In 2015, gastroenterologist Edwin Liu set to work on a clinical and genetic data set that had been growing for more than 20 years. The data pertained to celiac disease, a lifelong condition involving bouts of severe gastrointestinal distress and other symptoms, triggered by ingestion of gluten proteins that are found in wheat and several other grains. In a two-decade collaboration with researchers at Children’s Hospital Colorado in Denver, Liu’s predecessors and colleagues at the University of Colorado kept track of 1,339 babies born in the city who were deemed at risk of developing the disease due to mutations in celiac-linked genes. The researchers carried out yearly tests to see whether or not the children developed the disease, hoping to better define the risk associated with each of the genetic variants.
Not far into his analyses, however, Liu found something in the data that undermined a much larger assumption in the celiac field. “Usually, when we quote numbers for celiac disease, we’re quoting around 1 percent” prevalence in the US population, he says. But using data from this cohort along with estimated frequencies of each genotype across the Denver metro area to extrapolate the incidence of celiac disease to the general population, Liu found that the true prevalence of celiac disease had to be much greater—more than 3 percent by age 15. “It was a surprise,” he says. “These numbers are much higher than anything else quoted in the U.S.”
Researchers reading the paper, which was published online earlier this year in Gastroenterology,1 were similarly taken aback. “If you look at the rates, it’s frightening,” says Joseph Murray, a celiac researcher at the Mayo Clinic in Rochester, Minnesota. Of course, the statistic could be specific to the Denver cohort, he notes, but it does fit in with similar trends reported both in the U.S. and around the world.
Celiac symptoms, which include abdominal pain and distension, diarrhea and flatulence, nausea, and fatigue, are brought on by ingestion of gluten—a protein complex present in wheat, barley, and rye. Unlike food allergies, which are often primarily mediated by an overreaction of adaptive immune responses such as immunoglobulin E antibody production and mast cell activation, celiac disease engages both innate and adaptive immune pathways, and produces antibodies that target not only gluten, but the body’s own proteins. As a result, the disease is generally considered an autoimmune condition. (See illustration.) Triggered by even tiny amounts of gluten, these immunological attacks lead to T cell–mediated atrophy of the gut wall, which can be characterized via a biopsy of the small intestine for celiac diagnosis (see “Diagnosing Celiac Disease”).
As the use of biopsy and other diagnostic methods have improved in recent decades, celiac disease has become easier to detect. So when the first reports of increasing numbers of celiac cases in the U.S. came out in the early 2000s, many researchers attributed the uptick to progress in disease recognition. But closer scrutiny of the data suggested there was more going on. “We weren’t just better at finding celiac disease,” Murray says. “There was a lot more of it to go around.”
By comparing blood samples taken from young adults in the Air Force around 1950 with matched samples from residents of a Minnesota county collected since 1995, for example, Murray’s group estimated an increase in prevalence from 0.2 percent to nearly 1 percent.2 Sweden, meanwhile, experienced what is now referred to as a celiac epidemic in the late 1980s and early 1990s, with one study estimating that as many as 3 percent of children born at the height of the epidemic had developed celiac disease by the age of 12—though rates dropped back down to just over 2 percent for children born in 1997.3 And several studies based on blood tests suggest increasing numbers of people are developing celiac disease in wheat-eating areas of northern India, with a prevalence in children of around 1 percent and some researchers warning of an impending epidemic there too.
The cause of this apparently global trend remains a mystery, not least because, while the immunopathology of celiac disease has been studied for decades, just what causes people to develop the ailment in the first place remains unclear. Almost all diagnosed patients have mutations in at least one of the two genes coding for HLA-DQ, a membrane receptor on antigen-presenting cells that helps the immune system distinguish self from non-self and coordinate T-cell activity. But not everyone who has such risk genes gets celiac. “Around 40 percent of people have the genes predisposing them to celiac disease,” Liu explains. “The big question is why some people get it and others don’t.” Hypotheses abound, with many pointing the finger at a gluten-rich diet, but evidence to support these ideas remains far from conclusive.
Getting to the bottom of this question will be necessary not only to curb the concerning trend, but also to help doctors better detect and manage the multifarious disease, for which the only current treatment is a gluten-free diet. In addition to celiac’s sometimes-debilitating symptoms, the disease is associated with a heightened risk for numerous conditions, including autoimmune diseases such as diabetes and hypothyroidism, and myriad other disorders, from infertility to small-bowel cancer. Overall, celiac patients have up to a twofold increased mortality risk compared with the general population.
“The stakes are high,” says Murray. “If this disease has gone from being a truly rare disease in some geographies to being a common disease affecting 2 or 3 percent of children, that’s no longer a small disorder.”
Why the rise?
One thing celiac researchers agree on is that the direct cause of the rise in the disease likely resides outside of our DNA. “Over decades, it’s just too quick for genetic changes to occur,” Liu says. “We have to assume that this is based on environmental factors.” There’s still little in the way of concrete answers as to what these factors might be, however. “I’ve heard every type of hypothesis that’s been thrown out there,” says Murray, “but most are them are not easily testable.”
Some of the more unusual candidates blamed for triggering celiac disease include microwaves, plasticware, and diatomaceous earth—an abrasive powder applied to flour containers as an insecticide—although scientific evidence to incriminate these supposed culprits is scant. Other factors that have more reliably been tied to increased celiac risk in genetically predisposed infants include delivery by Caesarian section, and intestinal infections by pathogens such as reovirus (recently implicated in a mouse model)4—although their impacts are likely minor, Murray says.
The role of gluten itself—the immediate trigger for the immune responses in celiac patients and therefore, researchers have long assumed, a crucial player in the epidemiology of the disease—has also remained frustratingly elusive. During the 2000s, for example, several observational studies pointed to a suite of dietary factors, including age of gluten introduction, as influencing the development of celiac disease. But the findings suffered a blow in 2014 when two randomized clinical trials failed to find any effect of the timing of gluten introduction.5,6 The studies also found no evidence for a link with the duration of breast-feeding—a factor that had previously been touted as protective against developing celiac disease. Gastroenterologist Alessio Fasano of Massachusetts General Hospital, a coauthor on one of the publications, says researchers realized then that “the story is much more complex than we thought.”
A more recent hypothesis is that the amount of gluten consumed, if not the timing, could play a role in triggering celiac disease in children. The US Food and Drug Administration (FDA) notes that wheat consumption increased rapidly in the second part of the 20th century as people began to eat less meat and consume increasing amounts of readily available, wheat-containing fast foods. (A more controversial idea is that the composition of wheat has changed significantly during this time—search “Frankenwheat.”) And rising incidence of celiac disease in South Asia tracks with the widening adoption of Westernized diets, although data on gluten consumption per se is lacking.
Some evidence that these dietary changes could be tied to the rise in celiac disease comes from a retrospective 2016 study of Swedish infants, which suggested that genetically susceptible children consuming more than 5 grams of gluten per day—the equivalent of about one slice of whole wheat bread—before 2 years of age were up to two times more likely to develop celiac disease than those consuming less than that amount.7 “[The result] tells me that the amount of gluten matters,” says Murray. “I think we have to go back and revisit what’s happening with gluten—how much are we eating, and is it a potential risk factor?”
Reactions to these findings have been mixed, however. “The evidence was fairly weak,” notes celiac researcher Detlef Schuppan of Johannes Gutenberg University Mainz in Germany. In terms of the global rise in celiac prevalence, “the amount of gluten ingested does not explain it,” he adds.
Amidst uncertainty about gluten’s part in the celiac trend, many researchers are quick to point out that it’s not just our diet that has changed in the last century. One factor now under scrutiny across digestive diseases and beyond is humans’ usage of antibiotics and, consequently, the composition of bacteria making up the gut microbiome. Bacteria living in the gut play important roles in metabolism and in the regulation of immune responses to food, so for many researchers, these microbes are likely suspects in celiac disease pathogenesis. According to this line of thinking, “maybe the bugs we’ve now got are not as happy when they interact with gluten,” Murray says. “Or the results are not as good for us when these bugs interact with gluten as when our old bugs did.”
See “The Sum of Our Parts”
The last two decades have seen a number of observational studies report abnormal microbiome composition in the guts of celiac patients compared with healthy controls. Patients with celiac disease show a higher proportion of gram-negative bacteria such as Bacteroides and E. coli, for example, and some evidence suggests that those displaying gastrointestinal symptoms also have higher levels of Proteobacteria. Another bacterium, Helicobacter pylori, has been associated with protection from celiac disease, and declines in the number of adults carrying this microbe in their guts appear to have coincided with increases in the number of celiac cases in the U.S.—although research on this subject remains inconclusive.
Whether these differences in microbiome composition are the cause or the consequence of celiac disease remains unanswered, says Bana Jabri, director of research at the University of Chicago Celiac Disease Center. “You could think about it in several ways,” she explains. “Maybe there’s a difference in the microbiota from the beginning, and this has a causative role. Or maybe what you’re seeing is just a secondary effect. At this point, we really don’t know.”
Nevertheless, circumstantial evidence is accumulating to suggest more than a passive role for these microbes. For starters, it’s known that changes in the microbiota can induce different types of immune responses, and celiac patients often continue to show abnormal gut flora even after adopting a gluten-free diet. Additionally, Jabri’s group showed last year that mice engineered to overexpress interleukin-15—a cytokine involved in celiac disease pathogenesis—had restructured microbiota as well as altered production of certain fatty acids, mirroring precursors of intestinal inflammatory diseases in humans.8 “When you put all this together, you could say that there really is enough evidence to believe in a causative role for the microbiota,” says Jabri. “But the critical experiments still need to be done.”
With so many factors being investigated, Liu says, it’s unlikely the explanation for an increase in celiac disease incidence will be simple. “I don’t think we’re going to be able to find a single environmental trigger,” he says. “It’s going to be a combination.” Murray takes a similar view. “There are so many things going on, so many moving parts,” he says. “The challenge for us as scientists is to reduce it down to testable hypotheses.”……