Case Based Pediatrics For Medical Students and Residents
Department of Pediatrics, University of Hawaii John A. Burns School of Medicine
Chapter III.8. Neonatal Hypoglycemia
Joel Ruff, MD
February 2002

Return to Table of Contents

This is a newborn infant male delivered to a 25 year old G5P3, A+ mother at 37 weeks gestation by C section (for non-reassuring fetal heart tones). The pregnancy is notable for an antenatal ultrasound diagnosis of cleft lip and palate. Maternal serologies are unremarkable and her prenatal glucose tolerance test is normal. At delivery, blow-by oxygen is given for about 2 minutes for poor color and respiratory effort. Apgar scores are 6 (-2 color, -1 tone, -1 respiratory effort) and 9 (-1 color) at one and five minutes, respectively.

Exam: Vital signs are normal. Oxygen saturation is 99% in room air. Weight is 3950 gms, height is 54 cm and head circumference is 37.5 cm (all >95th percentile for gestational age). The infant is active and jittery with an obvious left sided cleft lip and palate. Heart is regular without murmurs. Lungs are clear. Abdomen is soft, without masses or hepatosplenomegaly. The remainder of the initial exam is normal.

The infant is transferred to the term nursery for transitioning. A bedside glucose is obtained and the blood sugar is read as close to 0 mg/dl. A serum sample is then sent STAT to the lab. An IV is started and 8 ml (2 mg/kg) of 10% dextrose in water (D10W) is given IV. The infant is transferred to the intermediate nursery where a repeat blood sugar 30 minutes after the bolus is still <20 mg/dl. A second IV bolus of D10W is given. The earlier serum glucose sent to the lab, comes back at <2 mg/dl. IV D10 is infusing at 80 ml/kg/day, which is a glucose infusion rate of 5.6 mg/kg/min. A blood sugar obtained 30 minutes after a second bolus is 31 mg/dl. The fluid is changed to 12.5% dextrose (D12.5W), the maximum concentration that can run in a peripheral IV. A third glucose is still below 40 mg/dl and a third bolus is given. The D12.5W infusion is increased to 100 ml/kg/day, which is a glucose infusion rate of almost 9 mg/kg/minute, and the baby is offered some formula. The next blood glucose is 48 mg/dl. The infant requires 100 ml/kg/day of dextrose 12.5% for about 24 hours, and then is able to be slowly weaned off the IV fluids.

Neonatal hypoglycemia is a very common condition. The stated incidence is estimated at 1 to 5 per 1000 births, but it is significantly higher in certain subgroups, 8% in LGA (large for gestational age) infants and about 15% in SGA (small for gestational age) infants (i.e., those with intrauterine growth retardation) (1). Neonatal hypoglycemia can be easily treated in most cases if it is recognized, but untreated hypoglycemia can have serious consequences for the infant as glucose is the major substrate for energy in all organs and almost exclusively used for cerebral metabolism (1).

In the fetus, serum glucose levels are about 70% of those in the mother (i.e., baby's glucose is about 70, when mother's glucose is 100) and almost all of this comes from facilitated diffusion across the placenta. Stores of glycogen in the liver accumulate slowly through gestation with a marked increase during the last trimester.

After delivery, the maternal supply of glucose is interrupted. Fetal glycogen storage is temporarily inactivated and glycogen phosphorylase breaks down hepatic glycogen stores to supply glucose. A term infant is estimated to have only enough hepatic glycogen to support metabolic demands for about 10 hours without an exogenous energy source. At the same time, synthesis of enzymes involved in gluconeogenesis increases and catecholamine levels are high (stimulating the release of substrates in the form of free fatty acids and free amino acids). Blood glucose levels in all infants take a physiologic dip in the first 30 to 60 minutes of life and then increase to a stable level at about 1.5 to 3 hours after birth.

Since Van Creveld recognized that premature infants had lower levels of blood sugar than term infants in 1929 (2) and Hartmann and Jaudon defined groups of "mild," "moderate" and "extreme" hypoglycemia in 1937 (3), the concept of what level of hypoglycemia is physiologically significant has been evolving. Animal studies suggest that hypoglycemia causes brain injury via multiple mechanisms which include excess glutamate, an excitatory amino acid neurotransmitter, free fatty acid release and increased mitochondrial free radicals. Glutamate action on its main receptor, the N-methyl-D-aspartate (NMDA)-type glutamate receptor may transport excess sodium and calcium into neuronal cells and cause selective neuronal necrosis via multiple mechanisms. Although there is an understandable lack of controlled studies in human infant subjects, it is known that the neonatal brain is more tolerant of low blood sugars than the adult brain. Long term effects on cognition and development are difficult to define from a case report literature base. There is a lack of correlation between the glucose level alone and permanent neurodegeneration (4, 5).

Studies in humans, limited to follow up evaluations of infants with hypoglycemia, are complicated by confounding problems such as hypoxia or prematurity, non-uniform definitions of hypoglycemia and lack of control groups. Some studies show adverse neurodevelopmental outcomes and some do not. The article by Cornblath and Ichord (6) is an excellent review of the current state of the literature on this topic. Recently, new imaging modalities including positron emission tomography (PET) and magnetic resonance imaging (MRI) are being used to evaluate function and structure of infant brains affected by hypoglycemia. One study stated that the "prognostic value of these techniques remains still obscure" but the possibilities are exciting (7).

Practically, however, an operational threshold of hypoglycemia for evaluation and treatment needs to be available for the practitioner. It is important to realize that the definition of hypoglycemia may vary from patient to patient. A healthy term infant in no distress may tolerate 30 mg/dl well but a stressed, premature infant may be symptomatic at 50 mg/dl. Symptomatic and very low blood sugars (<20 mg/dl) deserve quick assessment and treatment.

Some infants deserve special attention because of their high risk of hypoglycemia (8,9,10). These include: 1) Infants of diabetic mothers many of whom are LGA, 2) All macrosomic (LGA) infants (mothers may have occult diabetes), 3) SGA, (i.e., intrauterine stressed) infants, 4) Stressed infants (i.e., difficult delivery, low Apgar <7 at five minutes), 5) Infants of mothers on tocolytics (terbutaline, ritodrine), oral hypoglycemics or propranolol within 72 hrs before delivery, 6) Premature infants (<37 weeks gestational age), 7) Postmature infants (>42 weeks gestational age).

Hypoglycemic newborn infants may present with a variety of symptoms or be entirely asymptomatic. Symptoms may include: apnea, jitteriness, exaggerated Moro, irritability, poor sucking or feeding, cyanosis, hypotonia, lethargy or coma, temperature instability, seizure, tachypnea (rare), heart rate abnormalities (slow or fast), abnormal cry or vomiting. Because these signs and symptoms are fairly nonspecific, the differential should include other entities such as sepsis, hypocalcemia or intracranial hemorrhage, as appropriate for the clinical setting (1,8).

The differential diagnosis of hypoglycemia is wide, but a few diagnoses make up the bulk of cases. A convenient way to consider the differential is to divide it into transient hypoglycemia or persistent/recurrent hypoglycemia. The latter is defined as requiring either parenteral glucose for more than 7 days or high IV glucose infusion rates (>12-16 mg/kg/min).

Transient hypoglycemia can be due to maternal or neonatal conditions. Refer to the table below.

1. Maternal conditions causing transient hypoglycemia:
. . . . a. Intrapartum glucose given at too high a rate to the mother.
. . . . b. Drug treatment (terbutaline, ritodrine, propranolol, oral hypoglycemics).
. . . . c. Intrauterine growth retardation (IUGR): placental insufficiency resulting in SGA infant.
. . . . d. Diabetes in pregnancy (hyperinsulinism).

2. Neonatal conditions causing transient hypoglycemia:
. . . . a. Failure to adapt to extrauterine life: "Transient developmental immaturity of critical metabolic pathways" (6).
. . . . b. Birth asphyxia: Hypermetabolism following acute brain injury and/or energy metabolism shifting from aerobic to anaerobic pathways (exact mechanism unknown) (6,10).
. . . . c. Infection: Hypoglycemia due to overwhelming sepsis. Intracranial infection may also be associated with hypoglycemia (6,10).
. . . . d. Hyperviscosity: There is an inverse correlation with hematocrit and blood glucose levels (10) .
. . . . e. Congenital heart disease: May be related to metabolic demands of congestive failure (10).
. . . . f. Erythroblastosis fetalis: Previously reported to be associated with erythroblastosis fetalis from Rh incompatibility, which is uncommon today (10).
. . . . g. Hypothermia: Increased calories expended to thermoregulate (6).
. . . . h. Inadequate provision of calories.
. . . . i. Other iatrogenic causes: Associated with exchange transfusion and/or low lying umbilical artery catheter which selectively perfuses the pancreas (T11-L1) with a high glucose solution resulting in hyperinsulinism (8, 10).
. . . . j. Decreased glycogen stores: Associated with pre and postmature infants (8).

Persistent and/or recurrent hypoglycemia is defined as hypoglycemia for more than 7 days or requiring high levels of support >12-16 mg/kg/min (6,8). You should consider obtaining help from an endocrinologist at this point. Causes include hyperinsulinism, endocrine deficiency, inborn errors of metabolism and neurohypoglycemia (a rare condition in which the subject lacks a transport protein (GLUT1) that facilitates glucose transport across brain microvesicles) (10).

1. Hyperinsulinism conditions which cause persistent/recurrent hypoglycemia:
. . . . a. Nesidioblastosis: A very rare condition in which the pancreas has beta cell hypertrophy. This condition is on a continuum with islet cell adenoma and usually requires subtotal or total pancreatectomy to treat (8,10).
. . . . b. Beckwith-Wiedemann syndrome: Visceromegaly, macroglossia, hypoglycemia.

2. Endocrine deficiency conditions which cause persistent/recurrent hypoglycemia:
. . . . a. Pituitary insufficiency: Associated with syndromes such as septo-optic dysplasia, craniofacial defects and anencephaly (the case patient had a cleft lip and palate which has been associated with patients who have pituitary insufficiency) (10). Deficiency of pituitary hormones such as growth hormone and ACTH, results in hypoglycemia.
. . . . b. Cortisol deficiency/adrenal failure.
. . . . c. Congenital glucagon deficiency.
. . . . d. Epinephrine deficiency: Extremely rare (10).

3. Inborn errors of metabolism conditions which cause persistent/recurrent hypoglycemia (6):
. . . . a. Carbohydrate metabolism: Galactosemia, glycogen storage disease, fructose intolerance.
. . . . b. Amino acid metabolism: Maple syrup urine disease, propionicacidemia, methylmalonic aciduria, tyrosinemia, glutaric acidemia.
. . . . c. Fatty acid metabolism: Carnitine metabolism defect, Acyl-CoA dehydrogenase defect.

Your first evaluation of such an infant should be of airway, breathing and circulation (ABCs). Sepsis may present with hypoglycemia. Most nurseries use a glucose oxidase/peroxidase chromogen test to do bedside determination of blood glucose (Chemstrip or others). This method is quick and fairly easy, but variability in the amount of blood on the strip or residual isopropyl alcohol on the baby's skin may affect readings. Values are given in a range (i.e., 20-40 mg/dl or 40-80 mg/dl). Estimates of sensitivity of this method are 85% but the false positive rate may be as high as 25% (1). The strips are less accurate in the lower ranges (40 mg/dl or less). Because these lower values are of greater clinical concern and because of the variability of reading these strips, these visual methods of estimating blood glucose have been replaced by more precise electronic bedside glucose measurement devices. A stat serum level should be sent to the lab to verify low glucose values. A complete blood count may be a helpful screen for infection and to evaluate the possibility of polycythemia (8). In the lab, plasma or serum is separated from the blood sample and the glucose concentration is measured on the plasma or serum. Thus, a "blood" glucose is really a serum or plasma glucose. The terms blood, serum and plasma glucose can be used interchangeably since they are numerically identical.

The goal of treatment is to establish normoglycemia, usually defined as a stable glucose value above 40 or 50 mg/dl. Definitions vary depending on the clinical situation. All symptomatic infants should be treated with intravenous glucose. Asymptomatic infants in the 20-40 mg/dl range may have a trial of oral feeding, but if the glucose fails to normalize, an IV glucose infusion should be started. Although dextrose 5% (D5W) oral solution is occasionally used for treatment, formula has the advantage of containing fats and proteins which are metabolized slowly and provide a more sustained level of substrates for glucose production (1). Blood glucose should be rechecked in 30-60 minutes after feeding.

For symptomatic infants or asymptomatic infants with severe hypoglycemia (<20 mg/dl), a small IV bolus of dextrose has been shown to raise blood sugar levels safely and more quickly to adequate levels than intravenous infusion alone. The usual bolus is 2 ml/kg of a dextrose 10% (D10W) solution given intravenously followed by a glucose infusion rate of 6 to 8 mg/kg/min. The first part of the infusion, i.e., the bolus is given as a dose of ml/kg, but the second part of the infusion is given in mg/kg/min. This makes it a more difficult conversion because the user must convert grams of glucose to ml, then they must convert ml per minute to ml/hour, since ml/hr is the unit used on IV pumps. A simplified formula for glucose infusion rate (in mg/kg/min) is:

Glucose infusion rate (GIR) = (dextrose%concentration X ml/kg/d) / 144

So if dextrose 10% is used at 80 ml/kg/day that gives us:

GIR = (10 x 80) / 144 = 800 / 144 = 5.6 mg/kg/min

A faster way to figure this out is to use one of the following formulas to achieve a glucose infusion rate of 7 mg/kg/minute.

D5W: IV rate (in ml/hr) = 8.4 X Body Wt (in kg)

D10W: IV rate (in ml/hr) = 4.2 X Body Wt (in kg)

For a 3 kg newborn infant, using D5W would result in an IV rate of 25 ml/hr, which results in 600 ml/day, or 200 cc/kg/day, which is too much. This is why D10W must be used instead. The D10W infusion rate using the above formula would still give 100 cc/kg/day.

The glucose utilization of healthy infants is 5 to 8 mg/kg/min so the above mimics the endogenous requirements. A healthy term infant typically requires only about 60 ml/kg/day of fluids (on the first day of life). In this case, the increased fluids are being used as a vehicle for adequate glucose administration. Giving 15 ml (1/2 ounce) of a standard 20 calorie per ounce formula provides roughly 1.1 gm of carbohydrate (plus protein and fat, which provides additional calories). A 15 ml IV bolus of D5W provides 0.75 grams of glucose, while 15 ml of D10W, provides 1.5 grams of glucose. Formula provides more glucose equivalent than D5W and about as much glucose equivalent as D10W.

What if the follow up glucose (which should be obtained 30-60 minutes after the infusion) is still low, as it was in the patient described above? The options include either increasing the IV rate or increasing the dextrose concentration (i.e., increasing the GIR). The next higher dextrose concentration that is readily available is D12.5W. Because of hypertonicity, peripheral veins cannot tolerate more than 12.5% dextrose. This means that an infant requiring a higher glucose concentration needs a central line. Some infants may require glucose infusion rates as high as 16-20 mg/kg/min, but any infant in this range needs further evaluation and an endocrinology consult.

If the plasma glucose cannot be raised by glucose infusion alone, other options include a trial of corticosteroids (hydrocortisone 5-15 mg/kg/day IV in 2-3 divided doses or prednisone 2 mg/kg/day by mouth). If this fails, other drugs that may be used to raise the plasma glucose include human growth hormone, diazoxide, glucagon or long acting synthetic somatostatin (octreotide) (8).

The first step in evaluating persistent/recurrent hypoglycemia is to obtain serum glucose, insulin, and ketone levels. If the ratio of insulin to glucose (I/G ratio) is >0.3, then the cause is hyperinsulinism. Ketones should be low or absent in hyperinsulinism (4). Ketones are normally generated in hypoglycemic states because the body breaks down fat to acetyl CoA and other ketone bodies, in an effort to generate more substrate for the Krebs cycle. However, in a hyperinsulin state, insulin stimulates lipid synthesis (the opposite of fat breakdown) and thus, ketone levels will be low or absent. Other labs that can be helpful in the diagnosis include growth hormone levels, serum cortisol, free fatty acids, free T4, TSH, uric acid, glucagon, lactate, alanine, amino acids, and somatomedins. Imaging of the pancreas, heart or brain may also be indicated (8).

If the patient is stable, the blood sugar is steady at 50 mg/dl or above, and a more serious condition is not suspected, the frequency of blood glucose measurements can be reduced to every 4 to 6 hours. The intravenous glucose infusion may be weaned after the glucose has been stable and in the normal range for 12-24 hours (1). Some clinicians prefer 2 to 3 days of stable blood sugars before weaning (10). The infusion should be weaned 10-20% every several hours. Enteral feedings may be started concurrently if the infant is otherwise stable and fluid overload is not a concern. Failure to wean should prompt the above evaluation.


1. True/False: The level of hypoglycemia resulting in serious sequelae is well defined by scientific studies.

2. The advantage of using formula over 5% dextrose water (oral) to feed a moderately hypoglycemic term infant is:
. . . . a. More sustained rise in blood sugar.
. . . . b. A much faster rise in blood sugar than with dextrose 5% oral.
. . . . c. Infants less than 3 hours old cannot take formula yet.
. . . . d. One ounce of standard formula is equivalent gm per gm to a 2 ml/kg intravenous bolus of 5% dextrose.

3. When evaluating a hypoglycemic infant, the first thing to assess is:
. . . . a. Ballard exam.
. . . . b. Presence or absence of symptoms.
. . . . c. Airway, breathing, circulation.
. . . . d. Presence or absence of a suck reflex.

4. What is the formula to calculate the glucose infusion rate and at what level should you start?

5. Which of the following infants are at risk for hypoglycemia and should have a screening blood sugar performed in the term nursery? (more than one answer)
. . . . a. Infant of diabetic mother.
. . . . b. A jittery infant.
. . . . c. Small for gestational age infant status post difficult delivery.
. . . . d. 37 week infant born to a GBS positive mother.


1. McGowan JE. Neonatal hypoglycemia. AAP Journals 1999;20:6e-15e.

2. Van Creveld S. Carbohydrate metabolism of premature infants: The blood sugar during fasting. Am J Dis Child 1929;88:912-926

3. Hartmann AF, Jaudon JC. Hypoglycemia. J Pediatr 1937;11(1):1-36.

4. Auer RN, Olsson Y, Siesjo BK. Hypoglycemic brain injury in the rat. Correlation of density of brain damage with the EEG isoelectric time: a quantitative study. Diabetes 1984;33(11):1090-1098.

5. Siesjo BK. Hypoglycemia, brain metabolism and brain damage. Diabetes Metab Rev 1988;4(2):113-144.

6. Cornblath M, Ichord R. Hypoglycemia in the neonate. Semin Perinatol 2000;24(2):136-149.

7. Kinnala A, Korvenranta H, Parkkola R. Newer techniques to study neonatal hypoglycemia. Semin Perinatol 2000;24(2):116-119.

8. Gomella TL. Chapter 43 - Hypoglycemia. In: Gomella TL, Cunningham MD, Fabien GE, et al (eds). Neonatology, fourth edition. 1999, New York: Lange Medical Books, pp.247-251.

9. Hypoglycemia Protocol, Kapiolani Medical Center 1996.

10. Kalhan SC, Saker F. Chapter 47 - Metabolic and Endocrine Disorders. In: Fanaroff AA, Martin RJ (eds). Neonatal/Perinatal Medicine, sixth edition. 1997, St. Louis: Mosby, pp.1443-1461.

11. Schwartz RP. Neonatal hypoglycemia: How low is too low? J Pediatr 1997;131(2):171-173.

12. Cowett RM. Neonatal hypoglycemia: A little goes a long way. J Pediatr 1999;134(4):389-391.

Answers to questions

1. False

2. a

3. c

4. GIR = (dextrose % x ml/kg/d) / 144. Start at 6-8 mg/kg/min and titrate.

5. a, b and c are all correct.

Return to Table of Contents

University of Hawaii Department of Pediatrics Home Page