Predicting the organoleptic stability of beer from chemical data using multivariate analysis

摘要

Delaying flavour staling, to prolong the shelf life of beer, is one of the greatest challenges facing the brewing industry today. In this study, a series of lager beers were brewed using five European barley varieties and four chemical indicators of the degree of beer ageing were correlated with the sensory evaluation: E-2-nonenal, beta-damascenone, 2-furaldehyde and 5-hydroxymethyl-2-furaldehyde (5-HMF). A statistical strategy using principal component analysis and multiple linear regression was applied to draw relationships between the sensory and chemical data sets. Additionally, the relative significance of each of the chemical data on the organoleptic stability of beer was evaluated within the method. 5-HMF was the only studied carbonyl compound whose concentration cannot be used for predicting the total taste score of beers. E-2-nonenal, in contrast, was found to be the most discriminant carbonyl compound under consideration for predicting the flavour stability of beer.