Clinician evidence review
CGM in Gestational Diabetes
An Emerging Paradigm Shift
CGM expands what we can see. Research must establish how best to act on that information.
Measurement
More of the glucose pattern becomes visible.
Clinical utility
Outcome benefits vary across trials.
Treatment targets
GDM-specific percentage targets remain unresolved.
Chukwuma Onyeije, MD · OpenMFM
Targeted literature check: September 7, 2026 · Clinical review pending
01 / Learning objectives
Three questions frame the clinical decision
- What is measured?Distinguish a timed blood glucose value from a sensor-derived pattern.
- What improves?Evaluate the primary endpoint and the care pathway tested in each trial.
- What should trigger action?Separate established guidance, expert consensus, and exploratory CGM targets.
Identify the decision before interpreting the trace.
02 / Scope
Diabetes populations require separate evidence
| Population | Evidence and guidance | Implication |
|---|---|---|
| Type 1 diabetes | CONCEPTT supports pregnancy benefit; ADA recommends CGM | Use pregnancy-specific guidance and individualized therapy |
| Type 2 diabetes | Pregnancy evidence and percentage targets are less established | Individualize CGM use; avoid borrowing the magnitude of T1D benefit |
| Gestational diabetes | Recent randomized trials have mixed results | Discuss potential value alongside treatment, access, and burden |
03 / Measurement
Scheduled checks sample a larger glucose pattern
Duration, timing, trajectory, and overnight exposure become visible between fasting and postprandial checks.
04 / Biological rationale
A glucose–risk association does not validate a dosing target
- AssociationHAPO found graded relationships between maternal glucose and adverse pregnancy outcomes.
- Candidate metricCGM makes exposure over time measurable. Timing and duration may add information.
- Intervention testA treatment strategy must show benefit and acceptable harms before its threshold becomes a validated target.
HAPO was observational and used an oral glucose tolerance test. It was not a trial of CGM-guided treatment.
05 / Evaluating technology
Accuracy and clinical utility answer different questions
| Construct | Question |
|---|---|
| Analytical accuracy | How closely does the sensor agree with a reference measurement? Correlation alone is insufficient. |
| Trend accuracy | Does it correctly represent direction and rate of change? |
| Clinical accuracy | Could measurement error lead to a harmful decision? |
| Clinical utility | Does using the device within a care pathway improve outcomes? |
Reliable treatment requires accuracy even when the long-term value lies in patterns.
06 / Discordant readings
A discrepant pair can matter for an immediate decision
Immediate assessment
Consider symptoms, glucose trajectory, recent treatment, sensor pressure, and device status. Use meter confirmation when indicated.
Overall assessment
One pair does not characterize accuracy across a sensor’s wear period. Recurrent unexplained discrepancies deserve review.
ADA advises finger-stick checks in circumstances including symptoms that disagree, after hypoglycemia correction, and on day 1 when accuracy can be lower. Follow the device’s instructions.
07 / Dexcom G7 validation
Pregnancy validation is device- and population-specific
Enrolled: 59 T1D, 21 T2D, 25 GDM
Participants contributing accuracy data
Overall 20/20 agreement across 2,102 pairs
Comparator: arterialized venous glucose measured with YSI. The study evaluated a specific device across mixed diabetes types.
How to interpret these findings
20/20 agreement: within 20% of comparator readings ≥100 mg/dL, or within 20 mg/dL below 100 mg/dL. Agreement varied by wear day: 78.6% on day 1, 96.3% on days 4 and 7, and 97.3% on day 10.
Funding: Dexcom. This accuracy study does not establish neonatal benefit or validate every sensor in every pregnancy setting.
08 / The randomized evidence
Compare the primary endpoint and the care pathway
| Trial | Randomized | Primary endpoint | Comparator |
|---|---|---|---|
| GRACE | 375 | Large for gestational age (LGA) | SMBG plus standard care |
| DipGluMo | 302 | Composite perinatal outcome | SMBG six times daily |
| Steady Sugar | 128 | TIR 63–140 mg/dL | SMBG + monthly blinded CGM, with clinician review |
| Valent | 111 | TIR 60–140 mg/dL | CBG four times daily + periodic blinded CGM |
The glucose range, intervention, and clinical endpoint are not uniform.
09 / Primary endpoint: large for gestational age
GRACE found fewer LGA births with CGM
Scale: 0–20%
Reported OR 0.32 (95% CI 0.10–0.87; p = 0.014)
Study design, safety, and limitations
375 randomized (190 CGM / 185 SMBG); primary endpoint available for 345 (170 / 175). Open-label, four university hospitals in Austria, Germany, and Switzerland. Randomization at mean 28.6 weeks; Dexcom G6 through delivery versus SMBG. Control participants wore blinded CGM for assessment.
SGA: 33/170 (19%) versus 23/175 (13%); OR 1.59 (95% CI 0.86–2.99). This imprecise safety signal merits study, without proving overtreatment. LGA used customized GROW percentiles.
Funding: Dexcom. Limitations include open treatment, missing primary data, setting-specific care, and low event counts. Absolute percentages above are calculated from reported counts.
10 / Primary endpoint: composite perinatal outcome
DipGluMo found no improvement in its composite endpoint
Scale: 0–50%
Adjusted OR 1.02 (95% CI 0.63–1.66)
Study design, correction, and limitations
302 randomized (157 / 145); three withdrew before baseline, leaving 299 (156 / 143). Primary data: 297 (154 / 143). Single-center, open-label study in Bern. Dexcom G6 versus SMBG six times daily; blinded CGM assessments in controls.
Composite: LGA, macrosomia, polyhydramnios, neonatal hypoglycemia, or stillbirth. Unadjusted OR 1.06 (95% CI 0.66–1.71). The interval permits both benefit and harm; the study did not establish equivalence.
Funding: University of Bern and Swiss Diabetes Foundation. Limitations: single center, composite endpoint, and missing comparative sensor data. The published correction changes control-group TAR in Table 4 to 2.3%; it does not change this primary result. Percentages above are calculated from counts.
11 / Primary endpoint before secondary findings
Steady Sugar’s primary TIR result was nonsignificant
CGM TIR
95% CI 82.5–94.3
Control TIR
95% CI 79.1–93.9
Range 63–140 mg/dL · p = 0.37
Secondary LGA and NICU findings favored CGM. They require interpretation alongside the primary result and the CGM-informed comparator.
Endpoints, denominator changes, and funding
Single-center early GDM, diagnosed after 8 and before 26 weeks. 128 randomized (87 / 41); 120 with CGM data (80 / 40). Dexcom G6 with recommended SMBG twice daily versus SMBG four times daily plus monthly blinded G6 Pro. Clinicians reviewed CGM in both groups.
LGA: 4/80 (5.0%) vs 7/38 (18.4%), p = 0.019. NICU: 18/80 (22.5%) vs 17/38 (44.7%), p = 0.013. Between-group confidence intervals for these binary outcomes were not reported in the outcome table. Two controls were excluded from delivery analyses, including one stillbirth.
Group TIR intervals above are the paper’s reported mean 95% CIs, not the CI for a treatment difference. Funding and supplies: Dexcom; several authors had company affiliations. Multiple secondary comparisons and a single center limit inference.
12 / A glycemic primary endpoint
Valent found higher TIR with real-time CGM
CGM TIR, mean ± SD
Control TIR, mean ± SD
Range 60–140 mg/dL · p = 0.027
The trial does not establish a definitive neonatal benefit.
Trial context and limitations
111 randomized at ≥20 weeks (74 CGM / 37 control), single center at Oregon Health & Science University. Continuous Dexcom G6 plus adjunctive capillary blood glucose (CBG) versus CBG four times daily with periodic blinded CGM.
Primary outcome: percentage of time in 60–140 mg/dL from enrollment to admission for delivery. The abstract reports group means and SDs; a between-group 95% CI is not supplied there. Numbers shown are percentages, despite a unit inconsistency in the abstract display.
Funding: Dexcom. Open-label, small study designed around glycemia. Sensor availability and adherence affect interpretation; neonatal outcomes require larger trials.
13 / Clinical inference
Trial differences generate hypotheses about benefit
Comparator care
Frequent SMBG and structured review may narrow the added benefit of continuous data.
Population and timing
Early GDM, baseline risk, and gestational age may affect the opportunity to intervene.
Treatment response
Education, adherence, and the response to data may influence outcomes.
None of these explanations was isolated experimentally. Missing data and chance also remain plausible contributors.
14 / Evidence synthesis
Meta-analyses depend on what they pool
| Review | Evidence set | Interpretation |
|---|---|---|
| Gautam 2026 | 11 GDM RCTs; 1,225 participants | Lower mean birth weight; no significant improvement in multiple other clinical endpoints |
| Yang 2026 | 21 GDM studies; 5,650 participants; RCTs and observational studies | Broader favorable associations; pooled causal interpretation is limited by mixed designs |
| Balaji 2026 | 35 studies; mixed designs; search through January 2025 | Useful context on patterns; not a synthesis of all later trials |
Why these summaries can differ
Yang searched through April 2026 and reported reductions in selected glycemic and pregnancy outcomes. Gautam’s RCT-only synthesis found no significant differences in macrosomia, neonatal hypoglycemia, or NICU admission. Compare included trials, study design, sensor generation, and endpoint definitions before comparing effect estimates.
A lower mean birth weight is not automatically a net clinical benefit. SGA, hypoglycemia, burden, and costs belong in the benefit–harm assessment.
15 / Guidance and unresolved targets
GDM blood glucose goals and T1D CGM goals are distinct
Blood glucose goals in pregnancy
Fasting <95 mg/dL
Either 1-hour postprandial <140
or 2-hour postprandial <120
Treatment goals, not diagnostic OGTT thresholds. Individualize for safety.
T1D pregnancy CGM percentage goals
63–140 mg/dL: TIR >70%
>140 mg/dL: TAR <25%
<63 mg/dL: TBR <4%
<54 mg/dL: TBR <1%
Routine GDM and T2D percentage targets remain unresolved. The T1D percentage goals should not be presented as validated GDM targets.
16 / Illustrative case
Overnight elevation prompts contextual review
A patient’s sensor shows repeated overnight elevation despite a reassuring scheduled morning check.
Reveal the clinical reasoning
Review several nights and trace quality. Reconcile symptoms, pressure on the sensor, meter checks when indicated, late meals, and current therapy. Then evaluate whether treatment needs adjustment within the patient’s plan.
This invented trace illustrates a review process. It does not supply a validated nocturnal threshold or an automatic insulin rule.
17 / Illustrative comparison
An overall TIR can conceal timing-specific concerns
Trace A: 90 mg/dL all day
TIR in 63–140 mg/dL
Trace B: 109 mg/dL overnight, 90 later
TIR in the same range
Both invented profiles stay in range. Their overnight patterns differ.
Reveal the clinical reasoning
A 63–140 mg/dL summary cannot determine whether fasting blood glucose meets its separate goal. Evaluate fasting and meal-specific readings with the clinical context. Sensor and blood glucose values are not interchangeable at every moment.
These simplified profiles are hypothetical, not physiologic predictions or a dosing algorithm.
18 / Implementation proposal
CGM needs a defined review and response plan
- Select togetherDiscuss treatment, preferences, access, device suitability, and out-of-pocket cost.
- Teach and prepareExplain alerts, symptoms, checking rules, and backup meter supplies.
- Assign responsibilitySpecify who reviews data, how often, and how urgent concerns reach the team.
- ReassessReview usefulness, treatment response, hypoglycemia, skin effects, and burden.
19 / Five take-home messages
Clinical Pearls
- Separate GDM, T1D, and T2D evidence.
- Distinguish better measurement from demonstrated clinical benefit.
- Assess discordance when it matters to an immediate decision.
- Read the primary endpoint before favorable secondary findings.
- Review fasting and meal-specific patterns alongside summary metrics.
20 / What remains unresolved
Evidence & Controversies
Known
CGM reveals patterns between scheduled checks. Some trials show benefit, and others show no primary clinical improvement.
Uncertain
Which GDM patients benefit most, and which treatment algorithms improve neonatal outcomes safely?
Practice varies
Selection, escalation thresholds, review frequency, and reimbursement remain context-dependent.
Next studies
Compare actionable algorithms. Measure neonatal outcomes, SGA, hypoglycemia, patient burden, cost, and equitable access.
The strongest next step is a validated action strategy for the additional information.
21 / Evidence trail
References: guidance and randomized trials
- GuidelineAmerican Diabetes Association Professional Practice Committee. 15. Management of Diabetes in Pregnancy: Standards of Care in Diabetes—2026. Diabetes Care. 2026;49(Suppl 1):S321–S338. doi:10.2337/dc26-S015.
- Expert consensusBenhalima K, Durnwald C, Sweeting A, et al. Application of continuous glucose monitoring and automated insulin delivery technologies for pregnant women with type 1, type 2, or gestational diabetes: an international consensus statement. Lancet Diabetes Endocrinol. 2026;14(2):157–177. doi:10.1016/S2213-8587(25)00335-3.
- GDM randomized trialLinder T, Dressler-Steinbach I, Wegener S, et al. Glycaemic control and pregnancy outcomes with real-time continuous glucose monitoring in gestational diabetes (GRACE): an open-label, multicentre, multinational, randomised controlled trial. Lancet Diabetes Endocrinol. 2026;14(1):50–61. doi:10.1016/S2213-8587(25)00288-8.
- GDM randomized trialAmylidi-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(7):591–599. doi:10.1016/S2213-8587(25)00063-4.
- Published correctionCorrection to Lancet Diabetes Endocrinol 2025;13:591–99. Lancet Diabetes Endocrinol. 2026;14(3):e6. doi:10.1016/S2213-8587(25)00403-6.
- GDM randomized trialElkind-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;28:691–700. doi:10.1111/dom.70254.
- GDM randomized trialValent AM, Rickert M, Huerta Pagan C, Ward L, Dunn E, Rincon M. Real-Time Continuous Glucose Monitoring in Pregnancies With Gestational Diabetes Mellitus: A Randomized Controlled Trial. Diabetes Care. 2025;48(9):1581–1588. doi:10.2337/dc25-0115.
22 / Evidence trail
References: measurement and evidence synthesis
- Device validationPolsky S, Valent AM, Isganaitis E, et al. Performance of the Dexcom G7 Continuous Glucose Monitoring System in Pregnant Women with Diabetes. Diabetes Technol Ther. 2024;26(5):307–312. doi:10.1089/dia.2023.0516.
- Observational evidenceHAPO Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med. 2008;358(19):1991–2002. doi:10.1056/NEJMoa0707943.
- T1D randomized trialFeig DS, Donovan LE, Corcoy R, et al. Continuous glucose monitoring in pregnant women with type 1 diabetes (CONCEPTT): a multicentre international randomised controlled trial. Lancet. 2017;390(10110):2347–2359. doi:10.1016/S0140-6736(17)32400-5.
- GDM RCT meta-analysisGautam N, de Souza BL, Abramowitz J, Mourão DM, Mirfakhraee S, Abreu M. Continuous glucose monitoring versus self-monitoring of blood glucose in gestational diabetes: an updated systematic review and meta-analysis of randomized controlled trials. Acta Diabetol. 2026;63(7):1147–1157. doi:10.1007/s00592-026-02644-1.
- Mixed-design GDM meta-analysisYang S, Boon ET, Yang Q, et al. Comparing efficacy in glycaemic control and pregnancy outcomes between continuous glucose monitoring and self-monitoring of blood glucose in pregnancies complicated by gestational diabetes mellitus: a systematic review and meta-analysis. Diabetes Res Clin Pract. 2026;239:113488. doi:10.1016/j.diabres.2026.113488.
- Mixed-design systematic reviewBalaji 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. 2026;20(1):173–183. doi:10.1177/19322968251357873.
- Device instructionsDexcom. What is the Dexcom G7 Continuous Glucose Monitoring System? US patient information and safety statement. Accessed September 7, 2026. Device-specific instructions; not a class-wide recommendation.
Chukwuma Onyeije, MD / OpenMFM · Clinical review pending
Targeted search and source checks: September 7, 2026. This is an educational review, not a systematic search.