Home › Clinical Guideline Hubs › Section 4: Endocrinology › Chapter 4.5
Chapter 4.5 · Section 4: Endocrinology

Persistent Hypoglycemia

Beyond 48–72 hours · Diagnostic triggers, critical sample protocol (plasma glucose <50 mg/dL), etiology table, bedside clues by mechanism, monitoring targets, discharge safety fast, and specialist consultation guidance.
Critical Sample Workup Hyperinsulinism Safety Fast Baylor Ed. 33 cross-checked Sept 2026

1. Purpose

Persistent hypoglycemia = glucose support required beyond 48–72 hours of life to maintain blood glucose >60 mg/dL.

Goals:

  • Recognize infants who need diagnostic evaluation
  • Obtain a critical sample during hypoglycemia (plasma glucose <50 mg/dL)
  • Identify the mechanism driving hypoglycemia
  • Prevent recurrent brain fuel deprivation
  • Involve endocrinology/metabolic specialists before long-term therapy decisions

2. When to Suspect Persistent Hypoglycemia

SituationWhy It Matters
Glucose support needed beyond 48–72 hoursMeets the definition of persistent hypoglycemia — transitional hypoglycemia should have resolved
Frequent recurrent hypoglycemiaSuggests a persistent endocrine or metabolic disorder rather than transitional physiology
GIR >10 mg/kg/minStrongly suggests hyperinsulinism or another pathologic cause — transitional hypoglycemia rarely requires this level of support
Hypoglycemia beyond the first weekFurther increases concern for an underlying hypoglycemia disorder requiring diagnosis
Micropenis, undescended testes, midline defects, or recurrent hypoglycemiaThink pituitary disease (not only insulin or metabolic causes) — evaluate all pituitary axes

3. Important Rule

Do not send random broad lab panels without clinical context

History, physical examination, timing, feeding state, illness severity, and associated findings should guide the differential diagnosis before ordering tests.


Example: a newborn with micropenis and undescended testes needs pituitary evaluation — not only insulin and acylcarnitine testing.

4. Timing of Diagnostic Testing

  • The most useful endocrine critical sample is obtained when plasma glucose is <50 mg/dL
  • Sample may be obtained during: spontaneous hypoglycemia, planned GIR weaning, or monitored fasting
  • Testing done when glucose is normal is usually not helpful for diagnosing congenital hypoglycemia disorders
Why the timing matters

Endocrine responses to hypoglycemia (cortisol, GH, suppression of insulin) can only be accurately interpreted during actual hypoglycemia. A normal-glucose critical sample misses the diagnostic window entirely and may falsely reassure.

5. Glucose Targets

Infant GroupTarget Glucose
Suspected congenital hypoglycemia disorder>70 mg/dL
High-risk neonate without suspected congenital disorder (SGA, IDM)>60 mg/dL
Before discharge in known genetic or persistent hypoglycemiaConsider safety fast to confirm glucose stays >70 mg/dL if a feed is missed

6. Discharge Safety Fast

  • Before discharge, infants with known genetic or persistent hypoglycemia may need a specialist-reviewed fasting test
  • Goal: confirm blood glucose can remain >70 mg/dL if a feeding is missed
  • Baylor minimum safety fast duration: 6–8 hours in these infants
Never discharge without a safety plan

Infants with known or suspected persistent hypoglycemia should not be discharged without a specialist plan, confirmed feeding schedule, outpatient monitoring plan, and ideally a completed safety fast demonstrating adequate fasting tolerance.

7. Major Causes of Persistent Hypoglycemia

CategoryExamples
Insulin secretion / production disorders Congenital hyperinsulinism, infants of diabetic mothers (prolonged), perinatal stress hyperinsulinism, erythroblastosis fetalis, Beckwith-Wiedemann syndrome
Endocrine abnormalities Hypopituitarism, central adrenal insufficiency, GH deficiency, panhypopituitarism, primary adrenal insufficiency, CAH, congenital adrenal hypoplasia, adrenal hemorrhage
Ketogenesis and fatty-acid oxidation disorders MCAD deficiency, other fatty-acid oxidation disorders, carnitine transport disorders, carnitine deficiency
Amino-acid metabolism disorders Maple syrup urine disease, propionic acidemia, methylmalonic acidemia
Inborn errors of glucose production Glycogen storage disease, gluconeogenesis disorders, pyruvate carboxylase deficiency, PEPCK deficiency, fructose-1,6-bisphosphatase deficiency, hereditary fructose intolerance

8. Bedside Clues by Etiology

Bedside ClueMore Likely Diagnosis
GIR >10 mg/kg/min, low ketones, low free fatty acids Hyperinsulinism — insulin suppresses ketone and FFA production
Micropenis, cryptorchidism, midline defects, hypoglycemia Hypopituitarism, GH deficiency, ACTH deficiency
Hyponatremia + hyperkalemia + shock Primary adrenal insufficiency or CAH
Hypoketotic hypoglycemia with illness or fasting intolerance Fatty-acid oxidation disorder or hyperinsulinism
Acidosis, elevated lactate, hyperammonemia, abnormal urine organic acids Organic acidemia or other metabolic disease
Hepatomegaly or fasting intolerance Glycogen storage disease or gluconeogenesis defect

9. Critical Sample — What to Send When Plasma Glucose <50 mg/dL

Draw when plasma glucose <50 mg/dL — before treating if infant is stable
TestWhy It Matters
Laboratory plasma glucoseConfirms true hypoglycemia — not POCT alone
Serum insulinEvaluates hyperinsulinism, but insulin alone is insufficient for diagnosis
Plasma beta-hydroxybutyrateDetermines whether ketones are appropriately produced (suppressed = hyperinsulinism or FAO disorder)
Free fatty acidsIdentifies insulin effect (suppressed FFA) or fatty-acid oxidation problems
Plasma cortisolScreens adrenal response during hypoglycemia
Growth hormoneScreens GH response during hypoglycemia

Glucagon IM After Critical Sample (When Used Diagnostically)

Post-Sample Glucagon Protocol
1. Draw all critical blood samples at plasma glucose <50 mg/dL
2. Give glucagon IM
3. Repeat lab glucose at 30 minutes
4. Repeat cortisol, growth hormone, and lab glucose at 60 minutes

⚠ If infant is unstable: give D10W 2 mL/kg IV bolus immediately — do not delay treatment for sampling

10. Monitoring During Evaluation

Glucose ValueMonitoring Action
Routine persistent hypoglycemia evaluationCheck glucose every 3 hours
Blood glucose <60 mg/dLCheck every 1 hour
Blood glucose <50 mg/dLSend critical sample before treatment if clinically safe
Use STAT laboratory glucose during diagnostic workup

Use STAT laboratory glucose rather than POCT alone during the diagnostic evaluation. POCT has limited accuracy at low glucose values and may not accurately confirm the hypoglycemic episode needed to time sample collection.

11. Additional Testing for Non-Endocrine Causes

TestPossible Use
C-peptideHelps evaluate endogenous insulin secretion (C-peptide is co-secreted with insulin; absent in exogenous insulin administration)
Acylcarnitine profileScreens for fatty-acid oxidation disorders (e.g., MCAD, LCHAD, VLCAD)
Plasma amino acidsScreens amino-acid metabolism disorders
Pyruvic acidHelps evaluate disorders of energy metabolism (pyruvate carboxylase deficiency, PEPCK deficiency)
Ammonia and lactateScreens organic acidemias, mitochondrial disease, and other metabolic causes
Urine organic acids and ketonesIdentifies organic acidemias and the adequacy of ketotic response
Targeted gene panelConfirms genetic hypoglycemia disorders (ABCC8, KCNJ11, GCK, GLUD1, HADH, HNF4A mutations, etc.)
Timing of metabolic and genetic testing

Timing should be discussed with genetics/metabolic specialists. Many of these tests (acylcarnitine profile, plasma amino acids, gene panel) do not require the patient to be hypoglycemic at the time of collection — unlike the endocrine critical sample.

12. Hyperinsulinism: Key Point

Do NOT diagnose hyperinsulinism by insulin level alone

Diagnosis of hyperinsulinemic hypoglycemia requires interpretation of the full critical sample:

  • Insulin level (in context)
  • Beta-hydroxybutyrate (appropriately suppressed?)
  • Free fatty acids (appropriately suppressed?)
  • Glucose response to glucagon
  • Clinical context (GIR requirement, timing, associated features)
Why insulin alone misleads

Insulin has a very short half-life and insulin assays have poor precision at low concentrations. A "detectable" insulin level during hypoglycemia may appear normal by assay but still be inappropriately elevated for the glucose level. The ketone/FFA pattern and glucagon response are often more diagnostically useful.

13. Immediate Treatment While Evaluating

SituationTreatment
Stable infant during planned evaluation Obtain critical sample at plasma glucose <50 mg/dL — treat after sampling
Unstable infant with hypoglycemia Treat immediately with D10W 2 mL/kg IV bolus — do not delay for sampling
Maintaining glucose after sampling Start or adjust IV dextrose to maintain glucose >60 mg/dL (or >70 mg/dL if congenital disorder suspected)

14. Practical Bedside Algorithm

Persistent Hypoglycemia — Bedside Approach
STEP 1 — Identify persistent hypoglycemia: glucose support needed >48–72h to maintain >60 mg/dL
↓
STEP 2 — Review history and exam before ordering broad testing
   → GIR >10? SGA/IDM/perinatal stress? Beckwith-Wiedemann features?
   → Micropenis/undescended testes? Hepatomegaly? Acidosis? Electrolyte abnormalities?
↓
STEP 3 — Check glucose every 3 hours; when <60 mg/dL → check hourly with STAT lab glucose
↓
STEP 4 — When plasma glucose <50 mg/dL:
   → Stable: draw critical sample FIRST, then treat
   → Unstable: D10W 2 mL/kg IV immediately
↓
STEP 5 — Critical sample: plasma glucose, insulin, beta-hydroxybutyrate, FFA, cortisol, GH
↓
STEP 6 — Give glucagon IM after critical sample when used diagnostically
   → Repeat lab glucose at 30 min + cortisol/GH/glucose at 60 min
↓
STEP 7 — Add non-endocrine tests as indicated: C-peptide, acylcarnitine, amino acids, pyruvate, NH3, lactate, urine organic acids, gene panel
↓
STEP 8 — Consult endocrinology after initial results; before disease-specific treatment when clinically possible
↓
STEP 9 — Safety fast (6–8h) before discharge in known genetic or persistent hypoglycemia — confirm glucose >70 mg/dL

15. Common Mistakes

MistakeBetter Action
Sending labs when glucose is normalDraw the endocrine critical sample when plasma glucose is <50 mg/dL — normal-glucose samples are not diagnostic
Treating before obtaining a critical sample in a stable infantIf stable, draw critical sample first, then treat
Delaying treatment in an unstable hypoglycemic infantIf unstable, give D10W 2 mL/kg IV immediately — never delay for sampling in an unstable infant
Diagnosing hyperinsulinism by insulin level aloneInterpret insulin with ketones, free fatty acids, glucagon response, and clinical context
Ignoring physical cluesMicropenis and undescended testes should trigger pituitary evaluation — not only metabolic testing
Discharging persistent hypoglycemia without a specialist planConsider specialist consultation, safety fast, and outpatient monitoring plan before discharge
Using only POCT during diagnostic workupUse STAT laboratory glucose — POCT accuracy is insufficient to time critical sample collection

16. Parent Explanation

What to Tell Families

"Most newborn low blood sugar improves in the first 1–2 days, but if the baby still needs sugar supplementation after that, we need to look for a hormonal or metabolic reason."


"The best time to find the cause is during an actual low sugar episode, so we collect special blood tests before giving treatment — as long as the baby is stable enough for us to do that safely."


"If the baby is sick or unstable, we treat the low sugar right away without waiting, and gather information during the next episode."


"Before going home, some babies need a supervised fasting test to make sure blood sugar stays safe if a feeding is delayed. A specialist will help guide these decisions."

Key Takeaways — Chapter 4.5

  • Persistent hypoglycemia = glucose support needed beyond 48–72 hours to maintain >60 mg/dL
  • GIR >10 mg/kg/min strongly suggests hyperinsulinism or another pathologic cause
  • Do not send random broad panels — use clinical clues to guide the differential first
  • The endocrine critical sample must be drawn when plasma glucose <50 mg/dL
  • Critical sample: plasma glucose, insulin, beta-hydroxybutyrate, free fatty acids, cortisol, GH
  • Testing at normal glucose = not useful for diagnosing congenital hypoglycemia disorders
  • Stable infant → draw critical sample first, then treat
  • Unstable infant → treat immediately with D10W 2 mL/kg, do not delay for sampling
  • Hyperinsulinism: diagnose using full critical sample (suppressed ketones + FFA + glucagon response) — not insulin level alone
  • Bedside clues: low ketones/FFA → hyperinsulinism; micropenis + midline → pituitary; Na/K/shock → adrenal; hepatomegaly → glycogen storage; acidosis + lactate → organic acidemia
  • Suspected congenital disorder → target glucose >70 mg/dL; high-risk neonates without congenital disorder → >60 mg/dL
  • Use STAT laboratory glucose (not POCT) during diagnostic workup
  • Metabolic/genetic tests (acylcarnitine, gene panel) do not require hypoglycemia at time of collection
  • Safety fast (6–8 hours) before discharge in known or suspected persistent hypoglycemia
  • Consult endocrinology before starting disease-specific treatment when clinically possible

References

  • Baylor College of Medicine — Guidelines for Acute Care of the Neonate, current edition: Chapter 4.5
  • Thornton PS, et al. Recommendations from the Pediatric Endocrine Society for evaluation and management of persistent hypoglycemia in neonates, infants, and children. J Pediatr. 2015
  • Stanley CA, et al. Hyperinsulinism and hyperammonemia in infants with regulatory mutations of the glutamate dehydrogenase gene. N Engl J Med. 1998
  • De León DD, Stanley CA. Mechanisms of disease: advances in diagnosis and treatment of hyperinsulinism in neonates. Nat Clin Pract Endocrinol Metab. 2007
  • Lord K, De León DD. Monogenic hyperinsulinemic hypoglycemia: current insights into the pathogenesis and management. Int J Pediatr Endocrinol. 2013
  • Palladino AA, et al. Hypoglycemia: a risk factor for brain injury in infants with hyperinsulinism. J Pediatr. 2009
  • Kapoor RR, et al. Hyperinsulinaemic hypoglycaemia. Arch Dis Child. 2009
  • Huidekoper HH, et al. Fasting in children with fatty acid oxidation disorders. Orphanet J Rare Dis. 2016