Talk:Glyphosate/Archive 20

Archive 15

←

Archive 18

Archive 19

Archive 20

Archive 21

Lede: update needed?

I noticed that the 2015 EFSA assessment is cited in the lede. Should this be updated to the EFSA's 2023 assessment? --Leyo 13:04, 9 November 2023 (UTC) PS. Also the EFSA-specific information in Glyphosate#Europe seems to be partially outdated.

True, also ECHA is outdated (confirmed 2022, that there is no change proposed to hazard classification). --Julius Senegal (talk) 13:57, 9 November 2023 (UTC)

True, thanks. Does anyone volunteer to draft a revised lede? --Leyo 10:25, 16 November 2023 (UTC)

Absolutely agree it needs updating. EFSA is the best source we have on this. I tried updating the lede two months ago but the environment was quite toxic so this will be a hard pass for me. Good luck to some brave editor out there. {{u|Gtoffoletto}}  ^talk 11:25, 16 November 2023 (UTC)

In depth source on the "Glyphosate case": is it true that regulatory agencies all agree?

Zaller, J.G., 2020. Daily Poison. Pesticides—An Underestimated Danger; Springer Nature: Cham, Switzerland, p.315. (published by Springer and accessible via the Wikipedia Library WP:WIKILIB) https://link.springer.com/book/10.1007/978-3-030-50530-1

The author is Associate Professor at the University of Natural Resources and Life Sciences Vienna with an h index of 43 https://scholar.google.de/citations?user=o08CivAAAAAJ&hl=en

It covers extensively the "Glyphosate case" examining the reasons why (quotes are from the source): A dispute has arisen over glyphosate in recent years especially after experts from the International Agency for Cancer Research (IACR) a division of the World Health Organization of the United Nations classified glyphosate as probably carcinogenic to humans (classification 2A). However, the EU authority EFSA and the German Federal Institute for Risk Assessment (BfR), the agency particularly responsible for glyphosate registration in Europe, and several other national agencies did not confirm this IACR classification.

And:

the US EPA considers glyphosate as “not likely to be carcinogenic to humans.” EPA asserts that there is no convincing evidence that “glyphosate induces mutations in vivo via the oral route.” IARC concludes there is “strong evidence” that exposure to glyphosate is genotoxic through at least two 1 What Is the Problem? Pesticides in Our Everyday Life 71 mechanisms known to be associated with human carcinogens (DNA damage, oxidative stress). How can it be that national agencies and IARC reach such different conclusions?

I think we should include a similar overview in the article. Multiple important sources are cited highlighting three main issues for why EPA/EFSA and IARC reached diametrically opposed conclusions on glyphosate genotoxicity:

• The selection of studies considered in the review
• The distinction between the exposure to the active ingredient glyphosate, which occurs via residues in food and feed, and occupational exposure to the formulated glyphosate-based herbicide (people who apply the pesticide are more exposed and have sued Monsanto after getting sick)
• The many forms of glyphosate that are used and the different compositions of the final herbicide: glyphosate and the commercial product are not the same and yield different toxicity results

We have different important bodies (EPA/EFSA/IARC) reaching different conclusions. We should probably report the conclusions of all while describing accurately the ongoing debate.

For example in the lede we state: The consensus among national pesticide regulatory agencies and scientific organizations is that labeled uses of glyphosate have demonstrated no evidence of human carcinogenicity but this is false and contradicted by the next sentence that follows: In March 2015, the World Health Organization's International Agency for Research on Cancer (IARC) classified glyphosate as "probably carcinogenic in humans" (category 2A) based on epidemiological studies, animal studies, and in vitro studies.. This is unclear and confuses reader. Obviously not all agencies agree here. We need to describe the "Glyphosate case" in more detail. {{u|Gtoffoletto}}  ^talk 12:04, 19 October 2023 (UTC)

Oppose.

First of all Johann Zaller is more an activist not a simple key opinion leader in that field, the BOKU he is working at is in touch with much mumbo jumbo junk science and was repeatably criticized for that. The h index does not grant him credibility (argument from authority). Zaller tried to bring his POV in German speaking Wikipedia and failed fortunately. He is so awkward (books like "Daily Poison: Pesticides") that no one would ever consider him reputable.

So his article does not say anything useful and cannot be used.

Even worse is your misunderstanding what IARC stands for and what the other national regulatory agencies / scientific orgas are doing. They are not the "same". There is therefore no contradiction, IARC's purpose is just a complete different one from that of the others. In addition, IARC's decision itself is highly controversial btw.

So no, please get familiar with our rules and the topics we are discussing about. --Julius Senegal (talk) 13:28, 19 October 2023 (UTC)

A citation is needed for all of your claims. Do you have them? {{u|Gtoffoletto}}  ^talk 10:12, 24 October 2023 (UTC)

Nice try. First you have to prove that a single particular author with a scorcher is reliable acc. to our rules (e. g. WP:FRINGE, WP:DUE, WP:MEDRS). Do you have it? --Julius Senegal (talk) 10:57, 24 October 2023 (UTC)

"Nice try" of what? As I have stated above the author is highly reputable professor from what I can see and the publisher is highly reputable. That book has 34 citations according to Google Scholar: https://scholar.google.com/scholar?cites=3614147560922018039&as_sdt=2005&sciodt=0,5&hl=en On the contrary, your challenge is totally unsubstantiated at the moment. {{u|Gtoffoletto}}  ^talk 13:16, 24 October 2023 (UTC)

Thank you very much, you have confirmed me - unintentionally that is.

If you looked on those citations you would have noticed that vast majority of citations are from Zaller papers (party Zaller is not the first author, but still). So Zaller likes citing himself - made my day. And the few other non-Zaller citations only using this book as a side note for non-glyphosate statements (and only once ofc).

Taken together, this book is only interesting for Zaller himself. --Julius Senegal (talk) 06:53, 25 October 2023 (UTC)

It is pretty common for authors to cite their previous work. So your other statements are your own opinions? {{u|Gtoffoletto}}  ^talk 22:47, 29 October 2023 (UTC)

No, it is not pretty common for reputable scientists to cite their own book that nearly no one is interested in.

But thank you again to strengthen my point. --Julius Senegal (talk) 12:00, 2 November 2023 (UTC)

Once again: you have not proven that Johann Zaller is not reputable. All the evidence posted here prove otherwise (college associate professor, published author, high h-index etc.). I will retract the source if sources confirming your claim exist. {{u|Gtoffoletto}}  ^talk 12:36, 2 November 2023 (UTC)

Once again: you have not proven that Zaller, especially this "book" is reputable. The "evidence" you have provided shows clearly that his book is not an established source in the scientific field. If Zaller himself desperatly needs this book to cite himself and to boost his h-index - tough luck. If you cannot show this, this discussion is not worth my time (since you fail to understand what the difference between IARC and the national agencies is).--Julius Senegal (talk) 12:43, 2 November 2023 (UTC)

Is sourcing independent? And dietary carcinogenity vs. overall Carcinogenicity

We currently source the sentence: The consensus among national pesticide regulatory agencies and scientific organizations is that labeled uses of glyphosate have demonstrated no evidence of human carcinogenicity. with [1]. The corresponding author is Jose V. Tarazona of the European Food Safety Authority. So that would not be an independent source in assessing why IARC and the EFSA have reached different conclusions.

From the abstract of that paper (emphasis mine): ...the International Agency for Research on Cancer (IARC) concluded in March 2015 that it is probably carcinogenic. The IARC conclusion was not confirmed by the EU assessment or the recent joint WHO/FAO evaluation, both using additional evidence. Glyphosate is not the first topic of disagreement between IARC and regulatory evaluations, but has received greater attention. The review concludes that the differences were due to: Use of different data sets, particularly on long-term toxicity/carcinogenicity in rodents, could partially explain the divergent views; but methodological differences in the evaluation of the available evidence have been identified. that would be completely in line with what is claimed by the source I have presented above and also does not support the statement we include in the lead. They also cite WHO/FAO that concluded: the Meeting concluded that glyphosate is unlikely to pose a carcinogenic risk to humans from exposure through the diet.[2]. WHO/FAO confirmed dietary exposure is not a problem but was not assessing the overall carcinogenicity of Glyphosate.

Another source compares the EPA and IARC conclusions [3] saying: EPA and IARC reached diametrically opposed conclusions on glyphosate genotoxicity for three primary reasons: (1) in the core tables compiled by EPA and IARC, the EPA relied mostly on registrant-commissioned, unpublished regulatory studies, 99% of which were negative, while IARC relied mostly on peer-reviewed studies of which 70% were positive (83 of 118); (2) EPA’s evaluation was largely based on data from studies on technical glyphosate, whereas IARC’s review placed heavy weight on the results of formulated GBH and AMPA assays; (3) EPA’s evaluation was focused on typical, general population dietary exposure assuming legal, food-crop uses, and did not take into account, nor address generally higher occupational exposures and risks. IARC’s assessment encompassed data from typical dietary, occupational, and elevated exposure scenarios. More research is needed on real-world exposures to the chemicals within formulated GBHs and the biological fate and consequences of such exposures.

Basically our blanket statement "labeled uses of glyphosate have demonstrated no evidence of human carcinogenicity" is misleading. There is broad agreement that Glyphosate is not genotoxic through the dietary/oral route that would interest most consumers and the public. But carcinogenic risks have been identified and potentially affect other exposure routes (e.g. occupational exposure etc.).

Also, several studies have claimed that US and European regulatory agencies have relied excessively on industry sponsored studies ([4] In the industry documents, we found that members of the EFSA and the EPA panel were in contact with Monsanto representatives, and included Monsanto-funded studies in their analysis, while IARC did not include industry-funded studies and deterred assessment committee members from private communication with chemical industry representatives..

We should be more careful/more precise here and improve our sourcing to ensure independence of our sources while presenting the complete views of the highest level assessments (WHO, FAO, IARC, EPA, EFSA, ECHA, others?). Thoughts? {{u|Gtoffoletto}}  ^talk 13:33, 2 November 2023 (UTC)

Claiming now that EFSA or other agencies are biased is becoming getting ridiculous.

Ofc you are lacking to provide strong evidence for this strong claim.

This is some sort of a sole repeatment of this this discussion thread, please read WP:POVPUSH and follow it (not only on this page). --Julius Senegal (talk) 14:39, 2 November 2023 (UTC)

The sources I provided claim it. Not me. Please focus on the content instead of constantly attacking me per WP:FOC {{u|Gtoffoletto}}  ^talk 17:53, 4 November 2023 (UTC)

Please stick to our guidelines as several times indicated.

"The sources" are just single papers from unknown authors, so this is a clear violation of WP:BALANCE.

I will not discuss this unfounded claim any further acc. to WP:FRINGE. --Julius Senegal (talk) 18:47, 4 November 2023 (UTC)

If you refer to the Hendlin source published on Environmental Epidemiology: Dr. Yogi Hale Hendlin is an environmental philosopher and public health scientist. Hendlin is assistant professor at the Erasmus School of Philosophy, and core faculy of the Dynamics of Inclusive Prosperity Initiative at Erasmus University Rotterdam, as well as research associate in the Environmental Health Initiative at the University of California, San Francisco[5]. So definitely not unknown. {{u|Gtoffoletto}}  ^talk 18:58, 4 November 2023 (UTC)

Sources on link with Non-Hodgkin lymphoma

Strengthening my fears I have just removed this source [6] by Geoffrey Kabat ("mainly known for the discredited BMJ study funded by the tobacco industry"...Yikes) that we were using to cast doubt on a 2019 meta-analysis on high cumulative exposures (e.g. occupational exposure) we have in the article that found Overall, in accordance with findings from experimental animal and mechanistic studies, our current meta-analysis of human epidemiological studies suggests a compelling link between exposures to GBHs and increased risk for NHL.. This article needs a thorough and level headed review. {{u|Gtoffoletto}}  ^talk 13:56, 2 November 2023 (UTC)

And @Bon courage just reverted the removal (seriously!?) [7]. Can you justify how that content does not egregiously violate WP:MEDRS and WP:DUE? {{u|Gtoffoletto}}  ^talk 14:00, 2 November 2023 (UTC)

Err what? How does MEDRS disqualify this source, which is a meta-analysis in a high-quality journal. Per WP:BLP you need to be pretty careful throwing words like "disgraced" around on Wikipedia. Bon courage (talk) 14:01, 2 November 2023 (UTC)

Sorry I meant "discredited" as a direct quote from the lede of our entry on Geoffrey Kabat (I've amended my comment above). Are you sure you have reviewed my edit correctly? You may be conflating the source I removed (a "meta-analysis" by Geoffrey Kabat [8] that no longer holds any affiliation as far as I can tell) with the high level meta-analysis he was trying to discredit [9]? {{u|Gtoffoletto}}  ^talk 14:13, 2 November 2023 (UTC)

The paper was discredited, not the person. That somebody wrote one bad paper 20 years ago does not discredit every other piece of work they have ever produced. MrOllie (talk) 14:15, 2 November 2023 (UTC)

There are (very rare) cases on Wikipedia where the very name of an author calls into question the probity of a publication, but it would need some super-strength sourcing, particularly given the strength of this source. Doubling down on the BLP issue is not good. Bon courage (talk) 14:17, 2 November 2023 (UTC)

I will stick to the direct quote. Thank you for pointing out the issues @MrOllie and @Bon courage. My mistake out of haste. {{u|Gtoffoletto}}  ^talk 14:23, 2 November 2023 (UTC)

There are definitely huge WP:MEDRS issues with this source. From our entry on Kabat: "He is mainly known for the discredited BMJ study funded by the tobacco industry, that failed to find an association between secondhand smoke and health problems." {{u|Gtoffoletto}}  ^talk 14:30, 2 November 2023 (UTC)

Which bit of MEDRS? The relevant text is "Editors should not perform detailed academic peer review". The article you removed was peer-reviewed in a high-quality journal, but you are saying you - a random editor on Wikipedia - knows better because of something you read on Wikipedia about one of its authors. Sorry, that's not the way it works. If there is sourcing that casts this work in doubt, you need to produce it. Bon courage (talk) 14:34, 2 November 2023 (UTC)

I must admit I didn't even read the source after I saw: the author had no affiliations, was involved in the tobacco lobbying scandal and currently supports unscientific views such as claiming second hand smoking doesn't cause cancer. I (maybe hastily) assumed WP:MEDBIAS would be quite clear and uncontentious here. I guess I was wrong. MEDBIAS suggests Obvious or overt bias in a source is a difficult problem for Wikipedia...It may be simpler to find a "better" source – either a higher quality study type or a more specific source instead (see WP:MEDASSESS). If no high-quality source exists for a controversial statement it is best to leave it out; this is not bias. Can we verify we are accurately reflect[ing] current knowledge on this subject? Do we have other "similar level" sources on the link with NHL? Might be better to use those. {{u|Gtoffoletto}}  ^talk 17:51, 4 November 2023 (UTC)

So apart from much text you are producing here you don't have any proof that this meta analysis is fringe or biased? Guilt by association is clearly not enough.

We have also the EFSA review not indicating any concerns regarding non-Hodgkin lymphoma. We have also other reliable sources, indeed echoing the authors, e. g. PMID 32096774.--Julius Senegal (talk) 18:54, 4 November 2023 (UTC)

I think we should probably consider rewriting this entire section on human carcinogenicity. We give the impression that debate on a possible link between Glyphosate and NHL is under investigation while it seems most recent sources concur that the evidence is inconclusive? We seem to be giving WP:UNDUE weight to a study that found a link with NHL. And we counterbalance that statement with a questionable source while better sources exist (such as the EFSA or other meta-reviews/analysis). {{u|Gtoffoletto}}  ^talk 19:29, 4 November 2023 (UTC)

No, not inconclusive. Reliable sources do not see any association (at least if used as indicated). Otherwise this would have been issued, i.e. in the recent EFSA review and especially by IARC. --Julius Senegal (talk) 19:55, 4 November 2023 (UTC)

What do you think are the best sources for this statement and what do they say exactly? {{u|Gtoffoletto}}  ^talk 13:00, 6 November 2023 (UTC)

I've found this 2022 Meta-Analyses (Note:after publishing I see ClinMed International Library appears on https://beallslist.net/ so I guess this source is out) that includes the disputed Kabat paper as well as the 2019 review by Zhang et al to conclude: Confidence in potential a causal relationship between glyphosate exposure and NHL was considered low.. It might be better than what we have unless we can use EFSA 2023 directly. We probably need to clean up this section and rely on newer and high level sources per WP:MEDASSESS.

Also we should maybe provide some "history" on the process that led to this assessment on Glyphosate's carcinogenicity so that readers can understand how it came to be. Something like (I'll attempt an extreme summary here with what I've seen so far):

1. Some initial evidence appeared to indicate a potential link with cancers (e.g. with high exposure, animal experiments) (Main sources?)
2. Scientific debate, conflicting assessments by agencies and the IARC 2015 case (Main sources?)
3. Current consensus that there isn't a link (Main source: EFSA 2023?)

{{u|Gtoffoletto}}  ^talk 10:53, 9 November 2023 (UTC)

Yes, a predatory journal and so completely useless for us. If editors want to avoid the Kabat paper, PMID:34142676 is a (slightly) more recent meta-analysis, in a reputable journal. I have not read it yet. (Add: I have now read it and in fact added it to the article to replace the Kabat paper. See what you think.) Bon courage (talk) 11:19, 9 November 2023 (UTC); amended 11:58, 9 November 2023 (UTC)

I have added a very recent meta review and of course a report of the EFSA, which is reputable.

However, I have removed this weak maybe association (doi:10.23749/mdl.v112i3.11123}) as "annot be ruled out" is - sorry to say that - a useless phrase (see also FRIN).--Julius Senegal (talk) 12:15, 9 November 2023 (UTC)

I think we are going in the right direction. I prefer @Bon courage's edit as it closely follows the conclusions of the study. However one point isn't very clear.

Boffetta et al 2021 was an update on a previous work (Donato et al 2020) after they received some strong criticism here Rana et al. 2020

In Rana et al. they discuss (after finding several issues in the Donato 2020 paper) (emphasis mine): One interesting question posed by this endeavor is whether or not GBH contributes to risk of all NHL, or a particular subtype. DLBCL is the major subtype (~25%) of NHL ... GBH may be more strongly associated with DLBCL than other NHL subtypes; 2) if true, analyzing NHL as a whole would attenuate potential associations, which is exactly what was reported [3]; and, 3) as a result, our meta-RR of NHL would be further underestimated. They conclude that: our calculated meta-RR of 41% increase in NHL among GBH-exposed workers is very likely underestimated, with recent studies indicating that subtype (DLBCL) specific associations may be stronger than the heterogeneous class of disease (NHL) as a whole..

Diffuse large B-cell lymphoma (DLBCL) is the most common subtype of NHL (but this isn't clear in Bon courage's edit and it reads as if it is a different kind of cancer). For this subtype Boffetta 2021 found: The meta-RR for diffuse large B-cell lymphoma (DLBCL) was 1.29 (95% CI 1.02-1.63; 4 studies) suggesting a modest, potentially significant increased risk but noted that the result for DLBCL deserves attention, despite the fact that it would not reach the canonical level of statistical significance once the p-value is adjusted for multiple comparisons. I think Bon Courage was right in including the last sentence on DLBCL as the study concludes:

• lack of an association between exposure to glyphosate and risk of NHL overall = no association if you look at overall NHL rates
• although an association with DLBCL cannot be ruled out. = but there might be a potential association if just the DLBCL sub-type of NHL is considered.

We may need to express this point more precisely to correctly explain to readers what was found. How about adding something like: However, a modest association between glyphosate and diffuse large B-cell lymphoma (a common form of non-Hodgkin lymphoma) cannot be ruled out. {{u|Gtoffoletto}}  ^talk 21:17, 9 November 2023 (UTC)

Nope.

You also cannot rule out that there is somewhere cheese on the moon. If there was an association regarding DLBCL (!) then the authors of that particular paper would have stated it. But they didn't, so there is no association.

Besides of that the EFSA (also in newer publications) clearly state:

The available epidemiological studies currently do not provide sufficient indication that glyphosate exposure is associated with any cancer related health effect. ANY cancer related health effect, i.e. also DLBCL. The EFSA review is the highest level of publication we can use currently. --Julius Senegal (talk) 07:34, 10 November 2023 (UTC)

I agree with Julius Senegal's logic, but there might be an argument to say that we could just mirror what the source says, daft as it is. Overall, I don't feel strongly either way. Bon courage (talk) 07:41, 10 November 2023 (UTC)

Well, as pointed out: EFSA is newer and of higher quality (see also Peer Review Report on Glyphosate (AIR V) - Part 3 of 6 from July 2023). Unconstructive FRINs (cannot be ruled out) do not help and this single statement of a single paper contradicts WP:WEIGHT. --Julius Senegal (talk) 08:37, 10 November 2023 (UTC)

I agree with @Bon courage we should stick to what the sources say as closely as possible. And I agree with @Julius Senegal that EFSA is of higher level per WP:MEDASSESS. I would add the sentence I proposed as well as a the sentence that Julius is quoting just after it with a direct citation to the EFSA. They are compatible statements: we currently don't have "sufficient indication that glyphosate exposure is associated with any cancer related health effect" but we also cannot exclude an association with DLBCL at the moment as there is some evidence that there might be a modest association with high exposure. {{u|Gtoffoletto}}  ^talk 11:04, 10 November 2023 (UTC)

Something like: However, a modest association between glyphosate and diffuse large B-cell lymphoma (a common form of non-Hodgkin lymphoma) cannot be ruled out. According to the 2023 EFSA review of glyphosate, the available epidemiological studies currently do not provide sufficient indication that glyphosate exposure is associated with any cancer related health effect {{u|Gtoffoletto}}  ^talk 11:06, 10 November 2023 (UTC)

There is no association, why do you stick to that and especially claim sth about now "modest" (!) association?

You even agree on that EFSA is the best meta-meta-review ("highler level" than other meta reviews). So there is no need for using wiggling phrases like "cannot be ruled out". If the best source state there is no association, then adding such statesments is a clear violation of WP:CHERRYPICKING and WP:WEIGHT. Since DLBCL is a from of NHL this is even contradictive. --Julius Senegal (talk) 13:20, 10 November 2023 (UTC)

It isn’t as simple as “no association” as we have been discussing. I’m trying to stick as closely as possible to the sources as also suggested by Bon Courage. “Modest” (as in “small”) is an addition by me for context but I agree we could get rid of it. The conclusions of those high quality reviews are not contradictory. What does @Bon courage think? {{u|Gtoffoletto}}  ^talk 19:10, 12 November 2023 (UTC)

My view hasn't changed. Bon courage (talk) 07:48, 13 November 2023 (UTC)

"Addition by me" vs. "as closely as possible" - funny. What high quality reviews? You haven't shown why this one particular review with the weasel statement is of high-quality, or even of higher quality. In doubt EFSA >>> all other revies. --Julius Senegal (talk) 09:13, 13 November 2023 (UTC)

@Julius Senegal we have discussed the sources at length in this discussion. They are linked above (Boffetta 2021, Donato 2020, Rana 2020 etc.). I see you have edited Bon courage's edit [10] to remove the source discussed here (Boffetta 2021) and have instead included a new source [11] in addition to the EFSA. Is this new review better than what we had and if so why? I see it is more recent but I don't know the quality of the journal and if it should be given more weight. The current text also appears to me to be suffering from WP:SYNTH issues as it is mixing two sources together. Also by removing the source used by Bon courage the text does not follow the source closely at all anymore.

Also worth mentioning that the EFSA source (https://www.efsa.europa.eu/sites/default/files/2023-01/glyphosate-peer-review-minutes-nov-dec-2022.pdf) states: The available epidemiological studies currently do not provide sufficient indication that glyphosate exposure is associated with any cancer- related health effect. Open point: RMS to consider in a revised RAR the study by De Roos et al. 2022 identified after the public consultation period (see experts’ consultation 2.37): Herbicide use in farming and other jobs in relation to non-Hodgkin’s lymphoma (NHL) risk. Occupational and Environmental Medicine, 79(12), 795-806.

The 2022 study by De Roos et al https://oem.bmj.com/content/79/12/795# concludes that Most of the herbicides examined were not associated with NHL risk. However, associations of phenoxy herbicides and glyphosate with particular NHL subtypes underscore the importance of estimating subtype-specific risks.. So both the EFSA and this additional study seem to mention the fact that certain NHL subtypes have shown an association although the evidence is insufficient.

I would go back to something closer to Bon Courage's edit [12] plus the EFSA:

Although some small studies have suggested an association between occupational exposure to glyphosate and non-hodgkin lymphoma, the association has not been demonstrated in more robust studies.[13] According to the 2023 EFSA review of glyphosate, available epidemiological studies currently do not provide sufficient indication that glyphosate exposure is associated with any cancer-related health effect.[14]

I think this is clear while not saying anything more than what the cited sources say. Also by separating and attributing clearly the statements it removes the WP:SYNTH issues. I have also made it clear that we are talking here about "occupational exposure" only and not risks to the general population. {{u|Gtoffoletto}}  ^talk 01:00, 16 November 2023 (UTC)

"The available epidemiological studies currently do not provide sufficient indication that glyphosate exposure is associated with any cancer- related health effect"

That is the take home here. Not your continuing attempt to make something out of nothing. Nangaf (talk) 06:53, 16 November 2023 (UTC)

@Nangaf if you read more carefully what I wrote, you would see that it is exactly the text I am proposing to add to the article. {{u|Gtoffoletto}}  ^talk 10:20, 16 November 2023 (UTC)

I went ahead and added the full EFSA quote to the paragraph: [15] {{u|Gtoffoletto}}  ^talk 11:14, 16 November 2023 (UTC)

@Gtoffoletto: Three people have said they disagree with adding this, so why did you go ahead and add it? SmartSE (talk) 12:22, 16 November 2023 (UTC)

@Smartse sorry for the lack of clarity (my fault I guess, but the fact that the text keeps changing doesn't help). All editors involved in this discussion agree on the quote from the EFSA I added (so I assumed simply adding it would not be a problem.) I'll try to clarify further my edit:

This is the text we currently have in the article:

Although some small studies have suggested an association between glyphosate and non-hodgkin lymphoma, subsequent work confirmed the likelihood this work suffered from bias, and the association could not be demonstrated in more robust studies. Sourced to this EFSA link https://open.efsa.europa.eu/study-inventory/EFSA-Q-2020-00140

Problems:

• That text was written by Bon courage based on this study [16] (it is almost a direct quote) but that source was removed and changed
• The EFSA never said anything about bias etc. at all so it is now completely unsupported/misattributed
• The current link is a generic link to a file repository (almost 1GB in files) so it cannot be verified by readers

Proposed fix (changes highlighted)

Although some small studies have suggested an association between occupational exposure to glyphosate and non-hodgkin lymphoma, subsequent work confirmed the likelihood this work suffered from bias, and the association could not be demonstrated in more robust studies.[17] According to the 2023 EFSA review of glyphosate, available epidemiological studies currently do not provide sufficient indication that glyphosate exposure is associated with any cancer-related health effect.[18]

• It fixes the "mismatched source to text" problem
• It specifies that the only link that was found was for occupational exposure (=very high exposure) and not the general population (this was an addition by me)
• It uses a direct link to the source of the EFSA statement as provided by Julius Senegal in the previous edits

{{u|Gtoffoletto}}  ^talk 19:52, 16 November 2023 (UTC)

I think the most appropriate 2023 EFSA source link to use to source the quote is the following (from the final published report on the EFS journal): https://doi.org/10.2903/j.efsa.2023.8164 {{u|Gtoffoletto}}  ^talk 20:13, 16 November 2023 (UTC)

Regarding the supporting material from EFSA - it is only 61 MB, not 1 GB, as it refers to the "Peer Review Report" (Part 3 of 6). But ofc we can go back to the 2022 EFSA report as direct link until EFSA makes it easier. --Julius Senegal (talk) 08:30, 17 November 2023 (UTC)

I can confirm that I clicked on the link and it takes you to a directory with almost 1GB in files. Unclear what you are referring to exactly. {{u|Gtoffoletto}}  ^talk 15:16, 20 November 2023 (UTC)

It is the first zip file ("Peer Review Report", published date 25/08/2023, 61.5 MB), and therein part 3 of 6 (page 59-78). --Julius Senegal (talk) 17:20, 20 November 2023 (UTC)

That's not very clear. It isn't easy to WP:VERIFY at all. Why not use the direct link I provided above that contains the quote in question directly and is the published final work? Also I would include the quote in the article directly instead of using the source to support unrelated text (see my proposal above). {{u|Gtoffoletto}}  ^talk 20:09, 20 November 2023 (UTC)

The current version of the ref template should also redundantly spell that out for anyone, since it points out the main document titles, where to find it within the document, pages, and the quote now.

VERIFY specifically has a section on WP:SOURCEACCESS. It's similar to citing a book as a ref. You're not always going to have a link to a book's exact page, but there's enough information there for anyone to go and verify it. Though honestly in this case, it's less effort than having to go to a library and check out a physical book. KoA (talk) 21:31, 20 November 2023 (UTC)

The citation currently reads: "Peer Review Report on Glyphosate (AIR V): Report of Pesticide Peer Review TC 80". EFSA. July 1, 2023. pp. 59–78. Retrieved November 16, 2023. The available epidemiological studies currently do not provide sufficient indication that glyphosate exposure is associated with any cancer-related health effect.

And links to https://open.efsa.europa.eu/study-inventory/EFSA-Q-2020-00140

What file am I supposed to be looking at? Am I supposed to download everything and look for it? This is not clear and the citation fails verification in my view.

This direct link to the final EFSA paper published on 26 July 2023 and contains the quote "The currently available human epidemiological studies do not provide conclusive evidence that glyphosate exposure is associated with any cancer-related health effect." {{u|Gtoffoletto}}  ^talk 14:11, 9 December 2023 (UTC)

In multiple previous discussions, there were issues with the last link you are giving (a summary of the summary) being taken out of context. Discussion was focused on using the full report instead to avoid that. ^{context was mixed up with previous discussion as I see Nangaf ran into too} The currently cited full report instead gives full context that addresses the editor concerns from previous discussions and isn't being cited only for the quote. As mentioned before, WP:VERIFY is clear about sources and not rejecting them when they are a slight inconvenience to access (again this one is much, much simpler than the library book example). What matters is that the report is cited with the correct name, etc., not that it has a direct link. Technically a link isn't even needed, but it is still being included.

Right now anyone should be able to follow the citation with only one link mentioning glyphosate peer-review and a relevant publication date. Once in that folder, a reader simply just needs to use the table of contents if they are uncertain of where to go. EFSA uses multiple subheaders, so mentioning part 3 of 6 can add even more redundancy to the citation. In terms of tweaks to the citation template for more redundancy, this is really all that was needed to address any lingering access questions. Bon courage, you made the initial edit, so what do you think of the tweak I just made in this context? KoA (talk) 15:47, 11 December 2023 (UTC)

I can find the document pointed-to (unfortunately there is no way using established Standards to specify a URL to a target within a ZIP, something I failed to get traction for in my professional life, but I digress). It does strike me as a slightly odd source to be citing, seemingly a procedural document from part of the process of achieving a final view. Do we need to see the sausage being made?

On the wider point, there is no need for a document to be easily linkable, or even linkable at all, so long as it is cited correctly.

Myself, I'd cite the final report using the conventional WP:MEDREF format, i.e.:

• Álvarez F, Arena M, Auteri D, Binaglia M, Castoldi AF, et al. (July 2023). "Peer review of the pesticide risk assessment of the active substance glyphosate". EFSA J (Review). 21 (7): e08164. doi:10.2903/j.efsa.2023.8164. PMC 10369247. PMID 37502013.

Bon courage (talk) 16:17, 11 December 2023 (UTC)

Whoops, I was re-reading the talk discussion and thought it was you who made the edit to include the report, but it was actually Julius Senegal who added this specific report to what you included. Julius could probably say more of what their intent was, but something like Pesticide Peer Review TC 80 is what was considered the first level secondary source in this. It came up in previous discussion, so I can see why Julius was looking at adding it.

That said, I'm good adding this ref to the current pool. I got wires crossed on which specific source was being discussed earlier since we had issues about that earlier, but this is the full publication. TC 80 is referenced in that and previous discussion, so it doesn't hurt to keep it, but unless Julius had something to add about that, I don't know if we need to dive into where that older discussion was going. KoA (talk) 16:55, 11 December 2023 (UTC)

Much better now. Agree with @Bon courage that the other source is strange and not very helpful at this point. It just makes things more messy and hard to verify/access. I would remove it.

I still see WP:SYNTH issues with the current text. The EFSA never talked about bias yet we source it to them.

Also, out of all the reviews we discussed (and we discussed many), the one we currently have in the article is the only one that was funded by Bayer US - Crop Sciences, Monsanto Company, Chesterfield, MO, USA and whose author John Acquavella worked for Monsanto from 1989 to 2004 and has consulted for them on glyphosate-related epidemiology issues, including litigation, since January 2015. Not very WP:INDEPENDENT... We have other high quality sources that can support those statements so I would definitely avoid this one. {{u|Gtoffoletto}}  ^talk 19:30, 11 December 2023 (UTC)

I wish that would have been brought up a month ago on Acquavella. When I looked at it I was only looking at authorship and the text. If it was a university professor who wrote it, it would have been a better situation because universities have conflict of interest checks built in for that kind of funding, but consultants really do not. That just means though that we'd go back to the previous source being used by Bon Courage, Boffetta et al. that everyone at least appeared to not have issue with (aside from Julius replacing it with the more recent source).

All that said, the EFSA sources, especially the harder to link ones to deal with the subject discussing bias and issues in methodology, so this doesn't really change the content at all. Part 3 even discusses DLBCL mentioned in Boffetta, only Leon et al. finding a positive association, but still saying Overall, the EFSA WG on glyphosate concluded that the strength of the evidence assessed does not support an association between human exposure to glyphosate and cancer at any site. It looks like EFSA addressed many of the literature reviews coming up here in some fashion, so if lesser evidence is contradicting EFSA, it's probably best to leave it out. KoA (talk) 02:18, 12 December 2023 (UTC)

When I looked at it I was only looking at authorship and the text. what do you mean you were only looking at "authorship"? {{u|Gtoffoletto}}  ^talk 19:09, 12 December 2023 (UTC)

• Julius Senegal, I just saw this edit. My suggestion would be to restore the Acquavella source since it didn't really need to be replaced. If there's something to supplement that text with the new EFSA source, it's probably best to at least include a quote in the source and just add the reference at the end of the sentence. Probably best to clarify what you're paraphrasing here though first to see what could be done for source formatting/text if needed though. KoA (talk) 20:54, 16 November 2023 (UTC)

My thought was to use the EFSA source as this is currently the highest quality. --Julius Senegal (talk) 08:30, 17 November 2023 (UTC)

And it's good to have a higher quality source, but the Acquavella source looks good too (and is more accessible), so that one should remain at least regardless of EFSA. Is there specific text from the EFSA source that mirrors what we've paraphrased from the Acquavella source though on bias and more robust studies? I think the main thing at this point is just being clear on what exactly is being sourced from EFSA given the amount of text in it, but otherwise I think the rest of the content is fine as-is. KoA (talk) 15:02, 17 November 2023 (UTC)

I went and dug a bit, and I think I found what you were referring to tweak the ref template, though let me know if I was off with that. That should be enough info now for the citation since you gave page number, etc. earlier, and the title of the specific report where page numbers restart is now mentioned too. I added Acquavella back though while doing that since it seems complimentary from what I can see. The quote at least gets to the point now that there's no association and avoids the issues with other inclusions discussed extensively above. KoA (talk) 15:38, 17 November 2023 (UTC)

Above, Rana et al. (2020) is discussed. I've just noticed that there is a new systematic review by Rana and co-workers (doi:10.1016/j.chemosphere.2023.139572). Could it be useful? --Leyo 00:04, 13 December 2023 (UTC)

That would be:

• Rana I, Nguyen PK, Rigutto G, Louie A, Lee J, Smith MT, Zhang L (October 2023). "Mapping the key characteristics of carcinogens for glyphosate and its formulations: A systematic review". Chemosphere (Systematic review). 339: 139572. doi:10.1016/j.chemosphere.2023.139572. PMID 37474029.

Which is a strong source (particularly if one disregards author COIs, as one should). However the conclusion is a bit 'meh'. It might be used to state that the authors found reasons to suspect that glyphosate has characteristics which suggest it will prove to be carcinogenic to humans. Bon courage (talk) 05:03, 13 December 2023 (UTC)

LZ and MTS are retained as experts on behalf of plaintiffs in litigation pertaining to the development of cancer following glyphosate exposure. Being hired to be involved in that litigation would be an issue in terms of COI. If the tables were flipped with a review from scientists retained by Bayer, I'd expect similar issues too. Like the Acquavella source, it's probably best to avoid it due to all those hurdles and stick to sources not associated with the litigation.

Setting COI aside though, we have other arguably stronger sources like EFSA that say does not support an association between human exposure to glyphosate and cancer at any site, so it becomes a WP:DUE issue with a single review like this too. At the least, we wouldn't be in a great position to cite Rana as-is. KoA (talk) 16:18, 13 December 2023 (UTC)

I agree with the conclusions of Bon courage and KoA. I just stumbled across it and wanted to bring it up for discussion (as Rana et al. 2020 was mentioned above). Interestingly, no such COI is declared in Rana et al. (2020), at least according to my incomplete "snippet view". They may have become involved in that litigation after 2020. --Leyo 00:46, 14 December 2023 (UTC)

References