The Organic Consumers Association (among others) has gleefully announced: New Study Links GMO Food To Leukemia (also saved as a PDF). This article by Sayer Ji was originally published on Green Med Info on 12 May 2013. The paper is Hematotoxicity of Bacillus thuringiensis as Spore-crystal Strains Cry1Aa, Cry1Ab, Cry1Ac or Cry2Aa in Swiss Albino Mice by Mezzomo et al, published first in November 2012 then again in March 2013 (more on that later).
The article states: “Bt toxins are capable of targeting mammalian cells, particularly the erythroid (red blood cell) lineage, resulting in red blood cell changes indicative of significant damage, such as anemia. In addition, the study found that Bt toxins suppressed bone marrow proliferation creating abnormal lymphocyte patterns consistent with some types of leukemia.”
Except, the article doesn’t actually say that at all. I have a few quick observations, then a rebuttal provided by our own Dr. David Tribe, and I wrap up this latest edition of whack-a-mole with a revelation. There’s a lot more to say about this paper, and hopefully we will learn more as time goes by, but this post is just a fast attempt to help people know that this is not a good paper to be pointing at.
Are organic Bt sprays dangerous?
The strangest thing about the rush by OCA and other anti-GMO folks to call this the latest “GMO” scandal is that the study didn’t test Bt toxins expressed by biotech crops. It didn’t even test the Bt toxins expressed in bacteria then purified. Instead, the paper says they used “spore-crystals Cry1Aa, Cry1Ab, Cry1Ac and Cry2Aa from B. thuringiensis var. kurstaki were obtained in lyophilized form”. Some Bacillus thuringiensis bacteria can go dormant, forming a spore, and form crystals around the outside of the bacteria to protect themselves.
In other words, this study used the whole bacteria – the exact same thing that is used in organic Bt sprays. According to Organic Consumers Association, Bt sprays are used by “at least 57 percent of organic farmers” and “does not have detrimental effects on mammals, birds or non-target insect species and microorganisms. In addition, Bt sprays leave no poisonous residue on crops or trees and are readily degraded into the environment.” Yet somehow the exact same active ingredient, when expressed in a transgenic plant, lasts a long time in the environment and is now toxic to everything, according to OCA et al (despite the pesky science that says it doesn’t).
Anyway, the authors resuspended the whole bacteria in water, then orally administered them to the mice in incredibly high concentrations: 27, 136, and 270 mg of bacteria per kg of body weight.
The most important thing that sticks out to me here is that there could be all sorts of other things about the bacteria besides the presence of the Bt toxin that could be affecting the mice. Since the negative control was water instead of culture medium with Bt bacteria that don’t express the toxin, we have no idea what it is in the solution that might have affected the mice.
Strangely, they never mention this and while they are careful to conclude “results showed that the Bt spore-crystals… can cause some hematological risks to vertebrates” rather than saying it was the Bt crystals themselves, they then conclude there is “increased risk of human and animal exposures to significant levels of these toxins, especially through diet”.
So, I have to wonder – are the authors worried about organic farming? To reiterate, genetically engineered plants have only the Bt toxin, not the whole bacteria, while organic farmers use the whole bacteria as a spray. Realistically the dose of “spore-crystals” from a Bt spray residue is much lower than what these researchers claim harm from, but still. According to the Pesticide Information Profile published by Cornell, Bt sprays show every indication of being safe (and so do Bt toxins themselves).
David Tribe’s Analysis
David Tribe wrote a great but short analysis over at Reddit and I wanted to make sure more people were able to see it:
- They used spore/crystals from modified Bt strains, rather than isolated protein. As per Schnepf, 2012a, such sporulating cultures are about ~20% Bt crystal proteins by dry weight. In other words, 80% of what the mice were fed is God knows what.
- The results seen are not due to use of too high a level of protein (270 mg/kg) , as there are reports of up to 5280 mg/kg of purified proteins fed to mice without any adverse effects (Schnepf, 2012b)
- The real negative control should have been wild type Bt strains, not water, as water lacks the other stuff floating around the culture medium
- The amounts fed the mice do not reflect human dietary levels– they were some ~ 106 to 108 time higher than exposure from GM or organic crops ( as per Hammond, 2012)
- The use of 3 mice/sex/group, as compared to OECD standards that call for 5 mice per sex per group.
- The historical incidence for these pathologies in the mice used was not reported. Thus, it is not clear if they are looking at treatment effects or natural biological variability for the mice
- The results did not all show the expected dose-response, with the lower dose resulting in the highest symptoms, as opposed to the other way around (eg, the 136 Cry1Ab had reduced MCH, but not the 270 mg/kg group)
- The reduction in MCV and RDW in all groups suggests a contaminant is involved rather than different Bt proteins
- For dose response, they get U shaped or inverse U shaped curves, a la Seralini. These could be a result of small animal numbers, not of hormesis.
- In the lit review, they say that Thomas and Ellar (1983) previously showed that cry proteins were hemolytic. However, Thomas and Ellar (1983) tested cyt proteins, not cry proteins. Even then, the cyt results were in vitro– there were no results when mice were fed the cyt proteins, presumably because they are digested.
- They also cite Aris and LeBlanc (Bt in mother and fetal blood) to support their work. (a rubbish paper)
- Their results are at odds with all the other studies whereby Bt (the bacterium) and cry (the bt protein) have not had adverse results in feeding studies.
Also — the journal is not a high quality journal.
I think we’ve established that this paper has some methodological issues. It’s not just us, though. The paper was originally published in Food Chem Toxicol, a generally reliable journal that has been around since 1982, and then withdrawn. Surely there is another story about that – Why was it published in the first place? Why was it withdrawn? Who requested the withdrawal? But the point is that the study as currently published is not in a “high quality journal”, as David puts it.
The Journal of Hematology & Thromboembolic Diseases is brand new, without any impact factor or history of quality to support it. This study appeared in issue 1, volume 1. The National Library of Medicine reviews scientific journals based on their scientific merit, strength of their review process, etc for inclusion in PubMed. OMICS, the parent company for this journal, is pointedly not indexed in PubMed, and has been dubbed “fake journals” by many scientists.
Why does the journal matter? High quality journals have a high quality review process. They seek out reviewers with the relevant expertise to do a proper science-based review of the methods, results, and conclusions. Low quality journals, simply put, do not.
Finally, I will end with a note from Karl Haro von Mogel, who is currently looking a little deeper into this paper’s curious history: “There’s something going on here that matters to everyone who cares about the peer review process that we depend upon to sift reliable studies from specious claims. Paired with rapid amplification through political advocacy, the reputation of science is at risk of being hijacked to obscure the truth.”