Re: Healthy Worker Effect and Nuclear S

[JMUCKERHEIDE@delphi.com]

Will,

Yes, I agree. I'm sorry for not being more precise. If the raw population data 
is used, then there is a difference. That compares "workers" to a group that
includes people who are already dead!? :-)  I note below that this is in
comparing workers to a "comparable age-adjusted population", but I should have 
been more precise in stating/recognizing the obverse difference of that from
data when making comparisons to raw population data. But maybe this should be
the "living population effect" since it is not exclusive to "workers", though
maybe it doesn't apply to politicians and "voters" (at least in Chicago)  :-)  
 

The same effect would exist with a followup of medical patients if compared to 
the general population. That's OK as long as the basis for comparison is
internal, like in the study you report, the important comparison is between
the 2 worker cohorts, not with the general population. 

When a comparison to the general population is made, then it should be to a
"comparable" group if the difference is to mean anything. The problem comes
from referring to studies that ARE compared to adjusted populations and still
chalking differences up to "healthy worker effect". (And I think people should 
be careful to note that nuclear facility workers need to be distinguished
between those who worked in "process chemical industrial facilities," who have 
a lot more chemical and non-chemical carcinogen confounding exposures,
including internal rad contamination, and those who have less such exposure
and more distinct external gamma exposure (like the shipyard workers, with
welding and asbestos, but not like a process chemical worker). 

Would you please provide the all-cancer and leukemia (other radiogenic
cancers?) SMRs in this study? 

Thanks for your note.

> Jim Muckerheide wrote:
> 
> >As Jerry expects, in the NSWS, the 33,352 non-nuclear workers (NNW) had poorer 
> >health compared to the 39,004 NW. However, the NNW also had an all-cause SMR
> >of 1.00 (+- 0.03 at 95% C.I.) compared to the general population. THERE IS NO
> >HEALTHY WORKER EFFECT (which of course there can NOT be if you think about
> >comparing to *age-adjusted comparable populations*). 
> >
> >The leukemia SMR to the population was 0.97 (+-0.42), lymphatic and
> >hematopoietic cancers 1.10 (+-0.27), lung cancer 1.15 (+-0.13) [significant],
> >and mesothelioma 2.54 (1.16, 4.43) [significant] from asbestos. So much for
> >"healthy worker effect", which is a fallacious "mantra" for the troops and the 
> >gullible unscientific public to ignore data. 
>
> There are other studies Jim.  For example, Carpenter et al 1994,
> Radiation Research 138, 224-238. 'Combined analysis of mortality in
> three UK nuclear industry workforces, 1946-1988'.
> 
> In a study of 40,771 radiation workers and 34,441 non-radiation workers
> from the same establishments they obtained an all-cause SMR of 82
> (p<0.001) in non-rad workers and 81 (p<0.001) in rad workers.
> 
> Looks like a Healthy Worker Effect to me!
> 
> Will
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
> Will Atkinson                                Internet: will.atkinson@aeat.co.uk
> Health Effects                               
> AEA Technology, 364 Harwell, Didcot          Phone:    +44 1235 434370
> Oxfordshire, OX11 0RA, U.K.                  FAX:      +44 1235 432134
> - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Thanks.

Regards, Jim Muckerheide
jmuckerheide@delphi.com
Radiation, Science, and Health