Hyperglycaemia and the very preterm baby

https://doi.org/10.1016/j.siny.2005.04.008Get rights and content

Summary

Neonatal hyperglycaemia, as usually defined (a whole blood glucose of >7 mmol/L), is common in the first week of life in babies born more than 12 weeks early. However, a review of a cohort of all such births in the north of England suggests that significant glycosuria is uncommon, and that there is no threat of an osmotic diuresis until the urine contains 2% glucose (by which time the blood glucose level almost always exceeds 15 mmol/L). The current statistical or epidemiological definition of hyperglycaemia (derived from data on term babies) needs to be replaced, for clinical purposes, by a more operationally relevant definition.

Introduction

Much has been written about neonatal hypoglycaemia since clinicians first came to recognize, more than 50 years ago, the harm that this may cause. In contrast, little has been written about the hazards of neonatal hyperglycaemia since it first became clear, 30 years ago,1, 2, 3, 4 that this too could be a problem in the intravenously fed, very low birth weight infant. The review that follows will look at what has been written about this condition, summarize what is known about the renal handling of glucose in the very preterm baby, and report the outcome of a prospective study of blood glucose levels in 781 babies of less than 32 weeks' gestation in the first 10 days of life. It will end by suggesting that some of the ‘hype’ needs to be taken out of our current concern for ‘hyperglycaemia.’

Much of the confusion that has developed relates to our failure to define what we mean by the words we use when we talk about a specific medical condition. Clarity of thought is difficult until a term has been delineated with some precision, and meaningful communication is impossible until the reader also knows how the term is being used. The late Marvin Cornblath5 did more than anyone else to make clinicians aware of the many different ways in which the word ‘hypoglycaemia’ has come to be used, and these issues are discussed in some detail in the article by Williams in this issue. The term can be employed to describe a recognizable clinical condition, but most usage has suggested that authors have been invoking a statistical or functional definition. However, in the last 5 years it has been suggested that an ‘operational’ definition would be more clinically useful.6 No comparable attempt had yet been made to look, with comparable rigor, at the way the word ‘hyperglycaemia’ is being used.

Section snippets

Evolution of the use of the word hyperglycaemia

Applying a statistical approach, a level of more than 7 mmol/L (∼125 mg/dL) is unusual in the healthy term baby,5 although some would argue for a value closer to 8.3 mmol/L (150 mg/dL), especially in the preterm baby.7 However, the glucose present in any laboratory sample of whole blood falls quite rapidly after collection if it is not kept ‘on ice’ prior to measurement, even when the specimen is collected into a tube containing fluoride (unless the red cells are removed by prompt

Hyperglycaemia as a statistically defined entity

But are we using statistics appropriately? People do not judge the blood pressure of a baby using adult norms. Should the blood glucose level of a preterm baby be defined using the norms that are appropriate to a term baby? Should we be using the norms that describe what is usually found in an enterally fed baby to define what is ‘normal’ in a baby receiving nothing except intravenous glucose? And, given that the incidence of ‘hyperglycaemia’, as conventionally defined, is very much higher in

Functional hyperglycaemia

In fact, almost all the clinicians who have written about the high incidence of hyperglycaemia in the intravenously fed very low birth weight baby have actually invoked a functional rather than a statistical definition of hyperglycaemia, because they have all stressed the risk of glycosuria, of an osmotic diuresis, and of dehydration when the blood glucose level rises above ‘normal’ and used this as an implied justification for defining hyperglycaemia as a whole blood glucose in excess of 7

Hyperglycaemia: an ‘operational’ definition

Nobody doubts that hyperglycaemia can be dangerous. What does not seem to have been established is just how high the blood glucose level has to be before it triggers an osmotic diuresis. If such a level could be identified it might be possible to develop an operational definition of neonatal hyperglycaemia, similar to the recently proposed operational definition for hypoglycaemia—a level chosen so that it keeps unnecessary intervention to a minimum but prompts timely avoidance action before the

Renal handling of glucose in the very preterm baby

Contrary to general belief, the urine always contains glucose, unless it has become bacterially contaminated, but tubular reabsorption of what has filtered into the urine in the renal glomerulus is generally so efficient that what remains (often only 0.1% of the filtered load) goes undetected during side-ward testing. Reabsorptive capacity seems to be proportionately much the same in infancy as in later life.45 Reabsorption is almost complete at low filtered loads, but it becomes increasingly

Severe hyperglycaemia in babies of less than 32 weeks' gestation

Severe hyperglycaemia (a whole blood glucose ≥20 mmol/L) was identified in the first 10 days of life in 28 of the 781 liveborn babies of <32 weeks' gestation born in the north of England in 1990–1991 (at a time when insulin was never used to combat neonatal hyperglycaemia), and it was five times commoner in babies of <28 weeks' gestation than in those older than this. Only three of the 28 were getting more than 10% of their fluid by mouth at the time, and one of these three became

Why do very preterm babies become hyperglycaemic?

Our understanding of the hormonal mechanisms that control glucose metabolism in the neonate remains fragmentary and incomplete. Insulin increases the uptake and utilization of glucose, simultaneously inhibiting new glucose production (gluconeogenesis), while a range of counter-regulatory hormones—including adrenaline, noradrenaline, glucagon and cortisol—have the opposite effect. However, insulin also promotes amino acid entry into muscle and protein synthesis, enhances fatty synthesis in the

Hyperglycaemia can be factitious or iatrogenic

It is worth checking that the sample has not been collected from a line down which glucose was being infused, or shortly after the infusion of an inadvertent bolus of glucose (Fig. 2).

Hyperglycaemia can be a sign of stress

Hyperglycaemia can be caused by clinical stress,13, 15 and its sudden appearance can be the first sign of infection, especially in a previously stable baby.28, 40 It can also be a sign of some intracerebral catastrophe, or of incipient necrotizing enterocolitis. It is extremely common following surgery,56 even

Conclusions

The dangers associated with hyperglycaemia in the intravenously fed preterm baby of less than 28 weeks' gestation have almost certainly been over-‘hyped’ in the past. Levels that are statistically uncommon in the term baby (7 mmol/L or 125 mg/dL) are certainly not uncommon in the preterm baby. However, they are not hazardous, and levels twice as high as this never seem to trigger a diuresis.

That does not mean that the risk of hyperglycaemia can be ignored. We may, however, have reached the

Acknowledgements

The previously unpublished information on the incidence of hyperglycaemia in babies born more than 8 weeks early reported here comes from a collaborative regional study undertaken by neonatal nurses in the north of England as part of the work of the Northern Neonatal Nursing Initiative (NNNI) Trial Group in 1990 and 1991.72, 73 The information on the later developmental progress of these children was provided by Dr Win Tin, now a consultant neonatologist in Middlesbrough. Dr Tin saw each and

References (73)

  • D.H. Adamkin

    Nutrition in the very low birth weight infant

    Clin Perinatol

    (1986)
  • S.G. Shaffer et al.

    Fluid requirements in the preterm infant

    Clin Perinatol

    (1992)
  • R.D. Leake

    Perinatal nephrobiology: a developmental perspective

    Clin Perinatol

    (1977)
  • R.S. Pildes

    Current literature and clinical issues: neonatal hyperglycemia

    J Pediatr

    (1986)
  • N.D. Binder et al.

    Insulin infusion with parenteral nutrition in extremely low birth weight infants with hypernatremia

    J Pediatr

    (1989)
  • J.W. Collins et al.

    A controlled trial of insulin infusion and parenteral nutrition in extremely low birth weight infants with glucose intolerance

    J Pediatr

    (1991)
  • B.B. Poindexter et al.

    Exogenous insulin reduces proteolysis and protein synthesis in extremely low birth weight infants

    J Pediatr

    (1998)
  • F. Thabet et al.

    Continuous insulin infusion in hyperglycaemic very-low-birth-weight infants receiving parenteral nutrition

    Clin Nutr

    (2003)
  • L. Finberg

    The relationship of intravenous infusions and intracranial hemorrhage – a commentary

    J Pediatr

    (1977)
  • G. Srinivasan et al.

    Plasma glucose changes in preterm infants during oral theophylline therapy

    J Pediatr

    (1983)
  • H.S. Dweck et al.

    Glucose intolerance in infants of very low birth weight. I. Incidence of hyperglycaemia in infants of birth weights 1,100 grams and less

    Pediatrics

    (1974)
  • Y.W. Brans et al.

    Feeding the low birth weight infant: orally of parenterally? Preliminary findings of a comparative study

    Pediatrics

    (1994)
  • H.S. Dweck et al.

    Glucose intolerance in infants of very low birth weight: II. Intravenous glucose tolerance tests in infants of birth weight 500–1380 grams

    Biol Neonat

    (1994)
  • M. Cornbath et al.

    Disorders of carbohydrate metabolism in infancy

    (1991)
  • M. Cornblath et al.

    Controversies regarding definition of neonatal hypoglycaemia: suggested operational thresholds

    Pediatrics

    (2000)
  • J.M. Hawdon et al.

    Metabolic disease. Part I. Disorders of glucose homeostasis in the neonate

  • M. Zarif et al.

    Insulin and growth hormone responses in neonatal hyperglycemia

    Diabetes

    (1976)
  • R.M. Cowett et al.

    Glucose disposal of low birth weight infants: steady state hyperglycemia produced by constant intravenous glucose infusion

    Pediatrics

    (1979)
  • C. Louik et al.

    Risk factors for neonatal hyperglycemia associated with 10% dextrose infusion

    Am J Dis Child

    (1985)
  • S.A. Deshpande et al.

    Measuring blood glucose in neonatal units: how does HemoCue compare?

    Arch Dis Child

    (1996)
  • M. Cornblath et al.

    Blood glucose in the neonate and its clinical significance

    N Engl J Med

    (1965)
  • R.K. Ditchburn et al.

    Plasma glucose levels in infants weighing 2,500 g and less fed immediately after birth with breast milk

    Biol Neonat

    (1967)
  • M. Cornblath et al.

    Studies of carbohydrate metabolism in the newborn infant. III Some factors influencing the capillary blood sugar and the response to glucagon during the first hours of life

    Pediatrics

    (1961)
  • S. Singhi

    Effect of maternal intrapartum glucose therapy on neonatal blood glucose levels and neurobehavioral status of hypoglycaemic term newborn infants

    J Perinat Med

    (1988)
  • W. McGuire et al.

    ABC of preterm birth. Care in the early newborn period

    Br Med J

    (2004)
  • A.A. Fanaroff et al.

    Incidence, presenting features, risk factors and significance of late onset septicaemia in very low birth weight infants

    Pediatr Infect Dis J

    (1998)
  • Cited by (73)

    • Pathophysiology and Management of Disorders of Carbohydrate Metabolism and Neonatal Diabetes

      2020, Maternal-Fetal and Neonatal Endocrinology: Physiology, Pathophysiology, and Clinical Management
    • Pathophysiology and Management of Disorders of Carbohydrate Metabolism and Neonatal Diabetes

      2019, Maternal-Fetal and Neonatal Endocrinology: Physiology, Pathophysiology, and Clinical Management
    • The state of variability: A vision for descriptors of glycaemia

      2019, Annual Reviews in Control
      Citation Excerpt :

      They are based on observed distributions of BG in term infants (Hey, 2005), where less common extremes are used to define abnormal BG. However, these statistical definitions may simply describe measured BG in this cohort, rather than what may be clinically desirable, beneficial, or obtainable via GC (Hey, 2005). Neonatal hypoglycaemia is controversial, where agreement has not been reached on either thresholds or treatment (Harris, Weston, & Harding, 2012; Harris, Weston, Signal, Chase, & Harding, 2013; McKinlay et al., 2015; Rozance & Hay, 2010), in part because it is often asymptomatic.

    View all citing articles on Scopus
    View full text