Trans unsaturated fatty acids : a study of methodology and levels in New Zealand food fats including milkfat : a thesis presented in partial fulfilment of the requirements for the degree of Masters of Science in Biochemistry at Massey University, New Zealand

Trans fatty acids (TFAs) occur naturally in small amounts in foods such as milk, butter and tallow as a result of biohydrogenation by ruminant gut microflora. They are formed in much larger quantities during chemical hydrogenation of fats and oils. The relationship between dietary TFAs and blood cholesterol has been investigated over the last 30 years with equivocal results because of methodological limitations, including difficulties of quantifying the consumption of TFAs. The present study was conducted to investigate the methodologies used to quantify TFAs in fat samples. Two methodologies, based on infrared spectrophotometry and argentation-thin layer chromatography/gas chromatography (Ag-TLC/GC), were optimised for TFA quantification. Improvements in the infrared methods were made using a calibration standard made up with two non-trans components (stearin and olein) in order to mimic the fatty acid background in the samples. Further improvements were made using a spectral subtraction technique where the non-trans background spectrum was subtracted from the sample spectrum using Fourier-transform infrared spectrophotometry software. Results from the improved infrared methods were compared with TFA measurements by the more detailed Ag-TLC/GC method. The spectral subtraction technique for the methyl ester samples produced results that were closest to those of the Ag-TLC/GC method. This Ag-TLC/GC method gives information about the individual trans isomers (C18:1 trans positional isomers and C18:2 and C18:3 trans isomers) that is not available by infrared. The present study was also conducted to determine, as accurately as possible, the TFA content in 18 manufactured foods commonly available in New Zealand using the TFA methods mentioned above. The TFA contents in some of the foods determined by the Ag-TLC/GC method were, margarine (15.43-15.57%), butter (6.58%), milk (5.26-6.03%), meat patties (3.42%), plain sweet biscuits (3.65%) and white bread (4.41 %). Using these TFA data and the food consumption data from a Life in New Zealand (Horwarth et al., 1991), the estimated TFA intakes in the average New Zealand diet were approximately 3.99 and 5.75 g/person/day for females and males respectively. These figures were similar to or lower than those estimated for Northern Hemisphere countries. The predominant TFA isomer in the New Zealand diet was identified as the C18:1 Δ11t positional isomer (30-33%). Further studies were made on the total TFA content in New Zealand milkfat. These studies indicated that the total TFA levels in New Zealand milkfat were influenced by seasonal variations, with the highest TFA content recorded in spring (September, 6.7%) and the lowest in summer (January, 5.3%). The C18:1 Δ11t isomer was found to be the predominant isomer in milkfat, making up 49-60% of the total TFA. Similar ranges were observed for several overseas butter samples. However, major differences were observed with the distribution of the C18:1 trans positional isomers. These differences are currently suspected to be influenced by the feed and animal husbandry methods used in some Northern Hemisphere countries, where cows are mainly grain fed in the winter months. The seasonal variation of TFAs in New Zealand butter and possible effects of feed and animal husbandry methods on the C18:1 trans positional isomer distribution are important factors that the New Zealand dairy industry could exploit for the production of low trans milkfat and/or other dairy products in which the levels of specific trans isomers implicated to be "harmful" to humans could be minimised. Margarines display a trans isomer distribution that is quite distinct from that of butter. Unlike milkfat where the predominant trans isomer is C18:1 Δ11t, in margarines and hydrogenated fats and oils the positional isomers show a normal distribution around the C18:1 Δ10t-11t isomers. The predominant isomers for the margarines analysed in this study were Δ9t-Δ12t (90%) with the polyunsaturated C18:2 and C18:3 trans making up less than 2%. The distinct distribution of C18:1 trans positional isomers could serve as an additional tool for the identification of animal or hydrogenated vegetable oils used in food fats