A new study has been published that claims that the herbicide glyphosate is linked to an increased risk of non-Hodgkin lymphoma, a cancer of the lymph tissue. Of course, once a study like this hits the interwebs, everyone becomes panicked that glyphosate causes cancer.
I want to take a look at this paper because I am generally distrustful of any claims that “XYZ causes cancer!!!!!!!!!!!!”
Cancer myths are pervasive, and a lot of fear of cancer is based on those myths. So let’s take a critical eye and examine the peer-reviewed paper that claims that there is a link between glyphosate and non-Hodgkin lymphoma. There is a lot there, but there is also a lot of overreaction.
First of all, what is glyphosate?
Glyphosate (Monsanto’s Roundup and Ranger PRO) is a broad-spectrum systemic herbicide used to kill weeds, especially annual broadleaf weeds and grasses that are known to compete with commercial crops grown around the world.
It has several advantages over many herbicides in that it breaks down in the soil into non-toxic organic molecules, reducing or eliminating contamination of groundwater and lower soils. This is a large improvement over other types of herbicides.
Monsanto has developed genetically modified (GMO) corns that are resistant to glyphosate so that farms can apply the herbicide to kill the competitive weeds while not harming the crop. This allows farmers to suppress the weeds while allowing better production out of the grain crop. There is robust evidence that glyphosate has no effect on crop yield while reducing competing weeds at a lower cost than other modalities.
Whatever the benefits of Monsanto glyphosate, GMOs and the herbicide are tied together in many minds. And there is a “chemophobia” amongst many people that all chemicals are bad – glyphosate being the target of hatred for reasons.
Although maybe some fear of “chemicals” can be justified, we need to remember a key point – the dose makes the poison In toxicology, there are doses of any “chemical” that can be safe or unsafe. Water, for example, can be toxic if consumed at certain doses – sure, the dose is high for water toxicity, but it exists at a level that might surprise you.
Our culture’s chemophobia makes no sense to people, like me, who understand chemicals – every single thing we consume is made of evil chemicals, some with complex and indecipherable names.
Of course, as a result of this chemophobia, and misinterpreted scientific research, there has been an ongoing effort by many people to claim that glyphosate causes cancer. This, of course, has led to a Monsanto bashing across the internet along with several lawsuits. Although there have been a few robust studies that dismiss any link between glyphosate and cancer in general.
This article is agnostic about Monsanto – I just don’t care one way or another. All I care about is the quality of evidence that either supports or refutes the hypothesis that glyphosate is linked to non-Hodgkin lymphoma.
The glyphosate and non-Hodgkin lymphoma paper
The meta-analysis (considered to be the top of the hierarchy of biomedical research, and see Note 1) was published in Mutation Research/Reviews in Mutation Research, a moderate impact factor journal. Because of the use of glyphosate has increased significantly worldwide, including a nearly 16X increase in the USA alone between 1992 and 2009, exposure to the chemical may be widespread in food, water, and dust. So an examination of potential links to non-Hodgkin lymphoma (NHL) is important.
The analysis included six studies, which included over 65,000 participants, that met the inclusion criteria:
- One cohort study from the 2018 Agricultural Health study.
- Five case-control studies, one from the USA, one from Canada, two from Sweden, and one from France.
The key results were:
- The primary analysis showed a 41% increase in overall meta-relative risk (meta-RR) for NHL in individuals exposed to glyphosate.
- In a secondary analysis, the researchers showed a 45% increase in meta-RR for NHL in groups with high exposure to glyphosate.
The authors also laid out a compelling and strong case that shows the biological plausibility that glyphosate could be linked to NHL, such as being an endocrine disruptor which could be a causal factor for non-Hodgkin lymphoma. Biological plausibility is one of the necessary steps to move from whether correlation suggests causation or not.
Let’s be very clear with these results – they do not show a 41% increase with general exposure or a 45% increase with high exposure in relative risk to all cancers, just non-Hodgkin lymphoma.
If we look at these results without any critical analysis, we would assume that glyphosate is a dangerous carcinogen. But this is far from settled science, so let’s look at some important issues. regarding the study.
I’m going to list out my concerns below in a numbered list from what I feel is the least important to the most important. It makes it easier to pick out what may or may not be important to the analysis.
- All of this research is based on observational data, which makes it very difficult to conclude that correlation is equivalent to causation.
- There is limited available data. This analysis is based on just six studies spread over 15 years, which feels a bit weak.
- The study lacks a dose-response analysis, which is critical to understanding whether there is an increased risk of cancer with increasing dose. The authors specifically avoided this analysis, because there just isn’t enough data to include in a meta-analysis.
- The study doesn’t include several studies that have found no link between glyphosate and NHL. For example, this large study found no association between glyphosate and NHL (and other cancers). Now a meta-analysis can exclude studies, but with so few studies that were included, to exclude one that does not support their conclusion seems to be troublesome.
- The relative risk of cancer is a common term, but it does not tell a full story. Let’s take an example – the risk of getting hit by a meteorite is about 1 in 1.6 million, as rare as many of the 200 or more cancers. If you heard that your relative risk of getting hit by a meteorite increased by 45%, some of you might be hiding under your desk. Except that means that the absolute risk of getting hit by a meteorite is now about 1.45 in 1.6 million, meaning it’s still a low risk. The overall risk (in the USA) for non-Hodgkin lymphoma is about 19.7 per 100,000 people. If everyone in the USA was exposed to moderate levels of glyphosate and the relative risk, as shown in the article, was 45%, then the absolute risk for NHL is now about 28.6 per 100,000 people. What I’ve noticed is that a lot of people think that we have a 45% chance of contracting NHL from glyphosate, which isn’t the case. Now, many of us would say out loud that “I’d rather avoid being in that additional 8.9 per 100,000 who get NHL.” That’s a valid point, except let’s look at #6.
- There is a whole list of medical and environmental agents that may be linked to NHL. For example, age, family history, gender, ethnicity, geography, HIV, autoimmune diseases, chemotherapy, infections, body weight, and diet may all have an influence on the risk of NHL. So, if one is obese, male, and 65, what causes NHL, those medical conditions or glyphosate? Or is it some or all of it? These other factors are called “confounders,” and it does not appear that any of the research adjusted for that. They did control for other pesticides and herbicides, but that’s not the only causal factor for NHL. This is highly problematic with not only this study but with many “XYZ causes cancer” studies that I’ve read.
This meta-analysis provides a reasonable amount of evidence that glyphosate may be linked to non-Hodgkin lymphoma. That being said, it is not a completely convincing study. It is too small, it ignores contradictory published evidence, and it lacks robust separation of confounders that may be influencing the data.
Moreover, even if the study is accurate, the absolute risk of non-Hodgkin lymphoma is still small with or without glyphosate. Avoiding other risk factors for this particular cancer may be more critical than glyphosate.
Does that mean I’d want to drink a vat of glyphosate or bathe in it? No way. Do I think that agricultural workers should be protected from the chemical while working? Absolutely, since their exposure can be high.
As for the rest of us, who don’t drink or inhale large amounts, the amount found in foods and airborne dust (unless you live adjacent to agricultural areas, and even then, the data is weak) is probably not going to increase your risk of non-Hodgkin lymphoma or any other cancer. There are probably numerous other strategies one could undertake to reduce risks of cancer that might show higher levels of benefits.
So for now, I would like to see much more robust, powerful data that shows a link between glyphosate and non-Hodgkin lymphoma. This new study is interesting, but it does not rise to the level of conclusive.
- I keep stating that meta-analyses, meta-reviews, and systematic reviews are at the top of the hierarchy of biomedical research. It is because they combine data from a larger number of clinical and epidemiological studies, remove biases, give weight to higher quality studies, and provide a more powerful piece of evidence for science-based medicine. That being said, these type of studies are not infallible. They were not given to us by the Q continuum as dogma. They should, and must be, subject to reasonable criticism based on science, irrespective if they support or refute your beliefs. A good scientist should always treat all research, as long as it is within their broad knowledge base (so don’t ask me to critique quantum mechanics), with a good bit of skepticism until they feel comfortable with the hypothesis, the materials and methods, the statistical analysis, and the conclusions. It’s hard work.
- Andreotti G, Koutros S, Hofmann JN, Sandler DP, Lubin JH, Lynch CF, Lerro CC, De Roos AJ, Parks CG, Alavanja MC, Silverman DT, Beane Freeman LE. Glyphosate Use and Cancer Incidence in the Agricultural Health Study. J Natl Cancer Inst. 2018 May 1;110(5):509-516. doi: 10.1093/jnci/djx233. PubMed PMID: 29136183; PubMed Central PMCID: PMC6279255.
- De Roos AJ, Zahm SH, Cantor KP, Weisenburger DD, Holmes FF, Burmeister LF, Blair A. Integrative assessment of multiple pesticides as risk factors for non-Hodgkin’s lymphoma among men. Occup Environ Med. 2003 Sep;60(9):E11. PubMed PMID: 12937207; PubMed Central PMCID: PMC1740618.
- De Roos AJ, Blair A, Rusiecki JA, Hoppin JA, Svec M, Dosemeci M, Sandler DP, Alavanja MC. Cancer incidence among glyphosate-exposed pesticide applicators in the Agricultural Health Study. Environ Health Perspect. 2005 Jan;113(1):49-54. PubMed PMID: 15626647; PubMed Central PMCID: PMC1253709.
- Eriksson M, Hardell L, Carlberg M, Akerman M. Pesticide exposure as risk factor for non-Hodgkin lymphoma including histopathological subgroup analysis. Int J Cancer. 2008 Oct 1;123(7):1657-63. doi: 10.1002/ijc.23589. PubMed PMID: 18623080.
- Hardell L, Eriksson M, Nordstrom M. Exposure to pesticides as risk factor for non-Hodgkin’s lymphoma and hairy cell leukemia: pooled analysis of two Swedish case-control studies. Leuk Lymphoma. 2002 May;43(5):1043-9. PubMed PMID: 12148884.
- McDuffie HH, Pahwa P, McLaughlin JR, Spinelli JJ, Fincham S, Dosman JA, Robson D, Skinnider LF, Choi NW. Non-Hodgkin’s lymphoma and specific pesticide exposures in men: cross-Canada study of pesticides and health. Cancer Epidemiol Biomarkers Prev. 2001 Nov;10(11):1155-63. PubMed PMID: 11700263.
- Orsi L, Delabre L, Monnereau A, Delval P, Berthou C, Fenaux P, Marit G, Soubeyran P, Huguet F, Milpied N, Leporrier M, Hemon D, Troussard X, Clavel J. Occupational exposure to pesticides and lymphoid neoplasms among men: results of a French case-control study. Occup Environ Med. 2009 May;66(5):291-8. doi: 10.1136/oem.2008.040972. Epub 2008 Nov 18. PubMed PMID: 19017688; PubMed Central PMCID: PMC2728754.
- Zhang L, Rana I, Shaffer RM, Taioli E, Sheppard L. Exposure to glyphosate-based herbicides and risk for non-Hodgkin lymphoma: a meta-analysis and supporting evidence. Mutation Res/Rev Mutation Res; 10 Feb 2019. doi: