Re: Healthy Worker Effect

["Robert M. Loesch" <loesch@spok.eh.doe.gov>]

At 09:38 AM 3/20/96 -0600, Al Tschaeche wrote:

>Why can't the observed data derived from epidemiological studies of persons
>exposed to ionizing radiation that demonstrate cancer incidence or cancer
>mortality lower than that for the controls be interpreted as a hormetic
>effect?  Why is it always interpreted as a "healthy worker effect?"  Any
>comments?  Al Tschaeche xat@inel.gov
>
>
Al,

Not being an epidemiologist, my guess is the following:

To be able to interpret a hormetic effect, wouldn't the cohort being studied
have to be identical in composition to the cohort used in determining the base
cancer incidence or mortality rate?  In other words, the stimulus being studied
(radiation) would have to be statistically proven to be the cause of the effect
(lower cancer rates) and not somehow related to the cohort composition.

If the cohort being studied is biased towards a healther population or age 
distribution, then the "healthy worker effect" might be an attempt to explain
the results.  However, wouldn't the epidemiologists also have the burden of
proof
that this is indeed the case?

Bob Loesch

********************************************************
Robert M. Loesch
DOELAP Administrator
U.S. Department of Energy
Germantown, MD 20874
(301) 903-4443
loesch@spok.eh.doe.gov
********************************************************
Random number generation is too important a task
            to  be left to chance!