Perinatal factors associated with hypoglycemic brain injury were studied by review of medical records in 60 hypoglycemic neonates at Tottori University, Yonago, Japan. Patients were classified in 2 groups: Group I, 12 patients, abnormal, with mental retardation, developmental delay, cerebral palsy or epilepsy; and Group II, 48 patients, normal at follow-up. Proportion of infants small for gestational age (<10th percentile) was high in both groups (75% vs 58%) but was not associated with brain injury. Very low blood glucose levels (<15 mg/dl) occurred in 50% of Group I vs 14.6% of Group II (p=0.015). Duration of hypoglycemia was longer in Group I (median, 14 h) than in Group II (median, 1.75 h) (p<0.001). Associated factors more frequent in Group I than in Group II included toxemia (33.3% vs 8.3%, p=0.043), fetal distress (58.3% vs 14.5%, p=0.004), Apgar score <5 at 1 min (33.5% and 6.4%, p=0.025), neonatal seizures (53.8% vs 4.3%, p<0.001), and pathological jaundice (41.7% vs 6.4%, p=0.006). Eight of 9 patients in Group I had abnormal MRI at follow-up, showing cortical atrophy and white matter lesions, with occipital and parietal predominance. Apgar scores were partially correlated with the extent of brain lesions. Brain injury in neonates with prolonged hypoglycemia may be exacerbated by associated factors such as hypoxia, seizures, and jaundice. 
COMMENT. Neonatal seizures with hypoglycemia are correlated with duration of hypoglycemia and neurological outcome. Seizures may begin at onset of hypoglycemia but usually appear after 12 h of continuous hypoglycemia. 
Of 27 infants and children with seizures associated with hypoglycemia reported from the Mayo Clinic, only 2 had an onset of seizures in the neonatal period, and in 20 the etiology of hypoglycemia was unknown. Neurologic disease preceded the onset of symptoms in 50% of the 20 patients with cryptogenic hypoglycemia. Evidence for a primary neurological cause for seizures included birth injury, kernicterus, hydrocephalus, and cerebral dysgenesis. Level of blood sugar at time of seizure in patients with primary neurologic disorder was significantly lower than in patients with normal neurologic findings. Occurrence of seizures was not closely correlated with the level of blood sugar. A primary cerebral lesion should be considered as an etiologic factor in some neonatal and childhood hypoglycemic seizures.