NOTE: Please see our bulletin of yesterday for the background to this article:
 http://www.gmwatch.org/index.php?option=com_content&view=article&id=14335
---
---
The Seralini GMO study - a bone of contention to the Academy of Sciences
Paul Deheuvels
Member of the Academy of Science
Le Nouvel Observateur
19 October 2012
Article in French:  http://leplus.nouvelobs.com/contribution/661194-l-etude-de-seralini-sur-les-ogm-pomme-de-discorde-a-l-academie-des-sciences.html
English translation by GMWatch

This Friday, the National Academies of Agriculture, Medicine, Pharmacy, Science, Technology, and Veterinary Science issued a statement about the study of Gilles-Eric Seralini on Monsanto's NK603 GM maize. Their verdict? It contains "many shortcomings in methodology and interpretation". Paul Deheuvels, himself a member of the Academy of Sciences, is indignant.

Edited by Helene Decommer
Author sponsored by Guillaume Malaurie

I have just learned that a small number of representatives of the six Academies (Science, Medicine, Technology, ...) have come together to release a joint statement regarding the study by Professor Seralini (published recently in the journal, Food and Chemical Toxicology).

The Academies are not engaged in their entirety

Even without having read their statement, I have to draw the public's attention to the fact that the said statement can not engage any of these academies in their entirety. Indeed, a group of experts was convened in an emergency, we do not know by whom, no one knows how, with a total lack of transparency in the selection of its members, and on the basis of two representatives from each academy. These people have seen fit to write in a very short space of time an opinion highly critical of this study. They cannot claim to embody the opinion of the entire French scientific world, and it would be a crime to suggest that they do.

As the only member of the Academy of Sciences representing the discipline of statistics, it would have been normal for me to be consulted, yet that was not the case.

It appears from the conversations I've subsequently had about this statement that representatives of the five [GMW: should be six?] academies mentioned above criticized the descriptive part of the study by Professor Seralini on tumours, accusing him of not having established statistical significance. On the other hand, they ignored the toxicological part of the article - which was dealt with in a sophisticated manner using modern [statistical] methods (called PLS). Note that the article by Gilles-Eric Seralini is correctly published in a journal of toxicology.

A bad trial of Seralini's study

In any article of this type, the descriptive part simply describes [the findings], without drawing a demonstrated conclusion. This is what is happening here, and I cannot find fault with the professional design of this component, whatever the small group of signatories of the motion may say.

In contrast, the toxicology component of the article has the great merit of studying the complete body of data gathered from analyses over a period of about 16 months. This part of the article leads, in turn, to the highlighting of statistically significant differences, subject, of course, to the condition that the analysis has been properly done, and I see no reason to think that this was not the case.

It may be objected that these are precisely the descriptive parts of the article by Seralini that drew media attention, since they talk of tumours suffered by the experimental animals. But that's not the point. I think the study has suffered a bad trial, using bad arguments, carried out with a fury that is absolutely suspect, given the huge financial interests at stake. It is not a matter of what the media think, but rather to judge the technical quality of this work.

I want to say emphatically that the article by Professor Seralini is at a high level of quality among articles of this category. You cannot reproach its scientific value, which is undeniable. You cannot attack its descriptive part, which does not seek to establish proof in a statistical sense, but to present the findings in a readable, purely factual manner. It is even more unusual that the release of the five [GMW: six?] academies is not interested (if my information is correct) in the toxicological analysis part of the study, which makes use of confidence levels established by the usual statistical rules.

For a healthy exchange of ideas

I therefore reject in advance any text that is presented on behalf of five academies on this subject, because of the obvious fact that the committee that signed it only represents itself, independently of the eminence of its members.

This is also a process to limit the scandal, in which the few pretend to speak for all. Whatever the quality of the signatories, they express an opinion that cannot claim to be a universal truth, as the problem has not really been discussed openly and substantively.

The article by Professor Seralini has the immense merit of highlighting the fact that there are not enough studies on the long-term effects of GMO-based diets. Instead of criticizing this study for what it is not, while failing to appreciate its truly innovative components, organizations such as INRA or ANSES should undertake comprehensive studies on larger sample sizes. They would come out on top of a quarrel with clearly biased parameters. We need the debate of ideas to take place in a calm atmosphere, without being pressured by lobbyists who give their views for or against without even taking the time to discuss them.

I express my personal opinion based on my experience, and I do not insult those who have a contrary opinion. The value of science above all lies in a healthy exchange of ideas and arguments, that is, if possible, without bias. Currently I fear that we are very far from such a situation.