Forty-five insulin-dependent diabetics were randomized to 1 yr treatment with either continuous subcutaneous insulin infusion (CSII), multiple insulin injections (MI), or continued conventional treatment. The CSII group used regular insulin only, the MI group used 4-6 premeal injections of regular insulin and intermediate insulin at night, and the conventional group used two daily injections of combined regular and intermediate insulin. Only highly purified porcine insulin was used. Near normoglycemia was obtained during CSII and MI but not during conventional treatment.
Antibodies against insulin were measured in serum samples by measuring the binding of iodinated porcine insulin to serum after removal of free and antibodybound insulin from the samples by acid charcoal. The percent binding of 125I-labeled insulin increased significantly during MI and CSII, in contrast to conventional treatment.
Nineteen patients had sufficient binding capacity for Scatchard analysis. In the CSII and MI groups, high- or low-affinity antibodies or both were induced.
When insulin was administered subcutaneously during MI or CSII for 1 yr, the insulin antibody production increased, in contrast with conventional treatment.