Three reviews of GM safety

May 2012

GMO maize testing. Photo by I, Yann [GFDL (http://www.gnu.org/copyleft/fdl.html), CC-BY-SA-3.0 
(http://creativecommons.org/licenses/by-sa/3.0/or CC-BY-SA-2.5-2.0-1.0
(http://creativecommons.org/licenses/by-sa/2.5-2.0-1.0)], via Wikimedia Commons

Three reviews of the published science on the 'safety' or 'health impact' of GM plants appeared in 2011.

One of these was the third periodic examination of the scientific literature by Spanish scientist, José Domingo. His reviews have all focused specifically on the “potential adverse health/toxic effects” of GM food plants.

His first attempt in 2000, found the numbers of citations “very scant” (there wasn't really anything for him to review). His second attempt in 2007 still found the evidence too “limited” to demonstrate safety.

The most recent of Domingo's reviews found a “dramatic increase” in citations, and that there is now “an equilibrium” between the numbers suggesting safety and those “still raising serious concerns”. It was noted that most of the studies had been conducted by the biotech industry (a factor recognised as biassing the findings). He concluded that the GM debate is still controversial and “remains completely open at all levels”.

The second review was carried out by a French team, lead by Gilles-Eric Séralini. It examined 19 published rodent feeding studies on GM maize and soya. These experiments used feed derived from plants genetically transformed to tolerate or to produce pesticides, and which are already in our food chain.

The evidence was examined from a statistical and biological point of view, and included a re-analysis of the raw data from standard 90-day rat feeding trials.

It was found that convergent data indicated liver and kidney problems as the end points of a GM diet. (The liver and kidney are the major organs which deal with toxins in the body)

Séralini's team's conclusions were that the current 90-day feeding tests were “insufficient” and should be “improved and prolonged”, “made compulsory” and include sexual hormones and reproduction (aspects not routinely assessed).

The third review appeared later in 2011. Another French team, lead by Agès Ricroch (with lead author, Snell Chelsea of Nottingham University) sought to establish the value of carrying out feeding trials longer than the standard 90-days. It critically examined 24 selected GM feeding studies: 12 long-term ones (more than 90 days), and 12 multigenerational ones. Note was also made of eight 90-day rat feeding trials “for which long-term or multigenerational studies were conducted” (although these don't seem to have been published).

This review highlighted some “major insufficiencies” in the science due to a failure to adhere to standard procedures. For example:

1. Seventeen out of the 24 papers were based on a comparison between whatever GM and non-GM feed the researchers could get hold of, rather than between the genetically transformed and the equivalent non-transformed ('isogenic') strain grown side-by-side. (The latter conditions are necessary to minimise genetic and environmental confounding factors, leaving the novel DNA construct and its insertion as the only differences between the 'test' and 'control' diets. Commercially available feed ingredients will come from a pool of crops exposed to different soil and weather conditions, and contaminated by a wealth of different chemical pesticides and fertilizers.)
2. Sixteen of the 24 studies were exploratory in nature and measured parameters not previously investigated during toxicological investigations (Such unusual assays require validation to clarify their implications). These studies had not been repeated by others, and therefore had been over-interpreted.

It was also noted that “No long-term rodent studies are available for GM maize”. As a compromise, a 23-month study on dairy cows was included in the review.

Conclusions reached were:

1. “The studies reviewed present evidence to show that GM plants are nutritionally equivalent to their non-GM counterparts and can be safely used in food and feed”
2. Long term studies “do not reveal any new effect that has been found in 90-day studies: it could be assumed, therefore, that standard protocols are efficient enough to detect adverse effects and there is no need to design new protocols ...”
3. “The observations of major flaws in some papers highlight the urgent need to improve the reviewing process before publication of papers addressing this subject.”
4. “Now, for us, the debate on GMOs from a health point of view is closed” (statement made to the press).

OUR COMMENT

The pro-GM lobby has of course been hyping the conclusions of this third review of the year, supporting as it does the industry-led assessment system for GM food.

However, French newspaper, Le Monde, was less impressed and quoted a leading statistician as saying the study was “biased” and “extremely slanted”.

Indeed, do the authors' conclusions really seem to follow from the evidence reviewed?

The Ricroch review pronounced GM safe and the standard 90-day rodent feeding tests sufficient. It identified the peer reviewers as being to blame for the presentation of data from new, unestablished, techniques in the published literature.

The conclusion of safety was drawn despite the “major insufficiencies” in the science and a resultant lack of robust data. It's clear from the statement that “No long-term rodent studies are available for GM maize” that the authors' stated aim of establishing the value of feeding trials longer than the standard 90-days was simply not possible for this very major source of GM in our food chain: the data were not there to review. This lack also gives no basis for the conclusion that the 90-day feeding protocol, which was developed for purposes other than the assessment of novel foods, is sophisticated enough to assure safety. Indeed, according to these reviewers the quoted papers had such major flaws that they should not have passed the peer review and should not have been published at all. This seems a blatantly anti-science and anti-scientist stance.

The Ricroch team found their fellow scientists guilty of “over-interpretation” of new techniques, yet were happy to include in their 'safety' review feeding studies using fish, ruminants and birds. One such study (described by the German Federal Agricultural Research Centre as 'pivotal') was included in this review despite being clearly a test of meat production performance and certainly not of human 'health impact' . After criticising other scientists for compromising in their choice of experimental feed, the team was happy to make a compromise of its own by comparing rats with cows. This also raises a question Ricroch is happy to leave unasked: is a 'short-term', 3-month, trial on a rat whose normal life-span is 24-35 months actually shorter than a 2-year trial on cows whose normal life-span is over 20 years? These strangely unscientific comparisons and the inclusion of material irrelevant to the stated aim of the study seem to have got past the Ricroch team's own peer reviewers.

The key problem here is that, if you discard all the experiments with “major insufficiencies” (i.e. those for which no isogenic control feed was available, plus those for which the GM event was not specified, plus those carried out on animals irrelevant to humans, plus those carried out using crops not commercialised so we'll never eat them) there are no data whatsoever left to review. Apart from being breathtakingly nonsensical, the conclusion that the GM debate “from a health point of view is closed” is the biggest case of over-interpretation ever.

It must be mentioned also, that contrary to the political insinuations in the paper that the lack of suitable test feed is caused by activists and politicians, GM crops are obviously grown extensively in industry research establishments and on the vaste farms in North and South America: test materials could easily be made available.

Most scientists have used this difficulty in obtaining valid test feed to call for commercial GM crop-plants and their isogenic lines grown together to be made available for independent science to be carried out. So long as scientists do not have access to the appropriate materials and the biotech industry is not required to provide them, safety of GM foods will not be established and modern sophisticated toxicological techniques will not be developed.

'No evidence' doesn't mean 'no evidence of harm', nor does it mean 'evidence of safety' no matter what way you look at it.

However, there's a bigger picture to consider here. Early in 2011, two reviews based on data carefully selected for comparability and for the presence of disease-indicators both came to very negative conclusions on GM safety. At the end of the year, a third review appeared, this time based on an odd assortment of data, some of which didn't exist. This one came to the very positive, but unsupported, conclusion that GM food is safe, plus the dismissive, but unsupported, conclusions that modern methods of safety testing and longer studies are unnecessary. Let's face it, if you were a biotech industry PR department and you wanted something to counter all the very inconvenient negative information coming out of the published science, Ricroch's paper is tailor-made.

SOURCES

• José L. Domingo, 2007, Toxicity Studies of Genetically Modified Plants: A Review of the Published Literature, Critical Reviews in Food Science and Nutrition 47
• José L. Domingo and Jordi Giné Bordonaba, 2011, A literature review on the sfety assessment of genetically modified plants, Environment International 37
• Gilles-Eric Séralini et al., 2011, Genetically modified crops safety assessments: present limits and possible improvements, Environmental Sciences Europe 23:10
• GM Feed Toxic, New Meta-Analysis Confirms, Institute of Science in Society Report 5.09.11
• Sness Chelsea et al., 2011, Assessment of the health impact of GM plant diets in long-term and multigenerational animal feding trials: A literature review, Food and Chemical Toxicology
• Study concludes GM food is safe! GM Watch 29.12.11
• Gilles van Kote, Impact of GmOs on animal health: the debate is still not resolved, Le Monde, 16.12.11
• No health problems for animals fed on GM crops (study), Ricroch Press Release 13.12.11
• G. Flachowsky et al., 2007, Studies on feeds from genetically modified plants (GMP) - contributions to nutritional and safety assessment, Animal Feed Science and Technology, 133