NMR spectroscopy as a tool to discriminate between organic and conventional cow milk

The organic farming system involves higher production costs and economic risks which must be converted into an adequate market price. The premium price of organic products increases the risk of frauds. Nuclear magnetic resonance (NMR) spectroscopy is a robust analytical technique with high reproducibility, minimal and cost-effective sample preparation. NMR technique is commonly used in metabolomics, allowing to obtain information on the state of the organism originating the sample. NMR spectroscopy and metabolomics have been applied in the dairy sector for the identification of the geographical area and frauds, the determination of nutritional and technological traits, and the authentication of production systems. The present study evaluated the sensitivity of NMR spectroscopy for the authentication of organic bovine milk from conventional milk. Bulk milks of 24 herds (12 conventional and 12 organic) were sampled between September 2019 and May 2020 in the North of Italy. Three bulk milk samples per farm were collected in different months for a total of 72 samples. Milk samples were analysed for traditional composition traits using mid-infrared spectroscopy and for metabolites using NMR. Among quantified metabolites, the acetate was significantly greater in organic than conventional milk. Multivariate and univariate approaches were used to assess the capability of NMR spectroscopy to discriminate between the farming systems. Models built using the Random Forest learning method exhibited an accuracy between 63.9 and 76.4% using the entire NMR spectra and the areas of the identified metabolites, respectively.