Food authentication of almonds (Prunus dulcis mill.). Origin analysis with inductively coupled plasma mass spectrometry (ICP-MS) and chemometrics

Food fraud is a pervasive issue; products with high sales and region-dependent prices are strongly affected. In this study, isotopolomic fingerprinting was used to distinguish the origin of almonds in order to prevent food fraud. For this purpose, 250 almond samples consisting of more than 30 varieties from six different countries and four crop years were analysed and the chemical profiles were evaluated using chemometric methods. By using log10 as pre-treatment and support vector machine (SVM), a prediction accuracy of 92.2% ± 0.7% was achieved. The respective harvest year and the underlying almond variety had no significant effect on the origin discrimination.