Pregnancy and child

Decongestant sprays linked to rare birth defects

“Pregnant women who use nasal sprays for colds and hayfever increase the risk of rare birth defects,” reports the Mail Online.

This headline was based on a large study of infants with and without birth defects. It looked at whether their mothers said they had used decongestants during the first 12 weeks of their pregnancy.

It found a possible link between birth defects and two ingredients commonly found in decongestants (phenylephrine and phenylpropanolamine).

Most strikingly, the risk of a birth defect where the walls separating the chambers of the heart are poorly formed or absent was eight times higher in mothers taking phenylephrine. While this sounds alarming, as the Mail rightly points out, the increase should be placed into the wider context. The increase in risk translates into a 2.7 in 1,000 (0.27%) chance the baby would have the defect.

It is also worth noting that this study did not set out to prove that decongestant medicines cause birth defects, only whether the two may be linked. We can’t say for definite, on the basis of this study, that nasal sprays cause birth defects.

It is generally advisable to use as few over-the-counter medicines as possible during pregnancy.

Read more advice about using medication during pregnancy.

Where did the story come from?

The study was carried out by researchers from Harvard University, the National University of Singapore, and Boston University and was funded by Eunice Kennedy Shriver National Institute of Child Health and Human Development.

The study authors declare that the Pharmacoepidemiology Program at the Harvard School of Public Health (Harvard University) and the Slone Epidemiology Center (Boston University) receive support from various pharmaceutical companies, some of which could manufacture products included in these analyses. However, the report their current study analyses was not supported by any pharmaceutical manufacturer.

The study was published in the peer-reviewed American Journal of Epidemiology.

The Mail Online’s coverage was broadly accurate and included useful information highlighting how the risk increases uncovered in the study should be kept in perspective. As the birth defects involved were rare, any small changes in numbers can create a large sounding increase in risk.

What kind of research was this?

This was a case-control study looking at possible reasons for birth defects in infants born in the US or Canada.

The researchers were following up previous research suggesting that exposure to specific oral decongestants may increase the risk of several birth defects. They wanted to further investigate this possible link to see if it was true.

A case-control study is the favoured study design for investigating rare diseases such as birth defects. However, they cannot prove cause and effect. Instead, they can uncover potential links between how certain exposures during pregnancy may relate to differences in the chance of a birth defect occurring.

The ‘gold standard’ research design, a randomised controlled trial, could never be used to assess the possible harm of using nasal decongestants due to the obvious ethical implications of potentially placing a pregnancy at risk.

It would also be very impractical as you’d have to include a lot of pregnant women in the study to be able to include just a few cases of birth defects, because they are rare.

What did the research involve?

The researchers studied 12,734 infants with birth defects (the “cases”) and 7,606 without (the “controls”). Mothers of the infants were interviewed and filled in questionnaires within six months of delivery.

These interviews and questionnaires gathered information on reproductive, medical and lifestyle factors including details of exposure to all medications – prescription or over-the-counter – in the two months before pregnancy, and during pregnancy.

Detailed, highly structured interviews were used to boost the recall of the different medications used during this period. The specific medicines taken were identified so the researchers could look up their ingredients.

The mothers of the infants and the nurses who were interviewing them were not aware that the research was investigating decongestants (presumably to avoid the issue of recall bias).

Once they had all the information, the researchers compared the cases and controls to see if they had any noticeable differences in their exposure to different medicines, including decongestants.

The birth defects under investigation were:

  • eye defects
  • ear defects
  • ventricular septal defect (a defect in the wall dividing the two large chambers of the heart, the right, which pumps deoxygenated blood to the lungs, and the left, which pumps oxygenated blood around the body)
  • coarctation of the aorta (a narrowing of the large artery that leaves the left side of the heart and carries oxygenated blood to the body)
  • endocardial cushion defect (a defect where the walls separating the four chambers of the heart are poorly formed or absent)
  • pyloric stenosis (a condition where there is a narrowing of the pylorus – the stomach outlet – through which food needs to pass to reach the lower parts of the digestive system. This leads to vomiting, dehydration and failure to gain weight)
  • small-intestinal atresia/stenosis (where there is obstruction in the small bowel – either due to part of the small bowel not developing properly, or being narrowed)
  • clubfoot (a foot deformity where the foot points down and inwards)
  • gastroschisis (a defect in the abdominal wall meaning that the infant’s bowel and sometimes other organs are visible on the outside of the body)
  • hemifacial microsomia (where part of the face has not developed properly)

The analysis took account of a range of factors that could disrupt any potential link between decongestant medications and these birth defects, such as whether the mother smoked and the mother’s age.

As there were a lot of medications and a lot of birth defects under investigation, many comparisons and statistical tests were performed to find differences. This increases the chance that some statistically significant differences will be found by random chance alone. Using normal cut-offs around one in every 20 results will be statistically significant even if there is no real difference between the groups being compared.

What were the basic results?

From the large number of comparisons made between certain medicine ingredients and specific birth defects, three statistically significant links were found:

  • Mothers who used phenylephrine during the first trimester (first 12 weeks) of pregnancy were eight times more likely to have given birth to an infant with endocardial cushion defect than mothers who hadn’t used it (odds ratio (OR) 8.0, 95% confidence interval (CI) 2.5 to 25.3).
  • Mothers using phenylpropanolamine during the first trimester of pregnancy were 7.8 times more likely to have given birth to an infant with ear defects than mothers not using it (odds ratio 7.8; 95% CI 2.2 to  27.2) and 3.2 times more likely to have given birth to an infant with pyloric stenosis (odds ratio 3.2; 95%CI 1.1 to 8.8).

Both phenylephrine and phenylpropanolamine are commonly used in decongestant medicines.

To give a sense of the scale of the risks, the study authors were quoted in the Mail Online article explaining how “the risk of an endocardial cushion defect among babies whose mothers did not take decongestants is about 3 per 10,000 live births” and that “even the eight-fold increase in risk indicated by the study results, while it sounds large, would translate to a 2.7 in 1,000 chance the baby would have the defect”.

How did the researchers interpret the results?

The authors’ main conclusion was that, “accumulating evidence supports associations between first-trimester use of specific oral and possibly intranasal decongestants and the risk of some infrequent specific birth defects.”


This study suggests a possible link between phenylephrine and phenylpropanolamine – found in decongestant medicines – and an increased risk of three specific birth defects (endocardial cushion defect, ear defects, pyloric stenosis). Numerous other medications were tested but were not found to be associated with birth defects.

However, the study performed many statistical comparisons looking for links with many different birth defects. These three defects were the ones where significant links were found, but it is possible that some of the results may be due to chance alone.

The authors clearly recognised the limitations of their study, correctly reporting how their “hypothesis generating analysis involving multiple comparisons” found only a “small number of associations with oral and intranasal decongestants” and birth defects. This tells us the study was not looking to prove anything – and it does not. It was instead looking to uncover possible links that could be tested more rigorously in other research. To this end it achieved its aims.

The authors also usefully pointed out that “the associations identified involved defects that generally affect less than 1 per 1,000 infants. Some of them may require surgery, but not all are life-threatening.” So even though some of the relative risk increases sound large (such as eight times the risk) the chance of having a baby with a birth defect remained low in both groups.

Nonetheless, the study recruited a large number of infants (which is a strength) and confirmed past research that has suggested a similar link. This means it is likely to be investigated further to see if it the link holds up using more robust study designs.

The bottom line is that if you are pregnant and you have recently used a decongestant then there is no need to panic. Any potential risk to your pregnancy is likely to be minimal.

For more information, read advice about medication and pregnancy.

NHS Attribution