Early Human Developmenl, 30 (1992) 163- 170 Elsevier Scientific Publishers Ireland Ltd.
163
EHD 01332
Comparison of enteral and intramuscular vitamin A supplementation in preterm infants Jacob Landman*,
Department
of Paediatrics
(Received
Alan Sive, Hans De V Heese, Clive Van Der Elst and Renee Sacks and Child Health,
18 January
1992; revision
University of Cape Town, Cape Town (South Africa) received
5 July 1992; accepted
13 July 1992)
Summary Vitamin A deficiency associated with preterm delivery is not readily reversible using the recommended supplement of 1500 IU per day. It has been reported that 2000 IU of intramuscular vitamin A administered on alternate days for 28 days will correct the deficiency. The objective of this study was to compare this regime with the practice in our nursery of giving 5000 IU of vitamin A per day with the early introduction of feeds. The vitamin A status of ten preterm infants (mean gestation 30.5 weeks) who received intramuscular vitamin supplementation was compared with that of nine infants (mean gestation 30.7 weeks) given enteral vitamin A. Vitamin A status was evaluated on the 32nd day of life using plasma retinol and retinol-binding protein (RBP) concentrations and a modified relative dose response (RDR) test. Plasma retinol and RBP concentrations were similar in the two groups shortly after birth revealing vitamin A deficiency. By the 32nd day of life, plasma retinol and RBP concentrations had risen significantly in both groups and in 70% the modified RDR was normal. Differences between the groups were not observed irrespective of the method of vitamin A administration. None of the infants developed clinical or biochemical vitamin A toxicity. In most preterm infants who tolerate feeds, vitamin A deficiency can be corrected safely by supplementing the feeds with 5000 IU of vitamin A per day. Key words: vitamin A deficiency; vitamin A supplementation; retinol-binding protein.
relative dose response;
Correspondence to: A. Sive, Department of Paediatrics and Child Health, University tute of Child Health, Red Cross War Memorial Children’s Hospital, Rondebosch, *Present address: Neonatal Bene Baraq, Israel.
Department,
Bene Baraq
0 1992 Elsevier Scientific 0378-3782/92/$05.00 Printed and Published in Ireland
Hospital,
Publishers
‘Ma’aynei
Ireland
Ltd
of Cape Town, Insti7700, South Africa.
Hayeshua’,
17 Sharet
Street,
164
Introduction
Preterm infants are born with limited hepatic reserves of vitamin A and low circulating concentrations of vitamin A and retinol-binding protein (RBP) [1,2,11,12]. Their poor vitamin A status has been postulated to increase their susceptibility to bronchopulmonary dysplasia (BPD) [ 13,7]. Administration of supraphysiological amounts of vitamin A may decrease the risk of developing chronic lung disease in these infants. Indeed, Shenai et al. [14] were able to reduce substantially the incidence of BPD in their nursery by giving infants at risk 2000 IU of vitamin A intramuscularly on alternate days in addition to the recommended daily allowance of 1500 IU of oral vitamin A when feeds were introduced. Although this dose of vitamin A is higher than that recommended by the American Academy of Pediatrics [5], it was not associated with vitamin A toxicity nor any other adverse effects. Moreover, these high doses of vitamin A repleted hepatic vitamin A stores in the majority of infants as determined by the plasma RBP response to vitamin A administration [15]. In our nursery it is customary to introduce enteral feeds via nasogastric tubes within the first few days of life in preterm infants and to supplement these feeds with 5000 IU of vitamin A per day. This dose of vitamin A has not been noted to cause clinical vitamin A toxicity nor has it been associated with plasma vitamin A concentrations in the toxic range. We hypothesised that the vitamin A status of preterm infants would not differ substantially whether they were given 2000 IU of vitamin A by intramuscular injection on alternate days or whether they received 5000 IU of enteral vitamin A daily. We measured the plasma vitamin A and retinol-binding protein responses to a vitamin A load in preterm infants who received either enteral or intramuscular vitamin A during the first month of life. Methods Subjects
Twenty-eight preterm infants who weighed less than 1500 g and who were appropriately grown for gestational age were eligible for the study. None had major anatomical malformations or overt congenital infections. Within 36 h of birth each infant was randomly assigned to receive either intramuscular (IM Group) or enteral (E Group) vitamin A. Nine babies did not complete the study leaving nine in the E Group and ten in the IM Group. Infants were withdrawn from the study if they developed septicaemia or hepatitis or if they were not given the appropriate dose of vitamin A by the appropriate route. Three infants died before the end of the study. General procedures
Informed consent was obtained from the custodial parent of each infant admitted to the study which was approved by the Ethics and Review Committee of the University of Cape Town. Infants alloted to the E Group were given 5000 IU of vitamin A daily via nasogastric tubes starting within 48 h of the introduction of feeds. In the IM Group 2000 IU of vitamin A (retinyl pahnitate, Roche, Switzerland) were given by deep intramuscular injection into the antero-lateral aspect of the thigh
165
on alternate days starting on the 4th day of life. Fourteen doses of IM vitamin A were administered to each baby in total. The IM group of infants also received 1500 IU of enteral vitamin A per day when feeds were introduced. Feeds were started at 3.1 days (range 2-7 days) in the E Group and 3.4 days (range 2-5 days) in the IM Group. Parenteral nutrition was used to supplement the energy intake where indicated. All babies also received vitamin A from their milk and parenteral nutrition solutions. The milk (Prenan, Nestle) and parenteral nutrition solutions contained 200 IU per 100 ml and 110 IU per 100 ml of vitamin A, respectively. Care was taken at all times to avoid photodegradation of the vitamin A by ensuring that fresh vials of vitamin A were used, that tubing was protected from light and that the oral preparation was flushed down the nasogastic tube with milk immediately after administration. Between 24 and 36 h after birth 0.5-l ml of blood was taken from a peripheral vein for the measurement of plasma vitamin A and retinol-binding protein concentrations. On the 32nd day after birth a modified relative dose response (RDR) test for vitamin A reserves was performed. Blood for vitamin A and RBP was sampled. Immediately thereafter, each infant was given 2000 IU per kg of vitamin A by intramuscular injection and 5 h later the measurements were repeated. Blood samples were collected into heparinized light impenetrable tubes and centrifuged in the dark to separate the plasma which was stored at -70°C until the samples were assayed in batches. Plasma vitamin A was measured as retinol using high performance liquid chromatography (Beckman System Gold, Beckman Instruments, San Ramon, CA) according to the method of Catignani and Bieri [3]. The intra- and inter-assay coefticients of variation were 4.4 and 7.0%, respectively. Retinol-binding protein was measured by radial immunodiffusion with a commercial kit (LC-Partigen, Behringwerke, Marburg, Germany). The modified RDR was calculated as the percent change in the plasma RBP concentration (A RBP) from the base line value by the equation A RBP (%) = [RBP (5 h) - RBP (baseline)]/RBP(baseline) x 100. A RBP was accepted as zero if the value was negative. The change in the vitamin A concentrations (A Vit A) was calculated in the same way. Statistical methocis Populations were compared using the Students t-test. All values are expressed as means f S.D. Differences in groups over time were compared using paired t-tests. Significance was accepted at the 5% level. Results The clinical characteristics of the infants are shown in Table I. Gestational age, weight and sex distribution were comparable in the two groups at delivery. The infants who were withdrawn from the study were similar to the study infants with respect to clinical data and vitamin A status. The mean vitamin A and RBP concentrations at birth in the infants who did not complete the study were 0.42 i 0.15 and 0.66 f 0.21 Fmol/l, respectively. These concentrations were similar to the mean values of 0.40 & 0.17 pmol/l for vitamin A and 0.67 f 0.25 pmolil for RBP in the
166
TABLE I Clinical characteristics of study subjects. Results reported as means f SD. or means f S.D. with ranges in parentheses. Characteristics
Enteral group
Intramuscular group
n
9 514 30.7 f 1288 f 1590 zt 300 f
416 30.5 + 1108 + 1385 f 275 f
Male/female Gestation (weeks) Birth weight (g) Weight day 32 Increase (g) Head Circumference Day 1 (cm) Day 30 (cm) Increase (cm)
10
1.03 (29-32) 137 (1000-1500) 132 (1450-1849) 65
27.5 f 0.5 (25-28.5) 30.4 zt 1.7 (28-32) 2.5 zt 1.2
1.36 (28-32) 199 (800-1450) 297 (1000-1800) 138
26.1 zt 2.2 (23-28) 28.6 + 2.2 (26-31) 2.5 + 0.77
patients who completed the study. The mean birth weight of 1104 i 182 g and gestational age of 30.2 f 1.6 weeks were also equivalent to those in the study patients. Four of these babies were assigned to the IM group and the remainder to the E group. Clinical outcome and the rates of growth remained similar in the two groups who completed the study until Day 32 when the dose response test was performed. Vitamin A intake Vitamin A intakes were similar in the groups during the first week of life but thereafter they were significantly higher in the E group (Table II) The average daily vitamin A consumption in the E group was 4564 f 420 IU with a range of 3708-5004 IU per day compared to 2531 f 246 IU, range 2071-2831 IU in the IM group. The IM group received 60% of their total vitamin A via the enteral route, 25% in the formula and the remainder as the standard enteral supplement of 1500 IU per day.
TABLE II Vitamin A intake of study
subjects.
Vitamin A (IU) Source
Enteral group
Intramuscular group
Feeds Oral supplements Intramuscular Total Average per day
8936 f 1449 118889 zt 12444
8510 f 33 900 f 28060 + 70470 f 2531 f
127825 zt 12534b 4564 zt 420
aIncludes TPN as a source. bInfants received significantly more vitamin A (P < 0.00001).
1753a 6165 1532 7280 246
167
TABLE III Vitamin A status of study subjects. Plasma concentrations (pmoW
Enteral group
Intramuscular group
Day 1 Vitamin A RBP Vitamin A/RBP ratio
0.36 f 0.147 0.66 zt 0.23 0.58 zt0.24
0.43 f 0.18 0.69 f 0.27 0.71 *0.22
Day 32 Vitamin A RBP Vitamin A/RBP ratio A Vitamin A % A RBP %
0.87 zt 0.27* 1.12 * 0.3; 0.78 f 0.21 7.6 zk 11.4 5.9 f 9.9
0.92 f 1.16 f 0.8 f 9.3 f 9.0 zt
0.44; 0.31* 0.28 15.9 13.2
*Significant increase (P = 0.0001) from day 1.
Vitamin A status
The plasma vitamin A and retinol binding protein concentrations, the vitamin A/RBP ratios and the dose response results are shown in Table III. The mean plasma vitamin A and RBP concentrations were similar in the two groups shortly after birth and on Day 32. After 28 days of vitamin A supplementation, the plasma vitamin A concentrations had risen significantly in both groups (E group P < 0.001, IM group P < 0.03) although three infants in each group still had hyporetinaemia (C 0.7 pmol/l). Retinol binding protein concentrations and the vitamin A/RBP molar ratio were similar in the groups at all times. Three infants in the E group and four in the IM group had positive A RBP responses to vitamin A administration indicating low liver stores but significant differences between the groups was not demonstrated. The individual A RBP responses in the patients concerned were 16, 30, 7.5% in the E group and 35, 11.5, 11.1 and 33% in the IM group. Values for the A vitamin A were similar being 23, 31 and 15% in the E group and 5.6, 33, 39% and 43% in the IM group, respectively.
The low plasma vitamin A and RBP concentrations within 24 h of delivery indicate that the infants in this study were vitamin A deficient at birth. Only one infant had a plasma retinol concentration above 0.7 pmolll. Indeed the mean vitamin A and RBP concentrations of 0.40 f 0.10 and 0.67 f 0.06 pmol/l, respectively for the combined study group was even lower than the concentrations reported by others for infants of similar gestation and birth weight [ 1,3,7]. These low levels may reflect the poor vitamin A status of the mothers many of whom come from extremely poor environments. Term infants born to mothers with low serum vitamin A levels have been shown to have low cord blood retinol concentrations [ 161. Plasma vitamin A
168
concentrations in the mothers of our infants were not measured but it is likely that they were low. Our unpublished observations in pregnant women living under the same environmental conditions show that 30-40% have hyporetinaemia. Thus prematurity coupled to maternal hyporetinaemia may have accounted for the exceedingly low vitamin A concentrations in these infants. Overt vitamin A deficiency does not occur in our area and none of the mothers in this study had clinical signs to support the diagnosis. Further studies are needed to ascertain whether vitamin A supplementation to pregnant mothers from poor socio-economic circumstances would improve the vitamin A status of the foetus and infant. High dose vitamin A is teratogenic in experimental animals [4] and vitamin A congeners are known to cause birth defects in humans especially when exposure occurs in the first trimester [lo]. Studies which examine the value of vitamin A supplementation in mothers and infants would be obliged to avoid high dose administration during the first half of pregnancy. Vitamin A supplements were given in one of two ways to the infants in this study. One group received oral supplementation as is the practice in our nursery whereas the others were given intramuscular vitamin A at the dose recommended by Shenai et al. [14]. In both groups there was a significant and similar rise in the vitamin A concentrations. However, the mean concentration achieved in our patients was lower than that reported by Shenai et al. even though a similar percentage had plasma vitamin A concentrations above 0.7 rmolll(70 versus 73%) [14]. The reason for this discrepancy is not clear but it may be the result of lower plasma vitamin A levels in our patients at the outset, to different practices in the protection of vitamin A from light or to differences in the handling of the blood samples. Since the RBP concentrations in our patients were also lower than those reported by Shenai et al. it would appear that the discrepancies between the two studies cannot be ascribed to differences in the processing of samples or laboratory technique. Despite lower retinol levels than in the infants reported by Shenai et al. liver stores of vitamin A on Day 32 in the infants in the two studies were similar as judged by the modified relative dose response test [ 151. A high plasma RBP or vitamin A response to oral vitamin A administration has been reported to be useful for identifying patients with low vitamin A status and low hepatic vitamin A reserves [8,9]. Experience with the test in preterm infants is limited [ 131 but a modified test using intramuscular as opposed to oral vitamin A it has been used to show that hepatic vitamin A stores are reduced and that vitamin A status improves with intramuscular vitamin A supplementation [ 151. In the majority of our patients, the RBP responses to vitamin A administration indicated adequate vitamin A stores irrespective of whether patients received oral or intramuscular supplements. This suggests that in infants who tolerate enteral feeds, there is no advantage to using intramuscular vitamin A supplements to improve vitamin A status provided enough enteral vitamin A is given. Whether high dose enteral vitamin A will prevent bronchopulmonary dysplasia was not addressed or answered in this study. Unlike Woodruff et al. [ 181 who reported significant vitamin A deficiency using 3000 IU of enteral vitamin A per day, we achieved satisfactory vitamin A status using this route but at a dose of 5000 IU per day. Because all infants could not be fed ab initio, the average daily amount of vitamin A received was less than intended
169
but more than that given previously. This high dose was also given for a longer period. Woodruff et al. [ 171 administered 3000 IU daily for 14 days whereas in our study the supplement was given for 28 days. Compared to Shenai et al. [ 14,151, who started enteral feeds on average on the 12th day of life and enterally administered vitamin A 5 days later, our infants received their enteral supplement from day 5. Hence, the cumulative dose of vitamin A in our study was much larger than in other studies. Formula appears to be a better vehicle for vitamin A administration than parenteral feeds [18] in which retinol may be lost by binding to the plastic tubing of the intravenous administration set [6]. It is feasible however that retention is better when the vitamin A is given enterally. With intramuscular injection, loss of vitamin to plastic surfaces is less of a problem. It is unfortunate that we were unable to use breast milk in place of formula feeds. Most mothers of preterm infants in our nursery are indigent and it is not feasible for them to travel to the hospital often enough to provide breast milk. To ensure uniformity in this study, we only included infants who were unlikely to be breast fed. High doses of vitamin A given for a relatively long time may result in vitamin A toxicity. None of the infants in this study had plasma retinol concentrations in the toxic range and the highest plasma retinol concentration of 1.88 pmol/l was recorded in an infant in the IM group. Furthermore, none of the infants showed clinical evidence of toxicity as determined by an abnormal increase in the head circumference, a raised fontanelle or findings on ultrasound examination of the head that were not in keeping with their clinical course. Irrespective of the method of administration, preterm infants require large doses of vitamin A to correct their deficiency. Whether this relates mainly to the decreased tranplacental acquisition of retinol, to decreased bioavailability or to increased demands is unclear. However, this study has shown that in the majority of preterm infants who tolerate feeds, vitamin A deficiency can be corrected safely using 5000 IU of enterally administered vitamin A per day for 28 days. Acknowledgement This study was supported by the Rita and Bernard Brodie Research Associateship in Nutrition. The authors are grateful to Mrs Glynnis Kossew for her technical assistance and MS D Du Toit for typing the manuscript.
1 2 3 4 5
Bhatia, J. and Ziegler, E.E. (1983): Retinol-binding protein and prealbumin in cord blood of term and preterm infants. Early Hum. Dev., 8, 129-133. Brandt, R.B., Mueller, D.G., Schroeder, J.R., Guyer, K.E., Kirkpatrick, B.V., Hutcher, N.E. and Ehrlich, F.E. (1978): Serum vitamin A in premature and term neonates. J. Pediatr., 92, 101-104. Catignani, G.L. and Bieri, J.G. (1983): Simultaneous determination of retinol and alpha-tocopherol in serum or plasma by liquid chromatography. Clin. Chem., 29, 708-712. Cohlan, S.Q. (1954): Congenital anomalies in the rat produced by excessive intake of vitamin A during pregnancy. Pediatrics, 13, 556-567. Committee on Nutrition, American Academy of Pediatrics (1985): Nutritional needs of low-birthweight infants. Pediatrics, 75, 976-986.
170
10 11 12 13 14 15
16 17
18
Hartline, J.V. and Zachman, R.D. (1976): Vitamin A delivery in total parenteral nutrition solution. Pediatrics, 58, 448-451. Hustead, V.A., Gutcher, G.R., Anderson, S.A. and Zachman, R.D. (1984): Relationship of vitamin A (retinol) to lung disease in the preterm infant. J. Pediatr., 105, 610-615. Loerch, J.D., Underwood, B.A. and Lewis, KC. (1979): Response of plasma levels of vitamin A to a dose of vitamin A as an indicator of vitamin A reserves in rats. J. Nutr., 109, 778-786. Mobarahan, S., Russell, R.M., Underwood, B., Wallingford, J., Mathieson, R.D. and Al-Midani, H. (1981): Evaluation of the relative dose response test for vitamin A nutriture in cirrhotics. Am. J. Clin. Nutr., 34, 2264-2270. Rosa, F.W., Wilk, A.L. and Kelsey, F.O. (1986): Teratogen update: vitamin A congeners. Tetratology, 33, 335-364. Shenai, J.P., Chytil, F., Jhaveri, A. and Stahlman, M.T. (1981): Plasma vitamin A and retinolbinding protein in premature and term neonates. J. Pediatr., 99, 302-305. Shenai, J.P., Chytil, F. and Stahlman, M.T. (1985): Liver vitamin A reserves of very low birth weight neonates. Pediatr. Res., 19, 892-893. Shenai, J.P., Chytil, F. and Stahlman, M.T. (1985): Vitamin A status of neonates with bronchopulmonary dysplasia. Pediatr. Res., 19, 185-188. Shenai, J.P., Kennedy, K.A., Chytil, F. and Stahlman, M.T. (1987): Clinical trial of vitamin A supplementation in infants susceptible to bronchopulmonary dysplasia. J. Pediatr., 111, 269-277. Shenai, J.P., Rush, M.G., Stahlman, M.T. and Chytil, F. (1990): Plasma retinol-binding protein response to vitamin A administration in infants susceptible to bronchopulmonary dysplasia. J. Pediatr., 116, 607-614. Shirali, G.S., Oelberg, D.G. and Mehta, K.P. (1989): Maternal-neonatal serum vitamin A concentrations. J. Pediatr. Gastroenterol. Nutr., 9, 62-66. Woodruff, C.W., Latham, C.B., Mactier, H. and Hewett, J.E. (1987): Vitamin A status of preterm infants: correlation between plasma retinol concentration and retinol dose response. Am. J. Clin. Nutr., 46, 985-988. Woodruff, C.W., Latham, C.B., James, E.P. and Hewett, J.E. (1986): Vitamin A status of pretetm infants: the influence of feeding and vitamin supplements. Am. J. Clin. Nutr., 44, 384-389.
163
EHD 01332
Comparison of enteral and intramuscular vitamin A supplementation in preterm infants Jacob Landman*,
Department
of Paediatrics
(Received
Alan Sive, Hans De V Heese, Clive Van Der Elst and Renee Sacks and Child Health,
18 January
1992; revision
University of Cape Town, Cape Town (South Africa) received
5 July 1992; accepted
13 July 1992)
Summary Vitamin A deficiency associated with preterm delivery is not readily reversible using the recommended supplement of 1500 IU per day. It has been reported that 2000 IU of intramuscular vitamin A administered on alternate days for 28 days will correct the deficiency. The objective of this study was to compare this regime with the practice in our nursery of giving 5000 IU of vitamin A per day with the early introduction of feeds. The vitamin A status of ten preterm infants (mean gestation 30.5 weeks) who received intramuscular vitamin supplementation was compared with that of nine infants (mean gestation 30.7 weeks) given enteral vitamin A. Vitamin A status was evaluated on the 32nd day of life using plasma retinol and retinol-binding protein (RBP) concentrations and a modified relative dose response (RDR) test. Plasma retinol and RBP concentrations were similar in the two groups shortly after birth revealing vitamin A deficiency. By the 32nd day of life, plasma retinol and RBP concentrations had risen significantly in both groups and in 70% the modified RDR was normal. Differences between the groups were not observed irrespective of the method of vitamin A administration. None of the infants developed clinical or biochemical vitamin A toxicity. In most preterm infants who tolerate feeds, vitamin A deficiency can be corrected safely by supplementing the feeds with 5000 IU of vitamin A per day. Key words: vitamin A deficiency; vitamin A supplementation; retinol-binding protein.
relative dose response;
Correspondence to: A. Sive, Department of Paediatrics and Child Health, University tute of Child Health, Red Cross War Memorial Children’s Hospital, Rondebosch, *Present address: Neonatal Bene Baraq, Israel.
Department,
Bene Baraq
0 1992 Elsevier Scientific 0378-3782/92/$05.00 Printed and Published in Ireland
Hospital,
Publishers
‘Ma’aynei
Ireland
Ltd
of Cape Town, Insti7700, South Africa.
Hayeshua’,
17 Sharet
Street,
164
Introduction
Preterm infants are born with limited hepatic reserves of vitamin A and low circulating concentrations of vitamin A and retinol-binding protein (RBP) [1,2,11,12]. Their poor vitamin A status has been postulated to increase their susceptibility to bronchopulmonary dysplasia (BPD) [ 13,7]. Administration of supraphysiological amounts of vitamin A may decrease the risk of developing chronic lung disease in these infants. Indeed, Shenai et al. [14] were able to reduce substantially the incidence of BPD in their nursery by giving infants at risk 2000 IU of vitamin A intramuscularly on alternate days in addition to the recommended daily allowance of 1500 IU of oral vitamin A when feeds were introduced. Although this dose of vitamin A is higher than that recommended by the American Academy of Pediatrics [5], it was not associated with vitamin A toxicity nor any other adverse effects. Moreover, these high doses of vitamin A repleted hepatic vitamin A stores in the majority of infants as determined by the plasma RBP response to vitamin A administration [15]. In our nursery it is customary to introduce enteral feeds via nasogastric tubes within the first few days of life in preterm infants and to supplement these feeds with 5000 IU of vitamin A per day. This dose of vitamin A has not been noted to cause clinical vitamin A toxicity nor has it been associated with plasma vitamin A concentrations in the toxic range. We hypothesised that the vitamin A status of preterm infants would not differ substantially whether they were given 2000 IU of vitamin A by intramuscular injection on alternate days or whether they received 5000 IU of enteral vitamin A daily. We measured the plasma vitamin A and retinol-binding protein responses to a vitamin A load in preterm infants who received either enteral or intramuscular vitamin A during the first month of life. Methods Subjects
Twenty-eight preterm infants who weighed less than 1500 g and who were appropriately grown for gestational age were eligible for the study. None had major anatomical malformations or overt congenital infections. Within 36 h of birth each infant was randomly assigned to receive either intramuscular (IM Group) or enteral (E Group) vitamin A. Nine babies did not complete the study leaving nine in the E Group and ten in the IM Group. Infants were withdrawn from the study if they developed septicaemia or hepatitis or if they were not given the appropriate dose of vitamin A by the appropriate route. Three infants died before the end of the study. General procedures
Informed consent was obtained from the custodial parent of each infant admitted to the study which was approved by the Ethics and Review Committee of the University of Cape Town. Infants alloted to the E Group were given 5000 IU of vitamin A daily via nasogastric tubes starting within 48 h of the introduction of feeds. In the IM Group 2000 IU of vitamin A (retinyl pahnitate, Roche, Switzerland) were given by deep intramuscular injection into the antero-lateral aspect of the thigh
165
on alternate days starting on the 4th day of life. Fourteen doses of IM vitamin A were administered to each baby in total. The IM group of infants also received 1500 IU of enteral vitamin A per day when feeds were introduced. Feeds were started at 3.1 days (range 2-7 days) in the E Group and 3.4 days (range 2-5 days) in the IM Group. Parenteral nutrition was used to supplement the energy intake where indicated. All babies also received vitamin A from their milk and parenteral nutrition solutions. The milk (Prenan, Nestle) and parenteral nutrition solutions contained 200 IU per 100 ml and 110 IU per 100 ml of vitamin A, respectively. Care was taken at all times to avoid photodegradation of the vitamin A by ensuring that fresh vials of vitamin A were used, that tubing was protected from light and that the oral preparation was flushed down the nasogastic tube with milk immediately after administration. Between 24 and 36 h after birth 0.5-l ml of blood was taken from a peripheral vein for the measurement of plasma vitamin A and retinol-binding protein concentrations. On the 32nd day after birth a modified relative dose response (RDR) test for vitamin A reserves was performed. Blood for vitamin A and RBP was sampled. Immediately thereafter, each infant was given 2000 IU per kg of vitamin A by intramuscular injection and 5 h later the measurements were repeated. Blood samples were collected into heparinized light impenetrable tubes and centrifuged in the dark to separate the plasma which was stored at -70°C until the samples were assayed in batches. Plasma vitamin A was measured as retinol using high performance liquid chromatography (Beckman System Gold, Beckman Instruments, San Ramon, CA) according to the method of Catignani and Bieri [3]. The intra- and inter-assay coefticients of variation were 4.4 and 7.0%, respectively. Retinol-binding protein was measured by radial immunodiffusion with a commercial kit (LC-Partigen, Behringwerke, Marburg, Germany). The modified RDR was calculated as the percent change in the plasma RBP concentration (A RBP) from the base line value by the equation A RBP (%) = [RBP (5 h) - RBP (baseline)]/RBP(baseline) x 100. A RBP was accepted as zero if the value was negative. The change in the vitamin A concentrations (A Vit A) was calculated in the same way. Statistical methocis Populations were compared using the Students t-test. All values are expressed as means f S.D. Differences in groups over time were compared using paired t-tests. Significance was accepted at the 5% level. Results The clinical characteristics of the infants are shown in Table I. Gestational age, weight and sex distribution were comparable in the two groups at delivery. The infants who were withdrawn from the study were similar to the study infants with respect to clinical data and vitamin A status. The mean vitamin A and RBP concentrations at birth in the infants who did not complete the study were 0.42 i 0.15 and 0.66 f 0.21 Fmol/l, respectively. These concentrations were similar to the mean values of 0.40 & 0.17 pmol/l for vitamin A and 0.67 f 0.25 pmolil for RBP in the
166
TABLE I Clinical characteristics of study subjects. Results reported as means f SD. or means f S.D. with ranges in parentheses. Characteristics
Enteral group
Intramuscular group
n
9 514 30.7 f 1288 f 1590 zt 300 f
416 30.5 + 1108 + 1385 f 275 f
Male/female Gestation (weeks) Birth weight (g) Weight day 32 Increase (g) Head Circumference Day 1 (cm) Day 30 (cm) Increase (cm)
10
1.03 (29-32) 137 (1000-1500) 132 (1450-1849) 65
27.5 f 0.5 (25-28.5) 30.4 zt 1.7 (28-32) 2.5 zt 1.2
1.36 (28-32) 199 (800-1450) 297 (1000-1800) 138
26.1 zt 2.2 (23-28) 28.6 + 2.2 (26-31) 2.5 + 0.77
patients who completed the study. The mean birth weight of 1104 i 182 g and gestational age of 30.2 f 1.6 weeks were also equivalent to those in the study patients. Four of these babies were assigned to the IM group and the remainder to the E group. Clinical outcome and the rates of growth remained similar in the two groups who completed the study until Day 32 when the dose response test was performed. Vitamin A intake Vitamin A intakes were similar in the groups during the first week of life but thereafter they were significantly higher in the E group (Table II) The average daily vitamin A consumption in the E group was 4564 f 420 IU with a range of 3708-5004 IU per day compared to 2531 f 246 IU, range 2071-2831 IU in the IM group. The IM group received 60% of their total vitamin A via the enteral route, 25% in the formula and the remainder as the standard enteral supplement of 1500 IU per day.
TABLE II Vitamin A intake of study
subjects.
Vitamin A (IU) Source
Enteral group
Intramuscular group
Feeds Oral supplements Intramuscular Total Average per day
8936 f 1449 118889 zt 12444
8510 f 33 900 f 28060 + 70470 f 2531 f
127825 zt 12534b 4564 zt 420
aIncludes TPN as a source. bInfants received significantly more vitamin A (P < 0.00001).
1753a 6165 1532 7280 246
167
TABLE III Vitamin A status of study subjects. Plasma concentrations (pmoW
Enteral group
Intramuscular group
Day 1 Vitamin A RBP Vitamin A/RBP ratio
0.36 f 0.147 0.66 zt 0.23 0.58 zt0.24
0.43 f 0.18 0.69 f 0.27 0.71 *0.22
Day 32 Vitamin A RBP Vitamin A/RBP ratio A Vitamin A % A RBP %
0.87 zt 0.27* 1.12 * 0.3; 0.78 f 0.21 7.6 zk 11.4 5.9 f 9.9
0.92 f 1.16 f 0.8 f 9.3 f 9.0 zt
0.44; 0.31* 0.28 15.9 13.2
*Significant increase (P = 0.0001) from day 1.
Vitamin A status
The plasma vitamin A and retinol binding protein concentrations, the vitamin A/RBP ratios and the dose response results are shown in Table III. The mean plasma vitamin A and RBP concentrations were similar in the two groups shortly after birth and on Day 32. After 28 days of vitamin A supplementation, the plasma vitamin A concentrations had risen significantly in both groups (E group P < 0.001, IM group P < 0.03) although three infants in each group still had hyporetinaemia (C 0.7 pmol/l). Retinol binding protein concentrations and the vitamin A/RBP molar ratio were similar in the groups at all times. Three infants in the E group and four in the IM group had positive A RBP responses to vitamin A administration indicating low liver stores but significant differences between the groups was not demonstrated. The individual A RBP responses in the patients concerned were 16, 30, 7.5% in the E group and 35, 11.5, 11.1 and 33% in the IM group. Values for the A vitamin A were similar being 23, 31 and 15% in the E group and 5.6, 33, 39% and 43% in the IM group, respectively.
The low plasma vitamin A and RBP concentrations within 24 h of delivery indicate that the infants in this study were vitamin A deficient at birth. Only one infant had a plasma retinol concentration above 0.7 pmolll. Indeed the mean vitamin A and RBP concentrations of 0.40 f 0.10 and 0.67 f 0.06 pmol/l, respectively for the combined study group was even lower than the concentrations reported by others for infants of similar gestation and birth weight [ 1,3,7]. These low levels may reflect the poor vitamin A status of the mothers many of whom come from extremely poor environments. Term infants born to mothers with low serum vitamin A levels have been shown to have low cord blood retinol concentrations [ 161. Plasma vitamin A
168
concentrations in the mothers of our infants were not measured but it is likely that they were low. Our unpublished observations in pregnant women living under the same environmental conditions show that 30-40% have hyporetinaemia. Thus prematurity coupled to maternal hyporetinaemia may have accounted for the exceedingly low vitamin A concentrations in these infants. Overt vitamin A deficiency does not occur in our area and none of the mothers in this study had clinical signs to support the diagnosis. Further studies are needed to ascertain whether vitamin A supplementation to pregnant mothers from poor socio-economic circumstances would improve the vitamin A status of the foetus and infant. High dose vitamin A is teratogenic in experimental animals [4] and vitamin A congeners are known to cause birth defects in humans especially when exposure occurs in the first trimester [lo]. Studies which examine the value of vitamin A supplementation in mothers and infants would be obliged to avoid high dose administration during the first half of pregnancy. Vitamin A supplements were given in one of two ways to the infants in this study. One group received oral supplementation as is the practice in our nursery whereas the others were given intramuscular vitamin A at the dose recommended by Shenai et al. [14]. In both groups there was a significant and similar rise in the vitamin A concentrations. However, the mean concentration achieved in our patients was lower than that reported by Shenai et al. even though a similar percentage had plasma vitamin A concentrations above 0.7 rmolll(70 versus 73%) [14]. The reason for this discrepancy is not clear but it may be the result of lower plasma vitamin A levels in our patients at the outset, to different practices in the protection of vitamin A from light or to differences in the handling of the blood samples. Since the RBP concentrations in our patients were also lower than those reported by Shenai et al. it would appear that the discrepancies between the two studies cannot be ascribed to differences in the processing of samples or laboratory technique. Despite lower retinol levels than in the infants reported by Shenai et al. liver stores of vitamin A on Day 32 in the infants in the two studies were similar as judged by the modified relative dose response test [ 151. A high plasma RBP or vitamin A response to oral vitamin A administration has been reported to be useful for identifying patients with low vitamin A status and low hepatic vitamin A reserves [8,9]. Experience with the test in preterm infants is limited [ 131 but a modified test using intramuscular as opposed to oral vitamin A it has been used to show that hepatic vitamin A stores are reduced and that vitamin A status improves with intramuscular vitamin A supplementation [ 151. In the majority of our patients, the RBP responses to vitamin A administration indicated adequate vitamin A stores irrespective of whether patients received oral or intramuscular supplements. This suggests that in infants who tolerate enteral feeds, there is no advantage to using intramuscular vitamin A supplements to improve vitamin A status provided enough enteral vitamin A is given. Whether high dose enteral vitamin A will prevent bronchopulmonary dysplasia was not addressed or answered in this study. Unlike Woodruff et al. [ 181 who reported significant vitamin A deficiency using 3000 IU of enteral vitamin A per day, we achieved satisfactory vitamin A status using this route but at a dose of 5000 IU per day. Because all infants could not be fed ab initio, the average daily amount of vitamin A received was less than intended
169
but more than that given previously. This high dose was also given for a longer period. Woodruff et al. [ 171 administered 3000 IU daily for 14 days whereas in our study the supplement was given for 28 days. Compared to Shenai et al. [ 14,151, who started enteral feeds on average on the 12th day of life and enterally administered vitamin A 5 days later, our infants received their enteral supplement from day 5. Hence, the cumulative dose of vitamin A in our study was much larger than in other studies. Formula appears to be a better vehicle for vitamin A administration than parenteral feeds [18] in which retinol may be lost by binding to the plastic tubing of the intravenous administration set [6]. It is feasible however that retention is better when the vitamin A is given enterally. With intramuscular injection, loss of vitamin to plastic surfaces is less of a problem. It is unfortunate that we were unable to use breast milk in place of formula feeds. Most mothers of preterm infants in our nursery are indigent and it is not feasible for them to travel to the hospital often enough to provide breast milk. To ensure uniformity in this study, we only included infants who were unlikely to be breast fed. High doses of vitamin A given for a relatively long time may result in vitamin A toxicity. None of the infants in this study had plasma retinol concentrations in the toxic range and the highest plasma retinol concentration of 1.88 pmol/l was recorded in an infant in the IM group. Furthermore, none of the infants showed clinical evidence of toxicity as determined by an abnormal increase in the head circumference, a raised fontanelle or findings on ultrasound examination of the head that were not in keeping with their clinical course. Irrespective of the method of administration, preterm infants require large doses of vitamin A to correct their deficiency. Whether this relates mainly to the decreased tranplacental acquisition of retinol, to decreased bioavailability or to increased demands is unclear. However, this study has shown that in the majority of preterm infants who tolerate feeds, vitamin A deficiency can be corrected safely using 5000 IU of enterally administered vitamin A per day for 28 days. Acknowledgement This study was supported by the Rita and Bernard Brodie Research Associateship in Nutrition. The authors are grateful to Mrs Glynnis Kossew for her technical assistance and MS D Du Toit for typing the manuscript.
1 2 3 4 5
Bhatia, J. and Ziegler, E.E. (1983): Retinol-binding protein and prealbumin in cord blood of term and preterm infants. Early Hum. Dev., 8, 129-133. Brandt, R.B., Mueller, D.G., Schroeder, J.R., Guyer, K.E., Kirkpatrick, B.V., Hutcher, N.E. and Ehrlich, F.E. (1978): Serum vitamin A in premature and term neonates. J. Pediatr., 92, 101-104. Catignani, G.L. and Bieri, J.G. (1983): Simultaneous determination of retinol and alpha-tocopherol in serum or plasma by liquid chromatography. Clin. Chem., 29, 708-712. Cohlan, S.Q. (1954): Congenital anomalies in the rat produced by excessive intake of vitamin A during pregnancy. Pediatrics, 13, 556-567. Committee on Nutrition, American Academy of Pediatrics (1985): Nutritional needs of low-birthweight infants. Pediatrics, 75, 976-986.
170
10 11 12 13 14 15
16 17
18
Hartline, J.V. and Zachman, R.D. (1976): Vitamin A delivery in total parenteral nutrition solution. Pediatrics, 58, 448-451. Hustead, V.A., Gutcher, G.R., Anderson, S.A. and Zachman, R.D. (1984): Relationship of vitamin A (retinol) to lung disease in the preterm infant. J. Pediatr., 105, 610-615. Loerch, J.D., Underwood, B.A. and Lewis, KC. (1979): Response of plasma levels of vitamin A to a dose of vitamin A as an indicator of vitamin A reserves in rats. J. Nutr., 109, 778-786. Mobarahan, S., Russell, R.M., Underwood, B., Wallingford, J., Mathieson, R.D. and Al-Midani, H. (1981): Evaluation of the relative dose response test for vitamin A nutriture in cirrhotics. Am. J. Clin. Nutr., 34, 2264-2270. Rosa, F.W., Wilk, A.L. and Kelsey, F.O. (1986): Teratogen update: vitamin A congeners. Tetratology, 33, 335-364. Shenai, J.P., Chytil, F., Jhaveri, A. and Stahlman, M.T. (1981): Plasma vitamin A and retinolbinding protein in premature and term neonates. J. Pediatr., 99, 302-305. Shenai, J.P., Chytil, F. and Stahlman, M.T. (1985): Liver vitamin A reserves of very low birth weight neonates. Pediatr. Res., 19, 892-893. Shenai, J.P., Chytil, F. and Stahlman, M.T. (1985): Vitamin A status of neonates with bronchopulmonary dysplasia. Pediatr. Res., 19, 185-188. Shenai, J.P., Kennedy, K.A., Chytil, F. and Stahlman, M.T. (1987): Clinical trial of vitamin A supplementation in infants susceptible to bronchopulmonary dysplasia. J. Pediatr., 111, 269-277. Shenai, J.P., Rush, M.G., Stahlman, M.T. and Chytil, F. (1990): Plasma retinol-binding protein response to vitamin A administration in infants susceptible to bronchopulmonary dysplasia. J. Pediatr., 116, 607-614. Shirali, G.S., Oelberg, D.G. and Mehta, K.P. (1989): Maternal-neonatal serum vitamin A concentrations. J. Pediatr. Gastroenterol. Nutr., 9, 62-66. Woodruff, C.W., Latham, C.B., Mactier, H. and Hewett, J.E. (1987): Vitamin A status of preterm infants: correlation between plasma retinol concentration and retinol dose response. Am. J. Clin. Nutr., 46, 985-988. Woodruff, C.W., Latham, C.B., James, E.P. and Hewett, J.E. (1986): Vitamin A status of pretetm infants: the influence of feeding and vitamin supplements. Am. J. Clin. Nutr., 44, 384-389.
