Dr. Lubin's response to Drs. Cohen and Isenhower.

[FIELDRW@aol.com]

In a message dated 3/3/00 11:14:25 AM Central Standard Time, blc+@pitt.edu 
writes:

The only point that Lubin seems to ignore here is that my study was not 
claimed to give the risk to individuals -- it was only offered as a test of 
the linear-no threshold theory (LNT) and LNT grossly failed in that test. 

-----------------
Dr. Lubin asked me to post his response to Drs. Cohen and Isenhower.  
I thank the list for their patience.  I know this thread is not of interest 
to everyone. A good bit of the discussion has moved to private email.  I hope 
the subject heading is descriptive enough so those folks, who are not 
interested can delete the messages.  Bill Field        
mailto:bill-field@uiowa.edu   The response follows:
----
If it is true, as Dr. Cohen states, that his "study was not claimed to give 
the
risk to individuals", then by definition no valid inference about individuals
can be drawn from his results.  I agree with this entirely.  However, the
logical consequence of this admission by Dr. Cohen must be that the form of 
the
regression relationship of county lung cancer rates on county characteristics
carries no implication for disease patterns for individuals.  Consequently, 
the
allegation that there is no LNT relationship between county mortality rates 
for
lung cancer and mean county radon concentrations - whether true or not -
provides no information about the nature of the relationship between lung 
cancer
risk and radon exposure for an individual.  However, the opposite is also 
true.
If I specify a functional relationship for lung cancer and radon for
individuals, then you cannot "scale up" that function (unless it's linear in 
all
risk factor, which is not the case with radon) to the county level, unless 
there
is information on the joint distribution of all risk factors for all counties.
Thus, Cohen's "test" equally carries no inference for individuals.  As a 
health
scientist, my concern is with the increasing lung cancer risk with increasing
radon exposure in individuals and not with cancer rates in counties.

Finally, regarding the comments by Dr. Isenhower, the discussion about
ecological studies is not about "scientific theories", but about relevance.  
The
question is whether Cohen's results are relevant - my claim is that they are
not.

Jay Lubin 


Jay Lubin, PhD
National Cancer Institute
Biostatistics Branch, EPS/8042
6120 Executive Blvd
Bethesda, MD 20892-7244
Tel: 301-496-3357
Fax: 301-402-0081
Email: lubinj@exchange.nih.gov <mailto:lubinj@exchange.nih.gov> 
************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html