Prediction of astringency

摘要

The capacity to predict astringency has been the subject of considerable research. In our review of red wine astringency in 2011 (Trends Food So'. Nutr. vol. 27, pp. 25-36), we summarised six different approaches that have been used. In reality, most of these studies were not 'predictive'. Rather, they simply determined the correlation between the taster-assessed astringency and a chemical assay for astringency. One approach towards a predictive model has used the measurement of the turbidity that resulted following the addition of mucin to a wine and the consequent formation of a mucin-polyphenol precipitate. While the authors initially used grapeseed extract to develop their method, they found a high correlation (r2 = 0.95) between the measured turbidity and astringency rating for 18 wines (Condelli et aL, Food Qual. Pref. 2006, vol. 17, pp. 96-107).