No abstract

Antibodies to the insulin receptor are insulinomimetic in vitro, although they generally induce insulin resistance in vivo. We report the novel case of a patient who presented with fasting hypoglycemia as the sole manifestation of autoantibodies to the insulin receptor. Prednisone therapy (120 mg per day) produced a rise in fasting glucose to more than 100 mg per deciliter (6 mmol per liter) within 48 hours, although there was no detectable change in the titer of antireceptor antibodies. After 10 weeks of therapy, the titer of antireceptor antibodies had fallen approximately 100-fold, and prednisone could be discontinued without recurrence of hypoglycemia. This case demonstrates that antireceptor antibodies must be considered in the differential diagnosis of hypoglycemia, especially in patients with other manifestations of autoimmunity.

Twenty-six patients with symptoms suggestive of postprandial hypoglycemia were investigated by oral glucose tolerance test (OGTT). During the OGTT, symptomatic hypoglycemia occurred in 10 (38.5%). Nine of these 10 sugjects were given mixed meal tolerance tests (MMTT) and symptomatic hypoglycemia failed to occur in any case. During the OGTT the nadir glucose was significantly lower than that during MMTT (44.1 +/- 1.5 vs. 77.3 +/- 4.8 mg/dl +/- SEM, respectively; p less than 0.0005). Serum insulin during MMTT peaked significantly earlier than during OGTT (46.7 +/- 7.3 vs. 86.7 +/- 11.7 minutes (SEM, respectively; p less than 0.0125). The early secretion of insulin during MMTT may explain the lack of symptomatic hypoglycemia in these patients. We conclude that reactive hypoglycemia, when tested by a more natural stimulus (such as mixed meal) rather than by OGTT, is uncommon.

Phenformin is now an accepted therapeutic agent for the treatment of patients with diabetes mellitus of the adult-onset type. In spite of widespread evaluation, a t both the clinical and the laboratory levels, the exact mechanism of action remains an enigma. In contrast to the sulfonylureas,' phenformin therapy does not cause an increase in serum insulin levels. Long-term therapy, however, with either phenformin or the sulfonylureas, produces an improvement in carbohydrate tolerance and a concomitant decrease in circulating serum insulin levels.Grodsky and coworkers3 called attention to the presence of striking increases in serum insulin levels in the presence of obesity and demonstrated that phenformin therapy resulted in a significant decrease in these insulin levels. Subsequently Roddam and his associates4 noted that hyperinsulinemia was not present in all obese patients and that the phenformin effect on serum insulin levels was significantly apparent only in patients with hyperinsulinemia.Kipnis has reported that although hyperinsulinemia is present in obese patients with both normal and abnormal carbohydrate tolerance, there is a relative insulin deficiency in patients with abnormal carbohydrate tolerance when the level of blood glucose is taken into a c~o u n t .~ This observation is in contrast to that recently reported from our laboratory. Furthermore, Grodsky has stated that the hyperinsulinemia found in obesity disappears after significant weight reduction is a~h i e v e d .~The purpose of this paper is to review these points briefly, to further assess the mechanism whereby phenformin attacks hyperinsulinemia, and to explore its usefulness in diabetes and obesity. Materials and Methods 6The patients for this study were selected from a group referred to a special obesity research clinic. Several were hospitalized in the clinical research center of the University of Alabama Medical Center. Glucose tolerance tests were performed, using 1.75 g of glucose per kg of ideal body weight. The eight-hour glucose infusions were administered with a constant infusion pump, using 1 g of glucose per kg of ideal body weight per hour.For the intravenous tolbutamide tolerance tests, 1 g of sodium tolbutamide, diluted to a volume of 20 ml with sterile distilled water, was injected over a three-minute period.* Thirty grams of arginine were infused over a 30-minute period for arginine tolerance tests. A 15% alcohol solution was infused a t a rate of 2 ml/min for the alcohol tolerance tests.756