate insulin sensitivity. The gold standard for measurement of insulin sensitivity is the euglycemic-hyperinsulinemic clamp [1]. The minimal model [2] calculated from the frequently sampled intravenous glucose tolerance test (FSGTT) is another methodology that yields estimates of insulin sensitivity with a good correlation to the euglycemic-hyperinsulinemic clamp [3]. However, both the clamp and FSGTT are time-consuming, require multiple samples, and require steady-state for the clamp. Furthermore, the clamp and the FSGTT methods do not estimate insulin sensitivity in the physiologic state. They both measure insulin sensitivity in response to an intravenous rather than an oral stimulus of glucose.

Surrogate measures of insulin resistance, such as the HOMA-IR [4] and quantitative insulin sensitivity check index (QUICKI [5]) are easier to perform but based on fasting levels of glucose and insulin. HOMA-IR and QUICKI show reasonable correlation with euglycemic-hyperinsulinemic clamp in normal subjects, but each has failed to predict insulin sensitivity in subjects with impaired beta-cell function [6]. Thus, the use of these surrogate measures of insulin sensitivity remains controversial.

The oral glucose tolerance test (OGTT) remains the gold standard to diagnose diabetes and prediabetes. Since this is a more physiologic glucose stimulus, surrogate measures of insulin sensitivity using the OGTT have been developed. Underlying genotype-phenotype abnormalities can have different impact on fasting and postprandial glucose levels. The Matsuda index, a composite surrogate measure using 5 points from the OGTT, was developed to account for the heterogeneity of the diabetes phenotypes and have a broader measure of insulin sensitivity [7]. Another measure of insulin sensitivity from the OGTT is the oral minimal model by Dall Man et al [8]. The original model was developed using 20 samples over 300 minutes. A subsequent study showed that a 2-hour 7-sample OGTT correlates well with the 300 minutes 11-sample OGTT measurement of insulin sensitivity [9]. The oral minimal model is widely used. However, in most studies, especially those with large populations, a collection of 5- or 7-samples over 2 hours can be demanding on the research subject and may not be feasible in studies of large populations.

Our investigation aimed to develop a sparse sample method of measuring insulin sensitivity that will enable the use of the Dalla Man’s model in studies where less than 7 samples have been collected. When insufficient samples have been obtained per subject, a common practice in pharmacokinetic/pharmacodynamic (PK/PD) analysis has been to combine the data to yield a population model [10]. Population modeling extends the kinetic mathematical model to include fixed and random effects and, as such, are often known as mixed-effects models and can be nonlinear when the fixed part is not linear and referred as nonlinear mixed effects, NLME, models [11]. NLME models have previously been used to obtain estimates of minimal model parameters using FSGTT simulated sparse sampling data by using 13 samples from the FSGTT [12,13] and with 10 samples for estimation of Si from OGTT data [12]. These multiple samples are not feasible for large populations. Therefore, a quantitative estimate of insulin sensitivity that uses fewer time points during an OGTT and characterizes insulin sensitivity on a cohort level will be useful for studies assessing large numbers of people at high risk for type 2 diabetes.