Editorial NexoScience
The biggest killer of mothers worldwide is not an infection, not a stroke and not blood that refuses to clot. It is a muscle that refuses to squeeze.
Postpartum hemorrhage complicates more than one in ten births and remains the foremost cause of maternal death on the planet, yet its biology has stayed stubbornly opaque. Genetic data from more than 300,000 women in six national biobanks across Northern Europe, reported July 22, 2024 in the journal Nature Genetics, has now placed five specific spots in the human genome behind it. Out of some 18 million points of DNA compared, only those five stood out, and the genes they appear to switch on and off (HAND2, TBX3, RAP2C and FRMD7) had never been linked to the condition before.
All five signals sat in places where progesterone gives the orders. None of the five variants alters a protein directly. Every one of them falls inside an enhancer, a stretch of DNA that works like a volume dial for a gene located somewhere else. When the researchers checked which proteins dock at those exact spots, progesterone receptors turned up far more often than chance allows. The tissues that lit up in the analysis were the endometrium, the lining of the uterus, and smooth muscle.
Progesterone keeps the womb calm, which is precisely the wrong reflex after delivery. All through pregnancy, progesterone relaxes the myometrium, the layer of smooth muscle that produces contractions, and stops it from pushing the fetus out too early. Minutes after birth, that same muscle has to become the tourniquet. It clamps down over the raw patch where the placenta was attached and pinches the open vessels shut. When it fails, the bleeding does not stop on its own. Uterine atony, the inability of the uterus to contract after delivery, accounts for roughly 70% of cases, with retained placental tissue, tears and clotting disorders splitting what is left. The genetic map suggests that in some women the hormonal brake is simply harder to release.
The clotting genes the field expected never showed up. Earlier candidate gene studies concentrated almost entirely on the coagulation cascade, and the oxytocin receptor was the other favorite suspect. Neither appears among the five loci. Women with diagnosed coagulation disorders had been excluded from the analysis, so what the study is describing is the bleeding that happens to women with no obvious reason to bleed, which is also the bleeding that current risk scores miss.
Big babies and long pregnancies leave behind an overstretched muscle. At the genetic level, the risk of hemorrhage travelled together with birth weight and with the length of the pregnancy, and moved in the opposite direction to uterine fibroids. Both fibroids and endometriosis, which share genetic territory with two of the five loci, are already treated with drugs that act on the progesterone receptor. A uterus stretched further and for longer, the authors reason, has a harder time contracting once it is empty.
Early bleeding and postpartum bleeding turned out to be strangers. Light bleeding in the first weeks of pregnancy produced no genetic signal at all, and showed no genetic overlap with hemorrhage after delivery. What it did correlate with was a scattered mix of reproductive, cardiovascular, socioeconomic and psychiatric traits, post-traumatic stress disorder among them. The authors read that pattern as a highly polygenic condition tied to a woman’s underlying health and circumstances, a different kind of problem wearing a similar name.
Knowing the genes is not the same as predicting the bleeding. When a genetic risk score was bolted onto a standard clinical model built from age, body mass index, previous births and previous cesareans, the model’s ability to tell who would bleed moved from 0.60 to 0.61, on a scale where 0.50 is a coin toss. That is a rounding error, not a test. The value of the finding is in the biology it exposes, not in the prediction it currently buys.
The study points at a pathway without proving the mechanism. A genome-wide scan of this kind is observational: it shows that certain DNA variants travel with the condition, not that they cause it. The gene targets were predicted from regulatory maps rather than confirmed in the laboratory. The effect appeared to run through the mother’s own genome rather than the baby’s, though the authors noted they could not fully rule out a fetal contribution. And all six cohorts came from Northern European countries with universal healthcare, while the deaths from postpartum hemorrhage fall overwhelmingly on women in poorer parts of the world, a limitation the authors flagged themselves.
What makes the result worth following is that drugs targeting the progesterone receptor already exist and are already prescribed for other uterine conditions. And the clinical need is specific: a recent review of prognostic models for postpartum hemorrhage found that nearly half of them rely on information that only becomes available after the birth, which is to say after it is too late to prepare.
Source: Westergaard D, et al., “Genome-wide association meta-analysis identifies five loci associated with postpartum hemorrhage,” published in Nature Genetics on 22 July 2024. DOI: 10.1038/s41588-024-01839-y
This article summarizes published scientific research for general information. It is not medical advice and does not replace consultation with a qualified health professional.
