Abstract
Glucose homeostasis is the foundation of energy supply and maintenance during the transition period of fetal to newborn life. Despite this important position, our understanding of its vital role and the threshold between physiologic requirements and the turn to pathologic injury remains incomplete. Requirements for the developing fetus are largely met by placental transfer from the mother; yet with the increasing frequency of premature delivery, even late preterm birth, before 36 weeks, the newborn must become self-reliant. It is often this transition period, when it occurs during the developmental trajectory of the fetus, and complications around gestation and the birth process that make the definition and treatment of glucose homeostasis a difficult moving target. In this chapter, we try to clarify some of these issues. The newborn is most definitely resilient and is able to use alternative substrates, particularly lactate and building its glucose reserves, as glycogen, in the last 4 weeks of gestation. Nonetheless, metabolic stresses such as seizures and/or asphyxia tip the scales of supply and demand, requiring additional substrate to maintain energy. It is at these times, or when the duration of reduced glucose concentrations become prolonged after birth, that the clinicians’ threshold for treating neonatal hypoglycemia must be lowered.
Keywords
brain injury, glucose, hypoglycemia, neonatal
- •
Neonatal hypoglycemia remains a common though frequently overlooked complication of the newborn.
- •
In the term, otherwise normal newborn, brain injury requires prolonged episodes of glucose values below 1.1 mmol/L.
- •
High levels of sensitivity for hypoglycemia must be maintained in newborns with complex histories of diabetic mothers, growth restriction, fetal inflammation, or hypoxia-ischemia.
- •
In infants with prolonged or intractable hypoglycemia, insulin-secreting tumors, pituitary abnormalities, or metabolic disorders must be sought.
Hypoglycemia remains a common though controversial problem of the newborn infant. Such controversy persists around issues related in the first place, to definition and subsequent diagnosis, the relevance of “asymptomatic” versus “symptomatic” hypoglycemia, incidence rates, underlying pathophysiology, treatment, and of course, neurodevelopmental outcome. Importantly, contributing to the persistence of neonatal hypoglycemia as a cause of morbidity in the newborn is the increasing prevalence of both type 1 and type 2 diabetes worldwide and diagnosis in younger women of childbearing age. Confounding these issues are improved obstetric and neonatal intensive care, which have allowed for the survival of low-birth-weight infants, with their attendant complications of prematurity and lack of nutritional stores, respiratory distress, altered metabolism, and higher risks of disorders such as hypoxia-ischemia, seizures, and sepsis.
The adult acts as a completely independent being with respect to nutritional requirements. The fetus, on the other hand, is fully dependent on the mother and the placental transfer of glucose and its other nutritional requirements. The newborn exists in the transition phase between these two states of complete dependence and independence. For normal cerebral development and consequent function to proceed, an adequate amount of metabolizable substrate must be supplied to the brain during the perinatal period. Glucose is the primary energy substrate for both the adult and newborn brain under physiologic conditions. However, other organic substrates are capable of supplementing glucose during conditions whereby the normal balance of supply and demand for energy production are superseded.
At birth, the previously consistent supply of maternal glucose is abruptly terminated. Immediately after birth, hepatic glycogen stores are broken down to maintain adequate nutritional support. Glucose-6-phosphatase is the rate-limiting enzyme for this to occur; it is expressed at low levels in the newborn, increasing to adult values within the first few days of life. To rapidly adapt, an endocrine stress response involving insulin and glucagon drives hepatic glycogenolysis, lipolysis, and fatty acid oxidation that generate lactate and ketone bodies as alternative fuels important in maintaining cerebral energy metabolism. Estimated rates of glucose metabolism in the 1-day-old newborns are threefold greater than older newborns and infants. Moreover, measured rates of glucose oxidation suggest that only ∼70% of the energy needs of the brain are met through the metabolism of glucose. Hence, the newborn is adapted for using ketone bodies, which can be 5- to 40-fold greater than the adult, and lactate, which contributes significantly in the first few hours of life.
Despite the obvious importance of glucose for cerebral energy utilization, particularly during the complex transition from fetal to newborn life, questions remain regarding the role of hypoglycemia per se to brain damage and neurodevelopmental outcome. It is, therefore, the intent of this chapter to provide the reader with a general review of glucose metabolism and its alternate substrates to the newborn brain and to describe the recognized derangements associated with hypoglycemia. Finally, we review clinical aspects of hypoglycemia and present case examples that exemplify neonatal hypoglycemia, highlighting questions and controversies around this still complex issue.
Glucose Metabolism in the Fetus and Newborn
In most species studied, including humans, glucose serves as the primary organic fuel for energy production under physiologic circumstances. In the fetus, a linear relationship has been observed between the glucose level in the mother and that of the fetus. At term birth, blood glucose concentrations in the newborn are at about 80% to 90% of those of the mother. This linear relationship has been seen during all states of maternal euglycemia, hyperglycemia, and hypoglycemia and is important in that it implies at least a one-to-one relationship between maternal and fetal glucose needs.
At the time of birth, glucose concentrations in the term healthy newborn fall within the first hour of life, then recover and become more stable by 3 hours of age, and gradually increase for at least the first 96 hours, when infants receive exogenous nutrition. In preparation for birth, a doubling of the glycogen stores occurs from 36 weeks’ gestation to term. At birth, plasma insulin levels fall, together with a marked surge in glucagon levels, leading to a mobilization of glycogen stores, which are rapidly depleted within the first 12 to 24 hours of life. Glucagon levels remain elevated through the first week of life. Subsequent glucose concentrations in the normal newborn depend on feeding practices. Although some studies have suggested feeding intervals to be a major determinant of blood glucose concentrations, others have not found this to be the case. Irrespective, “low” blood glucose concentrations in appropriately fed normal term infants are very rare.
Preterm Infants
It is a generally held belief that blood glucose concentrations in the preterm infant are lower than those of the term infant. Although recent studies suggest that this is not likely the case, given more recent policies of early feeding and intravenous glucose supplementation, the theoretical risks certainly apply. In this regard, the preterm infant has not as yet had the opportunity afforded the term infant to build up glycogen stores, typically occurring in the last 4 weeks of gestation. Moreover, the rate-limiting enzyme for glyconeogenesis is significantly lower in the preterm compared with the term infant ; hence the ability to break down even these limited stores of glycogen are restricted. The capability of the preterm to mount a response with alternate substrates may also be impaired. Hawdon et al. compared 156 term infants with 62 preterm infants and found that although blood glucose concentrations were not statistically different, preterm infants were unable to mount a significant ketone body response at the lower end of the blood glucose values. Others have found preterm infants to variably mount an inadequate glycemic response to glucagons, suggesting features of insulin resistance.
Intrauterine Growth Restriction
Clearly, intrauterine growth restriction (IUGR), occurring as the fetus is exposed to an environment when nutrition is restricted either because of placental insufficiency or maternal lack of nutrition, predisposes the newborn to hypoglycemia. Previous reports have indicated a higher prevalence of hypoglycemia in the small infant compared with those within the normal weight range. Other reports are more controversial; some suggest similar glucose concentrations in small-for-gestation age (SGA) versus appropriate-for-gestation age (AGA) infants, again likely the result of more aggressive nutritional management. Still others continue to show differences between the two weight groups with SGA infants displaying lower glucose concentrations compared with AGA infants. In either case, infants with IUGR display altered metabolic profiles that include reduced glycogen stores, limited oxidation of free fatty acids, and functional hyperinsulinism. Combined with a relatively larger brain size, one can see the predisposition of these infants to neurologic injury from a hypoglycemic insult.
Cerebral Metabolism of Glucose
The ontogeny of regional changes in cerebral glucose utilization (CGU) has important implications regarding the sensitivity of the immature brain to hypoglycemia. Using the 2-deoxyglucose (2-DG) technique, regional cerebral glucose utilization (rCGU) in the perinatal animal has been shown to be high in brainstem gray matter structures, declining in a caudal to rostral fashion toward the cerebral cortex. Using positron emission tomography (PET) scanning with fluorine 18–2-DG ([ 18 F]–2-DG) as the isotope, Chugani et al. and others measured rCGU in humans from birth through adulthood. In infants 5 weeks of age, rCGU was highest in the sensorimotor cortex, thalamus, midbrain-brainstem, and cerebellar vermis. By 3 months of age, maximal glucose utilization had shifted to the parietal, temporal, and occipital cortices and in the basal ganglia, with subsequent increases in frontal and various association regions of the cerebral cortex occurring by 8 months of age. Little further change in rCGU was observed between 8 and 18 months, with adult values reached by 2 years.
Alternate Substrates to Glucose
The perinatal brain is capable of incorporating and metabolizing alternate substrates, most notably lactic acid and the ketone bodies, β-hydroxybutyrate and acetoacetate. In vitro studies of regional energy status and the availability of alternate substrates in rats have shown lactate concentrations to be elevated sixfold in newborn brain compared with the adult brain, and the β-hydroxybutyrate level to be double that of the adult, mature brain.
With respect to the latter, both animal and human studies of the newborn have shown an enhanced capacity for the cerebral extraction of ketone bodies from blood compared with older infants and adults. The investigation of ketone body utilization in suckling rats suggests they may account for between 20% and 35% of cerebral energy metabolism in this age group. Ketone body utilization peaks at postnatal day (PD) 14, and subsequently diminishes by PD21, at a time during which CGU is increasing and glucose becomes the major substrate for energy metabolism. These findings coincide with the capacity of the immature blood-brain barrier to transport ketone bodies at threefold greater rate compared with glucose. Furthermore, those enzymes linked to ketone body metabolism in the brain display a rapid increase in activity after birth and a subsequent decline after weaning, in contrast to the pattern displayed by the key enzymes of glycolysis, whose activity increases with advancing age in an inverse relation to those of ketogenesis.
Although ketone bodies appear to play a role in normal energy metabolism of the immature brain, whether they do under pathophysiologic circumstances of glucose deprivation seems unlikely. Data from human infants suggest that the capacity for hepatic ketone synthesis in the neonate is restricted. The findings demonstrate (1) low blood ketone levels, (2) a failure of ketone bodies to rise with fasting, and (3) a failure of ketone bodies to rise with hypoglycemia. In contrast, lactic acid has been shown to be an important source of energy during hypoglycemia. Elegant studies in the newborn dog during normoglycemia show that 95% of cerebral energy requirements are met by glucose with ketone bodies and lactate contributing 1% and 4%, respectively. With insulin-induced hypoglycemia and a concomitant reduction in CGU, lactate was able to support 58% of cerebral oxidative metabolism. Subsequent experiments showed that, under these conditions, there was no significant decline in brain high-energy phosphate levels.
Recent findings from 35 newborns, at risk and hypoglycemic, show that the infants had glucose, ketone bodies, and lactate concentrations measured in their blood within the first 48 hours of birth. In this study, Harris et al. (2015) found that with glucose concentrations in the severe to moderate hypoglycemic range, β-hydroxybutyrate levels remained low (2.03 mean mmol/L), whereas lactate concentrations generally increased (3.06 mean mmol/L). The authors suggested that under circumstances of hypoglycemia, lactate may serve as a better fuel for neuroprotection than ketone bodies. Other investigators have also shown a preferential utilization of lactate over either glucose or ketone bodies in the newborn rat and dog and a sparing effect on glucose utilization during hypoglycemia.
Glucose Transporters
The mechanism by which glucose is transported from blood into brain across cell membranes occurs by a Na + -glucose cotransporter protein that is energy independent. These facilitative glucose transporter proteins are a family of structurally related proteins. Twelve glucose transporters (GLUTs) have been identified and labeled as GLUT1 through GLUT12. Within the brain, GLUT1 and GLUT3 are predominant. GLUT1 is the most prevalent of the glucose transporters and is highly expressed in all blood-tissue barriers, including the blood-brain barrier. GLUT3 is the predominant isoform in neurons. GLUT5 has been detected in the microglia of both humans and rats.
The expression of glucose transporter proteins, not surprisingly, reflects the energy demands of the brain. Hence, analysis of cerebral cortical microvessels and membranes in the newborn rat demonstrates that all GLUT proteins are low during the first week of life. During the second and third postnatal weeks, GLUT proteins increase, particularly in the deep gray matter structures of the thalamus and hypothalamus, coincident with enhanced utilization of glucose as a fuel. Similarly, and in an almost linear fashion, GLUT proteins in the cortex and hippocampus increase from 20% to 100% of adult values between 7 and 30 PDs during a recognized period of rapid neuronal maturation and synaptogenesis.
Definitions
Ambiguity surrounding a precise definition of “neonatal hypoglycemia” continues and has been emphasized by a number of reviews in the recent past. Koh et al., as far back as 1988, surveyed 36 pediatric textbooks and 178 pediatric consultants searching for agreement on the definition of neonatal hypoglycemia. Perhaps not surprisingly, there was none, with definitions ranging from less than 1 mmol/L to less than 4 mmol/L. In 1937, Hartmann and Jaudon published a series of 286 neonates and infants with “significant hypoglycemia” as determined by recurrent or persistent low “true” blood glucose values. Only those infants with clinical manifestations were considered. These authors defined hypoglycemia as follows: mild (2.2–2.78 mmol/L), moderate (1.11–2.22 mmol/L), or extreme (<1.11 mmol/L). Their approach incorporated the important concept that the definition of hypoglycemia must represent a continuum of values that deviate from the biologic norm. This latter concept is particularly relevant today as definitions of “treatable” hypoglycemia take into account gestational age, multisystem organ complications, and neurophysiologic and/or clinical symptomatology.
Difficulty in arriving at an absolute value for hypoglycemia in the newborn stems from the obvious factors that encompass a dynamic and vulnerable biologic process. Hypoglycemia simply refers to an abnormally low blood glucose concentration . In this context, the definition of “abnormal” becomes relevant, given that hypoglycemia—or euglycemia, for that matter—is an evolving, dynamic process, itself dependent on a large number of variables.
Absolute glucose concentrations below which the term hypoglycemia can be applied have been defined based on statistical measures (within two standard deviations of the mean). Hence, serial plasma glucose determinations in term healthy newborn infants revealed an initial drop to 55 to 60 mg/dL (3.05 mmol/L) within the first 2 hours of life, followed by a rise to 70 mg/dL (3.88 mmol/L) from 3 to 72 hours, and levels in excess of 80 mg/dL (4.44 mmol/L) beyond the third day. Values below the fifth percentile were therefore considered by these authors as representing statistical hypoglycemia.
Lubchenco and Bard studied the incidence of hypoglycemia as determined by gestational age and birth weight. Their work showed that preterm AGA infants played a mean glucose concentration of 48 mg/dL (2.6 mmol/L) compared with 54 mg/dL (3.0 mmol/L) in the term AGA infants. In the SGA infants born at term, there was a further shift to the left with mean glucose concentrations of 44 mg/dL (2.4 mmol/L).
In more recent studies of glucose concentrations in healthy term infants, Hoseth et al. studied 223 term breast-fed newborns serially over the first 96 hours and found the lowest blood glucose values occurred within the first hours of life, with an overall range of 1.4 to 5.3 mmol/L (25.6–97 mg/dL; median 3.1 mmol/L, 57 mg/dL). Similar results were reported by a study of more than 200 term healthy newborns in whom a mean glucose concentration of 2.8 mmol/L (51 mg/dL) was found. In both of these studies, 12% to 14% of the children had blood glucose concentrations less than 2.6 mmol/L (47 mg/dL ), mostly during the first day of life.
A meta-analysis analyzing 10 studies, inclusive of 723 healthy term AGA infants, suggested parameters that are less than the fifth percentile of norm for the definition of neonatal hypoglycemia. In this regard, thresholds for hypoglycemia would be on a sliding scale based on time after birth and include values less than 1.6, 2.2, and 2.67 mmol/L (<29, 40, and 49 mg/dL) at 1 to 2, 3 to 47, and 48 to 72 hours of age, respectively.
Based on the preceding findings, it is reasonable to state that normal glucose concentrations in term healthy infants have a wide range, with the lowest concentrations occurring during the first few hours of life. Within this range, the risk of neurologic sequelae is remote, and routine testing for blood glucose concentrations has been suggested to be unnecessary.
Controversy and Question
These data do not, however, direct themselves to the more controversial and clinically relevant questions that remain somewhat unanswered. Hence, the definition of neonatal hypoglycemia remains nonspecific and is dependent on gestational age, appropriateness of fetal growth, the age of the newborn at the time of sampling, and whether or not the infant has fed. Given these parameters, current data suggest that hypoglycemia is not clinically evident, nor perhaps relevant, until values are less than 1.1 mmol/L (<20 mg/dL). We therefore need to ask the following question:
- 1.
Are there other parameters or markers of hypoglycemia that suggest an association with resultant encephalopathy?
Symptomatic Versus Asymptomatic Hypoglycemia
Most common among the features of hypoglycemic encephalopathy is an alteration in the level of consciousness, described as lethargy or somnolence. Irritability, high-pitched cry, or exaggerated primitive reflexes may also be found. Newborns are often described as being jittery, and this may progress to seizures, apnea, hypotonia, and coma.
In this regard, a number of investigators have shown that perhaps the more relevant way to define hypoglycemia is to do so based on whether or not the infant is symptomatic. Koivisto et al. reported on 151 children, divided into three groups described as follows: the (1) symptomatic-convulsive group (n = 8), (2) symptomatic-nonconvulsive group (n = 77), and (3) asymptomatic group (n = 66). In this group of patients, feeding was not initiated for the first 24 hours of life. Symptoms were characterized by the presence of tremor, cyanosis, pallor, limpness, irritability, apathy, or tachypnea, which disappeared with glucose therapy. Hypoglycemia was defined as a glucose concentration less than 20 mg/dL (1.1 mmol/L). The findings indicated that 50% of the symptomatic convulsive group and 12% of the symptomatic nonconvulsive group had neurologic abnormalities on follow-up compared with only 6% of both the asymptomatic and control groups. In a study by Singh et al., 107 infants with severe hypoglycemia (<25 mg/dL) were evaluated over 15 months. Symptoms were present in 40%. Neurodevelopment in asymptomatic infants was normal.
Moore and Perlman described three cases of profound hypoglycemia in term breast-fed newborns in whom seizures developed following discharge from hospital. All were symptomatic with pallor, jitteriness, poor feeding, but had nevertheless been sent home on early discharge. All of the patients showed glucose concentrations less than 1.1 mmol/L (<20 mg/dL). Late follow-up suggested that two of the three infants had normal findings, and one was significantly delayed.
Alkalay et al. reviewed reports of hypoglycemia over the past four decades. Their criterion for inclusion, albeit retrospective, was the presence of neurologic sequelae, considered to be directly or primarily the result of hypoglycemia. The study was inclusive of both AGA and SGA infants as well as preterm infants. Their findings indicated that, of the study patients reported, more than 95% had plasma glucose concentrations less than 25 mg/dL (1.4 mmol/L). The incidence in this group with neurologic abnormality was 21%.
To correlate a critical threshold of blood glucose concentration with neurologic dysfunction, several studies have evaluated neurophysiologic parameters in association with hypoglycemia. Koh et al. reported abnormalities in sensory evoked potentials in children when blood glucose concentrations fell below 2.6 mmol/L (<47 mg/dL). A glucose concentration of less than 18 mg/dL (<1.0 mmol/L) was associated with an isoelectric electroencephalogram (EEG), suggesting the potential for injury. Unfortunately, only four of these children were younger than 1 month of age. Cowett et al., who studied term and preterm infants, found no such correlation, and Pryds et al. also found no correlation between hypoglycemic glucose concentrations and brain auditory evoked response (BAER) and EEG patterns.
In a very recent study by Caksen et al. the authors examined 110 infants with hypoglycemia using magnetic resonance imaging (MRI). There seemed to be no difference in glucose concentrations between the symptomatic versus asymptomatic patients, all of whom had mean values less than 1.0 mmol/L (<18 mg/dL). However, the symptomatic infants were more likely to have abnormal MRI findings compared with the asymptomatic infants.
Duration of Hypoglycemia
Alkalay et al. found that the minimal age at which hypoglycemia was detected was 10 hours, suggesting that a prolonged period of hypoglycemia was required before neurologic sequelae or symptomatology become evident. Others have similarly suggested prolonged hypoglycemia as a prerequisite for damage. Lucas et al. determined the neurologic outcome of 661 preterm infants. Moderate hypoglycemia, defined in their study as less than 2.6 mmol/L (<47 mg/dL), occurred in 433 infants, of whom 104 displayed recurrent events on 3 or more separate days. A strong correlation existed between the number of separate days in which hypoglycemia was recorded and reduced mental and motor development scores at 18 months corrected age. When hypoglycemia was present on 5 or more days, the incidence of cerebral palsy or developmental delay was increased by a factor of 3.5.
A more recent study conducted by Duvanel et al. illustrated similar results. Eighty-five SGA preterm newborns were tested for hypoglycemia (defined as <2.6 mmol/L; <47 mg/dL). In their cohort, 73% met the criteria for hypoglycemia, and recurrent episodes were once again strongly correlated with persistent neurodevelopmental and physical growth deficits to 5 years of age.
A primate study determining the effect of prolonged insulin-induced hypoglycemia on outcome also showed that the longer the duration of hypoglycemia, the greater the degree of abnormal behavioral outcome. However, even in those in whom hypoglycemia was produced for 10 hours, the effects were transient and reversible when training was done for the behavioral task. Blood glucose concentrations were less than 25 mg/dL. Unfortunately, no neuropathologic examination was reported for this group of animals.
In our own laboratory, we have recently completed a study in newborn rat pups undergoing insulin-induced hypoglycemia to various degrees and for variable durations. Neuropathologic assessment was determined following a series of behavioral tests. In this experiment, the mortality rate was significantly increased among rat pups that achieved profound levels of hypoglycemia, below 1.0 mmol/L (<18 mg/dL). In survivors of prolonged hypoglycemia (>12 hours), behavioral assessments were no different than controls. However, brain pathology indicated a significant increase in cell death among those achieving severe hypoglycemia in specific nuclei of the thalamus ( Fig. 9.1 ). Moreover, these pathologic alterations were accompanied by alterations in excitatory amino acid release and free radical production (unpublished data, 2008).
In a more mature model of recurrent moderate hypoglycemia, rat pups received 5 U/kg of insulin twice daily from PD10 to PD19. Although glucose levels were not recorded, those pups receiving insulin displayed heightened levels of anxiety during juvenile age equivalent and diminished social play behavior as adolescents. Though it is not a model of neonatal hypoglycemia, the study confirms the need for a prolonged period of hypoglycemia to obtain an abnormal phenotype.
To provide basic guidelines by which to define hypoglycemia, Cornblath issued a consensus statement regarding “operational thresholds” for blood glucose concentrations. These researchers defined operational threshold as that concentration of plasma or blood glucose at which clinicians should consider intervention. In that regard, they felt that term, healthy, asymptomatic full-term infants need not have routine monitoring of their glucose concentrations. On the other hand, any infant with clinical manifestations compatible with hypoglycemia should be tested, and intervention should be taken for those with values less than 45 mg/dL (2.5 mmol/L). For infants at risk of hypoglycemia owing to alterations in maternal metabolism, intrinsic neonatal problems, or endocrine or metabolic disturbances, glucose monitoring should begin as soon after birth as possible. For values less than 36 mg/dL (2.0 mmol/L), close surveillance should be maintained and intervention is recommended if concentrations remain low, regardless of the presence or absence of symptoms. In those infants in whom very low concentrations are detected (<20–25 mg/dL; 1.1–1.4 mmol/L), therapeutic intervention should be initiated immediately. Newborns being fed by continuous parenteral nutrition will have persistently high insulin levels. As a result, their ability to manifest significant ketogenesis and the means by which to use alternate substrates will be impaired. Under these circumstances, prudent caution suggests maintaining glucose concentrations in the higher therapeutic ranges (>45 mg/dL; 2.5 mmol/L).
More recent recommendations from the American Academy of Pediatrics (AAP) and the Pediatric Endocrine Society (PES) essentially substantiate Cornblath’s early work, though controversy continues to persist ( Box 9.1 ).
Hyperinsulinism
- •
Beta-cell hyperplasia
- •
Nesidioblastosis
- •
Macrosomia
- •
Beckwith-Weidmann syndrome
Endocrine Abnormalities
- •
Panhypopituitarism
- •
Hypothyroidism
- •
Growth hormone deficiency
- •
Cortisol deficiency
Hereditary Metabolic Disorders
- •
Abnormalities of carbohydrate metabolism
- •
Amino acid disorders (maple syrup urine disease)
- •
Organic acid disorders
- •
Fatty acid oxidation defects
Glucose Transporter Defects
While this consensus certainly addresses many of the concerns regarding the circumstances under which clinicians should be vigilant in their approach to the diagnosis and treatment of hypoglycemia, it does not answer several other important questions regarding the newborn in particular. Before attempting to answer these questions, further understanding of the epidemiology and pathophysiology of hypoglycemia would be in order.
Causes of Hypoglycemia
Although it is not the intent of this chapter to discuss the underlying causes of hypoglycemia, a list of etiologies is found in Box 9.1 .
Incidence
The reported incidence of hypoglycemia depends on those variables also specific to the definition. Sexson et al. found 8.1% of 232 infants had glucose values less than 30 mg/dL (1.6 mmol/L) in the first hours of life, with 20.6% having glucose concentrations less than 40 mg/dL (2.2 mmol/L). When a value less than 30 mg/dL was used as the definition, Lubchenco et al. studied the incidence of hypoglycemia according to gestational age and weight and found the overall incidence to be 32% for SGA infants, and 10%, and 11.5% for AGA and LGA infants, respectively. With a definition of less than 20 mg/dL, the incidence obviously decreases. Of the 73% of SGA infants studied by Duvanel et al. who had hypoglycemia, 30% would have had 6 or more episodes with glucose values between 1.6 and 2.6 mmol/L.
Pathophysiology of Hypoglycemia
Cerebral Blood Flow, Glucose Utilization, and Cerebral Energy Metabolism
As the major metabolic fuel for cerebral energy production is glucose, there is an inextricable link between the demands for energy production and the supply and extraction of substrate. In that regard, studies in newborn dogs displayed an inverse linear relationship between blood glucose concentrations and cerebral blood flow (CBF). Therefore increases in CBF ranging from 150% to 450% of normal occurred as blood glucose concentrations decreased from 40 mg/dL to less than 5 mg/dL (2.2 to <0.3 mmol/L). Increases followed ontogeny and were more predominant in brainstem structures compared with other major regions of the brain. The same phenomenon is seen in human infants. Pryds et al. found CBF increased by 200% above normal at levels of blood glucose below 30 mg/dL (1.6 mmol/L).
Vannucci’s group of collaborators found similar results and, importantly, looked at the alterations in white matter glucose utilization during hypoglycemia. In this study blood glucose concentrations were reduced to ∼1.0 mmol/L. Hypoglycemia was associated with increases in regional CBF (rCBF) ranging from 170% (white matter) to 250% (thalamus). In both of the former studies, there was a direct relation between CBF and mean arterial pressure. Regional CGU, was unchanged in 11 of the 16 structures measured, but significantly, was reduced by 30% to 45% in the occipital white matter structures and cerebellum. Calculations of the extent to which glucose transport into the brain during hypoglycemia was enhanced by the increases in CBF suggested that glucose delivery contributed minimally (<10%) to the maintenance of CGU. Earlier studies had also shown that with hypoglycemia to levels as low as less than 1.0 mmol/L, the cerebral metabolic rate for oxygen (CMRO 2 ) decreased to 50% of normal. At this same level, cerebral metabolic rates for lactate increase 10-fold and became the dominant fuel for oxidative metabolism in the newborn dog brain. Later experiments showed that using a similar experimental paradigm of hypoglycemia, high-energy phosphate reserves (phosphocreatine, adenosine triphosphate [ATP]) remained within normal concentrations. From this group of studies, the authors concluded that CBF autoregulation is lost during hypoglycemia in the newborn, and that rather than glucose delivery, low energy demands serve to maintain glucose homeostasis and preclude tissue glucose deficiencies. They further hypothesized that alternate cerebral energy fuels, predominantly in the form of lactate, substitute for glucose at levels of blood glucose below 1 mmol/L.
Cerebral Biochemical Alterations During Hypoglycemia
Little work has been done on the biochemical perturbations that arise as a result of hypoglycemia in the immature brain. Much of what is known in this regard is derived from experiments done in the adult animal exposed to insulin-induced hypoglycemia. However, given the information discussed in the previous section, important comparisons between the adult and newborn brain can be made, and perhaps some tentative conclusions drawn.
As in the newborn, CBF increases during hypoglycemia. In the adult, this is where the similarities end. Hence in adult models of hypoglycemia, cerebral high-energy phosphate levels (ATP, phosphocreatine) plummet to levels less than 20% of normal as blood glucose concentrations fall below 1 mmol/L (<20 mg/dL). In concert with this depletion of high-energy reserves, neurophysiologic monitoring reveals an isoelectric EEG, a marked increase in intracellular Ca 2+ , and a 10-fold and 4-fold rise in extracellular concentrations of the excitatory amino acids (EAAs) aspartate and glutamate, respectively. Hence, at least in the adult, it now appears that the mechanism of neuronal death as a result of hypoglycemia is similar to that of hypoxia-ischemia (energy depletion → EEA → Ca 2+ influx → free radical production → cell death).
As indicated earlier, the preservation of cerebral energy status in the newborn, even at very low concentrations of glucose, is accompanied by a preservation of neurophysiologic function as demonstrated by EEG. A possible underlying cause of cellular injury in the newborn may, however, be related to the release of EEAs. Silverstein et al. induced hypoglycemia by insulin injection in 7-day-old immature rats. Blood glucose concentrations gradually diminished over time to 30 to 40 mg/dL in the first hour after injection, 20 mg/dL in the second hour, and to less than 5 mg/dL by 3 to 4 hours after injection. Their results indicated a direct correlation between decreasing glucose levels and increasing concentrations of extracellular glutamate. In hypoglycemic newborn human infants, Aral et al. also reported on their finding of increased concentrations of glutamate and aspartate in the cerebrospinal fluid.
In more recent studies, neonatal rats of 7 days of age were made hypoglycemic, and antioxidant factors and cell death factors were measured in the cerebral cortices at 1 month of age. Anju et al. found that levels of superoxide dismutase and glutathione peroxidase, both powerful endogenous antioxidants, were significantly reduced in the hypoglycemic groups compared with controls. Apoptotic cell death, as measured by Bax, caspase-3 and caspase-8, were activated and increased in the cerebral cortices of these rat pups as well. Glucose concentrations in the rats were maintained at less than 40 mg/dL. The authors point to this experiment as indicating the role of oxidative stress in brain injury following hypoglycemia and the role of reduced glucose concentrations in inhibiting neuronal antioxidant capacity.
Beyond the previous experimental data, there is very little information regarding the underlying mechanisms of hypoglycemic brain injury in the newborn. In comparing the newborn response to hypoglycemia with that of the adult ( Table 9.1 ), several important differences become evident, particularly in relation to the controversy surrounding hypoglycemic brain injury. In this regard, then, the newborn appears to respond to hypoglycemia with physiologic alterations that appear to protect the brain from damage. Hence, CBF increases, glucose utilization is reduced, alternate substrates are able to substitute for demands of energy production, and as a result, cerebral energy reserves are preserved, at least in experimental animals. These metabolic adaptations beg the question as to whether or not hypoglycemia per se does cause brain damage, certainly an area of ongoing controversy.
Related posts:
Cerebral Circulation and Hypotension in the Premature Infant: Diagnosis and Treatment
Hypothermia for Neonatal Hypoxic-Ischemic Encephalopathy: Different Cooling Regimens and Infants Not Included in Prior Trials
General Supportive Management of the Term Infant With Neonatal Encephalopathy Following Intrapartum Hypoxia-Ischemia
Neonatal Meningitis: Current Treatment Options
Hyperbilirubinemia and the Risk for Brain Injury
Amplitude-Integrated EEG and Its Potential Role in Augmenting Management Within the NICU
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
