About a month ago, my sister’s blood glucose was over 900 mg/dl [50 mmol/L], so the doctor asked her to have tests for anti-insulin, anti-tyrosine and anti-GAD antibodies. The first two tests were good, but the anti-GAD was bad. The doctor said that she would have to have “bad” results for all three tests to be diagnosed with type 1 diabetes. So, the doctor cannot figure out if the pancreas has been destroyed completely or partially because of the anti-GAD autoantibodies. He said that we should wait for several month to see if she will get fat from injecting insulin. This would mean that maybe she does not need insulin, just pills. What is your opinion? Can my sister be cured? What can you tell me about the anti-GAD? Currently, her blood glucose levels are between 50 and 100 mg/dl [2.8 and 5.6 mmol/L].

There must be some mix-up in what you have written since this is very complicated. In type 1 diabetes, antibodies are only positive about 60 to 80% of the time. So, it is possible to have type 1 diabetes 20 to 40% of the time even with negative antibodies. Any positive antibodies suggest type 1 diabetes, however. If the blood glucose levels are mostly normal but only occasionally abnormal, then this could be a very early stage of type 1 diabetes. Usually, over months, sometimes years, the blood glucose levels rise as the beta cells of the pancreas “die” and less and less insulin becomes available. Some diabetes specialists would start insulin now with the hope of resting the damaged pancreas and keeping the glucose levels normal longer. Others would just use a diet to stop simple sugars and thus not need as much insulin. There is no good science to suggest that one is better than others. My own personal belief is that such decisions should be made based upon frequent before and after meal blood glucose readings for a week or two. If the values are entirely normal, then I advice just the dietary approach. If they are not, then insulin is started. If the blood glucose levels are high pre-breakfast, then a bedtime insulin such as Lantus makes sense. If they are mostly high after eating, then decreasing carbohydrates becomes a dietary possibility and, if this does not work, then using insulin analogs such as NovoLog or Humalog would be my preference to control the post-food high sugar levels.

Antibody levels are only useful for research and minimally helpful for treatment decisions if negative; if positive, they suggest that insulin will be needed sooner or later.

SB