Relation Between Oxidative Stability and Composition in Argentinian Olive Oils

Abstract

The relation between oxidative stability and composition in 58 virgin olive oils from different cultivars and Argentinian regions was studied over four harvest years. The oxidative stability of the oils was assessed using the OSI index (110 °C, 20 L/h air flow). A multiple linear regression model is proposed using OSI values as the dependent variable (multiple R = 0.933, p = 1 x 10-15), with positive contributions of the independent variables: fatty acid composition [oleic acid/(linoleic acid ? linolenic acid), 55.3%, p = 1 x 10-15], total polyphenols (24.1%, p = 1.8 x 10-9), carotenes (4.8%, p = 6.1 x 10-5), β-toco- pherol (1.9%, 6.0 x 10-3) and other compounds (13.9%). Highly significant correlation was observed between oxidative stability indexes estimated by the compositional model and those experimentally determined by Rancimat method (b = 0.981, R = 0.924). Chlorophylls and Δ-5- avenasterol contributions to the model were non-significant when variables related with fatty acids and polyphenols were included. The results suggest that the fatty acid composition and the polyphenol content are the main factors that affect the oxidative stability of olive oils. The proposed model allows the estimation of the oxidative stability in olive oils independently of the cultivar. The model was obtained also taking into account samples that lie ossut of the international legal limits in some compositional values due to natural variations.