Abstract
Aims
Early detection and management of gestational diabetes mellitus (GDM) benefits both mother and infants. We aimed to compare continuous glucose monitoring (CGM) to routine self‐monitoring of blood glucose (SMBG) in a randomised controlled trial on glycaemic management and pregnancy outcomes in women with GDM diagnosed in gestational Weeks 8–26.
Materials and Methods
From October 2021 to December 2023, 128 women with GDM were enrolled and 120 completed the trial after 2:1 randomisation to either the intervention (Dexcom G6 CGM, n = 80) or the control group (SMBG 4×/day and monthly blinded CGM [Dexcom G6 Pro], n = 40). Both groups received lifestyle and medication advice based on the CGM interpretation, clinical follow‐up and glucose monitoring until delivery. CGM metrics were evaluated with the automated glucose data analysis (AGATA) toolbox (https://github.com/gcappon/agata) and maternal and neonatal outcomes were collected from medical records. Participants completed the CGM Satisfaction survey prior to delivery.
Results
While no significant differences between groups in time spent in the glucose target range (TIR) 63–140 mg/dL (primary outcome) were found (p = 0.37), the intervention group experienced significantly lower unscheduled caesarean section rates (20.0% vs. 44.4%, p = 0.046, V = 0.19) and preterm deliveries (6.8% vs. 18.4%, p = 0.041, V = 0.19). Neonates experienced lower large‐for‐gestational‐age rates (5.0% vs. 18.4%, p = 0.019, V = 0.22) and neonatal intensive care unit admissions (22.5% vs. 44.7%, p = 0.013, V = 0.23) compared to the control group. Participants in the intervention group reported high CGM satisfaction and endorsed the behavioural and psychosocial benefits.
Conclusions
For women with early GDM, despite no improvement in TIR, CGM improved maternal and neonatal outcomes versus SMBG. The effect may be mediated by improved awareness of glycaemic excursions and by facilitation of treatment regimen adjustments.
Keywords: continuous glucose monitoring (CGM), gestational diabetes, patient reported outcomes, randomised trial
1. INTRODUCTION
Gestational diabetes mellitus (GDM) is defined as carbohydrate intolerance resulting in hyperglycaemia, with onset or first recognition during pregnancy, and is associated with significant complications in both mothers and infants. 1 Women with GDM have an increased risk of numerous complications of pregnancy, including an increased risk of developing preeclampsia, shoulder dystocia and the necessity of undergoing a caesarean delivery, as well as developing chronic diabetes mellitus (DM) and cardiovascular disease in subsequent years. 2 , 3 The neonatal outcomes attributed to maternal hyperglycaemia can be even more severe, including macrosomia, neonatal intensive care unit (NICU) stays, delayed foetal lung maturity, birth trauma, hypoglycaemia, hyperbilirubinemia and stillbirth. 2 , 3 The risk of complications in women with GDM increases proportionally with the worsening of glucose levels. Glucose monitoring and therapy are indispensable to prevent these complications as even small increases in maternal glucose are related to poorer clinical outcomes. Two noteworthy prospective randomised controlled trials (RCTs) have confirmed that detection and management of GDM carry benefits for both mother and baby in terms of immediate pregnancy outcomes. 4 , 5
Glucose monitoring targets in pregnancy need to be tight 6 , 7 with low thresholds for commencing pharmacotherapy or increasing insulin dosage. As such, frequent glucose monitoring is essential. Self‐monitoring of blood glucose (SMBG) is the most‐used method and standard of care to monitor glucose levels during pregnancy. However, the task is often neglected, and does not detect all glycaemic fluctuations, particularly those occurring overnight. Missing glucose data may contribute to suboptimal treatment decisions and outcomes.
The utilisation of continuous glucose monitoring (CGM) during pregnancy is preferred because it provides a detailed profile of fluctuations in glucose values for optimal glucose regulation and permits a better understanding of how glucose variability may influence acute and long‐term complications of diabetes. CGM offers a potential source of data required to improve the detection and management of glucose levels in pregnant women with diabetes. Recent attention has therefore focused on evaluating the role of CGM in pregnancy with studies providing normative data in pregnancies complicated by diabetes. 8 , 9 , 10 However, the effectiveness of CGM on improving pregnancy outcomes complicated by GDM is still understudied, with a limited number of currently published studies. 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 Therefore, this study aimed to investigate whether CGM improved glycaemic management, and maternal and foetal outcomes compared with SMBG‐blinded CGM use in women diagnosed with GDM.
2. MATERIALS AND METHODS
2.1. Study design and participants
The ‘Steady Sugar’ trial was a randomised single‐centre controlled trial assessing real‐time CGM (Intervention) versus SMBG‐blinded CGM use (Control) conducted at the Woman's Hospital in Baton Rouge, LA, USA. The study was registered at ClinicalTrials.gov (NCT04948112). One hundred twenty‐eight women with GDM diagnosed in gestational weeks >8 and <26 were enrolled and 120 completed the trial. Participants were diagnosed before 26 weeks of gestation using either a 2‐ or 3‐h oral glucose tolerance test (OGTT) following the American Diabetes Association (ADA) guidelines. 16
After diagnosis, participants were educated on diet, exercise and how to use SMBG following standard of care guidelines. 17 Participants were recruited from the Diabetes Clinic of Woman's Hospital, who had been referred by obstetric practices. The Woman's Hospital Institutional Review Board (WHIRB) approved both the protocol and the informed consent form.
Key inclusion criteria included female gender and >8 to <26 weeks gestation with a singleton pregnancy. Exclusion criteria included: (i) prior diagnosis of DM; (ii) presence of infection, significant systemic disease or other severe metabolic, endocrine or medical co‐morbidities; (iii) history of bariatric surgery or other surgeries that induce malabsorption; (iv) long‐term use (>2 weeks) of systemic steroids prior to enrollment; (v) multiple pregnancy; (vi) prior use of insulin before diagnosis of GDM; (vii) foetal growth restriction due to placental dysfunction at study entry; (viii) history of major depressive or other severe psychiatric disorders or inpatient psychiatric treatment up to 1 year before enrollment and (ix) inability or refusal to comply with the protocol or currently participating or having participated in a clinical trial in the previous 3 months.
2.2. Procedures
At baseline, maternal age, parity, history of GDM, family history, ethnicity, preconception diseases and obstetric history were recorded, along with height and weight. Participants received medical nutrition advice (isocaloric diet containing 40%–50% carbohydrate, 20% protein and 30%–35% fat) and physical activity recommendations (30 min/day). They were taught SMBG using glucose metres and were advised on capillary blood glucose measurement (fasting as well as 1 h after starting each meal).
At study entry, each participant was randomly allocated from a computer‐generated sequence to either the intervention (n = 87) or control (n = 41) group in a 2:1 ratio using a permuted block design (random block sizes of 3). Participants were randomised in a 2:1 ratio (CGM:SMBG), as we believed this would facilitate participant recruitment. The intervention group received antenatal care plus CGM (Dexcom G6; Dexcom, Inc., San Diego, CA) with recommended SMBG (2×/day), while the control group received antenatal care plus standard SMBG (4×/day) and blinded (10 days/month) CGM (Dexcom G6 Pro; Dexcom, Inc., San Diego, CA). The intervention group was allowed to view their CGM glucose concentration in real time or retrospectively via their smart phone and were encouraged to react appropriately. Glucose concentrations via CGM were not visible to participants in the control group either in real time or retrospectively (blinded CGM). The control group CGM was blinded via the CGM being set up via a receiver, which was then not provided to the participant. CGM and SMBG data were reviewed periodically (approximately every 10 days in the intervention group and at each study visit in the control group) and treatment plans were adjusted accordingly.
Follow‐up visits occurred monthly and included glucose data review, medication adjustments (if necessary), body weight measurement and routine ultrasounds (foetal biometry and umbilical artery Doppler, evaluated by medical staff). The maternal–foetal medicine clinicians managed all participants in the trial using glycaemic metrics derived from uncomplicated pregnancies targeting fasting glucose levels ≤90 mg/dL and 1 h postprandial ≤120 mg/dL to replicate the physiological glucose profile 18 as closely as possible.
Participants in the intervention group completed the CGM satisfaction scale (CGM‐SAT) 19 , 20 at the end of trial participation prior to delivery, a validated 44‐item questionnaire designed to measure the impact of using CGM on diabetes management and satisfaction with the emotional, behavioural and cognitive effects of CGM use over the past 6 months. Participants rated their experience with CGM on a scale of 1–5, with higher scores indicating a more favourable impact of and satisfaction with CGM use.
Maternal and neonatal outcomes were collected from medical records, including pregnancy outcomes and laboratory data. Gestational age at the time of sampling for OGTT and the time of delivery were calculated from the last menstrual period (LMP). Gestational age was changed to the date calculated from the earliest available ultrasound if different by >7 or 14 days in the first or second trimester, respectively, or if the LMP was unknown. Delivery mode was categorised as either repeat caesarean section (C‐section), failed induction with C‐section (primary C‐section) or vaginal delivery. No participants attempted vaginal delivery after a previous C‐section. Neonatal outcomes recorded included size for gestational age (defined by the 90th (LGA) and 10th (Small for gestational age [SGA]) percentiles for our population, respectively), hypoglycaemia (<47 mg/dL), NICU admission and Apgar scores at 1 and 5 min.
2.3. Outcomes
The primary objective of this study was to evaluate maternal glycaemic management in participants with GDM measured by CGM. Data were analysed using the automated glucose data analysis (AGATA) toolbox, 21 which computed customised glucose management metrics following the definitions recommended by an international consensus on CGM‐derived metrics for clinical trials involving pregnant women. 22 CGM metrics were calculated for the entire monitoring period, as well as for the daytime (06:00 AM to 23:59 PM) and nocturnal (00:00 to 05:59 AM) periods and were analysed across all data recorded during pregnancy. Our primary outcome was the percentage of time in the CGM glucose target range (TIR), defined as 63–140 mg/dL from enrollment to admission for delivery. Other CGM metrics can be seen in Tables 2 and S1.
TABLE 2.
Continuous glucose monitoring (CGM) outcomes.
| Parameter | Intervention (n = 80) | Control (n = 40) | p Value |
|---|---|---|---|
|
Time in range (TIR, 63–140 mg/dL) (%) |
88.6 [82.5–94.3] | 87.4 [79.1–93.9] | 0.37 |
|
Time in tight range (TITR, 63–120 mg/dL) (%) |
71.5 [63.1–81.4] | 70.2 [58.6–77.3] | 0.31 |
|
Time above range (TAR, >140 mg/dL) (%) |
10.4 [5.3–17.1] | 11.0 [5.2–20.6] | 0.52 |
|
Time below range (TBR, <63 mg/dL) (%) |
0.6 [0.3–0.9] | 0.7 [0.2–1.3] | 0.85 |
| Hyperglycaemic events per week | 13.3 [8.6–18.9] | 10.5 [4.9–12.4] | 0.03 |
| Hypoglycaemic events per week | 1.3 [0.6–2.0] | 0.9 [0.3–2.2] | 0.22 |
| Mean glucose (mg/dL) | 109.1 [103.9–116.2] | 112.9 [105.3–119.4] | 0.25 |
| Coefficient of variation (CV) (%) | 21.5 ± 4.4 | 21.2 ± 4.1 | 0.73 |
| Proportion of participants with | |||
| ≥70% TIR (%) | 95.0 | 85.0 | 0.08 |
| <4% TBR (%) | 100 | 97.5 | 0.33 |
| <25% TAR (%) | 95.0 | 82.5 | 0.04 |
Note: Results are given over the entire CGM monitoring period as mean [95% confidence intervals], mean ± SD or categorical data as percentages. Bold = statistically significant (p < 0.05). A hypoglycaemic event was defined as >15 consecutive minutes of sensor glucose <63 mg/dL and a hyperglycaemic event was defined as >15 consecutive minutes of sensor glucose >140 mg/dL.
Abbreviations: SD, standard deviation; TAR, time above range; TBR, time below range; TIR, target range; TITR, time in tight range.
Perinatal and neonatal outcomes of interest included mode of delivery, preeclampsia, diabetes therapy (metformin, insulin with or without metformin) at the time of delivery, birth weight (g), gestational age at delivery, Apgar score, LGA rate, birth weight centile, 23 neonatal hypoglycaemia and NICU admission. Participant‐reported outcomes—the CGM‐SAT Benefits Subscale 20 total and individual item scores, were evaluated for the intervention group participants.
2.4. Statistical analysis
Statistical analyses were performed using SPSS version 29.0.1. Personnel who performed the statistical analyses were not blinded to participant grouping. Data were analysed according to randomisation assignments. For CGM metrics, the Lilliefors test evaluated normality. For continuous metrics in primary CGM endpoints, a two‐sample t test was used if both samples were Gaussian; otherwise, the Mann–Whitney U‐test was applied. CGM endpoints are reported as means (95% confidence intervals (CI)) or means ± standard deviation (SD). To achieve 80% power at a two‐sided 5% significance level, we determined a sample size of 106 participants was required to detect a significant between‐group difference of at least 5% points in TIR, the primary outcome. To account for 20% attrition, the enrollment goal was 128 participants. For categorical metrics in secondary CGM endpoints (reported as percentages), the Fisher exact test determined group associations and the metric under study. Spearman correlation analysis was used to further explore the relationships between CGM metrics and pregnancy outcomes, such as gestational age at birth, birth weight and birth weight centile.
For maternal and neonatal outcomes, continuous variables were tested for normality using the Kolmogorov–Smirnov test and transformed if necessary. Continuous parameters are presented as means ± SD, and categorical data as counts and percentages. Descriptive statistics were used to summarise demographic data. The χ 2 test was used for comparisons of categorical variables, and one‐way analysis of variance was used for differences in mean continuous parameters between study groups. For all analyses, statistical significance was defined as p < 0.05. Where statistical significance was detected, appropriate effect sizes are reported (i.e., Cohen's d for continuous CGM endpoints, eta squared (η 2) for continuous maternal and neonatal outcomes, and Cramer's V for categorical variables). Standard interpretations of Cohen's d are: small effect, 0.20; medium effect, 0.50; large effect, 0.80; for η 2 are: small effect, 0.01; medium effect, 0.06; large effect, 0.14; and for Cramer's V are: small effect, 0.10; medium effect, 0.30; large effect, 0.50. CGM‐SAT Benefits Subscale total and item scores are presented as means ± SD.
3. RESULTS
3.1. Participants
Between October 2021 and December 2023, a total of 128 women with GDM diagnosed <26 weeks' gestation were enrolled in the study with the last participant completing the study in May 2024. Eight participants withdrew from the study: one in the control group (noncompliance) and seven in the intervention group (four noncompliant, two insurance changes and one adhesive skin reaction). Eighty intervention and 40 control participants had sufficient CGM data for analysis (Figure 1). Maternal baseline characteristics are shown in Table 1. Apart from a higher family history of diabetes in the control group (p = 0.03, V = 0.21), no significant baseline differences were found between groups. Family history of diabetes had no significant effect on outcome measures (p ≥ 0.19). Two participants (both in the control group) were not included in the analysis of delivery and pregnancy outcomes due to stillbirth (n = 1) and delivery occurring at an alternative hospital and inability to obtain delivery records (n = 1).
FIGURE 1.
Flow of participants in the trial. A total of 128 women with gestational diabetes mellitus diagnosed <26 weeks enrolled in the study with 120 completers and 8 participants withdrawn from the study (1 in the control group [noncompliance] and 7 in the intervention group) (4 noncompliant, 2 insurance changes and 1 adhesive skin reaction). CGM, continuous glucose monitoring; RT‐CGM, realtime continuous glucose monitoring; SMBG, self‐monitoring of blood glucose.
TABLE 1.
Baseline characteristics.
| Parameter | Intervention (n = 80) | Control (n = 40) | p Value |
|---|---|---|---|
| Age, years (mean ± SD) | 33.1 ± 5.4 | 31 ± 5.9 | 0.50 |
| Race/ethnicity, n (%) | |||
| White | 53 (66.0%) | 22 (55.0%) | 0.49 |
| Black/African American | 24 (30.0%) | 16 (40.0%) | |
| Other (Hispanic, Asian and Middle Eastern) | 3 (4.0%) | 2 (5.0%) | |
| Pre‐pregnancy BMI (kg/m2) (mean ± SD) | 36.0 ± 8.9 | 39.0 ± 6.9 | 0.06 |
| BMI at study start (kg/m2) (early second trimester) (mean ± SD) | 37.5 ± 8.6 | 41.0 ± 6.6 | 0.07 |
| Week of pregnancy for GDM diagnosis (mean ± SD) | 19.6 ± 4.6 | 19.2 ± 4.5 | 0.80 |
| Hypertension history, n (%) | 37 (46.0%) | 17 (43.0%) | 0.85 |
| Family history of diabetes, n (%) | 59 (75.0%) | 37 (92.5%) | 0.03 |
| Primiparous, n (%) | 25 (31.0%) | 9 (22.5%) | 0.32 |
| Nulliparous, n (%) | 55 (69.0%) | 31 (77.5%) | 0.42 |
| Previous diagnosis of GDM, n (%) | 38 (69.0%) | 20 (64.5%) | 0.55 |
| Baseline pharmacological treatment | |||
| Diet, n (%) | 40 (50.0%) | 23 (57.5%) | 0.27 |
| Diet and metformin, n (%) | 23 (29.0%) | 7 (17.5%) | |
| Diet and insulin, n (%) | 13 (16.0%) | 5 (12.5%) | |
| Diet, metformin and insulin, n (%) | 4 (5.0%) | 5 (12.5%) |
Note: Results are given as means ± SD or categorical data as counts and percentages. Bold = statistically significant (p < 0.05).
Abbreviation: GDM, gestational diabetes mellitus; SD, standard deviation.
3.2. CGM metrics
The proportion of CGM values in various glucose concentration ranges and the proportion of participants meeting TIR‐, time below range (TBR)‐ and time above range (TAR)‐based goals are given in Tables 2 and S1. No statistically significant differences occurred between groups for most CGM outcomes, including TIR (p = 0.37), with the proportion of participants achieving ≥70% TIR showing non‐significant (p = 0.08) trends favouring the intervention group. However, a significantly higher proportion of participants in the intervention group achieved <25% TAR (95.0% vs. 82.5%, p = 0.04, V = 0.20). The proportion of subjects with <4% TBR (night) was also significantly higher in the intervention group (100% vs. 92.5%, p = 0.035, V = 0.19). More hyperglycaemic episodes per week were able to be detected in the intervention group compared to the control group (13.30 [95% CI 8.60–18.90] vs. 10.45 [4.94–12.38], p = 0.03, d = 0.15), particularly in the daytime period (p = 0.028, d = 0.43), throughout pregnancy. No significant differences were found in hypoglycaemic events per week or in glucose variability measures (coefficient of variation [CV], SD) across the entire pregnancy (p > 0.05 for all comparisons).
3.3. Delivery features and pregnancy outcomes
Maternal delivery features and neonatal outcomes are summarised in Table 3. Intervention group participants were less likely to have a failed induction leading to a C‐section than the control group, where the C‐section was a planned repeat delivery (20.0% vs. 44.4%, p = 0.046, V = 0.19). Gestational age at delivery was significantly higher in the intervention group (38 ± 1 weeks vs. 37 ± 2 weeks, p = 0.015, η 2 = 0.05), with a lower preterm delivery rate (6.8% vs. 18.4%, p = 0.041, V = 0.22). The intervention group had a lower incidence of LGA infants (5.0% vs. 18.4%, p = 0.019, V = 0.22) and fewer NICU admissions (22.5% vs. 44.7%, p = 0.013, V = 0.23).
TABLE 3.
Maternal and neonatal outcomes.
| Parameter | Intervention (n = 80) | Control (n = 38) | p Value |
|---|---|---|---|
| Mode of delivery | |||
| Vaginal, n (%) | 34 (66.0%) | 20 (55.0%) | 0.33 |
| C‐section, n (%) | 46 (30.0%) | 18 (40.0%) | |
| C‐section type‐scheduled, n (%) | 37 (80.0%) | 10 (55.6%) | |
| Failed induction, n (%) | 9 (20.0%) | 8 (44.4%) | 0.046 |
| Pregnancy disorders | |||
| Preeclampsia, n (%) | 9 (11.2%) | 4 (10.5%) | 0.91 |
| Shoulder dystocia, n (%) | 7 (8.8%) | 3 (7.9%) | 0.88 |
| Pharmacological treatment at time of delivery | |||
| Diet, n (%) | 16 (20.0%) | 8 (21.0%) | 0.93 |
| Diet and metformin, n (%) | 12 (15.0%) | 6 (15.8%) | |
| Diet and insulin, n (%) | 30 (37.5%) | 12 (31.6%) | |
| Diet, metformin and insulin, n (%) | 22 (27.5%) | 12 (36%) | |
| Neonatal outcomes | |||
| Infant female sex, n (%) | 45 (56.0%) | 25 (66.0%) | 0.42 |
| Gestational age at birth, weeks (mean ± SD) | 38 ± 1 | 37 ± 2 | 0.015 |
| Birth weight (g) (mean ± SD) | 3173 ± 468 | 3216 ± 527 | 0.49 |
| LGA >90 percentile, n (%) | 4 (5.0%) | 7 (18.4%) | 0.019 |
| Apgar at 1 min (mean ± SD) | 7.71 ± 1.4 | 7.5 ± 1.6 | 0.17 |
| Apgar at 5 min (mean ± SD) | 8.8 ± 0.4 | 8.74 ± 0.45 | 0.14 |
| Neonatal hypoglycaemia, n (%) | 36 (46.0%) | 23 (65.0%) | 0.062 |
| Preterm delivery (<37 weeks), n (%) | 5 (6.8%) | 7 (18.4%) | 0.041 |
| NICU hospital stay, n (%) | 18 (22.5%) | 17 (44.7%) | 0.013 |
| Hyperbilirubinemia, n (%) | 9 (11.3%) | 3 (7.9%) | 0.57 |
Note: Results are given as means ± SD or categorical data as counts and percentages. Bold = statistically significant (p < 0.05). n = 38 in the control group due to n = 2 not being included in the analysis as a result of stillbirth (n = 1) and delivery occurring at an alternative hospital and inability to obtain delivery records (n = 1).
Abbreviation: C‐section, caesarean section; NICU, neonatal intensive care unit; SD, standard deviation.
3.4. Associations between CGM metrics and pregnancy outcomes
In the control group, significant negative correlations were found between glucose variability (as measured by SD and CV) and gestational age at birth (ρ = −0.35, p = 0.030 for SD; ρ = −0.35, p = 0.031 for CV). Conversely, SD and CV were positively correlated with birth weight centile (ρ = 0.34, p = 0.036 for SD; ρ = 0.32, p = 0.050 for CV). In the intervention group, mean glucose and TAR were negatively correlated with gestational age at birth (ρ = −0.25, p = 0.027 for mean glucose; ρ = −0.23, p = 0.043 for TAR).
When considering the entire cohort, mean glucose (ρ = −0.27, p = 0.004), SD (ρ = −0.23, p = 0.012) and TAR (ρ = −0.25, p = 0.006) were negatively correlated with gestational age at birth, while these metrics also exhibited positive correlations with birth weight centile (ρ = 0.24, p = 0.009 for both mean glucose and SD; ρ = 0.20, p = 0.027 for TAR). TIR and time in tight range (TITR) showed positive correlations with gestational age at birth (ρ = 0.23, p = 0.012 for TIR; ρ = 0.24, p = 0.009 for TITR) and negative correlations with birth weight centile (ρ = −0.21, p = 0.020 for TIR; ρ = −0.23, p = 0.014 for TITR).
3.5. Participant‐reported outcomes
Participants in the intervention group reported high satisfaction with CGM and strongly endorsed the benefits of CGM (CGM‐SAT Benefit Subscale total score: 4.3 ± 0.8, n = 79 with available CGM‐SAT data). The five highest‐scoring items in the CGM‐SAT Benefits Subscale included: Teaches me how eating affects blood sugar (4.7 ± 0.5), Has taught me new things about diabetes that I didn't know before (4.6 ± 0.6), Makes it clearer how some everyday habits affect blood sugar levels (4.6 ± 0.5), Helps me to be sure about making diabetes decisions (4.6 ± 0.8) and Shows patterns in blood sugars that we didn't see before (4.5 ± 0.6).
4. CONCLUSIONS
In this RCT, despite no significant difference in the primary endpoint, TIR, we found that CGM was associated with improvements in certain maternal and neonatal outcomes compared to SMBG.
We also found that CGM in pregnant women with GDM may be more effective in detecting hyperglycaemia and reducing nocturnal hypoglycaemia. Importantly, women in this study reported high satisfaction with CGM and strongly endorsed the behavioural and psychosocial benefits.
Consistent with previous work, 24 we hypothesise that CGM contributed positively to treatment effects due to clinicians having greater visibility of patients' glucose levels to moderate the strategies of their medication by altering dosage and frequency in response to glycaemic values. The information provided by CGM may have empowered the expectant mothers to make behavioural changes and participate in their treatment planning.
Although non‐significant differences in TIR were observed between the CGM and SMBG groups, the use of CGM in pregnant women with GDM was associated with significant improvements in secondary maternal and neonatal outcomes, suggesting enhanced management of glycaemic parameters beyond what is captured by TIR alone. We also found that CGM users had significantly lower rates of unplanned C‐sections and resulted in neonates having a significantly later gestational age at delivery. Furthermore, in comparing foetal outcomes, there were lower rates of LGA, preterm deliveries and NICU admissions with CGM use. These findings may have significant health‐economic implications. 25
Data from this study agrees with the findings of at least four prior studies 24 , 26 , 27 , 28 but contrasts with a recent report from Switzerland 11 in which the use of CGM did not improve perinatal outcomes among women with GDM. Key differences between the aforementioned 11 and the present study which may explain the contradictory findings include the recommended SMBG monitoring in the control groups (6× vs. 4×/day), mean BMI at diagnosis (25.8 vs. 37.0 kg/m2), and the notable differences in the prevalence of adverse pregnancy and neonatal outcomes (e.g., overall NICU admission 3% vs. 30%). Furthermore, our findings agree with a recent systematic review aimed at assessing the effectiveness of CGM compared with SMBG in managing glycaemic control in women with GDM. 26 CGM detected more hyperglycaemic and hypoglycaemic events, particularly nocturnal fluctuations, leading to the conclusion that CGM offers substantial advantages over SMBG for managing GDM by enabling timely treatment adjustments. 26
High glycaemic variability is detrimental to pregnancies complicated by diabetes. 29 Importantly, however, the present study highlights that the magnitude of glucose excursions may be quantified to a greater extent using CGM compared to SMBG alone. Unlike SMBG or intermittent CGM use, continuous monitoring provides a more comprehensive assessment of glycaemic patterns, capturing fluctuations that would otherwise remain undetected. In this study, participants in the intervention group wore CGM continuously from enrollment to delivery, whereas the control group used CGM for only 10 days/month, offering a limited view of glycaemic variability. As anticipated, continuous CGM identified a greater number of hyperglycaemic episodes; however, a higher proportion of women achieved <4% TBR overnight. These findings suggest the important value of continuous CGM in detecting clinically relevant glucose excursions and improving overall glycaemic assessment.
In agreement with our findings, Yu et al. 10 observed that CGM shortens the time spent in hypoglycaemia and Chen et al. 30 established that CGM detected hyper‐ and hypoglycaemic events better than SMBG in women with GDM. In another study, 31 CGM detected postprandial hyperglycaemia that was underestimated by SMBG. Furthermore, Yu et al. 10 reported that in a study with 340 women with GDM, CGM resulted in less glucose variability and better glycaemic control with the infants of mothers using CGM having significantly better outcomes compared with SMBG infants. The pregnant women using CGM had significantly lower rates of preeclampsia and C‐sections versus those using SMBG alone. 10 Taken together, the current evidence base advocates for the superior use of CGM over SMBG for glycaemic control in women with GDM.
Women in this study reported high satisfaction with CGM and strongly endorsed the benefits of using CGM. Individual CGM‐SAT items illustrated the impact of CGM on supporting knowledge about lifestyle glucose management behaviours, including how eating and exercise impact glucose. Participants agreed that CGM helped identify patterns in glucose, supporting both patients and providers in identifying optimal glycaemic management behaviours and approaches which often change throughout pregnancy. Additionally, participants agreed that CGM helped them feel more confident in their diabetes decisions, which indicates CGM's impact on peace of mind in this population who were managing glucose while navigating pregnancy, a high cognitive burden in a time already known to be stressful. Overall, the CGM‐SAT indicated high satisfaction alongside CGM's support of knowledge, confidence and peace of mind in the GDM population. In agreement with our data, Amylidi‐Mohr et al. 11 also reported high satisfaction with CGM use and a preference over SMBG in women with GDM.
The present study has several strengths. First, we prospectively randomised participants to treatment. We included many participants with GDM who were connected to a CGM sensor during the second and third trimesters of pregnancy, and the CGM‐derived metrics provided comprehensive profiling of glucose homeostasis. Second, we collected a variety of maternal and neonatal parameters, which enabled us to explore associations of CGM‐derived metrics with common and major adverse maternal and neonatal outcomes, adjusting for a wide range of potential confounders.
There are also several limitations. First, this study included predominantly Caucasian and African American participants from a single centre, and thus the findings may not be generalisable to other populations or ethnicities, although we applied the ADA criteria for GDM diagnosis, which increases our comparability with other GDM populations. Second, our study only included women with early GDM (without overt diabetes in pregnancy but who had been diagnosed early in pregnancy at less than 26 weeks) and did not have a true control group, as they received advice based on blinded CGM which would not have been part of the standard of care. This may have contributed to the lack of statistical significance in our primary outcome and some individual adverse pregnancy outcomes.
While TIR did not differ significantly between groups, the study may have been underpowered to detect subtle differences. Given the anecdotally low SMBG adherence observed (SMBG adherence not reported), CGM provided more complete glycaemic data, capturing postprandial spikes and nocturnal hypoglycaemia. Real‐time feedback enabled timely lifestyle and treatment adjustments, which likely contributed to improved secondary maternal and neonatal outcomes. We also did not consistently collect maternal weight gain in pregnancy or glycosolated hemoglobin levels, and we are therefore unable to report on the impact on these pre‐specified variables. Both groups received lifestyle and medication advice based on the CGM data, and this may have led to a lack of statistically significant between‐groups differences in CGM metrics. Finally, the CGM devices were connected to participants after GDM diagnosis, which may have potentially altered their dietary habits or lifestyle factors. Based on the above limitations and as we report on positive secondary outcomes which were not statistically powered a priori, a multicentre and large‐scale prospective clinical trial with a true control group (receiving only standard of care) is needed to further establish the efficacy and utility of CGM for all women with GDM to validate our findings.
In conclusion, while CGM did not improve the primary outcome TIR greater than SMBG, secondary outcomes from this RCT suggest CGM may improve maternal and neonatal outcomes. The effect may have been mediated by improved awareness of glycaemic excursions with CGM, and by facilitation of treatment regimen adjustments. These data support the use of CGM in all expectant mothers diagnosed with GDM.
AUTHOR CONTRIBUTIONS
KE‐H, MA and EV were involved in the conception, design and conduct of the study. MA, EV, SG and JB were responsible for clinical care and data collection. KE‐H, CG and JBW worked on the analysis and interpretation of the results. GC performed mathematical analyses of the CGM data, critically revised the manuscript and provided technical support. MAC, AM and HS performed all behavioural data analysis and contributed to the methods, results and discussion of this manuscript. KE‐H wrote the first draft of the manuscript, and CG, and JBW edited and reviewed the final version of the text. All the authors approved the final version of the manuscript. KE‐H is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
CONFLICT OF INTEREST STATEMENT
Funding and CGM supplies were provided by Dexcom. The study was funded by Dexcom as an Investigator Initiated Trial (IIT) during a period when the lead author was acting as an independent investigator. This trial was driven by the investigator's scientific curiosity and aimed to address gaps in existing medical knowledge, rather than pursue commercial objectives. The protocol was developed and the study conducted by Karen Elkind‐Hirsch and Edward Veillon at Woman's Hospital prior to Karen Elkind‐Hirsch's employment at Dexcom. Corbin Griffen, John B. Welsh, Margaret A. Crawford, Angelo Monroy and Harsimran Singh are employees of Dexcom and assisted in the writing and submission of this publication. During the conduct of this study, Karen Elkind‐Hirsch received grant support from Dexcom and Novo Nordisk and served on advisory boards for Eli Lilly and Novo Nordisk. No other potential conflicts of interest relevant to this article were reported.
Supporting information
Table S1. CGM outcomes over the daytime (06:00 AM to 23:59 PM) and nighttime (00:00 to 05:59 AM) periods.
ACKNOWLEDGEMENTS
The authors thank all of the staff of the Woman's Diabetes Center and Maternal Fetal Medicine Department for their unrelenting efforts in patient care and assistance with this trial. We thank the study participants without whom this investigation and these analyses would not have been possible.
Elkind‐Hirsch K, Armatta M, Griffen C, et al. Continuous glucose monitoring in early gestational diabetes improves maternal and neonatal outcomes—The Steady Sugar trial. Diabetes Obes Metab. 2026;28(1):691‐700. doi: 10.1111/dom.70254
This work was presented in part at the American Diabetes Association 85th Scientific Sessions, Chicago, IL, 20–23 June 2025.
DATA AVAILABILITY STATEMENT
Some or all datasets generated during and/or analysed during the current study are not publicly available but are available from the corresponding author on reasonable request.
REFERENCES
- 1. Metzger BE, Coustan DR. Summary and recommendations of the Fourth International Workshop‐Conference on Gestational Diabetes Mellitus. The Organizing Committee. Diabetes Care. 1998;21(suppl 2):B161‐B167. [PubMed] [Google Scholar]
- 2. Buchanan TA, Xiang AH, Page KA. Gestational diabetes mellitus: risks and management during and after pregnancy. Nat Rev Endocrinol. 2012;8:639‐649. doi: 10.1038/nrendo.2012.96 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Kim C. Maternal outcomes and follow‐up after gestational diabetes mellitus. Diabet Med. 2014;31:292‐301. doi: 10.1111/dme.12382 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med. 2005;352:2477‐2486. doi: 10.1056/NEJMoa042973 [DOI] [PubMed] [Google Scholar]
- 5. Landon MB, Spong CY, Thom E, et al. A multicenter, randomized trial of treatment for mild gestational diabetes. N Engl J Med. 2009;361:1339‐1348. doi: 10.1056/NEJMoa0902430 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. American Diabetes Association Professional Practice Committee . 15. Management of diabetes in pregnancy: standards of care in diabetes‐2025. Diabetes Care. 2025;48:S306‐S320. doi: 10.2337/dc25-S015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Metzger BE, Gabbe SG, Persson B, et al. International Association of Diabetes and Pregnancy study groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care. 2010;33:676‐682. doi: 10.2337/dc09-1848 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Wei Q, Sun Z, Yang Y, Yu H, Ding H, Wang S. Effect of a CGMS and SMBG on maternal and neonatal outcomes in gestational diabetes mellitus: a randomized controlled trial. Sci Rep. 2016;6:19920. doi: 10.1038/srep19920 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Voormolen DN, DeVries JH, Sanson RME, et al. Continuous glucose monitoring during diabetic pregnancy (GlucoMOMS): a multicentre randomized controlled trial. Diabetes Obes Metab. 2018;20:1894‐1902. doi: 10.1111/dom.13310 [DOI] [PubMed] [Google Scholar]
- 10. Yu F, Lv L, Liang Z, et al. Continuous glucose monitoring effects on maternal glycemic control and pregnancy outcomes in patients with gestational diabetes mellitus: a prospective cohort study. J Clin Endocrinol Metab. 2014;99:4674‐4682. doi: 10.1210/jc.2013-4332 [DOI] [PubMed] [Google Scholar]
- 11. Amylidi‐Mohr S, Zennaro G, Schneider S, Raio L, Mosimann B, Surbek D. Continuous glucose monitoring in the management of gestational diabetes in Switzerland (DipGluMo): an open‐label, single‐centre, randomised, controlled trial. Lancet Diabetes Endocrinol. 2025;13:591‐599. doi: 10.1016/S2213-8587(25)00063-4 [DOI] [PubMed] [Google Scholar]
- 12. Burk J, Ross GP, Hernandez TL, Colagiuri S, Sweeting A. Evidence for improved glucose metrics and perinatal outcomes with continuous glucose monitoring compared to self‐monitoring in diabetes during pregnancy. Am J Obstet Gynecol. 2025;233:162‐175. doi: 10.1016/j.ajog.2025.04.010 [DOI] [PubMed] [Google Scholar]
- 13. Durnwald C, Beck RW, Li Z, et al. Continuous glucose monitoring profiles in pregnancies with and without gestational diabetes mellitus. Diabetes Care. 2024;47:1333‐1341. doi: 10.2337/dc23-2149 [DOI] [PubMed] [Google Scholar]
- 14. Valent AM, Rickert M, Pagan CH, Ward L, Dunn E, Rincon M. Real‐time continuous glucose monitoring in pregnancies with gestational diabetes: a randomized controlled trial. Diabetes Care. 2025;48:1581‐1588. doi: 10.2337/dc25-0115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Kusinski LC, Atta N, Jones DL, et al. Continuous glucose monitoring metrics and pregnancy outcome in women with gestational diabetes mellitus: a secondary analysis of the DiGest trial. Diabetes Care. 2025. doi: 10.2337/dc25-0452 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. American Diabetes Association Professional Practice Committee . 2. Diagnosis and classification of diabetes: standards of care in diabetes‐2025. Diabetes Care. 2025;48:S27‐S49. doi: 10.2337/dc25-S002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Committee on Practice Bulletins‐Obstetrics . ACOG practice bulletin No. 190: gestational diabetes mellitus. Obstet Gynecol. 2018;131:e49‐e64. doi: 10.1097/AOG.0000000000002501 [DOI] [PubMed] [Google Scholar]
- 18. Carlson AL, Beck RW, Li Z, et al. Glucose levels measured with continuous glucose monitoring in uncomplicated pregnancies. BMJ Open Diabetes Res Care. 2024;12:e003989. doi: 10.1136/bmjdrc-2023-003989 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Couch H, Pearson A, Malhotra N, et al. Continuous glucose monitoring in the management of congenital hyperinsulinism: a National User‐Satisfaction Survey, UK. J Clin Endocrinol Metab. 2025. doi: 10.1210/clinem/dgaf313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group . Validation of measures of satisfaction with and impact of continuous and conventional glucose monitoring. Diabetes Technol Ther. 2010;12:679‐684. doi: 10.1089/dia.2010.0015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Cappon G, Sparacino G, Facchinetti A. AGATA: a toolbox for automated glucose data analysis. J Diabetes Sci Technol. 2024;18:1109‐1121. doi: 10.1177/19322968221147570 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Battelino T, Alexander CM, Amiel SA, et al. Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. Lancet Diabetes Endocrinol. 2023;11:42‐57. doi: 10.1016/S2213-8587(22)00319-9 [DOI] [PubMed] [Google Scholar]
- 23. Norris T, Seaton SE, Manktelow BN, et al. Updated birth weight centiles for England and Wales. Arch Dis Child Fetal Neonatal Ed. 2018;103:F577‐F582. doi: 10.1136/archdischild-2017-313452 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kestila KK, Ekblad UU, Ronnemaa T. Continuous glucose monitoring versus self‐monitoring of blood glucose in the treatment of gestational diabetes mellitus. Diabetes Res Clin Pract. 2007;77:174‐179. doi: 10.1016/j.diabres.2006.12.012 [DOI] [PubMed] [Google Scholar]
- 25. Cheah IGS. Economic assessment of neonatal intensive care. Transl Pediatr. 2019;8:246‐256. doi: 10.21037/tp.2019.07.03 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Balaji B, Hannah W, Popova PV, et al. The use of continuous glucose monitoring in comparison to self‐monitoring of blood glucose in gestational diabetes: a systematic review. J Diabetes Sci Technol. 2025;19322968251357873. doi: 10.1177/19322968251357873 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Law GR, Ellison GT, Secher AL, et al. Analysis of continuous glucose monitoring in pregnant women with diabetes: distinct temporal patterns of glucose associated with large‐for‐gestational‐age infants. Diabetes Care. 2015;38:1319‐1325. doi: 10.2337/dc15-0070 [DOI] [PubMed] [Google Scholar]
- 28. Scholl TO, Sowers M, Chen X, Lenders C. Maternal glucose concentration influences fetal growth, gestation, and pregnancy complications. Am J Epidemiol. 2001;154:514‐520. doi: 10.1093/aje/154.6.514 [DOI] [PubMed] [Google Scholar]
- 29. Taslimi MM, Navabi K, Acosta R, Helmer A, El‐Sayed YY. Concealed maternal blood glucose excursions correlate with birth weight centile. J Diabetes Sci Technol. 2008;2:456‐460. doi: 10.1177/193229680800200315 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Chen R, Yogev Y, Ben‐Haroush A, Jovanovic L, Hod M, Phillip M. Continuous glucose monitoring for the evaluation and improved control of gestational diabetes mellitus. J Matern Fetal Neonatal Med. 2003;14:256‐260. doi: 10.1080/jmf.14.4.256.260 [DOI] [PubMed] [Google Scholar]
- 31. McLachlan K, Jenkins A, O'Neal D. The role of continuous glucose monitoring in clinical decision‐making in diabetes in pregnancy. Aust N Z J Obstet Gynaecol. 2007;47:186‐190. doi: 10.1111/j.1479-828X.2007.00716.x [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1. CGM outcomes over the daytime (06:00 AM to 23:59 PM) and nighttime (00:00 to 05:59 AM) periods.
Data Availability Statement
Some or all datasets generated during and/or analysed during the current study are not publicly available but are available from the corresponding author on reasonable request.