Metformin vs Insulin for GD: 2026 Evidence Update

The debate about whether to use metformin or insulin for gestational diabetes has produced more research than nearly any other question in perinatal medicine. And in 2025-2026, the body of evidence grew significantly—including important long-term follow-up data on children exposed to metformin in the womb. Here's what the current evidence shows and what the 2026 guidelines recommend.

The Clinical Situation

Approximately 15-25% of women with gestational diabetes cannot achieve adequate blood glucose control through diet and exercise alone. When lifestyle measures fail, medication becomes necessary. Two options are available:

• Insulin: FDA-approved for type 2 diabetes and gestational diabetes. Gold standard. Does not cross the placenta. Requires injections.
• Metformin: FDA-approved for type 2 diabetes. Used off-label for GD. Oral medication. Crosses the placenta.

A third option—glyburide—was once common but has fallen significantly out of favor (see our article on glyburide in pregnancy 2026).

Efficacy: How Well Does Each Work?

Glycemic Control

Multiple randomized controlled trials and meta-analyses have compared glycemic outcomes between metformin and insulin in GD:

• A large meta-analysis (>2,000 women) found similar rates of achieving glycemic targets with both agents
• A 2024 JAMA randomized trial comparing oral glucose-lowering agents (primarily metformin) vs. insulin found comparable neonatal composite outcomes (32.0% vs. 32.2%)
• Approximately 25-46% of women started on metformin require supplemental insulin to achieve adequate control—this "failure rate" must be considered when choosing initial therapy

Neonatal Outcomes

A comprehensive 2023-2024 meta-analysis comparing metformin and insulin for GD found:

Maternal Outcomes

The Critical Issue: Placental Transfer

Insulin does not cross the placenta in significant amounts. This is a fundamental difference between the two drugs:

Insulin: Exogenous insulin molecules are too large to cross the placental barrier. The insulin you inject stays in your circulation and does not directly affect your baby's insulin levels or development.

Metformin: Metformin is a small molecule that readily crosses the placenta. Umbilical cord blood metformin levels can be as high as or higher than simultaneous maternal blood levels. The fetus is directly exposed to metformin throughout the treatment period.

This difference is why the ADA recommends insulin as first choice: the long-term effects of prenatal metformin exposure on fetal development are still being studied.

The MiG TOFU Data: Long-Term Offspring Follow-Up

The most important new data informing the 2026 metformin debate comes from the Metformin in Gestational Diabetes: The Offspring Follow-Up (MiG TOFU) study, which has been tracking children born to mothers enrolled in the original MiG trial:

At 2 years:

No significant differences in BMI, body fat, or other metabolic measures between metformin-exposed and insulin-exposed children.

At 7-9 years:

In one cohort (Auckland, New Zealand), children exposed to metformin had higher BMI, larger waist circumference, and higher waist-to-height ratio compared to insulin-exposed children. In a second cohort (Adelaide, Australia), these differences were not observed.

The inconsistency between cohorts means definitive conclusions cannot be drawn. But the Auckland findings—showing potential programming of increased adiposity in metformin-exposed children—raise a concerning possibility that has led most major guidelines to maintain insulin as first-line therapy.

Potential Mechanisms

Researchers are investigating several ways prenatal metformin exposure might affect fetal development:

• Metformin inhibits mitochondrial complex I—mitochondrial function is critical for fetal energy metabolism
• Animal (primate) models have shown metformin initiated early in gestation leads to fetal bioaccumulation, growth restriction, and renal dysmorphology at clinical exposure levels
• Epigenetic programming: metformin may alter gene expression patterns related to energy metabolism

2026 ADA Recommendation

The 2026 ADA Standards state:

"Insulin is the preferred pharmacological treatment for hyperglycemia in gestational diabetes mellitus. Metformin may be used as an alternative when insulin is not available or acceptable to the patient, after discussion of the limited long-term safety data for offspring."

Key elements of this recommendation:

• Insulin is preferred, not mandatory—shared decision-making is appropriate
• Metformin is a valid alternative, not a prohibited option
• Informed consent requires discussing the limited long-term offspring data
• Women who strongly prefer to avoid injections can use metformin with appropriate counseling

Making the Decision: Questions to Discuss With Your Provider

1. How far above target are my glucose levels, and does that affect which drug is more appropriate?
2. What are your thoughts on the long-term offspring data—how much does that influence your recommendation?
3. If I start metformin, what is the probability I'll need to add insulin anyway?
4. What injection training and support do you offer for insulin?
5. How does the post-pregnancy plan differ between these options?
6. Are there any aspects of my specific situation (BMI, glucose levels, gestational age) that favor one option?

The Bottom Line

For short-term pregnancy outcomes, metformin and insulin are broadly comparable—and metformin has real advantages in maternal tolerability, lower hypoglycemia risk, and patient preference. However, insulin remains the first-line recommendation in 2026 guidelines because:

1. It is FDA-approved for GD
2. It does not cross the placenta
3. It has more data (no long-term offspring concerns)
4. It works for everyone (no "failure" requiring supplemental medication)

This is a genuinely nuanced decision that should be made in dialogue with your healthcare team, not a one-size-fits-all recommendation.

For related reading, see our article on glyburide safety in pregnancy 2026 and our overview of the 2026 GD guidelines.

This content is for informational purposes only and is not medical advice. Always consult your healthcare provider or registered dietitian for personalized guidance.