Vaginal swabs could be used to predict likelihood of preterm births
We often don’t know the causes of preterm birth – but analysing metabolic substances in the vaginal microbiome may be a way to predict the risk of premature labour
12 January 2023
Certain metabolic substances in the vagina may predict if a baby will be born premature, which is one of the leading causes of newborn death. Testing the vaginal microbiome during pregnancy could help healthcare providers identify and manage high-risk pregnancies.
Tal Korem at Columbia University in New York and his colleagues assessed 232 vaginal swab samples collected during the second trimester of pregnancy, 80 of which came from pregnancies ending in preterm birth. About 75 per cent of the samples came from people who identified as Black, a group disproportionately affected by premature births in the US.
Previous studies have implicated the vaginal microbiome in preterm birth, but unlike past work, Korem and his colleagues identified both the microbes and the metabolites in each sample. Metabolites are small substances made when foods, drugs or other chemicals degrade. The team used this data to train machine learning models to predict whether a pregnancy would end in preterm birth and found that certain metabolites, but not microbes, had the strongest association with preterm birth.
A model using data on metabolites and race had a 78 per cent chance of correctly identifying whether a sample was from a person whose pregnancy ended in preterm birth or not, whereas a model using data on microbes, race and other clinical factors had a 59 per cent chance.
Notably, many of the metabolites associated with preterm birth weren’t created by microbes or humans, says Korem. “They’re coming from an external source,” he says. While further analysis is needed to identify the source, some of these metabolites, like ethyl glucoside and EDTA, are present in certain cosmetic and hygienic products.
“This is far from proof,” says Korem, who notes these findings are only an association and don’t prove a causal relationship. Therefore, no recommendations can be made to avoid certain products. “This is very preliminary, but we’re sort of raising a flag and saying, look, there’s something suspicious here that we need to look into,” he says.
“This study, because they really focus on recruiting Black women, did a very good job in over-representing the underrepresented,” says Christine Metz at the Feinstein Institutes for Medical Research in New York. However, because the microbiome can change depending on various factors such as environmental exposures, it is possible the model could produce different results for samples collected from the same person on different days, she says.
Sign up to our free Health Check newsletter for a round-up of all the health and fitness news you need to know, every Saturday
More on these topics: