No abstract

Background and Aims Wine fraud has been practised since ancient times, continuing up to the present day and taking many forms. In this study, detection of the addition of unpermitted sugar and water to wines was evaluated using a univariate method, officially adopted, and several more effective multivariate methods. Methods and Results The results are based on 5220 Italian wine samples belonging to the European Union (EU) Wine Databank collected in the period 2000–2010. This databank contains stable isotope ratios for wine samples from many regions over many years. The addition of variable amounts of sugar and water was simulated according to experimental results, and univariate and multivariate methods were compared. It was shown that this type of fraud can be discovered much more easily by the application of multivariate methods. Conclusions It can be difficult to detect the addition of unpermitted sugar and water to wines. With multivariate tests based on the EU Wine Databank, the likelihood of detecting this kind of fraud is significantly increased. Significance of the Study This research evaluates and proposes a more efficient method for statistical interpretation of the EU Wine Databank for evaluation of wine authenticity.

No abstract

A common fraud in the dairy field is the addition of sheep's milk to goat's cheeses, because it has a very similar taste to goat's milk, but is more available, and is commonly considered to have a better capacity to curdle. For similar reasons, and due to economic convenience, sheep's cheeses may also contain fraudulent cow's milk. In order to detect this fraud, an EU official method may be used, but it is only a qualitative method (presence/absence of cow's milk). A method able to quantify the presence of sheep's milk during cheese production in goat's and cow's cheeses was developed. The method is based on liquid chromatography/electrospray ionization tandem mass spectrometry (LC/ESI-MS/MS) analysis of peptides of a casein extract from the cheese. By a simple procedure, caseins are extracted from cheeses, solubilized, digested with plasmin, and subsequently analyzed by LC/ESI-MS/MS. A typical sheep's peptide produced by plasmin hydrolysis (m/z = 860) was accurately selected and analyzed to understand if, and by how much, a declared pure goat's cheese contains sheep's milk. By analyzing the same peptide it is also possible to detect if, and by how much, a declared pure sheep's milk contains, or not, cow's milk. The method was applied to several goat's and cow's cheese samples. Quantitation was performed with a calibration curve obtained by analyzing curd cheeses containing different percentages of sheep's milk. The method detection limit and method quantitation limit were evaluated. This method appears accurate and suitable for detecting up to 2% of sheep's milk in cheeses.