John A. Widness, MD
Peer Review Status: Internally Peer Reviewed

Jump to: Hypoglycemia | Hyperglycemia | Normoglycemia | Glucose Infusion Rates Table



Plasma glucose < 40 mg/dL in both term or preterm infants.


The definition of neonatal hypoglycemia has been based on statistical criteria.2 The incidence of this condition in term AGA infants is approximately 2%.

Infants at risk immediately following birth:

IDMs, IGDMs (especially those whose mothers received oral hypoglycemic agents), LGA (>90%ile), SGA (IUGR <10%ile), post-asphyxiated, APGAR < 5 at five minutes, polycythemic, immune hemolytic disease, suspected sepsis, hypothermia (rectal temperature <35¾C), congenital anomalies, Beckwith-Wiedman syndrome, infants ≤ 36 wks gestation, infants ≥42 wks gestation, & those whose mothers received large amounts of i.v. glucose prior to delivery.


Non-specific, including tremulousness, twitching, jitteriness, irritability, exaggerated Moro reflex, high pitched cry, seizures, apnea, limpness, poor feeding, cyanosis, temperature instability, and coma.

Screening of infants at risk (see table at end of this section):

Screen by plasma glucose measurements at 1, 2, 4, 8 and 24 hours of age if not receiving glucose containing i.v. fluids, or when symptomatic.


By plasma glucose measurements only. For diagnostic purposes it is imperative that plasma glucose be measured in the NICU or Hospital laboratory by quantative chemical analysis and NOT by a semi-quantitative reagent strip method (e.g., Chemstrip bG®). Although use of reagent strips is widespread, their use, even for glucose screening purposes, is controversial1. This is because of their inaccuracy and imprecision relative to accepted laboratory determined glucose values. Treatment decisions in infants without signs ("asymptomatic") of hypoglycemia should not be based on Chemstrip data.


Asymptomatic Infants
Severity Plasma Glucose Treatment objective is to raise the plasma glucose concentration to 50-200 mg/dL
Mild <40 mg/dL Oral: If developmentally capable of feeding orally, and if clinical condition permits (i.e., no significant respiratory distress, etc.), immediately offer ad lib oral glucose solution (D5W), infant formula, or allow breast-feeding. If enteral feeding is not possible, treat parenterally (see next).
Severe <20 mg/dL Parenteral: Treat with i.v. 4-8 mg glucose/kg/min using D10W (see Figure at end) as for "symptomatic" infants.

Symptomatic infants:

Immediate bolus: 0.20 grams of glucose/kg, i.e., 2 ml/kg of i.v. 10% glucose, given over 1-2 minutes. (Do not use D25W or D50W.)

Continuous infusion: 6-8 mg glucose/kg/min using D10W (or D5W); this is equivalent to 90-120 ml/kg-day as 10% dextrose in water--see Figure at end of section.

Guidelines for monitoring plasma glucose in hypoglycemic infants

In hypoglycemic infants, plasma glucose values should be repeated 1-2 hours after initiation of treatment if the infant remains asymptomatic, or every 20-30 minutes if symptomatic (see table at end of this section). If signs of hypoglycemia persist or recur, or if plasma plasma glucose concentration as determined by the neonatal or hospital laboratory remains below 40 mg/dL, increase glucose infusion rate to 10 to 12 mg/kg/min. If further hypoglycemia persists, IMMEDIATELY notify staff neonatologist for consideration of further treatment and diagnosis.

After the plasma glucose has been normal for 24 hours, and enteral feedings have been started, taper i.v. glucose infusion rate by 1-2 mg/kg/min every 4 to 8 hours as tolerated. (This may necessitate a reduction in the concentration of dextrose in the i.v. fluid being used.) Intravenous glucose intake usually cannot be reduced below 4-6 mg/kg/min until enteral feedings are begun.



Unknown and controversial, although the neonatal staff has suggested a definition of plasma glucose as >200 mg/dL.2 Hyperglycemia almost always occurs in the first hours to days of life.



Infants at risk:

Infants with birth weights > 1 kg receiving intravenous fluids at high dextrose infusion rates >8mg/kg/min); VLBW infants < 1 kg at moderate glucose infusion rates (4-8 mg/kg/min); infants with congenital diabetes mellitus (rare); infants receiving, or whose mothers received, selected drugs, e.g. diazoxide.


Polyuria due to glycosuria (rare if blood glucose < 250 mg/dL); intracranial hemorrhage if hyperglycemia occurs rapidly as a result of an abrupt increase in plasma glucose concentration, e.g., following an i.v. D25W or D50W glucose bolus. As noted above, use of either of these high concentration glucose solutions is to be avoided.


Same as hypoglycemia, i.e., plasma glucose measurement is needed--NOT Chemstrip.


If plasma glucose > 200 mg/dL.

Reduce glucose infusion rate.

IV insulin administration (0.1 U/kg/hr) if reducing the rate of glucose infusion is not effective, or is not possible. This treatment should not be undertaken without first consulting the staff neonatologist.

Guidelines for monitoring plasma glucose in hyperglycemic infants:

Plasma glucose measurements should be determined every 1 to 4 hours depending on the degree of hyperglycemia, with therapy adjusted according to plasma plasma glucose results. Glucose determinations should be done on capillary blood from an extremity, or from a non-glucose-containing indwelling catheter.



Plasma glucose 50-100 mg/dL in infants not receiving i.v. fluids.

Guidelines for monitoring plasma and urine glucose in infants receiving I.V. fluids:

Blood glucose monitoring (see table at end of this section):

Monitoring of plasma glucose concentrations in sick and premature infants receiving parenteral fluids is indicated. When clinically feasible, try to do this at times when other blood work is being done. This will reduce the number of heelsticks.

Urine should be tested by the nursing staff once a shift for glucosuria. Some infants, especially those <1,000g, may not be able to fully utilize high glucose infusion rates, and will spill glucose in their urine. Although glucose-induced osmotic diuresis is rare in neonates, it should be looked for once glucosuria is present. The presence of glucosuria is an indication to perform a chemical determination of plasma glucose for the purpose of defining an individual infant's renal glucose threshold. If glycosuria is present and the plasma glucose is <200 mg/dL, it is NOT necessary to decrease the glucose infusion rate.

In the management of extremely small infants weighing <700 g, it may be advisable to perform less frequent plasma glucose monitoring (once daily), and to rely more on the results of urine glucose testing. In these situations, if plasma glucose testing documents the absence of hypoglycemia, urine glucose testing may be substituted for blood glucose testing for the purpose of avoiding hyperglycemia. Infants whose blood glucose values are within the normal range, and whose urine glucoses are negative, may have their glucose infusion rate increased to increase energy intake if deemed appropriate.

Newly admitted infants without risk factors for hypoglycemia should have a plasma glucose value measured between 1 and 4 hours after birth and again between 8-16 hr unless clinical circumstances mandate earlier glucose testing.

Subsequently, infants at high risk who are receiving only i.v. fluids, should have plasma glucose determinations done every shift for 3 days and once a day for days 4-7 after birth. In infants not at high risk, consider monitoring plasma glucose levels once or twice a day in the first week of life.

In infants who have been on i.v. fluids for over 1 week, and whose plasma glucose values have been within the normal range monitoring of plasma plasma glucose, may be decreased to twice a week.
Use of blood glucose screening using semi-quantitative reagent strips, i.e. Chemstrips, is to be discouraged.2 In addition, do not use blood from arterial lines for glucose determination.

Infants whose plasma glucose values are < 200 mg/dL may have their infusion rate of glucose (and lipid) increased to provide additional energy if their caloric intake is deemed suboptimal (< 90 kcal/kg/dL). It is prudent to increase glucose infusion rate slowly (1-3 mg/kg/min) with change made once/day. For infants receiving long-term i.v. fluids, the upper limit of the glucose infusion rate used may be defined by the infant's plasma glucose values.

An infant whose glucose has been stable on plain IV fluids may show hyperglycemia with the addition of lipids. Conversely, the addition of protein (NVN) may require an increase in the role of dextrose administration.



Cornblath M, Schwartz R, Aynsley-Green A, Lloyd JK. Hypoglycemia in infancy: the need for a rational definition. Pediatrics 1990;85:834-837.

Cornblath M, Schwartz R. Disorders of Carbohydrate Metabolism in Infancy. 3rd ed. Philadelphia: W.B. Saunders, 1991, pp. 87-124, 225-246.

Glucose Infusion Rates

This graph may be used in your management of neonates as an aid for determining:

  1. the i.v. rate needed to achieve a desired glucose infusion rate, i.e., in mg/kg/min as is needed for writing orders; or
  2. determining the glucose infusion rate of an existing i.v. to determine an infant's caloric intake.

As an example, a 2.5 kg infant whom you would like to have receive 6 mg/kg/min of glucose should be receiving 9.5 cc/hr of D10W (equivalent to 90 cc/kg of i.v. fluid).