Effect of dietary vegetable oils on milk yield, composition and CLA isomer profile in milk from dairy cows

The objective of this study was to determine the effect of adding vegetable oils (1.5% DM) to dairy cow rations on the content of unsaturated fatty acids in milk, including CLA isomers. The experiment was conducted on 16 Red-and-White cows in 4 × 4 Latin square design. The cows were divided into four groups and fed rations composed of grass silage, maize silage, spent brewer’s grain and concentrate. No significant differences were found in milk yield or the amount of standardized fat and protein content milk (FPCM). The addition of linseed oil significantly increased the milk fat content, while sunflower oil, significantly decreased it. Cows given rapeseed or linseed oil produced more fat than in the remaining groups. No significant differences were found in the protein percentage of milk or in milk traits (acidity, coagulability, density) among groups. Feeding cows rapeseed and sunflower seed oils increased the content of CLA isomers in milk by an average of 16%. The highest content of cis-9 trans-11 CLA isomers was found in the milk of cows receiving sunflower seed oil, the lowest, linseed oil. The daily production of cis-9 trans-11 CLA in milk ranged from 8.8 to 10.5 g/d.


INTRODUCTION
The possibility that the composition of cow's milk can be modified through dietary interventions seems very promising (Thomas and Chamberlain, 1984). It has been found that humans can synthesize the long-chain fatty acids EPA and DHA from their precursors, linoleic and α-linolenic acids, which are present in large amounts in rapeseed and soyabean oils. Both of these fatty acids inhibit HMG CoA reductase (β-hydroxy-β-methylglutaryl-CoA) in the liver, the enzyme responsible for endogenous cholesterol synthesis in mammals. Unsaturated fatty acids are found in high amounts in oil seeds, especially in the seeds of evening primrose, linseed, sunflower, rapeseed, soyabean, whereas rich sources of EPA and DHA are fish meals and fish oil (Bartnikowska and Kulasek, 1994).
Much interest has arisen in the possibility of increasing the conjugated linoleic acid (CLA) content of milk. This acid has been called the "rumen acid". CLA has been found to have an anti-cancer effect in laboratory animals (NRC, 1996). Moreover, its cis-9 trans-11 CLA isomer has been associated with preventing heart disease and diabetes, modulating the immune system and reducing lipogenesis (Noone et al., 2002). It has been shown that grazing increases the CLA content of milk fat (Kelly et al., 1998;Dhiman et al., 1999), whereas feeding cows silage lowers it (White et al., 2001).
It can be supposed that feeding cows vegetable oils as a source of unsaturated fatty acids may intensify the hydrogenation of fatty acids, thus increasing the concentration of CLA and precursors of CLA isomers in rumen fluid. If it is assumed that part of the CLA and CLA isomer precursors are absorbed in the duodenum and small intestine, it would be possible for CLA to pass into the udder, thus increasing its content in milk. The products arising in the process of hydrogenating linoleic acid should increase the endogenous synthesis of CLA in the mammary gland and other tissues. The objective of the presented study was to test the above hypothesis.

Animals and feeding
The study was conducted on 16 Red-and-White cows divided into 4 groups in 4 × 4 Latin square design with four 21-day periods each. The cows were selected for the experiment from a herd numbering 120 animals with an average milk yield of 6.5 thousand kg in the last lactation. The experimental cows produced from 24 to 25.5 kg milk/d between the 30 th and 60 th days of lactation. The ration was composed of maize and grass silages, wet spent brewer's grain, concentrate and vegetable oils mixed with wheat bran. The concentrate was fed during the morning and evening milkings at a dose of 0.28 kg/kg obtained milk. At the same time the cows were given the vegetable oil. The grass and maize silages were fed after the morning milking, whereas the spent brewer's grain, in the afternoon before the evening milking. Meadow grass was cut for silage in the head emergence stage (the dominant grasses were Dactylis glomerata, Festuca pratensis, and Lolium perenne). Maize was harvested at the wax stage. The rations were supplemented with rapeseed oil (RPO) in the control group and linseed oil (LIO), soyabean oil (SYO) or sunflower oil (SFO) in the experimental groups. The oils were given in the amount of 320 g daily mixed with about 3.2 kg of wheat bran per cow. The concentrate was composed of crushed barley and wheat grain, rapeseed oilmeal and a premix. The cows were fed according to INRA standards (2001). Feed intake was monitored at the end of each experimental period.

Measurements and chemical analysis
Milk yield was measured in the last three days of each experimental period (Mini F.V. Milkometer, Tru-Test Distributors Ltd., New Zealand). Milk for determination of fat and protein contents was preserved by adding 2-bromo-2nitro-1,3-propanediol and frozen. Penicillin was added to the milk used for fatty acid determinations and frozen. Milk traits (acidity, density, specific gravity) were determined in fresh milk from milkings on the last day of each period.
Blood samples were drawn from the cow's jugular vein on the last day of the experiment after the morning milking. Glucose was determined in plasma 2 h after sampling, the remaining assays were performed on frozen samples.
The dry matter content, proximate analysis, calcium and phosphorous contents of feeds were determined by standard methods. The nutritional value of feeds and rations was calculated using WINWAR software, calcium was determined by atomic absorption spectrometry (ASA) using a Philips PU 9400 apparatus, phosphorous, by a colorimetric method according to Polish Standards (PN-76/R-64781).
The fat and protein contents of milk were determined by NIR spectroscopy using a MilkoSkan model FT 120 from Foss-Electric. Acidity, density, and specific gravity of milk were assayed according to Polish Standards (PN-68/ A-86112). The nitrogen fraction in milk was determined using the procedure described by Gordon and Kolan (1983), milk fatty acids, as methyl esters, were analysed according to Atwal et al. (1990) using a gas chromatograph (GC Varian 3400) and a CP-Wax 58 column (60 m; 0.53 mm; 1.0 micron). Fatty acids were esterified following the method described by Folch et al. (1956).
The blood plasma contents of glucose, total protein, triglycerides, total cholesterol and high density lipoproteins (HDL) were determined enzymatically using diagnostic kits from Cormay Diagnostyka, whereas total lipids, using tests from Biochemtest.
The milk yield of cows was corrected for the content of fat (4%) and protein (3.2%); net energy intake (NEL) and protein digested in the intestine (PDI) were calculated. Milk production efficiency and the PDI-to-NEL ratio were calculated using regression equations (Subnel et al., 1994).
The results were subjected to statistical analysis of variance and a multiple range test (Tukey's test) using the Statgraphics 6.0 computer program.

RESULTS
The composition and nutritive value of the rations are given in Table 1, whereas the fatty acid composition of the vegetable oils, in Table 2. Adding vegetable oils to rations for cows in an amount approximately equivalent to 1.5% of the dry matter content of the ration (320 g/d) did not significantly affect the intake of dry matter (Table 3), energy (NEL) or protein digested in the small intestine (PDI), hence there were no significant differences in milk yield or the amount of milk standardized for fat and protein content (FPCM). Feeding cows linseed oil significantly increased the fat content of milk, whereas the addition of sunflower oil significantly lowered it (P<0.01). The differences in the protein content of milk did not reach statistical significance, with the exception of the cows given rapeseed oil (P<0.01). No significant differences were found in milk traits such as acidity, coagulability, density, or in milk production efficiency and the PDI/NEL ratio. The weight of the cows ranged from 575 to 599 kg, and daily weight gains averaged 144.5 g.
The N content of true protein, whey protein, and non-protein N was similar in all of the groups (Table 4), the casein N content was, however, significantly lower in cows fed rapeseed oil (P<0.01), and the share of casein N in total N was lower in the milk of cows given sunflower oil.  (Subnel i in., 1994), 2 efficiency for milk production (Subnel i in., 1994), 3 liveweight, 4 body mass, RPO -rapeseed oil, LIO -linseed oil, SYO -soyabean oil, SFO -sunflower oil, a,b values in the same rows with different letters differ significantly (P<0.05), A,B values in the same rows with different letters differ significantly (P<0.01) The basic fatty acid profile of milk was typical for the milk of Red-and-White cows (Table 5). Saturated acids dominated among the fatty acids, with myristic, palmitic, and stearic acids (C 14:0, C 16:0, C 18:0 ) accounting for about 59 g/100 g of fatty acids. No significant differences were found among the groups of cows in the contents of short-and medium-chain fatty acids or of mono-, di-, and tri-unsaturated fatty acids. Feeding rapeseed and sunflower oils to cows significantly (P<0.01) increased the content of conjugated CLA in milk. Significant differences were found in the content of cis -9, trans-11 CLA isomers: the highest concentration was in the milk of cows receiving sunflower oil, the lowest, in those given linseed oil. The difference in relative amounts came to 29.6%. The content of the cis-9 trans-11 CLA isomer in milk amounted to 82.1% of the total CLA content of milk, with the remaining isomers, including trans-10 cis-12, cis-9 cis-11 and trans-9 trans-11, accounting for 3.1; 3.2, and 11.5%, respectfully, of the total amount of CLA; the differences in their contents were not significant.
The saturated fatty acid content in cow's milk came to about 70 g/100 g, that of monounsaturated fatty acids (MUFA), to about 25 g/100 g.
Supplementing rations with soyabean oil significantly (P<0.01) lowered the production of medium-chain fatty acids, while adding soyabean and sunflower oils decreased production of long-chain fatty acids (Table 6). The production of CLA, including its cis-9 trans-11 isomer, was the lowest in cows receiving sunflower and rapeseed oils, whereas the content of the remaining CLA isomers did not differ significantly among groups. The production of unsaturated fatty acids in milk was significantly lower when cows were fed soyabean oil (P<0.01) in comparison with the remaining oils, but PUFA production remained unchanged. Significantly lower production of fatty acids having hypocholesterolaemic properties was found in cows fed soyabean oil (P<0.01).
The differences in blood plasma parameters (glucose content, total protein, lipids, triglycerides, cholesterol and the lipoprotein fractions HDL and LDL) in cows given vegetable oils were not significant (Table 7).

DISCUSSION
Studies carried out to date point to the anti-cancer effects of CLA, its properties as an immune modulator, its anti-diabetic actions and reduction of atherogenesis (Parodi, 1997). CLA in milk comes mainly from endogenous synthesis (from 61 to 78%) in which the enzyme, Δ 9 desaturase, plays a role , the remaining part comes from the rumen, taken up directly from the bloodstream and lipoproteins transporting fatty acids to the mammary gland Griinari et al., 2000). In the rumen, CLA is an intermediate in the process of hydrogenating unsaturated fatty acids, mainly linoleic and α-linolenic acids, by rumen bacteria (Kepler and Tove, 1967). According to Doreau and Ferley (1994) the extent of hydrogenation of fatty acids in the rumen is very high and comes to about 85-100% for linoleic acid, and 70-85% for linolenic acid.
The vegetable oils used in this study differ in their fatty acid compositions. The dominant fat in rapeseed oil is oleic acid (C 18:1 ), in linseed oil, linolenic acid (C 18:3 ), whereas in soyabean and sunflower oils, linoleic acid (C 18:2 ). Hence the MUFA/ SFA, PUFA/SFA and MUFA/PUFA ratios in these oils differ.
The results of this study indicate that feeding cows vegetable oils in amounts equivalent to about 1.5% of the ration's dry matter content (320 g/d) did not negatively affect milk yield or the basic traits of milk (acidity, density, and specific gravity). No significant differences were found either in milk protein production efficiency or the ratio of protein digested in the small intestine to the net energy of milk (PDI/NEL).
Depending on the type of added oils, which differed in their saturated fatty acid content, the CLA content of milk differed significantly; its content in the milk of cows fed sunflower or rapeseed oils was about 16% higher than in cows given linseed and soyabean oils. It can, therefore, be supposed that the fatty acids in both of these oils were hydrogenated to a high extent, including transvaccenic acid (C 18:1 trans-11, TVA), which is considered to be the main substrate for the endogenous synthesis of CLA (C 18:2 cis-9 trans-11) (Wonsil et al., 1994;Salminen et al., 1998).
It was shown that the cis-9, trans -11 isomers of C 18:2 acid account for about 90% of the total CLA (Griinari and Bauman, 1999;Parodii, 1999). We found lower values in our study. An increase in the content of C 18:2 cis-9 trans-11 CLA and transvaccenic acid C 18:1 trans-11 was obtained in the milk of cows receiving fish oil supplements, extruded soyabeans, or their mixture (AbuGhazaleh et al., 2002).
On the basis of studies carried out to date the consensus is that the most effective way of increasing the content of CLA isomers in milk is to give lactating cows and goats a mixture of polyunsaturated fish oil or mixtures of other oils (Whitlock et al., 2002;Lock and Garnsworthy, 2002;Szumacher-Strabel, 2005). Ramaswamy et al. (2001) fed cows fish oil (2.0% of the ration's dry matter), extruded soyabeans (10.6%) and a mixture of both products and found 1.2-2.2 g CLA cis-9 trans-11/100 g of fatty acids in the milk of these cows. Baer et al. (2001) gave cows fish oil in an amount equivalent to 2.0% of the ration's dry matter and obtained 2.4 g/100 g fat of CLA cis-9 trans-11 isomer in comparison with 0.7 g/100 g in control rations. White et al. (2001) found variability among breeds in the CLA content of cow's milk, which shows that the ability to synthesize CLA is highly dependent on genetic traits, and also showed that the milk of grazing Holstein and Jersey cows contained about 0.7% and 0.6 g CLA/100 g fatty acids, respectively. It was also found that grazing in comparison with confined feeding increases the CLA content in milk by about 45%. The content of CLA isomers drops rapidly when switching from grazing to feeding silage in the winter. The studies conducted by Elgersma et al. (2004) show that within 14 days from ending grazing the content of cis-9 trans-11 CLA isomers in milk fell from 2.30 to 0.37 g/100 g fatty acids. The results of our study indicate that at a level of about 1.0-1.2 g total CLA/100 g fatty acids, the daily production of CLA equaled about 8.8-10.5 g/cow. The respective values for the C 18:2 cis-9 trans-11 isomer were about 7.0-8.6 g, at an average production of 952 g fat/day and 838 g fatty acids /day. On the basis of the studies conducted to date it can be said that the extreme values of CLA content in the milk of cows range from 0.3 to 0.6 g/100 g fat in cows that do not receive vegetable or fish oils to 1.0-2.5 g/100 g fat when oil seeds or moderate amounts of fish or vegetable oils are fed, or when the cows are pasture fed.
The amounts of CLA possible to obtain in milk given above are far from the amounts showing therapeutic activity. Wiliams (2000) reports that humans need 3.0-3.5 g/d of CLA in their diets to achieve an anti-cancer effect and 15-20 g/day to inhibit lipogenesis and fat deposition. This is about 0.04 g/kg body weight and 0.05% of a diet. These are amounts that would be difficult to obtain in cow's milk by feeding the animals in a way promoting the synthesis of CLA isomers and their excretion into milk.
A relatively high content of long-chain fatty acids, eicosapentaenoic (EPA) and arachidonic (AA), was found in the cows' milk, but the differences between the groups of cows and the oils used in their diets were not significant. EPA and DHA belong to the ω-3 family and are synthesized by the animals from linoleic and linolenic acids in a process catalysed by the enzymes ∆ 5 and ∆ 6 desaturase, which increase the number of double bonds, as well as by elongase, which is responsible for lengthening the fatty acid carbon chain (Givens et al., 2000). Both of these acids, as well as arachidonic acid (AA) found in the milk of cows, make up eicosanoids which are classified as tissue hormones, from which prostaglandins, leukotrienes and thromboxanes are synthesized in the bodies of people who drink milk. These substances lower blood pressure and exhibit anti-inflammatory, anti-thrombotic and anti-atherosclerotic properties (Bartnikowska and Kulasek, 1994).
Adding vegetable oils to diets for cows did not lead to an increase in total unsaturated fatty acids in milk, in contrast to cows given rations containing rumen-protected fatty acid calcium salts (Aii et al., 1991;Brzóska et al., 1999 a,b). This confirms the opinion that nearly the entire unsaturated fatty acid content of vegetable oils undergoes hydrogenation in the rumen and is not transported directly to the mammary gland in the form of protein-bound unsaturated fatty acids, i.e. lipoproteins. This conclusion is supported by in vitro studies (Gulati et al., 2000). A comparison of the fatty acid profiles of cows from the same herd (Brzóska et al., 1999 b) showed that feeding vegetable oils distinctly increased the content of palmitic oil (C 16:0 ), as well as that of oleic, linoleic, including CLA (C 18:2 ) and linolenic acids. Total unsaturated fatty acids and stearic acid (C 18:0 ) were higher in the control group of cows that did not received a fat-supplemented diet.
Feeding cows vegetable oil did not cause significant changes in blood plasma parameters. The glucose, total protein, total lipid and triglyceride levels were within normal physiological ranges (Kłopocki and Winnicka, 1992). This indicates that the addition of vegetable oils to diets did not change the metabolism of carbohydrates, protein or fat, whereas it distinctly lowered the blood cholesterol level (Brzóska et al., 1999 a,b). The addition of soyabean and sunflower oil significantly reduced total cholesterol and lipoprotein LDL. These results support the previously shown negative correlation between the level of unsaturated fatty acids in the diet and the synthesis of endogenous cholesterol in the liver of cows.
In summary, it can be said that feeding vegetable oils in an amount equivalent to about 1.5% of the ration's dry matter content (320 g/d) to cows characterized by a daily yield of about 25 kg of milk does not affect the content of saturated and unsaturated fatty acids in milk, but does increase the content of total CLA isomers to 1.0-1.2 g/100 g fatty acids and of its cis-9, trans-11 isomer to 0.80-1.05 g/100 g fatty acids in milk. This corresponds to a daily synthesis of both of these compounds in amounts of 8.7-10.4 g (CLA), respectively, and 7.0-8.6 g/cow (CLA cis-9, trans-11).