Original Research Communications

trans Fatty acids in human milk are inversely associated with concentrations of essential all-cis n-6 and n-3 fatty acids and determine trans, but not n-6 and n-3, fatty acids in plasma lipids of breast-fed infants^1,2,3

² Supported by a grant from the Medical Research Council of Canada.

³ Address reprint requests to SM Innis, BC Research Institute for Children's and Women's Health, 950 West 28th Avenue, Vancouver, British Columbia V5Z 4H4, Canada. E-mail: sinnis{at}unixg.ubc.ca.

Key Words: Arachidonic acid • trans fatty acids • docosa-hexaenoic acid • human milk • breast-fed infants • breast milk • plasma lipids • n-6 fatty acids • n-3 fatty acids

Recent interest in the health effects of trans fatty acids has centered largely around potential adverse effects of trans fatty acids on lipid risk factors for cardiovascular disease, as well as metabolism of the essential (all-cis) n-6 and n-3 fatty acids, particularly in relation to infant growth and development (1–7). The major dietary sources of trans fatty acids in Western countries are hydrogenated fats and oils. Smaller amounts are consumed in foods from ruminants (eg, dairy fats, beef, and lamb) due to biohydrogenation of unsaturated fatty acids in the rumen. The extent of hydrogenation of dietary fats and oils depends on the unsaturated fatty acid content of the oil and the desired stability and physical properties of the product. Hydrogenation increases stability and the melting point by reducing the content of naturally occurring cis double bonds to give a complex mixture of geometric and positional isomers (8). The resulting fats and oils, in addition to containing trans fatty acids, have reduced amounts of linoleic acid (18:2n-6) and α-linolenic acid (18:3n-3). trans Fatty acids are well absorbed and are incorporated into tissue lipids, although predominantly into triacylglycerols rather than into phospholipids (8, 9). Estimates of average daily intakes by adults in the United States, Canada, Europe, and Australia based on food usage, food-frequency questionnaires, or duplicate portion analysis range from ≈3 to 17 g/person (8, 10–14).

Considerable evidence from published studies shows that the proportions of fatty acids in human milk are influenced by the trans, n-6, and n-3 fatty acid composition of the maternal diet (15–17). The composition of fatty acids in breast milk consumed by young infants is of concern because of the important roles of n-6 and n-3 fatty acids in infant growth and development (18–20). Some studies have noted an inverse relation between concentrations of trans fatty acids in fetal or infant tissue and measures of growth (5–7). It has also been suggested that inverse associations between trans fatty acids and arachidonic acid (20:4n-6), as well as the ratio of 20:4n-6 to 18:2n-6, reflect inhibition of desaturation of 18:2n-6 to 20:4n-6 (5–7). Although some animal and in vitro studies are consistent with the hypothesis that trans fatty acids may interfere with the desaturation of 18:2n-6 (4, 21, 22), it is possible that diets high in trans fatty acids are also low in all-cis n-6 and n-3 fatty acids.

Several studies have documented the presence of trans fatty acids in human milk (15, 16, 23–26), with concentrations possibly higher in Canada (25) than in Europe (23, 24, 26). Whether high milk concentrations of trans fatty acids result in high trans fatty acids and reduced 18:2n-6, 18:3n-3, 20:4n-6, or docosahexaenoic acid (22:6n-3) concentrations in the plasma lipids of breast-fed infants has not been clearly established. The objective of this study, therefore, was to determine in ≈100 breast-feeding mothers and their term infants the relations between all-cis n-6 and n-3 and trans fatty acid concentrations in the mothers' milk and in the plasma phospholipids and triacylglycerols of the breast-fed infants.

Subjects
This study involved 103 mothers who were exclusively breast-feeding their term (born at 37–41 wk gestation) infants. Infants with any evidence of a metabolic or physical abnormality, or who were given bottle feedings of formula or cow milk, were ineligible. A subset of the mothers (n = 62) agreed to allow collection of a venous blood sample from their infants at age 2 mo, concurrent with the provision of a milk sample. A separate subset of 21 mothers agreed to provide dietary information. No personal dietary instructions were given to any of the mothers. The procedures and methods used in this study were reviewed and approved by the Committee for Ethical Review of Research Involving Human Subjects at the University of British Columbia. Written, informed consent was obtained from all the women who participated.

Methods
Samples of breast milk (60–100 mL) were collected from all the women during the course of a feeding, at approximately the midpoint of the feeding, on the same day that the infants' blood samples or the mothers' 24-h dietary recalls were collected from subjects in the subsets providing this information.

The mothers who provided dietary information were interviewed by a nutritionist, who recorded their food intake for the preceding 24 h on the day when samples were collected. The 24-h dietary recall was used to train the mothers in how to keep a 3-d weighed food record. The importance of recording in detail the methods of food preparation, recipes, and brand names for all foods eaten was emphasized. The 3-d food records were completed during the following week, returned by mail, and reviewed by the nutritionist.

Venous blood samples were collected from the infants (n = 62) 2–3 h after a feeding, as close as possible to the anticipated time of the next feeding. Plasma was separated by low-speed centrifugation (2000 g x 15 min at 5°C) immediately after blood collection. Milk and plasma samples were stored at -70°C until analyzed. The milk samples were thawed in ice-cold water and were directly transmethylated to avoid potential losses of medium-chain saturated fatty acids (27).

The infants' plasma total lipids were extracted and then the triacylglycerols, phospholipids, and cholesteryl esters were separated by using thin-layer chromatography (TLC). The lipid fractions were recovered and the fatty acids were converted to their respective methyl esters (28) for separation by gas chromatography (GC) with a Varian 3400 gas chromatograph (Varian Canada Inc, Mississauga, Ontario), as described by Chen et al (25), that used an SP-2560 capillary column (100 m x 0.25 mm internal diameter, 20-µm film thickness) (Supelco, Bellefonte, PA). The CV of the GC method for the milk and infant plasma triacylglycerol and phospholipid fatty acid analyses was <1%. The GC procedure that we used does not separate all possible trans and cis-trans positional isomers (25, 29). Therefore, for the purposes of this report, concentrations of trans, cis-trans, and unusual cis positional isomers of the naturally occurring cis unsaturated fatty acids were calculated by summation and designated as total trans fatty acids.

The 3-d food records were analyzed to determine the average daily intake of total energy, fat, and trans fatty acids for each individual by using the FOOD PROCESSOR program (version 7.02; ESHA Research, Salem, OR), modified to include the total fat and fatty acid contents of ≈300 specific brand name food products, which were analyzed in the present study.

Statistical analysis
The significance of any relation between the concentration of a given fatty acid in milk and the concentration of the same fatty acid, or other fatty acids, in the infants' plasma triacylglycerols or phospholipids was examined by using correlation analysis for the results of matched analyses for milk and infants' plasma. The relations among the concentrations of trans fatty acids, 18:2n-6, 18:3n-3, 20:4n-6, and 22:6n-3 in the milk samples were similarly explored by using correlation analysis. A P value ≤0.05 was considered significant. Because the fatty acid concentrations were not normally distributed, the data were logarithmically transformed before statistical analysis. All statistical analyses were performed with the Statistical Package for the Social Sciences (version 7.5; SPSS Inc, Chicago).

Human milk and infant plasma triacylglycerol and phospholipid fatty acids
The composition of the major saturated, cis n-9 and n-7 monounsaturated, n-6, and n-3 fatty acids, conjugated linoleic acid (CLA), and other trans, unusual cis, and cis-trans isomers of unsaturated fatty acids in breast milk are shown in Table 1. The percentage distribution of fatty acids in the Canadian milk samples reported here are generally similar to those reported by Chen et al (25) for milk collected from women in 9 provinces in Canada. Our results show wide ranges for the concentrations of all fatty acids in milk; the samples were collected from ≈100 women who were all at a similar stage postpartum and were all exclusively breast-feeding their infants. Whereas mean percentages of 18:2n-6, 18:3n-3, 20:4n-6, and 22:6n-3 were 12.1%, 1.4%, 0.4%, and 0.2% of the total fatty acids, respectively, individual values varied widely among the women; eg, from 6.0% to 21.5% for 18:2n-6, from <0.1% to 4.1% for 18:3n-3, from 0.2% to 0.8% for 20:4n-6, and from 0.1% to 2.6% for 22:6n-3. Similarly, whereas the mean concentration of total trans fatty acids was 7.1%, the concentrations varied from 2.2% to 18.7% of milk fatty acids among the 103 women studied. The results (Table 1) also clearly show similar, wide interindividual variation in the concentrations of 14:0, 16:0, 18:0, cis 18:1n-9, and cis 18:1n-7 in the milk fatty acids.

The analyses of plasma triacylglycerol and phospholipid fatty acids in the 2-mo-old exclusively breast-fed infants also showed considerable interindividual variability in the concentrations of all-cis n-6 and n-3 fatty acids and trans fatty acids (Table 2). Mean (±SEM) trans fatty acid concentrations were higher in the infants' plasma triacylglycerols (6.5 ± 0.33%; range: 1.9–15.6%) than in the phospholipids (3.7 ± 0.16%; 1.7–8.3%), consistent with the higher incorporation of trans fatty acids into human tissue triacylglycerols than into phospholipids (9). In contrast, the concentrations of all-cis n-6 and n-3 fatty acids, with the exception of 18:3n-3, were consistently higher in the infant plasma phospholipids than the triacylglycerols. Comparison of the results for the human milk and infant plasma phospholipid and triacylglycerol fatty acid analyses (shown in Tables 1 and 2, respectively) showed that the means and ranges for the percentages of the major saturated, cis unsaturated, and trans fatty acid, and CLA concentrations were similar.

Relation between milk and infants' plasma fatty acids
The concentrations of total trans fatty acids and cis 18:1, 18:2n-6, 18:3n-3, 20:4n-6, and 22:6n-3 in human milk each showed a significant positive relation to the concentration of the same fatty acid in both the plasma triacylglycerols and phospholipids of the infants (Table 3, Figure 1). Similarly, the milk concentrations of the major saturated fatty acids (16:0 and 18:0) and other cis unsaturated fatty acids (16:1n-9, 16:1n-7, 20:2n-6, 20:3n-6, 22:4n-6, and 20:5n-3) each showed a significant positive relation to the concentration of the same fatty acid in both the plasma triacylglycerols and plasma phospholipids of the infants (Table 3).

The concentrations of trans fatty acids in milk showed significant inverse relations to the milk concentrations of 18:2n-6 (P = 0.02) and 18:3n-3 (P = 0.02) but not cis 18:1n-9, 20:4n-6, or 22:6n-3 (data not shown). There were no significant relations between the concentrations of trans fatty acids in milk and the concentrations of 20:4n-6 or 22:6n-3 in the infants' plasma triacylglycerols or phospholipids. Similarly, there were no significant relations between the milk concentration of 18:2n-6 and the infants' plasma triacylglycerol or phospholipid concentrations of 20:4n-6 or 22:6n-3, or between milk 18:3n-3 concentrations and the infants' plasma triacylglycerol 20:4n-6 or 22:6n-3 concentrations. In contrast, the concentration of 18:3n-3 in the milk showed a significant inverse relation to the 20:4n-6 concentration (P = 0.05) and a positive relation to the 22:6n-3 concentration (P = 0.02) in the infants' plasma phospholipids.

Major dietary sources of trans fatty acids
Mean (±SEM) concentrations of the major saturated and unsaturated fatty acids were not significantly different between the mothers who provided 3-d weighed dietary records and the mothers who did not (data not shown). Total trans fatty acids in the milk of women who provided dietary records was 8.0 ± 0.80% (range: 3.4–18.7%) of total fatty acids. Fat represented ≈31.8% of the total daily energy intake (: 10496 ± 597 kJ) with a range of 12.1–44.8% of daily energy (27.7–140.6 g) from fat among the mothers who provided dietary records (Table 4). Two of the 21 mothers reported that they ate no meat, fish, shellfish, or eggs and 1 mother reported no meat or eggs but did eat fish during the 3 d for which records were kept. Nine of the 21 mothers had at least one serving of fish or other seafood. The estimated daily intake of trans fatty acids was 6.87 g/person, representing ≈7.7% of total fat and ≈2.46% of daily energy intake. Of the total trans fatty acid intake, ≈78% (5.34 g/d) was derived from prepared and processed foods, with the major sources being bakery products such as cakes, cookies, pies, and muffins (≈22% of daily trans fatty acid intake), snack foods (≈14.4%), fast foods (≈11.1%), and breads and rolls (≈10.5%). Margarines contributed ≈11.1% of daily trans fatty acids, but this varied widely among the mothers. The use of tub (soft) margarines as a table spread or in cooking ranged from 0 g fat/d (n = 9) to 23.64 g fat/d.

This study showed a wide range (up to 18.7%) of trans fatty acid concentrations in milk from breast-feeding Canadian mothers. The results also showed that the percentage of trans fatty acids in milk was closely paralleled (P < 0.001) by the percentages of trans fatty acids in the plasma triacylglycerols and phospholipids of the breast-fed infants. The percentage of trans fatty acids in plasma phospholipids was ≈40% lower than that in plasma triacylglycerols, possibly because we collected blood samples when infants were in the fed state and because of the specificity of the acyltransferases involved in phospholipid synthesis. Similarly, the percentages of 18:2n-6, 18:3n-3, 20:4n-6, and 22:6n-3 in human milk were each significantly related to the percentage of the same fatty acid in the infants' plasma triacylglycerols and phospholipids. The percentage of trans fatty acids in milk was inversely related (P < 0.05) to the percentages of 18:2n-6 and 18:3n-3 in milk. Similar inverse relations between trans fatty acids and 18:2n-6 and 18:3n-3 in milk were noted previously (25). It is possible that women with high intakes of foods containing hydrogenated oils have lower intakes of all-cis 18:2n-6 and 18:3n-3, either from salad and cooking oils or from margarines.

Similar to our study, another study found a range for trans fatty acids in human milk of 0.1–17.2% (: 7.2%) for breast-feeding women in different regions of Canada (25). However, studies in Spain (n = 38) found mean concentrations of ≈1% trans 18:1 (23), whereas milk from women in France (n = 10) had ≈2% (range: 1.2–3.0%) trans fatty acids (24) and milk from women in Germany had a median of 4.4% (range: 2.2–6.0%) trans fatty acids (26). The intake of hydrogenated fats and oils in baked and other prepared foods may possibly be higher in Canada than in Europe.

In contrast with our results, Chen et al (25) found an inverse relation between trans 18:1 and cis 18:1n-9 in Canadian human milk. One explanation for the lack of a relation between trans fatty acids and cis 18:1n-9 in our study may be incomplete separation of trans 18:1 from cis 18:1n-9 by GC. Studies using partially hydrogenated margarines and synthetic unsaturated fatty acids have shown 9–30% overlap of t-18:1Δ12 to t-18:1Δ16 with cis 18:1 in GC with 100-m capillary columns, with the degree of overlap dependent on the total trans fatty acid content (29). The most prevalent isomer of 18:1 in human milk is t-Δ11 (23% 18:1 isomers), followed by t-Δ10 (20%) and t-Δ9 (16%), with smaller amounts of t-Δ12 and t-Δ13. Similarly, ≈43% of 18:1 trans isomers in cow milk fat are t-18:1Δ11, with low proportions of other isomers (25, 29). Our analysis did not include separation of 18:1 isomers (eg, by AgNO₃ -TLC) before GC. However, because the most common 18:1 isomers in hydrogenated vegetable oils are t-Δ11, t-Δ10, and t-Δ9, with relatively small amounts of t- Δ12 to t-Δ16, it seems probable that any overestimation of cis 18:1n-9 (or underestimation of trans fatty acids) was relatively small. Indeed, comparison of direct GC and AgNO₃-TLC followed by GC of milk fatty acids showed mean (±SD) cis 18:1n-9 values of 32.7 ± 3.2% and 30.6 ± 2.7%, respectively, and total trans 18:1 values of 4.6 ± 2.0% and 5.9 ± 25% total fatty acids, respectively, with no differences in cis n-6 and n-3 fatty acids with the 2 methods (25). A more probable reason for the lack of a relation between milk trans fatty acids and cis 18:1n-9 was that the major food sources of trans fatty acids in the study were bakery products, snacks, and fast foods. It seems probable that in these foods, trans fatty acids replaced saturated fats (eg, butter, lard, or saturated vegetable oils) rather than cis 18:1n-9–rich margarines or oils.

In contrast with previous studies (5–7), we found no association between trans fatty acids and 20:4n-6 or 22:6n-3 in milk or plasma phospholipids or triacylglycerols of breast-fed infants. Chen et al (25) also found no differences in the percentages of 20:4n-6 or 22:6n-3 in human milk samples grouped as milks with low (<4%), medium, or high (>10%) percentages of trans fatty acids. In our study, the percentages of 20:4n-6 and 22:6n-3, but not of trans fatty acids, in milk were significantly related to the percentages of 20:4n-6 and 22:6n-3, respectively, in the infants' plasma lipids. It may be that intakes of preformed 20:4n-6 and 22:6n-3 are a more important determinant of milk and blood lipid 20:4n-6 and 22:6n-3 than are any potential effects of inhibition of 18:2n-6 and 18:3n-3 desaturation by trans fatty acids. Some, but not all, studies have reported a relation between blood lipid 20:4n-6 and 22:6n-3 concentrations of young infants and measures of growth and visual development, respectively (19, 20). Whether the wide ranges of concentrations of 20:4n-6 (8.1–15.8%) and 22:6n-3 (2.2–8.0%) in the plasma phospholipids of breast-fed infants have any physiologic significance regarding growth or development is not known. However, because 20:4n-6 and 22:6n-3 are found in animal but not plant foods, mothers with higher 20:4n-6 and 22:6n-3 concentrations in their milk are likely to have intakes of animal, fish, and egg protein that are different from those of women with lower milk 20:4n-6 and 22:6n-3 concentrations. This suggests that a detailed analysis of nutrient intakes of a larger group of women than our sample is needed to investigate any potential relations of maternal diet to infant growth and development.

Our results also showed a relation between CLA concentrations in milk and in the plasma lipids of breast-fed infants. CLAs are positional and geometric isomers of 18:2n-6 that occur naturally in several foods, particularly dairy products and beef, and that appear to have biological activity (30, 31). In contrast with other trans fatty acids, CLAs were preferentially accumulated, by ≥2-fold, in the infants' plasma phospholipids rather than in triacylglycerols. Animal studies have also suggested possible tissue specificity for CLA (32). Whether CLAs have any physiologic effects on breast-fed infants is not yet known.

Craig-Schmidt et al (16) found that the concentration of trans 18:1 in milk increased from 1.8% to 6.5% of total fatty acids in 8 women when trans 18:1 in the diet increased from 1.0% to 11.8% of fatty acids. Using the equation generated in that study (16) to describe the relation between the percentage of trans 18:1 in the milk and in the diet, Chen et al (25) estimated daily trans fatty acid intakes of 7.7%, 3.9%, and 1.1% of total energy (equivalent to 20.3, 10.1, and 3.0 g, respectively) for Canadian women with high, medium, and low concentrations of trans fatty acids in their milk. Our analysis of 3-d diet records estimated a mean (±SEM) trans fatty acid intake of 6.9 ± 1.1 g/d (range: 1.3–10.9 g/d), representing ≈2.5% of energy intake and 7.7% of total fat intake. The finding that the mean (±SEM) percentage of trans fatty acids in milk (7.1 ± 0.3%) was similar to that calculated for the diet suggests that the estimated value of 6.9 g trans fatty acids/d is a reasonable estimation of the trans fatty acid intake of the women in this study. The major food sources of trans fatty acids for the women in our study were bakery products, snacks, and fast foods, with margarine and shortening contributing only ≈11% of the mean daily trans fatty acid intake (0.76 g). Similarly, data for the United States suggest that breads, cakes, cookies, snacks, and fried foods are likely to be major, although variable, sources of dietary trans fatty acids (8).

Numerous studies have shown that human milk fatty acids, particularly monounsaturated, n-6, n-3, and trans fatty acids, change in response to changes in maternal dietary intake (15–17). Concerns about dietary fat and health or lifestyle issues that alter maternal food choices may therefore influence the quality of fatty acid nutrition of breast-fed infants. It is particularly important to note that our results identified baked and prepared foods and breads as major sources of trans fatty acids, providing ≈50% of the total trans fatty acid intake but <20% of the total fat intake. The lack of labeling regarding fat composition and the possibility that many of these foods may be perceived as healthy in the context of lower-fat diets warrant consideration.

We gratefully acknowledge the assistance of Ellen Mackay in the collection of the dietary information.

1. 1. ASCN/AIN Task Force on Trans Fatty Acids. Position paper on trans fatty acids. Am J Clin Nutr 1996;63:663–70.[Abstract/Free Full Text]

• 2. Ascherio A, Willett WC. Health effects of trans fatty acids. Am J Clin Nutr 1997;66(suppl):1006S–10S.[Abstract/Free Full Text]

• 3. Khosla P, Hayes KC. Dietary trans-monounsaturated fatty acids negative impact on plasma lipids in humans: critical review of the evidence. J Am Coll Nutr 1996;15:325–39.[Abstract]

• 4. Kinsella JE, Bruckner G, Mai J, Shimp J. Metabolism of trans fatty acids with emphasis on the effects of trans, trans-octadecadienoate on lipid composition, essential fatty acid, and prostaglandins: an overview. Am J Clin Nutr 1981;34:2307–18.[Abstract/Free Full Text]

• 5. Desci T, Koletzko B. Do trans fatty acids impair linoleic acid metabolism in children? Ann Nutr Metab 1995;39:36–41.[Medline]

• 6. Koletzko B. Trans fatty acids may impair biosynthesis of long-chain polyunsaturates and growth in man. Acta Paediatr 1992;81:302–6.[Medline]

• 7. Koletzko B. Potential adverse effects of trans fatty acids in infants and children. Eur J Med Res 1995;1:123–5.[Medline]

• 8. Emken EA. Trans fatty acids and coronary heart disease risk: physiochemical properties, intake, and metabolism. Am J Clin Nutr 1995;62(suppl):659S–69S.[Free Full Text]

• 9. Ohrigge JB, Gulley RM, Emken EA. Occurrence of octadecenoic fatty acid isomers from hydrogenated fats in human tissue lipid classes. Lipids 1982;17:551–7.[Medline]

• 10. Enig MG, Subodh A, Keeney M, Sampugna J. Isomeric trans fatty acids in the US diet. J Am Coll Nutr 1990;9:471–6.[Abstract]

• 11. Hunter JE, Applewhite TH. Reassessment of trans fatty acid availability in the US diet. Am J Clin Nutr 1991;54:363–9.[Abstract/Free Full Text]

• 12. Mansour MP, Sinclair AJ. The trans fatty acid and positional (sn-2) fatty acid composition of some Australian margarines, dairy blends and animal fats. Asia Pac J Clin Nutr 1993;3:155–63.

• 13. Nestel PJ. Effect of trans fatty acids on serum lipoprotein levels in man. World Rev Nutr Diet 1991;75:187–9.

• 14. Willett WC, Stampfer MJ, Manson JA, et al. Intake of trans fatty acids and risk of coronary heart disease among women. Lancet 1993;341:581–5.[Medline]

• 15. Chappell JE, Clandinin MT, Kearney-Volpe C. Trans fatty acids in human milk lipids: influence of maternal diet and weight loss. Am J Clin Nutr 1985;42:49–56.[Abstract/Free Full Text]

• 16. Craig-Schmidt MC, Weete JD, Faircloth SA, Wickwire MA, Livant EJ. The effect of hydrogenated fat in the diet of nursing mothers on lipid composition and prostaglandin content of human milk. Am J Clin Nutr 1984;39:778–86.[Abstract/Free Full Text]

• 17. Innis SM. Human milk and formula fatty acids. J Pediatr 1992;120(suppl):S56–61.[Medline]

• 18. Innis SM. Essential fatty acids in growth and development. Prog Lipid Res 1991;30:39–103.[Medline]

• 19. Innis SM. Polyunsaturated fatty acid nutrition in infants born at term. In: Dobbing J, ed. Developing brain and behaviour. New York: Academic Press, 1997:103–67.

• 20. Carlson SE. Long-chain polyunsaturated fatty acid supplementation of preterm infants. In: Dobbing J, ed. Developing brain and behaviour. New York: Academic Press, 1997:41–102.

• 21. Hoy CE, Holmer G. Influence of dietary linoleic acid and trans fatty acids on the fatty acid profile of cardiolipins in rats. Lipids 1990; 25:455–9.[Medline]

• 22. Zevenberger JL, Hadderman E. Lack of effects of trans fatty acids on eicosanoid biosynthesis with adequate intake of linoleic acid. Lipids 1989;24:555–63.[Medline]

• 23. Boatella J, Rafecas M, Codony R, et al. Trans fatty acid content of human milk in Spain. J Pediatr Gastroenterol Nutr 1993; 16:432–4.[Medline]

• 24. Chardigny J-M, Wolff RL, Mager E, Sebedio J-L, Martine L, Juaneda P. Trans mono- and polyunsaturated fatty acids in human milk. Eur J Clin Nutr 1995;49:523–31.[Medline]

• 25. Chen Z-Y, Pelletier G, Hollywood R, Ratnayke WMN. Trans fatty acid isomers in Canadian human milk. Lipids 1995;30:15–21.[Medline]

• 26. Koletzko B, Mrotzek M, Bremer HJ. Fatty acid composition of mature human milk in Germany. Am J Clin Nutr 1988;47:954–9.[Abstract/Free Full Text]

• 27. Lepage G, Roy CC. Direct transesterification of all classes of lipids in a one-step reaction. J Lipid Res 1986;27:114–20.[Abstract]

• 28. Innis SM, Dyer R, Nelson CM. Evidence that palmitic acid is absorbed as sn-2 monoacylglycerol from human milk by breast-milk infants. Lipids 1994;29:541–5.[Medline]

• 29. Ratnayake WMN, Beare-Rogers JL. Problems of analyzing C₁₈cis- and trans-fatty acids of margarine on the SP-2340 capillary column. J Chromatogr Sci 1990;28:633–9.

• 30. Chin SF, Liu W, Storkson JM, Ha YL, Pariza MW. Dietary sources of conjugated olienoic isomers of linoleic acid, a newly recognized class of anticarcinogens. J Food Composition Anal 1992;5:185–97.

• 31. Li Y, Watkins BA. Conjugated linoleic acids alter bone fatty acid composition and reduce ex vivo prostaglandin E₂ biosynthesis in rats fed n–6 or n–3 fatty acids. Lipids 1998;33:417–25.[Medline]

• 32. Watkins BA, Shen CL, McMurtry JP, et al. Dietary lipids after histomorphometry and concentrations of fatty acids and insulin-like growth factor in chick tibiotarsal bone. J Nutr 1997;127:1084–91.[Abstract/Free Full Text]