If you walk down the aisles of any supermarket, you’ll see what dietary emulsifiers can accomplish. This common class of food additives binds water and oils together, preventing mixtures of the two from splitting. They stabilise ice-cream and other frozen desserts, mayonnaise, salad dressings, and virtually every kind of processed food. “Anything that sits in a package on a supermarket shelf, and can stay there for a while, probably has emulsifiers in it,” says Andrew Gewirtz from Georgia State University.
These additives may confer stability to food, but they can also bring discord to the gut—at least in mice. Gewirtz has found that two common emulsifiers—caboxymethylcellulose (CMC) and polysorbate-80 (P80)—can change the roll call of bacteria in a mouse’s gut. They also make the gut more porous, allowing microbes to slip through its walls and reach the immune cells and blood vessels on the other side. As a result, the mice developed severe inflammation. They also put on weight, and their blood sugar went up.
The team only looked at laboratory mice, so it’s not clear if emulsifiers have the same effects in humans at the doses we normally eat (more on this later). Still, “this work cannot be ignored,” says Fergus Shanahan from University College Cork, who was not involved in the study. He doubts that most people would be significantly affected by occasionally eating foods with emulsifiers. But the calculus of risk might change for those who have a genetic predisposition to inflammatory bowel disease (IBD), or who eat lots of processed foods.
“The other implication is that current methods for testing food additives for safety are not adequate,” says Gewirtz. CMC and P80 are both “generally regarded as safe”, since they’re not toxic at the levels found in food, and they don’t cause cancer. But such tests say nothing about their ability to disturb the relationship between us and the microbes we carry—disturbances that have been linked to obesity, IBD, and other conditions.
Your immune system needs to spot and thwart infectious microbes, while maintaining a truce with trillions that live in your body and carry out important tasks. If it’s too twitchy, it will constantly go berserk whenever it notices our microbial companions and trigger chronic inflammation. If it’s too relaxed, it wouldn’t detect dangerous threats. It must react without overreacting.
It achieves this balance through a bewilderingly complex network of cells and molecules that I’m going to boil down into a single image. Think of the thermostat that stabilises the temperature of your room. Now picture an “immunostat” that, in a similar way, dictates how responsive the immune system is. Set it too low and you become vulnerable to infections; too high, and your run the risk of inflammatory diseases.
Many things affect where the immunostat is set, including genes, diet, infections, and more. Your gut microbes are also involved. Some groups provoke the parts of the immune system that exacerbate inflammation. Others stimulate the pacifying components that calm everything down. Physical barriers are also important. The simplest way of stopping the immune system from overreacting to microbes is to keep them separate. Our gut achieves this by keeping its cells tightly fused and covering them with a thick layer of mucus. Microbes sit on one side of this Great Wall of Mucus; immune cells on the other.
Gewirtz suspects that emulsifiers disrupt both the mucus and the microbe communities, pushing the immunostat towards a twitchier setting.
His postdoc Benoit Chassaing fed lab mice with either CMC or P80, by adding both substances to their food or water at one part per hundred. When he looked at their guts under a microscope, he saw that their mucus wall was thinner than usual, and bacteria had penetrated deep into what was once a No Microbe’s Land. Some were actually touching the gut itself. The gut had also become leakier, so many microbes found their way through to the immune cells and blood vessels on the other side.
The emulsifiers also changed the communities of microbes within the rodents’ guts. Chassaing saw a rise in species that excel at triggering inflammation, and in those that eat mucus like Ruminococcus and Akkermansia. Other microbes shrank away, including groups that produce anti-inflammatory substances by digesting dietary fibre.
These changes lead to a vicious cycle of even more inflammation, even leakier guts, and even thinner mucus. The result: low-grade inflammation in normal lab mice, and a more severe form—colitis—in mutant rodents that were genetically susceptible to IBD.
After swallowing the emulsifiers, both breeds of rodents ate more food. They put on body fat and gained 10 grams in weight (on top of their normal 140). Their blood sugar levels went up. They became less sensitive to the hormone insulin. In other words, they showed many symptoms of metabolic syndrome—a condition that increase the risk of diabetes and heart disease.
Do the microbes cause these problems, or are they just along for the ride? It’s probably the former. None of these changes—the thinner mucus, the inflammation, or the metabolic problems—happened to germ-free mice that were raised in sterile conditions. Without their microbes, those rodents ate emulsifiers to no effect. But when Chassaing loaded them with microbes from individuals that had eaten emulsifiers, they too developed all the same symptoms. Whatever the additives are doing, they’re doing it via gut microbes.
Inflammatory bowel disease was once very rare, but has become more common since World War II. Many things that change our relationship with our microbes could have contributed to that rise, including antibiotics, sanitation, and dietary shifts, including an abundance of fat, a lack of fibre, or the presence of artificial sweeteners.
What about emulsifiers? It’s hard to say. To state the obvious, mice aren’t people. “Many observations in mice tell us a lot about host-microbe interactions but either don’t translate to humans or have far less significance in humans,” adds Shanahan. “The microbiota of a lab mouse is very simple and much simpler than that of humans. It doesn’t take much to significantly disturb it.” Many things can, including drugs like aspirin, antibiotics, other bacteria, and more. To Shanahan, it’s not surprising that dietary emulsifiers join the list. “What is surprising is that this would occur at such low levels, including levels that humans may be exposed to,” he adds.
But that’s another limitation: it’s hard to compare the doses that Chassaing used to the levels of emulsifiers we eat, because no good measurements of those levels exist. According to one report from the Food Safety Commission of Japan said, “In Western countries, the daily intake of polysorbates, based on their usage in food, was estimated at 12-111 milligrams per person per day.” That’s proportionally much less than what Chassaing’s rodents ate, but we have no idea if the Japanese estimates are reliable—the report provides no data or sources for its figures.
In the absence of such data, the team used the limits set by the US Food and Drug Adminstration, which approves the use of P80 at up to 1 percent in foods, and CMC at up to 2 percent. Chassaing used 1 percent levels in his experiments, but he also found signs of inflammation at 0.1 percent. “We gave amounts that approximate the total consumption of a person who eats a lot of processed food,” says Gewirtz. “It’s the best we could do at this time, but we need better estimates.”
“I went over the data, and they did a thorough job,”says Eugene Chang from the University of Chicago Medical Centre, who studied IBD. “There’s also precedent for this.” He points to other studies showing that carrageenan—another common emulsifier, derived from seaweed—can cause inflammatory bowel disease in mice.
Then again, there’s also some conflicting evidence from other animal studies—none of them have looked at microbes but a few have measured body weight. A Dutch team showed that CMC doesn’t affect the body weight of broiler chickens, and the US National Toxicology Program found that P80 doesn’t change the body weight of rats. Meanwhile, a Japanese study found that pregnant rats actually lost weight when given P80.
The FDA certainly isn’t changing its position. In a statement, it said, “The FDA closely monitors the scientific literature for information that might indicate a potential public health concern with a food substance. At this time, the FDA does not have sufficient evidence to alter its previous conclusion that polysorbate 80 and carboxymethyl cellulose are considered safe under their intended conditions of use in food.”
Meanwhile, Gewirtz says, “We’re certainly eating less processed food since we’ve been doing this work. It took a lot of effort, but we did find one type of ice-cream in the supermarket that’s emulsifier-free.”
Reference: Chassaing, Koren, Goodrich, Poole, Srinivasan, Ley & Gewirtz. 2015. Dietary emulsifiers impact themouse gutmicrobiota promoting colitis and metabolic syndrome. Nature http://dx.doi.org/10.1038/nature14232