JAMA Sept 22, 1975 Vol. 233 No. 12

Practical Endocrinology

Postprandial Hypoglycemia

Fact or Fiction?

Fred D. Hofeldt, MD; Robert A. Adler, MD; Robert H, Herman, MD


HYPOGLYCEMIA enjoys a popular position in the public's eye as a nonspecific medical condition that frequently provides an explanation for the varied symptoms that occur in daily life. Patients with reactive hypoglycemia tend to manifest variations of the neurotic triad with abnormally high Minnesota Multiphasic Personality Inventory scales of hysteria and hypersomatization as part of their personality profile. The magnitude of these complaints may lead a physician to be overzealous in searching for a cause to account for the patient's symptoms. Clinical studies of reactive hypoglycemia have been hampered by its vagueness in definition, and by the remarkably different criteria used in selection of patients. Some investigators use clinical symptoms alone in choosing patients for study, while others include patients who have had blood glucose values below a certain arbitrary limit, whether or not concomitant symptoms of hypoglycemia exist. Very few investigators use objective evidence to verify the pathophysiologic nature of the hypoglycemic symptoms,

We employ strict criteria for diagnosing reactive hypoglycemia. We have defined the condition by the following clinical and biochemical indexes. Hypoglycemia exists as a bona fide state only when the patient's symptoms occur simultaneously with the nadir in the blood glucose level. Following this glucose nadir, evidence of hypotha-lamic-pituitary-adrenal responsiveness is manifest by a substantial rise in plasma cortisol to stress levels. With these rigid criteria, we have demonstrated abnormalities of insulin secretion in the majority of such patients. The symptoms are due to adrenergic mechanisms that provide an emergency back-up system for increasing blood glucose levels through hepatic glucose output. We have classified patients with bona fide reactive hypoglycemia into four categories: (1) diabetic reactive hypoglycemia, (2) alimentary reactive hypoglycemia, (3) hormonal reactive hypoglycemia (hypothyroidism, hypoadrenalism), and (4) idio-pathic reactive hypoglycemia.

The over-diagnosis of reactive hypoglycemia most likely occurs because of the frequency with which low blood glucose values normally occur during the postprandial state. Gahill and Soeldner1 described 23% of a normal population as having blood glucose levels of less than 50 mg/100 ml and an occasional patient as having a glucose value as low as 35 mg/100 ml while asymptomatic. With use of continuous monitoring of plasma glucose level, Burns et al2 showed that approximately 42% of normal asymptomatic subjects, have glucose values that fall below 50 mg/100 ml. We have randomly sampled an outpatient population with nonspecific complaints. Of 25 such patients, approximately 48% had a blood glucose value below 50 mg/100 ml and had no symptoms coincident with the nadir in the blood glucose curve. There was also a lack of evidence of pathophysiological stress at the glucose nadir, as demonstrated by a failure of cortisol level to rise following the low glucose value. We have seen four patients with blood glucose values ranging between 34 and 37 mg/100 ml who have remained asymptomatic. These low blood glucose values, occurring during a glucose tolerance test, represent a summation of metabolic events that occur when the body's intermediary metabolism is shifted from the fed to the fasting state. This nadir in the blood glucose curve occurs as a transitional point or "switch point" in intermediary metabolism between the fed and fasting states, as described. Because of the relatively frequent occurrence of this nadir in blood glucose level, and because of the connotation that hypoglycemia has with the public, we prefer the term "transitional low blood glucose state" to indicate this normal sequence of intermediary glucose metabolism.

This occurrence of an asymptomatic nadir in blood glucose value during a glucose tolerance test may cause the physician to misattribute clinical significance to the biochemical event. The diagnosis should be "transitional low blood glucose state of no clinical significance." It is because of this frequent occurrence that both physicians and of the public deserve major reeducation. The term "reactive hypoglycemia" should be restricted to that clinical entity in which there is a timely occurrence of symptoms coincident with a low blood glucose value. The cortisol responsiveness has been a useful research tool by which to identify these patients.



1. Cahili GK, Soeldner JS: A noneditorial on non-hypnglvcemia. N Eng J Med 291:905-6, 1974.

2. Burns TW, Bregnant R, V'an Teenen HJ, et al Observations on blood glucose concentrations of human subjects during continuous sampling. Diabetes 14:186-193, 1965.