Elusive GM safety evidence

December 2014

A review of the scientific evidence on the relationship between GM crops and animal health has just been published. This carefully constructed search for a robust body of studies concluded that the task was "impossible" to do "properly" because of the fragmented nature of the approach to testing and the gaps in the presentation of the methodology and data.

The studies had all been peer-reviewed and published. Notably, reports submitted to regulatory authorities were omitted due to insufficient detail.

The Australian team, lead author Dr. Zdziarski, focused its review on studies which were similar enough to be compared: they looked at rats (the accepted standard model), fed commercially available GM crops containing any number of the three most commonly-used artificial genes, for a minimum of 90 days. In particular, they were looking for key measurements of effects on the animals' digestive system.

Because the digestive tract is intimately exposed to GM food for a considerable period of time, both in whole form and in all stages of decomposition it is the most likely organ to react if harmful elements are present. Longer-term responses arising from adverse changes in gut microbial activity may also be present.

Digestive health is, of course, key to all aspects of health of the whole body. Assessments limited to, for example, organ weight, or visible lesions and tissue changes, will miss all but the grossest symptoms of disease. The Australian team was, therefore, looking for data which included histopathological analysis of gut tissues.

What they found was a breathtaking paucity of GM safety evidence.

Only 19% of the 47 commercialised GM crops containing the common artificial genes of interest (Roundup herbicide resistance, and/or 'Cry1Ab' and 'Cry3Bb1' insecticides) appear in published feeding studies. Three quarters of these experiments had been performed years after approval for consumption. Only nine studies fitted Zdziarski's criteria for comparability.

Even in these few papers, so little detail was included about why the particular study design had been chosen, nor about what exactly had been done, that none could have been repeated by other scientists. Indeed, there was a lack of clarity about the 'end-points' in other words what was actually being measured as a sign of toxicity or pathology.

Most notable was the lack of supporting data for conclusions. It was common to limit the 'results' to a statement such as that there were no treatment-related or diagnostically significant or toxicologically relevant observations, or no meaningful differences. The authors point out that any observation of a difference requires a note of the scale or prevalence in case an existing weakness in the test animals is being exacerbated by the test feed, or in case a longer exposure time is needed to fully reveal the problem.

Safety evaluations in most countries fail to recognise that a GMO is a novel material: assessments are mainly restricted to the novel substance produced by the artificial gene, ignoring any side-effects which might be present in the plant. As Zdziarski points out:

"... the process of developing GM crops may generate unintended effects. Furthermore, the plant developed is a novel entity with genes, regulatory sequences and proteins that interact in complex ways".

Thus, GM crops stacked with multiple novel genes, and feed consisting of a mixture of GMOs, have been allowed to enter the food chain without any additional safety checks for interactions between novel attributes. 

The Australian team concludes with a recommendation that GM safety testing be modelled on that used for drugs in which a complete assessment of risk (as well as benefits) must be made in the form, dosage and population of its intended use. 

It further points out that
"long-term animal feeding studies should be included in risk assessments of GM crops, together with thorough histopathological investigations using a variety of methods to better detect subtle changes or the beginning or presence of pathologies (such as inflammation or cell-death processes)".

This is the first step to establishing evidence-based guidelines for the safety evaluation of GM food and feed.


Note that as far as safety-testing of GM foods is concerned, we haven't actually reached the starting line yet. There is, effectively, a complete lack of evidence to support GM food safety.

The lack of methodological clarity, consistency and justification, plus lack of defined end-points made the current body of safety assessments of GM crops with the three most common artificial genes impossible to review or to repeat. 

Repeatability and review are the mainstay of science. Repeatability is evidence of truth and review is vital for the understanding of the bigger picture of which each experiment is just one small part.

Interestingly, we have gone full circle since the dawn of GM food when Dr Arpad Pusztai in Scotland was contracted by the government to develop safety tests. Pusztai's measurements of changes in gut cells in rats fed GM suggested an adverse reaction. It seems now that these were exactly the sort of tests we needed, but he was eliminated from the British scientific establishment and no such key studies have been performed since.

GM food safety is not established. The GM debate, far from being over, can't actually begin until there's some appropriately targetted and fully described science to debate.

Compare and contrast what's been said here with the FAST FAT ANIMALS (December 2014) being presented as proof of GM feed safety.



  • J.M. Zdziarski, et al., 2014, GM crops and the rat digestive tract: A critical review, Environment International 73

No comments:

Post a comment

Thanks for your comment. All comments are moderated before they are published.