Premium
Multivariate analysis of fatty acid cis and trans isomers in margarines determined by HRGC/FID/capillary column
Author(s) -
Oliveira M. Beatriz,
Alves M. Rui,
Ferreira Margarida A.
Publication year - 2001
Publication title -
journal of chemometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.47
H-Index - 92
eISSN - 1099-128X
pISSN - 0886-9383
DOI - 10.1002/1099-128x(200102)15:2<71::aid-cem596>3.0.co;2-1
Subject(s) - biplot , chemistry , multivariate statistics , gas chromatography , chromatography , principal component analysis , multivariate analysis , fatty acid , cis–trans isomerism , food science , organic chemistry , mathematics , biochemistry , statistics , genotype , gene
The fatty acid profile, including trans isomers of unsaturated fatty acids, of eight margarine brands purchased at random from the retail market, in a total of forty samples, was determined by HRGC/FID/capillary column. The results obtained were analysed by statistical multivariate analysis in order to achieve a better interpretation and understanding of the lipidic composition of margarines and to assess possibilities for the development of new products. Predictive biplots based on principal component analysis (PCA) were very helpful in determining the major differences between margarines in terms of levels of the most abundant saturated and cis unsaturated fatty acids, and offer great potential for product development. Predictive biplots based on canonical correlation analysis (CCA) enabled the establishment of a relationship between the patterns of variation in the levels of trans, cis–trans and trans–cis isomers of unsaturated fatty acids with saturated and cis unsaturated fatty acids. The results obtained also showed that some sort of legislation should be devised in order to make product labelling more useful to consumers. The techniques used for the construction of biplots showed that the modulating effect of multivariate analysis, such as PCA or CCA, was kept, but interpretations became easier owing to the possibility of relating statistical results directly to the original fatty acids with appropriate scales for measurement. Copyright © 2000 John Wiley & Sons, Ltd.