Main

Since the discovery of vitamin E by Evans and Bishop in 1922(1), interest has increased enormously in free radical research, and meanwhile vitamin E has been recognized as a major chain-breaking antioxidant. During the last 20 y, there has been much concern about a possible marginal vitamin E deficiency in premature infants and its role in causing or exacerbating neonatal oxygen radical diseases, such as respiratory distress, bronchopulmonary dysplasia, retinopathy of prematurity, intraventricular hemorrhage, or necrotizing enterocolitis(2–4).

It is well known that placental transfer of vitamin E is limited, and the expectant mother has little influence on the delivery of this substance to the fetus(5). Fetal tissue concentrations are gestation-related and "stores" of vitamin E are not accumulated in utero. Therefore, the premature infant may slip into a precarious situation with respect to nutritional adequacy(6). Vitamin E is taken up by several tissues, including liver, lung, heart, skeletal muscle, and adipose tissue(7). Thus, a nutritional deficiency may affect many organs, although the signs and symptoms may be subtle. There is also controversy over the issue of whether or not vitamin E deficiency should be defined by biochemical abnormalities or by clinical disturbances. So far, the evaluation of vitamin E status has depended on biochemical analysis of plasma, erythrocytes, adipose tissue biopsies, and organs obtained at autopsy(8). Recently, the assessment of vitamin E nutritional status in neonates, infants, and children on the basis of α-tocopherol levels in blood components and buccal mucosal cells as a substitute for determination in adipose tissue has been suggested(9).

Plasma tocopherol levels in premature infants have been reported to be much lower than in adults and their RBC are more susceptible to in vitro oxidative stress(10,11). Because of the marked influence of circulating lipids on tocopherol concentrations, tocopherol to lipid ratios have been used to estimate vitamin E nutritional status(12). An alternative way to assess vitamin E status is the determination of RBC tocopherol. It has been reported that the majority of neonates have RBC tocopherol levels comparable to those of adults(13–15). A marginal deficiency of vitamin E in neonates can be defined by determination of PLT tocopherol as well as MN, PMN, or BMC tocopherol(9), with BMC being a substitute for adipose tissue due to its similar fatty acid composition(16).

Because colostrum and preterm milk contain approximately two to three times more α-tocopherol than mature milk(17), it remains to be evaluated if premature infants fed nonsupplemented preterm milk develop biochemically evident vitamin E deficiency due to impaired intestinal resorption or if the higher content of vitamin E in preterm milk avoids deficiency. The most active vitamin E isomer, α-tocopherol, is also the most abundant in human milk. Moreover, considerable amounts ofγ-tocopherol are transferred by breast milk, but this isomer has only about 10% of the biologic activity of α-tocopherol(7). To determine the adequacy of vitamin E in preterm milk for vitamin E status in healthy preterm infants, we measured α- and γ-tocopherol in various cells as well as in preterm milk samples, using HPLC fluorescence detection, a method that enables the assay of very small amounts of tocopherol. We tested the hypothesis that healthy premature infants fed on their own mothers' milk do not need vitamin E supplementation.

METHODS

Subjects. Fourteen healthy, premature infants (seven boys and seven girls) appropriate for their gestational age with birth weights from 1090 to 2170 g (mean ± SD of 1439 ± 364 g), body lengths at birth from 36 to 44 cm (mean ± SD of 40 ± 3 cm), head circumferences at birth from 24.3 to 31 cm (mean ± SD of 27.5 ± 2.1 cm), and gestational ages from 27 to 32 wk (mean ± SD of 30± 1.7 wk) were studied. All infants received at least 90% of their oral feeding as their mothers' milk. None of the infants had severe complications due to prematurity, such as birth asphyxia, severe respiratory distress, or apneas and bradycardias requiring ventilation longer than 48 h. All infants were meanwhile followed up until the age of 18 mo; none showed impairment of psychomotor skills. Complete oral feeding (including breast feeding, bottle feeding, and gavage feeding) with mothers' milk was achieved at 12 ± 3 d (mean ± SD). By then full caloric intake of 120 kcal·kg-1 d-1 and a fluid intake of 160 mL·kg-1 d-1 was achieved. Weight gain averaged 20± 10 g/d. All infants received a protein/calcium/phosphorus supplement(FM 85/Nestlé; Vevey, Switzerland) when complete oral intake was achieved. Before full oral intake the infants received parenteral nutrition without fat or vitamin E supplements. Oral feeding was started at 4-12 h of birth with glucose and gradually changed to mothers' milk after meconium was produced at least once. Parenteral nutrition contained glucose and minerals, and if full oral nutrition was not completed by d 5, amino acids were added. None of the infants developed diseases possibly linked to oxygen radical disorders such as bronchopulmonary dysplasia, retinopathy of prematurity, intraventricular hemorrhage, or necrotizing enterocolitis.

α- and γ-tocopherol levels were determined in cord blood and on d 0 to 1, 7, 14, 28, and 42 after birth. Total lipid content of each plasma sample was also analyzed.

The study protocol was approved by the ethics committee of the Hospital Center of the University of Heidelberg, and informed consent was obtained from the parents.

Sample preparation. One milliliter of blood was drawn in plastic tubes with EDTA from the placenta directly after birth and by venipuncture from the premature infants within 48 h of birth (before milk feeding), as well as at 7, 14, 28, and 42 d after birth. The blood was centrifuged at 400 × g and 10°C for 10 min to separate two layers. The top layer was then transferred to a sterile tube containing 1 mM EDTA, and was centrifuged at room temperature for 10 min at 1500 ×g to obtain platelet-free plasma and a platelet pellet. Platelets were washed and suspended in 0.9% NaCl for the tocopherol assay. To isolate RBC, MN, and PMN from the bottom layer of the first centrifugation, the Hypaque-Ficoll method was applied as follows. The sample was gently placed on a Mono-Poly Resolving Medium (Flow Laboratories, Australia) and centrifuged at 800 × g and 10°C for 30 min to give three layers, i.e. a top layer containing MN, a middle layer containing PMN, and a bottom layer containing RBC. Each of the layers was transferred to a separate glass tube, washed three times with 0.9% NaCl, and resuspended. MN and PMN were cleansed of contaminating RBC by hypotonic lysis(18).

BMC was collected by gentle scratching with a spatula from premature infants before feeding. The first sample was obtained within 24 h after birth. Cells adhering to the spatula were suspended and washed three times with 0.9% NaCl. Finally, the cells were suspended in 1 mL 0.9% NaCl and used for the tocopherol assay(19).

Two milk samples (1 mL) were collected from each mother. At least 10 d elapsed between the two sample collections. Samples were collected on d 14(n = 1), d 21 (n = 1), d 28 (n = 12), d 35(n = 4), or d 42 (n = 10). Each sample was placed in a glass tube with EDTA and used for tocopherol assay(19).

Tocopherol assay. Extraction of α- and γ-tocopherol was carried out as follows(9). One milliliter of 6% pyrogallol in ethanol solution was added to 1 mL of the sample to prevent tocopherol oxidation. After preincubation at 70°C for 2 min, the mixture was saponified with 0.2 mL of 60% KOH at 70°C for 30 min. After saponification, the mixture was immediately cooled in an ice bath, and then 2.5 mL of distilled water and 5 mL of n-hexane were added. The mixtures were vigorously shaken for 5 min and centrifuged at 1500× g for 5 min to obtain a hexane layer. Four milliliters of the hexane layer were taken up and evaporated. The residue was reconstituted with an appropriate volume of ethanol, and an aliquot of it was injected into the HPLC system.

The instruments used in the study were as follows: the HPLC was a Beckmann Gold-Basic-System 477 131 equipped with a RP-18T, ODS column (4 × 250 mm). The eluent was methanol/water 1000/5, and the flow rate was 1 mL/min. The detector was a Beckmann fluorescence detector, α- andγ-tocopherol were determined at an extinction of 298 nm and an emission of 330 nm. Retention times were 8 min for γ-tocopherol and 9 min forα-tocopherol, respectively. An external standard (mixture of 100µg/dL α-tocopherol and 50 µg/dL γ-tocopherol in ethanol) was used.

The PL α- and γ-tocopherol levels were expressed asµg/dL, and the tocopherol content of RBC suspensions was corrected by the hematocrit to achieve a hematocrit of 50% (µg/dL of packed cells). The protein content of PLT and BMC was determined by the method of Lowry et al.(20), and tocopherol levels were expressed asµg/g of protein. The suspended MN and PMN cells were counted, and the tocopherol content was expressed as µg/109 cells.

RESULTS

Mothers' milk vitamin E. Table 1 gives the α- and γ-tocopherol levels of the milk samples analyzed.α-Tocopherol levels were approximately eight times higher thanγ-tocopherol levels, being 406 ± 106 µg/dL(α-tocopherol) versus 52 ± 31 µg/dL(γ-tocopherol). The range of γ-tocopherol was extremely high(2-144 µg/dL), lower α-tocopherol levels were associated with lower γ-tocopherol levels and vice versa.

Table 1 α- and γ-tocopherol contents in mothers' milk
Full size table

Changes in α-tocopherol during the first 6 wk after birth. Table 2 shows the changes of α-tocopherol values in PL, tocopherol to lipid ratio in PL, RBC, PLT, BMC, MN, and PMN. In all blood cell types, cord blood values were higher than the values determined within 48 h of birth (i.e. d 0 to d 1). It is interesting to note that, although PL α-tocopherol hardly changed from birth (cord blood) to the first postnatal blood sample, the tocopherol to lipid ratio showed a significant drop from 0.91 to 0.51 (p < 0.05), falling below the ratio of 0.8 defined as acceptable by Martinez et al.(21). The only other significant drop from cord blood to d 0 or 1 occurred in PLT(101 ± 45 µg/g of protein to 46 ± 23 µg/g of protein; p < 0.05); in RBC, MN, and PMN the decreases were not significant. In all cells except RBC, the lowest values were seen within the first 7 d of life. As previously described for neonates(9), RBC values did not fall during the first week after birth. Lowest values were reached on d 14, but RBC values never decreased below 162 µg/dL of packed cells, defined as the lower limit of a normal control group aged 3-16 y(9). Only BMC stayed below the lower limit of 22µg/g of protein(9) during the first week. Together with the low tocopherol to lipid ratio of 0.51 on d 2, the low BMC values during the first week suggest a marginal biochemically determined vitamin E deficiency in healthy, premature infants. A marked rise in α-tocopherol in all fractions was seen between d 28 and 42. α-Tocopherol rose in all cell types between d 28 and 42, but remained within normal ranges for children. α-Tocopherol doubled in PL, RBC, BMC, and MN and tripled in PLT and PMN compared with the lowest values. There never was a decrease below the lower limits of normal(9) for α-tocopherol in PLT, MN, and PMN, namely 36 µg/g of protein, 3.6 µg/109 cells, and 2.1 µg/109 cells.

Table 2 α-Tocopherol status in premature infants<33 wk of gestation fed mothers' milk without vitamin E supplementation during the first 6 wk of life
Full size table

Correlations between gestational age and α-tocopherol levels in cord blood and later on were calculated. In cord blood, there was a positive correlation only for PL α-tocopherol (r = 0.612), whereas in none of the cell types could a correlation be detected. On d 0 to 1, gestational age correlated with BMC α-tocopherol (r = 0.787), whereas there was no correlation between gestational age and the other fractions studied. From d 7 through 42, there was no correlation between gestational age and α-tocopherol.

Changes in γ-tocopherol during the first 6 wk after birth. Table 3 shows the changes of γ-tocopherol values in PL, RBC, PLT, BMC, MN, and PMN. γTocopherol in cord blood averaged 1.3, 1.4, 2.5, 3.7, and 4.2% of α-tocopherol in PL, RBC, PLT, MN, and PMN, respectively. Within the first 14 d of life, by the time all infants were on full enteral mothers' milk (mean ± SD of 12 ± 3 d), γ-tocopherol amounted to 6, 11, 12, 16, 10, and 9% ofα-tocopherol values in PL, RBC, PLT, BMC, MN, and PMN, respectively. In mothers' milk, γ-tocopherol amounted to 13% of γ-tocopherol, showing that the take-up in cells correlated well with the enteral take-up ratio of α- to γ-tocopherol. It is interesting to note thatγ-tocopherol in PL was 6% of PL α-tocopherol, being only half as high as the percentage of γ-tocopherol in mothers' milk. No correlation existed between gestational age and γ-tocopherol in any of the fractions during the entire study period.

Table 3 γ-Tocopherol status in premature infants < 33 wk gestation fed mothers' milk without vitamin E supplementation during the first 6 wk of life
Full size table

Comparison of α- and γ-tocopherol changes during the first 6 wk after birth. Figures 1–6 show the changes of α- and γ-tocopherol in PL, RBC, PLT, BMC, MN, and PMN. PL α-tocopherol rose steadily from d 1 to 42(p < 0.05). In contrast, PL γ-tocopherol rose significantly from d 14 to 28 (+38%) and again from d 28 to 42 (+56%). RBCα-tocopherol showed a gradual rise during the study period (p> 0.05). RBC γ-tocopherol, however, rose significantly from d 14 to 28 and d 28 to 42, similar to PL γ-tocopherol. PLT α-tocopherol increased 3-fold from d 28 to 42, whereas the rise of PLT γ-tocopherol was 2-fold. BMC α-tocopherol rose moderately from d 1 to 42 (p> 0.05), similar to BMC γ-tocopherol. In MN, α- andγ-tocopherol rose gradually with significant increases from d 7 to 42 for α-tocopherol and d 1 to 42 for γ-tocopherol (p > 0.05). Similar changes were found for α- and γ-tocopherol in MN and PMN from d 1 to 42 (p > 0.05).

Figure 1
figure 1

PL α- and γ-tocopherol during the first 42 d after birth in healthy, premature infants.

Full size image
Figure 6
figure 2

PMN α- and γ-tocopherol during the first 42 d after birth in healthy, premature infants.

Full size image

DISCUSSION

So far, many guidelines have been formulated to delineate the necessary vitamin E intake for premature infants, evolving from the fact that prematurely born infants are regarded to be marginally vitamin E-deficient(21,22). Human vitamin E deficiency has been defined as a PL α-tocopherol level below 500 µg/dL accompanied by a tocopherol to lipid ratio below 0.8 or abnormal hydrogen peroxide hemolysis of erythrocytes. The term "subclinical" or biochemical deficiency has evolved to describe laboratory evidence of low tocopherol levels in circumstances that are not clearly accompanied by clinically evident deficiency(8).

The total intake of vitamin E with the diet depends on the volume of feeding. The recommended daily allowance of α-tocopherol for preterm infants is 2800 µg·kg-1 d-1. The preterm infants of our study received feeds of about 160 mL·kg-1 d-1. Thus, their α-tocopherol intake was only 650 ± 170µg·kg-1 d-1. The AAP Committee on Nutrition has suggested a safe and effective plasma vitamin E level between 1000 and 2000 µg/dL(23), a level that was achieved in the infants only on d 42, when adding α- and γ-tocopherol levels together, although it must be kept in mind that γ-tocopherol amounts to only 10% of the biologic activity of that of α-tocopherol. Ifα-tocopherol alone is seen as relevant for vitamin E status, the recommended range of 1000 to 2000 µg/dL was not achieved in most infants during the first 42 d after birth (highest α-tocopherol 855± 263 µg/dL on d 42). The ESP-GAN report from 1991(24) recommends α-tocopherol equivalents ≥600µg/100 kcal of milk. When assuming that mothers' milk has approximately 70 kcal/dL, the milk studied has approximately 587 ± 155 µg/kcalα-tocopherol equivalents, being only slightly below the recommended value.

In the present study, we determined α- and γ-tocopherol in different blood cells and BMC. In most previous studies, vitamin E status in premature infants was evaluated by determination of PL α-tocopherol alone, sometimes in combination with the tocopherol to lipid ratio. When regarding these two parameters in our study, a supplementation of vitamin E in premature infants may seem justified due to the low PL α-tocopherol and tocopherol to lipid ratio on d 0 to 1 (329 ± 154 µg/dL and 0.51 ± 0.25, respectively, see Table 2). All subsequent measurements from d 7 to 42 show PL α-tocopherol and tocopherol to lipid ratios within the range regarded as vitamin E-sufficient(9). The critical time period in a premature infant's life to develop hazards related to vitamin E deficiency and oxidative stress is in the perinatal period, presumably in the first 24 h of life. Ventilation and oxygen supplementation may lead to an overload with reactive oxidative substances, thereby increasing the risk for severe sequelae such as retinopathy of prematurity, bronchopulmonary dysplasia, and intraventricular hemorrhage. Therefore, low PL α-tocopherol on the first postnatal day may require prophylactic vitamin E supplementation. In our study, we examined healthy premature infants without clinical signs of increased risk for oxidative injury. When regarding different blood cell types such as RBC, PLT, MN, and PMN, all values were within the normal range of healthy newborn infants(9), indicating vitamin E sufficiency during the first 2 d of life in these infants. BMC α-tocopherol on d 0 to 1 was only 13 ± 6 µg/g of protein, still being within the normal range of healthy newborn infants(9). On the other hand, it must be discussed if sick infants will profit from an artificial elevation of vitamin E levels above the values measured here.

It is well known that tocopherol is delivered to tissues via the high affinity receptor for LDL(25). Because neither RBC nor PLT have LDL receptors, the determination of RBC and PLT tocopherol alone is not sufficient to give a good assessment of vitamin E, considering the necessity of vitamin E in many body tissues. BMC, MN, and PMN are easily accessible cells that possess LDL receptors and can therefore be used as indicators of tissue that can be obtained by biopsies only. BMC are of special interest due to their similar fatty acid composition compared with adipose tissue(16).

It is interesting to note that the corresponding γ-tocopherol values were much lower in relation to the α-tocopherol values during the first 2 d than in the subsequent weeks. The γ- to α-tocopherol ratio in PL and RBC increased from 0.01 at birth to 0.14 and 0.15 at 6 wk. In PLT the ratio rose from 0.02 to 0.11 and in PMN from 0.04 to 0.24. It is unclear why the γ- to α-tocopherol ratios were much lower at birth than later on. This phenomenon may be due to a not yet detected placentalα-tocopherol transfer protein.

In conclusion, the simultaneous determination of α- andγ-tocopherol levels in cells with LDL receptors (BMC, MN, and PMN) as well as PL, RBC, and PLT reveal a supply with vitamin E, which can be regarded as adequate in healthy, premature infants fed mothers' milk not supplemented with vitamin E during the first 6 wk of life. In our opinion, supplementation should be considered only if the infants either show severe asphyxia directly after birth or if evidence for neonatal oxygen radical disease exists. Taking the increased peroxidative potential of biomembranes in premature infants into account(26–28), vitamin E supplementation may be required when premature infants are exposed to active oxygen species under highly oxidative conditions. Further studies are needed to depict those sick premature infants at the earliest time possible who would profit from early vitamin E supplementation.

Acknowledgment. The authors thank S. Wolfrum, B.Sc., for technical assistance.

Figure 2
figure 3

RBC α- and γ-tocopherol during the first 42 d after birth in healthy, premature infants.

Full size image
Figure 3
figure 4

PLT α- and γ-tocopherol during the first 42 d after birth in healthy, premature infants.

Full size image
Figure 4
figure 5

BMC α- and γ-tocopherol during the first 42 d after birth in healthy, premature infants.

Full size image
Figure 5
figure 6

MN α- and γ-tocopherol during the first 42 d after birth in healthy, premature infants.

Full size image