Fraud detection in hen housing system declared on the eggs' label: An accuracy method based on UV-VIS-NIR spectroscopy and chemometrics.

EU regulation classifies egg production in four hen housing systems: organic, free range, barn and cages. However, there are no analytical methods for a complete detection of the housing systems declared on the eggs' label. This work describes a new method for this task. A lipid extract was obtained, and eggs classified in the four housing systems by UV-VIS-NIR spectroscopy and chemometrics. A total of 192 spectra were analysed. Eggs were classified by Support Vector Machine classification and Linear and Quadratic Discriminant Analysis (LDA and QDA, respectively). The cholesterol concentration did not allow to classify correctly the four farming systems. However, the yolk lipid extract successfully classified the eggs of the hen housing system. Results showed 100% accuracy by UV-VIS-NIR spectrum of the yolk lipid extract with the QDA statistical analysis. These results indicate that this lipid extract is a promising tool for analytical verification of the farming system.