Ingredients created by food companies flavor what Americans eat each day — everything from juice drinks and potato chips to ice cream and canned soups. They give Cheetos their addictive cheesy taste and help distinguish Jolly Ranchers from other fruit-flavored candies.
But the organization responsible for the safety of most “natural” and “artificial” flavors that end up in foods and beverages isn’t part of the U.S. government. Rather, the Flavor and Extract Manufacturers Association — a secretive food industry trade group that has no in-house employees, no office of its own and a minuscule budget — serves as the de-facto regulator of the nation’s flavor additives.
The trade association, which operates with the U.S. Food and Drug Administration’s blessing, says that it makes research on the safety of various flavors available for public inspection.
“Oh, garbage,” said Susan Schiffman, an adjunct professor at North Carolina State University who studies sweeteners. “It’s not transparent.”
In late 2012, Schiffman and a colleague from the National Institutes of Health were months away from publishing a paper examining the biological effects of the popular artificial sweetener sucralose.
But first they needed to contact the Flavor and Extract Manufacturers Association.
The trade group, Schiffman learned, had recently approved the safety of a chemical compound that amplifies sweetness. She says she contacted the Flavor and Extract Manufacturers Association in December 2012 to learn the chemical’s name and identification number so that she could research its safety.
“They said, ‘We’ll call you back.’ No call back. So I called again. No call back,” Schiffman remembered. “So then I wrote a letter. Nothing happened.”
While her persistent requests went unanswered, Schiffman says she eventually received the information she needed from a flavor expert unaffiliated with the trade association. But Schiffman still had a problem: She couldn’t find safety data on the chemical anywhere.
When a scientist from the trade association finally called her back, Schiffman says she asked him for data supporting the safety of the sweetness-enhancing compound. But she says he refused to give it to her.
“‘There’s something in the food supply, and I can’t find out the toxicity of it?’” Schiffman recalled asking the flavor group scientist who declined her request. “They would not give me the safety data… It was absolutely astounding.”
The Flavor and Extract Manufacturers Association disputes Schiffman’s version of events.
John Cox, the trade group’s executive director, wrote in an emailed statement to the Center for Public Integrity that Schiffman requested only information on the safety status of the flavor, which was “promptly provided” to her. But he said she never asked for the data supporting the determination.
“If Dr. Schiffman had requested the safety data,” Cox wrote, “we would have provided it to her.”
Public interest groups, however, share Schiffman’s frustration with the trade association. They, too, report getting stonewalled by the flavor group when requesting information about the industry’s safety determinations.
A ‘black box’
The largely industry-run system for evaluating flavors is “fundamentally problematic because it’s so opaque,” said Erik Olson, the Natural Resources Defense Council’s senior strategic director for health and food. “It’s a black box.”
Yet the system factors into food products found in nearly every consumer’s grocery cart.
In recent years, activist groups and social media campaigns have been demanding that food companies become more accountable to consumers and transparent about what they are adding to their products.
But most Americans know as little about the decidedly low-profile Flavor and Extract Manufacturers Association and its safety assessments as they do about the more than 2,700 flavoring chemicals it has declared safe during the past five decades.
Moreover, public interest groups say the FDA’s recent response to a Freedom of Information Act request suggests that even the government may be blind to the science behind many of those flavors.
Much is at stake: The flavor industry’s system of self-policing helps it avoid government oversight, potentially saving companies significant amounts of time and money. In Europe, by contrast, companies must have their flavors and other ingredients reviewed for safety by an independent agency funded by the European Union.
The flavor industry makes its safety evaluations “behind closed doors” and then asks consumers to trust them, said Caroline Cox, research director for the Center for Environmental Health. “We just have enough experience with all kinds of toxic chemicals to know not to want to trust an evaluation if someone says, ‘Trust us, it’s all OK.’”
“There is some real need for reform here,” Rep. Chellie Pingree, D-Maine, who serves on a subcommittee that oversees the FDA, said in an emailed statement. “We can do better than just letting the flavor industry decide for themselves which chemicals they can put in food without any oversight.”
Industry officials say flavors used in food sold in the United States are safe, pointing out that they pose very little health risk because they are used in such small doses. But identifying health concerns would seem to be difficult with so little scientific information publicly available behind some of the trade group’s safety decisions.
A Center for Public Integrity review of documents provided by the Flavor and Extract Manufacturers Association found that four of the group’s most recent safety assessments depended largely on studies that were not published in scientific literature. Public interest groups say that is problematic because it doesn’t allow the scientific community to vouch for the industry’s safety decisions.
The Center for Public Integrity contacted two dozen flavor companies to discuss their ingredients and the process by which they evaluate their safety. Most of them either declined to comment or ignored interview requests.
Two company officials who responded in writing both stressed that flavor companies are very guarded about discussing such information.
“Given the competitive landscape in our industry, many companies, including ours, regard our flavor formulations as valuable proprietary business information,” Donald Wilkes, president and chief executive officer of California-based Blue Pacific Flavors, emailed in response to questions. “We typically do not share this information unless required to do so for legal reasons.”
Nevertheless, the public should be assured food flavor safety standards are high, said Kevin Renskers, president-elect of the Flavor and Extract Manufacturers Association and vice president for corporate safety and regulatory affairs at Takasago International Corp.
“FEMA operates the premier global program to assure the safety of flavor materials,” he responded in an email.
Lucrative business
The companies that make up the flavor industry — including international manufacturers such as Givaudan, Firmenich and Sensient — are not household names. But they make their money by selling flavors to big food companies such as Kellogg, Kraft and Nestlé.
Last year, Switzerland-based Givaudan reported 4.4 billion Swiss francs (roughly $4.8 billion) in sales of flavor ingredients. The company leads the industry with about 25 percent of the global market share in flavors and fragrances.
“The modern processed food industry could not flourish without the flavor industry,” said Kantha Shelke, a food scientist and spokeswoman for the Institute of Food Technologists, a society of food science professionals.
Today, Shelke said, the flavor industry is “big, it’s complicated and it’s sophisticated” — to the point where companies can create a product that tastes like guacamole without even using avocado as an ingredient. The goal, one industry scientist told CBS’ 60 Minutes in 2011, is to develop addictive flavors that consumers “want to go back for again and again.”
Shelke says flavor companies don’t like to discuss their flavor compounds, partly because they are worried about scaring “chemophobic” consumers who might be frightened by long, unfriendly-sounding chemical names.
Consumers’ growing unease with chemicals in food has recently led companies to remove some controversial ingredients from their products. Fast food chain Subway, for example, announced last year that it was removing azodicarbonamide from its bread after a popular food blogger revealed that the chemical, used as a dough conditioner, was also used in making yoga mats.
Within the last month, Subway, Taco Bell and Pizza Hut each announced plans to remove artificial colors and flavorings from their products.
By law, flavor ingredients only have to be listed on food labels as “natural” or “artificial” — and rarely do labels volunteer additional information. Natural flavors are derived from plants and animals, while artificial flavors are synthetic chemicals. Both are produced in labs by scientists.
When it comes to questions about safety, flavor companies typically defer to the Flavor and Extract Manufacturers Association. The group is managed by Verto Solutions — a Washington, D.C., firm that provides scientific consulting, communications and government relations services to the flavor group, as well as other similar trade associations, including the International Association of Color Manufacturers, the International Organization of the Flavor Industry and Pickle Packers International.
While the Flavor and Extract Manufacturers Association reports no lobbying, it understands Congress’ power to regulate the food industry. The group is hosting its second annual congressional “fly-in” event on June 16-17 to “increase its visibility” and meet with representatives “to ensure that they are aware of the U.S. flavor manufacturing industry and its importance to the economy,” according to its website.
‘The highest standards for transparency’
The Flavor and Extract Manufacturers Association’s program for evaluating the safety of flavors has won praise from some corners of government, namely the FDA and the Government Accountability Office.
Its flavor safety program launched in 1960 — two years after Congress passed the first law regulating ingredients added to food.
The law allows food companies to bypass a lengthy government-led safety review if they can establish that their ingredients are “generally recognized as safe,” or GRAS, for their intended use.
In other words, companies using the so-called GRAS process must demonstrate that there is a consensus among scientific experts that their ingredients are safe.
Companies have the option of involving the FDA in the process, but most ingredient manufacturers choose to make safety determinations without government oversight.
In Europe, the industry doesn’t have the option to police itself.
The European Union requires companies to have their new flavors and other additives reviewed by the European Food Safety Authority, an independent agency funded by the European Union.
The Flavor and Extract Manufacturers Association’s member companies — which produce 95 percent of all flavors on the market in the United States — typically forgo FDA review and instead choose to submit their flavors to the trade association for a safety review.
A standing panel of six to eight scientific experts oversees the trade group’s safety program and determines whether ingredients are generally understood to be safe by the scientific community. These experts, who are paid by the trade association, review published and unpublished data before making a conclusion on the safety of an ingredient’s use.
“From Takasago’s perspective, to have GRAS approval by the FEMA Expert Panel provides us with a high level of confidence that our material can be used safely,” said Renskers, the company official and Flavor and Extract Manufacturers Association president-elect.
Industry officials stress that flavor ingredients generally pose very little safety risk because they are used in such small amounts in food. The trade group used this argument to support its conclusion that methyl eugenol is safe, despite studies finding that it causes cancer in animals.
The naturally occurring flavor, which the FDA has approved as a food additive, is used in foods including jellies, baked goods and chewing gum. The Flavor and Extract Manufacturers Association first determined methyl eugenol to be GRAS in 1965, but the group revisited that decision in 2001 after studies by the National Toxicology Program found “clear evidence of carcinogenic activity” of the substance in male and female rats and mice.
But the trade group still concluded that methyl eugenol “does not pose a significant cancer risk” to humans because it is used in food at such low levels.
After the trade group reconfirmed that it was safe, the International Agency for Research on Cancer found that methyl eugenol is “possibly carcinogenic to humans.”
The Flavor and Extract Manufacturers Association notes that many of the flavors it has reviewed over the years have also been evaluated for safety by other scientific bodies, including the European Food Safety Authority.
The Government Accountability Office praised the Flavor and Extract Manufacturers Association’s program in a 2010 report about the FDA’s limited oversight of food additives, specifically highlighting its procedures to prevent conflicts of interest and its practice of voluntarily informing the FDA about each of its safety determinations.
Scientists who serve on the flavor group’s expert panel are forbidden from having financial ties to companies submitting flavors for review. To that end, panelists’ stipends come from the trade association, rather than from companies directly. Panelists also review the safety of ingredients without knowing which company submitted them.
For these reasons and more, the trade group bills its GRAS program as “fully objective” and “not subject to bias.”
The Government Accountability Office report further applauded the Flavor and Extract Manufacturers Association for informing the FDA of flavor names, properties and reasons for the expert panel’s conclusion when notifying the agency of new safety determinations.
But if the FDA has such information, it can’t find it, according to the agency’s response to a recent Freedom of Information Act request.
Last October, the Center for Science in the Public Interest requested “all information that the Flavor and Extract Manufacturers Association (FEMA) has supplied to FDA since 1960 as support for its determinations that flavors it has evaluated are generally recognized as safe, or GRAS.”
The FDA’s May 21 response: “We have searched our files and find no responsive information.”
Laura MacCleery, the attorney for the public interest group who filed the records request, said she was “shocked” that the FDA’s response appears to contradict the flavor trade group’s public statements about sharing information with the FDA.
“We were worried about how we would handle the volume of documents,” MacCleery said. “I guess that’s not a concern.”
In an email, FDA spokeswoman Lauren Sucher said that the trade group “provides information about its GRAS lists to the FDA,” but she did not explain why this information was not produced in response to the public records request.
In addition, she said that a separate organization, the Joint FAO/WHO Expert Committee on Food Additives, which is affiliated with the World Health Organization, “has evaluated nearly 2,000 FEMA flavoring agents.” Those evaluations are based on information provided to them by the trade group, according to the FDA. Since the mid-1990s, FDA staff has participated in the evaluations conducted by the international organization, Sucher added.
The GAO said that it based its 2010 report on interviews with the trade group and the FDA. Steve Morris, GAO’s director for food safety and agriculture, said in an email that he was not sure why the FDA could not produce the documents requested by the Center for Science in the Public Interest.
Overall, the FDA praises the trade group’s safety program.
“In the FDA’s experience,” Sucher wrote in an email, “FEMA makes scientifically rigorous, credible determinations that have stood up to the scrutiny of the scientific community of qualified experts.”
For its part, the Flavor and Extract Manufacturers Association says it has long provided the FDA “with the basis for [each] FEMA GRAS determination … including all of the safety data.” Providing the Center for Science in the Public Interest “with all of the information that FEMA has provided to FDA since 1960 would likely fill five hundred boxes,” the trade group’s Cox said in a written response.
The trade association periodically announces its determinations in a food industry trade magazine, but those announcements do not include the data supporting the safety assessments.
However, the trade group, which boasts that its GRAS program “meets the highest standards for transparency,” says it will share safety data with any member of the public for only a copying fee.
Safety process questioned
Schiffman isn’t the only scientist who has recently cried foul on the Flavor and Extract Manufacturers Association.
Last September, the Natural Resources Defense Council, the Center for Science in the Public Interest and other public interest groups argued in a letter to the FDA that the trade group’s flavor safety program “does not comply with FDA policies.”
Their chief complaint: that the flavor trade association bases its safety decisions for new flavor chemicals on unpublished safety data.
To establish ingredients as “generally recognized as safe,” determinations typically rely on published studies to show that qualified scientists generally agree that the ingredients won’t harm consumers. This process allows companies to avoid subjecting their ingredients to an extensive FDA-led safety review.
The consumer groups’ letter points to comments Michael Taylor, the FDA’s deputy commissioner for food, made to The Washington Post last August: “The [GRAS] assessments need to be based on publicly available information where there is agreement among scientists,” the Post quoted Taylor as saying. “It has got to be more than three employees in a room looking at information that is only available to them.”
But the Flavor and Extract Manufacturers Association, the public interest groups charge, doesn’t always follow such guidance.
In their letter to the FDA, they said they could find “no relevant published safety data” establishing the safety of a handful of flavors declared safe by the trade group last July.
“We asked FEMA for the published data on the four substances and it said there was none,” the letter stated, adding that the trade group offered to make 7,000 pages of unpublished data available for a $1,000 processing fee. “This practice is contrary to FDA guidance … and a common sense definition of general recognition.”
The Center for Public Integrity sent the FDA a detailed list of questions about the trade group’s flavor safety program. The list included inquiries about the advocacy groups’ complaints, the agency’s interpretation of regulations regarding the use of published data and the extent to which the agency reviews each of the trade association’s safety determinations.
After nearly two weeks, FDA spokeswoman Sucher emailed that “there are so many questions that we do not currently have the resources available to provide individual responses.”
Olson, of the Natural Resources Defense Council, said FDA officials met with representatives of the public interest groups late last year to discuss their concerns about the trade association’s flavor safety program.
While agency officials listened to their complaints, he said they made no commitment to requiring the trade group to change how it runs its program.
Officials from the Flavor and Extract Manufacturers Association’s management firm would only answer questions in writing. The expert panelists paid by the trade group to assess the safety of flavors either declined to comment or did not respond to interview requests.
The trade association’s expert panel “reviews all available information relevant to its GRAS assessment to assure the safety of the candidate flavoring substance under its conditions of intended use whether the data are published or unpublished,” Cox wrote in response to questions.
He cited a section of the regulations governing GRAS assessments that notes that they “shall ordinarily” be based upon published studies.
In his response, however, Cox didn’t clearly answer a question about whether the trade group used only unpublished data to conclude that those four approved ingredients were generally recognized as safe. The Center for Public Integrity asked the trade group in an email to clarify how exactly those four safety determinations were made. Two days later, late on a Friday afternoon, two boxes of documents arrived at its newsroom.
Inside: 9,000 pages — 75 pounds — of published and unpublished safety data supporting four of the most recently approved flavors, including what the trade association described as a fruity flavor, a savory flavor and a minty, cooling flavor.
The trade association declined a request to review the documents with reporters.
A ‘daisy chain of inference’
A Center for Public Integrity review of the documents revealed that the four safety determinations were based mostly on unpublished studies, many of which redacted the name of the chemical being studied to comply with the trade group’s conflict of interest procedures.
The published reports used as safety evidence for three of the four chemicals evaluated not the specific chemical at hand, but a component of it or a structurally related substance. After reviewing the documents at the Center for Public Integrity’s request, researchers who have authored several reports on the GRAS system for the Natural Resources Defense Council were skeptical.
“It’s a daisy chain of inference,” Tom Neltner, an attorney and chemical engineer, said as he looked over the documents.
Without published studies assessing the safety of the specific chemical in question, “It’s hard to say it’s ‘generally recognized as safe,’” added Olson.
As he leafed through documents, Olson said that the four flavors in question should have instead been submitted to the FDA for a thorough evaluation.
But the Flavor and Extract Manufacturers Association contends that it is operating according to government policies. In a written response to questions, Cox pointed out that safety assessments can involve “data on structurally related substances.”
FDA guidance states that such information can be used to support a safety determination, “depending on the circumstances.”
The FDA, however, did not respond to specific questions asking if the agency agreed that the trade association was following the rules on published data.
“When you have an industry like this making safety decisions, we are dependent on this trade association getting it right in order to protect public health,” Neltner said. “But the system is so opaque.”
Little is known about many of these chemicals outside the trade association, said Maricel Maffini, a scientist who has co-authored several Natural Resources Defense Council reports on food additives.
“You have to basically trust their judgment."
This story was co-published with TIME and Huffington Post.