[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: LNT models -Constancy of Radon levels over time??



Bill,
Having read your preliminary article, I commend your study in many respects
because it addresses numerous potential confounding factors. By chosing
only women you mitigate the effect illegal drugs and inhalants may have. It
appears you'll also address medical effects, particularly pulmonary
history, which is likely a confounder. The study looks like well designed
case-control study and I believe it to be in very capable hands. Certainly
you'll explain why Iowa has one of the highest general radon concentration
among U.S. states, yet is much lower than the national average (39th of 51)
for lung cancer.

Case-control studies can be misleading even if they are well designed. The
corrections applied for confounding factors are mathematical assumptions
based on co-varying tests. The best test is a cohort study, preferably of
non-smokers. But as you've already pointed out, such a study is cost
prohibitive. If I were to design a study, I'd look at non-smokers. We know
non-smokers get lung cancer. Your study will no doubt show that. We also
know that non-smokers in areas with very low concentrations (e.g. tidewater
Virginia& FL) of radon also contract lung cancer. So if radon causes the
cancer in non-smokers in Iowa, what causes lung cancer in non-smokers in
areas low in radon. One logistical problem with a case-control study of
lung cancer for non-smokers is that there are so few of them.

Without prospective studies, I think Bernie Cohen's study fills a huge
void. If radon is a potent carcinogen, then its effect should be visible in
demographic data, regardless how coarse. His data should not be dismissed
off-hand. I understand that it does not have the rigorous parameters of
case-control or prospective studies. It should still raise some concern
that the model may not be an accurate description of reality. Further, as
I've noted previously, the collective case-control studies show a threshold
effect for radon up to about 350 Bq/m3, which agrees with Dr. Cohen's
conclusion from his test data.  

I would like nothing more than radon to be the second leading cause of
cancer. We could mitigate homes with elevated radon levels and lung cancer
could drop from 160,000 to either 157,000 or 130,000 depending on current
LNT projections. Realistically, would those decreases be measurable
considering the statistical variation in lung cancers and the number of
smokers from year-to-year? {I'd really like your opinion here.} How do you
measure the effectiveness of the radon mitigation program on lung cancer?
After all, once you identify a hazard and institute corrective actions, you
need to verify those corrective actions are effective. {I'd really like
your opinion here too.}

Tom


FIELDRW@aol.com wrote:
> 
> In a message dated 00-03-08 13:41:10 EST, Tom Mohaupt writes:
> 
> Radon is not constant. The concentration has been changing over the years due
> to buttoning up homes and filtering air. This point can easily applied to
> residential radon studies. It also applies to the Iowa study, which does not
> correct for  this effect. See "Residential Radon and Lung Cancer Case-Control
> Study" at  http://www.cheec.uiowa.edu/misc/radon.html. The paper describes
> the Iowa radon study methodology for those of you who are interested.
> 
> Tom,
> --------------------------------
> Tom,
> 
> Your statement above is incorrect.  The report you reference above is our
> findings for the "radon gas phase" of the Iowa Radon Lung Cancer Study.  The
> paper will be published in the May issue of the American Journal of
> Epidemiology.   There is less radon gas variation over time in the Iowa Study
> than other studies, because the study was limited to individuals who lived in
> the same house for 20 years (median 32 years).  One of the predominant
> factors that affect temporal variation of indoor radon concentrations is
> occupant behavior (opening windows, etc.) with new homeowners.  The occupant
> behavior in the Iowa study is less of a concern, since at least one of the
> occupants have stayed in the same home for extended periods (>20 years).
> 
> We have examined the temporal radon gas variation in a subset of the homes
> over time and can assess how that affects our risk estimates.  We also looked
> at factors that may have affected radon concentrations over time (furnace
> change, addition of window or central air, etc.).  We can stratify the homes
> that may have had changes to see if it affects the risk analyses.  This can
> not be done with an ecologic study.
> 
> Let me ask you this.  Do you think you would get a better idea of the radon
> concentration in your home from a measurement performed in your home five
> years ago or using the county average radon concentration (as used in
> ecologic studies)?
> 
> In addition, we performed retrospective determination of radon progeny
> exposure using glass-based detectors (I would be glad to provide references
> on these devices), which minimizes the problems associated with temporal
> radon variation.  if you would like to call me, I would be glad to discuss
> this is more detail.
> 
> How would you design a study to examine the risk posed by residential radon?
> 
> How would you test the validity of the LNT?
> 
> Bill Field
> College of Public Health
> University of Iowa
> bill-field@uiowa.edu
> 
> ************************************************************************
> The RADSAFE Frequently Asked Questions list, archives and subscription
> information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html

-- 
Thomas Mohaupt, M.S., CHP
University Radiation Safety Officer

104 Health Sciences Bldg
Wright State University
Dayton, Ohio 45435
tom.mohaupt@wright.edu
(937) 775-2169
(937) 775-3761 (fax)

"An investment in knowledge gains the best interest." Ben Franklin
************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html