Fatty Acid Composition of Human Milk in Western Iran

ORIGINAL COMMUNICATION Fatty acid composition of human milk in Western Iran

G Bahrami1* and Z Rahimi2

1Department of Pharmacology, Medical Biology Research Center, Medical School, Kermanshah University of Medical Sciences, Kermanshah, Iran; and 2Department of Biochemistry, Medical Biology Research Center, Medical School, Kermanshah University of Medical Sciences, Kermanshah, Iran

Objective: To investigate the fatty acid composition of mature human milk in Western Iran with special focus on trans fatty acids. Design: Observational study. Methods: Milk samples were collected from 52 lactating mothers aging 19–39 y, from Western Iran. Subjects were asked to complete a diet questionnaire. Milk fatty acids were measured as 2-nitrophenylhydrazide derivatives by high-performance liquid chromatography. Results: Saturated fatty acids were the main fraction of human milk (41.3%). Medium-chain fatty acids (C8:0-C14:0) constituted 24%, oleic acid (C18:1o9) accounted for 30.9% and elaidic acid (C18:1T), the trans isomer of oleic acid, comprised 11.3% of the total milk fatty acids. Linoleic (C18:2o6) and linolenic (C18:3o3) acid contents were 13.8 and 1.1%, respectively. The level of the polyunsaturated fatty acids was 1.4% for arachidonic (C20:4o6) and 0.2% for eicosapentaenoic (C20:5o3) acid. Conclusions: The milk from Iranian lactating mothers, as compared to that from the American or European mothers, contained high levels of medium-chain and trans fatty acids. This difference may be attributed to the maternal diet with low animal protein and animal fat but with high carbohydrate and partially hydrogenated vegetable oils that carry large amounts of trans fatty acids. As the detrimental effects of trans fatty acids on blood lipids and cardiovascular diseases have been emphasized in the literature, a reduction of trans fatty acid content in the diet of Iranian mothers is suggested. Sponsorship: Kermanshah University of Medical Sciences. European Journal of Clinical Nutrition (2005) 59, 494–497. doi:10.1038/sj.ejcn.1602099 Published online 12 January 2005

Keywords: fatty acids; lactation; human milk; trans fatty acids

Introduction milk (Francois et al, 1998). The influence of diet on milk fatty Fat in the breast milk is a major source of energy, essential acid composition has been established by several studies fatty acids and fat-soluble vitamins required for growth and (Aitchison et al, 1977; Boersma et al, 1991; Lawrence, 1994; development of infants (Finley et al, 1985; Broschel et al, Francois et al, 1998). 1986; Lawrence, 1994). Despite the individual and diurnal It would be expected that mothers on low-fat and low- variation in fat content of human milk, the fatty acid energy diets would produce a milk richer in endogenous composition shows great similarity among well-nourished fatty acids as compared to mothers receiving more dietary women from different populations. Dietary intake, mobiliza- proteins and fats. It is known that mammary cells are capable tion of body fat stores and endogenous synthesis by of synthesizing predominantly short- and medium-chain mammary glands are the sources of fatty acids in human length fatty acids. Therefore, these fatty acids would be expected to be found with higher proportions in milk of mothers on low-energy diets (Aitchison et al, 1977; Boersma *Correspondence: G Bahrami, Kermanshah University of Medical et al, 1991). Sciences, Pharmacology, Kermanshah 6714869914, Iran. The presence of trans fatty acids in human milk must be of E-mail: [email protected] Guarantor: G Bahrami. dietary origin, because they can not be synthesized by Contributors: GB has designed the project and was responsible for the humans (Aitchison et al, 1977). Detrimental effects of these study. GB and ZR were involved in analysis of the samples and writing fatty acids have frequently been reported (Aitchison et al, of the paper. ZR performed samples collection and statistical 1977; Jensen et al, 1992; Sardesai, 1992; Lawrence, 1994) and calculation of the data. Received 19 April 2004; revised 15 September 2004; accepted 25 October include alteration in membrane permeability, inhibition of 2004; published online 12 January 2005 certain enzymatic reactions in lipid metabolism, prevention Fatty acid in milk G Bahrami and Z Rahimi 495 of fatty acid elongation, increased total and LDL-cholesterol, tively. Serum BUN level was determined by diacetyl mono- decreased HDL-cholesterol, and disturbance in arachidonic xime method (Evans, 1968). Total lymphocyte count was acid and prostaglandin metabolism. Also impairment of performed using a cell counter (Baker system 9000). growth and development of infants has been reported to be Total lipids of milk were extracted by chloroform–metha- caused by trans fatty acids (Craig-Schmidt, 2001). nol based on the Folch method (Folch et al, 1957). Analysis Since the trans fats have been estimated to exist in of fatty acids was performed by the HPLC method, which is relatively high levels (up to 38%) in hydrogenated vegetable described in detail elsewhere (Miwa & Yamamoto, 1991) oils used in common Iranian food, the level of these fatty subsequent to derivatization by 2-nitrophenylhydrazide. acids in human milk is of interest. Therefore, the present Briefly, gradient elution of methanol and water as the mobile study was aimed at the fatty acid composition of mature phase and ODS column kept at 601C as the stationary phase human milk in Western Iran with special focus on trans fats were used and the signals were quantified by UV-VIS detector as well as short- and medium-chain fatty acids considering operated at 370 nm after correction of the response factor. the influence of dietary habit of the mothers.

Results Subjects and methods Anthropometry and biochemical characteristics of all sub- Subjects jects are shown in Table 1. Based on anthropometric A total of 52 healthy lactating women aged 19–39 y characteristics, severe overweight (BMI 435 kg/cm2) did participated in this study. They were randomly selected from not exist in our volunteers, however mild overweight (BMI mothers referring to health centers of Kermanshah Uni- 30–35 kg/cm2) and mild underweight (BMI 18–20 kg/cm2) versity of Medical Sciences for periodical growth evaluation were present in five and nine individuals, respectively. With of their infants. All the subjects were in the 45th to 135th respect to the biochemical parameters and clinical malnutri- day of their lactation; living in Kermanshah city (in Western tion definition (Miranda et al, 1983; Weinsier et al, 1989), Iran) and all babies were full-term infants (37–40 weeks of maternal malnutrition did not exist in the subjects. gestation). None of the subjects were dieting or lost weight All subjects were interviewed and asked questions about during the study. The study protocol was approved by the their eating habits and consumed edible oils. Complex Ethical Committee of the university and all the subjects gave carbohydrates including bread, rice and grains were the informed written consent prior to start of the study. They most common constituents (about 77% of energy) in the diet were asked to complete a qualitative questionnaire contain- of subjects, while animal fats and proteins including beef ing questions on the main constituents of their diet and type and chicken meats, diary products and fat-rich meals of oils consumed. Anthropometric characteristics of weight comprised a minor portion in their diet (around 8% energy and height were measured for all mothers. To determine if from protein and 15% from fat mostly from vegetable the women were undernourished, blood sampling and sources). As edible oil, only hydrogenated vegetable oils analysis of serum albumin, blood urea nitrogen (BUN) and were consumed by 85% of subjects, although, nearly all lymphocyte count were carried out in all the subjects. subjects received considerable amounts of hydrogenated oil through various foods prepared by these oils (snacks, confectionery, biscuit and the foods which served in Milk sampling restaurant). Milk sample (about 10 ml) was expressed directly into plastic The fat content in the milk samples was 3.871.5%. Fatty containers and placed in an ice bucket for transfer to the acid composition of the milk for all mothers is depicted in laboratory. All the samples were collected at a midmorning Table 2. The medium-chain fatty acids including C:8 to C:14, feeding before lunch time and due to differences in the fat and trans isomer of oleic acid (elaidic acid) accounted for 24 content of the fore and hind milk, mothers expressed their and 11.3% of total milk fatty acids, respectively. The total breasts as much as possible at the time of sampling. Each polyunsaturated fatty acids (PUFA) including linoleic acid milk sample was immediately divided into five different tubes, and stored frozen at À401C until analysis. To minimize oxidative decay of polyunsaturated fatty acids, long-term Table 1 Anthropometry and biochemical characteristics of subjects storage and exposure of the samples to high temperature and Mean7s.d. Minimum Maximum bright light were avoided and the samples were analyzed as Characteristics n ¼ 52 value value soon as possible. Age (y) 25.777.2 19 39 BMI (kg/cm2) 23.1172.72 17.19 32.42 Albumin (g/dl) 4.670.82 3.1 6.45 Analysis Protein (g/dl) 6.470.78 5.09 8.19 Total serum protein and albumin concentrations were BUN (mg/dl) 11.074.1 5.46 26.5 7 measured using the procedures of Lowry (Lowry et al, Lymphocyte 2399 663 1406 3996 count (mm3) 1951) and bromcresol green (Doumas et al, 1971), respec-

European Journal of Clinical Nutrition Fatty acid in milk G Bahrami and Z Rahimi 496 Table 2 Levels of fatty acids in breast milk Americans (Broschel et al, 1986; Jensen et al, 1992) to low Fatty acids N % of total fatty acids amounts (4.4%) as in Western German (Koletzko et al, 1988), Spanish and French women (Boatella et al, 1993; Chardigny C6:0 28 1.971.7 et al, 1995). Linoelaidic (18:2TT) and elaidic acid (18:1T) are 7 C8:0 51 1.2 1 the main trans fatty acids produced during partial hydro- C10:0 52 4.271.9 C12:0 52 11.374.5 genation of vegetable oils; however, linoelaidic acid (18:2TT) C14:0 52 7.372.6 was not detectable in about half of our samples. Thus, elaidic C16:0 52 11.473.6 is considered as the trans fat (Table 2). The mean trans fatty 7 C18:0 51 4.0 1.5 acid content in the milk from our subjects was 11.3% of the C18:1T 49 11.373.4 C18:1o9 52 30.976.5 total fatty acids. This is high as compared with other C18:2o6 52 13.874.1 populations and is explained by the consumption of C18:3o3521.170.4 partially hydrogenated vegetable oils with high content of o 7 C20: 4 6401.40.35 trans fats (up to 38% of total fatty acids) in our samples. A C20:5o3310.270.13 large number of metabolic and epidemiologic studies have N ¼ number of mothers who specific fatty acid is detected in their milk. indicated the adverse effects of trans fatty acids with respect to coronary artery diseases (Aitchison et al, 1977; Jensen et al, 1992; Sardesai, 1992; Lawrence, 1994) as well as growth and (C18:2o6), linolenic acid (C18:3o3), arachidonic acid development of infants (Innis & King, 1999). Thus, the (C20:4o6) and eicosapentaenoic acid (C20:5o3) constituted intake of hydrogenated fats should be restricted as much as 16.5% of the total milk fatty acids. possible in the Iranian diets and general population should be aware of the adverse health effects of hydrogenated oils. In contrast to medium-chain and trans fatty acids, palmitic Discussion (C16:0) and stearic (C18:0) acids tend to be somewhat lower Results from this study show that the milk from Iranian in our study (11.4 and 4.0%, respectively) than in others women contains high level of medium-chain and trans fatty (20.9–22.5 and 7.3–7.44%, respectively, Finley et al, 1985; acids as compared to the milk from American and European Broschel et al, 1986; Lawrence, 1994). This probably reflects a women. It is well established that the fatty acid composition lower intake of animal protein and fat but higher consump- of the breast milk is affected by the diet and endogenous tion of complex carbohydrates in our individuals (Finley et al, synthesis in mammary glands. It is known that a maternal 1985). diet low in protein and fat but rich in carbohydrate causes an The content of essential fatty acids, linoleic (C18:2o6) and enhanced synthesis of short- and medium-chain length fatty linolenic (C18:3o3) acids, in human milk is also related to acids by the mammary gland. It would therefore be expected the maternal diet (Finley et al, 1985; Sanders & Reddy, 1992; to find these fatty acids in a higher proportion in milk of Lawrence, 1994). For milk linoleic acid content, the range of mothers on low-fat, low-energy diets (Aitchison et al, 1977; 1–19% was found in a comparative study of Cuthbertson Boersma et al, 1991). Higher levels of medium-chain fatty (1976) on different populations and in another study the acids (24%) and lower amounts of palmitic (11.4%) and range reported to be 1–45% with an average value of 10–15% stearic acids (4.0%) were found in the milk samples from our (Finley et al, 1985). In the present study, however, the mean subjects as compared to those values reported for the of 13.8% (9.7–17.9%) was obtained. For linolenic acid, the Americans, Egyptians (Broschel et al, 1986) and also for range values of 0.3 to 0.7 has been reported by others vegetarian mothers (Finley et al, 1985). As the diets, (Broschel et al, 1986, Koletzko et al, 1992) and it was 1.1% in consumed by lactating mothers in our study, were low in our samples. The monounsaturated fatty acid (C18:1o9) animal protein and animal fat but mainly consisted of content in our study was lower than European human milk complex carbohydrates, the milk tend to have high content (Koletzko et al, 1992). This may reflect a lower consumption of short- and medium-chain fatty acids. In the milk of of monounsaturated fatty acid from animal fats and olive or African women also high levels of C12 and C14 fatty acids Canola oils by Iranians (Finley et al, 1985, Koletzko et al, (20–30%) have been reported (Koletzko et al, 1992). 1992). Arachidonic acid (C20:4o6) is synthesized from Trans fatty acids are mainly produced during industrial elongation and desaturation of C18:2o6 in mammary glands hydrogenation of vegetable oils and their presence in the and is highly concentrated in the central nervous system human milk comes mainly from the diet (Aitchison et al, (Carlson, 2001). The level of C20:4o6 in our samples was 1977; Lawrence, 1994; Ratnayake & Chen, 1996; Mojska, 1.4% (Table 2) of the total milk fatty acids, which is higher 2003). In communities with high consumption of hydro- than that reported for human milk by others (Finley et al, genated vegetable oils, milk levels of trans fatty acids are 1985; Broschel et al, 1986; Koletzko et al, 1992). Eicosapen- expectedly high. Therefore, there is a wide variations in the taenoic acid (C20:5o3) existed in small amount (0.2%) in the trans fatty acid content of the breast milk in different milk of our subjects. populations ranging from high as in Canadians (Chen et al, In conclusion, the findings from this study indicate 1995; Ratnayake & Chen, 1996) with a mean of 7.2%, and that the milk from Iranian women contains high levels of

European Journal of Clinical Nutrition Fatty acid in milk G Bahrami and Z Rahimi 497 short- and medium-chain fatty acids and trans fats as Evans RT (1968): Manual and automated methods for measuring compared to the American and European women. The high urea based on a modification of its reaction with diacetyl J. Clin. Path. percentage of trans fatty acids in the milk of our subjects monoxime and thiosemicarbazide. 21, 527. Finley DA, Lonnerdal B, Dewey KG & Grivetti LE (1985): Breast appeared to have been of dietary origin and attributed to the milk composition: fat content and fatty acid composition consumption of hydrogenated vegetable oils with high in vegetarians and non-vegetarians. Am. J. Clin. Nutr. 41, content of trans fatty acids. Since a lot of data have been 787–800. Folch J, Less M & Sloane-Stanley GH (1957): A simple method for the published on detrimental effects of trans fatty acids on blood isolation and purification of total lipids from animals tissue. lipids and cardiovascular diseases, reduction of trans fatty J. Biol. Chem. 226, 497–509. acid availability in the diet of Iranian mothers is proposed by Francois CA, Connor SL, Wander RC & Connor WE (1998): Acute this study. effects of dietary fatty acids on the fatty acids of human milk. Am. J. Clin. Nutr. 67, 301–308. Innis SM & King J (1999): Trans fatty acids in human milk are inversely associated with concentrations of essential all-cis n-6 and Acknowledgements n-3 fatty acids and determine trans, but not n-6 and n-3, fatty acids Am. J. Clin. Nutr. The assistance of Kermanshah University of Medical Sciences in plasma lipids of breast-fed infants. 70, 383–390. for financial support of this study is appreciated. We thank Jensen RG, Lammi-Keefe CJ, Henderson RA, Bush VJ & Ferris AM the mothers for their time and effort in providing milk (1992): Effect of dietary intake of n-6 and n-3 fatty acids on the samples and diet records. fatty acid composition of human milk in North America. J. Pediatr. 120, S87–S92. Koletzko B, Mrotzek M & Bremer HJ (1988): Fatty acid composition of mature human milk in Germany. Am. J. Clin. Nutr. 47, References 954–959. Aitchison JM, Dunkley WL, Canolty NL & Smith LM (1977): Koletzko B, Thiel I & Abiodun PO (1992): The fatty acid composition Influence of diet on trans fatty acids in human milk. Am. J. Clin. of human milk in Europe and Africa. J. Pediatr. 120, S62–S70. Nutr. 30, 2006–2015. Lawrence RA (1994): Breast Feeding: A Guide for the Medical Profession Boatella J, Rafecas M, Codnoy R, Gibert A, Rivero M, Tormo R, 4th Edition. Toronto: Mosby-year Book. Infante D & Sachez-Valverde F (1993): Trans fatty acid content of Lowry OH, Rosebrough NJ, Farr AL & Randall RJ (1951): Protein human milk in Spain. J. Pediatr. Gastroenterol. Nutr. 16, 432–434. measurement with the Folin-phenol reagent. J. Biol. Chem. 193, Boersma ER, Offringa PJ, Muskiet FAJ, Chase WM & Simmons IJ 265–275. (1991): Vitamin E, lipid fractions, and fatty acid composition of Miranda R, Saravia NG, Ackerman R, Murphy N, Berman S & colostrum, transitional milk, and mature milk: an international McMurray DN (1983): Effect of maternal nutritional status on comparative study. Am. J. Clin. Nutr. 53, 1197–1204. immunological substances in human colostrum and milk. Am. J. Broschel MW, Elkin RG, Kirksey A, Story A, Galal O, Harrison GG & Clin. Nutr. 37, 632–640. Jerome NW (1986): Fatty acid composition of mature human milk Miwa H & Yamamoto M (1991): High-performance liquid chromato- of Egyptian and American women. Am. J. Clin. Nutr. 44, 330–335. graphic analysis of fatty compositions of platelet phospholipids as Carlson SE (2001): Docosahexaenoic acid and arachidonic acid in their 2-nitrophenylhydrazides. J. Chromatogr. 568, 25–34. infant development. Semin. Neonatol. 6, 437–449. Mojska H (2003): Influence of trans fatty acids on infant and fetus Chardigny JM, Wolff RL, Mager E, Sebedio JL, Martine L & Juaneda P development. Acta Microbiol. Pol. 52, S67–S74. (1995): Trans mono- and polyunsaturated fatty acids in human Ratnayake WMN & Chen ZY (1996): Trans, n-3, and n-6 fatty acids in milk. Eur. J. Clin. Nutr. 49, 523–531. Canadian human milk. Lipids. 31, S279–S282. Chen ZY, Pelletier G, Hollywood R & Ratnayake WMN (1995): Trans Sanders TAB & Reddy S (1992): The influence of a vegetarian diet on fatty acid isomers in Canadian human milk. Lipids 30, 15–21. the fatty acid composition of human milk and the essential fatty Craig-Schmidt MC (2001): Isomeric fatty acids: evaluating status and acid status of the infant. J. Pediatr. 120, S71–S77. implications for maternal and child health. Lipids 36, 997–1006. Sardesai VM (1992): Nutritional role of polyunsaturated fatty acids. Cuthbertson WFJ (1976): Essential fatty acid requirement in infancy. J. Nutr. Biochem. 3, 154–166. Am. J. Clin. Nutr. 29, 559–568. Weinsier RL, Heimburger DC & Butterworth CE (1989): Hand Book of Doumas BT, Watson WA & Biggs HG (1971): Albumin standards and Clinical Nutrition: Clinician’s Manual for the Prevention, Diagnosis, the measurement of serum albumin with bromcresol green. Clin. and Management of Nutritional Problems, 2nd Edition. Toronto: C.V. Chem. Acta. 31, 87–96. Mosby Co.

European Journal of Clinical Nutrition