Seralini, author of the infamous study alleging to show increased rates of tumors in rats fed GM food, the one that was retracted by the journal and then later republished in a separate journal, has published another controversial study.

The study, published in PLOSone, looks at the feed that is fed to lab rodents, the kinds used in GM research. They found:

All diets were contaminated with pesticides (1-6 out of 262 measured), heavy metals (2-3 out of 4, mostly lead and cadmium), PCDD/Fs (1-13 out of 17) and PCBs (5-15 out of 18). Out of 22 GMOs tested for, Roundup-tolerant GMOs were the most frequently detected, constituting up to 48% of the diet.

The implication is that all prior research looking at GMO and pesticide toxicity is now called into question because the control rodents would also have been fed a diet that contains some GMO, pesticides, and also heavy metal contaminants. The concept here is valid – control groups need to be proper controls. If you are testing the effects of a pesticide on rats, and the control rats are also getting the pesticide in their food, then the comparison is compromised. This would dilute out the effects of the test substance by increasing the background rate of tumors and other negative outcomes, the “noise” in the study. This would further mean that studies would have to be more powerful (contain more subjects in each group) in order to detect the diluted signal.

Although the logic of their paper is sound, the devil is always in the details. How much pesticides and other contaminants did they find? They found 1-6 pesticides out of the 262 they tested for in each feed studied. This does not sound impressive. Further, they did not demonstrate that the small amounts detected were biologically relevant. The dose makes the toxin, and trace contaminants are almost ubiquitous, but are well below biologically active levels.

What they did not demonstrate is that there is any difference between lab rodents fed the foods they tested vs a control without any pesticide, heavy metal, or GMO contaminants. That, of course, is the ultimate question, and without that data they are just speculating.

Also consider that until 20 years ago there was no GMO feed. Therefore we have historical controls prior to the use of GMO, and any original GMO safety testing would by necessity have had no GMO in the control feed. There also does not appear to have been any significant change in the baseline rate of tumors and other effect from prior to the introduction of GMO varieties to after.

Seralini’s latest efforts are already being highly criticized by scientists around the world. The Genetic Literacy Project summarizes many of them here. One point that several scientists repeat is that Seralini used recommended limits on daily intake as his threshold for toxicity, but he used the limits for humans. Typically a 100 fold buffer is added when determining safety levels for humans. The more scientifically appropriate level to use would have been the no observed effect level (NOEL). Therefore the levels he detected in the food was still likely far below that which would cause any effect.

The GLP also points out that in the original version of the paper, sent to the media two weeks prior to publication, Seralini did not disclose a major conflict of interest that was later added to the published version (likely because PLOSone required it). The study was partly funded by a company, Sevene, that produces and sells homeopathic potions. This in itself is odd – any scientist allying themselves with a company whose product is pure pseudoscience is showing poor judgement, in my opinion. But worse – Sevene sells a homeopathic detox treatment for the pesticides glyphosate and atrazine. This is a clear conflict of interest, as they have a stake in the outcome of this study.

Conclusion

In science data is king, but data is only as good as the methods used to collect it. We can evaluate those methods by reading the formal presentation that is published in the technical literature, but this still does not give us a complete picture. As I have discussed many times here, there is a great deal that goes on behind the scenes of a scientific study that may not be reflected in the final paper. For this reason some journal editors are considering requiring submissions to include raw data and any notebooks used during the study, to get a peek behind the scenes and help root out fraud or just sloppy technique.

This all means that the reputation of the researcher, their team, and their lab is important, as are full disclosure of any potential conflicts of interest. This is also why independent replication is critical.

It is not uncommon for one researcher or team, or a small group of researchers, to be out of step with the rest of the scientific community. Often these researchers have a particular ideology that they seem to be promoting through their research. There are, for example, anti-vaccine researchers who seem to be the only ones to come up with data calling into question the safety of vaccines. The same is true for the safety of cell phones, or wifi.

Seralini has earned a reputation as being a scientific outlier in this sense. The quality of his studies are generally highly criticized, his results questionable, and his conclusions tend to go beyond the evidence he presents and to have a decidedly anti-GMO theme. It is therefore difficult to have faith in any of his studies.

Even putting his reputation aside, the quality of this latest study, as published, has major flaws and is simply not compelling. That will not stop anti-GMO activists from exploiting it to dismiss GMO safety studies, which seems to be the intent.