Talk:Linear no-threshold model

Major Rewrite

I'm working on a major rewrite. Considering how much work this article needs, I'm not making all of the changes at once. There are three major issues: 1) The overall organization is awful. The sections are overlapping, repetitive, irrelevant, etc., 2) The description of the position of various scientific organizations is quite outdated and, 3) in addition to describing the position of major organizations, the article is peppered (in no organized manner) with references to individual research studies. This is a complex scientific issue, random wikipedia editors do not have the time or expertise to adjudicate hundreds of research studies. The article should primarily be relying on the opinion of major research organizations. On that note, I'm focusing on #2 first. I've made significant edits to describe the updated positions of the various organizations. In general they agree that LNT has the most evidence, or is at least not inconsistent with the evidence and that no other dose-response model has more evidence. Accordingly, I've removed the "Support" vs. "Opposed" framing and added a summary of what is/is not agreed on (they mostly support LNT though some caution against estimating health risks at low doses). Now that that's done, I suggest largely removing the numerous references to individual studies, probably including almost the entire "Fieldwork" section and the first few paragraphs of the "Controversy" section. It could be reasonable to leave a few of the major studies that are driving epidemiological research, but that would be a fair bit of effort and is not at all what's there now. I guess I'll hold of on that for now and give others a chance to weigh in on my recent edits first.wagsbags (talk) 21:54, 2 March 2024 (UTC)

I've made another major edit. I removed a lot of repetitive statements and continued refocusing the article on statements made by major scientific organizations, rather than authors of individual studies. In that vein, I removed the entire "Fieldwork" section. It included an overwhelming array of various individual studies, with no particular focus on the major ones. If we really want to reference individual studies, which I do not think is a good idea for a topic this complicated, we should be pointing to one of the overview reports from a major scientific organization that has reviewed such studies. Also that would probably be better placed in the "radiation-induced cancer" page.wagsbags (talk) 03:32, 7 May 2024 (UTC)

Rewrite requested

Article should define the LNT is and say who developed it. Also it should explain policy implications of the model.

Has LNT been applied to hazards other than radiation, such as cyanide in drinking water?

I'd especially like to see any studies which treat LNT as a hypothesis and have made predictions from it. Have any scientists made or tested such predictions? --Uncle Ed 18:42, 27 October 2006 (UTC)

Yeah, i think the article mentions that EPA uses this for all carcinogens. Pdbailey 03:45, 28 October 2006 (UTC)

That wasn't what I asked. You answered the question, "Does anyone rely on LNT as a guiding principle?" (no longer wondering whether it's true or not). My question was whether anyone is testing the model. (actively trying to determine whether it's true or not).

I guess you mean that the EPA bypassed the stage of determining whether it's true and went right ahead into using it as a guiding principle.--Uncle Ed 16:03, 15 November 2006 (UTC)

I think we still need some more explanation about LNT.
In its very pure form LNT means, that you basically integrate the radiation rate over your lifetime. The result is the complete life-long radiation dose to which you were exposed (in Sieverts). In its pure form LNT says that this number is proportional to the risk of getting radiation induced cancer.
Now already it is clear that this is an oversimplification which is wrong. Basically getting 4 Sv spread out over your whole lifetime (>50 years) is dangerous, but your chances of survival are good; especially when you are not irridated in your first few years. WP says if you get this dose in a short time-frame your chance of surviving the next month is < 50%.
So it is already widely accepted that long-term low rate radiation is far less dangerous then short-term high-rate radiation. This is already a departure from pure LNT.
Now for "middle" dose radiation LNT seems to hold (even if I really have no clue about what kind of time-frame we are talking about. 10 years 20 years life-long ?). So basically you add up the doses per year. The result seems to give a pretty good estimate about radiation induced cancer.
For "low" rate radiation, statistical data is so difficult to obtain, that it simply is not known if LNT is completely appropriate.
—Preceding unsigned comment added by 84.154.23.40 (talk) 23:50, 9 September 2007 (UTC)
Potassium 40
There actually is one statistic that is known for chronic irradiation of the entire biosphere. ⁴⁰K, isotope 40 of the biologically essential element potassium, is chronically radioactive, i.e. the isotope is long lived, and low but persistent in its radioactivity. An adult human of about 70 kg requires enough potassium that he or she is internally bombarded at a rate of 4000 or 5000 Bq, which sounds a lot smaller in Sieverts. Every living organism on the planet has ancestors that evolved to repair damage well enough at this level that RNA and DNA have a vastly better record of retaining important information than any medium known for human communication. So at this level of radiation, the LNT is flat wrong. As for chemical mutagens, presumably the same usually holds.
I am very dubious about conclusions for lifelong exposure to any particular threat.
The fact that human population has more than doubled in the last fifty years means that the total global human deaths per year for the next fifty must exceed twice what it was when I was a young man.
The risk of death from lifelong exposure to the world is 100%, by definition. DaveyHume (talk) — Preceding undated comment added 15:19, 25 March 2015 (UTC)

84.154.23.40, There are two exposure regimes: acute and chronic. Acute exposure is measured in Gy, not Sv as you suspect. Chronic exposure is measured in Sv. The joke is that the problem with the Sv is that if you get one, it's not the right unit anyway, but this isn't quite true because 1 Sv is not equal to 1 Gy for all radiation.

As for chronic exposure, the NAS endorses two regions of response and defines how they interact. Maybe you could be bold and add information about that. 0¹⁸ (talk) 13:45, 2 August 2010 (UTC)

United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR).

Didn't fancy getting involved in a likely edit war, but the UN - specifically UNSCEAR - have now confirmed that LNT has no scientific basis. Article needs significant changes to reflect this final nail in LNT's coffin. 92.238.128.94 (talk) 13:46, 15 January 2013 (UTC)

The link to the article is dead (reference 6). Any update available? Bacillus subtilis (talk) 02:42, 19 December 2023 (UTC)

Straw Man hypotheses

There are two straw man hypothesis that are only placed in the article to make it seem that some people believe that LNT is an underestimate. Nobody believes that. The link is to a hypothetical scenerio that is not supported by any studies, and is not held by any serious researcher. To put it in the same paragraph as radiation hormesis just makes LNT look like a safe middle ground, a place which it does not occupy. LNT is the extreme upper limit on low-level radiation damage in the literature, and should not be made to look like a middle of the road estimate.

The second unsupported claim is that "some cancers might respond linearly while others might not". This is also unsupported by any studies, and is unreferenced in the article, and is probably wrong given the current state of knowledge of cancer biology. All cancers, in current understanding, have a certain threshold of mutation damage that are required to defeat apoptosis, and it is likely that the response to radiation damage is very similar across groups. While it is possible that this is false, it should not be stated, because it is also a straw-man position designed to make LNT look better than it is.

This is why I removed the statements. Here they are for future reference:

1. Other alternatives include those in which response to radiation increase more than linearly at low doses or that the LNT model underestimates risk at low radiation exposure. In the later hypothetical case, below a certain threshold the subtle damage caused maybe missed and not repaired by the body, leading to a greater risk of disease then indicated by the LNT model.
2. It is also possible that some cancers respond linearly while others do not.

[unsigned] Uhm, isn't the first of the statements you removed a plausible consequence of hormesis? The response to low level exposure could reasonably be larger than the LNT model would estimate BECAUSE the exposure was below a threshold needed to activate cellular repair mechanisms? Obviously, if that were so, then the LNT model would already be broken, since LNT implicitly rejects the notion of a damage-activated repair mechanism. (Or so it seems to me.) Above this hypothetical activation level, the LNT model would over-estimate the risk, but there might still be a region at the far low end of exposure (below the repair activation threshold) where it would give an under-estimate? --76.102.148.93 (talk) 19:34, 24 March 2021 (UTC)

LNT is the extreme upper limit in the literature.

How can a rate be an upper limit? Doesn't it depend on the coefficient? When I compare Ax to Bx², I can't come to a conclusion without knowing A, B and the range of x you're interested in. 129.132.209.62 (talk) 19:22, 25 October 2014 (UTC)

It is the upper limit in the literature because it assumes absolutely any dose of ionizing radiation contributes additively to harmful mutation, while it is plausible that extremely low dosages contribute not at all. It predicts the highest risk in this sense. You could suppose that there is some greater than 1 multiplier for extremely low dosages (such that one thousand 1μSv exposures have the effect of a single 5mSv exposure), but I don't think you will find it taken seriously in the literature. nakomaru (talk) 08:20, 20 April 2016 (UTC)

References

What are the actual numbers?

At this point, I do not want to participate in the discussion whether the LNT-model is valid. The article does not give any numbers for the risk of cancer involved with any particular equivalent dose. Is it 0.5% deaths per 100mSv or 1% deaths per 100mSv?--Hokanomono ✉ 09:19, 15 April 2016 (UTC)

Per radiation-induced cancer, which references ICRP Publication 103, it is 0.5% per 100mSv. nakomaru (talk) 08:03, 20 April 2016 (UTC)

Possible conflict of interest - extreme bias against linear no-threshold model

This article is extremely biased against the linear no-threshold model, which is the model endorsed by mainstream consensus.

As it stands, almost every paragraph of this article grudgingly acknowledges that the LNT model even exists, followed by extensive debunking of it using a combination of primary research, Wall Street Journal interviews, and basically any old quote from a small-fry not-for-profit org that doesn't like LNT.

If a regular person read this article they would walk away not properly informed about the reputation of LNT.

Note that the minoritarian effort to get LNT overturned continues to be rejected by the authorities. This Wikipedia article sounds like it was written by the petitioners in this case.

Our article even includes an entire section devoted to one of the cornerstones of the anti-LNT position: that the psychological effects of radiation fear are riskier than low-dose radiation exposure.

It is difficult to assume good faith with such persistent bias against LNT in the article. -- Hunan201p (talk) 18:51, 1 April 2022 (UTC)

Sorry, I don't see extreme bias in the article. I see haphazard organization. There is no clearly written passage which outlines the case for or against LNT. Mostly it reads like WP:OR.

I'd rather hear about the major agencies that have taken a position and what activities or substances they're regulating, along with any research they've done to confirm or contradict their theoretical stance. --Uncle Ed (talk) 17:24, 10 January 2024 (UTC)

Recent edit seems biased/unprofessional

A recent edit mentions that the US should proclaim to the world that background radiation causes increased cancer and that "Amazon sales of lead lined clothing would skyrocket." This seems like a personal opinion that should be removed from this article. 66.74.95.161 (talk) 20:23, 26 November 2023 (UTC)

Let's focus on the data to resolve this controversy

From the lead: "Scientific organizations generally support use of the LNT model ..." Really !? Maybe we should change this to "Government organizations ..." It seems to me that the battle is between scientists that want to abandon LNT, and government organizations that want to keep it.
Here are two scientific organizations that say LNT is bunk:
HPS https://hps.org/hpspublications/historylnt/episodeguide.html
SARI X-LNT https://www.x-lnt.org/evidence-for-radiation-hormesis
The videos from Health Physics Society explore the reasons why governments have clung to LNT.
The X-LNT website, sponsored by the Scientists for Accurate Radiation Information, has summaries of 7 studies supporting radiation hormesis. I find section 7 to be the most convincing. "Lung cancer rates decrease with increasing residential radon levels". There is just no other way to interpret this data.
I'm searching for any good data that contradicts this conclusion. Simple science is more convincing than an author's credentials or the prestige of a publication. Here is a plot from the European Code, but I can't find the source of the data:
https://www.facebook.com/groups/2081763568746983/posts/3726936500896340?comment_id=3728618867394770
David MacQuigg 01:25, 10 June 2024 (UTC)

Yes really. I updated the controversy section with the latest positions of the major scientific organizations, not just government agencies. The edits are inconsistent with those positions. The phrase "many scientists 'now' dispute..." is used to try to make it seem like more evidence is coming out against LNT. That is not the case. In fact, HPS' and ANS' positions seems to have softened over the years, with the uncertainty now noted as below background rather than around 5 rem. The HPS position does not say the LNT is "bunk" which is pretty biased language. I've never heard of SARI X-LNT but they appear to be a tiny organization of a few random people that really doesn't belong in the same conversation as international bodies of dozens of experts in the field. Regardless, they don't seem to have convinced the experts in the field. I'll not weigh in on the facebook reference.
wagsbags (talk) 03:19, 18 July 2024 (UTC)

This controversy reminds me of the debate over Einstein's theory of relativity. When confronted by a reporter with 100 papers disproving his theory, Einstein responded - Why 100? One would be enough. I am disappointed that you won't focus on the data I suggested, and continue to push the "latest positions" of organizations supporting LNT. Did you look at the latest from the Health Physics Society (linked above)? The quote from the Health Physics Society in your "updated controversy section" has only their vague statement from 2019 about "statistical uncertainties". Yes, there is a lot of uncertainty in trying to disprove a very small effect (radiation induced cancer) in a large background (cancers from other causes), yet that is exactly what the Cohen study shows (one of the seven studies summarized on the X-LNT website). Don't tell me they are a "tiny organization" or label them as "random people". Read their summary and tell me what is wrong with it. Simple science is more convincing than an author's credentials or the prestige of a publication.

Sorry about that link to the discussion on FaceBook. You have to scroll through it to find that plot from the European Code. Here is a better link to a summary of this radon controversy, showing both plots and what I believe are the best arguments from both sides. https://citizendium.org/wiki/Fear_of_radiation/Debate_Guide#LNT_and_radon,_Controversy_over_Figure_4

This talk page is not the right place for ongoing discussion. That is the role of the FaceBook forum. Please join me there for further discussion.
David MacQuigg 15:52, 18 July 2024 (UTC)

@Macquigg: please review WP:MEDRS. The sources you're presenting here are not usable. VQuakr (talk) 17:58, 18 July 2024 (UTC)

HPS is the same source cited earlier in this article to support LNT. I have simply cited a later publication from that same source.

X-LNT.org provides a summary of seven studies, each with citations to reliable sources.
David MacQuigg 18:40, 18 July 2024 (UTC)

X-LNT is synthesizing a non-peer reviewed position not supported by those studies. It is not a meta analysis and not a reliable source for medical claims. HPS in its current usage in the article is on history of LNT and an WP:ABOUTSELF citation to the HPS's own position. It is not usable for a claim that the LNT is discredited in general. For below ~100 mSv, the state of the field remains "we don't know, so we use a conservative model." VQuakr (talk) 19:38, 18 July 2024 (UTC)

Bowing out I don't have the time or inclination to debate Wikipedia rules. If anyone is interested in resolving this controversy by discussing the science, I recommend the (spam removed) ongoing at Citizendium.
David MacQuigg 18:50, 20 July 2024 (UTC)

Wikipedia's rules are that we follow what Reliable Sources say, and that means that if the consensus in Reliable Sources is factually wrong, than it should appear as factually wrong in Wikipedia until the consensus in RS's changes. Though some of us may have the background to interpret data and studies, that is not the job of Wikipedia editors: that is specifically banned here: WP:NOR, with good reason: as anonymous editors, we don't know whether anyone's claimed expertise is valid and how that compares to the claimed expertise of other editors. Many times experts with equal qualifications/experience disagree, and when that happens, we REPORT ON the disagreement, we don't adjudicate it.

I need to review the edits to the article, but I agree that it was a mess and could use improvement. Meta-analyses are better than single studies.---Avatar317^(talk) 01:46, 23 July 2024 (UTC)