CGM vs Fingerstick Monitoring in Gestational Diabetes: The DipGluMo Trial

Clinical Context

Gestational diabetes mellitus (GDM) affects 6-9% of pregnancies globally and is associated with significant maternal and fetal complications. Poor glycemic control during pregnancy increases risks of fetal macrosomia, birth trauma, neonatal hypoglycemia, respiratory distress syndrome, and long-term metabolic programming effects in offspring. For mothers, GDM increases cesarean delivery rates and substantially elevates lifetime risk of developing type 2 diabetes.

Traditional management relies on self-monitoring of blood glucose (SMBG) with fingerstick testing, typically 4-7 times daily. However, SMBG captures only snapshots of glucose levels, missing postprandial excursions, nocturnal variations, and the overall glycemic pattern. Many women with GDM have glucose values that appear acceptable on spot checks yet experience significant hyperglycemia between measurements.

Continuous glucose monitoring (CGM) technology has revolutionized diabetes management in type 1 and type 2 diabetes, but its role in GDM has been less established. CGM provides near-continuous glucose data (readings every 5-15 minutes), trend information, and alerts for impending hypo- or hyperglycemia. This study (DipGluMo) examined whether real-time CGM improves outcomes in GDM compared to standard fingerstick monitoring.

Study Summary (PICO Framework)

Summary:

In pregnant women with gestational diabetes mellitus (GDM), real-time continuous glucose monitoring (CGM) did not improve the primary composite perinatal outcome (odds ratio 1.02, 95% CI 0.63 to 1.66) compared to self-monitoring of blood glucose (SMBG), although participants reported a higher preference for the CGM device.

Clinical Pearls

1. CGM reveals hidden hyperglycemia that SMBG misses. Women with GDM may have acceptable fasting and pre-meal glucose values yet experience significant postprandial spikes lasting 1-2 hours. CGM captures these excursions, enabling targeted dietary modifications or medication adjustments that would not be prompted by normal fingerstick results.

2. CGM did not reduce macrosomia or the perinatal composite in this trial. Large-for-gestational-age infants face increased risk of shoulder dystocia, birth injury, and cesarean delivery, so a genuine reduction in macrosomia would be clinically valuable. In DipGluMo, however, macrosomia and the wider composite perinatal outcome were no different with real-time CGM than with self-monitoring, so the case for CGM in gestational diabetes rests on usability and patient preference rather than on improved neonatal outcomes.

3. CGM empowers patient self-management. Real-time glucose feedback allows women to immediately see the impact of food choices, portion sizes, and physical activity. This biofeedback accelerates learning and promotes sustained dietary adherence better than delayed fingerstick results can achieve.

4. Device burden and skin irritation are manageable concerns. Modern CGM sensors are smaller and more comfortable than earlier generations. Skin irritation can be minimized with proper site rotation and barrier products. Most women adapt to device wear within the first week, and the information gained typically outweighs the inconvenience.

Practical Application

Patient selection for CGM in GDM: Consider CGM for women with GDM who have suboptimal control despite dietary management, require insulin therapy, have difficulty achieving target glucose values, or express interest in more detailed glucose information. CGM may also benefit women with prior GDM and adverse outcomes who are highly motivated to optimize this pregnancy.

CGM initiation and education: Provide hands-on training for sensor insertion and app/reader use. Explain that the first 24 hours may show inaccurate readings during sensor warm-up. Teach interpretation of trend arrows (rising, falling, stable) and appropriate responses. Set realistic expectations that CGM reveals more glucose variability than expected; this is information, not failure.

Target glucose ranges in GDM: Typical CGM targets for GDM include fasting glucose <95 mg/dL (5.3 mmol/L), 1-hour postprandial <140 mg/dL (7.8 mmol/L), and 2-hour postprandial 70% time in the 63-140 mg/dL range.

Insurance and cost considerations: CGM coverage for GDM varies by payer and region. Some insurers cover CGM only for insulin-requiring GDM. The cost of sensors (approximately $75-150 per 10-14 day sensor without insurance) may be a barrier. Because this trial did not show improved perinatal outcomes, the value proposition for routine CGM in GDM rests on usability and patient preference rather than on cost offset from prevented complications.

How This Study Fits Into the Broader Evidence

The CONCEPTT trial (2017) demonstrated CGM benefits in pregnant women with type 1 diabetes, with reduced large-for-gestational-age births and neonatal complications. DipGluMo shows that this benefit does not carry over to gestational diabetes: in a population with milder hyperglycemia and lower baseline risk, real-time CGM did not improve the composite perinatal outcome. That contrast is itself informative, suggesting the gains seen in type 1 diabetes reflect that higher-risk setting rather than a universal effect of continuous monitoring.

The ADA Standards of Care 2025 acknowledge CGM as an option for GDM management, particularly for women on insulin therapy. However, routine CGM for all GDM is not universally recommended due to cost and limited outcomes data, and this trial reinforces that caution by showing no perinatal benefit.

Real-world implementation studies have shown that intermittently scanned CGM (FreeStyle Libre) is more commonly used in pregnancy than real-time CGM (Dexcom) due to lower cost and simpler use, though both provide more glucose data than SMBG alone.

Limitations to Consider

This was a single-center study in Switzerland, which may limit generalizability to other healthcare systems. The open-label design means participants knew their treatment assignment, potentially introducing bias in self-management behaviors and in the reported device preference. Long-term offspring outcomes were not assessed. Additionally, the specific CGM system used may influence results, as sensor accuracy varies between devices.

Bottom Line

In this open-label single-centre randomised trial, real-time continuous glucose monitoring did not improve the primary composite perinatal outcome in gestational diabetes compared with self-monitoring of blood glucose (odds ratio 1.02, 95% CI 0.63 to 1.66). The clearest difference was that participants preferred the CGM device, so real-time CGM may reasonably be offered to simplify monitoring rather than to improve perinatal outcomes. On this evidence, routine real-time CGM cannot be recommended to achieve better maternal or neonatal outcomes in unselected gestational diabetes, and standard self-monitoring remains an appropriate default.

Source: Amylidi-Mohr S, et al. 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. Read article here.