[Front Page] [Features] [Departments] [Society Home] [Subscribe]

Glyphosate and Cancer

Jim Barrow

When I read in a recent newsletter of the Wildflower Society of Western Australia that there was a "clear link" between glyphosate and a form of cancer, I naturally went to the internet to see what was there. I entered the search words "glyphosate" and "cancer" and found a plethora of sites. After several hours of browsing, I came to two inescapable conclusions.

1. Glyphosate is the most dangerous chemical ever released into the environment: acutely toxic, carcinogenic, persistent, and yet mobile so that water is also contaminated.
2. Glyphosate is the most benign and beneficial chemical ever discovered: of low toxicity, non-carcinogenic, rapidly decomposed in soil, and immobile.

There didn't seem to be any intermediate position!

Almost the only aspect on which there was agreement was on its chemical formulation and mode of action. Its correct chemical name is N-(phosphonomethyl) glycine. Glycine is the simplest amino acid. It is the left side of the diagram. The name tells us that there is a methyl (CH2) stuck on the N atom of the glycine. Linked to that with a C-P bond is the phosphono group.

Glyphosate acts by interfering with a biochemical pathway which plants and bacteria use to manufacture certain amino acids. Animals don't have this pathway. They rely on ingesting these amino acids and are therefore not affected in the same way. Some plants can break down glyphosate. They chop the molecule on the left side of the N of the central line. Some soil organisms use this pathway. Others chop between the P and the C and this seems to be the favoured pathway because organisms capable of doing this build up in treated soil. Monsanto has found a way to introduce into plants a bacterial gene which produces an enzyme that is resistant to glyphosate - rather than causing the molecule to break down.

One of the problems in assessing the dangers of this, or other pesticides, is that the two sides view the situation differently. One side, the "cons", says: the surfactant used in Roundup is more acutely toxic than glyphosate itself. The other side, the "pros", says: glyphosate is even less toxic than the surfactant used in Roundup. Another problem is that the two sides of the debate use the same word but assign different meanings. So the cons say: glyphosate is acutely toxic. Which sounds dreadful. The pros say: yes that's true but this just means that one dose, if sufficiently large can kill. Acute toxicity tests have been performed on animals and it is a requirement of the US EPA that doses be increased until deaths occur. Based on these tests, the acute toxicity is low: you need a lot to kill with one dose. It seems bizarre, but there is also information on acute toxicity to humans - mostly obtained from suicide attempts. The lethal dose is about 200 ml - three quarters of a cup full of the concentrated reagent.

Other laboratory tests measure chronic effects. They require the feeding of sub-lethal doses to laboratory animals over a long period. The US EPA requires that the dose must be high enough to produce effects. If a tested dose is not high enough the study must be repeated at a higher dose. The animals are then killed and just about every tissue checked for damage and especially for cancer. Another problem with these studies is that so many abnormalities are looked for that something is bound to be statistically "significant". (If you look at say 20 abnormalities it is probable that one will be "significant" at 1 in 20 - the test used in much science). So you find an "increase" in one kind of tumour in one study but an increase in a different tumour in the next study - and so on. The effects are often small so that in one study a dispute about the existence of an extra tumour in the untreated group determined whether an effect was significant or not. The cons list all of the observed effects as caused by glyphosate. The pros say such variability is just what is to be expected when there is no real effect.

Does glyphosate cause genetic effects?

It depends who you believe. The pros say a large battery of tests showed no effect; the cons list half a dozen reported effects. Two things seem relevant here. One is the simple philosophical point that you can never prove no effect. It is always possible that you haven't looked in the right place. On the other hand, let me tell you a secret. There is a lot of poor science about! You would really have to spend a lot of time evaluating the several reports to see which ones seemed kosher.

It is accepted that glyphosate reacts strongly with soil - presumably via the P end of the molecule. This means that only a small proportion can be in the soil solution and so free to move - to leach into rivers or groundwater. However, here in Western Australia, we have some of the sandiest soils in the world. I would expect retention to be low in such soils, but I am not aware of any local measurements. On the other hand, the cons are worried that sorption is reversible. This merely means that unlike phosphate, the molecule does not bury itself inside the soil particle. This is not necessary to prevent leaching.

Soil micro-organisms are a pretty voracious lot and will usually learn to gobble up anything they can digest. And if supplied with something unusual the organisms that can deal with it build up their numbers. On some soils, glyphosate has been shown to break down very quickly - sometimes in a matter of days. On other soils it may take longer. Again there are two relevant points. One is that I am not aware of any study on breakdown rates in Western Australian soils. The other is that some reports of very long persistence arose because glyphosate labelled with carbon 14 was used and this meant that the isotope was detected in the crop - not the chemical.

Now we come to the big one.

Laboratory tests can only show whether a chemical seems safe. They can't anticipate all the possible effects in practical use - witness the effects of DDT on bird eggs. So the real test is whether there are problems in practice. There are two kinds of problems: short term "irritant" effects, and long term effects such as possible cancer. There are certainly reports of irritant effects with glyphosate but the slant on these reports differs widely. One site says that glyphosate is the most commonly reported cause of pesticide related illness in California - the USA state with the most comprehensive reporting program. Mostly these were reports of effects on eyes and skin but there were also reports of blurred vision, peeling of skin, nausea, headache, vomiting , diarrhoea, chest pain, numbness, burning of the genitals (!), and wheezing. (You do find yourself wondering about some of these.) Another site reports that 80% of the cases involved minor and reversible eye irritation - and points out that the major cause of eye irritation in agriculture is sulfur dusting. It claims that glyphosate is used on millions of acres by many people with very few incidents.

But what about cancer? This is of course the scary one. And some sites make it sound really serious. "Researchers in Sweden have linked pesticides to one of the most rapidly increasing cancers in the Western world.....The patients were 2.3 times more likely to have had contact with glyphosate". "Swedish study finds exposure to glyphosate .. increases risk for non-Hodgkin's lymphoma". And our own newsletter: "clear links between......glyphosate and non-Hodgkin's lymphoma". So I thought I'd better get hold of the original report and evaluate it myself.

The Swedish researchers thoroughly quizzed people diagnosed with one of the several forms of this cancer - or their relatives for those who had died. They matched the individuals with people chosen from the general population to provide controls, including a similar number of relatives of people who had died of other causes. They had 404 "cases" - all of whom had the cancer and 741 controls. Many people in both groups had had exposure to phenoxyacetic acids - the herbicides 2,4-D, 2,4,5-T, and MCPA. For these exposures, the score was 51 exposures out of 404 for the cancer group and 71 out of 741 for the controls. The "odds ratio" was 1.5, which they reckoned was pretty significant - especially as it was in line with several other studies. For glyphosate the numbers were 4/404 and 3/741. So the odds ratio is pretty large - the "2.3" quoted in some reports. But the numbers are very small: one case more or less would make a big difference. The researchers quite properly write that glyphosate "might be of concern" (italics mine). They go on to say it deserves further study; and who could disagree. But this seems to me to be rather a long way from much of the sensational reporting.

So what should you do?

There are two kinds of risk assessments. One is the risk/benefit assessment by the individual. This is the sort of thing we all do many times every day....Shall I drive the car today? Shall I have another slice of cream cake?......If we have a weeding job for which we might use glyphosate we can each assess the risks/benefits involved and make our own decision. And if we do decide to use it, then to be very careful of exposure. The other risk assessment occurs when someone else makes the decision. Usually we are less happy about this situation. We don't like others exposing us to some perceived risk because we don't feel in control. Here the only option available is to campaign against the perceived risk. If you feel this way, then there are appropriate organizations within which you can so campaign.

From the newsletter of the Wildflower Society of Western Australia, May 2000.

[Front Page] [Features] [Departments] [Society Home] [Subscribe]

Australian Plants online - December 2000
Association of Societies for Growing Australian Plants