Re: Re: Risks of low level radiation - New Scientist Article

[BERNARD L COHEN <blc+@PITT.EDU>]

On Mon, 17 Dec 2001, John Williams wrote:



> Dr. Cohen,

>

> This is my last post on this issue.  You said in a previous post that

> you would look at any area of the U.S. to see if your inverse

> association holds.  You make statements and don't back them up.  I

> just want to again document that you have been asked to do this yet

> again.



	--My results for each area of U.S. and for each state are given in

my published papers, most recently in

	Updates and extensions to tests of the linear-no threshold theory.

 		Technology 7:657-672;2000

I am sure I have stated this any time I was challenged. The above paper is

also posted on my web site at 	www.phyast.pitt.edu/~blc



	Members of RADSAFE are probably all-too-aware of the fact that I

do not ignore challenges to my research.



> Why have you not responded to numerous posts and publications

> by Field and others to provide a definitive explanation why your

> inverse association does not hold in Iowa when they used all your

> data, and merely updated your lung cancer incidence data?



	--I have responded to every publication by Field et al and to

every one of their e-mail messages. References to my published responses

are in my publication list on my web site under "curriculum vita".

	 The difference between my results in

Iowa and those of Field et al is that they used incidence data and I used

mortality data -- why is incidence data better than mortality data? In any

case, in my data for Iowa the inverse association is not statistically

significant. Another factor is that the time periods covered by my

mortality data are not the same as those covered by the SEER data.



>  The

> inverse association disappeared when they used better SEER data.



	--If the SEER data is better, I would like to know why. I

could not use SEER data because my study covered all states and Seer data

is not available for all states



   If

> epidemiologists are too believe your findings for the whole country,

> they should hold for a state with the highest radon, finest gradation

> of ecologic units (1/16 of the total units you use)and Iowa only has

> 1% of the U.S. population.



	--If there is something you don't "believe" about my findings,

please tell me what specifically. Or better still, submit a letter to the

Editor about it.



> The only answer I saw that you gave was

> that there must be some minority population causing Smith et al

> findings.  I do not find it plausible given the lack of a large

> ethnic minority population in Iowa.  Don't say you can explain it for

> any geographic region when you failed to respond to Field's challenge.



	--I did not say anything like that. I said that there are problems

with data for a single state that are smoothed over when considering the

nation as a whole. I believe I gave several potential examples, including

the one you mention. A few other examples are statistical variations,

correlations between medical facilities and radon levels, correlations

between radon levels and other causes of lung cancer or other factors that

influence lung cancer rates, etc. If you would read my papers, you would

see that there is substantial variation in lung cancer vs radon

regressions in the various states, presumably due to such factors. But if

you average the results for the various states, the result is similar to

that for the nation as a whole.



************************************************************************

You are currently subscribed to the Radsafe mailing list. To unsubscribe,

send an e-mail to Majordomo@list.vanderbilt.edu  Put the text "unsubscribe

radsafe" (no quote marks) in the body of the e-mail, with no subject line.