View More 2020, Scientific Sessions, Session Coverage

Definition of hypoglycemia impacts clinical trial design, results


Estimated Read Time:

4 minutes

Simon R. Heller MD, FRCP
Simon R. Heller, MD, FRCP

Hypoglycemia is a major barrier to achieving glucose targets that prevent diabetes complications, but there’s no consensus whether 54 mg/dL or 70 mg/dL is the best standard for hypoglycemia in clinical trials.

Three experts addressed this issue during the Scientific Sessions symposium Reduction of Hypoglycemia by New Therapies in Clinical Trials—Overestimated, Underestimated, or Both? The presentation can be viewed by registered meeting attendees at through September 10, 2020. If you haven’t registered for the Virtual 80th Scientific Sessions, register today to access all of the valuable meeting content.

Simon R. Heller, MD, FRCP, Professor of Clinical Diabetes at the University of Sheffield, United Kingdom, said different guidelines and hypoglycemia classifications cause confusion among researchers and clinicians while also hampering education efforts. Dr. Heller made the case that internationally agreed upon definitions should use a glucose level of less than 54 mg/dL because that threshold identifies episodes that indicate increased risk and also guides regulators toward identifying the most effective and cost-effective treatments.

The 2017 position statements from the ADA and the European Association for the Study of Diabetes (EASD) indicate that Level 1, at or below 70 mg/dL, serves as an alert value for patients and clinicians and may signal alterations in insulin type or dosage. Patients at this level should have their glucose level rechecked and are often asymptomatic.

Level 2, at or below 54 mg/dL, denotes impaired cognitive function. Repeated episodes at this level cause reduced awareness and predict severe episodes, cardiac arrhythmias, and mortality.

Level 3, or severe hypoglycemia, does not have a specific blood glucose range. Dr. Heller suggested it would be defined as 20 mg/dL to 40 mg/dL.

Dr. Heller said that while the ADA/EASD-defined Level 1 holds high relevance as an alert to potential problems, there was little evidence that it impacted quality of life or had consequences on health economics.

“As a result of that, it was very difficult to persuade reimbursement authorities to make decisions based on clinical trial data that included only less than 70 mg/dL as a relevant level,” he said. “Indeed, those reimbursement authorities required differences to be shown in severe hypoglycemia.”

Aaron J. Kowalski, PhD
Aaron J. Kowalski, PhD

Aaron J. Kowalski, PhD, President and CEO of JDRF, provided a different perspective. Measuring time below 70 mg/dL offers clinically meaningful results, he said, capturing the short-term variations and daily challenges that A1C can’t. Dr. Kowalski did acknowledge that using the 70 mg/dL threshold requires more use of continuous glucose monitoring, which provides the time-in-range data critical to offer benefits.

“We need treatment options, as clinicians and people with diabetes, that are more representative of our ultimate goal, and our ultimate goal is to drive people into a more euglycemic range with less work,” Dr. Kowalski said. “We need to drive to time-in-range—or ranges—that represent both the short-term and long-term risk of diabetes. This provides a more physiologic target for diabetes management. It provides the best representation of diabetes glycemic health.”

Peter A. Senior, MBBS, PhD, Professor of Medicine and Director of the Division of Endocrinology at the University of Alberta, said the constraints of trial design determine hypoglycemia outcomes.

Peter A. Senior, MBBS, PhD
Peter A. Senior, MBBS, PhD

Regulators have significant influence in the design and execution of trials and what the endpoints should be, he said. Regulators set a common template used for clinical trial design, which leads industry to consider what can be approved and what is necessary for clinical trials.

Clinical trials studying hypoglycemia often underestimate benefit, Dr. Senior said, because they use low-risk populations, conservative targets and titration algorithms, blinded treat-to-target designs, and rigid application of statistics to pre-defined hypotheses.

“Many trials, particularly of insulin therapies, have often excluded people with hypoglycemia or a history of hypoglycemia unawareness for ethical reasons of safety, potentially influencing the rate of hypoglycemia detected in the trial,” Dr. Senior said. “And if you have a population at low risk, the ability to show a difference between two treatments will be much lower because of statistical power.”



Already registered?
View this presentation at

Not yet registered?
Register now to access all presentations from the Virtual 80th Scientific Sessions.