Suggestion for Dr. Cohen

[FIELDRW@aol.com]

Dr. Cohen,

I certainly do not feel you adequately addressed our concern #2 below or our 
previous concerns in the joint publications. I am still puzzled why your 
estimated mean county smoking rates are so inversely related to your mean 
county radon concentrations?   However, in keeping with the spirit of your 
most recent request, I will rephrase one of my concerns according to your 
format.   

I suggest that your observation that lung cancer rates in U.S. counties 
decreases with increasing radon exposure in those counties, in sharp contrast 
to the prediction of BEIR-IV (or BEIR VI) linear-no threshold theory that 
lung cancer rates should increase might be explained by the fact that you are 
not testing the BEIR-IV (or BEIR VI) linear-no threshold theory.    
 
Respectfully,

Bill Field

R. William Field, Ph.D.
College of Public Health
Research Scientist - Department of Epidemiology
Adjunct Professor - Department of Occupational and 
Environmental Health
N222 Oakdale Hall
University of Iowa
Iowa City, Iowa 52242

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

    Some examples of X-Y-Z that would receive this reward if it were
retroactive (which it is not) are:
    1. Lubin's suggestion (Health Physics 75:4-10;1998) that
individuals who smoke may have lower average radon exposures than
individuals who do not smoke
    2. The suggestions by Smith, Field, and Lynch (Health Physics
75:11-17;1998)that exposures outside the home were ignored, that % of time
spent inside the home may vary, and that measurements were made of radon
gas whereas lung cancer is caused by radon progeny.
    3. Archer's suggestion (Health Physics 75:652-653;1998) that total
county population may be an important confounding factor because it is
strongly correlated with both radon levels and lung cancer rates, for
unrelated reasons.
    4. Goldsmith's suggestions (Health Physics 76:553-557;1999) that
the radon exposure data may not be reliable, and that population density
may be an important confounding factor.
    5. The BEIR-VI suggestion that no account was taken of variations
in the intensity of smoking, as 1 pack/day vs 2 pack/day.
    6. Doll's suggestion (J. Radiol. Prot. 19:65;1999) that no account
was taken of the proportion of ex-smokers who have stopped for different
periods, the proportion of smokers who smoked cigarettes, cigars, and
pipes, and the age at which smoking began.

     Incidently, I believe that my published analyses of suggestions
1-6 above showed that these explanations were completely implausible.
Others can judge whether or not they agree, but that is irrelevant to the
reward issue. I just want opportunities to analyze suggested explanations
of our discrepancy, and am willing to pay for such opportunities. I
believe that it is only by accumulating suggestions and their analyses
that the issue of validity of my test of the linear-no threshold theory
can be resolved. 


Bernard L. Cohen
Physics Dept.
University of Pittsburgh
Pittsburgh, PA 15260
Tel: (412)624-9245
Fax: (412)624-9163
e-mail: blc+@pitt.edu

************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html