review article glutaminesupplementationinsickchildren ...rats with experimental nec, gln reduces...

Hindawi Publishing CorporationJournal of Nutrition and MetabolismVolume 2011, Article ID 617597, 41 pagesdoi:10.1155/2011/617597

Review Article

Glutamine Supplementation in Sick Children: Is It Beneficial?

Elise Mok1, 2, 3 and Régis Hankard1, 2

1 INSERM Centre D’Investigation Clinique 802, Centre Hospitalier Universitaire de Poitiers, 86021 Poitiers Cedex, France2 Pédiatrie Multidisciplinaire-Nutrition de l’Enfant, Centre Hospitalier Universitaire de Poitiers, 86021 Poitiers Cedex, France3 Child Health Clinical Research Centre, The Montréal Children’s Hospital, McGill University Health Centre,Montréal, QC, Canada H3H 1P3

Correspondence should be addressed to Régis Hankard, [email protected]

Received 1 August 2011; Accepted 28 September 2011

Academic Editor: Johannes B. van Goudoever

Copyright © 2011 E. Mok and R. Hankard. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

The purpose of this review is to provide a critical appraisal of the literature on Glutamine (Gln) supplementation in variousconditions or illnesses that affect children, from neonates to adolescents. First, a general overview of the proposed mechanisms forthe beneficial effects of Gln is provided, and subsequently clinical studies are discussed. Despite safety, studies are conflicting, partlydue to different effects of enteral and parenteral Gln supplementation. Further insufficient evidence is available on the benefitsof Gln supplementation in pediatric patients. This includes premature infants, infants with gastrointestinal disease, childrenwith Crohn’s disease, short bowel syndrome, malnutrition/diarrhea, cancer, severe burns/trauma, Duchenne muscular dystrophy,sickle cell anemia, cystic fibrosis, and type 1 diabetes. Moreover, methodological issues have been noted in some studies. Furthermechanistic data is needed along with large randomized controlled trials in select populations of sick children, who may eventuallybenefit from supplemental Gln.

1. Introduction

Glutamine (Gln) is the most abundant amino acid in themuscle and plasma of humans [1]. Although Gln is a nones-sential neutral amino acid, it is necessary for optimal growthof mammalian cells in tissue culture [2] and has importantphysiological functions. Apart from providing nitrogen forprotein synthesis, Gln is a precursor for nucleic acids,nucleotides [3], hexosamines [4], the nitric oxide precursor-arginine (Arg) [5], and the major antioxidant-glutathione[4, 6]. Gln is also an important oxidative fuel for rapidlyproliferating cells such as those of the gastrointestinal tract[7] and immune system [3], reticulocytes [8], fibroblasts[9], and so on. It plays a central role in nitrogen transportbetween tissues [10], specifically from muscle to gut, kidney,and liver. In addition to its role as a gluconeogenic substratein the liver, kidney [11], and intestine [12], Gln is involved inthe renal handling of ammonia, serving as a regulator of acid-base homeostasis [13]. Present data also indicate that Glnfunctions as a signalling molecule [14], particularly undercatabolic conditions.

Traditionally Gln is considered a nonessential aminoacid, because it is synthesized by most tissue (skeletalmuscle being the main producer and storage site) [15].Gln synthetase catalyzes the terminal step in Gln de novosynthesis and is a key enzyme in Gln metabolism [16, 17].In mammals, Gln synthetase expression is regulated bytranscriptional and posttranscriptional mechanisms, that is,increasing Gln synthetase mRNA in response to stress (e.g.,glucocorticoids) and regulation of Gln synthetase proteinturnover in response to its product (Gln concentrations)[18]. The importance of Gln at the whole body level ishighlighted by the report of severe brain malformationresulting in multiorgan failure and neonatal death in 2 unre-lated newborns with congenital Gln synthetase deficiency, inwhom Gln was largely absent in plasma, urine, and cerebralspinal fluid [19].

Under normal conditions, Gln is released into circu-lation for consumption by other tissue, whereas duringcatabolic stress the production of Gln may be insufficientto meet the increased requirements by the gut, immunesystem/inflammatory cells, liver, and kidneys. Demands are

2 Journal of Nutrition and Metabolism

partly met by skeletal muscle proteolysis and release of largeamounts of Gln to maintain normal concentrations in theplasma, resulting in depletion of Gln stores. Based on thisabundant evidence, Lacey and Wilmore [10] suggested thatGln may become a conditionally essential amino acid for thecritically ill.

In paediatrics, several researchers have studied the effi-cacy of supplemental Gln in premature infants of low birthweight (LBW), who are highly stressed and have low energyand protein reserves [20]. Similar to premature neonates, Glnsupplementation may also be beneficial for other childhoodconditions including gastrointestinal disease, malnutrition,cancer, severe burns/trauma as well as chronic diseases ofchildhood. However, less data is available on the effects ofsupplemental Gln in older infants and children with variousdiseases.

In addition to being sick and highly stressed, childrenare also in the process of growth and development. Hence,specific research on the role of Gln in pediatric patients isnecessary. The main purpose of this manuscript is to providea critical review of the literature on Gln supplementationin various conditions/illnesses that affect children (fromneonates to adolescents). First the proposed mechanisms ofGln action are reviewed in a general context, followed by adetailed description and critique of the clinical research onGln supplementation in children.

1.1. Glutamine Mechanisms of Action. While it is well estab-lished that Gln is a protein precursor as well as a major fueland nucleotide substrate for rapidly proliferating cells (e.g.,gut and immune system) [3, 7], additional mechanistic datahas emerged to explain the apparent benefits of Gln. Gln canregulate the expression of many genes related to metabolism,signal transduction, cell defense, and repair and can activateintracellular signaling pathways [14]. In brief, Gln seemsto affect antioxidant capacity, tissue protection, immune,and metabolic function [21] as well as protein synthesisand degradation [14] (Figure 1). The postulated mechanismsremain speculative and are by no means mutually exclusive,since Gln can provoke a number of different effects thatinteract with one another.

1.2. Antioxidant Capacity

1.2.1. Glutathione. Gln is a precursor of the glutamate (Glu),for glutathione (L-γ-glutamyl-L-cysteinyl-glycine) synthesis,an important antioxidant in many cell types [22]. Glu-tathione is present in the cell in reduced (GSH) and oxidized(GSSG) forms. The ratio of reduced-to-oxidized glutathioneis the major regulator of the cellular redox potential thatdetermines the antioxidant capacity of the cell [14, 22, 23].The effectiveness of glutathione protection in individualtissue depends on the tissue concentration of glutathione aswell as the capacity of the tissue to import GSH and to exportGSSG [24].

In vivo experiments in rats demonstrate that adminis-tration of Gln before ischemia/reperfusion injury or surgicalmanipulation can enhance GSH concentrations and provide

Tissue protection

Antioxidant capacity

Protein degradation

Metabolic function

Protein synthesis

Glutamine

Immune function

Figure 1: Schematic representation of the mechanism of glutamineaction.

protection against oxidative stress in various tissues (e.g., car-diac, intestinal, and lung) [25, 26]. Further, the effectivenessof Gln in preventing liver damage in neonatal sepsis appearsto be mediated via glutathione synthesis [27]. In humans,Gln supplementation can attenuate GSH depletion in skeletalmuscle following surgical trauma [28].

During critical illness, muscle concentrations of GSHdecrease and a change in the redox status occurs, indicativeof an elevated GSSG [24]. Moreover, there is a correlationbetween the concentrations of Gln and GSH [24]. Shiftingthe GSH/GSSG redox toward oxidizing conditions acti-vates several signaling pathways, such as c-Jun N-terminalkinase (JNK), apoptosis signal-regulated kinase-1 (ASK-1),mitogen-activated protein kinase (MAPK), and the tran-scription factor nuclear factor-kappaB (NF-κB: a stimulatorof the synthesis of proinflammatory cytokines and adhesionmolecules) [22, 23, 29]. Evidence also implicates oxidativestress as a potential regulator of NF-κB transactivation byMAPKs (in particular extracellular signal-regulated kinase1/2 (ERK1/2) and p38) [30] which could lead to increasedproteolysis in muscle [31]. Moreover, the cellular redoxstatus seems to be related to the degree of muscle proteindegradation [32, 33]. Likewise, there is significant literatureon the role of inflammatory cytokines (interleukin-1, -6,tumour necrosis factor-alpha (TNF-α)) and muscle wasting[34].

Gln metabolism via entry into the citric acid cycle mayallow the activation of malic enzyme which will resultin an increase in NADPH [15] and subsequently increasethe GSH/GSSG ratio [14]. Administration of Gln leads toan increased ratio of GSH/GSSG and reduces the activityof redox sensitive kinases subsequently preventing NF-κBactivation and thus inhibiting the inflammatory response[23].

Experiments from our group in the mdx mouse modelof muscular dystrophy (a condition associated with severemuscle wasting) showed that in vivo Gln administrationcan reduce GSSG in dystrophic skeletal muscle, hence,increasing the ratio of GSH/GSSG [35]. This was associated

Journal of Nutrition and Metabolism 3

with decreased activation of MAPK (ERK1/2) pathway [35].Similar effects were observed in muscle of control mice;however, the magnitude was less [35]. Thus, in muscle tissue,Gln might affect the cellular redox state involving MAPKpathway.

1.3. Tissue Protection

1.3.1. Heat Shock Protein (HSP). The HSPs serve as molec-ular chaperones that appear to repair denatured/injuredproteins or promote their degradation following irreparableinjury. Gln has cell-protective effects, as a potent enhancerof the expression of HSP25, HSP70, HSP72, and heme-oxygenase-1 in cell culture [36], in multiple organs of bothstressed and unstressed animal models [37], as well as inhumans [38, 39]. However, Gln depletion during the stressresponse can impair the expression of the major inducibleHSP (HSP70), as shown in human lymphocytes [40]. Andrecent evidence suggests that HSP70 expression is requiredfor Gln’s protection against tissue injury and for attenuationof NF-κB activation and proinflammatory cytokine release[41].

In rats, preoperative administration of Gln inducesHSP70 expression and attenuates the inflammatory responseby regulating nitric oxide synthase (NOS) activity in heart,lung, and liver [42]. Gln is a well-known precursor forArg [5], which can increase nitric oxide formation as aresult of enhanced NOS activity [6]. However, in variousmodels of human intestinal cells, Gln does not furtherincrease nitric oxide production or inducible NOS mRNAfollowing proinflammatory cytokine stimulation [43]. Thus,Gln’s effects on NOS activity might be tissue- or condition-specific.

The survival-promoting effects of HSP70 can also beattributed in part to the suppression of apoptosis, sincereduced HSP expression in Gln-deprived cells together withtheir impaired antioxidant capacity may make them moresusceptible to apoptosis [29].

1.3.2. Apoptosis. Gln starvation has been shown to induceapoptosis in intestinal epithelial cells [44] and also ren-ders human monocytic cells more susceptible to apoptosisinduced by Fas ligand, heat shock, or TNF-α stimulation[45]. In HeLa cells, Gln might also suppress ASK-1 andJNK/stress-activated protein kinase (SAPK) activation byFas ligand [46]. The effect of Gln in delaying spontaneousapoptosis in neutrophils may be mediated by the antioxidanteffects of glutathione [47]. Furthermore, Gln may protectactivated T cells from apoptosis, partially by upregulatingglutathione and Bcl-2 expression and inhibiting Fas [48].

1.3.3. Intestinal Barrier Function. As an important fuel forintestinal tissue and gut-associated lymphoid tissue, Glnmay contribute to gut barrier function [49]. Experimentaldata from neonatal animal models also support the ben-eficial effects of Gln on gastrointestinal development andfunction [50–59]. Gln is also involved in the biosynthesisof hexosamines which are important for maintaining gutwall integrity via surface mucin and glycoprotein-forming

intracellular tight junctions and thus may protect againstbacterial translocation [4, 60].

1.4. Immune Function. As a major fuel for immune cells,Gln is known to modulate immune function. More recently,Gln has also been shown to have anti-inflammatory effects,modulating cytokine production, both in vitro [36, 61–63]and in vivo [41, 51, 64], possibly through decreased NF-κBactivation [41, 62], a major transcription factor regulatingimmune and inflammatory responses. In neonatal mice andrats with experimental NEC, Gln reduces intestinal injury[54, 55], via mechanisms inhibiting inflammatory cytokinerelease [54].

Gln’s anti-inflammatory effects may also be related toenhanced HSP expression [36, 41]. The induction of HSPresponse can attenuate proinflammatory cytokine release,which in turn depends on the cellular redox potential andconsequently is regulated by the intracellular GSH/GSSGratio [14] (as previously described).

1.5. Tissue Metabolic Function. Gln can preserve tissue-meta-bolic function in stress states. For instance, Gln enhan-ces myocardial tissue-metabolic function after ischemia/reperfusion injury in rats [26]. Gln can also enhance ATPlevels in oxidant stressed endothelial cells [65].

1.5.1. Glucose Metabolism. While Gln plays an importantrole in gluconeogenesis [11, 12, 66], evidence also suggeststhat Gln can improve insulin sensitivity and glucose disposalin patients suffering from critical illness [21, 67], a conditionfrequently associated with insulin resistance and subsequenthyperglycemia. Gln also plays a role as a signaling moleculein amino acid- and glucose-stimulated insulin secretion[68]. Interestingly in rats with diet-induced obesity, Glnsupplementation induces insulin resistance in adipose tissueand reduces adipose mass, consequently attenuating insulinresistance and activation of JNK and inhibitory kappaBkinase subunit beta in liver and muscle, thus improvinginsulin signaling [69]. These data suggest that Gln can bene-ficially influence insulin-dependent glucose metabolism.

1.6. Protein Synthesis and Degradation. Gln appears toregulate protein turnover in cultured rat skeletal myotubes,stimulating protein synthesis in stressed myotubes whileinhibiting protein degradation in long-lived proteins. Thismay be related to the increase in HSP70 [70]. There is alsoabundant literature to suggest that amino acids affect proteinturnover via the mammalian target of rapamycin (mTOR)pathway [71, 72].

1.6.1. Protein Synthesis. Amino acids, particularly branchedchain amino acids, for example, leucine (Leu) stimulateskeletal muscle protein synthesis via the activation of mTORwhich in turn activates p70 ribosomal S6 kinase (p70S6K) anddephosphorylates eukaryotic initiation factor 4E-bindingprotein 1 (4E-BP 1), stimulating translation and proteinsynthesis [71, 72]. Gln can induce growth and maturationof neonatal rat cardiomyocytes, which is associated with


an increase in the mRNAs-encoding contractile proteins andmetabolic enzymes via the activation of protein kinase Aand mTOR [73]. However, the action of Gln seems to becell-type specific. For instance, in C2C12 myogenic cells,Gln and Leu have opposite effects on the mTOR pathway[74]. Whereas Leu activates this pathway, Gln inhibits it bydecreasing the phosphorylation states of mTOR (on serine(Ser)2448), p70S6K, and 4E-BP1, with no effect on proteinsynthesis [74].

1.6.2. Protein Degradation. The major proteolytic pathwaysin organs such as the liver, muscle, and intestine includethe autophagic/lysosomal (cathepsins), the calcium activated(calpains), and the ATP-ubiquitin-proteosome pathway [14,34]. The proteosome system (26S) is a highly selective prote-olytic pathway [71]. In visceral tissues (e.g., liver), autophagyis, however, the major proteolytic pathway and the onlypathway known to be regulated by plasma amino acids (inliver and skeletal muscle) [71, 72]. In autophagic proteolysis,several amino acids have direct regulatory potential, possiblyvia a plasma membrane amino acid receptor/sensor andsubsequent intracellular signaling [71].

Another line of evidence suggests that amino acidsactivate mTOR pathway which in turn suppresses proteinbreakdown by the autophagy/lysosomal pathway [71, 72].Amino acids can also control autophagic lysosomal proteoly-sis by inhibiting MAPK (ERK1/2) phosphorylation [75]. Glnmay cause its antiproteolytic effect through osmotic swelling[76], involving p38 MAPK pathway [77]. An increase incellular hydration acts as an anabolic signal, whereas cellshrinkage is catabolic, and there is a close relation in theregulation of cell volume, Gln, and protein catabolism [23].

On the other hand in the human gut, enteral Gln mayattenuate ubiquitin-dependent proteolysis as demonstratedby decreased ubiquitin mRNA, whereas mRNAs encoding forcathepsins or calpains were not affected [78]. Furthermore,in lung and muscle, Gln can also regulate its own productionthrough a posttranscriptional mechanism in which Glnregulates Gln synthetase protein degradation [16, 17, 79],by facilitating its degradation by the 26S proteosome [18].Thus, the presence of Gln could have a protein-sparing effect,sparing amino acids for protein accretion [80–82].

1.7. Glutamine in the Neonatal Period. Gln is the predomi-nant amino acid supplied to the fetus through the placentaand is specifically suited for its rapid development [83, 84].While normally present in the enteral diet, Gln has beenexcluded from parenteral nutrition (PN) because of lowsolubility and instability in solution. In the first weeks of life,however, premature infants receive most of their nutrientsfrom PN which is Gln-free [85]. The sudden cessation ofGln supply from the mother to premature infants, whoare already stressed and undergoing rapid growth, maybe detrimental [86]. Whereas plasma Gln concentrationsnormally increase during the first days of life in newborninfants breastfed ad libitum [87], selective amino aciddeficiencies have been reported in neonates suffering fromacute illness, including reductions in serum Gln and Arg in

infants who have necrotizing enterocolitis (NEC) that maypredispose them to the illness [88]. It has been suggestedthat in catabolic conditions premature infants are not ableto synthesize sufficient Gln to meet demands, and in theseconditions Gln may become a conditionally essential aminoacid [10].

1.8. Glutamine in Breast Milk. In addition to providing anideal nutritional composition for the neonate, breast milkcontains specific nutrients such as Gln that may influencegastrointestinal development and can modulate immune,metabolic, and inflammatory responses [89]. Deprivationof dietary or endogenously synthesized Gln results in abreakdown in the intestinal epithelium of artificially rearedneonatal rats, whereas Gln supplementation may help tomaintain intestinal integrity [59].

In extremely low birth weight (ELBW) infants, thebeneficial effects of breast milk ingested in the neonatalintensive care unit (NICU) on developmental outcomes at18 months of age [90] persist at 30 months [91]. WhilePN is Gln-free, Agostoni et al. [92] reported lower freeGln concentrations in standard infant formulas comparedto breast milk collected from 40 healthy lactating mothersafter delivery of term infants at age 1 month. And similarto previous reports [93], they observed that glutamic acidand Gln accounted for most of the free amino acids inbreast milk [92]. Moreover, heat sterilization of infantformulas can further lower the concentrations of Gln bymore than 60% [94]. Thus, suggesting that enrichmentof infant formulas with nonprotein nitrogen components(particularly Gln and glutamic acid) could be beneficial. Thesame group followed 16 healthy exclusively breastfeedingmothers after delivery of term infants and showed that theconcentrations of free glutamic acid and Gln increased by2.5 and 20 fold, respectively, with progressing lactation,representing >50% of total free amino acids by 3 months[95]. To assess the potential influence of gestational ageand duration of lactation, Jochum et al. [96] measuredthe content of free and protein-bound Gln in transitionaland mature breast milk of 40 healthy mothers after termand preterm delivery. While the time of delivery had noinfluence on the free Gln or total Gln concentration, free Glnconcentrations increased during maturation of lactogenesis,similar to previous reports [95, 97, 98]. In contrast, total Glnand protein-bound Gln concentrations decreased with theduration of lactation, and this correlated with the decrease intotal protein concentration in mature breast milk [96]. Thesedata highlight the need to better define the role of Gln inthe neonatal period and the possible benefit of supplementalGln.

2. Materials and Methods

The following sections describe the clinical studies thatexamined the effects of parenteral and enteral Gln supple-mentation in premature neonates as well as older infantsand children with various diseases. The search methods foridentification of studies consisted of searches of PubMed


(1966–June 2011). The database was searched using thesearch term: “glutamine.” The search output was limitedwith the search filter for ages: all children 0–18 years. Therewere no language restrictions. References in previous reviewsand studies were examined also. The title and abstract of allstudies identified by the above search strategy were screened,and the full text for all potentially relevant studies publishedin English was obtained. The full text of any potentiallyrelevant studies was assessed by the first author. Studies thatincluded only adult participants were excluded. The sameauthor extracted data from the published studies.

3. Results and Discussion

3.1. Parenteral Glutamine Supplementation in

Premature Neonates

3.1.1. Clinical Outcomes. The first evidence to suggest thatparenteral Gln appears safe and may be considered a con-ditionally essential amino acid in premature infants was putforth more than a decade ago by Lacey et al. [99] (Table 1).Although efficacy was not demonstrated in the entire cohortof very low birth weight (VLBW) infants (N = 44), subgroupanalysis in the infants with birth weight


Table 1: Glutamine-supplemented parenteral nutrition in premature neonates.

Reference Subjects Design Gln Control Outcomes Results

Lacey et al.1996 [99]

44 VLBWprematureneonates age


Table 1: Continued.

Reference Subjects Design Gln Control Outcomes Results

Poindexteret al. 2003∗

[103]

141 ELBW (≤72 hafter birth); birthwt: 401–1000 g;GA: 26± 2 wk

Multicentre-randomizeddouble blind

Isonitrogenousstudy AA solutionwith 20% of AAas Gln in PN for∼10 d, AA intake≤3–3.5 g/kg/d(n = 72)

Standard AAsolution withoutGln(TrophAmine) inPN for ∼10 d, AAintake≤3–3.5 g/kg/d(n = 69)

Safety as assessedby plasmaconcentrations ofAA and ammoniaafter infants hadreceived study PN(2.3± 1.0 g/kg/dAA) for ∼10 d

Increased plasma Gln(Gln group only) withno apparentbiochemical risk,increased plasmaessential AA (bothgroups) whereas Pheand Tyr decreased withgreater decrease in Tyr(Gln group), no changein plasma ammonia

Kalhan et al.2005 [80]

20 LBW clinicallystable (24–48 hafter birth); birthwt: 694–1590 g;GA ≤32 wk

Randomizeddouble blind

IsonitrogenousAA mixturesupplementedwith Gln0.6 g/kg/d in PNfor 3–5 d, AAintake∼3.0 g/kg/d(n = 10)

AA mixturewithout Gln inPN for 3–5 d, AAintake∼3.0 g/kg/d(n = 10)

Whole bodyprotein and Glnkinetics (IVinfusions of[2H5]Phe, L-[1–13C,15N]Leu,[15N2]urea,L-[5–15N]Gln)on d6 or d7 of lifewhile receivingAA mixturecontinuouslyafter 3–5 d

Lower endogenousrates of appearance ofPhe and Leu N (indicesof proteolysis), lowerGln de novo synthesis,no differences in rate ofappearance of Leu C,urea turnover orplasma AAconcentrations

Li et al.2007 [101]

53 LBWpremature infants(48–72 h afterbirth); birth wt:2 wk, AA intakestarted at0.5–1.0 g/kg/d to≤3.0 g/kg/d(n = 25)

Growth,biochemicalindices, feedingtolerance, andinfective episodesthroughouthospitalization

Fewer d to regain birthwt, no differences in wtor headcircumferences, fewer don PN, fewer episodesof hospital-acquiredinfection, shorter LOSin hospital, safe

Wang et al.2010 [104]

30 VLBW;median age(interquartilerange) (d):2.5(1.0-2.0) (Gln)and 2.2(1.0-2.0)(control); birthwt: 70% ofrecommendedintake from EN(n = 13)

6% pediatric AAcompoundinjection, averageAA dosage2.0 g/kg/d, PNdecreased whenEN increased andPN withheldwhen >70% ofrecommendedintake from EN(n = 15)

(1) Mortality andchanges inhepatic function(bile acid, ALT,AST, totalbilirubin, directbilirubin,prealbumin,albumin), (2)time to full EN(d), episodes ofgastric residual,total duration ofPN (d), wt gain(g/d), headcircumference(cm), LOS, d onventilator

(1) Decreased AST anddirect bilirubin, nodifferences in bile acid,ALT, total bilirubin,prealbumin, oralbumin, (2) nodifferences in time tofull EN, episodes ofgastric residual, totalduration of PN, wtgain, headcircumference, LOS ord on ventilator

VLBW: very low birth weight; PN: parenteral nutrition; wt: weight; GA: gestational age; AA: amino acid; EN: enteral nutrition; LOS: length of stay; NICU:neonatal intensive care unit; LBW: low birth weight; IV: intravenous; ELBW: extremely low birth weight; NEC: necrotizing enterocolitis; AST: aspartateaminotransferase; ALT: alanine aminotransferase.∗Originating from the same cohort.


(such as mortality) or multiple endpoints. Furthermore,no deaths occurred and only 28/30 infants randomizedcompleted the study and thus analysis was not by intentionto treat. Interestingly, hepatic function improved as assessedby serum aspartate aminotransferase and direct bilirubin,which both decreased after PN in the Gln-supplementedgroup (P < 0.05). It should be noted, however, that nodifferences were observed for other measures of hepaticfunction (bile acid, alanine aminotransferase, total bilirubin,prealbumin, or albumin) or other secondary outcomes (timeto full EN, episodes of gastric residuals, total duration of PN,weight gain, head circumference, length of stay, or days onventilator).

3.1.2. Protein Metabolism. Two small randomized con-trolled trials in LBW infants examined the effects of Gln-supplemented PN on whole body protein metabolism usingprimed continuous IV infusions of essential amino acidtracers [80, 105] (Table 1). Des Robert et al. [105] studied13 LBW neonates on postnatal d-3, while they receivedexclusive PN that was supplemented with Gln (0.5 g/kg/d)or an isonitrogenous Gln-free amino acid solution for 24 h.Compared to an isonitrogenous amino acid supplement,Gln decreased the rate of plasma Leu appearance, Leurelease from protein breakdown (an index of whole bodyproteolysis; −16%, P < 0.05), and rate of Leu oxidation(−35%, P < 0.05). There was also, however, a decrease innonoxidative Leu disposal (an index of whole body proteinsynthesis; −20%, P < 0.05), and, thus, net Leu balance(protein balance) did not differ between groups. Plasma Glnconcentrations were higher in Gln versus control, whereasplasma ammonia did not differ. Although parenteral Glnfailed to enhance estimates of protein synthesis, Gln maypreserve body protein as it suppressed Leu oxidation andprotein breakdown in LBW infants. In addition to the smallsample size, the failure to enhance protein synthesis mayhave also been due to insufficient amino acid availabilitysince whole body protein kinetics were assessed on d-4 of life(when amino acid intake was 2 g/kg/d in both groups) beforeinfants received an optimal amino acid intake of 3 g/kg/d[106].

Kalhan et al. [80] examined the effect of 0.6 g/kg/dGln-supplemented PN for 3–5 d on whole body proteinand Gln kinetics in a carefully selected population of 20clinically stable LBW infants, between d-1 and -2 after birth.Compared to an isonitrogenous control, Gln-supplementedPN resulted in significantly lower rates of appearance ofphenylalanine (Phe) and Leu nitrogen and a nonsignificantdecrease in the rate of appearance of Leu carbon. Gln alsosuppressed the endogenous rate of Gln synthesis. There wasno significant difference in urea turnover between the 2groups. The results suggest that parenteral Gln supplemen-tation at 0.6 g/kg/d decreases whole body protein breakdownand Gln de novo synthesis in clinically stable LBW infants andmay be beneficial in selected populations of LBW infants.The carefully selected population of clinically stable infantslimits the application of the results to other groups ofpremature neonates. Moreover, the use of a higher dose ofGln makes comparisons with other studies difficult.

Interestingly, the same group demonstrated that thesuppression of proteolysis and protein oxidation in responseto an acute increase in parenteral amino acids (withoutGln) was not evident when the amino acid infusion wascontinued for a prolonged period in both acutely ill [81] andclinically stable LBW infants [82]. The only exception waswhen amino acids were supplemented with Gln, wherebya prolonged infusion resulted in a sustained inhibition ofwhole body proteolysis and reduced Gln de novo synthesis.Taken together with previous studies [80, 105], Gln supple-mentation may have a protein-sparing effect in prematureinfants decreasing whole body protein breakdown and Glnde novo synthesis thereby “sparing” the increased amino acidsfor protein synthesis.

3.2. Enteral Glutamine Supplementation in

Premature Neonates

3.2.1. Clinical Outcomes. Neu et al. [107] conducted adouble-blind randomized trial to test whether enteral Glnsupplementation for VLBW infants decreases morbidity(Table 2). Sixty-eight premature neonates were assigned toa Gln-supplemented premature formula or a nonsupple-mented standard premature formula between postnatal d-3 to d-30. The Gln supplemented group initially received adose of 0.08 g/kg/d Gln which was increased to a maximumof 0.31 g/kg/d Gln by d-13. The Gln group had bettertolerance to enteral feedings (fewer % of d with no oral intakein Gln: 8.8% versus controls: 23.8%, P < 0.01). Episodesof hospital-acquired sepsis were 4/35 and 10/33 in Gln andcontrol group, respectively. Moreover, when controlling forbirth weight, the estimated odds of developing sepsis was3.8 times higher for control versus Gln (95% CI: 1.01–14.18). Analysis of T cell subsets showed a blunting of therise in HLA-DR+ and CD16/CD56 in the Gln group. Therewere, however, no significant differences between groupsfor cases of NEC, growth, or length of stay. Whereas theplasma concentrations of alanine (Ala), glycine (Gly), Ser,threonine (Thr), Phe, and total nonessential amino acidswere lower in the Gln-supplemented infants after 2-weeksupplementation, there were no differences between groupsfor plasma concentrations of Gln, Glu, or ammonia [108].The authors speculated that the lower plasma amino acidconcentrations in infants fed Gln were the result of enhanceduptake of these amino acids for gluconeogenesis and provideevidence of reduced tissue catabolism. A secondary analysisof the initial trial also provided evidence for decreasedhospital costs [109]. While the control used is comparableto routine clinical practice, the study design cannot ensurethe specific effect of Gln as the differences between feedinggroups might result from higher intakes of nitrogen orenergy with Gln supplementation. The authors, however,chose not to use a third group with an isonitrogenous controldue to recruitment constraints. Although the trial was small,these initial results provided evidence for better toleranceto enteral feedings and lower sepsis rates in VLBW infantsreceiving enteral Gln supplementation.

Barbosa et al. [110] conducted a randomized controlledpilot study to evaluate the tolerance and clinical impact of

Journal of Nutrition and Metabolism 9T

abl

e2:

Glu

tam

ine-

supp

lem

ente

den

tera

lnu

trit

ion

inpr

emat

ure

neo

nat

es.

Ref

eren

ceSu

bjec

tsD

esig

nG

lnC

ontr

olO

utc

omes

Res

ult

s

Neu

etal

.19

97∗

[107

]

68V

LBW

age

<3

dre

ceiv

ing

PN

;bir

thw

t:50

0–12

50g;

GA

:24

–32

wk

Ran

dom

ized

dou

ble-

blin

d

From

d3to

d30

oflif

ere

ceiv

edpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)su

pple

men

ted

wit

hG

lnat

0.08

g/kg

/don

d3an

dre

ach

ed0.

31g/

kg/d

byd1

3,P

NA

Ast

arte

dat

0.5

g/kg

/don

d3u

nti

l3

g/kg

/dby

d8(n=

35)

From

d3to

d30

oflif

ere

ceiv

edpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)al

one,

PN

AA

star

ted

at0.

5g/

kg/d

ond3

un

til3

g/kg

/dby

d8(n=

33)

(1)

Hos

pita

l-ac

quir

edse

psis

,to

lera

nce

toen

tera

lfee

din

gs,

and

LOS

inh

ospi

tal;

(2)

anal

ysis

ofT

-cel

lsu

bset

sin

per

iph

eral

bloo

ddu

rin

gw

k-4

oflif

e,N

EC

,gro

wth

,re

spir

ator

yst

atu

s,an

dsa

fety

(1)

Red

uce

dh

ospi

tal-

acqu

ired

seps

is(p

osit

ive

bloo

dcu

ltu

re)

wh

enco

ntr

olle

dfo

rbi

rth

wt,

few

er%

dw

ith

no

oral

inta

ke,a

nd

no

diff

eren

ces

inLO

Sin

hos

pita

l;(2

)bl

un

ted

the

rise

inH

LA-D

R+

and

CD

16/C

D56

subs

ets,

no

diff

eren

ces

inN

EC

,wt,

len

gth

,hea

dci

rcu

mfe

ren

ce,

mec

han

ical

ven

tila

tion

orde

ath

,saf

e

Roi

get

al.

1996

∗[1

08]

68V

LBW

age

<3

dre

ceiv

ing

PN

;bir

thw

t:50

0–12

50g;

GA

:25

–32

wk

Ran

dom

ized

dou

ble-

blin

d

From

d3to

d30

oflif

ere

ceiv

edpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)su

pple

men

ted

wit

hG

lnat

0.08

g/kg

/don

d3an

dre

ach

ed0.

31g/

kg/d

byd1

3,P

NA

Ast

arte

dat

0.5

g/kg

/don

d3u

nti

l3

g/kg

/dby

d8(n=

35)

From

d3to

d30

oflif

ere

ceiv

edpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)al

one,

PN

AA

star

ted

at0.

5g/

kg/d

ond3

un

til3

g/kg

/dby

d8(n=

33)

Pla

sma

AA

con

cen

trat

ion

sat

3ti

me

poin

ts(p

refe

edin

g,tr

ansi

tion

tofu

llfe

eds

and

full

EN

)

Dec

reas

edpl

asm

aco

nce

ntr

atio

ns

ofA

la,G

ly,

Ser,

Th

r,P

he,

and

tota

ln

ones

sen

tial

AA

afte

r∼2

wk

supp

lem

enta

tion

(i.e

.,at

tran

siti

onto

full

EN

),n

odi

ffer

ence

sbe

twee

ngr

oups

for

plas

ma

con

cen

trat

ion

sof

Gln

,Glu

,or

amm

onia

duri

ng

stu

dy

Dar

mau

net

al.1

997∗

[125

]

11V

LBW

age

<3

dre

ceiv

ing

PN

;bir

thw

t:<

1250

g;G

A:

24–3

2w

k

Ran

dom

ized

dou

ble-

blin

d

Rec

eive

dpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)su

pple

men

ted

wit

hG

lnat

0.08

g/kg

/don

d3an

din

crea

sed

to0.

2g/

kg/d

ond1

0,P

NA

Ast

arte

dat

0.5

g/kg

/don

d3u

nti

l3

g/kg

/dby

d8(n=

6)

Rec

eive

dpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)al

one

ond3

oflif

eu

nti

ld10

,PN

AA

star

ted

at0.

5g/

kg/d

ond3

un

til3

g/kg

/dby

d8(n=

5)

Wh

ole

body

prot

ein

met

abol

ism

/Leu

kin

etic

san

dG

lnki

net

ics

(IV

infu

sion

sof

L-[2

H]L

euan

dL-

[13C

]Gln

)on

d10

oflif

ein

fed

stat

e,th

atis

,du

rin

gco

nti

nu

ous

EN

(Gln

supp

lem

ente

dor

non

supp

lem

ente

d)an

dP

N

No

diff

eren

ces

inov

eral

lrat

eof

app

eara

nce

ofLe

uan

dG

ln,n

odi

ffer

ence

sin

Leu

and

Gln

rele

ase

from

prot

ein

brea

kdow

n,n

odi

ffer

ence

sin

Gln

deno

vosy

nth

esis

orpl

asm

aG

lnco

nce

ntr

atio

ns

Dal

las

etal

.19

98∗

[109

]

68V

LBW

age

<3

dre

ceiv

ing

PN

;bir

thw

t:50

0–12

50g;

GA

:24

–32

wk

Ran

dom

ized

dou

ble-

blin

d

From

d3to

d30

oflif

ere

ceiv

edpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)su

pple

men

ted

wit

hG

lnat

0.08

g/kg

/don

d3an

dre

ach

ed0.

31g/

kg/d

byd1

3,P

NA

Ast

arte

dat

0.5

g/kg

/don

d3u

nti

l3

g/kg

/dby

d8(n=

35)

From

d3to

d30

oflif

ere

ceiv

edpr

emat

ure

form

ula

(Sim

ilac

Spec

ial

Car

eG

rou

p)al

one,

PN

AA

star

ted

at0.

5g/

kg/d

ond3

un

til3

g/kg

/dby

d8(n=

33)

Hos

pita

lcos

tsan

alys

edby

log-

ran

kte

sts

and

Kap

lan

-Mei

erpl

ots

Red

uce

dm

edia

nco

sts

for

hos

pita

lisat

ion

,rad

iolo

gy,

phar

mac

y,la

bora

tory

,an

dN

ICU

,red

uce

dm

edia

nn

um

ber

ofu

tiliz

atio

nu

nit

s


Ta

ble

2:C

onti

nu

ed.

Ref

eren

ceSu

bjec

tsD

esig

nG

lnC

ontr

olO

utc

omes

Res

ult

s

Bar

bosa

etal

.19

99[1

10]

9cr

itic

ally

illin

fan

tsag

ed1–

24m

ore

quir

ing

inte

nsi

veca

rean

dto

lera

tin

gE

N

Ran

dom

ized

dou

ble-

blin

d

Gln

supp

lem

ente

d0.

3g/

kg/d

EN

for

5d

(n=

5)

Cas

ein

supp

lem

ente

d0.

3g/

kg/d

EN

for

5d

(n=

4)

Sept

icm

orbi

dity

,mor

talit

y,LO

Sin

ICU

,LO

Sin

hos

pita

lan

ddu

rati

onof

mec

han

ical

ven

tila

tion

,an

dto

lera

nce

No

sign

ifica

nt

diff

eren

ces

inse

ptic

com

plic

atio

ns,

mor

talit

y,d

inIC

U,d

inh

ospi

talo

rd

wit

hve

nti

lati

on,w

ellt

oler

ated

,an

dsa

fe

Mer

cier

etal

.20

03[1

12]

25LB

Wpr

emat

ure

neo

nat

esag

e≥1

4d

wit

hou

tac

ute

illn

ess

and

rece

ivin

gex

clu

sive

EN

;bi

rth

wt:

980–

1890

g;G

A:

27–3

5w

k

Ran

dom

ized

dou

ble-

blin

d

Exc

lusi

veen

tera

lpre

term

infa

nt

form

ula

supp

lem

ente

dw

ith

Gln

0.7

g/kg

/d(1

7%of

tota

lpr

otei

nin

take

asG

ln)

for

21d

(n=

12)

Exc

lusi

veen

tera

lpre

term

infa

nt

form

ula

supp

lem

ente

dw

ith

ison

itro

gen

ous

lact

oser

um

extr

acte

dpr

otei

n0.

7g/

kg/d

(17%

ofto

talp

rote

inin

take

asla

ctos

eru

m)

for

21d

(n=

9)

Para

met

ers

ofsu

per

ior

mes

ente

ric

bloo

dfl

owve

loci

ties

mea

sure

din

fed

stat

eby

puls

edD

oppl

eru

ltra

sou

nd

afte

r21

dsu

pple

men

tati

on

No

effec

ton

supe

rior

mes

ente

ric

bloo

dfl

ow

Vau

ghn

etal

.20

03[1

11]

649

VLB

Wag

e<

7d

rece

ivin

gP

N;b

irth

wt:

500–

1250

g

Mu

ltic

entr

era

ndo

miz

eddo

ubl

e-bl

ind

En

tera

lGln

supp

lem

ent

(0.3

g/kg

/d,3

%G

lnin

ster

ilew

ater

)gi

ven

atth

esa

me

tim

ebu

tse

para

tefr

omfe

edin

gsfo

rth

efi

rst

28d

oflif

e;A

Ain

take

∼3–3

.5g/

kg/d

byd7

(n=

314)

En

tera

lpla

cebo

(ste

rile

wat

er)

give

nat

the

sam

eti

me

but

sepa

rate

from

feed

ings

for

the

firs

t28

dof

life;

AA

inta

ke∼3

–3.5

g/kg

/dby

d7(n=

335)

(1)

Nu

mbe

rof

epis

odes

ofbl

ood

cult

ure

-pro

ven

nos

ocom

ials

epsi

sbe

twee

n7

dto

36w

kof

age;

(2)

susp

ecte

dse

psis

,pn

eum

onia

,UT

I,m

enin

giti

s,N

EC

,IV

H,P

VL

,re

tin

opat

hyof

prem

atu

rity

,u

seof

oxyg

enat

36w

k,ga

stro

inte

stin

aldy

sfu

nct

ion

,gr

owth

,age

and

wei

ght

atdi

sch

arge

,an

dde

ath

(1)

No

diff

eren

ces

inn

osoc

omia

lsep

sis;

(2)

decr

ease

dga

stro

inte

stin

aldy

sfu

nct

ion

and

seve

ren

euro

logi

cse

quel

aeam

ong

surv

ivor

s(G

rade

s3

and

4IV

Han

dP

VL

),n

odi

ffer

ence

sin

susp

ecte

dse

psis

,pn

eum

onia

,UT

I,m

enin

giti

s,N

EC

,re

tin

opat

hyof

prem

atu

rity

,u

seof

oxyg

enat

36w

k,gr

owth

,age

and

wt

atdi

sch

arge

,or

deat

h

Pari

mie

tal

.20

04[1

27]

26L

BW

ingo

odh

ealt

han

dga

inin

gw

tag

ed10

–74

d(>

23d

ofag

e);b

irth

wt:

693–

1846

g;G

A<

32w

k

Ran

dom

ized

dou

ble-

blin

dE

nte

ralG

ln0.

6g/

kg/d

for

5d

(n=

9)

(1)

En

tera

lGly

0.6

g/kg

/dfo

r5

d(n=

9)or

(2)

un

supp

lem

ente

den

tera

lfo

rmu

lafo

r5

d(n=

8)

Wh

ole-

body

prot

ein

,Gln

and

ure

aki

net

ics

onst

udy

d-6

(IV

infu

sion

sof

L-[1

–13C

,15N

]Leu

,[2

H5]

Ph

e,[1

5N2]

ure

a,L-

[5–1

5N]G

ln)

afte

r5

dsu

pple

men

tati

ondu

rin

gfa

stin

g(3

haf

ter

the

last

mea

l)an

dfe

dst

ate,

and

plas

ma

AA

con

cen

trat

ion

sdu

rin

gfa

stin

g

Hig

her

rate

ofu

rea

syn

thes

is,

no

diff

eren

ces

inra

tes

ofap

pea

ran

ceof

Ph

e,Le

uC

orLe

uN

,no

diff

eren

ces

inra

teof

app

eara

nce

ofG

lnor

plas

ma

Gln

con

cen

trat

ion

s


Ta

ble

2:C

onti

nu

ed.

Ref

eren

ceSu

bjec

tsD

esig

nG

lnC

ontr

olO

utc

omes

Res

ult

s

Kor

kmaz

etal

.200

7[1

23]

69V

LBW

prem

atu

ren

eon

ates

rece

ivin

gP

N;

birt

hw

t:<

1500

g

Pro

spec

tive

inte

rven

tion

From

8d

to12

0d

oflif

ere

ceiv

edG

lnsu

pple

men

t0.

15g/

kg/d

BID

mix

edw

ith

ster

ilew

ater

and

give

nat

the

sam

eti

me

but

sepa

rate

from

feed

ings

byan

orog

astr

ictu

beor

oral

ly(n=

36)

From

8d

to12

0d

oflif

ere

ceiv

edst

erile

wat

er(p

lace

bo)

give

nat

the

sam

eti

me

but

sepa

rate

from

feed

ings

byan

orog

astr

ictu

beor

oral

ly(n=

33)

Gro

wth

para

met

ers

atbi

rth

,1,

2,3,

and

4m

oan

dbi

och

emic

alpa

ram

eter

sat

4m

o

No

diff

eren

ces

ingr

owth

para

met

ers

atbi

rth

and

2m

o,h

igh

erw

t,le

ngt

h,h

ead

circ

um

fere

nce

,MU

AC

and

mid

-th

igh

circ

um

fere

nce

at3

and

4m

o,n

oto

xici

tysi

gns

Van

Den

Ber

get

al.2

005∗

∗

[113

]

102

VLB

W<

48h

afte

rbi

rth

rece

ivin

gP

N;

birt

hw

t:<

1500

g;G

A<

32w

k

Ran

dom

ized

dou

ble-

blin

d

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

52)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

50)

(1)

Feed

ing

tole

ran

ce;(

2)gr

owth

,in

fect

iou

sm

orbi

dity

,an

dsh

ort-

term

outc

ome

(1)

No

diff

eren

ces

inti

me

tofu

llen

tera

lfee

din

g;(2

)n

odi

ffer

ence

sin

age

PN

disc

onti

nu

ed,d

ofn

oen

tera

lfe

edin

g,N

EC

orw

tz-

scor

es,

decr

ease

din

cide

nce

of≥1

seri

ous

infe

ctio

ns

(sep

sis,

men

ingi

tis,

pyel

onep

hri

tis,

pneu

mon

ia,a

nd

arth

riti

s),

no

diff

eren

ces

inP

DA

trea

ted

wit

hin

dom

eth

acin

orsu

rgic

allig

atio

n,

mec

han

ical

ven

tila

tion

,su

pple

men

talo

xyge

n,

reti

nop

athy

,age

atdi

sch

arge

from

NIC

U,a

geat

disc

har

gefr

omh

ospi

talo

rde

ath

Van

Den

Ber

get

al.2

005∗

∗

[114

]

102

VLB

W<

48h

afte

rbi

rth

rece

ivin

gP

N;

birt

hw

t:<

1500

g;G

A<

32w

k

Ran

dom

ized

dou

ble-

blin

d

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

52)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

50)

Safe

tyby

dete

rmin

atio

nof

plas

ma

AA

con

cen

trat

ion

sm

easu

red

at4

tim

ep

oin

ts:

<48

h,d

7,d1

4,an

dd3

0of

life

Incr

ease

dpl

asm

aA

Aco

nce

ntr

atio

ns

ofm

ost

esse

nti

alan

dn

ones

sen

tial

AA

incl

udi

ng

Gln

and

Glu

(bot

hgr

oups

),de

crea

sed

Ph

ean

dTy

r(b

oth

grou

ps)

Van

Den

Ber

get

al.2

006∗

∗

[116

]

90V

LBW

<48

haf

ter

birt

hre

ceiv

ing

PN

;bi

rth

wt:<

1500

g;G

A<

32w

k

Ran

dom

ized

dou

ble-

blin

d

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

45)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

45)

Inte

stin

alp

erm

eabi

lity

(uri

nar

yla

ctu

lose

/man

nit

olra

tio)

asse

ssed

at4

tim

epo

ints

:<48

h,d

7,d1

4,an

dd3

0of

life

Dec

reas

edla

ctu

lose

/man

nit

olra

tio

(bot

hgr

oups

),de

crea

sed

uri

nar

yla

ctu

lose

con

cen

trat

ion

s(b

oth

grou

ps),

incr

ease

du

rin

ary

man

nit

olco

nce

ntr

atio

ns

(bot

hgr

oups

)


Ta

ble

2:C

onti

nu

ed.

Ref

eren

ceSu

bjec

tsD

esig

nG

lnC

ontr

olO

utc

omes

Res

ult

s

Van

Den

Ber

get

al.2

007∗

∗

[115

]

86V

LBW

<48

haf

ter

birt

hre

ceiv

ing

PN

;bi

rth

wt:<

1500

g;G

A<

32w

k

Ran

dom

ized

dou

ble-

blin

d

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

43)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

43)

Inte

stin

al(f

aeca

l)m

icro

flor

a(b

ifido

bact

eria

,lac

toba

cilli

,E

sch

eria

coli,

stre

ptoc

occi

,cl

ostr

idia

)at<

48h

,at

d7,

d14,

and

d30

oflif

e,as

anal

ysed

byfl

uor

esce

nt

insi

tuhy

brid

izat

ion

No

diff

eren

cein

prev

alen

ceof

inte

stin

alm

icro

flor

a

Van

Den

Ber

get

al.2

007∗

∗

[117

]

77V

LBW

surv

ivin

gin

fan

tsat

the

corr

ecte

dag

eof

1y;

birt

hw

t:<

1500

g;G

A<

32w

k

Follo

wu

pof

alls

urv

ivin

gpa

rtic

ipan

tsof

ara

ndo

miz

eddo

ubl

e-bl

ind

stu

dy

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

37)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

40)

Inci

den

ceof

alle

rgic

and

infe

ctio

us

dise

ase

duri

ng

the

firs

tye

arof

life

asas

sess

edby

valid

ated

ques

tion

nai

res

Low

erri

skof

atop

icde

rmat

itis

,no

diff

eren

ces

inin

cide

nce

ofbr

onch

ial

hype

ract

ivit

y,in

fect

ion

sof

upp

erre

spir

ator

y,lo

wer

resp

irat

ory,

orga

stro

inte

stin

altr

acts

Van

Zw

olet

al.2

008∗

∗

[122

]

72V

LBW

infa

nts

atth

eco

rrec

ted

age

of2

y;bi

rth

wt:<

1500

g;G

A<

32w

k

Follo

wu

pof

part

icip

ants

ofa

ran

dom

ized

dou

ble-

blin

dst

udy

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

40)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

32)

Neu

rode

velo

pmen

tal

outc

ome

(neu

rode

velo

pmen

tal

impa

irm

ent,

cere

bral

pals

y,vi

sion

,hea

rin

g,M

DI,

and

PD

Iof

the

Bay

ley

Scal

eof

Infa

nt

Dev

elop

men

tII

)ev

alu

ated

atth

eco

rrec

ted

age

of2

y

No

diff

eren

ces

inth

ein

cide

nce

ofan

MD

In

ora

PD

I≤

85,n

odi

ffer

ence

sin

the

inci

den

ceof

neu

rode

velo

pmen

tal

impa

irm

ent

orce

rebr

alpa

lsy

Van

Den

Ber

get

al.2

009∗

∗

[120

]

63V

LBW

infa

nts

<48

haf

ter

birt

hre

ceiv

ing

PN

;bi

rth

wt:<

1500

g;G

A<

32w

k

Ran

dom

ized

dou

ble-

blin

d

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

25)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

38)

Th

1(I

F-ga

mm

a,T

NF-

alph

a,IL

-2)

and

Th

2cy

toki

ne

resp

onse

s(I

L-10

,IL-

5,IL

-4)

at48

h,d

7,an

dd1

4of

life

follo

win

gin

vitr

ow

hol

ebl

ood

cell

stim

ula

tion

asan

alyz

edby

cyto

met

ric

bead

arra

y

No

diff

eren

ces

inT

h1

orT

h2

cyto

kin

ere

spon

ses

Van

Zw

olet

al.2

009∗

∗

[119

]

59V

LBW

infa

nts

atag

eof

1y;

birt

hw

t:<

1500

g;G

A<

32w

k

Follo

wu

pof

part

icip

ants

ofa

ran

dom

ized

dou

ble-

blin

dst

udy

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

28)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

31)

Th

1(I

F-ga

mm

a,T

NF-

alph

a,IL

-2)

and

Th

2cy

toki

ne

profi

les

(IL-

10,I

L-5,

IL-4

)at

1y

ofag

efo

llow

ing

invi

tro

wh

ole

bloo

dst

imu

lati

on

No

diff

eren

ces

inT

h1

orT

h2

cyto

kin

epr

ofile

s


Ta

ble

2:C

onti

nu

ed.

Ref

eren

ceSu

bjec

tsD

esig

nG

lnC

ontr

olO

utc

omes

Res

ult

s

Van

Zw

olet

al.2

010∗

∗

[121

]

72V

LBW

infa

nts

atth

eco

rrec

ted

age

of1

y;bi

rth

wt:<

1500

g;G

A<

32w

k

Follo

wu

pof

part

icip

ants

ofa

ran

dom

ized

dou

ble-

blin

dst

udy

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

34)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

38)

Inte

stin

al(f

aeca

l)m

icro

biot

a(b

ifido

bact

eria

,Ch

is/l

itgr

oup,

Esc

her

iaco

li)at

age

1y

asan

alys

edby

flu

ores

cen

tin

situ

hybr

idiz

atio

n

No

diff

eren

cein

prev

alen

ceof

inte

stin

alm

icro

biot

a

Van

Zw

olet

al.2

011∗

∗

[118

]

76V

LBW

infa

nts

at6

yof

age;

birt

hw

t:<

1500

g;G

A<

32w

k

Follo

wu

pof

part

icip

ants

ofa

ran

dom

ized

dou

ble-

blin

dst

udy

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

hG

lnin

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

38)

From

3d

to30

dof

life

rece

ived

ente

ralp

rete

rmfo

rmu

laor

brea

stm

ilksu

pple

men

ted

wit

his

onit

roge

nou

sA

lain

incr

easi

ng

dose

sto

≤0.3

g/kg

/d(n=

38)

Alle

rgic

(ato

pic

derm

atit

is,

hay

feve

r,re

curr

ent

wh

eeze

,an

das

thm

a)an

din

fect

iou

sdi

seas

e(u

pper

and

low

erre

spir

ator

ytr

act

infe

ctio

n,

gast

roin

test

inal

trac

tin

fect

ion

,uri

nar

ytr

act

infe

ctio

n)

at6

yof

age

asas

sess

edby

valid

ated

ques

tion

nai

res

Dec

reas

edri

skof

atop

icde

rmat

itis

,no

diff

eren

ces

inh

ayfe

ver,

recu

rren

tw

hee

zeor

asth

ma,

decr

ease

dri

skof

gast

roin

test

inal

trac

tin

fect

ion

,no

diff

eren

ces

inu

pper

orlo

wer

resp

irat

ory

trac

tin

fect

ion

oru

rin

ary

trac

tin

fect

ion

VLB

W:v

ery

low

birt

hw

eigh

t;P

N:p

aren

tera

lnu

trit

ion

;wt:

wei

ght;

GA

:ges

tati

onal

age;

AA

:am

ino

acid

;EN

:en

tera

lnu

trit

ion

;IV

:in

trav

enou

s;LO

S:le

ngt

hof

stay

;NE

C:n

ecro

tizi

ng

ente

roco

litis

;IC

U:i

nte

nsi

veca

reu

nit

;LB

W:l

owbi

rth

wei

ght;

UT

I:u

rin

ary

trac

tin

fect

ion

;IV

H:i

ntr

aven

tric

ula

rh

emor

rhag

e;P

VL:

peri

ven

tric

ula

rle

uko

mal

acia

;PD

A:p

aten

tdu

ctu

sar

teri

osu

s;N

ICU

:neo

nat

alin

ten

sive

care

un

it;B

ID:t

wic

ea

day;

MU

AC

:mid

upp

erar

mci

rcu

mfe

ren

ce;M

DI:

Men

talD

evel

opm

enta

lIn

dex;

PD

I:Ps

ych

omot

orD

evel

opm

enta

lIn

dex;

Th

1:T

hel

per

typ

e1;

Th

2:T

hel

per

typ

e2;

IF:i

nte

rfer

on;T

NF:

tum

orn

ecro

sis

fact

or;

IL:i

nte

rleu

kin

.∗ ,∗∗

Ori

gin

atin

gfr

omth

esa

me

coh

ort.


enteral formula supplemented with 0.3 g/kg/d Gln for 5 daysversus an equal dose of casein in 9 critically ill infants aged 1–24 months. Although Gln was well tolerated, the study wasunderpowered to detect differences in septic complications(control: 3/4 versus Gln: 1/5, P = 0.10), mortality (control:2/4 versus Gln: 0/5, P = 0.10) or other outcomes (ventilatoruse, length of stay in intensive care unit (ICU) or in hospital).It was also not reported in the inclusion criteria whether thestudy population of infants were premature or term.

Vaughn et al. [111] conducted a large multicentre trial totest whether enteral Gln supplement decreases the incidenceof hospital-acquired infection and other morbidities in 649VLBW infants. Within the first 7 d of age, infants wererandomly assigned to enteral Gln supplement (0.3 g/kg/d,3% Gln in sterile water) or placebo (sterile water) given atthe same time but separate from feedings for the first 28 d.There were no differences between groups for the primaryoutcome (nosocomial sepsis between 7 d and 36 weeks post-menstrual age; Gln: 30.9% versus control: 33.7%). However,gastrointestinal dysfunction (2.5 versus 7.5%, P < 0.01)and severe neurological sequelae among survivors (Grades3 and 4 intraventricular hemorrhage and paraventricularleukomalacia; 10.4 versus 15.1%, P = 0.08) were lessfrequent in Gln versus control, respectively. There were nodifferences in the occurrence of suspected sepsis, pneumonia,urinary tract infection, meningitis, NEC, retinopathy ofprematurity, oxygen use at 36 weeks, or mortality. Growth,age, and weight at discharge were also similar. Whereasenteral Gln does not appear to decrease nosocomial sepsisin VLBW infants, the study may have been underpoweredto detect a significant difference, as dropout rate was higherthan anticipated (i.e., 105 infants exited the study beforecompletion). Also the centre-to-centre variation in this andother multicentre trials [102] may have blunted differences inoutcomes (e.g., sepsis), since nutrition and infection controlpractices may differ among centres and may mask some ofthe differences that might be apparent in a single facility.While the study provides further evidence to suggest thatenteral Gln improves feeding tolerance and may preventcentral nervous system (CNS) morbidity, these positiveresults are based on secondary endpoints and subgroupanalyses. Furthermore, the Gln dose was based on birthweight and was not adjusted for interval changes in weight.Therefore, the dose administered may have been inadequatedue to rapid growth during the early neonatal period.

The apparent improved feeding tolerance in VLBWinfants receiving enteral Gln in previous studies [107, 111]cannot be explained by enhanced mesenteric blood flow[112]. It seems that, in premature infants without acuteillness and tolerating exclusive EN, mesenteric blood flowremains stable after 14 d of age and does not appear to beinfluenced by enteral Gln.

In contrast to previous reports on Gln-enriched EN inVLBW infants [107, 111], Van Den Berg et al. [113] found noimprovement in feeding tolerance, as assessed by the mediand to reach full enteral feeds (Gln: 13 d versus control: 13 d;hazard ratio [95% CI]: 1.19 [0.79–1.79]). In this randomizedcontrolled trial, 102 VLBW infants received either enteral Glnsupplementation or an isonitrogenous control (Ala) added to

breast milk or preterm formula in increasing doses from d-3to d-30 of life to a maximum dose of 0.3 g/kg/d Gln. Therewere also no differences between groups for other variablesof feeding tolerance (age at which PN was discontinued, dof no enteral feeding), NEC, or growth. However, the Gln-supplemented group had a lower incidence of ≥1 seriousinfections (sepsis, meningitis, pyelonephritis, pneumonia,and arthritis) compared with the isonitrogenous controlgroup (Gln: 50% versus control: 76%; OR [95% CI]: 0.32[0.14–0.74]. Other short-term outcomes (patent ductusarteriosus treated with indomethacin or surgical ligation,mechanical ventilation, supplemental oxygen, retinopathy,age at discharge from NICU, age at discharge from hospital,or death) were not significantly different. Gln did not alterplasma concentrations of Gln, Glu, or other amino acids[114]. Although safe at the dose provided, Gln-enrichedEN did not improve feeding tolerance or other short-term outcomes in VLBW infants. However, because Glnreduced infectious morbidity, the use of Gln-enriched EN inVLBW infants deserves further consideration. Comparison,however, with other studies is made difficult because of theuse of different feeding guidelines for the introduction orwithholding enteral feeds. Whereas the choice of isonitroge-nous control prevented the removal of amino acids fromthe Gln supplement in the present study, groups were madeisonitrogenous by adding more amino acid (Ala) to thecontrol group. The control group then received additionalamino acid/nitrogen, which is not representative of dailypractice. However, given the limited sample size, a thirdcomparison group (enteral formula routinely used in anNICU) was not feasible. The beneficial effect of enteral Glnon infection rate could not be explained by an increasednumber of bifidobacteria or lactobacilli in the intestinalmicroflora as demonstrated by a secondary analysis in asubset of 86 VLBW infants [115]. Furthermore, Gln did notenhance the postnatal decrease in intestinal permeability,as assessed by the urinary lactulose/mannitol ratio in asubset of 90 VLBW infants. Specifically, supplementationwith Gln or isonitrogenous control equally decreased urinaryconcentrations of lactulose and increased urinary mannitol[116]. More recently, followup of all surviving participants(n = 77) revealed that Gln-enriched EN in VLBW infantsmay lower the incidence of atopic dermatitis (OR [95%CI]: 0.13 [0.02–0.97]) during the first year of life but hasno effect on the incidence of bronchial hyperactivity orinfectious diseases [117]. Further followup of this cohortof VLBW infants (n = 76) at 6 y of age also found adecreased risk of atopic dermatitis (adjusted OR [95% CI]:0.23 [0.06–0.95]) and gastrointestinal infections (adjustedOR [95% CI]: 0.10 [0.01–0.93]) in the Gln-supplementedgroup [118]. Although outcomes were assessed by validatedquestionnaires in these 2 followup studies, parental reportof physician diagnosis of disease could be subject to report-ing/information bias. Furthermore, the lower incidence ofatopic dermatitis and infection rates with Gln were notrelated to changes in cytokine profiles [119]/responses [120]or to changes in intestinal bacterial species at age 1 y[121]. Finally, the same group studied neurodevelopmentaloutcome at 2 y corrected age in a subgroup of 72 VLBW


infants and found no beneficial effect of Gln-enriched ENduring the neonatal period [122].

The effect of Gln supplementation on long-term out-come of VLBW infants has also been reported by Korkmazet al. [123] who studied the effect of 4-month enteral Glnsupplementation on growth. From d-8 through d-120 of life,69 VLBW infants were assigned to enteral Gln (0.3 g/kg/d)supplement or placebo (sterile water) according to the orderof admission to the NICU. Whereas growth parameters didnot differ during the first 2 months of life, by the end ofthe third and fourth month, infants treated with Gln showedhigher weight, length, head circumference, mid-upper-armcircumference, and midthigh circumference compared tocontrols. The authors concluded that long-term enteral Glnin VLBW infants may lead to improvements in growth in atime-dependent manner without any signs of Gln toxicity.Although this was a prospective interventional study, a com-plete description of masking, blinding, or randomizationprocedures was not provided. Also, because the placebo(sterile water) was neither isocaloric nor isonitrogenous tothe Gln treatment, the enhanced growth could have resultedfrom the effect of increased amino acid/nitrogen, since earlyprovision of parenteral amino acids (without Gln) has beenshown to improve growth parameters in VLBW infants[124].

3.2.2. Protein Metabolism. Darmaun et al. [125] determinedthe effect of enteral Gln on Leu and Gln metabolism ina subset of 11 VLBW neonates from the larger trial [107](Table 2). Enteral Gln supplementation provided at lowdoses (≤0.2 g/kg/d) from d-3 to d-10 of life did not inhibitwhole-body protein breakdown in VLBW infants. Leu releasefrom protein breakdown (an index of whole-body proteinbreakdown) was slightly but not significantly lower in theGln group versus controls. Plasma Gln concentration, Glnrelease from protein breakdown, or Gln de novo synthesis didnot significantly differ between groups. However, there was atrend toward lower rates of Gln de novo synthesis in infantsreceiving Gln supplement. Although the number of patientswas small, the failure to detect a significant effect of Gln on itsown metabolism or on whole-body protein breakdown couldalso be due to the different effects of enteral and parenteralGln supplementation. Importantly, the majority of enteralGln is used in first pass in premature infants [126]. This islikely a significant factor in the different effects of enteral andparenteral Gln. Moreover, the dose of Gln used in the currentstudy was lower than that (0.5 g/kg/d Gln) previously shownto inhibit proteolysis in LBW infants [105].

Parimi et al. [127] examined the effect of enteralGln on whole-body Gln and nitrogen kinetics in healthygrowing LBW infants during the fasting (3 h after the lastmeal) and fed state. This study was the only to have 3groups, where Gln-supplemented group was compared withan isonitrogenous control and enteral formula routinelyused in the NICU. Between 10 and 74 d of age, infantswere randomly assigned to formula supplemented with Gln(0.6 g/kg/d; n = 9), isonitrogenous amounts of Gly (n =9), or unsupplemented formula (n = 8) for 5 d. During

fasting, the rate of appearance of Phe, Leu carbon, andLeu nitrogen (measures of proteolysis) were not significantlydifferent between groups. Compared with controls, enteralGln resulted in an increased rate of urea synthesis, no changein Gln rate of appearance, or plasma Gln concentrations.Similar effects were observed with Gly supplement, but themagnitude was less. The authors concluded that enteral Glndoes not affect Gln rate of appearance or whole-body proteinturnover in a specific group of healthy growing LBW infants,thus, suggesting that Gln is primarily metabolized in the gut(and liver) [126] and is associated with an increased rate ofurea synthesis. Alternatively, because, in healthy prematurenewborns, there is already a high rate of Gln turnover (85%of which is contributed by de novo synthesis) [128], thisspecific population of neonates could have been less sensitiveto enteral Gln. This is in contrast to premature neonates withacute illness, whereby catabolic stress may provoke a greaterneed for exogenous Gln.

In summary, although methodologically sound random-ized trials consistently demonstrate safety in VLBW infants[80, 99, 100, 102, 105, 107, 111, 113, 127], parenteralor enteral Gln supplementation does not appear to affectmortality [100, 102, 104, 107, 111, 113], NEC [100, 102, 107,111, 113], length of stay [99, 100, 102, 104, 107, 111, 113],or growth [100, 102, 104, 107, 111, 113]. Moreover, theresults are conflicting for other short-term clinical outcomessuch as feeding tolerance [99, 100, 102, 104, 107, 111, 113],serious infections/sepsis [100, 102, 107, 111, 113], ventilatoruse [99, 102, 104, 107, 113], and severe neurological sequelae[111]. Few data have been reported for the effects of Glnsupplementation in VLBW infants on long-term clinicaloutcomes such as growth at 4 months [123], allergic andinfectious morbidity at 1 y [117] and 6 y [118], and neu-rodevelopmental outcomes at 2 y corrected age [122]. Largerwell-controlled studies are needed. A systematic review of7 randomized controlled trials showed that parenteral orenteral Gln supplementation in premature infants of VLBWdoes not affect mortality (RR [95% CI]: 0.98 [0.80–1.20])or other clinical outcomes including invasive infection, NEC,time to achieve full EN, duration of hospital stay, growth, orneurodevelopmental outcomes at 18 months corrected age[129].

In contrast, studies on protein metabolism showed thatparenteral Gln may have a protein-sparing effect decreasingwhole body proteolysis and Gln de novo synthesis inpremature infants of LBW [80, 105]. However, the beneficialeffects on whole-body protein metabolism have not beenreproduced for enteral Gln [125, 127]. Hence, the routeof administration (enteral versus parenteral) should beconsidered in interpreting the effect of Gln on outcome inpremature infants [126].

Parenteral and enteral Gln supplementation is apparentlysafe in premature neonates; however, the lack of anyconsistent benefit(s) does not support its routine use in thispopulation as a whole. It is possible that any beneficial effectsof Gln are limited to specific subgroups of premature infantssuffering from acute stress (e.g., NEC, who are perhaps Glnor Arg deficient [88]) whereby increased Gln utilizationexceeds the body’s synthetic capacity [10]. Future studies


are needed to better define the role of Gln in the neonatalperiod and its mechanism of action. Large prospectivelystratified trials are needed to identify the specific subgroupsof premature neonates, who may have a greater need for Glnand who may eventually benefit from Gln supplementation.

3.3. Glutamine Supplementation in Pediatric Patients with

Gastrointestinal Disease

3.3.1. Glutamine Supplementation in Infants with SurgicalGastrointestinal Disease. While several trials have been con-ducted in VLBW infants, only 2 small double-blindedrandomized trials tested whether supplemental Gln might beof benefit in critically ill infants with surgical gastrointestinaldisease [130, 131] (Table 3). Duggan et al. [130] comparedenteral Gln to an isonitrogenous mix of nonessential aminoacids in 20 neonates and infants younger than 12 monthsreceiving PN after gastrointestinal surgery and found noapparent effect on the duration of PN (Gln: 39 d versuscontrol: 21 d, NS) or days to achieve 80% energy requirementby EN (Gln: 24 d versus control: 13 d, NS). Secondaryoutcomes (energy absorption, clinical infections, or growth)were also not affected by enteral Gln. Albers et al. [131]compared standard PN to isonitrogenous Gln-supplementedPN in 80 newborns and infants (


Ta

ble

3:G

luta

min

esu

pple

men

tati

onin

ped

iatr

icpa

tien

tsw

ith

gast

roin

test

inal

dise

ase.

Ref

eren

ceSu

bjec

tsD

esig

nG

lnC

ontr

olO

utc

omes

Res

ult

s

Alb

ers

etal

.20

05[1

31]

80n

ewbo

rns

and

infa

nts

age

≤2y

requ

irin

gP

Naf

ter

maj

ordi

gest

ive

trac

tsu

rger

y(G

A≥3

0w

k)

Ran

dom

ized

dou

ble-

blin

d

Pare

nte

ralG

ln;

ison

itro

gen

ous

isoc

alor

icG

ln(≤

0.4

g/kg

/d)

supp

lem

ente

dP

Nst

arte

don

d-2

afte

rsu

rger

yan

dre

ach

ed90

%of

reco

mm

ende

dA

Ain

take

of1.

5–2.

5g/

kg/d

(90%

ofG

lnta

rget

dose

of0.

4g/

kg/d

)by

d-4

un

tild

-31,

full

EN

,dis

char

geor

deat

h;

wit

hta

peri

ng

ofP

Nst

arte

don

oraf

ter

d-6

(n=

41)

Stan

dard

PN

star

ted

ond-

2af

ter

surg

ery

and

reac

hed

90%

ofre

com

men

ded

AA

inta

keof

1.5–

2.5

g/kg

/dby

d-4

un

tild

-31,

full

EN

disc

har

geor

deat

review article glutaminesupplementationinsickchildren ...rats with experimental nec, gln reduces...

Documents