View More 2022, Scientific Sessions, Session Coverage

Automated insulin delivery systems require manual adjustments for exercise, high-fat meals


Estimated Read Time:

5 minutes

Boris Kovatchev, PhD
Boris Kovatchev, PhD

Using automated insulin delivery (AID) systems to control glycemic levels makes life easier for patients with diabetes, but that doesn’t mean they are infallible. Exercise and eating high-fat meals require patients to pay extra attention to the devices and make adjustments.

A panel of experts looked at the differences in various types of AID on the market and ways to educate patients about personally tailoring the systems to their lifestyles during the session, Clinical Pearls in the Practical Use of Automated Insulin Delivery Systems, on Friday, June 3. The session was livestreamed and can be viewed on-demand by registered meeting attendees at If you haven’t registered for the 82nd Scientific Sessions, register today to access the valuable meeting content.

Boris Kovatchev, PhD, Director, University of Virginia Center for Diabetes Technology, spoke about the benefits and limitations of five closed-loop systems that have U.S. Food and Drug Administration approval or a CE Mark: OmniPod, Control IQ, CamAPS FX, and both the MiniMed 670G and 780G.

Based on the largest real-life closed-loop database available to date, patients using closed-loop systems typically increase their time in range by 10% in the first month of use over sensor-based therapies. But then they plateau.

The database also indicates that, when stratified by baseline glycemic control, all age groups improve uniformly, and bolus behavior changes immediately after system initiation in all subgroups.

David Maahs, MD, PhD
David Maahs, MD, PhD

“So there’s one unresolved problem with all the systems that we’re reviewing,” Dr. Kovatchev said. “We know that the effect is instantaneous. We know that it’s lasting for up to a year or even more. We don’t know why there is no improvement after the first month of use.”

He proposed that a treatment plan combining drug and device therapies could be the key to better outcomes with fully automated closed-loop systems.

David Maahs, MD, PhD, the Lucile Salter Packard Professor of Pediatrics, Division Chief of Pediatric Endocrinology, and Associate Chair for Academic Affairs in Pediatrics, Stanford University and the Lucile Packard Children’s Hospital, stressed that diabetes education needs to be continually adjusted as new AID systems become available.

Dr. Maahs said that because the systems can be adjusted, physicians should consider what targets they ask their patients to aim for and how they teach about hypoglycemia. Dr. Maahs said he tries to start all his newly diagnosed patients on continuous glucose monitoring (CGM) within the first month and teach them how to use a pump at one to three months.

“This does require a little bit extra upfront work, but we think that there’ll be benefits both to the patient and the diabetes team as patients get started on these systems early and are able to achieve better care and tighter control early on,” he said.

Laurel Messer, PhD, RN, CDCES
Laurel Messer, PhD, RN, CDCES

Research needs to begin looking at interoperability—combining one type of system with a different pump, a different CGM, and a different algorithm, Dr. Maahs added.

Laurel Messer, PhD, RN, CDCES, Assistant Professor of Pediatrics, Barbara Davis Center, University of Colorado School of Medicine, Anschutz Medical Campus, discussed strategies to prevent exercise-induced dysglycemia while using open- and closed-loop systems.

She pointed to research published in The Lancet Diabetes & Endocrinology that recommends reducing basal insulin 90 minutes before exercise and suspending pump use at the start of exercise, but for no longer than 60 minutes. After exercise, patients should reduce bolus for the first meal and either reduce overnight basal insulin or consume a bedtime snack without insulin, she said.

The literature also highlighted how aerobic exercise can lead to decreased glucose within 45 minutes from onset, how mixing aerobic exercise with intervals offers better glucose stability, and how purely anaerobic exercise produces increased counterregulatory hormone output, which can cause a rise in glucose.

These are arbitrary, Dr. Messer noted, because all exercise is a spectrum between aerobic and anaerobic activity and the primary dysglycemia issue with exercise is hypoglycemia across the spectrum.

Stuart A. Weinzimer, MD
Stuart A. Weinzimer, MD

She also cited studies showing that reducing bolus is nearly identical for closed- and open-loop systems.

“Exercise management is trial and error and, unfortunately, always will be as long as we have exogenous insulin that we’re trying to manipulate,” Dr. Messer said. “It’s most important to reduce insulin on board all around, and closed-loop exercise targets are the most common feature utilized to prevent exercise-induced hypoglycemia.”

Stuart A. Weinzimer, MD, Professor of Pediatrics, Yale School of Medicine, addressed how to tackle hyperglycemia associated with high-protein and high-fat meals. He discussed several studies comparing consumption of meals with various amounts of fat and protein but the same number of carbohydrates. They showed the profound impact that fat and protein have on glucose levels.

Physicians must know the capabilities of automated insulin delivery systems that their patients are using, as some of them allow for extended boluses and late boluses rather than just late increases in basal insulin delivery, Dr. Weinzimer noted.

“Three-hour post-meal correction for prolonged highs for most of my patients, that’s the job,” he said. “The superiority of extended boluses versus simple correction in closed-loop systems has really not been established yet. And food choices still matter. Even with all these automation systems, it’s advisable not to eat a lot of high-fat meals, and that’s good for cardiovascular health as well.”