Center for Food Safety & Applied Nutrition Office of Premarket Approval May 1993 (Effective June 18, 2001, Office of Premarket Approval is now Office of Food Additive Safety. See updated contact information) |
The latest version of this guidance issued in March 2006. Below is an earlier version.
This document describes the types of chemical and technological data that FDA's Office of Premarket Approval in the Center for Food Safety and Applied Nutrition considers necessary for the evaluation of petitions seeking regulation of the safe use of direct food additives or for the affirmation of the use of food ingredients as generally recognized as safe (GRAS).
As prescribed by the Federal Food, Drug, and Cosmetic Act (the Act) and Title 21, Code of Federal Regulations (21 CFR) 170.3(h)(i), food additives must be shown to be safe (i.e., to have a reasonable certainty of no harm) under their intended conditions of use before they can be intentionally added to food. Food additives generally fall into two broad categories: (1) those that are added directly to food, as codified in 21 CFR Part 172, and (2) those that are added indirectly to food through contact of the food with packaging materials, processing equipment, or other food-contact materials, as codified in 21 CFR Parts 174-178. In addition, there is another class of food additives that does not neatly fit into either the direct or indirect food additive categories. These substances are called secondary direct food additives, and are codified in 21 CFR Part 173. These are substances whose functionality is required during the manufacture or processing of food and are ordinarily removed from the final food; although residuals might carry over to the final product, these residuals are not expected to exhibit any technical effect in food. Examples of secondary direct additives include enzyme immobilizing agents, ion exchange resins, boiler water additives and other processing aids.(1)
Section 201(s) of the Act exempts substances that are GRAS from the definition of a food additive. There are two broad categories of GRAS substances codified in 21 CFR: (1) those listed as GRAS in 21 CFR Part 182 during 1958-1962 without having been subjected to a detailed scientific review of all information available on the safety of these substances; and (2) those in Part 184 that have been affirmed as GRAS by the Agency since the Presidential Directive of 1969. This Directive required the Agency to initiate a safety review of the status of all ingredients that were on the GRAS list, with the intention of affirming their use as GRAS, determining that they were subject to prior sanction, or concluding that they should be regulated as food additives. (Substances subject to prior sanction are listed in 21 CFR Part 181.) To affirm the use of a food ingredient as GRAS, the ingredient must be generally recognized among experts qualified by training and experience to evaluate its safety. Safety, under the conditions of intended use, may be shown through (1) scientific procedures or (2) experience based on common use in food prior to January 1, 1958.
Section 409(b)(2) of the Act describes the statutory requirements for food additive petitions. Briefly, these encompass five general areas of information: (1) the identity of the additive; (2) the proposed use of the additive; (3) the intended technical effect of the additive; (4) a method of analysis for the additive in food; and (5) full reports of all safety investigations with respect to the additive. In addition, the petitioner shall furnish, upon request, a complete description of the methods, facilities, and controls used in or for the production of the additive (21 CFR 409(b)(3)) and/or samples of the additive and of food in which the additive will be used (21 CFR 409(b)(4)).
21 CFR 171.1(c) describes in greater detail the data requirements for food additive petitions, including the five basic areas of information and scientific data noted above, as well as other administrative information and environmental assessment requirements. This document addresses chemistry-related issues only.
As described in 21 CFR 171.1(h), certain data and information contained in food additive petitions are available for public disclosure, while other data are not. Questions in this regard should be directed to the Office of Premarket Approval.
The Act does not provide specific statutory requirements for GRAS affirmation petitions. As noted above, Section 201(s) of the Act exempts GRAS substances from the meaning of the term "food additive" and, therefore, from the statutory requirements for food additive petitions. Section 201(s) also provides that general recognition of safety must be established either through (1) scientific procedures or (2) experience based upon common use in food prior to January 1, 1958. Under 21 CFR 170.35, the Agency has established a voluntary GRAS affirmation petition process in order to minimize controversy over whether the Agency agrees that the use of a substance is GRAS.
The eligibility requirements for classification of a substance as GRAS are described in 21 CFR 170.30. For a substance to be eligible for GRAS affirmation based upon its history of use in food prior to 1958, certain criteria must be met. If the substance was commonly used in food in the United States prior to 1958, then information documenting this use must be generally available (21 CFR 170.30(c)(1)). If the substance was only used in food outside the United States prior to 1958, then published documents, or other information, that shall be corroborated by information from a second, independent source that confirms the history and conditions of use, must be readily available in the country in which the history of use occurred, as well as to experts in the United States (21 CFR 170.30(c)(2)). If the substance did not have a common history of use in food prior to 1958, then the substance can be considered for affirmation as GRAS only based upon scientific procedures, as set forth in 21 CFR 170.30(b).
The regulatory requirements for GRAS affirmation petitions differ from food additive petitions in terms of the format, data requirements, and administrative requirements. In addition, there are distinctions in data requirements for GRAS petitions based upon history of use and those based upon scientific procedures. The data requirements for GRAS affirmation of a substance are set forth in 21 CFR 170.30(b) and (c). GRAS affirmation based upon scientific procedures requires "the same quantity and quality of scientific evidence as is required to obtain approval of a food additive" (21 CFR 170.30(b)); thus, in preparing a petition for GRAS affirmation based upon scientific procedures, consideration shall also be given to the requirements set forth in 21 CFR 171.1(c) for food additive petitions, as well as those in 21 CFR 170.35(c)(1). The data requirements for a petition for GRAS affirmation based upon history of use are set forth in 21 CFR 170.35(c)(1); these petitions do not require the same quantity and quality of scientific evidence that is required for approval of a food additive regulation (21 CFR 170.30(c)(1)).
Under 21 CFR 170.35(c)(3), all GRAS affirmation petitions are placed on view in the office of the Dockets Management Branch, where they are available for public inspection; thus, any trade secret information in a GRAS affirmation petition would not be protected under 21 CFR 20.61. Hence, trade secret information should not be included in GRAS affirmation petitions. Additionally, scientific information essential to a GRAS affirmation petition must be available to the public in the open literature.
The following is intended to explain to prospective petitioners the statutory and regulatory data requirements as they apply to substances added directly to food; differences between requirements for GRAS substances and food additives are noted.
Information allowing the unequivocal identification and characterization of the food additive or GRAS substance should be provided. Such items include:
Information on the method of manufacture of an additive or food ingredient is necessary not only for identification of that substance, but also for identification of any impurities in the additive or food ingredient that have no functional effect in food. For substances of natural origin containing known toxicants, the ability of the manufacturing process to reduce toxicants to acceptable levels should be addressed.
A discussion of the manufacturing process should include a list of the reagents, solvents, catalysts, processing aids, purification aids, etc. used, as well as a detailed description of the process itself, including all reaction conditions and production controls. A description of alternative methods of manufacture, if known, should also be included.
Specifications for identity and purity of the petitioned additive or ingredient should be proposed, preferably, in the format of the Food Chemicals Codex, 3rd or current edition.(2) If published specifications for the added substance are available, then these should be cited and appropriately referenced. The petitioner may also recommend changes to specifications in the Food Chemicals Codex monographs (cf. 21 CFR 170.35(c)(1)(i)(f)). Specifications should include the following:
In light of the Agency's continuing concern over the level of lead in the food supply, special attention should be paid to the proposed specification for lead. Actual lead levels should be determined and specifications should be proposed accordingly, keeping in mind that the proposed specification should be as low as technically feasible taking into account the source and manufacturing process of the substance, availability of appropriate analytical methodology, and probable human exposure to the substance.
In order to demonstrate conformance with the proposed specifications, at least five batches of the added substance should be analyzed. If the analytical method is a common standard test (e.g., Food Chemicals Codex or Association of Official Analytical Chemists (AOAC) method), only the reference needs to be provided. If the method is not common, if a common method is applied to a new substance, or if a modified standard method is used, a detailed description of the method and validation data for the method should be provided (see Section F for a discussion of analytical data requirements).
Specifications as discussed above serve only as minimal criteria for establishing that a substance intended for use in food is "food-grade". For a substance to be considered food-grade, it shall also have been manufactured, packaged, stored, and transported to the end-user under sanitary conditions that preclude, for example, contamination, formation of undesirable degradation products, and mislabeling. Controls or systems to avoid such problems should be established by manufacturers, processors, packers, and distributors throughout their operations. These practices, combined with adherence to the specifications for identity and purity constitute "good manufacturing practice" or GMP. A further elaboration on GMP can be found in the Food Chemicals Codex.(2).
Data demonstrating the stability of the additive or food ingredient should be included, particularly if the substance is sensitive to environmental conditions, decomposes with time, or otherwise has a limited useful shelf-life or stability. A description of the methodology and all raw data should be provided.
A number of technical effects for which added substances may be used, such as antimicrobial agents, humectants, flavoring agents, surface-active agents, stabilizers and thickeners, are defined in 21 CFR 170.3(o). A substance added to food might have legitimate technical effects in food other than that intended; however, the quantity of the substance added to food shall not exceed the amount reasonably required to accomplish its intended technical effect(s) in the food (21 CFR 172.5(a)(1), 182.1(b)(1), and 184.1(b)).
A discussion of the intended use and use level of the additive or ingredient should include:
Data should be submitted to demonstrate the amount of the additive required to achieve the intended technical effect. To demonstrate the minimum level required to accomplish the intended technical effect, the functionality of the added substance should be evaluated at several levels above and below the proposed use level. Some substances may have a technologically self-limiting use level; that is, the substance has a maximum concentration in food above which the food becomes unpalatable, unappealing, or otherwise unfit for human consumption. In such a case, data should be provided that demonstrate the effect of the substance on the food at several levels above and below the proposed self-limiting level.
The depth and degree of the data required to demonstrate the intended technical effect can vary, depending upon the additive. For example, these data may range from simple organoleptic evaluation to complex chemical analysis. Regardless of the degree of sophistication of the method, the experiments should be carefully designed and all appropriate controls included. For relatively new substances or novel technical effects, samples of food, with or without the added substance, might be requested for analysis; however, such samples should not be provided unless expressly requested during the petition evaluation.
GRAS affirmation petitions based upon common use in food prior to 1958 should contain the following historical information with regard to the use of the substance in food:
If assurance of safe use depends on a limitation imposed on the amount of a substance (additive, associated impurities, or degradation products) that a food can contain, a method is needed that can quantify the substance in food for the purpose of enforcing the limit. Quantifying the levels of a substance in food requires a practical analytical method that can be readily performed in a properly equipped laboratory by appropriately trained personnel. The method must be specific, precise, accurate and reliable; it must also be able to withstand the scrutiny of courtroom cross-examination, if necessary. However, the method must not be so sophisticated or complicated that it can be carried out only in laboratories with highly specialized equipment or by specially trained personnel.
The following is a suggested general format for such an analytical method.
FDA may require samples of the additive or ingredient for a trial of the method, but these samples should be furnished only upon request. If such samples are requested, instructions will be provided as to the types and amounts of samples required by the analyst.
The region for quantitation of the analyte should clearly be above the limit of detection. The limit of quantitation is located ten standard deviations above the average blank signal.(3)
((a - b) / c) x 100
where "a" is the level of the analyte analytically determined in the spiked sample, "b" is the "background" level (i.e., the level determined in the appropriate blank sample, typically the food without additive), and "c" is the amount of analyte added to the spiked sample.
Spiking and recovery experiments should be performed using fortified samples, each analyzed in triplicate. These samples should consist of food that contains the additive or ingredient, and is also fortified with a known amount of analyte. Typically, the samples are fortified ("spiked") at levels of one-half of, equal to, and twice the typical additive use level or analyte concentration (or limit of detection, if applicable). The range of spiking levels should be broad enough to encompass or bracket the total range of use levels, as well as any regulatory tolerances. Results of spiking experiments should demonstrate between 80 and 110% recovery for analyte concentrations above 0.1 ppm, and between 60 and 110% for concentrations below 0.1 ppm.
In some cases, it might also be necessary to confirm the identity of the substance being analyzed by an independent analytical technique. This information would be specifically requested in those cases.
As part of the petition process for regulation of a food additive or ingredient, the consumer exposure to the added substance, or any of its by-products, resulting from ingestion of food containing the substance, must be estimated. This estimate, known as the "estimated daily intake", or EDI, is intended to represent chronic daily intake of the added substance. The EDI is typically calculated for the "average" consumer and the "high" consumer. The Agency considers the "high" consumer as one belonging to that subgroup of the population that consumes, on a regular basis, greater quantities of a particular food than the "average" consumer. For purposes of the estimate, this subgroup is generally represented by the consumers of a given food at the 90th percentile in an "eaters-only" population. The Office of Premarket Approval typically consults several sources for the evaluation of food consumption. These include the USDA/National Food Consumption Survey (NFCS) 1977-1978 and 1987-1988 3-day surveys, the Market Research Corporation of America (MRCA; 1982-1987 Menu Census) 14-day frequency of food consumption survey, and USDA "Situation and Outlook" publications for various food commodities (e.g., "Sugar and Sweeteners"). It should be noted that in the USDA/NFCS 1987-1988 survey, certain sub-populations were under-represented; therefore, this database, in particular, must be used with caution.(4)
Typically, the Office of Premarket Approval will calculate an EDI for the additive based upon the following information provided by the petitioner:
Although the Agency will determine the EDI for a substance, the petitioner may wish to include their own EDI calculation in the petition. The petitioner may consult the above-mentioned references for food consumption data, or develop alternate sources of food intake data, and, at least, provide a mean EDI (to represent the "average" consumer) and the 90th percentile EDI (to represent the "high" consumer), assuming a typical use level of the added substance. However, it is important to note that the EDI must be based on reasonable consumption scenarios that would be permitted under the petitioned regulation; it cannot be based solely on the petitioner's current marketing plans.
Special cases may arise, particularly in the area of substances that could become macro-ingredients in the diet, for which the food consumption and use level information necessary for the estimation of exposure is inadequate or unavailable. Difficulties can arise in estimating intake when current eating habits cannot be reliably extrapolated to predict the probable intake of the new substance. In these cases, new approaches to the premarket estimate must be derived. The petitioner may wish to consult with the Office of Premarket Approval in agreeing to a suitable approach. The potential for large exposures to new macro-ingredients (e.g., fat replacements) reduces the amount of conservatism that can be included in the exposure estimate.
These recommendations do not address administrative, toxicological, microbiological, nutritional, environmental assessment, or labeling requirements that a petitioner might have to satisfy. Information on these or other requirements, as well as specific inquiries regarding GRAS affirmation or food additive petitions can be obtained by contacting the Office of Premarket Approval, HFS-200, 200 C Street, SW, Washington D.C., 20204. (See updated contact information)
The latest version of this guidance issued in March 2006.
Hypertext updated by ear/dms/emw/hrw/msl/cjm 2006-MAR-24