Ever heard of azodicarbonamide? You probably have, but not by that name. It made headlines as “the yoga mat chemical.” It’s an additive used in bread to whiten flour and aid in gluten development. The concern isn’t azodicarbonamide (often nicknamed AZO) itself, but one of its byproducts: ethyl carbamate, also known as urethane.

It certainly sounds unappetizing. Worse, according to the Environmental Protection Agency, “No information is available” on its chronic effects. That seems like a very good reason to get it out of your bread, but a fuller picture helps temper any fear. There’s urethane in all bread, and you can easily get an entire loaf’s worth of the chemical in one glass of wine. Although AZO can increase bread’s urethane by two-thirds, you can increase it by three to eight times by toasting the bread.

The AZO brouhaha got traction, in the absence of evidence of harm, when Vani Hari, who blogs as Food Babe, pointed out that the chemical is used in yoga mats. The image is vivid and the reaction was visceral. Subway was pressured into removing AZO from its bread, but it’s hard to see that as a triumph for public health. If urethane is risky, shouldn’t we be going after wine? Or toasters?

Michael Siegrist studies risk perception and consumer behavior at the Swiss Federal Institute of Technology in Zurich, and he explains why people fear food additives more than some other potential hazards: “Exposure to food additives [is] not voluntary, they cannot be controlled, and consumers may not be aware that they are exposed to food additives. [These] risk characteristics of food additives may result in a higher risk perception than justified based on the available evidence.”

Which doesn’t mean additives are risk-free. Although it’s unlikely that AZO is a meaningful threat, a 2013 report by the Pew Charitable Trusts documents more than 10,000 additives that are allowed in our food and the varying levels of safety information about them. Moreover, we have almost no information about the aggregate risk, as the Food and Drug Administration evaluates each individually.

The evidence we do have points generally, although not exclusively, to safety. Michael Jacobson, head of the Center for Science in the Public Interest, says that “most food additives are safe; some should be better tested; and several pose risk.” Salt, sugar and hydrogenated oils top his risky list, and he also points to the connection between food dyes and children’s behavior, the issue that drove Kraft to eliminate yellow dye from its iconic macaroni and cheese. Fergus Clydesdale, director of the Food Science Policy Alliance at the University of Massachusetts at Amherst, says there’s no hard evidence of harm from additives “in modern regulatory times,” with the exception of the dye issue that Jacobson raises. Even there, Clydesdale calls the evidence for the link “questionable,” and Jacobson acknowledges the difficulties of studying behavioral changes objectively but says he is convinced by a growing number of studies showing a link.

Another problem with food additives involves something called generally recognized as safe (GRAS) status. It’s a designation for ingredients that have been well established, either by research or by a long history of consumption, to pose no threat. The problem is that companies are allowed to determine GRAS status without involving, or even alerting, the FDA. Between 1997 and 2012, about a third of the GRAS determinations that were filed with the FDA (the filing is voluntary) were made by an employee of or consultant to the additive manufacturer. (Many others were made by expert panels, like that of the Flavor and Extract Manufacturers Association, with protocols for avoiding conflict of interest.)

There’s sense in the idea that, to lighten the load of an overburdened regulatory agency, we should start by relaxing rules on substances least likely to be dangerous. But freeing manufacturers to declare their own ingredients safe raises serious concerns about who’s watching the henhouse. I think the Natural Resources Defense Council, which recently released a report on GRAS substances, is right in pointing out that an undisclosed additive can’t be “generally recognized” as anything.

These problems are real and need to be addressed, but “yoga mat” headlines make the threat look much larger than it is. The fuss-to-risk ratio for food additives is off.

In that way, food additives are a bit like genetically modified organisms, or GMOs. Both fit Siegrist’s description of what’s most likely to scare us, and concerns about them often seem like proxies for big, important, complex issues that are difficult to tackle — a reason not to dismiss those concerns, even if they overstate risk. It’s hard to find a toehold on the slippery problem of corporate power in government or agricultural dependence on monocrops. Even though the scientific consensus is that there’s no evidence GMOs pose a health threat, the issue seems to offer such a toehold.

The chemicals in our food, although they probably pose only small risks, are likewise a way in to a more complex issue: the domination of processed foods in our diet. Jacobson notes that even safe additives “make possible the production of the thousands of nutritionally worthless or harmful foods that populate supermarket shelves.” Those processed foods pose a much-better-established risk than food additives do.

The difficulty of tackling processed foods starts with the most basic question: What are they? The lack of an agreed-upon definition can make straightforward advice like “avoid processed foods” troublesome. Connie Weaver, professor of nutrition science at Purdue University, told me that people should eat “the most nutrient-dense foods, regardless of degree of processing.” When I asked for examples of nutrient-dense foods that are highly processed, she listed wine, cheese and bread. Although I plan to make a meal of Weaver’s advice as soon as the sun passes the yardarm, I don’t think those are the foods people think of when they hear “processed.”

Justice Potter Stewart famously said that we know pornography when we see it. We know processed food, too, and it doesn’t look like canned tomatoes. It looks like instant ramen, or hot dogs, or Kraft Macaroni & Cheese, with or without yellow dye.

There’s compelling evidence that highly processed foods pose a health risk, but not necessarily because of the ingredients you can’t pronounce. What’s not in them might matter more. Marion Nestle of New York University’s department of nutrition, food studies and public health told me in an e-mail that “relatively unprocessed foods contain nutrients and other components (antioxidants, for example) that might have benefits for health, singly and in concert. People who eat relatively unprocessed foods, especially vegetables, tend to be healthier than people who don’t.”

Carlos Monteiro at the University of Sao Paulo in Brazil has published a classification system for processed foods and has found a correlation between the intake of “ultra-processed” food and obesity. He’s now working with U.S. data.

Meanwhile, in just the past few weeks, researchers in England reported that people who ate seven or more servings of fruits and vegetables daily had a 42 percent lower risk of dying — of any cause — during the 71 / 2-year study period. A study of many different diets — low carb, low fat, Mediterranean, Paleo, vegan — done by David Katz, at Yale’s School of Medicine, concludes that “a diet of minimally processed foods close to nature, predominantly plants, is decisively associated with health promotion and disease prevention.”

Every conversation we’re having about the risk of AZO is a conversation we’re not having about the risk of processed foods. It also shifts the discussion to ground where food manufacturers have science on their side and gives them an opportunity to look reasonable, safety-minded, even magnanimous. Kraft gets kudos for eliminating yellow dye, but it’s the mac and cheese that remains the problem.

Haspel, a freelance writer, farms oysters on Cape Cod and writes about food and science. On Twitter: @TamarHaspel. She’ll join today’s Free Range chat at noon: live.washingtonpost.com.