Scientific Jigsaw Puzzle: Fitting the Pieces of the Low-Level Radiation Debate

New submitter Lasrick writes "Skip past the dry abstract to Jan Beyea's main article for a thorough exploration of what's wrong with current 'safe' levels of low-level radiation exposure. The Bulletin is just releasing its 'Radiation Issue,' which is available for free for two weeks. It explores how the NRC may be changing recommended safe dosages, and how the studies for prolonged exposure have, until recently, been based on one-time exposures (Hiroshima, etc.). New epidemiological studies on prolonged exposure (medical exposures, worker exposures, etc.) are more accurate and tell a different tale. This is a long article, but reads well." Here's the free, downloadable PDF version, too.

Comments

Short summary

[Anonymous Coward]

Ionizing radiation causes cancer. More ionizing radiation causes more cancer. There is no "safe dose", though there is a certain unavoidable dose. So we're all at risk of cancer if we live long enough.

Re:Short summary

[History's Coming To]

Precisely - "low-level" at Sellafield in the UK used to mean "lower than the background level", and people still got hysterical about it. We need to stop with the wooly-language descriptions and simply use the established units, or units-above-background where applicable. Is low level gamma worse than high level alpha? Is holding a piece of uranium for 5 minutes more or less dangerous than sleeping 10ft away from it for a week? People have no idea, including most of the media, we need to throw out the "levels" model and actually educate people so they can understand the risks properly.

Re:Short summary

[vlm]

Ionizing radiation causes cancer... There is no "safe dose"

You seem to have not read the abstract, the whole point of which is at ultra unrealistically low levels, practically homeopathic low levels, the mechanism, the cause/effect seems to not make much sense or is under debate, both real scientific debate and crackpot astroturfing debate. But the article points out that at any realistic dosage level there is not much debate by anybody. So the article pragmatically suggests to only apply real world numbers to real world exposures and ignore the whole topic of unattainably asymptotic low levels. The article argument is the opposite of yours in some ways.

An example of a realistic question at the ultra-low end is, looking at how naturally radioactive some of our high potassium food is, you'd think we'd evolve a way to pee the bad stuff away. Presumably people evolved in granitic-source / volcanic-source soil would be better at it than people evolved in sedimentary-source soil. Another realistic area of cancer research is proving the presence or absence of two-step or catalysts of cancer. Your body is pretty good at dealing with mutant cells, except when it fails and then you die of cancer, whoops. So figuring out why your body fails to kill cancer cells is in many ways more important than trying to figure out how to reduce the number of cells caused by radiation because even if you zero that, you're still going to have random biochemical accidents. Its an interesting theoretical area of research but the article points out for normal human beings its at a level that doesn't matter.

Re:Short summary

[mhajicek]

Bananas.

Re:Short summary

[MozeeToby]

Exposure is always expressed in amounts over the background rate. So "Lower than background level" effectively means exposure to up to 2x the background level (background level + artificial); there's nothing illogical about being worried about it (though I wouldn't personally be concerned about a ~.0025 Sv per year exposure rate).

As for the rest of your comment, if you read the paper the summary links to, you'll see that all the evidence is pointing toward all exposure (presumably below radiation poisoning levels) carrying approximately the same relative risk. It doesn't matter high or low energy, it doesn't matter if you're exposed in 10 minutes or 10 years. Your total exposure level linearly maps to your risk of cancer (and, new information to me at least, heart attack and stroke).

Re:Short summary

[vlm]

Ah I'm not talking about cancer being like evolution, talking about evolution if you live in a niche of really high radioactive potassium consumption from eating bananas all the time, after a bazillion generations you'd expect the survivors to be better than the average human about excreting radioactive or otherwise K and/or getting by with as little of that nasty stuff inside them as possible, despite it being a big part of their diet.

Re:Short summary

[History's Coming To]

Your best bet is to read a high-school level introduction. Concisely, there are three types of radiation from radioactive atoms, alpha, beta and gamma. Alpha is a He nucleus, two protons and two neutrons - it can do a huge amount of damage to living cells, but is easily stopped by, eg, a sheet of paper. Beta is a high speed electron, less damaging but will penetrate clothing etc. Gamma is nasty - it can travel through a reasonable thickness of lead and still do harm.

If we look at the Uranium example, it gives off alpha, so you'd probably be quite safe with it on the other side of the room. Handling it, on the other hand, is an easy way to accidentally ingest some, which would probably be more harmful because it's then inside the body (this goes for any ionizing radiation source). When you see people being showered off after radiation exposure it doesn't stop any harm thats already been done, just reduces the chances that they are still in contact with a source.

This all ignores the fact that Uranium decays into several other isotopes which give off their own idiosyncratic radiation in turn, and a bunch of other things.

Low level radiation

[girlintraining]

There is no threshold below which radiation is 'safe'. There is a threshold below which is become statistically indistinguishable from random events, but that is not the same thing. We've known even "low" levels of radiation can be dangerous -- look at the cancer clusters showing up in TSA screeners. The scanners were declared 'absolutely safe' and had a 'low' level of radiation. There is a long history in the medical field of radiology where equipment, engineering, or our understanding of underlying principles failed and led to death or serious injury. The fact is, there is no such thing as "safe". That doesn't mean don't use the equipment -- it's often the only way to get the information needed (note: full body scanners NOT included, there are alternatives which provide the same information). But it does mean use the least amount of radiation necessary, only use it when necessary, and carefully track a person's exposure -- time, dosage, etc., to identify trends.

Radiation is a daily reality in our lives. Go outside, look up. There it is; the biggest source of radiation in your life (most likely). We can't avoid it... but we can limit it.

Re:Low level radiation

[InterGuru]

There might be a level at which radiation is beneficial. This is called hormesis

From Wikipedia

Hormesis (from Greek hórmsis ...) is the term for generally favorable biological responses to low exposures to toxins and other stressors. A pollutant or toxin showing hormesis thus has the opposite effect in small doses as in large doses

The concept is vigorously debated, but has been shown to work in some animal experiments. In humans, small doses of alcohol, a toxin, seems to improve heart health.

Humans, as all life, have evolved under low level background radiation. We may be adapted to it.

Re:Low level radiation

[XiaoMing]

There is no threshold below which radiation is 'safe'. There is a threshold below which is become statistically indistinguishable from random events, but that is not the same thing. We've known even "low" levels of radiation can be dangerous -- look at the cancer clusters showing up in TSA screeners.

Unfortunately, what you say is at best inconclusive, but at worst wrong. Google "hormesis".
Studies "including for example the respected "Iowa Radon Lung Cancer Study" of Field et al. (2000), which also used sophisticated radon exposure dosimetry....argue that radon exposure is negatively correlated with the tendency to smoke and environmental studies need to accurately control for this; people living in urban areas where smoking rates are higher usually have lower levels of radon exposure due the increased prevalence of multi-story dwellings".
http://en.wikipedia.org/wiki/Radiation_hormesis

I know that hormesis sounds like a crackpot theory along with holistic super-diluted medicinal honey therapy, but some of the greatest minds in Medical Physics believe it exists. It is basically the hypothesis that low levels of additional radiation can actually make you healthier than no additional radiation at all (including daily dosage of cosmic rays). Hence the quote about high background radon studies and inverse correlations with health outcomes.

One of the main mechanisms that is thought to possibly explain it is that while the additional radiation exposure is not enough to cause significant DNA damage, it still activates certain dormant mechanisms for DNA repair, resulting in a healthier-than-average individual.

So in short, there is at least very suggestive evidence for a "safe" (and even moreso than safe) level of radiation.

Re:Low level radiation

[rgbatduke]

I think that you aren't giving the possibility of positive reaction mechanisms enough credit. Humans have co-evolved with alcohol for at least 6 to 10 thousand years. Over the overwhelming bulk of that time, if you did not drink alcohol your life was ugly, nasty, brutish, parasite ridden, and short. There is an interesting program on Netflix you might want to watch entitled "How Beer Saved the World" -- tongue in cheek but not really. It's really only been safe to drink the water for less than 100 years, in some levels of wealthy and scientifically advanced society, in countries where it is safe to drink the water, which isn't most countries even now, presuming you think drinking halogenated water is "safe". I grew up in India, and used to drink beer on the road when we travelled at age seven or eight, because it was one of precisely three safe options once you ran out of boiled water or iodine tablets. Tea (boiled water, no milk). Coca Cola -- because even if you dropped a cockroach into Coke as it was bottled, you'd just eat/drink down an acid-pickled cockroach and be perfectly fine. And beer. Golden Eagle beer, to be specific, is the earliest beer I can recall tasting. Back then they didn't have bottled water for sale.

For the most part the body metabolizes alcohol in moderation harmlessly. It isn't particularly directly toxic to the liver (although fermentation adjucts may be), it's just that the liver tends to get fat, just as it does (as you observe) if you eat enough carbs or the wrong sugars and have metabolic syndrome or type 2 diabetes. Alcohol acts as a mild blood thinner -- not unlike aspirin, but not as strong -- and hence may be directly beneficial at moderate levels for precisely the same reasons that aspirin is (and aspirin has its own toxicity and side effect issues, although they are rare in adults). As you note, it is a fairly harmless relaxant. The Mayo clinic lists it -- with warnings -- as being "possibly good for" reducing risk of heart disease, dying of a heart attack, risk of strokes (especially ischemic strokes), lowers your risk of gallstones (my grandmother was prescribed one beer a day, which she drank very religiously and dutifully being the wife of a Methodist Minister who did not hold with drinking, for this very reason, way back in the 1960s, as an alternative to taking a wad of horrendous-sized pills), and diabetes. Their guideline is one drink for women and two for men, but of course this depends on body size. Women are at greater risk then men (relative to any benefits) because of their higher risk of breast cancer, BTW.

As for statistical studies, they are how one proves that prayer and astrology do not work. You got that one backwards, thought I'd help you out. You're thinking of "anecdotal evidence", not double blind placebo controlled statistical studies. Even in physics (where I'm a physicist) correlation may not be causality but it is often all one has until one maybe eventually formulates a theory that might explain it, and that theory has as its ultimate foundation what? Evidence in the form of statistical correlation, of course. What else is there?

With that said, given the mass of Bayesian priors (a.k.a. "laws of nature" and the like) we have arrived at that are reasonably statistically sound, I totally agree that one should look for reaction mechanisms and explanations, but don't forget what they are explaining -- the statistically sound results obtained from the data. On a really good day, you come up with both the mechanism and the data and they are consistent and the mechanism predicts other things as well and then you get your Nobel prize and everything. Other days you are up against multifactorial effects and sparse data and trying to make sound inferences of cause is, well, "challenging" even though the statistical correspondence itself may be as sound as you like.

rgb

protracted exposure

[sl4shd0rk]

Like fallout from nuclear testing and nuclear disasters.

Cost benefit analysis

[hoggoth]

Low level radiation may be dangerous, but we have to weight that against the benefit to the corporations that sell airport scanners. Some amount of harm is worth it.

Anti-nuclear publication

[bradley13]

Do note that the "Bulletin of the Atomic Scientists" is a generally an anti-nuclear, scare-mongering publication. These are the people whose count-down to nuclear disaster has been just a few minutes before midnight for decades. Whatever they publish should be viewed with this in mind.

Scanning RFTA, in the end, it says basically nothing at all. They did no studies themselves, but just looked around at ones already done. The key points seem to be:

• The same total exposure in the form of long-term exposure may be slightly (20%) more dangerous than the same dosage in a short, high-intensity form.
• They desperately search for something to say about low-level, long-term exposure. They spend pages talking about the competing theories, from "Supralinear response" (really dangerous) to "Adaptive response" (a little radiation is healthy). In the end, they find no convincing evidence one way or the other, because the uncertainty bounds at such low levels include essentially all possibilities.
• Based on this lack of evidence, they conclude that low levels of radiation are really dangerous, and that all Western populations are "primed for radiation-induced, delayed cancers from releases from nuclear reactors".

In the end, given the publication, the conclusion was obvious.

Re:Anti-nuclear publication

[Tyler Durden]

Do note that the "Bulletin of the Atomic Scientists" is a generally an anti-nuclear, scare-mongering publication

Utter nonsense. I remember perusing the print version of the Bulleting in my college library a few years ago, and it was anything but a knee-jerk, "scare-mongering" publication on nuclear issues. The articles were extremely informed and detailed.

There are two great articles that spring to mind. One was regarding a project run by the US government regarding how difficult it would be for countries without nuclear weapons to develop one. To test this, they found a physicist who had just gotten his PhD, making sure he that he wasn't someone with two much particular knowledge on nuclear physics. By using research from publicly-available sources he was able to eventually come up with a working design for a nuclear weapon. Just to be thorough he even designed a more complicated implosion design rather than a the simpler bullet design. The point of the article was that the difficult part for a country aspiring to create a nuclear arsenal is accumulating the proper uranium or plutonium. Creating the bomb is relatively simple.

The other article examined whether using depleted uranium for ammunition had lasting effects because of radioactivity. If I recall correctly, the radioactive aspect was not a concern. However, uranium can be poisonous without any consideration of its (limited) radioactivity. Since DU rounds piercing armor can cause the outer shell of them to vaporize, this could be a problem.

The Bulletin's conclusion was not obvious. Judging them just because of the Doomsday Clock is rash.

Re:Anti-nuclear publication

[imjustmatthew]

Do note that the "Bulletin of the Atomic Scientists" is a generally an anti-nuclear, scare-mongering publication. These are the people whose count-down to nuclear disaster has been just a few minutes before midnight for decades. Whatever they publish should be viewed with this in mind.

As a strong supporter of nuclear power I feel this attitude is exactly what makes it so easy to scare up opposition to nuclear power. That article was extremely well written and researched. IMO it presented a fairly balanced view of the existing studies and the overall challenges to new research and regulation. Yes their are concerns about low and protracted doses, and yes the industry has tried to downplay and bury that research. Just like the "green power" industry doesn't want anyone to look at the lifecycle costs on those PV cells and LiPo batteries.

The only way to stop fear mongering and get new power plants is with open and honest research - not making attacks on an article that tries to present the facts.

Re:Anti-nuclear publication

[Tyler Durden]

In other words, they rediscovered something that any professional or serious amateur in the field has known for decades.

Umm. The Bulletin is not, nor does not pretend to be, a scientific journal. Many serious amateurs or all professionals would not use it as a source for the latest information in nuclear science. And whether the conclusion of the article is known by the reader or not is completely irrelevant. The story behind it would definitely be of interest to a serious amateur or professional.

The conclusion regarding depleted uranium ammunition may be obvious to you, but I remember the mainstream media of the time had a lot of knee-jerk scare-stories regarding the harmful effects from "all that radiation". Since the Bulletin had a more informed and balanced article on the topic than other sources at the time, this means that the OP's assessment of it as "scare-mongering" in regards to all things nuclear (and your defense of him) is wrong.

But when you're not completely missing the point I'm sure you can come up with a post with more substance than insults.

Re:Extended exposure is riskier, and no superpower

[vlm]

The standard /. car analogy breaks down in that running my car engine up to 80% of redline RPM for a half hour a day is a pretty stupid idea that will only wear it out faster. Yet daily aerobic exercise seems to be a brilliant idea for long term cardiovascular health.

You can also have hilarious fun making vaccine analogies. "You mean, you'd intentionally inject small amounts of possibly fatal microbes into a healthy body? Madness I tell you! Madness!" Sadly there are highly educated actresses and pr0n models who pretty much use this argument when providing their valuable medical advice, along with the usual folks doing the FUD-for-profit thing.

Bad article, little information [Re:Short summary]

[Geoffrey.landis]

Ionizing radiation causes cancer. More ionizing radiation causes more cancer.

Of course. The question is, how much more cancer is caused by a given dose of radiation?

Unfortunately, this is a question that the paper in question does not answer, because it completely neglects to mention actual numbers. (The pretty colored graphs have units of "excess relative risk." How do you convert that to deaths? You can't. What are the units-- per year? Per lifetime? they don't say. Relative to what? They don't say.) I'd like to see a number, like "excess cancers per year per sievert of exposure," but they don't give one. They compare different studies, but never discuss whether the differences are statistically significant.

There is no "safe dose", though there is a certain unavoidable dose.

That is a question. That is what is known as the "linear no threshold" model-- but although these authors assert the validity of that model, you can't tell it from the data they show. Figure 1 shows too much scatter below 0.3 Sv to give much information about thresholds, and Figure 2 sure looks like it would be well fit by a threshold model.

In short, I'd like to have seen an article with real information.

Re:Bad article, little information [Re:Short summa

[AdamHaun]

Of course. The question is, how much more cancer is caused by a given dose of radiation?

Unfortunately, this is a question that the paper in question does not answer, because it completely neglects to mention actual numbers. (The pretty colored graphs have units of "excess relative risk." How do you convert that to deaths? You can't. What are the units-- per year? Per lifetime? they don't say. Relative to what? They don't say.) I'd like to see a number, like "excess cancers per year per sievert of exposure," but they don't give one. They compare different studies, but never discuss whether the differences are statistically significant.

As the article states, the graph is taken from another study, Preston et al (2007) Solid Cancer Incidence in Atomic Bomb Survivors: 1958–1998. You can find many tables with actual numbers there. The caption on the graph also answers some of your questions:

FIG. 3. Solid cancer dose–response function. The thick solid line is
the fitted linear gender-averaged excess relative risk (ERR) dose response
at age 70 after exposure at age 30 based on data in the 0- to 2-Gy dose
range. The points are non-parametric estimates of the ERR in dose categories.
The thick dashed line is a nonparametric smooth of the categoryspecific
estimates and the thin dashed lines are one standard error above
and below this smooth.

Altitude [Re:Bad article, little information]

[Geoffrey.landis]

Well, it would correlate with latitude as well.

That is, if cosmic radiation were in fact the main location-dependent factor that caused cancer.

But since cosmic radiation dose is something on the order of 0.5 millisievert per year, it's probably not significant enough to see the signal over the noise, assuming that there are other sources of cancer.

Re:Altitude [Re:Bad article, little information]

[HornWumpus]

They have the whole population as a sample. Cosmic radiation is typically about 10% of background. Deltas are fairly large. No problem finding the UV related cancers.

Also note: There are geographic areas with high Radon levels etc. None have been found to have higher then average cancer rates.

Yet they boldly assert they have proven there is no threshold. Show me the data!

Berkely Lab study suggests LNT is wrong

[JSBiff]

From what I understand, this is not absolutely definitive, but cancer researchers at Lawrence Berkeley Lab published a paper where they used imaging of cellular responses to radiation damage to show that at low levels, it appears that cells repair DNA damage due to radiation very effectively.

Seriously, follow that link, and learn.