To explore the flavor characteristics of human milk, we constructed a three-tiered human milk flavor wheel based on 53 sensory descriptors belonging to different sensory categories. Fifteen sensory descriptors were selected using M-value and multivariate statistical methods, and the corresponding references were set up to realize qualitative and quantitative sensory evaluation of the human milk samples. To ensure the accuracy and reliability of the sensory evaluation, the performance of the sensory panelists was also tested. The sensory profile analysis indicated that the established sensory descriptors could properly reflect the general sensory properties of the human milk and could also be used to distinguish different samples. Further investigation exposed that the fat content might be an important factor that influence the sensory properties of human milk. To the best of our knowledge, this is the first report on the flavor wheel of human milk.
Keywords: descriptive analysis; flavor wheel; human milk; sensory characteristics; sensory lexicon; sensory profile descriptors.