A Framework to Model Functional Food Formulation for Nutrient Bioavailability

Thesis Abstract

The study addresses the persistent gap between conventional functional food formulations and the actual bioavailability of micronutrients in diverse populations, acknowledging that nutrient release, absorption, and metabolism are modulated by food matrices, processing, and consumer-specific factors. Despite advances in formulation science, there remains a lack of an integrative framework that simultaneously optimizes matrix design, processing conditions, and predicted in vivo bioavailability. The aim is to develop a comprehensive framework for modeling functional food formulations that enhance nutrient bioavailability, with objectives to (i) identify key physicochemical determinants of nutrient release and absorption within representative food matrices, (ii) integrate in vitro digestion, permeability assays, and physiological variability into a unified modeling approach, (iii) validate the framework using case studies on iron, zinc, and folate in fortified cereal bars, and (iv) evaluate the framework’s predictive accuracy against in vivo data reported in the literature. The research adopts a mixed-methods, theory-driven design grounded in the frameworks of the Biopharmaceutical Classification System (BCS), the Food Matrix–Bioavailability Interaction (FMBI) model, and the Nutrient Interaction Theory. A sequential exploratory design guides the study, beginning with model development and in vitro validation, followed by computational simulation and a limited in vivo corroboration. The population comprises representative food matrices (fortified cereal bars and plant-based beverages) and human physiological variability across three demographic strata (young adults, middle-aged adults, and older adults). A purposive sample of five formulated products per matrix will be prepared under standardized processing conditions to produce 15 primary formulations. In vitro experiments will simulate gastric and intestinal phases using standardized INFOGEST protocols and will quantify nutrient release and solubility using Inductively Coupled Plasma Mass Spectrometry (ICP-MS) for minerals and High-Performance Liquid Chromatography (HPLC) for folates. Permeability will be assessed with Caco-2 cell monolayers to estimate transepithelial transport. Instrumental data will be complemented by computational modeling using multivariate regression, partial least squares (PLS) path modeling, and Bayesian networks to integrate physicochemical descriptors, digestion outcomes, and absorption likelihood. Model specification will incorporate interaction terms to reflect matrix–nutrient and processing–bioavailability effects, with validation against published in vivo bioavailability data (n ? 15 human studies) and sensitivity analyses to determine parameter influence. Expected findings include (i) identification of matrix components and processing parameters that significantly enhance or impede mineral and folate bioavailability, (ii) a validated, scalable framework that can predict in vivo bioavailability from in vitro and in silico indicators with acceptable predictive accuracy (R2 > 0.70 in cross-validation), (iii) quantification of inter-individual variability effects on formulation performance, and (iv) a set of design rules linking specific processing conditions to target bioavailability outcomes. The study will contribute to knowledge by delivering an integrative, theory-informed framework that unifies food matrix science, digestion dynamics, and absorption physiology into a practical modeling tool for formulation scientists. It will also extend the FMBI model with quantitative metrics suitable for industry deployment and regulatory dialogue. The main conclusion anticipated is that a structurally explicit framework combining matrix design, processing, and biological variability can reliably forecast nutrient bioavailability and guide formulation strategies. Recommendations include integrating the framework into standard product development pipelines, expanding validation with in vivo trials across broader populations, and developing user-friendly software that automates data input, model calibration, and scenario analysis for practitioners in food technology and nutrition science.