|
Abstract:
|
The reformulation to reduce calories from beverages was motivated by consumer-driven health trends and regulatory pressures. The challenging goal was to replace sugar with other ingredients to fit a range of region-specific consumer tastes yet to conserve the mouthfeel, a series of sensory attributes that determine the overall flavor. The study compared 26 variants and a control product already on market in terms of 23 sensory attributes. Subjective scores on attributes were assigned to each variant by panelists. This topic has been studied using traditional methods such as DOE, PLS and PCA. They are useful in selecting a variant significantly different from the control. However, to find a variant with similarity to the control we need less commonly used yet powerful statistical methods, which help achieving product changes with a myriad of functional materials. We explored some advanced models including Mahalanobis Distance which quantifies the separation from the control. We’ve found several qualified low sugar variants that contain commercially available ingredients. More confirmation studies were planned because changing the mouthfeel could lead to the loss of the consumers.
|
