Reminded of something that Geraldine Thomas, founder of the Chernobyl Tissue Bank, wrote when the Fukushima drama was going on, "Look at the science – smoking and obesity are more harmful than radiation":
> I can assure you that none of us are in the pay of the nuclear industry. I was anti-nuclear until I worked on the after effects of the Chernobyl accident – now I am very pro-nuclear as I realise that we have an unwarranted fear of radiation – probably due to all the rubbish about a nuclear winter we were fed during the Cold War.[10]
Paradoxically active ongoing research suggests that the same dose of (medically determined) radiation delivered over a very short period of time increases its efficacy against the cancer compared to damage to the surrounding tissue.
It's interesting that the core of this effect - that large doses delivered on timescales of a second or less significantly reduce normal tissue toxicity - was actually observed over 40 years ago, e.g.:
Some of the authors of that work were still active when the "Flash" concept came around again, and to hear them tell the story they tried to get funding to test it in tumour-bearing mice, but both their lab heads and external funders were sure the effect would be the same in tumours as normal tissue, and so wouldn't fund the work.
They eventually moved on to other things, and it needed a few decades for someone with enough soft money to give it a go and kick off this new research field.
Even before Linear No Threshold was a thing scientists were doing experiments showing that dosing fruit flies with radiation all at once would lead to a highly mutated second generation, but spreading that radiation out over the course of a month wouldn't.
Most radiobiological effects for acute exposures are actually quadratic in dose - this applies to things like mutations, chromosome aberrations, and lethal events.
The quadratic term relates (loosely) to interactions between damage from multiple ionising particles which are present in cells at the same time. When you protract exposures, more damage is repaired before subsequent ionising particles arrive, and you see a reduction in the quadratic term, to effectively a linear form at very low dose rates.
So it's not surprising that spreading the radiation out significantly reduces the yield of biological events, but actually supports a linear trend for mutation yield in low-dose and low dose-rate conditions. (Whether that tracks linearly in turn to cancer risk, on the other hand, is a topic of much more debate.)
At the time (the 1950s) evidence from fruit flies was essentially all the evidence there was. But as the article points out we have pretty good data showing it works the same way in humans.
I do not agree that they have good data showing it works the same in humans. I see a lot of stretching of datasets, a lot of squinting, and above all a lot of hope!
"Lie" is such a click-baity triggering word to use here. Sounds more like they overstated the harm of cumulative small doses of radiation.
But, I mostly just skimmed through the beginning of the article, so maybe it gets better, like maybe the author reveals an international cabal of influential anti-nuclear activists who are holding human progress back.
> Sounds more like they overstated the harm of cumulative small doses of radiation.
Not even that, they simply didn’t know because they couldn’t measure, so they took a conservative approach.
Btw you can count me in to the cabal of anti-nuclear activists. Humans simply are too greedy and incompetent to manage the technology responsibly over the long term. We’ve already irreversibly altered the biosphere with the nuclear activity we’ve engaged in so far. Time for it to stop.
Your not wrong, but you seem to be missing one significant detail: we have altered the biosphere even more by not engaging in nuclear activity, instead opting for less-scary-but-worse alternatives like coal, oil and gas.
The problem is that stopping nuclear activity has ancillary effects -- like increased carbon emissions -- that are potentially much more harmful than the radiation. The results of technological decisions are never independent of one another.
Check out the interview with Dr Bernie Cohen, who did a lot of the early epidemiological work. The interviewer is rather woo, but the professor is as hard-nose a scientist as you could hope for. It makes a good pair because it let him correct misconceptions.
Long story short, Dr Cohen became unpopular after his data showed home radon levels to be negatively correlated with lung cancer risk. The more radon, the lower your risk of lung cancer.
> Drinking one beer a night for a year is a lot less harmful than drinking 365 beers in one go. The same applies to radiation exposure, but regulation doesn’t agree.
Stating something confidently doesn't make it true. Show me the data.
The article is long on emotion, exposition, but very short on the data.
There's a big concerted effort to change this regulation, but it's not based on data, it's based on feelings.
It's quite likely that there's non-linear response, but it could just as easily be that the dose that's tolerated well in a 1 day exposure, might have higher risk when spread out over 365 days. When they say something like:
> nor any major chromosomal aberrations.
They don't have the technology to measure DNA damage that might be significant. I've spent some time in the past examining the REBC dataset of whole-genome sequencing of tumors of thyroid cancers from Chornobyl, where you actually do see the types of translocations that cause cancer from radiation.
We can't detect these types of translocations in non-cancerous tissue. The only reason we can see them in cancer is that the cancer has replicated billions of times, giving us many many many copies of the translocation to put through DNA sequencing. Doing the type of sequencing where we identify translocations that happen in individual cells, before the cell has become cancerous, would require a good amount of engineering effort, and I've never seen anything like it. And in 2006, when the study was published, we barely had any of the latest sequencing technologies.
> Chen interpreted this as evidence of the health benefits of radiation. This theory, known as hormesis, holds that low doses of stressors, including ionizing radiation, can improve health (in this case, reducing cancer risk) by triggering the body’s repair systems in much the same way that exercise improves fitness by stressing the cardiovascular system. While popular among a small community of researchers, it has not gained widespread acceptance due to limited and conflicting evidence in humans.
Yes, limited and conflicting evidence in humans. Yet these sorts of propaganda efforts are pushing hard on the idea being present, being obvious.
This article is not science, despite trying to put on airs of science. The data does not support their claims.
Let's see actual review articles published making these claims that aggregate over large numbers of small data. Let's see whether such aggregation claims hold up on scrutiny from those that have spent a lot of time thinking about this.
The active regulatory push to invalidate LNT should follow the science, not be ahead of the science.
Plus, the whole goal of this, to somehow how make nuclear construction cheaper, does not seem to be well served by changing LNT. The costs of nuclear are massive because it's a big constructuon project with lots of coordination. Making concrete walls 50% as thick is going to do very little to lessen the massive costs, which are related to construction productivity, or rather the lack of it in the West.
It seems like the nuclear industry tries to focus on anything except the one thing that will actually make it succeed: get really good at construction.
in a world where there is no safe low radiation dose, it would be quite easy to generate the data to demonstrate this. so either low doses cause no harm or cause such minimal harm as to be safely disregarded.
> in a world where there is no safe low radiation dose, it would be quite easy to generate the data to demonstrate this.
This is the classic fallacy seeing an absence of evidence and using that as evidence of absence!
And the lack of evidence goes both ways, it should be easy to show that current regulations are fully safe by doing epidemiology to show that living close to a nuclear power plant carries no additional risk!
So let's go looking for those epidemiological studies...
> May 19 2026 - Does Proximity to Nuclear Power Plants Increase Cancer Risk? New research finds correlation between disease and living close to a facility
> Koutrakis says that his advisee’s research is notable because it is the first series of studies to systematically demonstrate associations between residential proximity to nuclear power plants and cancer outcomes across multiple settings using large, population-based datasets. “This work fills a critical gap in the literature by providing large-scale, systematic evidence on a question that has remained unresolved for decades.”
> Using nationwide mortality data from 2000-2018, we assess long-term spatial patterns of cancer mortality in relation to proximity to nuclear facilities while accounting for socioeconomic, demographic, behavioral, environmental, and healthcare factors. Cancer mortality is higher across multiple age groups in both males and females, with the strongest associations among older adults, males aged 65–74 and females aged 55–64.
So there's a dose-response curve for cancer based on living close to a nuclear power plant. This survives correction for other confounders.
Notably, this is correlation not causation, but the only evidence getting close to disproving LNT actual leans towards super-linear, rather than sub-linear, correct?
Then you are not understanding it. Looking at a map of people not close to power plants would show the same rough picture. People live where people live, of course! But proximity to nuclear power plants has higher incidence of cancer.
What is different about this study that's worthy of a national map is that it's an evaluation of national data, after having first found the discovery on smaller state level datasets.
My problem with radiation: the units of exposure are so clumsy. We should have standardized on nanoSieverts as the main unit. The normal background radiation is 200 nSv per hour, and you get acute radiation sickness at 1_000_000_000 nSv. The lethal dose is 5_000_000_000. It really puts things into perspective.
E.g. even 10x the normal background is still ridiculously low.
Also, the LNT model is good enough. It's really the most conservative model that we have, so it makes sense to keep using it. We just need to quantify the risk increases properly.
Wow, this was a cherry-picking peace of misinformation, conveniently ignoring the actual counts of people affected by Chernobyl.
There's no safe dose of radiation, there's only statistics.
And I'm not sure what this article is supposed to justify?
Building power generation technology with the potential to make whole regions unlivable is ok now?
Willfully creating hazards that can affect people for thousands of years, starting with Uranium mining & processing to nuclear waste is a good idea?
Having to fortify a nuclear plant so it can withstand a plane crash (most won't withstand double plane crashes), securing it against terrorist - and then still have it fall into enemy hands that can use it as a bargaining chip (Russians are controlling Zaporizhzhia) is a good idea?
You know what the engineers of Three Mile Island, Chernobyl and Fukushima had in common?
They thought their plants were safe.
So even if "Radiation totally not bad, actually healthy" is the point here: It is still a tremendously stupid idea to build nuclear power plants when there are much better and cheaper alternatives.
There are studies that show cancer risk is higher near a nuclear plant. The reason is likely that poorer people live near a nuclear plant; it's _probably_ not because of radiation. My point is: Just having nuclear plants nearby lowers the market price of the property. If there _is_ an accident, the market price of many properties drops to zero. That's why no insurance company will insure the full risk of a nuclear accident: the remaining risk is on the population and land owners. (Property owners may get compensated - paied by taxes.)
In Switzerland there is now again the idea to build nuclear plants, by some (I'm pretty sure the political party that initiated this gets a lot of money from the nuclear lobby - unfortunately the money flow is intransparent in Switzerland.) A recent study in Switzerland [1] has shown nuclear plant are not competitive with solar, wind, hydro, and batteries, not even taking into account that accidents are not fully insured.
They studied the population in Hiroshima and Nagasaki for generations and found that in the first generations after the bombs there were elevated levels of hard-tissue cancers and in later generations elevated levels of soft-tissue/blood cancers. They're still dealing with the population effects of radiation 75-years later.
No one will be able to live in Chernobyl or Fukushima for hundreds of years. Or, well, they could but it would be stupid.
Much of the Fukushima area is inhabited again (the exclusion zone has shrunk from an original 1250km2 down to 371) and there is ample evidence that the overreaction evacuation did a lot more harm than good.
Reminded of something that Geraldine Thomas, founder of the Chernobyl Tissue Bank, wrote when the Fukushima drama was going on, "Look at the science – smoking and obesity are more harmful than radiation":
* https://www.theguardian.com/environment/2011/apr/26/obesity-...
> I can assure you that none of us are in the pay of the nuclear industry. I was anti-nuclear until I worked on the after effects of the Chernobyl accident – now I am very pro-nuclear as I realise that we have an unwarranted fear of radiation – probably due to all the rubbish about a nuclear winter we were fed during the Cold War.[10]
* https://en.wikipedia.org/wiki/Geraldine_Thomas
Paradoxically active ongoing research suggests that the same dose of (medically determined) radiation delivered over a very short period of time increases its efficacy against the cancer compared to damage to the surrounding tissue.
Here's one: https://pmc.ncbi.nlm.nih.gov/articles/PMC10640654/
It's interesting that the core of this effect - that large doses delivered on timescales of a second or less significantly reduce normal tissue toxicity - was actually observed over 40 years ago, e.g.:
https://radiation-research.kglmeridian.com/view/journals/rar...
Some of the authors of that work were still active when the "Flash" concept came around again, and to hear them tell the story they tried to get funding to test it in tumour-bearing mice, but both their lab heads and external funders were sure the effect would be the same in tumours as normal tissue, and so wouldn't fund the work.
They eventually moved on to other things, and it needed a few decades for someone with enough soft money to give it a go and kick off this new research field.
How's that paradoxical? Wouldn't TFA would seem to support findings of that sort?
Even before Linear No Threshold was a thing scientists were doing experiments showing that dosing fruit flies with radiation all at once would lead to a highly mutated second generation, but spreading that radiation out over the course of a month wouldn't.
Most radiobiological effects for acute exposures are actually quadratic in dose - this applies to things like mutations, chromosome aberrations, and lethal events.
The quadratic term relates (loosely) to interactions between damage from multiple ionising particles which are present in cells at the same time. When you protract exposures, more damage is repaired before subsequent ionising particles arrive, and you see a reduction in the quadratic term, to effectively a linear form at very low dose rates.
So it's not surprising that spreading the radiation out significantly reduces the yield of biological events, but actually supports a linear trend for mutation yield in low-dose and low dose-rate conditions. (Whether that tracks linearly in turn to cancer risk, on the other hand, is a topic of much more debate.)
Cancer cured (in mice)!
Radiation response shown to be sub-linear (in flies)!
At the time (the 1950s) evidence from fruit flies was essentially all the evidence there was. But as the article points out we have pretty good data showing it works the same way in humans.
I do not agree that they have good data showing it works the same in humans. I see a lot of stretching of datasets, a lot of squinting, and above all a lot of hope!
"Lie" is such a click-baity triggering word to use here. Sounds more like they overstated the harm of cumulative small doses of radiation.
But, I mostly just skimmed through the beginning of the article, so maybe it gets better, like maybe the author reveals an international cabal of influential anti-nuclear activists who are holding human progress back.
> Sounds more like they overstated the harm of cumulative small doses of radiation.
Not even that, they simply didn’t know because they couldn’t measure, so they took a conservative approach.
Btw you can count me in to the cabal of anti-nuclear activists. Humans simply are too greedy and incompetent to manage the technology responsibly over the long term. We’ve already irreversibly altered the biosphere with the nuclear activity we’ve engaged in so far. Time for it to stop.
Your not wrong, but you seem to be missing one significant detail: we have altered the biosphere even more by not engaging in nuclear activity, instead opting for less-scary-but-worse alternatives like coal, oil and gas.
This could be a convincing argument 70 years ago but we have other options now - mostly batteries and renewables.
"But what about the cobalt mines?" - damage limited in both space and time
Historically, yes. We have good alternatives now, though. What’s stopping us moving off carbon fuel is not the viability of alternatives.
The problem is that stopping nuclear activity has ancillary effects -- like increased carbon emissions -- that are potentially much more harmful than the radiation. The results of technological decisions are never independent of one another.
Don't forget
https://en.wikipedia.org/wiki/Naturally_occurring_radioactiv...
considerable amounts of low-level radiation is emitted by fossil fuel production and use as well as and construction materials.
Skeptical? I was too.
Check out the interview with Dr Bernie Cohen, who did a lot of the early epidemiological work. The interviewer is rather woo, but the professor is as hard-nose a scientist as you could hope for. It makes a good pair because it let him correct misconceptions.
Long story short, Dr Cohen became unpopular after his data showed home radon levels to be negatively correlated with lung cancer risk. The more radon, the lower your risk of lung cancer.
Part 1: https://www.youtube.com/watch?v=xhkBLhw-8pk
Part 2: https://www.youtube.com/watch?v=SuUFiUoynPo
> Drinking one beer a night for a year is a lot less harmful than drinking 365 beers in one go. The same applies to radiation exposure, but regulation doesn’t agree.
Stating something confidently doesn't make it true. Show me the data.
The article is long on emotion, exposition, but very short on the data.
There's a big concerted effort to change this regulation, but it's not based on data, it's based on feelings.
It's quite likely that there's non-linear response, but it could just as easily be that the dose that's tolerated well in a 1 day exposure, might have higher risk when spread out over 365 days. When they say something like:
> nor any major chromosomal aberrations.
They don't have the technology to measure DNA damage that might be significant. I've spent some time in the past examining the REBC dataset of whole-genome sequencing of tumors of thyroid cancers from Chornobyl, where you actually do see the types of translocations that cause cancer from radiation.
We can't detect these types of translocations in non-cancerous tissue. The only reason we can see them in cancer is that the cancer has replicated billions of times, giving us many many many copies of the translocation to put through DNA sequencing. Doing the type of sequencing where we identify translocations that happen in individual cells, before the cell has become cancerous, would require a good amount of engineering effort, and I've never seen anything like it. And in 2006, when the study was published, we barely had any of the latest sequencing technologies.
> Chen interpreted this as evidence of the health benefits of radiation. This theory, known as hormesis, holds that low doses of stressors, including ionizing radiation, can improve health (in this case, reducing cancer risk) by triggering the body’s repair systems in much the same way that exercise improves fitness by stressing the cardiovascular system. While popular among a small community of researchers, it has not gained widespread acceptance due to limited and conflicting evidence in humans.
Yes, limited and conflicting evidence in humans. Yet these sorts of propaganda efforts are pushing hard on the idea being present, being obvious.
This article is not science, despite trying to put on airs of science. The data does not support their claims.
Let's see actual review articles published making these claims that aggregate over large numbers of small data. Let's see whether such aggregation claims hold up on scrutiny from those that have spent a lot of time thinking about this.
The active regulatory push to invalidate LNT should follow the science, not be ahead of the science.
Plus, the whole goal of this, to somehow how make nuclear construction cheaper, does not seem to be well served by changing LNT. The costs of nuclear are massive because it's a big constructuon project with lots of coordination. Making concrete walls 50% as thick is going to do very little to lessen the massive costs, which are related to construction productivity, or rather the lack of it in the West.
It seems like the nuclear industry tries to focus on anything except the one thing that will actually make it succeed: get really good at construction.
> There's a big concerted effort to change this regulation, but it's not based on data, it's based on feelings.
Is the regulation based on hard, systematic and replicated data? Looks based on emotion (fear, greed) too
in a world where there is no safe low radiation dose, it would be quite easy to generate the data to demonstrate this. so either low doses cause no harm or cause such minimal harm as to be safely disregarded.
luckily the government is moving away from your position: https://www.eenews.net/articles/nrc-considers-eliminating-ha...
not having cheaper nuclear energy imposes a far greater cost on society.
No safe low radiation dose, you say? Well then, you had better stay away from red meat, brazil nuts and even bananas.
Consumed them already, you say? Well I guess you're screwed then.
Luckily? The NRC is considering it, hopefully they follow science rather than popular propaganda.
https://www.nature.com/articles/s41467-026-69285-4
> in a world where there is no safe low radiation dose, it would be quite easy to generate the data to demonstrate this.
This is the classic fallacy seeing an absence of evidence and using that as evidence of absence!
And the lack of evidence goes both ways, it should be easy to show that current regulations are fully safe by doing epidemiology to show that living close to a nuclear power plant carries no additional risk!
So let's go looking for those epidemiological studies...
> May 19 2026 - Does Proximity to Nuclear Power Plants Increase Cancer Risk? New research finds correlation between disease and living close to a facility
> Koutrakis says that his advisee’s research is notable because it is the first series of studies to systematically demonstrate associations between residential proximity to nuclear power plants and cancer outcomes across multiple settings using large, population-based datasets. “This work fills a critical gap in the literature by providing large-scale, systematic evidence on a question that has remained unresolved for decades.”
https://gsas.harvard.edu/news/does-proximity-nuclear-power-p...
And what do they see?
> Using nationwide mortality data from 2000-2018, we assess long-term spatial patterns of cancer mortality in relation to proximity to nuclear facilities while accounting for socioeconomic, demographic, behavioral, environmental, and healthcare factors. Cancer mortality is higher across multiple age groups in both males and females, with the strongest associations among older adults, males aged 65–74 and females aged 55–64.
https://www.nature.com/articles/s41467-026-69285-4
So there's a dose-response curve for cancer based on living close to a nuclear power plant. This survives correction for other confounders.
Notably, this is correlation not causation, but the only evidence getting close to disproving LNT actual leans towards super-linear, rather than sub-linear, correct?
I am looking at their county level distance-to-power-plant map and it's literally xkcd 1138.
Then you are not understanding it. Looking at a map of people not close to power plants would show the same rough picture. People live where people live, of course! But proximity to nuclear power plants has higher incidence of cancer.
What is different about this study that's worthy of a national map is that it's an evaluation of national data, after having first found the discovery on smaller state level datasets.
My problem with radiation: the units of exposure are so clumsy. We should have standardized on nanoSieverts as the main unit. The normal background radiation is 200 nSv per hour, and you get acute radiation sickness at 1_000_000_000 nSv. The lethal dose is 5_000_000_000. It really puts things into perspective.
E.g. even 10x the normal background is still ridiculously low.
Also, the LNT model is good enough. It's really the most conservative model that we have, so it makes sense to keep using it. We just need to quantify the risk increases properly.
Wow, this was a cherry-picking peace of misinformation, conveniently ignoring the actual counts of people affected by Chernobyl.
There's no safe dose of radiation, there's only statistics.
And I'm not sure what this article is supposed to justify? Building power generation technology with the potential to make whole regions unlivable is ok now?
Willfully creating hazards that can affect people for thousands of years, starting with Uranium mining & processing to nuclear waste is a good idea?
Having to fortify a nuclear plant so it can withstand a plane crash (most won't withstand double plane crashes), securing it against terrorist - and then still have it fall into enemy hands that can use it as a bargaining chip (Russians are controlling Zaporizhzhia) is a good idea?
You know what the engineers of Three Mile Island, Chernobyl and Fukushima had in common? They thought their plants were safe.
So even if "Radiation totally not bad, actually healthy" is the point here: It is still a tremendously stupid idea to build nuclear power plants when there are much better and cheaper alternatives.
There are studies that show cancer risk is higher near a nuclear plant. The reason is likely that poorer people live near a nuclear plant; it's _probably_ not because of radiation. My point is: Just having nuclear plants nearby lowers the market price of the property. If there _is_ an accident, the market price of many properties drops to zero. That's why no insurance company will insure the full risk of a nuclear accident: the remaining risk is on the population and land owners. (Property owners may get compensated - paied by taxes.)
In Switzerland there is now again the idea to build nuclear plants, by some (I'm pretty sure the political party that initiated this gets a lot of money from the nuclear lobby - unfortunately the money flow is intransparent in Switzerland.) A recent study in Switzerland [1] has shown nuclear plant are not competitive with solar, wind, hydro, and batteries, not even taking into account that accidents are not fully insured.
[1] https://www.20min.ch/story/akw-debatte-neue-atomkraftwerke-l...
> There's no safe dose of radiation,
The article claims the opposite with sources.
> there's only statistics.
You forgot about lies.
> ignoring the actual counts of people affected by Chernobyl.
Do you know what's the count?
Very odd risk calculations seem to be afoot here
> there are much better and cheaper alternatives.
Such as? (of course with the same energy density and 24/7 capacities)
They studied the population in Hiroshima and Nagasaki for generations and found that in the first generations after the bombs there were elevated levels of hard-tissue cancers and in later generations elevated levels of soft-tissue/blood cancers. They're still dealing with the population effects of radiation 75-years later.
No one will be able to live in Chernobyl or Fukushima for hundreds of years. Or, well, they could but it would be stupid.
Much of the Fukushima area is inhabited again (the exclusion zone has shrunk from an original 1250km2 down to 371) and there is ample evidence that the overreaction evacuation did a lot more harm than good.
Around 98% of Fukushima is inhabited again, unless of course you meant the NPP itself, but people were not living in a power plant to begin with.
That’s the entirety of the prefecture, most of which had not been evacuated.