For the production of high quality meat products, processors demand firm pig fat. About 20 years ago, Swiss slaughter houses established a titrimetric method to determine the so-called fat score (an index of double bonds in the fat) as a quality control measure. Since 2005, an NIR method is applied to determine the fat score in two processing plants. The use of toxic solvents and reagents was thus avoided and the work needed considerably reduced. The result, nevertheless, is an unspecific measure of the sum of double bonds. The aim of this study was to evaluate the potential of determining the fatty acid composition more specifically or even to predict fat consistency traits directly in pig fat without extracting the lard. For this purpose a total of 155 samples of pig fat was scratched from the outer layer of the backfat at the slaughter line according to the official fat score method. The collected fat was mixed with a rotor–stator homogeniser, before NIR spectra were recorded. The fatty acid composition was analysed using gas chromatography and the solid fat content was measured by differential scanning calorimetry. The proportion of saturated fatty acids (SFA), monounsaturated (MUFA) and polyunsaturated fatty acids (PUFA) was validated with relative standard errors of prediction (SEP) of 0.9%, 1.6% and 4.7% and ­coefficients of determination (r²) of 0.98, 0.88, and 0.96, respectively. The validation of the proportion of very long chain PUFA resulted in a SEP of 7.6% and an r² of 0.88. The solid fat content (at 10°C and 20°C) was also estimated with sufficient precision and accuracy (n = 80; SEP = 2.9% and 3.2%, r² = 0.94 and 0.96, respectively). NIR, therefore, provides the potential to establish more specific measures for pig fat quality than the original fat score method.

Keywords: NIR spectroscopy, pig, fatty acid composition, consistency, pork quality