A multivariate approach to verify correlations between sensory quality indices and analytical quality indices in traditional Brazilian cheese

Artisanal Minas cheese (AMC) is a traditional Brazilian cheese that combines particular production techniques and maturation conditions resulting in cheeses with unique sensory properties. The objective was to use multivariate statistical to determine a quality index based on the correlation of sensory and instrumental analyses of AMC. 25 samples were obtained from a cheese contest, which were sensorially evaluated by judges. Composition, lipid oxidation, fatty acid and volatile compound profiles were analyzed. The data were analyzed using principal component analysis (PCA) and canonical correlation analysis (CCA). PCA helped identify the key variables that distinguished different types of cheeses based on their qualities. CCA showed that hardness and lipid oxidation are indicators of the sensorial quality of cheeses, parameters that classified the cheeses in the competition, with the best evaluated having higher levels of these attributes.