As a physicist and full professor in medical imaging I can ensure you somewhat authoritatively: You can simply disregard every clown on the internet who dismisses linear no threshold out of hand. This is fairly sure sign that this person relies on pseudoscientific sources, also promoted by certain industries and shills which generally lobby against anti-pollution laws and sadly adopted by some pro-nuclear lobbyists. (of course, nuclear is far better than coal and one can make very reasonable arguments for nuclear, but simply dismissing LNT puts you directly into the pseudoscience world)
back
2 comments
Okay great, you've established some measure of credibility. But given the number of medical treatments and established methods of with zero actual backing evidence, could you cite some decent support for LNT?
Currently TFA links to some interesting arguments in favor of a time-rate dependent factor that incorporates cellular repair mechanisms. At first glance that seems plausible, especially below a certain threshold. There appears to be no statistical evidence of higher cancer rates around many nuclear areas, etc. LNT is not a topic I've dived into but seen pop up a few times so I'm curious why LNT is considered unassailable.
Claims based on arguments similar to "because I said so, and I should know" are dubious in my opinion, because how is one supposed to distinguish those claims from mere opinion? Instead arguments should point to evidence that provide new information or counter examples, if they are to have a chance to be convincing.
It is particularly difficult with nuclear as there are so many different professions involved. You may hear from a nuclear engineer about safety. But they are not a materials expert, or a human factors specialist, or a geologist, or a hydrologist. And safety depends on all of these things. And it depends on intimate understanding of the specific design and practices of a power station. Much of which will be confidential.
Basically, there is lot of empirical data which can be described by a linear relationship. Of course, if you go to even lower doses then the effect size becomes smaller and it becomes harder to show this effect in epidemiological studies. But large scale studies show that even for the rather low doses from medical imaging, the data is consistent with this model (e.g. Pearce et al., Lancet 2012;380:499–505). A lot of effort therefor goes into reducing radiation from medical imaging. Conceptually, we assume that there has to be some tiny risk with any damage caused by ionizing radiation because physiological repair mechanisms can not be assumed to be perfect. So the assumption that additional radiation (there is always a background) causes a linear increase of risk is natural. It is, of course, possible that additional effects can cause some nonlinear effects, e.g. additional repair mechanism get activated at some dose or repair gets overwhelmed at some point. But something like this would not really be expected at extremely low dose close to background levels. It is also not clear whether such an effect would lead to a lower or higher risk estimate for low doses.
Finally, one can also observe that some of the most prominent critics of LNT are connected to industry lobbyism. We have seen this before also in the tobacco industry who downplayed the risk of smoking, sometimes even the same persons appear in this context. Terminology such such "LNT worshipping" is a clear tell sign where this is coming from. Similar to all pseudoscience it has to come with a theory how established science is wrong because scientists are stupid, corrupt, worshipping some dogma, etc.
LNT is the only data we have but that doesn't make it right. All the data points for figuring radiation risk are based on high level exposures, LNT is the best fit line that passes through 0,0. The problem is that this makes the assumption that the line is straight.
The problem with this is we have many data points for lower level exposure that seem to almost all lie well below the line, but none have the statistical power to prove it even with very large sample sizes. Even the hottest of these data points shows nothing--why isn't there a cancer cluster in the people who live the lives in 20 rem/year??
And note that we aren't dismissing the LNT out of hand, we are questioning it in a certain realm. While I question the homeostasis argument it seems completely reasonable that the rate has something to do with it. In chemistry the norm is the dose makes the poison (you can breathe 3 psi of oxygen for a lifetime, 30 psi of oxygen and you'll be in trouble in less than an hour), what's so unreasonable about thinking that it might apply in this case? (And note that I don't think it applies much in your realm--while medical imaging isn't all that much radiation it's delivered quickly. Other than my one encounter with Tc-99m everything I've encountered has been in hold-your-breath type times.)
There's also the possibility the true line is a parabola. We have so few data points at the high end that the current data is completely incapable of excluding this case and it would fit the low data points better.
There are many possible reasons why LNT may not be exactly right, but misrepresenting it as anything else than the current best assumption is dishonest in my opinion. I am not aware of hard evidence contradicting it. Overall, evidence is consistent with linear risk at low doses and this is also consistent with plausible mechanisms. Even with LNT we do not expect "cancer clusters" in regions higher background radiation. We would expect a small excess number of cancers on top a huge amount of cancer we have anyway. This is statistically difficult to detect and there many other possible confounders. Regarding the original article, terminology such as "LNT worshipping" is a clear sign that sources are pseudoscientific.