Friday, September 28, 2012
Fatal flaws alleged in rat-tumor-GMO study
by Larry Geller
Last week French scientists published a paper in a peer-reviewed journal that appeared to show that rats fed GMO corn developed a huge number of tumors.
This instantly inflamed the controversy over whether GMO food is safe to eat and also whether food containing GMO ingredients should be labeled.
Then it turned out that the study was not as conclusive as it should have been. In fact, it’s very problematic. Now, that doesn’t mean that GMO corn is safe to eat. But this study does not show it to be unsafe, either.
First of all, the press was manipulated into reporting the study uncritically when it was released. In exchange for an advance copy of the report, media was required to sign a confidentiality agreement that (according to the radio program On The Media) prohibited any publication until the middle of the press conference. This prevented reporters from checking the results with other scientists. The BBC, according to On The Media, refused to sign the agreement.
Although criticism arose soon after the media reports emerged, many people will find the original articles and not see later ones questioning the study. This ensures that the study will have a life of its own, regardless of flaws reported later.
The presented results look like a toxicologist’s nightmare. The authors reported high rates of tumor development in the rats fed Roundup and the Roundup-tolerant maize. There are figures of rats with visible tumors, and graphs showing death rates that appear to begin early in the rats’ lifespan. The media of course picked up on it, and one site in particular has spawned some reports that sound like mass hysteria. It was the first study showing that genetically modified foods could produce tumors at all, let alone the incredibly drastic ones shown in the paper.
But can GMOs really produce such huge tumors? This paper isn’t convincing. Following the release of the study, numerous scientists questioned the findings, citing “anomalies throughout the paper that normally should have been corrected or resolved through the peer-review process.” In particular, there are problems with the statistics performed on the data, the way the data were presented, and the numbers and types of animals used in the study.
[Discover Magazine, Under Controlled: Why the New GMO Panic Is More Sensational Than Sense, 9/21/2012]
For one thing, the sample size was much too small in each group, as the linked article points out. Another problem is that the rats chosen were special rats that have a high incidence of tumors ( the Sprague Dawley rat strain used for biomedical research). The control group suffered a 30% death rate, a problem not discussed, apparently.
It appears that dead rats were removed from the samples even though the experiment was a mortality study. Also, no standard errors were given.
In other words, as discussed in the critical article above, nothing much was proved after all.