Advanced search
Start date
Betweenand
Related content
(Reference retrieved automatically from Web of Science through information on FAPESP grant and its corresponding number as mentioned in the publication by the authors.)

Lipid classification of fish oil omega-3 supplements by 1H NMR and multivariate analysis

Full text
Author(s):
Lima Amorim, Tatiane [1, 2] ; Granato, Alisson Silva [3] ; Mendes, Thiago de Oliveira [4] ; Leal de Oliveira, Marcone Augusto [1] ; Amarante, Giovanni Wilson [3] ; Angel de la Fuente, Miguel [2] ; Gomez-Cortes, Pilar [2]
Total Authors: 7
Affiliation:
[1] Univ Fed Juiz de Fora, Grp Quim Analit & Quimiometria GQAQ, BR-36036900 Juiz De Fora, MG - Brazil
[2] Inst Invest Ciencias Alimentac CIAL CSIC UAM, Nicolas Cabrera 9, Madrid 28049 - Spain
[3] Univ Fed Juiz De Fora, Grp Pesquisas Metodol Sintet GPMS, BR-36036900 Juiz De Fora, MG - Brazil
[4] Univ Brasil, Inst Cient & Tecnol, BR-08230030 Sao Paulo, SP - Brazil
Total Affiliations: 4
Document type: Journal article
Source: Journal of Food Composition and Analysis; v. 102, SEP 2021.
Web of Science Citations: 0
Abstract

The worldwide advent of concentrated supplements containing omega-3 fatty acids (FA) in the form of triacylglycerols (TAG) or ethyl esters (EE) has increased the interest in developing methods to classify these products. The quality control based on their lipid composition has become necessary since EE bioavailability has been proved to be lower when compared to the TAG. In this preliminary study, eight models based on 1H NMR and supervised discriminant analysis (PLS-DA/OPLS-DA) were applied to classify omega-3 fish oil in TAG or EE forms. The 4.0-4.5 ppm region was selected for modeling since it bracketed spectral features to discriminate TAG and EE. The non-supervised principal component analysis was employed to visually evaluate the distribution of samples and revealed a clear separation of TAG from EE marine oils along PC1. In addition, representative TAG and EE samples were 100 % correctly classified using any of the eight supervised models studied. The developed models resulted in high R2Y (> 0.977) and Q2 (> 0.953), and low root mean square error for prediction (<= 0.009), which demonstrates the high potential of this rapid and straightforward procedure to evaluate the lipid form of supplements and mislabeling. (AU)

FAPESP's process: 14/50867-3 - INCT 2014: National Institute of Science and Technology in Bioanalysis
Grantee:Marco Aurelio Zezzi Arruda
Support Opportunities: Research Projects - Thematic Grants