Metformin with insulin during pregnancy for type 2 diabetes does not appear to increase the risk of serious adverse events compared to use of insulin alone, according to the largest study of its kind.
The results of the Australian-Canadian research have been described as reassuring in an area where evidence to date has been patchy and contradictory.
More than 500 pregnant women with type 2 diabetes from 29 centres in Canada and Australia were randomised to receive either metformin 1000mg twice daily or placebo, in addition to a standard regimen of insulin, between May 2011 and October 2018.
The average patient age was 35. In both groups most women (70%) were non-European, 7% were Aboriginal, and 43% were considered socioeconomically disadvantaged (either having migrated within five years of study entry, being single, or having a highest educational attainment of secondary school or less).
Slightly less than a fifth of the cohort had polycystic ovary syndrome, more than 75% were obese, and roughly 11% reported ever smoking.
The team, led by the University of Toronto in Canada, found no significant difference between groups in the primary outcome: a composite score including measures of pregnancy loss, preterm birth, birth injury, moderate or severe respiratory distress syndrome, neonatal hypoglycaemia, and NICU admission lasting more than 24 hours.
They also found women in the metformin group achieved better glycaemic control, required less insulin (1.1 vs 1.5 units per kg per day), gained less weight (7.2kg vs 9kg on average), and had fewer caesarean births (53% vs 63%) compared to the placebo group.
Despite the increasing use of metformin during pregnancy for women with type 2 diabetes, the authors said their literature review found only five small randomised control trials involving metformin use in pregnancy, two of which investigated neonatal outcomes.
“Little data exist on the benefits and harms of metformin use on pregnancy outcomes in these women,” they wrote in the paper published in The Lancet Diabetes & Endocrinology.
“Notably, by contrast with some previous studies, we did not find a difference in serious neonatal outcomes when metformin was added to insulin. Given the far greater statistical power and rigour of the MiTy trial compared with previous trials, this discrepancy is most likely explained by previous studies having small sample sizes, no masking, and excluding women already on insulin.”
Babies in the metformin group tended to be smaller: there were less likely to be extremely large for gestational age and had reduced adiposity measures.
However, they were also more likely to be small for gestational age (13% of infants in the metformin group and 7% in the placebo group).
While the authors said it was unclear whether the downwards shift in birthweight in the metformin group represented smaller but healthier babies or pathologically growth-restricted babies, they said further research was crucial to understand long-term effects.
“This finding was reported in one previous study of women with type 2 diabetes and is important, because infants who are small for gestational age might have an increased risk of perinatal complications and potential long-term consequences such as neurodevelopmental delay affecting school performance, obesity, hypertension, diabetes, and cardiovascular disease, similar to infants of mothers with under-nutrition,” they wrote.
“Regardless of the cause, understanding the relative implications of these findings is important to properly advise patients who are contemplating the use of metformin during pregnancy.”
Co-author Professor David Simmons from Western Sydney University said the team had secured Canadian funding to investigate the implications for the cohort’s offspring for two years.
But metformin was an option that should be considered for pregnant women with type 2 diabetes, he said, since their babies were at such significant risk of stillbirth and other severe adverse outcomes.
“By managing the glucose with metformin, we get better adherence, better weight management, better glucose control, and on balance, it’s a good addition to our armamentarium,” said Professor Simmons, who is head of endocrinology at Campbelltown Hospital.
He stressed that care for pregnant women with type 2 diabetes involved balancing multiple risks, which should be handled by a multidisciplinary specialist team.
“Metformin diverts fuel, so if a woman is hyperglycaemic you want to divert the fuel because we know that if a baby is hyperglycaemic that also leads to worse outcomes, including bigger babies and an increased chance of obesity later in life,” he said.
“[But] whenever I see growth tailing off, I stop metformin and put them on insulin. It’s kind of a U-shape and we’re trying to manage that.”
He added that multiple factors including maternal weight management might limit fetal growth and lead to babies being small for gestational age, and that fat mass vs lean mass were not measured in the current study.
“Those who were on metformin had a lower weight gain, and better glucose control. It may be that the metformin itself is actually perfectly reasonable and in fact it’s the diet that we’re giving and the wider aspects of our glucose targets and whatever else that we’re doing that we need to have more research on.”
Endocrinologist Associate Professor Alison Nankervis, who was not associated with the work, said the results of the study had been “eagerly awaited”.
“It’s in the context of divergent and often strongly held views in the diabetes and pregnancy area,” said Professor Nankervis, who is the clinical head of diabetes at Royal Melbourne Hospital.
“This is a seminal study, well designed and executed, which has found mainly positive outcomes for both mother and offspring with the use of metformin in type 2 diabetes mellitus in pregnancy.
“The only negative is the small-for-gestational-age neonates. I find the results reassuring overall, but absolutely agree with the need to follow these offspring to determine long-term sequelae.”
Professor Nankervis said her unit already used metformin, almost always with insulin, in most women with type 2 diabetes mellitus, so the results were welcome but unlikely to change their practice.