Nutrition and Selected Disorders of the Gastrointestinal Tract

Total Parenteral Nutrition

The early postnatal period for the preterm infant represents a critical developmental stage with significant nutritional needs. In fact, it can be viewed as a nutritional emergency. Because these VLBW infants are unable to receive adequate nutrition enterally because of GI immaturity and critical illness, parenteral nutrition serves as a bridge. Over the past decade, clinicians have come to recognize the safety of and need for immediate provision of amino acids with the appropriate amount of energy to promote positive nitrogen balance and to minimize the protein deficit that would accumulate without it. Consideration of the correct timing to add lipids as well as the best dose are less well understood. Inadequate nutritional support, especially over the first days to weeks of life, may set the stage for postnatal growth faltering, which may have implications for long-term neurodevelopmental outcomes.

Amino Acids

The fetus receives a continuous supply of amino acids via active transport by the placenta. Animal studies have shown that fetal amino acid uptake is higher than that needed solely for protein accretion. When that supply of amino acids ends with preterm delivery, the infant immediately experiences a reduction in protein accretion that, unless addressed, will result in a significant deficit of protein by hospital discharge. Early institution of parenteral nutrition with sufficient amino acid content to meet fetal growth requirements (2.0–4.0 g/kg/day) is important to minimize such deficits, improve glucose tolerance, avoid metabolic shock, and prevent negative nitrogen balance. Even the administration of as little as 1 g/kg/day of amino acids has been shown to eliminate negative nitrogen balance, whereas 3 g/kg/day leads to protein accretion.

Numerous studies have shown the benefits of early amino acid administration in the VLBW infant. In addition to preventing the catabolism of body protein, the provision of amino acids in the first administered fluids smooths the metabolic transition to extrauterine life. Early amino acids may stimulate insulin secretion and improve glucose tolerance. A study from Rivera et al of early amino acid infusion in preterm infants showed that VLBW infants receiving early amino acids with glucose were able to receive more glucose (and consequently more energy) than infants receiving glucose alone. Mahaveer et al compared data from a clinical nutrition practice change (early initiation of amino acids at a higher dose) to historical experiences with glucose tolerance in infants less than 29 weeks. Cumulative protein intake (mean ± standard error of the mean) over the first 7 days was statistically higher in the change group (15.3 ± 0.4 g/kg vs. 11.8 ± 0.4). The proportion of infants that required insulin in the change group was half that of the historical group (26% vs. 53%). The mechanism for this finding is that hypoaminoacidemia leads to decreased insulin secretion and down-regulation of glucose transporters at the cell membrane, leading to a decrease in Na ⁺ , K ⁺ ATPase activity, and intracellular energy failure. Data from Radmacher et al showed that in the first 5 days of life, ELBW infants receiving earlier and higher doses of amino acids (∼3 g/kg/day) had lower glucose levels than those initiating amino acids at a lower dose. Glucose tolerance was enhanced with a higher dose of amino acids.

The predicted daily rate of protein accretion in a fetus at 70% of gestation (∼28 weeks) is around 2 g/kg. Studies have shown that infusions of glucose alone without amino acids will actually lead to a loss of body protein that is inversely proportional to gestational age. Fig. 5.2 compares the degree of protein loss in infants at gestational ages of 26 weeks, 32 weeks, and term when infused with glucose alone. How this protein loss is manifest in the 26-week 1-kg infant is shown in Fig. 5.3 . With body protein stores of around 88 g at birth and losses of around 1.5% each day without exogenous amino acids, this infant’s body protein pool shrinks by more than 10% in the first week of life. The comparable fetus in utero would be increasing body protein by around 2% each day or up to 15% for the week. The provision of amino acids at a dose as little as 1 to 1.5 g/kg/day can meet losses, but higher doses are necessary for protein accretion.

Studies have shown that early amino acid administration in VLBW infants can improve protein balance, minimize protein deficits, and improve growth. Initiating amino acids at doses of 2 to 3.5 g/kg/day within hours of birth has been shown to be safe. In short-term toxicity studies, acid–base status in VLBW infants receiving early amino acids has not been negatively affected. Studies have reported no rise in ammonia levels in infants receiving amino acids compared with those receiving glucose alone.

Many clinicians follow BUN levels as an indicator of amino acid tolerance. Multiple investigators have reported that BUN concentrations do not correlate with the dose of intravenous amino acids. Modest increases in BUN with no change in serum creatinine or other indicators of metabolic disturbance have been seen with early amino acid administration. BUN reflects not only intake but also acuity of illness, renal function, hepatic synthesis, and hydration status. Thus a comprehensive evaluation of the infant, not a single blood test, is necessary to determine a need to reduce amino acid intake in an infant with an “elevated” BUN.

Most amino acids formulations have not changed in over 25 years, and the optimal intake of individual parenteral amino acids for premature infants remains unknown. Pediatric amino acid solutions are more appropriate for the preterm infant than those designed for adults. However, the original safety data were compiled in the 1990s and may need to be revisited, as amino acid dosing was much lower, was usually initiated later, and did not include the higher doses being used today. Conditionally essential amino acids such as tyrosine, glutamine, and cysteine are unstable in current products and have been shown to be in low plasma concentrations in parenterally nourished infants.

Morgan et al conducted a single-center randomized controlled trial to compare a standard neonatal parenteral nutrition plan to one with enhanced macronutrient content. The protocol was followed for up to the first 28 days of life as long as parenteral nutrition was required. Clinical care (fluid management, enteral nutrition introduction, and biochemical monitoring) for both groups was by unit protocol. Parenteral nutrition was discontinued when enteral intake exceeded 75%. Plasma profiles were analyzed for 20 specific amino acids. Among the conditionally essential amino acids, there was little difference between groups, despite the higher total amino acid intake in the enhanced group. The median plasma levels of glutamine, arginine, and cysteine were all below reference values.

Lipids

Preterm birth leads to early termination of placental delivery of fatty acids. This coincides with a period of rapid brain growth and significant lipid accretion in VLBW infants. Coupled with scarce amounts of adipose tissue, these infants are fully dependent on postnatal nutritional strategies to meet their needs during this period because they are not able to form sufficient quantities of needed long-chain polyunsaturated fatty acids (LCPUFA) from precursor fatty acids ( Fig. 5.4 ). Intravenous lipid infusions serve as a source of LCPUFA, including the n-3 (linolenic LNA, 18:3n-3) and n-6 (linoleic LA, 18:2n-6) fatty acids, to prevent essential fatty acid deficiency (EFAD) and be a source of energy as a partial replacement for glucose. However, soybean oil–based emulsions are devoid of other important LCPUFAs, such as docosahexaenoic acid (DHA, 22:6n-3), eicosapentaenoic acid (EPA, 20:n-5), and arachidonic acid (ARA, 20:4n-6). The proper balance of lipid classes is critical, especially for the developing brain and retina, where these fatty acids serve as structural components of lipid-based cell structures. Soybean oil–based products, originally designed for use in adults, have been the most commonly used products in the United States, even in preterm infants. Recent research has shown the importance of providing a broader mix of LCPUFA to meet the needs of infants at critical windows of development.

Martin et al have shown the rapid decrease in DHA and ARA levels in preterm infants in the first postnatal week. This rapid loss of structural lipids cannot be corrected with soybean oil–based products. Research from multiple groups in both animals and humans has shown that deficiencies in LCPUFAs may have roles to play in the development of such neonatal complications as chronic lung disease, NEC, and retinopathy of prematurity (ROP).

When to start intravenous lipids is not universally agreed among clinicians, especially in critically ill VLBW infants with respiratory disease. Concerns related to potential adverse effects on gas exchange and bilirubin displacement from albumin because of increases in free fatty acids (FFA) have been raised. A number of studies have shown that with slow infusion rates over longer periods of time (up to 24 hours) and with careful monitoring of triglycerides, modest amounts of lipids can be infused in most infants, even in the first days of life, without adverse effects. In vitro, the displacement of bilirubin from binding sites on serum albumin by the increased FFA depends on the relative concentrations of all three compounds and may occur even with adequate metabolism of infused lipid. Andrew et al found no free bilirubin (B _f ) generated in vivo if the molar FFA/albumin remained less than 6. Adamkin et al also found acceptable FFA/albumin (<3) when lipids were infused over long periods (18–24 hours).

Amin et al conducted a prospective, nested study of bilirubin binding during intravenous lipid infusions in infants 24 to 33 weeks’ gestational age. Infants were stratified into two cohorts based on gestational age less than or equal to 28 or greater than 28 weeks. Intralipid 20% (IL; Fresenius Kabi, Uppsala, Sweden) was infused over 16 to 20 hours at a rate less than 0.15 g/kg/hour in increasing doses (0.5–3 g/kg/day) over the first 10 days of life. Triglycerides were monitored, and IL was held less than or equal to 2 g/kg/day in the presence of hyperbilirubinemia. Total serum bilirubin (TSB) and B _f were analyzed with each increase in IL; albumin was analyzed by the third postnatal day. The bilirubin/albumin equilibrium association binding constant ( K ) was calculated at IL intakes of 1.5, 2.0, 2.5, and 3.0 g/kg/day.

For infants greater than 28 weeks’ gestational age, TSB, K, and B _f did not change significantly as the IL dose increased to 2.5 or 3.0 g/kg/day. For infants 28 weeks or lower, the median differences in B _f and K at 1 gm/kg/day were not statistically significant compared with that in the more mature infants. However, as the dose increased above 1.5 g/kg/day, K was significantly lower ( P < .01), and B _f was significantly higher ( P <.05) in the more immature infants. Although there are no data that define the intake of lipid at which clinically significant bilirubin displacement occurs, these results suggest there may be a maturational effect.

Concerns about adverse effects of intravenous lipid on pulmonary function have been raised but have not generally been found to be an issue. For infants with significant respiratory disease, providing lipid at low dosages sufficient to prevent EFAD may be more prudent during the height of pulmonary disease, although it will not meet the needs of the preterm infant. It should only be a short period of time before lipids can be advanced.

The most commonly used lipid emulsions in the United States are Intralipid 20% (Fresenius Kabi USA, Lake Zurich, IL) and Liposyn III 20% (Hospira, Lake Forest, IL). These are 100% soybean oil and provide alpha-linolenic and linoleic acids, but no DHA, EPA, or ARA. Alternative lipid source emulsions have been developed, as evidence increasingly suggests that excessive polyunsaturated fatty acid (PUFA) and linoleic acid content may have harmful effects, especially regarding the inflammatory response and oxidative stress. Few large randomized trials are available to evaluate these newer products. Table 5.2 lists commercial products both in and outside the United States. Smoflipid, a mixed-oil product (soybean, medium-chain triglycerides [MCT], olive, and fish; Fresenius Kabi, Bad Homburg, Germany) was approved for use in adults by the U.S. Food and Drug Administration (FDA) in 2016. However, it comes with a warning regarding use in preterm infants. Two studies with Smoflipid have been published from Europe.

Tomsits et al randomized 60 premature infants to Intralipid or Smoflipid and stratified them into three birth weight groups. The primary safety parameter was serum triglyceride concentrations. Some 51 infants completed the study with mean durations of 11 and 10 days (study vs. control, respectively).

Triglyceride values were similar between groups at the beginning and end of the study period, showing a slight rise with lipid infusion but remaining within expected limits. No significant differences between groups were noted for vital signs, growth, occurrence of adverse events, and hematological and clinical chemistry lab results, with the exception of gamma-glutamyl transferase, which remained relatively constant in both groups through study day 8 but was significantly increased in the control group ( P < .05) at the end of treatment. Alpha-tocopherol levels were significantly increased in the study group compared with the control group ( P < .05). Red blood cell phospholipid fatty acid patterns were similar at baseline, but significant differences were seen at study termination for linoleic, alpha-linolenic, and EPA fatty acids that reflected the differences in lipid profile of the two emulsions: a significant increase in EPA and smaller decrease in DHA in the study group. Linoleic and alpha-linolenic acids significantly increased in the control group. Overall, Smoflipid was well tolerated.

Rayyan et al conducted a randomized trial of Smoflipid and Intralipid in 53 preterm infants less than 34 weeks’ gestation and stratified into three weight groups. Lipid solutions were infused 1 g/kg/day for 3 days and then increased to 2, 3, and 3.5 g/kg/day from day 6 on. The maximum infusion rate was 0.17 g/kg over 18 hours/day. Routine labs were collected as well as blood for fatty acid analysis (plasma and red blood cells). The primary safety parameter was change in serum triglycerides from baseline to study day 8.

There were no significant differences in nutrient intake or occurrence of adverse/serious events between groups. The rise in triglycerides was similar between groups (baseline to day 8). With the exception of total and LDL cholesterol, lipid parameters were similar between groups. Cholesterol values were higher in the study group but remained within expected ranges. In the control group, the direct bilirubin increased significantly from baseline to final observation ( P = .036). Plasma phospholipids were significantly different between groups at the end of the study period: total n-6 PUFA (control > study, P < .001), total n-3 PUFA (study > control P < .001), n-6/n-3 ratio (control > study, P < .001).

Omegaven (Fresenius Kabi, Bad Homburg, Germany) is 100% fish oil, approved for use in Germany, but only in expanded access protocols in the United States. Fish oil is rich in n-3 fatty acids, especial EPA and DHA. Gura et al showed that using a fish oil–based emulsion helped reverse cholestasis that developed in infants with short bowel syndrome after receiving a soybean-based emulsion for a prolonged period of time. The investigators noted the absence of EFAD, hypertriglyceridemia, coagulopathy, infections, or growth delay in these infants. Other investigators have noted a role for alternative lipid emulsions in preventing/ameliorating liver compromise related to use of soybean oil products.

Carnitine is an essential cofactor required for the transport of long-chain fatty acids (LCFA) across the mitochondrial membrane for beta-oxidation. Because of the limited carnitine reserves and low plasma carnitine levels in preterm infants, some investigators have recommended supplementing parenteral nutrition. A secondary analysis of metabolic profile data by Clark et al showed that supplementation with L -carnitine during parenteral nutrition significantly increased plasma free carnitine concentrations within the first week compared with infants not supplemented. During the transition to enteral nutrition, free carnitine content decreased but remained higher in the most immature infants who received supplementation. Thus carnitine supplementation is recommended only for VLBW infants who require prolonged parenteral nutrition (2–3 weeks). Dosage at 8 to 10 mg/kg/day has been used without observable side effects.

Parenteral Vitamins

Adequate supplies of vitamins are essential for normal growth and development. The optimal requirement for vitamins in neonates has not been determined, and little additional information has been developed in the past 20 years. Current recommendations are generally based on expert opinion. With commercially available products, dosing is generally done by adding one-third, two-thirds, or a full vial to total parenteral nutrition (TPN) based on weight groups (<1 kg or 1–3 kg) (see Table 5.3 ).

Parenteral vitamin preparations are exposed to light, oxygen, and the lipophilic surfaces of tubing materials during administration, which may reduce the amounts delivered to VLBW infants. For these reasons, vitamins are usually added to TPN shortly before infusing. Some nurseries cover the TPN bags and tubing with foil or opaque materials. Others choose to add vitamin preparations to the lipid emulsion. See Table 5.4 for recommended parenteral and enteral intakes.

Trace Minerals

As with vitamins, precise requirements for individual trace minerals remain unknown, especially in the low-birth-weight (LBW) infant who has inadequate stores and rapid growth demands. Parenteral preparations are designed to provide sufficient amounts to prevent deficiencies and to match in utero accretion rates. Contaminants such as aluminum and chromium require monitoring in cases of long-term parenteral nutrition. Copper levels may need to be monitored in infants with cholestasis to avoid toxicity. See Table 5.5 for parenteral and enteral recommended intakes.

Calcium and Phosphorus

Preterm infants require increased intakes of calcium and phosphorus for optimal bone mineralization. The fetus accretes calcium at a rate of around 104 to 125 mg/kg/day at 26 weeks’ gestation; phosphorus uptake is around 63 to 86 mg/kg/day. It is difficult to deliver adequate amounts of these mineral in conventional TPN solutions because of Ca/P precipitation. The addition of cysteine hydrochloride lowers the pH of the solution and improves solubility. Suggested parenteral intake of calcium is 60 to 80 mg/kg/day and phosphorus 45 to 60 mg/kg/day. See Table 5.5 and chapter on glucose calcium and magnesium.

Carbohydrates

In the fetus, glucose is the main carbohydrate used for fuel and is provided by the placenta at a rate of around 7 g/day in the last trimester of pregnancy. In the early postnatal period, carbohydrates continue to serve as the primary energy substrate for the preterm infant receiving parenteral nutrition, supplying metabolic fuel for all major organs, especially the brain.

The glucose infusion rate should maintain euglycemia. Depending on the degree of immaturity, 5% or 10% glucose is commonly used; higher concentrations may require a central line for infusion. Glucose intolerance, defined as an inability maintain euglycemia at glucose administration rates of less than 6 mg/kg/minute, is a frequent problem with VLBW infants, especially those less than 1000 g birth weight, and must be avoided. Endogenous glucose production is elevated in VLBW infants (8 mg/kg/minute) compared with term infants and adults. Additionally, high glucose production rates are found in VLBW infants who receive only glucose compared with those receiving glucose plus amino acids and/or lipids. Clinical experience with glucose intolerance suggests that glucose alone does not always suppress glucose production in these infants. Although not all of the contributing factors to hyperglycemia are entirely clear, it appears likely that persistent glucose production is the main cause, fueled by ongoing proteolysis that is not suppressed by physiologic concentrations of insulin. Hyperglycemia can also occur in the presence of nonoliguric hyperkalemia.

Strategies to treat early hyperglycemia include:

• 1.

  decreasing the glucose infusion rate until the hyperglycemia resolves;

• 2.

  administering parenteral amino acids, which results in lowering glucose concentrations, presumably by enhancing endogenous insulin secretion;

• 3.

  initiation of exogenous insulin at rates to control hyperglycemia ; and

• 4.

  infusion of insulin as a nutritional adjuvant (to control hyperglycemia and to increase nutrient uptake).

The first and third options limit the amount of energy available to the infant, and the last approach has been shown to lead to lactic acidemia. The second strategy is preferred because it more closely mimics the fetal circumstance and has been shown to work. The decline of specific amino acids at the time of birth may trigger the starvation response, which includes an increase in endogenous glucose production. By smoothing the metabolic transition to extrauterine life with early amino acid infusion, one may forestall the starvation response and improve glucose tolerance. With the current practice of earlier initiation of parenteral amino acids, the frequency of early hyperglycemia can be reduced.

Excessive glucose infusion may result in increased energy expenditure, oxygen consumption, serum osmolality, osmotic diuresis, and fat deposition. A study by Poindexter et al examined the effect of insulin using a hyperinsulinemic-euglycemic clamp in normoglycemic ELBW infants receiving only glucose. They reported significant increases in plasma lactate concentrations and metabolic acidosis. A study by Beardsall et al showed in a randomized trial in VLBW infants that 20% glucose coupled with insulin in the first 7 days of life led to increased episodes of hypoglycemia and increased mortality at 28 days compared with standard practice. Thus routine use of insulin is not recommended.

(For further details on glucose management, see Chapter 11 .)

Energy

Energy needs are dependent on age, weight, rate of growth, thermal environment, activity, hormonal activity, nature of feedings, and organ size and maturation ( Table 5.6 ). Measurement of a true basal metabolic rate requires a prolonged fast and cannot ethically be determined in VLBW infants. Therefore resting metabolic rate (RMR) is used to estimate energy needs, dietary-induced thermogenesis, minimum energy expended in activity, and the metabolic cost of growth. The metabolic rate increases during the first weeks of life from an RMR of 40 to 41 kcal/kg/day during the first week to 62 to 64 kcal/kg/day by the third week of life. The extra energy expenditure is primarily as a result of the energy cost of growth related to various synthetic processes. The metabolic rate of the nongrowing infant is approximately 51 kcal/kg/day, which includes 47 kcal/kg/day for basal metabolism and 4 kcal/kg/day for activity. Excessive energy intake may result in hyperglycemia, increased fat deposition, fatty liver, and other complications.

The contribution of activity to overall energy expenditure seems to be small, between 3 and 5 kcal/kg/day. Because of the large amount of time spent in the sleep state, energy expenditure in muscular activity in immature infants is relatively small in comparison to their resting metabolism. As infants mature, they become more active; therefore energy expenditure from activity increases.

The exposure of infants to a cold environment affects energy expenditure with small alterations in the thermal environment, making a significant contribution to energy expenditure. Infants nursed in an environment just below thermal neutrality increase energy expenditure by 7 to 8 kcal/kg/day; any handling adds to this energy loss. A daily increase of 10 kcal/kg/day should be allowed to cover incidental cold stress in the preterm infant. Infants who are intrauterine growth restricted, particularly the asymmetrical type, have a higher RMR on a per-kilogram body weight basis because of their relatively high proportion of metabolically active mass. Other factors that may increase metabolic rate include the effects of fever, sepsis, and surgery. The degree to which the infant’s energy requirements are increased is uncertain.

Caloric intake above maintenance contributes to growth. On average, for each 1-g increment in weight, approximately 4.5 kcals above maintenance energy needs are required. Therefore to attain the equivalent of the third-trimester intrauterine weight gain (10–15 g/kg/day), a metabolizable energy intake of approximately 45 to 70 kcal/kg/day above the 51 kcal/kg/day required for maintenance must be provided, or approximately 100 to 120 kcal/kg/day. Increasing metabolizable energy intakes beyond 120 kcal/kg/day with energy supplementation alone does not result in proportionate increases in weight gain. However, when energy, protein, vitamins, and minerals are all increased, weight gain with increases in rates of protein and fat accretion can be realized. The higher the caloric intake, the more energy that is expended through excretion, dietary-induced thermogenesis, and tissue synthesis. The energy cost of weight gain at 130 kcal/kg/day has been shown to be 3.0 kcal per gram of weight gain. However, at an intake of 149 kcal/kg/day and 181 kcal/kg/day, the energy cost of weight gain has been estimated to be 4.9 and 5.7 kcal/g of weight gain, respectively. In summary, to increase lean body mass accretion and limit fat mass deposition, an increase in protein-to-energy ratio in enteral diets is necessary.

The energy needs of the parenterally nourished infant differ from the enterally fed infant in that there is no fecal loss of nutrients. Preterm infants who are appropriately grown for gestational age are able to maintain positive nitrogen balance when receiving 50 nonprotein calories (NPCs)/kg/day and 2.5 g protein/kg/day. At an NPC intake of greater than 70 NPC/kg/day and a protein intake of 2.7 to 3.5 g/kg/day, preterm infants exhibit nitrogen accretion and growth rates similar to in utero levels. In the ELBW infant with minimal respiratory disease but requiring mechanical ventilation, energy expenditure may be as high as 85 kcal/kg/day in early postnatal life. There are few data that describe energy expenditure in these infants, as a result of technical difficulties and methodologic limitations affecting the interpretation of data.

The primary sources of energy for parenteral nutrition in infants are either glucose or lipid or a combination of the two. Although both glucose and fat provide equivalent nitrogen-sparing effects in the neonate, studies have demonstrated that a nutrient mixture using IV glucose and lipid as the nonprotein energy source is more physiologic than supplying glucose alone. The amount of glucose required to meet the total energy needs approximates 7 mg/kg/min (10 g/kg/day). Excess glucose is converted to fat or triglycerides. When lipids supply 60% to 63% of the NPCs given to LBW infants, nitrogen retention is decreased, and temperature control is adversely affected. A moderate IV fat intake, comprising approximately 35% of the NPCs, is preferred.

Practical hints for fluid and TPN management in the first few days of life:

• Provide sufficient fluid to result in a urine output of 1–3 mL/kg/hour, a urine specific gravity of 1.008 to 1.012, check for sugar and protein.

• Target a weight loss of 5% or less in full term or 15% or less in VLBW infants.

• Weigh infants twice daily in the first 2 days, then daily.

• Use birth weight to calculate intake until birth weight is regained.

• Record fluid intake, output, and weight.

• Use only 20% lipid emulsions. Infuse lipids slowly, not to exceed 0.17 g/kg/hour. If the infant is hyperbilirubinemic, limit lipids to 0.5 to 1.0 g/kg/day. The maximum lipid dose should be 3 to 4 gm/kg/day while maintaining serum triglyceride less than 200 mg/dL. Check before starting lipids, when lipids are advanced, and weekly thereafter;

• Aim to provide a parenteral energy intake of 90 to 100 kcal/kg/day and an amino acid dose up to 3.5 gm/kg/day. The NPC:N ratio should be 150 to 250.

• (Lipid calories + dextrose calories) ÷ (protein [grams] ∗ 0.16) = NPC/N (grams)

Carbohydrate

Carbohydrate provides 41% to 44% of the calories in human milk and most infant formulas. In human milk and standard infant formulas, it is present as lactose, which has been shown to enhance calcium absorption. Once the infant’s condition stabilizes, the requirement for carbohydrate is estimated at 40% to 50% of calories, or approximately 10 to 14 g/kg/day. In soy and other lactose-free formulas, the carbohydrate is in the form of sucrose, maltodextrins, and glucose polymers (corn syrup solids or modified starches). The three major disaccharidases responsible for the digestion of disaccharides are lactase, maltase, and sucrase-isomaltase. Maltase and sucrase-isomaltase first appear at 10 weeks’ gestation, reaching approximately 70% of newborn levels at 28 weeks. However, by 28 to 34 weeks’ gestation, lactase has only 30% of the activity found in the term infant. However, in clinical settings, lactose intolerance is rarely a problem. Human milk is usually well tolerated, possibly because preterm infants acquire a relatively efficient capacity to hydrolyze lactose in the small intestine at an earlier developmental stage that do infants in utero.

When lactose is not hydrolyzed in the small intestine, bacterial fermentation of the undigested portion occurs in the colon, producing short-chain fatty acids, which enhance mineral and water absorption and may stimulate growth and cell replication in the gut lumen. Thus colonic salvage is apparently important in disposal of unabsorbed lactose; however, its exact quantitative contribution remains unknown. Colonic bacterial fermentation of unabsorbed lactose to absorbable organic acids enables the infant to reclaim this carbohydrate energy and appears to prevent clinical symptoms of diarrhea.

Although pancreatic alpha-amylase, the major enzyme in starch hydrolysis, is either absent or in very low concentrations in the first 6 months of life, newborns are capable of tolerating small amounts of starch without side effects, and preterm infants are able to hydrolyze glucose polymers. Several enzymes may compensate for the physiologic pancreatic amylase deficiency in infancy. Glucoamylase, an enzyme found in the brush border of the small intestine, is present in the neonate in concentrations similar to those in adults. Also, salivary and human milk amylases may provide additional pathways for glucose polymer digestion in infancy.

Because lactase is found only at the tip of the villus, it is very sensitive to mucosal injury. Lactose intolerance may develop in infants with diarrhea, those suffering from undernutrition, or those recovering from NEC, necessitating temporary use of a lactose-free formula. In contrast, glucoamylase is able to survive partial intestinal atrophy because it is located at the base of the villi, thus enabling glucose polymers to be an alternative carbohydrate source when enteritis is present and lactase may be in low concentrations.

In premature infant formulas, lactose has been partially replaced by glucose polymers, polysaccharides with chains of 5 to 10 glucose residues joined linearly by 1, 4-alpha linkages to decrease the osmolality of the formula and to decrease the lactose load in the diet. Glucose polymers are well tolerated by preterm infants with glucose and insulin responses similar to those of a lactose feeding. Because glucose polymers add fewer osmotic particles to the formula per unit weight than does lactose, they permit the use of a high-carbohydrate formula with an osmolality less than 300 mOsm/kg of water. Special formulas for preterm infants contain approximately 40% to 50% lactose and 50% to 60% glucose polymers, a ratio that does not impair mineral absorption.

Protein

The protein requirement of the preterm infant is estimated to be 3.2 to 4.2 g/kg/day for VLBW infants and 3.5 to 4.4 g/kg/day for ELBW infants. One study suggests that in VLBW infants, a formula with higher protein content (3.6 g/100 kcal) versus 3.0 g/100 kcal in standard formula results in increased protein accretion and weight gain without evidence of metabolic stress. The quality and quantity of protein that the infant receives are important. Although weight gain and growth of VLBW infants fed protein intakes of 2.2 to 4.5 g/kg/day of either a casein- or whey-predominant formula have been shown to be no different from those receiving pooled human milk, the metabolic responses can be significantly different. Serum BUN, ammonia, albumin, and plasma methionine and cysteine concentrations were higher in the infants receiving high-protein formulas. Elevated levels of phenylalanine and tyrosine were seen in infants fed the casein-predominant, high-protein formula, and lower concentrations of taurine were noted in infants fed casein-predominant formulas regardless of quantity. Preterm infants fed soy protein formula supplemented with methionine exhibit slower weight gain and lower serum protein and albumin concentrations than infants fed a whey-predominant formula. Thus premature infant formulas are whey predominant with a 60 to 40 whey-to-casein ratio; soy protein-based formulas are not recommended for the preterm infant.

Human milk is considered to have the ideal amino acid distribution for the human infant. Preterm infants fed their OMM have more rapid growth than infants fed pooled, banked human milk with accretion of protein and fat similar to that of the fetus. Human milk is lower in mineral content, especially magnesium, calcium, phosphorus, sodium, chloride, and iron. To attain intrauterine growth rates in larger preterms, large volumes (180–200 mL/kg/day) of human milk must be fed. Fortification of human milk is discussed below.

The recommended dietary allowances for infants and children to age 12 months are listed in Table 5.7 .

Lipids

Fat is a major source of energy for the infant, with approximately 50% of the calories in human milk derived from fat. In commercial formulas, fat provides 40% to 50% of the energy. These feedings provide 5 to 7 g of fat per kg per day. The saturated fat of human milk is well absorbed by the preterm infant, in part because of the distribution pattern of fatty acids on the triglyceride molecule. Palmitic acid is present in the beta position in human milk fat and more easily absorbed than palmitic acid in the alpha position, which occurs in cow milk.

Human milk contains a bile salt–activated lipase that enhances lipid digestion in the duodenum. The special formulas for preterm infants contain a mixture of medium-chain triglycerides and vegetable oils rich in polyunsaturated long-chain triglycerides (LCTs), both of which are well absorbed by preterm infants.

MCTs are oils with an 8 to 12 carbon chain length. Unlike LCTs, MCTs do not require bile for emulsification. MCTs are rapidly hydrolyzed in the gut and pass directly to the liver through the portal circulation, whereas LCFAs must be re-esterified once absorbed and are transported via the lymph system into the blood circulation, where they are hydrolyzed by lipoprotein lipase. Medium-chain fatty acid (MCFA) metabolism differs from that of LCFA in that it does not require carnitine for transport into the mitochondria and is not regulated by cytosolic acyl-CoA synthetase. MCFAs enter the mitochondria directly and are rapidly oxidized. Formulas with MCTs have been shown to improve nitrogen, calcium, and magnesium absorption. Preterm infant formulas have been developed with approximately half the fat as MCTs.

Vitamins

Calcium, Phosphorus, Magnesium, and Vitamin D

The amount of enteral calcium, phosphorus, and magnesium intake required to match intrauterine accretion rates is high: calcium 185 to 200 mg/kg/day, phosphorus 100 to 113 mg/kg/day, and magnesium 5.3 to 6.1 mg/kg/day. VLBW infants with minimal illness may require lower intakes. The AAP recommends calcium intakes of 185 to 210 mg/kg/day, phosphorus 123 to 140 mg/kg/day, and magnesium 8.5 to 10 mg/kg/day. However, magnesium intake at this level with such high calcium and phosphorus intake results in negative magnesium balance. Therefore a higher intake of magnesium, approximately 20 mg/kg/day, may be needed.

The recommendation for vitamin D, which is required for normal metabolism of calcium, phosphorus, and magnesium, has ranged from 200 to 2000 IU per day for the preterm infant. VLBW infants can maintain normal vitamin D status with 400 IU/day; high-dose vitamin D supplementation does not decrease the incidence of rickets in VLBW infants.

Human milk has concentrations of calcium and phosphorus that are appropriate for full-term infants but not for the VLBW infant. Breast milk should be supplemented with additional calcium, phosphorus, and vitamin D, which can easily be done with human milk fortifiers. Fortification yields better mineral accretion than breast milk alone, similar to that of VLBW infants fed a premature infant formula.

Inadequate intakes of calcium, phosphorus, and vitamin D result in metabolic bone disease of prematurity, also called rickets of prematurity. This disease is characterized by reduced bone mineralization and, in severe cases, frank radiologic evidence of rickets and spontaneous fractures. The biochemical findings, although not highly sensitive, include an elevated alkaline phosphatase (>500 IU/L), decreased serum phosphorus (<4 mg/dL), and normal serum calcium. The 25-hydroxycholecalciferol (25-OH vitamin D) level is usually normal, but 1,25 dihydroxycholecalciferol (1,25-OH vitamin D) levels may be elevated as a result of increased parathyroid hormone levels and low serum phosphorus levels. The incidence of rickets was higher before institution of the current practices of higher calcium and phosphorus levels in parenteral nutrition solutions and early enteral feedings. The etiology of rickets remains unclear but is thought to be primarily because of an insufficient intake of calcium and phosphorus. Risk factors for rickets are listed in Box 5.3 . Confirming the diagnosis requires radiologic evidence of osteopenia.

Fortified human milk or premature infant formula is the preferred feeding for VLBW infants. The use of soy formulas is not recommended for infants with birth weights less than 1800 g. If continuous infusion feeding of human milk is necessary, the syringe and the pump should be placed upright to prevent loss of calcium, phosphorus, and milk fat by separation and adherence to the tubing.

Vitamin B ₁₂ and Folate

Vitamin B ₁₂ requires intrinsic factor for its absorption in the distal ileum. Therefore particular attention to this vitamin is necessary in infants who have had gastric resection or resection of the terminal ileum (e.g., NEC surgery). The potential neurologic complications of vitamin B ₁₂ deficiency are irreversible.

Serum folate levels may be low in the preterm infant. Folate is supplemented in the pediatric IV multivitamin preparation and in infant formulas. It is not available in the infant multivitamin drops because of its instability in the liquid form. Folate plays an important role in DNA synthesis. Deficiency of this vitamin may result in megaloblastic anemia, neutropenia, thrombocytopenia, and growth failure. Requirements for fat- and water-soluble vitamins in VLBW and ELBW infants are shown in Table 5.7 . Advisable intakes for infants 0 to 12 months are shown in Table 5.8 .

Iron

There has been increased interest in iron deficiency, with data suggesting that mental and developmental test scores are lower in infants with iron deficiency anemia and that iron therapy sufficient to correct the anemia is insufficient to reverse the behavioral and developmental disorders in many infants. This indicates that certain ill effects are persistent depending on the timing, severity, and/or degree of iron deficiency anemia during infancy.

Iron stores in the preterm infant are lower than in the term baby because iron stores are relatively proportional to body weight. Iron depletion occurs around the time the baby doubles her/his birth weight, and thus iron therapy should begin by 4 weeks of life in the preterm infant when enteral feedings are tolerated. Smaller preterm infants may need as much as 4 to 6 mg/kg/day, with about 2 mg/kg/day provided by iron-fortified formula and the remainder as iron supplementation at 2 to 4 mg/kg/day. A higher dose is also necessary for infants being given erythropoietin. Oral iron supplementation can interfere with vitamin E metabolism in the VLBW infant, thereby further increasing the need for vitamin E in an infant who is at risk for low serum tocopherol levels. Although premature infant formulas, both with and without iron fortification, are manufactured with ample amounts of vitamin E and a PUFA-to-E ratio of 6 or greater, premature infants on human milk and receiving supplemental iron can be also supplemented with 4 to 5 mg (6–8 IU) of vitamin E per day. This can be readily accomplished by use of an oral multivitamin with iron.

The impression that low-iron formulas are associated with fewer GI disturbances is not supported by controlled studies. Because the bioavailability of iron from iron-fortified infant cereals is somewhat low, it is recommended that iron-fortified formulas or daily iron supplements be continued through the first year of life.

Among term infants, breast feeding usually provides adequate iron intake during the first 4 to 6 months of life, and supplementation during this time is not necessary. Although the iron content of human milk is low, averaging 0.8 mg iron/L, the bioavailability is high, with term infants absorbing about 49% of the iron content compared with 10% to 12% from iron-fortified cow milk formula. Infants who are exclusively breast fed can maintain normal hemoglobin and ferritin levels, and do not need iron supplementation until 4 to 6 months. See Table 5.5 .

Fluoride

Because of reports of dental fluorosis in infants and toddlers, fluoride supplementation is no longer recommended in the infant younger than 6 months of age. The supplementation schedule ( Table 5.9 ) recommended by the AAP and the American Dental Association should be followed according to the fluoride content of the local water supply.