Overnight Counter-Regulatory Hormone Levels and Next Day Glycemia in Adults with Type 1 Diabetes During Closed-Loop Insulin Delivery Versus Sensor-Augmented Pump with Low-Glucose Suspend

Dear Editor:

When overnight use of a closed-loop system at home is compared with sensor-augmented pump (SAP), an improvement in nocturnal glycemia has been reported that may extend into the following day upon reversion to open loop in both adults ¹ and adolescents. ^1,2 Recently, our group published the results of a study involving 16 adults and 12 adolescents with type 1 diabetes, in which participants in random order completed four consecutive nights at home using an Android-based hybrid closed-loop system (Android-HCLS) and four nights using an SAP with low-glucose suspend (SAP-LGS). ³ The protocol (ClinicalTrials.gov ID: NCT02040571) has been previously described. ⁴ The combined group, as well as adult and adolescent subgroups, experienced a reduction in glycemic variability and nocturnal hypoglycemia. However, there were no significant differences in percentage time in target range (72–144 mg/dL) overnight for the group as a whole or the adolescent subgroup, nor were there any differences during the next day (data not shown). In contrast, the adult subgroup using Android-HCLS overnight spent a greater time in target range than those using SAP-LGS, ³ and this persisted into the first half of the following day between 08:00 and 13:00 postreversion to open-loop insulin delivery (59 [23]% vs. 45 [29]%; P = 0.046). There was no benefit in adults in the late afternoon (percentage time in target range between 13:00 and 17:00 after overnight Android-HCLS compared with that in overnight SAP-LGS of 49 [18]% versus 43 [21]%; P = 0.219.

Our prestudy hypothesis was that differences in next morning glycemia after overnight Android-HCLS and SAP-LGS may be accounted for by decreased excursions of overnight counter-regulatory hormones secondary to reductions in glycemic variability and hypoglycemia. Plasma catecholamines, metanephrines, cortisol, glucagon, and growth hormone were measured at 08:00 on days 1 and 5 (on the morning before the first night and on the morning after the last night of each study phase). In the adult group, urinary catecholamines and cortisol were measured on 12-h timed urine collections performed overnight (20:00–08:00) on the first two nights during Android-HCLS and SAP-LGS. Catecholamines were measured by high-performance liquid chromatography, metanephrines by liquid chromatography-tandem mass spectrometry, glucagon by radioimmunoassay, and cortisol and growth hormone by automated immunometric chemiluminescence. Counter-regulatory hormone levels during the study were quantified, with all of a subject's samples in the same analytical run, with intraassay CVs <7%. Nonparametric Wilcoxon signed rank test was used to evaluate the differences in these hormones.

Overnight closed loop was not associated with a significant reduction in counter-regulatory hormone levels. No differences were observed between days 1 and 5 plasma counter-regulatory hormone levels or the average of days 1–2 and days 2–3 overnight urine collections, when comparing overnight Android-HCLS with SAP-LGS in the adult (Table 1), adolescent, or combined groups (data not shown).

The absence of any differences in counter-regulatory hormones between Android-HCLS and SAP-LGS phases may be explained by the fact that, although statistically significant, the absolute differences observed in percentage time spent in the hypoglycemic range overnight were very low during both closed-loop and open-loop study phases without any episodes of severe nocturnal hypoglycemia. ³ Nevertheless, in the adult group, although there were no differences in counter-regulatory hormones measured, the improved percentage time in target range overnight with closed loop use did have a positive impact upon glycemic control in the first half of the next day. It is possible that starting the next day with stable glucose levels in target range had an independent beneficial impact of limited duration that extended to lunchtime.

We acknowledge the limitations of this analysis, including the short study duration and limited number of blood samples per subject. Frequent overnight sampling would have provided a more detailed assessment, especially given the pulsatile nature of counter-regulatory hormone secretion, which may partly explain the very wide standard deviations reported. ⁵ However, frequent overnight venous sampling would have required admission of the participants to a clinical trial center, which would have likely affected sleep patterns and counter-regulatory hormone production. Blunting of the counter-regulatory hormone response with antecedent hypoglycemia may also have impacted the study findings. ^6,7

In conclusion, our hypothesis was not supported in this group of well-controlled adult patients. It is unlikely that the improvement observed in next day glycemia after overnight closed loop in adults was related to differences in the overnight counter-regulatory hormone response. Additional studies of longer duration in a less well-controlled group, incorporating a different comparator, may reveal different outcomes.