Metformin provides glycemic, neonatal benefits for women with gestational diabetes
Click Here to Manage Email Alerts
Key takeaways:
- Women with gestational diabetes who received metformin had lower fasting glucose than those who received placebo.
- Fewer infants of women in the metformin group were born large for gestational age.
Metformin may lengthen the time until insulin initiation, lower fasting glucose and improve neonatal outcomes for pregnant women with gestational diabetes, according to data from a randomized trial.
In findings presented at the European Association for the Study of Diabetes annual meeting and simultaneously published in JAMA, researchers compared glycemic, maternal and neonatal outcomes for women with gestational diabetes randomly assigned up to 2,500 mg metformin daily with those receiving placebo. The metformin group had a lower mean fasting glucose at 32 and 38 weeks of gestation, and the offspring of women receiving metformin had a lower mean birth weight than the offspring of those receiving placebo.
“This trial shows that metformin is a safe and effective alternative treatment that women now have the choice to take if they have gestational diabetes,” Fidelma Dunne, MD, PhD, MMed, professor in the College of Medicine, Nursing and Health Sciences at the University of Galway in Ireland, told Healio.
Dunne and colleagues conducted a phase 3 randomized, double-blind, placebo-controlled trial of pregnant women aged 18 to 50 years diagnosed with gestational diabetes. Participants were enrolled at two sites in Ireland. At diagnosis, women were randomly assigned, 1:1, to metformin and placebo daily until delivery. Participants performed seven-point glucose testing at mealtimes and before bed each day. Laboratory testing was conducted at 32 and 38 weeks of gestation. Birth visits took place within 72 hours of delivery to assess maternal and neonatal outcomes. The primary outcome of the study was a composite of the need to initiate insulin before delivery or having a fasting glucose of 5.1 mmol/L or higher at 32 or 38 weeks of gestation.
Metformin reduces fasting glucose
Of 150 women in the metformin group and 167 women in the placebo group with data available, the risk of meeting the composite outcome of insulin initiation or a fasting plasma glucose level of 5.1 mmol/L or higher was similar between the groups. However, in an alternative time-to-event analysis, women receiving metformin were less likely to initiate insulin than women receiving placebo (HR = 0.66; 95% CI, 0.51-0.85; P = .001). When fasting glucose was analyzed on its own, women in the metformin group had a lower mean fasting glucose at 32 weeks of gestation (mean difference, –0.1 mmol/L; 95% CI, –0.19 to –0.01; P = .03) and 38 weeks of gestation (mean difference, –0.2 mmol/L; 95% CI, –0.28 to –0.09; P < .001) compared with placebo. Postprandial glucose was also lower in the metformin group after lunch and dinner at 32 weeks of gestation and after dinner at 38 weeks compared with placebo. Women receiving metformin gained 0.8 kg of body weight from randomization to delivery compared with a mean 2 kg weight gain for those receiving placebo (P = .003).
“Although the primary composite outcome was similar between groups, there were very important positive impacts on maternal weight gain and fasting and post-meal sugars at weeks 32 and 38,” Dunne told Healio.
Lower birth weight with metformin
Infants from mothers in the metformin group had a lower mean birth weight than offspring of mothers from the placebo group (3,393 g vs. 3,506 g; P = .005). The percentage of infants born large for gestational age was lower in the metformin group vs. placebo (6.5% vs. 14.9%; P = .003). Mean crown-to-heel length was shorter in offspring of mothers from the metformin group compared with placebo (51 cm vs. 51.7 cm; P = .02).
“Caution should continue with metformin and small for gestational age, especially in those where small for gestational age may be more likely, so those with hypertension or nephropathy,” Dunne said during the presentation.
Gastrointestinal adverse events were reported by 24.3% of the metformin group compared with 4.1% of the placebo group (P < .001). Thirteen women in the metformin group discontinued the trial due to gastrointestinal adverse events.
Dunne said more studies are needed to analyzed more metabolic and cardiovascular variables as well as body composition data for infants.
“Future trials should be randomized and placebo-controlled when we’re talking about metformin in pregnancy,” Dunne said during the presentation. “Women should be consented for long-term follow-up. We should give consideration to bio-banking in all of these trials for interrogation of disease mechanisms. Going forward, we should continue to use, as much as possible, core outcome sets and patient-reported outcomes for better data synthesis and meta-analysis.”
Reference:
Perspective
Back to Top
Linda A. Barbour, MD, MSPH, FACP
Accolades to Dunne and the EMERGE investigators in this JAMA publication reporting results of a double-blind, placebo-controlled trial with 510 participants at two centers in Ireland which randomly assigned predominantly white European women to metformin (mean dose of about 2 g) or placebo, at the time of gestational diabetes diagnosis (about 27 weeks). The investigators found no difference in the primary composite outcome (initiation of insulin or fasting glucose greater than 92 mg/dL) at 32 or 38 weeks. The initiation of insulin was 12.7% less in the metformin vs. placebo group (38.4% vs. 51.1%), although the mean insulin dose was actually not statistically different between groups. Other secondary outcomes included a modest but statistically lower fasting glucose at 32 weeks of 1.8 mg/dL (88.2 vs. 90 mg/dL) and 3.6 mg/dL at 38 weeks (81 vs. 84.6 mg/dL). Whether this difference could have been due to a reluctance in up-titration of insulin given that insulin dose was no different between groups is unknown.
Gestational weight gain was modestly less (1.2 kg) in the metformin group (0.8 vs. 2 kg). Consistent with some other trials, the mean birth weight was lower in the metformin group by 113 g (3393 vs. 3506) and large for gestational age was modestly lower by 8.4% (6.5% vs. 14.9%) as was a 7.2% difference in fetal macrosomia (7.6% vs 14.8%). Overall, the large for gestational age rate was relatively low, likely due to a median BMI of only about 30 kg/m2 and because WHO diagnostic criteria were used to diagnose gestational diabetes, which includes a population with less severe hyperglycemia than those diagnosed by Carpenter and Coustan, widely used in the U.S.
There was a non-significant trend for low birthweight (less than 2500 g) to be higher in the metformin group (6.1% vs. 3.4%) as well as small for gestational age (5.7% vs. 2.7%), but the overall rate of small for gestational age in both groups was low. Noteworthy, the mean crown-heel length was statistically lower in the metformin group, which raises some concern for a reduction in linear growth or lean body mass. There were no differences between groups in NICU admissions, RDS requiring support, or neonatal hypoglycemic (less than 47mg/dL).
Metformin crosses the placenta readily, is concentrated in mitochondria 1,000-fold, and suppresses mTOR, a growth-related kinase, nutrient sensor and a driver of anabolic processes in the cells. Suppression of mTOR can result in decreased cell proliferation, increased apoptosis and possibly, as the investigators mention, decreased amino acid transport across the placenta, which might compromise the development of lean body mass.
The EMERGE investigators appropriately call out for caution in using metformin in pregnant individuals with hypertension, preeclampsia or risk for fetal growth restriction, also supported by the finding of a higher small for gestational age rate in the Metformin in Type 2 Diabetes (MiTY) randomized controlled trial (Fieg DS, et al. Lancet Diabetes Endocrinol 2020;doi:10.1016/S2213-8587(20)30310-7). Small for gestational age infants, like large for gestational age infants, are also at risk for future childhood obesity, thought to be due to intrauterine nutrient restriction followed by postnatal exposure of nutrient excess. Metformin exposure in pregnancy was associated with a higher risk of overweight at age 5 to 10 years in some (Rowan JA, et al. BMJ Open Diabetes Res Care. 2018;doi:10.1136/bmjdrc-2017-000456, Hanem LGE, et al. Lancet Child Adolesc Health. 2019;doi:10.1016/S2352-4642(18)30385-7), but not all randomized controlled trials (Paavilainen E, et al. Diabetes Obes Metab. 2022;doi:10.1111/dom.14589).
Given the global and rapidly increasing impact of gestational diabetes, options for treatments other than insulin are of high importance. Whether the modest short-term benefits of metformin in this trial outweigh any potential increased risk of growth restriction or a reduction in infant lean body mass can only be answered in the long-term follow-up of this and other maternal-child randomized cohorts, as the investigators appropriately underscore.
Collapse
Dunne F. Metformin for the management of GDM: Friend or foe? Presented at: European Association for the Study of Diabetes Annual Meeting; Oct. 2-6, 2023; Hamburg, Germany (hybrid meeting).
Click Here to Manage Email Alerts