The rapid growth of plant-based milk consumption, driven by health awareness, lactose intolerance, and environmental sustainability concerns, has increased the demand for products with stable physical properties and acceptable mouthfeel. Meeting these quality expectations requires precise control of formulation parameters, particularly the type and concentration of hydrocolloid stabilizers used to govern texture and stability. This study aimed to develop and validate a non-linear regression modelling framework, guided by rheological characterization of commercial plant-based milk benchmarks, to predict optimal concentration ranges of five commonly used hydrocolloid stabilizers (gellan gum, xanthan gum, guar gum, carrageenan, and carboxymethyl cellulose). Rheological properties were evaluated using modular compact rheometer, and the relationships between stabilizer concentration and key rheological parameters—flow behavior index (n), consistency index (K), and apparent viscosity at 50 s⁻¹—were modelled using logarithmic and power-law regression. The models demonstrated strong predictive performance (R² > 0.97 for most stabilizers) and the predicted optimal concentration ranges for each stabilizer were determined as follows: carrageenan (0.0583–0.3613%), CMC (0.0148–0.1121%), guar gum (0.0718–0.1859%), xanthan gum (0.0596–0.2249%), and gellan gum (0.0581–0.1373%). The predicted concentration ranges provide practical guidance for reducing trial-and-error formulation, shortening product development time, and improving formulation reproducibility in the food industry. Nevertheless, model validation was limited to soy milk, and further studies are required to extend its applicability to other plant-based matrices.
Copyrights © 2026