Nutrient Requirements and Provision of Nutritional Support in the Premature Neonate

Brenda B. Poindexter and Richard A. Ehrenkranz

Current recommendations for the provision of parenteral and enteral nutrition to the infant born prematurely are based on the goal of approximating the rate and composition of weight gain of a normal fetus at the same postmenstrual age. For a number of reasons this goal is seldom achieved and postnatal growth failure remains a nearly universal complication of extreme prematurity. Infants who experience one or more major morbidities, such as bronchopulmonary dysplasia, necrotizing enterocolitis (NEC), or late-onset sepsis in the neonatal period, demonstrate slower growth than their counterparts who do not experience these conditions. Of particular concern is the association between suboptimal postnatal growth and adverse neurodevelopmental outcomes. To emphasize the importance of a combined approach to optimize outcomes in the premature neonate, this chapter will first describe overall nutrient requirements and then provision of intense nutritional support with both early parenteral and enteral nutrition.

Growth and Neurodevelopmental Outcomes

At the time of birth, approximately 20% of very low birth weight (VLBW) infants are small for gestational age (defined as weight less than the 10th percentile).³² After a stay in the neonatal intensive care unit, however, the vast majority of these infants will have experienced poor growth and weigh less than the 10th percentile at 36 weeks’ postmenstrual age. The Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) Neonatal Research Network has published growth outcomes of VLBW infants over the past several decades. Among VLBW infants enrolled in the Neonatal Research Network generic database of morbidity and mortality, the incidence of growth failure at 36 weeks’ postmenstrual age was 97% between 1995 and 1996⁴⁸ and 91% between 1997 and 2002.³² In the most recent report between 2003 and 2007, the incidence of postnatal growth failure decreased to 79%,⁸⁵ possibly reflecting widespread use of early parenteral nutrition at most neonatal intensive care units as well as a decline in the use of postnatal corticosteroids.

The association between weight gain in the neonatal intensive care unit and neurodevelopmental outcomes was evaluated in a cohort of extremely low birth weight (ELBW) infants admitted to centers participating in the NICHD Neonatal Research Network between 1994 and 1995.²⁹ Infants were divided into quartiles based on in-hospital growth velocity; infants in the highest quartile gained an average of 21 g/kg per day, and those in the lowest quartile gained 12 g/kg per day. Compared with infants in the highest quartile of in-hospital weight gain, the odds of cerebral palsy, Bayley Scales of Infant Development (BSID) II mental developmental index (MDI) less than 70, and neurodevelopmental impairment were significantly higher in infants in the lowest quartile of in-hospital weight gain. Similar findings were also observed with in-hospital head growth. A recent evaluation using a more contemporary population and the new edition of the Bayley (BSID III) have reported findings consistent with the earlier work by Ehrenkranz. Other investigators have recently described the association between poor linear growth and neurodevelopmental outcomes.⁶⁸ As evidence continues to accumulate that early nutritional inadequacies have long-term consequences, optimizing provision of both parenteral and enteral nutrition to high-risk neonates is crucial to ensure the best possible outcomes.

Among potential causes of postnatal growth failure are significant protein and energy deficits that occur in the early neonatal period that are difficult to recoup in extremely premature infants.³¹ Early deficiencies in protein are a particularly important contributor to the poor growth outcomes observed in this population.^11,60 Consequently, provision of adequate nutrition to avoid these deficits altogether is necessary to optimize growth in premature infants.

Nutrient Requirements

Protein and Amino Acid Requirements

Duplicating rates of in utero protein accretion remains a difficult clinical challenge. Failure to provide adequate protein, either in quantity or in quality, can significantly impact the long-term outcome of extremely premature infants. A variety of methods have been used to quantitate protein requirements in human infants: fetal accretion rates, nitrogen balance studies, plasma amino acid concentrations, and stable isotope studies investigating the kinetics of labeled essential amino acids. Clearly, the gold standard needs to be that which safely optimizes growth and neurodevelopment.

Normal human fetal development is characterized by rapid rates of growth and accretion of protein. In fact, the greatest rate of relative protein gain throughout life occurs prior to birth. At 26 weeks’ gestation, the human fetus gains approximately 1.8 to 2.2 g of body protein per day (Figure 43-1), with the placenta supplying about 3.5 g/kg per day of amino acids to the developing fetus. The placental supply of amino acids to the fetus is in excess of that needed for accretion of protein. The extra amino acids are oxidized by the fetus and contribute significantly to fetal energy production.

Figure 43-1 Change in body protein over the first week of postnatal life for a theoretical 1000-gram birth weight, 26 weeks’ gestation infant provided with 1 g/kg/day *(blue line with dashes)* and 3 g/kg/day *(solid blue line)* of intravenous amino acids.88,93 Rate of in utero protein gain for the reference fetus *(yellow line)* and extrapolated rates of protein loss for glucose alone²⁴ *(red line)* are shown for comparison.

In contrast to the high rate of protein gain in utero, protein losses in extremely premature infants are approximately twofold higher than in term infants.²⁴ In the absence of intravenous amino acids, extremely premature infants lose approximately 1.2 g/kg of protein each day, which corresponds to a daily loss of 1% to 2% of total endogenous body protein stores (see Figure 43-1).

Protein requirements and recommendations for protein intake in premature infants have been made based on several different approaches (Table 43-1). Ziegler has quantified protein and energy requirements in relation to body weight using the factorial approach (Table 43-2).¹⁰³ In the factorial approach, nutrient requirements are determined as the sum of the needs for growth plus needs for replacement of losses. The disadvantage of this approach is that it does not account for nutrient requirements for catch-up growth. It is important to note that protein requirements are inversely related to body weight. Based on the factorial method, protein requirements for infants who weigh less than 1200 grams are estimated at 4.0 g/kg per day. Other investigators have utilized empirical approaches to estimate the amount of protein intake required to duplicate fetal growth.^43,69 An expert panel convened by the Life Sciences Research Office (LSRO) concluded that the minimum protein intake for premature infants is 3.4 g/kg per day, and that an intake of 4.3 g/kg per day is tolerated without adverse consequences.⁴⁵

TABLE 43-1

Protein Requirements and Recommended Intakes

	Weight <1200 g		Weight >1200 g
	g/kg/d	g/100 kcal	g/kg/d	g/100 kcal
Ziegler¹⁰³	4.0	3.7	3.6	2.8
Kashyap & Heird⁴³	—	—	3.0	2.5
Tsang⁶⁹	3.8-4.2	3.3	3.4-3.6	2.8
LSRO⁴⁵	3.4-4.3	2.5-3.6	3.4-4.3	2.5-3.6
ESPGHAN³	4.0-4.5	3.6-4.1	3.5-4.0	3.2-3.6

TABLE 43-2

Protein and Energy Requirements of Premature Infants Determined by the Factorial Approach

Body Weight (g)	Protein (g/kg/d)	Energy (kcal/kg/d)	Protein/Energy (g/100 kcal)
500-700	4.0	105	3.8
700-900	4.0	108	3.7
900-1200	4.0	119	3.4
1200-1500	3.9	125	3.1
1500-1800	3.6	128	2.8
1800-2200	3.4	131	2.6

Although duplication of in utero weight gain can be achieved at protein intakes of 3 to 3.5 g/kg per day, the European Society of Pediatric Gastroenterology and Nutrition (ESPGHAN) has recently recommended an even higher range of protein intake, taking into account the need to make up for accumulated protein deficits observed in nearly all extremely premature infants. Consequently, the recommendation by ESPGHAN for protein intake in infants weighing up to 1000 g is 4.0 to 4.5 g/kg per day and 3.5 to 4.0 g/kg per day for infants weighing 1000 to 1800 g.³ The ESPGHAN committee further commented that protein intake can be reduced toward discharge if the infant’s growth pattern allows for this.

Protein Composition of Human Milk and Formula

The protein content and composition of human milk changes throughout lactation; the concentration diminishes from about 2 g/dL at birth to about 1 g/dL for mature milk. Qualitative changes also occur during lactation, resulting in a whey-casein ratio of 80 : 20 at the beginning of lactation, changing to 55 : 45 in mature milk. Although the levels of casein, α-lactalbumin, albumin, and lysozyme remain constant, the levels of secretory immunoglobulin A and lactoferrin decrease. Because these different protein fractions have different amino acid profiles, the content of the individual amino acids is also affected.

Casein-predominant cow milk formulas have the same whey-casein ratio as cow milk—that is, 18 : 82. Adding bovine whey makes whey-predominant formulas, such that the whey-casein ratio becomes similar to that of human milk (60 : 40). Nevertheless, the protein and amino acid profile remains very different from that of human milk. Compared with human milk, whey-predominant formulas have higher levels of methionine, threonine, lysine, and branched amino acids. These differences in amino acids have not resulted in any apparent clinical consequences in either term or preterm infants. Currently, most commercially available preterm formulas are whey predominant.

Energy and Carbohydrate Requirements

Fetal Glucose Metabolism

Because the supply of glucose to the fetus depends solely on maternal glucose, cord clamping at the time of birth requires that a number of events occur in order to maintain glucose homeostasis in the newborn. Fetal glucose use in utero matches umbilical glucose uptake, implying that glycogenolysis and gluconeogenesis are minimal in the fetus. Several factors promote glycogen deposition in utero: blunted pancreatic β cell regulation of insulin secretion, high insulin receptor density, and relative glucagon resistance. In late gestation, the fetus begins to prepare for the transition to postnatal life by increasing hepatic glycogen stores and brown fat deposits. Hepatic glycogen synthesis increases in response to increases in adrenal corticosteroid production, also characteristic of late gestation. At the time of delivery, glucagon levels rise, and insulin levels fall. Higher levels of plasma catecholamines, both epinephrine and norepinephrine, directly stimulate increases in hepatic glucose output. The increased levels of epinephrine and glucagon stimulate lipolysis and the activity of phosphorylase, a key enzyme in glycolysis. The increased level of glucagon also results in increased activity of phosphoenolpyruvate carboxykinase, a rate-limiting enzyme in gluconeogenesis. The newborn must be able to initiate gluconeogenesis, because glycogen stores can sustain glucose production only for several hours after birth. All of these changes act together to preserve glucose homeostasis after the infant’s maternal source of glucose is removed with cord clamping.

Neonatal Glucose Homeostasis

Because neural tissue makes up a greater proportion of body weight, newborns have higher rates of glucose oxidation than adults, with glucose being the primary energy substrate for the brain. The rate of glucose production in term newborns is approximately 3 to 5 mg/kg per minute,²³ whereas extremely premature infants have an even higher rate of glucose production of approximately 8 to 9 mg/kg per minute.⁴⁰ Historically, glucose intolerance in infants with extremely low birth weights was attributed to persistent endogenous hepatic glucose production in the face of increased exogenous supply, insufficient insulin production, or tissue insensitivity to insulin. However, ELBW infants are able to suppress endogenous glucose production when given parenteral glucose.⁴⁰

Energy Expenditure

Preterm infants have very low energy reserves, owing to limited body fat and glycogen stores in the liver. Maintaining these limited energy stores requires an energy intake that approximates energy expenditure. Energy expenditure in premature infants is thought to be in the 50 to 60 kcal/kg per day range, but it must be noted that data for ventilated infants and infants with extremely low birth weights are limited. Thermal stress can substantially increase energy expenditure. Conversely, activity contributes relatively little to energy expenditure in premature infants. The energy cost of growth in these infants has been estimated at about 5 kcal/g. To achieve the equivalent of the estimated third-trimester in utero weight gain of 14 to 18 g/kg per day, theoretically, an additional energy intake of about 70 kcal/kg per day is necessary.

Energy balance is a delicate equilibrium between energy intake and energy loss plus storage. Positive energy balance is achieved when exogenous metabolizable energy intake is greater than energy expenditure. Growth is then possible, with the excess energy stored as new tissue, usually fat. If exogenous energy intake is less than expenditure, energy balance is negative, and body energy stores must be mobilized to meet ongoing needs. During the acute phase of disease, the primary goal is avoidance of catabolism. This is difficult for infants with very low birth weights, owing to their higher maintenance energy requirements, lower energy stores, and often reduced intake. The components of estimated energy requirements for growing premature infants is shown in Table 43-3.

TABLE 43-3

Estimated Energy Requirements for Growing Premature Infants

Energy Expenditure	kcal/kg/d
Resting metabolic rate	40-60
Activity	0-5
Thermoregulation	0-5
Synthesis/energy cost of growth	15
Energy stored	20-30
Energy excreted	15
Total energy requirement (estimated)	90-120

Energy is lost either by excretion or by expenditure. Energy excreted is lost primarily as fecal fat. In preterm infants receiving full enteral feeding, approximately 90% of energy intake is absorbed. Usual measurements of total energy expenditure include the energy used to maintain basal metabolic rate (BMR), as well as the postprandial increase in energy expenditure (diet-induced thermogenesis), physical activity, and energy for the synthesis of new tissue. The BMR is the largest component of energy expenditure and includes energy requirements for basic cellular and tissue processes. In a critically ill patient, it also includes a “disease factor.” Little is known about this component in neonates, but it may contribute significantly in neonates with fever, sepsis, and chronic hypoxia. Because the BMR can be measured only after overnight fasting, the resting metabolic rate (RMR) has been accepted as an alternative. The RMR of preterm infants on a per kilogram basis is higher than that of term infants, and the nutritional requirements on a per kilogram basis are correspondingly greater. The energy required for thermoregulation can be minimized by keeping the infant in a thermoneutral environment and limiting stimulation. For example, energy requirements for thermoregulation are negligible under thermoneutral conditions, but routine nursing procedures can increase oxygen consumption or energy expenditure by as much as 10% in stable preterm infants. Both of these components may be of considerable magnitude in a critically ill infant.

The level of energy intake and diet composition determine the magnitude of diet-induced thermogenesis, which represents the energy required for transport, metabolism, and conversion of nutrients into stored energy. Estimates of diet-induced thermogenesis in enterally fed preterm infants vary considerably, but this component of expenditure is probably small.

Total energy expenditure is affected by several factors, assuming a thermoneutral environment and minimal interference from nursing procedures. Increases in metabolic rate with postnatal age are influenced primarily by energy intake and weight gain. Energy expenditure increases with increases in metabolizable energy intake, indicating increased substrate oxidation or tissue synthesis. If the preterm infant is growing at the same rate as the fetus during the third trimester—that is, gaining approximately 15 g/kg per day—then about 15% of the total energy intake is used for synthesis of new tissue.

Energy storage is a linear function of metabolizable intake, and the accretion rate for energy is related more to the level of metabolizable energy intake than to diet composition. Energy requirements for energy storage are difficult to predict. The increase in tissue mass during growth includes the energy stored as protein, carbohydrate (usually less than 1% of body weight), and fat. Therefore, the energy stored can be assumed to equal the sum of the cost of protein plus fat gain. The energy storage component of the energy balance equation is a function of the composition of weight gain, which in turn is a function of protein and energy intake and is likely to be quite variable. Therefore, the energy intake required to produce a specific rate of weight gain cannot be predicted without specifying the composition of that weight gain. From the point of view of energy storage, protein is a poor material because a small quantity of energy is stored per gram of weight gain. Approximately the same amount of energy is deposited in 1 g of fat tissue as in about 8 g of lean tissue.

Carbohydrates

Lactose is the predominant carbohydrate in human milk (6.2-7.2 g/dL) and supplies 40% to 50% of the caloric content. Lactose is hydrolyzed to glucose and galactose in the small intestine by β-galactosidase (lactase). Intestinal lactase activities in premature infants at 34 weeks’ gestational age are approximately 30% of those of term infants.⁴⁴ Despite low lactase activities in premature infants, lactose is well tolerated by premature infants, and stable isotope data suggest efficient lactose digestion. However, most premature infant formulas include glucose polymers as a significant source of carbohydrate; these glucose polymers are digested by α-glucosidases, which achieve 70% of adult activity between 26 and 34 weeks’ gestation. In addition, salivary and mammary amylases may contribute to glucose polymer digestion. Glucose polymers have the advantage of increased caloric density without a rise in osmolality, and they may also enhance gastric emptying.

Recommended Energy Intake

The American Academy of Pediatrics ¹ has recommended an average energy intake of 105 to 130 kcal/kg per day for preterm infants. The ESPGHAN Committee on Nutrition recommendation for energy intake for growing, premature infants with adequate protein intake is 110 to 135 kcal/kg per day.³ Infants who are small for gestational age or infants with diseases that increase energy requirements may need higher intakes to achieve the same growth rates. Infants with growth restriction often require an increased caloric intake for growth because of both higher maintenance energy needs and higher energy costs of new tissue synthesis.

Lipid Requirements

In humans, linoleic and linolenic acids cannot be endogenously synthesized and are therefore essential fatty acids. Biochemical evidence of essential fatty acid deficiency can develop in preterm infants within 72 hours. Essential fatty acid deficiency can be avoided if a minimum of 0.5 to 1.0 g/kg/day of intravenous lipid is provided. To meet energy requirements, additional intravenous lipid is required in early postnatal life.

Fat provides the major source of energy for growing preterm infants. At birth, the digestive function of premature infants is not fully developed; preterm infants have decreased gut absorption of lipids because of low levels of pancreatic lipase, bile acids, and lingual lipase. The fact that term and preterm infants absorb fat reasonably well is due to the development of alternative mechanisms for the digestion of dietary fat. One important mechanism is intragastric lipolysis, in which lingual and gastric lipases compensate for the low pancreatic lipase concentration. By 25 weeks’ gestation, lingual lipase is secreted by the serous glands of the tongue, and gastric lipase is secreted from gastric glands. The fatty acids and monoglycerides resulting from intragastric lipolysis compensate for low bile acid concentration by emulsifying lipid mixtures. Lingual lipase can also penetrate the core of the human milk lipid globule and hydrolyze the triglyceride core without disrupting the globule membrane. Human milk provides lipoprotein lipase, bile salt stimulated esterase, and nonactivated lipase to further aid lipolysis.

Lipid digestion and absorption are also affected by dietary fat composition. Fatty acid absorption increases with decreasing chain length and with the degree of unsaturation, meaning that medium chain triglycerides (MCT) with chain lengths of 6 to 12 carbons are hydrolyzed more readily than long chain triglycerides (LCT), and that fatty acids with more double bonds are absorbed more efficiently. In an attempt to increase the fat absorption of premature infants, the fat in commercial formulas contains relatively high levels of MCTs that can be absorbed without the need for lipase or bile salts. Standard commercial formulas for healthy term infants do not contain MCTs, and human milk typically contains 8% to 12% of fat as MCTs. Unlike LCTs, MCTs are readily hydrolyzed in the gut, and the released fatty acids are transported across the gut barrier without the need for bile acids. Then MCTs are transported directly to the liver via the portal vein as nonesterified fatty acids. In addition, MCTs can enter mitochondria and be oxidized without the need for carnitine-mediated transport through mitochondrial membranes. However, inclusion of MCTs in infant formula remains controversial, because the available data do not support the assertion of improved fat absorption or improved growth in preterm infants.

In human milk, fat is transported in globules consisting of a membrane composed of a polar mixture of proteins, phospholipids, triglycerides, cholesterol, glycoproteins, and enzymes surrounding a triglyceride core containing 98% of the fat in milk. The milk fat globules are among the largest structural components of milk, having a diameter of 4 mm in mature milk. The size of the globules increases with both length of lactation and length of gestation, with colostrum having smaller globules (especially in milk of women who deliver prematurely) than mature milk. After birth, as the total fat content of human milk increases, the percentage of cholesterol and phospholipids, both of which reside primarily in the milk fat globule membrane, decreases; in addition, the total phospholipid content decreases as lactation progresses. During the first weeks of lactation, preterm milk is also richer in membranous material compared with term or mature milk, resulting in a higher content of cholesterol, phospholipids, and very-long-chain polyunsaturated fatty acids (PUFA) with chain lengths of 20 to 22 carbons (C20-C22). Because these membranes act as emulsifiers that allow fat dispersion in an aqueous phase and limit lipolysis and oxidation, heat treatment or addition of fortifiers and supplements might disrupt this emulsion.

The milk fat content and nutritional value of human milk vary with time, and it does not always provide a complete source of nutrients for infants with very low birth weights. Its composition and energy content may vary in a pumping session and during subsequent changes throughout lactation. The total fat content of human milk at 3 days’ lactation is approximately 2 g/dL; the fat content of mature milk is approximately 4 to 5 g/dL, with large individual variation possible. The triglyceride content of human milk is its most variable component, changing with gestational and postnatal age, time of day, duration of individual feeds, and maternal diet. Shifts in the dietary practices of a population result in changes in the fatty acid composition of human milk, because the type and amount of fat in the maternal diet affect the composition of milk fat. Maternal diets low in fat and high in carbohydrate lead to de novo synthesis of fatty acids within the mammary gland, which results in high concentrations of fatty acids of less than 16 carbons. Therefore, although the total amount of fat present in the milk remains in the normal range, the fat is more saturated.

Fatty acids represent about 85% of the triglycerides and therefore are the principal component of human milk lipids. Fatty acids in human milk are derived from the maternal diet, de novo synthesis by the mammary gland, and mobilization from fat stores. The fatty acid composition of human milk fat reflects the fatty acid composition of the maternal diet. The long-chain polyunsaturated fatty acid (LCPUFA) composition of the milk of women in the United States, Europe, and Africa is quite similar, with the exception of higher amounts of n-3 LCPUFA in the milk of women whose diets contain a large quantity of fish. Medium-chain fatty acids (C8-C10) do not normally account for more than 2% of the fats, even in milk from women who have delivered preterm. Arachidonic acid (C20:4n-6) is the main LCPUFA, and eicosapentaenoic acid (C20:5n-3) is found in small quantities in human milk. Docosahexaenoic acid (C22:6n-3) is the main LCPUFA of the n-3 series.

Fatty acid composition changes with progressing lactation and with gestational age. Most striking is the higher content of C8-C14 fatty acids and of LCPUFAs in preterm milk as compared with term milk; the content of LCPUFA decreases with increasing postnatal age. This may be an advantage for preterm infants because shorter fatty acids are easier to digest, and LCPUFAs are essential for brain and retinal development.

The LCPUFAs play an important role in the development of the infant’s brain during the last trimester of pregnancy and also during the first months of life.⁹⁰ The precursor C18 fatty acids for the n-6 and n-3 LCPUFA series are linoleic acid (C18:2n-6) and α-linolenic acid (C18:3n-3). These are further elongated and desaturated to form other fatty acids, of which arachidonic acid (AA) and docosahexaenoic acid (DHA) are essential for normal growth and development. Although the capacity for endogenous synthesis of LCPUFA from precursor fatty acids in preterm and term infants was thought to be limited, stable isotope studies demonstrated that both term and preterm infants have the capacity to synthesize DHA and AA.^18,73 However, it remains unclear whether DHA and AA can be biosynthesized in quantities sufficient to meet the needs of these infants. In utero, LCPUFAs are supplied to the fetus across the placenta. After birth, breastfed infants receive sufficient preformed dietary LCPUFA with human milk.

Although contained in human milk, DHA and AA have only recently been added to most infant formulas. Currently, most formula manufacturers have elected to add DHA and AA to their premature formulas. Multiple randomized controlled trials evaluating the addition of DHA and AA to preterm formulas have been conducted. Added DHA and AA have generally resulted in positive or neutral changes in growth, although there are some reports of a negative effect. Findings of improved visual acuity have been inconsistent. Formula supplemented with DHA and AA has produced positive changes in neurodevelopment measured in infancy in some, but not all studies. A meta-analysis demonstrated no improvement in visual acuity or neurodevelopment.⁷⁹ Because studies evaluating added DHA and AA have not been conclusive, the LSRO Report assessing nutrient requirements for preterm infant formulas recommended a minimum content of zero and a maximum concentration of DHA (0.35% of total fatty acids) and AA (0.6% of total fatty acids).

Cholesterol is a major component of cell membranes and a precursor in the synthesis of bile acids and some hormones. It is present in human milk in concentrations ranging from 10 to 15 mg/dL, although commercial formulas contain only trace amounts of cholesterol (approximately 1 to 2 mg/dL). The high cholesterol content of breast milk relative to formula is maintained at this level regardless of maternal diet. The groups that assessed the nutrient requirements for term and preterm infant formulas did not recommend addition of cholesterol to infant formulas because there was no convincing evidence of a beneficial short- or long-term effect of such an addition.45,59 Furthermore, there is no evidence that added cholesterol would be equivalent to the cholesterol in a human milk globule.

Carnitine mediates the transport of long-chain fatty acids into mitochondria for oxidation and the removal of short-chain fatty acids that accumulate in mitochondria. Preterm infants may be at risk for carnitine deficiency because they are heavily dependent on lipids as an energy source and because the plasma carnitine concentration of preterm infants is low, owing to limited endogenous synthetic ability. In preterm infants not receiving supplemental carnitine, plasma and tissue carnitine levels fall even in the presence of adequate precursor amino acid concentrations. Carnitine is found in human milk and is currently added to standard term and preterm formulas in amounts somewhat higher than in human milk.

Provision of Nutritional Support

The extremely premature neonate is born with glucose stores of only 200 kcal and loses 1% of body protein per day when provided with intravenous glucose alone. Consequently, extreme prematurity should be viewed as a nutritional emergency. A number of observational studies have described the influence of nutritional practices on growth and have found that differences in caloric and protein intake in the first weeks account for the largest difference in growth among premature infants.⁶⁰ In addition, there is a growing body of data that shows the association between early nutrient intake and growth and neurodevelopmental outcomes. This section reviews the basis of recommendations for nutritional support of premature infants. Practice decisions related to provision of early nutritional support provided to ELBW infants seem to be related to the perceived severity of illness of the infant by clinicians.²⁸ The use of standardized protocols for feeding (both parenteral and enteral) extremely premature infants can lessen variation in practice and improve outcomes.

Standardized Feeding Guidelines

Enteral feeding practices vary considerably among different centers.⁴⁶ The use of evidence-based standardized feeding guidelines has been shown to improve nutritional outcomes (such as time to reach full enteral feeds), reduce number of days on parenteral nutrition, and encourage growth. In addition, standardized feeding guidelines, regardless of the content of the guideline, have been shown to reduce the incidence of necrotizing enterocolitis in premature infants.^33,63

Evidence Supporting Early Parenteral Amino Acids

Numerous randomized clinical trials of amino acid dose and advancement strategy to evaluate short-term tolerance, protein balance, and safety have been conducted in premature infants. Despite differences in study populations and composition of amino acid solutions, all studies demonstrated positive nitrogen balance in response to parenteral amino acids and improved protein balance with higher amino acid intake.41,43,71,93,95 It is also important to point out that positive nitrogen balance was found despite low total caloric intake (approximately 50 kcal/kg per day). Consequently, the initial goal of limiting catabolism and preserving endogenous protein stores in premature infants can be accomplished with provision of as little as 1.0 to 1.5 g/kg per day of intravenous amino acids, whereas delivery of 3 g/kg per day of amino acids will result in net protein gain that approximates that of the reference fetus (see Figure 43-1).

Stable isotope techniques have also been used to evaluate the effect of amino acids on protein metabolism in premature infants. In these studies, stable isotope tracers of one or more essential amino acids are used to reflect whole body protein kinetics. Rivera and colleagues found that administration of 1.5 g/kg per day of amino acids (Aminosyn-PF with cysteine added) beginning on the first day of life improved protein balance as a result of increased protein synthesis (as reflected by leucine kinetics).⁷¹ Using similar techniques, van den Akker and colleagues demonstrated that protein anabolism produced by administration of early amino acids is accomplished through an increase in protein synthesis and not from a decrease in proteolysis (protein breakdown).⁹²

Other investigators have assessed the safety and efficacy of different doses of amino acids. Thureen and colleagues conducted a randomized trial of low (1 g/kg per day) versus high (3 g/kg per day) amino acid intake in infants with extremely low birth weights immediately after birth.⁸⁸ The higher amino acid intake produced significantly greater protein accretion. In an open-label trial, te Braake and colleagues randomized VLBW infants to two different parenteral amino acid regimens.⁸⁷ Infants in the intervention group received 2.4 g/kg per day of amino acids starting immediately after birth. Infants in the control group received glucose alone on the first day of life, with a stepwise increase in amino acid intake thereafter (1.2 g/kg on day 2 and 2.4 g/kg on days 3 and 4). Infants who received amino acids on the first day of life were found to have positive nitrogen balance without any major adverse effects, whereas infants in the control group were in negative nitrogen balance on day 2. Ibrahim and colleagues evaluated an even higher initial dose of intravenous amino acids, randomizing VLBW premature infants to either 2 g/kg or 3.5 g/kg per day of intravenous amino acids on the first day of life. Similar to the studies discussed previously, the higher amino acid dose resulted in greater improvement in nitrogen balance without evidence of adverse effects.⁴¹

The safety of administration of early intravenous amino acids, particularly for premature infants, has been established in a variety of studies. Normal plasma amino acid concentrations have been reported using TrophAmine.³⁹ Rivera and colleagues studied infants with very low birth weights given amino acids in the first days of life and found no abnormal elevations of plasma amino acids, BUN, or ammonia.⁷² Thureen and colleagues found that 3 g/kg per day of early amino acids appears to be as safe as 1 g/kg per day, based on BUN and plasma aminograms as indicators of acute amino acid toxicity.^61,88 Although most studies have not demonstrated any relationship between amino acid intake and BUN in the first days of life, two studies reported increased BUN levels in infants receiving higher amino acid intakes at 7 days of age.^12,20 As mentioned earlier, a significant proportion of the amino acids supplied to the developing fetus are oxidized and serve as a significant energy source for the fetus. Urea production is a byproduct of amino acid oxidation. In premature infants, rates of urea production are higher than in term neonates and adults, consistent with high rates of protein turnover and oxidation. Some investigators have even suggested that azotemia might be evidence of the effective utilization of amino acids as an energy supply rather than of protein intolerance, but this contention remains unproven. There are no data to support the need to advance amino acid intake slowly.

Although the initial goal of providing intravenous amino acids to premature infants to limit catabolism and preserve endogenous protein stores can be accomplished even if total caloric intake is low, ultimately, both protein and energy must be supplied in quantities sufficient to support optimal growth. Protein quantity is the primary determinant of protein accretion. This is true regardless of whether parenteral nutrition is used exclusively or as a bridge to full enteral feedings. In a series of studies, Zlotkin and colleagues evaluated the effect of intravenous energy and nitrogen intake on nitrogen retention.¹⁰⁴ At constant nitrogen intake, increasing nonprotein energy intake from 50 to 80 kcal/kg per day resulted in increased nitrogen retention and weight gain. However, at low energy intake (~50 kcal/kg per day), increasing nitrogen intake from 494 to 655 mg/kg per day (3 to 4 g/kg of amino acids per day) had no effect on nitrogen retention or weight gain. However, at higher energy intakes (~80 kcal/kg per day), the same increase in nitrogen intake resulted in a significant increase in the rate of both nitrogen retention and weight gain.

Many conditions and interventions commonly encountered in extremely premature infants are known to increase protein requirements. Underlying disease states such as sepsis or surgical stress increase catabolism and can negatively impact protein accretion. In addition medications such as systemic steroids, fentanyl, and insulin can also impact protein accretion. Dexamethasone has been shown to increase protein catabolism by increasing protein oxidation and proteolysis, resulting in decreased accretion of protein.⁹⁴

Intravenous Amino Acid Mixtures

The first parenteral amino acid solutions used in neonates were hydrolysates of fibrin or casein. Concerns about these first-generation solutions included high concentrations of glycine, glutamate, and aspartate; the presence of unwanted peptides; and high acidity. Reports of hyperammonemia and acidosis in the early 1970s were associated with the use of these first-generation solutions in neonates. Although amino acid solutions have been significantly modified, the perceived risks associated with the protein hydrolysates linger, contributing to the hesitancy by some clinicians to administer early parenteral amino acids.

The second generation of amino acid solutions consisted of crystalline amino acid mixtures (FreAmine III, Travasol, Aminosyn). The amino acid pattern of these mixtures reflects that of high-quality dietary proteins with large amounts of glycine and alanine, absence of glutamate and aspartate, and absence or poor solubility of tyrosine and cysteine.

The newest solutions include modifications of crystalline amino acids for use in pediatric patients. The currently available solutions include modifications of crystalline amino acids for use in pediatric and neonatal patients (Table 43-4). TrophAmine was originally formulated to match plasma amino acid concentrations of healthy term, breastfed infants; Premasol is identical in composition to TrophAmine. The composition of Primene, available outside the United States, was derived from fetal and neonatal cord blood concentrations. Both TrophAmine and Premasol supply a mixture of L-tyrosine and N-acetyltyrosine. The bioavailability of N-acetyltyrosine, however, has been questioned. Neither Aminosyn-PF nor Primene supplies a substantial amount of tyrosine. Cysteine is not supplied by most amino acid solutions because it is not stable for long periods of time in solution. However, cysteine hydrochloride can be added during the compounding process just prior to delivery of the solution.

TABLE 43-4

Composition of Parenteral Amino Acid Solutions *

	Aminosyn-PF	TrophAmine Premasol	Primene
Histidine	312	480	380
Isoleucine	760	820	670
Leucine	1200	1400	1000
Lysine	677	820	1100
Methionine	180	340	240
Phenylalanine	427	480	420
Threonine	512	420	370
Tryptophan	180	200	200
Valine	673	780	760
Alanine	698	540	800
Arginine	1227	1200	840
Proline	812	680	300
Serine	495	380	400
Taurine	70	25	60
Tyrosine	44	240^†	45
Glycine	385	360	400
Cysteine	—	<16	189
Glutamic Acid	820	500	1000
Aspartic Acid	527	320	600

^*Amino acid concentration in mg/dL; all amino acid mixtures shown are 10% solutions.

^†Mixture of L-tyrosine and N-acetyltyrosine.

It is no surprise that the ideal composition of intravenous amino acid mixtures is unknown. Although these solutions are widely used in the neonatal intensive care unit, normative data on plasma amino acid concentrations, particularly in ELBW infants, has not been established. Whether the goal should be to match amino acid concentrations of term, breastfed infants or some other standard is not known. Clearly, the ultimate goal is to achieve plasma amino acid concentrations in response to provision of parenteral nutrition that optimize both growth and neurodevelopment without toxicity. To optimize nutrition and growth, particularly in a premature infant, the requirements for specific amino acids need to be more precisely defined.

Several amino acids may be “conditionally essential” in premature infants. That is, the infant’s ability to synthesize these amino acids de novo may be less than needed for functional metabolic demands. Cysteine, tyrosine, and arginine are often considered conditionally essential amino acids for premature infants.

Tyrosine is not present in appreciable amounts in currently available amino acid solutions because of its low solubility. Snyderman found lower rates of weight gain, nitrogen retention, and plasma concentrations of tyrosine in premature infants given a tyrosine-deficient diet.⁸⁰ Tyrosine is synthesized endogenously from phenylalanine by phenylalanine hydroxylase. The activity of this enzyme in premature infants was thought to be inadequate for growth and nitrogen retention without tyrosine supplements. However, stable isotope studies have demonstrated active phenylalanine hydroxylation in very premature (26 weeks) and premature (32 weeks) infants.^21,24 Therefore, in the strictest sense, tyrosine is not an essential amino acid. However, it remains unclear whether enough tyrosine can be endogenously produced from phenylalanine in premature infants to support normal rates of protein accretion. N-acetyl tyrosine, although currently added to TrophAmine, is not highly bioavailable. Nonetheless, several studies provide indirect evidence that N-acetyl tyrosine improves protein accretion in preterm infants.^36,38 In addition, it is unclear whether premature infants can adequately catabolize tyrosine via oxidation by the enzymes tyrosine aminotransferase and 4-hydroxyphenylpyruvate dioxygenase. Inability to catabolize tyrosine can lead to transient neonatal tyrosinemia. Further studies are needed to better define premature infants’ ability to catabolize tyrosine and to determine whether an alternative source of tyrosine is needed in parenteral amino acid solutions.

Cysteine may be a conditionally essential amino acid for premature infants, but is not contained in currently available amino acid solutions. Some studies have shown that the fetal liver lacks the enzyme system to convert methionine into cysteine and that infants on a cysteine-free diet demonstrate impaired growth and low plasma cysteine levels. Other studies have demonstrated that there is enough cystathionase in extrahepatic tissues of the fetus and premature infant to synthesize cysteine when an adequate amount of methionine is provided. Studies using stable isotope techniques have demonstrated active endogenous cysteine synthesis in low birth weight infants.⁷⁶ Nevertheless, there is evidence to support that when cysteine hydrochloride supplements are added to parenteral nutrition, nitrogen retention is improved in premature infants.⁸¹ Further, the addition of cysteine hydrochloride improves the solubility of calcium and phosphorus in parenteral nutrition solutions. However, it is important to note that cysteine hydrochloride supplements can produce metabolic acidosis unless appropriately buffered with acetate.

Glutamine is one of the most abundant amino acids in both plasma and human milk, yet it is not supplied by currently available amino acid solutions because glutamine is unstable in aqueous solution. Glutamine is a major energy substrate for small intestinal mucosa, as proved by a high glutamine uptake from the lumen and from arterial blood during the newborn period in rats. Adding glutamine to the parenteral nutrition (PN) solutions of animals prevents atrophy of small intestinal mucosa and smooth muscle, improves the gut immune function, and reduces the incidence of fatty infiltration of the liver. Several studies suggest that parenteral glutamine supplementation is of benefit in selected populations of critically ill adults. However, a large, multicenter, randomized clinical trial of parenteral glutamine supplementation found that parenteral glutamine supplementation did not decrease mortality or the incidence of late-onset sepsis in ELBW infants.⁶⁵

Finally, taurine is synthesized endogenously from cysteine and is not part of structural protein. It is present in large concentrations in the retina and brain of the fetus, reaching a peak concentration at birth. When newborn nonhuman primates are fed taurine-deficient formula, growth is depressed, but this does not occur in human preterm infants, despite declining plasma and urine taurine levels. Nevertheless, there is some limited evidence that taurine supplementation might influence auditory brainstem evoked responses. Several pediatric IV amino acid solutions (TrophAmine, Primene, Aminosyn-PF) contain one to three times the amount of taurine found in breast milk.

It is important to note that currently available amino acid solutions have not been modified for more than 20 years and that none were designed specifically to meet the needs of extremely premature infants. Future research efforts should be directed at designing a fourth generation of amino acid solutions to optimize amino acid nutrition provided to the most vulnerable infants.

Only gold members can continue reading. Log In or Register to continue