One expert explains why the GRAS discussion is much more complex—and why it should be ongoing.
The Generally Recognized as Safe (GRAS) regulation governing food substances in the U.S. is no stranger to criticism. Critics in particular focus on two issues.
One is the regulation’s independent GRAS conclusion provision. Before an ingredient can be considered safe and legal to market in food, GRAS regulations require that a marketer first determine that the substance is reasonably certain not to be harmful under the conditions of its intended use, based on common scientific knowledge. Marketers can notify FDA of their GRAS conclusions by submitting a GRAS notice to the agency—but they don’t have to. Alternatively, they can take the regulation’s independent conclusion route by still performing the safety review but not submitting the information to FDA before proceeding to market. Groups like the Center for Science in the Public Interest (CSPI) have sued FDA in the past over the independent-conclusion provision, criticizing the process for not requiring FDA review.
Another criticism of the GRAS process is that it is an “NDI loophole.” Regulations state that GRAS substances do not need to also submit an NDI notification to the agency in order to be used in a dietary supplement, as long as the substance is chemically in the same form in both food and supplement. During an April 28 webcast moderated by the Council for Responsible Nutrition (CRN) and hosted by Informa, guest speaker Laura MacCleery, director of strategy and program for CSPI, called this alternate route to the dietary supplement market problematic because an ingredient could theoretically be used in a supplement at a much higher dosage than it might be used in a food. In her opinion, the GRAS safety evaluation might not necessarily adequately assess the ingredient’s safe use at a higher concentration in a supplement.
Following the aforementioned webcast, I interviewed a regulatory expert who explained that the issue is actually a lot more complex than that. John Endres is chief science officer at AIBMR Life Sciences Inc. (Seattle, WA), a scientific and regulatory consulting and compliance firm specializing in GRAS and NDI work.
The issue to consider, he says, is not so much the individual dosage used in a single food compared to the dosage used in a single supplement; rather, the focus should be on the overall daily safety level set for the substance itself and what consumers are exposed to.
When performing a GRAS review, experts first determine what the intended uses of a substance will be. Based on the intended uses, they determine what the exposure to the substance will likely be, taking into account average and high-end users, as well as age and gender. They then set an acceptable daily intake value (ADI)—the threshold above which there is no data to support safety at this time—based on the no-observed-adverse-effect level (NOAEL) and with a broad, 100-fold uncertainty factor built in. Through this, they are able to determine what level of exposure to the substance is “reasonably certain to be safe for chronic consumption every day for the rest of your life,” Endres explains.
“The ADI, by definition, should not be exceeded, as it is the threshold determined by toxicological safety assessments,” he states. (For what it’s worth, he mentions that while one should never exceed the ADI, the broad safety margin built into an ADI likely protects consumers even against uses that slightly exceed the ADI.)
GRAS conclusions are meant to account for the substance’s intended uses—and intended uses should adhere to the safe level. For instance, says Endres, “If it is found that the exposure is too high, the intended uses need to be modified so that the daily exposure falls below the ADI.” Companies must take those intended uses into account and dose and design their products’ use accordingly.
Of course, the elephant in the room is how consumers are going to use products, at the end of the day, and what their exposure to a substance will realistically end up being as a result of those uses. Will they consume a 500-mg dose in a supplement—plus additional doses in multiple food servings in a day? In such a case, their intake would be higher than the ADI. As Endres says, “What it all boils down to is whether a consumer who eats the food will also take the dietary supplement and exceed the ADI—or whether, for many reasons such as cost, etc., they will either eat the foods or the dietary supplement, but not both. That is the crux of the argument and a very important consideration with respect to safety.”
He continues: “A series of toxicology studies might, for example, conclude an ADI (acceptable maximum daily exposure) of 500 mg/day. This means the EDI (estimated daily exposure from all sources) must be less than this ADI. For foods, the EDI is calculated using very sophisticated predictive software that calculates exposure at the 50th (average) and 90th (high end) percentile based upon the intended food uses and specific concentrations stated in the Independent GRAS Conclusion. Now, a dietary supplement company may choose to market the ingredient in a dietary supplement. In this case, they will essentially be making an assumption that consumers will either buy and consume the dietary supplement or purchase conventional food products, but not consume both—and therefore remain below the EDI from all sources. The high prices of functional ingredients may lead to support for such a conclusion because consumers can knowingly make a choice between the food or the supplement.”
An ingredient like hemp cannabidiol (CBD) has put the spotlight on this consideration. For instance, some have surmised that FDA—although the agency has not said so publicly—might be more open to one day allowing CBD in dietary supplements but might be hesitant to do the same for foods because there’s a higher likelihood that consumers might eat multiple servings of a food per day—and perhaps also a supplement—and eclipse a safe usage level versus taking a single daily supplement.
If GRAS conclusions are supposed to affirm an ingredient’s safe use, then it’s imperative that they reflect how products are most likely to be used—and if it’s found that daily exposures are higher in practice, then safety data should certainly be reassessed. That’s an ongoing discussion.
But, going back to the GRAS regulation itself, this doesn’t mean that the GRAS process is broken, Endres emphasizes.
On the contrary, he says, the GRAS procedure works very well. “For the most part, industry and regulators find the GRAS process quite credible,” he says, adding, “The program is working very well in my opinion.” As evidence, he points out that that there are few adverse events with regard to substances in food compared to, for instance, adverse effects reported from taking pharmaceuticals.
Is the GRAS process allowing dangerous substances to widely circulate in foods? It’s unlikely, he says. And, while he does acknowledge that there are “probably some ‘bad players’ out there, or the perception that there could be ‘bad players’ out there performing GRAS assessments,” the fact is that if FDA sees a widespread problem, it will act. “The FDA will quickly take action when they see a ‘signal’ building from serious adverse event reporting and therefore a concern for public safety,” he says. “Fortunately, cases like this are fairly rare.”
For those concerned about the potential consumer exposure to a substance like CBD, conversations about intended uses and actual uses are certainly relevant and bear discussion. After all, if we’re talking about safety, we have to consider exposure from all sources. There may also be room to discuss whether the GRAS process—particularly the independent GRAS conclusion process—could be more transparent. Could there one day be a public government database of independent GRAS conclusions, similar to how FDA makes public a database of GRAS notices it has reviewed and responded to? Both FDA and dietary supplement industry leaders have already discussed whether a mandatory product listing would benefit the dietary supplements industry. So why not GRAS?
Endres is open to the idea. “I think a system could easily be developed to have somewhat greater transparency to weed out the rare or inadvertent ‘bad player,’” he says. “A requirement to be listed on a GRAS Independent Conclusion database”—like the one AIBMR itself maintains on its own website as a public service—“managed by FDA would be a good start. This could include information such as: company, ingredient, simple manufacturing description, intended uses, resulting exposure, and a few bullet points about why this is safe for chronic consumption every day for the rest of your life (such as the NOAEL from the pivotal 90-day toxicology study). This should be on the FDA website and publicly available. Then, if a company tries to sell an ingredient and it’s not on that list, something might be wrong.”
So, maybe the GRAS system isn’t broken, but maybe we should be talking about ongoing transparency and a need to continually assess whether ADIs reflect what’s actually going on in the market. The discussion, as some would say, is evergreen.