Abstract #383

# 383
Evaluation of an automatic fat analysis system for determination of fat in dairy products.
H. Jiang*1, J. Romero1, S. Govindasamy-Lucey1, 1Wisconsin Center for Dairy Research, University of Wisconsin-Madison, Madison, WI.

The official standard (reference) methods used for testing fat in dairy products involves acid hydrolysis and organic solvent extraction steps. These methods are labor intensive and time consuming; a simpler and rapid testing system would be desirable. This study aimed to validate an automatic fully enclosed fat analysis system (Hydrotherm-Soxtherm) for the analysis of fat content of dairy products by comparison to a standard method (modified Röse-Gottlieb). The automatic system consists of acid hydrolysis (Hydrotherm), which is based on Weibull-Berntrop gravimetric method (ISO 8262–3:205), and fat extraction by soxhlet method (Soxtherm). A single laboratory validation (SLV) was performed; linearity, precision, limit of detection (LOD), and limit of quantification (LOQ) were determined. As ultrafiltered (UF) permeate stream contains (almost) no fat, it was used as a blank sample to calculate LOD and LOQ. Fat quantification for fluids were performed by creating multi-point calibration with increasing fat contents (0.5–35%) by mixing UF permeate with cream samples. Trueness (how close a result is to the true value) was determined on milk samples (n = 5) and commercial cheeses (n = 48) containing fat contents ranging from 0.5 to 48%. A double-blind test was performed on 18 varieties of commercial cheese on 4 different days. Acceptability of this automatic system with respect to precision was determined by calculating Horwitz Ratios (HorRatR). The LOD and LOQ values were determined to be 0.312g and 0.7839g per 100g, respectively. Linearity (R2 ≥ 0.99) was observed in the multi-point calibration method. Fat data obtained by the automated system were in very close agreement with the standard method for the commercial cheeses; there was no significant difference (P > 0.05) between the results obtained by the 2 methods, showing that the validated method is a suitable alternative. The HorRatR were found to be 0.3–1.3 with an average of 0.9, which is within the acceptable range for SLV. The automated method was less labor intensive and easier to use for routine standard fat analysis of dairy products.

Key Words: single lab validation, automatic fat analysis system, dairy products