Multivariate statistical analysis methods to investigate interindividual glucose dynamics for subjects with type 1 diabetes mellitus

Abstract

This paper investigates the interindividual variability of underlying glucose dynamics using multivariate statistical analysis methods for subjects with type 1 diabetes mellitus. Here two types of glucose dynamics are defined, the general dynamics and the output-relevant predictive dynamics. The concerned important issues are whether the underlying glucose dynamics change from subject to subject? Can a global (or universal) empirical model be developed from glucose data for a single subject and then used to explain the glucose dynamics for other subjects? These and related issues are investigated using multivariate statistical analysis methods based on clinical data for two groups of subjects. Together, these findings provide insights into more efficient development of data-driven models to understand and capture the glucose information in diabetes subjects.