Chemicals may be added to foods for a number of reasons.  Typically, these reasons may be broadly classified as consumer-related or distributor-related.  However, the goals of each group may be closely intertwined and the distinctions are often less obvious upon close examination of food distribution and purchase.

One purpose for food additives is to increase the quality of the food.  Preservatives decrease the likelihood of degradation and bacterial growth thus extending the shelf-life of a food and decreasing the risk of food poisoning.  Additionally, this practice provides a consumer service in that food may be purchased at the convenience of the customer (less trips to the market).

A second reason additives are used is to enhance to appearance (colouring) or taste appeal (flavourings and flavour enhancers) of the food.  Moreover, the use of artificial flavourings will decrease the cost of production of some foods.

Species differences in the effect of the additive

Species differences in the absorption, distribution, metabolism, and excretion of the additive -- As an example, the elimination pathways of the additive may become saturated at very high doses, leading to artificially high body loads of the additive in laboratory animals.  This may potentially lead to erroneous "false" toxicity that is unlikely to occur in humans.  A situation of this type was encountered in the testing for saccharin, as discussed below.

Poor design correlation -- Humans will typically ingest very small amounts often sporadically during a lifetime and doses may vary.  This type of administration in laboratory settings is difficult, if not impossible, to simulate.

Another potential problem with testing the toxicity of foods is the mistaken idea that "natural" foods (foods that have not been subjected to chemical additives or processing) are safe foods.  As illustrated below, "natural" does not always mean safer and these situations are almost impossible to test.

Hyperkinetic behaviour in children -- there is a correlation between an ADD-type reaction in some children who eat tartrazine-containing foods.  The evidence, however is circumstantial and tenuous.  In many studies in which children were switched to tartrazine-free diets, the hyperactivity disappeared.  In other studies, there was no change in the behaviour.  Some of these studies were poorly controlled, used subjective measurements, or did not screen the children adequately.  Therefore, there still exists a controversy over the real incidence of hyperkinesis in response to tartrazine.

Dermatologic response -- a better correlation is found between tartrazine and rash.  It has been demonstrated that tartrazine will increase histamine release, causing wheal development and itching.  There is cross-sensitivity with other dyes including erythrosine and sunset yellow.  There is a weak correlation between asthma and tartratzine (patients with asthma show a higher incidence of tartrazine sensitivity and tartrazine may precipitate an asthma attack) and a stronger correlation with aspirin sensitivity.  It appears that the sensitivity is not mediated by antibody formation and may therefore NOT represent a true allergic reaction, but rather is the result of a hypersensitivity to the compound.  The highest reported incidence of tartrazine sensitivity in the general population is 13%.  Interestingly, the naturally occurring dye annatto has a reported incidence of sensitivity of 26%.  Therefore a food that is labelled as free of artificial colouring may contain the natural colouring and exhibit a greater chance of adverse reaction.

Monosodium Glutamate (MSG) -- This compound is used as a flavour enhancer in prepared foods and restaurants.  There is a relatively high incidence of mild reaction to the additive.  In these people it most often produces a "pressure" in the head, "tightness" in the face, headache, and facial flushing.  It has occasionally produced seizures in epileptic individuals.  Extremely high doses have been shown to produce hypothalamic and retinal lesions and reproductive toxicity in both males and females.

Aflatoxin -- This mycotoxin may present as a food contaminant.  It is produced by the mould Aspergillus, which may grow on legumes (peanuts) and grains (corn).  Testing for Aspergillus is routine in the U.S.A. (exposure to ultraviolet light will expose the fungus) and batches of grain that contain the mould are destroyed.  Additionally, there are strict import requirements which reduce the incidence of aflatoxin toxicity.  The toxicity was discovered when turkeys were fed forage that had been infected with Aspergillus and later developed liver tumours.  The toxicity presents as both hepatoxicity (liver failure) and liver cancer.  Mechanistically, both effects are thought to be mediated by an epoxide intermediary.  Another example of "natural" foods that may be more toxic than processed is the case where peanut butter that had not been processed (the peanuts were not treated with a fungicide) was contaminated with aflatoxin, which resulted in hepatoxicity.

Case Study, The Spanish Oil Syndrome -- This food contamination incident is a classic case involving widespread toxicity to a contaminated food product.  It occurred in 1981 around Madrid, Spain and affected over 20,000 persons with at least 351 deaths.  The cases presented acutely with transient, severe pulmonary œdema, exanthema, and eosinophilia (there was a two-week latency before the onset of these symptoms, but this was not discovered until the causative agent was identified).  The patients then demonstrated a second, chronic phase of toxicity that presented with muscle atrophy, skin lesions, weight loss, and vasculitis.  The chain of events appears to be that rapeseed (canola) oil was imported into Spain.  As prescribed by Spanish law, it was adulterated with aniline and therefore was not suitable for consumption as a food.  It was then refined and sold as cooking oil (apparently a common practice with adulterated oils -- heretofore performed with no adverse reactions).  However, separate batches were refined and processed differently and also were diluted with different oils.  The exact contaminant and/or process that resulted in toxicant formation was never identified (the processes employed could not be duplicated) nor could the toxicity be reproduced in the laboratory.  The exact causative agent is still not know, nor will it probably ever be identified.  This case illustrates the importance of standard operating procedures for handling and processing foods and toxicity testing of ALL additives and end products.

Two additional cases of food contamination involve the use of mercurial fungicides to preserve grains that were used as food sources for livestock

One such case involved the use of  alkylmercury fungicides on grain which was fed to cattle, which was in turn ingested by people in Iraq in 1971-2, resulting in 6000 cases of toxicity and 500 deaths.

Additionally, in 1969, a family in New Mexico used mercury-based fungicides on their grain, which was fed to pigs, which were eaten by the family.  Three of the ten children exhibited signs of CNS toxicity and one child was exposed in utero, resulting in cerebral palsy.