A discussion meeting was held on October 17, 1989, to address the current status of the definition of significant hypoglycemia in infancy, especially in the normal- and low-birth-weight neonate.
Robert Schwartz introduced the complexity of the problem by indicating the multiple variables (duration, severity, cerebral blood flow, rates of glucose uptake, availability of alternate substrates, oxygen, etc) in equating a plasma glucose value with neurodevelopmental consequences.
Marvin Cornblath reviewed the various definitions of hypoglycemic blood sugar levels reported since 1911 which depend upon the method of blood sugar analysis, clinical recognition and concerns. Severe symptomatic hypoglycemia that persisted or recurred was first reported in neonates in 1937. Lower blood sugar levels, documented since the 1920s in both full-term and premature newborns, had been considered physiologic. The recognition of transient significant hypoglycemia first in symptomatic and then in asymptomatic small-for-gestational-age, neonates required new definitions in the 1960s. These definitions were later modified as changes in treating both the mother in labor and at delivery and the neonate occurred. Intensive care and the survival of very-low-birth-weight newborns have compounded the problem of definition. Currently methods are available to correlate plasma glucose concentrations and glucose metabolism in vivo in the brain with specific neurologic dysfunctions. This should permit a better definition of the continuum of significant hypoglycemia than has been available before.
William Hay analyzed studies of the various rapid bedside glucose oxidase stick techniques used to screen for blood glucose concentrations. He concluded that their dependence on the hematocrit, their requirements for precision in performance and timing, great variance (±5 to 15 mg/dL), and lack of reproducibility, especially at blood glucose values less than 50 mg/dL, made their use in the neonate unsatisfactory (whether read by eye or by meter).