William V. Tamborlane, MD, Professor and Chief of Pediatric Endocrinology, Yale School of Medicine, New Haven, Connecticut, addressed factors affecting hypoglycemia in children with type 1 diabetes at the 76th Scientific Sessions of the American Diabetes Association, June 10-14, in New Orleans. Dr. Tamborlane began the session by asserting that counterregulatory hormone responses to hypoglycemia differ between children and adults in some ways. In other ways, children and adults respond similarly to hypoglycemia.
Similarities in Counterregulatory Hormone Responses to Hypoglycemia between Children and Adults with Type 1 Diabetes
In children and adults with type 1 diabetes:
• Plasma glucagon responses to hypoglycemia are lost early in the disease.
• Both children and adults rely primarily on catecholamine responses to counteract insulin and induce symptoms.
Differences in Epinephrine Responses in Nondiabetic Children and Adults
In 1991 and 1995, Timothy W. Jones et al reported the epinephrine response to hypoglycemia was earlier and higher in healthy nondiabetic children than in nondiabetic adults. This difference was exaggerated in poorly controlled adolescents with type 1 diabetes.
The 1994 Diabetes Control and Complications Trial (DCCT)
In the 1994 DCCT, being an adolescent with type 1 diabetes was an independent risk factor for serious hypoglycemia with intensive treatment, despite this cohort's higher hemoglobin A1c levels than adults. Dr. Tamborlane hypothesized that hyper- and hypoglycemia in adolescents in the trial resulted from a perfect storm of the insulin resistance of puberty meeting the delayed and prolonged pharmacokinetic and pharmacodynamics properties of large doses of human regular insulin, especially during the overnight period.
In 1991, Stephanie A. Amiel and colleagues observed that the insulin resistance of puberty is characterized by decreased stimulation of peripheral insulin uptake, despite normal suppression of hepatic glucose production. Dr. Tamborlane noted that adolescents with type 1 diabetes require large (that is, 0.2 – 0.3 U/kg) premeal bolus doses of rapid-acting insulin to overcome the peripheral insulin resistance of puberty.
A delayed peak in the glucose infusion rate occurs with regular insulin (the only fast acting insulin available at the time), and this causes post-meal glucose to rise. This prolonged duration of glucose infusion leads to reduced hepatic glucose production as well as to late post-meal hypoglycemia. In adolescents, the greatest susceptibility to hypoglycemia is after bedtime, at least 6 hours after the last meal bolus of the day.
Overshoot hyperinsulinemia is a particular problem at night during sleep because:
• Sympathetic activity is decreased during sleep.
• Epinephrine responses to falling glucose levels are markedly blunted during sleep.
The Diabetes Research in Children Network (DirecNet) study demonstrated that the risk of nocturnal hypoglycemia is increased after antecedent exercise in the afternoon in children and adolescents.
The DirecNet Study
This 2007 study assessed 50 adolescents with type 1 diabetes on a sedentary day and one with exercise. Median patient age was 14.8 ± 1.7 years. Median type 1 diabetes duration was 7.0 ± 3.7 years. Median hemoglobin A1c was 7.8 ± 0.8%. The exercise protocol was:
• Exercise at 3-4 pm to simulate after-school activity,
• Four periods of brisk walking to target heart rate for 15 minutes, plus 5 minutes of rest,
Twenty-eight percent (28%) of patients experienced nocturnal hypoglycemia on nights following sedentary days, and this increased to 48% on nights following afternoon exercise (P=.009).
The group from Perth, Australia, used the glucose clamp technique to show that glucose requirements needed to maintain euglycemia after moderate-intensity afternoon exercise rose 7-11 hours post exercise.
Managing Young Children with Type 1 Diabetes
Dr. Tamborlane noted that the goal of managing young children with type 1 diabetes had been to prevent moderate to severe hypoglycemia at all costs by setting higher glucose and hemoglobin A1c targets to prevent abnormal brain growth and development. He pointed out that the premises that higher hemoglobin A1c lowers the risk of hypoglycemia, and that hyperglycemia does not adversely affect brain development are false.
DirecNet Examination of Neuroanatomical and Cognitive Development in Young Children with Type 1 Diabetes
In this 2015 DirecNet study, the impact of dysglycemia on neuroanatomical growth and cognitive development was prospectively examined in young children with type 1 diabetes diagnosed at an early age.
A total of 144 children with type 1 diabetes and 72 controls without the disease were studied. Median age was 4- to under 10-years. Onset of type 1 diabetes was observed at >6 months of age, with positive islet cell antibodies.
Baseline and 18-month assessments of cognitive function were performed using measurements of intelligence quotient, memory, and executive function. Voxel-based morphology, diffusion tensor imaging, and functional MRI were performed as well. Every 3 months, hemoglobin A1c and continuous glucose monitoring profiles were examined.
Type 1 diabetic and control subjects differed with respect to total and regional gray and white matter volumes and brain growth. Data from continuous glucose monitoring suggested that chronic hyperglycemia is detrimental to the developing brain.
Impact of Increased Use of Insulin Analogs, Closed System II, and Continuous Glucose Monitoring on Risks of Hypoglycemia
Unlike in the DCCT, better control of type 1 diabetes can now be achieved in children and adults without increasing the frequency of severe hypoglycemic events. Nevertheless, no type 1 diabetes treatment will eliminate the risk of hypoglycemia unless insulin delivery is controlled in a feedback mechanism, especially in the overnight period.
Effect of Closed-Loop Glucose Control on Exercise-Associated Hypoglycemia
In this 2013 trial by Jennifer L. Sherr and colleagues, subjects were admitted to the clinical research center on two separate occasions: 2 days of closed-loop and 2 days of open-loop control.
During each admission, one day was an exercise day and the other a sedentary day. Reference blood glucose levels were drawn every half hour. The hypoglycemia was defined as a blood glucose level ≤60 mg/dL. The exercise protocol was exercise at 3-4 pm to simulate after-school activity. Subjects walked briskly to target heart rate for 15 minutes and rested for 5 minutes.
Closed-loop control had a marked impact in reducing overnight hypoglycemia on both sedentary and exercise days. Because residual insulin was present, however, closed-loop control did not prevent hypoglycemia events during exercise (five vs six events with closed- vs open-loop control, respectively). In a DirecNet study, even complete suspension of the pump prior to starting exercise could not prevent hypoglycemia.
Snacking was assessed to prevent hypoglycemia with closed-loop insulin delivery. The hypothesis was that a proactive snacking strategy during closed-loop control would provide a simple and effective means to decrease the risk of hypoglycemia during exercise without causing post-exercise hyperglycemia.
A randomized crossover design (closed-loop alone vs closed-loop with proactive snacking) was employed to assess differences in the change in blood glucose levels during exercise between the two study conditions. Twelve subjects performed a treadmill exercise with heart rate targeted to achieve 50% to 55% maximal effort. Gatorade quantity depended on blood glucose prior to and midway through exercise.
On non-snacking days, delays in automatically shutting off basal insulin delivery resulted in a >60 mg/dL fall in blood glucose levels during exercise. In contrast, the drop in blood glucose was virtually eliminated by simple snacking. Late, post-exercise hyperglycemia was not observed.
Dr. Tamborlane concluded, "The data suggest that closed-loop insulin delivery systems for unsupervised home use offers the potential to allow patients with type 1 diabetes to improve overall glycemic control, while simultaneously reducing or even eliminating the risk of severe hypoglycemia."
"The bottom line," he added, "is that we need feedback control of insulin delivery to really put a dent into the risk of severe hypoglycemia. That's why the recent, rapid progress toward an artificial pancreas that is practical for home use is so important."
Tamborlane WV. Session: Hypoglycemia in Special Situations in Diabetes. 1-AC-SY03 - Factors Affecting Hypoglycemia in Children. ADA 2016: 76th Scientific Sessions of the American Diabetes Association, June 10-14, New Orleans, LA.
W.V. Tamborlane: Consultant; Speaker; AstraZeneca LP, Halozyme Therapeutics, Novo Nordisk, Inc., sanofi-aventis, Takeda Pharmaceutical Company, Ltd.