Mathematical model of in vitro digestion of infant formulas
Résumé
The prediction of the digestion kinetics of infant formulas depending on their lipid and protein composition would be of great interest (i) to limit time-consuming and costly in vitro or in vivo experiments, (ii) to optimize their digestibility and (iii) to predict the fate of newly developed formulas in the gastrointestinal tract.
This study aimed to establish a mechanistic model of the kinetics of proteolysis and lipolysis during the gastric and intestinal digestion of two different infant formulas using an in vitro dynamic simulator with two-compartments: stomach and small intestine1. Experimental data were available from a previous study2.
The model approach was based on mass and volume balance equations on the different proteins ( -lactalbulmin, -lactoglobulin and caseins), triacylglycerides (TAG) and digestive enzymes (pepsin, lipase and pancreatin), assuming the stomach and small intestine as two perfectly stirred reactors displayed in series and at steady state. For each infant formula, a system of 11 differential equations with 6 unknown parameters was implemented in Matlab® software for resolution. These unknown parameters, which correspond to the enzymatic reaction rates, were estimated by finding the minimum of constrained nonlinear multivariable function between the solutions of the mathematical model and experimental data with the fmincon function.
The mathematical model will be extended to four other experimental data sets of in vitro digestions of infant formulas presenting various fat and protein compositions and different rheological properties, in order to test its ability to predict their kinetics of digestion. A proper modelling of dynamic in vitro digestion phenomena occurring in the gastrointestinal tract is a challenging task, but could be of great help to better understand the key factors involved in the final nutritional quality of infant milk constituants.