For decades it’s been thought that preconception use of the oral antidiabetic metformin by mothers and fathers might result in adverse fetal outcomes, including congenital malformations and stillbirths.
Women with type 2 diabetes (T2D) are often advised to switch to insulin before or during early pregnancy out of concern for fetal safety. But two studies from the Harvard T.H. Chan School of Public Health in Boston, Massachusetts — one in mothers, the other in fathers — report that metformin, a common and cost-effective antidiabetic agent, is not associated with a significant increased risk of teratogenicity and negative perinatal outcomes. The studies appear in Annals of Internal Medicine.
The studies may make it easier for physicians to reassure diabetic parents-to-be about the safety of metformin use before conception and in early pregnancy,
In the context of sparse existing safety data, the maternal analysis looked at Medicaid data on 12,489 mothers (mean age, about 30) receiving metformin for pregestational T2D during the period 2000-2018. “Many women become pregnant while still taking noninsulin oral antidiabetics, mostly metformin, and one safety concern is whether metformin could cause birth defects,” lead author Yu-Han Chiu, MD, ScD, an epidemiologist, said in an interview, commenting on the impetus for the study.
Dr. Yu-Han Chiu
“On the one hand, metformin can cross the placenta and might directly affect the fetus. On the other hand, poor blood sugar control is a risk factor for birth defects,” she continued. “Insulin in combination with metformin might control blood sugar better than using insulin alone, which may lower the risk of birth defects.”
Switched to insulin monotherapy or prescribed additional insulin within 90 days of their last menstrual period, mothers were assessed for nonchromosomal fetal malformations and nonlive births, spontaneous abortion, and termination. Continuing metformin or adding insulin to metformin in early pregnancy resulted in little to no increased risk for major malformations in infants.
The estimated risk for nonlive birth was 32.7% with insulin monotherapy and 34.3% with insulin plus metformin polytherapy, for a risk ratio (RR) of 1.02 (95% confidence interval (CI), 1.01-1.04).
In addition, the estimated risk for live birth with congenital malformations was 8.0% (5.70-10.2) under insulin monotherapy and 5.7% under insulin plus metformin (95% CI, 4.5-7.3), amounting to a risk ratio of 0.72 (0.51-1.09).
While the results may involve residual confounding by participants’ glycemic control and body mass index, Dr. Chiu said, “Our findings suggest that the current clinical recommendations to switch from metformin to insulin before pregnancy, due to concerns about birth defects, may require reconsideration.”
She noted that previous trials showed adding metformin to insulin in mid-late pregnancy also improved blood sugar control with no increase in risk of birth defects. “However, most of these studies started treatment too late — between 10 and 34 weeks of pregnancy — to determine if metformin could cause birth defects.”
Observational studies found that women with pregestational diabetes who used noninsulin antidiabetics (mainly metformin) in the first trimester had a lower risk of birth defects, compared with those who used insulin, Dr. Chiu added. “However, comparing metformin with insulin may have some biases because women who used metformin generally have less severe diabetes than those who used insulin.”
Aligning with these reassuring findings, a randomized, placebo-controlled trial reported that adding metformin to insulin did not lead to a higher incidence of neonatal morbidity and mortality and was associated with better maternal glycemic control and reduced maternal weight gain. Metformin-exposed offspring, however, had lower birth weights and a higher incidence of being small for gestational age.
Similarly, a recent Nordic register study of more than 3.7 million infants also found no evidence of an increased risk of major defects with the use of metformin vs insulin in the first trimester.
Despite such reassuring findings, however, Dr. Chiu stressed the need to study other pregnancy and infant outcomes as well as the safety of other oral antidiabetics during pregnancy.