The Fear of Fukushima Radiation

The website What Doctors Don’t Tell You operates on the premise that there is a body of health knowledge that you are not going to find out from your doctor because they are ‘locked into conventional paradigms’, ignorant, in league with Big Pharma, or unlikely to make money from this knowledge. The mundane truth is that Doctors do not tell you the things on this website because they are wrong, misleading, conspiratorial or unnecessary. But the website is successful in driving that wedge between doctors and the patients who are prone to believe such things and, in so doing so, prime them to purchase a range of quack products and services. Today’s newsletter tells us that “Radiation from Fukushima meltdown will cause 2500 cancer cases”,

Radiation from the Fukushima nuclear disaster in Japan last year will claim around 1,300 deaths and cause 2,500 cases of cancer, say researchers.  This is in stark contrast to government assurances that the leak wouldn’t have any impact on health.

Although UK based, WDDTY have an American target audience. They warn that,

Around 80 per cent of the radioactive material was dumped into the Pacific Ocean, and some has washed up on the shores of North America and Europe.  The main death toll will still be in Japan, estimate Stanford University researchers, and there will be only a small risk to health among people living in Asia and North America.  Indeed, the researchers estimate that just 30 Americans will die from exposure to the dumped material.

A small industry has cropped up in the US with the likes of homeopaths selling their sugar pills to protect worried Americans against the dangers of Japanese radioactivity. Chief American homeopathy propagandist, Dana Ullman, has written in the Huffington Post advocating “Homeopathy For Radiation Poisoning”.

But what do the authors of the new research actually say? It is a somewhat different message from the “1,300 deaths”.

In fact, the best estimate from the paper is that there will be 130 extra deaths, ten times less than the ‘around 1,300’ that WDDTY want to scare you with.

So, where  do WDDTY get their figure from?

The paper* used a computer simulation of the global atmosphere to calculate what the total human exposure to the radioactive material that escaped from Fukushima after the tsunami hit the reactors. A lot of assumptions go into the calculations which means there are lots of uncertainties. Indeed, there are large ranges in the expected values,

We estimate an additional 130 (15–1100) cancer-related mortalities and 180 (24–1800) cancer-related morbidities incorporating uncertainties associated with the exposure–dose and dose–response models used in the study.

Even this is still short of the 1,300 deaths figure.  The paper looks at some assumptions about the rate of emissions, evacuation radiuses and the nature of the releases and concludes that such uncertainties could push the death figure up to 1,300.

So, the WDDTY claim that the tsunami disaster will “claim around 1,300 deaths” is not true. The upper limit, given a number of worst-case assumptions may mean that up to 1,300 have a fatal cancer. But the paper also notes this could be as low as 15 deaths. It is difficult to not conclude that WDDTY has picked the worst case figure and presented it as the likely figure in order to further its own aims at the expense of accuracy.

Let’s put these figures into perspective.

The Japanese Tsunami claimed nearly 18,000 lives. The attributable deaths to the reactor accident that add to this figure is, at the moment, zero.

Radiation-induced cancer can take decades to materialise. In the United States alone, there are annually 1,638,910 new cancer cases and 577,190 deaths. Over 40 years, and these extra deaths (and the fraction that would occur in the US and not in Japan) would have to be taken seen in the context of 23 million cancer deaths.

But there is a very big caveat even in these figures for cancer deaths from Fukushima. The modeling takes into account something called the linear no-threshold (LNT) model of human exposure. What this means is that the same quantity of human exposure to radiation has the same probability of inducing a cancer no matter how many people are exposed.

An example. Let us say that there will be on average 1 new cancer amongst 100 people exposed to 100 mSv. Now, let us take 1000 people and expose them to the same sum total of radiation. According to the LNT model, we would still expect 1 death even though now each person individually receives ten times less the dose.

It’s something of a counter-intuitive concept. If we give ten people 1000 paracetamol tablets shared amongst them, we are likely to end up with ten dead people.  Give a thousand people 1 tablet each and we would not expect any dead people. If paracetamol followed the same LNT model, we would still end up with ten dead people.

Part of the reason is the stochastic nature of harm from radiation exposure. That is, given an exposure to a certain amount of radiation (at high doses), the best we can say is that there is a probability you will get cancer. The thinking is that ionizing radiation damages DNA. Only if the DNA is damaged in a way that cannot be repaired properly is there a chance that the resultant damaged DNA will give rise to a cancer inducing mutation. Since, DNA will be damaged at random, the chances of you getting cancer are also probabilistic. The LNT model assumes we can measure the rates of cancer at high doses, given relatively quickly, and then extrapolate that back to minute doses given over long periods of time.

The researchers looking at cancer deaths are taking the amount of material released from the reactors and then trying to calulate how much of that ends up as causing radiation exposure in humans. Not all the material will end up being inhaled, ingested or landing near people. Much will end on the bottom of the sea, or will decay before reaching anywhere potentially harmful. Once the total dose has been calculated, then the LNT model allows an estimate to be made of how many deaths will follow, whether it exposed 1000 people or 100 million over several continents.

But there are good reasons to think this simple linear model does not extend down to the low doses seen as a result of nuclear accidents, occupational exposures and such activities as flying in aircraft. If an unrepairable break in DNA requires both strands to be broken at similar locations from two seperate ionising events, then we can see that low level events, happening infrequently, will give a cells repair mechanisms time to act before more damage occurs. This will result in a threshold, below which harmful mutation is near impossible, and above which the cells repair mechanisms can become overwhelmed by radiation induced damaging events.

Given the extreme difficulty of measuring tiny increases in cancer amongst large populations as a result of their radiation exposure, we simple do not know what the dose-response curve looks like at such low exposure levels. We can be pretty sure that the risk cannot be worse than the LNT model. This would mean that radiation doses became relatively more harmful as you reduced them and gave them over longer periods. Whilst a homeopath with skull water** for brains might not have an issue with this, the rest of us might see this as somewhat implausible.

The LNT model is used in a precautionary way by radiation health authorities. Many researchers are worried about the use of such a model as it gives rise to huge costs associated with nuclear safety. If we could assume that below certain levels, radiation adds no extra risk burden on societies then we need not spend large amounts of money preparing for contingencies that envisage large populations being exposed to relatively small amounts of radiation. It is quite easy to get silly about the LNT model. My professor at University calculated what the extra cancer risk was from sleeping with someone. (Humans are full of naturally occurring Potassium-40, a radioactive isotope with a half-life of about a billion years.) Being next to someone for eight hours a day will increase your total radiation exposure. Over populations of millions, this extra dose will result in cancers. The lesson? Do not sleep with people if you want to avoid cancer. (There may be other risks that turn out to be higher, such as you partner turning out to be an axe-wielding psychopath.)

Anyway, if you are interested in the sources of radiation you are exposed to and their relative amounts, I recommend XKCD as the definitive source.

Given this background, it is quite rational to make an argument that the extra radiation exposure to human populations from Fukushima  results in no extra deaths. The populations that should be of concern are those that were very close to the accident or were involved in the immediate clear up and rescue. Here, with relatively high doses given over short periods of time, the LNT model is probably quite valid. Homeopaths on the West Coast of America have other risks that ought to concern them – such as believing sugar pills can treat serious illness, or homeopathic sugar pills can prevent vaccine preventable disease.

Nuclear Power and its radiation have sunk deep into our collective consciousness as something that needs to be feared. In fact, ionising radiation is a very poor carcinogen. You need to be exposed to a lot of it before you might stand a chance of getting cancer. Other things are much better at causing cancer – like smoking, being overweight and vaccine preventable infections such as human papilloma virus.

But we will not see quacks focusing on these issues. The real money is in fear-mongering. And radiation is a great bogey man to scare your punters with.

* Worldwide health effects of the Fukushima Daiichi nuclear accident, John E. Ten Hoeve and Mark Z. Jacobson; Energy Environ. Sci., 2012, Advance Article DOI: 10.1039/C2EE22019A

** According to Robin Ince, Skull Water is what homeopaths have in their heads- it contains the memory of brains.