French scientists find signs of toxicity to liver and kidney in Monsanto’s study on its controversial GM maize. Dr. Mae-Wan Ho
In 2003, Monsanto submitted a request to market the GM maize MON 863 with resistance to corn rootworm to the regulatory authority in Germany. Monsanto was seeking approval for importing the maize into the Europe for processing and use as feed, but not as food or for cultivation.
The German authority concluded that there was no scientific evidence of risk for human health or the environment. But other Member States raised and maintained objections over molecular characterisation, the potential for allergic reactions, toxicity and other aspects. Most controversially, the feeding study submitted by Monsanto turned up many adverse effects [1] that were dismissed by Monsanto as “not biologically meaningful.”
Monsanto, supported by European Food Safety Authority (EFSA), kept the study from public scrutiny under a false claim of confidential business information until a German court order a year later forced Monsanto to release the full report.
Preliminary analysis by Giles-Eric Séralini and colleagues of Criigen (Committee for Independent Research and Information on Genetic Engineering, France) [2] found serious flaws in the study at every stage, from experimental design, to data collection, analysis, and reporting. The GM maize fed group was compared, not just to the group fed the non-GM isogenic line as it should have been, but also to five more ‘control’ groups fed other non-GM varieties. This had the effect of increasing the range of variation and making the treatment group of animals too small, thus considerably reducing the sensitivity of the test. The results were analysed with the wrong statistical tests, and despite having compared many variables, the correct standard statistical tools (multivariate and principal component analyses) were not used. Instead, in comparing one variable at a time, the researchers failed to note significant trends in body weight differences between experimental and control animals. Statistically significant differences that nevertheless turned up were then all dismissed as biologically insignificant.
The EFSA, when consulted, agreed with Monsanto, and gave MON 863 maize a ‘positive opinion’. In August 2005, the European Commission gave approval to maize MON 863 [3], despite a persistent failure to reach an agreement by all regulatory authorities.
Monsanto published a report on the study in 2006 [4], restating its position that the significant differences between transgenic and non-transgenic fed groups were “not biologically meaningful.” But Séralini and colleagues reanalysed the data in detail, using the appropriate statistical tests [5] and declared MON 863 maize “unfit for consumption” [6] as it shows signs of being toxic to the liver and kidney of rats. They called for “a moratorium on other approved GMOs while the efficacy of current health testing methods is reassessed.” They also wrote to the EFSA [7] to “urgently ask for a moratorium on MON 863.”
Séralini and colleagues found that the transgenic maize affected the two sexes differently, which is often the case for effects due to pesticides that disrupt sex hormones. A standard multivariate analysis showed a highly significant and sustained 3.3 percent decrease in body weight in males, and a 3.7 percent increase in females. In addition, the kidneys were reduced in weight in males and the ionic composition was modified in urine, as consistent with a disturbance in kidney function. In females, there was a significant rise in blood sugar and triglycerides as well as an increase in liver weight, also indicative of kidney and liver malfunction.
Out of a total of 494 comparisons besides body weight between GM fed and control groups, 40 (8 percent) showed significant differences, whereas only 25 differences (5 percent) were expected by chance alone.
The tests were insufficiently long term, and the hormonal levels of the rats should have been investigated, Séralini said [6].
There are large gaps in current health assessment of GM food and feed [8] (GM Food Nightmare Unfolding in the Regulatory Sham, ISIS scientific publication, alsoSiS 33).
As Séralini and colleagues point out in the case of MON 863, a new artificial protein Cry3Bb1 is produced at relatively high levels (49 – 97 microgram/gram). Its mechanism of action is not known in mammals because it was never studied, and the target receptor has not even been characterised in insects. Most of all, unlike ordinary drugs or pesticides, which have to be tested for three-months in three mammalian species, then with one mammalian species for one year, and yet another for 2 years, current regulation does not require such tests for ‘biopesticides’ produced continuously in open fields; nor for the herbicides and herbicide residues accumulated by herbicide-tolerant GM crops. The two traits, biopesticides and herbicide tolerance now account for practically all GM crops grown in the world today.
Article first published 03/05/07
Got something to say about this page? Comment