Principal component and multivariate analysis of milk long-chain fatty acid composition during diet-induced milk fat depression.

The objective of this study was to assess the relationship between individual milk fatty acids (FA) and diet-induced milk fat depression (MFD) using principal component analysis (PCA) and multivariate analysis (MA). Cow treatment observations (n = 63) from 3 published feeding experiments with lactating dairy cows were used in the analyses. In the PCA, principal component loading plots 1 (PC1) and 2 (PC2) described 55.9% of the total variation in milk FA and fat concentrations. Saturated FA (14:0, 16:0, and 17:0) and milk fat percentage showed negative loading for PC1. Trans-18:1 isomers (trans-6+7+8 to trans-15), trans-7, cis-9 conjugated linoleic acid (CLA), and trans-10, cis-12 CLA showed positive (opposite) loading, suggesting a negative relationship between these isomers and milk fat percentage. Cis-11, trans-13 CLA and cis-9, trans-11 CLA were associated with the PC2 axes (neutral), indicating that they were not associated with MFD. Multivariate analysis with milk fat percentage as the dependent variable and individual PC1 positive loading variables showed a breakpoint relationship for trans-6+7+8-, trans-9-, trans-10-, and trans-13+14-18:1 and a linear relationship for trans-11-, trans-12-, trans-15-18:1, trans-10, cis-12 CLA, and trans-7, cis-9 CLA. Subsequent MA was conducted on 41 treatment means from 12 independent experiments from the literature, in which concentrations of trans-6+7+8-, trans-9-, trans-10-, and trans-11-18:1, and cis-9 trans;-11, and trans-10, cis-12 CLA were reported. Significant negative effects of trans-9-18:1, trans-10-18:1, and trans-10, cis-12 CLA on milk fat percentage were observed. In this study, the PCA and MA showed that among trans-18:1 isomers, trans-10-18:1 was the most negatively correlated to milk fat percentage. However, the threshold concentration related to maximum MFD indicated that the relative potency was greatest for trans-6+7+8- and lowest for trans-10-18:1. These results suggested that trans-6+7+8-18:1 might be more important than trans-10-18:1 in MFD. Principal component analysis also showed that trans-10, cis-12 and trans-7, cis-9 CLA were the isomers most negatively correlated to milk fat percentage, implying a possible role of trans-7, cis-9 CLA in MFD. Additional experiments are needed to establish whether trans-7-18:1 is involved in MFD or that its effects are mediated via the endogenously synthesized trans-7, cis-9 CLA.