The FDA indicated general agreement with the GAO recommendations but has never enacted changes in response to the report.
Last year, the Pew Health Group added to the debate by publishing a report with estimates on the number of substances the FDA has reviewed (7,000) since 1958, those it hasn't (3,000) and those it has never been notified about (1,000). These include ingredients added directly to food but also indirect ingredients, which come into contact with food during processing and in packaging.
Pew plans to publish more studies in coming months focusing on how other countries treat U.S. GRAS ingredients and how the program deals with conflicts of interest between scientists and manufacturers, among other topics.
Although Pew researchers acknowledge that the 1997 rule change encouraged more manufacturers to submit notifications about new ingredients to the FDA, very few of those manufacturers ask for an FDA review, preferring to make their own determinations. The change also limits the opportunity for the public to provide input on new ingredients.
"The FDA no longer writes specific regulations for the use of the substance or puts those rules up for public comment by consumers, academics and competitors," said Tom Neltner, who heads the Pew Health Group project.
The GAO report noted that a few substances initially considered GRAS have later been found to pose health risks and were subsequently banned by the FDA. The most recent of these decisions was in 1986, when the FDA prohibited the use of sulfites as a preservative on raw produce due to severe allergic reactions in those sensitive to sulfites.
The GAO investigated the GRAS system at the request of Congress in 2010, and Pew formed a group to dig further shortly after that. More recently, the heart association joined the GRAS debate as part of its effort to reduce salt consumption.
Earlier this year, the heart association presented a 19-page statement to the FDA on salt, asking the agency to modify salt's GRAS status and set limits on its safe use in products. The letter also called on the FDA to strengthen the GRAS process, noting that it currently "relinquishes too much of the agency's authority to food manufacturers and does not do enough to ensure the safety of substances that are added to food."
Those who favor the FDA's current approach often point to the power of the marketplace as form of regulation: Unsafe ingredients are bad for the bottom line.
"Sick customers are not good business," said Tony Pavel, an attorney who assists clients with regulatory applications to the FDA.
"Are there areas in the program where things could be improved?" he asked. "I'm sure there are, as with any program. But I would be hard-pressed to name a more successful program with such an excellent safety record in the U.S., if not anywhere in the world."
The federal program, while it could be improved, "generally works well," said John Endres, chief scientific officer of AIBMR Life Sciences, which helps companies prepare GRAS submissions. Even if some manufacturers could cut corners for GRAS self-affirmation, Endres said they would have a hard time selling an ingredient backed by weak science to a food company.
"Who'd want to take on an ingredient that could result in a lawsuit or, God forbid, hurt someone?" he asked.
Matulka, Burdock Group's director of toxicology, argues that even if the FDA doesn't assess or even know about new ingredients on the market, it can still review food labels and issue warning letters to companies when necessary.
His recent article cited FDA warning letters sent to the makers of foods containing melatonin, stevia and rhodamine B as well as one that added caffeine to alcoholic beverages.
He wrote that such letters indicate that the "FDA has, at any time, the ability to review food ingredients and to request additional information on the scientific basis used to substantiate use in food ... or to request the removal of an ingredient if adequate documentation is not provided."
But Neltner said that if the FDA doesn't know a chemical is in use or if an ingredient is not required to be listed on a label, the agency can't exercise its authority. "Most of these things never appear on labels," he said.
FDA officials said they are doing the best they can to juggle priorities in the face of financial constraints.
Dennis Keefe directs the FDA Office of Food Additives Safety and calls the 1997 change an effort to "streamline" a "cumbersome and burdensome process" that the FDA did not have the funds to continue.
"When we think about where the FDA should expend limited resources, we try to prioritize our efforts to focus on areas where there has been an identified public health issue," Keefe said.
Most observers agree there is inadequate congressional funding to return to a system in which the FDA tests all new ingredients. But O'Reilly has suggested adopting a model used for pharmaceutical testing in which the manufacturer pays a testing fee to the FDA, whether it affirms safety or not. The change would require congressional approval.
Until then, O'Reilly said consumers are left with little more than the right to sue if an ingredient ends up causing a problem.
"The problem is you don't know until a sufficient number of people have a problem," O'Reilly said. "…So there will be instances in which the manufacturer is held liable, but it will be done through the tort injury system, three, four or five years later. It won't be the government doing what it said it would do in 1958, which is stopping the product before it reaches the consumer and demanding adequate testing and data."
Neltner advises consumers to consider the source when making food decisions.
"Ask yourself if you can trust that company, because it ultimately comes down to them making the safety decisions," he said. "Do they see themselves in it for the long term or short term? Some take their responsibilities very seriously and are very conscientious and are going to be very careful. But it's also a global market out there today."
firstname.lastname@example.org Twitter @monicaeng