Re: Epidemiology vs. Radiation Biology

[John Goldsmith <gjohn@bgumail.bgu.ac.il>]

(Opinion)  Speaking as an environmental epidemiologist, there seem to be 
two gaps in the discussion of epidemiology vs radiation biology.  The 
first is that there are many other agents to which human popualtions are 
exposed, both occupationally and in the community, and for each of them 
there are policy aspects similar to those for radiation, dealing with 
dose-response data, risk acceptance, risk recognition, integration of 
toxicological and epidemiolgical information, and use of scientific 
information as a basis for regulation.  For example carbon monoxide as a
vehichular pollutant is regulated on the basis of clear physiological 
risks and  a small number of epidemiological studies.  No one seems 
interested in thresholds for CO, because increasing uptake is known to
specifically impair oxygen transport function.  The second gap is in
focussing on cancer, rather than on associated conditions, like hematological
changes, mutations, respiratory or hepatic function, evidence of oxygen 
radical toxicity, or unfavorabloe reproductive outcome.  Cancer takes 
decades to manifest itself, may affect only a small portion of the exposed 
populations, and has many causal factors.  In short, broader experience 
in environmental epidemiology and environmental biologiy seems needed to
more clearly define approaches to the use of valid information for 
management of radiation risks.  The next meeting of the International 
Society for Environmental Epidemiology will be in Edmonton, August 17-20.
Contact Colin Soskolne, 13-103 Clinical Sciences Bldg, U. Alberta, 
Edmonton T6G 2G3 Lberta CANADA; FAX 403 492-0364.
gjohn@bgumail.bgu.ac.il   John Goldsmith, M .D., M.P.H. Professor
                          of Epidemiology, Ben Gurion University 

On Tue, 26 Mar 1996, Scott O. Schwahn wrote:

> At 05:21 PM 3/25/96 -0600, Alex Zapantis wrote:
> ..
> >It seems that we will probably never be able to determine with any measure
> >of certainty the effects of low dose/dose rate radiation using
> >epidemiological studies. Perhaps researchers should move away from
> >epidemiology and place more emphasis on radiation biology in an attempt to
> >resolve this problem. Do we really have a clear understanding of the
> >processes involved in radiocarcinogenesis? Do our current models take into
> >account environmental factors and exposures to other agents which may have a
> >synergistic or antagonistic effect. Do we know why there is a latency period
> >between exposure and the onset of cancer? I think not (OPINION).
> 
> I agree (OPINION).  One of the things that statistics shows us is that there
> is no good way to determine a parameter (threshold in this case) in a regime
> in which there is no data (or the data is immeasurable).  In order to do
> that, we have to make assumptions, which is what we try to avoid.  It is
> fairly easy to determine a linear (or otherwise), least-squares fit to data
> where there are measureable effects, but one cannot assume that the linear
> (or other) function continues in the same manner below this level.  So, in
> order to figure this all out, we need to be able to measure smaller
> effects... and how they relate to the "big picture".
> 
> Scott O. Schwahn, CHP
> Operational Health Physicist
> Continuous Electron Beam Accelerator Facility
> (804)249-7551 (w)
> (804)249-7363 (fax)
> schwahn@cebaf.gov
>