No clear evidence hormone treatment in early pregnancy helps prevent miscarriage

Pregnancy and child
"Hormone 'can reduce chances of miscarriage'," reports BBC News. The rather misleading headline follows a ...

"Hormone 'can reduce chances of miscarriage'," reports BBC News. The rather misleading headline follows a ...

"Hormone [treatment] 'can reduce chances of miscarriage'," reports BBC News. The rather misleading headline follows a trial looking at whether giving the hormone progesterone to women with bleeding in early pregnancy could prevent miscarriage.

Miscarriage is defined as loss of pregnancy before 24 weeks, and affects 1 in 5 pregnancies. There are many potential causes. Progesterone maintains the lining of the uterus and supports the placenta during pregnancy.

Previous studies have looked at whether giving women progesterone supplements in early pregnancy can:

  • prevent miscarriage in women who've experienced miscarriage before, or
  • "rescue" a pregnancy when women have started to see vaginal bleeding

However, these studies have been small and the results inconclusive.

This recent study included more than 4,000 UK women randomised to progesterone or a dummy treatment (placebo). The main finding was that progesterone did not make a significant difference to the number of women who went on to have a baby, which was 75% of the progesterone group and 72% of the placebo group.

Delving into the results, researchers found that progesterone might be helpful for women who'd had 3 or more previous miscarriages – but this was based on a small number of women and needs looking into further. The smaller the amount of people in a group often means the result is less likely to be reliable.

Therefore there's no clear evidence that progesterone supplements can prevent miscarriage.

Find out more about miscarriage.

Where did the story come from?

The doctors who carried out the research came from 23 universities, hospitals or charities in the UK, led by a researcher at the University of Birmingham, and one in Australia and one in the US. The study was funded by the UK's National Institute of Health Research and the charity Tommy's. It was published in the peer-reviewed New England Journal of Medicine on an open access basis so is free to read online.

The positive headlines in the UK media focus on results from only one small subgroup of the 4,153 women in the study – the 301 women who had 3 or more miscarriages before taking part in the study. The Mail Online headline says that "thousands of babies' lives could be saved" and that the treatment "slashes a woman's risk of a miscarriage".

This reporting ignores the main conclusion from the study, which was that the treatment did not make a difference to the risk of miscarriage for most women with bleeding in early pregnancy.

What kind of research was this?

This was a randomised controlled trial (RCT), which is the best type of study to find out whether a treatment works or not as it should be able to balance any differences in characteristics between groups.

This trial also has additional strengths, including its very large sample size and the fact that it was double blinded, meaning neither the women nor researchers would know whether the women were taking progesterone or placebo.

What did the research involve?

Researchers recruited women across the UK who were in early pregnancy (less than 12 weeks), aged 16 to 39 and had bleeding from the vagina (which can occasionally be a warning sign for miscarriage; though vaginal bleeding is common during the first trimester of pregnancy). All the women had an ultrasound scan to check the pregnancy was visible in the womb.

The women were assigned randomly to twice-daily pessaries (tablets to put into the vagina), containing either 400mg progesterone or placebo. They were told to continue with them until they had completed 16 weeks of pregnancy.

Researchers followed them up to see how many gave birth successfully after 34 weeks. They also recorded birthweights, gestational age and any miscarraiges, stillbirths or congenital abnormalities.

They looked at 10 sub-groups of women, to see if their results differed from the overall results. These included groupings by age, body mass index (BMI), weeks of pregnancy, bleeding and number of previous miscarriages.

What were the basic results?

Of the 4,153 women recruited:

  • 1,513 of the 2025 (75%) women who received progesterone had a baby
  • 1,459 of the 2013 (72%) women who did not receive progesterone had a baby

The difference between the two groups was too small to be sure it was not down to chance (relative risk 1.03, 95% confidence interval [CI] 1.0 to 1.07.

There was no difference in adverse outcomes, such as congenital abnormalities.

The only 1 of the 10 subgroup analyses to show any difference from the overall results was analysis by previous miscarriage.

The results showed that women who'd never had a miscarriage – or only 1 or 2 miscarriages – did not benefit from receiving progesterone. However, there was a difference for those women who’d had 3 or more previous miscarriages:

  • 98 of the 137 women (71.5%) who received progesterone had a baby
  • 85 of the 148 women (57.4%) who did not receive progesterone had a baby

The chances of having a baby for these women increased by 28% (risk ratio 1.28, 95% CI 1.08 to 1.51) if they were given progesterone.

Because the numbers in this group are small and multiple calculations have been performed, we have to be cautious about drawing firm conclusions from the results of this subgroup.

How did the researchers interpret the results?

The researchers concluded: "Treatment with progesterone did not result in significant improvement in the incidence of live births among women with vaginal bleeding during the first 12 weeks of pregnancy."

Of the results in women with 3 or more previous miscarriages, they said there was "a suggestion of benefit" but that "we did not identify this subgroup as one of special interest a priori [before the study started] in our statistical analysis plan, and multiple comparisons were performed (without adjustment for multiplicity); thus, this observation requires validation".


The headlines are misleading and do not reflect the main findings of this study.

The most reliable result of any trial is always the one that it was designed to look at. In this case, for women with bleeding in early pregnancy, the researchers wanted to see whether progesterone affected the chances of having a live birth. This study suggests that progesterone treatment is unlikely to make a difference to their chances of having a miscarriage or of their pregnancy having a successful outcome.

The results do suggest that women who have had 3 or more miscarriages may benefit from progesterone treatment. However, as the researchers acknowledge, they did not design the study to look at the effects of progesterone in this group of women. This means the study may not have included sufficient women with this experience to be a reliable finding for this very small subgroup.

And it should be noted that while these women had an improved chance of having a baby compared with others in this group who did not take progesterone, they still had a lower chance of having a baby than women with fewer previous miscarriages.

The researchers are quoted as saying that national UK guidelines on miscarriage prevention could now be updated due to their findings. But it's arguably hard to justify this based on the weak strength of evidence presented in this study.

Further research is needed before we know whether progesterone could benefit some women with recurrent miscarriages.

Sadly, miscarriage is fairly common. The cause of miscarriage is not known in most cases, and most could not have been prevented. There are things you can do to improve your chances of a successful pregnancy, however.

Find out more about lowering the risk of miscarriage.

Article Metadata Date Published: Do, 9 Mai 2019
Author: Zana Technologies GmbH
NHS website