7.2.1 Hypoglycaemia: assessment and management

Hypoglycaemia is defined as a blood glucose level less than 2.6 mmol/l. This is based on the consistent impairment of central nervous system function observed in subjects when blood glucose levels are below this (1). Glucose homeostatic mechanisms should maintain blood glucose level to preserve cognitive function. Hypoglycaemia triggers protective glucose homeostatic mechanisms and persistent hypoglycaemia is the result of a failure of homeostasis. This is a medical emergency with serious short- and long-term consequences, which result from a reduced supply of glucose to the brain. Recurrent and persistent hypoglycaemia does cause significant morbidity and death due to brain damage. In an adult, after recovery of glucose levels, neurological impairment usually recovers over minutes to hours. In children, the duration of hypoglycaemia leading to permanent damage is not known, but is presumed to depend on the age of the child, the frequency of hypoglycaemia, the degree and the rapidity of the fall in glucose, concurrent circumstances such as infection, trauma and hypoxia, the degree of resilience of the brain tissue at the current stage of development. and the energy demands of the particular parts of the brain. The reasons for the increased sensitivity in children appear to relate to the higher energy requirements and immaturity of the homeostatic mechanisms of the brain. In congenital hyperinsulinism of infancy (CHI) the rates of severe neurological impairment remain high at 20–50%, permanent neurological impairment with damage occurring mainly in the cerebral cortex, hippocampus, and caudate putamen. Appropriate long term management of hypoglycaemia requires the correct diagnosis, and this depends on obtaining ‘critical blood and urine samples’ during a hypoglycaemic episode. In the first 48 h of life 20% of normal full–term infants have a blood glucose level <2.6 mmol/l (2), after this it is relatively uncommon in infancy and childhood with the incidence of various underlying diagnoses varying with age. The causes of hypoglycaemia can be classified into five groups:

A high level of suspicion is required in children because vague, nonspecific symptoms and signs of hypoglycaemia may be overlooked. In the neonatal period symptoms/signs of hypoglycaemia are also vague and nonspecific: jitteriness, apnoea, cyanosis, floppiness, and jaundice. In contrast adults develop symptoms and signs in a relatively predictable way (Table 7.2.1.1). The age of onset and duration of fasting before onset of symptoms should be noted. A review of the past history and medical records may suggest a more prolonged duration of hypoglycaemia than first thought (e.g. the diagnosis of idiopathic epilepsy may need to be revised). The child may have been sweaty, shaky, cold, and clammy before breakfast with mood and cognition improving with food. Babies born premature, born small for gestational age, born after pregnancies affected by diabetes of any type, particularly if glycaemic control was suboptimal, born by a traumatic and/or difficult delivery associated with hypoxia, or born with polycythaemia are at risk of hypoglycaemia. There may be a family history of inherited causes of hypoglycaemia or adrenal dysfunction. Hypoglycaemia may be triggered after certain types of food, such as high protein load, high fructose content, toxin of tropical fruit (as seen in Jamaican vomiting sickness after consumption of unripe Ackee fruit) and high glycaemic index foods (which may lead to rebound hypoglycaemia). Certain drugs or chemicals if ingested/administered may cause hypoglycaemia: alcohol, aspirin, oral hypoglycaemic agents, injected insulin, β-blockers, and quinine. The examination should include height, weight, body mass index, and an evaluation of size for gestational age, as well as looking for macrosomia, organomegaly and signs of Beckwith–Wiedemann Syndrome, hyperpigmentation, ambiguous genitalia, cortisol deficiency, hypopituitarism and midline defects, hyperventilation, and dysmorphic features associated with inborn errors of metabolism. The child may also have evidence of damage from previous episodes of hypoglycaemia, such as delayed development, behavioural disorders, hemiplegia, or visual impairment/blindness.

Critical samples should be taken at the time of hypoglycaemia (Table 7.2.1.2) and forwarded to the laboratory immediately on ice. Decisions regarding which analyses are required can be made at a more convenient time. If samples are not taken during the hypoglycaemia, a formal fast (24–72-h study) should be performed in a paediatric investigation unit with a 24-h paediatric laboratory service. Tests should be performed on approximately 1 ml blood in a fluoride oxalate tube, 1.5 ml of blood in lithium heparin, and 2.5 ml of clotted blood in neonatal microcollection tubes to maximize the yield of plasma and serum. A 5 ml urine sample should be collected on first void after the event. If additional blood is available, samples should be taken for urea and electrolytes, liver function tests, C-peptide, ACTH, and transferrin isoforms. Extended neonatal screening programmes available in some countries may identify patients likely to be affected by hypoglycaemia, e.g. medium chain acyl coenzyme A dehydrogenase deficiency (MCAD). Results may not be available, however, before the first episode of hypoglycaemia and false negative results do occur.

The emergency treatment of hypoglycaemia begins with resuscitation for apnoea, unconsciousness, unprotected airway, and generalized convulsions followed by a rapid assessment of the child. Intravenous access is required immediately for critical blood and urine samples to be taken. If the child is fully conscious and cooperative then carbohydrate should be offered in an appropriate form whereas if uncooperative or semi-consciousness a glucose bolus of 0.25 g/kg can be administered intravenously followed by an infusion adjusted to maintain blood glucose at a level higher than 3.6 mmol/l. If venous access is not possible, glucagon can be given either by nasal spray (1.0 mg for children), subcutaneous injection (0.02–0.15 mg for conscious infants to adolescents and 0.3–1.0 mg for unconscious infants to adolescents) or intramuscular injection (0.50–1.0 mg for unconscious infants to adolescents). The child should be re-evaluated once the situation is stable in order to make the diagnosis and a management plan with instructions on the frequency of glucose monitoring, and when and how to intervene. Prior to discharge, the family should also have an emergency plan regarding feeding, the duration of fasting, glucose monitoring frequency, and a sick day action plan.

In some infants increased frequency or volume of feeds, or fortification of feeds is insufficient to maintain plasma glucose levels. A continuous glucose infusion should be started and titrated to maintain plasma glucose levels at a level higher than 3.6 mmol/l. Glucose infusion requirements lower than 10 mg/kg per min suggest the diagnosis is substrate lack or failure of counter-regulatory hormones while requirements greater than 10 mg/kg per min (above normal physiological requirements) suggest increased glucose utilization driven by insulin. In transient conditions <5 days) alternative methods can be used individually or in combination to normalize glucose levels, and allow early reintroduction of oral feeding: glucagon infused at a starting dose 1.0 µg/kg per h; hydrocortisone oral or infused (10–30 mg/m² per day); Octreotide subcutaneously at a starting dose of 5–20 µg/kg per day) (3, 4). If the hypoglycaemia is prolonged >5 days) and due to hyperinsulinism, diazoxide can be given at a starting dose of 5–10 mg/kg per day in divided doses 8 hourly, to a maximum of 20 mg/kg per day together with a thiazide (chlorthiazide 7–10 mg/kg per day in 2 divided doses). Toxic effects of diazoxide include cardiac or cardiopulmonary failure, fluid retention and electrolyte imbalance (5–7). Calcium channel blockers, such as nifedipine, which also suppresses insulin secretion have been successful in the management of a small proportion of patients (8, 9), and trials of glucagon-like peptide 1 (GLP1) receptor antagonists are in progress. When a child is unresponsive to diazoxide a referral should be made to a unit designated for the care of hyperinsulinism. In this setting the options for treatment are conservative management, (long–term subcutaneous infusion of glucagon and/or octreotide and continuous day/night gastrostomy feeding supplemented with cornstarch (4)) or surgery (subtotal or focal pancreatic resection).

The plasma glucose should be confirmed by reliable laboratory methods as the result may be affected by factors independent of the ‘true’ glucose level of the patient:

When making the diagnosis, the following factors should be considered: the clinical circumstances, such as the age of onset of hypoglycaemia and the duration of fasting; the pattern of results obtained; the response of the patient to the therapeutic intervention. Central to the diagnostic assessment is the presence or absence of ketonuria/ketonaemia (Fig. 7.2.1.1). In the first 2 days of life, coexistence of hyperinsulinism and high ketones can occur due to the presence of ketones of maternal origin associated with hyperketosis of labour (Table 7.2.1.2).

Low or undetectable levels of ketones and free fatty acid suggest suppression by insulin or more rarely a defect in fatty acid metabolism. In the presence of hypoglycaemia, insulin secretion should be suppressed and, if detectable (1.0–3.0 mU/l) then the likely diagnosis is hyperinsulinism. Hypoglycaemic neonates with hyperinsulinism may fail to generate an adequate serum cortisol counter–regulatory response making the diagnosis difficult to resolve (10). A diagnosis of hyperinsulinism will be supported by a metabolic and endocrine profile of low levels of ketones, β-hydroxybutyrate, free fatty acids, and IGFBP-1 together with high proinsulin and C–peptide levels and high intravenous glucose infusion rates 10 mg/kg per min). The biochemical methods used in the analysis of insulin should be reliable and have a lower detection limit of 0.1 mU/l (0.6 pmol/l).

High levels of ketones and free fatty acids, together with low or undetectable insulin levels suggest the diagnosis of either an abnormality in counter-regulatory hormones or an abnormality in glucose release. If the counter-regulatory hormone levels are low (growth hormone <10 mU/l and/or cortisol <500 nmol/l) then the diagnosis is either hypopituitarism (both low) or adrenal failure (cortisol low and elevated ACTH levels). If the counter-regulatory hormone response is appropriate, the diagnosis is likely to be idiopathic ketotic hypoglycaemia once the diagnosis of inborn errors of metabolism has been excluded. A child repeatedly presenting with ‘ketotic hypoglycaemia’ should be re-evaluated.

When performing a fasting study the child should be admitted, and remain in hospital until fully recovered with normal postprandial blood glucose levels. The duration of the fast (6–72 h) should be determined by the physician in charge and the fast should be terminated either because of hypoglycaemia or at the completion of the planned duration. The critical blood and urine samples are taken even if no hypoglycaemic event takes place to provide clues to the diagnosis.

At the completion of the fast or during an episode of hypoglycaemia a glucagon stimulation test can be performed to determine the extent of glycogen available for release of glucose. Glucagon 0.03 mg/kg is administered intravenously or intramuscularly, and the change in glucose levels is monitored from before (0 min) and 10, 20, and 30 min after the injection. A positive response, considered to be an increase in glucose levels of at least 1 mmol/l suggests hyperinsulinism.

An oral glucose tolerance test (OGTT) (1.75 g/kg to a maximum of 75 g) will demonstrate the lactate and insulin response to a glucose load. In glycogen synthase deficiency (glycogen storage disease GSD type 0) there is an exaggerated rise in lactate. Rebound hypoglycaemia may occur after gastric bypass surgery due to failure to suppress insulin secretion.

A standard protein load (1.0–1.5 g/kg of protein, given as an amino acid hydrolysate drink with no carbohydrate) after a period of fasting may induce hyperinsulinaemic hypoglycaemia in subjects affected by glutamate dehydrogenase deficiency or protein sensitivity (leucine or glutamate) (11). Plasma glucose, ammonia and insulin levels are measured over 180 min, or until glucose levels fall to less than 2.6 mmol/l. Protein sensitivity will trigger an abnormal response with a fall in glucose, a rise in insulin levels and in glutamate dehydrogenase deficiency there will be a rise in plasma ammonia levels.

Some of the causes of hypoglycaemia including defects in insulin secretion: glucokinase, glutamate dehydrogenase, SUR, KIR, MODY 1; adrenal insufficiency: 21-hydroxylase, DAX1, triple A syndrome; and aldolase B have been sufficiently characterized to consider routine genetic analysis as part of the work-up.

The fetus is provided with a steady glucose supply from the mother to allow growth and development. If nutrient supply is limited, growth is restricted and the baby is born small for gestational age (SGA). The fetus does not express the key enzymes in the gluconeogenic pathway. At birth, the continuous supply of nutrients ceases and the fall of plasma glucose during the first 4 h of life stimulates a counter regulatory hormone response as part of the homeostatic response. The transition to a fed/fasted cycle is accomplished with little consequence in the normal term infant, but in the premature or SGA infant this transition is often compromised. The term baby adapts to the fed/fasted cycle with induction of the glycogenolytic and gluconeogenic pathways and changes in the regulation of insulin secretion. Plasma glucose concentration is regulated by homeostatic mechanisms that balance glucose production and glucose utilization involving interaction between plasma glucose, insulin, and the counter–regulatory hormones. Insulin decreases glucose production and increases glucose utilization. Glucagon stimulates glucose release from liver glycogen. Cortisol and growth hormone play permissive roles in setting the sensitivity of the peripheral tissues to glucagon and insulin. Circulating glucose is available from three main sources: from food ingestion and digestion of carbohydrate, from the breakdown of glycogen (glycogenolysis), and from de novo manufacture from amino acid or fat (gluconeogenesis). Both glycogenolysis and gluconeogenesis result in the production of glucose-6-phosphate, which is hydrolysed by glucose-6-phosphatase in the liver to glucose, which can then enter the circulation. During fasting insulin secretion decreases and counter-regulatory hormones increase (Table 7.2.1.3). Glycogenolysis occurs as a result of the actions of several enzymes: glycogen phosphorylase, phosphoglucomutase, glycogen debrancher enzyme. The process is controlled by the activities of glycogen synthase and phosphorylase, and insulin and glucagon are the major hormones controlling these enzymes. As fasting proceeds, glucagon secretion with reduced insulin allows stored fats to be converted to glycerol, and fatty acids and proteins to be converted to amino acids. Gluconeogenesis involves the synthesis of glucose from lactate, alanine, glutamine, glycerol, and pyruvate. The process uses a number of key enzymes: pyruvate carboxylase, phosphoenolpyruvate carboxykinase, and fructose 1,6 bisphosphatase. Pyruvate carboxylase is regulated by the mitochondrial acetyl CoA and ADP concentrations, and is induced by alterations in plasma insulin, glucagon, and cortisol levels during the postnatal starvation after birth. The liberated free fatty acids are transported to the liver to undergo β oxidation to yield ketones. Muscle and other tissues become progressively more dependent on free fatty acids and ketone bodies for energy requirements. Healthy infants from 1 week to 1 year of age can usually tolerate 15–18 h of fasting increasing to 24 h between 1 and 5 years. The basal rate of glucose output by the liver is precisely matched to tissue uptake. Children will develop hypoglycaemia after a relatively short time of 36 h;, in contrast, adults can survive without food for a number of weeks due to the glucose sparing effect of ketones, and free fatty acids that allow the limited capacity of gluconeogenesis to provide glucose for key glucose-dependent tissues (brain, red blood cells, and renal tubules). Glucose uptake by the insulin-independent tissues, such as the brain and splanchnic organs accounts for 80% of total body glucose utilization under fasting conditions, mainly by the brain (50% of the total). Muscle, an insulin-dependent tissue, is responsible for most of the remaining glucose utilization in the fasting state. As fasting progresses tissue glucose utilization decreases, while utilization of free fatty acids and ketone bodies increases. Children have higher glucose production rates in comparison with adults in order to meet the increased metabolic demands. Brain size is the principal determinant of factors that regulate hepatic glucose output throughout life. Glucose requirements change as the child grows and the relative weight of brain mass to body mass changes (Table 7.2.1.4).

The transport of glucose into tissues is by facilitated diffusion and depends on the specific glucose transporters:

In the embryo, both exocrine cell types and islets are derived from a common pool of precursor endodermal cells derived from the dorsal and ventral portions of the embryonic mid-gut. The dorsal and ventral primordia fuse by 7 weeks gestation with the ventral area forming the inferior and posterior parts of the head of the pancreas, and the dorsal area forming the remainder of the pancreas. Islets comprise 20% of the pancreatic tissue in newborns, 7.5% in children (1 ½–11 years) and 1% in adults. Endocrine cells are present from 9–10 weeks gestation, and islet formation commences by 13 weeks. Islet formation continues during the neonatal period when both the fetal and adult type islets are observed (12). In the adult, islets consist of: β cells (48–59% of the islet cell population), α cells (33–46%), and δ cells (8–12%). There are a large number of factors that have been demonstrated to regulate pancreatic and islet development. In the β cell, glucose causes a dose-dependent release of insulin and C-peptide, predominantly through calcium-dependent exocytosis of preformed storage granules, as well as stimulating insulin biosynthesis and up-regulating the rate of transcription of insulin mRNA by cAMP, partly through phosphorylation of PDX1, the homoeodomain protein that binds to regulatory elements of the insulin gene promoter. There is a basal release of insulin, which accounts for approximately 50% of insulin secreted by the pancreas in a 24-h period with the remainder secreted in response to meals. There is also a small amount of ‘unregulated’ or constitutive release of insulin (1–2% of the total). Glucose enters the β cell via the GLUT1 and GLUT2, is metabolized by glucokinase and the glycolytic and oxidative phosphorylation pathways. This increases the intracellular concentration of ATP, changing the ATP/ADP ratio, triggering closure of the KATP channel. This channel consists of an octameric complex constituted by four sulphonylurea (SUR1) proteins that surround four Kir6.2 proteins, through which the potassium ions are moved across the cell membrane (Fig. 7.2.1.2). The cessation of this efflux by closure of the KATP channel results in depolarization of the β–cell membrane. This, in turn, leads to the influx of calcium through voltage–gated channels, which triggers insulin exocytosis (Fig, 7.2.1.3) (13). Drugs interact with the insulin secretory pathway. Sulphonylurea drugs interact with the KATP channel to enhance insulin release, diazoxide via a similar process inhibits insulin release. Drugs like quinine may also increase insulin release via an unknown process.

In neonates hypoglycaemia is a relatively common, although short-lived (<4 h), adjustment disorder to ex utero life. It is classified as transient when recurring over the first 5–10-day period of life. The various diagnoses made in those in whom hypoglycaemia lasted more than 4 h is shown (Table 7.2.1.5). The estimated rate of hypoglycaemia warranting investigation was 5 per 1000 deliveries.

Hyperinsulinism of infancy (HI), which covers a spectrum of conditions, is the commonest cause of severe, recurrent hypoglycaemia at this time of life (14). Hyperinsulinism is characterized by excessive and inappropriate secretion of insulin in relation to the prevailing blood glucose concentration.

The cause of the majority of transient HI in neonates is unknown, but is known to be associated with a mother with diabetes mellitus (IDM), intrauterine growth retardation, perinatal asphyxia, erythroblastosis fetalis, Beckwith–Wiedemann syndrome (BWS), and maternal administration of some drugs (e.g. sulfonylureas). Infants with transient HI due to IDM often present with macrosomia, selective organomegaly (liver, heart; skeletal length), and congenital anomalies including anencephaly, meningomyelocele, holoprosencephaly, sacral agenesis, small left colon syndrome, and a number of structural abnormalities of the heart. Some infants with transient HI due to IDM have more prolonged and severe hypoglycaemia, but all regain normal blood glucose control within 10 days of birth. BWS, that is a congenital overgrowth syndrome associated with organomegaly, hemihypertrophy, omphalocele, ear lobe anomalies, renal tract abnormalities, and an increased risk of embryonal tumours (liver and kidney), is caused by dysregulation of imprinted growth regulatory genes within the 11p15 region. The incidence of hyperinsulinaemic hypoglycaemia in children with BWS is about 50% (15).

Whatever the cause of transient HI there are a relatively large number of infants with this diagnosis requiring intensive monitoring and a number require diazoxide treatment, which can be reduced and stopped after the first few months of life.

Congenital hyperinsulinism of infancy should be considered if the hypoglycaemia is severe, recurrent, and persists longer than 10 days. Estimates of the incidence of CHI vary from 1 per 40 000 live births to 1 in 2500 live births in communities with high levels of consanguinity (16, 17). The age at presentation varies from the newborn period or during the first few months of life after birth (18). CHI associated with mutations in ABCC8 or KCNJ11 genes, which code for the inwardly rectifying potassium channel (KATP channel), is the most severe form. Reduced or absent function of the KATP channel results in unregulated insulin release, even in the presence of low glucose concentration. Partially functional channels may be associated with milder disease (19). A number of genetic causes for CHI have been identified (Table 7.2.1.6), although no cause is found in up to 50% of patients. Glucose levels are extremely labile and the primary aim of early management is to protect the brain with the aim to maintain levels higher than 3.6 mmol/l. CHI can be classified into diffuse and focal disease. Attempts to diagnose focal disease are important because surgical resection of the focal area is curative and preoperative detection is desirable. Positron emission tomography (PET) using fluro-dopa with CT or MRI scan is a relatively noninvasive, accurate method to detect focal cases including rare ectopic focal lesions (20–22). The pathogenesis of focal and diffuse disease differs. Focal disease is caused by a focal clonal expansion of β cells lacking the maternal 11p15.1 allele due to loss of maternal chromosome 11p material. When these cells have a paternal mutation in the ABCC8 gene, the result is a clone of β cells unable to express the KATP channel and, hence, this clone has unregulated insulin secretion and as this increases in size hypoglycaemia occurs. Diffuse disease is usually caused by the presence of two mutations in ABCC8 or (rarely) KCNJ11, resulting in an abnormal insulin secretion throughout the pancreas. Dominant mutations and de novo mutations have been reported. Diffuse disease is classically characterized by the presence of enlarged islet cell nuclei with abnormal architecture throughout the pancreas. However, ‘atypical’ histology has been described, with departures from normal in terms of islet and acinar architecture, and no evidence of enlarged islet cell nuclei. Diffuse CHI that cannot be controlled adequately by medical means requires surgical resection of the pancreas. The operation most commonly performed is a 95% pancreatectomy. Some children still remain hypoglycaemic postoperation, and may require ongoing diazoxide and/or octreotide, or may require a second or even third operation perhaps because of regeneration of the pancreatic remnant (23). Most infants who have undergone pancreatectomy develop diabetes mostly in the peripubertal period. Older age at surgery, greater extent of resection, and previous pancreatectomy increase the risk of insulin dependence immediately following pancreatectomy (24, 25). Nonpancreatectomized infants with CH HI due to ABCC8 mutations have also been reported to develop impaired glucose tolerance and diabetes in childhood or young adulthood (26), although reports of long-term follow-up from other centres, where no medically treated patients have developed diabetes, suggest that the risk in nonpancreatectomized patients is low (4).

Infants with HI/HA have an activating mutation in the gene encoding the enzyme glutamate dehydrogenase leading to increased intracellular concentrations of ATP, which trigger insulin secretion, with a mild, persistent hyperammonaemia. Most mutations occur de novo, but families with dominant modes of transmission have been reported. Hypoglycaemia due to inappropriate insulin concentration may occur with fasting or postprandially. The hypoglycaemia may be intermittent and milder than in infants with ABCC8 or KCNJ11 mutations. It responds well to diazoxide. Lifelong therapy is usually required (27).

Mutations resulting in abnormal activation of glucokinase increase intracellular concentrations of ATP and activate insulin secretion. They have been reported in a small number of families and are transmitted in an autosomal dominant pattern. The mutations result in increased affinity of glucokinase for glucose and inappropriate insulin secretion. The hypoglycaemia is relatively mild, of variable severity within families, and may be controlled with food or diazoxide if necessary (28, 29).

Hyperinsulinism can be caused by a mutation resulting in reduced activity of the HADH gene encoding the enzyme short chain 3-hydroxy-acyl-CoA dehydrogenase (SCHAD) involved in mitochondrial fatty acid metabolism. Mutations are inherited in an autosomal recessive mode. The disease varies from mild to severe forms. Patients present in the neonatal or infant period and the hypoglycaemia is diazoxide responsive (30, 31). The presence of 3-hydroxyglutaric acid in urine and raised plasma levels of 3-hydroxybutyryl-carnitine in plasma may aid the diagnostic evaluation (31).

Congenital disorders of glycosylation (CDG) are a large family of genetic diseases that result from defects in glycan metabolism, associated with a wide variety of different clinical presentations, such as hypoglycaemia, neurological impairment, gastrointestinal problems, hypertrophic cardiomyopathy, seizures, short stature and dysmorphism, and some have a characteristic physical or biochemical profile. Hyperinsulinaemic hypoglycaemia has been reported as the presenting feature of CDG in the absence of other signs. Abnormal serum transferrin isoforms are a screening test for CDG, but should be delayed until the infant is over 1 month of age as the test is unreliable before this. A positive transferrin test should be followed by enzyme assay in cultured fibroblasts to confirm the type of CDG, which will guide prognosis and clinical management. CDG associated HI is responsive to diazoxide (32, 33).

Mutations in the gene encoding the hepatocyte nuclear factor 4α (HNF4α) cause maturity onset diabetes of the young type 1 (MODY1). Infants with these mutations may present with severe hypoglycaemia and macrosomia. The mode of inheritance is autosomal dominant. The hypoglycaemia may be transient or persistent, and is responsive to diazoxide. The natural history of the disease is hyperinsulinaemic hypoglycaemia in infancy, progressing to impaired insulin secretion, and diabetes in adolescence or young adulthood (33).

Usher Syndrome is due to mutations in the chromosomal region contiguous with ABCC8 and KCNJ11, with affected infants also presenting with sensorineural deafness and retinopathy (34, 35). Mutations in the INSR gene have been associated with HI, but only reported in adults (36). Similarly, exercise-induced hypoglycaemia associated with abnormalities of the monocarboxylate pathway have been reported, but only in adults to date (37). HI can be caused by surreptitious administration of sulphonylurea drugs (causing a rise in endogenous insulin secretion, together with high levels of C-peptide) or insulin injection that will be associated with inappropriately low levels of C-peptide. HI presenting for the first time in children over 5 years of age may be due to an insulinoma and, if proven, then this is likely to be associated with the overall diagnosis of multiple endocrine neoplasia type 1 (MEN 1). Genetic studies of the MEN1 gene (OMIM 131100) should be performed in the child and family. Nonislet cell tumour hypoglycaemia is caused by abnormal insulin-like activity due to the production of preproIGF-2 by the tumour (38). Hirata’s disease, a syndrome seen in Japan (where it is the third most common cause of hypoglycaemia in adults) is characterized by persistent hypoglycaemia associated with high insulin antibody levels, low ketone levels, low free fatty acid levels, and low insulin levels.

Deficiency of the counter–regulatory hormones in particular growth hormone and cortisol may cause hypoglycaemia. In combination in disorders of pituitary function, such as congenital hypopituitarism as high as 20% of patients present with hypoglycaemia together with other features of panhypopituitarism. Standard replacement doses of hydrocortisone and growth hormone prevent further hypoglycaemia (dosage of hydrocortisone will need to be increased during ‘stress’).

Adrenal disorders can cause hypoglycaemia with the patient presenting with increased pigmentation, low cortisol levels, and high ACTH levels. Congenital adrenal hypoplasia can be due to an autosomal recessive form, an X-linked (OMIM 300200) form, mutations in the dosage-sensitive sex reversal-adrenal hypoplasia gene 1 (DAX1) (OMIM 300473), ACTH resistance caused by familial glucocorticoid deficiency (OMIM 607397), and triple A (AAA) syndrome (OMIM 231550), which is associated with achalasia, alacrima, and autonomic neuropathy. Deficiency of steroidogenic factor 1 (SF1) (OMIM 184757) can lead to adrenal failure with complete XY sex reversal due to testicular dysgenesis. Congenital adrenal hyperplasia also often presents with hypoglycaemia due to 21-hydroxylase deficiency (OMIM 201901) and in the male infant the first clue to the condition may be a presentation with collapse in the first 1–8 weeks of life with hypoglycaemia, hypotension, and hyperkalaemia. Secondary adrenal failure may be due to autoimmune disease (with adrenal autoantibodies detected), adrenoleukodystrophy (OMIM 300100) in males, or adrenal destruction due to haemorrhage or ischaemia.

Defects in the breakdown of hepatic glycogen cause hypoglycaemia with hepatomegaly. (Table 7.2.1.7). Glucose-6-phosphatase deficiency (OMIM 232200) is the commonest of the glycogen storage diseases causing hypoglycaemia. The two other glycogen storage diseases causing hypoglycaemia result from deficiencies of the enzymes amylo1,6 glucosidase (OMIM 232400) and liver phosphorylase (OMIM 232700). Deficiency of liver phosporylase kinase (OMIM 306000), which is required to activate liver phosphorylase results in variable hypoglycaemia and is inherited in an X-linked manner.

Glycogen synthase (OMIM 240600) plays an important role in the storage of glycogen in the liver, and deficiency of it is a rare cause of hypoglycaemia in childhood. Mutations in the hepatic isomer of glycogen synthase that result in an inability to form α1,4-linkages between glucose molecules to form glycogen are associated with fasting hypoglycaemia and postprandial hyperglycaemia together with elevated lactate and triglyceride levels.

Patients with deficiencies of each of the four unique enzymes of the gluconeogenic pathway that ensure a unidirectional flux from pyruvate to glucose [pyruvate carboxylase (OMIM 608786), phosphoenolpyruvate carboxykinase (PEPCK) (OMIM 261650), fructose 1,6 bisphosphatase (OMIM 229700), and glucose-6-phosphatase] present with fasting hypoglycaemia and lactic acidosis.

The commonest disorder of fatty acid β oxidation is medium chain acyl CoA dehydrogenase (MCAD) (OMIM 201450), which may be severe and even fatal in young patients, is autosomal recessive and is characterized by recurrent episodes of hypoglycaemic coma, impaired ketogenesis, a characteristic urine organic acid profile, increased octanylcarnitine on the plasma acylcarnitine profile, and low plasma and tissue carnitine levels (30). The condition is managed by avoidance of fasting, dietary manipulation, and carnitine therapy.

Other disorders of carnitine metabolism and fatty acid oxidation may similarly restrict ketone body synthesis, depriving the body of this alternate fuel source (Table 7.2.1.7). The disorders result in hypoglycaemia associated with low ketone levels despite high plasma free fatty acids. Plasma acylcarnitines and urine organic acids can help elucidate the diagnosis.

Ketone bodies are synthesized from the combination of acetyl CoA and acetoacetyl CoA by liver mitochondrial HMG CoA synthase to form hydroxymethylglutaryl CoA (HMG CoA). This is split by HMG CoA lyase to yield acetoacetate in the liver, which is then converted to B hydroxybutyrate. In the peripheral tissues, acetoacetate is activated back to acetoacetyl CoA by succinyl CoA:3 oxoacid CoA transferase (SCOT). Hypoglycaemia may occur as a result of defects in either the synthesis or the utilization of ketone bodies.

Hypoglycaemia can occur as a result of a number of metabolic conditions: galactosaemia a deficiency of galactose-1-phosphate uridylyltransferase (OMIM 606999); hereditary fructose intolerance (OMIM 229600), caused by catalytic deficiency of aldolase B (fructose 1,6 bisphosphonate aldolase), present after taking foods containing fructose or sucrose; and organic acidaemias including methylmalonic aciduria (OMIM 251000), propionic acidaemia (OMIM 606054), glutaric aciduria type 2 (OMIM 231680), maple syrup urine disease (OMIM 248600), tyrosinaemia (OMIM 276700), and in mitochondrial respiratory chain defects. Hypoglycaemia may be triggered after certain types of food, such as high protein load, high fructose content, the toxin in unripe ackee fruit, and high glycaemia index foods. There are a number of drugs and chemicals that if ingested/administered may, via interruption of the intermediate metabolism, lead to episodes of hypoglycaemia: alcohol, aspirin, oral hypoglycaemic agents, insulin injection, B-blockers, and quinine.

Idiopathic ketotic hypoglycaemia is common, although the pathogenesis is not understood, and is a diagnosis of exclusion (39) as the differential diagnosis includes inborn errors of metabolism. There is an association with low birth weight, poor weight gain, and male gender. The age of presentation is 18 months to 5 years, usually resolving by 8–9 years of age (when the brain to bodyweight ratio is decreasing and endogenous substrate availability is increasing). Ketotic hypoglycaemia may also be seen in various situations of lack of substrate and increased metabolic demands in which there is an obvious cause, such as severe illness, sepsis, malaria, and liver disease. Teenage girls prone to eating disorders may also present in this way particularly after ingestion of alcohol.