J. Dairy Sci. 85:190197 American Dairy Science Association, 2002.

Occurrence of trans-C18:1 Fatty Acid Isomers in Goat Milk: Effect of Two Dietary Regimens
M. LeDoux,* A. Rouzeau, P. Bas, and D. Sauvant
curite Sanitaire des Aliments (AFSSA), Laboratoire dEtudes et de *Agence Franc aise de Se et lHygie ` ne des Aliments, 94704 Maisons-Alfort, France Recherches sur la Qualite UMR INRA-INAPG Physiologie de la Nutrition et Alimentation, 75231 Paris Cedex 05, France

ABSTRACT Trans-octadecenoic acid composition of goat milk fat was studied by using silver-ion thin layer chromatography combined with gas-liquid chromatography. This analytical procedure also was used to investigate the effect of diet on trans-C18:1 fatty acid content in goat milk. Thirty-two goats were used in a 2 2 factorial arrangement and treatments. Two groups of goats received alfalfa hay at either a high or a low level of forage and two other groups received Rumiluz (dehydrated lon en Champagne, alfalfa from France Luzerne, Cha France) at either a high or a low level of forage. TransC18:1 isomer proportions (relative to total fatty acids) were, respectively, 2.02% for the Rumiluz low-level group and 1.75% for the Rumiluz high-level group versus 1.71% for the alfalfa low-level group and 1.21% for the alfalfa high-level group. Goats fed on Rumiluz thus produced relatively higher levels of trans-C18:1 fatty acids than animals fed alfalfa hay. The results also showed that production of trans-C18:1 fatty acids increased when the level of forage in the diet decreased. Moreover, goat milk trans-C18:1 composition appeared similar to the cow milk prole. Vaccenic acid, trans-11C18:1, was the major component and represented about 36.2% of total trans-C18:1 isomers. (Key words: trans-C18:1 fatty acids, milk fat, diet, goat) Abbreviation key: FA = fatty acid, FAIPE = fatty acid isopropyl ester, HA and LA = alfalfa hay at either a high level (60% DM basis) or a low level (30% DM basis) of forage, HR and LR = dehydrated alfalfa (Rumiluz) at either a high level (60% DM basis) or a low level (30% DM basis) of forage, PUFA = polyunsaturated fatty acid.. INTRODUCTION In recent years, high intake of mono-trans fatty acids (FA) has been associated with the risk of coronary heart

disease and myocardial infarction (Ascherio et al., 1994; Willet et al., 1993). In the human diet, the main source of mono-trans FA involved in heart disease risks is partially hydrogenated vegetable oils. Mono-trans isomers also occur naturally in ruminant milk fat, but at lesser rates and different isomer distribution than in partially hydrogenated oils (Precht and Molkentin, 1995b; Wolff, 1994). Several studies have reported cow milk fat trans FA composition and the effects of feed and dietary regimen on trans isomer production by cows (Jiang et al., 1996; Kalscheur et al., 1997; Precht and Molkentin, 1997; Stanton et al., 1997; Wolff, 1995). But, only a few articles have provided information on ovine or caprine milk fat related to trans FA content and its variations. Bickerstaffe et al. (1972) initially reported the existence of cis- and trans-C18:1 acids in goat milk fat. Recently, Wolff (1994) analyzed several goat cheeses for their trans-C18:1 acid composition, and Alonso et al. (1999) provided some data on FA composition of goat milk fat. Although the world production of goat milk seems relatively minor compared with the cow milk market, goat cheeses are widely consumed in some parts of the world such as France and some Mediterranean countries (Wolff, 1994). Thus, it is useful to obtain more information about composition of goat milk fat, and especially about its trans FA composition and the factors causing variation. Therefore, the aim of this work was to determine the trans-C18:1 isomer composition of goat milk fat and to study the range of variations in this composition related to the dietary regimen. MATERIALS AND METHODS Goats, Experimental Procedure, and Milk Sampling Thirty-two multiparous Saanen or Alpin goats in last third of the lactation period (DIM = 200; milk yield, 3.2 kg/d, SD = 0.80 kg/d; milk fat content, 31.8 g/kg, SD = 4.77 g/kg; milk protein content, 31.5 g/kg, SD = 3.85 g/ kg) were randomized to a 2 2 factorial arrangement and treatments (four groups and eight goats per group). Dietary factors were roughage type (alfalfa hay versus dehydrated alfalfa) and forage content (high versus low

Received March 12, 2001. Accepted September 27, 2001. Corresponding author: M. paris.afssa.fr.

LeDoux;

e-mail:

m.ledoux@

190

TRANS-C18:1 FATTY ACIDS IN GOAT MILK Table 1. Ingredient and chemical composition of diets. Diet Alfalfa hay Low DM (%) Ingredients Alfalfa Beet pulp Concentrate2 Nutrients CP NDF ADF Fatty acids (FA) C16:0 C18:0 C18:1 C18:1 isomers4 C18:2 C18:3 Others5 Total FA
1 2

191

Rumiluz1 High 91.68 (% of DM) Low 92.85 High 91.46

92.08

30 35 35

60 20 20 (% of DM)

30 35 35

60 20 20

13.69 41.99 23.15

13.75 50.49 32.19

14.37 40.92 20.30

14.19 45.11 26.26

(FAME3 % of DM) 0.542 0.062 0.350 0.021 0.957 0.267 0.039 2.238 0.554 0.063 0.245 0.017 0.766 0.277 0.094 2.016 0.643 0.085 0.434 0.025 1.111 0.322 0.098 2.718 0.611 0.089 0.333 0.019 0.924 0.617 0.126 2.719

lon sur Champagne, France. France Luzerne, Cha Mixture of corn, barley, and soybean meal, including vitamin and mineral mixture (3 and 4.5% of concentrate mixture for low and high level of forage, respectively. 3 FAME = Fatty acid methyl ester. 4 C18:1 isomers other than 9-cis oleic acid, included trans isomers. 5 Includes C14:0, C16:1, C17:0, C20:0, and so on.

proportions of forage). Two groups of goats received alfalfa hay at either a high level (HA group; 60% DM basis) or a low level (LA; 30% DM basis) of forage and two other groups received dehydrated alfalfa (Rumiluz, lon sur Champagne, France) at France Luzerne, Cha either a high level (HR; 60% DM basis) or a low level (LR; 30% DM basis) of forage. The remainder of the diet was composed of beet pulp silage and a concentrate mixture of corn, barley, soybean meal, and minerals and vitamins (Table 1). The 32 goats were housed individually and were milked twice a day. Goats were fed for ad libitum intake for 4 wk. Milk was randomly sampled from some goats before the experiment to check mean trans-C18:1 FA proportions and isomer distribution in milk fat of the herd. A second sampling of milk from all goats was done at the end of the fourth week of the experiment to study the effect of diet on the occurrence of transC18:1 isomers in milk fat. Samples were taken from two consecutive milkings (evening and morning) of each goat and mixed together in proportions (vol/vol) to the volume of each milking. All samples were kept frozen (<28C) and in the dark until lipid extraction.

Chemical Analysis Samples of the four TMR were oven-dried (model UL60; Memmert, Schwabach, Germany), ground in a s/ mill (1 mm screen; Gondard Productions, La Ferte Jouarre, France), and analyzed for NDF and ADF following the procedure outlined by Goering and Van Soest (1970) without using sodium sulte or decalin, and for CP by the Kjeldahl procedure (AOAC, 1990). The four TMR were also analyzed for long-chain FA by GLC as described by Gaynor et al. (1994) and Kalscheur et al. (1997), except that extracts were ltered on a sodium sulfate bed before evaporation and chromatography. The composition of the TMR are presented in Table 1. Lipid Extraction from Milk Milk fat extraction was based on the method described by Bas (1985). A 4-ml milk sample was transferred into a test tube, and 1.6 ml of ethanol, 0.4 ml of concentrated HCl, and 28 ml of hexane were added. After vigorous shaking, the solution was centrifuged at 200 g, and the upper phase was ltrated with a
Journal of Dairy Science Vol. 85, No. 1, 2002

192

LEDOUX ET AL.

Whatman P/S paper lter (Whatman, Maidstone, U.K.) into a ask. The lower phase was then extracted again with 20 ml of hexane. The supernatant was ltered and pooled with the previous extract. The hexane layer was then evaporated with a vacuum rotary evaporator at 30C. Preparation of FA Isopropyl Esters This procedure and the following ones were adapted mainly from Wolff et al. (1995) and from Precht and Molkentin (1995a). For preparation of FA isopropyl esters (FAIPE) of milk fat, the dry extract residue was dissolved in 2.5 ml of hexane/isopropanol (2:1, vol/vol) and transferred into a reaction tube. Then, 250 l of concentrated H2SO4 and 2 ml of isopropanol were added and the tube was tightly capped. After vigorous shaking, the reaction tube was incubated for 90 min at 100C. At the end of the reaction, the tube was cooled under tap water, and 5 ml of 1% aqueous solution of potassium chloride was added. The tube was shaken thoroughly and allowed to stand for about 1 min. The upper phase was decanted into a vial; the lower phase was then extracted twice with 2 ml of hexane. The three hexane fractions were pooled into the same vial and homogenized. A solution of 4 ml of FAIPE was evaporated under nitrogen and the dry residue was dissolved in 500 l of hexane. This FAIPE concentrated solution was used for TLC fractionation; the remainder of the FAIPE extract was used for total FA analysis. Fractionation of FAIPE by Using Silver-Ion TLC The FAIPE were fractionated according to the number and geometry of double bonds by silver-ion TLC. Commercial silica-gel plates (Merck, Darmstadt, Germany) were dipped in a 10% (wt/vol) AgNO3 solution in acetonitrile for 30 min and then partially air-dried and activated at 120C for 20 min. A 100-l solution of FAIPE was applied to the activated silver-ion TLC plate in a narrow band. The plate was developed in a tank containing hexane and diethyl ether (90:10, vol/vol) for up to 15-cm migration distance. The plate was air-dried and sprayed with a 0.2% (wt/vol) mixture of 95% ethanol solution of 2,7-dichlorouorescein and examined under UV light. The spots corresponding to the saturated and trans-monoenoic FAIPE were scraped off separately into centrifuge asks. The FAIPE were then extracted by adding successively 1.5 ml of methanol, 2 ml of hexane, and 1.5 ml of a 5% (wt/vol) aqueous solution of sodium chloride. The asks were shaken vigorously and then centrifuged at 200 g for 5 min. The upper phases were transferred into vials. The lower phases were then extracted again with 2 ml of hexane. The two hexane phases of both
Journal of Dairy Science Vol. 85, No. 1, 2002

saturated and trans FAIPE fractions were pooled respectively together and evaporated under nitrogen. The dry residue of the saturated FAIPE fraction was dissolved in 800 l of hexane; the dry residue of trans FAIPE fraction was dissolved in 100 l of the saturated FAIPE solution to use margaric acid (C17:0) as an internal standard for quantitation of trans FAIPE. GLC Analysis of Total and trans FA Total FAIPE chromatography was performed on an Autosystem model GC (Perkin-Elmer, Norwalk, CT) tted with a split injector (split ratio, 1:25) heated at 250C and a ame-ionization detector held at 270C. Analysis was performed with a BP21 capillary column (25 m 0.32 mm inside diameter, 0.25 m lm thickness; SGE France, Villeneuve, France) by using a temperature program. After injection, temperature was maintained at 45C for 2 min, then increased at a rate of 5C/min to 60C, then increased again at a rate of 8C/min to 250C and maintained at this value for 10 min. Helium was the carrier gas with a head pressure of 70 kPa. Injection volumes were 1 l. Tentative identication of FAIPE peaks was deduced from comparisons with CRM-164 reference material (Community Bureau of Reference, Brussels, Belgium) chromatograms, with saturated FA (C16:0, C17:0, C18:0, C19:0, and C20:0; Sigma, St. Louis, MO) standard solution chromatograms, and from spiking goat milk fat with the saturated FA standard solution (all FA as FAIPE form). Data from literature were helpful to identify FA peaks as well (Dorey et al., 1988). Trans FAIPE fractions were injected into a 3400CX chromatograph (Varian, Palo Alto, CA) equipped with a split injector (split ratio, 1:16) maintained at 250C and a ame-ionization detector held at 280C. Separations were performed on a CP-Sil-88 capillary column (100 m 0.25 mm inside diameter, 0.25 m lm thickness; Chrompack, Middelburg, The Netherlands) operated isothermally at 160C. The inlet pressure of the carrier gas (helium) was 220 kPa. Injection volumes were 1 l. Identication of trans-C18:1 isomer peaks was done by using the equivalent chain length measured with transC18:1 isomers (as FAIPE) kindly supplied by Dr. Robert L. Wolff, ISTAB, University of Bordeaux. To quantify the total proportions of trans-C18:1 isomers, the ratio of C17:0 to total trans-C18:1 was determined in the saturated plus trans-monoenoic FAIPE fraction and related to the C17:0 content of total FAIPE. Statistical Analysis ANOVA was applied to experimental data by using SAS software (SAS/STAT, 1987), according to the following model:

TRANS-C18:1 FATTY ACIDS IN GOAT MILK

193

Table 2. Equivalent chain length (ECL) values for trans-C18:1 fatty acid isomers (as fatty acid isopropyl esters at 160C on 100-m CPSil88). trans-C18:1 isomers trans-5 trans-6 trans-7 trans-8 trans-9 trans-10 trans-11 trans-12 trans-13 trans-14 trans-15 trans-16 Figure 1. Partial chromatogram of trans-monoenoic fatty acids from silver-ion TLC fractions showing trans-C18:1 positional isomers of goat milk fat (as isopropyl esters; 100-m CP-Sil 88, Chrompack, Middelburg, The Netherlands; 160C). Trans-4- (4t) and trans-5-C18:1 (5t) were detected above the detection limit but below the quantitation limit of the analytical method. ECL 18.25 18.29 18.30 18.30 18.32 18.34 18.40 18.43 18.46 18.48 18.54 18.77

Yijh = + i + j + ij + ijh i : [1,2] j : [1,2] where = overall mean, i = effect of forage type i (Rumiluz versus alfalfa hay), j = effect of forage content j (high versus low), ij = interaction between i and j, and ijh = residual error. The level of signicance of each analysis is included either in text or in tables. RESULTS A capillary GLC chromatogram of the trans-C18:1 FAIPE fraction is shown in Figure 1. The equivalent chain length values were determined for each trans-C18:1 isomer by using individual standards. The equivalent

chain length values are summarized in Table 2. Almost all trans-C18:1 FA isomers were sufciently separated to allow for an individual quantitation, except trans-6-8C18:1 and trans-13-14-C18:1, which were respectively quantied as groups. The coefcient of variation for the overall repeatability of trans-C18:1 FAIPE quantitation was 10% as measured according NF V03-110 AFNOR guidelines (Agence Franc aise de Normalisation, 1998). On the chromatogram of the total FA (not shown), the peak corresponding to butyric acid (C4:0) was completely separated from solvent peaks, and the margaric acid peak (internal standard for quantitation) was individualized as well, which was necessary for accurate quantitations (Wolff, 1995). Because no peak appeared on the blank chromatogram except for the solvent front, all peaks on total FA chromatograms were assumed to be FAIPE except solvent ones. Total FA in Goat Milk Fat The composition of FA in goat milk fat are shown in Table 3. The six most important FA in weight percentage (C10:0, C12:0, C14:0, C16:0, C18:0, and C18:1) accounted for almost 80% of total FA. These results do not differ very much from those reported by Alonso et al. (1999). Caprylic (C8:0) and capric (C10:0) acid contents were characteristics of goat milk FA pattern in all groups, higher than those of cow milk. Change from alfalfa hay to Rumiluz had small effects on C18:0 and on C18:1 group proportions and a signicant effect (P < 0.001) on C18:3 proportions. Modications of the dietary forage content had signicant effects (P < 0.01) on C10:0, C12:0, C18:1, and C18:3 percentages. Thus, the goats fed a high level of forage had lower values for C10:0, C12:0, and C18:2 and higher proportions for C18:1 and C18:3 than those fed the low level of forage.
Journal of Dairy Science Vol. 85, No. 1, 2002

Figure 2. Mean isomeric distribution (n = 24) of trans-C18:1 positional isomers in goat milk fat before experimentation (weight % of total trans-C18:1 fatty acid isopropyl esters). Trans-4- (4) and trans5-C18:1 (5) appeared only as traces.

194

LEDOUX ET AL. Table 3. Total fatty acid (FA) composition of goat milk fat in each group. Alfalfa hay Fatty acid Low (n = 8) 3.34 3.21 3.34 12.58 6.45 12.42 33.70 1.14 3.84 11.33 2.68 0.27 0.13 1.59 0.84 2.83 High (n = 8) 4.02 3.50 3.30 11.80 5.49 11.90 33.30 1.13 4.22 12.32 2.37 0.38 0.15 1.64 1.05 3.12 Rumiluz Low (n = 8) 3.74 3.69 3.42 12.55 6.04 11.84 32.00 1.07 4.13 11.84 2.69 0.45 0.12 1.57 0.89 2.71 High (n = 8) 3.54 3.37 3.13 11.31 4.79 11.56 33.41 1.04 5.25 13.40 2.66 0.66 0.15 1.41 0.85 2.75 RMSE1 Forage type Effects (P) Forage content Interaction

(% of total isopropyl esters) C4:0 C6:0 C8:0 C10:0 C12:0 C14:0 C16:0 C16:1 C18:0 C18:1 C18:2 C18:3 C20:0 iso-FA anteiso FA odd FA
1

0.58 0.54 0.35 0.98 1.00 0.94 3.09 0.26 0.86 1.14 0.27 0.10 0.04 0.18 0.15 0.30

NS NS NS NS NS NS NS NS * NS *** NS NS *

NS NS NS ** ** NS NS NS * ** ** NS NS NS

* NS NS NS NS NS NS NS NS NS NS NS NS * NS

RMSE = Root means standard error. < 0.10. *P < 0.05. **P < 0.01. ***P < 0.001.

Overall trans-C18:1 FA Composition of Goat Milk Fat Before Experimentation The average content of trans-C18:1 isomers in goat milk before the experiment (rst sampling) was 1.52% 0.14% (from 1.32 through 1.79%). Trans-C18:1 FA proportions in goat milk fat before experimentation and the distribution prole of these isomers are illustrated in Figure 2. Vaccenic acid (trans-11-C18:1) was found to be the major trans-C18:1 isomer in goat milk because it comprised 36.2% of total trans-C18:1 isomers (mean value, n = 24). Other isomers appeared in smaller proportions, from 5 through 12% depending on the isomer. Trans-4- and trans-5-C18:1 isomers were detected above the detection limit of the analytical method but below its quantitation limit. For this reason, these two minor isomers were not considered for quantitation and statistical treatments, but they will appear on gures as traces. Effect of Diet on trans-C18:1 Isomer Content in Goat Milk Mean and standard deviation values of trans-C18:1 content of milk fat for each group of eight animals are shown in Table 4. Except for the HA group, the values were higher than before the experiment. The highest value (2.02 g of trans-C18:1 FAIPE/100 g of total FAIPE) was obtained for the goat group that received Rumiluz at a low level of forage (LR group). Goats that
Journal of Dairy Science Vol. 85, No. 1, 2002

presented the lowest proportions of trans-C18:1 acids (1.24 %) were fed on alfalfa hay associated with a high level of forage (HA group). The ANOVA revealed a highly signicant effect (P = 0.002, Table 4) of forage type on trans FA proportions in milk fat. Mean values were 1.46 g of trans-C18:1 FAIPE/100 g of total FAIPE for the goat group fed the alfalfa hay (HA 1.71% and LA 1.24%) versus 1.88 g of trans-C18:1 FAIPE/100 g of total FAIPE for animals fed Rumiluz (HR 2.02% and LR 1.75%). Therefore, the Rumiluz diet led to a 29% increase of trans-C18:1 acid proportions in milk fat compared with an alfalfa hay diet. There was also a signicant effect of dietary forage content (P = 0.004, Table 4) on trans-C18:1 acid proportions. Average values were, respectively, 1.86 g of trans-C18:1 acids/100 g of total FAIPE for low level forage diets (LA 1.71% and LR 2.02%) and 1.48 g of trans-C18:1 FAIPE/100 g of total FAIPE for diets that contained a high level of forage (HA 1.24% and HR 1.75%). Thus, increasing the forage portion in diets led to a decrease in trans-C18:1 FA content in goat milk fat. Effect of Diet on Trans-C18:1 Isomer Prole in Goat Milk Despite noticeable variations, the overall trans-C18:1 distribution in milk fat of goats fed on different diets was comparable to the initial prole. Vaccenic acid

TRANS-C18:1 FATTY ACIDS IN GOAT MILK Table 4. Effect of forage type and forage content in the diet on total trans-C18:1 fatty acid (FA) content and on trans-C18:1 isomer proportions in milk fat. Alfalfa hay trans-C18:1 Total trans-68 trans-9 trans-10 trans-11 trans-12 trans-13-14 trans-15 trans-16
1 2

195

Rumiluz Low
2

Effects (P) RMSE

1

Low

High

High

Forage type

Forage content

Interaction

(% of total FAIPE ) 1.71 1.24 2.02 1.75 (% of total trans-C18:1 FAIPE2) 8.45 9.98 15.57 30.92 9.28 13.90 5.05 6.36 8.90 11.47 12.51 30.69 10.57 13.03 5.49 6.92 7.21 8.91 13.10 33.53 9.22 15.17 5.52 7.08 6.18 7.90 9.44 38.56 7.89 14.92 6.01 8.51 0.350 1.011 0.874 1.503 2.726 1.570 1.974 0.539 1.496 *** *** *** *** *** * * * * *** NS NS *** * NS NS * NS * ** NS * NS NS NS NS

RMSE = Root means standard error. FAIPE = Fatty acid isopropyl ester. P < 0.10. *P < 0.05. **P < 0.01. ***P < 0.001.

still remained the major component and represented from 28 to 45% of total trans-C18:1 FA. But, as shown in Figure 3, appreciable differences in individual isomer proportions occurred between the four groups. Change from the alfalfa hay diet to Rumiluz had a signicant effect (P < 0.05) on trans-C18:1 isomer proportions (Table 4). The Rumiluz diet led to a 6% increase in vaccenic acid; trans-13-14-, trans-15-, and trans-16-C18:1 followed slightly the same pattern.

Other isomers tended to decrease when goats were fed Rumiluz. Thus, unsaturation on the methyl side of the carbon chain seemed to be favored in the case of the Rumiluz diet. The forage content showed a signicant effect on only the following four trans-C18:1 isomers: 10, 11, 15, and 16. Vaccenic acid and, to a lesser extent, trans-15 and trans-16-C18:1 isomer proportions increased as dietary forage content increased, to a detri-

Figure 3. Mean isomeric distribution (n = 8 per group) of trans-C18:1 positional isomers (weight % of total trans-C18:1 fatty acid isopropyl esters) in milk fat of goats fed either an alfalfa diet at low (LA) or high (HA) forage content or Rumiluz diet at low (LR) or high (HR) forage content. Trans-4- (4) and trans-5-C18:1 (5) appeared only as traces. Journal of Dairy Science Vol. 85, No. 1, 2002

196

LEDOUX ET AL.

ment of trans-10-C18:1 FA, which decreased at the same time. When goats were fed diets rich in concentrate, trans-10-C18:1 isomer proportions increased signicantly, but vaccenic acid still remained the major isomer in any case. Interpretations of these ndings will necessitate complementary investigations of metabolic pathways of minor C18 isomer biohydrogenation. DISCUSSION The average trans-C18:1 FA content in goat milk fat before the experiment (1.52%) was lower than the values reported in literature. Alonso et al. (1999) observed a higher value for goat milk content of transC18:1 isomers (2.12%). Slightly higher values (2.68%, minimum 1.75% and maximum 4.50%, n = 7) were reported for goat cheeses by Wolff (1995). But it is not certain if the cheese FA content was strictly representative of milk fat composition, considering the differential lipolytic microbial activity during cheese process. LeDoux et al. (2000) recorded values from 2.5 to 3.8% reported in the literature for cow milk fat. The distribution of each trans-C18:1 isomer observed in goat milk fat before the experiment was close to those reported for goat cheeses by Wolff (1995) and for goat milk fat by Alonso et al. (1999). A similar prole was found in cow milk fat by Precht and Molkentin (1995a). Trans FA occurrence in milk is partly a consequence of rumen microbial activity. Rumen ora metabolize polyunsaturated fatty acids (PUFA), such as linoleic (C18:2) and linolenic (C18:3) acids from the diet, to stearic acid (C18:0) via a trans-C18:1 acid formation step. Linoleic acid is biohydrogenated to stearic acid via conjugated linoleic acid and trans-C18:1 formation (Chilliard et al., 2000; Griinari and Bauman, 1999). Linolenic acids, which are not direct precursors of conjugated linoleic acid, also increased the yield of ruminal vaccenic acid (trans-11 C18:1) by using another pathway including a step with production of conjugated C18:3 acids (Chilliard et al., 2000; Griinari and Bauman, 1999). The most likely hypothesis to explain the difference between the Rumiluz and the alfalfa hay diet on the effects of trans-C18:1 production in milk seems to be the initial substrate lipid composition. Lipid composition of the diet has a signicant inuence on ruminant milk fat (Chilliard et al., 2000; Kennelly, 1996). The Rumiluz diets had higher lipid contents than the alfalfa hay diets. In particular, Rumiluz contained more PUFA, similar to all early harvested grass. On the other hand, alfalfa hay came from a later cut in which lipid (and especially PUFA) stocks were slightly exhausted. Moreover, young plants (such as the ones
Journal of Dairy Science Vol. 85, No. 1, 2002

used in Rumiluz forage) are not rich in bers, which is another factor that contributes to an increase in trans FA content of milk fat (Chilliard et al., 2000; Jiang et al., 1996). These differences in initial substrate composition explain only part of the variation in trans-C18:1 acid content in both high and low forage diets. Low forage diets contained more linoleic acid than diets with high forage content (Table1). However, the linolenic acid contents of LA and HA diets were similar and the HR diet contained more C18:3 than the LR diet. Some additional complex mechanisms may also explain trans FA variations according to the forage level. The low forage diets led to increased trans-C18:1 acids for both alfalfa hay and Rumiluz. These results are in agreement with those of Jiang et al. (1996), who reported an increase from 0.24 to 0.33% of trans-C18:1 acids in milk fat of cows fed a diet with 50 and 65% concentrate, respectively. Decreasing the ber content and increasing the grain part of the diet (case of low forage LA and LR diets) would lead to a slowing down of the last biohydrogenation step of FA (C18:1 C18:0) and because of that, increasing the trans-C18:1 acid accumulation. High grain diets may favor low ruminal pH, which is also a factor that could result in the inhibition of the last step of PUFA biohydrogenation (Kalscheur et al., 1997). So, we checked the effect of forage content on stearic acid (C18:0) content in milk. Goats fed high forage diets (HA or HR groups) produced more stearic acid (signicant effect) than those receiving high concentrate diets (LA or LR) (Table 3). CONCLUSIONS The analytical method used for trans-C18:1 FA identication and quantitation was useful to determine and quantify these isomers in goat milk fat. We noticed variations in total trans-C18:1 content and in transC18:1 isomer proportions in goat milk fat in accordance with forage type and forage content of diets. What would explain the differences in milk trans-C18:1 isomer contents between Rumiluz and alfalfa hay groups was the initial PUFA content in diets. The lipid composition of diets had similar effects, but to a lesser extent, when comparing high vs. low forage contents. ACKNOWLEDGMENTS The authors thank Dr. R. L. Wolff (ISTAB, Bordeaux) for providing trans-C18:1 standards and for his precious advice in the analytical part of this work. We also address special thanks to Dr. P. Morand-Fehr and Dr. Ph. Schmidely (INRA, INA-PG, Paris) for the fruitful discussions on the topic and to M. Joseph Tes-

TRANS-C18:1 FATTY ACIDS IN GOAT MILK

197

sier (INRA, INA-PG, Grignon) for his technical work ` vrerie. Thanks also to Dr. Jayne Ireland in la che (AFSSA CIQUAL) for her valuable assistance in the preparation of the manuscript. REFERENCES
dure de Agence Franc aise de Normalisation (AFNOR). 1998. Proce thode alternative par rapvalidation intralaboratoire dune me ` une me thode de re fe rence. NF V03-110 AFNOR Guideport a lines, AFNOR, Saint-Denis, France. Alonso, L., J. Fontacha, L. Lozada, M. J. Fraga, and M. Juarez. 1999. Fatty acid composition of caprine milk: Major, branchedchain, and trans fatty acids. J. Dairy Sci. 82:878884. Ascherio, A., C. H. Hennekens, J. E. Buring, C. Master, M. J. Stampfer, and W. W. Willet. 1994. Trans-fatty acids intake and risk of myocardial infarction. Circulation 89:94101. Association of Ofcial Analytical Chemists (AOAC). 1990. Ofcial Methods of Analysis. 15th ed. AOAC, Arlington, VA. ` re grasse du lait pour Bas, P. 1985. Extraction rapide de la matie termination de sa composition en acides gras. Cah. Tech. la de INRA 9:2124. Bickerstaffe, R., D. E. Noakes, and E. F. Annison. 1972. Quantitative aspects of fatty acid biohydrogenation, absorption and transfer into milk fat in lactating goat, with special reference to the cisand trans-isomers of octadecenoate and linoleate. Biochem. J. 130:607617. Chilliard, Y., A. Ferlay, R. M. Mansbridge, and M. Doreau. 2000. The ruminant as a supplier of fatty acids for human consumption. Ann. Zootech. (Paris) 49:181205. Dorey, F., D. Brodin, J. F. Le Querler, and S. Kuzdzal-Savoie. 1988. e avec la Analyse des acides gras du beurre par CPG couple trie de masse. IAA juin:437442. spectrome Gaynor, P. J., R. A. Erdamn, B. B. Teter, J. Sampugna, A. V. Capuco, D. R. Waldo, and M. Hamosh. 1994. Milk fat yield and composition during abomasal infusion of cis or trans octadecenoates in Holstein cows. J. Dairy Sci. 77:157165. Goering, H. K., and P. J. Van Soest. 1970. Forage Fiber Analyses (Apparatus, Reagents, Procedures, and Some Applications). Agric. Handbook No. 379. ARS-USDA, Washington, DC. Griinari, J. M., and D. E. Bauman. 1999. Biosynthesis of conjugated linoleic acid and its incorporation into meat and milk in rumi-

nants. Pages 180200 in Advances in Conjugated Linoleic Acid Research. Vol. 1. M. P. Yurawecz, M. M. Mossoba, J. K. G. Kramer, M. W. Pariza, and G. J. Nelson, Ed. AOCS Press, Champaign, IL. Jiang, J., L. Bjoerk, R. Fonden, and M. Emalnuelson. 1996. Occurrence of conjugated cis-9, trans-11-octadecadienoic acid in bovine milk: Effect of feed and dietary regimen. J. Dairy Sci. 79:438445. Kalscheur, K. F., B. B. Teter, L. S. Piperova, and R. A. Erdman. 1997. Effect of dietary concentration and buffer addition on duodenal ow of trans-C18:1 fatty acids and milk fat production in dairy cows. J. Dairy Sci. 80:21042114. Kennelly, J. J. 1996. The fatty acid composition of milk fat as inuenced by feeding oilseeds. Anim. Feed Sci. Technol. 60:137152. ` res trans LeDoux, M., L. Laloux, and D. Sauvant. 2000. Les isome sence dans lalimentation. Sci. des acides gras: Origine et pre Aliments 20:393411. Precht, D., and J. Molkentin. 1995a. Variation of trans fatty acids in milk fat. Z. Ernaehrwiss. 34:2729. Precht, D., and J. Molkentin. 1995b. Trans fatty acids: Implication for health, analytical methods, incidence in edible fats and intake. Narhung 39:343374. Precht, D., and J. Molkentin. 1997. Effect of feeding on trans positional isomers of octadecenoic acid in milk fats. Milchwissenschaft 52:564568. SAS/STAT. SAS Users Guide: Statistical Analysis System, Version 6 Edition. 1987. SAS Inst., Inc., Cary, NC. Stanton, C., F. Lawless, G. Kjellmer, D. Harrington, R. Devery, J. F. Connoly, and J. Murphy. 1997. Dietary inuences on bovine milk cis-9, trans-11-conjugated linoleic acid content. J. Food Sci. 62:10831086. Willet, W. C., M. J. Stampfer, J. E. Manson, G. A. Colditz, F. E. Speizer, B. A. Rosner, L. A. Sampson, and C. H. Hennekens. 1993. Intake of trans fatty acids and risk of coronary heart disease among women. Lancet 341:581585. Wolff, R. L. 1994. Contribution of trans-18:1 acids from dairy fat to European diets. J. Am. Oil Chem. Soc. 71:277283. Wolff, R. L. 1995. Content and distribution of trans-18:1 acids in ruminant milk and meat fats. Their importance in European diets and their effect on human milk. J. Am. Oil Chem. Soc. 72:259272. Wolff, R. L., C. C. Bayard, and R. J. Fabien. 1995. Evaluation of sequential methods for the determination of butterfat acid composition with emphasis on trans-18:1 acids. J. Am. Oil Chem. Soc. 72:14711483.

Journal of Dairy Science Vol. 85, No. 1, 2002