Melanie See’s first bout of odd symptoms began in 2005. Suddenly she started sweating a lot. She rapidly lost 10 pounds. She got dizzy walk ing from the bedroom to the couch. She started lactating even though she was not nursing a baby. After a slew of laboratory tests, See, then 45, was diagnosed with Graves’ disease, an autoimmune disorder that makes thyroid hormones surge.
Three years later, when See’s symptoms from Graves’ were under control with medication, her health took another rapid downturn. She lost more weight. She felt extremely tired. Her doctors diagnosed her with celiac disease, another autoimmune disease, which in affected people is set off by eating foods with gluten. Then, in 2015, See, who lives in Chapel Hill, N.C., began experiencing terrible digestive symptoms and muscle pain. This time her doctors were stumped. “Initial diagnoses were all over the place—vasculitis, lupus, I can’t remember what all,” See says. “My bloodwork showed something was going on, as did the muscle biopsy I had in June 2016, but I didn’t fit into any particular box.”
After many tests, See was diagnosed with yet a third autoimmune illness: mixed connective tissue disease, a rare ailment that shares some features of lupus.
Women account for an estimated—and astonishing—78 percent of people who have these disorders, which include See’s afflictions, as well as lupus, multiple sclerosis, rheumatoid arthritis, and other illnesses in which the body’s immune system mistakenly attacks its own cells and tissues. Autoimmune diseases are now the fifth-leading cause of death in women younger than 65.
Why women are so much more likely than men to be plagued by autoimmunity has long been a mystery, but researchers are beginning to narrow down the causes: the different effects of sex hormones, of women’s X chromosomes, and even of the community of microbes inside us, which develops differently depending on sex. Evolution may also play a role in the staggering differences observed in autoimmunity, some scientists think. Because autoimmunity is much more common in women, researchers have suggested it might be an evolutionary relic—immune hypervigilance could have given women a reproductive advantage by improving the chances of a successful pregnancy, even if it came at the cost of increased disease risk.
“It is important to understand the underlying biology of these sex differences,” says Shannon Dunn, an immunologist at the University of Toronto. “If we can unravel this, we will not only better comprehend how autoimmune diseases get started and find new ways for intervention, but we will also shed light on the sex differences in how humans respond to infection, vaccination, injury and cancer.”
The wildly uneven burden of autoimmune diseases is not a new observation. Well over a century ago, when doctors first began diagnosing these conditions, they noticed that women were much more likely to develop such illnesses than men. But back then doctors tended to think of individual autoimmune diseases as distinct entities with their own unique causes. There was little awareness that they might all be connected in fundamental ways and that they might affect women more often for these shared biological reasons.
Everything changed in the early 1990s, when scientists found that some autoimmune diseases have biological mechanisms in common. Among other things, researchers discovered that immune cells known as CD4+ T helper cells were involved in rheumatoid arthritis, multiple sclerosis and type 1 diabetes. And in 1991 a woman with lupus named Virginia Ladd founded the American Autoimmune Related Diseases Association after discovering that a number of autoimmune diseases plagued various members of her family, suggesting a shared genetic inheritance.
Once researchers began to think of autoimmune diseases as a collection, they started to notice interesting patterns. One was that some of these conditions arise in women after key life transitions. (Almost all of this research has involved cisgender women.) Lupus and multiple sclerosis, for instance, tend to first appear during the childbearing years. Other diseases, such as rheumatoid arthritis, most commonly arise after menopause. Big autoimmune changes can also take place during pregnancy: symptoms in women with rheumatoid arthritis, multiple sclerosis and Graves’ disease often wane during pregnancy, whereas in women with lupus symptoms often get worse.
What do all these transitions—puberty, pregnancy and menopause—have in common? They all involve major changes in the hormones estrogen, progesterone and testosterone. Estrogen levels go up, for instance, during puberty and pregnancy. It is now clear that, although there are exceptions, many autoimmune diseases “are driven by estrogen,” says DeLisa Fairweather, a microbiologist and immunologist at the Mayo Clinic in Jacksonville, Fla. Indeed, the use of oral contraceptives and hormone-replacement therapy, both of which add estrogen to the body, has been linked to an increased risk for lupus.
Estrogen, like the other sex hormones, directly influences the expression of a number of genes involved in immunity. For instance, it attaches to and turns on the gene that codes for interferon gamma, a chemical that orchestrates immune responses against pathogens but that can also escalate autoimmune responses. Estrogen also activates B cells, which produce antibodies, proteins that mark and attack foreign substances. But some, known as autoantibodies, can also attack cells made by the body.
Hormones that play key roles in pregnancy, such as progesterone, have dramatic immune effects, too. Many critical immune cells, including T cells and macrophages, feature receptors for progesterone on their surface. When progesterone binds to these receptors, it shifts the body toward a kind of immune response that favors the production of antibodies and autoantibodies. This reaction is known as a Th2 immune response, for type 2 T helper cells. It contrasts with Th1 immune responses, which move the body away from antibody manufacturing and instead activate cells that attack other cells directly.
The rise in progesterone during pregnancy could explain why symptoms of rheumatoid arthritis and multiple sclerosis often wane when women are expecting—these diseases are driven by Th1, not Th2, immune responses, so the progesterone-induced shift eases their immune burden. But “women with multiple sclerosis are at a much higher risk of having a relapse shortly after delivery. And that has to do with the dramatic change and reduction in sex hormones,” says Tanuja Chitnis, a neurologist at Brigham and Women’s Hospital in Boston.
Testosterone, which women produce but to a lesser degree than men, is another important hormone when it comes to autoimmunity. Receptors for testosterone are found on the surface of B and T cells, and the hormone is largely immunosuppressive. It decreases the responses of immune cells, including neutrophils, natural killer cells and macrophages—which could be one reason that men tend to have lower rates of autoimmune disease. Research has found that men with multiple sclerosis often have lower than normal levels of testosterone and that men with low testosterone because of a condition known as hypogonadism are at increased risk for lupus and rheumatoid arthritis.
All these sex hormones can also affect the expression of key immune genes. In 1997 a consortium of Finnish and German scientists discovered a gene that plays a crucial part in autoimmunity. This gene, which they named AIRE, for “autoimmune regulator,” is expressed by cells in the thymus, an organ that makes T cells. AIRE ensures that key body proteins are shown to developing T cells, and these encounters teach T cells that the proteins are friends, not foes. Also thanks in part to AIRE, T cells that start attacking these friendly proteins are destroyed in the thymus before they can be released into the rest of the body, where they could do damage.
Not surprisingly, people in whom AIRE is missing or mutated are more likely to develop certain autoimmune diseases. That is because T cells that should be eliminated are not, and “they end up going out into your body and causing autoimmune disease,” Dunn says.
As it turns out, AIRE’ s activity—and that of other, similar genes—is partially controlled by sex hormones. In a 2016 study, researchers at the Sorbonne in Paris showed that in mice, estrogen and progesterone turn down AIRE expression, meaning they cause less of the protein it encodes to be made, where as testosterone ensures that more AIRE protein is made. The researchers also found that after puberty, women tend to make less AIRE than men do, perhaps because of the influence of sex hormones. Less AIRE means that more self-reactive T cells can escape from the thymus and cause autoimmune disease.
Yet despite their influential roles, sex hormones cannot be the whole story. Autoimmune diseases, including lupus and multiple sclerosis, sometimes develop in childhood, before hormones such as estrogen and progesterone ramp up during puberty. This means that other processes must be involved. To find them, some researchers are studying a primary difference between men and women that arises well before birth: the presence or absence of a second X chromosome.
Biological dogma holds that women have two X chromosomes, but one copy is turned off in every cell very early in embryonic development in a process known as X inactivation. The extra X chromosome becomes a dark, misshapen mass that persists silently in each cell lineage. This shutdown ensures that the body does not express more X-linked genes than it should. But in recent years scientists have discovered that X inactivation does not happen the way they thought it did. Studies show that at least 15 percent of the genes on the supposedly inactivated X chromosome are still turned on, which means, essentially, that those genes tell women’s bodies to make twice the amount of certain proteins compared with levels in men. In women with lupus, for instance, some genes are active on both X copies, and this higher activity correlates with disease severity: sicker lupus patients have more active X-linked genes than women with milder disease.
In fact, many X-linked genes have been directly tied to autoimmune disease. One of them is a gene for toll-like receptor 7, or TLR-7, a protein that has been implicated in autoimmune disorders such as lupus, polymyositis, scleroderma and Sjogren’s syndrome. TLR-7’s job is to recognize pathogens and alert other immune cells to their presence; it also increases the production of inflammatory immune chemicals known as interferons, which can ramp up the autoimmune response. Another gene that is often activated on supposedly inactivated X chromosomes in women is TASL, and it, too, increases interferon production, to the point where women have at least twice as much of the protein, says Hal Scofield, a physician scientist at the University of Oklahoma Health Sciences Center who studies the role of X inactivation in autoimmune disease.
Recently scientists uncovered something bizarre about X inactivation that also supports its role in autoimmunity. Women’s inactive X chromosome is maintained in an especially strange way in T and B cells, which are actively involved in immune responses. In 2019 Montserrat Anguera, a biomedical scientist at the University of Pennsylvania, and her colleagues observed that when young immune cells in female mice mature, the cellular mechanisms in place to cover and inactivate their second X chromosome undergo significant, dynamic changes that could make it easier for X-linked genes in these cells to get turned on when they should be off. It was a “crazy discovery,” Anguera says.
No one thought that females’ immune cells did anything different with regard to X inactivation than other cells did, but it turns out they do—in ways that could directly shape the risk for autoimmunity. In June 2021 Anguera and her team found that B cells in girls and women with lupus evade the normal cellular mechanisms for X inactivation, which likely allows the cells to make more X-linked proteins than they should.
What happens to people with unusual numbers of X chromosomes also points to their important role in autoimmunity. Men with Klinefelter’s syndrome, for instance, have two X chromosomes along with a Y chromosome, and they are 14 times more likely than other men to develop lupus. Similarly, women with trisomy X, who have three X chromosomes, are 2.5 and 2.9 times more likely than other women to develop lupus and Sjogren’s syndrome, respectively.
Why do women’s bodies have these strange mechanisms that increase the risk of disease? Typically over time evolution eliminates processes that make it harder for species to reproduce and thrive, and X-linked autoimmunity definitely hinders thriving. This paradox suggests to evolutionary biologists that the phenomenon might also provide some significant benefit.
In a 2019 paper in Trends in Genetics, Melissa Wilson, a computational and evolutionary biologist at Arizona State University, and her colleagues outlined their pregnancy compensation hypothesis, which is based in part on evolutionary observations. The evolution of the placenta—an organ that provides oxygen and nutrients to fetuses during pregnancy—occurred at the same time mammals evolved sex chromosomes, and it also coincided with the sudden addition of many more genes to the X chromosome. These three developments could all be related.
During pregnancy, women have to tolerate the growth of the fetus, in which half the DNA is foreign because it comes from the father. This outside origin creates cells that the immune system would normally attack. Women also must tolerate the placenta, which is made by the fetus. Perhaps, Wilson says, X-linked genes and incomplete X inactivation evolved as a way for a woman’s body to flexibly respond to the strange new immune requirements of pregnancy. During pregnancy, immunity shifts in dynamic ways: Early in pregnancy, certain healthy immune responses increase, which helps the placenta grow new blood vessels; in the middle of pregnancy, immunity decreases. Then immune responses and inflammation increase again in anticipation of labor.
Other observations align with the predictions of the pregnancy compensation hypothesis. For instance, women today spend a much smaller proportion of their lives pregnant than they did hundreds of years ago, which means that women’s immune systems are not suppressed as often as they used to be. This could help explain why autoimmune diseases are increasing among women today, as well as why they were less of a burden in the past. Although validating this hypothesis requires much more research, Wilson says it is possible that “placentation and pregnancy are critical in shaping maternal immune systems, which in turn could suggest why we have these sex differences in disease.” Put another way, autoimmunity may be an unfortunate by-product of the complex immune response women need to bear children.
Gutting it Out
Not everything in the body is determined by genetics—far from it. One identical twin may develop an autoimmune disease while the other twin, who shares the same genome, does not. The environment is a big piece of the puzzle. It is still unclear which outside exposures might be most important, but research is starting to implicate microbial infections, chemicals such as endocrine disruptors, smoking, diet, stress and the “good” commensal bacteria that live in the intestines.
Some fascinating work in animals points to gut bacteria—collectively referred to as the intestinal microbiome—as a driver of excessive autoimmune disease risk. Jayne Danska, an immunologist and biophysicist at the University of Toronto, has spent much of her career trying to understand the relation between sex and the genetics of autoimmune disease—essentially, whether genes that increase the risk for autoimmunity have varying effects on men versus women. But in 2012 she made a serendipitous discovery that launched her work in a surprising new direction. “It’s one of the adages of science that you find the best things that you weren’t looking for,” she says.
Danska and her team were trying to find risk genes for type 1 diabetes, an autoimmune disease in which the body attacks the insulin-producing beta cells in the pancreas. They were using a lab-bred type of rodent known as nonobese diabetic (NOD) mice. The mice are good models for the human disease, with one striking exception: men and women are equally likely to develop type 1 diabetes—it is one of the few autoimmune diseases that does not predominantly affect women—but in NOD mice, the disease is twice as likely to arise in females.
Danska knew that environmental factors sometimes interact with genes, and she had been looking at gut bacteria as a risk factor. She started to wonder whether, in her mice, gut bacteria differences might be related to the skewed diabetes ratio. To find out, she and her colleagues grew a subset of NOD mice in a germ-free environment, devoid of bacteria and viruses, including the commensal bacteria that normally populate the intestines.
That is when Danska made her first surprising discovery. When she looked at how many of the germ-free animals developed diabetes as adults, “the sex difference went away completely,” she recalls. The males were suddenly just as likely to develop diabetes as the females were. “This was a huge finding that we hadn’t expected. I just couldn’t believe it was true.”
But repeating the experiment showed the same effect. Then more work led to more surprises. The researchers took the bacteria from adult male NOD mice and put them into young female NOD mice that had not yet developed diabetes. The female mice then grew into healthy adults without the disease.
Danska’s findings, which were published in 2013 in Science, provided the first evidence that “the microbes in the gut can influence female-biased autoimmunity,” says Martin Kriegel, a rheumatologist and clinical immunologist at the University of Münster in Germany. It is an important finding, he says, that scientists are still working to understand.
No one knows yet why males’ gut microbes seem protective. One thing Danska and her team have determined, though, is that testosterone is crucial: When they drew blood from germ-free NOD mice, they found that the diabetes-prone males had lower levels of circulating testosterone than microbe-laden males usually do. And when female mice were colonized with microbes from males and were apparently protected from disease, they had higher circulating levels of testosterone than females with microbes usually do.
All of this suggests that there is something about the microbes in males that increases testosterone and is protective. When Danska and her colleagues took gut microbes out of male mice and put them into the guts of female mice, then blocked testosterone signaling in their bodies, they were again at an increased risk for type 1 diabetes. The findings align with lupus research in men showing that suppressed testosterone appears to raise their risk of that disorder. (The research also lines up with work in a strain of mice in which females are especially prone to lupus. Removing gut bacteria from these female rodents lowers their risk, scientists at the Medical University of South Carolina reported in 2020 in the Journal of Immunology.)
It is unclear how microbes might regulate testosterone, or vice versa. Danska’s research suggests that the composition of commensal microbes diverges in male and female mice around puberty, so something seems to happen to the bacteria around that time. This may even explain why there is not much of a sex difference in the prevalence of type 1 diabetes in people; the disease typically develops before puberty, before microbes would have a chance to shape risk based on sex. It could be that the microbes are affected by puberty’s sudden influx of sex hormones, but it is almost certainly “a two-way road,” Kriegel says—the microbes respond to sex hormones, and the sex hormones respond to microbes.
Toward a Better Balance
Of course, mice are not people. But Danska believes that her findings have significant implications for autoimmune diseases that do skew toward women. Perhaps some gut bacteria in women are critical in the development of autoimmunity. If so, tinkering with gut microbes might enable us to thwart disease.
Danska and Kriegel hope it may be possible to develop targeted microbe-based therapies for women at high risk for autoimmune diseases—therapies that can shape the microbiome in protective ways. Other researchers are looking at ways to tweak sex hormone signaling to temper risk. The more scientists learn about why women are vulnerable, the more chances there may be to intervene before diseases develop.
Given that X chromosomes, female sex hormones and female gut bacteria all appear to increase the risk for autoimmunity, it might seem as though biology is somehow conspiring against the female sex. But this autoimmune burden can be seen in another way, too: as a reflection of the importance of women for the survival of our species. “Females have to do all kinds of absolutely remarkable things from an immunological perspective that males just aren’t called on to do,” Danska says. Autoimmunity may be the cost women’s bodies pay for their dynamism—but it is at least a burden that science might eventually be able to eliminate.