Assessment of β-cell function during the oral glucose tolerance test by a minimal model of insulin secretion

Objective To characterise the performance of beta -cell during a standard oral glucose tolerance test (OGTT). Design Fifty-six subjects were studied. A minimal analogic model of beta -cell secretion during the OGTT was applied to all OGTTs (see below). The amount of insulin secreted over 120' in response to oral glucose (OGTT-ISR; Insulin Units 120'(-1) m(-2) BSA) and an index of beta -cell secretory 'force' (beta -lndex; pmol.min(-2).m(-2) BSA) were computed with the aid of the model. In protocol A, 10 healthy subjects underwent two repeat 75 g OGTT with frequent (every 10(/)-15(/)) blood sampling for glucose and C-peptide to test the reproducibility of OGTT-ISR and beta -Index with a complete or a reduced data set. In protocol B, 7 healthy subjects underwent three OGTTs (50, 100 or 150 g), to test the stability of the beta -Index under different glucose loads. In protocol C, 29 subjects (15 with normal glucose tolerance, 7 with impaired glucose tolerance and 7 with newly diagnosed type 2 diabetes) underwent two repeat 75 g OGTT with reduced (every 30' for 120') blood sampling to compare the reproducibility and the discriminant ratio (DR) of OGTT-ISR and beta -index with the insulinogenic index (IG-Index: Delta Insulin (30' - Basal)/Delta Glucose (30' - Basal)). In protocol D, 20 subjects (14 with normal glucose tolerance, 5 with impaired glucose tolerance and 1 with newly-diagnosed type 2 diabetes) underwent a 75 g OGTT and an intravenous glucose tolerance test (IVGTT) on separate days to explore the relationships between acute (0'-10') insulin response (ATR) during the IVGTT and beta -index and OGTT-ISR during the OGTT. Results In all protocols, the minimal analogic model of C-peptide secretion achieved a reasonable fit of the experimental data. In protocol A, a good reproducibility of both beta -index and OGTT-ISR was observed with both complete and reduced (every 30') data sets. In protocol B, increasing the oral glucose load caused progressive increases in OGTT-ISR (from 2.63 +/- 0.70 to 5.11 +/- 0.91 Units.120'(-1).m(-2) BSA; P < 0.01), but the <beta>-index stayed the same (4.14 +/- 0.35 vs. 4.29 +/- 0.30 vs. 4.30 +/- 0.33 pmol.min(-2).m(-2) BSA). In protocol C, both OGTT-ISR and beta -index had lower day-to-day CVs (17.6 +/- 2.2 and 12.4 +/- 2.4%, respectively) and higher DRs (2.57 and 1.74, respectively) than the IG-index (CV: 35.5 +/- 6.3%; DR: 0.934). OGTT-ISR was positively correlated to BMI (P < 0.03), whereas <beta>-index was inversely related to both fasting and 2 h plasma glucose (P < 0.01 for both). In protocol D, <beta>-index, but not OGTT-ISR was significantly correlated to ATR (r = 0.542, P < 0.02). Conclusions Analogically modelling <beta>-cell function during the OGTT provides a simple, useful tool for the physiological assessment of beta -cell function.