Roughly 70 percent of pregnant people experience morning sickness: bouts of nausea or vomiting, or both, that put them off their food and send them running for the toilet. Despite its name, the miserable condition can strike at any time of day—or last all day. It usually subsides after the first trimester, though it can sometimes linger throughout an entire pregnancy.
Up to 3 percent of people who are pregnant experience a severe and sometimes life-threatening form of morning sickness known as hyperemesis gravidarum (HG), which makes it extremely difficult to keep down food or liquid. This can cause severe dehydration and can sometimes lead to a hospital stay. Catherine, Princess of Wales (formerly Kate Middleton) and comedian Amy Schumer have both suffered from the condition.
Marlena Fejzo, an assistant professor of population and public health sciences at the Keck School of Medicine of the University of Southern California, was recently awarded the BioInnovation Institute & Science Translational Medicine Prize for Innovations in Women’s Health for her work on the genetics of HG. “We put men on the moon decades ago, but women are still dying from severe nausea and vomiting during pregnancy,” Fejzo wrote in an essay accompanying the award.
On supporting science journalism
If you’re enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Fejzo got interested in studying HG after suffering from the condition during both of her pregnancies. In collaboration with the consumer genetics company 23andMe, she conducted an analysis of people with HG that identified two genes involved: GDF15 and IGFBP7.
Fejzo and her colleagues published a 2023 study in Nature that confirmed the role of the hormone GDF15—whose production is controlled by the GDF15 gene—in HG and milder morning sickness. Most people produce the GDF15 hormone in response to physiological stress, even when they’re not pregnant, but people with HG have a version of the gene that prevents this. During pregnancy, the placenta—which develops from embryonic tissue—produces GDF15 at high levels that can trigger morning sickness. People with HG are hypersensitive to GDF15, so the effects are severe.
Fejzo is currently focused on investigating potential treatments for HG, including the diabetes drug metformin, as well as more targeted antibody therapies.
Scientific American spoke with Fejzo about hyperemesis gravidarum, the discovery of GDF15 and progress toward treatments for the debilitating condition.
[An edited transcript of the interview follows.]
What is hyperemesis gravidarum?
It is nausea and vomiting that are debilitating. It affects your daily routine, and you can’t eat or drink normally, and patients become dehydrated, so they generally will have weight loss and need to have IV fluids for dehydration and electrolyte imbalances—and they will, in more severe cases, need hospitalization and nutrition. Most cases need intervention with antinausea medications that currently don’t work very well and are pretty much used off-label to treat the condition [meaning such drugs are generally not officially approved for such treatment].
It is the primary cause of hospitalization in early pregnancy and the second leading cause of hospitalization in pregnancy overall after preterm birth (tied with gestational hypertension). For it to be so common, it’s really shocking how little is known.
What interested you in studying this condition?
I had already been studying women’s health as my focus in my career. And then I had hyperemesis in my pregnancy. It was so severe in my second pregnancy that I couldn’t move without violently vomiting. It really was torture. I just had to lie down and stare at the ceiling. I couldn’t get up. I couldn’t even sit up. I couldn’t get up to go to the bathroom or brush my teeth or shower or anything for weeks and weeks. My doctor gave me seven different medications at once and IV fluids, but nothing helped get me to be able to eat. Eventually, I was put on a feeding tube. And I ended up losing the baby in the second trimester.
After that, I looked into what was known about the condition, and there was so little known. I decided to start looking into whether it was genetic. I didn’t have it in my family, so I didn’t know. I partnered with Kimber MacGibbon, director of the Hyperemesis Education and Research [HER] Foundation, which had a great website with information on HG that people from all over the world were going to. We posted surveys on the HER Foundation website and started to look for answers to questions about HG, such as what the recurrence risk was and whether it ran in families. Those two things would provide evidence as to whether it was genetic or not.
What were some of your findings about the genetic basis of HG?
We found a high recurrence risk. More than 80 percent of the patients in our study had HG in a second pregnancy after the first one. Then we did a familial aggregation study and found that it did aggregate in families, so there’s a 17-fold increased risk of having it if your sister has it. We started a study to collect DNA saliva samples in addition to survey data from patients. We asked them to recruit unaffected controls—friends or acquaintances who did not have HG in their pregnancies—so that we could do a genetic study. I applied for funding to do the study on my cohort but was denied. But I was fortunate in that I got a kit for my brother from 23andMe (which just went out of business [last month]). They had a brilliant model to invite their customers whose DNA was already sequenced to participate in surveys. I called 23andMe and asked them to add hyperemesis questions to their survey, and they agreed. It was really fruitful, and we published our first paper that showed the link between both hyperemesis and normal nausea and vomiting.
It showed that both hyperemesis and regular nausea and vomiting are very strongly associated with this nausea and vomiting hormone GDF15. After I published that, I was able to get my study participants sequenced by the pharmaceutical company Regeneron. And in that cohort, we found a mutation in GDF15 that increased risk of hyperemesis almost 10-fold. So that really helped to solidify the association between this hormone and hyperemesis, because it was a rare mutation that you’re born with, and then you get the disease. But when I was looking at those patients with the mutation, some of them didn’t have hyperemesis in every pregnancy. My hypothesis was that whether the mother would have hyperemesis depended on whether the baby inherited the mutation or not. So I started sequencing the children from these women who had the mutation, and I got the surprising result that they were less likely to have hyperemesis if the baby inherited the mutation.
So even though some people have the gene that prevents production of GDF15 before pregnancy, if their baby’s placenta doesn’t produce it either, the pregnant person wouldn’t experience HG symptoms?
Exactly. That’s why, even though the mothers had lower levels of GDF15 before pregnancy that made them hypersensitive to the hormone, they would have less chance of getting hyperemesis—because they had lower levels during pregnancy if the baby inherited the gene.
At that time, I partnered with Steven O’Rahilly, an endocrinologist at the University of Cambridge, and we worked together to solve this really perplexing finding. The three genetic variants I had identified were actually associated with producing lower levels of the nausea and vomiting hormone rather than higher levels, which was surprising—but then also exciting because it meant that maybe there was desensitization going on in people who had higher levels, and that could be protective. In our Nature paper, we proved that is likely the case, both in a mouse model that O’Rahilly did and also with evidence from humans. It’s long been known, for example, that chronic smokers have a lower risk of nausea and vomiting in pregnancy and hyperemesis. It’s also known that chronic smokers have high levels of circulating GDF15 because it is actually a stress-response hormone. It’s produced whether you’re pregnant or not, and whether you’re male or female, from cells or in tissues that are under stress. That’s not to say people should start smoking before pregnancy! We’re looking into other methods that might be safer.
Is there any evolutionary reason for why we have more GDF15 in our body during pregnancy?
Back in ancient times, millions of years ago, even hundreds of years ago, going out to hunt for food was fraught with risk. So I think this evolved as a mechanism, when you are in some kind of sensitive state, to stop you from going out to eat, to stay in your cave, rest and recover or to get through that first part of pregnancy—rather than wandering far and going out to get foods where you could be attacked by a predator in a weakened state or you could, in pregnancy, eat something poisonous.
I always use the example of the octopus that lays its eggs and then starves to death and dies. And the other gene that I found, IGFBP7, is the same gene that gets [dialed up] in that octopus. It’s like an age-old mechanism where, in the case of the octopus, it’s clearly evolutionary beneficial for the mother to die in order to save the eggs. It’s an extreme example, but it still goes on in nature. But some animals, and I think humans, just don’t need it anymore. And I think it also wears off at a time when the nutritional needs of the fetus outweigh the risk to the fetus from not eating.
Is it true that morning sickness symptoms are associated with a lower risk of miscarriage? And does that suggest the ability to produce GDF15 is a good thing?
I think it’s just showing that the pregnancy is progressing. The placenta produces GDF15, so the more cells you have of the placenta, the more GDF15 you’re going to produce, right? So in those studies where they show that morning sickness is good, it’s more that no morning sickness may mean that your placenta and the fetus are not growing, and so you’re more likely to lose the baby.
There are actually human knockouts [people who lack the gene entirely] out there. There is a population of people who married their first cousins, in which both partners carry the mutation, and then their offspring have the knockout of the hormone. And they’re fine. They have a normal lifespan and normal fertility. That suggests that we really don’t need this hormone anymore.
Has there been any progress on treatments for HG? I believe you’ve studied the diabetes drug metformin?
I haven’t published it yet, but I have done a retrospective study of metformin use prior to pregnancy and risk of hyperemesis, and that showed positive results. Metformin increases GDF15 levels. It’s also been used, for example, to increase fertility in polycystic ovarian syndrome patients. So it’s been used in the same time frame we would want to use it in people who have a history of hyperemesis: prior to pregnancy. It’s also used to treat gestational diabetes, so there’s quite a lot of evidence out there on its safety, though more study is needed to understand the potential effects on fetal growth. And it’s also available in generic form, so that’s also great, as far as it being an equitable approach.
I just initiated a prospective study of metformin in patients. I don’t have the results of that yet. The company [NGM Biopharmaceuticals], which I have been working with, just announced that they treated their first patient with a drug that is an antibody to the receptor for GDF15. And they’ve initiated a clinical trial, too. We are on the cusp of finding out whether these approaches are going to work. So it’s an exciting time.