Determination of milk fat authenticity in ultra-filtered white cheese by using Raman spectroscopy with multivariate data analysis

•Discrimination of foreign fat/oil types was achieved using Raman spectroscopy.•PLS-DA and PLS chemometric methods were used for analysing Raman spectra.•Variability in Raman data was correlated to fatty acid content of cheeses.•LODs for adulteration of margarine, and corn and palm oils were less than 4%. Cheese is one of the most widely consumed food products in the world. However, the increasing demand for nutritionally enhanced or functional products by the cheese industry has created new approaches that partially or fully replace milk fat. With this, new methods of adulteration have also been noted, potentially leading to these fully/partially-replaced products being offered as cheese. In this study, Raman spectroscopy was used to determine origins of fats in margarine, corn, and palm oils present in white and ultra-filtered cheese samples. Raman spectra were evaluated with partial least square-discriminant (PLS-DA) and PLS to identify fat/oil origins and adulteration ratios. The coefficients of determination and limits of detection for margarine, and corn and palm oil adulteration were found to be 0.990, 0.993, 0.991 and 3.38%, 3.36% and 3.59%, respectively.