Temporal Changes of Human Breast Milk Lipids of Chinese Mothers

Fatty acids (FA), phospholipids (PL), and gangliosides (GD) play a central role in infant growth, immune and inflammatory responses. The aim of this study was to determine FA, PL, and GD compositional changes in human milk (HM) during lactation in a large group of Chinese lactating mothers (540 volunteers) residing in Beijing, Guangzhou, and Suzhou. HM samples were collected after full expression from one breast and while the baby was fed on the other breast. FA were assessed by direct methylation followed by gas chromatography (GC) analysis. PL and GD were extracted using chloroform and methanol. A methodology employing liquid chromatography coupled with an evaporative light scattering detector (ELSD) and with time of flight (TOF) mass spectrometry was used to quantify PL and GD classes in HM, respectively. Saturated FA (SFA), mono-unsaturated FA (MUFA), and PL content decreased during lactation, while polyunsaturated FA (PUFA) and GD content increased. Among different cities, over the lactation time, HM from Beijing showed the highest SFA content, HM from Guangzhou the highest MUFA content and HM from Suzhou the highest n-3PUFA content. The highest total PL and GD contents were observed in HM from Suzhou. In order to investigate the influence of the diet on maternal milk composition, a careful analyses of dietary habits of these population needs to be performed in the future.


Introduction
Human milk (HM) is considered the optimal form of nourishment for infants during the first six months of life [1] and among its macronutrients, the lipid fraction is crucial, representing approximately 50% of the energy supplied to the newborn infant [2]. Lipids (2%-5%) occur in milk in the form of fat globules mainly composed of triacylglycerols (TAG) (~98% of total lipids) surrounded by a structural membrane composed of phospholipids (PL) (0.8%), cholesterol (0.5%), enzymes, proteins, glycosphingolipids (e.g., gangliosides (GD)), and glycoproteins [3,4].

Subjects
This study was part of MING, a cross-sectional study designed to investigate the dietary and nutritional status of pregnant women, lactating mothers, and young children aged from birth up to three years living in urban areas of China [26]. In addition, the HM composition of Chinese lactating mothers was characterized. The study was conducted between October 2011 and February 2012. A multi-stage milk sampling from lactating mothers in three cities (Beijing, Suzhou, and Guangzhou) was performed for breast milk characterization. In each city, two hospitals with maternal and child care units were selected and, at each site, mothers at lactation period 0-240 days were randomly selected based on eligibility criteria. Subjects included in the period 0-5 days were recruited at the hospital, whereas the other subjects were requested by phone to join the study; if participation was dismissed a replacement was made. The response rate was 52%. Recruitment and milk, as well as baseline data collection, were done in separate days. A stratified milk sampling of 540 lactating mothers in six lactation periods of 0-4, 5-11, and 12-30 days, and 1-2, 2-4, and 4-8 months was obtained in the MING study.

Inclusion and Exclusion Criteria
Eligibility criteria included women between 18-45 years of age giving birth to a single, healthy, full-term infant and exclusive breastfeeding at least until four months of age. Exclusion criteria included gestational diabetes, hypertension, cardiac diseases, acute communicable diseases, and postpartum depression. Lactating women who had nipple or lacteal gland diseases, who had been receiving hormonal therapy during the three months preceding recruitment, or who had insufficient skills to understand study questionnaires were also excluded.

Ethical and Legal Considerations
The study was conducted according to the guidelines in the Declaration of Helsinki. All of the procedures involving human subjects were approved by the Medical Ethics Research Board of Peking University (No. IRB00001052-11042). Written informed consent was obtained from all subjects participating in the study. The study was also registered in ClinicalTrials.gov with the number identifier NCT01971671.

Data Collection
All subjects responded to a general questionnaire including socio-economic and lifestyle aspects of the mother. Self-reported weight at delivery, number of gestational weeks at delivery, and delivery method were also recorded. Additionally, a physical examination evaluated basic anthropometric parameters (height, weight, mid-arm circumference) blood pressure, and hemoglobin. Data collection was done through face-to-face interviews the day of HM sample collection. In addition, the date of birth and gender information of the baby was collected after the data collection, since the data was not included in the initial questionnaires. Subjects were contacted by phone and were asked to clarify these two aspects retrospectively.

HM Sampling
Breast milk sampling was standardized for all subjects and an electric pump (Horigen HNR/X-2108ZB, Xinhe Electrical Apparatuses Co., Ltd., Beijing, China) was used to sample the milk. Samples were collected at the second feeding in the morning (9:00-11:00 a.m.) to avoid circadian influence on the outcomes. A single full breast was emptied and aliquots of 10 mL for colostrum and 40 mL for the remaining time points was secured for characterization purposes. The rest of the milk was returned to the mother for feeding to the infant. Each sample was distributed in freezing tubes, labelled with subject number, and stored at −80 • C until analysis. Figure 1 shows the study flowchart for the subjects' recruitment.

FA Quantification
FA profile was determined by preparing the methyl esters of FA (FAMEs). A direct transesterification of HM was performed with methanolic chloridric acid solution, as described by Cruz-Hernandez et al. [27]. Briefly, into a 10 mL screw cap glass test tube, milk (250 µL) was added and mixed with 300 µL of internal standard FAME 11:0 solution (3 mg/mL) and 300 µL of internal standard TAG 13:0 solution (3 mg/mL). After addition of 2 mL of methanol, 2 mL of methanolic chloridric acid (3 N), and 1 mL of hexane, the tubes were heated at 100 • C for 90 min. To stop the reaction 2 mL of water were added and after centrifugation (1200× g for 5 min) the upper phase (hexane) was transferred into gas chromatography vials. The analysis of FAMEs was performed by GC using a CP-Sil 88 capillary column (100 m, 0. 25

Gangliosides Quantification
GD were quantified as previously described by Giuffrida et al. [2]. Briefly, HM (0.2 mL) was dissolved in water (1 mL) and mixed with 4 mL methanol/chloroform (2/1). After centrifugation (3000× g, for 10 min), the upper liquid phase was quantitatively transferred into a 15 mL centrifuge tube. The residue was mixed with water (1 mL), 2 mL of methanol/chloroform (2/1), shaken, put into an ultrasonic bath at 25 • C for 10 min, centrifuged (3000× g, for 10 min), and upper liquid phases polled together; the volume was adjusted to 12 mL with methanol 60% and pH to 9.2 by adding Na2HPO4 30 mmol/L (0.2 mL). The extract solution was loaded on an Oasis HLB VAC RC SPE cartridges (30 mg, 15 mL, Waters) previously conditioned with methanol (2 mL) and methanol 60% (2 mL). The sample was passed through the cartridge at maximum flow rate 2-3 mL /min. The sorbent was washed with 2 mL of methanol 60% and dried by vacuum suction for a few seconds; the analyte was eluted with methanol (2 mL). Solvent was evaporated to dryness under a nitrogen flow at 30 • C and the residual lipids were re-dissolved in 0.2 mL of methanol 70% and analysed by liquid chromatography (LC) coupled with quadrupole time of flight (QTOF), using an Aquity BEH C18 column (1.7 µm; 150 × 2.1 mm i.d.; Waters). All chromatography was performed at 50 • C. Solvent A was composed of water/methanol/ammonium acetate (1 mmol/L) (90/10/0.1 v/v/v) and solvent B of methanol/ammonium acetate (1 mmol/L) (100/0.1 v/v). Gradient conditions were as follows: time = 0 min 10% solvent A; time = 0.2 min 10% solvent A; time = 8.2 min 5% solvent A; time = 12.2 min 5% solvent A; time = 12.4 min 0% solvent A; time = 18.4 min 0% solvent A; time = 18.6 min 10% solvent A; time = 21 min 10% solvent A. Flow rate was 0.2 mL/min. Injection volume was 0.01 mL for GD3 and 0.005 mL for GM3. The mass spectrometer was equipped with an electrospray ionization (ESI) ion source. The ESI mass spectra were recorded in the negative ion mode under the following conditions: ion spray voltage (IS) −4000 V, temperature of the source 400 • C, declustering potential (DP) −40 V, ion source gases one and two at 40 and 35 psi, respectively, curtain gas at 15 psi, collision energy −40 V. GD3 and GM3, were monitored by transitions of the precursor ions to the m/z 290. Quantification was performed by the standard addition method.

Demographics and Anthropometrics of Study Subjects
In the current study we analyzed HM from 539 mothers (Figure 1), collected in a cross-sectional design over eight months postpartum. Milk obtained for analyses was a single, whole breast milk sample to have a comprehensive view on nutrient content. The details of the demographics and anthropometrics of the study subjects are outlined in Table 1. Groups of mothers, which delivered either a male or a female infant, were comparable for their age and anthropometric and demographic characteristics. Gestational age at birth (average 39 weeks) were also comparable between groups. The details of demographics and anthropometrics of the study subjects for the time period 0-4 days are not available.

FA
FA were determined by gas chromatography coupled with flame ionization detector (GC-FID), as previously described by Cruz-Hernandez et al. [27] and the results are listed in Table 2.
In our study total SFA content increased significantly from colostrum (35.7% of total FA) to transitional milk (38.9% of total FA) and decreased in mature milk (36.2% of total FA), with palmitic acid (16:0) being the most abundant FA and decreasing significantly (p < 0.05) from 23.2% in colostrum to 19.8% of total FA in mature milk ( Table 2). Stearic acid (18:0) content was constant along the lactation period, i.e., colostrum, transitional, and mature milk, at about 5% of total FA, and medium-chain (MC) FA (10:0-14:0) content was low in colostrum (6.8% of total FA) compared to transitional (13.1% of total FA), and mature milk (11.0% of total FA) ( Table 2). Arachidic (20:0) and lignoceric acids (24:0) were constant along the lactation time at about 0.2 and 0.1% of total FA, respectively. No significant differences (p > 0.05) on total SFA content were observed among cities in colostrum, and transitional milk ( Table 2). SFA content was significant lower (p < 0.05) in mature milk from Suzhou (34.5% of total FA). Palmitic (22.5%, 19.4%, and 18.5% of total FA in colostrum, transitional, and mature milk, respectively) and stearic (4.9%, 4.5%, and 4.8% of total FA in colostrum, transitional and mature milk, respectively) FA also showed the lowest content in mature milk from Suzhou.
In the total population the MUFA content of HM decreased from 40.7% in colostrum to 36.9% of total FA in mature milk, with oleic acid (18:1n-9) being the most abundant FA and decreasing along the lactation time from 34.2% in colostrum to 31.9% of total FA in transitional and mature milk. Other MUFA (i.e., 17:1n-7, 20:1n-9, 22:1n-9. and 24:1n-9) also decreased over the lactation period ( Table 2). The highest level of total MUFA content was found in colostrum (43.1% of total FA), transitional (39.3% of total FA), and mature milk (38.3% of total FA) from Guangzhou ( Table 2). The lowest level of total MUFA content was found in colostrum (38.4% of total FA), transitional (34.7% of total FA), and mature milk (34.3% of total FA) from Beijing ( Table 2). HM samples obtained from mothers in Guangzhou contained the highest level of Oleic acid whereas milk obtained from mothers in Beijing contained the lowest level, respectively: colostrum (37.1% vs. 32.6% of total FA), transitional (34.0% vs. 30.3% of total FA), and mature milk (33.4% vs. 30.1% of total FA).
In the total population, total PUFA n-6 increased from 21.7% in colostrum to 24.1% of total FA in mature milk with linoleic acid (18:2n-6) being the most abundant FA and increasing along the lactation time from 18.9% in colostrum to 22.8% of total FA in mature milk. ARA (20:4n-6) content decreased from 0.9% to 0.5% of total FA from colostrum to mature milk. Beijing and Suzhou showed higher total PUFAn-6 content in colostrum (23.3% and 22.8% of total FA, respectively), transitional (22.5% and 22.9% of total FA, respectively), and mature milk (26.6% and 25.3% of total FA, respectively) than Guangzhou (Table 2).
Total PUFA n-3 in HM from total population slightly increased from 1.4% in colostrum to 1.9% of total FA in mature milk with linolenic acid (18:3n-3) being the most abundant and increasing along the lactation time from 0.9% in colostrum to 1.5% of total FA in mature milk. DHA (22:6n-3) slightly decreased over lactation period from 0.5% in colostrum to 0.3% of total FA in mature milk, and EPA (20:5n-3) was present in a small amount (<0.1% of total FA in colostrum, transitional, and mature milk). The highest level of total PUFA n-3 content was found in colostrum (1.8% of total FA), transitional (2.1% of total FA), and mature milk (2.4% of total FA) from Suzhou (Table 2), which, as a consequence, showed the lowest n-6 to n-3 ratio (12.7% in colostrum, 10.9%in transitional milk, and 10.5% of total FA in mature milk). Table 2. Median fatty acid composition of HM expressed as g/100 g of total FA.

Phospholipids
PL classes were determined by LC-ELSD, as previously described by Giuffrida et al. [28] and the results are listed in Table 3.
We did not measure minor constituents, such as lysophosphatidylcholine, which may contribute only to small amounts of the infant's diet.
From the total population, total PL content in HM decreased along the lactation period from 33.0 in colostrum to 24.2 mg/100 mL in mature milk, being significant lower (p < 0.05) in mature milk (Table 3). PtdCho was the most abundant PL in HM (from 12.0 mg/100 mL in colostrum to 8.2 mg/100 mL in mature milk) followed by CerPCho (from 9.1 mg/100 mL in colostrum to 7.2 mg/100 mL in mature milk), PtdEtn (from 8.5 mg/100 mL in colostrum to 6.4 mg/100 mL in mature milk), PtdIns (from 1.8 mg/100 mL in colostrum to 1.5 mg/100 mL in mature milk), and PtdSer (from 1.5 mg/100 mL to 1.0 mg/100 mL in mature milk). The PL class distribution was similar in colostrum, transitional, and mature milk (Figure 2).

Gangliosides
Gangliosides were determined by LC-MS/MS as described by Giuffrida et al. [29] and the results are listed in Table 4.
From the total population, the amount of GD changed during the lactation period (Table 4), with GM3 significantly increasing (p < 0.05) from 3.8 mg/mL in colostrum to 10.1 mg/L in mature milk and GD3 significantly decreasing (p < 0.05) from 4.1 mg/mL in colostrum to 1.0 mg/mL in mature milk. Total gangliosides increased significantly (p < 0.05) from 8.0 mg/L in colostrum to 11.0 mg/L in mature milk (Table 4). However, variability was high and total ganglioside content ranged from 1.     Among the different cities, GM3 content was comparable (p > 0.05) in colostrum; GM3 highest content (p < 0.05) in transitional milk (7.7 mg/L) was observed in HM of lactating mothers from Guangzhou and in mature milk in lactating mothers from Guangzhou and Suzhou, at 10.5 and 10.8 mg/L, respectively (Table 4). Within mature milk (Figure 4) Beijing, Guangzhou, and Suzhou showed the highest GM3 content at 4-8 months (11.0 ± 3.9, 12.3 ± 5.5, and 15.6 ± 6.1 mg/L, respectively). The highest content (p < 0.05) of GD3 was observed in colostrum of lactating mothers from Suzhou (8.6 mg/L); GD3 content was comparable (p > 0.05) in transitional milk among the different cities and between Guangzhou and Suzhou in mature HM (Table 4). However, when considering mature milk at different lactation stages (Figure 4), Beijing, Guangzhou, and Suzhou showed the highest GD3 content at 12-30 days (0.9 ± 1.3, 1.1 ± 1.1, and 1.5 ± 2.2 mg/L, respectively). Suzhou showed the highest content (p < 0.05) of total GD in colostrum and mature milk (12.6 and 11.9 mg/L, respectively), the   Among the different cities, GM3 content was comparable (p > 0.05) in colostrum; GM3 highest content (p < 0.05) in transitional milk (7.7 mg/L) was observed in HM of lactating mothers from Guangzhou and in mature milk in lactating mothers from Guangzhou and Suzhou, at 10.5 and 10.8 mg/L, respectively (Table 4). Within mature milk (Figure 4) Beijing, Guangzhou, and Suzhou showed the highest GM3 content at 4-8 months (11.0 ± 3.9, 12.3 ± 5.5, and 15.6 ± 6.1 mg/L, respectively). The highest content (p < 0.05) of GD3 was observed in colostrum of lactating mothers from Suzhou (8.6 mg/L); GD3 content was comparable (p > 0.05) in transitional milk among the different cities and between Guangzhou and Suzhou in mature HM (Table 4). However, when considering mature milk at different lactation stages (Figure 4), Beijing, Guangzhou, and Suzhou showed the highest GD3 content at 12-30 days (0.9 ± 1.3, 1.1 ± 1.1, and 1.5 ± 2.2 mg/L, respectively). Suzhou showed the highest content (p < 0.05) of total GD in colostrum and mature milk (12.6 and 11.9 mg/L, respectively), the Among the different cities, GM3 content was comparable (p > 0.05) in colostrum; GM3 highest content (p < 0.05) in transitional milk (7.7 mg/L) was observed in HM of lactating mothers from Guangzhou and in mature milk in lactating mothers from Guangzhou and Suzhou, at 10.5 and 10.8 mg/L, respectively (Table 4). Within mature milk (Figure 4) Beijing, Guangzhou, and Suzhou showed the highest GM3 content at 4-8 months (11.0 ± 3.9, 12.3 ± 5.5, and 15.6 ± 6.1 mg/L, respectively). The highest content (p < 0.05) of GD3 was observed in colostrum of lactating mothers from Suzhou (8.6 mg/L); GD3 content was comparable (p > 0.05) in transitional milk among the different cities and between Guangzhou and Suzhou in mature HM (Table 4). However, when considering mature milk at different lactation stages (Figure 4), Beijing, Guangzhou, and Suzhou showed the highest GD3 content at 12-30 days (0.9 ± 1.3, 1.1 ± 1.1, and 1.5 ± 2.2 mg/L, respectively). Suzhou showed the highest content (p < 0.05) of total GD in colostrum and mature milk (12.6 and 11.9 mg/L, respectively), the highest content (p < 0.05) of total GD in transitional milk was observed in Guangzhou (10.7 mg/L) ( Table 4).
highest content (p < 0.05) of total GD in transitional milk was observed in Guangzhou (10.7 mg/L) ( Table 4).

Discussion
This study measured the FA, PL, and GD content and the profile of 539 HM samples from Beijing, Guangzhou, and Suzhou.

Discussion
This study measured the FA, PL, and GD content and the profile of 539 HM samples from Beijing, Guangzhou, and Suzhou.
Among different cities, over lactation time, HM from Beijing showed slightly higher SFA content ( Table 2), Guangzhou the highest MUFA content (Table 2), and Suzhou the highest n-3PUFA content ( Table 2).
It is known that the type of fat/oil in the maternal diet influences the FA composition of breast milk. Francois et al. [22] showed that the consumption of six different dietary fats, each providing a specific FA, caused an acute response in HM FA composition, especially within 24 h, and that the response remained significantly elevated for 1-3 days after consumption of dietary fat. Therefore, difference observed in HM FA composition may reflect variation in maternal diet [33].
However, a careful analyses of dietary habits of Guangzhou, Beijing, and Suzhou needs to be performed for correlating to HM composition.

Phospholipids
Several studies have recognized the importance of PL for infant growth [39][40][41]. At the same time, PL are involved in immunity and inflammatory responses [42], and in neuronal signaling [43].
PL content in HM significantly (p < 0.005) decreased along the lactation period from 33.0 in colostrum to 24.2 mg/100 mL in mature milk, in agreement with previous studies performed elsewhere [12,44]. The PL class distribution was similar in colostrum, transitional and mature milk ( Figure 2). PL as emulsifiers are essential for the solubilization of dietary fats and as a consequence for their digestion and absorption. In this regard, the higher content of PL in colostrum and transitional HM compared to mature milk might explain the good fat absorption from HM by the newborn, despite poor pancreatic secretion, as suggested by Harzer et al. [11]. A decrease in PL content in HM along the lactation stage might occur because the diameter of the milk fat globule membrane increases [11,45], decreasing the PL/TAG ratio [7,8].
Our study showed that PtdCho was the most abundant PL in HM (Figure 2), followed by CerPCho and PtdEtn, and PtdIns and PtdSer, in agreement with previous studies [3,11,12,44,46]. PtdCho and CerPCho are important sources of choline considered as an essential nutrient for infants. Choline is a precursory amino alcohol of the neurotransmitter acetylcholine, it acts by regulating the transduction signal, and serves as a source of methyl groups in intermediate metabolism, being considered essential for optimum development of the brain [7,8]. In addition, CerPCho can reduce cholesterol absorption between 20.4%-85.5%, depending on the ingested dose (0.1% and 5.0%, respectively) [47], being possibly involved in cholesterol regulation programming.
Among the different cities, Suzhou showed the highest total PL and PtdEtn levels in colostrum, transitional, and mature milk (Table 3). Dietary sources of PtdEtn may be lecithin from rapeseed oil, whose consumption may explain also the higher content of ALA in HM from Suzhou. However, a careful analyses of dietary habits of this region needs to be performed for correlating to HM composition.
It is well known [58] that lipid and liposoluble nutrients content increases towards the latter part of a feeding session, a phenomenon that has been corroborated by biochemical analyses of total milk fat in fore-milk, and hind-milk [59,60]. Therefore, in order to assure sample homogeneity in our study all efforts have been made to collect fully-expressed milk. Among the cited studies, only Bitman et al., Thakkar et al., Holmes et al., and Fischer et al. [44,48,54,56] refer to full breast milk samples, Sala-Vila et al. [12] to fore-milk, and no detailed sampling procedure is described in the other studies. Analysis performed in fore-milk and hind-milk rather than fully-expressed milk could explain the discrepancy among results.

Gangliosides
GD are widely distributed in almost all human tissues, with the highest amount found in neural tissue and extra-neural organs, such the lung, spleen, and gut. It has been reported that during the first stages of life, dietary GD may have an important role in preventing infections [61] and in cognitive development functions [10,62].
Ma et al. [64] suggested that the ganglioside concentrations in HM at any time point may be influenced by the mother's dietary intake of gangliosides or their precursors. It was demonstrated [67] that GD3 and GM3 are transferred across the human placenta using an ex vivo model of dually-perfused isolated human placental lobules, suggesting that they are available to the developing fetus. Therefore, a careful analysis of dietary habits in this region needs to be performed for correlating to HM GD composition.

Conclusions
In this study, FA, PL, and GD contents and compositions of HM from lactating women living in Suzhou, Guangzhou, and Beijing were evaluated.
HM was collected over a period of eight months, allowing the observation of lipid compositional changes during lactation.
SFA, MUFA, and PL content decreased during lactation, PUFA and GD content increased. Among different cities, over lactation time, HM from Beijing showed the highest SFA content, HM from Guangzhou showed the highest MUFA content, and HM from Suzhou showed the highest n-3PUFA content. The highest total PL and GD contents were observed in HM from Suzhou. In order to investigate the influence of the diet on maternal milk composition, a careful analysis of dietary habits of these population needs to be performed in future work.