Low Dietary Exposure: Basis for a Determination That a Substance Is GRAS

December 1, 2002

One means of establishing a generally recognized safe (GRAS) status for a food-contact material is to demonstrate that the intended use of the substance results in an extremely low dietary exposure, and conclude, where not foreclosed by adverse toxicology data, that the intended use of the substance is GRAS.

As discussed more fully below, articles in peer-reviewed scientific journals, as well as the U.S. Food and Drug Administration's own analysis that led to the development of the Threshold of Regulation policy, support the conclusion that the intended use of a particular substance can properly be considered GRAS if such use results in an extremely low dietary concentration.

Threshold of Regulation Rule

Although the FDA Threshold of Regulation process has been largely supplanted by the Food Contact Notification (FCN) program, the agency's analysis that led to the development of the Threshold of Regulation rule provides support for the proposition that substances can be considered safe on the basis of low dietary exposure.1 The Threshold Rule represents the agency's determination of a de minimis dietary level for food-contact materials, where the risk presented from the low dietary exposure to a particular substance is so slight as to obviate the need for regulation of the substance as a food additive.

Specifically, the Threshold of Regulation rule enables the agency to exempt from regulation food-contact substances upon a showing that the dietary concentration of the substance does not exceed 0.5 parts per billion (ppb). (Substances that are demonstrated carcinogens, and substances that, based on their chemical structure, are suspected to be carcinogens, are excluded from clearance under the Threshold of Regulation process.)

The rule also exempts those substances already regulated as direct food additives that, when used as intended in food-contact applications, do not present a dietary exposure in excess of 1 percent of the acceptable daily intake established for the substance's direct addition to food. The exemption is applicable so long as the substance does not contain any carcinogenic constituents or impurities with a TD50 of less than 6.25 mg/kg body weight per day.2 Although the Threshold of Regulation clearance program has been largely superceded by the FCN program, FDA's analysis supporting the threshold policy remains valid and, indeed, has been largely incorporated into the FCN guidelines.

In developing the Threshold of Regulation policy, FDA recognized that of all classical toxicological endpoints, carcinogenesis occurs in animals at exposure levels generally lower than for other toxic effects. Thus, FDA determined that a threshold based on the carcinogenic end point would provide a conservative measure of protection.

To set the appropriate threshold value, the agency reviewed the carcinogenic potency data for compounds in the Carcinogenic Potency Database (CPDB) compiled by Gold et al.3 and established a threshold level that balanced the risk of exposure to compounds with unknown toxicity (considering the possibility that the compound could later be found to be a carcinogen) with the need to reach a level that was technologically achievable. If the threshold level was set too low as to not be achievable for any reasonable use of a compound, the policy would not have the intended effect of reducing the regulatory burden for the agency. If the threshold level was set too high, the proposed use of an untested compound could result in exposure at a potentially unsafe level.

Although initially proposing a threshold of 50 parts per trillion (ppt) (the level that resulted in an 85 percent probability that the increased risk from cancer would be less than one in one million if the compound were later found to be carcinogenic), the agency ultimately determined that the 0.5 ppb level represented a reasonable balance between the necessary conservatism and practical utility.4

In the Threshold of Regulation rule, FDA acknowledged the right of a manufacturer to determine independently that a substance used in a particular application does not make the substance a food additive, but FDA reserved to itself the authority to issue formal Threshold of Regulation exemptions. Because FDA's development of the Threshold of Regulation policy was based on recognized scientific procedures, establishing that the expected dietary concentration of 0.5 ppb, even for compounds with unknown toxicity, is a sound basis upon which to conclude that the intended use of a substance is generally recognized as safe.

Further Scientific Analysis of Low Dietary Exposure

Additional scientific evidence supports the conclusion that substances with extremely low dietary concentrations can be considered safe. In 1990, a panel of renowned independent experts convened by the Canadian Center for Toxicology published its conclusion that substances present in the diet at concentrations of 1.0 ppb or below can be considered safe even if no toxicity testing has been performed on the specific chemical, provided that there is no reason to believe that the substance demonstrates unusual toxicological properties.5

The panel also concluded that, for substances with toxicity data that indicate a lack of genotoxic potential, a higher level (on the order of 10 to 15 ppb) may be determined safe based upon classical toxicological principles. Based on the principles discussed in the Canadian report, it is reasonable for a company to take the position that dietary exposure levels of nontoxic substances on the order of approximately 1.0 ppb may be considered GRAS.

More recently, three well-respected officials at FDA's Center for Food Safety and Applied Nutrition published a paper providing a scientific basis for an expanded threshold scheme.6 The chemists examined data on 709 carcinogens listed in the Gold database and found that the structure of an untested substance can be a strong indicator of whether it is likely to be a carcinogen.7 Further, the authors found that the results of short-term toxicity data and genotoxicity tests-i.e., Ames assays and LD50 tests-can be strong indicators of carcinogenic potency.8

As a result of their findings, the FDA personnel concluded that the dietary threshold could be expanded to as high as 15 ppb for certain substances, depending on the structure of the substance and available toxicity data. Specifically, the authors concluded that the following tiered threshold scheme was supported by their review of the data:

  1. A dietary threshold of 4 ppb to 5 ppb for substances lacking structural alerts (i.e., those substances not belonging to the following structural classes of substances known to be carcinogens: N-nitroso compounds, strained heteronuclear rings, alpha-nitro-furans, polycyclic amines, hydrazines/triazenes/azides/azoxy compounds, organophosphorous compounds and heavy metal-containing compounds).
  2. A dietary threshold of 4 ppb to 5 ppb for substances with structural alerts (other than N- nitroso and benzidine-like compounds) testing negative in the Ames assay.
  3. A dietary threshold of 10 ppb to 15 ppb for those substances testing negative in the Ames test and having an LD50 above 1000 mg/kg.

The authors also noted that the use of a substance that results in extremely low dietary concentrations (i.e., on the order of 50 to 100 ppt) is not likely to pose a significant carcinogenic risk even if it contains an alerting structure.9 While the threshold scheme discussed in this article does not represent the promulgation of new threshold criteria by the agency, it provides further guidance for a manufacturer's or supplier's evaluation of food-contact materials. Indeed, it may be considered a suitable basis for taking self-determined GRAS position for a dietary exposure at levels higher than 0.5 ppb, or even 1 ppb, for substances that qualify under the analysis set forth in the paper.


If the resulting exposure from the intended use of an uncleared food-contact substance is sufficiently low, the manufacturer or supplier may want to consider whether such use of the substance may properly be considered GRAS. Because the FD Act specifically excludes GRAS substances from the premarket approval requirements applicable to food additives, a manufacturer or user of a substance may legitimately conclude without approaching FDA that a particular use of the substance does not require a food additive regulation on the basis that it is generally recognized as safe among experts qualified by scientific training and experience to evaluate its safety. If such a conclusion is reached, the substance can be marketed into food-contact applications without specific premarket clearance or review by FDA.


1See 60 Fed. Reg. 36,582 (July 17, 1995).

2Id. The TD50 is the calculated dose that causes cancer in 50 percent of the test subjects.


4See Alan M. Rulis, Threshold of Regulation: Options for Handling Minimal Risk Situations, Chapter 14, in FOOD SAFETY ASSESSMENT (American Chemical Society, 1992).

5See Munro, et al., Safety Assessment Procedures for Indirect Food Additives: An Overview, 12 REGULATORY TOXICOLOGY AND PHARMACOLOGY 2 (August, 1990).

6See M.A. Cheeseman et al., A Tiered Approach to Threshold of Regulation, 37 FOOD AND CHEMICAL TOXICOLOGY 387 (1999).

7When FDA reviewed the available data and developed the 0.5 ppb Threshold of Regulation, the Gold database contained far fewer compounds (477 as compared to 709).

8The LD50 is the acute dose that causes the death in half (50 percent) of a group of test animals. It is a measure of the acute toxicity of a substance.

9See Cheeseman, et al., at p. 401.