PREDICTING TEXTURE ATTRIBUTES FROM MICROSTRUCTURAL, RHEOLOGICAL AND THERMAL PROPERTIES OF HAZELNUT SPREADS

Abstract

ABSTRACT Eight trained assessors evaluated one visual, eight texture and four gustative attributes of 15 commercial spreads. In addition, 100 consumers evaluated the acceptability of the spreads. Rheological behavior of the samples was evaluated in transient (stress–relaxation test) and dynamic (strain and frequency sweep test) shear mode. Thermal measurements were also performed. Finally, the microstructure of the samples was analyzed by scanning electron microscopy. Partial least squares regression was used to predict acceptability from sensory attributes and texture from instrumental data. Scanning electron micrographs indicated that spreads differed in morphology, size and distribution of hazelnut particles. Overall liking of the spreads was found to be related to hazelnut flavor and texture attributes. Prediction of the most important texture properties from instrumental measurements was quite successful, especially for spreadability which was found negatively related to the consistency and flow indices and, secondly, for meltability which was found to be strongly dependent on measured thermal parameters. PRACTICAL APPLICATIONSThis study contributes to knowledge development in the research area of acceptability, sensory and instrumental correlation for semi‐solid foods. The results confirm that texture properties of spreadable foods play a most important role in consumer acceptability. The research also allows better understanding as to why no single measurement is able to predict all texture attributes adequately and why more appropriate tests need to be developed for an accurate prediction of some sensory attributes as perceived by the human senses.