European Food Safety Authority Analyzes the new Seralini Paper

GMO Panel deliberations on the paper by de Vendômois et al. (2009, A Comparison of the Effects of Three GM Corn Varieties on Mammalian Health, International Journal of Biological Sciences, 5: 706-726) - EFSA/GMO/578 – part of the Minutes 55th Plenary Meeting of the GMO Panel Adopted part of the minutes1 of the 55th plenary meeting of the Scientific Panel on Genetically Modified Organisms held on 27-28 January 2010 to be published at http://www.efsa.europa.eu/en/events/event/gmo100127.htm

The EFSA GMO Panel has considered the paper by de Vendômois et al. (2009, A Comparison of the Effects of Three GM Corn Varieties on Mammalian Health, International Journal of Biological Sciences, 5: 706-726), a statistical reanalysis of data from three 90-day rat feeding studies already assessed by the GMO Panel (EFSA, 2003a,b; EFSA 2004a,b; EFSA 2009b,c). The GMO Panel concludes that the authors’ claims, regarding new side effects indicating kidney and liver toxicity, are not supported by the data provided in their paper. There is no new information that would lead it to reconsider its previous opinions on the three maize events MON810, MON863 and NK603, which concluded that there were no indications of adverse effects for human, animal health and the environment.

The GMO Panel notes that several of its fundamental statistical criticisms (EFSA, 2007a,b) of the authors' earlier study (Seralini et al., 2007) of maize MON863 are also applicable to the new paper by de Vendômois et al. In the GMO Panel's extensive evaluation of Seralini et al. (2007), reasons for the apparent excess of significant differences found for MON863 (8%) were given and it was shown that this raised no safety concerns. The percentage of variables tested reported by de Vendômois et al. that were significant for NK603 (9%) and MON810 (6%) were of similar magnitude to that for MON863.

The GMO Panel considers that de Vendômois et al.: (1) make erroneous statements concerning the use of reference varieties to provide estimates of variability that allow equivalence testing to place statistically significant results into biological context as advocated by EFSA (2008, 2009a); (2) do not use the available information concerning normal background variability between animals fed with different diets, to place observed differences into biological context; (3) do not present results using their False Discovery Rate methodology in a meaningful way; (4) give no evidence to relate wellknown gender differences in response to diet to claims of effects due to the respective GMOs; (5) estimate statistical power based on inappropriate analyses and magnitudes of difference.

The significant differences highlighted by de Vendômois et al. have all been considered previously by the GMO Panel in its previous opinions on the three maize events MON810, MON863 and NK603. The study by de Vendômois et al. provides no new evidence of toxic effects. The approach used by de Vendômois et al. does not allow a proper assessment of the differences claimed between the GMOs and their respective counterparts for their toxicological relevance because: (1) results are presented exclusively in the form of percentage differences for each variable, rather than in their actual measured units; (2) the calculated values of the toxicological parameters tested are not related to the normal range for the species concerned; (3) the calculated values of the toxicological parameters tested are not compared with ranges of variation found in test animals fed with diets containing different reference varieties; (4) the statistically significant differences did not show consistency patterns over endpoint variables and doses; (5) the inconsistencies between the purely statistical arguments of de Vendômois et al., and the results for these three animal feeding studies which relate to organ pathology, histopathology and histochemistry, are not addressed.

Regarding claims made by de Vendômois et al. concerning the inadequacy of the experimental design of these three animal feeding studies, the GMO Panel notes that they were all carried out to agreed internationally-defined standards consistent with OECD protocols.