Original Research
Analysis of Prevalence, Magnitude and Timing of the Dawn Phenomenon in Adults and Adolescents With Type 1 Diabetes: Descriptive Analysis of 2 Insulin Pump Trials

https://doi.org/10.1016/j.jcjd.2019.08.003Get rights and content

Abstract

Objectives

To better understand the dawn phenomenon in type 1 diabetes, we sought to determine its prevalence, timing and magnitude in studies specifically designed to assess basal insulin requirements in patients using insulin pumps.

Methods

Thirty-three participants from 2 sensor-augmented insulin pump studies were analyzed. Twenty participants were obtained from a methodologically ideal semiautomated basal analysis trial in which basal rates were determined from repeated fasting tests (the derivation set) and 13 from an artificial pancreas trial in which duration of fasting was variable (the “confirmation” set). Prevalence was determined for the total cohort and for individual trials using the standard definition of an increase in insulin exceeding 20% and lasting ≥90 minutes. Among cases, time of onset and percent change in the magnitude of basal delivery were determined.

Results

Seventeen participants (52%) experienced the dawn phenomenon (11 of 20 [55%] in the derivation set and 6 of 13 [46%] in the confirmation set). Time of onset was 3 AM (interquartile range [IQR], 3 to 4:15 AM) in the derivation set and 3 AM (IQR, 3 to 4 AM) in the confirmation set. The magnitude of the dawn phenomenon was a 58.1% (IQR, 28.8% to 110.6%) increase in insulin requirements in the derivation set and 65.5% (IQR, 45.6% to 87.4%) in the confirmation set.

Conclusions

The dawn phenomenon occurs in approximately half of patients with type 1 diabetes; when present, it has predictable timing of onset (generally 3 AM) and a substantial, but highly variable, magnitude. These findings imply that optimization of glycemic control requires clinical emphasis on fasted overnight basal insulin assessment.

Résumé

Objectifs

Pour mieux comprendre le phénomène de l'aube dans le diabète de type 1, nous avons cherché à déterminer la prévalence, le moment et l'ampleur à partir d’études conçues particulièrement pour évaluer les besoins en insuline basale chez les patients qui se servaient de pompes à insuline.

Méthodes

L'analyse a porté sur 33 participants de 2 études sur les pompes à insuline reliées à un capteur. Vingt participants provenaient d'un essai sur les analyses semi-automatisées de l'insuline basale idéal sur le plan méthodologique dans lequel les taux d'insuline basale étaient déterminés à partir des épreuves à jeun répétées (la cohorte de dérivation), et 13 patients provenaient d'un essai sur le pancréas artificiel dans lequel la durée du jeûne était variable (la cohorte de validation). Nous avons déterminé la prévalence pour la cohorte entière et pour les essais individuels à l'aide de la définition normalisée d'une augmentation de l'insuline excédant 20 % et durant ≥90 minutes. Parmi les cas, nous avons déterminé le moment de survenue et la différence en pourcentage de l'ampleur de la distribution de l'insuline basale.

Résultats

Dix-sept participants (52 %) ont subi le phénomène de l'aube (11 sur 20 [55 %] dans la cohorte de dérivation, et 6 sur 13 [46 %] dans la cohorte de validation). Le moment de survenue a été observé à 3 h (intervalle interquartile [IIQ], de 3 h à 4 h 15) dans la cohorte de dérivation et à 3 h (IIQ, de 3 h à 4 h) dans la cohorte de validation. L'ampleur du phénomène de l'aube a montré une augmentation de 58,1 % (IIQ, de 28,8 % à 110,6 %) des besoins en insuline dans la cohorte de dérivation et de 65,5 % (IIQ, de 45,6 % à 87,4 %) dans la cohorte de validation.

Conclusions

Le phénomène de l'aube survient chez environ la moitié des patients atteints du diabète de type 1. Lorsqu'il est présent, son moment de survenue est prévisible (généralement à 3 h), et son ampleur est importante, mais varie grandement. Ces résultats signifient que l'optimisation de la régulation de la glycémie exige qu'on accorde une importance particulière sur le plan clinique à l’évaluation de l'insuline basale à jeun pendant la nuit.

Introduction

The dawn phenomenon is a major cause of morning hyperglycemia in patients with type 1 diabetes 1, 2, 3. Its treatment requires sufficient basal insulin replacement to suppress an overnight fasted increase in blood glucose levels (4). Basal insulin is the amount of insulin required to maintain stable glucose levels and prevent excess ketone production during a fasted state 5, 6. This contrasts with food bolus (or “prandial”) insulin, which maintains stable glucose levels during meals, and correction bolus insulin, which lowers glucose levels from hyperglycemia to a target range. The dawn phenomenon has also been observed in patients with type 2 diabetes with and without treatment by insulin 7, 8.

The increase in overnight fasting basal insulin requirements due to the dawn phenomenon is generally felt by patients and practitioners to be paradoxical. The counterintuitive nature of the dawn phenomenon combined with a fear of hypoglycemia leads to many patients not receiving sufficient overnight basal insulin 9, 10. Supported by studies that demonstrate suppression by somatostatin infusion, the pathogenesis of the dawn phenomenon is fundamentally believed to relate to hepatic and peripheral insulin resistance induced by the nocturnal physiological peak of growth hormone 11, 12, 13, 14. Quantitatively, the existence of clinically relevant dawn phenomenon in those with diabetes has been defined as an increase in insulin requirements by >20% lasting at least 90 minutes during overnight hours 15, 16.

When present, the dawn phenomenon has been found to be correlated with glycemic instability; other early morning glucose excursions, including those after the first meal; and higher glycated hemoglobin 1, 17. Thus, longstanding early morning hyperglycemia has major implications for overall glycemic control and the development of diabetic complications. Morning hyperglycemia can be appropriately resolved by continuous subcutaneous insulin infusion via “open-loop” insulin pump therapy—which does not require continuous glucose monitoring—because basal insulin doses can be precisely preprogrammed to reflect physiological requirements 18, 19, 20, 21. However, standardized guidelines on how to adjust the dose and timing of hourly basal insulin rates to treat the dawn phenomenon are lacking. This is in part due to the difficulty of ascertaining patients' overnight basal insulin requirements, which is done through a fasting process referred to as basal assessment. In clinical practice, basal assessments are not systematically performed owing to the cumbersome nature of the evaluation. The test requires the patient to have a low-fat dinner at 5 PM containing a known amount of carbohydrate; avoidance of any further food or bolus insulin until breakfast; avoidance of exercise, caffeine and alcohol; and repeated capillary blood glucose testing or the use of continuous glucose monitoring (CGM) for examination of changes in glucose concentration. Rises in glucose concentration indicate insufficient basal insulin, whereas decreases represent excess basal insulin.

Observational studies have demonstrated high intraindividual variability in overnight basal insulin requirements (22), suggesting that, for certain patients, a single basal insulin profile is insufficient. Fully automated closed-loop insulin delivery systems offer a solution to account for the dawn phenomenon and this day-to-day variability. However, superior efficacy of these closed-loop systems has not been proven (23), and widespread implementation of these systems into routine clinical care and practice may not be imminent. To better manage multiple daily injections and conventional open-loop insulin pump therapy, we must gain better insight into the descriptive characteristics of the dawn phenomenon.

Two previous trials in type 1 diabetes conducted by our group evaluated algorithms that performed either real-time, or delayed adjustments to patients' insulin infusion based on analysis of their CGM to optimize glycemic control 24, 25. Insulin data from these studies provided calculated estimates of these patients' true physiological insulin requirements, thus offering a unique opportunity to perform a detailed study of the dawn phenomenon. Using one of these studies as a derivation set and the other as a confirmation set, we performed a retrospective, descriptive analysis of the prevalence, timing and magnitude of the dawn phenomenon in type 1 diabetes using sensor-augmented insulin pump therapy to help guide clinicians in making appropriate adjustments to patients' insulin regimens to avoid morning hyperglycemia and optimize glycemic control.

Section snippets

Study design

This study is a retrospective descriptive secondary analysis of 2 independent clinical trials that were previously used to evaluate the efficacy of 2 different algorithm-based insulin dose calculators. The first study, entitled “Evaluation of a Clinical Tool to Test and Adjust the Programmed Overnight Basal Profiles for Insulin Pump Therapy: A Pilot Study,” and subsequently referred to as the Basal Rate Analyzer Trial (BRAT), evaluated 20 adult participants with type 1 diabetes (24). The second

Results

Table 1 shows the baseline characteristics of all participants in the derivation set and the confirmation set. In the derivation set, there were 11 dawn phenomenon cases, and 9 controls---dawn phenomenon prevalence was 55%. In the confirmation set, there were 6 dawn phenomenon cases and 7 controls---dawn nonprevalence was 46%. Although we did not notice any substantial differences between cases and controls in both the derivation and confirmation set, there was a longer duration of diabetes

Discussion

In this study, we have sought to determine the prevalence, timing and magnitude of the dawn phenomenon in 33 participants from 2 studies designed to evaluate overnight insulin requirements in patients receiving sensor-augmented insulin pump therapy. Twenty participants were taken from a methodologically ideal trial in which basal insulin requirements were determined using 5 iterative fasting overnight assessments, and 13 additional participants were from an artificial pancreas trial in which

Acknowledgments

The study was supported by the Charles Hollenberg Summer Studentship Program (to I.O.), run by the Banting and Best Diabetes Centre at the University of Toronto. The BRAT study was funded by the JDRF Canadian Clinical Trials Network (CCTN Grant 80-2010-585). The JDRF CCTN is a public‒private partnership, including JDRF International, JDRF Canada and the Federal Economic Development Agency for Southern Ontario. For the CLASS04 study, the authors acknowledge Diabetes Canada (Grant

References (29)

  • D.R. Owens et al.

    Basal insulin analogues in the management of diabetes mellitus: What progress have we made?

    Diabetes Metab Res Rev

    (2014)
  • F. Porcellati et al.

    Thirty years of research on the dawn phenomenon: Lessons to optimize blood glucose control in diabetes

    Diabetes Care

    (2013)
  • L. Monnier et al.

    Magnitude of the dawn phenomenon and its impact on the overall glucose exposure in type 2 diabetes: Is this of concern?

    Diabetes Care

    (2013)
  • P. Aschner et al.

    Practical steps to improving the management of type 1 diabetes: Recommendations from the Global Partnership for Effective Diabetes Management

    Int J Clin Pract

    (2010)
  • Cited by (7)

    View all citing articles on Scopus
    View full text