Secondary Logo

Journal Logo

Original Articles: Nutrition

Dietary Effects on Plasma Glycerophospholipids

Uhl, Olaf*; Hellmuth, Christian*; Demmelmair, Hans*; Zhou, Shao J.; Makrides, Maria; Prosser, Colin; Lowry, Dianne; Gibson, Robert A.§; Koletzko, Berthold*

Author Information
Journal of Pediatric Gastroenterology and Nutrition: September 2015 - Volume 61 - Issue 3 - p 367-372
doi: 10.1097/MPG.0000000000000783
  • Free


What Is Known

  • Human breast milk contains a complex mixture of lipids.
  • Standard infant formula fat is based on simple vegetable oils.
  • New developed infant formula is based on complex animal (goats) lipids.

What Is New

  • Despite the different dietary lipid sources of infant formulas, differences between the plasma glycerophospholipid profiles of infants are low.
  • Plasma glycerophospholipid species containing myristic acid (14:0) or palmitoleic acid (16:1) are increased in the infants-fed formula based on goats’ milk compared with the standard infant formula.
  • Breast-fed infants show markedly higher contents of plasma glycerophospholipid species containing LC-PUFAs compared with formula-fed infants.

Nutrition in the first months of life is crucial for the healthy development of infants and may affect long-term outcomes (1). Breast-feeding is the preferred choice of infant feeding (2). For infants that are not fully breast-fed, infant formulae are needed which should be of the best possible quality (3).

In human milk, triglycerides contribute approximately 99% of total fat content (4). Human breast milk contains a complex mixture of lipids with a broad range of fatty acids. Standard infant formula fat is based on vegetable oils (VIF) to ensure provision of essential fatty acids, but to approximate the composition of human milk mammalian fat sources such as cow milk or goat milk (GIF) can be introduced. There is an increasing consumer demand for goat milk formula and the European Food Safety Authority has recently approved the use of whole goat's milk in the manufacture of infant formula (5). Dietary fatty acid composition strongly affects the fatty acid status of the infant including the fatty acid composition of different tissue lipids, which are important for early postnatal development. Long-chain polyunsaturated fatty acids (LC-PUFAs) are found in high amounts in neuronal tissue and in the retina, and their inclusion in formula significantly enhances the visual acuity maturation and cognitive functions in infants (6). LC-PUFAs have been considered as conditionally essential substances, because endogenous synthesis is not sufficient to compensate for the absence of an exogenous supply (7). In contrast to vegetable oils, human milk lipids are characterized by a high content of palmitic acid preferentially esterified in the sn-2 position of triacylglycerols (8). During digestion, sn-1 and sn-3 bound fatty acids are cleaved from the glycerol backbone, leaving palmitic-acid-rich sn-2 monoacylglycerols which because of the polarity of glycerol are better absorbed than free palmitic acid. After the absorption, the sn-2 monoacylglycerols are re-esterified to triacylglycerols or glycerophospholipids (GPLs) (9). Thus, both fatty acid composition as well as the molecular structure of dietary fatty acids matter for infant supply.

In the present study, we aimed to investigate the effects of different dietary milk lipid supplies to infants on the composition of infant plasma GPL, which serve as a source of membrane lipids and hence may affect cell and tissue functions. The plasma GPL pattern of infants fed GIF was expected to be more similar to breast-fed infants than to infants fed VIF.


Patients and Study Procedure

We analyzed the plasma samples of infants who participated in a multicentre, double-blind, controlled feeding trial in Australia (Australian New Zealand Clinical Trials Registry ACTRN12608000047392). Details on the study design and participating patients have been previously published (10). Written informed consent was obtained from all participating mothers. Healthy term infants (37–42 weeks of gestation) with birth weights between 2.5 and 4.75 kg were included, and randomised to receive either VIF or GIF from the age of maximal 2 weeks. The GIF was manufactured using whole goats’ milk without added whey proteins and a blend of approximately 60% goats’ milk fat and 40% vegetable oils. The VIF contained cows’ skim milk and whey proteins, and vegetable oils as the sole source of fat. Macronutrients and further details about the infant formula composition were previously published (10). The total fatty acid composition of both infant formulas is shown in Table 1. Plasma samples were taken after 4 months of exclusively feeding of the corresponding study formula. A reference group of exclusively breast-fed infants was included.

Total fatty acid composition (mol%) of the formula based on whole goat's milk (GIF) and of the control infant formula based on vegetable oils (VIF)

Sample Preparation and Analysis

GPL species were analyzed as described previously (11). Briefly, proteins of 10 μL plasma were precipitated with 90 μL methanol containing internal standards. After filtration, 400 μL methanol were added. Diluted samples were injected to an Agilent 1200 high-performance liquid chromatography and GPL species were separated on a Kinetex C18 high-performance liquid chromatography column, 100 mm × 2.1 mm, 2.6-μm particle size (Phenomenex, Aschaffenburg, Germany). The mobile phases (A, 40% methanol/60% water; B, 90% isopropanol/10% methanol) contained 10 mmol/L ammonium acetate and 1 mmol/L acetic acid, respectively. The gradient elution started from 40% B, held for 4 minutes, rose to 65% B within 2 minutes, held for further 8 minutes, before the gradient rose up to 95% B within 4 minutes and held for further 2 minutes. Finally, the initial gradient of 40% B was equilibrated for 4 minutes. The triple quadrupole mass spectrometer (AB Sciex API4000; Applied Biosystems, Darmstadt, Germany) was operated in negative-scheduled multiple reaction monitoring mode using two fragments of each analyte. A fatty acid–containing fragment of each analyte was utilized for quantification and a second fragment of the molecule, depending on the kind of analyte, was used as qualifier ion. Data were post processed with Analyst 1.5.1 software (Applied Biosystems, Darmstadt, Germany) and Excel 2010 (Microsoft, Redmond, WA). The GPLs are labeled as lysophosphatidylcholine (LPC), lysophosphatidylethanolamine (LPE), phosphatidylethanolamine (PE), or phosphatidylcholine (PC) with XX:Y/XX:Y as description of the sn-1/sn-2 fatty acids, whereas XX means number of carbon atoms and Y the number of double bonds, for example, PC(16:0/22:6) for 1-palmitoyl-2-docosahexaenoyl-sn-glycero-3-phosphocholine.


Plasma samples were measured for the total number of 116 GPL species with a detection limit of 1 μmol/L. A censored regression model was performed to estimate the mean log GPL levels below detection threshold. A maximum of 37% missing values per analyte was accepted, resulting in a number of 52 GPL species with given mean concentrations in micromol per liter. Differences between groups were calculated from mean log GPL levels of the censored regression model. As a result of multiple testing, the significant threshold for P values was Bonferroni corrected with P values under 0.00096 considered as significant.


Some 167 plasma samples (62 GIF, 53 VIF, and 52 breast-fed) of the study were analyzed for 116 GPL species. After statistical adjustment for levels below the detection threshold, mean values of 52 GPL species for each group were calculated (Table 2).

Estimates of geometric mean glycerophospholipid levels from censored regression models of infants fed formula based on whole goat's milk (GIF), an infant formula based on vegetable oils (VIF) or breast-fed infants

GIF versus VIF

Ten GPL species were found with significantly different levels between plasma samples of infants fed GIF compared with infants fed VIF. All of them contained a choline head group. A total of 9 of 10 significant GPL species were between 27% and 72% higher in GIF. Only PC(16:0/18:3) was approximately 20% lower in infants fed GIF. All GPL species containing myristic acid (14:0) and all GPL species with palmitoleic acid (16:1) were increased in the infants fed GIF.

GIF versus Breast-fed

Sixteen GPL species were found with lower levels in goats’ milk fed infants, and seventeen GPL species were found with higher levels, compared to infants fed breast milk. Thirteen out of the 16 species with higher values in breast-fed infants contained long-chain polyunsaturated fatty acids (>20 carbon atoms and >1 double-bound PUFAs). In addition, PC(16:0/16:0) and PC(18:0/16:0), containing palmitic acid at sn-2 position, were higher in infants fed breast milk. A total of 15 of 17 GPL species with higher levels in the GIF group contained oleic (18:1), linoleic (18:2), or alpha-linolenic acid (18:3).

VIF Versus Breast-fed

A total of 18 of 23 GPL species with higher values in breast-fed infants contained LC-PUFAs. The levels of species with LC-PUFAs in formula-fed infants were up to 60% lower compared with breast-fed infants. Again, PC(16:0/16:0) and PC(18:0/16:0), containing palmitic acid at sn-2 position, were higher in the breast milk group. All of the 13 GPL species with higher levels in the VIF-fed infants contained oleic (18:1), linoleic (18:2), or alpha-linolenic acid (18:3).


The plasma GPL composition is influenced by both dietary intake and endogenous synthesis, similar to other lipid classes. In contrast to other lipids, for example, triglycerides, GPLs are less influenced by the last meal and postprandial absorption state. Dietary fats are digested and absorbed by the small intestine. Neutral triacylglycerides are re-assembled within the endoplasmic reticulum of the small intestine cells. These lipid droplets are coated by phospholipids and apolipoprotein B48 and further lipoproteins to form water-soluble chylomicrons, catalyzed by the microsomal triglyceride transfer protein (12,13). After absorption, LC-PUFAs seem to be incorporated preferentially into phospholipids, rather than in triacylglycerols, as suggested by the experiments in rats feeding radiolabeled arachidonic acid (AA) and linoleic acid (14). Chylomicrons are released into the lymphatic system and transferred as chyle via the chylothorax into the subclavian vein, and thus the blood circulation. Triacylglycerols are transferred by chylomicrons or VLDL to tissue, and GPLs are metabolized by lecithin-cholesterol acyltransferase to cholesterol esters and lyso-GPL. Cholesterol esters are further transported by HDL and lyso-GPL are bound to albumin. Lyso-GPLs are in discussion as the preferred form of transport for AA and docosahexaenoic acid (DHA) to all kinds of tissue, including the brain (15). Recently, in Mfsd2a, a transporter through the blood–brain barrier could be identified, specific for LPC with long-chain fatty acids (<14 carbon atoms), especially for DHA (16). Fatty acids at position sn-1 are cleaved and LPC with LC-PUFAs are left over for transport. Plasma GPLs are the most relevant source of LC-PUFAs for neuronal tissue, because LC-PUFAs cannot be synthesized de novo after the blood–brain barrier. Thus, the plasma GPL composition is important for the optimal development of cognitive functions; however, analyzed plasma lipids represent a mix of all kind of lipoproteins, including last-meal-derived chylomicrons, as infants could not be studied in a fasting state. Thus, plasma GPL reflects endogenous and to a smaller extent dietary fatty acids. In view of all these aspects, GPLs can be considered a suitable marker for the pool of fatty acids available for metabolism or tissue incorporation. We hypothesized that differences in the GPL status of infants fed GIF, VIF, or breast milk would reflect the different dietary fat sources.

Our results show that the GPL species most markedly influenced by the different formulas are those containing myristic acid [LPC(14:0), PC(14:0/18:1), PC(16:0/14:0)], which were significantly higher in the GIF group than in the VIF group. Similarly, in adults in whom the major source of myristic acid is exogenous supply from dairy fat rather than endogenous synthesis (17). Myristic acid modulates the activity of several mammalian enzymes. Via an N-myristoyltransferase, myristic acid is bound irreversible to the N-terminal glycine of nascent polypeptides. This modification is a regulator of the Δ4-desaturation of dihydroceramide to ceramide as well as an activator of the Δ6-desaturation of polyunsaturated fatty acids (18). In a 1-year study comparing different myristic acid intakes (1.2% vs 1.8% of total energy intake), higher myristic acid intakes enhanced the level of n-3 fatty acids, which supports the hypothesis of a stimulating effect of myristic acid on Δ6-desaturation of polyunsaturated fatty acids (19). In contrast, we did not detect any enhancement of infant docosahexaenoic acid levels in GIF compared with VIF.

Significantly increased values of GPL species containing palmitoleic acid [60% LPC(16:1), 46% PC(16:0/16:1), and 66% PC(16:1/18:1)] were also found in GIF compared with VIF. Milk fat is a rich source of palmitoleic acid (20); however, the major portion of blood palmitoleic acid is derived from palmitic acid via endogenous desaturation mediated by the enzyme stearoyl-CoA desaturase rather than dietary intake. Adipose tissue seems to act as a source for the nonesterified fatty acid to provide signals to other organs (21). Because palmitoleic acid was not detectable in the formula (Table 1), the higher levels of GPL containing palmitoleic acid in GIF may be a metabolic response to some compositional aspects of the formula which deserves further exploration.

On the basis of the fatty acid composition of both infant formula, we expected larger differences for GPL species containing fatty acids with a number of carbon atoms less than 14, but these fatty acids are hardly found in GPL. The lesser content of palmitic acid in the GIF was not reflected in the plasma GPL profile.

Human breast milk contains high concentrations of palmitic acid bound to the sn-2 position of triacylglycerols. Palmitic acid bound to the sn-1 and sn-3 position of triacylglycerols, as found in vegetable oils and in most infant formula, was related to a reduced fat absorption as a result of the formation of insoluble calcium dipalmitate (22). During the process of digestion, triacylglycerols are hydrolyzed to 2-monoacylglycerols and nonesterified fatty acids, which are absorbed and re-esterified in the intestinal epithelial cells (23). Consequentially, the sn-2 position of the palmitic acid is conserved. Re-esterification may occur in large amounts again to triacylglycerols, but synthesis of GPLs from monoacylglycerols also occurs to a lesser extent. Our results showed significantly higher content of GPL species containing sn-2 palmitic acid [PC(16:0/16:0) and PC(18:0/16:0)] in breast-fed infants compared with both formula groups. The lower content of PC(16:0/18:2) in breast-fed infants indicates a different reason for higher PC(16:0/16:0) and PC(18:0/16:0), apart from higher palmitic acid in breast milk. The majority of PC(16:0/16:0) and PC(18:0/16:0) is synthesized de novo via the Kennedy pathway, but higher values in breast-fed infants may be an indication for the endogenous synthesis of GPLs from sn-2 palmitic acid monoacylglycerol, conserved from dietary triacylglycerols. Thus, the diet is not only reflected in the fatty acid composition of GPLs but also in the molecular constitution. Despite higher sn-2 palmitic acid concentrations in GIF compared with VIF (24), only PC(18:0/16:0) was found at a slightly higher level in GIF. In contrast, in breast-fed infants, PC(18:0/16:0) was 37% higher than in GIF.

The GPL species pattern shows a strong separation between both formula groups and the breast-fed group, because 32 (GIF) and 36 (VIF) of 52 measured species showed significant differences. Figure 1 shows smaller contents of GPL species containing LC-PUFAs (mainly AA and DHA) in the profiles of all formula-fed infants compared with the breast-fed ones. Both formulas were manufactured according to the international recommendations containing a ratio of linoleic acid to alpha-linoleic acid between 5 and 15:1 (3), but there was no added DHA or AA in either formula. Linoleic acid and alpha-linoleic acid are essential fatty acids and precursors for the endogenous synthesis of LC-PUFAs. The absence of exogenous LC-PUFA supply in both formula groups is reflected in the plasma levels of LC-PUFA containing GPLs (Fig. 1), with slightly higher values for GIF, and it clearly shows that endogenous synthesis of LC-PUFAs was not sufficient to reach breast-fed levels (7,25). The slightly higher values in the GIF group may be explained by the use of approximately 60% goats’ milk fat which contains small amounts of LC-PUFAs; however, these GPLs reached only approximately 40% [LPE(22:6)] of the levels found in breast-fed infants. Looking at the species of AA and DHA in formula-fed infants compared with breast-fed ones, no relevant differences between PC and PE species could be seen. All PC and PE species containing AA or DHA were approximately 50% less in formula-fed infants than in breast-fed infants. In the hepatic GPL synthesis, LC-PUFAs are preferentially incorporated into PE. The phosphatidylethanolamine-N-methyltransferase (PEMT) pathway is in part responsible for the transformation of LC-PUFA rich PE to PC species, by transferring 3 methyl groups to the ethanolamine head group (26). Our results did not show a relevant difference of single GPL species between PC and PE, and thus we conclude that PEMT activity or specificity was not affected by the different diets. This is in line with the previous results showing a homogenous incorporation of DHA into LPC, LPE, PC, and PE species (27). LC-PUFAs seem to be equilibrated between PC and PE in mature lipoproteins; however, nascent lipoproteins are enriched in PE compared with plasma lipoproteins (28). During the process of maturation within liver cells, PE is metabolized by the hepatic lipase, which is specific for the sn-1 position, and thus may generate LC-PUFA-rich LPE that also may supply AA and DHA to the brain. LPE22:6 was significantly higher in breast-fed infants, compared to both formula groups and thus a higher supply of DHA to the brain seems possible.

Contrasts between formula based on whole goats’ milk (black bars) and a control infant formula based on vegetable oils (white bars) to breast-fed infants. Values are given as relative difference of geometric means.


Despite the different dietary lipid sources of the infant formula based on whole goats’ milk and the infant formula based on vegetable oils, there were only small differences in the plasma glycerophospholipid profile of infants. Formula based on complex animal lipids had comparable effects to formula with vegetable oils, but showed small improvements in the plasma GPL profile, for example, more bioavailable palmitoleic acid. GPLs in both formula groups were clearly different from the breast-fed infants, with smaller contents of species containing LC-PUFAs.


1. Koletzko B, Boey CC, Campoy C, et al Current information and Asian perspectives on long-chain polyunsaturated fatty acids in pregnancy, lactation, and infancy: systematic review and practice recommendations from an Early Nutrition Academy Workshop. Ann Nutr Metab 2014; 65:49–80.
2. Agostoni C, Braegger C, Decsi T, et al ESPGHAN Committee on Nutrition. Breast-feeding: a commentary by the ESPGHAN committee on nutrition. J Pediatr Gastroenterol Nutr 2009; 49:49112–49125.
3. Koletzko B, Baker S, Cleghorn G, et al Global standard for the composition of infant formula: recommendations of an ESPGHAN coordinated international expert group. J Pediatr Gastroenterol Nutr 2005; 41:584–599.
4. Koletzko B, Agostoni C, Bergmann R, et al Physiological aspects of human milk lipids and implications for infant feeding: a workshop report. Acta Paediatr 2011; 100:1405–1415.
5. EFSA NDA Panel. (EFSA panel on dietetic products NaA scientific opinion on the essential composition of infant and follow-on formulae. EFSA J 2014; 12:3760–3866.
6. Uauy R, Hoffman DR, Peirano P, et al Essential fatty acids in visual and brain development. Lipids 2001; 36:885–895.
7. Larque E, Demmelmair H, Koletzko B. Perinatal supply and metabolism of long-chain polyunsaturated fatty acids: importance for the early development of the nervous system. Ann N Y Acad Sci 2002; 967:299–310.
8. Martin JC, Bougnoux P, Antoine JM, et al Triacylglycerol structure of human colostrum and mature milk. Lipids 1993; 28:637–643.
9. Mu H, Hoy CE. The digestion of dietary triacylglycerols. Prog Lipid Res 2004; 43:105–133.
10. Zhou SJ, Sullivan T, Gibson RA, et al Nutritional adequacy of goat milk infant formulas for term infants: a double-blind randomised controlled trial. Br J Nutr 2014; 111:1641–1651.
11. Uhl O, Glaser C, Demmelmair H, et al Reversed phase LC/MS/MS method for targeted quantification of glycerophospholipid molecular species in plasma. J Chromatogr B Analyt Technol Biomed Life Sci 2011; 879:3556–3564.
12. Olofsson SO, Bostrom P, Andersson L, et al Lipid droplets as dynamic organelles connecting storage and efflux of lipids. Biochim Biophys Acta 2009; 1791:448–458.
13. Xiao C, Lewis GF. Regulation of chylomicron production in humans. Biochim Biophys Acta 2012; 1821:736–746.
14. Nilsson A, Landin B, Jensen E, et al Absorption and lymphatic transport of exogenous and endogenous arachidonic and linoleic acid in the rat. Am J Physiol 1987; 252 (6 Pt 1):G817–G824.
15. 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 1999; 72:338–345.
16. Nguyen LN, Ma D, Shui G, et al Mfsd2a is a transporter for the essential omega-3 fatty acid docosahexaenoic acid. Nature 2014; 509:503–506.
17. Rioux V, Legrand P. Saturated fatty acids: simple molecular structures with complex cellular functions. Curr Opin Clin Nutr Metab Care 2007; 10:752–758.
18. Rioux V, Pedrono F, Legrand P. Regulation of mammalian desaturases by myristic acid: N-terminal myristoylation and other modulations. Biochim Biophys Acta 2011; 1811:1–8.
19. Dabadie H, Motta C, Peuchant E, et al Variations in daily intakes of myristic and alpha-linolenic acids in sn-2 position modify lipid profile and red blood cell membrane fluidity. Br J Nutr 2006; 96:283–289.
20. Ceballos LS, Morales ER, Adarve GDT, et al Composition of goat and cow milk produced under similar conditions and analyzed by identical methodology. J Food Compos Anal 2009; 22:322–329.
21. Hodson L, Karpe F. Is there something special about palmitoleate? Curr Opin Clin Nutr Metab Care 2013; 16:225–231.
22. Bar-Yoseph F, Lifshitz Y, Cohen T. Review of sn-2 palmitate oil implications for infant health. Prostaglandins Leukot Essent Fatty Acids 2013; 89:139–143.
23. Nelson CM, Innis SM. Plasma lipoprotein fatty acids are altered by the positional distribution of fatty acids in infant formula triacylglycerols and human milk. Am J Clin Nutr 1999; 70:62–69.
24. Prosser CG, Svetashev VI, Vyssotski MV, et al Composition and distribution of fatty acids in triglycerides from goat infant formulas with milk fat. J Dairy Sci 2010; 93:2857–2862.
25. Heird WC. The role of polyunsaturated fatty acids in term and preterm infants and breastfeeding mothers. Pediatr Clin North Am 2001; 48:173–188.
26. Pynn CJ, Henderson NG, Clark H, et al Specificity and rate of human and mouse liver and plasma phosphatidylcholine synthesis analyzed in vivo. J Lipid Res 2011; 52:399–407.
27. Uhl O, Demmelmair H, Klingler M, et al Changes of molecular glycerophospholipid species in plasma and red blood cells during docosahexaenoic acid supplementation. Lipids 2013; 48:1103–1113.
28. Hamilton RL, Fielding PE. Nascent very low density lipoproteins from rat hepatocytic Golgi fractions are enriched in phosphatidylethanolamine. Biochem Biophys Res Commun 1989; 160:162–173.

glycerophospholipids; infant formula; long-chain polyunsaturated fatty acids; palmitoleic acid

© 2015 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,