Biology Fortified recently reported that “an Italian research group run by Professor Federico Infascelli at the University of Naples “Federico II” has just had a peer-reviewed paper forcibly retracted by the journal for plagiarism”. The retracted paper, entitled “Gamma-Glutamyl Transferase Activity in Kids Born from Goats Fed Genetically Modified Soybean”, has been used as evidence of GMO harm (see here as an example). I reviewed the paper on my personal blog last year and was struck by it’s flawed methodology. Given the fact that the paper is in the news again due to its retraction and the scientific team is being investigated due to allegations of fraud, I thought I’d share some highlights of the paper on the Biofortified Blog.
First, let’s examine the quality of the journal that it was published in: Food and Nutrition Sciences. The journal is not indexed by the NIH’s database of scientific publications, which immediately raises a red flag because it indicates that the journal does not meet the NIH’s criteria for a quality publication. Next, I searched for the journal in Beall’s list of predatory publications: sure enough, the publisher “Scientific Research Publishing” is listed. Predatory journals will publish nearly anything, as long as you pay their hefty publication fee. There have been several exposés on these journals (see here and here) and the crazy papers that they’ve accepted for publication (my favorite one: the journal accepted a paper entitled “Get Me Off Your F*ing Mailing list“. I particularly enjoyed the diagrams from the paper).
Aims of Study
The authors start by outlining and defining RoundUp Ready Soy. They state that “the majority of animal feeding trials using GM feeds indicated no clinical effects”, but that there is data indicating liver and kidney problems. Their reference for this last statement is the infamous Seralini paper, which was retracted and later republished. They also point to a few other papers as the basis for their study:
- a paper which “hypothesized that cell metabolism of several enzymes was altered in rabbits fed GM soybean” (this paper is authored by the Italian team)
- a paper that found pieces of DNA from the transgene in “goat milk but also the kids organs when mothers are fed GM soybean”. The paper also found higher levels of an enzyme in these animals which serves as an indicator of injury and disease (LDH) (this paper is also authored by the Italian team).
The aim of the paper I’m reviewing is to find out if there are:
- DNA fragments from the GM soybean
- changes in the activity of an enzyme named gamma-glutamyltransferase (GGT) which can serve as an indicator of liver disease, in the kid goats whose mothers were fed RoundUp Ready soybean.
Flawed materials and methods
The authors state that they performed the experiment on 20 male kids born from goats fed a soybean extract from conventional or RoundUp Ready soy. After this point, nothing else really matters because the authors didn’t do an analysis of the feed. Dr Anastasia Bodnar and I recently made an infographic and wrote an article to accompany it, outlining how animal feeding studies should be properly designed. The analysis of the feed is a crucial aspect of such studies. Here’s why: if you are going to do an experiment to determine if a single variable impacts a system (in this case, that the presence of the protein that confers RoundUp Ready resistance causes harm in goats), then you have to be as sure as possible that nothing else is different. Studies have shown that the location and environment where a crop is grown create greater variability in the crop than the presence of the transgene. For example, if I take an ear of corn from Northern Ontario and compare it to an ear of corn from Southern Ontario (of the same variety), they will be more different in terms of nutrients/amino-acids/minerals than taking an ear of corn from Southern Ontario that’s a GMO and comparing it to a non-GMO variety that doesn’t have the transgene grown in Southern Ontario.
The goat feeding study didn’t do an analysis on the composition of the soybeans used in the study. It provides no information on the variety of soybean used, the location where they were grown, or even if they were from the same season, etc. In general, well designed feeding studies perform an analysis of the composition of the feed that is given and then make the feed equivalent by adding supplements, so that the variable in the feed is the presence/absence of the transgenic protein (and the gene that encodes for it).
Here’s a hypothetical example: the scientists conducting a study buy regular soybeans and transgenic soybeans and do a nutritional analysis. They find that the regular soybeans have 12% less calcium and that the transgenic soybeans have 8% less of the amino acid lysine. Maybe it rained a bit more in the farm where the regular soybeans were grown, causing this difference. Maybe they added more fertilizer to the soil where the transgenic soybeans were grown, leading to these differences. The researchers will add calcium to the feed that consists of regular soy and add lysine to the feed that consists of transgenic soy. Otherwise, they won’t be able to conclude if any differences observed are due to the calcium, the lysine, or the transgenic protein.
That’s where the paper falls apart.
As expected, the authors detected the presence of DNA from the transgene in organs and blood. Does it matter? Not really: if I took two people and fed one strawberries and the other blueberries, I’d possibly detect small pieces of DNA from the cells of the strawberry in the blood of one person and small pieces of DNA from the cells of the blueberry in the blood of the other person. Does it mean that the individual will turn blue or red? No. So why would DNA from a transgenic crop be any different? Additionally, the presence of DNA from a transgene in someone’s blood does not mean that it has become part of that person’s DNA (see this post and infographic for more information on this point).
It’s also surprising that the authors have several papers that are very similar to one another, yet their findings such as impact on kid growth/size are not reproducible (see here, here, and here).
Quality of Studies
When Anastasia and I started working on the infographic for animal feeding studies, I was motivated out of the sheer frustration of seeing poorly designed studies which were then used in campaigns against GMOs. Even if this particular paper had not been retracted, even if the authors had not been accused of fraud, the research was not well-designed and its findings were not solid. Even if it had been conducted ethically, it still should not have been published, let alone touted as evidence of GMO harm because it’s not a good paper. It should not have passed peer review, and it’s possible that it never did, leading the authors to publish in a predatory journal. Yet the paper, and all other papers from this group, have been used in campaigns to warn us of the dangers of GMOs.
I challenge all of us to demand evidence when we see memes and catchy slogans. If you do not feel comfortable evaluating a particular study and are curious about its findings, I think it’s safe to say that someone at Biology Fortified would be happy to help you out.
My final comment is this: I’m not a very strong animal advocate, but I believe that they should be treated with care. I’m not a vegetarian, let alone a vegan. I’m not a member of PETA. But I believe that scientists have a responsibility to perform animal studies judiciously. If it’s not necessary, if there’s another way to reach the same conclusion, then give it a shot before you resort to animals. Unfortunately, this experiment was not an example of the judicious use of animals.