Modelling sweetness and texture perception in model emulsion systems

Model emulsion samples, prepared subject to a d-optimal response surface design, were used to investigate the effect of rapeseed oil, sugar and hydroxypropyl methyl cellulose thickener (HPMC) on perceived sweetness, textural attributes (thickness, stickiness, mouth-coating, dispersing), instrumental measures of apparent viscosity 50 s⁻¹ and Kokini oral shear stress. An increase in oil, sugar and HPMC resulted in an increase in perceived thickness, stickiness and mouth-coating, and a decrease in dispersion; sweetness was enhanced by the addition of both sugar and oil but suppressed by the addition of HPMC. Viscosity and Kokini oral shear stress were well correlated with oral thickness (r ² > 0.9). Validated multiple linear regression models highlighted several 2-factor interactions between ingredients. Model statistics indicated that the variation in data was well explained; the models were predictive and could be used to navigate the design space. Samples predicted to be iso-thick and iso-sweet could not be discriminated (P > 0.10) in a 3-alternative forced choice (3-AFC) test using 35 panellists.