Sunday, February 10, 2008

The Linear-No-Threshold Hypothesis

In a recent article we discussed the BEIR VII report's conclusion with respect to the linear-no-threshold (LNT) hypothesis concerning low-level radiation's possible health effects. It's worthwhile to compare it with other reports' findings, all from professional organizations in the US.

First, here's the pertinent statement in BEIR VII
"At doses of 100 mSv or less, statistical limitations make it difficult to evaluate cancer risk in humans. A comprehensive review of available biological and biophysical data led the committee to conclude that the risk would continue in a linear fashion at lower doses without a threshold and that the smallest dose has the potential to cause a small increase in risk to humans." [A typical person in the US receives 3 milliSieverts per year.]

That's a tepid justification for retaining LNT, but compare that with the statement from the National Institutes of Health:

"It is very difficult to detect biologic effects in animals or people who are exposed to small doses of radiation. Based on studies in animals and in people exposed to large doses of radiation such as the atomic bomb survivors, scientists have made conservative estimates of what might be the largest doses that would be reasonably safe for a person over a lifetime. But these calculations are estimates only, based on mathematical models. Low-level exposures received by the general public have shown no link to cancer induction. Even so, the U.S. Government uses these estimates to set the limits on all potential exposures to radiation for workers in jobs that expose them to ionizing radiation. International experts and various scientific committees have, over the years, examined the massive body of knowledge about radiation effects in developing and refining radiation protection standards."

And with the statement from the Health Physics Society"

"There is substantial and convincing scientific evidence for health risks following high-dose exposures. However, below 5–10 rem (which includes occupational and environmental exposures), risks of health effects are either too small to be observed or are nonexistent."

"In view of the above, the Society has concluded that estimates of risk should be limited to individuals receiving a dose of 5 rem in one year or a lifetime dose of 10 rem in addition to natural background." [5 rems would be 50 milliSieverts.]

Professor Bernard Cohen goes on to estimate what would be the health effects of low-level exposures and compares them with other health risks, using the LNT model even though he shows in his analyses that it overstates the adverse effects and probably understates the beneficial (hormesis) effects of low-level radiation.

As an exercise we'll do something simple here. The BEIR report says ten million mSv would cause 1140 deaths. And it says that, on average, 304 million Americans receive 3 mSv per year, so the total would be 912 million mSv. So all of the radiation-induced deaths add up to 104,000 per year. Of that number, according to the report, 0.2% are due to nuclear energy, the rest mainly being due to natural radiation. If the LNT hypothesis is right, 208 deaths per year can be attributed to nuclear energy.

In comparison, every study done shows that tens of thousands of Americans die every year from the pollution generated by coal-fired power plants. The most comprehensive study done so far puts the range between 33,000 and 121,000 per year, just counting adults over 25. In 2006, according to DOE, coal generated 1930 billion KWH of electricity and nuclear generated 787 billion KWH, so if nuclear replaced coal an additional 510 deaths would take place, but at least 50,000 lives would be saved.

And all of the radiation-related deaths depend on a hypothesis that hasn't been proved and which specialized professionals don't believe.

Here's the kicker: Coal plants emit more than ten times as much radioactivity as nuclear power plants. If the LNT hypothesis were true, 5000 of the coal-related deaths would be avoided by converting to nuclear energy just because of reducing radioactive emissions.

If some form of renewable energy could provide full-time power, this might be a harder decision to make. As we saw in an earlier article, though, there aren't any that could.

So those are the two options. We can let over 50,000 Americans die every year from coal or we can switch to nuclear energy and start cleaning up the environment while minimizing the threat of global warming. What to do, what to do.


Jason Ribeiro said...

Regarding the "kicker" of nuclear plants emit 10 times less radioactivity compared to coal - Are you basing this one the concept of the LNT hypothesis for the sake of the argument?

I ask because from every credible source I've read, the background radiation at plants and distances (1000 yrds, etc) are virtually the same, meaning of course there is no additional detectable radiation emitting from the plant.

Or it wouldn't it be safe and accurate to say that coal plants emit 10 times more than normal background radiation in your opinion?

Red Craig said...

Jason, I think it would be totally misleading to claim that coal plants emit 10 times more than normal background radiation. The increment in radioactivity around coal plants is slight and certainly doesn't increase radioactivity by a factor of ten. The point made in the article is that the increment around nuclear plants is only a tenth as much.

Anti-nukes aren't impressed by comparisons between background radiation and the additional radiation that comes from ventilating the buildings. In their view, all radioactivity is lethal and an increment in radioactivity, no matter how small, causes additional deaths. They base this view on the LNT hypothesis.

There are many reasons to oppose coal plants. Radioactivity releases is way down the list. The information is included here to show the absurdity of the anti-nuclear position.

Anonymous said...

Cohen's anti LNT claims have been discredited.


“A number of criticisms have arisen regarding this use of an ecologic study (NAS 1999). Aside from the biological implausibility of the results and the apparent disagreement with the results from miner cohort studies and residential case-control studies (see below), the most serious of these revolve around the question of possible confounding with smoking, which contributes to a very high percentage of lung cancer cases. In particular, if radon levels were inversely correlated with smoking across counties, it is easy to see that one can have a spurious inverse correlation between average radon level and lung cancer rate. A more subtle bias can arise from the synergism between radon and smoking in causing lung cancer if smoking and radon levels are correlated within counties (Greenland and Robins 1994, Lubin 1998). Cohen has argued that the likely magnitude of these kinds of biases is too small to explain his negative correlation, and the controversy continues (Smith et al. 1998, Cohen 1998, Cohen 1998a, Lubin 1998a, Field et al. 1998, Goldsmith 1999). The BEIR VI committee sided with the critics and concluded that Cohen’s inverse correlation “was considered to have resulted from inherent limitations of the ecologic method” and “was considered to be an inappropriate basis for concluding that indoor radon is not a potential cause of lung cancer.” Most recently, Puskin (2003) found that Cohen’s radon levels have quantitatively similar, strongly negative correlations with cancer rates for cancers strongly linked to cigarette smoking, weaker negative correlations for certain cancers weakly dependent on smoking, and no such correlation for cancers not linked to smoking. These results support the hypothesis that the negative trend reported by Cohen for lung cancer can be largely accounted for by a negative correlation between smoking and radon levels across counties.”

Red Craig said...

Anonymous, thanks for taking the trouble to comment and for the link to the interesting EPA paper.

As I understand it, Cohen's work considered just radon while the EPA considered radon in conjunction with cigarette smoking. If I have that wrong, please correct me.

Continuing on the premise that my understanding is correct, the EPA's response can hardly be called a renunciation of Cohen's conclusions. It argues only that smokers could tend to live in places that have less radon and that would explain Cohen's negative correlation between radon and lung cancer. That's a very tenuous supposition. Is there a basis for believing smokers live in places with less radon?

The EPA paper mentions some studies that show a positive correlation between radon and lung cancer, Puskin's in particular. But there are studies that don't. For example, shows muddled results: higher rates for the lowest and the highest radon exposures. For the in-between exposures the rates are not consistent but they are lower. I'm not qualified to draw conclusions but this other paper supports Cohen's thesis more than it rebuts it.

It makes sense that smoking would change the parameters. Cohen doesn't say high exposures to radon are healthy, only that normal exposures aren't a significant threat. Cigarette smoking would naturally be expected to amplify the harmful effects of radon without amplifying any hormetic effects it might have, on top of its own toxic effects. And the same would be said for high radon exposures: it doesn't seem reasonable that hormetic effects could ever compensate for high exposures.

Cohen shows that even if the LNT hypothesis is correct, which he doesn't believe, nuclear energy saves many lives. I included a link to his book so interested readers could learn more about his analysis. Following his example, but at a more rudimentary level, I found the same result.

The Puskin links merely make the same argument as the EPA paper, that smokers live in low-radon areas. I'm puzzled by your mention of the Bond paper, in which the data plots clearly show a negative correlation between radon (at normal levels, of course) and lung cancer and a distinctly positive correlation for cigarette smoking. The authors of that paper declined to reach a conclusion about Cohen's thesis.