Replies

Here in Emmett, residents have spent the last several days making lists of relatives and neighbors - living and dead - with cancer. They have rushed to a local bakery, aptly named the Rumor Mill, where the owner, Tona Henderson, has produced a form letter of her own that residents can fill out and send to the National Academy of Sciences.
A committee of the academy is taking public comment for a study on nuclear fallout and public health to be submitted to the Department of Health and Human Services, said Bill Kearney, a spokesman for the academy. While many scientists and medical experts have said there is a connection between exposure to Iodine-131 and greater risk of thyroid disease and thyroid cancer, a link between the fallout and other diseases has not been established.
Still, in Emmett, dozens of residents have gathered in coffee shops and farmhouses to talk about cancer. Many furiously said they suspected their radiation exposure was connected to their cancers.

"This whole thing is wrong," said Richard Rynearson, 62, who is dying of colon and liver cancer, and who ran a heating and air-conditioning business until he became too sick to work. "Somebody needs to own up to the fact that they messed up."

Junk science alert.

Let's review the facts. The residents of this area got about 15 rads in the space of several years (some few may have been exposed to as much as 100 rads). A Rad is a unit of absorption, a Rem is a unit of exposure (a more accurate measure of how threatening the absorbed radiation is). The most dangerous radiation exposures tend to have 1 Rad = 1 Rem. We'll use that as our worst case scenario.

Short term exposures to radiation are the most dangerous, while long term exposure to low levels has no observably deleterious effect. Every American absorbs between .06 and .6 Rems of radiation per year just from his natural surroundings. Studies have shown that doses below 100 rems have no observable effect on cancer rates, and that doses as high as 300 rems are necessary to get significant (50%) cancer rates. So, what does this story tell us? Most exposures were far below the threshold for any effect at all, while a handful may have been exposed to the very lowest amount that might have an effect.

Second, epidemiology is a very strict science, and what we have is very anecdotal information in the story. Radioactive iodine causes mostly thyroid cancer, yet the residents blame many different varieties of cancer on the fallout. This is a perfect example of the post hoc, ergo propter hoc logical fallacy ("after this, therefore because of this"). Fact: nearly three in four Americans will get cancer of some kind and nearly one in four will die of it. Cancer is a disease of the elderly, and it is a natural function of the body's cellular systems breaking down and malfunctioning due to aging. Epidemiologists track disease and cancer threats based on quantifiable and measurable increases in effects. For example, when coupled with specific exposure to a known substance, higher rates of rare forms of cancer, cancer localized in patients (such as in the liver or thyroid), or cancer occuring at young ages all are valid indicators of a possible effect. But clusters of cancers occur in then population at random as well (that's what "random" means), so rigorous statistical methodology is necessary. Besides, someone has to be the person that gets the one-in-a-million cancer...

Continued below ->

Knowing this, re-read the article. What you have is one person that had thyroid cancer writing a letter, and groups of people now using hindsight (as opposed to careful study and statistics) to try and group a large number of disparate illnesses together. The story quickly glosses over the lack of substatiation from scientific sources (see the bold part of the quote), but that's because it is more interested in "scare" than "fact."

This is Love Canal or DDT redux. But expect our tax dollars to get drained to pay people whose cancers have nothing to do with the fallout, simply because big government wants to show it cares (and cover its own a** from the scientific illiterates out there who easily get stirred up about these enviro-scares).

But clusters of cancers occur in the population at random as well (that's what "random" means), so rigorous statistical methodology is necessary.

Once I was discussing one of these cancer clusters with a physician. He wrote it off as a natural statistical fluctuation. But this got me thinking. How can one tell if a cluster is a normal fluctuation or due to some definite cause? Suppose one computes the number of clusters of a given size to be expected in the US population. Perhaps the number is at least one. How do you then know that there are not other clusters that have not been accounted for? And how do you know that the underlying "random" cancer incidence level is not all or in part due to causes which are merely undetermined? The combination of the difficulty in analyzing this type of situation with the tendency of each side to want to bend the analysis to their pre-determined point-of-view must make it very difficult to get at the truth.