Characteristics of included studies
| Author (year) | Country | Period of study | Study design | Patients | Metformin exposure (n=sample size) | Control (n=sample size) | Total number of LB in metformin-exposed group1; MCM | Total number of LB in control group2; MCM |
| Coetzee and Jackson (1984)45 | South Africa | Not reported (duration: 5.5 years) | Retrospective | Pregnant women with established non-insulin-dependent diabetes | In the beginning, 1.5–3 g/day; later 1750–2550 mg/day during first trimester (n=20) | No treatment in first trimester (n=89) | 20; 0 | 89; 5 |
| Hellmuth et al (1994)46 | Denmark | 1966–1991 | Retrospective | Pregnant women with T2DM | 250–2000 mg/day at the time of conception and during first 8 weeks of pregnancy (n=7, out of them two continued metformin until delivery and five were changed to insulin) | Treated with sulphonylurea during the first trimester (one continued until delivery, 2 changed to metformin until delivery, 15 changed to insulin) (n=18) | 7; 0 | 16; 0 |
| Jakubowicz et al (2002)47 | Venezuela | 1996–2000 | Retrospective | Non-diabetic pregnant women with PCOS | 1000–2000 mg/day throughout pregnancy (n=65) | No treatment at the time of conception or during pregnancy (n=31) | 61; 0 | 18; 0 |
| Palomba et al (2005)43 | Italy | 2003 (6 months) | RCT | Non-obese primary infertile anovulatory women with PCOS | 850 mg twice daily until confirmation of pregnancy (n=31) | Placebo (n=16) | 28; 0 | 10; 0 |
| Moll et al (2006)44 | The Netherlands | 2001–2004 | RCT | Women with PCOS | Metformin 500–2000 mg/day until confirmation of pregnancy (n=111) | Clomifene citrate plus placebo (n=114) | 29; 1 (atresia) | 42; 1 (anencephaly) |
| Hameed et al (2011)48 | Egypt | 2008–2010 | Observational | Infertile women with PCOS | Patients conceived while taking metformin (1000–2500 mg/day) with/without other ovulation inducing agents and continued metformin during pregnancy (n=31) | Patients who conceived without metformin and did not take it during pregnancy (n=26) | 30; 0 | 19; 1 (atrial septal defect (A.SD)) |
| Diav-Citrin et al (2018)31 | Israel | 2000–2013 | Prospective observational cohort study | Pregnant women with PCOS or PGDM | Median daily dose: 1700 mg, at least in the first trimester of pregnancy, (21/170 (12.4%) continued throughout gestation; for PGDM or PCOS (n=170) (51 PCOS+119 PGDM) | Pregnant women with pregestational diabetes treated with insulin (n=93); non-teratogenic exposure (n=519) | 135; 6 | 599; 14 |
| PCOS | ||||||||
| 45; 3 | 519; 11 | |||||||
| PGDM | ||||||||
| 90; 3 | 80; 3 | |||||||
| Panchaud et al (2018)35 | Multicenter (Europe) | 1993–2015 | Prospective observational cohort study | Pregnant women with a PGDM and other (obesity, ovary stimulation, insulin resistance, glucose intolerance, hyperglycemia) | Pregnant women on metformin during the first trimester of pregnancy for different indications relative to a matched unexposed reference group (n=458; of that 219 indication PGEM and 173 PCOS) | Randomly selected pregnant women who did not use metformin, insulin or other hypoglycemic agent at any time during pregnancy (n=479) | 392; 20 | 431; 9 |
| PCOS | ||||||||
| 173; 3 | 431; 9 | |||||||
| PGDM | ||||||||
| 219; 17 | 431; 9 | |||||||
| Scherneck et al (2018)29 | Germany | 2004–2014 | Prospective observational cohort study | Pregnant patients affected by PCOS, T2DM and/or insulin resistance | Patients with metformin exposure (median dosage: 1500 mg/day (IQR 1000–2000)) at least between gestational week 2+0 to 12+6 days after first day of last menstrual period (n=336) (225 first, 71 first and second and 40 all three trimesters) (69 T2DM and 163 PCOS) | Matched controls were randomly selected from all eligible pregnant women (n=1011) | 232; 11 | 913; 38 |
| PCOS | ||||||||
| 163; 8 | 913; 38 | |||||||
| PGDM | ||||||||
| 69; 3 | 913; 38 | |||||||
| Kelty et al (2020)49 | Australia | 2003–2012 | Retrospective cohort study | Pregnant women with gestational or T2DM) | Women treated with metformin (four 0.5 g tablets, two 1.0 g tablets or two 0.85 g tablets, daily) during pregnancy (n=108) | Women who were dispensed gliclazide during pregnancy (n=108) | 108; 6 | 108; 8 |
| Lin et al (2020)50 | Taiwan | 2003–2014 | Retrospective cohort study | Women with pre-existing T2DM and singleton pregnancies | Metformin before and during pregnancy (n=626) | Insulin and no oral antidiabetic drugs before and during pregnancy (n=222) | 2.72 stillbirth incidence; 609 live births; 33 | 3.60 stillbirth incidence; 214 live births; 21 |
1 and 2: these numbers were used for the meta-analysis.
LB, live births; MCM, major congenital malformations; PCOS, polycystic ovary syndrome; PGDM, pregestational diabetes mellitus; RCT, randomized controlled trials; T2DM, type 2 diabetes mellitus.