Hacker News Reader

Comment by zosima | Hacker News Reader

If you look at the relative risk curve (the very last page of the supplementary data), most of the exposure levels contain 1 in the relative risk estimate (also please notice that for some reason the confidence interval is 90%, if choosing the customary 95% all relative risk confidence intervals would probably overlap relative risk of 1.0):

https://www.bmj.com/content/bmj/suppl/2023/08/16/bmj-2022-07...

The fact that ~10mGy seems to not have 1.0 in the confidence interval is kind of bothering to me. Normal annual background radiation should be about ~3.5mGy (so 10mGy is about 3 years normal background radiation) and getting an increased solid tumor mortality risk from normal radiation levels, does suggest that something else has not quite been controlled for.

Also the same group did the same paper on the same cohort before: https://www.bmj.com/content/351/bmj.h5359?ijkey=9f35e31bf918...

This just seems to be updated with a slightly older population.

View on HN

Comparing background radiation to exposure from artificial radiation sources does not sound right to me. The background radiation is absorbed relatively evenly by the body surface. However, various artificial radiation exposure is local, such as dental x-rays or CT, which, for example, will increase the thyroid cancer risk at seemingly small exposure levels compared to cosmic radiation exposure.

I wondered if the high safety standards in nuclear plants was affecting the outcome. If industrial workers aren't dying in things like construction accidents then they have to die of something else and sometimes that will be cancer. Because that graph does make it look like being employed in the nuclear industry increases cancer risk even without any radiation dose at all. It isn't though, because it looks like we're talking about ~10,000 otherwise unexplained deaths over ~300,000 people and even being pessimistic I think industrial accidents would only be <50 deaths. I wouldn't completely rule out some industry-specific effect like that even now, looking at the figure you reference.

The nature of the study make it hard to be confident, there are a lot of potential influences that could throw the conclusions. It is challenging to interpret too, table C in the supplementary data makes it look like risk drops with higher doses and I am sure I'm reading that wrong.

Thanks for linking the graph, that's kind of wild. I agree with you that the lowest datapoint seems crazy. I can think of a few explanations.

- Random bad luck.

- As you say, failing to control for something -- although, if you then treat the lowest datapoint as being effectively the default risk, this would suggest support for radiation hormesis (that people who got a bit more than background radiation actually did better.)

- Some kind of data collection artifact. Perhaps the people with the absolute lowest dose, in a radiation-worker dataset, are selected for being ones who are not getting an accurate measurement (i.e. sloppy about wearing dose badges or something), and those people genuinely do have worse outcomes.

Reply on news.ycombinator.com