Main

The supply and metabolism of polyunsaturated fatty acids in infancy is receiving considerable attention owing to a proposed importance of LCP status for infant growth and development(1). Relatively large amounts of long-chain metabolites of the polyunsaturated fatty acids, primarily AA and DHA, are incorporated during the perinatal period into membrane lipids of the developing nervous system and other organs(2). The availability of LCP in early life has been related to visual and cognitive development in some intervention studies in preterm and term infants. In utero, LCP are supplied by a preferential materno-fetal transfer(3), and after birth, breast-fed infants receive LCP from human milk lipids(4). In contrast, infants fed conventional formulas based on vegetable oils without preformed LCP depend on an endogenous synthesis from the precursors(1). The enzymatic biosynthesis of LCP involves consecutive desaturation and carbon chain elongation by C2 units. It has been questioned whether the activity of this complex enzyme system is sufficient to meet the LCP needs in young infants, because plasma and red blood cell lipid analyses revealed markedly lower LCP levels in term infants fed formulas without preformed metabolites than in infants fed human milk or formulas with LCP(5,6). Using stable isotope methodology, we could demonstrate that full-term infants convert LA to AA at an age of 18 ± 4 d after birth(7), and active conversion of essential fatty acids to n-6 and n-3 LCP has since been confirmed in further isotope studies(8–10). Because no information has been available thus far on essential fatty acid turnover and LCP synthesis in human infants immediately after birth, we studied the metabolism of LA during the first week of life.

METHODS

Apparently healthy, full-term newborn infants were enrolled into the study at the second day of life at the department of neonatology, Thomayer Hospital, Prague. Enrollment criteria included gestational age between 37 and 42 wk and birth weight between the 25th and 75th percentile for gestational age, planned exclusive breast-feeding throughout the study period, and the absence of any apparent disease, organ, or metabolic dysfunction or history of perinatal asphyxia. The study protocol was reviewed and approved by the ethical committee, Charles University of Prague, and informed parental consent was obtained.

On d 3 after childbirth, all mothers of the participating infants manually expressed ≥5 mL each of both fore and hind milk. Equal volumes of fore and hind milk samples were pooled and stored at -80°C.

In the morning of d 2 of life, 10 participating infants each received orally before breast-feeding 1 mg/kg body weight of uniformly labeled [13C]LA (Martek Biosciences Corp., Columbia, MD) dissolved in a phospholipid emulsifier (E-Mulsin, Mucos Pharma, Geretsried, Germany). In two additional infants, samples were collected according to the same study protocol but without administration of labeled LA to obtain information on the variation of 13C enrichment in expired air and blood lipids during the first days of life.

Samples of exhaled breath were obtained before and at 30-min intervals after tracer administration over a period of 6 h. Using a subnasal prong connected via a three-way valve to a 20-mL syringe, exhaled air was manually aspirated during the second half of the expiratory phase and transferred into Vacutainers (10 mL, Labco Ltd., Wycombe, England) for storage until analysis. On the same morning, total CO2 production was determined during a 30-min period at 1-2 h after feeding by indirect calorimetry with a Deltatrac II metabolic monitor (Deltatrac II MBM-200 Metabolic Monitor, Datex/Division of Instrumentarium Corp., Helsinki, Finland) using the baby spontaneous breathing-canopy mode.

Blood samples were obtained in EDTA-coated vials before feeding in the morning of d 2, i.e. before tracer administration, and thereafter in 24-h intervals for three consecutive days. Plasma phospholipid fatty acid and isotopic analyses were performed as previously described(7). Briefly, lipids were extracted from 0.5 mL of plasma with chloroform/methanol (1:1, vol/vol), and lipid classes were separated on silica gel plates by thin-layer chromatography. The bands containing phospholipids were identified by comparison with appropriate standards and removed by scraping, and the fatty acids were trans-esterified with 3 M methanolic hydrochloric acid. After neutralization, fatty acid methyl esters were extracted with 3 × 1 mL of hexane, and the pooled residues were taken to dryness and finally dissolved in 100 µL of hexane containing butylhydroxytoluene (2 g/L). Aliquots of this solution were used for GC analysis, which was performed on a Hewlett-Packard 5890 series II system (Hewlett Packard, Waldbronn, Germany). After on-column injection, separation of individual fatty acid methyl esters was achieved on a BPX 70 column (SGE, Weiterstadt, Germany) of 50-m length and 0.32-mm inner diameter. The oven temperature program started at 130°C, followed by an increase of 3°/min until 200°C and from there on with 1°/min up to the final temperature of 210°C. For identification of fatty acid methyl esters and determination of response factors, commercially available standards were used (NuChek Prep, Elysian, MN). For quantitative analysis, flame ionization detection was used, and 13C content was determined by GC-combustion-isotope ratio mass spectrometry (delta S, Finnigan MAT, Bremen, Germany), as described in detail elsewhere(7). For the determination of the fatty acid composition of human milk, lipids from 1-mL samples were extracted into chloroform/methanol (2:1), dried, and derivatized with the same procedure as used for plasma lipids.

CO2 from breath samples was purified by GC (column: Poraplot Q, 25-m length, 0.32-mm inner diameter, Chrompack, Frankfurt/Main, Germany) and introduced into the ion source of the isotope ratio mass spectrometer via the combustion interface. Results on 13C-enrichment are expressed as δ-over baseline (DOB, %‰) values based on calculated δ13C values relative to the PeeDeeBelemnite carbonate standard(11).

Changes in δ13C values were evaluated by Wilcoxon and Friedman tests using SPSS for Windows, Version 6.1.3. Probability values ≤0.05 were considered statistically significant.

RESULTS

Ten healthy infants with a birth weight of 3.4 ± 10.2 kg (mean ± SD) and a gestational age of 40.5 ± 1.1 wk participated in the stable isotope study, and two infants with birth weights of 3.7 kg and 3.3 kg and gestational ages of 39 and 40 wk, respectively, were studied for variation of 13C enrichment after birth without tracer application.

The essential fatty acid composition of colostral human milk lipids that we determined here (Table 1) is similar to previously published results on human milk in central Europe(4,12).

Table 1 Fatty acid composition (% wt/wt of total fatty acids) of human milk samples expressed by the 10 mothers of infants participating in the stable isotope study
Full size table

From d 2 to 5 of life there were significant increases of total plasma phospholipid fatty acid content as well as total saturated and total monounsaturated fatty acids, LA, C20:2n-6, and C20:5n-3. During the same period, DGLA, AA, and DHA showed insignificant trends to increase (Table 2). Percentage contributions of LA to plasma phospholipid fatty acids increased from d 2 to d 5 of life (Table 2), whereas percentages of DGLA decreased and the percentage of AA showed an insignificant trend to decrease. Percentage values of DHA decreased significantly during the study period.

Table 2 Plasma phospholipid fatty acid composition (% wt/wt) and concentrations (mg/L) in 10 newborn infants fed human milk on d 2-5 of life (mean ± SE)
Full size table

Infant CO2 production was 16.7 ± 10.57 mL/min (mean ± SE). In the two infants not receiving tracer, we found a variation (SD) of the δ13C-value in breath CO2 of 0.18 and 0.28‰, respectively, throughout the observation period. In contrast, infants who received [13C]LA showed a rapid and marked increase of 13C enrichment in breath CO2, with maximal DOB values in expired air 4.2 ± 1.1 h after tracer administration (Fig. 1). Six hours after tracer administration, some 7.4 ± 0.6% (mean ± SE) of administered label, which can reasonably be assumed to be well absorbed(13), was recovered in expired CO2 (Fig. 2).

Figure 1
figure 1

Increase of 13C enrichment in breath CO2 (DOB, [‰, mean ± SE]) of 10 newborn infants after oral application of 1 mg/kg body weight of uniformly 13C-labeled linoleic acid.

Full size image
Figure 2
figure 2

Cumulative recovery of tracer 13C in breath CO2 (%, mean ± SE) over 6 h after oral application of 1 mg/kg body weight of uniformly 13C-labeled linoleic acid on d 2 of life in 10 newborn infants.

Full size image

δ13C values of phospholipid fatty acids showed a variation of 0.3 and 1.7%‰ for LA, 0.4 and 0.8%‰ for DGLA, and 0.3 and 0.5%‰, respectively, for AA in the infants who did not receive tracer. In the infants who received labeled LA, plasma phospholipid LA showed significant (p < 0.001) enrichment at all points with a maximum at 24 h after tracer application (Table 3 and Fig. 3). Enrichment of the intermediary metabolite DGLA showed a gradual, significant (p < 0.001) increase from baseline to 72 h after tracer application (DOB at 24 h 2.1 ± 0.5%‰, 48 h 3.7 ± 0.9%‰, 72 h 4.4 ± 1.0%‰). Although AA enrichment tended to increase, there was a large interindividual variation (Table 3), and no significant change over baseline was observed (DOB at 24 h 0.1 ± 0.2%‰, 48 h 0.2 ± 0.2%‰, 72 h 0.3 ± 0.2%‰). The maximal DOB values for LA reached at any time correlated with the maxima of DGLA (r = 0.62, p < 0.005), but the percentage of tracer recovery in breath was not related to the enrichment in LA metabolites.

Table 3 Maximal 13C enrichments (DOB, %) of LA, DGLA, and AA from phospholipids of the neonates studied and the calculated AUC (µmol tracer 13C × h/L). The relative contribution of each fatty acid to the total AUC of the n-6 fatty acids is given in parenthesis.
Full size table
Figure 3
figure 3

13C enrichment (mean + SE) of n-6 fatty acids in serum phospholipids of 10 newborn infants after oral application of 1 mg/kg body weight of uniformly 13C-labeled LA, at 24-h intervals for three consecutive days after tracer application, (DOB [‰], LA values divided by 10, AA values multiplied by 10; * p < 0.002 vs baseline).

Full size image

Tracer concentrations (µmol 13C/L plasma) were calculated by multiplying the concentration of micromoles of carbon found in the form of a given fatty acid times the percentage 13C increase over basal 13C content. AUC were calculated from the tracer concentrations by averaging neighboring values, multiplying by the time span between the measuring points, and summing these values during the observation time. No further curve-fitting or extrapolation to infinity was performed. AUC for individual fatty acids are given as micromoles of tracer 13C × h/L plasma and are expressed as a percentage of the total AUC of all n-6 fatty acids studied (Table 3). By calculating the AUC, the dilution of the tracer in the various pools of the n-6 fatty acids with their different pool sizes is taken into account, but it is not considered that these pools might well have different turnover rates. Furthermore the results are dependent on the time schedule and duration of blood sampling.

DISCUSSION

With the tracer method used here we demonstrated active synthesis of n-6 LCP in newborn infants immediately after birth. The application of this stable isotope study protocol during the first days of life proved to be feasible and did not lead to any major interference with the daily routines of the participating infants and their mothers or their well-being. No adverse effects of the tracer application or other aspects of the study were observed.

The proportion of dietary LA oxidized in these infants during the first days of life is similar to the rate of LA oxidation in healthy lactating women (10.7 ± 2.4% [mean ± SE] recovery of LA label 24 h after tracer application)(14). Thus, newborn infants after birth appear not to have a proportionally greater conservation of dietary LA than adults in spite of their high essential fatty acid requirements and higher energy expenditure. This is in agreement with studies in LA-deficient growing rats, which showed partial oxidation of exogenously supplied LA(15). As we found no relation of the proportion of LA oxidized to LCP enrichments, it appears that there is no direct competition of LA oxidation and conversion under the conditions of this study. However, only about 3% of the LA tracer dose was converted to LCP based on the calculated AUC values, which might be the reason for the absence of any significant correlation between the tracer recovery in breath and the enrichments detected in the LCP metabolites. As apparently only a minor portion of the exogenous LA is transferred into LCP, the amount of substrate available should not be a limiting factor.

AA enrichments were not significantly different from zero, and hence endogenous synthesis seems to contribute little to the plasma AA pool of breast-fed infants in the first week of life. In accordance with this observation, we have previously estimated that only about 6% of the total plasma AA pool is renewed each day by endogenous synthesis in infants at an average age of about 3 wk(7). In human milk-fed preterm infants, Carnielli et al.(16) found that 6% of the total dose of [13C]LA was converted into plasma phospholipid AA. In the infants studied here the intraindividual variation of n-6 fatty acid enrichments was similar to the variation reported by Sauerwald et al.(17) for term infants at an older age.

It has been suggested that one factor limiting AA synthesis might be a low enzymatic activity of Δ-5-desaturation(14). Indeed, in vitro studies in human epidermal keratinocyte cell cultures demonstrated relatively high activities of Δ-6-desaturase and elongase, but a low activity of Δ-5-desaturase(18). The markedly higher 13C enrichment of DGLA, compared with AA (Fig. 3), could be interpreted as an indicator of relatively low conversion of DGLA to AA and hence a low Δ-5-desaturation activity. Previously we also found such a markedly higher 13C enrichment of DGLA, compared with AA, in older infants(7). However, the higher 13C enrichment of DGLA than AA may also be related to a dilution of endogenously synthesized, 13C-labeled AA by the considerable exogenous intake of preformed AA with human milk (Table 1) and by the relatively large amount of AA in the phospholipid pool (Table 2). Also, the 72-h duration of this study might have been too short to detect maximal AA enrichment. In contrast to the difference in enrichment between DGLA and AA, their AUC of the tracer concentrations are similar because the concentration of AA is about 5 times higher than the concentration of DGLA. As the AUC is proportional to the average tracer concentration during the 72 h period of sampling, this leads us to assume that a major portion of the labeled DGLA is converted to AA. Although this is only an approximate estimate because of the simplifications introduced by the AUC calculation and the limited period of observation, it seems that only a small portion of the applied [13C]LA is converted to [13C]DGLA via Δ-6-desaturation and elongation whereas a substantial amount of the [13C]DGLA is Δ-5-desaturated to yield [13C]AA.

The changes of plasma phospholipid fatty acid values in the first days of life observed in our subjects, i.e. the increase of total fatty acids, LA, and AA, are in accordance with results of Carnielli et al.(16). The increasing trend of DGLA, AA, and DHA concentrations in our subjects appears to reflect the onset of postnatal supply with human milk. However, the percentage contribution of these LCP shows a decreasing trend (Table 2), which suggests that the plasma disapperance rate, relative to endogenous production and intake, might be higher for LCP than for other fatty acids, such as LA, which is found in higher concentrations in breast milk.

The results obtained here in breast-fed infants with a dietary LCP supply cannot be directly extended to infants fed formulas without these compounds, as the total dietary balance of fatty acids seems to influence the metabolism of individual fatty acids(19). A marked postnatal decrease of AA contents in plasma lipids was observed in infants fed formulas without preformed LCP and was interpreted as evidence of a limited LCP synthesis(16). Further studies should investigate whether the activity of LCP formation is modulated by their dietary supply, because it is conceivable that preformed dietary LCP may modulate the rate of infant synthesis.

In conclusion, these results on plasma phospholipid fatty acid enrichment clearly demonstrate active endogenous synthesis of n-6 LCP in full-term neonates already during the first days after birth. Based on calculated tracer mass in plasma phospholipids, a relatively low portion of exogenous LA is converted to DGLA whereas a major portion of DGLA is desaturated to AA.