Journal of Pediatric Gastroenterology & Nutrition:
Original Articles: Hepatologyand Nutrition
French Mothers’ Milk Deficient in DHA Contains Phospholipid Species of Potential Interest for Infant Development
Garcia, Cyrielle*; Millet, Véronique†; Coste, Thierry Charles*; Mimoun, Myriam*; Ridet, Audrey*; Antona, Claudine*; Simeoni, Umberto†; Armand, Martine*
*INSERM, U476 “Nutrition Humaine et Lipides,” Marseille, F-13385 France, INRA, UMR1260, “Nutriments Lipidiques et Prévention des Maladies Métaboliques,” Marseille, F-13385 France, Université de la Méditerranée Aix-Marseille 2, Faculté de Médecine, IPHM-IFR 125, Marseille, F-13385 France
†Division of Neonatology, Hôpital de la Conception, AP-HM, Marseille, France.
Address correspondence and reprint request to Martine Armand, PhD, INSERM Research Scientist, UMR INSERM Unité 476/INRA 1260/Université de la Méditerranée, Faculté de Médecine de la Timone, 27 Boulevard Jean Moulin, F-13385 Marseille cedex 05, France (e-mail: firstname.lastname@example.org).
Received 28 June, 2010
Accepted 21 February, 2011
This work was supported by a grant from the Institut Benjamin Delessert (M.A.), the Conseil Régional Provence-Alpes-Côte d’Azur and Application Santé des Lipides, A.S.L. (thesis of C.G.).
The authors report no conflicts of interest.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Website (www.jpgn.org).
Objectives: An insufficient human milk docosahexaenoic acid (DHA) level was reported worldwide, which leads to the question of the sufficiency of the DHA supply for infant development in the French Mediterranean area. Also, among milk lipids, phospholipids may be of high potential interest for infant brain development, being a specific vector of DHA and providing plasmalogens. We aimed to estimate the consumption of such milk compounds by preterm and term infants.
Materials and Methods: Milk samples from 22 lactating French women living in a port city, Marseille, were collected in a neonatology department from a single full-breast expression using an electric pump. Amounts of triglycerides, total phospholipids and plasmalogens, and fatty acid profile were determined by gas chromatography, and cholesterol by enzymatic assay.
Results: Depending on the infant dietary guidelines we referred to, 46% or 82% of milk samples were below the recommended DHA level (0.4% or 0.7%), and a majority exhibited high linoleic acid/α-linolenic acid and n-6/n-3 ratios, probably resulting from high linoleic acid together with low fish and seafood products consumption. DHA carried by phospholipids in a majority of specimens met the requirements for brain development for term but not for premature infants. Milk plasmalogen levels ranged from 3.4 to 39.2 mg/L.
Conclusions: Our results support the recommendation of DHA supplementation to French mothers living in a Mediterranean port city, and of decreased linoleic acid intake, to reach optimal milk composition for infant health. DHA-containing phospholipids including plasmalogen species may represent important bioactive human milk compounds.
Human milk is recognized as the optimal food for infant growth and development partly because of its polyunsaturated fatty acid (PUFA) composition (1). After birth, preterm and full-term newborns are able to synthesize long-chain PUFA (LC-PUFA), that is, arachidonic (AA) and docosahexaenoic (DHA) acids from linoleic (LA) and α-linolenic (ALA) acids, respectively (2), but endogenous synthesis does not fully meet the needs of organs, particularly the brain (1). Levels of LA and ALA but also AA and DHA found in human milk are considered “criterion standard” values for optimal infant health, and were used by several committees on nutrition to settle dietary recommendations in pediatrics (3–6). Thus, formula, as human milk, should contain 7% to 26% LA, 1% to 4% ALA, and up to 1% AA, expressed as percent of total fatty acids, depending on the country (6,7). The recommended level of DHA is more controversial, varying from 0.2 to 0.8 (5,6) or even 0.7% to 1% total fatty acids for an optimal effect on nervous system development (1,7,8). Based on the literature, the French Food Security Agency recommends 9% to 22% LA, 1% to 3% ALA, 0.4% to 1% AA, and 0.4% to 1% DHA (9). A clear consensus exists for LA/ALA ratio, which should be between 5 and 15 (3,4) or preferentially lower than 10 (9) to maintain the synthesis of DHA de novo from ALA (8) and possibly to prevent hyperplasia of adipose tissue (10), by avoiding excess of LA. A balanced AA/DHA ratio around 1 is also recommended (4,9); however, PUFA status in human milk is not always close to reference values, and a deficiency in n-3 fatty acids, especially DHA, was reported in several countries, except in coastal or island populations with high fish consumption (5,8,11,12). The literature is scarce concerning milk PUFA status in French women compared to other countries (not updated since 1999), and available from women living in the northern and central but not in the French Mediterranean area (13–16). Furthermore, human milk contains other bioactive lipid molecules with interesting properties that need to be explored extensively, such as phospholipids (17–19). Human milk DHA is carried by triglycerides (TG) and phospholipids, but because the latter are present at a low proportion its potential health interest was not raised. Human milk seems to contain plasmalogens (17) that represent phospholipids of possible health interest because of their bioavailability through the digestive tract (20), importance in brain development (21), and antioxidant properties (22).
The aim of our study was thus to analyze the lipid composition of human milk collected from a group of lactating women living in southern France, Marseille, a port city where fish and other seafood products are highly available, and to estimate the DHA consumption, especially through phospholipids, by preterm and term infants.
MATERIALS AND METHODS
Milk Sample Collection
Human milk samples (5 colostrum at 3–4 days postpartum, 7 transitional at 6–10 days postpartum, and 16 mature at about 30 days postpartum) were obtained from breast-feeding donor women from the Marseille area who delivered at term (n = 8, 37–40 weeks’ GA, mean ± SD 38.7 ± 1.5) or prematurely (n = 14, 28–33 weeks’ GA, mean ± SD 30.4 ± 2.2) in the local neonatology department of the Conception Hospital. Samples represented a single full-breast expression collected by electric pump. Samples were stored at −20°C and analyses were performed soon after collection.
Milk Lipid-droplet Size Determination
The milk lipid droplet size was measured by laser light scattering (Mastersizer Microplus, Malvern Instruments, Malvern, UK). The emulsion mean and median diameters (micrometer) were calculated by the particle-sizer software from the droplet size distribution.
Milk Lipid Composition Analysis
Total lipid extraction (from 1 to 4 mL milk samples depending on the analysis to be done) was performed using chloroform/methanol (2/1, v/v) containing 0.01% butylated hydroxytoluene and the organic solvent phases were partitioned with 20% (v/v) 0.15 mol/L aqueous NaCl containing 2% glacial acetic acid (v/v, pH 3.0), as already used (23). Milk TG and total phospholipids (PL) were separated by 1-dimensional thin-layer chromatography on silica gel plates (ready plastic sheet F 1500, Schleicher & Schuell, Serflam, Marseille, France) developed in chloroform/methanol/acetic acid (98/2/0.1, v/v/v) as already described (23). The resulting TG and total PL bands were visualized by spraying the plate with 6-p-toluidino-2-naphthalenesulfonic acid (Fluka, Saint Quentin Fallavier, France) and viewing the plate in UV light; then, TG and total PL individual bands were scraped into screw-cap tubes. Total lipid extracts from each milk sample and silica bands of TG and total PL were treated with boron trifluoride-methanol (BF3-methanol, Sigma, St Louis, MO) for 60 minutes at 100°C before analysis of fatty acid composition by gas chromatography (GC) (24). Internal standards were added for quantification of milk TG and total PL amounts (tripentadecanoin and phosphatidylcholine diheptadecanoyl, respectively, Sigma). Fatty acid methyl esters (FAME) were analyzed by GC (Autosystem XL, Flame Ionization Detector, Turbochrom software, Perkin Elmer, Courtaboeuf, France) using an Omegawax 250 capillary column (30 m × 0.25 mm inner diameter, 0.25-μm film thickness) (Sigma-Supelco, Saint Quentin Fallavier). Carrier gas was hydrogen and oven temperature ranged from 60°C to 215°C with a temperature rise of 45°C/min. FAME were identified by their retention times on the column using standards (PUFA 2, Sigma-Supelco) and expressed as weight percent of total fatty acids. Total cholesterol (TC) level was measured in milk lipid extracts using enzymatic procedure (Cholesterol CHOD-PAP, Roche Diagnostic, Meylan, France) after lipid pellet was suspended with isopropanol.
Before plasmalogen content determination, total PL were purified pouring a total milk lipid extract contained in 500 μL of chloroform on a silica cartridge (maxi clean SPE 900 mg, Alltech, Discoverysciences, Deerfield, IL). TG were eluted by 2 times 10 mL chloroform wash and total PL were collected using 3 times 10 mL methanol. Methanol was evaporated under nitrogen, and dried PL underwent methylation using hexane/methanol solution 4/1 and acetyl chloride, 60 minutes at 100°C (25). This method allows the analysis of FAME and of dimethyl acetal (DMA) species with no peaks overlap between these compounds (25). DMA peaks attribution was verified by comparing with retention time of a phospholipid standard rich in plasmalogens and pure 16:0 DMA (phosphatidylethanolamine from bovine brain and hexadecanal dimethyl acetal, respectively, Sigma). The same GC program as reported above was used for determination. Plasmalogens were calculated from 16:0, 18:0, and 18:1 DMA species generated during the methylation process from the vinyl ether–linked alkyl chains and using an internal standard (phosphatidylcholine diheptadecanoyl) with the following equation:
Equation (Uncited)Image Tools
where DMA and internal standard are expressed as percentage of total fatty acids (values found on chromatogram), quantity of internal standard is expressed as total microgram added in each sample assay, and milk volume is expressed as milliliter of volume used for PL extraction. The value is then multiplied by 2 to take into account the fact that 1 molecule of plasmalogen contains 1 fatty acid involved in an ester bond (that gives FAME) and 1 fatty acid involved in a vinyl ether bond (that gives DMA).
Dietary data were collected using a quantitative food-frequency questionnaire with food portion photographs adapted from a nutritional French study (26). Twenty-five lactating mothers, 16 of whom delivered prematurely, recruited in the neonatology department (Conception Hospital) and provided agreement for dietary interview, were questioned about their diet of the last 3 months of pregnancy to assess nutritional habits and, more specifically, PUFA (LA, ALA, AA, eicosapentaenoic [EPA], and DHA) consumption. The food composition table used to analyze the dietary records was completed on the basis of the one used in the SUVIMAX study (27).
Data were expressed as mean ± SD. Analyses were done in duplicate (FAME analysis, TG and PL content) or in triplicate (CT, droplet size measurement). We analyzed all of the data using a nonparametric Kruskal-Wallis test and differences between groups were identified using the Dunn posttest. The Spearman rank test was used to set up correlations. P < 0.05 was considered significant. All of the analyses were completed by using the Statview software (Abacus Concepts, Berkeley, CA). Because no significant difference was obtained between term or preterm milk samples, the data were pooled.
Milk Physicochemical Characteristics
Lipid parameters and droplet size varied greatly between milk samples and were not significantly different between lactation stages (Table 1). Overall, human milk samples contained 33.6 ± 12.4 g/L of total lipids with 33.1 ± 12.4 g/L of TG, 0.53 ± 0.32 g/L of PL, and 0.28 ± 0.10 g/L of TC. TG, PL, and TC represented, respectively, 95.3% to 99.3%, 0.4% to 3.8%, and 0.4% to 1.7% of total lipids. The mean level of plasmalogens was 18.2 ± 10.0 mg/L milk (range 3.4–39.2 mg/L corresponding to 4.6–52.2 μmol/L), and is close to data recently obtained using 31P nuclear magnetic resonance (27.3 ± 11.1 mg/L, n = 23, Garcia et al, unpublished). DMA species–related plasmalogens were not significantly different between lactation stages (Fig. 1). All milk specimens average were 2.70 ± 1.80, 1.93 ± 1.25, and 1.44 ± 0.86 μg/mL for DMA 16:0, 18:0, and 18:1, respectively, and distribution was 44.3 ± 12.1%, 28.9 ± 13.3%, and 26.7 ± 5.4%, respectively. Within each milk category, lipid concentrations varied greatly, from 2 to 5 times for TG, 5 to 7 times for total PL, 3 to 6 times for plasmalogens, and 2 to 6 times for TC.
Fatty Acid Composition of Total Lipids in Human Milk Samples
The main differences in total fatty acid profile between milk groups were a higher concentration of palmitoleic, oleic, AA, and nervonic acids in colostrum compared with mature milk, whereas the ALA level was higher in mature milk samples and showed intermediate values for transitional milk (Table 2). By comparing obtained average values to French recommended levels of PUFA for infant nutrition, we found that LA and AA were appropriate, ALA was always lower, and DHA barely reached the minimal recommended value or was below in colostrum and mature milk, respectively, and reached appropriate levels in transitional milk samples. As a consequence, LA/ALA ratio averages were higher than recommended, as were the AA/DHA ratio (in colostrum and mature milk samples). Overall n-6/n-3 fatty acids were 1.7- to 2.2-fold higher than the recommended ratio of 5; however, because a great variability was observed between samples and because infants are priority fed with their own mother's milk rather than pooled milk, we focused on individual data to conclude the fatty acid quality of milk. Regarding LA and AA recommendations, only 32% of the milk samples were below recommended levels; ALA was worse as 82% samples failed to meet the recommended values, and for DHA, 46% to 82% samples were deficient using 0.4% to 1% (9) or 0.7% to 1% (1,7,8) as the recommended level, respectively. Moreover, 11%, 32%, and 57% samples demonstrated a LA/AA ratio close to 5 to 10, to 10 to 15, or superior to 15, respectively. The AA/DHA ratio reached a value close to 1 for 36% of the samples, whereas 39% of the samples exhibited superior and 25% inferior values. The n-6/n-3 ratio was close to 5 for 32% of the samples, between 6 and 10 for 39% of the samples, and above 10 for 29% of the samples. No significant difference was obtained between our term or preterm milk samples (data not shown) in agreement with most of the published studies (28).
Fatty Acid Composition of Milk Triglycerides Versus Total Phospholipids
The fatty acid composition of TG (online-only Supplemental Digital Table 1, http://links.lww.com/MPG/A44) was relatively close to that of total lipids (Table 2). Fatty acid compositions of TG and PL (online-only Supplemental Digital Table 2, http://links.lww.com/MPG/A44) were in accordance with the scarce existing literature (19). Regarding the specific PUFA distribution between lipid classes, LA was located in both TG (11.6% ± 4.9% of total fatty acids) and PL (15.0% ± 4.4%), but with a significant higher proportion in PL (P = 0.027), whereas ALA was preferentially found in TG (P < 0.0001). The LC-PUFAs, AA and DHA, were proportionately higher in PL than in TG (P < 0.0001). Mean relative proportions of AA in TG and PL, but not of DHA, were significantly lower in mature milk compared to colostrum and transitional milk samples (Table 3). The specific contribution of TG and PL to AA and DHA supply was calculated taking into account the proportion of each PUFA expressed as a percentage of total fatty acid and the total amount of each lipid class in milk (Table 3). TG is the main quantitative lipid class in human milk, and despite a low relative proportion of LC-PUFA, the total amount of AA and DHA carried by TG was 5- to 17-fold greater than that carried by PL, representing about 80% to 90% of total human milk supply.
Dietary Habits of Pregnant Women
Daily dietary PUFA intake in pregnant women varied widely, but most reported an insufficient quality of lipids with excess SFA and n-6 PUFA compared with MUFA or n-3 PUFA (Table 4). Considering n-6 PUFA, spontaneous diet provided <8 g/day of LA for 20% of the women, between 8 and 11 g/day for 32%, between 12 and 20 g/day for 32%, and >20 g/day for 16%; thus, approximately 50% of the women consumed more LA than recommended. For n-3 PUFA, however, ALA was lacking, with only 8% of the women adhering to recommendations leading to an LA/ALA ratio >5, <10 for 56%, and >10 for 44% of the mothers interviewed. The recommended intake of DHA was also hard to satisfy as 56% of women consumed <0.2 g/day, 24% between 0.2 and 0.3 g/day, and only 20% >0.3 g/day. This is linked to an insufficient consumption of fish and seafood products and especially oily fish because 12% of women ate seafood products less than once per week, 40% between 1 and 3 times per week, and 48% more than 3 times per week. The quality of seafood products rather than frequency of consumption is responsible for meeting the need in DHA (Fig. 2). Of note, only 24% of the pregnant women were taking dietary supplements containing fish oil for at least 1 month.
The main goals of this preliminary study were to document the actual PUFA status of milk from a group of French Mediterranean mothers recruited in a neonatology department for at-risk pregnancy, to quantify milk plasmalogen content, and to assess whether breast-fed premature and full-term infants have an adequate consumption of DHA in this specific population. Even if the number of breast-milk samples analyzed was low, some interesting information was addressed.
Regarding physicochemical properties, human milk samples were consistent with the literature reporting TG, PL, and TC contents from 10 to 55, 0.4 to 0.55, and 0.15 to 0.65 g/L, respectively (18), and lipid droplet mean diameter from 3.5 to 5 μm (29,30).
Concerning PUFA, milk LA levels were in the range of European recommendations, close to values reported from French mothers living in the northern and central areas (13–16) and from other countries (12). Because ALA levels were lower than recommended as already reported in several countries (6,12–16), a high LA/ALA ratio was often obtained that raises questions regarding infant health because of its link with adipogenicity and prevalence of many inflammatory and immune disorders (10,31). Concerning LC-PUFA, a common feature observed in human breast milk worldwide was also found herein with a relatively constant AA level close to published data (5,6,11,13–16) with a maximal variation of 3 (6,15) and a wider distribution of DHA (5,6,8,13–15). This is generally explained by milk DHA content being closely linked to maternal dietary DHA intake, and milk AA concentration depends on LA conversion and maternal storages (8,11). A clear deficiency in DHA was detected in at least half of our milk samples, and average levels were not so much higher than for French women living in a noncoastal areas (13–15). The needs for DHA in premature newborns estimated from fatty acid accretion in fetal tissue during the last trimester of gestation are between 40 and 70 mg/day (1,6,9), compared to about 25 mg/day for AA (9). With a daily milk intake of 170 mL (23), average quantities of DHA provided by our milk samples were 25 ± 23 mg (range 7–96 mg) versus 29 ± 19 mg for AA (range 12–87 mg). Assuming an average daily intake of 780 mL human milk by a full-term infant (1), the averages of DHA intake among breast-fed infants were 115 ± 105 mg (range 37–442 mg) (averages 131 ± 85 mg, range 55–400 mg for AA), and the recommended intake was about 106 mg/day for the first month of life (74 mg/day for AA) (9). Thus, only 15% and 44% of our milk samples successfully met the DHA requirements for premature and full-term newborns (vs 63% and 74% for AA), respectively.
The relative proportion of DHA and AA was markedly higher in PL compared to TG, suggesting that milk specimens with the highest PL concentrations may be more efficient for brain LC-PUFA accretion because phospholipids are the best carriers (32,33). Estimating a daily need by the brain of about 3 mg for each DHA and AA from the last trimester of gestation to 2 years of age (34), and estimating a daily milk intake of 780 or 170 mL for full-term or premature newborns (1,23), 81% of milk samples meet both fatty acid requirements for term infants via PL, whereas only 18.5% and 44.5% meet the needs for DHA and AA, respectively, for premature infants.
Human milk contains plasmalogens in amounts representing 1% to 15.4% (average 4.9%) of total PL, which is consistent with the only data published in the literature on ethanolamine plasmalogens coeluted with lysophosphatidylethanolamine in human milk (17). This may be important for infants as far as plasmalogens can be absorbed as intact molecules from the intestines and delivered to tissues (20) to possibly meet the requirements of the brain during development (21). Also, plasmalogens may provide some protection against PUFA peroxidation (22). Interestingly, we found that the total concentration of milk plasmalogens correlates positively and specifically with total milk concentrations of AA (rho = 0.625, P = 0.0194) and DHA (rho = 0.7, P = 0.0088) (not with LA and ALA), and with AA (rho = 0.657, P = 0.0139) and DHA (rho = 0.711, P = 0.0078) carried by TG (no correlation with AA and DHA from phospholipids); this suggests that plasmalogens, among other antioxidant molecules present in human milk, could ensure protection of LC-PUFA, especially carried by TG, against peroxidation. Moreover, the hydroxylamine group of phosphatidylcholine and phospatidylethanolamine possesses antioxidant activity that may also protect PUFA (35). All of this could explain in part why peroxidation products, 4-hydroxyhexenal derived from DHA and 4-hydroxynonenal from AA, and malondialdehyde were lower in human milk during storage compared with infant formula (30).
Of note, calculations of minimal and maximal cholesterol intakes, using the lowest and the highest cholesterol concentrations found in our milk samples, gave 15 to 95 mg/day for premature infants and 70 to 437 mg/day for full-term infants. Cholesterol is known to play a key role in cell structure and hormone synthesis, and it may be important to consider such intakes for infant development.
A low level of DHA in a majority of samples surprised us because Marseille, a port city, enjoys easy access to fish and other seafood products. We investigated the cause using a quantitative food-frequency questionnaire of the last trimester of pregnancy because the relative proportions of milk fatty acids are largely dependent on the dietary habits and nutritional status of the mother (18). The low DHA level in milk may thus be explained by a relatively high consumption of LA versus ALA (ratio 6–22 vs 5 recommended (9)), reducing markedly the conversion of ALA to DHA through substrate competition for the Δ6 desaturase enzyme (8). Another explanation is clearly the low intake of DHA resulting from an insufficient consumption of oily fish (fresh, canned, or smoked) and an infrequent use of DHA supplements. About 60% of our studied lactating mothers consumed less than the minimal amount recommended (200–250 mg/day) to reach DHA concentrations of 0.7% in milk (1,6,9), which is a problem in Western populations (1). Several nutritional strategies may be proposed to improve the DHA level in milk. A decrease in LA consumption could help increase the biosynthesis of DHA from ALA, but then even if the conversion rate of ALA to DHA is more efficient in women of childbearing age (8), the quantity of DHA produced may not be enough to obtain milk with the recommended concentration. An increase in ALA intake, using linseed oil for example, is not of interest because this does not lead to an increase in the DHA level in milk (36). A direct increase in DHA consumption through naturally rich food (fish and seafood products) or enriched food obtained after animal nutritional manipulation seems the most powerful way (1,6,8,29). Seafood products should not be overeaten because of contaminants (methyl mercury, dioxin, and dioxin-like compounds), and this leads to the recommendation of no more than 1 portion of oily fish per week (6), preventing the desired daily intake of DHA (1,6,9). The use of DHA supplements is a good alternative (1,6–9), but human studies are necessary to select the most efficient ones; DHA can be provided as ethyl esters, TG (microalgal and fish oils), or PL (from egg), but DHA seems transferred to milk more efficiently when using the phospholipid form (37). Furthermore, genetic susceptibility for dietary DHA transfer into milk (38,39) should be taken into account.
In conclusion, this preliminary Marseille study provides the following information: living in a port city where fish and seafood are available and affordable does not necessarily lead to a higher, adequate consumption of these products because a majority of the analyzed milk samples are n-3 PUFA deficient and do not meet total DHA needs for premature and full-term infants. DHA vectors are either TG and PL, but milk rich in PL-DHA may be more efficient for brain development. Human milk appears to be a good plasmalogen dietary source for infants. We emphasize the need for the development of dietary fat recommendations or even for DHA supplementation during pregnancy and lactation, the first key stages for infant health, in the French coastal population group studied. We also question the benefit of addition of milk lipids including phospholipids enriched in DHA, plasmalogens, and cholesterol in infant formula.
We thank Dr N. Darmon and Mr M. Maillot for their kind gift of the food composition table (Excel database); C. Péchaire and L. Labbadi for their help in dietary records; S. Pitel, A. Dekkari, and C. Pastega for technical assistance; and Dr J. Philpott for reading the manuscript for proper English usage.
1. Innis S. Omega-3 fatty acids and neural development to 2 years of age: do we know enough for dietary recommendations? J Pediatr Gastroenterol Nutr
2. Salem N Jr, Wegher B, Mena P, et al. Arachidonic and docosahexaenoic acids are biosynthesized from their 18-carbon precursors in human infants. Proc Natl Acad Sci U S A
3. Aggett PJ, Haschke F, Heine W, et al. ESPGHAN Committee on Nutrition: committee report. Comments on the content and composition of lipids in infant formulas. Acta Paediatr Scand
5. Kuipers RS, Smit EN, van der Meulen J, et al. Milk in the island of chole [Tanzania] is high in lauric, myristic, arachidonic and docosahexaenoic acids, and low in linoleic acid. Reconstructed diet of infants born to our ancestors living in tropical coastal regions. Prostaglandins Leukot Essent Fatty Acids
6. Koletzko B, Lien E, Agostini C, et al. The roles of long-chain polyunsaturated fatty acids in pregnancy, lactation and infancy: review of current knowledge and consensus recommendations. J Perinat Med
7. Gibson RA, Neumann MA, Makrides M. Effect of increasing breast milk docosahexaenoic acid on plasma and erythrocyte phospholipid fatty acids and neural indices of exclusively breast fed infants. Eur J Clin Nutr
8. Arterburn LM, Hall EB, Oken H. Distribution, interconversion, and dose response of n-3 fatty acids in humans. Am J Clin Nutr
9. Martin A. Lipides. In: Apports Nutritionnels Conseillés pour la population Française
. 3rd ed. AFSSA, CNERNA-CNRS. Paris: Editions Tec & Doc; 2001: 63–82.
10. Ailhaud G, Massiera F, Weill P, et al. Temporal changes in dietary fats: role of n-6 polyunsaturated fatty acids in excessive adipose tissue development and relationship to obesity. Prog Lipid Res
11. Brenna JT, Varamini B, Jensen RG, et al. Docosahexaenoic and arachidonic acid concentrations in human breast milk worldwide. Am J Clin Nutr
12. Yuhas R, Pramuk K, Lien EL. Human milk fatty acid composition from nine countries varies most in DHA. Lipids
13. Guesnet P, Antoine JM, Rochette de lempdes JB, et al. Polyunsaturated fatty acid composition of human milk in France: changes during the course of lactation and regional differences. Eur J Clin Nutr
14. Martin JC, Bougnoux P, Fignon A, et al. Dependence of human milk essential fatty acids on adipose stores during lactation. Am J Clin Nutr
15. Chardigny JM, Wolff RL, Mager E, et al. Trans mono- and polyunsaturated fatty acids in human milk. Eur J Clin Nutr
16. Pugo-Gunsam P, Guesnet P, Subratty AH, et al. Fatty acid composition of white adipose tissue and breast milk of Mauritian and French mothers and erythrocyte phospholipids of their full-term breast-fed infants. Br J Nutr
17. van Beusekom CA, Martini IA, Rutgers HM, et al. A carbohydrate-rich diet not only leads to incorporation of medium-chain fatty acids (6:0-14:0) in milk triglycerides but also in each milk-phospholipid subclass. Am J Clin Nutr
18. Jensen RG, Bitman J, Carlson SE, et al. Human milk lipids. In: Handbook of Milk Composition
. Jensen RG, ed. New York: Academic Press; 1995: 495–575.
19. Sala-Vila A, Castellote AI, Rodriguez-Palmero M, et al. Lipid composition in human breast milk from Granada (Spain): changes during lactation. Nutrition
20. Nishimukai M, Wakisaka T, Hara H. Ingestion of plasmalogen markedly increased plasmalogen levels of blood plasma in rats. Lipids
21. Green P, Glozman S, Kamensky B, et al. Developmental changes in rat brain membrane lipids and fatty acids. The preferential prenatal accumulation of docosahexaenoic acid. J Lipid Res
22. Yavin E, Brand A, Green P. Docosahexaenoic acid abundance in brain: a biodevice to combat oxidative stress. Nutr Neurosci
23. Armand M, Hamosh M, Mehta NR, et al. Effect of human milk or formula on gastric function and fat digestion in the premature infant. Pediatr Res
24. Ohta A, Mayo MC, Kramer N, et al. Rapid analysis of fatty acids in plasma lipids. Lipids
25. Masood A, Stark KD, Salem N Jr. A simplified and efficient method for the analysis of fatty acid methyl esters suitable for large clinical studies. J Lipid Res
26. Astorg P, Arnault N, Czernichow S, et al. Dietary intakes and food sources of n-6 and n-3 PUFA in French adult men and women. Lipids
27. Hercberg S. Table de Composition des Aliments SUVIMAX
. Paris: Edition INSERM; 2004.
28. Bokor S, Koletzko B, Decsi T. Systematic review of fatty acid composition of human milk from mothers of preterm compared to full-term infants. Ann Nutr Metab
29. Favé G, Coste TC, Armand M. Physicochemical properties of lipids: new strategies to manage fatty acid bioavailability. Cell Mol Biol
30. Michalski MC, Calzada C, Makino A, et al. Oxidation products of polyunsaturated fatty acids in infant formulas compared to human milk—a preliminary study. Mol Nutr Food Res
31. Calder PC, Krauss-Etschmann S, de Jong EC, et al. Early nutrition and immunity: progress and perspectives. Br J Nutr
32. Bernoud N, Fenart L, Moliere P, et al. Preferential transfer of 2-docosahexaenoyl-1-lysophosphatidylcholine through an in vitro blood-brain barrier over unesterified docosahexaenoic acid. J Neurochem
33. Wijendran V, Huang MC, Diau GY, et al. Efficacy of dietary arachidonic acid provided as triglyceride or phospholipid as substrates for brain arachidonic acid accretion in baboon neonates. Pediatr Res
34. Martinez M. Tissue levels of polyunsaturated fatty acids during early human development. J Pediatr
35. Song JH, Imoue Y, Miyazawa T. Oxidative stability of docosahexaenoic acid-containing oils in the form of phospholipids, triacylglycerols, and ethyl esters. Biosci Biotech Biochem
36. François CA, Connor SL, Bolewicz LC, et al. Supplementing lactating women with flaxseed oil does not increase docosahexaenoic acid in their milk. Am J Clin Nutr
37. Valenzuela A, Nieto S, Sanhueza J, et al. Tissue accretion and milk content of docosahexaenoic acid in female rats after supplementation with different docosahexaenoic acid sources. Nutr Metab
38. Xie L, Innis SM. Genetic variants of the FADS1 FADS2 gene cluster are associated with altered (n-6) and (n-3) essential fatty acids in plasma and erythrocyte phospholipids in women during pregnancy and in breast milk during lactation. J Nutr
39. Molto-Puigmarti C, Plat J, Mensink RP, et al. FADS1 FADS2 gene variants modify the association between fish intake and the docosahexaenoic acid proportions in human milk. Am J Clin Nutr
docosahexaenoic acid; human milk; nutrition; pediatrics; phospholipids
Supplemental Digital Content
Copyright 2011 by ESPGHAN and NASPGHAN
Highlight selected keywords in the article text.