This article was downloaded by: [ECU Libraries] On: 23 April 2015, At: 00:12 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Journal of the American College of Nutrition Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uacn20
Intravenous vitamins for very-low-birth-weight infants. a
a
a
a
a
a
H L Greene , R Smith , P Pollack , J Murrell , M Caudill & L Swift a
Vanderbilt University School of Medicine, Department of Pediatric Nutrition, Nashville, Tennessee 37232-2576. Published online: 02 Sep 2013.
To cite this article: H L Greene, R Smith, P Pollack, J Murrell, M Caudill & L Swift (1991) Intravenous vitamins for very-lowbirth-weight infants., Journal of the American College of Nutrition, 10:4, 281-288, DOI: 10.1080/07315724.1991.10718154 To link to this article: http://dx.doi.org/10.1080/07315724.1991.10718154
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Intravenous Vitamins for Very-Low-Birth-Weight Infants Harry L. Greene, MD, Rita Smith, BS, Paul Pollack, MD, Joel Murrell, MT, Melinda Caudill, MT, and Larry Swift, PhD Vanderbilt University School of Medicine, Department of Pediatrie Nutrition, Nashville
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Key words: very-low-birth-weight infants, vitamins, vitamin A, vitamin E, vitamin B:, vitamin B6, total parenteral nutrition Term infants and children appear to adapt to large variations in vitamin intakes. This is supported by thefindingof similar blood levels of vitamins despite several-fold differences in intake on a body weight basis. By contrast, the finding of marked elevation of some vitamins and low levels of others seen in very-low-birth-weight ( VLBW) infants (< 1500 g) suggest that this group has less adaptive capacity to high- or low-dose intakes. This indicates that their vitamin intakes need to be more closely aligned with actual needs. This paper reviews previously published data on vitamins A, E, B2, and B6 from VLBW infants receiving total parenteral nutrition (TPN). Vitamin A. VLBW infants are relatively deficient in retino! (R) at birth. During TPN large losses of R onto the delivery sets result in a further decline in stores of R as reflected in a progressive decline in plasma R during TPN. Because of the reported lower incidence of bronchopulmonary dysplasia associated with intramuscular vitamin A treatment, alternative methods of vitamin A delivery during TPN have been suggested. First, the vitamins were mixed with Intralipid (IL) and, second, retinyl palmitate (RP) rather than R was used. There was little in vitro loss of R when mixed with IL, and in vivo treatment resulted in higher blood levels after 1 month of rctinol administration in IL than seen previously (21.4 ± 4.2 vs 14.1 ± 3.7 μg/dl). Use of RP in VLBW infants resulted in high RP levels (40 ± 6 μg/dl), although R levels were similar to that seen with R added to IL (21.1 ±4 μg/dl). Using these data and those from other publications, currently suggested intravenous intake of vitamin A as R is 500 μg/kg/day. Vitamin E. The TPN solution for pediatrie patients contains oc-tocopherol acetate. Little of the vitamin is lost to the plastic infusion sets. Infusion of four different dosage levels suggests that doses of 2.8-3.5 mg/kg/day will maintain most infant blood levels between 1 and 2 mg/dl. Vitamin B2. Vitamin B2 is activated to its active cofactor forms flavin mononucleolidc (FMN) and flavin adenine dinucleotide (FAD). Three doses of riboflavin (0.68, 0.56, and 0.34 mg/kg/day) resulted in elevated blood levels. Using these blood response doses, a projected intake of 0.15 mg/kg/day appears more appropriate to maintain blood levels in the range of those seen in formula-fed term infants. Vitamin Bft. Vitamin B(1 is converted in vivo to pyridoxal and activated to its cofactor form, pyridoxal phosphate (PLP). Three doses of pyridoxine (from 0.2 to 0.5 mg/kg/day) resulted in elevated blood levels. Using the blood response to these doses, an intake of 0.18 mg/kg/day is projected to maintain PLP levels within the range of that seen in plasma samples from formula-fed term infants. Abbreviations: AA = amino acid, AAP = American Academy of Pediatrics, BW = birth weight, EGOT = erythrocyte glutamic-oxaloacetic transaminase, EGR = erythrocyte glutathione reductase, FAD = flavin adenine dinucleotide, FMN = flavin mononucleotide, HPLC = high-performance liquid chromatography, IL = Intralipid, PLP = pyridoxal phosphate, R = retinol, RP = retinyl palmitate, TPN = total parenteral nutrition, VLBW = very low birth weight
INTRODUCTION Nutritional support of preterm infants requires special considerations compared to term infants for several reasons, including immaturity of the organs required for assimilation of nutrients, low stores of most nutrients in preterm compared to term infants, and an increased ten
dency for systemic infections which may further alter their metabolism and assimilation of nutrients. Although the needs of energy and protein for verylow-birth-weight (VLBW) infants (< 1500 g) are being investigated by a number of laboratories, few studies have been performed to evaluate vitamin needs of this group of high risk infants. Our laboratory has developed
Address reprint requests to Harry L. Greene. M.D., Vanderbilt University Medical Center. Department of Pediatrics. Division of Nutrition, D-4130 Medical Center North, Nashville. Tennessee 37232-2576.
Journal of the American College of Nutrition, Vol. 10. No. 4. 281-288 ( 1991 ) © 1991 John Wiley & Sons, Inc.
CCC 0731 -5724/91/040281-08$04.00
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins in Premature Infants methods for vitamin analysis using the small quantities of blood available from VLBW infants. The purpose of this paper is to review our published findings evaluating two fat-soluble (A and E) and two water-soluble (B2 and B6) vitamins from VLBW infants who required total parenteral nutrition (TPN) during the first 3-4 weeks of life [1-7]. Since parenteral feedings are necessary initial ly in most VLBW infants and since an accurate measure of vitamin intake is possible with parenteral feedings, these findings should represent an initial step toward determining vitamin needs during this period of rapid growth and development. Detailed methods and results on individual vitamins have been published in several independent papers and represent the work of a number of collaborators who are authors of individual manuscripts [1-7]. Pertinent methods and results have been extracted from these papers to provide a cohesive discussion of vitamins in VLBW infants through the illustration of findings with these four vitamins.
METHODS All studies were approved by the Committee for the Protection of Human Subjects, and informed consent was given by the subject or guardian. Patients Studies were begun at the time parenteral feedings were initiated, which was always within the fist 6 days of life, and no infants were given oral supplemental vitamins. Vitamins were added to the TPN solutions within 2 hours of initiating the infusion. With the excep tion of vitamin A, no substantial differences in vitamin content were present between the beginning and end of the continuous infusion of TPN containing vitamins. Forty-two infants comprised the study population. Birth weights were all < 1530 g, except three whose weights were between 1500 and 1750 g; all were appropriate for gestational age. Study infants received energy intakes between 56 and 78 kcal/kg/day during the first week and progressed to 63-96 kcal/kg/day thereafter. Infants who had vitamins added to the lipid infusion received Intralipid (IL) (KabiVitrum, Stockholm) initially at 0.5 g/kg/day on days 3-6, which was increased progressive ly to 2.5 g/kg/day by days 6-8 after birth before increas ing to a maximum dose of 3 g/kg/day. All vitamins were given in the IL; however, since only vitamin A delivery was significantly affected by this modification, data from all studies of vitamins E, B,, and B6 have been grouped together. All TPN solutions contained a trace element
282
mixture which met the American Medical Association guidelines. Mineral intakes were adjusted according to the needs of the individual infants.
Target Blood Levels Since the usual "normative data" from orally fed in fants was not available, venous blood from the placenta was obtained at the time of delivery or blood from the infants obtained within 48 hours from both term infants as well as VLBW infants. Although much of the data on target levels has been reported as "normative" data in prior publications, the numbers of measurements are in creased and data from the larger sample size will be presented. This blood was assayed for each of the vitamins and used as a guide to approximate target ran ges for vitamins B, and B6. Since the lipid-soluble vitamins A and E are handled differently in vivo than vitamins B, and B6, and since tissue levels of vitamins A and E are considered to be suboptimal in prêterai infants, plasma levels in growing healthy term infants between 1 and 3 months of age were used as a target range for vitamin A. For vitamin E, blood levels suggested by the American Academy of Pediatrics (AAP) Committee on Nutrition were used as the target range [8]. Vitamin Formulations Two vitamin formulations were used for the studies: MVI Pediatrie (Rorer Pharmaceuticals, Kankakee, IL) was used for most studies and mixed with the glucose/amino acid mixture just prior to infusion. Since there is generally a 10-20% overage of vitamins in the bottle compared to that listed on the label, vitamin con tent was measured in random TPN solutions. The amount stated to be given was calculated from these vitamin measurements. In one study, MVI Pediatrie was added directly to IL using the IL as a diluent for the lyophilized MVI Pediatrie [6]. In another study, 40% of the vial of Berrocca PN, the vial containing A, E, D (LyphoMed, Hoffman-LaRoche, NJ) was given as the source of lipid-soluble vitamins to study the ester, retinyl palmitate. The remainder of the vitamins was MVI Pediatrie given at 40% of the vial/kg body weight. Vitamin Measurements Erythrocyte glutathione reductase (EGR) from vitamin B·, and erythrocyte glutamic-oxaloacetic transaminase (EGOT) for vitamin B6 were measured as described previously [1]. All determinations were carried out at the Vanderbilt University Clinical Nutrition Re search Unit. At the time of the assay, stored samples (-70"C) were thawed in the dark and 0.2-0.4 ml aliquots
VOL. 10, NO. 4
Vitamins in Premature Infants were processed for analysis. Retinol (R), retinyl palmitate (RP), and oc-tocopherol measurements were made by high-performance liquid chromatography (HPLC) with retinyl acetate and D-a-tocopherol acetate as inter nal standards [2-4]. Riboflavin and pyridoxine measurements in plasma and urine were performed by HPLC using reverse-phase chromatography as described previously using the methods of Pietà and Sampson [6,9,10]. Statistical analyses used repeated-measures analysis of variance with the Newman-Keuls post hoc test. Level of sig nificance was p < 0.05. Results are given as mean ± SEM.
RESULTS
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Target Ranges Vitamin A. Thirty-two infants weighing < 1500 g had cord blood assayed for R concentration. Of the 32 in fants, 16 were < 1000 g, with a mean weight of 858 ± 36 g, and the remaining infants weighed 1308 ± 92 g. There was no significant difference in cord blood R levels be tween the two groups (16.8 + 1.2 vs 17.4 ± 0.9 μ§Μ1). Twenty-eight term infants who were consuming either breast milk or formula milk between 1 and 3 months of age showed blood retinol levels of 25.4 ± 5.2 μg/dl. This level was significantly higher than that present in the preterm infants. Vitamin E. The same infants and children used for vitamin A target ranges also had vitamin E (αtocopherol) concentrations measured. Plasma vitamin E levels were not significantly different in the VLBW in fants weighing > 1000 g compared to those weighing < 1000 g and were 0.26 ± 0.06 mg/dl. Term infants showed a concentration of 0.37 ± 0.09 mg/dl. This was significantly higher than in the preterm infants (p < 0.01). Vitamin B2. The erythrocyte glutathione reductase measurements performed in 16 VLBW infants showed an activity coefficient of 1.03 + 0.023, consistent with that present in term infants as well as their mothers and other normal adults. Vitamin B, (riboflavin) concentra tions in cord blood plasma from 14 VLBW infants were 64.1 ± 6 ng/ml and were similar to the cord blood con centrations in 10 term infants of 61.4 ± 1.0 ng/ml. Erythrocyte riboflavin levels were expressed per gram of hemoglobin and showed a concentration of 71.9 ± 14.0 ng/g hemoglobin. Maternal erythrocyte riboflavin was 67.3 ± 13.0 ng/g hemoglobin and, in contrast to plasma riboflavin, did not differ significantly from that seen in infants. The initial infant erythrocyte flavin
mononucleotide (FMN) and flavin adenine dinucleotide (FAD) values were 89.8 + 20.7 and 1342 ± 133.0 ng/g hemoglobin, respectively. The maternal FAD value was 1400.0 ± 123.0 ng/g hemoglobin. Maternal erythrocyte FMN was not measured. The levels in 38 infants who were formula fed in creased to 74 + 17 and 71 ± 12 ng/ml by 7 and 14 days, then declined progressively to 44.7 + 7 and 38.4 + 6.9 ng/ml by 3 and 4 months of age. However, 20 agematched growing infants who were exclusively breast fed showed a rapid decline to 28 + 9 ng/ml by 1 week, and this level was maintained by most infants for 4 months (21.4 ± 6.7 ng/ml). Three of the infants had levels below 15 ng/ml [7]. Vitamin B6. EGOT measurements in 16 infants were similar in all infants as well as their mothers and are grouped together as a whole. The glutamic-oxaloacetic transaminase activity coefficient was 1.17 ± 0.05. Vitamin B6 cord blood measurements were performed in the VLBW infants, and there was no statistical difference between the 16 weighing < 1000 g, the 12 weighing 1000-1500 g, and 18 term infants. The groups were therefore combined to give mean cord levels of pyridoxal phosphate (PLP) of 93 ± 13 ng/ml and of pyridoxal 38 ± 14 ng/ml; pyridoxine levels were undetectable. The mothers' levels of PLP for preterm (n = 18) vs term (n = 21 ) infants were also not different and together were 22.1+8.1 ng/ml. Blood Levels During TPN Vitamin A. In the original studies of vitamin A in infants receiving TPN [2], a progressive decline in plas ma R concentration was noted during the month of TPN despite the daily addition of R to TPN at 400-500 μg; these intakes were consistent with dietary recommenda tions. The reported losses of vitamin A on the tubing as well as some degree of light degradation during the period of infusion was such that only an estimated 15% of the prescribed dose was actually received by the in fants. In a second study, infants were divided into two groups < 1000 g (n = 24) and 1000-1500 g (n = 17). Blood levels in both groups prior to initiating TPN (age 1-4 days) were similar (14.8 ± 0.9 and 13.5 ± 0.7 μg/dl, respectively), but the decline in vitamin A levels after 3 weeks of TPN was greater in the smaller infants (< 1000 g). The combined groups showed a progressive decline in R by the second week of life to 10.5 ± 1.7 μg/dl (Fig. 1 ). These concentrations persisted during the period of total parenteral nutrition of 28 days. These in fants were followed after being weaned completely to oral feedings. After 1 week of oral feedings, infants < 1000 g increased their levels to 13.4 ± 2 μ^αΐ; their
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
283
Vitamins in Premature Infants
ï o
s
Downloaded by [ECU Libraries] at 00:12 23 April 2015
• pH -t->
Days Fig, 1. Comparison of blood plasma levels of patients receiving vitamins (MVI Pediatrie) as 420 μg/day of retinol added either to the glucose/am ino acid mixture (G + AA) or Intralipid (IL). The dose in Inlralipid was 280 μg/day and the dose in the glucose/amino acid mixture was 460 μg/clay.
60 S
50
tao
40
20
>
RBP (mg/dl) day0-1.4±.4 day28-l.e+.4
10
—r-
5
10
—i—
—i—
—i—
15
20
25
30
Days Fig. 2. Blood plasma levels of rctinol (ROL) and relinyl palmitatc (RP) of five infants who received 500 μg/kg/day of vitamin A as relinyl palmitatc.
levels did not increase further during the subsequent 3 weeks of feeding. In the second group of infants > 1000 g, there was a significant increase in serum R levels by the fourth week of oral feeding, although several infants in both groups had levels < 10 μg/dl, a
284
level associated with signs of R deficiency in older children [2]. Because of the large losses of vitamin A on the in travenous tubing, a third study was performed [5] with vitamin A concentrations measured in infants who were
VOL. 10, NO. 4
Vitamins in Premature Infants
2000τ
τ100 3 A
ί
,2 o
1500--
75
1000--
50
?
3
σ E
m _σ CL
ta
500-
-■ 25
3
CD Q
3] 3* CD
12
16
20
24
28
32
36
40
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Age (days)
Fig. 3. General trend of plasma riboflavin concentrations and urinary excretion in an infant (BW 1210 g) who received 0.66 mg/kg/day of riboflavin during TPN for 38 days. Other patients showed similar trends with a marked increase in plasma content during the first 8-12 days, a progressive decline over the next 2-3 weeks, and a progressive increase in urinary excretion after thefirstweek.
receiving the vitamin mixture in a lipid emulsion (Intralipid). In this study, infants received 280 μg/kg/day of R in the lipid emulsion. The plasma vitamin A concentra tions increased significantly (Fig. 1) from a mean level of 11 ± 0.7 before IL was infused, to a level of 19.2 ± 0.97 μ ^ Ι after the vitamin had been infused for 1 month (p < 0.01). Since all infants do not receive IL in the first week of life, there was a need to develop a solution which could have the vitamin A added to the glucose/amino acid mix ture. For this reason a fourth study was performed [10]. In this study, intravenous RP was given at a dose of 400 μg/kg/day and R was given at a dose of 80 μg/kg/day. Results of this study indicated that total vitamin A con centrations increased substantially from the initial level of 18.4 ± 7.8 μς/άί to a total concentration of 69.7 μ^αΐ. However, the amount of R was maintained at basically the same level as had been seen in the previous study using R with IL: the total increase resulted primarily from the RP (Fig. 2). Vitamin E. In the initial study of vitamin E, pretreat ment levels of oc-tocopherol were 0.31 ± 0.03 mg/dl. After 3 weeks of treatment with 4.6 mg/kg/day, the levels increased to 3.08 ± 0.27 mg/dl. It was noted, how ever, that several infants received oral supplements of vitamin E (80-100 mg/day); these patients had levels much higher than those receiving no oral supplement (3.4 ± 0.6 vs 2.0 ± 0.2 mg/dl) [2]. Subsequent studies indicated that doses between 2.8 and 3.5 mg/kg/day were
sufficient to maintain blood levels at 1.0-2.5 mg/dl [6], Vitamin B,. Thirteen infants had EGR measurements performed during 28 days of TPN. The activity coeffi cient was within the normal reference range (< 1.2) for all infants (0.92 ± 0.02 to 0.89 ± 0.01) during the 28 days. Interpretation of these results was difficult since all infants had received blood transfusions. Thus, assay of B2 levels was performed. Blood and urine riboflavin concentrations during TPN showed a large degree of variation between patients. During the first few days of parenteral feedings, when 0.23 mg/kg/day of riboflavin was given, three patients had levels of 222, 300, and 392 ng/ml, respectively, and a dose of 0.41 mg/day for 2 days gave levels of 681, 652, and 763 ng/ml. A higher dose of 0.66 mg/kg/day resulted in a marked, although variable, increase in plasma riboflavin concentrations in plasma in all patients [6]. During the period of 21 days, the individual levels dif fered substantially, although a common pattern of plasma and urinary excretion was seen (Fig. 3). Following the initial marked increase in blood levels during the first 1-2 weeks, there was a progressive decline in levels. Coincident with the decline in plasma levels, urinary ex cretion increased markedly by the second week of life. This increased excretion was greater than the TPN dose and thus accounts for the decline in blood levels. The striking finding in this study was the remarkably elevated plasma riboflavin levels. There was concern in itially that these extremely high levels of riboflavin
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
285
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins
in Premature
Infants
might interfere with the conversion of riboflavin to its active coenzyme FMN and FAD. This did not appear to be the case, however, since the FAD levels prior to treat ment in red cells was 972 ± 1 2 1 and, after 28 days of treatment, it had increased to a level of 2005 ± 294 ng/g hemoglobin. Using the data from the three doses of vitamin B,, an estimate of the dose which would result in blood levels close to the target range was 0.15 mg/kg body weight. It is suggested that intakes of this range be tested in future studies. Vitamin B6. The recommended dose of vitamin B6 in current practices of TPN is 0.6 mg/day. Using this dose, the EGOT activity coefficient in 13 infants during 28 days of TPN remained normal (< 1.84). This dose per kg of body weight for VLBW infants was approximately three times the recommended dietary allowance of 0.4 mg/day for term infants (approximately 0.05-0.1 mg/kg/day). We postulated, therefore, that the dose given in TPN to VLBW infants would result in elevated blood levels similar to those seen with vitamin B2. After receiv ing the dose of 0.6 mg/day (0.4-0.8 mg/kg/day) in TPN for 2 weeks, the PLP level in seven patients was 87 ±11 nmol/L, similar to the cord blood level. Pyridoxic acid, however, had increased from 69 ± 5 to 989 ± 328 nmol/L, and the pyridoxal level increased from 23 ± 3 to 708 ± 140 nmol/L with a pyridoxine level of 251 ± 38 nmol/L. These indicate that the dose of pyridoxine which is presently given to VLBW infants results in extremely high levels in the blood. The infants also appear to be able to convert the intravenously administered pyridoxine to pyridoxal and its active cofactor, PLP. A lower dose of 0.06 and 0.2 mg/kg/day was given to three patients for the first 48 hours; then in the subsequent 48 hours the mean pyridoxal levels were 210 and 418 nmol/L. Using these dose responses, an estimated dose which would result in blood pyridoxal levels closer to the target range is 0.18 mg/kg/day. Future studies should evaluate this dosage.
DISCUSSION The preceding results suggest that the amounts of vitamins recommended in the package insert of MVI Pediatrie for prêterai infants caused a substantial devia tion from the target blood levels of each of the four vitamins tested. Vitamin A levels tended to be low, whereas vitamins E, B2, and B 6 tended to be high. Al though the target ranges may not necessarily represent optimal levels for VLBW infants, they do represent the expected levels consistent with normal growing neonates
286
or approximations of the levels expected during in trauterine growth. These levels therefore appear to be reasonable to use as guides for approximating a safe in take for intravenous vitamins [11]. Since the results indi cate a need to discuss the findings with regard to each vitamin, this will be provided below. Vitamin A. Vitamin A is present in the diet primarily as its ester, retinyl palmitate. This compound is first hydrolyzed then reesterified to RP in the intestine, secreted with the chylomicron fraction into lymphatics and circulated to the liver. The liver then hydrolyzes the RP to R and either stores it as RP in stellate cells or secretes the R into plasma complexed with R—binding protein and prealbumin. To accurately interpret the blood measurements of R and RP, the various stages of absorp tion, transport to the liver, hepatic uptake, and secretion into plasma must be taken into account. In this regard, a number of conditions such as infection or liver disease may alter one or more of the processes involved in liver or tissue uptake or release of vitamin A into plasma. These problems preclude the use of plasma vitamin A as a measure of vitamin A stores. This problem of inter pretation is probably magnified in VLBW infants be cause of immaturity, although the ontogeny of hepatic vitamin A uptake and secretion has not been inves tigated. Given the above limitations in interpretation of blood levels in response to intravenous vitamin A, there is an additional complication: substantial amounts of R added to parenteral solutions are lost in the delivery sets. Thus, the fact that blood levels do not reflect tissue stores of vitamin A in humans is even further compli cated in VLBW infants for two reasons: (1) the dosage received from TPN solutions is not known, and (2) the ontogeny of hepatic assimilation of vitamin A is not known. The first problem may be corrected by mixing the R with IL or using RP. These two options provide a basis for investigation of the question related to on togeny. Despite the apparent limitations in the interpreta tion of blood vitamin A concentrations, our data and that reported from other laboratories suggest that at birth most VLBW infants have insufficient vitamin A stores to meet their needs and that the dose recommended by the package insert for MVI Pediatrie should be increased. The American Society for Clinical Nutrition has publish ed a newer set of recommendations [12], suggesting that the intake of R in low-birth-weight infants should be approximately 500 μg/kg/day and that efforts to prevent losses during infusion should be instituted. It is hoped that this dose will maintain an adequate supply of R for the infant's needs during this critical period of develop ment. This recommendation is based on the above dis cussion of vitamin A as well as data which have been obtained by others [13-16].
VOL. 10, NO. 4
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins in Premature Infants Vitamin E. Tocopherol exists as several isomers, the most active being α-tocopherol. Intravenous lipid preparations contain enough tocopherols to affect local blood tocopherol levels. The majority of tocopherol in soybean oil emulsions is γ-tocopherol, which has sub stantially less biologic activity than cc-tocopherol, present in safflower oil emulsions. Thus, blood tocopherol measurement by HPLC is preferred over the total tocopherol level given by the calorimetrie assay because HPLC separates tocopherol into its various isomers. Needs of vitamin E are usually related directly to the level of dietary polyunsaturated fatty acids. Studies using mathematical or statistical modeling techniques, which compare a sensitive assay for deficiency (red blood cell peroxide hemolysis) with tocopherol levels in children, suggest that tocopherol levels < 0.44 mg/dl are as sociated with increased hemolysis. This level compares favorably with the conventionally accepted level of 0.5 mg/dl as the discriminator of vitamin E adequacy in adults. The influence of circulating lipids on tocopherol levels makes it preferable to express plasma cctocopherol levels as a ratio to ß-lipoprotein, cholesterol, or total lipids. However, in most instances newborn or premature infants show no abnormalities in circulating lipid levels that would lead to inappropriate interpreta tion of vitamin E, and levels are usually expressed in terms of plasma concentration. Neither the intravenous dose nor the optimal blood level of tocopherol in preterm infants has been defined. Because there is no proven benefit for pharmacologie doses of vitamin E and because several premature infants receiving large intravenous doses (25-100 mg/day) developed liver failure, the AAP Committee on Nutrition has suggested that safe and effective target blood levels are between 1 and 2 mg/dl [8]. Our findings using vitamin E at 2.1 and 4.6 mg/day in infants < 1500 g suggested that blood levels vary directly with the dose of vitamin E per kilogram of body weight. From these data, a dose response indicates that a daily dose of between 2.8 and 3.5 mg cc-tocopherol acetate/kg should maintain blood levels within the target range [16]. Vitamin B2. Riboflavin is phosphorylated to riboflavin-5-phosphate (FMN). As part of two coenzymes, FMN and FAD, riboflavin serves in cellular me tabolism for hydrogen transfer. When given in excess of needs, the vitamin is excreted primarily unchanged in the urine. Riboflavin is inactivated by light, especially that emitted from phototherapy lights, but FMN is more stable. For this reason, FMN is the form used for TPN. Using the measurement of erythrocyte glutathione reductase activity coefficient as a marker in breast-fed infants
receiving phototherapy shows a higher incidence of erythrocyte FAD depletion [17]. Despite the high rate of photodegradation expected from TPN solutions normally exposed to light during delivery, deficiency of riboflavin has not been described in children maintained on TPN. Our assessment of riboflavin status using activity of erythrocyte glutathione reductase before and after the addition of FAD indicated that there was not a deficiency of riboflavin. Since the enzyme assay will not detect marginal levels or high levels of riboflavin, methods to quantitate serum ribo flavin level as well as red blood cell riboflavin, FMN, and FAD content were developed. Our data indicate that a parenteral riboflavin intake of 0.66 mg/kg/day results in a 20- to 400-fold elevation in blood riboflavin levels in VLBW infants, and lower doses provided some estimate of a dose response [6]. Based on these estimates of riboflavin in blood, a dose of 0.15 mg/kg/day has been recommended [12]. Vitamin B6. Vitamin B6 is a generic descriptor for the three naturally occurring pyridines — pyridoxine, pyridoxal, and pyridoxamine — and their phosphorylated coenzyme forms of the vitamin — pyridoxine phosphate, PLP, and pyridoxamine phosphate, respectively. PLP is the primary cofactor form of the vitamin which functions in a wide range of metabolic reactions, primarily related to nitrogen metabolism. Although no toxicity of the vitamin has been ob served with medically indicated doses of vitamin B6, a sensory neuropathy in adults ingesting megadoses of pyridoxine has been described. The functional assay for vitamin B6 status uses an in vitro assay of erythrocyte aspartate aminotransferase (glutamic-oxaloacetic transaminase) before and after the addition of PLP. Using this measure of vitamin B6 status following doses of 0.3-0.7 mg/kg/day, the glutamic-oxaloacetic transferase activity coefficient was normal. This assay does not, however, measure blood concentrations of the vitamin. Assays of blood levels of pyridoxine, pyridoxal, and PLP, as well as urinary losses during TPN, should provide a better measure of intravenous needs. When 0.4 mg/kg/day was given, pyridoxine levels increased by > 10-fold over cord blood and maternal levels [5]. Lower doses of the vitamin resulted in data which allowed an estimate of a dose response, suggesting that 0.18 mg/kg should result in blood levels closer to the target range. This dose has been recommended for future formulations to be given to VLBW infants. In summary, these data from VLBW infants indicate that the currently formulated vitamin preparations does not maintain blood levels of all vitamins within an ac ceptable range. Our findings indicate that new formula tion should be made specifically for high-risk preterm
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
287
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins in Premature Infants infants. The following steps are recommended for the development of a more appropriate mixture of parenteral vitamins for preterm infants. (1) Separate mixtures of the water-soluble vitamins and the lipid-soluble vitamins should be provided. Because of the possibility of im paired metabolism of polysorbate and propylene glycol in preterm infants, efforts should be made to provide lipid-soluble vitamins without the addition of emulsifiers such as polysorbate. If a lipid emulsifier is essential, it should be a chemical that has been established to be nontoxic when administered intravenously to preterm in fants. (2) The water-soluble vitamin mixture should be tested in vitro to determine the most appropriate method of delivery to minimize photodegradation in the usual setting of an intensive care neonatal nursery when highintensity lighting is present on a continuous basis. (3) Both the water-soluble and the lipid-soluble vitamins should be evaluated in vitro to insure stability and bioavailability in TPN solutions under the conditions of an intensive care nursery. (4) In vivo studies of a few preterm infants who receive the new recommended doses should be evaluated prior to marketing a new preparation.
REFERENCES 1. Moore MC, Greene HL, Phillips B, Shulman RJ, Murrell JE, Ament ME: Evaluation of a pediatrie multiple vitamin preparation of total parenteral nutrition in infants and children, I. Blood levels of water-soluble vitamins. Pediatrics 77:530-538, 1986. 2. Greene HL, Moore MC, Phillips B, Franck L, Shulman RJ, Ament ME, Murrell JE, Chan MM, Said HM: Evalua tion of a pediatrie multiple vitamin preparation for total parenteral nutrition, II. Blood levels of vitamins A, D, and E. Pediatrics 77:539-547, 1986. 3. Greene HL, Phillips BL, Franck L, Fillmore CM, Said HM, Murrell JE, Courtney-Moore ME, Briggs R: Persist ently low blood retinol levels during and after parenteral feeding of very low birth weight infants: examination of losses into intravenous administration sets and a method of prevention by addition to a lipid emulsion. Pediatrics 79:892-900, 1987. 4. Phillips B, Franck LS, Greene HL: Vitamin E levels in premature infants during and after intravenous multivitamin supplementation. Pediatrics 80:680-683, 1987.
288
5. Greene HL. Baeckert PA, Murrell J, Oelburg DG, Adcock EW III: Blood pyridoxinc levels in preterm infants receiv ing TPN. Pediatr Res 25(4): 113A, 1990. 6. Baeckert PA, Greene HL, Oelberg DG, Adcock EW III: Blood levels of vitamins A, D, E, and riboflavin in very low birth weight infants receiving parenteral nutrition with vitamins added lo the lipid emulsion. J Pediatr 113:10571065, 1988. 7. Greene HL, Specker BL, Smith R, Murrell J, Swift LL: Plasma riboflavin concentrations in infants fed human milk versus l'onnula. J Pediatr 117(6):916-920, 1990. 8. Poland RL: Vitamin E: what should we do? Pediatrics 77:787-788. 1986. 9. Pietta P, Calatroni A, Rava A: Hydrolysis of riboflavin nucleoiides in plasma monitored by high performance liq uid chonnaiography. J Chromatogr 229:445-449, 1982. 10. Sampson D, O'Connor DK: Analysis of B 6 vitamers and pyridoxic acid in plasma, tissues, and urine using high pcrfomiance liquid chormatography. Nutr Res 9:259-272, 1989. 11. Pollack P. Murrell J, Caudill M, Swift L, Adcock E, Greene HL: Vilamin A in parenteral feedings for very low birth weight infants. Pediatr Res 27-.289A, 1990. 12. Greene HL, Hambidge KM, Schanler R, Tsang RC: Guidelines for the use of vitamins, trace elements, cal cium, magnesium, and phosphorus in infants and children receiving total parenteral nutrition: Report of the Subcom mittee on Pediatrie Parenteral Nutrient Requirements from the Committee on Clinical Practice Issues of the American Society for Clinical Nutrition. Am J Clin Nutr 48:13241342, 1986. 13. Shcnai JP, Chylil F, Jhaveri A, Stahlman MT: Plasma vitamin A and retinol binding protein in premature and terni neonates. J Pedialr 99:302-305, 1981. 14. Shenai JP, Kennedy KA, Chytil F, Stahlman MT: Clinical trial of vitamin A supplementation in infants susceptible to bronchopulmonary dysplasia. J Pediatr 111:269-277, 1987. 15. Brandt RB, Mueller DG, Shroeder JR, Guyer KE, Kirkpatrick BV, Hutcher NE, Ehrlich FE: Serum vitamin A in premature and lenn neonatcs. J Pediatr 92:101-105, 1978. 16. Husicatl VA, Gutchcr GR, Anderson SA, Zachman RD: Relationship of vilamin A (retinol) status to lung disease in Ihc prcicmi infant. J Pediatr 105:610-615, 1984. 17. Hovi L, Hckali Rand Sûmes MA: Evidence of riboflavin depletion in breasi-lcd newborns and its further accelera tion during treatment with phototherapy. Acta Pediatr Scand 68:567-570, 1979. Received August 1990; revision acceptedDecember
1990.
VOL. 10, NO. 4
Journal of the American College of Nutrition Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uacn20
Intravenous vitamins for very-low-birth-weight infants. a
a
a
a
a
a
H L Greene , R Smith , P Pollack , J Murrell , M Caudill & L Swift a
Vanderbilt University School of Medicine, Department of Pediatric Nutrition, Nashville, Tennessee 37232-2576. Published online: 02 Sep 2013.
To cite this article: H L Greene, R Smith, P Pollack, J Murrell, M Caudill & L Swift (1991) Intravenous vitamins for very-lowbirth-weight infants., Journal of the American College of Nutrition, 10:4, 281-288, DOI: 10.1080/07315724.1991.10718154 To link to this article: http://dx.doi.org/10.1080/07315724.1991.10718154
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Intravenous Vitamins for Very-Low-Birth-Weight Infants Harry L. Greene, MD, Rita Smith, BS, Paul Pollack, MD, Joel Murrell, MT, Melinda Caudill, MT, and Larry Swift, PhD Vanderbilt University School of Medicine, Department of Pediatrie Nutrition, Nashville
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Key words: very-low-birth-weight infants, vitamins, vitamin A, vitamin E, vitamin B:, vitamin B6, total parenteral nutrition Term infants and children appear to adapt to large variations in vitamin intakes. This is supported by thefindingof similar blood levels of vitamins despite several-fold differences in intake on a body weight basis. By contrast, the finding of marked elevation of some vitamins and low levels of others seen in very-low-birth-weight ( VLBW) infants (< 1500 g) suggest that this group has less adaptive capacity to high- or low-dose intakes. This indicates that their vitamin intakes need to be more closely aligned with actual needs. This paper reviews previously published data on vitamins A, E, B2, and B6 from VLBW infants receiving total parenteral nutrition (TPN). Vitamin A. VLBW infants are relatively deficient in retino! (R) at birth. During TPN large losses of R onto the delivery sets result in a further decline in stores of R as reflected in a progressive decline in plasma R during TPN. Because of the reported lower incidence of bronchopulmonary dysplasia associated with intramuscular vitamin A treatment, alternative methods of vitamin A delivery during TPN have been suggested. First, the vitamins were mixed with Intralipid (IL) and, second, retinyl palmitate (RP) rather than R was used. There was little in vitro loss of R when mixed with IL, and in vivo treatment resulted in higher blood levels after 1 month of rctinol administration in IL than seen previously (21.4 ± 4.2 vs 14.1 ± 3.7 μg/dl). Use of RP in VLBW infants resulted in high RP levels (40 ± 6 μg/dl), although R levels were similar to that seen with R added to IL (21.1 ±4 μg/dl). Using these data and those from other publications, currently suggested intravenous intake of vitamin A as R is 500 μg/kg/day. Vitamin E. The TPN solution for pediatrie patients contains oc-tocopherol acetate. Little of the vitamin is lost to the plastic infusion sets. Infusion of four different dosage levels suggests that doses of 2.8-3.5 mg/kg/day will maintain most infant blood levels between 1 and 2 mg/dl. Vitamin B2. Vitamin B2 is activated to its active cofactor forms flavin mononucleolidc (FMN) and flavin adenine dinucleotide (FAD). Three doses of riboflavin (0.68, 0.56, and 0.34 mg/kg/day) resulted in elevated blood levels. Using these blood response doses, a projected intake of 0.15 mg/kg/day appears more appropriate to maintain blood levels in the range of those seen in formula-fed term infants. Vitamin Bft. Vitamin B(1 is converted in vivo to pyridoxal and activated to its cofactor form, pyridoxal phosphate (PLP). Three doses of pyridoxine (from 0.2 to 0.5 mg/kg/day) resulted in elevated blood levels. Using the blood response to these doses, an intake of 0.18 mg/kg/day is projected to maintain PLP levels within the range of that seen in plasma samples from formula-fed term infants. Abbreviations: AA = amino acid, AAP = American Academy of Pediatrics, BW = birth weight, EGOT = erythrocyte glutamic-oxaloacetic transaminase, EGR = erythrocyte glutathione reductase, FAD = flavin adenine dinucleotide, FMN = flavin mononucleotide, HPLC = high-performance liquid chromatography, IL = Intralipid, PLP = pyridoxal phosphate, R = retinol, RP = retinyl palmitate, TPN = total parenteral nutrition, VLBW = very low birth weight
INTRODUCTION Nutritional support of preterm infants requires special considerations compared to term infants for several reasons, including immaturity of the organs required for assimilation of nutrients, low stores of most nutrients in preterm compared to term infants, and an increased ten
dency for systemic infections which may further alter their metabolism and assimilation of nutrients. Although the needs of energy and protein for verylow-birth-weight (VLBW) infants (< 1500 g) are being investigated by a number of laboratories, few studies have been performed to evaluate vitamin needs of this group of high risk infants. Our laboratory has developed
Address reprint requests to Harry L. Greene. M.D., Vanderbilt University Medical Center. Department of Pediatrics. Division of Nutrition, D-4130 Medical Center North, Nashville. Tennessee 37232-2576.
Journal of the American College of Nutrition, Vol. 10. No. 4. 281-288 ( 1991 ) © 1991 John Wiley & Sons, Inc.
CCC 0731 -5724/91/040281-08$04.00
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins in Premature Infants methods for vitamin analysis using the small quantities of blood available from VLBW infants. The purpose of this paper is to review our published findings evaluating two fat-soluble (A and E) and two water-soluble (B2 and B6) vitamins from VLBW infants who required total parenteral nutrition (TPN) during the first 3-4 weeks of life [1-7]. Since parenteral feedings are necessary initial ly in most VLBW infants and since an accurate measure of vitamin intake is possible with parenteral feedings, these findings should represent an initial step toward determining vitamin needs during this period of rapid growth and development. Detailed methods and results on individual vitamins have been published in several independent papers and represent the work of a number of collaborators who are authors of individual manuscripts [1-7]. Pertinent methods and results have been extracted from these papers to provide a cohesive discussion of vitamins in VLBW infants through the illustration of findings with these four vitamins.
METHODS All studies were approved by the Committee for the Protection of Human Subjects, and informed consent was given by the subject or guardian. Patients Studies were begun at the time parenteral feedings were initiated, which was always within the fist 6 days of life, and no infants were given oral supplemental vitamins. Vitamins were added to the TPN solutions within 2 hours of initiating the infusion. With the excep tion of vitamin A, no substantial differences in vitamin content were present between the beginning and end of the continuous infusion of TPN containing vitamins. Forty-two infants comprised the study population. Birth weights were all < 1530 g, except three whose weights were between 1500 and 1750 g; all were appropriate for gestational age. Study infants received energy intakes between 56 and 78 kcal/kg/day during the first week and progressed to 63-96 kcal/kg/day thereafter. Infants who had vitamins added to the lipid infusion received Intralipid (IL) (KabiVitrum, Stockholm) initially at 0.5 g/kg/day on days 3-6, which was increased progressive ly to 2.5 g/kg/day by days 6-8 after birth before increas ing to a maximum dose of 3 g/kg/day. All vitamins were given in the IL; however, since only vitamin A delivery was significantly affected by this modification, data from all studies of vitamins E, B,, and B6 have been grouped together. All TPN solutions contained a trace element
282
mixture which met the American Medical Association guidelines. Mineral intakes were adjusted according to the needs of the individual infants.
Target Blood Levels Since the usual "normative data" from orally fed in fants was not available, venous blood from the placenta was obtained at the time of delivery or blood from the infants obtained within 48 hours from both term infants as well as VLBW infants. Although much of the data on target levels has been reported as "normative" data in prior publications, the numbers of measurements are in creased and data from the larger sample size will be presented. This blood was assayed for each of the vitamins and used as a guide to approximate target ran ges for vitamins B, and B6. Since the lipid-soluble vitamins A and E are handled differently in vivo than vitamins B, and B6, and since tissue levels of vitamins A and E are considered to be suboptimal in prêterai infants, plasma levels in growing healthy term infants between 1 and 3 months of age were used as a target range for vitamin A. For vitamin E, blood levels suggested by the American Academy of Pediatrics (AAP) Committee on Nutrition were used as the target range [8]. Vitamin Formulations Two vitamin formulations were used for the studies: MVI Pediatrie (Rorer Pharmaceuticals, Kankakee, IL) was used for most studies and mixed with the glucose/amino acid mixture just prior to infusion. Since there is generally a 10-20% overage of vitamins in the bottle compared to that listed on the label, vitamin con tent was measured in random TPN solutions. The amount stated to be given was calculated from these vitamin measurements. In one study, MVI Pediatrie was added directly to IL using the IL as a diluent for the lyophilized MVI Pediatrie [6]. In another study, 40% of the vial of Berrocca PN, the vial containing A, E, D (LyphoMed, Hoffman-LaRoche, NJ) was given as the source of lipid-soluble vitamins to study the ester, retinyl palmitate. The remainder of the vitamins was MVI Pediatrie given at 40% of the vial/kg body weight. Vitamin Measurements Erythrocyte glutathione reductase (EGR) from vitamin B·, and erythrocyte glutamic-oxaloacetic transaminase (EGOT) for vitamin B6 were measured as described previously [1]. All determinations were carried out at the Vanderbilt University Clinical Nutrition Re search Unit. At the time of the assay, stored samples (-70"C) were thawed in the dark and 0.2-0.4 ml aliquots
VOL. 10, NO. 4
Vitamins in Premature Infants were processed for analysis. Retinol (R), retinyl palmitate (RP), and oc-tocopherol measurements were made by high-performance liquid chromatography (HPLC) with retinyl acetate and D-a-tocopherol acetate as inter nal standards [2-4]. Riboflavin and pyridoxine measurements in plasma and urine were performed by HPLC using reverse-phase chromatography as described previously using the methods of Pietà and Sampson [6,9,10]. Statistical analyses used repeated-measures analysis of variance with the Newman-Keuls post hoc test. Level of sig nificance was p < 0.05. Results are given as mean ± SEM.
RESULTS
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Target Ranges Vitamin A. Thirty-two infants weighing < 1500 g had cord blood assayed for R concentration. Of the 32 in fants, 16 were < 1000 g, with a mean weight of 858 ± 36 g, and the remaining infants weighed 1308 ± 92 g. There was no significant difference in cord blood R levels be tween the two groups (16.8 + 1.2 vs 17.4 ± 0.9 μ§Μ1). Twenty-eight term infants who were consuming either breast milk or formula milk between 1 and 3 months of age showed blood retinol levels of 25.4 ± 5.2 μg/dl. This level was significantly higher than that present in the preterm infants. Vitamin E. The same infants and children used for vitamin A target ranges also had vitamin E (αtocopherol) concentrations measured. Plasma vitamin E levels were not significantly different in the VLBW in fants weighing > 1000 g compared to those weighing < 1000 g and were 0.26 ± 0.06 mg/dl. Term infants showed a concentration of 0.37 ± 0.09 mg/dl. This was significantly higher than in the preterm infants (p < 0.01). Vitamin B2. The erythrocyte glutathione reductase measurements performed in 16 VLBW infants showed an activity coefficient of 1.03 + 0.023, consistent with that present in term infants as well as their mothers and other normal adults. Vitamin B, (riboflavin) concentra tions in cord blood plasma from 14 VLBW infants were 64.1 ± 6 ng/ml and were similar to the cord blood con centrations in 10 term infants of 61.4 ± 1.0 ng/ml. Erythrocyte riboflavin levels were expressed per gram of hemoglobin and showed a concentration of 71.9 ± 14.0 ng/g hemoglobin. Maternal erythrocyte riboflavin was 67.3 ± 13.0 ng/g hemoglobin and, in contrast to plasma riboflavin, did not differ significantly from that seen in infants. The initial infant erythrocyte flavin
mononucleotide (FMN) and flavin adenine dinucleotide (FAD) values were 89.8 + 20.7 and 1342 ± 133.0 ng/g hemoglobin, respectively. The maternal FAD value was 1400.0 ± 123.0 ng/g hemoglobin. Maternal erythrocyte FMN was not measured. The levels in 38 infants who were formula fed in creased to 74 + 17 and 71 ± 12 ng/ml by 7 and 14 days, then declined progressively to 44.7 + 7 and 38.4 + 6.9 ng/ml by 3 and 4 months of age. However, 20 agematched growing infants who were exclusively breast fed showed a rapid decline to 28 + 9 ng/ml by 1 week, and this level was maintained by most infants for 4 months (21.4 ± 6.7 ng/ml). Three of the infants had levels below 15 ng/ml [7]. Vitamin B6. EGOT measurements in 16 infants were similar in all infants as well as their mothers and are grouped together as a whole. The glutamic-oxaloacetic transaminase activity coefficient was 1.17 ± 0.05. Vitamin B6 cord blood measurements were performed in the VLBW infants, and there was no statistical difference between the 16 weighing < 1000 g, the 12 weighing 1000-1500 g, and 18 term infants. The groups were therefore combined to give mean cord levels of pyridoxal phosphate (PLP) of 93 ± 13 ng/ml and of pyridoxal 38 ± 14 ng/ml; pyridoxine levels were undetectable. The mothers' levels of PLP for preterm (n = 18) vs term (n = 21 ) infants were also not different and together were 22.1+8.1 ng/ml. Blood Levels During TPN Vitamin A. In the original studies of vitamin A in infants receiving TPN [2], a progressive decline in plas ma R concentration was noted during the month of TPN despite the daily addition of R to TPN at 400-500 μg; these intakes were consistent with dietary recommenda tions. The reported losses of vitamin A on the tubing as well as some degree of light degradation during the period of infusion was such that only an estimated 15% of the prescribed dose was actually received by the in fants. In a second study, infants were divided into two groups < 1000 g (n = 24) and 1000-1500 g (n = 17). Blood levels in both groups prior to initiating TPN (age 1-4 days) were similar (14.8 ± 0.9 and 13.5 ± 0.7 μg/dl, respectively), but the decline in vitamin A levels after 3 weeks of TPN was greater in the smaller infants (< 1000 g). The combined groups showed a progressive decline in R by the second week of life to 10.5 ± 1.7 μg/dl (Fig. 1 ). These concentrations persisted during the period of total parenteral nutrition of 28 days. These in fants were followed after being weaned completely to oral feedings. After 1 week of oral feedings, infants < 1000 g increased their levels to 13.4 ± 2 μ^αΐ; their
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
283
Vitamins in Premature Infants
ï o
s
Downloaded by [ECU Libraries] at 00:12 23 April 2015
• pH -t->
Days Fig, 1. Comparison of blood plasma levels of patients receiving vitamins (MVI Pediatrie) as 420 μg/day of retinol added either to the glucose/am ino acid mixture (G + AA) or Intralipid (IL). The dose in Inlralipid was 280 μg/day and the dose in the glucose/amino acid mixture was 460 μg/clay.
60 S
50
tao
40
20
>
RBP (mg/dl) day0-1.4±.4 day28-l.e+.4
10
—r-
5
10
—i—
—i—
—i—
15
20
25
30
Days Fig. 2. Blood plasma levels of rctinol (ROL) and relinyl palmitatc (RP) of five infants who received 500 μg/kg/day of vitamin A as relinyl palmitatc.
levels did not increase further during the subsequent 3 weeks of feeding. In the second group of infants > 1000 g, there was a significant increase in serum R levels by the fourth week of oral feeding, although several infants in both groups had levels < 10 μg/dl, a
284
level associated with signs of R deficiency in older children [2]. Because of the large losses of vitamin A on the in travenous tubing, a third study was performed [5] with vitamin A concentrations measured in infants who were
VOL. 10, NO. 4
Vitamins in Premature Infants
2000τ
τ100 3 A
ί
,2 o
1500--
75
1000--
50
?
3
σ E
m _σ CL
ta
500-
-■ 25
3
CD Q
3] 3* CD
12
16
20
24
28
32
36
40
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Age (days)
Fig. 3. General trend of plasma riboflavin concentrations and urinary excretion in an infant (BW 1210 g) who received 0.66 mg/kg/day of riboflavin during TPN for 38 days. Other patients showed similar trends with a marked increase in plasma content during the first 8-12 days, a progressive decline over the next 2-3 weeks, and a progressive increase in urinary excretion after thefirstweek.
receiving the vitamin mixture in a lipid emulsion (Intralipid). In this study, infants received 280 μg/kg/day of R in the lipid emulsion. The plasma vitamin A concentra tions increased significantly (Fig. 1) from a mean level of 11 ± 0.7 before IL was infused, to a level of 19.2 ± 0.97 μ ^ Ι after the vitamin had been infused for 1 month (p < 0.01). Since all infants do not receive IL in the first week of life, there was a need to develop a solution which could have the vitamin A added to the glucose/amino acid mix ture. For this reason a fourth study was performed [10]. In this study, intravenous RP was given at a dose of 400 μg/kg/day and R was given at a dose of 80 μg/kg/day. Results of this study indicated that total vitamin A con centrations increased substantially from the initial level of 18.4 ± 7.8 μς/άί to a total concentration of 69.7 μ^αΐ. However, the amount of R was maintained at basically the same level as had been seen in the previous study using R with IL: the total increase resulted primarily from the RP (Fig. 2). Vitamin E. In the initial study of vitamin E, pretreat ment levels of oc-tocopherol were 0.31 ± 0.03 mg/dl. After 3 weeks of treatment with 4.6 mg/kg/day, the levels increased to 3.08 ± 0.27 mg/dl. It was noted, how ever, that several infants received oral supplements of vitamin E (80-100 mg/day); these patients had levels much higher than those receiving no oral supplement (3.4 ± 0.6 vs 2.0 ± 0.2 mg/dl) [2]. Subsequent studies indicated that doses between 2.8 and 3.5 mg/kg/day were
sufficient to maintain blood levels at 1.0-2.5 mg/dl [6], Vitamin B,. Thirteen infants had EGR measurements performed during 28 days of TPN. The activity coeffi cient was within the normal reference range (< 1.2) for all infants (0.92 ± 0.02 to 0.89 ± 0.01) during the 28 days. Interpretation of these results was difficult since all infants had received blood transfusions. Thus, assay of B2 levels was performed. Blood and urine riboflavin concentrations during TPN showed a large degree of variation between patients. During the first few days of parenteral feedings, when 0.23 mg/kg/day of riboflavin was given, three patients had levels of 222, 300, and 392 ng/ml, respectively, and a dose of 0.41 mg/day for 2 days gave levels of 681, 652, and 763 ng/ml. A higher dose of 0.66 mg/kg/day resulted in a marked, although variable, increase in plasma riboflavin concentrations in plasma in all patients [6]. During the period of 21 days, the individual levels dif fered substantially, although a common pattern of plasma and urinary excretion was seen (Fig. 3). Following the initial marked increase in blood levels during the first 1-2 weeks, there was a progressive decline in levels. Coincident with the decline in plasma levels, urinary ex cretion increased markedly by the second week of life. This increased excretion was greater than the TPN dose and thus accounts for the decline in blood levels. The striking finding in this study was the remarkably elevated plasma riboflavin levels. There was concern in itially that these extremely high levels of riboflavin
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
285
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins
in Premature
Infants
might interfere with the conversion of riboflavin to its active coenzyme FMN and FAD. This did not appear to be the case, however, since the FAD levels prior to treat ment in red cells was 972 ± 1 2 1 and, after 28 days of treatment, it had increased to a level of 2005 ± 294 ng/g hemoglobin. Using the data from the three doses of vitamin B,, an estimate of the dose which would result in blood levels close to the target range was 0.15 mg/kg body weight. It is suggested that intakes of this range be tested in future studies. Vitamin B6. The recommended dose of vitamin B6 in current practices of TPN is 0.6 mg/day. Using this dose, the EGOT activity coefficient in 13 infants during 28 days of TPN remained normal (< 1.84). This dose per kg of body weight for VLBW infants was approximately three times the recommended dietary allowance of 0.4 mg/day for term infants (approximately 0.05-0.1 mg/kg/day). We postulated, therefore, that the dose given in TPN to VLBW infants would result in elevated blood levels similar to those seen with vitamin B2. After receiv ing the dose of 0.6 mg/day (0.4-0.8 mg/kg/day) in TPN for 2 weeks, the PLP level in seven patients was 87 ±11 nmol/L, similar to the cord blood level. Pyridoxic acid, however, had increased from 69 ± 5 to 989 ± 328 nmol/L, and the pyridoxal level increased from 23 ± 3 to 708 ± 140 nmol/L with a pyridoxine level of 251 ± 38 nmol/L. These indicate that the dose of pyridoxine which is presently given to VLBW infants results in extremely high levels in the blood. The infants also appear to be able to convert the intravenously administered pyridoxine to pyridoxal and its active cofactor, PLP. A lower dose of 0.06 and 0.2 mg/kg/day was given to three patients for the first 48 hours; then in the subsequent 48 hours the mean pyridoxal levels were 210 and 418 nmol/L. Using these dose responses, an estimated dose which would result in blood pyridoxal levels closer to the target range is 0.18 mg/kg/day. Future studies should evaluate this dosage.
DISCUSSION The preceding results suggest that the amounts of vitamins recommended in the package insert of MVI Pediatrie for prêterai infants caused a substantial devia tion from the target blood levels of each of the four vitamins tested. Vitamin A levels tended to be low, whereas vitamins E, B2, and B 6 tended to be high. Al though the target ranges may not necessarily represent optimal levels for VLBW infants, they do represent the expected levels consistent with normal growing neonates
286
or approximations of the levels expected during in trauterine growth. These levels therefore appear to be reasonable to use as guides for approximating a safe in take for intravenous vitamins [11]. Since the results indi cate a need to discuss the findings with regard to each vitamin, this will be provided below. Vitamin A. Vitamin A is present in the diet primarily as its ester, retinyl palmitate. This compound is first hydrolyzed then reesterified to RP in the intestine, secreted with the chylomicron fraction into lymphatics and circulated to the liver. The liver then hydrolyzes the RP to R and either stores it as RP in stellate cells or secretes the R into plasma complexed with R—binding protein and prealbumin. To accurately interpret the blood measurements of R and RP, the various stages of absorp tion, transport to the liver, hepatic uptake, and secretion into plasma must be taken into account. In this regard, a number of conditions such as infection or liver disease may alter one or more of the processes involved in liver or tissue uptake or release of vitamin A into plasma. These problems preclude the use of plasma vitamin A as a measure of vitamin A stores. This problem of inter pretation is probably magnified in VLBW infants be cause of immaturity, although the ontogeny of hepatic vitamin A uptake and secretion has not been inves tigated. Given the above limitations in interpretation of blood levels in response to intravenous vitamin A, there is an additional complication: substantial amounts of R added to parenteral solutions are lost in the delivery sets. Thus, the fact that blood levels do not reflect tissue stores of vitamin A in humans is even further compli cated in VLBW infants for two reasons: (1) the dosage received from TPN solutions is not known, and (2) the ontogeny of hepatic assimilation of vitamin A is not known. The first problem may be corrected by mixing the R with IL or using RP. These two options provide a basis for investigation of the question related to on togeny. Despite the apparent limitations in the interpreta tion of blood vitamin A concentrations, our data and that reported from other laboratories suggest that at birth most VLBW infants have insufficient vitamin A stores to meet their needs and that the dose recommended by the package insert for MVI Pediatrie should be increased. The American Society for Clinical Nutrition has publish ed a newer set of recommendations [12], suggesting that the intake of R in low-birth-weight infants should be approximately 500 μg/kg/day and that efforts to prevent losses during infusion should be instituted. It is hoped that this dose will maintain an adequate supply of R for the infant's needs during this critical period of develop ment. This recommendation is based on the above dis cussion of vitamin A as well as data which have been obtained by others [13-16].
VOL. 10, NO. 4
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins in Premature Infants Vitamin E. Tocopherol exists as several isomers, the most active being α-tocopherol. Intravenous lipid preparations contain enough tocopherols to affect local blood tocopherol levels. The majority of tocopherol in soybean oil emulsions is γ-tocopherol, which has sub stantially less biologic activity than cc-tocopherol, present in safflower oil emulsions. Thus, blood tocopherol measurement by HPLC is preferred over the total tocopherol level given by the calorimetrie assay because HPLC separates tocopherol into its various isomers. Needs of vitamin E are usually related directly to the level of dietary polyunsaturated fatty acids. Studies using mathematical or statistical modeling techniques, which compare a sensitive assay for deficiency (red blood cell peroxide hemolysis) with tocopherol levels in children, suggest that tocopherol levels < 0.44 mg/dl are as sociated with increased hemolysis. This level compares favorably with the conventionally accepted level of 0.5 mg/dl as the discriminator of vitamin E adequacy in adults. The influence of circulating lipids on tocopherol levels makes it preferable to express plasma cctocopherol levels as a ratio to ß-lipoprotein, cholesterol, or total lipids. However, in most instances newborn or premature infants show no abnormalities in circulating lipid levels that would lead to inappropriate interpreta tion of vitamin E, and levels are usually expressed in terms of plasma concentration. Neither the intravenous dose nor the optimal blood level of tocopherol in preterm infants has been defined. Because there is no proven benefit for pharmacologie doses of vitamin E and because several premature infants receiving large intravenous doses (25-100 mg/day) developed liver failure, the AAP Committee on Nutrition has suggested that safe and effective target blood levels are between 1 and 2 mg/dl [8]. Our findings using vitamin E at 2.1 and 4.6 mg/day in infants < 1500 g suggested that blood levels vary directly with the dose of vitamin E per kilogram of body weight. From these data, a dose response indicates that a daily dose of between 2.8 and 3.5 mg cc-tocopherol acetate/kg should maintain blood levels within the target range [16]. Vitamin B2. Riboflavin is phosphorylated to riboflavin-5-phosphate (FMN). As part of two coenzymes, FMN and FAD, riboflavin serves in cellular me tabolism for hydrogen transfer. When given in excess of needs, the vitamin is excreted primarily unchanged in the urine. Riboflavin is inactivated by light, especially that emitted from phototherapy lights, but FMN is more stable. For this reason, FMN is the form used for TPN. Using the measurement of erythrocyte glutathione reductase activity coefficient as a marker in breast-fed infants
receiving phototherapy shows a higher incidence of erythrocyte FAD depletion [17]. Despite the high rate of photodegradation expected from TPN solutions normally exposed to light during delivery, deficiency of riboflavin has not been described in children maintained on TPN. Our assessment of riboflavin status using activity of erythrocyte glutathione reductase before and after the addition of FAD indicated that there was not a deficiency of riboflavin. Since the enzyme assay will not detect marginal levels or high levels of riboflavin, methods to quantitate serum ribo flavin level as well as red blood cell riboflavin, FMN, and FAD content were developed. Our data indicate that a parenteral riboflavin intake of 0.66 mg/kg/day results in a 20- to 400-fold elevation in blood riboflavin levels in VLBW infants, and lower doses provided some estimate of a dose response [6]. Based on these estimates of riboflavin in blood, a dose of 0.15 mg/kg/day has been recommended [12]. Vitamin B6. Vitamin B6 is a generic descriptor for the three naturally occurring pyridines — pyridoxine, pyridoxal, and pyridoxamine — and their phosphorylated coenzyme forms of the vitamin — pyridoxine phosphate, PLP, and pyridoxamine phosphate, respectively. PLP is the primary cofactor form of the vitamin which functions in a wide range of metabolic reactions, primarily related to nitrogen metabolism. Although no toxicity of the vitamin has been ob served with medically indicated doses of vitamin B6, a sensory neuropathy in adults ingesting megadoses of pyridoxine has been described. The functional assay for vitamin B6 status uses an in vitro assay of erythrocyte aspartate aminotransferase (glutamic-oxaloacetic transaminase) before and after the addition of PLP. Using this measure of vitamin B6 status following doses of 0.3-0.7 mg/kg/day, the glutamic-oxaloacetic transferase activity coefficient was normal. This assay does not, however, measure blood concentrations of the vitamin. Assays of blood levels of pyridoxine, pyridoxal, and PLP, as well as urinary losses during TPN, should provide a better measure of intravenous needs. When 0.4 mg/kg/day was given, pyridoxine levels increased by > 10-fold over cord blood and maternal levels [5]. Lower doses of the vitamin resulted in data which allowed an estimate of a dose response, suggesting that 0.18 mg/kg should result in blood levels closer to the target range. This dose has been recommended for future formulations to be given to VLBW infants. In summary, these data from VLBW infants indicate that the currently formulated vitamin preparations does not maintain blood levels of all vitamins within an ac ceptable range. Our findings indicate that new formula tion should be made specifically for high-risk preterm
JOURNAL OF THE AMERICAN COLLEGE OF NUTRITION
287
Downloaded by [ECU Libraries] at 00:12 23 April 2015
Vitamins in Premature Infants infants. The following steps are recommended for the development of a more appropriate mixture of parenteral vitamins for preterm infants. (1) Separate mixtures of the water-soluble vitamins and the lipid-soluble vitamins should be provided. Because of the possibility of im paired metabolism of polysorbate and propylene glycol in preterm infants, efforts should be made to provide lipid-soluble vitamins without the addition of emulsifiers such as polysorbate. If a lipid emulsifier is essential, it should be a chemical that has been established to be nontoxic when administered intravenously to preterm in fants. (2) The water-soluble vitamin mixture should be tested in vitro to determine the most appropriate method of delivery to minimize photodegradation in the usual setting of an intensive care neonatal nursery when highintensity lighting is present on a continuous basis. (3) Both the water-soluble and the lipid-soluble vitamins should be evaluated in vitro to insure stability and bioavailability in TPN solutions under the conditions of an intensive care nursery. (4) In vivo studies of a few preterm infants who receive the new recommended doses should be evaluated prior to marketing a new preparation.
REFERENCES 1. Moore MC, Greene HL, Phillips B, Shulman RJ, Murrell JE, Ament ME: Evaluation of a pediatrie multiple vitamin preparation of total parenteral nutrition in infants and children, I. Blood levels of water-soluble vitamins. Pediatrics 77:530-538, 1986. 2. Greene HL, Moore MC, Phillips B, Franck L, Shulman RJ, Ament ME, Murrell JE, Chan MM, Said HM: Evalua tion of a pediatrie multiple vitamin preparation for total parenteral nutrition, II. Blood levels of vitamins A, D, and E. Pediatrics 77:539-547, 1986. 3. Greene HL, Phillips BL, Franck L, Fillmore CM, Said HM, Murrell JE, Courtney-Moore ME, Briggs R: Persist ently low blood retinol levels during and after parenteral feeding of very low birth weight infants: examination of losses into intravenous administration sets and a method of prevention by addition to a lipid emulsion. Pediatrics 79:892-900, 1987. 4. Phillips B, Franck LS, Greene HL: Vitamin E levels in premature infants during and after intravenous multivitamin supplementation. Pediatrics 80:680-683, 1987.
288
5. Greene HL. Baeckert PA, Murrell J, Oelburg DG, Adcock EW III: Blood pyridoxinc levels in preterm infants receiv ing TPN. Pediatr Res 25(4): 113A, 1990. 6. Baeckert PA, Greene HL, Oelberg DG, Adcock EW III: Blood levels of vitamins A, D, E, and riboflavin in very low birth weight infants receiving parenteral nutrition with vitamins added lo the lipid emulsion. J Pediatr 113:10571065, 1988. 7. Greene HL, Specker BL, Smith R, Murrell J, Swift LL: Plasma riboflavin concentrations in infants fed human milk versus l'onnula. J Pediatr 117(6):916-920, 1990. 8. Poland RL: Vitamin E: what should we do? Pediatrics 77:787-788. 1986. 9. Pietta P, Calatroni A, Rava A: Hydrolysis of riboflavin nucleoiides in plasma monitored by high performance liq uid chonnaiography. J Chromatogr 229:445-449, 1982. 10. Sampson D, O'Connor DK: Analysis of B 6 vitamers and pyridoxic acid in plasma, tissues, and urine using high pcrfomiance liquid chormatography. Nutr Res 9:259-272, 1989. 11. Pollack P. Murrell J, Caudill M, Swift L, Adcock E, Greene HL: Vilamin A in parenteral feedings for very low birth weight infants. Pediatr Res 27-.289A, 1990. 12. Greene HL, Hambidge KM, Schanler R, Tsang RC: Guidelines for the use of vitamins, trace elements, cal cium, magnesium, and phosphorus in infants and children receiving total parenteral nutrition: Report of the Subcom mittee on Pediatrie Parenteral Nutrient Requirements from the Committee on Clinical Practice Issues of the American Society for Clinical Nutrition. Am J Clin Nutr 48:13241342, 1986. 13. Shcnai JP, Chylil F, Jhaveri A, Stahlman MT: Plasma vitamin A and retinol binding protein in premature and terni neonates. J Pedialr 99:302-305, 1981. 14. Shenai JP, Kennedy KA, Chytil F, Stahlman MT: Clinical trial of vitamin A supplementation in infants susceptible to bronchopulmonary dysplasia. J Pediatr 111:269-277, 1987. 15. Brandt RB, Mueller DG, Shroeder JR, Guyer KE, Kirkpatrick BV, Hutcher NE, Ehrlich FE: Serum vitamin A in premature and lenn neonatcs. J Pediatr 92:101-105, 1978. 16. Husicatl VA, Gutchcr GR, Anderson SA, Zachman RD: Relationship of vilamin A (retinol) status to lung disease in Ihc prcicmi infant. J Pediatr 105:610-615, 1984. 17. Hovi L, Hckali Rand Sûmes MA: Evidence of riboflavin depletion in breasi-lcd newborns and its further accelera tion during treatment with phototherapy. Acta Pediatr Scand 68:567-570, 1979. Received August 1990; revision acceptedDecember
1990.
VOL. 10, NO. 4
