Highlights
· CGM significantly improved hemoglobin A1c (HbA1c) compared with self-monitoring blood glucose in Korean children and adolescents with type 1 diabetes mellitus. Sustained HbA1c reduction over 12 months highlights continuous glucose monitoring's real-world effectiveness and potential to enhance long-term glycemic control.
Introduction
Type 1 diabetes mellitus (T1DM) presents significant challenges in children and adolescent populations due to the complexity of blood glucose management and the risk of both acute and chronic complications. Especially hemoglobin A1c (HbA1c) levels can become higher during adolescence, potentially due to physiologic insulin resistance [1]. Continuous glucose monitoring (CGM) technology has emerged as a promising tool to improve glycemic control by providing real-time feedback on glucose levels. Korean CGM users have been expected to increase with reimbursement of CGM sensors for T1DM patients by the Korean National Health Insurance Service since 2019 [2,3]. Recent studies have demonstrated that CGM can significantly improve glycemic outcomes by increasing time in range (TIR) and reducing HbA1c levels in adults [4].
Implementation of CGM in the management of T1DM has shown promising results in various clinical settings. Sanderson et al. [5] reported significant improvements in glycemic outcomes among children with T1DM using real-world data from a population-based clinic. Furthermore, CGM has been associated with a reduced risk of severe hypoglycemia and diabetic ketoacidosis (DKA) in children, adolescents, and young adults with T1DM, as demonstrated by Karges et al. [6] in a comprehensive population-based study. The efficacy of CGM in improving glycemic control has been attributed to its ability to provide continuous, real-time glucose data and a standardized report called an ambulatory glucose profile [2,7,8].
However, the adoption and consistent use of CGM technology are not without challenges. Hilliard et al. [9] identified several barriers to CGM use in young children with T1DM, including device fatigue, discomfort, and adherence issues. These factors underscore the importance of structured education programs and ongoing support to maximize the benefits of CGM technology [10].
In the Korean context, an increasing trend in CGM usage has been observed among children and adolescents with T1DM [2,3], but there was no study that evaluated the effects of CGM on glycemic control in Korean pediatric populations in real-world settings.
This study aims to evaluate the effect of CGM on glycemic outcomes in Korean children and adolescents with T1DM and to examine changes in CGM metrics of TIR and hypoglycemia in a real-world setting over a 12-month period based on ambulatory glycemic profiles (AGPs).
Materials and methods
1. Study design and subjects
This retrospective study was conducted at Korea University hospitals from January 2019 to May 2024. The study included 66 Korean children and adolescents diagnosed with T1DM, divided into a CGM group (n=22) and a self-monitoring blood glucose (SMBG) group (n=44) as matched controls based on age, sex, and disease duration. All participants in the CGM group initiated CGM use at least 6 months after diagnosis and performed SMBG until CGM initiation. Patients who initiated CGM within 6 months of diagnosis were excluded due to the potential influence of the honeymoon period.
2. Data collection
Data were collected every 3 months and included demographic, clinical, and AGP data. Demographic and clinical data were age; sex; and HbA1c levels at diagnosis, at initiation of the study, and every 3 months thereafter for 1 year. HbA1c at diagnosis was defined as the initial HbA1c level measured at the time of T1DM diagnosis, while HbA1c at baseline was the HbA1c level at the start of CGM or SMBG monitoring. Also we included the time from diagnosis to the initiation of study, insulin delivery method, and history of DKA and severe symptomatic hypoglycemia accompanied by loss of consciousness or hypoglycemic seizure after diagnosis of T1DM. AGP data for the CGM group were CGM activation time, mean glucose, glucose management indicator (GMI), coefficient of variation (CV), TIR, and time in hypoglycemia, separated into durations for glucose levels of 55–70 mg/dL and <55 mg/dL. CGM data were collected using the Dexcom G6 system. Patients were instructed to maintain a consistent sensor wear schedule throughout the study period.
3. Outcome measures
The primary outcome was the change in HbA1c levels from baseline to 3, 6, 9, and 12 months in the CGM and SMBG groups. Secondary outcomes were the percentage of patients achieving an HbA1c level less than 7.0% and the percentage of patients with reductions in HbA1c of 0.5% and 1.0% or more at 3, 6, 9, and 12 months. Glycemic metrics in the AGP, including CGM activation time, mean glucose, CV, TIR, and hypoglycemia, were monitored at 3, 6, and 12 months in the CGM group. We analyzed duration of hypoglycemia as 'mild hypoglycemia' (glucose 55–70 mg/dL) and 'severe hypoglycemia' (glucose <55 mg/dL).
4. Statistical analysis
Statistical analyses were performed using IBM SPSS Statistics ver. 27.0 (IBM Co., USA). Continuous variables were presented as mean±standard deviation, and categorical variables were presented as frequencies and percentages. Independent t-test, chi-square test, and repeated measures analysis of covariance (ANCOVA) were used to compare continuous variables between the CGM and SMBG groups and to assess changes in glycemic metrics over time within the CGM group. A P-value less than 0.05 was considered statistically significant.
Results
1. Demographic and clinical characteristics
The baseline characteristics of the study participants are summarized in Table 1. The CGM group had a mean age of 16.63 years and the SMBG group had a mean age of 17.85 years. There were no significant differences in age and sex between CGM and SMBG users. The mean HbA1c level at diagnosis was 13.33%±1.90% in the CGM group and 13.28%±2.32% in the SMBG group (P=0.930). Time from diagnosis to initiation of the study was 4.19±3.77 years and 5.19±3.90 years, respectively (P=0.326), and baseline HbA1c measurements were conducted between January 2019 and April 2023 for all patients in both groups, with the groups assessed within the same study period. Only one patient in the CGM group used an insulin pump, and the others received insulin through multiple daily injections. Neither DKA nor symptomatic severe hypoglycemia occurred during the 1 year of glycemic monitoring.
2. Glycemic outcomes
Table 2 presents primary and secondary glycemic outcomes at baseline and 3, 6, 9, and 12 months. The primary outcomes showed that HbA1c level decreased from 8.68% at baseline to 7.58% at 3 months, 7.70% at 6 months, 7.88% at 9 months, and 7.92% at 12 months in the CGM group (P=0.004, P=0.034, P=0.045, P=0.011, respectively). But HbA1c level increased from 8.44% at baseline to 9.07% at 3 months and 9.13% at 6 months and then decreased to 8.96% at 9 months and 8.93% at 12 months in the SMBG group (P<0.001, P<0.001, P<0.001, P<0.001, respectively) (Fig. 1). The changes in HbA1c from baseline to 3, 6, 9, and 12 months all showed significant difference between the groups: -1.11 vs. 0.61 at 3 months, -0.99 vs. 0.67 at 6 months, -0.80 vs. 0.49 at 9 months, and -0.76 vs. 0.47 at 12 months (P<0.001, P<0.001, P<0.001, P=0.001, respectively).
The secondary outcomes showed baseline HbA1c <7.0% in 13.6% of both the CGM and SMBG groups, and 27.3%, 31.8%, 22.7%, and 31.8% of the CGM group achieved an HbA1c <7.0% at 3, 6, 9 and 12 months, respectively, compared to 0%, 4.5%, 0%, and 2.3% at 3,6, 9, and 12 months in the SMBG group. The difference was significant (P<0.001, P=0.005, P=0.003, P=0.001).
The percentage of patients with reduction in HbA1c ≥0.5% was 72.7% at 3 months, 59.1% at 6 months, 54.5% at 9 months, and 59.1% at 12 months in the CGM group, respectively, compared to 15.9% at 3 months and 6 months, 20.5% at 9 months, and 18.2% at 12 months in the SMBG group (P<0.001, P<0.001, P=0.006, and P=0.001). The percentage of patients with reductions in HbA1c ≥ 1.0% was significantly higher in the CGM group compared to the SMBG group at all time points: 59.1% vs. 9.1% at 3 months, 50.0% vs. 9.1% at 6 months, 36.4% vs. 11.4% at 9 months, and 40.9% vs. 6.8% at 12 months (P<0.001, P<0.001, P=0.021, and P=0.001).
Table 3 presents the results of repeated measures ANCOVA, with a significant effect of group on HbA1c reduction (F=6.139, P=0.016) after adjusting for sex, age, insulin delivery method, and time from diagnosis to initiation of the study. However, other covariates, including sex (P=0.341), age (P=0.471), insulin delivery method (P=0.475), and time from diagnosis to the initiation of study (P=0.349), were not significantly associated with HbA1c changes.
3. Variations in glycemic metrics in the AGP
The glycemic metrics in AGP, including CGM activation time, mean glucose, CV, and range (TIR), were monitored over 3, 6, and 12 months, as shown in Table 4.
Although there were significant improvements in HbA1c, there were no significant changes in TIR, mean glucose, CV, and hypoglycemia over time in the CGM group. Meanwhile, CGM activation time showed a slight decrease, from 89.09% at 3 months to 81.40% at 6 months and 79.24% at 1 year of CGM application (P=0.090).
Mean glucose was 183.9 mg/dL at 3 months, 189.3 mg/dL at 6 months, and 192.6 mg/dL at 12 months (P=0.336), and the GMI was 7.69%, 7.83%, and 8.01%, respectively (P=0.072). TIR was 52.73%, 51.68%, and 48.41%, respectively (P=0.124), and CV was 39.71%, 39.45%, and 40.72%, which did not change significantly (P=0.365). Time spent in mild hypoglycemia was 2.14% at 3 months, 2.05% at 6 months, and 2.0% at 1 year of CGM application (P=0.851). The number of patients experiencing mild hypoglycemia for 5% or more of the time was 2, 4, and 1, respectively (P=0.327). Time spent in severe hypoglycemia was 0.36% at 3 months, 0.64% at 6 months, and 1.47% at 1 year (P=0.406), while the number of patients with severe hypoglycemia for 1% or more of the time was 4, 6, and 5, respectively (P=0.772).
Discussion
This is the first study to demonstrate that use of CGM in Korean children and adolescents with T1DM is associated with significant improvements in HbA1c levels compared to SMBG in the real-world settings. HbA1c levels at both diagnosis and baseline were comparable between the CGM and SMBG groups, minimizing the potential effect of initial metabolic differences on subsequent glycemic outcomes. Time from diagnosis to study initiation was not significantly different between the 2 groups, and baseline HbA1c measurements were conducted within the same time period.
Despite this similarity at baseline, the comparison of glycemic outcomes showed significant decreases in HbA1c from baseline in the CGM group, with a mean reduction of 1.11% at 3 months, 0.99% at 6 months, 0.8% at 9 months, and 0.76% at 12 months. In contrast, the SMBG group experienced an increase in HbA1c levels. The sustained improvement in HbA1c levels over 12 months in our study is particularly noteworthy, as it suggests that the benefits of CGM persist beyond the initial adoption period.
Also, the analysis of glycemic outcomes further underscores the effectiveness of CGM in achieving and maintaining better glycemic control. A significantly higher percentage of patients in the CGM group achieved an HbA1c level less than 7.0% at 3, 6, 9, and 12 months compared to the SMBG group. Moreover, a larger proportion of CGM users experienced clinically significant reductions in HbA1c (≥0.5% and ≥1.0%) compared to SMBG users. These findings align with previous research highlighting the benefits of CGM in improving glycemic control and reducing HbA1c levels by providing real-time glucose feedback in adults [11-13]. The ability of CGM to facilitate such substantial improvements in glycemic control is particularly important for pediatric populations, as it may help reduce the risk of long-term complications associated with T1DM [14,15].
After adjusting for potential confounders including sex, age, insulin delivery method, and time from diagnosis to initiation of the study, the CGM group continued to exhibit significantly lower HbA1c levels compared to the SMBG group. This highlights the independent effect of CGM in improving long-term glycemic outcomes.
By providing real-time glucose monitoring, CGM offers patients the ability to make informed decisions regarding insulin dosing, exercise, and diet [7,8]. This continuous feedback loop likely contributes to sustained improvement in glycemic outcomes, particularly during the first year of CGM use [12,16]. The significant reductions in HbA1c observed in our study support the integration of CGM into routine clinical practice for better management of T1DM in younger populations, as recommended by recent clinical practice guidelines [17,18].
Despite significant improvements in HbA1c, glycemic metrics in AGP showed unexpected outcomes. Unlike a previous study that showed significant improvements in TIR [19], in our study, CGM activation time, mean glucose, GMI, CV, and TIR showed no significant change over time. Meanwhile, CGM activation time showed a decreasing trend, which suggests it as the reason for the increasing trend in GMI and no significant changes in TIR and mean glucose. Patients might have experienced difficulties in consistently using the device, which could be due to factors such as device fatigue, discomfort, or lack of adherence, despite the initial benefits of CGM [9]. So, maintaining long-term glucose control may require additional interventions, such as support for CGM use and enhanced education on CGM interpretation and consistent use [10,13], which can maximize the full benefits of CGM technology.
The CV is an important metric for assessing glycemic variability. A lower CV indicates more stable blood glucose levels, while a higher CV indicates greater variability. CGM use was reported previously to significantly reduce CV compared to that without CGM [20], but that in the present study remained relatively stable and the mean CV almost reached the target of 36% [19]. This suggests that, while CGM use did not significantly reduce glycemic variability, it might have helped maintain a stable level of variability around the target over time.
Some previous studies reported a significant decrease in hypoglycemia [6,16,21,22], while others did not demonstrate such change [23,24]. Interestingly, analysis of hypoglycemia outcomes showed no significant differences over time in our study. We thought these results might be due to the small sample size and slight decreasing trend in CGM activation time mentioned earlier. Our results regarding CGM metrics suggest that its use is not sufficient for improving all aspects of glycemic control. This underscores the importance of structured education programs and regular adjustments to the diabetes management plan, which can help patients interpret CGM data and adjust their insulin regimens accordingly [10].
This study has some limitations that should be acknowledged. While we were able to demonstrate improvements in HbA1c, the relatively small sample size may have limited the detection of significant changes in other glycemic metrics such as TIR, CGM activation time, and hypoglycemia. Also, we did not assess the impact of other factors, such as patient education, psychological support, or socioeconomic factors, which may influence adherence to CGM use.
Future studies with larger, more diverse cohorts and longer follow-up periods are needed to fully assess the long-term impact of CGM on glycemic control. Additionally, evaluating the role of structured educational programs in enhancing CGM adherence and outcomes would be beneficial. Interventions designed to improve patient engagement with CGM and to address the psychosocial barriers to its consistent use are critical areas for future investigation.
In conclusion, CGM use in Korean children and adolescent populations with T1DM significantly improves HbA1c levels over 12 months compared to SMBG. The implementation of CGM may provide valuable benefits in glycemic control and potentially reduce the risk of diabetes-related complications.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Print
