- William W. Hay Jr, MD*
Ordinarily, invited authors will write the review articles and commentaries for NeoReviews. However, as the coeditor primarily responsible for this inaugural edition, I wanted to take the prerogative to write an introduction that will set the stage for what each edition will include and explain how David Stevenson and I have developed topics, authors, content, and format. We chose the topics of hypoglycemia and hyperglycemia for this first edition for several reasons that should illustrate our vision of NeoReviews.
Neonatal hypoglycemia is both a historical and a current “hot topic.” It has been studied thoroughly and has been the subject of extensive debate, yet it remains a common problem that is difficult to define, understand, and manage. The evidence is very clear from human and animal studies that severe, prolonged hypoglycemia leads to acute neurologic injury that often can have permanent sequelae. Frequently overlooked because of this ultimate serious outcome is the more common problem of hypoglycemia leading to cellular energy failure and generalized organ dysfunction.
Despite such serious and common consequences and years of investigation into its characteristics, the definition of hypoglycemia remains elusive. Fifty years ago, Norval, Kennedy, and Berkson defined the normal concentration range of blood sugar (glucose) in term newborns as 0.83 to 6.66 mmol/L (15 to 120 mg/dL), with a mean of 3.39±0.77 mmol/L (SD) (61.0±13.8 mg/dL). Ninety-five percent of infants had a concentration of 2.89 to 3.89 mmol/L (52 to 70 mg/dL). Such values are perfectly acceptable and within the normal range today. Why, therefore, is there still controversy over what is considered the normal lower limit of glucose concentration in blood or plasma of newborns and what values constitute hypoglycemia?
As discussed by Jane E. McGowan, MD, in this issue, many infants experience transiently low and highly variable glucose concentrations, but no data link such conditions directly to lasting neurologic impairment. Also, early surveys of preterm infants documented statistically lower “normal” glucose concentrations (ie, 1.11 to 2.22 mmol/L[ 20 to 40 mg/dL]) that often have been interpreted as clinically acceptable for prolonged periods despite clinical signs of hypoglycemia. Furthermore, rather inaccurate reagent strip methods of measuring glucose concentrations have led many to accept a very broad range of normal values for glucose concentrations in term or preterm newborns. In addition, animal studies show considerable variability among plasma glucose concentrations, rates of brain glucose uptake and metabolism, and neuronal function. There also is considerable variability among associated factors that should affect the relationship between plasma glucose concentration and neuronal function. These include, but certainly are not limited to, the rate of cerebral plasma flow, expression and activity of cerebral endothelial cell and neuronal glucose transporters, circulating concentrations of alternate substrates for neuronal metabolism, developmental stage of the infant, and functional aspects of the mechanisms of synaptic transport, excitotoxic neurotransmitter levels, and neuronal ion channel activity.
Thus, there appears to be no consistent threshold or lower limit single value for plasma glucose concentration below which neurologic impairment or injury invariably begins or actually develops. Not surprisingly, clinicians remain uncertain about when to say an infant is“ hypoglycemic” and how aggressive to make treatment in the presence of relatively lower glucose concentrations or clinical signs of hypoglycemia.
In contrast, recent studies and management trends have indicated that the normal lower limit of glucose concentration in preterm or term newborns should be above 2.78 mmol/L (50 mg/dL). More accurate and reliable bedside methods now are available to measure blood glucose concentrations rapidly and are part of standard management. A retrospective study by Lucas et al showed greater incidences of mental and motor delays associated with increased episodes of blood glucose concentrations below about 2.66 mmol/L (48 mg/dL) in preterm infants. In support of this possible threshold value for neonatal glucose concentration, a survey study by Srinivasan et al provided mean and 95% confidence interval values of plasma glucose concentrations in normal, healthy preterm infants. Their data defined a lower 95% CI value of about 3.33 mmol/L (60 mg/dL) by the second to third day of postnatal life. Finally, a study by Marconi et al showed that healthy, normal human fetuses had umbilical venous glucose concentrations above 2.78 mmol/L (50 mg/dL) from 20 weeks’ gestation to term. These data clearly demonstrate that normal human fetuses grow and develop in an environment that has a glucose concentration much closer to that in the plasma of normal term infants in postnatal life.
The clinical outcome of neonatal hypoglycemia also remains uncertain, especially in relation to its variable clinical pattern, severity, and association with other conditions or diseases. We still do not have quantitative evidence, in general or in specific cases, of how simultaneous hypoxemia, hyperviscosity, cerebral blood flow rate, sepsis, and many other factors interact with plasma glucose concentration to produce abnormal cellular glucose uptake, metabolism, or function that can lead to impaired neurodevelopmental outcome.
Even the causes of neonatal hypoglycemia, although generally well understood, remain an enigma in many clinical situations. We know that extremely low-birthweight (ELBW) and small for gestational age infants have limited glycogen reserves and, thus, a reduced capacity for glucose production from glycogenolysis. They also can have limited gluconeogenesis due to low fat stores, resulting in limited release of free fatty acids, limited production of glycerol, and limited conversion of amino acids to glucose. However, we do not know how these processes are regulated abnormally when severe and persistent hypoglycemia occurs. Additionally, abnormally persistent high insulin concentrations can lead to horrendously difficult cases of persistent, severe hypoglycemia, but we do not know how this occurs most of the time.
The exciting developments in understanding the mechanisms of pancreatic beta-cell glucose sensing, potassium and calcium ion channel activity, and insulin secretion processes shine brightly on what has been a rather dark and poorly understood pathologic process. The fascinating findings that diabetologists are learning about insulin secretion in adults may provide important clues about how neonatal hypoglycemia can develop and get so out of hand.
Neonatal hyperglycemia also is understood poorly, yet it is encountered commonly and frequently is frustrating. In fact, it is much more common than hypoglycemia in very preterm infants. Just when it is appropriate to increase glucose supply to provide energy and brain nutrient substrate to very small infants, the supply must be restricted in those who have hyperglycemia to prevent or treat excessively high plasma glucose concentrations. Most importantly, but often overlooked, is the strong association of hyperglycemia in the early neonatal period with increased mortality rates and poor neurodevelopmental outcomes among survivors. No one knows how this happens. Is it even the hyperglycemia that is partly at fault or is the hyperglycemia simply a marker for other, more damaging pathologic processes? If the latter is the case, what are those other processes and how are they related to hyperglycemia?
Despite evidence defining many causes of neonatal hyperglycemia, which are well reviewed by Drs Hemachandra and Cowett in this issue, it is difficult or impossible to determine the specific cause(s) in individual cases. How much is due to excessive glucose infusion? How much is due to inadequate insulin secretion, and how does inadequate insulin secretion develop? How much is due to glucose intolerance, insulin resistance, or both, and what are the mechanisms responsible for these pathologic processes? How much is due to excessive glucose production? What are the roles of hormones and other substrates such as amino acids and fatty acids in the regulation of insulin secretion, glucose production, and glucose utilization?
Questions about treatment abound as well. Should glucose supply be restricted, and if so, how and to what rate? Should insulin treatment be used? If so, when and how? What do amino acids, especially leucine, glutamine, and arginine, provide to promote insulin secretion and insulin action? Does early, more aggressive infusion of intravenous amino acid actually decrease the incidence and severity of hyperglycemia in ELBW infants? If so, does it do this by promoting insulin secretion and action or by some other mechanism?
If we knew the answers to the many questions about neonatal hypoglycemia and hyperglycemia, we might be much further along in more specific and beneficial prevention and treatment. We then might be able to determine how aggressive to be in preventing and treating low glucose concentrations, and what the lower limit of plasma glucose concentration should be. We also could establish whether reducing hyperglycemia is beneficial, and if so, how best to accomplish this. Clearly, neonatal hypoglycemia and hyperglycemia are both well understood and yet poorly defined, and in most clinical situations, they are managed without a clue to the specific cause(s) or what will happen following specific treatments. There is much need for further basic and clinical research.
Hopefully this inaugural edition of NeoReviews will initiate the processes that could lead to resolution of such important clinical problems. I have provided case studies of both hypoglycemia and hyperglycemia that include relevant questions for readers to consider. Each case study is followed by the review article written by experts in the field. Future editions will take issue with similarly important problems and often “hot topics.” Basic science underpinnings, clinical management, and future research will be part of each edition. We look forward to a stimulating exchange of facts and ideas from both our authors and our readers.
Lucas A, Morley R, Cole TJ. Adverse neurodevelopmental outcome of moderate neonatal hypoglycemia. Br Med J. 1988;297:1304-1308
- Copyright © 1999 by the American Academy of Pediatrics