Gestational diabetes mellitus (GDM), defined as glucose intolerance with onset or first recognition during pregnancy, is now recognized as one of the most prevalent pregnancy-related complications worldwide (1, 2). GDM can increase the risk of perinatal complications and adversely affect long-term metabolic health in both mothers and their offspring. Metabolic diseases pose a serious threat to human health and have become a major global public health issue (3). Therefore, effective management of blood glucose levels in mothers with GDM is essential to prevent intergenerational transmission of metabolic diseases. There are a variety of clinical treatments for GDM, including dietary modifications, increased physical activities, oral hypoglycemic drugs, and other pharmacological therapies (4, 5). Among these, metformin is widely used due to its efficacy, safety, and low cost. As a first-line therapy, metformin has many advantages, including improvements in hemoglobin A1C (HbA1c), weight, and cardiovascular outcomes. Unless contraindicated, metformin is considered an initial pharmacological option for managing diabetes (6, 7). With the rising prevalence of GDM, the use of metformin during pregnancy has become increasingly common. However, recent evidence suggests that metformin crosses the placenta through passive diffusion, potentially reaching fetal concentrations equal to or exceeding maternal levels, thereby leading to concerns about potential effects on fetal and placental growth (8, 9). Animal studies have indicated that fetal accumulation of metformin is linked to growth restriction, as seen the emergence of SGA in the offspring of primates exposed to the drug within 30 days after conception (10). In human studies, the association between maternal metformin use and small-for-gestational-age (SGA) infants, defined as birth weight below the 10th percentile for gestational age, remains poorly characterized. Current evidence is conflicting, with some studies suggesting an increased risk of SGA (11), while others report no significant association (10). This inconsistency may stem from heterogeneity across study populations, diagnostic criteria, and treatment protocols. Furthermore, the mechanisms underlying potential effects of metformin on fetal growth require further elucidation. Given these uncertainties, it is necessary to systematically evaluate the evidence regarding the effects of metformin on fetal growth outcomes. Therefore, we conducted a meta-analysis to assess the association between maternal metformin use during pregnancy and SGA risk in offspring. The findings of this study will provide evidence-based recommendations to help optimize clinical decision-making in the management of GDM and reduce adverse neonatal outcomes.

This meta-analysis, encompassing data from 115,192 pregnancies across 19 studies, provides the most comprehensive evidence to date regarding the association between metformin use for GDM and the risk of SGA. The principal finding is that metformin does not significantly increase the risk of SGA compared to insulin, placebo, or an unclassified treatment population.

The comparable SGA risk between metformin and insulin aligns with a previous meta-analysis, reinforcing the reliability of this finding (32). Although some individual studies mentioned a potential increased risk of SGA with maternal metformin use, our statistical analysis did not support this concern. The reason might be that metformin improves maternal insulin sensitivity without substantially impairing placental function, thus exerting minimal adverse effects on fetal growth. However, as metformin crosses the placenta, the long-term effects on fetal growth and metabolism remain unclear, necessitating further study (8, 9).

A noteworthy finding is the significantly lower risk of SGA observed with metformin compared to diet modification alone (OR 0.50). This contrasts with the null findings in other comparisons and has not been highlighted in prior reviews.

This finding highlights the potential limitations of dietary interventions in achieving optimal glycemic control during pregnancy. This conclusion is based on limited data, as studies comparing metformin with diet are few and methodologically heterogeneous, warranting cautious interpretation (13, 28).

By incorporating a larger number of studies than previous reviews, this analysis provides a more comprehensive assessment of the association between metformin use and SGA risk. In the process of searching, we used the keywords with metformin, SGA, GDM and the synonymous substitutions, to assure more and complete articles being included.

A major strength of this study was the comprehensive analysis of multiple comparative groups. No previous meta-analysis has specifically compared metformin with multiple alternative treatments regarding SGA incidence, highlighting the unique contribution of our study. Conversely, insulin therapy should remain the preferred option in high-risk populations, including those with pre-existing fetal growth concerns or contraindications to metformin use.

However, there were several limitations. Variations in GDM diagnostic criteria, treatment protocols, metformin dosing (ranging from 500 mg to 2500 mg daily), patient baseline characteristics (e.g., age, BMI, ethnicity—with approximately 45% of studies not reporting ethnicity) contributed to the moderate-to-high statistical heterogeneity observed (I² up to 55%) (9, 33) (Table 4). It suggests that the overall “no significant difference” finding may mask variable effects in different patient subgroups or clinical settings. Additionally, the complete absence of long-term follow-up data in the included studies represents a major evidence gap. While our analysis focused on neonatal SGA, it provides no information on potential long-term developmental, metabolic, or cardiovascular outcomes in children exposed to metformin in utero. Therefore, the absence of an overall significant association should not be interpreted as uniform safety across all clinical contexts, and individualized treatment decisions remain essential. This limitation precludes a comprehensive safety assessment and means that current clinical recommendations based on short-term neonatal outcomes must be made with caution, acknowledging this unknown. In clinical practice, this means that while metformin may be appropriate for many women with GDM, insulin may remain preferable in pregnancies with existing concerns about fetal growth or in populations not well represented in current studies. The limitations weaken the robustness and generalizability of these specific comparisons, and prevents a comprehensive evaluation of the potential developmental and metabolic effects of prenatal metformin exposure on offspring, preventing a definitive evaluation of metformin's long-term safety profile.

To address the limitations of current evidence and translate these findings into clearer clinical guidance, future research should prioritize the following: Standardization of Study Protocols: prospective studies, especially randomized controlled trials, should adopt consensus-based, standardized protocols for GDM diagnosis, metformin dosing, and comparator treatments. This will reduce methodological heterogeneity and facilitate more robust and comparable meta-analyses in the future. Long-term Offspring Follow-up: the focus of research must extend beyond birth outcomes. Large-scale, longitudinal cohort studies are imperative to evaluate the long-term cardiometabolic health, neurodevelopment, and growth trajectories of children exposed to metformin in utero. Such data are essential for a comprehensive risk-benefit assessment of metformin use in pregnancy.

Overall, maternal oral metformin can indeed reduce newborn weight to a certain extent (metformin can prevent LGA to a certain extent), but the data tend to presents a milder effect that has not reached the level of SGA. The potential for its broader clinical application remains promising. Some studies provided proof of concept that treatment of preterm pre-eclampsia and pregnancy-related hypertension is possible (34, 35). Mechanistic studies exploring how metformin influences placental function and fetal development may further elucidate its safety profile. Through this analysis, we currently agree on the safety of metformin, it is reasonable to use metformin in those with type 2 diabetes and chronic hypertension or nephropathy in pregnancy (36).

This meta-analysis suggests that metformin is a cost-effective and generally safe treatment option for GDM, with no significant association with increased SGA risk in the general population. Metformin presents a practical alternative for patients who cannot tolerate insulin or face significant financial constraints. Its affordability and ease of use make it an attractive option for low-risk pregnancies or resource-limited settings. Despite these promising findings, the need for further research remains critical to fully understand the long-term effects of metformin and to optimize GDM management strategies. A patient-centered approach that considers individual risk factors, treatment preferences, and socioeconomic circumstances is crucial for tailoring GDM management strategies (6, 8, 9, 36, 37).