Published on December 26th, 2013 | by Claresa Levetan MD, FACE

When glucose sensors first became available in clinical trials some 2 decades ago, I decided to wear a sensor to compare my glucose levels as a non-diabetic individual with glucose levels of my patients. I was excited to have this new tool, which measured 288 glucose readings a day and could be used 3 days at a time, as a resource for patients, particularly for those with blood glucose levels that had been difficult to control.

During my training as an endocrinologist, we didn’t have glucose meters. In the hospital, we used glucose strips that required drawing a significant amount of blood, then waiting and wiping the blood off. You determined glucose range based on how dark the strip became.

While wearing my sensor, I did my best to exercise and eat correctly, but I also indulged in chocolate donuts. Still, when I compared my daily readings with that of a patient with a fabulous A1C of 6.3%, I was shocked. Even with the donuts, my highest glucose over the three days was 103 mg/dL, whereas my patient’s A1C levels could bounce from 50 to 500 mg/dL in moments. There was no comparison.

Here I was patting myself on the back for thinking that I was taking such great care of my patients, when in reality, I learned that comparing A1Cs for people with and without diabetes was like comparing apples and oranges. Even patients who used insulin pumps, ate the same meals every day, exercised daily and “did everything right” can hardly hope to achieve the glucose levels of a person without diabetes. I realized I needed to stop blaming my patients or thinking myself superior.

Research has played a vital role in helping others come to the same conclusion. It can be easy to overlook that our knowledge of glucose control is still in its infancy. One of the landmark studies that helped us understand glucose control was the 1993 Diabetes Control and Complications Trial (DCCT), the first randomized prospective clinical trial that provided the data to show glucose control mattered.  Prior to this study, while many of us in the medical community knew there was a link, we didn’t have the data to demonstrate the relationship between glucose control and diabetes outcomes.

What we learned most from the DCCT was that diabetes was not as simple as we thought. It’s a great example of a study teaching us much in its failure. Diabetes thought leaders who designed the DCCT said the goal of the study was to restore normal glucose levels among patients with Type 1 diabetes; a major treatment goal was to establish an average hemoglobin A1C over the 10-year study period of ≤6.05% without an increased risk for hypoglycemia.

Researchers reported that they couldn’t achieve this goal with just insulin therapy, whether it was administered by many injections a day or through pumps. Patients in the intervention group of DCCT were only able to achieve an average hemoglobin A1C of 7.2%, with significantly more hypoglycemia than those in the control groups. However, they did have a 50% reduction in rates of common diabetes complications over that of the control group. Even if A1C levels couldn’t be brought below the 6.0% threshold, the study results showed, the act of controlling it yielded benefits to patient outcome.

There have been many groundbreaking advances in pumps since then, but modern technology can have its limits. Insulin pumps with sensors were recently shown only to decrease A1C levels from 8.3% to 7.5% over 12 months, with further reduction to 7.4% after an additional 6 months of treatment. Despite technological advances in sensors and pumps, sensor-augmented pump therapy did not improve hemoglobin A1C levels as much as pump therapy in the DCCT did decades ago. That’s the bad news; the good news is that these achievements were made without the associated weight gain or hypoglycemia observed in the DCCT cohort.


Diabetes is a difficult physiological condition to control, and insulin therapy or lifestyle changes will only take us so far in controlling glucose. We are now only just beginning to understand that there are 4 different cell types within the islet that make 5 different hormones, all of which impact glucose control (insulin, amylin, glucagon, somatostatin, pancreatic polypeptide, and islet ghrelin). All four cell types become dysfunctional over time with diabetes. Glucose homeostasis requires an adequate number of completely functional islets, as illustrated by the failure of even intensive insulin regimens to restore normoglycemia (yes, it’s a word) among diabetic patients.

I believe that insulin independence for Type 1 patients will occur sooner than believed possible through new research in the human proteome that generates new islets from stem cells residing in the pancreas of children and adults. Those with Type 1 diabetes will also utilize safe, tried and true immune agents that are currently used in a multitude of diseases in order to protect new islets from autoimmune attack.

But I no longer blame my patients, their diets, activity level, or even the cake, candy or snacks for their A1C levels. It comes down to physiology. At the time of diagnosis of Type 1 diabetes, there is a reduction of 90% of the beta cell mass; Type 2 patients experience a reduction of as much as 75%. Diabetes is the fault of a pancreas that has too few functioning islets.

The term “non-compliant patient” doesn’t exist in my vocabulary when it comes to the treatment of diabetes; I now know better.

About the Author

Claresa Levetan, M.D., is a noted endocrinologist with expertise spanning diverse areas of diabetes. She has been internationally recognized for the translation of science and technology into improved health care options for patients with diabetes. She served as an associate editor of the journals Clinical Diabetes and Diabetes Forecast.

Share This
Skip to content