Grist's new food writer fails to see beyond industry spin in his series of articles on GM. Report: Claire Robinson

The online magazine Grist is well known for its coverage of environmental issues and for its incisive writing on food, agriculture, and GM, from the likes of Tom Philpott (now with Mother Jones) and Tom Laskawy, the founder and executive director of the Food and Environment Reporting Network.

Recently, Grist has offered us a new food writer, Nathanael Johnson, a Philpott replacement who has started to pen a series about GMOs. Johnson says a friend asked him before he started, "So are you for them, or against?" to which he answered, "I'm trying to figure that out."

His figuring out has given considerable pleasure to GM supporters. Jon Entine of the Cato Institute, whom Tom Philpott has called an "agribiz apologist", has celebrated in the columns of the business magazine Forbes "the sudden and surprising turn” of Grist. Entine uses Johnson’s pieces to pontificate on what constitutes good science journalism – the creators of which, in his view, include himself, Mark Lynas, the pro-GM blogger Keith Kloor, and the bloggers at Biofortified, who are keen admirers of the GM plant scientist Pam Ronald.

Ronald might be said to typify the problems with Johnson's series exploring the GM issue. In the third piece in this series, Johnson set out to explore the differences between conventional breeding and genetic engineering, with Ronald as his principal guide. For additional information he turned to another pro-GM plant scientist, Margaret Smith of Cornell. Unsurprisingly, the resulting article was supportive of GM’s safety.

The only real balance he offered to these two keen GM supporters were some things he had heard at a talk given by the microbiologist Dr Ignacio Chapela. This juxtaposition would seem to suggest a severe lack of balance: on the one hand, direct quotes arising out of hours apparently spent meeting and talking to Ronald and her assistant, and a direct interview with Margaret Smith, contrasted with notes from a lecture Chapela gave “years ago”. Again unsurprisingly, it's Smith who gets the final word.

Sadly, Johnson's latest article, "Is extremism in defense of GM food a vice?", is full of the mistakes and misleading statements that I have come to expect from his supposedly open-minded series.

The most depressing thing is that Johnson could have avoided all these mistakes if he had done more in-depth research and asked some not-too-difficult questions. I suspect the problem was that once again he was relying on pro-GM scientists for his sources, since his mistakes follow exactly the lines pushed by GM industry spin doctors.

Let's take a few examples.

1. Johnson misses the most serious scientific error of all

Johnson presupposes that scientists whose research has uncovered risks associated with GM crops have committed "errors". And yes, apparently he means all of them.

Sadly he fails to address the one serious type of error in GM safety research that has the power to endanger public health: the type II error, or false negative - when a toxic effect exists but is missed in the test. Critics of GM crops among the public and the scientific community are preoccupied with these errors, which they believe have compromised GM crop approvals from the beginning.

Instead, Johnson focuses on a supposed "error" by the researcher Prof Arpad Pusztai, who found ill health effects in rats fed GM potatoes back in the 1990s. But the error turns out not to be Pusztai's, but Johnson's, as explained below.

2. Johnson misrepresents Pusztai’s research

Johnson claims the differences found in Prof Arpad Pusztai's GM potato-fed rats were due to genetic shifts caused by Pusztai's failure to compare the GM potatoes with their non-GM parent line, with the same genetic background, and grown in the same conditions (what Johnson inaccurately calls the "unmodified clone"). Thus, Johnson implies that the differences in the GM-fed rats were due to genetic changes in the potatoes, which in turn were caused by environmental factors like different growing conditions.

The problem with this characterization of Pusztai’s work is that it is 100% wrong. Far from not understanding the importance of environmental conditions in creating variability in plants, Pusztai was one of the first scientists to educate me and countless others about this fact.

Pusztai recognised that if you are trying to find out if a GM plant is more or less toxic than the non-GM plant it was developed from, then the only valid comparator is the non-GM parent plant, grown side-by-side in the same conditions.

For this reason, Pusztai and his research collaborators took great care to grow the GM potatoes side-by-side in the same conditions with the non-GM parent potato.

Incidentally, while Pusztai was conscious of this scientific principle back in 1998, it is still routinely ignored by the GM industry. Industry pushes its GM crops through the regulatory system by comparing the GM plant *not* just with the non-GM parent, as the scientifically rigorous method propounded by Pusztai and other independent scientists requires. This is arguably because GM industry scientists know very well, given the imprecision and crude nature of the genetic engineering process, that the GM plant will not be the same as the non-GM parent. Instead, industry broadens the comparison to a whole range of different varieties of the crop, grown at different times, in different conditions. "Look," the industry says, "there are all these differences between soy varieties grown all over the place, and our GM variety fits somewhere in the range of all those. So it must be OK!"

Thus the industry declares its GM crop as "equivalent" to non-GM crops, the regulators rubber stamp the industry's view, and another badly assessed GM product is released onto the market. Badly assessed, because in this game of pretended "equivalence", only a few basic components like fats and proteins are analysed. Unexpected toxins and allergens will be missed.

In truth, what shocked Pusztai most about the GM potatoes he tested was that there was every reason why they should have been the same as the non-GM parent, apart from the intended change of insect resistance – but they were not. Why? There's only one possible explanation – the one Pusztai reached. It was something to do with the genetic engineering process.

Granted, the fact that Pusztai used the correct comparator is not stated in his published paper in The Lancet. Perhaps this was because at the time, few people (apart from Pusztai) accepted that it was important. Pusztai did, however, make it clear in an interview with GeneWatch magazine.

It took me a full 30 seconds of googling to find the interview. Perhaps Johnson felt he did not need to check the 'line' he had got from his source – he refers, for instance, to an "excellent analysis" of the Pusztai affair by the "plant geneticist and prominent GM food supporter Nina Fedoroff". But when the credibility of your story hangs on a certain line of argument, you need to do a bit of basic fact checking. This is even more important when the source – in this case Fedoroff – has strong biotech interests – something else that Tom Philpott has previously flagged up on Grist.

Fedoroff is certainly keen to let GM off the hook as the cause of the toxicity of the GM potatoes in Pusztai's experiment. But interestingly, she ends up doing the exact opposite. In trying to neutralise Pusztai's conclusion that the GM process was to blame for the unexpected toxicity of the GM potatoes, Fedoroff correctly points out that the GM process isn't just a matter of inserting the transgene, but also includes a period of growing on the genetically transformed plant cells in tissue culture. Tissue culture in itself is a highly mutagenic process which can produce unexpected changes. Fedoroff thinks that the tissue culture, rather than the transgene insertion, might have created the unexpected changes in the GM potatoes. Thus far, she is correct – but oddly, she seems to think this should reassure us about the safety of GM.

This is illogical, since all GM plants have to go through the tissue culture phase. Far from being reassuring, the fact that the tissue culture phase could have been responsible for the toxicity of Pusztai's GM potatoes is yet another reason – on top of the disruptive effects of inserting the transgene – to doubt GM's safety. http://earthopensource.org/files/pdfs/GMO_Myths_and_Truths/GMO_Myths_and_Truths_1.3b.pdf (Chapter 1)

So in defending GM, Fedoroff is drawing attention to another hazard of the GM process. This hazard too seems to have passed Johnson by.

3. Johnson fails to understand disease processes

Johnson claims that "the only real difference" found in the GM-fed rats in Pusztai's research was "a slight change in gut cells". Again, Johnson seems to have been misled, perhaps by plant scientists with no knowledge of toxicology or pathology (disease development). Most pro-GM scientists fall into this category, though they are nevertheless quite happy to sit in judgment on the work of scientists who are qualified in those fields.

What Pusztai actually found was proliferative growth of the cells of the digestive tract in the GM-fed rats. This is concerning because, as a pathologist such as Pusztai’s research collaborator Stanley Ewen would know, such changes may be a prelude to cancer. Then again, they may have been harmless. If Pusztai had been allowed to continue his studies, we would know, one way or the other. We do not know because the research was summarily shut down and a gagging order was imposed on the researchers. Pusztai's experiments were never repeated.

Here, for Johnson's benefit, is a lesson in scientific method. It is simple, though he won't learn it from his preferred advisors on GM. The way to show that Pusztai – or any other researcher who has found hazards associated with GM foods – was wrong is to repeat the experiments and get a different result. Last time I looked, no one has done this, either with Pusztai's experiments or any others that have found problems with GM foods. That means, those results stand.

4. Johnson believes it’s easy for critical researchers to gain access to GM companies’ crops

Johnson then turns his attention to Dr Judy Carman, whose recent research found more severe stomach inflammation and heavier uteri in pigs fed GM feed. He says Carman and colleagues "repeated Pusztai's error" in failing to use the non-GM parent crops as comparators – an “error” that Pusztai did not, however, make, as I show above.

In fact Carman did not use non-GM parent crops as comparators because she, in common with other researchers, couldn't get hold of them, since they are the property of GM companies. Johnson asserts – on the basis of no evidence – that Carman and her team "could have partnered with any researcher at a university in the United States, which have permission to study patented seeds".

This seems touchingly naive. Johnson apparently believes that a US university that has established hard-won research agreements with Monsanto (and may even depend on research funds from the company) will happily hand over the company’s proprietary materials to a researcher who is known to ask critical questions about GM crop safety. He also believes that somehow the university will have access to the correct research materials. That means not only the GM crop material, all grown at the same time in the same conditions (to avoid the confusion caused by different environmental conditions that Johnson referred to earlier), but the non-GM parent variety, grown side-by-side in the same conditions.

Accessing the non-GM parent variety seems a particularly onerous ‘ask’, since the GM industry has told the European Food Safety Authority that this variety may not be “available”, even to the regulator.

Presumably Johnson also believes that the university, once told about the planned research, would keep it totally confidential from Monsanto and its allies, a condition that history has taught us is necessary to enable the research to be safely completed and published.

Johnson is entitled to his beliefs, but they seem to bear little relation to the real world.

5. Johnson fails to notice he’s hoisted himself on his own petard

Johnson claims, "When long-term feeding trials are done so as to compare apples to apples, they’ve reinforced the conclusion that genetically modified foods are safe." He gives a reference to the review of the scientific literature on this issue by Snell and colleagues[1].

In fact the Snell review provides no such evidence.

Some of the studies cited by Snell:

• do not examine health effects but are animal production studies of interest to the food industry and farmers, looking at effects like milk yield and body weight gain
• are not conducted over the lifetime of the animal and so do not test for the effects of lifetime consumption
• do in fact find signs of toxicity, which are dismissed either by the authors of the original studies or by Snell and colleagues as not being biologically significant (without scientific justification)
• have serious methodological shortcomings that limit the conclusions that can be drawn, such as not giving basic information like numbers of animals used; using very small groups of animals that are insufficient to prove safety; or not using the non-GM isogenic (genetically the same) variety as the comparator for the GM crop. In other words, they do NOT, to borrow Johnson's phrase, "compare apples to apples".

Here I come back to the problem so casually dismissed by Johnson as an issue of the past that's been wafted away by the biotech industry's magnanimous research agreements with some US universities. That is: the GM industry generally refuses access for independent researchers to the necessary research materials to do comparative studies: the GM crop and the non-GM parent crop grown side-by-side in the same conditions. Snell and colleagues correctly point out in their review that these methodological shortcomings affect studies that find risk and safety alike.

But then, in an example of the double standards that we've come to expect of GM supporters, Snell and colleagues accept the findings of safety from these methodologically compromised tests at face value, while rejecting the findings of harm! This sleight-of-hand has been condemned by the European Network of Scientists for Social and Environmental Responsibility (ENSSER).[2]

It's especially bizarre, given Johnson's concern (misplaced, as it turns out) that Pusztai failed to use the correct comparator, that he is happy to accept Snell and colleagues' conclusion of GM safety, based on studies that didn't use the correct comparator. So apparently, in Johnson's view, a less than ideal study methodology is fine if the conclusion is that the GMO is safe, but unacceptable if the conclusion is that GMO is not safe.

6. Johnson ignores GM industry bias in safety studies

Another interesting fact – but perhaps not surprising, given the overall slant of Johnson's series of articles – is that Johnson chooses to reference the Snell review but fails to cite two other peer-reviewed reviews of animal feeding studies with GMOs, which came to a very different conclusion. The authors of these reviews found that while some studies concluded that GM foods were safe and others identified hazards, there was one fundamental difference between the two sets of studies. Those that found the GM food tested was safe were more likely to be authored by scientists affiliated with the GM industry, whereas those that found hazards were more likely to be authored by researchers independent of the industry.[3,4] This is due to funder bias, a phenomenon of which Johnson appears blissfully unaware.

Johnson's intention of taking a fresh open-minded look at the GM debate is entirely commendable. But, as a newcomer to the GM issue, he appears to have placed an unwarranted degree of trust in a limited subset of pro-GM scientists, and as a consequence he has been badly misled and exploited by his advisors. It's worrying that he is in a position to pass on this slanted information to Grist's readers as the careful conclusions of an even-handed investigator.

Notes

1. Snell, C., et al. (2011). "Assessment of the health impact of GM plant diets in long-term and multigenerational animal feeding trials: A literature review." Food and Chemical Toxicology.
2. European Network of Scientists for Social and Environmental Responsibility (ENSSER) (2012). Questionable biosafety of GMOs, double standards and, once again, a "shooting-the-messenger" style debate. http://bit.ly/SHCfvm
3. Diels, J., et al. (2011). Association of financial or professional conflict of interest to research outcomes on health risks or nutritional assessment studies of genetically modified products. Food Policy 36: 197–203.
4. Domingo, J. L. and J. G. Bordonaba (2011). A literature review on the safety assessment of genetically modified plants. Environ Int 37: 734–742.

Update 28 Aug 2013: Since this article was published, it has been pointed out to me that Pusztai was the first scientist to use and advocate a second control group in testing GMOs for safety, in which the gene product (in this case the insect-resistant lectin) was added to the parent potatoes at the same concentration as it was present in the GM potato. This way, the difference in composition and the unexpected toxicity found from the GM potatoes could only originate from the process of genetic modification. – Claire Robinson