A Simplified Assessment of Insulin Secretion without Measurement of C-Peptide Provides an Evaluation of Functional Beta-Cell Mass

Research Article

Austin Diabetes Res. 2016; 1(2): 1010.

A Simplified Assessment of Insulin Secretion without Measurement of C-Peptide Provides an Evaluation of Functional Beta-Cell Mass

Brun JF¹*, Villard O1,2, Wojtusciszyn A2, Berney T3, Benhamou PY4, Raynaud de Mauverger E1, Fedou C1, Cristol JP5 and Mercier J1

1PHYMEDEXPInserm U1046-CNRS UMR 9214– Université de Montpellier Physiologie et Médecine Expérimentale du coeur et des muscles-PHYMEDEXP “CHU Arnaud de Villeneuve34295 Montpellier cedex 05, France

2Department of Endocrinology, Diabetes and Nutrition, France

3Department of Surgery, Cell Isolation and Transplant Center, Switzerland

4Inserm, U1055, Grenoble, France

5Department of Biochemistry, Université de Montpellier I, Montpellier, France

*Corresponding author: JF Brun, PHYMEDEXP Inserm U1046-CNRS UMR 9214-Université de Montpellier « Physiologie et Médecine Expérimentale du coeur et des muscles-PHYMEDEXP “CHU Arnaud de Villeneuve 34295 Montpellier cedex 05, France

Received: November 23, 2016; Accepted: December 22, 2016; Published: December 27, 2016


Calculation of pre-hepatic insulin secretion ISR(t) during a standardized meal uses the measurement of C-peptide Cp(t). Serum insulin concentration I(t) is less reliable for calculating ISR(t), due to the variability of its clearance (n) and its hepatic extraction (F). We propose a simplified prediction of ISR(t) using only I(t) and theoretical values of n and F. After a standardized breakfast test (495 kCal, 76 g of carbohydrates), ISR(t) was calculated from Cp(t) according to can Carter’s bicompartmental model and from I(t) by the formula ISR(t)=[dI(t)/ dt+nI(t)]/F with n=0.36 and F=0.45. In 396 subjects representing the whole range of glucoregulation values of ISR (t) given by insulin and C-peptide were well correlated (r=0.842). The coefficients of variation of beta-cell function parameters are 2 to 3% with the C-peptide model and 5 to 6% with the insulin model. In islet-transplanted patients the transplanted beta cell mass is well correlated to beta-cell function parameters calculated with the insulin model (beta cell sensitivity to glucoser=0.538 p<0.01; index Φ(oral) of Cobelli’s model r=0.740 p<0.001). Overall, the simplified approach with insulin alone provides a reasonably acceptable approximation of the results of the reference technique using C-peptide, in spite of a twofold lower intra-subject reproducibility. It provides an evaluation of the beta-pancreatic functional mass. We propose this easy to use model as a reasonable alternative to the reference method, when C- peptide values are not available.

Keywords: Insulin secretion; Model; C-peptide; Beta-cell mass


Due to the current development of diabetes treatments targeting insulin secretion and to the progress in pancreatic islets transplantation [1], it becomes interesting to precisely assess insulin secretion in various clinical situations. For this purpose, several sophisticated procedures developed over the last decades are available [2,3], based on the classical modeling of pre-hepatic insulin release from C-peptide kinetics after a glucose oral load. Such measurements have been shown to provide a prediction of the transition from prediabetic states to overt diabetes [4,5] and to reflect functional islet mass [6].

What makes such approaches very attractive is that, despite the complexity of calculations, they are in fact based on a procedure which is technically very simple, i.e. OGTT or meal test with repeated venous blood samplings over 3 or 4 hrs. However, this approach remains underutilized, due perhaps to the sophistication of the mathematical procedures used for analyzing the crude data and also to the need to assess C-peptide which slightly increases its cost compared to the usual OGTT procedure.

Currently, most studies presenting insulin secretion data employ much simpler measurements for example areas under the curve, insulinogenic index [7,8] or surrogates such as the Homeostasis Model Assessment Beta-cell index (HOMA B) [9,10] whose validity is often discussed [11-13].

We made the hypothesis that between the highly sophisticated mathematic approaches using C-peptide indicated above and the very simplistic indices most commonly used (that are only based on insulin concentrations), a modeling approach derived from the C-peptide based methods but using only insulin data can be reasonably accurate, cheaper and easier to employ and provide much more information on beta-cell function than the more simplistic procedures.

In a classical paper by Tura, et al. [14] serum insulin concentration I(t) is described as a very simple function of prehepatic Insulin Secretion Rate (ISR) and the derivative of I(t) with two parameters n (systemic insulin fractional clearance) and F (posthepatic insulin fractional appearance). We used this equation in order to calculate prehepatic Insulin Secretion Rate (ISR) from serum insulin concentration I(t) and the derivative of I(t), as explained below. We then assessed the reproducibility of the method, its agreement with values calculated with the C-peptide method and finally the correlation of the index of second phase insulin release it yields with beta pancreatic mass.

Subjects and Methods


Three samples of subjects were studied, see (Table 1). We first used a database of 396 subjects undergoing for whatever reason a breakfast test with complete assessment of glucoregulation with the C-peptide models and further analysis with Cobelli and Mari’s models (see below). Another sample of 9 subjects underwent the breakfast test in triplicate in the context of a trial in which they were control subjects and thus provide a possibility to assess reproducibility. Another sample, 7 type 1 diabetic islets transplanted subjects were explored with the breakfast test in order to test the correlation between the number of transplanted islets and insulin secretion (islet mass infused ranged from 9619 to 16304 islet equivalent/kg). This sample of patients was previously studied with the C-peptide model and with this model it was shown that indexes of phase 2 insulin release provided by both the models of Mari [2] and Breda and Cobelli [3] were well correlated to beta cell mass. Data and results of this study obtained with the C-peptide model are presented elsewhere. All studies were performed according to the ethical standards with informed consent and ethical committee advice. The study used for reproducibility was a randomized controlled trial (“DialphaDCT1401”) registered in the EU Clinical Trials Register (Eudract) under the N° ID-RCB: 2014- A00692-45.