Food and Drug Administration

The Food and Drug Administration (FDA), which resides within the Department of Health and Human Services, has a number of divisions. The primary one of interest to this report is the Center for Food Safety and Applied Nutrition (CFSAN); most of the FDA's assessment of potential human health risks from exposure to chemical substances is conducted by CFSAN in conjunction with its regulatory responsibility over additives and contaminants of foods and cosmetics.

The principal legislation on which FDA's authority is based is the Federal Food, Drug, and Cosmetic Act (FFDCA). Although it has been much amended over the years, the original act dates to 1906, making it by far the oldest among federal laws concerned with the regulation of public health risks from toxic substances. As such, much of the methodology for safety evaluation and risk assessment had its origin and early evolution in the implementation of parts of the FFDCA. The act had its origin in response to widespread scandals and "muckraking" exposés of poisonings from dangerous patent medicines, unwholesome meat packing, adulterated foods, and misrepresentations in labeling. Accordingly, the provisions of the act stress avoidance of "filthy, putrid, or decomposed" ingredients, sanitary conditions for processing and packing, proper identification and labeling, and strict limits to prevent "adulteration" of foodstuffs. It is in these adulteration provisions that toxicological risk assessment issues arise--foods are considered adulterated under the act when they contain "added substances" that are poisonous or injurious to health. The application of the act becomes somewhat arcane because the law distinguishes several categories of added substances: food additives, color additives, pesticides, and animal drugs. The question of pesticides is further complicated by the fact that regulatory authority over pesticides is shared by FDA under the FFDCA and the Environmental Protection Agency under the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA).

"Food additives" (regulated under §409) exclude adequately tested substances listed by the agency to be recognized as safe "among experts qualified by scientific training and experience to evaluate its safety" (§201); otherwise, the safety of additives is established by the agency's granting of a petition by the would-be user (although agency initiative is also allowed and pursued in practice). The petition must contain experimental and toxicological data bearing on the evaluation together with a statement of the conditions of proposed use. In its response, the agency specifies conditions of permissible use (which may differ from those proposed) and maximal concentrations that may remain in the food when marketed. Section 409 specifies that, in considering what uses are safe, "the Secretary shall consider among other relevant factors...the probable consumption of the additive,...the cumulative effect of such additive in the diet..., taking into account pharmacologically related substances,...[and] safety factors which in the opinion of experts qualified by scientific training and experience...are generally recognized as appropriate for the use of animal experimentation data." (Although this is phrased quite generally, this still ranks as one of the more specific statements about risk assessment methods to be found among environmental laws.) Section 409 also stipulates that tolerances should be set no higher than is "reasonably required to accomplish the physical and other technical effect for which such additive is intended" notwithstanding the fact that higher levels might be deemed safe. "Color additives" are regulated under a separate section of the act (§721); other than some procedural differences, however, the risk assessment provisions are similar to those applying to additives.

This methodologic prescription applies only to non-cancer toxic effects, however, because at §409(c)(3)(A) the FFDCA contains a very specific statement about how the safety of potentially carcinogenic food additives is to be treated. This is the well known "Delaney Clause," named after the sponsor of the 1958 amendment under which the provision was included in the act. It states that "no additive shall be deemed to be safe if it is found to induce cancer when ingested by man or animal, or if it is found, after tests which are appropriate for the evaluation of the safety of food additives, to induce cancer in man or animal." The rationale cited at the time of the Delaney Clause's adoption was that carcinogens may be without a threshold concentration of toxic action; thus no exposure level could be declared "safe." This stipulation prohibits consideration of the quantitative level of risk that an additive might pose, effectively avoiding the quandary faced under other environmental laws of defining "acceptable" levels of cancer risk.

The Delaney Clause specifically exempts "the use of a substance as an ingredient of feed for animals which are raised for food production" if it is found that "no residue of the additive will be found (by methods of examination prescribed or approved by the Secretary...) in any edible portion of such animal after slaughter...or in any food...derived from the living animal" [§409(c)(3)(A)]. This so-called "DES proviso" was added (in 1962) to allow the use of potentially carcinogenic animal drugs (such as diethylstilbestrol, or DES) as long as they did not harm the treated animals and left "no" residues in the derived food products. The weakness of this formulation became evident as methods for detection of chemical residues became more and more able to detect tiny, even infinitesimal amounts. This led to a quandary: the Secretary could fail to specify the most sensitive existing methods (thereby technically avoiding "detection" of chemicals known scientifically to be present) or he could specify that technical advances in detection should be used (thereby indirectly reversing decisions about "safety" of additives even though knowledge about their safety was not what was changing). Debate about the Sensitivity of Method standards produced the realization that the true issue was not about changing detectability, but about the potential for minute quantities of the agent to cause meaningful risk. This debate led to the development of the first methods for quantitative risk assessment of carcinogens at the FDA.

As with most environmental laws, the mandates in the FFDCA about risk are phrased generally and depend on interpretation. Section 409, applying to additives, requires that only uses that may be demonstrated to be "safe" be permitted. Soon after this section's addition to the FFDCA in 1958, the agency officially defined "safe" as meaning "that there is a reasonable certainty in the minds of competent scientists that the substance is not harmful under the intended conditions of use" but recognized that absolute safety could not be definitively guaranteed (21 CFR 170.3). (This has commonly been codified into the phrase "a reasonable certainty of no harm," which is widely regarded as a quotation from §409, although it does not in fact appear in the act.) Under §409, consideration of benefits and costs is not allowed.

Section 408, applying to non-concentrating pesticide residues, requires setting tolerances "to the extent necessary to protect the public health," but also states that "appropriate consideration" be given "to the necessity for the production of an adequate, wholesome, and economical food supply." That is, costs and benefits are to be weighed, albeit in an unspecified way.

As with other environmental laws with generally phrased mandates about risk, the specifics of how risk assessment is conducted in practice at the FDA depends on the particular procedures put in place to implement the mandate. Remarkably little of this implementation is firmly documented in citable policy documents, guidelines, or standard operating procedures. This is particularly true of the FDA. Some ascribe this to a desire to maintain as much flexibility as possible in the face of the rigidity and draconian nature of decisions mandated under the Delaney Clause, but it is perhaps more reasonable to note that the history of risk assessment at FDA is long and represents a period of considerable evolution of the role of risk considerations in regulation, from qualitative, ad hoc, and peripheral to quantitative, codified, and central. Much of the methodology was invented in attempts to respond to new and emerging needs from the regulatory process. In any case, the methods are codified largely in the history of evolving practice at the agency and in the documentation of regulatory actions (e.g., in the preambles to rules, laying out methods of analysis, in Federal Register notices).

To a great extent, the FDA relies on seminal publications outlining risk assessment principles as the grounding for its methods. These include the Red Book and the OSTP Principles. These expert consensus documents largely reflect compilation of insights and approaches first developed at FDA along with their elaboration and further development by the agency and other risk-assessing institutions. Unlike the EPA, however, the FDA has no officially published "guidelines" that establish standard methods for conducting risk assessment.