GLP-1 medicines in early pregnancy: the InPreSS birth defect study

Why this study matters

More people of childbearing age are using GLP-1 medicines, both for type 2 diabetes and for weight management. Some pregnancies are unplanned, so a baby may be exposed to the medicine before the pregnancy is known. This study is one of the largest to look at whether that early exposure is linked to major birth defects. It does not include Canadian data, but its findings are relevant to anyone in Canada who becomes pregnant while using one of these medicines.

Who was in the study

The study included pregnant people with type 2 diabetes whose babies were born alive. Data came from four Nordic countries (2009 to 2020), the US MarketScan insurance database (2012 to 2021), and the Israeli Maccabi Health Services database (2009 to 2020). Babies were followed for up to one year after birth.

In total there were 51,826 infants born to people with type 2 diabetes. Of these, 938 had been exposed to a GLP-1 receptor agonist and 5,078 to insulin. Exposure was counted if there was at least one prescription fill from 90 days before pregnancy to the end of the first trimester, the first three months of pregnancy.

What the researchers did

This was an observational cohort study. Cohort means the researchers followed groups of people who were already taking different medicines, rather than assigning them by chance. Because there was no random assignment, the groups can differ in ways that affect the result. The researchers adjusted for known differences using statistical models, then combined the results from each country.

The main outcome was major congenital malformations, serious birth defects present at birth. Insulin was the active comparator, meaning GLP-1 medicines were compared with insulin rather than with no treatment. This helps because everyone in both groups had diabetes that needed treatment. The study also looked at three other classes of diabetes medicines.

What they found

Major birth defects occurred in 3.7% of all infants in the source populations (3,514,865 babies) and in 5.3% of infants born to people with type 2 diabetes. Among exposed infants the standardized rates were:

A relative risk of 1.0 would mean the same risk as insulin. A confidence interval is the range in which the true value most likely sits. For GLP-1 medicines the interval runs from 0.72 to 1.26, which includes 1.0, so the study did not find a clear increase or decrease in risk compared with insulin (R32-C123).

The authors conclude that the results did not indicate a large increased risk of birth defects above the risk already linked to having type 2 diabetes that needs second-line treatment. They describe the findings as reassuring but call for confirmation from other studies.

Side effects and people who stopped

This study looked only at birth defects in babies. It did not measure side effects in the pregnant person, and the idea of stopping treatment does not apply in the same way as in a trial. The extracted data do not report discontinuation.

What this study does not tell you

• Only live births were included. Pregnancies that ended in miscarriage, stillbirth or termination were not counted, which could hide some effects.
• Exposure was based on prescription fills, not on whether the medicine was actually taken.
• Everyone in the study had type 2 diabetes. The results do not directly apply to people using a GLP-1 medicine for weight management without diabetes.
• The groups were not randomized, so other differences between people on GLP-1 medicines and people on insulin may have influenced the result.
• The confidence interval for GLP-1 medicines, 0.72 to 1.26, is fairly wide. The study cannot rule out a modest increase in risk.
• The study looked at exposure around conception and in the first trimester. It cannot say whether continuing a GLP-1 medicine throughout pregnancy is safe.

What it means in Canada

This study did not include Canadian data, but the populations studied are broadly comparable. It describes what was seen in babies after early exposure and does not tell anyone what to do about their own medicine before or during pregnancy. For information about specific medicines, see semaglutide in Canada and tirzepatide in Canada. For what provincial plans pay for, see public drug coverage by province.

Related reading

Other studies in this library that bear on the same question.

Source