Analytical shortcomings and other considerations related to the identification of biomarkers of dairy fat intake

Abstract

Over the last decade, the concentrations of certain odd-chain saturated fatty acids (for example, C15:0 and C17:0) and trans-palmitoleic acid (trans-9 C16:1) in plasma and erythrocytes have been proposed as biomarkers of dairy fat intake,1, 2, 3 which is expected to overcome major issues associated with dietary assessment methods that rely on self-report, such as food frequency questionnaires. However, some overlooked analytical issues may have resulted in misidentification or even precluded other candidate biomarkers of dairy fat intake from being identified in several previous studies.1, 2, 3, 4, 5 The separation of fatty acids in biological and foods matrices is usually performed by gas chromatography after lipid extraction and methylation/transesterification using either acid or alkali-catalysis.6 Fatty acid methyl esters (FAME) are routinely identified by retention time comparisons with authentic commercial standards using a gas chromatograph equipped with a flame ionization detector (GC-FID). Methyl esters (except geometric isomers) not available as commercial standards can be identified by GC-MS analysis of 4,4-dimethyloxazoline (DMOX) derivatives prepared from FAME as described elsewhere.7 However, achieving proper separation and unequivocal identification of FAME in foods containing a very complex lipid composition such as ruminant milk fat has proved particularly challenging due to partial or complete overlapping of a number of fatty acids even when the most appropriate columns (for example, CP-Sil 88, Varian Inc.; SP 2560, Supelco Inc.) and different GC temperature programs are used.8 The same shortcomings should be expected in the GC analysis of fatty acid composition from biological tissues collected from individuals consuming diets containing ruminant-derived foods such as dairy products, which could possibly lead to inaccurate analytical results and misleading conclusions.