Pharmacologic therapy is commonly used in gestational diabetes mellitus (GDM), yet comparative evidence across drug classes has remained limited. A network meta-analysis of randomized controlled trials published in Diabetes, Obesity and Metabolism provides a comparative evaluation of multiple pharmacologic options, assessing both glycemic control and maternal and neonatal outcomes.
The analysis included 71 trials with 14,877 patients identified through systematic searches of PubMed, Embase, Cochrane Central Register of Controlled Trials, and Web of Science up to May 16, 2025. Random-effects models were applied for pairwise and frequentist network meta-analyses, and the Confidence in Network Meta-Analysis framework was used to assess certainty of evidence. Seven drug classes were evaluated, including insulin, biguanides, sulfonylureas, α-glycosidase inhibitors, sodium-glucose cotransporter-2 inhibitors, dipeptidyl peptidase-4 inhibitors, and glucagon-like peptide-1 receptor agonists, with all effects compared against standard care.
Among therapies supported by moderate to high certainty evidence, sulfonylureas showed the greatest reduction in fasting glucose (mean difference −0.33 mmol/L; 95% confidence interval −0.55 to −0.10), followed by insulin (−0.30 mmol/L; 95% confidence interval −0.40 to −0.21) and biguanides (−0.20 mmol/L; 95% confidence interval −0.28 to −0.12). Biguanides showed the greatest reduction in glycated hemoglobin (HbA1c) (−0.10%; 95% confidence interval −0.16 to −0.03), but were associated with increased odds of low birth weight (odds ratio 2.04; 95% confidence interval 1.04 to 4.01).
For neonatal outcomes, insulin (odds ratio 0.51; 95% confidence interval 0.34 to 0.75), biguanides (0.39; 95% confidence interval 0.26 to 0.59), and sulfonylureas (0.50; 95% confidence interval 0.31 to 0.79) were associated with lower odds of macrosomia. Sulfonylureas were associated with higher rates of preterm delivery and neonatal hypoglycemia compared with biguanides, while evidence for other drug classes, including α-glycosidase inhibitors, sodium-glucose cotransporter-2 inhibitors, dipeptidyl peptidase-4 inhibitors, and glucagon-like peptide-1 receptor agonists, remained limited.
These findings indicate differences in glycemic and neonatal outcomes across pharmacologic options in GDM. Insulin showed consistent glycemic control with a favorable safety profile, whereas biguanides were associated with increased odds of low birth weight.
