Early Human Development, 29 (1992) 345-350 Elsevier Scientific Publishers Ireland Ltd.
345
EHD 01301
Is mother’s milk the most suitable food for very low birth weight infants? W. Heine Children’s Hospital of the University of Restock (Germany)
Human milk feeding (HMF) as compared with formula feeding (FF) has the advantage of more effective utilization of proteins, fat, minerals and trace elements. HMF provides passive immunologic protection and active immunostimulation. It prevents the VLBWI from antigenic and toxic loads. The disadvantage of HMF is the high volume required tomeet the energy and protein needs of the VLBWI and the growing potential risk of AIDS, hepatitis and cytomegaly infections which makes human milk banking increasingly difficult. The current concept of VLBWI formula feeding (FF) is based on high protein, energy and mineral concentrations to compensate for the lower biological value, for lower bioavailability and for side effects related to the antigenicity of food proteins. FF as compared with HMF results is increased mineral and water retention, in high renal load and in a completely different body composition. The risk of necrotizing enteritis is significantly higher. All this has to be considered a challenge to further adapt LBWI formulas to the amino acid composition of human milk protein to induce bifidogenic effects and to provide sufficient amounts of essential fatty acids and carbohydrates which serve as building stones for normal brain development. Key words: very low birth weight infant; nutrient requirements; human milk; formula feeding
The answer to this question is yes and no. Human milk is a unique, highly sophisticated nutrient developed by the human species over millions of years. As compared to bovine milk based formulas human milk has numerous beneficial effects. Raw mother’s milk contains cellular elements such as macrophages and lymCorrespondence 20: W. Heine, Children’s Hospital of the University of Restock, 2500 Restock 1, Rembrandt Str. 16117, Restock, Germany. 0378-378U9USO5.00 0 1992 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland
346
phocytes which are responsible for phagocytosis, the production of complement factors, lysozyme, lactoferrin, for cell-associated immunity, sIgA and other immunoglobulins [l-3]. Bovine milk does not contain any of these cell products. The protective shield provided by human milk comprises immunoglobulins and other antibacterial substances, which may inhibit the growth of potential pathogenic intestinal microbes. Most of the cysteine of the human milk proteins is distributed among immunoglobulins and lysozyme. Cysteine is essential for bifidobacterial growth. It is apparently provided to a high extent to the intestinal microbes in the breast-fed baby due to the low digestibility of these protective proteins as they pass through the small intestines. The carbohydrates in human milk consist mainly of lactose, which serves as a source of energy for the baby as well as for the bifidobacteria and their growth. Beyond this lactose, however, human milk is rich in amino sugars such as sialic acid, which, in free and bound form serve as antimicrobial factors and as building blocks for gangliosides and glycoproteins of the brain. Lipids, fatty acids and monoglycerides found in human milk were also reported to possess antimicrobial effects. The entirety of the cellular and biochemical factors as well as the overall nutrient composition of human milk are in harmony with optimal thriving of the term-born infant. Nature has not provided for any comparable chances for the survival and growth of infants of very low birth weight. The golden standard for the requirements of these infants is the nutrient uptake from the placental flow and the corresponding accretion of nutrients in the fetal body mass. The order of magnitude of this process has been suspected of being not fully congruent with the composition of human milk [6]. The mass accretion of the fetal body does not take place at the same continuous pace. Rather, there are periods of rapid growth and high requirements of protein and energy alternating with periods of lower requirements. The highest net protein gain, amounting to 3.0 g kg-’ day-‘, is attained between the 28th and 32nd weeks of gestation [7]. After this period, the protein requirement drops from 3 g kg-’ day-’ to 1.0 g kg-’ day-’ in the 40th week of gestation, respectively. It is evident that an intake of 3.0 g protein kg-’ as required between the 28th and 32nd week of gestation cannot be met by the usual daily volume of human milk taken by the infant. Milk of mothers who have delivered prematurely has been shown to have higher protein concentrations than milk of mothers who have delivered on time [8,9]. This has led some investigators to speculate that this milk might be better suited for meeting the higher protein requirements of the low birth-weight infant. However, the higher protein content of preterm milk is mainly due to fractions such as sIgA; lactoferrin and lysozyme [lo] which are known to be poorly absorbed in the neonate. Their significance is related to their immunological function, protecting the preterm infant from microbial infections. Milk of mothers delivered of preterm infants has a somewhat higher sodium concentration as compared to mature mothers’ milk. By contrast, calcium and phosphorus concentrations are too low to meet the high requirements of the fast-growing preterm infant.
347
However, the chances that mothers who have given birth to infants with a low birth weight can produce sufficient milk to feed their babies over a long period are usually limited. It follows from these observations that mother’s milk cannot be considered fully satisfactory for the dietary management of the very low birth-weight infant. To meet the requirements of these infants, the milk has to be enriched with protein equivalents, energy, minerals and vitamins. Human milk modified in this way can be considered the most suitable mode of preterm infant nutrition, since it combines all the benefits of human milk feeding with the higher requirements of protein, energy, calcium, phosphorus and vitamins. One of the most convincing advantages of human milk feeding is the low incidence of necrotizing enterocolitis. Human milk for preterm infant nutrition used to be collected in human milk banks. This has since been given up as the risk of infection of AIDS and other contaminants had become increasingly higher. Meanwhile, preterm infants worldwide tend to be fed industrially produced preterm infant formulas rather than enriched human milk. From the point of view of pediatric nutritionists, the preterm infant formulas currently on the market do not represent the state of the art. They contain several components in excess and thus constitute a metabolic load, whereas other essential substrates are not or not sufficiently present in these formulas which may cause metabolic disturbances in low birth-weight infants. The protein content of adapted formulas for preterm infant nutrition varies between 1.7 and 2.3%. This protein concentration, which is comparatively high, is a consequence of an insufficient biological value of the protein components, which in comparison with human milk protein, has concentrations of some essential amino acids which are too low. The calculation of the amino acid scores as based on the single most limiting amino acid reveals a deficit in sulfur-containing amino acids and tryptophan in most formulas [l 11. Whey-dominated formulas contain excess threonine, whereas caseindominated formulas contain more aromatic amino acids than human milk protein. The order of magnitude of protein accretion is determined by the limiting amino acid within the protein mixture. A chemical score of a whey protein/casein mixture of 80% would mean that in a formula containing 1.8% protein only 1.4% of this protein would be available for protein synthesis. The excess protein will be degraded to ammonia and urea. The difference in amino acid composition between infant formulas and human milk are reflected in the plasma amino acid pattern of term-born infants [ 12-141 and accentuated in premature infants, whose protein intake is higher. The intake of bovine milk formula protein results in a specific elevation in the plasma concentrations of valine, phenylalanine and methionine; supplementing this with whey protein leads to further elevation in threonine, lysine, leucine and isoleucine concentrations [12]. Lowering the protein concentration of adapted formulas to correct these deviations in the homeostasis of the amino acids is then associated with low tryptophan and taurine concentrations and does not necessarily correct the plasma amino acid imbalances [ 131. Extremely low birth weight infants were shown to encounter particular difficulties in converting methionine to cystine [7]. The consequences of imbalances resulting from an excessive or deficient amino acid intake
348
for the mental and physical development of these infants are unknown. It is one of the most urgent tasks of the formula-manufacturing industry to adapt the amino acid pattern of preterm infant formulas to the composition of human milk protein [151. Human milk contains a considerable amount of long-chain polyunsaturated fatty acids. There is evidence that long chain polyunsaturated n-3 and n-6 fatty acids play an important role in normal brain development. There is a considerable enrichment of n-6 and n-3 long chain polyenoic fatty acids in the fetal brain during the last trimester of fetal life [16]. Preterm infant formulas should therefore be enriched with nutrients such as fish oil or egg yolk lipids, which contain these fatty acids in excess. However, recent studies conducted by Liu and coworkers [ 171 have revealed that the addition of fish oils to preterm infant formulas may supply n-3 long chain polyunsaturated fatty acids but does not improve the n-6 LPC status. This problem requires further attempts at finding a satisfying solution
V81.
Unlike bovine milk, human milk is known to contain amino sugars in free and bound form at strikingly high concentrations. Amino sugars such as Nacetylneuramic acid are important constituents of brain glycoproteins and gangliosides. They also play an important role as growth factors of bilidobacteria. The synthesis of amino sugars from carbohydrates such as glucose requires numerous ATP dependent enzymatic steps, which are presumably not fully developed in the very low birth-weight infant. Preterm infant formulas should therefore contain these sugar substances in sufficient amounts [19]. Another point at issue in preterm infant nutrition is the calcium, magnesium and phosphorus requirement. The fetal accretion rate of a 1000-g infant gaining weight at a rate of 20 g day -’ is 150 mg calcium, 90 mg phosphorus and 4.7 g magnesium [20]. This would correspond to an intake of 400-600 ml human milk per kg per day. Taking into account the limited intestinal absorption, the advisible intake would even be higher, namely 200-250 mg kg-’ for calcium and 1lo- 125 mg kg-’ for phosphorus. Commercially available formulas for preterm infants contain much lower quantities of calcium and phosphorus. It seems to be more rational to avoid excessive fecal losses of insoluble calcium phosphates by switching over to better absorbed calcium compounds instead of increasing the calcium and phosphate intake to such a high level. Summing up the current state of art, it has to be said that the majority of formulas designed for preterm infant nutrition does not represent an optimal substitute for human milk enriched with protein equivalents, energy and minerals. If the intrauterine protein accretion rate is taken as a standard, there is a high requirement for protein only between the 28th and 32nd weeks of gestation. Thereafter, preterm infants can thrive, like termborn infants, on low protein formulas [21,22]. The chemical score and the bioavailability of formula proteins as recommended for preterm infants is frequently below 80%. The low biological value of the food proteins is generally connected with excessive ammonia and urea formation which increases if, in catabolic states, the stabilizing effect of the protein synthesis is lost. Disturbances in intestinal motility are one of the most urgent practical problems in oral feeding of extremely low birth-weight infants. Commercially available preterm
349
infant formulas rich in protein and minerals tend to cause constipation and ileus symptoms in these infants. The reasons for these complications are not known in detail. There is, however, some evidence that extremely low birth-weight infants require a special composition of formulas, containing predigested protein and phosphates in relatively low concentration and relatively high concentrations of lactose in order to reduce the consistency of their feces. For creation of improved preterm infant formulas, attention will also have to be focussed on enrichment with long chain polyunsaturated fatty acids and essential carbohydrates, which serve as building blocks for normal brain development. The highly sophisticated composition of human milk can be considered the most reliable basis for further progress towards optimizing the composition of preterm infant formulas. References 1 Braun, O.H. (1976): Uber die infektionsverhiltende Wirkung der Muttermilch und deren mogliche Ursachen. Klin. Pildiatr., 188, 297-310. Goldman, AS. and Garca, C. (1987): Future research in human milk. Pediatr. Res., 22, 493-496. Gruttner, R. (1983): Neues iiber Frauemnilchemahnmg. Monatsschr. Kinderheilkd, 131,420-423. Schreier, K. (1983): EmPhrung und Immunologic. Monatsschr. Kinderheilkd, 131, 483-487. Heine, W., Mohr, Ch., Wutzke, K.D. (1992): Host-microflora correlations in infant nutrition. Prog. Food Nutr. Sci. (in press). 6 Widdowson, E.M., Southgate, O.A.T. and Hey, E.N. (1979): Body composition of the fetus and infant. In: Nutrition and Metabolism of the Fetus and Infant, pp. 169-179. Editor: H.A.K. Visser, Martinus Nijhoff. The Hague. Boston-London. Pohlandt, F. and Kupferschmid, Ch. (1985): The protein requirements of the preterm infant. Klin. Piidiatr., 197, 164-166 Atkinson, S.A. anderson, G.H. and Bryan, M.H. (1980): Human milk: comparison of the nitrogen composition in milk from mother’s of premature and full term infants. Am. J. Clin. Nutr., 81 l-815 9 Gross, S.J. (1980): Nutritional composition of milk produced by mothers delivering pretetm. J. Pediatr., 96, 641-644. 10 Gross, S.J., Buckley, R.H., Wakil, S.S., McAllister, DC., David, R.J. and Faix R.G. (1981): Elevated IgA concentration in milk produced by mothers delivered of preterm infants. J. Pediatr., 99, 389-393. 11 Sarwar, G., Botting, H.G. and Peace, R.W. (1989): Amino acid rating method for evaluating protein adequacy of infant formulas. J. Ass. Off. Anal. Chem., 72, 622-626. 12 Janas, L.M., Picciano, M.F. and Hatch, T.F. (1987): Indices of protein metabolism in term infants fed either human milk or formulas with reduced protein concentration and various whey/casein ratio. J. Pediatr., 110, 838-848. 13 Janas, L.M., Picciano, M.F. and Hatch, T.F. (1985): Indices of protein metabolism in term infants fed human milk, whey predominat formula or cow’s milk formula. Pediatrics, 75, 775-784. 14 Jiirvenpiil, A.L., Rassin, O.K., Raiha, N.C.R. and Gaull, G.E. (1982): Milk protein quantity and quality in the term infant. Effects on acidic and neutral amino acids. Pediatrics, 70, 221-230. 15 Heine, W., Klein, P.O. and Reeds, P. (1991): The significance of alphalactalbumin in infant nutrition. J. Nutr., 121, 277-283. 16 Clandinin, M.T. and Chappell, J.E. (1985): Long chain polyenoic essential fatty acids in human milk: are they of benefit to the newborn? In: Composition and physiological properties of human milk, pp. 213-224. Editor: J. Schaub. Ekevier, Amsterdam. 17 Pohlandt, F., Wagner, M., Rhein, R. and Obladen, M. (1990): Eine neue Aminosaurenliisung fur die parenterale Emlhnmg von Friihgeborenen, Neugeborenen und Siiuglingen. Infusionsther., 17, 40-46. 18 Liu, C.C.F., Carlson, S.E., Rhodes, P.G., Rao, V.S. and Meydrech, E.F. (1987): Increase in plasma
350
phospholipid docosahexaenoic and eicosapentaenoic acids as a reflection of their intake and mode of administration. Pediatr. Res., 22, 292-296. 19 Heine, W. (1991): Kohlenhydrate in parenteralen Nihrlosungen fur die Pldiatrie - Eine kritische Bewertung. Infusionsther., 18, 160-164. 20 Forbes, G.B. and Woodruff, C.B. (1985): Pediatric nutrition handbook. Acad. Pediatr., Elk Grove Village, Illinois. 21 Heine, W., Gassmann, B. and Plenert, W. (1968): Vergleichende Bilanxunersuchungen an jungen Siiuglingen unter EmHhrung mit eiwe$reichen und eiweifiarmen Fertignahrungen. PHdiatr Grenzgeb.,
7, 301-316.
22 Bhatia, 3. and Rassin, D.K. (1991): Feeding the premature infant after hospital discharge: growth and biochemical responses. J. Pediatr., 118, 515-519.
345
EHD 01301
Is mother’s milk the most suitable food for very low birth weight infants? W. Heine Children’s Hospital of the University of Restock (Germany)
Human milk feeding (HMF) as compared with formula feeding (FF) has the advantage of more effective utilization of proteins, fat, minerals and trace elements. HMF provides passive immunologic protection and active immunostimulation. It prevents the VLBWI from antigenic and toxic loads. The disadvantage of HMF is the high volume required tomeet the energy and protein needs of the VLBWI and the growing potential risk of AIDS, hepatitis and cytomegaly infections which makes human milk banking increasingly difficult. The current concept of VLBWI formula feeding (FF) is based on high protein, energy and mineral concentrations to compensate for the lower biological value, for lower bioavailability and for side effects related to the antigenicity of food proteins. FF as compared with HMF results is increased mineral and water retention, in high renal load and in a completely different body composition. The risk of necrotizing enteritis is significantly higher. All this has to be considered a challenge to further adapt LBWI formulas to the amino acid composition of human milk protein to induce bifidogenic effects and to provide sufficient amounts of essential fatty acids and carbohydrates which serve as building stones for normal brain development. Key words: very low birth weight infant; nutrient requirements; human milk; formula feeding
The answer to this question is yes and no. Human milk is a unique, highly sophisticated nutrient developed by the human species over millions of years. As compared to bovine milk based formulas human milk has numerous beneficial effects. Raw mother’s milk contains cellular elements such as macrophages and lymCorrespondence 20: W. Heine, Children’s Hospital of the University of Restock, 2500 Restock 1, Rembrandt Str. 16117, Restock, Germany. 0378-378U9USO5.00 0 1992 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland
346
phocytes which are responsible for phagocytosis, the production of complement factors, lysozyme, lactoferrin, for cell-associated immunity, sIgA and other immunoglobulins [l-3]. Bovine milk does not contain any of these cell products. The protective shield provided by human milk comprises immunoglobulins and other antibacterial substances, which may inhibit the growth of potential pathogenic intestinal microbes. Most of the cysteine of the human milk proteins is distributed among immunoglobulins and lysozyme. Cysteine is essential for bifidobacterial growth. It is apparently provided to a high extent to the intestinal microbes in the breast-fed baby due to the low digestibility of these protective proteins as they pass through the small intestines. The carbohydrates in human milk consist mainly of lactose, which serves as a source of energy for the baby as well as for the bifidobacteria and their growth. Beyond this lactose, however, human milk is rich in amino sugars such as sialic acid, which, in free and bound form serve as antimicrobial factors and as building blocks for gangliosides and glycoproteins of the brain. Lipids, fatty acids and monoglycerides found in human milk were also reported to possess antimicrobial effects. The entirety of the cellular and biochemical factors as well as the overall nutrient composition of human milk are in harmony with optimal thriving of the term-born infant. Nature has not provided for any comparable chances for the survival and growth of infants of very low birth weight. The golden standard for the requirements of these infants is the nutrient uptake from the placental flow and the corresponding accretion of nutrients in the fetal body mass. The order of magnitude of this process has been suspected of being not fully congruent with the composition of human milk [6]. The mass accretion of the fetal body does not take place at the same continuous pace. Rather, there are periods of rapid growth and high requirements of protein and energy alternating with periods of lower requirements. The highest net protein gain, amounting to 3.0 g kg-’ day-‘, is attained between the 28th and 32nd weeks of gestation [7]. After this period, the protein requirement drops from 3 g kg-’ day-’ to 1.0 g kg-’ day-’ in the 40th week of gestation, respectively. It is evident that an intake of 3.0 g protein kg-’ as required between the 28th and 32nd week of gestation cannot be met by the usual daily volume of human milk taken by the infant. Milk of mothers who have delivered prematurely has been shown to have higher protein concentrations than milk of mothers who have delivered on time [8,9]. This has led some investigators to speculate that this milk might be better suited for meeting the higher protein requirements of the low birth-weight infant. However, the higher protein content of preterm milk is mainly due to fractions such as sIgA; lactoferrin and lysozyme [lo] which are known to be poorly absorbed in the neonate. Their significance is related to their immunological function, protecting the preterm infant from microbial infections. Milk of mothers delivered of preterm infants has a somewhat higher sodium concentration as compared to mature mothers’ milk. By contrast, calcium and phosphorus concentrations are too low to meet the high requirements of the fast-growing preterm infant.
347
However, the chances that mothers who have given birth to infants with a low birth weight can produce sufficient milk to feed their babies over a long period are usually limited. It follows from these observations that mother’s milk cannot be considered fully satisfactory for the dietary management of the very low birth-weight infant. To meet the requirements of these infants, the milk has to be enriched with protein equivalents, energy, minerals and vitamins. Human milk modified in this way can be considered the most suitable mode of preterm infant nutrition, since it combines all the benefits of human milk feeding with the higher requirements of protein, energy, calcium, phosphorus and vitamins. One of the most convincing advantages of human milk feeding is the low incidence of necrotizing enterocolitis. Human milk for preterm infant nutrition used to be collected in human milk banks. This has since been given up as the risk of infection of AIDS and other contaminants had become increasingly higher. Meanwhile, preterm infants worldwide tend to be fed industrially produced preterm infant formulas rather than enriched human milk. From the point of view of pediatric nutritionists, the preterm infant formulas currently on the market do not represent the state of the art. They contain several components in excess and thus constitute a metabolic load, whereas other essential substrates are not or not sufficiently present in these formulas which may cause metabolic disturbances in low birth-weight infants. The protein content of adapted formulas for preterm infant nutrition varies between 1.7 and 2.3%. This protein concentration, which is comparatively high, is a consequence of an insufficient biological value of the protein components, which in comparison with human milk protein, has concentrations of some essential amino acids which are too low. The calculation of the amino acid scores as based on the single most limiting amino acid reveals a deficit in sulfur-containing amino acids and tryptophan in most formulas [l 11. Whey-dominated formulas contain excess threonine, whereas caseindominated formulas contain more aromatic amino acids than human milk protein. The order of magnitude of protein accretion is determined by the limiting amino acid within the protein mixture. A chemical score of a whey protein/casein mixture of 80% would mean that in a formula containing 1.8% protein only 1.4% of this protein would be available for protein synthesis. The excess protein will be degraded to ammonia and urea. The difference in amino acid composition between infant formulas and human milk are reflected in the plasma amino acid pattern of term-born infants [ 12-141 and accentuated in premature infants, whose protein intake is higher. The intake of bovine milk formula protein results in a specific elevation in the plasma concentrations of valine, phenylalanine and methionine; supplementing this with whey protein leads to further elevation in threonine, lysine, leucine and isoleucine concentrations [12]. Lowering the protein concentration of adapted formulas to correct these deviations in the homeostasis of the amino acids is then associated with low tryptophan and taurine concentrations and does not necessarily correct the plasma amino acid imbalances [ 131. Extremely low birth weight infants were shown to encounter particular difficulties in converting methionine to cystine [7]. The consequences of imbalances resulting from an excessive or deficient amino acid intake
348
for the mental and physical development of these infants are unknown. It is one of the most urgent tasks of the formula-manufacturing industry to adapt the amino acid pattern of preterm infant formulas to the composition of human milk protein [151. Human milk contains a considerable amount of long-chain polyunsaturated fatty acids. There is evidence that long chain polyunsaturated n-3 and n-6 fatty acids play an important role in normal brain development. There is a considerable enrichment of n-6 and n-3 long chain polyenoic fatty acids in the fetal brain during the last trimester of fetal life [16]. Preterm infant formulas should therefore be enriched with nutrients such as fish oil or egg yolk lipids, which contain these fatty acids in excess. However, recent studies conducted by Liu and coworkers [ 171 have revealed that the addition of fish oils to preterm infant formulas may supply n-3 long chain polyunsaturated fatty acids but does not improve the n-6 LPC status. This problem requires further attempts at finding a satisfying solution
V81.
Unlike bovine milk, human milk is known to contain amino sugars in free and bound form at strikingly high concentrations. Amino sugars such as Nacetylneuramic acid are important constituents of brain glycoproteins and gangliosides. They also play an important role as growth factors of bilidobacteria. The synthesis of amino sugars from carbohydrates such as glucose requires numerous ATP dependent enzymatic steps, which are presumably not fully developed in the very low birth-weight infant. Preterm infant formulas should therefore contain these sugar substances in sufficient amounts [19]. Another point at issue in preterm infant nutrition is the calcium, magnesium and phosphorus requirement. The fetal accretion rate of a 1000-g infant gaining weight at a rate of 20 g day -’ is 150 mg calcium, 90 mg phosphorus and 4.7 g magnesium [20]. This would correspond to an intake of 400-600 ml human milk per kg per day. Taking into account the limited intestinal absorption, the advisible intake would even be higher, namely 200-250 mg kg-’ for calcium and 1lo- 125 mg kg-’ for phosphorus. Commercially available formulas for preterm infants contain much lower quantities of calcium and phosphorus. It seems to be more rational to avoid excessive fecal losses of insoluble calcium phosphates by switching over to better absorbed calcium compounds instead of increasing the calcium and phosphate intake to such a high level. Summing up the current state of art, it has to be said that the majority of formulas designed for preterm infant nutrition does not represent an optimal substitute for human milk enriched with protein equivalents, energy and minerals. If the intrauterine protein accretion rate is taken as a standard, there is a high requirement for protein only between the 28th and 32nd weeks of gestation. Thereafter, preterm infants can thrive, like termborn infants, on low protein formulas [21,22]. The chemical score and the bioavailability of formula proteins as recommended for preterm infants is frequently below 80%. The low biological value of the food proteins is generally connected with excessive ammonia and urea formation which increases if, in catabolic states, the stabilizing effect of the protein synthesis is lost. Disturbances in intestinal motility are one of the most urgent practical problems in oral feeding of extremely low birth-weight infants. Commercially available preterm
349
infant formulas rich in protein and minerals tend to cause constipation and ileus symptoms in these infants. The reasons for these complications are not known in detail. There is, however, some evidence that extremely low birth-weight infants require a special composition of formulas, containing predigested protein and phosphates in relatively low concentration and relatively high concentrations of lactose in order to reduce the consistency of their feces. For creation of improved preterm infant formulas, attention will also have to be focussed on enrichment with long chain polyunsaturated fatty acids and essential carbohydrates, which serve as building blocks for normal brain development. The highly sophisticated composition of human milk can be considered the most reliable basis for further progress towards optimizing the composition of preterm infant formulas. References 1 Braun, O.H. (1976): Uber die infektionsverhiltende Wirkung der Muttermilch und deren mogliche Ursachen. Klin. Pildiatr., 188, 297-310. Goldman, AS. and Garca, C. (1987): Future research in human milk. Pediatr. Res., 22, 493-496. Gruttner, R. (1983): Neues iiber Frauemnilchemahnmg. Monatsschr. Kinderheilkd, 131,420-423. Schreier, K. (1983): EmPhrung und Immunologic. Monatsschr. Kinderheilkd, 131, 483-487. Heine, W., Mohr, Ch., Wutzke, K.D. (1992): Host-microflora correlations in infant nutrition. Prog. Food Nutr. Sci. (in press). 6 Widdowson, E.M., Southgate, O.A.T. and Hey, E.N. (1979): Body composition of the fetus and infant. In: Nutrition and Metabolism of the Fetus and Infant, pp. 169-179. Editor: H.A.K. Visser, Martinus Nijhoff. The Hague. Boston-London. Pohlandt, F. and Kupferschmid, Ch. (1985): The protein requirements of the preterm infant. Klin. Piidiatr., 197, 164-166 Atkinson, S.A. anderson, G.H. and Bryan, M.H. (1980): Human milk: comparison of the nitrogen composition in milk from mother’s of premature and full term infants. Am. J. Clin. Nutr., 81 l-815 9 Gross, S.J. (1980): Nutritional composition of milk produced by mothers delivering pretetm. J. Pediatr., 96, 641-644. 10 Gross, S.J., Buckley, R.H., Wakil, S.S., McAllister, DC., David, R.J. and Faix R.G. (1981): Elevated IgA concentration in milk produced by mothers delivered of preterm infants. J. Pediatr., 99, 389-393. 11 Sarwar, G., Botting, H.G. and Peace, R.W. (1989): Amino acid rating method for evaluating protein adequacy of infant formulas. J. Ass. Off. Anal. Chem., 72, 622-626. 12 Janas, L.M., Picciano, M.F. and Hatch, T.F. (1987): Indices of protein metabolism in term infants fed either human milk or formulas with reduced protein concentration and various whey/casein ratio. J. Pediatr., 110, 838-848. 13 Janas, L.M., Picciano, M.F. and Hatch, T.F. (1985): Indices of protein metabolism in term infants fed human milk, whey predominat formula or cow’s milk formula. Pediatrics, 75, 775-784. 14 Jiirvenpiil, A.L., Rassin, O.K., Raiha, N.C.R. and Gaull, G.E. (1982): Milk protein quantity and quality in the term infant. Effects on acidic and neutral amino acids. Pediatrics, 70, 221-230. 15 Heine, W., Klein, P.O. and Reeds, P. (1991): The significance of alphalactalbumin in infant nutrition. J. Nutr., 121, 277-283. 16 Clandinin, M.T. and Chappell, J.E. (1985): Long chain polyenoic essential fatty acids in human milk: are they of benefit to the newborn? In: Composition and physiological properties of human milk, pp. 213-224. Editor: J. Schaub. Ekevier, Amsterdam. 17 Pohlandt, F., Wagner, M., Rhein, R. and Obladen, M. (1990): Eine neue Aminosaurenliisung fur die parenterale Emlhnmg von Friihgeborenen, Neugeborenen und Siiuglingen. Infusionsther., 17, 40-46. 18 Liu, C.C.F., Carlson, S.E., Rhodes, P.G., Rao, V.S. and Meydrech, E.F. (1987): Increase in plasma
350
phospholipid docosahexaenoic and eicosapentaenoic acids as a reflection of their intake and mode of administration. Pediatr. Res., 22, 292-296. 19 Heine, W. (1991): Kohlenhydrate in parenteralen Nihrlosungen fur die Pldiatrie - Eine kritische Bewertung. Infusionsther., 18, 160-164. 20 Forbes, G.B. and Woodruff, C.B. (1985): Pediatric nutrition handbook. Acad. Pediatr., Elk Grove Village, Illinois. 21 Heine, W., Gassmann, B. and Plenert, W. (1968): Vergleichende Bilanxunersuchungen an jungen Siiuglingen unter EmHhrung mit eiwe$reichen und eiweifiarmen Fertignahrungen. PHdiatr Grenzgeb.,
7, 301-316.
22 Bhatia, 3. and Rassin, D.K. (1991): Feeding the premature infant after hospital discharge: growth and biochemical responses. J. Pediatr., 118, 515-519.
