Is Glyphosate (Roundup) a simple weedkiller or big cause of cancer?
So unless you’ve been living under a rock for the past few days you might have noticed a recent US court ruling awarding damages of US$289m to a man with a terminal diagnosis of a blood cancer called Non-Hodgkin’s Lymphoma claimed to be caused by a common weedkiller called glyphosate, known to you and I as Roundup.
This has led to lots of media discussion about whether we should all be worried about this agent causing us all to die from cancer and whether the agents should be banned.
I had to prepare a brief for my colleagues in the cancer charity that I work in so I thought I’d use this info as the basis for an examination of whether our fears are well grounded or not.
The bottom line — for the millennial reader who needs an instant answer.
For the average food consumer no studies have found evidence of the levels of glyphosate around us or in our food or water causing any reason to be worried about increased risk of cancer.
The scientific studies on this agent are mixed and there is evidence both for and against cancer-causing claims.
What claims there are supporting this agent being carcinogenic indicate that you’d need to be exposed to very (and I mean very!) large amounts for long periods and, even then, the risk, if any, would be very small and associated with a very narrow set of uncommon cancers.
Agricultural workers and folks like groundskeepers etc., who could be exposed to larger amounts over the course of their working lifetime, might be advised, just in case, to follow the instructions for using this agent and wear appropriate protective equipment (like gloves etc) and wash and change after using large amounts to avoid any potential cancer risks.
There, that’s not so scary now is it?
Ok, so how did I and much of science arrive at these conclusions and where is all this coming from?
For the non-millennial & fact nerds among you (poor millennials, get such a hard time!!) here’s a bit of background.
Glyphosate (also known in the shops as Roundup) is a weed killer commonly used around the world. While many may think of weed killers for their landscape uses (stopping weeds growing on garden paths etc), weed killers (aka herbicides) are vital in food production as competition from weeds greatly reduce the yields of food crops and can lead to contamination of food with the products and seeds of weeds, some of which might not be good for us at all. Modern agriculture simply could not feed us all if it were not for agents such as weed killers, fertilisers etc. which give sufficient yields to produce enough crops for us all to eat at a price we can afford. Glyphosate, as one of the most common weed killers in use, plays a big role in that. There’s an interesting economic analysis here
Glyphosate has a number of attractive features for agriculture. It’s very effective, works against lots of different plant species. When compared to other pesticides its relatively non-toxic and easy to handle and it breaks down reasonably quickly in nature (thereby allowing crop plants to be sowed quite soon after initial application to remove pesky weeds)
So weed killers are arguably a necessary evil?
Well, yes but the amounts of glyphosphate produced and used are staggering and there is little question that in some parts of the world, lack of regulation and guidance leads to overuse, misuse and the potential for more ending up in our environment than needs to be. That’s not a good thing at all, no matter how safe the agent might be. It’s also important that we continue to find new ways to produce high quality crops with minimal impact on the environment.
Another complication is formulation. Glyphosate on its own doesn’t get into plants too easily so its made up as various salts and with other ingredients like surfactants which make it easier and more efficient to use. The available scientific evidence consensus though indicates that’s not as important as some might have you believe.
How does glyphosate work?
Simply put, glyphosate gets into plants (and some microbes) and blocks a key enzyme that is responsible for the production of certain amino acids only in these organisms. Humans and other don’t make use of this enzyme. The lack of these amino acids causes plant growth to stop rapidly and then there are knock on effects throughout the whole plant ultimately leading to it withering and dying. Here’s a really scientific description of that.
So, time for me to stop faffing around and come talk about the cancer stuff!!
Firstly we need to quickly mention that different levels of evidence and methods are used to try and work out the risks from substances we might worry about.
On the lab side we have mechanistic studies where pure chemicals interact with biological models we might have across a broad range of concentrations.
We can expose various lab animals to the substance and see if any cancers emerge.
We usually don’t experiment on human beings unless the substance we are testing for is a putative medicine so human studies of environmental agents normally rely on association studies. Even then, adult cancers can take decades to appear so in practice, even if it were ethical you simply couldn’t give a substance and wait 30–40 years to see if those people get more cancer than the rest.
To get around this a few different type of associative studies are undertaken. We can look at the general population and see if exposure to the agent in any way correlates to the emergence of cancer. Usually this would be expected to be specific forms of cancer. There is no way to accurately measure someone’s exposure over time, so the only work around is to ask people their recollection of how much they worked with, for how long etc. That’s a problem because, its hugely biased to perceptions and recollections; even then if a chemical caused cancer and we both worked with similar amounts, you might be the careful type and wore protective gloves etc. while I might be more cavalier and sloshed it around, giving me more exposure for using the same amount. There is a second type of study where we can ask patients with specific types of cancers to describe their exposure to chemicals as best they can remember and you can also look at their livelihood (so say a butcher wouldn’t be expected to have much exposure to weed killers but someone like a farm hand might, so is there a difference in cancer rates between the two careers?).
Now we know that roughly 6/10 cases of invasive cancer occur independent of environment (excess exposure to the UV in sunlight is always the biggest avoidable cause of cancers overall if you take into account non-invasive cancers). So, unless the agent causes quite a lot of a specific cancer (which glyphosate clearly doesn’t), it can be complicated to dissect the role and contribution of an agent we are interested in out of background cancer rates. So for example, those working in agriculture work outdoors, often get lots of sun and have higher rates of skin cancers. However, usually if something causes cancer, we see rises in very specific types of cancer, not in ALL cancers. Examples of this would be smoking-related cancers, where we see specific and dramatic rises in lung cancer cases in people who smoke regularly over decades of life.
Ok, so if you are still with me and I haven’t bored you to tears, lets summarise what has been found by all of these different studies.
The World Health Organisation supports a specialist group to make determinations on possible causes of cancer through its organisation called the International Agency for Research on Cancer (IARC).
One vitally important (and very OFTEN missed) point is that IARC classify carcinogens on how strong the evidence for them causing cancer NOT on how strong a carcinogen it is.
There is an excellent summary of many research findings about links between glyphosate and cancer in the IARC monograph which was used to make their determination for the agent being a class 2a, probable carcinogen here. I’ll try and summarise :-
Laboratory studies where the agent was fed in high concentrations to animals in their food or water have given mixed results. In a few animals some increases in specific types of cancers were found, however, in other animals no increases were found. Now that’s kinda confusing but its even more complicated in practice. Animals are different to humans (I touched on this revelation in my previous article) but can be metabolically very different to humans. This can mean that they get cancers when humans, with a different metabolism don’t, or, as happened with thalidomide testing in rabbits, humans get the effect while the animals don’t. Animals are usually exposed to concentrations of the agent, by food water or injection, that are vastly higher than human beings. While this speeds up the process of testing, toxicity and carcinogenicity are not linear functions. To give a practical illustration, if I take up to 4–6 g of paracetamol (acetaminophen if you live in the US) for an elevated temperature I’ll be fine. If I take 40g, it is very likely I will die from liver damage a week to two weeks later. One could summarise the animal studies collectively as being inconclusive but giving cause to be concerned that there could be a cancer risk at really high concentrations.
Under lab conditions, exposure of some (but not all) cultured human cells to high levels did show evidence of damage, however, this does not mean that the agent causes cancer, since several common chemicals, such as caffeine, can yield similar results, yet we know they do not cause human cancer. Again, this could give cause for concern though.
Useful studies in humans have largely looked at agricultural workers (who would be expected to have the highest exposure-risk profile) and looked at how many developed specific cancers and how this associated with their recollection of having used glyphosate. Again findings have been mixed with many studies finding no associations and some studies finding links. Where links were found the impact was small and associative studies merely show associations, they do not establish that one thing necessarily causes another (correlation is not the same as causation). So, for example, agricultural workers have proportionately high exposures to many different chemicals so it can be difficult to pinpoint which ones might cause an effect (and remember several key studies found no effect). The only evidence that comes out of the human studies is a potential small association between exposure and a specific cancer called Non-Hodgkins Lymphoma (NHL). NHL is an umbrella term for a collection of blood cancers so unfortunately even then its not as clear cut as we might like.
So perhaps one can see now why a “black and white” organisation like IARC might conclude from the evidence that at high concentration over prolonged periods, there are reasons to be concerned that glyphosate could cause specific forms of cancer. Remember, IARC merely assess the strength of evidence for carcinogenicity. So they classified it as a probable cause of cancer, meaning that if you were exposed to very high concentrations over long periods the risk of a specific form of cancer (NHL) might increase. (This classification group of putative carcinogens includes some medicines and practices such as regular night shift work by the way).
However, some other expert bodies have weighted into the discussions with probably more useful information. More recently the European Food Safety Authority (EFSA) weighted in with their expert opinion and they had two findings both of which indicated no particular worries about cancer being caused by the weed killer. But how can two different and highly respected agencies have divergent views?
The first finding is fairly uncontrovertial, essentially the EFSA found no evidence of cancer risks from glyphosate use among livestock. To be honest, that’s interesting but we’re probably more worried about ourselves getting cancer than “Larry the lamb”! As explained above, there are plausible reasons that sheep and cattle mightn’t get cancer where humans might. Their second other assessment though took things from a more detailed and practical human perspective. The EFSA found no compelling evidence for cancer being caused in human beings once levels in our environment and diet were kept below a reasonable threshold.
This threshold thing is really important. In toxicology (essentially the study of poisons) the biggest mantra is “the dose makes the poison”. Again, a topic I’ve mentioned before, because important principles become useful in lots of situations. My previous paracetamol example works again here too. Reminder, at normal doses, paracetamol is good at bringing my temperature down if I have say a viral infection. However, if I take too much, then it could kill me.
Likewise we might rationalise the apparently divergent opinions of the different bodies to say that the normal amounts of glyphosate in the diet and that workers etc are exposed to is not likely to raise the risk of any cancer. However, in a pure situation of high exposure for long periods, its “probable” from the limited models available that the agent slightly raises the risk of NHL cancer.
Separately and independently, The Food and Agriculture of the United Nations (also part of the WHO) also concluded that glyphosate in the diet did not cause cancer.
The FAO concluded that “ The overall weight of evidence indicates that administration of glyphosate and its formulation products at doses as high as 2000 mg/kg body weight by the oral route, the route most relevant to human dietary exposure, was not associated with genotoxic effects.”
To put it another way, based on their analysis, an average adult could eat 150g (6 teaspoons) of the stuff and not have to worry about cancer. Seriously folks, do not try this at home! Thinking about eating 150 g of weed killer just to prove me right or wrong, indicates a troubled mind but hopefully you get the point of what the experts are actually saying.
What can we take from this expert opinion?
For the average consumer of food or gardener, there is no risk of cancer being caused by glyphosate exposure.
For those working regularly, for long periods of their lifespan with large amounts and high concentrations of the agent (inc gardeners and agricultural workers) out of an abundance of caution, they should make full use of appropriate protective equipment (gloves, handwashing after use etc.) lest there might be any possibility of the agent increasing cancer risk.
Glyphosate may cause cancer but if it does so (and it’s quite a big if) the absolute risk is tiny, limited to a specific type of cancer and only in people who are exposed to large amounts for prolonged periods of their lifetime.
but I repeat, don’t eat 6 teaspoons of the stuff just because you disagree with me!
An important concluding thought:
With the continuous feed of media controversy over health, its easy to loose sight of genuine facts that give us the opportunity to make practical use of research findings and stack the odds in our favour. Glyphosate carcinogenicity is a theoretical concept, irrelevant to everyday life. We have cast iron proof of the role of many other behaviours on health risks on our population as a whole. If you really want to practically reduce your and your family’s risk of many different cancers this internationally agreed code is really the only proven intervention summarising all that we know. You can find a nice visual here.