A critique of the Seralini Study.

In debates about Genetically Modified Organisms ( More commonly known as GMOs. ) the study done by Séralini comes up fairly often.
For those unfamiliar, the study was done by a group led by Gilles-Éric Séralini that fed Roundup ( Here after referred to as Glyphosate when referring to the herbicide, and used as the market name of Roundup ready when referring to maize. ) Ready maize to rats over a two year period.
It was originally published in the journal Food and Chemical Toxicology in 2012 without peer review.  The published study, which cost in the neighborhood of $4,500,000 ( €3,200,000. ) to run, purported to show health risks associated with the corn itself, and the herbicide Glysophate.
Note 1The Seralini study, posted on his website.
http://www.gmoseralini.org/wp-content/uploads/2012/11/GES-final-study-19.9.121.pdf/Note 1
It was retracted in 2013 by the journal, after Seralini refused to withdraw it himself, due to the data not supporting his conclusions, and numerous methodological problems.
It was eventually republished in 2014 by Environmental Sciences Europe, without edit to the study or further peer review, though he ( Seralini. ) did release some more data than he originally did in the original publication.
I have been asked a number of times to explain what, precisely, the problems are with it and, after having written it out a number of times in various places, I decided to collect my thoughts in one location and give the various problems with the study individual attention.
The first problem is a conflict of interest.
Seralini's work, both the study in question and previous works, have been regularly funded by groups that have a vested interest in seeing the demise of GMOs, most notably groups such as Greenpeace.
Note 2There are some notable reasons why studies funded by Monsanto and various other companies that fund genetic research are a false comparison.
The most notable of which is the fact that it harms them in the long run to market something as safe when it isn't if the data shows that it is in fact unsafe.
When companies release a dangerous product to the public, they open themselves up to litigation due to negligence when it comes to light that it is harmful to the general public.  ( We can see this with the numerous class action suits filed against asbestos and cigarette manufactures when it was demonstrated that their products were harmful. )
If it can be demonstrated ( Which it can, quite easily. ) that the company knew about the danger before hand, that turns it from negligence into malevolent intent, which takes the potential litigation from a couple hundred thousand dollars into multiple million dollar suits.
It *will* kill a company if such a thing comes to light, period.  Not only in momentary loss due to the litigation and legal costs, but momentary loss due to decreased sales revenue and the extremely real possibility of extended prison sentences for various member of the company, ranging from executives to scientists to control managers.  ( And considering how cushy many of their lives are, any period of time incarcerated is bound to be hell. )
Not only will their professional lives effectively be ruined, their personal lives will as well, along with their company.
Further reasons it's a false comparison is that the company has to demonstrate to the scientific community at large, which features millions of people across numerous disciplines, that their product is safe.
Contrary to public opinion, this isn't an easy process.
The genome of the things being modified is very, very well mapped.  We know the genetic sequence of the things being altered, and the things being used to alter them.  We have mapped them.
It's why we are able to say that taking genes from a plant that normally causes allergies in some people won't cause an allergic reaction when transferred over to this new plant.
An example would be taking genes from soy that make it a hardy plant and able to thrive in a variety of soils and climates and transferring them over to maize to give it that hardy ability won't cause an allergic reaction in a person allergic to soy because the genes being transferred are ( Here I will be making arbitrary designations, as I am not familiar enough with the genetic sequence of either soy or maize to give the actual designations. ) A9, B081, and T62 which causes the root structure of the plant to be more resilient by causing the plant to secret a specialized protein, while the genes responsible for the protein that causes the allergic reaction are C67, B42, and U89.
This transfer will be noted and the scientists reviewing the data, due to having said information, will know precisely where to look and, consulting existing scientific literature will know that the specific genes and the protein in question aren't responsible for the allergic response will give it a green light.
From here, it will progress on the next stage, with often involves animal feed trials, most notable of which is mice due to their similar physiology to humans, rapid life span and high reproduction rates.  Other animals that will be fed the food will be pigs, cows, and a variety of other specimens, though the largest is by and far rodents, for reasons listed above.
All along the way the data will be freely available to the scientific community upon request, with extreme care taken to document every stage of development, with particular emphasis on controls.  ( Controls in science are everything.  Without them, science could not progress at all. )
/Note 2
Seralini's study did not follow any of these procedures.
Quite the contrary.  Seralini would not release his work, or commentary on his work, to anyone unless they agreed to publish *without* peer review.
Now, this doesn't necessarily mean his work is flawed in and of itself.  I can not think of any work that would require this, but I suppose it might be possible.
However, it is a very, very serious red flag in the scientific community, as peer review is second only to replication.  ( Peer review is so others can provide input that allows the work to possibly proceed in originally unsought directions, and to point out various methodological flaws in work that might otherwise go unnoticed or unaccounted for.  Replication is to ensure that the results aren't a one off fluke, or product of some hitherto unknown process. )
To the best of my knowledge at this time, no one in the scientific community has attempted to replicate his results.  The reasons are very, very compelling.
To put it short and sweet, his methodology was very deeply flawed.
The first, and greatest, flaw was the particular rat species he used.
He used groups of rats known as Sprague-Dawley rats.  A particular feature of this particular rat species is it's preponderance towards spontaneous benign tumors.
A study authored by J. D. Prejean, J. C. Peckham, A. E. Casey, et al for the Kellering-Meyer Laboratories, published in the scientific journal Cancer Research  issue 33, 2768-2773, in November 1973, and since replicated in various other studies which necessitated the discontinuation of Sprague-Dawley rats as a viable strain for general experimentation purposes, found that the species exhibited a 45% spontaneous benign tumor rate.
Female rates were twice that of male rates, with the majority of the difference lying in mammary tumors.  The most common place for tumors in both sexes, following the mammary tumors in females, was in the pituitary and adrenal gland in the brain.  Males suffered tumors in the urinary tract, while females didn't.
The two locations outside of the brain correlate to the images Seralini released of the tumors that formed during his study.
Note 3Study in question :http://cancerres.aacrjournals.org/content/33/11/2768.full.pdf
Pictures of rat tumors.  ( Possibly disturbing to some.  Caution advised. )http://www.gmoseralini.org/wp-content/uploads/2012/12/3-rats-seralini-e1354631752195.jpg/Note 3
The second major flaw lay in the diet he fed the rats, which tainted his findings.
Part of the diet of the rats he fed the GM corn he also gave them direct Glyphosate in their drinking water.
I can already hear you going, 'Wait a minute here, but this corn is sprayed with this stuff, so he needs to include it!'
Well... yes, during the growing it is indeed sprayed with the herbicide Glysophate.  However, what is very, very important to note, is that Glyphosate has two very specific conditions that make this a very different instance.
The first is that it does not bioaccumulate in animals.  Animals ( And humans are an animal. ) excrete any acquired Glyphosate in their urine and feces.
This is very important to note due to the fact that it isn't the chemical that harms you, so much as the dose.  Some chemicals are very dangerous, but the lethality of a particular chemical is entirely irrelevant if the dose is too small.
The second, and this is the most important one, is that round up has a very short half-life of, on average ( It varies based upon certain local conditions, though it remains relatively short. ) of 47 days.
Note 4What, precisely does this mean?
A half-life is a chemical term for when something is half the mass it was at the beginning, due to breaking down into more stable elements.
Glyphosate, like most herbicides, is typically applied during planting time.  This part it becomes a bit trickier to work out, as growing seasons vary by area and by crop.
Since this is dealing with maize, I will use the growing season of April to October, since that falls within the growing seasons of America, China, and the European Union, which accounts for 68% of the worlds annual Maize production.
Going from the end of April to October 15th, that is a total of approximately 168 days, which gives the half life 3 1/2 complete cycles.
That would reduce original amounts down to about 9% of the original total, provided the entirety of it remains there.
*However* something else to take into consideration is that not all of it remains there during that time.  Due to various environmental phenoma, such as rain, much of it is washed off.
Further, all crops go through multiple thorough cleaning stages between harvest and delivery to your local grocery store, so any potential residue that could have survived the entire growing season ( Which is very, very unlikely that any is there between the environmental factors and it's natural decay, but if it is it is in such infinitesimal amounts as to be negligible. ) will be cleaned off before it every gets to your grocery store./Note 4
Seralini's study did not take such things into consideration, and instead gave it to the specimens directly, which is a very big no-no, due to it's harmful properties when directly ingested in sizable quantities.  ( The water solution that used was 450 grams per liter, which is a very sizable quantity given the toxicity of large quantities of ingested Glyphosate. )
The third flaw ( Which some would argue is the greatest flaw, though I still attribute the greatest flaw as the selection of the particular strain of rats as it is directly responsible for this particular flaw. ) is the size of his groups, control and otherwise.
Seralini divided 200 animals ( 100 of each gender, which is something I can give him credit for as gender is very important when dealing with species such as Sprague-Dewley rats, though it only further undermines his claim to intellectually integrity in this. ) into 10 different groups.
6 of these groups were control groups ( Fed a non-GMO, non-Glyphosate diet. ) and the remaining 4 groups were divided into a different diet each, each containing different concentrations of GM corn, though all 4 consumed the same water with the Glyphosate dilution.
Why is this particularly concerning?
When dealing with statistics, you want to get as large of a sample pool as you can.  It's why when dealing with pharmaceutical testing, we typically use groups that number in the thousands.
Yes, you read that right, thousands.  Some of the larger groups can be tens of thousands.
Why do we do this?
Well, to give you a scenario, think of this :
Imagine breast cancer.  In Canada, it is estimated, based upon previous years numbers, and various trends such as population growth, previous diagnostic numbers, and trends, that this year 24,400 women will be diagnosed with it, and approximately 5,000 will die from it.
These numbers, while certainly not inconsequential ( I would put forward one person dying is one too many, for this and a number of other diseases, such as heart disease, which claims approximately 75,000 lives in Canada every year. )
What this means is that breast cancer will get diagnosed in 0.069% of the Canadian population, and kill 0.014% of the Canadian population this year.
Doesn't sound quite so scary now, correct?
If we were to look at a smaller sampling size of, say, 10000 people, to get somewhat comparable results, 6 people would be diagnosed with it and 1 would die of it.
Using Seralini's sampling sizes, if just one person were to be diagnosed with breast cancer of that group, we could conclude based upon that data that 1,744,000 would be diagnosed with breast cancer this year.
1,744,000 is more than a little different from 24,400.
About 71,000% different, in fact.
This lists some of, and in my mind the important ones worth discussing, problems with the Seralini study.
We live in a world where facts and fiction get blurred
In times of uncertainty you need journalism you can trust. For only R75 per month, you have access to a world of in-depth analyses, investigative journalism, top opinions and a range of features. Journalism strengthens democracy. Invest in the future today.
Subscribe to News24
Lockdown For
Voting Booth
Children between 12 and 17 can now be vaccinated against Covid-19. As a parent I am:
Please select an option Oops! Something went wrong, please try again later.
Relieved, it's been a long wait to this point
48% - 245 votes
A bit hesitant, and will wait to see how it goes
19% - 100 votes
Completely against my child getting the shot
33% - 170 votes
Rand - Dollar
Rand - Pound
Rand - Euro
Rand - Aus dollar
Rand - Yen
Brent Crude
Top 40
All Share
Resource 10
Industrial 25
Financial 15
All JSE data delayed by at least 15 minutes Iress logo
Editorial feedback and complaints

Contact the public editor with feedback for our journalists, complaints, queries or suggestions about articles on News24.