Determination of authenticity of pure milk fat. linear discriminant analysis of triacylglycerides

Abstract

The addition of non-milk fat of animal or plant origin to dairy products such as fluid milk, milk powder, cheese or butter is considered to be an adulteration.
The purpose of this research was is to propose a methodology using linear discriminating analyses for the detection of adulteration of milk fat with non-milk fat (animal origin) at levels below 10%, with at least 95% of the analysis correctly classified. The derived functions are applicable to pure milk fat samples of the VIII^th, IX^th and X^th regions. Similarly partial discriminating functions applicable to pure milk fat of the said regions were also proposed. The percentage composition of triacylglycerides (C₂₆ to C₅₄) in 72 samples of pure milk fat of milk stored in tanks of six dairy plants was determined. The pure milk fat samples were adulterated with non-milk fat according to a theoretical model at 0 %, 5 %, 10% and 15% within 4 groups (n groups = 72).
The best discriminant function F1 found for the southern zone was:
F1 = 2.4633C₃₄ – 2.69418C₃₆ + 0.895685C₃₈ – 2.09497C₄₀+1.773789C₄₂-1.57736C₄₄
where C₃₄, C₃₆, C₃₈, C₄₀, C₄₂ and C₄₄ are the weight percentages of the weight of triacylglycerides with carbon numbers 34, 36, 38, 40, 42 and 44. This equation allowed the correct classification of 97.2 or 79.1% of the samples adulterated with 10 or 5% of non-milk fat of animal origin, respectively.