Continuous Glucose Monitoring in Healthy Children Aged 2

Abstract

Background:

The purpose of this study was to add the missing information on glycemic levels and patterns as measured by continuous glucose monitoring (CGM) in the daily life of healthy children aged 2–8 years. These data are needed when studying glycemic patterns and treatment outcome in children aged 2–8 years with diabetes.

Methods:

Each of the 15 healthy children aged 2–7.99 years used a CGM device (Dexcom G4 Platinum) for 7 days.

Results:

A total of 15 children (10 girls) aged 5.4 ± 1.6 years registered a mean of 1976 ± 15 counts. Mean sensor glucose was 5.3 ± 1.0 mmol/L (95 ± 18 mg/dL) and 89% of values were in the range 4–7.8 mmol/L (72–140 mg/dL), 9% of sensor glucose values were <4.0 mmol/L (72 mg/dL), and 2% of sensor glucose values were >7.8 mmol/L (140 mg/dL).

Conclusion:

We present glycemic data as measured by CGM in healthy children aged 2–8 years.

Introduction

Children with diabetes need ongoing insulin treatment. Treatment of children with diabetes aims to avoid long-term and short-term complications. To achieve this, it is important to strive for normoglycemia from the time of diabetes onset.¹ Subcutaneous continuous glucose monitoring (CGM) is used to evaluate, monitor, and guide insulin treatment.¹ Normal glycemic patterns have been described in older children² and adults.^2
–4 CGM studies have also been performed in older children⁵ and teenagers⁶ to compare the difference in glycemic pattern between young persons with and without multiple diabetes-related autoantibodies.

The purpose of this study was to add the missing information on glycemic levels and patterns as measured by CGM in the daily life of healthy children aged 2–8 years.

Methods

The study was conducted at The Queen Silvia Children's Hospital, Gothenburg, Sweden, after approval by the regional ethical board. This study was initiated, designed, performed, analyzed, and reported here by the authors with no influence from the sponsors.

Siblings of insulin-treated children with diabetes using CGM were invited to participate in this study. Parents were given information about the study both in written and verbal forms. Children were then given verbal information by their parents, and the parents were invited to ask their children whether they wanted to participate in the study. Parents provided written informed consent on behalf of their children.

Inclusion criteria were age 2–7.99 years, body mass index-standard deviation score (SDS) ±2SDS according to Swedish growth charts,⁷ no illness and on no medication, HbA1c <42 mmol/mol (<6.0%), fasting plasma glucose <5.6 mmol/L (<100 mg/dL), and a postprandial 2-h plasma glucose <7.8 mmol/L (<140 mg/dL) after a challenge with a carbohydrate-rich meal consisting of pancakes and jam. The carbohydrate content of the meal was at least 1.75 g carbohydrates per kilogram body weight. Fasting and postprandial glucose values were measured with HemoCue 201 RT (HemoCue AB, Ängelholm, Sweden).

Height and weight of each child were measured at the diabetes clinic by a research nurse. A child health questionnaire was filled in by the parent before glycemic screening.

HbA1c was analyzed with DCA Vantage (Siemens Healthcare Diagnostics, Inc., Tarrytown, NY) and quality controlled in accordance with Equalis (External quality assurance in laboratory medicine in Sweden, www.equalis.se).

Subjects used a Dexcom G4 (Dexcom, Inc., San Diego, CA) for CGM for 7 days. After anesthetic with topical lidocaine (EMLA; AstraZeneca PLC, London, UK), the sensor was applied by a trained research nurse.

The sensor was calibrated by the parents in accordance with instructions from the manufacturer. Capillary blood samples (“finger pricks”) were used for CGM calibration. A plasma glucose monitor (14 Contour; Bayer Consumer Care AG, Basel, Switzerland and one Freestyle lite; Abbott Diabetes Care, Inc., Alameda, CA) was used for calibration of the CGM system. After the CGM registration, one parent filled in a report on behalf of the child to exclude data from days with illness or medication (especially fever, vomiting, or intake of paracetamol).

The monitoring period was up to 166 h (mean 165 h or 6.9 days).

The mathematical formulas of the assessment methods for glucose variability were taken directly from the article by Hill et al.³ and estimated in the EasyGV^© open source computer program developed by the Nuffield Department of Primary Care Health Sciences at the University of Oxford.⁸ The EasyGV computer program was used to calculate 10 different measures of glycemic variability:

M value, mean amplitude of glycemic excursions, lability index, average daily risk ratio, J index, low blood glucose index and high blood glucose index, continuous overlapping net glycemic action, mean of daily differences, glycemic risk assessment in diabetes equation, and mean absolute glucose.

The article by Hill et al.³ explains in detail the measures and mathematical formulas and we applied the default settings in EasyGV when calculating the glycemic variability measures. We used the sample standard deviation (SD) to calculate the variability of the mean glucose and the glycemic variability measures already listed.

In addition, we calculated the mean glucose, SD of the mean, and the distribution of glucose levels for the entire study population stratified by time of day in three intervals (00:00–08:00, 08:00–16:00, and 16:00–24:00).

Results

Each of the 15 participating children (Table 1) provided valid data from more than 6 days. No data had to be excluded due to concurrent illness or medication. Sensor glucose levels were mainly within 4–7.8 mmol/L (72–140 mg/dL) and rarely <3.5 mmol/L (63 mg/dL) or >9.0 mmol/L (162 mg/dL) (Table 2). Zero percent of sensor glucose values was >11.1 mmol/L (200 mg/dL). Three children provided more low sensor glucose values than the rest of the group. The glycemic pattern was very stable (Table 3). Sensor glucose levels were slightly higher in the evening and lower in the morning.

Table 1.

Study Participants

N = 15	Mean	SD
Gender (girls 10: boys 5)
Age (years)	5.4	1.6
Height (cm)	117	11
Weight (kg)	22	4.2
BMI-SDS	0.18	0.86
HbA1c, IFCC (mmol/mol)	32	2.5
HbA1c, NGSP (%)	5.1	0.23
Fasting plasma glucose (mmol/L)	4.9	0.42
Fasting plasma glucose (mg/dL)	89	7.6
2-h postprandial glucose (mmol/L)	5.8	0.75
2-h postprandial glucose (mg/dL)	104	13

BMI, body mass index; IFCC, International Federation of Clinical Chemistry; NGSP, National Glycohemoglobin Standardization Programme; SDS, standard deviation score.

Table 2.

Sensor Glucose Levels at Different Periods of the Day

	Overall	00:00–08:00	08:00–16:00	16:00–24:00
Mean sensor glucose (mmol/L)	5.3	4.7	5.4	5.7
±SD (mmol/L)^*	1.0	0.8	0.9	1.0
Mean sensor glucose (mg/dL)	95	83	97	103
±SD (mg/dL)^*	18	18	16	18
Distribution of sensor glucose levels (%)^†
4–7.8 mmol/L (72–140 mg/dL)	89	81	91	93
<4 mmol/L (<72 mg/dL)	9	18	7	4
<3.5 mmol/L (<63 mg/dL)	3	6	2	1
<3 mmol/L (<54 mg/dL)	1	2	0	0
>7 mmol/L (>126 mg/dL)	6	2	6	8
>7.8 mmol/L (>140 mg/dL)	2	1	2	3
>9 mmol/L (>162 mg/dL)	1	0	1	1
>11.1 mmol/L (>198 mg/dL)	0	0	0	0
AUC for glucose <4 mmol/L (mmol^min/L)^*	93.1
AUC for glucose >7.8 mmol/L (mmol^min/L)^*	70.6

SD is calculated as sample SD.

†

Observe that the distribution of sensor glucose levels does not sum to 100% due to overlap in ranges.

AUC calculated by the trapezoidal rule.

AUC, area under the curve.

Table 3.

Different Measurements of Glycemic Variability

n = 15	Mean	SD
CGM count	1976	15
Mean glucose (mmol/L)	5.3	0.4
Mean glucose (mg/dL)	95	0.42
SD	1.0	0.2
CONGA	4.7	0.3
LI	0.9	0.4
J index	12.9	1.9
LBGI	3.0	1.2
HBGI	0.8	0.6
GRADE	0.6	0.2
MODD	0.9	0.1
MAGE	1.5	1.0
ADRR	2.8	1.8
M value	4.2	2.4
MAG	1.9	0.2

ADRR, average daily risk ratio; CGM, continuous glucose monitor; CONGA, continuous overlapping net glycemic action; GRADE, glycemic risk assessment in diabetes equation; HBGI, high blood glucose index; LBGI, low blood glucose index; LI, lability index; MODD, mean of daily differences; MAG, mean absolute glucose; MAGE, mean amplitude of glycemic excursions.

Conclusion

Our data describe glycemia as measured by subcutaneous CGM in healthy children aged 2–8 years.

Insulin treatment of preschool children with diabetes aims to avoid long-term and short-term complications. To achieve this, it is important to strive for normoglycemia from the time of diabetes onset.¹ Hopefully the tools to fine-tune insulin treatment in young children will become more accurate, which means that we must know the normal glucose levels in children. Using age-appropriate data regarding glycemic levels and patterns is also of importance for evaluating abnormal glucose levels from any cause in young children and for evaluating children at risk for diabetes.

We have not found any previously published data on healthy children in the age group 2–7.99 years to compare our results with. Three studies describe CGM data in healthy older children and adolescents.^2,5,6

The participants in the Steck et al.⁶ study were on average 14 years old and thus probably most of them pubertal, which also is probable regarding a large proportion of the participants in the JDRF² study. The on average 10-year-old cases and controls in the Helminen et al.⁵ study were sex and age matched, allowing for a difference of ±1 year of age. No information is available on pubertal status regarding the participants in the Helminen et al.⁵ study. Experience from insulin treatment of children⁹ shows that glycemic patterns and insulin need differ with age and maturity during childhood and adolescence, and this impairs comparisons of our data with the JDRF,² Steck et al.,⁶ and Helminen et al.⁵ studies.

Steck et al.⁶ and Helminen et al.⁵ both designed their studies for exploring CGM as a tool for early detection of dysglycemia in children with known autoantibody positivity, which indicates an extremely high risk of developing clinical diabetes within 10–15 years. A total of 5 of 14 autoantibody positive participants in the Steck et al.⁶ study developed overt diabetes within 6 months from the studied CGM registration.

The descriptive study of JDRF² used older versions of CGM (Guardian Clinical and Freestyle Navigator) and reports that children aged 8–15 years had 1.8% of sensor glucose values ≤3.9 mmol/L and 1.3% of sensor glucose values ≥7.8 mmol/L with a glucose variability of 0.9 mmol/L. Helminen et al.⁵ used Dexcom G4 Platinum⁵ and reports 0.4% versus 5.6% of values above 7.8 mmol/L while Steck et al.⁶ used Dexcom seven plus and reports 9% versus 18% when comparing time spent with sensor glucose >7.8 mmol/L when comparing participants without or with multiple autoantibodies. Helminen et al.⁵ show that the autoantibody positive participants have higher glucose values in the evening than the autoantibody negative participants.

Neither Steck et al.⁶ nor Helminen et al.⁵ present results on sensor glucose values <3.9 mmol/L.

Research in very young children raises ethical considerations affecting the study design. Siblings of children using CGM for monitoring glycemia in insulin treatment were chosen as study subjects since this group of children and their parents already know what CGM is and how the related procedures are performed. This improves the child's ability to accept or reject participation in the study. Another consideration is that the parents are already skilled in performing CGM in a child-friendly manner.

For ethical reasons, there was an active decision to minimize the number of invasive procedures, thus only one sensor was used per child. The number of low plasma glucose values was unevenly distributed, with three children having more values <4.0 mmol/L than the rest of children. With this study design, it is impossible to distinguish between outlying values caused by physiological variations, sensor quality, and registration-related problems such as calibrations. This design limitation is shared with all the mentioned CGM studies in older children.^2,5,6

Of the 15 participants, 13 were siblings of children with type 1 diabetes, 2 were siblings of one child with another rare cause of insulin-treated diabetes (status postpancreatectomy). Siblings of children with type 1 diabetes have a higher risk of getting diabetes than other children.^10,11 The ethical decision to limit the number of painful procedures also led to the decision not to measure diabetes-related autoantibodies, which is an obvious limitation of the study. In contrast, all families were followed at our diabetes outpatient clinic and, for 18 ± 4.4 months of mean follow-up time, none has developed diabetes so far.

Footnotes

Acknowledgments

The authors thank study nurse Agneta Johansson for practical work with the continuous glucose monitoring (CGM) registrations, Jens Olsen and Sandra Stallknecht at Incentive (www.incentive.dk) for statistical analyses, the Swedish Childhood Diabetes Foundation for economical support, and Dexcom for providing equipment (sensors, transmitters, and monitors) for CGM.

Author Disclosure Statement

The authors (F.S. and G.F.) both state that no competing financial interests exist.

References

Sundberg

, Barnard

, Cato

, et al.: Managing diabetes in preschool children. Pediatr Diabetes, 2017; 18:499–517.

JDRF (Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group): Variation of interstitial glucose measurements assessed by continuous glucose monitors in healthy, nondiabetic individuals. Diabetes Care, 2010; 33:1297–1299.

Hill

, Oliver

, Choudhary

, et al.: Normal reference range for mean tissue glucose and glycemic variability derived from continuous glucose monitoring for subjects without diabetes in different ethnic groups. Diabetes Technol Ther, 2011; 13:921–928.

Zhou

, Li

, Ran

, et al.: Establishment of normal reference ranges for glycemic variability in Chinese subjects using continuous glucose monitoring. Med Sci Monit, 2011; 17:CR9–CR13.

Helminen

, Pokka

, Tossavainen

, et al.: Continuous glucose monitoring and HbA1c in the evaluation of glucose metabolism in children at high risk for type 1 diabetes mellitus. Diabetes Res Clin Pract, 2016; 120:89–96.

Steck

, Dong

, Taki

, et al.: Early hyperglycemia detected by continuous glucose monitoring in children at risk for type 1 diabetes. Diabetes Care, 2014; 37:2031–2033.

, Albertsson-Wikland

, Karlberg

: Population-based body mass index reference values from Goteborg, Sweden: birth to 18 years of age. Acta Pediatr, 2000; 89:582–592.

University of Oxford, Nuffield Department of Primary Care Health Sciences: EasyGV. www.phc.ox.ac.uk/research/technology-outputs/easygv (accessed April 19, 2017 ).

DiMeglio

, Boyd

, Pottorff

, et al.: Preschoolers are not miniature adolescents: a comparison of insulin pump doses in two groups of children with type 1 diabetes mellitus. J Pediatr Endocrinol Metab, 2004; 17:865–870.

10.

Dahlquist

, Blom

, Tuvemo

, et al.: The Swedish childhood diabetes study - Results from a nine-year case register and a one-year case-referent study indicating that type 1 (insulin-dependent) diabetes mellitus is associated with both type 2 (non-insulin-dependent) diabetes mellitus and autoimmune disorders. Diabetologia, 1989; 32:2–6.

11.

Gillespie

, Gale

EAM

, Bingley

: High familial risk and genetic susceptibility in early onset childhood diabetes. Diabetes, 2002; 51:210–214.

Continuous Glucose Monitoring in Healthy Children Aged 2–8 Years