parenteral nutrition in infants and children

21
Invited Review Parenteral Nutrition in Infants and Children Robert J. Shulman and Sarah Phillips Department of Pediatrics, Baylor College of Medicine, USDA/ARS Children’s Nutrition Research Center, Texas Children’s Hospital, Houston, Texas, U.S.A. JPGN 36:587–607, 2003. Key Words: Parenteral nutrition— Intravenous—Feeding—Neonates—Energy—Protein—Amino acids—Lipid emulsion. © 2003 Lippincott Williams & Wilkins, Inc. INTRODUCTION Parenteral nutrition (PN) came of age in 1964 with the demonstration that beagle puppies could be nourished successfully from 12 weeks of age to maturity by pro- viding all nutrients intravenously (1). The first total par- enteral nutrition of an infant with extreme short bowel syndrome followed in 1967 (1). Since them many lessons have been learned as a result of complications of PN. These have included nutrient deficiencies and excesses, infections, complications of inadequate or excessive en- ergy and protein intake, liver disease, and toxicities from product contamination. Our patients have paid the price for these lessons, but fortunately improved survival has also resulted. These incidents have reinforced to us the physicians the axiom that “good judgment comes from experience . . . and experience comes from bad judg- ment.” (2). Along the way we have learned much about nutrition, infection, liver pathophysiology, and child de- velopment (e.g., yes, infants really can “unlearn” how to suck and swallow). Have we learned so much that ad- ministration of PN can now be placed on autopilot? The question posed to the authors when this review was solicited was whether the art and science of PN had reached the stage where administration could be treated as a routine clinical algorithm such as diarrhea or croup. Do we only need to check the weight, calculate the ad- ministration rate and check the box next to “Standard Child Solution” on the PN pharmacy order sheet? Is there really anything new in PN? Our immediate reaction was to take offense, feel hurt, and shake our heads in disbelief that anyone could be so foolish as to think such things. However, as we thought about it, we realized that the question is pertinent and perceptive. Like the clinical pathways set out for com- mon illnesses such as diarrhea and croup, the question is not so much how to follow the roadmap but rather, when not to follow it. Thus, our review will focus on two areas: 1) Common misconceptions surrounding the use of PN; and 2) When should “standard PN” (i.e., checking the weight and the box on the order sheet) not be used. These questions have lead us to review recent developments in PN (in the past 5–7 years). Whenever possible we will take an evidenced-based approach to the data. Unless specified, the studies we will discuss were carried out in pediatric patients. The risk of any review is that it is old hat to some and very new to others. We have done our best to strike a middle ground. Because it could be a subject unto itself, we will not review PN-associated cholestasis, although when pertinent, some comments will be made. Energy Requirements, Energy Sources, and the Consequences of Overfeeding Estimating the appropriate energy intake is a funda- mental step in prescribing PN. Although we have long acknowledged the risks of underfeeding our patients, more recently we have come to reevaluate energy re- quirements in particular clinical scenarios and to appre- ciate the problems associated with excessive energy in- take. Received November 25, 2002; accepted February 13, 2003. Address correspondence to Robert J. Shulman, 1100 Bates St., Hous- ton, Texas 77030, U.S.A. (e-mail: [email protected]). Journal of Pediatric Gastroenterology and Nutrition 36:587–607 © May 2003 Lippincott Williams & Wilkins, Inc., Philadelphia 587

Upload: others

Post on 12-Sep-2021

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Parenteral Nutrition in Infants and Children

Invited Review

Parenteral Nutrition in Infants and Children

Robert J. Shulman and Sarah Phillips

Department of Pediatrics, Baylor College of Medicine, USDA/ARS Children’s Nutrition Research Center, Texas Children’sHospital, Houston, Texas, U.S.A.

JPGN 36:587–607, 2003. Key Words: Parenteral nutrition—Intravenous—Feeding—Neonates—Energy—Protein—Amino

acids—Lipid emulsion. © 2003 Lippincott Williams &Wilkins, Inc.

INTRODUCTION

Parenteral nutrition (PN) came of age in 1964 with thedemonstration that beagle puppies could be nourishedsuccessfully from 12 weeks of age to maturity by pro-viding all nutrients intravenously (1). The first total par-enteral nutrition of an infant with extreme short bowelsyndrome followed in 1967 (1). Since them many lessonshave been learned as a result of complications of PN.These have included nutrient deficiencies and excesses,infections, complications of inadequate or excessive en-ergy and protein intake, liver disease, and toxicities fromproduct contamination. Our patients have paid the pricefor these lessons, but fortunately improved survival hasalso resulted. These incidents have reinforced to us thephysicians the axiom that “good judgment comes fromexperience . . . and experience comes from bad judg-ment.” (2). Along the way we have learned much aboutnutrition, infection, liver pathophysiology, and child de-velopment (e.g., yes, infants really can “unlearn” how tosuck and swallow). Have we learned so much that ad-ministration of PN can now be placed on autopilot?

The question posed to the authors when this reviewwas solicited was whether the art and science of PN hadreached the stage where administration could be treatedas a routine clinical algorithm such as diarrhea or croup.Do we only need to check the weight, calculate the ad-ministration rate and check the box next to “StandardChild Solution” on the PN pharmacy order sheet? Isthere really anything new in PN?

Our immediate reaction was to take offense, feel hurt,and shake our heads in disbelief that anyone could be sofoolish as to think such things. However, as we thoughtabout it, we realized that the question is pertinent andperceptive. Like the clinical pathways set out for com-mon illnesses such as diarrhea and croup, the question isnot so much how to follow the roadmap but rather, whennot to follow it. Thus, our review will focus on two areas:1) Common misconceptions surrounding the use of PN;and 2) When should “standard PN” (i.e., checking theweight and the box on the order sheet) not be used. Thesequestions have lead us to review recent developments inPN (in the past 5–7 years). Whenever possible we willtake an evidenced-based approach to the data. Unlessspecified, the studies we will discuss were carried out inpediatric patients. The risk of any review is that it is oldhat to some and very new to others. We have done ourbest to strike a middle ground. Because it could be asubject unto itself, we will not review PN-associatedcholestasis, although when pertinent, some commentswill be made.

Energy Requirements, Energy Sources, and theConsequences of Overfeeding

Estimating the appropriate energy intake is a funda-mental step in prescribing PN. Although we have longacknowledged the risks of underfeeding our patients,more recently we have come to reevaluate energy re-quirements in particular clinical scenarios and to appre-ciate the problems associated with excessive energy in-take.

Received November 25, 2002; accepted February 13, 2003.Address correspondence to Robert J. Shulman, 1100 Bates St., Hous-

ton, Texas 77030, U.S.A. (e-mail: [email protected]).

Journal of Pediatric Gastroenterology and Nutrition36:587–607 © May 2003 Lippincott Williams & Wilkins, Inc., Philadelphia

587

Page 2: Parenteral Nutrition in Infants and Children

Energy Requirements

A few terms must be defined in order to discuss energyexpenditure (3). Basal metabolic rate is the energy ex-penditure of a recumbent child or adult in a thermoneu-tral environment after a 12- to 18-hour fast just when theindividual has awakened but before daily activities havecommenced. Basal metabolic rate is a reflection of theenergy expenditure required for vital processes. Restingenergy expenditure refers to the energy expenditure of aperson at rest in a thermoneutral environment. Basalmetabolic rate and resting energy expenditure usually donot differ by more than 10% (3). Total energy expendi-ture is the sum of requirements for basal metabolism,thermic effect of ingested food, thermoregulation, andactivity. In older children, activity accounts for a largeproportion of total energy expenditure. Thus, the totalenergy expenditure of a child who is hospitalized andlying in bed is reduced.

Historically, it has been thought that the surgical pa-tient requires an energy intake proportional to the sever-ity of the illness. By definition, the energy intake wouldbe increased compared with that of a “non-stressed” pa-tient. However, from a practical standpoint this conceptdoes not seem to be the case as the increased energyexpenditure is short-lived. For example, although new-borns have a 20% increase in resting energy expenditureafter major surgery, this elevation returns to baselinewithin 12-24 hours (4). Infants who remain critically illand require PN also do not appear to require increasedenergy intakes. Jaksic et al. measured energy expenditurein eight non-ventilated surgical neonates (gastroschisis,atresia, volvulus) on postoperative day 16 (± 12, SD) andcompared them with ten infants on extracorporeal lifesupport studied at 7 ± 3 days of age (5). There were nodifferences in energy expenditure between the groups(53 ± 5 vs. 55 ± 20 kcal � kg−1 � d−1) (5). This level ofenergy expenditure is comparable to that of “non-stressed” infants (see below). The similarity in energyexpenditure was present despite the finding that IL-6 andC-reactive protein levels were significantly greater in theextracorporeal life support group reflecting an increaseddegree of illness, and ultimately, mortality (30% vs. 0)(5). This same group of investigators recently has shownthat preterm infants (25 weeks gestation) behave simi-larly (6).

Lloyd recently reviewed the data regarding energy re-quirements in surgical newborns and children (7). Theevidence reveals that the increase in energy expenditureassociated with surgery only lasts for 24 hours after theprocedure (7). Estimating energy expenditure during thefirst 24 hours after surgery will overestimate the energyrequirements for the entire postoperative period, poten-tially resulting in excess energy intake with its potentialconsequences (see below).

Energy requirements also have been examined in non-operated children in the intensive care unit. In a group of

mechanically ventilated critically ill children (N � 33; 6± 5 years of age) whose mean length of stay in theintensive care unit was approximately 2 weeks, measuredenergy expenditure was only about 8 kcal � kg−1 � d−1

(17%) above expected based on more recent data (8,9).Joosten et al. found even smaller increases in energyexpenditure in a more heterogeneous group of 36 infantsand children in an intensive care unit (10). Finally, Turiet al. measured energy expenditure in 21 patients in theintensive care unit with systemic inflammatory responsesyndrome or sepsis and compared their energy expendi-tures to a group of hospitalized control children (11). Nodifferences were noted between the groups (11).

Taken together, these data support the idea that inmost cases there is little need to provide post operative orcritically ill infants and children with much more thantheir resting energy expenditure, which in most caseswill be an amount similar to their basal metabolic rate.The information in Table 1 can be used to calculate basalmetabolic rate in normal children (9). A recent report byDuro et al. used the Enhanced Metabolic Testing Activ-ity Chamber (EMTAC) to predict REE in 50 normalinfants up to 7 months of age with a maximum weight of9 kg (12). Their data suggest that for this age group thevalues of Shofield may be an underestimation (9,12).However, part of this difference may be explained by thefact that the values for Shofield are for BMR and thoseof Duro et al. are for REE (9,12). The equations de-scribed by Duro et al. are:

REE (kcal/d) � [84.5 x Weight (kg)] – 117.33(R2 � 0.65, P < 0.01)

The prediction is improved somewhat by includinglength:

REE (kcal/d) � [10.12 x Length (cm)]+ [61.02 x Weight (kg)] – 605.08

(R2 � 0.7, P < 0.01)

It is evident in the study by Jaksic et al. that there maybe significant interindividual variations in energy expen-diture (5,7). This should not be surprising, as evenamong normal newborns energy requirements do not fallwithin narrow margins (13). White et al. have shown thatin pediatric intensive care unit patients that there is littlewithin-day variation in energy expenditure but day-to-day variation is as high as 21 ± 16% (mean ± SD) (14).

Ideally, energy expenditure should be measured incritically ill patients. However, it may not be practicalbecause of the expensive equipment and the expertiserequired. At the very least, one should be flexible regard-ing the “appropriate” energy intake and monitor the pa-tient for weight gain (fluid vs. tissue) and evidence ofoverfeeding (e.g., CO2 production; see below). Two re-cent studies may be of value in calculating resting energyexpenditure for surgical infants and intensive care pa-tients.

R. J. SHULMAN AND S. PHILLIPS588

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 3: Parenteral Nutrition in Infants and Children

Using indirect calorimetry and a number of covariatesin a group of surgical infants, Pierro et al. developed theformula shown below to calculate energy expenditure(15,16).

Resting energy expenditure (kcal/d) �−74.436 +[34.661 x weight (kg)]+ [0.496 x heart rate (beats/min)]+ [0.178 x age (days)] x [1.44]

The prediction has an error of – 0.95 ± 3.65%; an R2

� 0.85; and a P < 0.00001)White et al devised the following formula for estimat-

ing energy expenditure in intensive care patients (17):

Energy expenditure (kcal/day) �[(17 × age in months) + (48 × weight in kg)+ (292 × body temperature in °C) – 9677]

x [0.239] (R2 � 0.867)

Obese adolescents are an exception to the basal meta-bolic rate calculations provided in Table 1. The basalmetabolic rate for obese children can be calculated asfollows (18):

For Boys: BMR �[16.6 x weight (kg)] + [77 x height (meters)] + 572

For Girls: BMR �[7.4 x weight (kg)] + [482 x height (m)] + 217

The clinical significance of energy expenditure, par-ticularly in the critically ill patient, also has an impact onthe timing of PN initiation. If very ill patients are goingto be “hypermetabolic”, available data suggest that it willbe early in the disease process. PN administration itselfincreases metabolic rate (19,20). The more malnourishedthe patient, the greater the stimulatory effect of PN onmetabolic rate (20). Thus, administration of PN whichprovides all nutritional requirements should be avoidedin the first couple of days of a patient’s critical illness, asit will further increase the metabolic rate and do little tostem the protein catabolic response to the illness. An

exception is the preterm infant. PN is required after birthbecause of the premature infant’s limited nutritionalstores (21). A recent randomized controlled study com-pared preterm infants (27 weeks gestational age, N �125) receiving PN on the first day of life to those inwhom it was started in the first few days of life andadvanced more slowly (22). Infants in the early PN grouphad fewer infections during hospitalization and weremore likely to be above the 10th percentile for weight orheight at the time of discharge (22).

Energy Sources

There is controversy around the choice of nutrients toprovide energy requirements via PN. There are glucoseadvocates, lipid believers, and argument as to whetherprotein (i.e., amino acids) should be counted as part ofthe total administered energy. At the expense of over-simplification and wounding the true believers in eachcamp (the authors are active participants in these contro-versies), we submit that a few basic principles can beagreed upon. 1) There is a minimum amount of glucosethat must be provided in order to prevent hypoglycemiaand a maximum that results in the production of exces-sive CO2 and/or hepatic steatosis (see below); 2) There isa minimum requirement (both dose and frequency) forintravenous fat emulsion to prevent essential fatty aciddeficiency but a maximum beyond which intravenous fatmay have deleterious effects; and 3) Amino acids mustbe provided in adequate amounts to prevent hypoprotein-emia but there are adverse consequences of giving anexcess. Both the size and the age of the pediatric patientare important in determining the appropriate quantities ofglucose, fat and amino acids administered in PN.

Glucose

Estimates of glucose utilization by the brain are shownin Table 2 (23). These estimates vary with age and, atfirst glance, reflect the minimum amount of glucose thatmust be provided to prevent hypoglycemia (23). Someinvestigators have argued that gluconeogenesis providesa significant amount of glucose (even in preterm infants)and suggest that not all the glucose shown in Table 2need be provided exogenously (24,25). However, these

TABLE 1. Equations for estimating basal metabolicrate (MJ/day)*

Age(years) Male Female

<3 0.249 wt − 0.127 0.244 wt − 0.130<3 0.0007 wt + 6.349 L − 2.584 0.068 wt + 4.281 L − 1.730

3–10 0.095 wt + 2.110 0.085 wt + 2.03310–18 0.074 wt + 2.754 0.056 wt + 2.89818–30 0.063 wt + 2.896 0.062 wt + 2.036

* Notes1. To convert to kcal/day multiply by 2392. Wt � weight in kilograms3. L � Length in meters. Only predictions for children <3 years of

age are improved by including length in the equation.From (9).

TABLE 2. Estimates of glucose consumption by the brain

Age

Utilization

(mg � kg−1 � min−1) (g � kg−1 � d−1)

Newborn 8.0 11.51 year 7.0 10.15 years 4.7 6.8Adolescent 1.9 2.7Adult 1.0 1.4

From (23).

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 589

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 4: Parenteral Nutrition in Infants and Children

studies have used isotopic tracers to measure gluconeo-genesis and it has been shown that this method underes-timates the contribution of glucose to CO2 when com-pared with data obtained from indirect calorimetry (23).Thus, the data in Table 2 probably do reflect the amountof exogenous glucose required to prevent hypoglycemiaunder most circumstances (23). Kalhan and Kiliç re-viewed the evidence that this amount of exogenous glu-cose also is sufficient to minimize nitrogen loss (23).

What about the maximum rate of glucose oxidation?Lafeber et al. found a glucose oxidation rate of 6.6 ± 1.2mg � kg−1 � min−1 (9.5 g � kg−1 � d−1) in appropriate forgestational age preterm infants (birth weight: 1613 ± 151g, gestational age: 31.1 ± 1.5 wk; mean ± SD)(26). Therate was minimally greater than that found in small forgestational age infants (26). In contrast, in term surgicalinfants Jones et al. measured a maximal glucose oxida-tion rate of 12.5 mg � kg−1 � min−1 (18 g � kg−1 � d−1)(27). This rate is comparable to that of stable patientsboth on long-term PN (N � 7, 30 ± 41 months of age)or short-term PN (N � 36, 6 ± 4 months of age) (28,29).

Clinical status can modify glucose oxidative capacitysignificantly. For example, Sheridan et al. noted that incritically burned children the maximal rate of glucoseoxidation was 5 mg � kg−1 � min−1 (30).

Consequences of Overfeeding with Glucose

In critically ill individuals, glucose intakes above themaximal glucose oxidation rate will result in the non-oxidative production of fat, hence, are unlikely to en-hance energy balance, or even reduce protein catabolism(23). On the other hand, glucose intakes above the maxi-mal oxidation rate promote fat deposition, which may bea nutritional goal in specific clinical situations (e.g., pre-term infants). We will use the term overfeeding to meanproviding calories in excess of the amount required fornormal weight gain.

A number of recent reviews have addressed issuesrelated to carbohydrate overfeeding in the intensive careunit patient (31–35). Unfortunately, the relationship be-tween carbohydrate feeding and impaired pulmonaryfunction (in the form of increased CO2 production) aswell as the development of fatty liver often is oversim-plified.

When glucose is administered in excess of the amountthat can be directly oxidized for energy production andglycogen production, the excess is directed to fat syn-thesis (lipogenesis) (Fig. 1) (36,37). This conversion isinefficient and probably accounts, in part, for the in-crease in energy expenditure seen with high rates of glu-cose infusion (Fig. 1) (20). This inefficiency can haveclinical consequences besides simply producing hyper-glycemia.

Nose et al. studied energy expenditure and CO2 pro-duction in 7 infants receiving long term PN (28). Three

PN regimens were tested: 87:5:8 (glucose:fat:amino ac-ids as percent of total energy), 60:32:8, and 34:58:8 Totalenergy intake was comparable to the recommended dailyallowance. The increase in resting energy expenditure(although reported as basal metabolic rate) induced bythe high glucose regimen was nearly 5 times greater thanthat of the other two regimens (28). Bresson et al. alsonoted increased energy expenditure in relationship to in-creasing glucose intake, albeit to a lesser extent, amount-ing to 16% of the energy value of infused glucose (29).

In the study by Nose et al. the respiratory quotient (RQ� CO2 production/O2 consumption) was significantlygreater on a high glucose regimen compared to lowerglucose regimens (28). The data suggest that in thispopulation of stable patients on long-term PN, glucoseintakes in excess of 10 mg � kg−1 � min−1 result inconversion of glucose to fat (RQ >1) (28). As seen inFigure 1, lipogenesis from glucose results in a large in-crease CO2 production relative to O2 consumption (i.e.,high RQ).

Talpers et al. (among others) examined the effect ofdifferent glucose/fat ratios on CO2 production in adults(19,38). In contrast to the studies of Nose et al (and otherinvestigators), they were unable to detect a change inCO2 production as a consequence of alterations in theglucose/fat ratios (19,38). It is possible that the discrep-ancy between these studies is a result of the higher pro-portion of total energy provided by amino acids in Talp-ers’ study (20% of total energy) than in Nose’s study (8%of total energy) (28,38). Amino acids also can affectventilatory drive and the response to CO2. Increasingamino acid intake (with energy intake constant) leads toan increase in minute ventilation and more importantly,an enhanced response to CO2 (i.e., minute ventilationincreases in a more responsive fashion to increasing lev-els of CO2 (39,40). Thus, the higher percentage of aminoacids provided in the adult studies may have blunted thedifferences in CO2 production at different glucose/fatratios.

Another factor with an impact on CO2 production istotal energy intake. There is a linear relationship betweenenergy intake and CO2 production and increased minuteventilation (presuming glucose is providing a portion ofthe energy) (38,40). Indeed, the patients reported to havedeveloped respiratory failure (i.e., CO2 retention and in-creased minute ventilation) as a consequence of PN withhigh concentrations of glucose were receiving excessiveenergy intakes (see reference 35 for a review). Undernormal circumstances the increased CO2 production ishandled easily by increasing the respiratory rate and/ordepth. Problems potentially arise in patients with respi-ratory compromise. The respiratory response to enteralcarbohydrate intake and enteral overfeeding mirrorswhat is seen with PN (41).

In summary, the greater the proportion of glucose inthe PN energy mix, the larger the increase in CO2 pro-duction and minute ventilation. Malnourished patients

R. J. SHULMAN AND S. PHILLIPS590

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 5: Parenteral Nutrition in Infants and Children

are least and hypermetabolic individuals are most sus-ceptible. Normally nourished and metabolically normalpatients have an intermediate response (42). The greaterthe total energy intake, the greater the effect.

In selected patient groups, giving glucose in amountsabove the maximal oxidative rate may be appropriate(e.g., preterm infants who need to deposit fat). However,the greater the amount of glucose, the greater the risk ofadverse consequences. Unfortunately, the absolute intakeat which the adverse consequences occur is poorly de-fined.

Overfeeding of glucose also can affect liver functionalthough its contribution to the development of cholesta-sis in humans is unclear. Burke et al. observed a rela-tionship between glucose infusion rate and the develop-ment of fatty infiltration of the liver in necropsies ofchildren who died of severe burns (43). In a study in 37catabolic adults, Tulikoura et al. administered PN usingglucose and glucose and fat as energy sources (44).There was a significant increase in hepatic steatosis inthe glucose group but none when glucose and fat were

used in combination (44). Steatosis was associated witha rise in serum transaminases but no cholestasis wasnoted (44). Studies in normal adult volunteers suggestthat high carbohydrate feeding leads to an increase intotal VLDL triglyceride secretion rate from de novo syn-thesis primarily due to stimulation of the secretion ofpreformed fatty acids (45). The results imply that theliver derives all its energy from carbohydrate oxidationas opposed to fatty acid oxidation such that fatty acidstaken up by the liver are channeled into VLDL triglyc-erides (45). Hepatic steatosis results when export of theVLDL triglycerides does not keep pace with production(34,45).

Another potential complication of overfeeding withglucose is an increase in infectious complications. PNhas been associated with an increased risk of infectiouscomplications compared with enteral feeding or no nu-tritional support in adult studies (46–49). Recent datasuggest that this finding may actually be explained byoverfeeding with glucose (50). Hyperglycemia is a riskfactor for infection (51,52). Recently, hyperglycemia

FIG. 1. Respiratory quotient (CO2 production / O2 consumption) and caloric yield from the oxidation of glucose, fat, amino acids, and theconversion of glucose to fat. *Energy production values shown are theoretical. Actual energy produciton is less efficient. For conversionof palmitate it is about 41% (980/2398) whereas glucose conversion to fat is only about 20% efficient. Adapted from Flatt JP: Energeticsof intermediary metabolism. In Assessment of Energy Metabolism in Health and Disease, p. 77 (ed.) JM Kinney, Report of the First RossConference on Medical Research. Columbus, Ohio, Ross Laboratories, 1980.

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 591

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 6: Parenteral Nutrition in Infants and Children

was associated with an increased risk of infection inchildren with burns (53). In a prospective randomizedtrial in adults there was no difference between groupsreceiving hypocaloric intravenous fluids and those re-ceiving PN that did not overfeed the patients or inducehyperglycemia (54). These intriguing results require fur-ther verification but may have great implications for pa-tient care.

Clinical Implications

Patients with respiratory compromise, particularlyCO2 retention (e.g., cystic fibrosis) are at risk for expe-riencing a worsening of their pulmonary status if glucoseprovides a large proportion of their energy needs. Therisk increases as energy intake surpasses REE. Malnour-ished patients may not manifest this response until theybegin to be nutritionally rehabilitated. A hypermetabolicpatient will exhibit this reaction most quickly. Concernsabout overfeeding have been addressed recently in a con-sensus statement on the care of the adult intensive careunit patient (55).

Intravenous Fat Emulsion

Unless otherwise noted, the term intravenous fat emul-sion refers to the currently available (in the UnitedStates) soy- or soy/safflower-based emulsions. A keyrole for intravenous fat emulsion is the prevention ofessential fatty acid deficiency. The usual preterm infant(< 2 kg birthweight) on PN without intravenous fat emul-sion or on enteral feedings begins to demonstrate bio-chemical evidence of essential fatty acid deficiency (in-creased triene/tetraene ratio) within seven days of birth ifnot provided with adequate intravenous fat emulsion(56). In the absence of intravenous fat emulsion, endog-enous adipose tissue and the liver are probably the sourceof essential fatty acids (57). In order to prevent the mo-bilization of fatty acids for energy, essential fatty acidsmust be provided in amounts necessary to both replacedeficits and supply ongoing needs for metabolism andgrowth. In other words, essential fatty acid deficienciesare cumulative (58,59). Given that intravenous fat emul-sions contain approximately 50% essential fatty acids byweight, about 0.5 g � kg−1 � d−1 of intravenous fat emul-sion is required to prevent essential fatty acid deficiencyin patients on total PN (56,60). However, there is con-troversy regarding this figure, particularly in adults (andpresumably adolescents) with some investigators sug-gesting the recommendation is more than adequate (e.g.,1.5 g/kg twice weekly) and others stating it is too low(61–64). There appears to be a fair amount of inter-individual variation in essential fatty acid requirementsbased upon age, disease, and nutritional status (56,61,62,65). Thus, one should not presume that giving the small-est recommended amount of essential fatty acid, such as

1.5 g/kg (approximately 500 mL of 20% intravenous fatemulsion) twice a week in an adolescent, would be ad-equate for all such patients. In patients receiving no in-travenous fat emulsion, some studies suggest that fattyacid deficiency develops faster when total energy intakeis low than it does when total energy intake is high(66,67).

It often is suggested that intravenous fat emulsion beadministered over less than 24 hours (e.g., 20 hours) toallow time for the fat to clear from the blood. In fact,studies suggest that it is better to administer the dailydose over 24 hours in patients who have difficulty tol-erating the infusion. Preterm infants should receive theinfusion over 24 hours to avoid hypertriglyceridemia.For patients without hypertriglyceridemia or other evi-dence of intolerance to intravenous fat emulsion, a lessthan 24-hour infusion schedule may be tolerated well.One of the major determinants of the rate of clearance offat from the blood is the amount of intravenous fat emul-sion infused per unit time (68). The longer the infusiontime, the less likely the patient will develop hypertriglyc-eridemia (68,69). Intravenous fat emulsions with highconcentrations of phospholipids (i.e., 10% emulsions)should be avoided as they carry a higher risk of produc-ing high serum levels of triglycerides, cholesterol, andphospholipids than other emulsions (i.e., 20% and 30%emulsions) (70–72).

A number of other factors decrease the rate of clear-ance of intravenous fat emulsions. The more malnour-ished the patient, the slower the rate of clearance will be.Slower clearance in malnourished patients is a result oflower levels of lipoprotein lipase. This enzyme, which isresponsible for releasing fatty acids from the fat emul-sion, resides within the capillary system. The lower thecapillary tissue mass (a situation found in preterm infantsand malnourished patients), the slower the rate of intra-venous fat emulsion clearance. Studies in adult volun-teers have shown that the muscle, splanchnic, myocardi-al, and subcutaneous fat capillary systems clear 47%,25%, 14%, and 13% of the total amount of lipid infused,respectively (73). Under normal circumstances the liverclears less than 1% of infused lipid (73).

Another factor that alters the tolerance to intravenousfat emulsion is the concurrent administration of drugs.Because of their lipolytic effect, steroids are the proto-type. Patients simultaneously receiving steroids and in-travenous fat emulsions are prone to hypertriglyceride-mia (74). Additionally, a number of medications containlipid or are microsomal formulations. Drugs such as pro-pofol and amphotericin B may contribute significantamounts of (fat) energy to the total daily intake andwhich should be taken into account in nutritional calcu-lations.

Patients who are metabolically stressed (i.e., sepsis,trauma) or who have organ dysfunction (e.g., liver and/orrenal disease) also are likely to develop hypertriglyceri-demia during intravenous fat emulsion infusion. These

R. J. SHULMAN AND S. PHILLIPS592

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 7: Parenteral Nutrition in Infants and Children

patients have an outpouring of cortisol, catecholamines,and cytokines that promote the lipolysis of endogenouslipid stores (75–77). Liver and renal disease themselvesare associated with hypertriglyceridemia. However, adult(and presumably pediatric) patients with sepsis and liverdisease can tolerate and utilize intravenous fat emulsions(78,79). Infants with ventricular septal defects or trans-position of the great vessels (i.e., cyanotic heart disease)do not have inordinate increases in serum triglycerides inresponse to intravenous fat emulsion (80).

Serum triglyceride levels should be measured fourhours after starting an intravenous fat infusion or fourhours after any increase in infusion rate. It is at this timethat hypertriglyceridemia is most likely to occur (81–83).Ideally the triglyceride level should be �100 mg/dL(84). However, this upper limit varies depending on themethod used to measure serum triglycerides. For ex-ample, some common methods of measuring triglycer-ides measure the free glycerol released from the triglyc-eride molecule (85). Given that intravenous fat emul-sions contain free glycerol in addition to triglycerides,the serum triglyceride level measured by evaluatingglycerol will be an overestimation of the true serum tri-glyceride level. A value of 150 mg/dL may correspond toan actual triglyceride level of only 100 mg/dL (Buffoneand Shulman, unpublished data). Nephelometry shouldnot be used to monitor serum triglyceride levels as it isunreliable (85,86). Heparin administration lowers serumtriglyceride levels but does not affect the rate of fattyacid oxidation (83,87,88).

Clinicians frequently ask what level of hypertriglyc-eridemia is acceptable in a patient receiving PN. Giventhe need to provide essential fatty acids to patients whomay have hypertriglyceridemia for prolonged periods(e.g., the critically ill patient), at what serum level do thetriglycerides start having an adverse impact on the pa-tient? Unfortunately, to our knowledge, there are no datadefining the level at which one should be concerned.

Many investigators have studied the impact of intra-venous fat emulsion on pulmonary function. Although anin depth review of this topic is beyond the scope of thisarticle, a few points can be touched upon and the readeris referred to two reviews (89,90).

After a fatty meal, the serum triglyceride of normaladults may transiently approach 500 mg/dL (91). Evenvalues above 1500 mg/dL do not appear to decrease thecarbon monoxide diffusing capacity in healthy adults(91). This begs the question as to whether elevated tri-glycerides cause some of the adverse oxygenation effectsreported with intravenous fat emulsions that when severeis referred to as the “fat overload syndrome” (92). It wasfirst believed that the sometimes seen adverse effects ofintravenous fat emulsions on pulmonary function (e.g.,decreased diffusion capacity, oxygenation, intrapulmo-nary shunting) were related to the serum triglyceridelevel (89,90,93). It appears more likely that thesechanges, when seen, are due to the conversion of the

polyunsaturated fatty acids in the emulsions to prosta-glandins, which then can cause vasoconstriction or va-sodilatation depending on the rate of infusion and theclinical state of the lungs (85,86,93). Peroxides are an-other contaminant that can be found in the fat emulsion(94). Evidence suggests that they also promote increasedprostaglandin levels (95).

Interpretation of the data linking intravenous fat emul-sions and pulmonary dysfunction is complicated. Theadverse effects appear to depend on the dose and rate ofadministration, presence of peroxides, and clinical stateof the lungs (89,90,93–95). Are the effects clinically rel-evant? In preterm infants the balance of data suggeststhat administration of intravenous fat emulsion at normalrates does not affect oxygenation (72). Studies to thecontrary have usually used inappropriately rapid rates ofinfusion (96–98). Because intravenous fat emulsion caninduce vasoconstriction that is not limited to the pulmo-nary vasculature, observed decreases in transcutaneouspO2 may be related to local declines in subcutaneousperfusion and not reflect systemic oxygenation (95,99).The effects of intravenous fat emulsion on the develop-ment of chronic lung disease have been reviewed re-cently (72). The balance of the data suggests no adverseimpact and possibly a beneficial effect (72,100). How-ever, much more work is needed in this area before wecan be certain that there are no clinically significant ad-verse effects.

Even more confusing than the issue of pulmonaryfunction is the controversy regarding the effect of intra-venous fat (101–105) emulsions on immune function.Some studies show a beneficial effect, some no effect,and some an adverse effect. A complete discussion isbeyond the scope of the present article and the reader isreferred to the supplied references. It should be pointedout that reviews in this area sometimes give a prejudicedview in that only selected publications are quoted. Inter-pretation of publications often is blurred by the use ofinappropriate doses (both in vivo and in vitro), differ-ences in patient groups or in vitro models (102). Similarto the situation in pulmonary function studies, many ofthe effects of intravenous fat emulsions on immune func-tion appear to be related to the generation of leukotrienesand other polyunsaturates from the emulsions. There alsois evidence the long chain fats may alter cellular functionby changing membrane fluidity (104). Studies on pedi-atric patients and tissues have not revealed evidence sug-gesting an impairment of immune function by intrave-nous fat emulsions (106–113).

Do intravenous fat emulsions promote infections?Data are limited. A study in adult and pediatric bonemarrow transplant patients found no increase in the riskof bacteremia or fungemia in patients receiving intrave-nous fat emulsions (114). In contrast, a retrospectivestudy of preterm infants identified intravenous fat emulsionas potentially contributing to the development of coagulase-negative staphylococcal infection (115). However,

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 593

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 8: Parenteral Nutrition in Infants and Children

the lack of a temporal relationship between intravenouslipid administration and infection weakens the conclu-sions of this study (115). The lack of knowledge on thisimportant issue underscores the importance of monitor-ing and using intravenous fat emulsions judiciously.Newer intravenous fat emulsions not yet available in theUnited States using olive oil or structured lipids mayobviate some of these immune-related concerns (116).

The use of intravenous fat emulsions in patients withcoagulation abnormalities requires comment. In rarecases, intravenous fat emulsion has been associated withthrombocytopenia. However, prospective and retrospec-tive reports suggest that under normal circumstancesthey do not induce thrombocytopenia (117,118). Itshould be kept in mind that essential fatty acid deficiencyitself is a cause of hematologic abnormalities (119).

Clinical Implications

There is debate as to the appropriate amount of intra-venous fat emulsion required to prevent essential fattyacid deficiency. In infants (including preterms) who aremost susceptible, at least 0.5 g � kg−1 � d−1 should beused. In older children and adolescents the minimumdose is probably in the range of 1.5 g/kg (approximately500 mL of 20% intravenous fat emulsion) twice a week.For individuals on long term parenteral nutrition whoeither have little or no enteral intake of fat or who havesevere fat malabsorption (e.g., severe short bowel) es-sential fatty acid status should be assessed after a fewmonths of therapy. It has been suggested that measure-ment of n-6 and n-3 long chain polyunsaturated fattyacids is more reliable than the more commonly employedtriene/tetraene ratio (57).

The American Gastroenterological Association Tech-nical Review on Parenteral Nutrition states that intrave-nous fat emulsion should not be administered to patientswhose serum triglyceride levels exceed 400 mg/dL (21).No data are provided to support this statement but theauthors agree that serum triglyceride levels � 400 areunlikely to cause clinically relevant problems. How theduration hypertriglyceridemia might affect the complica-tion rate is unclear. In preterm infants pulmonary vascu-lar lipid deposition was more common in infants whoreceived intravenous fat emulsion and correlated with theduration of lipid therapy although some infants whonever received intravenous fat emulsion also demon-strated lipid deposition (120). Liver or pancreatic dis-eases are not in themselves contraindications to the useof intravenous fat emulsion. It is helpful to measure se-rum triglyceride levels prior to starting intravenous fatemulsion in patients who are likely to be hypertriglyc-eridemic and when their clinical status deteriorates sig-nificantly.

Amino Acids

Use of amino acids in pediatric PN has recently beenreviewed and the reader is referred to these references formore details (121,122). Amino acid requirements varyfrom � 2.5 g � kg−1 � d−1 for preterm infants to 0.75 g �kg−1 � d−1 for adolescents (123,124). Similar to energyrequirements, amino acid requirements vary from indi-vidual to individual under normal circumstances. Theimmunodeficiency associated with malnutrition is to agreat extent related to protein malnutrition as opposed toenergy malnutrition. Consequently, amino acid require-ments should be attended to first in prescribing paren-teral nutrition. The intake of energy and amino acids incombination result in the greatest protein gain. Aminoacid intake is the main determinant of protein gain (Fig-ure 2). That is, there is a much larger gain in proteinaccretion with increases in amino acid intake than withincreases in energy intake (Fig. 2).

Since amino acids are generally not metabolized tosupply energy but to provide structural and visceral pro-teins and enzymes, some clinicians do not include themin the total energy calculation. In practice, to includethem or not usually does not make much difference. Forexample, a term infant who weighs 4 kg requires about 8g/d of amino acids, which is equivalent to 32 kcal (8 g x4 kcal/g). The total energy requirement is about 400kcal/d (4 kg x 100 kcal/d). Thus, adding the calories fromprotein changes the total energy calculation by less than10%. Interindividual energy requirements easily can varyby this much. The bottom line is that both amino acidsand energy need to be titrated to the patient’s require-ments.

Although serum prealbumin (transthyretin) and trans-ferrin are often used as short-term (1–2 day) measures ofthe response to amino acid intake, the blood urea nitro-gen (BUN) can be also be used in many cases. If renalfunction and hydration are normal, a low BUN (< 5)reflects inadequate amino acid intake and conversely, ahigh BUN (> 20) suggests that the amino acid intake maybe too great. For many patients a BUN between 10 and15 mg/dL is a good target and reflects an amino acidintake that is adequate for protein anabolism (Laine andShulman, unpublished).

Although it has been stated in the literature that theamino acid intake should be gradually increased over thefirst few days of PN administration, we know of no goodscientific documentation for this view. For example, weuse intakes of 2.5 mg � kg−1 � d−1 even in 25-weekgestation infants without complications starting within acouple days after birth. A recent review of the literatureon the use of early amino acid introduction found nometabolic abnormalities (elevated BUN, serum pH, am-monia, serum amniogram) associated with this practiceeven in preterm infants (123). Gradually increasing theamino acid concentration or volume only postpones thetime at which the patient receives adequate intake.

R. J. SHULMAN AND S. PHILLIPS594

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 9: Parenteral Nutrition in Infants and Children

Evidence suggests that amino acids contribute to thedevelopment of PN-associated cholestasis (125,126).Even enteral protein intake can decrease bile flow (127).If amino acids are removed from the PN when the patientdevelops cholestasis and no other source of protein isprovided, one then is faced with a patient who in additionto having liver disease has hypoproteinemia or evenkwashiorkor. There are some data to support the idea thatprovision of protein enterally while other nutrients (pri-marily carbohydrate) are provided intravenously can re-duce the risk of or lessen the severity of PN-associatedcholestasis (125). Protein is usually well digested and

absorbed, even in patients with short bowel syndrome(128). Consequently, this is an option for many patientsreceiving PN. Of course, whenever possible other nutri-ents should be provided enterally as well.

There are very limited data in pediatric patients re-garding the use of specialized amino acid solutions forspecific conditions such as critical illness and liver orrenal failure. A crossover trial in children with end stageliver disease suggested a benefit to a branched chainsupplemented enteral formulation in that the special for-mula promoted more rapid nutritional rehabilitation(129). Studies in adults with liver disease have been in-

FIG. 2. Differential effect of amino acid andenergy intake on protein gain in preterm in-fants. Expected protein gain in a preterm in-fant related to amino acid and energy intake.If amino acid intake is doubled from 1 g �kg−1 � d−1 to 2 g � kg−1 � d−1 and energyintake is held constant at 60 kcal � kg−1 � d−1

(Top Panel) the corresponding increase inprotein gain is greater than if the energy in-take is doubled from 60 kcal � kg−1 � d−1 to120 kcal � kg−1 � d−1 and amino acid intake isheld constant at 1 g � kg−1 � d−1 (MiddlePanel). The greatest protein gain is achievedwith an increase in amino acid intake andenergy intake (Bottom Panel). Adapted from(123).

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 595

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 10: Parenteral Nutrition in Infants and Children

consistent. Most studies indicate that if used at all,branched chain formulae should be reserved for patientswho are malnourished and/or who are encephalopathic(130,131). The benefit of specialized amino acid solu-tions for patients with renal disease also is not clear(132,133). Similarly, the nutritional benefit of parenteralnutrition given during dialysis requires further study(134,229). The two commercially available (in theUnited States) pediatric amino acid solutions have con-centrations of essential amino acids, including branchedchain amino acids, which are fairly close to those in thespecialized amino acids solutions.

Clinical Implications

Provision of amino acids should be a priority in cal-culation of PN requirements. In situations where fluidintake is significantly restricted, energy intake should besacrificed at the expense of maintaining a greater aminoacid intake, particularly when more than resting energyexpenditure can be provided. The PN concentration ofamino acids can be increased to 4%–8% (g/dL) depend-ing on the amino acid brand used (concentration of thestock amino acid solutions range from 10%–20%). TheBUN is a readily accessible and cheap method of titrat-ing amino acid intake if renal function and hydration arenormal. Consideration of enteral (as much as possible)rather than parenteral administration of protein should beconsidered, particularly in patients at risk for cholestasis.If not limited by intolerance (e.g., enteral carbohydrate ina patient with short bowel syndrome) other nutrientsshould be provided enterally as well.

CYCLIC PARENTERAL NUTRITION

The term cyclic PN refers to the administration ofintravenous fluids intermittently with regular breaksfrom infusion. Cyclic PN offers the advantage of increas-ing the mobility of the patient and family. Much has beenwritten regarding its other presumed advantages butthere are scant scientific data (135,136). One of the pro-posed benefits is a lower risk for the development of liverdisease although reports have been anecdotal (137). Re-cently, Hwang et al. carried out a prospective study inadults on PN exhibiting various degrees of presumedPN-associated liver disease (138). Patients who devel-oped hyperbilirubinemia were randomized to either re-main on continuous PN or were placed on cyclic PN.Patients with initial serum bilirubin less than 20 mg/dL,who remained on continuous PN, had a significant rise inserum bilirubin compared with the cyclic PN groups(138). There was no apparent advantage of cyclic PN inpatients with serum bilirubin greater than 20 mg/dL(138). These results are intriguing and beg for similarstudies in pediatric patients. For information on PN-

associated cholestasis the reader is referred to recent re-views (139–141).

REFEEDING SYNDROME

Refeeding syndrome is a potentially lethal conditioncharacterized by severe electrolyte and fluid shifts asso-ciated with metabolic abnormalities in malnourishedpatients undergoing refeeding orally, enterally, or paren-terally (139). Clinical features include fluid-balance ab-normalities, abnormal glucose metabolism, hypophos-phatemia, hypomagnesemia, hypokalemia, and some-times thiamine deficiency (142).

Refeeding syndrome can occur with either parenteralor enteral feeding. Because it is a preventable condition,patients who are at risk should receive a PN prescriptionfrom the outset that will decrease the likelihood of itsoccurrence. A detailed review of the condition is beyondthe scope of this article but the reader is referred tocomprehensive reviews (142,143). The syndrome can beprevented by adequate provision of phosphorus, and to alesser extent magnesium and potassium at the time thepatient begins to become anabolic. A common miscon-ception is that additional phosphorus and potassium pro-vided at the onset of PN will prevent the syndrome. Thisis true only if adequate nutrition is being provided to thepatient so that they are actually regaining lost tissue mass(i.e., muscle and viscera). If the amount of nutrition (en-ergy) being provided is less than that required to main-tain weight or promote catch-up weight gain, the syn-drome will not develop. Consequently, the patient maybe at risk for hypophosphatemia and hypokalemia weeksafter the initiation of PN (or whenever nutrient intakeallows for catch-up weight gain).

Refeeding syndrome can be prevented by providingadditional phosphorus and potassium (usually as KxPOx)at the time PN is started. The amount of additional phos-phorus and potassium required depends on the degree ofmalnutrition and the adequacy of total energy intake. Ageneral rule of thumb in the authors’ experience is thattwice the recommended daily allowance (RDA) of phos-phorus is a reasonable starting point. Usually the risk forclinically significant declines in these minerals lasts for7–10 days. During this period, it is important to followserum phosphorus and potassium closely (e.g., everyother day or more often if serum levels are tenuous).After this time, supplementation above the RDA cangradually be reduced. Calcium and phosphorus shouldnot be given independently as administering one withoutthe other leads to renal wasting of the infused mineral(144).

A rate-limiting factor in providing additional phospho-rus is the risk of calcium/phosphorus precipitation. Table3 provides guidelines for the maximal amounts of cal-cium and phosphorus that can be administered in PN(145–147). Note that the addition of cysteine (HCl) be-

R. J. SHULMAN AND S. PHILLIPS596

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 11: Parenteral Nutrition in Infants and Children

cause of its low pH enhances the solubility of the min-erals. Computer software is available to assist in assuringthat calcium and phosphorus solubilities are not ex-ceeded (148). In preterm infants and possibly other pa-tients prone to acidosis, the addition of cysteine mayrequire that the patient be supplemented with a source ofbase (e.g., acetate) (149). Calcium and phosphorus re-quirements for preterm infants have been fairly well de-lineated in contrast to older children (150,151).

PARENTERAL NUTRITION ADMIXTURES

PN admixtures (also called Total Nutrient Admixturesor TNA, Three-in-One solutions, or All-in-One solu-tions) make use of the fact that under certain conditionsthe glucose/amino acid solution may be mixed with in-travenous fat emulsion and administered to the patient inone bag. There are a number of potential advantages toPN admixtures but also some drawbacks (Table 4; seebelow) (152).

In addition to reducing cost, use of a PN admixturesimplifies the patient’s life in that only a single infusionpump with less tubing and other accessories are needed(153). There also is evidence that it reduces the risk ofbacterial growth in intravenous fat emulsion even after24 hours (154). In a randomized trial in which adultpatients (N � 96) received either PN admixtures or tra-ditional (two-in-one) PN there was no difference in therate of infection between the two groups (155).

Recently, a bag has been developed that allows thesolution to be made with the three components (aminoacids, glucose, fat) stored in separate compartmentsseparated by inner seals (Kabiven Multichamber Paren-

teral Nutrition Packaging System, Fresenius Kabi,Uppsala, Sweden). The bag and components can bestored up to 24 months with the seals between the cham-bers intact. Seals are opened just before use. PN admix-tures find their greatest usefulness in older patients, par-ticularly those on cyclic PN and those at home (see be-low).

Because the admixture is a physical mix of naturallyincompatible substances (oil and water, calcium andphosphorus), their utilization requires strict attention topharmaceutical guidelines for preparation, storage, anduse (156). Particulate matter (mobile, undissolved sub-stances that are unintentionally present in products) re-sulting from inappropriate preparation or storage can belife threatening. For example, deaths have been reportedpresumably as a result of precipitation of calcium andphosphorus in admixtures (157,158). Table 5 lists poten-tial areas of concern in the preparation and storage of PNadmixtures.

There have been multiple publications regarding howPN admixtures should be prepared and their stability.Preparation of an admixture requires exquisite attentionto detail. The commercial product used, the sequence ofmixing, the type and concentration of additives, and stor-age conditions are critical for maintaining lipid dropletsize and distribution and preventing the development ofparticulate matter (159). For example, ideally all water-soluble additions should be mixed together whereas li-pophilic additives (e.g., fat soluble vitamins) should beadded to the intravenous fat emulsion (159). The last stepshould be mixing the intravenous fat emulsion (and itscomponents) with the water-soluble components (159).Undiluted hypertonic glucose solution should not be

TABLE 3. Maximal calcium and phosphorus concentrationsin parenteral nutrition solutions

g/dL (%)Amino acids

Without cysteine With cysteine#

CalciummEq/L

Phosphorusmmol/L

CalciummEq/L

Phosphorusmmol/L

1.0 50 12 50 232 50 2 50

1.5 50 11 50 185 50 12 50

2.0 50 12 50 257 50 12 50

2.5 50 22 50 386 50 28 50

* Presumes the base amino acid solution has a pH of �5.5Administration of intravenous fat emulsion into the same intrave-

nous catheter may reduce solubility by 10%–20% (145, 146). Additionof acetate without the addition of cysteine also may reduce solubility by5–10% (147). # 40 mg/g of amino acids.

Note: This table is intended as a general guideline and is not to beused in lieu of pharmacist verification of the compatibility of the PNsolution. Modifications to the formulation and alterations in the methodand order of preparation may result in different calcium/phosphorussolubilities.

TABLE 4. Advantages and disadvantages of parenteralnutrition admixtures

Advantages Disadvantages

Nursing time for administrationof PN is decreased

Particulate matter is difficult tosee prohibiting visualinspection of solution

Rate of extrinsic touchcontamination potentiallyreduced

Filtration must be performedwith a larger filter (1.2micron vs. 0.22 micron)

Pharmacy preparation timedecreased

Admixtures support microbialgrowth less than fat emulsionalone

Support the growth ofmicroorganisms better thanglucose-amino acid solutions

A two pump system iseliminated

Computer interfaced automatedcompounder expensive

Increased compliance ofadministering fat emulsion inthe home patient population

Emulsion instability can bedisrupted with highconcentrations of electrolytesor when base solutioncomponent amounts exceedcompatibility limits

The cost of ‘Y’-site tubing andadditional supplies is saved

From (152).

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 597

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 12: Parenteral Nutrition in Infants and Children

added to intravenous fat emulsion (160). Solutions maybe stable when refrigerated and become unstable at roomtemperature. Another example, heparin stability, hasbeen a subject of great debate. Contradictions regardingits risk of precipitating with calcium probably are re-lated to differences in the studied heparin prepara-tions and the other additives and their concentrationsin the admixtures (159). The reader is referred to com-prehensive reviews regarding the factors related to PNadmixture stability as well as methodologies used totest stability (159–163). In general, final concentra-tions range from 2% to 5% for amino acids, 5% to 23%for glucose, and 1.5% to 5% for intravenous fat emul-sion (161).

In order to see through the morass of information it isbest to follow a simple rule: after identifying a PN ad-mixtures whose stability has been well documented, usethem exactly as described. Unlike standard PN solutions,alterations of components, volumes, and/or concentra-tions should not be made. A case in point is a recentreport of episodes of respiratory distress and death pos-sibly related to changing from one commercial aminoacid brand to another despite (or it should be said, be-cause of) not modifying any other aspect of the admix-ture (164). A Food and Drug Administration (FDA) alertwas issued in 1994 regarding the hazards of precipitation(165). It is recommended that information on the properpreparation of a PN admixture be obtained directly fromthe manufacturer(s) of the amino acids and intravenousfat emulsion that are to be used. As a general rule, theamino acids, glucose, and intravenous lipid are added infixed ratios by volume (e.g., 2:1:1, 1:1:1, 1:1:0.5, etc.)(166). To reduce the risk of precipitates reaching thepatient an inline filter should be used. This should be a1.2 micron air-eliminating filter (in contrast to the 0.22micron air-eliminating filter recommended for nonlipid-containing PN) (165). The importance of filtration re-cently has been reviewed and revalidated (167,168). APN admixture is considered inappropriate for adminis-tration if > 0.4% of the total fat contains particles > 5microns in size (162,163).

What patients are candidates for PN admixtures? Itmay be more important to ask whether there are PN

mixtures that are inappropriate for certain patients. Be-cause of the limitations in the amount of calcium andphosphorus that can be used in PN admixtures (evenmore severe than in non- admixture PN), their use maybe inappropriate in young infants. One report has sug-gested that PN admixtures can be used in former preterminfants and newborns, however, in this study it was notclear that the PN admixtures had been validated as stablenor was there stability testing reported after the admix-tures were prepared (169). Because the emulsion isopaque (as a result of fat emulsion) a precipitate may beinvisible to the naked eye. The calcium and phosphorusrequirements for neonates is 3–4 mEq � kg−1 � d−1 and1–2 mmol � kg−1 � d−1, respectively (162). Based uponstudies using a commonly employed pediatric aminoacid solution (TrophAmine, B. Braun Medical, Inc., Mel-sungen, Germany) the PN admixture would provide in-adequate amounts of these minerals (170). As an infantapproached 4–6 months of age it is conceivable that theircalcium and phosphorus needs could be met using a PNadmixture (161,162,170). However, it must be stressedthat this is dependent upon the amino acid, glucose, in-travenous fat emulsion, calcium, and phosphorus prod-ucts (as well as the other components) used as well as themixing protocol. If a PN admixture is to be used in aninfant less than a year of age, it should be done withparticularly careful consideration regarding mineral re-quirements.

PN admixtures only should be used in patients who areclinically stable. Changes in the formulation of PN ad-mixtures are more costly than changes in traditional(two-in-one) PN because of the wastage of both thedextrose/amino acid and its components, and the intra-venous fat emulsion.

A number of questions remain about PN admixtures.These include the appropriate dosing of some vitamins,the true risks related to particulate matter and fat dropletsize, and drug compatibilities. For example, there is lessloss of fat-soluble vitamins such as vitamin A than intraditional (two-in-one) PN because the intravenous fatemulsion is protective (see below) (171). Some drugs arecompatible with PN admixtures but not with traditionalPN and visa versa (172).

TABLE 5. Three potential issues in the use of PN admixtures and contributing factors

Chemical precipitation Intravenous fat stability Vitamin stability

Calcium and phosphorus Low pH Thiamine reduction bymetabisulphite

Sulphur-containing amino acidsand copper

Intravenous fat concentration Lack of exclusion of oxygenduring compounding

Sulphur-containing amino acidsand calcium

Formation of peroxides Lack of use ofoxygen-impermeablecontainers during storage

Ascorbic acid and seleniumIron and phosphateOrder of Mixing

From (159).

R. J. SHULMAN AND S. PHILLIPS598

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 13: Parenteral Nutrition in Infants and Children

VITAMINS

Guidelines for pediatric parenteral vitamin and min-eral supplementation have been previously reviewed, al-beit quite some time ago (173). Despite subsequent pub-lications that have provided additional support for theserecommendations, there has not been a recent significantevaluation or reformulation of parenteral vitamin prod-ucts for premature infants, infants, or children (less than11 years of age) (174,175).

Recommendations from the 1998 National AdvisoryGroup on Standards and Practice Guidelines emphasizedthe need for establishing optimal trace element and vita-min formulations for both adult and pediatric patients

(162). In April 2000, the Food and Drug Administrationamended the adult multivitamin formulation to bring itinto accordance with the 1988 recommendations of theAmerican Medical Association-FDA Public WorkshopCommittee (Table 6) (176). To date, the optimal paren-teral vitamin and mineral requirements for children andneonates have not been determined and a parenteral mul-tivitamin formulation specifically for preterm infants hasnot been developed.

While there are several parenteral vitamin prepara-tions available for older children (> 11 years) and adults(Table 6), few are available for neonates and children.InFuvite Pediatric™ (Sabex, Inc., Boucherville, Canada)and M.V.I. Pediatric™ (aaiPharma, Inc., Wilmington,NC) are approved for use in prematures, infants, andchildren < 11 years (Table 7). The adult formulations arenot recommended for use in low birth weight infants lessthan 1500 grams because of concerns about the toxicityof the propylene glycol and polysorbate additives(177,178). Limitations in the availability of the pediatricproducts have made children vulnerable to shortages(179). Table 7 gives the doses of vitamins recommendedfor infants by the American Society of Clinical NutritionSubcommittee on Pediatric Parenteral Nutrient Require-ments (). This recommendation differs from the packageinserts (180).

Clearly, there are patients whose needs do not fit thecurrent vitamin formulations. For example, preterm in-fants, children with liver or renal disease or short bowelsyndrome, or who are severely malnourished requireclose attention to vitamin nutriture. Some adult parenter-al vitamin products that are used for children >11 yearsof age may put younger patients on long term PN at riskfor excessive vitamin intakes (see Table 6).

TABLE 6. Adult and children (>11 years of age) parenteralmultivitamin preparations

VitaminInfuvite™*

(Sabex)M.V.I.-12™(aaiPharma)

A (IU) 3300 IU 3300D IU (�g) 200 (5) 200 (5)E (IU) 10 10K (�g) 150 —C (mg) 200 200Thiamin (mg) 6 3Riboflavin (mg) 3.6 3.6Niacin (mg) 40 40Pyridoxine (mg) 6 4Folate (�g) 600 400B12 (�g) 5 5Pantothenic Acid (mg) 15 15Biotin (�g) 60 60

* Conforms to the FDA amended formula for adult multivitaminpreparations (165).

TABLE 7. Infant and child parenteral multivitamin requirements andcommercial preparations*

Vitamin (amount/d)Best estimate#preterm infant

Pediatric parenteralmultivitamins#

<2.5 kg40% of the vial (2 ml)

Pediatric parenteralmultivitamins*>2.5 kg–11 yrs

100% of the vial (5 ml)

A (�g)** 500 280 700C (mg) 25 32 80D (IU) 160 160 400E (mg)## 2.8 2.8 7K (�g) 80 80 200Thiamin (mg) 0.35 0.48 1.2Riboflavin (mg) 0.15 0.56 1.4Niacin (mg) 6.8 6.8 17Pyridoxine (mg) 0.18 0.40 1Folate (�g) 56 56 140B12 (�g) 0.3 0.4 1Pantothenic Acid (mg) 2.0 2.0 5Biotin (�g) 6 8 20

* Infuvite Pediatric™/MVI-Pediatric™.# See reference (180).** 500 �g � 1643 IU.## 1 mg � 1 IU.

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 599

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 14: Parenteral Nutrition in Infants and Children

Vitamin K

One potential problem with the introduction of vita-min K into the new adult formulations is the possibleinterference with anticoagulants such as warfarin. Addi-tionally, intravenous fat emulsions contain vitamin K invarying amounts (approximately 13–70 �g/dL) that canmake titrating anticoagulant therapy even more problem-atic (181,182).

Vitamin A

A recent Cochrane review suggests that an increasedvitamin A intake (via the intramuscular route) is benefi-cial in the preterm infant (183). Whether dosing of vita-min A in PN will provide a similar benefit is unclear.Dosing of vitamin A is complicated by its adherence tothe bags and tubing and from photodegradation (184).Evidence suggests that when administered in theglucose/amino acid solution the current recommendationfor vitamin A is too low (162,185,186).

Loss of several lipid soluble (and water-soluble) vita-mins occurs during their delivery in PN (187,188). Evi-dence suggests that using dark tubing and placing fat-soluble vitamins into the intravenous fat emulsion (seePN Admixtures, above) mitigates the losses of these vi-tamins (189–192,230). However, multivitamin manufac-turers in the US do not recommend mixing vitamins inlipid emulsions (see package inserts). Obviously it iscommonly done in the case of PN admixtures (seeabove). More research in this area is vital.

Clinical Implications

Our knowledge regarding the appropriateness of cur-rent vitamin preparations and intake levels is less thanoptimal. Prudence would suggest checking vitamin lev-els (especially vitamin A and possibly riboflavin) in pa-tients on long term PN because of potential losses in theadministration set, particularly when enteral intakeand/or absorption is poor (191). Consideration alsoshould be given to monitoring fat-soluble vitamin levelsin patients whose vitamins are placed into the lipid emul-sion.

Carnitine

The use of carnitine in PN for preterm infants recentlyhas been the subject of a Cochrane Database review(193). Using weight gain as a primary outcome, therewas no apparent effect of carnitine supplementation(193). However, the studies available did not assess theuse of carnitine in long term PN, a situation that is morelikely to result in carnitine deficiency. Tissue carnitinelevels decline in infants receiving carnitine-free PN(194). An adult patient who received PN for over a year

has been described with symptomatic carnitine defi-ciency although the level of enteral intake was not de-scribed (195). It is possible that the effects of carnitinedeficiency in patients on long term PN may be too subtleto be identified easily but are still metabolically relevantgiven the central role of carnitine in metabolism (196).Addition of carnitine does enhance intravenous fat emul-sion oxidation although it has been argued that the in-crease may not be clinically relevant (193). In long termPN patients it may be reasonable to follow the suggestionthat carnitine (as pure L-carnitine) be added at a dose of2 to 5 mg � kg−1 � d−1 and that plasma and red blood cellconcentrations be measured at baseline and at 4-monthintervals until levels have stabilized, then yearly there-after (196). Too high a carnitine dose may be detrimental(196).

Iron

Parenteral nutrition patients at risk for iron deficiencyare premature infants, long-term PN patients with little orno enteral intake, and patients with significant malab-sorption or fluid losses. Iron may be administered orally,intramuscularly, or parenterally. Parenteral iron can beinfused as an intermittent intravenous infusion or as adiluted total dose infusion. Iron dextran has been used inchildren (197–200). Despite its fairly widespread use,total dose infusion is not approved by the Food and DrugAdministration.

The use of iron dextran can be complicated by adversereactions including a small but real risk of anaphylaxis.A test dose is recommended prior to administration ofthe required amount (see package insert). Iron dextran inPN can be efficacious (201,202). Although often em-ployed in PN, the stability of iron dextran and its poten-tial interaction with other nutrients remain as concerns(202). Its use in PN admixtures should be undertakenwith extreme caution if at all (163).

Until recently, iron dextran was the only formulationavailable for parenteral iron therapy. Two new iron prod-ucts free of dextran have recently been approved for usein the United States. Sodium ferric gluconate complex insucrose (Ferrlecit, Watson, Inc., Morristown, NJ) hasbeen used extensively in Europe and became availablefor use in the United States in 1999. Iron sucrose (Ve-nofer, Luitpold Pharmaceuticals, Inc., Shirley, NY) alsopreviously available in Europe was introduced in late2000. These products appear to be safe and have fewerside effects than those of iron dextran, although like irondextran, they can cause hypotension (203,204). Whilethe safety and efficacy of these products has not beenextensively studied nor are they approved by the FDAfor use in children, there is evidence that they are safeand effective in children (205).

R. J. SHULMAN AND S. PHILLIPS600

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 15: Parenteral Nutrition in Infants and Children

Trace Minerals

Trace mineral requirements and metabolism in gen-eral, including concerns regarding aluminum contamina-tion and toxicity, recently have been reviewed elsewhere(180,206–209). We briefly will touch on current issuesregarding specific trace minerals in PN. First, it shouldbe remembered that PN solutions usually contain someof the various trace minerals due to contamination of thecomponents (e.g., amino acids, calcium gluconate, mul-tivitamins) with trace metals (210). In one report Zn, Cu,Mn, and Se were found in greatest concentration butcontamination will vary depending on the commercialproducts used to prepare the PN (210). Heat and storagetime can modestly reduce the levels of some trace ele-ments as well (210).

The importance of Se as an antioxidant is well known.A recent report suggests that the recommended intake ofSe for preterm infants (N � 29, gestational age 26weeks) may not be adequate for all infants (211). Incontrast, a study of children and adults on long term PNsuggested Se status is well maintained even with low orno Se intake (212). However, in this study there wasgreat variability in Se serum levels as well as glutathioneperoxidase activity with some patients actually showingdeficiency (212). Given the reports of heart failure be-cause of Se deficiency and our limited knowledge re-garding adequacy of intake, it is prudent to check serumSe levels in the short term (particularly in preterm in-fants) and serum Se levels and glutathione peroxidaseactivity in older infants and adolescents (207,211–213).

Some studies have suggested that chromium toxicitymay be a concern. Mouser et al. noted in infants andchildren (up to 12 years of age) on long term PN thatserum Cr levels were elevated despite following the rec-ommended intake (180,214). These data are consistentwith other studies and imply that the addition of Cr to PNat recommended levels in combination with Cr contami-nation in the PN fluids raises serum Cr levels above thatdesired (210,215–217). Again, patients on long term PNshould be closely observed. Whether the impaired glo-merular filtration rate noted in children on long term PNis related to Cr excess requires further investigation(217).

Zn has the widest range of functions of all the traceminerals and is particularly important in wound healingand immune function. Despite its intentional and non-intentional addition to PN, many patients are at risk forZn deficiency because of the number of predisposingclinical situations (206,207). For example, patients withgastrointestinal losses due to diarrhea, ileostomies, oreven nasogastric suction are at risk for Zn deficiencybecause of the high Zn content of gastrointestinal fluids(218). Serum Zn levels, although not entirely reliable, areeasily obtained and Zn intake can be titrated appropri-ately.

Of great concern are the numerous observations re-garding Mn retention. This essential trace mineral nor-mally is excreted in bile and has a special affinity for theextrapyramidal system (206,207). Mn often contami-nates PN solution components in amounts sufficient fordaily requirements (206,207,219). Manganese intoxica-tion causes parkinsonian-like symptoms with muscularweakness, stiffness, tremors, ataxia, abnormal gait, as-thenia, and difficulty with speech (220). Psychologicalchanges have been reported including mental irritability,headaches, nervousness, compulsive actions, and hallu-cinations (220). Manganese accumulation can be de-tected in the basal ganglia as symmetric, increased signalintensity on T-1 weighted magnetic resonance images(220,221). These changes are reversible when Mn is re-moved from the PN (222,223). Clinical symptoms areusually but not always reversible (224).

It is not surprising that Mn accumulates in individualswith cholestatic liver disease given its hepatobiliary se-cretion (220). Of particular concern for pediatric patientsis the widely held suspicion that Mn accumulation inpatients on PN may cause cholestasis. A recent studyrandomized 244 children on PN to receive either 1.0(Group 1) or 0.0182 �mol � kg−1 � d−1 (Group 2) of Mn(224). When all patients were considered, those in Group1 showed a trend towards higher peak manganese anddirect bilirubin levels. The two groups did not differ inthe occurrence of cholestasis but Group 1 patientsshowed a trend towards increased incidence and severityof hyperbilirubinemia (225). Of the 160 children whoreceived >75% of their daily fluid intake from PN for>14 days, peak whole blood manganese and peak serumdirect bilirubin concentrations were significantly higherin Group 1. Significantly more infants in Group 1 devel-oped a more severe degree of direct hyperbilirubinemia(225). These data implicate Mn as a contributor to thedevelopment and/or the severity of PN-associated chole-stasis (225). Based on these and other data, it is sug-gested that patients with any cholestasis (some would sayany liver disease) should not receive Mn in their PN(220,225,226). Others would go so far as to say that Mnshould not be given to individuals on PN for <30 daysand that levels should be monitored in patients receivingPN with Mn for >30 days (207,226).

There is disagreement regarding the best measure ofMn status (220). Whole blood Mn appears to correlate(in adults) with MRI-documented Mn deposition (226).Most investigators use whole blood Mn, red blood cellMn, or Mn superoxide dismutase as measures of tissuedeposition (220).

SUMMARY

Our knowledge regarding PN reflects the same patternwe find in all of medicine: what we do is based part uponscience and part upon best judgment. It turns out that

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 601

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 16: Parenteral Nutrition in Infants and Children

practices that seem clear (i.e., based on science), oftenturn out to be wrong or not so clear. We now realize thataggressive nutritional support carries risks. The risk ofinfection may be related more to hyperglycemia (or hy-pertriglyceridemia) than to PN per se, and essential fattyacid requirements may not be as well defined as previ-ously thought. When it seems as if knowledge is fornaught, the following quote comes to mind: “I reservethe right to be smarter tomorrow than I am today”(227,228).

In a way, we are victims of our own success. PN canprovide effective nutritional support but, as is evidentfrom this review, there are large holes in our knowledgeand little likelihood of abundant future research to fillthem. For the most part, these deficiencies in knowledgeare more problematic for pediatric patients than foradults. Unfortunately, because pediatric PN makes upsuch a small share of the market, manufacturers are re-luctant to spend the resources to carry out the studiesneeded to define the child-specific impact of new lipidemulsions, reformulated vitamins, etc. Money spent onthese questions would not be recouped quickly. The re-sponsibility for improving PN safety and efficacy lieswith all concerned: industry, funding agencies, physi-cians, pharmacists, nurses, and nutritionists.

PN is still the life-saving, far-from-perfect therapy ithas been for over 40 years. Like all therapies it must beused and monitored appropriately. Using it means morethan just checking the patient’s weight, and monitoringnow means more than checking the serum electrolytes.Checking the box on the order sheet will never beenough.

Acknowledgments: The authors thank Dr. W.C. Heird forhis very helpful comments. This study was supported by theNational Institute of Child Health and Human Development,Grant No. RO1 NR05337-01A2, the Daffy’s and HenriksenFoundations, and the USDA/ARS under Cooperative Agree-ment No. 58-6250-1-003. This work is a publication of theUSDA/ARS Children’s Nutrition Research Center, Departmentof Pediatrics, Baylor College of Medicine and Texas Children’sHospital, Houston, TX. The contents of this publication do notnecessarily reflect the views or policies of the USDA, nor doesmention of trade names, commercial products, or organizationsimply endorsement by the US Government.

REFERENCES

1. Spencer CT, Compher C. The development of TPN: an interviewwith pioneer surgical nutritionist Jonathan E. Rhoads, MD. J AmDiet Assoc 2001:101;747–50.

2. Barry Lepatner 1983 in testimony before Congress; personalcommunication 2002.

3. Forchelli ML, Gura KM, Shulman RJ. Metabolism estimation,recommended energy, protein and fluid needs by age in pediatricnutrition in your pocket. Corkins MR, Shulman RJ, ed., AmericanSociety for Parenteral and Enteral Nutrition (in press).

4. Pierro A, Jones MO, Donnell SC. Total parenteral nutrition insurgical infants. Biochem Soc Trans 1998 May;26:131–6.

5. Jaksic T, Shew SB, Keshen TH, Dzakovic A, Jahoor F. Do criti-

cally ill surgical neonates have increased energy expenditure? JPediatr Surg 2001;36:63–7.

6. Garza JJ, Shew SB, Keshen TH, Dzakovic A, Jahoor F, Jaksic T.Energy expenditure in ill premature neonates. J Pediatr Surg2002;37:289–93.

7. Lloyd DA. Energy requirements of surgical newborn infants re-ceiving parenteral nutrition. Nutrition 1998;14:101–04.

8. Coss-Bu JA, Klish WJ, Walding D, Stein F, Smith EO, and Jef-ferson LS. Energy metabolism, nitrogen balance, and substrateutilization in critically ill children 1-3. Am J Clin Nutr 2001;74:664–69.

9. Schofield W. Predicting basal metabolic rate, new standards andreview of previous work. Hum Nutr Clin Nutr 1985;39C S1;5–41.

10. Koen FM, Joosten KFM, Verhoeven JJ, and Hazelzet JA. EnergyExpenditure and Substrate Utilization in Mechanically VentilatedChildren. Nutrition 1999;15:444–48.

11. Turi RA, Petros AJ, Eaton S, Fasoli L, Powis M, Basu R, Spitz L,Pierro A. Energy metabolism of infants and children with sys-temic inflammatory response syndrome and sepsis. Annals of Sur-gery 2001; 233: 581–87.

12. Duro D, Rising R, Cole C, Valois S, Cedillo M, Lifshitz F. Newequations for calculating the components of energy expenditure ininfants. J Pediatr 2002;140:534–9.

13. http://www.nal.usda.gov/fnic/dga/rda.pdf14. White MS, Shepherd RW, McEniery JA. Energy expenditure

measurements in ventilated critically ill children: within- andbetween-day variability. J Parenter Enteral Nutr 1999;23:300–4.

15. Pierro A, Jpnes MO, Hammond P, Donnell SC, Lloyd DA. A newequation to predict the resting energy expenditure of surgicalinfants. J Pediatr Surg 1994;29:1103–8.

16. Pierro A, Jones MO, Donnell SC. Total parenteral nutrition insurgical infants. Biochem Soc Trans 1998;26:131–6.

17. White MS, Shepherd RW, McEniery JA. Energy expenditure in100 ventilated, critically ill children: Improving the accuracy ofpredictive equations. Crit Care Med 2000; 28: 2307-2312.

18. Dietz WH, Bandini LG, Schoeller DA. Estimates of metabolicrate in obese and nonobese adolescents. J Pediatr 1991;118:146–49.

19. Mondejar EF, Duro Lombardo M, Perez de la Cruz AJ, MeridaMorales A, Torres Ruiz JM, Ferron Orihuela JA. Variations inoxygen consumption and carbon dioxide production during par-enteral nutrition. Intensive Care Med 1982;8:169–72.

20. Elwyn DH, Askanazi J, Kinney JM, Gump FE. Kinetics of energysubstrates. Acta Chir Scand Suppl 1981;507:209–19.

21. Koretz RL, Lipman TO, Klein S. AGA Technical Review onParenteral Nutrition. Gastroenterology 2001;121:970–1001.

22. Wilson DC, Cairns P, Halliday HL, Reid M, McClure G, DodgeJA. Randomised controlled trial of an aggressive nutritional regi-men in sick very low birthweight infants. Arch Dis Child FetalNeonatal Ed 1997;77:F4–11.

23. Kalhan SC, Kilic I. Carbohydrate as nutrient in the infant andchild: range of acceptable intake. Europ J Clin Nutr 1999;53:S94–S100.

24. Denne SC, Karn CA, Wang J, Liechty EA. Effect of intravenousglucose and lipid on proteolysis and glucose production in normalnewborns. Am J Physiol 1995;269:E361–7.

25. Sunehag AL, Haymond MW, Schanler RJ, Reeds PJ, Bier DM.Gluconeogenesis in very low birth weight infants receiving totalparenteral nutrition. Diabetes 1999;48:791–800.

26. Lafeber HN, Sulkers EJ, Chapman TE, Sauer PJ. Glucose pro-duction and oxidation in preterm infants during total parenteralnutrition. Pediatr Res 1990;28:153–7.

27. Jones MO, Pierro A, Hammond P, Nunn A, Lloyd DA. Glucoseutilization in the surgical newborn infant receiving total paren-teral nutrition. J Pediatr Surg 1993;28:1121–5.

28. Nose O, Tipton JR, Ament ME. Effect of the energy source onchanges in energy expenditure, respiratory quotient, and nitrogenbalance during total parenteral nutrition in children. Pediatr Res1987;21:538–41.

R. J. SHULMAN AND S. PHILLIPS602

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 17: Parenteral Nutrition in Infants and Children

29. Bresson JL, Narcy P, Putet G, Ricour C. Energy substrate utili-zation in infants receiving total parenteral nutrition with differentglucose to fat ratios. Pediatr Res 1989;25:645–8.

30. Sheridan RL, Yu YM, Prelack K, Young VR, Burke JF, Tomp-kins RG. Maximal parenteral glucose oxidation in hypermeta-bolic young children: a stable isotope study. J Parent Ent Nutr1998;22:212–6.

31. Schears GJ, Deutschman CS. Common nutritional issues in pe-diatric and adult critical care medicine. Crit Care Clinic 1997;13:669–90.

32. Cerra FB, Benitez MR, Blackburn GL, Irwin RS, Jeejeebhoy K,Katz DP, Pingleton SK, Pomposelli J, Rombeau JL, Shronts E,Wolfe RR, Zaloga GP. Applied nutrition in ICU patients. Chest1997;111:769–78.

33. Chan S, McCowen KC, Blackburn GL. Nutrition Management inthe ICU. Chest 1999; 115: 145–48.

34. Klein CJ, Stanek GS, Wiles CE 3rd. Overfeeding macronutrientsto critically ill adults: metabolic complications. J Am Diet Assoc1998;98:795–806.

35. Shikora SA, Benotti PN. Nutritional support of the mechanicallyventilated patient. Respir Care Clin N Am 1997;3: 69–90.

36. Robin AP, Carpentier YA, Askanazi J, Nordenstrom J, KinneyJM. Metabolic consequences of hypercaloric glucose infusions.Acta Chir Belg 1981;80:133–40.

37. Elwyn DH, Askanazi J, Kinney JM, Gump FE. Kinetics of energysubstrates. Acta Chir Scand Suppl 1981;507:209–19.

38. Talpers SS, Romberger DJ, Bunce SB, Pingleton SK. Nutrition-ally associated increased carbon dioxide production. Excess totalcalories vs high proportion of carbohydrate calories. Chest 1992;102:551–5.

39. Askanazi J, Weissman C, LaSala PA, Milic-Emili J, Kinney JM.Effect of protein intake on ventilatory drive. Anesthesiology1984;60:106–10.

40. Rodriguez JL, Askanazi J, Weissman C, Hensle TW, RosenbaumSH, Kinney JM. Ventilatory and metabolic effects of glucoseinfusions. Chest 1985;88:512–8.

41. Angelillo VA, Bedi S, Durfee D, Dahl J, Patterson AJ,O’Donohue WJ Jr. Effects of low and high carbohydrate feedingsin ambulatory patients with chronic obstructive pulmonary dis-ease and chronic hypercapnia. Ann Intern Med 1985;103:883–5.

42. Askanazi J, Rosenbaum SH, Hyman AI, Silverberg PA, Milic-Emili J, Kinney JM. Respiratory changes induced by the largeglucose loads of total parenteral nutrition. JAMA 1980;243:1444–7.

43. Burke JF, Wolfe RR, Mullany CJ, Mathews DE, Bier DM. Glu-cose requirements following burn injury. Parameters of optimalglucose infusion and possible hepatic and respiratory abnormali-ties following excessive glucose intake. Ann Surg 1979;190:274–85.

44. Tulikoura I, Huikuri K. Morphological fatty changes and functionof the liver, serum free fatty acids, and triglycerides during par-enteral nutrition. Scand J Gastroenterol 1982;17:177–85.

45. Asle Aarsland, David Chinkes, and Robert R. Wolfe. Contribu-tions of De Novo Synthesis of Fatty Acids to Total VLDL-Triglyceride Secretion during Prolonged Hyperglyce-mia/Hyperinsulinemia in Normal Man. J Clin Invest 1996;98:2008–17.

46. Perioperative total parenteral nutrition in surgical patients. TheVeterans Affairs Total Parenteral Nutrition Cooperative Studygroup. N Engl J Med 1991;325:525–32.

47. Kudsk KA, Croce MA, Fabian TC, Minard G, Tolley EA, PoretHA, Kuhl MR, Brown RO. Enteral versus parenteral feeding.Effects on septic morbidity after blunt and penetrating abdominaltrauma. Ann Surg 1992;215:503-11.

48. Moore FA, Moore EE, Jones TN, McCroskey BL, Peterson VM.TEN versus TPN following major abdominal trauma–reducedseptic morbidity. J Trauma 1989;29:916–22.

49. Moore FA, Feliciano DV, Andrassy RJ, McArdle AH, Booth FV,Morgenstein-Wagner TB, Kellum JM Jr, Welling RE, Moore EE.

Early enteral feeding, compared with parenteral, reduces postop-erative septic complications. The results of a meta-analysis. AnnSurg 1992;216:172–83.

50. Bistrian BR. Hyperglycemia and infection: which is the chickenand which is the egg? J Parenter Enteral Nutr 2001;25:180–1.

51. Engelich G, Wright DG, Hartshorn KL. Acquired disorders ofphagocyte function complicating medical and surgical illnesses.Clin Infect Dis 2001;33:2040–8.

52. Khaodhiar L, McCowen K, Bistrian B. Perioperative hypergly-cemia, infection or risk? Curr Opin Clin Nutr Metab Care 1999;2:79–82.

53. Gore DC, Chinkes D, Heggers J. Association of hyperglycemiawith increased mortality after severe burn injury. J Trauma 2001;51:540–44.

54. McCowen KC, Friel C, Sternberg J, Chan S, Forse RA, BurkePA, Bistrian BR. Hypocaloric total parenteral nutrition: effective-ness in prevention of hyperglycemia and infectious complica-tions–a randomized clinical trial. Crit Care Med 2000;28:3606–11.

55. Cerra FB et al. Applied Nutrition in ICU Patients: A ConsensusStatement of the American College of Chest Physicians. Chest1997;111:769–78.

56. Gutcher GR, Farrell PM. Intravenous infusion of lipid for theprevention of essential fatty acid deficiency in premature infants.Am J Clin Nutr 1991;54:1024–8.

57. Foote KD, MacKinnon MJ, Innis SM. Effect of early introductionof formula vs fat-free parenteral nutrition on essential fatty acidstatus of preterm infants. Am J Clin Nutr 1991;54:93–7.

58. O’Neill JA Jr, Caldwell MD, Meng HC. Essential fatty acid de-ficiency in surgical patients. Ann Surg 1977;185:535–42.

59. Farrell PM, Gutcher GR, Palta M, DeMets D. Essential fatty aciddeficiency in premature infants. Am J Clin Nutr 1988;48:220–9.

60. Friedman Z. Essential fatty acids revisited. Am J Dis Child 1980;134:397–408.

61. Mascioli EA, Lopes SM, Champagne C, Driscoll DF. Essentialfatty acid deficiency and home total parenteral nutrition patients.Nutrition 1996;12:245–9.

62. Jeppesen PB, Hoy CE, Mortensen PB. Differences in essentialfatty acid requirements by enteral and parenteral routes of admin-istration in patients with fat malabsorption. Am J Clin Nutr 1999;70:78–84.

63. Jeejeebhoy KN, Langer B, Tsallas G, Chu RC, Kuksis A, Ander-son GH. Total parenteral nutrition at home: studies in patientssurviving 4 months to 5 years. Gastroenterology 1976;71:943–53.

64. Barr LH, Dunn GD, Brennan MF. Essential fatty acid deficiencyduring total parenteral nutrition. Ann Surg 1981;193:304–11.

65. Wolfram G, Eckart J, Walther B, Zollner N. Factors influencingessential fatty acid requirement in total parenteral nutrition(TPN). J Parenter Enteral Nutr 1978;2:634–9.

66. Anderson TL, Muttart CR, Bieber MA, Nicholson JF, Heird WC.A controlled trial of glucose versus glucose and amino acids inpremature infants. J Pediatr 1979;94:947–51.

67. Nube M, Bos LP, vd Boomgaard DM, Hekkens WT. Energyintake and the appearance of 5,8,11-eicosatrienoic acid in serumlipids during parenteral nutrition without fat. J Parenter EnteralNutr 1982;6:134–9.

68. Spear ML, Stahl GE, Hamosh M, McNelis WG, Richardson LL,Spence V, Polin RA, Pereira GR, Hamosh P. Effect of heparindose and infusion rate on lipid clearance and bilirubin binding inpremature infants receiving intravenous fat emulsions. J Pediatr1988;112:94–8.

69. Kao LC, Cheng MH, Warburton D. Triglycerides, free fatty acids,free fatty acids/albumin molar ratio, and cholesterol levels inserum of neonates receiving long-term lipid infusions: controlledtrial of continuous and intermittent regimens. J Pediatr 1984;104:429–35.

70. Haumont D, Deckelbaum RJ, Richelle M, Dahlan W, CoussaertE, Bihain BE, Carpentier YA. Plasma lipid and plasma lipopro-

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 603

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 18: Parenteral Nutrition in Infants and Children

tein concentrations in low birth weight infants given parenteralnutrition with twenty or ten percent lipid emulsion. J Pediatr1989;115:787–93.

71. Haumont D, Richelle M, Deckelbaum RJ, Coussaert E, Car-pentier YA. Effect of liposomal content of lipid emulsions onplasma lipid concentrations in low birth weight infants receivingparenteral nutrition. J Pediatr 1992;121:759–63.

72. Putet G. Lipid metabolism of the micropremie. Clin Perinatol2000;27:57–69.

73. Rossner S, Eklund B, Freyschuss U, Hallberg D, Kaijser L, Ols-son A. Elimination of parenterally administered fat. Studies onremoval sites for intralipid in normo-and hyperlipidaemic sub-jects. Acta Chir Scand Suppl 1976;466:56–7.

74. Amin SB, Sinkin RA, McDermott MP, Kendig JW. Lipid intol-erance in neonates receiving dexamethasone for bronchopulmo-nary dysplasia. Arch Pediatr Adolesc Med 1999;153:795–800.

75. Hasselgren PO. Catabolic response to stress and injury: implica-tions for regulation. World J Surg 2000;24:1452–9.

76. Plank LD, Hill GL. Sequential metabolic changes following in-duction of systemic inflammatory response in patients with severesepsis or major blunt trauma. World J Surg 2000;24:630–8.

77. Dahn MS, Kirkpatrick JR, Blasier R. Alterations in the metabo-lism of exogenous lipid associated with sepsis. J Parenter EnteralNutr 1984;8:169–73.

78. Druml W, Fischer M, Ratheiser K. Use of intravenous lipids incritically ill patients with sepsis without and with hepatic failure.J Parenter Ent Nutr 1998;22:217–23.

79. Nordenstrom J, Carpentier YA, Askanazi J, Robin AP, ElwynDH, Hensle TW, Kinney JM. Metabolic utilization of intravenousfat emulsion during total parenteral nutrition. Ann Surg 1982;196:221–31.

80. Lundell K-H, Sabel K-G, Eriksson BO. Plasma metabolites aftera lipid load in infants with congenital heart disease. Acta Paediatr1999;88:718–23.

81. Zaidan H, Dhanireddy R, Hamosh M, Pramanik AK, ChowdhryP, Hamosh P. Effect of continuous heparin administration onIntralipid clearing in very low-birth-weight infants. J Pediatr1982;101:599–602.

82. Griffin EA, Bryan MH, Angel A. Variations in intralipid toler-ance in newborn infants. Pediatr Res 1983;17:478–81.

83. Benderly A, Rosenthal E, Levi J, Brook G. Effect of heparin onlipoprotein profile during parenteral fat infusions. J Parenter En-teral Nutr 1983;7:37–9.

84. Forget PP, Fernandes J, Begemann PH. Utilization of fat emul-sion during total parenteral nutrition in children. Acta PaediatrScand 1975;64:377–84.

85. Adamkin DH, Gelke KN, Andrews BF. Fat emulsions and hy-pertriglyceridemia. J Parenter Enteral Nutr 1984;8:563-7.

86. D’Harlingue A, Hopper AO, Stevenson DK, Shahin SM, KernerJA Jr. Limited value of nephelometry in monitoring the admin-istration of intravenous fat in neonates. J Parenter Enteral Nutr1983;7:55–8.

87. Zaidan H, Dhanireddy R, Hamosh M, Bengtsson-Olivecrona G,Hamosh P. Lipid clearing in premature infants during continuousheparin infusion: role of circulating lipases. Pediatr Res 1985;19:23–5.

88. Paust H, Park W, Brosicke H, Knoblach G, Helge H. Fat utiliza-tion in newborn infants with and without heparin administration.Comparative study with the 13C-triolein breath test. InfusionstherKlin Ernahr 1985;12:85–7.

89. Skeie B, Askanazi J, Rothkopf MM. Intravenous fat emulsionsand lung function: a review. Crit Care Med 1988;16:183–94.

90. Hageman JR, Hunt CE. Fat emulsions and lung function. ClinChest Med 1986;7:69-77.

91. Newball HH, Friedewald WT, Roberts B, Levy RI, Lenfant CJ.Effect of elevated triglycerides on the diffusing capacity of man.Am Rev Respir Dis 1975;112:83–8.

92. Wesson DE, Rich RH, Zlotkin SH, Pencharz PB. Fat overload

syndrome causing respiratory insufficiency. J Pediatr Surg 1984;19:777–8.

93. Hwang TL, Huang SL, Chen MF. Effects of intravenous fat emul-sion on respiratory failure. Chest 1990;97:934–8.

94. Helbock HJ, Motchnik PA, Ames BN. Toxic hydroperoxides inintravenous lipid emulsions used in preterm infants. Pediatrics1993;91:83–7.

95. Lavoie JC, Chessex P. The increase in vasomotor tone induced bya parenteral lipid emulsion is linked to an inhibition of prostacy-clin production. Free Radic Biol Med 1994;16:795–9.

96. Periera GR, Fox WW, Stanley CA, Baker L, Schwartz JG. De-creased oxygenation and hyperlipemia during intravenous fat in-fusions in premature infants. Pediatrics 1980;66:26–30.

97. Marks KH, Turner MJ, Rothberg AD. Effect of Intralipid infusionon transcutaneous oxygen and carbon dioxide tension in sickneonates. S Afr Med J 1987;72:389–91.

98. Sun SC, Ventura C, Verasestakul S. Effect of Intralipid-inducedlipaemia on the arterial oxygen tension in preterm infants. Resus-citation 1978;6:265–70.

99. Pia Lundman P, Eriksson M, Schenck-Gustafsson K. Transienttriglyceridemia decreases vascular reactivity in young, healthymen without risk factors for coronary heart disease. Circulation1997;96:3266–68.

100. Alwaidh MH, Bowden L, Shaw B, Ryan SW. Randomised trial ofeffect of delayed intravenous lipid administration on chronic lungdisease in preterm neonates. J Pediatr Gastroenterol Nutr 1996;22:303–6.

101. Waitzberg DL, Lotierzo PH, Logullo AF, Torrinhas RS, PereiraCC, Meier R. Parenteral lipid emulsions and phagocytic systems.Br J Nutr 2002;87:S49–57.

102. Shulman RJ. The effect of Intralipid on the cytotoxicity of leu-kocytes for herpes simplex virus-infected cells: a clinical con-cern? J Infect Dis 1983;148:181–2.

103. Adolph M. Lipid emulsions in parenteral nutrition. Ann NutrMetab 1999;43:1–13.

104. Guillou PJ. The effects of lipids on some aspects of the cellularimmune response. Proc Nutr Soc 1993;52:91–100.

105. Pomposelli JJ, Bistrian BR. Is total parenteral nutrition immuno-suppressive? New Horiz 1994;2:224–9.

106. Strunk RC, Murrow BW, Thilo E, Kunke KS, Johnson EG. Nor-mal macrophage function in infants receiving Intralipid by low-dose intermittent administration. J Pediatr 1985;106:640–5.

107. Wheeler JG, Boyle RJ, Abramson JS. Intralipid infusion in neo-nates: effects on polymorphonuclear leukocyte function. J Pedi-atr Gastroenterol Nutr 1985;4:453–6.

108. Usmani SS, Harper RG, Usmani SF. Effect of a lipid emulsion(Intralipid) on polymorphonuclear leukocyte functions in the neo-nate. J Pediatr 1988;113:132–6.

109. Dahlstrom KA, Goulet OJ, Roberts RL, Ricour C, Ament ME.Lipid tolerance in children receiving long-term parenteral nutri-tion: a biochemical and immunologic study. J Pediatr 1988;113:985–90.

110. Sirota L, Straussberg R, Notti I, Bessler H. Effect of lipid emul-sion on IL-2 production by mononuclear cells of newborn infantsand adults. Acta Paediatr 1997;86:410–3.

111. Usmani SS, Harper RG, Sia CG, Suntharalingam K, RobesonWR. In vitro effect of Intralipid on polymorphonuclear leukocytefunction in the neonate. J Pediatr 1986;109:710–2.

112. English D, Roloff JS, Lukens JN, Parker P, Greene HL, GhishanFK. Intravenous lipid emulsions and human neutrophil function.J Pediatr 1981;99:913–6.

113. Puri P, Reen DJ, Browne O, Guiney EJ. Immune status of theneonate maintained on total parenteral nutrition. Arch Dis Child1981;56:283–6.

114. Lenssen P, Bruemmer BA, Bowden RA, Gooley T, Aker SN,Mattson D. Intravenous lipid dose and incidence of bacteremiaand fungemia in patients undergoing bone marrow transplanta-tion. Am J Clin Nutr 1998;67:927–33.

115. Avila-Figueroa C, Goldmann DA, Richardson DK, Gray JE, Fer-

R. J. SHULMAN AND S. PHILLIPS604

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 19: Parenteral Nutrition in Infants and Children

rari A, Freeman J. Intravenous lipid emulsions are the majordeterminant of coagulase-negative staphylococcal bacteremia invery low birth weight newborns. Pediatr Infect Dis J 1998;17:10–7.

116. Carpentier YA, Dupont IE. Advances in intravenous lipid emul-sions. World J Surg 2000;24:1493–7.

117. Spear ML, Spear M, Cohen AR, Pereira GR. Effect of fat infu-sions on platelet concentration in premature infants. J ParenterEnteral Nutr 1990;14:165–8.

118. Dahlstrom KA, Goulet OJ, Roberts RL, Ricour C, Ament ME.Lipid tolerance in children receiving long-term parenteral nutri-tion: a biochemical and immunologic study. J Pediatr 1988;113:985–90.

119. Bistrian BR, Bothe A Jr, Blackburn GL, DeFriez AI. Low plasmacortisol and hematologic abnormalities associated with essentialfatty acid deficiency in man. J Parenter Enteral Nutr 1981;5:141–4.

120. Shulman RJ, Langston C, Schanler RJ. Pulmonary vascular lipiddeposition after administration of intravenous fat to infants. Pe-diatrics 1987;79:99–102.

121. Heird WC. Amino acids in pediatric and neonatal nutrition. CurrOpin Clin Nutr Metab Care 1998;1:73–8.

122. Brunton JA, Ball RO, Pencharz PB. Current total parenteral nu-trition solutions for the neonate are inadequate. Curr Opin ClinNutr Metab Care 2000;3:299–304.

123. Brown MR, Thunberg BJ, Golub L, Maniscalco WM, Cox C,Shapiro DL. Decreased cholestasis with enteral instead of intra-venous protein in the very low-birth-weight infant. J Pediatr Gas-troenterol Nutr 1989;9:21–7.

124. National Advisory Group on Standards and Practice Guidelinesfor Parenteral Nutrition. Safe practices for parenteral nutritionformulations. J Parent Ent Nutr 1998;22:49–66.

125. Brown MR, Thunberg BJ, Golub L, Maniscalco WM. Decreasedcholestasis with enteral instead of intravenous protein in the verylow-birth-weight infant. J Pediatr Gastroenterol Nutr 1989;9:21–7.

126. Moss RL, Amii LA. New approaches to understanding the etiol-ogy and treatment of total parenteral nutrition-associated chole-stasis. Semin Pediatr Surg 1999;8:140–7.

127. Senger H, Boehm G, Beyreiss K, Braun W, Raiha N. Evidencefor amino acid induced cholestasis in very-low-birth-weight in-fants with increasing enteral protein intake. Acta Paediatr Scand1986;75:724–8.

128. Woolf GM, Miller C, Kurian R, Jeejeebhoy KN. Nutritional ab-sorption in short bowel syndrome. Evaluation of fluid, calorie,and divalent cation requirements. Dig Dis Sci 1987;32:8–15.

129. Chin SE, Shepherd RW, Thomas BJ, Cleghorn GJ, Patrick MK,Wilcox JA, Ong TH, Lynch SV, Strong R. Nutritional support inchildren with end-stage liver disease: a randomized crossover trialof a branched-chain amino acid supplement. Am J Clin Nutr1992;56:158–63.

130. Fabbri A, Magrini N, Bianchi G, Zoli M, Marchesini G. Over-view of randomized clinical trials of oral branched-chain aminoacid treatment in chronic hepatic encephalopathy. J Parenter En-teral Nutr 1996;20:159–64.

131. Marchesini G, Bianchi G, Rossi B, Brizi M, Melchionda N. Nu-tritional treatment with branched-chain amino acids in advancedliver cirrhosis. J Gastroenterol 2000;35 Suppl 12:7–12.

132. Heyman MB. General and specialized parenteral amino acid for-mulations for nutrition support. J Am Diet Assoc 1990;90:401–8.

133. Koretz RL. Does nutritional intervention in protein-energy mal-nutrition improve morbidity or mortality? J Ren Nutr 1999;9:119–21.

134. Brewer ED. Pediatric experience with intradialytic parenteral nu-trition and supplemental tube feeding. Am J Kidney Dis 1999;33:205–7.

135. MacFie J. Cyclic parenteral nutrition. Nutrition 1997;13:46–55.136. Friel C, Bistrian B. Cycled total parenteral nutrition: is it more

effective? Am J Clin Nutr 1997;65:1078–9.

137. Takehara H, Hino M, Kameoka K. A new method of total par-enteral nutrition for surgical neonates: it is possible that cyclicTPN prevents intrahepatic cholestasis. Tokushima J Exp Med1990;37:97–102.

138. Hwang TL, Lue MC, Chen LL. Early use of cyclic TPN preventsfurther deterioration of liver functions for the TPN patients withimpaired liver function. Hepatogastroenterology 2000;47:1347–50.

139. Teitelbaum DH. Parenteral nutrition-associated cholestasis. CurrOpin Pediatr 1997;9:270–5.

140. Sokol RJ. Total parenteral nutrition-related liver disease. ActaPaed Sin 1997;38:418–28.

141. Kaufman SS. Prevention of parenteral nutrition-associated liverdisease in children. Pediatr Transplant 2002;6:37–42.

142. Crook MA, Hally V, Panteli JV. The importance of the refeedingsyndrome. Nutrition 2001;17:632–7.

143. Brooks MJ, Melnik G. The refeeding syndrome: an approach tounderstanding its complications and preventing its occurrence.Pharmacotherapy 1995;15:713–26.

144. Kimuru S, Nose O, Seino Y. Effects of alternate and simultaneousadministrations of calcium and phosphorous on calcium metabo-lism in children receiving total parenteral nutrition. J Parent EntNutr 1986;10:513–16.

145. Fitzgerald KA, MacKay MW. Calcium and phosphate solubilityin neonatal parenteral nutrient solutions containing TrophAmine.Am J Hosp Pharm 1986;43:88–93.

146. Lenz GT, Mikrut BA. Calcium and phosphate solubility in neo-natal parenteral nutrient solutions containing Aminosyn-PF orTrophAmine. Am J Hosp Pharm 1988;45:2367–71.

147. Dunham B, Marcuard S, Khazanie PG, Meade G, Craft T, Ni-chols K. The solubility of calcium and phosphorus in neonataltotal parenteral nutrition solutions. J Parenter Enteral Nutr 1991;15:608–11.

148. Porcelli PJ, Block SM. Increased parenteral nutrition calcium andphosphorus for very-low-birth-weight infants using computersoftware assisted ordering. J Am Coll Nutr 1997;16:283–7.

149. Laine L, Shulman RJ, Pitre D, Lifschitz CH, Adams J. Cysteineusage increases the need for acetate in neonates who receive totalparenteral nutrition. Am J Clin Nutr 1991;54:565–67.

150. Schanler RJ, Shulman RJ, Prestridge LL. Parenteral nutrientneeds of very low birth weight infants. J Pediatr 1994;125:961–8.

151. Prestridge LL, Schanler RJ, Shulman RJ, Burns PA, Laine LL.Effect of parenteral calcium and phosphorus therapy on mineralretention and bone mineral content in very low birth weight in-fants. J Pediatr 1993;122:761–8.

152. Dickerson RN. Parenteral nutrition. http://dcp.utmem.edu/lectures/TPNsolutionslecture.pdf

153. Rollins CJ, Elsberry VA, Pollack KA, Pollack PF, Udall JN Jr.Three-in-one parenteral nutrition: a safe and economical methodof nutritional support for infants. J Parenter Enteral Nutr 1990;14:290–4.

154. Didier ME, Fischer S, Maki DG. Total nutrient admixtures appearsafer than lipid emulsion alone as regards microbial contamina-tion: growth properties of microbial pathogens at room tempera-ture. J Parenter Enteral Nutr 1998;22:291–6.

155. D’Angio RG, Riechers KC, Gilsdorf RB, Constantino JM. Effectof the mode of lipid administration on parenteral nutrition-relatedinfections. Ann Pharmacother 1992;26:14–7.

156. http://www.fda.gov/cder/fdama/difconc.htm157. United States Pharmacopeia, Quality Assureance for Sterile Drug

Compounding, An Educational Source, 1998.158. McKinnon BT. FDA Safety Alert: Hazards of Precipitation As-

sociated With Parenteral Nutrition. Nutr Clin Prac 1996;11:59–65.

159. Hardy G, Ball P, McElroy B. Basic principles for compoundingall-in-one parenteral nutrition admixtures. Curr Opin Clin NutrMetab Care 1998;1:291–6.

160. Driscoll DF, Baptista RJ, Bistrian BR, Blackburn GL. Practical

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 605

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 20: Parenteral Nutrition in Infants and Children

considerations regarding the use of total nutrient admixtures. AmJ Hosp Pharm 1986;43:416–9.

161. Brown R, Quercia RA, Sigman R. Total nutrien admixture: areview. 1986;10–650–58.

162. National Advisory Group on Standards and Practice Guidelinesfor Parenteral Nutrition. Safe practices for parenteral nutritionformulations. J Parent Ent Nutr 1998;22:49–66.

163. Driscoll DF, Bhargava HN, Li L, Zaim RH, Babayan VK, Bis-trian BR. Physicochemical stability of total nutrient admixtures.Am J Health Syst Pharm 1995;52:623-34.

164. Shay DK, Fann LM, Jarvis WR. Respiratory distress and suddendeath associated with receipt of a peripheral parenteral nutritionadmixture. Infect Control Hosp Epidemiol 1997;18:814–17.

165. Department of Health and Human Services, Public Health Ser-vice; Food and Drug Administration, FDA Safety Alert: Hazardsof Precipitation Associated with Parenteral Nutrition, April 18,1994, http://www.fda.gov/cdrh/parnutr.pdf

166. Parry VA, Harrie KR, McIntosh-Lowe NL. Effect of variousnutrient ratios on the emulsion stability of total nutrient admix-tures. Am J Hosp Pharm 1986;43:3017–22.

167. Bethune K, Allwood M, Grainger C, Wormleighton C. Use offilters during the preparation and administration of parenteralnutrition: position paper and guidelines prepared by a Britishpharmaceutical nutrition group working party. Nutrition 2001;17:403-8.

168. Driscoll DF, Bacon MN, Bistrian BR. Effects of in-line filtrationon lipid particle size distribution in total nutrient admixtures. JParenter Enteral Nutr 1996;20:296–301.

169. Rollins CJ, Elsberry VA, Pollack KA, Pollack PF, Udall JN Jr.Three-in-one parenteral nutrition: a safe and economical methodof nutritional support for infants. J Parenter Enteral Nutr 1990;14:290–4.

170. Bullock L, Fitzgerald JF, Walter WV. Emulsion stability in totalnutrient admixtures containing a pediatric amino acid formula-tion. J Parenter Enteral Nutr 1992;16:64–8.

171. Smith JL, Canham JE, Kirkland WD, Wells PA. Effect of In-tralipid, amino acids, container, temperature, and duration of stor-age on vitamin stability in total parenteral nutrition admixtures. JParenter Enteral Nutr 1988;12:478–83.

172. Trissel LA, Gilbert DL, Martinez JF, Baker MB, Walter WV,Mirtallo JM. Compatibility of medications with 3-in-1 parenteralnutrition admixtures. J Parenter Enteral Nutr 1999;23:67–74.

173. American Medical Association Department of Foods and Nutri-tion. Multivitamin preparations for parenteral use. A statement bythe Nutrition Advisory Group. J Parent Ent Nutr 1979;3: 258–62.

174. Green HL, Moore ME, Phillips B. Evaluation of a pediatric mul-tiple vitamin preparation for total parenteral nutrition. II. Bloodlevels of vitamins A,D, and E. Pediatrics 1986;77:539–47.

175. Moore MC, Greene HL, Phillips B, Franck L, Shulman RJ, Mur-rell JE, Ament ME. Evaluation of a pediatric multiple vitaminpreparation for total parenteral nutrition in infants and children. I.Blood levels of water-soluble vitamins. Pediatrics 1986;77:530–8.

176. Department of Health and Human Services Food and Drug Ad-ministration. Parenteral multivitamin products; Drugs for HumanUse; Drug efficacy study implementation; Amendment. FederalRegister. April 20, 2000;56:21200–21201.

177. Alade SL, Brown RE, Paquet A Jr. Polysorbate 80 and E-Feroltoxicity. Pediatrics 1986;77:593–7.

178. MacDonald MG, Getson PR, Glasgow AM, Miller MK, BoeckxRL, Johnson EL. Propylene glycol: increased incidence of sei-zures in low birth weight infants. Pediatrics 1987;79:622–5.

179. Shulman RJ, Phillips SM, Maciejewski S. An update on a globalshortage of intravenous multivitamins. J Pediatr GastroenterolNutr 1999;28:361.

180. Greene HL, Hambidge KM, Schanler R, Tsang RC. Guidelinesfor the use of vitamins, trace elements, calcium, magnesium, andphosphorus in infants and children receiving total parenteral nu-trition: report of the Subcommittee on Pediatric Parenteral Nutri-

ent Requirements from the Committee on Clinical Practice Issuesof the American Society for Clinical Nutrition. Am J Clin Nutr1988;48:1324–42.

181. Drittij-Reijnders MJ, Sels JP, Rouflart M, Thijssen HH. VitaminK status and parenteral nutrition; the effect of Intralipid on plasmavitamin K1 levels. Eur J Clin Nutr 1994;48:525–7.

182. Lennon C, Davidson KW, Sadowski JA, Mason JB. The vitaminK content of intravenous lipid emulsions. J Parenter Enteral Nutr1993;17:142–4.

183. Darlow BA, Graham PJ. Vitamin A supplementation for prevent-ing morbidity and mortality in very low birthweight infants. Co-chrane Database Syst Rev 2000;(2):CD000501.

184. Henton DH, Merritt RJ. Vitamin A sorption to polyvinyl andpolyolefin intravenous tubing. J Parenter Enteral Nutr 1990;14:79–81.

185. Pediatric Parenteral Multivitamin Products; Drug Efficacy StudyImplementation; Announcement of Marketing Conditions; Fed-eral Register: January 26, 2000 (Volume 65, Number 17)] [No-tices] [Page 4253–4256].

186. Porcelli PJ, Greene HL, Adcock EW. Retinol (vitamin A) andriboflavin (vitamin B2) administration and metabolism in verylow birth weight infants. Semin Perinatol 1992;16:170–80.

187. Smith JL, Canham JE, Kirkland WD, Wells PA. Effect of In-tralipid, amino acids, container, temperature, and duration of stor-age on vitamin stability in total parenteral nutrition admixtures. JParenter Enteral Nutr 1988;12:478–83.

188. Smith JL, Canham JE, Wells PA. Effect of phototherapy light,sodium bisulfite, and pH on vitamin stability in total parenteralnutrition admixtures. J Parenter Enteral Nutr 1988;12:394–402.

189. Greene HL, Phillips BL, Franck L, Fillmore CM, Said HM, Mur-rell JE, Moore ME, Briggs R. Persistently low blood retinol levelsduring and after parenteral feeding of very low birth weight in-fants: examination of losses into intravenous administration setsand a method of prevention by addition to a lipid emulsion. Pe-diatrics 1987;79:894–900.

190. Dahl GB, Svensson L, Kinnander NJ, Zander M, Bergstrom UK.Stability of vitamins in soybean oil fat emulsion under conditionssimulating intravenous feeding of neonates and children. JParenter Enteral Nutr 1994;18:234–9.

191. Silvers KM, Sluis KB, Darlow BA, McGill F, Stocker R, Win-terbourn CC. Limiting light-induced lipid peroxidation and vita-min loss in infant parenteral nutrition by adding multivitaminpreparations to Intralipid. Acta Paediatr 2001;90:242–9.

192. Baeckert PA, Greene HL, Fritz I, Oelberg DG, Adcock EW.Vitamin concentrations in very low birth weight infants givenvitamins intravenously in a lipid emulsion: measurement of vita-mins A, D, and E and riboflavin. J Pediatr 1988;113:1057–65.

193. Cairns PA, Stalker DJ. Carnitine supplementation of parenterallyfed neonates. Cochrane Database Syst Rev 2000;(4):CD000950.

194. Penn D, Schmidt-Sommerfeld E, Pascu F. Decreased tissue car-nitine concentrations in newborn infants receiving total parenteralnutrition. J Pediatr 1981;98:976–8.

195. Worthley LI, Fishlock RC, Snoswell AM. Carnitine deficiencywith hyperbilirubinemia, generalized skeletal muscle weaknessand reactive hypoglycemia in a patient on long-term total paren-teral nutrition: treatment with intravenous L-carnitine. J ParenterEnteral Nutr 1983;7:176–80.

196. Borum PR. Should carnitine be added to parenteral nutrition so-lutions? Nutr Clin Prac 2000;15:153–4.

197. Reed MD, Bertino JS Jr, Halpin TC Jr. Use of intravenous irondextran injection in children receiving total parenteral nutrition.Am J Dis Child 1981;135:829–31.

198. Halpin TC Jr, Bertino JS, Rothstein FC, Kurczynski EM, ReedMD. Iron-deficiency anemia in childhood inflammatory boweldisease: treatment with intravenous iron-dextran. J Parenter En-teral Nutr 1982;6:9–11.

199. Surico G, Muggeo P, Muggeo V, Lucarelli A, Martucci T,Daniele M, Rigillo N. Parenteral iron supplementation for the

R. J. SHULMAN AND S. PHILLIPS606

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003

Page 21: Parenteral Nutrition in Infants and Children

treatment of iron deficiency anemia in children. Ann Hematol2002;81:154–7.

200. Mamula P, Piccoli DA, Peck SN, Markowitz JE, Baldassano RN.Total dose intravenous infusion of iron dextran for iron-deficiency anemia in children with inflammatory bowel disease. JPediatr Gastroenterol Nutr 2002;34:286–90.

201. Burns DL, Mascioli EA, Bistrian BR. Parenteral iron dextrantherapy: a review. Nutrition 1995;11:163–8.

202. Kumpf VJ. Parenteral iron supplementation. Nutr Clin Pract1996;11:139–46.

203. Charytan C, Levin N, Al-Saloum M, Hafeez T, Gagnon S, VanWyck DB. Efficacy and safety of iron sucrose for iron deficiencyin patients with dialysis-associated anemia: North Americanclinical trial. Am J Kidney Dis 2001;37:300–7.

204. Michael B, Coyne DW, Fishbane S, Folkert V, Lynn R, Nissen-son AR, Agarwal R, Eschbach JW, Fadem SZ, Trout JR, StrobosJ, Warnock DG. Sodium ferric gluconate complex in hemodialy-sis patients: adverse reactions compared to placebo and iron dex-tran. Kidney Int 2002;61:1830–9.

205. Pollak A, Hayde M, Hayn M, Herkner K, Lombard KA, Lubec G,Weninger M, Widness JA. Effect of intravenous iron supplemen-tation on erythropoiesis in erythropoietin-treated premature in-fants. Pediatrics 2001;107:78–85.

206. Hardy G, Reilly C. Technical aspects of trace element supple-mentation. Curr Opin Clin Nutr Metab Care 1999;2:277–85.

207. Van Gossum A, Neve J. Trace element deficiency and toxicity.Curr Opin Clin Nutr Metab Care 1998;1:499–507.

208. Allwood MC. Aluminium in parenteral nutrition admixtures: anunnecessary risk? Nutrition 1999;15:958–9.

209. Klein GL. Aluminum in parenteral solutions revisited–again. AmJ Clin Nutr 1995;61:449–56.

210. Pluhator-Murton MM, Fedorak RN, Audette RJ, Marriage BJ,Yatscoff RW, Gramlich LM. Trace element contamination oftotal parenteral nutrition. 1. Contribution of component solutions.J Parenter Enteral Nutr 1999;23:222–7.

211. Klinger G, Shamir R, Singer P, Diamond EM, Josefsberg Z,Sirota L. Parenteral selenium supplementation in extremely lowbirth weight infants: inadequate dosage but no correlation withhypothyroidism. J Perinatol 1999;19:568–72.

212. Hatanaka N, Nakaden H, Yamamoto Y, Matsuo S, Fujikawa T,Matsusue S. Selenium kinetics and changes in glutathione per-oxidase activities in patients receiving long-term parenteral nu-trition and effects of supplementation with selenite. Nutrition2000;16:22–6.

213. Yusuf SW, Rehman Q, Casscells W. Cardiomyopathy in associa-tion with selenium deficiency: a case report. J Parenter EnteralNutr 2002;26:63–6.

214. Mouser JF, Hak EB, Helms RA, Christensen ML, Storm MC.Chromium and zinc concentrations in pediatric patients receivinglong-term parenteral nutrition. Am J Health Syst Pharm 1999;56:1950–6.

215. Bougle D, Bureau F, Deschrevel G, Hecquard C, Neuville D,Drosdowsky M, Duhamel JF. Chromium and parenteral nutritionin children. J Pediatr Gastroenterol Nutr 1993;17:72–4.

216. Moukarzel AA, Song MK, Buchman AL, Vargas J, Guss W,McDiarmid S, Reyen L, Ament ME. Excessive chromium intakein children receiving total parenteral nutrition. Lancet 1992;339:385–8.

217. Hak EB, Storm MC, Helms RA. Chromium and zinc contamina-tion of parenteral nutrient solution components commonly used ininfants and children. Am J Health Syst Pharm 1998;55:150–4.

218. Moukarzel AA, Ament ME, Buchman A, Dahlstrom KA, VargasJ. Renal function of children receiving long-term parenteral nu-trition. J Pediatr 1991;119:864–8.

219. Shulman RJ. Zinc and copper balance studies in infants receivingtotal parenteral nutrition. Am J Clin Nutr 1989;49:879–83.

220. Dickerson RN. Manganese intoxication and parenteral nutrition.Nutrition 2001;17:689–693.

221. Quaghebeur G, Taylor WJ, Kingsley DP, Fell JM, Reynolds AP,Milla PJ. MRI in children receiving total parenteral nutrition.Neuroradiology 1996;38:680–3.

222. Fell JM, Reynolds, AP, Meadows N. Manganese toxicity in chil-dren receiving long-term parenteral nutrition. Lancet 1996;347:1218–21.

223. Bertinet DB, Tinivella M, Balzola FA, de Francesco A, Davini O,Rizzo L, Massarenti P, Leonardi MA, Balzola F. Brain manga-nese deposition and blood levels in patients undergoing homeparenteral nutrition. J Parenter Enteral Nutr 2000;24:223–7.

224. Masumoto K, Suita S, Taguchi T, Yamanouchi T, Nagano M,Ogita K, Nakamura M, Mihara F. Manganese intoxication duringintermittent parenteral nutrition: report of two cases. J ParenterEnteral Nutr 2001;25:95–9.

225. Fok TF, Chui KK, Cheung R, Ng PC, Cheung KL, Hjelm M.Manganese intake and cholestatic jaundice in neonates receivingparenteral nutrition: a randomized controlled study. Acta Paediatr2001;90:1009–15.

226. Fitzgerald K, Mikalunas V, Rubin H, McCarthey R, VanagunasA, Craig RM. Hypermanganesemia in patients receiving totalparenteral nutrition. J Parenter Enteral Nutr 1999;23:333–6.

227. Takagi Y, Okada A, Sando K, Wasa M, Yoshida H, Hirabuki N.On-off study of manganese administration to adult patients un-dergoing home parenteral nutrition: new indices of in vivo man-ganese level. J Parenter Enteral Nutr 2001;25:87–92.

228. C. Garza, personal communication, 1980.229. Pupim LB, Flakoll PJ, Brovillette JR. Intradialytic parenteral nu-

trition improves protein and energy homeostasis in chronic he-modialysis patients. J Clin Invest 2002;110:483–92.

230. Haas C, Genzel-Boroviczeny O, Koletzko B. Losses of VitaminA and E in parenteral nutrition suitable for premature infants. EurJ Clin Nutr 2002;56:96–12.

PARENTERAL NUTRITION IN INFANTS AND CHILDREN 607

J Pediatr Gastroenterol Nutr, Vol. 36, No. 5, May 2003