Controversies regarding definition of neonatal hypoglycemia: suggested operational thresholds.

The definition of clinically significant hypoglycemia remains one of the most confused and contentious issues in contemporary neonatology. In this article, some of the reasons for these contentions are discussed. Pragmatic recommendations for operational thresholds, ie, blood glucose levels at which clinical interventions should be considered, are offered in light of current knowledge to aid health care providers in neonatal medicine. Future areas of research to resolve some of these issues are also presented.

Hypoglycemia in the neonate.

After a brief history of the development of neonatal hypoglycemia, this review emphasizes the current approach to the anticipation, diagnosis, and management of the neonate with a low plasma glucose concentration. Current techniques for studying the neurophysiological and endocrine-metabolic effects of significant hypoglycemia provide new approaches for establishing relevant definitions of significant hypoglycemia, its prognosis, and pathogenesis. The inadequacy of glucose oxidase strips for screening, the definition of high-risk infants, new definitions for low plasma glucose concentrations, and their treatment are presented as well as the ability of the neonate to respond to significantly low glucose values. New data concerning the hereditary aspects of hyperinsulinemia (Glaser, this issue), hereditary defects in branched-chain amino acid, 3-methylglutaconic aciduria and mitochondrial betaoxidation, and degradation of fatty acids (Ozand, this issue), the role of glucose transporters (Vannucci and Vannucci, this issue), and the newer computed tomography and magnetic resonance imaging techniques (Kinnala, this issue) to study neonatal hypoglycemia are reviewed elsewhere in this issue.

An international symposium on neonatal hypoglycemia: thirty years later. Does it injure the neonatal brain?

A discussion of neonatal hypoglycemia was held on 18 November 1995 as a satellite symposium of the 40th Annual Meeting of Japan Society for Premature and Newborn Medicine and continued in closed session of 19 November to address neonatal hypoglycemia in the 21st century. This represented a 30-year follow-up of a discussion of carbohydrate and energy metabolism in the newborn held in Tokyo on 10 November 1965. This follow-up was prompted by the incredible advances in clinical care in perinatal medicine and in basic knowledge in the neurosciences, neonatal physiology and metabolism that have occurred in Japan and around the world throughout these 3 decades.

Neonatal hypoglycemia 30 years later: does it injure the brain? Historical summary and present challenges.

Since 1911, blood sugars have been measured in newborn infants. Significant neonatal hypoglycemia was first reported in 1937. In 1959, the report of transient symptomatic neonatal hypoglycemia generated worldwide reports. This, along with the ongoing advances in studies of energy metabolism, thermal control and oxygen requirements, led to the first conference on Energy and Carbohydrate Metabolism in the newborn in Tokyo, 1965. Subsequently, a number of hypoglycemia syndromes were discovered. Concurrently, pre-, peri- and neonatal care changed dramatically with the survival or very tiny and very sick newborns. These advances in care made previously derived statistical definitions of hypoglycemia irrelevant. New functional definitions are needed to define abnormal glucose concentrations. Significant hypoglycemia is a continuum of low glucose concentrations of varied duration and severity. Its impact depends upon other risk factors as well. In addition, new hypoglycemic syndromes have appeared. These include deficiencies of blood-brain glucose transporters, the association of hyperinsulinemic hypoglycemia with isoimmune thrombocytopenia and a variety of acyl CoA dehydrogenase deficiencies. Concurrently, carbohydrate disorders in infancy appear to be changing. Neonatal diabetes mellitus, previously transient and benign, now shows a high frequency of recurrence and remaining as a permanent condition. Idiopathic ketotic hypoglycemia of infancy has disappeared in the USA. Familial hyperinsulinemic hypoglycemic syndromes of infancy appear to have a good prognosis, respond to medical intervention and have had their genetic defect localized to a specific gene. Current advances promise reliable bedside techniques to measure central nervous system function, cerebral blood flow, endocrine hormones and receptors as well as glucose transporters and specific genetic defects. These data, when correlated with plasma glucose concentrations and central nervous system function and development, should provide a better understanding of the impact of prolonged and profound hypoglycemia on long-term outcome.

Hypoglycemia in the neonate.

After a brief history of the development of neonatal hypoglycemia, this review emphasizes the current approach to the anticipation, diagnosis, and management of the neonate with a low plasma glucose concentration. Both transient and recurrent or persistent hypoglycemia are discussed. Current techniques for studying the neurophysiologic and endocrine-metabolic effects of significant hypoglycemia provide new approaches for establishing relevant definitions of significant hypoglycemia, its prognosis, and pathogenesis. The inadequacy of glucose oxidase strips for screening, the definition of high risk infants, new definitions for low plasma glucose concentrations and their treatment are discussed. New data concerning the hereditary aspects of hyperinsulinemia, the role of glucose transporters, and the ability of the neonate to respond to significantly low glucose values are presented as well.

Familial hyperproinsulinemia due to a proposed defect in conversion of proinsulin to insulin.

Familial hyperproinsulinemia is a genetic disorder characterized by elevated plasma levels of proinsulin-like material. In two previously described kindreds this has been shown to be due to a structural abnormality in the proinsulin molecule. We have identified a third family with hyperproinsulinemia in which there appeared to be a different defect. The propositus, a 12-year-old girl, had borderline glucose intolerance and markedly elevated immunoreactive-insulin levels on oral glucose-tolerance testing. Gel filtration of plasma revealed that 66 per cent of circulating insulin immunoreactivity was accounted for by the proinsulin-like components. Two of four siblings, the father, and the paternal grandfather also had elevated fasting insulin immunoreactivity in the presence of normal plasma glucose concentrations and elevated levels of proinsulin-like material. In vitro tryptic digestion of plasma proinsulin-like material from an affected family member revealed that proinsulin was converted to insulin in a manner indistinguishable from that in the control. Similarly, proinsulin and insulin exhibited normal activity in a radioreceptor assay. These findings suggest that the proinsulin molecule in this family was normal and that hyperproinsulinemia was due to a defect in the conversion of proinsulin to insulin.

Neonatal 'brain damage'-an analysis of 250 claims.

Advances in perinatal care have resulted in decreased neonatal mortality. Increasingly, damage in survivors has been attributed to alleged negligence. We analyzed the 250 claims (1957 to 1982) from one major insurance company for factors to characterize high-risk pregnancies and then to distinguish preventable from nonpreventable causes within the group. Using predetermined criteria, 77 (31%) were classified preventable, 105 (42%) nonpreventable and 68 (27%) indeterminate. Preventable actions could be attributed to family members as well as health care providers. Twenty risk factors were significantly increased in the study group compared with those in a general population and included maternal, gestational, delivery and postdelivery risks. Furthermore, 13 of 25 factors differed significantly between preventable and nonpreventable cases. Those with significantly higher prevalence in preventable cases included prolonged gestation, the use of mid or high forceps, cesarean sections, meconium staining, low one- and five-minute Apgar scores, birth weight exceeding 4.5 kg (10 lb), poor tone, seizures and transfers to neonatal intensive care units. Increased in prevalence in the nonpreventable cases were congenital infections and malformations and the late onset of neurologic abnormalities. These findings suggest preventive measures to reduce unwarranted litigation and certain cases of neonatal brain damage.

Segmental hyperlucent defects in the liver.

When lucent defects in the liver have a segmental configuration, they may be on an ischemic basis and related to decreased vascular perfusion. Portal venous inflow, by virtue of its low pressure, is particularly susceptible to diversion by focal intrahepatic masses, intravenous thrombi, or external compression. Innovative operative techniques for tumor enucleation may also result in lucent defects that can be confused with, or conceal, pathology. A hypothesis relating such defects to diminished portal inflow and reduced glycogen content is proposed.

The use of the plasma epinephrine response in the diagnosis of idiopathic postprandial syndrome.

Patients with idiopathic postprandial syndrome (IPS) report recurrent postprandial episodes that resemble the clinical manifestations of hypoglycemia. In an effort to find objective criteria for diagnosis of IPS, we studied a group of patients with IPS and controls during an oral glucose tolerance test. Patients with IPS had a significantly lower mean glucose nadir and higher hypoglycemic index than controls. Although 74% of patients with IPS had glucose nadirs higher than 50 mg/dL, their responses of epinephrine, cortisol, growth hormone, glucagon, and norepinephrine were significantly higher than the respective changes in the controls. Of the five hormones, only the epinephrine response separated the patients from the controls without overlap. The epinephrine response may represent a valuable diagnostic criterion for this disorder.

Influence of glycogen on liver density: computed tomography from a metabolic perspective.

The liver is a metabolically active organ with a radiographic density that can be modified by its glycogen and fat content. In rhesus monkeys an increase in liver glycogen induced by glucose loading was accompanied by an increase in attenuation values on computed tomography and a decrease in total liver fat. Conversely, fasting depleted glycogen, increased fat, and decreased liver attenuation. Acute glycogen depletion without significant change in fat was induced by administration of glucagon and accompanied by a decrease in attenuation. These results along with in vitro measurements of glycogen solutions suggest that an increase of approximately 3 Hounsfield units can be expected for each percent increase in liver glycogen content.

Insulin receptors in the developing fetal lung.

Fetal lung insulin receptor numbers and affinities were studied in rat pregnancies from 15 to 22 days gestation. Insulin receptor binding capacities were found to increase six-fold from approximately 100 fmoles insulin bound/mg lung DNA at 15 days gestation to approximately 600 fmoles bound/mg DNA at 22 days gestation. However, the affinity constants of the receptors were unchanged during this same period (high affinity, 1.9 +/- 0.4 S.E. and low affinity, 0.03 +/- 0.01 S.E.). The results suggest that the lung may become increasingly more sensitive to insulin as development progresses.

Computed tomography of the liver and kidneys in glycogen storage disease.

Glycogen, in concentrations encountered in von Gierke's disease, has computed tomography (CT) attenuation coefficients in the 50 to 70 Hounsfield unit (HU: 1,000 scale) range and accounts for the increased density of the liver. However, in eight patients with Type I glycogen storage disease, simultaneous hepatic infiltration with fat and glycogen led to a range of liver CT densities from 13 to 80 HU. Fatty infiltration may facilitate the demonstration of hepatic tumors in older patients with this disease. Half the patients showed increased attenuation coefficients of the renal cortex, indicating glycogen deposition in the kidneys.

Increased 125I-insulin receptor binding to erythrocytes of hypoglycemic infants and children.

The measurement of 125I-insulin specific binding to erythrocytes obtained from seven infants and children with various hypoglycemic syndromes showed a significant increase in six patients with recurrent, documented, symptomatic hypoglycemia (percent specific 125I-insulin binding 9 to 14 versus 6.1 +/- 1.4% mean +/- 2 S.D. for 13 controls). The increase was due to an increased number of receptors sites per cell rather than to increased affinity for insulin. The patients included three children with nesidioblastosis, all after 90% pancreatectomy, two with leucine sensitivity, and two with glycogen storage disease type I. One of the patients with leucine sensitivity, who for 2 years before the study had no hypoglycemia, had normal insulin values (less than 10 microunits/ml). Thus, a symptomatic hypoglycemia correlated better with increased 125I-insulin binding than with plasma insulin values. Furthermore, Diazoxide therapy in two patients caused a mild but consistent decrease in the number of insulin receptor sites, and the institution of continuous nocturnal nasogastric feedings in a patient with glycogen storage disease type I was followed by amelioration of the hypoglycemia and a marked increase in 125I-insulin specific binding from 5.2 to 9.5%.