A forty-two-year-old man with hyperglycemia and pronounced reactive hypoglycemia after meals and oral glucose is reported. He produced anti-insulin antibodies without previous known immunization.

Labeled insulin, when mixed with the patient's serum, passed with the serum proteins on gel filtration, and was found among the 7S globulins on preparative ultracentrifugation, in the anodic γ-region on agar electrophoresis, and among the anodic γG-globulins on immunoelectrophoresis. The clearance of labeled insulin in vivo was very slow, and the glucose response to intravenous insulin was absent. The free binding capacity in serum was several thousand μU./ml. Acid splitting of the insulin—anti-insulin complexes revealed 7,000 μU./ml. or more of endogenous insulin, and antibodies which bound human and pig insulin better than bovine insulin. After oral glucose the patient's serum insulin increased by 2,300 μU./ml. Some new principles are introduced in the quantitative studies on serum insulin and antibodies. Theoretical restrictions on these and other quantitations are discussed.

The hyperglycemia may be explained by the antibodies binding secreted insulin, and the hypoglycemia by an insulin overshoot from maximally stimulated β-cells when saturation of the binding capacity was approached. Carbohydrate restriction prevented the hypoglycemic attacks. The cause of the antibody production and the properties of the antibodies are discussed.

This content is only available via PDF.