Analysis of PK/PD risk factors for development of type 2 diabetes in high risk population using Bayesian analysis of glucose-insulin kinetics.

This study was designed to investigate the pharmacokinetics/pharmacodynamics (PK/PD) risk factors preceding the onset of type 2 diabetes using a population-based Bayesian nonlinear hierarchical model to describe the glucose-insulin kinetics. One hundred fifty-two healthy subjects with a family history of type 2 diabetes were recruited. Each subject received an intravenous glucose tolerance test (IVGTT) and the data of glucose and insulin was collected when entering the study. After the test, subjects were followed up to 25 years and further divided into the diabetic outcome group and non-diabetic outcome group according to the follow-up results. A glucose-insulin kinetic model was developed to account for the physiology and molecular biology of the insulin biphasic secretion and glucose-insulin interactions with a minimal structure. The population PK/PD parameters of the two groups were estimated from the proposed glucose-insulin kinetic model. The relationships between the population PK/PD parameters and the diabetic follow-up results were evaluated. A high insulin baseline concentration, a lower maximum insulin-dependent glucose removal and a lower insulin removal rate constant were found associated with the development of type 2 diabetes in the high risk population. The study shows that the very early pre-diabetic pharmacokinetic differences exist and can be helpful for prediction of development of type 2 diabetes.