Truth: GM foods are safety tested by the developer companies and regulation varies from non-existent to weak

Myth at a glance

Claims that GM foods are extensively tested and strictly regulated are false. At best, they are tested for safety by the companies that want to commercialize them. The tests are weak and inadequate to show safety.

GM foods were first allowed into the human food supply in the US, based on the claim that they are Generally Recognized as Safe (GRAS) – despite the fact that none has ever fulfilled the strict legal criteria that define GRAS status.

In many countries, GM foods are approved by regulators as “substantially equivalent” to non-GM crops, but when this assumption is tested scientifically, GM crops are often found to have unexpected and unintended differences.

Examples of regulatory failure are common and include unscientific procedures, sloppy practices, and the failure to recognize and address important types of risk. Regulatory lapses are often linked to conflicts of interest among regulators.

Industry and some government sources claim that GM foods are strictly regulated.1,2 But GM food regulatory systems worldwide vary from voluntary industry self-regulation (in the US) to weak (in Europe). None are adequate to protect consumers’ health. All rely on safety testing done by the GMO developer company that wishes to commercialize the genetically modified organism (GMO) in question.

As criticism has mounted of the deficiencies in GM food regulatory systems, the message from pro-GM lobbyists has shifted, from “GM foods are strictly regulated” to “GM foods are no more risky than non-GM foods, so why regulate them at all?” They point out that each time a plant breeder develops a new variety of apple or beetroot through conventional breeding, we do not demand that it be tested toxicologically, and there is no reason to think that GM foods will be any more toxic.

But this argument is spurious. Humans have co-evolved with their food crops over millennia and have learned by long – and doubtless sometimes bitter – experience which plants are toxic and which are safe to eat. There would have been casualties along the way, but the survivors would have learned from any mistakes and would only have developed their food crops from plants that were proven safe over many years of use.

With GM foods, we do not have the luxury of long periods of experimentation by our ancestors. And unlike our ancestors, we show no sign of learning from the mistakes of genetic engineering, since signs of toxicity in animal feeding experiments with GM foods are routinely dismissed (see Chapter 3).

How GMOs first entered world markets

“One thing that surprised us is that US regulators rely almost exclusively on information provided by the biotech crop developer, and those data are not published in journals or subjected to peer review… The picture that emerges from our study of US regulation of GM foods is a rubber-stamp ‘approval process’ designed to increase public confidence in, but not ensure the safety of, genetically engineered foods.” – David Schubert, professor and head, Cellular Neurobiology Laboratory, Salk Institute3,4

GM foods were first commercialized in the US in the early 1990s. The US Food and Drug Administration (FDA) allowed the first GM foods onto world markets in spite of its own scientists’ warnings that genetic engineering is different from conventional breeding and poses special risks, including the production of new toxins or allergens that are difficult to detect.5,6,7,8,9,10

PRIBYL1-1992-first-page-1200px Comments on the FDA’s policy to release GMOs into the food supply by FDA microbiologist Dr Louis Pribyl. Dr Pribyl castigates the FDA for the lack of a scientific basis to its GMO policy. This document is one of many that were released as a result of a lawsuit brought against the FDA by the Alliance for Bio-Integrity (

For example, FDA microbiologist Dr Louis Pribyl stated: “There is a profound difference between the types of unexpected effects from traditional breeding and genetic engineering”. He added that several aspects of genetic engineering “may be more hazardous”.10

Dr E. J. Matthews of the FDA’s Toxicology Group warned that “Genetically modified plants could … contain unexpected high concentrations of plant toxicants”.7

Gerald Guest, director of FDA’s Center for Veterinary Medicine (CVM), called for GM products to be demonstrated safe prior to marketing, on the grounds that “animal feeds derived from genetically modified plants present unique animal and food safety concerns.”6

FDA official Linda Kahl protested that the agency was “trying to fit a square peg into a round hole” by “trying to force an ultimate conclusion that there is no difference between foods modified by genetic engineering and foods modified by traditional breeding practices.” Kahl stated: “The processes of genetic engineering and traditional breeding are different, and according to the technical experts in the agency, they lead to different risks.”5

Several FDA scientists called for more rigorous scientific data to be presented by the companies before GMOs were released onto the market, and specifically for safety and toxicological testing.6,7,10

However, FDA administrators, who expressly admitted that the agency had been following a government agenda to “foster” the growth of the biotech industry,11 disregarded their scientists’ concerns, refused to regulate GM foods, and permitted them to enter the market without any testing or labelling.

The creation of this policy was overseen by the FDA’s deputy commissioner of policy, Michael Taylor, who was appointed to the post in 1991. Prior to joining the FDA, Taylor had been in private practice at King & Spalding, a law firm that represented the GM crop developer company Monsanto. In 1998 he became Monsanto’s vice president for public policy.12,13 By 2010 he was back at the FDA as its deputy commissioner for foods.14

Taylor’s career is often cited as an example of a type of conflict of interest known as the “revolving door”. The term describes the movement of personnel between roles as regulators and the industries affected by the regulation.

The US regulatory process for GMOs

Contrary to popular belief, the US FDA does not have a mandatory GM food safety assessment process and has never approved as safe any GM food that is currently on the market. It does not carry out or commission safety tests on GM foods. Instead, the FDA operates a voluntary pre-market review programme, in which it looks at whatever data the manufacturer chooses to provide.

Although all GM foods commercialized to date have gone through this lenient process, there is no legal requirement for them to do so. Companies are allowed to put any GMO on the market that they wish without even notifying the FDA. And even though they might theoretically be held liable for any resulting harm to consumers, it would be extremely difficult to prove such harm in court.

The outcome of the FDA’s voluntary assessment is not a conclusion, underwritten by the FDA, that the GMO is safe. Instead it consists of the FDA sending the company a letter stating that:

  • The company has provided the FDA with a summary of research that it has conducted assessing the GM crop’s safety
  • Based on the results of the research done by the company, the company has concluded that the GMO is safe
  • The FDA has no further questions
  • The company is responsible for placing only safe foods in the market
  • If a product is found to be unsafe, the company may be held liable.15

This process does not guarantee – or even attempt to scientifically investigate – the safety of GM foods. Therefore although it may protect the image of GM foods, it does not protect the public.

Letter from the US FDAs Center for Food Safety Letter from the US FDA’s Center for Food Safety and Applied Nutrition to Monsanto regarding its GM glyphosate-tolerant soybean. The letter confirms that the FDA is not liable if any safety concerns are identified with the soybean.

The US government is not impartial regarding GM crops and foods

The US government cannot be relied upon to regulate GMOs. It is not an impartial authority, given its aim to “foster” the growth of the biotechnology industry.11And not only is the US Department of Agriculture (USDA) influenced by that same policy, it even has financial interests in GM technology, owning 1.2% of all public-sector US agricultural biotechnology patents granted between 1982 and 2001.16

Through its embassies and agencies, the US government promotes GM crops globally and sometimes even pressures other governments to accept them. This was made clear by diplomatic cables disclosed by Wikileaks, which revealed that:

  • The US embassy in Paris recommended that the US government launch a retaliation strategy against the EU that “causes some pain” as punishment for Europe’s reluctance to adopt GM crops.17
  • The US embassy in Spain suggested that the US government and Spain draw up a joint strategy to help boost the development of GM crops in Europe.18
  • The US State Department is trying to steer African countries towards acceptance of GM crops.19,20

This strategy of exerting diplomatic pressure on national governments to adopt GM crops is undemocratic as it interferes with their ability to represent the wishes of their citizens. It is also inappropriate to use US taxpayers’ money to promote patented products owned by individual private companies to further the companies’ economic goals. A 2003 paper found that nearly three-quarters (74%) of agricultural biotechnology patents were privately owned.16

FDA presumes that GMOs are “generally recognized as safe”

The US FDA claims that GM foods can be marketed without prior testing or oversight because they are “generally recognized as safe” or GRAS.21

However, GM foods do not meet the GRAS criteria, which are strict. According to US statutory law and FDA regulations, a food that does not have a history of safe use prior to 1958 cannot qualify as GRAS unless it satisfies two requirements:

  • There must be an overwhelming expert consensus that it is safe; and
  • This consensus must be based on scientific evidence generated through “scientific procedures”, which “shall ordinarily be based upon published studies”.22

Because GM foods have never met either requirement, they cannot legally be classified as GRAS. At the time the FDA made its presumption that all GM foods are GRAS, there was not even expert consensus about their safety within the FDA (as attested by the statements of the agency’s scientists detailed above). The FDA’s biotechnology coordinator admitted there was no such consensus outside the agency either.23

Moreover, no such scientific consensus has emerged since then. For instance, in 2001 an expert panel of the Royal Society of Canada issued an extensive report declaring that it is “scientifically unjustifiable”to presume that GM foods are safe.24 Over the following years, many hundreds of experts have signed various formal statements declaring that the safety of GM foods has not been established and is subject to reasonable doubt. In 2013 nearly 300 scientists and experts signed a statement rejecting claims of a scientific consensus on GMO safety, either for human or animal consumption or for the environment.25

Even if there had been such a consensus, GM foods would still have failed to meet the GRAS standard because there has never been adequate technical evidence to establish that even one GM food is safe, especially because the law requires that the data must demonstrate a “reasonable certainty” the food will not be harmful.22

The sham of substantial equivalence

“The concept of substantial equivalence has never been properly defined; the degree of difference between a natural food and its GM alternative before its ‘substance’ ceases to be acceptably ‘equivalent’ is not defined anywhere, nor has an exact definition been agreed by legislators. It is exactly this vagueness that makes the concept useful to industry but unacceptable to the consumer… Substantial equivalence is a pseudo-scientific concept because it is a commercial and political judgment masquerading as if it were scientific. It is, moreover, inherently anti-scientific because it was created primarily to provide an excuse for not requiring biochemical or toxicological tests.”
– Erik Millstone, professor in science and technology policy, University of Sussex, UK, and colleagues26
“Substantial equivalence is a scam. People say that a potato has vaguely the same amount of protein and starch and stuff as all other potatoes, and therefore that it is substantially equivalent, but that is not a test of anything biological.”
– Professor C. Vyvyan Howard, medically qualified toxicopathologist, then at the University of Liverpool, in testimony to the Scottish Parliament Health and Community Care Committee27
“In one interpretation, to say that the new [GM] food is ‘substantially equivalent’ is to say that ‘on its face’ it is equivalent (i.e. it looks like a duck and it quacks like a duck, therefore we assume that it must be a duck – or at least we will treat it as a duck). Because ‘on its face’ the new food appears equivalent, there is no need to subject it to a full risk assessment to confirm our assumption. This interpretation of ‘substantial equivalence’ is directly analogous to the reasoning used in approval of varieties derived through conventional breeding. In both cases, ‘substantial equivalence’ does not function as a scientific basis for the application of a safety standard, but rather as a decision procedure for facilitating the passage of new products, GM and non-GM, through the regulatory process.”
– The Royal Society of Canada24

Worldwide, regulators approve GM foods as safe based on the concept of “substantial equivalence”. Substantial equivalence assumes that if a GMO contains similar amounts of a few basic components such as protein, fat, and carbohydrate as its non-GM counterpart, then the GMO is substantially equivalent to the non-GMO and no rigorous safety testing is required.

The concept of substantial equivalence as applied to GMOs was first put forward by the industry and the Organization for Economic Cooperation and Development (OECD), a body dedicated not to protecting public health but to facilitating international trade.28,29

Until recently there has been no legal or scientific definition of substantial equivalence. For example, it has not been established how different a GM crop is allowed to be in its constituents from the non-GM parent line, or how different it can be from other varieties of the crop, before it is deemed non-substantially equivalent and regulatory action is triggered.30 Such regulatory action could comprise a ban or a requirement for in-depth, long-term toxicological testing.

In 2013, after years of criticism over the lack of scientific definition of substantial equivalence, the EU instituted a regulation defining limits on the extent to which a GMO can differ from the non-GM comparator and still qualify as equivalent.31

Claims of substantial equivalence for GM foods have been widely criticized and revealed as scientifically inaccurate by independent researchers32,33,34,35 and by the Royal Society of Canada.24A useful analogy to help us understand what is meant by substantial equivalence is that of a BSE-infected cow and a healthy cow. They are substantially equivalent to one another, in that their chemical composition is the same. The only difference is in the shape of a protein (prion) that constitutes a minute proportion of the total mass of the cow. This difference that would not be picked up by current substantial equivalence assessments. Yet few would claim that eating a BSE-infected cow is as safe as eating a healthy cow.

When claims of substantial equivalence are tested, they are often found to be untrue. Using molecular analytical methods, GM crops have been shown to have a different composition to their non-GM counterparts. This is true even when the two crops are grown under the same conditions, at the same time and in the same location – meaning that the changes are not due to different environmental factors but to the genetic modification.

Examples include:

  • GM soy had 12–14% lower amounts of isoflavones, compounds that play a role in sex hormone metabolism, than non-GM soy.36
  • Experimental GM rice varieties had unintended major nutritional disturbances compared with non-GM counterparts, although they were grown side-by-side in the same conditions. The structure and texture of the GM rice grain was affected and its nutritional content and value were dramatically altered. The authors said that their findings provided “alarming information with regard to the nutritional value of transgenic rice” and showed that the GM rice was not substantially equivalent to non-GM.39
  • GM soy had 27% higher levels of a major allergen, trypsin-inhibitor, than the non-GM parent variety, despite the Monsanto authors’ claim that the GM soybean was “equivalent” to the non-GM soybean. In order to reach the conclusion of “equivalence”, the Monsanto authors compared plants grown at different locations and different times, increasing the range of variability with irrelevant data. Good scientific practice in a test of substantial equivalence requires the GM plant to be compared with the non-GM isogenic (with the same genetic background) variety, grown at the same time in the same conditions.37
  • Canola (oilseed rape) engineered to contain vitamin A in its oil had much reduced vitamin E and an altered oil-fat composition, compared with non-GM canola.38
  • Experimental GM insecticidal rice was found to contain higher levels of certain components than non-GM rice. Differences were caused by both genetic manipulation and environmental factors. However, differences in sucrose, mannitol, and glutamic acid were shown to have resulted specifically from the genetic manipulation.40
  • Commercialized MON810 GM maize had a markedly different profile in the types of proteins it contained compared with the non-GM counterpart when grown under the same conditions.35These unexpected compositional differences also showed that the MON810 maize was not substantially equivalent to the non-GM isogenic comparator, even though worldwide regulatory approvals of this maize had assumed that it was.41
  • Bt maize of the variety MON810 Ajeeb YG showed significant differences from its isogenic non-GM counterpart, with some values being outside the range recorded in the scientific literature. Some fatty acids and amino acids present in the non-GM maize were absent in the Bt maize. The researchers concluded that the genetic modification process had caused alterations in the maize that could result in toxicity to humans and animals.42

Altered nutritional value is of concern for two reasons: first, because it could directly affect the health of the human or animal consuming it by providing an excess or shortage of certain nutrients; and second, because it is an indicator that the genetic engineering process could have altered biochemical processes in the plant. This could signify that other unexpected changes have also occurred that might impact human or animal health, such as altered toxicity or allergenicity.

Indeed, the Bt maize MON810 Ajeeb YG and its non-GM counterpart, which were found to be compositionally different,42 were tested in a rat feeding study and the GM variety was found to cause organ toxicity.43,44

Different environmental conditions produce wide variations in protein expression

A comparison of GM maize MON810 and the isogenic non-GM parent variety grown in two different locations revealed a total of 32 different proteins that were expressed at significantly different levels in fresh leaf tissue from GM maize compared to non-GM. These proteins belonged to three main functional categories: (1) carbohydrate and energy metabolism, (2) genetic information processing, and (3) stress response.45

The differences were influenced by environmental conditions, since different proteins were expressed differentially in the two locations studied. The evidence also suggested that gene expression in non-GM maize was more stable, less influenced by environmental factors, than in GM maize.45

This study did not measure specific parameters related to food safety or environmental impact, but identified 32 differences in the expression of specific proteins in GM and non-GM maize plants.45 However, it would be informative to extend this study by carrying out additional research to assess whether health impacts of MON810 maize reported by other researchers44,46,47,48 might be linked to one or more of the protein (proteomic) changes observed in this study.

Herbicide residues in GM herbicide-tolerant crops mean they are not substantially equivalent to non-GM crops

Over 80% of GM crops worldwide are engineered to tolerate glyphosate herbicides. These GM crops are approved by regulators on the grounds that they are substantially equivalent to the non-GM parent crops. This assumption was tested in a comparative analysis of GM glyphosate-tolerant soy, non-GM soy cultivated under a conventional “chemical” regime, and non-GM soy grown organically. All crops tested were grown in Iowa, USA.49

The GM soy was found to contain high residues of glyphosate and its breakdown product AMPA. Conventional and organic soybeans contained neither of these chemicals.49

Organic soybeans showed the healthiest nutritional profile, with more sugars, such as glucose, fructose, sucrose and maltose, and significantly more protein and zinc and less fibre than conventional and GM soy. Organic soybeans also contained less total saturated fat and omega-6 fatty acids than conventional and GM soy.49

Using 35 different nutritional variables to characterise each soy sample, the researchers were able to discriminate GM, conventional and organic soybeans without exception.49

The study showed that GM glyphosate-tolerant soy is not substantially equivalent to non-GM soy, not only because of the herbicide residues in the GM soy, but because of the different nutritional profile.49

Europe’s comparative safety assessment: Substantial equivalence by another name

Europe has controversially adopted the concept of substantial equivalence in its GM food assessments – but under another name. The European Food Safety Authority (EFSA) does not use the discredited term “substantial equivalence” but has allowed industry to replace it with another term with the same meaning: “comparative assessment” or “comparative safety assessment”.

The story of how the comparative safety assessment made its way into Europe’s GMO regulatory system is, like the formation of the US FDA’s biotech policy, a tale of revolving doors and conflicts of interest with industry.

The change of name from “substantial equivalence” to “comparative safety assessment” was suggested in a 2003 paper on risk assessment of GM plants.50 The paper was co-authored by Harry Kuiper, then chair of EFSA’s GMO Panel, with Esther Kok. In 2010 Kok joined EFSA as an expert on GMO risk assessment.51 In their 2003 paper, Kuiper and Kok freely admitted that the concept of substantial equivalence remained unchanged and that the name change was in part meant to deflect the “controversy” that had grown up around the term.50

At the same time that Kuiper and Kok published their 2003 paper, they were part of a task force of the GMO industry-funded International Life Sciences Institute (ILSI), that was working on re-designing GMO risk assessment.29 In 2004 Kuiper and Kok co-authored an ILSI paper on the risk assessment of GM foods, which defines comparative safety assessment. The other co-authors include representatives from GM crop companies that sponsor ILSI, including Monsanto, Bayer, Dow, and Syngenta.52

EFSA has followed ILSI’s suggestion of treating the comparative safety assessment as the basis for GM safety assessments. EFSA has promoted the concept in its guidance documents on assessment of environmental risks of GM plants53 and of risks posed by food and feed derived from GM animals,54 as well as in a peer-reviewed paper on the safety assessment of GM plants, food and feed.55

In 2013 the EU Commission incorporated the industry- and EFSA-generated concept of the comparative safety assessment into its new regulation on GM food and feed.31 A major problem with the comparative safety assessment is that, as the name suggests, the authorities are beginning to treat it as a safety assessment in itself, rather than as just the first in a series of mandatory steps in the assessment process. In other words, EFSA and the EU Commission are moving towards a scenario in which if the GMO passes this weak test – and many have, in spite of having significant differences from the non-GM comparators – then they may not be subjected to further rigorous testing.

What is the comparative assessment?

This comparative assessment consists of a comparison of the newly developed GM variety with its closest non-GM relative, normally the parent variety. The non-GM relative has the same genetic background as the GMO, but without the genetic modification, so it is called the isogenic (genetically the same) variety.

A comparison is made of the composition of the GMO compared with the non-GM isogenic variety, with regard to the levels of certain basic components such as carbohydrate, protein, and fat. If they fall roughly within the same range, the GMO is deemed substantially equivalent to the non-GM isogenic variety. The effects of feeding the GMO and its non-GM isogenic variety to animals are also compared in a short animal feeding study.

The right and wrong way to do a comparative assessment

The proper scientific method of carrying out a comparative assessment is to grow the GM crop and its non-GM isogenic comparator side-by-side under the same conditions. This method ensures that any differences found in the GM crop, or in animals that eat it in a feeding trial, are understood to arise from the genetic modification and not from environmental factors such as different growing conditions. It also fulfills the intent of the EU Directive, which is to enable differences “arising from the genetic modification” to be identified and assessed.56

If differences are found between the GM crop and the correct comparator, this is a sign that the genetic engineering process has caused disruption in the structure and/or function of the native genes of the host plant. Further investigations should then be carried out to look for other unintended changes. These would include in-depth toxicological testing and “stress testing”, in which the crop is subjected to challenges in the laboratory that it might encounter in the field, such as exposure to crop diseases and simulated adverse weather conditions.

In contrast, comparisons with unrelated or distantly related varieties grown at different times and in different locations introduce and increase external variables and serve to mask rather than highlight the effects of the genetic engineering process. Such practices undermine the aim of the GMO comparative assessment, which is to identify any unintended disruption to gene structure and function – and consequent biochemical composition – brought about by the genetic engineering process.

This, however, is the method favoured by the GMO industry, both in the compositional analyses it performs on its products37,57 and in the animal feeding trials it carries out on its GMOs for regulatory authorizations. In these animal feeding trials, it compares the GMO diet not only with the non-GM isogenic comparator diet, but also with a range of “reference” diets containing varieties grown in different locations.58,59 The effect is to hide the effects of the genetic modification on the plant amid the “noise” created by the external variables.

GMOs would not pass an objective comparative safety assessment

Scientists and even the Royal Society of Canada have heavily criticized the use of substantial equivalence and the comparative safety assessment as the basis of safety assessments of GM crops.4,24,26,60

Yet if the comparative safety assessment were applied objectively and systematically with proper controls, most GMOs would not pass even this weak test of safety. This is because as explained above (“The sham of substantial equivalence”), many studies on GM crops show that they are not substantially equivalent to the non-GM counterparts from which they are derived. There are often significant differences in the levels of certain nutrients and types of proteins, which could impact allergenicity, toxicity, or nutritional value.

The GMO industry and its supporters have sidestepped this problem by widening the range of comparison. Adopting a method used by Monsanto in analyses of its GM soy,37,57 they no longer restrict the comparator to the GM plant and the genetically similar (isogenic) non-GM line, grown side-by-side under the same conditions and at the same time. Instead they use as comparators a range of non-isogenic varieties grown at different times and in different locations.

In some cases the spurious comparators are modern varieties that have been recently grown and analyzed, but in other cases they are historical varieties on which data has been gathered in the literature. Some of this “historical” data even dates back to before World War II.60 It may have been analyzed by different researchers using methods that vary in sensitivity, accuracy and reliability. Anyone familiar with the fundamental principles of the experimental sciences will recognize that comparisons to such data are not meaningful.

Despite the loose approach taken in these comparative assessments, they often reveal significant differences in composition between the GMO and the diverse comparator dataset used by the company applying for approval of the GMO. This reveals that the properties of the GMO are outside the range of the non-GMO comparator data, including even the historical data. But even in these extreme cases, according to scientists who have served in regulatory bodies, the differences are dismissed as not being “biologically relevant”.60

The ILSI database

The industry-funded International Life Sciences Institute (ILSI) has created a database of crop varieties,61 including historical or unusual varieties that have untypically high or low levels of certain components. It appears that the primary purpose of this database is to provide “comparative data” that allow industry to argue that the constituents of their GMOs are within the normal range of variation, regardless of how deviant they are from the norm and from the appropriate comparator, which is the relevant non-GM isogenic line grown in the same conditions. EFSA experts use this industry database as the basis of the compositional comparison in GMO risk assessments.29

If, on the basis of this “comparative safety assessment”, EFSA experts judge the GM crop to be equivalent to the comparator non-GM crops, it is assumed to be as safe.29,62 Further rigorous tests that could reveal unexpected differences, such as long-term animal feeding trials and environmental stress tests, are not required.29 Instead, a limited check is carried out.

EFSA disregards advice of its own head of GMO risk assessment

Joe Perry, the chair of EFSA’s GMO Panel, has admitted that the ILSI database cannot be relied upon for risk assessment purposes. Perry said: “At the present time we can’t trust the ILSI database. There is not sufficient environmental information from where these trials were done and that’s why we insist that the commercial reference variety should be planted simultaneously with the GM and the non-GM. Otherwise I think we are in an unsafe situation and I would worry that the limits would be too wide.”63

Although Perry’s statement implies that comparison with the isogenic line is EFSA policy, this does not seem to be the case, since EFSA used the ILSI database as the basis of the risk assessment of SmartStax, a stacked trait GM maize to which eight genetically modified genes had been added.64 Moreover, EFSA did not confine its comparison of the GM maize with a commercial reference variety “planted simultaneously”, as Perry said EFSA requires. Instead, EFSA compared one of the parent GM maize varieties used to develop the stacked trait crop and its non-GM isogenic parent variety grown in “various field trials” in “different field trial locations”, on two different continents, and at different times.64 This is antithetical to good scientific practice, which tests one variable at a time.

In spite of all the “noise” introduced by these irrelevant data, statistically significant differences in composition were still found between the parent GM maize and the non-GM comparator. But EFSA dismissed these differences on the basis that the values fell within the “natural variation” found in unspecified “literature” and in the ILSI database. EFSA was able to conclude that the GM stacked maize was “equivalent” to existing “commercial maize varieties”,64 with the result that it was not considered necessary to perform further detailed risk assessment on this stacked trait crop.

EFSA weakens the comparative assessment by widening the range of comparators

An EU Directive of 2001 was strict in stipulating that the comparator against which the GMO should be assessed for safety should be the non-GM genetically similar (isogenic) parent – “the non-modified organism from which it is derived”.56 The non-GM isogenic parent would have the same genetic background as the GM crop, but without the GM transformation. This would enable differences “arising from the genetic modification” to be identified and assessed, without the confounding factor of different environmental conditions in which the crops are grown.

In line with this Directive, the EU Regulation of 2003 on GM food and feed stipulated that the comparator against which the GMO should be assessed for safety should be the non-GM “conventional counterpart”.65

Until 2011 EFSA followed the principle of using the correct comparator in its Guidance documents and Opinions. But in a Guidance document published in late 2011,66 EFSA legitimized unscientific industry practice by widening the range of acceptable comparators beyond the non-GM isogenic comparator. In doing so, EFSA arguably departed from EU legislative requirements.65,56

EFSA even proposed to allow the use of other GM crops, rather than the usual non-GM isogenic line, as comparators for stacked trait crops containing multiple GM traits. And remarkably, EFSA stipulated that in some cases, plants from different species could be accepted as comparators.66

EFSA’s approach is in line with industry’s practices.But whether it complies with EU regulation is questionable.

The result of this lax regulatory process is that almost any GMO could pass through the regulatory process unchallenged. This forces consumers and farmers into the role of experimental guinea pigs. Any unexpected effects of a GMO that has entered the market via this channel will only be revealed post-commercialization, in the form of ill effects on humans or animals that eat the GMO, or poor crop performance in farmers’ fields.

Industry-backed lobbying to weaken the criteria for comparative assessment

There has been intense lobbying pressure on regulators to allow a wider range of comparison for GMOs beyond the non-GM isogenic variety. As part of this drive, some scientists have published papers in scientific journals explaining away significant alterations in a GM plant compared with the non-GM isogenic comparator by widening the range of comparison and recommending this practice to regulators. They compare the GM plant not only with the non-GM parent plant from which it is derived, but with a wide range of different varieties of the plant. Two examples of such papers follow.

1. Catchpole and colleagues (2005)

This study evaluated levels of certain metabolites (breakdown products) in GM potatoes and compared these levels not only with levels in the non-GM parent lines but also with levels in other non-GM potato varieties. The authors found significant differences in the levels of one metabolite, rhamnose, in a GM potato variety, as compared with the levels in the non-GM isogenic parent line grown in the same conditions. But they believed that this was not important because the GM variety had rhamnose levels that were “typical of potato cultivars”.67

The authors were explicit about the lobbying purpose of their study: “The cultivar-based compositional heterogeneity [differences] we describe emphasizes the importance of comparison with a range of equivalent cultivars and not solely the parental line.”67 They were recommending widening the range of comparison used in the comparative assessment of GM crops to a range of different varieties. This effectively hides the significant difference between a GM crop and its non-GM isogenic control.

The authors also emphasized the conclusion that regulators were supposed to reach: that the GM potatoes were “substantially equivalent to traditional cultivars”.67

2. Ricroch and colleagues (2011)

This review of safety assessment methods for GM crops68 took the same approach as Catchpole and colleagues (above). Ricroch and colleagues disagreed with the principle of EU Directive 2001/18, which states that the non-GM isogenic line should be used as the comparator for the risk assessment.56 They argued that the natural range of variation of certain components in different non-GM lines was greater than the variation between the GM and the isogenic non-GM parent line.

Also, the authors argued that compared with any differences brought about by the genetic engineering of a crop compared with the isogenic non-GM parent line, different “environmental conditions usually have a larger impact”.68

This is entirely our point – environmental conditions create large differences in plants. But the aim of the comparative assessment in EU regulatory practice is to exclude differences caused by environmental conditions, so that any differences caused by the GM process (“arising from the genetic modification”, as EU Directive 2001/18 states) can be identified.56 The differences caused by different environmental conditions are confounders, or confusing elements, in this process. With that in mind, the proper comparator for the GMO is the non-GM isogenic variety, grown side-by-side under the same conditions.

The lobbying point made by Ricroch and colleagues is the same as that of Catchpole and colleagues: “These observations indicate that the current regulatory burden on GE crops should be lowered… the time may have come to simplify the risk assessment of modern biotechnology products, and therefore reduce cost.” Like Catchpole and colleagues, Ricroch and colleagues affirmed the “validity” of the concept of substantial equivalence – the basis for the non-regulation of GM crops by the US government.68

Both sets of authors did not want the GMO to be compared with the non-GM isogenic counterpart, arguably because of the significant differences that are generally found. Instead they wanted to compare it with a range of other non-GM plants – masking the differences in the GM plant compared with the non-GM isogenic variety – amid the “noise” created by irrelevant data on a wide range of varieties grown in a range of conditions.

In summary, these authors are in conflict with the spirit and letter of EU legislation as well as scientific rigour. What they are recommending is the equivalent in chemicals risk assessment of:

  1. Carrying out a toxicological experiment that finds that a certain chemical causes a certain type of cancer in 40% of the test group of animals as compared with control rates of 0-5%, then…
  2. Dismissing the significance of the finding on the grounds that in a certain town where carcinogenic chemicals are manufactured, 40% of the population has this type of cancer, and…
  3. Concluding that the cancer incidence in this experiment is within the natural range of variation and that therefore the chemical is safe.

Such a conclusion would rightly be derided. But it is no different in principle from invoking the “natural range of variation” to conclude that GM crops are safe.

Comparative assessment does not directly assess safety

Comparative assessment or assessment of substantial equivalence measures the composition of the GMO and of some comparators and on that basis comes to a conclusion on whether the GMO is significantly different from the comparators. This compositional analysis says nothing directly about the safety of a GMO for human or animal consumption or about its potential impacts on the environment.

Even if the comparative assessment were correctly carried out using the isogenic non-GMO variety as comparator, it still would not be able to establish the safety or otherwise of a GM crop. It can only find what the researcher is looking for. It cannot find unexpected toxins or allergens or changes in nutrients that may have been caused by the GM process. The only way to look for such unexpected changes is long-term toxicological and nutritional testing in animals. Such testing screens broadly for harmful impacts of consuming the GMO. When such tests are carried out on GM foods, as discussed in Chapter 3, they often expose problems with the GM food tested against the non-GM food.

The comparative assessment also cannot predict the responses of a GM crop to environmental stresses. Such responses can only be known by testing the GMO under different environmental stress conditions. Similarly, it is not possible to predict environmental impacts of the GMO from a comparative assessment, and these too must be tested.

Such tests should be carried out in controlled, enclosed conditions in order to prevent the introduction of the GMO into the wider environment until evidence is obtained that it is stable and safe.

Masking effects of a GM diet

In parallel with the trend of widening the range of comparators used in the comparative assessment of GMOs, industry and regulators have adopted an equally unscientific approach to assessing the health effects of a GMO in animal feeding trials. When, as is often the case, a feeding trial reveals statistically significant differences between the animals fed a GM diet, as compared with those fed a non-GM diet, these changes are often dismissed as being “not biologically meaningful” or as being within the range of normal biological variation (see Chapter 3 for a discussion of this practice and how it places public health at risk).

These practices run counter to good scientific method and seem to be part of a strategy for masking the effects of the GMO by introducing into the data analysis additional data from other experiments, often carried out under different and non-comparable conditions. This greatly widens the apparent “natural range of variation” to the point where the results for the GMO fall within this artificially widened range. This generates a convenient answer – that the GMO is no different from non-GM comparators – but in no way assures safety for the consumer or the environment.

Proof of equivalence not required in Europe until 2013

Before 2013, the degree of similarity that a GMO needed to have to its non-GMO counterpart in order to pass the comparative assessment was never defined. Previous to that time, all GMOs were approved without any objective criteria for similarity or dissimilarity being in place. A regulation passed in 2013 changed this situation and demands proof of equivalence within defined limits.31 However, the GMOs commercialized previous to this date have not been subject to this requirement and the regulation will not be retrospectively applied to them.

Regulatory process is based on industry studies

Many governments, including those of the EU, Japan, Australia, and New Zealand, have an agency that reviews GM crops before commercialization. Some agencies make a judgment regarding the safety of those crops for consumption and the environment. Others, for example, the US FDA, make no such judgment. In Europe, the relevant agency is the European Food Safety Authority (EFSA) and the final decision to approve or reject the GMO is made by a vote by representatives of the governments of the member states. In Australia and New Zealand, the agency is FSANZ.

Worldwide, safety assessments of GMOs by government regulatory agencies are not scientifically rigorous. Nowhere in the world do the relevant agencies carry out or commission their own safety tests prior to GMO commercialization. Instead, they make decisions regarding the safety of the GMO based on studies commissioned and controlled by the very same companies that stand to profit from the crop’s approval.

The problem with this system is that industry studies have an inbuilt bias. Published reviews that evaluate studies assessing the safety and hazards of risky products or technologies have shown that industry-sponsored studies, or studies where authors are affiliated with industry, are much more likely to reach a favourable conclusion about the safety of the product than studies carried out by scientists independent of industry.

The most notorious example is industry studies on tobacco, which succeeded in delaying regulation for decades by sowing confusion about the health effects of smoking and passive smoking.69,70 A similar bias has been found in industry studies on other products, including pharmaceuticals,71,72 medical products,73 and mobile phone technology.74

The GMO field is no exception. A review of scientific studies on the health risks of GM crops and foods showed that either financial or professional conflict of interest (author affiliation to industry) was strongly associated with study outcomes that cast GM products in a favourable light.75

Grey literature and lack of transparency

Lack of transparency of industry data is a major problem with the GMO regulatory process. The animal feeding and other safety studies that companies submit to regulatory agencies are often unpublished at the time the GMO is approved. This means that they are not available for scrutiny by the public or independent scientists. Unpublished studies fall into the category of so-called “grey literature” and are of unknown reliability.

Such grey literature stands in stark contrast with the standard quality control method traditionally used by the scientific community: peer-reviewed publication. The peer-reviewed publication process is far from perfect and is subject to biases of various kinds. Yet it is still the best method that scientists have come up with to ensure reliability. Its strength lies in a multi-step quality control process:

  • The editor of the journal reads the study. If he judges it as potentially acceptable for publication in the journal, he sends it to qualified scientists (“peers”) to evaluate. They give feedback, including any suggested revisions, which are passed on to the authors of the study.
  • Based on the outcome of the peer review process, the editor publishes the study, rejects it, or offers to publish it with revisions by the authors.
  • Once the study is published, it can be scrutinized and repeated (replicated) or extended by other scientists. Replication is the cornerstone of scientific reliability, because if other scientists were to do the same experiment but come up with different findings, this could challenge the findings of the original study.

In the US, significant portions of the industry data on GMOs submitted to regulators are classified as confidential business information and are shielded from public scrutiny.76

The lack of access to industry studies has resulted in the public being deceived over the safety of GMOs. For example, in Europe, industry’s raw data on Monsanto’s GM Bt maize variety MON863 (approved for food and feed use in the EU in 2005) were only forced into the open through court action by Greenpeace. Scientists at the independent research organization CRIIGEN in France analyzed the raw data and found that Monsanto’s own feeding trial on rats revealed signs of liver and kidney toxicity that had been kept hidden from the public.77,78

Since this case and perhaps as a result of it, transparency has improved in Europe and the public can obtain industry toxicology studies on GMOs from EFSA on request, along with other safety data submitted by the developer company. Only a small amount of information, such as the genetic sequence of the GMO, can be kept commercially confidential.79

However, the problem of the lack of transparency of industry data in Europe is far from solved. In 2013 EFSA published the full Monsanto dossier of data on the GM maize NK603 as part of its transparency initiative80 after the safety of the maize was cast into doubt by a study carried out by the team of Professor Gilles-Eric Séralini at the University of Caen, France.81Monsanto responded by threatening legal action against EFSA for publishing its data.82(The French study was subsequently retracted by Food and Chemical Toxicology, the journal that published it, in highly questionable circumstances: see Chapter 3.)

Moreover, industry safety data on pesticides is still kept secret under commercial confidentiality agreements between industry and regulators.83 This is relevant to GMO safety because most GM crops are engineered to tolerate being sprayed with herbicide (herbicides are technically pesticides): that is, they can absorb the herbicide and survive.

Therefore GM crops are likely to contain higher specific pesticide residues.49 Yet the public cannot see the studies that form the basis of pesticide approvals. In Europe, all that is accessible to the public is the report on the industry studies drawn up by the authorities of the “rapporteur” member state, responsible for liaising between industry and the EU authorities for the application for authorization of that particular pesticide.83

This secrecy was challenged in a 2012 court case brought by Pesticide Action Network Europe and Greenpeace Netherlands to force disclosure of the industry studies on glyphosate. Astonishingly, however, the German court prioritized commercial interests over public health and ruled that the studies must remain secret.84

GMO assessment turns its back on science

The medical biologist and immunologist Dr Frédéric Jacquemart, president of the independent scientific research group Inf’OGM and a member of France’s High Council for Biotechnology, analyzed EFSA’s risk assessment of Monsanto’s insecticidal GM maize MON810 as an example of the unscientific nature of GMO evaluations. Unscientific practices used by Monsanto in its dossier and accepted by EFSA (and other regulators around the world) include:

  • Assuming that the Bt toxin protein expressed by the GM maize is the same as, and as safe as, the natural Bt protein, when in fact the protein in GM maize is a hybrid and truncated protein with different biological and toxicological properties.
  • Introducing irrelevant comparison data, from experiments carried out on a range of crops grown in a wide range of conditions, into studies on a GM crop. This has the effect of masking differences between the GM crop and the corresponding non-GM crop that were caused by the GM process and allows a false conclusion of equivalence to be drawn between the two.
  • Accepting claims of equivalence between the GMO and the non-GM comparator even though equivalence has not been proven. The tests performed by industry have historically not been capable of proving equivalence. A European regulation passed in 2013 addresses this problem by setting criteria for equivalence and non-equivalence,31 but this has not been applied to MON810.
  • Allowing industry to select which data it presents in order to reach the desired conclusion, without requiring industry to disclose all of the studies it has carried out or the criteria it used in selecting the data submitted.
  • Failing to require a power analysis in animal toxicological feeding studies. The power analysis ensures that the experiment uses the appropriate number of animals to enable the researchers to detect the effect that is being looked for. If a study finds no effect from the GM diet, without a power analysis that demonstrates that a sufficient number of animals was used, one cannot determine whether the negative result was because there truly was no effect or whether the study used too few animals to detect the effect.

The report notes that stating that nothing of concern was seen in a study is only valid “if one looks”, and points out that the evaluations done as part of the regulatory process create the appearance of having looked, but are “designed not to find anything”. The report concludes that while evaluations of GMOs are “passed off as rigorous studies, directly based on data”, in fact they are “a parody of science, aimed at political decision-makers and the public”.85

Industry and the US government design the GMO regulatory process worldwide

Agricultural biotechnology corporations have lobbied long and hard on every continent to ensure that the weak safety assessment models developed in the US are the norm globally. Working through the US government or groups that appear to be independent of the GMO industry, they have provided biosafety workshops and training courses to smaller countries that are attempting to grapple with regulatory issues surrounding GMOs. The result has been models for safety assessment that favour easy approval of GMOs without rigorous assessment of health or environmental risks.

For example, a report by the African Centre for Biosafety (ACB) described how the Syngenta Foundation, a nonprofit organization set up by the agricultural biotechnology corporation Syngenta, worked on “a three-year project for capacity building in biosafety in sub-Saharan Africa”. The Syngenta Foundation’s partner in this enterprise was the Forum for Agricultural Research in Africa (FARA), a group headed by people with ties to Monsanto and the US government.

The ACB identified the Syngenta Foundation/FARA project as part of an “Africa-wide harmonization of biosafety policies and procedures” that would “create an enabling environment for the proliferation of GMOs on the continent, with few biosafety checks and balances”.86

In India, the US Department of Agriculture led a “capacity building project on biosafety” to train state officials in the “efficient management of field trials of GM crops”87 – the first step towards full-scale commercialization. And in 2010, a scandal erupted when a report from India’s supposedly independent national science academies recommending release of GM Bt brinjal (eggplant/aubergine) for cultivation was found to contain 60 lines of text copy-pasted almost word for word from a biotechnology advocacy newsletter – which itself contained lines extracted from a GM industry-supported publication.88

Regulatory failures around the world

There is a constant stream of revelations about the lack of competence, objectivity, and transparency of GMO regulatory bodies around the world. Individuals who sit on GMO regulatory bodies are frequently found to have conflicts of interest in the form of professional or financial affiliation with the GMO industry or ownership of patents on GMO technology. A few examples of this compromised regulatory system follow.

India: “public sector” GM Bt cotton infected with Monsanto’s gene

A taxpayer-funded project of the Indian Council of Agricultural Research (ICAR) to commercialize a “public sector” variety of GM Bt cotton came to an ignominious end when the crop was found to carry a Monsanto-patented GM gene. The crop also failed in the field and was withdrawn.89

An inquiry revealed that the developers of the Bt cotton variety had submitted three different maps of the inserted GM gene unit sequence to different authorities. The maps showed that even the developers of the GMO did not understand its genetic makeup.90

It was also reported that the scientist responsible for conducting the molecular analysis of the GM Bt cotton variety, Ishwarappa S. Katageri from the University of Agricultural Sciences in Dharwad, did not do so because he did not have the technical skills to carry out such studies and was not even aware of any methodology to differentiate various events.90

Such tests are mandatory for regulatory assessments in India. Yet the regulators, the Genetic Engineering Appraisal Committee (GEAC) and the Review Committee on Genetic Manipulation (RCGM), seemingly did not notice these lapses. Indeed, Katageri had sat on the RCGM as a regulator for years.90

India: Regulatory breakdown left GM Bt cotton farmers vulnerable

In 2012, faced with conflicting reports of the performance and prospects of GM crops in India, an expert committee of the Indian Parliament was tasked with looking into the matter. The committee was especially concerned to investigate reports of an escalation in farmer suicides since the introduction of GM Bt cotton. Critics of GM crops in India have linked the suicides to failure of the Bt cotton crop and farmer indebtedness resulting from high seed costs.

After gathering evidence from all stakeholders, the committee visited villages in the cotton growing belt of Vidarbha in the state of Maharashtra to interview Bt cotton farmers. In spite of strenuous efforts by the Maharashtra state government to divert them elsewhere,91 the committee visited a Monsanto showcase village. According to a previously published article in The Times of India authored by a journalist on a Monsanto-sponsored field trip, thanks to Bt cotton, “not a single person” in this village had committed suicide.92

But the visiting committee members talked to farmers in the Monsanto model village and heard a very different account, according to an article in The Hindu by the award-winning journalist P. Sainath. The farmers said there had been 14 suicides in the village, most of them since Bt cotton was introduced. Many of the remaining farmers had given up farming altogether or switched to soybeans.91

In their final report, the committee noted that while seed companies had benefited from selling Bt cotton, “The poor and hapless farmers have received more of the costs than the benefits”. They concluded that there are better options than GM crops for increasing food production and demanded a ban even on GM crop field trials.93

It is reasonable to ask why, if this assessment is true, so many farmers in India adopted Bt cotton. The committee’s report addressed this question and partly blamed the “craze” for cultivating Bt cotton because of its “perceived advantages”, leading to a situation where traditional non-GM seeds had been “almost wiped out”.93

The “craze” interpretation is backed up by a peer-reviewed study by the anthropologist Glenn Davis Stone, who is not an opponent of GM Bt cotton. Stone concluded that seed “fads” were responsible for the widespread adoption of Bt cotton, helped along by “agricultural deskilling” and aggressive marketing campaigns by seed companies.94

According to the Indian Parliament expert committee, the other part of the answer lies in the failure of the government regulatory bodies, which should protect the interests of the public and farmers. The committee noted “with concern the grossly inadequate and antiquated regulatory mechanism for assessment and approval” of GM crops; the “serious conflict of interest of various stakeholders involved in the regulatory mechanism”; and “the total lack of post commercialization, monitoring and surveillance”.95

Worldwide: Lack of regulation of a new type of GMO based on gene-silencing technology

In 2013 a peer-reviewed study was published by Professor Jack Heinemann and colleagues suggesting that government GMO regulators are failing to consider important risks of a new type of GM plants and related technologies.96

While most existing GM plants are designed to make new proteins, these new-type GM plants and products are designed to make a form of genetic information called double-stranded RNA (dsRNA). The dsRNA molecules are short (21-23 base unit) gene function regulatory molecules which are designed to alter the way genes are expressed – by silencing or activating them. This process of gene expression alteration is broadly called RNA interference (RNAi) and is at the basis of post-transcriptional gene silencing (PTGS) in plants.

A number of GMOs have been made using dsRNA gene-silencing technology. Australia’s public research institute CSIRO has developed GM wheat and barley varieties where genes have been silenced in order to change the type of starch made by the plant in its grain. Another example is biopesticide plants, which produce a dsRNA molecule designed to silence a gene in insects that eat the plant. The insect eats the plant, and the dsRNA in the plant survives digestion in the insect and travels into the insect’s tissues to silence a gene. The insect dies as a result.96

Gene silencing may be inherited across generations through epigenetic mechanisms in plants and some kinds of animals that are exposed to gene-silencing dsRNA.96

Furthermore, there is massive ongoing investment to develop products that directly transfer dsRNA into the living cells of plants, animals and microbes via their food or by being absorbed through their “skin”.  This allows dsRNA molecules to be sprayed onto fields of crops to kill insects or weeds, or to be delivered as oral medicine to bees.97

Heinemann and colleagues reviewed their experience with three government safety regulators (for either food or the environment) in three different countries over the past ten years. They found that the safety of dsRNA molecules was usually not considered at all. If it was considered, the regulator simply assumed that any dsRNA molecules were safe, rather than requiring evidence that they were safe.96

The authors found that government regulators:

  • Dismissed the need for any assessment of the sequence of the base unit nucleotides in the dsRNAs produced by GM plants
  • Seemed to assume that dsRNAs produced by these plants are much the same as the more fragile single-stranded RNAs (for example, mRNA), and therefore would not survive cooking and digestion
  • Claimed that these new dsRNA molecules are safe because humans and non-target animals would not be exposed to them.96

On the basis of these assumptions, the regulators did not assess whether the gene regulatory dsRNAs could cause adverse effects by, for example, silencing or activating genes in people or animals that come into contact with the plant when it is grown commercially. Contact could include eating the crop or processed products derived from it, inhaling dust from the crop when harvesting it, or inhaling flour from the crop when baking with it. And regulators made that decision regardless of whether the dsRNA was generated intentionally or unintentionally by the crop. All three regulators decided that there were no risks to be considered, based on assumptions, rather than scientific evidence.96

The problem is that all these assumptions are incorrect, as shown by many scientific studies reviewed by Heinemann and colleagues.96

For example, a study by Zhang and colleagues showed that short dsRNA gene regulatory molecules produced in non-GM plants can be taken up into the bodies of people who eat the plant. The dsRNA from the plant was found circulating in blood, indicating that it survives cooking and digestion. Research has also shown that:

  • At least one dsRNA produced in plants (called MIR168a) can change the expression of genes in mice when the dsRNA is taken up through eating
  • One type of dsRNA (MIR168a) can change the expression of a gene in human cells growing in tissue culture.98

Another study found a wide range of RNA molecules from many different organisms, including bacteria and fungi as well as other species, in human plasma (a component of blood). The authors concluded that these RNA molecules may be able to influence cellular activities and may thus affect human health.99

According to Heinemann and colleagues, these studies show that there is a real risk that novel short dsRNA gene regulatory molecules produced by the new GM crops could survive digestion in people and change how those people’s genes are expressed. Therefore regulators should not ignore the specific risks posed by novel dsRNAs in GM foods.96

As a result of their analysis, Heinemann and colleagues developed and provided a safety testing procedure for all GM plants that may produce new dsRNA molecules, as well as for products where the active ingredient is dsRNA.96

Since the publication of the paper by Heinemann and colleagues, two more have appeared on the topic of dsRNA uptake through food. The first, by Witwer and colleagues,100 studied dsRNA uptake into primates. The concentration of dietary dsRNA was just at the detection limit, creating uncertainty about how common these molecules are. Therefore the authors encouraged more studies. That should be concerning to regulators, who for years have assumed that dsRNAs could not survive digestion. The new work further justified calls for testing of foods created using RNAi-based technology to confirm the safety of novel dsRNA molecules.

Witwer and colleagues reported poor reproducibility of detection, which they said suggested low levels of dsRNAs.100 Indeed, it is to be expected that dietary dsRNAs will be present at low levels. However, the paper did not address relevant risk assessment issues, such as:

  • Which concentration of dsRNA in blood (or other tissues) matters?
  • Which exposure routes (diet, inhalation, contact) matter most?

Witwer and colleagues used different animals and food sources than other investigators, their study had only two animals, and only a very small number (five) of potential dsRNAs were targeted. So while the authors concluded that effects were unlikely, they also carefully acknowledged that their study was too small and the strength of their positive detections too strong to exclude uptake of dsRNA into mammals through food.100

The second paper was written by employees of Monsanto and another company that produces dsRNA-containing products (Dickinson and colleagues, 2013).101 Dickinson and colleagues extended Monsanto’s previous study,102 but failed to find dsRNA of plant origin in mice fed the plants in question.101 An editorial in the journal Nature Biotechnology claimed that Monsanto’s new study facilitated “the process of self-correction” in the literature,103 effectively implying that Zhang and colleagues’ study98 was wrong.

However, the methodology of the second Monsanto study was severely criticized by some of the original authors of Zhang and colleagues’ study.104 Moreover, on the basis of the evidence in the study by Zhang and colleagues (2012)98 and the second Monsanto study by Dickinson and colleagues (2013),101 it is not possible to say either that Zhang and colleagues or the Monsanto authors are wrong. Different groups of researchers working on different groups of animals, using different methodologies and looking for different dsRNA molecules, may reach different conclusions. Both may be correct, or either or both may be wrong.

More importantly, there have been many more successful detections of short dsRNA gene regulatory molecules of plant origin in mammals than there have been failures to detect them, as recorded in a study by Monsanto102 and in the patent literature.105

Unexpected effects from gene-silencing technology

A study in honeybees revealed unexpected ecological risks of dsRNA molecules. The study found that the expression of nearly 1,400 of the bees’ genes was altered in response to a certain type of dsRNA administered in their food – representing around 10% of known honeybee genes. The findings were a surprise, since this particular dsRNA had been used as a control in honeybee experiments because its gene sequence does not exist in honeybees and thus it was not expected to trigger RNAi responses in the bees.106 Another demonstration, this time in humans, was published by Hanning and colleagues. They attempted to predict which genes would be silenced in human cells based on full knowledge of the sequence of the dsRNAs they were using –and failed. They concluded that information-based modeling tools (known as bioinformatics) are insufficient to predict the effects of dsRNAs without specific biological testing.107

Viral Gene VI

A paper published in 2012 by scientists at the European Food Safety Authority (EFSA) revealed that the most common genetic regulatory sequence in commercialized GMOs also encodes a significant fragment of a viral gene.108 Yet this viral gene, called Gene VI, has been missed in regulatory assessments worldwide, including by EFSA. Regulators failed to identify the gene, to investigate whether it is expressed, and to assess any risks it may pose to human and animal health.

The EFSA researchers discovered that of the 86 different GMOs commercialized to date in the United States, 54 contain portions of Gene VI. They include any with the widely used gene regulatory sequence called the CaMV 35S promoter (from the cauliflower mosaic virus, CaMV).108

Among the affected GMOs are some of the most widely grown, including Roundup Ready soybeans, NK603 maize, and MON810 maize.

The EFSA researchers did a computer search of Gene VI DNA sequences to see if there were any similarities to known toxins and found “no significant hits”. In fact they did find a similarity between parts of Gene VI and a known allergen, suggesting that it is a “potential allergen”. But the authors went on to conclude that Gene VI was probably not an allergen, based on database searches against known allergens.108

However, the databases that the EFSA authors used include the Allergy Research and Resource Program database (FARRP) at The objectivity of this database is questionable on the grounds that its staff and facilities at the University of Nebraska are funded by the six major biotech companies: Monsanto, Syngenta, Dow, Dupont Pioneer, Bayer, and BASF.110

More importantly, databases of allergens only contain information on known allergens. They are not useful for identifying unknown allergens, which would be missed in a computer search such as that which the EFSA authors carried out. And as there are no meaningful animal models for assessing the allergenicity of foods or isolated proteins, hitherto unknown allergens could only be revealed by extensive testing on humans.

Also, Gene VI could express differently, depending on the genetic context in the host plant into which it is inserted. Thus no conclusions of safety can be drawn from the computer exercise.

The EFSA researchers did, however, conclude that the presence of segments of Gene VI “might result in unintended phenotypic changes”108 – changes in the plant’s observable characteristics or traits. Such changes could include the creation of proteins that are toxic or allergenic to humans. The segments of Gene VI could also trigger changes in the plants themselves that could compromise their performance in the field.

The protein produced by Gene VI is known to be toxic to plants.111 Gene VI is also known to interfere with the basic mechanism of protein synthesis,112 which is common to humans, animals, and plants, and to disrupt RNA silencing – a biological mechanism shared by humans, animals, and plants. Thus it is reasonable to ask whether the protein produced by Gene VI could be toxic to humans. This question can only be answered by further experiments.

Jonathan Latham, a crop geneticist and plant virologist, and Allison Wilson, a molecular biologist and geneticist, commented that viral genes expressed in plants raise both agronomic and human health concerns because many viral genes function to disable their host in order to facilitate pathogen invasion. They concluded, “The data clearly indicate a potential for significant harm,” and recommended that all GM crops containing Gene VI should be recalled. These include numerous commercial GMOs containing a promoter from the figwort mosaic virus (FMV), which were not considered by the EFSA researchers.113

After Latham and Wilson’s article drew the EFSA researchers’ paper to public attention, EFSA published a statement defending its risk assessment of GMOs. But EFSA’s response was misleading. It stated, “The viral gene (Gene VI) belongs to a plant virus (cauliflower mosaic virus) that cannot infect animals or humans”.114

This seems to miss the point of the concerns raised. As Latham and Wilson pointed out in response, Gene VI as found in GM crops is not the same as the natural cauliflower mosaic virus found in vegetables: “Depending on the specifics of its genome integration into commercial GMOs, Gene VI DNA may produce either a simple viral protein fragment or a chimeric (part-viral) protein. In either case the result will not be equivalent in structure, cellular location, or quantity, to any protein produced by the virus.”115 Therefore the safety of Gene VI as found in GMOs cannot be deduced from the qualities or known behaviour of the natural cauliflower mosaic virus.

Safety questions about Gene VI could be answered by analyzing GM crops with CaMV-driven cassettes to see if they express Gene VI and produce a protein product containing it. If Gene VI is expressed, then more in-depth studies should be carried out to investigate the consequences to the plant and the animals and humans that eat it.

GM salmon

The biotechnology aquaculture company AquaBounty has developed a GM salmon called the AquAdvantage®. The GM salmon is intended to grow faster and reach the market more quickly than natural salmon.

Dr Michael Hansen, senior scientist with the Consumers Union, examined116 the US Food and Drug Administration’s (FDA) assessment of the company’s data on the AquAdvantage salmon.117

Hansen found that the company data, though “woefully incomplete”, raised concerns that the GM salmon could be more allergenic than non-GM salmon. The study used groups of fish that were far too small to enable reliable conclusions to be drawn – only six GM fish were used. Despite the small sample sizes, tests with blood serum from humans who were allergic to salmon still showed a highly statistically significant increase (52%) in allergenic potency of one type of GM salmon (“diploid”), compared with non-GM controls. This means that the process of genetic engineering led to an increase in allergenic potency, at least in this test.116

A smaller increase (20%) in allergenic potency was found in the second type (“triploid”) of GM salmon. These are the salmon that will be commercialized and eaten by consumers. The FDA stated that the increase was not statistically significant. However, lack of statistical significance could have been due to the very small sample sizes. Hansen believed the FDA should have demanded that the test be repeated with larger sample sizes.116

Instead the FDA stated that there were not enough data to enable it to draw a conclusion on the allergenicity of the diploid GM salmon and that the triploid salmon posed “no additional risk” compared with non-GM salmon.

Hansen found the FDA’s assessment of the allergenicity data “inadequate” and concluded that there was cause for concern that the salmon “may pose an increased risk of severe, even life-threatening allergic reactions to sensitive individuals”.116

Hansen highlighted other questionable practices by the FDA, such as reportedly manipulating data on levels of IGF-1, a growth hormone that is linked with cancer, which was found at an average of 40% higher levels in the GM fish compared with controls. The data manipulation, as reported by Hansen, enabled the FDA to conclude that there was no significant difference between the IGF-1 levels for the GM and non-GM salmon.116

The FDA even reached a conclusion about growth hormone levels in the salmon flesh, despite having no data at all on growth hormone levels, due to the use of insensitive test methodology. In addition, the FDA allowed the company to select fish for inclusion in studies without specifying that they were chosen randomly.116

The FDA also allowed the company to cull out deformed fish prior to selecting fish to include in the studies, on the grounds that it is standard practice in the industry.117 This may be true, but it is not acceptable scientific practice in a study that is supposed to be designed to examine the effects of genetic modification in salmon. Even the FDA admitted that the culling “may have skewed the population” of fish studied,117 but it failed to draw the only scientifically valid conclusion, which is to reject the results as insufficient and require additional more rigorous research.

Hansen concluded that the FDA’s assessment of the company data was an example of “sloppy science”.116


The regulatory regime for GM crops and foods is weakest in the US, the origin of most such crops, but is inadequate in most regions of the world, including Europe. The US assumes that GM foods are “generally recognized as safe” (GRAS), even though they do not meet the legal definition of GRAS. Worldwide, regulators assume that GM crops are safe if certain basic constituents of the GM crop are “substantially equivalent” to those of their non-GM counterparts – a term that has not been legally or scientifically defined. The European regime applies the same concept but terms it “comparative safety assessment”.

Often, however, when an in-depth scientific comparison of a GM crop and its non-GM counterpart is undertaken, the assumption of substantial equivalence is shown to be false, as unexpected differences are found.

Today, no regulatory regime anywhere in the world requires long-term or rigorous safety testing of GM crops and foods. Regulatory assessments are based on data provided by the company that is applying to commercialize the crop – the same company that will profit from a positive assessment of its safety.

The regulatory procedure for GM crops is not independent or objective. The GM crop industry, notably through the industry-funded group, the International Life Sciences Institute (ILSI), has heavily influenced the way in which its products are assessed for safety. ILSI has successfully promoted concepts such as the comparative safety assessment, which maximize the chances of a GMO avoiding rigorous safety testing and greatly reduce industry’s costs for GMO authorizations.

Examples of regulatory failure are common and include unscientific procedures, sloppy practices, and the failure to recognize and address important types of risk. Regulatory lapses are often linked to conflicts of interest among regulators.


  1. European Commission. GMOs in a nutshell. 2009. Available at:
  2. Monsanto. Commonly asked questions about the food safety of GMOs. 2013. Available at:
  3. Tokar B. Deficiencies in federal regulatory oversight of genetically engineered crops. Institute for Social Ecology Biotechnology Project; 2006. Available at:
  4. Freese W, Schubert D. Safety testing and regulation of genetically engineered foods. Biotechnol Genet Eng Rev. 2004:299-324.
  5. Kahl L. Memorandum to Dr James Maryanski, FDA biotechnology coordinator, about the Federal Register document, “Statement of policy: Foods from genetically modified plants.” US Food & Drug Administration; 1992. Available at:
  6. Guest GB. Memorandum to Dr James Maryanski, biotechnology coordinator: Regulation of transgenic plants – FDA Draft Federal Register Notice on Food Biotechnology. US Department of Health & Human Services; 1992. Available at:
  7. Matthews EJ. Memorandum to toxicology section of the Biotechnology Working Group: “Safety of whole food plants transformed by technology methods.” US Food & Drug Administration; 1991. Available at:
  8. Shibko SL. Memorandum to James H. Maryanski, biotechnology coordinator, CFSAN: Revision of toxicology section of the “Statement of policy: Foods derived from genetically modified plants.” US Food & Drug Administration; 1992. Available at:
  9. Pribyl LJ. Comments on the March 18, 1992 version of the Biotechnology Document. US Food & Drug Administration; 1992. Available at:
  10. Pribyl LJ. Comments on Biotechnology Draft Document, 2/27/92. US Food & Drug Administration; 1992. Available at:
  11. Sudduth MA. Genetically engineered foods – fears and facts: An interview with FDA’s Jim Maryanski. FDA Consum. 1993:11–14.
  12. Bittman M. Why aren’t GMO foods labeled? New York Times. Published February 15, 2011.
  13. Nestle M. Food Politics: How the Food Industry Influences Nutrition and Health. Revised 15 October 2007. University of California Press; 2002.
  14. US Food and Drug Administration (FDA). Meet Michael R. Taylor, JD, deputy commissioner for foods. 2013. Available at:
  15. US Food and Drug Administration (FDA). Biotechnology consultation agency response letter BNF No. 000001. 1995. Available at:
  16. Graff GD, Cullen SE, Bradford KJ, Zilberman D, Bennett AB. The public-private structure of intellectual property ownership in agricultural biotechnology. Nat Biotechnol. 2003;21:989-95. doi:10.1038/nbt0903-989.
  17. Vidal J. WikiLeaks: US targets EU over GM crops. The Guardian. Published January 3, 2011.
  18. US lobbied EU to back GM crops: WikiLeaks. Published January 4, 2011.
  19. EINNEWS. Wikileaks document pushes genetically modified food for African countries. Published December 1, 2010.
  20. Laskawy T. Wikileaks: State Dept wants intel on African acceptance of GMOs. Grist. Published November 30, 2010.
  21. US Food and Drug Administration (FDA). Statement of policy: Foods derived from new plant varieties. FDA Fed Regist. 1992;57(104):22984.
  22. US Food and Drug Administration (FDA). CFR – Code of Federal Regulations Title 21, Volume 3 (Revised 1 April 2013): 21CFR170.30. 2013. Available at:
  23. Maryanski J. Letter from Dr James Maryanski, Biotechnology Coordinator, to Dr Bill Murray, Chairman of the Food Directorate, Canada. Subject: the safety assessment of foods and food ingredients developed through new biotechnology. 1991. Available at:
  24. Royal Society of Canada. Elements of precaution: Recommendations for the regulation of food biotechnology in Canada. An expert panel report on the future of food biotechnology. 2001. Available at:
  25. European Network of Scientists for Social and Environmental Responsibility (ENSSER). Statement: No scientific consensus on GMO safety. 2013. Available at:
  26. Millstone E, Brunner E, Mayer S. Beyond “substantial equivalence.” Nature. 1999;401:525–6. doi:10.1038/44006.
  27. Howard CV. GM crops inquiry: Testimony of Prof C. Vyvyan Howard to the Scottish Parliament Health and Community Care Committee, meeting no. 31, 27 November 2002. 2002. Available at:
  28. Organisation for Economic Cooperation and Development (OECD). Safety evaluation of foods derived by modern biotechnology: Concepts and principles. OECD Publishing; 1993. Available at:
  29. Then C, Bauer-Panskus A. European Food Safety Authority: A playing field for the biotech industry. Testbiotech; 2010. Available at:
  30. Levidow L, Murphy J, Carr S. Recasting “substantial equivalence”: Transatlantic governance of GM food. Sci Technol Hum Values. 2007;32:26–64.
  31. European Parliament and Council. Commission implementing regulation (EU) no. 503/2013 of 3 April 2013 on applications for authorisation of genetically modified food and feed in accordance with Regulation (EC) No 1829/2003 of the European Parliament and of the Council and amending Commission Regulations (EC) No 641/2004 and (EC) No 1981/2006. Off J Eur Union. 2013. Available at:
  32. 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:
  33. Nodari RO, Guerra MP. Implications of transgenics for environmental and agricultural sustainability. Hist Cienc Saude Manguinhos. 2000;7(2):481-91.
  34. Zdunczyk Z. In vivo experiments on the safety evaluation of GM components of feeds and foods. J Anim Feed Sci. 2001;10:195-210.
  35. 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.
  36. Lappé M, Bailey B, Childress C, Setchell KDR. Alterations in clinically important phytoestrogens in genetically modified herbicide-tolerant soybean. J Med Food. 1999;1:241–245.
  37. Padgette SR, Taylor NB, Nida DL, et al. The composition of glyphosate-tolerant soybean seeds is equivalent to that of conventional soybeans. J Nutr. 1996;126:702-16. [Supplementary information from Puerto Rico trials, excluded from the main publication but deposited with the American Society for Information Science, National Auxiliary Publication Service (NAPS) and in the archives of the Journal of Nutrition.]
  38. Shewmaker C, Sheehy JA, Daley M, Colburn S, Ke DY. Seed-specific overexpression of phytoene synthase: Increase in carotenoids and other metabolic effects. Plant J. 1999;20:401–412X.
  39. Jiao Z, Si XX, Li GK, Zhang ZM, Xu XP. Unintended compositional changes in transgenic rice seeds (Oryza sativa L.) studied by spectral and chromatographic analysis coupled with chemometrics methods. J Agric Food Chem. 2010;58:1746-54. doi:10.1021/jf902676y.
  40. Zhou J, Ma C, Xu H, et al. Metabolic profiling of transgenic rice with cryIAc and sck genes: an evaluation of unintended effects at metabolic level by using GC-FID and GC-MS. J Chromatogr B Anal Technol Biomed Life Sci. 2009;877:725-32. doi:10.1016/j.jchromb.2009.01.040.
  41. European Food Safety Authority (EFSA) GMO Panel. Opinion of the scientific panel on genetically modified organisms on a request from the Commission related to the notification (reference C/DE/02/9) for the placing on the market of insect-protected genetically modified maize MON 863 and MON 863 x MON 810, for import and processing, under Part C of Directive 2001/18/EC from Monsanto. EFSA J. 2004;2004:1-25.
  42. Abdo EM, Barbary OM, Shaltout OE. Chemical analysis of Bt corn “Mon-810: Ajeeb-YG®” and its counterpart non-Bt corn “Ajeeb.” IOSR J Appl Chem. 2013;4(1):55–60.
  43. Gab-Alla AA, El-Shamei ZS, Shatta AA, Moussa EA, Rayan AM. Morphological and biochemical changes in male rats fed on genetically modified corn (Ajeeb YG). J Am Sci. 2012;8(9):1117–1123.
  44. El-Shamei ZS, Gab-Alla AA, Shatta AA, Moussa EA, Rayan AM. Histopathological changes in some organs of male rats fed on genetically modified corn (Ajeeb YG). J Am Sci. 2012;8(10):684–696.
  45. Agapito-Tenfen SZ, Guerra MP, Wikmark O-G, Nodari RO. Comparative proteomic analysis of genetically modified maize grown under different agroecosystems conditions in Brazil. Proteome Sci. 2013;11(1):46. doi:10.1186/1477-5956-11-46.
  46. De Vendomois JS, Roullier F, Cellier D, Séralini GE. A comparison of the effects of three GM corn varieties on mammalian health. Int J Biol Sci. 2009;5:706–26.
  47. Gu J, Krogdahl A, Sissener NH, et al. Effects of oral Bt-maize (MON810) exposure on growth and health parameters in normal and sensitised Atlantic salmon, Salmo salar L. Br J Nutr. 2013;109:1408-23. doi:10.1017/S000711451200325X.
  48. Finamore A, Roselli M, Britti S, et al. Intestinal and peripheral immune response to MON810 maize ingestion in weaning and old mice. J Agric Food Chem. 2008;56:11533–39. doi:10.1021/jf802059w.
  49. Bøhn T, Cuhra M, Traavik T, Sanden M, Fagan J, Primicerio R. Compositional differences in soybeans on the market: glyphosate accumulates in Roundup Ready GM soybeans. Food Chem. 2013. doi:10.1016/j.foodchem.2013.12.054.
  50. Kok EJ, Kuiper HA. Comparative safety assessment for biotech crops. Trends Biotechnol. 2003;21:439–444.
  51. European Food Safety Authority (EFSA). Annual declaration of interests – Esther Kok. 2010.
  52. International Life Sciences Institute (ILSI). Nutritional and safety assessments of foods and feeds nutritionally improved through biotechnology, prepared by a task force of the ILSI International Food Biotechnology Committee. Compr Rev Food Sci Food Saf. 2004;3:38–104.
  53. European Food Safety Authority (EFSA) GMO Panel. Guidance on the environmental risk assessment of genetically modified plants. EFSA J. 2010;8:1879–1990. doi:10.2903/j.efsa.2010.1879.
  54. European Food Safety Authority (EFSA). Guidance on the risk assessment of food and feed from genetically modified animals and on animal health and welfare aspects. EFSA J. 2012;10:2501. [43 pp.].
  55. European Food Safety Authority (EFSA) GMO Panel Working Group on Animal Feeding Trials. Safety and nutritional assessment of GM plants and derived food and feed: The role of animal feeding trials. Food Chem Toxicol. 2008;46:S2-70. doi:10.1016/j.fct.2008.02.008.
  56. European Parliament and Council. Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC. Off J Eur Communities. 2001:1–38.
  57. Taylor NB, Fuchs RL, MacDonald J, Shariff AR, Padgette SR. Compositional analysis of glyphosate-tolerant soybeans treated with glyphosate. J Agric Food Chem. 1999;47:4469-73.
  58. Hammond B, Dudek R, Lemen J, Nemeth M. Results of a 13 week safety assurance study with rats fed grain from glyphosate tolerant corn. Food Chem Toxicol. 2004;42:1003-14. doi:10.1016/j.fct.2004.02.013.
  59. Hammond B, Lemen J, Dudek R, et al. Results of a 90-day safety assurance study with rats fed grain from corn rootworm-protected corn. Food Chem Toxicol. 2006;44:147-60. doi:10.1016/j.fct.2005.06.008.
  60. Hilbeck A, Meier M, Römbke J, Jänsch S, Teichmann H, Tappeser B. Environmental risk assessment of genetically modified plants – concepts and controversies. Environ Sci Eur. 2011;23. doi:10.1186/2190-4715-23-13.
  61. International Life Sciences Institute (ILSI). ILSI crop composition database: Version 4. 2011. Available at:
  62. European Food Safety Authority (EFSA). Guidance on the submission of applications for authorisation of genetically modified food and feed and genetically modified plants for food or feed uses under Regulation (EC) No 1829/2003. EFSA J. 2011;9:1–27. doi:10.2903/j.efsa.2011.2311.
  63. Perry J. Comment by Joe Perry, chair of EFSA’s GMO Panel, at EFSA’s consultative workshop on its draft guidance for the selection of Genetically Modified (GM) plant comparators, Brussels, 31 March 2011. 2011. Available at:
  64. European Food Safety Authority (EFSA) Panel on Genetically Modified Organisms (GMO). Scientific Opinion on application (EFSA-GMO-CZ-2008-62) for the placing on the market of insect resistant and herbicide tolerant genetically modified maize MON 89034 x 1507 x MON 88017 x 59122 and all sub-combinations of the individual events as present in its segregating progeny, for food and feed uses, import and processing under Regulation (EC) No 1829/2003 from Dow AgroSciences and Monsanto. EFSA J. 2010;8(9):1–37.
  65. European Parliament and Council. Regulation (EC) No 1829/2003 of the European Parliament and of the Council of 22 September 2003 on genetically modified food and feed. Off J Eur Union. 2003;268:1–23.
  66. European Food Safety Authority (EFSA) Panel on Genetically Modified Organisms (GMO). Guidance document on selection of comparators for the risk assessment of GM plants: Adopted on 14 April 2011. EFSA J. 2011;9:1–20.
  67. Catchpole GS, Beckmann M, Enot DP, et al. Hierarchical metabolomics demonstrates substantial compositional similarity between genetically modified and conventional potato crops. Proc Natl Acad Sci USA. 2005;102:14458-62. doi:10.1073/pnas.0503955102.
  68. Ricroch AE, Berge JB, Kuntz M. Evaluation of genetically engineered crops using transcriptomic, proteomic, and metabolomic profiling techniques. Plant Physiol. 2011;155:1752-61. doi:10.1104/pp.111.173609.
  69. Michaels D. Doubt is Their Product: How Industry’s Assault on Science Threatens Your Health. Oxford University Press; 2008.
  70. Barnes DE, Bero LA. Why review articles on the health effects of passive smoking reach different conclusions. JAMA. 1998;279:1566-70.
  71. Lexchin J, Bero LA, Djulbegovic B, Clark O. Pharmaceutical industry sponsorship and research outcome and quality: systematic review. Br Med J. 2003;326:1167. doi:10.1136/bmj.326.7400.1167.
  72. Lexchin J. Those who have the gold make the evidence: How the pharmaceutical industry biases the outcomes of clinical trials of medications. Sci Eng Ethics. 2011. doi:10.1007/s11948-011-9265-3.
  73. Bekelman JE, Li Y, Gross CP. Scope and impact of financial conflicts of interest in biomedical research: a systematic review. JAMA. 2003;289:454-65.
  74. Huss A, Egger M, Hug K, Huweiler-Müntener K, Röösli M. Source of funding and results of studies of health effects of mobile phone use: Systematic review of experimental studies. Environ Health Perspect. 2007;115:1–4.
  75. Diels J, Cunha M, Manaia C, Sabugosa-Madeira B, Silva M. Association of financial or professional conflict of interest to research outcomes on health risks or nutritional assessment studies of genetically modified products. Food Policy. 2011;36:197–203.
  76. Waltz E. Under wraps – Are the crop industry’s strong-arm tactics and close-fisted attitude to sharing seeds holding back independent research and undermining public acceptance of transgenic crops? Nat Biotechnol. 2009;27(10):880–882. doi:10.1038/nbt1009-880.
  77. Séralini GE, Cellier D, Spiroux de Vendomois J. New analysis of a rat feeding study with a genetically modified maize reveals signs of hepatorenal toxicity. Arch Environ Contam Toxicol. 2007;52:596–602.
  78. CRIIGEN. Revelations on the toxicity of GMOs – CRIIGEN reveals serious anomalies observed in rats fed on GMOs. 2005. Available at:
  79. Dalli J. GMOs: Towards a better, more informed decision-making process. Published March 17, 2011.
  80. European Food Safety Authority (EFSA). EFSA promotes public access to data in transparency initiative [press release]. 2013. Available at:
  81. Séralini GE, Clair E, Mesnage R, et al. [RETRACTED:] Long term toxicity of a Roundup herbicide and a Roundup-tolerant genetically modified maize. Food Chem Toxicol. 2012;50:4221-4231.
  82. Starling S. Monsanto threatens to sue EFSA over publication of maize GM data. Food Navigator. Published March 8, 2013.
  83. Antoniou M, Habib MEM, Howard CV, et al. Teratogenic effects of glyphosate-based herbicides: Divergence of regulatory decisions from scientific evidence. J Env Anal Toxicol. 2012;S4:006. doi:10.4172/2161-0525.S4-006.
  84. Administrative Court of Braunschweig. Pesticide Action Network Europe and Greenpeace Netherlands vs the Federal Republic of Germany, represented by Bundesamt für Verbraucherschutz und Lebensmittelsicherheit (BVL): 2A1033/12. 2012.
  85. Jacquemart F. GMO expertise: Assessment turns its back on science. Inf’OGM; 2012. Available at:
  86. Swanby H. Ongoing concerns about harmonisation of biosafety regulations in Africa. Melville, South Africa: African Centre for Biosafety; 2009. Available at:
  87. Ministry of Environment and Forests (India). Genetic Engineering Approval Committee (GEAC) and Cartagena Protocol on Biosafety (CPB). 2010. Available at:
  88. Mudur GS. Experts admit GM brinjal report fault. The Telegraph (India). Published September 26, 2010.
  89. Jishnu L. Untangling India’s Bt cotton fraud. Down To Earth. Published February 15, 2012.
  90. Jishnu L. ICAR’s shoddy science. Down to Earth. Published January 15, 2013.
  91. Sainath P. Reaping gold through cotton and newsprint. The Hindu. Published June 16, 2012.
  92. Shrivastav S. Reaping gold through Bt cotton. Times of India. Published August 28, 2011.
  93. Ministry of Agriculture Committee on Agriculture 2011-2012 (India): Fifteenth Lok Sabha. Cultivation of genetically modified food crops – Prospects and effects: 37th Report. New Delhi, India; 2012.
  94. Stone GD. Agricultural deskilling and the spread of genetically modified cotton in Warangal. Curr Anthropol. 2007;48.
  95. Lok Sabha Secretariat Committee on Agriculture (2011–12). Thirty-seventh report of the committee on cultivation of genetically modified food crops – Prospects and effects [press release]. 2012. Available at:
  96. Heinemann J, Agapito-Tenfen SZ, Carman J. A comparative evaluation of the regulation of GM crops or products containing dsRNA and suggested improvements to risk assessments. Environ Int. 2013;55:43–55.
  97. Carman J, Heinemann J, Agapito-Tenfen S. A briefing document for non-specialists describing the lack of regulation of a new class of products and GM crops based on dsRNA technology. Published March 21, 2013.
  98. Zhang L, Hou D, Chen X, et al. Exogenous plant MIR168a specifically targets mammalian LDLRAP1: Evidence of cross-kingdom regulation by microRNA. Cell Res. 2012;22(1):107-126. doi:10.1038/cr.2011.158.
  99. Wang K, Li H, Yuan Y, et al. The complex exogenous RNA spectra in human plasma: An interface with human gut biota? PLoS ONE. 2012;7(12):e51009. doi:10.1371/journal.pone.0051009.
  100. Witwer KW, McAlexander MA, Queen SE, Adams RJ. Real-time quantitative PCR and droplet digital PCR for plant miRNAs in mammalian blood provide little evidence for general uptake of dietary miRNAs: Limited evidence for general uptake of dietary plant xenomiRs. RNA Biol. 2013;10. Available at:
  101. Dickinson B, Zhang Y, Petrick JS, Heck G, Ivashuta S, Marshall WS. Lack of detectable oral bioavailability of plant microRNAs after feeding in mice. Nat Biotechnol. 2013;31(11):965-967. doi:10.1038/nbt.2737.
  102. Zhang Y, Wiggins BE, Lawrence C, Petrick J, Ivashuta S, Heck G. Analysis of plant-derived miRNAs in animal small RNA datasets. BMC Genomics. 2012;13:381. doi:10.1186/1471-2164-13-381.
  103. The editors of Nature Biotechnology. Receptive to replication. Nat Biotechnol. 2013;31(11):943-943. doi:10.1038/nbt.2748.
  104. Chen X, Zen K, Zhang C-Y. Reply to Lack of detectable oral bioavailability of plant microRNAs after feeding in mice. Nat Biotechnol. 2013;31(11):967-969. doi:10.1038/nbt.2741.
  105. Lam E. Edible transgenic plants as oral delivery vehicles for RNA-based therapeutics. 2012. Available at:
  106. Nunes F, Aleixo A, Barchuk A, Bomtorin A, Grozinger C, Simões Z. Non-target effects of Green Fluorescent Protein (GFP)-derived double-stranded RNA (dsRNA-GFP) used in honey bee RNA interference (RNAi) assays. Insects. 2013;4(1):90-103. doi:10.3390/insects4010090.
  107. Hanning JE, Saini HK, Murray MJ, et al. Lack of correlation between predicted and actual off-target effects of short-interfering RNAs targeting the human papillomavirus type 16 E7 oncogene. Br J Cancer. 2013;108(2):450-460. doi:10.1038/bjc.2012.564.
  108. Podevin N, du Jardin P. Possible consequences of the overlap between the CaMV 35S promoter regions in plant transformation vectors used and the viral gene VI in transgenic plants. GM Crops Food. 2012;3:296–300. doi:10.4161/gmcr.21406.
  109. University of Nebraska-Lincoln. About AllergenOnline. 2010. Available at:
  110. Shetterly C. The bad seed: The health risks of genetically modified corn. Elle. 2013. Available at:
  111. Takahashi H, Shimamato K, Ehara Y. Cauliflower mosaic virus gene VI causes growth suppression, development of necrotic spots and expression of defence-related genes in transgenic tobacco plants. Mol Gen Genet. 1989;216:188–194.
  112. Park HS, Himmelbach A, Browning KS, Hohn T, Ryabova LA. A plant viral “reinitiation” factor interacts with the host translational machinery. Cell. 2001;106:723-33.
  113. Latham J, Wilson A. Regulators discover a hidden viral gene in commercial GMO crops. Indep Sci News. 2013. Available at:
  114. European Food Safety Authority (EFSA). FAQ on inserted fragment of viral gene in GM plants. 2013. Available at:
  115. Latham J, Wilson A. Is the hidden viral gene safe? GMO regulators fail to convince. Indep Sci News. 2013. Available at:
  116. Hansen M. Comments of Consumers Union on genetically engineered salmon, Food and Drug Administration docket no. FDA-201034-N-001, Veterinary Medicine Advisory Committee Meeting. 2010. Available at:
  117. US Food and Drug Administration (FDA) Veterinary Medicine Advisory Committee. Briefing packet: AquAdvantage salmon. 2010. Available at: