Thresholds for Allergenic Foods

Thresholds can be defined as the maximum amount of an allergenic food that can be tolerated without producing any adverse reaction. An individual threshold is the maximum amount of an allergenic food that can be tolerated by a specific food-allergic individual. A population threshold is the maximum amount of an allergenic food that can be tolerated by the entire population (or a representative sub-population) of individuals with a specific type of food allergy. Thresholds can also be referred to as NOAELs (no observed adverse effect levels) or MTDs (maximum tolerated dose).

Operationally, an individual's NOAEL is determined by a double-blind, placebo-controlled oral food challenge (DBPCFC) conducted at an allergy clinic. During DBPCFCs, the allergist will administer increasing doses of the allergenic food disguised in some other food matrix. The lowest dose that provokes a reaction is the LOAEL or lowest observed adverse effect level and the next lower dose is the NOAEL. Most food-allergic individuals do not know their individual threshold dose because they have never had a DBPCFC as part of their diagnostic work-up. However, knowledge of the individual NOAEL and LOAEL is quite beneficial to the individual because it will dictate the degree of care needed in the implementation of a safe and effective avoidance diet.

Population thresholds are more difficult to discern. Information is needed on the individual thresholds for a group with a specific type of food allergy e.g. peanut allergy. If information is available on a larger number of subjects, then the population threshold will be known with more certainty. Obviously, in the determination of population thresholds, it is quite important to have a representative group of individuals with the specific food allergy and especially to include individuals who are extremely sensitive to that particular food. The most sensitive individuals dictate the low end of the range for the population threshold. The population threshold is often described in terms of percentages e.g. ED10 (the eliciting dose predicted to provoke a reaction in 10% of individuals with a specific food allergy). Obviously, the ED10 is not a safe dose for the entire population with that food allergy because it predicts that 10% will have adverse reactions at that dose. But, the ED10 is useful for comparison of the potency of one allergenic food vs. another and also is useful in predicting how many individuals should respond to any specific exposure e.g. a food contaminated with an undeclared allergen.

The definition of the meaning of adverse reaction is also important to the understanding of this concept. DBPCFCs can be used to determine an individual's LOAEL, the lowest dose that provokes an observed reaction. Since the reaction must be observed by the physician, then the response must involve the elicitation of objective symptoms (symptoms that can be visually observed by others e.g. hives, dermatitis flares, vomiting, etc.). For the purposes of threshold doses, FARRP believes that subjective responses should not be used. Subjective responses are those that cannot be independently verified by the physician observer. Examples of subjective responses might include mouth itching. There is some controversy about this approach but principally the controversy surrounds how to handle the symptom of abdominal pain in infants and young children. Abdominal pain is a subjective response but clinical experience indicates that, especially in infants and young children, abdominal pain is a noteworthy symptom that should be considered. Thus, FARRP endorses the clinical viewpoint that an individual LOAEL can be established on the basis of any objective symptom or on the basis of abdominal pain in infants and children.

While food-allergic individuals can and do react adversely to exposure to small quantities of the offending food, it is now well-documented that food-allergic individuals have threshold doses below which they will not experience adverse reactions. Considerable variation occurs in the individual threshold dose among those with a given type of food allergy. For example, from controlled clinical challenge trials in a study of 450 peanut-allergic individuals, peanut thresholds (NOAELs) for peanut-allergic individuals ranged from 0.1 mg up to perhaps 8 g of whole peanut (an individual peanut can weigh from 500 - 800 mg). Furthermore, the threshold dose may vary from one allergenic food to another. For example, evidence indicates that the threshold dose for peanuts on a population-wide basis is considerably lower than the threshold dose for soybeans. From a risk assessment and regulatory perspective, a population threshold must be defined, which would be the largest amount of the allergenic food that would not cause a reaction when tested experimentally in a defined population of allergic individuals.

Food-allergic individuals are typically advised to totally avoid the offending food(s), essentially assuming that the threshold dose is zero. With increased awareness and regulatory attention paid to food allergy in the past 15-20 years and in the absence of consensus or guidance on threshold levels, the food industry exercises considerable caution in its labelling practices. Thus, ingredients derived from allergenic sources but containing only miniscule amounts of allergenic protein are labelled as "Contains X". And, voluntary precautionary (advisory) labelling has become extremely common e.g. "May Contain X" (or variations). Food-allergic consumers are advised to avoid all foods bearing such advisory labelling statements. Behavioral research indicates that the quality of life for these consumers suffers as a consequence. Research also indicates that some food-allergic consumers are increasingly ignoring precautionary labelling in particular and thus intentionally taking risks.

Regulatory guidance on thresholds for allergenic foods is sorely needed to improve the quality of life for food-allergic consumers and to focus limited regulatory resources on the genuine risks that exist to these consumers. Of course, regulatory (population) thresholds must be scientifically based but sufficient evidence exists to allow the development of regulatory thresholds for some of the most hazardous allergenic foods including peanuts and milk.