Modeling and predicting spoilage of cooked, cured meat products by multivariate analysis

A cooked, cured meat product is a perishable product spoiled mainly by lactic acid bacteria (LAB). LAB cause discoloration, slime formation, off-odors and off-flavors as the result of their metabolic activity producing various products. These microbial products in conjunction with the microbial population could be used to assess the degree of spoilage of this type of product. The spoilage evaluation was achieved by following a multivariate approach. Cluster analysis, principal component analysis and partial least square regression were employed to associate spoilage with microbiological and physicochemical parameters. The developed model was capable of giving accurate predictions of spoilage describing the spoilage associations. The study might contribute to the improvement of quality assurance systems of meat enterprises. (c) 2007 Elsevier Ltd. All rights reserved.