A numerical deconvolution method to estimate C-peptide secretion in humans after an intravenous glucose tolerance...

Quantitative assessment of pancreatic insulin secretion rate in individuals may help advance our understanding and treatment of diabetes. We describe for the first time the application of a long-established numerical deconvolution procedure in which a prescribed input function is used to represent first-phase pancreatic secretion response to an intravenous glucose challenge (intravenous glucose tolerance test [IVGTT]) in individual subjects. We identify that C-peptide secretory response to an IVGTT can be described by a basal secretion rate (S(b)) (picomoles per liter per minute) and a first-phase secretory response characterized by a Gaussian function. The Gaussian function contains 3 parameters: P(1) (picomoles per liter per minute), which represents the peak rate secretion; P(2) (per square minute), which is related to the inverse of peak width at half-peak height; and P(3) (minutes), which is the time of the peak secretion rate. When applied to data from 8 healthy Chinese subjects, the estimated parameter values (mean +/- SD) were 19.2 +/- 12.9 pmol L(-1) min(-1), 1548 +/- 1143 pmol L(-1) min(-1), 1.09 +/- 1.21 min(-2), and 2.94 +/- 1.13 minutes for S(b), P(1), P(2), and P(3), respectively. The Gaussian input functions are shown to have similar shapes and to be highly concordant in magnitude with secretory responses estimated by means of the method of Eaton et al (1980) and by the ISEC computer program. In conclusion, we have presented a simple, integrated, validated, and easily implemented method suitable for quantifying pancreatic C-peptide and insulin secretion in individual human subjects. The superiority of our method in comparison with other methods is that it uses as an input function the Gaussian, which has been experimentally verified as describing in vivo the profile of pulsatile insulin secretion. The particular strength of our method is that the Gaussian parameters and simple indices derived from them provide a standardized and interpretable means for carrying out comparative investigations aimed at quantifying how pancreatic secretory responses to an IVGTT differ in various demographic groups or in response to therapeutic treatments.