METFORMIN

CONGENITAL ANOMALIES IN THE CHILDREN OF WOMEN WITH PREGESTATIONAL TYPE 2 DIABETES MELLITUS OR GESTATIONAL DIABETES

Studies of congenital anomalies in the children of diabetic women who took metformin early in pregnancy are confounded by the teratogenic effects of maternal diabetes. The risks of congenital anomalies estimated from such studies may be useful in counseling pregnant women with diabetes who are being treated with metformin but cannot be used to determine the teratogenic potential of metformin treatment in human pregnancy.

Associations were observed with maternal metformin treatment of diabetes during the first trimester of pregnancy among 3570 infants with cardiac septal defects (odds ratio=2.1, 95% confidence interval 1.0-4.4), 748 infants with anorectal defects (odds ratio=3.2, 95% confidence interval 1.1-8.9), 3454 infants with oral clefts (odds ratio=2.3, 95% confidence interval 1.2-4.6), 1134 infants with craniosynostosis (odds ratio=2.4, 95% confidence interval 1.0-6.0), and 882 infants with limb reduction defects (odds ratio=2.9, 95% confidence interval 1.1-7.7) in the National Birth Defects Prevention Study (Dukhovny et al., 2018). With the exception of craniosynostosis, all of these associations were seen more strongly with maternal insulin treatment for diabetes.

A higher overall frequency of major congenital anomalies was observed among 219 infants whose mothers took metformin for diabetes during the first trimester of pregnancy when compared to infants and fetuses whose mothers did not take any hypoglycemic agent (odds ratio=3.95, 95% confidence interval 1.77-9.41) in a European multicenter teratogen information service study (Panchaud et al., 2018). The risk for major congenital anomalies in this study increased with greater severity of maternal diabetes.

The rate of malformations was not significantly different between 2852 infants of women with type 2 diabetes treated with metformin alone in the first trimester of pregnancy and 878 infants born to disease-matched women treated with insulin alone in a Nordic registry-based cohort study (4.4 vs 5.7%) (Kjerpeseth et al., 2023). The adjusted risk of malformations did not appear to be increased for either metformin alone or metformin combined with insulin when compared to insulin alone. Likewise, no increase in risk of congenital anomalies or nonlive birth was found among 1577 women with type 2 diabetes treated with metformin in combination with insulin during the first trimester when compared to 850 diabetic women treated with insulin monotherapy in a United States observational cohort study (Chiu et al., 2024). 

In a retrospective cohort study of pregnancies in women with gestational diabetes, the frequency of congenital anomalies was not significantly increased among 186 infants and fetuses whose mothers were treated with metformin during the first trimester when compared to 151 infants and fetuses whose mothers were treated with insulin any time during pregnancy (Vanlalhruaii et al., 2018). The frequency of major congenital anomalies was similar among 90 infants whose mothers had been treated with metformin and 80 infants whose mothers had been treated with insulin for diabetes during the first trimester of pregnancy in an Israeli teratogen information service study (Diav-Citrin et al., 2018). The rate of major congenital anomalies was similar among the infants of 85 women with type 2 diabetes who had been treated with metformin during the first trimester of pregnancy and among the infants of 121 type 2 diabetic women who received other kinds of treatment or no treatment during the first trimester in another study (Hughes & Rowan, 2006).

In a meta-analysis of randomized controlled trials and cohort studies, first trimester treatment with metformin was not associated with a meaningful increase in the risk of major congenital anomalies among infants of 1122 women with pregestational diabetes (Abolhassani et al., 2023). 

CONGENITAL ANOMALIES IN THE CHILDREN OF WOMEN WITH POLYCYSTIC OVARY SYNDROME OR OBESITY (BUT NOT DIABETES)

Studies of congenital anomalies in the children of women who took metformin early in pregnancy for indications other than diabetes may be confounded by the metabolic correlates of maternal obesity, which is associated with polycystic ovary syndrome and insulin resistance. The risks of congenital anomalies estimated from such studies may be useful in counselling non-diabetic pregnant women who are being treated with metformin but may not reflect the teratogenic potential of metformin treatment in human pregnancy.

In the National Birth Defects Prevention Study, associations were observed with maternal metformin treatment of subfertility during the first trimester of pregnancy among 1692 infants with secundum atrial septal defects (odds ratio=2.0, 95% confidence interval 1.0-3.9) and 882 infants with limb reduction defects (odds ratio=3.3, 95% confidence interval 1.50-7.2) (Dukhovny et al., 2018). Similar associations were seen with maternal metformin or insulin treatment during the first trimester for diabetes in this study.

A meta-analysis of epidemiological studies conducted prior to 2014 concluded that the risk of congenital anomalies does not appear to be increased among the infants of women treated for polycystic ovary syndrome with metformin during the first trimester of pregnancy (Cassina et al., 2014). No recurrent pattern of congenital anomalies was apparent among the affected children. Similarly, in an updated meta-analysis of randomized controlled trials and cohort studies, first trimester treatment with metformin was not associated with a meaningful increase in the risk of major congenital anomalies among infants of 529 women with polycystic ovary syndrome (Abolhassani et al., 2023).

The frequencies of major congenital anomalies and of cardiac defects were no greater than expected among 173 infants or fetuses whose mothers took metformin during the first trimester of pregnancy for indications other than pregestational diabetes in a European multicenter teratogen information service study (Panchaud et al., 2018). In another teratogen information service study, the frequency of major congenital anomalies did not appear to be substantially increased among 45 infants whose mothers had been treated with metformin for polycystic ovary syndrome (Diav-Citrin et al., 2018).

The frequency of congenital anomalies was not increased among 209 and 202 infants, respectively, in two randomized controlled trials of metformin treatment of obese, non-diabetic women during the second and third trimesters of pregnancy (Chiswick et al., 2015; Syngelaki et al., 2016).

CONGENITAL ANOMALIES IN THE CHILDREN OF WOMEN EXPOSED TO METFORMIN DURING PREGNANCY FOR VARIOUS INDICATIONS

The risk of congenital anomalies was not increased among 315 infants whose mothers were treated with metformin for polycystic ovary syndrome, diabetes or insulin resistance during the first trimester of pregnancy in a German teratogen information service study (Scherneck et al., 2018). The malformations differed in each infant in the metformin-exposed group. In a Finnish registry study, the overall frequency of major congenital anomalies was not increased among 3697 infants born to women prescribed metformin during pregnancy for any indication when compared to infants of mothers treated with insulin (Brand et al., 2022).

A signal suggesting a possible association with maternal metformin treatment during the first trimester of pregnancy was reported among infants with pulmonary valve atresia in an intergroup analysis of 50,167 infants with various congenital anomalies who had been reported to one of 11 European congenital anomaly registries (Given et al., 2018). No comparison to infants without congenital anomalies was available in this study, and the signal could have arisen by chance.

NEURODEVELOPMENTAL OUTCOMES

None of 126 infants born to women who had been treated with metformin for polycystic ovary disease during at least the first trimester of pregnancy was noted to have developmental delay by 18 months of age (Glueck et al., 2004). No differences in neurodevelopmental test scores at 18 months or two years of age were observed between children born to mothers who were treated with metformin and those born to mothers who were treated with insulin for gestational diabetes (Ijas et al., 2015; Tertti et al., 2015; Wouldes et al., 2016).

Cognitive and neuropsychological outcomes at nine years of age were similar between 82 children of women treated with metformin during pregnancy and children born to women treated with insulin for gestational diabetes in a prospective Finnish study (Paavilainen et al., 2023). Similarly, long-term cognitive functioning was no different between 52 children whose mothers were treated with metformin during pregnancy and those who received placebo in a follow-up study of two randomized controlled trials (Greger et al., 2020). In a meta-analysis of controlled studies, no association was found between maternal treatment with metformin at any time during pregnancy and adverse neurodevelopmental outcomes in 7641 adolescents (Gordon et al., 2024).

ADVERSE PREGNANCY, NEONATAL AND OTHER CHILDHOOD OUTCOMES

The frequencies of neonatal hypoglycemia, macrosomia, perinatal mortality, prematurity, respiratory distress, hyperbilirubinemia, and neonatal intensive care unit admission in infants of diabetic mothers who are treated with metformin during pregnancy have usually been found to be lower or similar to those of infants whose diabetic mothers were treated with insulin (Balani et al., 2009; Ruholamin et al., 2014; Refuerzo et al., 2015; Diav-Citrin et al., 2018; Panchaud et al., 2018; Simeonova-Krstevska et al., 2018; Vanlalhruaii et al., 2018; Picon-Cesar et al., 2021; Sheng et al., 2023; Wu et al., 2024). Metformin treatment during early pregnancy may reduce the risk of spontaneous abortion, fetal growth restriction, and preterm delivery in women with polycystic ovary syndrome or other non-diabetic indications (Lautatzis et al., 2013; Al-Biate, 2015; Zeng et al., 2016; Elmaraezy et al., 2017; Hyer et al., 2018; Panchaud et al., 2018).

The frequency of small-for-gestational age birth was higher among infants whose mothers were treated with metformin during pregnancy, as compared to insulin, in a registry-based cohort study and a prospective, randomized controlled trial (Feig et al., 2020; Brand et al., 2022). Among women with polycystic ovary syndrome, metformin use during pregnancy has been associated with larger newborn head circumference but such increases were within the normal range (Cao et al., 2021; Nilsen et al., 2024).

Nine-year-old children whose mothers were treated with metformin during pregnancy in a randomized controlled trial of women with gestational diabetes had greater weights, arm and waist circumferences, waist-to-height ratios, triceps skinfold thickness, fat mass, lean mass, and abdominal fat volumes than the children of insulin-treated mothers (Rowan et al., 2018). Similar effects have also been seen in some follow-up studies, but not in others (Carlsen et al., 2012; Ro et al., 2012; Ijas et al., 2015; Tertti et al., 2016; Hanem et al., 2018; van Weelden et al., 2018). No differences in anthropometric measurements, including weight, height, and skinfold thicknesses, were found between two-year-old children of women with diabetes treated with metformin during pregnancy as compared to placebo-treated diabetic women in a follow-up study of a randomized controlled trial (Feig et al., 2023).

Cardiovascular risk was not increased among the children of 19 obese women treated with metformin during pregnancy when compared to children of obese, placebo-treated women in a follow-up study of a randomized controlled trial (Yang et al., 2022). Moreover, improved cardiovascular outcomes (lower central hemodynamic and cardiac diastolic indices) were observed in 77 children born to obese women treated with metformin during pregnancy when compared to children of obese, placebo-treated women in a randomized controlled trial (Panagiotopoulou et al., 2020). No difference in systolic or diastolic blood pressure was found between two-year-old children whose mothers were treated with metformin for gestational diabetes in the second or third trimester of pregnancy and children whose mothers were treated with insulin in a randomized clinical trial (Battin et al., 2015).

In a meta-analysis of randomized controlled trials and prospective randomized controlled studies that focused on longitudinal growth, infants of women treated with metformin for gestational diabetes had lower average birth weights when compared to infants of insulin-treated mothers but showed accelerations in postnatal growth such that children were heavier and had higher body mass index by mid-childhood (Tarry-Adkins et al., 2019).

ANIMAL TERATOLOGY STUDIES

Increased fetal death but no significant increase in malformations was observed when pregnant rats were treated with 10-20 times the maximum human therapeutic dose of metformin (Tuchmann-Duplessis & Mercier-Parot, 1961). Oral administration of metformin to pregnant rats at a dose that was 6 times the maximum therapeutic dose in humans did not alter the weight of the offspring at birth or in adulthood (Vidigal et al., 2018).

No adverse effects of maternal metformin exposure during pregnancy on offspring growth, cardiovascular, or metabolic outcomes were identified in a meta-analysis of controlled animal studies (van Hoorn et al., 2024).