The use of multivariate analysis as a method for obtaining a more reliable shelf-life estimation of fresh-cut produce: a study on pineapple

For fresh-cut products, the definition of a representative target attribute for shelf-life estimation is very hard to assess because, during storage, a large number of chemical, sensorial and physical attributes degrade at the same time. The aim of this study was to obtain a more reliable shelf-life estimation of fresh-cut pineapple by applying multivariate accelerated shelf-life testing (MASLT). This approach is based on principal component analysis (PCA) and allows an estimate of shelf-life considering several degradation reactions. Fresh-cut pineapple pieces were packaged in PP-PE bags (45 μm, 17.5x15.5 cm in size; OTR=940 cm3 m2 d-1, β=3.3) in a passive modified atmosphere and stored at 0, 5, and 15°C. A total variance of 90.7% was explained by three principal components (PC). The PC scores were used to build a multivariate kinetic chart that summarized the degradation information from all studied quality attributes. Changes of PC1 as a function of time were well described by a first-order kinetic for samples stored at 0°C and by a zero-order kinetic for those at 5 and 15°C, showing correlation coefficients ranging between 0.88 and 0.95. The results showed that texture, color score, and appearance score were the most important variables affecting the PC model. Then, establishing a shelf-life limit for each of the attributes included in the model, a cut-off criterion of -1.33 was calculated, defining a shelf-life of ∼3, ∼7.9, and >11 days for fresh-cut fruit stored at 15, 5, and 0°C, respectively.