Truth: No thorough allergenicity assessment is conducted on GM foods

Myth at a glance

The public is told that GM foods are no more likely to cause allergic reactions than non-GM foods and that they are fully assessed for allergenicity. But this is untrue. Genetic engineering can create novel proteins that have no history of safe use in food, raising the potential for allergic reactions.

The absence of reliable methods for allergenicity testing and the lack of rigour in current allergy assessments mean that it is impossible to reliably predict whether a GMO will prove to be allergenic.

If a GMO did prove to be allergenic, it would be almost impossible to find out, as no post-commercialization monitoring is carried out anywhere in the world.

Most food allergies are caused by a reaction to a protein in a food. The DNA of an organism contains instructions for making proteins. Genetic engineering changes the DNA of a food, and the altered DNA can in turn create new proteins. GM foods could create new allergies in two ways: the new proteins could cause allergic reactions (be allergens) themselves, or the new proteins could sensitize people to existing food proteins.

The website GMO Compass, which is run by the public relations firm Genius GmbH,claims that GM plants pose no greater risk than new varieties of crops obtained through conventional breeding, or the importation of new exotic foods, which can also result in new allergens appearing in the diet.1

But independent scientists disagree. A 2003 review states that compared with conventional breeding, GM has a “greater potential to introduce novel proteins into the food supply” and increase the likelihood of allergic reactions.2

A study on humans confirmed the potential for GM to create novel and potentially allergenic proteins. One of the experimental subjects showed an immune response to GM soy but not to non-GM soy. GM soy was found to contain at least one protein that was different from the profile of proteins present in the non-GM variety.3 The study did not show that GM soy is more allergenic than non-GM soy, but it did show that a GM food can unexpectedly cause an allergic reaction in a person who is not allergic to the food in its non-GM form.

The EU system for assessing GM foods for allergenicity

Under European law, GM foods must be assessed for their potential to cause allergies before they are allowed onto the market. Proponents claim that any potentially allergenic GM foods are likely to be caught by these regulatory checks. The GMO Compass website calls these assessments “rigorous” and adds, “If a GM plant is found to contain a potential allergen, its chances of receiving approval in the EU are slim to none.”1

But in reality, the European regulatory process, though stronger than the US process, has no rigorous system for assessing the allergenic potential of GM foods. This is largely because reliable scientific tests to predict allergenicity have not been developed.

The process that EU regulators use to assess the allergenicity of GM foods1,4 is based on a system proposed in 2001 by the Food and Agriculture Organization (FAO) of the United Nations and the World Health Organization (WHO). This system was actually designed by two GM industry-funded groups, the International Life Sciences Institute (ILSI) and the International Food Biotechnology Council (IFBC), as the FAO/WHO freely states.5

The process begins with a comparison of the protein that the GM plant is designed to produce with known allergenic proteins. Depending on the outcome of this initial assessment, further investigations can include:

  • Tests to see if the new protein reacts with the blood serum of sensitive individuals
  • Artificial stomach tests to see if the protein is broken down easily. If it is, it is thought unlikely to be an allergen
  • Animal feeding trials.1

Why the allergy assessment process is ineffective

The EU’s allergy assessment is unlikely to reliably predict whether a GM food will cause allergic reactions.

The most important reason is that the new protein that is assessed in the regulatory process is normally not the protein as expressed in the whole GM plant. Instead, it is what is known as a surrogate protein. This surrogate protein is isolated from sources such as GM E. coli bacteria or occasionally, a different plant species.6 This is scientifically unjustifiable because the protein can change as a result of the genetic engineering process and according to the organism within which it is expressed – as in the case of StarLink maize (Myth 3.7).

In other words, the same GM gene introduced into a GM plant and into E. coli bacteria can produce proteins that can have very different effects on the people and animals that eat them. Plants and bacteria process newly synthesized proteins in different ways. In particular, the GM plant protein will undergo a process known as “post-translational modification” and will thus possess added sugar molecules (“glycosylation”). So even though the amino acid sequences of the GM plant and GM E. coli proteins may be identical, their functions and allergenic potential can be quite different.

Other reasons why the allergenicity decision tree model is unsatisfactory include:

  • A comparison of the new protein in the GM food with the database of known allergens will not detect new allergens.
  • Blood serum tests are problematic because allergenic sensitization is an allergen-specific process. So unless the transgenic protein expressed in the GM food is already a common allergen, there is unlikely to be a single sensitized person in the world whose blood serum would react with it.2
  • Blood serum tests are not useful in detecting uncommon allergens – substances that few people are allergic to2 – since it is unlikely that serum samples will be taken from these few people.
  • A phenomenon known as cross-reactivity can make it difficult to identify from blood serum testing which specific protein out of several is the allergen.2
  • The artificial stomach tests carried out for regulatory purposes are performed under unrealistic conditions. Levels of acidity and digestive enzymes are much higher than would be present in the digestive systems of individuals who would consume the GM food. This makes it likely that the new GM protein will be broken down into fragments that are too small to be potent allergens. In real life, however, the levels of acidity and digestive enzymes in people’s stomachs vary according to age, health status, length of time since they ate their last meal, and other factors. One study found that under the standard conditions used in artificial stomach tests, one of the insecticidal proteins commonly present in GM Bt crops was broken down. But when the researchers adjusted the acidity and enzymes to more realistic levels, the insecticidal protein was highly resistant to being broken down. The authors called for regulatory tests to be carried out in “more physiologically relevant” conditions of lower acidity and enzyme levels.7

One review concluded that the allergenicity assessment might be useful in assessing GM foods containing a known allergenic protein, but that assessing proteins of unknown allergenicity is “more problematic” and “the predictive value of such an assessment is unknown”.2 Another review agreed that the standard tests were “not always conclusive”, especially when the organism from which the GM gene is taken has no history of dietary use or has unknown allergenicity.8

The current allergy assessment system is not reliable because it relies heavily on in vitro tests (laboratory experiments in non-living systems, such as the blood serum and artificial stomach tests). Unfortunately, however, an effective alternative does not yet exist. In vivo tests (tests on living organisms such as animals or humans) are useful for detecting nutritional or toxicological effects of foods, but no reliable animal testing methods have yet been established for allergenicity testing of foods.9,2,8,10

The main problem is that the immune systems of humans and animals are different, so it is difficult to predict human allergenicity from animal responses. The most reliable assessment of allergenicity of a GM food would be to test the food prior to commercialization, on large numbers (around 5,000) of human volunteers. The large numbers avoid false negative results (where an allergenic effect exists but is missed because too few subjects are used) and provide statistical power.

In the absence of pre-market tests on humans, at present the only reliable approach to assessing the allergenicity of GMOs would be post-commercialization monitoring. Consumers would have to be clearly informed when they ate the new GMO and would be asked to report any adverse effects to designated authorities.

Such post-commercialization assessments are not required in any country. In countries such as the US and Canada, where GM foods are not labelled, the likelihood that allergenicity would be linked to a GMO is extremely low, unless it caused acute allergic reactions in a large portion of the population.

“There is more than a casual association between GM foods and adverse health effects. There is causation as defined by Hill’s Criteria in the areas of strength of association, consistency, specificity, biological gradient, and biological plausibility.The strength of association and consistency between GM foods and disease is confirmed in several animal studies… Multiple animal studies show significant immune dysregulation, including upregulation of cytokines [protein molecules involved in immune responses] associated with asthma, allergy, and inflammation.”
– American Academy of Environmental Medicine11

Studies on GM foods confirm existing allergy assessments are inadequate

Studies on GM foods confirm that current allergy assessments are inadequate to detect new allergens created by the genetic engineering process.

In a study on mice fed GM peas containing an insecticidal protein from beans, mice showed antibody immune reactions and allergic-type inflammatory responses to the GM protein and chicken egg white protein when it was fed to them with the GM peas.12

The mice did not show antibody immune reactions and allergic-type inflammatory responses to beans that naturally contain the insecticidal protein or to egg white protein when it was fed with the natural insecticidal protein obtained from beans. They also did not have an immune response to the egg white protein when it was fed on its own.12

These outcomes show that the GM insecticidal protein made the mice more susceptible to developing allergic-type inflammatory reactions to foods eaten with the GM food. This is called immunological cross-priming.

The results indicated that the reaction of the mice to the GM peas was caused by changes brought about by the genetic engineering process. The normally non-immunogenic and non-allergenic insecticidal protein naturally produced in beans was altered in structure and/or function when engineered into peas, in particular in the addition to the protein of sugar molecules (glycosylation) via post-translational modification processes, becoming a potent immunogen (substance that produces an immune response) and allergen.12

This was not a regulatory test and tests such as this are not required to be carried out for the regulatory assessment of GM foods. The allergenicity of the GM peas would likely not have been spotted by the EU’s screening process because the natural, non-GM version of the bean insecticidal protein is not a known allergen. Because of this, blood serum from sensitized individuals would not have been available for regulatory serum tests.

Overall, the study shows that GM foods can contain new allergens and cause new allergic reactions – and that the GMO’s allergenicity is unlikely to be detected using the current allergy assessment process.

Other studies confirm the inadequacy of the current allergy assessment process:

  • A study on a commercialized GM insecticidal maize, MON810, showed that the GM plant’s proteins were markedly altered compared with those in the non-GM counterpart. Unexpected changes included the appearance of a new form of the protein zein, a known allergen, which was not present in the non-GM maize variety. A number of other proteins were present in both their natural forms and in truncated and lower molecular mass forms.13 These findings suggest major disruptions in gene structure and function in this GM crop. The EU’s allergy assessment failed to pick up these changes and failed to detect the presence of the newly created allergen.
  • A GM soy variety modified with a gene from Brazil nuts was found to be capable of producing an allergic reaction in people who are allergic to Brazil nuts. The researchers had genetically engineered the Brazil nut gene into the soy in order to increase its nutritional value. When they tested the effect of this GM soy on blood serum from people allergic to Brazil nuts, they found that the serum produced an allergic response to the soy. Through scratch tests on skin, they confirmed that people allergic to Brazil nuts were allergic to the modified soybean.14 This study is often cited by GM proponents as evidence of the effectiveness of regulatory processes in identifying allergenic foods before they reach the marketplace. But this is untrue. As with the GM peas study,12 this was not a regulatory test and tests such as this are not required to be carried out for the regulatory assessment of GM foods in any country.


The absence of reliable methods for allergenicity testing and the lack of rigour in current allergy assessments mean that it is impossible to reliably predict whether a GMO will prove to be allergenic. If a GMO did prove to be allergenic, it would be almost impossible to find out, as no post-commercialization monitoring is carried out anywhere in the world.


  1. GMO Compass. The allergy check. 2006. Available at:
  2. Bernstein JA, Bernstein IL, Bucchini L, et al. Clinical and laboratory investigation of allergy to genetically modified foods. Env Health Perspect. 2003;111:1114-21.
  3. Yum HY, Lee SY, Lee KE, Sohn MH, Kim KE. Genetically modified and wild soybeans: an immunologic comparison. Allergy Asthma Proc. 2005;26:210-6.
  4. European Food Safety Authority Panel on Genetically Modified Organisms (GMO). Guidance document for the risk assessment of genetically modified plants and derived food and feed. EFSA J. 2006;99:1–100.
  5. Food and Agriculture Organization (FAO) and World Health Organization. Decision tree approach to the evaluation of the allergenicity of genetically modified foods. In: Evaluation of Allergenicity of Genetically Modified Foods: Report of a Joint FAO/WHO Expert Consultation on Allergenicity of Foods Derived from Biotechnology, 22–25 January 2001. Rome, Italy: Food and Agriculture Organization of the United Nations (FAO); 2001:5–15; 25–27.
  6. Friends of the Earth. Could GM foods cause allergies? A critique of current allergenicity testing in the light of new research on transgenic peas. London, UK; 2006.
  7. Guimaraes V, Drumare MF, Lereclus D, et al. In vitro digestion of Cry1Ab proteins and analysis of the impact on their immunoreactivity. J Agric Food Chem. 2010;58:3222-31. doi:10.1021/jf903189j.
  8. Penninks AH, Knippels LM. Determination of protein allergenicity: studies in rats. Toxicol Lett. 2001;120:171-80.
  9. Pusztai A, Bardocz S, Ewen SWB. Genetically modified foods: Potential human health effects. In: D’Mello JPF, ed. Food Safety: Contaminants and Toxins. Wallingford, Oxon: CABI Publishing; 2003:347–372. Available at:
  10. Pusztai A. Genetically modified foods: Are they a risk to human/ animal health? Published June 2001.
  11. American Academy of Environmental Medicine. Genetically modified foods. 2009. Available at:
  12. Prescott VE, Campbell PM, Moore A, et al. Transgenic expression of bean alpha-amylase inhibitor in peas results in altered structure and immunogenicity. J Agric Food Chem. 2005;53:9023–30. doi:10.1021/jf050594v.
  13. 13.      Zolla L, Rinalducci S, Antonioli P, Righetti PG. Proteomics as a complementary tool for identifying unintended side effects occurring in transgenic maize seeds as a result of genetic modifications. J Proteome Res. 2008;7:1850-61. doi:10.1021/pr0705082.
  14. 14.      Nordlee JA, Taylor SL, Townsend JA, Thomas LA, Bush RK. Identification of a Brazil-nut allergen in transgenic soybeans. N Engl J Med. 1996;334:688-92. doi:10.1056/NEJM199603143341103.