Seralini herbicides and adjuvants

By Grant Jacobs 22/02/2015 6

Gilles-Éric Séralini is best known for a research paper testing rats on GM feed and Roundup that was widely panned by the scientific community and eventually retracted.[1]

He has also tried to argue that it is other factors in Roundup (adjuvants) that causes toxicity, not the active ingredient glyphosate and that this calls for regulatory changes.

Below is a very brief selection of notes for others’ use. I’m not especially interested in this myself, odd as that might seem. I would much rather be writing about the really interesting things happening in genetics and genomics that feeling obliged out of public duty. I’m pooling these in case they are of use to those concerned about Séralini’s claims about adjuvants and testing, as there are a number of people who seem confused by the the claims made and of those I’ve seen, they seem unaware of these statements. With that in mind, please respect that this is not intended to be a formal breakdown of this idea, but a small collection of pointers others may wish to investigate for themselves.

Below linking to these sources I’ve given a potted account of a few general thoughts about this type of work, trying to outline for a non-scientist some things that might want to be borne in mind. I’d write something more cohesive, but I simply haven’t time. Better something that nothing, I hope?

First, other’s criticisms. In particular note these say that adjuvants are tested in Europe and are not considered ‘inert’ (i.e. contrary to Séralini’s implications/claims).

From the European Crop Protection Association (ECPA):

“… The authors’ direct exposure of in vitro cultured human cell lines to pesticide formulations circumvents the body’s most effective natural protective barrier, the skin, and does not reflect relevant in vivo exposure conditions which take into account the absorption, distribution, metabolism and excretion of a product within the body. Consequently the data presented in the publication are not relevant for the safety evaluation of pesticide products in relation to human health.

“Pesticide products, including the active substance and the pesticide formulations, are closely regulated in the European Union. No pesticide product can be authorised for sale in the EU unless data are provided and accepted by regulatory authorities showing that it does not pose an unacceptable risk to human health for the designated use. The assertion by the authors that industry or regulatory authorities assume pesticide co-formulants (called “inerts” in the publication) are non-toxic is incorrect.

“All of the pesticides studied in the publication have already been fully evaluated with in vivo toxicity studies. These studies confirm the absence of significantly increased toxicity of the pesticide formulation when compared with the toxicity of the active substance contained in those formulations. This in vivo data is publicly available and remains the most robust and relevant data upon which to assess to potential toxicity of pesticides. In addition, co-formulants, including “adjuvant” co-formulants, are specifically mentioned in the EU regulatory requirements for pesticide product approval and are part of the overall assessment of the potential risk of the use of a pesticide product.”

Coverage in Science magazine:

“Toxicologists have reservations about the study. “There are issues in terms of its design and execution, as well as its overall tone,” writes Michael Coleman, a toxicologist at Aston University in Birmingham, U.K., in an e-mail to ScienceInsider. “Anything is toxic in high concentration, the question is whether the toxicity is relevant to the levels of the agents we are ingesting. This paper does not seem to address this issue at all.” Martin van den Berg, a toxicologist at Utrecht University in the Netherlands, says the paper deserved to be reviewed. But he, too, questions the experimental design. “The endpoints observed are so general that we could probably find the same kind of toxicity with lemon juice or grapefruit extract,” he says. “It’s not new or shocking. It is what I would have expected at the level he is giving this to the cells.” Séralini dismisses the criticisms as biased. “I recognize the remarks of industry in that,” he tells ScienceInsider.”

(Note neither of the people quoted were from industry; they were from universities.)

From Val Giddings, consultant in science and regulatory policy relating to biotechnology innovations in agriculture and biomedicine:

He notes many issues with the science, also:

“One of the journal editors was so unhappy with the poor quality of the study and flawed peer review that he immediately resigned his post.”

“Contrary to the implications of this paper, all “inert ingredients” and adjuvants in pesticides require prior approval by EPA (see 40CFR (Part 180)).”

He’s got a fair bit to say as bullet points in response to quoted claims; perhaps a useful starter list to investigate.

More here:

Some quick comments of my own, intended for non-scientists

(These are general comments intended as pointers for those interested the topic, not a breakdown of the papers.)

There are a number statements by people concerned with his claims,[2] saying that adjuvants do nothing circulating. This is wrong and I strongly suspect arises from a poorly-worded sentence in Séralini’s abstract:

“Its adjuvants are generally considered as inert diluents.”

Read at face value this contradicts the definition of what an adjuvant is and claims them to be ‘inert diluents’. The problem is the word ‘considered’. Read in the abstract without reading further, it reads as defining what adjuvants are – but incorrectly.

Adjuvants are intended to assist the active ingredient to perform better. That’s what the term means.

Early in the introduction to the paper—if you can download it—you will find,

“adjuvants are developed at least to enhance their stability and penetration into cells”

The wording in the abstract is sloppy and this has created some very confused arguments. Adjuvants are not inert, they’re meant to aid the effect of the active ingredient.

This confusion has, in turn, led to some suggesting that the regulatory bodies must be ignorant of that and have overlooked it, that they don’t know that adjuvants contribute to the effect of a mixture. I’m not familiar with regulatory bodies, so I can’t comment on specifics of what they do or not in regulation, but I find it hard to believe they are unaware of what adjuvants are. I suggest people ask the regulatory bodies rather than speculate. (In NZ’s case, this might be the EPA.) Speculation doesn’t really do any good.

Amounts (dosages) matter. For an adjuvant to be an safety issue, adjuvants would have to both not been considered in the testing and to lift the effect of the active ingredient in non-target species (including us!) to a level that matters. If they lift the effect to a level that has no real effect, then the point is moot.

Everything is toxic at some amount. Even water. Yes, water can kill![3] Things aren’t really toxins or non-toxins; it is the dosage that is toxic, the amount. A point here is that you can’t talk about these things without considering the amounts involved, the word ‘toxin’ doesn’t mean anything without looking at the dosages involved.

I suspect the lack of formal complaint to Séralini’s claims in the research literature is simply that researchers in the field are already sick of Séralini and his claims (!), and that there has been limited previous public fuss about his claims about adjuvants. From others’ accounts his infamous ‘rat study’ weren’t the first time he’s been considered as offering poor science and researchers who are considered to repeatedly offer poor work (or worse) are usually ignored unless there is a public fuss that needs correcting. (There’s just simple pragmatism behind this: researchers’ time is limited & there are better things to do.)

I’ve skimmed the research paper. It’s not enough to comment specifically, but a few general points about this type of study might be worth considering by those trying to read about it.

Studies of cells in a dish can’t be extrapolated straight to effects on whole animals. It also follows that the study can’t be immediately argued as something to demand regulatory change.

The effect is to test cells from inside our bodies without considering how the substances would get to those cells in the body, or could, and in what amounts.

Directly exposing cells in a dish can make results misleading if talked about in terms of people or animals. (It also doesn’t consider processing of the food – washing, if the stuff breaks down over time, etc.)

The adjuvant is a detergent. Detergents dissolve the lipid part of the membranes that enclose the contents of cells. Bathe mammalian cells in a dish in detergent and they’ll break up if you keep at it. This is basic biology, high school level stuff.

It’s what is routinely done to get the contents out of cells – wash them in detergents. You want to extract DNA or other stuff from a cell? Start by breaking the cells using detergents. (And/or shearing the cells.) Researchers are very, very familiar with this; this activity itself is not a rare thing to be ‘discovered’.

I strongly suspect you’d get similar results using an equivalent concentration of household detergent, but we don’t seem to die from eating and drinking from washed dishes that weren’t rinsed enough. Never mind that shampoo, liquid bath soap and so on all have detergents. (This will want considering the dosages and biochemistry of the detergents, something I don’t have time for, but the general point should be considered.)

Adjuvants will appear to give a ‘large’ effects relative to the glyphosate in animal cells because glyphosate has little effect on animal cells. That’s a case of something being larger than a tiny number. The toxicity of glyphosate is very small (its toxic effect, in dosages that cause it, is specific to plant cells), so any effect by the adjuvants will appear to be large when applied directly to animal cells in an artificial setting.

When Séralini writes (paraphrasing) ‘greater than the active substance’ it can be read by non-scientists as if saying the glyphosate had an effect and the adjuvants had a higher-still effect. Glyphosate is reported as having little effect on non-plant cells.

Some of the controls needed to do this properly seem missing, to a brief reading. No comparing with non-starved cells for example. (Starved cells are used; these are already under respiratory stress, favouring poor outcomes under further stress.)

Readers are welcome to add to this points in the comments below, but please make these constructive. In particular, I would welcome comments from toxicologists. Please be aware that I may be unable to respond to comments over the next week.


1. The paper, largely unaltered, was republished with no further peer-review despite criticism in the research literature.

2. One looks to be a conference report. You’ll spot “ToxMix 2011: International Toxicology of Mixtures Conference. A selection of papers.” near the top. (These are a few notches down on a formal research paper. These aren’t peer-reviewed in the way that research papers but are, for most/better meetings, approved by the meeting committee. Their standard will depend on who was on the meeting committee.) Another is in a ‘pay-to-play’ journal, with the issues that can come with that. (For example, they can feature lax peer-review, in some even to the point of peer-review only in name i.e. none!)

3. This is off-topic, but that article is an interesting read, e.g.

Fatal water intoxication has been described in several different clinical situations. The most common of these is psychogenic polydipsia (compulsive water drinking), which is sometimes associated with either mental illness or mental handicap.4,5 The condition has also been described in young army recruits of good health who developed hyponatraemia after apparent overhydration following heat related injuries.

There are other articles on this, like this from Scientific American.

Other articles on Code for life:

Doggie ERVs

Rubella, not a benign disease if experienced during early pregnancy

Finding platypus venom

Coiling bacterial DNA

Preserving endangered species — of gut microbes

6 Responses to “Seralini herbicides and adjuvants”

  • Yeah, well done. I caught the starved cells trick way back in some other paper where they were trying to force Bt to kill cells in dishes. You could put BSA on the cells and they’d croak too.

  • For any potentially toxic chemical formula used in the environment and/or on/in our food, I want to know the toxicity of the full formula, not just one ingredient, even if that ingredient is said to be the main or active ingredient.

  • What Prof Seralini said at the Food Matters conference was that the full formula of Roundup is much more toxic than glyphosate on its own. The full formulae of any pesticides are what should be tested for toxicity, not just the active ingredient.

  • “the full formula of Roundup is much more toxic than glyphosate on its own”
    It’s unfortunate, then, that he didn’t actually test the full formulation. The adjuvants were tested & compared to glyphosate, but not to the actual full formulation.

    Another point, as Grant has said, is that in vitro tests won’t actually tell you a lot about how these substances might affect a whole organism.

    And this: “Anything is toxic in high concentration, the question is whether the toxicity is relevant to the levels of the agents we are ingesting. This paper does not seem to address this issue at all.”

  • Thanks Alison. Seralini and his team added the full formula of Roundup to the drinking water of one batch of rats in their two-year in vivo study, and found adverse health effects in those rats drinking the water with Roundup. They concluded more tests should be done.
    Do you know of other in vivo studies of pesticides using the full formula rather than the active ingredient?

  • From the ECPA quote: “No pesticide product can be authorised for sale in the EU unless data are provided and accepted by regulatory authorities showing that it does not pose an unacceptable risk to human health for the designated use.” – who provides the data, in the main? Companies making the pesticides? Independent scientists? Both?