Quantifying Food Texture Selectivity of Children Through Conjoint Analysis

Selective (picky) eating at an early age is very common. Food refusal due to certain sensory characteristics of a food, namely texture, can lead to inadequate intake of nutrients and a limited list of acceptable foods. This research set out to quantitatively measure selective eating in regard to food textures by the use of adaptive choice-based conjoint analysis. Two populations of children aged 8-12 were identified by a parent as either non-picky (n=30) or picky eaters based on texture (n=19). Hierarchical Bayes theorem was used to calculate utilities of 30 texture attributes and importance scores of 8 overarching categories to which attributes were assigned. Four metrics were identified to quantify pickiness: number of unacceptable attributes chosen, number of times a person selected a group of food textures as "a possibility" (meaning they were acceptable), the None utility score calculated from the groups of food textures chosen as "not a possibility", and the importance scores- which demonstrated the relevance of each category. On average, the non-picky children chose fewer attributes as unacceptable and had a lower None utility than the picky children (p=0.046 and p=0.0090, respectively). Similarly, there was a trend for non-picky children to identify more model foods as being "a possibility" (p=0.060). Unevenness in importance scores illustrated which categories drive food choices. The use of conjoint analysis allowed for the differentiation between picky and non-picky eaters on the basis of texture preferences. This research has the opportunity to provide insight to parents on what foods to provide or to avoid feeding their child.