Healthiness mapping The sensory quality of healthy bread according to consumers

Healthiness is a central credence consumer quality dimension that needs to be communicated, but the use of health claims is highly controlled. The present study explores how sensory attributes influence consumer perceptions of bread healthiness. Sensory and health-related properties of rye bread (n=24) on the Swedish market were described and a representative subset of 9 samples (varying in whole-grain, rye flour, added sugar and sourdough content) were included in a consumer test. Consumers (n=401) tasted the samples blindly in random order and rated their perceived healthiness by indicating degree of agreement with the statement “This bread seems healthy” on a 9-point scale. Background questions regarding socio-demographics and bread consumption habits were also included.Overall, the consumers classified the breads into four groups with significantly different perceived healthiness. This classification corresponded well with the objective health-related measurements of the samples. Healthiness mapping was performed using Principal Component Regression. Perceived healthiness were positively correlated to brown colour, sourness, rye flavour and texture, such as heterogeneity and chewiness while negatively correlated to sweetness, deformability and plasticity. Hierarchical cluster analysis followed by K-means clustering identified three consumer clusters, with different healthiness perceptions. Internal healthiness mapping by Principal component analysis showed that cluster 1 (36%, n=156) clearly and correctly differentiated between the samples while cluster 2 (37%, n=167) discriminated less between the samples and mainly used colour as a cue, cluster 3 (20%, n=78) rated healthiness lower for all samples. Cluster 2 displayed a lower educational level, ate mainly white bread and was more interested in general information about a breads health effects. This study can help facilitate identification of healthy bread in different consumer segments. Based on only intrinsic sensory information 36% discriminated well between the samples while in other segments, the discrimination could be improved by comprehensible extrinsic health-related cues.