Re: Hormesis and Chernobyl

[mgoldman@ucdavis.edu (Dr. Marvin Goldman)]

>I subscribe to a periodical called "Access to Energy" (available from Access
>to Energy, Box 1250, Cave Junction OR 97523 for $35 for 12 monthly issues
>[Canada $37, overseas, by air mail only, $40]).  The May, 1996 issue
>contained the following statements: "When this information (Marv Goldman's
>article "Cancer Risk of Low-Level Exposure" in Science 271, pp1821-1822,
>March 1996) spreads to the public media, it is likely to stimulate a
>technological revolution.  In the best interests of not getting too far
>ahead of this and becoming a negative influence, we refrain for now from
>answering an obvious question - will the lives saved by radiation hormesis
>from Chernobyl exceed the lives lost from the initial accident?  Hint: The
>likely answer is definitely not politically correct."  Has anyone out there
>done such a calculation?  Why should such calculations not be required in
>ALARA evaluations?  In Environmental Impact Statements? etc?  I think I know
>the answer, but tell me anyway.
>
>By the way, Access to Energy should be required reading for all RadSafers
>(OPINION).  The ideas therein are stimulating and useful in our battle
>against psudoscience and the anti-everythingers.  Try it for a year and see
>if you do not agree.   Al Tschaeche. xat@inel.gov

Reply:

Whatever "Access to Energy" read into my Science article, it was NOT an
endorsement of hormesis for stochastic health effects (cancer) risk
estimation.  I mentioned hormesis as one of the options, not a proven fact
in my book.  I'm pleased that I helped focus interest on this issue and
began and ended my article with a plea for less zealotry and better SCIENCE
in an objective forum, on a level playing field.  Despite all the wailing,
this has yet to happen! It's my impression that the extremists on each side
of this many sided issue, by their very selection of the data to review,
all fail to objectively address the data which does not support their pet
approach, including NRPB, ICRP, NCRP, the "hormesis club,"and even
sometimes UNSCEAR.

Every study likely has some inherent flaws, weaknesses and uncertainties.
What upsets me is that a nice theory with some good data support is too
often extended beyond reality as a universal fix for the broader problem of
cancer risk.  Correlation doesn't equal causality.  I want to see each of
the key studies reviewed in terms of strengths and weaknesses and an
assessment of what is needed to reduce the uncertainties for the specific
question of cancer risk assessment or mechanism of effect.  Until molecular
biology feeds us the answers to all the many steps in the cancer causation
scheme, the debate will continue, including the major and very significant
uncertainty about the low dose accuracy for the Hiroshima data base.
Anyone who has taken the first course in epidemiology must know that there
are practical limits to its quantitative resolution, however heroic the
effort; alone, it will not get you the exact answer to miniscule risks.

This electronic stage is not the forum to get it resolved, but I hope this
forum will raise your interest and somehow  assist in putting the needed
resources into helping resolve it.

Stay tuned!

Marvin Goldman
(mgoldman@ucdavis.edu)