LNT theories - Which model??

[FIELDRW@aol.com]

Prof. Raabe,

In response to your reply (to my previous question) below, I appreciate and 
respect Dr. Cohen's attempts to test the validity of the LNT.  The validity 
of the LNT is an extremely important question.  I have already had a debate 
with Dr. Cohen, which is well documented in the scientitific literature.  
Obviously, I do not feel that his study design is appropriate for testing the 
validity of either the BEIR IV or BEIR VI LNT model.  

However, you did not respond to my previous question, which was, "Which LNT 
theory do you feel Dr. Cohen is testing?"  I have previously heard Dr. 
Cohen's response to my question.  I would like to hear your response, 
especially given your interpretation of his work. 

Just in case you did not understand my previous question, I will ask the 
question in a different way.  Do you feel Dr. Cohen is testing either (or 
both) the BEIR IV OR BEIR VI LNT model?  

Regards, Bill Field

R. William Field, Ph.D.
College of Public Health
Department of Epidemiology
N222 Oakdale Hall
University of Iowa
Iowa City, Iowa 52242

319-335-4413 (phone)
319-335-4748 (fax)
mailto:bill-field@uiowa.edu
******************************


In a message dated 12/8/99 12:25:58 PM Central Standard Time, blc+@pitt.edu 
writes:

<< On Wed, 8 Dec 1999, Otto G. Raabe wrote:
 > 
 > Recognizing these limitations, Cohen has tested the linear model that
 > assumes that the occurrence of lung cancer as a result of exposure to radon
 > in homes is proportional to the average radon air concentration in these
 > homes. He has shown that it is not. His results cannot be discounted as
 > caused by the ecological fallacy since, as we can see inductively: if the
 > LNT model that he is testing is actually correct, then the distribution of
 > the radon levels among the exposed persons at a given average concentration
 > would not alter the overall occurrence of lung cancer. On the other hand,
 > if the LNT model that he is testing is wrong (as he has shown), then the
 > shape of the dose response curve that he has observed is not definitive and
 > it becomes apparent that we need to known dose to each person to quantify
 > the risk and the shape of the curve. Hence, the downward slope of his dose
 > response observation may be an artifact.
 > 
 > This is my interpretation of Cohen's test of LNT. I hope Dr. Cohen will
 > comment.
 
    --This is correct and is what I have been saying always,although I
 do not like the word "artifact". The negative slope in the dependence of
 lung cancer rates in U.S. counties on average radon exposure in those
 counties is certainly true, but it cannot be
 interpreted as a risk to individuals vs their cancer risk if the linear-no
 threshold theory fails (as I have shown). To understand this, consider the
 situation if there were a sharp threshold for radon-induced lung cancer:
 the lung cancer rates in counties would depend on the fraction of the
 population exposed above the threshold, which might vary from county to
 county in a very different way than the average exposures.
    In summary, my data shows that linear-no threshold theory fails
 very badly, grossly exaggerating the risks of low level radiation. (The
 last phrase requires more justification, which I can provide if this is an
 issue). I have never claimed that my data demonstrate hormesis which would
 be the interpretation of a negative slope in risk vs dose to individuals. >>
************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html