Protamine-containing insulins are strong risk factors, and human insulin analogues are possible risk factors for insulin autoantibody: case-control study

Insulin autoantibody is known to cause fluctuation of blood glucose. We examined whether medications for diabetes are risk factors for insulin autoantibody. Especially, we examined the associations between types of insulin and insulin autoantibody. We performed a case-control study. From April 2005 to March 2010, insulin autoantibody was measured 273 times in 217 patients in our hospital. Insulin autoantibody was positive (greater than 10%) 53 times in 19 patients (case), and was negative 220 times in 198 patients (control). Oral hypoglycemic agents were not risk factors for insulin autoantibody; the odds ratio was 0.0. In contrast, insulin use was a significant risk factor for insulin autoantibody; the odds ratio (95% confidence interval) was 56.3 (7.3-432.5). As for the types of insulin and insulin autoantibody, human insulins without protamine were not risk factors; the odds ratio was 0.0. For protamine-containing insulins, the odds ratio and adjusted odds ratio (adjusted by age, gender, and disease: type 1 diabetes mellitus, type 2 diabetes mellitus, and no diabetes) were 35.3 (9.6-129.5) and 29.6 (7.6- 115.4), respectively. For Aspart-containing insulins, they were 6.2 (2.2-17.9) and 3.8 (1.2- 12.0), respectively. For Glargine, they were 3.2 (0.6-16.7) and 1.3 (0.2-8.3), respectively. To decrease the problem of insulin antibody, avoiding the use of protamine-containing insulins and avoiding the use of human insulin analogues might be preferable for the patients with diabetes.