People are always saying the wrong thing to pregnant women.

Expectant mothers hear everything from the obnoxious (“You’re huge!”) to the outright bizarre (“If you eat that Sriracha, your baby will come out bald”).

Then there are the well-meaning—yet utterly unhelpful—superstitions and platitudes: “I can tell from how you’re carrying that it’s a girl.” (No, you can’t.) “At least the terrible sleep you’re getting now is great preparation for all those sleepless nights you’re going to have with baby!” (Bone-splitting exhaustion is not something you need to practice ahead of time.) “But morning sickness means your baby is healthy!”

Actually, there might be something to that last one.

Pregnant women have long been told that feeling miserable every single day for several months may indicate that a developing baby is doing well—especially in the first trimester, when nausea and vomiting are most common. Now, there’s more science to support the idea.

A new study from researchers at the National Institutes of Health, and published today in JAMA Internal Medicine, finds morning sickness is associated with a lower risk of miscarriage in pregnant women who have experienced previous losses. The NIH calls the research, which builds on similar studies, “the strongest evidence to date” that nausea and vomiting in pregnant women are associated with a lower risk of pregnancy loss.

The latest NIH study, which was a secondary analysis of data from a separate clinical trial, involved tracking symptoms logged daily by nearly 800 pregnant women. All of the women in the study had at least one previous pregnancy loss, with about one-third of the participants having experienced two losses.

About 84 percent of the women reported nausea, with or without vomiting, by the time they were eight weeks pregnant. (Smaller percentages of women had morning sickness earlier in pregnancy—with about 20 percent of them reporting illness at two weeks pregnant, and more than 50 percent of them reporting nausea or vomiting by five weeks.) Nearly one-quarter of the pregnancies resulted in miscarriage, many of which occurred before the eight-week mark.

Overall, the women who reported nausea by itself or nausea with vomiting were between 50 percent and 75 percent less likely to miscarry than those who didn’t feel sick.(Earlier research, including a 2014 meta-analysis of 10 separate studies conducted between 1992 and 2012, has also found that women who had morning sickness experienced fewer miscarriages and gave birth to larger, healthier babies with fewer birth defects.)

But even though earlier studies have found similar associations, few researchers have taken into account the other potential indicators for miscarriage among study participants—like the number of previous pregnancy losses a woman has experienced, alcohol intake during pregnancy, and fetal characteristics such as chromosomal abnormalities. Such factors might increase the likelihood of miscarriage, even among women who experience morning sickness. The NIH researchers controlled for these and other circumstances in their assessment, giving them a clearer picture that the association between morning sickness and reduced pregnancy loss is strong—without confounding factors getting in the way.

The NIH study was unusual, too, in that it began with women who were still trying to conceive—rather than first enrolling participants who were already pregnant. “This is important because it allowed us to get detailed data from diaries that women were keeping about their symptoms in the earliest weeks of pregnancy—even before most women knew they were pregnant,” said Stefanie Hinkle, a staff scientist at the NIH, and the lead author of the study. “We found that in the week after conception, one in five women were already experiencing some nausea symptoms.”

The protective benefit of morning sickness was stronger among women who were throwing up compared with those who just felt crummy—up to a point, anyway. Researchers didn’t include findings from women who experienced hyperemesis, a severe form of nausea and vomiting during pregnancy that can lead to hospitalization—meaning the study’s findings do not apply to women with the very worst symptoms. The study had some other limitations, too. The vast majority of participants were married, highly educated, white women—so the extent to which the findings would apply to women in other demographics is unclear.

Researchers still don’t understand what mechanism is responsible for the protective association between feeling miserable and having a healthy pregnancy, though they have some ideas about what might cause the nausea in the first place. Scientists have suggested that pregnancy illness is a byproduct of rapid hormonal changes, especially the increase of human chorionic gonadotropin, or hCG. (The presence of hCG in a woman’s urine is what makes a home pregnancy test positive.)  “Another possibility is that nausea and vomiting are markers for viable placental tissue,” Hinkle and her colleagues wrote in the NIH paper. “Thus, less nausea and vomiting may identify failing pregnancies.”

Scientists are also unsure about whether there’s an evolutionary reason for morning sickness to carry protective benefits. It’s possible, many have theorized, that nausea is a way of keeping a pregnant woman from doing (or eating) things that might harm her fetus. “First, symptoms may be part of an evolutionary advantage to change one’s dietary intake, increase consumption of carbohydrate-rich foods, or avert intake of potentially teratogenic substances,” the NIH researchers wrote, referring to drugs or other agents that can cause birth defects. But since their model accounted for smoking and drinking alcohol, it seems “the mechanism is likely not through avoidance of such substances.”

The NIH findings should be reassuring to women enduring morning sickness (which, as any pregnant woman can tell you, also happens at night, and in the afternoon, and sometimes around the clock)—but sickness isn’t a guarantee of a viable pregnancy. Some women who experience nausea and vomiting will still go on to have miscarriages or stillbirths. Similarly, Hinkle told me, an absence of symptoms isn’t automatically cause for concern.

“Not all pregnancies are the same and every individual is different,” she said, “So just because they do not have symptoms does not mean that they will go on to have a loss.”