Mehta

lines the evidence for a ketogenic response byterm infants to blood glucose concentrations aslow as <2-0 mmol/l. Therefore, it is not possibleto claim that counter-regulation has failed ifonly blood glucose concentrations are meas-ured. We feel that the presence of symptoms orthe features of an underlying disorder are moreimportant determinants of treatment than theblood glucose concentration alone, and wouldsuggest that treating all asymptomatic terminfants with blood glucose concentrations tran-siently below 2-0 mmol/l with intravenousglucose is over aggressive management.

If clinicians are not comfortable with bloodglucose concentrations as low as this, wesuggest that the first line of treatment shouldbe to increase enteral feed volumes andfrequency with formula supplementation ofbreast feeds, and check postprandial increasesin blood glucose concentration, rather than toseparate mother and baby and commenceinvasive management. In fact, we do not agreethat enteral feeds should be withheld fromhypoglycaemic infants who have previouslytolerated milk feeds, as there is no scientificsupport for Dr Mehta's speculation in section c(1) (iii) and he has not examined the effect ofwithholding enteral feeds on ketone body pro-duction. Rather, milk contains more energy/mlthan 10% dextrose and its high fat content willprovide substrate for ketogenesis. In addition,there is published evidence that enteral feeding

may stimulate glucagon release and ketonebody production.4 5We agree that adequate intakes of intra-

venous glucose or enteral milk prevent hypo-glycaemia in most preterm infants, but pointout that there is a high prevalence of hypogly-caemia in preterm infants who are fluidrestricted.5

In summary, we feel that it is inappropriateand potentially confusing to base themanagement of neonatal hypoglycaemia onhypotheses which are not fully scientificallytested. However, it should be noted that thereis agreement between the two papers on theidentification of groups at risk of failure ofmetabolic adaptation.

J M HAWDONM P WARD PLATT

A AYNSLEY-GREEN1 Hawdon JM, Aynsley-Green A, Ward Platt MP. The role of

pancreatic insulin secretion in neonatal glucoregulation. I.Healthy term and preterm infants. Arch Dis Child 1993; 68:274-9.

2 Hawdon JM, Aynsley-Green A, Bartlett K, Ward Platt MP.The role of pancreatic insulin secretion in neonatalglucoregulation. II. Infants with disordered blood glucosehomoeostasis. Arch Dis Child 1993; 68: 280-5.

3 Hume R, Burchell A. Abnormal expression of glucose-6-phosphatase in preterm infants. Arch Dis Child 1993; 68:202-4.

4 Lucas A, Bloom SR, Aynsley-Green A. Metabolic andendocrine consequences of depriving preterm infants ofenteral nutrition. Acta Paediatr Scand 1983; 72: 245-9.

5 Hawdon JM, Ward Platt MP, Aynsley-Green A. Patterns ofmetabolic adaptation for preterm and term infants in thefirst neonatal week. Arch Dis Child 1992; 67: 357-65.

6 Lucas A, Morley R, Cole T. Adverse neurodevelopmentaloutcome of moderate neonatal hypoglycaemia. BMJ7 1988;297: 1304-8.

Prevention and management of neonatalhypoglycaemia

JM Hawdon, M P Ward Platt, A Aynsley-Green

Institute of ChildHealth, University ofLiverpoolJ M Hawdon

Princess MaryMaternity Hospital,Newcastle upon TyneM P Ward Platt

Institute of ChildHealth, University ofLondonA Aynsley-Green

Correspondence to:Dr JM Hawdon, Institute ofChild Health, RoyalLiverpool Children'sHospital Alder Hey, EatonRoad, Liverpool LI 2 2AP.

Few topics in clinical paediatrics cause moreconfusion than the management of neonatalhypoglycaemia. Although it has been acknowl-edged as a clinical problem for many decades, 'its diagnosis and management continue to becontroversial issues, with disagreement evenbetween staff on the same nursery.2-4 Thisreview presents our conspectus of the clinicalsituations in which hypoglycaemia mostcommonly occurs and the management of thecondition. Rare disorders, such as inbornerrors of metabolism and complex hormonedeficiencies, which may present with hypo-glycaemia are not discussed.

Neonatal hypoglycaemia most commonlyoccurs when the normal processes of metabolicadaptation after birth fail to occur. Theseprocesses enable the healthy infant to meet twomajor metabolic challenges. First, the need tomaintain an adequate circulating concentra-tions of glucose or alternative fuels to supplythe body's organs, especially the brain, andsecond, the need to adapt to a new form ofnutrition, namely intermittent feeding with

milk. As the blood glucose concentration fallsafter the cutting of the umbilical cord anduntil milk feeding is established, the baby isentirely dependent upon its own resources tomaintain fuel supply. Glycogenolysis is initi-ated as the initial counter-regulatory process.Once the glycogen stores are exhausted,however, substrates must be mobilised frombody protein and fat stores so that gluco-neogenesis can occur. These processes areinduced by the counter-regulatory hormones(glucagon, catecholamines, growth hormone,and cortisol).

Although much is known of the basicbiochemical processes which underlie thesephenomena,5 6 clinical controversy exists sur-rounding the criteria for diagnosis of hypogly-caemia in the neonatal period, suitablemonitoring policies, and the practical implica-tions for management. This discussion firstrelates to the diagnosis of hypoglycaemia ingeneral neonatal practice, then deals withmonitoring and clinical management inspecific circumstances.

F60

on January 10, 2022 by guest. Protected by copyright.

http://fn.bmj.com

/A

rch Dis C

hild Fetal N

eonatal Ed: first published as 10.1136/fn.70.1.F

60 on 1 January 1994. Dow

nloaded from

Prevention and management of neonatal hypoglycaemia

Diagnosis ofneonatal hypoglycaemiaImmediate emphasis must be placed on theimportance of accurate diagnostic methods.Several authors have shown that reagent stripsare too inaccurate for the diagnosis of neonatalhypoglycaemia.' 2 7 We believe that, whereverpossible, accurate laboratory blood glucosemeasurements should be made for neonatalblood glucose monitoring and the diagnosis ofhypoglycaemia. The introduction of a benchtop blood glucose electrode (YSI 2300 StatPlus, YSI Ltd) in our neonatal nursery hasmarkedly facilitated blood glucose monitoringwithout adding an excessive technical work-load, in a way analogous to the instantaneous,accurate analysis of blood gas concentrationsin the nursery sideroom.

In the discussion of the diagnosis of hypo-glycaemia, it is essential to consider whichnumerical value represents hypoglycaemia,and in order to do this the significance of lowblood glucose concentrations must be under-stood. These are two important questionswhich are difficult to answer in the face ofcurrent knowledge. Of particular importance isthe recognition that previous definitions ofhypoglycaemia were based on statisticalanalysis of blood glucose concentrations in thefirst few days after birth, with values that werealmost certainly dependent on the manage-ment policies of the day.' 8 9 Improved obstet-ric management and the early introduction ofmilk feeding have created a very differentpattern of metabolic adaptation to that seen 30years ago.°0

In adults, a counter-regulatory hormonaland metabolic response occurs and symptomsof neuroglycopenia are evident at bloodglucose concentrations of 4 0 and 2-7 mmolIrespectively. "I In a small number of terminfants, neurological dysfunction was demon-strated when blood glucose concentrations fellbelow 2 6 mmolIl, whether or not the infanthad symptoms, and, moreover, resolutionof normoglycaemia did not restore normalneurological function immediately.'2 Of moreconcern is the finding in a retrospective studythat the neurological outcome of preterminfants was adversely affected if the babies hadexperienced blood glucose concentrationsbelow 2-6 mmolll on five or more days duringthe neonatal period.'3 From these data, it istempting to propose that a blood glucose con-centration below 2-6 mmol/l should be takenas the lowest which is acceptable in allneonates. However, we have challenged thisview, suggesting that hypoglycaemia is a con-tinuum, its biological effects varying accordingto the availability of alternative fuels, such asketone bodies, for brain metabolism, andaccording to the presence of other compound-ing clinical problems such as birth asphyxia,polycythaemia, hypocalcaemia, and hypercar-bia. 10 13These dilemmas may be resolved only by a

major research study in which glucose andother fuel concentrations are measured con-currently with direct measurements of brainfunction in different groups of newborninfants, and these data then related to long

term neurological outcome. Such a studywould allow the formulation of preciseguidelines for monitoring and interventionin each category of term and preterminfants. However, until such a study has beenperformed, we must consider a pragmaticapproach to the management of hypogly-caemia, and this is based upon our findingsregarding circulating fuel concentrations ininfants managed according to current policiesin our nurseries.

Metabolic adaptation in healthy,appropriate birth weight for gestationalage (AGA) infantsEarly and recent work has demonstrated thatafter the abrupt cessation of placental nutritionat birth, transient hypoglycaemia is almostuniversal in all newborn mammals. However inhuman infants, even if enteral feeds are with-held, this phenomenon is self limitingand blood glucose concentrations rise.9 10 14Therefore, this early self limiting period ofhypoglycaemia cannot be considered to bepathological, and there is little practical valuein measuring the blood glucose concentrationsof asymptomatic babies in the first two post-natal hours. After this period, blood glucoseconcentrations below 2-6 mmol/l may recur inmany healthy babies, particularly those whoare demand fed, with long between feed inter-vals, and those who are breast fed.'0 However,a marked ketogenic response to low bloodglucose concentrations has been demonstratedin these infants, and evidence from animalstudies suggests that this response is protectiveof neurological function as ketone bodies areimportant glucose sparing cerebral fuels.'0 15-18Therefore, we consider the monitoring ofblood glucose concentrations in healthy,appropriately grown neonates to be unneces-sary, and potentially harmful to parental well-being and the successful establishment ofbreast feeding.'9 20

Metabolic adaptation in preterm infants(<36 weeks' gestation)Like term babies, preterm babies will experi-ence a rapid fall in blood glucose concentrationafter delivery, but, unlike term infants they areless able to mount the counter-regulatoryresponse whereby gluconeogenesis and keto-genesis occur.10 21 In addition, preterm infants<32 weeks' gestation are at increased risk ofother neurological insults, such as cerebralischaemia. Therefore, clinical attention toblood glucose concentrations may be regardedas being more important when dealing withpreterm babies, and, in the light of the findingsof Lucas et al,'3 we propose that concentra-tions <2 6 mmolI should be avoided. Thisview is contrary to older beliefs that smallbabies can be expected to tolerate lower bloodglucose concentrations than large ones.'

However, current management policies ofearly provision of enteral and/or intravenousfluids, usually ensure that this early hypo-glycaemia is rapidly corrected, and, provided

F61

on January 10, 2022 by guest. Protected by copyright.

http://fn.bmj.com

/A

rch Dis C

hild Fetal N

eonatal Ed: first published as 10.1136/fn.70.1.F

60 on 1 January 1994. Dow

nloaded from

Hawdon, Ward Platt, Aynsley-Green

the exogenous supply of energy is sufficient,hypoglycaemia will not subsequently recur.10This supply of energy should be institutedimmediately after birth and the initial prescrip-tion of enteral or intravenous fluids shouldnot be influenced by an immediate postnatalblood glucose concentration. Therefore, themeasurement of blood glucose concentrationimmediately after birth adds little clinicalinformation and should not affect manage-ment. It is much more rational and of greaterimportance to ensure that the prescribedenergy supply is adequate by accuratelymeasuring the blood glucose concentration1-2 hours after intravenous or enteral feedinghas been commenced, so that necessary adjust-ments can be made.

In our experience, neonatal hypoglycaemiacan be prevented in preterm babies by the pro-vision of enteral and/or intravenous feeds (asformula milk and 10% dextrose respectively) atrates of at least 100 ml/kg/day.'0 However,some sick infants will not tolerate thesevolumes and, as described previously, bloodglucose monitoring should confirm that theirrates of glucose supply are adequate. If bloodglucose concentrations are low (<2.6 mmolI),this suggests that glucose demand exceedssupply. Therefore, glucose infusion ratesshould be increased, either by increasing theconcentration of the glucose solution or byincreasing the intravenous fluid infusion rate.Neonatal blood glucose concentrations corre-late closely with glucose infusion rates.22 Ifthere are symptoms of hypoglycaemia (fits orreduced level of consciousness) a small bolusdose of 10% intravenous glucose (3 ml/kg)may be given, but should always be followedimmediately by a glucose infusion at an ade-quate rate to meet the baby's ongoing glucoserequirement (at least 5 mg glucose/kg/min).Bolus injections of large volumes of hypertonicglucose solutions should be avoided becausethey cause a rapid rise in blood glucose con-centrations, which may be harmful to neuro-logical function and may be followed byrebound hypoglycaemia. In older childrensuch treatment has been shown to causecerebral oedema or even death.23 In additionclinical experience suggests that such highglucose concentrations are caustic to fragileneonatal veins.Once adequate blood glucose concentra-

tions are achieved, it is not necessary tomeasure blood glucose concentrationsrepeatedly in a clinically stable baby unlessthere is a reduction in exogenous energysupply. As there is a risk of 'rebound hypogly-caemia', reductions of intravenous glucoseinfusion rates should be gradual and 'tissued'drips should be resited promptly. Clearly, anychange in clinical condition or reduction influid or nutrient intake should be followed byblood glucose measurement to ensure suchchanges have not resulted in hypoglycaemia.

Finally, attention should be drawn to theimportance of enteral milk feeding. Milkformulas provide more energy/ml than 10%dextrose, and supply important non-glucosefuels, which have a glucose sparing role in

neurological function. There is evidence thatenteral milk feeding enhances the vitalprocesses of metabolic adaptation by promot-ing ketogenesis and gut maturation.10 24In immature babies who may have reducedmilk absorption secondary to delayed gastricemptying and reduced gut motility, theadequacy of exclusive enteral feeding to main-tain normoglycaemia should be confirmed byaccurate blood glucose measurement untilthese measurements are stable. Both prefeedand postprandial values should be considered,as there is a feed related cyclical variation inblood glucose concentrations.'5

Metabolic adaptation in small forgestational age (SGA) infantsWe propose that the SGA infants of most con-cern are those whose birth weight falls belowthe third centile or those who have clinicallyobvious wasting of adipose tissue and muscle(regardless of their birthweight centile), butwith preservation of head growth. For thelatter group of babies, measurement of themid-arm circumference: head circumferenceratio has been shown to be useful in predict-ing those who are at risk for metabolicproblems.25-27SGA infants are at risk of the adverse effects

of hypoglycaemia by virtue of impaired gluco-neogenesis and ketogenesis, both immediatelyafter birth and in response to persistenthypoglycaemia.27-30 As the patterns of meta-bolic adaptation and nutritional requirementsof preterm SGA infants are similar to thoseof preterm AGA infants, this section willfocus on the management of the term SGAneonate.SGA infants are at risk of adverse neuro-

logical sequelae of neonatal hypoglycaemia fora number of reasons: chronic hypoglycaemia,acidosis and hypoxia may have occurred inutero3l 32; perinatal asphyxia is more com-mon33; there is an inability to mount a keto-genic response to hypoglycaemia, so that thereis no provision of alternative fuels for brainmetabolism27; and enteral feed tolerance maybe poor, with an increased risk of necrotisingenterocolitis.33 Therefore, accurate bloodglucose monitoring should be performed inthis group of infants. Both prefeed and post-prandial values should be considered, as thereis a feed related cyclical variation in bloodglucose concentrations.'5As with preterm infants, it is important

to ensure the early provision of exogenouscalories, to curtail the early postnatal hypogly-caemic period, and to prevent subsequenthypoglycaemia. It is our experience that thiscan usually be achieved by regular, formulamilk feeds every three hours at volumes of atleast 120 ml/kg/day. As breast milk appears tobe more ketogenic than formula milk,'0 it isimportant that breast feeding is encouragedwhen this is the mother's wish, by allowing thebaby to suckle at each feed time, and then bysupplementing the breast feed with formulamilk, by gavage if necessary. It should be notedthat the energy content of formula milk is

F62

on January 10, 2022 by guest. Protected by copyright.

http://fn.bmj.com

/A

rch Dis C

hild Fetal N

eonatal Ed: first published as 10.1136/fn.70.1.F

60 on 1 January 1994. Dow

nloaded from

Prevention and management of neonatal hypoglycaemia

approximately 2750 kJ/l, while that of 10%dextrose is 1600 kJ/l. If necessary high energymilk formulas may be used, for example thosedesigned for preterm infants, or energy sup-plements (for example, Duocal, ScientificHospital Supplies) may be added to enteralmilk feeds. By these means, it should bepossible to meet the energy needs of most fullterm SGA babies with enteral milk feeds with-out recourse to intravenous glucose.When babies cannot tolerate these enteral

feed volumes, or if the above regimen fails toprevent hypoglycaemia (blood glucose concen-tration <2-6 mmol/1), an intravenous glucoseinfusion may be required. Hypoglycaemia isusually rapidly reversed by intravenous glucoseinfusion, at a rate of 5 mg/kg/min (= 72mi/kg/day of 10% dextrose) without a bolusglucose injection.34 This rate can then beadjusted, up or down, depending on bloodglucose concentrations. Because of the benefi-cial effects of enteral feeding, milk feeds mustnot be discontinued or reduced when intra-venous fluids are given (unless, of course, theinfant is developing necrotising enterocolitis).Our practice is to feed the baby with as muchmilk as is tolerated, and to infuse glucose at arate sufficient to prevent hypoglycaemia, sothat the fluid regimen is tailored to each baby.

Finally, there are some SGA infants who havepersistent hypoglycaemia or excessively highglucose requirements. Glucose infusions maybe required for up to one week, and rapidreductions or wide swings in glucose infusionrates should be avoided. By maximising enteralfeed intakes and gradually reducing intravenousinfusion rates, almost all infants eventuallyachieve the ability to control their own bloodglucose concentrations. Additional treatments,such as intravenous glucagon, may be of valueand are discussed in detail elsewhere.3436

Infants with hyperinsulinismThis phenomenon is believed to be commonlyfound in the infant of the poorly controlleddiabetic mother, but may occur in associa-tion with rarer conditions, such as Beckwith-Weidemann syndrome and islet cell dys-regulation syndrome (nesidioblastosis) whosemanagement is reviewed elsewhere.37 In theselatter conditions, hypoglycaemia is commonbut, despite plentiful adipose tissue stores, thelipolytic and ketogenic responses are suppressedby high circulating insulin concentrations.

Although many studies have outlined thepossible mechanisms of early hypoglycaemia inthe infant of the diabetic mother,38A0 therehave been no studies which have investi-gated the occurrence ofhypoglycaemia, and, ofgreater importance, the ketogenic responsebeyond the immediate postnatal period, par-ticularly in the context of improved maternaldiabetic control. However, it is our clinicalexperience that neonatal hypoglycaemia is nowrare in insulin dependent mothers managedaccording to current practices.

It should be pointed out that previousdiagnoses of 'neonatal hyperinsulinism' inthese insulin dependent mothers were made

with reference to insulin/glucose relationshipsin older subjects. We have recently demon-strated that even healthy, asymptomatic termneonates have insulin/glucose relationshipswhich differ markedly from those of olderchildren and that, for neonates, plasmainsulin and blood glucose concentrations arepoorly related and baseline insulin secretionappears to be high.4' Therefore, we suggestthat the diagnosis should only be made onclinical grounds, that is a high glucoserequirement to maintain normoglycaemia,together with biochemical evidence of anabsent lipolytic and ketogenic response tohypoglycaemia, with reference to data fromhealthy neonatal subjects. We suggest that,using these criteria hyperinsulinism is nowrare in insulin dependent mothers and thatthese babies could be treated as healthy termbabies, provided they are asymptomatic.This is the hypothesis to be addressed in anongoing research study.

Infants with perinatal asphyxiaLike small for gestational age infants, thesebabies may be at increased risk of hypogly-caemia which may compound other severeneurological insults. In addition, we havefound that under these circumstances thereis no ketogenic response to hypoglycaemia(unpublished data), and some of these infantshave unusually high glucose requirements(>12 mg/kg/min).42 The exogenous supply offuels to minimise cerebral dysfunction must beensured but often these infants are fluidrestricted and hypoglycaemia may occur.Therefore, as with preterm and SGA babies,accurate blood glucose measurement ismandatory and the supply of exogenousenergy, as concentrated glucose solutions ifnecessary, must be maintained. However, wehave found that these infants are also at risk ofhyperglycaemia which may be secondary toincreased stress hormone release (unpublisheddata), and suggest that blood glucose monitor-ing and carefully prescribed glucose infusionrates should be utilised to prevent bothextremes of loss of glycaemic control. Thesame general considerations apply to anyinfant who has a severe systemic illness.

In summary, there are some groups ofneonates who are at risk, first, of an increasedincidence of hypoglycaemia and, second,adverse neurological sequelae when hypogly-caemia occurs. Even within these groups, eachinfant will have a different level of risk, andsome of the protective mechanisms, such aschanges in cerebral blood flow or ketone bodyproduction, cannot be easily quantified.Therefore, at the present level of knowledge, werecommend that for the following groups ofinfants blood glucose concentrations should bemonitored, until they are stable within an opti-mal range, that is blood glucose concentrationsbelow 2-6 mmoJIl and above 10 mmol/l shouldbe avoided: preterm babies, especially whensick; SGA babies; infants with proved hyper-insulinism; and infants with perinatal asphyxiaor with other systemic illnesses (figure). It is

F63

on January 10, 2022 by guest. Protected by copyright.

http://fn.bmj.com

/A

rch Dis C

hild Fetal N

eonatal Ed: first published as 10.1136/fn.70.1.F

60 on 1 January 1994. Dow

nloaded from

Hawdon, Ward Platt, Aynsley-Green

Intention to milk feed Intravenous fluids only

Enteral milk feeds Infuse glucose >5 mg/kg/min>120 ml/kg/day (72 ml/kg/day 10% dextrose)

Unable to tolerate

full feed volume 4Blood glucose <2 6 mmol/l Blood glucose <2 6 mmol/l

Abletotolerate-_ NoNomore milk

Symptoms No symptoms

Yes 4

Increase glucose infused Glucagon(volume +/- strength).

Increase milk Recheck blood glucosevolume and feed

frequency.Recheck blood

glucose Bolus 10% dextrose (3 ml/kg)+ infuse >5 mg/kg/min

Outline for prevention and management ofhypoglycaemiain at risk groups.

possible that further scientific investigation,correlating circulating fuel availability withneurological function and neurodevelopmentaloutcome, will identify more accurately thosegroups who are most at risk, so that more appro-

priate management policies can be applied.Dr Hawdon was supported by the Scientific and ResearchCommittee of Newcastle Health Authority and by a Foundationfor the Study of Infant Deaths training fellowship.

1 Aynsley-Green A. Glucose: a fuel for thought! _7 PediatrChild Health 1991; 27: 21-30.

2 Comblath M, Schwartz R. Hypoglycemia in the neonate.

In: Cornblath M, Schwartz R, eds. Disorders of carbo-hydrate metabolism in infancy. Philadelphia: W B Saunders,1991.

3 Koh THHG, Eyre JA, Aynsley-Green A. Neonatal hypogly-caemia- the controversy regarding definition. Arch DisChild 1988; 63: 1386-8.

4 Cornblath M, Schwartz R, Aynsley-Green A, Lloyd JK.Hypoglycemia in infancy: The need for a rational defini-tion. A Ciba Foundation discussion meeting. Pediatrics1990; 85: 834-7.

5 Ogata ES. Carbohydrate metabolism in the fetus andneonate and altered neonatal glucoregulation. Pediatr ClinNorth Am 1986; 33: 25-44.

6 Girard J. Metabolic adaptation to change of nutrition atbirth. Biol Neonate 1990; 58 (suppl):3-15.

7 Holtrop PC, Madison KA, Kieckle FL, Karcher RE, BattonDG. A comparison of chromogen test strip (Chemstripbg) and serum glucose values in newborns. Am 7 Dis Child1990; 144: 183-5.

8 Cornblath M, Reisner SH. Blood glucose in the neonateand its clinical significance. N Engl _7 Med 1965; 273:378-80.

9 Srinivasan G, Pildes RS, Cattamanchi G, VioraS, LilienLD. Plasma glucose values in normal neonates: a new

look. JPediatr1986; 109: 114-7.10 Hawdon JM, Ward Platt MP, Aynsley-Green A. Patterns of

metabolic adaptation for preterm and term neonates inthe first postnatal week. Arch Dis Child 1992; 67:357-65.

11I de Feo P, Gallai V, Mazzotta G, et al. Modest decrements inplasma glucose concentration cause early impairment incognitive function and later activation of glucose counter-regulation in the absence of hypoglycemic symptoms innormal man..7 Clin Invest 1988; 82: 436-41.

12 Koh THHG, Aynsley-Green A, Tarbit M, Eyre JA. Neuraldysfunction during hypoglycaemia. Arch Dis Child 1988;63: 1353-8.

13 Lucas A, Morley R, ColeTJ. Adverse neurodevelopmentaloutcome of moderate neonatal hypoglycaemia. BM.1988; 297: 1304-8.

14 Cornblath M, Reisner SH. Blood glucose in the neonateand its clinical significance. N Engl _7 Med 1965; 273:378-80.

15 Lucas A, Boyes S, Bloom SR, Aynsley-Green A. Metabolicand endocrine responses to a milk feed in 6 day old terminfants: differences between breast and cow's milkformula feeding. Acta Paediatr Scand 1981; 70: 195-200.

16 Edmond J, Auestad N, Robbins RA, Bergstrom JD. Ketonebody metabolism in the neonate: development and theeffect of diet. Federal Proceedisgs 1985; 44: 2359-64.

17 Medina JM, Fernandez E, Bolans JP, Vicario C, ArizmendiL. Fuel supply to the brain during the early postnatalperiod. In: Cuezva JM, Pascual-Leone AM, Patel MS,eds. New York: Plenum, 1990.

18 Yager JY, Heitjan DF, Towfighi J, Vannucci RC. Effect ofinsulin-induced and fasting hypoglvcaemia on perinatalhypoxic-ischemic brain damage. Pediatr Res 1992; 31:138-42.

19 Hawdon JM. Neonatal hypoglycacmia: the consequences ofadmission to the special care nursery. Child Health 1993;Feb:48-51.

20 Hawdon JM, Ward Platt MP, Aynsley-Green A. Neonatalhypoglycaemia - blood glucose monitoring and infantfeeding. Midwifery 1993; 9: 3-6.

21 Cowett RM, Schwartz R. The role of hepatic control ofglucose homeostasis in the etiology of neonatal hvpo andhyperglycemia. Semnin Perinatol 1979; 3: 327.

22 Hawdon JM, Aynsley-Green A, Bartlett K, Ward Platt MP.The role of pancreatic insulin secretion in neonatal glu-coregulation. II Infants with disordered blood glucosehomeostasis. Arch Dis Child 1993; 68: 280-5.

23 Shah A, Stanhope R, Matthew D. Hazards of pharmacolog-ical tests of growth hormone secretion in childhood. BM.1992; 304: 173-4.

24 Lucas A, Bloom SR, Aynsley-Green A. Gastrointestinalpeptides and the adaptation to extrauterine nutrition. CanYPhysiolPharmacol 1985; 63: 527-37.

25 Sasanow SR, Georgieff MK, Pereira GR. Mid arm circum-ference and mid arm circumference/head circumferenceratios: standard curves for anthropometric assessmentof neonatal nutritional status. Y Pediatr 1986; 109:311-5.

26 Georgieff MK, Sasanow SR, Chockalingham UM, PereiraGR. A comparison of the mid arm circumference/headcircumference ratio and ponderal index for the evaluationof newborn infants after abnormal intrauterine growth.Acta PaediatrScand 1988; 77: 214-9.

27 Hawdon JM, Ward Platt MP. Metabolic adaptation in smallfor gestational age infants. Arch Dis Child 1993; 68:262-8.

28 Hay WW. Fetal and neonatal glucose homeostasis and theirrelation to the small for gestational age infant. SeciinPerinatol 1984; 8:101-16.

29 Jones RAK, Roberton NRC. Problems of the small for datesbaby. Cliii Obstet Gynzecol 1984; 11: 499-524.

30 de Leeuw R, de Vries Ii. Hypoglycemia in small for datesnewborn infants. Pediatnrcs 1976; 58: 18-22.

31 Economides DL, Nicolaides KH. Blood glucose and oxygentension levels in small for gestational age fetuses. Ams YObstet Gynzecol 1989; 160: 385-9.

32 Hawdon JM, Ward Platt MP, McPhailS, Cameron H,Walkinshaw SA. Prediction of impaired metabolic adapta-tion by antenatal Doppler studies in small for gestationalage fetuses. Arch Dis Child 1992; 67: 789-92.

33 Hackett GA, CampbellS, Gamsu H, Cohen-Oberlech T,Pearce JMF. Doppler studies in the growth retarded fetusand prediction of neonatal necrotising enterocolitis,haemorrhage and neonatal morbidity. BM3 1987; 294:13-6.

34 Hawdon JM, Aynsley-Green A, Ward Platt MP. Neonatalblood glucose concentrations: metabolic effects of intra-venous glucagon and intragastric medium chain triglyc-eride. Arch Dis Child 1993; 68: 255-61.

35 Mehta A, Wootton R, Cheng KL, Penfold P, Halliday D,Stacey TE. Effect of diazoxide or glucagon on hepaticglucose production rate during extreme hvpoglvcaemia.Arch Dis Child 1987; 62: 924-30.

36 Carter PE, Lloyd DJ, Duffty P. The use of glucagon in themanagement of hypoglycaemia in the small for gestationalage infant. A report of 25 cases. Arch Dis Child 1988; 63:1264-7.

37 Aynsley-Green A, Soltesz G. Disorders of blood glucosehomeostasis in the neonate. In: Roberton NRC, ed.Neonatology. Edinburgh: Churchill Livingstone, 1992.

38 Pedersen J, Bejsen-Moller B, Poulsen H. Blood sugar innewborn infants of diabetic mothers. Acta Enidonnlol(Copenh) 1954; 15: 33.

39 Kuhl C, Andersen GE, Hestel J, Molsted-Pedersen L.Metabolic events in infants of diabetic mothers duringfirst 24 hours after birth. I Changes in plasma glucose,insulin and glucagon. Acta Paediatr Scand 1982; 71:19-25.

40 Andersen 0, Hestel J, Scholonder L, Kuhl C. Influence ofmaternalplasmaglucose concentration at delivery on therisk of hypoglycaemia in infants of insulin dependentdiabetic mothers. Acta Paediatr Scanid 1985; 74:268-73.

41 Hawdon JM, Aynsley-Green A, Alberti KGMM, WardPlatt MP. The role of pancreatic insulin secretion in neo-natal glucoregulation. I. Healthy term and preterminfants. Arch Dis Child 1993; 68: 274-9.

42 Bhowmick SK, Lewandowski L. Prolonged hyperinsuli-naemia and hypoglycaemia in an asphyxiated small forgestation infant. Case management and literature review.Clin Pediatr (Phila) 1989; 28: 575-8.

F64

on January 10, 2022 by guest. Protected by copyright.

http://fn.bmj.com

/A

rch Dis C

hild Fetal N

eonatal Ed: first published as 10.1136/fn.70.1.F

60 on 1 January 1994. Dow

nloaded from