Estimating postprandial glucose fluxes using hierarchical Bayes modelling

Area of Science:

• Metabolic Research
• Computational Biology
• Endocrinology

Background:

• Accurate estimation of glucose fluxes is crucial for understanding glucose metabolism, especially in conditions like type 2 diabetes.
• Existing methods, such as maximum likelihood, may produce non-physiological results for endogenous glucose production.
• The meal tolerance test is a key diagnostic tool for assessing glucose regulation.

Purpose of the Study:

• To develop and validate a novel stochastic computational method for estimating glucose production, appearance, and disposal rates during meal tolerance tests.
• To compare the performance of the new method against the maximum likelihood method using both human subject data and simulations.
• To assess the accuracy and physiological plausibility of glucose flux estimates, particularly endogenous glucose production.

Main Methods:

• Development of a Bayes hierarchical model incorporating a prior probability distribution for smooth glucose fluxes.
• Application of the new stochastic method and the maximum likelihood method to data from 18 type 2 diabetes subjects ingesting a mixed meal with [U-¹³C]glucose.
• Validation using simulated triple-tracer experiments in 12 virtual subjects to compare accuracy metrics like root mean square error (RMSE).

Main Results:

• Both methods yielded similar mean estimates for endogenous glucose production, R(a meal), and R(d) in human subjects.
• The new stochastic method provided plausible endogenous glucose production estimates in all subjects, unlike the maximum likelihood method in two cases.
• The new method significantly improved the accuracy of glucose disposal (R(d)) estimates (P<0.01) and showed comparable accuracy for endogenous glucose production and R(a meal) compared to the maximum likelihood method in simulations.

Conclusions:

• The novel stochastic computational method enhances the physiological plausibility of endogenous glucose production estimates during meal tolerance tests.
• This new approach offers improved accuracy in quantifying glucose disposal compared to traditional maximum likelihood methods.
• The findings support the utility of this Bayes hierarchical model for more reliable glucose metabolism assessment in clinical and research settings.