OBJECTIVEWe report four patients whose cases resulted in our observation that profound hypoglycemia resulting from intermittent hyperinsulinism plays a significant role in patients with brainstem dysfunction from Chiari I or II malformations who have intermittent autonomic dysfunction (“blue spells”).METHODSThe records of four children with severe brainstem dysfunction associated with hindbrain herniation (Chiari I or II malformation) were reviewed retrospectively. Each patient had severe lower cranial nerve dysfunction that required tracheotomy and feeding tube placement. After we found that profound hypoglycemia had occurred during a spell of autonomic dysfunction in one patient, the charts of the other three patients were reviewed for evidence of hypoglycemia. Now, whenever one of them has evidence of autonomic dysfunction, prospective studies of glucose and insulin levels are performed. Three of the patients had Chiari II malformation in association with myelomeningocele, and one patient had a Chiari I malformation resulting from Pfeiffer’s syndrome.RESULTSHypoglycemia occurred in these patients episodically, and usually when their shunts were functioning. The hypoglycemia was associated with hyperinsulinemia in each patient. The brainstem structures of these children (presumably the dorsal motor nuclei of the vagus) were extremely sensitive to changes in local or regional intracranial pressure. These changes were triggered by intermittent shunt failure, agitation from pain, abdominal distention from constipation, and retention of CO2. In patients with Chiari malformations, even mild increases in intracranial pressure lead to brainstem dysfunction. One possible explanation is that pressure on the deformed Xth cranial nerve nuclei may lead to insulin release and life-threatening hypoglycemia. Continuous-drip feeds are necessary to prevent this complication.CONCLUSIONPatients with severe intermittent brainstem dysfunction after decompression of Chiari I or Chiari II malformations should have laboratory studies of glucose levels performed at the time of the episodes to rule out hypoglycemia.