-Not everybody with diabetes antibodies develops type 1 diabetes, and some people coincidentally have diabetogenic antibodies and non-type 1 diabetes. This is particularly true of people with only one positive diabetes antibody diagnosed with diabetes as adults.
- Many doctors incorrectly assume that diabetes in a child is type 1, and don't test anything else; or despite tests that don't support a type 1 diagnosis, don't know what else to think.
I always think of this when I remember a family I met on the DOC many years ago whose son was diagnosed with diabetes with a blood sugar over a thousand and was in a nonketotic hyperosmolar hyperglycemic coma at the time. He was antibody negative. But he was also ten years old. So they diagnosed him with type 1 diabetes anyways and put him on insulin. About a year (maybe two years?) later, he went off insulin and about six years later he's still off insulin.
- A lack of available other diagnoses. Type 2 diabetes is by definition diabetes caused by an insulin secretory defect on top of a background of insulin resistance. However... that description does not hold true for a very large portion of adults with type 2 diabetes diagnoses, and many doctors are reluctant to make a diagnosis of diabetes that's not type 1 or type 2, particularly because the other available diagnoses are other specific diabetes and gestational diabetes, and it's often impossible to say what specific other the diabetes is. So if they're pretty sure there's not an insulin secretory defect on top of insulin resistance, and they're pretty sure it's diabetes... then you may get a type 1 diagnosis.
- The c-peptide test of insulin production needs to be performed when a person is NOT in DKA, and has a blood sugar that's only slightly high to be a good measure of insulin producing capacity.
Based on studies of c-peptide levels in people with longstanding diagnoses of type 1 diabetes, I think that at least 5% of type 1 diabetes diagnoses are incorrect.