Thursday, July 24, 2008

Best Analogy for the Linear No-Threshold Theory

Many NEI NN readers here know that there are two competing theories on radiation: linear no-threshold (LNT) and hormesis. For those who don't, the LNT theory basically says that there is no safe dose of radiation whereas the hormesis theory says small doses of radiation are safe and that there is a threshold before radiation becomes harmful. Both are highly debatable but the latter makes a stronger case - at least in my opinion. (US nuclear plants, however, are regulated under the LNT theory.)

One of the best analogies that explains the LNT theory comes from a comment made by DV82XL over at Physical Insights:

If the LNT were applied to falling as it is to radiation, we might note that 100 percent of those falling onto concrete from 100 feet are killed, but only 50 percent of those falling from 50 feet die. With these data we would linearly extrapolate to say that 10 percent falling from 10 feet and one percent of those falling from one foot would die. Armed with this “linear no-threshold falling theory,” we could confidently assert that jumping rope should be banned on all school playgrounds since statistically anyone making 100 one-foot jumps would die.
Hat tip to Luke Weston.

16 comments:

Jim Muckerheide said...

30+ years ago we wondered how many people would die every day in NY city stepping off the curbs, and that we should check the NY Times obits every day (and to watch for gov't cover-ups of this mortality data).

Also, note that presuming jumping rope is at one foot constitutes a very conservative safety analysis. :-)

Brian said...

Hi David,

Actually, hormesis says that not only is there a threshold but some amount of radiation is beneficial. In between LNT is the linear-threshold model, which I believe is the one you meant to describe.

Although I tend to put more credibility on the threshold or a sublinear model, I caution the use of such analogies as critiques. Falls and radiation are fundamentally two different impactors on a biological system and I hesitate to draw a comparison.

The LNT model is basically an extrapolation of data. It is the best basis we have and it is conservative. Nonetheless, it fails to take into account WHY the effects are linear and does not consider non-linear biological repair mechanisms.

Kirk Sorensen said...

We've had a bit of an on-going discussion/argument on LNT going on on the thorium-forum:

http://www.energyfromthorium.com/forum/viewtopic.php?f=2&t=809

Brian Mays said...

Actually, DV8 is a little off on his statistics.

If we take the LNT as applied to falls, according to how he explains it, to mean that the chance of dying from a one-foot fall is 1%, then someone making 100 one-foot jumps would not be statistically certain to be dead. He or she would have only a 63% chance of being dead. That's a 1 in 3 chance of still being alive from this "highly risky" behavior.

In fact, if the LNT jump-rope theory is to be applied even reasonably conservatively, then the number of jumps should be limited to only 68, since after that one is statistically more likely to be dead from the activity than to be alive (i.e., > 50% chance of death). If we are willing to accept a 1 in 5 odds ratio (i.e., 20% of those skipping rope will die), limit the jumps to only 22.

His main point is right on target, however, and this all serves to show that overly simplistic assumptions -- without supporting evidence and sometimes in the face of contradictory evidence -- lead to extremely unrealistic results, sometimes laughably so.

DV8 2XL said...

Oh dear - This passage I am afraid is not mine. It belongs to Ed Hiserodt. I posted it to Depleted Cranium, and in haste forgot to attribute it to the original author. My apologies to him and those that re-quoted it thinking it was by me.

I have used it so often in verbal conversation that I hardly think of where I picked it up. Quite inexcusable, I know.

Charles Barton said...

A classic case of bad science. There is Powerful empirical evidence against the LNT as applied to workers at and neighbors of power reactors. All of gets tossed out the window. Anti-nuclear advocates commit a further absurdity by ignoring radiation from natural sources, and radiation exposures from non-nuclear human activities, like coal powered steam plants.

Ruth Sponsler said...

Very interesting post.

Exercise (by itself, assuming an adequate diet) is a perfect example of hormesis.

An hour's workout is good for you. Buh-bye - see ya tomorrow back at the gym. Two hours is a bit much or maybe "insane" but recoverable. See ya in a couple days...

Three hours of intense exercise is do-able in the sense of a marathon or similar, but takes a few days for recovery.

Mote than that in most cases is harmful.

NO exercise is also harmful.

Exercise generates free radicals and reactive oxygen species, which are biologically harmful. Yet...the person who never goes for a walk or bike and lies on the couch too long gets negative consequences from _too little_ exercise.

The body's response to the exercise stimulus, as long as the stimulus is moderate and not excessive, is beneficial. It seems that the body needs to have its repair mechanisms stimulated on a periodic basis. This is the case whether the exercise is aerobic (jogging, distance running, treadmill, bicycling) or anaerobic (sprinting, lifting weights, doing football team exercises).

Aerobic exercise in moderate quantities conditions the hesrt and cardiovascular system and burns fat, while anaerobic exercise in moderate amounts tears down muscle and stimulates the body to repair itself.

Both kinds of exercise are harmful if done to excess. Who would spend more than a couple hours lifting weights at the gym? Moderate cardio - like a long, slow bike ride - is more doable than marathon running.

Joe Average could tear up his shoulder lifting weights designed for Olympic medalists. Jane Average would get seriously dehydrated/fatigued if she tried to run a marathon without training for 26.2 miles of distance.

Radiation generates reactive oxygen species and free radicals as well. A little bit is OK and stimulates the body's repair mechanisms - but too much is like trying to bench press too many sets.

John Wheeler said...

The underlying basis for all nuclear plant regulation is the Linear No Threshold Theory, and it is a theory on the verge of being proven wrong! Can you Imagine....

... the incredible benefits to society that could be derived if regulatory bodies recognized the LNT theory for the "junk science" that it is and instead created regulations based on scientific observations of the affects of low exposures? The resultant changes to reactor design criteria would collapse emergency planning zones to the site boundaries, seismic design criteria would be virtually unnecessary, and the need for long-term geological repositories would be a thing of the past.

Nuclear plant construction would be cheaper because less concrete and steel would be needed. Less equipment and simpler designs would cause operation and maintenance costs to plummet. In all, I believe the cost of energy from nuclear plans would be cut in half or more.

I'm glad this issue is getting some air time. This single change to how the industry is regulated has more potential to benefit society than any other. It will be a difficult change, but it is one we can and should pursue.

Anonymous said...

Imagine what the public reaction will be like when the nuclear industry starts saying, "Hey, a little radiation isn't bad...in fact, it's GOOD for you!" Goodbye new plants. Overnight.

Anonymous said...

To the last anonymous, just because the majority of the "peepul" think radiation is bad doesn't make it so. A million men are no smarter than one man, and to this dictorship of the simple majority that we seem to have does nothing except prevent benefit to mankind. It's quite Demokratik, actually.

Anonymous said...

"just because the majority of the "peepul" think radiation is bad doesn't make it so."

No, but it certainly could affect the industry's ability to get new plants approved and built.

Look at all the opposition to Yucca Mountain because it might leak in 999,999 years. Now you want the industry to start saying radiation is good? It's PR suicide. Like Lewis Strauss's "too cheap to meter," we'll never hear the end of it.

robert merkel said...

While I appreciate that the evidence for LNT is very, very weak, as I understand it the people who study the health effects still take the view that it can't be definitively rejected.

Politically, it seems to me that the best argument is that the amount of radiation released to the environment is miniscule compared to natural sources. You can make the follow-up point that the evidence for the LNT model is almost non-existent, but most people are going to take the view that if you don't know, you should take a conservative approach to risk.

Anonymous said...

I've heard of another analogy regarding the LNT hypothesis as it is applied to estimating casualties from radiation releases, such as Chernobyl and TMI.

If 100 REM is sufficient to induce fatal cancer in one person, then 10 REM to 10 people would also statistically induce one fatal cancer among them. Extrapolating further, 1 millirem of exposure between 100,000 people should also statistically result in one radiation-induced fatal cancer.

It is possible the LNT hypothesis is true -- that even a small amount of radiation can sometimes be responsible for eventually inducing cancer. I don't think it should be entirely discounted. I do agree though, that the theory, as applied in a legislative way, is flawed. Nothing in life is without risk, and radiation exposure is disproportionately regulated for the low danger it poses.

DV8 2XL said...

Referring to the LNT as a hypothesis is giving it too much weight; it is in fact a statement of ignorance that has never been supported by observation.

The LNT model was first considered in the 1940s purely on the theoretical grounds that a single hit by ionizing radiation on a single cell could cause chromosome damage that could cause a mutation or cancer without any hard evidence to support that contention. The justification for using the LNT model was that too many test animals or too much time would be needed to evaluate chronic dose rates. If the LNT model is correct, there is no "no observed adverse effect level" (NOAEL) for regulators to observe, thus officials responsible for public health can claim justification in calling for minimization of exposures to ionizing radiation. Note that this is tantamount to saying that avoiding sunlight is justified on the grounds that nobody will get sunburns in the dark. Added to this, during the Cold War a number of people promoted the LNT model in an attempt to discourage nearly all uses of nuclear weapons and nuclear power, and used it as leverage in their campaigns.

As a result the radiophobes and the politicians took a handy but false rule of thumb and enshrined it in law and regulation. The second problem, related, is that this results in a lot of stupid but expensive procedures where people and vendors can make a lot of money thus entrenching this false standard through special interests.

Rod Adams said...

I have to disagree with my good friend Robert on this one. In the real world, we all know that there is no such thing as perfect, but you would never hear someone advertising a Volvo say anything other than it is a safe car.

Airlines tell people that air travel is safe, self service gasoline stations are considered to be safe, and people have been repeated told that they are safe when they ride public transport in big cities.

By all existing standards, a low level of radiation exposure is safe. Simple, straightforward and true.

Of course there will be people who disagree. Explaining why they disagree and what the data show should be their problem and challenge, not ours.

Kirk Sorensen said...

Joe Romm's attacking Charles Barton over LNT now at his anti-nuclear blog "Climate Progress". Maybe it's progress for coal...

Delayer/denier Joe.